, 'V'V'' The Journal of ARACHNOLOGY OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY Proceedings of the XIV International Congress of Arachnology and a Symposium on Spiders in Agroecosystems VOLUME 27 1999 NUMBER 1 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF: James W. Berry, Butler University MANAGING EDITOR: Petra Sierwald, Field Museum ASSOCIATE EDITORS: Matthew Greenstone, USDA; Robert Suter, Vassar College EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E. Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C. Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ.; C. D. Dondale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galia- no, Mus. Argentino de Ciencias Naturales; M. H. Greenstone, USDA, Stillwater, Oklahoma; C. Griswold, Calif. Acad. Sci.; N. V. Horner, Midwestern State Univ.; D. T. Jennings, Garland, Maine; V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ.; E. A. Maury, Mus. Argentino de Ciencias Naturales; N. I. Platnick, American Mus. Natural Hist.; G. A. Polis, Vanderbilt Univ.; S. E. Riechert, Univ. Tennessee; A. L. Rypstra, Miami Univ., Ohio; M. H. Robinson, U.S. National Zool. Park; W. A. Shear, Hampden-Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa Rica. The Journal of Arachno logy (ISSN 0160-8202), a publication devoted to the study of Arachnida, is published three times each year by The American Arach- no logical 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, $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: Ann L. Rypstra (1997-1999), Dept, of Zoology, Miami Univer- sity, Hamilton, Ohio 45011 USA. PRESIDENT-ELECT: Frederick A. Coyle (1997-1999), Department of Biology, Western Carolina University, Cullowhee, North Carolina 28723 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. BUSINESS MANAGER: Robert Suter, Dept, of Biology, Vassar College, Pough- keepsie, New York 12601 USA. SECRETARY: Alan Cady, Dept, of Zoology, Miami Univ., Middletown, Ohio 45042 USA. ARCHIVIST: Lenny Vincent, Fullerton College, Fullerton, California 92634. DIRECTORS: H. Don Cameron (1997-1999), Matthew Greenstone (1997- 1999), David Wise (1998-2000). HONORARY MEMBERS: C. D. Dondale, H. W. Levi, A. F. Millidge, W. Whit- comb. Publication date: 17 August 1999 @ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Proceedings of the XIV International Congress of Arachnology and a Symposium on Spiders in Agroecosystems held at the Field Museum of Natural History in Chicago, Illinois June 27 - July 3, 1998 at the 22nd Annual Meeting of the American Arachnological Society Congress organizer: Petra Sierwald (Field Museum, USA) Proceedings editor: James Berry (Butler University, USA) Proceedings review editor: Brent Opell (Virginia Polytechnic Institute and State University, USA) Symposium review editors: Matthew Greenstone (US Dept, of Agriculture, USA) Keith Sunderland (Horticulture Research International, UK) Sponsors: American Arachnological Society Centre International de Documentation Arachnologique Field Museum of Natural History ACKNOWLEDGMENTS An historic event transpired between the 27th of June and the 3rd of July in 1998. I am indebted to more than three hundred arachnologists who gathered at the Field Museum of Natural History in Chicago, Illinois, USA to make the combined meeting of the International Congress of Arachnology and the American Arachnological Society a giant success. I think I speak for the entire American Arachnological Society when I say how delighted we were to be able to combine our 22nd meeting with the XIV International Congress and host the event in North America. This volume, in one sense the culmination of the meetings, stands as lasting documentation of the vast array of research publications that formed the core of these meetings. The success of these meetings hinged on the hard work and dedication of a number of individuals who contributed in both large and small ways. Special thanks go to Petra Sierwald whose hard work and attention to detail ensured that we all had a good experience in Chicago. We also appreciate the cheerful support of the students and staff at the Field Museum through- out our visit. Thanks also go to Richard Bradley, Alan Cady, Karen Cangliosi, Jonathan Cod- dington, Fred Coyle, Matthew Greenstone, Gustavo Hormiga, Otto Kraus, Patricia Miller, Nor- man Platnick, Robert Raven, Gail Stratton, Robert Suter, Keith Sunderland, George Uetz, Betsy Berry, Rudiger Bieler, and Marius van der Merwe making significant contributions to the event. My apologies to those whom I have forgotten or whose contributions I failed to notice. Please know you have my gratitude as well. The quality of the meetings were greatly enhanced by the two symposia. My sincere appre- ciation goes to Matthew Greenstone and Keith Sunderland for organizing a symposium on “Spiders in Agroecosystems” and to Gustavo Hormiga and Jonathan Coddington for organiz- ing a symposium entitled “Higher Level Phylogenetics of Spiders.” Thanks also to all the symposium participants. It was fortunate that the American Arachnological Society was able to subsidize meeting costs, defray some of the travel expenses for participants, and support publication of this volume. The Centre International de Documentation Arachnologique also provided some sup- port for this publication. The diligent efforts of Matthew Greenstone and Keith Sunderland secured extra funding to subsidize travel for the participants in their symposium and contribute to the publication of the proceedings. The important and demanding task of assembling this volume cannot be overlooked. My heartfelt appreciation goes to James Berry, Brent Opell, and Matthew Greenstone who worked hard to bring us this high quality publication. Thanks also to all the reviewers who read and commented on all these manuscripts. Finally, I am gratified by the high level and excellent quality of the student participation at these meetings. The excitement and commitment I saw in students from all over the world at these meetings reassured me that the future of arachnology is safe. Ann Rypstra, President American Arachnological Society 2 1999. The Journal of Arachnology 27:3-6 HISTORIC OVERVIEW OF PAST CONGRESSES OF ARACHNOLOGY AND OF THE CENTRE INTERNATIONAL DE DOCUMENTATION ARACHNOLOGIQUE (C.LD.A.) Otto Kraus: Zoologisches Institut und Museum, Universitat Hamburg, Martin-Luther“King“Platz 3, D=20146 Hamburg, Germany ABSTRACT. In 1959, Hermann Wiehle encouraged junior colleagues to arrange a meeting for arach- nologists. This was held in 1960 at the University of Bonn, Germany, and is counted as the first in the series of international congresses. After a second conference, held just one year later in Saarbriicken, Germany, a third truly international congress followed in 1965 in Frankfurt am Main. On this occasion. Max Vachon presented his idea to form what was later called the C.LD.A. This institution was formally established at the occasion of the 4*^ congress, held in Paris in 1968. Detailed information on the origins, the series of congresses, and the C.LD.A. up to 1968 is presented. The state of arachnology during the first half of this century, and its condition follow- ing developments after World War II, differed strikingly. This is illustrated by two figures. Roewer’s original catalog, and later supple- mentary entries, lists a total of 71 papers on spider systematics published in 1939. The equivalent figure for 1995 (according to the entries in Platnick’s catalogue) is 268 — almost four times more. This difference can only par- tially be explained by a world-wide increase in the number of universities and other re- search institutions, combined with an increas- ing number of researchers. Two other factors seem to be much more important: Air travel to foreign countries and world-wide personal con- tacts between researchers had become relative- ly easy. At the same time, primarily European conferences on arachnology quickly evolved into international congresses. The foundation of the “Centre International de Documentation Arachnologique” — ^the C.LD.A. — in Paris, formed an integrative part of this develop- ment. Traveling. — Up to the 1950s, most arach- nologists did not know each other personally. They exchanged reprints, wrote kind (and oth- er) letters; and they mainly traveled by train. Collecting and research work, especially in tropical countries, necessitated the organiza- tion of expeditions. Due to increasing inter- national air travel facilities, this situation changed quite rapidly. American colleagues crossed the Atlantic, worked in European col- lections and studied type materials; and per- sonal contacts frequently resulted in firm friendships. As the western part of Europe re- covered from the war, more and more Euro- pean arachnologists were also able to go to other countries. The aircraft most frequently used in those early days was the “Lockheed Super Constel- lation” with a cruising speed of 255 mph. Some people called it the world’s safest three- propeller airplane, as there was a rumor that one of its four engines could occasionally be out of order. In these early days, Ernest Browning of the Department of Arachnology at the British Museum still prepared his end- less lists for the Zoological Record by hand. International organization and coopera- tion.— Most of those present in Chicago for the XIV International Arachnological Con- gress may not know how the flourishing series of International Arachnological Congresses originated, including the origins of C.LD.A. So, I was asked, how it is that the present meeting is the 14^^ conference in the series. For these reasons, I will concentrate on the early history. Late in 1959, the German arachnologist Hermann Wiehle (1884-1966, Fig. 1) con- ceived the idea that arachnologists should come together. Not only would this allow them to discuss scientific problems, but it would foster personal contacts and coopera- tion among arachnologists. Wiehle may be re- garded as the initiator of our congresses. As 3 4 THE JOURNAL OF ARACHNOLOGY Figure 1. — Hermann Wiehle, the initiator of the International Congresses of Arachnology (photo taken on 23 April 1965 at the 3'^'^ congress). he lived in the eastern part of Germany (which was dominated by the Soviet Union at that time) and did not hold any office, he encour- aged two junior arachnologists to arrange a meeting: Wolfgang Crome, curator at the Ber- lin Museum, and Ernst Kullmann, at that time scientific assistant at the University of Bonn. As the next annual meeting of the German Zoological Society was scheduled to be held in Bonn, Kullmann used this meeting, in 1960, as a platform for hosting the arachnol- ogy conference. Approximately 20 German arachnologists attended (Fig. 2). The presence of Father Chrysanthus from the Netherlands, who worked on spiders from New Guinea, provided a bare minimum of international par- ticipation. Wiehle himself contributed a paper on “Der Embolus des mannlichen Spinnen- tasters.” Despite the fact that this study, based on functional morphology including aspects of co-adaptation between male and female genital structures, was of considerable impor- tance and far ahead of its time, the topic re- mained almost completely neglected — even until now. Another contributor was Heinrich Homann who reported on retinal structures of spider eyes. Homann ’s landmark study con- tributed greatly to our present understanding of certain aspects of aranaeomorph phylogeny. Carl Friedrich Roewer, at this time nearly 80 years old, did not attend. Volume 2 of his “Katalog der Araneae,” comprising 1751 pages, had been published in two voluminous parts five years previously. The availability of Roewer’s completed catalog was a tremen- dous stimulus to araneologists. In principle, the information was arranged so perfectly that Paolo Brignoli, as well as Norman Platnick, had no reason to make major changes when they published four supplementary volumes in 1983, 1989, 1993, and 1997. Everybody felt that meetings of arachnol- ogists should be continued. Those present in Bonn resolved with enthusiasm that the next (second) conference should take place one year later (1961) in Saarbriicken. This was rel- atively close to the German/French border. Again, the annual meeting of the German Zoological Society was used as a vehicle. Kullmann had just accepted a position in Af- ghanistan for a couple of years. So I was asked to organize the meeting. The intention was to broaden the scientific basis and further to encourage attendance by colleagues from other European countries. This approach was successful: a large group of French arachnol- ogists attended, including Pierre Bonnet and Max Vachon; and Peter van Helsdingen from the Netherlands also joined the meeting. A gap of four years followed; but this was not at all a period of inactivity. In April 1964, Max Vachon came to Frankfurt for a couple of days and visited me, then curator at the Senckenberg Museum (Fig. 3). We discussed details of the forthcoming meeting of arach- nologists to be held one year later, in 1965, in Frankfurt and, more importantly, how inter- national cooperation in the field of arachnol- ogy could be improved. This was the concep- tion, but not yet the birth, of C.I.D.A. Vachon, who was the last representative of the great French tradition, proposed the formation of an international organization which later became known as the ‘'Centre International de Doc- umentation Arachnologique.” The Frankfurt congress was held from 21-25 April 1965. Originally, it was called ‘TIL Treffen euro- paischer Arachnologen” , but the designation “III. KongreB europaischer Arachnologen” was used in the printed congress report (Senckenbergiana Biol., 47(1): 1966). Fifty- KRAUS— HISTORIC OVERVIEW OF PAST CONGRESSES 5 Figure 2. — Group of arachnologists at the first meeting in 1960 in Bonn. 1, G. Schmidt; 2, H. Casemir; 3, H. Homann; 4, B. von Broen; 5, H. Nemens; 6, R. Braun; 7, Father Chrysanthus; 8, H. Wiehle; 9, E. Kullmann; 10, Mrs. Kullmann; 11, H. Hiebsch; 12, R. Lehmensick; 13, M. Grasshoff; 14, W. Engelhard!; 15, Mrs. Crome; 16, W. Crome; 17, O. Kraus; 18, B. Heydemann. three arachnologists were present. They came from Austria, Belgium, Czechoslovakia, Fin- land, France, Germany, Great Britain, Italy, Yugoslavia, The Netherlands, Poland, Ruma- Figure 3. — Max Vachon vigorously discussing the foundation of the C.I.D.A. (conference held on 17 April 1964 at the Senckenberg-Museum, Frank- furt am Main). nia and from Switzerland (Fig. 4). A business meeting was held, in which Vachon explained our joint ideas to establish the C.I.D.A. at the Museum National d’Histoire Naturelle. He further proposed to meet three years later, in 1968, in Paris. These plans were unanimously accepted. Vachon started work as the “Secre- taire general permanent du C.I.D.A.,” and I was elected as the first president of the orga- nization just created. The C.I.D.A. was bom! There are only a very few survivors who can still remember Wiehle’s unforgettable “Rede iiber die Freundschaft” [lecture on friend- ship], held at the occasion of the farewell din- ner. This perfectly reflected the general enthu- siastic atmosphere. Between 1965 and 1968, intensive contacts between Frankfurt and Paris followed. The world-wide system of C.I.D.A. correspon- dents was established, and details of the forth- coming meeting were carefully arranged. The Paris conference was officially called “I Verne Congres International d’ Arachnologie.” It took place from 8-13 April 1968, at the Mu- 6 THE JOURNAL OF ARACHNOLOGY Figure 4. — Pierre Bonnet (left) and G.H. Locket (right) in conversation (photo taken on 23 April 1965 at the 3"^^ congress). seum National d’Histoire Naturelle,” a sacred place in arachnology. A total of 109 arach- nologists from 23 nations attended, including participants from Algeria, Argentina, Canada, USA, India, Israel, Madagascar, Turkey, and Uruguay. The international scope was estab- lished step by step, but the fourth congress in Paris was the first which really had a world- wide scope. The program included no less than 60 contributions and the presentation of 1 1 scientific films (for details see the proceed- ings: Bull. Mus. Natn. Hist. Nat, (2)41, Suppl. 1; 1969). Roger Legendre was elected as president for the forthcoming period. The congress participants accepted my proposal that the next meeting should be held in Bmo, Czechoslovakia. The idea behind this was that more colleagues from eastern countries would obtain permission to attend. But this proved to be an error. The fifth “Brno Congress” was held from 30 August-4 September 1971. Ninety-seven participants from 23 countries attended. Com- pared to the preceeding congress in Paris, there was only a slight decline in the number of arachnologists present, but only 6 col- leagues came from countries outside Europe (Paris: 13). Officially, the congress was or- ganized by the “Institute of Vertebrate Zool- ogy,” Director J. Kratochvfl. However, it was Dr. Vladimir Silhavy, a physician and quali- fied amateur arachnologist specialized in the Phalangida who did all the work. He also pre- pared the congress proceedings. These were technically published in “1972” by the Insti- tute already mentioned, but the publication was not issued until May 1973 as a separate volume. From this point forward, a self-perpetuating cycle of international conferences was clearly established. Information on the succeeding congresses, numbered 6-13, is readily obtain- able as their proceedings were published reg- ularly and are accessible. So I simply mention places, countries and organizers: VI; Amster- dam (Netherlands), P. van Helsdingen. — -VII: Exeter (UK), A.F. Millidge; VIII: Vienna (Austria), H. Nemenz and J. Gruber; IX: Pan- ama City (Panama), M. Robinson; X: Jaca (Spain), M. Rambla; XI: Turku (Finland), P. Lehtinen; XII: Brisbane (Australia), R. Raven; XIII: Geneva (Switzerland), V. Mahnert. — It is noteworthy that only two out of these 13 congresses have been held outside Europe. This seems to be correlated with the geo- graphical distribution of those arachnologists with access to substantial institutional assis- tance, financial resources and appropriate pub- lishing facilities. Concluding remarks. — Against this back- ground, the present invitation by the American Arachnological Society to have this meeting, for the first time in the United States, is ap- preciated. We are extremely grateful to Petra Sierwald and her collaborators for undertak- ing the laborious task. But we should also ex- press our sincere thanks to the people of the C.I.D.A. Secretariat in Paris, to Mme. Jacque- line Heurtault and her co-workers. This insti- tution, established in 1964, has functioned perfectly since 1968. The annual information service provided by the C.I.D.A. was extraor- dinarily useful. It was this institution that helped safeguard the series of our internation- al congresses, which has not been interrupted since 1965. Now, the celebration of C.D.I.A’s 30^^ birthday coincides with the end of the original Paris secretariat — due to complica- tions caused by organizational and personal problems. But fortunately, the headquarters will move from the Paris museum to the Na- tional Museum of Natural History in Wash- ington, D.C., and the work will be continued by J. Coddington as Secretary General (Wash- ington), Robert Raven as current president (Brisbane, Australia) and N. Platnick as mem- bership secretary (New York). 1999. The Journal of Arachnology 27:7-15 THE GENUS ATTIDOPS (ARANEAE, SALTICIDAE) G.B. Edwards: Curator, Arachnida & Myriapoda, Florida State Collection of Arthropods, Division of Plant Industry, RO. Box 147100, Gainesville, Florida 32614- 7100 USAi ABSTRACT. The genus Attidops is resurrected from Ballus based on its strongly excavate cymbial tip, transverse embolar groove, flatter carapace with extended postocular area, and single retromarginal che- liceral tooth, which indicate a closer relationship to the genus Admestina. The type species, Ballus youngii Peckham & Peckham 1888, again transferred, becomes Attidops youngi (Peckham & Peckham). Two new species, Attidops nickersoni (sister species to A. youngi) and Attidops cutleri, are described. Icius cinctipes Banks 1900, previously transferred to Ballus, becomes a new combination, Attidops cinctipes (Banks). Lectotypes and paralectotypes are designated for Ballus youngii and Icius cinctipes. The genus is recorded from south-central Canada, eastern U.S. and eastern Mexico. Banks (1905) created the genus Attidops in a footnote, stating simply “Attidops, a new genus for Ballus youngi Peck.” The Peckhams (1909) refuted this genus, asserting that, . Youngii is so close to the type of Ballus (depressus) that Mr. Emerton, in a letter to us, has suggested that it may be identical. We think that it \youngr\ differs enough to be a good species, but it clearly belongs to the ge- nus Ballus, and hence we treat Attidops as a synonym.” Unfortunately, none of these au- thors offered any morphological evidence sup- porting or refuting the placement of B. youngi in Ballus. Kaston (1948) later addressed this issue: “although Ballus is considered a plur- identate genus, all the specimens of youngii seen by me have only a single retromarginal tooth. There may be justification for removing youngii to the genus Attidops set up for it by Banks (1905).” Quotes are as originally pub- lished, except name in brackets is my inser- tion. European species of Ballus C.L. Koch 1851 have been revised by Alicata & Cantarella (1987) with good illustrations of the genitalia, and Roberts (1985b: plate 59) illustrated in color the female dorsum of Ballus chalybeius (Walckenaer 1802) (as Ballus depressus). Ar~ anea depressa Walckenaer 1 802 (the type spe- cies of Ballus), which has page priority over Aranea chalybeia Walckenaer 1802, is pre- 'Contribution No. 866, Bureau of Entomology, Nematology and Plant Pathology-Entomology Sec- tion. occupied by Aranea depressa Razoumowsky 1789. Admestina Peckham & Peckham 1888, a genus consisting of three small, bark-dwell- ing species known from eastern North Amer- ica, has also been revised (Piel 1991). In order to ascertain relationships, I compared speci- mens of the type species of all three putative genera: Attidops youngi, Ballus chalybeius, and Admestina tibialis Peckham & Peckham 1888. I concluded that Attidops is more close- ly related to Admestina than it is to Ballus, therefore I resurrect Attidops. Icius cinctipes Banks 1900, which I previously transferred to Ballus (Edwards 1980), also belongs in Atti- dops. In addition, two new species belonging to Attidops are described below. METHODS Measurements (in mm) of carapace length (CL), carapace width (CW), body length (BL), and legs were made with a calibrated ocular reticule in a Leica MS5 binocular microscope. If available, five specimens of each sex were measured; primary type measurements are given in parentheses within the range of var- iation of each species. A camera attachment was used to photograph dorsal views of spec- imens. Male and female genitalia were dis- sected, mounted in depression slides, and pho- tographed through a Nikon Labophot compound microscope. All photographs were digitized, computer enhanced, printed, and re- touched with black and white inks to produce most of the figures. Other abbreviations used are: anterior median eyes (AME), anterior lat- 7 8 THE JOURNAL OF ARACHNOLOGY Map 1. — Distribution of Attidops species; • = A. youngi (divided circles = literature records), ★ = A. nickersoni new species; ■ = A. cutleri new spe- cies, o? = A. cinctipes. Quebec literature record of A. youngi not shown. eral eyes (ALE), posterior median eyes (PME), posterior lateral eyes (PLE), anterior eye row (AER), posterior eye row (PER), and ocular quadrangle (OQ; the area bounded by the ALE and PLE). Eyes are included in the measurements of eye row width and OQ length. All illustrations of male palpi are of the left palpus. I follow Platnick (1989) in considering Bonnet’s corrections of patronyms ending in -ii as valid emendations of incorrect original spellings. Possible sites of obscure localites were found with http://mapping.usgs.gov/ w w w/gni s/gni sform . html . TAXONOMY Attidops Banks 1905 Map 1 Attidops Banks 1905: 321 in footnote Attidops: Peckham & Peckham 1909: 586 (= Ballus) Type species. — Ballus youngii Peckham & Peckham by monotypy. Figures 1-5. — Prosomal views. 1-3, lateral view of female carapace. 1. Admestina tibialis; 2. Atti- dops youngi; 3. Ballus chalybeius; 4-5, Male Atti- dops youngi. 4. Posterolaterodorsal view of cara- pace showing surface reticulations; 5. Venter of prosoma minus legs, free segments of palpi, and one chelicera. Figures 1-4 to same scale. Diagnosis. — Carapace flat as in Admestina (Fig. 1), but postocular dorsum (from PLE to top of thoracic slope) extended about equal to length of OQ (Fig. 2). The cymbium is strong- ly excavate distally on the retrolateral side, and the embolar groove is transverse. The em- bolus is situated distally and makes 1.5-3 spi- rals. The lateral portion of the tegular duct has a well-developed median bend. The gono- pores of the epigynum are medial in trans- verse slits or a transverse depression. The epi- gynal plate is triangular, as in Admestina, although in Attidops the anterior epigynal edge is not well-defined. Tibia I not enlarged nor abdomen elongated as in Admestina. In Admestina, the postocular dorsum extends at least 1.25X length of OQ, the cymbium is only slightly excavate, the embolus is canted to the retrolateral side and is a single spiral, and the gonopores are anterior. In Ballus, the carapace is not as flat, the postocular dorsum extends only about 0.5 X length of OQ (Fig. 3), the cymbium is not excavate, the embolar groove is longitudinal, the embolus has more than 3 spirals mostly fused together, and the gonopores are in submedian longitudinal slits. Description. — Small spiders, 2. 1-3.1 in length, carapace length 1.0-1. 3, width 0.8-1. 1 EDWARDS— SALTICID GENUS ATTIDOPS 9 at widest point behind PLE and half as high as wide in middle, length of OQ plus AME almost half length of carapace; PME tiny, on line between dorsal edges of ALE and PLE, slightly closer to ALE (0.46 distance from ALE to PLE). Prosoma: carapace dark red- dish-brown, darker around eyes; vertical lat- erally, flat dorsally with slight slant downward from PLE forward; postocular dorsum extends about the length of the OQ behind the PLE; thoracic slope is abruptly steep (but not ver- tical) and slightly concave; carapace integu- ment (Fig. 4) entirely reticulate (reticulations polygonal on OQ, granular tending toward anastomosing striations laterally and posteri- orly), polygonal reticulations extending pos- terior to PER medially to top of thoracic slope; chelicerae small, two adjacent promar- ginal teeth, one opposing retromarginal tooth; endites semitruncate distally with corners rounded, with a slight flange or incipient cusp in males on outer distal comers; labium slight- ly wider than anterior margin of sternum, with minute, transverse, sclerotized ridge each side posterolaterally (Fig. 5); venter and legs red- dish- to yellowish-brown (pale yellow in A. cinctipes); legs IV, I, II, III in order of length, both sexes similar; legs I only slightly broader than legs II-IV, legs with brown maculations. All tarsi and metatarsi pale yellow with prox- imal brown annulae. Leg measurements given only for A. youngi, as all four species are sim- ilar, and legs are proportional to body size. The only macrosetae present (all ventral) are: leg I: metatarsus 2-2 (distal and median), tibia 2-1 or 1-0 (median and proximal); leg II: metatarsus 1 (median), tibia 1 (median). Male palp with femur concave ventrally; sperm duct, upon traversing distal haematodocha, en- ters a membranous to sclerotized apparent em- bolus base which forms a core which is sur- rounded by a spiralled, flat, mostly membranous to mostly sclerotized embolus. Opisthosoma (abdomen): 0. 8-1.4 wide; dor- sum reddish brown with an anterior pair of submedian pale integumental spots, followed by two or three complete, narrow, pale trans- verse bands, followed by a few partial median transverse bands (all partially obscured in A. cinctipes)', in males, dorsum completely cov- ered with a translucent scutum (not obscuring pattern); venter pale to reddish-brown, with- out maculations (except A. cinctipes). Entire body with sparse covering of short white setae and translucent clear to white “scales” (flat- tened adpressed setae), especially laterally. In addition to the above mentioned differ- ences, Ballus has two or three teeth on a raised portion of the cheliceral retromargin (“serrat- ed platelet”; Alicata & Cantarella 1987), the labium lacks modifications, anteriorly the car- apace is inclined noticeably downward from the PME, the posterolateral areas of the car- apace are concave, the thoracic slope is grad- ually inclined, the lateral portion of the tegular duct has the median bend barely noticeable, and the epigynum is rectangular with the an- terior margin of the epigynum well-defined. In size, species of Ballus are mostly longer than (2. 5-5.0: Roberts 1985a; Alicata & Cantarella 1987) and have more mass than (since they are not as flat) species of Attidops. KEY TO THE SPECIES OF ATTIDOPS 1. Integument mostly pale (except carapace), with symmetrical dark maculations on legs and on ab- domen (both dorsum and venter) (Figs. 12-13); embolus heavily sclerotized, 1.5 spirals (Figs. 24- 25); gonopores in submedian transverse slits (Figs. 26-27); Gulf states of U.S. (and Mexico?) cinctipes (Banks) - Integument mostly dark reddish brown, with narrow pale transverse bands on abdominal dor- sum; abdominal venter immaculate; embolus with 2 or 3 spirals; gonopores in transverse de- pression 2 2. Embolus mostly sclerotized, 2 spirals (Figs. 22-23); gonopores in posterior depression (Figs. 29-30); abdominal transverse bands broken and/or fused (Figs. 10-11); Texas, Gulf states of Mexico cutleri Edwards - Embolus mostly membranous, 3 spirals (Figs. 14-15, 18-19); gonopores in lobed median de- pression (Figs. 16-17, 19-20); abdominal transverse bands distinct 3 3. Three complete transverse abdominal bands, two pair small white spots on OQ (Figs. 6-7); south-central Canada, eastern U.S. except lower Southeast ....... youngi (Peckham & Peckham) - Two complete narrow transverse abdominal bands; no OQ spots (Figs. 8-9); central Rorida . . . nickersoni Edwards 10 THE JOURNAL OF ARACHNOLOGY Figures 6-13. — Female dorsum, male dorsal ab- domen of species of Attidops. 6-7. A. youngv, 8-9. A. nickersoni new species; 10-11. A. cutleri new species; 12-13. A. cinctipes. Attidops youngi (Peckham & PeckJiam 1888) Figs. 2, 4-7, 14-17 Ballus youngii Peckham & Peckham 1888:87, Plate 1 Fig. 66 ( $ abdominal dorsum), Plate 6 Fig. 66, 66a ((? palp), 66b (epigynum); Cotypes (d palp; 3 9 , one missing epigynum) from USA: Pennsyl- vania: Allegheny County, under bark (usually hickory and sycamore), November (J.J. Young, MCZ), examined [data from description]; $ lec- totype, 3 paralectotypes designated B. youngi: Marx 1890:576 (-//); Banks 1895:92, 1899:190; Simon 1901:485; Peckham & Peck- ham 1909:586 (-zz), Plate 49: Figs. 9, 9a (c? palp), Plate 51: Figs. 13, 13a (9 abdomen, epigynum); Petrunkevitch 1911; Comstock 1913:671 (-zz); Crosby & Bishop 1928; Kaston 1938:195 (-zz), 1948:447 (-zz). Figs. 1621 (epigynum), 1622 (palp); Roewer 1954:973 (-zz); Bonnet 1955:848; Levi & Field 1954:462 (-zz); Dorris 1968:36 (- zz); Berry 1970:104; Richman & Cutler 1978:83; Stietenroth & Homer 1987:237; Belanger & Hutchinson 1992:71 (-zz) Attidops youngi: Banks 1905:321, 1910:74; Bar- rows 1918:315 Diagnosis. — Carapace with two pairs of white spots. Abdomen with three white, nar- row, transverse abdominal bands, of which the second band is noticeably bent forward in the middle. Embolus mostly membranous and starts ventrally, with three spirals; embolus base membranous. Epigynum with a lobed, central, shallow pit containing the gonopores, which open near the center on the lateral edg- es of the lobe; ducts convoluted and diverted laterally. Description. — Female CL 1.14 (1.15)- 1.30, CW 0.89 (0.90)-1.00, BL 2.40 (2.65)- 2.70; male CL 1.15-1.25, CW 0.87-0.98, BL 2.35-2.75. Embolus sclerotized on outer edge, sperm duct on inner edge. Carapace with two pair of small white spots (patches of scales), one pair about the middle of the OQ, the other pair posteromedial to the PLE; AER about 0.87 the width of the PER. Legs with femora, patellae, and tibiae brown laterally. Leg seg- ment lengths of typical female (in order legs I-IV): femora (0.39, 0.32, 0.31, 0.42), patellae (0.23, 0.20, 0.16, 0.21), tibiae (0.19, 0.17, 0.17, 0.27), metatarsi (0.15, 0.15, 0.15, 0.24), tarsi (0.16, 0.16, 0.16, 0.17). First transverse abdominal band often broken in middle; rarely a nearly complete fourth transverse band pre- sent posteriorly. Distribution. — South-central Canada and eastern United States from Connecticut to North Carolina, west to Wisconsin, and south to eastern Texas; under the bark of deciduous hardwoods (elm, shagbark hickory, sycamore) and hemlock. Literature records for which 1 EDWARDS— SALTICID GENUS ATTIDOPS 11 Figures 14-30. — Attidops genitalia. 14, 18, 22, 24. Palpi, ventral view; 15, 19, 23, 25. Palpi, retrolateral view; 28. Palpus, distal, slightly retrolateroventral view, cymbium removed; 16, 20, 26, 29. Epigyna, ventral view; 17, 21, 27, 30. Epigyna, dorsal view; 14-17. A. youngi; 18-21. A. nickersoni new species; 22-23, 29-30. A. cutleri new species; 24-28. A. cinctipes. 12 THE JOURNAL OF ARACHNOLOGY have not seen specimens are those of Banks (1899) from Louisiana, Barrows (1918) from Ohio (Rockbridge), Belanger & Hutchinson (1992) from Quebec, Berry (1970) from North Carolina, Dorris (1968) from Arkansas, and Levi & Field (1954) from Wisconsin. Material examined. — CANADA: Ontario: Hal- ton County, Burlington, Lamb’s Hollow, 1 August 1985, 1 $ (W. Maddison, MCZ); USA: Connecticut: Fairfield County, Shelton, 7 April 1935, 19 (B.J. Kaston, USNM); New Haven County, Mt. Carmel, 15 April 1935, 1(519 (B.J. Kaston, USNM); South Meriden, 31 March 1935, 25 (H.L. Johnson, USNM); Kansas: Jefferson County, Nelson Envi- ronmental Area, 24 January 1994, 1 juv (H. Guar- isco, HGC); Maryland: Montgomery County, Silver Spring, 28 September 1944, 19 (M.H. Muma, FSCA); Michigan: Livingston County, E.S. George Reserve, June-August 1951-1957, 135199 4 juv (all H.K. Wallace, FSCA); New York: Nassau Coun- ty, Sea Cliff, 1519 2 juv (2 vials) (N. Banks, MCZ); Orange County, Harriman, Bear Mountain State Park, 25 April 1964, 15 (J.& W. Ivie, AMNH); Westchester County, Yonkers, 14 January 1935, 19 6 juv (R. Woodbury, USNM); Ohio: Knox County, Brinkhaven, on sandstone cliffs, 15 September 1917, 453 9 (WM. Barrows, OSU); Pennsylvania: Bucks County, northeast of Jamison, Horseshoe Bend, January-June and October 1954- 1958, 635779 14 juv (all W Ivie, AMNH); Texas: Brazoria County, Otey, February 1971, 15 (K. Ste- phan, FSCA); Virginia: Arlington County, Cherry- dale, 20 April 1935, 1519; 30 April 1935, 153 9 (all R. Woodbury, USNM); York County, Site 5, 5 January 1984, 1 juv (C. Stietenroth, MSU); Wis- consin: [? Waukesha County, Pine Lake], 1 5 (Peck- ham, MCZ); [? Alabama: Lawrence County or Tex- as: Anderson County]: Green’s Bluff, phoebe’s nest, 5 November 1949, 152 9 (FSCA). Brackets indicate possible resolutions of missing data. Attidops nickersoni new species Figs. 8-9, 18-21 Types. — Holotype 5, alloparatype 9 to- gether in one vial; 1569 paratypes in second vial, from USA: Florida: Marion County, Ocala National Forest, 1.8 miles west of FR 579 on FR 595 [older maps indicate as FR 79 and FR 95, respectively], under bark of living and dead longleaf pines, 13 November 1975 (G.B. Edwards and J.C.E. Nickerson, FSCA); 15 paratype, same locality, 6 November 1998 (RE. Skelley, ESCA); 153 9 paratypes, 2 miles west of FR 579 on FR 595, 29 Septem- ber 1976 (G.B. Edwards, FSCA) [collected as penultimates, all matured October 1976]. Etymology. — Named after the late Dr. Ev- erett Nickerson, fellow student of Dr. Willard H. Whitcomb, co-collector of the type series. Diagnosis. — Proportionately narrower than A. youngi, and there are only two pale yellow, very narrow, transverse abdominal bands (the second sometimes has a slight median for- ward bend), followed (in preserved speci- mens) by a pale posterior triangular spot (which in females has 3-5 partial, median transverse bands). Embolus like A. youngi, but it starts retrolaterally. Epigynum like A. youn- gi, but proportionately smaller. Description. — Female CL 1.13-1.20, CW 0.81-0.86, BL 2.17-2.40; male CL 1.10 (1.10)-1.15, CW 0.80 (0.80)-0.82, BL 2.25 (2.25)-2.30. Males with prolaterodorsal yel- low stripe on all patellae and tibiae; otherwise legs marked as in A. youngi, but paler ven- trally. The Gainesville specimen differs by having the most anterior band broken in the middle, and by having a third transverse band, also broken in the middle. Distribution. — Central Florida, most re- cords from under the bark of longleaf pine. Material examined.— USA: Florida: Alachua County, Gainesville, under live oak bark, 30 No- vember 1975, 19 (D.B. Richman, FSCA); Pinellas County, Dunedin, 1927, 1 9 (WS. Blatchley, MCZ). Attidops cutleri new species Figs. 10-11, 22-23, 29-30 Types. — Holotype 5 from USA: Texas: Travis County, Austin, 18 October 1967 (D. Simon, FSCA); paratype 5 from Texas: Cald- well County, Lockhart State Park, W97.40: N29.50, 13 April 1963 (WJ. Gertsch & W Ivie, AMNH). Etymology. — Named for Dr. Bruce Cutler, who first identified the AMNH specimens to genus. Diagnosis. — Shorter than A. youngi, and the first transverse band broken into two spots (making two pair of spots anteriorly), the sec- ond transverse band (apparent first band) bent forward not only in the middle but once on each side as well, and all following bands bent forward in the middle. Embolus more sclero- tized than membranous with two complete spirals, embolus base sclerotized. Epigynum with gonopores located in anterior part of large posterior depression; ducts less complex than A. youngi. Description. — Eemale CL 1.0- 1.0, CW EDWARDS— SALTICID GENUS ATTIDOPS 0.80-0.85, BL 2.2~2.2; male CL (1.05)-1.05, CW (0.86)-0.92, BL (2.15)-2.15. Legs marked as in A. youngi. Dorsum of abdomen with scattered symmetrical lateral pale mark- ings; second and third transverse bands (usu- ally the only two complete bands present) may be fused together laterally and somewhat bro- ken medially in females. Overall color pattern gives the impression of being intermediate be- tween A. youngi and A. cinctipes. Males and females have not been collected together, but the size, shape and color pattern of the spec- imens leads me to match them. The Tamau- lipas female has underdeveloped insemination ducts and may be immature, therefore I have illustrated the Campeche female. Distribution.- — Texas south into Mexico in states bordering the Gulf of Mexico. Material examined. — MEXICOi Tamaulipas: Llera Mesa (near summit), W98.59:N23.23, 16 April 1963, 1?(?) 2 juv (WJ. Gertsch & W. Me, AMNH); Campeche: Chicanna ruins, ca. 8 km w Xpujil, ca. 89°3TW, 18°32'N, dead branch in short tropical rainforest, 12-14 July 1983, 19 (W. Mad- dison, MCZ). Attidops cinctipes (Banks 1900), NEW COMBINATION Figs. 12-13, 24-28 Scius cinctipes Banks 1900:101 {Scius a lapsus cal- ami for IciusY, Cotypes (29) from USA: Louisi- ana: Baton Rouge Parish, Baton Rouge, May (H. Soltaw, MCZ), examined, lectotype and paralec- totype designated (subadult specimen also pre- sent) Icius cinctipes: Banks 1910:71; Petrunkevitch 1911: 661; Bryant 1933:193, Figs. 42 (epigynum), 47(9 dorsum); Roewer 1954:1222; Bonnet 1957: 2279 Ballus cinctipes: Edwards 1980:12 (n. comb.); Plat- nick 1993:738 Diagnosis. — As large to slightly larger than A. youngi, and the abdomen, while retaining remnants of the typical Attidops pattern (e.g., evidence of anterior pale spots and posterior light transverse bands), is dominated by pale coloration with symmetrical median dark maculations. Embolus with one and a half spi- rals, heavily sclerotized; embolus base scler- otized and extended dorsally to create shield with curved rectangular projection which cra- dles part of distal embolus. Epigynum with gonopores in two submedial, transverse slits; ducts like A. cutleri but larger. Description. — Female CL 1.11 (1.18)- 13 1.30, CW 0.90 (0.95)-0.95, BL 2.45 (3.00)- 3.10; male CL 1.21-1.40, CW 0.94-1.06, BL 2.50-2.90. Carapace brown, but lighter red- dish-brown dorsally behind eyes and onto tho- racic slope. Legs pale, with dark brown mac- ulations dorsal distally on patellae, dorsal proximally on tibiae (in addition to broken proximal rings on tarsi and metatarsi), and lat- erally (both sides) on femora and tibiae; the femora may have one or two maculations each side. Dorsal abdominal pattern may coalesce into median chevrons, especially posteriorly, and there are many small dark spots laterally, approaching the pattern found in Admestina. Abdominal venter pale with median gray stripe flanked by several pair of gray spots. Like A. cinctipes, the color pattern of Ad- mestina consists of pale integument (except for the dark carapace) with dark maculations on legs and abdominal venter, and the pale abdominal dorsum has a series of connected dark median chevrons or triangles (more re- stricted to the sagittal plane and sometimes totally coalesced into a black median stripe) and small dark lateral spots, although the body is much narrower. Distribution. — Florida to Louisiana, pos- sibly into eastern Mexico (where two penul- timate males with the appropriate color pat- tern were found); most found on laurel and water oaks. Material examined. — USA: Florida: Alachua County, Gainesville, 7 Febmary 1927, 1 9 (OSU); Lochloosa Wildlife Management Area, September- March, 1976-1985, 33219 (2 eggsacs with 3 and 5 eggs respectively) (G.B. Edwards, D.B. Richman; ESC A); Collier County, Royal Palm Park, March 1929, 19 (W.S. Blatchley, MCZ); Dade County, Homestead, 1 9 (AMNH); Highlands County, Lake Placid, Archbold Biological Station, 24 November 1961, 19 (A.M. Nadler, AMNH); Indian River County, Vero Beach, on Avicennia germinans, 10 July 1990, 1 juv (K. Hibbard & K. Dady, FSCA); St. Lucie County, on Ficus sp., 19 (E. Thompson 6 K. Hibbard, FSCA); Louisiana: Jefferson Parish, Harahan, 15 November 1944, 1 juv (EG. Werner, MCZ); Tammany Parish, Covington, 1 9 (N. Banks, MCZ); Mississippi: Claiborne County, on car, 13 April 1954, 13 (H.K. Wallace, FSCA); MEXICO: Tamaulipas: nr. Gomez Farias, woody plants, 6 June 1983, 1 juv 3 (W. Maddison, MCZ); Veracruz: Los Tuxtlas, dead palm fronds, 1 August 1983, ljuv3 (W. Maddison, MCZ). DISCUSSION Maddison (1996) mentions both Ballus and Admestina among a group of genera possibly 14 THE JOURNAL OF ARACHNOLOGY derived from the Dendryphantinae. Attidops can be included in this group. All three genera have the embolus a distal spiral set perpen- dicular to the tegulum, an unequally bilobed tegulum, and the carapace integument is en- tirely reticulate (with coarser reticulations in the OQ extending past the PLE), which may be characteristic of the group. Admestina and Attidops share a flattened carapace with a long postocular dorsum; cusp-like flange subdistal- ly on the anterior outer comer of male endites; the excavate cymbial tip and the transverse embolar groove; a broadly triangular epigynal plate; and the males have a complete dorsal abdominal scutum. Ballus has a small, sub- distal anterolateral cusp on the male endites. Both the cusp and the flanges in this position might support a relationship to the dendry- phantines (which have cusps), although not necessarily supporting a close relationship be- tween Ballus and the two other genera. Mar- pissa (Marpissinae) is the only other genus with male endite cusps of which I am aware (D. Logunov, pers. comm.); further study may demonstrate the character has phylogenetic value. Attidops cinctipes is the most distinct spe- cies in the genus; its color pattern and heavily sclerotized embolus are similar to Admestina, although otherwise it is like the other Attidops species. The embolus of A. cutleri is less sclerotized and its color pattern is intermedi- ate between A. cinctipes and the A. yoimgi-A. nickersoni pair. The emboli of A. youngi and A. nickersoni are mostly membranous, and the epigyna are barely distinct; A. nickersoni is likely a Florida isolate of A. youngi and the two are sister species. ACKNOWLEDGMENTS My sincerest thanks go the the following institutions and curators for loans of speci- mens: the American Museum of Natural His- tory (AMNH), Norman 1. Platnick and Louis N. Sorkin; Midwestern State University (MSU), Norman Horner; the Museum of Comparative Zoology (MCZ), Herbert W. Levi and Laura Liebensperger; the Ohio State University (OSU), Richard A. Bradley; the United States National Museum of Natural History (USNM), Jonathan Coddington and Scott Larcher; and the Florida State Collection of Arthropods (ESC A). Hank Guarisco (HGC) loaned a specimen from his private collection. Paula Cushing, Norman Horner, and Joan Jass helped find obscure localities. Jim Berry, Bruce Cutler, and an anonymous reviewer provided helpful comments. LITERATURE CITED Alicata, P. & T. Cantarella. 1987. The genus Bal- lus: A revision of the European taxa described by Simon together with observations on the other species of the genus. Animalia, 14:35-63. Banks, N. 1895. A list of spiders of Long Island, with description of new species. J. New York Entomol. Soc., 3:76-93. Banks, N. 1899. Some spiders from northern Lou- isiana. Proc. Entomol. Soc. Washington, 4:188- 195. Banks, N. 1900. Some new North American spi- ders. Canadian Entomol., 32:96-102. Banks, N. 1905. Synopses of North American in- vertebrates. XX. Families and genera of Aranei- da. American Natur., 39:293-323. Banks, N. 1910. Catalog of Nearctic spiders. Bull. United States Nat. Mus., 72:1-80. Barrows, W.M. 1918. A list of Ohio spiders. Ohio J. Sci., 18:297-318. Belanger, G. & R. Hutchinson. 1992. Liste annotee des Araignees (Araneae) du Quebec. Pirata, 1:2- 119. Berry, J.W. 1970. Spiders of the North Carolina Piedmont old-field communities. J. Elisha Mitch- ell Sci. Soc., 86:97-105. Bonnet, P 1955. Bibliographia Araneorum. Tome II A-B. Toulouse, France. Pp. 1-918. Bonnet, P 1957. Bibliographia Araneorum. Tome II G-M. Toulouse, France. Pp. 1927-3026. Bryant, E.B. 1933. New and little known spiders from the United States. Bull. Mus. Comp. Zool., 74:171-193. Comstock, J. 1913. The Spider Book. Doubleday, Doran & Co., Inc., Garden City, New York. 721 pp. Crosby, C.R. & S.C. Bishop. 1928. Araneae. Pp. 1034-1074. In A List of The Insects of New York. Cornell Univ. Agric. Exp. Sta. Mem. 101. Dorris, PR. 1968. A preliminary study of the spi- ders of Clark County Arkansas compared with a five year study of Mississippi spiders. Proc. Ar- kansas Acad. Sci., 22:33-37. Edwards, G.B. 1980. Jumping spiders of the Unit- ed States and Canada: Changes in the key and list (4). Peckhamia, 2:11-14. Kaston, B.J. 1938. Checklist of the spiders of Con- necticut. Bull. Connecticut Geol. Natur. Hist. Surv., 60:175-201. Kaston, B.J. 1948. Spiders of Connecticut. Bull. Connecticut Geol. Natur. Hist. Surv., 70:1-874. Koch, C.L. 1851. Uebersicht des Arachnidensys- tems. Ntimberg, 5:1-77. EDWARDS— SALTICID GENUS ATTIDOPS 15 Levi, H.W. & H.M. Field. 1954. The spiders of Wisconsin. American Midi. Natur., 51:440-467. Maddison, WR 1996. Pelegrina Franganillo and other jumping spiders formerly placed in the ge- nus Metaphidippus (Araneae: Salticidae). Bull. Mus. Comp. Zool., 154:215-368. Marx, G. 1890. Catalogue of the described Ara- neae of temperate North America. Proc. United States Nat. Mus., 12:497-594. Peckham, G.W & E.G. Peckham. 1888. Attidae of North America. Trans. Wisconsin Acad. Sci. Arts Let., 7:1-104. Peckham, G.W. & E.G. Peckham. 1909. Revision of the Attidae of North America. Trans. Wiscon- sin Acad. Sci. Arts Let., 16:355-646. Petrunkevitch, A. 1911. A synonymic index-cata- logue of spiders of North, Central and South America, etc. Bull. American Mus. Natur. Hist., 29:1-791. Piel, W.H. 1991. The Nearctic jumping spiders of the genus Admestina (Araneae: Salticidae). Psy- che, 98:265-282. Platnick, N.I. 1989. Advances in Spider Taxonomy 1981-1987. Manchester Univ. Press, UK. 673 pp. Platnick, N.I. 1993. Advances in Spider Taxonomy 1988-1991. New York Entomol. Soc. and Amer- ican Mus. Natur. Hist., New York. 846 pp. Razoumowsky, G. de. 1789. Araignees. Pp. 241- 247. In Histoire Naturelle du Jorat et de Ses En- virons. Lausanne. 2 volumes. Richman, D.B. & B. Cutler. 1978. A list of the jumping spiders (Araneae: Salticidae) of the United States and Canada. Peckhamia, 1(5):82- 110. Roberts, M.J. 1985a. The Spiders of Great Britain and Ireland: Vol. 1. E.J. Brill, The Netherlands. 229 pp., 100 text figs. Roberts, M.J. 1985b. The Spiders of Great Britain and Ireland: Vol. 3. E.J. Brill, The Netherlands. 256 pp., 237 color plates. Roewer, C.F. 1954. Katalog der Araneae, 2 Band, Abt. B. Institut Royal des Sciences Naturelles de Belgique, Brussels. Pp. 927-1751. Simon, E. 1901. Histoire naturelle des Araignees. Tome 2 fasc. 3. Paris 1901. Pp. 381-668. Stietenroth, C.L. & N.V. Horner. 1987. The jump- ing spiders (Araneae: Salticidae) of the Virginia peninsula. Entomol. News, 98:235-245. Walckenaer, C.A. 1802. Araignee. Pp. 187-250. In Faune Parisienne. Insectes. Paris. Manuscript received I May 1998, revised 10 Jan- uary 1999. 1999. The Journal of Arachnology 27:16-18 A NEW DISEMBOLUS (ARANEAE, LINYPHIIDAE) FROM CAPE COD, MASSACHUSETTS AND LONG ISLAND, NEW YORK Robert L. Edwards: Research Associate, United States National Museum, Box 505, Woods Hole, Massachusetts 02543 USA ABSTRACT. Disembolus bairdi new species is described from the coastal region of northeastern United States. Notes on the habitat, natural history and its associated spiders are provided. The genus Disembolus is one of the groups of small erigonine spiders that are all too often described from single individuals with little or no information on their habitat or natural his- tory. The genus is apparently found only in North America and was revised by Millidge in 1981, who recognized 22 species. The new species described here is common on the dunes adjacent to the salt and brackish water marshes of Cape Cod, Massachusetts and Long Island, New York. Disembolus bairdi new species Diagnosis. — Small spiders averaging 1.20 mm or less in total length. Male: Cymbium with stout knob engaging distal end of tibial apophysis; embolic spiral flat, large. Cephalic dome large and bulbous, sloped backwards, with sulci and pits. Female: Epigynum with broad mantle, posterior plate hyaline, with oval, bubble-like bilateral areas, spermathecae widely spaced. Etymology. — The species is named after Spencer F. Baird, who played the principal role in selecting Woods Hole, Massachusetts, as a center for marine research, and was the first director of the U.S. National Museum. Material examined. — Holotype male, al- lotype female and 24 paratype males and 29 paratype females, collected 23 December 1985 and 22 December 1986, under storm tide debris. West Falmouth Harbor, West Fal- mouth, Barnstable County, Massachusetts. Additional specimens collected at the same lo- cality: five adult females, 14 April 1990, eight adult females, 9 May 1989, and one adult male, 10 October 1990. A gravid female was collected 22 April 1990, in brackish marsh de- bris, Salt Pond, Falmouth. All above material was collected by R.L. Edwards. In December 1995, and in April and May 1996, Miss Jacqui Kluft collected adult male and female and pre- mature specimens in detritus at storm tide lev- els on the beach at Fire Island, New York. Holotype male and allotype female, para- type males and females deposited in the Unit- ed States National Museum, Washington, D.C.; paratype males and females in the Mu- seum of Comparative Zoology, Cambridge, Massachusetts, in the American Museum of Natural History, New York and in the British Museum of Natural History, London. Measurements were made with an ocular micrometer. The cephalic index is the length of the cephalothorx divided by the width. The Tml value is the ratio of the length of meta- tarsus I divided by the distance between the trichobothrium and the proximal end of the metatarsus. Description, — Measurements (mm), means in mm and SD. Female: (n = 12). Total length 1.20 ±0.057, cephalothorax length 0.51 ±0.031, cephalic index 1.29 ±0.082, Tml 0.49 ±0.022, epigynum width 0.16 ±0.009. Male: (n = 12). Total length 1.17 ±0.053, cephalothorax length 0.53 ±0.036, cephalic index 1.24 ±0.061, Tml ratio 0.47 ±0.024, TmlV absent. Tibia without spines in male, in female 0-0- 1-1. Cephalothorax broad, light yellow-brown with thin, dark margin. Clypeal area and dome slightly darker. Eyes margined with black. Vague radii on thoracic portion but otherwise without distinctive markings (Fig. 1). Male with large dome on cephalic portion of ceph- alothorax, sloped to the rear (Fig. 2). Large sulcus at base of lobe with pit immediately behind posterior lateral eye. Chelicera with 16 Figures 1-4. — Disembolus bairdi new species. 1, Male cephalothorax, dorsal; 2, Male cephalothorax, lateral; 3, Left palp, ectal. Note the apparent conjunction of the distal tip of the tibial apophysis and the knob on the cymbium; 4. Epigynum, ventral. Abbreviations: E, embolus; K, knob-like process on cym- bium; M, suprategular apophysis, membranous part. four promarginal teeth, two retromarginal. Abdomen grey to black, rarely with indistinct chevrons on posterior half, venter lighter. Ster- num with darker margin. Legs light yellow- orange. Palp (Fig. 3) with broad, relatively flat base of the embolus spiral with a recurved distal loop and projecting fan-like membra- nous suprategular process (Fig. 3). Cymbium with stout, darkened knob-like process that engages distal end of tibial apothesis (Fig. 3). Palpal tibia with a single trichobothrium, a stout seta, and recurved distal tip. Female cephalothorax without dome, colored essen- tially as male. Epigynum with broad, darker mantle anteriorally (Fig. 4). Posterior plate glassy, with bilateral swellings. Spermathecae widely spaced, clearly visible. Spermathecal ducts not clearly visible. Millidge (1981) provided partial keys for the males and females of Disembolus. For fe- males (table 1), D. bairdi new species fits best under line 4, “with posterior plate notably convex and glassy in appearance, and with dark colored bar anterior to plate.” For males (table 2), bairdi new species fits best under line 3, “Carapace with a lobe which has a hole and sulcus on each side,” and under line 3-iv, “tibial apophysis with small hook distally.” It 18 THE JOURNAL OF ARACHNOLOGY will be noticed that D. bairdi new species is much smaller (±1.2 mm) than the species list- ed in this line. Natural history. — This species is found under mats of organic debris without obvious webbing, usually Zostera and/or Spartina, found at or near the highest tide levels in areas otherwise protected from direct ocean influ- ence. The spider occupies the interface be- tween these mats and the underlying sand. When the mats are lifted the spiders will sometimes burrow into the sand. Immature in- stars were taken in March, September and Oc- tober, with most individuals mature in Decem- ber. Mature females have been taken through May. Apparently most of the species of this genus are found in the colder months of the year (Millidge 1981). It shares the highest tid- al zone habitat with Coloncus americana Chamberlin & Ivie 1944, another erigonine species that matures in the late fall. Coloncus was as abundant as Disembolus bairdi in the West Falmouth Harbor area and was also found by Miss Jacqui Kluft on Long Island with D. bairdi and Erigone brevidentata Emerton. Erigone aletris Crosby & Bishop 1928, and Grammonota trivittata Banks 1895 were also abundant in or on detrital mats on Cape Cod. These species also mature during the colder months of the year. Two additional species of Disembolus occur on Cape Cod. Disembolus sacerdotalis (Crosby & Bishop 1933) has been taken in grassy areas in mixed woodlands and several females of an appar- ently undescribed species have been taken in pine litter. Millidge (1981) notes the difficulty in see- ing the transparent spermathecal ducts of Di- sembolus species. In D. bairdi, they are vir- tually invisible, even in prepared material. Three epigyna were removed and mounted in a lactic-acetic acid mixture, PVA, and euparol. In PVA mounted material what appeared to be wide-mouthed spermathecal ducts could be discerned only with considerable uncertainty. Histological preparations may be necessary to fully describe the internal genitalia of Disem- bolus species. ACKNOWLEGMENTS Dr. Charles Dondale and Mr. James Redner, Research Branch, Agriculture Canada, exam- ined my specimens and confirmed my opinion that this was an undescribed species. Eric Ed- wards kindly critiqued the draft manuscript. Comments and corrections provided by re- viewers were much appreciated. LITERATURE CITED Millidge, A.E 1991. The erigonine spiders of North America. Part 4. The genus Disembolus Chamberlin and Ivie (Araneae: Linyphiidae). J. Arachnol., 9:259-284. Manuscript received 29 April 1998, revised 5 Oc- tober 1998. 1999. The Journal of Arachnology 27:19-24 SINOPODA, A NEW GENUS OF HETEROPODINAE (ARANEAE, SPARASSIDAE) FROM ASIA Peter Jager: Institute for Zoology, Johannes Gutenberg-University, Saarstr. 21, 55099 Mainz, Germany ABSTRACT. Sinopoda new genus (Araneae, Sparassidae, Heteropodinae) is described from Asia. It is recognizable only from genital characters. At present, the new genus comprises 25 species from Japan, Korea, China, Thailand, Malaysia and east India. Heteropoda campanacea, H. forcipata, H. hamata, H. koreana, H. licenti, H. marsupia (?), H. minschana, H. serrata, H. shennonga, H. stellata and Panaretidius microphthalmus are placed in Sinopoda new genus. Relationships to other heteropodine genera are dis- cussed. Jager (1998a) identified somatic characters, which are useful for distinguishing suprage- neric taxa in the Sparassidae. At present, nine heteropodine genera are known from Asia: Adrastis Simon 1880, Heteropoda Latreille 1804, Panaretidius Simon 1906, Panaretus Simon 1880, Pandercetes L. Koch 1875, Par- hedrus Simon 1887, Spariolenus Simon 1880, Torania Simon 1886, Yiinthi Davies 1994. Most of the species belonging to Sinopoda new genus were formerly described under Heteropoda. METHODS The following abbreviations are used in the text: AC, anterior width of carapace; AL, ab- domen length; ALE, PME, AME, PEE, ME, LE, AE, PE refer to anterior lateral eyes, pos- terior median eyes, etc.; AW, abdomen width; BE, body length; CH, carapace height; CL, carapace length; CW, carapace width; GL, gnathocoxae length; GW, gnathocoxae width; LL, labium length; LW, labium width; mm, millimeters; n, number of examined specimen; SL, sternum length; SW, sternum width. NHMB, Naturhistorisches Museum Basel, NHMW Naturhistorisches Museum Wien, SMF Senckenberg Museum Frankfurt, ZMB, Zoologisches Museum der Humboldt-Univer- sitat Berlin. Spine notation follows Davies (1994), ex- ceptions are given in brackets. Dissected epi- gyna were cleared in lactic acid. Measure- ments are in mm. The variation of measurements is given first, followed by mea- surements of the lectotype in brackets. For characterizing the new genus and giv- ing information of its distribution, 14 undes- cribed species were examined, which are rec- ognized by the author as species of Sinopoda new genus. SYSTEMATICS Sinopoda new genus Type species. — Sarotes forcipatus Karsch 1881. Diagnosis. — Close to Heteropoda but male palps with embolic apophysis. Conductor membranous, arising from the distal part of the tegulum (Figs. 2, 3). Tibial apophysis bi- furcate, dorsal branch longest (Fig. 1). Copu- latory ducts of vulva uncoiled, running from anterior genital orifice to posterior part of spermathecae. Typically they are fused along the median line (Fig. 7). Spermathecae divid- ed into a basal part and a head, situated lat- erally from the entrance of the copulatory duct into the spermathecae (Fig. 8). Etymology. — The name is an acronym of the prefix sino- (belonging to China) and Het- eropoda. The gender is feminine. Characters of checked Sinopoda-spe- cies. — Small to large spiders (3-25 mm) with laterigrade legs. Carapace: widest above coxa II to III; CL/CW 1.0 to 1.3; CW/AC in S 1.8-2. 3, in $ 1.4-1. 8; fovea and highest point of carapace mostly above coxa III; CL/ CH 2. 7-6. 2. Sternum broadest between coxa II; SL/SW 0.9-1. 1 (0.9 = BE < 6mm). LW/ LL 0.9-L3 (BL > 8mm), L3-1.7 (BL < 6mm). GL/GW 1. 5-2.4. Abdomen: oval, AL/ AW 1.2-3. 3 (mean l.l\ n = 24). Eyes: in two rows, viewed from above both rows re- curved. Chelicerae: with 3 anterior and 4 19 20 THE JOURNAL OF ARACHNOLOGY Figures 1-11. — 1-8. Genital characters of paralectotypes of Sinopoda forcipata NEW COMBINATION from Japan. 1. Lateral aspect of left palp; 2. Ventral aspect of left palp; 3. Ventral aspect of embolus and conductor; 4, 5. Cross sections of embolus; 6. Ventral aspect of epigynum; 7. Dorsal aspect of vulva; 8. Schematic course of right spermatheca; 9-11, Male genital characters of Heteropoda venatoria Linnaeus: 9. Ventral aspect of tegulum; 10, 11. Cross sections of embolus. Abbreviations: Bs, Basal part of sper- mathecae; cd, copulatory duct; co, conductor; ea, embolic apophysis; em, embolus; fd, fertilization duct; hs, head of spermathecae; le, ledges of epigynum; sd, sperm duct; st, subtegulum; te, tegulum. Scales for Figs. 1-3, 6, 9 = 1.0 mm; Figs. 4, 5, 10, 11 =0.1 mm. posterior teeth and several denticles in be- tween them; as in all other heteropodine spe- cies majority of denticles near the anterior teeth (Jager 1998a). Legs: length in most cas- es 2143 {n = 24), also 2413 {n = 7), 2431 {n = 3), 2134 {n 2). In cavernicolous spe- cies and in all males legs, especially tibiae and metatarsi, elongated relative to body length. Scopulae on tarsi and distal part of metatarsi, sparse in some species. Palpal claw in females present, with 5 to 11 long teeth. Color: variable, pale brown and pale yellow JAGER— NEW GENUS OF HETEROPODINAE FROM ASIA 21 in cavernicolous species to dark brown, with or without pattern. Male palp: Part of subtegulum visible in ventral view. Tegulum oval in general shape. Embolus more or less s-shaped, arising from prolateral part of tegulum in a 6 to 8 o’clock- position, basally embedded in a tegular flange (Fig. 2). Tip of embolus near membranous and flattened conductor (Fig. 3). Embolus broad- ened, containing just one cavity (sperm duct) as in Sinopoda forcipata NEW COMBINA- TION (Figs. 4, 5), some species with tooth in subdistal position. Embolic apophysis arising prolaterally from the basal half of the embo- lus, embedded distally in a subtegular furrow. By comparison, embolus of Heteropoda fili- form, not s-shaped (Fig. 9) and with two tub- uliform cavities, a narrow sperm duct and a large one, which is connected with the tegular cavity (Figs. 10, 11). Conductor thicker, sheath-like, tapering and arising on a prola- teral position of tegulum (Fig. 9). Epigynum: Characterized by copulatory or- ifices covered by two ledges running from a medio-anterior to latero-posterior position. Two posterior lobes in some species divided in the median line. The darkened field of epi- gynum sometimes extended in two anterior elongated bands, these sometimes separated from epigyneal field (Fig. 6). In Heteropoda rims aligned in direction of body axis, in most cases short. Copulatory ducts coiled, if touch- ing each other then only at their origin. Distribution. — Japan, Korea, China, Thai- land, Malaysia (Selangor, Sarawak), east India (Assam). Natural history — Little is known about the representatives of this genus. Deeleman (1998) collected specimens in Borneo (W-Sa- rawak, Matang reserve) from tree bark and grass. The author observed individuals in Chi- na (Shaanxi, Taibai Shan) in leaf litter, run- ning like lycosids during the day. Another species there inhabited natural rock fissures and man-made walls. This species came out at a certain time in the evening to ambush for prey always near the same spot (Jager 1998b). Other species are reported from caves, though not all of them appear to be restricted to caves. Other species included in Sinopoda. — Sinopoda campanacea (Wang 1990) NEW COMBINATION, S. hamata (Fox 1937) NEW COMBINATION, S. koreana (Paik 1968) NEW COMBINATION, 5. licenti (Schenkel 1953) NEW COMBINATION, 5. marsupia (Wang 1991)(?) NEW COMBINA- TION, S. minschana (Schenkel 1936) NEW COMBINATION, N. serrata (Wang 1990) NEW COMBINATION, S. shennonga (Peng, Yin & Kim 1996) NEW COMBINATION, Y stellata (Schenkel 1963) NEW COMBINA- TION, S. microphthalma (Fage 1929) NEW COMBINATION. Sinopoda forcipata (Karsch 1881) NEW COMBINATION (Figs. 1-8) Sarotes forcipatus Karsch 1881:38. (Syntypes: 33 (ZMB 2696, 2698/2), 29 (ZMB 2694, 2695), 1 9 with epigynum region dissected (epigynum is missing) and a subadult 3 (both ZMB 2696) - labeled: Japan, leg. Hilgendorf, det. Karsch -, ex- amined. 1 3 (ZMB 2698, both palps and legs en- tire, PJ 921) hereby designated as lectotype, others as paraiectotypes) Heteropoda forcipata: Bosenberg & Strand 1906: 276; Jarvi 1912:82, 113; 1914:209; Fox 1936: 127; 1937:7; Suzuki 1952:3, 14; Roewer 1954: 714; Bonnet 1957:2189; Yaginuma 1960:113; 1962a:52; 1962b:75; 1962c:130; 1963:51; 1971: 113; 1975:190, 1986:199; Chikuni 1989:130; Ya- ginuma 1990:270; Ono et al. 1995:128. Diagnosis. — Male with distal part of em- bolic apophysis bent at a right angle (Figs. 2, 3); dorsal branch of tibial apophysis broad and tapering (Fig. 1). 9 posterior lobes of epigyn- um point towards the median line, closest dis- tance between ledges about Vq of total width of epigynum (Fig. 6); head of spermathecae two times as long as broad, nearly constant in width (Fig. 7). Redescription of male. — CL 7. 3-9. 8 (9.1), CW 6.4-8.8 (7.9), AC 3.4-3.9 (3.7), CH 1.5- 2.4 (1.9), AL 8.0-11.0 (10.2), AW 4.7-6.3 (5.5). Color: Pale yellowish-brown, without distinct pattern. Eyes: Diameters AME 0.30- 0.44, ALE 0.41-0.53, PME 0.34-0.41, PLE 0.45-0.53, interdistances AME-AME 0.22- 0.27, AME- ALE 0.07-0.12, PME-PME 0.33- 0.47, PME-PLE 0.52-0.58, AME-PME 0.44- 0.49, ALE-PLE 0.36-0.47, clypeus AME 0.41-0.55, clypeus ALE 0.40-0.62; both rows recurved in dorsal view. Legs and palps: 2143 (measurements see Table 1). Spination: Palps 131, 101, 211(smaiy0)l. Femora I-II 323, III 333, IV 331. Patellae 101. Tibiae 2326. Meta- tarsi I 102(1)4, II 1(2)01(2)4, III 2026, IV 3036. 22 THE JOURNAL OF ARACHNOLOGY Table 1. — Leg and palp measurements in males {n = 8) of Sinopoda forcipata NEW COMBINATION. Leg segment Palp I II III IV Femur 4.3-5. 1 10.1-12.0 11.0-13.6 9.1-11.1 9.6-11.8 Patella 1. 5-2.4 3.5-5.0 3. 8-5.0 2.7-4.3 3. 1-4.1 Tibia 2.4-2.9 11.0-13.0 12.1-14.2 8.8-10.7 9.7-11.5 Metatarsus — 10.8-13.8 12.5-15.4 8.8-10.8 10.6-12.7 Tarsus 3.6-4.3 3.5-4.0 3.9-4.5 2.7-3.5 3.4-3.6 Total 11.8-14.7 39.0-47.1 53.3-52.4 32.6-39.1 36.7-43.1 Female.— CL 9.7-10.2, CW 8.2-8.8, AC 4.9-5.3, CH 2.4-2.6, AL 10.6-13.9, AW 6.0- 8.3. Color: As in male. Eyes: Diameters AME 0.29-0.44, ALE 0.48-0.55, PME 0.37-0.42, PLE 0.51-0.58, interdistances AME-AME 0.28-0.42, AME- ALE 0.10-0.22, PME-PME 0.48-0.55, PME-PLE 0.63-0.77, AME-PME 0.54-0.67, ALE-PLE 0.58-0.63, clypeus AME 0.63-0.74, clypeus ALE 0.56-0.69. Legs and palps: 2143 (measurements see Ta- ble 2). Spination: Palps 131, 1(0)01, 2121, 1014. Femora I-III 3(4)2(3)3(274), IV 33(2)1(2). Patellae 101. Tibiae I-II 22(0/1)26, III-IV 23(1/2)26. Metatarsi I-II 1014, III 2014(2016/2026), IV 3036. Palpal claw with 7-8 teeth. Other material examined. — 332$ (NHMB 551, labeled: Clubionidae, Heteropoda aulica (L. K.), Japan, Yokohama, G.v.R. Merian), 1 ? (ZMB 2396, labeled: Sarotes aulicus L. Koch. Nagasaki Westrh.), Id (SMF 4578, labeled: Heteropoda in- victa (L. Koch), Id, Japan: Saga, W. Donitz S.), Id (SMF 4595, labeled: Eusparassid No. 41, Het- eropoda forcipata (Ka.), Id, China, Roewer det. 1933), Id (NHMW 1879.11.20, labeled: holotype of Heteropoda invicta, Inv.-No. 17.863, Coll. Musei Vindobonensis, Sarotes invictus, Erber Tausch). Note. — In Japan two species exist, Sino- poda forcipata NEW COMBINATION and 5. stellata NEW COMBINATION. The descrip- tions by L. Koch (1878) of Heteropoda aulica and H. invicta have caused some problems. Individuals of S. forcipata NEW COMBI- NATION have been examined by the author, which were determined as H. aulica or H. in- victa. H. aulica clearly does not belong to Sin- opoda new genus. H. invicta cannot clearly be recognized from the original description. There are two specimens deposited in NHMW in one vial, which are labeled as ’holotype’, one male and one female. Koch described only one female. The female belongs to Het- eropoda venatoria Linnaeus 1767, the male to Sinopoda sp. The original drawing No. 32 by Koch does not permit a clear association with this genus. If the species belongs to Sinopoda new genus, the epigynum is turned upside down. Color variation. — In most other specimens examined color dark yellow to brown. Distribution. — Japan, China. DISCUSSION Autapomorphic characters of the new genus are: embolic apophysis, bifurcate tibial apoph- ysis, and course of epigyneal rims. The shape of the conductor resembles that of Barylestis (Africa), Pandercetes (Asia, Australia), Spa- riolenus (Asia) and two undescribed genera from Asia. In contrast to them, Heteropoda Table 2. — Leg and palp measurements in female {n = 5) of Sinopoda forcipata NEW COMBINATION. Leg segment Palp I II III IV Femur 4.0-4.6 9.5-10.4 10.6-11.5 9.1-10.3 9.7-10.8 Patella 2.1-2.3 4.4-4.9 4.6-4.9 3.8-4.2 3.7-3.9 Tibia 2.8-3.3 10.0-10.6 10.2-11.2 8.1-9.3 8. 8-9.7 Metatarsus — 9.0-9.5 9.1-10.0 7.3-8.4 9.0-9.8 Tarsus 4.1-4.5 2.8-3. 1 2.8-3.0 2. 1-2.9 2.2-3. 1 Total 13.0-14.4 36.9-38.4 36.8-40.5 30.9-34.8 34.2-37.3 JAGER— NEW GENUS OF HETEROPODINAE FROM ASIA 23 (all continents) and Parhedrus (Asia) possess sheath-like conductors. Jarvi (1912, 1914) recognized the female genital organs of S. for- cipata NEW COMBINATION as an extreme specialization and stated that in a hypothetical ancestral epigynum the lateral lobes and the ducts were medially fused. This conclusion can be confirmed, as in several species a me- dian furrow is present. Furthermore, he ho- mologized the head of spermathecae of Sino- poda for cipata NEW COMBINATION with the coils in Heteropoda spp., which is here rejected. In Heteropoda spp. the head of the spermathecae is absent. As S. forcipata has uncoiled ducts with a simple cavity, it is sup- posed that this type of vulvae has been divid- ed from Heteropoda s. str. and other hetero- podine genera a relatively long time ago. The most primitive species of Sinopoda new genus is known from W-Sarawak (Borneo). The typ- ical epigyneal ledges of Sinopoda spp. are ho- mologous with those of Heteropoda spp., but in the latter genus they either are covered by the lateral lobes or are situated in the anterior part of the epigynum. For these reasons, I con- sider Sinopoda new genus a monophyletic group, whose exact placement within the Het- eropodinae is not yet clear. To elucidate inter- generic relationships more taxa need to be ex- amined. ACKNOWLEDGMENTS Thanks are due to Dr. C. Deeleman, Dr. C. Rollard, Prof. C.-M. Yin, Dr. Z.-F. Chen, Dr. J. Coddington, Dr. J. Dunlop, Dr. M. Gras- shoff. Dr. J. Gruber, Dr. A. Hanggi, Dr. R Hill- yard, Dr. T Kronestedt, Prof. S. Li, Prof. Dr. J. Martens, Dr. H. Ono, Dr. N. Platnick, Dr. P. Schwendinger, Prof. D. Song and P. West for providing specimens, literature, comments on the manuscript and helpful discussion. This study was partly supported by the Stifterver- band fiir die Deutsche Wissenschaft and by a scholarship from the Bundesland Rheinland- Pfalz. LITERATURE CITED Bonnet, P. 1957. Bibliographia Araneoram, 2(3), Pp. 1927-3026. Bosenberg, W. & E. Strand. 1906. Japanische Spin- nen. Abh. Senckenberg. Naturf. Ges., 30:93- 422. Chikuni, Y. 1989. Pictorial Encyclopedia of Spi- ders in Japan. Kaisea-Sha Publishing Co, Tokyo. 308 pp. Davies, V.T 1994. The huntsman spiders Hetero- poda Latreille and Yiinthi gen. nov. (Araneae: Heteropodidae) in Australia. Mem. Queensland Mus., 35(1):75-122. Fage, L. 1929. Arachnida. In Fauna of the Batu caves, Selangor. J. Fed. Malay Stat. Mus., 14(3/ 4):356-364. Fox, I. 1936. Chinese spiders of the families Age- lenidae, Pisauridae, and Sparassidae. J. Washing- ton Acad. Sci., 26(3): 121-128. Fox, I. 1937. New species and records of Chinese spiders. American Mus. Novit., 907:1-9. Jager, P. 1998a. First results of a taxonomic revi- sion of the SE Asian Sparassidae (Araneae). Pp. 53-59. In Proc. 17th European Colloquium Ar- achnoL, Edinburgh, 1997. Burnham Beeches, Bucks. Jager, P. 1998b. An oldfashioned way to catch sparassid spiders. Newsl. British Arachnol. Soc., 82:4. Jarvi, T.H. 1912. Das Vaginalsystem der Sparas- siden. I, Allgemeiner Teil. Ann. Acad. Sci. Fen- nicae (A), 4(1): 1-1 17. Jarvi, T.H. 1914. Das Vaginalsystem der Sparas- siden. II. Spezieller Teil. Ann. Acad. Sci. Fen- nicae, 4(1): 118-248. Karsch, F. 1881. Diagnoses Arachnoidarum Japon- iae. Berliner Entomol. Zeitschr., 25:35-40. Koch, L. 1878. Japanesische Arachniden und My- riapoden. Verb, zool.-bot. Ges. Wien, 27:735- 798. Ono, H., E. Shinkai & K. Kato. 1995. The spider fauna of Fukushima prefecture. Northern Japan. Mem. Natn. Sci. Mus. Tokyo, 28:113-133. (In Japanese). Paik, K.Y. 1968. The Heteropodidae (Aranea) of Korea. Kyungpook Univ. Theses Coll., 12:167- 185. Peng, X.-J., C.-M. Yin & J.-P. Kim. 1996. One species of the genus Heteropoda and a descrip- tion of the female Heteropoda minschana Schen- kel, 1936 (Araneae: Heteropodidae). Korean Ar- achnol., 12:57-61. Roewer, C.R 1954. Katalog der Araneae, 2a, Pp. 1-923. Schenkel, E. 1936. Schwedisch-chinesische wis- senschaftliche Expedition nach den nordwes- tlichen Provinzen Chinas, unter Leitung von Dr. Sven Hedin und Prof. Sii Ping-chang. Araneae gesammelt vom schwedischen Arzt der Expedi- tion Dr. David Hummel 1927-1930. Ark. Zool., 29 A(l):l-314. Schenkel, E. 1953. Chinesische Arachnoidea aus dem Museum Hoangho-Peiho in Tientsin. Bol. Mus. Nac. Rio de Janeiro (N.S., Zool.), 119:1- 108. Suzuki, S. 1952. Cytological studies in spiders. II. Chromosomal investigation in the twenty two species of spiders belonging to the four families 24 THE JOURNAL OF ARACHNOLOGY Clubionidae, Sparassidae, Thomisidae and Oxy- opidae, which constitute Clubionoidea, with spe- cial reference to sex chromosomes. J. Sci. Hiro- shima Univ., (Ser. B, Div. 1), 13(1): 1-52. Wang, J.F. 1990. Six new species of the spiders of the genus Heteropoda from China (Araneae: Heteropodidae). Sichuan J. ZooL, 9(3):7-ll. Wang, J.F. 1991. Two new species and three sup- plemental descriptions of family Heteropodidae from China (Arachnida: Araneae). Sichuan J. ZooL, 10 (l):3-6. Yaginuma, T. 1960. Spiders of Japan in Colour. Hoikusha Publishing Co, Osaka. 186 pp. Yaginuma, T. 1962a. The spider fauna of Japan. Arach. Soc. East Asia, 1962:1-74. Yaginuma, T. 1962b. Cave spiders in Japan. Bull. Osaka Mus. Nat. Hist., 15:65-77. Yaginuma, T. 1962c. Spiders from Osumi penin- sula, Mt. Takakuma and Mt. Kirishima, Kyushu, Japan. Misc. Rep. Res. Inst, Nat. Resour., 56-57: 129-136. Yaginuma, T. 1963. Spiders from limestone caves of Akiyoshi plateau. Bull. Akiyoshi-dai Sci. Mus. 2:49-62. Yaginuma, T. 1971. Spiders of Japan in Colour. Hoikusha Publishing Co, Osaka. 197 pp. Yaginuma, T. 1975. The spider fauna of Japan (V). Fac. Let. Rev. Otemon Gakuin Univ., 9:187-195. Yaginuma, T 1986. Spiders of Japan in Colour (new ed.). Hoikusha Publishing Co, Osaka. Yaginuma, T. 1990. Check list of Japanese spiders (1989). Pp. 270-271. In Spiders -Etymology of Their Scientific and Japanese names. (T. Yagin- uma, Y. Hirashima & C. Okuma). Kyushu Univ. Press. Kyushu, Japan. Manuscript received 1 May 1998, revised 11 No- vember 1998. 1999. The Journal of Arachnology 27:25-36 CARBINEA, A NEW SPIDER GENUS FROM NORTH QUEENSLAND, AUSTRALIA (ARANEAE, AMAUROBIOIDEA, KABABININAE) Valerie Todd Davies: Queensland Museum, RO. Box 3300, South Brisbane, Australia ABSTRACT. The distribution of four species of Carbinea new genus in the Wet Tropics region of northern Queensland documents the species’ richness and local endemism. The new species are C Ion- giscapa, C. breviscapa, C. wunderlichi and C. robertsi. They are placed in the sub-family Kababininae which is removed from the Amphinectidae (Davies 1995) as there is evidence that it does not belong there. The placement of this clade remains problematical. Amaurobioid spiders abound in Australia. In rainforest surveys (Monteith & Davies 1984) they were usually found to have the highest number of species after ‘theridiids’ and salticids. From the number of identifiable species in these surveys it was estimated that probably less than 20% of the Australian spi- ders are described. At present about 40 amau- robioid genera are known; however, their placement in families is unresolved. In de- scribing Kababina (Davies 1995) I said that it belonged in a group of undescribed genera. One of these genera is ecribellate and is de- scribed here. METHODS Spiders were collected from rainforests in the Wet Tropics region of North Queensland between latitudes 16°16'S, 17°06'S. All ma- terial is deposited in the Queensland Museum (QM Brisbane, Australia). Measurements are in millimeters. Coordinates are given in square brackets when these were not included in the original label data. Lengths of epigynal scape, cymbium and tibial apophysis were measured by linear micrometer on a dissect- ing microscope and converted to millimeters. Length of the epigynal scape was measured from the posterior edge of the fossa. Notation of spines follows Platnick & Shadab (1975). Illustrations were drawn with the aid of a camera lucida. The left male palp is illus- trated. Collection methods include litter-sieving followed by heat extraction in funnels, pit-fall (PF) collection, pyrethrum (PY) spraying of tree-trunks and fallen logs, hand collecting from under logs in daylight and night collect- ing. Most spiders were collected by G.B. Monteith (GBM) and fellow collectors D. Cook (DC), D. Yeates (DY), G. Thompson (GT), H. Janetzki (HI), L. Roberts (LR) and the Australian New Zealand Scientific Explo- ration Society (ANZSES). Table 1 lists ana- tomical abbreviations used in Table 3 and the text; abbreviations on illustrations are ex- plained in the legends to figures. SYSTEMATICS Subfamily Kababininae Diagnosis. — Three-clawed spiders. Color- ation of the abdomen varies from pale to dark grey-black with pattern of light spots in vague chevron pattern. Carapace highest in foveal region (Fig. 1); eyes directed forwards (Fig. 2). Posterior eye row straight or slightly recurved, anterior row straight (Fig. 3); AME reduced. Two retromarginal and two promar- ginal cheliceral teeth (Fig. 5); prolateral fil- amentous seta at base of fang longer than other setae. Labium about as long as wide; sternum slightly longer than wide, pointed posteriorly (Fig. 6). Legs 1423/4123. Feath- ery hairs on legs (Fig. 24). Tarsal trichoboth- ria increasing in length distally; bothrium collariform (Fig. 25). Tarsal organ slit-like, broadening distally. Epigynum with medial 25 26 THE JOURNAL OF ARACHNOLOGY Table 1. — List of anatomical abbreviations. AL abdomen length ALE anterior lateral eyes ALS anterior lateral spinnerets AME anterior median eyes APOPH apophysis AW abdomen width CB cymbium CH cheliceral CL carapace length CR cribellum CW carapace width E embolic EPIG epigynal MAP major ampullate spigots MT metatarsal PCR paracribellar spigots PLE posterior lateral eyes PME posterior median eyes PLS posterior lateral spinnerets PMS posterior median spinnerets RTA retrolateral tibial apophysis TRICH trichobothria fossa wider than long (Fig. 8); spermathecae posterior (Fig. 9) or lateral to fossa. Male palp with rounded tegulum (Fig. 16); course of sperm duct showing clearly. Membranous conductor; embolus with or without proximal embolic apophyses; without median apoph- ysis. Tibial apophysis with ventral and dorso- retrolateral branches (Fig. 17). Cribellum (two fields) present or absent in females, ab- sent in males; proximal calamistrum with one row of setae. Large broad colulus (Figs. 4, 41) when cribellum is absent. Anterior spin- nerets largest with short conical terminal seg- ment; two major ampullate spigots on female ALS; one and a nubbin on male ALS. Pos- terior spinnerets with long slender terminal segment. Carbinea new genus Type species. — Carbinea longiscapa Etymology. — The generic name is from the Carbine Tableland, north Queensland, the geo- graphic area where three species have been collected. Diagnosis. — Ecribellate (cf. Kababina) spi- der. The epigynum has a long, medium or short posterior scape. The embolus and con- ductor arise antero-ventrally on the tegulum; the embolus has two elaborate brush-like apophyses (cf. Kababina) proximally. KEY TO CARBINEA SPP. 1. Posterior epigynal scape medium to long, 0.4-0. 7; S tibial apophysis half or more than half the length of cymbium 2 Posterior epigynal scape short, 0.3-0. 4; S tibial apophysis a third or less the length of cymbium 3 2. Epigynal scape very long, 0.6-0, 7. Tibial apophysis more than half the length of the cymbium longiscapa Epigynal scape medium length (0.4). Tib- ial apophysis about half the length of the cymbium robertsi 3. Epigynal scape short (0.3). Tibial apoph- ysis short, thick without heel . . . breviscapa Epigynal scape medium (0.4). Tibial apophysis with heel wunderlichi Carbinea longiscapa new species (Figs. 1-13, 22, 23, 42, 43. Tables 2, 4) Types. — Australia: North Queensland. Holotype $, Stewart Ck., 4 km NNE Mt. Spurgeon, Carbine Tableland, Camp 1, 16°24'S, 145°13'E, 1250-1300 m, 15-20 Oc- tober 1991, GBM, HJ, DC, LR (QM S30283). Paratypes: S, same data as holo- type (S30284); 4S, PF (S30285); $, c5 (S30286); <5 (S30287); 4d, PF (S30288); 2?, 7 km N Mt. Spurgeon, Camp 2, 16°22', 145°13', 1200-1250 m, 17-19 October 1991; GBM, HJ, DC, LR (S30289); ?, c?, PF, (S30290); 2$, (5, PF, (S30291); 9 (S30292); 29, Upper Whyanbeel Ck, 16°23', 145°17', 150 m, PY, 5 September 1992, GBM (S35247); 9, Black Mtn., 4.5 km N Mt. Spurgeon, 16°24', 145T2', 1240 m, PY, 17 October 1991, GBM, HJ (S35223); 26, Kar- nak-Devils Thumb, 16°24', 145T8', 8-12 km NW Mossman, 1120 m, PF, 26 December 1989-15 January 1990, ANZSES (S35231); 26, 16°23', 145°17', 1080 m, PF (S35232); 9,6, Stony Ck., 2.5 km NE Mt. Spurgeon, 16°25', 145°139 1200 m, 15-21 October 1991, PF, GBM, HJ, DC, LR (S30294); d. Head of Roots Ck., 12 km WNW Mossman 16°24', 145°15', 1200 m, PF, 28 December 1989-11 January 1990, ANZSES (S35230); 9, Upper Cow Ck., 1.5 km NE Mt. Spur- geon, 16°26', 145°13', 1180 m, 15-21 Octo- ber 1991, PF, GBM, HJ, DC, LR (S30293); 9, 2(3, Pauls Luck, 16°26', 145°15', 1100 m, PF, 28-30 November 1990, GBM, HJ, DC (S35224); 9, 2 km SE Mt. Spurgeon via Mt. Carbine 16°27', 145°12', 1100 m, 20 Decem- ber 1988, PY, GBM, GT (S30295); 9, 20 De- DAVIES— A NEW AUSTRALIAN AMAUROBIOID 27 Figures 1-13. — Carbinea longiscapa new species. 1-10, Female. 1, Cephalothorax, lateral view; 2, Eyes, chelicerae, frontal view; 3, Eyes, dorsal view; 4, Colulus, spinnerets; 5, Chelicera; 6, Endites, labium, sternum; 7-9, Epigynum, lateral, ventral, dorsal views; 10, Epigynum (Mossman Bluff); 11-13, Male. 11, Palp; 12, Tibial apophysis; 13, Tibial apophysis (Mossman Bluff). 28 THE JOURNAL OF ARACHNOLOGY Table 2. — Leg lengths of $((?) Carbinea longiscapa new species. Leg I Leg II Leg III Leg IV Femur 2.1 (2.3) 1.7 (1.8) 1.5 (1.7) 2.0 (2.2) Patella 0.6 (0.6) 0.6 (0.6) 0.5 (0.5) 0.6 (0.6) Tibia 2.1 (2.2) 1.4 (1.5) 1.3 (1.2) 2.0 (2.1) Metatarsus 2.0 (2.4) 1.5 (1.7) 1.5 (1.6) 2.1 (2.4) Tarsus 1.3 (1.3) 1.0 (1.0) 0.9 (0.9) 1.2 (1.2) Total 8.1 (8.8) 6.2 (6.6) 5.7 (5.9) 7.9 (8.5) cember 1988-4 January 1989, GBM, GT, ANZSES (S35222); ;2fe, <3, 9 km W. Moss- man, 16°28', 145n6', 1000 m, PY, 22 Decem- ber 1989, GBM, ANZSES (S35229); ?, 10 km W Mossman, 1100-1300 m, 17-18 De- cember 1988, GBM, GT (S16540); 2?, Mossman Bluff Camp, 16°28', 145°17', 1000 m, PF, 30 November 1990, GBM, HJ (S35226); S , Mossman Bluff Track, 5-10 km W Mossman, 1180 m, PF, 1-17 January 1989, GBM, GT, ANZSES (S27724); (3, 17- 31 December 1988, (S35227); (3, 10 km W Mossman, 1200 m, PY, 17 December 1988, GBM, GT (S35228); ?, (3, Mt. Demi sum- mit, 16°30', 145n9', 1000 m, PY, 16-17 De- cember 1995, GBM (S35225); $, Upper Figures 14-21. — Carbinea breviscapa new species. 14-15, Epigynum, ventral, dorsal 16-17, Male palp, ventral, retrolateral; 18, Tibial apophysis; 19-20, from Black Mountain; 19, Epigynum; 20, Tibial apophysis; 21, Tibial apophysis from Windsor Tableland, e = embolus, ea = embolic apophysis, c = conductor. DAVIES— A NEW AUSTRALIAN AMAUROBIOID 29 Figures 22-28. — 22-23, Carbinea longiscapa new species. 22, Expanded palp, prolateral; 23, Embolic region, anterior. 24-28, Carbinea breviscapa new species. 24, Feathery hair, leg I; 25, Trichobothrium, tarsal organ on tarsus I; 26, Epigynum; 27, 28, Embolic region male palp; 27, Prolateral view; 28, Embolic apophyses, embolus, conductor, retrolateral view, e = embolus. Leichhardt, Mt. Lewis, 16°35', 145°16', 840 m, stick brushing, 18 November 1997, GBM (S39197). Etymology. — The specific epithet is from the Latin, “longus”, long and “scapus”, stem in reference to the long posterior epigynal scape. Diagnosis. — Epigynum with very long (0.6-0. 7) posterior scape (Figs. 7-10). Tibial apophysis of male palp more than half the length of the cymbium (Figs. 11-13). Female: CL 1.8, CW 1.3, AL 2.0, AW 1.3. Straw-colored carapace with two dark lon- gitudinal bands; deep foveal groove. Viewed from the top, eye rows slightly recurved. Ra- tio of AME:ALE:PME:PLE is 6:11:11:11. Legs 1423 (Table 2) with dark pigmented bands. Notation of leg spines. Femora: I, 30 THE JOURNAL OF ARACHNOLOGY Figures 29-35. — Carbinea spp. 29- 32, C. wunderlichi new species. 29-30, Epigynum, ventral, dorsal; 31-32, Tibial apophysis. 31, From Lambs Head; 32, From Mt. Williams. 33-35, C. robertsi new species. 33-34, Epigynum, ventral, dorsal; 35, Tibial apophysis. DUO, poll, ROOl; II, DUO, poll, ROOl; HI, DUO, poll, ROOl; IV, DUO, pool, ROOl. Patellae: I, DlOl; II, DlOl; III, DlOl; IV, 001. Tibiae: I, V020; H, V020; HI,D001, PlOl; Vlll, ROll; IV, DOOl, Pill, Vlll, RlOl. Metatarsi spined with whorl 4-5 dis- tally. Epigynum with semi-divided fossa; long scape, short insemination ducts to sper- mathecae. Spinnerets: ALS with two major ampullate spigots, the anterior larger, and about 25 piriform spigots; PMS with a large anterior spigot(minor ampullate) and about 10 other spigots, two of which (cylindricals) have thicker shafts than the rest (aciniforms). PLS with about 25 (aciniform) spigots. Length 3. 5-4.0. Females collected from Mossman Bluff (Fig. 10), Pauls Luck and Mt. Demi showed a longer, more attenuated scape. Male: CL 1.8, CW 1.3, AL 1.9, AW 1.3. Coloration and eyes like female. Legs 1423 (Table 2). Notation of spines. Femora: I, DUO, POOL ROOl; H, DUO, poll, ROll; HI, DUO, POlLROll; IV, DUO, POOl, ROOL Patellae: I, DOOl; H, 001; HI, 001; IV, 001. Tibiae: I, DlOO, POll, V221, ROll; H, DlOl, POlO, V221, ROll; HI, DlOl, POll, V112, ROll; IV, DOOl, POll, V112, ROll. Metatarsi spined with distal whorl 4-5. Palp: short broad tibia; ratio of length to width is 1:0.86. Tegulum with anterior teg- ular groove. Sperm duct looping over retro- lateral tegulum before entering base of em- bolus. Elaborate embolic apophyses with brushes of plain and “knobbed” setae (Figs. 22-23). RTA more than half as long as cym- bium with short blunt retro ventral flange. Spinnerets: ALS with one major ampullate spigot and a nubbin. PMS with one large anterior spigot (minor ampullate) and about 10 aciniform spigots; PLS with about 17 aciniform. spigots. Length 3.5-4. 1. Distribution. — Carbinea longiscapa has been collected only on the Carbine Tableland (Fig. 42). Carbinea breviscapa new species (Figs. 14-21, 24-28, 41-43; Table 4) Types. — Australia: North Queensland. Holotype $, Stewart Ck., 4 km NNE Mt. Spurgeon, Carbine Tableland, Camp 1, 16°24^S, 145°13'E, 1250-1300 m, PF, 15-20 October 1991, GBM, DC, LR (QM S35235).Paratypes: d, 7 km N Mt. Spurgeon, DAVIES— A NEW AUSTRALIAN AMAUROBIOID 31 Figures 36-41. — Carbinea spp. 36- 37, C. wunderlichi new species from Mt. Williams. 36, Male palp, prolateral; 37, Conductor, base of embolus, embolic apophyses. 38-40, C. robertsi new species, female spinnerets. 38, ALS; 39, PMS; 40, PLS. 41, C. breviscapa new species, male spinnerets and colulus. 16°22', 145°13', 1200--1250 m, PF, 17-19 Oc- tober 1991, GBM, DC, LR (S35236); $, same data (S35237); $, same data as holotype (S35251); S , Whypalla State Forest, Windsor Tableland, 16°16', 145°02', 1060 m, PF, sum- mer 1992-3, Scott Burnett (S33163); 2?, Windsor Tableland, 1.2 km past barracks, 16°15', 145°02', 1060 m, stick brushing, 24 November 1997, GBM (S39201); d, Mt. Lewis Rd., 11 km from Hwy. 16°35', 145°17', 1000 m, PF, 18 December 1989-13 January 1990, GBM, GT, ANZSES (S35239); $,<3, Black Mtn., 17 km ESE Julatten, 16°39', 145°29', 1000 m, sieved litter and moss, 29 April 1982, GBM, DY, DC (S35240);$, 800- 1000 m, PY, 29-30 April 1982 (S35241); ?, Mt. Formartine South, 10 km N Kuranda 16°43\ 145°43', 700 m, PF, 23-24 November 1990, GBM, GT (S35242); c3, sieved litter (S35243). 32 THE JOURNAL OF ARACHNOLOGY Figure 42. — Map showing distribution of Car- binea species. Etymology. — The specific epithet is from the Latin “brevis”, short and “scapus” stem referring to the very short epigynal scape. Diagnosis. — Epigynum with short scape (0.3); S tibial apophysis about a quarter the length of the cymbium. Both characters dis- tinguish this species from C. longiscapa. Female: CL 1.8, CW 1.3, AL 2.1, AW 1.6. Coloration and pattern are similar to C. lon- giscapa. Ratio of AME: ALE: PME: PLE is 6:11:11:11. Legs, I, 7.2; II, 5.7; III, 5.3; IV, 7.1. Notation of spines. Femora: I, DllO, pool, ROOl; II, DllO, POOl, ROOl; III, DllO, pool; IV, DllO, pool, ROOl. Patellae: III, DOOl; IV, DOOl. Tibiae: I, V020; II, POOl, V020; III, DlOl, poll, VI 11, ROll; IV, DlOl, poll, VI 11, ROIL Metatarsi spined with whorl 4-5 distally. Epigynum (Figs. 14, 15, 26). Spinnerets: arrangement of spigots similar to $ C. longiscapa. Length 3. 3-3. 9. Male: (from same locality as $, lacking legs II, IV). CL 1.7, CW 1.4, AL 1.8, AW 1.3. Coloration, pattern and eyes are similar to 9. Legs (Black Mt. specimen) I, 8.1; II, 6.3; III, 5.7; IV, 7.7. Notation of leg spines. Femora: I, DllO(l), P00(1)1; II, DllO, POll, ROll; III, Dill, poll, ROll; IV, DIOO. Pa- tellae: I, DOOl; II, DOOl; III, D0(l)01; IV, DOOl. Tibiae: I, DIOO, POll, V221, ROOl; II, DOOl, POOl, V221, ROll; III, DlOl, POll, Vlll, ROll; IV, DlOl, POll, V012, ROll. Metatarsi spined with distal whorl 4- 5. Male palp (Figs. 16-18, 27, 28). Tegulum, conductor, embolus, brush-like apophyses and course of sperm duct similar to C. lon- giscapa. RTA about quarter length of cym- bium. Spinnerets: arrangement of spigots similar to S C. longiscapa. Length 3. 1-3.5. Males and females from Black Mountain (Figs. 18, 19), Mt. Formartine and the Wind- sor Tableland (Fig. 21) are considered to be- long in this species. Distribution, — Most specimens were from the Carbine Tableland (Fig. 42). Carbinea breviscapa was also found south of there in the Black Mountain region (Note: not the more northern Black Mountain near Mt. Spurgeon). Carbinea wunderlichi new species (Figs. 29-32, 36, 37, 42, 43; Table 4) Types. — Australia: North Queensland. Holotype 9 , Lambs Head via Mareeba, 17°02'S, 145°38'E, July 1992, J. Wunderlich (QM S35245). Paratypes: 9, (?, same data as holotype (S35273); S , Lambs Head, 10 km W Edmonton, 1200 m, PF, 10 December 1989-8 January 1990, GBM, JT, HJ (S35244); 9, Emerald Ck, Lamb Ra, 17°06', 145°37', 950 m, sieved litter, 11 October 1982, GBM, DY, GT (S35246); (3, Mt. Wil- liams, 16°559 145°40', 1000 m, sieved litter and moss, 3 December 1993, GBM, HJ (S35234). Etymology.— The specific name is a pa- tronym in honor of Dr. Jorg Wunderlich who collected the holotype. Diagnosis. — The epigynal scape is of me- dium (0.4) length. Tibial apophysis about a third length of cymbium (cf. C. longiscapa) with a posterior heel (cf. C. breviscapa). Female: CL 1.7, CW 1.5, AL 2.0, AW 1.6. Coloration, eye measurements, notation of spines similar to C. longiscapa. Legs, I, 7.5; II, 5.9; III, 5.3; IV, 7.3. Epigynum (Figs. 29, 30). Length 2.8-3.5. Male: CL 1.4, CW 1.1, AL 1.4, AW 0.8. Coloration, eyes, notation of spines similar to C. longiscapa. Legs, I, 6.8; II, 5.2; III, 4.9; IV, 6.5. Male palp (Figs. 31, 32, 36, 37). Te- gulum, conductor and embolus similar to C. DAVIES— A NEW AUSTRALIAN AMAUROBIOID 33 Wandella barbarella Gray 14 20 32 p-[]-|-|- Dictynidae sp. 1 16 24 26 28 29 34 ■f++++H 3 4 8 29 2 5 11 binnaburra Davies 161718 f-HHH 2 0 0 HH-H- rlH+ ^ ' ' ^Malala lubinae Davies 15 1920 [ I I -Desis sp. 8 13 29 HH-H 0 1 —Amphinecta milina Forster & Wilton 9 ~\~Storenosoma terranea Davies 19 25 33 4- 3 26 30 +++ 3 35 ++. Tasmarubrius milvinus (Simon) 0 1 1 1621 -^^Quesmusia aquilonia Davies '-^-^^Jalkaraburraalta Davies 2 0 0 -W-Paramatachia decorata Dalmas 2 1 2 13 29 35 r-^-f-^^Amaurobius fenestralis (Stroem) 0 111 1631 4-H Badumna longinqua (Koch) 0 1 \-^-\-Manjala plana Davies ++ 1 4 2931 4HH 0 10 22 -\^Procambridgea sp. r-ll- ' 0 nr 11162136 spinipes Davies 112 1 1 18 22 29 -^t\^t\~Stiphidion facetum (Simon) 0 2 0 2 34 -^Kababina alta Davies 3 12 i+ I forward change I forward change with homoplasy I reversal with homoplasy I] reversal 3 7 1021 ++++ / Kababininae / Carbine a breviscapa sp. nov. sp. nov. 16 17 1823 27 4+M+l r Carbinea wunderlichi ++ Carbinea gen. nov. Carbinea longiscapa sp. nov. Carbinea robertsi sp. nov. Figure 43. — The single most parsimonious tree showing the cladistic relationships of the Amaurobioi- dea. longiscapa. RTA about one-third length of cymbium. Length 2. 8-3. 5, Distribution. — From the Lamb Range, SW Cairns (Fig. 42). Carbinea robertsi new species (Figs. 33-35, 38-40, 42, 43; Table 4) Types. — Australia: North Queensland. Holotype ?, Mt. Lewis, [16°29'S, 145°15'E], 7 November 1975, V.E. Davies (QM S35233). Paratypes: 2$, d, same data as ho- lotype (S35272); 9, Mt. Lewis Rd. (Hut) 16°3r, 146°16', 1200 m, PY trees, 14 July 1996, GBM (39198). Etymology. — The specific epithet is a pa- tronym in honor of Lewis Roberts, a noted collector from Shiptons Flat, North Queens- land. Diagnosis. — Epigynal scape of medium 34 THE JOURNAL OF ARACHNOLOGY Table 3. — Characters and character states, with states in parentheses. Multistate (*) character treat- ed as unordered. No. Character 1 AME: as large or larger than ALE (0); smaller than ALE (1) 2 Retromarginal CH teeth: 2+ (0); 2 (1); 1 (2); 0 (3) 3 Promarginal CH teeth: 3+ (0); 3 (1); 2 (2); 0 (3) 4 Long prolateral seta at base of fang: absent (0) ; present (1) 5 Large frontal CH seta: absent (0); present (1) 6 CH lamina: absent (0); present (1) 7 Foveal area highest: absent (0); present (1) 8 $ leg I: shorter than leg IV (0); equal to or longer than leg IV (1) 9 Stridulatory ridges on (3 coxa I: absent (0); present (1) 10 Trochanteral notch: absent (0); present (1) 11 Large ventral spines on tibia and MT I, II: absent (0); present (1) 12 Feathery hairs: absent (0); present (1) 13 MT preening combs: absent (0); present (1) 14 MT TRICH: 2+ (0); 1 (1) 15 Tarsal TRICH: 0 (0); 2+ (1); double row (2) 16 CR: 2 spinning fields (0); 1 spinning field (1); absent (2) 17 CR spigots: absent (0); longitudinally ribbed (1); annulate (2) 18 Calamistrum: absent (0); proximal (1); proximo-medial (2) 19 MAP 9 ALS: 2 (0); 1 and nubbin (1); 1 (2) 20 Position of MAP 9 ALS: mesal (0); anterior (1) 21* PCR 9 PMS: one shaft per base (0); more than one shaft (1); absent (2) 22 Medial EPIG fossa: absent (0); present (1) 23 Posterior EPIG scape: absent (0); short (1); medium (2); long (3) 24 Insemination duct: absent (0); present (1) 25 EPIG lateral projections: absent (0); present (1) 26 E direction: straight (0); clockwise (1); an- ticlockwise (2) 27 Proximal E APOPH: absent (0); present (1) 28 Parembolic process: present (0); absent (1) 29* Conductor: absent (0); rounded (1); large T- shaped (2); s- shaped-falciform (3) 30 Secondary conductor: absent (0); present (1) 31 Median APOPH: absent (0); present (1) 32 Orientation of CB to bulb: dorsal (0); mesal (1) 33 Paracymbium: absent (0); present (1) Table 3. — Continued. No. Character 34 RTA to CB length: absent (0); quarter or less (1); third (2); half (3); more than half (4) 35 Dorsal branch of RTA: absent (0); present (1) 36 Palpal patellal APOPH: absent (0); present (1) (0.4) length (cf. C. longiscapa). Tibial apoph- ysis long (cf, C. breviscapa and C. wunder- lichi). Female: CL 1.9, CW 1.4, AL 2.1, AW 1.4. Coloration, eye measurements, notation of spines similar to C. longiscapa. Legs, I, 7.5; 11, 5.8; III, 5.4; IV, 7.3. Epigynum (Figs. 33, 34). Spinnerets (Figs. 38-40): ALS with two major ampullate spigots and about 25 piri- form spigots. PMS with a large anterior spig- ot (minor ampullate) and about 10 smaller spigots, 2-3 with thicker shafts from cylin- drical glands, the others from aciniform glands. PLS with about 25 aciniform spigots. Length 3. 6-4.0. Male: CL 1.7, CW 1.3, AL 1.8, AW 1.2. Legs, I, 7.9; II, 6.1; III, 5.6; IV, 7.5. Tegulum, conductor, embolus similar to C. longiscapa. RTA long, half length of cymbium (Fig. 35). Distribution. — Carbinea robertsi was found at one site on Mt. Lewis (Fig. 42). RELATIONSHIPS OF CARBINEA A cladistic analysis examined 36 charac- ters (Table 3) for relationships of Carbinea spp. and 18 other taxa (names and authors given on cladogram. Fig. 43). Voucher spec- imens of the taxa are deposited in the QM. Outgroup comparison was with the Austra- lian spiders Wandella barbarella, a filistatid and an undescribed dictynid. A data matrix (Table 4) was assembled in MacClade 3.01 (Maddison & Maddison 1992). Unknown characters are represented by “?”, inappli- cable characters by The data were ana- lyzed in PAUP version 3.1.1 (Swofford 1993) and replicated in Hennig. A heuristic search of the data with 10 random- addition sequenc- es and TBR branch swapping generated one most parsimonious tree (Fig. 43); length 103, Cl = 0.52, Cl excluding uninformative char- acters = 0.48, RI = 0.67, RC = 0.35. Char- DAVIES— A NEW AUSTRALIAN AMAUROBIOID 35 Table 4. — Data matrix 10 20 30 Wandeila 033001010 0000100212 00--000000 0000000 Dictynidae A 120000010 0000001112 1000101012 0010100 Badumna 010000010 0000110110 0110101013 0100100 Paramatachia 010000010 0000111120 0000101013 0000101 Desis 110000010 0000122002 1-00101013 0100100 Quemusia 111000010 0000110110 0200102013 1000100 Jalkaraburra 111000010 0000112000 0-00102013 1000100 Amphinecta 110000000 000111200? 0-00101011 0100100 Amaurobius 000000010 0001110110 0000101011 0100110 Storenosoma 112000001 0001112001 0-00101011 0100110 Tasmarubrius 112000000 0001112000 0-00111011 0101110 Procambridgea 100100010 1000110110 0100101011 0000100 Stiphidion 011100010 0010110120 0100101012 0000100 Midgee 102110000 0100112001 0-00101011 0000100 Dardurus 100100010 0100111110 0210101011 0000101 Manjala 110110010 0100110111 0710101013 0100100 Malala 100010010 0100112001 0-00101013 0000100 Kababina 112100110 1010110110 0010101011 0000400 Carbinea longiscapa 112100110 1010112000 0-13101111 0000400 C. breviscapa 112100110 1010112000 0-11101111 0000100 C. wunderlichi 112100110 1010112000 0-12101111 0000200 C. robertsi 112100110 1010112000 0-12101111 0000300 acters were mapped in CLADOS version 1.2 (Nixon 1992) with DELTRAN optimization. Conclusions,— Wandeiia and Dictynidae A appear as distinct from the ingroup which is regarded as the superfamily Amaurobioi- dea. This is composed of two clades, the first of which includes Desis (Desidae), AmphU necta (Amphinectidae) and Tasmarubrius (Davies 1998). The second clade includes Amaurobius (Amaurobiidae), Stiphidion (Sti^ phidiidae) and the metaltellines Quemusia and Jalkaraburra. Kababina and Carbinea are the only genera in this clade that form a well-supported monophyletic group Kababi- ninae, which was previously (Davies 1995) placed in the Amphinectidae and from which it is now withdrawn. The group appears clos- est to Stiphidion; however, separation of the Kababininae would render the base of the clade paraphyletic. Until further description and analysis of the basal members of this clade are extended, the family placements of the clade will remain problematic. ACKNOWLEDGMENTS I am indebted to my colleague Dr. G.B. Monteith for his and co-workers, collections from the Wet Tropics region of north Queens- land. Since 1993 the field trips have been supported by the Wet Tropics Management Authority which also supports Mrs. Kylie Stumkat, SEM technician. I thank the Coun- cil of the Australian Biological Research Studies for funding rainforest surveys during which some of this material was collected and for the financial support of illustrator, Mrs. Christine Lambkin, who also set up the program for phylogenetic analysis resulting in the cladogram. I am grateful for the sup- port of other members of the Queensland Museum, particularly Mrs. Audra Topping and Ms. Jennifer Ingram for their help in preparation of this paper. LITERATURE CITED Davies, V.T. 1995. A new spider genus (Araneae: Amaurobioidea; Amphinectidae) from the wet tropics of Australia. Mem. Queensland Mus., 38(2):463-469. Davies, V.T. 1998. A redescription and renaming of the Tasmanian spider Amphinecta milvina (Simon, 1903) with descriptions of four new species (Araneae: Amaurobioidea: Amaurobi- idae). Pp. 67-82. In Proceedings of the 17"^- European Colloquium of Arachnology, Edin- burgh 1977. Maddison, W.P. & D.R. Maddison. 1992. Mac- Clade: analysis of phylogeny and character co- 36 THE JOURNAL OF ARACHNOLOGY evolution. Version 3 documentation. Sinauer: Sunderland. Monteith, G.B. & V.T. Davies. 1984. Preliminary account of a survey of arthropods (insects and spiders) along an altitudinal rainforest transect in tropical Queensland. Pp. 402-412. In Aus- tralian National Rainforest Study. (G.L. Werren & A.P. Kershaw, eds.). World Wildlife Fund (Australia). Project 44 Report. Vol. I. Reprinted 1991, Pp. 345-362. In The Rainforest Legacy: Australian National Rainforest Study, Vol 2: Australian Government Publishing Service. Canberra. Nixon, K.C. 1992. Clados, Version 1.2. L.H. Bai- ley Hortorium, Cornell University, Ithaca. Platnick, N.I. & M.U. Shadab. 1975. A revision of the spider genus Gnaphosa (Araneae: Gna- phosidae) in America. Bull. American Mus. Nat. Hist., 155:1-16. Swofford, D.L. 1993. PAUP. Phylogenetic Anal- ysis Using Parsimony. Version 3.1. Illinois Nat. Hist. Surv. and Smithsonian Institution: Cham- paign and Washington. Manuscript received 29 April 1998, revised 4 De- cember 1998. 1999. The Journal of Arachnology 27:37-43 SPIDERS OF THE GENUS HEPTM^HELA (ARANEAE, LIPHISTIIDAE) FROM VIETNAM, WITH NOTES ON THEIR NATURAL HISTORY Hirotsugu Ono: Department of Zoology, National Science Museum, 3^23- 1 Hyakunin=cho, Shinjuku-ku, Tokyo, 169“0073 Japan ABSTRACT, Spiders of the family LipMstiidae collected from northern Vietnam are taxonomically studied. Two new species of the genus Heptathela are described under the names, H. abca (from Yen Bai) and H. cucphuongensis (from Cue Phuong National Park). Some natural history and zoogeographic notes of the new species are given. About 90 years ago, a liphistiid spider was recorded from Kha-le in the area of the river Song Luc Nam northeast of Hanoi (Simon 1908). This spider was erroneously identified as Liphistius birmanicus Thorell 1897, origi- nally described from Burma. Because this re- cord was based on a misidentification, Bris- towe (1933) gave a new name Liphistius tonkinemis for the same spider. Although the genus Heptathela was already known at that time (Kishida 1923), the Vietnamese species was misplaced in Liphistius for a long time. Haupt (1983) finally redescribed the species and transferred it to Heptathela. Heptathela tonkinensis was the first heptatheline record- ed. In 1995 and 1997, I had opportunities to participate in entomological expeditions to northern Vietnam, made by the National Sci- ence Museum, Tokyo. Although it was hard to find preserved forests in the country, re- searchers on our expeditions collected many spider specimens, including those of Heptath- ela. Heptathela tomokunii, the second species of the genus from Vietnam, was described from Mt. Tam Dao, about 60 km northwest of Hanoi, based on this material (Ono 1997a). The present paper deals with the results of a taxonomic study of liphistiids obtained dur- ing the second Vietnam expedition in 1997. Two further new species of the genus Hep- tathela are described from Yen Bai and Cue Phuong National Park. Some biological data of these spiders are noted. METHODS Between 26 September-25 October 1997, liphistiid spiders were collected at Yen Bai (elevation 120 m), and in the National Park of Cue Phuong (350 m) in northern Vietnam (Fig. 1). Some biological observations (for ex- ample, the shape of egg sac and the diameter of trapdoor) were made in the field. The spi- ders were kept in 75% alcohol and taxonom- ically studied in laboratory of the museum at Tokyo. After morphological observations, two new species were recognized. The type specimens of the new species are deposited in the collection of the Department of Zoology, National Science Museum, Tokyo (NSMT). The abbreviations herein used are as follows: ALE, anterior lateral eye; AME, an- terior median eye; PLE, posterior lateral eye; PME, posterior median eye. Morphology of genital organs chiefly follows Haupt (1979). DESCRIPTIONS OF NEW SPECIES Heptathela abca new species (Figs. 2-4, 14) Diagii,osis.^This new species seems to be related to Heptathela tomokunii Ono 1997, de- scribed from Mt. Tam Dao, Vinh Phu Prov- ince, Vietnam, having the same construction of female genitalia. However, the new species can be distinguished from the latter by the shape of spermathecae (receptac alula semin- is): main, lateral bursae of Heptathela abca are reniform and are set on their bases, while those of H. tomokunii are oval and close to the lamellar interior part of the genitalia; the tubular stems of median bursae of H. abca are much longer than those of H. tomokunii (cf Fig. 3 of this paper and fig. 6 in Ono 1997a). Etymology .^The specific epithet is an ar- bitrary combination of letters. 37 38 THE JOURNAL OF ARACHNOLOGY Figure 1. — Records of the spiders of the genus Heptathela in Vietnam. 1. Song Luc Nam, Ha Bac Province, Heptathela tonkinensis (Bristowe 1933); 2. Tam Dao, Vinh Phu Province, Heptathela to- mokunii Ono 1997; 3. Yen Bai, Yen Bai Province, Heptathela abca new species; 4. Cue Phuong, Ninh Binh Province, Heptathela cuephuongensis new species. [Small circle = Hanoi; upper line of the frame, 24°N, bottom, 19°N, left, 102°E, right, 108°E; scale = 100 km] Type series. — Female holotype, VIET- NAM, Yen Bai Province, Yen Bai (elevation 120 m), 13 October 1997, H. Ono leg. (NSMT-Ar 3830); paratypes: 29, same data as for holotype (NSMT-Ar 3831-3832), 3? and 2 juveniles, same locality and collector as for holotype, 14 October 1997 (NSMT-Ar 3833-3837). Description. — Female (male specimen not available). Measurements based on holotype: body length 17.5 mm; prosoma length 8.4 mm, width 7.1 mm; opisthosoma length 9.1 mm, width 6.8 mm; lengths of palp and legs [total length (femur + patella + tibia + meta- tarsus + tarsus)]: palp 15.3 mm (5.3 + 2.7 + 3.4 + - + 3.9); leg 1 17.8 mm (5.8 + 3.0 + 3.5 + 3.5 + 2.0); II 18.2 mm (5.8 + 2.9 + 3.4 + 3.9 + 2.2); III 19.9 mm (5.7 + 3.3 + 3.4 + 4.7 -f 2.8); IV 29.0 mm (8.0 + 3.7 + 5.1 + 8.2 + 4.0). Variation of body length: 12.1-20.2 mm. Head high; ocular tubercle wider than long, ALE > PLE > PME > AME (9. 6:8. 3:4. 6:1 in ratio), AME small, clypeus wider than ALE-ALE, median ocular area trapezoidal, wider than long. Chelicera with 11 teeth (3 large and 8 small) on promargin of fang furrow. Leg formula IV, III, II, I; su- perior claws of tarsi each with 2 teeth; claw of palp without tooth. Opisthosoma ovate, longer than wide; posterior median spinnerets reduced, completely fused, with setae (Fig. 4). Two pairs of spermathecae present (Fig. 2-3); main, lateral bursae set on thick bases, reni- form, with many middle-sized, granulate tu- bercles; median ones small, on long tubular stems. Coloration: Prosoma brown, cephalic part not darker, ocular tubercle black; chelicerae brown, basally lighter, fang and fang furrow reddish-brown, sternum and coxae of legs and palps light reddish-brown, other segments of legs and palps brown. Opisthosoma grayish- brown, dorsal sclerites blackish-brown, ven- tral sclerites and spinnerets yellowish-brown. Remarks. — A female specimen (NSMT-Ar 3839) collected at a village about 15 km northwest of Yen Bai on 6 October 1997 by myself tentatively identified as Heptathela abca. However, the shape of female genitalia of the spider (Figs. 5-6) is slightly different from that of holotype. Heptathela cuephuongensis new species (Figs. 7-9, 17) Diagnosis. — This new species is close to Heptathela hunanensis Song & Haupt 1984, described from Qianyang County, Hunan Province of China. In both the species, median spermathecae are fused and form a large bur- sa. However, the new species is distinguish- able from the Chinese spider by the shape of lateral bursae without bases {cf. Fig. 8 of this paper and fig. 3e in Song & Haupt 1984). Etymology. — The specific epithet is de- rived from the type locality. Type series. — Female holotype, VIET- NAM, Ninh Binh Province, Gia Vien, Cue Phuong (elevation 350 m), 30 September 1997, H. Ono leg. (NSMT-Ar 3822); para- types: 2$ and 7 juveniles, same data as for holotype (NSMT-Ar 3823-3827). Description. — Female (male specimen not available). Measurements based on holotype: body length 17.5 mm; prosoma length 7.8 mm, width 6.1 mm; opisthosoma length 8.1 mm, width 5.4 mm; lengths of palp and legs [total length (femur + patella + tibia + meta- tarsus + tarsus)]: palp 12.2 mm (4.4 -f 2.3 + 2.5 + - + 3.0); leg I 15.5 mm (5.3 + 2.5 T 3.1 -f 2.9 + 1.7); II 15.7 mm (5.2 + 2.7 T 2.8 + 3.3 + 1.7); III 16.6 mm (5.1 + 2.7 + O^O—HEPTATHELA FROM VIETNAM 39 Figures 2-9. — 2-4, Heptathela abca new species, female holotype, NSMT-Ar 3830; 5-6. Heptathela sp. (?//. abca) from 15 km NW of Yen Bai; 7-9. Heptathela cucphuongensis new species, female holotype, NSMT-Ar 3822. 2, 5, 7. Genital area, ventral view; 3, 6, 8. Spermathecae, dorsal view; 4, 9. Posterior median spinnerets, ventral view. [Scales = 0.25 mm] 2.9 + 3.8 + 2.1); IV 23.4 mm (6.8 + 3.2 + 4.2 + 6.4 + 2.8). Variation of body length: females 15.1=17.5 mm. Head high; ocular tu- bercle slightly longer than wide, ALE > PLE > PME > AME (8 : 7.6 : 4.3 : 1 in ratio), clypeus wider than ALE- ALE, median ocular area trapezoidal, slightly wider than long. Chelicera with 12 (right) or 13 (left) teeth (3 large and 9 or 10 small) on promargin of fang furrow. Leg formula IV, III, II, I; superior claws of tarsi I with 3 teeth, II-IV each with 2 teeth; claw of palp without tooth. Opistho- soma ovate, longer than wide; posterior me- dian spinnerets not fused (Fig. 9). Three sper- mathecae present (Figs. 7-8); lateral bursae without bases, globular, with many small granulate tubercles; median one unpaired, oval, granulate, without stem. Coloration: Prosoma blackish-brown, ce- phalic part darker, ocular tubercle black; chehc- erae blackish-brown, basally light yellowish- brown, fang and fang furrow reddish-brown, sternum and coxae of legs and palps grayish- brown, other segments of legs and palps dark gray, femora much darker. Opisthosoma dark gray, dorsal sclerites blackish-brown, ventral sclerites and spinnerets yellowish-brown. Color in life bluish-black. NATURAL HISTORY NOTES As an agricultural country in Asia, Vietnam has developed at the expense of deforestation for centuries. Sub-tropical and tropical low- lands are totally cultivated mainly for rice pro- duction. On the other hand, the people of var- ious tribes farm with primitive methods on temperate highlands, and cut trees for fire- wood. Thus, primary forests, the habitat of li- phistiid spiders, are only found scattered on the mountainous hinterland. Tam Dao (1230 m elevation at peak) is one of the typical areas preserved by the country. Spiders of Heptath- ela tomokunii live there (Ono 1997a). In Cue Phuong National Park, evergreen broad-leaved forests and occasional damp bushes are preserved in nature (Fig. 10). Spi- 40 THE JOURNAL OF ARACHNOLOGY Figures 10-19. — 10. Habitat of Heptathela cucphuongensis new species at Cue Phuong; 11. Habitat of Heptathela abca new species at Yen Bai; 12. Holotype female of Heptathela abca new species; 13-14. A retreat with grass-blades of Heptathela abca new species; 15-16. Egg sacs of Heptathela abca new species; 17. Holotype female of Heptathela cucphuongensis; 18-19. Egg sac of Heptathela cucphuongensis new species. [Scales = 10 mm] ONO—HEPTATHELA FROM VIETNAM 41 Figure 20. — Distribution of the species of the subfamily Heptathelinae in East Asia: The symbols correspond to the groups A-E in the text: □ = Group A, ■ = Group B, • = Group C, A = Group D, and ☆ = Group E. The open circles (o) indicate species whose female is unknown, a-e are diagrams of female genitalia. [Scale = 1000 km] ders of Heptathela cucphuongensis were found along roadsides at the forest edge. They built retreats in the soil about 15-25 cm deep, as is typical of the genus Heptathela. The trapdoor of the holotype female was the larg- est at 35 mm wide and 27 mm long. Globular egg sacs made with soil and 24-27 mm in diameter (Fig. 18) were found at the bottom of the tubular retreats, with 120 spiderlings closely packed in the sac (Fig. 19). As an unusual case, Heptathela abca was found in a village near the city of Yen Bai. The environmental difference between that village and other cultivated areas on lowlands, in which no liphistiids were found, is the ex- istence of trees left around the houses and rice fields (Fig. 11). The earth was kept moist. The openings of some retreats were decorated with fragments of grass (Figs. 13-14). The trap- doors of the spiders were measured: 43 X 33, 33 X 29, 30 X 26, 28 X 25, 24 X 21 and 17 X 15 mm. Egg sacs were semiglobular and about 35 mm in diameter (Figs. 15-16). Re- spectively, 201 and 221 spiderlings emerged from the sacs in November. ZOOGEOGRAPHIC NOTES Trapdoor spiders of the family Liphistiidae (Araneae, Mesothelae) are composed of two recent subfamilies, Liphistiinae and Heptath- elinae, both distributed in East Asia. More than 40 species of the single genus Liphistius Schiodte 1849 were described under the for- mer subfamily from Myanmar (Burma), Thai- land, the Malay Peninsula and Sumatra (new- est informations: Schwendinger 1996; Platnick, Schwendinger & Steiner 1997), while 29 species of two genera, Heptathela Kishida 1923 and Ryuthela Haupt 1983 were known in the latter subfamily in Japan, China and Vietnam (Haupt 1983; Song & Haupt 1984; Ono 1996, 1997a & b, 1998 and this paper). Based on characteristics of the female gen- italia (males are unknown in many species), the heptatheline species are classified into five groups as follows, which are grouped in an allopatric arrangement (Fig. 20). Their phy- logenetic relationships are not considered in this paper. Vietnamese species belong to two different groups. Group A.,— Ryuthela nishihirai Haupt 1979, R. ishigakiensis Haupt 1983, R. owadai Ono 1997, R. sasakii Ono 1997, R. secundaria Ono 1997, and R. tanikawai Ono 1997. A pair of monolobal spermathecae present, both the spermathecae close to each other, or fused 42 THE JOURNAL OF ARACHNOLOGY with one large opening (Fig. 20a). Distribu- tion: Japan (southern part of the Ryukyu Is- lands). [Although the genus Ryuthela was syn- onymized with Heptathela by Raven (1985), I follow the contrary treatment by Haupt (1990).] Group B. — Heptathela kimurai (Kishida 1920), H. amamiensis Haupt 1983, H. higoen- sis Haupt 1983, H. kanenoi Ono 1996, H. kik- uyai Ono 1998, H. nishikawai Ono 1998, H. yaginumai Ono 1998, H. yakushimaensis Ono 1998, H. yanbaruensis Haupt 1983. A pair of spermathecae present, spermathecae bilobal with secondary process laterally (Fig. 20b). Distribution: Japan (Kyushu and the northern part of the Ryukyu Islands). Group C. — Heptathela sinensis Bishop & Crosby 1932, H. bristowei Gertsch 1967, H. jianganensis Chen et al. 1988, H. schensiensis (Schenkel 1953), H. heyangensis (Zhu & Wang 1984), H. yunnanensis Song & Haupt 1984, H. tomokunii Ono 1997 and H. abca new species. Two pair of spermathecae pre- sent, the lateral bursae larger and usually on thick bases, the median ones small and on tu- bular stems (Fig. 20c). Distribution: China (from Hebei to Yunnan) and Vietnam. Group D. — Heptathela hangzhouensis Chen, Zhang & Zhu 1981 and H. cipingensis (Wang 1989). Two pair of spermathecae pre- sent, the main bursae situated in more median position, the median ones moved posteriorly and situated at the base of main bursae (Fig. 20d). Distribution: China (from Zhejiang to Hunan). Group E. — Heptathela hunanensis Song & Haupt 1984 and H. cucphuongensis Ono new species. Three spermathecae present, two lat- eral bursae and one median one in same size (Fig. 20e). Distribution: China (Hunan) and Vietnam. Female unknown. — Heptathela tonkinen- sis (Bristowe 1933) and H. hongkong Song & Wu 1997. ACKNOWLEDGMENTS I wish to express my sincere thanks to all the staff of the Department of Entomology, Ha-noi Agricultural University, especially to Prof. Ha Quang Hung, Dr. Nguyen Due Khiem and Mr. Tran Dinh Chien, to Mr. Dao Van Khuong, Director of the Cue Phuong Na- tional Park, and to Dr. Mamoru Owada and Mr. Wataru Abe, National Science Museum, Tokyo, for their kind aid in my field research. The manuscript of this paper was improved by comments from Drs. Brent D. Opell, Jer- emy Zujko-Miller, James Berry, and anony- mous reviewers. I am deeply indebted to the above persons. This study was supported by the Grants-in-aid No. 06041116 and 09041167 for Field Research of the Monbu- sho International Scientific Research Program, and No. 07640944 and 10640688 for Scien- tific Research from the Ministry of Education, Science, Sports and Culture, Japan. LITERATURE CITED Bristowe, W.S. 1933. The liphistiid spiders (chap- ters I-VIII, X). In The liphistiid spiders, with an appendix on their internal anatomy. (W.S. Bris- towe & J. Millot, ed.). Proc. Zool. Soc. London, 1932:1016-1045, 1055-1057, pis. I-IL Haupt, J. 1979. Lebensweise und Sexualverhalten der mesothelen Spinne Heptathela nishihirai n. sp. (Araneae, Liphistiidae). Zool. Anz., Jena, 202:348-374. Haupt, J. 1983. Vergleichende Morphologic der Genitalorgane und Phylogenie der liphistiomor- phen Webspinnen (Araneae: Mesothelae) I. Re- vision der bisher bekannten Arten. Z. Zool. Sys. Evol.-Forsch., 21:275-293. Haupt, J. 1990. Comparative morphology and phy- logeny of liphistiomorph spiders (Araneida: Me- sothelae). 111. Provisional diagram of relation- ships in Heptathelidae. Pp. 134-140. In Comptes Rendus du XIP Colloque Europeen d’Arachnologie. (M.-L. Celerier, J. Heurtault & C. Rollard, eds.) Societe Europeenne d’Arachnologie, Paris. Kishida, K. 1923. Heptathela, a new genus of li- phistiid spider. Annot. Zool. Japonenses, 10:235- 242. Ono, H. 1996. Two new species of the families Liphistiidae and Thomisidae (Araneae) from the Ryukyu Islands, Southwest Japan. Acta Arach- noL, 45:157-162. Ono, H. 1997a. A new species of the genus Hep- tathela (Araneae: Liphistiidae) from Vietnam. Acta Arachnol., 46:20-30. Ono, H. 1997b. New species of the genera Ryu- thela and Tmarus (Araneae, Liphistiidae and Thomisidae) from the Ryukyu Islands, South- west Japan. Bull. Natn. Sci. Mus., Tokyo, (A), 23:149-163. Ono, H. 1998. Spiders of the genus Heptathela (Araneae, Liphistiidae) from Kyushu, Japan. Mem. Natn. Sci. Mus., Tokyo, (30): 13-27. Ono, H. & Y. Nishikawa. 1989. Taxonomic revi- sion of the heptathelid spider (Araneae, Mesoth- elae) from Amami-d shima Island, the Ryukyus. Mem. Natn. Sci. Mus., Tokyo, (22): 1 19-125. OnO—HEPTATHELA FROM VIETNAM 43 Platnick, N.I., R Schwendinger & H. Steiner. 1997. Three new species of the genus Liphistius (Ara- neae, Mesothelae) from Malaysia. American Mus. Novitates, (3209): 1-13. Raven, RJ. 1985. The spider infraorder Mygalo- morphae (Araneae): cladistics and systematics. Bull. Amercan Mus. Nat. Hist., 182:1-180. Schwendinger, P.J. 1996. New Liphistius species (Araneae, Mesothelae) from western and eastern Thailand. Zool. Script., 25:123-141. Simon, E. 1908. Etude sur les arachnides du Ton- kin (F® partie). Bull. Sci. France Belgique, 42: 69-147. Song, D.-X. & J. Haupt. 1984. Comparative mor- phology and phylogeny of liphistiomorph spiders (Araneae: Mesothelae). 2. Revision of new Chi- nese heptathelid species. Verb. Naturwiss. Ver. Hamburg, (NF), 27:443-451. Manuscript received 1 May 1998, revised 10 Jan- uary 1999. 1999. The Journal of Arachnology 27:44-52 ON THE PHYLOGENETIC RELATIONSHIPS OF SISICOTTUS HIBERNUS (ARANEAE, LINYPHIIDAE, ERIGONINAE) Jeremy Zujko-Miller: Department of Biological Sciences, The George Washington University, Washington, D.C. 20052, and Department of Entomology, National Museum of Natural History, NHB-105, Smithsonian Institution, Washington, D.C. 20560 ABSTRACT. Carorita hiberna NEW COMBINATION, a species with many putative autapomorphies known from one sex and few specimens, is transferred from Sisicottus. This transfer is based on a modified version of a cladistic analysis of erigonine relationships by G. Hormiga which incorporated 43 spider taxa scored for 73 characters. The modified analysis features 46 taxa scored for 74 characters. The resulting cladogram placed C. hiberna sister to C limnaea, the type species of Carorita. It is concluded that C hiberna is better placed in Carorita than in either a new monotypic genus or in Sisicottus. Carorita hiberna is redescribed and the monophyly of Carorita as currently circumscribed is discussed. Carorita hiberna (Barrows 1945) NEW COMBINATION (Linyphiidae, Erigoninae) was inexplicably described as a member of the genus Sisicottus Bishop & Crosby 1938. Ca- rorita hiberna is a very unusual erigonine spe- cies that shares none of the synapomorphies that unite Sisicottus (Zujko-Miller 1999). Ca- rorita hiberna is known only from three male specimens from the Great Smoky Mountains National Park, North Carolina, USA. After re- vising Sisicottus (Zujko-Miller 1999), I was left with three alternatives as to the fate of C hiberna: I could keep it in Sisicottus, which would leave Sisicottus polyphyletic; I could erect a new monotypic genus for it; or I could transfer it to the genus which contains its clos- est relatives. Carorita hiberna features many apparently apomorphic character states in the form of the male palpus, and it was not ob- vious to me what genus contained its closest relatives. I chose to seek the closest relatives of C. hiberna using phylogenetic methods. By placing C. hiberna in a phylogenetic context, I was able to formulate a testable phylogenetic hypothesis. My cladistic analysis identified Carorita limnaea (Crosby & Bishop 1927), the type species of Carorita Duffey & Merrett 1963, as the sister taxon of C. hiberna. With the transfer of C. hibernus, Carorita currently contains three species. METHODS I cleared specimens in methyl salicylate (Holm 1979) and positioned them for illustra- tion using a temporary slide mount (Codding- ton 1983). I made sketches using a camera lucida fitted to a Leica DMRM compound mi- croscope at 400 X. Further observations were made using a Leica MZ APO dissecting mi- croscope. Museum acronyms for specimen de- positories appear in the acknowledgments. CLADISTIC ANALYSIS The cladistic analysis by Hormiga (in press) is the most rigorous hypothesis of erigonine relationships to date and is the logical starting point for questions of relationships within the Erigoninae. The original analysis incorporates 43 terminal taxa, including 31 erigonine gen- era, scored for 73 characters. My modified version of Hormiga’s analysis incorporates three additional taxa, one new character, and one recoded character. Carorita hiberna, C. limnaea, and Sisicot- tus montanus (Emerton 1882) were added to Hormiga’s (in press) matrix. Carorita limnaea was included because it appears to share some potentially synapomorphic character states with C. hiberna including a looped sperm duct in the tegulum, a tuberculate radical tailpiece, and a suprategulum separated from the tegul- um by a membranous region so that it appears to form a distinct sclerite. Sisicottus was in- cluded to test the implicit phylogenetic hy- pothesis of Barrows (1945) that C. hiberna plus Sisicottus form a monophyletic group. 44 ZUJKO-MILLER— m/co irro HIBERNUS 45 Several character states remain unknown for C. hiberna because females are unknown and males are rare and not available for irre^ versible methods of examination. The male cephalothorax was not examined using scae^ ning electron microscopy to search for cutic- ular pores in the clypeal region (character 50) or to examine details of the stridulatory striae on the chelicerae (character 56). No abdomens were digested for examination of the tracheal system (characters 51, 52). Carorita hiberna was coded as follows: 0001310110 1210110101 1201000101 3????????? 000000000? ??001?0??1 00011?0??1 1??1 Carorita iimnaea was coded as follows: 0001310110 1210110101 1501000101 300=000100 0000000000 0=00120111 0011100101 1??L Carorita Iimnaea was giv- en a unique character state for the shape of the radical tail piece (character 22). In C. Um- naea, the tailpiece extends both dorsally and ventrally from its origin distal to the origin of the embolus. Coding of C. Iimnaea was based on examination of the following specimens: UNITED STATES: Maine: Piscataquis County, 2.3 km ESE of Soubunge Mtn., dense spruce-fir forest, Line I, Stn, 1, T4 Rll, WELS, 1 June 1978, pitfall collection, Id, (D.T Jennings, M.W. Houseweart, USNM); New York: McLean, mud pond, 42°32'N, 76°18'W, 30 May 1921, 8dl7 9, (C.R. Cros- by, AMNH). Sisicottus montanus was coded as in Zujko- Miller (1999). Character 74 was scored with a zero. Coding of 5. montanus was based on examination of the following specimens: UNITED STATES: Massachusetts: Berk- shire County, Mt. Greylock, 3400 feet, decid- uous litter, 15 October 1990, 2dl 9, (R.L. Ed- wards, USNM). Hormiga’s (in press) analysis was modified by the addition of one new character and the recoding of an existing character. Both char- acters pertain to structures that are synapo- morphies of the Linyphiidae (the suprateguL um and the linyphiid radix) so character states for linyphiid taxa other than C Iimnaea, C. hiberna, and S. montanus were determined us- ing Hormiga (1994, in press). Character 74 is the texture of the radical tailpiece which is tuberculate in C. Iimnaea and C. hiberna. In all other taxa with a radical tailpiece, this sclerite is more or less smooth. Taxa without a radical tailpiece were coded as inapplicable for this character. The junction between the tegulum and the suprategulum (character 12) was recoded. Character 12 documents the membranous hinge between the tegulum and the suprategulum in Stemonyphantes Menge 1886 (van Helsdingen 1968; Hormiga 1994). In the context of Hormiga's analysis, this character state is autapomorphic and character 12 is not phylogeeetically informative. I have added a third state to character 12 and it is now coded as follows: Suprategulum: 0 = continuous with tegulum; 1 == articulated; 2 = separate from tegulum (Figs. 7, 9). In Ca- rorita Iimnaea, C. hiberna, Asthenargus pa- ganus (Simon 1884), Gongylieiium vivum (O. Pickard-Cambridge 1875) and Erigone psy- chrophila Thorell 1871, there is a membra- nous division between the sclerotized parts of the tegulum and the suprategulum so that the suprategulum appears to be a distinct sclerite rather than a more heavily sclerotized distal portion of the tegulum. This is the new char- acter state coded as 2. In most other liny- phiids, the tegulum and the suprategulum are joined by a region of continuous sclerotization and only part of the junction between the te- gulum and the suprategulum is membranous. In Stemonyphantes, the junction between the tegulum and the suprategulum is a wide flex- ible hinge (Hormiga 1994, fig. 2c). Also, the tegular- suprategular junction is unusual in Ste- monyphantes because it is on the ventral face rather than the mesal face of the palpal bulb. Aealysis.=I used PAUP version 3.1 (Swofford 1993), Hennig86 version 1.5 (Far- ris 1988) and NONA version 1.6 (Goloboff 1993) to search the data (46 taxa, 74 charac- ters) for the most parsimonious topology. In PAUP, I ran a heuristic search with 100 rep- licates of random taxon addition subjected to tree bisection-reconnection branch swapping. In Heneig86, I used the “mh*,bb*” search strategy. In NONA, I ran a search under the “amb=” setting (modified rule 3; see Cod- dington & Scharff 1994; Zujko-Miller 1999) with the “mult*” random taxon addition al- gorithm for 100 replicates followed by the “max*” branch-swapping algorithm. I used MacClade (Maddisoe & Maddison 1992) to analyze character optimization. Successive character weighting (Farris 1969; Carpenter 1988) by the maximum value of the rescaled consistency index was per- formed in PAUP with the base weight set to 46 THE JOURNAL OF ARACHNOLOGY Figures 1-6. — Palpus of male Carorita hiberna from Thomas Ridge, North Carolina. 1. Embolic di- vision, dorsal view; 2. Ventral view with embolic division removed; 3. Ventral view; 4. Mesal view; 5. Fetal view; 6. Palpal tibia, dorsal view. Abbreviations: DRP, dorsal radical process; DSA, distal suprate- gular apophysis; E, embolus; EM, embolic membrane; ETP, ectal tibial process; F, fundus; P, paracymbium; PT, protegulum; PTA, palpal tibial apophysis; R, radix; SD, sperm duct; SPT, suprategulum; ST, subtegu- lum; T, tegulum; TP, radical tail piece. 1000. Trees found by Hennig86 and NONA were imported into PAUP. NONA trees were saved using the “ksv*” command. PAUP will arbitrarily resolve polytomies in trees saved using the “sv” command in NONA. The so- lution set from all three programs (PAUP, Hennig86, and NONA) was combined. Dupli- cate trees were eliminated. The remaining unique trees were then filtered to exclude po- lytomous trees when more highly resolved compatible trees were found (Coddington & Scharff 1996). This set of trees was re-weight- ed and the data re-analyzed in PAUP. Results. — PAUP, Hennig86, and NONA and all found multiple trees of 236 steps (cal- culated after excluding uninformative charac- ZUJKO-MILLER— -m/COTTC/i' HIBERNUS 47 Figures 7-9. — Genitalia of Carorita species. 7. Male palpus of C. hiberna from Thomas Ridge, North Carolina, mesoventral view with embolic division removed; 8, 9. Carorita limnaea from McLean New York; 8. Cleared epigynum, ventral view; 9. Male palpus, ventromesal view. Abbreviations: CD, copu- latory duct; CDC, copulatory duct capsule; DP, dorsal plate of epigynum; DSA, distal suprategular apoph- ysis; E, embolus; EM, embolic membrane; FD, fertilization duct; P, paracymbium; PT, protegulum; R, radix; S, spermatheca; SPT, suprategulum; ST, subtegulum; T, tegulum; TP, radical tail piece; VP, ventral plate of epigynum. ters) with a consistency index (Cl) of 0377 and a retention index (RI) of 0.673. PAUP found 63 trees, HennigSO found 33 trees, and NONA found 45 trees. A total of 78 unique most parsimonious trees were found. Of these, 24 trees were more highly resolved but oth- erwise compatible with other most parsimo- nious trees. All trees place Carorita limnaea sister to Carorita hiberna. A strict consensus of all most parsimonious trees has a trichot- omy composed of the Carorita clade, Asthen- argus paganus and Gongylidiellum vivum. This clade is part of an 11 -tomy composed of various erigonine terminals and clades. Suc- cessive character weighting stabilizes on six trees. All six trees are 236 steps long under equal weights and were among the original set of 78 trees. Except for the additional taxa, this set of six trees is identical to the result found in Hormiga's (in press) original analysis. Dis- agreement among the six trees is found in two places: the relationships among the linyphi- ines, micronetines and all other linyphiids and the relationships among Drepanotylus Holm 1945, Sciastes Bishop & Crosby 1938 and the “distal erigonines” clade. The “distal erigon- ines” clade is fully resolved and identical in all six trees. The “distal erigoinines” clade features a monophyletic Carorita clade sister to Gongylidiellum vivum. Sisicottus remains in the position reported by Zujko-Miller (1999), sister to Oedothorax gibosus (Blackwall 1841). Implications,”-™The two species which for- merly composed Carorita have traditionally been united based on their chaetotaxy (espe- cially the presence of a prolateral macroseta on tibia I), the large paracymbium, the form of the suprategular apophysis, overall palpal conformation, the general form of the palpal 48 THE JOURNAL OF ARACHNOLOGY tibia, the relatively long tarsi, and proportional characteristics of the eyes (Duffey 1971; Mil- lidge 1977). Millidge (1977) considered C limnaea and C. paludosa Duffey 1971 to be members of a monophyletic group despite conspicuous differences in the form of the em- bolic division and cited it as evidence of how greatly the embolic division can vary within an otherwise “good” genus. Unfortunately, I have been unable to test this claim since I have not been able to examine specimens of C. paludosa and existing descriptions and il- lustrations are not adequate for scoring many of the characters in the matrix. Optimization* — This analysis indicates that the genus Carorita can be defined phy- logenetically on the basis of two unambiguous synapomorphies: the presence of a radical tail- piece (character 21) and the loss of a lamella characteristica (character 27). Although C. paludosa has a large mesal sclerite of the em- bolic division (Millidge 1977, fig. 159), I have been unable to determine from published de- scriptions whether it is a radical tailpiece or a lamella characteristica. Five characters can be optimized to provide additional support for the monophyly of Carorita. However, these five characters could also be placed on alter- native tree nodes without affecting tree length. These are the presence of papillae on the pro- tegulum (character 9; unknown in C paludo- sa), a long embolus (character 17; short in C. paludosa), the shape of the radical tailpiece (character 22; unknown in C. paludosa), en- capsulation of the epigynum (character 38; unknown in C. hiberna and C. paludosa), and a tuberculate radical tailpiece (character 74; unknown in C paludosa). Although C lim- naea and C hiberna are the only taxa in the analysis with a tuberculate radical tailpiece, their two closest relatives, Asthenargus pa- ganus, and Gongylidiellum vivum, both lack a radical tailpiece and were therefore coded as inapplicable for this character. A generic re- vision of Carorita with a strong phylogenetic component should be undertaken in the near future to address the placement of C palu- dosa. All three Carorita species feature a strong kink or loop in the path of the sperm duct in the tegulum near the junction between the te- gulum and the suprategulum (Figs. 7, 9; Mil- lidge 1977, fig. 159). Most other erigonines have a sperm duct that follows a smooth curve through the tegulum. However, when I at- tempted to code this as a cladistic character, I found a spectrum of conditions rather than a small number of discrete character states. Nevertheless, the presence of a kink in the sperm duct is a useful part of the diagnosis of Carorita. Missing data* — -Attributes of the tracheal system have long been relied upon as char- acters in linyphiid classification (Blest 1976; Millidge 1984, 1986). Hormiga (in press) demonstrated that these characters have high consistency. This analysis optimizes C hiber- na as having the haplotracheate condition, al- though no observation of the tracheal system of C hiberna has been made. If C. hiberna is speculatively coded as having the desmitra- cheate condition (1: character 51) with no taenidia (0: character 52), 328 unique most parsimonious trees of 237 steps result (192 trees in Hennig86, 246 trees each in NONA and PAUP; Cl = 0.376, RI = 0.670). A total of 102 of these trees are more re- solved than otherwise compatible trees. The strict consensus of these 102 trees has little phylogenetic structure. It features a polytomy of 21 clades and individual taxa at the node which represents the Erigoninae. Carorita limnaea and C. hiberna are not consistently monophyletic and form two of the branches in this 21-tomy. Successive character weight- ing of these 102 trees stabilizes on 30 trees (Cl = 0.372, RI = 0.666). These trees are two steps longer under equal weights than the most parsimonious trees. The strict consensus of these 30 trees forms a topology unlike that found by Hormiga (1999), especially in the deep nodes. Carorita limnaea and C. hiberna form a paraphyletic assemblage rather than a monophyletic group. Under the assumption that C hiberna is haplotracheate rather than desmitracheate without taenidia, the data are found to be more internally consistent and the resulting phylogenetic hypothesis that is con- sistent with the previous hypothesis (Hormiga in press), one which suffered from few miss- ing observations. DISCUSSION Phylogenetic hypotheses can be communi- cated either implicitly as higher taxonomic names or explicitly in the form of a dado- gram. If taxonomy is to reflect phylogenetic history, genera should serve as implicit hy- ZUJKO-MILLER— S/S'/COTTf/^ HIBERNUS pothesis that a particular group of species share a unique common ancestor. However, monotypic genera do not provide this group- ing information and thus cannot be considered phylogenetic hypotheses. On the other hand, cladistic analyses by their nature explicitly convey detailed phylogenetic hypotheses. The ranks and labels assigned to nodes in a clad- ogram are less critical since the hypothesis of relationships is fully expressed in the tree. As part of an explicit phylogenetic hypothesis, detailed grouping information is available and monotypic genera are less problematic. The placement of Carorita hiberna was dif- ficult because of the many apparently auta- pomorphic character states exhibited by this species. However, no matter how many auta- pomorphic character states are found in a par- ticular species, it must still have a sister taxon (Platnick 1976). Placing Carorita hiberna in a new monotypic genus without identifying its putative sister taxon would have been an ab- dication; a lack of a grouping hypothesis. As it stands, my circumscription of Carorita is based on explicit evidence and is subject to falsification given sufficient new evidence. Although these conclusions reflect the best available evidence, missing data has led to the optimization of some character states based on inference (Platnick et al. 1991). When addi- tional specimens are discovered, the predic- tions of this analysis should be tested. Between 1981 and 1995, over 70% of new linyphiid genera have been monotypic (Plat- nick 1989, 1993, 1997). The volume of new monotypic genera generated within the Liny- phiidae indicates a critical lack of phyloge- netic consideration. A phylogenetic approach based on morphological characters demands careful examination of comparative anatomy but the result is a rich hypothesis not only of taxonomic relationships but of character evo- lution. A broad phylogenetic context will be very helpful for associating monotypic genera with their relatives. This will no doubt lead to the synonymization of many existing genera. Ultimately, the utility of the linyphiid taxo- nomic system will be greatly improved. How- ever, I have attempted to demonstrate that we do not need to wait for some grand phyloge- netic hypothesis of the Linyphiidae. The basic context for applying phylogenetic criteria to new work in linyphiid systematics is already available. 49 TAXONOMY Carorita Duffey & Merrett 1963 Carorita Duffey & Merrett 1963:573-576, figs. 1- 8 ((?,?). Duffey 1971: 14-15, figs. 1-8 (3,9). C Locket, Millidge & Merrett 1974: 92-94, figs. 56a-g (3,9). Blest 1976: 188. Millidge 1977: 40, figs. 158, 159 (3); 1984: 245-246. Brignoli 1983: 328. Roberts 1987: 108, figs. 53d, 53e (3,9). Platnick 1989: 223; 1993: 252; 1997: 328. Heimer & Nentwig 1991: 124, figs. 352.1-353.4 (3,9). Type species by original designation and by monotypy, Oedothorax limnaeus Crosby & Bishop 1927: 149-150, figs. 11-14. Diagnosis. — Males of Carorita can be dis- tinguished from most other erigonines by the form of the suprategulum which has a distinct boundary at its origin and by the kinked or looped path of the sperm duct through the te- gulum (Figs. 7, 9). They are distinguished from males of other erigonines with these character states by the absence of a lamella characteristica and the presence of a radical tail piece (uncertain in C. paludosa). Females can be distinguished from other erigonines by the complex, anteriorly-projecting looped path of the copulatory ducts (Fig. 8). Description. — Tibial macrosetae 2-2-1-1 (except in C. hiberna: 2-2-2- 1); Tm IV absent. Tibia I with one distal prolateral macroseta (absent in C hiberna). At least in C. limnaea, chelicerae with imbricated stridulatory files and median tracheal trunks unbranched, short- er than laterals, confined to abdomen (Blest 1976; haplotracheate sensu Millidge 1984). Males: Palpal tibia with one prolateral and one retrolateral trichobothrium. Carorita lim- naea and C hiberna (but not C. paludosa) with long embolus in frontal plane, radix with tuberculate tailpiece and no anterior radical process (Figs. 3, 9). Distal part of cymbium and ectal side of palpal tibia with clusters of macrosetae. Females: Females of C. paludosa have not been examined and details of female genitalia cannot be unambiguously interpreted using published descriptions and illustrations; females of C. hiberna unknown; females of C limnaea with posteriorly oriented ventral plate invagination leading to copulatory open- ings; copulatory ducts encapsulated with loop at anterior maximum and again posterior to spermathecae near junction. Fertilization ducts project posteriomesally from spermathecae (Fig. 8). 50 THE JOURNAL OF ARACHNOLOGY Composition. — Three species: Carorita limnaea (Crosby & Bishop 1927), C. palu- dosa Duffey 1971 and C. hiberna (Barrows 1945). Distribution. — North America (C. lim- naea, C. hiberna) and Europe (C. limnaea, C. paludosa). Carorita hiberna (Barrows 1945) NEW COMBINATION Figs. 1-7 Sisicottus hibernus Barrows 1945: 74, figs. 1, 2 (d). Brignoli 1983: 356. Type . — Male holotype from United States: North Carolina, Great Smoky Mountains Na- tional Park, Mingus Creek, 1 February 1943, in OSU, examined. Diagnosis. — Males of C hiberna are read- ily distinguished from other Carorita species by the unusual, complex shape of the embolus which runs in a more or less frontal plane (Fig. 3E), by their long, ectally projecting, tu- berculate, radical tailpiece (Fig. 3TP), by the horn-like process on the ectal side of the pal- pal tibia (Fig. 6ETP), by the presence of a distal macroseta on tibia III, by the absence of a prolateral macroseta on tibia I, and by the presence of a dorsal radical process (Fig. 1, DRP). Description. — Male: (from Thomas Ridge, North Carolina). Carapace length = 0.6 mm. Tibial macrosetae weak; 2-2-2- 1; Tm I = 0.34. Chelicerae with 5-6 promarginal teeth; 5 retromarginal teeth with proximal 2 larger than distal 3. Embolus describing a semicir- cular arc in more or less frontal plane with outside of curve ectal; tip of embolus arced in nearly transverse plane with outside of curve dorsal (Figs. 1, 3E). Embolic membrane long, narrow, curved with outside ectal (Fig. 2EM). Radix with anteriomesally projecting tuber- culate tailpiece (Fig, 3TP) and ectodorsally projecting dorsal radical process (Fig. IDRP). Protegulum on ectal side of bulb (Fig. 2PT). Palpal tibia with long, straight apophysis with flat distal margin; horn-like process on ectal side of apophysis; ectal and mesal sides of palpal tibia both with regions of semitrans- parent chitin (Figs. 4-6). Female: Unknown. Distribution. — Known only from the Great Smoky Mountains National Park, North Car- olina. Material examined. — UNITED STATES: North Carolina: Swain County, Great Smoky Mountains National Park, Mingus Creek, 1 Febru- ary 1943, Id, (holotype, Barrows, OSU); Swain County, Great Smoky Mountains National Park, Thomas Ridge, ca. 200 m from trailhead at route 441, west-facing slope below trail, 4540 feet, old growth mixed hardwood, UTM: E3107, N39436, 22 September 1994, 1 m- litter sample. Id, (Cribbs team, GSMNP); Haywood County, Great Smoky Mountains National Park, Cataloochee, 150 m south of mouth of Palmer Branch at Caldwell Fork, 2800-3000 feet, old growth hemlock, UTM: E3107, N39436, 23-31 March 1997, pitfall trap. Id, (Coyle, Edwards & Wright, GSMNP) North Carolina, Great Smoky Mountains National Park. Natural history. — Carorita hiberna is known only from three male specimens col- lected in the Great Smoky Mountains National Park. The rarity of this species in the face of intensive collecting efforts by Dr. Frederick Coyle (Western Carolina University) and his collaborators within the park raise questions about the long term prospects for the contin- ued survival of this species. The rare diplurid spider Microhexura montivaga Bishop & Crosby 1925 is known only from spruce and Fraser fir forests in the southern Appalachians and is currently listed as a federally protected endangered species (Coyle 1981; Fridell 1995). The staphylinid beetle Dasycerus bi- color Wheeler & McHugh 1994 and the lin- yphiid spider Sisicottus montigenus Bishop & Crosby 1938 are both endemic to this same region and habitat type, have experienced re- cent and dramatic declines in their popula- tions, and may be worthy of similar protected status (Wheeler & McHugh 1994; Zujko-Mill- er 1999). However, C. hiberna has been found in both mixed hardwood and hemlock forests. It does not appear to be associated with the declining spruce-fir habitat and though rare, there is no evidence that its population has actually declined since its discovery. ACKNOWLEDGMENTS The following persons and institutions kindly loaned specimens for this study: R. Bradley, Ohio State University Museum of Biodiversity (OSU); J.A. Coddington, Nation- al Museum of Natural History, Smithsonian Institution (USNM); FA. Coyle, Western Car- olina University -Great Smoky Mountains Na- tional Park Collection (GSMNP); N.I. Plat- nick, American Museum of Natural History (AMNH). ZUJKO-MILLER— 5/5/COITC/5 HIBERNUS 51 G. Hormiga, J. Coddington, N. Scharff, N. Platnick, B. Opell, and Berry made helpful comments on an earlier draft of this manu- script. I would like to thank G. Hormiga, J. Coddington, E Coyle, N. Platnick, N. Scharff, C. Griswold, D. Lipscomb, M. Allard, and J. Clark for stimulating discussion and advice. G. Hormiga generously allowed me to use drawings and data from a manuscript in press in addition to unpublished data. F. Coyle kept an eye out for new specimens of Carorita ki- te rna. The George Washington University and National Museum of Natural History (Smithsonian Institution) provided valuable lab space, materials, and equipment. Equip- ment and financial support was provided by an NSF-PEET grant to G. Hormiga and J. Coddington (DEB-9712353). Additional fi- nancial support was provided by a Weintraub fellowship from The George Washington Uni- versity. LITERATURE CITED Barrows, W.M. 1945. New spiders from the Great Smokey [sic] Mountains National Park. Ann. En- tomol. Soc. America, 38:70-76. Blest, A.D. 1976. The tracheal arrangement and the classification of linyphiid spiders. J. Zool., London, 180:185-194. Brignoli, P.M. 1983. A Catalog of the Araneae De- scribed Between 1940 and 1981. Manchester Univ. Press, Manchester, 755 pp. Carpenter, J.M. 1988. Choosing among multiple equally parsimonious cladograms. Cladistics, 4: 291-296. Coddington, J.A. 1983. A temporary slide mount allowing precise manipulation of small struc- tures. Verb. Naturwiss. Ver. Hamburg, 26:291- 292. Coddington, J.A. & N.I. Scharff. 1994. Problems with zero-length branches. Cladistics, 10:415- 423. Coddington, J.A. & N.I. Scharff. 1996. Problems with “soft” polytomies. Cladistics, 12:139-146. Coyle, EA. 1981. The mygalomorph spider genus Microhexura (Araneae, Dipluridae). Bull. Amer- ican Mus. Nat. Hist., 170:64-75. Crosby, C.R. & S.C. Bishop. 1927. New species of Erigoneae and Theridiidae. J. New York En- tomoL, Soc., 35:147-157. Duffey, E. 1971. Carorita paludosa n. sp., a new linyphiid spider from Ireland and eastern Eng- land. Bull. British Arachnol. Soc., 2:14-15. Duffey, E. & P. Merrett. 1963. Carorita limnaea (Crosby & Bishop), a linyphiid spider new to Britain, from Wybunbury Moss, Cheshire. Ann. Mag. Nat. Hist., 6:573-576. Farris, J.S. 1969. A successive approximations ap- proach to character weighting. Syst. Zool., 18: 374-385. Farris, J.S. 1988. Hennig86, version 1.5. Program and documentation. Computer program distrib- uted by D. Lipscomb, Dept, of Biol. Sci., The George Washington Univ., Washington, D.C. Fridell, J. 1995. Endangered and threatened wild- life and plants; spruce-fir moss spider determined to be endangered. Federal Register, 60:6968- 6974. Goloboff, PA. 1993. NONA. Noname (a bastard son of Pee-Wee), version 1.6 (32 bit version). Program and documentation. Computer program distributed by J.M. Carpenter, Dept, of Entomol- ogy, American Museum of Natural History, New York. Heimer, S. & W. Nentwig. 1991. Spinnen Mitte- leuropas. Ein Bestimmungsbuch. Verlag Paul Parey, Berlin. 542 pp. Holm, A. 1979. A taxonomic study of European and east African species of the genera Pelecopsis and Trichopterna (Araneae, Linyphiidae), with descriptions of a new genus and two new species of Pelecopsis from Kenya. Zool. Scripta, 8:255- 278. Hormiga, G. 1994. Cladistics and the comparative morphology of linyphiid spiders and their rela- tives (Araneae, Araneoidea, Linyphiidae). Zool. J. Linnean Soc., 111:1-71. Hormiga, G. In press. Higher level phylogenetics of erigonine spiders (Araneae, Linyphiidae, Eri- goninae). Smithsonian Contrib. Zool. Locket, G.H., A.F. Millidge & P Merrett. 1974. British Spiders. Volume III. The Ray Society, London. 314 pp. Maddison, WP. & D.R. Maddison. 1992. Mac- Clade: Analysis of phylogeny and character evo- lution. Version 3.0. Sinauer Assoc., Sunderland, Massachusetts. Millidge, A.F. 1977. The conformation of the male palpal organs of linyphiid spiders, and its appli- cation to the taxonomic and phylogenetic anal- ysis of the family (Araneae, Linyphiidae). Bull. British Arachnol. Soc., 4:1-60. Millidge, A.F. 1984. The taxonomy of the Liny- phiidae, based chiefly on the epigynal and tra- cheal characters (Araneae: Linyphiidae). Bull. British Arachnol. Soc., 6:229-267. Millidge, A.F. 1986. A revision of the tracheal structures of the Linyphiidae (Araneae). Bull. British Arachnol. Soc., 7:57-61. Platnick, N.I. 1976. Are monotypic genera possi- ble? Syst. Zool., 25:198-199. Platnick, N.I. 1989. Advances in Spider Taxonomy 1981-1987. Manchester Univ. Press, Manchester. 637 pp. Platnick, N.I. 1993. Advances in Spider Taxonomy 1988-1991. New York Entomol. Soc. 846 pp. 52 THE JOURNAL OF ARACHNOLOGY Platnick, N.I. 1997. Advances in Spider Taxonomy 1992-1995. New York Entomol. Soc. 976 pp. Platnick, N.I., C.E. Griswold & J.A. Coddington. 1991. On missing entries in cladistic analysis. Cladistics, 7:337-343. Roberts, M.J. 1987. The Spiders of Great Britain and Ireland. Vol. 2. Linyphiidae and Check List. Harley Books, Martins. 204 pp. Swofford, D.L. 1993. Phylogenetic analysis using parsimony, version 3.1. Illinois Nat. Hist. Sur- vey, Champaign, Illinois. van Helsdingen, P.J. 1968. Comparative notes on the species of the Holarctic genus Stemonyphan- tes Menge (Araneida, Linyphiidae). Zool. Med- ed. Leiden, 43:117-139. Wheeler, Q.D. & J.V. McHugh. 1994. A new southern Appalachian species, Dasycerus bicolor (Coleoptera: Staphylinidae: Dasycerinae), from declining endemic fir forests. The Coleopts. Bull., 48:265-271. Zujko-Miller, J. 1999. Revision and cladistic anal- ysis of the erigonine spider genus Sisicottus (Ar- aneae, Linyphiidae, Erigoninae). J. ArachnoL, 27(2):000-000. Manuscript received 1 May 1998, revised 7 Decem- ber 1998. 1999. The Journal of Arachnology 27:53-63 "TOWARDS A PHYLOGENY OF ENTELEGYNE SPIDERS (ARANEAE, ARANEOMORPHAE, ENTELEGYNAE) Charles E. Griswold^, Jonathan A. Coddington^, Norman I. Platnick^ and Raymond R. Forster^: ^Department of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 USA; ^Department of Entomology, National Museum of Natural History, NHB-105, Smithsonian Institution, Washington, D.C. 20560, USA; ^Department of Entomology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024 USA; ^McMasters Road, R.D. 1, Saddle Hill, Dunedin, New Zealand ABSTRACT. We propose a phylogeny for all entelegyne families with cribellate members based on a matrix of 137 characters scored for 43 exemplar taxa and analyzed under parsimony. The cladogram confirms the monophyly of Neocribellatae, Araneoclada, Entelegynae, and Orbiculariae. Lycosoidea, Amaurobiidae and some included subfamilies, Dictynoidea, and Amaurobioidea (sensu Forster & Wilton 1973) axe polyphyletic. Phyxelidinae Lehtinen is raised to family level (Phyxelididae, NEW RANK). The family Zorocratidae Dahl 1913 is revalidated. A group including all entelegynes other than Eresoidea is weakly supported as the sister group of Orbiculariae. The true spiders or Araneomorphae (ara- neae verae of Simon 1892) comprise more than 30,000 described species. The classifi- cation of this group has undergone a revolu- tion in the last 30 years, sparked by Lehtinen’s (1967) comprehensive reassessment of ara- neomorph relationships and steered by Hen- nig’s phylogenetic systematics (Heneig 1966; Platnick & Gertsch 1976). Spider classifica- tion, portrayed by some authors as chaotic (Head 1995; Elgar et al. 1990; Vollrath & Parker 1997: Prenter et al. 1997) is actually one of the better-understood megadiverse or- ders (Coddiegton & Levi 1991): including the results reported here, 100 of the 108 currently recognized families (93%) have been placed cladistically, that is, in a higher taxon based on evidence assessed phylogenetic ally. New character systems compared across worldwide samples of taxa have led to many new and thought-provoking hypotheses in araneo- morph phylogeny. The strongest test of such hypotheses is how simply they can account for the available data, i.e., most parsimonious cladograms based on matrices of taxa by char- acters. Tests specifically designed at the fam- ily level and above are increasingly common: ^Presented as part of a symposium on ‘'Higher Level Phylogenetics of Spiders.” Raven (1985) and Goloboff (1993a) for My- galomorphae (15 families); Coddington (1986, 1990a, b) for Orbiculariae (13 families) and Entelegynae; Platnick et al. (1991) on haplo- gynes (17 families) and Araneomorphae; Gris- wold et al. (1994, 1998) for Araneoidea (12 families), and Griswold (1993) for Lycosoidea and related families (11 families). The latter studies targeted large but rela- tively well-defined lineages. It is now feasible to probe how these large lineages are related. We chose exemplars from all cribellate fami- lies, reasoning that taxa retaining this plesiom- Orphic feature are more likely to straddle the basal nodes of the phylogeny of higher groups than are their relatives that have lost the cri- helium: therefore they are most likely to re- flect phylogenetic groundplans. Although phylogeeetically ancient, the cribellum is a complex feature unlikely to have evolved more than once. Most major araneomorph clades have cribellate members (exceptions are Palpimanoidea and Dionycha). A phylog- eny of these basal taxa should mirror the re- lationships of the large clades they exemplify. As Lehtinen (1967: 202) declared, “because of the central position of the Cribellate groups in Araneomorphae, a detailed revision of them is a short cut to a rough classification of the whole suborder.’' 53 54 THE JOURNAL OF ARACHNOLOGY A parsimonious cladogram based on an ex- plicit taxon-character matrix is concise, logi- cal, and testable. Our analysis tests many su- prafamilial hypotheses of the last 30 years and is the first attempt to relate them using quan- titative phylogenetic techniques: Amaurobioi- dea {sensu Forster & Wilton 1973), Amauro- biidae and included subfamilies (sensu Lehtinen 1967), Dictynoidea and Desidae (sensu Forster 1970), Entelegynae (sensu Coddington 1990b; Coddington & Levi 1991), Lycosoidea (Homann 1971; sensu Griswold 1993); Orbiculariae (sensu Coddington 1986, 1990a, b); and the ’RTA clade’ (sensu Cod- dington & Levi 1991). TAXA AND CHARACTERS Table 1 comprises 43 exemplars from 21 of the 22 araneomorph families with cribellate members. As outgroups we included HYPO- CHILIDAE (Hypochilus), AUSTROCHILI- DAE (Hickmania and Thaida), and FILIS- TATIDAE (Filistata and Kukulcania, Filistatinae). From eresoids, we included OECOBIIDAE (Oecobius and Uroctea) and ERESIDAE (Eresus and Stegodyphus). From Orbiculariae we included DEINOPIDAE (Deinopis and Menneus), ULOBORIDAE (Octonoba and Uloborus) and an araneoid groundplan. Recent phylogenetic study of this superfamily (Griswold et al. 1998) gives us confidence that the reconstructed groundplan accurately reflects the primitive conditions for Araneoidea. From “dictynoids” we included DICTYNIDAE (Dictyna and Nigma, Lathys, and Tricholathys representing Dictyninae, Ci- curinae, and Tricholathysinae, respectively), DESIDAE (Badumna Candida, B. longinquua, and Matachia, formerly Matachiinae), and NI- CODAMIDAE (Megadictyna). From “amau- robioids” we included AMAUROBIIDAE (Amaurobius and Callobius (Amaurobiinae), Metaltella (Metaltellinae), Retiro and Pimus (Macrobuninae), Phyxelida, Vytfutia, and Xe- vioso (Phyxelidinae)), AMPHINECTIDAE (Maniho), NEOLANIDAE (Neolana), AGE- LENIDAE (Neoramia), and TITANOECI- DAE (Goeldia and Titanoeca). From lyco- soids and related groups we included CTENIDAE (Acanthoctenus), MITURGIDAE (Raecius and Uduba, Uliodoninae), PSE- CHRIDAE (Psechrus), STIPHIDIIDAE (Baiami and Stiphidion); TENGELLIDAE (Tengella), and ZOROPSIDAE (Zoropsis). We omitted Gradungulidae because the cri- bellate genera are extremely rare in collections and its placement in Austrochiloidea seems firm (Forster, Platnick & Gray 1987; Platnick et al. 1991). Voucher specimens for exemplars are deposited in the California Academy of Sciences (CAS) with the exception of Vytfutia (Deeleman coll.) and male Raecius (NHMV). Character data taken from the literature in- clude the suite of classical characters from spider internal anatomy (characters 43-49: Platnick 1977; ex Millot 1931, 1933, 1936; Marples 1968 [these have been recorded for hypochiloids, austrochiloids, and such a wide variety of haplogyne and entelegyne Araneo- clada that we are confident that the assumed states for entelegyne exemplars in Table 1 are justifiable]) and character 114, presence/ab- sence of the muscle M29 in the male palp (as- sumed for all taxa in Table 1 following Huber 1994). Silk ultrastructure data are taken from Eberhard & Pereira (1993) and from unpub- lished observations (R. Carlson in lit.). Characters, character states, and codings are listed in Table 1. Some features are most suc- cinctly described by reference to a taxon for which they are typical, e.g, ’dictynid conduc- tor.’ For figures of entelegyne genitalia see es- pecially Lehtinen (1967), Coddington (1990a) and Griswold (1993); for features of spinner- ets see especially Platnick et al. (1991) and Griswold et al. (1998). Character evolution is summarized on the cladogram (Fig. 1) opti- mized via Clados (Nixon 1992) and MacClade (Maddison & Maddison 1992). METHODS AND ANALYSIS Spigot classification follows Coddington (1989); all specimens were critical point dried before scanning electron microscope (SEM) examination of spinning organs. Behavioral observations were made on living animals in the field or lab. The matrix (all characters unordered and equally weighted) was analyzed with three phylogenetic packages: Nona 1.6 (Goloboff 1993b), Hennig86 1.5 (Farris 1988), and Pee- Wee 2.6 (Goloboff 1997), using a wide variety of randomized and directed search strategies. Nona (using both ‘amb =’ and ‘amb-’ options for clade support) and Hennig86 found the same three topologies, including Fig. 1 (length 376, ci 0.43, ri 0.69). The two alternate to- pologies involved local rearrangements of Ni- GRISWOLD ET AL.— ENTELEGYNE SPIDER PHYLOGENY 55 codamidae and Eresoidea. The strict consen- sus has one 4-tomy at the entelegyne node, otherwise identical to Fig. 1. We used successive and implied weighting (Carpenter 1988; Goloboff 1993c) to further evaluate the data. Successive weighting in Nona (length 16,346, ci 0.63, ri 0.80) pre- ferred Fig. 1. Successive weighting in Hen- nig86 (length 1 127, ci 0.79, ri 0.88) found Fig. 1 as well as two other trees one step longer. Pee-Wee at concavity functions of 3 and 4 (fits 962.6 and 1009.0, length 378) found one tree in which Retiro and Pimus swapped places, otherwise identical to Fig. 1. Concavity 5 (fit 1041.6, lengths 376, 378) found Fig. 1 as well as the tree found at concavities 3 and 4. Con- cavity 6 (fit 1067.8, length 376) found only Fig, 1. Because Fig. 1 was the only topology judged optimal by all criteria (equal, succes- sive, and implied weights), we recommend it as the working hypothesis for entelegyne re- lationships. Table 1 gives the number of steps, consistency index, retention index, weight {ex Hennig86), and fit {ex Pee-Wee, concavity 4) for all characters on Fig. 1. In addition to mapping character support at nodes, we also examined cladogram ro- bustness with branch support indices (Bremer 1994) calculated with Nona using the param- eters ’h25000; bsupport8;\ The “Bremer Support” (“Decay Index”) for a given node in the shortest unconstrained tree is the num- ber of additional steps required in the shortest trees for which that node collapses. The fol- lowing Bremer Support values were found for the clades on Fig. 1: Austrochiloidea (5), Araneoclada (1), Entelegynae (1), Haplogy- nae (8), Eresoidea (1), Stegodyphus-Eresus (8), Uroctea-Oecobius (4), Canoe Tapetum Clade (0), Orbiculariae (2), Deinopis-Octo- noba (3), Deinopis-Menneus (4), Uloborus- Octonoba (5), Megadictyna-Zoropsis (0), Di- vided Cribellum Clade (1), Titanoecoids (1), Titanoeca-Goeldia (2), Vytfutia-Phyxelida (2) , Xevioso-Phyxelida (2), RTA Clade (1), Dictynidae (2), Tricholathys-Nigma (1), Die- tyna-Nigma (3), Amaurobioids (1), Fused Paracribellar Clade (2), Stiphidioids (1), Sti- phidion-Baiami (5), Agelenoids (2), Maniho- Badumna c (2), Maniho-Metaltella (2), De- sidae (1), Badumna l-Badumna c (5), Retiro-Zoropsis (1), Amaurobiidae (1), Pi- mus-Callobius (2), Amaurobius-Callobius (3) , Tengella-Zoropsis (4), Raecius-Zoropsis (2), Raecius-Uduba (2), Lycosoidea (2), and Acanthoctenus-Zoropsis (3). RESULTS Status of the Lycosoidea and their kin.^ — Homann (1971, followed by Levi 1982) de- fined the Lycosoidea on the basis of a grate- shaped tapetum in the indirect eyes. Griswold (1993) produced a phylogeny for those fami- lies plus selected tengellids and miturgids that Table 1. — Character by taxon matrix. Rows represent characters. The first state listed is coded as "0", the second as "1", etc., "?" == unknown, = non-applicable. Columns represent taxa. The final five columns give the number of steps, the consistency index, the retention index, weight, and fit on Fig. 1. Taxon abbreviations are Ac = Acanthoctenus; Am = Amaurobius; AP = Araneoidea; Ba = Baiami; Bd = Badumna Candida', Ca = Callobius', De = Deinopis', Di = Dictyna', Er= Eresus', FP = Filistata', Go = Goeldia', Hi = Hickmania', Hy = Hypochilus', Ix — Badumna longinquua', Ku = Kukulcania', La = Lathys', Ma = Matachia', Me = Megadictyna', Mh = Maniho', Mn = Menneus', Mt = Metaltella', Ne = Neoramia', Ni = Nigma', N1 = Neolana', Oc = Octonoba', Oe = Oecobius; Ph = Phyxelida', Pi = Pimus', Ps = Psechrus', Ra = Raecius', Re = Retiro', Sg = Stegodyphus', St = Stiphidion', Te = Tengella', Th = Thaida', Ti = Titanoeca', Tr = Tricholathys', Ud = Uduba', U1 = Uloborus', Ur = Uroctea', Vy = Vytfutia', Xe = Xevioso', Zo = Zoropsis. Character abbreviations: ALS = anterior lateral spinnerets; annul. = annulate; C = conductor; cent. = central; CR’ed = cut and reeled; diet = dictynid; E = embolus; embr. = embraces; extra = in addition to C and MA; iLl = inside first leg; iL4 inside fourth leg; long. = longitudinal; L3 = third leg; L4 = fourth leg; MAP = major ampullate; mAP = minor ampullate; membr. = membraneous; met = metal- telline; Oe = oecobiid; oLl = outside first leg; opp. = opposite; papill. = papillate; PMS = posterior median spinnerets; post. = posterior; PY == pyriform; scl. = sclerotized; spinn. = spinneret; STP = sclerotized tegular process; squam. = squamate; strob. = strobilate; Th = Thaida', trans. = transverse; trich. = trichobothria; Ulo = uloborid. 56 THE JOURNAL OF ARACHNOLOGY Table 1. Legs 1 . Femoral trichobothria.: absent; present; 2 . Tarsal organ: exposed; capsulate; 3 . Tarsal trichobothria: absent; present; 4 . Tarsal trichobothrial rows: 1; 2+; 5. Metatarsal trichobothria: 1-2; >2; 6 . Tarsal trichobothria: normal; longer distally; 7 . Palpal tibial trichobothria: absent; present; 8 . Trich. base: smooth; trans. ridge; long, ridge; 9 . Trichobothrial base: simple; notched; 10 .Trochanteral notch: absent; present; 11 .Tarsal claws: 3; 2; 12 . Claw tufts: absent; present; 13 .Calamistral rows: 2; 1; 3: 14 . Calamistrum: linear; oval; 15 . Calamistrum origin: basal; median; 16 . Male palpal femoral thorns: absent; present; 17 . Female palpal femoral thorns: absent; present; 18 .Feathery hairs: absent; present; 1 9 . Hair: plumose; pseudoserrate; serrate; 20 . Metatarsal preening combs: absent; present; 21 .Tibial ventral spine number: <7; >7; 22 . Male metatarsus I: unmodified; modified; 2 3 . Male tibial crack: absent; present; 24 . Serrate accessory claw setae: absent; present; 25 . Female posterior leg scopula: absent; present; 2 6 . Deinopoid tarsal comb: absent; present; Carapace 27 . Carapace shape: oval; square; round; 28 . Clypeal hood: absent; present; 29 . Thickened setae at fang base: absent; present; 30 . Male chelicerae: normal; bowed; 31 . Chelicerae: normal; small; 32 . Chelicerae: free; fused at base; 33 . Cheliceral teeth: present; absent; 34 . Cheliceral chela: absent; present; 35 . Cheliceral boss: absent; present; 36 . Cheliceral boss: small; large; 37 . Chilum; absent; median; bilateral; 38 .Tapetum: primitive; canoe; grate; absent; 3 9 . Posterior eye row: straight; recurved; 4 0 . Serrula tooth rows: multiple; single; 4 1 . Sigilla: present; absent; 4 2 . Medial cheliceral concavity: present; absent; 4 3 . Venom gland: cheliceral; carapace; absent; 4 4 . Coxal gland duct: convoluted; simple; 4 5 . 5th endostemite invagination: present; absent; 4 6 . Midgut cheliceral diverticula: present; absent; 4 7 . Pharynx muscle origin: carapace; rostrum; Abdomen 4 8 . Intestine: M-shaped; straight; 4 9 . Heart ostia number: 4; 3 or 2; 50 . Ana! tubercle: small; large; 51 . Posterior spiracles: 2; 1 ; 52 . Posterior spiracles: wide; narrow; 53 . Post, respiratory system: booklungs; tracheae; 5 4 . Median tracheae; simple; branched; 5 5 . Lateral tracheae: simple; branched; absent; 56 . Epiandrous spigots: present; absent; 57 . Epiandrous spigots: dispersed; 2 bunches; Spinneret morphology 58 . Spinn. cuticle: annul.; ridged; squam.; papilL; 59 . Cribellum: present; absent; 60 . Cribellum; entire; divided; 61 . Cribellate spigots; uniform; clumped; 62 . Cribellate spigots: strobilate; claviform; 63 .Tartipores: absent; present; 64 . ALS field margin; narrow; broad; 65 . ALS MAP; clustered; dispersed; 66. ALS MAP number: >3; 3; 2; 1; 67 . ALS MAP nubbin; absent; present; 68 . ALS MAP nubbin: one; extra; 69 . ALS piriform margin: rounded; flat; sharp; MTHS'KSEUOEl«QB»WrLTGVXPHSB»®*ffllRPACTBI3PZA yhiiugrreenlcPeiiraioyehltaehtadxeimaeadsoc St Cl RI ' Wt Fit 0000000000011000000000000000000000000000000 1 1.00 1.00 10 10 0001111111111111111111111111111111111111111 1 1.00 1.00 10 10 0000000000000000011000001111111111111111111 2 .50 .95 4 8 00 0000000000000111011 2 .50 .75 3 8 0001100000000000011000001111111111111111111 3 .33 .89 2 6.6 10 1111011001111000000 4 .25 .66 1 5.7 0000000000? 0? 000000000000110111110111000000 3 .33 .77 2 6.6 00000110100000000001111100? 0000000111111111 5 .20 .75 1 5 0110000000000000000000000000000000000000000 1 1.00 1.00 10 10 0000000000000100000000000010101000000100001 6 .16 .00 0 4.4 0000000000000000000000000000000000000000011 1 1.00 1.00 10 10 0000000000000000000000000000000000000000111 1 1.00 1.00 10 10 0112211-11111-11111111111111111111111 2 1.00 1.00 10 10 ooooooo-ooooo-oooooooooooooooooooooboiiiiii 1 1.00 1.00 10 10 0110000-00? 00-00000000110000000000000000000 2 .50 .66 3 8 0100000000000000000001110000000001100000000 4 .25 .40 1 5.7 0000000000000000000001110000000001000000000 2 .50 .66 3 8 0100000001111000000000000110100000000010000 5 .20 .50 1 5 0000000001111200000000000000000000000000000 2 1.00 1.00 10 10 0000000000000000000000100001111000011000000 4 .25 .50 1 5.7 0000000000000000000000000000000000000000011 1 1.00 1.00 10 10 0000000000000000000001110000000000000000000 1 1.00 1.00 10 10 0000000000000000000000000000000000000011010 2 .50 .50 2 8 0000000001111110000000000000000000000000000 2 .50 .80 4 8 0000000000? 00000000000000000000000000111111 1 1.00 1.00 10 10 0000000001111010000000000000000000000000000 HTHFKSE001MOQ»OOTLTGyXPNSBt®®«4BIRPACTBUPZA 2 .50 .75 3 8 yhiiugrreenlcPeiiraioyehltaehtadxeimaeadsoc St Cl RI Wt Fit 0000011220000000000000000000000000000000000 2 1.00 1.00 10 10 0110011000000000000000000000000000000000000 2 .50 .66 3 8 0110000001? 00011111111111111110111111111111 4 .25 .70 1 5.7 0000000000000001100000000000000000000000000 1 1.00 1.00 10 10 0000000110000000000000000000000000000000000 1 1.00 1.00 10 10 0001100000000000000000000000000000000000000 1 1.00 1.00 10 10 0001100110000000000000000000000000000000000 2 .50 .66 3 8 0001100000000000000000000000000000000000000 1 1.00 1.00 10 10 0000011000? 01101111111111111111111111111111 4 .25 .66 1 5.7 00 ? -00-1111111111111111111111111111 1 1.00 1.00 10 10 0?? 00?? 00??? 020? 00? 11122211? 22122? 111222222 7 .28 .68 1 4.4 0000033003333111111111111221111111111111222 5 .60 .86 5 6.6 0111100110000111111111111011111111111111100 5 .20 .55 I 5 0111111111111111111111111111111111111111111 1 1.00 1.00 10 — 0110011111111111111111111111111111111111111 2 .50 .50 2 8 0111111111111111111111111111111111111111111 1 1.00 1.00 10 0111111111122111111111111111111111111111111 2 1.00 1.00 10 10 0111111111111111111111111111111111111111111 1 1.00 1.00 10 — 0111111111111111111111111111111111111111111 1 1.00 1.00 10 — 0111111111111111111111111111111111111111111 1 1.00 1.00 10 __ 0111111111111111111111111111111111111111111 1 1.00 1.00 10 — HTHFKSETOmO«®OTLTGVXPNSB!»®*®IKPAiCTKUPra. yhiiugrreenlcPeiiraioyehltaehtadxeimaeadsoc St Cl RI Wt Fit 0001111111111111111111111111111111111111111 I 1.00 1.00 10 10 0001111111111111111111111111111111111111111 1 1.00 1.00 10 10 0000000110000000000000000000000000000000000 1 1.00 1.00 10 10 0101111111111111111111111111111111111111111 2 .50 .00 0 8 -1-0011111111111111111111111111111111111111 1 1.00 1.00 10 10 OOOllllllllllllllllllllllllllllllll'llllllll 1 1.00 1.00 10 10 -0 00000? 11011111000110000001111000000000 6 .16 .58 0 4.4 00000? 00002222000000000001110000000000 2 1.00 1.00 10 10 00? 0000010??? 0011110? 100? 1? 1000001000? 11? 11 5 .20 .66 1 5 00? 00000-0??? 00 0? -11??? oil HTHFKSEtX)!»mM)im.Te?XPNSm«*fflIRPACTKDPZA 2 .50 .66 3 8 yhiiugrreenloPeiiraioyehltaehtadxeimaea^oc St Cl RI Wt Fit 0111133111111211111111111111111111111111111 3 1.00 1.00 10 10 0000000100000100000000000000000000000000000 2 .50 .00 0 8 0001111-10000-00000111111111110111111101111 6 .16 .61 1 4.4 00000? O-OOOOO-OOOOOOOOOOOOOOOOOOOOOOOOllOOl 2 .50 .50 2 8 00011? 0-0? 000-00000000000000000000000000000 1 1.00 1.00 10 10 0110011111111111111111111111111111111111111 2 .50 .50 2 8 000000000000000000000? 001111111110011000000 2 .50 .90 4 8 00011111-0000000000000000000000000000000000 2 .50 .75 3 8 0221100030033323333223222222222223222222222 9 .33 .60 2 4 0110010001111101110011000000000011101000000 19 .11 .50 0 3.3 -11—0—11000-000—00 101-1 4 .25 .50 1 5.7 0000000002111001100000000000000110000000020 5 .40 .57 2 5.7 GRISWOLD ET AL.— ENTELEGYNE SPIDER PHYLOGENY 57 Table 1. — Continued. 70 . PMS mAP; absent; present; 7 1 . PMS mAP number: 1 ; > 1 ; 7 2 . PMS mAP position: median/anterior; posterior; 7 3 . PMS mAP nubbins: absent; present; 7 4 .PMS mAP nubbins: 1; 2; 7 5 . PMS aciniform: present; absent; 7 6 . PMS aciniform number: >3; 1-3; 7 7 . PMS paracribellars: absent; present; 78 . PMS paracribellar: at apex; mid-fieid; 7 9 . PMS paracribellar: bunched; encircling; 80 . PMS paracribellar shafts: single bases; grouped; 81 . PMS paracribellar bases: round; long, narrow; 82 .PMS paracribellar: deinopoid; strobilate; floppy; 8 3. PMS cylindrical spigots: absent; present; 84 . PMS cylindrical spigot number: 1-2; many; 85 . PLS cuticular finger: absent; present; 8 6 . PLS aggregate gland spigot: absent; present; 87 . PLS paracribellar: absent; present; 8 8 . PLS paracribellar number: 1 ; 2-3 ; 8 9 . PLS paracribellar: separate; + modified spigot; 90 .PLS modified spigot: absent; present; 91 . PLS modified spigot: in field; segregated; 92 . PLS apical segment: domed; conical; elongate; 93 . PLS apical enlarged seta: absent; present; Male Genitalia 94 . Tibial retroapical process: absent; present; 95 . Tibial retroapical process: simple; complex; 96 . Tibial ventroapical process: absent; present; 97 .Tibial dorsal process: absent; apical; proximal; 98 .Tibial dorsal process: simple; complex, folded; 99 . Tibial prolateral process: absent; present; 100 . Paracymbium: absent; present; 101. Cymbial dorsal scopula: absent; present; 102 . Pyriform bulb; absent; present; 103 . Tegular locking lobe: absent; present; 104 . Subtegular locking lobe: absent; present; 105 . Conductor: present; absent; tegular groove; 106 . Conductor: sclerotized; hyaline; 107 . Scl. C: embr. E; opp. E; cent.; Ulo; Oe; Th; 108 .C embr. E: apex only; desid; diet; met; apical; 109. Median apophysis; present; absent; 110 . Median apophysis: convex; concave; 111 . Median apophysis; fixed; flexibly attached; 112 . ‘Extra’ tegular processes: absent; present; 113 . ‘Extra’ tegular process; conical; STP; membr.; 114 . Palpal tarsus M29 muscle: present; absent; Female Genitalia 115 . Female genitalia: haplogyne; entelegyne; 116. Epigynum: absent; present; 117. Epigynum teeth: absent; present; 118 .Convoluted vulva: absent; present; 119. Posterior receptaculum; absent; present; Behavior and Silk 120 . Cribellate silk axial lines: present; absent; 121. Cribellate silk reserve warp: present; absent; 122 .Cribellate silk nodules: absent; present; 123 . Cribellate fibrils; cylindrical; flat; 124 . Cribellate silk: uniform; puffed; 12 5 . Web posture; inverted; erect; 126. Orb web architecture: absent; present; 127 . Deinopid web architecture; absent; present; 128 . Combing leg support; fixed L3; mobile L4; 12 9. Frame construction: absent; present; 130 . Radius construction; absent; CR'ed; doubled; 0111111011111111111111111111111111111111111 -110011-00000000000000000000000000100000011 -000000-00000110100000000000000000000000000 -000011000000100000000110000000000000010111 10 0 00 0-000 0000011000000000000000000000000000000000000 00010—000000000001000000001000100000000010 0111100-01111-11111001111111111110111000000 -0000 0000-00010—000000111011-000 -1000 0000-111 111 0 00 -0000 0000-01110—000111111011-001 -0000 0000-00000—111 0 000 -0022 0000-11111—111111111111-111 00000111111111? 1011111111111110111111111111 110011110? 0-1000110000000-110000011111 0000000000000070000000100001000000000000000 0000000000000100000000000000000000000000000 0101100000000-11000110101110111110111000000 -0-11 00 01-0-1? 1-11011-111 -0 00 0-0-000-00011-010 11? 000000111111110010111111111111111101? 11? 00? 11111000-0-0000000000000000-0? 00? 00? 0000221? 1101000000000111011011-060? 00101 00? 0000000? 0000000000? 11000000000-000000000 HTHFKSEU0I»100M)NTLTGVXPNSB»SM®IRPACTKUPZA yhiiugrreenlcPeiiraioyehltaehtadxeimaeadsoc 0000000000000001011001001111111111111111-11 0-00—0—0001111010000000-00 0000000000000000000000000000000001111011000 0000000000000022100111112000220201111000000 000—110000 00-0-0000 0000000000000000000000000000000001011000000 1000000000000100000000000000000000000000000 0000000000000000000000000000000000000011111 0001100000000000000000000000000000000000000 0000000000000000000000000000000000010110111 0000000000000000000000000000000000000110111 0011100000000000000220000000000000000000000 00 00000000000000—0000000000001101111011 05 00442233320000—110200110000—1 1— 0 00 2222 1-14—3111 0011111001100011111000000110000000000000100 00 00—000 000000-0000110010010-01 00 00—001 111010-1101111111111-11 01 11—000 110100-0110001111011-10 -0 00 00-0 00 1111-11-2- 0000011111111111111111111111111111111111111 HTHFKSEUODMUQft^C)lm,TGVXPNSBN^M4BIRPAC^RUPZA yhiiugrreenlcPeiiraioyehltaehtadjceimaeadsoc 0000011111111111111111111111111111-111111111 0100011111111111111111111111111111111111111 -0 00000000000000000000001111110010000000 0000000000000000000000000000110000000000000 0110000000000000000000000000000000000000000 HTHFKSEtXMmrQAMDNTLTGVXPNSBl®®*4BIRPMrrBUPZA yhiiugrreenlcPeiiraioyehltaehtadxeimaeadsoc 0? 0000? -00000-011?? 0??? 0? 0?? 0010?? 00 0 0 07 ? 0? 0? 1000? -00011-0107 ? 0??? 0? 07? 0010?? 00000?? 0? 0? 1001? -11111-111?? 1??? 1? 17? 11117? 1111??? 1? 0? 0110? -000 0 0-0007 ? 0??? 0? 07? 0000?? 0000??? 0? 0? 0000? -? 1111-011?? 0??? 0? 07? 0 0 0 07 ? 0000??? 0? 0001117? 1000000117? 1?? 10000110111? 11111101- 000000000111110000007000000000000000000000- 0000000001100000000000000000000000000000000 0??? 01? -11111-11????? 111?- 007? 07? 0???0?? 1?111?????????0?' 1?221?????????0?' ?? 17? 17? 1111111? — 0????????0?0— — 0????????0?0 — 131 . Hub construction: absent; present; 0??? 07? 132 .Temporary spiral construction: absent; present; 0???0?' 133 . Sticky spiral construction: absent; present; 0???0?' 134 .Sticky silk localization; absent; oLl; iL4; 0???0?' 1 35 . L4 sticky silk shift; absent; present; 0??? 0? 136. Non-sticky-sticky line grip: otherwise; w L4; 0??? 0? 137. Wrap attack: absent; present; 007? 10? -? 1? 111????????? 0??? — 0???????? 0? 0 — 1? 111?????????0??? — 0????????0? 0 — 1? 111????????? 0??? — 0???????? 0? 0 — 2? 111????????? 0??? —0???????? 0? 0— -? 1? 111????????? 0??? — 0???????? 0? 0 — 1? 111????????? 0??? — 0???????? 0? 0 — ?? 11111100?????? 1? 0??? 07? 0? 000007? 0? 2 .50 .00 0 8 4 .25 .50 1 5.7 3 .33 .00 0 6.6 5 .20 .50 1 5 1 1.00 1.00 10 — 1 1.00 1.00 10 10 5 .20 .00 0 5 5 .20 .66 1 5 3 .33 .60 2 6.6 3 .33 .66 2 6.6 4 .25 .72 1 5.7 1 1.00 1.00 10 10 2 1.00 1.00 10 10 3 .33 .66 2 6.6 7 .14 .60 0 4 2 .50 .00 0 8 1 1.00 1.00 10 — 9 .11 .55 0 3.3 4 .25 .40 1 5.7 2 .50 .50 2 8 6 .16 .44 1 4.4 1 1.00 1.00 10 10 8 .25 .53 1 4 1 1.00 1.00 10 10 St Cl RI wt Fit 3 .33 .89 2 6.6 2 .50 .75 3 8 2 .50 .80 4 8 8 .25 .57 1 4 1 1.00 1.00 10 10 2 .50 .50 2 8 2 .50 .00 0 8 1 1.00 1.00 10 10 1 1.00 1.00 10 10 3 .33 .60 2 6.6 2 .50 .75 3 8 3 .66 .66 4 8 3 .33 .71 2 6.6 11 .45 .45 2 4 4 1.00 1.00 10 10 8 .12 .50 0 3.6 4 .25 .25 0 5.7 5 .20 .50 1 5 8 .12 .41 0 3.6 2 1.00 1.00 10 10 1 1.00 1.00 10 10 St Cl Rl wt Fit 1 1.00 1.00 10 10 2 .50 .66 3 8 2 .50 .83 4 8 1 1.00 1.00 10 10 1 1.00 1.00 10 10 St Cl RI Wt Fit 2 .50 .50 2 8 4 .25 .25 0 5.7 2 .50 .50 2 8 1 1.00 1.00 10 10 2 .50 .80 4 8 7 .14 .57 0 4 1 1.00 1.00 10 10 1 1.00 1.00 10 10 1 1.00 1.00 10 10 1 1.00 1.00 10 10 2 1.00 1.00 10 10 1 1.00 1.00 10 10 1 1.00 1.00 10 10 1 1.00 1.00 10 10 2 1.00 1.00 10 10 1 1.00 1.00 10 10 1 1.00 1.00 10 10 3 .33 .71 2 6.6 Hapjpgynae j-83mii5ii6i28 Entelegynae A usttDchiloidea - - 2 4i 49 51 53 se jsz A laneoclada 4.^4D42-^4z^^sszozzj^ Neocdbellatae Paleo :ribe!!atae 58 THE JOURNAL OF ARACHNOLOGY Hypochilus f-Thaida S ^2 « S F-Hiokmania Fillstata Kukuicania Stegodyphus Eresus tUroctea Oecobius f-^Araneoidea Deinopis Menneus Uloborus Octonoba Megadictyna Titanoeca Goeldia Vytfutia t;:.i-a,p|rXevioso M — 'Phyxeiida Lathy s Tricholathys ^ Dictyna R 5 r P f f; Nigma Neolana Stiphidion Baiami Neoramia HHManiho ^ ■ Metaltella Matachia Badumna I Badumna c t-Retfro Pimus Arreurobius Callobius Tenge lla H-Ra©ciiis Uduba Psechrus H-Acanthoctenus ■ ■ Zoropsis GRISWOLD ET AL.— ENTELEGYNE SPIDER PHYLOGENY 59 shared with Lycosoidea a derived, oval caL amistrum. Figure 1 corroborates the oval caL amistrum (14) as a synapomorphy for this group. The former miturgid genera Campos- dchomma, Zorocrates, Zorodictyna, Raecius, and Uduba (clade BB in Griswold 1993) com^ prise the family Zorocratidae Dahl 1913. Syn- apomorphies of Zorocratidae are: male tibial crack (23), clumped cribellar spigots (61), and a male palpal tibial ventroapical process (96). Zorocratidae are sister to Lycosoidea (Fig. 1) based on posterior median spinnerets with many cylindrical spigots (84) and a dorsal scopula on the cymbium (101). Stiphidiidae {Baiami and Stiphidion in Fig. 1) were for- merly included in Lycosoidea, but in this anal- ysis are more closely related to the Ageleni- dae, Amphinectidae, Desidae, and Neolanidae, based on a suite of rather homo- plasious characters. Perhaps the strongest ev- idence is the clumped rather than dispersed arrangement of the paracribellar spigots on the PMS (80). Jointly these characters overrule the grate- shaped tapetum, which therefore ap- pears to have evolved independently in sti- phidiids. Lycosoidea will probably be further modified by detailed study of lineages now included in Cteeidae. Amaurobiidae and included subfamilies (sensu Lehtinen 1967). -—Defined classically by a plesiomorphy (presence of the cribeL lum), perhaps Amaurobiidae was most obvi- ously in need of relimitatioe after the collapse of the old Cribellatae. Lehtinen (1967) pro- posed nine subfamilies, seven of which are treated here: Matachiinae (Badumna, Mata- chia), Desinae (Maniho), Phyxelidinae (Phy- xelida, Xevioso; also Vytfutia following Gris- wold 1990), Stiphidiinae {Baiami, Stiphidion), Macrobuninae (Pimus, Retiro), Metaltellinae (Metaltella) and Amaurobiinae (Amaurobius, Callobius). The cribellate Altellopsinae are known only from females (Lehtinen 1967: 338) and Rhoicininae are neither cribellate nor amaurobiids (Griswold 1993). Unless the lim- its of the family are expanded to include the lycosoids, one must conclude that Amaurobi- idae is the most seriously polyphyletic family discovered to date. Only Lehtinen’s macro- bunines (paraphyletic) and amaurobiines ar- guably remain in Amaurobiidae (Fig. 1). Leh- tinen’s desines and metaltellines are closely related and belong in Amphinectidae sensu Forster & Wilton 1973: this result corrobo- rates Davies (1998). As noted above, Stiphi- diidae sensu Forster & Wilton 1973 (Stiphi- dion and Baiami in Fig. 1) is sister to Neolanidae, not amaurobiids. Matachiines are desids sensu Forster & Wilton 1973 (Fig. 1); at least Matachia is strikingly similar to the ecribellate Desis. Phyxelidinae Lehtinen 1967 (formerly Amaurobiidae), which includes Am- bohima, Kulalania, Lamaika, Malaika, Ma- tundua, Namaquarachne, Phyxelida, Pongo- lania, Themacrys, Vidole, Vytfutia and Xevioso, constitutes a distinct family (Phyxel- ididae, NEW RANK). Phyxelididae (Vytfutia, Xevioso and Phyxelida in Fig. 1) is sister to Titanoecidae (Titanoeca and Goeldia in Fig. 1), not close to amaurobiids. Phyxelididae is corroborated by various synapomorphies: male (16) and female (17) palpal femur thorns, modified male metatarsus I (22), and long, narrow, closely packed and laterally flat- tened PMS paracribellar spigots (81). Amaurobioidea and Dictynoidea (semu Forster & Wilton 1973)*— Building upon an extensive study of the respiratory systems of Figure L — Cladogram for entelegyne spider exemplars. Character changes are noted on branches by character number, with ambiguous optimizations underlined. Characters optimized at the neocribellate node are ambiguous because Mygalomorphae, Mesothele, and Amblypygi are not considered in this matrix. Taxon names are to the right of their branch. Familial assignments of exemplars on this cladogram are: AGELENIDAE (Neoramia), AMAUROBIIDAE (Amaurobius, Callobius, Pimus, and Retiro), AMPHL NECTIDAE (Maniho and Metaltella), AUSTROCHILIDAE (Hickmania and Thaida), CTENIDAE (Acam thoctenus), DEINOPIDAE (Deinopis and Menneus), DESIDAE (Badumna c[andida), Badumna llonginquua], and Matachia), DICTYNIDAE (Dictyna, Nigma, Lathys, and Tricholathys), ERESIDAE (Eresus and Stegodyphus), FILISTATIDAE (Filistata and Kukulcania), HYPOCHILIDAE (Hypochilus), NEOLANIDAE (Neolana), NICODAMIDAE (Megadictyna), OECOBIIDAE (Oecobius and Uroctea), PHYXELIDIDAE (Phyxelida, Vytfutia, and Xevioso), PSECHRIDAE (Psechrus), STIPHIDIIDAE (Baiami and Stiphidion), TENGELLIDAE (Tengella), TITANOECIDAE (Goeldia and Titanoeca), ULOBORIDAE (Octonoba and Uloborus), ZOROCRATIDAE (Raecius and Uduba), and ZOROPSIDAE (Zoropsis). 60 THE JOURNAL OF ARACHNOLOGY spiders, Forster (1970) and Forster & Wilton (1973) defined two superfamilies that con- tained all the families treated here as well as others. The Amaurobioidea (unbranched, slen- der tracheae) included Agelenidae, Amauro- biidae, Amphinectidae, Ctenidae, Cyclocteni- dae, Neolanidae, Psechridae, and Stiphidiidae. Figure 1 suggests a much more limited ar- rangement: Amaurobiidae is sister to only ten- gellids, zorocratids, and lycosoids. The Dic- tynoidea (at least median tracheae branched) included Amaurobioididae, Anyphaenidae, Argyronetidae, Cybaeidae, Dictynidae, Desi- dae, Hahniidae, and Nicodamidae. The un- branched condition (54) is primitive and thus Amaurobioidea should not be expected to be monophyletic. Branched tracheae (54), how- ever, originates six times on Fig. 1 and al- though it helps to define families (Uloboridae, Dictynidae) it does not, as yet, clearly define a larger clade. Dictynidae is monophyletic and is sister (or part of the sister group) to most distal entelegynes, including Neolanidae, Sti- phidiidae, Amphinectidae, Amaurobiidae, De- sidae, Agelenidae, Tengellidae, Zorocratidae, and Lycosoidea. The ‘RTA’ Clade. — Coddington & Levi (1991) suggested an informal but informative grouping for those spiders having a retrolater- al tibial apophysis (RTA) on the male palp, including taxa thought to lack the RTA sec- ondarily. A variety of tibial apophyses on the male palp exist, sometimes on the same ani- mal, and here we code this diversity as four homologies rather than one. The RTA itself (94) still defines roughly the same lineage (Fig. 1), except that the absence of the RTA in Nicodamidae, Phyxelididae, and Titanoe- cidae is primitive, not secondary and thus ex- cludes them from the RTA clade. An addi- tional unambiguous synapomorphy is trichobothria on the tarsi (3). Vytfutia appar- ently evolved the RTA independently. Outgroup of the Orbiculariae. — With more than 10,000 described species and a great variety of documented webs and other behaviors, the Orbiculariae comprise one of the largest and most interesting clades of spi- ders. Coddington (1990b) implied Dicty no- idea as a possible Orbicularian sister group. Platnick et al. (1991) suggested that the Amaurobioidea (represented in their study by Amaurobius) and Dictynoidea (represented by Dicty na) together could be the sister group. Coddington & Levi (1991) suggested that the ‘RTA clade’ (including Dictynoidea, Amau- robioidea, Dionycha, and Lycosoidea) was the orbicularian sister group. The first two studies lacked many relevant taxa, and the last was a review, not a new analysis. This study omits palpimanoids, but suggests that the sister group to Orbiculariae is essentially all ente- legyne spiders other than eresoids. In retro- spect, the difficulty in finding the sister group of orbweavers is understandable. The answer, suggested by all of these studies in one way or another, is not one or a few classical fam- ilies, or even any pre-existing taxonomic hy- pothesis in spiders. It is, rather, a previously unknown suprafamilial clade whose precise characterization still requires much work. In one alternative parsimonious topology for this dataset, however, the orbicularian sister group is Nicodamidae (Megadictyna), based on ser- rate accessory claw setae (24), the entire cri- bellum (60), and inverted posture in the web (125). Given this possibility, further field stud- ies of nicodamid behavior and web construc- tion would be welcome. New entelegyne groups. — As before (Cod- dington & Levi 1991; Scharff & Coddington 1997; Griswold et al. 1998) we propose in- formal names for a few clades so that they may be discussed and tested by other workers prior to formal taxonomic recognition. All en- telegynes distad of eresoids we call the “ca- noe-tapetum clade” (Fig. 1). On this dado- gram the canoe tapetum arises unambiguously at this node and certainly represents an im- portant restructuring of the spider visual sys- tem. The clade is also supported by the ap- pearance of the modified silk spigot on the PLS (90), called the pseudoflagelliform in dei- nopoids, but now known to have homologs in many other lineages. This spigot presumably contributes additional axial fibers to the cri- bellate silk, as noted by Eberhard & Pereira (1993), and may represent an important event in the evolution of capture threads. It seems logical to redefine the Amauro- bioidea to include all families in the sister clade to Dictynidae (Fig. 1). Likewise, the clade including Titanoecidae and Phyxelididae could be called the “titanoecoids.” “Agele- noids” could refer to Agelenidae, Amphinec- tidae, and Desidae. Similarly it seems worthwhile to recognize the “fused paracribellar clade” as well as the GRISWOLD ET AL.— ENTELEGYNE SPIDER PHYLOGENY 61 “divided cribellum clade” (Fig. 1). The func- tional role of paracribellar fibrils in capture threads is not known with certainty, but these taxa have the paracribellar shafts fused so that many spigots emerge from the same shaft~a striking morphology (80). The same clade is also defined by wide ALS piriform field mar- gins (64)“another spinning field feature whose functional significance is still unknown. Like- wise, the divided cribellum (60) is scarcely free from homoplasy, but one of its origins does define a large clade of spiders (Fig. 1). DISCUSSION These results constitute the most detailed proposal to date for basic entelegyne relation- ships. Added to previous analyses (refs, in Coddington & Levi 1991), 100 of the 108 cur- rent spider families are now placed in higher taxa intermediate between suborder and su- perfamily. Incertae sedis families are only Cryptothelidae, Cybaeidae, Cycloctenidae, Hahniidae, Halidae, Homalonychidae, the re- maining Miturgidae, and Zodariidae. The higher taxa Palpimanoidea and Dionycha (if monophyletic) also need to be placed in the general schema. Both groups are entirely ecri- bellate and so many informative characters cannot be scored. Palpimanoidea was placed by Platnick et al. (1991) as sister group of the clade Orbiculariae plus the RTA clade, which group was supported by the presence of the PLS pseudoflagelliform gland spigot (90). Nothing in our additional data challenges this conclusion. On the whole, these results sharp- en rather than contest earlier work by provid- ing a much more detailed and factually based hypothesis for test. A notable result is the unavoidable homo- plasy in character systems traditionally relied upon in araneomorph classification. For ex- ample, branched median tracheae (54) arise six times, the divided cribellum (60) evolves three times and reverts to entire three times (Dictynidae, Matachia and Raecius). Loss and regain of epiandrous spigots (56) occurs. Al- though the median apophysis (109) is homol- ogous wherever it occurs, eight unambiguous losses are required. Once again understanding spider phylogeny seems to be, as succinctly put by Coddington Sl Levi (1991: 575), “not so much a question of finding characters as it is of allocating homoplasy.” Spider data, however, is not abnormally homoplasious. Based on regression coefficients calculated by Sanderson & Donoghue (1989) 43 taxa yield on average ci values of about 0.35; the value observed here (0.43) is rather better. Several tasks remain before the first, rough, cladistic reconnaissance of Araneae could be said to be “complete.” The major groups Pal- pimanoidea (Forster & Platnick 1984) and Dionycha (sensu Coddington & Levi 1991) as well as families mentioned above, are not placed on this cladogram. At infrafamilial lev- els, many cribellate enigmas remain unstud- ied, e.g., Poaka (Psechridae?) and Aebutina (Dictynidae?). The generality of these results is uncertain because in many cases the mono- phyly of families containing cribellate and ecribellate members is untested (especially Agelenidae and Dictynidae). Nevertheless, in its breadth of taxa and characters this study represents progress towards a comprehensive family-level phylogeny for the true spiders. ACKNOWLEDGMENTS Griswold wishes to acknowledge financial support from National Science Foundation grants BSR-9020439 and DEB-9020439, the Exline-Frizzell and In-house Research Funds of the CAS, and post-doctoral fellowships from the Smithsonian Institution and the American Museum of Natural History. Cod- dington wishes to acknowledge financial sup- port from National Science Foundation grants DEB-9712353 and DEB-97-07744, and the Neotropical Lowlands Program and Biotic Surveys and Inventory Program from the Smithsonian Institution. Robin Carlson shared new data on the fine structure of cribellate silk; her research was enabled by the CAS Summer Systematics Institute, itself supported by NSF grant BIR-9531307. We thank Per de Place Bjprn, L. Joy Bou- tin, Fred Coyle, Crista Deeleman, Gustavo Hormiga, Bill Peck, Barbara and the late Vin- cent Roth, Evert Schlinger, and Darrell Ubick for providing crucial specimens. A male Rae- cius was lent by Jurgen Gruber (Naturhisto- risches Museum, Zoologisches Abteilung, Vi- enna [NHMV]). D. Ubick (CAS) and Mrs. Susan Braden (Smithsonian) assisted with scanning electron microscopy. LITERATURE CITED Bremer, K. 1994. Branch support and tree stability. Cladistics, 10:295-304. 62 THE JOURNAL OF ARACHNOLOGY Carpenter, J.M. 1988. Choosing among multiple equally parsimonious cladograms. Cladistics, 4: 291-296. Coddington, J.A. 1986. The monophyletic origin of the orb web. Pp. 319-363. In Spiders: Webs, Behavior, and Evolution. (W.A. Shear, ed.). Stan- ford Univ. Press. Coddington, J.A, 1989. Spinneret silk spigot mor- phology: Evidence for the monophyly of orb- weaving spiders, Cyrtophorinae (Araneidae), and the group Theridiidae plus Nesticidae. J. Arach- noL, 17:71-95. Coddington, J.A. 1990a. Ontogeny and homology in the male palpus of orb-weaving spiders and their relatives, with comments on phylogeny (Ar- aneoclada: Araneoidea, Deinopoidea). Smithson- ian Contrib. ZooL, 496:1-52. Coddington, J.A. 1990b. Cladistics and spider clas- sification: araneomorph phylogeny and the monophyly of orb weavers (Araneae: Araneo- morphae; Orbiculariae). ActaZool. Eennica, 190: 75-87. Coddington, J.A, & H.W. Levi. 1991. Systematics and evolution of spiders (Araneae). Ann. Rev, Ecol. Syst., 22:565-592. Dahl, F. 1913. Vergleichende Physiologie und Morphologie der Spinnentiere unter besonderer Berucksichtigung der Lebensweise. 1. Die Bezie- hungen des Korperbaues und der Farben zur Um- gebung. Jena, Pp. 1-112. Davies, V.T 1998. A revision of the Australian me- taltellines (Araneae: Amaurobioidea: Amphinec- tidae: Meteltellinae). Invert. Taxon., 12:211-243. 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). J. Arachnol., 21:161-174. Elgar, M.A., N. Ghafer & A. Read. 1990. Sexual dimorphism in leg length among orb-weaving spiders: A possible role for sexual cannibalism. J. Zool. London, 222:455-470. Farris, J.S. 1988. Hennig86 1.5. Microcomputer program available from Dr. Arnold Kluge, Mu- seum of Zoology, University of Michigan, Ann Arbor, Michigan, 49109-1079 USA. Forster, R.R. 1970. The Spiders of New Zealand, III. Otago Museum Bulletin, 3:1-184. Forster, R.R. & N.I. Platnick. 1984. A review of the archaeid spiders and their relatives, with notes on the limits of the superfamily Palpima- noidea (Arachnida, Araneae). Bull. American Mus. Nat. Hist, 178:1-106. Forster, R.R., N.I. Platnick & M.R. Gray. 1987. A review of the spider superfamilies Hypochiloidea and Austrochiloidea (Araneae, Araneomorphae). Bull. American Mus. Nat. Hist., 185(1): 1-1 16. Forster, R.R. & C.L. Wilton. 1973. The spiders of New Zealand, Part IV. Otago Museum Bulletin, 4:1-309. Goloboff, P. 1993a. A reanalysis of mygalomorph spider families. American Mus. Nov., 3056:1- 32. Goloboff, P. 1993b. Nona 1.6. Computer program available from J.M. Carpenter, Dept. Entomolo- gy, American Mus. Nat. Hist., Central Park West at 79“^, New York, New York 10024 USA. Goloboff, P. 1993c. Estimating character weights during tree search. Cladistics, 9:83-92. Goloboff, P. 1997. Pee- Wee 2.6. Computer pro- gram available from J.M. Carpenter, Dept. En- tomology, American Mus. Nat. Hist., Central Park West at 79'*’, New York, New York 10024 USA. Griswold, C.E. 1990. A revision and phylogenetic analysis of the spider subfamily Phyxelidinae (Araneae, Amaurobiidae). Bull. American Mus. Nat. Hist., 196:1-206. Griswold, C.E. 1993. Investigations into the phy- logeny of the lycosoid spiders and their kin (Arachnida, Araneae, Lycosoidea). Smithsonian Contrib. Zool., 539:1-39. Griswold, C.E., J.A. Coddington, G. Hormiga & N. Scharff. 1994. Phylogeny of the orb web build- ing spiders (Araneomorphae, Orbiculariae). American Arachnol., 50:5. (abstract) Griswold, C.E., J.A. Coddington, G. Hormiga & N. Scharff. 1998. Phylogeny of the orb- web build- ing spiders (Araneae, Orbiculariae: Deinopoidea, Araneoidea). Zool. J. Linn. Soc., 123:1-99. Head, G. 1995. Selection on fecundity and varia- tion in the degree of sexual size dimorphism among spider species (Class Araneae). Evolu- tion, 49:776-781. Hennig, W. 1966. Phylogenetic Systematics. Univ. Illinois Press, Urbana. Homann, H. 1971. Die Augen der Araneae: Ana- tomie, Ontogenese und Bedeutung fiir die Sys- tematik (Chelicerata, Arachnida). Zeitschr. Morph. Tiere, 69:201-272. Huber, B. 1994. Genital bulb muscles in entele- gyne spiders. J. Arachnol., 22:75-76. Lehtinen, P.T. 1967. Classification of the cribellate spiders and some allied families. Ann. Zool. Fen- nica, 4:199-468. Levi, H.W. 1982. Araneae. Pp. 77-95. In Synopsis and Classification of Living Organisms. 2. (S.B. Parker, ed.). New York: McGraw Hill. Maddison, W.P. & D.R. Maddison. 1992. Mac- Clade, ver. 3.0. Sinauer Associates, Sunderland, Massachusetts. Millot, J. 1931. Les diverticules intestinaux du cephalothorax chez les Araignees vrais. Zeitschr. Morph. Okologie Tiere, 21:740-764. Millot, J. 1933. Notes complementaires sur I’anatomie des Liphistiides et des Hypochilides, GRISWOLD ET AL.— ENTELEGYNE SPIDER PHYLOGENY 63 a propos d’un travail recent de A. Petrankevitch. Bull. Soc. ZooL France, 58:217-235. Millot, J. 1936. Metamerisation et musculature ab- dominale chez les Araneomorphes. Bull. Soc. Zool. France, 61:181-204. Marples, B.J. 1968. The hypochilomorph spiders. Proc. Zool. Soc. London, 179:11-31, Nixon, K.C. 1992. Clados, version 1.2. Program and documentation, available from author at L.H. Bailey Hortorium, Cornell Univ., Ithaca, New York 14853 Platnick, N.I. 1977. The hypochiloid spiders: A cladistic analysis, with notes on the Atypoidea ■ (Arachnida, Araneae). American Mus. Nov., 2627:1-23. Platnick, N.I., J.A. Coddington, R.F Forster & C.E. Griswold. 1991. Spinneret evidence and the higher classification of the haplogyne spiders (Araneae, Araneomorphae). American Mus. Nov., 3016:1-73. Platnick, N.I. & W.J. Gertsch. 1976. The suborders of spiders: A cladistic analysis. American Mus. Nov., 2607:1-15. Prenter, J., W.I. Montgomery & R.W. El wood. 1997. Sexual dimorphism in northern temperate spiders: Implications for the differential mortal- ity model. J. Zool. London, 243:341-349. Raven, R.J. 1985. The spider infraorder Mygalo- morphae: Cladistics and systematics. Bull. American Mus. Nat. History, 182:1-180. Sanderson, M.J. & M.J. Donoghue. 1989. Patterns of variation in levels of homoplasy. Evolution, 42(8):1781-1795. Scharff, N. & J.A. Coddington. 1997. A phyloge- netic analysis of the orb-weaving spider family Araneidae (Arachnida, Araneae). Zool. J. Linn. Soc., 120:355-434. Simon, E. 1892. Histoire naturelle des Araignees. Paris, vol. 1, pt. 1. Pp. 1-256. Vollrath, F. & G.A. Parker, 1997. Reply to Cod- dington et al. Nature, 385:688. Manuscript received 1 May 1998, revised J October 1998. 1999. The Journal of Arachnology 27:64-70 HYPOTHESES FOR THE RECENT HISPANIOLAN SPIDER FAUNA BASED ON THE DOMINICAN REPUBLIC AMBER SPIDER FAUNA David Penney: Invertebrate Zoology, Manchester Museum, University of Manchester, Oxford Road, Manchester Ml 3 9PL United Kingdom ABSTRACT. The Dominican Republic amber fossil spider record is examined and hypotheses generated concerning the Recent Hispaniolan spider fauna which is, at present, poorly known. The families Cyr- taucheniidae, Microstigmatidae, Nemesiidae, Ochyroceratidae, Tetrablemmidae, Palpimanidae, Hersiliidae, Symphytognathidae s.l, Anapidae, Mysmenidae, and Hahniidae, known from the fossil, but not Recent, fauna are predicted to be components of the Recent fauna of Hispaniola. Based on a terrestrial invertebrate species longevity of less than ten million years, the presence of endemic and non-endemic species, and the assumption that Hispaniola has suffered no major ecological disruption that would cause the amber lineages to become extinct, the following hypotheses are made: Filistatidae and Desidae colonized His- paniola after the Miocene amber formation; Drymusidae, Amaurobiidae, and Deinopidae were present on Hispaniola during the Tertiary, but avoided capture, or have yet to be found in the amber; and Scytodidae, Oecobiidae, Uloboridae, Dictynidae and Clubionidae have colonized Hispaniola since the Miocene amber formation but these families, which were present on Hispaniola during the period of amber formation, contain undiscovered endemic species. Hispaniola is unique, in terms of its known spider fauna, in that more families are record- ed from fossil species in Miocene Dominican Republic amber than are recorded from extant species (Wunderlich 1988; Penney 1999). Pe- trunkevitch (1928) considered the Greater An- tillean spider fauna to represent an eastern outgrowth of the Central American fauna by way of a presumed earlier land connection and subsequent continent-island vicariance. How- ever, a land connection appears not to have existed (Ross & Scotese 1988). Based on a quantitative computer model of platetectonics, these authors proposed that the Proto-Greater Antillean (Fig. 1) and subsequently the Great- er Antillean landmass formed on the west of the Proto-Caribbean region during the late Lower Cretaceous. This landmass moved north-eastwards remaining close to the Yuca- tan Peninsula until the Eocene (Figs. 2, 3). During the Late Eocene-Oligocene this land- mass was contiguous with Cuba and Puerto Rico before undergoing island-island vicari- ance (MacPhee & Iturralde-Vinent 1995). There is no evidence of island size change subsequent to this vicariance. During the mid- Tertiary the North and South American land- masses moved westwards relative to the Ca- ribbean. During the period of amber-forming resin secretion (15-20 million years ago; Itur- ralde-Vinent & MacPhee 1996) the Haitian part of Hispaniola lay directly south of and close to the south-eastern part of Cuba. Since then, the separation of the islands has contin- ued until the far-western tip of Hispaniola was clear of the south-easternmost tip of Cuba (Fig. 4) (e.g., Ross & Scotese 1988). Spiders, in general, are renowned for their good dis- persal capabilities and presumably did not re- quire a land bridge in order to colonize His- paniola. There are 291 Recent species in 155 genera and 40 families recorded from Hispan- iola (Penney 1999), but this fauna has not been intensively investigated using a variety of collecting techniques (e.g., Banks 1903; Bryant 1943, 1945, 1948). The fossil fauna consists of approximately 200 species in 46 extant families (Penney 1999). Eleven of the families are recorded only from the fossil fau- na and five are recorded only from the Recent fauna which is an overlap of approximately 70%. Coddington et al. (1991) suggested that one hectare of typical neotropical forest prob- ably supported 300-800 different spider spe- cies, supporting the idea that the Recent His- paniolan spider fauna is poorly known. Species longevity. — Based on observations from the fossil record and/or Lyellian per- 64 PENNEY— RECENT FAUNA HYPOTHESES FROM DOMINICAN AMBER SPIDERS 65 Figures 1-4. — Palaeogeography of Central America and the Caribbean (after Ross & Scotese (1988)). 1, Early Aptian (118.7 Ma); 2, Middle Campanian (84.0 Ma); 3, Late Eocene (44.1 Ma); 4, Recent (0.0 Ma). centages (the percentage of species in the fos- sil record that exist today), Stanley (1985) suggested a few million years for a number of groups of terrestrial animals, whereas plants and some marine animals were found to have longer species durations. Proszyhski (1986), in a footnote, requested information regarding species longevity, and suggested that a salticid species may survive a few million years. Pro- szyhski’s estimate was speculative and based on the disjunct distributions of Recent species which were assumed to have been caused by the Pleistocene glacial and interglacial peri- ods. He has no additional data that would more accurately estimate the species longevity for this family (J. Proszyhski pers. comm. 1998). Decae (1986), however, suggested that two species of Cyrtocarenum Ausserer 1871 (Ctenizidae) may both have a minimum age of 23 million years. Decae (1986) only men- tioned that 23 million years was the minimum age of the species in the abstract of his pub- lication. Evidence was given for a possible speciation event 23 million years ago resulting from vicariance (the separation of the Greek mainland into western and eastern parts divid- ed by an oceanic trough). Testing paleobiogeographic hypothe- ses.— With spiders, the analysis of Recent bio- geographic patterns without evidence from the fossil record can be considered speculative at best. This is demonstrable by the numerous disjunct distributions between Recent spider families and genera and those preserved in the fossil record. Wunderlich (1994) discussed the biogeographic relationships of the extant and fossil central European spiders to the tropical and subtropical faunas. The families Archaei- dae C.L. Koch & Berendt 1854, Deinopidae C.L. Koch 1851 and Cyatholipidae Simon 1894 (the fossils attributed to this family may be incorrectly placed; C. Griswold pers. comm.) were discussed with respect to their fossil (central Europe) and Recent (tropics and southern hemisphere) distributions. The pres- ence of families and genera found in the fossil 66 THE JOURNAL OF ARACHNOLOGY record in central Europe, which today are only found in southern Europe, or are rare in cen- tral Europe, for example, Ctenizidae Thorell 1887, Dipluridae Simon 1889, Leptonetidae Simon 1890, Hersiliidae Thorell 1870, Oeco- biidae Blackwall 1862 and Orchestina Simon 1882 (Oonopidae Templeton 1835) was re- ported by Wunderlich (1994). DecaeN (1986) logic does not consider the ancestral area(s) of the taxa or their ances- tor(s). Three hypotheses for the Recent distri- butions of the Afrotropical genera of the fam- ily Archaeidae may be generated. Vicariance resulted from Madagascar separating from the African mainland. The following evolutionary events may have occurred: 1) the Madagascan genus Archaea C.L. Koch & Berendt 1854 may have evolved from a population of the south African (e.g., Dippenaar-Schoeman & Jocque 1997) genus Afrarchaea Forster & Platnick 1984 (one species A. godfreyi (Hewitt 1919) is found in Madagascar (Lotz 1996), is not endemic, and has probably been intro- duced from South Africa); or 2) vice versa; or 3) both genera evolved from a common an- cestor. Madagascar separated from northern Gond- wanaland and moved southwards to its present position over approximately 150 million years beginning prior to the Middle Jurassic initia- tion of sea-floor spreading in the Western So- mali Basin (Coffin & Rabinowitz 1987). The presence of fossil evidence, i.e., the genus Ar- chaea in Baltic amber (e.g., Eskov 1992) re- jects two of these hypotheses as follows. Mad- agascar is renowned for its unique fauna and flora; it is unlikely that Archaea would colo- nize the Baltic region so far north without also crossing the relatively narrow Mozambique Channel to colonize the African mainland. The fossil evidence and Recent distribution suggests a much wider distribution of this ge- nus in the past (e.g., Eskov & Golovatch 1986) probably prior to the separation of Mad- agascar from the mainland (the family is also recorded from the Jurassic of Kazakhstan (Es- kov 1987)). Thus two of the above hypotheses (1 and 3) are rejected because both genera may have evolved from a common ancestor, but prior to the vicariance event in question. There is no evidence to support the remaining hypothesis that Afrarchaea evolved from Ar- chaea', however, this hypothesis is subject to falsification through the fossil record in the same manner as hypothesis 1. Because fossils of Recent terrestrial animal species have not been found in rocks more than ten million years old, Eldredge (1985) proposed that all species alive more than about ten million years ago are extinct. All species described from Dominican Republic amber are extinct (with possibly a few excep- tions, which warrant re-examination; e.g., Poinar 1992). Therefore a terrestrial inverte- brate species longevity of less than 10 million years is a reasonable expectation. The obvious contraindication to this assumption are those Recent species considered to be living fos- sils’, but these belong to extant clades known in the fossil record to show long and narrow clade shapes, i.e., occupying a long range of geological time and with few branches (Stan- ley 1985). Hypotheses for the Hispaniolan spider fauna. — On the basis of the presence and ab- sence data of spider families in the Dominican Republic amber, the Recent Hispaniolan spi- der fauna, and the Recent Neotropical spider fauna (Table 1 -families known from all fau- nas and with both endemic and non-endemic Recent species not included) it is reasonable to expect that the families Cyrtaucheniidae Pocock 1903, Microstigmatidae Roewer 1942, Nemesiidae Simon 1892, Ochyroceratidae Page 1912, Tetrablemmidae O.P.-Cambridge 1873, Palpimanidae Thorell 1870, Hersiliidae, Symphytognathidae s.l. Hickman 1931, An- apidae Simon 1895, Mysmenidae Simon 1922 and Hahniidae Bertkau 1878, have Recent representatives on Hispaniola which have yet to be discovered. These families are known from the Dominican Republic amber but not from the Recent Hispaniolan fauna, and are components of the Recent Neotropical fauna. Many of the smaller species (e.g., Ochyrocer- atidae, Tetrablemmidae, Symphytognathidae, Anapidae, Mysmenidae, Hahniidae), cryptic species (e.g., Cyrtaucheniidae, Microstigma- tidae, Nemesiidae, Hersiliidae) or less com- mon species (e.g., Palpimanidae) may have been overlooked in the early stages of a spe- cies inventory of Hispaniola, in favor of the larger and more common species. In the in- ventories listed by Bryant (1948) for Cuba, Puerto Rico, St. Vincent and the Virgin Is- lands the above families were represented only by the Hersiliidae recorded from Cuba, PENNEY— RECENT FAUNA HYPOTHESES FROM DOMINICAN AMBER SPIDERS 67 Table L — Dominican Republic amber, Hispaniolan and Neotropical spider families considered in this paper, and the presence of Recent non-endemic and Recent endemic Hispaniolan species in those families. Family Dominican Republic (amber) Recent Hispaniola (non- (endemic) endemic) Recent Neo- tropical Reference Cyrtaucheniidae + - - + Wunderlich (1988) Microstigmatidae + - - + Wunderlich (1988) Nemesiidae + - - Schawaller (1981) Ochyroceratidae + - - + Wunderlich (1988) Tetrablemmidae + - - + Wunderlich (1988) Palpimanidae + - - Wunderlich (1988) Hersiliidae + - - + Wunderlich (1988) Symphytognathidae s.l. + - - + Schawaller (1981) Anapidae + - - + Wunderlich (1988) Mysmenidae + - - + Wunderlich (1998) Hahniidae + - - New amber record Filistatidae - - + Platnick (1993) Desidae - - + Platnick (1993) Deinopidae - - + + Bryant (1948) Drymusidae - + - + Bryant (1948) Amaurobiidae - + + -r Platnick (1997) Scytodidae - + + Wunderlich (1988) Oecobiidae + - + + Wunderlich (1988) Uloboridae + - + + Wunderlich (1988) Dictynidae + - -f + Wunderlich (1988) Clubionidae + — + + Wunderlich (1988) and the Palpimanidae recorded from all but the Virgin Islands. Subsequently, Tetrablem- midae was recorded from Cuba and the Virgin Islands and Anapidae from St. Vincent (Plat- nick 1989); Palpimanidae from the Virgin Is- lands (Platnick 1993); Ochyroceratidae and Mysmenidae from Cuba and St. Vincent (Plat- nick 1997). Endemic vs, non-endemic species. — Some, if not all, of the families known from the Recent, but not amber, Hispaniolan spider fauna (Table 1) may have colonized Hispan- iola since the period of amber-forming resin production in the Tertiary. The only known Hispaniolan filistatid, Kukulcania hibernalis (Hentz 1842), is widespread on the American mainland and the only known desid on His- paniola, Paratheuma insulana (Banks 1902), is found in America and the West Indies (Plat- nick 1993). The only Hispaniolan deinopid, Deinopis lamia MacLeay 1839, is distributed throughout the Antilles; the only Hispaniolan drymusid, Drymusa simoni Bryant 1948, and two amaurobiids: Tugana crassa (Bryant 1948) and Retiro gratus (Bryant 1948) are en- demic to Hispaniola. Tugana cavatica (Bryant 1940) is found on Cuba and Hispaniola (Alay- 6n-Garcia 1992). It is possible that those families containing species endemic to Hispaniola (Drymusidae Simon 1893 and Amaurobiidae Thorell 1870) were present on Hispaniola at the time of the Dominican Republic amber formation al- though this cannot be established unequivo- cally unless they are found in the amber or other fossils from the region. Assuming a species longevity of less than 10 million years, families with only non-en- demic species on Hispaniola (discovered and undiscovered) must have colonized Hispan- iola since the Tertiary amber-forming period or have colonized other regions from Hispan- iola since the Tertiary. It is more likely that most of the families known from only non- endemic species also have undiscovered en- demic species present, particularly those fam- ilies present in Dominican Republic amber, as detailed below. The families Scytodidae Blackwall 1864, Oecobiidae, Uloboridae Tho- rell 1869, Dictynidae O.P. -Cambridge 1871, and Clubionidae Wagner 1887, are recorded in Dominican Republic amber; Filistatidae 68 THE JOURNAL OF ARACHNOLOGY Ausserer 1867, Deinopidae, and Desidae Po- cock 1895, are not. Many Recent genera of Desidae live in the intertidal zones of rocky coasts and may have been present on Hispan- iola during the Miocene but avoided capture in resin because of their habitat. Eskov (1990) reported Filistatidae from the Upper Jurassic of Kazakhstan (but this mate- rial has yet to be published), before the for- mation of the Proto-Greater Antillean land mass. It is probable, then, that Filistatidae col- onized Hispaniola from the American conti- nent, possibly via Cuba. The same may be true for the Desidae, but the Recent distribu- tion of the deinopid species (Greater Antilles) suggests a colonization event originating from within the Greater Antilles, possibly Hispan- iola. Families that have colonized Hispaniola, but which lack endemic species, have proba- bly not been on Hispaniola long enough to speciate; these families must have colonized the island since the amber formation and with- in the last ten million years. Families repre- sented in the Dominican Republic amber and known from the Recent fauna of Hispaniola from only non-endemic species (Table 1) may either have colonized other regions from His- paniola, or have colonized Hispaniola from other regions. Uloboridae and Dictynidae include species with distributions restricted to the Greater An- tilles and these families may have colonized other regions from Hispaniola; the clubionid Elaver exceptus (L. Koch 1866) (possibly pre- sent on Hispaniola) is distributed from Cana- da, through the USA to the West Indies (Plat- nick 1993). On Hispaniola, Scytodidae is known only from pantropical species and, de- spite the lack of evidence, the probability of Hispaniola being scytodid ancestral area is un- likely due to the relatively young age and iso- lated nature of the island, and the cosmopol- itan distribution of Scytodidae. On Hispaniola, Oecobiidae is known from one species Oec- obius concinnus Simon 1892, collected from Port-au-Prince, Haiti; elsewhere in the region it has a distribution throughout the Caribbean islands, Peninsula Florida, coastal Mexico, Central America, Venezuela and Columbia (Shear 1970). Many oecobiids are synaethrop- ic, small, often overlooked, and are frequently inadvertently transported by man. All of the records of this species given by Shear (1970) are from coastal localities, so this was prob- ably the means of dispersal for this species. Only 1 1 specimens are recorded from Hispan- iola (Bryant 1948), compared with the hun- dreds of specimens collected from other re- gions, so this is probably an introduced species. Families known from the Dominican Re- public amber and recorded from the Recent Hispaniolan fauna from only non-endemic, presumably introduced species, unless their amber species lineages have become extinct since the Tertiary, might be expected to con- tain species endemic to Hispaniola that await discovery and description. The only known possible cause of major extinctions on His- paniola since the amber formation might be the Pleistocene glaciations. Hispaniola lay in a tropical-subtropical zone with an arid glacial climate (in part, more arid than at present), and there is good evidence of a cooler Pleis- tocene climate from sedimentary and geomor- phic data and alluvial terraces. However, ex- treme aridity and glaciation have not been documented for the Dominican Republic dur- ing the Pleistocene (Schubert 1988). Whilst the surrounding sea temperature dropped by approximately 2-3 °C during the glacial max- ima (Prell et al. 1976), the albedo of Hispan- iola was the same as it is at present (15=19%). The albedo increased during the last glacial maximum due to the expansion of savannah at the expense of tropical forest; e.g., Panama had a reflectivity of 15=19 percent during the last glacial maximum and at present has a re- flectivity of 10=14% (Schubert 1988). Gri- maldi (1996) presented a reconstruction of the Tertiary Dominican Republic amber-produc- ing forest, based on fossil evidence, which dif- fered little from a Recent Neotropical rainfo- rest. It can be concluded that the Dominican Republic rainforest has suffered no drastic changes since the Tertiary that would cause the spider lineages present in the amber to be- come extinct. Wunderlich (1988) recognized 25 Hispani- olan spider genera recorded only from fossil species. These genera may or may not be ex- tinct. Considering the poorly known nature of the Recent Hispaniolan spider fauna, the lack of these genera in the Recent fauna cannot be construed as evidence for considering these genera extinct; they may contain extant spe- cies which have yet to be discovered. PENNEY— RECENT FAUNA HYPOTHESES FROM DOMINICAN AMBER SPIDERS 69 ACKNOWLEDGMENTS Thanks to R Selden, J. Dunlop and referees for comments on the manuscript and to G. Poinar, Jr. for providing Dominican Republic amber spiders for study. LITERATURE CITED Alay6n=Garcia, G. 1992. El genero Tugana (Arach- nida, Araneae, Amaurobiidae). Poeyana, 416:1-8. Ausserer, A. 1867. Die Arachniden Tirols nach ihr- er horizontalen und verticalen Verbreitung, 1. Verb. Zool.-Bot. Ges. Wien, 17:137-170. Ausserer, A. 1871. Beitrage zur kenntniss der Ar- achniden-Familie der Territelariae Thorell (My- galidae Autor). Verh. Zool.-Bot. Ges. Wien, 21: 117-224. Banks, N. 1902. Some spiders and mites from the Bermuda Islands. Trans. Connecticut Acad. Arts Sci., 11:267-275. Banks, N. 1903. A list of Arachnida from Hayti, with descriptions of new species. Proc. Acad. Nat. Sci. Philadelphia, 55:340-345. Bertkau, R 1878. Versuch einer natiirlichen Anord- nung der Spinnen, nebst Bemerkungen zu ein- gelnen Gattungen. Arch. Naturg. (Berlin), 44: 351-410. Blackwall, J. 1862. Descriptions of newly-discov- ered spiders from the Island of Madeira. Ann. Mag. Nat. Hist., (Sen 3), 9:370-382. Blackwall, J. 1864. A history of the spiders of Great Britain and Ireland, Vol. 2. Ray Society, London, 209 pp. Bryant, E.B. 1940. Cuban spiders in the Museum of Comparative Zoology. Bull. Mus. Comp. Zool., 86:247-554. Bryant, E.B. 1943. The Salticid spiders of Hispan- iola. Bull. Mus. Comp. ZooL, 92:445-521. Bryant, E.B. 1945. The Argiopidae of Hispaniola. Bull. Mus. Comp. Zool., 95:357-442. Bryant, E.B. 1948. The spiders of Hispaniola. Bull. Mus. Comp. Zool., 100:331-459. Coddington, J.A., C.E. Griswold, D.S. Davila, E. Penaranda & S.F. Larcher. 1991. Designing and testing sampling protocols to estimate biodiver- sity in tropical ecosystems. Pp. 44-60, In The Unity of Evolutionary Biology: Proc. Intern. Cong. Syst. Evol. Biol. (E.C. Dudley, ed.). Dios- corides Press, Portland, Oregon. Coffin, M.F. & P.D. Rabinowitz. 1987. Reconstruc- tion of Madagascar and Africa: evidence from the Davie Fracture Zone and Western Somali Ba- sin. J. Geophys. Res., 92:9385-9406. Decae, A.E. 1986. Cyrtocarenum Ausserer, 1871, a living fossil? Pp. 39-44, In Proc. Ninth Intern. Cong. Arachnol. (Eberhard, W.G., YD. Lubin & B.C. Robinson, eds.). Smithsonian Inst. Press, Washington, D.C. Dippenaar-Schoeman, A. & R. Jocque. 1997. Af- rican Spiders, An Identification Manual. Agric. Res. Council, South Africa. 392 pp. Eldredge, N. 1985. Time Frames. Simon & Schus- ter, New York. 240 pp. Eskov, K.Y 1987. A new archaeid spider (Cheli- cerata, Araneae) from the Jurassic of Kazakhs- tan, with notes on the so-called “Gondwanan” ranges of Recent taxa. N. Jb. Geol. Palaont. Abh., 175(1):81-106. Eskov, K.Y. 1990. Spider palaeontology: present trends and future expectations. Acta Zool. Fen- nica, 190:123-127. Eskov, K.Y. 1992. Archaeid spiders from Eocene Baltic amber (Chelicerata, Araneida, Archaeidae) with remarks on the so-called “Gondwanan” ranges of Recent taxa. N. Jb. Geol. Palaont. Abh., 185(3):31 1-328. Eskov, K.Y. & S.L Golovatch. 1986. On the origin of trans-Pacific disjunctions. Zool. Jb. Syst., Jena, 1 13(2):265-285. Fage, L. 1912. Etudes sur les araignees cavemi- coles. 1. Revision des Ochyroceratidae (n. fam.). Biospelogica, 29. Arch. Zool. Exper. et Gen., 10: 97-162. Forster, R.R. & N.I. Platnick. 1984. A review of archeaid spiders and their relatives, with notes on the limits of the superfamily Palpimanoidea (Arachnida, Araneae). Bull. American Mus. Nat. Hist., 178:1-106. Grimaldi, D.A. 1996. Amber: Window to the Past. Harry N. Abrahams, Inc., New York. 216 pp. Hentz, N.M. 1842. Descriptions and figures of the Araneides of the United States. Boston J. Nat. Hist, 4:54-57; 223-231. Hewitt, J. 1919. Description of new South African spiders and solifuge of the genus Chelypus. Rec. Albany Mus., 3:196-215. Hickman, V.V. 1931. A new family of spiders. Proc. Zool. Soc. London, 1931:1321-1328. Iturralde-Vinent, M.A. & R.D.E. MacPhee. 1996. Age and palaeogeographical origin of Dominican amber. Science, Washington, 273:1850-1852. Koch, C.L. 1851. Uebersicht des Arachnidensys- tems. Vol. 5. Nurenberg, 104 pp. Koch, C.L. & G.C. Berendt. 1854. Die im Bern- stein befindlichen Crustaceen, Myriapoden, Ar- achniden und Apteren der Vorwelt. Berlin, 124 pp. Koch, L. 1 866. Die Arachniden-Familie der Dras- siden. Parts 1-6. Nurenberg, 304 pp. Lotz, L.N. 1996. Afrotropical Archaeidae (Ara- neae): 1. New species of Afrarchaea with notes on Afrarchaea godfreyi (Hewitt, 1919). Navors. nas. Mus., Bloemfontein, 12(5): 141-160. Macleay, WS. 1839. On some new forms of Arachnida. Ann. Mag. Nat. Hist. (Ser. 1), 2:1- 14. MacPhee, R.D.E. & M.A. Iturralde-Vinent. 1995. Origin of the Greater Antillean land mammal 70 THE JOURNAL OF ARACHNOLOGY fauna, 1: new Tertiary fossils from Cuba and Puerto Rico. American Mus. Nov., 3141:1-30. Penney, D. 1999. Dominican Republic amber spi= ders and their contribution to fossil and Recent ecology. Unpubl. PhD thesis, Univ. Manchester. 377 pp. Petrunkevitch, A. 1928. The Antillean spider fauna — a study in geographic isolation. Science, Washington, 68(1774):650. Pickard-Cambridge, O. 1871. Arachnida. ZooL Rec., 7:207-224. Pickard-Cambridge, O. 1873. On some new genera and species of Araneida. Proc. ZooL Soc. Lon- don, 1873:112-129. Platnick, N.I. 1989. Advances in spider taxonomy 1981-1987. A supplement to Brigeoli's A Cata- logue of the Araneae described between 1940 and 1981, Manchester Univ, Press, Manchester. 673 pp. Platnick, N.I. 1993. Advances in spider taxonomy 1988-1991 with synonymies and transfers 1940- 1980. New York Entomol. Soc., New York. 846 pp. Platnick, N.I. 1997. Advances in spider taxonomy 1992-1995 with redescriptions 1940-1980. New York Entomol. Soc., New York. 976 pp. Pocock, R.I. 1895. Description of two new spiders obtained by Messrs. J.J. Quelch and F. Mac- Connel on the summit of Mount Roraima, in De- merara; with a note upon the systematic position of the genus Desis. Ann. Mag. Nat. Hist. (Ser. 6), 16:139-143. Pocock, R.I. 1903. On the geographical distribu- tion of spiders of the order Mygalomorphae. Proc. ZooL Soc. London, 1903(l):340-368. Poinar, G.O., Jr. 1992. Life in Amber. Stanford Univ. Press, California. 350 pp. Prell, W.L., J.V. Gardner, A.W.H. Be & J.D. Hays. 1976. Equatorial Atlantic and Caribbean fora- miniferal assemblages, temperatures and circu- lation: interglacial and glacial comparisons. GeoL Soc. America Mem., 145:247-266. Proszyhski, J. 1986. What, if anything, is a genus in Salticidae (Araneae). Pp. 367-372. In Actas X Congresso Intemacional de Aracnologia, Jaca (Espana), VoL 1. (J.A. Barrientos, ed.). Barce- lona. Roewer, C.F. 1942. Katalog der Araneae von 1758 bis 1940, VoL 1. Bremen. 1040 pp. Ross, M.I. & C.R. Scotese. 1988. A hierarchical tectonic model of the Gulf of Mexico and Carib- bean region. Tectonophysics, 155:139-168. S cha waller, W. 1981. Ubersicht iiber Spinnen-Fam- ilien im Dominikanischen Bernstein und anderen tertiaren Harzen (Stuttgarter Bemsteinsammlung: Arachnida, Araneae). Stuttgarter Beitr. Naturk. Ser. B (GeoL und Palaont.), 77:1-10. Schubert, C. 1988. Climatic changes during the last glacial maximum in northern South America and the Caribbean: a review. Interciencia, 13(3): 128-137. Shear, W.A. 1970. The spider family Oecobiidae in North America, Mexico, and the West Indies. Bull. Mus. Comp. ZooL, 140:129-164. Simon, E. 1882. Etudes arachnologiques. 13® Me- moire. 20. Descriptions d'especes et de genres nouveaux de la famille des Dysderidae, Ann. Soc. Entomol. France, 2(6):201~240. Simon, E. 1889. Voyage de M.E. Simon au Ven- ezuela (Decembre 1887-Avril 1888). 4® Me- moire. Ann. Soc. Entomol. France, 9(6): 169- 220. Simon, E. 1890. Etudes arachnologiques. 22® Me- moire. 34. Etude sur les Arachnides de F Yemen. Ann. Soc. Entomol. France, 10(6):77-124. Simon, E. 1892. Histoire Naturelle des Araignees. VoL 1, part 1. Paris, pp. 1-256. Simon, E, 1893. Histoire Naturelle des Araignees. VoL 1, part 2. Paris, pp. 257-488. Simon, E. 1894, Histoire Naturelle des Araignees. VoL 1, part 3. Paris, pp. 489-760, Simon, E. 1895. Histoire Naturelle des Araignees. VoL 1, part 4. Paris, pp. 761-1084. Simon, E. 1922. Description de deux Arachnides cavemicoles do midi de la France. Bull. Soc. En- tomol. France, 15:199-200. Stanley, S.M. 1985. Rates of evolution. Palaeo- biology, 11 (1): 13-26. Templeton, R. 1835. On the spiders of the genus Dysdera Latr. with the description of a new allied genus. ZooL J. London, 5:400-408. Thorell, T. 1869. On European spiders. Part 1. Re- view of the European genera of spiders, preceed- ed by some observations on zoological nomen- clature. N. Act. Reg. Soc. Sci. Upsala, 7(3): 1- 108. Thorell, T. 1870. On European spiders. N. Act. Reg. Soc. Sci. Upsala, 7(3): 109-242. Thorell, T. 1887. Viaggio di L. Fea in Birmania e regioni vicine. 2. Primo saggio sui Ragni bir- mani. Ann. Mus. Civ. Stor. Nat., 5(2):5-4r7. Wagner, W.A. 1887. Copulations organe des Manechens als Criterium fiir die Systematik der Spinnen. Mem. Soc. Entomol. Russia, 22:3-132. Wunderlich, J. 1988. Die Fossilen Spinnen im Dominikanischen Bernstein. Beitr. AraneoL, 2:1- 378. Wunderlich, J. 1994. Bemerkenswerte Spinnen der rezenten und fossilen Faunen Mitteleuropas und ihre biogeographischen Beziehungen zu den Tro- pen und Subtropen (Arachnida, Araneae). Arach- noL Mitt., 7:53-55. Wunderlich, J. 1998. Beschreibung der ersten fos- silen Spinnen der Unterfamilien Mysmeninae (Anapidae) und Erigoninae (Linyphiidae) im Dominikanischen Bernstein (Arachnida, Ara- neae). Entomol. Z., 108(9):363-367. Manuscript received 29 April 1998, revised 10 Jan- uary 1999. 1999. The Journal of Arachnology 27:71-78 AN ADAPTIVE RADIATION OF HAWAIIAN THOMISIDAE: BIOGEOGRAPHIC AND GENETIC EVIDENCE Jessica E. Garb: Zoology Department, University of Hawaii at Manoa, Honolulu, Hawaii 96822 ABSTRACT. The Hawaiian Thomisidae are noted for being extremely species rich, as well as diverse in morphology and ecology. This exceptional diversity led early systematists to place the species into several genera with cosmopolitan distributions. It has been recently suggested that these species compose a single large adaptive radiation. Species-area relationships for all thomisid species and for Misumenops F.O. Pickard-Cambridge 1900 (Thomisidae) species for various island areas were generated. Further, a phylogenetic hypothesis was constructed based on genetic distances between the Hawaiian thomisids and various outgroups using a 450 bp region of the mitochondrial cytochrome oxidase I (COI) gene to test for close genetic relationships. Despite the extraordinary isolation of the Hawaiian islands, the numbers of Misumenops and total thomisid species were found to be significantly higher than predicted for an island system of its size. Phylogenetic analysis of COI suggests the Hawaiian thomisids are more closely related to each other than to representatives of genera to which they have been previously assigned. These results support the existence of a Hawaiian thomisid adaptive radiation, and merit further investigation using comparative methods. The biota of the Hawaiian archipelago is characterized by a number of large species ra- diations. These radiations result from the ex- treme geographic isolation and topographical diversity of the islands (Carson & Clague 1995). Natural colonization of the archipelago has been largely limited to organisms having exceptional dispersal capabilities, and those species that were sucessful colonizers experi- enced total genetic isolation from their source populations. Founders that made the long journey frequently underwent rapid evolution due to drift and selection in extremely small populations (Carson 1994). The Hawaiian Is- lands themselves are volcanic, formed at a “hot spot” connected to the Earth’s core. As the Pacific tectonic plate continually rolls northwestward over the hot spot, new islands are formed. Consequently, they are arranged in chronological order. After initial coloniza- tion of the archipelago, it appears that taxa have frequently progressed down the island chain in a step-wise manner, resulting in re- peated founder events as each new island is colonized (Wagner & Funk 1995). This unique combination of biogeographical events may explain some of the unusual aspects of the composition of the Hawaiian biota. For example, Hawaii’s terrestrial biota is consid- ered depauperate at higher taxonomic levels due to its extreme isolation (Howarth & Mull 1992). However, the islands are exceptionally diverse at the species level, due to extensive autochthonous speciation. The diversity of Hawaiian spiders closely follows this pattern. Of the 105 known spider families (Cod- dington & Levi 1991), only 10 occur naturally in the Hawaiian Islands, and many of these are represented by a single genus (Simon 1900; Suman 1964). Some of these genera have undergone extensive speciation (Gilles- pie et al. 1998). In particular, native species of the genera Tetragnatha Latreille 1831 (Te- tragnathidae), Argyrodes Simon 1864 and Theridon Walckenaer 1805 (Therididae), Mis- umenops F.O. Pickard-Cambridge 1900 (Thomisidae), Sandalodes Keyserling 1883 (Salticidae) and Cyclosa Menge 1866 (Ara- neidae) are known to be unusually diverse and most, if not all, are endemic to the Hawaiian Islands (Table 1). It is likely that many more species within these groups remain to be dis- covered. However, because many species have extremely small ranges of endemism, rapid degradation of natural areas along with in- creasing numbers of alien species, make these spiders exceptionally vulnerable to extinction (Gillespie & Reimer 1993). The family Thomisidae attracted attention early in the study of Hawaiian biology. R.C.L. 71 72 THE JOURNAL OF ARACHNOLOGY Table 1. — Species radiations of Hawaiian spiders. Family Genus Native species % Considered endemic Reference Tetragnathidae Tetragnatha >52 100 Gillespie et at (1998) Therididae Theridion >13 100 Simon (1900) Therididae Argyrodes >30 100 Gillespie et al (1998) Thomisidae Misumenops 17 100 Suman (1970) Araenidae Cyclosa >7 100 Simon (1900) Salticidae Sandalodes >9 88 Simon (1900) Perkins, a pioneering Hawaiian naturalist and collector for the Fauna Hawaiiensis spent lit- tle time in the pursuit of spiders. However, he recognized that “the Thomisidae are probably the most interesting and important group in the Hawaiian spiders” (Perkins 1913). His collections were examined by Eugene Simon (1900), who described 14 new species and a new genus of Thomisidae. These 14 species were placed in the cosmopolitan genera Mis- umena Latreille 1804, Diaea Thorell 1869 and Synaema Fabricius 1775 as well as the new endemic genus Mecaphesa Simon 1900. In a subsequent revision of Hawaiian Thomisidae, Suman (1970) noted that the genitalia of the Hawaiian Diaea was extremely similar to that of the Hawaiian Misumena and these two gen- era were likely the same. Further, examination of the setation and eyes of Hawaiian spiders previously assigned to Diaea and Misumena revealed that they were actually more similar to representatives of Misumenops, a cosmo- politan genus, than to other representatives of Diaea and Misumena. Accordingly, represen- tatives of Diaea and Misumena were placed in Misumenops. Suman considered the endem- ic Mecaphesa to be related to the genus Ozyp- tila Simon 1964, and further suggested that that all of the Hawaiian thomisids were likely derived from three separate colonizers, one that gave rise to the Hawaiian Misumenops, one that gave rise to the endemic genus Me- caphesa, and one that gave rise to the Ha- waiian Synaema (Suman 1970). Suman's re- vision also resulted in the recognition of several new species, for a total of 21 described species, 17 of which are in the genus Misu- menops. Recently, Lehtinen (1993) has suggested that all Hawaiian thomisids comprise an adap- tive radiation generated from a single founder. one in which extensive adaptive morphologi- cal evolution has played a significant role. Ex- amples of explosive adaptive radiations in is- land archipelagos illustrate how closely related species often display great morpholog- ical diversity (Grant & Grant 1989). This phe- nomenon is well documented in other Hawai- ian spiders (Gillespie 1994; Gillespie et at 1997), and other Hawaiian taxa (Roderick & Gillespie 1998). Consequently, if the Hawai- ian thomisids constitute an adaptive radiation, they may represent another exceptional op- portunity for understanding evolutionary and ecological mechanisms that generate rapid di- versification. In this study I examine the species richness of Hawaiian thomisids in a biogeographical framework. Classical biogeographical theory (Mac Arthur & Wilson 1967) states that island species diversity is a balance between immi- gration and extinction. What determines the rate of immigration and extinction depends largely on the size of an island and its distance from a source. In this context one would pre- dict that the isolation and small size of Hawaii would result in a species poor fauna. Here I examine the species-area relationship for Ha- waiian thomisids at the family and generic levels in comparison to other island areas. Also, I conduct a phylogenetic analysis of Ha- waiian thomisid species and representatives of the genera to which they have been assigned historically, in order to provide corroborative evidence for a within- archipelago radiation. METHODS Biogeographical analysis* — A species-area curve was generated for both total thomisid species and for species in the genus Misumen- ops. Data for the distribution of thomisids and Misumenops species have been assessed in garb— EVIDENCE FOR A HAWAIIAN THOMISIDAE RADIATION 73 Table 2. — Species used in sampling of a 450 bp region of mitochondrial cytochrome oxidase L Species Collection locality Misumenops anguliventris Simon 1900 Kauai, Oahu, Maui, Hawaii Island Misumenops cavatus Suman 1970 Hawaii Misumenops discretus Suman 1970 Kauai Misumenops editus Suman 1970 Oahu Misumenops facundus Suman 1970 Hawaii Island Misumenops insulanus Keyserling 1890 Oahu Misumenops imbricatus Suman 1970 Oahu, Maui Misumenops junctus Suman 1970 Molokai Misumenops nigrofrenatus Simon 1900 Kauai, Oahu, Hawaii Island Misumenops vitellinus Simon 1900 Kauai, Oahu, Maui, Hawaii Island Mecaphesa perkinsi Simon 1900 Oahu Mecaphesa semispinosa Simon 1900 Oahu Synaema naevigerum Simon 1900 Maui Synaema globulosum Fabricius 1775 Mt. Carmel, Israel Xysticus sp. Westchester, New York Ozyptila georgiana Keyserling 1880 Westchester, New York Misumena vatia Thorell 1870 Maryland Misumenops asperatus Hentz 1870 Maryland Misumenoides sp. Indiana Diaea sp. New Zealand various island systems worldwide by a num- ber of researchers: Andaman and Nicobar (Ti- kader 1977), Japan (Ono 1988), Britain and Ireland (Roberts 1985), Puerto Rico (Petrunk- evitch 1930), Cuba (Bryant 1940), Tonga (Marples 1959), Fiji (Marples 1957), Pitcairn Islands (Benton & Lehtinen 1995), Philip- pines (Barrion & Litsinger 1995), Samoa (Marples 1957), Rapa (Berland 1924), Maur- itius (Simon 18975), Galapagos Islands (Banks 1902), Tahiti (Berland 1934), and Marquesas Islands (Berland 1927). In the gen- eration of the Misumenops species area curve, Diaea species from Tonga, Fiji and Samoa were substituted for Misumenops, as these were likely incorrectly diagnosed and are closely related to the Hawaiian thomisids (Su- man 1970). Predictions of the total expected number of thomisid species and Misumenops species in Hawaii were calculated based on the species-area relationships generated. Collection and identification of Hawaiian Thomisidae. — Hawaiian thomisid species were collected from native ecosystems on Kauai, Oahu, Maui, Molokai and Hawaii Is- land (Table 2). Specimens were collected by beating vegetation. Sexually mature spiders were identified to species using the Bishop Museum reference collection and Suman’s (1970) key. Additional outgroup species were collected from North America and Israel and others were kindly supplied by Dr. J. Robin- son, Dr. G. Dodson, Cor Vink and Dr. W. Shipley. Collection and analysis of genetic data. — DNA was extracted from the 13 Hawaiian species and 7 non-Hawaiian species listed in Table 2, using a phenol-chloroform prepara- tion followed by ethanol precipitation (Pal- umbi et al. 1991). Voucher specimens were retained and will be deposited in the Bishop Museum, Honolulu, Hawaii. A 450 bp region of the mitochondrial gene cytochrome oxidase I (COI) was amplified for at least two indi- viduals per species using a thermal cycler, with universal primers Cl-J-1718 and Cl-N- 2191 (Simon et al. 1994). The COI gene was selected because it evolves rapidly in the third-codon position (Simon et al. 1994), pro- viding sufficient variation to determine rela- tionships between closely related species. This gene has been useful for assessing genetic di- versity and phylogenetic relationships in other Hawaiian arthropods (Roderick & Gillespie 1998). PCR products were sequenced using an ABI 377 automatic sequencer and were aligned using Sequencher 3.0. Pairwise ge- netic distance (percent nucleotide difference between sequences) was calculated for all spe- cies using the Kimura 2-parameter correction 74 THE JOURNAL OF ARACHNOLOGY Area (km^) Figure 1. — Species-area relationship for total species in the family Thomisidae. Linear regression and a 95% Cl on log-transformed data are shown. Equation of line is log(species) = 0.398 log(area) - 0.917, = 0.70, P < 0.0001. Expected number of species in the family Thomisidae is 7, while 21 species are observed. for nucleotide bias (Kimura 1980), and a phy- logenetic tree was constructed from these dis- tances using a neighbor-joining algorithm in PAUP* 4.0 (Swofford 1998). The Xysticus sp. sequence was used to root the resulting tree. RESULTS Biogeographical analysis. — For the spe- cies-area in the family Thomisidae, a highly significant linear increase was found between the number of species and area of the island system (log-transformations, slope = 0.39, R- = 0.70, see Fig. 1). Perhaps not unexpectedly, Hawaii has an unusually high diversity for its size, falling just above the 95% confidence in- terval for the regression. The slope of the line indicates that an archipelago the area of Ha- waii (28,314 sq. km), should have a total of 7 thomisid species, but 21 occur in Hawaii. The relationship for representatives of the ge- nus Misumenops is quite different (Fig. 2). Whether or not Hawaii is included in the re- gression, there is no significant relationship between island size and number of Misumen- ops species = 0.49, P = 0.43, for 15 is- lands not including Hawaii). By any measure, Hawaii appears to be an extreme outlier com- pared to other island systems examined. For example, Hawaii has the highest observed di- versity of Misumenops species (17), in con- trast to the 1-4 species observed in other is- land systems. The number of Misumenops species on the 15 non-Hawaiian islands ap- peared to be normally distributed, and assum- ing normality, Hawaii falls far above the 99.999th percentile of the distribution. These results suggest that although thomisid diver- sity in the Hawaiian islands is high in general, Misumenops constitutes a disproportionate amount of this diversity. Further, the restric- tion of 17 of Hawaii’s 21 thomisids to a single genus suggests that biodiversity in this island archipelago is a function of autochthonous speciation, rather than external colonization. Analysis of genetic data. — Construction of a distance based phylogeny from 450 bp of COI revealed that the smallest genetic dis- tances were found between Hawaiian thomisid species (Fig. 3). Thus the Hawaiian thomisids appear to be more closely related to each other than they are to representatives of Misumena, Misumenoides, Diaea, Synaema, and Ozyptila. The Hawaiian Synaema and non-Hawaiian S. garb— EVIDENCE FOR A HAWAIIAN THOMISIDAE RADIATION 75 Figure 2. — Species-area relationship for species in the genus Misumenops, along with a simple linear regression on log-transformed data. Equation of line is log(species) = 0.054 log(area) — 0.067, = 0.49, P = 0.43. Expected number of Hawaiian species in the genus Misumenops is 0.6, while 17 species are observed. globulosum are extremely distant while the Hawaiian Synaema is comparatively much more similar to the Hawaiian Misumenops, supporting Lehtinen’s (1993) claim that the Hawaiian 'Synaema' is not a true Synaema, Small genetic distances between non-Hawai- ian Misumenops asperatus and all Hawaiian Misumenops species sampled suggests that the Hawaiian species are likely descended from a Misumenops ancestor. However, because Mis- umenops asperatus falls within the clade con- taining all Hawaiian thomisids, more than one colonization event is implied for the Hawaiian thomisid fauna. Sampling of many more Mis- umenops species from other geographic re- gions must be conducted in order to firmly establish what proportion of Hawaii's present day thomisid diversity has arisen thi'ough au- tochthonous species radiation. DISCUSSION The high diversity of Misumenops species in Hawaii, coupled with the short genetic dis- tances between species, is a pattern similar to that found in other Hawaiian terrestrial arthro- pods that have undergone species radiation (Roderick & Gillespie 1998). While the is- lands are exceptionally diverse in both thom- isid and Misumenops species, the Misumenops species represent nearly all of the thomisid di- versity. It appears, therefore, that the number of successful colonizers is consistent with the island biogeography model. However, the model does not account for subsequent pro- cesses that occur over evolutionary time. Thus, despite few colonization events, autoch- thonous speciation raises the species diversity to levels higher than predicted by island area. Such a pattern might suggest that most of the available niche space for the family is being occupied by species in the genus Misumenops. Colonizing species may have undergone rapid ecological divergence in the absence of com- petition from other genera that exploit a va- riety of ecological strategies elsewhere. Although Hawaiian Misumenops species are extremely diverse (Fig. 2), it is possible that this result may be biased by data points representing inaccurate estimates of diversity. Many of the data points used to generate the relationships between area and diversity were extracted from surveys of unequal sampling effort. Further, current revisions of these is- lands’ spider fauna ought result in different generic diagnosis of species. Consequently, 76 THE JOURNAL OF ARACHNOLOGY Xysticus sp., Westchester, New York Ozyptila georgiana, Westchester, New York Synaema globulosum, Mt, Carmel. Israel Diaea sp, , New Zealand Misumenoides sp. , Maryland Misumena vatia, Maryland Misumenops cavatus, Hawaii, Hawaii Synaema naevigerum, Maui, Hawaii Misumenops asperatus, Maryland Mecaphesa perkinsi, Oahu, Hawaii Misumenops discretus, Kauai, Hawaii Misumenops junctus, Molokai, Hawaii Misumenops vitellinus, Oahu, Hawaii Misumenops nigrofrenatus, Kauai, Hawaii Misumenops imbricatus, Oahu, Hawaii Misumenops facundus, Hawaii, Hawaii Misumenops editus, Oahu, Hawaii Misumenops insulanus, Oahu, Hawaii Misumenops anguliventris, Kauai, Hawaii Mecaphesa semispinosa, Oahu, Hawaii Figure 3. — A distance based phytogeny constracted from 450 bp of COL Genetic distances were caL culated using Kimura’s (1980) 2-parameter corrected distance measure. many of these points may represent over or under estimates of true diversity. It is possible that a more complete data set would reveal that Hawaii is not overly diverse in Misumen- ops or total thomisid species. However, Japan (the most thoroughly studied island group), has an area 10 times the size of Hawaii yet only twice the number of thomisid species (Ono 1988). Moreover, only three of these species are Misumenops. The relatively small genetic distances among Hawaiian species are consistent with the hypothesis that the high morphological and ecological diversity of Ha- waiian thomisids is primarily the result of au- tochthonous radiation. The Hawaiian Misumenops possess excep- tional ecological diversity when compared to their continental congeners. Members of the family Thomisidae are well known for their employment of mimicry to ambush prey. The tribe Misumenini Simon 1895, including spi- ders in the genera Misumena, Misumenoides, and Misumenops (Ono 1988), are commonly known as flower spiders because they mimic the coloration of flowers on which they sit in order to capture pollinating insects (Gertsch 1979). Recent research reveals that certain species of this tribe have ecological roles as nectivores and possible pollinators (Pollard et al. 1995). While genetic data suggest that the Hawaiian thomisids are descendants of flower spiders, Hawaiian thomisids are extremely di- verse in their substrate affinities. For example, the green and brown speckled Misumenops editus, endemic to the summit of Mt. Kaala, Oahu, is perfectly camouflaged amongst its primary microhabitat of moss patches. Like- wise, M. aridus and M. nigrofrenatus are well hidden on their substrate of white filamentous lichen. Other species are more specific to garb— EVIDENCE FOR A HAWAIIAN THOMISIDAE RADIATION 77 green foliage. Many Hawaiian Misumenops that are ecologically separated as a result of differential substrate affinity are sympatric and may be very close relatives (Sumae 1970). In order to appreciate fully the extent of species radiation in the Hawaiian Misumen- ops, a complete phylogenetic construction is warranted. Only with a well- supported phy- logenetic hypothesis can the idea of species radiation among the Hawaiian thomisids be tested rigorously. From such a hypothesis, one can determine the number of colonization events and the ecological diversity of this group. It is clear from data presented here that the Hawaiian Misumenops do not fit the clas- sical model for island biogeography as they are species rich in an isolated and small island area. ACKNOWLEDGMENTS I thank Rosemary Gillespie, George Rod- erick, A. Bohonak, and two anonymous re- viewers for help in preparation of this manu- script, B. Thorsby, M. Rivera, and J. Gubili for laboratory assistance, M. Heddle, C. Ew- ing, B. Thorsby, M. Rivera, A. Vandergast, D. Preston, A. Asquith, A. Medieros and E. Lar- son for assistance in collecting Hawaiian spi- ders, W. Shipley, C. Vink, J. Robinson and G. Dodson for shipment of non-Hawaiian thom- isids, S. Swift at the Bishop Museum for ac- cess to collections, the managers of Volcanoes and Haleakala national parks, Hawaii State Dept, of Fish and Wildlife reserves and The Nature Conservancy of Hawaii’s preserves for permission to collect thomisids. This work is partly supported by a student research grant from the University of Hawaii Ecology, Evo- lutionary and Conservation Biology Program and from A AS, as well as NSF grants to R. Gillespie. LITERATURE CITED Banks, N. 1902. Papers from the Hopkins Stanford Galapagos Expedition. Proc. Washington Acad. Sci., 4:49-86. Barrion, A.T. & J.A. Litsinger. 1995. Riceland spi- ders of South and Southeast Asia. Center for Ag- riculture and Bioscience, Wallingford, UK, Benton, TG, & P.T Lehtinen. 1995. Biodiversity and origin of the non-flying terrestrial arthropods of Henderson Island. Bio. J. Linn. Soc., 56(1-2): 261-272. Berland, L, 1924. Araignees de File de Paques et des lies Juan Fernandez. The Natural History of Juan Fernandez and Easter Island, 3:419-437. Berland, L. 1927. Notes sur les Araignees recueil- lies aux eles Marquises par le R.P. Simen Del- mas. Bull. Du Museum National d’Histoire Na- turelle, Paris, 38:366-368. Berland, L. 1934. Les Araignees de Tahiti. Bull. Bernice P. Bishop Museum, 113:97-107. Bryant, E.B. 1940. Cuban spiders in the Museum of Comparative Zoology, Bull. Mus. Comp. ZooL, 86(7):247-554. Carson, H.L, 1994. Genetical processes of evolu- tion of high oceanic islands. Pp. 259-262. In A Natural History of the Hawaiian Islands. (E.A. Kay, ed.). Univ. Hawaii Press, Honolulu, Hawaii. Carson, H.L & D.A. Clague. 1995. Geology and biogeography of the Hawaiian Islands. Pp. 14- 29. In Hawaiian Biogeography, Evolution on a Hot Spot Archipelago. (W.L. Wagner & V.A. Funk, eds.), Smithsonian Institution Press, Wash- ington, D.C. Coddington, J.A. & H.W. Levi. 1991. Systematics and evolution of spiders (Araneae). Ann. Rev. EcoL Syst., 22:565-592. Gertsch, W.J. 1979. American Spiders. 2Ed ed. Van Nostrand Reinhold Co., New York. Gillespie, R.G. 1994. Hawaiian spiders of the ge- nus Tetragnatha: III. Tetragnatha acuta clade. J. ArachnoL, 22:161-168. Gillespie, R.G., H.B. Croom & L. Hasty. 1997. Phylogenetic relationships and adaptive shifts among major clades of Tetragnatha spiders (Ar- aneae: TetragnatMdae) in Hawaii. Pacific Sci., 51(4):380-394. Gillespie, R.G. & N. Reimer. 1993. The effects of alien predation of ants (Hymenoptera: Formici- dae) on the Hawaiian endemic spiders (Araneae: TetragnatMdae). Pacific Sci., 47(l):21-33. Gillespie, R.G., M.A. Rivera & J.E. Garb. 1998. Taxonomy and phylogeography of Hawaiian Ar- aneae. Bull. British ArachnoL Soc., 11:41-51. Grant, B.R,& P.R. Grant. 1989. Evolutionary Dy- namics of Natural Populations. Univ. Chicago Press, Chicago. Howarth, EG. & W.P. Mull. 1992. Hawaiian In- sects and Their Kin. Univ. Hawaii Press, Hono- lulu. Kimura, M. 1980. A simple method for estimating evolutionary rate of base substitution through comparative studies of nucleotide sequences. J. Mol. Evol. 2:87-90. Lehtinen, P.T. 1993. Polynesian Thomisidae — A meeting of old and new world groups. Mem. Queensland Mus., 33(2):585-591. MacArthur, R.H. & E.O. Wilson. 1967. The The- ory of Island Biogeography. Monog. Population Biology L Princeton Univ. Press, Princeton. Marples, B.J. 1957. Spiders from some Pacific is- lands II. Pacific Sci., 11:386-395. 78 THE JOURNAL OF ARACHNOLOGY Marples, B.J. 1959. Spiders from some Pacific is- lands III. The Kingdom of Tonga. Pacific Sci., 13:362-367. Ono, H. 1988. A revisional study of the spider family Thomisidae of Japan. National Science Museum, Tokyo, Japan. Palumbi, S.R., A. Martin, S. Romano, W.O. McMillian, L. Stice & G. Grabowski. 1991. The simple fool’s guide to PCR. Dept, of Zoology, University of Hawaii, Honolulu. Perkins, R.C.L. 1913. Introduction. Pp. 15-228. In Fauna Hawaiiensis, Vol. 1. (D. Sharp, ed.) Cam- bridge Univ. Press, Cambridge. Petrunkevitch, A. 1930. The spiders of Puerto Rico. Trans. Connecticut Acad. Sci., 30:1-355; 31:1-191. Pollard, S.D., W.M. Beck & G.N. Dodson. 1995. Why do male crab spiders drink nectar? Anim. Behav., 49:1443-1448. Roberts, M.J. 1985. The Spiders of Great Britain and Ireland. Vol. 1. Atypidae-Theridiosomati- dae, 229 pp. Harley Books, Colchester, UK. Roderick, G.K. & R.G. Gillespie. 1998. Speciation and phylogeography of Hawaiian terrestrial ar- thropods. Mol. Ecol., 7:309-322. Simon, C., E Fratti, A. Beckenbach, B. Crespi, H. Liu & P. Flook. 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene se- quences and a compilation of conserved poly- merase chain reaction primers. Ann. Entomol. Soc. America., 87:651-701. Simon, E. 1897. Arachnological records par Al- luaud a File Maurice. Ann. Soc. Entomol. France, 66:271-276. Simon, E. 1900. Arachnida. Pp. 443-519. In Fauna Hawaiiensis, Vol. 2. (D. Sharp, ed.), Cambridge Univ. Press, Cambridge, UK. Suman, TW. 1964. Spiders of the Hawaiian Is- lands: catalog and bibliography. Pacific Insects, 6:665-687. Suman, TW. 1970. Spiders of the Family Thom- isidae in Hawaii. Pacific Insects, 12(4):773-864. Swofford, D.L. 1998. PAUP*. Phylogenetic Anal- ysis Using Parsimony (*and other methods). Ver- sion 4. Sinauer Associates, Sunderland, Massa- chusetts. Tikader, B.K. 1977. Studies of the spider fauna of Andaman and Nicobar Islands, Indian Ocean. Rec. Zool. Surv. India, 72:153-212. Wagner, WL. & V.A. Funk. 1995. Hawaiian Bio- geography. Evolution On a Hot Spot Archipela- go. Smithsonian Institution Press, Washington, D.C. Manuscript received 1 May 1998, revised 4 Novem- ber 1998. 1999. The Journal of Arachnology 27:79-85 COMPARISON OF RATES OF SPECIATION IN WEB-BUILDING AND NON-WEB-BUILDING GROUPS WITHIN A HAWAIIAN SPIDER RADIATION Rosemary G. Gillespie: Department of Zoology & Center for Conservation Research and Training. University of Hawaii, 3050 Maile Way, Honolulu, Hawaii 96822 USA ABSTRACT. The isolation of the Hawaiian archipelago has resulted in a fauna that shows high levels of endemism. I examined the role of lifestyle, as inferred from web-building versus non-web-building behavior, in dictating the rate of differentiation and species formation within a lineage of spiders in the genus Tetragnatha from the Hawaiian Islands. This genus comprises a group of morphologically, ecolog- ically and behaviorally diverse taxa. Included in the radiation is a ‘spiny-leg’ clade which never builds webs and is relatively loosely associated with a specific habitat, and a large group of web-building species which are generally more tightly associated with a given substrate and habitat. Sequences of mitochondrial cytochrome oxidase DNA provided relative estimates of the age of a clade. Both linear and logarithmic models were used to estimate rates of speciation and the relative time required for speciation for each clade. The results showed that several small clades of web-building species have a greater rate of speciation as compared to the ‘spiny-leg’ clade. One explanation is that the web-building species may be capable of differentiation between more closely contiguous habitats, which would be consistent with the hypothesis that ecological differentiation promotes diversification and species formation. Possible alternative expla- nations for the results include differences in rates of molecular evolution, for example as a consequence of differences in metabolic activity. The Hawaiian archipelago provides a nat- ural laboratory for studies of speciation (Si- mon 1987). First, the extreme isolation of the islands has caused accentuation and accelera- tion of evolution in the archipelago, with nu- merous examples of rampant species prolif- eration (for reviews see Wagner & Funk 1995; Roderick & Gillespie 1998). Further, the islands are a series of volcanoes arranged within an identifiable chronological time frame (Carson & Clague 1995). The currently high islands range from Kauai, the oldest and most eroded, to Hawaii, the youngest, highest, and largest, with five separate volcanoes. Each volcanic mountain therefore shows a different stage in the evolutionary history of a clade, and allows determination of the nature of the relationship between evolutionary time and the abundance and distribution of a set of spe- cies. Radiations of spiders in the Hawaiian Is- lands include the genera Tetragnatha (as de- scribed below), Argyrodes (Simon 1900), Theridion (Simon 1900), and species in the family Thomisidae (Simon 1900; Suman 1970; J.E. Garb, this volume), among others (Gillespie et al. 1998). I have been examining patterns of specia- tion (Gillespie 1991a, b, 1992a, Gillespie 1993; Gillespie & Croom 1995; Gillespie et al. 1994) and extinction (Gillespie 1992b, Gil- lespie & Reimer 1993) in a radiation of Ha- waiian spiders in the long-jawed orb- weaving genus Tetragnatha (Tetragnathidae). Outside the archipelago, Tetragnatha is of worldwide distribution (Levi 1981), yet it is also one of the most homogeneous genera of spiders, in both morphology (elongate form [Kaston 1948]) and ecology (Caraco & Gillespie 1986; 1987a, b; Gillespie & Caraco 1987). Until 1991, information on the endemic Hawaiian tetragnathids was based on descriptions of only nine species (Karsch 1880; Simon 1900) in the genus Tetragnatha and one Doryony- chus. Over the last 11 years I have collected native Tetragnatha in every native habitat type on all of the Hawaiian Islands. I have now described an additional 25 species of Ha- waiian Tetragnatha (Gillespie 1991a, 1992a, 1994) and more than 60 additional new taxa have been collected, of which descriptions for many are near completion. The tetragnathid radiation spans a tremendous spectrum of col- 79 80 THE JOURNAL OF ARACHNOLOGY ors, shapes, sizes, ecological affinities, and be- haviors (Gillespie 1991a, b; 1992a, b; 1994). Many species are web-building, with striking patterns, colors and structural modifications of the abdomen that allow concealment within specific microhabitats. Some species have structural modifications which appear to allow specialization on specific prey types. One en- tire clade (‘spiny-leg’ clade) has abandoned web building, with the concomitant develop- ment of long spines on the legs and adoption of a cursorial predatory strategy (Gillespie 1991a). This clade was originally described as 12 species. Recent molecular evidence indi- cates that an additional six species belong in this clade. For the ‘spiny-leg' clade I have generated hypotheses of evolutionary relationships among some species (Gillespie 1993; Gilles- pie & Groom 1995; Gillespie et al. 1994; Gil- lespie et al. 1997) and among populations within widespread species (Gillespie & Rod- erick unpubl. data.) based on molecular and morphological characters. I have also begun to study phylogenetic relationships among groups of web-building species, in which the component species are more sedentary and tend to be more tightly associated with a given habitat (Gillespie & Groom 1995; Gillespie et al. 1997). Here, I quantify the relationship between the lifestyle of a clade and the rate of speci- ation. I focus on two monophyletic lineages, the 'spiny-leg' clade and a large web-building group, because of the contrast in vagility that their species display. The majority of the web- building species have become extreme habitat specialists, while the ‘spiny-leg' species have abandoned web-building to become cursorial predators and tend to be less specific in habitat preference (Gillespie & Groom 1995). I use molecular sequence data to compare patterns of speciation among web-building and non- web-building lineages on different islands. I then examine relative rates of sequence diver- gence to determine the extent to which life- style (i.e., web-building or cursorial) is asso- ciated with the rate of speciation and relative time to species formation. Because the web- building species tend to be more tied to spe- cific habitat types, the influence of disruptive selection may be enhanced relative to non- web-building species, and new species may form more rapidly (Thoday 1972; Bush & Howard 1986). METHODS DNA sequences.— ”DNA sequence data have been generated based on part of the cy- tochrome oxidase subunit I (COI) mitochon- drial DNA for almost all known representa- tives of Hawaiian Tetragnatha, and mitochondrial 12S ribosomal DNA sequences and allozyme data have been obtained for some representatives (Gillespie et al. 1997; R.G. Gillespie unpubl. data). For the current study I present molecular data from COI mi- tochondrial DNA only. A 450 base pair piece of COI was amplified for a minimum of two individuals of each species using primers Cl- J-1718 and Cl-N-2191 (designed by R. Har- rison lab, Simon et al. 1994). Amplification was done with the following profile: 94 °C (60 sec), 48 °C (35 sec) and 72 °C (45 sec) for 40 cycles. Automated cycle sequencing was used to run and score the sequencing gels (ABI 377). Each sequence plot was inspected in Se- quencher 3.1 (Gene Codes Corporation 1998). Alignments, which are straightforward for this protein-coding region, were performed by eye. Species sampling* — The current study ex- amines relative rates of speciation. According- ly, I had to identify monophyletic clades, with all representatives included, prior to the anal- ysis. For each of the clades chosen, the spe- cies group is recovered consistently by max- imum parsimony and maximum likelihood analyses, and has bootstrap support of > 50% (most much higher; R.G. Gillespie unpubl. data). For the current study I obtained se- quence data as follows: ‘Spiny-leg' clade: I obtained additional data from COI so as to have sequences for each of the following spe- cies, all of which have been described: T. pi- losa Gillespie, T. kauaiensis Simon, T. per- reirai Gillespie, T, tantalus Gillespie, T. polychromata Gillespie, T. waikamoi Gilles- pie, T. brevignatha Gillespie, T macracantha Gillespie, T. restricta Simon, T. kamakou Gil- lespie and T. quasirnodo Gillespie. Sequences were also obtained for an additional six rep- resentatives of the ‘spiny-leg' clade which are undescribed. These 17 species encompass all known representatives of the ‘spiny-leg’ clade except for T. mohihi from Kauai, a small spe- cies that is known only from a single male (Gillespie 1991a). GILLESPIE— SPECIATION RATES IN HAWAIIAN TETRAGNATHA 81 Web-building species: I examined all known web-building species, most of which are undescribed. However, the analysis of spe- ciation rates focused only on groups with > 50% bootstrap support (most much higher). One clade (54% bootstrap support) is from Maui and includes three described species: T. stelarobusta Gillespie, T. tritub erculata Gil- lespie and T. filiciphilia Gillespie. A second clade (89% bootstrap support) is from Oahu, and includes four undescribed species which are readily identified on the basis of mor- phology (gross morphology and genitalic structure, Gillespie unpubl. data) and ecology: “Elongate Palikea” (Waianae Mountains, mid-elevation mesic forest), “Elongate Tan- talus” (Koolau Mountains, mid-elevation me- sic forest), “Slender Elongate” (Waianae Mountains, low elevation dry forest) and “Elongate Kaala” (Waianae Mountains, high elevation wet forest). A third clade (95% bootstrap support) is found on Kauai (“Elon- gate Kauai”), high elevation habitats on Maui (“Elongate Maui Crater”) and Hawaii (“Elongate Mauna Kea”), and mid-elevation wet forest on Hawaii (“Elongate Hawaii Sad- dle”). Phylogenetic reconstruction. — Phyloge- nies were reconstructed using maximum like- lihood as implemented in PHYLIP (Felsen- stein 1993). This method can accommodate the heavy AT bias in the nucleotide compo- sition (approximately 65% of the bases were A or T), using base frequencies estimated from the data. Maximum parsimony (PAUP* 4.0, Swofford 1998) was also used to test the robustness of the phylogenetic reconstruction. Trans versions were weighted 4X transitions, roughly approximating the greater frequency of base changes that involved transitions (4X) between closely related species. Branches having maximum length zero were collapsed to yield polytomies. Rates of speciation. — I used both linear and logarithmic models to estimate speciation rates as described by McCune (1998). Under the linear model, which assumes a “comb- shaped” tree, the rate of speciation (SRun) is calculated as: SRii„ = nit, where n is the number of known species in a monophyletic clade and t is the age of the clade. The time required for speciation (TFS,J is: TFS,i„ = tld = ti{n-l). Under the logarithmic model, which assumes a symmetrical tree, the rate of speciation (SRJ is: Sr,n = In niXxil, with the time required for speciation (TFS,n) as: TFS,n = tid ^ t ln2/lnn. RESULTS Molecular phylogenetic analysis. — For the range of genetic distances encompassing the major radiation of Hawaiian Tetragnatha both transitions and transversions increased linearly when plotted against Tamura distance (Tamura 1992) suggesting that both transitions and transversions are phylogenetically infor- mative at this level and that the data, even at third positions, are not phylogenetically satu- rated (Gillespie et al. 1997). For the greater distances between the major Hawaiian radia- tion and species in the ‘J. hawaiensis' clade (a separate introduction into Hawaii, Gillespie et al. 1994) transversions are still informative, although transitions show evidence of satura- tion. For the major radiation, sequences for the 450-base-pair homologous region were compared and variation was found at 204 dif- ferent sites, with 111 being phylogenetically informative. Phylogenetic relationships among represen- tatives of the ‘spiny-leg’ clade of Hawaiian Tetragnatha based on COI sequences for each of the described species are shown in Fig. lA. The relationships are similar to those de- scribed previously (Gillespie et al. 1997). The six recently described species all form a monophyletic group with T. quasimodo. How- ever, because relationships between taxa in this clade were not resolved, these species are not included in the analysis of speciation. Re- lationships between species within the three selected web-building clades are shown in Fig. IB. Rates of speciation. — For each island group in the ‘spiny-leg’ clade, I summed the number of base changes from the base of the clade (marked * in Fig. lA for each group) to the branch tip for each species in the clade. Then, for each clade, the average number of base changes for species within a clade was divided by the total number of bases to give the percent sequence divergence from the base of the clade to the present. An estimate of 2% A, Spiny Leg Clade B. Different Clades of Web-builders 82 THE JOURNAL OF ARACHNOLOGY Figure 1. — Estimates of phylogeny for (A) representatives of the ‘spiny-leg’ clade of Hawaiian Tetragnatha (Gillespie et al. 1997) and (B) Different clades of web-building species. Numbers below branches indicate actual numbers of bases changes associated with that branch. * indicates the branch that was used as the “base” for each clade, and from which the numbers of base changes to the terminal nodes was counted. GILLESPIE— SPECIATION RATES IN HAWAIIAN TETRAGNATHA 83 Table 1. — Age of each clade, number of species, speciation rates, and times to speciation for: (A) the ‘spiny-leg’ clade, and (B) three different clades of web-building species, of Hawaiian Tetragnatha. * calculated assuming 2% sequence/10^ years (DeSalle et al. 1987). ^ represents divergence from Oahu. Island Kauai A Oahu Maui Kauai- Hawaii B Oahu Maui Age (myrs) 5.1 3.7 1.9 5.1-0 3.7 1.9 Number of species 2 3 5 4 4 3 Av. % seq divergence 4.00 4.45 5.03 3.16 1.61 2.37 Age of clade (myrs)* 2.0011 2.23 2.51 1.58 0.81 1.19 SRlin LOO 1.35 1.99 2.54 4.97 2.53 TFSlin 2.00 1.11 0.63 0.53 0.27 0.59 SRln 0.35 0.49 0.64 0.88 1.72 0.93 TFSln 2.00 1.40 1.08 0.79 0.40 0.75 base change/ 10^ years was used to assess the age of a clade (DeSalle et al. 1987). This is an approximate measure, and is likely to de- viate from the actual age of the clade because of the inconstancy of the molecular clock (J.H. Gillespie 1986). Accordingly, the age of the clade is used here as a relative measure only. Because the current study focuses on the comparison of two clades within the same ge- nus in comparable habitats, base change dif- ferences between the two clades are likely to provide reliable relative estimates of differ- ences in evolutionary rates. The age of each clade, the number of species, and both linear and logarithmic estimates of rates of specia- tion and relative times to speciation are given in Table lA. Similar estimates were generated for the three clades of web-building species described above (Table IB). Rates of specia- tion (both linear and logarithmic) were con- sistently higher for web-building species, and times to speciation were consistently lower for web-building species; these differences were all significant (Mann- Whitney C/-test, P < 0.05). DISCUSSION The results show that the patterns of spe- ciation relative to island age are similar in both ‘spiny-leg’ and web-building species groups: In both cases, speciation appears to have occurred largely within an island (Gil- lespie et al. 1997). However, the rate of spe- cies formation for the web-building clades contrasts with that of the ‘spiny-leg’ clade. In particular, differentiation between taxa within each of the web-building clades appears to oc- cur much more rapidly. The higher rate of species formation in this clade may arise part- ly from their web-building habit. Based on current theories, groups that are only loosely associated with habitat types, such as the Ha- waiian Drosophila and the cursorial ‘spiny- leg’ clade of Hawaiian Tetragnatha, may re- quire longer periods of isolation in order to initiate divergence (Mayr 1963; Carson 1986; Bush & Howard 1986). However, groups comprising taxa with more rigorous ecological associations could diverge more rapidly be- tween contiguous habitats through the action of forces such as disruptive selection (Rausher 1984; Rosenzweig 1990). Spiders that build webs tend to demonstrate stronger habitat af- finities than cursorial species (Gillespie & Croom 1995), and consequently may be ca- pable of differentiating more rapidly. There are alternative explanations that might account for the differences in relative rates of species formation between the ‘spiny- leg’ and web-building species. In particular, both metabolic rate and generation time are known to affect rates of molecular evolution (Martin & Palumbi 1993): Higher metabolic rate and shorter generation time cause accel- eration of molecular evolution. Although we have no evidence to suggest differences in generation time between the ‘spiny-leg’ and web-building species, it may be that the met- abolic rate is higher among representatives of the more active ‘spiny-leg’ clade. The age of the islands can be compared with estimates of the ages of the different clades based on sequence divergence for the 84 THE JOURNAL OF ARACHNOLOGY 'spiny-leg' clade, assuming a constant substi- tution rate of 2% per million years (DeSalle et al. 1987; Juan et al. 1996) (Table lA). The Kauai and Oahu clades might be expected to match the age of Oahu (3.7 myrs), as the for- mation of this island would have allowed di- vergence to be initiated. However, these two clades are considerably younger (2.0 and 2.2 myrs), suggesting perhaps that colonization and divergence started well after the forma- tion of Oahu, or that molecular evolution is more rapid than that of other arthropod taxa for which calibrations have been made. On the other hand, divergence of the Maui species in the 'spiey-leg" clade are considerably older than the oldest of the islands in the Maui Nui complex (Molokai, Maui, Lanai and Kahoo- lawe). This result suggests that divergence of the Maui Npiey-leg' clade was initiated prior to the colonization of Maui. Alternatively, rates of evolution may vary more than ex- pected, and provide only very poor estimates of the age of a clade (J.H. Gillespie 1986; Kambhampati & Rai 1991). Within a lineage, acceleration in rates of sequence divergence may be associated with the formation of new species (Carson & Templeton 1984). Accord- ingly, the greater apparent age of the Maui species may be merely a reflection of the greater number of species. Whether or not the DNA sequences provide an indication of actual age of a clade does not affect the major conclusions of this study. Here, the estimates of sequence divergence are used as a relative measure only, to com- pare groups of spiders that differ in lifestyle (web-building or non-web-building). The re- sults suggest that there are much smaller ge- netic distances involved in species formation in web-building as compared to non-web- building species of Hawaiian Tetragnatha. It appears, therefore, that lifestyle, as indicated by the web-building habit, dictates in part the rate at which differentiation and divergence can occur within a lineage, ACKNOWLEDGMENTS The work reported here was supported by grants from the National Science Foundation and the University of Hawaii Research Coun- cil. I thank J. Gubili for technical assistance and G. Roderick for helpful comments and suggestions on the manuscript. I am indebted to the following individuals for facilitating field work: Art Medeiros and Lloyd Loope (Haleakala National Park), Ed Misaki and Mark White (Nature Conservancy of Hawaii), Randy Bartlett (Maui Land and Pineapple), and David Preston (Bishop Museum). LITERATURE CITED Bush, G.L. &. DJ. Howard. 1986. Allopatric and non-allopatric speciation: assumptions and evi- dence. Pp. 411-438. In Evolutionary Processes and Theory (S. Karlin & E. Nevo, eds.). Aca- demic Press, New York. Caraco, T. & R.G. Gillespie. 1986. Risk sensitivity: foraging mode in an ambush predator. Ecology, 67:1180-1185. Carson, H.L. 1986, Sexual selection and specia- tion. Pp. 391-409. In Evolutionary Processes and Theory (S. Karlin & E. Nevo, eds.). Aca- demic Press, New York. Carson, H.L. & D.A. Clague. 1995. Geology and biogeography of the Hawaiian Islands. Pp. 14- 29. In Hawaiian Biogeography Evolution on a Hot Spot Archipelago (W.L. Wagner & V.A. Funk, eds.). Smithsonian Institution Press, Wash- ington, D.C. Carson, H.L. & A.R. Templeton. 1984. Genetic revolutions in relation to speciation phenomena: The founding of new populations. Ann, Rev. EcoL Syst., 15:97-131. DeSalle, R., T. Freedman, E.M. Prager & A.C. Wil- son. 1987. Tempo and mode of sequence evo- lution in mitochondrial DNA of Hawaiian Dro- sophila. J. Mol. EvoL, 26:157-164. Felsenstein, J. 1993. PHYLIP. University of Wash- ington, Seattle. Gillespie, J.H. 1986. Variability of evolutionary rates of DNA. Genetics, 113:1077-1091. Gillespie, R.G. 1987a. The mechanism of habitat selection in the long jawed orb weaving spider Tetragnatha elongata (Araneae, Tetragnathidae). J. ArachnoL, 15:81-90. Gillespie, R.G. 1987b. The role of prey in aggre- gative behaviour of the long jawed orb weaving spider Tetragnatha elongata. Anim. Behav., 35: 675-681. Gillespie, R.G. 1991a. Hawaiian spiders of the ge- nus Tetragnatha: 1. Spiny Leg Clade. J. Arach- noL, 19:174-209. Gillespie, R.G. 1991b. Predation through impale- ment of prey: The foraging behavior of Doryon- ychus raptor (Araneae, Tetragnathidae). Psyche 98:337-350. Gillespie, R.G. 1992a. Hawaiian spiders of the ge- nus Tetragnatha II. Species from natural areas of windward East Maui. J. ArachnoL, 20:1-17. Gillespie, R.G. 1992b. Impaled prey. Nature, 35: 212-213. Gillespie, R.G. 1993. Biogeographic pattern of phylogeny among a clade of endemic Hawaiian GILLESPIE— SPECIATION RATES IN HAWAIIAN TETRAGNATHA 85 spiders (Araneae, Tetragnathidae). Mem. Queensland Museum, 33:519-526. Gillespie, R.G. 1994. Hawaiian spiders of the ge- nus Tetragnatha: III. T. acuta clade. J. Arach- nol., 22:161-168. Gillespie, R.G. & T. Caraco. 1987. Risk sensitive foraging strategies of two spider populations. Ecology, 68:887-899. Gillespie, R.G. & H.B. Groom. 1995. Comparison of speciation mechanisms in web-building and non- web-building groups within a lineage of spi- ders. Pp. 121-146. In Hawaiian Biogeography: Evolution on a Hot Spot Archipelago. (W.L. Wagner & V.A. Funk, eds.). Smithsonian Insti- tution Press, Washington. Gillespie, R.G. & N. Reimer. 1993. The effect of alien predatory ants (Hymenoptera, Formicidae) on Hawaiian endemic spiders (Araneae, Tetrag- nathidae). Pacific Science, 47:21-33. Gillespie, R.G., H.B. Groom & G.L. Hasty. 1997. Phylogenetic relationships and adaptive shifts among major clades of Tetragnatha spiders (Ar- aneae: Tetragnathidae) in Hawaii. Pacific Sci- ence, 51:380-394. Gillespie, R.G., S.R. Palumbi & H.B. Groom. 1994. Multiple origins of a spider radiation in Hawaii. Proc. Natl. Acad. Sci., 91:2290-2294. Gillespie, R.G., M. Rivera & J. Garb. 1998. Sun, surf and spiders: taxonomy and phylogeography of Hawaiian Araneae. Pp. 41-51. In Proceedings of the 17th European Colloquium of Arachnol- ogy, Edinburgh 1997. British Arachnological So- ciety, Burnham Beeches, Bucks. Juan, C., K.M. Ibrahim, P. Oromi & G. Hewitt. 1996. Mitochondrial DNA sequence variation and phylogeography of Pimelia darkling beetles on the island of Tenerife (Canary Islands). He- redity, 77:589-598. Kambhampati, S. & K.S Rai. 1991. Variation in mitochondrial DNA of Aedes species (Diptera: Culicidae). Evolution, 45:120-129. Karsch, E 1880. Sitzungs-Berichte der Gesellschaft Naturforschender freunde zu Berlin. Jahrgang. Sitzung vom 18:76-84. Kaston, B.J. 1948. How To Know The Spiders. 3rd. ed., Wm. C. Brown Co., Dubuque, Iowa, Levi, H.W. 1981. The American orb- weaver genus Dolichognatha and Tetragnatha north of Mexico (Araneae: Araneidae, Tetragnathinae). Bull. Mus. Comp. Zool. Harvard. 149:271-318. Martin, A.P. & S.R. Palumbi. 1993. Body size, metabolic rate, generation time, and the molec- ular clock. Proc. Natl. Acad. Sci., 90:4087-4091. Mayr, E. 1963. Animal Species and Evolution. Harvard Univ. Press, Cambridge. McCune, A.R. 1998. How fast is speciation? Mo- lecular, geological, and phylogenetic evidence from adaptive radiation in fishes. Pp. 585-610. In Molecular evolution and adaptive radiation. (T.J. Givnish & K.J. Systsma, eds.). Cambridge Univ. Press, Cambridge. Rausher, M.D. 1984. The evolution of habitat pref- erence in subdivided populations. Evolution, 38: 596-608. Roderick, G.K. & R.G. Gillespie. 1998. Patterns of speciation and phylogeography of Hawaiian ar- thropods. Mol. Ecol., 7:309-322. Rosenzweig, M.L. 1990. Ecological uniqueness and loss of species: commentary. Pp. 188-198. In The Preservation and Valuation of Biological Resources. (G.H. Orians, G.M. Brown, Jr., WE. Kunin & J.E. Swierzbinski, eds.). Univ. of Wash- ington Press, Seattle. Simon, C. 1987. Hawaiian evolutionary biology: An introduction. Tr. Ecol. EvoL, 2:175-178. Simon, C., F. Frati, A. Beckenbach, B. Crespi, H. Liu & P. Flook. 1994. Evolution, weighting, and phylogentic utility of mitochondrial gene se- quences and a compilation of conserved poly- merase chain reaction primers. Ann. Entomol. Soc. America, 87:651-701. Simon, E. 1900. Arachnida: Fauna Hawaiiensis 2(5):443-519, pis. 15-19. Suman, T.W. 1970. Spiders of the family Thomis- idae in Hawaii. Pacific Insects, 12:773-864. Swofford, D.L. 1998. PAUP*. Phylogenetic anal- ysis using parsimony, version 4.0. 1 . Smithsonian Institution, Washington, D.C. Tamura, K. 1992. Estimation of the number of nu- cleotide substitutions when there are strong tran- sition-transversion and G+C content biases. Mol. Biol. EvoL, 9:678-687. Thoday, J.M. 1972. Disruptive selection. Proc. Roy. Soc. London, 182:109-143. Wagner, W.L. & V.A. Funk. 1995. Hawaiian Bio- geography: Evolution on a Hot Spot Archipela- go. Smithsonian Institution Press, Washington, D.C. Manuscript received 1 May 1998, revised 24 Feb- ruary 1999. 1999. The Journal of Arachnology 27:86-93 FOSSIL EVIDENCE, TERRESTRIALIZATION AND ARACHNID PHYLOGENY Jason A. Dunlop: Institute fiir Systematische Zoologie, Museum fur Naturkunde, D- 10 115 Berlin, Germany Mark Webster: Department of Earth Sciences, University of California, Riverside, California 92521 USA ABSTRACT. Geological and morphological evidence suggests that the earliest scorpions were at least partially aquatic and that terrestrialization occurred within the scorpion clade. Scorpions and one or more other arachnid lineages are therefore likely to have come onto land independently. The phylogenetic position of scorpions remains controversial and we question Dromopoda, in which scorpions are placed derived within Arachnida, as this is not supported by scorpions’ lateral eye rhabdomes, embryology and sperm morphology. We propose a synapomorphy for scorpions + eurypterids, a postabdomen of five segments as part of an opisthosoma of 13 segments. Scorpions and tetrapulmonates must have evolved their book lungs convergently while fossil evidence indicates that a stomotheca, synapomorphic for Drom- opoda, is probably convergent too. ‘Arachnid’ characters such as Malpighian tubules, the absence of a carapace pleural margin, and an anteriorly directed mouth may also be convergent, although their status as synapomorphies can be defended using parsimony. Convergence is difficult to prove unequivocally, but when there are strong grounds for suspecting it, such characters are questionable evidence for arachnid monophyly. Living arachnids are predominantly terres- trial, with those groups found in aquatic hab- itats, such as the water spider Argyroneta or halacaroid mites, assumed to have returned to the water secondarily. It is also reasonable to assume that the earliest chelicerates were aquatic, including the ancestors of arachnids (e.g., Kraus 1976). Terrestrialization was a key event in arachnid evolution and as there is good evidence for aquatic fossil scorpions (Jeram 1998), terrestrialization probably oc- curred independently within at least two arachnid lineages. The position of the scorpi- ons within Chelicerata has proved to be con- troversial (e.g., Weygoldt 1998) and we do not accept that scorpions are a derived group within arachnids, nor that Arachnida is mono- phyletic. A number of characters which have been used to support arachnid monophyly could also be interpreted as adaptations for life on land (e.g., Kraus 1976). We consider these to include book lungs, Malpighian tu- bules, absence of carapace pleural margin, and an anteriorly directed mouth. Parsimony ana- lyses (e.g., Wheeler & Hayashi 1998) suggest that these characters should be treated as ho- mologous and apomorphic. However, if at least two arachnid lineages moved onto land independently it would be surprising to find terrestrial adaptations in the common ancestor of all arachnids, irrespective of whether or not they were monophyletic, as this was an aquat- ic animal. These arguments are expanded be- low. THE AQUATIC NATURE OF PRIMITIVE SCORPIONS A number of authors have suggested that some of the oldest fossil scorpions were par- tially or wholly aquatic (e.g., Pocock 1901; Kjellesvig-Waering 1986). This proposal has not gone unquestioned. Petrunkevitch (1953) found no evidence for scorpion gills and ar- gued that all other arachnids have book lungs or tracheae. More recently Weygoldt & Paulus (1979) again questioned whether scorpions re- ally were aquatic, with Weygoldt (1998) find- ing Kjellesvig-Waering’s (1986) evidence for scorpion gills unconvincing. Stqrmer (1970) and Brauckmann (1987) also described scor- pion gills, although these structures projecting beyond the body margin are equivocal as re- spiratory organs and do not resemble xipho- suran book gills. Gills in scorpions have not, 86 DUNLOP & WEBSTER— ARACHNID PHYTOGENY 87 therefore, been proven, but in his recent phy- logeny of Silurian and Devonian scorpions Jeram (1998) presented both sedimentological and morphological evidence that at least some of these early scorpions were aquatic. His most basal scorpion taxon comes from the De- vonian Hunsriickschiefer, a fully marine se- quence (Bartels et al 1998), while other Si- lurian taxa come from marginal marine deposits. Morphological evidence for an aquatic habitat includes the presence of gna- thobases in some taxa, lack of an oral tube for liquid feeding, single-clawed, digitigrade tarsi, and abdominal plates lacking book lung spi- racles. Jeram (1998) further commented that two or more scorpion lineages might have come onto land independently, but that this would be difficult to detect as many of the characters he analyzed were likely to have al- tered state during terrestrialization. Scorpions have obvious autapomorphies (pectines, sting, pedipalpal claws) and there is no evidence to derive any other arachnid order directly from the scorpion clade, e.g., Ander- son (1973); something which would require autapomorphy reversal. It is conceivable that scorpions and other arachnids evolved on land from a common terrestrial ancestor and that some scorpions then re-entered the water and lost their terrestrial adaptations. However, as the fossil record shows an accumulation of terrestrial-related features through the Palaeo- zoic (e.g., Selden & Jeram 1989, fig. 4), this remains an unlikely scenario. We therefore suggest that if the oldest scorpions were aquatic (Jeram 1998), they must have di- verged from the other arachnids while still in the water and moved onto land independently. SCORPIONES + EURYPTERIDA The phylogenetic position of scorpions re- mains controversial with the three principal cladistic analyses (Weygoldt & Paulus 1979; Shultz 1990; Wheeler & Hayashi 1998) in- cluding them in a monophyletic Arachnida, al- though differing in placement of the order. Weygoldt & Paulus (1979) and Weygoldt (1998) placed scorpions as a sister group to all other arachnids. However, the analyses of Shultz (1990) and Wheeler & Hayashi (1998) (primarily using Shultz’s morphological data) placed scorpions higher in the Arachnida as a sister group to Haplocnemata (Pseudoscorpi- ones + Solifugae) with Opiliones as sister group to all three, forming the taxon Dromo- poda. Several authors have suggested that scorpions are most closely related to euryp- terids (e.g., Versluys & Demoll 1920; Bris- towe 1971; Kjellesvig-Waering 1986). These studies generally relied on overall similarities rather than discrete synapomorphies, and were justifiably criticized for this by Shultz (1990). While Shultz (1990) found scorpions to be de- rived arachnids, he excluded from his analysis Weygoldt & Paulus’ (1979) character 21 (star- shaped lateral eye rhabdomes in scorpions and xiphosurans, quadratic in all other arachnids bearing lateral eyes). Wheeler & Hayashi (1998) did include this character. Both Wheel- er & Hayashi (1998) and Shultz ignored An- derson’s (1973) observation that the embryo- logical development of scorpions and xiphosurans is similar in possession of a growth zone giving rise to both the prosoma and opisthosoma, while in all other arachnids this growth zone gives rise to the opisthosoma only. The distribution of these characters does not favor placement of scorpions as derived arachnids with Xiphosura as an outgroup. Fur- thermore, the analysis of Shultz (1990) found spermatozoa with a coiled flagellum axoneme to be synapomorphic in Arachnida, despite re- tention of a free flagellum in scorpion sper- matozoa (the presumed plesiomorphic state). Weygoldt & Paulus (1979) proposed a coiled flagellum as synapomorphic only for non- scorpion arachnids. Alternatively, the eye rhabdome, growth zone character and sperm flagellum could all be reversals in the scorpion clade. Dunlop (1998) stated that the clearest po- tential synapomorphy for scorpions and eu- rypterids is the five-segmented postabdomen, an obvious character (Fig. 1) ignored by cla- distic studies. Shultz (1990) and Weygoldt (1998) argued that similarities between euryp- terids and scorpions were probably symple- siomorphic, as they occurred in xiphosurans, too. Outgroup comparison with trilobites and other arachnates indicates that lack of a post- abdomen is the plesiomorphic state with re- spect to Chelicerata, while in synziphosurines (primitive xiphosurans; Anderson & Selden 1997) and some arachnids there is a postab- domen of three segments (the pygidium of Shultz 1990). In support of the homology of the postabdomen we argue that scorpions and eurypterids have a groundplan of 13 opistho- 88 THE JOURNAL OF ARACHNOLOGY Figure 1. — Proposed homology of the ventral opisthosoma of a Paleozoic scorpion (left) a scorpion- like eurypterid (center) and a synziphosurine (right). This model differs in its placement of the pectines from Dunlop (1998, fig. 4) and it should be added that the chilaria in at least some synziphosurines may have been fully pediform (L. Anderson pers. comm.). Our model lines up four morphological reference points: the genital segment (A), the last opisthosomal appendage pair (B), the first postabdominal segment (C) and the posteriormost, 13th segment (D). Points A-C, i.e., the preabdomen, are seen in Scorpiones, Eurypterida and Xiphosura; C-D, i.e., the 5-segmented postabdomen, is synapomorphic for Scorpiones + Eurypterida. Opisthosomal segments numbered. Abbreviations: T = telson, st = sternum, emb = embry- onic appendages, mt = metastoma, go = genital operculum, ego = eurypterid genital operculum formed from two fused appendage pairs, pt = pectines, ga = genital appendage, ch = chilaria, op = appendage- derived operculum, a structure usually termed Blatfuss (literally ‘sheet foot’) in eurypterids. somal segments: A preabdomen (segments 1- 8) and a postabdomen (segments 9-13). THE QUESTION OF OPISTHOSOMAL SEGMENTATION In scorpions. — The groundplan of arach- nids has widely been quoted as comprising an opisthosoma of 12 segments (e.g., Kraus 1976; Shultz 1990). However, there is evi- dence for a transitory pair of pregenital limb buds described in scorpion embryos (Brauer 1895) representing an ‘extra’ segment, the nerve ganglion of which is retained (Anderson 1973). A scorpion with 13 segments, includ- ing this embryonic one, was figured by Millot (1949, figs. 52, 53). Shultz (1990) coded scor- pions as lacking appendages on opisthosomal segment 1 despite evidence for their presence embryologically in the less derived, apoiko- genic scorpions (Anderson 1973). Scorpion opisthosomal segmentation has therefore been proposed as: (1) embryonic, (2) genital oper- cula, (3) pectines, (4-7) book-lungs, (8) last preabdominal segment and (9-13) postabdo- men, plus a telson (e.g., Dunlop 1998). In some of the earliest scorpions there is an additional abdominal plate (Jeram 1998) ap- parently representing an additional segment. While modern scorpions have four append- age-derived book lungs on the preabdomen, some early derivative scorpions have five ap- pendage-derived opercula (Fig. 1) posterior to the genital operculae and the pectines. This could be interpreted as evidence that scorpi- DUNLOP & WEBSTER— ARACHNID PHYLOGENY 89 ons have a groundplan of 14 opisthosomal segments, e.g., (1) embryonic, (2) genital opercula (3) pectines, (4-8) book=lungs, (9) last preabdominal segment and (10-14) post- abdomen, plus a telson. However, all known fossil and Recent scorpions express only 12 tergites, the last five of which are fused with the sternites into the postabdomen (Fig. 1). It should be possible to match all tergal and ster- nal elements. Weygoldt & Paulus (1979, fig. 2) accepted that there are 13 ventrally expressed structures in Recent scorpions, but cautioned that there is no musculature evidence for an additional tergal element; i.e., they interpreted scorpions as having only 12 segments. They suggested that the ‘extra’ segment in scorpions is the result of the division of the ventral elements of opisthosomal segment 2; for them a scor- pion autapomorphy. Essentially they argued that both the genitalia and pectines belonged to opisthosomal segment 2, a proposal which we support (Fig. 1). However, these authors were not aware of the additional abdominal plate in fossil scorpions, a fully expressed structure which appears to bring us back to a body plan of 13 opisthosomal segments: (1) embryonic, (2) genital opercula + pectines, (3-7) abdominal plates, (8) last preabdominal segment, (9-13) postabdomen, plus a telson (Fig. 1). Our model is tentative, but we believe it fits the available data and means that, like xiph- osurans, the scorpion groundplan consists of eight preabdominal segments with appendages on segments 1-7 (Fig. 1). We are still ‘miss- ing’ a tergite in scorpions, but the first tergite in chelicerates is often reduced and has been overlooked in eurypterids (below). This mod- el requires loss of one abdominal plate in more derived scorpion clades. This is reflected in the cladogram of Jeram (1998, fig. 2, node F). In eurypterids. — Eurypterids are typically reconstructed with 12 opisthosomal segments. However, Raw (1957) cited evidence found by Holm (1898, pp. 8-9, fig. 15) for an ‘extra’ segment in eurypterids, a proposal overlooked by recent authors. Raw (1957, pp. 160-161, fig. 8c) proposed that the membranous fold between the carapace and the opisthosoma formed a discrete but highly reduced segment in eurypterids. Raw (1957) argued that what was traditionally interpreted as the first tergite of eurypterids does not tuck under the poste- rior margin of the carapace, as the tergite does in xiphosurans. Instead, the membrane behind the eurypterid carapace is reflexed forward and then doubles back on itself in a manner consistent with it being a reduced and poorly sclerotized tergite (Fig. 2). One of Holm’s eu- rypterid preparations (British Museum Natural History specimen I. 3406 (6)) was examined and shows these reflexed membranes (Fig. 3). What has traditionally been called tergite 1 in eurypterids is in contact with the carapace at its lateral margins, but has a slightly concave anterior margin forming a narrow gap on the midline in which there is a membrane con- taining thin fragments of sclerotized cuticle (Fig. 3). This, we believe, is consistent with it being a highly reduced tergite; our tergite 1. Reduced first opisthosomal tergites are common in chelicerates as shown by xipho- suran microtergites (Anderson & Selden 1997) and the pedicels of some arachnids (Shultz 1990). These eurypterid cuticle frag- ments could be dismissed as having simply sutured off from the carapace or adjacent ter- gite, and we have noted Weygoldt & Paulus’ (1979) evidence against an extra tergite in scorpions above. However, eurypterid ventral anatomy can also be homologized with our scorpion model (Fig. 1). The eurypterid me- tastoma is interpreted as opisthosomal seg- ment 1, though it is unclear whether it is a sternal or appendicular structure. Jeram (1998) suggested that what is traditionally called the scorpion sternum is homologous with the eurypterid metastoma, and as such both may be appendage-derived elements, ho- mologous with xiphosuran chilaria (Fig. 1). It is conceivable that the embryonic limb buds in scorpions actually become the scorpion sternum. Both are in approximately the same position and the fate of the limb buds and or- igins of the sternum are equivocal. This limb bud/sternum question merits investigation. Assuming homology (segment 1), the next segment in both scorpions and eurypterids bears the gonopore (segment 2). This fits the general chelicerate pattern of a gonopore on opisthosomal segment 2, which appears to be a valuable marker for homologizing segments (e.g., Millot 1949). The eurypterid gonopore on opisthosomal segment 2 lies at the base of the genital ap- pendage. The appendage is part of the large 90 THE JOURNAL OF ARACHNOLOGY Figures 2-3. — Eurypterid dorsal segmentation. 2. Schematic adapatation of Holm’s (1898) observation that the dorsal membrane between the prosoma and opisthosoma in eurypterids folds back on itself. This differs from the way that the other tergites connect to each other. Raw (1957, fig. 8C) cited this as evidence that this membrane represents a highly reduced first opisthosomal tergite giving eurypterids 13, not 12, opisthosomal segments; 3. Camera lucida drawing of BMNH I. 3406 (6) a small, translucent specimen of Baltoeurypterus acid-etched from the matrix. This specimen supports Raw’s interpretation by showing the folding of the membranes at the prosoma-opisthosoma junction (fl) and the slight sclerotization within this membrane (sc), which we interpret as opisthosomal tergite 1. Scale = 5 mm. genital operculum which is divided by a trans- verse suture, suggesting it is formed from the fused appendages of segments 2 and 3. That scorpion pectines possibly belong to the gen- ital segment is interesting in this context. Si- mon Braddy (pers. comm.) suggested that the paired pectines of scorpions could be homol- ogous with the paired furcae at the end of the eurypterid genital appendage; a structure which also appears to belong to opisthosomal segment 2 (see also Braddy & Dunlop 1997). In eurypterids, segments 3-7 bear gill tracts (probably non-homologous with book gills [Manning & Dunlop 1995]), and the last preabdominal segment is segment 8. There is a five-segmented postabdomen (segments 9- 13) plus a telson (Fig. 1). This segmentation pattern merits further investigation. Our mod- el (Fig. 1) suggests homology of several ref- erence points: The genital segment (A), the posteriormost opisthosomal appendages (B), the start of the postabdomen (C), and the five segments to the 13th, posteriormost segment (D). Points A, B and C also match the body plan of xiphosurans (Fig. 1). Scorpions, eu- rypterids, and xiphosurans share an 8-seg- mented preabdomen with appendages on seg- ments 1-7, while scorpions + eurypterids show an apomorphic 5-segmented postabdo- men. No other chelicerates show an 8-seg- mented preabdomen. Thirteen opisthosomal segments could be interpreted as a synapomorphy for Scorpiones + Eurypterida, though in reality this and the postabdomen are probably best treated as ex- pressions of the same character. We have ar- gued that the postabdominal segments (9-13) are homologous in these taxa and that the character state is derived. No outgroup shows a 5-segmented postabdomen. The three-seg- mented postabdomen of some arachnids (Shultz 1990) is unlikely to be derived from this condition by loss of two segments, as this would reduce the total number of opisthoso- mal segments to only 11 (arachnids such as thelyphonids have 12 segments). PROBABLE CONVERGENT CHARACTERS We have proposed one morphological syn- apomorphy shared by scorpions and eurypter- ids. Other characters have been used to sup- port a monophyletic Arachnida and/or a derived position for scorpions within arach- nids. Some of these characters appear to be adaptations for, or associated with, life on land. If arachnids terrestrialized more than once, then terrestrial adaptations are likely to DUNLOP & WEBSTER— ARACHNID PHYLOGENY 91 be convergent. It would be surprising to find terrestrial characters in the aquatic common ancestor predicted by arachnid monophyly and the models of Weygoldt & Paulus (1979), Shultz (1990) and Wheeler & Hayashi (1998). Book lungs. — Book lungs are a Textbook’ arachnid character. However, their distribution indicates that although book lungs in scorpi- ons and other arachnids are clearly homolo- gous with respect to the pre-existing abdom- inal appendages, the book lungs of scorpions are not directly homologous with those of te- trapulmonates (contra Wheeler & Hayashi 1998). Weygoldt (1998) and Dunlop (1998) independently noted that only tetrapulmonate arachnids retain a respiratory organ on the genital segment. It is absent in xiphosurans, scorpions and probably eurypterids. Weygoldt (1998) regarded loss of a respiratory organ on the genital segment as most likely being con- vergent. Dunlop (1998) noted that outgroups such as trilobites have a respiratory organ on all opisthosomal segments, and proposed that the most parsimonious interpretation of this character is to treat it as plesiomorphically re- tained in tetrapulmonates and synapomorphi- cally lost in a xiphosuran, scorpion and eu- rypterid clade. In either case, book lungs in spiders and scorpions are not directly homologous, be- longing to opisthosomal segments 2-3 in te- trapulmonates and 4-7 in scorpions (Dunlop 1998; Kraus 1998). Probable independent ter- restrialization of scorpions and other arach- nids indicates that their book lungs evolved independently from gills in response to the de- mands of breathing on land. Differences in de- tailed lung anatomy might be predicted be- tween scorpions and other arachnids. This represents an interesting line of research which could be applied to other morphologi- cal structures which may be terrestrial adap- tations, e.g., trichobothria, and details of limb morphology and mouthpart structure. Stomotheca. — Shultz’s (1990) taxon Dromopoda is supported by a number of ap- pendicular characters coded primarily from Recent terrestrial forms. The strongest appears to be presence of extensor muscles, but spe- cializations of the femorpatellar and patello- tibial leg articulations, and a stomotheca formed from coxal endites are included. The stomotheca creates a preoral cavity where ex- traintestinal digestion takes place. This feed- ing process is less likely (although possible) in an aquatic animal. Weygoldt (1998) reject- ed Dromopoda, arguing that a stomotheca is clearly absent in many fossil scorpions (see also Jeram 1998, character 23), in solifuges, and in pseudoscorpions. Weygoldt (1998) concluded that either the stomotheca is con- vergent (supported here) or that scorpions must be paraphyletic. This example illustrates the dangers of ignoring fossil data when cod- ing characters. Malpighian tubules. — Malpighian tubules are endodermal extensions of the gut found in most arachnids, but not in palpigrades, opi- lionids and pseudoscorpions (Shultz 1990). They also occur convergently in insects. Their function is to remove excretory products such as guanine and uric acid from the body (e.g., Seitz 1987). They are not present in xipho- surans and their presence or absence is un- known in eurypterids. The tubules could be convergent terrestrial adaptations for remov- ing dry, low-toxicity excretory products (e.g., guanine), given the importance of water con- servation for animals on land (Kraus 1976). Poorly developed carapace pleural mar- gin.— The carapaces of xiphosurans, and to a lesser extent eurypterids, project laterally and form a cavity around the coxosternal region (Shultz 1990). This projection is not seen in arachnids. Its association primarily with taxa that masticate food using gnathobases may be significant; it is also seen in outgroups such as trilobites, although it is not apparent in fos- sil or Recent scorpions. Shuster (1950) noted that xiphosurans feed by burying themselves into the substrate in pursuit of worms and mollusks. The carapace pleural margin, and the associated cavity it creates, forms a semi- enclosed chamber within the substrate in which the gnathobases are free to masticate food. With a move towards terrestrialization and away from feeding in the substrate, a car- apace pleural margin could become non-func- tional and lost. In contrast, many arachnids show a trend towards developing a preoral cavity (e.g., Selden & Jeram 1989) which sur- rounds the food during cheliceral mastication and extraintestinal ingestion. Anteroventrally directed mouth. — In xiphosurans the mouth is directed posteroven- trally, towards the postoral gnathobases from where they receive masticated food (Shultz 1990). Eurypterid mouths have also been re- 92 THE JOURNAL OF ARACHNOLOGY constructed with this orientation, and although supportive fossil evidence is weak, the inter^ pretation is probably correct on functional grounds. Mouth orientation of fossil scorpions is unknown. Trilobites were also gnathobasic feeders and had posteroventrally directed mouths. Recent arachnid mouths are all di- rected anteroventrally, towards the preoral chelicerae. Terrestrialization could have caused a shift from gnathobasic mastication in water to cheliceral mastication on land. Post- oral gnathobases along the length of the pro- soma are of little use for mastication on land where food would drop from between the cox- ae. However, a similar function appears to have been retained in the palpal coxae of spi- ders, adjacent to the mouth, which are used to manipulate food (e.g., Bristowe 1971). Unlike xiphosurans, which trap their food in soft sed- iments beneath them (Shuster 1950), most arachnids live on an essentially solid substrate and generally catch their food in front of them using anteriorly directed, preoral chelicerae and/or anterior appendages. An anteriorly di- rected mouth can be interpreted as an adap- tation for receiving prey captured preorally by a terrestrial animal. PARSIMONY AND THE BURDEN OF PROOF Arachnid monophyly is supported by a number of characters which may represent convergence in response to terrestrialization. These characters can be defended by parsi- mony and assumed to be homologous and synapomorphic until 'proved’ otherwise by a parsimony analysis. Kraus (1998) discussed the limitations of parsimony analysis and ar- gued that characters for phylogeny should be selected and weighted a priori based on struc- tural and functional considerations. We have presented functional arguments to question some of the characters supporting arachnid monophyly. Unfortunately, unequivocal proof for convergent evolution of characters can be difficult to establish, especially when they in- volve characters in fossil taxa where function- al morphology often has to be inferred (e.g., feeding methods in eurypterids) or where em- pirical data is not available. Our concern is that parsimony is being used to defend weak or inappropriate characters to which consid- erable objections can be raised. We support Kraus (1998) in questioning whether charac- ters should be assumed homologous unless detailed arguments in favor of their homology are presented (e.g., our postabdomen charac- ter). We worry that in an attempt to compile ever larger databases of characters for parsi- mony analysis, homologies are being assumed at face value without assessment of character validity. Which is more important, the quan- tity or the quality of the data used? ACKNOWLEDGMENTS We thank Simon Braddy, Lyall Ander- son, Otto Kraus, Andy Jeram, Roger Farley, Scott Stockwell, Paul Selden, David Walter, Peter Weygoldt & Manfred Ade for useful dis- cussions and correspondence and Bill Shear for valuable criticisms of an earlier version of the manuscript. LITERATURE CITED Anderson, D.T. 1973. Embryology and Phylogeny in Annelids and Arthropods. Pergamon, New York. Anderson, L.L & P.A. Selden. 1997. Opisthosomal fusion and phylogeny of Palaeozoic Xiphosura. Lethia, 30:19-31. Bartels, C., D.E.G. Briggs & G. Brassel. 1998. The Fossils of the Hunsrack Slate. Marine Life in the Devonian. Cambridge Univ. Press, Cambridge. Braddy, S.J. & J.A. Dunlop. 1997. The functional morphology of mating in the Silurian eurypterid, Baltoeurypterus tetragonopthalamus (Fischer, 1839). Zool. J. Linn. Soc., 121:435-561. Brauckmann, C. 1987. Neue Arachniden-Funde (Scorpionida, Trigonotarbida) aus dem west- deutschen Unter-Devon. Geol. Palaeont., 21:73- 85. Brauer, A. 1895. Beitrage zur Kenntnis der En- twicklungsgeschicte des Skorpiones, IL Z. Wiss. ZooL, 59:351-433. Bristowe, W.S. 1971. The World of Spiders. 2nd ed, Collins, London. Dunlop, J.A. 1998. The origins of tetrapulmonate book lungs and their significance for chelicerate phylogeny. Pp. 9-16. In Proc. 17th Europ. Coll. ArachnoL, Edinburgh 1997. Holm, G. 1898. Uber die Organisation von Euryp- term fischeri Eichw. Acad. Imp. Sci. St. Peters- burg, Mem., ser. 8, voL 8, no, 2. Jeram, A.J. 1998. Phylogeny and classification of Palaeozoic scorpions. Pp. 17-31. In Proc. 17th Europ. Coll. ArachnoL, Edinburgh 1997. Kjellesvig-Waeiing, E.N. 1986. A restudy of the fossil scorpions of the world. Palaeontogr. Amer- icana, 55:1-287. Kraus, O. 1976. Zur phylogenetischen Stellung und Evolution der Chelicerata. EntomoL Ger- manica, 3:1-12. DUNLOP & WEBSTER— ARACHNID PHYLOGENY 93 Kraus, O. 1998. Elucidating the historical process- es of phylogeny: Phylogenetic systematics versus cladistic techniques. Pp. 1-7. In Proc. 17th Eu- rop. Coll. ArachnoL, Edinburgh 1997. Manning, PL. & J.A. Dunlop. 1995. The respira- tory organs of eurypterids. Palaeontology, 38: 287-297. Millot, J. 1949. Classe des Arachnides. Morphol- ogie generale et anatomic interne. Pp. 261-319. In Traite de Zoologie, Tome VI, (P.-P. Grasse, ed.). Masson et Cie, Paris. Petrunkevitch, A.I. 1953. Paleozoic and Mesozoic Arachnida of Europe. Mem. Geol, Soc. America, 53:1-128. Pocock, R.I. 1901. The Scottish Silurian scorpion. Q. J. Microscop. Sci., 44:291-311. Raw, E 1957. Origin of chelicerates. J. Paleont., 31:139-192. Seitz, K.-A. 1987. Excretory organs. Pp. 239-248. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin. Selden, PA. & A.J. Jeram. 1989. Palaeophysiology of terrestrialisation in the Chelicerata. Trans. R. Soc. Edinburgh: Earth Sci., 80:303-310. Shultz, J.W. 1990. Evolutionary morphology and phylogeny of Arachnida. Cladistics, 6:1-38. Shuster, C.N. 1950. Observations on the natural history of the American horseshoe crab, Limulus poiyphemus. Woods Hole Oceanograph. Inst., 564:18-23. St0rmer, L. 1970. Arthropods from the Lower De- vonian (Lower Emsian) of Aiken an der Mosel, Germany. Part 1. Arachnida. Senckenbergiana Lethaea, 51:335-369. Versluys, J. & Demoll, R. 1920. Die Verwandt- schaft der Merostomata mit den Arachnida und den anderen Abteilugen der Arthropoda. Kon. Ak. Wetenschap. Amst., 23:739-765. Weygoldt, P. 1998. Evolution and systematics of the Chelicerata. Exp. App. Acarol., 22:63-79. Weygoldt, P. & H.F. Paulus. 1979. Untersuchungen zur Morphologic, Taxonomic und Phylogenie der Chelicerata. Z. E Zool. Syst. Evolutionsforsch., 17:85-116, 177-200. Wheeler, W.C. & C.Y. Hayashi. 1998. The phylog- eny of the extant chelicerate orders. Cladistics, 14:173-192. Manuscript received 20 April 1998, revised 23 Jan- uary 1999. 1999. The Journal of Arachnology 27:94-102 CEPHALOTHORACIC SULCI IN LINYPHHNE SPIDERS (ARANEAE, LINYPHHDAE, LINYPHIINAE) Gustavo Hormiga: Department of Biological Sciences, George Washington University, Washington, D.C. 20052, USA ABSTRACT. Pore-bearing cephalothoracic sulci (pits) are described and illustrated for the first time in several linyphiine spiders (Linyphiidae, Linyphiinae). Sulci are reported in members of the genera Bath- yphantes Menge, Diplostyla Emerton, Kaestneria Wiehle, Pacifiphantes Eskov & Marusik, Porrhomma Simon, and Vesicapalpus Millidge. The phylogenetic implications of the presence of sulci in linyphiines are discussed. Members of two lineages of linyphiid spi- ders (Linyphiidae) typically exhibit cephalo- thoracic sulci, that is, a pair of grooves or pits of varying shape and depth in the cephalic part of the prosoma (see Hormiga 1994, in press, for references). These two lineages are the Mynogleninae and the Erigoninae. The mynoglenines are a relatively small clade distributed in Africa (Holm 1968), New Zealand (Blest 1979), and Tasmania and some southern Pacific islands (Hormiga, unpubl. data). Mynoglenine sulci are located on the clypeus (Fig. 1-3), below the anterior lateral eyes and are lined with cuticular openings (pores) that connect to glands with secretory cells that exhibit a unique strategy of mem- brane amplification (Blest & Taylor 1977; Blest & Pomeroy 1978; Blest 1979). The sulci of mynoglenines are present in adults of both sexes and in juveniles (Blest & Taylor 1977); all known mynoglenines have clypeal sulci. These sulci do not seem to play any active role during the courtship (at least in the spe- cies studied by Blest & Pomeroy 1978). It has been suggested that the glands that serve these sulci might elaborate defensive secretions be- cause the unique ultrastructure of the clypeal secretory cells is consistent with the synthesis of a toxic product (Blest & Taylor 1977). The Erigoninae are the largest clade of lin- yphiids and are distributed throughout the world except New Zealand (Millidge 1988) and Australia {contra Wunderlich 1995; see Platnick 1997:419) which lack native species. One of the most conspicuous characteristics of many (but not all) species of erigonines is the presence in males of a vast morphological di- versity of cephalic modifications, including lobes, turrets, sulci, pits and modified setae. Erigonine sulci usually have a post-ocular po- sition (Figs. 4, 5) and usually have pores as- sociated with glands that are cytologically dif- ferent from those of the mynoglenine sulci (Blest & Taylor 1977; Schaible et al. 1986; Schaible & Gack 1987). The sulci and other cephalic modifications of erigonines are found only in adult males and play an active me- chanical role during courtship: the sulci are gripped by the female cheliceral fangs, as first noted by Bristowe (1931, 1958). In this paper I report on the discovery of cephalothoracic sulci in several members of the linyphiid subfamily Linyphiinae, includ- ing the well-known Holarctic genus Bathy- phantes Menge. This paper brings to light the presence of these structures in linyphiines but does not attempt to cover exhaustively the dis- tribution and morphological variation of sulci across the Linyphiidae. METHODS The morphological observations were car- ried out using a Leica MZAPO dissecting mi- croscope and a Leica DMRM compound mi- croscope. All illustrations were done using a camera lucida and inked on drafting film. Specimens examined with the SEM were first transferred to a vial with 75% ethanol and then cleaned ultrasonic ally for 1-3 nfinutes. They were then transferred to absolute ethanol and left overnight. Specimens were air dried and then glued to rounded rivets using an ac- etone solution of polyvinyl resin. Specimens were coated with a carbon base coat followed by a gold-palladium coat for SEM examina- 94 HORMIGA— CEPHALOTHORACIC SULCI IN LINYPHIINE SPIDERS 95 tion with the AMRAY 1810 of the Smithson- ian Institution’s SEM Facility. All the speci- mens studied (see Table 1) are deposited at the National Museum of Natural History (Smithsonian Institution, Washington, D.C.) unless otherwise stated. In addition to those specimens listed in Table 1, I examined Hap- Unis diloris (Urquhart): New Zealand, Fiord- land Cascade, in Raoulia, 16 January 1975 (col. & det. A.D. Blest), (Otago Museum, Dunedin) and Lophomma punctatum (Black- wall): United Kingdom, Killhope haw., Dur- ham, 2000’, 15 June 1966 (J.A.L.C.), (AMNH). RESULTS Cephalothoracic sulci were found in all spe- cies examined of the linyphiine genera Bath- yphantes (Figs. 6-11), Diplostyla Emerton (Figs. 17, 18), Kaestneria Wiehle (Figs. 12, 14), Pacifiphantes Eskov & Marusik (present only in P. zakharovi Eskov & Marusik), Por- rhomma Simon, and Vesicapalpus Millidge (Figs. 15, 16). The sulci were absent in Lin- yphantes Chamberlin & Ivie (five species ex- amined) and in Pacifiphantes magnificus (Chamberlin & Ivie) (the males of this latter species remain unknown). The species and specimens studied are given in Table 1. In all these taxa the sulci are located anteriorly in the margin of the carapace, between the che- licerae and the pedipalpal trochanters. The sulcus is a relatively shallow pore-bearing cu- ticular depression that opens ectoventrally and has an elliptical perimeter (margin). The pres- ence/absence of these pores requires SEM to be discerned. The sulcus is best seen after ex- cision of the pedipalp (e.g.. Figs. 6, 9, 15), but in general it can be easily seen without re- moval of any appendages, particularly from an ecto ventral angle. When present, the sulci were found in both males and females of all the 21 species for which both sexes were available for study (see Table 1); no sexual dimorphism in the sulci was discerned with the dissecting microscope. The morphological details of the sulci of both sexes of Bathyphantes pallidus (Banks) (Figs. 6-11) and Kaestneria pullata (O.R-C.) (Figs. 12-14) were examined with the SEM. In the female of Bathyphantes pallidus (Figs. 6-8) the sulci are elliptically shaped (in one specimen the sulcus measured 75 X 45 p,m). The dorsal and lateral margins of the sulcus have a slightly sharper edge than the ventral. The sulcus is provided with numerous cutic- ular pores, particularly in the dorsal half. These pores are not grouped into clusters, as in mynoglenines (Fig. 3), but scattered around, and have a very distinctive wide edge around them (Fig. 8). No cuticular pores can be seen beyond the margin of the sulcus. The sulci of the male of B. pallidus (Figs. 9-11) are very similar to those found in females (in one female specimen. Fig. 11, the sulcus mea- sured 78 X 52 p.m). The sulci of Kaestneria pullata are shallow (Figs. 12-14), particularly in the female, and have a more ventral position under the margin of the carapace. In the females of this species SEM is needed to ascertain the presence of the sulcus which consists of a very shallow depression bearing about 20 cuticular pores (under the dissecting microscope the female sulcus is seen as a slightly darker spot on the cuticle). The male sulcus (Fig. 12) can be seen with the dissecting microscope because the cuticular depression is somewhat deeper than that of the female; in the other aspects these pits are similar to those of the females (Figs. 13-14). In Porrhomma convexum (Westring) the sulci, although shallow, can be clearly seen when viewed under the dissecting microscope. In Porrhomma borealis (Banks) (only one fe- male was available for study), P. montanum Jackson, P. microphtalmum (O.R-C.) (only males were examined), and P. pygmaeum (Blackwall) the sulci are very shallow and ori- ented ventrally and in both respects are some- what similar to those of Kaestneria pullata (Figs. 12-14). The sulci of Pacifiphantes zakharovi are present in both sexes and are similar, both in position and morphology, to the sulci of Bath- yphantes pallidus. Males and females of Lepthyphantes zebra (Emerton) were examined with the SEM. No cuticular pores were found on the area of the carapace where the sulcus is found in other linyphiines. DISCUSSION The discovery of cephalothoracic sulci in linyphiines, such as the members of the ge- nus Bathyphantes, is remarkable because this genus includes some very common spiders (in the Holarctic region) and because these Table 1. — Taxonomic distribution of cephalothoracic sulci (pits) in the linyphiid taxa studied and specimens examined. “ + ” indicates sulci present and absent. A “?” indicates that no specimens were available for study; “ + ?” or “ — indicates that the observation needs corroboration with scanning electron microscopy (see text for details). Museum abbreviations: AMNH (American Museum of Natural History, New York), CAS (California Academy of Sciences, San Francisco), JZM (J. Zujko-Miller private collection), USNM (National Museum of Natural History, Smithsonian Institution, Washington, D.C.), ZMUK (Zoological Museum, University of Copenhagen, Copenhagen). 96 THE JOURNAL OF ARACHNOLOGY b ^ a 0^0 U ^ £ ^ c/) £ cd O ^ ^ § ^ ^ ^ D D N ^ B o ? Q ^ O CO cd s * • S 3 Sh GO OQ PQ S c: o o ^ a M o :a ^ <11 00 § Q K & 05 0^ ^ 5 c/D C/D ^ D H CO ON ON 00 NO S-i ON ON O ^ - s Cfl fli P P O bO W) 0) P P Q 00 S ^ S S ^ ^ ^ ^ ^ g 5 C/D OO 00 GO > ;d p N y NO NO ON ON ON ON 0^ S ^ 00 ^ ON ON 00 ^ ^ ^ ^ 3 S § ^ S o ON Cu r- 00 o I o ir^ — < -H GO U H p o p X Ph p bp g p B £ o p GO ■x p -3 u o > (U U o 5 P O s a .B 0 p X X ’d' o 5 'p >-~i B o o p >> 4h o pq p - p TO (U c« (D < m m p b • + + G e NU X > X ^ -2 p o t: 0) S w to •S- K Q ^ Q B B CQ CQ DQ OQ CSS Q Q <3 ^ ^ ^ Ci, Ci, ^ + + + + + + I 'S'o w g*! . o t! I •a ^ W ^05 5 3^ S V. So -V 5 5 5 Q ^3 Q -5-5-5 Cg ^ 5, S ^ ^ "S "S "S 5 5 Q CQ QQ «5 I < (U GO Q P + + + C-. bC p •G o § s •I ^ 'C 5 O 5 5 'a 5 -5 -5 ^ 5, •S Q a CQ CQ < GO W ^-1 .2 c M g c« P P X X fi o O bQ o GO .« t/5 Ai .5 .S - -i I > > O S o I < < < < ’S 'i < (UGOGOGOGO ^ pGO q^:d^Pc^p^d + 4-^- + + + + + o- + -p e-. + (>• + + c- + e-. + c- 5 a -5 I a CQ 2 3 ^ c: ^ >> N fo O ^ Co Eo ^ S 5 X Pc' Ui, 6 d O S ^ ^ -S s NU ^ 5 So ^5 5 - 5 P O P 0 a p m o Ci bfi — 9 ^ O 1 P W2 fe;S O G 5 ^ -5 -5 -5 5 a CQ oq 5 5 5 Q 5 3 -5-5-5 5g S S -5-5-5 5 ^ ^ CQ CQ QQ 5 5 Q ^ -5 -5 -§ CQ CQ ^ 5 CQ s o s ^ 5 CS So S G ^ s 5 c § ^ “S ^ ^ Xi 5 ^ Cl X •-^<3 3 5 Q ^ ^ cS Table 1. — Continued HORMIGA— CEPHALOTHORACIC SULCI IN LINYPHIINE SPIDERS 97 s U U U u o ;z; ;d pD G 00 OO ^ s N N N N G O ^ .-H ^ || G O G bill O -a a 0) -o G G •Q. -Q. p O (B O < O q=l 00 G G O G (D 'm G ;-i tZ) > > o O G 0) G o > ’G 0) o O B -a 2 G W) a 'o '% ^ PC ffi -O ^ ^ B B B ^ § fi O (U 0) BOO + + ^• + + + Os Os 1— I Ul o >s -5 ^ a > o o Os fa 0) o ^ ’p IS V § H G T3 ^ G P i3 G B G N eu 00 OS os I CD M '> fa G U + + Co S I I a I I- § I Is N B o ^ B u G G 2 O O ■p G bO a G W) a ■& 2 a 2 s' 8 o !Z) 1 a o tzi "Th a 'n < < < < p 'o 00 00 00 00 « U + + 1 1 1 1 C CD K P, Linyphantes pualla Chamberlin & - - USA, Washington Mason Co., Brown Creek 15 October 1993 Ivie Horse Camp Linyphantes victoria Chamberlin & — — USA, Washington Thurston Co., 25-27 October 1992 Ivie Evergreen State College 98 THE JOURNAL OF ARACHNOLOGY Figures 1--5. — Cephalothoracic sulci in Linyphiidae. 1. Haplinis diloris (Urquhart) (Mynogleninae), female, lateral; 2, 3. Detail of Haplinis diloris male lateral subocular sulcus (right); 4. Lophomma punc- tatum (Blackwall) (Erigoninae), male, lateral; 5. Detail of Lophomma punctatum male postocular sulcus (left). pits are easily observable with the dissecting microscope. Consequently I was very reluc- tant to believe that these distinct structures remained undescribed. The only taxonomic revision of Bathyphantes is that of Ivie (1969) for Nearctic species. In his mono- graph Ivie described and illustrated 27 spe- cies of Bathyphantes (some of them are cur- rently classified in other genera), but he made no reference to the sulci. None of the many isolated taxonomic descriptions of Bathy- phantes, Diplostyla, Porrhomma, and Vesi- capalpus species I have checked mention these pits. It is difficult to explain how such HORMIGA— CEPHALOTHORACIC SULCI IN LINYPHIINE SPIDERS 99 Figures 6-14. — Cephalothoracic sulci in Linyphiidae. 6. Bathyphantes pallidus (Banks), female, lateral; 7. Bathyphantes pallidus, female sulcus (right); 8. Detail of Bathyphantes pallidus female sulcus (right); 9. Bathyphantes pallidus, male, ventrolateral (arrow points to sulcus); 10. Bathyphantes pallidus, male sulcus (right); 11. Detail of Bathyphantes pallidus male sulcus (right); 12. Kaestneria pullata (O.P-C.), male cuticular pores (right); 13. Kaestneria pullata, female cuticular pores (right); 14. Kaestneria pullata, female, ventrolateral (arrow points to location of cuticular pores). a conspicuous structure has remained undoc- umented for so long. The presence of cephalothoracic pore-bear- ing cuticular depressions in linyphiines is of relevance for the study of the evolution of ce- phalic specializations. Unfortunately, detailed understanding of the evolutionary history of these structures is severely hindered by the lack of explicit hypotheses on the phyloge- netic structure of the Linyphiidae. In my opin- ion, one of the main questions concerning the evolution of linyphiid sulci is whether myn- oglenine and erigonine sulci are homologous structures, as argued by Blest (1979), Blest & Taylor (1977), Blest & Pomeroy (1978) and Millidge (1993). In an initial assessment of this problem (Hormiga 1994) I concluded that although the non-homology hypothesis was 100 THE JOURNAL OF ARACHNOLOGY more parsimonious, the question could not be rigorously tested “until more data (taxa, par- ticularly those with any type of sulci and/or glands, and information on the glands)” were studied. Recent progress in the phylogenetics of erigonines (Hormiga in press) supports the non-homology hypothesis of the erigonine and mynoglenine sulci. This is a result of the basal cladistic placement (within the Erigoni- nae) of a number of lineages in which neither the cuticular pores nor the sulci are present. Given such a cladogram topology parsimony requires independent origins (i.e., non-homol- ogy) of the sulci in erigonines and mynoglen- ines; this occurs even when the cuticular pores of erigonines and mynoglenines are coded as homologous in the cladistic analysis. The dis- covery of sulci in linyphiines suggests that this “character” is even more homoplasious than initially argued because several indepen- dent origins of sulci in the Linyphiidae are required: one in mynoglenines, at least two in erigonines, and one or more instances in lin- yphiines (unless mynoglenine and linyphiine sulci are homologous). All described myno- glenines have subocular sulci (Blest 1979), and the presence of such sulci (and their as- sociated glands) is synapomorphic for this subfamily (Hormiga 1994, in press). The phy- logenetic patterns of sulci and cuticular pores in erigonines are much more complex (Hor- miga in press) and our knowledge is still very preliminary. Despite this, it seems that based on the taxonomic distribution of these char- acters and on my cladistic reconstructions, there are multiple origins of the prosomic cu- ticular pores and at least two independent or- igins for sulci in erigonines (pores are more widely distributed across taxa than sulci). In linyphiines prosomic cuticular pores had been described in Bolyphantes (Blest & Taylor 1977). In B. luteolus (Blackwall) pores are found both in the male and the female {contra Blest & Taylor 1977: 91) (Hormiga in press). So far no cephalothoracic sulci have been doc- umented in linyphiines (other than those de- scribed in this paper). Because there is no ex- plicit cladistic hypothesis for the genera (or an adequate subset of genera) of Linyphiinae, at the present time it is not possible to provide a robust hypothesis about the origin of the sul- ci within this subfamily. Nevertheless the gen- era Bathyphantes, Diplostyla and Kaestneria have been traditionally considered close rela- tives (this has been largely based on overall similarity; e.g., Wiehle (1956) or Millidge (1984)), and several species of these genera have changed generic placement within this cluster of genera. The presence of cephalo- thoracic sulci in Bathyphantes, Diplostyla, Kaestneria, and Porrhomma is a derived trait and thus a potential putative synapomorphy supporting the monophyly of these genera. The genitalic morphology of these four genera seems to be consistent with this hypothesis (e.g., Millidge 1977), but this requires further study and testing. Chamberlin & Ivie (1942: 45) and Ivie (1969: 2) suggested that Bathy- phantes and Linyphantes were closely related because they agree in “in most general char- acters.” The five species of Linyphantes that I have examined, including the type species L. aeronauticus (Petrunkevitch), lack sulci and their palp morphology shares little in common with the palp morphology of Bathyphantes. To this date no synapomorphies have yet been proposed that could potentially support the monophyly of Linyphantes plus Bathyphantes. Pacifiphantes magnificus, recently transferred from Bathyphantes by Eskov & Marusik (1994) using phenetic criteria, also lacks sulci (although males remain unknown) although the pits are present in the both sexes of the type species, Pacifiphantes zakharovi. The phylogenetic placement of Vesicapalpus re- mains even more obscure. This monotypic Neotropical genus is known in the literature after a single male specimen (the holotype of Vesicapalpus simplex Millidge). Millidge (1991) did not document the presence of sulci in his sparse description of V. simplex. Sulci are present in both sexes of Vesicapalpus sim- plex (Figs. 15, 16) and in an undescribed spe- cies from Colombia (Hormiga unpubl. data). Whether the sulci of Vesicapalpus are or not homologous to the sulci of other linyphiines cannot be answered in absence of a phyloge- netic hypothesis for the relationships of these taxa. Outside Linyphiidae, but within Araneo- idea, similar pore-bearing sulci are found in most members of the family Anapidae (Plat- nick & Forster 1986, 1989, 1990). Platnick & Forster (1989: 135) suggest that the presence in both sexes of glandular openings at the an- terolateral corners of the carapace may be a synapomorphy of Anapidae. The cuticular openings of anapids are located in a pit on the HORMIGA— CEPHALOTHORACIC SULCI IN LINYPHIINE SPIDERS 101 Figures 15-18. — Cephalothoracic sulci in Linyphiidae. 15. Vesicapalpus simplex (Millidge), male (ho- lotype) from Argentina (right side reversed); 16. Vesicapalpus simplex female from Brazil; 17. Diplostyla concolor (Wider), male from Massachusetts; 18. Diplostyla concolor, female from Massachusetts. (Scale bars = 0.5 mm) edge of the carapace, just above the endites (see figs. 1-4 in Platnick & Forster 1986). In Minanapis Platnick & Forster there is no pit and the pores open directly on the surface of the carapace. In other anapids, such as Gert~ schanapis Platnick & Forster and Maxanapis Platnick & Forster, the cephalic pit has shifted onto a separate sclerite that is reflexed under the lateral margin of the carapace (Platnick & Forster 1989, 1990) in an analogous situation to the condition found in Kaestneria pullata (cf figs. 271 and 272 in Platnick & Forster 1989). In sum, the pore-bearing sulci of linyphi- ines provide another instance of homoplasy in the evolution of cephalic specializations in linyphiids. Preliminary data presented here suggests that the study of this character will be important for phylogenetic reconstruction. Progress in understanding the evolutionary chronicle of these complex character systems will have to wait for more data on their bio- logical role and function (at present, the func- tion of the sulci in mynoglenines and linyphi- ines remains unknown), for more information on its taxonomic distribution, and for a more detailed understanding of the higher level phylogenetics of this large group of araneoid spiders. ACKNOWLEDGMENTS I would like to thank the following curators for the loan of specimens: Jonathan Codding- ton (National Museum of Natural History, Smithsonian Institution), Charles Griswold (California Academy of Sciences), Norman Platnick (American Museum of Natural His- tory), Nikolaj Scharff (Zoological Museum, University of Copenhagen), and Anthony Har- ris (Otago Museum). Yuri Marusik (Institute of Biological Problems of the North, Russian Academy of Sciences) made available for 102 THE JOURNAL OF ARACHNOLOGY study specimens of Pacifiphantes zakharovi. I am very grateful to the numerous people who assisted in my fieldwork in Colombia with J. Coddington and J. Zujko-Miller, including Ed- uardo Florez, Fernando Fernandez, Federico Escobar, Claudia Medina, and Dario Correa. I am thankful to Norman Platnick for pointing out to me the presence of cephalothoracic pits in anapids. Jeremy Zujko-Miller made avail- able to me his collection of Pacific Northwest spiders. Comments on an earlier draft of this manuscript were provided by Kefyn M. Ca- tley, Jeremy Zujko-Miller and an anonymous reviewer. This research has been funded in part by a National Science Foundation PEET grant (NSF DEB-9712353) and by a George Washington University Facilitating Grant. LITERATURE CITED Blest, A.D. 1976. The tracheal arrangement and the classification of linyphiid spiders. J. Zool. (London), 180:185-194. Blest, A.D. 1979. Linyphiidae-Mynogleninae. In The Spiders of New Zealand. Part V. (R.R. For- ster R.R. & A.D. Blest, eds.), Otago Mus. Bull., 5:95-173. Blest, A.D. & H.H. Taylor. 1977. The clypeal glands of Mynoglenes Simon and some other lin- yphiid spiders. J. Zool. (London), 183:473-493. Bristowe, W.S. 1931. The mating habits of spiders: a second supplement, with the description of a new thomisid from Krakatau. Proc. Zool. Soc. London, 4:1401-1412. Bristowe, W.S. 1958. The World of Spiders. Lon- don: Collins. Chamberlin, R.V. & W. Ivie. 1942. A Hundred New Species of American Spiders. Bull. Univ. Utah, 32:1-117. Eskov, K.Y. & YM. Marusik. 1994. New data on the taxonomy and faunistics of North Asian lin- yphiid spiders (Aranei, Linyphiidae). Arthropoda Selecta, 2:41-79. Holm, A. 1968. Spiders of the families Erigonidae and Linyphiidae from East and Central Africa. Ann. Musee Roy. de I’Afrique Centrale (Zoolo- gie), 171:1-49. Hormiga, G. 1994. Cladistics and the comparative morphology of linyphiid spiders and their rela- tives (Araneae, Araneoidea, Linyphiidae). Zool. J. Linnean Soc., 111:1-71. Hormiga, G. In press. Higher level phylogenetics of erigonine spiders (Araneae, Linyphiidae, Erigon- inae). Smithsonian Contrib. Zool. Ivie, W 1969. North American Spiders of the ge- nus Bathyphantes (Araneae, Linyphiidae). Amer- ican Mus. Nov., 2364:1-70. Millidge, A.E 1977. The conformation of the male palpal organs of Linyphiid spiders and its appli- cation to the taxonomic and phylogenetic anal- ysis of the family (Araneae: Linyphiidae). Bull. British Arachnol. Soc., 4:1-60. Millidge, A.E 1984. The taxonomy of the Liny- phiidae, based chiefly on the epigynal and tra- cheal characters (Araneae: Linyphiidae). Bull. British Arachnol. Soc., 6:229-267. Millidge, A.E 1988. The relatives of the Linyphi- idae: phylogenetic problems at the family level (Araneae). Bull. British Arachnol. Soc., 7:253- 268. Millidge, A.E 1991. Further linyphiid spiders (Ar- aneae) from South America. Bull. American Mus. Nat. Hist., 205:1-199. Millidge, A.E 1993. Further remarks on the tax- onomy and relationships of the Linyphiidae, based on the epigynal duct conformation and other characters (Araneae). Bull. British Arach- nol. Soc., 9:145-156. Platnick, N.I. 1997. Advances in Spider Taxonomy 1992-1995. New York Entomol. Soc., with The American Museum of Natural History. New York, 976 pp. Platnick, N.I. & R.R. Forster. 1986. On Teutoniel- la, an American genus of the spider family Mi- cropholcommatidae (Araneae, Palpimanoidea). American Mus. Nov., 2854:1-10. Platnick, N.I. & R.R. Forster. 1989. A revision of the temperate South American and Australasian spiders of the family Anapidae (Araneae, Ara- neoidea). Bull. American Mus. Nat. Hist., 190: 1-139. Platnick, N.I. & R.R. Forster. 1990. On the spider family Anapidae (Araneae, Araneoidea) in the United States. J. New York Entomol. Soc., 98: 108-112. Schaible, U., & C. Gack. 1987. Zur Morphologic, Histologic und biologischen Bedeutung der Kopfstrukturen einiger Arten der Gattung Diplo- cephalus (Araneida, Linyphiidae, Erigoninae). Verb. Natur. Vereins Hamburg, 29:171-180. Schaible, U., C. Gack & H.E Paulus. 1986. Zur Morphologic, Histologic und biologischen Be- deutung der Kopfstrukturen mannlicher Zwerg- spinnen (Linyphiidae: Erigoninae). Zoologische Jahrb. (Systematik), 113:389-408. Schlegelmilch, B. 1974. Zur biologischen bedeu- tung der kopffortsatze bei zwergspinnenmanchen (Micryphantidae). Diplomarbeit, Univ. Freiburg. Wiehle, H. 1956. Spinnentiere oder Arachnoidea. X. 28. Familie Linyphiidae. Tierwelt Deutsch- lands, 44:1-337. Wunderlich, J. 1995. Primerigonina n. gen., the first endemic Australian spider genus of the sub- family Erigoninae (Arachnida: Araneae: Liny- phiidae). Beitr. AraneoL, 4:535-537. Manuscript received 1 May 1998, revised 4 Decem- ber 1998. 1999. The Journal of Arachnology 27:103-116 SPERMATOPHORES AND THE EVOLUTION OF FEMALE GENITALIA IN WHIP SPIDERS (CHELICERATA, AMBLYPYGI) Peter Weygoldt: Albert-Ludwigs-Universitat, Institut fur Biologic I (Zoologie), HauptstraBe 1, D-79104 Freiburg, Germany ABSTRACT. Whip spiders use stalked spermatophores for sperm transfer. These are complex structures, and their morphology varies among genera and families. Usually, the paired sperm masses hidden within the spermatophores are small, and there has been a co-evolution of spermatophores and those parts of the female genitalia which are used to pick up the spermatozoa and to store spermatozoa. These are structures like specialized sclerotizations, glands or, in a few species, seminal receptacles which are hidden inside the genital atrium (or uterus extemus). In most species there are paired erectile bodies, homologous to genital appendages, which are attached to the dorsal side of the genital operculum which also is part of an appendage homologon. All these structures vary among genera and families. The comparison of sper- matophores and genitalia of different species belonging to most genera and families suggest that the female gonopods consist primarily of paired cushion-like structures, each equipped with a small finger-like ap- pendage vestige. These appendage vestiges are retained in many species, particular in the Charinidae and Charontidae. They are erectile by increase in blood pressure, and they are thereby probably bent in characteristic ways and thus can pull off the sperm masses from the spermatophore. In some Charinidae, and in some species of Damon and Phrynichus (Phrynida, Phrynichidae) these appendage vestiges are totally lost. In the Phrynidae, on the other hand, they have become sclerotized and hard. They form the well-known claw-like sclerites, and an invagination at the base of each sclerite has been shaped to form a true seminal receptacle. Similar genitalia have evolved convergently in the genus Trichodamon (Phryn- ida, Phrynichidae). Spermatophores and the corresponding female genitalia and their mechanisms of a number of genera from most families are described and illustrated. Whip spiders transfer spermatozoa by means of stalked spermatophores (Alexander 1962a, b; Klingel 1963; Weygoldt 1969). Af- ter a prolonged courtship dance which the male performs in front of the female, he turns around until facing the same direction as the female and standing in front of her. In this position he deposits a spermatophore and at- taches its stalk to the substratum. Thereafter he turns around to face the female and lures her toward the spermatophore. The female then steps over the spermatophore and picks up the sperm. The spermatophores are large and complex structures. They consist of a stalk, a sper- matophore head and paired sperm masses or, in other species, sperm packages. Each indi- vidual spermatozoon is rolled up and encap- sulated, and the globular cells are either glued together or surrounded by secretion. The sperm masses are small when compared to the size of the total spermatophore. Spermato- phore morphology varies between species, genera and families, and the same is true for the female genitalia, in particular for those structures which are used to pick up the sper- matozoa. After sperm transfer, an empty sper- matophore is left behind or, in a few species, is eaten either by the male or the female. The distal genitalia of a whip spider of ei- ther sex are composed of a large genital atri- um, homologous to the uterus extemus of spi- ders, a large genital operculum, and paired erectile bodies attached to the dorsal or inner side of the genital operculum (Weygoldt et al. 1972). These erectile bodies are considered to be homologous to genital appendages, to the endopods of the opisthosomal appendages of eurypterids or xiphosurids (Pocock 1894; Werner 1935; Weygoldt 1970); they are there- fore termed gonopods here. The genital oper- culum with its lateral book lungs is homolo- gous to the genital operculum of xiphosurids and eurypterids which is part of the same pair of appendages, the book lungs or, in xipho- surids the book gills, representing the exo- pods. 103 104 THE JOURNAL OF ARACHNOLOGY In the male, the gonopods are two-seg- mented. They form an unpaired complex structure provided with muscles and haemo- lymph spaces and a complex central cavity which acts as a mold for the formation of the spermatophore head. It also contains grooves through which the secretions from several large glands can be lead to the exterior in or- der to attach the spermatophore stalk to the ground. Kraus (1970) suggested that the shape of the male genitalia could provide useful characters for taxonomy. The problem is that the shape of these organs depend heavily on the state of preservation. Genitalia of two males of the same species sometimes appear much more different from each other than those from two males of two different species. The spermatophores formed in these male genitalia are complex and diverse. Their sperm masses are located at various positions within the spermatophore head. Unfortunately the spermatophores of only a limited number of species are known. They will be described subsequently. The female genitalia are much simpler. The genital atrium contains specialized sclerotiza- tions which vary among species, glandular structures and various structures used for sperm storage. Even true seminal receptacles have evolved convergently in some groups. The gonopods are cushion-like structures with or without vestiges of the appendage telopod- ids. These telopodids, here termed appendage vestiges, are used to pick up the sperm mass from the spermatophores. Their morphology varies considerably between species, genera and families. Just as with male copulatory or- gans, there has been a co-evolution of sper- matophores and female genitalia. All whip spiders will mate several times if they have the chance. The females become unreceptive once oogenesis has reached a cer- tain stage. All whip spiders are long-lived and continue to molt and to grow after having reached sexual maturity. During molting, the females loose all stored spermatozoa — which remain in the shed storage organs. The fe- males become receptive again soon after molt- ing. In the following discussion I will demon- strate different types of spermatophores and the corresponding female genitalia and de- scribe how these are used to pick up the sperm masses and how they have evolved within the taxon Amblypygi. As to the function, most of my description is inferred from the morphol- ogy of the structures. Many whip spiders are unable to walk and to mate on glass; therefore, direct observation is impossible. As a base for the discussion of the evolu- tion of genitalia I use the cladogram and sys- tem of Weygoldt (1996a, b) (Table 1). In this system, the African genus Paracharon is con- sidered the sister group of the remaining am- blypygids, the Euamblypygi. This group is di- vided into two taxa, the Charinidae and the Neoamblypygi. The Charinidae is mainly characterized by plesiomorphies; I have not found convincing synapomorphies. All char- inid genera are in urgent need of revision. In particular, it is not clear whether Charinus Si- mon 1892 is a monophyletic group and whether Charinides Gravely 1911 should be considered a junior synonym of Charinus as Delle Cave (1986) assumes. The Neoambly- pygi, however, is united by several synapo- morphies. It contains the taxa Charontidae as restricted by Quintero (1986) and the Phryn- ida or Apulvillata, and these are divided into the Phrynidae and Phrynichidae, both char- acterized and united by convincing synapo- morphies. RESULTS Paracharon. — According to the taxonomic analysis, Paracharon caecus Hansen 1921 is the most plesiomorphic species and the adel- photaxon of all remaining Amblypygi, the Euamblypygi. Its female gonopods are simple soft cushions (Fig. 1). Spermatphores are not known, therefore the function of the gonopods are unclear; and it is also unclear whether the simple gonopods are plesiomorphic or the re- sult of simplification. Charinidae and Charontidae. — Gonopods with a soft, finger-like appendage vestige (Fig. 2, 3) are found in many charinid and charontid whip spiders. They are probably synapo- morphic for the Euamblypygi or, if the geni- talia of Paracharon are secondarily simpli- fied, for all Amblypygi. Such finger-like appendage vestiges as shown in Fig. 2, 3 for Charinus koepkei Wey- goldt 1972c may be short and pointed as in this case and in Charinides bengalensis Gravely 1911 or much longer as in several other Charinus species, e.g., Charinus afri~ canus Hansen 1921 (Weygoldt 1972a). The WEYGOLDT— SPERMATOPHORES AND GENITALIA IN THE AMBLYPYGI 105 Table 1. — The amblypygid genera, their relationships according to Weygoldt (1996a and 1996b), their distribution, and species numbers. The species numbers for many genera are guesses based on descriptions from the last century. Only for the Phrynidae and Phrynichidae can reliable data be given; they are based on the revisions of Mullinex (1975), Quintero (1981, 1983, a few more species have been described since then), Weygoldt (1998) and the author’s unpublished data on Damon. Genera Distribution No. of species n ET g B Cl O ' 3 ~ CL P o r— 9 f B o' oi p a> Paracharon -Sarax *“ Phrynichosarax f-Charinus Charinides L-Tricharinus — Catageus Charon Stygophrynus Acanthophrynus pPhrynus Paraphrynus Heterophrynus Xerophrynus Phrynichodamon •Damon r ? —Musicodamon Trichodamon Phrynichus Euphrynichus W. Africa S.-E. Asia S.-E. Asia world wide circumtropic neotropic S.-E. Asia S.-E. Asia S.-E. Asia Mesoamerica Mesoamerica Mesoamerica S. America S.-E. Africa S.-E. Africa Africa N. Africa S. America Africa, Asia Africa 4(?) 5(?) >20 5(?) 3 1 >4 6(?) 1 >16 >12 10(?) 1 10 1 2 14 2 106 THE JOURNAL OF ARACHNOLOGY Figures 1-7. — Female genitalia and spermatophores of some charinid and charontid Amblypygi. 1. Female genitalia of Pracharon caecus; dorsal aspect; 2. Female genitalia of Charinus koepkei, dorsal aspect; 3. Same, posterior aspect (from Weygoldt 1972c); 4. Spermatophore of Sarax sarawakensis in lateral view; 5. Female genitalia of Sarax sarawakensis, dorsal aspect; 6. Spermatophore of Stygophrynus longispina; 7. Female genitalia of Stygophrynus longospina, dorsal aspect, (s = sperm mass). actual appearance may vary between sped- mens, depending on the state of preservation or on haemolymph pressure during preserva- tion. For most species, spermatophores are not known. It is most likely that the finger-like appendage vestiges of the gonopods can be elongated or erected by an increase in hae- molymph pressure and withdrawn by muscles, and that they are bent and strengthened in a species specific way and thus can pull off the protruding sperm masses during sperm trans- fer. However, a few examples have been stud- ied. In Sarax sarawakensis (Thorell 1888) WEYGOLDT— SPERMATOPHORES AND GENITALIA IN THE AMBLYPYGI 107 (Charinidae), the spermatophore head is com- plex, with paired wing-like appendages of which the functional significance is obscure. They probably provide the necessary stimuli for the gonopods to find the two protruding sperm masses (Fig. 4) (Weygoldt 1990). The inactive gonopods are rounded in this species (Fig. 5), but it is likely that they can also be elongated and pull off the sperm masses dur- ing sperm transfer. The situation is similar in Stygophrynus longispina Gravely 1915 (Neoamblypygi, Charontidae). Its spermato- phore (Fig. 6) (Weygoldt 1990) also carries two protruding sperm masses. They are mounted on the distal ends (viewed from the male) of the spermatophore head. The sper- matophore stalk is inserted at about the center of the spermatophore head. Thus, when the female presses down the proximal end of the spermatophore head with her genital opercu- lum, the distal end will raise and move the sperm masses into the female gonopore. The folded appearance of the female genitalia and its appendage vestiges (Fig. 7) suggest that these can be inflated considerably and can tear off the sperm packages during sperm transfer. Charinus is a large genus distributed cir- cumtropically over all continents and also oc- curring on islands, even volcanic ones like Galapagos. Some species have evolved differ- ent spermatophores, different genitalia and different means of sperm transfer. In the Bra- zilian species C. brasilianus Weygoldt 1972 and C. montanus Weygoldt 1972, the append- age vestiges of the genitalia are enlarged and thickened and have the appearance of sucker- like or prehensile structures (Weygoldt 1972a, b) (Fig. 10). They can be extended by increase in haemolymph pressure and retracted by strong muscles. The spermatophores are also different. They are quite simple (Figs. 8, 9), and the spermatozoa do not form compact, protruding sperm masses but flat layers at the base of the spermatophore head (Weygoldt 1972b, 1974a). The female picks up the sper- matozoa by means of her sucker-like genitalia. They are then stored in two distal cavities of the genital atrium directly behind the ends of the genitalia. The situation is even more different in Charinus seychellarum Kraepelin 1898. The genitalia are reduced to flat cushions in front of which the floor of the genital atrium and its roof are strongly sclerotized. There is no erectile appendage vestige. Further, the pos- terior margin of the genital operculum is transparent and forms a hard and sharp edge (Fig. 12). The spermatophore is unique among amblypygids (Fig. 11). There is a strong, tri- angular stalk which firmly attaches the struc- ture to the ground, even to sand. The sper- matophore head consists of a flat plate carrying two strong sperm packages, each with a spacious sperm reservoir and an open- ing at its tip. This tip is bent upwards and forms an embolus armed with two small hooks. The reservoirs of both sperm packages join proximally; here they contain no sper- matozoa but a swelling substance which, on contact with aqueous solutions, presses out the spermatozoa stored distally. This is one mech- anism. There is another, more important mechanism: The flat plate carrying the sperm packages acts as a spring. If the whole struc- ture is bent upwards distally, it arrests at about 45°, and two rod-like structures at the upper part of the spermatophore stalk act as pistons pressing out the sperm masses. During sperm transfer, the female attaches the sharp edge of the margin of her genital operculum under the hooks of the emboli and then bends the sperm package upwards. The spermatozoa are there- by emptied into the genital atrium and stored between the roof of the atrium and a dorsal fold. Phrynida. — The situation in the Phrynida or Apulvillata is much clearer. This taxon con- tains two families, the Phrynidae and the Phrynichidae. Phrynidae. — -In the Neotropical Phrynidae, the female gonopods are equipped each with a claw-like, hard and dark sclerite (Fig. 13). These sclerites have long been known and termed cocoon-holders by Bomer (Werner 1935). They have, however, nothing to do with the transportation of the egg sac. The sclerites can be elevated by increase in hae- molymph pressure. They are further equipped with a strong adductor muscle which, by su- perficial view, seems to be attached to a deep apodeme. However, at the base of each sclerite there is an invagination leading into a spa- cious seminal receptacle (Fig. 14). The aper- tures of these receptacles are covered by the sclerite bases, and the adductor muscles are attached to the walls of the receptacles (Fig. 15). Contraction of these muscles, thus, leads to the adduction of the sclerites and at the 108 THE JOURNAL OF ARACHNOLOGY Figures 8-12. — Spermatophores and female genitalia of two species of Charinus. 8. Spermatophore of C brasilianus, anterior view; 9. Same, lateral aspect; 10. Female genitalia of C. brasilianus, dorsal aspect (from Weygoldt 1972b); 11. Spermatophore of C. seychellarum, lateral view; 12. Female genitalia of C. seychellaraum, dorsal aspect. Abbreviations: e = embolus, h = hooks, rd = rod-like structure which compresses the sperm package when the tip is lifted upwards. Abbreviations: s = sperm mass, sp = sperm package). same time widens the seminal receptacles. The walls of the receptacles are punctured by many glandular pores. Nutrients or other sub- stances are probably released through these pores and nourish or otherwise maintain the spermatozoa (Weygoldt et al. 1972). The males of the Phrynidae produce large spermatophores with triangular, heavily sculp- tured spermatophore heads (Weygoldt 1969, 1972b, 1974b, 1977) (Figs. 16, 17, 19, 20). The formation of these complex spermato- phores takes quite long, 10-20 minutes. After spermatophore formation, the male turns to- ward the female again and touches for another 10 minutes the spermatophore with his pedi- palps and chelicerae. The meaning of this be- havior is still obscure. The attachments of a pheromone may be one possibility, the depo- sition of an enzyme to soften the sperm pack- ages another. The spermatophore contains two comparatively small sperm-packages hidden deeply among the sculpturing (Figs. 17, 18, 19, 21), and two arm-like distal extensions act as conductors leading towards the sperm- packages. The female pulls out these sperm- packages by means of her claw-like sclerites, and the sperm is thereby sucked into the sem- inal receptacles. In Phrynus marginemacula- tus C.L. Koch 1841, the sperm packages are attached to small plates, and these plates are visible pressed underneath the claw-like scler- ites after sperm transfer. The sculpturing varies among species and also the shapes of the arm-like appendages; and they may even be forked or T-shaped. The functional significance of these differences is obscure, in particular since the female geni- talia are quite uniform. We may assume that the different sculpturings aid the female in recognizing the spermatophore and finding the sperm packages. Another point may be that the sculpturing creates a large, hard, elastic WEYGOLDT— SPERMATOPHORES AND GENITALIA IN THE AMBLYPYGI 109 Figures 13-15. — Female genitalia of Phrynus marginemaculatus; 13. Dorsal aspect of genitalia of an exuvia with the claw-like sclerites; 14. Longi- tudinal section through one of the gonopods with the entrance to the seminal receptacle (arrow); 15. Cross section through one of the gonopods with the seminal receptacle containing spermatopzoa. Ab- breviations: m = muscle, rs = seminal receptacle (from Weygoldt et al. 1972). spermatophore head with a minimum of ma- terial. Even the stalk is created with a mini- mum of material; it is not solid as may seem from Figs. 16 and 20 but its cross section is V-shaped instead. This unique system featuring spermato- phores with complexly sculptured spermato- phore heads and female gonopods equipped with claw-like sclerites and seminal recepta- cles is one of the autapomorphies of the Phrynidae. The features are found with little variation in all four genera and in all species (Mullinex 1975; Quintero 1981; Weygoldt 1972b, 1974b, 1977). Phrynichidae* — The Phrynichidae are much more variable as far as their spermato- phores and genitalia are concerned. Phrynichodamon scullyi (Purcell 1911) is a primitive species and the sister taxon of all other Phrynichidae (Weygoldt 1996a) (with the exception of Xerophrynus Weygoldt 1996, which is tentatively considered a basal offshot of the Phrynichidae, Table 1). Spermatophore s and genitalia of this species resemble the sit- uation found in charinid and charontid whip spiders. In the small and simple spermato- phores, the spermatozoa form large, protrud- ing sperm masses (Weygoldt 1998a). The fe- male gonopods are equipped with soft, finger-like appendage vestiges which can probably be extended by haemolymph pres- sure and tear off the sperm masses. These are then stored underneath the appendage vestiges (Weygoldt 1996a) (Fig. 23). It is easy to conceive that a remote ancestor of the Phrynidae had a similar system and that the appendage vestiges became sclerotized and hard and the place at the base of these vestiges invaginated to better store the sper- matozoa. The remaining phrynichids, however, evolved into another direction. In Damon, and convergently in most Phrynichinae, the ap- pendage vestiges were lost. Some of the western species of Damon still have appendage rudiments. In Damon john- stonii (Pocock 1894) (Fig. 24) and in Damon tibialis (Simon 1876) there is a small append- age rudiment which has, perhaps, a sensory function; nothing is known about spermato- phores and sperm transfer in these species. Another undescribed species from Came- roon has evolved very different genitalia. The gonopods are enlarged hook-like structures which are strongly sclerotized and black (Fig. 25). Again, nothing is known about spermato- phores and sperm transfer, but these genitalia strongly suggest that the spermatophore s are very different from those of other Damon spe- cies. This species may be the sister taxon of the Phrynidae, in which case the Phrynichidae form a paraphyletic group. But this is unlike- ly. The Damon species are united by clear synapomorphies, and the genitalia of this Da- mon species and of the Phrynidae are only no THE JOURNAL OF ARACHNOLOGY Figures 16-21. — Spermatophores of two species of the Phrynidae. 16. Spermatophore of Heterophrynus longicornis , lateral view; 17. Anterior view of spermatophore head; 18. Right sperm package enlarged (from Weygoldt 1972b); 19. Spermatophore of Phrynus marginemaculatus, anterior view; 20. The same, lateral view; 21. One of the sperm packages enlarged (from Weygoldt 1969). Abbreviations: a = arm-like distal extension, s = spermatozoa, sp = sperm package. superficially similar. There is no claw-like sclerite on a soft cushion-like gonopod with seminal receptacles, but the tip of the gonopod is sclerotized. It is more likely that this simi- larity is the result of convergent evolution. Damon medius (Herbst 1797), another West African species, has its lost gonopodial ap- pendage vestiges. The gonopods are large cushions with a deep dorsal depression. They look more like a depression surrounded by large walls which join in the midline. In all East African species of Damon, the female genitalia are flat cushions without any appendage vestiges. They are supported by an anterior sclerotized plate or bar (Fig. 29). These cushions are used to detach large sperm-packages from the spermatophore (Weygoldt 1998a; Weygoldt & Hoffmann 1995) (Figs. 26-28). The sperm packages are so large that they fill out nearly the complete genital atrium. This is a specialty of this Da- mon variegatus species group or of all Damon species; the reproductive biology of the west- ern species is unknown. The larger parts of these sperm packages consist of a secreted mass which serves perhaps as a matrix to hold and retain the spermatozoa within the female genital atrium; seminal receptacles are miss- ing (Weygoldt & Hoffmann 1995). The female genitalia of the Old World WEYGOLDT— SPERMATOPHORES AND GENITALIA IN THE AMBLYPYGI 111 Figures 22-29. — Spermatophores of Phrynichodamon and Damon. 22. Spermatophore of Phrynicho- damon scullyi, lateral view (from Weygoldt 1998a); 23. Female genitalia of Phrynichodamon scullyi, dorsal aspect (from Weygoldt 1996a); 24. Female genitalia of Damon johnstonii, dorsal aspect; 25. Female genitalia of undescribed Damon species, dorsal aspect; 26. Spermatophore of Damon diadema, anterior view; 27. Anterior view of spermatophore head of emptied spermatophore; 28. One of the sperm packages; 29. Female genitalia of Damon diadema, dorsal aspect (from Weygoldt & Hoffmann 1995). Phrynichinae, the genera Phrynichus and Eu~ phrynichus, are simple cushion-like elevations which, in some species, have partly sclero- tized walls or tooth-like sclerotized tips (Fig. 33). There are no seminal receptacles; the spermatozoa are stored inside the genital atri- um. The spermatophores of all Phrynichinae are complex and unique. Their spermatophore heads consist of an outer frame and two inner bars, each carrying a compact sperm mass at its proximal end (Figs. 30, 31, 34, 35). At the distal end of the spermatophore head there are arms or levers and, at the basis of these, cush- ions. If these levers or the cushions are pressed down, the bars carrying the sperm masses are elevated (Weygoldt 1998a; Wey- 112 THE JOURNAL OF ARACHNOLOGY Figures 30-35. — Spermatophores of Phrynichus and Euphrynichus. 30. Lateral view of total spermato- phore of Phrynichus ceylonicus, 31. Dorsal aspect of spermatophore head of same; 32. Spermatophore head of emptied spermatophore (from Weygoldt & Hoffmann 1995); 33. Female genitalia of Phrynichus ceylonicus, dorsal aspect (from Weygoldt 1998b); 34. Lateral view of total spermatophore of Euphrynichus bacillifer; 35. Dorsal aspect of spermatophore head (from Weygoldt 1998a). Abbreviations: b = bar carrying sperm mass, c = cushion, h = lateral horn, s = sperm mass. goldt & Hoffmann 1995) (Fig. 32) and the sperm masses can be grasped by the female gonopore. Trichodamon f roe si Mello Leitao 1940, a member of the only New World phrynichid genus, has female genitalia which bear some similarity to those of the Phrynidae (Weygoldt 1977), The gonopods are soft cushions with a hook“like appendage vestige, and close to each hook there is s small seminal receptacle WEYGOLDT— SPERMATOPHORES AND GENITALIA IN THE AMBLYPYGI 113 Figures 36-40. — Spermatophore and female genitalia of Trichodamon froesi (from Weygoldt 1977). 36. Lateral view of total spermatophore; 37. Dorsal aspect of spermatophore head; 38. Lateral view of emptied spermatophore head; 39. Female genitalia, dorsal aspect; 40. Longitudinal section through one of the gonopods. Abbreviations: b = bar carrying sperm mass, c = cushion, h = lateral horn, ho = hook-like structure, rs = seminal receptacle, s = sperm mass. (Figs. 39, 40). The spermatophore of this spe- cies is composed of the same parts as those of the Old World Phrynichinae and have the same mechanism, but they are stronger, with a short and stout stalk and a strong outer frame (Figs. 36-38). If the large cushions are pressed down, the bars carrying the sperm masses are lifted and finally arrested (Fig. 38). The hook-like appendage vestiges are proba- bly used to grasp around the bases of the sperm masses and lead them into the seminal receptacles. DISCUSSION Courtship behavior is similar in all whip spiders observed; and, although there are spectacular variations, it will not discussed here in detail. One of the characteristic fea- tures of the Amblypygi is the fact that the male turns away from the female during sper- matophore formation. The Amblypygi share this behavior with the Uropygi; in other arach- nids which deposit spermatophore s, the male faces the female during spermatophore pro- duction. This characteristic behavior can be assumed to be synapomorphic either for the Pedipalpi (Uropygi and Amblypygi) or for the Megoperculata (Uropygi, Amblypygi and Ar- aneae). I believe, and Alexander & Ewer (1957) do also, that this latter possibility is the correct one. As a consequence, we have to assume that spiders lost this behavior — per- 114 THE JOURNAL OF ARACHNOLOGY haps better, changed this behavior-=in the course of the evolution of their characteristic indirect -direct method of sperm transfer. Al- exander (1962a, b) was the first to observe mating in amblypygids. She described sper- matophore formation as a two step process: The male first deposits an empty spermato- phore, then turns to the female again and fills the spermatophore with spermatozoa. Alex- ander & Ewer (1957) used this two step pro- cess to understand the evolution of mating be- havior in spiders. In some spiders, the male courts the female for a while, then interrupts courtship in order to fill his palpal copulatory organs. Thereafter he turns toward the female again and resumes courtship. In other species, the filling of the copulatory organs and court- ship are two completely separated behavior acts. The behavior of the male whip spider, in which he turns away from the female to pro- duce an empty spermatophore, was hypothe- sized by Alexander & Ewer (1957) to have been the initial step leading to the reproduc- tive behavior of spiders. Flowever, the obser- vation of Alexander is incorrect. I have now observed and videotaped the behavior of sev- eral species of Damon^ including D. variega- tus, the same species Alexander observed. In all amblypgid species observed, the sperma- tozoa are firmly built into the spermatophore as soon as the male lifts his body and starts to turn towards the female again. We can, of course, still assume that a behavior by which the male turns away from the female before spermatophore formation was the initial step from which spider mating behavior evolved. Male and female genitalia, or spermato- phores and female genitalia, have evolved as means to successfully transfer sperm and thus ensure insemiation. I assume that they have been shaped by sexual selection in the sense of Eberhard (1985). It is evident from these few examples of Amblypgi that the co-evo- lution of spermatophores and female genitalia has led to different structures and mecha- nisms. It is also evident that the structures vary among genera and families and that they can be used as characters in systematic re- search. The comparative approach demonstrated here helps to understand the origin and evo- lution of complex genitalia such as those of Trichodamort or of the Phrynidae with their claw-like sclerites and seminal receptacles. The genitalia also provide useful characters for taxonomy. For example, all Phrynidae are characterized by goeopods with claw-like sclerites and seminal receptacles. Species without this morphological arrangement can- not belong to the family Phrynidae unless it can be shown that the species in question shares other synapomorphies with the Phryn- idae and has not yet evolved the typical gen- italia, or that it has reduced these structures. Thus, the Namibian Xerophrynus machadoi (Purcell 1901) which had been described as Paraphrynus machadoi, has female genitalia different from those of all phrynids and, in fact, different from those of all other species. Because of other characters, this species is probably a remote plesiomorphic member of the Phrynichidae (Weygoldt 1996a). Unfortu- nately, the reproductive biology of this desert- adapted species is not known. In captivity it refuses to produce a spermatophore; perhaps reproductive behavior is triggered by a com- plex set of environmental changes. Therefore the mechanism of sperm transfer in this spe- cies has yet to be discovered. Genitalia with claw-like sclerites and sem- inal receptacles are an autapomorphy of the Phrynidae. Quintero (1980) assumed that such claw-like sclerites are an autapomorphy of the Phrynida (Phrynichidae and Phrynidae) and that the Phrynichidae, with the exception of the undescribed Damon species, have reduced these sclerites. But this is unlikely. The Da- mon species and the Phrynichinae are united by clear synapomorphies, and the genitalia of this Damon species and of the Phrynidae are only superficially similar. It is more likely that this similarity is the result of convergent evo- lution. There are many more open questions-=”in fact, there are more questions than answers. For example, the reproductive biology and the exact systematic position of Musicodamon at- lanteus Fage 1939 within the Phrynichidae is unknown. This species is known from only four badly-preserved museum specimens. The situation is even worse in the Charin- idae. Although many species have the typical gonopods with a finger-like appendage ves- tige, some lack them. The Brazilian species of Charinus possess sucker-like gonopods, and those of Charinus seychellarum are even more different. The spermatophre and female geni- talia of this species are unique among ambly- WEYGOLDT— SPERMATOPHORES AND GENITALIA IN THE AMBLYPYGI 115 pygids. The spermatophore resembles those of some pseudoscorpions or whip scorpions (Uropygi). It came as a surprise to find such different spermatophore s within one genus, and it is hard to believe that species with such different genitalia should be found in the same genus. The gonopods of Tricharinus Quintero 1986 are also different. Quintero (1986) pub- lished SEM pictures which reveal similar de- tails to my own unpublished light microscopy studies. The mechanism of these gonopods re- main unknown, as the spermatophore s are not known; and there are no histological data. The various spermatophores and genitalia in dif- ferent Charinus species and their relatives may suggest that this genus is a paraphyletic or even polyphyletc assemblage, and the fact that the Charinidae as a whole are not char- acterized by synapomorphies shows that stud- ies of the reproductive biology and the asso- ciated structures of these genera are urgently needed and that there is ample work to do for the next generation. Another unsolved question is the functional significance of various part of the complex spermatophore heads. Perhaps studies with larger numbers of specimens, some of which may produce spermatophores with slight mor- phological differences, may lead to an under- standing of female choice and sexual selec- tion. Sexual selection and sperm competition have never been studied in any whip spider, and the meaning or information content of the different behavior elements during courtship or fighting are completely obscure. ACKNOWLEDGMENTS I am grateful for all those who helped to collect whip spiders in different countries; they have been thanked in previous papers. William Eberhard, Brent Opell and an anon- ymous reviewer helped to improve the man- uscript. My studies were sponsored by the Deutsche Forschungsgemeinschaft. LITERATURE CITED Alexander, A.J. 1962a. Courtship and mating in amblypygids (Pedipalpi, Arachnida). Proc. ZooL Soc. London, 138:379-383. Alexander, A.J. 1962b. Biology and behavior of Damon variegatus Perty of South Africa and Ad= metus barbadensis Pocock of Trinidad, W.L (Arachnida, Pedipalpi). Zoologica (New York), 47:25-37. Alexander, R.D. & D.W. Ewer. 1957. On the origin of mating behavior in spiders. American Nat., 91:311-317. Delle Cave, L. 1986. Biospeleology of the Soma- liland Amblypygi (Arachnida, Chelicerata) of the caves of Showli Beerdi and Mugdile (Bardera, Somaliland). Redia (Florence), 69:143-170. Eberhard, W.G. 1985. Sexual selection and animal genitalia. Harvard Univ. Press, Cambridge, Lon- don. Klingel, H. 1963. Paarungsverhalten bei Pedipal- pen (Thelyphonus caudatus L., Holopeltidia, Uropygi, und Sarax sarawakensis Simon, Char- ontidae, Amblypygi). Verb. dt. zool. Ges., 1962: 452-459. Kraus, O. 1970. Genitalmorphologie und Syste- matik der Amblypygi (Arachnida). Bull. Mus. Nat. Hist. Natur. 2^ ser., 41, Suppl. no. 1, 1969: 176-180. Mullinex, C.L. 1975. Revision of Paraphrynus Moreno (Amblypygida: Phrynidae) for North America and the Antilles. Occ. Papers California Acad. Sci., 116:1-80. Pocock, R.I. 1894. Notes on the Pedipalpi of the family Tarantulidae contained in the collection of the British Museum. Ann, Nat. Hist. Ser. 6, 14: 273-258. Quintero, D. 1980. Systematics and evolution of Acanthophrynus Kraepelin (Amblypygi, Phryni- dae). Pp. 341-347. In 8th Intern. Congr. Arach- noL, Wien. Quintero, D. 1981. The amblypygid genus in the Americas (Amblypygi, Phrynidae). J. Ar- achnoL, 9:117-166. Quintero, D. 1983. Revision of the amblypygid spiders of Cuba. Studies Fauna Curasao and oth- er Caribbean Islands, 196:1-54. Quintero, D. 1986. Revision de la classification de Amblypygidos pulvinados: Creation de subor- denes, una nueva familia y un nuevo genero con tres nuevas especies (Arachnida, Amblypygi). Pp. 203-212. In Proc. 9"" Intern. Congr. Arach- noL, Panama 1983, Smithsonian Inst. Press. Werner, E 1935 Scorpiones, Pedipalpi. Pp. 1-490. In Bronns Klassen und Ordnungen des Tierreich- es, 5. Bd, IV Abt. 8. Buch, Akad. Verlagsgesell- schaft, Leipzig. Weygoldt, P. 1969. Beobachtungen zur Fortpflan- zungsbiologie und zum Verhalten der GeiBelspinne Tarantula marginemaculata C.L. Koch (Chelicerata, Amblypygi). Z. Morph. Ti- ere, 64:338-360. Weygoldt, P. 1970. Lebenszyklus und postem- bryonale Entwicklung der GeiBelspinne Taran- tula marginemaculata C.L. Koch (Chelicerata, Amblypygi) in Laboratorium. Z. Morph. Tiere, 67:58-85. Weygoldt, P. 1972a. Charontidae (Amblypygi) aus Brasilien: Beschreibung von zwei neuen Chari- nus-P^Qu, mit Anmerkungen zur Entwicklung, 116 THE JOURNAL OF ARACHNOLOGY Morphologic und Tieregeographie und mit einem Bestimmungsschliissel fiir die Gattung Charinus. ZooL Jb. Syst., 99:107-132. Weygoldt, R 1972b. Spermatophorenbau und Sa- meniibertragung bei Uropygen {Mastigoproctus brasilianus C.L. Koch) und Amblypygen {Char- inus brasilianus Weygoldt und Admetus pumilio C.L. Koch) (Chelicerata, Arachnida). Z. Morph. Tiere, 71:23-51. Weygoldt, P. 1972c. Charinus koepkei n, sp. aus Peru (Abmlypygi: Charontidae). Senckenbergi- ana BioL, 53:281-286. Weygoldt, P. 1974a. Kampf und Paarang bei der GeiBelspinne Charinus montanus Weygoldt (Arachnida, Amblypygi, Charontidae). Z. Tier- psychoL, 34:217-223. Weygoldt, P. 1974b. Vergleichende Untersuchun- gen an zwei Heterophrynus {Admetus)- ArtGH, H. longicornis Butler und H. batesii Butler (Arach- nida, Amblypygi, Tarantulidae). Zool. Anz. Jena., 192:175-191. Weygoldt, P. 1977. Kampf, Paarangsverhalten, Spermatophoren-morphologie und weibliche Genitalien bei neotropischen GeiBelspinnen (Amblypygi, Arachnida). Zoomorphologie, 86: 271-286. Weygoldt, P. 1990. Arthropoda -Chelicerata: Sperm transfer. Pp. 77-119. In Reproductive Bi- ology of Invertebrates. Vol. IV, Part B. Fertiliza- tion, development, and parental care. (K.G. & R.G. Adiyodi, eds.). Oxford & IBH Publishing Co., New Dehli, Bombay, Calcutta. Weygoldt, P. 1996a. The relationships of the South East African whip spiders Hemiphrynus macha- doi Fage, 1951 and Phrynichus scullyi Purcell, 1901: Introduction of the new generic names Xe- rophrynus and Phrynichodamon (Chelicerata: Amblypygi). Zool. Anz., 235:117-130. Weygoldt, P. 1996b. Evolutionary morphology of whip spiders: towards a phylogenetic system (Chelicerata: Arachnida: Amblypygi). J. Zoo. Syst. Evol. Research, 34:185-202. Weygoldt, P. 1998a. Mating and spermatophore morphology in whip spiders {Phrynichodamon scullyi (Purcell, 1901), Damon gracilis nov. spec., Damon variegatus (Perty, 1834), and Eu- phrynichus bacilUfer (Gerstaecker, 1873)) (Arachnida: Amblypygi: Phrynichidae). Zool. Anz., 236:259-276. Weygoldt, P. 1998b. Revision of the species of Phrynichus Karsch, 1879 and Euphrynichus Weygoldt, 1995 (Chelicerata, Amblypygi). Pp. 1-65. In Zoologica (Stuttgart), Heft 147, E. Schweizerbart’sche Verlagsbuchhandlung, Stutt- gart. Weygoldt, R & P. Hoffmann. 1995. Reproductive behavior, spermatophores, and female genitalia in the whip spiders Damon diadema (Simon, 1876), Phrynichus cf. ceylonicus (C.L. Koch, 1843) and Euphrynichus alluaudi (Simon, 1936) (Chlicerata: Amblypygi). Zool. Anz., 234:1-18. Weygoldt, R, A. Weisemann & K. Weisemann. 1972. Morphologisch-histologische Untersu- chungen an den Geschlechts-organen der Ambly- pygi unter besonderer Berucksichtigung von Ta- rantula marginemaculata C.L. Koch (Arach- nida). Z. Morph. Tiere, 73:209-247. Manuscript received 1 May 1998, revised 6 Septem- ber 1998. 1999 The Journal of Arachnology 27:117-122 ONTOGENY OF CHARACTERISTIC LEG MACROSETAE IN MIMETUS (ARANEAE, MIMETIDAE) Bruce Cutler: Electron Microscopy Laboratory and Department of Entomology, University of Kansas, Lawrence, Kansas 66045-2106 USA Hank Guarisco: RO. Box 3171, Lawrence, Kansas 66046 USA Daniel J. Mott: Department of Biological and Physical Sciences, Lincoln Land Community College, 5250 Shepherd Road, Springfield, Illinois 62794“9256 USA ABSTRACT. The distinctive prolateral spination of the metatarsi and tibiae of the first two legs in Mimetus is obscure in the first post-eggsac eclosion instar. Only one of the small, acuminate tipped macrosetae appears in the first instar, small macroseta numbers increase in the second instar, and outnumber the large macrosetae by the third instar. The high variability in adult macroseta counts occurs in the third instar as well. The characteristic macrosetae have a socketed base and longitudinally grooved shafts. The large macrosetae are characterized by numbers of small pustules on the base below the emergence of the shaft and the tips of the macrosetae are round. The small macrosetae have fewer pustules or none, and the tips of the macrosetae are falcate and acuminate. Both the large and small macrosetae morphologically resemble presumptive mechano-receptive setae on the legs, and may have a sensory function. The Mimetidae is a worldwide family of araneophagic spiders, although they will also feed on insect prey captured by other spiders and, rarely, on non-snared insects (Cutler 1972; Jackson & Whitehouse 1986; Lawler 1972). The family is characterized by the dis- tinctive spination of the prolateral surfaces of metatarsi and tibiae of the first two pairs of legs in all females and most males (Forster & Platnick 1984; Heimer 1986; Platnick & Shadab 1993). This spination consists of two different types of macrosetae, referred to as spines by earlier authors. In the adults of Mi- metus, this spination consists of a series from the distal part of the segment to the proximal part of smaller macrosetae growing smaller in length, followed by a distinctly larger macro- seta, another series of small macrosetae de- creasing in length, a large macroseta, and so forth with the numbers of series dependent on the leg segment and the species (Fig. 1). Pre- viously, we noted that spiderlings of Mimetus emerging from the eggsac lacked the charac- teristic spination, which provided the impetus for this study. METHODS Spiderlings were reared from the first post- eggsac eclosion instar through the third post- eggsac eclosion instar. Eggsacs were collected in the field and were also produced by females in the laboratory. The eggsacs of Mimetus no- tius Chamberlin 1923 and M. puritanus Chamberlin 1923 are morphologically distinc- tive (Guarisco in press; Guarisco & Mott 1990) and can be readily distinguished. Specimens of M. notius were obtained from Bexar and Medina Counties, Texas and of M. puritanus from Douglas County, Kansas. Spi- derlings were kept in glass scintillation vials (45 mm tall X 25 mm diameter) at ambient indoor room temperatures (about 20 °C) and varying light conditions. Food was predomi- nantly first and second instar Achearanea tep- idariorum (C.L. Koch 1841) (Theridiidae), augmented by first and second instar Latro- dectus mactans (Fabricius 1775) (Theridi- idae), Argiope aurantia Lucas 1833 and Neos- cona sp. (Araneidae), and Agelenopsis sp. (Agelenidae). Two to three days after eggsac emergence (first instar) or after molting (second and third instar) specimens were preserved in 70% eth- anol. Samples for scanning electron micros- copy (SEM) observations were dehydrated in a graded ethanol series to acetone, air-dried out of acetone, mounted on conductive glue tabs on stubs, sputter coated with 40 nm of 117 118 THE JOURNAL OF ARACHNOLOGY Figure 1. — Mimetus puritanus, female from Douglas County, Kansas. Tibia, leg II, top is prolateral, left is distal. Scale bar == 400 jxm. gold: palladium alloy (60:40) and examined by using a Hitachi S570 SEM with a LaB^ filament. Characteristic spine patterns were determined by examining all specimens avaiL able using a dissecting microscope with max- imum magnification of 120X. For difficult de- terminations, specimens were prepared for SEM examination as above. In the tables and text, the smaller macrosetae are indicated by an S, the larger macrosetae by an L, and the order is from distal to proximal. RESULTS All micrographs are of Mimetus puritarius, because of better overall preservation of the material. Unless specified, all macrosetae re- ferred to in this paper are the prolateral ma- crosetae that form the characteristic spination. As can be seen in Table 1, only one small macroseta was found in the first instar of both species. All of these macrosetae are about the same length, and there are no interspersed ma- crosetae of any type. The numbers of these macrosetae did not vary within the species for the specimens examined (Figs. 2-“5). In the second instar (Figs. 6-9) the number of large macrosetae outnumbers the number of small macrosetae. There is no variation in pattern in M. notius, but some variation in pattern in M. puritanus (Table 1). In the third instar one sees for the first time the typical adult macro - seta pattern, albeit in a reduced form, and Fig. 10 shows an example. Compared to the earlier instars, the number of small macrosetae in- creases, the relative lengths of the small ma- crosetae versus the large macrosetae is that of the adult pattern, and the number of small ma- crosetae per segment is greater than that of the large macrosetae. A total of seven third instar M. notius and five third instar M. puritanus was examined. The amount of variability was so great that displaying the information in a Table 1 . — Macroseta types and counts in the two first instars of Mimetus notius and Mimetus puritanus. Numbers in parentheses are numbers of individuals with a particular macroseta count (if not specifically indicated, the counts are for all «); macroseta counts are listed from distal to proximal positions on leg segment. Each specimen had the same spination on right and left corresponding leg segments S = small macroseta, L = large macroseta. Species Instar n Metatarsus 1 Tibia Metatarsus 2 Tibia 2 M. notius first 23 S, 3L 4L 3L 2L M. notius second 17 2S, L, S, 2L, S, L 6L 2S, 3L S, 2L, S M. puritanus first 8 S, 3L 4L 3L 3L M. puritanus second 9 2S, L, S, 3L 5L(8) 2S, 3L(8) S, 2L(4) Damage(l) 2S,2L(1) S, 3L, S(2) S, 2L, S(2) S,L, S,L(1) CUTLER ET AL.— LEG MACROSETAE ONTOGENY IN MIMETUS 119 Figures 2-5. — Mimetus puritanus, first instar from left is distal. 2, Metatarsus I, small macroseta socket I, scale bar = 50 [xm; 4, Metatarsus II, scale bar = large macrosetae. Douglas County, Kansas. Top of image is prolateral, indicated by arrowhead, scale bar = 50 (xm; 3, Tibia 40 jxm; 5, Tibia II, scale bar — 40 |xm. Asterisks = table would be unwieldly. In 32% of the seg- ments examined, the corresponding left and right segments on the same specimen had dif- ferent macrosetae type counts. As might be expected there were also differences among the individuals. In all instars examined there are differences in ultrastructural morphology between large and small macroseta of Mimetus. The macro- setae emerge from a socket base, both the base and the macroseta shaft have linear ridges, and some of the ridges are hackled (Figs. 11, 12, 15). On the base of the large macrosetae below the emergence of the macroseta shaft there is a large number of approximately 0.5- 1.0 p,m pustules (Fig. 11). These occur in smaller numbers on the larger of the small macrosetae, but are absent in most of the small macrosetae and the sensory setae on the leg segments (Figs. 12, 15). Another differ- ence between the large macrosetae and the small is that the tips of the large macrosetae are rounded and the macroseta shaft is straight or gently curved, while small macrosetae have falcate, acuminate tips and the macroseta shaft is strongly curved (Figs. 13, 14). The differ- ence in tip shape becomes more pronounced in the later instars. DISCUSSION Leg segment macroseta counts in adult Mi- metus are very variable. Total macroseta counts for Leg I and II metatarsi and tibiae are given in Table 2 (from Mott 1989). The counts are for all macrosetae, not just those that form the characteristic spination. Since the other macrosetae are constant in number, the variation results from the macrosetae mak- ing up the characteristic spination. There are no observations, including the most detailed study of mimetid behavior (Jackson & Whi- tehouse 1986), that indicate a specific function for the macrosetae. Anecdotal observations of the three authors indicate that the macrosetae form a trapping basket in drawing the prey’s leg to the chelicerae. They may have a sen- sory function, since they strongly resemble the smaller presumed, mechano-receptive se- tae on the legs. These serrate setae are mor- phologically very similar to the closed tactile 120 THE JOURNAL OF ARACHNOLOGY Figures 6-10. — Mimetus puritanus, from Douglas County, Kansas. Top of image is prolateral, left is distal. 6-9, Second instar. 6, Metatarsus I, scale bar = 60 pm; 7, Tibia I, scale bar = 70 pm; 8, Metatarsus II, scale bar = 60 pm; 9, Tibia II, scale bar = 80 pm. Asterisks = large macrosetae, arrowheads = small macrosetae; 10, Third instar metatarsus I, macrosetae types obvious, reduced version of adult pattern, scale bar = 90 pm. setae of Amaurobius (reported as Ciniflo, Har= ris & Mill 1977). The shape of the socket ba- ses and the ultrastructure of the shaft is very similar in the setae and macrosetae (Fig. 15). The macrosetae are not simple cuticular pro- jections since they are socketed. Harris & Mill (1977) showed through manipulation that erecting the leg macrosetae in Amaurobius cause an electro-physiological response. These macrosetae are morphologically differ- ent from those of Mimetus; however, some sort of tactile response seems a likely func- tion, although probably different from that of Amaurobius. The differences in the pustule details and the shape of the macroseta tips provide a way to distinguish the large and small macrosetae in early instars where the discrepancy in the lengths of the macrosetae is much less than in the later instars and adults, and results in difficulty in determining CUTLER ET AL.— LEG MACROSETAE ONTOGENY IN MIMETUS 121 Figures 11-15. — Mimetus puritanus, from Douglas County, Kansas. 11, Base of large macroseta, meta- tarsus II, third instar, note pustules on base of macroseta socket at lower left, scale bar = 4 pm, inset, pustules from large macroseta of adult female, scale bar =1.5 pm; 12, Base of small macroseta, metatarsus II, third instar, note lack of pustules on macroseta socket base, scale bar = 3.5 pm; 13, Tip of large macroseta, metatarsus I, adult female, scale bar = 3.5 pm; 14, Tip of small macroseta, metatarsus I, adult female, scale bar = 3.5 pm; 15, Large macroseta (top), closed tactile seta (below), metatarsus II, first instar, scale bar = 3.6 pm. 122 THE JOURNAL OF ARACHNOLOGY Table 2. — Total macrosetae per leg segment in adult female Mimetus notius (n = 10) and Mimetus puritanus (n = 10). Specimens from eastern United States. S. E. = standard error of the mean (from Mott 1989). Meta- tarsus 1 Tibia Meta- tarsus 2 Tibia 2 M. notius Range 35-48 31-47 21-30 20-32 Mean 41.9 39.9 26.2 26.4 S. E. 1.760 1.560 1.052 1.046 M. puritanus Range 33-53 29-44 21-33 19-29 Mean 40.2 34.5 25.4 23.9 S. E. 2.081 1.614 1.204 1.048 which macroseta type is present. Useful ma- croseta characteristics in separating the two species occur in the first instar on the second tibia, i.e., there are two large macrosetae in M. notius and three in M. puritanus. In the second instar there are six large macrosetae on the first tibia in M. notius, but only five in M. puritanus. Since other species of Mimetus occur in the range of the two species dis- cussed here (Mott 1989) and no descriptions of the early instars in these species exist, at this point the macroseta patterns do not have diagnostic value for field collected material. However, once patterns for the early instars of other species becomes available, then these patterns may have diagnostic value. ACKNOWLEDGEMENTS We wish to thank B. Opell and an anony- mous reviewer for valuable comments on the manuscript. LITERATURE CITED Cutler, B. 1972. Notes on the biology of Mimetus puritanus Chamberlin (Araneae: Mimetidae). American Midi. Nat., 87:554-555. Forster, R.R. & N.I. Platnick. 1984. A review of the archeid spiders and their relatives, with notes on the limits of the super family Palpimanoidea (Arachnidae, Araneae). Bull. American Mus. Nat. Hist., 178:1-106. Guarisco, H., In press. Description of the eggsac of Mimetus notius (Araneae: Mimetidae) and a case of egg predation by Phalacrotophora epeirae (Diptera, Phoridae). J. ArachnoL, 00:000-000. Guarisco, H. & D.J. Mott. 1990. Status of the ge- nus Mimetus (Araneae: Mimetidae) in Kansas and a description of the eggsac of Mimetus pur- itanus Chamberlin. Trans. Kansas Acad. Sci., 93: 79-84. Harris, D.J. & PJ. Mill. 1977. Observations on the leg receptors of Ciniflo (Araneida: Dictynidae). 1. External mechanoreceptors. J. Comp. Physiol., 119:37-54. Heimer, S. 1986. Notes on the spider family Mi- metidae with description of a new genus from Australia. Entomol. Abh. Mus. Tierk. Dresden, 49:113-137. Jackson, R.R. & M.E.A. Whitehouse. 1986. The biology of New Zealand and Queensland pirate spiders (Araneae, Mimetidae): aggressive mim- icries, araneophagy and prey specialization. J. Zool., London, 210:279-303. Lawler, N. 1972. Notes on the biology and behav- ior of Mimetus eutypus Chamberlin and Ivie (Ar- aneae: Mimetidae). Notes ArachnoL Southwest, 3:7-10. Mott, D.J. 1989. A revision of the genus Mimetus in North America (Araneae, Mimetidae). Ph.D. Thesis, Southern Illinois Univ., Carbondale. 182 pp. Platnick, N.I. & M.U. Shadab. 1993. A review of the pirate spiders (Araneae, Mimetidae) of Chile. American Mus. Nov., 3074:1-30. Manuscript received 28 April 1998, revised 24 No- vember 1998. 1999. The Journal of Arachnology 27:123-128 VENTRAL MESOSOMAL CHANGES IN EMBRYOS FROM THREE SCORPION FAMILIES: lURIDAE, BUTHIDAE AND VAEJOVIDAE Roger D. Farley; Department of Biology, University of California, Riverside, California 92521 USA ABSTRACT. The scanning electron microscope was used to examine embryos at a stage when book- lungs and spiracles are forming. Earlier studies with scorpion fossils suggest there was ventral mesosomal transition from gills or booklungs above ventral plates to stemites, booklungs and spiracles. In Hadrurus arizonensis (luridae), ventral plates and then sternites are formed on the ventral surface of mesosomal segments before spiracles appear. Bilateral invaginations in body segments XII-XV apparently give rise to the booklungs, with spiracles formed lateral to the site of invagination. Sternites with bilateral depres- sions were also present before spiracles in embryos of the buthid Centruroides exilicauda. In the devel- opmental stages herein examined, spiracles were formed in embryos of Paruroctonus mesaensis (Vaejov- idae); but there was no indication of ventral plates or sternites on the ventral mesosoma. Spiracles appear in the intersegmental area posterior to body segments XIII-XV. Booklungs may form later from primordia associated with bilateral depressions observed in a later stage in these segments. The earliest scorpion fossils (Silurian) sug- gest these animals were aquatic, while all sur- viving species are terrestrial (Selden & Jeram 1989; Sissom 1990; Jeram 1994). A critical stage in scorpion evolution was the change (ventral mesosoma) from gills to booklungs, probably in the Permian and Carboniferous periods. Kjellesvig-Waering (1986) provided some evidence that aquatic scorpions had gills above ventral plates in the ventral mesosoma. He proposed that there was gradual reduction of these plates and formation of stemites, booklungs and spiracles. Selden & Jeram (1989) and Jeram (1990) described a fossil Carboniferous scorpion with booklungs rather than gills above ventral plates. Scorpion embryos were examined with the possibility they might provide some informa- tion about the water-to-land transition (Farley 1999a, b). These initial observations showed some differences in the ventral mesosoma during spiracle and booklung formation in embryos of the vaejovid, Paruroctonus me- saensis, and the iurid, Hadrurus arizonensis. The present study is an extension of that work, including embryos of Centruroides exilicau- da, a buthid. The latter was examined since buthids are considered most primitive among extant scorpion families (Stockwell 1989, 1992; Sissom 1990), and mesosomal changes may reflect the ancestral condition. METHODS The composition of physiological saline and the procedures for collection and main- tenance of specimens were described in an earlier publication (Farley 1987). Specimens of Paruroctonus mesaensis Stahnke 1957 were collected in the Colorado Desert near In- dio and Palm Springs, California. Specimens of Hadrurus arizonensis Ewing 1928 (Wil- liams 1970; Francke & Soleglad 1981) and Centruroides exilicauda Wood 1863a (Wood 1863b; Ewing 1928; Williams 1980) were col- lected in Arizona. Specimens of all three spe- cies are in the California Academy of Science, San Francisco. Tissues were flushed with saline to remove debris as animals were dissected with micro- scissors and forceps. The ovariuterine tubules were opened and embryos removed. Sur- rounding membranes (amnion, serosa) were pulled away with microprobe and forceps. Tissues were fixed (6-10 h, 23-25 °C) with 4% glutaraldehyde in 0. 1 M cacodylate buffer with one drop of calcium chloride for each 10 ml of solution (Lane et al. 1981). The tissues were washed in cacodylate buffer-NaCl solu- tion and postfixed (2 h, 23-25 °C) in 1% os- mium tetroxide in 0.2 M cacodylate buffer with NaCl. The concentrations of these solu- tions were adjusted to approximate the os- 123 124 THE JOURNAL OF ARACHNOLOGY molality of scorpion blood (630 mOsm; Yok“ Ota 1984). Tissues were dehydrated in acetone, critically^point dried (Balzers, CDD 020) and sputter-coated (EMscope SC500) with 20 nm thickness of gold/palladium. Tis- sues were examined at 12-15 KV with a Phil- ips 15 scanning electron microscope (SEM). RESULTS At a stage before spiracle and booklung for- mation, embryos of H. arizonensis have plates demarcated on the ventral surface of meso- somal segments. Initially, only a narrow ridge outlines the ventral plates, with the delineated region much smaller than the ventral surface of the segment. The outlined region becomes a flap-like structure (Fig. 1) fused to the body wall anteriorly but free at the lateral and pos- terior margins. The early ventral plates do not extend the full width of the mesosoma nor overlap antero-posteriorly. Embryos were not sectioned, but no indications of an opening or gill-like structures were observed at the pos- terior margin of the ventral plates. Paired in- dentations in body segments XII-XV (Hjelle 1990) are presumably booklung primordia. In later stages, the invaginations in seg- ments XII-XV become more prominent (Fig. 2), and the ventral cuticle increases in length and width, forming structures that resemble adult stemites with the perimeter joined to pleural or intersegmental integument. The stemites extend the full width of the meso- soma and overlap in the longitudinal axis. In- trastemal spiracles eventually form at the adult location (Farley 1990a, b), just lateral to the site of booklung invagination. Booklungs do not develop in segment XVI; the indenta- tions (Fig. 2) eventually disappear, leaving no external trace. Mesosomal development in the buthid, C. exilicauda, appears to be similar to that of H. arizonensis. In Fig. 3, an embryo of the former species has stemites with bilat- eral depressions, presumably for booklungs. In embryos of P. rnesaensis, there is no de- marcation of ventral plates or stemites at the time when spiracles first appear (Fig. 4). In the stages observed in this study, spiracles were seen near the mesosomal midline in the intersegmental tissue posterior to segments XIII-XV. In later stages, bilateral depressions were seen in segments XII-XVI, but there was still no indication of ventral plates or stemites. Advanced embryos of P. mesaensis were not available to determine if booklungs and new spiracles form at these depression sites, or if the initial spiracles (Fig. 4) move to the adult position, farther anterior and lateral in the seg- ment (Farley 1990a, b). The early spiracles differed in shape among the embryos, but usu- ally had a smooth, apparently cuticular margin and a slit-like opening (Fig. 4), in comparison with the oval shape in the adult. Although ventral plates or stemites are not evident in embryos of P. mesaensis when spi- racles first appear (Fig. 4), the ventro- poste- rior margin of each mesosomal segment was examined for indications of invagination or gill-like structures. In some embryos, the lat- eral intersegmental area shows differentiation suggestive of infolding, with vertical striations (Fig. 4). The spiracles form in the medial as- pect of this distinctive intersegmental area. Embryos were not sectioned, but during dissection in transmitted light, some internal structures can be seen. In embryos of P. me- saensis, there was no indication of a thick- ening or density anterior to the spiracles, as would be expected if booklungs were forming. The spiracle site in the intersegmental area (Fig. 4) did not appear to be a region of in- vagination as occurs in the bilateral depres- sions seen in the mesosomal segments of the iurid and buthid embryos (Figs. 1-3). DISCUSSION In H. arizonensis and C. exilicauda, book- lung and spiracle formation appears to be like that described by earlier workers in species from the families Buthidae (Abd-el Wahab 1951), Chactidae (Laurie 1890; Brauer 1895) and Scorpionidae (Metschnikoff 1871; Laurie 1892). The bilateral depressions evident in mesosomal segments in Figs. 1-3 appear to be sites of invagination, and spiracles are later formed here at the location seen in adults (Farley 1990a, b). The early demarcation of flap-like structures (Fig. 1) supports the notion that ventral plates preceded (Kjellesvig-Waer- ing 1986) or occurred with booklungs in an- cient scorpions (Selden & Jeram 1989; Jeram 1990). Differences were reported among scorpion species in the shape and texture of the cuticle of adult booklung lamellae (Lankester 1885; Berteaux 1889; Laurie 1896a, b). These were proposed as taxonomic criteria, but other fea- tures subsequently found acceptance (Stock- FARLEY— VENTRAL MESOSOMA IN SCORPION EMBRYOS 125 Figures 1, 2. — SEMs of ventral surface of mesosoma of embryos of Hadrurus arizonensis. 1. Flap-like ventral plates (P) are fused to the body wall anteriorly and free at the posterior margin. Bilateral invagi- nations (arrows) are present where spiracles and booklungs will eventually form. Teeth (T) are evident at the posterior edge of the pectines. XIV, body segment; 2, Later stage. The ventral cuticle of each segment has broadened and is now a stemite (S) attached around the entire perimeter. Bilateral invaginations (black arrows) have deepened. Shallow depressions occur in body segment XVI, but booklungs do not develop in this segment. The white arrow indicates a pair of small, transitory appendages of unknown significance between gonopore and pectine. A, remnants of amnion not removed during preparation; T, pectinal teeth. Scales, 0,5 mm. 126 THE JOURNAL OF ARACHNOLOGY Figures 3, 4. — SEMs of ventral surface of mesosoma of embryos. 3, Centruroides exilicauda. Each segment has a sternite (S) like that of the irurid embryo of Figure 2. Bilateral invaginations (arrows) are presumably the site of booklung formation. Depressions are evident in body segment XVI although book- lungs do not form in this segment. L, fourth walking leg. T, pectinal teeth. Scale, 0.5 mm; 4, Paruroctonus mesaensis, left side of ventral mesosoma. No ventral plates or stemites are evident, but spiracles (arrows) are present at the posterior margin of body segments XIII-XV. The spiracles are at the medial end of an invaginated intersegmental region (★) with vertical striations. L, fourth walking leg. Scale, 0.2 mm. FARLEY— VENTRAL MESOSOMA IN SCORPION EMBRYOS 127 well, 1989, 1992; Sissom 1990)= Developing booklungs were previously described as bilat- eral invaginations in the ventral mesosoma (Metschnikoff 1871; Laurie 1890, 1892; Brauer 1895; Abd-el Wahab 1951). Tissue sections showed that sac-like invaginations extend anteriorly in the segment from the ini- tial site of ingress, which remains open to be- come the spiracle. A few lamellae are initially formed in the horizontal plane. These later ro- tate 90° to the dorso-ventral axis, along with development of many more lamellae (Laurie 1890, 1892). There may be absence or delay of ventral plates, and stemites may form late in embryos of P. mesaensis in comparison with the iurid and buthid embryos. Among scorpion fami- lies, heterochrony occurs in embryogenesis in relation to the mode of maternal nourishment of the embryos (Matthew 1959; Farley 1999a, b). All extant scorpions have adaptations for terrestrialization {i.e., oral tube, booklungs, latterly compressed podomeres), but may be polyphyletic with convergent evolution (Jer- am 1994). The possibility of a different vae- jovid derivation is raised in the present studies by the delay or absence of ventral plates (Fig. 4) and the development of spiracles at the me- dial end of lateral intersegmental specializa- tions that may be indicative of ancestral re- spiratory structures. Fossils of British Triassic scorpions have slit-like spiracles in the inter- segmental membrane of mesosomal segments or in the latero-posterior margin of the abdom- inal plates (Wills 1947). Tissue sections are needed to determine if booklung formation is also distinctive in P. mesensis. The lack of tissue invagination at the place where spiracles first appear in the intersegmental area (Fig. 4) suggests this is not the site of booklung primordia. These spi- racles may migrate from the intersegmental area to the adult position more anterior and lateral in the segment (Farley 1990a, b). An- other possibility is that the early spiracles in Fig. 4 are transitory, and new spiracles form later with booklungs more anterior in the seg- ments. Kjellesvig-Waering (1986) proposed that ventral plates were abdominal flaps or ap- pendages that overlay the body wall beneath, and stemites developed as the abdominal plates were reduced and eventually lost. From their review of fossil evidence, Selden & Jer- am (1989) considered it more likely that ven- tral plates later became stemites by fusion with the body wall. The latter proposal is sup- ported in the present study in embryos of H. arizonensis. Small regions, initially outlined by a ridge on the ventral surface of mesoso- mal segments, become flap-like plates (Fig. 1) and then the ventral cuticle is broadened to form stemites (Fig. 2). There was no indica- tion of reduction or loss of the ventral plates, resulting in exposure of overlying stemites. ACKNOWLEDGMENTS The author wishes to thank Kari J. Me West for providing gravid specimens of C exilicau- da. Thanks also to Morice A. Izmane for as- sistance in collecting the other species. This research was supported by intramural funds from the University of California. LITERATURE CITED Abd-el Wahab, A. 1951. Some notes on the seg- mentation of the scorpion Buthus quinquestriatus (H.E.). Proc. Egyptian Acad. Sci., 7:75-91. Berteaux, L. 1889. Le poumon. La Cellule, 5:255- 316. Brauer, A. 1895. Beitrage zur kenntnis der en- twicklungsgeschichte des skorpions. IT Zeit- schrift fur wissenschaftliche Zoologie, 59:351- 435. Ewing, H.E. 1928. The scorpions of the Western part of the United States. U.S. Natl. Mus. Wash- ington, 73:27-30. Farley, R.D. 1987. Postsynaptic potentials and con- traction pattern in the heart of the desert scorpi- on, Paruroctonus mesaensis. Comp. Biochem. Physiol., 86A:121-131. Farley, R.D. 1990a. Functional organization of the respiratory and circulatory systems in the desert scorpion, Paruroctonus mesaensis. Acta Zool. Fennica, 190:139-145. Farley, R.D. 1990b. Regulation of air and blood flow through the booklungs of the desert scor- pion, Paruroctonus mesaensis. Tissue «& Cell, 22:547-569. Farley, R.D. 1999a. Scorpiones. Pp. 117-222. In Microscopic Anatomy of Invertebrates (F. W. Harrison, ed.), Vol. 8A. Chelicerate Arthropoda (EW Harrison & R.E Foelix, eds). Wiley-Liss, New York. Farley, R.D. 1999b. Structure, reproduction and development. Pp. 25-98. In Scorpion Biology and Research. (P.H. Brownell & G.A. Polls, eds.). Oxford University Press, Oxford/New York. Francke, O.E & M.E. Soleglad. 1981. The family luridae Thorell (Arachnida, Scorpiones). J. Ar- achnoL, 9:233-258. 128 THE JOURNAL OF ARACHNOLOGY Hjelle, J.T 1990. Anatomy and morphology. Pp. 9-63. In The Biology of Scorpions (G.A. Polls, ed.). University Press, Stanford, California. Jeram, A.J. 1990. Book-lungs in a lower Carbon- iferous scorpion. Nature, 343:360-361. Jeram, A.J. 1994. Scorpions from the Visean of East Kirkton, West Lothian, Scotland, with a re- vision of the infraorder Mesoscorpionina. Trans. Roy. Soc. Edinburgh: Earth Sci., 84:283-288. Kjellesvig-Waering, E.N. 1986. A restudy of the fossil scorpionida of the world. Palaeontographi- ca Americana, 55:1-287. Lane, N.J., J.B. Harrison & R.F. Bowerman. 1981. A vertebrate-like blood-brain barrier with intra- ganglionic blood channels and occluding junc- tions, in the scorpion. Tissue & Cell, 13:557- 576. Lankester, E.R. 1885. Notes on certain points in the anatomy and generic characteristics of scor- pions. Trans. Zool. Soc. (London), 11:372-384. Laurie, M. 1890. The embryology of a scorpion {Euscorpius italicus). Quart. J, Microsc. Sci., Ser. 2., 31:105-141. Laurie, M. 1892. On the development of the lung- books in Scorpio fulvipes. Zool. Anz., 15:102- 105. Laurie, M. 1896a. Notes on the anatomy of some scorpions, and its bearing on the classification of the order. Ann. Mag. Nat. Hist., ser. 6, 17:185- 193. Laurie, M. 1896b. Further notes on the anatomy and development of scorpions, and their bearing on the classification of the order. Ann. Mag. Nat. Hist, ser. 6, 18:121-133. Mathew, A.R 1959. Some aspects of the embry- ology of scorpions. J. Zool. Soc. India, 11:85- 88. Metschnikoff, E. 1871. Embryologie des scorpi- ons. Zeits. fiir Wissenschaft, ZooL, 21:204-232. Selden, P.A. & A.J. Jeram. 1989. Palaeophysiology of terrestialization in the Chelicerata. Trans. Roy. Soc. Edinburgh: Earth Sci., 80:303-310. Sissom, W.D. 1990. Systematics, biogeography and paleontology. Pp. 64-160. In The Biology of Scorpions (G.A. Polis, ed,). University Press, Stanford, California. Stahnke, H.L. 1957, A new species of scorpion of the Vejovidae: Paruroctonus mesaensis. Ento- mol. News, 68:253-259. Stockwell, S.A. 1989. Revision of the phylogeny and higher classification of scorpions (Cheiicer- ata). Ph.D. dissertaion, Univ. California, Berke- ley. Stockwell, S.A. 1992. Systematic observations on North American scorpionida with a key and checklist of the families and genera. J. Med. En- tomoL, 29:407-422. Williams, S.C. 1970. A systematic revision of the giant hairy- scorpion genus Hadrurus (Scorpion- ida: Vaejovidae). Occas. Pap. California Acad. Sci., 87:1-62. Williams, S.C. 1980. Scorpions of Baja California, Mexico, and adjacent islands. Occas. Pap. Cali- fornia Acad. Sci., 135:1-127. Wills, L.J. 1947. A monograph of British Triassic scorpions. Monogr. PalaeontoL Soc. (London), vol. 100/101. 137 pp. Wood, H.C. 1863a. Descriptions of new species of North American pedipalpi. Proc. Acad. Nat. Sci. Philadelphia 1863, pp. 107-112. Wood, H.D. 1863b. On the pedipalpi of North America. J. Acad. Nat. Sci. Philadelphia, 2nd sen, 5:357-376. Yokota, S.D. 1984. Feeding and excretion in the scorpion Paruroctonus mesaensis: water and ma- terial balance. J. Exp. Biol., 110:253-265. Manuscript received 24 April 1998, revised 25 Sep- tember 1998. 1999. The Journal of Arachnology 27:129-134 THE USE OF MORPHOMETRIC CHARACTERISTICS FOR THE RECOGNITION OF SPECIES AMONG GONIOSOMATINE HARVESTMEN (ARACHNIDA, OPILIONES, GONYLEPTIDAE) Pedro Gnaspini: Departamento de Zoologia, Instituto de Biociencias, Universidade de Sao Paulo, Caixa Postal 11461, 05422-970, Sao Paulo, SP, Brasil ABSTRACT. Morphometric data from males of six species of Goniosoma are presented and their im- portance in characterization and recognition of the species is discussed. Data presented show that it is important to use intraspecific variation during descriptions of these harvestmen. Harvestmen (Arachnida, Opiliones) are worldwide in distribution. Most papers on har- vestmen treat taxonomic aspects. Biological, ecological and behavioral surveys are some- what rare, especially those dealing with de- velopmental characterization. Most taxonomic papers that describe species present single measurements of body structures. This is, in part, because many species are known by only a single specimen. As discussed by McGhee (1977), the structures may vary among spec- imens of different populations or even of the same population. Gnaspini (1995) also showed that measurements vary due to devel- opmental differences. In the present paper, variation among different species is analyzed and discussed. The main goal of this paper is to show the importance of taking into account these aspects when diagnosing and describing species of harvestmen. Toward this goal, the morphometric characterization of adults of six different species of Goniosoma is presented and discussed. METHODS The species treated here were collected dur- ing a study focusing on the cavernicolous spe- cies Goniosoma spelaeum (Mello-Leitao 1933) in the Ribeira Valley, Sao Paulo State, southeastern Brazil (Gnaspini 1993). The spe- cies included in this study were collected in caves in Sao Paulo State and/or in the neigh- borhood of the study area, as follows: Gon- iosoma sp. 1 aff. badium - caves near Curitiba, Parana State; Goniosoma sp. 2 aff. badium - Guaricana Dam, near Curitiba, Parana State; Goniosoma longipes (Roewer 1931) - Caves near Ipeuna, Sao Paulo State; Goniosoma proximum (Mello-Leitao 1933) - forest in the Ribeira Valley (except when in two specific caves within the distribution of G. spelaeum - see Gnaspini 1996), Sao Paulo State; Gonio- soma spelaeum (Mello-Leitao 1933) - caves in the Ribeira Valley, Sao Paulo State; Gon- iosoma varium Perty 1833 - forest and inside the first 0.5 m of caves in the Ribeira Valley, Sao Paulo State. Collected animals were fixed in 40% ethyl alcohol and, after some hours, transferred to flasks with 70% alcohol. This procedure avoided hardening of the specimens, especial- ly their leg articulations, and facilitated easy measurements. A series of voucher specimens of all species treated herein is deposited in the Museu de Zoologia da Universidade de Sao Paulo (MZSP). Morphometrical and meristic characteristics were observed with a Wild M5A stereomicro- scope. The characteristics used in the analysis were the number of tarsal segments and the measurement of body width (maximum width of dorsal scutum), body length (length of the dorsal scutum), total length of the pedipalp and of the walking legs. The raw data have been statistically ana- lyzed using the Mols’ method (Mdls 1987 apud Neet 1993; Gnaspini 1995), which tests if two given sets have their values homoge- neous or heterogeneous, i.e., if they should be considered the same or different from each other. Neet (1993) provided an example of use with spiders and an algorithm for the test. Briefly described, the test first orders the val- ues analyzed and then groups them by values 129 130 THE JOURNAL OF ARACHNOLOGY Table 1. — Results of the Mols’s test for each pair of species studied, for body length measurements taken in adult males of Goniosoma. “ + ” = “securely heterogeneous” (a < 0.01), “±” = “probably heterogeneous” (0.01 < a < 0.05), “ — ” = “possibly homogeneous” (a > 0.05). The value of a is given between parentheses following the respective symbol. G. sp. 2 G. varium {n = 4) G. spelaeum G. proximum {n = 8) (u = 10) G. longipes in = 5) aff. badium {n = 8) G. sp. 1 aff. badium (n = 8) -(2.13) -(0.685) -(0.575) ±(0.041) -(1.46) G. sp. 2 aff. badium -(2.78) -(1.05) + (0.002) -(0.317) G. longipes -(0.709) -(0.167) -(2.31) G. proximum -(0.079) -(0.105) G. spelaeum -(3.23) as if to prepare a graph of distribution. FoU low^ing this the data are treated as two curves of distribution that may overlap. The percent- age of overlap is represented by a. If a is smaller than 5%, the test considers the distri- bution heterogeneous, and provides a limit value. Individuals with values smaller than this limit are considered as part of a group, and those with larger values are part of a sec- ond group. A series of these pair-wise com- parisons then determines the similarities of species. RESULTS The number of tarsal segments showed mostly the same ranges among the species studied (global range of 8-12 segments in Leg I, 14-25 in Leg II, 9-14 in leg III, and 9-15 in leg IV), although G. spelaum showed the larger values and G. varium showed the small- er values. Because the variation was not sta- tistically significant, the number of tarsal seg- ments proved not to be useful in the recognition among species. Mols’ test of body width measurements showed that this feature is not also useful in distinguishing the species. The resulting a ranged from 0.154 to 27.55. Figure 1 shows large overlaps among the measurements of the species. On the other hand, Mols’ analyses (Tables 1-6) showed that body length, the length of legs I-IV, and palpal length are useful mea- surements in distinguishing Goniosoma spe- cies. Figure 1 shows graphically the differ- ences in the means and variances of these measurements. However, the lengths of body and of palp are useful only for given pairs of species. For example, from Table 6 (palpal length), a (interval of significance of the test) is mostly larger than 5% for the pairs of spe- cies analyzed, being coded as a in the table. This means that the differences are not statistically significant. For pair G. spelaeum - G. longipes, a is smaller than 5%, being statistically significant, and coded as a “±” in the table. Finally, a is smaller than 1%, being coded as a “ + ” in the table, showing that the differences are highly significant for Table 2. — Results of the Mols’s test for each pair of species studied, for leg I length measurements taken in adult males of Goniosoma. “ + ” = “securely heterogeneous” (a < 0.01), “±” = “probably heterogeneous” (0.01 < a < 0.05), “ — ” — “possibly homogeneous” (a > 0.05). The value of a is given between parentheses following the respective symbol. G. sp. 2 G. varium in = 4) G. spelaeum {n = 8) G. proximum {n = 10) G. longipes aff. badium {n = 5) (« = 8) G. sp. 1 aff. badium (n — 8) G. sp. 2 aff. badium G. longipes G. proximum G. spelaeum + (0.0002) + (0.0018) -(0.407) + (0.000007) + (0.0008) ±(0.018) -(19.18) + (0.0012) -(2.76) + (0.0008) -(44.68) + (0.00002) + (0.0015) + (0.0048) + (0.007) GNASPINI— MORPHOMETRIC CHARACTERIZATION OE HARVESTMEN 131 Table 3. — Results of the Mols's test for each pair of species studied, for leg II length measurements taken in adult males of Goniosoma. “ + ” = “securely heterogeneous” (a < 0.01), “±” = “probably heterogeneous” (0.01 < a < 0.05), “ — ” = “possibly homogeneous” (a > 0.05). The value of a is given between parentheses following the respective symbol. G. varium {n = 4) G, spelaeum {n = 8) G. proximum {n = 10) G. sp. 2 G. longipes aff. badium (n = 5) (n = 8) G. sp. 1 aff. badium {n = 8) G. sp. 2 aff. badium G. longipes G. proximum G. spelaeum ±(0.037) -(0.059) -(1.12) + (0.00002) ±(0.054) -(0.075) -(8.38) + (0.0069) -(0.134) + (0.0000004) -(4.08) + (0.000001) + (0.0007) + (0.0036) ±(0.019) pairs G. proximum ~ G. sp. 2 aff. badium, G. proximum - G. spelaeum, and G. spelaeum ~ G. varium. When values of length of legs I, II, III and IV do not allow recognition, other structures can be used. One good example is the case of G. sp. 2 aff. badium and G. proximum, which can be easily distinguished by the length of the body and especially of the palp. The only two pairs of species which could not be statistically recognized using morpho- metries were G. sp, 2 aff. badium and G. spe- laeum, and G. longipes and G. varium. There- fore, other characters should be used, such as color, as treated below. DISCUSSION These data show that there are morphomet- ric, morphological and meristic variations be- tween and within species of harvestmen, as previously discussed by McGhee (1977) and Gnaspini (1995). These intraspecific varia- tions may be even very large, and may be due either to variation within the same develop- mental stage or between stages. In addition. adding measurements of single specimens during descriptions (even when series are available) does not allow precise species rec- ognition. For instance, if a given specimen of the same species is checked against measure- ments of only one described specimen, it may not fit, and would not be recognized as such species. However, if a range of variation was given, it would probably fit. In addition, data also show that the range of variation may help identify species. Of course, genitalia characterization and color patterns can also help distinguish species from each other. Sometimes, they do so alone; how- ever, in other cases they are conservative and misleading. In her unpublished revision of the genus Goniosoma, Stefanini- Jim (1985, 1995, pers. comm.) proposed synonymizing several spe- cies under G. badium, due to the conserva- tive shape of their penis. She also considers G. spelaeum to belong in a ’badium-group’, and possibly being synonymous with G. bad- ium, again because of the similar penis. She Table 4.— Results of the Mols’s test for each pair of species studied, for leg III length measurements taken in adult males of Goniosoma. “ + ” = “securely heterogeneous” (a < 0.01), “±” = “probably heterogeneous” (0.01 < a < 0.05), “ — ” = “possibly homogeneous” (a > 0.05). The value of a is given between parentheses following the respective symbol. G. sp. 2 G. varium {n = 4) G. spelaeum {n ^ 8) G. proximum {n = 10) G. longipes {n = 5) aff. badium (n = 8) G. sp. 1 aff. badium {n = 8) G. sp. 2 aff. badium G. longipes G. proximum G. spelaeum +(0.0052) +(0.0072) -(0.149) +(0.00004) + (0.0015) ±(0.011) -(30.80) + (0.0041) -(4.37) + (0.0003) -(99.45) + (0.0002) -(0.069) ±(0.018) + (0.0027) 132 THE JOURNAL OF ARACHNOLOGY Table 5. — Results of the Mols’s test for each pair of species studied, for leg IV length measurements taken in adult males of Goniosoma. “T” = “securely heterogeneous” (a < 0.01), “±” = “probably heterogeneous” (0.01 < a < 0.05), “ — ” = “possibly homogeneous” (a > 0.05). The value of a is given between parentheses following the respective symbol. G. sp. 2 G. varium {n = 4) G. spelaeum in = 8) G. proximum (n = 10) G. longipes (n = 5) aff. badium in = 8) G. sp. 1 aff. badium {n — 8) G. sp. 2 aff. badium G. longipes G. proximum G. spelaeum + (0.0003) +(0.0002) -(0.716) + (0.00002) + (0.0007) ±(0.038) -(1.06) + (0.0038) ±(0.014) + (0.0008) -(3.35) + (0.0002) ±(0.023) + (0.0025) + (0.0001) also identified the two species, treated herein as G. sp. 1 and 2 aff. badium, as belonging to G. badium. However, the morphometric data available showed that these species can be separated. Moreover, they differ in color, both in nature and preserved specimens. Goniosoma spelaeum and G. badium {sensu Stefanini-Jim) are yellowish-brown (howev- er, the latter has very contrasting dark brown legs), whereas the others are darker - G. sp. 2 aff. badium being dark brown, and G. sp. 1 aff. badium being dark greenish-brown. Goniosoma sp. 2 aff. badium has the pleura of articulations between coxae and trocanters pink-colored, whereas the others are white. In addition, this species and G. badium {sen- su Stefanini-Jim) have a series of internal spines on femur IV from base to the medial portion, whereas the others have from one to three medial internal spines on femur IV. In conclusion, this paper is intended to show that morphometric al variation is im- portant when describing harvestmen and can provide useful information and avoid mis- identification of species. Therefore, although it would take a long time to be done, I sug- gest that systematists include ranges of var- iation in their descriptions of new species. These would give a holistic understanding of the species being described, and a larger number of characters to be used in species recognition. ACKNOWLEDGMENTS This study was supported by grant #91/ 2818-0 from FAPESP (Fundagao de Amparo a Pesquisa do Estado de Sao Paulo). This re- port is part of my Ph.D. thesis, which was supervised by Dr. S.A. Vanin (IB-USP), to whom I am deeply indebted. The author has a research fellowship from CNPq (Conselho Nacional de Desenvolvimento Cientffico e Tecnologico). Fundagao Florestal do Estado de Sao Paulo allowed and supported field trips to Parque Estadual Intervales, where a large part of this study was conducted. Dr. B.D. Opell (Virginia Tech, USA) and the referees are thanked for their useful comments on the manuscript. Table 6.— Results of the Mols’s test for each pair of species studied, for palp length measurements taken in adult males of Goniosoma. “ + ” = “securely heterogeneous” (a < 0.01), “±” = “probably heterogeneous” (0.01 < a < 0.05), “ — ” = “possibly homogeneous” (a > 0.05). The value of a is given between parentheses following the respective symbol. G. sp. 2 G. varium {n = 4) G. spelaeum in = 8) G. proximum G. longipes aff. badium {n = 10) {n = 5) {n = 8) G. sp. 1 aff. badium {n = 8) -(0.144) -(0.078) -(1.56) -(1.30) -(2.07) G. sp. 2 aff. badium -(0.076) -(2.12) + (0.0007) -(1.05) G. longipes -(0.115) ±(0.042) -(5.56) G. proximum -(0.938) + (0.000004) G. spelaeum + (0.0050) GNASPINI— MORPHOMETRIC CHARACTERIZATION OF HARVESTMEN 133 Figure 1. — Mean ± standard deviation (rectangles) and amplitude of variation (lines) for the morpho- metric measurements taken in adult males of Goniosoma spelaeum (Gsp, n = 8), G. sp. 1 aff. badium (Gbl, n = 8), G. sp. 2 aff. badium (Gb2, n = 5), G. proximum (Gpr, n = 10), G. longipes (Gig, n = 5), and G. varium (Gvr, n = 4). Values are in millimeters. 134 THE JOURNAL OF ARACHNOLOGY LITERATURE CITED Gnaspini, R 1993. Biologia de opilioes cavemL colas da Provmcia Espeleologica do Vale do Ri- beira, SP/PR (Arachnida: Opiliones). Universi- dade de Sao Paulo, Instituto de Biociencias, Sao Paulo, PhD Thesis. 101 + v pp. Gnaspini, P. 1995. Reproduction and postembry- onic development of Goniosoma spelaeum, a cavernicolous harvestman from southeastern Brazil (Arachnida: Opiliones: Gonyleptidae). In- vert. Reprod. Develop., 28(2): 137-151. Gnaspini, P. 1996. Population ecology of Gonio- soma spelaeum, a cavernicolous harvestman from south-eastern Brazil (Arachnida: Opiliones: Gonyleptidae). J. Zool., 239(3):417-435. McGhee, C.R. 1977. Observations on the use of measurements in the systematic study of Leiob- unum (Arachnida: Phalangida). J. ArachnoL, 5: 169-178. Neet, C. 1993. Contribution a la methode biomet- rique de determination du nombre de mues au cours du developpement des araignees: separa- tion statistique des stades par le test de Mols. Rev. ArachnoL, 10(l):9-20. Stefanini-Jim, R.L. 1985. Gonyleptidae da subfa- mflia Goniosomatinae Mello-Leitao, 1935 (Opi- liones). Universidade Estadual Paulista “Julio de Mesquita Filho”, Instituto Basico de Biologia Medica e Agricola, Botucatu, MS Dissertation. 200 pp. Stefanini-Jim, R.L. 1995. Estudo sistematico de Goniosomatinae (Opiliones, Laniatores, Gony- leptidae). Universidade Estadual Paulista “Julio de Mesquita Filho”, Instituto Basico de Biologia Medica e Agricola, Botucatu, PhD Thesis. 84 pp. Manuscript received 29 April 1998, revised 16 Jan- uary 1999. 1999. The Journal of Arachnology 27:135-141 SEXUAL SELECTION IN PHOLCID SPIDERS (ARANEAE, PHOLCIDAE): ARTFUL CHELICERAE AND FORCEFUL GENITALIA Bernhard A. Huber: Department of Entomology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024 USA ABSTRACT. Two aspects of pholcid reproductive biology are reviewed and appear best explained by sexual selection by female choice: the rapid and divergent evolution of male chelicerae (and clypei in some groups) which contact the female epigynum during copulation and probably act as copulatory court- ship devices; and the often exceptionally strong pedipalps in males, which possibly function in correlation with the ‘valve’ in the internal female genitalia. The last decades have seen a promising in- crease of studies examining spider reproduc- tion from an evolutionary perspective (review: Elgar 1998). In most cases, the mechanisms of sexual selection in spiders are much the same as those documented in insects and other major groups (see e.g., Eberhard 1996, where almost every spider example used to docu- ment a specific mechanism of cryptic female choice is accompanied by at least one insect or mammal example). Some details, however, make spiders either especially useful (e.g., the pairedness of genitalia for studies of fluctu- ating asymmetry - Huber 1996b), or especial- ly interesting (e.g., the apparent lack of both muscles and nerves in the male intromittent genitalia - Eberhard & Huber 1998). (For fur- ther, though less unique, spider characteristics, see Elgar 1998.) In the present paper I will briefly review some recent advances in one particular spider family, the pholcids. Pholcids are the only non-entelegyne spiders whose reproductive biology has been carefully studied in several species (Eberhard 1992; Eberhard & Briceno 1983, 1985; Huber 1994, 1995, 1996a, b, 1997a, b, 1998a, b, c; Huber & Eberhard 1997; Raster & Jakob 1997; Uhl 1993, 1994; Uhl et al. 1995; Yoward 1998). Further ad- vantages for the study of sexual selection are the number of synanthropic species that are available worldwide and readily maintained in the laboratory for in depth single-species stud- ies, and a rich and diverse (mainly tropical) fauna for comparative studies. For reasons of space, I will focus on two particular aspects: on non-genitalic contact structures which appear to evolve under se- lection similar to that acting on genitalia, and on the unusual phenomenon of copulatory courtship associated with vigor. Voucher specimens of all unnamed species are deposited at the American Museum of Natural History, New York, and labeled with an I.D. number (“B.A.H. 1999 I.D.# 1-6”)- ARTFUL CHELICERAE Pholcids are not unique in having species- specific copulatory contact structures (Eber- hard 1985). However, pholcids are unique, at least among spiders, with respect to the wide range of non-intromittent male structures that are sexually modified (practically the entire palp is sexually dimorphic in most pholcids, including coxa and trochanter). Two male structures deserve special attention: the che- licerae and the clypeus. At least one of them contacts the female during copulation in all species studied (Huber 1994, 1995, 1997b, 1998b; Huber & Eberhard 1997; Uhl et al. 1995), and the chelicerae in particular are of- ten the most species- specific and taxonomi- cally useful structures. Modifications range from hairs of different shapes to cones, round- ed, pointed, hooked and blade-shaped apoph- yses, and even to sexually dimorphic fangs (Figs. 1-4, 10, 11, 13, 14). Several hypotheses might explain this phenomenon: (1) reproduc- tive isolation hypotheses (lock-and-key and genitalic recognition - reviewed in Eberhard 1985); (2) the “conflict of interest hypothe- sis” (Alexander et al. 1997); (3) sexual selec- 135 136 THE JOURNAL OF ARACHNOLOGY Figures 1-7. — Sexually dimorphic structures in male pholcids, SEM. 1. Cheliceral apophyses in Uthina sp. (I.D. #1); 2. Modified hairs on the chelicerae of Modisimus dominical Huber; 3. Modified hairs on the chelicerae of Spermophora senoculata (Duges); 4. Sclerotized cones on the chelicerae of Physocyclus guanacaste Huber; 5-6. Eye turret of Modisimus culicinus (Simon), in lateral and frontal view, showing frontal lobe; 7. Femur of Modisimus tortuguero Huber, showing a spine, a “normal” tactile sensillum, and several almost perpendicular hairs that cover the femora of only male walking legs. Scale bars: 0.01 mm (1-3); 0.05 mm (4, 7); 0.1 mm (5, 6). HUBER— SEXUAL SELECTION IN PHOLCID SPIDERS 137 tion by male-male competition (Eberhard & Bricefio 1985); (4) the “sperm holder hypoth- esis” (Brignoli 1973); (5) sexual selection by female choice (Eberhard 1985, 1996). Reproductive isolation hypotheses assume that species-specific differences in pholcid chelicerae evolved because they prevent hy- bridization. The lock-and-key version does this on a mechanical level, the genitalic rec- ognition hypothesis on a sensory level. Both seem unlikely to account for the phenomenon in a general way. Often there is no female “lock”, for instance in most Modisimus spe- cies where the epigynum against which the male chelicerae are pressed during copulation is just a flat plate, but the chelicerae are nev- ertheless species-specific (Huber 1998a). Even in cases with a lock-and-key like fit, the hy- pothesis that such fit evolved to avoid cross- specific pairings is dubious because natural selection should favor early species recogni- tion (Eberhard 1985, and references therein). However, transitory selection on cheliceral morphology in a species-isolation context can- not be ruled out, and may have been important in the past (Shapiro & Porter 1989). The “conflict of interest hypothesis” (Al- exander et al. 1997), applied to pholcid che- licerae, would explain their often complex and species-specific design by physical coercive mating during which the male has to over- come female resistance and for this purpose uses his cheliceral modifications. One predic- tion is that one sex (usually the male) changes to increase the match and the other sex evolves to either decrease it or it does not evolve at ail in that context (Alexander et al. 1997, p. 5). The data available for pholcids do not support this scenario. To the contrary, fe- male epigyneal structures usually appear ei- ther neutral (flat plates) or even cooperative (hoods, grooves, pits, scapes) in that they help the male to lodge his chelicerae and thus to position his body correctly. It should be em- phasized that conflict per se is of course not a distinguishing characteristic between the “conflict of interest” and the “female choice” hypotheses. Conflict is a necessary result of female choosiness, and some morphological details reflect this conflict particularly clearly: in a Peruvian pholcid (LD. #2) the male pre- sumably (judging by male and female mor- phology) lodges his cheliceral apophyses into a hood in the female epigynum. However, the epigynum also carries a large apophysis on each side of the hood (Figs. 8, 9), so that only males with exceedingly long cheliceral apoph- yses can reach the hood (Figs. 10, 11). Thus, “genitalic arms races” of this sort may reflect selective female cooperation (the female pro- vides a hood for those males able to overcome her obstructive apophyses) rather than female resistance to coercive males. Revealingly, the females of several putative close relatives have the cooperative structure (the hood) but not the barrier (the apophyses) (Fig. 12). Ac- cordingly, their males’ cheliceral apophyses are rather inconspicuous (Fig. 13), but nev- ertheless species-specific in form. Male-male competition is equally unlikely to provide a general explanation (Huber 1994). In most species the modifications seem highly inefficient for combat, and in some cas- es where their shape might suggest such a function (as in the Ecuadorian pholcid illus- trated in Fig. 14; I.D. #3) the female mor- phology strongly indicates their use during copulation (the epigynum is unusually broad and at the lateral extremes provides two blind- ended cuticular tubes, whose location and spacing indicates that they are used to accom- modate the male fang-apophyses: Figs. 15, 16). However, inter- and intrasexual functions are not mutually exclusive, and fighting with chelicerae must be considered a possibility in single cases. Brignoli’s (1973) speculation that pholcid chelicerae may function to hold the sperm during sperm uptake is probably based on the observations of Gerhardt (1921-1933, refer- ences in Huber 1998d) that pholcid males make no sperm webs but transfer the drop of sperm to the chelicerae and take it up from there. This hypothesis obviously fails to ex- plain why males should evolve such a variety of modifications to perform the same simple task of holding a drop of sperm. Thus, by elimination, the hypothesis that seems to fit the available data best is cryptic female choice (as also for many other non- genitalic male contact structures - Eberhard 1985). Much like genitalia, chelicerae may function as copulatory courtship devices, whose elaborate morphology is used to stim- ulate or fit the female in a way that increases the male’s chances of fathering her offspring. In this hypothesis two main factors account for the diversity and relatively rapid evolution 138 THE JOURNAL OF ARACHNOLOGY Figures 8-19. — Characters discussed in the present paper. 8-9. Epigynum of a Peruvian pholcid (I.D. #2) in ventral and lateral (anterior side on left) view, showing the median hood and the lateral apophyses; 10-11. Male chelicerae of the same species, lateral and frontal view; 12-13. Epigynum, ventral view, and male chelicerae, frontal view, of 'Blechroscelis' cyaneotaeniata (Keys.); 14. Portrait of an Ecuadorian species (I.D. #3), with modified fangs; 15-16. Epigynum of the same species, in ventral and lateral (anterior side on left) view, showing the blind-ended tubes into which the male apophyses are presumably inserted during copulation; 17-18. Clypeal modifications in two Metagonia species from Peru (17; I.D. #4) and Brazil (18; I.D. #5); 19. Right pedipalp of a Bolivian species (I.D. #6), in which the patella is reduced. Drawn to different scales. HUBER— SEXUAL SELECTION IN PHOLCID SPIDERS 139 of male chelicerae: the unpredictability of fe- male criteria and the never ending competition among males for access to female eggs. How females evaluate the minimal differences among conspecific males’ chelicerae, i.e,, the sensory and neuroanatomical basis for doing so, remains an open question. Sexual modifications of the male clypeus are less common in pholcids, but have appar- ently evolved several times convergently (e.g., in Metagonia: Figs. 17, 18, Holocneminus - Huber 1997b, Deeleman-Reinhold 1994). Like the chelicerae, the clypeal modifications are highly species-specific and in one species {Metagonia rica Gertsch 1986) it has been shown that they also contact the female gen- ital area during copulation (Huber 1997b). A special case of non-genitalic contact structure is the frontal lobe in male Modisimus culicinus (Simon 1893) (Figs. 5, 6). Clypeal glands open at the lobe, and during copulation the female mouth is in contact with the lobe, suggesting gustatorial courtship (Huber 1997a). However, the nature and function of the gland products are unknown (trigger fe- male responses that are favorable to male? - signal the female that copulation has oc- curred? - nourish the female?) meaning that a decision between natural and sexual selection is not yet possible (see Eberhard 1996 for ar- guments linking sexual selection and male seminal products). FORCEFUL GENITALIA It has been noted that “details of copulatory courtship often seem to have little relationship to male size or vigor” (Eberhard 1997: 35). If this is the rule, then many pholcids might be exceptional: their genitalia are obviously their strongest organs (provided with the largest muscles), and in Physocyclus globosus (Tacz. 1873) this force is apparently used to rhyth- mically squeeze parts of the female genitalia during copulation (Huber & Eberhard 1997). Moreover, a morphometric study of genitalic and non-genitalic structures in the same spe- cies also apparently supported the notion that there is sexual selection on male vigor: fluc- tuating asymmetry (FA: deviations from per- fect bilateral symmetry that are thought to re- flect the degree of developmental stability) in large (strong) genitalia tended to be lower than in small genitalia (Huber 1996b). In the recent literature on FA such a negative re- gression of FA on size is often interpreted as evidence for handicap models of sexual selec- tion, in which only genetically “good” males can produce display structures that are both large and symmetric (Mpller & Pomiankowski 1993; Watson & Thornhill 1994). From a mechanical point of view, the phol- cid male pedipalp works like a clamp, with the most distal segment (cymbium with pro- cursus) acting against the femur. The economy of such a clamp is decreased by the two seg- ments in-between (patella, tibia) and could be improved by elimination of one or both seg- ments. In fact, in many pholcids (e.g., in P. globosus), the patella is functionally reduced in that part of the muscles of the femur insert in the tibia (Huber & Eberhard 1997) and not as usually in the patella (Ruhl & Rathmayer 1978). And in at least one species (sp. n. from Bolivia; I.D. #6) the reduction is complete, with the femur directly articulating with the tibia and no external trace of the patella left (Fig. 19). Yet another characteristic apparently functioning to increase the force applied to one critical point is realized in P. globosus (and probably in all Physocyclus species and in Artema atlanta Walckenaer 1837): the pro- cursi are locked to each other, but the tip of only one is inserted into the female, moved by the muscular power of both pedipalps (Huber & Eberhard 1997). Thus, there seems to be an ultimate advan- tage for males with strong palps, but the prox- imate function of this vigor is poorly under- stood. A possible solution may be in a structure of the female genitalia that is appar- ently unique to pholcids: the so-called “valve”, an often complex “three dimension- al zipper” between copulatory pouch (uterus extemus) and oviduct (uterus internus). An apparent correlation has been documented be- tween the strength and complexity of the “valve” and the strength of the male pedipalp (Huber 1998c). The correlation may be phy- logenetically biased, however, so it is difficult to interpret. It is not surprising that the recently inten- sified research on pholcids has raised more questions than it has answered. Thus, I would like to close this short review with yet another riddle. The males (but not females) of most species of several mainly Central American genera have the femora of their walking legs covered with short, almost perpendicular hairs 140 THE JOURNAL OF ARACHNOLOGY (Huber 1998a), resembling taste hairs (Foelix & Chu-Wang 1973) (Fig. 7). Nothing is known of these hairs, apart from the approx- imate systematic and geographical distribution of the character, the improbability of taste hairs being concentrated on femora, and the apparent lack of terminal pores necessary for chemosensory function. ACKNOWLEDGMENTS I thank James Berry, William Eberhard, Gustavo Hormiga and Brent Opell for valu- able comments on a previous draft. This study was supported by a Theodore Roosevelt Fel- lowship from the American Museum of Nat- ural History, New York. LITERATURE CITED Alexander, R.D., D.C. Marshall & J.R. Cooley. 1997. Evolutionary perspectives on insect mat- ing. Pp. 4-31. In The Evolution of Mating Sys- tems in Insects and Arachnids (J.C. Choe & B.J. Crespi, eds.). Cambridge Univ. Press, U.K. Brignoli, PM. 1973. Notes on spiders, mainly cave-dwelling of Southern Mexico and Guate- mala (Araneae). Accad. Naz. Lincei, 171:195- 238. Deeleman-Reinhold, C.L. 1994. Redescription of Holocneminus multiguttatus Simon and descrip- tion of two new species of pholcid spiders from Australia (Arachnida: Araneae: Pholcidae). Beitr. AraneoL, 4:31-42. Eberhard, W.G. 1985. Sexual Selection and Animal Genitalia. Harvard Univ. Press, Cambridge, Mas- sachusetts. Eberhard, W.G. 1992. Notes on the ecology and behaviour of Physocyclus globosus (Araneae, Pholcidae). Bull. British Arachnol. Soc., 9:38- 42. Eberhard, W.G. 1996. Female Control: Sexual Se- lection by Cryptic Female Choice. Princeton Univ. Press, New Jersey. Eberhard, W.G. 1997. Sexual selection by cryptic female choice in insects and arachnids. Pp. 32- 51. In The Evolution of Mating Systems in In- sects and Arachnids (J.C. Choe & B.J. Crespi, eds.). Cambridge Univ. Press, U.K. Eberhard, W.G. & D. Briceno L. 1983. Chivalry in pholcid spiders. Behav. Ecol. SociobioL, 13: 189-195. Eberhard, W.G. & D. Briceno L. 1985. Behavior and ecology of four species of Modisimus and Blechroscelis (Araneae, Pholcidae). Rev. Arach- nol., 6:29-36. Eberhard, W.G. & B.A. Huber. 1998. Possible links between embryology, lack of innervation, and the evolution of male genitalia in spiders. Bull. British Arachnol. Soc., ll(2):73-80. Elgar, M.A. 1998. Sperm competition and sexual selection in spiders and other arachnids. Pp. 307- 339. In Sperm Competition and Sexual Selection (TR. Birkhead & A.R Mpller, eds.). Academic Press. Foelix, R.E & Chu-Wang I-Wu. 1973. The mor- phology of spider sensilla. II. Chemoreceptors. Tissue & Cell, 5(3):461-478. Huber, B.A. 1994. Genital morphology, copulatory mechanism and reproductive behaviour in Psil- ochorus simoni (Borland, 1911) (Pholcidae; Ar- aneae). Netherlands J. ZooL, 44(l-2):85-99. Huber, B.A. 1995. Copulatory mechanism in Hol- ocnemus pluchei and Pholcus opilionoides, with notes on male cheliceral apophyses and stridu- latory organs in Pholcidae (Araneae). Acta Zool., Stockholm, 76:291-300. Huber, B.A. 1996a. On the distinction between Modisimus and Hedypsilus (Araneae, Pholcidae), with notes on behavior and natural history. Zool. Scripta, 25:233-240. Huber, B.A. 1996b. Genitalia, fluctuating asym- metry, and patterns of sexual selection in Phy- socyclus globosus (Araneae: Pholcidae). Rev. Suisse Zool., suppl. 1996:289-294. Huber, B.A. 1997a. Evidence for gustatorial court- ship in a haplogyne spider (Hedypsilus culicinus: Pholcidae: Araneae). Netherlands J. Zool., 47: 95-98. Huber, B.A. 1997b. On American ‘'‘’Micromerys'" and Metagonia (Araneae, Pholcidae), with notes on natural history and genital mechanics. Zool. Scripta, 25:341-363. Huber, B.A. 1998a. Notes on the neotropical spider genus Modisimus (Pholcidae, Araneae), with de- scriptions of thirteen new species from Costa Rica and neighboring countries. J. Arachnol. 26(1): 19-60. Huber, B.A. 1998b. Genital mechanics in some neotropical pholcid spiders (Araneae; Pholcidae), with implications for systematics. J. Zool., Lon- don, 244:587-599. Huber, B.A. 1998c. On the “valve” in the genitalia of female pholcids (Pholcidae, Araneae). Bull. British Arachnol. Soc., ll(2):41-48. Huber, B.A. 1998d Spider reproductive behaviour: a review of Gerhardt’s work from 1911-1933, with implications for sexual selection. Bull. Brit- ish Arachnol. Soc., 11(3):81-9L Huber, B.A. & W.G. Eberhard. 1997. Courtship, copulation, and genital mechanics in Physocyclus globosus (Araneae, Pholcidae). Canadian J. Zool., 74:905-918. Raster, J.L. & E.M. Jakob. 1997. Last-male sperm priority in a haplogyne spider (Araneae: Pholci- dae): correlations between female morphology and patterns of sperm storage. Ann. Entomol. Soc. America, 90:254-259. M0ller, A.P & A. Pomiankowski. 1993. Fluctuat- HUBER— SEXUAL SELECTION IN PHOLCID SPIDERS 141 ing asymmetry and sexual selection. Genetica, 89:267—279. Ruhl, M. & W. Rathmayer. 1978. Die Beinmusku- latur und ihre Innervation bei der Vogelspinne Dugesiella hentzi (Ch.) (Araneae, Aviculariidae). Zoomorphologie, 89:33-46. Shapiro, A.M. & A.H. Porter. 1989. The lock-and- key hypothesis: evolutionary and biosystematic interpretation of insect genitalia. Ann. Rev. En- tomoL, 34:231-245. Uhl, G. 1993. Mating behaviour and female sperm storage in Pholcus phalangioides (Fuesslin) (Ar- aneae). Mem. Queensland Mus., 33(2);667-674. Uhl, G. 1994. Genital morphology and sperm stor- age in Pholcus phalangioides (Fuesslin, 1775) (Pholcidae; Araneae). Acta ZooL, Stockholm, 75:1-12. Uhl, G., B.A. Huber & W. Rose. 1995. Male ped- ipalp morphology, and copulatory mechanism in Pholcus phalangioides (Fuesslin, 1775). Bull. British Arachnol. Soc., 10:1-9. Watson, RJ. & Thornhill, R. 1994. Fluctuating asymmetry and sexual selection. TREE, 9:21-25. Yoward, P. 1998. Sperm competition in Pholcus phalangoides (Fuesslin, 1775) (Araneae, Pholci- dae) - Shorter second copulations gain a higher paternity reward than first copulations. Pp. 167- 170. In Proc. Arachnol. Congress, Edinburg. (P.A. Selden, ed.). Manuscript received 29 April 1998, revised 6 Sep- tember 1998. 1999. The Journal of Arachnology 27:142-148 A COMPARISON OF THE RESPIRATORY SYSTEMS IN SOME CAVE AND SURFACE SPECIES OF SPIDERS (ARANEAE, DYSDERIDAE) Matjaz Kuntner^ \ Boris Sket^ and Andrej Blejec-: ‘Department of Biology, Biotechnical Faculty, University of Ljubljana, SI-1111 Ljubljana, Slovenia; -National Institute of Biology, FOB 141, SLIOOI Ljubljana, Slovenia ABSTRACT. We tested the hypothesis that the respiratory system of hypogean spiders is subject to regressive evolution by examining representatives of the family Dysderidae. This comparison included the epigean species Dysdera ninnii Canestrini 1868, and Harpactea lepida (C.L. Koch 1838), and the hypo- gean species Stalita taenaria Schiodte 1847, and Parastalita stygia (Joseph 1882). Both tube tracheae and book lungs of these species were measured and compared using 10 indices. Both the tracheal system and book lungs of the hypogean species were less developed than those of the epigean ones. We suggest that the cause is reduction of the respiratory system as a part of general structural reductions in the troglobites. This is consistent with the lower respiratory rates that characterize many troglobites. True cave animals, called “troglobites,” are adapted to living in a very different environ- ment from their surface relatives. Ecological conditions of the terrestrial underground en- vironment are summarized here according to Vandel (1965) and Sket (1996). The most ob- vious and constant factor is the absence of light, consequently the absence of green plants and near absence of primary production, re- sulting in an energy-poor hypogean (subter- ranean) environment. The nearly constant temperature roughly equals the yearly average of the region. The chemical composition of the air in the caves that are well ventilated is similar to the surface atmosphere with a very slight increase of CO, concentration. Howev- er, in some caves CO2 concentration may be significantly increased and O2 concentration may be low (Vandel 1965; James et al. 1975; Whitten et al. 1987). Relative humidity in caves is normally 95-100%, so troglobites are hygrophilic and more sensitive to drying than epigean species (Vandel 1965). Morphological adaptations of animals to the underground environment, or troglomorph- isms, can be seen as gains or reductions. Typ- ical hypogean arthropod gains are larger bod- ies and longer appendages, and increases in the number of “nonvisual” sensory organs. Current address: Department of Biological Scienc- es, George Washington University, 2023 G Street, N.W, Washington, DC 20052, USA Typically, cuticle features such as wings, pig- mentation and eyes are reduced (Vandel 1965; Sket 1985). In European cave spiders, all pos- sible stages of depigmentation, eye reduction, and weakening of the integument can be ob- served (Deeleman-Reinhold 1975) along with gains such as elongation of appendages. Regressive evolution or degenerative evo- lution in cave organisms, reviewed by Fong & Culver (1985), Kane & Richardson (1985), Poulson (1985), Romero (1985) and Sket (1985) is not restricted to morphological re- gression but can also be met with physiolog- ical and ethological changes. Typical changes in hypogean animals include decreased met- abolic rate, slower ontogenetic development, and pedomorphosis. The respiratory metabolic rates of the studied cave species were as low as 3% of that of related surface species in iso- pod crustaceans and as low as 14% in trog- lophilic spiders (Vandel 1965; Hiippop 1985). The possible reasons of this reduction include relative ecological stability of the under- ground environment, lack of predators, low food availability (Hiippop 1985) and possibly higher CO2 concentration (Whitten et al. 1987). A number of authors present data about spi- der metabolism and respiratory physiology (e.g., Anderson 1970; Anderson & Prestwich 1980, 1982, 1985; Bromhall 1987; Dresco- Derouet 1969; Greenstone & Bennett 1980; 142 KUNTNER ET AL.— RESPIRATORY SYSTEMS IN SPIDERS 143 Opell 1987, 1990, 1992, 1998; Paul et al 1987, 1989; Paul & Fincke 1989; Paul 1992; Prestwich 1983a, 1983b; Strazny & Perry 1984, 1987). Energetic adaptations allow spi- ders to have roughly half the value of the met- abolic rate present in other poikilothermic an- imals of the same weight. During starvation periods, which in spiders can be prolonged, they have the ability to lower metabolic rates below the resting values. Spiders that have the lowest metabolic rates are adapted to living in the energy-poorest environments. The mea- sured metabolic rates show positive correla- tion to the respiratory surface and volume. Al- though the cited studies did not include any troglobitic species, we can still assume that in the energy-poor cave environment the low metabolism in spiders can have impact on the structure of their respiratory system. We studied the structure of the respiratory system in hypogean and epigean species of the spider family Dysderidae, which has many representatives in the Mediterranean, ranging from the most xerophylic epigean forms to the blind troglobites. Dysderoidea includes families Dysderidae, Segestriidae, Oonopidae and Orsolobidae. Forster & Platnick (1985) claim that for Dys- deroidea that their representatives have re- duced book lungs, in the extreme case only four respiratory lamellae in some oonopids, and well developed paired tube tracheae, opening immediately behind the book lungs on the ventrolateral part of the abdomen. Levi (1967) and Winkler (1955) claim the above statement to be particularly adequate for Dys- deridae. In Dysderidae the tracheae extend into pro- soma and enter the appendages (Winkler 1955; Novak 1967; Bromhall 1987; Foelix 1992). The dysderid heart is relatively small (Kaestner 1969). Spiders possessing prosomal tracheae have lower heart rates than spiders with tracheae limited to the abdomen (Brom- hall 1987). The conclusion is that in Dysder- idae the tracheae have a larger role in gas ex- change than the book lungs, the latter being a less functional remnant of the evolution of this family. We tested the hypothesis that reductions in the respiratory system occur during the course of evolution in phylogenetically old hypogean spider species. Thus, H, claims that the hy- pogean species have reduced respiratory sys- tems relative to the epigean species. In con- trast, H() states that there are no differences between hypogean and epigean species respi- ratory systems. The evidence supporting hy- pothesis H, is: (1) With few exceptions trog- lobites show lower metabolic rates than their epigean relatives; (2) High relative humidity in underground air coupled with thin integu- ment of troglobitic species might allow addi- tional gas exchange through their body sur- face, so their respiratory system need not to be strongly developed; (3) The majority of European karst caves are well aerated (Gams 1974: p. 123), and the differences between the cave and the surface atmospheres in such cas- es are small (James et al. 1975). A significant increase of CO2 concentration and corre- sponding O2 concentration decrease in karst caves is a rather unusual phenomenon. METHODS Species studied. — For the study of the re- spiratory system four different dysderid spe- cies belonging to three subfamilies (Deele- man-Reinhold & Deeleman 1988) were chosen (number and sex of the studied spec- imens in parentheses): the epigean species Dysdera ninnii Canestrini 1868 of the Dys- derinae (69) and Harpactea lepida (C.L. Koch 1838) of the Harpacteinae (5 <3, 1$), and the hypogean species of Rhodinae Stalita taenaria Schiodte 1847 (3(3, 3 9), and Par- astalita stygia (Joseph 1882; 3(3, 29, 1 im- mature). All these species were collected in Slovenia. The exact locality and habitat data of the examined material is given elsewhere (Kuntner 1998). Preparation. — Spiders were dissected in 70% ethanol in a petri dish. Dorsal surfaces of the prosoma and of the abdomen were care- fully removed. The animals were then gently heated in 10% KOH for 1 hour. They were then placed in a vial filled with distilled water and the vial was rigorously shaken. All soft tissues, eroded by KOH were thus removed. The chitinous cuticle was then stained over- night in chlorazol black mixed with glycerol, all integumental structures (including both components of the respiratory system) being colored black. Later the preparation was ex- amined in water or further dissected and mea- sured in glycerol. Parameters measured and indices calcu- lated.— Abbreviations of measured parame- 144 THE JOURNAL OF ARACHNOLOGY Table 1.^ — Parameters measured. Units in millimeters or mm^ (*) except in NBL. Abbre- via- tion Parameter Description PL Prosoma length dorsal view OL Opisthosoma length dorsal view F4 Femur IV length prolateral view BSW Book lung stigma width ventrolateral view TSW Tracheal stigma width ventrolateral view TLl Cranial tracheal trunk length dorsal view of the outer length of the curved main tracheal trunk TC Cranial tracheal curvature width dorsal view TL Actual cranial tracheal trunk length calculated from TLl and TC TW Cranial tracheal trunk width width in the middle of the trunk TV Cranial tracheal trunk volume* calculated from TL and TW CTL Caudal tracheal tube length dorsal view TP Cranial tracheal profile circumference of the terminal part of the cranial tracheal trunk was calculated from the measurement using drawing and curvimeter, taking into account only the profile from which the tracheolae originate CTP Caudal tracheal profile circumference of the terminal part of the caudal tracheal trunk using the same method as above NBL Number of book lung lamellae examined under light microscope laterally ters, their names and short descriptions are given in Table 1. They were measured for each specimen, using dissecting and com- pound light microscope with a micrometer. For parameters TP and CTP we used scale drawings and curvimeter to calculate circum- ferences of tracheal parts. Parameters TL and TV were calculated using formulae from Bronstejn & Semendjajev (1984). Ten indices were devised (Table 2) to compare the size and development of the respiratory systems between species and to reduce the influence of body size on studied parameters. Features tend to vary with size, surface and volume of the spider (Harpactea lepida was smaller in size than the rest of the species). This was taken into account in construction of indices as we tried to reduce variability of measured parameters within the pairs of surface and cave species. Statistical analyses. — For analysis of dif- ferences among the surface and cave species, several statistical tests were applied: Wilcoxon rank sum (Mann- Whitney U-test), Kolmogo- rov-Smirnov, and Student t-test. Since differ- ent methods gave essentially the same results, only Wilcoxon rank sum is reported. There were generally no significant differences Table 2. — Indices for comparison of the respiratory systems in species examined. Parameter abbrevia- tions given in Table 1. Structure-index composition: L = length, S = surface, V = volume, N = number. Quotient Description Structure 11 TL/PL3 Relative cranial tracheal trunk length LW 12 TW/PL Relative cranial tracheal trunk width L/L 13 TV/PL' Relative cranial tracheal trunk volume V/(S.V) 14 CTL/PL3 Relative caudal tracheal tube length LW 15 TP/PL2 Relative extent of tracheolae branching in the prosoma L/S 16 CTP/OL2 Relative extent of tracheolae branching in the opisthosoma L/S 17 TP/F42 Relative extent of tracheolae branching in the appendages L/S 18 TSW/PL2 Relative tracheal stigma width L/S 19 BSW/PL2 Relative book lung stigma width L/S 110 NBL/PL Relative number of book lung lamellae N/L KUNTNER ET AL.— RESPIRATORY SYSTEMS IN SPIDERS 145 Table 3. — Wilcoxon rank sum test of differences between species and species groups: — Not signif- icant, * P < 0.05, ** P < 0.01, *** P < 0.001. Comparison Index Stalita: Para- stalita Dysdera: Harpactea Cave: Surface TL/PL3 — — — TW/PL — — TV/PL5 — — CTL/PL3 * — TP/PL2 — — CTP/OL2 — — — TP/F42 ** — TSW/PL2 — — BSW/PL2 — — * NBL/PL — — among the cave or surface species (Table 3), so combined samples were eventually used to test the differences between cave and surface groups of species. RESULTS Both hypogean species lack eyes and have longer legs than both epigean species. In ad- dition they are depigmented, have more setae on their legs, and appreciably more delicate cuticle. Dysdera ninnii has a well-developed tracheal system, but both cave species have less extensive tracheae. Although Harpactea lepida has relatively stout cranial tracheal trunks, these are relatively shorter than those of D. ninnii. However, no significant differ- ence in all the indices between both surface species was observed (Table 3). Similarly, there was no significant difference in most of the indices between both cave species, the in- dices 14 and 17 being an exception (Table 3). Figure 1 shows the values of indices II to 110 for the studied species in the same order as listed in Table 2 (in the graphs the names of genera are given). In Table 3, Wilcoxon rank sum test of differences between both species pairs are presented. Relative cranial tracheal trunk length (II) was largest in D. ninnii, followed by H. lepida and both cave species. However, there was no significant difference between the epigean and hypogean pairs of species. Relative cranial tracheal trunk width (12) showed similar val- ues in both epigean species and was lower in both troglobites, both pairs of species showing significant difference. Relative cranial tracheal trunk volume (13) showed a similar result to the previous index. Relative caudal tracheal tube length (14) showed highly significant dif- ference between both species pairs, and so did the next index- — relative extent of tracheolae branching in the prosoma (15). Relative extent of tracheolae branching in the opisthosoma (16) was lower in the epigean pair of species, both species pairs showing no significant dif- ference. Relative extent of tracheolae branch- ing in the appendages (17) was highest in D. ninnii, followed by H. lepida, S. taenaria and was lowest in the longlegged (more troglo- morphic) P. stygia. There was a highly sig- nificant difference between the pairs of spe- cies. Relative tracheal stigma width (18) was similar within the cave and surface species groups and showed a significant difference be- tween them. The values for the relative book lung stigma width (19) were again significant- ly higher in the epigean pair of species. Rel- ative number of book lung lamellae (110) showed again significantly higher values for the surface versus cave species pairs. DISCUSSION Dysdera ninnii has the most extensive tra- cheal system of the examined species. Both cranial and caudal tracheal trunks are strongly developed, and they branch into numerous tracheolae that supply with oxygen the pro- somal organs, appendages and opisthosoma. Book lungs are also well developed, having up to 23 lamellae, but show considerable var- iability in their size (Kuntner 1998). The sec- ond epigean species, Harpactea lepida, shows similarly developed tracheal system and book lungs to D. ninnii, despite its smaller size. Al- though H. lepida and D. ninnii are both forest species, different ecological factors might in- fluence their anatomy and physiology. How- ever, as hypothesized, the examined surface species exhibit a well-developed respiratory system, even though they belong to different subfamilies. The troglobites, Stalita taenaria and Par- as tali ta stygia, both show reductions in the tracheal as well as book lung systems, com- pared to both surface species. Their cranial tracheal trunks are relatively shorter, narrower, and not as curved as in D. ninnii. They extend further into prosoma through the petiole be- 146 THE JOURNAL OF ARACHNOLOGY Figure 1. — Relative size and development of the respiratory system in epigean (Dysdera ninnii, Har- pactea lepida) and hypogean {Stalita taenaria, Parastalita stygia) dysderid spiders from Slovenia, mea- sured in six specimens each. For explanation of indices (II -1 10) see Table 2. (o = Individual data, % = Mean value ± SD, p = Wilcoxon rank sum significance level for differences between the surface and cave groups of species, groups indicated by the lines connecting mean values). KUNTNER ET AL.— RESPIRATORY SYSTEMS IN SPIDERS 147 fore branching into tracheolae. The tracheolae bundle is much weaker in prosoma and fewer were observed to enter the legs. Their caudal tracheal tubes are greatly reduced compared to the ones in surface species, but the extent of the tracheolae branching in the opisthoso- ma shows no difference. The book lungs of both troglobites are also reduced and have a slightly different general appearance from those in the epigean species. Both hypogean species showed very similar values of all the indices. Since they belong to a different subfamily than Dysdera and Har~ pactea, we cannot be sure that the supposedly epigean ancestors of hypogean Rhodinae had a stronger developed respiratory system, sim- ilar to that of Dysdera and Harpactea. How- ever, we speculate that in both cave species it has been subject to regressive evolution. There are no comparable data on cave spider respiratory morphology in available araneo- logical literature (e.g., Nentwig 1987). Yet, this seems to be another example of structural reduction in troglobites, similar to the reduc- tion of other originally integumental structures in troglobitic spiders (Deeleman-Reinhold 1975) and other arthropods (Vandel 1965; Sket 1985). The cause for the reduction of the respiratory system in cave spiders still needs to be investigated further. Opell (1990, 1998) states that tracheae and book lungs in the spider family Uloboridae are complementary respiratory structures; when one system is better developed the other is reduced. Opell concludes that the develop- ment of the two systems is governed by both spider’s total respiratory demands and by the specificity of these demands. The more active species (with reduced webs) have relatively better developed tracheae, and the less active (orb-weaving) species have relatively better developed book lungs. If this is true for Dys- deridae, future research could focus on pos- sible compensating changes in both systems. Are both systems reduced in the troglobites or is there a shift in relative development of each system? As our study primarily treated the dysderid tracheae, future studies may reveal that the book lungs in the hypogean environ- ment are more useful than in the epigean one. ACKNOWLEDGMENTS We thank Irena Sereg and Slavko Polak for providing a part of the studied material. Jon- athan Coddington, Frederick Coyle, Christa Deeleman-Reinhold, Kazimir Draslar, Gusta- vo Hormiga, Ivan Kos, Norman Platnick, Dennis Radabaugh, Paul Selden, and Jorg Wunderlich helped with advice or literature. France Velkovrh, Luka Malensek and Domen Komac supported us technically. Sonja Kunt- ner and Ian Baxter kindly corrected the En- glish manuscript. We thank Brent Opell, Mar- tin Ramirez, and two anonymous reviewers for numerous comments to our early draft of the manuscript. Finally, the first author thanks the American Arachnological Society and Pe- tra Sierwald for an award to attend the XIV International Congress of Arachnology in Chi- cago. This project was submitted by the first author as partial fulfillment for the degree of Biology from the University of Ljubljana. LITERATURE CITED Anderson, J.E 1970. Metabolic rates of spiders. Comp. Biochem. Physiol., Pergamon Press, 33: 51-72. Anderson, J.E & K.N. Prestwich. 1980. Scaling of subunit structures in book lungs of spiders (Ar- aneae). J. Morphol., 165:167-174. Anderson, J.E & K.N. Prestwich. 1982. Respira- tory gas exchange in spiders. Physiol. ZooL, 55(l):72-90. Anderson, J.E & K.N. Prestwich. 1985. The phys- iology of exercise at and above maximal aerobic capacity in a theraphosid (tarantula) spider, Bra- chypelma smithi (EO. Pickard-Cambridge). J. Comp. Physiol., B, 155:529-539. Bromhall, C. 1987. Spider heart-rates and loco- motion. J. Comp. Physiol., B, 157:451. Bronstejn, J.N. & K.A. Semendjajev. 1984. Mate- maticni prirocnik. Tehniska zalozba Sloveniie. 699 s. Deeleman-Reinhold, C.L. 1975. Distribution pat- terns in European cave spiders. Proc. Int. Symp. Cave Biol., Oudtshoom (South Africa), Pp. 25- 36. Deeleman-Reinhold, C.L. & PR. Deeleman. 1988. Revision des Dysderinae. Tijdschrift voor Ento- mologie, 131:141-269. Dresco-Derouet, L. 1969. Etude d’Araignees et d'Opilions cavernicoles dans leur milieu. I. In- tensite respiratoire, premiers resultats. Annales de Speleologie, 24(3):529-532. Eincke, T. & R. Paul. 1989. Book lung function of arachnids. III. The function and control of the spiracles. J. Comp. Physiol., B, 159:433-441. Foelix, R.E 1992. Biologic der Spinnen, 2. Auf., Thieme, Stuttgart. 331 pp. Fong, D.W. & D.C. Culver. 1985. A reconsidera- tion of Ludwig’s differential migration theory of 148 THE JOURNAL OF ARACHNOLOGY regressive evolution. NSS Bulletin, 47(2): 123- 127. Forster, R.R. & N.I. Platnick. 1985. A review of the Austral spider family Orsolobidae (Arachni- da, Araneae), with notes on the superfamily Dys- deroidea. Bull. American Mus. Nat. Hist., 181(1), 229 pp. Gams, L 1974. Kras =■ Zgodovinski, naravoslovni in geografski oris. Slovenska matica, Ljubljana. 358 pp. Greenstone, M.H. & A.F. Bennett. 1980. Foraging strategy and metabolic rate in spiders. Ecology, 61(5): 1255-1259. Hiippop, K. 1985. The role of metabolism in the evolution of cave animals. NSS Bulletin, 47(2): 136-146. James, J.M., A.J. Pavey & A.F. Rogers. 1975. Foul air and the resulting hazard to cavers. Trans. Brit- ish Cave Res. Assoc., 2(2):79-88. Kaestner, A. 1969. Lehrbuch der Speziellen Zool- ogie, Band 1: Wirbellose 1, 3. Auf., Gustav Fi- scher Verlag, Stuttgart. 898 pp. Kane, T.C. & R.C. Richardson. 1985. Regressive evolution: An historical perspective. NSS Bul- letin, 47(2):71-77. Kuntner, M. 1998. Primerjava zgradbe dihal pri jamskih in povrsinskih vrstah pajkov (Araneae: Dysderidae). Graduation Thesis, Univ. of Lju- bljana. 84 pp. Levi, H.W. 1967. Adaptations of respiratory sys- tems of spiders. Evolution, 21:571-583. Nentwig, W. (ed.). 1987. Ecophysiology of Spi- ders. Springer Verlag, Berlin. 448 pp. Novak, V. 1967. Stalita taenaria Schiodte s poseb- nim ozirom na dihala. Graduation Thesis, Univ. of Ljubljana, 20 pp. Opell, B.D. 1987. The influence of web monitoring tactics on the tracheal systems of spiders in the family Uloboridae. Zoomorphology, 107:255- 259. Opell, B.D. 1990. The relationship of book lung and tracheal systems in the spider family Ulo- boridae. J. MorphoL, 206:211-216. Opell, B.D. 1992. Influence of web-monitoring tactics on the density of mitochondria in leg mus- cles of the spider family Uloboridae. J. MorphoL, 213:341-347. Opell, B.D. 1998. The respiratory complementarity of spider book lung and tracheal systems. J. Mor- phoL, 236:57-64. Paul, R.J. 1992. Gas exchange, circulation, and en- ergy metabolism in arachnids. Pp. 169-197. In Physiological Adaptations in Invertebrates. Res- piration, Circulation, and Metabolism (S.C. Wood, R.E. Weber, A.R. Hargens & R.W. Mil- lard, eds.). Dekker, New York. Paul, R., T. Fincke & B. Linzen, 1987. Respiration in the tarantula Eurypelma califomicum: evi- dence for diffusion lungs. J. Comp. Physiol., B, 157:209. Paul, R., T. Fincke & B. Linzen. 1989. Book lung function in arachnids. I. Oxygen uptake and re- spiratory quotient during rest, activity and recov- ery— relations to gas transport in the haemo- lymph. J. Comp. Physiol., B, 159:409-418. Paul, R. & T. Fincke. 1989. Book lung function in arachnids. 11. Carbon dioxide release and its re- lations to respiratory surface, water loss and heart frequency. J. Comp. Physiol., B, 159:419- 432. Poulson, T.L. 1985. Evolutionary reduction by neutral mutations: Plausibility arguments and data from amblyopsid fishes and linyphiid spi- ders. NSS Bulletin, 47(2): 109-1 17. Prestwich, K.N. 1983a. Anaerobic metabolism in spiders. Physiol. ZooL, 56(1):112-12L Prestwich, K.N. 1983b. The roles of aerobic and anaerobic metabolism in active spiders. Physiol. ZooL, 56(1): 122-132. Romero, A. 1985. Can evolution regress? NSS Bulletin, 47(2):86-88. Sket, B. 1985. Why all cave animals do not look alike — a discussion on adaptive value of reduc- tion processes. NSS Bulletin, 47(2):78-85. Sket, B. 1996. The ecology of the anchihaline caves. Trends EcoL EvoL, 1 1(5):221-225. Strazny, E & S.E Perry. 1984. Morphometric dif- fusing capacity and functional anatomy of the book lungs in the spider Tegenaria spp. (Age- lenidae). J. Morph., 182:339-354. Strazny, E & S.F. Perry. 1987. Respiratory system: structure and function. Pp. 78-94. In Ecophysi- ology of Spiders (W. Nentwig, ed.), Springer Verlag, Berlin. Vandel, A. 1965. Biospeleology. Pergamon Press. 524 pp. Whitten, A.J., M. Mustafa & G.S. Henderson. 1987. The Ecology of Sulawesi. Gadjah Mada Univ, Press, Yogyakarta. 777 pp. Winkler, D, 1955. Das Tracheensystem der Dys- deriden. Mitt. ZooL Mus., Berlin, 31:25-43. Manuscript received 1 May 1998, revised 3 March 1999. 1999. The Journal of Arachnology 27:149-153 A NEW ALL-FEMALE SCORPION AND THE FIRST PROBABLE CASE OF ARRHENOTOKY IN SCORPIONS Wilson R, Lourengo: Laboratoire de Zoologie (Arthropodes), Museum National d’Histoire Naturelle, 61 rue de Buffon 75005 Paris, France Orlando Cuellar: P.O. Box 17074, Salt Lake City, Utah 84117-0074 USA ABSTRACT. A new parthenogenetic species of scorpion, Ananteris coineaui Louren^o, is reported from French Guyana. Parthenogenesis is based on the production of an all-female brood (thelytoky) by a wild virgin female. Conversely, the first probable case of male parthenogenesis (arrehnotoky) in scorpions is reported based on the production of two successive all-male broods by a wild caught virgin female of Tityus metuendus Pocock from Peru. Both species were found in isolated palm trees within the rain forest, conforming with the insular theory of parthenogenesis. With the exception of mites, all-female re- production is quite rare within the order Arachnida (Taberly 1987; Palmer & Norton 1991; Norton & Palmer 1991; Nagelkerke & Sabelis 1991), but has also been demonstrated in a few species of harvestmen (Tsurusaki 1986), spiders (Lake 1986; Deelman-Reinhold 1986; Camacho 1994) and scorpions (Lour- engo & Cuellar 1994). Among the almost 1500 species of scorpions throughout the world, only five are known to be partheno- genetic (Louren^o & Cuellar 1994). The first case was reported by Matthiesen (1962) in the Brazilian species Tityus serrulatus Lutz & Mello. Wild pregnant females were collected and their all-female progeny reared individu- ally, giving virgin birth to a second all-female generation several months later. Matthiesen’s findings were later confirmed by San Martin & Gambardella (1966). Since then, T. serru- latus has been relegated to Tityus stigmurus (Thorell) (Lourengo & Cloudsley-Thompson 1996) a parthenogenetic species consisting of at least four distinct all-female morphs (Lour- engo & Cloudsley-Thompson this volume) of which the original T. serrulatus is one. The other four parthenogenetic species are Tityus uruguayensis Borelli from Uruguay and Bra- zil, Tityus columbianus (Thorell) from Colom- bia, Hottentota hottentota (Fabricius) from West Africa, and Liochelis australasiae (Fa- bricius) from the South Pacific (Lourengo & Cuellar 1994). Tityus trivittatus Kraepelin from Argentina is also suspected of parthe- nogenesis (Peretti 1994; Maury 1997). In this paper, we report an additional parthenogenetic scorpion {Ananteris coineaui Louren^o from French Guyana), and the first observation of all-male broods in scorpions (Tityus metuen- dus Pocock from Peru). METHODS Scorpions were raised individually in plas- tic boxes terraria, with different sizes ranging from 6/5/4 to 36/14/24 cm. The botton of each terrarium was covered with a soil layer and water was supplied in Petri dishes. Food, con- sisting on crickets and spiders of the genus Pardosa, was supplied once a week. The ter- raria were placed in a room where temperature was maintained at 25 °C ±2 °C. Humidity ranged from 60-70%. The sex of individuals was determined by the examination of the size and sexual dimor- phism of the pectines. For details and illustra- tions see Farzanpay & Vachon (1979) and Lourengo (1983). When life cycles are completed, voucher material will be deposited in the Natural His- tory Museum, Paris (A. coineaui) and in the Zoologisches Museum of the University of Hamburg (T metuendus). RESULTS Ananteris coineaui Louren^o. — Ananteris coineaui was described from a rain forest near 149 150 THE JOURNAL OF ARACHNOLOGY ; • Ananteris coineaui ^ O Tityus metuendus 540 km 330 mi Figure 1 . — Map showing the areas where the parthenogenetic females of Ananteris coineaui and Tityus metuendus have been collected. the Arataye river in French Guyana, based on three adult females collected in a palm tree of the species Astrocaryum paramaca Martins (Louren9o 1982; Kahn 1997). Since then, only one additional specimen was collected from Saiil (close to the original locality), also in a palm tree. Within about two weeks, the female molted; by March 30, she gave birth to 16 Figure 2. — The parthenogenetic female of Tityus metuendus. LOURENgO & CUELLAR— PARTHENOGENESIS IN SCORPIONS 151 young, which remained on her back until April 6 when they all died before molting. An examination of the size and sexual dimor- phism of the pectines revealed that the entire brood consisted of females, suggesting parthe- nogenesis. Ananteris coineaui is probably en- demic to the central region of French Guyana, and based on the rarity of field specimens, probably has a very low population density. Since this genus was created by Thorell in 1891, the number of species described is now 20 (Louren^o 1997a). The number of speci- mens representing these species remains very small (less than 200), suggesting that most species are rare. Only the original species de- scribed by Thorell, A. balzani, seems abun- dant and has a much larger range of distri- bution (Lourengo 1993). Nearly 50 specimens of A. balzani were collected since 1975, with a sex ratio of approximately ld:2$. Of the remaining 19 species, 12 are represented by less than 5 specimens each, and 4 by only a single specimen. Males are rare, having been found in only 10 of the 19 species. As noted by Camacho (1994) for spiders, female-biased sex ratios may be taken as ev- idence of parthenogenesis. This could also be true for the genus Ananteris and other mi- croscorpions in which males are rare or ab- sent. With the exception of a recent study on reproductive effort between sexual and par- thenogenetic populations of Tityus columbi- anus (Lourengo et al. 1996), virtually nothing is published on the life history and behavior of parthenogenetic scorpions. The occupancy of isolated palm trees with- in vast areas of rain forest or savanna con- forms with the concept of insular partheno- genesis proposed by Cuellar (1977, 1994). Most well studied parthenogenetic animals oc- cur in insular habitats such as isolated caves, spring heads, bogs, termite nests, rotting logs, tree trunks, hibemacula and oceanic islands (Cuellar 1994). Most parthenogens are also characterized by small size, low mobility, and low population density (Cuellar 1994). The rarity of A. coineaui, its occurrence on isolat- ed palm trees, and the absence of males all suggest a parthenogenetic mode of reproduc- tion. This may hold as a rule for other species in this genus. Tityus metuendus Pocock. — Tityus me- tuendus is a rain-forest species distributed mainly in western Amazonia between Brazil and Peru. In the vicinity of Manaus, Brazil, specifically the Ducke Reserve, the popula- tions of T. metuendus are strictly sexual with a sex ratio of 1/1 (Lourengo 1983, 1997b). During some recent collections (1996) in the Amazonian region of Peru, near Iquitos (town of Jenaro Herrera), a single pre-adult female of Tityus metuendus was collected by Dr. G. Couturier (Orstom-Museum) from a palm tree of the species Astrocaryum chambira (Kahn 1997) and brought to one of us (WRL). On 18 October 1996, about seven months after the last molt, this female gave birth to a brood of 21 neonates, of which only three survived to the adult stage, all males. An examination of the pectines of the remaining preserved im- matures revealed that the entire brood was male. On 29 September 1997, the same female produced a second brood of 32 neonates, of which three did not complete embryological development and 29 were normal. The normal ones all died a few days later after the first molt. As with the previous brood, examination of the pectines revealed only males. A third all-male brood was observed on the 30 April 1998. The production of three consecutive all- male broods by this virgin female may well represent the first case of arrehnotoky in scor- pions, and possibly among Arachnida other than Acari (Nagelkerke & Sabelis 1991). No data are presently available in scorpions either to explain the meiotic mechanisms of arrhe- notoky or its evolutionary significance (Bull 1983), as exist for other groups such as the Hymenoptera (Waage 1986; Cuellar 1987) and mites. According to Taylor & Sauer (1980), a major selective advantage of arrhe- notoky compared to diploidy is that mothers can precisely determine sex ratio by control- ling the fertilization of each egg. This is par- ticularly advantageous in species with finite mating groups, in which the probablity is high that some clutches may contain no males (Na- gelkerke & Sabelis 1991), or that the sex ratio may be biased in favor of females (Charnov 1982). Precise sex ratios have been docu- mented in several arrhenotokous species of parasitic wasps (Waage 1986), which lay their eggs either in a single host or a clumped group of hosts. In phytoseiid mites, pseudo-arrhe- notoky has apparently arisen as a consequence of low mobility and a subdivided population structure. Their dominant prey form patchy in- 152 THE JOURNAL OF ARACHNOLOGY festations which are probably invaded by only a few females, leading to very small mating groups (Nagelkerke & Sabelis 1991). Similar mating conditions may exist for T. metuendus, but extensive field work is needed to under- stand its life history and behavior, ACKNOWLEDGMENTS We are most grateful to Dr. G. Couturier, ORSTOM-Museum for the use of Fig. 2. LITERATURE CITED Bull, JJ, 1983. The evolution of sex chromosomes and sex determining mechanisms. Benjamin & Cummings, Menlo Park, California. 316 p, Camacho, J.P. 1994. Female-biased sex ratio in spi- ders caused by parthenogenesis. Hereditas, 120: 183-185. Chamov, E.L. 1982. The Theory of Sex Alloca- tion. Monographs in Population Biology. Prince- ton Univ. Press, Princeton, No. 18. 355 p. Cuellar, O. 1977. Animal parthenogenesis. Sci- ence, 197:837-843. Cuellar, O. 1987. The evolution of parthenogene- sis: a historical perspective. Pp. 43-104, In Mei- osis. (P.B. Moens, ed,). Academic Press, Inc. New York. Cuellar, O. 1994. Biogeography of parthenogenetic animals. Biogeographica, 70(1): 1-13. Deeleman-Reinhold, C.L. 1986. Dysdera hungar- ica Kulczynski - a case of parthenogenesis? Ac- tas X Congr. Inst. AracnoL, Jaca/Espana, 1:25- 31. Farzanpay, R. & M. Vachon. 1979. Contribution a r etude des caracteres sexuels secondaires chez les scorpions Buthidae (Arachnida). Rev. Arach- noL, 2(4):137-142. Kahn, E 1997. Les palmiers de F Eldorado. OR- STOM Editions, Paris, 252 p. Lake, D.C. 1986. Possible parthenogenesis in the huntsman spider Isopoda insignis (Araneae, Sparassidae). J. ArachnoL, 14:129. Lourengo, W.R. 1982. Revision du genre Anariteris Thorell, 1891 (Scorpiones, Buthidae) et descrip- tion de six especes nouvelles. Bull. Mus. Natn. Hist. Nat, Paris, 4e ser. 4 (Al/2):1 19-151. Lourengo, W.R, 1983. Contribution a la connaiss- ance du Scorpion amazonien Tityus metuendus Pocock, 1897 (Buthidae). Stud. Neotrop. Fauna Environ., 18(4): 185-193. Lourengo, W.R. 1993. A review of the geographi- cal distribution of the genus Ananteris Thorell (Scorpiones; Buthidae), with description of a new species. Rev. Biol. Trop., 41(3):697-701. Lourengo, W.R. 1997a. A reappraisal of the geo- graphical distribution of the genus Ananteris Thorell (Scorpiones, Buthidae). Biogeographica, 73(2):81-85. Lourengo, W.R. 1997b. Additions a la faune de scorpions neotropicaux (Arachnida). Rev. Suisse ZooL, 104(3):587-604. Lourengo, W.R. & J.L. Cloudsley-Thompson. 1996. Effects of human activities on the envi- ronment and the distribution of dangerous spe- cies of scorpions. Pp. 49-60, In Envenomings And Their Treatments. (C. Bon & M. Goyffon, eds.). Edit. Fondation M. Merieux, Lyon. Lourengo, W.R. & O. Cuellar. 1994. Notes on the geography of parthenogenetic scorpions. Biogeo- graphica, 70(1): 19-23. Louren^o, W.R., O. Cuellar & ER. Mendez de la Cruz. 1996. Variation of reproductive effort be- tween parthenogenetic and sexual populations of the scorpion Tityus coiumbianus. J, Biogeogr., 23:681-686. Matthiesen, FA. 1962. Parthenogenesis in scorpi- ons. Evolution, 16(2):255-256. Maury, E.A. 1997. Tityus trivittatus en la Argen- tina. Nuevos datos sobre distribucion, parteno- genesis, sinantropia y peligrosidad (Scorpiones, Buthidae). Rev. Mus. Argentino. Cienc. Nat., 24: 1-24. Nagelkerke, C.J. & M.W Sabelis. 1991. Precise sex-ratio control in the pseudo-arrhenotokous phytoseiid mite Typhlodromus occidentalis Nes- bitt. Pp. 193-207, In The Acari. Reproduction, development and life-history strategies. (R. Schuster & P.W. Murphy, eds.). Chapman & Hall, London. Norton, R.A. & S.C. Palmer. 1991. The distribu- tion, mechanisms and evolutionary significance of parthenogenesis in oribatid mites. Pp, 107- 136, In The Acari, Reproduction, Development and Life-history Strategies. (R. Schuster & P.W. Murphy, eds.). Chapman & Hall, London. Palmer, S.C. & R.A. Norton. 1991. Taxonomic, geographic and seasonal distribution of thelyto- kous parthenogenesis in the Desmonomata (Ac- ari: Oribatida). Exper. Appl. AcaroL, 12:67-81. Peretti, A. 1994, Comportamiento de relacion mad- re-cria de Tityus trivittatus Kraepelin, 1898 (Scorpiones, Buthidae). Bol. Soc, Biol. Concep- cion, 65:9-21. San Martin, P. & L.A. Gambardella. 1966. Nueva comprobacion de la partenogenesis en Tityus ser- rulatus (Lutz y Mello-Campos, 1922 Scorpioni- da, Buthidae). Rev. Soc. Ent. Argentina., 28(1- 4):79-84. Taberly, G. 1987. Recherches sur la parthenoge- nese thelytoque de deux especes d'Acariens Or- ibates: Trhypochthonius tec to rum (Berlese) et Platynothrus peltifer (Koch), 1. Acarologia, 28(2):187-198. Taylor, P.D. & A. Sauer. 1980, The selective ad- vantage of sex-ratio homeostasis. American Nat., 116:305-310. Tsurasaki, N. 1986. Parthenogenesis and geograph- LOURENgO & CUELLAR— PARTHENOGENESIS IN SCORPIONS 153 ic variation of sex ratio in two species of Leiob- unum (Arachnida, Opiliones). Zool. Sci., 3(3): 517-532. Waage, J.K. 1986. Eamily planning in parasitoids: adaptive patterns of progeny and sex allocation. Pp. 63-95, In Insect Parasitoids. (J.K. Waage & D. Greathead, eds.). Academic Press, London. Manuscript received 3 April 1998, revised 6 Sep- tember 1998. 1999. The Journal of Arachnology 27:154-158 DISCOVERY OF A SEXUAL POPULATION OF TITYUS SERRULATUS, ONE OF THE MORPHS WITHIN THE COMPLEX TITYUS STIGMURUS (SCORPIONES, BUTHIDAE) Wilson R. Lourengo: Laboratoire de Zoologie (Arthropodes), Museum National d’Histoire Naturelle, 61 rue de Buffon 75005 Paris, France John L. Cloudsley-Thompsonj 10 Battshill Street, Islington, London N1 ITE, United Kingdom ABSTRACT. Tityus serrulatus Lutz & Mello 1922 (in fact, the form confluenciata within the Tityus stigmurus complex) is an extremely toxic scorpion of considerable medical importance in Brazil. Its rapid spread is partially due to parthenogenesis. Speculation regarding the occurrence of sexual individuals has been resolved by the discovery of a population, described here, having a male-female sex ratio of 1/2.5. Four color morphs of the T. stigmurus complex are described, and it is concluded that T. serrulatus and Tityus lamottei Lourengo 1981 are junior synonyms of T. stigmurus (Thorell 1877). Scorpionism is well known in Brazil and has been documented there since the end of the 19th century. Two species are currently associated with most incidents of medical im- portance, Tityus serrulatus Lutz & Mello 1922 and Tityus bahiensis (Perty 1934). The first comprehensive study of the phenomenon of scorpionism was that of Maurano (1915); this work, however, dealt primarily with Tityus bahiensis (Perty), the second most toxic spe- cies in South America. This species was orig- inally described from Brazil. In recent years, the geographic range of the Brazilian scorpion T. serrulatus has increased considerably (Lourengo & Cloudsley-Thompson 1996). In Brazil, this species poses an exceptional health problem due to its rapid expansion in urban areas, its sudden proliferation and its great toxicity. Since the description of Tityus serrulatus, most attention has been focused on its medical importance. However, it was observed that males are absent from all known populations, and Matthiesen (1962) first demonstrated that this species reproduces by parthenogenesis. This phenomenon although rare, has been demonstrated in other species of scorpions, also (Lourengo & Cuellar 1994). Tityus ser- rulatus was considered to be an obligate par- thenogenetic species because bisexual popu- lations had not been detected. Moreover, the apparent absence of related bisexual individ- uals within its known geographic distribution suggested that the generating species either had been eliminated after giving rise to par- thenogenesis, or that T. serrulatus evolved elsewhere and has since occupied an extensive region from which its bisexual progenitors were absent (Lourengo & Cuellar 1994; Lour- engo & Cloudsley-Thompson 1996; Lourengo et al. 1996). METHODS The specimens of the sexual population of Tityus serrulatus were collected during day- time and were found under logs and bark. The area where this population was located pre- sents a transitional vegetation type ranging from dry forests and cerrados to caatingas. In contrast, parthenogenetic populations are only known from modified sites and are often found inside cities and towns, where they live inside houses, but are also frequent in ceme- teries and even in the sewer system. For de- tails see (Lourengo & Cuellar 1995; Lourengo & Cloudsley-Thompson 1996; Lourengo et al. 1996). Identification of sex was based both on ex- ternal features and on dissection of adult males and females. The voucher material con- cerning the sexual population of Tityus ser- rulatus is partially deposited in the Natural History Museum in Paris, but also in the Eze- quiel Dias Foundation in Belo Horizonte, Bra- zil. 154 LOURENgO & CLOUDSLEY-THOMPSON— SEXUAL POPULATION, TITYUS 155 RESULTS AND DISCUSSION Lourengo (1981) suggested that T. serru- latus was closely related to Tityus stigmurus (Thorell 1877), a bisexual species with a range of distribution further north of T. ser- rulatus. Several other authors had discussed the status of these two species and their pos- sible relationship. In the opinion of some (Pessoa 1935; Mello-Leitao 1939; Eickstedt 1983), both species should be considered dis- tinct. Others asserted that, before about 1920, Tityus stigmurus had been a common species in the central and southern regions of Brazil (States of Minas Gerais, Sao Paulo and Goias), and that the two species are varieties of a single species (Mello-Campos 1924; Vel- lard 1932). A few years ago, Lourengo (un- publ.) checked the notes of Vellard and some of his collected material. This confirmed that Tityus stigmurus was undoubtedly a common species in the State of Minas Gerais and south of Goias, until at least the 19th Century (Lour- en^o & Cloudsley-Thompson 1996). Lourengo & Cloudsley-Thompson (1996) and Lourengo et al. (1996) suggested that the sexual and the parthenogenetic populations of a complex T. stigmurus/T. serrulatus might correspond respectively to the northern range of T. stigmurus and the southern range of T. serrulatus in Brazil. However, recent unpub- lished field observations by Lourengo show that both the morphs T. serrulatus (= con- fluenciata) and T. stigmurus {= unifasciata) reproduce by parthenogenesis. Moreover, the sexual individuals of T. stigmurus occur in an undisturbed region of Exu in the State of Per- nambuco, whereas the parthenogenetic popu- lations are found among human communities along the coastal regions of its northern range. Two other sexual morphs {confluenciatai maculata and trisfaciata) occur in undisturbed regions. The first corresponds to the species described by Lourengo (1981) as Tityus la- mottei. It occurs in a transitional zone between two major natural ecosystems, Cerrados and Caatingas in the western part of the State of Bahia, whereas trisfaciata occurs in the State of Ceara at the extreme northern end of its range. Although these species are at present sexual, we speculate that future human distur- bance could possibly lead to their replacement by parthenogenetic counterparts (see Louren- go & Cuellar 1995; Lourengo & Cloudsley- Thompson 1996) (Fig. 1). Recent field observations on the polymor- phic patterns of pigmentation (Lourengo in press) suggested that the above distinction be- tween T stigmurus {= sexual) and T. serru- latus (= parthenogenetic) forms, is not suffi- ciently comprehensive. The following classification of color morphs is therefore pro- posed: (a) Morph unifasciata: with a single median longitudinal dark stripe over the body as observed in the species named T. stigmurus (Thorell 1877). (b) Morph confluenciata: with confluent dark spots over the tergites as ob- served in the species named T. serrulatus Lutz & Mello 1922. (c) Morph confiuenciata/ma- culata: with the same pattern as in b, but with dark spots over the pedipalps and legs, as ob- served in the species named T. lamottei Lour- engo 1981. (d) Morph trifasciata: with three longitudinal dark stripes over the body as ob- served on an undescribed population from the State of Ceara. Other patterns probably exist but have not yet been documented. Where the sexual populations of the con- fluenciata {= T serrulatus) form are distrib- uted remains unsolved. Recently a sexual pop- ulation was located by one of the authors (WRL) in the north of the State of Minas Ger- ais, Brazil, in the region of Irape close to the Jequitinhonha river (Fig. 1). A sample of 39 specimens containing 12 males, 27 females and immature individuals of different instars was collected, giving a sex- ratio of 1 to 2.25 in favor of females. This is close to the sex-ratios observed in other spe- cies of Tityus (Lourengo, 1980). Detailed ex- amination of all these specimens revealed few differences from the morphology of the par- thenogenetic population. The pattern of col- oration is very similar in both males and fe- males, and in both sexual and parthenogenetic populations. The general morphology of the females is also similar in both the sexual and the parthenogenetic populations, with the ex- ception of body size which seems to be slight- ly larger and bulkier in the sexual females. The general morphology of the males is quite different, however, from that of the females. The pedipalps are longer and more slender (Figs. 2, 3). The pectines are larger although the total number of teeth is almost the same in both sexes: males, 22-27; females, 22-26. 156 THE JOURNAL OF ARACHNOLOGY O Confluenciata/maculata form ▼ Trifasciata form # Sexual population of conflueneiata form Figure L — Present geographical distribution of Tityus stigmurus (morph unifasciata) and Tityus ser- rulatus (morph conflueneiata). The black represents the contact zone between the two populations. In detail, the localities where Tityus lamottei (morph confluenciata/maculata), morph trifasciata and the sexual population of T. serrulatus have been collected. The first two differences characterize males of many species (Polis & Sissom 1990). One major morphological difference was, however, observed between the sexual and parthenogenetic populations. The typical trait which characterizes the parthenogenetic pop^ ulation, as suggested by the Latin name ""ser- rulatus”, is the presence of granules modified as spines on the posterior region of the dorsal keels of metasomal segments III and IV. The number of spines varies from 2-10. In the sex- ual population, these spines were present in the immature instars, but disappeared after the last molt and were absent in adults (Figs. 4- 6). This difference cannot yet be explained but is almost certainly associated with genetic dif- ferences between the two populations corre- lated with their modes of reproduction. Fur- ther studies on the biology of the sexual population are required. Diagnosis of the sexual comfiuenciata form.— Scorpions of medium size ranging from 55-70 mm in total length. General col- oration yellowish. Metasoma: segments I to V yellowish with blackish lateral and ventral spots: 10-10-8-8-5 keels present. The dorsal keels of segments III and IV with the posterior granules modified as spines varying in num- ber from 8-10, only present in immature ie- stars; absent from adults. Dentate margins of pedipalp-chela fingers composed of 13-17 oblique rows of granules. Telsoe with a long curved aculeus; subaculear tooth strong and spinoid. Pectines with 22-27 teeth; slightly LOURENgO & CLOUDSLEY-THOMPSON— SEXUAL POPULATION, TITYUS 157 Figures 2-6. — Morphological characters of sexual and parthenogentic populations of the confluenciata form. 2, 3. Pedipalps in dorsal view of male and female sexual individuals showing dimorphism; 4. Metasomal segments IV, V and telson in lateral view of an adult sexual female showing the absence of typical spines; 5. Metasomal segments III to V and telson, in lateral view, of an adult parthenogenetic female population showing the typical spines; 6. Same with metasomal segments III and IV in detail. 158 THE JOURNAL OF ARACHNOLOGY greater than that observed in the parthenoge^ netic population (19-24) (see Lourengo & Ei- ckestedt 1981). For nomenclatural correctness, only the name Tityus stigmurus (Thorell 1877) should be retained, while Tityus serrulatus Lutz & Mello 1922 and Tityus lamottei Lourengo 1981 must be considered as junior synonyms of Tityus stigmurus. The designations unifas- data, confluendata, confluendata/maculata and trifasdata provide a practical way of identifying the different forms. Phylogenetic work at the molecular level would be of great value for a better definition of the genetic relationships among the differ- ent forms of the Tityus stigmurus complex. A beginning has already been made and this mo- lecular work is in progress with research teams in Mexico and Brazil. ACKNOWLEDGMENTS We are most grateful to Mr. Jacques Rebi- ere, Laboratoire de Zoologie Arthropodes, for preparing several illustrations. LITERATURE CITED Eickstedt, V.R.D. 1983. Escorpionismo por Tityus stigmurus no nordeste do Brasil (Scorpiones; Buthidae). Mem. Inst. Butantan, 47/48:133-137. Louren^o, W.R. 1980. Contribution a la connaiss- ance systematique des scorpions appartenant au (( complexe )> Tityus trivittatus Kraepelin, 1898 (Buthidae). Bull. Mus. Natn. Hist. Nat., Paris, 4e ser., 2 (A3):793~843. Lourengo, W.R. 1981. Sur la systematique des scorpions appartenant au complexe Tityus stig- murus (Thorell, 1877) (Buthidae). Rev. Brasilei- ra Biol., 41(2):35 1-362. Lourengo, W.R. (in press). A guide to the scorpions of Brazil. Mem Soc. Biogeogr., Paris. Lourengo, W.R. & J.L. Cloudsley-Thompson. 1996. Effects of human activities on the envi- ronment and the distribution of dangerous spe- cies of scorpions. Pp. 49-60, In Envenomings and their treatments. (C. Bon & M. Goyffon, eds.). Edit. Fondation M. Merieux, Lyon. Lourengo, W.R., J.L. Cloudsley-Thompson, O. Cuellar, V.R.D. von Eickstedt, B. Barraviera & M.B. Knox. 1996. The evolution of scorpionism in Brazil in recent years. J. Venom. Anim. Tox- ins, 2(2): 121-134. Lourengo, W.R. & O. Cuellar, 1994. Notes on the geography of parthenogenetic scorpions. Biogeo- graphica, 70(1): 19-23. Lourengo, W.R. & O, Cuellar. 1995. Scorpions, scorpionism, life history strategies and parthe- nogenesis. J. Venom. Anim. Toxins, l(2):51-62. Louren^o, W.R. & V.R.D. von Eickestedt. 1981. A proposito da indicagao de um neotipo para Tityus serrulatus Lutz e Mello, 1922 (Scorpiones; Buth- idae). Mem. Inst. Butantan, 44/45:181-190. Matthiesen, EA. 1962. Parthenogenesis in scorpi- ons. Evolution, 16(2):255-256. Maurano, H.R. 1915. O escorpionismo. These Fa- cul. Medic., Rio de Janeiro. Mello-Campos, O. 1924. Os escorpioes brazileiros. Mem. Inst. Osw. Cruz, 17(2):237-363. Mello-Leitao, C. 1939. Revisao do genero Tityus. Physis, 17:57-76. Pessoa, S.B. 1935. Nota sobre alguns escorpioes do genero Tityus e Bothriurus. Ann. Paulista Med. Cir., 29(5):429-436. Polis, G.A. & W. D. Sissom. 1990. Life History. Pp. 161-223, In The Biology of scorpions. (G.A. Polis, ed.), Stanford Univ. Press, Stanford. Vellard, J. 1932. Scorpions. In Mission scientifique au Goyaz et au Rio Araguaya. Mem. Soc. ZooL France, 29(6):539-556. Manuscript received 3 April 1998, revised 6 Sep- tember 1998. 1999. The Journal of Arachnology 27:159-164 ACTIVITY RHYTHMS AND BEHAVIORAL CHARACTERIZATION OF TWO EPIGEAN AND ONE CAVERNICOLOUS HARVESTMEN (ARACHNIDA, OPILIONES, GONYLEPTIDAE) Sonia Hoenen and Pedro Gnaspini: Departamento de Zoologia, Institute de Biociencias, Universidade de Sao Paulo, Caixa Postal 11461, 05422-970, Sao Paulo, SP, Brasil ABSTRACT. The activity rhythms, feeding behavior, and reaction to light of two epigean (surface inhabitant) species of harvestmen (Iporangaia pustulosa and Iguapeia melanocephala) and of one cav- emicolous species {Pachylospeleus strinatii) have been recorded. Both the epigean and the cavemicolous species showed a highly pronounced circadian rhythmicity. The cave species showed a bimodal pattern. Whereas the epigean species carried food away to feed, the cave species fed where they found the food. The time of reaction to light did not differ statistically between species. However, when exposed to light, the cave species walked much longer distances after it started walking. These differences are probably due to cave adaptation. The cave species may have to wander further for food (and maybe mates) because of the scarcity of resources and, therefore, show greater activity and also a tendency to exploit a resource wherever they find it. The cave environment is characterized by darkness, at least in deep regions, and the con- sequent reduction or absence of photoautotro- phic organisms, high relative humitidy (al- most 100%), and small temperature variations over both a daily and annual basis (Barr 1968; Poulson & White 1969; Howarth 1983). These peculiar conditions promote the establishment of a characteristic fauna which may possess morphological, physiological and behavioral modifications that allow them to find food and sexual partners (Culver 1982; Parzefall 1986; Holsinger 8l Culver 1988), especially among species restricted to the subterranean environ- ment (the troglobites). These modifications may include reduction or loss of eyes and pig- mentation, improvement of other sensorial or- gans, etc. Behavioral patterns that have lost their biological meaning in the new habitat (e.g., avoidance of predators, which may be absent) or for which a cue is absent (e.g., those behaviors related to vision) could be su- pressed in the troglobites. In turn, new behav- ioral patterns related to the new habitat con- ditions may arise. One of the most conspicuous biological dif- ferences between troglobitic and epigean spe- cies is related to endogenous, self-sustained biological rhythms. Because surface environ- mental conditions oscillate cyclically, the abil- ity to anticipate temporal changes in the en- vironment would enable an organism to be prepared, both physiologically and behavior- ally, to perform specific activities when the environmental conditions are most favorable to the species. This confers to the organism the property of being continuously adjusted to the cyclic changes of the environment and, therefore, of being temporally adapted (Marques & Waterhouse 1994). In non-troglobitic cavemicolous species (those which may and those which must leave the cave during their life), the presence and synchronization of endogenous rhythms would guarantee the time adjustment of exits and returns to the cave. Indeed, after Saunders (1982), studies focusing on temporal patterns with animals of each one of these categories show that activity patterns appear to reflect the relationship that each animal has with the cave. In contrast, it is generally accepted that the internal clocks of troglobites have been supressed (Lamprecht & Weber 1991). Nev- ertheless, as cycles (even attenuated ones) are present in some deep regions of caves, it is possible that some rhythmic characteristics persist. Although there are numerous troglobitic 159 160 THE JOURNAL OF ARACHNOLOGY species, few studies on biology of opilionids have been conducted (e.g., Juberthie 1964; Gnaspini 1996). We present herein informa= tion on activity rhythms and general behavior of three species of harvestmen, one troglobitic and two epigean, aiming the comparison of strategies in different habitats and the contri- bution to the knowledge of biology of opi- lionids in general. METHODS Three species of Laniatores harvestmen (family Gonyleptidae) were used in this study. The two epigean species, Iporangaia pustu- losa Mello-Leitao 1935 and Iguapeia melan- ocephala Mello-Leitao 1935, belong to the subfamily Progonyleptoidellinae, and the trog- lobitic species, Pachylospeleus strinatii SiL havy 1974, belongs to the monotypic subfam- ily Pachylospeleinae. They were chosen for two reasons. First, they are abundant and were easily available for our study. Second, one species is a restricted cave species and the others live only outside of caves. Thus, they have contrasting characteristics. The species were collected in the Ribeira Valley, Sao Paulo State, southeastern Brazil. This area is a humid subtropical region with- out a dry season; total rainfall is 1500 mm; and the annual average temperature ranges be- tween 17-19 °C (Setzer 1966; Monteiro 1973; see also Gnaspini 1996 for description of ar- eas). Recent phylogenetic analysis of the family Gonyleptidae (R. Pinto-da-Rocha, pers. comm.) has shown that Progonyleptoidellinae is the sister group of Sodreaninae + Caelo- pyginae. Species of this whole clade (three subfamilies) could be considered “diurnal” because they can easily be seen during the day, active or inactive. On the other hand, Pa- chylospeleinae belongs to a “nocturnal” clade. Epigean species of this second clade completely hide during the day, leaving their shelters only after dusk. Iporangaia pustulosa and /. melanocephala are epigean species which live on tree trunks and on or under leaves in rainforests from southeastern Brazil. Our observations have shown that they are not gregarious species be- cause the specimens have been always ob- served alone or, at most, wandering near co- specifics. Specimens of /. pustulosa are generally seen walking on low vegetation near damp areas during the morning, whereas spec- imens of /. melanocephala are generally seen walking on tree trunks on the same areas. These two species were collected in the area of Parque Estadual Intervales (PEI), a moun- tainous region with elevation ranging from 70-1000 m (collections were made at about 800 m). Nevertheless, their geographical dis- tribution includes the area where the third spe- cies studied is found. The troglobitic species (P. strinatii) was collected in the Aguas Quentes cave (SP-016, elevation 180 m), located in Parque Estadual Turistico do Alto Ribeira (PETAR), at lower elevations than PEI. It is restricted to one sys- tem of caves (“Sistema das Areias”) and has been studied primarily at the population level by Pinto-da-Rocha (1996). It does not seem to be gregarious either. Its endemic distribution, allied with troglomorphisms (depigmentation and increased segmentation of tarsus of the sensorial leg II), have led Pinto-da-Rocha (1996) to consider it a troglobite. Silhavy (1974) has also considered it a troglobite, and has added to the list of troglomorphisms the reduction of eyes. However, the species has eyes, which seem to have “normal” size. We maintained all harvestmen species in terraria with very damp soil on the bottom. Seven individuals of /. pustulosa and one of /. melanocephala were kept under natural day- light : dark cycles, and four individuals of the cavernicolous P. strinatii were kept in a base- ment laboratory. Inside the laboratory the ter- raria were left in a chamber with a dark and humid environment which approximates the cave environment. All species were fed once a week with pieces of Tenebrio obscurus lar- vae. Iporangaia pustulosa was also fed with pieces of carrot and lettuce. During feeding, we observed behavior under natural illumi- nation or with fluorescent bulbs. We conducted the tests of reaction to light in a box 35 cm long with a small side retreat in the back. A glass with water provided a thermal barrier between the front of the box and an incandescent bulb (light intensity at this point = 340 lux). An individual was placed in the box 15 minutes before the test, in the end nearest the lamp (which was off) under very dim light (4 lux). During tests, the light was switched on, and the time elapsed for the animal to start walking (time of reac- tion) as well as the total distance walked dur- HOENEN & GNASPINI— ACTIVITY RHYTHMS AND BEHAVIOR OF HARVESTMEN 161 24/Oh 18h 12h 6h 24/Oh 1 2 3 Figures 1-3. — Results of the circular analysis of the activity rhythms of the harvestmen. 1. Iporangaia pustulosa; 2. Iguapeia melanocephala; 3. Pachylospeleus strinatii. The circumference represents 24 hours; each line represents the grouped data (all days recorded) of percentile activity per hour. ing 15 minutes were recorded. Tests were con- ducted in different hours of the day, and, in each test, all animals have been tested at the same time. This procedure avoided both re- action differences due to endogenous timing, and differences due to specific different peri- ods of rest/activity. The results were compared using analysis of variance (ANOVA) (Zar 1996). Continuous records of activity rhythms were made using a system that detects vibra- tions and sends the data to a computer. Through acoustic sensors connected to the walls, the system recorded overall activity, by detecting all body movements of the animals, including walking, grasping and chewing the food, cleaning legs with mouth parts, etc. All records were made under continuous red light (2 lux, 680 nm), because it is well accepted among entomologists and arachnologists that this type of light does not affect the behavior of these arthropods. Moreover, previous tests showed that this light condition do not disturb the behavior of these animals. The animals were provided with food and water in the be- ginning of the experiment and after one week. Data were analysed using the Rayleigh test, which tests if there is a preferential direction for a circular unimodal distribution. When Rayleigh was inappropriate, the Hodges- Ajne test, which can be applied to samples with any distribution, even multimodal, was used (Zar 1996). RESULTS Activity rhythms. — The results of the re- cords of activity rhythms are presented in Figs. 1-3. The circumference represents 24 h; each line represents the grouped data (all days recorded) of percentile activity per hour. Only the clearest record for each species is shown here. The record of /. pustulosa was made over 7 days. The circular analysis (Fig. 1) indicates that the activity rhythm has a circadian peri- odicity (Rayleigh test: R = 24.18; r = 0.41; n = 59; P < 0.05; this test is more significant the closer r gets to 1.0), with the major phase of activity occuring around 1700 h. A second and much lower activity peak occurs during the morning. The record of I. melanocephala continued for 1 1 days. The circular analysis (Fig. 2) also shows a circadian periodicity (Rayleigh test: R = 350.43; r - 0.693; n = 506; P < 0.05), with an activity peak around 1400 h. The record of P. strinatii was made for 10 days. The circular analysis (Fig. 3) shows a circadian periodicity in the activity rhythm. However, the activity pattern was clearly bi- modal, thus the Rayleigh test could not be ap- plied. In order to statistically analyze these data, we utilized the test of Hodges- Ajne, that indicates a very significant circadian period- icity (Median = 21 h; n = 120; m = ^; P < 0.05). This means a bimodal preference for 162 THE JOURNAL OF ARACHNOLOGY Table 1. — Results of the reaction to light of two individuals of Iporangaia pustulosa and two of Pa- chylospeleus strinatii. TR. = Time to start Reaction (min); T.D.W. = Total Distance Walked (cm). represents no displacement. Date (Time) /. pustulosa P. strinatii ind. 1 ind. 2 ind. 1 ind. 2 TR. T.D.W. TR. T.D.W. TR. T.D.W. TR. T.D.W. 13/Jan/98 (lOOOh) 4.83 10 5.38 25 2.97 257 7.5 253 16/Jan/98 (lOOOh) — — 4.833 15 12.17 41 — — 19/Jan/98 (0400h) 1.5 18 1.9 93 0.42 203 17.8 55 22/Jany98 (1600h) — — - 0.5 63 1.2 184 — — 29/Jan/98 (1400h) — — 7.65 27 3.6 28 10.4 27 activity at 0900 and 2100 h. The record also shows that this species walks long distances, because the activity is very intense. Feeding behavior. — Neither the epigean nor the cavernicolous species showed any pat- tern of intraspecific aggressive behavior, even when feeding occurred after a long time under starvation conditions. Instead, individuals of /. pustulosa stayed very close, touching each other with their sensorial legs while feeding. Even interspecific aggressiveness among /. pustulosa, I. melanocephala and P. strinatii was not observed. In turn, in previous obser- vations, individuals of Goniosoma proximum (Mello-Leitao 1922) actually expelled /. me- lanocephala from the food (S. Hoenen, pers. obs.). The former are much larger than the latter. G. proximum (Goniosomatinae) can be found either in the same area where the two epigean species occur, either in some caves. The animals studied have been collected from a granitic cave. Whether or not it is a proper cavernicolous species or if it is an accidental in caves is difficult to assure, because they may or may not inhabit caves in areas where they occur in the forest (see Gnaspini 1996 for discussion). We observed that all epigean tested species remove pieces of food to ingest away from the source. This movement away from the food source appears to be mediated by contact. The harvestmen seem to stop for ingestion only after they leave the area where they touch one another. This could be a behavior to avoid fighting for food among conspecifics. How- ever, no aggressive pattern was observed when individuals casually meet each other. Moreover, this behavior occurred only when small pieces of food (e.g., pieces of beetles and of carrot) were offered; when lettuce was offered, possibly because pieces were bigger, they ingested it in the same site, even if touch- ing one another. Although they may rest close to each other, either intra- or interspecifically, they do not seem to be gregarious because resting close together is not the general rule. Because the animals did not exclusively re- treat under shelters either before or after food capture, we do not believe that movement away from the food source is related to pred- ator avoidance. In contrast, this behavior of carrying food did not happen frequently in the troglobitic harvestmen. This could be related to life in an environment lacking predation pressure; the animal does not need to hide while feeding. In addition, it is probably advantageous to im- mediately consume food when it is patchy and scarce, like in a cave. Reaction to light.— The results of the tests, made with two individuals of /. pustulosa and two of P. strinatii, are shown in Table 1. In order to evaluate possible differences in these responses, all values of time of reaction ob- tained for /. pustulosa were compared with those obtained for P. strinatii using an ANO- VA. The same test was used for comparison of the total distance walked. No statistical dif- ference concerning time of reaction was ob- served between the species (F-ratio = 1.670; df — 1; P > 0.05). However, there is a statis- tical difference for the distance walked (F-ra- tio “ 5.514; df = 1; P = 0.03), indicating that P. strinatii walks for significantly greater dis- tances than /. pustulosa. This is an interesting result because it suggests a greater vagility in the former species, which is an expected char- acteristic for a troglobite that lives in an en- HOENEN & GNASPINI— ACTIVITY RHYTHMS AND BEHAVIOR OF HARVESTMEN 163 vironment with a poor food supply. Moreover, considering the apparatus used for tests with only 35 cm length, the great distances walked by P. strinatii implies that the animal would walk back and forth in the chamber, some- times towards the light. This may suggest that this species has less photophobic reaction than the epigean one. However, the fact that the animals react immediately to a light source could imply reaction to a sudden stimulus, and not necessarily a photophobic reaction. We expect that any other stimulus, in addition to light (mechanic, magnetic, electric, etc.), strong enough to be detected by the animal, would promote start of activity. However, we have not tested it yet. DISCUSSION Both of the epigean species (/. pustulosa and /. melanocephala) show strong circadian activity rhythms, as expected facing the ubiq- uity of circadian clocks among surface organ- isms (Biinning 1967; Menna-Barreto 1997). Both species could also be characterized as “diurnal” because of the main expression of activity during the day. Accordingly, species of the whole clade that includes both Ipor~ angaia and Iguapeia are considered “diurnal” because they are mostly seen during the day. There are two peaks for I. pustulosa, one after dawn and one around sunset, the latter being greater. This pattern may be explained by the temporal distribution of their “food,” i.e., in- sects are much more available during these hours of the day, especially near sunset. Although it is generally accepted that trog- lobites have lost temporal organization, at least for circadian frequencies (Lamprecht & Weber 1991), because they live in an environ- ment without the light/dark cycle, some trog- lobitic species appear to maintain a circadian rhythm (e.g., Wilkens et al. 1990; Trajano & Menna-Barreto 1995). This is also the case of P. strinatii, which also shows circadian rhyth- micity. This seems to be a rather intriguing result, and there are several evolutionary traits that may lead to this result. As pointed out by Husson (1971): “Cave fauna are heteroge- neous, differing from one another in the age of their existence, their origins, their reactions towards environment, in the reasons for their presence in caves. On account of this hetero- geneity in the cave fauna it is not reasonable to apply identical laws to all cave animals and to hope to find the same biological rhythms.” Therefore, it is not surprising that the tem- poral aspects of cave species, mainly the trog- lobites, are not as universal as those of the epigean species. The main cause for this dif- ference is probably the diversity of ecological origins of troglobites and their adaptive char- acteristics (Vermeij 1987). Our data point to an endogenous control underlying the expression of the activity of P. strinatii; and, although bursts of locomotion can happen at anytime, there are two main in- tervals within which activity seems to occur. A circadian rhythm in troglobites may be maintained either because it is advantageous, or because it is a relictual feature from an “old” epigean relationship which has not been lost. If a given troglobitic species feeds on material ruled by epigean cycle, it is ex- pected that this species would keep circadian rhythmicity. This would apply, for example, either to a predatory troglobite which feeds on non-troglobitic organisms (which would prob- ably show circadian rhythmicity), or to a de- tritivorous/scavenger/omnivorous troglobite which feeds on material which comes from the epigean environment following circadian rhythms (such as regular floods, regular wind flows). However, this does not seem to be the case for P. strinatii, as it seems to be mainly detritivorous/omnivorous and the input of or- ganic material in the caves where it lives does not seem to be related to any daily event. On the other hand, the circadian rhythmicity of P. strinatii could be considered a relictual feature because Pachylospeleinae (which includes Pa- chylospeleus) belongs to a “nocturnal” clade, in which epigean species completely hide dur- ing the day and leave their shelters only after dusk. Based on our data, it seems that any given stimulus, be it internal (e.g., hunger), or ex- ternal (e.g., turning a light on or handling the animal), causes the start of activity of P. stri- natii, which continues for long time intervals. This happens probably because of the scarcity of food and mates in the environment where the species lives. It is likely that, because food is patchy and scarce and should be exploited promptly, this species feeds wherever it finds food and does not take it away. In addition, the bimodal pattern of activity of P. strinatii may be a result of its life in the cave environ- ment; i.e., the scarcity of food may have led 164 THE JOURNAL OF ARACHNOLOGY the species to look for food in a more frequent and more intensive way, causing the original nocturnal expression of the activity to become duplicate, resulting in bimodality. However, it awaits testing. Circadian periodicities seem to be impor- tant, not only for adjustment to that habitat, but also for the maintenance of internal tem- poral organization (Marques et al 1997), which is responsible for the regulation of dif- ferent and sometimes incompatible physiolog- ical systems. Thus, an additional hypothesis is that P. strinatii could be maintaining a circa- dian rhythmicity for some internal and yet un- known physiological feature. ACKNOWLEDGMENTS We thank Funda9ao Flore stal do Estado de Sao Paulo which allowed collections. This study was supported by grant #96/2494-3 from FAPESP. The junior author has a re- search fellowship from CNPq (Conselho Na- cional de Desenvolvimento Cientifico e Tec- nologico). EH.S. Santos (IBUSP) is thanked for collecting specimens used for study; and T.C. Ramos (IBUSP) helped in the statistical analysis. LITERATURE CITED Barr, T.C. 1968. Cave ecology and the evolution of troglobites. Evol. Biol., 2:35-102. Biinning, E. 1967. The Physiological Clock. The English Universities Press, Springer- Verlag New York, Inc. Culver, D.C. 1982. Cave Life: Evolution and Ecol- ogy. Harvard Univ. Press, Cambridge. Gnaspini, P. 1996. Population ecology of Gonio- soma spelaeum, a cavernicolous harvestman from south-eastern Brazil (Arachnida: Opiliones: Gonyleptidae). J. ZooL, 239(3):417-435. Holsinger, J.R. & D.C. Culver. 1988. The inverte- brate cave fauna of Virginia and a part of eastern Tennessee: Zoogeography and ecology. Brimley- ana, 14:1-162. Howarth, EG. 1983. Ecology of cave arthropods. Ann. Rev. EntomoL, 28:365-389. Husson, R. 1971. Rythmes biologiques et vie cav- ernicole. Bull. Soc. Zool. France, 96(3):301-316. Juberthie, C. 1964. Recherches sur la biologic des Opilions. Ann. Speleol., 19:5-238. Lamprecht, G. & E Weber. 1982. A test for the biological significance of circadian clocks: evo- lutionary regression of the time measuring ability in cavernicolous animals. Pp. 151-178. In En- vironmental Adaptation and Evolution (D. Mos- sakowski & G. Roth, eds.). Gustav Fischer, Stutt- gart. Lamprecht, G. & F. Weber. 1991. Spontaneous lo- comotion behavior in cavernicolous animals: the regression of the endogenous circadian system. Pp. 225-262. In The Natural History of Biospe- leology (A.I. Camacho, ed.). Monografias del Museo Nacional de Ciencias Naturales, Madrid. Marques, M.D., D. Golombek & C. Moreno. 1997. Adapta9ao temporal. Pp. 45-84. In Cronobiolo- gia: Principios e Aplica96es (N. Marques & L. Menna-Barreto, eds.). EDUSP, Sao Paulo. Marques, M.D. & J.M. Waterhouse. 1994. Mask- ing and the evolution of circadian rhythmicity. Chronobiol. Int., 1 1(3): 146-155. Menna-Barreto, L. 1997. O tempo na biologia. Pp. 17-21. In Cronobiologia: Principios e Aplica96es (N. Marques & L. Menna-Barreto, eds.). EDUSP, Sao Paulo. Monteiro, C.A.F. 1973. A dinamica climatica e as chuvas no Estado de Sao Paulo: Estudo geografi- co sob a forma de atlas. Universidade de Sao Paulo, Instituto de Geografia, Sao Paulo. Parzefall, J. 1986. Behavioural ecology of cave- dwelling fishes. Pp. 433-458. In The Behaviour of Teleost Fishes. (TJ. Pitcher, ed.). Croom Helm, London & Sydney. Pinto-da- Rocha, R. 1996. Biological notes on and population size of Pachylospeleus strinatii Sil- havy, 1974 in the Gruta das Areias de Cima, Iporanga, south-eastern Brazil (Arachnida, Opi- liones, Gonyleptidae). Bull. British Arachnol. Soc., 10(5):189-192. Poulson, T.L. & T.C. Jegla. 1969. Circadian rhythms in cave animals. Proc. IV Int. Congr. Speleol., 4-5:193-195. Poulson, T.L. & WB. White. 1969. The cave en- vironment. Science, 165:971-981. Saunders, D.S. 1982. Insect Clocks. Pergamon Press, Oxford. Setzer, J. 1966. Atlas climatico e ecologico do Es- tado de Sao Paulo. Comissao Interestadual da Bacia Parana-Uruguai, Sao Paulo. Silhavy, V. 1974. A new subfamily of Gonylepti- dae from Brazilian caves, Pachylospeleinae sub- fam. n. (Opiliones, Gonyleptomorphi). Rev. Suisse Zool., 81(4):893-898. Trajano, E. & L. Menna-Barreto. 1995. Locomotor activity pattern of Brazilian cave catfishes under constant darkness (Siluriformes, Pimelodidae). Biol. Rhythm Res., 26(l):341-353. Vermeij, G.J. 1987. Evolution and Escalation. Princeton Univ. Press, Princeton. Wilkens, H., J. Parzefall & A. Ribowski. 1990. Population biology and larvae of the anchialine crab Munidopsis polymorpha (Galatheidae) from Lanzarote (Canary Islands). J. Crust, Biol., 10: 667-675. Zar, J.H. 1996. Biostatistical Analysis. Prentice Hall, New Jersey. Pp. 407-445. Manuscript received 29 April 1998, revised 22 Jan- uary 1999. 1999. The Journal of Arachnology 27:165-170 COURTSHIP AND MATING BEHAVIOR OF BRACHYPELMA KLAASI (ARANEAE, THERAPHOSIDAE) Martha Yanez and Arturo Locht: Laboratorio de Acarologia, Facultad de Ciencias, Universidad Nacional Autonoma de Mexico, Coyoacan 04510, D.E Mexico Rogelio Macias-Ordonez: Departamento de Ecologia y Comportamiento Animal, Instituto de Ecologia, A.C., Xalapa, Veracruz 91000, Mexico ABSTRACT. Courtship and mating behavior of Brachypelma klaasi, heretofore unknown, is described on the basis of three courtship and mating sequences, one in captivity and two in the field. Adult males perform courtship movements (pedipalp drumming, leg drumming, push-up and shaking) when they locate a female’s burrow, probably in order to avoid female aggression. After some physical contact, the female raises the prosoma and extends her chelicerae. The male then grasps her chelicerae with his tibial apoph- yses and the female arches her body backwards leaving the epigynum exposed. The male starts boxing the female’s sternum and presumably inserts his pedipalps and inseminates the female. In two cases the female vigorously attacked the male immediately after mating and probably would have killed him had observers not intervened; the other pair separated more slowly and peacefully. Males appear to use chem- ical and/or tactile cues from the female’s silk around the burrow during short-range searching behavior. Males begin courtship behavior by drumming on the silk to signal to the female that he is present. One male of B. klaasi observed in the field laid silk over the female’s silk around the burrow, possibly to prevent subsequent matings by other males. A second male did not detect the burrow after this act. RESUMEN. Se describe el cortejo y apareamiento de Brachypelma klaasi, hasta ahora desconocidos con base en tres secuencias de cortejo y apareamiento, una en cautiverio y dos en campo. Los machos adultos realizan movimientos de cortejo (tamborileo con pedipalpos, tamborileo con patas, lagartijas y temblado) cuando localizan nidos de hembras, probablemente para evitar la agresion de las mismas. Despues de un periodo de contacto fisico la hembra levanta el prosoma y evierte los queliceros. El macho prende los queliceros de la hembra con sus apofisis tibiales y la hembra se arquea hacia atras exponiendo el epigineo. El macho boxea contra el estemon de la hembra y se asume que inserta sus pedipalpos y la insemina. En dos casos la hembra ataco al macho inmediatamente despues del apareamiento y probable- mente lo hubiera matado de no haber inter venido el observador, la tercera pareja se separo mas lenta y pacificamente. Aparentemente los machos utilizan senales quimicas o tactiles de la seda de la hembra alrededor del nido durante la busqueda de corto alcance. Los machos inician el cortejo tamborileando en la seda, probablemente para anunciar su presencia a la hembra. Un macho de B. klaasi observado en el campo deposito seda sobre la de la hembra alrededor del nido, posiblemente para evitar copulas subse- cuentes de otros machos. Un segundo macho no parecio detectar el nido despues de la conducta men- cionada. Tarantulas (Mygalomorphae, Theraphosi- dae) are highly diverse in Mexico, with many species distributed in restricted, endemic areas (Yanez & Locht 1997). The genus Brachy- pelma contains nine species distributed along the Pacific coast of Mexico, eight of which have small, discrete ranges. At least one of these species, Brachypelma klaasi (Schmidt & Krause), and possibly more, are being consid- ered for inclusion as endangered species under CITES. Consequently, studies of reproductive be- havior are important to aid in their reintro- duction to their natural environment. Further- more, few studies have been conducted on the reproductive behavior of tarantulas in general, and little of the literature that has been pub- lished contains detailed behavioral descrip- tions (though see Stradling 1994; Shillington & Verrell 1997). Studies that describe repro- ductive behavior either partially or in detail have been undertaken on the following gen- era: Dugesiella (Baerg 1958; Petrunkevitch 1911), Eurypelma (Baerg 1928), Cyrtopholis (Petrunkevitch 1934), Aphonopelma (Bucherl 1971; Herrero & Valerio 1986; Minch 1979; 165 166 THE JOURNAL OF ARACHNOLOGY Shillington & Verrell 1997), Grammostola (Perez-Miles 1988), Ceropelma (Perez-Miles 1992). Here we provide a detailed description of courting and mating behavior in Brachy- pelma klaasi in the field and in captivity. Species studied. — The female lays a single egg sac containing 400-800 eggs in her bur- row in April-May. The female guards the egg sac for 2-3 months before the spiderlings emerge and disperse. In the juvenile stage, spiders produce temporary burrows until a suitable site is found for a permanent burrow which the spider inhabits for many years. Adult females reach reproductive maturity be- tween 7-9 years and live for up to 30 years. Adult female body size ranges from 50-75 mm and female weight ranges from 19.7-50 g. Males mature earlier (between 6-8 years) and live between 4-6 months. Male weight ranges from 10-45 g (Yanez pers. obs.). The female’s burrow varies in length from 0.15-2 m, depending on the site and the age of the spider. The burrow complex consists of a horizontal tunnel leading from the burrow entrance to a primary chamber where molting usually takes place, and an inclined tunnel that connects the primary chamber to a larger, sec- ondary chamber where the spider rests during the night and where prey is consumed. The female puts a few silk strands at the entrance of the burrow probably so that a male can de- tect that a female is present. Once the male has detected the silk strands, courtship behav- ior may be initiated. Brachypelma klaasi was originally placed in a new genus, Brachypelmides, by Schmidt & Krause (1994). However, a recent compar- ative study of morphology and distribution of species within Brachypelma provides much evidence for including klaasi in the Brachy- pelma group and suggests that Brachypelmi- des be used as a synonym of Brachypelma (Locht et al. 1999). In order to avoid confu- sion, we use the name Brachypelma klaasi (Schmidt & Krause). METHODS Field studies were conducted at La Estacion de Biologia “Chamela”, Jalisco, Mexico, sit- uated on the Pacific coast in a tropical decid- uous forest (19°30'N, 105°03'W, 200 m). Courtship and mating behavior were observed in two pairs in the field on 23 November 1997. The first description of courtship and mating (pair 1) was made at 1030 h (temper- ature 27 °C and 89% relative humidity) when a male (weight = 10 g) was found 4 m away from a female burrow (female weight = 40 g) and placed 10 cm in front of it in order to encourage reproductive behavior. The second description (pair 2) was made at 1700 h (tem- perature 27 °C and 87% relative humidity) us- ing a male (weight — 30 g) caught several kilometers away from the second female bur- row (female weight = 45 g). A third pair (pair 3) was observed in cap- tivity on 11 February 1998 in an environmen- tal chamber at the Facultad de Ciencias of the National Autonomous University of Mexico in Mexico City (27 °C, 60% RH, 12:12 light cycle) using a male (weight = 27 g) caught on 24 November 1998 and a female (weight = 35 g) caught on 2 October 1997 at the field location. Prior to pairing, the male was kept isolated in a 37.5 1 aquarium with soil from the collection site and small pieces of logs. The female was kept in similar conditions in a 50 1 aquarium. The male was placed in the female’s tank to provoke reproductive behav- ior. Encounters were videotaped using a Sony Handycam® Video 8 recorder. Video records were observed at varying speeds in order to accurately describe behavioral patterns. RESULTS Five behavioral patterns were observed in the male. Pedipalp drumming (PD): pedipalps are alternately raised and lowered, about 5 mm off the ground, each cycle lasts between 0. 5-0.8 sec. Pedipalp boxing (PB): the male alternately strikes the female sternum with his pedipalps, each cycle lasts between 0.5-0.8 sec. Boxing cycles could not be quantified given the angle of view, since the bodies ob- structed vision of ventral interactions between males and females. Leg drumming (LD): the two legs of the same pair are rapidly (0. 1 sec) raised and lowered, between 5-20 mm off the substrate; only pairs I (LDI) or II (LDII) are involved in this pattern. Shaking (S): quick (< 1 sec) vibratory movements of the entire body. Push-ups (PU): an instantaneous raise and lowering of the body. No behavioral pat- terns were observed in the female, except for shaking, which was similar to that of the male. Although the three reproductive events var- ied greatly, four stages could be defined: Male YANEZ ET AL.— COURTSHIP AND MATING OF BRACHYPELMA KLAASl 167 approach (MA) begins when the male is placed by the observer near the female’s bur- row and ends when the female is observed in the entrance of the burrow. Female response (FR) ends when physical contact is estab- lished between male and female. Physical contact (PC) ends when the pair separates. Post-mating behavior (PM) comprises any be- havior pattern performed immediately after separation. Pair 1. — Courtship began with the male leg-drumming on the silk surrounding the bur- row and shaking his body. The female came out of the burrow 64 sec after the male started drumming. As she approached, the male con- tinued leg-drumming and drunrmed his pedi- palps once. Female response lasted 58 sec from exiting the burrow to engaging in frontal physical contact. During physical contact, the male drummed the substrate and boxed the fe- male, after which she arched backwards and presumably was inseminated (insemination could not be confirmed in any pair given the angle of view since the bodies obstructed vi- sion of ventral interactions between males and females). The female then started pushing down the male, and he retreated gradually, facing her. Within a second of breaking phys- ical contact (which lasted 84 sec) the female attacked the male vigorously, at which point the observer intervened. A drop of sperm was recovered from the female, and microscopic analysis showed a dense mass of what could be reserve substances mixed with sperm cells. Pair 2. — The male slowly approached the burrow, frequently shaking, and entered the burrow after 261 sec, at which point the fe- male could not be observed. They remained out of sight inside the burrow for 153 sec, after which they came out, engaged in frontal physical contact and remained like this for an- other 196 sec. During this period the female seemed to be highly receptive, arching the body backwards while the male boxed her, and presumably inseminated her. After a slow separation, the male started to groom his che- licerae while the female returned to the bur- row. Then, 128 sec after separation, the male started to spin a thread immediately next to the female burrow for another 60 sec. A sec- ond male placed on the silk surrounding the burrow did not seem to locate the female’s burrow. Pair 3. — Immediately after being placed in- side the female’s aquarium, the male started drumming and pushing-up on silk threads spun by the female. He then approached the female from behind, at which point she turned and engaged in frontal physical contact. Dur- ing physical contact, which lasted for 67 sec, the male boxed the sternum of the female though the number of bouts could not be quantified. The female pushed vigorously down on the male, an apparently aggressive act that made insemination difficult. The fe- male then suddenly attacked the male, at which point the observer intervened. A drop of sperm was observed after separation in the left embolous palp of the male. However, in- semination probably did not occur. Of the three pairs observed. Pairs 1 and 3 displayed similar behavioral patterns com- pared to Pair 2 (Table 1). The female in Pair 2 was more receptive than the other females observed: the male went into her burrow and brought the female out, she arched completely during mating, and afterwards made no ag- gressive attempt before returning to her bur- row. The other two pairs had shorter physical contact before disengaging, females arched less and pushed the male forward. In both cas- es the female attacked the male. The two fe- males observed in the field, females 1 and 2, have remained within their burrows, closed with leaves and silk, to the date of submission of this paper (late April 1998) in a fashion similar to that described for other species when they are developing an egg sac (Baerg 1928). DISCUSSION Observations of short-range male-searching behavior suggest that males might use chem- ical or tactile cues from silk spun around the female’s burrow. Once in contact with the fe- male’s silk, males begin courtship behavior by drumming on the silk to signal to the female that he is present. This behavior has been ob- served in other theraphosids (e.g., Minch 1979; Costa & Perez-Miles 1992; Shillington & Verrell 1997). An interesting observation for one male of B. klaasi observed in the field was the laying down of silk over the female’s silk around the burrow. It appears that this may be a method of interfering with chemical or tactile cues that may be used by subsequent males to locate the female. Although minor differences in courtship be- 168 THE JOURNAL OF ARACHNOLOGY Table 1 . — Duration and behavioral patterns in each stage of the three reproductive interactions observed. Abbreviations are described in the Results section, numbers are frequencies of male behavior patterns, except for F5: female shaking. Pedipalp drumming PD: pedipalps are alternately raised, about 5 mm off the ground, each cycle lasts between 0.5-0. 8 seconds. Pedipalp boxing PB: the male alternately strikes the female sternum with his pedipalps, each cycle lasts between 0.50-0.8 seconds. Leg drumming LD: the two legs of the same pair are rapidly (0.1 second) raised and lowered, between 5-20 mm off the substrate; only pairs I (LDI) or II (LDII) are involved in this pattern. Shaking (S): quick (<1 second) vibratory movements of the entire body. Male approach Female response Physical contact Post-mating Pair 1 64 sec 58 sec 84 sec None (field) LDI:2, S:1 PD:1, LDI:5 PD:1, PB:3, LDII:3, S:1 (female attack) Pair 2 201 sec Inside female 196 sec 252 sec (field) LDI:20, S:10 burrow (not ob- served) PD:1, PB:2, S:6 LDI:2, S:3 Pair 3 153 sec 116 sec 67 sec None (captivity) PD:32, LDI:1, S:4, PU:4 S:20, PD:11, LDI: 9 PB (female attack) havior were observed among the three pairs of B. klaasi, general aspects of the behavior were similar to those known for other thera= phosid species. In particular, the “aggressive” posture adopted by the female by raising her prosoma, followed by the grasping of her che- licerae with his tibial apophyses is character- istic of many theraphosids (Baerg 1958; Minch 1979; Raven 1988; Costa & Perez- Miles 1992; Shillington & Verrell 1997). It is worth noting that courtship behavior and mat- ing occur outside the female’s burrow in B. klaasi as in other tarantula species (e.g., Costa & Perez-Miles 1992), where there is sufficient space for the female to adopt the raised pos- ture. Male courting may serve various functions (Coyle 1971, 1985; Jackson & Pollard 1990; Costa & Perez-Miles 1992; Shillington & Ver- rell 1997). Courting behavior by the male may inhibit female attack (Barth 1993). Agressive female behavior towards males before, during and after mating is well known in many spider groups (Elgar 1992), as males represent a po- tential food resource as well as a mating op- portunity for females. In the Theraphosidae, cannibalism after mating has been document- ed for several species (e.g., Biicherl 1951; Shillington & Verrell 1997). However, many studies have observed no sexual cannibalism (e.g., Costa & Perez-Miles 1992; Stradling 1994), and even in species where cannibalism has been recorded, such events are often rare (Shillington & Verrell 1997). Jackson & Pol- lard (1990) suggested that sexual cannibalism is generally rare in theraphosids, and that male grasping behavior may be due to factors other than avoidance of female attack. Other poten- tial factors include communication related to mate choice (Coyle 1971, 1985), or simply as a way of maneuvering the female for success- ful sperm transfer (Coyle 1971; Jackson & Pollard 1990). However, low rates of sexual cannibalism in theraphosids may be a result of male grasping behavior. Without this be- havior, cannibalism rates might be signifi- cantly higher. While cannibalism is rarely ob- served in theraphosids, aggresssive female behavior directed towards the male was ob- served in two of the three pairs of B. klaasi; though whether the attacks would have re- sulted in cannibalism is unknown as, in both cases, the authors intervened before the male could be injured. Brachypelma klaasi is en- demic and has small, isolated populations with limited distributions in parts of the Pacific coast (Yanez & Locht 1998). The interactions described in this paper were staged given the rarity of the species and the difficulty of ob- serving courtship and mating behavior under natural conditions. Compounding the rarity of B. klaasi, the high value placed on tarantulas in the pet trade has led to high rates of col- lection and trafficking of species from Mexi- YANEZ ET AL.— COURTSHIP AND MATING OF BRACHYPELMA KLAASI 169 CO, although the extent of trafficking in B. klaasi is unknown. Consequently CITES is considering giving B. klaasi (and other species in the Brachypelma group) endangered status. Captive breeding and reintroduction of B. klaasi is an important means of sustaining nat- ural populations. The studies presented here suggest that mating B. klaasi in captivity is not difficult and the production of eggs in the laboratory should be successful under a cap- tive-breeding program. Furthermore, if B. klaasi, one of the rarest Mexican tarantulas, can be mated successfully in captivity, and studies of other species have produced similar results (e.g., Shillington & Verrell 1997) re- introductions of captive-bred individuals may be a successful technique for increasing pop- ulation levels of other tarantula species. ACKNOWLEDGMENTS We are grateful to Anita Hoffmann, Juan Morales-Malacara and Graham Floater and two anonymous reviewers for useful sugges- tions in early versions of this manuscript. James Berry made useful style corrections. La Estacion de Biologia “Chamela”, Instituto de Biologia, Universidad Nacional Autonoma de Mexico, provided support during the field work. The Controlled Environment Chambers laboratory in The Facultad de Ciencias, Univ- ersidad Nacional Autonoma de Mexico, pro- vided captivity facilities. Financial assistance was provided by Direccion General de Asun- tos del Personal Academico, UNAM, Grant No. IN217397. LITERATURE CITED Baerg, W.J. 1928. The life cycle and mating habits of the male tarantula. Quart. Rev. Biol., 3:109- 116. Baerg, W.J. 1958. The Tarantula. Univ. of Kansas Press, Lawrence, Kansas. Barth, EG. 1993. Sensory guidance in spider pre- copulatory behaviour- a mini review. Comp. Biochem. Physiol. 104'*:7 17-733. Biicherl, W. 1951. Estudios sobre a biologia e a sistematica do genero Grammostola Simon, 1892. Monog. Inst. Butantan, 1:1-126. Biicherl, W. 1971. Los Aracnidos. Pp. 341-344. In Zoologia Hispanoamericana de los Invertebrados (L. Cendrero, ed). Brazil. Costa, F. & F. Perez-Miles. 1992. Notes on mating and reproductive success of Ceropelma longis- ternalis (Araneae, Theraphosidae) in captivity. J. ArachnoL, 20:129-133. Coyle, EA. 1971. Systematics and natural history of the mygalomorph spider genus Antrodiaetus and related genera (Araneae: Antrodiaetidae). Bull. Mus. Comp. ZooL, 141:269-402. Coyle, EA. 1985. Observations on the mating be- haviour of the tiny mygalomorph spider, Microh- extura montivaga Crosby & Bishop (Araneae: Dipluridae) Bull. British ArachnoL Soc., 6:320- 330. Elgar, M.A. 1992. Sexual cannibalism in spiders and other invertebrates. Pp. 128-153. In Canni- balism: Ecology and Evolution Among Diverse Taxa (M.A. Elgar & B.J. Crespi, eds). Oxford Univ, Press, Oxford. Herrero, M. & C. Valerio. 1986. Analisis de la ac- tividad diaria de Aphonopelma seemanni (Ara- neae, Theraphosidae) en Costa Rica. J. Arach- noL, 14:79-82. Jackson, R.R. & S.D. Pollard. 1990. Intraspecific interactions and the function of courtship in my- galomorph spiders: a study of Porrhothele anti- podiana (Araneae: Hexathelidae) and a literature review. New Zealand J. ZooL, 17:499-526. Locht, A., Yanez, M. & Vazquez, I. (in press). Notes on the distribution and natural history of the genera Brachypelma and Brae hyp elmides (Theraphosidae, Theraphosinae) with comments on morphology and generic affinities. J. Arach- noL Minch, E.W. 1979. Reproductive behavior of the tarantula Aphonopelma chalcodes Chamberlin (Araneae: Theraphosidae). Bull. British Arach- noL Soc., 4(9):416-420. Perez-Miles, E 1988. Variacion relativa de carac- teres somaticos y genitales en Grammostola mol- licoma (Araneae, Theraphosidae). J. ArachnoL, 17:263-274. Perez-Miles, F. & F. Costa. 1992. Interacciones in- tra e intersexuales en Grammostola mollicoma (Araneae: Theraphosidae) en condiciones exper- imentales. Bol. Soc. ZooL Uruguay., 7:71-72. Petrunkevitch, A. 1911. Sense of sight, courtship, and mating in Dugesiella hentzi (Girard), a ther- aphosid spider from Texas. ZooL Jb. (Syst.), 31: 355-376. Petrunkevitch, A. 1934. New observations on molting and mating in tarantulas. New York En- tomol. Soc., 42:289-296. Raven, R. 1988. Preliminary observations on the mating behaviour of the Australian mygalo- morph spider Australothele jamiesoni (Dipluri- dae, Araneae, Arachnida). Mem. Queensland Mus., 25:471-474. Shillington, C. & P. Verrell. 1997. Sexual strategies of a North American “tarantula” (Araneae: Theraphosidae). Ethology, 103:588-598. Stradling, D.J. 1994. Distribution and behavioral ecology of an arboreal “tarantula” spider in Trinidad. Biotropica, 26(1): 84-97. Yanez, M. & A. Locht. 1997. El Infraorden My- 170 THE JOURNAL OF ARACHNOLOGY galomorphae en Mexico: Una recopilacion a ni- vel mundial. Tesis Conjunta. Facultad de Cien- cias, UNAM. Yanez, M. & A. Locht. 1998. Ensayos de apaream- iento inducido en Brachypelma klaasi (Schmidth & Krause, 1994) (Araneae: Theraphosidae). So- ciedad Mexicana de Entomologia. Memorias del XXXIII Congreso Nacional de Entomologia. Pp. 37-41. Manuscript received 1 May 1998, revised 8 January 1999. 1999. The Journal of Arachnology 27:171-175 LOCATION OF SUCCESSFUL STRIKES ON PREY BY JUVENILE CRAB SPIDERS MISUMENA VATIA (ARANEAE, THOMISIDAE) Douglass H. Morse: Department of Ecology and Evolutionary Biology, Box G-W, Brown University, Providence, Rhode Island 02912 USA ABSTRACT. Second-instar crab spiderlings Misumena vatia (ca. 0.6 mg) that had never previously fed made killing attacks on pomace flies Drosophila melanogaster (ca. 1.0 mg) in direct proportion to the surface areas of the flies’ body parts: abdomen, 50%; thorax, 29%; head, 20%. They retained this pattern over their next six encounters with these flies. They also attacked the different surfaces of these body parts (front, side, above, below, behind) with a frequency predicted by the respective areas of these surfaces. All of the spiderlings tested more than once successfully attacked prey on more than one body part. Fifth and sixth-instar Misumena (ca. 7-15 mg) attacked small (4 mg) syrphid flies Toxomerus mar- ginatus more frequently on the head than the second instars attacked Drosophila heads. This difference may result from subsequent experience, greater activity of the syrphid flies than the Drosophila, or mat- uration of the spiders. A wide variety of animals employ a sit-and- wait predatory strategy, ranging from spiders and insects to lions (Curio 1976; Morse 1980). Sit- and- wait predators depend primarily on prey coming to them and consequently will encounter many of these prey either head-on or tangentially, even though some of these predators may orient to their prey and even pursue them for short distances. These pred- ators often increase their proficiency with ex- perience (Bailey 1985; Cloarec 1991), which may result from the development and refine- ment of a particular repertoire (Papaj & Pro- kopy 1989), and may include new prey spe- cies as the predator grows, or as the season changes (e.g., Erickson & Morse 1997). The body part (head, thorax, abdomen) of the prey struck by the predator may form an important part of a developing repertoire exhibited by sit-and-wait predators. Little information exists on the initial part of the body struck successfully by spiders, in- cluding classic sit-and-wait predators (Foelix 1996a, 1996b), despite the oft-cited “neck- bites” found in general sources (e.g., Bristowe 1958; Main 1976). To my knowledge, infor- mation on strike sites does not exist for naive spiderlings of any species making their first kill. Spiderlings are excellent subjects for such an investigation, because they can be easily obtained in large numbers and can be easily run in enclosures using readily available prey. In this paper I report the body parts of prey, wild-type pomace flies Drosophila melano- gaster Meigen, successfully struck by just- emerged, second-instar crab spiders Misu- mena vatia (Clerck 1757) (Thomisidae) making their first captures, as well as the body parts successfully attacked in several subse- quent captures by these spiderlings. Drosoph- ila approximate the size and activity patterns of small Diptera encountered by the spider- lings in the field and often constituting their first captures (Morse 1993). I then compare these results with those of fifth and sixth-in- star Misumena attacking small syrphid flies Toxomerus marginatus (Say), an important prey item of older Misumena in the field. These results provide important insights into the development of prey-capture behavior in Misumena and also provide the basis for fu- ture comparisons with other species. METHODS I obtained all spiders and syrphid flies from old fields and roadsides in South Bristol, Lin- coln County, Maine in August of 1995 and 1996, and Drosophila came from wild-type laboratory stocks. All second-instar spider- lings used in this study had emerged from their egg sacs within the preceding two days at the time of their first observation. The egg sacs themselves had been collected from the field shortly before emergence. Laying dates 171 172 THE JOURNAL OF ARACHNOLOGY of these clutches were known, so approximate emergence dates could be calculated. Middle- instar spiders and syrphid flies were collected from goldenrod {Solidago spp.) flowers, which both the spiders and the flies frequented during August. All spiderlings were weighed before their first experimental run, and their similar masses (0.4-0. 7 mg: Morse 1993) en- sured that they had not cannibalized their sibs [a rare event occurring in less than 10% of the broods (DHM unpubl. data)], and, therefore, had not previously fed. The spiderlings were tested seven times, at three-day intervals. Since a few were lost during handling, a few died, and most refused to feed on one or more occasions, I obtained the maximum seven data points for only four of the 32 individuals run in this protocol. I placed approximately 10 Drosophila (ca. 1.0 mg, 3 mm body length) in a petri dish (6 cm diameter) and then added a second-instar spiderling to the dish. Although a high-density setting, this density is frequently approximat- ed when just-emerged, second-instar Misu- mena recruit onto goldenrod inflorescences that contained large numbers of small dance flies (Empididae) (Morse 1993). The Dro- sophila were lightly chilled to immobilize them sufficiently for convenient handling, and then allowed to recover before adding the spi- derling. I recorded the body part where the spiders first successfully struck their prey (head, thorax, abdomen) and the surface of the body part they struck (anterior, lateral, dorsal, ventral, and posterior). As soon as a spiderling captured either a Drosophila or syrphid fly, I viewed it under a dissecting microscope to verify the site of the successful attack. Since these individuals were carefully observed for up to 30 min, I missed few predation events. The close observations ensured that none of the spiderlings shifted their positions on their prey before being recorded. Spiders often re- position their prey subsequent to capture (Foe- lix 1996a), necessitating this close attention. These spiderlings required a few minutes to shift from the original kill site (pers. obs.), although attention to obtaining original kill sites made it impossible to record exact shift times as well. 1 calculated expected frequencies of attacks on the head, thorax, and abdomen as the rel- ative surface area of each of these body parts; excluding the posterior surface of the head, anterior and posterior surfaces of the thorax and anterior surface of the abdomen; surfaces largely occluded from strikes by surrounding structures. I calculated the areas from mea- surements of length and width of the head and thorax, estimating them to be cylinders. The anterior surface of the head was calculated as the area of a circle. The anterior part of the abdomen, back to the point at which it ta- pered, was also treated as a cylinder, and the remaining posterior part as a cone. This cal- culation assumed that the spaces separating head and thorax, and the thorax and abdomen, were too narrow to permit a successful strike and deleted these surfaces from the areas cal- culated. Since only four of 148 successful at- tacks struck these sites between the body parts, the criteria seem appropriate. I also gathered similar data on the prey cap- ture of small (ca. 4.0 mg, 5 mm: Morse 1979, 1998) syrphid flies Toxomerus marginatus by older, wild-caught, juvenile female Misumena weighing 6.9-15.6 mg (probably fifth and sixth instars). Relative proportions of area on the three body parts of Toxomerus were cal- culated as for Drosophila. These observations were made in 7-dram vials (5 cm long, 3 cm diameter), which also permitted me to observe initial capture sites. However, I did not record the part of the head, thorax, or abdomen where the syrphids were struck by the spiders. RESULTS Second instars attacking Drosophila. — In their first run, naive second instars made more of their first killing attacks to the abdomen than to the thorax or head of Drosophila, and more killing attacks to the thorax than to the head (abdomen > thorax > head) (Table 1). This distribution of killing attacks to the dif- ferent body parts did not differ significantly from the number predicted as a consequence of the different surface areas of these body parts (Table 1) (G = 1.09, df = 2, P > 0.5 in a G-test), since the surface area of the abdo- men considerably exceeded that of the thorax, which in turn exceeded that of the head (Table 1). Likewise, the sites of attack in the original trial and in the mean of the combined subse- quent trials did not differ (Table 1) (G = 0.36, df ^ 2, P > 0.8 in a G-test). Neither did the original and last trials (Table 1) differ in a G- test (G = 0.90, df = 2, P > 0.5). In fact, comparisons of only two pairs of trials (2 and MORSE— CRAB SPIDER ATTACKS ON PREY 173 Table 1 . — Successful strikes (kills) of Misumena vatia on body parts of prey and percentage of total surface area of each body part. Predicted number of strikes (in parentheses), based on percentage of total surface area. Head Thorax Abdomen Drosophila strikes First run 5 (4.4) 7 (6.9) 12(12.7) Second run 3 (5.4) 10(8.3) 16(15.3) Third run 4 (5.2) 14 (8.0) 10(14.8) Fourth run 7 (3.9) 5 (6.0) 9(11.1) Fifth run 5 (3.2) 6 (4.8) 6 (9.0) Sixth run 6 (2.0) 2 (3.2) 3 (5.8) Seventh run 6 (3.3) 5 (5.2) 7 (9.5) % surface area 18.6 28.6 52.8 Toxomerus strikes First run 15 (5.8) 12 (8.9) 4(16.3) % surface area 17.5 28.2 54.3 6, 3 and 6) differed at P < 0.05 (2 vs. 6: G = 8.21, df = 2, P < 0.02 in G-test; 3 vs. 6: G = 6.84, = 2, P < 0.05 in same test), and their validity is highly suspect, because of the small sample sizes in two cells of Trial 6. Fur- ther, neither comparison is significant when a sequential Bonferroni adjustment (Rice 1989) is applied to accommodate for the multiple comparisons carried out. Successful strikes in runs 2-7 continued to follow the order abdo- men > thorax > head in most instances, con- sistent with the different surface areas of the three body parts. Thus, no significant shift in sites occurred over the period during which these spiderlings killed their first several prey. None of the spiderlings specialized strongly on a particular body part; in fact, none of the 29 individuals tested more than once confined their kills to a single body part (P < 0.001 in a binomial test). The pattern of attack thus showed little sign of specialization, at the in- dividual or population level. As no clear shifts in killing patterns emerged in the analysis of consecutive kills, I pooled the data from the different runs in or- der to establish how the spiderlings directed their killing attacks to the different surfaces of the body parts (Table 2). With 15 total surfac- es recognized (Table 2), the sample of kills from any single run or pair of runs was not large enough to test statistically. The results can, however, establish where a predator most frequently attacks a prey species, an aspect that may serve to drive selection of prey-cap- ture techniques of the predator, and corre- sponding selection on the prey species. The spiderlings showed little tendency to capture prey by striking between the body parts, with only four such successful strikes, these being directed to the rear of the head (3) and the rear of the thorax (1). Deleting the areas of these four surfaces largely covered by adjacent body parts, successful attacks were carried out to the 1 1 remaining surfaces of the three body parts at rates that did not differ from the predicted (G = 10.81, df = 10, P > 0.3 in a one-sample G-test). Thus, the areas of the various surfaces of the different body parts also accurately predicted the rates at which these sites were successfully struck. Later instars attacking syrphid flies. — Middle-instar spiders successfully struck Tox- omerus on the head and thorax far more often than predicted by chance, based on the re- spective surface areas of the body parts (Table 1) (G = 31.13, df = 2, P < 0.001 in a G- test). This tendency differed significantly from that of the second instars capturing their first prey item (G = 12.42, df = 2, P < 0.01 in a G-test). I did not record the surfaces of the body parts struck that resulted in kills by these middle-instar spiders. DISCUSSION These spiders must be able to capture a broad range of prey over their lifetimes, both Table 2. — Strikes of second-instar Misumena vatia on different surfaces of Drosophila body parts. Predicted number of strikes in parentheses, based on percentage of total surface area. Body part Surface area Front Side Above Below Behind Head 12 (8.4) 8 (9.2) 4 (4.6) 9 (4.6) 3 — Thorax 0 — 24 (20.5) 9(10.3) 15 (10.3) 1 — Abdomen 0 — 19 (27.8) 9(13.9) 22(13.9) 13 (20.5) 174 THE JOURNAL OF ARACHNOLOGY as a consequence of their change in size and with the progression of the season. Opportu- nities will also differ with the habitat, and these sit=and“Wait predators will also experi- ence changes associated with the flower hunt- ing sites experienced here. It is thus not sur- prising that the spiderlings do not exhibit a highly programmed repertoire upon initial ex- perience with prey. Species with such varied demands often learn to perfect foraging rep- ertoires appropriate to their context; where pa- rental care is involved, this procedure often involves extensive information passed on from parent to offspring (e.g., Altmann 1998); where not, extensive trial- and-error may be re- quired (e.g., Heinrich 1976). It is of interest that the spiderlings did not exhibit any clear pattern of change in surfaces struck over seven runs. Clearly, they caught these prey with little difficulty, mostly captur- ing a Drosophila in a few seconds to several minutes (DHM pers. obs.), and thus they probably never accumulated information that favored shifts in prey-capture patterns. These spiderlings’ high success rates differ markedly from that of second instars attacking Toxo- merus flies (Erickson & Morse 1997), or that of adults on bumble bees Bombus spp. (Fritz & Morse 1985), both far more formidable prey than Drosophila. Although the condi- tions experienced in this experiment clearly differ from many situations experienced by novice foragers, such conditions are not un- usual for naive Misumena spiderlings, as they typically recruit onto goldenrod inflorescenc- es, which have wide, platform-like surfaces and, often, dozens of dance flies of 0.7-0. 8 mg mass within a single small group of inflo- rescences. These flies are slow-moving and show little sign of responding evasively to the spiderlings (Morse 1993), and spiderlings probably experience little selection to position their site of attack more precisely on these small prey. The tendency of the spiderlings to approx- imate predictions of strike sites based on sur- face areas of the prey, and the stronger ori- entation to the anterior part of the body in the larger spiders, suggest that the spiders modify their patterns somewhat with experience, al- though maturation could also account for the change. The failure of spiderlings to confine their activities to one body part or another may simply be a consequence of the substan- tial proportions of prey taking trajectories that place them both face-on and lateral to the spi- ders, as occurs routinely when foraging on flowers in the field (Morse 1986). The older spiders probably also encounter higher pro- portions of prey moving directly toward them, as frequently occurs with active prey (Curio 1976), which would further enhance the prob- ability of striking the anterior parts of a prey item. Although the spiders attacked these flies in laboratory containers rather than on flow- ers, the frequency in the field seems unlikely to change greatly because of the spiders’ pri- mary foraging strategy of waiting for such in- sects to approach them. ACKNOWLEDGMENTS K.S. Erickson supplied some of the syrphid fly data. I thank M. Hodge, B. Opell, and an anonymous reviewer for their comments on the manuscript. I also thank K. Eckelbarger, TE. Miller, and other staff members of the Ira C. Darling Marine Center of the University of Maine for facilitating the field work on their premises. In particular, TE. Miller helped in many ways to expedite research at the site. Partially supported by the National Science Foundation IBN93- 17652. LITERATURE CITED Altmann, S.A. 1998. Foraging for survival. Univ. of Chicago Press, Chicago. Bailey, P.C.E. 1985. “A prey in the hand”, multi- prey capture behaviour in a sit-and-wait predator Ranatra atra (Heteroptera: Nepidae), the water stick insect. J. EthoL, 3:105-112. Bristowe, W.S. 1958. The world of spiders. Col- lins, London. Cloarec, A. 1991. Handling time and multi-prey capture by a water bug. Anim. Behav., 42:607- 613. Curio, E. 1976. The ethology of predation. Spring- er Verlag, New York. Erickson, K.S. & D.H. Morse. 1997. Predator size and the suitability of a common prey. Oecologia, 109:608-614. Foelix, R.F. 1996a. How do crab spiders (Thom- isidae) bite their prey? Rev. Suisse ZooL, hors serie: 203-2 10. Foelix, R.F. 1996b. Biology of spiders, 2nd ed. Ox- ford Univ. Press, New York. 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. Heinrich, B. 1976, The foraging specializations of MORSE— CRAB SPIDER ATTACKS ON PREY individual bumblebees. Ecol. Monogr., 46:105- 128. Main, B.Y. 1976. Spiders. Collins, Sydney. Morse, D.H. 1979. Prey capture by the crab spider Misumena calycina (Araneae: Thomisidae). Oec- ologia, 39:309-319. Morse, D.H. 1980. Behavioral mechanisms in ecol- ogy. Harvard Univ. Press, Cambridge. Morse, D.H. 1986. Predatory risk to insects for- aging at flowers. Oikos, 46:223-228. Morse, D.H. 1993. Some determinants of dispersal by crab spiderlings. Ecology, 74:427-432. 175 Morse, D.H. 1998. The effect of wounds on des- iccation of prey: implications for a predator with extra-oral digestion. Oecologia, 115:184-187. Papaj, D.R. & R.J. Prokopy. 1989. Ecological and evolutionary aspects of learning in phytophagous insects. Ann. Rev. EntomoL, 34:315-350. Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution, 43:223-225. Manuscript received 1 May 1998, revised 11 No- vember 1998. 1999. The Journal of Arachnology 27:176-182 SAMPLING METHOD AND TIME DETERMINES COMPOSITION OF SPIDER COLLECTIONS Jan Green: Department of Zoology and Entomology, The University of Queensland, Brisbane, Queensland, Australia 4072 ABSTRACT. Sampling methods and times can misrepresent components of spider assemblages found in tree crops. I collected 2561 spiders, including 20 families, 77 genera and 140 species, from inland and coastal south-east Queensland citrus orchards maintained under Integrated Pest Management programs. Spider assemblages, collected diumally and noctumally using vacuum and pit-trap sampling methods over four seasonal periods (spring, summer, autumn and winter), were compared using Simpson and Shannon- Wiener diversity indices and Morisita-Horn similarity index. Significantly different spider assemblages were collected by the two sampling methods in all orchards and seasons. Nocturnal and diurnal sample data differed for spider abundance (similarity) and diversity for several orchards. These results indicate the need to conduct nocturnal and diurnal sampling using a combination of sampling methods to reduce misinterpretation of the composition of spider assemblages. Such misinterpretations may underestimate the predatory importance of spiders in agricultural ecosystems. Spiders are gaining favor in ecological studies as indicators of environmental quality (Clausen 1986; Maelfert et al. 1990; Churchill 1997), and as biological control agents in ag- ricultural ecosystems (Riechert & Lockley 1984; Young & Lockley 1985; Nyffeler & Benz 1987; Bishop & Riechert 1990). Knowl- edge of field populations of spiders, and the sampling techniques for gaining that knowl- edge, are therefore of great importance. Different collecting methods can misrepre- sent certain components of spider assemblages (Merrett & Snazell 1983; Churchill 1993). For instance, pitfall traps, which are commonly used for spider collecting, are effective for ground-dwelling spiders but underestimate the diversity and abundance of the foliage-dwell- ing fauna. Many surveys of spiders in agri- cultural ecosystems employ pit-traps alone (Alderweireldt & Desender 1990; Vangsgaard et al. 1990). Canopy fogging (Basset 1990; Russell-Smith & Stork 1995) underestimates web-building and web-producing spiders which can remain attached to their webs or suspended in foliage after the insecticide treat- ment. Branch beating can under-represent web-building spiders. For instance, Neoscona oaxacensis (Keyserling) did not constitute a high percentage of spiders collected by branch-beating in vineyards, although the webs and spiders were numerous and highly visible between rows of vines (Costello & Daane 1995). Fewer spiders were collected by vacuum sampling than pitfall trapping in heathland (Merrett 1983). Fogging is not an option in orchards under Integrated Pest Management as imported biological control insects may be unnecessarily destroyed, and branch-beating at night is not successful as escaping spiders are not easily seen in poor light (Green un- publ. data). Consequently, vacuum suction, in conjunction with pit-fall trapping, is chosen for this study. Merrett & Snazell (1983) rec- ommend a combination of vacuum and pit- trapping for sampling spiders in heathland and De Barro (1991) advocates the use of a two stroke gasoline-driven blower vacuum for aphids on wheat. The temporal dimension to spider foraging behavior must also be considered. Diurnal and nocturnal sampling appear necessary to effec- tively sample all of the spider fauna as many spiders are nocturnal (Coddington et al. 1990). Most studies which use several methods, such as hand collecting, sweep nets or vacuum samples, are usually conducted during day- light hours (Young & Lockley 1990; Mason 1992; Breene et al. 1993a, b). Some studies include nocturnal samples or observations of spider assemblages (Coddington et al. 1990; Coddington et al. 1996; Dobyns 1997). Sampling methods should be kept to a mini- 176 GREEN— SAMPLING METHOD AND TIME DETERMINE SPIDER COLLECTIONS 177 mum to reduce complexity in the sampling protocol, and methods chosen should mini- mize species overlap by collecting different spider assemblages (Coddington et ah 1990). Here I demonstrate the importance of a com- bination of two sampling methods, in this case vacuum and pit-trap, and sampling times, di- urnal and nocturnal, for determining the nu- merically abundant spider species in citrus or- chards. STUDY AREA AND METHODS Study locations.— Spiders were collected from two inland (Mundubbera 25°35'S, 151°18'E, 300 km from the coast) and two coastal (Coochin Creek 26°54'S, 53°05'E) cit- rus orchards in south-east Queensland, Aus- tralia; the orchards are under an Integrated Pest Management (IPM) program which has been developed around biological control and the limited use of pesticides. Consequently, higher numbers of native natural enemies like spiders are conserved than in chemically man- aged orchards. Sampling was conducted di- urnally and nocturnally over four seasons from Spring 1993 to Winter 1994. Sampling took place over the middle month of each sea- son. Replicates in each orchard were sampled once per season. Sampling occurred under suitable weather conditions for spider collec- tion, temperatures between 5-38 °C and no rain. Ellendale mandarins and Navel oranges from inland orchards, and Valencia and Navel oranges from the coastal orchards were sam- pled. Sampling methods.^ — Four groups of six trees each were randomly selected in each or- chard in each season, giving a total of 24 trees sampled per orchard per season. Within each group of six trees, three trees were sampled diumally and three were sampled nocturnally. Vacuum-sampling was carried out for 15 min- utes per tree between 0630-1030 h for diurnal samples and between 1800-2200 h for noc- turnal samples. Nocturnal sampling was ef- fected by wearing a headlamp. Foliage, trunk and branches were sampled on each tree to the height of the author’s reach plus the length of the sampler (about 2.5 m). A Little Wonder Power Blower® (Model 9444E, Korditz, Japan) powered by a two- stroke gasoline motor was used for suction sampling. The only modification, a net sleeve, was placed inside the muzzle of the vacuum to facilitate spider collection. After suction sampling, spiders were placed into labelled killing jars containing ethyl acetate before be- ing transferred, in the laboratory, to labelled glass vials containing 70% EtOH. Pit-traps consisted of plastic food contain- ers (115 mm diameter, 80 mm deep) which were three-quarters filled with detergent and water (1:40). The traps were placed in the ground, so that the soil was flush with the rim, one trap under each of three trees (about 7.3 m apart) in each block of six trees. The con- tainers were left open for one week. Speci- mens collected from pit-traps were placed into labelled glass vials containing 70% alcohol. Data analysis.— Diversity analysis, using 10,000 randomizations, determined the signif- icance of observed differences in community structure between two sampling methods and two sampling times based on species abun- dance distributions (Solow 1993). Two diver- sity indices used are the Shannon- Wiener in- dex, which is sensitive to changes in the abundance of rare species in a community, and the Simpson index, which is sensitive to changes in the most abundant species in a community (Solow 1993). Shannon- Wiener index, which increases with the number of species in the community, is an ordinal scale. An index of 2 does not suggest that commu- nity is twice as diverse as a community with an index of 1 . The values for Simpson’s index vary between 0 (for a sample with high di- versity) and 1 (for a sample dominated by a few species) (Solow 1993). Shannon- Wiener index is defined as: H = -S.logjr^, where: p, = the observed relative abundance of a particular species (Solow 1993). Simpson index is defined as: F) = [N(N - 1)] where: — the number of individuals of spe- cies /, and N = Xw, (Solow 1993). Two-tailed tests were used to test the hypotheses that the two sampling methods (pit-trapping and vac- uum sampling) and sampling at different times of the day (diurnal and nocturnal) col- lect different abundance and composition of spider assemblages. The Morisita-Horn index (Wolda 1981; Krebs 1989) was used to calculate similarity 178 THE JOURNAL OF ARACHNOLOGY Table 1. — Shannon- Wiener (H) and Simpson (D) diversity indices and Morisita-Hom similarity indices (MH) for vacuum and pit- trap samples in three IPM orchards during four seasons, n = total number of genera collected by each sampling method; Diff. = difference between diversity indices for vacuum and pit-trap sampling methods. Season Orchard n H Diff. D Diff. MH Pit Vac Pit Vac Pit Vac Summer Coastal 1 14 0.00 2.16 -2.16 1.00 0.18 0.82 0.000 Inland 1 4 25 0.93 2.17 -1.24 0.51 0.18 0.33 0.004 Inland 2 6 31 0.57 2.38 -1.81 0.77 0.20 0.57 0.002 Autumn Coastal 4 26 0.63 2.36 -1.73 0.69 0.18 0.51 0.003 Inland 1 6 17 1.26 2.58 -1.32 0.41 0.09 0.32 0.010 Inland 2 6 28 0.38 2.62 -2.24 0.86 0.11 0.75 0.000 Winter Coastal 2 22 0.44 2.81 -2.37 0.73 0.07 0.65 0.000 Inland 1 3 18 0.60 2.00 -1.40 0.68 0.25 0.43 0.003 Inland 2 4 20 0.56 2.52 -1.97 0.75 0.11 0.64 0.013 Spring Coastal 5 25 1.23 2.33 -1.11 0.35 0.15 0.20 0.000 Inland 1 7 24 1.18 2.65 -1.47 0.41 0.09 0.32 0.008 Inland 2 6 43 1.09 2.80 -1.71 0.47 0.09 0.38 0.106 (or non- similarity) between spider populations from two sampling methods and two sampling times. The index is independent of sample size and diversity (Wolda 1981) and is an ap- propriate measure to compare community structure in day and night sampling using vac- uum and pit- trap sampling methods. The Mor- isita-Horn index was calculated from: MH (X, + \2)N,N^ where MH = Morisita-Horn index of similarity between sampling methods j and k, n,, = the number of individuals of species / in sample 7, Nj = the total number of individuals of all species in sample 7, and Xj = ~W Diversity and similarity indices were achieved using spider abundance at the genus level to minimize false results from rare spe- cies. This study is part of a larger project which investigated the potential of spiders as natural pest control agents in citrus orchards in south-east Queensland. RESULTS I collected a total of 2561 spiders, including 20 families, 77 genera and 140 species. All spiders, including immatures (29%), were identified to genus or species with the help of Dr. Robert Raven, Queensland Museum. Effect of sampling method. — For each or- chard in each season, diversity indices for genera differed significantly between vacuum sampling and pit- trapping using either Shan- non-Wiener (H) or Simpson (D) analyses (P < 0.0001, Table 1). Differences were also ap- parent at family and species level (Table 1). Similarity values differed markedly for each orchard in each season; all values were below 0.01 (Table 2). Generic composition was markedly differ- ent for each sampling method. Only 13 spe- cies (10%), 13 genera (18%) and 8 families (40%) were common to both methods. Com- bined data for all orchards in all seasons showed that at all taxonomic levels, more taxa were collected by vacuum sampling than by pit-traps (Fig. lA). No lycosids or zodariids, and few gnaphosids and corinnids, were col- lected by vacuum sampling. Spiders from these families are ground-dwelling spiders. Some salticids were collected in pit-traps but the vast majority were found in the upper stratification of the orchard. Different spider communities were seen in the orchard for the two main stratification layers — trees (81-97% of total taxa) and ground (29-57% of total taxa). Effect of sampling time. — Pit-traps were left open for one week continuously; conse- quently these data are not included in the di- urnal/nocturnal analysis. Generic richness was significantly different between diurnal and GREEN— SAMPLING METHOD AND TIME DETERMINE SPIDER COLLECTIONS 179 Table 2. — Shannon-Wiener (H) and Simpson (D) diversity indices, their P-values, and Morisita-Horn similarity indices (MH) for diurnal and nocturnal samples in three IPM orchards during four seasons, n = total number of genera collected in each time period; Diff = difference between diversity indices for diurnal and nocturnal sampling periods. Season Orchard n H Diff. P D Diff. P MH AM PM AM PM AM PM Summer Coastal 6 10 1.42 2.02 -0.6 <0.0001 0.32 0.17 0.15 <0.0001 0.645 Inland 1 17 19 1.95 2.46 -0.51 <0.0001 0.21 0.13 0.08 <0.0001 0.748 Inland 2 23 22 2.26 2.24 -0.02 <0.7 0.19 0.21 0.02 <0.0001 0.963 Autumn Coastal 18 17 2.15 2.08 0.07 <0.3 0.18 0.22 -0.03 <0.0001 0.869 Inland 1 15 10 2.12 2.46 -0.34 <0.001 0.13 0.09 0.04 <0.0004 0.510 Inland 2 21 21 2.42 2.36 0.06 <0.3 0.14 0.14 -0.004 <0.5 0.821 Winter Coastal 15 15 2.54 2.56 -0.02 <0.7 0.09 0.08 0.004 <0.6 0.663 Inland 1 14 14 1.94 1.91 0.04 <0.4 0.25 0.25 -0.009 <0.06 0.983 Inland 2 14 17 2.35 2.43 -0.07 <0.2 0.12 0.12 0.003 <0.6 0.825 Spring Coastal 21 22 2.56 2.61 -0.04 <0.5 0.10 -0.11 -0.01 <0.7 0.912 Inland 1 18 19 2.48 1.5 0.99 <0.0001 0.11 -0.42 -0.31 <0.0001 0.283 Inland 2 24 24 2.56 2.48 0.08 <0.05 0.10 -0.11 -0.01 <0.01 0.701 nocturnal sampling in 42% (Shannon- Wiener) and 58% (Simpson) of samples over four sea- sons {P < 0.05) (Table 2). Although Shannon- Wiener indices for diurnal and nocturnal col- lections from two orchards (Summer Inland 2, Autumn Coastal) showed no difference (P < 0.3), Simpson indices were significantly dif- ferent (P < 0.05). Numerical dominance by some species (i.e., Zenodorus orbiculatus (Keyserling), Salticidae, and Cyrtophora mol- uccensis Doleschall, Araneidae) was consid- erably higher in diurnal than nocturnal sam- ples in Summer Inland 2, while dominance varied between diurnal and nocturnal collec- tions in Autumn Coastal. The sensivity to dominance of the Simpson’s index may ac- count for this result. Inland 1 showed signif- icant differences between the two collection times in all seasons (Table 2). Greater num- bers of spiders collected in the warmer months provide a possible explanation for greater dif- ferences in species richness in Summer, Spring and Autumn than in Winter (Table 2). Morisita-Horn similarity indices (MH) for diurnal and nocturnal sampling were relatively high in comparison with those for sampling methods. Most MH values for sampling times corresponded with the diversity indices, i.e., similar MH, similar H and D (Table 2). How- ever, while MH and D for Summer Inland 2 showed similarity between the two sampling times, H showed a significant difference. Nine rare species (i.e., < 2 individuals collected) which were not common to both times were collected in this orchard; the significant dif- ference between the two sampling times is a result of the Shannon- Wiener sensitivity to rare species. Most families were included in collections at both sampling times. Nocturnal spiders, like Eriophora transmarina, the clubionid Cheir- acanthium sp. ‘a’, and Heteropoda sp. ‘a’, were collected in greater numbers at night than in the day time. Families collected only at night were Deinopidae, Gnaphosidae, Lam- ponidae and Mimetidae. Combined vacuum data for each orchard in each location in each season show 57-75% of taxa were collected diumally; 69-87% were collected nocturnally (Figure IB). DISCUSSION Effect of sampling method. — In contrast to this study, pitfall traps collected more spi- der taxa in heathland than sweep-netting and visual searching (Churchill 1993) or vacuum sampling (Merrett & Snazell 1983). Vegeta- tional architecture plays a major role in the species composition found within a habitat (Scheidler 1990), and vegetation which is structurally more complex can sustain a high- er abundance and diversity of spiders (Hatley & MacMahon 1980). Diversity in web-build- ing spiders is significantly correlated with vegetation height (Greenstone 1984) and high species diversity in wandering ground-dwell- 180 THE JOURNAL OF ARACHNOLOGY A B FAMILY 86% GENUS 97% SPECIES VACUUM PIT-TRAP DIURNAL NOCTURNAL Figure 1. — (A). Total number of taxa caught by vacuum and pit-trap sampling methods (Percentages are of total taxa collected). (B). Total number of taxa caught by diurnal and nocturnal sampling (Per- centages are of total vacuum samples — pit-trap data are not included in this analysis). ing spiders is correlated with large amounts of litter (Uetz 1975; Koch & Majer 1980). The low shrubs and abundant ground cover in Tas- manian heathland (Churchill 1995) differ markedly from the mature (> 2 m high) citrus trees with little understorey. Differences in vegetational architecture at the two sites ac- count for the different community structures seen in foliage and ground-dwelling spiders. Effect of sampling time. — Although other studies included observation or sampling of nocturnal species, for various reasons such as results were not quantified (Provencher et al. 1988) or results were combined (Costello & Daane 1995), these studies could not be com- pared with the present study in terms of dif- ferences in species diversity, abundance or similarity between nocturnal and diurnal col- lections. Coddington et al. (1996) found time of day had no significant bearing on the taxonomic composition of the samples from temperate forests. However, these authors recommend both day and night collecting to maximize the species richness of the samples. In the same temperate forest, abundance of adult spiders GREEN— SAMPLING METHOD AND TIME DETERMINE SPIDER COLLECTIONS 181 differed significantly between day and night samples, but similarity indices were similar for the two time periods (Dobyns 1997). Trop- ical forests produced significantly more spe- cies in nocturnal samples than the temperate forests (Coddington et al. 1991). This agrees with the present study which was conducted in sub-tropical citrus, suggesting that species differences are greater in tropical forests than in temperate forests and, consequently, that nocturnal predation is higher in sub-tropical and tropical zones. Sampling times and methods showed dif- ferent profiles at the family level in spider as- semblages. Similarity (MH) indices for differ- ences in sampling times did not all show differences in abundance. However, the results demonstrate the need for different sampling times to provide a more extensive estimate of spider diversity and abundance. Spiders from 4 of 21 families were collected only noctur- nally. Had sampling been limited to daylight hours these families would not have been in- cluded in the overall composition of the spider assemblage. In conclusion, this study has established that a combination of sampling methods over two time periods, diurnal and nocturnal, is es- sential for a comprehensive assessment of the spider fauna to be made, particularly in sub- tropical areas. The vegetational architecture of a habitat must be taken into consideration be- fore sampling commences. Each sampling method was oriented to different strata of the vegetation and so spiders with contrasting for- aging behavior and habitats were collected. Vacuum sampling collected considerably more representatives of each taxonomic level but missed one group of hunting spiders, the ground-dwellers. Pit-trapping is necessary to collect these spiders. This research has important ramifications in terms of assessing biodiversity of native nat- ural enemies in agricultural ecosystems for pest management, and sampling agricultural crops in general to provide a greater under- standing of the composition of all invertebrate fauna including pests and beneficials. A com- bination of vacuum sampling and pit-trapping, used diurnally and nocturnally is recommend- ed for spider collection in tropical or sub-trop- ical orchards to sample a greater percentage of the spider fauna. ACKNOWLEDGMENTS I thank Myron Zalucki (The University of Queensland) for writing the statistical pro- grams and for valuable discussion on the data analysis. Graham White (Cooperative Re- search Centre for Tropical Pest Management), Myron Zalucki (The University of Queens- land), and Robert Raven (Queensland Muse- um) provided valuable comments on the man- uscript. Robert Raven provided help with, and facilities for, spider identification. I am grate- ful to citrus growers in Mundubbera and Coochin Creek for allowing me to collect on their properties. The study is part of a doctoral dissertation funded by a postgraduate research scholarship from Cooperative Research Cen- tre for Tropical Pest Management, The Uni- versity of Queensland. LITERATURE CITED Alderweireldt, M. & K. Desender. 1990. Micro- habitat preference of spiders (Araneae) and ca- rabid beetles (Coleoptera, Carabidae) in maize fields. Med. Fac. Landbouww. Rijksuniv. Gent., 55:501-510. Basset, Y. 1990. The arboreal fauna of the rainfo- rest tree Argyrodendron actinophyllum as sam- pled with restricted canopy fogging: Composi- tion of the fauna. The Entomologist, 109:173- 183. Bishop, L. & S.E. Riechert. 1990. Spider coloni- zation of agro-ecosystems: mode and source. Environ. Entomol., 19:1738-1745. Breene, R.G., D.A. Dean & R.L. Meagher, Jr. 1993a. Spiders and ants of Texas citrus groves. Florida. Entomol., 76:168-170. Breene, R.G., R.L. Meagher, Jr. & D.A. Dean. 1993b. Spiders (Araneae) and ants (Hymenop- tera: Formicidae) in Texas sugarcane fields. Flor- ida Entomol., 76:645-650. Churchill, T.B. 1993. Effects of sampling method on composition of a Tasmanian coastal heathland spider assemblage. Mem. Queensland Mus., 33: 475-481. Churchill, TB. 1995. Scales of spatial and tem- poral variation in a Tasmanian heathland spider community. Ph.D. dissertation, Griffith Univer- sity, Brisbane, Australia. Churchill, T.B. 1997. Spiders as ecological indi- cators: an overview for Australia. Mem. Mus. Victoria, 56:331-337. Clausen, I.H.S. 1986. The use of spiders (Araneae) as ecological indicators. Bull. British Arachnol. Soc., 7:83-86. Coddington, J.A., C.E. Griswold, D.S. Davila, E. Penaranda & S.E Larcher. 1990. Designing and testing sampling protocols to estimate biodiver- 182 THE JOURNAL OF ARACHNOLOGY sity in tropical ecosystems, Pp. 44-60. In The Unity of Evolutionary Biology. Proc. Fourth In- tern. Congress of Systematic and Evolutionary Biology, Vol. 1. (E.C. Dudley, ed.), Dioscorides Press, Portland, Oregon. Coddington, J.A., L.H. Young & EA. Coyle. 1996. Estimating spider species richness in a southern Appalachian cove hardwood forest. J. ArachnoL, 24:111-128. Costello, M.J. & K.M. Daane. 1995. Spider (Ara- neae) species composition and seasonal abun- dance in San Joaquin Valley grape vineyards. Environ. EntomoL, 24:823-831. De Barro, P.J. 1991. A cheap lightweight efficient vacuum sampler. J, Australian EntomoL Soc., 30: 207-208. Dobyns, J.R. 1997. Effects of sampling intensity on the collection of spider (Araneae) species and the estimation of species richness. Environ. En- tomoL, 26:150-162. Greenstone, M.H. 1984. Determinants of web spi- der species diversity: vegetation structural diver- sity vs. prey availability. Oecologia, 62:299-304. Hatley, C.L. & J.A. MacMahon. 1980. Spider com- munity organization: seasonal variation and the role of vegetational architecture. Environ. Ento- moL, 9:632-639. Koch, L.E. & J.D. Majer. 1980. A phonological investigation of various invertebrates in forest and woodland areas in the south-west of Western Australia. J. Roy. Soc. Western Australia, 63:21- 28. Krebs, C.J. 1989. Ecological Methodology. Harper Collins, New York. Maelfait, J.R, R. Jocque, L. Baert & K. Descender. 1990. Heathland management and spiders. Acta Zool. Fennica, 190:261-166. Mason, R.R. 1992. Populations of arboreal spiders (Araneae) on Douglas-firs and true firs in the in- terior Pacific North-west. Environ. EntomoL, 21: 75-80. Merrett, P. 1983. Spiders collected by pitfall trap- ping and vacuum sampling in four stands of Dor- set heathland representing different growth phas- es of heather. Bull. British ArachnoL Soc., 6:14- 22. Merrett, P, & R. Snazell. 1983. A comparison of pitfall trapping and vacuum sampling for assess- ing spider faunas on heathland at Ashdown For- est, south-east England. Bull. British ArachnoL Soc., 6:1-13. Nyffeler, M. & G. Benz. 1987. Spiders in natural pest control: A review. J. Appl. EntomoL, 103: 321-339. Provencher, L., D. Coderre & C.D. Dondale. 1988. Spiders (Araneae) in cornfields in Quebec. Ca- nadian EntomoL, 120:97-100. Riechert, S.E. & T. Lockley. 1984. Spiders as bi- ological control agents. Ann. Rev. EntomoL, 29: 299-320. Russell-Smith, A. & N.E. Stork. 1995. Composi- tion of spider communities in the canopies of rainforest trees in Borneo. J. Trop. EcoL, 11:223- 235. Scheidler, M. 1990. Influence of habitat structure and vegetation architecture on spiders, ZooL Anz., 5/6:333-340. Solow, A.R. 1993. A simple test for change in community structure. J. Anim. EcoL, 62:191- 193. Uetz, G.W. 1975. Temporal and spatial variation in species diversity of wandering spiders (Araneae) in deciduous forest litter. Environ. EntomoL, 4: 719-724. Vangsgaard, C., E. Gravesen & S. Toft. 1990. The spider fauna of a marginal agricultural field (Ar- aneae). EntomoL Meddr., 58:47-54. Wolda, H. 1981. Similarity indices, sample size and diversity. Oecologia, 50:296-302. Young, O.P. & T.C. Lockley. 1985. The striped lynx spider, Oxyopes salticus (Araneae: Oxyopi- dae), in agro-ecosystems. Entomophaga, 30:329- 346. Young, O.P. & T.C. Lockley. 1990. Autumnal pop- ulations of arthropods on aster and goldenrod in the delta of Mississippi. J. EntomoL Soc., 25: 185-195. Manuscript received 1 May 1998, revised 6 March 1999. 1999. The Journal of Arachnology 27:183-188 NOTES ON THE BIOGEOGRAPHY AND NATURAL HISTORY OF THE ORB WEAVING SPIDER CAREPALXIS (ARANEAE, ARANEIDAE), INCLUDING A GUMNUT MIMIC FROM SOUTHWESTERN AUSTRALIA Barbara York Main: Department of Zoology, University of Western Australia, Nedlands, Western Australia 6907 Australia ABSTRACT. The biogeography of the Gondwanan orbweaving spider Carepalxis is reviewed. The genus occurs in Central and northern South America, Australia and New Guinea. It is recorded for the first time from Western Australia. Mimicry of a gumnut (eucalypt seed capsule) is described and illustrated for a southwestern Australian species. It is postulated that the mimicry protects the spiders from bird predation. The Australian spider fauna includes a sig- nificant Gondwanan component. Even amongst taxonomically rich, cosmopolitan families such as the Araneidae, there are some genera with characteristic Gondwanan geo- graphic ranges. The orbweaving genera Par- araneus Caporiacco 1940 and Carepalxis L. Koch 1872 are examples. Pararaneus (like the mygalomorph Paramiginae) is distributed in Africa and Australia where it is possibly con- fined to southwestern Western Australia (Main unpubl. data). Carepalxis (as with the myga- lomorph Actinopodidae, Goloboff & Platnick 1987) occurs in Central and South America and Australia (Levi 1992; Bonnet 1956; Roewer 1942) and New Guinea (Chrysanthus 1961; Brignoli 1983). Within Australia, Carepalxis has been re- corded previously only from eastern Australia. Some years ago in southwestern Australia I observed an interesting specimen of Carepa- lxis which appeared to mimic a gumnut (seed capsule of a eucalypt) and this aroused my interest in the genus. Prior to this and more recently the Western Australian Museum has acquired a small collection of specimens of Carepalxis comprised of several species from widely scattered localities in Western Austra- lia. This paper presents preliminary remarks on the systematics and biogeography of Austra- lian Carepalxis species, records for the first time the occurrence of the genus in Western Australia, and describes the natural history and the apparent mimicry of a southwestern Australian species. SYSTEMATICS AND BIOGEOGRAPHY Carepalxis L. Koch Carepalxis L. Koch 1872: 123. Type species by monotypy C. montifera L. Koch 1872: 123. Ho- lotype female from Mackay, Queensland, Austra- lia, originally in Godeffroy Museum, now in Zoologisches Institut and Museum, Hamburg (not examined). Carepalxis: Davies 1988: 300, PI. 22. Carepalxis: Levi 1992: 252. See also Bonnet (1956), Roewer (1942) and Brig- noli (1983) for species lists and synonymies. Comments.— Levi examined the type of montifera and reported that it has a shrivelled abdomen and that the scape of the epigynum was tom off (Levi 1992). Davies (1988) illus- trated a specimen, including the epigynum, which she identified as C. tuberculata Key- serling. Diagnostic notes.— The most characteristic feature of Carepalxis is the pair of large humps on the carapace of the female. These have the appearance of two subdued horns (see Davies 1988). As stated by Levi (1992) Gasteracantha also has carapace humps but they are much smaller. The abdomen is often tuberculate, is high in the front and overhangs the carapace where it fits snugly into the space behind the caput humps. The epigynum in Australian species is basally broad with a short, pointed or long, finger like scape hinged at the front then directed backwards (see Da- vies 1988). Levi (1992) redescribed the males 183 184 THE JOURNAL OF ARACHNOLOGY identified by Simon (1896) as of C. tubercu- latua Keyserling and noted that the carapace had only slight swellings in place of the humps of the female. Other features of the male included concave chelicerae associated with a large palpus, a single macroseta on the palpal patella, a tooth on the endite, and a hook on the first coxa. He also gave some details of the palpal organ. Davies (1988) had earlier stated that many “unmatched” males of Carepalxis were known and included male features such as the spur (“hook” of Levi) on coxa I and a single spine (“macroseta” of Levi) on the palpal patella in her generic key of Australian orbweaving spiders. Biogeography and taxonomy of spe- cies.— Levi (1992) recognized three species from Central and South America in his review of the American species. There are currently eight species recognized from Australia and two from New Guinea (Roewer 1942; Bonnet 1956; Chrysanthus 1961). The occurrence of Carepalxis in Australia and Central and South America but not in Af- rica parallels that of the South American Tac- zanowskia Keyserling 1879 and the Australian Celaenia Thorell 1868 (the latter also occurs in New Zealand) which are regarded as sister genera (Simon 1895; Eberhard 1981; Levi 1996). Such a geographic disjunction suggests that Carepalxis was established as a genus at least by the early Tertiary before the separa- tion of South America and Australia but after the breakaway of Africa post- Jurassic. All currently recognized Australian species (Bonnet 1956; Roewer 1942; Brignoli 1983) are from eastern Australia. All types are fe- males and their localities and depositions fol- low. Carepalxis beelzebub (Hasselt) 1873, Melbourne Victoria, Amsterdam Netherlands; C. bilobata Keyserling 1886, Peak Downs Queensland, Museum Godeffroy now in Zool- ogisches Institut und Museum (ZMH) Ham- burg); C. coronata (Rainbow) 1896, New England New South Wales, Australian Muse- um (AM) Sydney; C furcula Keyserling 1886, Peak Downs, Queensland, ZMH; C lichensis Rainbow 1916, Gordonvale, Queens- land, AM; C. montifera L. Koch 1872, Mack- ay, Queensland, ZMH; C poweri Rainbow 1916, Narabeen New South Wales, AM; C tuberculata Keyserling 1886, Sydney New South Wales, Rockhampton and Peak Downs Queensland, ZMH. Figure 1 . — Distribution of Carepalxis in Western Australia. Of the American species, Levi (1992) con- sidered C. camelus Simon from Paraguay and Argentina to be the most similar to the type species, C. montifera from Queensland, Aus- tralia. This opinion appeared to be based on similarity of the long scape of the epigynum of C. camelus and illustrations by Davies of a specimen identified by her as tuberculata. Records of Carepalxis from Western Aus- tralia: Carepalxis is now recorded from West- ern Australia for the first time. There are 44 specimens of the genus in the Western Aus- tralian Museum and three in my collection (BYM) held at the Zoology Department, Uni- versity of Western Australia. These appear to belong to seven species, three of which show similarities to C. tuberculata, C. furcula and C. beelzebub. Locality records are indicated on the map in Fig. 1 showing the scattered, wide distribution in Western Australia from the tropics and arid interior to the mesophytic and coastal southwest. The apparent central gap in the continental distribution is anomalous. Although there is a tendency to think of Gondwanan distributions as pertaining to southern Australia (and there are many such relictual distributions), clearly some Gondwanan genera occur also in the tropics and even arid regions. Within the Ar- aneidae, Celaenia Thorell (sister genus of the South American Taczanowskia Keyserling) has a continental distribution, including arid habitats. Considering the wide (although ap- parently disparate) geographic range of Car- epalxis on both “sides” of the Australian con- main— BIOGEOGRAPHY AND MIMICRY OF CAREPALXIS 185 tinent it would be expected to range also from Darwin to Adelaide through the center. NATURAL HISTORY OBSERVATIONS AND DISCUSSION Predation. — Thirty of the 44 specimens of Carepalxis in the WAM collection were found in dissected mud wasp nests. Twenty-eight of these, from Sabina River near Busselton, were from eight nests (which also included other Araneids) made by solitary wasps of a Polis sp. (J. Waldock pers. comm.). Of the other two specimens, one from Kathleen Valley was re- ported by the collector to have come from “a hornet’s nest packed with spiders,” the other from Darlington (near Perth) was recorded as “prey of a wasp.” It is well known that var- ious mud-dauber pompilids prey on araneid spiders which they seek out while the spiders are resting during the day. Also of interest and direct relevance is Rainbow’s record of a spec- imen of Carepalxis bilobata Keyserling 1886 “from nest of Scelephron sp.” from Cook- town, Queensland (Rainbow 1916). Other predators probably include birds such as hon- eyeaters which are well known for their habit of searching out spiders as prey. Mimicry and web. — On 24 May 1980 in West Cape Howe National Park I was sur- prised to observe a specimen of a Carepalxis species sitting among some buds of a jarrah tree {Eucalyptus marginata Donn. ex Sm.) (see Main 1988 for description of the site). The spider was unnoticed, until it flexed and retracted its legs. This was due to its striking resemblance to the seed capsules (gumnuts) borne abundantly on the jarrah tree along with developing flower buds. This resemblance re- sulted from the shape and colour of the spi- der’s abdomen. The “folium” or leaf pattern on the back of the abdomen was much darker than the general background color. The folium was not posteriorly attenuated as in many ar- aneids such as Araneus Clerck 1757 and Er- iophora Simon 1864 species, but was longi- tudinally compressed and roughly circular in outline. This, combined with the characteristic squat and dorsally flattened shape of the ab- domen, presented in profile an urn shape (like a smooth- walled, woody jarrah “nut”) and dorsally the dark, truncated pseudo-folium re- sembled the opening of the nut from which the seeds are shed (Figs. 2, 3 & 4). Web construction: The spider was kept Figure 2. — Jarrah “nut” (seed capsule) and Car= epalxis mimic, oblique profile view. (Scale bar = 5.0 mm). alive in the laboratory for a month to observe its behavior. Dry jarrah twigs with gumnuts were stood in a small container of soil and over these was placed a large glass battery jar (18 X 20 X 28cm). The spider constructed a horizontal line between the ends of two ter- minal twigs and hung suspended from the line each night for 15 nights. The spider ap- peared to hold the thread with the first legs outstretched in front and with the fourth legs stretched behind. Legs 2 and 3 were bent close to the body but also held the thread. It varied its position at either end of the thread on different nights. The first day after con- structing the thread the spider rested on the floor of the cage. Thereafter, when away from the thread it took up a position against a gumnut to which it showed a remarkable re- semblance. After 15 days the spider con- structed, during one night, a complete verti- cal orbweb and hung suspended upside down at the hub. The sticky spiral began some dis- tance from the hub which appeared to be “open.” It is remarkable that the spider was able to construct the web in the absence of any appreciable air currents below the battery jar. There are few references to the web of Car- epalxis. However, Rainbow (1909) referred to the typical “orbicular” web of C. tuberculata. Rainbow (1916) also described (from a col- lector’s notes) three spherical egg cocoons of 186 THE JOURNAL OF ARACHNOLOGY Figures 3, 4, — Jarrah “nut” and Carepalxis mim- ic “perched” on stem below nut. 3. Profile view of spider; 4. Dorsal view (or “rear” view of abdomen) of spider. (Scale bars ==5.0 mm). Carepalxis Uchensis Rainbow 1916, “sus- pended in a horizontal line in forest tree.” Mimicry in araneid spiders: Over a century ago Rainbow (1896) in his dissertation on pro- tective coloration and mimicry of spiders cited several examples of mimicry in Australian Ar- aneidae. These included “mimicry” and con- cealment of egg cocoons. However, most ex- amples were of spiders fitting one of the following three categories. (1) Those which resembled parts of plants through a combi- nation of color and pattern; such mimicry was interpreted as helping the spiders to avoid at- tack by birds and reptiles. Rainbow listed the following instances in this category: (a) Epei- ra ficta (= Araneus fictus (Rainbow 1896)) and E. similaris (= Zealaranea crassa (Wal- ckenaer 1842)) which “mimics” leaves exhib- iting patterns of holes caused by insect attack. (b) Acrosoma Perty 1833, which bears a like- ness to thorny leaves and “knots” of shrubs. (c) Tetragnatha Latreille 1804, Phono gnatha Simon 1894, and Epeira higginsii L. Koch 1871 (= Arachnura higginsii (L. Koch)) (all at that time included in the Argiopidae (= Ar- aneidae)) which have a likeness to twigs and/ or leaves. Not mentioned by Rainbow is the notable blend of Dolophones Walkenaer 1837 with twigs around which it wraps the broad, flattened abdomen thereby eluding visual de- tection. (2) Those spiders which bear a like- ness to insects that are unpalatable to birds (i.e., Batesian mimicry), e.g., Cyrtarachne caliginosus (Rainow 1894) (= Ordgarius fur- catus O.P. Cambridge 1877)) in which the male has hairs that mimic in appearance the irritating hairs of certain caterpillars. (3) Those spiders which attract insect prey, e.g., Celaenia excavata (L. Koch 1867) which through its resemblance to a bird dropping, Rainbow suggested lured potential insect prey in “quest of food.” ''Gumnut” mimicry as protection against predators: There do not seem to be docu- mented examples of araneids with such pre- cise specific plant models as that of the “gumeut” Carepalxis recorded here, the ad- vantage of which is assumed to be protection against bird predators. In this light it is of interest that four of the Carepalxis specimens in the WAM collections found in dissected wasp nests show a remarkable similarity to the West Cape Howe gumnut mimic. The specimens are not in ideal condition and some are shrivelled but show a distinctly dark, modified folium roughly circular in out- line on a lighter ground. In that such wasps are diurnal hunters the spiders must have been searched out while resting rather than when exposed in a web. Hence although the mimicry may protect the spiders from birds such as honeyeaters which are very persistent in their foraging amongst foliage, clearly some wasps are able to detect them. MAIN— BIOGEOGRAPHY AND MIMICRY OF CAREPALXIS 187 Development of “gumnut” mimicry, — Many araneid genera, e.g., Araneus Clerck 1857 and Eriophora Simon 1864 have a leaf- like (folium) pattern on the dorsum of the ab- domen. The folium margin is frequently sten- cilled and the folium itself may be a different color to the background color of the integu- ment. Some species exhibit a distinct poly- morphism for color patterns which may be fixed throughout life. For example, E. biapi- cata (L. Koch 1871) shows a range of color combinations through greys and browns of fo- lium and background, with the folium toning or contrasting with the abdomen background color. Some individuals appear to lack pig- ment in the folium area with the guanin de- posits showing through the integument as a white folium patch or arrowhead (see Main 1976, PI. 36; as Araneus transmarinus (Key- serling 1865)). Spiders rest away from their webs during the day on foliage, seed heads, bark and build- ings. Matching colors of spiders and back- ground is often striking. Nevertheless it is doubtful whether these spiders can change color abruptly or at all after the first few in- stars. It is more likely that selection occurs on spiderlings during their early instars. Howev- er, there is certainly scope for manipulation and experimentation with spiderlings on dif- ferent backgrounds to test colour modifica- tion. Species of Carepalxis possibly exhibit color variations amongst individuals like other ara- neids, particularly Eriophora. Thus selection could favor those individuals with a dark ab- dominal patch (gumnut mimics) which would be less conspicuous to birds (at least for spi- ders hosted in jarrah foliage). The questions to ask now are: Is there variation in initial color patterns within cocoon clutches? If so, is there selection according to differential fo- liage settlement, i.e., on jarrah versus other vegetation in relation to color pattern (gumnut versus non-gumnut mimics)? Observations and museum collections sug- gest that Carepalxis spiders, of any design, although geographically widespread, are rare. Such perceived rarity poses real difficulties for a biologist’s analysis of the mimicry. How- ever, the wasps know otherwise! ACKNOWLEDGMENTS I thank Dr. Mark Harvey and Julianne Wal- dock of the Western Australian Museum for giving access to the collection of Carepalxis. Many of the spiders from the wasp nests were collected by Julianne and her colleagues. I thank Vic Smith for a spider from Goode Beach (Albany). Permission from the Depart- ment of Conservation and Land Management to collect spiders in National Parks is grate- fully acknowledged. LITERATURE CITED Bonnet, P. 1956. Bibliographia Araneorum, Tome 2, pp.9 19-925. Toulouse: Douladoure. Brignoli, P.M. 1983. A catalogue of the Araneae described between 1940 and 1981.(P. Merrett, ed.). Manchester: Manchester Univ. Press, 755 pp. Chrysanthus, P. 1961. Spiders from South New Guinea IV. Nova Guinea (N.S.), 10 (ZooL): 195-214. Davies, V.T. 1988. An illustrated guide to the gen- era of orb-weaving spiders in Australia. Mem. Queensland. Mus., 25:273-332. Eberhard, W.G. 1981. Notes on the natural history of Taczanowskia sp. (Araneae: Araneidae). Bull. British Arachnol. Soc., 5:175-176. Goloboff, P.A. & N.I. Platnick. 1987. A review of the Chilean spiders of the superfamily Migoidea (Araneae, Mygalomorphae). American Mus. Nov., 2888:1-15. Koch, L. 1872. Die Arachniden Australiens. Bauer and Raspe: Niimberg. Levi, H.W. 1992. The American species of the orb- weaver genus Carepalxis and the new genus Rubrepeira (Araneae: Araneidae). Psyche, 98: 251-264. Levi, H.W. 1996. The genus Taczanowskia of the orb- weaver spider family Araneidae (Araneae). Anales Inst. Biol. Univ. Nac. Aut. Mexico, Ser. ZooL, 67:183-195. Main, B.Y 1976. Spiders. Collins: Sydney. Main, B.Y. 1988. The biology of a social thomisid spider. Pp. 55-73. In Australasian Arachnology. (A.D. Austin & N.W. Heather, eds.). Australian Entomol. Soc., Brisbane. Main, B.Y. 1991. Occurrence of the trapdoor spi- der genus Moggridgea in Australia with descrip- tions of two new species (Araneae: Mygalomor- phae: Migidae). J. Nat. Hist., 25:383-397. Rainbow, W.J. 1896. Descriptions of some new Ar- aneidae of New South Wales. No. 6. Proc. Linn. Soc. New South Wales, 21:320-344. Rainbow, W.J. 1909. Notes on the architecture, nest habitat and life history of Australian Ara- neidae. Rec. Australian Mus., 7:212-234. Rainbow, W.J. 1916. Arachnida from Northern Queensland, Part II. Rec. Australian Mus., 11: 78-119. 188 THE JOURNAL OF ARACHNOLOGY Roewer, C.F. 1942. Katalog der Araneae von 1758 bis 1940. Bremen: Natura, 1:1-1040. Simon, E. 1895. Histoire Naturelle des Araignees, 1. 2nd ed. Pp. 761-1084. Paris: Roret. Simon, E. 1896. Liste der Arachniden der Se- mon’schen Sammlung in Australien und dem Malayischen Archipel. Jenaische Denkschrift., 8: 341-352. Manuscript received 1 May 1998, revised 3 March 1999. 1999. The Journal of Arachnology 27:189-195 EFFECTS OF SHORT-TERM SAMPLING ON ECOLOGICAL CHARACTERIZATION AND EVALUATION OF EPIGEIC SPIDER COMMUNITIES AND THEIR HABITATS FOR SITE ASSESSMENT STUDIES Uwe Riecken; Federal Agency for Nature Conservation, MallwitzstraBe 1-3, D-53177 Bonn, Germany ABSTRACT, Epigeic invertebrates such as spiders are of increasing importance for habitat character- ization and for assessments within environmental plannings in Germany and other European countries. Due to high costs for spider sampling (e.g., with pitfall traps), proposals for a limited sampling effort are required for the practical use. The results of a two-year study with continuous sampling are compared to results of short-term sampling and to results of a reduced number of traps. The same data set is used for all evaluations. Decreasing sampling effort generally reduced the number of recorded species and led to a biased ecological characterization of the spider communities. Reducing the number of pitfall traps used provided a more representative sample than did reducing the duration of sampling. In general, errors based on reduced sampling were lower for agricultural than for natural habitats. These results offer practical use of spiders for bioindication in future environmental planning. The number of epigeic arthropod species which are collected in a specific area depends mainly on the sampling effort, such as on the number of traps or on the length of the sam- pling period (Stein 1965). One reason for this phenomenon is the finding that rare species or species with short activity periods (but also species living in adjacent habitats) are more likely to be caught with increasing sampling intensity. Therefore, sampling by means of pitfall traps is usually carried out during the entire growing season (in Germany: March- October) and is often repeated in subsequent years to obtain data for a reliable analysis of the species composition of the arthropod com- munity. Unfortunately, there are often limited financial resources for these studies and the results are often required within a short period of time. Therefore, there have been several proposals for a limited investigation program concerning pitfall traps, including the recom- mendations for sampling periods of only six weeks (Duelli et al. 1990) or 10 weeks (Finck et al 1992) or the reduction of the sampling period to only one season (spring or summer; Maelfait & Desender 1990). Alternatively, sampling efforts can be reduced by limiting the number of pitfall traps per habitat. However, there is little knowledge about the effects of a reduced sampling effort on the quality of the results, and on the conclusions based on these results. This study tests the ef- fects of short term sampling by: (1) comparing data from an eight week trapping period to data from continuous trapping throughout the season (28 weeks; March-October) and (2) analyzing the results obtained by a reduced number of traps. Data are analyzed to examine both the impact of the reduced sampling effort on species numbers and on the ecological characterization of the spider communities of 20 different study sites. STUDY AREA This study was conducted in a typical ag- ricultural landscape south of Bonn (North- Rhine-Westphalia, Germany), which is char- acterized by intensively-used arable land, meadows, orchards and patchily distributed small forests. Semi-natural landscape ele- ments include small river valleys with adja- cent wet grassland, small riparian alder for- ests, river banks and small patches of abandoned formerly wet pastures. A set of 20 different habitats representing the most im- portant habitat types was investigated along two transects across two valleys (transect I near the village of Pech; transect II near the village of Zuellighoven). These transects ranged from semi-natural to agricultural areas. 189 190 THE JOURNAL OF ARACHNOLOGY Table 1. — List of the investigated sites (I = transect I (near the village of Pech), II = transect II (near the village of Zuellighoven), a = additional site). Code Tran- sect Investigation period Habitat fori I 3/90-10/91 beech-oak forest on acid soil with poor herb vegetation for22 I 3/92-10/93 beech-oak forest on acid soil with poor herb vegetation mixed with Pinus silvestris and Ilex shrubs for24 a 3/92-10/93 beech-oak forest on acid soil with poor herb vegetation alfll II 3/90-10/91 pastured red alder forest with springs alfl4 II 3/90-10/91 red alder forest with natural flood dynamic rib5 I 3/90-10/91 shady river bank with red alder riparian forest, mixed with Prunus padus rib 13 II 3/90-10/91 river bank with red alder riparian forest partly mixed with Urti- ca dioica stands rib26 a 3/92-10/93 muddy river bank with red alder riparian forest rib27 a 3/92-10/93 top of river bank 26 with mesotrophic grassland Molinio-Ar- rhenatheretea-commumiy pla25 a 3/92-10/93 young plantation of Quercus petraea, mixed with blackberry bushes and birch trees on acidic soil falls a 3/90-10/91 mesophilic fallow surrounded by forests, partly covered with blackberry bushes and young trees (aspen) rib4 I 3/90-10/91 linear red alder riparian forest close to the river bank exposed to the sun with rich tall herb vegetation wfal2 I 3/90-10/91 wet fallow {Convolvuletalid), smaller parts with Filipendulion- and Magnocaricion-VQgQidiiion wfall7 a 3/90-10/91 wet fallow with sedges, Carex acutiformis-community, Magno- caricion wpasl2 II 3/90-10/91 wet pasture with Juncus ejfusus pas7 I 3/90-10/91 intensively managed mesophilic pasture Lolio-Cynosuretum pas 15 II 3/90-10/91 intensively managed mesophilic pasture Lolio-Cynosuretum pas 19 a 3/90-10/91 intensively managed mesophilic pasture Lolio-Cynosuretum with apple trees, surrounded by forests field 10 II 3/90-10/91 extensively managed crop field with rich stands of weeds, Aphano-Matricarietum fields I 3/90-10/91 intensively managed crop field with poor or no weeds Additionally, some samples were collected in adjacent localities with characteristic habitat types not covered within the transects (Table 1; for details see Riecken 1998). METHODS Spiders were sampled by means of pitfall traps (350 ml honey-glasses, opening diameter 7 cm), filled with 125 ml of formaldehyde so- lution (2%) and protected by a roof of a clear acrylic plastic (20 cm X 20 cm). Four traps were exposed at each site (in line, distance 5 m) for two years (for two different periods: 1990, 1991, and 1992, 1993 March to October every year; Table 1). All traps were emptied every two weeks. Duelli et al (1990) originally proposed two sampling periods of five weeks a year, with the traps being emptied once a week. Further analysis should include only data from those three weeks of each period during which the greatest number of specimens were caught. In this study, traps were emptied every two weeks. As it was impossible to take data from three-week periods, two four-week periods seemed to be a good approximation of Duel- li’s method. Applying this protocol for a lim- ited sampling period resulted in a short-term data set for the following time periods (two four- week periods from both years): sites in- vestigated 1990 and 1991 (see Table 1): 18 May- 12 June 1990, 9 August-5 September 1990, 16 May-11 June 1991 and 8 August-5 September 1991; sites investigated 1992 and 1993 (see Table 1): 21 May-16 June 1992, 13 August-9 September 1992, 19 May- 15 June 1993 and 12 August-8 September 1993. Parametric t~tests were used for compari- RIECKEN— EFFECTS OF SHORT TERM SPIDER SAMPLING 191 short-term sampling 2 out of 4 pit-fall-traps study sites Figures 1, 2, — Percentage of species numbers in different sampling protocols. 1. Results from a short- term sampling (eight weeks a year); 2. Results from a reduced data set (average species number from all possible pairs of two out of four pitfall traps) in comparison to the complete data set from sampling throughout two growing seasons (March to October) and all traps. sons of percentage values (Jongman et al. 1987). All data sets were tested for a normal distribution. All comparisons were made between the re- sults of the complete data set over two seasons (28 weeks each = 100%) and reduced data sets. I first compared the results from two four-week periods (short-term sampling), and then the results of a reduced number of pitfall traps. In the case of the reduced trap numbers, 192 THE JOURNAL OF ARACHNOLOGY 0 O C m ■o £ □ m m > m 0 100% 80% 60% 40%- 20%o- forl for24 alf11 rib5 rib26 pla25 nb4 wfal2 wpas12 pas7 pas19 fields for22 alf14 ribIS rib27 fal18 wfallJ paslB fieldIO Study sites ■i mesophilic forests wet forests stenotopic forests euiytopic wet open habitats [Zj dry open habitats □□ eurytopic species Figure 3. — Composition of the spider communities based on classification of habitat affinity. the arithmetic means of the results for each trap (n = 4) or each possible pair of traps (n ” 6) were calculated (bars in Figs. 2 and 5). Habitats were classified by a cluster anal- ysis based on the percentage similarity (Ren- konen 1938), using the computer program COMM (Piebenburg & Piatkowski 1992) and the “unweighted pair group method using arithmetic means” (UPGM-linkage). In this method, the distances between clusters are calculated from arithmetic means of the dis- tances between the objects within the com- pared clusters (Legendre & Legendre 1983). RESULTS A general analysis, based on a total catch of 50,471 adult spiders belonging to 169 spe- cies, showed that Linyphiidae (75.4% of all specimens) and Lycosidae (18.3%) were the most abundant families. Agelenidae (2.3%), Tetragnathidae (1.9%), and Amaurobiidae (1.6%) also occurred regularly. The remaining 17 families comprised only 1.5% of the total catch, but 27% of the recorded species. Influence of short-term sampling and number of pitfall traps on species num- bers.— In the present study, a short-term trap- ping period as proposed by Duelli et al. (1990) would have reduced the number of recorded species to 64.4% of the initial sample (Fig. 1). In two habitats (forest 22 and 24), less than 50% of all species were included. By contrast, the reduced data set from the intensively used pasture 19 and from the fields contained more than 70% of all species recorded there. If data from only two of four pitfall traps were used (i.e., a reduction of number instead of time), a significantly higher proportion (P < 0.001) of species was included (on average 77.9% of the total number; Fig. 2) in com- parison to Duelli’s proposal. Even in the worst case (forest 22 and 24), approximately 70% of all species were included. Influence of reduced data sets on the eco- logical characterization of the spider com- munities.=-Bioindication or planning pro- cesses related to nature conservation often require a classification of the habitat prefer- ences or ecological characters of the recorded species. To determine whether a reduced data set would have an impact on ecological char- acterization of the spider communities, all species were classified based on literature data (Hanggi et at 1995; Platen et al. 1991; Reinke & Irmler 1994; Roberts 1985, 1987, 1995; for further details, see Rieckee 1998). The fol- lowing six habitat affiliations were distin- RIECKEN— EFFECTS OF SHORT-TERM SPIDER SAMPLING 193 short-term sampling 2 out of 4 pit-fall-traps Figures 4, 5. — Summarized differences (percentage dissimilarities) in relative abundance of six types of habitat affinity. 4. Affinity resulting from a short-term sampling in comparison to the complete data set from sampling throughout two growing seasons (March-October) and all traps; 5. Affinity resulting from a reduced pitfall number (average of all possible pairs of two out of four pitfall traps). guished: (1) species restricted to mesophilic forests, (2) species restricted to wet forests, (3) species preferring forests without being re- stricted to them, (4) species preferring wet open habitats such as bogs, grassland or shores, (5) species preferring dry open habi- tats, such as meadows or heathers, and (6) eu- rytopic species that cover a broad range of open habitats, e.g., all types of meadows, fields and fallows. Based on these classifica- tion, the community compositions were deter- mined (Fig. 3). 194 THE JOURNAL OF ARACHNOLOGY rib4 wpas12 aif14 rib27 ribs rib13 pasIS pas7 6 7 Figures 6, 7. — UPGM-linkage dendrogram based on the “percentage similarity” (RENKONEN-index) classifying the study sites (r = resemblance). 6. Similarities based on the complete data set from sampling throughout two growing seasons; 7. Similarities based on data resulting from short-term sampling. The dissimilarities (based on the ecological classifications) between short-term data sets and the full data set varied between 1.7% (pasture 7) and 36.6% (river bank 5). The av- erage dissimilarity (pooling all sites) was 12.5% (Fig. 4). When analyzing only 1 out of 4 pitfall traps, the dissimilarities varied be- tween 1.0% (pasture 19) and 20.4% (river bank 4), with an average of 9.6% for all sites. This result did not differ significantly from short-term sampling (P > 0.05). Considering data from two pitfall traps (Fig. 5), the results were significantly more similar to the com- plete data set than the results from short-term sampling were (P < 0.001). Here, the dissim- ilarities varied between 0.6% (pasture 19) and 12.8% (alder forest 11), with an average of 5.6%. There are two major reasons for the rela- tively high dissimilarities resulting from a re- duced sampling period: the phenology of the dominant species and, depending on it, the differences in phenology of ecological types. Thus, the spider communities are dynamic during the season, both in species composition and in the relative abundance of ecological types. Therefore, different results can be ex- pected depending on the time frame for sam- pling, leading to assessment errors and inap- propriate nature conservation measures based on bioindication. In general, errors based on reduced sam- pling were lower for agricultural habitats (pas- tures, fields) than for semi-natural sites. The main reason for this finding is the generally low percentage of stenotopic species in all pastures and fields (except the wet pasture 12, see Fig. 3). Influence of short-term sampling on coenotic comparisons.-— The results were also strongly influenced by short-term sam- pling when different spider coenoses were compared by cluster analysis (UPGM-linkage) based on the “percentage similarity” index (Renkonen 1938). Using the complete data set, five clusters of habitats could be distin- guished at a similarity level > 40% (Fig. 6). This result confirms the expected pattern based on the studied habitat types. For ex- ample, all forests and all agricultural sites RIECKEN— EFFECTS OF SHORT-TERM SPIDER SAMPLING 195 were clustered together. The reduced data set, however, produced a completely different re- sult. Even the three quite similar forest sites or the pastures were grouped to different clus- ters then. CONCLUSIONS Short-term sampling reduces the number of recorded species by as much as 50% of the full set. An ecological characterization based on these results is weak, as is a characteriza- tion based on a reduced number of pitfall traps, taking only one out of four traps. In contrast to results for carabid beetles (Maelfait & Desender 1990), this reduced data set also leads to important failures in habitat classifi- cation and habitat differentiation. Consequent- ly, there will be considerable errors in site as- sessment. Also, conclusions for planning or for nature conservation activities will be bi- ased if these results are used. Short-term sam- pling seems to be acceptable only in agricul- tural habitats. Site assessment studies of epigeic spiders should be carried out through- out the whole growing season (in Germany: March-October). If financial resources are limited, a reduction of the number of pitfall traps will be more appropriate than a reduc- tion of the sampling period. ACKNOWLEDGMENTS I thank my colleagues Dr. Manfred Klein, Dr. Axel Ssymank and Dr. Wolfgang Volkl and Dr. Brent D. Opell for their valuable com- ments on the manuscript. LITERATURE CITED Duelli, R, M. Stader & E. Katz. 1990. Minimal- programme ftir die Erhebung und Aufbereitung zoookologischer Daten als Fachbeitrage zu Plan- ungen am Beispiel ausgewahlter Arthropoden- gruppen. Schriftenreihe Landschaftspflege und Naturschutz, 32:211-222. Finck, P, D. Hanuner, M. Klein, A. Kohl, U. Rieck- en, E. Schroder, A. Ssymank & W. Volkl. 1992. Empfehlungen ftir faunistisch-okologische Date- nerhebungen und ihre naturschutzfachliche Bew- ertung im Rahmen von Pflege- und Entwicklung- splanen fiir NaturschutzgroBprojekte des Bundes. Natur -1- Landschaft, 67 (7/8):329-340. Hanggi, A., E. Stockli & W. Nentwig. 1995. Le- bensraume mitteleuropaischer Spinnen.-Charak- terisierung der Lebensraume der haufigsten Spin- nenarten Mitteleuropas und der mit diesen vergesellschafteten Arten. Centre suisse de car- tographie de la faune, Neuchatel (= Miscellanea Faunistica Helvetiae, 4), 459 pp. Jongmann, R.H. Ter Braak, C.J.f. & O.f.R. van Ton- geren. 1987. Data Analysis in Community and Landscape Ecology. Centre f. Agricultural Publ. and Document. (PUDOC), Wageningen. 299 pp. Legendre, L. & P Legendre. 1983. Numerical Ecology. Elsevier, Amsterdam. 419 pp. Maelfait, J.-P & K. Desender. 1990. Possibilities of short-term Carabid sampling for site assess- ment studies. Pp. 217-225. In The Role of Ground Beetles in Ecological and Environmental Studies (N.E. Stork, ed). Intercept, Andover, Hampshire. Piepenburg, D. & U. Piatkowski. 1992. A program for computer-aided analyses of ecological field data. CABIOS, 8(6);587-590. Platen, R., M. Moritz & B.v. Broen. 1991. Liste der Webspinnen-und Weberknechtarten (Arach.: Araneida, Opilionida) des Berliner Raums und ihre Auswertung fiir Naturschutzzwecke (Rote Liste). Landschaftsentwicklung und Umweltfor- schung, S 6:169-205. Reinke, H.-D. & U. Irmler. 1994. Die Spinnenfau- na (Araneae) Schleswig-Holsteins am Boden und in der bodennahen Vegetation. Faun.-okol. Mitt., Suppl. 17:1-148. Renkonen, O. 1938. Statistisch-okologische Unter- suchungen fiber die terrestrische Kaferwelt der finnischen Bruchmoore. Ann. Zool. Soc. ZooL- Bot. Fennica “Vanamo”, 6:1-231. Riecken, U. 1998. The importance of semi-natural landscape structures in an agricultural landscape as habitats for stenotopic spiders. In Proceedings of the 17th European Colloquium of Arachnol- ogy, Edinburgh 1997:301-310. Roberts, M.J. 1985. The Spiders of Great Britain and Ireland. Vol. 1: Atypidae to Theridiosomath- idae. Harley Books, Colchester. 229 pp. Roberts, M.J. 1987. The Spiders of Great Britain and Ireland. Vol. 2: Linyphiidae and Check List. Harley Books, Colchester. 204 pp. Roberts, M.J. 1995. Spiders of Britain & Northern Europe. Collins, New York. 383 pp. Stein, W. 1965. Die Zusammensetzung der Cara- bidenfauna einer Wiese mit stark wechselnden Feuchtigkeitsverhaltnissen. Z. Morph. Okol. Ti- ere, 55:83-89. Manuscript received 1 May 1998, revised 6 March 1999. 1999. The Journal of Arachnology 27:196-200 DISTRIBUTION AND NATURAL HISTORY OF MEXICAN SPECIES OF BRACHYPELMA AND BRACHYPELMIDES (THERAPHOSIDAE, THERAPHOSINAE) WITH MORPHOLOGICAL EVIDENCE FOR THEIR SYNONYMY A. Locht, M. Yanez and I. Vazquez: Laboratorio de Acarologia “Anita Hoffmann,” Facultad de Ciencias, Universidad Nacional Autonoma de Mexico, Coyoacan 04510, D.E, Mexico ABSTRACT. This comparision of Brachypelmides and Brachypelma species is based on newly collected spiders and more than 100 specimens from five museum collections. The results show that there are six endemic species of Brachypelma in western Mexico {B. auratum, B. baumgarteni, B. boehmei, B. emilia, B. pallidum, B. smithi), presenting a gap in their distribution only where Brachypelmides klaasi is found. Brachypelma vagans is distributed along both coasts of Mexico and Brachypelmides ruhnaui is found in the central part of Mexico. Notes on natural history, a morphological comparison of 27 characters of these genera, and a discussion of the generic affinities are included. RESUMEN. De junio de 1997 a Octubre de 1998 se hizo un estudio comparativo de Brachypelmides y de las especies de Brachypelma. Se revisaron especimenes de ambos generos obtenidos en el campo recientemente y mas de 100 especimenes de cinco diferentes colecciones para realizar este estudio. Los resultados muestran que hay seis especies endemicas al Pacifico mexicano de Brachypelma {B. auratum, B. baumgarteni, B. boehmei, B. emillia, B. pallidum, B. smithi), presentando una distribucion continua a lo largo de la costa del Pacifico, siendo interrumpida por la distribucion de B. klaasi. Brachypelma vagans se distribuye en ambas costas y Brachypelmides ruhnaui en el centro del pais. Se incluyen notas de historia natural, una comparacion morfologica de 27 caracteristicas de estos generos y una discusion de las afin- idades genericas. The subfamily Theraphosinae is a speciose group from the New World, representing some of the most beautiful species of the family Theraphosidae (Perez-Miles 1992; Schmidt 1993; Smith 1993; Perez-Miles et al. 1996). The genus Brachypelma can be found from Mexico to Costa Rica (Valerrio 1980; Smith 1994). The species from the west coast of Mexico are particularly docile and colorful. These traits have led to their being collected in large numbers for the pet trade. The de- struction of the natural habitat and the high mortality before sexual maturity (99%) (Baerg 1958) are two factors that affect the popula- tions of these species, and combined with the illegal trade that normally involves the capture of preadult and adult tarantulas, can cause the extinction of these tarantulas. To regulate this trade and prevent their endangerment, all the species of this genus have been listed in ap- pendix II of CITES. In 1856 White described the first Brachy- pelma species, B. emilia, that is endemic to the Pacific coast of Mexico. Since then, an- other five species have been described that are endemic to this area {B. auratum, B. baum- garteni, B. boehmei, B. pallidum, B. smithi) (Schmidt 1992; Smith 1993; Schmidt & Klaas 1994; F.O.P Cambridge 1897). Brachypelma vagans Ausserer 1875 inhabits the same area but populations also exist along the Gulf of Mexico down to Costa Rica. Brachypelma ep- icureanum (Chamberlin) 1925 is found only in the Yucatan peninsula (Smith 1994). Schmidt & Krause (1994) described a new species of Theraphosinae from the west coast of Mexico. Although this tarantula is very similar to those of Brachypelma, they argued that it should be placed in a new genus be- cause it has a pad of plumose hairs on the femur IV, the males present a sharp tapered embolus, and the females have a bipartite and wide spermatheca. The species was named Brachypelmides klaasi. However, in the same 196 LOCHT ET AL.— NOTES ON BRACHYPELMA AND BRACHYPELMIDES 197 Figure 1. — Distribution of the species of Brachypelma and Brachypelmides in Mexico. year Smith (1994), after examining the types, concluded that B. klaasi belongs to the Bra- chypelma group, being only “its most extreme form.” PereZ“Miles et al. (1996) made a sys- tematic revision and a cladistic analysis of Theraphosinae, but they did not include Bra- chypelmides in their analysis. Schmidt (1997) described another new Brachypelmides spe- cies from the central region of Mexico, B. ruhnaui, adding support to his idea that Bra- chypelmides is a valid genus. Our research of the species of Brachypelma and Brachypelmides brings more data to the question of the distinctness of these genera. It also adds information on the natural history and distribution data for these tarantulas. METHODS The collections visited and used for the study included the following: the American Museum of National History, the California Academy of Sciences, Field Museum, Insti- tuto de Biologia, UNAM, in Mexico City (IBUNAM), and Estacion de Biologia, Cha- mela, in Jalisco. The five collections have to- gether more than 100 specimens of both gen- era. All collection data were recorded. No types were studied. Fifteen field trips were made to the west coast of Mexico from June 1997 to October 1998, and two more to the east coast in this same period. Ecological and geographical data were taken, and the specimens were brought to the lab in Mexico City Laboratorio de Acarologia “Anita Hoffmann.” Live speci- mens were put in controlled environment chambers where their reproductive behavior was studied (Yanez & Locht 1998). The mor- phological characteristics of preserved speci- mens were analyzed in more detail, and the figures presented in this work were made from them. The specific collecting data for speci- mens are not given, but a range map (Fig. 1) is included because we wish to protect the species from the illegal pet trade. DISTRIBUTION Brachypelma is a common genus in the Pa- cific and Gulf coasts of Mexico; the distribu- tion of species along the west coast is only interrupted by that of Brachypelmides klaasi, and in the central part by B. ruhaui (Fig. 1). 198 THE JOURNAL OF ARACHNOLOGY Figures 2-5. — Frontal view of the right bulbs of four species of tarantulas. 2. Brachypelma vagans right bulb; 3. Brachypelmides ruhnaui right bulb; 4. Brachypelmides klaasi right bulb; 5. Brachypel- ma smithi right bulb. Brachypelma smithi has a disjunct distribu- tion, and B. vagans has the largest distribu- tion. The only distribution that we could not verify was that for B. epicureanum, which is endemic to the Yucatan peninsula (Smith 1994). Field and collection data corresponded in all other species. The distributions given by Smith (1994) did not correspond with ours in all cases, but they are generally not contradic- tory. The other collecting data found in the descriptions of the species coincide with the areas shown in Fig. 1 (White 1856; Ausserer 1875, F.O.P. -Cambridge 1897; Schmidt 1992; Schmidt 1993; Smith 1993; Schmidt & Klaas 1994; Schmidt 1997; Locht et al. 1998). NATURAL HISTORY The natural history of Brachypelma species differs little if at all from the Brachypelmides species. The following data are field and lab- oratory observations of the two genera. Burrow Construction. — All the species studied live in burrows found in the soil, sometimes near rocks or trees, sometimes in open field, but not far from vegetation. They have only one entrance, a little wider than the tarantula’s body size. When the tarantula is active this entrance is clean and some silk can be found. When the tarantula is inactive for a long period the entrance is covered by soil and leaves that the tarantula gathers with silk. A horizontal tunnel leading from the entrance is normally three times larger than the tarantula. This tunnel is followed by a chamber 2-3 times bigger than the tarantula, where it molts, then a vertical tunnel shorter than the first one that ends with a larger chamber, where the ta- rantula rests and eats its prey. The mature fe- male’s burrow in the reproductive season has more silk in the entrance than usual. Phenology. — The tarantulas of these genera are long-lived. The males can reach maturity in 7-8 years, living only one year or less after the last molt, while the females reach maturity in 9-10 years, then live 10 more years. Com- pared to other genera of the same subfamily, they grow slowly (Smith 1994). In all species pre- adults and adults molt at the end of the dry season (June-November), males begin to wander in search of the females after they molt, and the females lay an egg-sac before they molt. The egg-sac hatches 3-4 weeks be- fore the rainy season begins. Males of all west coast species wander during daylight, partic- ularly in the morning and in the evening, while the species of the east coast and center wander at night. Color pattern. — Brachypelma klaasi col- oration is very similar to that of the six spe- cies of Brachypelma that are endemic to the west coast. Brachypelma boehmei is the more similar, having, like B. klaasi, black tarsi, or- ange-yellow metatarsi, tibias and patellas, black femora and coxae and orange-yellow hairs on the opistosoma. It differs only in the carapace, which is yellow-orange in B. boeh- mei and black in B. klaasi. Brachypelma baumgarteni is also very similar, but it has a more reddish patella. Brachypelmides ruhnaui has the same coloration of B. vagans and B. epicureanum, differing only in having a yel- low carapace, rather than black as in the oth- ers. Although all the species have striking col- LOCHT ET AL.— NOTES ON BRACHYPELMA AND BRACHYPELMIDES 199 Figures 6-9. — Dorsal view of the cleared spermathecae of four species of tarantulas. 6. Brachypelma auratum; 7. Brachypelma emilia; 8. Brachypelma vagans; 9. Brachypelmides klaasi. oration, they are in fact cryptic within their native habitat, making it very difficult to see the tarantulas, even when they are out of their burrows in daylight. DISCUSSION OF GENERIC AFFINITIES The data on distribution and natural history provide support for the hypothesis that these species are closely related. In the cladistic analysis of Perez-Miles et al. (1996) 27 char- acters were used. The characters in Brachy- pelma are: palpal bulb with concave/convex apical region; relative width of sclerites II + III (measured at 20% of its length, from the apex) wide (equal to or more than 10% of the length of the bulb); lack of the paraembolic and digitiform apophysis; presence of smooth peripheric and supernumerary keels; and a large subtegulum. The male’s tibia lacks a lat- eral process in the retrolateral region, a retro- lateral cluster of spines and a prolateral pro- cess. It has two tibial apophyses; metatarsus I lacks a basal process and is not strongly curved, its flexion provided by the outer side of the tibial spurs. Spermatheca widely fused and with unilobulated receptacles; femur III and tibia IV not incrassate; femur IV without a retrolateral scopula; urticating hairs type I and III present, type IV are absent. Trochan- teral and coxal stridulatory hairs absent; coxal spinules are absent; numerous labial cusps present; fovea without a spherical process. We compared all 27 characters among the 10 species analyzed and found that Brachy- pelmides has only one character that distin- guishes it from Brachypelma. This character is the presence of a spermatheca with two re- ceptacles separated or only partly fused. How- ever, in the genus Brachypelma some sper- mathecae are widely fused {B. smithi, B. auratum), some semi-divided {B. emilia, B. baumgarteni, B. boehmei) and some only part- ly fused {B. vagans) (Figs. 6-9). The palpal bulb morphology, principally in the embolus, is distinctive in all the species. Brachypelmides klaasi and B. runhaui have the embolus sharper and more tapered, but do not differ in the characteristics above listed from Brachypelma. The palpal bulb of B. klaasi, being wider, is more similar to that of Brachypelma species from the Pacific coast. The bulb of B. ruhnaui bulb is more similar to the thinner bulb of B. vagans (Figs. 2-5). The diagnosis of Brachypelma (Perez-Miles et al. 1996) shows that this genus does not have retrolateral scopulae on femur IV We ex- amined the two species of Brachypelmides, and we found no scopulae on the retrolateral face of the femur IV. We found plumose hairs in both genera. The patch of plumose hairs mentioned for B. klaasi as a new character separating these genera is not mentioned in B. ruhnaui (Schmidt & Krause 1994; Schmidt 1997); and we could not find it in this species, so this would only be a characteristic that dis- tinguishes B. klaasi. Another characteristic that Brachypelmides klaasi shares with the 200 THE JOURNAL OF ARACHNOLOGY species of Brachypelma is that it is very pop- ular in the pet trade, but Brachypelmides is not listed in appendix II of CITES. Although, the distribution and morphology likely provide strong evidence that Brachy- pelmides and Brachypelma are one and the same genus, a revision of all the species of these genera, not just those from Mexico, us- ing cladistic analysis will provide a strong ba- sis for placing the two species of Brachypel- mides in the genus Brachypelma. ACKNOWLEDGMENTS We thank Dr. Norman I. Platnick, American Museum of Natural History; Dr. Charles E. Griswold, California Academy of Sciences; and Dr. Tila M. Perez, Laboratorio de Acar- ologia, Institute de Biologia, UNAM for giv- ing us the facilities to study the collections of tarantulas of which they are in charge. We are grateful to Graham Floater for helping with the English translation. We are specially thankful to Dr. Anita Hoffmann for encour- aging us to write this paper. Thanks to DGA- PA, UNAM for the financial support grant IN- 217397. LITERATURE CITED Baerg, W.L. 1958. The Tarantula. Univ. of Kansas Press, Lawrence. 88 pp. Cambridge, EO.P. 1897. Arachnida-Araneida. Pp. 1-40, In Biologia Centrali-Americana. (ED. Godman & O. Salvin, ed.) London, Vol. 2. Locht, A., I. Vazquez & M. Yanez. 1998. Las “Ta- rantulas” (Araneae, Theraphosidae) de la Esta- cion de Biologia Chamela, Jalisco. Sociedad Mexicana de Entomologia. Memorias del XXXIII Congreso Nacional de Entomologia. Pp. 47-49. Perez-Miles, E 1992. Analisis cladistico preliminar de la subfamilia Theraphosinae (Araneae, Ther- aphosidae). Bol. Soc. Zool. Uruguay, (2" Epoca). Perez-Miles, E, S.M. Lucas, PI. da Silva, Jr. & R. Bertani. 1996. Systematic revision and cladistic analysis of Theraphosinae (Araneae, Theraphos- idae). Mygalomorph, 1:33-68. Raven, R.J. 1985. The spider infraorder Mygalo- morphae (Araneae): Cladistics and systematics. Bull. American Mus. Nat. Hist., 182:1-180. Schmidt, G. 1992. Brachypelma auratum sp. n., die so genannte Hochlandform von Brachypelma smithi (Araneida: Theraphosidae: Theraphosi- nae). Arach. Anz. 3(8):9-14. Schmidt, G. 1993. Vogelspinnen. Landbuch Ver- lag. Miinchen. 146 pp. Schmidt, G. & P. Klaas. 1994. Eine neue Brachy- pc/ma-Spezies aus Mexico Brachypelma boeh- mei sp. n. (Araneida: Theraphosidae: Theraphos- inae). Arach. Mag., 7(2):7-15. Schmidt, G. & R.H. Krause. 1994. Eine neue Vo- gelspinnen-Spezies aus Mexico, Brachypelmides klaasi sp. n. (Araneida, Theraphosidae, Thera- phosinae). Stu. Neotrop. Fauna and Environ., 29(1):7-10. Schmidt, G. 1997. zweiic Brachypelmides- Axi aus Mexico: Brachypelmides ruhnaui n. sp. (Arachnida: Araneae: Theraphosidae: Theraphos- inae). Entomol. Z. (Essen), 107(5):205-208. Smith, A. 1993. A New Mygalomorph Spider From Mexico {Brachypelma, Theraphosidae, Arachnida) Brachypelma baumgarteni n. sp. British Tarantula Soc. J., 8(4): 14-19. Smith, A. 1994. Tarantula Spiders: Tarantulas of the USA and Mexico. Fitzgerald Publishing, London. 196 pp. Valerio, C. 1980. Aranas Terafosidas de Costa Rica (Araneae, Theraphosidae). 1. Sericopelma y Bra- chypelma. Brenesia, 18:259-288. Yanez, M. & A. Locht. 1998. Ensayos de apaream- iento inducido en Brachypelma klaasi (Schmidt & Krause 1994) (Araneae, Theraphosidae). Soc. Mex. Entomol., Mem. XXXIII Congr. Nac. En- tomol., Pp. 37-41. Manuscript received 1 May 1998, revised 30 March 1999. 1999. The Journal of Arachnology 27:201-204 COMMON GROUND-LIVING SPIDERS IN OLD TAIGA FORESTS OF FINLAND Seppo Koponen: Zoological Museum, University of Turku, FIN-20014 Turku, Finland ABSTRACT. Spiders living on the forest floor in six old taiga forests were studied using pitfall traps in 1994 (in Suomussalmi) and 1995 (in Puolanka), central-eastern Finland, ca. 65° N. Seventy-seven species belonging to eleven families were caught. Linyphiidae (s. lat.) dominated both in species and individual numbers. The most common species were Lepthyphantes alacris, Agyneta ramosa, Lepthy- phantes antroniensis, Centromerus arcanus and Agyneta subtilis. The fauna found is, in general, typical of old Finnish boreal forests. The spider fauna of the boreal (taiga) for- ests in Finland has been studied by many au- thors. Basic studies were carried out by Huhta (1965, 1971). Investigations on spiders in old, primeval forests of Finland include, e.g., pa- Figure 1. — The study areas in central-eastern Finland. 1, Suomussalmi; 2, Puolanka. pers by Palmgren & Bistrom (1979), Bistrom & Vaisanen (1988), Vaisanen & Bistrom (1990), Niemela et al. (1994) and Pajunen et al. (1995). These are all from more southern areas of Finland. The spiders of taiga forests in central-east- ern Finland have not been studied. The aim of this brief paper is to present the abundant spe- cies (of the early-midsummer period) on the floor in six old taiga forests. Some compari- sons with previous studies will also be made. These old forests are in the interests of the pulp and paper industry, and, on the other hand, there are plans to protect these areas. This study was supported by the Kainuu Park Area of the Finnish Forest and Park Service in order to provide some basic data for plan- ning the use of these old forests. The data from the research in Suomussalmi have been partly published (Koponen 1995). Table 1. — Percentage of Linyphiidae (s.lat.) of all specimens (A) and of all species caught (B). Site A B Suomussalmi: Luolakangas 95.1 85.0 Likoaho 93.3 72.9 Heinavaara 87.3 71.2 Puolanka: Paijakka 97.4 90.5 Kuirivaara 95.3 73.1 Siikavaara 84.1 74.2 201 202 THE JOURNAL OF ARACHNOLOGY Table 2. — Percentage of the ten most abundant species at Suomussalmi sites, 1994. Site Percent Luolakangas Lepthyphantes alacris (Blackwall 1853) 22.4 L. antroniensis Schenkel 1933 21.7 Centromerus arcanus (O.P.-Cambridge 1873) 11.1 Diplocentria bidentata (Emerton 1882) 7.5 Agyneta ramosa Jackson 1912 5.3 Lepthyphantes tenebricola (Wider 1834) 4.6 Latithorax latus (Holm 1939) 3.1 Hilaira herniosa (Thorell 1875) 2.7 Agyneta subtilis (O.P.-Cambridge 1863) 2.6 Robertas lividus (Blackwall 1836) 2.4 Likoaho Agyneta subtilis (O.P.-Cambridge 1863) 18.7 Lepthyphantes antroniensis Schenkel 1933 18.7 Agyneta conigera (O.P.-Cambridge 1863) 7.9 A. ramosa Jackson 1912 7.6 Lapthyphantes alacris (Blackwall 1853) 6.5 Centromerus arcanus (O.P.-Cambridge 1873) 6.1 Robertas lividus (Blackwall 1836) 5.4 Lepthyphantes tenebricola (Wider 1834) 5.0 Alopecosa pinetorum (Thorell 1856) 1.6 Walckenaeria dysderoides (Wider 1834) 1.5 Heinavaara Agyneta ramosa Jackson 1912 20.8 Lepthyphantes antroniensis Schenkel 1933 13.8 Agyneta subtilis (O.P.-Cambridge 1863) 11.2 A. conigera (O.P.-Cambridge 1863) 10.3 Lepthyphantes alacris (Blackwall 1853) 9.9 Centromerus arcanus (O.P.-Cambridge 1873) 7.4 Lepthyphantes tenebricola (Wider 1834) 5.5 Robertas lividus (Blackwall 1836) 4.7 Diplocentria bidentata (Emerton 1882) 3.6 Hilaira herniosa (Thorell 1875) 1.2 STUDY AREA AND METHODS The study areas are old, more or less natural primeval forests, surrounded by cutting areas and by young planted tree formations. The majority of large pines {Pinus sylvestris) and spruces (Picea abies) have a diameter of 35- -45 cm. Dead standing and ground-lying trees are not very common. Field layer is mainly dominated by Vaccinium vitis-idaea and V. myrtillus, and the ground layer by the mosses of genera Pleurozium, Dicranum and Hylo- comium. Elevation of the study sites varies be- tween 160 and 270 m. There were three study forests situating in the northern boreal forest zone (northern tai- ga) in Suomussalmi (1994) and three in Puo- lanka (1995), as follows (Fig. 1): Suomussab mi (64°45'N, 29°40'E): - Luolakangas: spruce-dominated, mosaic type (diversified) forest; - Likoaho: relatively dry, pine-domi- nated forest; - Heinavaara: pine-dominated, more moist than the two previous sites; Puo~ lanka (65°N, 28°E): - Paljakka: spruce-domi- nated forest; - Kuirivaara: spruce-dominated, more moist than the two other sites in Puo- lanka; - Siikavaara: dry, pine-dominated mixed forest. PitfaU trapping periods were 13 June-21 July 1994 in Suomussalmi and 14 June-2 Au- gust 1995 in Puolanka. The traps were plastic cups (mouth diameter 65 mm). Ethylene gly- col with detergent was used as the preserva- tion liquid, and the traps were provided with covers agaist rain and litter. Altogether, about 5600 identifiable spider specimens were col- lected. The material has been deposited in the Zoological Museum, University of Turku. The usefulness of pitfall technique in spider studies has been discussed by many authors (e.g., Lowrie 1985). As pitfall data are not in- KOPONEN— GROUND-LIVING SPIDERS IN OLD TAIGA FORESTS 203 Table 3. — Percentage of the ten most abundant species at Puolanka sites, 1995. Site Percent Paljakka Lepthyphantes alacris (Blackwall 1853) 33.1 Centromerus arcanus (O.P.-Cambridge 1873) 22.3 Agyneta subtilis (O.P.-Cambridge 1863) 7.8 Macrargus rufus (Wider 1834) 7.4 Agyneta ramosa Jackson 1912 5.6 Lepthyphantes antroniensis Schenkel 1933 5.2 Diplocentria bidentata (Emerton 1882) 3.4 Hilaira herniosa (Thorell 1875) 3.0 Cryphoeca silvicola (C.L. Koch 1834) 2.2 Walckenaeria nudipalpis (Westring 1851) 1.9 Kuirivaara Lepthyphantes alacris (Blackwall 1853) 44.1 Centromerus arcanus (O.P.-Cambridge 1873) 19.3 Agyneta subtilis (O.P.-Cambridge 1863) 5.0 Macrargus rufus (Wider 1834) 4.0 Agyneta ramosa Jackson 1912 3.7 Asthenargus paganus (Simon 1884) 3.1 Lepthyphantes antroniensis Schenkel 1933 3.0 L. tenebricola (Wider 1834) 3.0 Pardosa lugubris (Walckenaer 1802) 1.9 Walckenaeria nudipalpis (Westring 1851) 1.6 Siikavaara Lepthyphantes alacris (Blackwall 1853) 24.5 Agyneta subtilis (O.P.-Cambridge 1863) 15.3 A. ramosa Jackson 1912 13.8 Zornella cultrigera (L. Koch 1879) 10.7 Pardosa lugubris (Walckenaer 1802) 10.4 Centromerus arcanus (O.P.-Cambridge 1873) 9.2 Walckenaeria nudipalpis (Westring 1851) 4.0 Lepthyphantes tenebricola (Wider 1834) 2.0 Agyneta cauta (O.P.-Cambridge 1902) 1.7 Lepthyphantes antroniensis Schenkel 1933 1.4 dicating real population densities, percentages (not individual numbers) are used here when comparing the sites. RESULTS Altogether 77 species were collected. Lin- yphiidae (s. lat.) clearly dominated in terms of both species and individual numbers (Table 1). This is typical of old, closed and shady forests. Other marked families were Lycosi- dae, Theridiidae and Gnaphosidae. The ten most abundant species at each site are shown in Tables 2 and 3. These lists in- clude 14 and 15 species at Suomussalmi and Puolanka sites, respectively. The dominant species in all forests in Puolanka was Lepthy- phantes alacris, in Suomussalmi the domi- nants included L. alacris, L. antroniensis, Agyneta ramosa and A. subtilis. Non-liny- phiids among the abundant species were Par- dosa lugubris, Robertas lividus, Alopecosa pi- netorum and Cryphoeca silvicola. In Suomussalmi, three Lepthyphantes (L. antroniensis, L. alacris, L. tenebricola) and three Agyneta (A. ramosa, A. subtilis, A. con- igera) species accounted for 65% of the total material. In Puolanka, Lepthyphantes alacris, Centromerus arcanus, Agyneta subtilis and A. ramosa formed 68% of the total material. Only 14 of the 77 species caught were found in all studied six forests (Table 4). The most common (and evenly occurring) of these linyphiid species were Lepthyphantes alacris, Agyneta ramosa, Lepthyphantes antroniensis, Centromerus arcanus and Agyneta subtilis. Also the following species were found at all sites but in smaller numbers: Macrargus ru- fus, Walckenaeria nudipalpis, Diplocentria bi- dentata, Tapinocyba pallens (O.P.-Cambridge 1872), Lepthyphantes tenebricola, Porrhom- 204 THE JOURNAL OF ARACHNOLOGY Table 4. — Species found at all six forest sites; average rank = mean of species’ abundance rank (e.g., Lepthyphantes alacris: 1st, 5th, 5th, 1st, 1st, 1st = 2.3). Species Average rank Lepthyphantes alacris 2.3 Agyneta ramosa 3.8 Lepthyphantes antroniensis 4.0 Centromerus arcanus 4.2 Agyneta subtilis 5.5 Macrargus rufus approx. 12 Walckenaeria nudipalpis 13 Diplocentria bidentata 15 Tapinocyba pallens 16 Lepthyphantes tenebricola 18 Porrhomma pallidum 20 Zornella cultrigera 22 Walckenaeria cuspidata 23 Hilaira herniosa 25 ma pallidum Jackson 1913, Zornella cultri- gera, Walckenaeria cuspidata Blackwall 1833 and Hilaira herniosa. DISCUSSION The fauna found is relatively typical of Finnish boreal coniferous forests, i.e., taiga. Many of the abundant species have also been observed in old forests in previous studies in southern and central Finland. The northern lo- cation (ca. 65°N) of the study areas resulted in the occurrence of several northern species, along with the absence of some species with a more southern range. Vaisanen & Bistrom (1990) listed the eight most abundant spiders found (however, collected with dry funnels) at Saarijarvi (62°50'N; about 300 km SW of the present study area) in central Finland. Of these eight species, only Centromerus arcanus was both abundant and common, six other species were found in smaller numbers and one was absent in the present material. Some of the present abundant species have northern general range in Finland being rare or absent in earlier studies on old forest spi- ders carried out in more southern areas of southern or central Finland (e.g., Palmgren & Bistrom 1979; Bistrom & Vaisanen 1988; Vaisanen & Bistrom 1990; Niemela et al. 1994; Pajunen et al. 1995). These include, e.g., Lepthyphantes antroniensis, Latithorax latus, Zornella cultrigera and Hilaira hernio- sa. ACKNOWLEDGMENTS I am grateful to Anja Finne, Director of the Kainuu Park Area of the Finnish Forest and Park Service, for help and support in organiz- ing this work. LITERATURE CITED Bistrom, O. & R. Vaisanen. 1988. Ancient-forest invertebrates of the Pyhan-Hakki National Park in Central Finland. Acta Zool. Fennici, 185:1- 69. Huhta, V. 1965. Ecology of spiders in the soil and litter of Finnish forests. Ann. Zool. Fennici, 2: 260-308. Huhta, V. 1971. Succession in the spider commu- nities of the forest floor after clear-cutting and prescribed burning. Ann. Zool. Fennici, 8:483- 542. Koponen, S. 1995. Ground-living spiders (Ara- neae) of old forests in eastern Finland. Memo- randa Soc. Fauna Flora Fennica, 71:57-62. Lowrie, D.C. 1985. Preliminary survey of wan- dering spiders of a mixed coniferous forest. J. ArachnoL, 13:97-110. Niemela, J., T. Pajunen, Y. Haila, P. Punttila & E. Halme. 1994. Seasonal activity of boreal forest- floor spiders (Araneae). J. ArachnoL, 22:23-31. Pajunen, T, Y. Haila, E. Halme, J. Niemela & P. Punttila. 1995. Ground-dwelling spiders (Arach- nida, Araneae) in fragmented old forests and sur- rounding managed forests in southern Finland. Ecography, 18:67-72. Palmgren, P & O. Bistrom. 1979. Populations of Araneae (Arachnoidea) and Staphylinidae (Co- leoptera) on the floor of a primeval forest in Mantyharju, southern Finland. Ann. Zool. Fen- nici, 16:177-182. Vaisanen, R. & O. Bistrom. 1990. Boreal forest spiders and the preservation of biotic diversity: results from Finnish primeval forests. Acta Zool. Fennici, 190:373-378. Manuscript received 1 May 1998, revised 25 Sep- tember 1998. 1999. The Journal of Arachnology 27:205-210 ABUNDANCE AND PHENOLOGY OF SCHIZOMIDA (ARACHNIDA) FROM A PRIMARY UPLAND FOREST IN CENTRAL AMAZONIA J. Adis\ J. ReddelF, J. Cokendolpher^ and J.W. de Morais"^: 'Max-Planck- Institute for Limnology, Tropical Ecology Working Group, Postfach 165, D-24302 Plon, Germany; ^Texas Memorial Museum, University of Texas, Austin, Texas 78705, USA; ^2007 29th St., Lubbock, Texas 79411, USA; '^Instituto Nacional de Pesquisas da Amazonia (INPA), C.R 478, 69.011-970 Manaus, AM, Brazil ABSTRACT. There were 193 schizomids (hubbardids) collected from the soil (0-7 cm depth) during a 12 month study of a primary upland forest (37.5 ± 16.8 ind/mVmonth) near Manaus. They were represented by Surazomus brasiliensis (ICraus 1967) and an undescribed species of a new genus (96% and 4% of the total catch, respectively). About 68% of all specimens of S. brasiliensis inhabited the organic soil layer (0-3.5 cm depth) where monthly catches of juveniles were positively correlated with soil tem- perature. Females were twice as abundant as males. The lack of a distinct reproductive period and the presence of juveniles (in particular the first nymphal instar) and adults (both sexes) throughout the year indicate a plurivoltine mode of life. Few specimens were caught on the soil surface, and none were on tree trunks or in the canopy. Abundance of S. brasiliensis is compared to that of the Palpigradi (micro- whip scorpions) and Thelyphonida (vinegaroons) from the same study site. The order Schizomida is comprised by about 180 described species. Few studies have been conducted on their ecology and biology. Schizomids are considered to be hygrophi- lous, photophobic, hemiedaphic inhabitants of soils, particularly in the tropics and subtrop- ics. Some species are termitophiles, myrme- cophiles, nidicoles or troglobites. {cf. Moritz 1993; Humphreys et al. 1989; Reddell & Cok- endolpher 1995; Rowland 1972). In Central Amazonian forests, schizomids represent less than 1% of the soil arthropods which mostly inhabit the top 7 cm (cf. Adis et al 1987, 1989). Our material, obtained in a primary upland forest over a 12 month pe- riod, represents the very first contribution on the abundance and phenology of a Neotropical schizomid species: Surazomus brasiliensis (Kraus, in Kraus & Beck 1967). STUDY AREA AND METHODS Schizomids were collected between 1981- 1983 in the course of ecological studies on Central Amazonian arthropods from a previ- ously investigated and fully-described prima- ry upland forest at Reserva Florestal A. Ducke (= Reserva Ducke; 2°55'S, 59°59'W; Penny & Arias 1982). The reserve is located on the Manaus-Itacoatiara highway (AM-010), about 26 km from Manaus. The forest is subject to a rainy season (December-May: average pre- cipitation 1550 mm; 258.9 ± 36.8 mm/month) and a “dry” season (June-November: average precipitation 550 mm; 91.8 ± 43.8 mm/month and each month with some rain events; Ri- beiro & Adis 1984). The yellow latosol (= ferrasol in Jordan 1984) of the primary upland forest had a 2-3 cm thick humus layer, inter- spersed with fine roots, and a thin surface cov- ering of leaf-litter. One ground photo-eclector (emergence trap) and one arboreal photo-eclector for trunk ascents (funnel trap) were installed in the for- est (cf Adis & Schubart 1984) and remained there from December 1981 to December 1982. Distribution of schizomids in the soil was studied between September 1982 and August 1983 (Morals 1985). Twelve soil samples were taken once a month every 2 m along a randomly selected transect. The split corer, composed of a steel cylinder with lateral hing- es (diameter 21 cm, length 33 cm), was driven into the soil by a mallet. Each sample of 7 cm depth was then divided into two subsamples of 3.5 cm each for extraction of animals, fol- lowing a modified Kempson method (Adis 205 206 THE JOURNAL OF ARACHNOLOGY E E £ q ‘>5 CO [a ‘5 0) Surazomus brasiliensis (Schizomida) — Eukoenenia janetscheki (Palpigradi) Thelyphonellus amazonicus (Thelyphonida) Figure 1. — Distribution of S. brasiliensis (Kraus 1967) (Schizomida), E. janetscheki Conde 1993 (Pal- pigradi) and T. amazonicus (Butler 1872) (Thelyphonida) in the soil. Samples taken monthly at 0-7 cm depth between September 1982-August 1983 in a primary upland forest near Manaus. (N = total number of specimens). Total precipitation per month given between sampling dates (= at the end of each month). The low rainfall observed in early 1983 was due to a strong El Nino-event {cf. Adis & Latif 1996). 1987). The combined area of the 12 samples represented 0.42 m^. Calculated average abun- dances per m^ are given with sample standard deviation. The monthly collection data of schizomids from the two soil layers in relation to changing abiotic conditions (precipitation, temperature and humidity of the air near the forest floor; moisture content, temperature and pH of the soil) were statistically evaluated with a linear, parametric correlation test (Cav- alli-Sforza 1972) using the original field data (Morals 1985). In addition, the presence of schizomids in tree crowns of the primary up- land forest was tested by fogging canopies with pyrethrum during the dry and rainy sea- sons (August 1991-July 1994; c/ Adis et al 1997a). All Schizomida sampled were classified as juveniles, subadults and adults (males and fe- males, respectively; cf. Reddell & Cokendol- pher 1995). Juveniles were tentatively as- signed to three size classes, based on the length of the cephalo thorax. The size classes presumably represent the three development stages in nymphs, apart from the subadult stage {cf Brach 1976; Dumitrescu 1973; Row- land 1972). Voucher specimens have been deposited at the Systematic Entomology Collections of the Instituto Nacional de Pesquisas da Amazonia (INPA) in Manaus, Brazil, at the Texas Me- morial Museum, Austin, Texas and at the Mu- seum d’histoire naturelle in Geneve, Switzer- land. RESULTS Schizomida obtained from the primary up- land forest at Reserva Ducke were represented ADIS ET AL.— SCHIZOMIDA OF CENTRAL AMAZONIA 207 J = juveniles S = subadults A = adults 0 - 3.5 3.5 - 7.0 Soil Depth (cm) Figure 2. — Distribution of Surazomus brasiliensis in the soil according to soil depth, and percentage of developmental stages in a primary upland forest near Manaus. (Total catch = 100%) Samples taken monthly at 0-3.5 and 3.5-7 cm depths over a 12 month period. (N = total number of specimens). by Surazomus brasiliensis (body length ^4.3 mm without flagellum; cf. Kraus & Beck 1967; Reddell & Cokendolpher 1995) and an undescribed species (Reddell & Cokendolpher 1999) of a new genus (96% and 4% of the total catch, respectively). A total of 193 schizomids was collected. Out of these, 99% could be identified to their developmental stages. Schizomids were most- ly found in the soil and never caught on tree trunks or in the canopy. Only three specimens (adults of S. brasiliensis), were captured in pitfall traps inside the ground photo-eclector, while active on the soil surface. Schizomids represented 0.4% of the total arthropods ex- tracted from soil samples within 12 months if Acari and Collembola are omitted (Morais 1985) and < 0.1% when they are included (Adis unpubl. data). The abundance of Schi- zomida in 0-7 cm soil depth was higher than that of the Palpigradi (455 vs. 351 ind/m^), whereas abundance of the Thelyphonida (7 ind/m^) was much lower (corrected data of Fig. 1 in Adis et al 1997b). This is also con- sistent for the dominant species in each order (Fig. 1). An average abundance of 37.5 ± 16.8 schizomids/mVmonth was recorded in 0-7 cm soil depth {S. brasiliensis: 36.1 ± 16.8 ind/ mVmonth; new genus, new species: 1.4 ± 1.7 ind/mVmonth). Most specimens of S. brasiliensis inhabited the organic soil layer (Fig. 2: 0-3.5 cm) and a few (32%) the mineral subsoil (3. 5-7.0 cm depth). About 70% (25.0 ± 13.7 ind/mV month) of the total catch was represented by juveniles (Fig. 2), and 15% each by subadults and adults (5.5 ±4.1 and 5.5 ± 4.6 ind/mV month, respectively). Sex ratio of adult males to females was 1:2.4 but instars of juveniles could not be sexed. No significant difference was found for the cephalothorax length be- tween subadult males and subadult females {x^ test: P < 0.05). The monthly abundance of juveniles in S. brasiliensis obtained from the organic soil layer (0-3.5 cm depth) was positively corre- lated with soil temperature (17.6-26.6 °C; av- erage 23.8 ± 2.4 °C) (total catch: r = +0.77097, P < 0.01; n = 12). The total catch- es of specimens obtained during the dry sea- son and the rainy season were similar: 48% versus 52%. However, there was no distinct 208 THE JOURNAL OF ARACHNOLOGY Figure 3. — Temporal occurrence of developmen- tal stages of Surazomus brasiliensis in the soil (n/ m^ in 0-7 cm depth) of a primary upland forest near Manaus. reproductive period because juveniles, in par- ticular the first nymphal instar, as well as adults of both sexes, occurred throughout the year (Figs. 3, 4). These results indicate a plu- rivoltine mode of life. DISCUSSION The low number of schizomids in samples from the ground photo-eclector at Reserva Ducke indicates that these two species were rarely active on the soil surface. This conclu- sion is supported by another study at the re- serve, in which apparently no schizomids were collected in 20 baited pitfall traps and in three or more ground photo-eclectors during a sample period of 12 months (Penny & Arias 1982). The depth to which schizomids occur in the soil of the Central Amazonian upland forests is unknown. Our studies in various forest types near Manaus (Adis et al. 1987, 1989, 1997b, c) revealed their presence to a soil depth of 14 cm. The vertical distribution of S. brasiliensis is influenced by soil temperature in that catch numbers increased with rising temperatures. Schizomids are easily mistaken for young spiders, particularly if their flagellum or front legs are missing. This might explain their “ab- sence” in other studies on the Neotropical ar- thropod fauna in 0-30 cm soil depth (e.g., Harada & Bandeira 1994a, b; Macambira 1997; Serafino & Merino 1978). Parthenogenesis has been reported for sev- eral schizomid species (Reddell & Cokendol- pher 1995). In S. brasiliensis both sexes were present. However, more than twice as many females as males were captured. This was also observed in the euedaphic palpigrad Euko- enenia janetscheki Conde 1993 from the same study site (Adis et al. 1997b). Predominance of females assures the continuation of a spe- cies. This was also found for three species of Symphyla from the primary upland forest at Reserva Ducke and from a secondary upland forest at Rio Taruma Mirim near Manaus where the number of females was 2-4 X high- er than of males (Adis et al. 1997c). Surazomus brasiliensis is the only Amazo- nian schizomid species for which observations on the biology are available (Kraus & Beck 1967). Beck (1968) observed that animals in whitesand soils at Reserva Ducke predomi- nantly feed on Collembola and Symphyla. Prey is searched by actively running around in a jerky manner {cf. Humphreys et al. 1989; Sturm 1973) and exploring the surroundings with the long and highly mobile front legs, which also serve as tactile instruments. Once the prey is perceived, the pedipalps are used to seize and transfer it to the chelicerae where it is cut during longitudinal and vertical move- ments. After ingestion and deposition of the remains on the soil, animals often groom themselves, particularly the long front legs and the flagellar region which is reached by folding the abdomen over the cephalothorax (= opisthosoma and prosoma, respectively, in Beck 1968). The grooming procedure is con- cluded by cleaning the pedipalps and the che- licerae. A similar behavior was reported for ADIS ET AL.— SCHIZOMIDA OF CENTRAL AMAZONIA 209 ■ 0.50 -0.69 mm 110.70 -0.89 mm 00.90 -1.10 mm Figure 4. — Occurrence of three size classes in juveniles of Surazomus brasiliensis (based on the length of the cephalothorax). Specimens, obtained from 0-7 cm soil depth, presumably represent the three de- velopmental stages in nymphs, apart from the subadult stage. (Number of specimens examined per month = 100%; 122 (97.6%) out of 125 juvenile specimens measurable). two other schizomids: Draculoides vinei (Har- vey 1988) from caves in western Australia (Humphreys et al. 1989) and for Surazomus sturmi (Kraus 1957) from the surroundings of Bogota, Colombia (Sturm 1973). According to Beck, the mating behavior and indirect trans- fer of the spermatophore in S. brasiliensis is similar to that observed in S. sturmi (Kraus & Beck 1967; Sturm 1958, 1973). ACKNOWLEDGMENTS This study resulted from a cooperation be- tween the National Institute for Amazonian Research (INPA) at Manaus, Brazil and the Tropical Ecology Working Group at the Max- Planck-Institute (MPI) for Limnology in Plon, Germany (Projeto INPA/Max-Planck). We wish to acknowledge the valuable support re- ceived from PD Dr. W.J. Junk, Head of the Tropical Ecology Working Group. Dr. J. Mark Rowland, University of New Mexico, Albu- querque, and Prof. Dr. Otto Kraus, University of Hamburg, Germany are thanked for valu- able comments on the manuscript. Berit Han- sen (MPI Plon) is thanked for making the drawings. Dr. Johann Bauer, MPI for Bio- chemistry (Martinsried, Germany) assisted us with literature. LITERATURE CITED Adis, J. 1987. Extraction of arthropods from Neo- tropical soils with a modified Kempson appara- tus. J. Trop. Ecol., 3(2): 13 1-138. Adis, J. & M. Latif. 1996. Amazonian arthropods react to El Nino. Biotropica, 28(3):403-408. Adis, J. & H.O.R. Schubart. 1984. Ecological re- search on arthropods in Central Amazonian for- est ecosystems with recommendations for study procedures. Pp. 1 1 1-144. In Trends in Ecological Research for the 1980s. NATO Conference Se- ries, Series I: Ecology, Vol. 7. (J.H. Cooley & EB. Golley, eds.). Plenum Press, New York, London. 344 pp. Adis, J., J.W. de Morais & H.G. de Mesquita. 1987. Vertical distribution and abundance of arthropods in the soil of a Neotropical secondary forest dur- ing the rainy season. Stud. Neotrop. Fauna & En- viron., 22(4): 189-197. Adis, J., W. Paarmann, C.R. da Fonseca & J.A. Ra- fael. 1997a. Knock-down efficiency of natural pyrethrum and survival rate of arthropods ob- tained by canopy fogging in Central Amazonia. Pp. 67-81. In Canopy Arthropods. (N.E. Stork, J. Adis & R.K. Didham, eds.). Chapman & Hall, London. 576 pp. Adis, J., E.E Ribeiro, J.W. de Morais & E.T.S. Cav- alcante. 1989. Vertical distribution and abun- dance of arthropods from white sand soil of a Neotropical campinarana forest during the dry 210 THE JOURNAL OF ARACHNOLOGY season. Stud. Neotrop. Fauna & Environ., 24(4): 201-211. Adis, J., U. Scheller, J.W. de Morais & J.M.G. Rod- rigues. 1997b. On the abundance and phenology of Palpigradi (Arachnida) from Central Amazo- nian upland forests. J. ArachnoL, 25:326-332. Adis, J., U. Scheller, J.W. de Morais, C. Rochus & J.M.G. Rodrigues. 1997c. Symphyla (Myriapo- da) from Amazonian non-flooded upland forests and their adaptations to inundation forests. In Many-legged animals. A collection of papers on Myriapoda and Onychophora. (H. Enghoff, ed.). Entomol. Scandinavica SuppL, 51:115-119. Beck, L. 1968. Aus den Regenwaldem am Ama- zonas II. Natur und Museum, 98(2):71-80. Brach, V. 1976. Development of the whip scorpion Schizomus floridanus, with notes on behaviour and laboratory culture. Bull. South. California Sci., 74:270-274. Cavalli-Sforza, L. 1972. Grundzuge biologisch- medizinischer Statistik. G. Fischer, Stuttgart. 212 pp. Dumitrescu, M. 1973. Deux especes nouvelles du genre Schizomus (Schizomida), trouvees a Cuba. Resultats des expeditions biospeleologiques cu- bano-rumaines a Cuba, 1:279-292. Harada, A.Y. & A.G. Bandeira. 1994a. Estratifi- ca9ao e densidade de invertebrados em solo ar- enoso sob floresta e plantios arboreos na Ama- zonia central durante a esta^ao seca. Acta Amazonica, 24(1/2): 103-1 18. Harada, A.Y. & A.G. Bandeira. 1994b. Estratifi- cagao e densidade de invertebrados em solo ar- giloso sob floresta e plantios arboreos na Ama- zonia central durante a estagao seca. Bol. Mus. Par. Emflio Goeldi, ser. Zool., 10(2):235-251. Humphreys, W.F., M. Adams & B. Vine. 1989. The biology of Schizomus vinei (Chelicerata: Schi- zomida) in the caves of Cape Range, Western Australia. J. Zool., London, 217:177-201. Jordan, C.F. 1984. Soils of the Amazon rainforest. Pp. 83-105, In Amazonia (G.T. Prance & T.E. Lovejoy, eds.). Pergamon Press, Oxford. 442 pp. Kraus, O. & L. Beck. 1967. Taxonomic und Biol- ogic von Trithyreus brasiliensis n. sp. (Arach.: Pedipalpi: Schizopeltidia). Senckenbergiana Biol., 48:401-405. Macambira, M.L.J. 1997. A fauna de invertebrados do solo. Pp. 355-360. In Caxiuana (P.L.B. Lisoa, ed.). Museu Paraense E. Goeldi, Belem, Brazil. 446 pp. Morais, J.W. de. 1985. Abundancia e distribuigao vertical de Arthropoda do solo numa floresta pri- mMa nao inundada. M. Sc. thesis, CNPq/INPA/ FUA. Manaus, Brazil. 92 pp. Moritz, M. 1993. 3. Ordnung Schizomida. Pp. 158-164. In Lehrbuch der Speziellen Zoologie. Bd. I: Wirbellose Tiere. 4. Teil: Arthropoda (ohne Insekten). (H,-E. Gruner, ed.). G. Fischer Verlag. 1279 pp. Penny, N.D. & J. Arias. 1982. Insects of An Am- azon Forest. Columbia Univ. Press, New York. 269 pp. Reddell, J. & J. Cokendolpher. 1995. Catalogue, bibliography, and generic revision of the order Schizomida (Arachnida). Speleol. Mono., (Texas Memorial Museum), 4:1-170. Reddell, J. & J. Cokendolpher. 1999. Additional Schizomida (Arachnida) from South America. Stud. Neotrop. Fauna & Environ., In Press. Ribeiro, M. de N.G. & J. Adis 1984. Local rainfall variability — a potential bias for bioecological studies in the Central Amazon. Acta Amazonica, 14(1/2):159-174. Rowland, J.M. 1972. The brooding habits and ear- ly development of Trithyreus pentapeltis (Cook) (Arachnida, Schizomida). Entomol. News, 83(3): 69-74. Serafino, A. & J.F. Merino. 1978. Poblaciones de microartropodos en diferentes suelos de Costa Rica. Rev. Trop. Biol., 26(1):139-15L Sturm, H. 1958. Indirekte Spermatophoreniibertra- gung bei dem Geisselskorpion Trithyreus sturmi Kraus (Schizomidae, Pedipalpi). Naturwiss., 45(6): 142-143. Sturm, H. 1973. Zur Ethologie von Trithyreus stur- mi Kraus (Arachnida, Pedipalpi, Schizopeltidia). Z. TierpsychoL, 33:113-140. Manuscript received 1 May 1998, revised 22 Feb- ruary 1999. 1999. The Journal of Arachnology 27:211-216 RELATIONSHIP OF HABITAT AGE TO PHENOLOGY AMONG GROUND-DWELLING LINYPHHDAE (ARANEAE) IN THE SOUTHEASTERN UNITED STATES Michael L. Draney: Department of Biology, RO. Box 30001, New Mexico State University, Las Cruces, New Mexico 88003 USA D.A. Crossley, Jr.: Institute of Ecology, University of Georgia, Athens, Georgia 30602 USA ABSTRACT. Ground-dwelling Linyphiidae from eight South Carolina inner coastal plain habitats were sampled for one year using pitfall traps. Habitats formed an age gradient, from a field disturbed yearly and pine stands aged 5, 25 and 40 years, to xeric, mesic and hydric hardwoods (50-75 years) and an old- growth forest (200 years). Sixteen of the 55 trapped species were represented in sufficient numbers (n adults > number of sampling periods, 26) to examine patterns of correlation between phenology and habitat distribution. Half of the species are multivoltine, characterized by adults present throughout the year, continuous reproduction, and overlapping generations. Adult abundance of these species peaked during spring through autumn. Other species were univoltine, with adults present briefly, indicating syn- chronous reproduction and non-overlapping generations. Adult abundance of these species always peaked during winter months. This study examines relationships between observed voltinism patterns and char- acteristic habitat (distribution among the habitats) among the 16 most abundant species. Species from older habitats tend to be univoltine, whereas species inhabiting more recently disturbed habitats were more likely to be multivoltine. Stenochronous winter reproduction (univoltines) probably increases survivorship by limiting individuals’ exposure to the harsh conditions of the southeastern summer during vulnerable periods of immaturity and reproduction. This phenological specialization appears optimal in this region except in frequently disturbed habitats, where rapid multivoltine reproduction is most advantageous. Knowledge of cyclic temporal aspects of organisms’ life cycles (phenology) is crucial for understanding population dynamics and community ecology, and lends realism to evo- lutionary and ecological hypotheses (Lieth 1974; Tauber & Tauber 1981). This basic data can be time-consuming to obtain, as it re- quires sampling a population repeatedly throughout the year, and phenology may vary geographically or from one year to the next. Although Linyphiidae is by far the most di- verse spider family in North America (Cod- dington et al. 1990), little is known about the phenology of most North American species. This is particularly true of the ground-dwell- ing forms, whose small size (1-3 mm), high diversity (ca. 800 species in North America), and cryptic microhabitats within litter inter- stices, etc., make them difficult to observe in the field. This paper is part of a research program examining life history variation among ground-dwelling Linyphiidae. As phenology and other life history information is time-con- suming to obtain, our goal is to determine the extent to which more easily obtainable kinds of information (such as the habitats used) are predictive of life history variation among ground-dwelling linyphiids. By understanding this, we may begin to apply information de- rived from careful phenological and labora- tory studies to unstudied taxa as they are en- countered in sampling. Although linyphiids are most diverse in Northern Hemisphere mid-temperate latitudes and increasingly dominate spider assemblages farther north, they are still fairly diverse and form a conspicuous portion of spider assem- blages in more southerly humid temperate re- gions, such as the southeastern Atlantic coast- al plain of the United States (Draney 1997a, b). The present study examines species from the inner coastal plain of South Carolina (ap- proximately 33° N). Pitfall trapping during the course of this re- search yielded 55 species, of which 16 were 211 212 THE JOURNAL OF ARACHNOLOGY judged abundant enough to use in elucidating phenological patterns. Two general phenology ical patterns emerged (Draney 1997a): In about half of the species, the adults are eu- rychronous (present during most of the year); these species appear to be multivoltine, with overlapping generations and continuous repro- duction. For all these species, adults trapped peaked during the warm season. The other species have stenochronous adults (present during only a short time during the year); these species appear to be univoltine in our region, with non-overlapping generations and a winter mating period. Although other phe- nological patterns doubtless exist in our re- gion, these two general patterns of voltinism appear to be very common among ground- dwelling linyphiids in the southeastern US. The objective of the present article is to ex- amine the extent to which these observed vol- tinism patterns are correlated with the distri- bution of individuals of a species among our sampled habitats. It might be expected that habitats differing in time since last distur- bance would favor different life history re- sponses, corresponding to r- and K-selection models (Stearns 1992). Our sampled habitats were specifically selected to provide a wide gradient of habitat age, as measured by time since the soil/litter layer has been significantly disturbed (as by clear-cutting, burning, mow- ing, plowing, etc.). We expect that species oc- curring in younger, more recently disturbed habitats would be more likely to be multivol- tine than species occurring in older, more per- manent habitat types, for reasons discussed more fully in the Results and Discussion sec- tion. METHODS Study sites. — The study areas were located on the Savannah River Site (SRS), a 780 km^ area adjacent to the Savannah River in Aiken, Allendale, and Barnwell Counties, South Car- olina. SRS has been maintained by the US Department of Energy since 1951. Because a primary objective of this study was to deter- mine whether habitat age is related to life his- tory parameters of the linyphiid inhabitants, the primary criterion for site selection was to locate sites that vary widely in the frequency with which the soil stratum is disturbed. In addition, several relatively mature habitats of the same age but with different vegetative communities were chosen. In all, eight areas were selected to represent the major terrestrial habitat types on the inner Atlantic coastal plain, here listed in order from the youngest, most recently disturbed to the oldest, least fre- quently disturbed habitat. For more detailed site information, see Draney (1997a). 1). Old Field: forb-grassland with Opuntia and lichen. Mowed or herbicided annually. 2). Young pines: 5 year-old plantation overgrown with Rubus and Prunus spp. 3). Medium pines: 25 year-old plantation with a sparse understory. 4). Mature Pines: 40 year-old volunteer pine stand with young oak and pine understory. 5). Scrub-Oak/Pines: 50-75 year-old xeric upland oak/pine stand. 6). Upland hardwood: 50-75 year-old mesic oak-hickory stand. 7). Riparian hardwood: 50-75 year-old hydric hardwood stand. 8). Riparian old growth: ca. 200 year- old pines within riparian hardwood stand. Sampling methods. — A study area of about 1 ha was subjectively delimited to rep- resent each selected habitat. A 0.25 ha (50 m X 50 m) plot was randomly located within this area, and 10 pitfall traps were randomly located within each plot. Pitfalls each consist- ed of an 8.5 cm diameter plastic cup buried with the lip flush to the soil surface, housed under a concrete building block (39 X 19 X 9 cm) propped up at one end by a 7 cm brick. Traps contained 4% formalin with a trace of detergent to decrease surface tension. Traps were run continuously for a year (366 days, 1-4 May 1995 to 1-4 May 1996) and emptied at approximately biweekly intervals (11-17 days; mean = 14.0). At each of 26 sampling periods, all traps were collected from each of the eight sites. Contents were washed into ajar and the trap was refilled with formalin solution. Samples were sieved through a 250 jjim mesh sieve to remove the formalin, and stored in 70% ethanol for sub- sequent identification. Nomenclature follows Buckle et al. (1993) and Platnick (1996). Voucher specimens of all species are depos- ited at AMNH and UGCA. Data analysis. — Pitfall counts from each site during each trap period were expressed as number of organisms/ 140 trap-days (10 traps X 14 days), to correct for variable number of days per trap period, and to correct for the 22 traps lost to animal (dog or coyote) distur- bance at the old field site. A constraint of phenological data is that DRANEY & CROSSLEY— LINYPHIID PHENOLOGY AND HABITAT 213 Table 1. — Phenology and habitat indices of 16 ground-dwelling linyphiids from the South Carolina inner coastal plain, n = Number of adults trapped. Is = Index of seasonality. Pm = proportion of adults trapped at modal sampling period. Ih = Index of habitat range. Ha = Habitat age score. See text for explanation. Species are listed by voltinism (see Draney 1997a), and then by numbers of adults trapped. Species n Is Pm Ih Ha Winter active, univoltine Pelecopsidis frontalis (Banks 1904) 199 2.6 0.307 3.64 4.18 unidentified species, cf. Walckenaeria 191 1.7 0.351 2.04 3.59 Walckenaeria Carolina Millidge 1983 110 2.1 0.355 3.66 3.90 Centromerus latidens (Emerton 1882) 110 3.8 0.200 5.20 4.35 Lepthyphantes sabulosus (Keyserling 1886) 51 3.2 0.275 3.88 3.99 Ceraticelus laetabilis (L. Pickard-Cambridge 1874) 39 2.9 0.333 1.48 5.80 Origanates rostratus (Emerton 1882) 33 3.3 0.273 2.84 3.77 Scylaceus pallidus (Emerton 1882) 33 3.3 0.242 4.35 4.73 Ceraticelus alticeps (Fox 1891) 31 1.9 0.419 2.20 4.61 Eurychronous, multivoltine Meioneta sp. n. #1 805 9.5 0.083 2.22 2.48 Meioneta sp. n. #3 563 4.5 0.224 1.95 5.37 Ceratinops crenatus (Emerton 1882) 230 6.3 0.161 1.00 1.00 Meioneta barrowsi Chamberlin & Ivie 1944 72 4.5 0.181 2.91 3.46 Eperigone maculata (Banks 1892) 61 4.8 0.230 3.47 4.61 Erigone autumnalis Emerton 1882 32 5.8 0.125 1.84 1.43 Meioneta micaria (Emerton 1882) 31 3.7 0.226 3.22 3.90 many parameters are dependent on sample size. As the number of individuals sampled increases, the apparent temporal span of the species will tend to increase as well. Param- eters pertaining to distribution of individuals across sampling periods were only calculated on categories (species, or sexes/stages within species) when the category sample size equaled or exceeded the number of sampling periods {n = 26; Table 1). Thus, phenological parameters were calculated only for 16 of the 55 linyphiid species trapped during this study (Draney 1997a). We used several indices to compare the complex temporal and spatial distribution pat- terns of different species and sexes (Table 1). We calculated Pm, the proportion of adults trapped during the modal sampling period (that is, the period during which the maximal number of individuals of that species was trapped) as a simple indicator of each species’ temporal distribution; higher proportions in- dicate more stenochronous populations. We also calculated an index of seasonality (Is; Curtis 1978) for each species. This index uses the proportion of individuals of a species cap- tured in each month to determine how evenly the species is distributed throughout the year. As our sampling periods were biweekly in- stead of monthly, the index was converted to a fraction and then multiplied by twelve to standardize the index to “months”: Is = 12[(l/5:pp/s] summed over all sampling periods, where pj = proportion caught during sampling period j, and s = total number of sampling periods (always 26 in this study). The index varies from 12/s for a highly stenochronous species found only during one sampling period, to 12, for a completely eurychronous species evenly distributed across all sampling periods. A modified index can be used to examine the distribution of spiders across sampled hab- itats. This index of habitat range, Ih, can be used to categorize species as eurytopic (oc- curring in many habitats) or stenotopic (oc- curring in a narrow range of habitat types): Ih = l/2p^ summed over all habitats sampled, where p^ = proportion caught in habitat k. This index varies from 1 for a stenotopic species found in only one habitat, to the number of habitats sampled (in this case, 8). Both of the above indices assume that each sampling period or habitat is sampled equally, an assumption met 214 THE JOURNAL OF ARACHNOLOGY in the present study when the data are stan- dardized to number/ 140 trap-days. Because we are interested in the extent to which phenology and life history patterns may be predicted by habitat distribution, a habitat age score was calculated for each species. This score summarizes the characteristic hab- itat of each species with respect to habitat age, and is essentially a weighted average of a spe- cies’ presence at each habitat, giving more weight to individuals found in older habitats: Pi "" [(Ini) + (2n2) + (SnO + (4n4) + (Sn^) -f (SnJ + (5n7) T (6n8)]/ntot,j, where = Number of adults from old field, scored as 1 (disturbed yearly); n2 number from young pines, scored as 2 (ca. 5 years old); n3 = num- ber from medium pines, scored as 3 (ca. 25 years old); n4 = number from large pines, scored as 4 (ca. 40 years old); n5,n6,n7 = num- ber from scrub-oak/pines, upland hardwoods, and riparian hardwoods, all scored as 5 (hab- itats were forested 40 years ago and so have been forested at least ca. 75 years); ng = num- ber from old growth habitat, scored as 6 (for- ested for ca. 200 years); and ntot^i = sum of individuals from all sites. This species score varies from 1 to 6. For example, species found only at the most frequently disturbed habitat have a score of 1.00. Single factor regression analyses were used to examine whether any habitat distribution indices (Ih or Ha) are related to phonological indices (Pm or Is) among the examined taxa. RESULTS AND DISCUSSION Species characteristically trapped at youn- ger habitats tended to be more eurychronous and species at older habitats tended to be more stenochronous. Species habitat range (Ih) was not correlated with phonological indices, but the age of the habitats occupied by the species was correlated with phenology. Specifically, the adult index of seasonality. Is, was nega- tively correlated with habitat age score, Ha (U = 0.328, P = 0.0204) and percent of adults at the species’ modal date, Pm, was positively correlated with Ha (r^ = 0.345, P = 0.0167; Table 1). These trends did not hold when the sexes were examined independently. This was probably due to the decrease in number of taxa examined: sample size was adequate (n > 25) to examine total adults (males and fe- males combined) of 16 species, but males of only 1 1 species, and females of only five spe- cies. It should be noted that Draney (1997a) found no evidence that the phenology patterns of individual taxa vary among the habitats (Draney 1997a), so it can be assumed that spe- cies are either univoltine or multivoltine with- in this region, regardless of the habitat they occupy. In Figs. 1 and 2, species points in the scat- terplots are labeled as multivoltine or univol- tine, as determined by phenological indices, graphs of age/sex distribution over time, and other supporting evidence (Draney 1997a; Ta- ble 1). Among the species studied, those oc- curring mainly in younger, more frequently disturbed habitats (species with habitat age scores < 3.5) were all multivoltine, whereas species characteristically inhabiting older hab- itats were predominantly, but not exclusively, univoltine. The apparent increased likelihood of find- ing eurychronous, multivoltine species in younger, less permanent habitats is consistent with the advantages which, we postulate, ac- company this phenological strategy. Eury- chronous spiders would be capable of repro- ducing opportunistically when the habitat is favorable or when they arrive at a favorable habitat. While the habitat remains favorable, continuous reproduction allows for individu- als to maximize their instantaneous rate of re- production (and thus, probably, their fitness). Finally, overlapping generations and the mixed age structure resulting from continuous reproduction mean that when the habitat changes, some individuals of the life history stage which is best able to survive by toler- ance or dispersal should already be present. We postulate that these eurychronous species are phenological “generalists” with a flexible strategy that can result in successful repro- duction even in unpredictable or impermanent habitats. It is interesting to note that Merrett’s (1969) phenological study of 90 linyphiid spe- cies found that all eurychronous species were “common aeronauts,” species commonly ob- served or collected ballooning. Stenochronous species, conversely, appear to be phenological “specialists,” finely adapt- ed to completing various life history stages at specific times when conditions are most fa- vorable, or to avoiding unfavorable condi- tions. All stenochronous species we examined are winter-reproducing, which seems to strengthen this specialization hypothesis. For DRANEY & CROSSLEY— LINYPHIID PHENOLOGY AND HABITAT 215 Figures 1-2. — Relationships between age of hab- itats in which linyphiid species were trapped, and indices of phenology. Each circle is data from all adults of one species (Table 1). Closed circles (•) are postulated to be multivoltine and open circles (o), univoltine (Draney 1997a). See text for r^ and P values, and explanations of indices. 1, Negative correlation between habitat age score (Ha) and in- dex of seasonality (Is); 2, Positive correlation be- tween habitat age score (Ha) and proportion of adults trapped at modal sampling period (Pm). three reasons, the mild southeastern US winter may be favorable for reproduction by ground- layer Linyphiidae. First, the cool, moist con- ditions that prevail in winter are probably more favorable both to survivorship of im- matures, which are especially susceptible to desiccation (based on pers. obs., MLD), and to the longevity of adults, which is negatively correlated with temperature and positively correlated with fitness in spiders (Li & Jack- son 1996). Using such a stenochronous strat- egy, only adults would encounter the harsher summer conditions. In contrast, a eurychron- ous population would include many imma- tures during the summer. Second, Collembola, which are a major component of the prey of ground-dwelling linyphiids (Nyffeler & Benz 1988; Nentwig 1980, 1983, 1987; Alderwei- reldt 1994), are most abundant during the cool season (pers. obs. MLD). Third, we believe that both predation and competition for prey resources would be lower in the winter, since most other arthropod groups (including most spiders) are more active during the warm sea- son. For organisms from a lineage that pre- sumably evolved at higher latitudes (liny- phiids are most diverse in mid-latitude temperate regions; van Helsdingen 1983), the southeastern US winter may indeed be a tem- poral “island” of favorable conditions, which may itself select for a stenochronous life cy- cle. It appears that stenochronous winter repro- duction is the predominant strategy for ground-dwelling linyphiids in this region, ex- cept in the younger habitats. We hypothesize that in such frequently disturbed habitats, the advantages of multivoltinism outweigh uni- voltinism’s postulated advantage of avoiding the harsh conditions which ground-dwelling linyphiids would encounter during the south- eastern summer. ACKNOWLEDGMENTS We thank B.E. Taylor for critical reading of earlier drafts of this work, and for providing facilities and guidance to M. Draney during his tenure at the Savannah River Ecology Lab- oratory, South Carolina. We thank M.I. Saar- isto for deternfination of linyphiine material, J. Zujko-Miller for examining new species, cf. Walckenaeria, and the staff at SREL, AMNH and NMNH for much help. This article ben- efited from comments by Jim Berry, Brent Opell and two anonymous reviewers. This re- search was supported by Financial Assistance Award Number DE-FC09-96SR 18546 from the Department of Energy to the University of Georgia Foundation. 216 THE JOURNAL OF ARACHNOLOGY LITERATURE CITED Alderweireldt, M. 1994. Prey selection and prey capture strategies of linyphiid spiders in high- input agricultural fields. Bull. British Arachnol. Soc., 9:300-308. Buckle, D.J., D. Carroll, R.L. Crawford & V.D. Roth. 1993. Linyphiidae of America north of Mexico: Checklists, synonymy, and literature. Version 2.1, July 1994. Available on disk or printed from third author, Washington State Mu- seum, Univ. Washington, Seattle, Washington 98198. 106 pp. Coddington, J.A., S.E Larcher & J.C. Cokendol- pher. 1990. The systematic status of Arachnida, exclusive of Acari, in North America north of Mexico. Pp. 5-20. In Systematics of the North American Insects and Arachnids: Status and Needs. (Kosztarab, M. & C.W. Schaefer, eds.). Virginia Agricul. Exp. Stat. Inform. Sen, 90-1, Virginia Polytechnic Institute, Blacksburg, Vir- ginia. Curtis, D.J. 1978. Community parameters of the ground layer araneid-opilionid taxocene of a Scottish Island. Symp. Zool. Soc. London, 42: 149-159. Draney, M.L. 1997a. Life history variation among ground-dwelling Linyphiidae (Araneae). Ph.D. thesis, Univ. of Georgia, Athens. 237 pp. Draney, M.L. 1997b. Ground-layer spiders (Ara- neae) of a Georgia piedmont floodplain agroe- cosystem: Species list, phenology, and habitat se- lection. J. Arachnol., 25:333-351. Helsdingen, EJ. van. 1983. World distribution of Linyphiidae. Proc. 9th IntT Cong. Arachnol., Panama. Smithsonian Inst. Press, 1986, Pp. 121- 126. Li, D. & R.R. Jackson. 1996. How temperature affects development and reproduction in spiders: A review. J. Thermal Biol., 21:245-274. Lieth, H.(ed.). 1974. Phenology and Seasonality Modeling. Ecol. Stud., vol. 8. Springer- Verlag, New York. Merrett, P. 1969. The phenology of linyphiid spi- ders on heathland in Dorset. J. Zook, London, 157:289-307. Nentwig, W. 1980. The selective prey of linyphiid- like spiders and of their space webs. Oecologia, 45:236-243. Nentwig, W. 1983. The prey of web-building spi- ders compared with feeding experiments (Ara- neae: Araneidae, Linyphiidae, Pholcidae, Age- lenidae). Oecologia, 56:132-139. Nentwig, W. 1987. The prey of spiders. Pp. 249- 263. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, New York. Nyffeler, M. & G. Benz. 1988. Prey and predatory importance of micryphantid spiders in winter wheat fields and hay meadows. J. Appl. Ento- mol., 105:190-197. Platnick, N.I. 1996. Advances In Spider Taxonomy 1992-1995, With Redescriptions 1940-1980. New York Entomol. Soc., New York, 976 pp. Steams, S.C. 1992. The Evolution of Life Histo- ries. Oxford Univ. Press, New York. 249 pp. Tauber, C.A. & M.J. Tauber. 1981. Insect seasonal cycles: Genetics and evolution. Ann. Rev. Ecol. Syst., 12:281-308. Manuscript received I May 1998, revised 6 Septem- ber 1998. 1999. The Journal of Arachnology 27:217-221 HOUSE SPIDERS OF KANSAS Hank Guarisco: Kansas Biological Survey, University of Kansas, Lawrence, Kansas 66047 USA ABSTRACT* Spiders found in and around buildings may be divided into three categories; 1) true synanthropes, which can establish breeding populations in houses, seldom occur locally in the natural environment, and have broad ranges because they may be accidentally transported to new locations, 2) spiders which are seasonally abundant in natural habitats as well as in houses, but don’t establish breeding populations in houses, 3) spiders which are rarely found in houses because they are locally rare or spiders that are locally common but are rarely found indoors. Fifteen species, including the venomous Loxosceles relusa and Cheiracanthium miidei are true synanthropes in Kansas. Category 2 contains 26 species, in- cluding the venomous species Latrodectus hesperus, L. rnactans, and L. variolus. There are 33 species which are rarely found indoors in Kansas. Most species listed have been reported from buildings across the United States. The interest of a homeowner is usually aroused when a spider is discovered inside the house. Since there are at least five species of Kansas spiders which can inflict bites serious enough to require medical attention, common questions that immediately come to mind are: “Is this spider venomous?”, and “Can it hurt me?”. Another concern is whether the house is “infested” with spiders, and what can be done to eliminate an infestation. In response to these questions, information on spiders found in and around houses, buildings, and other structures in Kansas was gathered from the author’s personal collection and field notes, the spider collection of the University of Kansas Snow Entomological Museum, and the literature. Most records are from the Lawrence area in northeastern Kansas because this region has been more intensively studied than other parts of the state. The scientific names of spiders and their order in Tables 1- 3 follow Platnick (1997). The relative preva- lence of spiders in Kansas homes ranges from very common, common, occasional, uncom- mon to rare. Very common and common spe- cies are routinely found in buildings, while rare, uncommon and occasional species have been encountered on 1-5, 6-10, and 11-15 occasions, respectively. Because Kansas is located at the center of the continental United States and contains a wide variety of natural habitats, including eastern deciduous forest and tallgrass and shortgrass prairies, it is the home of a diverse array of plants and animals. The Kansas flora, for example, exhibits geographical affinities with eastern, southern, northern, western, and interior plant communities (Bare & McGregor 1970). Preliminary studies of the spider fauna of the state reveal a similar pattern (Fitch & Fitch 1966; Guarisco & Fitch 1995; Guarisco & Kinman 1990; Scheffer 1904, 1905). There- fore, many of the spiders encountered in Kan- sas homes can also be found in other parts of the continent. Some building-inhabiting spe- cies have even wider distributions, with pop- ulations on several continents. House spiders can be conveniently divided into three categories. True synanthropes are associated with houses, can establish breeding populations in these locations, and usually have very wide distributions because they are often accidentally transported to new areas. They seldom occur locally in the natural en- vironment. The second category includes spe- cies which are seasonally abundant in natural habitats and in houses. Although some may hibernate indoors, and the emergence of large numbers of spiderlings from an occasional eggsac may give the impression of an infes- tation, these species do not establish popula- tions inside houses. The third category con- tains species which are rarely found in and around buildings. They are either common in natural situations and rarely frequent houses, or they are locally rare species. As Edwards & Edwards (1997) have indicated in their study of the spiders of rural delivery mailbox- 217 218 THE JOURNAL OF ARACHNOLOGY Table 1. — Synanthropic (associated with people) spiders in Kansas (Category 1). VC = very common, C = common, O = occasional, U = uncommon, R = rare. Species Distribution Abundance Loxosceles reclusa Gertsch & Mulaik 1940 US VC Scytodes thoracica (Latreille 1802) Cosmopolitan R Pholcus phalangioides (Fuesslin 1775) Cosmopolitan C Spermophora senoculata (Duges 1836) Holarctic R Dysdera crocata C.L. Koch 1838 Cosmopolitan O Oecobius cellariorum (Duges 1836) US, Europe O Octonoba sinensis (Simon 1880) eastern US & Orient O Achaearanea tepidariorum (C.L. Koch 1841) Cosmopolitan VC Steatoda triangulosa (Walckenaer 1802) Cosmopolitan VC Tegenaria domestica (Clerck 1757) Cosmopolitan c Amaurobius ferox (Walckenaer 1830) eastern US, Europe o Cheiracanthium mildei L. Koch 1864 US, Europe, north Africa c Urozelotes rusticus (L. Koch 1872) Cosmopolitan R Salticus scenicus (Clerck 1757) N. & S. America, Europe, north Africa u Sitticus fasciger (Simon 1880) North America, Asia R es in Massachusetts, eventually individuals of all local species will be found in or around buildings. In addition to these three catego- ries, alien species have been occasionally found in Kansas. These species are usually found in produce, such as bananas, and sel- dom establish breeding populations. RESULTS The list of true synanthropic spiders in Kan- sas presently contains 15 species (Table 1). The bites of two of these, Loxosceles reclusa (Sicariidae) and Cheiracanthium mildei (Mi- turgidae) can produce necrotic lesions that re- quire medical attention (Gorham 1968; Spell- man & Levi 1970), and the former is often present in large numbers in Kansas buildings. A sticky trap survey in the University of Kan- sas Museum of Natural History yielded 231 spiders in a two year period, 46.7% of these were L. reclusa. Two very common species occasionally play beneficial roles in houses. Achaearanea tepidariorum (Theridiidae) was observed preying upon the lone star tick. Am- blyomma americanum (Linnaeus) (Acarina: Ixodidae), and Steatoda triangulosa (Theridi- idae) fed upon the brown recluse, L. reclusa (Guarisco 1991). Studies in New Zealand re- vealed that Pholcus phalangioides (Pholcidae) actively invades spider webs and preys upon its occupants (Jackson & Brassington 1987). There are approximately 26 spider species which are seasonally abundant in natural en- vironments and edificarian habitats (Table 2). The bites of Herpyllus ecclesiasticus (Gna- phosidae) (Majeski & Durst 1975), Argiope aurantia (Argiopidae) (Gorham & Rheney 1968), Trachelas tranquillus (Corinnidae) (Oehler 1971; Uetz 1973), and Phidippus au- dax (Salticidae) (Gorham 1968) have pro- duced mostly local reactions; however, those of black widow spiders (Latrodectus sp.) can have much more serious consequences (White et al. 1995). In Kansas, black widows are sometimes discovered in outbuildings, garages and carports. Mimetus puritanus (Mimetidae) and M. no- tius are specialized spider predators which are occasionally found on house eaves near or within host webs. Euryopis limbata (Theridi- idae) occurs on walls and eaves of houses where females produce semicircular, tufted eggsacs. Two Kansas feeding records and ob- servations elsewhere in its range (Archer 1946; Carico 1978) indicate this species’ diet consists of ants. Larinioides cornutus (Ar- giopidae) and L. patagiatus build orbwebs on buildings and bridges near water. The large orbwebs of Argiope aurantia, A. trifasciata (Argiopidae), and Neoscona crucifera (Ar- giopidae) attract attention when located on porches and windows. The large fishing spi- der, Dolomedes tenebrosus (Pisauridae) is of- ten found in sheds, basements and houses. Funnelweb weavers {Agelenopsis sp.) are of- GUARISCO— HOUSE SPIDERS OF KANSAS 219 Table 2. — Seasonally common spiders in Kansas homes (Category 2). VC = very common, C = common, O = occasional, U = uncommon, R = rare. Species Distribution Abun- dance Mimetus notius Chamberlin 1923 eastern US O Mimetus puritanus Chamberlin 1923 eastern US O Euryopis limbata (Walckenaer 1841) eastern US, Canada o Latrodectus hesperus Chamberlin & Ivie 1935 western US, Israel o Latrodectus mactans (Fabricius 1775) North America o Latrodectus variolus Walckenaer 1837 eastern US, Canada u Steatoda borealis (Hentz 1850) US, Canada, Alaska o Theridion murarium Emerton 1882 North America c Argiope aurantia (Lucas 1833) North America to Costa Rica o Argiope trifasciata (Forskal 1775) Cosmopolitan o Larinioides cornutus (Clerck 1757) Holarctic VC Larinioides patagiatus (Clerck 1757) Holarctic o Neoscona crucifera (Lucas 1839) North America, Canary Isl. VC Hogna caroUnensis (Walckenaer 1805) southern Canada, US u Dolomedes tenebrosus Hentz 1843 eastern US, Canada c Agelenopsis naevia (Walcknaer 1841) US, Canada VC Agelenopsis pennsylvanica (C.L. Koch 1843) US VC Hibana gracilis (Hentz 1847) eastern US, Canada c Elaver excepta (L. Koch 1866) eastern US, West Indies o Castianeira variata Gertsch 1942 eastern US, southern Canada u Trachelas tranquillus (Hentz 1847) eastern US, southern Canada o Herpyllus ecclesiasticus Hentz 1832 North America c Philodromus vulgaris (Hentz 1847) Holarctic VC Maevia inclemens (Walckenaer 1837) US, Canada c Phidippus audax (Hentz 1845) US, Canada VC Platycryptus undatus (De Geer 1778) North & Central America VC ten located in comers, windows, and porches, while Hibana gracilis (Anyphaenidae), Elaver excepta (Clubionidae), and Philodromus vul- garis (Philodromidae) occur most often on ceilings and walls. The robust, hairy jumping spider, Phidippus audax, is often mistaken for a black widow by the homeowner because of its coloration. The last category of spiders includes 33 species which occur in natural habitats and have rarely been found in or on buildings in Kansas (Table 3). Further investigation, es- pecially in other sections of the state, would undoubtedly add species to this list. Two alien species have been found in northeastern Kan- sas. A tropical spider belonging to the genus Cupiennius (Ctenidae) was discovered in a produce shipment at the local community mercantile (Cutler pers. comm.). An adult fe- male Hibana cambridgei (Bryant 1931) (An- yphaenidae) was found inside a Lawrence, Kansas residence. The owner may have acci- dentally imported this spider from Arkansas when returning from a weekend trip. DISCUSSION The present study provides baseline data concerning house spiders in northeastern Kan- sas. The 15 synanthropic species and most of the seasonally common species found in northeastern Kansas homes have been report- ed from houses across the United States (Cut- ler 1973; Kaston 1983). Amaurobius ferox (Amaurobiidae) (Guarisco 1989), Scytodes thoracica (Scytodidae), Cheiracanthium mild- ei (Guarisco 1991), and Sitticus fasciger (Sal- ticidae) were first found in the Lawrence area in the late 1980s. Today, C. mildei is one of the most common local house spiders. Since its arrival in North America during the late 1940s (Bryant 1951), this old world species has spread from Boston and New York to southern Ontario, Illinois, California, and Al- abama (Dondale & Redner 1982), and has re- 220 THE JOURNAL OF ARACHNOLOGY Table 3. — Spiders of rare occurrence in/on Kan- Table 3. — Continued sas homes (Category 3). — — — Species Distribution Mimetus epeiroides Emerton 1882 eastern US Theridion goodnightorum Levi 1957 western US Stemonyphantes blauveltae Gertsch 1951 US, Canada Araneus pegnia US to Ecuador, (Walckenaer 1841) Jamaica Hogna helluo (Walckenaer 1837) Pirata sp. US, Canada Rabidosa punctulata (Hentz 1844) US Schizocosa ocreata (Hentz 1844) North America Pisaurina dubia (Hentz 1847) US Pisaurina mira (Walckenaer 1837) US, Canada Agelenopsis Oklahoma (Gertsch 1936) US Anyphaena fraterna (Banks 1896) US Castianeira descripta (Hentz 1847) US, Canada Castianeira variata Gertsch 1942 US, Canada Drassyllus lepidus (Banks 1899) US Drassyllus novus (Banks 1895) US, Canada Sergiolus montanus (Emerton 1890) North America Zelotes hentzi Barrows 1945 US, Canada Zora pumila (Hentz 1850) US Philodromus keyserlingi Marx 1889 US, Canada Philodromus marxi Keyserling 1884 US Thanatus formicinus (Clerck 1757) Holarctic Thanatus rubicellus Mello-Leitao 1929 US, Canada Bassaniana versicolor (Keyserling 1880) North America Misumenops oblongus Canada to (Keyserling 1880) Guatemala Xysticus auctificus Keyserling 1880 US, Canada Xysticus ferox (Hentz 1847) US, Canada Species Distribution Xysticus texanus Banks 1904 US, Mexico Habrocestum pulex (Hentz 1846) US, Canada Phidippus insignarius C.L. Koch 1846 US Phidippus putnami (Peckhams 1883) US, Mexico Phidippus whitmani (Peckhams 1909) US, Canada Tutelina elegans (Hentz 1846) US placed the native C inclusum (Hentz 1847) (Clubionidae) in houses in the northeastern United States (Gertsch 1979). The mobility of today’s society has hastened the spread of syn= anthropic spiders. Other notable instances of spiders being found well beyond their native ranges include, the western black widow (L. hesperus) and the brown recluse {Loxosceles reclusa) in Maine (Jennings & McDaniel 1988; McDaniel & Jennings 1983). The native Steatoda borealis (Theridiidae) has been dis- placed by the European species, S. bipunctata (Linnaeus 1758), in buildings in the north- eastern United States (Nyffeler et al. 1986). Recent surveys of the spider fauna of Cape Cod, Massachesetts revealed the presence of two possible European immigrants, Trochosa ruricola (De Geer 1778) (Lycosidae) andLep- thyphantes tenuis (Blackwall 1852) (Linyphi- idae) (Edwards 1993). ACKNOWLEDGMENTS I gratefully acknowledge the following for help in the completion of this project: the Uni- versity of Kansas Department of Entomology for providing laboratory space, the Kansas Bi- ological Survey for providing valuable equip- ment, and for critically reviewing the manu- script, Bruce Cutler and Henry S. Fitch of the University of Kansas, LITERATURE CITED Archer, A.F. 1946. The Theridiidae or comb-footed spiders of Alabama. Alabama Mus. Nat, Hist. Pa- per, 22:5-67. Bare, J.E. & R.L. McGregor. 1970. An introduc- tion to the phytogeography of Kansas. Univ. Kansas Sci. Bull., 48(26):869-949. GUARISCO— HOUSE SPIDERS OF KANSAS 221 Bryant, E.B. 1951. Redescription of Cheiracan- thium mildei L. Koch, a recent spider immigrant from Europe. Psyche, 58:120-123. Carico, J.E. 1978. Predatory behaviour in Euryopis funebris (Hentz) (Araneae: Theridiidae) and the evolutionary significance of web reduction. Zool. Soc. London Symp., 42:51-58. Cutler, B. 1973. Synanthropic spiders Araneae of the Twin Cities area. Minnesota Acad. Sci. J., 39: 38-39. Dondale, C.D. & J.H. Redner. 1982. The insects and arachnids of Canada. Part 9. The sac spiders of Canada and Alaska Araneae: Clubionidae and Anyphaenidae. Biosystematics Res. Instit., Otta- wa, Canada. Pub. 1724, 194 pp. Edwards, R.L. 1993. New records of spiders (Ar- aneae) from Cape Cod, Massachusetts, including two possible European inunigrants. Entomol. News, 104(2):79-82. Edwards, R.L. & E.H. Edwards. 1997. Behavior and niche selection by mailbox spiders. J. Ar= achnoL, 25(l):20-30. Fitch, H.S. & V.R. Fitch. 1966. Spiders from Mea- de county, Kansas. Kansas Acad. Sci. Trans., 69(1): 11-22. Gertsch, W.J. 1979. American Spiders (2nd ed.). Van Nostrand Reinhold Co., New York. 274 pp. Gorham, J.R. 1968. The brown recluse spider and some other venomous spiders in Georgia. San- script, 11(1):1, 9-13. Gorham, J.R. & T.B. Rheney. 1968. Envenomation by the spiders Chiracanthium inclusurn and Ar- giope aurantia. Observations on arachnidism in the United States. J. American Med. Assoc., 206: 1958-1962. Guarisco, H. 1989. Amaurobius ferox (Araneae: Amaurobiidae), a new addition to the Kansas fauna. J. Kansas Entomol. Soc., 62(1): 127-128. Guarisco, H. 1991. Three species of house spiders first recorded in Kansas: Dysdera crocata (Dys- deridae), Scytodes thoracica (Scytodidae), and Cheiracanthium mildei (Clubionidae). Kansas Acad. Sci. Trans., 94(l-2):73-76. Guarisco, H. & H.S. Fitch. 1991. Spiders of the Kansas Ecological Reserves. Kansas Acad. Sci. Trans., 98(3-4):! 18-129. Guarisco, H. & K.E. Kinman. 1990. Annotated list of the spider family Gnaphosidae in Kansas. Kansas Acad. Sci. Trans., 93(l-2):47-54. Jackson, R.R. & R.J. Brassington. 1987. The bi- ology of Pholcus phalangioides (Araneae, Phol- cidae): predatory versatility, araneophagy and aggressive mimicry. J. Zook, London, 211:227- 238. Jennings, D.T. & I.N. McDaniel. 1988. Latrodectus hesperus (Araneae: Theridiidae) in Maine. En- tomol. News, 99(l):37-40. Kaston, B.J. 1983. Synanthropic spiders. Pp. 221- 245. In Urban Entomology: Interdisciplinary Perspectives. (Frankie & Koehler, eds.). Majeski, J.A. & G.G. Durst, Sr. 1975. Bite by the spider Herpyllus ecclesiasticus in South Caroli- na. Toxicon, 13:377. McDaniel, I.N. & D.T. Jennings. 1983. Loxosceles reclusa (Araneae: Loxoscelidae) found in Maine, USA. J. Med. Entomol., 20:316-317. Nyffeler, M., C.D. Dondale & J.H. Redner. 1986. Evidence for displacement of a North American spider, Steatoda borealis (Hentz), by the Euro- pean species S. bipunctata (Linnaeus) (Araneae: Theridiidae). Canadian J. Zook, 64:867-874. Oehler, C. 1971. Two reports of envenomation by the spider Trachelas tranquillus (Hentz). Cincin- nati J. Med., 52:194. Platnick, N.I. 1997. Advances in spider taxonomy 1992-1995 with redescriptions 1940-1980. New York Entomol. Soc. and American Mus. Nat. Hist., New York. 976 pp. Scheffer, T.H. 1904. A preliminary list of Kansas Spiders, Industrialist (Kansas State Agricultural College) 30(24):37 1-386. Scheffer, TH. 1905. Additions to the list of Kansas spiders. Industrialist (Kansas State Agricultural College) 31(28):435-444. Speilman, A. & H.W. Levi. 1970. Probable enven- omation by Chiracanthium mildei: a spider found in houses. American J. Trop. Med. Hyg., 19: 729-732. Uetz, G.W. 1973. Envenomation by the spider Trachelas tranquillus (Hentz). J. Med. Entomol., 10:227. White, J., J.L. Cardoso & H.W. Fan. 1995. Clinical toxicology of spider bites. Pp. 259-329. In Handbook of Clinical Toxicology of Animal Venoms and Poisons. (J. Meier & J. White, eds.). CRC Press, Boca Raton, New York, & London. Manuscript received 25 April 1998, revised 26 No- vember 1998. 1999. The Journal of Arachnology 27:222-228 SPIDER AND HARVESTMAN COMMUNITIES ALONG A GLACIATION TRANSECT IN THE ITALIAN DOLOMITES Vito Zingerle: Institute of Zoology and Limnology, University of Innsbruck, Technikerstrasse 25, 6020 Innsbruck, Austria ABSTRACT. Arachnid communities of alpine grassland, of screes and woodlands near the timberline and of the nival zone have been compared along a transect from the northern to the southern border of the Dolomites. The region is zoogeographically interesting because of differences of the ice cover during glaciation, which was less severe in the southern area. Along the whole transect spider communities in grasslands and at the timberline zone show approximately the same composition. Endemic species, e.g, Harpactea grisea (Canestrini 1868), Amaurobius rujfoi Thaler 1990, Coelotes mediocris Kulczynski 1887, Cybaeus intermedins Maurer 1992 and Eudasylobus ligusticus Roewer 1923 occur mostly on the south- ernmost station, which remained free of ice. Re-immigrants over short distance are scarce, e.g., Coelotes mediocris at Passo Rolle and Coelotes solitarius L. Koch 1868 in the Puez area. Endemic species were not found in the alpine grassland of the northern Dolomites, which suggests severe impact of glacial events on the local fauna. Central alpine species, i.e., Erigonella subelevata (L. Koch 1869), Metopobactrus nadigi Thaler 1976, Meioneta orites (Thorell 1875), Pardosa blanda (C.L. Koch 1833) and Pardosa mixta (Kulczynski 1887) are still present at the southernmost boundary of the Alps. Nunataks in the northern and central area of the Dolomites allowed speciation effects within the nival fauna: Lepthyphantes brunneri Thaler 1984, Lepthyphantes merretti Millidge 1974, Megabunus annatus (Kulczynski 1887). Further zoo- geographically interesting records are Cryphoeca nivalis Schenkel 1919 and Xysticus bonneti Denis 1938. RIASSUNTO. E stata studiata la composizione della fauna aracnologica della zona subalpina, alpina e nivale lungo un transetto che parte dalle Dolomti settentrionali (Parco Naturale Puez-Odle) e porta fino al bordo meridionale delle Alpi (Monte Grappa). II versante meridionale delle Alpi e di grande importanza ai fini di studi zoogeografici, essendo queste regioni in parte rimaste libere dai ghiacciai durante le epoche glaciali. Sul Monte Grappa sono state riscontrate piu specie endemiche p.es. Harpactea grisea (Canestrini 1868), Amaurobius rujfoi Thaler 1990, Coelotes mediocris Kulczynski 1887, Cybaeus intermedins Maurer 1992 e Eudasylobus ligusticus Roewer 1923. Alcune specie reimmigranti a breve distanza hanno ricon- quistato parti delle Dolomiti raggiungendo regioni piu a nord: Coelotes mediocris Kulczynski 1887 e stato catturato anche a Passo Rolle, Coelotes solitarius L. Koch 1868 anche nel Parco Naturale Puez-Odle. Specie endemiche sembrano essere assenti nella prateria alpina delle Dolomiti settentrionali, dimostrando Peffetto distruttivo dei ghiacciai sulla fauna del suolo. In questa zona si possono trovare specie endemiche sulle cime piu alte rimaste libere dai ghiacciai: Lepthyphantes merretti Millidge 1974, Lepthyphantes brunneri Thaler 1984 e Megabunus armatus (Kulczynski 1887). Altre specie rare e di notevole interesse zoogeografico catturate nell’ambito di questo studio sono Cryphoeca nivalis Schenkel 1919 e Xysticus bonneti Denis 1938. E particolarmente sorprendente la presenza sul Monte Grappa di specie tipiche delle Alpi centrali che sembrano spingersi fino al bordo piu meridionale delle Alpi, p.es. Erigonella subelevata (L. Koch 1869), Metopobactrus nadigi Thaler 1976, Meioneta orites (Thorell 1875), Pardosa blanda (C.L. Koch 1833) e Pardosa mixta (Kulczynski 1887). The southern Alps are interesting for zoo- geographical research. During glaciation pe- riods most of the Alps were covered with a thick ice layer (Klebelsberg 1935; Husen 1987), but ice-free conditions (massivs de ref- uge) along the southern and south-eastern bor- der made survival of animal and plant species possible (Holdhaus 1954). A few species also survived on summits rising above the ice shield (nunataks, Janetschek 1956). Ice-free regions played an important part in differen- tiation of new species as well as in recoloni- zation after glaciation. Therefore widely dis- tributed faunal elements are present in the central Alps, whereas in the southern Alps also endemic species occur. Thaler (1976) rec- ognized the main families which form endem- ic species on the southern border of the Alps: Dysderidae, Linyphiidae, Agelenidae and Amaurobiidae. Maurer (1982a) and Maurer & 222 ZINGERLE— SPIDERS AND HARVESTMEN OF THE DOLOMITES 223 Thaler (1988) studied living conditions in re- fugial areas and discussed the possible dura- tion of speciation and migration. Still little is known about spider commu- nities living in the Dolomites. Koch (1876) and Kulczynski (1887) presented the first lists of the arachnids from this area. Further results concerning spiders or harvestmen of the Do- lomite region were published by Janetschek (1957), Denis (1963) and Marcellino (1988). Marcuzzi (1956, 1975) provided general sur- veys of the fauna of the Dolomites. Noflatsch- er (1996) and Hellrigl (1996) summarized the spider and harvestman fauna of the South-Ty- rol territory, including the northern part of the Dolomites. Lists of the spiders and harvest- men of northern Italy were given by Pesarini (1994) and Chemini (1994). Recently the spi- der fauna of the Puez region was studied by Zingerle (1997). This paper compares arachnid communities from sites near the timberline, from alpine grasslands and from nival habitats along a transect between the northern Dolomites and the southern border of the Alps. It is part of the results of a larger study on the spider fau- na of the Dolomites performed by the author. METHODS Study sites,~“The study sites are situated in the northeastern corner of Italy, about 60 (Monte Grappa) to 140 km (Puez Nature Park) north of Venice in the regions Veneto and Trentino-South Tyrol (Fig. 1). The Dolomite Mountains are mainly composed of limestone and dolomite rock. Several summits rise up to more than 3000 m, the highest elevation is Mount Marmolada with 3342 m. Different forest types can be found due to decreasing rainfall and mediterranean influence between the southern border of the Dolomites and the central Alps. In southern sites the timberline occurs at 1700 m elevation and rises contin- uously up to 2300 m towards the north. In the timberline zone of the southern Dolomites mostly spruce and beech forest (Picea abies, Fagus sylvatica) are found; in northern sites larch (Larix decidua), cembra-pine {Pinus cembra) and mountain pine (Pinus mugo) ex- ist. Ditches and humid places are mostly cov- ered by associations of Alnus viridis and of Salix spp. Alpine grasslands on limestone and dolomite rock are dominated by Car ex sem- pervirens and Sesleria albicans. On bare rocks Figure 1. — Map showing position of the sam- pling areas in the Dolomites (Northern Italy) and ice-cover situation during glaciation (according to Klebelsberg 1935). Shaded area indicates ice-free regions on the border of the south-eastern Alps and nunataks (see text). Study sites: G = Monte Grappa (MG); P = Puez Nature Park (PU); S = Passo Sella (SE); R = Passo Rolle (RO); X = Sesto/Sexten Nature Park (SX). Main Cities in the vicinity of the study areas: B = Bolzano (Bozen); I = Innsbruck; U = Udine; Vn = Venice; Vr = Verona. of the mountain tops there grow patches of mosses and grasses, like Carex firma and Car- ex rupestris. For a general view of the area, see Fig. 2. Five study areas were selected (see Fig. 1): Monte Grappa, elevation 1775 m, 45°55'N, ir49'E, at the border between the Provinces Belluno, Vicenza and Treviso. Pas- so Rolle, Paneveggio Pale-S. Martino Nature Park, elevation 1970 m, 46°18'N, 11°48'E, Province Trento. Passo Sella, elevation 2244 m, 46°30'N, 11°48'E, at the border between Provinces Trento and South-Tyrol. Puez Na- ture Park, Antersasc Valley, elevation 2000 m, 46°37'N, ir52'E, Province South-Tyrol. Sesto/Sexten Nature Park, Gsell area, eleva- tion 2000 m, 46°40'N, 12°22'E, Province South-Tyrol. Collection methods.— -In each of the five study areas mentioned above, three habitats types (grassland, scree and forest at the tim- berline) were sampled by pitfall traps. Four covered traps containing a formalin/ water so- lution and a small amount of detergent were 224 THE JOURNAL OF ARACHNOLOGY Figure 2. — Typical landscape of the Dolomites in the vicinity of Passo Rolle (RO). Alpine grasslands and subalpine woodlands sampled are visible on the left side of the photo, nival habitats on the right side. placed a few meters apart from each other in each habitat. Altogether 15 sites were sampled by 60 traps. The traps remained in the area throughout a whole year and were emptied in intervals of 3^4 weeks during the vegetation period. In the Puez area sampling was carried out from Spring 1995 to Spring 1996, in the other areas from Spring 1997 to Spring 1998. For details about the traps see Zingerle (1997). The mean number of specimens per trap, number of species, diversity value (^log) according to Shannon-Weaver and family composition are given for each site. Additionally, alpine and nival habitats were sampled during Summer 1995, 1996 and 1997 by hand collecting. This sampling technique is effective, due to the low species number in these habitats. Hand collecting was performed at the following localities: Puez (2600-2900 m). Sella (2900-3150 m), Sasso Lungo (Langkofel) (2700 m), Passo Falzarego (2400 m), Tofana (2950-3220 m), Cristallo (2400- 2950 m), Sesto/Sextee (2450 m), Marmolada (3000-3300 m) and Pale di S. Martino (2700- 3000 m). Voucher specimens will be deposited in the Natural History Museum, Vienna (Austria) and in Naturmuseum Sudtirol, Bozen (Italy). RESULTS Spiders and harvestmen of alpine grass- lands and timberline zone.— The total ma- terial captured by pitfalls comprises 3640 adult specimens from 164 species and 15 fam- ilies. The dominant families are: Lieyphiidae (41% of specimens), Lycosidae (39%), Age- lenidae (5%) Gnaphosidae (5%), Thomisidae (5%). Characteristics of the spider fauna found at each site are shown in Table 1. On the whole transect family composition of spiders in grassland sites is approximately the same. In grasslands Lycosidae reach 48- 74% of the total spiders. The most dominant species are Pardosa blanda (C.L. Koch 1833), Pardosa o reop hi la Simon 1937, Alopecosa taeniata (C.L. Koch 1848), Pardosa riparia (C.L. Koch 1833), Trochosa terricola Thorell 1856 and Pardosa ferruginea (L. Koch 1870). Fluctuation of abundance among these species in different areas is probably coincidental. The high abundance of Lycosidae in alpine grass- land habitats is a well-known phenomenon. The secoedmost important family is Linyphi- idae (14-49%), mostly species with a prefer- ence for sites with an open canopy cover and little shade, like Centromerus pabulator (O.R- ZINGERLE— SPIDERS AND HARVESTMEN OF THE DOLOMITES 225 Table 1. — Structure of spider communities studied with pitfall traps in the Dolomites (Southern Alps, Italy) from spring 1995 to spring 1998. Abbreviations: x = mean number of individuals; S = number of species; H' = Shannon- Weaver diversity (^log). Percentages = number of specimens from a specific family the total number of spiders captured in a habitat during the whole sampling period. Habitats: G = alpine grassland; S = scree; T = forest sites near the timberline. Study areas: MG = Monte Grappa; RO = Passo Rolle; SE = Passo Sella; PU = Puez Nature Park (see also Zingerle 1997); SX = Sesto/Sexten Nature Park. Habitat Study areas MG RO SE PU SX G X 98 102 80 79 93 S 40 27 32 24 24 H' 3.6 3.3 2.2 2.7 3.1 Family Lycosidae % 67 63 74 65 48 composition Linyphiidae % 20 14 18 19 49 Gnaphosidae % 10 11 4 7 1 Thomisidae % 1 11 1 7 0 others % 2 (4 fam.) 1 (4 fam.) 3 (5 fam.) 2 (4 fam.) 2 (5 fam.) S X 32 10 26 14 25 S 30 9 16 15 16 H' 4.2 2.4 2.9 3.5 3.0 Family Thomisidae % 2 44 31 30 46 composition Linyphiidae % 24 39 42 24 23 Lycosidae % 16 10 26 26 23 Gnaphosidae % 29 2 0 9 0 Dysderidae % 17 0 0 0 0 Theridiidae % 0 0 0 7 1 others % 12 (4 fam.) 5 (1 fam.) 1 (1 fam.) 4(1 fam.) 7 (2 fam.) T X 35 65 37 115 99 S 30 28 31 39 37 H' 3.8 3.4 3.5 3.9 3.8 Family Linyphiidae % 75 75 46 55 76 composition Lycosidae % 9 14 42 12 10 Agelenidae % 9 4 1 29 2 Gnaphosidae % 4 1 9 2 1 Theridiidae % 0 5 0 1 6 others % 3 (5 fam.) 1 (2 fam.) 2 (2 fam.) 1 (3 fam.) 5 (3 fam.) Cambridge 1875), Tiso vagans (Blackwall 1834) and Bolyphantes alticeps (Sundevall 1832). High numbers of Linyphiidae (e.g., in Sesto) occur in the proximity of a dense veg- etation cover. Shannon- Weaver diversity in grasslands reaches values between 2.2 and 3.6. The harvestman Eudasylobus ligusticus Roewer 1923, endemic to the southern border of the Alps, occurs in grasslands on Monte Grappa. The absence of an endemic Coelotes species in my collection in grasslands of this area is quite interesting (Maurer 1982a, b). Screes are generally characterized by lower numbers of individuals and species. High di- versity values (e.g., Monte Grappa) are found, when the influence of adjacent habitats is great. So the dominant thomisid species (Xy.?- ticus lanio C.L. Koch 1824, Xysticus desidio- sus Simon 1875 and Xysticus audax (Schrank 1803)) also occur in grassland areas. Typical inhabitants of alpine screes are Lepthyphantes variabilis Kulczynski 1887, Tiso aestivus (L. Koch 1872), Rugathodes bellicosus Simon 1873 and Pardosa nigra (C.L. Koch 1834). In screes on Monte Grappa exist also endemic spiders, Harpactea grisea (Canestrini 1868) and Amaurobius rujfoi Thaler 1990, and the endemic harvestman Eudasylobus ligusticus. All these are invaders from the neighboring forests. Forests near the timberline are dominated by linyphiid spiders, which reach 46-76% of 226 THE JOURNAL OF ARACHNOLOGY Table 2. — Presence ( + ) or absence (— ) of zoogeographically interesting spider and harvestman species in alpine grassland and woodlands near the timberline collected by pitfall traps in the Dolomites (Southern Alps, Italy) from Spring 1995 to Spring 1998. Study areas: PU = Puez Nature Park; SX = Sesto/Sexten Nature Park; SE = Passo Sella; RO = Passo Rolle; MG = Monte Grappa. Species Study areas PU SX SE RO MG Harpactea grisea (Canestrini 1868) - - - - + Amaurobius rujfoi Thaler 1990 - - - - + Cybaeus intermedius Maurer 1992 - - - - Eudasylobus ligusticus Roewer 1923 - - - - + Coelotes mediocris Kulczynski 1887 - - - + + Coelotes solitarius L. Koch 1868 + - - + - Metopobactrus nadigi Thaler 1976 + - + - + Meioneta orites (Thorell 1875) + - - + Pardosa blanda (C.L. Koch 1833) + - - + Lepthyphantes cf.fragilis (Thorell 1875) + + + - Troglohyphantes tirolensis Schenkel 1950 + + + + - Pardosa mixta (Kulczynski 1887) + - + + + Erigone Ha subelevata (L. Koch 1869) + -f- + + + the spiders collected there. In addition, a high abundance of Agelenidae was found. Timber- line sites are interesting because at this eco- tone species from forest and alpine habitats occur together. Accordingly Lycosidae are also quite numerous and amount to 42% of all individuals from these habitats. Species num- bers are generally higher than in other habitats and the Shannon-Weaver diversity index reaches values between 3.4 and 3.9, indicating a “mixed” fauna. Typical and abundant spe- cies from forest sites, which occur in the tim- berline zone are C. pabulator, Diplocephalus latifrons (O. P.-Cambridge 1863), Lepthyphan- tes monticola (Kulczynski 1882), Lepthyphan- tes jacksonoides Helsdingen 1977, Cybaeus tetricus (C.L. Koch 1839) and Cryphoeca sil- vicola (C.L. Koch 1834); species from alpine grassland are P. blanda, A. taeniata, P. or- eophila and P. riparia. Endemic spiders and harvestmen collected on Monte Grappa are: Amaurobius rujfoi, Coelotes mediocris Kul- czynski 1887, Cybaeus intermedius Maurer 1992 and Eudasylobus ligusticus. Remarkable is the occurrence of C mediocris at timberline sites of Passo Rolle. Hand catches in alpine and nival zones. — Spider and harvestman communities in these zones are mainly composed of few specialists which, however, occur abundantly. Thaler (1981, 1988) collected 27 spider spe- cies in the nival zone of the central Alps and 49 nival species on 58 summits of the eastern Alps. In the present study a total of 29 spider (299 individuals) and 4 harvestman species (11 individuals) were collected on 16 summits of the Dolomites between 2400-3300 m ele- vation. Twenty-two of these species belong to Linyphiidae, 2 to Theridiidae, 2 to Thomisi- dae, 1 to Agelenidae, Lycosidae and Philod- romidae, respectively. Several of these species collected in high numbers on most summits show a rather continuous distribution in the studied area. Only two of them, Hilaira mon- tigena (L. Koch 1873) and Erigone tirolensis (L. Koch 1872), occur mainly in the nival zone, the rest, (e.g., L. variabilis, Meioneta gulosa (L. Koch, 1869), Diplocephalus helleri (L. Koch 1869) and Oreonetides glacialis (L. Koch 1872)) are also present in the alpine zone. The ballooning lowland species Mei- oneta rurestris (C.L. Koch 1836) also occurs frequently in the nival zone. The harvestmen {Dicranopalpus gasteinensis Doleschall 1852, Mitopus glacialis (Heer 1845) and Megabunus armatus (Kulczynski 1887)) occur in the nival and in the alpine zone, whereas Ischyropsalis kollari C.L. Koch 1839 is also present in the montane woodland. Furthermore, two endem- ic spider species (i.e., Lepthyphantes merretti Millidge 1974 and L. brunneri Thaler 1984) with a mainly nival distribution were found quite numerously at certain localities (see Ta- ble 3). The endemic harvestman Megabunus armatus was also collected several times. These species demonstrate that the highest ZINGERLE— SPIDERS AND HARVESTMEN OF THE DOLOMITES 227 Table 3. — Presence ( + ) or absence (— ) of zoogeographically interesting spider and harvestman species in the alpine zone and the nival zone of the Dolomites (Southern Alps, Italy) collected by hand during Summer 1995, 1996 and 1997. Sampled localities: SE = Sella; PU = Puez; MA = Marmolada; PA = Pale di S. Martino; PF = Passo Falzarego; SL = Sasso Lungo/Langkofel; TO = Tofane; SX = Sesto/ Sexten; CR = Cristallo. Sampled localities Species SE PU MA PA PF SL TO SX CR Lepthyphantes merretti Millidge 1974 + + -j- - - - - - Lepthyphantes brunneri Thaler 1984 - - - - - - + + + Megabunus armatus (Kulczynski 1887) 4- + - - 4- + + - - Cryphoeca nivalis Schenkel 1919 - - - - - - - - Xysticus bonneti Denis 1938 — — - - + — — - — summits of the Dolomites were spared by gla- ciation events, allowing speciation to take place. Two remarkable nival spiders species were captured only at one site each: Cryphoe- ca nivalis Schenkel 1919 on Pale di S. Mar- tino (2700 m) and Xysticus bonneti Denis 1938 in the vicinity of Passo Falzarego (2400 m). DISCUSSION The fauna of spiders and harvestman from the alpine and the timberline zone of the Do- lomites shows a similar composition in the northern and in the southern area. Neverthe- less, a few species indicate a minor impact of glaciation events in the border region and give evidence for peripheral isolation in the south- ern Dolomites (e.g., Harpactea grisea, Amau- robius rujfoi, Cybaeus intermedins, Coelotes mediocris and Eudasylobus ligusticus). Whether these species occur at northern sites or not depends on their capabilities to re-col- onize the area after glaciation (see Table 2). Some species, like H. grisea, A. rujfoi, C. in- termedins and E. ligusticus, seem to be re- stricted to the southernmost border, occurring exclusively on Monte Grappa. Others, (e.g., Coelotes mediocris), re-immigrated into the area up to Passo Rolle and show a higher mo- bility. Coelotes solitarius probably survived glaciation periods near the south-eastern bor- der of the Alps; nevertheless it was found all the way up to the northernmost site, the Puez area. Populations of Lepthyphantes cf . fragilis (Thorell 1875) and Troglohyphantes tirolensis Schenkel 1950 were also influenced by gla- ciation, but further taxonomic work is needed to realize their precise status. Maurer (1982a, b) reported the occurrence of 10 species of the Coelotes pastor-gmwp from alpine grasslands between Liguria (Italy) and Slovenia, e.g., C. pastor lessinensis Maurer 1982 on Monti Les- sini and Coelotes alpinus Polenec 1972 in the eastermost area. Species from this group were not found during this study. Remarkable is the presence of central- alpine spiders, (i.e., Eri- gonella subelevata, Metopobactrus nadigi, Meioneta orites, Pardosa blanda and P. mix- ta) at the southermost limit of the Alps at Monte Grappa. The nival fauna of the Dolomites demon- strates isolation effects and speciation on nun- ataks during glaciations. The endemic Liny- phiidae Lepthyphantes merretti and L. brunneri were found on only a few summits, L. brunneri being distributed in the eastern and L. merretti in the western area of the Do- lomites (see Table 3; Thaler 1988). The en- demic harvestman Megabunus armatus is re- stricted to the south-eastern Alps where it lives on rocks above the timberline. The vi- cariance pattern in this genus in the Alps probably reflects the effects of glaciation (Martens 1978; Chemini 1985). The agelenid Cryphoeca nivalis was previously known only from the Adamello and Brenta area in Italy and from the central Swiss Alps (Thaler 1978; Maurer & Hanggi 1990). The station in the Dolomites is probably close to its eastermost boundary of distribution. Xysticus bonneti is a rarely found species (Thaler 1981), which shows a very patchy distribution in the alpine zone of the western palearctic mountains. ACKNOWLEDGEMENTS I thank Dr. Konrad Thaler for help with species identifications, discussion and manu- script review. For suggestions I am also grate- 228 THE JOURNAL OF ARACHNOLOGY ful to Dr. Peter Schwendinger. The following institutions provided equipment, support and permits: Institute of Zoology, University of Innsbruck (Austria); Museo Scienze Naturali Alto Adige, Bolzano/Naturmuseum Siidtirol, Bozen (Italy); Ufficio Parchi Naturali, Bolza= no/Amt fiir Naturparke, Bozen (Italy); Ente Parco Paneveggio - Pale di S. Martino (Italy). LITERATURE CITED Chemini, C. 1985. Megabunus bergomas n. sp. dalle Alpi Italiane (Arachnida, Opiliones). Boll. Soc. Ent. Italiana, 117:4-7. Chemini, C. 1994. Arachnida, Scorpionida, Pal- pigradi, Solifugae, Opiliones. In Checklist delle specie della fauna italiana, 23. (A. Minelli, S. Ruffo & S. La Posta, eds.). Calderini, Bologna. Denis, J. 1963. Araignees des Dolomites. Atti 1st. Veneto Sc. Lett. Arti, 121:253-271. Hellrigl, K. 1996. Opiliones - Weberknechte, Af- terspinnen. Pp. 205-210. In Die Tierwelt Siidti- rols. (K. Hellrigl, ed.). Veroff. Naturmus. Siidti- rol, Bozen, 1. Holdhaus, K. 1954. Die Spuren der Eiszeit in der Tierwelt Europas. Abh. Zool.-Bot. Ges. Wien, 18:1-493. Husen, D. 1987. Die Ostalpen in den Eiszeiten. Popularwiss. Veroff. Geol. Bundesanst., Wien. Janetschek, H. 1956. Das Problem der inneralpinen Eiszeittiberdauemng durch Tiere (Ein Beitrag zur Geschichte der Nivalfauna). Osterreichische Zool. Zeitschr., 6:421-506. Janetschek, H. 1957. Zur Landtierwelt der Dolom- iten. Der Schlem, 31:71-86. Klebelsberg, R. 1935. Geologie von Tirol. Bom- traeger, Berlin. Koch, L. 1876. Verzeichniss der in Tirol bis jetzt beobachteten Arachniden. Zeitschr. Ferdinan- deum, Innsbruck, 20:219-354. Kulczynski, V. 1887. Przyczynek do Tyrolskiej fauny Pajeczykow. Rozpr. Spraw, Wydz. Mat. Przyrod. Akad. Umiej., 16:245-356. Marcellino, 1. 1988. Opilionidi (Arachnida, Opili- ones) di ambienti montani ed alpini delle Dolom- iti. Studi Trentini Sc. Nat. Acta Biol., 64 suppl., pp. 441-465. Marcuzzi, G. 1956. Fauna delle Dolomiti. - Mem. 1st. Veneto Sci. Lett. Arti, Cl. Sci. Mat. Nat., 31: 1-595. Marcuzzi, G. 1975. La Fauna delle Dolomiti. Man- frini, Galliano (Trento). Martens, J. 1978. Spinnentiere, Arachnida: Weber- knechte, Opiliones. Tierwelt Deutschlands, 64, Fischer, Jena. Maurer, R. 1982a. Zur Kenntnis der Gattung Coe- lotes (Araneae: Agelenidae) in den Alpenlan- dem. 1. Rev. Suisse Zool., 89:313-336. Maurer, R. 1982b. Zur Kenntnis der Gattung Coe- lotes (Araneae: Agelenidae) in den Alpenlan- dem. II. Boll. Mus. Civ. St. Nat., Verona, 8:165- 183. Maurer, R. & J. E. Hanggi. 1990. Katalog der Schweizerischen Spinnen. Doc. Faun. Helvetiae, 12, Neuchatel. Maurer, R. & K. Thaler. 1988. Uber bemerken- swerte Spinnen des Parc National de Mercantour (F) und seiner Umgebung (Arachnida: Araneae). Mit Anmerkungen zum Endemismus in der Spin- nenfauna der Meeralpen. Rev. Suisse Zool., 95: 329-352. Noflatscher, M.T. 1996. Ordnung Aranei - Spinnen, Webspinnen. Pp. 211-228. In Die Tierwelt Siid- tirols. (K. Hellrigl, ed.). Veroff. Naturmus. Siid- tirol, Bozen, 1. Pesarini, C. 1994. Arachnida Araneae. Pp. 1-42. In Checklist delle specie della fauna italiana, 23. (A. Minelli, S. Ruffo & S. La Posta, eds.). Cald- erini, Bologna. Thaler, K. 1976. Endemiten und arktoalpine Arten in der Spinnenfauna der Ostalpen (Arachnida: Araneae). Entomol. Germanica, 3:135-141. Thaler, K. 1978. Die Gattung Cryphoeca in den Alpen (Arachnida, Aranei, Agelenidae). Zool. Anz., Jena, 200:334-346. Thaler, K. 1981. Neue Arachniden-Funde in der nivalen Stufe der Zentralalpen Nordtirols (Oster- reich) (Aranei, Opiliones, Psudoscorpiones). Ber. Nat. -Med. Ver. Innsbruck, 68:99-105. Thaler, K. 1988. Arealformen in der nivalen Spin- nenfauna der Ostalpen (Arachnida, Aranei). Zool. Anz., Jena, 220:233-244. Zingerle, V. 1997. Epigaische Spinnen und Weber- knechte im Naturpark Puez-Geisler (Dolomiten, Siidtirol) (Araneae, Opiliones). Ber. Nat.-Med. Ver. Innsbruck, 84:171-226. Manuscript received 29 April 1998, revised 6 Sep- tember 1998. 1999. The Journal of Arachnology 27:229-235 SALTICIDAE (ARACHNIDA, ARANEAE) OF ISLANDS OFF AUSTRALIA Barbara Patoleta and Marek Zabka: Zaklad Zoologii WSRP, 08-110 Siedlce, Poland ABSTRACT. Thirty nine species of Salticidae from 33 Australian islands are analyzed with respect to their total distribution, dispersal possibilities and relations with the continental fauna. The possibility of the Torres Strait islands as a dispersal route for salticids is discussed. The studies of island faunas have been the subject of zoogeographical and evolutionary research for over 150 years and have resulted in hundreds of papers, with the syntheses by Carlquist (1965, 1974) and Mac Arthur & Wil- son (1967) being the best known. Modem zoogeographical analyses, based on island spider faunas, began some 60 years ago (Borland 1934) and have continued ever since by, e.g., Forster (1975), Lehtinen (1980, 1996), Baert et al. (1989), Zabka (1988, 1990, 1991, 1993), Baert & Jocque (1993), Gillespie (1993), Gillespie et al. (1994), Proszyhski (1992, 1996) and Berry et al. (1996, 1997), but only a few papers were based on verified and sufficient taxonomic data. The present contribution is mostly based on material collected by one of us (MZ) while visiting Queensland Museum (Brisbane), Aus- tralian Museum (Sydney), Western Australian Museum (Perth) and Australian National In- sect Collection (Canberra). The main purposes of this paper are (1). To analyze the species composition in respect to their origin, total distribution and dispersal abilities; (2). To es- timate the expansiveness of Australian conti- nental faunas towards studied islands; (3). To evaluate the role of Torres Strait islands in faunistic exchange between Australia and New Guinea. THE AREA The islands are of coral, volcanic or con- tinental origin and are (with few exceptions) located along the NE coast of Australia (Fig. 1). Their surfaces are rather flat, either barren or vegetated, mostly by Eucalyptus, wattles, palms and ferns. Few have developed rain- forest or mangrove communities. Due to ocean level fluctuations over the last 50,000 years, at least some islands have been sub- merged or formed land bridges with the con- tinent (e.g., Torres Strait islands). All these circumstances and the human occupation make it rather unlikely for the majority of islands to have developed their own endemic salticid faunas. When one of us (MZ) began research on the Australian and New Guinean Salticidae over ten years ago, close relationships be- tween the faunas of these two regions were expected. Consequently, it was hypothesized that the Cape York Peninsula and Torres Strait islands were the natural passage for dispersal/ expansion. In fact, the parts of this area cov- ered with savannah and Eucalyptus forests do form such a passage zone within these habi- tats, but mostly in one direction — from Aus- tralia to South Papua, and no further north be- cause of rainforest barrier. During glacial cooling, aridization and rainforest regression, habitats were further enhanced in favor of the Australian fauna. Thus, for northern (Oriental and New Guinean) rainforest dwellers, the Cape York Peninsula and Torres Strait islands should be treated as filters rather than a dis- persal route. THE SALTICIDS Continental fauna — ^the source. — About 340 salticid species have been reported from Australia so far (Davies & Zabka 1989; Zab- ka 1990, 1991, unpubl. data). Of them 286 belong to 63 verified genera, others are clas- sified as incertae sedis. Approximately 60% of species are endemic and these increase in number towards southern and central-western Australia. The long-term isolation of the con- 229 230 THE JOURNAL OF ARACHNOLOGY 130 140 150 Figure 1 : Map showing the geographical location of the analyzed islands along the coast of Australia (for detailed information see Table 1). tinent and uniqueness of the Australian biota made the speciation so successful. Further- more, inhabiting various Eucalyptus com- munities, remote desert and semi-desert areas and/or microhabitats (e.g., under bark, in leaf litter), particular species have biological and structural limitations to expansion. The sec- ond largest group of continental salticids, but smaller than expected especially in compar- ison with other spider families (Araneidae, Theridiidae, see Main 1981), is formed of tropical immigrants from the Oriental Region and New Guinea. They spread to north and north-eastern coastal rainforest remnants, and decrease in number to the south. Finally, the third group is made up of cosmopolitan/pan- tropical species, distributed by human activ- ity. Island fauna. — During the last ten years substantial progress has been made in studies of the Pacific island Salticidae (Zabka 1988; Proszynski 1992, 1996; Berry et al. 1996, 1997). In our research, we analyzed 39 spe- cies, though no island had more than eight species. Being aware of the limitations, we distinguish three groups of species (Tables 2, 3). Group 1: The largest (24 species) is made up of Australian endemics. Although some of them have also spread to south Papua (savan- nah, Eucalyptus forests) they seem to be of Australian origin and belong to Australian en- demic genera (Abracadabrella, Astia, Holo- platys, Ligurinus, Mopsus, Mopsolodes, Si~ maetha, Tauala). Four species of this group {Ergane cognata, E. insulana, Simaetha atyp- ica and Tauala minutus) are known exclu- sively from the islands. However, their en- demic island status seems doubtful due to the young age of the inhabited islands. Group 2: At least 11 species are of wide distribution, ranging from west Africa through Sri Lanka to western Pacific islands (in one case even to Hawaii) and belong to genera of alien (outside Australian) origin— usually SE Asian and New Guinean. Cytaea plumbeiv- entris, reported from 12 islands, was the most common here. This species can be found in gardens and parks of NE Queensland and as such has probably been dispersed by man. Cosmophasis thalassina has a similar biology and distribution, though it is less common (three islands). Group 3: Four island species have cosmo- politan/p antropical distribution, and all live in human habitations and are spread by man. Dispersal. — For the analyzed case two dis- persal methods, aerodispersal and antropod- ispersal, should be considered. Rafting, though theoretically possible, is not discussed because of lack of published or other data re- garding Salticidae. Aerodispersal: Salticidae occupy various habitats, each providing different aerodisper- sal possibilities. Leaf-litter or bark dwellers, for instance, are poorer candidates for bal- looning than those living in open areas, tree canopies or human habitats. Salticidae con- stitute only 1.5-7% of all spiders in aero- plankton (Horner 1975; Salmon & Horner 1977; Greenstone et al. 1987). It is widely known that juveniles are more effective bal- looners than adults; and in our research they constituted 50.7% of all specimens which seems to support the aerodispersal hypothe- sis. Some indirect data from the analyzed area were provided by Zabka (1991) from tree canopies of NE Queensland. Amongst 70 specimens found there, the most common were representatives of Tara, Simaethula, Opisthoncus, Prostheclina. Except for the latter, those genera have also been recorded in our study. Tara and Simaethula have not been considered as identified to the genus (not species) level only. Helpis minitabunda (found in tree canopies) is spread from Aus- tralia and New Guinea to adjacent archipel- PATOLETA & ZABKA— SALTICIDS OF AUSTRALIAN ISLANDS 231 Table 1. — Number of species recorded on individual islands. Number of species Island Geographical location S E 8 Fitzroy 16°56', 146°00' Queensland 8 Masthead 23°32', 15r43' Queensland 7 Horn 10°37', 142°17' Torres Strait 5 Heron 23°26', 15r55' Queensland 4 Barrow 20°46', 115°24' Western Australia 4 Lizard 14°40', 145°28' Queensland 4 Motmot 3 Caimcross West iri5', 142°55' Torres Strait 3 Hannibal East ir36', 142°56' Torres Strait 2 Campbell 9°34', 143°29' Torres Strait 2 Damley 9°35', 143°46' Torres Strait 2 Fraser 25°22', 153°07' Queensland 2 Friday 10°36', 142°10' Torres Strait 2 Murray 9°56', 144°04' Torres Strait 2 North West 23°18', 15U42' Queensland 2 Pellew 15°3L, 136°53' Northern Territory 2 Pethebridge 14°44', 145°05' Queensland 2 Stephens 9°3r, 143°32' Torres Strait 2 Thursday 10°35', 142°13' Torres Strait 2 Tryon 23°15', 151°46' Queensland 2 Yam 9°53\ 143°45' Torres Strait 1 B instead 13°13', 143°33' Queensland 1 Gannett Cay 21°59', 152°28' Queensland 1 Little Fitzroy 16°55', 146°01' Queensland 1 Low 22°03', 150°06' Queensland 1 Moreton 27°11', 153°24' Queensland 1 Percy 2r42', 150°20' Queensland 1 Rocky 15°36', 145°21' Queensland 1 Saibai 9°23', 142°40' Queensland 1 Tana 1 Wharton Reef 14°08', 144°00' Queensland 1 Wilson 23n8', 151°55' Queensland 1 Yorke 9°44', 143°25' Torres Strait agos and to New Zealand, and has also been found in our research. Anthropodispersai: This way of dispersal is typical for species occupying human habita- tions, and their distribution is world- wide. Four such species (Hasarius adansoni. Me- nemerus bivittatus, Plexippus paykulli, P. pe- tersi) are found on the islands. It is likely that also other island species (e.g., Cytaea plum- beiventris, Cosmophasis thalassind) can dis- perse this way. CONCLUSIONS Only 10% of all continental Australian saL ticid species are found on the analyzed is- lands, indicating they are either poorly stud- ied, scanty in species and/or ecologically inappropriate. Even some large continental genera are missing on the islands or are rep- resented by single species only (Table 4). This supports the idea (quite obvious for resident Australian arachnologists) that the enormous- ly diverse Australian spider/salticid fauna is largely the result of habitat variability and flo- ristic diversity. The islands, being poor in plant communities, are mostly inhabited by eury topic species. However, until the material is more complete, it is premature to reliably discuss such “island problems” as size effect, distance from the source of the fauna, island age, plant communities and topographic influ- ence. For the majority of islands only one or two species are listed. Even for the richest (Fi- tzroy) only eight species are recorded. Of all 232 THE JOURNAL OF ARACHNOLOGY Table 2. — The distribution of species recorded on islands off Australia. WA = Western Australia, NT = Northern Territory, SA = South Australia, TAS = Tasmania, QLD = Queensland, NSW = New South Wales, NG = New Guinea, PNG = Papua New Guinea, C = central, M = middle, S = south, W = west, NE = north-east, E = east, N == north. Records in continental Australia Species Islands WA NT QLD NSW Other records Abracadabrella elegans Binstead NE, E E Astia hariola Fraser E E NG Bavia aericeps Horn, Campbell NE NG, C and W Pacific Archipelagoes Bianor maculatus Gannett Cay, Motmot S E New Caledonia, Sa- moa, Vietnam Clynotis severus Yam, Horn + E E S PNG Cosmophasis bitaeniata Fitzroy E E PNG, Am Is. Cosmophasis micarioides Motmot NE Cosmophasis thalassina Cairncross West, Fitzroy, N Malay Arch., NG Hannibal East Cyrba ocellata Barrow, Masthead from Africa to Oriental Region and Australia Cytaea mitellata Campbell Am Is., Yule Is., Sun- da Arch. Cytaea frontaligera Darnley E N PNG, Am Is. Cytaea plumbeiventris Fitzroy, Hannibal East, NE Am Is., PNG, New Heron, Horn, Little Fitz- roy, Lizard, Low, Mast- head, Murray, Pethe- bridge, Stephens, Tryon Mecklenburg Cytaea severa Barrow, Lizard, Masthead, + Yam Ergane cognata Pellew Ergane insulana Pellew Euryattus bleekeri Cairncross West, Fitzroy, NE, M NG, Ambon, Am, Ma- Thursday laysia Evarcha infrastriata Horn NE, E Gangus longulus Motmot + Hasarius adansoni Heron, Masthead, North Pantropical West, Percy, Wilson Helpis minitabunda Fitzroy SE Holoplatys colemani Lizard, Masthead + + Holoplatys complanata Fitzroy, Masthead, Tryon N E PNG Ligurinus bipenicilatus Fraser + + Menemerus bivittatus Barrow, Heron, Masthead Pantropical Mopsolodes australiensis Horn + N, NE, SE Mopsus mormon Fitzroy, North West, Saibai + N NG Opisthoncus abnormis Wharton Reef -1- Plexippus paykulli Hannibal East Pantropical Plexippus petersi Thursday Pantropical Servaea vestita Yorke E Tasmania Simaetha atypica Pethebridge + Simaetha robustior Stephen NE Am Simaetha tenuidens Friday, Heron, Horn, More- E PNG ton Simaetha tenuior Barrow, Heron, Masthead E PATOLETA & ZABKA— SALTICIDS OF AUSTRALIAN ISLANDS 233 Table 2. — Continued Species Islands Records in continental Australia WA NT QLD NSW Other records Tauala minutus Murray ''Trite"' longula Caimcross West, Damley, Motmot, Rocky NE Zenodorus arcipluvius Tana New Hebrides Zenodorus metallescens Horn E E Zenodorus orbiculatus Fitzroy, Fraser, Friday, Liz- ard E E 39 species, three zoogeographic groups are distinguished: Australian endemics, Oriental and New Guinean immigrants, and cosmopol- itan/pantropical elements. We hypothesize that no island endemics are found, unless con- firmed by further research. Ballooning and man agency seem possible ways of dispersal; however, it is more likely that, at least some islands were colonized via past land bridges. The Torres Strait islands are the barrier for northern tropical (rainforest) species and the passage for southern savannah and Eucalyptus forest inhabitants. ACBCNOWLEDGMENTS Drs. V.T. Davies, R. Raven, M. Gray, R.B. Halliday and Ms. J. Waldock are acknowl- edged for their generous help and cooperation during stay of one of us (MZ) in their de- partments. V. Davies and R. Raven provided useful data on the studied islands and Drs. R Selden, J. Berry and Mr. G. Wishart (Gerrin- gong) corrected the manuscript. LITERATURE CITED Baert, L. & R. Jocque. 1993. A tentative analysis of the spider fauna of some tropical oceanic is- lands. Mem. Queensland Mus., 33(2):447-454. Table 3. — Island species and their zoogeographic distribution. Australian endemics Widely distributed Cosmopolitan/pantropical Abracadabrella elegans Bavia aericeps Hasarius adansoni Astia hariola Bianor maculatus Menemerus bivittatus Clynotis severus Cosmophasis bitaeniata Plexippus paykulli Cosmophasis micarioides Cosmophasis thalassina Plexippus petersi Cytaea severa Cyrba ocellata Ergane cognata Cytaea frontaligera Ergane insulana Cytaea mitellata Evarcha infrastriata Cytaea plumbeiventris Gangus longulus Euryattus bleekeri Helpis minitabunda Zenodorus arcipluvius Holoplatys colemani Holoplatys complanata Mopsus mormon Ligurinus bipenicilatus Mopsolodes australiensis Opisthoncus abnormis Servaea vestita Simaetha atypica Simaetha robustior Simaetha tenuidens Simaetha tenuior Tauala minutus "Trite" longula Zenodorus metallescens Zenodorus orbiculatus 234 THE JOURNAL OF ARACHNOLOGY Table 4. — Island genera in comparison with the continental fauna (after Zabka 1991, unpubL). Australian genera Number of species on the Continent the Islands Abracadabrella 3 1 Adoxotoma 2 Afraflacilla 5 Arasia 2 Ascyltus 1 Astia 2 1 Bavia 3 1 Bianor 2 1 Canama 1 Clynotis 1 1 Cocalus 1 Coccorchestes 1 Copocrossa 1 Cosmophasis 6 3 Cyrba 1 1 Cytaea 5 4 Damoetas 1 Diolenius 1? Ergane — 2 Euryattus 4 1 Evarcha 1 1 Erigga 1 Gangus 2 1 Grayenulla 5 Harmochirus 1 Hasarius 2 1 Helpis 3 1 Holoplatys 36 2 Hypoblemum 3 Jacksonoides 7 Jotus 1 Lauharulla 1 Ligonipes 4 Lycidas 22? Maratus 7 Margaromma 1 Megaloastia 1 Menemerus 2 1 Mintonia 1 Mopsolodes 1? 1 Mopsus 1 1 Myrmarachne 10 Ocrisiona 8 Omoedus 1 Opisthoncus 31 1 Palpelius 2 Paraplatoides 5 Plexippus 2 2 Portia 1 Prostheclina 1 Pseudomaevia 1 Pseudosynagelides 6 Rombonatus 1 Table 4. — Continued Australian genera Number of species on the Continent the Islands Servaea 3 1 Simaetha 10 4 Simaethula 8 Sondra 11 Tara 3? Tauala 7 1 Trite 5? 1 Zebraplatys 4 Zenodorus 14 3 Baert, L., J.R Maelfait & K. Desender. 1989. Re- sult of the Belgian 1986-expedition: Araneae, and provisional checklist of the spiders of the Galapagos archipelago. Bull. Inst. Sci. Nat. Bel- gique, 58:29-54. Berland, L. 1934. Les Araignees du Pacifique. In Contribution a 1’ etude du peuplement zoologique et botanique des “les du Pacifique.” Publ. Soc. Biogeogr., 4:155-180. Berry, J.W., J.A. Beatty & J. Proszynski. 1996. Sal- ticidae of the Pacific Islands. I. Distribution of twelve genera, with descriptions of eighteen new species. J. Arachnol., 24:214-253. Berry, J.W., J.A. Beatty & J, Proszynski. 1997. Salticidae of the Pacific Islands. II. Distribution of nine genera, with descriptions of eleven new species. J. Arachnol., 25:109-136. Carlquist, S. 1965. Island Life. A Natural History of the Islands of the World. The Natural History Press, New York. Carlquist, S. 1974. Island Biology. Columbia Uni- versity Press, New York. Davies, V.T. & M. Zabka. 1989. Illustrated keys to the genera of jumping spiders (Araneae: Saltici- dae) in Australia. Mem. Queensland Mus., 27: 189-266. Forster, R.R. 1975. The spiders and harvestmen. Pp. 493-506. In Biogeography and Ecology of New Zealand, Junk, Den Haag. Gillespie, R.G. 1993. Biogeographic pattern of phylogeny in a clade of endemic Hawaiian spi- ders (Araneae, Tetragnathidae). Proc. XII Int. Congr. Arachnol. Brisbane. Mem. Queensland Mus., 33(2):5 19-526. Gillespie, R.G., H.B. Croom & S.R. Palumbi. 1994. Multiple origins of a spider radiation in Hawaii. Proc.. Nat. Acad. Sci., 91:2290-2294. Greenstone, M.H., C.E. Morgan, A.L. Hultsch, R.A. Farrow & J.E. Dowse. 1987. Ballooning spiders in Missouri, USA, and New South Wales, Aus- tralia: family and mass distribution. J. Arachnol., 15:163-170. PATOLETA & ZABKA— SALTICIDS OF AUSTRALIAN ISLANDS 235 Homer, N.V. 1975. Annual aerial dispersal of jumping spiders in Oklahoma (Araneae, Saltici- dae). J. ArachnoL, 2:101-105. Lehtinen, P.T. 1980. Arachnological zoogeography of the Indo-Pacific Region. Proc. 8th Int. Congr. ArachnoL, Pp. 499-504. Lehtinen, P.T. 1996. Origin of the Polynesian spi- ders. Rev. Suisse ZooL, hors SOTe:383-397. Mac Arthur, R.H. & E.O. Wilson. 1967. The The- ory of Island Biogeography. Princeton Univ. Press, Princeton. Main, B.Y. 1981. Australian spiders: Diversity, dis- tribution and ecology. Pp. 808-852, In Ecologi- cal Biogeography of Australia. (A. Keast, ed.). Junk, The Hague, Boston, London. Proszyhski, J. 1992. Salticidae (Araneae) of the Old World and Pacific Islands in several US col- lections. Ann. Zook, 44:87-163. Proszynski, J. 1996. Salticidae (Araneae) distribu- tion over Indonesian and Pacific Islands. Rev. Suisse ZooL, hors serie:53 1-536. Salmon, J.T & N.V. Homer. 1977. Aerial disper- sion of spiders in North Central Texas. J. Arach- noL, 5:153-157. Zabka, M. 1988. Salticidae (Araneae) of Oriental, Australian and Pacific Regions, III. Ann. ZooL, 41:421-479. Zabka, M. 1990. Remarks on Salticidae (Araneae) of Australia. Ann. ZooL Fennici, 190:415-418. Zabka, M. 1991. Studium taksonomiczno-zoogeo- graficzne nad Salticidae (Arachnida: Araneae) Australii. Rozprawa naukowa nr 32., WSR-P Siedlce. Zabka, M. 1993. Salticidae (Arachnida: Araneae) of New Guinea, Australia and adjacent areas. Boll. Acc. Gioenia Sci. Nat., 26:389-394. Manuscript received 1 May 1998, revised 24 Oc- tober 1998. 1999. The Journal of Arachnology 27:236-240 PSEUDOSCORPIONS IN FIELD MARGINS: EFFECTS OF MARGIN AGE, MANAGEMENT AND BOUNDARY HABITATS James R. BelF, Simon Gates^, Alison J. Haughton^ David W. Macdonald, and Helen Smith^: Wildlife Conservation Research Unit, Department of Zoology, Oxford University, South Parks Road, Oxford 0X1 3PS, U.K. C. Philip Wheater and W. Rod Cullen: Department of Environm.ental and Geographical Sciences, Manchester Metropolitan University, Manchester Ml 5GD, U.K. ABSTRACT. Pseudoscorpions (Chthonius ischnocheles (Hermann) and C. orthodactylus (Leach) sensu strictus) were collected using a D-Vac over two-years from 60 field margins at Oxford University farm at Wytham, U.K. Old and new grassland margins were subjected to six different treatments involving spray- ing, non-intervention and four different cutting intensities. Significantly more pseudoscorpions were found in old compared to new margins, suggesting they may be attracted to litter build-up over time. Pseudo- scorpion numbers were reduced on treatments subjected to two cuts annually, particularly when a summer cut was included, although this effect was ameliorated when the cuttings were left. However, pseudo- scorpions were most numerous on treatments which involved no management because of the increase in leaf litter which may replicate a woodland environment. Adjacent hedges appear to buffer the effects of management: margins with adjacent hedges (rather than ditches or tracks) having more individuals. In contrast to results for other invertebrate groups, sowing wildflower seed did not significantly increase the abundance of pseudoscorpions. The effect of different treatments on pseudoscorpion numbers demonstrates that they are useful indicators of the effects of management practice. Leaf litter is the ancestral habitat of pseu- doscorpions, with deeper litter, such as that in woodlands, providing an ideal stable environ- ment (Jones 1970). Few studies have exam- ined them in grasslands despite some species occurring commonly in this ecosystem. Fewer still have considered the role of habitat man- agement on pseudoscorpions in grassland margins. Thus, stable woodland environments have provided nearly all the ecological and taxonomic work for common pseudoscorpion families, such as the Chthonidae, in Britain 'Current address: Department of Environmental and Geographical Sciences, Manchester Metropol- itan University, Manchester Ml 5GD, U.K. -Current address: National Perinatal Epidemiolo- gy Unit, Radcliffe Infirmary, Woodstock Road, Ox- ford 0X2 6HE, U.K. -^Current address: Crop & Environmental Re- search Centre, Harper Adams Agricultural College, Newport, Shropshire TFIO 8NB, U.K. "'Current address: Waveney Cottage, Redgrave Road, South Lopham, Diss, Norfolk IP22 2HL, U.K. (e.g., Gabbutt & Vachon 1963; Gabbutt 1967; Goddard 1976; Wood & Gabbutt 1978). Par- ticular attention has been paid to the ecology and life history of Chthonius ischnocheles (Hermann) and to a lesser extent Chthonius orthodactylus (Leach) sensu strictus, both of which are widespread species. It is not sur- prising that there has been little work on grassland pseudoscorpions since only the Nanolpium species (Garypoidea, Olpiidae) from Africa are consistently collected from grass, probably because grasslands are too ex- posed for many other pseudoscorpions (Jud- son & Heurtault 1996). In the only British ref- erence to grassland pseudoscorpions, Salt et al. (1948) found high densities (19.56 m'^) in agricultural fields in Cambridgeshire, but un- fortunately did not report the species’ names. Grasslands are a marginal habitat for several British species including Chthonius ischno- cheles and C. orthodactylus (Legg & Jones 1988), probably because deep litter is found in large amounts only under hedges and in man-made piles. Despite this, Rapp (1978) 236 BELL ET AL.— PSEUDOSCORPIONS IN FIELD MARGINS 237 compared the distribution of Microbisum con- fusum Hoff, in two contrasting areas of tall grass and mixed prairie. He determined that grazing caused a decline in the number of pseudoscorpions because of adverse changes to moisture and litter depth. Our paper is the first to examine changes in pseudoscorpion numbers within a unique farm scale environ- ment that has been block designed and repli- cated to test for successional processes over a long period of time. Over the last ten years, the Oxford University farm at Wytham has been used to test the effect of management on a variety of other grassland arthropods such as spiders (Baines et al. 1998), butterflies (Fe- ber et al. 1996) and the Hemiptera (Smith et al. 1993), producing a better understanding of the habitat requirements of these groups in an agricultural context. We use this field experi- ment, designed to evaluate the wildlife con- servation benefits of a variety of management techniques for arable field margins, to exam- ine the effects of age, management and prox- imity to mature boundary habitats on pseu- doscorpions. Our hypothesis that low intensity management and stable conditions found in hedges will support more individuals is based on a pseudoscorpion’s sensitivity to litter depth and mutable habitat structures. METHODS Site management. — The surveys took place at the site of an experiment established in 1987 on 60 field margins at Oxford Uni- versity farm (SP 472 097) at Wytham (Smith et al. 1993; Feber et al. 1996). Old margins (n = 60) around the fields existed pre-1987 and are about 1 m wide. In 1988 these old margins were extended 1 m into the field to a total width of 2 m and producing new margins (n = 60). In 1990 sterile strips were cultivated between the margin and the crop, allowing ac- cess to the margins without trampling. Six management treatments in 50 m strips were randomized in a block design across the 60 margins. Treatment A had no management (n = 12); B was cut in spring and autumn and the hay collected (n = 12); C was cut in spring and summer and the hay collected (n = 12); D was cut in summer only and the hay col- lected (n = 12); E was cut in spring and sum- mer and the hay left (n = 6); F was sprayed with herbicide (glyphosate) in the same way as the crop (n = 6). Those treatments with a sample size of 12 (A~D) were subdivided into margins which were sown with wildflower seed in March 1988, and those which were left to colonize without sowing (n = 6 in each case). Cutting took place in spring (April), summer (June) and autumn (September) using an Allen scythe and brush cutters. The cut- tings were left to dry and the hay was col- lected or left depending upon the treatment. Glyphosate (Roundup Biactive®, Monsanto) was applied to six margins (F) at a rate of 3 liters ha“^ at a field volume of 200 liters ha“^ using an Oxford precision sprayer. The habi- tats adjacent to the margins were recorded: ditches (n = 56), hedges (n = 38), and tracks (n = 26). Sampling. — ^Pseudoscorpion species (Chthon- ius ischnocheles and C. orthodactylus) were collected for two field seasons in 1995 and 1996 using a Dietrick Vacuum suction sam- pler (D-Vac). During each season, samples were collected in spring (May), summer (July) and autumn (September). Five sucks of 30 seconds were made, each at a 10 m interval along a 50 m margin, and were aggregated to give one sample. The samples were taken sep- arately from old and new margins. Specimens were separated from debris in the laboratory and stored in 70% methanol. No attempt was made to collect silken chambers and this may have underestimated the true number of fe- males of both species and protonymphs of C. orthodactylus (Gabbutt 1970; Goddard 1976). D-Vac sampling is biased towards collecting samples from the litter layer. We made no at- tempt to collect samples from the soil which may have underestimated actual numbers within each margin. Adult specimens were identified using Legg & Jones (1988), and nymphal stages were identified using Gabbutt & Vachon (1963) and unpublished drawings by Mark Judson (Museum National D’Histoire Naturelle, Paris). Statistical analysis. — Both species, and the six collections over two years, were aggregat- ed to simplify analysis. The Kolmogorov- Smimov two sample test indicated that it was not possible to transform the data to normal- ity, therefore we used non-parametric statistics to test for difference. Non-specific Meddis rank means tests were used, blocking by treat- ments not under investigation and using post hoc analyses to identify ideal rank orders where significant results were obtained (Med- 238 THE JOURNAL OF ARACHNOLOGY dis 1984). The level of significance for all the tests was where P < 0.05. RESULTS AND DISCUSSION A total of 247 specimens (Chthonius is- chnocheles: 57 d' , 47 $ , 54 tritonymphs, 8 deu- tonymphs = 166. C. orthodactylus: 3AS , 249, 21 tritonymphs, 2 deutonymphs = 81) was recorded over the two year period. As a proportion of the total number of invertebrates collected from the field margins during this study (~6 X 10^, excluding Collembola and Acari), the pseudoscorpioe population was less than 0.05%. This highlights the relatively minor role of pseudoscorpions in grassland liL ter compared with woodland litter and soil, in which pseudoscorpions are at relatively high densities (Goddard 1976; Gabbutt 1967). Chthonius ischnocheles were twice as abun- dant and twice as widely distributed over the field margins compared to C. orthodactylus. Overall, pseudoscorpions exploited 43.4% of the available field margin habitat (n = 120) but in only 9.6% of samples (n = 120) did the two species occur together. The ratio of males to females was slightly biased in favor of males for both C. ischnocheles (1.2:1) and C orthodactylus (1.4:1). In both species, tri- tonymphs contributed substantially to the total numbers (C ischnocheles = 32.5%; C, ortho- dactylus = 25.9); but deutonymphs were rare (C ischnocheles = 4.8% ; C. orthodactylus = 2.5). No protonymphs were recorded. Age of margies«“— Significantly more pseu- doscorpions were found in old margins com- pared with new {H = 9.471, P = 0.002, df = 1). This may be a due to litter build-up over time, since deeper litter supports more indi- viduals and suitable prey (Jones 1970). Both Baines et al. (1998) and Frank & Nentwig (1995) recorded more spider species, and more individuals, on older field margins com- pared to new. Although spiders and pseudo- scorpions differ in their habitat requirements, it may be that the longer the field margin has to develop, the more stable and ultimately more suitable the litter environment becomes. Management.- — The timing and the inten- sity of management had a significant impact on the abundance of pseudoscorpions (see Table 1) found in the field margins {H = 12.712, P = 0.026, df = 5, rank order: A>D,E,B>F>C). No management (A) was associated with higher abundances (Table 1) when compared Table 1. — Comparison of the effect of treatment on total pseudoscorpion numbers expressed as a rank. Species counts are also given but were not tested separately and should not be compared with the post hoc rank scores. The test using a non-spe- cific Meddis rank means test was significant {P = 0.026). The post hoc ranks range from 1 (the high- est) to 4 (the lowest) scores; D, E, B are not sig- nificantly different and therefore have the same rank score. Treat- ment Chthonius Chthonius Combined ortho- ischno- spp. dactylus cheles total Post hoc rank A 37 58 95 1 B 4 25 29 2 C 9 10 19 4 D 25 28 53 2 E 4 26 30 2 F 2 19 21 3 Totals 81 166 247 — with other treatments, probably because it al- lowed litter build-up, provided cover and a comparatively stable microclimate. Timing and frequency of cutting are critical for other invertebrate groups (e.g., Morris & Rispin 1988) with summer cuts being deleterious to spiders (Baines et al. 1998). When a summer cut is combined with a spring cut and the clip- pings are collected on both occasions (C), it has a more serious effect on pseudoscorpioe numbers (Table 1). Effectively this manage- ment treatment produces a grassland sward that is short, with minimal litter development and an unstable microclimate. Rapp (1978) re- corded a similar effect under grazing where the abundance of Microbisum confusum was determined by the thickness of the litter and available soil moisture. Frequent cutting of grassland creates a high degree of disturbance and structural alteration through removal of standing vegetation (Morris 1979). Plant ar- chitecture and floral composition are governed by management (Brown & Gibson 1990) which in turn causes the microclimate within the sward to change (Morris 1968; Morris & Rispin 1988). Dramatic fluctuations in the mi- croclimate, particularly humidity and temper- ature, will have a negative impact on pseu- doscorpions, (Weygoldt 1969; Rapp 1978). Chthonius ischnocheles can recover lost water with changes in humidity quickly and will mi- grate in adverse conditions (Caplin 1974; BELL ET AL.— PSEUDOSCORPIONS IN FIELD MARGINS 239 Legg & Jones 1988) but probably not over a sustained period with little vegetation cover. Management treatments B, D and E were similar to each other in terms of the abun- dance of pseudoscorpions (Table 1), but sup- ported fewer individuals than did A. These re- sults suggest that a single summer cut (D) or two cuts which avoided the sunrmier period (B) have less impact than a spring and sum- mer cut with removal of the vegetation (C). When two cuts close together are necessary, leaving the hay on the margins appears to ameliorate the effects of this intensive regime (E). Ideally, we advocate that, in order to achieve a balance with the habitat require- ments of other invertebrate groups, manage- ments B, D and E are acceptable compromis- es, and leaving grass piles on the margin would increase litter availability, but probably encourage weeds (Smith pers. comm.). Our study suggests that pseudoscorpions are sen- sitive to herbicide applications (F), although this is not as deleterious to pseudoscorpions as a spring and summer cut with removal of vegetation (C) (Table 1). Such spraying gen- erates a deeper litter over a period of months, creating an ideal habitat for pseudoscorpions (Jones 1970) but it is suggested that conse- quent changes in microclimate may render conditions unsuitable. Boundary habitats. — The three types of boundary habitat adjacent to the field margins had a significant impact on the numbers of pseudoscorpions found in the field margin {H = 7.286, P — 0.025, df == 2, rank order: H > D > T). Margins adjacent to a hedge (H) had more individuals than those adjacent to a ditch (D), but the poorest adjacent habitat was a track (T). Moisture, light and food availability are often relatively stable in hedgerows (Hance et al. 1990), but in field systems which have a high level of disturbance from man- agement (Morris 1979) these must be in a state of flux (although soil moisture may not change significantly: White & Hassall (1994)). The hedge and, to a lesser extent, the ditch could act as a buffer and possible over- wintering site and refuge from extremes of management. Tracks are less likely to buffer populations as they are highly disturbed by farm traffic and subject to substantial diurnal variations in temperature. Wildflower seeded margins. — No signifi- cant difference was detected in abundance be- tween margins sown with a wildflower seed mixture and those left to be colonized natu- rally {H - 0.969, P = 0.674, df = 1). As most pseudoscorpions live within the litter, their numbers are not as directly affected by the diversity of vegetation in field margins, as are the numbers of spiders, which are dependent on vegetation structural complexity and di- versity (White & Hassall 1994; Baines et al. 1998). However, there are benefits of sowing to other groups of invertebrates, e.g., Maniola jurtina L. (Lepidoptera) (Feber et al. 1996). Conclusions.— “Although pseudoscorpions are not an abundant group in grasslands, they are good indicators of mutable habitat struc- tures. The differences detected among differ- ent management regimes, age structures and adjacent habitats supports our hypothesis that the management of field margins for arthro- pods should be consistent over time and of low intensity creating a suitable litter environ- ment. Hedges clearly buffer the populations of pseudoscorpions within field margins and should be conserved. Although sowing field margins with wildflowers had no detectable effect, this practice benefits other invertebrate groups and should be encouraged. ACKNOWLEDGMENTS We would especially like to thank Mark Judson of the Museum National D’Histoire Naturelle, Paris for his guidance and help with identification. Mark Judson and Mark Harvey (Western Australia Museum, Perth) provided useful information on pseudoscorpion grass- land ecology. We are especially grateful to all those at WildCRU who lent a hand with man- agement, and we would particularly like to thank Paul Johnson and the farm manager, Da- vid Sharpe, for their help. The work was fund- ed by the Ministry of Agriculture Fisheries and Food and the Ernest Cook Trust. LITERATURE CITED Baines, M., C. Hambler, P.J Johnson, D.W. Mac- donald, & H. Smith. 1998. The effects of arable field margin management on the abundance and species richness of the Araeeae (spiders). Ecog- raphy, 21:74-86. Brown, V.K. & C.W.D. Gibson. 1990. The mech- anisms controlling insect diversity in calcareous grasslands. Pp. 128-133. In Calcareous grass- land- ecology and management. (S.H. Hillier, D.W.H. Walton & D.A. Wells, eds.). Bluntisham Books, Bluntisham. 240 THE JOURNAL OF ARACHNOLOGY Caplin, G. 1974. General physiology of some Brit- ish pseudoscorpions in relation to feeding and environmental factors. PhD thesis, University of Manchester. Feber, R.E., H. Smith, & D.W. Macdonald. 1996. The effects on butterfly abundance of the man- agement of uncropped edges of arable fields. J. AppL EcoL, 33:1191-1205. Frank, T. & W. Nentwig. 1995. Ground dwelling spiders (Araneae) in sown weed strips and ad- jacent fields. Acta Ecologica, 16:179-193. Gabbutt, P.D. 1967. Quantitative sampling of the pseudoscorpion Chthonius ischnocheles from beech litter. J. ZooL London, 151:469-478. Gabbutt, P.D. 1970. Sampling problems and the va- lidity of life history analyses of pseudoscorpions. J. Nat. Hist, 4:1-15. Gabbutt, RD. & M. Vachon. 1963. The external morphology and life history of the pseudoscor- pion Chthonius ischnocheles (Hermann). Proc. ZooL Soc. London, 140:75-98. Goddard, S.J. 1976. Population dynamics, distri- bution patterns and life cycles of Neobisium mus- corum and Chthonius orthodactylus (Pseudo- scorpions: Arachnida). J. ZooL London, 178: 295-304. Hance, TH., C. Gregoire-Wibo & P.H. Lebrun. 1990. Agriculture and ground beetle popula- tions. Pedobiologia, 34:337-346. Jones, RE. 1970. The occurrence of Chthonius is- chnocheles (Hermann) (Chelonethi: Chthoniidae) in two types of hazel coppice leaf litter. Bull. British Arachnol. Soc., 1:72-79. Judson, M.L.I. & J. Heurtault. 1996. Nanolpium species (Garypoidea, Olpiidae) on grasses in southern Africa - a new niche for pseudoscorpi- ons. Rev. Suisse ZooL, 00:321-326. Legg, G. & R.E. Jones. 1988. Pseudoscorpions. Synopses of the British Fauna (new series). No. 40:1-159. Meddis, R. 1984. Statistics Using Ranks: A Uni- fied Approach. Blackwell Scientific, London. Morris, M.G. 1968. Differences between the in- vertebrate faunas of grazed and ungrazed chalk grassland: II. The faunas of sample turves. J. AppL EcoL, 5:601-611. Morris, M.G. 1979. Responses of grassland inver- tebrates to management by cutting. II. Heterop- tera. J. AppL EcoL, 16:417-432. Morris, M.G. & W.E. Rispin. 1988. Beetle fauna of Oolitic limestone and the responses of species to conservation management by cutting regimes. Biol. Conserv., 43:87-105. Rapp, W. 1978. Preliminary studies on pseudo- scorpion populations in the soil-grass interface as observed in the Nebraska prairies of the U.S.A. NewsL British Arachnol. Soc., 23:5-7. Salt, G., ES.J. Hollick, F. Raw, & M.V. Brian. 1948. The arthropod population of pasture soil. J. Anim. EcoL, 17:139-150. Smith, H., R.E. Feber, P.J. Johnson, K. McCallum, S. Plesner Jensen, M. Younes, & D.W. Macdon- ald, 1993. The Conservation Management of Arable Field Margins. English Nature, Peterbor- ough. Weygoldt, P, 1969, The Biology of Pseudoscorpi- ons. Harvard University Press. Massachusetts. White, P.C.L. & M. HassalL 1994. The effects of management on spider communities on head- lands in cereal fields. Pedobiologia, 38:169-184. Wood, P.A. & P.D. Gabbutt. 1978. Seasonal verti- cal distribution of pseudoscorpions in beech lit- ter. Bull. British Arachnol. Soc., 4:176-183. Manuscript received 29 April 1998, revised 6 Sep- tember 1998. 1999. The Journal of Arachnology 27:241-248 COMPARATIVE ANALYSES OF EPIGEIC SPIDER ASSEMBLAGES IN NORTHERN HUNGARIAN WINTER WHEAT FIELDS AND THEIR ADJACENT MARGINS Ferenc Toth and Jozsef Kiss: Godollo University of Agricultural Sciences, Department of Plant Protection, H~2103 Godollo, Hungary ABSTRACT. Pitfall trapping was carried out in northern Hungarian winter wheat fields and their ad- jacent margins during the growing seasons of three consecutive years, 1992-1994. The dominant species of both habitats was the wolf spider Pardosa agrestis (Westring). A total of 8403 adult individuals of 19 families of 149 spider species was identified: 118 species from the winter wheat and 118 from the margins with fewer traps. The efficiency of detecting species by trapping was 90%, according to the Baule- Mitscherlich extrapolation model. Provided that the sampling effort is the same in both habitats, traps in the margin may catch higher number of individuals and species, than traps located within the field. Calculations, however, indicate that the field, with an area more than a hundred times larger than that of the margins, has a higher total number of species. Although the spider species spectrum of the field and of the margin had a considerable overlap, the Renkonen similarity index indicates that the spider fauna of the two types of habitats were different. Spider assemblages of the margins were more diverse (Renyi diversity), than those of the fields. The species richness of epigeic spiders in our Hungarian winter wheat fields was high, and it was increased by the presence of margins. Thus, for the purposes of the protection of our fauna and promotion of integrated pest management, establishment and maintenance of margins is strongly desirable. Recent development of the Hungarian ag- riculture shows an increased attention to land use in general. Re-evaluation of former land use (share of field crops, reforestation of areas that are not suitable and econonfical for crop production), implementation of the basic prin- ciples of the “National Strategy for Conser- vation of Biodiversity” (Hungarian Academy of Sciences) reflects the importance of agrar- ian biotopes. Parallel to this, the present de- velopment of plant protection is focusing on the potential of natural enemies in integrated pest management (IPM), which involves maintaining their habitats and applying man- agement practices that have minimal adverse effect. Winter wheat and corn are the two most important crops grown in Hungary. Winter wheat covers about 25% of the arable land. Only a few data sets concerning the spider assemblages of arable lands in Hungary are available. Balogh & Loksa (1956), Samu et al. (1996) and Nemeth (1996) have examined the spider community in alfalfa fields. Ac- cording to a recent bibliography of Hungarian arachnological studies (Szineter & Samu 1995), the present research is the first to study the spider fauna of winter wheat in Hungary. Our preliminary surveys in winter wheat showed that spiders are among the dominant epigeic predators in winter wheat in Hungary (Kiss et al. 1993, 1994, 1998). Thus our study aimed to analyze the spider assemblages of winter wheat fields and their adjacent margins with respect to biotic diversity and the devel- opment of IPM. METHODS Description of study area and traps. — The study area was located in northern Hun- gary in the vicinity of the village Kartal (lat- itude 47°40'). Three winter wheat fields, Kartal 1 (Kl), Kartal 2 (K2) and Jozsefmajor (JM) and their adjacent margins less then 6 km apart were surveyed by pitfall traps in three consecutive years. Diameter of the pit- fall traps was 10 cm. A 2% formalin solution with a drop of detergent was placed in traps as a preservative. The traps were run contin- uously and were emptied weekly, except in winter (in K2) when they were emptied monthly. Hereafter when we refer to a trap row we mean 5 traps placed in a row parallel to the closest field margin. 241 242 THE JOURNAL OF ARACHNOLOGY Figure 1. — A pair of directional pitfall traps in Jozsefmajor field (JM) (1994), Annual precipitation in the region is about 600-650 mm. Direction of the prevailing winds is highly variable. The topsoil is Luvic chernozem, developed on loess, mixed with weathered local andesitic material, that ex- plains the more clayey texture than the loess origin would suggest. During the dry season in summer the topsoil in the fields opens l“-2 cm wide deep cracks. Field margins were abandoned, uncultivated and untreated strips at the edge of the fields, covered by an her- baceous undergrowth and containing a few separate trees {Robinia pseudoacacia) and shrubs (Robinia pseudoacacia, Sambucus ni- gra, Rubus caesius, Prunus spinosa). Coen- ological details are given in Kiss et al. (1997). The K1 field measured 131 hectares. The margin was 2““3 m wide. Fifteen pitfall traps were operated in three parallel rows from late April until harvest in early July 1992. One row of traps was in the margin, parallel to the ecotone (5 m apart from each other), and two more were positioned into the field at incre- ments of 30 and 250 m from the margin. The area of the K2 field was 250 hectares. The margin was 2-3 rn wide. Twenty pitfall traps were operated in four rows parallel to the ecotone from early November 1992 throughout winter until harvest in early July 1993. Five traps were placed in the margin (5 m apart from each other), and the other trap rows were in the field at increments of 20, 50 and 250 m from the margin. The area of the JM field was 61 hectares and the margin was 4-5 m wide. Twenty pit- fall traps were operated in four rows parallel to the ecotone from mid-March until harvest in mid- July 1994. Five traps were placed in the margin (10 m apart from each other) and the other trap lines were placed in the field at increments of 20, 50 and 250 m from the mar- gin. Three pairs of directional pitfall traps, 20 m apart were placed in the JM field, Im from the margin (Fig. 1). A pair of directional traps consisted of two pitfall traps, separated by two transparent U-shaped plastic plates. The plates were 1 meter long, 30 cm tall and were sunk into the ground to a depth of 10 cm. The upper edge of the plates was smeared with a thick layer of vaseline to inhibit climbing by ar- thropods. Traps facing the margin are called 'Dir. M’, whereas traps facing the field are called ‘Dir. Fk Directional traps enabled us to determine whether the spider assemblage of the immediate area of the margin (1 meter within the field) is similar to that of the mar- gin or of the field. Data analyses.— Since immature spiders are difficult to identify, only adults were taken into account in the analyses of extrapolation, similarity and diversity models, and all were identified to species level. Extrapolation: The potential number of species caught in traps was estimated with the Baule-Mitscherlich function (Svab 1981; Samu & Lovei 1995). The equation is: y -= T*(l“e“-*-'’^) where T means the potential number of spe- cies (saturation level), a and b are parameters, X (independent variable) is the cumulative number of individuals, and y (dependent var- iable) equals the cumulative number of spe- cies. Trapping results were randomly sorted and successively simulated an increase in sampling effort. Saturation level of the func- tion best fitting the points was calculated us- ing an iterative least squares grid searching method. Randomization of the trapping data and calculation of the saturation level (T) was repeated 50 times and the means were used in each case as the potential number of species. Potential number of species was calculated for both habitat types separately (field or margin), and combined (field + margin). Similarity: Similarity of the spider assem- blages of different trap rows was calculated with the Reekonee index (Reekonen 1938). The equation is: R = S min(pi;qi) where p^ and qj mean the relative frequency of species number i in habitats p and q. Since operating with relative frequencies, the cap- tures are re-scaled between 0 and 1. So the Renkonen index enables us to compare results of different sampling efforts. In order to make the comparison of spider assemblages of the TOTH & KISS— SPIDERS OF WINTER WHEAT FIELDS AND MARGINS 243 Table 1. — Number of spider species and adult individuals captured in pitfall traps in three Hungarian winter wheat fields and in their adjacent margins (1992-94). Numbers in brackets imply that the results of directional pitfall traps has been added. (K1 == Kartal field 1; K2 == Kartal field 2; JM = Jozsefmajor field. Distance of field traps from the margin is indicated.) Field (year) Margin 20(30) m 50 m 250 m Wheat total Total Number of species K1 (1992) 57 38 31 54 77 K2 (1992-93) 72 34 35 37 55 91 JM (1994) 77 56 57 44 83 (97) 103 (111) Total 118 107 (118) 145 (149) Number of adult individuals K1 (1992) 1073 738 764 1502 2575 K2 (1992-93) 706 245 286 378 909 1615 JM (1994) 882 882 819 719 2420 (3331) 3302 (4213) Total 2661 4831 (5742) 7492 (8403) three different fields more reliable, autumn and winter catches in K2 were not included in this analysis. The similarities were also illus- trated by a dendrogram, which shows the re- sults of a hierarchical cluster-analysis using single linkage and the Renkonen index as dis- tance measure data. Diversity: The Renyi-function (Tothmeresz 1993) was used to characterize species diver- sity of different trap rows. Renyi-diversity: H,. = (InX (N,/NT-)“)/(l-a) where 0 < a, a ^ 1, means the number of individuals of the species number i, T means the total number of species, Nj means the total number of individuals, and a is a scale param- eter. Where the scale parameter is low, the function is more sensitive to rare species, whereas high values of the scale parameter suggests that the function is more sensitive to the dominant species. If a-Al, then (Hs : Shannon diversity). If a ~ 0, then = InT Species richness was expressed by the Margalef-index (Margalef 1958). The equa- tion is: d - (S-l)/ln N where S means the number of species, N means the number of individuals. RESULTS The dominant spider species in our Hun- garian winter wheat fields was Pardosa agres- tis (Westring) (43% of wheat total), followed by Oedothorax apicatus (Blackwall) (16%), Meioneta rurestris (C.L. Koch) (11%), Xysti- cus kochi Thorell (3%), Trichoncoides pisca- tor (Simon) (3%) and Zelotes mundus (Kul- czynski) (3%). The dominant species of field margin spiders was also Pardosa agrestis (17% of margin total), followed by Pardosa prativaga (L. Koch) (7%), Zelotes pedestris (C.L. Koch) (6%), Aulonia albimana (WaL ckenaer) (5%), Hahnia nava (Blackwall) (5%) and Xysticus kochi Thorell (4%). A total of 8403 adult individuals of 149 spi- der species was identified. From the winter wheat, 118 species were collected and simi- larly (with fewer traps), 118 species were col- lected from the margin. There were 87 species (58.4% of total) which occurred in both hab- itats. Margin trap rows collected larger num- ber of individuals and species than field trap rows in the same field. Directional traps at JM collected 911 adults of 69 species. This in- creased the number of species in wheat total (K1 + K2 + JM) by 11 species. The direc- tional trapping added only 4 species to the to- tal (Table 1). According to the Baule-Mitscherlich ex- trapolation model, the potential number of species caught with pitfall traps in these fields given the trap numbers and configuration was 164 (r^ ^ 0.981) for the total area (field + margin), 135 (r^ “ 0.979) for the wheat, and 130 (r^ = 0.980) species for the margin. The model suggests that 116 species is predicted to occur in both habitats. This means a 70.7% potential overlap between the species spec- trum of field and margin, compared to the ob- served overlap of 58.4%. In all the three fields, species composition of trap rows of the same field were highly 244 THE JOURNAL OF ARACHNOLOGY Table 2. — Renkonen similarity indices comparing adult ground spider assemblages captured in pitfall traps positioned in three Hungarian winter wheat fields and in their adjacent margins (1992-94). Indices were computed from relative frequency of species. (K1 = Kartal field 1; K2 = Kartal field 2; JM = Jozsefmajor field. Distance of field traps from the margin is indicated. Dir. M./Dir. F. = directional traps facing the margin/field.) K1 Mar- gin K1 30 m K1 250 m K2 Mar- gin K2 20 m K2 50 m K2 250 m JM Mar- gin JM Dir. M. JM Dir. F. JM 20 m JM 50 m K1 30 m 0.50 K1 250 m 0.45 0.83 K2 Margin 0.32 0.18 0.16 K2 20 m 0.40 0.56 0.51 0.22 K2 50 m 0.38 0.56 0.50 0.19 0.83 K2 250 m 0.41 0.58 0.52 0.21 0.78 0.80 JM Margin 0.40 0.25 0.23 0.44 0.32 0.27 0.30 JM Dir. M. 0.53 0.58 0.57 0.29 0.65 0.59 0.65 0.43 JM Dir. E 0.49 0.52 0.50 0.27 0.55 0.49 0.55 0.46 0.79 JM 20 m 0.44 0.62 0.59 0.20 0.66 0.60 0.65 0.34 0.79 0.79 JM 50 m 0.43 0.60 0.57 0.18 0.67 0.61 0.69 0.31 0.75 0.75 0.89 JM 250 m 0.42 0.57 0.56 0.18 0.54 0.47 0.55 0.30 0.72 0.78 0.80 0.79 similar (R = = 0.72- ■0.89) (Table 2; Fig. 2). Spe- of the margins. Overlap between the number cies composition of margins differed more ei- ther from that of the other margins (R = 0.32- 0.44) or from that of the field trap rows (R = 0.16-0.53). The lowest field-field similarity (R = 0.47) was higher than the highest mar- gin-margin similarity (R = 0.44). Directional trap catches, oriented to capture spiders mov- ing across the edge were highly similar to each other (R = 0.79) and to the field catches (R = 0.72-0.79). The Renyi-diversity of margin trap rows were higher than those of the field trap rows of the same field, regardless of concentrating on rare (when scale parameter is low) or dom- inant (when scale parameter is high) species (Fig. 3). Species richness of the total capture is characterized by a Margalef-index of d = 16.4, whereas the index for wheat is d = 13.5, and d = 14.8 for margin. DISCUSSION Field margin seems to be a more dense and rich habitat than field, since traps in the mar- gins usually catch higher number of individ- uals and species, than traps located within the fields (Kromp & Steinberger 1992; A1 Hus- sein & Liibke-Al Hussein 1995; Samu et al, 1996), This experience, however, does not necessarily means that the total number of species is higher in the margins because the total area of the fields is much larger than that of spider species in the margin vs. field may be as high as 70%. The proportion of species occurring in both habitats is influenced by the sampling effort, so comparability is more re- liable with the Renkonen index than with the species list overlap. Similarity between mar- gin and field was found R = 0.18 (Kromp & Steinberger 1992), R = 0.21-0.53 (Al Hussein & Liibke-Al Hussein 1995), and R - 0.62 (Janssens & De Clercq 1986), whereas in-field similarity was R = 0.82-0.95 (Al Hussein & Liibke-Al Hussein 1995). These data and our findings suggest that margin-field or margin- margin similarity usually remains under 0,5, while field-field similarity values in most cas- es exceed this level. This is explained by that the fields are strongly and repeatedly dis- turbed every year by tillage, harvest, pesticide application and other field works, while oc- casional disturbance in the margins (mowing, pesticide drifting) does not destroy the habitat basically. As a consequence, pioneer spider species, such as Oedothorax spp., Meioneta spp., Erigone spp., Pardosa spp., Trochosa spp., Pachygnatha spp. dominate the Euro- pean arable fields, resulting in a relative uni- formity. Most of these pioneer species are fre- quent in the margins as well, but spider assemblages of the margins are more diverse than those of the fields ( Kromp & Steinberger TOTH & KISS— SPIDERS OF WINTER WHEAT FIELDS AND MARGINS 245 Figure 2. — Similarity of the spider assemblages of three Hungarian winter wheat fields and of their adjacent margins. The dendrogram illustrates the result of a hierarchical cluster analysis using single linkage and the Renkonen index, calculated from relative frequency data of species. (K1 = Kartal field 1; K2 = Kartal field 2; JM = Jozsefmajor field. Distance of field traps from the margin is indicated. Dir. M./Dir. F. = directional traps facing the margin/field.) 1992; A1 Hussein & Liibke-Al Hussein 1995; Samu et al. 1996). Ground spider species rich- ness in these northern Hungarian winter wheat fields and their adjacent margins were higher than those found in other such surveys in Eu- ropean agricultural areas (Table 3). For the ex- planation of this phenomenon further investi- gations are needed. According to the lOBC Technical Guide- lines for arable crops (Boiler et al. 1997) eco- logical compensation areas (reservoirs of pest antagonists, like flowering field margins, groups of trees, ponds, haystacks) have to cover at least 5% of the entire farm surface excluding forests. It means that calculating with a 5m wide field margin, the average field size should not exceed 1-2 ha. As a result of the large scale farming, that was characteristic of Hungary until 1989 (Kiss et al. 1997) we estimate that field margins cover around 1% 246 THE JOURNAL OF ARACHNOLOGY — >« — margin — o — 30m 250m — « — margin — 0— 20m 50m 250m margin 20m 50m 250m Figure. 3. — Renyi-diversity (H„) in three Hungarian winter wheat fields and in their adjacent margins, calculated from relative frequency data of spider species. Where the scale parameter (a) is low, the function is sensitive to the rare species, whereas increasing the scale parameter results in a higher sensitivity to the dominant species. If a^l, then H<,-^Hs (Hg : Shannon diversity). If a = 0, then H„ = InT (T = number of species). (K1 = Kartal field 1; K2 = Kartal field 2; JM = Jozsefmajor field. Distance of field traps from the margin is indicated.) TOTH & KISS— SPIDERS OF WINTER WHEAT FIELDS AND MARGINS 247 Table 3. — Spider species richness in different European pitfall trap studies, according to the Margalef- index (d), computed from the number of individuals (n) and species (S). Habitat n S d Reference Total catch 8403 149 16.4 Present study (Toth & Kiss 1999) Winter wheat 5742 118 13.5 Margin 2661 118 14.8 Winter wheat and sugar beet 101,213 122 10.5 Janssens & De Clercq (1986) Maize 2018 19 2.4 Alderweireldt & Desender (1990) Winter wheat, peas and maize 1235 47 6.5 Gajdos (1992) Winter wheat 4460 80 9.4 Kromp & Steinberger (1992) Winter wheat 5069 41 4.7 Topping & Sunderland (1992) Potato and margin 5145 75 8.7 Steinberger & Kromp (1993) Winter wheat, winter barley and margin 9510 82 8.8 A1 Hussein & Lfibke-Al Hussein (1995) of our arable lands. Owing to the rich epigeic spider fauna, Hungarian agroecosystems have a considerable potential to enhance natural en- emy populations and their diversity by in- creasing the number, width and quality of field margins. ACKNOWLEDGMENTS We would like to express our special thanks to Dr. Konrad Thaler for checking the identi- fication of many problematic species, and Franciska T. Bogdanyi for her useful com- ments on the manuscript. This work was sup- ported by the Soros Foundation, ANKA Foun- dation and OTKA Grant F 017691. LITERATURE CITED Alderweireldt, M. & K. Desender. 1990. Microhab- itat preference of spiders (Araneae) and carabid beetles (Coleoptera, Carabidae) in maize fields. Med. Fac. Landbouww. Rijksuniv. Gent, 55(2b): 501-510. A1 Hussein, LA. & M. Liibke-Al Hussein. 1995. Zur Webspinnenfauna (Arachnida; Araneae) in Getreidefeldem und angrenzenden Feldrainen im Mitteldeutschen Raum. Hercynia N. F. Halle, 29: 227-240. Balogh, J. & 1. Loksa. 1956. Untersuchungen tiber die Zoozonose des Luzemenfeldes. Acta Zool. Hungarica, 2:17-114. Boiler, E.F., C. Malavolta & E. Jorg (eds.). 1997. Guidelines for Integrated Production of Arable Crops in Europe, lOBC Technical Guideline III. lOBC/wprs Bull., 20:3-18. Gajdos, P. 1992. Communities of epigaeic spiders (Araneae) in agricultural cenosis of Malanta and Janikovce near Nitra. Spravy Slovenskej Ento- mol. Spolocnosti, 3:10-17. Janssens, J. & R. De Clerq. 1986. Distribution and occurrence of Araneae in arable land in Belgium. Med. Fac. Landbouww. Rijksuniv. Gent, 51(3a): 973-980. Kiss, J., E. Kozma, I. Toth & F. KMar. 1993. Im- portance of different habitats in agricultural land- scape related to integrated pest management. Landscape and Urban Planning, 27(2-4): 191- 198. Kiss, J., E Kadar, 1. Toth, E. Kozma & F. Toth. 1994. Occurrence of predatory arthropods in winter wheat and in the field edge. Ecologie, 25(2): 127-132. Kiss, J., K. Penksza, F. Toth & F. Kadar. 1997. Eval- uation of fields and field margins in nature pro- duction capacity with special regard to plant pro- tection. Agric. Ecosyst. Environ., 63:227-232. Kiss, J., E Toth, F KMar & R. Barth. 1998. Pred- atory arthropods in winter wheat in northern Hungary. lOBC/wprs Bulletin, 2 1(8): 8 1-90. Kromp, B. & K.H. Steinberger. 1992. Grassy field margins and arthropod diversity: a case study on ground beetle and spiders in eastern Austria (Co- leoptera: Carabidae; Arachnida: Aranei, Opili- ones). Agric. Ecosyst. Environ., 40:71-93. Margalef, R. 1958. Information theory in ecology. General Systematics, 3:36-71. Nemeth, J. 1996. Egy kiserleti lucemas pokfauna- janak kialakulasa es a savos kaszalasi elj^as ha- tasa populacioikra. MSc Thesis, ELTE, Buda- pest. 75 pp. Renkonen, O. 1938. Statistisch-okologische Unter- suchungen fiber die terrestrische Kaferwelt der finnischen Bruchmoore. Ann. Zool. Soc. Zool. Bot. Fennini, 6:1-226. Samu, F. & G.L. Lovei. 1995. Species richness of a spider community: extrapolation from simulat- ed increasing sampling effort. European J. En- tomoL, 92:633-638. Samu, E, G. Voros & E. Botos. 1996. Diversity and community structure of spiders of alfalfa fields and grassy field margins in south Hungary. Acta Phytopath. Entomol. Hungarica, 31(3-4):253- 266. 248 THE JOURNAL OF ARACHNOLOGY Szinet^, Cs. es F. Samu. 1995. Bibliography of ar- achnological articles on the arachnofauna of the Carpathian Basin by Hungarian zoologists. Folia Entomol. Hungarica, 56:241-256. Steinberger, K.H & B. Kromp. 1993. Barberf alien- fange von Spinnen in biologisch und konventi- onell bewirtschafteten Kartoffelfeldem und einer Feldhecke bei St. Veit (Kamten, Osterreich) (Arachnida: Aranei). Carinthia 11, 183/103:647- 656. Svab, J. (1981): Biometriai modszerek a kutatasban. Mezogazdasagi Kiado, Budapest. 557 pp. 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. J. Appl. EcoL, 29:485-491. Tothmeresz, B. 1993. DivOrd 1.50: A program for diversity ordering. Tiscia, 27:33-44. Manuscript received 25 April 1998, revised 20 May 1999. 1999. The Journal of Arachnology 27:249-254 THE EFFECTS OF DIFFERENT RATES OF THE HERBICIDE GLYPHOSATE ON SPIDERS IN ARABLE FIELD MARGINS Alison J. Haughton‘, James R. Bell-, Nigel D. Boatman^ and Andy Wilcox*: 'Crop & Environment Research Centre, Harper Adams University College, Newport, Shropshire TFIO 8NB UK; ^Department of Environmental & Geographical Sciences, Manchester Metropolitan University, Chester St., Manchester Ml 5GD. UK; ^Allerton Educational & Research Trust, Loddington House, Loddington, Leicestershire LE7 9XE. UK ABSTRACT. Field margins are susceptible to agro-chemical spray drift, and the effects of herbicide on spiders in semi-natural habitats have been little studied. In this experiment, an arable field margin was sprayed with three rates of glyphosate (90 g active ingredient/hectare (a. i/ha), 180 g a.i./ha & 360 g a.i./ ha) and control plots left unsprayed. Spiders were sampled monthly (June-October) using a converted garden-vac and adult spiders were identified to species. A total of 23,393 spiders was sampled with the web-spinners representing more than 90% of the individuals. The effects of glyphosate application on the abundance of wandering and web-spinning prey-capture guilds, and the two most abundant species {Gona- tium rubens and Lepthyphantes tenuis) were analyzed using ANOVA F tests. The highest rate of glyphosate consistently reduced the total number of spiders, the numbers of web-spinners, G. rubens and L. tenuis, but not numbers of wandering spiders. Changes in vegetation structure and microclimate caused by the glyphosate are implicated in the reduction of numbers of spiders in plots receiving the highest rate of glyphosate. We conclude that glyphosate drift at rates of more than 360 g a.i./ha (active ingredients per hectare) into arable field margins could result in significant losses of important arthropod predators in farmland and a reduction in spider biodiversity in agroecosystems. In the United Kingdom, arable field mar- gins commonly comprise a boundary (hedge, fence, wall, or ditch) and a grass-dominated boundary strip, and these constituent parts have been shown to be beneficial in enhancing flora, mammals, game birds and insects on ar- able farmland (e.g., Boatman 1994). Arable field margins are important as overwintering sites (Bayram & Luff 1993), permanent hab- itats (Alderweireldt 1994a) and refuges for re- covery (Thomas et al. 1991) for spiders in the agroecosystem. Not only do arable field mar- gins increase the opportunity for enhancing spiders as predators (Alderweireldt 1994a), but they are able to increase spider-biodiver- sity within biologically-impoverished arable land (Duelli et al. 1990). Field margins are susceptible to direct her- bicide applications (Boatman 1989) and also to spray drift by the virtue of their proximity to high-input cropped areas. Glyphosate is a commonly used herbicide and with the devel- opment of herbicide resistant crops, the use of non-selective herbicides like glyphosate is likely to increase (Mueller & Womac 1997). Research into the optimum width of buffer zones for reducing spray drift into sensitive areas has recommended margins in the order of 6 m wide for reduction of the most toxic effects of various pesticides (Marrs et al. 1992; de Snoo 1997). Although the impact of insecticides on spi- der behavior (Samu & Vollrath 1992) and mortality (Everts et al. 1989) has been studied, the effects of herbicide contamination on spi- ders remain little-researched (Raatikainen & Huhta 1968; Asteraki et al. 1992). Spiders are sensitive to changes in vegetation structure, where a highly variable structure provides web-spinners with increased web-site oppor- tunities. Availability of structural support for webs and a suitable micro-climate (ameliorat- ed fluctuations in humidity and temperature) are the most important factors in web site se- lection (Samu et al. 1996). The intrinsic action of herbicide on plants alters both the vegeta- tion structure and therefore microclimate con- ditions, and so it is likely that changes in spi- 249 250 THE JOURNAL OF ARACHNOLOGY der fauna would occur when a habitat is exposed to a herbicide application. Here, we subjected an arable field margin to a herbicide application to establish whether relative spider population, prey-capture guild and number of individuals of abundant species were affected. METHODS Site. — A well established arable field mar- gin was selected on the Allerton Research and Educational Trust’s Loddington Estate in Lei- cestershire, UK. The field margin was domi- nated by couch grass (Elymus repens (L.)) and false oat grass (Arrhenatherum elatius (L.)) and lay adjacent to a dense uncut hawthorn {Crataegus monogyna Jacq.) and blackthorn (Prunus spinosa L.) hedge. The field margin was east-south-east facing on slightly stoney clay soils from the Hanslope Series and the field was sown to winter barley (cultivar: Fighter). Treatments. — Eight replicates of four treatments (90 g active ingredient/hectare (a.i./ ha), 180 g a.i./ha & 360 g a.i./ha glyphosate and control) were randomly applied to adja- cent field margin plots, which measured 12 m X 1 m. The glyphosate (Roundup Biactive®, Monsanto) was applied to the plots at a vol- ume rate of 200 liters/ha and a pressure of 25 bars using an Oxford Precision sprayer on 30 May 1997. Sampling.— “Spiders were sampled using a modified garden- vac (g-vac) (Ryobi RSV3100E). As a relatively new arthropod sampling de- vice, the g-vac has received critical attention. Its sampling efficiency has been reviewed and the machine used in this experiment has been considered to be an effective method of sam- pling spiders (Samu et al. 1997). The g-vac samples comprised 10 sub-samples of 30 sec- ond ‘sucks’ at 1 m intervals along each ex- perimental plot. This approximated to a total sampling area per plot of 0.13 m^. The inver- tebrate samples were cooled immediately and then extracted with an aspirator into 70% al- cohol before being identified. All adult spiders were identified to species, whilst immatures were included in total number of spiders. Spiders were sampled prior to the herbicide application to confirm that plots did not sup- port different abundances of spiders. Spiders were then sampled two weeks post-herbicide application and monthly thereafter. Spiders were sampled from June to October inclusive. Statistical analysis. — Total spider abun- dance data and prey-capturing guild data were log(x + 1) transformed while spider species abundance data were square-root (x + 0.5) transformed. Two-way univariate repeated measures ANOVAs were used to test for dif- ferences in mean number of spiders between treatments because the samples of spiders through the season could not be considered to be independent of each other (Von Ende 1993). Where an interaction between treat- ment and date existed, indicating that the ef- fect of treatment differed between dates, a one-way univariate ANOVA was used to test for differences in mean number of spiders be- tween treatments in each month. Planned comparisons were used to test for differences implicit in the experimental design: we used Least Significant Difference (LSD) tests to de- termine differences between means (Sokal & Rohlf 1995). RESULTS A total of 23,393 spiders from 11 families and 67 species was recorded and the dominant family was the Linyphiidae. Specimens have been deposited at the Liverpool Museum, UK. Pre- treatment. — Spider abundance did not differ between plots prior to treatment appli- cation (ANOVA F(3,28) = 1-31; P = 0.2901). We therefore considered the plots to be similar in spider fauna composition and proceeded with analysis. Total spider abundance.-Two-way re- peated measures ANOVA indicated that there was a significant date X treatment interaction (^(15. 140) ^ 2.69; P < 0.0013), so we analyzed data from individual months. Total abundance of spiders was only significantly different be- tween treatments in September (one-way AN- OVA F(3,28) = 4.01; P < 0.0171), where sig- nificantly fewer spiders were found in the 360 g a.i. /ha treatment than in all other treatments (Table 1). Prey-capture guilds. — Web- spinning adult spiders from the Tetragnathidae, Theridiidae and Linyphiidae and wandering adult spiders from the Thomisidae, Clubionidae, Pisauridae, Zoridae, Oonopidae and Lycosidae were grouped to investigate the treatment effects on these two prey-capture guilds. Table 2 shows the mean number of individuals from the fam- ilies in the two guilds in each of the treat- ments. Web- spinning spiders were dominated HAUGHTON, ET AL.— HERBICIDE AND FIELD MARGIN SPIDERS 251 Table 1 . — Mean total number of spiders in treatments and LSD P values for differences between means in all treatments and 360 g active ingredient/hectare (a.i./ha). Control 90 g a.i./ha 180 g a.i./ha 360 g a.i./ha mean 220.50 205.13 209.38 152.13 P <0.0037 <0.0174 <0.0136 — by the Linyphiidae and were more abundant than wandering spiders, where they represent- ed more than 90% of individuals in these two guilds. Wandering spiders were not found to differ between treatments (repeated measures ANOVA F(3,28) ^ 0-67; P < 0.5779). Two-way repeated measures ANOVA in- dicated that there was a significant treatment by date interaction (F(i2, i^) “ 2.61; P < 0.0042) for web-spinners, so we analyzed data from individual months. The number of web- spinners was significantly different among treatments in August, September and October, where more spiders were found in the control plots than in the 360 g a.i./ha in September and October only (Table 3). Species data.— Only species which oc- curred in sufficient numbers (mean number in- dividuals > 1.5 in each month) were analyzed individually. Only two linyphiid species ful- filled this criterion and showed significantly different mean abundances among treatments. Gonatium rubens (Blackwall 1833) showed different abundances in different treatments (repeated measures ANOVA 4.41; P < 0,0116) in months August to October, where the control and 90 g a.i./ha plots had Table 2. — Mean number of individuals in each family from treatments (June to October), a.i./ha = active ingredient/hectare. Con- trol 90 g a.i./ha 180 g a.i./ha 360 g a.i./ha Wandering spiders Thomisidae — — 0.4 0.3 Clubionidae 0.4 0.4 0.3 0.4 Pisauridae — 0.3 — — Zoridae — 0.1 0.1 0.1 Oonopidae — — 0.1 — Lycosidae 2.6 2.4 1.8 4.3 Web spinners Tetragnathidae 0.8 0.6 0.8 1.3 Theridiidae 6.8 6.3 10.9 7.3 Linyphiidae 29.2 27.3 32.7 24.1 significantly more individuals (LSD P < 0.0043; LSD P < 0.0072 for control and 90 g a.i./ha respectively) than the 360 g a.i./ha treatment. Lepthyphantes tenuis (Blackwall 1852) showed different abundances in different treatments in September and October (repeat- ed measures ANOVA 28) “ 7.63; P < 0.0007), where each of the other treatments had significantly more individuals than the 360 g a.i./ha treatment (Table 4). DISCUSSION General effect of treatment. — Applica- tions of glyphosate at 360 g a.i./ha signifi- cantly reduced the abundance of total spiders, web-spinners, Gonatium rubens and Lepthy- phantes tenuis, but not of wandering spiders. The lower rates of herbicide had little or no effect on the abundance of spiders per se; however, this study does not take into effect possible changes in wandering and mating. The initial effects of the herbicide on the total number of spiders and prey-capture guilds were insignificant, but became more profound as the season progressed. Therefore, it is assumed that spiders are not affected di- rectly by glyphosate (which is generally non- toxic to animals), but indirectly by modifica- tions of other factors, such as habitat, prey availability and microclimatic conditions. The time taken for the herbicide to act on vege- tation and change the habitat sufficiently for spiders to exert preferences clearly takes months rather than weeks. Where such effects are widespread, numbers of spiders may be low in the following spring, which is a time when spiders are a determining factor in aphid population dynamics in wheat crops (Coc- quempot & Chambon 1990). Thus, our single season study indicates that the longer term ef- fects of herbicide on spiders as biocontrol agents and spider species diversity in agroe- cosystems are of concern. Wandering spiders. — The highest rate of herbicide did not significantly reduce the 252 THE JOURNAL OF ARACHNOLOGY Table 3. — Comparison of mean number of web-spinners between different treatments & LSD P values for differences between means in all treatments and 360 g active ingredient/hectare (a.i./ha). Treatment August September October mean P mean P mean P control 10.00 ns 28.0 <0.0007 85.25 <0.0061 90 g a.i./ha 8.25 ns 24.75 <0.0059 82.13 <0.0159 180 g ai./ha 19.25 <0.0226 29.86 <0.0004 82.50 <0.0230 360 g a.i./ha 7.25 — 16.38 — 66.50 — number of wandering spiders. Wandering spi- ders generally contain few examples of sten- ophages (Nentwig 1986) and they may be more adept at finding suitable food items in disturbed habitats due to their prey-capture strategy (Young & Edwards 1990). Thus, a combination of feeding strategy and an avail- able diverse prey source may not have suffi- ciently deterred the wandering spiders from using the herbicide treated plots. Vegetation structure can influence not only wandering spider prey recognition (Rovner 1980) but also mate detection (Uetz & Strat- ton 1982). The indirect effects of herbicide on the ability of spiders to detect mates was not recorded, and we suggest that long-term ex- periments should concentrate on mating suc- cess and feeding ability to investigate any cor- relations with herbicide use. Web-spinners. — The action of herbicide on vegetation results in sparse cover and re- duced vegetation height (Raatikainen & Huhta 1968) as plants lose their vigor. Web-spinning spiders rely on vegetation structure to provide both web-attachment sites and appropriate hu- midity (Greenstone 1984; Young & Edwards 1990; White & Hassall 1994). Unlike wan- dering spiders, web- spinning linyphiids tend Table 4. — Comparison of mean number of Lep- thyphantes tenuis between different treatments & LSD F values for differences between means in all treatments and 360 g active ingredient/hectare (a.i./ ha). Treatment September & October mean P control 34.38 <0.0002 90 g a.i./ha 34.19 <0.0005 180 g a.i./ha 31.75 <0.0072 360 g a.i./ha 24.38 — to have preferences for specific prey type (Al- derweireldt 1994b). Many web-spinning spi- ders, therefore, may not utilize sub-standard habitat with a poor prey availability, since they invest energy in web-building (Uetz 1991). The web-spinners in this study repre- sented the dominant prey-capture guild and indicated that higher levels of herbicide re- sulted in unfavorable habitat. Such losses of important farmland spiders from herbicide misapplications could be significant in terms of conservation of spiders in agroecosystems and in enhancing spiders as predators. Gonatium rubens: This linyphiid is a litter species (McFerran et al. 1994) and it showed a preference away from heavily sprayed plots. Although autecological literature about G. rubens is sparse, as a web-spinning spider it has similar habitat requirements as those out- lined above. It must be concluded that all, or a combination of, abundance of web-building sites, availability of prey and level of humid- ity were sub-standard. Lepthyphantes tenuis: The most abundant spider in the British agroecosystem is L. ten- uis (Topping & Lovei 1997). This linyphiid builds webs at 10 cm above the ground and is completely dependent upon web-building for prey ( Alder weireldt 1994b). As vegetation height is reduced under exposure to herbicide (Raatikainen & Huhta 1968), the ideal web- building height for L. tenuis may become dis- placed to a height with reduced humidity. Aphids form a large part of the diet of L. ten- uis (Alderweireldt 1994b) and the spider can reduce the aphid {Rhopalosiphum pad! L.) population on wheat plants by 34% (Maesour & Heimbach 1993). Thus, Lepthyphantes ten- uis is an important predator in farmland and reductions caused by herbicide applications should be considered against the benefits of biocontrol. HAUGHTON, ET AL.— HERBICIDE AND FIELD MARGIN SPIDERS 253 Conclusions. — Herbicide applications at higher rates reduce the abundance of impor- tant predators. Field margins, which are val- ued as refuges for farmland spiders during winter and periods of disturbance, are suscep- tible to herbicide spray drift and may suffer losses in spider fauna. Reduced herbicide use in and near field margins is suggested here and elsewhere (Young & Edwards 1990) as a way of enhancing spider populations in agroe- cosystems not only for biocontrol but also for conservation of spider biodiversity. ACKNOWLEDGMENTS We wish to thank the farm manager at the Loddington Estate, Phil Jarvis for access to the field site and Paul Johnson at the Univer- sity of Oxford for statistical advice. A.J. Haughton was funded by the Higher Educa- tion Funding Council for England. LITERATURE CITED Alder weireldt, M. 1994a. Habitat manipulations increasing spider densities in agroecosystems: possibilities for biological control? J. Appl. En- tomol., 118:10-16. Alderweireldt, M. 1994b. Prey selection and prey capture strategies of linyphiid spiders in high in- put agricultural fields. Bull. British Arachnol. Soc., 9:300-308. Asteraki, E.J., C.B. Hanks & R.O. Clements. 1992. The impact of the chemical removal of the hedge-base flora on the community structure of carabid beetles (Col., Carabidae) and spiders (Araneae) of the field and hedge bottom. J. Appl. Entomol., 113:398-406. Bayram, A. & M.L. Luff. 1993. Winter abundance and diversity of Lycosids (Lycosidae, Araneae) and other spiders in grass tussocks in a field mar- gin. Pedobiologia, 37:357-364. Boatman, N.D. 1989. Selective weed control in field margins. Brighton Crop Protection Confer- ence - Weeds, 1989:785-794. Boatman, N.D. (ed.). 1994. Field Margins: Inte- grating Agriculture and Conservation. BCPC Famham, UK. Cocquempot, C. & J-P. Chambon. 1990. L’activite des Araignees et son incidence sur la limitation des populations de pucerons des biocenoses cer- ealieres. Rev. Ecol. Sol., 27:205-219. De Snoo, G.R. 1997. Arable flora in sprayed and unsprayed crop edges. Agric. Ecosystems and Environ., 66:223-230. Duelli, R, M. Studer, I. Marchand & S. Jacob. 1990. Population movements of arthropods be- tween natural and cultivated areas. Biol. Con- serv., 54:193-207. Everts, J.W., B. Aukema, R. Hengeveld & J.H. Koeman. 1989. Side-effects of pesticides on ground-dwelling predatory arthropods in arable ecosystems. Environ. Poll., 59:203-225. Greenstone, M.H. 1984. Determinants of web spi- der species diversity: vegetation structural diver- sity vs. prey availability. Oecologia, 62:299-304. McFerran, D.M., W.I. Montgomery & J.H. Mc- Adam. 1994. The impact of grazing on com- munities of ground-dwelling spiders (Araneae) in upland vegetation types. Proc. Royal Irish Acad., 94B: 119-126. Mansour, F. & U. Heimbach. 1993. Evaluation of lycosid, micryphantid and linyphiid spiders as predators of Rhopalosiphum padi (Horn: Aphid- idae) and their functional response to prey den- sity - Laboratory experiments. Entomophaga, 38: 79-87. Marrs, R.H., A.J. Frost, R.A. Plant & P. Lunnis. 1992. The effects of herbicide drift on semi-nat- ural vegetation: the use of buffer zones to mini- mize risks. Aspects of Appl. Biol., 29:57-64. Mueller, T.C. & A.R. Womac. 1997. Effect of for- mulation and nozzle type on droplet size with isopropylamine and trimesium salts of glyphos- ate. Weed TechnoL, 11:639-643. Nentwig, W. 1986. Non-webbuilding spiders: prey specialists or generalists? Oecologia, 69:571- 576. Raatikainen, M. & V. Huhta. 1968. On the spider fauna of Finnish oat fields. Ann. Zool. Fennica, 5:254-261. Rovner, J.S. 1980. Vibration in Heteropoda vena- toria (Sparassidae): a third method of sound pro- duction in spiders. J. Arachnol., 8:193-200. Samu, F. & E Vollrath. 1992. Spider orb web as bioassay for pesticide side effects. Entomol. Exp. App., 62:117-124. Samu, E, K.D. Sunderland, C.J. Topping & J.S. Fenlon. 1996. A spider population in flux: se- lection and abandonment of artificial web-sites and the importance of intraspecific interactions in Lepthyphantes tenuis (Araneae: Linyphiidae) in wheat. Oecologia, 106:228-239. Samu, E, J. Nemeth & B. Kiss. 1997. Assessment of the efficiency of a hand-held suction device for sampling spiders: improved density estima- tion or oversampling? Ann. Appl. Biol., 130: 371-378. Sokal, R.R. & EJ. Rohlf. 1995. Biometry. 3rd ed. W.H. Freeman & Co., New York. Thomas, M.B., S.D. Wratten & N.W Sotherton. 1991. Creation of ‘island’ habitats in farmland to manipulate populations of beneficial arthro- pods: predator densities and emigration. J. Appl. EcoL, 28:906-917. Topping, C.J. & G.L. Lovei. 1997. Spider density and diversity in relation to disturbance in agroe- 254 THE JOURNAL OF ARACHNOLOGY cosystems in New Zealand with a comparison to England. New Zealand J. EcoL, 21:121-128. Uetz, G.W. 1991. Habitat stracture and spider for- aging. Pp. 325-348. In Habitat Stracture: The Physical Arrangement of Objects in Space. (S.S. Bell, E.D. McCoy & H.R. Muschinsky, eds.). Chapman & Hall, London. Uetz, G.W., & G.E. Stratton. 1982. Acoustic com- munication and reproductive isolation in spiders. Pp. 123-159. In Spider Communication: Mech- anisms and Ecological Significance. (P.N. Witt & J.S. Rovner, eds,). Princeton Univ. Press, Prince- ton, New Jersey. Von Ende, C.N. 1993. Repeated measures analysis: growth and other time-dependent measures. Pp. 113-137. In Design and Analysis of Ecological Experiments. (S.M. Scheiner & J. Gurevitch, eds.). Chapman & Hall, London. White, P.C.L. & M. HassalL 1994. Effects of man- agement on spider communities of headlands in cereal fields. Pedobiologia, 38:169-184. Young, O.P. & G.B. Edwards. 1990. Spiders in United States field crops and their potential effect on crop pests, J. ArachnoL, 18:1-27. Manuscript received 29 April 1998, revised 11 No- vember 1998. 1999. The Journal of Arachnology 27:255-261 A FAUNISTIC AND ZOOGEOGRAPHICAL REVIEW OF THE SPIDERS (ARANEAE) OF THE BALKAN PENINSULA Christo Deltshev: Institute of Zoology, Bulgarian Academy of Sciences, Blvd. Tsar Osvoboditel 1, 1000-Sofia, Bulgaria ABSTRACT. The Balkan Peninsula is home to 1409 species, included in 337 genera and 47 families. This number was established after a critical review of the existing literature and taxonomic revision of some available collections containing spider material from this region. The highest number of species is recorded for the territories of Bulgaria (775), Greece (642), Croatia (615) and Serbia (508). This biodi- versity depends not only on the size of the regions, but also on the degree of exploration by researchers. The territories of Albania, Turkey, Montenegro and Bosnia are less well investigated. According to their current distribution, the established 1409 species can be classified into 24 zoogeograpical categories, grouped into four complexes (widely distributed, European, Balkan endemics, and Mediterranean). The largest number of species belongs to the widely distributed complex, but the most characteristic are the Balkan endemics. Their established number (379 species) is high and reflects the local character of the fauna. This phenomenon can be attributed to the relative isolation of the mountains compared with the lowlands, in the context of paleo-environmental changes since Pliocene. Their high percentage (26.9%) suggests an important process of autochtonous speciation. Thus, the Balkan Peninsula can be considered as a main center of speciation for the European araneofauna. The spider fauna of the Balkan Peninsula is comparatively well- studied due to the efforts of many araneologists from different coun- tries; but the first large, significant work con- cerning the spiders of all territories of the re- gion came from Drensky (1936). He reported 1066 species from 35 families which consti- tuted a review of all literature available at that time. Some years later Hadjissarantos (1940) compiled all faunistic data about the spiders of continental Greece. Nikolic & Polenec (1981) combined the data concerning Yugo- slavian spiders and reported 1022 species from this country. More recent publications list the fauna of Bulgaria, Greece, Serbia, Macedonia, Montenegro and part of Turkey (Brignoli 1968, 1971, 1972, 1974a, b, 1976, 1977, 1979, 1984, 1986; Deeleman 1976, 1978, 1988, 1993; Deltshev 1979a, b, 1983a, b, 1985, 1988, 1990, 1993, 1996, 1997a, b; Deltshev & Curcic 1997; Deltshev & Paraschi 1990; Thaler 1996; Thaler & Knoflach 1991, 1993, 1995; Wunderlich 1980, 1985, 1994a, b, c). These contributions are a result of in- tensive faunistic research; and the accumula- tion of new data makes possible a critical tax- onomic and faunistic review, together with a zoogeographical analysis. STUDY AREA The Balkan Peninsula is situated in the southeastern part of Europe. The northern bor- der follows the rivers Danube (including its delta), Sava and Soca, and through Gorizia and Monfalcone reaches the line of the Gulf of Trieste. Its western border follows the line of Adriatic and Ionian coast including the is- lands. The eastern border passes to the east of the Aegean Islands Sirina, Astipalea, Amor- gos, Miconos, Tinos, Andros, Skiros, Limnos, and Imros, continues along the Dardanelles, goes across the Marmara Sea and, through the Bosphorus, reaches the Black Sea coast. The southernmost point of the Balkan Peninsula region is Crete and the islands of Gavdos, Ai- duronisi, and Kufonisi (Fig. 1). The material treated herein can be divided into two major parts: the first comprises a crit- ical incorporation of all available records from the literature concerning the distribution of spiders on the Balkan Peninsula; the second concerns a revision of all of the existing ma- terial from Drensky’s collection. The geographical areas and their abbrevia- tions used in the text, are as follows: A1 == Albania, BG — Bulgaria, CT = Crete, CR = Croatia, GR = Greece, BS = Bosnia, MA = 255 256 THE JOURNAL OF ARACHNOLOGY Figure 1. — Map of the Balkan Peninsula. Abbreviations: A1 = Albania; BG = Bulgaria; CT = Crete; CR = Croatia; GR = Greece; BS = Bosnia; MA = Macedonia; MNG = Montenegro; RO = Romania; SB = Serbia; SL = Slovenia; TR = Turkey. Macedonia, MNG = Montenegro, RO = Ro- mania, SB = Serbia, SL = Slovenia, and TR = Turkey. The data concerning the general zoogeographic al distribution are taken mainly from Platnick (1989, 1993, 1997). The zoo- geographical categories used and their abrev- iations are as follows: WD = widely distrib- uted, COS = cosmopolitan, PPT = Palearctic-Paleotropic, H = Holarctic, OW = Old World, P = Palearctic, WP = west Pale- arctic, ECA = European-central Asian, E = European, HE = Holoeuropean, MEE = mid- dle east-European, MSE = middle south Eu- ropean; MSEE = middle southeast European, EE = east European; SE = south European, SEE = southeast European, PO ^ Pontic, BK = Balkans, M = Mediterranean, HM = Hol- omediterranean, EM = east Mediterranean, NM — north Mediterranean, NEM = north- east Mediterranean, SEM = southeast Medi- terranean, BKMA = Balkan-Asia Minor, POM = Pontic-Mediterranean. RESULTS AND DISCUSSION Species composition. — The spider fauna of the Balkan Peninsula is represented by 1409 species, included in 337 genera and 47 fami- lies (Table 1). This number is established after DELTSHEV— SPIDERS OF THE BALKAN PENINSULA 257 Table 1. — The families and number of species and endemic taxa. Family Number of species according to Drensky (1936) Actual number of species Actual number of endemic species Actual number of endemic genera Atypidae 2 2 Ctenizidae 5 6 5 Nemesidae 4 7 4 Filistatidae 4 3 Sicaridae 1 1 Scytodidae 1 1 Leptonetidae 18 18 3 Pholcidae 6 16 9 Segestridae 4 5 2 Dysderidae 40 106 81 9 Oonopidae 1 5 2 Palpimanidae 1 Mimetidae 4 5 Eresidae 2 3 Oecobiidae 1 4 Uloboridae 5 6 Nesticidae 6 9 6 2 Theridiidae 80 91 9 Theridiosomatidae 1 1 Anapidae 1 2 1 Mysmenidae 1 Linyphiidae 172 330 112 3 Tetragnathidae 17 26 1 Araneidae 60 55 Lycosidae 80 83 1 Pisauridae 4 3 Oxyopidae 3 3 Zoropsidae 1 Agelenidae 32 64 36 1 Cybaeidae 2 4 1 Argyronedidae 1 1 Desidae 1 Hahnidae 8 12 5 1 Dictynidae 15 23 1 Amaurobiidae 24 37 19 Titanoecidae 6 6 Anyphoenidae 2 2 Liocranidae 22 19 4 Clubionidae 29 36 3 2 Corinnidae 3 3 Zodariidae 11 23 9 Prodidomidae 2 Gnaphosidae 110 131 22 Zoridae 4 10 2 Heteropodidae 6 5 2 Philodromidae 45 37 2 Thomisidae 55 70 4 Salticidae 140 130 14 258 THE JOURNAL OF ARACHNOLOGY Table 2. — Distribution of species and endemic taxa in different regions of Balkan Peninsula. Region Number of species according to Drensky (1936) Actual number of species Actual number of endemic species Actual number of endemic genera Solvenia 108 216 26 3 Croacia 466 615 66 4 Bosnia 42 87 42 2 Serbia 453 508 10 Montenegro 15 102 29 3 Macedonia 455 394 21 1 Albania 7 73 10 2 Bulgaria 697 775 55 2 Romainia 35 47 7 1 Turkey 83 5 Greece 241 642 156 Crete 59 42 4 a critical review of all available records from the literature concerning the spiders in the Balkan Peninsula and a revision of all existing materials of Drensky’s collection. The number of species is high compared with the number of spiders recorded from oth- er parts of Europe: France -1400 (Jones et ah 1990); Russian Plain -1(X)1 (Michailov 1997); Alps -1000 (Thaler 1980); Germany -925 (Koponen 1993); Switzerland -875 (Maurer & Hanggi 1990); England & Wales -624 (Rob- erts 1987). The number of families is also high compared with the data for the world - 95 (Platnick 1997); Switzerland -39 (Maurer & Hanggi 1990); Russian Plain -35 (Michai- lov 1997). Best represented are the families Linyphiidae (327 species or 23.4%), Saltici- dae (130 species or 9.3%), Gnaphosidae (129 species or 9.2%) and Dysderidae (106 species or 7.6%). The genera with the highest number of species are: Troglohyphantes (53), Lepthy= phantes (49), Dysdera (38), Zelotes (38), Xys- ticus (37), Pardosa (35) and Tegenaria (31). The genus Troglohyphantes is a remarkable faunistic phenomenon since from all 53 spe- cies 52 are the Balkan endemics, distributed mainly in caves. Deeleman- Reinhold (1978) concluded that the present distribution and morphological diversity of Troglohyphantes in the Balkan Peninsula represents of a re- peated processes of expansion and contraction of its range. The representation of the genera Dysdera (28 endemics of 38 species), Lepthy- phantes (18 endemics of 49 species) and Te- genaria (17 endemics of 31 species) is also due to expansion in caves, woodlands and highlands. Present-day examples of cave pen- etration are the species Lepthyphantes cen- tromeroides and L. spelaeorum, comparative- ly widespread in the Balkan peninsula. They occur in caves but also in the humus and ground detritus, and active subterranean col- onization is indicated (Deeleman-Reinhold 1978). The highest number of species is recorded for the territories of Bulgaria (775), Greece (642), Croatia (614) and Serbia (508). This richness, however, depends not only on the size of the regions, but also on the degree of exploration by araneologists (Table 2). The territories of Albania, Turkey, Montenegro and Bosnia are less well-explored. Zoogeographical analysis. — According to their current distribution, the established 1409 species can be classified into 24 zoogeograp- ical categories, grouped into 4 complexes (Figs. 2, 3). Best represented is the complex of widely distributed species (WD) (COS + PPT -f- H + OW + P + WP + EC A), represented by 533 species (38.1%). Within the WD complex, Palearctic species are dominant (75.4%), fol- lowed by Holarctic (19.9%), Cosmopolitan (3.8%) and Palearctic-Paleotropic (0.2%). The complex includes especially widespread spe- cies associated with lowlands, buildings, woodlands and high altitude zones of moun- tains. The Balkan endemics complex (BK) forms the second largest group and comprises 379 species (26.9%). The established number is high and reflects the local character of the fau- DELTSHEV— SPIDERS OF THE BALKAN PENINSULA 259 Figure 2. — The main zoogeographic complexes in the spider fauna in Balkan Peninsula, showing the number of species represented in each. na. The endemics are best represented in Greece (156), Croatia (66), Bulgaria (55), Bosnia (42) and Crete (40). It should be em- phasized that of the established 14 endemic genera (Antrohyphantes, Barusia, Cryphoeci- na, Fageiella, Folkia, Icariella, Lasconia, Macedoniella, Minotauria, Protoleptoneta, Parastalita, Rhodera, Stalagtia, Sulcia) for the Balkan Peninsula, only three of them {An- throhyphantes, Macedoniella, Protoleptoneta) are distributed in the east of the Balkan Pen- insula. Especially interesting is the distribu- tion of the genus Antrohyphantes, found only at high altitude zones and caves of the eastern part of the region (Bulgaria). It is related to the genus Fageiella, an endemic from the caves of the western part of the Balkan Pen- insula (Bosnia, Montenegro). Their allopatric distribution indicates that they had already separated before the establishment of the Var- dar tectonic zone (Deltshev 1996). This sug- gests that these two genera are paleoendemics. The largest fraction of endemics was en- countered mainly in caves, coastal sites, woodlands and high altitude zones. According to their ranges, the endemics belong to two principal faunistic complexes: Mediterranean and European. The Mediterranean elements are distributed in caves, forests, coastal sites and high altitudes, while the European ele- ments are distributed mainly in high altitude sites and forests. This phenomenon can be re- garded as a result of the relative isolation of the mountains compared with the lowlands, in the context of paleo-environmental changes since the Pliocene (Deltshev 1996). The European complex (E) (HE + MEE + MSE + MSEE + EE + SEE + PO) includes 300 species (21.3%). Within it, the Holoeu- ropean species are dominant (72.7%), wide- spread mainly in mountains. The middle southeast European (9.0%), southeast Euro- pean species (9.0%), and east European spe- cies (7.4%) are comparatively well represent- ed. The complex comprises widespread spiders in Europe and the Balkan Peninsula which inhabit both lowlands and mountains. Figure 3. — Zoogeographical types in the spider fauna in Balkan Peninsula, showing the number of species represented in each. 260 THE JOURNAL OF ARACHNOLOGY Interesting is the group of European mountain species, best represented in the forest and sub- alpine belts. The last complex (M) includes 195 species that occur in the Mediterranean area (HM + EM + WM A NM + SE + NEM + SEM + BKMA + POM) or a part of it. This complex forms 13.8% of the total spider fauna of the Balkan Peninsula, but the real percent is prob- ably much higher because a large part of the Balkan endemics have a Mediterranean origin. Most of these species are widely-distributed in the Mediterranean region. Very interesting are the mountain-Mediterranean species {Acu- lepeira taiishia and Pardosa incerta), which may be regarded as ancient elements in the high mountains. Conclusioiis. — The faunistic diversity of the 1409 spider species shows that the Balkan Peninsula is a territory of considerable species richness. This conclusion is supported also by the existence of 379 endemic species. The un- even species richness in different parts of the Balkan Peninsula is due mainly to the degree of exploration by researchers. In a zoogeo- graphical respect, the widely distributed spi- ders (WD) are dominant. However, the most characteristic faunal element is the Balkan en- demics (BK). Their number is high, and their faunistic composition reflects the local char- acter of the fauna. This phenomenon can be explained by the relative isolation of the mountains compared with the lowlands, in the context of paleo-enviromental changes that have occurred since the Pliocene. The high percentage of the Balkan endemics (26.9%) suggests an important process of autochtho- nous speciation. Thus, the existing data sug- gest that the Balkan Peninsula represents one of the main centers of speciation in Europe. ACKNOWLEDGMENTS I am especially indebted to my colleagues G. Blagoev, Dr. D. Dobrev and S. Lazarov for the help with computerizing the data. LITERATURE CITED Brignoli, P.M. 1968. Ueber grechische Leptoneti- dae. Senckenbergiana BioL, 49:259-264. Brignoli, P.M. 1971. Su alcuni Leptyphantes di Creta. Fragm. EntomoL, 7:231-241. Brignoli, P.M. 1972. Su alcuni ragni cavemicoli di Corfu. Rev. Suisse ZooL, 79:861-869. Brignoli, PM. 1974a, Ragni di Grecia VI. Specie nuove o interessanti delle isole lonie e della Mo- rea. Rev. Suisse ZooL, 81:155-175. Brignoli, P.M. 1974b. Ragni di Grecia VII. Rac- colte in grotte delF Attica del Dr. P. Strinati. Rev. Suisse ZooL, 81:493-499. Brignoli, P.M. 1976. Ragni di Grecia IX. Specie nuove o interessanti delle famiglie Leptonetidae, Dysderidae, Pholcidae ed Agelenidae (Araneae). Revue Suisse ZooL, 83:539-578. Brignoli, P.M. 1977. Ragni di Grecia X. Nuovi dati SjLilla Grecia continentale ed insulare. Rev. Suisse ZooL, 84:937-954. Brignoli, P.M. 1979. Ragni di Grecia XI. Specie nuove o interessanti, cavemicole et epigee. Rev. Suisse ZooL, 86:181-202. Brignoli, P.M. 1984. Ragni di Grecia XII. Nuovi dati su varie famiglie (Araneae). Rev. Suisse ZooL, 91:281-321. Deeleman-Reinhold, C.L. 1976. Distribution pat- terns in European cave spiders. First Intern. Symp. Cave BioL Cave Paleont., Cape Town. Pp. 25-35. Deeleman-Reinhold, C.L. 1978. Revision of the cave-dwelling and related spiders of the genus Troglohyphantes Joseph (Linyphiidae), with spe- cial reference to the Yugoslav species. Slov Akad. Znat. Umet., Ljubljana. 218 pp. Deeleman-Reinhold, C.L. 1993. The genus Rhode and harpacteine genera Stalagtia, Folkia, Mino- tauria and Kaemis (Araneae, Dysderidae) of Yu- goslavia and Crete, with remarks on the genus Harpactea. Rev. ArachnoL, 10:105-135. Deltshev, C. 1979a. The origin, formation and zoo- geography of troglobitic spiders of Balkan Pen- insula. Symp. ZooL London, 42:345-351. Deltshev, C. 1979b. Contribution to the study of cave spiders (Araneae) in Greece. Four new spe- cies of Crete and Thera. Acta ZooL Bulgarica, 13:78-92. Deltshev, C. 1983a. A contribution to the taxonom- ical study of sylvaticus group of genus Centrom- erus E Dahl (Araneae, Linyphiidae) in Bulgaria. Acta ZooL Bulgarica, 21:53-58. Deltshev, C. 1983b. A contribution to the taxo- nomical and faunistical study of genus Lepthy- phantes Menge (Araneae, Linyphiidae) from Pi- rin Mountain. Acta ZooL Bulgarica, 23:25-32. Deltshev, C. 1985. New data concerning cave spi- ders (Araneae) in Greece with description of a new Leptonetela (Araneae, Leptonetidae). Acta ZooL Bulgarica, 27:41-45. Deltshev, C. 1988. The genus Fageiella and An- trohy phantes in cave of Balkan Peninsula. TUB- Documentation, 38:293-299. Deltshev, C. 1990. A critical review of genus Coe- lotes Blackwall in Bulgaria with description of a new species {Coelotes drenskii sp. n.) (Araneae, Agelenidae). Acta ZooL Bulgarica, 40:29-43. Deltshev, C. 1993. The genus Tegenaria Latreille DELTSHEV— SPIDERS OF THE BALKAN PENINSULA 261 in Bulgaria: A critical review with description of two sibling species (Arachnida, Araneae: Age- lenidae). Ben Naturwiss.-med. Ver. Innsbruck, 80:167-174. Deltshev, C. 1996. The origin, formation and zoo- geography of endemic spiders of Bulgaria (Ara- neae). Rev. Suisse. ZooL, hors serie 1:141-151. Deltshev, C. 1997a. A New Cybaeus from the mountains of Balkan Peninsula. Reichenbachia, 32:1-4. Deltshev, C. 1997b. Cryphoecina deelemanae gen. n., sp. n., a remarkable spider from the moun- tains of Montenegro (Yugoslaavia) (Arachnida, Araneae, Hahniidae). Rev. Suisse ZooL, 104: 485-489. Deltshev, C. & B.P.M. Curcic. 1997. Contribution to the knowledge of the group europaeus of Cen- tromerus Dahl (Linyphiidae, Araneae) in the Bal- kan Peninsula. Rev. Suisse ZooL, 104(l):49-55. Deltshev, C. & L. Paraschi. 1990. A contribution to the study of spiders (Araneae) in Greece, with a description of a new species (Malthonica spi- nipalpis Deltshev, sp. n., Agelenidae). Biol. Gal- lo-hellenica, 17:3-12. Drensky, P. 1936. Katalog der echten Spinnen (Ar- aneae) der Balkanhalbinsel. Shorn. Bulg. Akad. Nauk., 32:1-223. Hadjissarantos, H, 1940. Les araignees de FAttique. Athens. 132 pp. Jones, D.J., C. Ledoux & M. Emerit. 1990. Guide des araignees et Opiliones d’ Europe, Delachaux & Niestle. 384 pp. Koponen, S. 1993. On the biogeography and faun- istics of European spiders: latitude, altitude and insularity. Bull. Soc. Neuchatel. Sci. Nat., 116: 141-152. Maurer, R. & A. Hanggi. 1990. Katalog der Schweizerischen spinnen. Documenta Faunistica Helvetiae, 12. Mikhailov, K.G. 1997. Katalogue of the spiders of the territories of the former Soviet Union (Arach- nida, Aranei). Moscow: ZooL Mus. Moscow St. Univ. 416 pp. Nikolic, F. & A. Polenec. 1981. Catalogus faunae Jugoslaviae. Ljubljana. 135 pp. Platnick, N. 1989. Advances in Spider Taxonomy 1981-1987. A Supplement to Brignoli’s A Cat- alogue of the Araneae Described Between 1940 and 1981. Manchester Univ. Press, New York. 673 pp. Platnick, N. 1993. Advances in Spider Taxonomy 1988-1991. With Synonymies and Transfers 1940-1980. New York EntomoL Soc., American Mus. Nat. Hist. 846 pp. Platnick, N. 1997. Advances in Spider Taxonomy 1992-1995. With Redescriptions 1940. New York EntomoL Soc.; American Mus. Nat. Hist. 976 pp. Roberts, M.J. 1987. The spiders of Great Britain and Ireland 2, Colchester. 204 pp. Thaler, K. 1980. Die Spinnenfauna dwes Alpen: Ein Zoogeogrtafischer Versuch. Verb. Int. Arach- nol. Kong. 8 (Wien, 1980). Pp. 389-404. Thaler, K. 1996. Three Walckenaeria species from Peloponese, Greece (Araneae: Linyphiidae). Bull. British Arachnol. Soc., 10(4): 156-160. Thaler, K. & B. Knoflach. 1991. Eine neue Amau- robius-Art aus Griechenland (Arachnida: Ara- neae, Amaurobiidae). Mitt. Schweizerischen En- tomoL Ges., 64:265-268. Thaler, K. & B. Knoflach. 1993. Two new Amau- robius species (Araneae: Amaurobiidae) from Greece. Bull. British Arachnol. Soc., 9(4): 132- 136. Thaler, K. & B. Knoflach. 1995. Ueber Vorkom- men und Verbreitung von Amaurobius-Art&n in Peloponnes und Aegaeis (Araneida: Amaurobi- idae). Rev. Suisse ZooL, 102(l):41-60. Wunderlich, J. 1980. Linyphiidae aus Sued-Afrika (Arach.: Araneae). Verb. Naturwiss. Ver. Ham- burg, 23:319-337. Wunderlich, J. 1985. Pachygnatha clerckoides n. sp. aus Jugoslavien (Arachnida: Araneae: Tetrag- nathidae). Senckenbergiana Biol., 65:325-328. Wunderlich, J. 1994a. Zwei bisher unbekannte Mediterrane Arten der Gattung Pholcus Wal- ckenaer 1805 (Arachnida: Araneae: Pholcidae). Beitrage AraneoL, 4:625-628. Wunderlich, J. 1994b. Beschreibung bisher unbek- anter Arten der Baldachinspinnen aus der oestli- chen Mediterraneen (Arachnida: Araneae: Liny- phiidae). Beitrage AraneoL, 4:655-686. Wunderlich, J. 1994c. Beschreibung einer bisher unbekanten Art der Gattung Amaurobius C.L. Koch 1837 von Kreta (Arachnida: Araneae: Amaurobiidae). Beitrage AraneoL, 4:729-730. Manuscript received 24 April 1998, revised 30 May 1999. : .. „•%. Iv.., . Ji Sf . > '. M {^1- i'' .-'^■i^ll^|i#^ » ■' *rfl f'/r r ',.(■ - •■■1 >.*.'2* ■« '.,-> '. .'. i- *i ^,v* ,^*j1f/ .fv;^j:A- ■ . %.ny 4^ itf^ r * ' V. -> ai^TJI M^f , i'4«f .r:‘^vr;r- . • v«:-V .«• •■'''' IJ’lJ' ' ^ ^ .;y^ ^ ■?;♦ * '’^' ' i- . *-* I -‘ « * ■■ .i>'"^'^ii* * .rf c 3^ ‘ -''irf 'ViJ»^ 99% of visible light. The vertical force transducer (2) was calibrated against the hot-wire anemometer (1) in a low- velocity wind tunnel prior to the collection of field data. The array was attached to the top of a tripod and adjusted so that the sensor ar- ray was at the level of, and surrounded by but not touching, the tops of nearby inflorescences making up the canopy of Solidago sp. Prior to collecting data, the array was rotated so that an imaginary line connecting the two air ve- locity sensors would be perpendicular to the predominant wind direction and the light sen- sor was downwind from both. Analog signals from the sensor array were digitized by an analog-to-digital (A/D) con- verter (National Instruments Corporation model NB-MIO-16L) driven by a custom pro- gram (National Instruments software LabView 3) on a microcomputer (Apple Corporation model Power Macintosh 7100/80AV) located 5 m cross-wind from the sensor array. The software handled calibration, converted the signals into units of velocity (m/sec) and in- solation (arbitrary units of intensity), and SUTER— BALLOONING IN A CHAOTIC ATMOSPHERE 285 Figure 1 , — Titanium dioxide (Ti02) smoke streams elucidate the turbulent structure of moving air just above the goldenrod canopy. The four smoke streams (top left; unmodified photograph), originating at 10 cm intervals horizontally and at 15 cm intervals vertically, indicate that even small cross-wind displace- ments result in large differences in flow patterns, even to the extent that one smoke stream may rise while its nearest neighbor falls (top right; hand tracings of plumes added to enhance clarity). A negative image of one plume (bottom), photographed during a period of roughly horizontal air flow, indicates that even at quite small scales air movement appears to be turbulent. In each image, the length of the horizontal portion of the smoke tubing is 3 cm. stored the data to disk. Trios of data were re- corded at 0.1 Hz, resulting in accumulations of 360 trios of data per hour and 4320 trios per 12 hour day. Given the variability inherent in a turbulent atmosphere at all scales (Schlichting 1979), this rate of data collection must underestimate the total atmospheric var- iability (see results). RESULTS I collected data on three days during which I also observed ballooning (by unidentified spiders) from Solidago inflorescences at the study site. Photographic images (Fig. 1) of smoke plumes during this period indicated, at least qualitatively, that air movement just above the goldenrod canopy was remarkably variable at several scales. I report here quan- titative data and analyses based on only one of the sample days (15 September 1997) pri- marily because data from the three days were very similar. Raw data for the subject day show that the sun was visible throughout the day with the exception of brief intervals when small clouds obscured it, that wind velocity varied from 0-2.5 ml sec, and that the vertical component of air movement was generally up- ward prior to 1300 h and downward during 286 THE JOURNAL OF ARACHNOLOGY Figure 2. — (A) Raw data for 15 September 1997 depicting insolation (dotted line), total wind veloc- ity (black line), and the vertical component of air movement (gray line) at the top of a canopy of Sol- idago sp. (B) Mean wind speed (black line, 30- point running average) ± S.D. (gray lines) derived from the raw data. (C) Mean ± S.D. values for the vertical component of air movement. Brief down- ward displacements of the insolation line represent periods when the sun was partly or entirely ob- scured by small clouds. The change from upward to downward air motion at approximately 1300 h (A & C) provided the rationale for later analytical emphasis on the period prior to 1300 h. the remainder of the day (Fig. 2A). Thirty- point running averages ± S.D. for total wind velocity (Fig. 2B) show that average wind ve- locity was <1.0 m/sec throughout the day, and similarly processed data for the vertical component of air movement (Fig. 2C) empha- size the difference between pre-1300 h and post- 1300 h conditions. A sample-by-sample trigonometric combi- nation of total wind velocity with the vertical component of air motion yielded both the in- clinations (negative indicates an upward incli- nation, in keeping with the biological conven- tion in which geotaxis is positive when toward the earth) and the magnitudes of the resultant air motion vectors (Fig. 3A). Because bal- looning could occur only rarely after 1300 h, those data are ignored in the analyses that fol- low. The velocity vector magnitudes for pre- 1300 h are distributed approximately log-eor- mally (Fig. 3B) as -”0.93 ± —0.77 (logio m/ sec; equivalent of 0.117 ± 0.170 m/sec, mean velocity ± S.D.; = 0.93), although the ide- alized distribution overestimates very low ve- locities and underestimates velocities between 0.2 and 0.5 m/sec. The velocity vector incli- nations for pre-1300 h are distributed approx- imately as a circular normal distribution (Fig. 3C; Batschelet 1981) with an upward-directed mean of —19° ± 27° (mean ± angular devi- ation, mean vector loading, r = 0.885; = 0.87). A linear regression of inclination on magnitude for the pre-1300 h data had a sig- nificantly positive slope (r^ = 0.022, F = 19.58, P = 0.0001), indicating a slight ten- dency (--2% of the variance explained) for the inclination to become more downward as magnitude increased (negative values indicate upward inclination). The chaotic character of the turbulent at- mosphere at the canopy of a field means that, from moment to moment at a single location, air velocities and directions are apparently random (Panofsky & Dutton 1984; and note ratios of mean to S.D. for both velocities and inclinations, above). This apparent random- ness allows one to treat frequency distribu- tions as probability functions. Accordingly, a surface that shows the probabilistic structure of air motions at the sensor array, in terms of magnitude (Fig. 3B) and inclination (Fig. 3C), can be constructed as the product of the in- clination and magnitude frequency distribu- tions (Figs. 4 A and 4B). Note that, because of the relatively low sampling rate (0.1 Hz). used in the collection of the data from which these analyses were derived, and the consequent un- derestimation of the variability in atmospheric motion at small scales, the calculated proba- bilities (Figs. 4”6) are overestimates of the true probabilities experienced by ballooning spiders. An exploration of this unpredictable but statistically defined situation for spiderlings of known sizes is instructive. The terminal ve- SUTER— BALLOONING IN A CHAOTIC ATMOSPHERE 287 270.0° 0.05 0.35 0.65 0.95 1.25 1.55 1.85 Velocity (m/sec) C 0.10-1 c o o Q. O £L Angle (°) Figure 3. — (A) Trigonometric calculations based on air movement data (Fig. 2 A) resolved the data into its vector components (magnitude and inclination). Pre-1300 h data (filled circles) are used in subsequent analyses to the exclusion of the post- 1300 h data (crosses). (B) In the pre-1300 h data, air movement magnitudes (bars) were approximately log-normally distributed (open circles) with a mean of —0.93 log units (0.117 m/sec). (C) The inclinations of the pre-1300 h data (bars; negative angles are upward) formed an approximately circular normal distribution (open circles) with a mean direction of —19° and angular deviation of 27.1°. 288 THE JOURNAL OF ARACHNOLOGY 1.5 Angie (°) Velocity of Wind (m/sec) B X >» Velocity of Wind (m/sec) Angle C) Figure 4. — Probability distributions derived from the data in Fig. 3 for (A) the actual data distributions and (B) the idealized (log-normal and circular normal) distributions. locity of a 0.4 mg spider trailing 1 m of silk is about 0.24 m/sec (Equation 7 in Suter 1991) if both its body and its silk are in the same column of air. The vertical component of air velocity (speed- sin(inclination angle)) must then exceed 0.24 m/sec if the spider is to be- come, and remain, airborne. (Because the spi- der standing in the “tiptoe” posture at the top of a plant is in air moving more slowly than the air surrounding the silk, this estimate of minimum vertical magnitude is an underesti- mate for becoming airborne; Suter 1991.) The part of the angle- velocity plane shown in Fig. 4B that meets the criterion of having the ver- tical component of wind velocity > 0.24 m/ sec is quite limited (Fig. 5). The sum of all probabilities in the elevated part of the surface (Fig. 5, 0.4 mg), P, is 0.101. Thus the 0.4 mg spiderling, at any moment, has F < 0.1 that in the next 10 sec conditions will be momen- tarily suitable for becoming airborne, and the probability that those conditions will persist for 20 seconds is < 0.01. This analysis is sensitive both to the mass of the spider and to the length of silk in use as a “balloon.” A spider of twice the mass, SUTER— BALLOONING IN A CHAOTIC ATMOSPHERE 289 0.1 mg 0.4 mg 1 .6 mg Figure 5. — A spider can become airborne only if the upward component of the air velocity vector exceeds the terminal velocity of the spider with its silken “balloon.” For a 0.1 mg spider with aim length silk (top), the probability of becoming airborne in any 10- sec period is the volume under the elevated part of the surface, about 0.28. The probability falls steeply with increasing mass, so that a 1.6 mg spider has a probability near zero for the conditions that prevailed during the pre-1300 h period of 15 September 1997. using the same 1 m of silk, would have a ter- minal velocity of 0.43 m/sec, would encounter sufficient vertical air movement at P < 0.044, and could count on 20 s of those conditions at P2 < 0.002. In terms of probabilities, small- er spiders attempting to balloon have dispro- portionately greater access to aerial dispersal than do larger spiders (Fig. 6A). The release of additional silk, a behavioral tactic, also in- fluences P but the effectiveness of increases in silk length decline with length (Fig. 6B). DISCUSSION A spider attempting to balloon from the highest point on a plant is bathed in air the motion of which is demonstrably unpredict- able (Fig. 2A), chaotic (Panofsky & Dutton 1984), and best described probabilistically (Figs. 3-6). Thus the spider’s perception of the current state of its microclimate, at least with respect to air direction and speed, has almost no predictive value and cannot con- tribute, except in a statistical sense, to the spi- der’s decision-making. Behavioral consequences. — Faced with the stochastic atmospheric environment im- posed by turbulence in the air surrounding a ballooning site, what tactics can a would-be aeronaut adopt to maximize its probability of success? (1) Because favorable micro-scale 290 THE JOURNAL OF ARACHNOLOGY 0.3n 0.0' 0.0 0.5 1.0 Mass (mg) 1.5 0.5 0.4 0.3 0.2 0.1 0.0 — 1 I I I I 0 2 4 6 8 10 Silk Length (m) Figure 6. — The cumulative probability of becom- ing airborne (A) decreases strongly with increasing mass (at a constant silk length of 1 m) and (B) increases with the length of the balloon silk (for a spider of mass 0.4 gm). The mass relationship pro- vides a partial explanation for the data on the mas- ses of actual aeronauts (see references) and suggests a developmentally coupled decay in a spider's pro- clivity for ballooning. The silk-length relationship, in contrast, indicates that a spider can have some behavioral control over the probability of becoming and remaining aloft. conditions rarely persist for more than a few seconds, the spider should deploy silk rapidly and release from the substrate as soon as some vertical acceleration is assured (a testable pre- diction because p is an inverse function of mass [Fig. 6A], so that silk deployment rate should increase with mass and latency to re- lease should decrease with mass). Put another way, a 0.2 mg spider can more easily afford to wait for improved conditions than can a 0.4 mg spider, but both should be relatively re- sponsive to marginal conditions. (2) Because B ..Q cc n o £ ffi .> 3 E 3 o post-release posture strongly influences drag (Suter 1992), the spider should assume a spread-eagle posture (all unoccupied legs ex- tended) immediately upon release, a tactic more important for larger spiders because of their increased terminal velocities. This be- havior would also be beneficial if air flow were laminar (which it never is at the canopy), but its importance under turbulence is in- creased because favorable conditions are so transitory. Finally, (3) spiders climbing to- ward a site from which to balloon should be selective if their sensory capabilities permit: the rarity of conditions (i.e., high magnitude and inclination) that could rapidly extract a ballooning spider from the vicinity of obsta- cles favors selection of the highest local prom- ontory, and again, selectivity should be most evident in larger spiders. Consequences for larger spiders.—Fluid dynamic calculations (Humphrey 1987; Suter 1991) make clear that ballooning is a dispersal mechanism that is primarily available to small spiders. The bias against large ballooners is strengthened by other considerations as well: climbing is more energetically expensive for large than for small spiders (Thompson 1942; Price 1984), and large spiders, given their higher terminal velocities, have to climb more frequently than small ones to achieve the same horizontal displacement; larger spiders are more conspicuous to predators; and favorable conditions under turbulent conditions vanish rapidly as spider size rises (Fig. 6A). It is not surprising, therefore, that this pronounced bias is mirrored in data on the sizes of aeronauts (Dean & Sterling 1985; Greenstone et al. 1987; Bishop 1990). Complications.— In this paper I have ana- lyzed micrometeorological data from a single day and, had I chosen another day, the data would surely have been different in detail. But ballooning did occur during the morning of the subject day, the day was similar to those described by others as prime for ballooning (Vugts & Van Wingerden 1976), and the char- acteristics of turbulent flow over a single hab- itat within a specific range of velocities (e.g., 0-=3 m/sec) are remarkably consistent (Panof- sky & Dutton 1984; Anderson et al. 1986). More problematic is the consideration of only a single site within the field -other sites, being closer to a tree line or hedgerow or more re- mote from a patch of field dominated by low SUTER— BALLOONING IN A CHAOTIC ATMOSPHERE 291 grasses, could have quite different atmospher- ic characteristics. A study of site-specific aerodynamics, and particularly the identifica- tion of features (such as down-wind barriers) that cause consistent and stable updrafts, is certainly warranted. Much of the analysis used in this study has been statistical and probabilistic rather than strictly mathematical. This is a necessary con- sequence of the structure of a turbulent at- mosphere and its chaotic (and apparently ran- dom) behavior as it moves past a fixed point from which a spider might attempt to balloon. From the perspective of an airborne spider, the situation is equally complex but very different because the spider and its silk are now incom- pletely entrained in air that is characterized by eddies of many sizes (e.g., Gao et al. 1989) that are relatively coherent as they move downwind (Zhang et al. 1992). A full under- standing of the motion of airborne spiders in the turbulent air above agroecosystems is still a long way off. Finally, during a particular ballooning at- tempt, the risk of failure (going only a short distance horizontally) is high (Fig. 3), but that kind of risk is relatively cost-free for very small spiders: returning to an exposed tip of a plant, even repeatedly, is energetically cheap (Thompson 1942), whereas the risk of preda- tion during a cursorial or drop-and-swing (Barth et al. 1991) dispersal of the same hor- izontal distance must be considerably greater. ACKNOWLEDGMENTS I thank Matthew Greenstone for stimulat- ing, once again, my interest in the physics of ballooning, Gary Lovett for educating me both in the dynamics of turbulent atmospheres and in the physics of deposition of small par- ticles, and Christopher Smart for helping to construct a device that allowed me to visualize turbulence at the study site. The study was supported in part by funds provided by Vassar College through the Class of ’42 Faculty Re- search Fund. LITERATURE CITED Anderson, D.E., S.B. Verma, R.J. Clement, D.D. Baldocchi & D.R. Matt. 1986. Turbulence spec- tra of CO2, water vapor, temperature and velocity over a deciduous forest. Agric, Forest Meteor., 38:81-99. Andow, D.A. 1991. Yield loss to arthropods in ve- getationally diverse agroecosystems. Environ. EntomoL, 20:1228-1235. Batschelet, E. 1981. Circular Statistics in Biology. Academic Press, London. 371 pp. Barth, EG., S. Komarek, J.A.C. Humphrey & B. Treidler. 1991. Drop and swing dispersal behav- ior of a tropical wandering spider: experiments and a numerical model. J. Comp. Phys., 169: 313-322. Bishop, L. 1990. Meteorological aspects of spider ballooning. Environ. EntomoL, 19:1381-1387. Bishop, L. & S.E. Riechert. 1990. Spider coloni- zation of agroecosystems: mode and source. En- viron. EntomoL, 19:1738-1745. Bristowe, W.S. 1939. The Comity of Spiders, Vol. 1. Johnson Reprint Corp., New York. Bruce, R. & D. Howard (eds.). 1990. Species dis- persal in agricultural habitats. Belhaven Press, London. 288 pp. Carter, RE. & A.L. Rypstra. 1995. Top-down ef- fects in soybean agroecosystems: spider density affects herbivore damage. Oikos, 72:433-439. Cox, D.L. & D.A. Potter. 1986. Aerial dispersal behavior of larval bagworms, Thyridopteryx ephemeraeformis (Lepidoptera: Psychidae). Ca- nadian EntomoL, 118:525-536. Coyle, EA., M.H. Greenstone, A-L. Hultsch & C.E. Morgan. 1985. Ballooning mygalomorphs: es- timates of the masses of Sphodros and Ummidia ballooners (Araneae: Atypidae, Ctenizidae). J. ArachnoL, 13:291-296. Darwin, C. 1839. Journal of Researches into the Geology and Natural History of the Various Countries Visited by H.M.S. Beagle under the Command of Captain Fitzroy, R.N. from 1832 to 1836. Henry Colburn, London. Dean, D.A. & W.L. Sterling. 1985. Size and phe- nology of ballooning spiders at two locations in eastern Texas. J. ArachnoL, 13:111-120. Duelli, R, M. Studer, I. Marchand & S. Jacob. 1990. Population movements of arthropods be- tween natural and cultivated habitats. Biol. Con- serv., 54:193-207. Duffey, E. 1956. Aerial dispersal in a known spider population. J. Anim. EcoL, 25:85-111. Eberhard, W.G. 1987. How spiders initiate airborne lines. J. ArachnoL, 15:1-9. Edwards, J.S. 1988. Life in the allobiosphere. Trends EcoL EvoL, 3:111-114. Emerton, J.H. 1908. Autumn flights of spiders. Psyche, 15:121. Gao, W., R.H.Shaw & K.T Paw U. 1989. Obser- vation of organized structure in turbulent flow within and above a forest canopy. Boundary- Layer MeteoroL, 44:349-377. Geiger, R. 1965. The Climate Near the Ground. Harvard Univ. Press, Cambridge. 611 pp. Gertsch, W.J. 1979. American Spiders (2nd Ed.). Van Nostrand Company, New York, 274 pp. 292 THE JOURNAL OF ARACHNOLOGY Click, RA. 1939. The distribution of insects, spi- ders, and mites in the air. USDA Tech. Bull., 673:1-15. Greenstone, M.H. 1982. Ballooning frequency and habitat predictability in two wolf spider species (Lycosidae: Pardosa). Florida EntomoL, 65:83- 89. Greenstone, M.H. 1990. Meteorological determi- nants of spider ballooning: the roles of thermals vs. the vertical windspeed gradient in becoming airborne. Oecologia, 84:164-168. Greenstone, M.H., R.L. Eaton & C.E. Morgan. 1991. Sampling aerially dispersing arthropods: a high- volume, inexpensive, automobile- and air- craft-bome system. J. Econ. EntomoL, 84:1717- 1724. Greenstone, M.H., C.E. Morgan & A.-L. Hultsh. 1987. Ballooning spiders in Missouri, USA, and New South Wales, Australia: family and mass distributions. J. ArachnoL, 15:163-170. Halley, J.M., C.F.H. Thomas & P.C. Jepson. 1996. A model for the spatial dynamics of linyphiid spiders in farmland. J. Appl. EcoL, 33:471-492. Hardy, A.C. & L. Cheng. 1986. Studies in the dis- tribution of insects by aerial currents. III. Insect drift over the sea. Ecol. EntomoL, 11:283-290. Henschel, J.R., J. Schneider & Y.D. Lubin. 1995. Dispersal mechanisms of Stegodyphm (Erisidae): Do they balloon? J. ArachnoL, 23:202-204. Hoelscher, C.E. 1967, Wind dispersal of brown soft scale crawlers, Coccus hesperidium (Ho- moptera: Coccidae), and Texas citrus mites, Eu- tetranychus banksi (Acarina: Tetranychidae) from Texas citrus, Ann. EntomoL Soc. America, 60:673-678. Humphrey, J.A.C. 1987. Fluid mechanical con- straints on spider ballooning. Oecologia, 73:469- 477. Johnson, D.T & B.A. Croft. 1976. Laboratory study of the dispersal behavior of Amblyseius faP lads (Acarina: Phytoseiidae). Ann. EntomoL Soc. America, 69:1019-1023. Kemp, J.C. & G.W. Barrett. 1989. Spatial pattern- ing: impact of uncultivated corridors on arthro- pod populations within soybean agroecosystems. Ecology, 70:114-128. Mansour, E, D.B. Richman & W.H. Whitcomb. 1983, Spider management in agroecosystems: habitat manipulation. Environ. Manag., 7:43-49. Mauremooto, J.R., S.D, Wratten, S.P. Womer & G.L.A. Fry, 1995. Permeability of hedgerows to predatory beetles. Agric. Ecosys. Environ., 52: 141-148. McCook, H.C. 1877. The aeronautic flight of spi- ders, Proc. Philadelphia Acad. Nat. Sci., 29:308- 312. McCook, H.C. 1878. Note on the probable geo- graphic distribution of a spider by the trade winds. Proc. Philadelphia Acad. Nat. Sci,, 30: 136-147. McManus, M.L. & C.J. Mason. 1983. Determina- tion of the settling velocity and its significance to larval dispersal of the gypsy moth (Lepidop- tera: Lymantridae). Environ. EntomoL, 12:270- 272. Meijer, J. 1977. The immigration of spiders (Ara- neida) into a new polder. Ecol. EntomoL, 2:81- 90. Morse, D.H. 1993. Some determinants of dispersal by crab spiderlings. Ecology, 74:427-432. Panofsky, H.A. & J.A. Dutton. 1984. Atmospheric Turbulence; Models and Methods for Engineer- ing Applications. John Wiley & Sons, New York. 397 pp. Platnick, N.L 1976. Drifting spiders or continents?: Vicariance biogeography of the spider subfamily Laroniinae (Araneae: Gnaphosidae). Syst. ZooL, 25:101-109. Price, P.W. 1984. Insect Ecology. 2nd Ed. Wiley- Interscience, New York. 607 pp. Richter, C.J.J. 1970. Aerial dispersal in relation to habitat in eight wolf spider species (Pardosa, Ar- aneae, Lycosidae). Oecologia, 5:200-214. Richter, C.J.J. 1971. Some aspects of aerial dis- persal in different populations of wolf spiders, with particular reference to Pardosa amentata (Araneae, Lycosidae). Misc. Pap. Landbouwho- gesch. Wageningen, 8:77-88. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in garden test systems. Ecology, 17:1441-1450. Riechert, S.E. & T Lockley. 1984. Spiders as bi- ological control agents. Annu. Rev. EntomoL, 29:299-320. Rodenhouse, N.L., G.W. Barrett, D.M. Zimmerman & J.C. Kemp. 1992. Effects of uncultivated cor- ridors on arthropod abundances and crop yields in soybean agroecosystems. Agric. Ecosys. & Environ., 38:179-191. Rypstra, A.L., RE. Carter, R.A. Balfour & S.D. Marshall. 1999. Architectural modifications of agricultural habitats and their impact on the spi- der inhabitants. J. ArachnoL, 27(l):371-377. Salmon, J.T. & N.V. Homer. 1977. Aerial disper- sion of spiders in North Central Texas. J. Arach- noL, 5:153-157. Schlichting, H.T. 1979. Boundary Layer Theory. 7th Ed. McGraw-Hill, New York. Smitley, D.R. & G.G. Kennedy. 1985. Photo-ori- ented aerial-dispersal behavior of Tetranychus urticae (Acari: Tetranychidae) enhances escape from the leaf surface. Ann. EntomoL Soc. Amer- ica, 78:609-614. Southwood, T. 1962. Migration in terrestrial ar- thropods in relation to habitat. Biol. Rev., 37: 171-214. Stamps, J.A., M. Buechner & V.V. Krishnae. 1987. SUTER— BALLOONING IN A CHAOTIC ATMOSPHERE 293 The effects of edge permeability and habitat ge- ometry on emigration from patches of habitat. American Nat., 129:533-552. Suter, R.B. 1991. Ballooning in spiders: results of wind tunnel experiments. Ethol. EcoL & EvoL, 3:13-25. Suter, R.B. 1992. Ballooning: data from spiders in freefall indicate the importance of posture. J. Ar- achnoL, 20:107-113. Thomas, C.EG. 1996. Modelling aerial dispersal of linyphiid spiders. Aspects Appl. Biol., 46:217- 222. Thomas, C.F.G., P. Brain & PC. Jepson. In press. Measuring and modelling the aerial activity and dispersal distances of ballooning spiders. J. Anim. Ecol. Thomas, C.EG., E.H.A. Hoi & J.W. Everts. 1990. Modelling the diffusion component of dispersal during recovery of a population of linyphiid spi- ders from exposure to an insecticide. Funct. Ecol., 4:357-368. Thompson, D’A.W. 1942. On Growth and Form. 2nd Ed. Cambridge Univ. Press, Cambridge. 1116 pp. Thornton, I.W.B., T.R. New, D.A. McLaren, H.K. Sundarman & P.J. Vaughan. 1988. Air-bome ar- thropod fall-out on Anak-Krakatau and a possi- ble pre-vegetation pioneer community. Phil. Trans. R. Soc. London, Biol. Sci., 322:471-479. Tolbert, W.W. 1977. Aerial dispersal behavior of two orb weaving spiders. Psyche, 84:13-27. van Wingerden, W.K.R.E. 1980. Aeronautic dis- persal of immatures of two linyphiid spider spe- cies (Araneae, Linyphiidae). Int. Arachnol. Kon- gre, Wein, 8:91-95. van Wingerden, W.K.R.E. & Vugts, H.E 1974. Factors influencing aeronautic behavior of spi- ders. Bull. British Arachnol. Soc., 3:6-10. Vugts, H.E & W.K.R.E. van Wingerden. 1976. Me- teorological aspects of aeronautic behaviour of spiders. Oikos, 27:433-444. Washburn, WO. & L. Washburn. 1984. Active ae- rial dispersal of minute wingless arthropods; ex- ploitation of boundary-layer velocity gradients. Science, 223:1088-1089. Weyman, G.S. 1993. A review of the possible causative factors and significance of ballooning in spiders. Ethol. Ecol. & EvoL, 5:279-291. Weyman, G.S. & PC. Jepson. 1994. The effect of food supply on the colonisation of barley by ae- rially dispersing spiders (Araneae). Oecologia, 100:386-390. Weyman, G.S., PC. Jepson & K.D. Sunderland. 1995. Do seasonal changes in numbers of aeri- ally dispersing spiders reflect population density on the ground or variation in ballooning moti- vation? Oecologia, 101:487-493. Wickler, W. & U. Seibt. 1986. Aerial dispersal by ballooning in adult Stegodyphus mimosarum. Na- turwilL, 73:628-629. Wissinger, S. 1997. Cyclic colonization in predict- ably ephemeral habitats: a template for biological control in annual crop systems. Biol. Control, 10: 4-15. Yeargen, K.V 1975. Factors influencing the aerial dispersal of spiders (Arachnida: Araneida). J. Kansas Entomol. Soc., 48:403-408. Zhang, C., R.H. Shaw & K.T. Paw U. 1992. Spatial characteristics of turbulent coherent structures with and above an orchard canopy. Pp. 741-752. In Precipitation Scavenging and Atmosphere- Surface Exchange. (S.E. Schwartz & WG.N. Slinn, eds.). Hemisphere Publishing Corp., Washington, D.C. Manuscript received 24 March 1998, revised 24 September 1998. 1999. The Journal of Arachnology 27:294-300 DIFFERENTIAL AERIAL DISPERSAL OF LINYPHIID SPIDERS FROM A GRASS AND A CEREAL FIELD C.F.G. Thomas^ and P.C. Jepson^: Department of Biology, University of Southampton, Bassett Crescent East, Southampton S09 3TU, U.K. ABSTRACT. Ground and aerial populations of linyphiid spiders were sampled in and above a grass and a cereal field, weekly from June- August 1991. Aerial activity of inunature, adult male and adult female spiders was significantly higher over the senescing cereal field than the grass field. Water-trap catches and wind-speed data were used to calculate indices of aerial activity to show differences in the timing of dispersal by adult male, female and immature spiders. Some aerial dispersal occurred every week with highest adult dispersal in July and highest immature dispersal in August. Aerial activity indices were higher for males than females, and the dispersal peak occurred earlier for males than females. Immatures dispersed from the cereal field in July and August, and from the grass field mainly in August. Differences in aerial activity are discussed with reference to dispersal strategies that might maximize spider survival in the patchy, disturbed agricultural landscape. Several species of linyphiid spider are widespread and abundant natural predators of pests in the agricultural ecosystem (Sunder- land et al. 1986). In farmland, the patchwork of annual and perennial crops fragments spi- der populations into patches of habitat among which resource quality and risks of habitat disturbance vary in space and time. In such spatially structured populations or metapopu- lations, dispersal behavior is critical for re- founding locally extinct populations (Gilpin & Hanski 1991) and has a major effect on pop- ulation size and persistence at the landscape scale (Halley et al. 1996). Little is known of the ancestral habitats of species of linyphiid spider which are now abundant in farmland, or the pattern of dis- turbance they experienced. Both immature and adult linyphiid spiders are, however, able to disperse over large areas by “ballooning” (Thomas 1996), and this ability is a necessary pre- adaptation for survival in disturbed farm- land habitats (Halley et al. 1996). The param- eters of dispersal — distance, frequency, timing and the proportion of a population that under- goes dispersal — are all likely to be under ' Current address: lACR -Long Ashton Research Station, Dept, of Agricultural Sciences, University of Bristol, Long Ashton, Bristol, BS41 9AF, U.K. ^ Current address: Department of Entomology, Cordley Hall, Oregon State University, Corvallis, Oregon 97331 USA adaptive pressure to maximize survival in agroecosystems by spreading risks or optimiz- ing foraging among the shifting mosaic of available habitats. Throughout history, agri- cultural fields have been periodically dis- turbed by harvesting and cultivations. The fre- quency, severity and predictability of such catastrophic events depends on patterns of land use and methods of production. However, in recent decades, insecticides have become an additional hazard to linyphiid spiders. On an evolutionary time- scale, these changes may have been so sudden and widespread that lin- yphiid spiders are unable to adapt; and there is some evidence that the abundance of liny- phiid spiders has been declining in UK arable farmland since the 1970s (Aebischer 1990). In order to understand linyphiid spider re- sponses to changing agricultural practices and patterns of land use, a simulation model of spatially dynamic spider populations has been developed (Halley et al. 1996). All the dis- persal parameters mentioned above can be varied in the model. However, in that model, the dispersal process is simplified such that the proportion of dispersers in the population is the same across all habitats. Other param- eters apply equally to both sexes and all age classes. This decision was based on the as- sumption that meteorological factors were the most important constraint on dispersal, affect- ing all spiders uniformly. The critical con- straint is wind-speed which must be below 3 294 THOMAS & JEPSON— SPIDER DISPERSAL FROM GRASS & CEREALS 295 ms“‘ for ballooning to occur (Vugts & van Wingerden 1976). However, the model has shown that small differences in the duration and timing of dispersal can have large impacts on population size and persistence (Halley et al. 1996; Thomas 1997). Wind tunnel studies have shown that ap- proximately 40% of linyphiid spiders attempt to balloon when wind conditions are suitable (Legel & van Wingerden 1980). Computer simulations have also shown 40% to be the optimum dispersal rate that maximizes popu- lation size in a wide range of agricultural landscapes (Halley et al. 1996). Age or sex specific differences in the proclivity to dis- perse from different crop types under the same meteorological conditions may, therefore, re- sult in different risks to different sections of the population; and these aspects require fur- ther investigation. This paper describes field observations on the aerial activity of male, female and imma- ture linyphiid spiders dispersing from known populations in a cereal and a grass field. The results are discussed in terms of possible dif- ferences in the relative importance of dispersal in foraging and risk-spreading strategies thought to be relevant to each group. METHODS In June, July and August 1991, during the most active phase of summer reproduction and dispersal, linyphiid spiders were sampled at weekly intervals from a winter wheat field and a grass field. Eleven samples, totaling 5.28 m^, were taken from near the center of each field with a suction sampler (D-vac), each consisting of five sub-samples of a 10- 15 second application of the suction head (0.096 m^) to the soil surface enclosing any foliage present in a net 1.5 m long. Over the same period, aerial activity was measured with two water traps in each field. Traps were constructed from square, galva- nized steel trays 1.2 m on each side and sup- ported on a metal frame one meter above the soil. Eight plastic seed trays (46 cm X 27 cm, Stewart Plastics, UK) were placed within the steel tray. The inner trays were filled with 50% ethylene glycol in water containing 2.5% detergent and had a total surface capture area of 1 m^. The outer steel tray was also filled with water and detergent to prevent access to the inner trays by spiders climbing in from the crop. The inner trays thus trapped only bal- looning spiders landing from the air, giving a measure of aerial dispersal activity. Parallel studies on the change of spider aerial density with height (Thomas 1992) indicated that the water trap catches were dominated by spiders from low altitudes attempting to disperse from the field in which the trap was situated. Spi- ders descending from higher altitudes, having dispersed from more distant fields, are likely to form only a small proportion of the trap catch. Water traps were emptied at weekly in- tervals by sieving the contents through a fine nylon mesh of the same material as the D-vac net. Samples were sorted in the laboratory un- der a dissecting microscope, and adult liny- phiid spiders were sexed and identified to spe- cies. Because immature spiders could not be identified to species, data are presented for all species combined. Because aerial dispersal does not occur when wind-speeds are greater than 3 ms“^ (Vugts & van Wingerden 1976) an anemom- eter (Lambrecht, Germany) was used to record wind- speed on paper chart to quantify the amount of time suitable for dispersal during each sampling period. This was defined as the total number of hours between 0600-1800 h GMT with wind-speeds below the 3 ms“* bal- looning threshold. In each trapping period, the total water trap catch was converted to an ae- rial activity index, expressed as the total num- ber of spiders trapped in each field, per hour of available ballooning time. RESULTS Linyphiid spiders comprised more than 95% of the sampled population. Other fami- lies were not considered in this study. The samples were dominated by five taxa: Erigone atra (Blackwall), E. dentipalpis (Wider), Mei- oneta rurestris (C.L.Koch), Lepthyphantes tenuis (Blackwall) and Oedothorax spp. A sixth taxon “other species” comprised a few individuals of a number of species. An earlier study (Thomas & Jepson 1997) showed no significant difference between the species composition of water trap samples and D-vac samples (Fig. 1). Data are therefore presented as total linyphiids, divided into the categories “immature,” “adult male” and “adult fe- male”. Table 1 shows the total numbers of imma- ture, adult male and adult female linyphiid 296 THE JOURNAL OF ARACHNOLOGY Figure 1 . — Percentage population composition of dominant taxa from suction samples (total ground population) and water trap samples (total aerial population) based on data from a previous study (Thomas & Jepson 1997) at the same field site. Ab- breviations: e.atr = Erigone atra\ e.den = E. den- tipaipis; m.rur = Meioneta rurestris; l.ten = Lep- thyphantes tenuis; o.spp = Oedothorax spp.; others = other linyphiid species. spiders, captured in the water traps and sam- pled on the ground in the two fields. Signifi- cant differences between captures in the grass and cereal field on each date, tested by and the total number of available hours for bal- looning are also given. Aerial activity indices (number of trapped spiders divided by number of hours available for ballooning during trap- ping period) for immature, adult male and adult female spiders, in the grass and cereal fields, are given in Fig. 2. During June and early July aerial dispersal was too low on some dates to test for signif- icant differences between numbers of spiders captured in the grass and cereal field, either on the ground or from the air (Table 1). In mid- July there were significantly more im- mature spiders captured over the cereal crop when the population density on the ground was significantly higher in the grass field than the cereal field. In August, there were signif- icantly more immature spiders from the grass field compared to the cereal field, in both ae- rial and ground samples. A 2 X 2 contingency test on total raw counts of air and ground sam- ples from the grass and cereal field, over the entire experimental period, was highly signif- icant by G-test using Williams’ correction (Sokal & Rohlf 1981): = 871.5; df = 1; P < 0.001, indicating a significantly higher proportion of captures of airborne immature spiders over the cereal crop than expected from the respective ground population densi- ties in the two fields. During late July and August, there were significantly higher numbers of adult males and females caught in the water traps in the cereal field, compared with the grass field. Be- tween early June and mid-July, there were sig- nificantly higher ground population densities of adult males and females in the cereal field compared to the grass field, or no significant difference between the two populations. After the end of July and during August, following a week with the highest number of ballooning hours (week ending 3 August: 69 hours of wind-speed below 3ms~') the pattern re- versed; and there were significantly higher ground population densities of adult males and females in the grass field compared with the cereal field, suggesting a net emigration from the cereal field and a net immigration into the grass field. A 2 X 2 contingency test on total raw counts of air and ground samples from the grass and cereal field over the entire experimental period was highly significant by G-test for both male (G^jj = 37.0; df = 1; P < 0.001) and female spiders (G^dj = 32.4; df = 1; P < 0.001). As for the immature spiders, these results indicate that higher proportions of adult male and female airborne spiders were taken over the cereal crop than expected from the relative ground population densities in the grass and cereal field. Figure 2 shows the water trap catches in each week divided by the number of hours of available ballooning time. In both the grass and cereal fields, there was little ballooning activity up to July 3. Thereafter, the majority of dispersal by male spiders occurred over a period of three weeks, with a peak occurring during the week ending July 18. On each date. THOMAS & JEPSON— SPIDER DISPERSAL FROM GRASS & CEREALS 297 Table L — Aerial and ground captures of linyphiid spiders in a grass and cereal field. Significant dif- ferences tested by chi square. P < 0.05 = *; P < 0.01 = **; p < 0.001 = ***; ns = not significant. Dates with expected captures <5 not tested. Sample period week ending (ballooning hours) Immatures Adult males Adult females Grass Cereal Grass Cereal Grass Cereal Aerial activity (spiders/trapping period) 4 June (8) 22 9* 4 2 — 1 2 — 13 June (20) 8 20* 1 1 — 1 1 — 18 June (7) 1 3 — 1 0 — 0 0 — 25 June (6) 3 2 — 2 1 — 0 0 — 3 July (32) 3 5 — 12 2** 8 1 — 9 July (7) 2 47 54ns 16 16ns 18 July (10) 12 45*** 112 259*** 20 92*** 25 July (14) 49 64ns 79 no* 65 252*** 3 August (69) 506 292*** 72 2 24*** no 53** 10 August (19) 585 270*** 20 38* 38 39ns 17 August (24) 809 521*** 41 38ns 58 44ns Ground population density (spiders/5.28 sq m) 4 June 82 27*** 1 8 — 0 4 — 11 June 59 36* 0 22*** 2 24** 18 June 105 104ns 11 15ns 3 23*** 26 June 148 158ns 32 39ns 26 27ns 4 July 136 88** 95 90ns 76 58ns 9 July 60 54ns 132 109ns 77 69ns 16 July 153 g4*** 72 85ns 61 80ns 23 July 759 128*** 59 122*** 48 260*** 30 July 2316 9g*** 177 43*** 154 52*** 6 August 3454 667*** 115 65*** 96 87ns 13 August 3335 2g9*** 117 55*** 172 103*** higher numbers were trapped over the cereal field than the grass field. Female spiders also showed increased dispersal during this period, although this was generally lower than the males and the dispersal peak occurred one week later. The aerial activity indices of fe- males were generally much higher over the cereal field than the grass field. Immature spi- ders had higher aerial activiy indices over the grass field in August, reflecting the higher population densities on the ground in that field. However, immatures began to disperse from the cereal field earlier than from the grass field. DISCUSSION The D-vac and water traps are not 100% efficient sampling devices. However, they are the most reliable and cost-efficient methods available. The water traps are of comparable efficiency, regardless of where they are sited. The efficiency of the D-vac, however, is de- pendent on the vegetation structure and den- sity (Duffey 1980). In this study, the popula- tions sampled from the dense grass sward are likely to have been underestimated in com- parison with the populations in the much less dense stand of cereal. Thus, where the densi- ties of linyphiid spiders in the grass field are shown to be higher than the cereal field (Table 1), the true differences are likely to have be even greater. Although water traps can be left unattended in the field where they catch dispersing spi- ders effectively, the number of spiders they trap is not directly related to the number of spiders ballooning. A ballooning spider takes off and lands a number of times during a dis- persal episode, dependent on the amount of atmospheric turbulence on a given day (Thomas 1992). Thus, water trap catches are a function of the size of the source population on the ground, the proportion of the popula- tion that engages in dispersal, the number of 298 THE JOURNAL OF ARACHNOLOGY Date 4 June 18 June 3 July 18 July 3 August 17 August Date Figure 2. — Aerial activity indices expressed as the total number of spiders in water traps per hour of available ballooning time (wind-speed < 3ms“') for immature (filled bars), adult male (hatched bars) and adult female (open bars) spiders in a grass field (top), and a cereal field (bottom). hours of available ballooning time, and the take-off and landing rates on different days during a trapping period, which might com- prise several ballooning episodes. Quantitative estimates of differential dispersal rates from different habitats, therefore, can only be ob- tained by simultaneously measuring the num- bers of spiders initiating dispersal behavior from given areas of ground in different crops and relating these to the respective ground population densities over the same area. In spite of the limitations of the sampling methods employed, there is good evidence that dispersal from senescing cereal fields by linyphiid spiders is significantly higher than from grass fields. Similar results have been demonstrated by Weyman et al. (1995). There are also good theoretical reasons why differ- ential dispersal should be expected. Cereal and grass fields provide different quality hab- itat, especially in the late summer when pe- rennial grass, if left uncut, provides a cool, humid microclimate, while senescing annual cereals provide a hot, dry microclimate (Geig- er et al. 1995). These micro-climatic differ- ences, and the different resources provided by a lush perennial and a senescing annual crop, are also reflected in the fauna sampled with the spiders from the two habitats. In the grass field, high prey density (mostly Collembola, aphids and Diptera) was found while in the cereal field, low prey density was found (pers. obs.). Further evidence of differences in hab- itat quality between the grass and the senesc- ing cereal field comes from the observation of significantly higher production of immature linyphiid spiders in the grass field compared with the cereal field (Table 1). These factors are likely to affect the level of satiation of spiders, which is known to affect their pro- pensity to disperse (Legel & van Wingerden 1980; Weyman et al. 1994). Linyphiid spiders have also been shown to be retained in ex- perimental plots with high prey density (Wey- man & Jepson 1995), presumably by reduced emigration. A number of dispersal strategies may op- erate in this system and be adopted to differ- ent degrees by different sections of the pop- ulation. When suitable meteorological conditions prevail, opportunistic dispersal be- tween habitat patches of differing quality in the farmland mosaic might form the basis of a foraging strategy. When annual crops ripen and senesce, causing gradual deterioration of habitat quality, a resource-assessment strategy (Parker & Stuart 1976) might operate, e.g., spiders may respond to a marginal value (Chamov 1976) of food availability or thresh- old of environmental stress and initiate dis- persal only when this is reached. Some as- pects of these behavioral strategies have been reviewed by Janetos (1986) and Riechert & Gillespie (1986). In agroecosytems where habitat patches have a probability of undergoing unpredict- able catastrophic disruptions, e.g., harvesting or insecticide applications, a risk-spreading strategy (den Boer 1968) might operate. The risk to a sedentary spider of succumbing to unpredictable catastrophic events in the habi- tat patch in which it is resident needs to be balanced against the probabilities of dispers- ing to a more favorable habitat patch, dis- THOMAS & JEPSON— SPIDER DISPERSAL FROM GRASS & CEREALS 299 persing to a less favorable habitat patch, and the risk of dying during dispersal. Male and female spiders might also respond differently to the same resources and risks. Adult males feed little when sexually active (Alderweireldt & Lissens 1988) and are there- fore partially released from constraints of re- source availability. High male dispersal may simply reflect the relative importance of mate- finding over feeding. Females, on the other hand, require more resources for egg produc- tion. They may, therefore, be less likely than males to disperse from a high quality habitat where resources and microclimate may also increase the survival of eggs and early im- mature stages. Some female dispersal from high quality grass fields may still be expected as a strategy to increase overall survival prob- ability of offspring by laying egg sacs in sev- eral patches, thus spreading risks among the patchwork of field types in the landscape. The dispersal strategies of linyphiid spiders in agroecosystems are likely to be complex. More detailed field studies and computer sim- ulations are required in order to resolve the relative importance of different dispersal strat- egies, predict their effect on the metapopula- tion dynamics of the group, and determine whether dispersal strategies can adapt fast enough for populations to persist in the face of increased risks associated with insecticide use and other management practices in mod- em agroecosystems. ACKNOWLEDGMENTS This work formed part of a project ‘Spatial Dynamics of Spiders in Farmland’ funded un- der the Joint Agriculture and Environment Programme administered by NERC (Grant GST/02/478). Thanks are due to B. Gibbons and staff of Leckford Estates for permission to conduct the fieldwork and their co-opera- tion throughout. LITERATURE CITED Aebischer, N.J. 1990. Twenty years of monitoring invertebrates and weeds in cereal fields in Sus- sex. Pp. 305-331. In The Ecology of Temperate Cereal Fields. (L.G. Firbank, N. Carter, J.E Dar- byshire, & G.R. Potts, eds). 32nd Symposium of the British Ecol. Soc., Blackwell Sci. Publ., Ox- ford. Alderweireldt, M. & A. Lissens. 1988. Laborato- riumwaarnemingen van de ontwikkeling en re- productie bij Oedothorax apicatus (Blackwall, 1850) en Oedothorax retusus (Westring, 1851). Nieuwsbr, Belg. ArachnoL Ver., 9:19-26. Chamov, E.L. 1976. Optimal foraging, the margin- al value theorem. Theoret. Pop. Biol., 9:129- 136. den Boer, PJ. 1968. Spreading of risk and stabili- zation of animal numbers. Acta Biotheoretica, 18:165-194. Duffey, E. 1980. The efficiency of the Dietrick vacuum sampler (D-vac) for invertebrate popu- lation studies in different types of grassland. Bull. Ecol., 11:421-431. Geiger, R., R.H. Aron & P. Todhunter. 1995. The Climate Near the Ground. Friedr. Vieweg & Sohn Verlagsgesellschaft mbH, Braunschweig/ Wiesbaden. Gilpin, M. & I. Hanski. 1991. Metapopulation dy- namics: empirical and theoretical investigations. Biol. J. Linn. Soc., 42(1) & (2). Academic Press, London. Halley, J.M., C.FG. Thomas & PC. Jepson. 1996. A model for the spatial dynamics of linyphiid spiders in farmland. J. Appl. Ecol., 33:471-492. Janetos, A.C. 1986. Web-site selection: Are we asking the right questions? Pp. 9-22. In Spiders: Webs, Behavior and Evolution. (WA. Shear, ed). Stanford Univ. Press, Stanford, California. Legel, G.J. & WK.R.E. van Wingerden. 1980. Ex- periments on the influence of food and crowding on the aeronautic dispersal of Erigone arctica (White 1852) (Araneae Linyphiidae). Proc. 8th Intern. ArachnoL Congr. Pp. 1-6. Parker, G.A. & R.A. Stuart. 1976. Animal behavior as a strategy optimizer: Evolution of resource as- sessment strategies and optimal emigration thresholds. American Nat., 110:1055-1076. Riechert, S.E. & R.G. Gillespie. 1986. Habitat choice and utilization in web-building spiders. Pp. 23-48. In Spiders: Webs, Behavior and Evo- lution. (WA. Shear, ed). Stanford Univ. Press, Stanford, California. Sokal, R.R.& EJ. Rohlf. 1981. Biometry. 2nd ed., WH. Freeman & Co., New York. Sunderland, K.D., A.M. Fraser, & A.F.G. Dixon. 1986. Field and laboratory studies on money spi- ders (Linyphiidae) as predators of cereal aphids. J. Appl. Ecol., 23:433-447. Thomas, C.F.G. 1992. Spatial dynamics of spiders in farmland. Ph.D. Thesis, University of South- ampton. Thomas, C.F.G. 1996. Modelling aerial dispersal of spiders. Aspects Appl. Biol., 46:217-222. Thomas, C.F.G. 1997. Modelling the effect of available dispersal time on the size and persis- tence of linyphiid spider populations in an agri- cultural landscape. Pp. 127-134. In Species Dis- persal and Land Use Processes. (A. Cooper & J. Power, eds.). Proc. 6th annual I ALE (UK) con- ference, University of Ulster, Coleraine. 300 THE JOURNAL OF ARACHNOLOGY Thomas, C.F.G. & RC. Jepson. 1997. Field-scale effects of farming practices on linyphiid spider populations in grass and cereals. Entomol. Exp. Applic., 84:59-69. Vugts, H.F & W.K.R.E van Wingerden. 1976. Me- teorological aspects of aeronautic behaviour of spiders. Oikos, 27:433-444. Weyman G.S., K.D. Sunderland & J.S. Fenlon. 1994. The effect of food deprivation on aero- nautic dispersal behavior (ballooning) in Erigone spp. spiders (Araneae, Linyphiidae). Entomol. Exp. Applic., 73:121-126. Weyman, G.S. & P.C. Jepson. 1995. Effect of food supply on the colonization of barley by aerially dispersing spiders (Araneae). Oecologia, 100: 386-390. Weyman, G.S., P.C. Jepson, & K.D. Sunderland. 1995. Do seasonal changes in numbers of aeri- ally dispersing spiders reflect population density on the ground or variation in ballooning moti- vation? Oecologia, 101:487-493. Manuscript received 1 May 1998, revised 13 Sep- tember 1998. 1999. The Journal of Arachnology 27:301-307 PREY CHOICE AND SPIDER FITNESS S0ren Toft: Department of Zoology, University of Aarhus, Building 135 DK-8000 Arhus, Denmark ABSTRACT. Although spiders in general are polyphagous, indiscriminate feeding is not advantageous because prey vary enormously in quality due to toxicity or nutrient deficiency. Active prey selection serves to find the optimal compromise between three “nutritional goals”: maximize energy intake, balance nutrient composition of the body, and minimize toxin consumption. Consumption of toxic prey is reduced by more or less specific induced aversions, probably associated with both prey taste and behavior. Spiders’ ability to avoid toxic prey seems limited because aversions are short-lasting and some toxic prey do not induce an aversion. Such prey may be lethal. Toxic prey in a mixed diet may inhibit feeding on and utilization of good prey. Induced tolerance to toxic prey may be possible, however. Nutritional balance may be obtained through consumption of high-quality prey or through mixing of prey types. It is argued that nutrient balance is more important than maximization of energy intake for fitness. Foraging decisions influence individual fit= ness in a variety of ways. Choice of foraging habitat (patch) has been recognized as being of primary importance through its effect on feeding rates, with derived benefits to growth (size) and reproduction (Riechert 1981; Morse & Stephens 1996). Once in a feeding patch, the spider is confronted with an array of po- tential prey species. Prey selection is the con- sumption of prey relative to the composition of prey available in the spider’s microhabitat. So defined, selectivity has two aspects {cf. Pastorok 1981; Sih & Moore 1990): 1) passive selection or capture success, and 2) active se- lection (“choice”), i.e., acceptance/rejection of the prey. Both are clearly important for de- termining the diet composition in the field. The capture success reflects the co-evolution- ary balance between prey and predator in a broad sense (Malcolm 1992) and is largely outside the individual spider’s control. Active selection reflects the spider’s decisions (choic- es) whether positive (acceptance) or negative (rejection). Spiders are selective if they choose differentially between the prey species according to their preferences. Such prefer- ences by individuals may change over short time spans and, contrary to expectations, they are not necessarily related in a simple way to the nutritional value of the prey (see later). Active and passive selection can be distin- guished if prey is presented with and without possibilities for escape (Onkonbury & For- manowicz 1997; Lang & Gsodl unpubl. data). This review is concerned only with active se- lection or “prey choice,” and in our own ex- periments we have confined spiders and prey to small containers. Additionally, I have large- ly neglected the phenomenon of size selectiv- ity and focus on choice between prey species. Finally, my treatment is biased towards wolf spiders (Lycosidae) and money spiders (Lin- yphiidae). These two families account for most of the species and individuals of spiders found in agricultural fields of Northern and Central Europe (Sunderland 1987; Nyffeler & Benz 1988a, b; Toft 1989), and most of the studies reviewed were completed with the specified goal of analyzing potentially impor- tant trophic links of the agroecosystem. The experimental prey types were likewise select- ed to represent the most important prey groups as revealed by studies of European cereal fields, i.e., Collembola, Diptera and aphids (Nyffeler & Benz 1988a, b; Sunderland et al. 1986, 1987; Alderweireldt 1994). Food contains three main components: en- ergy, nutrients, and toxins. Viewed in isolation, feeding behavior should aim at the following three “nutritional goals”: maximize energy in- take, balance body nutrient composition, and minimize toxin consumption. They cannot all be realized at the same time. Prey choice is expected to achieve the optimal compromise that maximizes the fitness of the spider. As spiders are generalist predators, we might a priori expect that energy gain would be maximized by accepting all kinds of prey. 301 302 THE JOURNAL OF ARACHNOLOGY i.e., by being non-selective. In line with this, the broad polyphagy of spiders is considered to follow from the prevailing siu and- wait strategy and a temporally varying food supply (Riechert & Luczak 1982; Riechert & Harp 1987). In the notion of polyphagy it is usually understood that most kinds of potential prey are of approximately equal value to the pred- ator and may substitute each other in the pred- ator’s diet {cf. Slansky & Scriber 1985; Wald- bauer & Friedman 1991; Wise 1993). Rejection of toxic, dangerous or difficult prey (Riechert & Luczak 1982; Nentwig 1987) as well as novel (unfamiliar) prey (Turnbull 1960; Riechert & Luczak 1982) has, however, been recognized. Prey selection aimed at ob- taining a specific composition of amino acids was indicated by Greenstone (1979); but, gen- erally, nutrient balance is considered to be achieved through a mixed diet (Riechert & Harp 1987; Uetz et al. 1992). In conclusion, though the existence of active choice has been acknowledged, most reviews have considered it to be of limited importance. The questions raised here are: 1) Does ac- tive prey choice improve fitness? 2) What mechanisms influence prey choice? 3) Are the three nutritional goals equally important? 4) Do the answers to these questions confirm the inferior role of active selection for spider nu- trition? PREY QUALITY Prey differ enormously in quality as food for spiders, as indicated by the effects on fit- ness parameters of keeping spiders on single- species diets (Toft 1995, 1996; Sunderland et al. 1996a, b; Toft & Wise 1999a). Based on laboratory experiments with the wolf spider Schizocosa sp., the latter authors establish five quality categories: 1) high-quality prey (e.g., the collembolan Tomocerus bidentatus) are nutritionally complete; single-species diets al- low complete development (possibly full life cycle); 2) intermediate-quality prey (e.g., lab- oratory fruit flies Drosophila melanogaster) give initially high growth rates, but are insuf- ficient for full development and the spiders die before maturity; 3) low-quality prey (e.g., sciarid midges and conspecifics) allow very little growth and development and spiders die in an early instar; 4) poor-quality prey (e.g., several aphid species) allow neither growth nor development, and performance is no bet- ter than for starved controls; 5) toxic prey (e.g., the collembolan Folsomia Candida) re- sult in the spiders dying faster than starved controls. Prey species may have roughly the same quality characteristics for all polypha- gous predators, as basically similar results have been obtained for lycosids, linyphiids and carabid beetles (Bilde & Toft 1994, 1997a, b). With single-species diets no selection is possible. Furthermore, prey species may be insufficient as single prey but make positive contributions to fitness as parts of mixed diets. DIETARY MIXING Experiments with mixed diets allow us to analyze the extent to which spiders can choose an optimal diet, given varying prey availabil- ities. Uetz et al. (1992) demonstrated in- creased survival and growth rate in Lycosa spp. on a mixed diet when compared to a sin- gle-prey diet. Toft & Wise (1999a) found the same for Schizocosa sp., when fruit-flies and the high-quality collembolan Tomocerus were mixed. However, other mixed diets have re- vealed conflicting results. Thus, performance of Schizocosa was below that of the starvation control with two diets mixing two toxic Col- lembola with higher-quality prey. Toft (1995) found improved reproductive success in the linyphiid spider Erigone atra when females were given a mixed diet of fruit-flies and the poor-quality aphid Rhopalosiphum padi. From these results it is hypothesized that the posi- tive effect of dietary mixing depends on the quality of the prey species being mixed in the following way: mixing of higher-quality prey may or may not be beneficial; mixing of high- quality prey with prey of inferior quality may be beneficial as long as toxic prey is not in- cluded; nfixed diets including toxic prey may also be toxic even if higher-quality prey is in- cluded. Toft (unpubl. data) tested these generaliza- tions in a comparable experiment with Par- dosa prativaga and prey mainly from Euro- pean agricultural fields. This study revealed no positive effects of dietary mixing, but con- firmed the low quality of nfixed diets consist- ing of toxic and higher-quality prey. Thus, the most consistent outcome in these experiments was not a positive mixing effect, but that of toxic prey elinfinating the possible benefits of higher-quality prey. toft— PREY CHOICE AND SPIDER FITNESS These results clearly show that under the experimental conditions, the spiders were un- able to select the most profitable diet from the mixture of good and bad prey available. In some cases (F. Candida and fruicflies) accep- tance of the toxic prey led to a quick death of the spider in spite of high availability of good prey. So far, there is no information available on the impact of toxic prey on spiders in the field, or on whether spiders in the field are better able to avoid consuming these prey. ACQUIRED AVERSION AGAINST DETERRENT OR TOXIC PREY What mechanisms do spiders have to re- duce or avoid consumption of poor-quality or toxic prey? Clearly, smell or taste may act as deterrents, perhaps as signals of unpalatability or toxicity. Thus, Bilde & Toft (1994) found reduced acceptance by a carabid beetle of fruit flies coated with homogenate of an aphid or fungus gnat. The response is rarely all-or- none, however. Prey species may be classified according to the spiders’ response to them: (a) Some potential prey types are never even at- tacked by spiders; spiders may have an inher- ent selectivity against them (c/. Nentwig 1987). (b) Other prey types are attacked but discarded uneaten. The difference between this group and the first may be only the strength of the signal that informs the spider about the unpalatability of the prey: if per- ceived at a distance attack is prevented, (c) Many prey types are accepted readily and eat- en on several encounters, but eventually re- jected: an aquired aversion has developed. These prey must be moderately deterrent at most, (d) Prey that do not induce aversions: palatable prey. Most often these are high- or intermediate quality prey, (e) Prey that do not induce aversions but are nevertheless toxic (e.g., F. Candida). These prey may be non- deterrent (i.e., palatable) and spiders (at least young individuals) may die even if the prey are only part of a mixed diet. Tolerance may eventually be induced (see below). Acquired prey aversions occur when the spider’s preference for a prey is reduced fol- lowing consumption of similar prey, and are probably the main mechanism for limiting consumption of potentially toxic food (Ber- nays 1993). Several questions arise concern- ing their specificity and duration. Is an aver- sion associated with a certain taste of the prey. 303 with its morphology/behavior or with both? Do spiders learn to associate smell/taste and other prey characteristics in order to be able to avoid poor prey? General answers cannot be given at the moment, but two examples indicate that both taste and behavior can be important for prey recognition. Toft (1997) studied intra- and interspecific aversions of P. prativaga against three species of cereal aphids that are similar in behavior and perhaps also in chemical deterrency. Induction of aver- sion was graded in terms of the number of aphids needed to create it, reflecting a differ- ence in palatability of the aphids to the spider. Also, the duration of the aversion (i.e., the time until the next aphid was accepted) de- pended on which aphid induced the aversion. However, no matter which aphid induced the aversion, the “aversive” spider showed no differential response in encounters with new aphids. Thus, the motivation to attack was de- termined by the aversion rather than by the aphid actually confronting the spider, i.e., ir- respective of its palatability. In another experiment with Schizocosa sp., the spiders’ responses to low-quality fungus gnats were recorded (Toft & Wise 1999b). Spiders were offered prey sequentially and al- lowed to eat them one at a time. The next prey was offered when the spider had eaten the pre- vious one completely or repeatedly rejected it. In one series only fungus gnats were offered; in a second series only fruit-flies; in a third series fungus gnats and fruit flies were offered alternately. Both prey types were accepted ini- tially, and fruit flies continued to be accepted and eaten completely until satiation in both series. Fungus gnats, however, were often re- jected after 4-6 had been consumed, with no difference between the single-prey and two- prey series. In the series given only fungus gnats, most spiders eventually ignored the fungus gnats completely. Presumably, they re- lied on their experience from the last several captures that only fungus gnats were available and used behavior (flying insect) as the cue to prey recognition. In the mixed treatment, where there was a 50% chance that a flying insect would be a palatable one, the spiders continued to catch fungus gnats, only to re- lease them (mostly alive) when they recog- nized their identity by taste. Aversions may modify active prey choice at any of the successive stages of the capture 304 THE JOURNAL OF ARACHNOLOGY process (c/. Endler 1991). Following recog- nition of a potential prey, the spider may com- pletely ignore it, or attack-and-retreat if the prey’s unpalatability is recognized during at- tack (but before or at the bite). In the subju- gation phase (following bite or wrapping) the prey may be left dead or released alive. In the consumption phase, partial consumption may signify an aversion. Notice that acceptance/re- jection is not all-or-none but a graded re- sponse, which reflects its conditional nature. Spiders should become more selective when prey availability is high (Riechert 1981; Riechert & Luczak 1982; but see Riechert 1991), or selectivity may depend on an ac- quired aversion which may develop gradually. As argued by Riechert & Luczak (1982) re- jection should occur as early as possible in the predatory sequence. However, acquired aver- sions indicate that knowledge about the prey depends on experience which takes time to gain, and the information obtained may be un- certain. In the fungus gnat experiment de- scribed above (Toft & Wise 1999b), the spi- ders at first accepted and ate the prey. However, as experience accumulated, they stopped eating (discarded partly eaten prey; released captured prey, very often alive), and eventually stopped attacking the prey (retreat- ing, ignoring the prey). Whether refusing to eat subsequently leads to ignoring of the prey probably depends on the certainty with which the spider is able to identify the prey at a dis- tance. The example indicates that this ability depends not only on the spiders’ “knowl- edge” of (experience with) the prey charac- teristics, but also on the spider’s experiences (expectations) with respect to what prey is available. An aversion reduces the amount of a poor- quality prey that a spider consumes. However, some poor-quality prey may contribute posi- tively in mixed diets. Thus, complete exclu- sion from the diet is only advantageous if the prey is always detrimental. A limited duration of an aversion may thus serve to secure a con- stantly low intake rate, which balances nutri- ent benefits and toxic damage to the positive side, thus creating a synergistic effect. Toft (1997) measured the duration of aphid aver- sions in a wolf spider to be mostly < 24 hours. Since only about two aphids were needed to create the aversion in the first place, and probably fewer are needed to reestablish one, the daily rate of feeding on aphids will be kept quite low. In quantitative estimates with P. prativaga, consumption of aphids were only */l0 or less of the spider’s food de- mand as determined with fruit flies (Toft 1995). For some prey types, however, aversions do not protect the spider against toxic overload. Duration of the aversion against the collem- bolan F. Candida (animals from USA) was of the same order of magnitude as for aphids, up to ca. 24 hours. This was too short to prevent chronic inhibition of feeding and growth in Schizocosa (Toft & Wise 1999b). In similar tests with P. prativaga and (presumably) the same collembolan species from Europe, an aversion could not be induced (D. Mayntz un- publ. data). With constant availability this Collembola is lethal to P. prativaga. However, surviving hatchlings raised on fruit flies with a limited supplementation of F. Candida even- tually developed a partial resistance to F. Can- dida. After 5-6 weeks of inhibited growth the young spiders showed compensatory growth and caught up with the fruit fly controls in a few weeks’ time (Toft unpubl. data). Such in- duced resistance to toxic prey may explain why large juvenile Schizocosa were not inhib- ited by F. Candida, even though small juve- niles were (Toft & Wise 1999b). Nentwig (1985) described digestive modifications in a spider following prolonged feeding on KCN- treated prey. Turnbull (1960) and Riechert & Luczak (1982) noted neophobia, i.e., reluctance to ac- cept unfamiliar prey that were later readily ac- cepted, in two web-spinning spiders. We have not observed this in wolf spiders, but it may be more prevalent in webspinners in which the web may intercept large and dangerous prey. NUTRITIONAL QUALITY OF PREY Greenstone’s (1979) work on prey selection in Pardosa ramulosa, indicating nutritional self-selection (cf. Waldbauer & Friedman 1991) for essential amino acids, is still unique. The huge divergence in food quality of poten- tial prey species demonstated above should leave us more open to accept this possibility. If prey are deterrent or toxic, this fact may completely override any differences in nutri- tional composition. However, the improved hatching success of Erigone atra eggs from females given a mixed fruit fly-aphid diet in- toft— PREY CHOICE AND SPIDER FITNESS dicates that even low consumption of the poor-quality aphid gave a significant nutrient supplement to the fruit flies (Toft 1995). Also, palatable prey may differ in nutrient quality. Several authors have stated that fruit flies are nutritionally insufficient for complete devel- opment of spiders (Miyashita 1968; Riechert & Harp 1987). Recent studies show that nu- trient additions to the standard fruit fly me- dium create flies of enhanced quality to spi- ders (Kristensen & Toft unpubl. data; Mayntz & Toft unpubl. data). Wolf spider hatchlings fed flies raised on standard and nutrient-im- proved media, respectively, show differences in survival and growth rates after six weeks. Thus, spiders may enhance their fitness by se- lecting the most nutritious prey species and even the most nutritious individuals of each species, if that is possible. It can be hypoth- esized that nutritional selectivity increases with degree of nutritional imbalance, which is most likely to develop when a limited range of prey species is available, as is the case in most feeding experiments. Greenstone’s (1979) results were obtained in a situation where diversity of potential prey was limited to three species. Toft (1996) proposed a graphical model predicting the relationship between a preda- tor’s tolerance (i.e., maximal consumption ca- pacity) to various types of prey and its per- formance if the diet is restricted to one prey type. If prey are deterrent or toxic and con- sumption capacity is therefore low, the pred- ator is unable to reach satiation on this prey and fitness will be low. If prey is non-toxic and palatable, the predator may reach satiation on this prey type. If it is also of optimal nu- trient composition, fitness (growth or repro- duction) is maximal on a feeding rate equiv- alent to the maximal rate at which food can be converted into spider tissue (or eggs). If the prey is deficient in essential nutrients, the spider may to some degree compensate by in- creasing consumption at the expense of food utilization. Thus, the model predicts an in- creased consumption rate of nutritionally de- ficient prey associated with a lowered benefit. A test of these predictions was provided by Marcussen et al. (in press) in experiments with the linyphiid spider Erigone atra. A col- lembolan (Folsomia fimetarid) was clearly toxic. Both consumption and reproduction were low, and supplementing a fruit fly diet 305 with this collembolan reduced reproduction compared to the pure fruit fly diet. Another collembolan (Isotoma anglicana) was found to be of very high quality, significantly better than fruit flies, in terms of both number of egg-sacs and eggs/sac produced by the fe- males. However, quantification of daily con- sumption rates showed that the spiders con- sumed > 1.5X more fruit fly mass than /. anglicana mass. Thus, the spiders attained the maximal reproductive rate by a moderate con- sumption rate of nutritionally high-quality prey. A result that can be interpreted in the same vein is the finding that limited and un- limited rations of fruit flies gave the same re- productive output in E. atra (Toft 1995). Pre- sumably the limited feeding rate was nutritionally superior due to more efficient nu- trient extraction of each prey item (Toft 1996). An interesting consequence of these results is that fitness maximization is achieved by op- timal balancing of nutrients rather than by maximization of energy consumption. CONCLUSIONS This review has demonstrated that two con- ditions for the evolution of prey selectivity of spiders, viz. the differences in food quality of various prey and the consequences of prey quality for spider fitness, are met. In several experiments, however, poor performance of the spiders proved that what the spiders chose was not the optimal selection from the avail- able prey mixture. Benefits of prey mixing were found, but so far there has been no ex- perimental demonstration that spiders can choose the optimal diet when a mixture of prey of various qualities is available. Also, nu- tritional self-selection still needs to be exper- imentally established. Selectivity is evident in the spiders’ responses to some (but not all) low- and poor-quality prey types, which are accepted at a rate much below availability. Acquired aversions are probably the main be- havioral mechanism for reducing consumption of toxins and are the most prominent expres- sion of selective prey choice. A short aversion memory may serve to balance toxic damage with the nutritional benefit of a diverse diet. ACKNOWLEDGMENTS I am indebted to several students and col- leagues for their contribution to the develop- ment of the ideas presented in this paper — in 306 THE JOURNAL OF ARACHNOLOGY particular, Jprgen Axelsen, Trine Bilde, Hel- ene Bracht Jprgensen, Bente Marcussen, Da- vid Mayntz, Keith Sunderland, Ove Flemming Sprensen and David Wise. Most of the work was made possible through a grant from the Danish Environmental Research Program to the Centre for Agricultural Biodiversity. LITERATURE CITED Alderweireldt, M. 1994. Prey selection and prey capture strategies of linyphiid spiders in high- input agricultural fields. Bull. British Arachnol. Soc., 9: 300-308. Bemays, E.A. 1993. Aversion learning and feed- ing. Pp. 1-17. In Insect Learning. Ecology and Evolutionary Perspectives. (D.R. Papaj & A.C. Lewis, eds.). Chapman and Hall, New York. Bilde, T. & S. Toft. 1994. Prey preference and egg production of the carabid beetle Agonum dorsale. Entomol. Exp. AppL, 73:151-156. Bilde, T. & S. Toft. 1997a. Limited predation ca- pacity by generalist arthropod predators on the cereal aphid, Rhopalosiphum padi. Biol. Agric. Hort., 15:143-150. Bilde, T. & S. Toft. 1997b. Consumption by cara- bid beetles of three cereal aphid species relative to other prey types. Entomophaga, 42:21-32. Endler, J.A. 1991. Interactions between predators and prey. Pp. 169-196. In Behavioral Ecology. An Evolutionary Approach (J.R. Krebs & N.B. Davies, eds.). Blackwell, Oxford. Greenstone, M.H. 1979. Spider feeding behaviour optimises dietary essential amino acid composi- tion. Nature, 282:501-503. Malcolm, S.B. 1992. Prey defence and predator foraging. Pp. 458-489. In Natural enemies (M.J. Crawley, ed.). Blackwell, Oxford. Marcussen, B., J.A. Axelsen & S. Toft. In press. The value of two collembola species as food for a cereal spider. Entomol. Exp, Appl. Miyashita, K. 1968. Growth and development of Lycosa Tdnsignita Boes. et Str. (Araneae: Ly- cosidae) under different feeding conditions. Appl. Entomol. ZooL, 3:81-88. Morse, D.H. & E.G. Stephens. 1996. The conse- quences of adult foraging success on the com- ponents of lifetime fitness in a semelparous, sit and wait predator. Evol. EcoL, 10:361-373. Nentwig, W. 1985. Spiders eat crickets artificially poisoned with KCN and change the composition of their digestive fluid. Naturwissenschaften, 72: 545-546. Nentwig, W. 1987. The prey of spiders. Pp. 249- 263. In Ecophysiology of Spiders (W. Nentwig, ed.). Springer- Verlag, Berlin. Nyffeler, M. & G. Benz. 1988a. Feeding ecology and predatory importance of wolf spiders {Par= dosa spp.) (Araneae, Lycosidae) in winter wheat fields. J. Appl. Entomol., 106:123-134. Nyffeler, M. & G. Benz. 1988b. Prey and preda- tory importance of micryphantid spiders in win- ter wheat fields and hay meadows. J. Appl. En- tomol., 105:190-197. Onkonbury, J. & D.R. Formanowicz. 1997. Prey choice by predators: effect of prey vulnerability. EthoL EcoL EvoL, 9:19-25. Pastorok, R.A. 1981. Prey vulnerability and size selection by Chaoborus larvae. Ecology, 62: 1311-1324. Riechert, SE. 1981. The consequences of being territorial: spiders, a case study. American Nat., 117:871-892. Riechert, S.E. 1991. Prey abundance vs. diet breadth in a spider test system. Evol. EcoL, 5: 327-338. Riechert, S.E. & J.M. Harp. 1987. Nutritional ecol- ogy of spiders. Pp. 645-672. In Nutritional Ecol- ogy of Insects, Mites, Spiders and Related In- vertebrates (F. Slansky & J.G. Rodriguez, eds.). John Wiley & Sons, New York. Riechert, S.E. & J. Luczak. 1982. Spider foraging: behavioral responses to prey. Pp. 353-385. In Spider Communication. Mechanisms and Eco- logical Significance (P.N. Witt & J.S. Rovner, eds.). Princeton Univ. Press, Princeton. Sih, A. & R.D. Moore. 1990. Interacting effects of predator and prey behaviour in determining diets. Pp. 771-796. In Behavioural Mechanisms of Food Selection (R.N. Hughes, ed.). Springer- Ver- lag, Heidelberg. Slansky, F. & J.M. Scriber. 1985. Nutrition. In Comprehensive Insect Physiology, Biochemistry and Pharmacology. VoL 4. Regulation, Diges- tion, Nutrition, Excretion (G.A. Kerkut & L.I. Gilbert, eds.). Pergamon Press, Oxford. Sunderland, K.D. 1987, Spiders and cereal aphids in Europe. SROP/WPRS Bulk, 10:82-102. Sunderland, K.D., A.M. Fraser & A.F.G. Dixon. 1986. Distribution of linyphiid spiders in rela- tion to capture of prey in cereal fields. Pedo- biologia, 29:367-375. Sunderland, K.D., N.E. Crook, D.L. Stacey & B.J. Fuller. 1987. A study of feeding by polyphagous predators on cereal aphids using ELISA and gut dissection. J. Appl. EcoL 24:907-933. Sunderland, K.D., T. Bilde, L.J.M.E den Nijs, A. Dinter, U. Heimbach, J.A. Lys, W. Powell & S. Toft, 1996a. Reproduction of beneficial preda- tors and parasitoids in agroecosystems in relation to habitat quality and food availability. Pp. 1 17- 153. In Arthropod Natural Enemies in Arable Land. II. Survival, reproduction and enhance- ment (K. Booij & L. den Nijs, eds.). Aarhus Univ. Press, Arhus. Sunderland, K.D., C.J. Topping, S, Ellis, S. Long, S. Van de Laak & M. Else, 1996b. Reproduction and survival of linyphiid spiders, with special reference to Lepthyphantes tenuis (Blackwall). TOFT— PREY CHOICE AND SPIDER FITNESS 307 Pp. 81-95. In Arthropod Natural Enemies in Ar- able Land. IL Survival, reproduction and en- hancement (K. Booij & L. den Nijs, eds.). Aar- hus Univ. Press, Arhus. Toft, S. 1989. Aspects of the ground-living spider fauna of two barley fields in Denmark: species richness and phenological synchronization. En- tomoL Meddr., 57:157-168. Toft, S. 1995. Value of the aphid Rhopalosiphum padi as food for cereal spiders. J. AppL EcoL, 32:552-560. Toft, S. 1996. Indicators of prey quality for arthro- pod predators. Pp. 107-116. In Arthropod Nat- ural Enemies in Arable Land. IL Survival, repro- duction and enhancement (K, Booij & L. den Nijs, eds.). Aarhus Univ. Press, Arhus. Toft, S. 1997. Acquired food aversion of a wolf spider to three cereal aphids: intra- and interspe- cific effects. Entomophaga, 42:63-69. Toft, S. & D.H. Wise. 1999a. Growth, development and survival of a generalist predator fed single- and mixed-species diets of different quality. Oec- ologia, 119:191-197. Toft, S. & D.H. Wise. 1999b, Behavioral and eco- physiological responses of a generalist predator fed single- and mixed-species diets of different quality. Oecologia, 119:198-207. Turnbull, A.L. 1960. The prey of the spider Liny- phia triangularis (Clerck) (Araneae, Linyphi- idae). Canadian J. ZooL, 38:859-873. Uetz, G., J. Bischoff & J. Raver. 1992. Survivorship of wolf spiders (Lycosidae) reared on different diets. J. ArachnoL, 20:207-211. Waldbauer, G.P. & S. Friedman, 1991. Self-selec- tion of optimal diets by insects. Annu. Rev. En- tomoL, 36:43-63. Wise, D.H, 1993. Spiders in Ecological Webs. Cambridge University Press, Cambridge. Manuscript received 1 May 1998, revised 30 Sep- tember 1998, 1999. The Journal of Arachnology 27:308-316 MECHANISMS UNDERLYING THE EFFECTS OF SPIDERS ON PEST POPULATIONS Keith Sunderland; Department of Entomological Sciences, Horticulture Research International, Wellesboume, Warwick CV35 9EF, U.K. ABSTRACT. Assemblages of spider species can make significant reductions in pest numbers that are of value to the farmer. A group of spider species with complementary niches leaves few refuges for the pest in space or time. Spiders usually exert an influence on pest numbers in concert with other natural enemies, and spiders are sometimes the dominant component. In addition to killing pests by direct attack, spiders cause pest mortality by dislodging them from plants or trapping them in webs. If the pest is distasteful, or if it is the dominant prey type available, spiders may kill more than they consume, which increases the rate of pest kill per unit of spider food demand. The implications for pest control, of various types of interaction between spiders and other natural enemies, are explored in this paper. Interactions with specialist natural enemies usually result in complementary effects, enhancing pest control. Specialists reduce the density of pests to levels where spiders can prevent resurgence. Specialists foraging on the crop may flush pests off the plant to be killed by ground=zone spiders. Although hyperpredation (i.e., predators killing other predators) may disrupt biological control occasionally, it is considered that the wide range of competitive interactions between natural enemies, in general, promotes diversity and stability of the natural enemy community and generates a robust basis for pest control. There are currently in excess of 3000 de^ scribed genera of spiders (Coddington in lit), more than 50,000 species are predicted to be living on the planet, and they are the dominant insectivores in some terrestrial ecosystems (Thompson 1984). They are of economic val- ue to man because of their ability to suppress pest abundance in agroecosystems. Faced with the need to reduce pesticide usage on the world’s crops and optimize natural biological control, full investigation of the means by which spiders influence pest abundance is long overdue. Also, in recent years, there has been a realization by ecologists that compo- nents of agroecosystems are tractable to ma- nipulate and that spiders are convenient model organisms. Consequently, there are a growing number of investigations in which spiders in agroecosystems are used as tools to gain fun- damental insights into the role of generalist predators in community and ecosystem func- tion. This review is a brief exploration of the ma- jor routes by which spiders can influence the abundance of invertebrate pests. DIRECT PREDATION BY SPIDERS By spider assemblages. — Individual spe- cies of spider may, occasionally, make suffi- cient impact on a herbivore population for measurable effects on plant production to be registered (Louda 1982), but there is currently no evidence for this in agriculture. Consider- ation of some aspects of the biology of spiders (e.g., generation times and functional and nu- merical responses) leads to the conclusion that individual spider species are unlikely to reg- ulate pest populations (Riechert & Lockley 1984; Riechert 1992; Wise 1993). However, when assembled into groups of species, as is the norm in agriculture (Sunderland et al. 1997), they often contribute to significant re- ductions in pest numbers that are of value to the farmer. In an individual-based model, in- creasing the number of spider species contrib- uted significantly to prey limitation (Prov- encher & Riechert 1994). Nentwig (1982) found indications that the size-frequency structure of the aracheofauna matches that of their potential prey, and he suggested a sig- nificant role for spiders as a multi-predator complex for reducing a multi-prey complex. Experimental field manipulations, of the mean abundance of spider assemblages, has dem- onstrated a significant level of impact of spi- ders on leafhoppers in rice (Graze & Grigarick 1989), on caterpillars in taro (Nakasuji et al. 1973), cotton (Mansour 1987) and orchards 308 SUNDERLAND— MECHANISMS UNDERLYING EFFECTS OF SPIDERS ON PESTS 309 (Mansour et al. 1980), scale insects in or- chards (Mansour & Whitcomb 1986), and var- ious pests in vegetables (Riechert & Bishop 1990) and old fields (Provencher & Riechert 1994; Riechert & Lawrence 1997). The effect of spider predation on pest populations can be sufficient to reduce significantly levels of crop damage (Riechert & Bishop 1990; Carter & Rypstra 1995). Spider species often have com- plementary niches (Whitcomb 1974; Nyffeler & Sterling 1994), segregating in terms of di- mensions such as vertical location, diel cycle and foraging mode (Marc & Canard 1997), and, as an assemblage, may be able to kill all growth stages of a pest, including the eggs (Nyffeler et al. 1990). This means that there are fewer refuges for the pest in time or space (less “enemy-free space” — Jeffries & Lawton 1984), and this is to the benefit of pest control (Murdoch 1990). By a group of natural enemies including spiders.- — Spiders need not act alone to be of value in agriculture. They also have a valid role as one component of a larger complex of natural enemies with the potential to keep pests at non-damaging levels. They play this role in the control of Colorado potato beetle (Cappaert et al. 1991) and mosquitoes (Ser- vice 1973), for caterpillar control on cotton (Gravena & Da-Cuhna 1991) com (Coll & Bottrell 1992; Clark et al. 1994) and in for- estry (Mason et al. 1983), against aphids on cotton (Chen et al. 1994) and apple (Wyss et al. 1995), and against hoppers and other pests on rice (Sawada et al. 1993; Kamal & Dyck 1994; Settle et al. 1996). As would be ex- pected, the relative contribution of spiders compared to other natural enemies varies with crop and season and in response to many other factors. Spiders were not the dominant ele- ment in the predatory complex controlling Helicoverpa spp. caterpillars on cotton in Australia (Bishop & Blood 1981). However, on cotton in Texas more than 80% of preda- tors observed to kill cotton fleahopper {Pseu- datomoscelis seriatus) were spiders (Nyffeler et al. 1992); and they accounted for 73% of the net value of predators, compared to 27% for insects (Sterling et al. 1992). Spiders are the most abundant natural enemies in cotton fields throughout China (55-81% of all natu- ral enemies); and they play a key role, togeth- er with the other dominant natural enemies, in suppressing pest populations (Zhang 1992). ADDITIONAL ATTRIBUTES OF VALUE FOR PEST CONTROL Pest dislodgement.— The foraging behav- ior of spiders on crop vegetation may disturb pest aggregations and may also cause the dis- turbed pests to walk or fall off the plants. This can reduce the pest population if the physical conditions on the ground cause rapid mortality (e.g., aphid survival times of only a few min- utes at ground temperatures above 40 °C in North American field crops; Dill et al. 1990), or if they cannot easily regain the plants, or if they move into danger zones with greater probabilities of attack by natural enemies. It is possible that pest species belonging to many Orders are dislodged by spiders, but the lit- erature emphasizes an effect on Lepidoptera. In manipulative experiments in apple orchards (Mansour et al. 1981) and in taro fields (Na- kasuji et al. 1973; Yamanaka et al. 1973) ap- proximately one third of caterpillars were dis- lodged, and the authors considered that the majority of dislodged individuals would be unable to regain the plant. In these examples dislodgement resulted in death of the pest. Under less extreme conditions the loss of feeding time resulting from dislodgement may be expected to reduce plant damage and also to reduce the rate of increase of the pest pop- ulation. Death of pests in webs not caused by spi- der predation. — Small pests, such as thrips, midges and aphids, may die by being caught in the webs of large spiders, even when they are ignored by the spider (Nentwig 1987). Al- der weireldt (1994) identified 319 prey items in webs of linyphiid spiders in maize fields in Belgium. Spiders were feeding on only 184 of these prey items. Linyphiidae, Dictynidae, Theridiidae and Agelenidae do not renew their webs daily, and feed infrequently (Nyffeler et al. 1994a), so these families may contribute to pest control by the action of their webs. First instars of the cereal aphid Sitobion avenae did not escape from webs of non-attacking sati- ated adult female linyphiid spiders, Lepthy- phantes tenuis (Sunderland et al. 1986). The proportion of S. avenae falling into webs (or sticky traps to simulate webs) that are first in- stars is typically 10-20% (Fraser 1982; Ken- nedy 1990). The mean duration of web-site tenacity in this species is less than two days (Samu et al. 1996), so the number of webs 310 THE JOURNAL OF ARACHNOLOGY may exceed the number of web-makers. In this study (Samu et al. 1996) nearly all L. ten- uis webs contained uneaten S. avenae and none of the 60 observed spiders were feeding. Thus the potential of webs to kill pests, in the absence of spider attack, can be a relevant consideration for biological control. Wasteful killings partial consumption and the wounding of pests. — Under certain circumstances the predator may kill a pest but subsequently ingest little (partial consump- tion) or none (variously referred to in the lit- erature as “superfluous killing” and “waste- ful killing”) of the pest’s biomass. This is advantageous for pest control because it will result in more pests being killed per unit of spider food demand. These behaviors are usu- ally observed when prey are plentiful (or when a small spider is able to overcome a large prey) and the spider is nearly or com- pletely satiated. The seemingly inappropriate behavioral overshoot of continuing to kill when enough food to induce satiation has al- ready been secured may be due to the time lag between prey capture and ingestion associated with the spider’s extra-oral digestion system (Riechert & Lockley 1984). This is the arach- nid equivalent of the gut compartmentaliza- tion theory proposed for insects (Johnson et al. 1975). There are examples of wasteful kill- ing at high prey density for Clubionidae and Linyphiidae against aphids (Provencher & Coderre 1987; Mansour & Heimbach 1993), for Linyphiidae (DeKeer &. Maelfait 1988) and Lycosidae (Samu & Biro 1993) against flies, and for Araneidae against hymenopteran parasitoids (Smith & Wellington 1983). Partial consumption has been recorded for Thomisi- dae and Lycosidae at high densities of Dro- sophila prey (Haynes & Sisojevic 1966; Samu 1993). These examples all refer to laboratory studies and it is not known how prevalent wasteful killing and partial consumption are under field conditions. Predators are not 100% efficient, and it is known that wounded prey may die following unsuccessful attacks by co- leopteran (Doane et al. 1985) and dipteran (Griffiths et al. 1984) predators. It is likely that spiders, also, cause some pest mortality by wounding leading to fatal infection or loss of haemolymph, but this will be difficult to quantify in the field. Mansour & Heimbach (1993) recorded a high rate of wasteful killing of the cereal aphid Rhopalosiphum padi by spiders, even at low aphid density. R. padi, in common with other species of cereal aphid, is a poor quality food for spiders (Toft 1995). They may find it distasteful and can develop an aversion to it, but such aversions persist for only a few hours (Toft 1997). Prey that cause an aversion re- sponse by the spider may be ignored, or at- tacked and released intact, or released wound- ed or dead, or killed and partially consumed (Nentwig 1985), depending on many vari- ables, including spider hunger and degree of naivete (Toft 1997). Such behavior would not be expected from a specialist natural enemy. When food availability is dominated by a non- preferred pest species (e.g. aphids constituting 83% of prey items in webs of L. tenuis in maize; Alder weireldt 1994) the spider popu- lation might even kill more pests than if the pests were a high-quality preferred food, be- cause spiders would remain unsatiated. This prey sampling-aversion-wasteful killing syn- drome also raises questions about our meth- odologies for determining the kill rate of spi- ders on pests, especially for species that do not construct webs. Post-mortem methods, such as electrophoresis, radio-tracers and an- tibody techniques (Sunderland 1988; Green- stone 1996), would fail because no ingestion has taken place. Quantitative methods based on direct observation of food being eaten by spiders in the field (e.g., Edgar 1970) would underestimate the impact on pests, because the probability of observing a kill and rejection incident is much lower than that of a kill and consume incident (the former being of shorter duration than the latter). SPIDERS IN COMMUNITIES Spiders in agroecosystems are components of species-rich communities of herbivores, detritivores and natural enemies. The effect of a spider species on a pest population may be enhanced if the spider population increases rapidly in response to a rich supply of nutri- tious alternative prey (Jeffries & Lawton 1984; Axelsen et al. 1997). However, if the pest species is less-preferred than the alter- native prey, the net effect of these opposing processes on the level of pest control will be difficult to predict (Bilde & Toft 1994). Se- lective predation by spiders in relation to the size of pest taken (Nentwig & Wissel 1986) can alter the mean body size of the pest pop- SUNDERLAND— MECHANISMS UNDERLYING EFFECTS OF SPIDERS ON PESTS 311 ulation, modifying its vulnerability to other size-dependent natural enemies in the com- munity (Strauss 1991). Some additional ex- amples of interactions between spiders and other natural enemies, with implications for pest control, are described below. Predation of moribund pests. — Predation by spiders of moribund parasitized pests (i.e., living pests that will be killed eventually by the developing parasitoid) is counter-produc- tive to biological control because the mortality of these pest individuals is already assured, and spider predation will reduce the size of the next generation of parasitoids. Predation of moribund parasitized pests could, however, be of value to the farmer in cases where the moribund pest continues to damage the plant significantly, or reproduces before death (Sun- derland 1996). Moribund pests often have phenologies, distributions, activity, defenses and palatability that are different from the healthy pest (Sunderland 1996), but there is very little information in the literature con- cerning how this influences the probability of capture by spiders (Coll & Bottrell 1992). Pre- dation of moribund diseased pests could be beneficial if the spider spreads the disease to other individuals in the pest population. For example, it is considered likely that Oxyopes salticus is an important disseminator of Anti- carsia gemmatalis Nuclear Polyhedrosis Virus in USA soybean (Bering et al. 1988). It is re- grettable that interactions between predators and moribund pests are rarely taken into ac- count in comparisons of the relative effective- ness of predators, parasitoids and pathogens in biocontrol (Hawkins et al. 1997). Interactions between spiders and special- ist predators. — Pest density: Spiders, as gen- eralist predators, can be present in a crop, feeding on alternative prey, before the pest ar- rives. High spider-pest ratios early in the sea- son (e.g., Wheeler 1973; Zhang 1992) may reduce the pest increase rate sufficiently to en- able later-arriving specialist natural enemies to suppress the pest population below the eco- nomic threshold, a conclusion also reached from metabolic pool modeling exercises (Ax- elsen et al. 1997). This is a finely-balanced relationship that can also have a negative out- come if synchronization is inadequate. For ex- ample, if generalists depress pest density be- low the oviposition threshold of immigrant specialists (Honek 1980; Ghanim et al 1984), the specialists are likely to leave the field in search of higher pest densities elsewhere, and the generalists may then be unable to prevent a pest outbreak occurring. Generalists and specialists do, however, work together har- moniously, if not synergistically, in many agroecosystems (e.g., Zhang 1992). In peren- nial systems, this is sometimes because spe- cialists reduce pest density, in one year, to a level from which generalists are able to pre- vent resurgence in later years (Mason & Tor- gersen 1987; Roland & Embree 1995). Spe- cialist predators, by themselves, tend to be unreliable (except for crops with economics that permit rear-and-release strategies) be- cause their densities are highly variable from year to year (Aebischer 1991). In contrast, it has been suggested (Sunderland et al. 1996) that the densities of generalists (e.g., spider assemblages) are buffered, in that a deficiency in the numbers of any one species in a given year is very likely to be counterbalanced by a superabundance of another species within the same guild. Spatial effects: Pests, such as aphids and caterpillars, are dislodged by foraging parasit- oids and predators, and especially by special- ists such as aphidophagous coccinellids (see review in Sunderland et al. 1997). Spider as- semblages are often vertically stratified in crops (e.g., Provencher et al. 1988; Marc & Canard 1997), and many spider species are confined to the ground zone or lower strata of vegetation (Wheeler 1973; Leathwick & Win- terbourn 1984; Heong et al. 1990). In U.K. winter wheat, the proportion of fallen aphids that climb back onto plants is negatively re- lated to the density of ground predators (Winder 1990; Duffield et al. 1996). Sixty-one out of 109 species of spider in this crop are confined to the ground zone (Sunderland et al. 1988), and the webs of linyphiids can cover 50% of the ground surface below the crop (Sunderland et al. 1986). Thus it is clear that spiders, and other ground predators, will make a greater contribution to aphid control in this crop in situations where aphids are flushed off the crop by specialist natural enemies. Competitive interactions between preda- tors.^ — Cannibalism, intra-specific competi- tion and territoriality (Wise 1993) may result in self-limitation of the density of a given spi- der species, and inter-specific interactions, including interference competition (Spiller 312 THE JOURNAL OF ARACHNOLOGY 1984; Moran & Hurd 1994), can result in fur- ther reductions in density. Complete elimina- tion of a competing species from the crop may be averted if the intensity of competition is ameliorated by the action of a top predator reducing the density of the dominant compet- itor (i.e,, exploiter-mediated coexistence, as applied to predators). An example of a top predator that might fulfil this role, in USA cot- ton, is the green lynx spider (Peucetia viri- dans), which is strongly araneophagous (Nyf- feler et al. 1987). Interactions such as these promote spider biodiversity and, in addition, cannibalism and hyperpredation (i.e., preda- tors killing predators) may buffer the spider community (i.e., prevent localized species ex- tinctions) during short-term dearth of herbi- vore and detritivore prey. These mechanisms reduce the availability of enemy-free space to pests, and their effects are enhanced by a be- havioral flexibility on the part of predators that permits them to make short-term niche shifts (Jeffries & Lawton 1984; Polls et al. 1989). For example, hunting spiders in USA field crops are highly polyphagous, but can narrow their feeding niche significantly when a suitable prey species reaches high numbers (Nyffeler et al. 1994b). Some species of spider have been shown to make little impact on other predators (Nen- twig 1975; Lockley & Young 1987; Jmhasly & Nentwig 1995), but others are significant predators of spiders, ants, lacewings, lady- birds, and predatory Heteroptera (Nentwig 1986; Nyffeler et al. 1987; Sengonca & Klein 1988; Heong et al. 1992; Nyffeler et al. 1994b; Dinter 1998). Hyperpredation is valu- able in promoting diversity and stability of the natural enemy community, but is occasionally detrimental to pest control when intense pre- dation of one predator by another releases a pest from a former level of satisfactory bio- logical control (Rosenheim et al. 1995). Spi- ders are included in the natural enemy com- plex implicated as reducing the effectiveness of lacewing release for leafhopper control in vineyards (Daane et al. 1996) and of penta- tomid release for suppression of Colorado po- tato beetle (Hough-Goldstein et al. 1996). CONCLUSIONS There are indications from the literature of many mechanisms whereby spiders can affect the abundance of invertebrate pests. Direct predation, pest dislodgement and wasteful killing (by both spider and web) reduce pest abundance, whilst predation of moribund pests and IGP may destabilize existing natural control and trigger indirectly an increase in the pest population. The relative importance of these various pathways in any given agroe- cosystem, and whether major pathways differ between agroecosystems, is not known. An- swering these questions is consistent with the development of a “community approach” to biological pest control and spiders are espe- cially apt subjects of study in this context be- cause they are known to exert their influence on pest populations as species assemblages and in concert with other groups of natural enemies. ACKNOWLEDGMENTS This work was funded by the U.K. Ministry of Agriculture, Fisheries & Food. I am very grateful for the help of the HRI Library staff, and for financial assistance to attend the Sym- posium from the USD A Cooperative State Re- search, Education and Extension Service Na- tional Research Initiative Competitive Grants Program. LITERATURE CITED Aebischer, N.J. 1991. Twenty years of monitoring invertebrates and weeds in cereal fields in Sus- sex. Pp. 305-331. In The Ecology of Temperate Cereal Fields (eds. L.G. Firbank, N. Carter, J.F. Darbyshire & G.R. Potts), Blackwell Scientific Publications, Oxford. Alderweireldt, M. 1994. Prey selection and prey capture strategies of linyphiid spiders in high- input agricultural fields. Bull. British Arachnol. Soc., 9:300-308. Axelsen, J.A., P. Ruggle , N. Holst & S. Toft. 1997. Modelling natural control of cereal aphids. III. Linyphiid spiders and coccinellids. Acta Jutl., 72:221-231. Bilde, T. & S. Toft. 1994. Prey preference and egg production of the carabid beetle Agonum dorsale. Entomol. Exp. AppL, 73:151-156. Bishop, A.L. & P.R.B. Blood. 1981. Interactions between natural populations of spiders and pests in cotton and their importance to cotton produc- tion in southeastern Queensland. Gen. Appl. En- tomol., 13:98-104. Cappaert, D.L., EA. Drummond & P.A. Logan. 1991. Population dynamics of the Colorado po- tato beetle (Coleoptera: Chysomelidae) on a na- tive host in Mexico. Env. Entomol., 20:1549- 1555. Carter, RE. & A.L. Rypstra. 1995. Top-down ef- SUNDERLAND— MECHANISMS UNDERLYING EEEECTS OE SPIDERS ON PESTS 313 fects in soybean agroecosystems: spider density affects herbivore damage. Oikos, 72:433-439. Chen, PR., Z.Q. Zhang, K. Wang, X.Y. Wang, WL. Xu & Z.L. Gao. 1994. Allothrombium pulvinum Ewing (Acari, Trombidiidae), an important early- season natural enemy of Aphis gossypii Glover (Horn., Aphididae) in cotton. J. Appl. EntomoL, 117:113-121. Clark, M.S., J.M. Luna, N.D. Stone & R.R. Young- man. 1994. Generalist predator consumption of armyworm (Lepidoptera, Noctuidae) and effect of predator removal on damage in no-till com. Env. EntomoL, 23:617-622. Coll, M. & D.G. Bottrell. 1992. Mortality of Eu- ropean com borer larvae by natural enemies in different com microhabitats. Biol. Cont., 2:95- 103. Daane, K.M., G.Y Yokota, Y Zheng & K.S. Hagen. 1996. Inundative release of common green lace- wings (Neuroptera: Chrysopidae) to suppress Er- ythroneura variabilis and E. elegantula (Homop- tera: Cicadellidae) in vineyards. Environ. EntomoL, 25:1224-1234. De Keer, R. & J.R Maelfait. 1988. Laboratory ob- servations on the development and reproduction of Erigone atra Blackwall, 1833 (Araneae, Lin- yphiidae). Bull. British Arachnol. Soc., 7:237- 242. Dill, L.M. A.H.G. Fraser & B.D. Roitberg. 1990. The economics of escape behaviour in the pea aphid, Acyrthosiphon pisum. Oecologia, 83:473- 478. Dinter, A. 1998. Intraguild predation between eri- gonid spiders, lacewing larvae, and carabids. J. Appl. EntomoL, 122:163-167. Doane, J.R, PD. Scotti, O.R.W. Sutherland & R.R Pottinger. 1985. Serological identification of wireworm and staphylinid predators of the Aus- tralian soldier fly {Inopus rubriceps) and wire- worm feeding on plant and animal food. Ento- moL Exp. Appl., 38:65-72. Duffield, S.J., PC. Jepson, S.D. Wratten & N.W Sotherton. 1996. Spatial changes in invertebrate predation rate in winter wheat following treat- ment with dimethoate. EntomoL Exp. Appl., 78: 9-17. Edgar, W.D. 1970. Prey and feeding behaviour of adult females of the wolf spider Pardosa amen- tata (Clerck). Netherlands J. Zool. 20:487-491. Fraser, A.M. 1982. The Role of Spiders in Deter- mining Cereal Aphid Numbers. Ph.D. Thesis, University of East Anglia, U.K. Ghanim, A.E., B. Freier & T. Wetzel. 1984. Zur Nahmngsaufnahme und Eiablage von Coccinella septempunctata L. bei unterschiedlichem Ange- bot von Aphiden der Art Macrosiphum avenae (Fabr.) und Rhopalosiphum padi (L.). Arch. Phy- topathol. U. Pflanzenschutz, Berlin., 20:117-125, Gravena, S. & H.E Da-Cuhna. 1991. Predation of cotton leafworm first instar larvae Alabama ar- gillacea (Lepidoptera: Noctuidae). Entomopha- ga, 36:481-491. Greenstone, M.H. 1996. Serological analysis of ar- thropod predation: past, present and future. Pp. 265-300. In The Ecology of Agricultural Pests. (W.O.C. Symondson & J.E. Liddell, eds.). Chap- man & Hall, London. Griffiths, C., J.B. Carter & J. Overend. 1984. Phaonia signata (Meigen)(Diptera: Muscidae) larvae predatory upon leather] ackets, Tipula paP udosa (Meigen) (Diptera: Tipulidae) larvae. En- tomoL Gaz., 35:53-55. Hawkins, B.A., H.V. Cornell & M.E. Hochberg. 1997. Predators, parasitoids and pathogens as mortality agents in phytophagous insect popula- tions. Ecology, 78:2145-2152. Haynes, D.L. & R Sisojevic. 1966. Predatory be- haviour of Philodromus rufus Walckenaer (Ara- neae: Thomisidae). Canadian EntomoL, 98:113- 133. Heong, K.L., S. Bleith & E.G. Rubia. 1990. Prey preference of the wolf spider, Pardosa pseu- doannulata (Boesenberg et Strand). Res. Popul. EcoL, 32:179-186. Heong, K.L., G.B. Aquino & A.T Barrion. 1992. Population dynamics of plant- and leafhoppers and their natural enemies in rice ecosystems in the Philippines. Crop Protection, 11:371-379. Honek, A. 1980. Population density of aphids at the time of settling and ovariole maturation in Coe- cinella septempunctata [CoL: Coccinellidae]. En- tomophaga, 25:427-430. Hough-Goldstein, J., J.A. Janis & C.D. Ellers. 1996. Release methods for Perillus bioculatus (E), a predator of the Colorado potato beetle. Biol. Cont., 6:114-122. Jeffries, M.J. & J.H. Lawton. 1984. Enemy free space and the structure of ecological communi- ties. Biol. J. Linn. Soc., 23:269-286. Jmhasly, R & W Nentwig. 1995. Habitat manage- ment in winter wheat and evaluation of subse- quent spider predation on insect pests. Acta Oec- oL, 16:389-403. Johnson, D.M., B.G. Akre & PH. Crowley. 1975. Modeling arthropod predation: wasteful killing by damselfly naiads. Ecology, 56:1081-1093. Kamal, N.Q. & V.A. Dyck. 1994. Regulations of whitebacked planthopper, Sogatella furcifera Horvath populations by predators. Bangladesh J. ZooL, 22:61-67. Kennedy, T.F. 1990. A Study of the Spider Fauna of Irish Cereal Fields, with Particular Reference to the Role of Linyphiidae as Aphid Predators. Ph.D. Thesis, National Univ. of Ireland. Kring, TJ., S.Y. Young & WC. Yearian. 1988. The striped lynx spider, Oxyopes salticus Hentz (Ar- aneae: Oxyopidae), as a vector of nuclear poly- hedrosis virus in Anticarsia gemmatalis Hubner 314 THE JOURNAL OF ARACHNOLOGY (Lepidoptera: Noctuidae). J. EntomoL Sci., 23: 394-398. Leathwick, D.M. & MJ. Winterboum. 1984. Ar- thropod predation on aphids in a lucerne crop. New Zealand EntomoL, 8:75-80. Lockley, T.C. & O.R Young. 1987. Prey of the striped lynx spider Oxyopes salticus (Araneae, Oxyopidae) on cotton in the Delta area of Mis- sissippi. J. ArachnoL, 14:395-397. Louda, S.M. 1982. Infloresecence spiders: a cost/ benefit analysis for the host plant Haplopappus venatus Blake (Asteraceae). Oecologia, 55:185- 191. Mansour, E 1987. Spiders in sprayed and un- sprayed cotton fields in Israel, their interactions with cotton pests and their importance as pred- ators of the Egyptian cotton leaf worm, Spodop- tera littoralis. Phytoparasitica, 15:31-41. Mansour, E & U. Heimbach. 1993. Evaluation of lycosid, micryphantid and linyphiid spiders as predators of Rhopalosiphum padi (Horn.: Aphi- didae) and their functional response to prey den- sity — laboratory experiments. Entomophaga, 38: 79-87. Mansour, E & 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, E, D. Rosen, A. Shulov & H.N. Plant. 1980. Evaluation of spiders as biological control agents of Spodoptera littoralis larvae on apple in Israel. Acta OecoL, 1:225-232. Mansour, E, D. Rosen & A. Shulov. 1981. Dis- turbing effect of a spider on larval aggregations of Spodoptera littoralis. EntomoL Exp. AppL, 29:234-237. Marc, P. & A. Canard. 1997. Maintaining spider biodiversity in agroecosystems as a tool in pest control. Agr. Ecosyst. & Env., 62:229-235. Mason, R.R. & TR. Torgersen. 1987. Dynamics of a nonoutbreak population of the Douglas-fir tus- sock moth, (Lepidoptera: Lymantriidae) in south- ern Oregon. Env. EntomoL, 16:1217-1227. Mason, R.R., TR. Torgersen, B.E. Wickman & H.G. Paul. 1983. Natural regulation of Douglas- fir tussock moth (Lepidoptera: Lymantriidae) population in the Sierra Nevada. Env. EntomoL, 12:587-594. Moran, M.D. & L.E. Hurd. 1994. Short-term re- sponses to elevated predator densities: noncom- petitive intraguild interactions and behavior. Oecologia, 98:269-273. Murdoch, W.W. 1990. The relevance of pest-ene- my models to biological control. Pp. 1-24. In Critical Issues in Biological Control (M. Mack- auer, L.E. Ehler & J. Roland, eds). Intercept, An- dover, U.K. Nakasuji, E, H. Yamanaka & K. Kiritani. 1973. The disturbing effect of micryphantid spiders on the larval aggregation of the tobacco cutworm, Spodoptera litura (Lepidoptera: Noctuidae). Kontyu, 41:220-227. Nentwig, W. 1975. Why are spiders highly selec- tive in what they catch? Proc. 6^*^ Int. Arach. Cong. 1974. Pp. 177-179. Nentwig, W. 1982. Epigeic spiders, their potential prey and competitors: relationship between size and frequency. Oecologia, 55:130-136. Nentwig, W. 1985. Spiders eat crickets artificially poisoned with KCN and change the composition of their digestive fluid. Naturwiss., 72:545-546. Nentwig, W. 1986. Non-webbuilding spiders: prey specialists or generalists? Oecologia, 69:571- 576. Nentwig, W. 1987. The prey of spiders. Pp. 249- 263. In Ecophysiology of Spiders (W. Nentwig, ed.). Springer- Verlag, Berlin. Nentwig, W. & C. Wissel. 1986. A comparison of prey lengths among spiders. Oecologia, 68:595- 600. Nyffeler, M. & W.L. Sterling. 1994. Comparison of the feeding niche of polyphagous insectivores (Araneae) in a Texas cotton plantation: Estimates of niche breadth and overlap. Environ. EntomoL, 23:1294-1303. Nyffeler, M., D.A. Dean & W.L. Sterling. 1987. Predation by green lynx spider, Peucetia viridans (Araneae: Oxyopidae), inhabiting cotton and woolly croton plants in East Texas. Environ. En- tomoL, 16:355-359. Nyffeler, M., W.L. Sterling & D.A. Dean. 1992. Impact of the striped lynx spider (Araneae: Ox- yopidae) and other natural enemies on the cotton fleahopper (Hemiptera: Miridae) in Texas cotton. Environ. EntomoL, 21:1178-1188. Nyffeler, M., W.L. Sterling & D.A. Dean. 1994a. How spiders make a living. Environ. EntomoL, 23:1357-1367. Nyffeler, M., W.L. Sterling & D.A. Dean. 1994b. Insectivorous activities of spiders in United States field crops. J. AppL EntomoL, 118:113- 128. Nyffeler, M., R.G. Breene, D.A. Dean & W. Ster- ling. 1990. Spiders as predators of arthropod eggs. J. AppL EntomoL, 109:490-501. Graze, M.J. & A. A. Grigarick. 1989. Biological control of aster leafhopper (Homoptera: Cicadel- lidae) and midges (Diptera: Chironomidae) by Pardosa ramulosa (Araneae: Lycosidae) in Cal- ifornia rice fields. J. Econ. EntomoL, 82:745- 749. Polls, G.A., C.A. Myers & R.D. Holt. 1989. The ecology and evolution of intraguild predation: Potential competitors that eat each other. Ann. Rev. Ecol. Syst., 20:297-330. Provencher, L. & D. Coderre. 1987. Functional re- sponses and switching of Tetragnatha laboriosa Hentz (Araneae: Tetragnathidae) and Clubiona SUNDERLAND— MECHANISMS UNDERLYING EFFECTS OF SPIDERS ON PESTS 315 pikei Gertsch (Araneae: Clubionidae) for the aphids Rhopalosiphum maidis (Fitch) and Rho- palosiphum padi (L.) (Homoptera: Aphididae). Env. EntomoL, 16:1305-1309. Provencher, L,, D. Coderre & C.D. Dondale. 1988. Spiders (Araneae) in com fields in Quebec. Ca- nadian EntomoL, 120:97-100. Provencher, L. & S.E. Riechert. 1994. Model and field test of prey control effects by spider assem- blages. Env. EntomoL, 23:1-17. Riechert, S.E. 1992. Spiders as representative “sit- and-wait” predators. Pp. 313-328. In Natural Enemies: The Population Biology of Predators, Parasites and Diseases (M.J. Crawley, ed.), Blackwell Scientific Publications, Oxford. Riechert, S.E. & T. Lockley. 1984. Spiders as bi- ological control agents. Ann. Rev. EntomoL, 29: 299-320. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in garden test systems. Ecology, 71:1441-1450. Riechert, S.E. & K. Lawrence. 1997. Test for pre- dation effects of single versus multiple species of generalist predators: spiders and their insect prey. EntomoL Exp. Applic., 84:147-155. Roland, J. & D.G. Embree. 1995. Biological con- trol of the winter moth. Ann. Rev. EntomoL, 40: 475-492. Rosenheim, J.A., H.K. Kaya, L.E. Ehler, J.J. Marois & B.A. Jaffee. 1995. Intraguild predation among biological-control agents: theory and evidence. Biol. Cont., 5:303-335. Samu, F. 1993. Wolf spider feeding strategies: op- timality of prey consumption in Pardosa horten- sis. Oecologia, 94:139-145. Samu, F. & Z. Biro. 1993. Functional response, multiple feeding and wasteful killing in a wolf spider (Araneae: Lycosidae). European J. Ento- moL, 90:471-476. Samu, E, K.D. Sunderland, C.J. Topping & J.S. Fenlon. 1996. A spider population in flux: se- lection and abandonment of artificial web-sites and the importance of intraspecific interactions in Lepthyphantes tenuis (Araneae: Linyphiidae) in wheat. Oecologia, 106:228-239. Sawada, H., A. Kusmayadi, S.W.G. Subroto & E. Suwardiwijaya. 1993. Comparative analysis of population characteristics of the brown planthop- per, Nilaparvata lugens Stal. between wet and dry rice cropping seasons in West Java, Indone- sia. Res. Popul. EcoL, 35:113-137. Sengonca, C. & W. Klein. 1988. Beutespektmm und Frassaktivitat der Apfelanlagen haufig vor- kommenden Kreuzspinne, Araniella opistogra- pha (Kulcz.) und der Laufspinne, Philodromus cespitum (Walck.) im Labor. Z. Angew. ZooL, 75:43-54. Service, M.W. 1973. Mortalities of the larvae of the Anopheles gambiae Giles complex and de- tection of predators by the precipitin test. Bull. EntomoL Res., 62:359-369. Settle, W.H., A. Ariawan, E.T. Astuti, W. Cahyana, A.L. Hakim, D. Hindayana, A.S. Lestari & Pa- jamingsih. 1996. Managing tropical rice pests through conservation of generalist natural ene- mies and alternative prey. Ecology, 77:1975- 1988. Smith, R.S. & WG. Wellington. 1983. The func- tional response of a juvenile orb- weaving spider. Proc. 9“^ Int. Congr. ArachnoL, Smithsonian Inst. Press, Panama, Pp. 275-279. Spiller, D.A, 1984. Seasonal reversal of competi- tive advantage between two spider species. Oec- ologia, 64:322-331. Sterling, W.L., A. Dean & N.M. Abd El-Salam. 1992. Economic benefits of spider (Araneae) and insect (Hemiptera: Miridae) predators of cotton fleahoppers. J. Econ. EntomoL, 85:52-57. Strauss, S.Y. 1991. Indirect effects in community ecology: their definition, study and importance. TREE, 6:206-210. Sunderland, K.D. 1988. Quantitative methods for detecting invertebrate predation ocurring in the field. Ann. AppL BioL, 112:201-224. Sunderland, K.D. 1996. Progress in quantifying predation using antibody techniques. Pp. 419- 455. In The Ecology of Agricultural Pests (WO.C. Symondson & J.E. Liddell, eds.). Chap- man & Hall, London. Sunderland, K.D., A.M. Fraser & A.F.G. Dixon. 1986. Field and laboratory studies on money spi- ders (Linyphiidae) as predators of cereal aphids. J. AppL EcoL, 23:433-447. Sunderland, K.D., R.J. Chambers & O.C.R. Carter. 1988. Potential interactions between varietal re- sistance and natural enemies in the control of ce- real aphids. Pp. 41-56. In Integrated Crop Pro- tection in Cereals (R. Cavalloro & K.D. Sunderland, eds.). A. A. Balkema, Rotterdam. Sunderland, K.D., T. Bilde, L.J.M.F Den Nijs, A. Dinter, U. Heimbach, J.A. Lys, W Powell & S. Toft. 1996. Reproduction of beneficial predators and parasitoids in agroecosystems in relation to habitat quality and food availability. Acta JutL, 71:117-153. Sunderland, K.D., J.A. Axelsen, K. Dromph, B. Freier, J.-L. Hemptinne, N.H. Holst, P.J.M. Mols, M.K. Petersen, W Powell, P Ruggle, H. Triltsch & L. Winder. 1997. Pest control by a commu- nity of natural enemies. Acta JutL, 72:271-326. Thompson, J.N. 1984. Insect diversity and the tro- phic structure of communities. Pp. 591-606. In Ecological Entomology. (C.B. Huffaker & R.L. Rabb, eds.). John Wiley, New York. Toft, S. 1995. Value of the aphid Rhopalosiphum padi as food for cereal spiders. J. AppL EcoL, 32:552-560. Toft, S. 1997. Acquired food aversion of a wolf 316 THE JOURNAL OF ARACHNOLOGY spider to three cereal aphids: intra- and interspe- cific effects. Entomophaga, 42:63-69. Wheeler, A.G. 1973. Studies on the arthropod fau- na of alfalfa. V. Spiders (Araneida). Canadian EntomoL, 105:425-432. Whitcomb, W.H. 1974. Natural populations of en- tomophagous arthropods and their effect on the agroecosystem. Pp. 150-169. In Proceedings of the Summer Institute on Biological Control of Plant Insects and Diseases. (EG. Maxwell & EA. Harris, eds.). Univ. Press of Mississippi, Jackson. Winder, L. 1990. Predation of the cereal aphid Si- tobion avenae by polyphagous predators on the ground. Ecol. EntomoL, 15:105-110. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge Univ. Press, 328 pp. Wyss, E., U. Niggli & W. Nentwig. 1995. The im- pact of spiders on aphid populations in a strip- managed apple orchard. J. Appl. Entomol., 119: 473-478. Yamanaka, H., F. Nakasuji & K. Kiritani. 1973. Life tables of the tobacco cutworm Spodoptera litura and the evaluation of effectiveness of nat- ural enemies. Japanese J. Appl. Entomol. ZooL, 16:205-214. Zhang, Z.Q. 1992. The natural enemies of Aphis gossypii (Horn., Aphididae) in China. I. Appl. Entomol., 114:251-262. Manuscript received 1 May 1998, revised 25 Sep- tember 1998. 1999. The Journal of Arachnology 27:317-324 PREY SELECTION OF SPIDERS IN THE FIELD Martin Nyffeler: Zoological Institute, Division of Ecology, University of Berne, Baltzerstr. 3, CH“3012 Berne, Switzerland ABSTRACT. In this article, an overview of the general feeding patterns of common agroecosystem spiders is presented. Five groups of web-weavers (Tetragnathidae, Araneidae, Theridiidae, Linyphiidae, Dictynidae) and five groups of hunters (small-sized Oxyopidae, large-sized Oxyopidae, Thomisidae, Sal- ticidae, Lycosidae) are analyzed comparatively (based on 40 prey analyses previously published by various European and US authors). Fewer than 10 insect orders, as well as the order Araneae, make up the bulk of the prey of these spiders. Web- weavers and hunters both basically feed on the same prey orders, but in different proportions. The observed differences reflect in part the very diverse range of life styles and foraging modes exhibited by the various spider groups and, to some extent, differences in prey availability. Web- weavers are almost strictly insectivorous (insects constituting > 99% of total prey). Hunters, however, exhibit a mixed strategy of insectivorous and araneophagic foraging patterns (insects constituting —75- 90% of total prey). Diet breadth computed with the Inverted Simpson Index was, on average, significantly higher in the hunting spiders than the web spiders. There seems to be a consistent trend of greater diet breadth of the hunters compared to the web- weavers in agroecosystems. Overall, spider individuals of small size (including large percentages of immatures) numerically dominate the faunas of field crops, and these feed primarily on tiny prey (< 4 mm in length). Information on how prey selection in the field operates is a prerequisite to a quantitative assessment of the spiders’ potential as biolog- ical control agents in agroecosystems. Prey se- lection has been defined by Hassell (1978) as follows: “Preference for a particular prey is normally measured in terms of the deviation of the proportion of that prey attacked from the proportion available in the environment.” Most authors who studied the prey of spiders failed to record the availability of potential prey in the environment probably due to tech- nical difficulties. Thus, corresponding data on the actual and potential prey are scarce; and, consequently, only a limited number of prey selection studies on spiders following Has- sell’s approach exist (e.g., Uetz et al. 1978). Another approach to searching for patterns of prey selection is to analyze a large set of data on the actual prey of different spider groups (with very differing life styles and for- aging modes) and to compare the degree to which utilization of the various prey taxa dif- fers. Numerous published field studies on the actual prey of spiders are available for such an investigation (see reviews by Nyffeler 1982; Nentwig 1987; Riechert & Harp 1987; Wise 1993; Nyffeler el al. 1994a, b). In the current investigation, five groups of web- weavers (Tetragnathidae, Araneidae, Theridi- idae, Linyphiidae, Dictynidae) and five groups of hunters (small-sized Oxyopidae [i.e., Oxy- opes salticus^ large-sized Oxyopidae [i.e., Peucetia viridans}, Thomisidae, Salticidae, Lycosidae), representing nine families, are an- alyzed comparatively. These selected groups are among the most common spider predators in agroecosystems (Nyffeler et al. 1994b) and, thus, are of particular interest from the point of view of biological control. Descriptions of the life styles and foraging modes of these 10 spider groups are given by Rypstra (1982), Nentwig (1987), Wise (1993), and Nyffeler et al. (1994a, b). METHODS For each of the 10 spider groups the relative taxonomic composition of the diets (mean ± SE of 4 different prey analyses) was assessed (Tables 2, 3). Overall, 40 different prey ana- lyses (based on observational data from 31 published studies [see Table 1]) have been processed. To determine relative feeding spe- cialization, the diet breadth B diversity of arthropod orders in the diet) was computed for each spider group by means of the Inverted Simpson Index (see Levins 1968; Colwell & Futuyma 1971) (Table 4). Diet breadth is in- versely related to ecological specialization 317 318 THE JOURNAL OF ARACHNOLOGY Table 1. — Field studies used for the assessment of the relative taxonomic composition of the diets of ten spider groups. Habitats: SO = soybean, CO = cotton, PE = peanuts, AA = alfalfa, WW = winter wheat, OA = oats, MA = maize, MM = mown meadow, VE = vegetables, NC = noncrop. Spider group Habitat Area Author(s) Tetragnathidae Tetragnatha laboriosa SO USA LeSar & Unzicker (1978) Tetragnatha laboriosa SO USA Culin & Yeargan (1982) Tetragnatha laboriosa CO USA Nyffeler et aL (1989) Tetragnatha Extensa WW Europe Nyffeler & Benz (1979) Araneidae Acanthepeira stellata CO USA Nyffeler et al. (1989) Argiope aurantia CO USA Nyffeler et al. (1987a) Neoscona arabesca CO USA Nyffeler et al. (1989) Neoscona arabesca SO USA Culin & Yeargan (1982) Theridiidae Latrodectus mactans CO USA Nyffeler et al. (1988a) Achaearanea riparia WW Europe Nyffeler & Benz (1988a) Theridion impressum WW Europe Nyffeler (1982) Theridion impressum OA Europe Nyffeler & Benz (1979) Linyphiidae various Erigoninae MA Europe Alderweireldt (1994) various Erigoninae WW Europe Sunderland et al. (1986) various Erigoninae WW Europe Nyffeler & Benz (1988b) various Erigoninae MM Europe Nyffeler (1982) Dictynidae Dictyna segregata CO USA Nyffeler et al. (1988b) Dictyna arundinacea WW Europe Heidger & Nentwig (1989) Dictyna arundinacea NC Europe Heidger & Nentwig (1986) Dictyna montana NC Africa Nentwig (1987) Oxyopidae (small-sized) Oxyopes salticus CO USA Nyffeler et al. (1987b) Oxyopes salticus CO USA Nyffeler et al. (1992a) Oxyopes salticus CO USA Lockley & Young (1987) Oxyopes salticus PE USA Agnew & Smith (1989) Oxyopidae (large-sized) Peucetia viridans CO USA Nyffeler et al. (1987c) Peucetia viridans CO USA Nyffeler et al. (1992a) Peucetia viridans NC USA Turner (1979) Peucetia viridans NC USA Randall (1982) Thomisidae Misumenops spp. PE USA Agnew & Smith (1989) Misumenops spp. CO, NC USA Dean et al. (1987) Xysticus emertoni NC USA Morse (1983) Xysticus spp. MM Europe Nyffeler & Breene (1990a) , Salticidae Phidippus audax CO, NC USA Dean et al. (1987) Phidippus audax CO, NC USA Young (1989) Phidippus audax VE USA Riechert & Bishop (1990) Phidippus johnsoni NC USA Jackson (1977) Lycosidae Pardosa ramulosa AA USA Yeargan (1975) Pardos a spp. PE USA Agnew & Smith (1989) Pardosa spp. WW Europe Nyffeler & Benz (1988c) Pardosa amentata NC Europe Hallander (1970) NYFFELER— PREY SELECTION IN THE FIELD 319 Table 2. — Relative taxonomic composition of the diets of various web- weavers [for each spider group a mean ± SE, based on 4 different prey analyses has been computed]. ' LeSar & Unzicker (1978); Nyffeler & Benz (1979); Culin & Yeargan (1982); Nyffeler et al. (1989). ^ Culin & Yeargan (1982); Nyffeler et al. (1987a); Nyffeler et al. (1989) [data for 2 species]. ^ Nyffeler (1982); Nyffeler & Benz (1979, 1988a); Nyffeler et al. (1988a). “^Nyffeler (1982); Sunderland et al. (1986); Nyffeler & Benz (1988b); Alderwei- reldt (1994). ^ Nentwig (1987); Nyffeler et al. (1988b); Heidger & Nentwig (1986, 1989). Diet item (in %) Tetragna- thidae' (Tetragnathd) Araneidae^ {Acan- thepeira, Argiope, Neoscona) Theridiidae^ (Latrodectus, Achaearanea, Theridion) Linyphiidae"^ (Erigoninae) Dictynidae^ (Dictyna) Overall mean Homoptera 51 ± 16 36 ± 7 26 ± 9 33 ± 8 21 -1- 13 33 ± 5 Diptera 40 ± 17 21 ± 6 15 ± 7 9 ± 2 64 -f- 15 30 ± 6 Hymenoptera 3 ± 1 7 ± 2 32 ± 17 2 ± 1 7 -+- 3 10 ± 4 Collembola 0 ± 0 0 ± 0 0 ± 0 48 ± 8 0 + 0 10 ± 5 Coleoptera 1 ± 1 24 ± 9 13 ± 4 <1 ± 0.2 1 -1- 1 8 ± 3 Heteroptera 5 ± 4 3 ± 1 1 ± 1 1 ± 0.5 <1 -1- 0 2 ± 1 Lepidoptera <1 ± 0.7 3 ± 1 1 ± 1 0 ± 0 0 -i- 0 <1 ± 0.4 Araneae 0 ± 0 <1 ± 0.2 <1 ± 0.5 <1 ± 0.2 <1 0.2 <1 ± 0.1 Others 0 ± 0 5 ± 4 11 ± 4 6 ± 2 7 2 6 ± 2 Total 100 ± 0 100 ± 0 100 ± 0 100 ± 0 100 -1- 0 100 ± 0 (Colwell & Futuyma 1971; Turner 1979). Thus, high .S-values are characteristic for ex- ceedingly polyphagous predators, whereas low values indicate a more specialised feeding behavior. [Here a specialist feeder is defined as one that exhibits narrow diet breadth in a particular environment.] RESULTS AND DISCUSSION Overall, fewer than 10 arthropod orders (Diptera, Homoptera, Hymenoptera, Heterop- tera, Collembola, Coleoptera, Lepidoptera, and Araneae) make up the bulk of the prey of common agroecosystem spiders all of which are polyphagous predators (generalists) (Ta- bles 2, 3). Dietary mixing seems to be advan- tageous by optimizing a balanced nutrient composition needed for survival and repro- duction (Greenstone 1979; Uetz et al. 1992; Toft 1995). The various spider groups feed ba- sically on the same orders, but in different proportions. The observed differences reflect. Table 3. — Relative taxonomic composition of the diets of various hunters [for each spider group a mean ± SE, based on 4 different prey analyses has been computed]. • Lockley & Young (1987); Agnew & Smith (1989); Nyffeler et al. (1987b; 1992a). ^ Turner (1979); Randall (1982); Nyffeler et al. (1987c, 1992a). 3 Morse (1983); Dean et al. (1987); Agnew & Smith (1989); Nyffeler & Breene (1990a). ^ Jackson (1977); Dean et al. (1987); Young (1989); Riechert & Bishop (1990). Hallander (1970); Yeargan (1975); Nyffeler & Benz (1988c); Agnew & Smith (1989). Thomisidae^ Diet item Oxyopidae* " Oxyopidae^ {Misumenops, Salticidae'^ Lycosidae^ Overall (in %) {Oxyopes) (Peucetia) Xysticus) (Phidippus) (Pardosa) mean Heteroptera 30 10 18 -+- 4 18 -1- 11 21 + 11 16 13 21 4 Diptera 14 -1- 3 13 5 28 -t- 8 17 -+- 6 21 7 19 + 3 Araneae 11 -1- 4 13 -t- 6 9 3 16 6 24 9 15 -1- 3 Hymenoptera 11 -1- 5 35 20 16 -h 6 5 5 3 1 14 -+- 3 Homoptera 18 4 1 -1- 0.5 2 -f- 1 14 -t- 3 17 -1- 5 10 -+- 2 Lepidoptera 8 6 9 -f- 2 16 7 10 -i- 4 3 2 9 -+- 2 Coleoptera <1 0.3 6 -f- 1 6 -H 2 13 + 7 3 -+- 2 6 2 Collembola 0 -h 0 0 -4- 0 <1 0.2 0 + 0 8 ■+- 6 2 1 Others 7 1 5 2 5 -f- 2 4 -+- 2 5 2 5 + 1 Total 100 0 100 -t- 0 100 -1- 0 100 -+- 0 100 0 100 0 320 THE JOURNAL OF ARACHNOLOGY in part, the diverse range of life styles and foraging modes exhibited by the various spi- der groups, and to some extent differences in prey availability (see Riechert & Luczak 1982; Nentwig 1987; Nyffeler et al. 1994b). Web-weavers are almost strictly insectivo- rous (insects constituting > 99% of total prey) (Table 2). Aggressive encounters among web- weavers occur quite frequently, but rarely re- sult in predation. In a web, the potential vic- tim gets advanced vibrational warning and can flee or be ready to repulse the attacker. During such encounters between web-weavers the in- ferior individual is usually chased away by its opponent (see Wise 1993). Under conditions of suitable food supply in the form of insects the web-weavers seem to minimize feeding on “dangerous prey” such as spiders. Hunters, however, exhibit a mixed strategy of insectiv- orous and araneophagic foraging patterns (in- sects constituting « 75-90% of total prey) (Ta- ble 3). Field populations of several species of hunters had been found to be in a state of undernourishment (see Nyffeler & Breene 1990b). Thus, araneophagy including canni- balism (as an additional feeding strategy to insectivory) may be crucial in sustaining the hunter populations during periods of food shortage (see Wise 1993). “Eating other spi- ders appears to be an opportunistic occur- rence, a larger or faster individual overpow- ering another in a chance encounter” (Jackson 1992). Based on the data presented in Tables 2 and 3, the diet breadth {B) for spiders was com- puted with the Inverted Simpson Index (Table 4). The highest value was approximately five times higher than the minimum (5 = 1.13 vs. 5.58), which indicates considerable between- species differences in diet breadth. Evidently the hunters exhibit on average a less special- ized feeding behavior (overall mean diet breadth = 4.20 ± 0.20) compared to the web- weavers (overall mean = 2.61 ± 0.22) (Table 4), the difference between the two overall means being statistically significant (Mann- Whitney U test; U, = 52.5; df = 20, 20; P < 0.002). The data in Table 3 are almost exclusively based on US sources (3 out of 20 references from Europe), whereas those in Table 2 are from both European and US sources (10 out of 20 references from Europe). The US studies are generally from more southern and warmer Table 4. — Diet breadth {B) of five groups each of web- weaving spiders and hunting spiders; higher values indicate a less specialized feeding behavior (same data used as in Tables 2, 3). Diet breadth B Spider group Mean ± SE Range Web- weavers: Tetragnathidae 1.87 0.40 1.24-2.96 Araneidae 3.42 -1- 0.28 2.86-4.19 Theridiidae 3.20 0.60 1.70-4.52 Linyphiidae 2.55 -H 0.30 1.85-3.20 Dictynidae 2.00 -H 0.42 1.13-3.00 Overall mean 2.61 -h 0.22 Hunters: Oxyopidae (Oxyopes) 4.42 -1- 0.58 2.76-5.44 Oxyopidae (Peucetia) 4.34 -+- 0.34 3.42-4.86 Salticidae 4.38 -h 0.33 3.45-4.89 Thomisidae 3.95 -h 0.44 3.09-5.17 Lycosidae 3.90 0.69 2.65-5.58 Overall mean 4.20 -h 0.20 latitudes than the European ones (so far, most studies on the natural diets of hunters in crops available in the literature are from the south- ern US). Furthermore, the majority of US studies were conducted in structurally com- plex crops such as cotton and soybean fields, whereas most European studies were from ce- real crops with a less complex (i.e., prevail- ingly vertical) vegetation structure. Differenc- es in geographic latitude as well as vegetation structure could influence the prey availabili- ties. Thus, the question arises whether the re- sult of a greater diet breadth of the hunters observed in this study (Table 4) eventually is due to biases in the data set (the web and hunting spiders being studied in different crops and continents). To rule out this possi- bility, hunters and web-weavers should be an- alysed under comparable conditions (i.e., in the same field with identical prey availabili- ties). Studies in which both hunters and web- weavers were evaluated in the same fields were published by Nyffeler (1982), Nyffeler & Sterling (1994), and Bardwell & Averill (1997). Based on these studies the diet breadth of web spiders and hunting spiders was as- sessed comparatively (Table 5). In Nyffeler’s (1982) study in winter wheat fields near Zu- rich, Switzerland, hunters (represented by Pardosa spp. wolf spiders) had a greater diet NYFFELER— PREY SELECTION IN THE FIELD 321 Table 5. — Diet breadth (B) of web- weaving spiders vs. hunting spiders in winter wheat, cotton, and cranberry, based on data from: ‘ Nyffeler (1982); ^ Nyffeler & Sterling (1994); ^ Nyffeler et al. (1992a); 4 Bardwell & Averill (1997). Crop Foraging strategy Spider species Diet breadth B WHEAT: Web-weavers Tetragnatha extensa^ 1.24 Theridion impressum^ 2.90 Erigoninae (pooled data)' 3.10 Achaearanea riparia^ 3.70 Hunters Pardosa spp. (pooled data)' 4.48 COTTON: Web-weavers Tetragnatha laboriosa^ 1.36 Latrodectus mactans^ 1.70 Dictyna segregata^ 2.37 Neoscona arabesca^ 2.86 Acanthepeira stellata^ 3.29 Hunters Oxyopes salticus^ 4.73 Oxyopes salticus^ 4.76 Peucetia viridans^-^ 4.86 CRANBERRY: Web-weavers (pooled data)"^ 3.17 Hunters (pooled data)"^ 4.69 breadth than the web-weavers (represented by orb weavers, sheet web-weavers, and tangle web-weavers) (Table 5). Likewise, in Texas cotton fields, the numerically dominant hunt- ers (Oxyopes salticus and Peucetia viridans) exhibited greater diet breadth than several species of web-weavers (Table 5) (see Nyf- feler et al. 1992a; Nyffeler & Sterling 1994). Furthermore, the data presented by Bardwell & Averill (1997) from cranberry bogs in Mas- sachusetts suggest that the hunting spiders ex- hibited greater diet breadth than the web- weavers (pooled data for all hunters vs. web-weavers) (Table 5). Thus, in agroecosys- tems there seems to be a consistent trend of greater diet breadth of hunters compared to web-weavers regardless of crop type or geo- graphic region investigated. How do we explain this difference? Web spiders are stationary predators that wait for food to come to them (i.e., ‘sit- and- wait’ strat- egy). The prime requirement for the ‘sit-and- wait’ strategy is a food that moves (Turnbull 1973). A large proportion of web spiders spin aerial webs, with which they filter the aerial plankton (see Kajak 1965; Chacon & Eber- hard 1980; Nentwig 1980). Others spin webs adapted to capture walking, crawling, or jumping prey (Turnbull 1973). Most web- weavers depend largely on relatively few prey groups available in high numbers in a partic- ular environment (see Bristowe 1941; Turn- bull 1960; Nyffeler & Benz 1979, Sunderland et al. 1986; Nentwig 1987; Alderweireldt 1994). In contrast, hunting spiders, by and large, seem to be less restricted in their diet (see Turnbull 1973). Representatives of vari- ous hunting spider families (e.g., Oxyopidae, Salticidae, Thomisidae, Lycosidae) have been reported to feed on both moving and motion- less prey, which is indicative of a more mobile foraging strategy (see Nyffeler et al. 1990; Jackson & Tarsitano 1993). It is quite possible that the greater diet breadth of the hunting spi- ders (Table 4) simply reflects their greater op- portunities to actively seek out suitable food due to their higher mobility (see Turnbull 1973). There is observational evidence that hunt- ing spiders can narrow their diet breadth sig- nificantly at times when a suitable prey type becomes locally superabundant relative to oth- er prey (see Kiritani et al. 1972; Dean et al. 1987; Nyffeler et al. 1992b, 1994b). Thus, the greater diet breadth observed in the hunters (Table 4) does not necessarily imply that they require a more diverse diet than the web- weavers. It may instead show that they have a better chance of finding suitable food than web-weavers in agroecosystems (Young & Edwards 1990). However, there are exceptions to the rule (Turner & Polls 1979). Several members of the hunter families Thomisidae, Salticidae, Clubionidae, Gnaphosidae and Zo- 322 THE JOURNAL OF ARACHNOLOGY dariidae are known to specialize on ants (see Nentwig 1986, 1987). Most spiders feed on prey that are small relative to their own size (prey length < spider length) (Wise 1993). Feeding experiments with a variety of spider species and a model prey (crickets) conducted in the laboratory re- vealed that the optimal prey length ranges from 50-80% of the spiders' own length (Nentwig 1987). Nentwig's laboratory data are fully supported by observations in the field (Hayes & Lockley 1990; Nyffeler et ah 1987b, c, 1992a). Overall, spider individuals of small size (including large percentages of immatures) numerically dominate the faunas of field crops, and these feed primarily on tiny prey organisms (< 4 mm in length) (LeSar & Unzicker 1978; Young & Edwards 1990; Nyf- feler et al. 1994a). ACKNOWLEDGMENTS I extend my gratitude to Matt Greenstone, Wolfgang Nentwig, Keith Sunderland, Soeren Toft, and an anonymous reviewer for their comments on earlier drafts of this paper. LITERATURE CITED Agnew, C.W. & J.W. Smith, Jr. 1989. Ecology of spiders (Araneae) in a peanut agroecosystem. En- viron. EntomoL, 18:30-42. Alder weireldt, M. 1994). Prey selection and prey capture strategies of linyphiid spiders in high- input agricultural fields. Bull. British ArachnoL Soc., 9:300-308. Bardwell, C.J. & A.L. Averill. 1997. Spiders and their prey in Massachusetts cranberry bogs. J. ArachnoL, 25:31-41. Bristowe, W.S. 1941. The Comity of Spiders. Ray Society, London. Chacon, P. & W.G. Eberhard. 1980. Factors af- fecting numbers and kinds of prey caught in ar- tificial spider webs, with considerations of how orb webs trap prey. Bull. British ArachnoL Soc., 5:29-38. Colwell, R.K. & D.J. Futuyma. 1971. On the mea- surement of niche breadth and overlap. Ecology, 52:567-576. Culin, J.D. & K.V. Yeargan. 1982. Feeding behav- ior and prey of Neoscona arabesca (Araneae: Araneidae) and Tetragnatha laboriosa (Araneae: Tetragnathidae) in soybean fields. Entomophaga, 27:417-424. Dean, D.A., W.L. Sterling, M. Nyffeler & R.G. Breene. 1987. Foraging by selected spider pred- ators on the cotton fleahopper and other prey. Southwest. EntomoL, 12:263-270. Greenstone, M.H. 1979. Spider feeding behaviour optimises dietary essential amino acid composi- tion. Nature, 282:501-503. Hallander, H. 1970. Prey, cannibalism and micro- habitat selection in the wolf spiders Pardosa che- lata (O.E. Muller) and P. pullata (Clerck). Oi- kos, 21:337-340. Hassell, M.R 1978. The Dynamics of Arthropod Predator-prey Systems. Monographs in Popula- tion Biology 13. Princeton Univ. Press, New Jer- sey. Hayes, J.L. & T.C. Lockley. 1990. Prey and noc- turnal activity of wolf spiders (Araneae: Lycosi- dae) in cotton fields in the Delta region of Mis- sissippi. Environ. EntomoL, 19:1512-1518. Heidger, C. & W. Nentwig. 1986. The prey of Dic- tyna arundinacea (Araneae: Dictynidae). Zool. Beitr. (NF), 29:185-192. Heidger, C. & W. Nentwig. 1989. Augmentation of beneficial arthropods by strip-management. 3. Artificial introduction of a spider species which preys on wheat pest insects. Entomophaga, 34: 511-522. Jackson, R.R. 1977. Prey of the jumping spider Phidippus johnsoni (Araneae: Salticidae). J. Ar- achnoL, 5:145-149. Jackson, R.R. 1992, Eight-legged tricksters (Spi- ders that specialize in catching other spiders). BioScience, 42:590-598. Jackson, R.R. & M.S. Tarsitano. 1993. Responses of jumping spiders to motionless prey. Bull. Brit- ish ArachnoL Soc., 9:105-109. Kajak, A. 1965. An analysis of food relations be- tween the spiders — Araneus cornutus Clerck and Araneus quadratus Clerck — and their prey in meadows. EkoL Polska (A), 13:717-764. Kiritani, K., S. Kawahara, T. Sasaba & F. Nakasuji. 1972. Quantitative evaluation of predation by spiders on the green rice leafhopper, Nephotettix cincticeps Uhler, by a sight-count method. Res. Popul. EcoL, 13:187-200. LeSar, C.D. & J.D. Unzicker. 1978. Life history, habits, and prey preferences of Tetragnatha la- boriosa (Araneae: Tetragnathidae). Environ. En- tomoL, 7:879-884. Levins, R. 1968. Evolution in changing environ- ments: Some theoretical explorations. Mono- graphs in Population Biology. 2. Princeton Univ, Press, New Jersey. Lockley, TC. & O.P. Young. 1987. Prey of the striped lynx spider, Oxyopes salticus (Araneae, Oxyopidae), on cotton in the Delta area of Mis- sissippi. J. ArachnoL, 14:395-397. 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. J. ArachnoL, 11:87-94. Nentwig, W. 1980. The selective prey of linyphiid- like spiders and of their space webs. Oecologia, 45:236-243. NYFFELER— PREY SELECTION IN THE FIELD 323 Nentwig, W. 1986. Non- webbuilding spiders: prey specialists or generalists? Oecologia, 69:571- 576. Nentwig, W. 1987. The prey of spiders. Pp. 249- 263. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin, New York. Nyffeler, M. 1982. Field studies on the ecological role of the spiders as insect predators in agroe- cosystems. Ph.D. Dissertation, Swiss Fed. Inst. TechnoL, Zurich, 174 pp. Nyffeler, M. & G. Benz. 1979. Studies on the eco- logical importance of spider populations for the vegetation of cereal and rape fields. J. Appl. En- tomoL, 87:348-376 (in German). Nyffeler, M. & G. Benz, 1988a. Prey analysis of the spider Achaearanea riparia (Blackw.) (Ara- neae, Theridiidae), a generalist predator in winter wheat fields. J. Appl. EntomoL, 106:425-431. Nyffeler, M. & G. Benz. 1988b. Prey and preda- tory importance of micryphantid spiders in win- ter wheat fields and hay meadows. J. Appl. En- tomol., 105:190-197. Nyffeler, M. & G. Benz. 1988c. Feeding ecology and predatory importance of wolf spiders {Par- dosa spp.) (Araneae, Lycosidae) in winter wheat fields. J. Appl. EntomoL, 106:123-134. Nyffeler, M. & R.G. Breene. 1990a. Spiders as- sociated with selected European hay meadows and the effects of habitat disturbance, with the predation ecology of the crab spiders, Xysticus spp. (Araneae: Thomisidae). J. Appl. EntomoL, 110:149-159. Nyffeler, M. & R.G. Breene. 1990b. Evidence of low daily food consumption by wolf spiders in meadowland and comparison with other cursorial hunters. J. Appl. EntomoL, 110:73-81. Nyffeler, M. & W.L. Sterling. 1994. Comparison of the feeding niche of polyphagous insectivores (Araneae) in a Texas cotton plantation: Estimates of niche breadth and overlap. Environ. EntomoL, 23:1294-1303. Nyffeler, M., R.G. Breene, D.A. Dean & W.L. Ster- ling, 1990. Spiders as predators of arthropod eggs. J. Appl. EntomoL, 109:490-501. Nyffeler, M., D.A. Dean & W.L. Sterling. 1987a. Feeding ecology of the orb-weaving spider Ar- giope aurantm (Araneae: Araneidae), in a cotton agroecosystem. Entomophaga, 32:367-375. Nyffeler, M., D.A. Dean & W.L. Sterling. 1987b. Evaluation of the importance of the striped lynx spider, Oxyopes salticus (Araneae: Oxyopidae), as a predator in Texas cotton. Environ, EntomoL, 16:1114-1123. Nyffeler, M., D.A. Dean & W.L. Sterling. 1987c. Predation by green lynx spider, Peucetia viridans (Araneae: Oxyopidae), inhabiting cotton and woolly croton plants in East Texas. Environ. En- tomoL, 16:355-359. Nyffeler, M., D.A. Dean & W.L. Sterling. 1988a. The southern black widow spider, Latrodectus mactans (Araneae, Theridiidae), as a predator of the red imported fire ant, Solenopsis invicta (Hy- menoptera, Formicidae), in Texas cotton fields. J. Appl. EntomoL, 106:52-57. Nyffeler, M., D.A. Dean & W.L. Sterling. 1988b. Prey records of the web-building spiders Dictyna segregata (Dictynidae), Theridion australe (Theridiidae), Tidarren haemorrhoidale (Theri- diidae), and Frontinella pyramitela (Linyphi- idae) in a cotton agroecosystem. Southwest. Nat., 33:215-218. Nyffeler, M., D.A. Dean & W.L. Sterling. 1989. Prey selection and predatory importance of orb- weaving spiders (Araneae: Araneidae, Ulobori- dae) in Texas cotton. Environ. EntomoL, 18:373- 380. Nyffeler, M., D.A. Dean & W.L. Sterling. 1992a. Diets, feeding specialization, and predatory role of two lynx spiders, Oxyopes salticus and Peu- cetia viridans (Araneae: Oxyopidae), in a Texas cotton agroecosystem. Environ. EntomoL, 21: 1457-1465. Nyffeler, M., W.L. Sterling & D.A. Dean. 1992b. Impact of the striped lynx spider (Araneae: Ox- yopidae) and other natural enemies on the cotton fleahopper Pseudatomoscelis seriatus (Hemip- tera: Miridae) in Texas cotton. Environ. Ento- moL, 21:1178-1188. Nyffeler, M., W.L. Sterling & D.A. Dean. 1994a. Insectivorous activities of spiders in LFnited States field crops. J. Appl. EntomoL, 118:113- 128. Nyffeler, M., W.L. Sterling & D.A. Dean. 1994b. How spiders make a living. Environ. EntomoL, 23:1357-1367. Randall, J.B. 1982. Prey records of the green lynx spider, Peucetia viridans (Hentz) (Araneae, Ox- yopidae). J. ArachnoL, 10:19-22. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in garden test systems. Ecology, 71:1441-1450. Riechert, S.E. & J.M. Harp. 1987. Nutritional ecol- ogy of spiders. Pp. 645-672. In Nutritional Ecol- ogy of Insects, Mites, and Spiders. (F. Slansky & J.G. Rodriguez, eds.). John Wiley, New York. Riechert, S.E. & J. Luczak. 1982. Spider foraging: behavioral responses to prey. Pp. 353-385. In Spider Communication: Mechanisms and Eco- logical Significance. (P.N. Witt & J.S. Rovner, eds.). Princeton Univ. Press, New Jersey. Rypstra, A.L. 1982. Building a better insect trap; an experimental investigation of prey capture in a variety of spider webs. Oecologia, 52:31-36. Sunderland, K.D., A.M. Fraser & A.F.G. Dixon. 1986. Distribution of linyphiid spiders in rela- tion to capture of prey in cereal fields. Pedo- biologia, 29:367-375. Toft, S. 1995. Value of the aphid Rhopalosiphum 324 THE JOURNAL OF ARACHNOLOGY padi as food for cereal spiders. J. AppL Ecol., 32:552-560. Turnbull, A.L. 1960. The prey of the spider Liny- phia triangularis (Clerck) (Araneae, Linyphi- idae). Canadian J. ZooL, 38:859-873. Turnbull, A.L. 1973. Ecology of the true spiders (Araneomorphae). Annu. Rev. EntomoL, 18: 305-348. Turner, M. 1979. Diet and feeding phenology of the green lynx spider, Peucetia viridans (Ara- neae: Oxyopidae). J. ArachnoL, 7:149-154. Turner, M. & G.A. Polis. 1979. Patterns of co- existence in a guild of raptorial spiders. J. Anim. EcoL, 48:509-520. Uetz, G.W., J. Bischoff & J. Raver. 1992. Survi- vorship of wolf spiders (Lycosidae) reared on different diets. J. ArachnoL, 20:207-211. Uetz, G.W., A.D. Johnson & D.W. Schemske. 1978. Web placement, web structure, and prey capture in orb- weaving spiders. Bull. British ArachnoL Soc., 4:141-148. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge Univ. Press, Cambridge, U.K. Yeargan, K.V. 1975. Prey and periodicity of Par- dosa ramulosa (McCook) in alfalfa. Environ. En- tomoL, 4:137-141. Young, O.P. 1989. Field observations of predation by Phidippus audax (Araneae: Salticidae) on ar- thropods associated with cotton. J. EntomoL Sci., 24:266-273. Young, O.P. & G.B. Edwards. 1990. Spiders in United States field crops and their potential effect on crop pests. J. ArachnoL, 18:1-27. Manuscript received 1 May 1998, revised 1 October 1998. 1999. The Journal of Arachnology 27:325-332 SCALE DEPENDENT DISPERSAL AND DISTRIBUTION PATTERNS OF SPIDERS IN AGRICULTURAL SYSTEMS: A REVIEW Ferenc Samu: Department of Zoology, Plant Protection Institute, Hungarian Academy of Sciences, Budapest, RO.B. 102, H-1525 Hungary Keith D. Sunderland: Horticulture Research International, Wellesbourne, Warwick CV35 9EF, U.K. Csaba Szinetar; Berzsenyi College, 4 Karolyi Caspar Sqr., Szombathely, H-9700 Hungary ABSTRACT. A conceptual framework is presented for the study of the factors affecting the distribution, dispersal and abundance of spiders in agricultural systems. It is useful to consider how factors operate at three levels of a spatial hierarchy, namely micro-habitat, habitat and landscape. The size and distribution of spider populations are determined by factors influencing survival, reproduction and dispersal. Modes of dispersal vary in terms of the efficiency of sampling new habitats and the level of risk. A literature survey of proximal factors (micro-climate, habitat structure, disturbance, prey availability, predation, and territoriality) affecting micro-habitat usage by spiders showed that the relative importance of these factors varied according to spider species. Spider abundance and diversity were found, in general, to be positively correlated with environmental diversity at different spatial scales. Within-field habitat diversifications were found to be more effective in increasing spider populations when interspersed throughout the crop (e.g., poly cultures and reduced tillage) than when spatially segregated (e.g., strip management). Two approaches (modeling and experimental) to studying the effects of landscape level phenomena on spider distribution and abundance are discussed. Manipulation of habitats at the edge of fields has not, in the main, resulted in increased spider density within fields. Opportunities were identified for increasing regional populations of spiders, and optimizing pest control, by management of the annual shift in the crop mosaic to maximize spider transfer rates from senescing crops to young crops. Spiders are ubiquitous predators in terres- trial ecosystems and they have a substantial presence in the agricultural landscape. Species distributions of spiders in various agricultural habitats provide strong evidence that those as- semblages are not just randomly collected from the local species pool (Topping & Lovei 1997). Spider diversity varies from being im- poverished under intensive culture (Nyffeler et al. 1994) to being, under favorable agricul- tural management, even greater than in natural habitats (Toft 1989). Agricultural systems are, however, characterized by a relatively small number of highly dispersive dominant agro- biont species, which thrive under disturbed conditions (Luczak 1979). The potential role of spiders in controlling pest populations in agriculture has already been reviewed (Riech- ert & Lockley 1984; Nyffeler & Benz 1987). Here we depart from the proposition that spi- ders are useful components of agroecosys- tems. Our aims here are to determine, from the literature, how spiders are distributed in agricultural systems, to discover the factors which bring about the observed distributions, and to assess whether it might be feasible to manipulate some of these factors to increase the abundance and effectiveness of spider populations as antagonists of pests. For the discussion of the different factors that influence spider distribution it may be useful to consider the distribution patterns in relation to three nested scales (Wiens 1989; Juhasz-Nagy 1992): the micro -habitat (e.g., a weedy patch within a field, bare ground be- tween rows of crop plants, or the air space between foliage), the habitat (e.g., the whole crop, the adjacent hedge, or an abandoned field), which comprises a collection of micro- habitats, and the landscape, which comprises 325 326 THE JOURNAL OF ARACHNOLOGY a collection of habitats. For each scale, we will review dispersal modes, specific factors and farming practices that are relevant to the given level. MICRO-HABITAT SCALE Selection of micro-habitat by individual spiders is likely to be in relation to a specific biological need or collection of needs or may reflect avoidance of some factor, such as in- terspecific encounters (Post & Riechert 1977). The spider may, for example, assess a micro- habitat as a potential web site, oviposition site, overwintering site or as a safe haven from predators during the inactive phase of a diel cycle. Harsh physical conditions are common in agricultural habitats and the spider may need to seek temporary refuge in a favorable micro-habitat in order to maintain its physio- logical integrity. From this it follows that in- dividuals of any spider population are har- bored by specific micro-habitats which might vary over time, according to the current needs of the individuals. The numbers of spiders to be found, at any instant of time, in each of the micro-habitats are determined by site selection (immigration into the micro-habitats), site-related rates of survival and reproduction (Sunderland & Top- ping 1993), and site abandonment (Gillespie & Caraco 1987) which, in turn, are deter- mined by various abiotic and biotic factors. Studies aimed at the examination of specific factors can be useful in devising agricultural practices which, typically acting at the habitat scale, might create an improved quality and distribution of micro-habitats for the enhance- ment of natural enemies. Abiotic factors.— Structural complexity is usually determined by the vegetation. It pro- vides support for webs and its degree of com- plexity has a bearing on the costs of explo- ration and web building (Zschokke 1996). The differential preference of spider species for various structural features can be demonstrat- ed by the strong relationship between struc- ture and the richness and density of spider as- semblages (Rypstra & Carter 1995). Manipulations of shrub and tree structure were found to influence spider species diver- sity and abundance (Hatley & MacMahon 1980). Bultman & Uetz (1982) separated the effects of forest litter as a nutritional base for spider prey from its role as a spatially com- plex substrate by the use of artificial leaves. Web-builders were more abundant in struc- tured artificial litter but hunting spiders pre- ferred prey-rich natural litter. Other studies employing artificial micro-habitat structures showed that salticids preferred open geome- tries whilst theridiids selected dense ones (Robinson 1981). Web-builders are not entire- ly reliant on vegetation, and irregular ground surface features have also been found to meet the requirements of some species. Depressions in the soil of arable fields, for example, are attractive web sites for some linyphiids, and species segregate in relation to the diameter of such depressions (Alderweireldt 1994; Samu et al. 1996). Preference for different structures is also a size related phenomenon. Field data by Gunnarsson (1992) suggested a relationship between spider mean size and vegetation fractal dimension within a habitat, while Riechert (1974) observed change in the structural needs with growth within one spe- cies. Although micro-climate and structure are often correlated (Cady 1984), manipulative experiments have been carried out in an at- tempt to separate the two factors. Such exper- iments demonstrated the strong effect of mi- cro-climate: web-site selection occurred in relation to humidity for araneid, tetragnathid and linyphiid spiders (Enders 1977; Gillespie 1987; Samu et al. 1996) and in relation to temperature for the funnel web spider. Age- lenopsis aperta (Riechert 1985). Web destruction is often a precursor to web-site abandonment (Hodge 1987). Some micro-habitats may be more prone than others to destructive forces that endanger the web, such as foraging animals and meteorological factors (Enders 1976). Agrobiont species can cope with disturbances by life-history strate- gies compatible with disturbance patterns (Toft 1989; Samu et al. 1998) and by their high dispersal power. Mobility gives spiders the flexibility to vacate a locally disturbed area and re-invade later. This contrasts with many of their intra-guild competitors, such as carabid and staphylinid beetles, which often have eggs, larvae and pupae in earthen cells highly vulnerable to mechanical disturbance (Sunderland et al. 1996). Biotic factors. — Cues from prey can (but do not always) play a role in micro-habitat selection and retention. The cues may be vi- SAMU ET AL.— DISPERSAL AND DISTRIBUTION OF SPIDERS IN AGRICULTURE 327 brational (Pasquet et al. 1994), olfactory (Riechert 1985) or visual (Persons & Uetz 1996). The relative importance of these cues varies with species, but Persons & Uetz (1997b) found visual cues to be the most sig- nificant for the wolf spider Schizocosa ocrea- ta. Many authors (Gillespie 1987; Weyman & Jepson 1994) have reported spiders (belonging to a range of families) to have shorter resi- dence times in micro-habitats where prey is scarce compared with sites where food is abundant. Riechert (1984) showed that if prey was experimentally supplemented in a poor quality web site of Agelenopsis aperta, the web owner made more effort to defend its web in territorial disputes. However, often weak or no relationship was observed between site quality and the tenacity of spiders to web- sites. In such cases spiders were demonstrated to follow a fixed probability random leaving strategy (Vollrath & Houston 1986; Persons & Uetz 1997a). An important factor affecting web-site tenacity and responsiveness to prey availability is the energetic cost of web con- struction, which varies between spider fami- lies (Janetos 1982). A micro-habitat chosen by a spider can act as a refuge from its own natural enemies. In predator exclusion experiments, Gunnarsson (1996) demonstrated that, in the presence of bird predators, the abundance and mean size of spiders were greater on spruce branches where a high needle density provided a refuge from predation. Intraguild predation and can- nibalism might affect micro-habitat selection as well. After spiders have reproduced, spatial separation of parents and offspring is often re- corded, and this may be a mechanism to re- duce cannibalism. Some species of adult ly- cosid, for example, move horizontally to occupy micro-habitats away from their off- spring (Edgar 1971; Greenstone 1983), whilst age-specific vertical migration in Clubiona phragmitis may serve the same purpose (Nen- twig 1982). Spider territoriality may be generally un- common (Wise 1993), but cases are known where agonistic interactions lead to spacing out of the spider population (Riechert 1981; Marshall 1995). In other instances, intra- and interspecific contests for webs and web-sites were observed frequently, without obvious in- fluence on spider aggregation in a micro-hab- itat. Intraspecific contests between adult fe- male Lepthyphantes tenuis for webs constructed in hollows in the earth of a wheat field were observed regularly (Samu et al. 1996), and they resulted in departure of more than 30% of web-owners from the web-site. An opposite trend can occur amongst less ag- gressive spiders, where a reduction in web construction costs can be obtained by attach- ing webs to each other, as was recorded for Zygiella x-notata (Leborgne & Pasquet 1987), and Hypochilus thorelli (Hodge & Storfer Is- ser 1997). Various degrees of communality are known in the Araneae, and this has con- comitant implications for micro-habitat usage (Rypstra 1986; Hodge & Uetz 1995). HABITAT SCALE Habitats are comprised of a number of mi- cro-habitats within a delimited area. Animals which move from one micro-habitat to anoth- er within a habitat can usually do so by low- risk dispersal modes. Agricultural fields with their artificial homogenous vegetation can typ- ically be viewed as habitats. Dispersal within habitats. — Micro-habitat relocations within a habitat are part of the for- aging strategy of actively hunting spiders (Ford 1978), but abandonment of web-sites can occur with high frequency in web spiders as well (Samu et al. 1996). To change micro- habitat, walking over the ground {cursorial dispersal) is relatively low-risk, as the spider can withdraw rapidly if it accidentally enters inimical territory. However, in extreme envi- ronments, even short distance movements, such as within-habitat web-relocation, can sig- nificantly increase mortality in a desert widow spider (Lubin et al. 1993). Increased daily movement rates had similar effects for a wolf spider species. In a tidal flood area Pardosa lapidicina migrates back and forth with the tides, and the mortality of the population in this habitat was higher than in a nearby salt marsh habitat where the animals moved less (Morse 1997). Cursorial movement could be ineffective for moving across or into large ar- eas of monoculture (Thomas et al. 1990). An alternative dispersal mode is that of “rig- ging.” This entails the spider climbing to the top of the vegetation, letting out strands of silk which fall onto the top of the canopy, then running along the silken line for a few meters and then repeating the process. It is likely to be relatively low-risk and is intermediate be- 328 THE JOURNAL OF ARACHNOLOGY tween aerial and cursorial dispersal in terms of habitat sampling rate. Farming practices. — Many farming oper- ations result in major habitat-scale disturbance for spiders. Harvesting, plowing, pesticide spraying and forest clearcutting are likely to affect most micro-habitats within a given hab- itat; and they are known to cause severe re- ductions in spider populations (Nyffeler et al. 1994; Thomas & Jepson 1997). Conversely, disturbances of intermediate strength and fre- quency may actually increase the diversity of a spider community (Johnson 1995). This ef- fect may operate by increasing the diversity of micro-habitats within a habitat. Another type of diversification might be achieved through the selection of appropriate farming practices which alter vegetation/structure in areas within habitats (fields) that are either spatially segregated (e.g., strip farming) or fully interspersed (e.g., intercropped polycul- tures, mulching). Interspersed diversification is frequently at- tained by planting multiple crop species in one field. This in a number of instances resulted in spider densities greater than those found in monocultures, and an associated suppression of pest species (Letourneau & Altieri 1983; Coderre et al. 1989; Coll & Bottrell 1995). Lycosid abundance, for instance, was in- creased; and com borer (Ostrinia furnacalis) decreased, in peanut intercropped with maize, compared with monocultures (Altieri 1994). Reduced-tillage systems often provide a di- versification of interspersed micro-habitats by engendering a rough or heterogeneous soil surface, plus structural complexity in the form of plant residues conserved from previous- year crops (House & Stinner 1983; Clark et al. 1993; Robertson et al. 1994). Other authors (Thornhill 1983; Alderweireldt 1994; Samu et al. 1996) have experimentally demonstrated that linyphiid density can be increased by cre- ating depressions in the surface of arable soils. Clover, as a living mulch (Altieri et al. 1985), and mulches experimentally applied to a gar- den system (Riechert & Bishop 1990) have been found to increase spider densities signif- icantly, probably by simultaneous effects on structure, micro-climate and prey availability (see above). Strip management contributes to micro- habitat diversification within crops, but the spatial separation of micro-habitats is greater than for interspersed treatments. In a Swiss orchard, the density of spiders and their webs on the apple trees was greater in plots where weeds had been planted in strips below the trees, than in weed-free control plots (Wyss et al. 1995). However, in many cases, spider density on and under crop plants is unaffected by strip management (Nentwig 1989; Riechert & Bishop 1990; Samu et al. 1997), perhaps because spiders aggregate in the favorable mi- cro-habitats (such as weed and flower strips) and do not disperse out onto the crop plants. LANDSCAPE SCALE The distribution of spiders is least studied at the landscape scale. This is mostly because it is extremely labor intensive to obtain even a coarse picture of spider distribution over a large area. To study the effect of landscape level phenomena on spider distribution two approaches are possible. One is to model land- scape scale distribution of spiders using in- formation on the biology of specific species and incorporate that into spatially explicit metapopulation models (Topping & Sunder- land 1994b; Halley et al. 1996; Topping, this volume). The other approach is to select smaller scale landscape fragments, a meaning- ful subset of landscape structures such as a field and its margin, to experimentally study the distribution and movement of spider pop- ulations. Dispersal. — For both above-mentioned re- search strategies, knowledge of the scale-de- pendent dispersal of spiders is essential. In fact. Topping’s (1997) simulation model ap- peared to be more sensitive to assumptions about dispersal than to field size or timing of agricultural operations. Spiders can vary dis- persal modes, such as cursorial movement, rigging or ballooning, by applying the most effective mode for the given scale of move- ment, although not all dispersal modes are available in each stage or taxon (Plagens 1986). The most efficient dispersal mode at the landscape scale is ballooning {aerial dis- persal), which provides the individual spider with the potential to sample different widely- separated habitats in a short period of time (Weyman 1993). If it lands in a safe habitat the spider may have the option to re-balloon immediately, sample the new habitat and re- balloon after a short period of time, or it may choose to stay. As far as is known, the desti- SAMU ET AL.— DISPERSAL AND DISTRIBUTION OF SPIDERS IN AGRICULTURE 329 nation of the aeronaut is determined purely by meteorological factors (Bishop 1990; Thomas 1996). It is, therefore, a high-risk activity; in- dividuals which land at unfavorable destina- tion areas will not be able to reproduce, thus these places act as a reproductive sink (Meijer 1977; Crawford et al. 1995). Farming practices. — At the landscape scale the effect of basic landscape structure (size and distribution of different habitat types, e.g., fields) and the cumulative impact of field-scale farming practices, (including their timing and distribution), are of primary interest. Root’s “enemies hypothesis” (Root 1973), which predicts generalist and specialist natural enemies to be more abundant in di- versified agricultural systems, was tested by spatially explicit models at the landscape scale. The maintenance of grass habitats that are not demolished by crop rotation (Topping & Sunderland 1994b) and the presence of set- aside fields (Topping & Sunderland 1994a) significantly increased the viability of the modeled Lepthyphantes tenuis metapopulation (Topping 1997). In a simulated linear land- scape the inclusion of small amounts of grass- land considerably increased overall spider population sizes (Halley et al. 1996). These models were also useful at pointing out the importance of the pattern and timing of de- structive agricultural practices. Crop rotation was generally detrimental, but the re-sched- uling of plowing could decrease this negative effect on spiders (Topping & Sunderland 1994b). Using the experimental approach many at- tempts have been made to increase the abun- dance of natural enemies in field habitats by manipulation of habitats at the edges of fields. In the majority of such studies (including soy- bean, cereals and orchards), increases in spi- der densities at the edges were not translated into increases in the fields themselves, and es- pecially in the centers of large fields (Altieri & Schmidt 1986; Alderweireldt 1989; Kemp & Barrett 1989; Dennis & Fry 1992; Kromp & Steinberger 1992; Altieri 1994; Vangsgaard 1996; Toth & Kiss 1997). Landscape frag- ments studied by transect sampling show this phenomenon as the ‘edge effect.’ Edges are often considered as distinct ecological sys- tems, ecotones, where the local fauna consist of species specific to the ecotone, and a mix- ture of the two neighboring faunas which overlap there. The width of the overlap was typically not found to be greater than a few meters for farmland and forest spiders (Bed- ford & Usher 1994; Downie et al. 1996). Be- tween wheat and various grassy areas the pen- etration of spiders into neighboring habitats was also limited (Duelli et al. 1990; Kajak & Lukasiewicz 1994). For spiders the absence of certain species from specific habitats is usu- ally due to the lack of habitat suitability, rath- er than a limitation of dispersal capacity. Bish- op & Riechert (1990) found that about half of the spider species found in a garden system were not found in nearby habitats, but arrived by long-distance migration. Larger scale landscape models suggest that the maintenance of habitat diversity and pre- serve areas are important for the subsistance of spider metapopulations. On the other hand, smaller scale experimental studies reveal the overriding importance of within field habitat quality. Perhaps these two phenomena could be combined by finding ways to provide “time- specific habitat diversity,” such that natural enemy populations build up in favor- able non-crop habitats and micro -habitats ini- tially, but are forced to transfer to crop plants (at a time when pests start their increase) by strategically timed destruction of the favorable non-crop habitats. It will be a challenge to de- velop practical management systems to achieve these goals. Large numbers of natural enemies do, however, emigrate from senesc- ing crops (Whitcomb & Bell 1964); and, since the various crop species in a landscape tend to senesce asynchronously, there is an oppor- tunity to manage the annual shift in the crop mosaic to maximize transfer rates of benefi- cials from senescing crops to young crops (Burel & Baudry 1995). ACKNOWLEDGMENTS KDS was funded by the U.K. Ministry of Agriculture, Fisheries and Food. FS was fund- ed by a research grant from the Ecological Center of the Hungarian Academy of Scienc- es, FS and CS by the OTKA grant No. 17691. FS was a Bolyai Fellow of the Hungarian Academy of Sciences. LITERATURE CITED Alderweireldt, M. 1989. An ecological analysis of the spider fauna (Araneae) occurring in maize fields, Italian ryegrass fields and their edge 330 THE JOURNAL OF ARACHNOLOGY zones, by means of different multivariate tech- niques. Agric. Ecosyst. Environ., 27:293-306. Alderweireldt, M. 1994. Habitat manipulations in- creasing spider densities in agroecosystems: pos- sibilities for biological control? J. Appl. Ento- moL, 118:10-16. Altieri, M. 1994. Biodiversity and Pest Manage- ment in Agroecosystems. Food Products Press, New York. 185 pp. Altieri, M.A. & L.L. Schmidt. 1986. The dynamics of colonizing arthropod communities at the in- terface of abandoned, organic and commercial apple orchards and adjacent woodland habitats. Agric. Ecosyst. Environ., 16:29-43. Altieri, M.A., R.C. Wilson & L.L. Schmidt. 1985. The effect of living mulches and weed cover on the dynamics of foliage and soil arthropod com- munities in three crop systems. Crop Prot., 4: 2010-2213. Bedford, S.E. & M.B. Usher. 1994. Distribution of arthropod species across the margins of farm woodlands. Agric. Ecosyst. Environ., 48:295- 305. Bishop, L. 1990. Meteorological aspects of spider ballooning. Env. Entomol., 19:1381-1387. Bishop, L. & S.E. Riechert. 1990. Spider coloni- zation of agroecosystems mode and source. Env. Entomol., 19:1738-1745. Bultman, T.L. & G.W. Uetz. 1982. Abundance and conununity structure of forest floor spiders fol- lowing litter manipulation. Oecologia, 55:34-41. Burel, F. & J. Baudry. 1995. Farming landscapes and insects. Pp. 203-220. In Ecology and Inte- grated Farming Systems. (D.M. Glen, M.P. Greaves & H.M. Anderson, eds.). Chichester, John Wiley. Cady, A.B. 1984. Microhabitat selection and lo- comotory activity of Schizocosa ocreata. J. Ar- achnoL, 11:297-307. Clark, M.S., J.M. Luna, N.D. Stone & R.R. Young- man. 1993. Habitat preferences of generalist predators in reduced- tillage com. J. Entomol. Sci., 28:404-416. Coderre, D., L. Provencher & J. Champagne. 1989. Effect of intercropping maize-beans on aphids and aphidophagous insects in com fields of southern Quebec, Canada. Acta Phytopath. En- tomol. Hungarica, 24:59-63. Coll, M. & D.G. Bottrell. 1995. Predator-prey as- sociation in mono- and dicultures: Effect of maize and bean vegetation. Agric, Ecosyst. En- viron., 54:115-125. Crawford, R., P. Sugg & J. Edvards. 1995. Spider arrival and primary establishment on terrain de- populated by volcanic emption at Mount St. Hel- ens, Washington. American Midi. Nat., 133:60- 75. Dennis, P. & G.L.A. Fry. 1992. Field margins: can they enhance natural enemy population densities and general arthropod diversity on farmland? Agric. Ecosyst. Environ., 40:95-115. Downie, I.S., J.C. Coulson & J.E.L. Butterfield. 1996. Distribution and dynamics of surface- dwelling spiders across a pasture-plantation eco- tone. Ecography, 19:29-40. Duelli, R, M. Studer, I. Marchand & S. Jakob. 1990. Population movements of arthropods be- tween natural and cultivated areas. Biol. Cons., 54:193-207. Edgar, W. D. 1971. The life-cycle, abundance and seasonal movement of the wolf spider, Lycosa (Pardosa) lugubris, in Central Scotland. J. Anim. EcoL, 40:303-322. Enders, F. 1976. Effects of prey capture, web de- stmction and habitat physiognomy on web-site tenacity of Argiope spiders (Araneidae). J. Ar- achnol., 3:75-82. Enders, F. 1977. Web-site selection by orb-web spi- ders, particularly Argiope aurantia Lucas. Anim. Behav., 25:694-712. Ford, M. J. 1978. Locomotory activity and the pre- dation strategy of the wolf spider Pardosa amen- tata (Clerck) (Lycosidae). Anim. Behav., 26:31- 35. Gillespie, R.G. 1987. The mechanisms of habitat selection in long-jawed orb- weaving spider Te- tragnatha elongata (Araneae, Tetragnathidae). J. Arachnol., 15:81-90. Gillespie, R.G. & T. Caraco. 1987. Risk sensitive foraging strategies of two spider populations. Ecology, 68:887-899. Greenstone, M. H. 1983. Site specificity and site tenacity in a wolf spider: a serological dietary analysis. Oecologia, 56:79-83. Gunnarsson, B. 1992. Fractal dimension of plants and body size distribution in spiders. Funct. EcoL, 6:636-641. Gunnarsson, B. 1996. Bird predation and vegeta- tion stmcture affecting spmce living arthropods in a temperate forest. J. Anim. EcoL, 65:389- 397. Halley, J.M., C.F.G. Thomas & PC. Jepson. 1996. A model for the spatial dynamics of linyphiid spiders in farmland. J. Appl. EcoL, 33:471-492. Hatley, C.L. & J.A. MacMahon. 1980. Spider com- munity organization: seasonal variation and the role of vegetation architecture. Environ, Ento- moL, 9:632-639. Hodge, M.A. 1987. Factors influencing web site residence time of the orb weaving spider, Mi~ crathena gracilis. Psyche, 94:363-371. Hodge, M.A. & A. Storfer Isser. 1997. Conspecific and heterospecific attraction: A mechanism of web-site selection leading to aggregation for- mation by web-building spiders. Ethology, 103: 815-826. Hodge, M.A. & G.W. Uetz. 1995. A comparision of agonistic behaviour of colonial web-building SAMU ET AL.— DISPERSAL AND DISTRIBUTION OF SPIDERS IN AGRICULTURE 331 spiders from desert and tropical habitats. Anim. Behav., 50:963-972. House, G. & B. Stinner. 1983. Arthropods in no- tillage soybean agroecosystems: community composition and ecosystem interactions. Envi- ron. Manage., 7:23-28. Janetos, A.C. 1982. Foraging tactics of two guilds of web- spinning spiders. Behav. Ecol. SociobioL, 10:19-27. Johnson, S. 1995. Spider communities in the can- opies of annually burned and long-term un- bumed Spartina pectinata wetlands. Environ. EntomoL, 24:832-834. Juhasz-Nagy, P. 1992. Scaling problems almost ev- erywhere; an introduction. Abstr. Bot., 16:1-5. Kajak, A. & J. Lukasiewicz. 1994. Do semi-natural patches enrich crop fields with predatory epigean arthropods. Agric. Ecosyst. Environ., 49:149- 161. Kemp, J.C. & G.W. Barrett. 1989. Spatial pattern- ing: impact of uncultivated corridors on arthro- pod populations within soybean agroecosystems. Ecology, 70:114-128. Kromp, B. & K.-H. Steinberger. 1992. Grassy field margins and arthropod diversity: a case study on ground beetles and spiders in eastern Austria (Coleoptera: Carabidae; Arachnida: Aranei, Op- iliones). Agric. Ecosyst. Environ., 40:71-93. Leborgne, R. & A. Pasquet. 1987. Influence of conspecific silk-structures on the choice of web- site by the spider Zygella x-notata. Rev. Arach- noL, 7:85-90. Letoumeau, D.K. & M.A. Altieri. 1983. Abun- dance patterns of a predator Orius tristicolor (Hemiptera: Anthocoridae) and its prey prey, Frankliniella occidentalis (Thysanoptera: Tripi- dae); habitat attraction in poly cultures versus monocultures. Environ. EntomoL, 122:1464- 1469. Lubin, Y., S. Ellner & M. Kotzman. 1993. Web relocation and habitat selection in a desert widow spider. Ecology, 74:1915-1928. Luczak, J. 1979. Spiders in agrocoenoses. Polish Ecol. Stud., 5:151-200. Marshall, S.D. 1995. Mechanisms of the formation of territorial aggregations of the burrowing wolf spider Geolycosa xera archboldi McCrone (Ar- aneae, Lycosidae). J. ArachnoL, 23:145-150. Meijer, J, 1977. The immigration of spiders (Ara- neida) into a new polder. Ecol. EntomoL, 2:81- 90. Morse, D. 1997. Distribution, movement, and ac- tivity patterns of an intertidal wolf spider Par- dosa lapidicina population (Araneae, Lycosidae). J. ArachnoL, 25:1-10. Nentwig, W. 1982. Zur Biologic der Schilfsacspin- ne Clubiona phragmitis (Arachnida, Araneae, Clubionidae). EntomoL Abh, Mus. Tierk. Dres- den, 45:183-193. Nentwig, W. 1989. Augmentation of beneficial ar- thropods by strip management. II. Successional strips in a winter wheat field. J. Plant Diseases Prot., 96:89-99. Nyffeler, M. & G. Benz. 1987. Spiders in natural pest control: a review. J. Appl. EntomoL, 103: 321-339. Nyffeler, M., W.L. Sterling & D.A. Dean. 1994. Insectivorous activities of spiders in United States field crops. J. Appl. EntomoL, 118:113- 128. Pasquet, A., A. Ridwan & R. Leborgne. 1994. Presence of potential prey affects web-building in an orb-weaving spider Zygiella x-notata. Anim. Behav., 47:477-480. Persons, M. & G. Uetz. 1997a. The effect of prey movement on attack behavior and patch resi- dence decision rules of wolf spiders (Araneae: Lycosidae). J. Insect Behav., 10:737-752. Persons, M. & G. Uetz. 1997b. Foraging patch res- idence time decisions in wolf spiders: Is perceiv- ing prey as important as catching prey. Ecosci- ence, 4:1-5. Persons, M.H. & G.W. Uetz. 1996. The influence of sensory information on patch residence time in wolf spiders (Araneae, Lycosidae). Anim. Be- hav., 51:1285-1294. Plagens, M.J. 1986. Aerial dispersal of spiders in a Florida cornfield ecosystem. Environ. Ento- moL, 15:1225-1233. Post, WM. & S.E. Riechert. 1977. Initial investi- gation into the structure of spider communities. J. Anim. Ecol., 46:729-749. Riechert, S.E. 1974. The pattern of local web dis- tribution in a desert spider: mechanisms and sea- sonal variation. J. Anim. Ecol., 43:733-746. Riechert, S.E. 1981. The consequences of being territorial: spiders, a case study. American Nat., 117:871-892. Riechert, S.E. 1984. Games spiders play. Ill: cues underlying context-associated changes in agonis- tic behaviour. Anim. Behav., 32:1-15. Riechert, S.E. 1985. Decisions in multiple goal contexts: habitat selection of the spider, Agelen- opsis aperta (Gertsch). Z. Tierpsychol, 70:53- 69. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in garden test systems. Ecology, 71:1441-1450. Riechert, S.E. & T. Lockley. 1984. Spiders as bi- ological control agents. Ann. Rev. EntomoL, 29: 299-320. Robertson, L.N., B.A. Kettle & G.B. Simpson. 1994. The influence of tillage practices on soil macrofauna in a semi-arid agroecosystem in northeastern Australia. Agric. Ecosyst. Environ., 48:149-156. Robinson, J.V. 1981. The effect of architectural 332 THE JOURNAL OF ARACHNOLOGY variation in habitat on a spider community: an experimental field study. Ecology, 62:73-80. Root, R.B. 1973. Organization of plant-arthropod association in simple and diverse habitats: the fauna of collards {Brassica oleraced). Ecol. Monogr., 43:95-124. Rypstra, A.L. 1986. High prey abundance and re- duction in cannibalism: the first step to social spiders (Arachnida). J. Arachnol., 14:193-200. Rypstra, A.L. & RE. Carter. 1995. The web-spider community of soybean agroecosystems in south- western Ohio. J. Arachnol., 23:135. Samu, E, J. Nemeth, F. Toth, E. Szita, B. Kiss & C. Szinetar. 1998. Are two cohorts responsible for bimodal life history pattern in the wolf spider Pardosa agrestis in Hungary? Proc. 17th Euro- pean Coll. Arachnol., Edinburgh. Pp. 215-221. Samu, E, V. Racz, C. Erdelyi & K. Balazs. 1997. Spiders of the foliage and herbaceous layer of an IPM orchard in Kecskemet-Szarkas, Hungary. Biol. Agric. Horticult., 15:131-140. Samu, E, K.D. Sunderland, C.J. Topping & J.S. Fenlon. 1996. A spider population in flux: se- lection and abandonment of artificial web- sites and the importance of intraspecific interactions in Lepthyphantes tenuis (Araneae: Linyphiidae) in wheat. Oecologia, 106:228-239. Sunderland, K.D., T. Bilde, L.J.M.F. Den Nijs, A. Dinter, U. Heimbach, J.A. Lys, W. Powell & S. Toft. 1996. Reproduction of beneficial predators and parasitoids in agroecosystems in relation to habitat quality and food availability. Acta Jutlan- dica, 71:117-153. Sunderland, K.D. & C.J. Topping. 1993. The spa- tial dynamics of linyphiid spiders in winter wheat. Mem. Queensland Mus., 33:639-644. Thomas, C. 1996. Modelling aerial dispersal of lin- yphiid spiders. Asp. Appl. Biol., 46:217-222. Thomas, C.F.G., E.H.A. Hoi & J.W. Everts. 1990. Modelling the diffusion component of dispersal during recovery of a population of linyphiid spi- ders from exposure to an insecticide. Funct. Ecol., 4:357-368. Thomas, C.F.G. & PC. Jepson. 1997. Field-scale effects of farming practices on linyphiid spider populations in grass and cereals. Entomol. Exp. AppL, 84:59-69. Thornhill, W.A. 1983. The distribution and prob- able importance of linyphiid spiders living on the soil surface of sugar-beet fields. Bull. British Ar- achnol. Soc., 6:127-136. Toft, S. 1989. Aspects of the ground-living spider fauna of two barley fields in Denmark: species richness and phenological synchronization. En- tomol. Meddr., 57:157-168. Topping, C.J. 1997. Predicting the effect of land- scape heterogeneity on the distribution of spiders in agroecosystems using a population dynamics driven landscape-scale simulation model. Biol. Agricult. Horticult., 15:325. Topping, C.J. 1999. An individual-based model for dispersive spiders in agro/ecosystems: simula- tions on the effects of landscape structure. In Spiders in Agroecosystems: Ecological Processes and Biological Control. J. Arachnol., 27(1):378- 386. Topping, C.J. & G.L.Lovei. 1997. Spider density and diversity in relation to disturbance in agro- ecosystems in New Zealand, with comparison to England. New Zealand J. ZooL, 21:121-128. Topping, C.J. & K.D. Sunderland. 1994a. The po- tential influence of set-aside on populations of Lepthyphantes tenuis (Araneae: Linyphiidae) in the agroecosystems. Asp. Appl. Biol., 40:225- 228. Topping, C.J. & K.D. Sunderland. 1994b. A spatial population dynamics model for Lepthyphantes tenuis (Araneae: Linyphiidae) with some simu- lations of the spatial and temporal effects of farming operations and land-use. Agric. Ecosyst. Environ., 48:203-217. Toth, F. & J. Kiss. 1997. Occurrence of Pardosa (Araneae, Lycosidae) species in winter wheat and in the field margin. Proc. 16th Europ. Coll. Arachnol., Siedlce. Pp. 309-315. Vangsgaard, C. 1996. Spatial distribution and dis- persal of spiders in a Danish barley field. Rev. Suisse Zook, hors serie:67 1-682. Vollrath, F. & A. Houston. 1986. Previous experi- ence and site tenancity in the orb spider Nephila (Araneae, Araneidae). Oecologia, 70:305-308. Weyman, G. S. 1993. A review of the possible causative factors and significance of ballooning in spiders. Ethol. Ecol. EvoL, 5:279-291. Weyman, G.S. & P. Jepson. 1994. The effect of food supply on the colonisation of barley by ae- rially dispersing spiders (Araneae). Oecologia, 100:386-390. Whitcomb, WH. & K. Bell. 1964. Predaceous in- sects, spiders, and mites of Arkansas cotton fields. Arkansas Agric. Exp. Stn. Bull., 690:1- 83. Wiens, J.A. 1989. Spatial scaling in ecology. Funct. Ecol., 3:385-397. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge University Press, Cambridge, U.K. 328 pp. Wyss, E., U. Niggli & W. Nentwig. 1995. The im- pact of spiders on aphid populations in a strip- managed apple orchard. J. Appl. Entomol., 119: 473-478. Zschokke, S. 1996. Early stages of web construc- tion. Rev. Suisse Zook, hor serie:709-720. Manuscript received I May 1998, revised 24 Sep- tember 1998. 1999= The Journal of Arachnology 27:333-342 SPIDER PREDATION: HOW AND WHY WE STUDY IT Matthew H. Greenstone: USDA-ARS Plant Sciences and Water Conservation Laboratory, 1301 N. Western Street, Stillwater, Oklahoma 74075 USA ABSTRACT. Predation is of great ecological, evolutionary and behavioral interest. For our present purposes the primary reason for studying it is to determine the role of spiders in suppressing pest popu- lations. Research approaches have included laboratory studies of preference, feeding rate, and fitness; direct observation of predation events or accumulations of prey carcasses; gut analysis; and field experi- ments. Laboratory studies provide some uniquely useful kinds of information but cannot give reliable indications of the “biological control potential” of spiders against a given pest. Direct observation can be powerful; it has provided the best data on dietary range and predation rates in the field. Gut analytical methods include the use of radionuclides, electrophoresis, chromatography and serology. Serological tech- niques are preferred: antibodies can be made specific down to the level of prey stage or instar, and assays are simple, sensitive, and reliable. They can determine the relative importance of different predator species, and may be the most efficient methods to document predation on eggs. Problems in quantitation remain. Field experiments have demonstrated unequivocally that spiders can effectively reduce pest populations and the crop damage they cause. Spiders are ubiquitous in terrestrial ecosys- tems and abundant in both natural and agri- cultural habitats (Dondale 1970; Turnbull 1973; Nyffeler & Benz 1987). They also have a suite of adaptations that enable them to wait out periods of low prey abundance rather than dispersing like some other groups of arthro- pod predators (Ford 1977; Greenstone & Ben- nett 1980). It has therefore been assumed that spiders play a major role in suppressing insect pest populations (Riechert & Lockley 1984; Young & Edwards 1990). However their small size, cryptic habit, and mode of feeding have made it difficult to determine whether this is so (Kiritani & Dempster 1973; Stuart & Greenstone 1990). HOW PREDATION IS STUDIED Predation data can be obtained by four means: laboratory feeding studies, direct ob- servation in the field, gut analysis, and exper- imental field manipulations. Laboratory studies. — There is a large lit- erature in agroecosystem entomology and ar- achnology describing feeding trials of individ- ual predators confined with prey in small containers in order to determine their “pred- atory potential.” For spiders this approach is problematic since most are oligophagous or polyphagous (Nentwig 1986), and neither choose nor survive and reproduce well on sin- gle-species diets (Miyashita 1968; Van Dyke & Lowrie 1975; Hydhom 1976; Greenstone 1979; Lowrie 1987; Uetz et al. 1992; Toft 1996). Another difficulty is our lack of knowl- edge of the role of environmental variables in predator-prey behavioral interactions. Web spider feeding trials are least apt to be compromised, because many of the critical en- vironmental and behavioral elements are em- bodied in the web itself. But hunting spider feeding trials are generally performed in sim- ple open arenas — rarely made more realistic by provisioning plant bouquets (Lingren et al. 1968) — and the spiders are usually starved for a day or even a week (Young 1989a; Punzo 1991; Sadana & Kumari 1991) to increase the likelihood of a result. Since a starved spider in a small, featureless arena is apt to attack any but the most unsuitable (noxious, ven- omous, too large or too well armored) arthro- pod placed before it, it is hardly surprising that such spiders usually feed, and sometimes consume large numbers of prey. However, lengthy starvation inflates feeding rates (Toft 1996); furthermore, starvation causes reduc- tions in basal metabolic rate (Anderson 1974), which could change feeding latency or oth- erwise distort predatory behavior. Since spi- ders in the field usually consume about one appropriate- sized insect per day (Edgar 1969, 1970; Schaefer 1974; Morse 1979; Nyffeler 1982; Nyffeler & Benz 1988a, b; Nyffeler et 333 334 THE JOURNAL OF ARACHNOLOGY Table 1. — Studies in which direct observation was used to determine the spectrum of spider species attacking a pest, pest complex or biological control agent. Prey Reference Acari Metatetranychus ulmi (Koch) & Bryobia praetiosa Koch Chant 1956 Lepidoptera Helicoverpa zea (Boddie) & Heliothis virescens (Fabricius) Quaintance & Braes 1905 Fletcher & Thomas 1943 Whitcomb & Bell 1964 Whitcomb et al. 1963 Whitcomb 1967 Anticarsia gemmatalis Hiibner Elvin et al. 1983 Godfrey et al. 1989 Gregory et al. 1989 Hyphantria cunea (Drary) Whitcomb & Tadic 1963 Coleophora parthenica Meyrick Nuessly & Goeden 1983 Coleoptera Diaprepes abbreviatus (L.) Richman et al. 1983a,b Ips and Dendroctonus spp. Jennings & Pase 1975, 1986 Heteroptera Pseudatomoscelis seriatus (Reuter) Dean et al. 1987 al, 1987a, 1992a), starved spiders may not be- have normally (but see Bilde & Toft 1998). These same objections apply when more than one prey type is offered simultaneously to de- termine preferences (e.g., Provencher & Cod- erre 1987; Gillebeau & All 1989). Also, ap- parent preferences may be perversely misleading because a less preferred, even pat- ently unpalatable prey species may, in com- bination with others, provide greater fitness than a pure diet of a preferred species (Toft 1995, 1996). Space does not permit a discussion of the numerous laboratory studies to generate spi- der functional response curves, but the same objections apply. The non-congruence of field and laboratory functional response data for one intensively studied insect predator (O'Neill 1997) should give us pause in con- templating the initiation of such lab studies. Other feeding questions are well suited to laboratory study. Sunderland et ah (1986) de- termined escape rates of aphids from occupied linyphiid webs under varying conditions of falling frequency, aphid stage and spider sa- tiation, necessary for converting field web density estimates into predation potential. Ed- gar (1969, 1970), Kiiitani et ah (1972), Nyf- feler & Benz (1988b) and Nyffeler et al. (1987a, b) determined the time during which prey items were carried and fed upon by spi- ders, a parameter needed to convert field ob- servations on feeding to predation rates (see also below). A general case for which labo- ratory studies are defensible is where one needs to know about preferences of highly ste- nophagous predators (e.g., Morse 1984; Li & Jackson 1996). Direct observation*— Our most extensive data on spider prey spectra, prey preferences and predation rates are derived from extensive field observations of spiders feeding and from the identification of prey carcasses taken from spiders’ webs. Prey spectrum: These studies focus on ei- ther the spiders attacking a particular pest or pest complex, or the prey spectrum of partic- ular guilds of spiders. Table 1 lists studies of the first kind. In a paper on bollworm preda- tion, Whitcomb (1967) pioneered the place- ment of eggs or larvae at regularly spaced sta- tions to facilitate data collection, a method later used for studies of the velvetbean cater- pillar (Elvin et al. 1983; Godfrey et al. 1989) and sugarcane rootstalk borer (Richman et al. 1983a, b). Whitcomb & Tadic (1963) surveyed spider predators of the fall webworm, a task facilitated by the arresting power of the webs GREENSTONE— HOW AND WHY WE STUDY PREDATION 335 Table 2. — Rates at which field observations of feeding rate by hunting spiders have been collected by human observers. *Best estimate if investigators did not record exact number of hours. **Events/person- hour; in all cases only one author made observations (see Acknowledgments). Spider species Habitat Events Hours* Rate** Reference Phidippus audax cotton 58 10 5.80 Young 1989c Oxyopes salticus cotton 48 11.25 4.27 Lockley & Young 1987 Peucetia viridans woolly croton 68 25.5 2.67 Nyffeler et al. 1987b Lycosa anteleucana cotton 147 91.4 1.61 Hayes & Lockley 1990 Dolomedes triton ponds 625 -400 1.56 Zimmermann & Spence 1989 Pardosa spp. wheat 106 104.5 1.01 Nyffeler & Benz 1988b Oxyopes salticus cotton 64 85 0.75 Nyffeler et al. 1987a Phidippus audax woolly croton 19 25.5 0.75 Dean et al. 1987 Pardosa ramulosa salt marsh 32 ~50 0.64 Greenstone 1976 Oxyopes salticus cotton 63 108 0.58 Nyffeler et al. 1992b Misumenops celer woolly croton 11 25.5 0.43 Dean et al. 1987 Metaphiddipus galathea woolly croton 9 25.5 0.35 Dean et al. 1987 Peucetia viridans cotton 31 108 0.29 Nyffeler et al. 1992b Peucetia viridans cotton 25 85 0.29 Nyffeler et al. 1987b Misumena calycina old field 16 79.3 0.20 Morse 1979 Pardosa milvina cotton 14 91.4 0.15 Hayes & Lockley 1990 Pisaurina mira cotton 12 -300 0.04 Young 1989b Phidippus johnsoni various 33 -3,000 0.01 Jackson 1977 of the prey, which collected the spiders for the investigators' perusal. The guild-centered prey spectrum literature comprises thousands of person-hours of direct observation (see Nyffeler 1999, this volume, for a thorough review and analysis). Hunting spiders pose the biggest challenge because they are less easily found and do not leave the carcasses of their prey where they can be iden- tified and counted. Thirteen studies for which one can estimate the rate of discovery of hunt- ing spider predation events by a human ob- server are summarized in Table 2. They reveal a surprising range, from about 0.01 to almost 6 events/person-hour of observation; rates for one species in cotton (Oxyopes salticus) var- ied seven-fold. These data demonstrate that direct observation can sometimes be an effi- cient way to learn about the prey spectrum of hunting spiders, and may enable the investi- gator to assess the effort likely to be involved in such an undertaking. Predation rates: Predation rates for web spiders can be obtained directly from web densities and counts of prey carcasses in webs or in sticky traps, provided prey escape prob- abilities are determined; and an advantage of using traps is that they can obviate the need to work at night (Sunderland et al. 1986). De- termining predation rates of hunting spiders by direct observation requires ingenuity; an approach was outlined by Edgar (1969, 1970) and formalized by Nyffeler & Benz (1988b). Their formula contains an estimate of the hours per day spent hunting. Such an estimate is implicit in all predation rate estimates, and must be stated explicitly if the investigator limits the time during which data are taken (Jmhasly & Nentwig 1995). Published preda- tion rates for web and hunting spiders, vari- ously expressed, are presented in Table 3. Rates for individual spider species and spe- cies complexes suggest relatively low propor- tions of pest populations being destroyed, but one must remember that spiders constitute an assemblage of species that may, in aggregate, exert effective control (Riechert & Bishop 1990; Riechert & Lawrence 1997). Further- more, spiders kill many more insects than they consume (see Sunderland 1999, this volume). Finally, in conjunction with parasitoids, path- ogens, and other polyphagous predators, spi- ders may tip the balance in biological control. Intraguild predation: Prey spectrum studies reveal that some spiders consume large num- bers of beneficial arthropods, including other spiders and parasitic and predatory Hymenop- tera and Diptera (see Hodge 1999, this vol- ume); most notorious is the green lynx spider, Peucetia viridans (Turner 1979; Randall 1982; 336 THE JOURNAL OF ARACHNOLOGY Table 3. — Spider predation rates and prey population impacts derived by direct observation. *Data collection restricted to hours of daylight. **Based on calculation from raw data in Edgar 1969. Species or Complex Rate Impact Reference Web Spiders All foliage web 0.2-1. 2 X 10^ insects/ha/year Nyffeler (1982) spiders Araneidae only 0.2-1. 2 kg insects/ha/year 38 insects/mWay 150 fresh kg insects/ha/year Araneus spp. 12 insects/mVday Kajak 1965 Linyphiidae, Araneidae & 3. 5-5. 8 prey /mV9 h day* Jmhasly «fe Nentwig 1995 Tetragnathidae Linyphiidae only 1. 5-1.7 aphids/m2/9 h* 4% of aphid popu- lation Linyphiidae 0.023-31.2 aphids/mVday 105.6 aphids/mVseason Sunderland et al. 1986 Micryphantidae 42 insects/mVday 20 aphids/mVday 2% of aphid popu- lation Nyffeler & Benz 1988a Hunting Spiders Pardosa spp. — 1.3 insect/day 2 aphids/m^/week Nyffeler & Benz 1988b Pardosa lugubris 0.8 insect/day** Edgar 1969 Pardosa amentata 1.17 insect/day Edgar 1970 Peucetia viridans 0.25-0.5 insects/day Nyffeler et al. 1987b Oxyopes salticus 120,000 insects/ha/week 4.5% of available prey Nyffeler et al. 1987a Oxyopes salticus 0.9 insects/day 15-18% avail, flea- hoppers Nyffeler et al. 1992a Phidippus audax 5% of available prey Young 1989c Nyffeler et al. 1987b). Randall (1982) asserted that the green lynx is “counterproductive” as a biological control agent, but the only sure way to determine this would be to study the agroecosystem in its presence and absence. Louda (1982) performed just such a study of predation by P. viridans in a natural system and found that its net effect was beneficial to plants. Gut analysis. — Identifying and quantifying prey remains in the gut are the first steps in determining spider predation rates (Sunder- land 1996). However, because spiders are liq- uid feeders and the remains of several meals may be found concurrently, this presents for- midable technical problems (Stuart & Green- stone 1990). Radionuclides: Breene et al. (1988) showed that when mosquito larvae irradiated with were made available to three amphibious spi- der species in simulated ponds, spider feeding could be documented by acquired radioactiv- ity. Similar approaches were used to docu- ment spider predation on moth eggs and lar- vae (Buschman et al. 1977; McCarty et al. 1980; McDaniel & Sterling 1979; Elvin et al. 1983; Godfrey et al. 1989). To use this ap- proach in the field, one must label and release large numbers of potential prey and then de- termine the proportion of total prey that are labeled. One must also assume that the re- leased animals and those of the natural pop- ulation are equally susceptible to predation, and that radioactivity is not being made avail- able to the spiders by other routes. Finally, this approach would be difficult to employ given environmental concerns and regulations (Elvin et al. 1983); besides, there are better alternatives (below). Chromatography: In a unique application, GREENSTONE— HOW AND WHY WE STUDY PREDATION 337 Putnam (1967) used paper chromatography to detect the pigments of mites that had been consumed by spiders. Electrophoresis: Gel electrophoresis of prey allozymes has been used for insect and mite gut analysis but not yet for spider gut analysis. One must pay serious attention to the choice of enzyme system and gel medium, and be a competent bench scientist (Solomon et al. 1996). If it could be made to work, an advantage would be relative economy and, with luck, the ability to distinguish a wide range of prey species with a single analytical system. Nucleic acid probes: Probes employing species-specific DNA sequences have been used to diagnose a number of arthropod inter- specific interactions, including pathogen host and parasitoid host (Greenstone & Edwards 1998), and could, in principle, be used to iden- tify prey remains in a spider’s gut. Serology: Vertebrate antibodies have been used for spider gut analysis for 35 years (Loughton et al. 1963). I have recently re- viewed this approach (Greenstone 1996) and shall here emphasize just three points. First, these are proven technologies with stable, re- producible protocols. Second, assay technol- ogy is getting cheaper and simpler; and ELISA, which requires expensive equipment, could be replaced by the immunodot (Stuart & Greenstone 1990; Greenstone & Trowell 1994; Agusti in press). Finally, any level of prey specificity is achievable, down to stage and even instar (Ragsdale et al.l981; Green- stone & Morgan 1989; Greenstone & Trowell 1994; Goodman et al. 1997). Although most directly achieved by monoclonal antibody technology, the same specificities might be achieved more cheaply by affinity chromatog- raphy of conventional antisera (Greenstone 1996). Serological assays have been used to study spider predation on all manner of arthropod prey (Greenstone 1996). They are particularly useful for studying oophagy, a poorly docu- mented phenomenon because of the small size and cryptic habit of eggs and short spider han- dling times (Nyffeler et al. 1990). For exam- ple, a single monoclonal antibody has re- vealed the extent of bollworm egg predation by two Cheiracanthium species in India and North America (Sigsgaard 1996; Ruberson & Greenstone 1998). Two problems in quantitating serological data remain. First, predators differ in digestive rates and are therefore differentially likely to contain detectable remains of a prey item at any interval post-feeding. Weighting factors, based on temperature-dependent detectabili- ties, are necessary to determine the relative importance of different predator species. Max- imum detectability intervals (Sunderland et al. 1987) and detectability half-lives (Greenstone & Hunt 1993; Agusti in press) have been pro- posed as weighting factors. Second, due to the extraordinary sensitivity of contemporary as- says, one generally cannot know the number of prey items represented by a serological pos- itive (Greenstone 1996). If one assumes that the number of prey contained in any given predator gut is a Poisson variate, then the pro- portion of negatives can be used as the zero class to calculate the mean number of prey individuals per gut (Nakamura & Nakamura 1977; Greenstone 1979; Lister et al. 1987). Since Poisson assumptions may not always be met, the model needs to be tested. Other mod- els have been suggested by Ashby (1974), Sunderland & Sutton (1980), and Sopp et al. (1992). Field experiments. — Direct evidence for the effectiveness of spiders in biological con- trol comes from field experiments in which spider numbers are manipulated and the re- sultant pest populations and attendant crop damage are compared to those in controls. Mansour et al. (1980) removed all of the spiders from half of a sample of apple trees in an abandoned orchard and then infested them with Egyptian cotton leafworm egg mas- ses. After five days, damage to egg masses was significantly greater; and larval popula- tions and leaf feeding damage were signifi- cantly lower on the controls. Analogous ex- periments were performed with the same insect on cotton (Mansour 1987) and a scale insect on citrus (Mansour & Whitcomb 1986), with similar results. Ito et al. (1962) used heptachlor to reduce spider numbers in a rice ecosystem. Spider densities were lower and planthopper and leafhopper densities and population growth rates were greater where plots had been sprayed. In a northern California rice ecosys- tem, Graze & Grigarick (1989) used floating rings with sticky tops to manipulate the num- bers of Pardosa ramulosa within. Rings with 338 THE JOURNAL OF ARACHNOLOGY higher spider densities had significantly re- duced densities of the aster leafhopper. Carter & Rypstra (1995) added artificial web sites (crates) and an inoculum of Achaearanea tepidariorum to some soybean plots and removed all spiders and uninhabited webs from others. Crates increased spider density, which was significantly correlated positively with insects killed and negatively with leaf area damaged. Riechert & Bishop (1990) increased the hospitability of a vegetable garden ecosystem by adding mulch, which significantly in- creased spider density and decreased pest den- sities and plant damage. After spider removal, the mulch treatments were no longer different from bare ground controls. Riechert & Lawrence (1997) manipulated numbers of the entire spider assemblage and also of four in- dividual species (an abundant small lycosid and linyphiid and less abundant but large ly- cosid and araneid), in an old field ecosystem. The entire assemblage significantly reduced insect herbivore numbers relative to spider re- moval controls, but individually the abundant small species could also significantly reduce the densities of some insect taxa. CONCLUSIONS All four approaches to studying spider pre- dation have some value, but laboratory feed- ing studies are only useful in selected cases; before succumbing to the temptation to per- form them, one should ask whether the re- sulting data are likely to be informative (How- ell & Pienkowski 1971). Direct observation is a powerful tool that will continue to provide useful information on prey spectrum and feed- ing rates. Serological gut analysis is the most efficient and least disruptive method available for gathering large-scale spider predation data on selected prey species. Experimental field manipulations provide the most powerful demonstrations of the efficacy of spider spe- cies and assemblages as biological control agents, and they can also serve as realistic tri- als for proposed management approaches (e.g., Riechert & Bishop 1990). ACKNOWLEDGMENTS I thank Tim Lockley and Martin Nyffeler for technical details on their papers listed in Table 2, Martin Nyffeler and Mike Solomon for alerting me to several overlooked refer- ences, and Mike Solomon and Keith Sunder- land for thoughtful reviews of the manuscript. LITERATURE CITED Anderson, J. 1974. Responses to starvation in the spiders Lycosa lenta Hentz and Filistata hiber- nalis (Hentz). Ecology, 55:576-585. Agusti, N. In press. Immunological detection of Helicoverpa annigera (Lepidoptera: Noctuidae) ingested by heteropteran predators: time related decay and effect of meal size on detection period. Ann. Entomol. Soc. America. Ashby, J.W. 1974. A study of arthropod predation of Pieris rapae L. using serological techniques. J. Appl. EcoL, 11:419-426. Bilde, T. & S. Toft. 1998. Quantifying food limi- tation of arthropod predators in the field. Oec- ologia, 115:54-58. Breene, R.G., M.H. Sweet & J.K. Olson. 1988. Spider predators of mosquito larvae. J. Arach- noL, 16:275-277. Buschman, L.L., W.H. Whitcomb, R.C. Hemenway, D.L. Mays, N. Ru, N.C. Leppla & B.J. Smittle. 1977. Predators of velvetbean caterpillar eggs in Florida soybeans. Environ. Entomol., 6:403-407. Carter, P.Y. & A.L. Rypstra. 1995. Top-down ef- fects in soybean agroecosystems: spider density affects herbivore damage. Oikos, 72:433-439. Chant, D.A. 1956. Predaceous spiders in orchards in south-eastern England. J. Hort. Sci., 31:35-46. Dean, D.A., W.L. Sterling, M. Nyffeler & R.G. Breene. 1987. Foraging by selected spider pred- ators on the cotton fleahopper and other prey. Southwest Entomol., 12:263-270. Dondale, C.D. 1970. Spiders of Heasman’s Field, a mown meadow near Belleville, Ontario. Proc. Entomol. Soc. Ontario, 101:62-69. Edgar, W.D. 1969. Prey and predators of the wolf spider, Lycosa lugubris. J. ZooL, 159:405-411. Edgar, W.D. 1970. Prey and feeding behaviour of adult females of the wolf spider, Pardosa amen- tata (Clerck). Netherlands J. ZooL, 20:487-491. Elvin, M.K., J.L. Stimac & WH. Whitcomb. 1983. Estimating rates of arthropod predation on vel- vetbean caterpillar larvae in soybeans. Florida Entomol., 66:319-330. Fletcher, R.K. & EL. Thomas. 1943. Natural con- trol of eggs and first instar larvae of Heliothis annigera. J. Econ. Entomol., 36:557-560. Ford, M.J. 1977. Energy costs of the predation strategy of the web-spinning spider Lepthyphan- tes zbnmermanni Bertkau (Linyphiidae). Oecol- ogia, 28:341-349. Gillebeau, L.P. &. J.N. All. 1989. Geocoris spp. (Hemiptera: Lygaeidae) and the striped lynx spi- der (Araneae: Oxyopidae): cross predation and prey preferences. J. Econ. Entomol., 82:1106- 1110. Godfrey, K.E., W.H. Whitcomb & J.L. Stimac. GREENSTONE— HOW AND WHY WE STUDY PREDATION 339 1989. Arthropod predators of velvetbean cater- pillar, Anticarsia gemmatalis Hiibaer (Lepidop- tera: Noctuidae), eggs and larvae. Environ. En- tomoL, 18:118-123. Goodman, C.L., M.H. Greenstone & M.K. Stuart. 1997. Monoclonal antibodies to vitellins of boll- worm and tobacco budworm (Lepidoptera: Noc- tuidae): biochemical and ecological implications. Ann. EntomoL Soc. America, 90:83-90. Greenstone, M.H. 1976. A quantitative field study of predatory behavior in the wolf spider, Pardosa ramulosa. Ph.D. Thesis. Univ. of California, Berkeley. 161 pp. Greenstone, M.H. 1979. Spider feeding behaviour optimises dietary essential amino acid composi- tion. Nature, 181:501-503. Greenstone, M.H. 1996. Serological analysis of ar- thropod predation: past, present and future. Pp. 265-300. In The Ecology of Agricultural Pests: Biochemical Approaches (W.O.C. Symondson & J.E. Liddell, eds.). Chapman & Hall, London. Greenstone, M.H. & A.F. Bennett. 1980. Foraging strategy and metabolic rates in spiders. Ecology, 61:1255-1259. Greenstone, M.H. & C.E. Morgan. 1989. Predation on Heliothis zea (Lepidoptera: Noctuidae): an in- star-specific ELISA for stomach analysis. Ann. EntomoL Soc. America, 82:45-49. Greenstone, M.H. & J.H. Hunt. 1993. Determina- tion of prey antigen half-life in Polistes metricus using a monoclonal antibody-based immunodot assay. EntomoL Exp. AppL, 68:1-7. Greenstone, M.H. & S.C. TrowelL 1994. Arthro- pod predation: A simplified immunodot format for predator gut analysis. Ann, EntomoL Soc. America, 87:214-217. Greenstone, M.H. & M.J. Edwards. 1998. A DNA hybridization probe for endoparasitism by MP cropUtis croceipes (Hymenoptera: Braconidae). Ann. EntomoL Soc. America, 91:415-421. Gregory, B.M., Jr., C.S. Barfield & G.B. Edwards. 1989. Spider predation on velvetbean caterpillar moths (Lepidoptera, Noctuidae) in a soybean field. J. ArachnoL, 17:120-122. Hayes, J.L. & T.C. Lockley. 1990. Prey and noc- turnal activity in wolf spiders (Araneae: Lycosi- dae) in cotton fields in the Delta region of Mis- sissippi. Environ. EntomoL, 19:1512-1518. Hodge, M.A. 1999. The implications of intraguild predation for the role of spiders in biological control. J. ArachnoL, 27(l):00-00. Howell, J.O. & R.L. Pienkowski. 1971. Spider populations in alfalfa, with notes on spider prey and effect of harvest. J. Econ. EntomoL, 64:163- 168. Hydhom, S.B. 1976. Laboratory biology of Par- dosa ramulosa (McCook). Ph.D. Thesis, Univ. of California, Berkeley. 251 pp. ltd, Y., K. Miyashita & K. Sekiguchi. 1962. Stud- ies on the predators of the rice crop insect pests using the insecticidal check method, Japanese J. EcoL, 12:1-11. Jackson, R.R. 1977. Prey of the jumping spider Phidippus johnsoni (Araneae: Salticidae). J. Ar- acheoL, 5:145-149. Jennings, D.T. & H.A. Pase, III. 1975. Spiders preying on Ips bark beetles. Southwestern Natur., 20:225-229. Jennings, D.T. & H.A. Pase, III. 1986. Spiders preying on Dendroctonus frontalis (Coleoptera: Scolytidae). EntomoL News, 97:227-229. Jmhasly, P. & W. Nentwig. 1995. Habitat manage- ment in winter wheat and evaluation of subse- quent spider predation on insect pests. Acta Oec- oL, 16:389-403. Kajak, A. 1965, An analysis of food relations be- tween the spiders Araneus cornutus Clerck and Araneus quadratus Clerck and their prey in meadows. EkoL Polska, 13:717-761. Kiritaei, K. & J.P. Dempster. 1973. Different ap- proaches to the quantitative evaluation of natural enemies, J. AppL EcoL, 10:323-330. Kiritani, K,, S. Kawahara, T. Sasaba & F. Nakasuji. 1972. Quantitative evaluation of predation by spiders on the green rice leafhopper, Nephotettix cincticeps Uhler, by a sight-count method. Res. Popul. EcoL, 13:187-200. Li, D. & R.R. Jackson. 1996. Prey-specific capture behaviour and prey preferences of myrmicophag- ic and araneophagic jumping spiders (Araneae; Salticidae). Rev. Suisse ZooL, 423-436. Lingren, P.D., R.L. Ridgway & S.L. Jones. 1968. Consumption by several common arthropod predators of eggs and larvae of two Heliothis species that attack cotton. Ann. EntomoL Soc. America, 61:613-618. Lister, A., M.B. Usher & W. Block. 1987. Descrip- tion and quantification of field attack rates by predatory mites: an example using an electro- phoresis method with a species of Antarctic mite. Oecologia, 72:185-191. Lockley, T.C. & O.P, Young. 1987. Prey of the striped lynx spider Oxyopes salticus (Araneae, Oxyopidae), on cotton in the Delta area of Mis- sissippi. J. ArachnoL, 14:395-397. Louda, S.M. 1982. Inflorescence spiders: a cost/ benefit analysis for the host plant, Haplopappus venetus Blake (Asteraceae). Oecologia, 55:185- 191. Loughton, B.G., C. Derry & A.S. West. 1963. Spi- ders and the spruce budworm. Mem. EntomoL Soc. Canada, 31:249-268. Lowrie, D.C. 1987. Effects of diet on the devel- opment of Loxosceles laeta (Araneae: Loxoscel- idae). J. ArachnoL, 15:303-308. McCarty, N.T., M. Shepard & S.G. Turnipseed. 1980. Identification of predaceous arthropods in 340 THE JOURNAL OF ARACHNOLOGY soybeans by using autoradiography. Environ. En= tomol., 9:199-203. McDaniel, S.G. & W.S. Sterling. 1979. Predator determination and efficiency on Heliothis vires- cens eggs in cotton by using Environ. Ento- mol., 8:1083-1087. Mansour, F. 1987. Spiders in sprayed and un- sprayed cotton fields in Israel, their interactions with cotton pests and their importance as pred- ators of the Egyptian cotton leafworm, Spodop- tera littoralis. Phytoparasitica, 15:31-41. 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, E, D. Rosen & H.N. Plant. 1980. Eval- uation of spiders as biological control agents of Spodoptera littoralis larvae on apple in Israel. Acta Oecol., 1:225-232. Miyashita, K. 1968. Growth and development of Lycosa t-insignata Boes et. Str. (Araneae: Ly- cosidae) under different feeding conditions. Appl. Entomol. ZooL, 3:81-88. Morse, D.H, 1979. Prey capture by the crab spider Misumena calycina (Araneae: Thomisidae). Oec- ologia, 39:309-319. Morse, D.H. 1984. How crab spiders hunt at flow- ers. J. ArachnoL, 12:307-316. Nakamura, M. & K. Nakamura. 1977. Population dynamics of the chestnut gall wasp, Dryocosmus kuriphillus Yasumatsu (Hymenoptera: Cynipi- dae). Oecologia, 27:97-116. Nentwig, W. 1986. Non-webbuilding spiders: prey specialists or generalists? Oecologia, 69:571- 576. Nuessly, G.S. & R.D. Goeden. 1983. Spider pre- dation on Coleophora parthenica (Lepidoptera: Coleophoridae), a moth imported for the biolog- ical control of Russian thistle. Environ. Ento- moL, 12:1433-1438. Nyffeler, M. 1982. Field studies on the ecological role of the spiders as insect predators in agroe- cosystems (abandoned grassland, meadows, and cereal fields). Ph.D. Thesis, Swiss Fed. Inst. Tech., Zurich. 174 pp. Nyffeler, M. 1999. Prey selection in the field. J. ArachnoL, 27(l):00-00. Nyffeler, M. & G. Benz. 1987. Spiders in natural pest control: a review. J. Appl. Entomol., 103: 321-329. Nyffeler, M. & G. Benz. 1988a. Prey and predatory importance of micryphantid spiders in winter wheat fields and hay meadows. J. Appl. Ento- mol., 105:190-197. Nyffeler, M. & G. Benz. 1988b. Feeding ecology and predatory importance of wolf spiders {Par- dosa spp.) (Araneae, Lycosidae) in winter wheat fields. J. Appl. Entomol., 106:123-134. Nyffeler, M., D.A. Dean & W.L. Stirling. 1987a. Evaluation of the importance of the striped lynx Oxyopes salticus (Araneae: Oxyopidae), as a predator in Texas cotton. Environ. Entomol., 16: 1114-1123. Nyffeler, M., D.A. Dean & W.L. Stirling. 1987b. Predation by green lynx spider, Peucetia viridans (Araneae: Oxyopidae), inhabiting cotton and wooly croton plants in East Texas. Environ. En- tomol., 16:355-359. Nyffeler, M., R.G. Breene, D.A. Dean & W.L. Stir- ling. 1990. Spiders as predators of arthropod eggs. J. Appl. Entomol., 109:490-501. Nyffeler, M., W.L. Stirling & D.A. Dean. 1992a. Impact of the striped lynx spider (Araneae: Ox- yopidae) and other natural enemies on the cotton fleahopper (Hemiptera: Miridae) in Texas cotton. Environ. Entomol., 21:1178-1188. Nyffeler, M., D.A. Dean & W.L. Stirling. 1992b. Diets, feeding specialization, and predatory role of two lynx spiders, Oxyopes salticus and Peu- cetia viridans (Araneae: Oxyopidae), in a Texas cotton agroecosystem. Environ. Entomol., 21: 1457-1465. O’Neill, R.J. 1997. Functional response and search strategy of Podisus maculiventris (Heteroptera: Pentatonudae) attacking Colorado potato beetle (Coleoptera: Chrysomelidae). Environ. Ento- mol., 26:1183-1190. Oraze, M.J. & A. Grigarick. 1989. Biological con- trol of aster leafhopper (Homoptera: Cicadelli- dae) and midges (Diptera: Chironomidae) by Pardosa ramulosa (Araneae: Lycosidae) in Cal- ifornia rice fields. J. Econ. Entomol., 82:745- 749. Provencher, L. & D. Coderre. 1987. Functional re- sponses and switching of Tetragnatha laboriosa (Araneae: Tetragnathidae) and Clubiona pikei Gertsch (Araneae: Clubionidae) for the aphids Rhopalosiphum maidis (Fitch) and Rhopalosi- phum padi (L.) (Homoptera: Aphididae). Envi- ron. Entomol., 16:1205-1309. Punzo, F. 1991. Field and laboratory observations on prey items taken by the wolf spider, Lycosa lenta Hentz (Araneae, Lycosidae). Bull. British ArachnoL Soc., 8:261-264. Putnam, W.L. 1967. Prevalence of spiders and their importance as predators in Ontario peach or- chards. Canadian Entomol., 99:160-170. Quaintance, A.L. & C.T. Brues. 1905. The cotton bollworm. U.S. Dept. Agric. Bur. Entomol. Bull., 50:1-155. Ragsdale, D.W., A.D. Larson & L.D. Newsome. 1981. Quantitative assessment of the predators of Nezara viridula eggs and nymphs within a soybean agroecosystem. Environ. Entomol., 10: 402-405. Randall, J.B. 1982. Prey records of the Green Lynx spider, Peucetia viridans (Hentz) (Araneae, Ox- yopidae). J. ArachnoL, 10:19-22. GREENSTONE—HOW AND WHY W STUDY PREDATION 341 Richman, D.B., W.E Buren & WH. Whitcomb. 1983a. Predatory arthropods attacking eggs of Diaprepes abbreviatus (L.) (Coleoptera: Curcu- lionidae) in Puerto Rico and Florida. J. Georgia EntomoL Soc., 18:335-342. Richman, D.B., W.H. Whitcomb & WE Buren. 1983b. Predation on neonate larvae of Diaprepes abbreviatus (L.) (Coleoptera: Curculionidae) in Florida and Puerto Rico citrus groves. Florida Entomol., 66:215-222. Riechert, S.E. & T. Lockley. 1984. Spiders as bi^ ological control agents. Ann. Rev. Entomol., 29: 299-320. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in a garden test system. Ecology, 71:1441-1450. Riechert, S.E. & K. Lawrence. 1997. Test for pre- dation effects of single versus multiple species of generalist predators: spiders and their insect prey. Entomol. Exp, AppL, 84:147-155. Ruberson, J. & M.H, Greenstone. 1998. Predators of bollworm/budworm eggs in cotton: an im- munological study. Proc. 1998 Beltwide Cotton Conf., 2:1095-1098. Sadana, G.L. & M. Kumari. 1991. Predatory po- tential of the lyssomanid spider, Lyssomanes sik- kimemis Tikader on the mango hopper, Idiosco- pus clypealis (Lethierry). Entomon, 16:283-285. Schaefer, M. 1974. Experimentelle Untersuchun- gen zur Bedeutung der interspezifischen Konk- urrenz bei 3 Wolfspinnen-Arten (Araneida: Ly- cosidae) eiiier Salzwiese. Zool. Jb. Syst. Bd., 101:213-235. Sigsgaard, L. 1996. Serological analysis of preda- tors of Helicoverpa armigera Hiibner (Lepidop- tera: Noctuidae) eggs in sorghum-pigeonpea in- tercropping at ICRISAT, India: a preliminary field study. Pp. 367-382. In The Ecology of Ag- ricultural Pests: Biochemical Approaches (W.O.C. Symondson & J.E. Liddell, eds.). Chap- man & Hall, London. Solomon, M.G., J.D. Fitzgerald & R.A. Murray. 1996. Electrophoretic approaches to predator- prey interactions. Pp. 457-468. In The Ecology of Agricultural Pests: Biochemical Approaches (W.O.C. Symondson & J.E. Liddell, eds.). Chap- man & Hall, London. Sopp, P.I., K.D. Sunderland, J.S. Fenlon & S.D. Wratten. 1992. An improved method for esti- mating invertebrate predation in the field using an enzyme-linked immunosorbent assay (ELISA). J. AppL EcoL, 29:295-302. Stuart, M.K. & M.H. Greenstone. 1990. Beyond ELISA: a rapid, sensitive, specific immunodot assay for identification of predator stomach con- tents. Ann. Entomol. Soc. America, 83:1101- 1107. Sunderland, K.D. 1996. Progress in quantifying predation using antibody techniques. Pp.419- 455. In The Ecology of Agricultural Pests: Bio- chemical Approaches (W.O.C. Symondson & J.E. Liddell, eds.). Chapman & Hall, London. Sunderland, K.D. 1999. Effects of spiders on pest populations: mechanisms. J. ArachnoL, 27(1): 000-000. Sunderland, K.D. & S.L. Sutton. 1980. A serolog- ical study of arthropod predation on woodlice in a dune grassland ecosystem. J. Anim. EcoL, 49: 987-1004. Sunderland, K.D., A.M. Fraser & A.F.G. Dixon. 1986. Field and laboratory studies of money spi- ders (Linyphiidae) as predators of cereal aphids. J. Appl. EcoL, 23:433-447. Sunderland, K.D., N.E. Crook, D.L. Stacey & B.J. Fuller. 1987. A study of feeding by polyphagous predators on cereal aphids using ELISA and gut dissection. J. Appl. EcoL, 24:907-933. Toft, S. 1995. Value of the aphid Rhopalosiphum padi as food for cereal spiders. J. Appl. EcoL, 32:552-560. Toft, S. 1996. Indicators of prey quality for arthro- pod predators. Pp. 107-116. In Arthropod natural enemies in arable land. II. Survival, reproduction and enhancement. (K. Booij & L. den Nijs, eds.). Aarhus Univ. Press, Aarhus, Denmark. Turnbull, A.L. 1973. Ecology of the true spiders (Araneomorphae). Ann. Rev. Entomol., 18:305- 348. Turner, M. 1979. Diet and feeding phenology of the green lynx spider, Peucetia viridans (Ara- neae, Oxyopidae). J. ArachnoL, 7:149-154. Uetz, G.W., J. Bischoff & J. Raver. 1992. Survi- vorship of wolf spiders (Lycosidae) reared on different diets. J. ArachnoL, 20:207-211. Van Dyke, D. & D.C. Lowrie. 1975. Comparative life histories of the wolf spiders Pardosa ramu- losa and P. sierra (Araneae: Lycosidae). South- west Natur., 20:29-44. Whitcomb, W.H. 1967. Field studies on predators of the second-instar bollworm, Heliothis zea (Boddie) (Lepidoptera: Noctuidae). J. Georgia Entomol. Soc., 4:113-118. Whitcomb, W.H. & K. Bell. 1964. Predaceous in- sects, spiders, and mites of Arkansas cotton fields. Arkansas Agric. Sci. Exp. Sta. Bull., 690: 3-84. Whitcomb, W.H. & M. Tadic. 1963. Araneida as predators of the fall webworm. J. Kansas Ento- mol. Soc., 36:186-190. Whitcomb, W.H., H. Exline & R.C. Hunter. 1963. Spiders of the Arkansas cotton field. Ann. En- tomol. Soc. America, 56:653-660. Young, O.P. 1989a. Predation of the tarnished plant bug, Lygus iineolaris (Heteroptera: Miridae): laboratory evaluations. J. Entomol. Sci., 23:174- 179. Young, O.P. 1989b. Predation by Pisaurina mira (Araneae, Pisauridae) on Lygus Iineolaris (Het- 342 THE JOURNAL OF ARACHNOLOGY eroptera, Miridae) and other arthropods. J. Ar- achnoL, 17:43-48. Young, O.R 1989c. Field observations of predation by Phidippus audax (Araneae: Salticidae) on ar- thropods associated with cotton. J. Entomol. Sci., 24:266-273. Young, O.R & G.B. Edwards. 1990. Spiders in United States field crops and their potential effect on crop pests. J. ArachnoL, 18:1-27. Zimmermann, M. & J.R. Spence. 1989. Frey use of the fishing spider Dolomedes triton (Pisauri- dae, Araneae): an important predator of the neus- ton community. Oecologia, 80:187-194. Manuscript received 1 May 1998, revised 17 Sep- tember 1998. 1999. The Journal of Arachnology 27:343-350 SPIDER COMPETITION IN STRUCTURALLY SIMPLE ECOSYSTEMS Samuel D. Marshall: Department of Zoolo.gy, Miami University, Oxford, Ohio 45056 USA Anu L. Rypstra: Department of Zoology, Miami University, 1601 Peck Boulevard, Hamilton, Ohio 45011 USA ABSTRACT. Spider competition has long been an elusive phenomenon for ecological study. Because most spiders are generalist predators, they are predicted to overlap in resource use wherever they overlap in space use and activity periods. However, despite this obvious potential for competition, the empirical evidence for competition has been weak. Spider competition could potentially limit densities in agricultural ecosystems, which would limit their effectiveness as biological control agents. We summarize the results of five studies in a type of ecosystem which may be considered to be analogous to row crops in both the physiognomy of vegetation and cyclic disturbance regimes, namely, v/etlands. In addition, we summarize the results of our own work in a soybean, ecosystem. Interspecific competition occurs whenever the availability of a limiting resource, is re- duced by the presence of two or more species which rely on the same resource. Species which are more efficient or aggressive in re- source use are predicted to prevail over other species. Exploitative competition occurs when one species has a negative effect on individ- uals and/or populations of another because it is more efficient at using a limiting resource. However, competitive interactions can also be behavioral, such as when one species alters the space use or feeding behavior of a second species, resulting in a reduction in access to important resources by the second species. This is termed 'interference competition'. It seems intuitively obvious that resource limitation by the action of other species should be an important phenomenon influenc- ing the success of species populations and, ul- timately, community structure. However, just how prevalent and significant competition is has been the subject of extensive and often contentious debate for decades (Strong et al. 1984). One of the key assumptions in com- petition theory is that one of the species in a community is better at exploiting some lim- iting resource, and does so to the detriment of ^Present address: James H. Barron Field Station, Dept, of Biology, Hiram College, Hiram, Ohio 44234 USA. other species. Because both resource avail- ability and population densities of the com- petitors may vary for reasons that have noth- ing to do with the activities of the competing species, competition may not be a persistent phenomenon. Also, no two species will ever use the same resource in precisely the same way. Even small variations in resource ex- ploitation strategies will reduce the intensity of competitive interactions. The most generally accepted paradigm in spider ecology is that spiders are food-limited generalist predators of arthropods, particularly insects. Spiders rarely specialize in specific prey taxa; however, the prey spectrum of any spider species population will never be a per- fect reflection of all potential prey in their en- vironment. Some prey will either be rejected as too risky because they possess defenses, or will not be encountered at all as they are not active in the same places at the same times of the day as the spiders are foraging. However, given these constraints on prey use, there is still the potential for different species of gen- eralist predators like spiders to share prey taxa (Nyffeler & Sterling 1994). Intraspecific com- petition would be predicted to be the stron- gest, given that resource use should overlap 100% within a single species age/size class. However, because many spiders are territorial, and potentially cannibalistic, they will tend to be self-limiting. Because of these population 343 344 THE JOURNAL OF ARACHNOLOGY regulation mechanisms, spider densities are predicted to rarely be high enough to set the stage for density-dependent resource limita- tion (Riechert 1974). Evidence for interspecific competition in spiders has been rare enough that it may be viewed as the exception and not the rule (Wise 1993). However, there are similarities among the strongest examples of interspecific com- petition we found which suggest that compe- tition may be more prevalent in structurally simple ecosystems which exhibit a gradient of habitat structure than those ecosystems which are more uniform and exhibit greater vertical stratification. Examples of structurally simple successional ecosystems are estuaries, wet- lands, and an anthropogenic analog, agricul- tural ecosystems. Early research in spider community ecology documented the fact that early successional habitats had lower spider species diversity when compared to later serai stages nearby (Lowrie 1948; Barnes 1953). In these cases, this relatively low spider species diversity was functionally linked to the rela- tive structural simplicity of the vegetation. The link between vegetation structure and spi- der species diversity was explicitly tested by Greenstone (1984). This link between struc- tural simplicity and species poverty was ex- tended to the litter layer by Uetz (1979). The link between the structural complexity of hab- itat and spider species diversity is so well-es- tablished as to be a paradigm in spider ecol- ogy (Uetz 1991; Wise 1993). Wetland ecosystems are structurally simple because they are biotically simple. Tidal es- tuaries, in particular, have few plant species which in turn supports a species poor com- munity of herbivores and predators. In addi- tion, there is an inshore-outshore gradient of the intensity and cycle of disturbance which results from cyclical disturbance by flowing water. Along this disturbance gradient habitat complexity varies from barren soil at the wa- ter’s edge, to complex vegetation inshore (Cameron 1972; Uetz 1976). Agricultural eco- systems are structurally simple systems by de- sign. Intensive cyclic disturbance regimes in tandem with the mechanized dispersal of seeds are all techniques used to attempt to cre- ate landscapes colonized by one species: The crop plant. However, despite the strenuous ef- forts of agriculturists, many species of plants and animals do colonize agricultural fields. Because most crop plants have an annual crop cycle, and because of the disturbances inher- ent in crop management, most invading spe- cies of plants and animals are those naturally associated with early successional ecosystems (Wissinger 1997). The evolved ability of these pioneer species to invade and reproduce in a narrow window of time, tolerate abiotic ex- tremes, and benefit from disturbance prea- dapts them to life in an agricultural field (Luc- zak 1979). Spiders are common and often abundant in- cidental colonists of agricultural ecosystems. Young & Edwards (1990) found that 614 spi- der species are found in crop systems in North America. Studies of the spiders associated with crops found that the spiders found there are generally not species invading from local natural ecosystems, but those species found in the surrounding agricultural landscape (Duelli et al. 1990; Bishop & Riechert 1990). In our own fields at the Miami University Ecology Research Center in southwestern Ohio we found that the two dominant wolf spider spe- cies (Hogna helluo (Walckenaer) and Pardosa milvina Hentz) are absent from the native for- est adjacent to the fields, but common in ri- parian and pond edge habitats in the area (Marshall & Rypstra unpubl. data). Araneid competition in a California salt marsh. — Dave Spiller’s 1984 study of the competitive interactions between two ara- neids, Cyclosa turbinata (Walckenaer) and Metepeira grinneli (Coolidge), in a salt marsh ecosystem near San Francisco remains one of the clearest experimental demonstrations of interspecific competition between spiders to date (Wise 1993). The higher and drier parts of this marsh are dominated by two plant spe- cies: Salicornia pacifica and Baccharis Doug- lasii. Salicornia is a low-growing woody pe- rennial herb, and Baccharis a shrub (Cameron 1972). Spiller was motivated to study the po- tential for competition between these two spe- cies by both the high densities achieved by these spiders and the apparent absence of predators. Densities of these spiders are in- deed high. In ten 1 X 1 m plots established to study the phenology of the focal species, Me- tepeira densities approximated 1 per m^, and Cyclosa 10 per m^ (Spiller 1984). Metepeira was the larger of the two, having approxi- mately twice the body length of Cyclosa as an adult (Spiller 1984). MARSHALL & RYPSTRA— SPIDER COMPETITION 345 Spiller conducted his experimental studies in replicated 4 X 1 m plots, with three plots for each removal treatment and three controls. Across these plots the foliage height approx- imated 1.25 m. Spiller removed Cyclosa or Metepeira over a six month period and noted the response by the other species in numbers, web height, and fecundity. Spiller found evi- dence of both exploitative and interference competition in this pair of spiders. Removing the much more common Cyclosa resulted in an increase in the body size and fecundity of Metepeira. The mechanism to explain this re- sult is the increased foraging success of Me- tepeira in plots free of Cyclosa. The much larger Metepeira, on the other hand, engaged in interference competition with Cyclosa. Both Cyclosa densities and web placement were affected by Metepeira. Cyclosa numbers were higher, and their webs were higher in the vegetation in the Metepeira removal plots. Spider’s findings are significant because they represent clear evidence of how interspe- cific competition can influence both space use and fitness in spiders. However, it is also re- vealing that these results come out of a type of terrestrial ecosystem which is structurally simpler than most. The salt marsh ecosystem in which he worked was dominated by two plant species, and this vegetation had a low growth form. In addition, the densities of these two species were high enough that they were likely to have the opportunity to interact. Wolf spider competition in a German salt marsh. — Schaefer (1972, 1974; summa- rized in English in Schaefer 1975 and Wise 1993) studied how competition might explain the patterns of space use exhibited by three lycosid species across a shoreline-salt marsh ecotone on a bay of the Baltic Sea near Kiel, Germany. These lycosids were: Pirata pirati- cus (Clerck), Pardosa purbeckensis (EO.R- Cambridge), and Pardosa pullata (Clerck). The Pirata and Pardosa purbeckensis were more common in the salt marsh, and Pardosa pullata was more common on the dry ground fringing the salt marsh. However, all three overlapped in substratum use. Pirata was the largest of the three and would opportunisti- cally prey upon the two Pardosa spp. Schae- fer found that Pardosa purbeckensis com- prised up to 8.2% of the diet of Pirata. Schaefer conducted density manipulation experiments in the salt marsh as did Spiller (1984). His treatment consisted of reducing Pirata densities by applying insecticide (a Parathion® spray) to four 10 X 10 m plots. He had a fifth untreated control plot. The Parathion® treatment initially eliminated all spiders, however, as Pirata was much slower to reinvade the plots than the Pardosa spp., its populations were never able to recover from the treatments (as he intended). Addi- tionally, he sprayed the edges of the plots weekly to further limit recolonization by Pir- ata (Schaefer 1974, reported in Wise 1993). Schaefer (1975) reported a significant and im- pressive reduction in Pirata numbers as a re- sult of his manipulations. Mean densities of Pirata in the Parathion plots was 2.2 per m^, and in the control plot, 14 per m^ (Schaefer 1975, table 1). On the other hand, both Par- dosa spp. densities were higher in the Para- thion plots than in the control plots (P. pur- beckensis; Parathion, 9.6 per m^, control 8.3 per m^; P. pullata; Parathion 3.5 per m^, con- trol 2.8 per m^). Schaefer found no statistically significant increase of Pardosa spp. numbers in the Parathion (i.e., Pirata removal) plots. As Wise (1993) notes, it is hard to interpret these results in light of the fact that Schaefer had only one control plot. However, it is none- theless interesting to note that the two Par- dosa spp. were in higher numbers in the Para- thion plots despite the insecticidal application, when Pirata numbers were so low. Schaefer conducted a second experiment in six 2 X 2 m enclosures in the salt marsh. He had six treatments, with one replicate each (Schaefer 1975). In each he added either each Pardosa spp. alone, each alone with Pirata, the two Pardosa spp. alone together, and all three together. All spiders where added in nat- ural densities {Pirata: 2.5 per m^, P. purbeck- ensis: 2.0 per m^, P. pullata: 0.5 per m^). He found that the mean densities of P. purbeck- ensis were higher in the enclosures where it was alone or with P. pullata (9.7 per m^) than in the two enclosures with Pirata or the open control plots (6.5 per m^. He also found that P. pullata was more common in the enclo- sures where it was alone, or with P. purbeck- ensis only (3.3 per m^ than in the two enclo- sures with Pirata or in the single open control plot (2.5 per m^). Based on his statistical anal- ysis Schaefer concluded that he did not have any basis to invoke competition in the patterns he observed. However, in his reanalysis of the 346 THE JOURNAL OF ARACHNOLOGY data. Wise (1993) suggests that Schaefer was too conservative in drawing his conclusions and that in fact he had evidence for interspe- cific competition. Schaefer found limited, but compelling ev- idence that the larger Pirata piraticus had an impact on numbers of the two Pardosa spp. where they co-occurred. This interaction was primarily mediated by predation by Pirata on the Pardosa spp., as well as presumably some interference interactions. Like the two ara- neids in Spiller’s (1984) study, these combat- ive lycosids lived at relatively high densities in a structurally two-dimensional ecosystem. Space use by lycosids on the New Jersey shore. — Dobel et al. (1990) examined the re- lationship between vegetation structure and tidal flooding with spider community structure in a salt marsh near Tuckerton, New Jersey. This study was specifically designed to ex- amine species diversity as a function of hab- itat structure and disturbance. However, they also uncovered indirect evidence of competi- tion between two lycosid species, Lycosa mo- desta (Keyserling) and Pardosa littoralis Banks. This study was entirely correlative, and consisted of a series of D-VAC samples taken in two replicates of four categories of Spartina spp. vegetation. On the shore was S. patens, a low growth form type of grass with a thick thatch base. Below the mean high wa- ter mark were three growth forms of S. alter- niflora: short, intermediate, and tall. Every two weeks between early May and October they D-VAC ’d four sites within each of the eight plots. Each sample was four, 15 sec ap- plications of the D-VAC head to the vegeta- tion. They found that L. modesta and P. lit- toralis were the numerically dominant hunting spiders in this salt marsh ecosystem (Dobel et al. 1990). The authors concluded that the pat- terns they observed for most species were in fact correlated with the architecture of the Spartina, the depth of the thatch, and the cycle of tidal disturbance. One exception was for the habitat use patterns of L. modesta and P. littoralis. Pardosa littoralis, alone among the species graphed (Dobel et al. 1990), exhibits a dip in densities in the short form Spartina. This is the second of the four successive cat- egories of Spartina habitat going from higher to lower elevations. This is noteworthy as all the other species graphed exhibit single peak in abundance across the inshore-outshore gra- dient. This dip in densities exhibited by P. lit- toralis (to almost zero) corresponds to the sin- gle and very pronounced peak in L. modesta numbers. The authors note: . .it is likely that the larger (10-12 mm) and more aggres- sive L. modesta simply drives the smaller (6 mm) P. littoralis from short-form S. alterni- folia."' (Dobel et al. 1990). Unfortunately, the authors do not detail the basis for their as- sessment of the temperament of L. modesta. However, given the strength of the pattern they observed, it would be well worth an ex- perimental study of the interactions of these two lycosids. Lycosid competition in a hot springs eco- system.— John Moeur studied the foraging consequences of competition between two ly- cosids {Pirata maculatus Emerton and Par- dosa altamontis Chamberlin & Ivie) which live on bluegreen algal mats near hot springs in Serendipity Meadow at Yellowstone Na- tional Park (Moeur 1977). The ecosystem Moeur studied consists for the most part of the algal mats, a brine fly (Diptera: Ephydri- dae, Paracoenia turbida) which feeds on the algae, and the two lycosids which feed on the flies. Through direct observation, Moeur de- termined that the brine fly P. turbida com- prised at least half the diet of both spider spe- cies (Moeur 1977). He also found that they overlapped broadly in activity periods, al- though the Pirata was more active early and late in the day, and the Pardosa was more active near midday (Moeur 1977). Moeur ob- served agonistic interactions between the two spiders, which led him to investigate the po- tential for competition between them: “When two meet, the larger spider invariably drives off the smaller, though it rarely kills it. En- counters between equal-sized individuals evoke much leg waving and feints by each animal before one routs the other. After a con- frontation lasting a few seconds, P. maculatus usually drives away P. altamontis." (Moeur 1977, p. 30). Moeur conducted enclosure experiments in which he tested for the impact of Pirata on Pardosa survival and fecundity. He construct- ed five artificial streams 1.22 X 2.44 m in area. He added vegetation (i.e., algae) from the surrounding hot springs habitats. He set up different ratios of 20 adult female Pirata and Pardosa in each. These ratios were: 20:0, 15: 5, 10:10, 5:15, and 0:20 Pardosa.Pirata. MARSHALL & RYPSTRA— SPIDER COMPETITION 347 Moeur apparently added and removed the Pardosa from these systems on a weekly ba- sis, although it is hard to tell from his descrip- tion of the methods (Moeur 1977). Whatever the details of his methods, he found a clear pattern of reduced Pardosa survivorship in enclosures with more Pirata (Moeur 1977). He attributes this outcome to interference competition mediated by the agonistic inter- actions he observed. Field and lab test for lycosid competition in soybean agroecosystems. — We have been studying the ecology and interactions of two lycosids, Hogna helluo (Walckenaer) and Par- dosa milvina Hentz, in two soybean agroeco- systems since 1994. These two spiders seemed to exhibit elements of some of the systems we have reviewed above: A larger, less vagile, and less common species {Hogna) which we predicted might engage in interference com- petition and intraguild predation with the smaller, more vagile, and more common spe- cies {Pardosa). In the first field season (1994) we recorded both species in high densities. Using a restricted-area search sampling meth- od, we recorded a high of approximately 0.8 Hogna per m^, and 2.0 Pardosa per m^ (Mar- shall & Rypstra in press). We tested for com- petition in both the field and lab. Our field studies of competition were con- ducted in replicated field plots (Marshall et al., unpubl. data). In six, 0.42 ha soybean plots we created eight 6 X 6 m islands of enhanced wolf spider microhabitat using wild bird seed (to increase vegetation) and wheat straw mulch. We enhanced prey availability by add- ing composted vegetable waste to selected subplots. We had three treatments and a con- trol, repeated for each species (for a total of eight subplots): 1) spiders added, 2) prey at- tractants added, 3) spiders and prey attractants added, and 4) control. Each two weeks from June to September we added the compost and/ or the spiders. We added Pardosa and Hogna in numbers similar to natural densities we ob- served in the field (Marshall & Rypstra in press); 25 Hogna and 36 Pardosa per 36 m^ subplot). At the end of the season we censused a third of the area of each subplot using a restricted area search method. We found no significant decrease in Pardosa numbers in the Hogna addition plots, although these were the lowest Pardosa densities we recorded (Marshall et al. unpubl. data). Interestingly, we collected more Pardosa than we added to most plots, but recovered far fewer Hogna. This indicates that something was limiting Hogna densities, perhaps intraspecific com- petition. The extremely low densities of Hog- na when compared to Pardosa may also ex- plain the limited evidence for interspecific competition we found. We also conducted laboratory mesocosm tests for competitive interactions between the two species. We compared the weight gain of eight Pardosa held alone to the weight gain of six Pardosa in enclosures with a single Hogna. We applied a paraffin muzzle to the Hogna to prevent them from preying on the Pardosa. This was done using a protocol de- vised by Jerome Rovner (Rovner 1980). These mesocosms were 40 liter aquaria (25 cm X 50 cm X 32 cm). Each aquarium was filled with soil from the soybean fields to a depth of 3.5 cm. One half the soil surface in each aquarium was then covered in 3.5 cm of pine bark mulch. A house plant {Syngonium podophyl- lum), 17-25 cm, high was placed in the mulched half of the tank to offer vertical strat- ification. The tanks were lit on a 12:12 light: dark cycle by standard fluorescent lights sus- pended over the tops of the tanks. The Hogna were placed in randomly assigned tanks one day before the Pardosa were added. Pardosa were randomly assigned to treatments, and weighed and placed into their assigned tank. This was done in the early evening; data col- lection began the following morning and last- ed one week. Vestigial-winged fruit flies {Drosophila melanogaster) were introduced into the tanks twice during the week as food for the Pardosa. The flies were added in suf- ficient numbers to assure a constant supply of prey. We misted the soil often enough to keep the soil moist. At the end of seven days the spiders were removed from the tanks and weighed. We found a significant reduction in weight gain by Pardosa in the presence of Hogna. Pardosa alone gained 45.3 ± 10.6 (x ± 1 S.D.) percent of their body weight, compared to 24.3 ± 5.1% when housed with Hogna (one-tailed Mest on the residuals of a regres- sion of In transformed weight gain on In trans- formed carapace width, t = -1.784, df = 12, P < 0.05). Despite the small sample size, the results are clear: Pardosa alone gained almost twice the weight of Pardosa in the presence 348 THE JOURNAL OF ARACHNOLOGY of Hogna. This was interesting, given that the Hogna were hidden in refuges during the ‘daylight’ hours when the Pardosa were most active. We have good evidence that Hogna has a negative influence on a fitness associated be- havior, foraging, in Pardosa. Our field test did not reveal a strong negative effect of Hogna on Pardosa; however, Hogna numbers re- mained low in spite of repeated Hogna addi- tions. It may be that Hogna is more prone to engage in intraspecific interactions than Par- dosa, and so be more self-limiting. Our find- ing of a negative effect in the small laboratory mesocosms but not in the field coupled with the low density of Hogna in the field may mean that there is only a local negative effect of Hogna on Pardosa. CONCLUSIONS Spiders are conspicuous components of ag- ricultural ecosystems wherever found. How- ever, most work to date on spiders in agroe- cosystems has focused on the ecology of their predation on arthropod pests rather than com- petitive interactions. In general, the most com- pelling evidence for interspecific competition in spiders comes from studies undertaken in structurally simple successional habitats such as salt marshes. Can we draw any general con- clusions about the likelihood and consequenc- es of interspecific competition among spiders in agroecosystems? Agroecosystems and littoral ecosystems, like other early successional ecosystems, are structurally simple when compared to later serai stages. This is because both are subject to regular cycles of disturbance, be it plowing in agricultural fields, or tidal or seasonal flooding as in estuarine or littoral ecosystems. Disturbance will inevitably limit the species diversity, and select for colonists which are vagile, fecund, or in other words ‘weedy’ (Gibson et al. 1992; Wissinger 1997). Because of the nature of the community structure in these simple ecosystems, the commonest spe- cies found there tend to occur in high densi- ties. This, coupled with the structural simplic- ity of the vegetation, also increases the opportunity for interactions. The consequence is the potential for competition. Because lit- toral zones are ecotones between highly dis- turbed and less disturbed microhabitats, spider species with divergent but overlapping micro- habitat associations may come into contact. This kind of fine-scale microhabitat segrega- tion has been documented in lycosids (den Hollander & Lof 1972; Greenstone 1980). The zone of overlap is where we would be most likely to see competition occur, based on the studies reviewed herein. The wetland-crop- land analogy is more than functional, as Luc- zak (1979) reports that many of what she terms ‘agrobiont’ spiders in Poland are natu- rally found in littoral ecosystems. We have likewise found that Pardosa milvina and Hog- na helluo co-occur in littoral and riparian hab- itats as well as in the soybean agroecosystems in southwest Ohio. The studies we reviewed, as well as ours, indicate that both exploitation and interference competition can occur between spiders. Spill- er (1984) found that his species pair engaged in both, with the larger and rarer Metepeira dominating the smaller and more common Cy- closa in behavioral interactions. However, because Cyclosa occurred in much higher densities, it did engage in exploitation com- petition with Metepeira. This makes it unclear just which species is the competitive dominant of the pair. If we conclude that it is Cyclosa, by virtue of its greater densities, then the mechanism of its success is its smaller body size and apparent greater toleration of conspe- cifics allowing it to attain greater densities and so deplete prey. The dominant of the trio of lycosids Schaefer studied (1975) exhibited the highest densities, and engaged in interference competition with the two inferior competitors. However, in this study, as in Dobel et al. (1990), there was separation by substratum type. Competition, if observed, occurred in ar- eas where the preferred microhabitats over- lapped. This pattern of a commoner competi- tive dominant was not seen for the lycosids studied by Dobel et al. (1990) or Moeur (1977). In these two studies, the more aggres- sive competitive dominant was no more com- mon, or in fact rarer, than the less aggressive species. In our own studies with Hogna and Pardosa we found that while Hogna did have a local negative effect on Pardosa activity, it is the rarer species in the fields (Marshall & Rypstra in press). Which is the superior com- petitor? Hogna may engage in interference competition and intraguild predation with Par- dosa; however, we have no evidence that Par- MARSHALL & RYPSTRA— SPIDER COMPETITION 349 dosa exploits the prey base to the detriment of Hogna populations. What is the implication of interspecific competition for biological control of arthro- pod pests in agroecosystems? Because spiders are generally self-limiting via agonistic inter- actions and cannibalism, they may have lim- ited utility as biocontrol agents (Riechert & Lockley 1984). Between species competitive interactions could likewise limit species di- versity and abundance. However, as Breene et al. (1993) note, the ability to engage in can- nibalism and intraguild predation during times of low pest abundances may help maintain spider populations at some minimum in the crop fields. Spiller (1986) found that compe- tition limited the predatory efficiency of the two species salt marsh system he worked on. He proposed that Cyclosa alone would better limit herbivore numbers than a combination of Metepeira and Cyclosa because the inter- ference interactions between the two would limit Cyclosa densities. Luczak (1979) also proposes that spider competition should re- duce prey-limitation by spider populations. However, Luczak also suggests that spider competition should be lower in littoral and agroecosytems than in ecosystems with higher spider species diversity (1979). Riechert & Lawrence (1997) on the other hand, found ev- idence that a multispecies spider assemblage would better reduce pest insect abundances because a group of species would occupy a wider range of niches than any single species could. In our own studies we uncovered lim- ited evidence that Hogna could affect both Pardosa densities as well as foraging efficien- cy. We would predict that Pardosa alone would achieve greater abundances and prey population control. Spider competition potentially occurs in agroecosystems because the community struc- ture and physical structure of these engineered ecosystems may promote high densities of a few spider species. How important competi- tion is in moderating the impact of spider pop- ulations on pest insect populations will de- pend on the spider species present in the fields. ACKNOWLEDGMENTS We thank the members of the Miami Uni- versity Spider Group for assistance in the field and lab, M. Hodge for her help in reference spotting and moral support, and the comments of D. Wise on the manuscript. Our research was funded by Miami University’s Ecology Research Center, the Departments of Zoology of Miami University Oxford and Hamilton campuses, and the National Science Founda- tion grant DEB-9527710. LITERATURE CITED Barnes, R. 1953. The ecological distribution of spi- ders in non-forest maritime communities at Beaufort, North Carolina. Ecol. Mono., 23:315- 337. Bishop, L. & S. Riechert. 1990. Spider coloniza- tion of agroecosystems: mode and source. Envi- ron. EntomoL, 19:1738-1745. Breene, R., D. Dean, M. Nyeffeler & G. Edwards. 1993. Biology, predation ecology, and signifi- cance of spiders in Texas cotton ecosystems. Texas Agric. Exp. Stat., B-1711, College Station, Texas. Cameron, G. 1972. Analysis of insect trophic di- versity in two salt marsh communities. Ecology, 53:58-73. Dobel, H., R. Denno & J. Coddington. 1990. Spi- der (Araneae) community structure in an inter- tidal salt marsh: effects of vegetation structure and tidal flooding. Environ. EntomoL, 19:1356- 1370. Duelli, R, M. Studer, 1. Marchand & S. Jakob. 1990. Population movements of arthropods be- tween natural and cultivated areas. Biol. Con- serv., 54:193-207. Gibson, C., C. Hambler & V. Brown. 1992. Chang- es in spider (Araneae) assemblages in relation to succession and grazing management. J. Appl. EcoL, 29:132-142. Greenstone, M. 1980. Contiguous allotopy of Par- dosa ramulosa and Pardosa tuoba (Araneae: Ly- cosidae) in the San Francisco Bay region, and its implications for patterns of resource partitioning in the genus. American Midi. Nat., 104:305-311. Greenstone, M. 1984. Determinants of web spider species diversity: vegetation structural diversity vs. prey availability. Oecologia, 62:299-304. den Hollander, J. & H. Lof. 1972. Differential use of the habitat by Pardosa pullata (Clerck) and Pardosa prativaga (L. Koch) in a mixed popu- lation (Araneae, Lycosidae). Tijd. EntomoL, 115: 205-215. Lowrie, D. 1948. The ecological succession of spi- ders of the Chicago area dunes. Ecology, 29: 334-351. Luczak, J. 1979. Spiders in agrocoenoses. Polish Ecol. Stud., 5:151-200. Marshall, S. & A. Rypstra. In press. Patterns in the distribution of two wolf spiders (Araneae, Ly- cosidae) in two soybean agroecosystems. Envi- ron. EntomoL 350 THE JOURNAL OF ARACHNOLOGY Moeur, J. 1977. Bioenergetic consquences of in- terspecific competition between two wolf spider species (Araneae; Lycosidae). Unpubl. Ph.D. the- sis, Univ. of Georgia. Athens, Georgia. 115 pp. Nyffeler, M. & W. Sterling. 1994. Comparison of the feeding niche of polyphagous insectivores (Araneae) in a Texas cotton plantation: Estimates of niche breadth and overlap. Environ. EntomoL, 23:1294-1303. Riecheit, S. 1974. Thoughts on the ecological sig- nificance of spiders. BioScience, 24:352-356. Riechert, S. & T. Lockley. 1984. Spiders as bio- logical control agents. Ann. Rev. EntomoL, 29: 299-320. Riechert, S. & K. Lawrence. 1997. Test for pre- dation effects of single versus multiple species of generalist predators: spiders and their insect prey. EntomoL Exp. AppL, 84:147-155. Rovner, J. 1980. Morphological and ethological adaptations for prey capture in wolf spiders (Ar- aneae: Lycosidae). J. ArachnoL, 8:201-215. Schaefer, M. 1975. Experimental studies on the im- portance of interspecies competition for the ly- cosid spiders in a salt marsh. Pp. 86-90. In Proc. Sixth Intern. ArachnoL Cong. Spiller, D. 1984. Competition between two spider species: experimental field study. Ecology, 65: 909-919. Spiller, D. 1986. Interspecific competition between spiders and its relevance to biological control by generalist predators. Environ. EntomoL, 15:177- 181. Strong, D., D, Simberloff, L. Abele & A. Thistle. 1984. Ecological Communities: Conceptual Is- sues and the Evidence. Princeton Univ, Press. Princeton, New Jersey. Uetz, G. 1976. Gradient analysis of spider com- munities in a streamside forest. Oecologia, 22: 373-385. Uetz, G. 1979. The influence of variation in litter habitats on spider communities. Oecologia, 42: 29-42. Uetz, G. 1991, Habitat structure and spider forag- ing. Pp. 325-348. In Habitat Structure: The Physical Arrangement of Objects in Space, (S. Bell, E. McCoy & H. Mushinsky, eds.). Chap- man and Hall, New York. Wise, D. 1993. Spiders in Ecological Webs. Cam- bridge University Press. Cambridge, England. Wissinger, S. 1997. Cyclic colonization in predict- ably ephemeral habitats: a template for biological control in annual crop systems. Biol. Conte, 10: 4-15. Young, O. & G. Edwards. 1990. Spiders in United States field crops and their potential effect on crop pests. J. ArachnoL, 18:1-27. Manuscript received 1 May 1998, revised 17 Sep- tember 1998. 1999. The Journal of Arachnology 27:351-362 THE IMPLICATIONS OF INTRAGUILD PREDATION FOR THE ROLE OF SPIDERS IN BIOLOGICAL CONTROL Margaret A. Hodge: Department of Biology, The College of Wooster, Wooster, OHIO 44491 USA ABSTRACT. Evidence is growing that spiders can be effective biological control agents, particulary assemblages of several species. Other evidence finds that spiders prey on each other and other generalist predators, and as such are of limited value in biological control. Such predatory interactions between species which use similar resources have been dubbed intraguild predation (IGP) due to their potential to modify competition as well as cause direct mortality. IGP interactions can have unexpected effects at other trophic levels, and sometimes result in enhancement of a pest population. In this paper I review the evidence for intraguild predation interactions involving spiders in natural systems, and other generalist predators in agroecosystems. To date not much research has examined whether such interactions influence spider biological control potential. Some suggestions as to how we might begin to address these issues are presented. Given their generalist arthropod diet and abundance in most terrestrial habitats, spiders likely inflict substantial mortality on insect populations. While the mechanisms by which spiders limit insect prey populations have been debated (Riechert & Lockley 1984; Wise 1993), it is generally agreed that they are im- portant in reducing insect numbers, and as such are of potential value in biological con- trol (Riechert & Lockley 1984; Nyffler & Benz 1987; Young & Edwards 1990; Wise 1993). Several studies have shown that assem- blages of many predator species may be more effective at controlling agricultural pests than single species augmentation (Chiverton 1986; Riechert & Bishop 1990; Clark et al. 1994; Provencher & Riechert 1994; Chang 1996; Riechert & Lawrence 1997). On the other hand, different species of predators and/or parasitoids may compete with or prey on each other, potentially reducing their biological control potential (Force 1974; Ehler & Hall 1982; Spiller 1986; Briggs 1993; Rosenheim et al. 1995; Chang 1996; Ferguson & Stiling 1996; Kester & Jackson 1996; Cisneros & Ro- senheim 1997; Rosenheim 1998). The nature of the diet of spiders suggests that they can prey on each other and other arthropod predators (Polis 1981; Jackson 1992; Wise 1993), as well as overlap in prey taxa consumed, thus potentially competing for resources. Diet overlap is one distinguishing feature of a guild, a group of sympatric taxa that use similar resources (Root 1967; Polis et al. 1989; Simberloff & Dayan 1991). Preda- tory interactions among members of the same guild are termed intraguild predation (IGP). This is distinguished from predation as tradi- tionally defined because, by eating a guild member, an individual not only directly gains energy and nutrients, but also reduces poten- tial competition for food (Polis et al 1989; Po- lis & Holt 1992). letraguild predation and cannibalism (killing and eating a member of the same species), may have profound effects on community structure (Polis 1981, 1988; Polis et al. 1989; Polis & Holt 1992). Given their ubiquity in terrestrial ecosystems, spiders are model organisms to investigate the occur- rence and consequences of IGP IGP: predation among potential compet- itors.— Intraguild predation and cannibalism have been shown to directly limit predator populations (Polis & McCormick 1986, 1987; Spiller & Schoener 1988; Wissinger 1989; Leonardsson 1991; Finke 1994; Wagner & Wise 1996; Wissenger et al. 1996). Since un- successful predation attempts represent ex- treme forms of interference competition (Polis et al. 1989; Elgar & Crespi 1992), IGP can also lead to behavioral adaptations to reduce mortality and conflict, resulting in habitat and diet shifts by IG prey (Fox 1975; Turner & Polis 1979; Doncaster 1992; Sih 1982; Eben- man & Persson 1988; Foster et al. 1988; Polis et al. 1989; Polis 1993; Dong & Polis 1992; 351 352 THE JOURNAL OF ARACHNOLOGY Holt & Polls 1997). These changes in foraging and habitat distribution may in turn have ef- fects at other trophic levels (Polls 1984; Wil- bur 1988). The traditional view of feeding relation- ships has been to assign species in a com- munity to a “trophic-level”, such as second- ary consumer (predator), primary consumer (herbivore), primary producer (plant), and so forth, with each level feeding on the former (Krohne 1998). Thus, in classic biological control, insect herbivore populations are re- duced by addition of predators, and this in turn reduces damage to crop plants (van den Bosch et al. 1982). In reality, however, ani- mals may feed from a variety of trophic lev- els, especially generalist predators, which take prey of whatever size they can handle (Polls 1988; Polls et al. 1989; Spence & Carcamo 1991; Dong & Polls 1992; Finke 1994). If these prey include younger conspecifics or other predators, then control of the herbivore population is not guaranteed. Various studies suggest that direct effects of one predator on another can indirectly affect a shared prey species by releasing it from intense predation or competition (Press et al. 1974; Pacala & Roughgarden 1984; Hurd & Eisenburg 1990; Polls & Holt 1992; Rosenheim et al. 1993; Wissinger & McGrady 1993; Wootton 1993; Cisneros & Rosenheim 1997; Fagan & Hurd 1994). If shared prey are herbivores then the indirect effects could cascade to plants, influ- encing primary productivity, an issue of ag- ricultural relevance. The purpose of this paper is to review the theory and empirical evidence relevant to the implications of IGP for the po- tential role of spiders in biological control of herbivorous pests in agriculture. IGP between spiders and other generalist predators. — Several studies of IGP in natural communities have uncovered direct and indi- rect interactions involving spiders. Polis & McCormick (1986, 1987) investigated a desert community of arachnids including spiders, solpugids and scorpions, all generalist preda- tors that use similar prey and prey on each other. Scorpions were continually removed from experimental plots, but not from control plots, and the relative abundances of the spi- ders and solpugids were tracked over time. At the end of the experiment (29 months), sig- nificantly more spiders occurred in the scor- pion removal plots than in the control plots. Two alternative hypotheses could explain these results: removal of the scorpions could have resulted in competitive release by the spiders in experimental plots, or IGP (by scor- pions) in the control plots may have reduced spider population size. There was no evidence of competitive release in that there were no differences between the experimental and con- trol plots in insect prey abundance or spider reproduction. Release from scorpion predation was the most likely cause of the increased numbers of spiders. Two independent studies on Anolis lizards examined evidence for intraguild predation on spiders cascading to populations of shared in- sect prey. Pacala & Roughgarden (1984) ma- nipulated anole densities in a Carribean forest and found a direct effect of lizards on forest floor arthropods, their primary prey, and an indirect effect on flying insects, the prey of orb- weaving spiders. Since anoles also prey on orb-weavers, the increase in flying insects on the high density lizard plots was thought to be due to intraguild predation by the lizards on the spiders. On Bahamian islands Spiller & Schoener (1990) also found a direct effect of lizards on spiders, but no indirect effect on flying insects. They did, however, observe that more spiders were feeding on lizard removal plots than on plots where they co-existed with lizards. The authors hypothesized that inter- ference competition or predation by lizards may displace spiders from prime web-sites, resulting in a reduction in prey capture for the spiders. Although the generalist diet of most spider species suggests that exploitative competition for food should be important (Marshall & Rypstra, this volume), experimental tests have found little evidence (Schaefer 1978; Wise 1981; Horton & Wise 1983; Riechert & Cady 1983; see Spiller 1984 a, b for an exception). In spider removal experiments to test for ex- ploitative competition among four genera of web-building spiders, Riechert & Cady (1983) not only found no competitive release, but on some of their removal plots they observed a negative effect of spider removals on the spe- cies remaining. They hypothesized that this may have been due to the fact that by remov- ing the other species of spiders, they may have been removing potential prey. Hodge & Marshall (1996) tested Riechert & Cady’s hypothesis that intraguild predation HODGE— IGP, SPIDERS & BIOLOGICAL CONTROL 353 masked competitive release in their system of web-building spiders on rock outcrops in Ten- nessee. After 12 weeks of removing each of three species from experimental plots we found that one of the species, Hypochilus tho- relli (Araneae, Hypochilidae) had lower body condition indices (indicating lower fecundity, Jakob et al. 1996) on spider removal plots as compared to control plots. This species was the major intraguild predator in the system, with spiders comprising over 40% of its diet (Riechert & Cady 1983; Hodge & Marshall 1996). These results support Riechert & Cady’s interpretation of the lack of competi- tive release in their study. Another species, Achaearanea tepidariorum (Araneae, Theri- diidae) exhibited greater spiderling popula- tions on rock outcrops from which the other two species had been removed as compared to control plots, suggesting that the manipu- lation removed predators (Hodge & Marshall 1996). The fact that this study found IGP-re- lated effects was striking given that the re- movals occurred over a relatively short time frame. Hurd & Eisenberg (1990) examined how interactions between praying mantises {Ten- odera sinesis) and wolf spiders (Lycosa (Ra- bidosd) rabida) affected overall arthropod numbers in a temperate early successional field habitat. They established four treatments, each of which enclosed a cubic meter of old- field vegetation in screen cages: mantids alone, wolf spiders alone, mantids and wolf spiders, and a control with neither predator added, and then sampled the number and bio- mass of other arthropods after 10 days. The ‘mantids alone’ enclosures had lower arthro- pod biomass than any of the other treatments. Examining the arthropods on a taxa-by-taxa basis revealed that in the ‘wolf spiders alone’ treatments there was a significant increase in the density of crickets as compared to the oth- er treatments. The explanation for this coun- ter-intuitive result illustrates the complexity of direct and indirect effects that can result from IGP interactions. The authors concluded that interactions between wolf-spiders (as evi- denced by some cannibalism) decreased their effectiveness as predators on crickets, and that in mantid/lycosid enclosures this effect was ameliorated because of mantid impact on spi- der numbers (an indirect effect of mantids on crickets). The presence of lycosids on the ground caused the crickets to move upward on vegetation (a direct effect of spiders on crickets); but when present with both mantids and lycosids, crickets were captured by man- tids hunting in the vegetation. Finally, lyco- sids may have consumed other cricket preda- tors (other spiders; an indirect effect of spiders on crickets); this was supported by the finding that there were significantly fewer heterospe- cific spiders in the lycosid enclosures. These results should be of interest to biological con- trol because mantids are often augmented, at least in small scale crop systems. On the other hand one may question the reality of this ex- periment since many predators were packed into small enclosures, whereas normally they could flee from one another. Enclosure effects were not a factor in the open-plot studies performed by Moran & Hurd (1994) in the same system. They added first instar mantids to 2 m X 2 m plots sepa- rated by 2 m wide barriers of black plastic sheeting which had a band of insect trapping compound painted down the middle to inter- cept arthropods leaving the plots. By compar- ing arthropods captured around mantid addi- tion plots to those captured around control (no mantids) plots, they discovered a behavioral response by spiders to the presence of elevat- ed mantid densities. Spiders dispersed from plots in which mantids were augmented. Smaller spiders (< 8 mm) are prey of the mantids whereas larger spiders (primarily wolf spiders) prey on the mantid nymphs and smaller spiders. Larger wolf spiders may have departed mantid addition plots because small- er spiders had dispersed. In this case the threat of IGP caused smaller spiders to leave, and scarcity of this IG prey caused the larger spi- ders to leave. Addition of supplemental food {Drosophila) reduced the tendency for spiders to emigrate from mantid augmented plots (Moran & Hurd 1997). This in turn increased IGP by spiders on mantids, as their numbers tended to decline in the food supplemented plots. As the authors point out, alternative prey does not always benefit generalist pred- ators if they can prey on one another. These authors extended their investigation to include the possibility of trophic cascades (Moran et al. 1996; Moran & Hurd 1998). Of interest was how a diverse plant community would respond in the context of an assem- blage of many predator and herbivore species. 354 THE JOURNAL OF ARACHNOLOGY In these experiments, control plots had no mantids, and experimental plots had natural densities of mantids. Cursorial spiders emi- grated from mantid plots throughout the course of the study (2 months) (Moran et al. 1996). Early emigration was probably due to the threat of IGP by mantids, whereas later emigration may have been a result of com- petition for prey, since mantid numbers at this point in time were too low to cause predator avoidance. Herbivore biomass was signifi- cantly lower and plant biomass was 30% high- er in the mantid addition plots by the end of the experiment. Mantids therefore caused a trophic cascade that extended to plants. These studies demonstrate that, contrary to theoretical predictions that interactions be- tween trophic levels in complex communities will be diffuse and buffer the intensity of re- sponses of any given species to another (Strong 1992), a single predator in a speciose natural assemblage can indeed initiate a tro- phic cascade. Perhaps this bodes well for po- tential predator influences on primary produc- tion in less speciose agroecosy terns: that is, despite the potential for IGP, strong interac- tions can cascade through trophic levels in such a way as to benefit crops. On the other hand, these strong interactions could be such that IGP interactions disrupt rather than en- hance the control of herbivore populations (Rosenheim et al. 1993; Rosenheim 1998). IGP in agroecosystems. — The fact that the self-limiting nature of spiders (via interfer- ence, territoriality or araneophagy) can de- crease their potential as biological control agents has been recognized (Reichert & Lock- ley 1984; Wise 1993), but rarely quantified. The bulk of my review will therefore cover experimental studies involving IGP among ar- thropod generalist predators other than spi- ders, since they have as yet not been well studied. Rosenheim et al. (1993) examined interac- tions between three species of predatory hem- iptera (Geocoris spp., Nabis spp., and Zelus renardii) and green lacewing larvae (Chryso- perla carnea (Neuroptera)), all of which eat aphid pests {Aphis gossypii) of cotton. To de- termine whether hemipteran predators exert mortality on lacewings, they caged cotton plants with aphids alone (control), and aphids with various combinations of the hemipterans. Lacewing survival was significantly reduced in the presence of bugs. To isolate the influ- ences of predation cotton plants were caged with a variety of combinations of predators: each hemipteran species alone or in combi- nation with lacewing larvae (with appropriate controls). Lacewing survival was significantly lower in the Z renardii and Nabis spp. treat- ments. Comparing aphid population growth among the single-predator species treatments, only in those cages with lacewing larvae alone was there a significant impact on aphids, sug- gesting that of all of these predator species, lacewings are the most effective at aphid con- trol. Given these results, it is not suprising that cages with lacewing larvae and Z. renardii or Nabis spp. exhibited a non-additive effect on aphid population control. Not only were the effects non-additive, but aphid populations ac- tually increased in these treatments. There- fore, predator interference generated a trophic cascade, increasing the abundance of herbi- vores. They also examined the effect that nymphal hemipterans can have on lace wing eggs. The presence of hemipterans reduced the propor- tion of lacewing eggs surviving to larval stag- es. Cisneros & Rosenheim (1997) examined the effect of predation by Z renardii of dif- ferent age-size classes on control of cotton aphid populations by lacewing larvae. Surviv- al of lacewing larvae was significantly lower in the presence of larger, older Zelus, and this produced a significant disruption of lacewing control of aphid populations. Observations of freely foraging bugs in the field showed an ontogenetic shift in foraging height and for- aging behavior resulting in higher encounter rates between Zelus adults and other predators (Cisneros & Rosenheim 1998). Another study of aphidophagous predators examined interactions between generalists and specialists and evaluated predator mobility as a potential factor influencing vulnerability to IGP (Lucas et al. 1998). The predators were lacewings {Chrysoperla rufilabris), spotted lady beetles {Coleomegilla maculata), both generalists, and larvae of the gall midge {Ap~ idoletes aphidimyza), a specialist on the shared prey, potato aphids (Macrosiphum eu- phorbiae). The lace wing and lady beetles are very active foragers as larvae and adults, whereas gall midge larvae are slow-moving predators. IGP interactions between all three predators HODGE— IGP, SPIDERS & BIOLOGICAL CONTROL 355 were investigated in the absence of aphid prey. Various combinations of predators at dif- ferent developmental stages (egg-adult) yield- ed 37 different test combinations. Symmetric IGP occurred between lacewings and lady beetles; that is, larger developmental stages of one predator fed on smaller developmental stages of the other. A few exceptions were ex- plained by behavioral and morphological dif- ferences between the predators. Third instar lacewings were able to prey on larger fourth instar lady beetles as well as adult beetles. It may be that a more aggressive hunting style and effective grasping mouthparts of lace- wings allow them to defy the general trend that the larger predator wins (Lucas et al. 1998). Interactions between both lacewings and lady beetles with gall midges were asym- metric: gall midges were almost never IG predators. This confirmed the authors’ predic- tion that more mobile predators have an ad- vantage over slow moving predators. In the presence of shared or extraguild prey (potato aphids) IGP was lower in several of the predator/life- stage combinations. Some of the IGP interactions persisted though, except when extraguild prey densities were very high. Based on the outcomes of IGP interac- tions between their different predators at var- ious levels of extraguild prey, Lucas et al. (1998) developed some general predictions as to the effect of extraguild prey and predator characteristics on the direction and outcome of IGP interactions. In cases where both pred- ators forage randomly, IGP will decrease steadily with increasing extraguild prey. Ran- dom search will, in this case, bring predators into contact with extraguild prey more often. When IGP interactions are risky for both pred- ators, IGP should decrease exponentially as extraguild prey increases in density. Abun- dance of alternative prey has similarly been observed to influence the tendency towards cannibalism in many animals (Elgar & Crespi 1992). In some cases IGP may remain con- stant despite increasing extraguild prey, es- pecially if IG prey are vulnerable, sessile and/ or aggregated. Finally, IGP may remain high at low extraguild prey densities, and only de- cline at very high extraguild prey densities. When extraguild prey are at low density, IG predators may benefit from removing potential competitors, whereas at high prey density this benefit disappears. Overall, the theme that unifies all of these predictions, and all of the experimental studies presented above, is that a detailed understanding of the ontogeny, be- havior and ecology of predators and prey is required to understand the role that IGP plays in the dynamics of complex communities, in- cluding agroecosystems. IGP & spiders in agroecosystems. — Though some research has been conducted evaluating the effectiveness of spiders as bio- control agents in agroecosystems (Riechert & Bishop 1990; Clark et al. 1994; Provencher & Riechert 1994; Carter & Rypstra 1995; Riech- ert & Lawrence 1997), there has been scant research on their potential interactions with other predators. Fagan et al. (1998) discovered an unpredicted interaction between IGP, pes- ticide appfication and biological control. They set out to examine the compatibility of insec- ticide-based and natural enemy-based pest control methods in tropical rice. Using open- top cages (to ameliorate enclosure effects) they established four treatments: insecticide added, wolf spiders added, both insecticide and wolf spiders added, nothing added. As would be predicted, rice pests were lower in the insecticide and wolf spider treatments, and each reduced pest densities to similar levels. The combination of insecticide and wolf spi- der addition, however, resulted in an increase in pests such that these enclosures were indis- tinguishable from the controls. They attribute these results to the additive impact of spiders and insecticide on predatory hemipterans (me- soveliids) which are also important biological control agents of rice pests. The combination spider-insecticide treatment lowered the den- sities of these alternative predators below the threshold of effective biological control. This study has important implications for integrat- ed pest management, and further illustrates the importance of a clear understanding of the role of IGP in agricultural systems. Given the general lack of experimental studies, what evidence (beyond Fagan et al. 1998) do we have that IGP involving spiders might be important in agroecosystems? Sev- eral studies have documented that spiders do engage in IGP interactions with other gener- alist predators, and many of these observa- tions come from crop systems (Table 1). These data were gleaned from tables in pri- mary research papers and from several re- views of spider diets by Nyffeler and col- 356 THE JOURNAL OF ARACHNOLOGY Table 1 . — A survey of the literature containing field observations of the spectrum of prey captured by a variety of spider species focusing on taxa that are potentially intraguild prey. The percent of the total observed diet for the majority of species in the list is obtained from a total number of observed prey exceeding 50. IG predator IG prey taxon % of diet Habitat Source Araneidae: Argiope bruennichi Araneae 3.3 Grassland Nyffeler 1982 Linyphiidae: Oedothorax insecticeps Araneae 16.3 Rice Kiritani et al. 1972 Theridiidae: Latrodectus mactans Soienopsis in- victa 75.3 Cotton Nyffeler et al. 1988 Achaearanea tepidario- Araneae 22 Rock outcrop Hodge & Marshall 1996 rum Lycosidae: Paradosa spp. Araneae 6.8 Winter Wheat Nyffeler & Benz 1988 Paradosa ramulosa Araneae 19.6 Alfalfa Yeargan 1975 Lycosa pseudoannulata Araneae 8.9 Rice Kiritani et al, 1972 Lycosa antelucana Lycosidae 4 Cotton Hayes & Lockley 1990 Lycosa antelucana Staphylinidae 6.7 Cotton Hayes & Lockley 1990 Lycosa antelucana Carabidae 10.9 Cotton Hayes & Lockley 1990 unspecified Araneae 19.2 Peanuts Agnew & Smith 1989 Pardosa lugubris Araneae 24 Forest Edgar 1969 Pardosa lugubris Araneae 34 Forest Hallander 1970 Pardosa amentata Araneae 11 ? Edgar 1970 Pardosa puilata Araneae 38 Meadow Hallander 1970 Pardosa purbeckensis Araneae 23 Salt Meadow Nyffeler & Benz 1988 Pardosa ramulosa Araneae 20 Alfalfa Yeargan 1975 Pardosa hokkaido Araneae 12 Forest Suwa 1986 Pirata piraticus Araneae 22 Salt Meadow Schaefer 1974 Pirata piraticus Araneae 28 River Bank Gettmann 1977, 1978 Pardosa agrestis Araneae 16 ? Nyffeler 1982 Lycosa osceola Araneae 27.5 Florida scrub Hodge, unpublished Lycosa pseudoceratiola Araneae 9 Florida scrub Hodge, unpublished Oxyopidae: Oxyopes salticus Araneae 15.9 Cotton Nyffeler et al. 1992 Oxyopes salticus Araneae 14.1 Cotton Nyffeler et al. 1987a Oxyopes salticus Araneae 9 Cotton Nyffeler & Sterling 1994 Peucetia viridans Araneae 40 Cotton Nyffeler et al. 1987b Peucetia viridans Hemiptera 8 Cotton Nyffeler et al. 1987b Peucetia viridans Neuroptera 8 Cotton Nyffeler et al. 1987b . Peucetia viridans Araneae 16 Croton Nyffeler et al. 1987b Peucetia viridans Araneae 13.3 Peanuts Agnew & Smith 1989 Peucetia viridans Araneae 7 shrubs Turner 1979 various Araneae 15 ? Nentwig 1986 Thomisidae: Xy Stic us spp. Araneae 6.4 Meadow: Plants Nyffeler 1982 Xysticus spp. Araneae 26 Meadow: Soil Sur= face Nyffeler 1982 Misumenops spp. Araneae 16.7 Peanuts Agnew & Smith 1989 HODGE— IGP, SPIDERS & BIOLOGICAL CONTROL 357 Table 1. — Continued % of IG predator IG prey taxon diet Habitat Source Salticidae: Phidippus audax Araneae 22.2 Phidippus audax Araneae 15.5 Phidippus johnsoni 27 various Araneae 20 Amaurobiidae Coras montanus Araneae 24 Hypochilidae Hypochilus thorelli Araneae 46 Pisauridae Pisaura mirabilis Araneae 18 Pholcidae Pholcus phalangiodes Araneae 6 Scytodidae Scytodes longipes Araneae 17.4 Unspecified Araneae Hemiptera 14.5 Araneae 17.3 Wild Plants & Cotton ? ? ? Dean et al. 1987; Nyffeler et al. 1994 Young 1989 Jackson 1977 Nentwig 1986 Rock outcrop Hodge & Marshall 1996 Rock outcrop Hodge & Marshall 1996 ? Nitzsche 1981 Cellars Nentwig 1983 outside buildings Peanuts Peanuts Nentwig 1985 Agnew & Smith 1989 Agnew & Smith 1989 leagues. Higher levels of IGP might have been reported in some cases if more specific taxo- nomic categories were used, for example, breaking insect orders into families which of- ten exhibit characteristic feeding habits (e.g., Carabidae rather than Coleoptera). Even so, it is not uncommon to find spider diets consist- ing of almost one-fifth IG prey (mean for Ta- ble 1 = 18.3% ± 12.7%). It is hard to form any general conclusions based on the data in Table 1 since the list is not comprehensive, and the methodology and intensity of data collection vary among stud- ies. The most striking feature, however, is the number of studies involving lycosids and ox- yopids in agricultural systems, and the some- times high percentage of IG prey reported from these spiders (e.g., 40% Araneae in the diet of Oxyopes salticus). It may be that cur- sorial spiders dominate as IG predators due to their active hunting style (Lucas et al.l998; Cisneros & Rosenheim 1998). It would be quite informative to have greater taxonomic resolution to the IG prey reported, to see if they are represented disproportionately by less mobile predators. This type of resolution would also suggest the types of direct and in- direct effects that might cascade to herbivores and crops. CONCLUSIONS How can we determine the implications of IGP for the role of spiders in agroecosystems? How does one begin to identify which of a suite of predators present in a particular crop have the potential for IGP interactions? Clas- sification systems exist for spider guilds (Uetz et ah, this volume), predator and herbivore guilds in crops (Breene et al. 1993) and struc- tural zones in crop plants which may support distinct suites of predator and prey species (e.g., LeSar & Unzicker 1978). Using these as a starting point, one can begin to define po- tential predator-predator and predator-herbi- vore interactions in which IGP may be of con- sequence. Quantification of the potential for IGP and/or competition should be achieved by careful study of the relative densities, habitat use, activity period and space, and diet. From these measures one can calculate indices of the opportunity for predation (lOP) and the opportunity for competition (IOC), as derived by Wissinger (1992) from pre-existing indices of resource overlap (Hurlbert 1978), Wissin- 358 THE JOURNAL OF ARACHNOLOGY ger’s indices allow for comparisons of the rel- ative strengths of predation, cannibalism, and resource competition between and within spe- cies by quantifying these interactions in the field and laboratory (Wissinger 1992). In a sense, they simply involve collecting the rel- evant natural history information about each species, and quantifying this information to make specific predictions of the relative im- portance of cannibalism, IGP and intra- or in- terspecific competition. This allows the design of more rigorous and meaningful field exper- iments (Wissinger 1992). Future field experiments should heed les- sons from the past regarding the use of enclo- sures and the duration of experiments. Stock- ing closed cages with predators may not reveal information relevant to the real world; and responses in the short term may lead to very different conclusions than might be reached from experiments of duration more similar to the actual seasonality of the partic- ular system (Wise 1993; Moran & Hurd 1994, 1998), and should be repeated across years to detect the effects of temporal variability (Polis et al. 1998). Despite the recent revival of interest in food web interactions, (“top-down” versus “bot- tom-up” effects ) and the complex nature of feeding relationships (Strong 1992; Polis 1994; Polis & Strong 1996; Polis & Wine- miller, 1996; Holt & Polis 1997), the scenario still generally used for biocontrol is that of a 3 -tiered system in which herbivores eat plants, and in turn are eaten by predators. As the studies reviewed in this paper demonstrate, animals do not recognize these artificial tro- phic boundaries, and often feed from several trophic levels. This can generate a complex array of direct and indirect effects which can have important and unexpected consequences for the effectiveness of generalist predators as biological control agents. The paucity of ex- perimental research on the potential web of IGP interactions involving spiders is surpris- ing since they are widely recognized as model organisms for the types of manipulative field studies used to investigate these interactions (Polis 1993; Wise 1993). Other generalist predators studied to date (hemiptera, lace- wings, beetles) are similar in nature to spiders in that they include animals with a both sit and wait and active foraging hunting styles, and also involve animals with distinct size classes, generating possibilities for both can- nibalism and intraguild predation between dif- ferent life stages of different predator species. Given the variety of crop systems, manage- ment practices (e.g., tillage versus no-tillage), and diverse predator and prey assemblages, agricultural systems provide models for in- vestigating the role of IGP from both pure and applied perspectives. ACKNOWLEDGMENTS I would like to thank Sam Marshall for help in finding references and for valuable discus- sions of early drafts; Gary Polis and the Spi- der Lab of Miami University for editorial in- put; Beth Jakob and Karen Cangialosi for moral support; Matt Greenstone and Keith Sunderland for organizing the symposium. LITERATURE CITED Agnew, C.W. & J.W. Smith, Jr. 1989. Ecology of spiders (Araneae) in a peanut agroecosystem. En- viron. Entomol., 18:30-42. Breene, R.G., D.A. Dean, M. Nyffeler & G.B. Ed- wards. 1993. Biology, predation ecology, and significance of spiders in Texas cotton ecosys- tems with a key to the species. Report B-1711, Texas Agric. Exp. Stn., College Station, Texas. Briggs, C.J. 1993. Competition among parasitoid species on a stage structured host and its effect on host suppression. American Nat., 141:372- 397. Carter, RE. & A.L. Rypstra. 1995. Top-down ef- fects in soybean agroecosystems: Spider density affects herbivore damage. Oikos, 72:433-439. Chang, Gary C. 1996. Comparison of single versus multiple species of generalist predators for bio- logical control. Environ. Entomol., 25:207-212. Chiverton, P.A. 1986. Predator density manipula- tion and its effects on populations of Rhopalo- siphum padi (Horn.: Aphididae) in spring barley. Ann. Appl. Biol., 109:49-60. Cisneros, J.J. & J.A. Rosenheim. 1997. Ontoge- netic change of prey preference in the generalist predator Zelus renardii and its influence on pred- ator-predator interactions. Ecol. Entomol., 22: 399-407. Cisneros, J.J. & J.A. Rosenheim. 1998. Changes in the foraging behavior, within-plant vertical dis- tribution, and microhabtat selection of a gener- alist insect predator: an age analysis. Environ. Entomol., 27:949-957. Clark, M.S., J.M. Luna, N.D. Stone, & R.R. Young- man. 1994. Generalist predator consumption of army worm (Lepidoptera: Noctuidae) and effect of predator removal on damage in no-till com. Environ. Entomol., 23:617-622. Dean, D.A., W.L. Sterling, M. Nyffeler, & R.G. HODGE~IGP, SPIDERS & BIOLOGICAL CONTROL 359 Breene. 1987. Foraging by selected spider pred- ators on the cotton fleahopper and other prey. Southwest. EntomoL, 12:263-270. Doncaster, C.R 1992. Testing the role of intraguild predation in regulating hedgehog populations. Proc. R. Soc. London, B, 249:113-117. Dong, Q. & G.A. Polls. 1992. The dynamics of cannibalistic populations: a foraging perspective. Pp. 13-37. In Cannibalism: Ecology and Evolu- tion Among Diverse Taxa. (M.A. Elgar & B.J. Crespi, eds.), Oxford Univ. Press. Ebenman, B. & L. Persson,eds. 1988. Size-struc- tured Populations. Springer- Verlag, New York. Edgar, W.D. 1969. Prey and predators of the wolf spider Lycosa lugubris. J. ZooL, 159:405-411. Edgar, W.D. 1970. Prey and feeding behaviour of adult females of the wolf spider Pardosa amen- tata (Clerck). Netherlands J. Zook, 20:487-491. Ehler, L.E. & R. W. Hall. 1982. Evidence for com- petitive exclusion of introduced natural enemies in biological control. Environ. EntomoL, 11:1-4. Elgar, M.A. & B.J. Crespi, eds. 1992. Cannibalism: Ecology and Evolution Among Diverse Taxa. Oxford Univ. Press. Fagan, W.E & L.E. Hurd. 1994. Hatch density var- iation of a generalist arthropod predator: popu- lation consequences and community impact. Ecology, 75:2022-2032. Fagan, W.E, A. Lukman Hakim, H. Ariawan & S. Yuliy antiningsih. 1998. Interactions between bi- ological control efforts and insecticide applica- tions in tropical rice agroecosystems: the poten- tial role of intraguild predation. Biol. Control, 13:121-126. Ferguson, K.L & P. Stiling. 1996. Non-additive ef- fects of multiple natural enemies on aphid pop- ulations. Oecologia, 108:375-379. Finke, O.M. 1994. Population regulation of a trop- ical damselfly in the larval stage by food limi- tation, cannibalism, intraguild predation and hab- itat drying. Oecologia, 100:118-127. Force, D.C. 1974. Ecology of insect host-parasitoid communities. Science, 184:624-632. Foster, S.A., V.B. Garcia, & M.Y. Town. 1988. Cannibalism as the cause of an ontogenetic shift in habitat use by fry of the threespine stickle- back. Oecologia, 74:577-585. Fox, L.R. 1975. Cannibalism in natural popula- tions. Ann. Rev. Ecol. Syst., 6:87-106. Gettmann, W.W 1977. Okologische Untersuchun- gen zum Beutefang und Analyse der Beutefan- ghandlung bei Wolfspinnen der Gattung Pirata (Araneae: Lycosidae). Ph.D. thesis, Univ. of Kai- serslautern, Germany. Gettmann, W.W. 1978. Untersuchungen aum Nah- rungsspektrum von Wolfspinnen (Lycosidae) der Gattung Pirata. Mitt, dtsch. Ges. allg. angew. EntomoL, Keil. Pp. 63-66. Hallender, H. 1970. Prey, cannibalism and micro- habitat selection in the wolf spiders Pardosa che- lata (O.E. Muller) and P. pullata (Clerck). Oi- kos, 21:337-340. Hayes, J.L. & T.C. Lockley. 1990. Prey and noc- turnal activity of wolf spiders (Araneae: Lyco- sidae) in cotton fields in the delta region of Mis- sissippi. Environ. EntomoL, 19:1512-1518. Hodge, M.A. & S.D. Marshall. 1996. An experi- mental analysis of intraguild predation among three genera of web-building spiders: Hypochi- lus, Coras, and Achaearanea (Araneae: Hypo- chilidae, Amaurobiidae and Theridiidae). J. Ar- achnoL, 24:101-110. Holt, R.D. & G.A. Polls. 1997. A theoretical framework for intraguild predation. American Nat., 149:745-764. Horton, C.C. & D.H. Wise. 1983. The experimen- tal analysis of competition between two syntopic species of orb- web spiders (Araneae: Araneidae). Ecology, 64:929-944. Hurd, L.E. & R.M. Eisenberg. 1990. Arthropod community responses to manipulation of a bi- trophic predator guild. Ecology, 71:2107-2114. Hurlbert, S.H. 1978. The measurement of niche overlap and some relatives. Ecology, 59:67-77. Jackson, R.R. 1977. Prey of the jumping spider Phidippus johnsoni (Araneae: Salticidae). J. Ar- achnoL, 5:145-149. Jackson, R.R. 1992. Eight-legged tricksters: spi- ders that specialize at catching other spiders. Bioscience, 42:590-598. Jakob, E.M., S.D. Marshall & G.W Uetz. 1996. Estimating fitness: a comparison of body condi- tion indices. Oikos, 77:61-67. Kester, K.M. & D.M. Jackson. 1996. When good bugs go bad: intraguild predation by Jalysus wickhami on the parasitoid, Cotesia congregata. EntomoL Exp. AppL, 81:271-276. Kiritani, K.S., S. Kawahara, T. Sasaba & E Naka- suji. 1972. Quantitative evaluation of predation by spiders on the green rice leafhopper, Nepho- tettix cincticeps Uhler, by a sight-count method. Res. Popul. EcoL, 13:187-200. Krohne, D.T 1998. General Ecology. Wadsworth Publ. Co., Belmont, California. Leonardsson, K. 1991. Effects of cannibalism and alternative prey on population dynamics of Sadu~ ria entomon (Isopoda). Ecology, 72:1273-1285. LeSar, Charles D. & John D. Unzicker. 1978. Soy- bean Spiders: Species Composition, Population Densities, and Vertical Distribution. Illinois Nat. Hist. Surv., 107. Lucas, E., D. Coderre. & J. Brodeur. 1998. Intra- guild predation among aphid predators: charac- terization and influence of extraguild prey den- sity. Ecology, 79:1084-1092. Marshall, S.D. & A.L. Rypstra. 1999. Spider com- petition in structurally simple ecosystems. J. Ar- achnoL, 27(l):343-350. 360 THE JOURNAL OF ARACHNOLOGY Moran, M.D. & L.E. Hurd. 1994. Short-term re- sponse to elevated predator densities: noncom- petitive intraguild interactions and behavior. Oecologia, 98:269-273. Moran, M.D. & L.E. Hurd. 1997. Relieving food limitation reduces survivorship of a generalist predator. Ecology, 78:1266-1270. Moran, M.D. & L.E. Hurd. 1998. A trophic cas- cade in a diverse arthropod community caused by a generalist arthropod predator. Oecologia, 113:126-132. Moran, M.D., T.R Rooney & L.E. Hurd. 1996. Top down cascade from a bitrophic predator in an old field community. Ecology, 77:2219-2227. Nentwig, W. 1983. The prey of web-building spi- ders compared with feeding experiments (Ara- neae: Araneidae, Linyphiidae, Pholcidae, Age- lenidae). Oecologia, 56:132-139. Nentwig, W. 1985. Feeding ecology of the tropical spitting spider Scytodes longipes (Araneae, Scy- todidae). Oecologia, 65:284-288. Nentwig, W. 1986. Non- web building spiders: prey specialists or generalists? Oecologia, 69:571- 576. Nitzsche, R. 1981. Beutefang und Brautgeschenk bei der Raubspinne Pisaura mirabilis (Cl.) (Ar- aneae: Pisauridae). Diplomarbeit, FB Biologie der Universitat Kaiserslautern. Nyffeler, M. 1982. Field Studies on the Ecological Role of the Spiders as Predators of Insects in Agroecosystems. Ph.D. thesis; Swiss Fed. Inst. Technol. (ETH), Zurich. Nyffeler, M & G. Benz. 1987. Spiders in natural pest control: a review. J. Appl. EntomoL, 103: 321-339. Nyffeler, M. & G. Benz. 1988. Feeding ecology and predatory importance of wolf spiders {Par- dosa spp.) (Araneae, Lycosidae) in winter wheat fields. J. Appl. EntomoL, 106:123-134. Nyffeler, M. & W.D. Sterling. 1994. Comparison of the feeding niche of polyphagous insectivores (Araneae) in a Texas cotton plantation: estimates of niche breadth and overlap. Environ. EntomoL, 23:1294-1303. Nyffeler, M., D.A. Dean. & W.D. Sterling. 1987a. Evaluation of the importance of the striped lynx spider, Oxyopes salticus (Araneae: Oxyopidae), as a predator in Texas cotton. Environ. EntomoL, 16:1114-1123. Nyffeler, M., D.A. Dean. & W.D. Sterling. 1987b. Predation by green lynx spider, Peucetia viridans (Araneae: Oxyopidae), inhabiting cotton and woolly croton plants in East Texas. Environ. En- tomoL, 16:355-359. Nyffeler, M., D.A. Dean. & W.D. Sterling. 1988. The southern black widow spider, Latrodectus mactans (Araneae: Theridiidae), as a predator of the red imported fire ant, Solenopsis invicta (Hy- menoptera: Formicidae), in Texas cotton fields. J. Appl. EntomoL, 106:52-57. Nyffeler, M., D.A. Dean. & W.D. Sterling. 1992. Diets, feeding specialization, and predatory role of two lynx spiders, Oxyopes salticus and Peu- cetia viridans (Araneae: Oxyopidae), in a Texas cotton agroecosystem. Environ. EntomoL, 21: 1457-1465. Nyffeler, M., W.D. Sterling. & D.A. Dean. 1994. How spiders make a living. Environ. EntomoL, 23:1357-1367. Pacala, S. & J. Roughgarden. 1984. Control of ar- thropod abundance by Anolis lizards on St. Eu- statius (NetL. Antilles). Oecologia, 64:160-162. Polls, G.A. 1981. The evolution and dynamics of intraspecific predation. Annu. Rev. Ecol. Syst., 12:225-251. Polls, G.A. 1984. Age structure component of niche width and intraspecific resource partition- ing; can age groups function as ecological spe- cies? American Nat., 123:541-564. Polls, G.A. 1988. Exploitation competition and the evolution of interference, cannibalism and intra- guild predation in age/size structured popula- tions. Pp. 185-202. In Size Structured Popula- tions: Ecology and Evolution. (B. Ebenman & L. Persson, eds.). Springer- Verlag, New York. Polls, G.A. 1993. Scorpions as model vehicles to advance theories of population and community ecology: the role of scorpions in desert commu- nities. Mem. Queensland Mus., 33:401-410. Polls, G.A. 1994. Food webs, trophic cascades and community structure. Australian J. ZooL, 19: 121-126. Polls, G.A. & R.D. Holt. 1992. Intraguild preda- tion: the dynamics of complex trophic interac- tions. Trends Ecol. EvoL, 7:151-155. Polls, G.A. & D.R. Strong. 1996. Food web com- plexity and community dynamics. American Nat., 147:813-842. Polls, G.A. & K. Winemiller, eds. 1996. Food Webs: Integration of Patterns and Dynamics. Chapman & Hall, New York. Polls, G.A. ,C.A. Myers & R.D. Holt. 1989. The ecology and evolution of intraguild predation: potential competitors that eat each other. Annu. Rev. Ecol. Syst., 20:297-330. Polls, G.A., S.D. Hurd, C.T. Jackson & E Sanchez- Pinero. 1998. Multifactor population limitation: variable spatial and temporal control of spiders on Gulf of California islands. Ecology, 79:490- 502. Polls, G.A. & S.J. McCormick. 1986. Scorpions, spiders and solpugids: predation and competition among distantly related taxa. Oecologia, 71:111- 116. Polls, G.A. & S.J. McCormick. 1987. Intraguild predation and competition among desert scoipi- ons. Ecology, 68:332-343. HODGE— IGP, SPIDERS & BIOLOGICAL CONTROL 361 Press, J., R. Flaherty, & R. Arbogast. 1974. Inter- actions among Plodia interpuncteiia, Bracon he- betor, and Xylocoris flavipes. Environ. EntomoL, 3:183-184. Provencher, L. & S.E. Riechert. 1994. Model and field test of prey control effects by spider assem- blages. Environ. EntomoL, 23:1-17. Riechert, S.E. & A.B. Cady. 1983. Patterns of re- source use and tests for competitive release in a spider community. Ecology, 64:899-913. Riechert, S.E. & T. Lockley. 1984. Spiders as bi- ological control agents. Ann. Rev. EntomoL, 29: 299-320. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in a garden test system. Ecology, 71:1441-1450. Riechert, S.E. & K. Lawrence. 1997. Test for pre- dation effects of single versus multiple species of generalist predators: spiders and their insect prey. EntomoL Exp. AppL, 84:147-155. Root, R. 1967. The niche exploitation pattern of the blue-grey gnat catcher. Ecol. Monogr., 37: 317-350. Rosenheim, J.A. 1998. Higher-order predators and the regulation of insect herbivore populations. Annu. Rev. EntomoL, 43:421-447. Rosenheim, J.A., H.K. Kaya, L.E. Ehler, J.J. Ma- rois, & B.A. Jaffee. 1995. Intraguild predation among biological control agents: theory and ev- idence. Biol. Control, 5:303-335. Rosenheim, J.A., L.R. Wilhoit & C.A. Armer. 1993. Influence of intraguild predation among generalist insect predators on the suppression of an herbivore population. Oecologia, 96:439-449. Schaefer, M. 1974. Experimentelle Untersuchun- gen zur Bedeutung der interspezifischen Konk- urrenz bei 3 Wolfspinnen-Arten (Araneida: Ly- cosidae) einer Salzwiese. ZooL Jb., Abt. Syst., 101:213-235. Schaefer, M. 1978. Some experiments on the reg- ulation of population density in the spider Flo- ronia buccelenta (Araneida: Linyphiidae). Symp. ZooL Soc, London, 42:203-210. Sih, A. 1982. Foraging strategies and the avoid- ance of predation by an aquatic insect, Notonecta hojfmanni. Ecology, 63:786-796. Simberloff, D. & T. Dayan. 1991. The guild con- cept and the structure of ecological communities. Annu. Rev. EcoL Syst,. 22:115-143. Spence, J.R. & H.A. Carcamo. 1991. Effects of cannibalism and intraguild predation on pond- skaters (Gerridae). Oikos, 62:333-341. Spiller, D.A. 1984a. Competition between two spi- der species: experimental field study. Ecology, 65:909-919. Spiller, D.A. 1984b. Seasonal reversal of compet- itive advantage between two spider species. Oec- ologia, 64:322-331. Spiller, D.A. 1986. Interspecific competition be- tween spiders and its relevance to biological con- trol by general predators. Environ. EntomoL, 15: 177-181. Spiller, D.A. & TW. Schoener. 1988. An experi- mental study of the effects of lizards on web- spider communities. Ecol. Monogr., 58:57-77. Spiller, D.A. & T.W. Schoener, 1990. Lizards re- duce food consumption by spiders: mechanisms and consequences. Oecologia, 83:150-161. Strong, D.R. 1992. Are trophic cascades all wet? Differentiation and donor-control in speciose ecosystems. Ecology, 75:747-754. Suwa, M. 1986. Space partitioning among the wolf spider Pardosa amentata species group in Hok- kaido, Japan. Res. Popul. Ecol., 28:231-251. Turner, M. 1979. Diet and feeding phenology of the green lynx spider, Peucetia viridans (Ara- neae: Oxyopidae). J. ArachnoL, 7:149-154. Turner, M. & G. Polls. 1979. Patterns of co-exis- tence in a guild of raptorial spiders. J. Anim. Ecol., 48:509-520. Uetz, G.W., J. Halaj & A.B. Cady. 1999. Guild structure of spiders in major crops. J. ArachnoL 27(l):270-280. van den Bosch, R., P.S. Messenger & A.P. Gutier- rez. 1982. An Introduction to Biological Con- trol. Plenum Press, New York. Wagner, J.D. & D.H. Wise. 1996. Cannibalism reg- ulates densities of young wolf spiders: evidence from field and laboratory experiments. Ecology, 77:639-652. Wilbur, H. 1988. Interactions between growing predators and growing prey. Pp. 157-172. In Size-Structured Populations. (B. Ebenman & L. Persson, eds.) Springer- Verlag, New York, N.Y. Wise, D.H. 1981. Inter- and intraspecific effects of density manipulations upon females of two orb- weaving spiders (Araneae: Araneidae). Oecolo- gia, 48:252-256. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge University Press. Wissinger, S.A. 1989. Seasonal variation in the in- tensity of competition and predation among dragonfly larvae. Ecology, 70:1017-1027. Wissinger, S.A. 1992. Niche overlap and the po- tential for competition and intraguild predation between size- structured populations. Ecology, 73:1431-1444. Wissinger, S.A. & J. McGrady. 1993. Intraguild predation and competition between larval drag- onflies: direct and indirect effects on shared prey. Ecology, 74:207-218. Wissinger, S.A., G.B. Sparks, G.L. Rouse, W.S. Brown, & H. Steltzer. 1996. Intraguild predation and cannibalism among larvae of detrivorous caddisflies in subalpine wetlands. Ecology, 77: 2421-2430. Wootton, J.T. 1993. Indirect effects and habitat use in an intertidal community: interaction chains 362 THE JOURNAL OF ARACHNOLOGY and interaction modifications. American Nat., 141:71-89. Yeargan, K.V. 1975. Prey and periodicity of Par- dosa ramulosa (McCook) in alfalfa. Environ. En- tomoL, 4:137-141. Young, O.P. 1989. Field observations of predation by Phidippus audax (Araneae: Salticidae) on ar- thropods associated with cotton. J. Entomol. Sci., 24:266-273. Young, O.P. & G.B. Edwards. 1990. Spiders in United States field crops and their potential effect on crop pests. J. ArachnoL, 18:1-27. Manuscript received 1 May 1998, revised 25 No- vember 1998. 1999. The Journal of Arachnology 27:363-370 SPIDERS IN DECOMPOSITION FOOD WEBS OF AGROECOSYSTEMS: THEORY AND EVIDENCE David H. Wise, William E. Snyder and Patchanee Tuntibunpakul; Department of Entomology, University of Kentucky, Lexington, Kentucky 40546 USA Juraj Halaj*: Department of Zoology, Miami University, Oxford, Ohio 45056 USA ABSTRACT. The involvement of spiders in decomposition food webs has the potential to affect agri- cultural productivity through two quite different types of interactions: (1) cascading, top-down effects of spider predation on rates of nutrient mineralization — spider-initiated trophic cascades in the detrital food web that could alter rates of decomposition and release of nutrients to plants; and (2) a bottom-up linkage, through spiders, between decomposition and grazing food webs — energy from the detrital web contributing to elevated spider densities, which in turn might reduce pests and enhance net primary production. Scant experimental evidence exists to refute or support either hypothesis. The first set of interactions is most likely to be of significance in no-till and conservation tillage farming. In theory, spiders have the potential to enhance productivity by increasing rates of mineralization, but theory also predicts that spiders, by preying on important detritivores and fungivores, depress rates of litter decomposition. Field experiments by Kajak and her colleagues have uncovered such negative effects of spiders in mown pastures. Although this negative effect could reduce plant growth, the expected time lags in most types of crops suggest that the overall impact of spiders on plant production will be determined more by the interactions comprising the second hypothesis. However, the later hypothesis, that bottom-up control processes in the decompo- sition web affect crop productivity via energy subsidies to spiders and other generalist predators in the grazing web, remains conjecture without clear experimental confirmation. This hypothesis should be tested in agroecosystems in which detritus-based food webs can feasibly be manipulated. A major goal of agriculture is to maximize net primary production, which is the ultimate source of energy for both grazing and decom- position food webs. Biocontrol practitioners have focused on the grazing web because ag- ricultural pests and humans compete directly for the living products of photosynthesis. Thus, research on the roles of spiders in agroecosystems has focused primarily on the extent to which these predators suppress den- sities of grazing herbivores. Spiders also be- long to decomposition food webs of agroeco- systems, yet arachnid connections to such webs have stressed acarine cousins; and spi- ders have been largely ignored. Is this neglect justified, or might knowledge of how spiders function in decomposition food webs be uti- lized to increase agricultural yields? Two quite different sets of interactions be- tween spiders and the detritus-based food web are potentially relevant. The first set involves ’ Current address: Department of Entomology, University of Kentucky, Lexington, Kentucky 40546 USA cascading, top-down effects of spider preda- tion on rates of nutrient mineralization— spi- der-initiated trophic cascades in the detrital food web that could affect rates of decompo- sition and release of nutrients to plants. The second set of interactions relies on a linkage, through spiders, between decomposition and grazing food webs — energy from the detrital web contributing to elevated spider densities, which in turn might cause lower pest numbers and enhanced net primary production. Here we examine the evidence that these two dif- ferent sets of interactions affect, or have the potential to influence, agricultural productivi- ty. TOP-DOWN TROPHIC CASCADES AND RATES OF DECOMPOSITION Theory. — Decomposition food webs are often categorized as “donor-controlled” sys- tems (Pimm 1982) because the rate at which detritus is consumed does not immediately in- fluence the rate of supply of this energy to the system, but the amount of detrital input can influence densities of detritivores and their 363 364 THE JOURNAL OF ARACHNOLOGY predators. Viewing detrital food webs as pri- marily donor-controlled leads to the conclu- sion that bottom-up control processes predom- inate in such webs, i.e., that most linkages of indirect effects are responses to changes in rates of input at the base of the food web. This view is an over-simplification for at least two reasons. First, the rate at which the primary decomposers — the bacteria and fungi, collec- tively known as the microflora — decompose detritus depends on their population growth rates, which in turn are potentially influenced by their natural enemies (Swift et al. 1979). The second oversimplification is that, ulti- mately, decomposition processes affect net primary production by altering rates of min- eralization, i.e., the rates at which nutrients locked in detritus become available to auto- trophs. As a consequence, spiders that feed on detritivores have the potential to influence in- directly the growth of autotrophs by generat- ing trophic cascades in the decomposition food web, thus affecting rates at which nitro- gen and other nutrients required by plants en- ter the nutrient pool. Such linked indirect effects most likely oc- cur in those agroecosystems in which a sub- stantial fraction of litter decomposition occurs on the soil surface, and in which the major consumers of detritus and the microflora are preyed upon by spiders. Ploughing disrupts the litter layer and distributes organic matter throughout the soil; thus, cultivation encour- ages below-ground, bacterial-based decom- position food webs (Hendrix et al. 1986). The role of spiders in such webs is likely insig- nificant. Under no-till and conservation tillage conditions, however, litter accumulates above ground, fungi are a major component of the microflora, and microarthropods, primarily mites and Collembola (springtails), are major detritivores/fungivores (Hendrix et al. 1986; Stinner & House 1990; Robertson et al. 1994). Spiders readily consume Collembola, which constitute a substantial portion of the diet of many species (Hallander 1970; Schaefer 1975; Yeargan 1975; Wingerden 1975, 1978; Green- stone 1980, 1983; Nentwig 1987; Dobel & Denno 1994; Nyffeler et al. 1994). Hence spi- ders are most likely to exert a trophic cascade affecting mineralization in no-till annual crops; and in orchards, pastures and other pe- rennial crops Spiders have the potential to generate either positive or negative impacts on rates of min- eralization, even if one ignores interactions between spiders and other predators. The sim- plest abstract food chain in a spider-influenced system would consist of three effective tro- phic levels: detritus, Collembola (realizing that other detritivores play roles similar to Collembola, but Collembola appear to be ubiquitous and abundant), and spiders. The in- direct effects of spiders in such a food chain would be to enhance the standing crop of de- tritus, i.e., to retard the rate of litter decom- position (Fig. lA). This model, however, is greatly oversimplified. Fungi play a critical role in decomposing plant litter; and although Collembola consume plant material, many Collembola species are primarily fungivorous (Peterson 1971; Chen et al. 1996). Thus a four-level food chain model is more realistic (Fig. IB). In grazing food chains of four tro- phic levels, predators can induce a trophic cascade that negatively impacts the base tro- phic level — the primary producers. Reasoning by analogy, one might predict that an increase in spider density should lead to a decrease in the amount of detritus, i.e., an increase in de- composition rate, by relieving predation pres- sure on fungi (Fig. IB). However, does an in- crease in fungal biomass always lead to increased rates of mineralization? Not neces- sarily, because nutrients can become immo- bilized in senescent fungal hyphae. Thus, Col- lembola at intermediate densities can enhance rates of mineralization by consuming senes- cent fungal hyphae (Parkinson et al. 1977; van der Drift & Jansen 1977; Wamock et al. 1982; Finlay 1985; Verhoef & de Goede 1985; Vis- ser 1985). Collembola also enhance decom- position by more indirect pathways, i.e., by comminuting the litter (Anderson et al. 1984). Therefore, depression of Collembola popula- tions by spiders could negatively impact rates of litter decomposition and mineralization (Fig. 1C). It is clear that the potential relationship be- tween spider densities and mineralization rates is complicated because links between Collem- bola density and rates of litter decomposition are complex. Field experiments in which spi- der densities are reduced is the most direct way to determine which of the competing hy- potheses about possible spider-induced trophic cascades is correct. Evidence. — Kajak and her colleagues have WISE ET AL.— ROLES OF SPIDERS IN DETRITAL FOOD WEBS 365 (A) Spiders i Collemboia o Detritus ° + o (B) Spiders i Collemboia Y' Detritus (C) Spiders 1 r 1- 1 Collemboia o A 1 [■ 1 Fungi |. ' > - r ; -/+ o 1 > Detritus o ( > < 3 O + energy flow interaction coefficient Figure 1. — Three different hypothesized sets of trophic cascades in which predation by spiders on Collemboia might either enhance or retard rates of litter decomposition. (A) A 3-level food chain in which indirect effects of predation by spiders retard decomposition, indicated by a net 4- influence of spiders on the amount of detritus [(— ) X (— ) = ( + )]. (B) A simple food chain of 4 trophic levels in which spider predation enhances litter decomposition, i.e., decreases litter amount [( — ) X (— ) X (— ) = ( — )]. (C) A more realistic food chain model, in which indirect effects of spider predation are predicted to retard decomposition rate. This model incorporates the positive non-trophic effects of Collemboia feeding, such as litter comminution, on rates of fungal growth; and the fact that the relationship between fungal density and rates of fungal breakdown is non-linear. studied the impact of spiders and other gen- eralist arthropod predators — -primarily carabid beetles — on rates of litter decomposition in mown meadows (Kajak & Jakubczyk 1975, 1976 1977; Kajak & Kaczmarek 1988; Kajak et al. 1991; Kajak 1997). Their basic approach has been to exclude epigeic macrofauna from small field cages, and then compare trapping rates of predators; densities of Collemboia, other fungivores/detritivores, and the micro- flora; and rates of litter decomposition, with corresponding rates and densities in partially open cages or open areas. In some of their studies not all differences were statistically significant, and some varia- tion occurred between experiments conducted in different meadows and different years. Nevertheless, a clear pattern emerges. Exclud- ing macrofaunal predators, of which spiders comprise a substantial fraction, frequently causes an increase in Collemboia and other mesofauna and an increase in rates of litter decomposition. Thus ambient densities of spi- ders and other macrofaunal predators inhibit rates of decomposition in grasslands. These studies were conducted in a managed system in which the diversity of vegetation, fauna and microflora probably is more similar to that of more natural ecosystems than to highly disturbed crop systems. To date avail- able evidence supports the hypothesis that spi- ders inhibit rates of decomposition and min- eralization (Fig. 1C); and evidence does not support the alternative hypothesis, that spiders increase crop productivity through trophic cascades within the decomposition food web. Clearly more experiments of the type con- ducted by Kajak and her colleagues are need- ed in a variety of agroecosystems. DETRITAL SUBSIDY OF THE GRAZING FOOD WEB Theory. — Spiders, because they prey upon both detritivores/fungivores and herbivores, simultaneously belong to decomposition and grazing food webs. The connection of spiders to both webs opens the possibility that in- creased input of detritus to the agroecosystem could enhance detritivorous and fungivorous prey of spiders, leading to elevated spider 366 THE JOURNAL OF ARACHNOLOGY I Detrital input I Primary Production Figure 2. — If decomposition and grazing food webs are linked by common top predators, it is pos- sible that increased input of detritus could elevate the biomass of primary producers through a com- plex linkage of trophic interactions (i.e. apparent competition between detritivores and grazing her- bivores). Because in this model the predators are feeding on more than one trophic level and there is an external input of energy to the grazing food chain, this type of effect on the grazing food web has been termed an allochthonous energy subsidy via multichannel omnivory (Polis 1994; Polis & Strong 1996). densities which in turn would increase pres- sure on herbivores, thereby enhancing rates of net primary production. In this way increased input of detritus could, in principle, make spi- ders more effective biocontrol agents by in- troducing an energy subsidy to the grazing food web. This complex series of indirect ef- fects involving bottom-up control processes in one food web and top-down processes in an- other is an example of an allochthonous en- ergy subsidy via multichannel omnivory (Polis 1994; Polis & Strong 1996; Fig 2). An ex- ample from eon- agricultural systems is the in- put of detrital energy from the marine envi- ronment that leads to elevated densities of spiders and other predators in coastal areas, with subsequent declines in herbivores and re- duced plant damage (Polis & Hurd 1996). Food-web interconnections in tropical rice fields, in which detritus is both alio- and au- tochthonous, constitute a probable example from agricultural systems. Settle et al. (1996) argue that abundant populations of detritivores and planktivores early in the season maintain high densities of generalist predators, which are then able to suppress pest populations on rice later in the season. Trophic cascades induced by spiders in a grazing food web via energy subsidies from a decomposition web will occur only if certain conditions are met: (1) Bottom-up control pro- cesses in the decomposition web must affect spider densities; i.e., spiders must be food- limited, and fungivores/detritivores must be readily consumed by spiders if numbers of the former increase; (2) Increased availability of prey in the decomposition web must not cause spiders to switch from feeding on herbivorous prey, or if such changes occur, the increased numbers of spiders must compensate for this dietary shift; (3) Spider-generated trophic cas- cades in the grazing food web must be strong enough to enhance crop yield. No direct experimental evidence exists to support the hypothesis that energy subsidies from decomposition food webs enhance crop yield via multichannel omnivory in agroeco- systems. In the absence of direct evidence, the only alternative is to evaluate the evidence that the proposed conditions for such a detrital subsidy are satisfied. Below we summarize briefly some of this evidence, which comes from a mixture of studies with noe-agricul- tural and agricultural systems. Evideece.=—Field experiments have dem- onstrated that spider populations frequently are food-limited (Wise 1993). Although most studies have been conducted with grazing food webs, some experiments have demon- strated bottom-up limitation of spider numbers in decomposition food webs in non-agricui- tural systems (Spiller 1992; Chen & Wise 1999). The importance of Collembola num- bers in influencing spider population dynam- ics is particularly crucial to the argument. Ex- perimentally enhancing detrital input to the forest-floor increases densities of Collembola and other fuegivores, which is accompanied by a doubling of densities of many families of spiders (Chen & Wise 1999). Indirect evidence suggests that Collembola help maintain high densities of spiders and other generalist predators in some agroecosys- tems. For example, two species of wolf spi- ders in Swiss wheat fields feed predominantly on Collembola (> 35% of their diet) early in the season. As the season progresses, how- ever, Lepidoptera larvae and cereal aphids gradually become the spiders’ main prey WISE ET AL.— ROLES OF SPIDERS IN DETRITAL FOOD WEBS 367 (Nyffeler 1982). Field experiments have dem- onstrated that carabids and spiders can limit aphid numbers in cereal crops (Chiverton 1986; Edwards et al. 1979). Although this paL tern supports the hypothesis that a detrital sub- sidy promotes top-down control in the grazing food web, the relationship between decom- position and grazing food webs in cereal crops is likely complex. In addition to subsidizing spider populations and inducing trophic cas- cades, high numbers of alternative prey from the decomposition web might weaken such a cascade because some aphid species are low- quality food for spiders (Toft pers. comm.; Toft 1995); and spiders might shift from aphids to higher-quality alternatives. However, although spiders can develop aversions to low-quality prey, limited consumption of sub- optimal food provides nutritional benefits to these predators (Toft 1995). Hence spider pre- dation on aphids at low prey levels, common early in the season, may limit aphid outbreaks if the presence of superior-quality prey boosts spider numbers enough to ensure that the overall impact of spider predation on aphids is high (Toft 1995). An additional complicat- ing issue is the finding that not all Collembola are high-quality prey for spiders (Toft & Wise 1999a, b). Clearly more information is needed before accurate predictions can be made about the general impact of increased densities of Collembola and other fungivores/detritivores on the total predation pressure exerted by spi- ders on agricultural pests. Increasing evidence suggests that spiders can depress densities of agricultural pests in several types of crop; however, effects on crop yield have not been widely documented. Enough evidence exists to implicate spiders as potentially important biocontrol agents to jus- tify research into the question of whether or not a detrital subsidy could increase their ef- fectiveness. No experimental evidence exists to support the prediction that such a detrital subsidy could be used to enhance the biocon- trol effectiveness of spiders. Experimental studies are needed with agroecosystems in which a detrital subsidy is likely to have an impact. Below we discuss one such system. Vegetable crops.— Field experiments of Riechert & Bishop (1990) have revealed that spiders can limit densities of pest insects in vegetable gardens. Their use of fencing to ex- clude spiders could have also reduced carabid beetles, which are abundant predators in agroecosystems. We have recently expanded their experiments to examine explicitly the combined impacts of spiders and carabids, their separate impacts, and whether or not in- traguild predation limits their effectiveness in biocontrol. In one set of experiments we employed fence barriers, pitfall trapping and hand re- moval to reduce densities of ground spiders, foliage spiders and carabid beetles (Tuetibun- pakul & Wise unpubl. data). In the first ex- periment, conducted in mixed-vegetable gar- dens, reducing spiders and carabids led to elevated densities of squash bugs in cucum- bers and Colorado potato beetles (CPB) in po- tatoes. The presence of beetles and spiders marginally increased total cucumber yield, significantly improved the individual weight of marketable cucumbers, and marginally in- creased the yield of one of two varieties of potato. In an experiment conducted the fol- lowing year, manipulating spiders and cara- bids in a garden planted solely in potatoes had no impact on densities of CPB, which were higher than the previous year, and did not af- fect potato production. In another set of experiments, densities of wolf spiders and carabids were manipulated together and separately by continuously alter- ing rates of immigration into fenced plots in conjunction with removal by pitfall trapping (Snyder & Wise in press). In the first year, simultaneously decreasing lycosids and cara- bids had no impact on pests or yield of cu- cumber in a spring garden; however, reducing colonization by spiders and carabids of a sum- mer squash garden harmed squash production (Snyder & Wise in press). In the second year (unpubl. data) we manipulated immigration rates of lycosids and carabids both singly and together in order to separate their impact as biocontrol agents and to uncover effects of in- traguild predation on their total biocontrol ef- fectiveness. In the spring garden carabids did not affect cucumber production, but in the fall garden of squash they reduced squash bugs and increased fruit production, Lycosids in- creased cucumber production by feeding on striped cucumber beetles. In marked contrast, in the summer garden wolf spiders harmed squash production by causing an increase in squash bugs at the critical early stage of squash growth. This indirect negative effect of 368 THE JOURNAL OF ARACHNOLOGY lycosids likely was caused by their feeding on important insect predators of squash bug nymphs. In the summer gardens, allowing ca- rabids to immigrate into the plots compensat- ed for the negative impact of lycosids on yield, so that the combined effect of the as- semblage of wolf spiders and carabid beetles was to increase squash production. Thus spiders and carabids have the poten- tial to depress pest numbers and increase yield in potatoes and cucurbits, but the pattern is complex. Densities of immigrating spiders and carabid beetles may not always be high enough either to reduce pest numbers or to lower them enough to improve yield. Further- more, lycosids can substantially reduce crop production by preying on other predators of insect pests. Whether or not the enhancement of spider numbers will improve vegetable yield depends upon both the densities and phenologies of pests and other predators. Enhancement of spider numbers, and the provision of alternative prey to reduce their feeding on other predators of insect pests, has the potential to increase the impact of spiders in vegetable systems. Providing straw mulch increases the density of spiders ca. lOX com- pared to plots with bare ground (Riechert & Bishop 1990; Tuntibunpakul & Wise unpubl. data). Similarly, the use of straw refugia in soybeans can dramatically elevate the abun- dance (20-30 X) and diversity of resident spi- der fauna (Halaj, Cady & Uetz unpubl. data). Such effects could be due primarily to altered structure of the physical habitat; however, some of the increase in spider numbers could have resulted from an energy subsidy derived from the decomposition food web based on the added straw. This effect could be partic- ularly important later in the season, after the mulch has partially decomposed, enhancing detritivore populations. It would be worthwhile to devise field ex- periments in which the rate of input of detritus to the decomposition food web of vegetable gardens is explicitly manipulated in order to test the hypothesis that a detrital subsidy can enhance the biocontrol impact of spiders (Fig. 3). Increased detritus could improve plant pro- ductivity simply by providing more nutrients, so a complete test of the hypothesis would require documentation that the detrital subsidy does in fact cause higher spider numbers and higher densities of Collembola and other fun- \ Detrital Input ^ | Vegetable Production Figure 3. — Hypothesized increase in vegetable production via a detrital energy subsidy of a grazing food web in which spiders are top generalist pred- ators. givores, and that increased productivity of marketable vegetables can be explained as the result of decreased feeding by insect pests. The impact of a detrital subsidy could be in- vestigated most directly by experimentally in- troducing high quality detritus to the system (e.g., Chen & Wise 1997, 1999). Support of the detrital- sub sidy hypothesis would then justify detailed studies of how farnfing tech- niques could be modified to encourage the de- composition food web in vegetable produc- tion. CONCLUSIONS In theory, spiders have the potential to en- hance productivity by increasing rates of min- eralization through cascading top-down ef- fects on rates of decomposition. Such cascading indirect effects are most likely to be of significance in no-till and conservation tillage farming. The actual impact of spiders on mineralization rates is difficult to predict. It is possible that spider predation also could negatively impact the rate of litter decompo- sition, as has been demonstrated for grassland systems. Such an effect might lower rates of plant production, but the expected time lags before such an effect would be expressed sug- gest that the overall impact of spiders on plant production will be determined more by their interactions in the grazing food web. It is more likely that bottom-up control pro- cesses in the decomposition web can affect crop productivity via multichannel omnivory. Energy subsidies from the decomposition sub- system may cause spider-induced trophic cas- WISE ET AL.— ROLES OF SPIDERS IN DETRITAL FOOD WEBS 369 cades that are strong enough to increase crop production. This conjecture remains a hypoth- esis, but well worth testing in agroecosystems in which detritus-based food webs can feasi- bly be enhanced. ACKNOWLEDGMENTS Unpublished results referred to in this paper are from studies supported by EPA Grant G71A0056 (DHW), NSF DDI Grant DEB- 9701180 (WES and DHW), NSF Training Grant DGE-9355093 (DHW and WES), USDA/Kentucky Agricultural Experiment Station Hatch Project KY-008005 (DHW), a graduate fellowship from the Thai Institute for the Promotion of Teaching Science and Tech- nology (PT), and a State of Ohio Postdoctoral Academic Challenge Fellowship (JH). This is publication #98-08-174 of the Kentucky Ag- ricultural Experiment Station. LITERATURE CITED Anderson, J.M., A.D.M. Rayner & D.WH. Walton (eds). 1984. Invertebrate-microbial interactions. Cambridge Univ. Press, New York. Chen, B. & D.H. Wise. 1997. Responses of forest- floor fungivores to experimental food enhance- ment. Pedobiologia, 41:316-326. Chen, B. & D.H. Wise. 1999. Bottom-up limitation of predaceous arthropods in a detritus-based ter- restrial food web. Ecology, 80:761-772. Chen, B., R.J. Snider & R.M. Snider. 1996. Food consumption by Collembola from northern Michigan deciduous forest. Pedobiologia, 40: 149-161. Chiverton, P.A. 1986. Predator density manipula- tion and its effects on populations of Rhopalo- siphum padi (Horn.: Aphididae) in spring barley. Ann. Appl. Biol., 109:109:49-60. Dobel, H.G. & R.F. Denno. 1994. Predator-plant- hopper interactions. Pp. 325-399. In Planthop- pers: Their Ecology and Management. (R.F. Den- no & T.J. Perfect, eds.). New York: Chapman and Hall. Drift, van der, J. & E. Jansen. 1977. Grazing of springtails on hyphal mats and its influence on fungal growth and respiration. Ecol. Bull., (Stockholm), 25:203-209. Edwards, C.A., K.D. Sunderland & K.S. George. 1979. Studies on polyphagous predators of ce- real aphids. J. Appl. Ecol., 16:811-823. Finlay, R.D. 1985. Interactions between soil micro- arthropods and endomycorrhizal associations of higher plants. Pp. 319-331. In Ecological Inter- actions in Soil. Plants, Microbes and Animals. (A.H. Fitter, D. Atkinson, D.J. Read & M.B. Usher, eds). Boston: Blackwell Sci. Publ. Greenstone, M.H. 1980. Contiguous allotopy of Pardosa ramulosa and Pardosa tuoba (Araneae: Lycosidae) in the San Francisco Bay Region, and its implications for patterns of resource partition- ing in the genus. American Midi, Nat., 104:305- 311. Greenstone, M.H. 1983. Site-specificity and site te- nacity in a wolf spider: a serological dietary analysis. Oecologia, 56:79-83. Hallander, H. 1970. Prey, cannibalism, and micro- habitat selection in the wolf spiders Pardosa che- lata O.E Muller and P. pullata Clerck. Oikos, 21:337-340. Hendrix, P.E., R.W Parmelee, D.A. Crossley, Jr., D.C. Coleman, E.P. Odum & P.M. Groffman. 1986. Detritus food webs in conventional and no-tillage agroecosystems. BioScience, 36:374- 380. Kajak, A. 1997. Effects of epigeic macroarthro- pods on grass litter decomposition in mown meadow. Agric. Ecosys. Environ., 64:53-63. Kajak, A. & H. Jakubczyk. 1975. Experimental studies on spider predation. Pp. 82-85. In Proc. Sixth Intern. Arachnol. Cong., Amsterdam. (L. Vlijm, ed.). Kajak, A. & H. Jakubczyk. 1976. Trophic relation- ships of epigeic predators. Polish Ecol. Stud., 2: 219-229. Kajak, A. & H. Jakubczyk. 1977. Experimental studies on predation in the soil-litter interface. Ecol. Bull. (Stockholm), 25:493-496. Kajak, A. & M. Kaczmarek. 1988. Effect of pre- dation on soil mesofauna: An experimental study. XL Europaisches Arachnol. Colloq., Technische Univ. Berlin, Dokumentation Kon- gresse und Tagungen, 38:188-198. Kajak, A., K. Chmielewski, M. Kaczmarek & E. Rembialkowska. 1991. Experimental studies on the effect of epigeic predators on matter decom- position processes in managed peat grasslands. Polish Ecol. Stud., 17:289-310. Nentwig, W. 1987. The prey of spiders. Pp. 249- 273. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin. Nyffeler, M. 1982. Field Studies on the Ecological Role of the Spiders as Insect Predators in Agro- ecosystems (Abandoned Grassland, Meadow, and Cereal Fields). Ph.D. Thesis., Swiss Fed. Inst. Tech. Zurich, Switzerland. Nyffeler, M., WL. Sterling & D.A. Dean. 1994. How spiders make a living. Environ. EntomoL, 23:1357-1367. Parkinson, D.S., S. Visser & J.B. Whittaker. 1977. Effects of collembolan grazing on fungal colo- nization of leaf litter. Ecol. Bull. (Stockholm), 25:75-79. Peterson, H. 1971. The nutritional biology of Col- lembola and its ecological significance. A review 370 THE JOURNAL OF ARACHNOLOGY of recent literature with a few original observa- tions. Entomol. Meddr., 39:97-118. Pimm, S. 1982. Food Webs. Chapman and Hall, London. Polls, G.A. 1994. Food webs, trophic cascades and community structure. Australian J. EcoL, 19: 121-136. Polls, G. & S.D. Hurd. 1996. Allochthonous input across habitats, subsidized consumers, and ap- parent trophic cascades; examples from the ocean-land interface. Pp. 275-285. In Food Webs: Integration of Patterns and Dynamics. (G.A. Polls & K.O. Winemiller, eds.). Chapman and Hall, New York. Polls, G.A. & D.R. Strong. 1996. Food web com- plexity and community dynamics. American Nat., 147:813-846. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: Spiders in garden test systems. Ecology, 71:1441-1450. Robertson, L.N., B.A. Kettle & G.B. Simpson. 1994. The influence of tillage practices on soil macrofauna in a semi-arid agroecosystem in northeastern Australia. Agri. Ecosyst. Environ. 48:149-156. Schaefer, M. 1975. Experimental studies on the im- portance of interspecies competition for the ly- cosid spiders in a salt marsh. Pp. 86-90. In Proc. Sixth Intern. Arachnol. Cong., Amsterdam. (L. Vlijm, ed.). Settle, W.H., H. Ariawan, E.T Astuti, W. Cahyana, A. L. Hakim, D. Hindayana, A. S. Lestari & P. Sartanto. 1996. Managing tropical rice pests through conservation of generalist natural ene- mies and alternative prey. Ecology, 77:1975- 1988. Snyder, WE. & D.H. Wise. In press. Predator in- terference and the establishment of generalist predation populations for biocontrol. Biol. Con- trol. Spiller, D.A. 1992. Numerical response to prey abundance by Zygiella x~notata. J. Arachnol., 20: 179-188. Stinner, B.R. & G.J. House. 1990. Arthropods and other invertebrates in conservation-tillage agri- culture. Annu. Rev. Entomol., 35:299-318. Swift, M.J., O.W Heal & J.M. Anderson. 1979. Decomposition in terrestrial ecosystems. Univ. California Press, Berkeley. 372 pp. Toft, S. 1995. Value of the aphid Rhopalosiphym padi as food for cereal spiders. J. Appl. EcoL, 32:552-560. Toft, S. & D.H. Wise. 1999a. Growth, develop- ment, and survival of a generalist predator fed single- and mixed-species diets of different qual- ity. Oecologia, 119:191-197. Toft, S. & D.H. Wise. 1999b. Behavioral and eco- physiological responses of a generalist predator to single- and mixed-species diets of different quality. Oecologica, 119:198-207. Verhoef, H.A. & R.G.M. de Goede. 1985. Effects of collembolan grazing on nitrogen dynamics in a coniferous forest. Pp. 367-376. In Ecological Interactions in Soil. Plants, Microbes and Ani- mals. (A.H. Fitter, D. Atkinson, D.J. Read & M.B. Usher, eds). Boston: Blackwell Sci. Publ. Visser, S. 1985. Role of the soil invertebrates in determining the composition of soil microbial conununities. Pp. 297-317. In Ecological Inter- actions in Soil. Plants, Microbes and Animals. (A.H. Fitter, D. Atkinson, D.J. Read & M.B. Usher, eds). Boston: Blackwell Sci. Publ. Wamock, A.J., A.H. Fitter & M.B. Usher. 1982. The influence of the springtail Folsomia Candida (Insecta, Collembola) on the mycorrhizal asso- ciation of leek Allium Glomus fasciculatus. New PhytoL, 90:285-292. Wingerden, W.K.R.E. van. 1975. Population dy- namics of Erigone arctica (White) (Araneae, Linyphiidae). Pp. 71-76. In Proc. Xlth Intern. Cong. Arachnol., Amsterdam. (L. Vlijm, ed.). Wingerden, W.K.R.E. van. 1978. Population dy- namics of Erigone arctica (White) (Araneae, Linyphiidae). 11. Symp. Zool. Soc. London, 42: 195-202. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge Univ. Press, Cambridge. 328 pp. Yeargan, K.V 1975. Prey and periodicity of Par- dosa ramulosa (McCook) in alfalfa. Environ. En- tomol., 4:403-408. Manuscript received I May 1998, revised 5 October 1998. 1999. The Journal of Arachnology 27:371-377 ARCHITECTURAL FEATURES OF AGRICULTURAL HABITATS AND THEIR IMPACT ON THE SPIDER INHABITANTS Ann L. Rypstra: Department of Zoology, Miami University, 1601 Peck Blvd., Hamilton, Ohio 45011 USA Paul E. Carter, Robert A. Balfour, and Samuel D. Marshall: Department of Zoology, Miami University, Oxford, Ohio 45056 USA ABSTRACT. The density and diversity of the spider community has been closely tied to the structural complexity of the local environment. For instance, soil dwelling spiders increase dramatically when the litter layer is enhanced because there are more retreats and hiding places and because temperature and humidity extremes are moderated. Web-building spiders are directly linked to the configuration of the vegetation because of specific web attachment requirements. Both correlative and experimental data sup- port a tight relationship between spider density and habitat structure. Most of the available data show that agricultural practices which enhance the structural complexity of the environment (such as intercropping, mulching, and conservation tillage practices) enhance the density and diversity of the spider community. The key question regarding spiders in agroecosystems is, of course, whether they are in any way sup- pressing the activity of herbivores. Some studies uncovered a strong link between habitat complexity, spider abundance and plant productivity; but others have not, and the mechanisms by which spiders could exert a top-down effect are not clear. More investigation into the specifics of how habitat structure influ- ences the predator-prey interactions in agroecosystems is needed in order to truly understand and manage agricultural production in a responsible manner. Some of the earliest studies of spider hab- itat selection and community structure focused on the importance of architectural features of the environment. Clear relationships have been revealed between the physical complex- ity of the environment and spider abundance and diversity both across successional gradi- ents (Lowrie 1948; Barnes 1953; Duffey 1978; Hurd & Fagan 1992) and across geo- graphical regions (Greenstone 1984; Rypstra 1986). Surveys as well as manipulative stud- ies have demonstrated that spiders respond to the diversity and complexity of the vegetation (Rypstra 1983, 1986; Robinson 1981; Green- stone 1984; Gunnarsson 1990; Halaj et al. 1998) and that cursorial spiders, in particular, respond to the depth and complexity of the litter layer (Uetz 1976, 1979; Bultman & Uetz 1982, 1984; Hurd & Fagan 1992). As pointed out by Uetz (1991), there are several reasons why spiders should be more sensitive to struc- ture than other organisms. As a group, spiders perceive their environment using vibratory cues which are mediated through the substrate on which they live. Web spiders must anchor their prey capture device to the appropriate substratum and complex habitats provide ap- propriate sites for a greater range of sizes and types of webs. Finally, since all spiders are predators that can potentially consume one an- other, the extent to which they can coexist may strongly depend on their ability to move around and hide in a complex environment. The literature on the responses of individual species of spiders and the community as a whole to habitat structure and complexity was comprehensively reviewed in the early 1990s (Uetz 1991; Wise 1993). The conclusions of both of these reviews suggest that, although clear associations exist, the specific aspects of the environment to which the spiders respond have not been adequately teased out. Diverse habitats provide a greater array of microcli- mate features, alternative food sources, and a greater number of possible retreat sites that can encourage colonization and the establish- ment of robust populations. More specifically, plant diversity and plant composition may in- fluence predator diversity by changing forag- ing efficiency (Strong et al. 1984; Andow & Prokym 1990) and/or the nutritional quality of the herbivore prey (Price et al. 1980). Varia- 371 372 THE JOURNAL OF ARACHNOLOGY tions in the manner in which habitat structure affect interactions between predators and prey limit our ability to make broad generalizations regarding the specific mechanisms by which habitat diversity influences the spider com- munity. Numerous ecological models suggest that diverse producer communities will support more diversity at higher trophic levels. A re- cent study of Siemann (1998) demonstrates how complex testing this concept can be. He studied grassland communities with historical differences in fertilization into which he nest- ed a more recent fertilization treatment. Past fertilization caused more than a four-fold de- crease in plant species but resulted in no de- tectable differences in herbivore or detritivore species richness. Interestingly, arthropod pred- ator (including spider) and parasite species richness was significantly higher in these plots. Siemann (1998) suggested that the pred- ators might have become more abundant be- cause they were more successful in foraging in the physically simpler environment but also mentioned that the shift in producer species composition may have changed the nutritional quality of the herbivores for the predator. This example runs counter to generalizations tra- ditionally made by spider ecologists about the response of spider diversity to increased plant diversity (Uetz 1991; Wise 1993 and refer- ences therein). Siemann’s results (1998) chal- lenged our prior conceptions further in that his more recent fertilization treatments increased the diversity and abundance of all trophic lev- els in the grasslands. In this case, he explained the increase in herbivores and detritivores by suggesting that the increase in plant produc- tivity allowed rarer species to persist. How- ever, it is not at all clear why his two fertilizer treatments, which produced differences in the plant, herbivore and detritivore communities, had no apparent impact on predator diversity and/or abundance. Obviously the interactions in this trophic web cannot be explained by the basic principles we think we understand re- garding how arthropod predators, including spiders, respond to habitat complexity. The only way to truly understand the responses of the predators in Siemann’s (1998) experiments is to focus more directly on how the specific interactions between predators and their prey are mediated by the plant community. Studies such as Siemann’s (1998) may serve to make us more pessimistic regarding our ability to predict the potential impact of shifting agricultural practices on predatory ar- thropods. Since more than half of the preda- tory fauna in agroecosystems are spiders (Fer- guson et al. 1984; Young & Edwards 1990), and it is known that changes in spider density can impact pest populations (Mansour et al. 1983; Riechert & Lockley 1984; Nyffeler & Benz 1987; Nyffeler et al. 1994), it would seem logical that the spider community would be a key component of integrated pest man- agement strategies. Even though significant control of prey populations by assemblages of spiders has been suggested repeatedly (Clarke & Grant 1968; Riechert & Lockley 1984; Chiverton 1986; Agnew & Smith 1989), pest control strategies in North America rarely in- clude them (Young & Edwards 1990). In order to increase the emphasis on spiders as agents of biological control, it is imperative to deci- pher exactly how shifting agricultural practic- es that change the habitat structure within rel- atively homogeneous fields influence the density, diversity and foraging behavior of spiders. Although the agricultural literature was not specifically addressed in the reviews of Uetz (1991) and Wise (1993), a rich body of work has demonstrated that vegetation diversity of agroecosystems provides some measure of plant protection (Risch et al. 1983; Andow 1991a). Root (1973) proposed two hypotheses to explain the lower levels of herbivorous in- sects and pest damage in diverse systems. The resource concentration hypothesis suggests that specialist herbivores respond strongly to homogeneous systems of their host plant and cannot reach high levels in diverse systems. More critical to arachnologists is the enemies hypothesis which suggests that predators and parasites are more effective in diverse systems where alternative prey are present. Apparently the idea that biological diversity promotes community stability (the diversity-stability hypothesis of Mac Arthur 1955; Elton 1958) captured the attention of agriculturists study- ing the response of arthropods to diversity (Goodman 1975; Risch et al. 1983; Coll & Bottrell 1995). Thus, the notion that habitat diversification impedes the build up of pest populations became a paradigm even though empirical evidence in support of it is not any more rigorous than the identification of the RYPSTRA ET AL.— ARCHITECTURAL COMPLEXITY AND SPIDERS 373 specific features of a complex habitat to which the spider community responds. Although the tendency over recent decades has been toward the simplification of the ag- ricultural landscape, diversification within ag- ricultural fields can easily be attained by in- tercropping, cover cropping, changing planting strategies and tolerating weedy cul- ture. Overall, these practices tend to increase predator abundance (including spiders) and thus provide support for the enemies hypoth- esis (Ferguson et al. 1984; Coll & Bottrell 1995; Rypstra & Carter 1995; Balfour & Ryp- stra 1998; Costello & Daane 1998). However, it is not clear whether vegetation that provides abundant resources will act as a source or a sink for natural enemies in agroecosystems (Bugg et al. 1987; Kemp & Barrett 1989; Cor- bett & Plant 1993; Coll & Bottrell 1995; Cos- tellO' & Daane 1998). For example, the greater availability of alternative food sources may reduce predation rate on a target pest (Abies et al. 1978). Non-host plants and high struc- tural complexity may interfere with predator movement and alter the interactions between natural enemies and their prey (Perrin 1980; Andow & Prokym 1990). Alternative crops or weeds may actually draw predators away from the crop and thus reduce their impact on the herbivores (Bugg et al. 1987; Kemp & Barrett 1989; Rodenhouse et al. 1992). The specific manner in which diverse agricultural systems impact natural enemies in general or the spi- der community in particular needs to be quan- tified. Only then can we begin to make pre- dictions about how the habitat changes that accompany diversification affect the role that spiders play in the food web. There are a few examples of habitat manip- ulations in which the spiders appear to exert a top-down effect in the food web and in- crease plant production. In 1982, a USDA re- port mentioned that farmers in the Hunan re- gion of China used straw bundles as retreats for spiders during irrigation of rice fields and that this minor habitat manipulation was as- sociated with a 50-60% reduction in pesticide use. Kobayashi (1975) provided alternative prey in the form of fruit flies for spiders in- habiting rice paddies and observed an increase in spider populations and a decline in rice pests. However, the decrease in pest insects apparently came too late in the season to af- fect the amount of damage experienced by the plants. Critical experiments regarding the impor- tance of habitat manipulations to spiders were conducted in a mixed vegetable garden (Riechert 1990; Riechert & Bishop 1990). The habitat for the spiders was altered by adding mulch, which provides structure and moder- ates physical conditions for spiders. Prey den- sity was altered by planting flowering plants that were meant to attract pollinators. These manipulations increased spiders densities, re- duced pest insect densities, and led to reduced plant losses to herbivory. Further experimen- tation with spider removals and separation of the mulch and flower treatments demonstrated that the spiders that invaded the mulch were responsible for the observed increase in plant productivity (Riechert 1990; Riechert & Bish- op 1990). Garden systems tend to be more diverse than standard agricultural fields and small scale manipulations such as the addition of mulch are relatively easy for motivated gar- deners to implement if economic or produc- tion benefits were to be accrued. What is not clear from these experiments is the specific feature of the mulch (i.e., microhabitat mod- eration, structure, protection from predators, increased levels of prey, etc.) to which the spi- ders were responding. Planting ground cover under emergent ag- ricultural crops, such as vineyards and citrus groves, has been shown to increase spider abundance and diversity (Altieri & Schmidt 1985; Wyss et al. 1995; Costello & Daane 1998) but little effort has been invested in un- derstanding how it affects pest control. Cos- tello & Daane (1998) compared changes in the spider community in California grape vine- yards with and without ground cover and at- tempted to relate it to the abundance and di- versity of pest insects. Although there was no significant difference in the total spider abun- dance on vines with or without ground cover, Trachelas pacificus (Chamberlin & Ivie) 1932 (Araneae; Corineidae), was significantly more abundant on vines planted with ground cover (Costello & Daane 1998). Even though Cos- tello & Daane (1998) noted that T. pacificus is a major predator of the common leafhopper pests, they are pessimistic about the ability of ground cover to reduce pest populations by enhancing spider abundance. This question will not be resolved without a more mecha- 374 THE JOURNAL OF ARACHNOLOGY nistic approach to understanding the specific interactions among the species that inhabit the different components of the habitat. For ex- ample, more detail regarding the effects of multiple predators and how they might take advantage of the movement of potential prey between the vines and the ground cover plants might explain effects that are not obvious from descriptive information about the distri- bution of organisms. Losey & Denno (1998a; 1998b) have described a situation where the escape response of an aphid upon encounter- ing a predator foraging up on the vegetation made it more vulnerable to another predator on the soil surface. It is these kinds of com- plexities that must be incorporated more ex- plicitly into our attempts to understand how spiders in complex habitats affect the food web. Recently tilled and planted fields are barren habitats that are inhospitable for many arthro- pods, including spiders, yet this may be a crit- ical time for establishing a community of predators capable of impacting plant produc- tion, In the spring, many spider species are actively engaged in widespread dispersal (Bishop & Riechert 1990), thus habitat ma- nipulations that make the fields more attrac- tive to them at this time may be particularly critical. Carter & Rypstra (1995) attempted to encourage spider establishment by placing crates in soybean fields just after planting. Their idea was that these crates would provide shade and some habitat structure while the plants were still small and then, as the plants grew, the spiders would move out into the vegetation where their impact on pest insects would be greater. Although their crates were colonized by spiders, the predominant species in the crates was a species that is not naturally abundant in soybean fields; and there was no evidence that these spiders moved out of the crates after the plants were mature. Neverthe- less, across three seasons the biomass of in- sects consumed by the spiders in the boxes was negatively correlated with the amount of herbivory experienced by adjacent plants (Carter & Rypstra 1995). In two of three years, the herbivory experienced by plants in the vicinity of the boxes was significantly lower than in the fields at large or in areas where spiders were systematically removed. Although this manipulation was rather artifi- cial and unlikely to be practical on a large agricultural scale, it demonstrates that small manipulations that enhance spider populations can have significant effects on herbivory in a conventional agroecosystem. However, it is again the case that no attempt was made to reveal the specific mechanisms by which the spiders interacted with the herbivores to cause the reduction in leaf damage observed. Existing data provide strong evidence that simple habitat manipulations can affect spider populations and impact plant production in agroecosystems. Further experimentation must focus on how specific shifts in actual agricultural practices impact the spiders and, ultimately, the damage inflicted by herbivores. One example in which the aforementioned studies may be particularly applicable is the shift to conservation tillage (no-till) that has been occurring in North America over the last few decades (Sprague & Triplett 1986; Geb- hardt et al. 1985; Ehrenfeld 1987). Fields managed under conservation tillage regimes experience lower levels of soil disturbance, which reduces erosion and allows the devel- opment of a much more complex litter layer (Gebhardt et al. 1985; Hendrix et al. 1986; Wardle 1995). Likewise, pressure to reduce the use of chemical herbicides may lead to increased invasion of weeds (Triplett & Lytle 1972; Wardle 1995; Pavuk et al. 1997). It is generally known that no-till fields support a more diverse resident arthropod community including pests and natural enemies (House & Stinner 1983; Stinner & House 1990; Tonhas- ca 1993). As mentioned above, spider com- munities respond to both soil litter (Bultman & Uetz 1982; 1984; Riechert & Bishop 1990) and plant diversity, including weed density, in no-till soybean systems (Rypstra & Carter 1995; Balfour & Rypstra 1998). Likewise, lower levels of insect damage have been ob- served in some no-till systems (House & Stin- ner 1983; Andow 1991b). Therefore, one would expect this to be a promising system, from both ecological and economic points of view, in which to study the impact of habitat changes on the spider community and how they may impact plant production. Research needs to proceed toward devel- oping a mechanistic understanding of how spiders and other natural enenfies respond to specific habitat manipulations and how the habitat manipulations mediate predatory in- tensity. We need to uncouple the linkages be- RYPSTRA ET AL.— ARCHITECTURAL COMPLEXITY AND SPIDERS 375 tween the structure itself and the specific fea- tures of the structure to which the spiders are responding so that we can quantify the ulti- mate effects on the food web. Although it is generally hypothesized that diversity enhances natural enemies by providing supplemental re- sources, few studies have actually document- ed this phenomenon experimentally. Given the variability of the community level responses observed, further investigations should incor- porate a broad spectrum of specific effects such as the importance of spatial variation, changes in survivorship and fecundity, more detail on mobility and dispersal patterns, and the dynamics of the predator-prey interactions that occur within the agricultural systems (Corbett & Plant 1993). Only this level of comprehension will provide a basis for un- derstanding the specific role of spiders in agroecosystems and ultimately enable us to predict the response of spiders to changes in agricultural practices. Support for this project was provided by NSF grant DEB 9527710, by several grants from Sigma Xi, The Scientific Research So- ciety, the American Arachnological Society, and Miami University’s Undergraduate Re- search Fund, Summer Scholar’s Program, Committee for Faculty Research, Department of Zoology, Ecology Research Center, and Hamilton Campus. LITERATURE CITED Able, J.R., S.L. Jones, & D.W. McCommas. 1978. Response of selected predator species to different densities of Aphis gossypii and Heliothis vires- cens eggs. Environ. Entomol., 7:402-404. Agnew, C.W. & J.W. Smith, Jr. 1989. Ecology of spiders (Araneae) in a peanut agroecosystem. En- viron. Entomol., 18:30-42. Altieri, M.A. & L.L Schmidt. 1985. Cover crop management in northern California orchards and vineyards: Effects on arthropod communities. Biol. Agric. Hortic., 3:1-24. Andow, D.A. 1991a. Vegetational diversity and ar- thropod population response. Ann. Rev. Ento- mol., 36:561-586. Andow, D.A. 1991b. Yield loss to arthropods in vegetationally diverse agroecosystems. Environ. Entomol., 20:1228-1235. Andow, D.A. & D.R. Prokym. 1990. Plant struc- tural complexity and host finding by a parasitoid. Oecologia, 82:162-165. Balfour, R.A. & A.L. Rypstra. 1998. The influence of habitat structure on spider density in a no-till soybean agroecosystem. J. Arachnol., 26:221- 226. Barnes, R.D. 1953. The ecological distribution of spiders in nonforest maritime communities at Beaufort, North Carolina. Ecol. Mongr., 23:315- 337. Bishop, L. & S.E. Riechert. 1990. Spider coloni- zation of agroecosystems: Mode and source. En- viron. Entomol., 19:1738-1745. Bugg, R.L., P.M. Kareiva, & L. Zia. 1987. Effect of common knotweed {Polygonum aviculare) on abundance and efficiency of insect predators of crop pests. Hilgardia, 55:1-51. Bultman, T.L. & G.W. Uetz. 1982. Abundance and community stmcture of forest floor spiders fol- lowing litter manipulation. Oecologia, 55:34-41. Bultman, T.L. & G.W. Uetz. 1984. Effect of stmc- ture and nutritional quality of litter on abundanc- es of litter-dwelling arthropods. American Midi. Nat., 111:165-172. Carter, RE. & A.L. Rypstra. 1995. Top-down ef- fects in soybean agroecosystems: Spider density affects herbivore damage. Oikos, 72:433-439. Chiverton, P.A. 1986. Predator density manipula- tion and its effects on populations of Rhopalo- siphum padi (Homoptera: Aphididae) in spring barley. Ann. Appl. Biol., 109:49-60. Clarke, R.D. & PR. Grant. 1968. An experimental study of the role of spiders as predators in a for- est litter community. Part I. Ecology, 49:1152- 1154. Coll, M. & D.G. Bottrell. 1995. Predator-prey as- sociations in mono- and dicultures: Effect of maize and bean vegetation. Agric., Ecosys. & Environ., 54:115-125. Corbett, A. & R.E. Plant. 1993. Role of movement in the response of natural enemies to agroeco- system diversification: A theoretical evaluation. Environ. Entomol., 22:519-531. Costello, M.J. & K.M. Daane. 1998. Influence of ground cover on spider populations in a table grape vineyard. Environ. Entomol., 23:33-40. Duffey, E. 1978. Ecological strategies in spiders including some characteristics of species in pio- neer and mature habitats. Symp. Zool. Soc. Lon- don, 42:109-123. Ehrenfeld, D. 1987. Implementing the transition to a sustainable agriculture. Bull. Ecol. Soc. Amer- ica, 68:5-8. Elton, C.S. 1958. The Ecology of Invasions by An- imals and Plants. Methuen, London. Ferguson, H.J., R.M. McPherson & W.A. Allen. 1984. Ground- and foliage-dwelling spiders in four soybean cropping systems. Environ. Ento- mol., 13:975-980. Gebhardt, M.R., TC. Daniel, E.E. Schweizer & R.R. Allmaras. 1985. Conservation tillage. Sci- ence, 230:625-630. Goodman, D. 1975. The theory of diversity-stabil- 376 THE JOURNAL OF ARACHNOLOGY ity relationships in ecology. Quart. Rev. Biol., 50:238-266. Greenstone, M.H. 1984. Determinants of web spi- der species diversity: Vegetation structural diver- sity vs. prey availability. Oecologia, 62:299-304. Gunnarsson, B. 1990. Vegetation structure and the abundance and size distribution of spruce-living spiders. J. Anim. Ecol., 59:743-752. Halaj, J., D.W. Ross, & A.R. Moldenke. 1998. Habitat structure and prey availability as predic- tors of the abundance and community organiza- tion of spiders in western Oregon forest cano- pies. J. ArachnoL, 26:203-220. Hendrix, RE, R.W. Parmalee, D.A. Crossley, D.C. Coleman, E.P. Odum & P.M. Groffman. 1986. Detritus food webs in conventional and no-tillage agriculture. BioScience, 36:374-380. House, G.J. & B.J. Stinner. 1983. Arthropods in no-tillage soybean agroecosystems: Community composition and ecosystem interactions. Envi- ron. Manag., 7:23-28. Hurd, L.E. & W.E Fagan. 1992. Cursorial spiders and succession: Age or habitat structure? Oecol- ogia, 92:215-221. Kemp, J.C. & G.W. Barrett. 1989. Spatial pattern- ing: Impact of uncultivated corridors on arthro- pod populations within soybean agroecosystems. Ecology, 70:114-128. Kobayashi, S. 1975. The effect of Drosophila re- lease on the spider population in a paddy field. Appl. Entomol. ZooL, 10:268-274. Losey, J.E. & R.R Denno. 1998a. The escape re- sponse of pea aphids to foliar-foraging predators: Factors affecting dropping behaviour. Ecol. En- tomol., 23:53-61. Losey, J.E. & R.E Denno. 1998b. Positive preda- tor-predator interactions: Enhanced predation rates and synergistic suppression of aphid pop- ulations. Ecology, 79:2143-2152. Lowrie, D.C. 1948. The ecological succession of spiders of the Chicago area dunes. Ecology, 29: 334-351. Mac Arthur, R.H. 1955. Fluctuations in animal pop- ulations and a measure of community stability. Ecology, 36:533-536. Mansour, E, D.B. Richman & W.H. Whitcomb. 1983. Spider management in agroecosystems: Habitat manipulation. Environ. Manag., 7:43-49. Nyffeler, M. & G. Benz. 1987. Spiders in natural pest control: A review. J. Appl. Entomol., 103: 321-339. Nyffeler, M., W.D. Sterling & D.A. Dean. 1994. How spiders make a living. Environ. Entomol., 23:1357-1367. Pavuk, D.M., EE Purrington, C.E. Williams, & B.R. Stinner. 1997. Ground beetle (Coleoptera: Carabidae) activity density and community com- position in vegetationally diverse com agroeco- systems. American Midi. Nat., 138:14-28. Perrin, R.M. 1980. The role of environmental di- versity in crop protection. Protection Ecol., 2:77- 114. Price, P.W., C.E. Bouton, P. Gross, B.A. McPherson, J.N. Thompson, & A.E. Weis. 1980. Interactions among three trophic levels: Influence of plants on interactions between insect herbivores and natural enemies. Annu. Rev. Ecol. Syst., 11:41- 65. Riechert, S.E. 1990. Habitat manipulations aug- ment spider control of insect pests. Acta Zool. Fennica, 190:321-325. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: Spiders in garden test systems. Ecology, 71:1441-1450. Riechert, S.E. & T. Lockley. 1984. Spiders as bi- ological control agents. Annu. Rev. Entomol., 29:299-320. Risch, S.J., D. Andow & M.A. Altieri. 1983. Agroecosystem diversity and pest control: Data, tentative conclusions and new research direc- tions. Environ. Entomol., 12:625-629. Robinson, J.V. 1981. The effect of architectural variation in habitat on a spider community: An experimental field study. Ecology, 62:73-80. Rodenhouse, N.L., G.W. Barrett, D.M. Zimmerman & J.C. Kemp. 1992. Effects of uncultivated cor- ridors on arthropod abundances and crop yields in soybean agroecosystems. Agric., Ecosys. & Environ., 38:179-191. Root, R.B. 1973. Organization of plant- arthropod association in simple and diverse habitats: The fauna of collards {Brassica oleracea). Ecol. Monogr., 43:94-125. Rypstra, A.L. 1983. The importance of food and space in limiting web- spider densities: A test us- ing field enclosures. Oecologia, 59:312-316. Rypstra, A.L. 1986. Web spiders in temperate and tropical forests: Relative abundance and environ- mental correlates. American Midi. Nat., 115:42- 51. Rypstra, A.L. & RE. Carter. 1995. The web-spider community of soybean agroecosystems in south- western Ohio. J. ArachnoL, 23:135-144. Siemann, E. 1998. Experimental tests of effects of plant productivity and diversity on grassland ar- thropod diversity. Ecology, 79:2057-2070. Sprague, M. A. & G. B. Triplett. 1986. No-Tillage and Surface Tillage Agriculture. John Wiley and Sons, New York. Stinner, B.R. & G.J. House. 1990. Arthropods and other invertebrates in conservation-tillage agri- culture. Ann. Rev. Entomol., 35:299-318. Strong, D.R., J.H. Lawton, & T.R.E. Southwood. 1984. Insects on Plants. Harvard Univ. Press, Cambridge, Massachusetts. Tonhasca, A. 1993. Carabid beetle assemblage un- der diversified agroecosystems. Entomol. Exp. Appl., 68:279-285. RYPSTRA ET AL.— ARCHITECTURAL COMPLEXITY AND SPIDERS 311 Triplett, G.B. & G.D. Lytle. 1972. Control and ecology of weeds in continuous com grown without tillage. Weed Science, 20:453-457. Uetz, G.W. 1976. Gradient analysis of spider com- munities in a streamside forest. Oecologia, 22: 373-385. Uetz, G.W. 1979. The influence of variation in lit- ter habitats on spider communities. Oecologia, 40:29-42. Uetz, G.W. 1991. Habitat stmcture and spider for- aging. Pp. 325-348. In Habitat Stmcture: The Physical Arrangement of Objects in Space. (S.S. Bell, E.D. McCoy & H.R. Mushinsky, eds.).. Chapman & Hall, London, U.K. United States Dept, of Agric. 1982. Biological Control of Pests in China. 201 pp. U.S.D.A., Washington, D.C. Wardle, D.A. 1995. Impacts of disturbance on de- tritus food webs in agro-ecosystems of contrast- ing tillage and weed management practices. Adv. EcoL Res., 26:105-185. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge Univ. Press, Cambridge, U.K. Wyss, E., U. Niggli, W. Nentwig. 1995. The im- pact of spiders on aphid populations in a strip managed apple orchard. J. Appl. EntomoL, 119: 473-478. Young, O.P. & G.B. Edwards. 1990. Spiders in United States field crops and their potential ef- fects on crop pests. J. ArachnoL, 18:1-27. Manuscript received 1 May 1998, revised 9 Novem- ber 1998. 1999. The Journal of Arachnology 27:378-386 AN INDIVIDUAL-BASED MODEL FOR DISPERSIVE SPIDERS IN AGROECOSYSTEMS: SIMULATIONS OF THE EFFECTS OF LANDSCAPE STRUCTURE C.J. Topping; Department of Landscape Ecology, NERI, Kal0, Grenavej 14, DK-8410 R0nde, Denmark ABSTRACT. A general individual-based model of spiders in agricultural land was constructed. The populations of spiders were simulated on landscapes which were defined from a set of landscape descrip- tors based on a Danish agricultural landscape. These descriptors gave the types of habitats present in the landscape together with their area and a frequency distribution of the size of individual habitat patches. The agricultural land was divided into crop types each with its own array of crop managements which were considered to influence the spiders via mortality. The dimensions of the model are relatively large, with the spider population able to grow to a size of one million individuals and with a spatial resolution of 10* landscape units. The effect of altering the spatial organization of the landscape elements was investigated together with the influence of the size of fields in the agricultural landscape. Results showed that the spatial arrangement of landscape elements did not affect spider population sizes, but that the effect of increasing habitat patch size, whilst maintaining a constant habitat area, was to increase population sizes, especially where dispersal was minimal. Thus stochastic events (e.g., mortality and the placement of set-aside), were not significant factors in the simulation results. Simulation results indicated that the optimal dispersal strategy for spiders in this system was one of high juvenile dispersal, although the extent to which these results can be translated to other systems is not yet known. These results indicate the potential for using models of this type for theoretical investigations of the life-history strategies used by spiders, especially where landscape heterogeneity and limited dispersal ability could result in complex spatial dynamic patterns. There is an increasing realization that land- scape-scale perspectives are important when considering the ecology of organisms (Dun- ning et al. 1992; Kupfer 1995). However, this new perspective brings with it a number of problems, not least of these is the difficulty of integrating the bewildering variety of data re- quired to interpret the responses of organisms at landscape scales. These data typically op- erate at a range of scales and across a range of disciplines. For instance, a model of a dis- persive spider may have to take large-scale topography into account because of the long- distance dispersal possible, but may also need detailed information at a field-scale to deter- mine the reproductive success or survival of the spider when not dispersing. In addition, weather, local management (such as grazing, plowing and spraying) and the spiders re- sponses to these factors must be considered. In this way the disciplines of ethology and ecology meet and their respective scales need to be reconciled (Lima & Zollner 1996). When these two are combined it may be pos- sible to wield considerable interpretative pow- er through the use of computer simulation models which are capable of integrating the range of data and scales required to investi- gate the determinants of spatial dynamics and distributions in heterogeneous landscapes. The typical agricultural landscape of west- ern Europe is, for the spiders that live there, a patchwork of more or less ephemeral habi- tats of varying quality. Survival in such land- scapes presents the spider with a problem of exploiting resources while conditions are good and minimizing the chance of being killed by agricultural operations. However, landscapes are highly variable and agricultural practices are ever-changing, thus prediction of the suc- cess of a spider population will depend upon the particular set of circumstances under con- sideration. To date, there have been attempts to model this type of system as a set of pop- ulations linked by dispersal. Topping & Sun- derland (1994) used a two-dimensional system of 30 X 30 squares to model the landscape whilst Halley et al. (1996) used a one-dimen- 378 TOPPING— AN INDIVIDUAL-BASED MODEL FOR DISPERSIVE SPIDERS 379 sional ribbon of fields. The one-dimensional ribbon was justified by the assumption that spiders always disperse on a scale which is much larger than one or a few fields, and hence local spatial influences will be negligi- ble. However, many spiders do not disperse on these scales and so it is pertinent to ask what is the effect of landscape structure on spiders with restricted dispersal ability. There are two aspects to take into account. The first is the actual landscape structure. This was im- proved by Topping (1997) to include much more realistic structures by hand mapping the landscape two-dimensionally in units of 50 X 50 m. However, this technique is awkward to use; and it is very time consuming to develop standardized landscapes which are significant- ly different from each other. The second as- pect is that it may be more realistic to consider spiders, not as populations arbitrarily classi- fied as being within a man-made field, but as individuals which are free to move between these artificial population boundaries. Thus the power of the individual-based modeling approach is invoked to take individual varia- tion and the influence of location into account (see DeAngelis Rose 1992). This paper presents a model which can be used to develop detailed individual-based sim- ulations to predict the impact of land-use and structure on the dynamics of spider popula- tions with differing life-history strategies. Since landscape structure is an essential ele- ment the model is designed to be able to cre- ate, automatically, different landscapes with the same basic landscape descriptors (e.g., mean field size and variance). The model is designed for maximum flexibility of life his- tory and landscape structure and management. The simulations presented here do not fully utilize these options, but they illustrate the po- tential of the approach by investigating the in- teraction between spider dispersal strategy and landscape structure. METHODS Model overview. — -The model was built using C++ and an object oriented approach; it is an i-state configuration model (Caswell & John 1992) with a discrete time interval. The time interval used in these simulations was one month. However, events involving the development of spiders occurring at a finer temporal resolution were incorporated by han- dling these events consecutively. Hence the development of eggs is considered before the development of juveniles. Mortality and dis- persal are also managed in the same way. As a result an individual egg could, under ex- treme conditions, develop into a juvenile, dis- perse, develop into an adult and then be killed all within the single time step. There are two main elements to the model, the landscape and the spider population: The landscape: The model landscape has a resolution of 10,000 X 10,000 units and is built of rectangles of varying size and dimen- sions which are classified into different habitat types. The basic landscape used in this study is based on the landscape of Arhus County in Denmark, which is probably typical of most of the intensive agricultural areas of north- western Europe. The landscape was construct- ed by first defining a set of landscape descrip- tors which were used to generate a landscape of a particular type. These descriptors list the number of different habitat types of which the landscape is comprised, together with the pro- portion of land area they cover and a statisti- cal description of the frequency distribution of size of individual patches of habitats. These habitat types typically include urban devel- opments, variously classified woodland, open water, agricultural fields and miscellaneous small or marginal habitats. Once the descrip- tors were defined, the landscape was con- structed automatically by a model which ran- domly selects individual habitats (defined as rectangles) from the described distributions and fits these together to form a landscape. Once the pre-defined total area for a habitat type is reached no more habitats of this type are selected by the model. At the end of this process the landscape consists of a patchwork of major habitat types, thus having an overall structure but little detail about the precise habitat types. This is reme- died by a second set of descriptors. These de- scriptors sub-divide the major habitat classi- fication into a sub-classification (e.g., agricultural fields will be allocated crop types). Again, the sub-habitat classification applied is based on the proportion of habitat occupied by that sub-habitat in terms of area. These sub-classifications can also be divided into management categories. For instance, each winter wheat field may be classified as one of up to 15 different management types 380 THE JOURNAL OF ARACHNOLOGY (combinations of different farming opera- tions). At the level of management categories each habitat patch in the landscape will have a management code attributed to it for each month of the year. These codes will determine when agricultural events will occur. The land- scape also simulates rotational set-aside which covers 7% of the arable area in Arhus County. Each year, the model allocates set-aside fields randomly to the landscape until 7% of the ar- able area is covered. However, no field may be in set-aside for two consecutive years. Crop rotation is implemented in a similar way by changing the crops allocated to fields whilst still maintaining the same area cover for each crop. An internal clock is incorpo- rated which allows the habitat patches to alter their properties at each time step, so as to sim- ulate the management operations described for their particular management category and habitat sub-classification combination. Weath- er data were also incorporated on a monthly basis in the form of a mean temperature and mean number of calm days (as a measure of dispersal potential via ballooning). The land- scape simulation therefore consists of habitat patches which have been classified into dif- ferent vegetation and management types, as described by a set of varying monthly states. Thus landscape structure and heterogeneity can be easily controlled and varied tremen- dously, both spatially and temporally. The spider population: The spider popula- tion modeled is, like a real population, com- prised of individual spiders in various states of development. The model recognizes indi- vidual spiders in three distinct development stages: egg, juvenile, and adult. In order to reduce the number of computations, only fe- male spiders are considered in the model. Thus the implicit assumptions are that male spiders are never a limiting resource and that they follow the same dispersal rules as the fe- males. Each stage has its own behavioral rules which govern the behavior of the individual in the simulation (Table 1). Population devel- opment is controlled by these behavioral rules and by the landscape’s time clock and land- scape data. Once set in motion the individual spiders modeled in the simulation reproduce, disperse and die according to their behavioral rules and data which they gather from the landscape. The model allows for a population size of a total of eggs, juveniles and adults not ex- ceeding one million individuals. In order to ensure that the population could not grow be- yond this limit, parameters representing car- rying capacities were scaled down in initial tests until simulated populations were of a suitable size & Kj, Table 2). Parameterization. — The landscape: The basic landscape was created using statistics re- lating to Arhus County, Denmark. Average weather parameters were obtained from Dan- ish national statistics (anonymous 1997) and temperature data loggers operating at Rpnde, Jutland. These weather patterns were incor- porated as standard weather for all simulation years. Topographical data used to generate es- timates of area coverage for lakes and forests were available from the Arhus County admin- istration. Urban area sizes were estimated from human population census data by re- gression against the area of 31 towns and vil- lages for which area data were available (re- gression equation km^ == O.OOOSx + 0.0433, where x is the human population; P < 0.0001). Field size data were obtained from ongoing agricultural studies in Denmark (T. Dalgaard pers. comm.). The area covered by agricultural crops was available from Danish National Statistics (anonymous 1996). Typical crop management for the area was obtained from Danish agricultural advisors. These stan- dard parameters were used to build the stan- dard landscape, ‘Landscape L’ The spiders: The spiders modeled in this study were designed to represent a generalized spider species of the type which commonly inhabits agricultural land in western Europe. Table 2 lists the parameters that are integral to the spider models and their functions. The parameter values relating to reproduction and development were based on studies of the lin- yphiid spider Lepthyphantes tenuis Blackwall (Topping & Sunderland 1994, 1996, 1998; Sunderland et al. 1996) and were of a fixed value throughout the study. Other integral pa- rameters were used as variables in this study and thus took on values according to the sim- ulation being undertaken. Thus, following Weyman et al. (1995), the probability that a spider will disperse under favorable condi- tions was held constant throughout the year, whilst weather conditions control the possi- bility to disperse at any given time (based on TOPPING— AN INDIVIDUAL-BASED MODEL FOR DISPERSIVE SPIDERS 381 Table 1. — Simulated behaviors of each spider life-stage modeled. Egg Behaviors: Develop: Mortality: Hatch: Juvenile Behaviors: Develop: Mortality: Dispersal: The egg develops according to the temperature experienced following a standard day-degrees equation using the stage specific parameter Determines whether the egg dies using a probability given by the mortality proba- bility Me, and the probability of mortality caused by management of the habitat at its present location. The egg hatches and becomes a juvenile. The juvenile develops according to the temperature experienced following a standard day-degrees equation using the stage specific parameter Tj. Determines whether the juvenile dies using a probability given by the mortality probabilities Mj, and M,2 (see Table 2), the density of population in the habitat patch the spider occupies (interpreted as a density above a threshold, K^^), and the proba- bility of mortality caused by management of the habitat at its present location. Determines whether a spider will attempt dispersal based on the probability given by the dispersal motivation (Wy) and the prevailing wind conditions obtained from the landscape. If dispersal can occur then a new location is generated, based upon a random direction and a distance traveled. The distance traveled is obtained from where ‘a’ is a random number between 0 and Dj, and Dj is the maximum distance traveled by a juvenile spider. This equation results in a center-weighted distribution of spiders from a point source. Thus a spider has a greater chance of traveling a short distance than a large and may travel up to Dj units from its starting position. Adult Behaviors: Mortality: Determines whether the adult dies using a probability given by the mortality prob- abilities M^, and M^2 (see Table 2), the density of population in the habitat patch the spider occupies (interpreted as a density above a threshold, K^), and the proba- bility of mortality caused by management of the habitat at its present location. In addition, the spider will die if it has no further reproductive potential. Dispersal: As for the juvenile but uses the adult-specific parameters D^ rather than D, and rather than Wj. Reproduction: Depending upon the time of year and adult density, the adult may produce one or more egg-sacs per month. Reproduction is density-dependent above K^. The chance of an individual producing and egg-sac decreases linearly with increasing density until it is zero at X 10. The gradient of this relationship is given by the parameter B. Each egg sac is assumed to have a fixed number of eggs and there are a fixed number of egg-sacs possible per spider (assuming no premature mortality). Egg-sacs are deposited at the spider’s current location. After producing an egg-sac the spider may disperse. The probability of dispersal here is separate from the normal course of dispersal and is controlled by another parameter {RS). Thus RS allows some flexibility in reproductive strategy by allowing a spider to either lay eggs in a single location or disperse before producing the next egg-sac. mean wind speeds); however, this parameter was varied between simulations. There were also parameters relating to the effect of management on the spiders. Poten- tially these parameters could be used to vary reproductive ability, dispersal ability and mor- tality; but in the simulations presented here only mortality and reproduction were directly controlled by these values. Reproduction was controlled by a binary switch which prevented reproduction in urban, water and forest habi- tats. Mortality was varied with farming oper- ation (e.g., 90% mortality was assumed as a result of insecticide application). Simulations. — The simulations investigat- ed variation in dispersal-related parameters 382 THE JOURNAL OF ARACHNOLOGY Table 2. — Parameters controlling simulated spi- der behaviours. Parameter Function M, Monthly egg mortality Mn Monthly juvenile mortality below Kj Mj2 Monthly juvenile mortality above Monthly adult mortality below Monthly adult mortality above Kj Juvenile density at which mortality becomes density-dependent Ka Adult density at which mortality be- comes density-dependent Wj Juvenile dispersal motivation W, Adult dispersal motivation Dj Juvenile maximum dispersal distance D, Adult maximum dispersal distance RS Reproductive Strategy — the Proba- bility of dispersal after laying eggs E Number of eggs per egg-sac S Number of egg-sacs per adult B The gradient of the linear relationship between density and reduction in probability of egg-sac production Te Day degrees required before eggs hatch T, Day degrees required before juve- niles mature over a range of landscapes with different structures and managements. The main land- scapes considered were based on topographi- cal and management information from Arhus County in Denmark. Thus the basic landscape comprised of 78% agricultural land of which one third is grassland, the remaining 22% is forest of various types, water and urban area. Three values (maximum, minimum and mid- range), were used for five dispersal related pa- rameters (reproductive strategy (RS), juvenile dispersal-motivation (VFy), adult dispersal-mo- tivation (W^), juvenile dispersal distance (Dj), adult dispersal distance All sensible combinations of the possible 243 combina- tions of these values were tested. Thus a zero dispersal motivation ability precluded the test- ing of dispersal distances of greater than zero. Testing of each combination was achieved by taking the mean population size during the last 15 years of simulation from 20 simulations of 20 years using the same parameters for the spider life-history and the same basic land- scape (i.e., the first five years were ignored because of the potential variance from differ- ent starting positions at the beginning of the simulation). Thus variation between runs was due to stochasticity in the model (e.g., dis- persal decisions, mortality, the position of set- aside), and the starting position for each sim- ulation (which was a randomly positioned population of 1000 adult individuals). All landscapes constructed for the simulation were constructed by the model randomly al- locating the pattern of landscape elements. In all cases only the stated change was made to the landscape so all other landscape descrip- tors were kept constant. All landscapes pro- duced were checked by eye for skewed dis- tribution of field sizes (e.g., all small fields clustered in one comer), and those showing such distributions were rejected. Simulations considered. — (A). The effect of varying agricultural field size whilst keep- ing other factors constant. Three mean field sizes were used, 2.6, 4.9, 8.7 ha respectively (Landscapes 1=3). In each case three replicate simulation mns were performed in order to establish the degree of between-mn variation. For the overall comparisons, these runs were combined by taking means. (B). The effect of different landscape stmctures whilst maintain- ing the same management and area coverage of habitats. Landscapes were re-created to give approximately the same mean field sizes and habitat coverage as Landscapes 1-3, but with a different spatial arrangement of actual habitat blocks (Landscapes 4-6). Again three replicate runs were combined for the overall comparisons. RESULTS Simulations. — -Note that all population siz- es given refer only to adults in the simulated population. The degree of between-run varia- tion was minimal. Landscape 3 produced the maximal between-run variation (Fig. 1). Var- iation between runs of other simulations was negligible. In all cases it proved impossible to repro- duce exactly the landscape construction hence there was some variation in mean field sizes (2.7 V5. 2.8; 4,9 v.s. 5.0; 8.8 v.s. 8.9). However, the effect of recombining landscape elements but maintaining the same overall landscape structure was to produce little more variation than found between runs. Regression analysis produced slopes of 1.000, 0.999 and 1.008 for small, medium and large fields respectively TOPPING— AN INDIVIDUAL-BASED MODEL FOR DISPERSIVE SPIDERS 383 Mean Annual Population - Landscape 3, Run 1 Figure 1. — The mean adult population size of two runs of the same simulation plotted against a third run. These simulations used ‘Landscape 3’ and resulted in the maximal between run variation of any simulations. This variation is caused by sto- chastic factors within the model not associated with landscape structure. (all correlations n ” 147, > 0.999, P < 0.000001). Figure 2a, b shows the results of compari- sons between simulation runs using land- scapes with different mean field sizes com- pared to simulations using Landscape 1. As the field size increases so the departure of the curve from linear also increases. Four distinct life-history strategy groups can be identified (1-4, Fig. 2b). By examining the input param- eters resulting in these four groups the follow- ing pattern emerged: Group 1: Low mean population level. In all cases there was little or no dispersal — either dispersal motivation or dispersal distance was zero for both adults and juveniles. Limited dispersal was only possible for some adults with an RS > 0. Group 2: Low to medium mean population levels. No juvenile dispersal, but adult dispersal distance > 0. This group can be further sub-divided (into sub-groups 2. 1-2.4), on the basis of in- creasing mean population size into four groups with the following pattern: 2.1 — RS = 0; 2.2 - RS = 50, = 50; 2.3 ~ RS = 100, = 50; 2.4 RS > 0, = 100; Group 3: Medium to high mean population levels. Intermediate juvenile dispersal, i.e., Wj ~ 50. Within the group the trend is for decreasing adult dispersal ability with increasing mean population size. Group 4: High mean popu- lation levels. High juvenile dispersal (Wj = 100). This group can be clearly sub-divided (into sub-groups 4. 1-4.4), on the basis of in- creasing mean population size. 4.1 — RS = 100 or = 50, > 0; 4.2 - RS = 50, = 50, > 0; 4.3 — either RS = 0 or = 0; 4.4 — 100. Thus the groups of increas- ing population size also correspond to de- creasing adult dispersal ability. Increasing field size led to an increase in mean population size, but the increase was not linear. Group 2 was most strongly affected, especially sub-groups 2.1 and 2.2. Group 3 demonstrated considerable variation within the group whilst Group 4 responded linearly. The groups can be separated into two distinct parts (groups 1-2 and groups 3-4), on the ba- sis of juvenile dispersal. Simulation results from Landscape No. 3 (large fields) show a distinct asymptotic curve over groups 1 and 2, but no such relationship for groups 3 and 4. There is a similar, but less pronounced pat- tern for Landscape 2 (medium-sized fields). Almost identical patterns can also be created by plotting the results of Landscapes 4-6 against Landscape 1, with Landscapes 5 and 6 similar to Landscapes 2 and 3, and Land- scape 4 being only very slightly curved over groups 1 and 2. DISCUSSION There was little between-run variation in mean population size when using the same landscape. This may not be too surprising be- cause the differences between runs would be entirely related to stochastic events (e.g., mor- tality and the placement of set-aside). How- ever, it does demonstrate that this stochasticity is not a significant factor in the simulation re- sults. The variation in model output between different landscapes with the same basic con- figuration was also rather limited. The exact spatial relationship between habitat patches was not therefore significantly influencing the outcome of these simulations. The effect of varying field size was, how- ever, much more noticeable, especially for those simulations where there was no juvenile dispersal. In these cases the effect of increas- ing field size was to preferentially increase the population means for those simulations where adults could not disperse between egg-laying (RS = 0, Group 2.1) and those simulations where the only dispersal possible was due to 384 THE JOURNAL OF ARACHNOLOGY A. B. Adult dispersal ability: Figure 2. — The mean adult population size resulting from simulations of two landscapes plotted against the results from the standard landscape ‘Landscape V with small field sizes. Deviations from a straight line with a slope of 1 indicate an interaction between landscape structure and life-history strategy. A. ‘Landscape 2% medium fields size. B. ‘Landscape 3\ large field size. Numbered boxes refer to groups of results (see text). Arrows show the direction of change in the magnitude of adult dispersal within the groups indicated. TOPPING— AN INDIVIDUAL-BASED MODEL FOR DISPERSIVE SPIDERS 385 RS (some of Group 1). In the most extreme case (low adult and no juvenile dispersal), the population increase was 400%. At first sight this seems like a counter-intuitive result. If in- creasing adult dispersal increases population size, then increasing field size should reduce the population because it effectively reduces the chance of dispersal to other habitats. But there are three interacting factors here: the patch, the effect of density and the ephemeral nature of the suitable habitats. The spider pop- ulations in the fields are always being reduced to low levels and then increasing again. When calculating the effect of density, the model as- sumes that spiders in a field space themselves more or less evenly throughout the field, thus density is field population size divided by field area. Hence, density dependent effects can cause two fields, one small and one large both being re-colonized by the same number of col- onists after a catastrophic event, to have dif- ferent growth curves. This effect will result in a feedback loop if dispersal is low, because few small fields can be colonized. Those that are colonized will get a relatively high density of colonists due to the fact that dispersal is spatially local, thus initial density will be high, compared to the same dispersers dis- persing into a large field. This high density results in a low rate of growth which will mean fewer colonists and so even more re- stricted spatial distributions. This behavior would not be exhibited by the spatially- sim- plistic population-based models of Topping & Sunderland (1994) and Halley et al. (1996). The implication of this effect is interesting if consideration is given to those life-history strategies with medium levels of dispersal. In these cases, for large fields, many variations in strategy (e.g., minimal adult and no Juve- nile dispersal compared with maximal juve- nile and maximal adult dispersal) all result in approximately equivalent mean annual popu- lations. However, the same strategies on a small-field landscape result in a > 100% dif- ference in mean annual population size. For such species landscape structure could be very important. In these simulations, juvenile dispersal is clearly the most important factor determining mean population size. In those simulations with juvenile dispersal there is a relationship between increasing juvenile dispersal ability and population size; however, within the groups with juvenile dispersal (Group 3-4), there is a negative relationship between adult dispersal ability and population size. This sug- gests that in this system dispersal is generally a beneficial thing, but that it is not an advan- tage to over-disperse. For the simulations without juvenile dispersal, generally the more adult dispersal the better; but the response is curvilinear such that maximal adult dispersal results in only marginally larger populations than medium adult dispersal levels, especially when fields are large. This is almost the op- posite to the strategy suggested by Van Win- gerden (1980), who believed that the best col- onization strategy was to disperse as mated females. However, in the case where juveniles could not disperse significantly, Van Winger- den’s strategy matches the model results. It should also be noted that the results may change for simulations of other combinations of life-history parameters and landscape struc- tures and managements. Hence, in agreement with the model results, high juvenile dispersal is the norm for most families (e.g.. Dean & Sterling 1985). However, there are also sig- nificant differences between studies and fam- ilies, for instance Duffey (1956) and Green- stone et al. (1987) found that in Linyphiidae there can be a relatively large proportion of adults dispersing by ballooning. This is almost certainly a contributory factor to the confusion around the causative factors of ballooning (see Weyman 1993). The choice of ballooning strategy will depend upon the particular situ- ation under consideration, and the combina- tion of factors needing to be considered results in a complex problem. The model presented here is still in an early stage of development. In particular, more work is needed to determine how important the effects of spatial heterogeneity in the land- scape are and how dividing the temporal as- pects into even finer time steps might influ- ence the model output. More importantly, an analysis of actual spider strategies might re- veal constraints to the number of possible pa- rameter combinations used. This would be particularly useful when considering varying developmental and reproductive parameters. Nevertheless, the preliminary results present- ed here suggest that this type of simulation modeling may render some of the complex as- pects of investigations into the spatial dynam- ics of spiders more tractable than previously 386 THE JOURNAL OF ARACHNOLOGY possible. In particular, the two-dimensional nature of the model together with the individ- uahbased approach permits the investigation of the effects of local conditions within a complex landscape. Such effects have not pre- viously been considered in models of spiders in farmland. ACKNOWLEDGMENTS The author was supported by the Danish Environmental Research Programme within the center ‘Changing Landscapes — The Cen- tre for Strategic Studies in the Cultural Envi- ronment, Nature and Landscape History’ and by ARLAS center under the ‘Area usage: The farmer as a landscape manager’ program. REFERENCES Anonymous. 1996. Danmarks Statistik, Landbmg. 1996. Kpbenhavn. Anonymous. 1997. Statistisk Arbog, Danmarks Statistik 1997. Kpbenhavn. Caswell, H. & A.M. John. 1992. From the individ- ual to the population in demographic models. Pp. 36-61. In Individual-based Models and Ap- proaches in Ecology: Populations, Communities and Ecosystems. (D.L. DeAngelis & L.J. Gross, eds.). Routledge, Chapman and Hall, New York. Dean, D.A. & W.L. Sterling. 1985. Size and phe- nology of ballooning spiders at two locations in eastern Texas. J. ArachnoL, 13:111-120. DeAngelis, D.L. & K.A. Rose. 1992. Which indi- vidual based approach is most appropriate for a given problem? Pp. 67-87. In Individual-based Models and Approaches in Ecology: Populations, Communities and Ecosystems. (D.L. DeAngelis & L.J. Gross, eds.). Routledge, Chapman and Hall, New York. Duffey, E. 1956, Aerial dispersal in a known spider population. J. Anim. EcoL, 25:85-111. Dunning, J.B., B.J. Danielson & H.R. Pulliam. 1992. Ecological processes that affect popula- tions in complex landscapes. Oikos, 65(1): 169- 175. Greenstone, M.H., C.E. Morgan, A-L. Hultsch, R.A. Farrow & J. E, Dowse. 1987. Ballooning spiders in Missouri, USA, and New South Wales, Australia: Family and mass distributions. J. Ar- achnoL, 15:163-170. Halley, J.M., C.F.G. Thomas & PC. Jepson. 1996. A model for the spatial dynamics of linyphiid spiders in farmland. J. Appl. EcoL, 33:471-492, Kupfer, J.A. 1995. Landscape ecology and bioge- ography. Prog. Phys. Geog., 19(1): 18-34. Lima, S.L. & PA. Zollner. 1996. Towards a be- havioural ecology of ecological landscapes. TREE, 11:131-134. Sunderland, K.D., C.J. Topping, S. Ellis, S. Long & S. van der Laak. 1996. Reproduction and sur- vival of linyphiid spiders with special reference to Lepthyphantes tenuis (Blackwall). Pp, 81-95. In Arthropods Natural Enemies in Arable Land II: Survival, reproduction and enhancement. (C.J.H. Booij & L.J.M.E den Nijs, eds.). Acta Jutlandica, Vol. 71(2). Topping, C.J, 1997. Predicting the effects of land- scape heterogeneity on the distribution of spiders in agroecosystems using a population dynamics driven landscape scale simulation model. Biol. Agric. Hortic., 15:325-336. Topping, C.J, & K.D. Sunderland. 1994. A spatial population dynamics model for Lepthyphantes tenuis (Araneae: Linyphiidae) with some simu- lations of the spatial and temporal effects of farming operations and land-use. Agric. Ecosyst. Environ., 48:203-217. Topping, C.J. & K.D. Sunderland. 1996. Estimat- ing the mortality rate of eggs and first free-living instar Lepthyphantes tenuis (Araneae: Linyphi- idae) from measurements of reproduction and de- velopment. Pp. 51-61 . In Arthropods Natural Enemies in Arable Land II: Survival, Reproduc- tion and Enhancement. (C.J.H. Booij & L.J.M.E, den Nijs, eds.). Acta Jutlandica, Vol. 71(2). Topping, C.J. & K.D. Sunderland. 1998. The pop- ulation dynamics and dispersal of Lepthyphantes tenuis in an ephemeral agro-ecosystem. Entomol. Exp. AppL, 87:29-41. van Wingerden, W.K.R.E. 1980. Aeronautic dis- persal of immature of two linyphiid spider spe- cies (Araneae, Linyphiidae). Proceedings, 8th In- tern. ArachnoL Congr. (J. Gruber, ed.), Pp. 91- 96. Egerman, Vienna. Weyman, G.S. 1993, A review of the possible causative factors and significance of ballooning in spiders. Ethol. EcoL EvoL, 5:279-291. Weyman, G.S., PC. Jepson & K.D. Sunderland. 1995. Do seasonal changes in numbers of aeri- ally dispersing spiders reflect population density on the ground or variation in ballooning moti- vation? Oecologia, 101:487-493, Manuscript received I May 1998, revised 2 Novem- ber 1998. 1999. The Journal of Arachnology 27:387-396 THE HOWS AND WHYS OF SUCCESSFUL PEST SUPPRESSION BY SPIDERS: INSIGHTS FROM CASE STUDIES Susan E. Riechert: Department of Ecology & Evolutionary Biology, University of Tennessee, Knoxville Tennessee 37996-1610 USA ABSTRACT. We can identify agricultural systems in which spiders might best be applied in pest sup- pression from study of the mechanisms by which spider populations influence prey in natural ecosystems. Theory predicts that prey control is achieved through the development of a stable interaction between predator and prey populations. Two models have been applied to predator control of prey, limit cycle and equilibrium point or focal control. Limit cycle control is exerted on a prey species population by a predator species that tracks the densities of its prey. Although the limit cycle approach is commonly applied to pest control situations, the long life cycles and generalist feeding habits of spiders limit their abilities to exhibit density-dependent tracking of their prey. Crops with short growing seasons and species-depauperate systems are the best candidates for limit cycle influences of spiders on prey. Spider populations that exhibit an uneven age-structure and have strong migratory/aggregational tendencies would offer the greatest pest suppression in these simple systems. Equilibrium point/focus control involves the limiting effects of an assemblage of polyphagous feeders on an assemblage of prey species. Spiders fit this model to a greater extent than they do a limit cycle model of prey control. Agricultural systems that conserve spider densities and species representation through minimal chemical application and the maintenance of ground cover are good candidates for equilibrium point control of prey by spiders. It is also important to recognize that many success stories in agroecosystems do not involve stable interactions between predator and prey populations. Indirect effects (e.g., the cessation of feeding in the presence of a predator) and superfluous killing of prey are two factors that augment the influence of spiders on targeted insect populations. Field biologists search for patterns in nat- ural ecosystems with the ultimate goal of ap- plying the knowledge gained to human bene- fit. Of particular concern to the arachnologist is maximizing the potential of spiders to con- trol insect pests in agroecosystems. Wang (1982) reports that as long as 2000 years ago, Chinese writing states that ‘Tf there is a large gathering of spiders, everything will be sat- isfactory.” This contribution deals with how spider abundance and the mechanisms of prey control they exhibit affect their influence on prey populations in natural and agricultural systems. SPIDER SIGNIFICANCE IN NATURAL ECOSYSTEMS Energy and nutrient flow studies. — One problem inherent in assessing the predatory role of spiders is the logistical difficulty of determining not only what spiders eat, but in what quantities relative to other predators in a system. Our best estimates of spider signif- icance in arthropod community dynamics ac- tually are available from radioisotope tagging experiments completed on a reservation at the Oak Ridge National Laboratories in east Ten- nessee (van Hook 1971; Moulder & Reichle 1972). Van Hook (1971) identified wolf spi- ders of the genus Lycosa, (i.e., Hogna and Ra- bidosa) as the most important predators of herbivorous insects in the Festuca-Andropo- gon old field system he analyzed. These ly- cosids were prominent throughout the plant growing season, while other biomass promi- nent predators (i.e., other Araneae) were most important only in spring and early summer. Van Hook attributed 21.1% of the total mor- tality of herbivorous insects (i.e., orthopterans, hemipterans and homopterans) to predation by Hogna and Rabidosa wolf spiders. (Actually, the percentage of consumer biomass cycling through the two spider populations was un- derestimated in the study, because insect ex- uviae and egg clutches were included as part of the estimate of mortality attributed to non spider sources (van Hook 1971). Spiders in this grassland ecosystem exhibited a mean density of 56 individuals/m^ and a mean bio- mass of 146 mg/m^; other arthropod predators in the system “were not present in sufficient 387 388 THE JOURNAL OF ARACHNOLOGY biomass to warrant consideration” in van Hook’s (1971) analyses. Moulder & Reichle (1972) followed the movement of the radionucleotide, cesium- 137, through cryptozoan food chains on the labo- ratory’s reservation. Forty spider species were represented in the samples collected from the litter community. They exhibited a mean den- sity of 126 individuals/m^ and a dry weight biomass of 43 mg/m^. Thus the greater spider densities in this system compared to the grass- land system van Hook (1971) investigated were offset by the smaller sizes of individual spiders. Centipedes (Chilopoda) and preda- ceous beetles (Coleoptera) were the other nu- merically and biomass prominent predatory groups in the forest floor community. Spiders, however, were both numerically more abun- dant (2.7 times that of either the centipedes or the predaceous beetles) and had a total bio- mass that was 1 .4 times greater than the other classes of invertebrate predators. Moulder & Reichle (1972) attributed the numerical and biomass prominence of spiders in the forest floor system to their greater predatory effect on herbivorous insects: spiders consumed 77.8% of the herbivorous prey biomass lost to predation, while centipedes consumed 14.6% and coleopterans 7.7%, respectively. Allochthonous food sourccs.^—The two classic studies described above are represen- tative of the field of ecology that deals with the cycling of energy and nutrients through the ecosystem, consisting of both biotic and abiotic components. Ecosystem ecology’s in- terest in top-down effects and trophic cas- cades (e.g., Hairston, Smith & Slobodkin 1960; Cohen et ah 1990; Hunter & Price 1992) has led to another example of the re- lationship between spider abundances and their effects on prey populations in natural systems. There is no general agreement as to the relative importance of top-down versus bottom-up control of food-web structure. Rather, the type of control appears to be sys- tem dependent. Polls & Strong (1996) pro- posed that trophic cascading (top-down ef- fects) is most pronounced where there is an allochthonous source of food to predators. They argue that external food inputs will aug- ment predator numbers to the extent that they can impose control on the lesser abundant res- ident prey and thus have cascading effects on the plants that these primary consumers forage on. In Polls & Hurd’s (1995) island food- webs, beached marine algae and carrion sup- ported extremely large populations of arthro- pod detritivores. The detritivores, in turn, provided more than 90% of the food to spider populations of sizes that were 1=2 orders of magnitude greater than those observed for similar areas not influenced by the detritus. These large populations of island spiders strongly limited terrestrial herbivorous insects and the plant damage they would ordinarily have imposed. Henschel et al. (1996) suggest that emergents from aquatic habitats can have a similar subsidizing influence on spider pop- ulations. They found that spider community richness and biomass are significantly higher in the vicinity of water bodies with emergent insects than in comparable habitats away from water bodies. The authors conclude that such subsidies might be used to augment the role of spiders in agroecosystems. POTENTIAL MECHANISMS OF SPIDER PREY LIMITATION Non-consumptive effects,=^In managing populations of pest insects and the damage they impose on crops, practitioners are inter- ested in predator control of prey. Modeling approaches indicate that prey control by pred- ators is achieved when a stable equilibrium is established between predator and prey popu- lation numbers. In practice, successful control commonly violates the assumptions of a stable equilibrium (Murdoch et al. 1985). 'Predator- induced effects’ and ‘superfluous killing’ are two effects that spiders may have on prey population dynamics that fall outside of stable population interactions as they are non-con- sumptive effects. Predator-induced effects oc- cur as a consequence of the fact that predators and prey are in an escalating evolutionary race. Predators become increasingly more ef- ficient at capturing prey, while prey have evolved responses to predatory cues that per- mit escape from predation. Predator presence thus causes pests to cease feeding, to forage at less favorable sites, and to drop off host plants altogether in an escape response. The resulting effect is usually a slowing of prey population growth, which delays the outbreak phase. However, dropping from a plant to the forest or field crop floor may result in mor- tality as well due to desiccation and predation by generalist predators (ants and spiders in the RIECHERT— HOWS AND WHYS OF SUCCESSFUL PEST SUPPRESSION 389 case of the hemlock woolly adelgid (McClure (1995)). Nakasuji et al. (1973) document the significance of predator-induced effects by lin~ yphiid spiders on tobacco cutworm larvae, Spodoptera litura, by comparing spider exclu- sion cages to open cages. Only 4% of the 60% mortality rate suffered by cutworm larvae was attributed to actual spider predation, another 18% was not related to spider causation, and 38% involved predator induced effects (i.e., larval dispersaUdislodgement from the forag- ing site caused by spider presence). Since there is no ground cover in tobacco, dislodged larvae suffered starvation (Nakasuji et al. 1973). There is evidence that trophic cascades can also be elicited through indirect predator-in- duced effects in which herbivores shift their foraging behavior in response to perceived predation risk. Schmitz et. al. (1997) found that ‘risk’ spider treatments (glued chelicerae) elicited similar avoidance behavior by grass- hoppers feeding on herbaceous plants and grasses in an old field as did ‘predation’ (in- tact predators) spider treatments. Both treat- ments decreased the impact grasshoppers had on grass biomass, evidence for the existence of a trophic cascade in each case. Superfluous killing, also referred to as wasteful killing and overkill, entails capture rates that significantly exceed rates of con- sumption: it includes the partial consumption of multiple prey items and the killing of prey that are never consumed at all, Samu & Biro (1993) observed killing without feeding and partial consumption of prey in the lycosid, Pardosa hortensis, when they offered test sub- jects high prey densities. Riechert & Maupin (1998) also observed high levels of these two facets of superfluous killing in all of the web spider species they tested: the theridiid Achaearanea tepidariorum (61%), the araneid Argiope trifasciata (49%), the dictynid Die- tyna volucripes (20%), the agelenid Agelen- opsis aperta (44%), and the linyphiid Flor- inda coccinea (43%). The numbers in parentheses following the test species names refer to the proportion of prey captured that were not consumed. Thus spider killing of prey was between 1.2 and 2.6 times greater than that required for feeding. Combined then, predator-induced effects and superfluous killing can account for in ex- cess of 80% of spider limiting effects on prey populations. These kinds of non-consumption influences must be considered in assessing the impact of spiders on pests in agroecosystems. Equilibrium models of predator-prey in- teractions.— In the reductionist approach commonly applied to agroecosystems, there is an interest in dealing with a single pest spe- cies problem. The addition of a single preda- tor or parasitoid species to control a particular pest is an attempt to establish a stable limit cycle between predator and prey population numbers (Hassell 1978). This reductionist ap- proach involves the tracking of the size of a prey population by the selected predator/par- asitoid population. Density-dependent track- ing requires that the predator/p arasitoid: 1) have a life span of similar length to that of its prey, 2) is a prey specialist, and 3) exhibits a search behavior pattern that concentrates for- aging in patches of high prey densities while allowing prey refuges to survive (Hassell 1978; Murdoch et al. 1985). In the food- web literature, top-down control and trophic cascades are achieved through an- other model involving stable predator-prey population interactions, stable equilibrium point or focal control. In this mathematical model, population sizes of predators and prey equilibrate at some relative level rather than cycling out of phase of one another, which is characteristic of limit cycle control (De- Angelis et al.l975; Tanner 1975). To achieve a stable equilibrium, there must be one or more polyphagous predator populations and an assemblage of prey types. As predator en- counter rates with prey change in space or time, individual predators will switch feeding concentration among these prey types (Mur- doch & Oaten 1975; Beddington et al.l978). In addition, the predators must not be limited by local prey availabilities in the immediate sense. Rather, they are expected to have some mechanism of self-damping that keeps their population numbers below the limits of prey availability (e.g., energy-based territoriality, cannibalism or forced migration) (DeAngelis et al. 1975; Tanner 1975; Post & Travis 1979; Erlinge et al. 1984). These behaviors often are evolutionarily adjusted to averages or lows in prey availabilities for particular habitats (e.g., the funnel-web spider, Agelenopsis aperta (Riechert 1981)). 390 THE JOURNAL OF ARACHNOLOGY IMPLEMENTING SPIDER CONTROL IN AGRICULTURAL SYSTEMS: THE NEED FOR SYSTEM SPECIFIC PROTOCOLS Single spider species on single pest spe- cies: limit cycle control. — Many pest insects are r-selected and thus have short generation times and high reproductive potentials. Be- cause spiders have much longer generation times and comparatively low fecundities, they generally will not develop stable limit-cycles with their insect prey. It may be possible to achieve stable cycling between a spider spe- cies and a particular pest in a simple system with just a few herbivores and/or a crop with a very short life cycle. An uneven age distri- bution exhibited by a spider population (e.g., the multivoltine lycosids (e.g., Pardosa lugub- ris (Walckenaer): Edgar 1971) and linyphiids (e.g., Erigone arctica (White): van Wingerden & Vugts 1974)) and a strong aggregational numerical response to prey densities (e.g., ae- rial ballooning by linyphiids in response to lo- calized weather conditions and densities; see review in Riechert & Gillespie (1986)) would also permit some density-dependent tracking of the prey population by a spider species population. There would be fewer generations of pest population build-up in the short-season crop, and the predators would be concentrat- ing foraging on encounter with the numeri- cally prominent prey, the pest. Obviously, the best case scenario for suc- cessful control by limit cycle would involve spider species with the characteristics listed above feeding on a pest in a crop with a short life cycle. An example of a system that meets the short-crop season criterion is spring barley in Sweden. Chiverton (1986) found that liny- phiids successfully control cereal aphids, Rho- palosiphum padi, in this northern region be- cause R. padi overwinters here only in the egg stage. The growing season is simply too short to permit the sowing of a fall grass or cereal crop that would permit the build-up of large R. padi populations of viviparae before spider emigration in May and June. Aphid densities in barley field plots enclosed early enough to prevent spider emigration were six times high- er than those observed in unenclosed plots in Chiverton’s (1986) study. Spider species assemblage control of pest species assemblages: stable equilibrium point or focal control. — There are objective reasons for implementing a holistic approach (stable equilibrium point or focus control) to pest suppression by spiders as opposed to the reductionist approach, a single predator spe- cies acting on a single pest species (limit cy- cle). Stable equilibrium point control or focus (Tanner 1975) is the predator-prey model that is associated with top-down and cascading ef- fects in natural food webs (Post & Travis 1979) and spiders exhibit the traits requisite to stable equilibrium point control of prey by predators (Riechert et al. 1999). They are, in fact the prominent predators of insects in nat- ural ecosystems and this occurs despite the fact that they are self-damped by territorial and cannibalistic behaviors (Edgar 1969; Riechert 1981; Wagner & Wise 1996). Self- damping behavior is actually a necessary con- dition of stable equilibrium-point control of prey by polyphagous predators. Thus spiders are well-suited to this community level ap- proach to prey control. In addition, the fact that suppression of a prominent pest is often followed by new problems with secondary pests favors a community approach where an assemblage of predators influences the entire assemblage of pest species in a local system. An example of a holistic approach to pest suppression by spiders is offered in Riechert & Bishop’s (1990) study of spider assemblage effects on herbivorous insects in mixed veg- etable garden systems. Riechert and Bishop observed a highly significant spider assem- blage effect on pest insects (60-80% reduc- tion in pest-induced plant damage) across a wide variety of vegetable types. The effect was achieved by grass-hay mulch applica- tions, which augmented spider population densities thirty-fold over those observed in tilled control plots. Contrasts completed on spider predation effects in the mulched plots indicated that spiders significantly suppressed insects and thereby afforded less plant damage in mulched than in bare-ground control plots. On the other hand, pest numbers and plant damage were not significantly different be- tween controls and mulched plots from which spiders were systematically removed. Although the analysis of variance was com- pleted on spider densities alone, Riechert & Bishop (1990) presented additional analyses of the quadrat sampling of spiders as well as the results of timed watches of foraging activ- ity within the same system. Calculations made RIECHERT— HOWS AND WHYS OF SUCCESSFUL PEST SUPPRESSION 391 on the data set from the quadrat sampling pro- duced an average spider diversity (Shannon- Wiener H’) in the tilled control plots of 0.94 compared to an average of 2.48 in the mulch and mulch + flowers plots (Pielou 1974). In a five guild system, the nocturnal running spi- der guild was totally absent from the tilled control plots, while all guilds were well rep- resented in the mulch and mulch + flowers plots. Fifteen families of spiders were ob- served during the course of the foraging ob- servations, six of which were web-building families. Most of the spider families were ob- served feeding on more than one pest species with six families feeding on almost the entire range of 13 insect pests observed during the course of the watches. Riechert & Bishop (1990) conclude from these results that the significant effect of spiders on pest insects in the mixed vegetable system was an assem- blage effect, rather than the effect of just a few prominent spider species. Augmentation: sheer numbers versus species richness.—^Regardless of the mecha- nisms by which control is achieved, all evi- dence indicates that successful pest suppres- sion by spiders will best be achieved through the maintenance of high spider densities and in many cases also high species diversities. Maximization of spider densities and species richness are steps that logically must be taken in agricultural systems to increase the bene- ficial functioning of spiders in them. Numerous studies support the idea that spi- der effects on prey are approximately a func- tion of spider versus prey densities/biomasses in a system (e.g., natural communities: Moul- der & Reichle 1972; Polls & Hurd 1995; Hen- schel et al. 1996; Kajak 1997; coconut: Sath- iamma 1995; com: Laub & Luna 1991; Clark et al. 1994; Coll & Bottrell 1995; cotton: Ster- ling et al. 1989; mixed vegetables: Riechert & Bishop 1990; old fields: Riechert & Lawrence 1997; pastures: DeBarro 1992; rice: Sasaba et al. 1973; Wang 1982; Graze & Grigarick 1989; Litsinger et al. 1994; soybeans: Carter & Rypstra 1995; wheat: Hausamman 1996). The significance of increasing species diver- sity is less clear. From summary analyses of the performance of prominent single spider species versus spider species assemblages, Provencher & Riechert (1994) and Riechert & Lawrence (1997) in different experiments found that more than 70% of total reduction in prey biomass and numbers (over that ex- hibited in spider removal controls) is contrib- uted by single spider species. This would seem to indicate that having different foraging strategies represented is less important than mere numbers or biomass. However, Riechert et al. (1999) encountered very different results when they considered the predatory perfor- mance of spider species assemblages versus those of single prominent spider species over time (i.e., a four month period). The spider species assemblage was far more temporally consistent in its predation effects on the broad spectrum of prey types encountered in the old field system than was any single spider spe- cies (four numerically/biomass prominent spe- cies tested). Further, no single spider species performed as well over time towards a partic- ular prey category as did the spider species assemblage, despite the fact that all of the predators used in the single species treatments were maintained at high densities throughout the four months of the study. Each spider spe- cies did show changes in linear dimensions, mass, and reproductive status that correspond- ed to its own unique life cycle at various times during the period of the study. Therefore, spi- ders show changes in their diets over time, a factor that makes it important to have a di- versity of spider species in longer-lived crop systems. Sathiamma (1995) reached a similar con- clusion in a study of natural enemy suppres- sion of the white spider mite, OUgonychus is- eilemae on coconut foliage (See also Agnew & Smith 1989 for spider suppression of pests in peanuts). The total predator density (seven prominent species) corresponded closely over time to the density of the mite and control was adequate to eliminate the need for chemical applications. No single predator, however, was abundant at all potential peak periods of mite density and Sathiamma concluded that the suppression effect of any single predator spe- cies by itself on the pest would be insignifi- cant. This is not to say that single spider species might not be able to exert sufficient control to eliminate the need for chemical intervention in some systems (see also section on limit cy- cle control). It may be that in exhibiting high population densities at critical times in the life cycle of a pest, a single spider population may suppress that pest sufficiently to require only 392 THE JOURNAL OF ARACHNOLOGY minimal chemical intervention during periods when the predator's life cycle is out of phase with the pest. Such a special case may account for the successful use of Lycosa pseudoan- nulata in controlling green rice leafhopper in Japan (Kiritani & Kakiya 1975). This predator imposes highly significant reductions of over- wintering leafhoppers that prevent the early season transmission of rice diseases. A sys- tems model has been developed that incor- porates the densities of this predator and those of the pest in determining when particular paddies need to be treated with insecticides (Kiritani & Sasaba 1978). It is important to note that while simulation models may key on the densities of prominent spider species in the management of a partic- ular pest, it is probable that any conservation scheme designed for a prominent spider will also foster other spider species populations as well. While these other species might have lesser individual roles, their cumulative effect on pests and crop damage can be significant. Spiders and agroecosystem practices. — • Successes with spider suppression of pests in agroecosystems are correlated most frequently with increased predator densities, though re- cent studies indicate that species richness may be an important component as well. Augmen- tation of spider densities and accompanying species richness in agroecosystems could in- clude the following three practices: 1) restric- tion of chemical pesticide applications to an as needed basis, 2) habitat diversification, and 3) maximization of allochthonous inputs. 1). Restriction of chemical pesticide use. — The literature on the deleterious effects of chemical insecticides on spider communi- ties is too substantial to attempt to cover here. I include only two systems (rice and cotton) for which simulation models have been de- veloped that favor the conservation of spiders through the monitoring of pest and predator ratios and selective use of pesticides where warranted. Spiders are used effectively in the control of rice pests in southeast Asia. Kiritani (1977) reported that the regular application of a broad spectrum insecticide to control a rice stem borer in Japan decimated spider popu- lations but had little effect on the leaf- and planthoppers that transmit viral diseases in rice. It took ten years for the spider popula- tions to sufficiently recover from exposure to the insecticide to be useful in suppressing the major new problem of insect transmitted viral diseases. In China, Wang (1982) found that when spiders moved into an area of planthop- per infestation, they reduced the pest: predator ratio from 9 to 1.5:1 within 10 days in one study area and from 5:1 to 0.03:1 within 5 days in another. After implementing conser- vation practices that fostered spider density increases (e.g., encouraging the movement of predators from early rice plantings to 2"^ plantings and using cultural practices that lim- ited the frequency of pesticide application and quantities of chemicals used), the need for chemical pesticides decreased as much as 80% with no measurable loss in rice yields. Monitoring programs of predator density/pest density ratios and weather are used in rice in Japan to determine when pesticide applica- tions are required (Sasaba et ah, 1973). A sim- ilar systems model has been developed for cotton by researchers at Texas A&M Univer- sity (e.g., Hartstack & Sterling 1989; Sterling et al. 1992). The TEXCIM model is widely used in Texas cotton to predict when it is eco- nomically beneficial to apply chemical insec- ticides rather than to rely on natural enemy (primarily spider) suppression of pests. Bree- ne et al. (1990) tested the predation compo- nent of the TEXCIM model for cotton, ob- taining evidence that natural enemies protect the Texas cotton crop for 95% of the crop days, whereas chemical pesticides provide control for only 5% of the season (See also Mansour (1987) on spider control of cotton pests in Israel). 2). Habitat diversification. — This problem has been addressed in a number of ways, in- cluding the addition of weed strips, the main- tenance of uncultivated borders, intercrop- ping, and the application of mulch and other ground covers. The first three approaches have produced mixed results. Jmhasly & Nentwig (1996) report that the addition of weed strips to winter wheat did lead to higher web-spider densities in the vicinity of the strips, but that this increase did not lead to protection of the wheat crop from pest damage. Riechert & Bishop (1990) did not find that alternation of rows of vegetables with flowering buckwheat augmented spider numbers compared to bare- ground controls, nor was plant damage less in the plots containing the flowering buckwheat. A potential increase in spider diversity is one pest suppression benefit of intercropping. RIECHERT— HOWS AND WHYS OF SUCCESSFUL PEST SUPPRESSION 393 but this suppression effect may be confounded by the fact that intercropping also reduces ovi- position sites for the pests. Roltsch & Gage (1990) report that tomatoes intercropped with beans in the control of the potato leafhopper, Empoasca fabae, did not increase natural en- emy densities nor species richness, but nev- ertheless suppressed pest densities. The avail- ability of oviposition sites was the important factor in this study. On the other hand. Coll & Bottrell (1995) found that spiders and nabid hemipterans had a greater effect on Mexican bean beetle (Epilachna varivestis) populations in dicultures of maize and bean vegetation than in bean monocultures. The lower bean beetle numbers were correlated with the high- er abundances of the predators in the dicul- ture. The problem I have with all three of the above habitat management techniques con- cerns the question of whether the augmented spiders will move into the crop where the pest problem exists versus stay in more structurally complex natural habitats. Bishop & Riechert (1990) did not find that spiders from surround- ing natural communities (i.e., old field, oak- hickory woods, and briar) colonized a mixed vegetable garden system. Over 60% of the species present in the garden system were not even collected in neighboring habitats and ex- perimental limitation of ground dispersal in- dicated that most of the colonization occurred through ballooning. Greater success has been achieved with the addition of ground cover and structure in an- nual crop systems (e.g., mulch in vegetables (Riechert & Bishop 1990), artificial habitat structure in soybeans (Carter & Rypstra 1995), and in no-till com through the mowing of the winter cover crop (Laub & Luna 1991)). In the first two studies, experimental increases in spider densities and suppression of pest numbers and crop damage were noted. Laub & Luna (1991) found that by mowing a winter rye cover crop rather than spraying it at com planting time, they achieved a signif- icant net economic benefit (US $91~$113/ hectare). They attribute the benefit of the mulch produced by mowing rye to suppres- sion of army worm populations through the conservation of natural enemies in the mowed treatment. (Other work by this lab (Clark et al, 1994) used predator exclusion arenas in com to demonstrate significant carabid and staphylinid beetle, ant, and spider predation effects on army worm damage.) The applica- tion of mulch to an agroecosystem early in the season may provide more favorable thermal environments to the spider populations before crop growth is sufficient to provide cover. It also may provide an abundance of early food in the form of subsidies from the detritivore food chain (e.g., collembola). 3). Maximization of allochthonous in- puts.— It is clear from Polis & Hurd’s (1995) study that allochthonous inputs from detriti- vore food chains may also subsidize spider population densities, permitting them to have greater cascading effects on crop production. Little is known about the detritivore commu- nity in temporary systems. Turnbull (1966) re- ported that 38% of the food of spider com- munities in overgrazed pastures was contributed by the detritivore food chain; and Kajak (1995) reports that linyphiid, araneid and lycosid spiders generally feed to a large extent on detritivorous dipterans. Sunderland (1975) and Sunderland et al. (1986) demon- strate the significance of the contribution of collembola from the detritivore food chain to linyphiid spiders in cereals. These serve as al- ternative food to polyphagous predators early in the growing season of the crop when pest numbers are low. The build up of the linyphiid populations supported by the collembola per- mits effective suppression of pest numbers lat- er in the season when warm temperatures caused the collapse of the collembola popu- lations and the predators switched to feeding on aphids. Although De Barro (1992) does not directly address the role of detritivores in supporting spider populations in irrigated perennial pas- tures in Australia, this system too is likely subsidized by input from decomposer food chains. De Barro (1992) experimentally dem- onstrated that lycosids and linyphiids signifi- cantly limit the population growth of a prom- inent pest, the cereal aphid Rhopalosiphum padi, in its summer pasture refuges. The re- moval of spiders led to a 15” 16 fold increase in R. padi numbers in experimental plots. The limiting effect spiders imposed on the aphids reduced the number of alates produced in the fall that could colonize and transmit diseases to cereal crops. Rhopalosiphum populations fluctuate greatly with local weather conditions in the pastures, as this aphid suffers 90% mor- 394 THE JOURNAL OF ARACHNOLOGY tality when exposed to temperatures in excess of 39 °C, whereas the spider populations are not affected by summer temperature extremes. The system has no specialist predators or par- asitoids, probably because of the degree to which population numbers of R. padi fluctuate during the summer months. During periods of low cereal aphid densities, the spider com- munity, which consists of approximately 11 species, is supported by alternative prey which I suggest comes from detritivores in the soil and litter. The reader should consult Wise et al. (this volume) for further discussion of the topic of external subsidies. Whatever the mechanism, spiders as agents of biological control can be used in combi- nation with no-till agriculture for the kind of whole ecosystem approach to habitat manage- ment that should be encouraged in modem ag- riculture. While it may be difficult to suggest that a grower decrease pesticide use and apply mulch for its benefits to natural enemies alone, a favorable cost/benefit ratio that might result from these practices in terms of reduced chemical costs; and greater water retention and organic matter accumulation may make the whole ecosystem approach both attractive and practical. LITERATURE CITED Agnew, C.W. & J.W. Smith, Jr. 1989. Ecology of spiders (Araneae) in a peanut agroecosystem. En- viron. EntomoL, 18:30-42. Beddington, J.R., C.A. Free & J.H. Lawton. 1978. Characteristics of successful enemies in models of biological control of insect pests. Nature, 273: 513-519. Bishop, L. & S.E. Riechert. 1990. Spider coloni- zation of agroecosystems: source and mode. En- viron. EntomoL, 19:1738-1745. Breene, R.G., W.L. Sterling & M. Nyffeler. 1990. Efficacy of spider and ant predators on the cotton fleahopper (Hemiptera: Miridae). Entomophaga, 35:393-401. Carter, RE. & A.L. Rypstra. 1995. Top-down ef- fects in soybean agroecosystems: spider density affects herbivore damage. Oikos, 72:433-439. Chiverton, P.A. 1986. Predator density manipula- tion and its effects on populations of Rhopalo- siphum padi (Horn.: Aphididae) in spring barley. Ann. Appl. Biol., 109:49-60. Clark, M.S., J.M. Luna, N.D. Stone & R.R. Young- man. 1994. Generalist predator consumption of armyworm (Lepidoptera: Noctuidae) and effect of predator removal on damage in no-till com. Environ. EntomoL, 23:617-622. Cohen, J.E., E Briand & C.M. Newman. 1990. Conununity food webs: data and theory. Bio- mathematics. Vol. 20. Springer Verlag, New York. Coll, M. & D.G. Bottrell. 1995. Predator-prey as- sociation in mono- and dicultures: effect of maize and bean vegetation. Agric. Ecosyst. & Environ., 54:115-125. DeAngelis, D.L., R.A. Goldstein & R.V. O’Neill. 1975. Stability and connectance in food web models. Ecology, 56:238-245. DeBarro, RJ. 1992. The impact of spiders and high temperatures on cereal aphid {Rhopalosiphum padi) numbers in an irrigated perennial grass pasture in south Australia. Entomologist, 122: 19-26. Edgar, W.D. 1969. Prey and predators of the wolf spider Lycosa lugubris. J. ZooL, London, 159: 405-411. Edgar, W.D. 1971. The life cycle, abundance, and seasonal movement of the wolf spider Lycosa (Pardosa) lugubris in central Scotland. J. Anim. EcoL, 40:303-322. Erlinge, S., G. Goransson, G. Hogstedt, G. Jansson, O. Liberg, J. Loman, I.N. Nilson, T. von Shantz & M. Sylven. 1984. Can vertebrate predators regulate their prey? American Nat., 123:125- 133. Hairston, N.G., EE. Smith & L.B. Slobodkin. 1960. Community stmcture, population control, and competition. American Nat., 94:421-425. Hassell, M.P. 1978. The Dynamics of Arthropod Predator-Prey Ssystems. Princeton Univ. Press, Princeton, New Jersey. Hartstack, A.W. & W.L. Sterling. 1989. TEX- CIM30: The Texas cotton insect model. Texas Agric. Exp. Stn. Publ. MP 1646. Hausammann, A. 1996. The effects of weed strip- management on pests and beneficial arthropods in winter wheat. J. Plant Dis. Prot., 103:70-81. Henschel, J., H. Stumpf, & D. Mahsberg. 1996. Increase of arachnid abundance and biomass at water shores. Proc. XIII Intern. Congr. Arach- noL, Geneva, Rev. Suisse ZooL, vol. hors serie. Pp. 269-278. Hunter, M.D. & RW Price. 1992. Playing chutes and ladders: heterogeneity and the relative roles of bottom-up and top-down forces in natural communities. Ecology, 73:724-732. Jmhasly, P. & W. Nentwig. 1995. Habitat manage- ment in winter wheat and evaluation of subse- quent spider predation on insect pests. Acta Oec- oL, 16(3):389-403. Kajak, A. 1997. Effects of epigeic macroarthro- pods on grass litter decomposition in mown meadow. Agric. Ecosyst. & Environ., 64:53-63. Kajak, A. 1995. The role of soil predators in de- composition processes. European J. EntomoL, 92:573-580. RIECHERT— HOWS AND WHYS OF SUCCESSFUL PEST SUPPRESSION 395 Kiritani, K. 1977. Recent progress in pest manage- ment for rice in Japan. Japanese Agric. Res. Quart., 11:40-49. Kiritani, K. & N. Kakiya. 1975. An analysis of the predator-prey system in the paddy field. Res. Po- pul. EcoL, 17:29-38. Kiritani, K. & T. Sasaba. 1978. An experimental validation of the systems model for predation of rice dwarf virus infection. Appl. Entomol. ZooL, 13(3):209-214. Laub, C.A. & J.M. Luna. 1991. Influence of winter cover crop suppression practices on seasonal abundance of army worm (Lepidoptera: Noctui- dae), cover crop regrowth, and yield in no-till corn. Environ. Entomol., 20:749-754. Litsinger, J.A., N. Chantaraprapha, A.T. Barrion & J.P. Bandong. 1994. Natural enemies of the rice caseworm Nymphula depunctalis (Guenee) (Lep- idoptera: Pyralidae). Insect Sci. Applic., 15(3): 261-268. Mansour, F. 19 87. Spiders in sprayed and unsprayed cotton fields in Israel, their interactions with cot- ton pests and their importance as predators of the Egyptian cotton leaf worm, Spodoptera littoralis. Phytoparasitica, 15(1):31-4L McClure, M.S. 1995. Diaterobates humeralis (Or- batida: Ceratozetidae): An effective control agent of the woolly adelgid (Homoptera: Adelgidae) in Japan. Environ. Entomol., 24(5): 1207-1215. Moulder, B.C. & D.E. Reichle. 1972. Significance of spider predation in the energy dynamics of forest-floor arthropod communities. Ecol. Mon- ogr., 42:473-498. Murdoch, W.W. & A. Oaten. 1975. Predation and population stability. Adv. Ecol. Res., 9:1-131. Murdoch, W.W., J. Chesson & PL. Chesson. 1985. Biological control in theory and practice. Amer- ican Nat., 125:344-366. Nakasuji, E, H. Yamanaka & K. Kiritani. 1973. The disturbing effect of micryphantid spiders on the larval aggregation of the tobacco cutworm Spodoptera litura (Lepidotera: Noctuidae). Kon- tyu, 41:220-227. Graze, M.J. & A. A. Grigarick. 1989. Biological control of aster leafhopper (Homoptera: Cicadel- lidae) and midges (Diptera: Chironomidae) by Pardosa ramulosa (Araneae: Lycosidae) in Cal- ifornia rice fields. J. Econ. Entomol., 82:745- 749. Polls, G.A. & S.Hurd. 1995. Extraordinarily high spider densities on islands: flow of energy from the marine to terrestrial food webs and the ab- sence of predation. Proc. Natl. Acad. Sci. USA, 92:4382-4386. Polls, G.A. & D. Strong. 1996. Food web com- plexity and community dynamics. American Nat., 147:813-846. Post, W.M. & C.C. Travis. 1979. Quantitative sta- bility in models of ecological communities. J. Theoret. Biol., 79:547-553. Provencher, L. & S.E. Riechert. 1994. Model and field tests of prey control effects by spider as- semblages. Environ. Entomol., 23:1-17. Riechert, S.E. 1981. The consequences of being territorial: spiders, a case study. American Nat., 117:871-892. Riechert, S.E. & R.G. Gillespie. 1986. Habitat choice and utilization in web-building spiders. Pp. 23-48. In Spiders: Webs, Behavior and Evo- lution. (WB. Shear, ed.). Stanford Univ. Press, Stanford. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators in a garden test system. Ecology, 71:1441-1450. Riechert, S.E. & K. Lawrence. 1997. Test for pre- dation effects of single versus multiple species of generalist predators: spiders and their insect prey. Entomol. Exp. Appl., 84:147-155. Riechert, S.E. & J. Maupin. 1998. Spider effects on prey: Tests for superfluous killing in five web builders. Pp. 203-210 In Proc. 17**’ European Colloquium Arachnol. (P.A. Selden, ed.). Bull. British Arachnol. Soc. Riechert, S.E., L. Provencher, & K, Lawrence. 1999. The potential of spiders to exhibit stable equilibrium point control of prey: Tests of two criteria. Ecol. Appl., 9:365-377. Roltsch, WJ. & S.H. Gage. 1990. Influence of bean-tomato intercropping on population dynam- ics of the potato leafhopper (Homoptera: Cica- dellidae). Environ. Entomol., 19:534-543. Samu, F. & Z. Biro. 1993. Functional response, multiple feeding and wasteful killing in a wolf spider (Araneae: Lycosidae). European J. Ento- mol., 90:471-476. Sasaba, T, K. Kiritani & T. Urabe. 1973. A prelim- inary model to simulate the effect of insecticides on a spider-leafhopper system in a paddy field. Res. Popul. Ecol., 15:9-22. Sathiamma, B. 1995. Biological suppression of the white spider mite Oligonychus iseilemae (Hirst) on coconut foliage. Entomon, 20:237-243. Schmitz, O.J., A.R Beckerman & K.M. O’Brien. 1997. Behaviorally mediated trophic cascades: effects of predation risk on food web interac- tions. Ecology, 78:1388-99. Sterling, W.L., K.M. El-Zik & L.T Wilson. 1989. Biological control of pest poulations. Pp. 155- 189. In Integrated Pest Management Systems and Cotton Production. (R. Frisbie, K. El-Zik & T. Wison, eds.). Wiley, New York. Sterling, WL., A. Dean & N.M. ABD El-Salam. 1992. Economic benefits of spider (Araneae) and insect (Hemiptera: Miridae) predators of cotton leafhoppers. J. Econ. Entomol., 85:52-57. Sunderland, K.D. 1975. The diet of some preda- 396 THE JOURNAL OF ARACHNOLOGY tory arthropods in cereal crops. J. Appl. Ecol., 12:507-515. Sunderland, K.D,, A.M. Fraser & A.F.G. Dixon. 1986. Distribution of linyphiid spiders in rela- tion to capture of prey in cereal fields. Pedo- biologia, 29:367-375. Tanner, J.T 1975. The stability and the intrinsic growth rates of prey and predator populations. Ecology, 56:855-867. Turnbull, A.L. 1966. A population of spiders and their potential prey in an overgrazed pasture in eastern Ontario. Canadian J. Zool., 44:557-583. van Hook, R.I. Jr. 1971. Energy and nutrient dy- namics of spider and orthopteran populations in a grassland ecosystem. Ecol. Monogr., 41(1): 1- 26. van Wingerden, K.R.E. & H.F. Vugts. 1974. Fac- tors influencing aeronautic behaviour of spiders. Bull. British Arachnol. Soc., 3:6-10. Wagner, J.D. & D.H. Wise. 1996. Cannibalism reg- ulates densities of young wolf spiders: evidence from field and laboratory experiments. Ecology, 77:629-652. Wang, H. 1982. Conservation and utility of paddy field spiders to control pests. Pp. 274-281. In Proc. Chinese Acad. Sci.-US Natl. Acad. Sci. Symposium on Biological Control of Insects. (P.L. Ackleson, ed.). Manuscript received 1 May 1998, revised 24 No- vember 1998. 1999. The Journal of Arachnology 27:397-400 AFTERWORD SUMMARY AND FUTURE DIRECTIONS FOR RESEARCH ON SPIDERS IN AGROECOSYSTEMS Keith D. Sunderland and Matthew H. Greenstone This symposium, using spiders as a focus, has explored a number of themes of general importance in ecology and biological control. These themes relate to grappling with the problems of scale in the distribution and dis- persal of invertebrates, to gaining behavioral insights into the trophic biology of generalist predators, to understanding the functioning of communities and ecosystems, and to launch- ing new initiatives in pest control based on ecosystem engineering. Each of these topics has been illuminated from a range of different angles by the participants, who each ap- proached the symposium with a unique view- point and expertise, and the net effect (ob- tained by reading the full set of Symposium papers) is to provide the reader with a mature and well-rounded appreciation of the subject. The symposium presentations were nothing if not innovative and forward-looking, and they record the forging of some exciting new ap- proaches to the study of ecological processes and the implementation of biocontrol. Distribution and dispersal: from micro- habitat to landscape.— Individual spiders can move between microhabitats within a habitat by walking. Some of the species characteristic of agricultural systems also have the ability to disperse over short and long distances through the air by ballooning, so they can move be- tween habitats within the landscape during the lifetime of an individual (Samu et ah, this vol- ume). Annual recolonization of crops by spi- ders owes more, in many cases, to aerial de- position of aeronauts than to cursorial invasion from refugia adjacent to fields. The aerial arachnofauna is taxonomically rich, its composition is not representative of ground- based communities, and the size distribution of aeronauts is skewed in favor of small spi- ders (Suter, this volume). The small size of aeronauts might not be entirely a result of physical constraints, since a modeling exer- cise predicted that the optimal dispersal strat- egy is one of high juvenile dispersal (Topping, this volume). The probability of ballooning can increase in response to a decline in habitat quality, such as crop senescence; and the tim- ing of these changes in habitat quality varies from habitat to habitat within the landscape (Thomas & Jepson, this volume). We have an increasingly detailed understanding of the proximate biological and meteorological con- straints and cues for the initiation of aerial dis- persal (Suter, this volume), but we know vir- tually nothing about the fate of aeronauts once they are airborne. Ballooning is risky for the individual. The unpredictable air movement at the time of ballooning could take the individ- ual into danger zones. For the species, how- ever, it is part of the equipment needed for efficient exploitation of the resources offered by ephemeral habitats (Thomas & Jepson, this volume). To understand why particular strategies of dispersal have evolved it is necessary to con- sider large spatial areas and long spans of time, and the only practicable way of doing this is through modeling. This aspect of mod- eling is growing rapidly in sophistication and power (Topping, this volume), but it continues to rely on good-quality biological data. In ad- dition to its contribution to developing eco- logical theory, landscape modeling could be of practical value if it can produce robust pre- dictions about the optimal patterning of hab- itats within landscapes, i.e., patterning that will maximize regional populations of natural enemies, including spiders (Samu et ah, this volume). Maybe it will also eventually be pos- sible to incorporate a flavor of the topograph- ical, political and economic factors that con- 397 398 THE JOURNAL OF ARACHNOLOGY tribute to the farmer’s decision making about the spatial distribution of crop types and other land use decisions. Prey selection and pest control. — Preda- tion and consumption of pests or alternative foods can be studied in the laboratory, or by direct observation in the field, by gut analyt- ical methods (e.g., radionuclides, electropho- resis, chromatography, serology), or by field experiments (Greenstone, this volume). Lab- oratory studies are of limited value, serologi- cal gut analysis has proved to be the most ef- ficient technique for large-scale studies of the consumption of selected prey species by spi- ders, and field experiments have demonstrated the value of spiders for biological control. Di- rect observation has provided good data on dietary range and predation rates in the field (Greenstone, this volume). This method has demonstrated that web-spiders are 99% insec- tivorous, whereas hunting spiders have a wid- er diet breadth (Nyffeler, this volume). It is not understood why, despite there being fre- quent agonistic encounters between web spi- ders, these rarely result in intra-guild preda- tion (IGP). Many of the hunters, on the other hand are strongly araneophagic (Nyffeler, this volume). Linyphiidae have been found to feed mainly on small Diptera, Homoptera and Col- lembola, but hunting spiders (Oxyopidae, Thomisidae and Salticidae) eat Diptera, Hy- menoptera, Heteroptera, Homoptera, Lepidop- tera and Araneae (Nyffeler, this volume). Some of the factors affecting prey selection have been studied in the laboratory. Active prey selection appears to be a compromise be- tween maximizing energy intake, balancing nutrients and minimizing toxin consumption. Aversions to distasteful, toxic prey (e.g., some aphid species) can be learned by spiders, yet forgotten in less than a day (Toft, this vol- ume). The intriguing possibility has been raised that in some situations pest control might even be improved if the pest is distaste- ful because of the operation of wasteful killing (= superfluous killing) and unsatisfied spider food demand (Sunderland, this volume). The role of prey quality in determining spider fit- ness is proving to be a very complex issue. This is a fertile area of current research that is revealing some unexpected facets of the tro- phic biology of spiders. For example, spiders do not always choose the optimal combination of prey from a mixture, toxic prey in a mixed diet may inhibit consumption of high quality prey, but if small quantities of other types of toxic prey are consumed they may even im- prove spider performance (Toft, this volume). “Limit cycle” control of pests, involving tracking of pest population density by the predator population, is not, generally, a mech- anism of pest control that can be attributed to spiders. “Stable equilibrium point or focal control,” on the other hand, is a pest control mechanism suitable for spiders and other po- lyphagous predators that display prey switch- ing and whose populations can be self-limited by cannibalism or territoriality. It relates more to pest control by assemblages of species, rather than to single species of spiders (Riech- ert, this volume). Spider species often have complementary niches and so an assemblage of species may be able to attack all growth stages of a pest, thus reducing “enemy-free” space and improving the prospects for effec- tive biological control (Sunderland, this vol- ume). Spiders have some additional attributes that increase their value as biocontrol agents. These include a) pest dislodgment, b) the ca- pacity of webs to kill pests even when the spider is absent or unmotivated to attack, and c) wasteful killing and partial consumption (Riechert, this volume; Sunderland, this vol- ume). Whatever the mechanism, solid evi- dence has accumulated, mainly from field ex- periments, that spider assemblages can be effective in reducing pest populations and the crop damage that they cause (Greenstone, this volume; Riechert, this volume; Rypstra et al. this volume; Sunderland, this volume; Wise et al., this volume). Communities and ecosystems. — A major theme here, picked up by various contributors (Riechert, this volume; Sunderland, this vol- ume; Wise et al., this volume) and treated in different ways, is the realization that spiders, as polyphagous predators, can get a subsidy from the detritivore food chain, and that this can boost their impact on herbivores, includ- ing pests. It can be argued that ways should be found to apply this principle in agriculture and that there should be research into farming- compatible techniques to increase the detriti- vore component in a wide range of crops. There could, however, be the penalty that more choice of food may mean that spiders and other generalist predators refuse to eat the pests, especially in cases where the pest spe- SUNDERLAND & GREENSTONE— AFTERWORD 399 cies are distasteful and toxic (Toft, this vol- ume). Clearly, there is a need for studies di- rected at determining the outcome of increasing the amount of prey biodiversity in agroecosystems, and at determining the mech- anism by which this affects pest control (Wise et ah, this volume). There is some evidence that spider predation on detritivores and fun- givores can depress rates of litter decompo- sition and nutrient mineralization in agroeco- systems, but this negative effect is expected to be more than offset by the positive effect of spider predation on pests (Wise et al., this volume). A promising approach to the study of how spider assemblages, as part of a community, affect pest populations, is to see how they fit into functional guilds, rather than always treating them taxonomically. There are indi- cations that this approach is already throwing up some commonalities of community orga- nization that apply across a spectrum of crops (Uetz et al., this volume), but knowledge of the mechanism(s) underlying these findings awaits further research. There is still a dearth of behavioral and life history information for many species, and this is hindering the devel- opment of guild classifications and of quan- titative comparative guild studies (Uetz et al., this volume). The suggestion that the exact taxonomic composition of these guilds de- pends heavily on the composition of assem- blages in nearby non-agricultural habitats (Uetz et al., this volume) underscores the need for a better understanding of the role of dis- persal in the assembly of spider guilds in agro-ecosystems (Samu et al., this volume), especially since the contrary has been ob- served in experimental garden systems (Riechert, this volume). A rich complexity of interactions (including various types of competition and intraguild predation (IGP)) can occur between natural enemies in agroecosystems. Some of these in- teractions are thought to buffer the conununity from change, while others have been shown to destabilize pest control (Sunderland, this volume; Wise et al., this volume). Both ex- ploitation and interference competition can occur between spider species, especially where preferred microhabitats overlap (Mar- shall & Rypstra, this volume). Subtle com- petitive interactions may also be occurring, but investigations of these are still at an early stage. For example, preliminary results from laboratory mesocosm experiments with two lycosid species suggested that foraging activ- ity of Pardosa milvina was reduced in the presence of Hogna helluo, even though the two species have contrasting diel activity cy- cles. A kairomone that alerts Pardosa to the presence of a potential predator may be in- volved (Marshall & Rypstra, this volume). It is hypothesized that complete elimination of a competing species from the crop may be averted if a top predator (such as the strongly araneophagic green lynx spider, Peucetia vir- idans) reduces the density of the dominant competitor (i.e., exploiter-mediated coexis- tence, as applied to predators) (Sunderland, this volume). Cannibalism and IGP may also enable populations of spiders and other pred- ators to persist in a habitat during periods of low abundance of herbivore and detritivore prey (Marshall & Rypstra, this volume). IGP involving spiders has been studied in natural communities, and IGP involving predators other than spiders has been studied in agroe- cosystems (sometimes demonstrating that in- tense predation by one predator on another may release a pest from a former level of sat- isfactory biological control); but IGP involv- ing spiders in agroecosystems has not yet been investigated experimentally (Hodge, this vol- ume). IGP by lycosid spiders on the insect predators of squash bug eggs was, however, suspected as the explanation for reduced squash production in summer vegetable gar- den experiments (Wise et al., this volume), and there is a wealth of observational data on the involvement of spiders in IGP relation- ships in agroecosystems (Nyffeler, this vol- ume). Quantification of density, activity and diet of spiders and co-occurring predators in agroecosystems will enable prediction of the probability of IGP and competition which can then be used to guide the design of rigorous and meaningful field experiments (Hodge, this volume). Modifications to agricultural practice. — Ways are being sought to promote the effec- tive use of spiders in biological control; but it should be noted that spiders will, for the fore- seeable future, be embedded in integrated management systems which are likely to con- tinue to include some use of pesticides. The selective use of pesticides so that they work with, rather than against, natural enemies 400 THE JOURNAL OF ARACHNOLOGY (Riechert, this volume), needs development, and can only be based on a sound understand- ing of the ecotoxicology of spiders and other natural enemies. Our knowledge of the eco- toxicology of spiders is lagging significantly behind that of some other generalist predators, such as carabid beetles. A strong relationship between spider density and habitat structure has been demonstrated by correlations and ex- perimental manipulations. Measures that in- crease the structural complexity of the habitat, such as intercropping, mulching and conser- vation tillage, are known to enhance spider density and diversity (Rypstra et ah, this vol- ume). Diversification is most likely to be ef- fective if it comes in the form of interspersed microhabitats, rather than spatially-segregated microhabitats or habitats (Samu et ah, this volume). How the specific details of habitat structure influence the effectiveness of spiders as biological control agents has yet to be worked out. Conservation tillage and mulches are examples of approaches that could simul- taneously provide spiders with a more diver- sified habitat structure and a nutrient and en- ergy boost from the detritivore food chain (Samu et ah, this volume; Riechert, this vol- ume; Rypstra et ah, this volume; Wise et ah, this volume). This topic justifies theoretical, experimental and applied research in the fu- ture (Wise et ah, this volume). Contents continued from outside back cover Effects of Short-Term Sampling on Ecological Characterization and Evaluation of Epigeic Spider Communities and Their Habitats for Site Assessment Studies by Uwe Riecken 189 Distribution and Natural History of Mexican Species of Brachypelma and Brachypelmides (Therophosidae, Theraphosinae) with Morphological Evidence for Their Synonymy by A. Locht, M. Yanez & I. Vazquez 196 Common Ground-Living Spiders in Old Tiaga Forests of Finland by Seppo Koponen 20 1 Abundance and Phenology of Schizomida (Arachnida) from a Primary Upland Forest in Central Amazonia by J. Adis, J. Reddell, J. Cokendolpher & J.W. de Morais . 205 Relationship of Habitat Age to Phenology Among Ground-Dwelling Linyphiidae (Araneae) in the Southeastern United States by Michael L. Draney & D.A. Crossley, Jr 211 House Spiders of Kansas by Hank Guarisco 217 Spider and Harvestman Communities Along a Glaciation Transect in the Italian Dolomites by Vito Zingerle 222 Salticidae (Arachnida, Araneae) of Islands Off Australia by Barbara Paloleta & Marek Zabka 229 Pseudoscorpions in Field Margins: Effects of Margin Age, Management and Boundary Habitats by J.R. Bell, S. Gates, A.J. Haughton, D.W. Macdonald, H. Smith, C.P. Wheater & W.R. Cullen 236 Comparative Analyses of Epigeic Spider Assemblages in Northern Hungarian Winter Wheat Fields and Their Adjacent Margins by Ferenc Toth & Jozsef Kiss 241 The Effects of Different Rates of the Herbicide Glyphosate on Spiders in Arable Field Margins by A.J. Haughton, J.R. Bell, N.D. Boatman & A. Wilcox 249 A Faunistic and Zoogeographical Review of the Spiders (Araneae) of the Balkan Peninsula by Christo Deltshev 255 Symposium on Spiders in Agroecosystems Why a Symposium on Spiders in Agroecosystems Now? by Matthew H. Greenstone & Keith D. Sunderland 267 Guild Structure of Spiders in Major Crops by G.W. Uetz, J. Halaj & A.B. Cady . . . 270 An Aerial Lottery: The Physics of Ballooning in a Chaotic Atmosphere by Robert B. Suter 281 Differential Aerial Dispersal of Linyphiid Spiders from a Grass and a Cereal Field by C. F.G. Thomas & P.C. Jepson 294 Prey Choice and Spider Fitness by Spren Toft 301 Mechanisms Underlying the Effects of Spiders on Pest Populations by Keith Sunderland 308 Prey Selection of Spiders in the Field by Martin Nyffeler 317 Scale-Dependent Dispersal and Distribution Patterns of Spiders in Agricultural Systems: A Review by F. Samu, K.D. Sunderland & C. Szinetar 325 Spider Predation: How and Why We Study It by Matthew H. Greenstone 333 Spider Competition in Structurally Simple Ecosystems by Samuel D. Marshall & Ann L. Rypstra 343 The Implications of Intraguild Predation for the Role of Spiders in Biological Control by Margaret A. Hodge 351 Spiders in Decomposition Food Webs of Agroecosystems: Theory and Evidence by D. H. Wise, W.E. Snyder, P. Tuntibunpakul & J. Halaj 363 Architectural Features of Agricultural Habitats and Their Impact on the Spider Inhabitants by A.L. Rypstra, P.E. Carter, R.A. Balfour & S.D. Marshall .... 371 An Individual-Based Model for Dispersive Spiders in Agroecosystems: Simulations of the Effects of Landscape Structure by C.J. Topping 378 The Hows and Whys of Successful Pest Suppression by Spiders: Insights from Case Studies by Susan E. Riechert 387 Summary and Future Directions for Research on Spiders in Agroecosystems by Keith D. Sunderland & Matthew H. Greenstone 397 CONTENTS The Journal of Arachnology Volume 27 Congress Papers Number 1 Historic Overview of Past Congresses of Arachnology and of the Centre International de Documentation Arachnologique (C.I.D.A.) by Otto Kraus 3 The Genus Attidops (Araneae, Salticidae) by G.B. Edwards 7 A New Disembolus (Araneae, Linyphiidae) from Cape Cod, Massachusetts and Long Island, New York by Robert L. Edwards 16 Sinopoda, a New Genus of Heteropodinae (Araneae, Sparassidae) from Asia by Peter Jager 19 Carbinea, a New Spider Genus from North Queensland, Australia (Araneae, Amaurobioidea, Karabininae) by Valerie Todd Davies . 25 Spiders of the Genus Heptathela (Araneae, Liphistiidae) from Vietnam, with Notes on their Natural History by Hirotsuga Ono 37 On the Phylogenetic Relationships of Sisicottus hibemus (Araneae, Linyphiidae, Erigoninae) by Jeremy Zujko-Miller 44 Towards a Phylogeny of Entelegyne Spiders (Araneae, Araneomorphae, Entelegynae) by C.E. Griswold, J.A. Coddington, NJ. Platnick & R.R. Forster . 53 Hypotheses for the Recent Hispaniolan Spider Fauna Based on the Dominican Republic Amber Spider Fauna by David Penney 64 An Adaptive Radiation of Hawaiian Thomisidae: Biogeographic and Genetic Evidence by Jessica E. Garb 71 Comparison of Rates of Speciation in Web-building and Non-web-building Groups Within a Hawaiian Spider Radiation by Rosemary G. Gillespie 79 Fossil Evidence, Terrestrialization and Arachnid Phylogeny by Jason A. Dunlop & Mark Webster 86 Cephalothoracic Sulci in Linyphiine spiders (Araneae, Linyphiidae, Linyphinae) by Gustavo Hormiga 94 Spermatophores and the Evolution of Female Genitalia in Whip Spiders (Chelicerata, Amblypygi) by Peter Weygoldt 103 Ontogeny of Characteristic Leg Macrosetae in Mimetus (Araneae, Mimetidae) by B. Cutler, H. Guarisco & D.J. Mott 117 Ventral Mesosomal Changes in Embryos from Three Scorpion Families: luridae, Buthidae and Vaejovidae by Roger D. Farley 123 The Use of Morphometric Characteristics for the Recognition of Species Among Goniosomatine Harvestmen (Arachnida, Opiliones, Gonyleptidae) by Pedro Gnaspini 129 Sexual Selection in Pholcid Spiders (Araneae, Pholcidae): Artful Chelicerae and Forceful Genitalia by Bernhard A. Huber 135 A Comparison of the Respiratory Systems in Some Cave and Surface Species of Spiders (Araneae, Dysderidae) by M. Kuntner, B. Sket & A. Blejec 142 A New All-Female Scorpion and the First Probable Case of Arrhenotoky in Scorpions by Wilson R. Louren^o & Orlando Cuellar 149 Discovery of a Sexual Population of Tityus serrulatus, One of the Morphs Within the Complex Tityus stigmurus (Scorpiones, Buthidae) by Wilson R. Louren^o & John L. Cloudsley-Thompson 154 Activity Rhythms and Behavioral Characterization of Two Epigean and One Cavemicolous Harvestmen (Arachnida, Opiliones, Gonyleptidae) by Sonia Hoenen & Pedro Gnaspini 159 Courtship and Mating Behavior of Brachypelma klaasi (Araneae, Theraphosidae) by M. Yanez, A. Locht & R. Macias-Orddnez 165 Location of Successful Strikes on Prey by Juvenile Crab Spiders Misumena vatia (Araneae, Thomisidae) by Douglass H. Morse 171 Sampling Method and Time Determines Composition of Spider Collections by Jan Green 176 Notes on the Biogeography and Natural History of the Orbweaving Spider Carepalxis (Araneae, Araneidae), Including a Gumnut Mimic From Southwestern Australia by Barbara York Main 183 Contents continued on inside back cover [ The Journal of ARACHNOLOGY OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY VOLUME 27 1999 NUMBER 2 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF: James W. Berry, Butler University MANAGING EDITOR: Petra Sierwald, Field Museum ASSOCIATE EDITORS: Matthew Greenstone, USDA; Robert Suter, Vassar College EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E. Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C. Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ.; C. D. Dondale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galia- no, Mus. Argentino de Ciencias Naturales; C. Griswold, Calif. Acad. Sci.; N. V. Horner, Midwestern State Univ.; D. T. Jennings, Garland, Maine: V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ.; N. I. Platnick, American Mus. Natural Hist.; G. A. Polis, Vanderbilt Univ.; S. E. Riechert, Univ. Tennessee; A. L. Rypstra, Miami Univ., Ohio; M. H. Robinson, U.S. National Zool. Park; W. A. Shear, Hampden-Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa Rica. 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, $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: Ann L. Rypstra (1997-1999), Dept, of Zoology, Miami Univer- sity, Hamilton, Ohio 45011 USA. PRESIDENT-ELECT: Frederick A. Coyle (1997-1999), Department of Biology, Western Carolina University, Cullowhee, North Carolina 28723 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. BUSINESS MANAGER: Robert Suter, Dept, of Biology, Vassar College, Pough- keepsie, New York 12601 USA. SECRETARY: Alan Cady, Dept, of Zoology, Miami Univ., Middletown, Ohio 45042 USA. ARCHIVIST: Lenny Vincent, Fullerton College, Fullerton, California 92634. DIRECTORS: H. Don Cameron (1997-1999), Matthew Greenstone (1997- 1999), David Wise (1998-2000). HONORARY MEMBERS: C. D. Dondale, H. W. Levi, A. F. Millidge, W. Whit- comb. Cover photo: Two mature female spiders, Holocnemus pluchei (Pholcidae), eating a single honey bee. Apis mellifera. (Photo by Rick Vetter) Publication date: 1 November 1999 @ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 1999. The Journal of Arachnology 27:401-414 FOSSIL ARANEOMORPH SPIDERS FROM THE TRIASSIC OF SOUTH AFRICA AND VIRGINIA Paul A. Selden: Department of Earth Sciences, University of Manchester, Manchester Ml 3 9PL, UK John M. Anderson and Heidi M. Anderson: National Botanical Institute, Private Bag XlOl, Pretoria 0001, South Africa Nicholas C. Fraser: Virginia Museum of Natural History, Martinsville, Virginia 24112 USA ABSTRACT. New fossil spiders from Triassic rocks of South Africa and Virginia are described. Though lacking synapomorphies of Araneomorphae, certain features suggest they belong in that infraorder, and possibly in the superfamily Araneoidea. Thus, they represent the oldest known fossil araneomorphs and extend the fossil record of the infraorder by approximately 40 Ma to 225 Ma. Few Mesozoic spiders have been described. Cretaceous mygalomorphs (Eskov & Zon~ shtein 1990), orbicularian araneomorphs (Sel- den 1990; Mesquita 1996) and indeterminable Araneae (Jell & Duncan 1986) have been de- scribed; and spiders from the Cretaceous Cra- to Formation of Brazil (Maisey 1991) and Ca- nadian amber (Me Alpine & Martin 1963) are currently being studied by PAS. Jurassic re- cords are equally sparse, consisting of a de- scribed archaeid (Eskov 1987), the araneoid Juraraneus Eskov 1984, and undescribed fil- istatids (Eskov 1989). Until now, Juraraneus was the oldest known fossil spider which could be considered an araneomorph. Only one Triassic spider, Rosamygale Selden & Gall 1992, has been described: it was placed in Hexathelidae and is the earliest mygalo- morph. In this paper, two new fossil spiders are described, both from rocks of Triassic age, thus tripling the number of known Triassic spiders. The specimens show features consis- tent with Araneomorphae (though synapomor- phies of that clade are not preserved), and rep- resent the oldest known fossil araneomorphs. Stratigraphy, paleoecology and locality infor- mation is provided for the South African spi- der by JMA and HMA, and for the Virginia specimen by NCF. All other discussion and systematics are the responsibility of PAS. METHODS Terminology and abbreviations. — In the specimens studied, the largest movable cutic- ular processes are termed bristles; they decline in width from base to tip and in this respect they differ from the spines which occur on many spiders which thicken between base and tip. Smaller and thinner cuticular hairs are termed setae. Abbreviations used in the text and figures: I-IV = first to fourth legs, car = carapace, fe = femur, mt = metatarsus, pa = patella, Pd == pedipalp, st = sternum, ta = tarsus, ti = tibia, trich = trichobothrium. All measurements are in mm. South African specimens. — The specimen from South Africa (Figs. 2-8) was discovered by JMA and very kindly sent to the senior author by HMA. It originates from the Upper Umkomaas locality in the Triassic Molteno Formation at Natal- Kwazulu, South Africa (Anderson & Anderson 1983, 1984). This is the first fossil spider to be found in South Af- rica. A second specimen of a possible spider, from the Telemachus Spruit locality in the same formation, was discovered recently; but it is less well-preserved and is not described here. The Molteno Formation (Fig. 1) was de- posited in an extensive, intracontinental fore- land basin bounded by rising fold mountains to the south and traversed by a system of braided rivers (Cairncross, Anderson & An- derson 1995). It reaches a maximum thickness of about 600 m and the erosional remnant ex- tends over an area roughly 400 km north to 401 402 THE JOURNAL OF ARACHNOLOGY Figure 1. — Location maps of the Solite Quarry, Virginia, in the Danville/Dan River basin (Triassic outcrops shown in black) and localities in the Molteno Formation (shown in black), South Africa, in relation to the late Triassic (—220 Ma) world (land shaded; Triassic outcrops in darker shading). Maps after Anderson et al. 1998; Fraser et al. 1996; Smith et al. 1994. south and 200 km west to east. The age of the formation is not tightly established, but on the basis of global biostratigraphic correlations (Anderson & Anderson 1983; Anderson et al. 1989) is considered to be Camian (late Tri- assic), 222-229 Ma BP. No absolute radio- metric ages are available. A 30-year collecting program (Anderson & Anderson 1983, 1989, 1993a, b; Anderson et al. 1989) has yielded 100 phytotaphocenoses (PTCs, fossil plant assemblages) from the for- mation. The flora — the richest known globally from the Triassic — includes 56 genera with 204 vegetative species. It is particularly char- acterized by some 20 species of the seed-fem Dicroidium Gothan 1912. There occurs a roughly equal diversity of gymnosperms, in- cluding conifers, cycads and ginkgos, along with several new orders, and ‘pteridophytes,’ primarily horsetails and ferns. Though rare, insects comprise by far the most frequently encountered element of the fauna. A remark- able diversity of 117 genera and 333 species in 1 8 orders is provisionally recognized in the over 2000 specimens at hand from 43 of the 100 plant assemblages. The beetles, cock- roaches and bugs clearly dominate. Concho- straca, from 20 PTCs, are represented by some 3 genera and 8 species. The remaining fauna is sparse: 3 species of fish (impressions only) from 3 PTCs, 2 species of bivalve from 1 PTC, and the 2 spider specimens documented here. Dinosaur trackways, but no skeletal re- mains, have been identified at a few (non- plant) sites. The Upper Umkomaas (‘WaterfalT) locali- ty: The fossiliferous bed consists of a dark grey, thinly laminated, carbonaceous shale with excellently preserved plant compressions (with cuticle) and rare insects. Exposed in the bed of a small mountain stream, it reaches 2.3 m in thickness and over 10 m in strike. The shale is interpreted as having accumulated in a non-aerated, abandoned river channel. With SELDEN ET AL.— TRIASSIC SPIDERS 403 Figure 2. — Triassaraneus andersonorum Selden new genus and species, camera lucida drawing of holotype part, PRE/F 18560a, from the Triassic (Camian) Molteno Formation, South Africa. See Figure 4. 23 genera and 73 species of plant (vegetative taxa), the Umkomaas site has produced by far the richest flora of the 100 Molteno PTCs. The assemlage is seen as representing a dense riv- erine forest dominated by Dicroidium. The insects from the site, mostly isolated wings and abdomens and- — far less common- ly— complete adults, now number 166 indi- viduals. The rate of yield is around one spec- imen per hour when scanning cleaved. bedding plane surfaces under the microscope. The insect fauna is strongly dominated by cockroaches (80 individuals, 4 species), bee- tles (63 individuals, 28 species) and bugs (12 individuals, 6 species). The remaining fauna consists of some 70 specimens of Conchostra- ca in 3 species and the single specimen of spider. The extreme rarity of the spider is clearly emphasized when it is considered that we (JMA & HMA) have to date spent 400 404 THE JOURNAL OF ARACHNOLOGY Figure 3. — Triassaraneus andersonorum Selden new genus and species, camera lucida drawing of holotype counterpart, PRE/F 18560b, from the Triassic (Camian) Molteno Formation, South Africa. See Figure 7. man-hours cleaving plant-fossiliferous slabs from the Umkomaas site, and that the entire curated collection of 2500 cataloged slabs has been carefully scanned under a binocular mi- croscope. The Telemachus Spruit locality: This site is somewhat different in character from Upper Umkomaas. The plant-fossiliferous bed, a 10 cm thick, buff mudstone, is exposed along a stream bank over approximately 10 m of strike and is interpreted as an abandoned channel-fill (Caimcross, Anderson & Ander- son 1995). The flora (vegetative) of 12 genera and 19 species is strongly dominated by the single coniferous species Heidiphyllum elon- gatum (Morris 1845) Retallack 1981. The as- semblage most likely represents two distinct plant communities: a mono-dominant stand of reed-like conifers colonizing sand bars in the braided river and, from farther afield, a Di- croiWiMw-dominated riparian forest occupying the river bank. The insect fauna at this site. SELDEN ET AL.— TRIASSIC SPIDERS 405 represented by only 17 fragmentary speci- mens, is dominated, as at Umkomaas, by bee- tles, cockroaches and bugs. The yield remains at around one individual per man-microscope hour. Conchostracans have not been found. The rarity of the single spider specimen is once again emphasized by the fact that JMA & HMA have spent 90 man-hours cleaving slabs at this site and that all 900 curated and cataloged slabs have been carefully scanned for insects or other faunal elements under the binocular microscope. Preservation: The Umkomaas specimen is preserved as brown, organic cuticle on a dark grey shale. Superficially, the cuticle appears black; but under ethanol and high magnifica- tion the brown color is evident, and the shale appears paler. The shale is splintery, and piec- es readily spall away, necessitating care while studying the fossils. Both part and counterpart eventually cracked after hours of study; the crack on the counterpart is shown in Fig. 7. Scattered throughout the shale are abundant plant remains, including leafy shoots (Fig. 4), spores, and unidentified coalified strands. The legs are outstretched in a relaxed man- ner, suggesting that the spider died in the wa- ter or was carried there soon after death, thus enabling the muscles to relax. The podomeres are flattened by compression of the shale ma- trix, but there is a little relief in the form of a pair of oval (long axes sagittal) humps in the anterior half of the prosomal body and a wider oval (long axis transverse) area in the poste- rior part of the prosomal body. These struc- tures occur on the part with equivalent de- pressions on the counterpart. Lateral to the prosomal body, the areas representing the cox- ae, trochanters and basal femora of the legs has a stepped appearance, with each leg over- lapping the more posterior leg slightly, and with a sliver of matrix between. The fact of the prosomal humps and the overlapping cox- ae suggest that the part represents the animal viewed lying on its back, presenting its ventral side to view; the counterpart is mainly an ex- ternal mold of the ventral surface. The humps on the prosomal body thus represent paired palpal endites with chelicerae beneath (as viewed with the spider on its back) and the sternum behind. The numerous sheets of cu- ticle, some well-sclerotized, at the anterior end of the prosomal body area, and the convexity of the humps, suggest that the chelicerae as well as the palpal endites are involved here. The detail of preservation is quite extraordi- nary, although it has been pointed out (Selden 1989) that the fossilization of a spider is a rare event. When one is recognizable to a collector, it usually turns out to be well-preserved; and the stratum from which it came is dubbed a Fossil-Lagerstatte. Under high magnification, long setae are seen to be abundant on the legs, especially on the more distal podomeres. Larger bristles are apparent, too, as well as the paired claws on some tarsi (Fig. 5), and the details of some joint articulations. Virginia specimen. — The Virginia speci- men (Figs. 9-14) was collected by NCF It originates from the late Triassic (Camian) of Virginia and is deposited in the Virginia Mu- seum of Natural History, Martinsville, Virgin- ia. Details of the paleoenvironment and asso- ciated fauna are given in Fraser et al. (1996). The early Mesozoic rocks of the Newark Su- pergroup of eastern North America were de- posited in a series of rift basins that formed as Pangaea started to separate (Fig. 1). Col- lectively, the sediments provide a continuous record from the middle Triassic (Anisian) through to the early Jurassic (Hettangian or younger) (Table 1). The sediments are poten- tially of enormous value in studies of terres- trial faunal and floral change at this crtitical period. However, despite their long-time fame for extensive dinosaur trackways (e.g., Hitch- cock 1836a, b, 1858; Lull 1915), documen- tation of other fossils is extremely limited. The long (167 km), but exceptionally nar- row (3-15 km), Danville/Dan River basin in Virginia and North Carolina is one of the more southern basins (Fig. 1). Trackways have been recorded from a variety of localities in this basin (e.g., Fraser & Olsen 1996); and isolated occurrences of tetrapods, fish and plants have also been reported (Olsen & Gore 1989). By far the most significant and pro- ductive locality to date is the Virginia Solite Quarry which straddles the Virginia-North Carolina state line. The sediments exposed at the Solite Quarry are referred to the Cow Branch Formation. Paleomagnetic studies in- dicate that they are equivalent in age to the Lockatong Formation of the Newark basin, most specifically the Nursery through to the Prahls Island members, with the main insect- producing unit probably age-equivalent with the basal Skunk Hollow member (Kent & Ol- 406 THE JOURNAL OF ARACHNOLOGY Figures 4-7. — Triassaraneus andersonorum Selden new genus and species, holotype from the Triassic (Camian) Molteno Formation, South Africa; incident light under ethanol. 4. Whole part, PRE/F 18560a; 5. Detail of tarsus and distal metatarsus of left leg I of part showing general preservation and tarsal claws (top), X 160; 6. Detail of body and proximal podomeres of part, X24; 7. Whole counterpart, PRE/F 18560b, X12. SELDEN ET AL.— TRIASSIC SPIDERS 407 Stratigraphy North America South Africa T Table 1 . — Stratigraphic correlation of the Triassic (Camian) Cow Branch Formation of Virginia and the Molteno Formation of South Africa. sen 1997). On this evidence, they are late Car- nian in age, which is in close agreement with biostratigraphic studies (Olsen & Gore 1989). Taken together, the three quarries of the So- lite Corporation at Cascade expose over 350 m of section. Like other lacustrine rocks of the Newark Supergroup, there is a very clear cyclical pattern of sedimentation which re- fleets fluctuating lake levels. Typically, each sequence (van Houten cycle) consists of three divisions interpreted as: 1) lake transgression, followed by 2) a high stand, and then 3) a regression and low stand. These fluctuations are attributed to climate changes which affect the rates of inflow and evaporation (van Hou- ten 1964; Olsen 1986). Each cycle is about 20 m thick. Division 3 facies contain footprints (including Rhynchosauroides, Gwyneddi- chnium, as well as those of small theropod and omithischian dinosaurs), and they also yield root traces and foliage fragments. Division 2 facies are the most fossiliferous. In the upper 408 THE JOURNAL OF ARACHNOLOGY portion Pagiophyllum and Brachyphyllum shoots are common together with cone scales. Towards the base of division 2 diversity levels increase, and there are abundant remains of cycadeoid foliage together with ferns, gym- nosperms and occasional ginkgophytes. Ver- tebrates are also present in these units. Fish are represented by a number of semionotids, redfieldiids, the paleoniscoid Turseodus and a coelacanth. The most abundant tetrapod is the prolacertiform Tanytrachelos, known from over 200 skeletons. In addition, phytosaurs and a smaller number of other tetrapods are represented by fragmentary remains. The in- sects occur almost exclusively at the base of division 2. Foliage fragments, particularly of cycadeoids, are fairly common in division 1. To date, only one area in the quarry has been substantially excavated. In the 1970s, teams from Yale University, under the direction of Paul E. Olson, first realized the potential of the locality and collected over 300 insects from a relatively small (approx. 15 m^) area. This excavation was extended recently by teams from the Virginia Museum of Natural History and Columbia University, but the total area exposed does not exceed 40 m^. Most of the time spent in the field is devoted to ex- posing the fossiliferous units. Once exposed, each man-hour yields, on average, 6 or 7 in- sects and literally thousands of conchostra- cans. The cleaved surfaces are scanned in the field using 7X magnification eye visors. The insects are found almost exclusively in two or SELDEN ET AL.— TRIASSIC SPIDERS 409 Figure 9. — Camera lucida drawing of Argyrarachne solitus Selden new genus and species, holotype, VMNH 782, part, from the Triassic (Camian), Solite Quarry, Cascade, Virginia. See Figure 10. three discrete, very thin beds—over 95% of the specimens were uncovered from a single 2.5 cm thick unit. The other units produce abundant plant and vertebrate remains and, while they are unlikely to produce insects, an equal portion of the excavation effort has been channeled towards these units. Some 2500 in- sects had been collected by mid- 1997, when a new insect-bearing unit was identified, con- taining water bugs in exceptionally high den- sities (1 cm“Q. Thus, the numbers of insects recovered from the Solite quarries is expected to increase. The great majority of the insect finds are of complete individuals. By contrast with the large number of insects, only two spi- ders have been recovered, making them a very rare component of the fauna, as they are in the Molteno assemblages. Preservation: The specimen is almost cer- tainly a juvenile. An additional specimen of a possible spider from the Solite Quarry has been seen by the author, but it is even smaller and less well preserved than the one described here. If it is a spider, then it is also a juvenile. While there is active collecting at the site, hope remains that further, mature specimens might turn up and allow better description of the species. 410 THE JOURNAL OF ARACHNOLOGY Figures 10-14. — Argyrarachne solitus Selden new genus and species, holotype, VMNH 782, from the Triassic (Camian), Solite Quarry, Cascade, Virginia; incident light under ethanol. 10. Whole part, X20; 11. Anterior carapace, palpal endites, chelicerae, and right pedipalp of part, details of endites and chelicerae are obscure. Note matrix bubbles due to mineral growth and patches where silvery cuticle is absent, X85; 12. Whole counterpart, X20; 13. Detail of distal parts of left legs I and II of part, note fine preservation of setae, dentate tarsal claws, X 100; 14. Detail of tarsus and distal metatarsus of right leg II of part, X 100. SELDEN ET AL.— TRIASSIC SPIDERS 411 The specimen is preserved, as are the in- sects in this deposit, as silver streaks on a black matrix. On the same slab are abundant bivalved crustaceans. The matrix is an ex- tremely fine black shale but contains abundant crystals (gypsum?) which form bubbles on the flat surface of the rock and disrupt the speci- men in places (Fig. 11). Finding, studying, and photographing the specimen require con- siderable manipulation of the light source and the use of ethanol to enhance the contrast be- tween silvery streaks and matrix. Neverthe- less, very fine details, such as leg setae, can be seen. The carapace and pedipalp endites are preserved as sheets of silvery cuticle. The ab- sence of patches of cuticle in the anterior car- apace region gives the appearance of large eyes, but higher magnification (Fig. 11) shows this not to be the case. Patches where cuticle is absent on the part (Fig. 10) can be matched with the presence of cuticle on the counterpart (Fig. 12). The carapace is approximately rect- angular; its anterior border is obscured by pre- sumed chelicerae and palpal endites. The ped- ipalps are preserved bent to the left in the part. The walking legs are setose but lack bristles or spines, and the proximal podomeres are poorly preserved. The tarsal claws are small and stout, and small teeth can be seen in Figs. 13 and 14. Leg III is very short; and legs I, 11, and IV are approximately the same length, giving a leg formula of 1243. The abdomen is not preserved. The specimen almost cer- tainly represents a juvenile, as evidenced by the short, undifferentiated podomeres. Pre- sumably, like the South African specimen, it fell into the lake waters and died there. DISCUSSION There is little doubt that the South African specimen is a spider; no other arachnid order presents the same arrangement of prosomal appendages, palpal endites and sternum. The broad sternum and presence of palpal endites (not known in mesotheles and most mygalo- morphs) precludes the fossil from Mesothelae. If the spider were a mesothele or a mygalo- morph, then the orthognath chelicerae would be expected to protrude conspicuously well beyond the front of the body, as they do in compression fossil mesotheles (Selden 1996) and mygalomorphs (Eskov & Zonshtein 1990; Selden & Gall 1992) but not araneomorphs except where specially enlarged, e.g., in some adult males (Eskov 1984; Selden 1990). Rath- er, the most anterior fragments of chelicerae occur only just anterior to the humps which represent the cheliceral bodies and palpal en- dites. The general appearance of the spider, with rather long and slender legs, a leg for- mula of 1243, lack of leg scopulae and with generally rather sparse bristles, are features suggestive of Araneomorphae rather than My- galomorphae. Furthermore, this leg shape and arrangement, the lack of spines (only bristles), the small tarsal claws and lack of scopulae, the paucity and arrangement of bristles, and the possible metatarsal trichobothrium seen on one leg are all suggestive of Araneoidea. While it is impossible to be precise about the number of leg bristles, there appears to be a pattern. All legs have a superior bristle near the distal edge of the patella (also on the ped- ipalp) and a row of three bristles on the su- perior side of the tibia. In addition, there are at least two bristles on the superior side of femur I. Griswold et al. (1998) demonstrated the existence of a clade of Araneoidea in which femoral spination is lacking — the spineless femora clade — which includes Ther- idiidae, Nesticidae, Cyatholipidae and Syno- taxidae. The South African fossil spider can- not belong in this clade. Circumstantial evidence may also be help- ful in determining the systematic placement of the South African spider. The Molteno shale is interpreted as having accumulated in a an- aerobic, abandoned river channel. Many mod- em spiders live close to such an environment, including lycosoids (e.g., lycosids and pisaur- ids) which favor damp habitats, and tetrag- nathids which build orb webs among water- side vegetation. Lycosoids are adept at walking on water, but orb-weavers are likely to drown if they actually fall into water. In- deed, a fossil tetragnathid and other orb-weav- ers are known from waterside situations in the Jurassic and Cretaceous (Eskov 1984; Selden 1990). The preservation of the Triassic my- galomorph, Rosamygale Selden & Gall 1992, was unusual in that sea-water inundation was involved. The identity of the Virginia specimen is less secure. Like the South African specimen, it is also clearly a spider because of the leg and podomere arrangement. The short, undiffer- entiated podomeres and undeveloped pedipalp suggest an immature. The very short third leg 412 THE JOURNAL OF ARACHNOLOGY is distinctive and typical (but not diagnostic) of Orbiculariae. The short, dentate tarsal claws do not give a clue to relationships but are common among web weavers. In the Treatise on Invertebrate Paleontol- ogy, Petrunkevitch (1955) listed five genera of spiders from the Carboniferous period tenta- tively referred to Araneomorphae. Three of these, Archaeometa Pocock 1911, Arachno- meta Petrunkevitch 1949, and Eopholcus Erie 1904 (family Archaeometidae Petrunkevitch 1949) were placed in a new superfamily Ar- chaeometoidea Petrunkevitch 1955, diagnosed as ‘Presumptive Trionychi with segmented ab- domen.’ The other two, Pyritaranea Erie 1901 and Dinopilio Eric 1904 (family Pyritaranei- dae Petrunkevitch 1953), were placed in a new superfamily Pyritaraneoidea Petrunkev- itch 1955 and diagnosed as ‘Presumptive Dionychi with laterigrade legs and segmented abdomen.’ Petrunkevitch (1955: 132) noted problems in regarding these specimens as ar- aneomorphs: abdominal segmentation and lack of araneomorph synapomorphies. All of these specimens are currently being studied by PAS. Preliminary studies indicate that Ar- chaeometa, Arachnometa, and Dinopilio are arachnids but not spiders, while Eopholcus and Pyritaranea may be spiders but are not sufficiently well preserved to determine their affinities. No other fossil araneomorph spiders are known from the Paleozoic era or the Tri- assic period of the Mesozoic era, thus the specimens described herein are the oldest known fossil spiders which can be referred to Araneomorphae with some degree of confi- dence. The discovery of araneomorph spiders in the Triassic period is not unexpected, since the existence of their sister group, Mygalomor- phae, in strata of similar age (Selden & Gall 1992) predicts this. Eurthermore, the discov- ery of a mesothele spider in rocks of Penn- sylvanian age (Selden 1996) is a predictor that Opisthothelae (Mygalomorphae + Araneo- morphae) occurred at that time too, so it is indeed possible that araneomorph spiders may be found in earlier strata. However, no fossil spiders have yet been found in strata of Perm- ian age (which immediately precedes the Tri- assic), despite the fact that a vast number of insect fossils are known from that period. Eos- sil spiders from this period should prove to be extremely interesting (Eskov 1990). The fos- sils described here are not primitive araneo- morphs, which suggests that a fair degree of radiation had occurrred among araneomorphs before the late Triassic. The existence of sim- ilar forms in South Africa and Virginia, wide- ly separated geographically and subject to dif- ferent climatic regimes (though on the same continent), is further evidence in support of this hypothesis. SYSTEMATICS [Note: Due to the lack of autapomorphic characters, the diagnoses given below are not comparative.] Infraorder Araneomorphae Smith 1902 ?Superfamily Araneoidea Latreille 1806 Trias saraneus Selden new genus Type and only species. — Trias saraneus andersonorum new species, see below. Etymology. — The genus name refers to Triassic, the stratigraphic period from which the specimen originates, and Araneus, a wide- spread genus of spiders which the fossil su- perficially resembles. Diagnosis. — Araneomorph spider, possibly araneoid; sternum wider than long and with straight anterior border. Triassaraneus andersonorum Selden new species Eigs. 2-8 Holotype. — PRE/E 18560a (part) and 18560b (counterpart), immature, or mature fe- male, from Member Z of the late Triassic (Camian) Molteno Eormation at the Upper Umkomaas ‘Waterfall Locality’ (UMK III), Natal- Kwazulu, South Africa; deposited in the National Botanical Institute, Pretoria, South Africa. Etymology. — ^The trivial name honors Drs. John and Heidi Anderson who found this specimen and kindly passed it on for descrip- tion. Diagnosis. — As for the genus. Description. — Carapace not visible. Ante- rior part of prosomal body shows a pair of humps representing palpal endites and chelic- erae beneath (as specimen is viewed). Layers of cuticle bearing well-sclerotized areas also suggest this is the case. Posterior to palpal en- dites another raised area represents sternum which is wider than long, posterior border gently procurved, anterior border straight, lat- SELDEN ET AL.— TRIASSIC SPIDERS 413 eral edges more vague but apparently curved outwards. Pedipalp fe, pa and ti/ta preserved, latter podomeres rather thickened, and visible on left side of part extending beyond cuticle fragment as a hump {cf. endites and chelicer- ae); pa bears superior bristle. Approximate lengths of Pd podomeres: fe 0.3, pa 0.1, ti/ta 0.4. All legs with superodistal bristle on pa, line of 3 superior bristles on ti (middle lon- gest), small paired ta claws (median claw not seen), no scopulae; possible mt trichoboth- rium visible on leg 3, leg I fe with row of at least 2 superior bristles. Approximate lengths of leg podomeres: leg I: fe 1.3, pa 0.2, ti 1.4, mt 1.2, ta 0.6, total 4.7; leg II: fe 1.3, pa 0.2, ti 1.2, mt 1.0, ta 0.3, total 4.0; leg III: fe 0.9, pa 0.2, ti 0.8, mt 0.7, ta 0.4, total 3.0; leg IV: fe 1.0, pa 0.2, ti 0.9, mt 0.8, ta 0.3, total 3.2. Leg formula 1243. Abdomen not visible. Argyrarachne Selden new genus Type and only species. — Argyrarachne so- litus new species, see below. Etymology. — Greek: argyros, silver, and arachne, a spider, referring to the appearance of the fossil spider as silvery streaks on a black rock. Diagnosis. — Araneomorph spider with sub- rectangular carapace and short, stout, dentate tarsal claws. Argyrarachne solitus Selden new species Figs. 9-14 Holotype. — VMNH 782 (part and counter- part), immature, from the late Triassic (Car- nian) Cow Branch Formation at Solite Quarry, Cascade, Virginia; deposited in the Virginia Museum of Natural History, Martinsville, Vir- ginia. Etymology. — The trivial name refers to the Virginia Solite Corporation, in whose quarry the specimen was found. Diagnosis. — As for the genus. Description. — Carapace roughly parallel- sided with straight posterior margin, anterior border obscured by presumed chelicerae and palpal endites. Total length of preserved car- apace + chelicerae/endites 0.7 mm. Pedipalps short, not swollen. Walking legs setose, lack- ing bristles, proximal podomeres poorly pre- served. Tarsal claws small, stout, dentate. Ap- proximate lengths of leg podomeres: leg I: ti 0.42, mt 0.43, ta 0.51; leg II: ti 0.38, mt 0.39, ta 0.51; leg III: ti 0.31, mt 0.29, ta 0.26; leg IV: ti 0.43, mt 0.49, ta 0.45. Leg III very short, legs I, II, and IV approximately the same length; leg formula 1243. Abdomen not pre- served. ACKNOWLEDGMENTS We are grateful to Arthur Cruickshank (Leicester) and Marion Bamford (Johannes- burg) for carrying the Triassaraneus specimen respectively from and back to South Africa, the Virginia Solite Corporation for their sup- port for paleontology, Jon Coddington for use- ful information on araneoids, and Bill Shear for his help in many ways with this work. LITERATURE CITED Anderson, J.M. & H.M. Anderson. 1983. Palaeo- flora of Southern Africa. Molteno Formation (Triassic) 1. Dicroidium. A. A. Balkema, Rotter- dam. Anderson, J.M. & H.M. Anderson. 1984. The fos- sil content of the Upper Triassic Molteno For- mation, South Africa. Palaeont. Africana, 25:39- 59. Anderson, J.M., H.M. Anderson & A.R.I. Cruick- shank. 1998. Late Triassic ecosystems of the Molteno/Lower Elliot biome of southern Africa. Palaeontology, 41:387-421. Caimcross, B., J.M. Anderson & H.M. Anderson. 1995. Palaeoecology of the Triassic Molteno Formation, Karoo Basin, South Africa sedimen- tological and palaeontological evidence. South African J. GeoL, 98:452-478. Eskov, K. 1984. A new fossil spider family from the Jurassic of Transbaikalia (Araneae: Chelicer- ata). N. Jb. GeoL Palaont. Mh., 1984:645-653. Eskov, K. 1987. A new archaeid spider (Chelicer- ata: Araneae) from the Jurassic of Kazakhstan, with notes on the so-called “Gondwanan” rang- es of recent taxa. N. Jb. Geol. Palaont. Abh., 175: 81-106. Eskov, K. 1989. Spider paleontology: present trends and future expectations. Acta Zool. Fen- nica, 190:123-127. Eskov, K. & S.L. Zonshtein. 1990. First Mesozoic mygalomorph spiders from the Lower Creta- ceous of Siberia and Mongolia, with notes on the system and evolution of the order Mygalomor- phae (Chelicerata: Araneae). N. Jb. Geol. Pa- laont. Abh., 178:325-368. Fraser, N.C., D.A. Grimaldi, RE. Olsen & B. Axsmith. 1996. A Triassic Lagerstatte from eastern North America. Nature, 380:615-619. Fraser, N.C. & RE. Olsen. 1996. A new dinosau- romorph ichnogenus from the Triassic of Virgin- ia. Jeffersoniana, 7:1-17. Eric, A. 1901. Fauna der Gaskohle und der Kalk- steine der Permformation Bohmens. Vol. IV, part 414 THE JOURNAL OF ARACHNOLOGY 2. Myriapoda pars 11. Arachnoidea: 56-63, text figures 359-368, pL 153-154. Prague. Fric, A. 1904. Palaeozoische Arachniden. Privately published by Anton Fric, Prague. Gothan, W. 1912. Uber die Gattung Thinnfeldia Et- tingshausen. Abh. naturhist. Ges. Niimberg, 19: 67-80. Griswold, C.E., J.A. Coddington, G. Hormiga & N. Scharff. 1998. Phylogeny of the orb- weaving spiders (Araneae, Orbiculariae: Deinopoidea, Ar- aneoidea). Zool. J. Linn. Soc., 123:1-99. Hitchcock, E. 1836a. Ornithichnites in Connecti- cut. American J. Sci., 1836:174-175. Hitchcock, E. 1836b. Description of the footmarks of birds {Ornithichnites) in New Red Sandstone in Massachusetts. American J. Sci., 1836:307- 340. Hitchcock, E. 1858. Ichnology of New England; A Report on the Sandstone of the Connecticut Val- ley, Especially in its Footmarks. Boston. Houten, EB. van 1964. Origin of red beds: some unsolved problems. Pp. 000-000, In Problems in Palaeoclimatology (A.E.M. Naim, ed.). Intersci- ence, London. Jell, PA. & P.M. Duncan. 1986. Invertebrates, mainly insects, from the freshwater. Lower Cre- taceous, Koonwarra Fossil Bed (Kommburra Group), South Gippsland, Victoria. Mem. Asso. Australasian Palaeontols, 3:311-205. Kent, D.V & PE. Olsen. 1997. Paleomagnetism of Upper Triassic continental sedimentary rocks from the Dan River-Danville rift basin (eastern North America). Bull. Geol. Soc. America, 109: 366-377. Latreille, P.-A. 1806. Genera Cmstaceomm et In- sectomm. Vol. 1, Araneides. A. Koenig, Paris. Pp. 82-127. Lull, R.S. 1915. Triassic life of the Connecticut Valley. Bull. St. Geol. Nat. Hist. Surv. Connect- icut, 81:1-285. Maisey, J.G. (ed.). 1991. Santana Fossils: An Il- lustrated Atlas. TFH Publications, Neptune City, New Jersey. McAlpine, J.F. & J.E.H. Martin. 1963. Canadian amber — a paleontological treasure-chest. Cana- dian Entomol., 101:819-838. Mesquita, M.V 1996. Cretaraneus martinsnetoi n. sp. (Araneoidea) da Forma^ao Santana, Cretaceo Inferior da Bacia do Araripe. Rev. Univ. Gua- mlhos, Ser. Geoc., 1:24-31. Morris, J. 1845. Fossil flora. Pp. 245-254, In Phys- ical Description of New South Wales and Van Diemans Land. (RE. de Strezelecki, ed.). Long- man, Brown & Green, London. Olsen, RE. 1986. A 40-million-year lake record of early Mesozoic orbital climate forcing. Science, 234:842-848. Olsen, RE. & P.J.W. Gore. 1989. Dan River-Dan- ville Basin, North Carolina and Virginia; Stop 2.2; Solite Quarry, Leakesville Junction, Virgin- ia. Pp. 37-44, In Sedimentation and Basin Anal- ysis in Siliciclastic Rock Sequences; Vol. 2, Tec- tonic, Depositional, and Paleoecological History of Early Mesozoic Rift Basins, Eastern North America. (RE. Olsen, ed.). Field Trips for the 28th International Geological Congress. Ameri- can Geophysical Union, Washington D.C. Petmnkevitch, A. 1949. A study of Palaeozoic Arachnida. Trans. Connecticut Acad. Arts Sci., 37:69-315. Petmnkevitch, A. 1953. Palaeozoic and Mesozoic Arachnida of Europe. Geol. Soc. America, Mem- oir 53. Geol. Soc. America, New York. Petmnkevitch, A. 1955. Arachnida. Pp. P42-P162, In Treatise on Invertebrate Paleontology. Part P. Arthropoda 2. (R.C. Moore, ed.). Geol. Soc. America and Univ. Kansas Press, Boulder, Col- orado and Lawrence, Kansas. Pocock, R.L 1911, A monograph of the terrestrial Carboniferous Arachnida of Great Britain. Mon- ogr. Palaeontogr. Soc., 64:1-84. Retallack, G.J. 1981. Middle Triassic megafossil plants from Long Gully, near Otematata, North Otago, New Zealand. J. Roy. Soc. New Zealand, 11:167-200. Selden, PA. 1989. Taxonomic and taphonomic bi- ases in fossil Lagerstatten. Abstr. Int, Geol. Congr., Washington D.C., 1989, 3:70-71. Selden, P.A. 1990. Lower Cretaceous spiders from the Sierra de Montsech, north-east Spain. Pa- laeontology, 33:257-285. Selden, P.A. 1996. Fossil mesothele spiders. Na- ture, 379:498-499. Selden, PA. & J.-C. Gall. 1992. A Triassic my- galomorph spider from the northern Vosges, France. Palaeontology, 35:211-235. Smith, A.G., D.G. Smith & B.M. Funnell. 1994. Atlas of Mesozoic and Cenozoic Coastlines. Cambridge Univ. Press, Cambridge. Smith, F.P 1902. The spiders of Epping Forest. Es- sex Nat., 12:181-201. Manuscript received II October 1996, revised 1 November 1998. 1999. The Journal of Arachnology 27:415-431 NEW SPECIES AND CLADISTIC REANALYSIS OF THE SPIDER GENUS MONAPIA (ARANEAE, ANYPHAENIDAE, AMAUROBIOIDINAE) Martin J. Ramirez: Laboratory of Arthropods, Department of Biology, FCEyN, Universidad de Buenos Aires, Pabellon II Ciudad Universitaria (1428) Buenos Aires, Argentina, and Museo Argentino de Ciencias Naturales, Av. Angel Gallardo 470 (1405), Buenos Aires, Argentina ABSTRACT. The known range of the South American genus Monapia, previously known only from temperate South American forests, is expanded to central and eastern Argentina and Uruguay. A mono- phyletic group of five species with spinose forelegs is proposed, including M. angusta, newly transferred from Arachosia, plus four new species: M. charrua, M. guenoana, M. fierro and M. Carolina. One new species, M. tandil (from Buenos Aires Province), is proposed to be the sister group of Monapia vittata. A data matrix with 43 characters for the 13 species of the genus (plus 9 amaurobioidine outgroups) was cladistically analyzed. Although relationships among species are mostly resolved, the basal phylogeny of the genus remains unclear. The previous hypothesis of relationships of Monapia alupuran is unsupported in this new analysis. Additional records are given for M. lutea and M. dilaticollis . The genus Monapia Simon 1897 was re- vised in a recent contribution (Ramirez 1995b). The seven species there included are endemic of the temperate forests of Southern Chile and adjacent Argentina. Mello-Leitao (1944) described Arachosia angusta from Buenos Aires, Argentina, a very peculiar spe- cies with elongate body, flat carapace and re- markably spinose forelegs. In recent years, two very similar undescribed species were collected on large riparian grasses from Buen- os Aires and adjacent regions, and because of the highly modified body and bizarre leg spi- nation they were thought to belong, together with Arachosia angusta, to a separate, undes- cribed genus. However, a reexamination of genitalic characters in Arachosia angusta and those two related species showed a depressed median area on the epigymum, a character previously considered as synapomorphic of Monapia (Ramirez 1995b), and the divided male conductor typical of that genus (i.e.. Fig. 11). These species have extremely cryptic habits. Their elongate body helps camouflage the spiders on grass leaves, while the yellow- ish and spotted coloration mimics that of the dry leaves. Monapia guenoana new species females were found covering their eggsacs with their own body; and even when exposed to view, they were discerned with difficulty. Besides these three elongate and very spi- nose spiders, I consider two more new species collected in grasslands of central Argentina. They also have spinose fore tibiae, but a typ- ical amaurobioidine appearance. As shown below, these five species form a monophyletic group distributed in a region generically known as Pampas, where the main plant co- munities are grasslands. Finally, I include one new species close to Monapia vittata (Simon 1884) which was previously considered the sister group of all other species of the genus. In the light of these additional taxa and sev- eral new characters, a cladistic reanalysis for all Monapia species was undertaken. This analysis challenges some of the conclusions of my previous revision. METHODS The format of the descriptions follows Ra- mirez (1995b). Spermathecae for scanning mi- croscopy were treated as in Sierwald (1990). All measurements are expressed in millime- ters. Specimens are deposited in the following institutions: CAS = California Academy of Sciences (Charles Griswold); IRSN == Institut Royal des Sciences Naturelles de Belgique, Brussels (Louis Baert); MACN = Museo Ar- gentino de Ciencias Naturales “Bernardino 415 416 THE JOURNAL OF ARACHNOLOGY Rivadavia,” Buenos Aires (Cristina Scioscia); MLP = Museo de La Plata (Carola Sutton de Licitra, Luis Pereira); ZMK = Zoologisk Mu= seum, Copenhagen (Henrik Enghoff). CLADISTIC ANALYSIS All characters considered in a previous re- vision (Ramirez 1995b) were taken into ac= count. Some descriptive terminology was modified as in Ramirez (1995a). Multistate characters were considered additive when the states are interpreted as intemested homolo- gies. This is not intended to express any as- sumption on the evolution of characters, but merely reflect degrees of similarity (Lipscomb 1992; Goloboff 1997). Morphoclines were in- terpreted as intemested homologies. Because this approach might be suspected by some au- thors as an unjustified assumption, additional runnings were made with all characters non- additive. This analysis produced identical trees and statistics, thus demonstrating that the obtained phylogeny does not depend on my interpretation of morphoclines. The root of the tree was placed according to the subfamilial analysis made in Ramirez (1995a). Anyphaen- inae was used as a more distant outgroup. Be- cause the relationships among the 32 genera of that subfamily are unknown (Brescovit 1997), most entries were coded as polymor- phisms, according to the variability found throughout genera. The resolution of the out- groups changed slightly from the previous analysis of the genus (Ramirez 1995b), ac- cording to new knowledge of the genera re- lated to Amaurobioides Hickman 1949 (Ra- mirez 1997). Character 0: Body pattern, 0 = dark patches or uniform, 1 = dark spots on light back- ground (Fig. 1). Character 1: Carapace out- line, 0 = oval (Fig. 35), 1 ^ lengthened (Fig. 8). Character 2: Carapace height, 0 == normal, the posterior slope begins near the thoracic fo- vea (Fig. 24), 1 = flattened, the slope begins well behind the thoracic fovea (Figs. 7, 14). Character 3: Posterior eye row, 0 = procurved or straight, 1 = recurved. Character 4: Ab- domen shape, 0 = oval, 1 = lengthened. Char- acter 5: Number of retromarginal cheliceral teeth, 0 = two, 1 = three, 2 = four or more. The character is considered additive (correct- ed additivity from Ramirez 1995b). Character 6: Long ventral hairs on male palpal tibia, 0 = absent, 1 = present. Re-examined from Ra- mirez (1995b) and scored as absent in Mon- apia alupuran Ramirez 1995 because the hairs in M. alupuran are shorter than those on M. vittata, and very similar to those in other Monapia. Character 7: Cymbial basal retro- lateral notch, 0 = absent, 1 ^ weak (Figs. 27, 36), 2 = strong (Ramirez 1995b: fig. 48). Some undescribed species that probably be- long to Oxysoma Nicolet 1849 have a strong cymbial notch. Considered additive. Character 8: Tegulum with a deep notch occupied by the median haematodocha, 0 = absent, 1 = pre- sent. Character 9: Trajectory of sperm duct, 0 = parallel to tegular notch, 1 = with a curve near the apical margin of tegulum. Coding and internal step for Oxysoma as in Ramirez (1995b: character 8). Character 10: Shape of paramedian apophysis, 0 ^ thick, 1 = thin, 2 = long and very thin. Considered additive. Character 1 1 : Amaurobioidine paramedian apophysis closely associated with median apophysis (Ramirez 1995a), 0 == absent, 1 present. Character 12: Length of basal portion of embolus, 0 = short, 1 = very long. The intermediate state “long” considered in Ra- mirez (1995b) is here ignored because after the addition of new species the distinction be- tween “short” and “long” is equivocal. Char- acter 13: Shape of basal portion of embolus, 0 = cylindrical, 1 = flattened. Character 14: Extension of basal embolar unsclerotized area, 0 = absent or small, 1 = wide. The scoring of “small” as an intermediate state made in the previous revision is not used here because it is scored in character 12 state 1 (the basal portion of the embolus is defined by the pres- ence of the unsclerotized area). The wide and folded unsclerotized membrane present in M. lutea (Nicolet 1849) and M. huaria Ramirez 1995 is instead very different from the basic pattern of embolar morphology. Character 15: Grooved primary conductor (Ramirez 1995a), 0 = absent, 1 — present. Character 16: Sec- ondary conductor, 0 ^ present, 1 = absent. Character 17: Groove in secondary conductor, 0 ^ absent (Figs. 38-39), 1 = present. In most Monapia species the secondary conductor is divided by an unsclerotized area, and the groove remains on the prolateral portion. Character 18: Division of secondary conduc- tor (Ramirez 1995b), 0 = entire, 1 = divided by an unsclerotized area (Fig. 11). Character 19: Unsclerotized area of secondary conductor with a lobe, 0 = absent, 1 = present. Char- RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA 417 acter 20: Prolateral portion of the secondary conductor displaced towards the base of the embolus, 0 = apical, 1 = displaced (Figs. 37, 39). Character 21: Denticles on prolateral por= tion of the secondary conductor, 0 — absent, 1 = present (Fig. 37). Character 22: Retrola- teral portion of the secondary conductor fused to tegulum, 0 = free (Fig. 16), 1 — fused (Fig. 10). Character 23: Denticles on retrolateral portion of secondary conductor, 0 = absent, 1 = present. Character 24: Shape of base of re- trolateral portion of secondary conductor, 0 — thick, 1 = thin and wide. Character 25: An- terior pouch on median epigynal field, 0 — absent, 1 = present (Fig. 4). Character 26: Shape of epigynal anterior pouch, 0 = pit like (Fig. 13), 1 = transversal furrow (Fig. 19). Monapia angusta is scored as uncertain be- cause the opening is circular but prolongued to the sides in a furrow between the median field and lateral lobes. Character 27: Cavities on epigynal anterior pouch, 0 == a single pit or furrow, 1 = two cavities (Figs. 5, 30). Character 28: Central depression on epigynal median field, 0 = absent, 1 = present (Fig. 4), 2 = vestigial. Polymorphic entries and non- additivity justified in Ramirez (1995b). Character 29: Median pouch on epigynal me- dian field, 0 = absent, 1 = present (Fig. 12). Variable through individuals of M. dilaticollis (Nicolet 1849) (see Ramirez 1995b: figs. 35- 39). The small foldings found on M. angusta (Fig. 22) and M. fierro new species (Fig. 4) might be homologous to the median pouch; the character is coded as uncertain for these species. Character 30: Position of epigynal lat- eral lobes, 0 = separate, 1 = contiguous, 2 = fused with a median suture, 3 = fused without suture. Considered additive. Character 31: De- gree of fusion of proximal copulatory ducts, 0 — separate, 1 = fused with median wall, 2 — totally fused with common lumen. Consid- ered additive. The “unpaired copulatory duct” considered in the previous analysis as an independent character is subsumed here in the state 2. Character 32: Copulatory plug in mated females, 0 — absent, 1 = present. M. angusta was coded uncertain because only one female is known, which lacks plug. En- tries coded as absent are based on numerous specimens. Character 33: Shape of spermathe- cae, 0 = irregular, 1 = spherical or oval. Char- acter 34: Shape of copulatory ducts, 0 = thick, outline of duct not well distinct from that of spermatheca, 1 = thin at least on distal por- tion, well distinct from outline of spermathe- ca. Character 35: Shape of anterior copulatory ducts, 0 = narrow (Figs. 19, 34), 1 = wide (Ramirez 1995b: figs. 61, 79). Character 36: Thickness of walls of proximal copulatory ducts, 0 — thick, 1 = thin. Character 37: Length of female tibia + metatarsus III, 0 = longer than tibia IV, 1 = shorter than tibia IV, 2 == shorter than 75% of tibia IV. Considered additive. Character 38: One strong anterior spine on chelicerae, 0 = absent, 1 = present. Scored as polymorphic in Oxysoma, because the spines are absent in Oxysoma valdiviensis (Simon 1897) but present in some undescribed species. Character 39: Ventral spines on fe- male palp, 0 = absent, 1 = present. Character 40: Prolateral/ventral spines on anterior fem- ora, 0 “ absent, 1 = one (Figs. 24, 32), 2 = several. M. charrua new species and M. an- gusta have an oblique line of thick setae (Figs. 7, 20) in similar position as the thick spines of M. guenoana (Fig. 14), with which are pre- sumed to be homologous. Considered addi- tive. Character 41: Number of ventral spines on anterior tibiae, 0 = three pairs or less, 1 = four pairs or more. Character 42: Thickness of ventral spines on anterior tibiae, 0 = normal, slender (Figs. 24, 32), 1 = strong (Figs. 7, 14, 20). Character 43: Apical ventral spines on anterior tibiae, 0 — present, 1 — absent. The data matrix of Table 1 was analyzed under parsimony using implied weights (Go- loboff 1993, 1995), using Pee- Wee version 2.5.1 (Goloboff 1996a). This program assigns lower weight to characters showing more ho- moplasy. Internal steps of characters were as- signed as implied by polymorphic terminals with command ccode—. Polymorphisms in Anyphaeninae were not taken into account for this purpose, because in such large group most characters are polymorphic, and it seems im- proper to decrease the weight of characters be- cause of variability in a group so distant. A heuristic search of 100 independent Wag- ner trees, each followed by TBR branch swap- ping (command mult*100;) produced the same two trees in all replications, for any val- ue of the constant of concavity K (1 < K < 6). A strict consensus of the two trees is shown in Fig. 6. The cladograms have a length of 79 steps, consistency index (for in- formative characters only) of 0.65, a retention index of 0.83, a fit (sum of implied weights) Table 1. — Data matrix for Monapia species and outgroups, x = [01], y = [012], z = [02], ? = unknown, — = uncertain. Commands for additivity and intemal steps: ccode — 7 9 29 38 * = 1 9 * — . +5 7 10 30 31 37 40;. 418 THE JOURNAL OF ARACHNOLOGY ^ o CN o m o XOOOOOOrHOOOOOOOOOrHrHTH XOO^OOOOOOOOOOOOO^^vH XOO^OOOOOOOOOOOOO^^H^ XOOOOOOOOOOOOOOOOrHCNIfNl OOOOOOOOOOOOOOOOOrHrHrH OOOOOOOO XOOOOOOOOOrH^ OOOOOOOOOOOOOOOOO^CMCM >lOOOOOOOOOOOOvHvH^^OOO Xoooooooooooo^^^^ooo X CD CD ^ — I ^ — I — I ^ — I \ — I ^ — f ^ — I ^ — I \ — t ^ — I \ — I ^ — I \ — I K — I K — I ^ — I CD CD CD CD ^ — I 'I — I ^ — I \ — I ^ — 1 ; — t ^ — I ^ — I ^ — I ^ — I \ — I ^ — I ^ — I ^ — I ^ — I ^ — 1 XOOOOOOOOO^vH^^^OO^^f^- OOOOOOOOOOOOO^^CMCMOOO Xooooo. OOOOOO^CMCNrOrOOOO I 0000 0-000000 X XOOOOOOOOO^^vH XI I 10-000000^0 XI I I^OOOOOrHrHrH X O' CD CD ^ — I ^ — I \ — I ^ — 1 % — I \ — I ^ — I ^ — I ^ — 1 O O O O ^ ^ I CNI CN X X ' — I ' — I ' — I ^ ^ o o o o o ^ — I t — 1 — 1 t — I CD ^ — I 0-® ! OOOOOOOOOOO^^rHOO^O^O® i OOOOO Xo X XOO^rHO^^OOO- I I IrHOOOOOOOO^^^vH^tHOO- 1 I 10000000^00000000 0- I I 100000^^^^000000^0- I I I I I I I I 1 I o oooooooooo^ I I I — ^vHO CD ^ — I \ — I ^ — I X — I ^ ^ — I ^ — I ^ — I ^ — I \ — I o o ^ ^ o o o- ^ — I ^ — 1 G® ^ o c® ^ — I ^ — I G® OrH^^OOOOOOOOOOOOOOOG® XOOOOOOOOOOOOOO^vHOOO- XOOOOOOOOOOO^^tH^tHOOO- XOOOOOOOOOOOO^^vHrHOOO- I I lOrH^OOOOOOOOOOOOOO I I lO^vHOOOOOrH^OrHCMCMOOO- I OOOOOOO^rH O 5 1 I 1 ^ 1 ^ 1 ^ 1 ! 1 \ 1 r— I 5 1 OOOOOOOO NO OOOOOOOOOO CNCN^CNOOOOOCN CNOOOOOOOOO- vHOOOOOOOOO- OOOOOOOOOO XOOOOOOOrHOOOOOOOOrH^vH OO^H^HOOOOOOOOOOOOOOOO OOOOOOOO^OOOOOOOO^rH^ OOOOOOOOOOOOOOOOOOrHtH O O O O O O O O — I rH ^ — I ^ — I ^ ^ — I \ — I ^ — j ^ — I ^ — I T — I g-§ S 52 a u ^ ■g I ‘C u S'*. Q S i ^-5 ^ Q Q Q ?s S' S' e s o o S' S' S' Q Q sec o o o Monapiafierro 10000 00111 OOOOO OHIO 00001 1111- 00111 OOOOO 1100 Monapia Carolina 10000 0???? ?0??? ????? ????? 11111 00111 OOOOO 1100 Monapia tandil 10000 00111 OOOOO 01010 11000 ????? ????? ??00? 0000 RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA 419 of 346 for K = 3, and a rescaled fit of 0.76. The same data were analyzed under equal weights, with the program NONA version 1.5.1 (Goloboff 1996b), and produced identi- cal results. The string of conunands poly-; max; apo[; of Pee- Wee/Nona was used to list the unam- biguous synapomorphies common to the 270 possible dichotomous trees corresponding to the equally parsimonious resolutions of the politomies of the two trees. A parsimony jack- nifing analysis (Farris et al. 1996) was made in order to have an estimation of support for clades. This procedure evaluates the stability of each node to a particular perturbation of the data set, which is the deletion of a portion of the characters. Over many replications with randomly deleted characters, the frequency at which a given node is monophyletic gives a measure of the support of the node: strongly supported nodes are more likely to be found even in absence of part of the data. The par- simony jacknifing is preferred over bootstrap- ing, because it produces a more direct relation between group frequence and support (Farris et al. 1996: 114). It is also preferred over the Bremer support (Bremer 1994), because in the latter, the values are absolute differences of steps (or fit), and a given figure might be orig- inated by widely different relations among supporting and conflicting characters (Golo- boff 1996a). During the jacknifing analysis, 200 pseu- doreplications were made deleting randomly 30% of the characters each time, as imple- mented in the JAK and FQ programs of Pee- Wee. Because the purpose was to evaluate support instead of quick searches, a more ex- haustive search was made in each replication, instead of the kind of Wagner tree proposed in the original description of the method (“a fast approximate procedure, similar to the hennig command of Hennig86” (Farris et al. 1996: 113)). This procedure yields a better sensitivity of the support measure (Goloboff pers. conun.). For each pseudoreplicate matrix a Wagner tree with randomized sequence of addition of taxa was calculated and submitted to extended (TBR) branch-swapping, saving up to 50 most parsimonious trees (command search=;). Clades with positive support have a jacknife frequency between 0.5 and 1. DISCUSSION As in the previous analysis (Ramirez 1995b), Monapia appears supported by four characters (clade h, char. 18, 26, 28 and 32), of which two (char. 18 and 32) are also pre- sent in some taxa not included in this analysis. Compared with the previous fully resolved cladogram of Monapia species, this new anal- ysis, with additional taxa and characters, is a bit more ambiguous. The consensus tree has a basal tetrachotomy; the resolutions place ei- ther clade j or Monapia alupuran as sister group of clade n. Monapia tandil new species and M. vittata (clade i) are united by the prolateral denticles on the secondary conductor (char. 21), and the loss of the groove on the same sclerite (char. 17). M. vittata was previously considered the sister group of all other species of the genus, but in the present analysis its relationships are uncertain (see clade i), as those of M. alupur- an, previously considered as the sister group of clade n. Clade j is united by the spinose forelegs and the median epiginal pouch (char. 29, 40 and 41), and clade k by the presence of a double anterior pouch (char. 27), with a convergence in clade p and in M. alupuran. Clade 1 is supported by four homoplasious characters: the flattened carapace and elongate abdomen (char. 2 and 4) are convergent in Ox- ysoma, the absence of apical ventral spines on fore tibiae and the strong ventral tibial spines (char. 42 and 43) are convergent in several groups, of which Liparotoma Simon 1884 and Ferrieria Tullgren 1901 are here included. The very similar species M. guenoana and M. angusta are united in the strongly supported clade m by five characters: the elongate car- apace (char. 1), short third leg (char. 37), che- liceral spine (char. 38), and several ventral femoral spines on palp (char. 39) and first leg (char. 40 state 2). Clades n-q are unchanged from the previous analysis, although there are slight changes in the characters supporting them: the fusion of the secondary conductor to the tegulum (char. 22) is added as support to the clade n, the lobe on the unsclerotized area of secondary conductor is ambiguously optimized in clade p, one unclear state was deleted from character 14, and two logically dependent characters were considered here as one (see char. 31 above). Clades 1-q are strongly supported, with a 420 THE JOURNAL OF ARACHNOLOGY jac knife frequency of 0.9 or greater. The rather weak support of clade k, might be related only to the circumstance that M. Carolina new spe- cies is known only from females. Some of the species here described have a small sclerite arising from behind the median apophysis. Its shape can be pointed and thin (Figs. 27, 39), or triangular, or the sclerite can be apparently reduced to an indistinct sclero- tized area. Tasata Simon 1903 species have a lamellar sclerite in that position, but its ho- mology is unclear in most amaurobioidine genera. I preferred not to score or name this structure until a more detailed study across more amaurobioidine genera is done. In M. fierro and M. tandil the sclerite is pointed and thin. If this character is added to the matrix the same cladograms are obtained. Areas of distribution of Monapia species might be of interest for studies of vicariance biogeography. M. vittata, M. alupuran and clade n are endemic to temperate forests of southern Chile and adjacent Argentina. Most species of clade n are sympatric. Only the sis- ter species M. lutea and M. huaria appear to have non-overlapping distributions (Ramirez 1995b: 87): M. lutea ranges from Curico Province to Chiloe, while M. huaria was found only in Valparaiso Province and around Santiago city. Clade 1 is endemic to riparian areas of eastern Argentina and Uruguay. Mon- apia Carolina was found in central Argentina, where no Monapia species lives. M. fier- ro and M. tandil are endemic to grasslands of east-central Argentina. Perhaps the most pe- culiar distribution is that of M. vittata and M. tandil, a pair of sister species known from widely disjunct areas. However, these data might change when additional specimens of M. tandil, known only from one individual, are collected. The cladogram presented here does not suggest a specific hypothesis of area relationship at both sides of the Andes be- cause of the lack of resolution of the polytomy h: the six possible dichotomous resolutions propose mutually exclusive relationships of areas. Monapia Simon 1897 Monapia Simon 1897: 93, 96, 97, 101. Gerschman de Pikelin & Schiapelli 1970: 131. Ramirez 1995b: 78. Synonymy, diagnosis and description given in Ra- mirez 1995b. Monapia charrua new species (Figs. 7-13) Types. — Female holotype (MACN 9576) and male paratype (MACN 9577) from Ar- gentina, Entre Rios Province, Rio Gualeguay- chu and RN 14, 14 July 1985, M. Ramirez. Etymology. — The specific name is a noun in aposition taken from the Charruas, an in- digenous ethnic group that lived in the region where this species occurs. Diagnosis. — Males, females and juveniles resemble those of M. angusta and M. guen- oana by having an elongate body, but are dis- tinguished by the chelicerae lacking spines. Female (holotype). — Total length 6.63. Carapace 2.52 long, 1.68 wide, wider between coxae II-III. Length of tibiae/metatarsi: I miss- ing; II 2.16/1.76; III 1.40/1.24; IV 2.72/2.72. Palpal tarsus 0.90 long. Sternum 1.50 long. Spines (female from Rosario del Tala): Leg I: Femur d 1-1-1, p 0-0-dl-l-dl and an apical oblique line of thick setae (Fig. 7), r 0-dl-dl; tibia V 2-2-2-2-0, p and r 1-1; metatarsus v 2- 2-2-0, p and r 1-0, d O-pl-2. II: Femur d 1-1- 1, p and r 0-1-1; tibia v 2-rl-2-0“2, p 1-1, r 1-vl; metatarsus v 2bas, p dl-1-0-1, r dl-vl- 0-1, d pl-2. Ill: Femur d 1-1-1, p 0-1-1, r lap; tibia V pl-pl-2, p and r dl-1, d rl-1; metatar- sus V 2-0-2, p and r dl-1-1, d pl-2. IV: Femur d 1-1-1, p and r lap; patella r 1; tibia v pi-2- 2, p and r 1-1, d rl-1; metatarsus v 2-2-2, p, r and d = III. Palp: Femur ventrally with me- dian line of 8 spines, d 0-0-l-p2, p lap; patella p 2-0, d 1-0-1; tibia p 2-2, dl-1; tarsus v 2ap, p 1-1, r 1-0, d 2bas. Color: yellow with brown spots on the dorsal axis of body and sparse spots on dorsum (Fig. 8) and legs. Chelicerae with anterior longitudinal brown spot. Epigyn- um with anterior pouch wide, almost hemi- spherical, lateral lobes separate, median de- pression large, extended behind the anterior pouch (Fig. 12), usually occupied by a mas- sive plug. Copulatory ducts with thick walls, accesory bulb with long duct parallel to mar- gin of lateral lobes (Fig. 13). Male (paratype). — Total length 4.12. Car- apace 1.98 long, 1.28 wide. Abdomen 2.30 long. Length of tibiae/metatarsi: I 2.40/2.02; II 1.80/1.54; III 1.18/1.04; IV 2.06/2.22. Spines: Leg I: Femur d 1-1-1, p 0-0-dl-l-dl and an apical oblique line of thick setae, r dlap; tibia v 2-2-2-2-0, p and r 1-1; metatar- sus V 2-2-2-0, p and r 1-0, d O-pl-2. II: Femur RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA 421 Figures 1-3. — Monapia guenoana new species. 1, Female on leaf of Panicum prionitis; 2, Same, posterior view; 3, Male. d 1-Dl, p O-dl-dl, r dlap; tibia v 2-r 1-2-0, p and r dl-1; metatarsus v 2-rl-rl, p and r dl- 1, d O-pl-2. Ill: Femur = II; tibia v pi -pi -2, p and rdl-l,drl-l; metatarsus v 2-0-2, p and r 0-dl-l, d O-pl-2. IV: Femur d 1-1-1, p and r dlap; patella r 1; tibia v pl-2-2, p and r dl- 1, d rl-1; metatarsus v 2-2-2, p and r dl-1-1, d O-pl-2. Palp: Femur d 0-0-1-2, p dlap; pa- tella p d2, d 1-0-1; tibia p 2-2, d rl-1; cym- bium p 1-1-1, d 2-0. Color: as in female but darker. Copulatory bulb (Figs. 9-11) with paramedian apophysis sinuous, embolus long and thick, retrolateral portion of secondary conductor fused to tegulum. Natural history. — Most specimens were collected on the large grass Panicum prionitis (“paja brava”) in temporarily flooded riparian areas, in close sympatry with Monapia guen- oana. Distribution. — Riparian zones of Entre Rios Province in Argentina and Departamento Rocha in Uruguay. Other material examined. — ARGENTINA: Entre Rios: Same locality as types, 1 d 1 9 (MACN); Gualeguay, 20 August 1989, M. Ramirez, 1 9 (MACN); Rosario del Tala, 20 November 1988, M. Ramirez, 1 9 (MACN). URUGUAY: Departa- mento Rocha: Arroyo Sarandi del Consejo, ruta 9 km 251, 18 May 1993, M. Ramirez & F. Perez Miles, lc33juv. (MACN). Monapia guenoana new species (Figs. 1-3, 14-19) Types. — Female holotype (MACN 9578) and male paratype (MACN 9579) from Ar- gentina, Entre Rios Province, Gualeguay, 20 August 1989, M. Ramirez. Etymology. — The specific name is a noun 422 THE JOURNAL OF ARACHNOLOGY in aposition taken from the Guenoanes, an in- digenous ethnic group related to the Chamias, that lived in the region where this species oc- curs. Diagnosis.-^Males, females and juveniles resemble those of M. angusta and M. charrua by having an elongate body, but are distin- guished by the anterior legs with several ven- tral spines on femora (Fig. 14), and numerous (more than 7 pairs) strong ventral spines on tibiae. Female (holotypej.—Total length 6.95. Carapace 2.38 long, 1.46 wide, wider on coxa II. Length of tibiae/metatarsi: I 2.96/1.54; II 1.60/1.04; III 0.90/0.68; IV 2.76/2.08. Palpal tarsus 0.80 long. Sternum 1.38 long. Spines: Chelicerae with 1 strong basal anterior. Leg I (Fig. 14): Femur d 1-1-1, p 1 ap, v 2-2-2-2ap or 2-2-2-2-rl-rl-rlap; tibia ventrally with a prolateral line of 10 or 11, and a retrolateral of 13 or 14; metatarsus v 2-2-2-0, p y r 1 ap. II: Femur d 1-1-1, p and r 1 ap; tibia v rl-rl- r 1-2-0; metatarsus v 2-rl-rl, p 0-1-dl, r dlap. Ill: Femur ^ II; tibia v pl-pl-0, p 1-1, r 0-1, d r 1-0-1; metatarsus v rl-0, d 2ap, r lap. IV: Femur d 1-1-1, p 0-1; tibia v pl-2-rl, p and r 1-1, d r 1-0-1; metatarsus v 2-pl-O, p 2-0-1, r 1-1-2. Palp: Femur ventrally with median line of 4 or 7, d O-O-l-l, p lap; patella d lap, p 2bas; tibia p 3-2, d 1-1; tarsus d 2bas, p 1, v 2ap. Color: yellow, with dorsal pattern of brown spots (Figs. 1, 2) and two dark parallel patches at each side of spinnerets. Legs with spots as follows: femora I-IV d l-l-l-l,IIIp vlap; patellae I-IV d 3-1, III v pi wide; tibiae I-III d Ibas, IV d 1-1, p 0-1, r 1-0; metatarsi I-III d 1-1, IV with a longitudinal band, ba- sally wider. The terminal dark spots on ab- domen and the lines on hind metatarsus are conspicuous from above when the spider is at resting position (Fig. 2). Epigynum with an- terior pouch wide, lateral lobes separate, me- dian area depressed behind the anterior pouch (Fig. 18), copulatory openings usually occu- pied by one plug each. Copulatory ducts with thick walls, duct of the accesory bulb curved in a transversal plane (Fig. 19). Male (paratype). — Total length 3.20. Car- apace 1.80 long, 1.16 wide. Abdomen 2.08 long. Length of tibiae/metatarsi: I 2.84/1.78; II 1.54/1.02; III 0.80/0.64; IV 2.48/1.88. Spine arrangement similar to female, but much weaker. Ventral foreleg spines thin and not erect. Color: as in female; some individuals have dark first tibiae. Copulatory bulb (Figs. 15-17) with paramedian apophysis sinuous, embolus short and thick, with basal membrane extending half of its length, retrolateral por- tion of secondary conductor small, with a ven- tro-basal peak, prolateral portion without groove, displaced towards the base of embo- lus. Natural history.— Specimens were collect- ed at the bases of the large grass (1.5--2 m tall) Panicum prionitis (“paja brava”) in tem- porarily or permanently flooded riparian areas. In some localities they occur in close sym- patry with Monapia charrua. Females make a flattened egg-sac on the concave side of the leaves, and the egg-sac is covered by its cryp- tic body. The grasses where the spiders live have thin, quite rigid and straight leaves, with a V-section. The spider uses to walk along the concavity of the leaf keeping its legs I, II and IV aligned with the body and leaf axis (Figs. 1-3). While doing this, the forelegs are usu- ally not used to walk, but to palpate the sub- strate, and the articulations femur-patella of legs II and IV are almost not moved. In these legs the movements are mostly achieved by the tibia-metatarsus joints. Distribution. -^-Riparian zones of Entre Rios and northeast of Buenos Aires Provinces in Argentina, and Departamento Rocha in Uruguay. Other material examined. — ARGENTINA; Entre Rios: Same data as the types, 2? (MACN); Arroyo El Palmar and RN 14, 14 October 1984, M. Ramirez, 1 9 (MACN); Arroyo Gualeyan and RN 14, near Gualeguaychu, 2 November 1996, M. Ra- mirez 2d 2 9 (MACN); Rio Gualegaychu and RN 14, 10 December 1982, M. Ramirez & P. Goloboff, 3 9 (MACN); Rosario del Tala, 20 November 1988, M. Ramirez, 3 9 (MACN). Buenos Aires: Delta, es- tacion experimental INTA, July 1968, A. Bach- mann, 1 9 (MACN); Isla Talavera, 2 km E Z^ate, 3 November 1996, M. Ramirez 1349 (MACN). URUGUAY; Departamento Rocha: Arroyo Sar- andf del Consejo, rata 9 km 251, 18 May 1993, M. Ramirez & F. Perez Miles, 1 9 3juv. (MACN). Monapia angusta (Mello-Leitao 1944) new combination (Figs. 20^23) Arachosia angusta Mello-Leitio 1944: 357 (juve- nile holotype from Argentina, Buenos Aires Province, Tigre, Rio Guayraca, MLP 16100, ex- amined). Note: The holotype is a badly preserved and slight- RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MON APIA 423 Figures 4-5. — Monapia fierro new species. 4, Epigynum (arrow indicates the copulatory plug on me- dian depression); 5, Vulva, dorsal. (AB == Accesory bulb. AP = epigynal anterior pouch. DP = “dic- tynoid” pore of spermatheca. FD = fertilization duct. LL = epigynal lateral lobe. MF = epigynal median field. MP = epigynal median pouch. Sp = spermatheca.) ly crushed juvenile. Although the cheliceral spines are lost, their insertions are clearly visible, as well as the ventral spines on the foreleg. Diagnosis. — Females and juveniles resem- ble those of M. guenoana by having an elon- gate body and basal anterior spines on the chelicerae, but can be distinguished by having only 4 pairs of ventral spines on the anterior tibiae. Female (Mar del Tuyu). — Carapace dam- aged, bowed, ~2.50 long, ~1.70 wide. Ab- domen elongated, deformed. Length of tibiae/ metatarsi: I 3.40/2.48; II 2.26/1.64; III 1.28/ 1.20; IV 3.32/2.56. Palps long (Fig. 20), pal- pal tarsus 1.30 long. Sternum 1.44 long. Spines: Chelicerae with 1 strong basal ante- rior. Legs I: Femur d 1-1-1, p 0-0-vl-dl and and an apical oblique line of thick setae, r lap; tibia V 2-2-2-2-0, d rl-1 setae; metatarsus v 2- 2-2-0-0, d pl-pl-rl-0-2. II: Femur d 1-1-1, p 0-1-1, r lap; tibia v 2-rl-2-rl-0, p 1-1, d rl-1 setae; metatarsus v 2-rl-rl, p 1-0, d pi -2. Ill: Femur II; tibia v pl-pl, p 0-1, d rl-1; meta- tarsus V 2-0-pl, d pl-2. IV: Femur d = I; tibia V 2-pl-rl, p and r 1-1, d rl-1; metatarsus v 2- pl-pl, p 1-1, r 0-1, d 2-2-2 or 2-pl-2-2. Palp: Femur with ventral and prolateral lines of sev- eral spines, d 0-0- 1-1-1, p 0-1; patella p 2-3, d 1-1; tibia with a ventral line of slender spines, p 3-1-dl-O-O, d 1-1-0; tarsus v 2-2-2, d p2bas. Color (immature from Castillos, Fig. 21): yellow, with dark reddish-brown spots on legs and dorsum; abdomen with dorsal white guanine reticulum under cuticle. Epigynum with lateral lobes widely separated, median field depressed in the center and behind the lateral lobes, folded in its anterior half and elevated over the anterior pouch (Fig. 22). An- terior pouch with spherical cavity; the opening is prolonged on each side in the anterior bor- der of the median field. Spermathecae oval, accesory bulbs with long and sinuous ducts (Fig. 23). Male.— Unknown. Natural history. — The specimens were collected on Pampas grass Cortaderia seU loana (“cortadera”). Distribution. — Margins of Rio de La Plata and Parana in Buenos Aires and Uruguay. Other material examined. — ARGENTINA: Buenos Aires: Mar del Tuyu, February 1984, M. Ramirez 1 9 (MACN); Parana de las Palmas, 7 April 63, M. Galiano, 2juv. (MACN); Reserva Ota- mendi, 10 June 1997, M. Ramirez, L. Compagnuc- ci, C. Grismado, F. Uehara, ljuv. (MACN). URU- GUAY: Departamento Rocha: Laguna de Castillos, 19 May 1993, M. Ramirez & F. Perez Miles, 4juv. (MACN). Monapia fierro new species (Figs. 4, 5, 24-30) Types.— Female holotype (MACN 9580) and male paratype (MACN 9581) from Ar- gentina, Buenos Aires Province, Sierra de la Ventana, Cerro Negro, April 1974, Cesari. Etymology. — The specific name is dedicat- ed to the brave gaucho Martin Fierro, who 424 THE JOURNAL OF ARACHNOLOGY alupnran vlttata tandll flerro Carolina cbarrua guBnoana angusta sllvatica lutea 32[0] Figure 6. — Cladogram of Monapia species and outgroups. Unambiguous synapomorphies are noted on branches, character states are “1”, otherwise noted in brackets. Jacknife frequency for each clade noted in parenthesis. gives the name to the most popular Argenti- nian poem, written by Jose Hernandez. Diagnosis. — This species is most similar to M. Carolina, but females can be distinguished by the wider opening of the anterior epigynal pouch; furthermore, they usually have 5 pairs of ventral spines on anterior tibiae. Males are recognized by the short and thick embolus. Female (holotype).— -Total length 6.00. Carapace 2.88 long, 2.10 wide, wider on cox- ae III. Length of tibiae/metatarsi: I 2.06/1.72; II 1.80/1.60; III 1.54/1.58; IV 2.20/2.56. PaL pal tarsus 1.00 long. Sternum 1.50 long. Spines: Legs I: Femur d UUl, p O-O-Udl, r lap; tibia with two lines of ventral spines, 4 or 6 prolateral, 5 retrolateral, (most commonly 2-2"2=-2“0=2), p Ul, r O-l; metatarsus v 2bas, p and r 1, d O-pl-2. II: Femur d Ul-1, p and r OHH; tibia v 2-2^2=-0^2 or 2=2^2^2^2, p M, r 0“1; metatarsus = 1. Ill: Femur " II; patella r 1, d lap; tibia v pl“2“2, p and r Ul, d rl-1; metatarsus v 2^0-2, p and r Ul, d 2-pU2. IV: Femur d UUl, p and r lap; patella r 1, d lap; tibia = III; metatarsus v 2-2-2, p and r 1-1, d 2-pl-2. Palp: Femur d 0-0- 1-2, p lap; patella d 1-1; tibia p 2-2, d 1-1; tarsus v p2ap, p 1- 1, r 1-0, d 2bas. Color: light brown, legs with brown spots, dorsum with brown pattern as in Fig. 25, sternum with three spots on each side, abdomen ventrally with three longitudinal brown stripes from epigastric furrow to the spiracle. Epigynum with central depression, lateral lobes with internal border sinuous, al- most touching on middle (Figs. 4, 29); median field elevated below anterior pouch, central depression prolonged in small foldings above central depression. Anterior pouch with two- fold cavity (Figs. 5, 30). Accesory bulbs with vertical ducts. Central epigynal depression of- ten occupied by hard plug (Fig. 4). Male (paratype).— -Total length 5.41. Car- apace 2.52 long, 1.90 wide. Abdomen 2.80 long. Length of tibiae/metatarsi: I 1.98/1.70; II 1.68/1.48; III 1.40/1.40; IV 1.98/2.20. Spines distributed as in female but weaker. Color: as in female. Cymbium with slight re- trolateral basal notch. Copulatory bulb (Figs. 26“28) with paramedian apophysis short and thick, embolus short and thick, with basal membrane extending half of its length, retro- lateral portion of secondary conductor with long and thin tip and small ventro-basal peak. There is a thin pointed sclerite of uncertain homology arising behind the median apoph- ysis. Natural history.- — Specimens were collect- ed in white retreats in bunches of grass. Distribution.— Known from grasslands in Buenos Aires and eastern Chubut Provinces. Other material examined. — ARGENTINA: Chubut: Isla de los Pajaros, Golfo de San Jose, 10 August 1975, M. Rumboll, 3 9 (MACN); Peninsula de Valdez, Punta Norte, 2 August 1972, M, Rum- boll, 1 9 (MACN). Buenos Aires: same locality as types, 2c529 ljuv. (MACN); Allen, August 1945, Concioti, 1 9 (MLP); Argerich, Villarino, June- July RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA 425 Figures 7-13. — Monapia charrua new species. 7, Female, lateral, 8, Female, dorsal; 9, Male palp, ventral; 10, Palp, retrolateral; 11, Palp, apical detail; 12, Epigynum, ventral; 13, Epigynum, dorsal, cleared. Abbreviations: AB = Accessory bulb of spermathecae; AP = epigynal anterior pouch; C2(p) = prolateral portion of the amaurobioidine secondary conductor; C2(r) = retrolateral portion of the amaurobioidine secondary conductor; E = embolus; LL = epigynal lateral lobe; MA = median apophysis; MH = median hematodocha; ME = epigynal median field; MP = epigynal median pouch; PMA = amaurobioidine paramedian apophysis; Sp = spermatheca; T = tegulum. Scales: Figs. 7-8 = 1 mm. Figs. 9-13 = 0.1 mm. 426 THE JOURNAL OF ARACHNOLOGY Figures 14-19. — Monapia guenoana new species. 14, Female, lateral; 15, Male palp, ventral; 16, Palp, retrolateral; 17, Palp, apical detail; 18, Epigynum, ventral; 19, Epigynum, cleared. Scales: Fig. 14 = 1 mm, Figs. 15-19 = 0.1 mm. 1958, H. Hepper, 2$ (MACN); Brandsen, no date, M. Biraben, 1 9 (MLP); D'Orbigny, November 1963, J.M. Gallardo, 1 9 (MACN); 15 Km W Lob= eria, 4 September 1972, 19 (MACN); Las Flores, 24 May 1931, J.M. Daguerre, 56149 (MACN 29951); Los Medanos, 8 April 1965, J.M. Gallardo & E. Maury, 26 5juv. (MACN); Mar del Plata, 20- 21 July 1984, M. Ramirez, 8 9 (MACN); Mar del Tuyu, 2 May 1981, M. Ramirez, 16 (MACN); Ola- varrfa, Sierra de la China, 19 November 1965, E. RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA All Figures 20-23. — Monapia angusta (Mello-Leitao). 20, Female, lateral; 21, Immature from Castillos; 22, Epigynum, ventral, 23, Epigynum, cleared. Scales: Figs. 20-21 = 1 mm. Figs. 22-23 ==0.1 mm. Maury, 3 9 ljuv. (MACN); Otamendi, 20 October 1979, P. Goloboff, 19 (MACN); Quequen, 7-12 May 1931, J.M. Daguerre, 1(369 (MACN 28871), June 1931, J.M. Daguerre, lc319 (MACN), 1941, E. Balech, 1 6 (MACN); Rio Lujan, estacidn FCGM, 14 November 1991, Pesce, 19 (MACN); Rosas (?), EC.S., 29 (MACN); between Tres Ar= royos and Pringles, November 1962, M.E. Galiano, 49 (MACN); Sierras de Azul, 1-2 October 1983, R Goloboff & A. Zanetic, 1 9 (MACN); Sierra de la Ventana, Pque. Prov. E. Tomquist, 18-20 No- vember 1982, M. Ramirez, 5 9 (MACN); October 1988, M.E. Galiano & C. Scioscia, 19 (MACN); Sierra de los Padres, 27 October 1984, M. Ramirez, 109 (MACN); Tandil, J.M. Viana, many c3 9 (MACN); Tomquist, estacion Fortin Chaco, January 1972, J. Arias, 1 9 (MACN). Monapia Carolina new species (Figs.’ 31-34) Types.™ Female holotype (MACN 9582) from Argentina, San Luis Province, Carolina, September 1970, J.M. Viana & Williner. Etymology.- — The specific name is a noun in aposition taken from the type locality. Diagnosis.™ This species is most similar to M. fierro, but females can be distinguished by the small opening of the anterior epigynal pouch; furthermore, the anterior tibiae usually bear 6 pairs of ventral spines (Fig. 32). Female (holotype). — Total length 6.38. Carapace 2.80 long, 2.00 wide, wider on cox- ae III. Abdomen 3.88 long. Length of tibiae/ metatarsi: I 1.92/1.58; II 1.70/1.44; III 1.44/ 1.48; IV 2.10/2.40. Palpal tarsus 0.94 long. Sternum 1.50 long. Spines as in M. fierro, but usually 2-2-2-2-2-0-2 ventral spines on tibia I. Color: light brown with brown spots and patches (Fig. 31), legs with brown spots at spine bases, and ventral spot lines on femora: one retrolateral on I and II, two on III, and one prolateral on IV; abdomen ventrally light, with short median band of brown spots. Epi- gynum with separate lateral lobes, median area elevated above anterior pouch, central de- pression deep, extended forward in central pouch (Fig. 33). Anterior pouch twofold, with conical cavities directed to each side (Fig. 34), accesory bulbs with ducts vertical to oblique. Male. — ^Unknown. Natural history. — Unknown. Distribution. — Known from Cordoba and San Luis Provinces. Other material examined. — ARGENTINA j 428 THE JOURNAL OF ARACHNOLOGY Figures 24-30, — Monapia fierro new species, 24, Female, lateral (arrow indicates prolateral/ventral spine on femur); 25, Female, dorsal; 26, Male palp, ventral; 27, Palp, retrolateral; 28, Palp, apical detail; 29, Epigynum, ventral; 30, Epigynum, cleared. Scales: Figs, 24-25 = 1 mm, Figs, 26-28 = 0.1 mm. RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA 429 Figures 31-34. — Monapia Carolina new species. 31, Female, dorsal; 32. Left foreleg (arrow indicates prolateral/ventral spine on femur); 33, Epigynum, ventral; 34, Epigynum, cleared. Scales: Figs. 31, 32 = 1 mm. Figs. 33, 34 = 0.1 mm. Cordoba: La Cumbre, 8 November 1991, M. Ra- mirez, 1 9 (MACN); Pampa de Achala, El Condor, 20 November 1983, M.E. Galiano, 1 9 (MACN). San Luis: same locality as types, 3 9 . Monapia tandil new species (Figs. 35-39) Types. — Male holotype (MACN 9583) from Argentina, Buenos Aires Province, Tan- dil, no date, J.M. Viana, Etymology. — The specific name is a noun in aposition taken from the type locality. Diagnosis. — This species is closest to M. vittata, as both have a flat apical extension on the amaurobioidine paramedian apophysis, and can be easily distinguished from M. vit- tata by the thin median apophysis. Male (holotype).- — Total length about 5.60. Carapace 2.50 long, 1.80 wide, wider on cox- ae III. Abdomen badly preserved, about 3.10 long. Length of tibiae/metatarsi: I 3.46/3.03; II 2.73/2.30; III 1.97/1.70; IV 2.87/2.97. Spines: I: Femur d 1-1-1, p and r 0-1-1; tibia V 2-2-2, p and r 1-1; metatarsus v 2bas, p and r 1, d O-pl-2. II=I. Ill: Femur d M-1, p 0-0- 1-1-1, r 0-1-1; tibia v pl-2-2, p and r 1-1, d 0- 1; metatarsus v2-pl-2, p and r dl-1-1, d 0- pl-2. IV: Femur d 1-1-1, p lap, r 0-1-1; pa- tella r 1; tibia and metatarsus = III. Palp: Fe- mur d 0-0-l-p3; patella d lap; tibia p 2-2, d 1- 1; cymbium p 0-1-1. Color: light brown with brown pattern on carapace as in Fig. 35, legs with brown spots, tibiae III and IV darker, with two longitudinal light lines; sternum with one oval brown patch in front to each coxae I-III, and median longitudinal line; abdomen light gray, with brownish-violet anterior me- dian patch, several inverted “V” through the median line, the three posterior darker, and a very dark patch upon the anal tubercle. Cym- bium with slight retrolateral basal notch (Fig. 36). Copulatory bulb (Figs. 37-39) with para- median apophysis short and thick, prolateral portion of secondary conductor without groove, displaced towards the base of embolus and bearing small denticles. There is a thin. 430 THE JOURNAL OF ARACHNOLOGY Figures 35-39. — Monapia tandil new species, 35, Male carapace; 36, Male palp, retrolateral; 37. Male palpal bulb, apical detail; 38, Palp, ventral; 39, Palp, retrolateral. Scales: Fig. 35 = 1 mm, Fig. 36 = 0.2 mm, Figs. 37-39 = 0.1 mm. curved and pointed sclerite of uncertain ho- mology arising behind the median apophysis. Natural history. — Unknown. Distribution.— Known only from type lo- cality. Other material examined. — None. Monapia dilaticolUs (Nicolet 1849) Clubiona dilaticolUs Nicolet 1849: 436. M. dilaticolUs, Ramirez 1995b: 78. Additional records.— CHILE: REGION IX: Malleco: Rio Blanco, Curacautm, 1-5 February 1959, L. Pena, 19 (IRSN LG. 19736). Unknown locality: El Coigo, 1-10 October 1960, L. Pena, 1 9 (IG 19736, IRSN). Monapia vittata (Simon 1884) Tomopisthes vittatus Simon 1884: 135. Monapia vittata, Ramirez 1995b: 81. Additional records. — ARGENTINA: Chubut: La Hoya, 42°54'S, 7ri9'W, 16 November 88. V. & B. Roth, 5 9 (CAS). Monapia lutea (Nicolet 1849) Clubiona lutea Nicolet 1849: 429. Monapia lutea, Ramirez 1995b: 86. Additional records. — ARGENTINA: Neu- quen: P. Nac. Lanm: 5 km E Hua Hum, 5 November 1981, Pucara, Nielsen & Karsholt, 5(389 (ZMK); February 1963, S. Schajovskoy, 1 9 (MACN); No- vember 1971, L. Yinoff, 19 (MACN); December 1973, S. Schajovskoy, 26 (MACN); San Martin de los Andes, 640 m, 17-31 October 1981, Nielsen & Karsholt, 2(3 (ZMK). Rw Negro: Bariloche, 12-20 November 1981, Nielsen & Karsholt, 19 (ZMK); 810 m, 22 November 1978, Nielsen & Karsholt, 1 9 (ZMK); El Bolson, 24 November 1962, Biraben, 6(369 (MACN); February 1965, Biraben, 19 (MACN); Chubut: R Nac. Lago Puelo, 220 m, 18 November 1978, Mision Cientifica Danesa, 1(3 (ZMK); Parque Nacional Los Alerces: March 1974, Bordon, 1 9 (MACN); Lago Futalaufquen, January 1990, M.J. RanuTez, 5 9 (MACN); Lago Menendez, RAMIREZ— NEW SPECIES AND CLADISTIC REANALYSIS OF MONAPIA 431 Rfo Arrayanes, February 1986, M. Ramirez, 69 (MACN); Villa Futalaufquen, 9 February 1986, M. Ramirez, 19 (MACN). CHILE: REGION VIII: Concepcion: Hualpen, 2 January 1989, M. Ramirez, 1 9 (MACN). REGION IX: Malleco: Fundo Maria Ester, 15 km W Victoria, 14 January 1989, M. Ra- mirez, 29 (MACN); Monumento Natural Contul- mo, 12 January 1989, M. Ramirez, 19 (MACN). Unknown locality: El Coigo, 1-10 October 1960, L. Pena, 1 9 (IG 19736, IRSN); 36, 1 9 (IG 15.765, IRSN). REGION X: Valdivia: 18.20 km NW Nel- tume, 25 November 1988, V. & B. Roth, 1 9 (CAS). Osomo: P. Nac. Puyehue, Aguas Calientes, 40°44'S, 72°19'W, 1440 m, 5-7 December 1988, V. & B. Roth, 29 (CAS). ACKNOWLEDGMENTS All curators of institutions listed above are greatefully acknowledged for the loan of spec- imens. Helpful comments on versions of the manuscript were provided by Maria Elena Ga- liano, Pablo Goloboff, Petra Sierwald, James Berry, and two anonymous reviewers. Mirta Arriaga provided valuable help in the identi- fication of grasses. Specimens were reared in laboratory by courtesy of Maria Elena Galia- no, and maintained by Susana Ledesma. Pa- tricia Sarmiento provided valuable assistance in scanning micrographs. Support for this pro- ject was provided by a graduate fellowship and EXO085 fund from the Universidad de Buenos Aires, LITERATURE CITED Bremer, K. 1994. Branch support and tree stability. Cladistics, 10:295-304. Brescovit. A. D. 1997. Revisao de Anyphaeninae Bertkau a mvel de generos na regiao neotropical (Araneae, Anyphaenidae). Rev. Brasileira Zook, 13(1):1-187. Farris, J.S, VA. Albert, M. Kallersjo, D. Lipscomb & A.G. Kluge. 1996. Parsimony jackknifing outperforms neighbor-joining. Cladistics, 12:99- 124. Gerschman de Pikelin, B.S. & R.D. Schiapelli. 1970. El genero Monapia Simon 1897 (Araneae, Anyphaenidae). Revta Mus. Argentino Cien. Nat., Zool. 10(9):131-144. Goloboff, RA. 1993. Estimating character weights during tree search. Cladistics, 9:83-91. Goloboff, P.A. 1995. Parsimony and weighting: a reply to Turner and Zandee. Cladistics, 11:91- 104. Goloboff, P.A. 1996a. Pee-Wee version 2.5.1, pro- gram and documentation . Goloboff, P.A. 1996b. Nonaversion 1.5.1, program and documentation . Goloboff, P.A. 1997. Self-weighted optimization: Tree searches and character state reconstructions under implied transformation costs. Cladistics, 13:225-245. Lipscomb, D.L. 1992. Parsimony, homology and the analysis of multistate characters. Cladistics, 8:45-65. Mello-Leitao, C.F. 1944. Aranas de la provincia de Buenos Aires. Rev. Mus. La Plata, Zool. 111:311- 393. Nicolet, H. 1849. Aracnidos. III., Pp. 319-543, In Historia Fisica y Pohtica de Chile. Zoologia. (C. Gay, ed.). Paris. Ramirez, M.J. 1995a. A phylogenetic analysis of the subfamilies of Anyphaenidae (Arachnida, Ar- aneae). Entomol. Scandinavica, 26:361-384. Ramirez, M.J. 1995b. Revision y filogenia del ge- nero Monapia (Araneae, Anyphaenidae), con no- tas sobre otras Amaurobioidinae. Boln. Soc. Biol. Concepcion, 66:71-102. Raimrez, M.J. 1997. Revision y filogenia de los generos Ferrieria y Acanthoceto (Araneae: An- yphaenidae, Amaurobioidinae). Iheringia, ZooL, (82): 173-203. Sierwald, P. 1990. Note on a technique for cleaning female copulatory organs in spiders. American ArachnoL, 41:2. Simon, E. 1897. Histoire Naturelle des Araignees, 2(1). Paris, Librairie Encyclopedique de Roret. 192 pp. Manuscript received 15 December 1997, revised 18 December 1998. 1999. The Journal of Arachnology 27:432-434 THE FEMALES OF ANELOSIMUS DUBIOSUS AND ANELOSIMUS JABAQUARA (ARANEAE, THERIDIIDAE) Marcelo de Oliveira Gonzaga and Adalberto Jose dos Santos: Pos-graduagao em Ecologia, Departamento de Zoologia -IB -Universidade Estadual de Campinas. C.R 6109. CEP: 13083-970 -Campinas, Sao Paulo, Brazil ABSTRACT. The females of Anelosimus dubiosus Keyserling 1891 and Anelosimus jabaquara Levi 1956 are described and illustrated based on specimens collected in Jundiai, Sao Paulo, Brazil. Keyserling (1891) described Anelosimus dubiosus based on a male collected in Nova Friburgo, Rio de Janeiro, Brazil. Levi (1956) described Anelosimus jabaquara also based on a male. This latter species, however, was considered a junior synonym of A. dubiosus by the same author in 1963 (Levi 1963). Levi & Smith (1982) revalidated A. jabaquara, but until now the two species were known only from males. Both species build communal webs of sim- ilar size and architecture and are sympatric in some places, as in Serra do Japi, a forest re- serve close to Jundiai, Sao Paulo, Brazil. Dur- ing studies on the ecology of social spiders in this area, we encountered these and three other species of the genus, A. jucundus (O.P.-Cam- bridge 1896), A. ethicus (Keyserling 1884) and A. studiosus (Hentz 1850). Adult males of A. jabaquara and A. dubio- sus are found in colonies only during the re- productive period. During the rest of the year the identification of the species is based on the females, which are described in this paper. METHODS The format of the description follows Levi (1963). Complete measurements were taken from one specimen of each species, and ad- ditional measurements of total length, cara- pace length and width were taken from six specimens of each species. All measurements are in mm. The epigyna were observed and drawn using an Olympus SZHIO dissecting microscope. For observation of the internal genitalia, the epigyna were immersed and ex- amined in clove oil and drawings were made using an Olympus Bx50 microscope with a camera lucida attached. Coloration was de- scribed using specimens that had been fixed for two days. The material is deposited in the collection of Instituto Butantan, Sao Paulo (IBSP, Curator: A.D. Brescovit). Anelosimus dubiosus (Keyserling 1891) (Figs. 1-3) Theridium dubiosum Keyserling 1891: 187, pi. 6, fig. 133 (Male holotype from Nova Friburgo, Rio de Janeiro, Brazil, in British Museum of Natural History, not examined) Anelosimus dubiosus: Levi 1963: 34; Levi & Smith 1982: 277, fig. 4 Diagnosis. — Anelosimus dubiosus males can be distinguished from all other species of Anelosimus, except A. jabaquara, by the pres- ence of a half-moon shaped tegular process. This species differs from A. jabaquara by the long and filamentous embolus (Levi & Smith 1982, fig. 4). The females are similar to A. jucundus and A. jabaquara in coloration and shape of the epyginum, but can be separated from these and other species of the genus by the lateral loops of copulatory ducts (Fig. 3). Description. — Female: Carapace red with black rings around the eyes, clypeus and che- licerae orange, sternum and labium red, en- dites orange. Legs light brown with distal ends of segments darker, tibiae and femora with dark rings in the middle. Abdomen light brown with a dorsal median band with black spots (Fig. 1) ending with four transverse red strips, venter with a central dark band with a black ring around spinnerets. Posterior median eyes little more than their diameter apart, slightly less from laterals. Epigynum a lightly sclerotized, folded plate, with a small poste- rior projection (Fig. 2). Total length 4.29, car- apace 1.65 long and 1.12 wide. First femur 432 GONZAGA & SANTOS— FEMALES OF ANELOSIMUS 433 Figure 1-5. — Anelosimus dubiosus and Anelosimus jabaquara. 1. A. dubiosus, dorsal view of carapace and abdomen; 2. A. dubiosus, epigynum, ventral view; 3. A. dubiosus, internal genitalia, dorsal view; 4. Anelosimus jabaquara, epigynum, ventral view; 5. A. jabaquara, internal genitalia, dorsal view. Scales: Figs. 1, 2, 4 = 0.5 mm; Figs 3, 5 = 0.25 mm. 434 THE JOURNAL OF ARACHNOLOGY L79, patella and tibia 2.04, metatarsus 1.34, tarsus 0.75. Second patella and tibia 1.42, third 1.2, fourth 1.7. Variation: Endite apex, distal ends of legs segments, lung plates and base of chelicerae may be red. Dorsal median band and sternum are sometimes totally black, mainly in males. Spermathecae occasionally visible externally. Measurements (mean and standard deviation, n = 7): Total length 4.27 ± 0.36, carapace length 1.37 ± 0.14, carapace width 1.22 ± 0.08. Material examined* — BRAZIL: Sao Paulo, Jun- dial, Serra do Japi, 13-16 November 1997 (M.O. Gonzaga), 21916 (IBSP 14403). Anelosimus jabaquara Levi (Figs. 4--5) Anelosimus jabaquara Levi 1956: 414, fig. 18 (Male holotype from Jabaquara, Sao Paulo, Sao Paulo, Brazil (H. Sick col.), in American Muse- um of Natural History, not examined); Levi & Smith 1982: 277 (revalidated). Anelosimus dubiosus (Key selling): Levi 1963: 34. Diagnosis. — ^Males of A. jabaquara can be distinguished from A. dubiosus by the short embolus and smaller half-moon tegular pro- cess (Levi 1963, fig. 18). The coloration and epigynum of females are similar to that of A. dubiosus and A. jucundus, from which it can be separated by the structure of the internal genitalia. Like A. domingo Levi 1963, this species has broad copulatory ducts, but differs by their lateral insertion on the spermathecae and their trajectory, which covers the fertil- ization ducts almost completely (Fig. 5). Description. — Female: Coloration as in A. dubiosus. Posterior median eyes little more than their diameter apart, slightly less from laterals. Epigynum a lightly sclerotized, fold- ed plate, with a posterior small projection (Fig. 4). Total length 4.29, carapace 1.79 long and 1.26 wide. First femur 2.24, patella and tibia 2.57, metatarsus 1.9, tarsus 0.89. Second patella and tibia 1.73, third 1.48, fourth 1.9. Variation: Coloration varies as in A. dubio- sus. Spermathecae occasionally visible exter- nally. Measurements (mean and standard de- viation, n = 7): Total length 4.1 ±0.18, carapace length 1.62 ±0.14, carapace width 1.2 ±0.07. Material examined. — BRAZIL: Sao Paulo, Jundiai, Serra do Japi, 13-16 November 1997 (M.O. Gonzaga), 18946, 2 juv. (IBSP 14404). ACKNOWLEDGMENTS The study was financed by graduate grants from FAPESP to M.O. Gonzaga and CAPES to A.J. Santos. The authors are grateful to J.A. Coddington, A.D. Brescovit and E.H. Buckup for their helpful comments on the manuscript, and to P.C. Eterovick for help in preparation of the figures. LITERATURE CITED Keyserling, E. 1891. Die Spinnen Amerikas: Bras- ilianische Spinnen. Niimberg. Bauer & Raspe, Bd 3. 278 pp. Levi, H.W. 1956 (1957). The spider genera Neot- tiura and Anelosimus in America (Araneae: Theridiidae). Trans. American Micros. Soc., 75: 407-422. Levi, H.W. 1963. The American spiders of the ge- nus Anelosimus (Araneae, Theridiidae). Trans. American Micros. Soc., 82:30-48, Levi, H.W. & D.R.R. Smith. 1982. A new colonial Anelosimus spider from Suriname (Araneae: Theridiidae). Psyche, 89:275-277. Manuscript received 10 January 1998, revised 20 October 1998. 1999. The Journal of Arachnology 27:435-448 REVISION OF THE GROENLANDICA SUBGROUP OF THE GENUS PARDOSA (ARANEAE, LYCOSIDAE) Charles D. Dondale: Eastern Cereal & Oilseed Research Centre, Research Branch, Agriculture and Agri-Food Canada, Ottawa, Ontario KIA 0C6 Canada ABSTRACT. The groenlandica subgroup, which currently stands as a component of the Pardosa mod- ica group, is characterized by a flat conductor tip in the male palpus and comprises P. groenlandica (Thorell 1872), P. dromaea (Thorell 1878), P. bucklei Kronestedt 1975, P. tristis (Thorell 1877), and P. prosaica Chamberlin & Ivie 1947. Neotypes are designated to stabilize each of the Thorellian names iracunda, dromaea and tristis, all original material relevant to these names having been lost or destroyed. Individuals of Pardosa groenlandica and related species are among the largest and darkest of the genus Pardosa C.L. Koch 1848. Living on exposed mountain slopes and sum- mits, stony beaches or open prairies in North America and Siberia, often in great numbers, they appear to be an integral part of the in- vertebrate food chains in these habitats (Levi & Levi 1951, 1955; Lowrie & Gertsch 1955; Schmoller 1970; Lowrie 1973; Moring & Stewart 1994). Taxonomic knowledge of these spiders has not kept pace, however, and be- haviorists and ecologists have found it impos- sible to identify their specimens with confi- dence. This applies particularly to collections from the Cordillera of western North America and the adjacent Great Plains, where popula- tions of two or more of the species apparently occur together and where additional, as yet unknown, species of the subgroup may occur. Pardosa groenlandica (Thorell 1872), the earliest known species of the assemblage, was originally described from the west coast of Greenland, but the range was later thought to extend across northern Canada to Alaska and the Russian Far East (Dondale & Redner 1990). Three additional species, P. iracunda (Thorell 1877), P. dromaea (Thorell 1878) (originally described under the name indaga- trix, which was preoccupied in the genus Ly- cosd) and P. tristis (Thorell 1877), had mean- while been described from Colorado. Thorell thought that his specimens of iracunda were extremely similar to and possibly conspecific with those described earlier as groenlandica, but he did not resolve the problem. Emerton (1894) examined the type material of iracun- da, dromaea and tristis, and could find no characters by which to distinguish the three species from P. groenlandica or from each other. The type of P. groenlandica exists (see below), but those of P. iracunda, P. dromaea and P. tristis then became lost or destroyed, making it impossible for any subsequent workers either to confirm or to refute Emer- ton’s conclusions. Progress toward a solution began when Kronestedt (1975) published illustrations of the palpus of a male of P. groenlandica which he had compared with the types. In the same paper he added a new prairie species, P. buck- lei Kronestedt 1975, to the assemblage; but he was unable to shed any light on the status of iracunda, dromaea or tristis. Dondale & Red- ner (1990) determined the probable identity of P. dromaea using data on the type locality, body size and female genitalia as given by Thorell (1877) or as illustrated by Emerton (1894). Dondale & Redner (1990) also showed that the range of P. bucklei extends into the Cordillera of western North America. The problems posed by the loss of type ma- terial and the failure to find diagnostic char- acters for P. groenlandica and related species are addressed in the present paper. HISTORY OF THE NAMES IRACUNDA, DROMAEA AND TRISTIS In the summer of 1875 the eminent New England entomologist A.S. Packard, Jr. made a month-long trip to the Front Range of the Rocky Mountains of Colorado and to the Great Salt Lake, Utah to collect arthropods. According to travel information relevant to the time (Holbrook 1947; Bowles 1977), 435 436 THE JOURNAL OF ARACHNOLOGY Packard probably travelled to Cheyenne, Wy- oming on the transcontinental Union Pacific Railroad, then southward to Boulder, Colora- do on the Cheyeene-Denver Railroad, which had opened in 1870. From the dates given by the identifier of Packard's specimens (Thorell 1877) we can infer that the collector thee pen- etrated the mountainous area west and south- west of Boulder, ascending '‘Arapaho Peak” (either North Arapaho Peak or South Arapaho Peak), “the Blackhawk” (probably Black Hawk Mountain), “Kelso cabin” (probably a miner’s shack on Kelso Mountain) and Grays Peak, with stops for collecting at Golden, “Idaho” (Idaho Springs) and Georgetown. Packard then descended to Denver, where he collected briefly, and proceeded southward, probably using the stagecoach, to “Manitou” (Manitou Springs), “Garden of the Gods” (probably Garden of the Gods Park) and Pikes Peak. His final arachnid collections on the trip were from American Fork Canyon and Great Salt Lake, Utah in late July. Packard sent his arachnids to Tamerlan Thorell in Sweden. Thorell identified 30 spe- cies of spiders, 23 of which he described as new to science (Thorell 1877). There was also a new species of harvestman. Nearly all of the spider material was later returned to Packard, who in turn placed it in the hands of J.H. Emerton. Emerton (1894), in a paper dealing with his own collections from the Lake Louise area of Alberta, mentioned that “Prof, Pack- ard has sent to me the spiders described by Thorell from the Rocky Mountains ...” Moreover, Emerton provided the first illustra- tions of any of ThorelFs types and, in an ad- dendum to ThorelFs paper, described two more species of spiders based on Packard’s Colorado material. Today the only known specimens are a male and female of Pardosa sinistra (Thorell 1877) (see Kronestedt 1981) and a female of Pardosa uncata (Thorell 1877) (see Lowrie & Dondale 1981) in the Swedish Museum of Natural History, Stock- holm, perhaps overlooked when Thorell re- turned the collection to Packard. Enquiries by me at the Swedish Museum of Natural History and the Museo Civico di Storia Naturale “Giacomo Doria” in Genoa (where some of ThorelFs types are deposited), at the Peabody Museum of Natural History in Boston (where Packard was custodian for some time) and at the Museum of Comparative Zoology, Har- vard University (where most of Emerton’s col- lections are now stored) failed to uncover any other relevant types. The original description of Pardosa iracun- da was based on one syntype male from the 13,000 ft. (3965 m) level on Pikes Peak, Col- orado and one syntype female from “Kelso Cabin”, Colorado. My search, with Jim Red- ner, of the upper levels of Pikes Peak in 1985 yielded only females, and we never found the locality “Kelso Cabin” on any highway map or gazetteer of Colorado. A male from a mountainous locality near Pikes Peak and de- posited in the American Museum of Natural History is selected as neotype of iracunda (see below). This specimen fits ThorelFs de- scription of iracunda. The holotype female of P. dromaea was collected at Denver, Colorado. Because of the difficulty of finding undisturbed habitat in that city, Redner and I concentrated our search along the banks of the South Platte River where it flows through the northern outskirts of the city. A single male specimen, which we judged to be conspecific with fe- males in turn agreeing with ThorelFs descrip- tion of dromaea, was found. This male, though of the sex opposite to that of the type, is designated neotype of dromaea (permitted by the rules of nomenclature where stability of nomenclature is thereby ensured) (Article 75(d) (4), ICZN). The original material of P. tristis consisted of two syntype females, one from “Idaho” (Idaho Springs), Colorado, the other from “Manitou” (Manitou Springs), Colorado. Our searches at these localities failed to produce specimens that represented any of the species here assigned to the groenlandica subgroup, probably because of severe destruction of hab- itat for road construction in those areas. A specimen from a locality in the same general area as Idaho Springs is therefore selected as neotype of tristis (see below). This specimen matches ThorelFs description of tristis. Court- ship and mating behavior observations of P. groenlandica were made at various locations in the field. METHODS The term retrolateral process of terminal apophysis is used here for a structure of the male palpus (see Figs. 1-5). This structure was illustrated for a male of P. wasatchensis DONDALE—PARDOSA GROENLANDICA SUBGROUP 437 Figures 1-5. — Distal part of male palpus of Pardosa spp., retrolaterobasal view. 1, P. groenlandica, Sondrestrom Air Base, west Greenland; 2, P. tristis, 32 km northwest of Weiser, Idaho; 3, P. bucklei. Bear Lake, Utah; 4, P. dromaea. Fountain Valley, Colorado; 5, P. prosaica. North Fork Pass, Yukon Territory. Abbreviations: CON = conductor, E = embolus, PTA = retrolateral process of terminal apophysis, TERM = terminal apophysis. Gertsch 1933 by Kronestedt (1993), who called it the retrolateral grooved process of the terminal apophysis. To see this process it is necessary to remove the apical division of the genital bulb (with the embolus, conductor and terminal apophysis intact) from the tegulum (see Dondale & Redner 1990 for definitions of terms). If the preparation is viewed prola- terobasally, the retrolateral process of terminal apophysis is seen as a tooth, ridge or similar structure on the side of the terminal apophysis near its base. A second character used frequently in this work is the epigynal ratio. This is calculated from piq X 100, where p is the distance from the anterior end of the median septum to the anterior margins of the atrial sclerites, and q is the total length of the median septum (Fig. 15). Differences in the means of epigynal ra= tios are compared using an ANOVA. The following abbreviations are used for museums: AMNH (American Museum of Natural History, New York, New York); CNC (Canadian National Collection of In- sects and Arachnids, Ottawa, Ontario); MCZ (Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts); NRS (Swedish Museum of Natural History, Stock- 438 THE JOURNAL OF ARACHNOLOGY Figures 6-15.' — Female genitalia of Pardosa spp., ventral view. 6, 7, 9, 14, 15, Epigynum; 8, 10-13, Median septum. 6, P. dromaea, Woolford Provincial Park, Alberta; 7, P. tristis, Salmon Arm, British Columbia; 8, F. tristis, Penticton, British Columbia; 9, P, bucklei. Mormon Lake, Arizona; 10, 11, P. prosaica, Magadan area, Russia; 12, P, prosaica, Hyndman Lake, Northwest Territories; 13, F. prosaica. East Chukotka Peninsula, Russia; 14, F. groenlandica, Sondrestrom Air Base, west Greenland; 15, F. prosaica. Old Crow, Yukon Territory. Abbreviations: ATS = atrial sclerite, MS = mediae septum, p = distance from anterior end of median septum to anterior margin of atrial sclerite, q — length of median septum. DONDALE— PA/?Z)OM GROENLANDICA SUBGROUP 439 holm, Sweden); IBPN (Institute for Study of Biological Problems of the North, Magadan, Russia). RELATIONSHIPS Dissections of male palpi indicate that a re- trolateral process of terminal apophysis is pre- sent in males of most members of the Pardosa modica group, of which the groenlandica sub- group is a component (for current composition of the modica group see Kronestedt 1975, 1981, 1986, 1988, 1993 and Dondale & Red- ner 1990). The retrolateral process of terminal apophysis is apparently absent in male P. modica (Blackwall 1846), and it apparently occurs in males of at least a few species in other groups of the genus Pardosa. In males of two lineages of modic a- species the retrolateral process of terminal apophysis occupies much of the space be- tween the base of the terminal apophysis and the free part of the conductor (Figs. 1-5). This expression of the process, called “large,” is restricted, in the first place, to males of P. groenlandica, P. dromaea, P. bucklei, P. tris- tis and P. prosaica. Males of all five of these species are characterized by the unique pos- session of a flat conductor tip (Figs. 1-5). To- gether these species comprise the groenlan- dica subgroup. The second lineage in which the males have a large retrolateral process of terminal apophysis comprises P. wasatchensis Gertsch 1933, and some related species that are the subject of a future investigation. Within the groenlandica subgroup one can discern two further lineages based on the shape of the 'retrolateral process of terminal apophysis. In males of P. groenlandica and P. dromaea this structure is a thin upright ridge that arises angularly from the free part of the conductor (Figs. 1, 4). Males Of P. bucklei, P. tristis and P. prosaica, on the other hand, have a retrolateral process of terminal apophysis with a long low margin (Figs. 2, 3, 5). SYSTEMATICS Pardosa groenlandica (Thorell) Figs. 1, 14, 16 Lycosa groenlandica Thorell 1872: 157. Syntype male from Disko Island, West Greenland (69°15'N, 3°32'W (Th. Fries), deposited in NRS (Thorell Collection No. 244/ 15 24a), examined and here designated LECTOTYPE. Syntype fe- male from the type locality, 3 July 1871, depos- ited in NRS, examined and here designated PAR- ALECTOTYPE. Lycosa iracunda Thorell 1877: 514. Syntype male from Pikes Peak (38°50'N, 105°02W), 3965 m elevation, El Paso County, Colorado, 14 July 1875 (A.S. Packard, Jr.), and syntype female from ‘Kelso Cabin’ (probably on Kelso Mountain, 39°33'N, 105°47'W), Clear Creek County, Colo- rado, 6 July 1875 (A.S. Packard, Jr.), both lost or destroyed. NEOTYPE male from Pikes Peak, 3660 m elevation, El Paso County, Colorado, 24 June 1940 (W.J. Gertsch and L. Hook), deposited in AMNH, here designated. Name iracunda first synonymized under groenlandica by Emerton (1894: 423), here confirmed. Pardosa groenlandica’. Emerton 1894: 423(part); Kronestedt 1975: 218, figs. 3c, 4C, 4c; Dondale & Redner 1990: 212, figs. 300-304. Diagnosis. — Males of P. groenlandica are distinguished from males of other species of Pardosa except P. dromaea by the possession of a ridgelike retrolateral process of terminal apophysis, the margin of which is angular (Fig. 1). I have not found a character in the male palpus by which to distinguish males of P. groenlandica from those of P. dromaea, but the former males are significantly larger (Table 1), and they court by approaching fe- males after drumming their bodies on the sub- strate (rather than while drumming). In addi- tion, individuals of P. groenlandica range in the western North American Cordillera, in northern and eastern Canada and in Greenland (rather than the Central Plains), and they oc- cupy stony or gravelly habitats (rather than prairie habitats). Females of P. groenlandica differ from those of other species of Pardosa except P. dromaea in the shape of the median septum: this structure widens in the posterior two-thirds or three-fourths and is parallel or somewhat convex at the lateral margins (Fig. 14). Females of P. groenlandica are signifi- cantly larger than those of P. dromaea (and those of P. bucklei) (Table 1), and the range and habitats differ as stated for males. The epigynal ratio of 71.1 ±4.6 (Colorado sample) is significantly larger (P < 0.05) than that of P. dromaea (Table 2). Description. — Male: Carapace black to dark reddish-brown, with pale median area, and with lateral bands represented by three or four yellowish spots. Legs black to reddish- brown, distally yellowish; femora, tibiae, and basitarsi often with broad pale rings. Abdo- men dull black, usually mottled with dull red. Table 1, — Measurements (in mm) of Pardosa groenlandica, P. dromaea, P. bucklei, P. tristis and P. prosaica. Sample size is 20 individuals of each sex. Data for P. bucklei are from Dondale & Redner (1990). Means significantly different within columns bear same superscript. 440 THE JOURNAL OF ARACHNOLOGY ■o 2 a 'S BO •h m (X ov A m CO CN| 04 04 d d d d d s « +1 +i +1 +1 +1 IT) Ov VO ri Ov q q q fb 04 04 cb cb 4) O Dh « n efl o 'C o O p, p u P-i .S ‘v i-i o •S 'K o d o & S O P. fl N S S u U 3 o p. -9 sd 3 ’C d « 3 m « 50 d bO s d © 'd >> d d 2 ^ &.0 J-T © ■§ a 4) Pn OT O 3 ^ 50 3 ^ B C M 3 ■§ O u "S ^ u 0 1 -o CO U o S a-S i e § a d « 8 3 o O a 441 442 THE JOURNAL OF ARACHNOLOGY Figure 16. — Collection localities of Pardosa groenlandica (•, circles) and P. dromaea (A, tri- angles). they hatched. It is possible that Schmoller in- cluded specimens of P. tristis in his work, as specimens of both that species and of P. groenlandica are now known to occur on Mount Evans. Repetition of his work is need- ed on known specimens. The same applies to an excellent study of P. groenlandica on a pebbly beach at Flathead Lake, Montana (Ri- cards 1967), where populations of P. tristis and of P. bucklei occur together with one of P. groenlandica. Courtship by males of P. groenlandica has been observed some 29 times by J.H. Redner and/or myself at localities on the Atlantic coast of Canada, at Bagotville, Quebec and in Colorado. On detecting a female, the male al- ways ceased moving for a time, drawing his body close to the substrate and extending his legs I stiffly forward. The body was then raised while the palpi were held forward and downward. While in this posture, the male be- gan to move his palpi in small circles a little above the substrate. The right palpus circled clockwise and the left one counterclockwise. Three or four circles was the usual number, both palpi moving at the same time. Then one palpus or the other was raised quickly to an angle of approximately 80° with the substrate, held aloft for an instant, and quickly lowered to the resting position. Both palpi then re- sumed circular motions for a second or two, after which the alternate palpus was raised, held, and lowered as the first had been. This sequence was repeated several times while the spider remained at one spot. After a number of such palpal sequences, the male suddenly raised his body to maxi- mum height, then lowered it in a rapid series of four or five taps against the substrate so that at the end of the series his body lay flat against the substrate and his legs were held outspread. The tapping series was accompa- nied by a rapid vibration of legs I, usually alternately but occasionally in unison; this leg action produced an audible rattle. At comple- tion of the tapping series, the male moved at moderate speed toward the female, his legs vibrating audibly on the substrate. This for- ward dash was usually interrupted one or more times by a resumption of palpal circling. Eventually the male was close enough to the female to vibrate his legs on her body and legs, and to mount and copulate. Courtship lasted about 15 minutes. In a typical mating the male inserted his right or left embolus into the corresponding copulatory tube of the female as follows: 1 insertion on left side, 1 on right, 2 on left, 1 on right, 1 on left, 1 on right, 1 on left, 2 on right, 1 on left, 1 on right, 1 on left, 2 on right, 1 on left, 1 on right, 1 on left. Each insertion was accompanied by one to five brief pulsa- tions of the palpal haematodocha and each pulsation was followed by a brief partial de- flation of that organ. The epigynal ratio of P. groenlandica ap- pears to vary geographically. This ratio was calculated for 16 females from each of seven parts of the range, with the following results: Utah (68.5 ± 5.8); Colorado (71.1 ± 4.6); Wyoming (73.5 ± 3.6); Montana (71.9 ± 4.8); Alberta/Northwest Territories (69.2 ± 5.2) Greenland (68.9 ± 4.3); Atlantic Provinces of Canada (66.6 ± 3.4). The sample mean for the Atlantic Provinces is smaller than means for the other six areas. This difference is statistically significant for the samples from Colorado {P < 0.01), Mon- tana {P < 0.01) and Wyoming {P < 0.01), and approaches the 5% level of significance for Alberta/Northwest Territories (P < 0.09). I in- DONDALE— PAi?DOSA GROENLANDICA SUBGROUP 443 fer that the population living on the Atlantic coast, which is somewhat isolated from the others, has developed small but measurable differences. I have not found a corresponding difference in the palpi of Atlantic coast males. A second point to note from the value for epigynal ratio is that the Wyoming sample, which gives the largest mean, is statistically different from those from Utah {P < 0.01), Alberta/Northwest Territories {P < 0.05), Greenland {P <0.01) and the Atlantic coast {P < 0.01), and approaches that level for the Montana {P < 0.08) and Colorado {P < 0.08) samples. This Wyoming population seems to invite further investigation, but was not so treated in this study. The mean epigynal ratio did not differ significantly between any pair of samples from Utah, Colorado, Montana, Alberta/Northwest Territories or Greenland. In the southern parts of the range of P. groenlandica, individuals occur on alpine tun- dra or among bare rocks at or above timber line. On the Atlantic coast, however, popula- tions thrive on pebbly or cobblestone beaches at or somewhat above sea level. Pardosa dromaea (Thorell) Figs. 4, 6, 16 Lycosa indagatrix Thorell 1877: 512. Holotype 9 from Denver (39°44'N, 104°59'W), Denver County, Colorado, 10 July 1875 (A.S. Packard, Jr.), lost or destroyed. NEOTYPE male from South Platte River at 88th Street, Denver, Denver County, Colorado, 20 June 1985 (C.D. Dondale & J.H. Redner), here designated, deposited in CNC. Name indagatrix preoccupied in genus Ly- cosa. Lycosa dromaea Thorell 1878: 395. New name for Lycosa indagatrix, preoccupied. Pardosa groenlandica: Emerton 1894: 423, fig. lb (pi. 4). Holotype female of P. indagatrix illus- trated. Pardosa nebraska Chamberlin & Ivie 1942: 30, figs. 69, 70 (pi. 7). Holotype d.from 6 km west of Lexington (40°50'N, 99°55'W), Dawson Coun- ty, Nebraska, 6 June 1933 (W. Ivie), deposited in AMNH, examined. Name nebraska first synony- mized by Dondale & Redner (1990). Pardosa dromaea: Simon 1898: 359; Dondale & Redner 1990: 209, figs. 305-307. Diagnosis.— "Males of P. dromaea are dis- tinguished from those of other species of Par- dosa except P. groenlandica by the posses- sion of a ridgelike retrolateral process of terminal apophysis, the margin of which is an- Figure 17. — North American collection localities of Pardosa prosaica (▼, solid triangles), P. bucklei (V, hollow triangles) and P. tristis (■, squares). gular (Fig. 4). Males are significantly smaller than those of P. groenlandica (Table 1), and they tap their bodies on the substrate while dashing toward the female rather than tapping while remaining stationary. They also occur in prairie habitats on the Central Plains rather than in the Cordillera and the northern parts of the North American continent. Females of P. dromaea differ from those of other species of Pardosa except P. groenlandica in pos- sessing a median septum that widens in the posterior two -thirds or three-fourths and has parallel or convex lateral margins (Fig. 6). Fe- males of P. dromaea are significantly smaller than those of P. groenlandica and significant- ly larger than females of P. bucklei (Table 1), and differ in range from the former (see Figs. 16, 17). The epigynal ratio of 67.9 ±4.0 is significantly smaller {P < 0.05) than that of P. groenlandica (Colorado sample), P. bucklei (P < 0.01), and P. tristis {P < 0.01) (Table 2). Description. — Male: Carapace black to dark reddish-brown, with median band often reduced to a small spot, and with submarginal bands represented by three or four small yel- lowish spots. Legs black to dark reddish- 444 THE JOURNAL OF ARACHNOLOGY brown, often with darker rings on femora and tibiae. Abdomen black dorsally, mottled with dull red, with pale heart mark; venter reddish- brown to yellowish. Palpus (Fig. 4) dark, hairy; tegulum protruding at base; median apophysis small, with slender sinuous basal process; embolus broad at base, with small flange at tip; terminal apophysis toothlike, curved, with tip straight or deflected some- what basad (retrolaterobasal view. Fig. 4); conductor broad, flat; retrolateral process of terminal apophysis large, ridgelike, with an- gular margin. Measurements: see Table 1. Female: Coloration essentially as in male, but legs somewhat paler. Epigynum (Fig. 6) with flask-shaped atrium; median septum slen- der anteriorly, broader in posterior two-thirds or three-fourths, with lateral margins parallel or convex; atrial sclerites broad, prominent; epigynal ratio: see Table 2; spermathecae long, curved, club-shaped, with small nodules (Dondale & Redner 1990, fig. 307). Measure- ments: see Table 1. Material examined. — There are 244 adult spec- imens, all bearing my label, deposited as VOUCH- ERS in the following institutions: AMNH (373469); CNC (88389 9); MCZ (3319); A pair of VOUCHER specimens from Moring & Stewart’s (1994) study in Colorado is deposited in CNC. Range. — Eastern foothills of the Rocky Mountains in Alberta, Montana, Wyoming, Colorado and New Mexico, east to southern Manitoba, Minnesota, Iowa and Nebraska (Eig. 16). The species is regarded as a member of the Great Plains fauna. Biology. — Specimens of P. dromaea have been collected from late April to early Sep- tember. I have observed male courtship only twice, but Donald J. Buckle (pers. comm.) has observed it many times. The specimens were respectively from the Lethbridge, Alberta area and from Saskatoon, Saskatchewan. The se- quence of palpal and leg movements appeared to be like that of male P. groenlandica (see above), but the body tapped the substrate while the male dashed toward the female rath- er than while stationary. Courtship lasted ap- proximately 20 minutes, copulation 25-30 minutes. The insertion series in one mating was as follows: 2 insertions on right side, 1 on left, 1 on right, 1 on left, 1 on right, 1 on left, 1 on right, 1 on left, 1 on right, 1 on left, 1 on right, 1 on left, 1 on right, 1 on left. Each insertion was accompanied by 1-8 brief pul- sations of the haematodocha. An ecological study was recently published on P. dromaea (reported as P. tristis) by Mor- ing & Stewart (1994). The principal habitats of P. dromaea along the Conejos River in south-central Colorado were reported as rock- cobble, grass-willow, and sand-cobble. Leaf litter near the river produced no specimens of P. dromaea. Pardosa bucklei Kronestedt Eigs. 3, 9, 17 Pardosa bucklei Kronestedt 1975: 224, figs. 2c, 3 Figure L — Critical thermal maxima (CTn^a^’s) of Misumenops asperatus, Misumenoides formosipes, noctumally-active spiders and diumally-active spi- ders. Data for nocturnal and diurnal spiders were taken from Table 2; winter- active species were not included in the analysis. Values with different let- ters are significantly different using post-hoc Mann- Whitney U tests with adjusted a = 0.009. Error bars represent one SE for nocturnal and diurnal spiders and one SD for M. asperatus and M. formosipes. Standard errors could not be used for M. asperatus and M. formosipes since species averages were cal- culated from individual values. In contrast, noctur- nal and diurnal averages were calculated from spe- cies averages, making use of the standard error appropriate. erature data for diurnally-active and noctur- nally-active spiders revealed significant dif- ferences (Kmskal- Wallace test: H = 23.878, df = 3, P < 0.0001). Preferred temperatures were similar between M. formosipes and M. asperatus, and between M. formosipes and nocturnal spiders (Fig. 2). However, Tp’s of M. asperatus and nocturnal spiders differed sig- nificantly, and Tp of diurnal spiders was sig- nificantly higher than that of M. formosipes, M, asperatus, or nocturnal spiders (Fig. 2), Relationship between and thermal preference.— Using the literature data shown in Table 2, a significant positive relationship was found between a spider species’ and its (F = 6.707, df = 1, 19, P - 0.018, = 26.1%) (Fig. 3). Because M. asperatus and M. formosipes had unusually high CTj^^^’s for their Tp’s, data for these two species were not included in this analysis. Addition of M. asperatus and M. formosipes data to the re- gression resulted in a loss of the significant relationship (F = 1.615, = 1, 21, F = 0.2177, F = 7.1%). For similar reasons, data for winter-active spiders were also excluded. Field temperature -—Misumenoides for- mosipes experienced preferred temperatures under field conditions more frequently than did M, asperatus. Ambient temperature fell within the preferred temperature range (PTR) of adult M. asperatus 43% of the time, ex- ceeded PTR 47% of the time, and fell below PTR 10% of the time (Fig. 4). In contrast, ambient temperature fell within PTR of adult M. formosipes 65% of the time, exceeded PTR 10% of the time, and fell below PTR 25% of the time (Fig. 4). During their adult phase, M. asperatus experienced an increase in average daily temperature of 5.3 °C, and M. formosi- pes experienced a decrease in average daily temperature of 6.5 °C. DISCUSSION Comparing the thermal tolerances and pref- erences of the spring-maturing M. asperatus with those of the summer-maturing M. for- mosipes yielded the expected results. Adult fe- male M. asperatus, which experienced lower ambient temperatures than did adult female M. formosipes, had a lower CT^^j^, while M. for- mosipes had a higher CT^a^- Thermal prefer- ences of the two crab spider species were sim- ilar. Of greater interest is a comparison of the thermal tolerances and preferences of these flower-dwelling spiders with those of other spider species. There is little information available regard- ing CTn,in in spiders. In the single study of which I am aware, Hagstrum (1970) reported a CT^jn of 6 °C for a southern California wolf spider, Alopecosa kochi (Keyserling 1877) (as Tarentula kochi in Hagstrum 1970). Much more data is available concerning lower lethal temperatures and temperature effects on de- velopmental rates (e.g., Almquist 1970; Li & Jackson 1996). Data from other studies indi- cates that temperate-climate spiders are gen- erally capable of activity at relatively low temperatures. Ford (1978) showed that a Eu- ropean wolf spider, Pardosa amentata Clerck 1757, remained active at 5 °C, and Moulder & Reichle (1972) obtained similar results for the litter-spider fauna of a Tennessee Lirio- SCHMALHOFER— CRAB SPIDER THERMAL ECOLOGY 475 dendron forest. Aitchison (1984) found that both winter-active and winter-inactive Cana- dian spiders fed at 2 °C, with winter-active species continuing to feed at temperatures as low as “5 °C. These studies, coupled with the CTn,in values calculated for M. asperatus and M. formosipes, suggest that temperate-zone spiders from moderate climates may generally be expected to have CTj^jn’s near 0 °C. Misumenops asperatus, and particularly M. formosipes, had high thermal tolerances. In general, CTn,ax’s of these flower-dwelling thomisids were more similar to the upper le- thal temperatures than to the CTj^^x’s of other spider species (see Table 2). Of the 27 species for which data was available, only six species had comparable PhuroUthus festivus (C.L. Koch 1835), Euophrys frontalis (Wal- ckenaer 1802), Cyrtophora citricola (ForskM 1775), Zelotes longipes (L. Koch 1866) (as Z serotinus in Almquist 1970), Hogna caroli- nensis (Walckenaer 1805) (as Lycosa caroli- nensis in Moeur & Eriksen 1972), and Seo- thyra henscheli (Dippenaar 1991). Misu~ menoides formosipes had the second highest CTn^a,, recorded for a spider; only S. henscheli, an eresid from the Namib desert (Lubin & Henschel 1990), had a higher thermal toler- ance. The natural histories of M. asperatus and M. formosipes may provide an explana- tion for their unusually high thermal toleranc- es. These thomisids do not stalk prey, but, rather, position themselves close to a flower’s nectaries and/or anthers (pollen-bearing struc- tures) in order to ambush flower-visiting in- sects (pers. obs.). On the plants used by M. asperatus and M. formosipes, the floral sur- face from which nectaries and anthers are ac- cessed by insects is typically exposed to the sun (pers. obs,). Under conditions of high ra- diant intensity and low wind speed, body tem- peratures of spiders on sun-exposed floral sur- faces can exceed ambient temperature by 15 °C or more (Schmalhofer 1996). A high ther- mal tolerance would allow M. asperatus and M. formosipes to continue hunting at ambient temperatures near 30 °C, when floral surface temperatures could be in excess of 40 °C, Am- bient temperatures approaching 30 °C are not an uncommon occurrence in late spring and summer in central New Jersey: from April through September in 1993-1995 there were, on average, 52 days per year having a daily high temperature of at least 30 °C. One would expect that because diumally- active spiders experience higher temperatures than noctumally-active species, diumally-ac- tive spiders would prefer higher temperatures. Evaluation of data from the literature showed that this was indeed the case (Fig. 2). How- ever, Tp’s of M. asperatus and M. formosipes were lower than those of other diumally-ac- tive species, and Tp of M. asperatus was also lower than that of noctumally-active species! Pulz (1987) suggested that, barring winter- ac- tive spiders, lower thermal preference corre- lates with lower thermal tolerance. Regression analysis of the available literature data sup- ported Pulz’s hypothesis of a positive relation- ship between and CTj^^. Interestingly, the CTjnax's of M. asperatus and M. formosipes predicted from the regression equation (40.2 °C and 41.3 °C, respectively) were much low- er than the measured values; alternatively, predicted Tfs (32.1 °C and 43.4 °C, respec- tively) were much higher than the measured values. Thus, depending on how one looks at it, M. asperatus and M. formosipes have ex- ceptionally high thermal tolerances or excep- tionally low thermal preferences. This com- bination of high thermal tolerance and low thermal preference is unusual for an ecto- therm; preferred temperature is usually nearer the upper than the lower tolerance limit (May 1985). Broad temperature tolerances and relatively low thermal preferences displayed by M. as- peratus and M. formosipes may be viewed as adaptations that facilitate their diurnal preda- tory lifestyles in potentially thermally stress- ful habitats (sun-exposed flowers), Hymenop- terans and dipterans comprise most of the prey captured by these thomisids (Schmalhofer 1996). Dipterans are well known for their ability to fly at low temperatures (reviewed in Heinrich 1993); and large hymenopterans, such as honeybees and bumblebees, require thoracic temperatures of 30-35 °C in order to fly (reviewed in Heinrich 1993). The capacity to endothermically generate heat by shivering wing muscles allows these bees to fly at low ambient temperatures; honeybees can fly when ambient temperature is as low as 15 °C, and some bumblebees can fly when ambient temperature is less than 10 °C (reviewed in Heinrich 1993). Both M. asperatus and M. formosipes prey on honeybees, and M. for- mosipes is also capable of capturing bumble- 476 THE JOURNAL OF ARACHNOLOGY O fli 2 I -S 3 f :s ej ^ ^ 4) 4h M o O ° 00 d ‘o I M g O 05 I .2 I o m H ■3 5 •= I ■§ -S eJ a 1 is 00 .2 s r- r- r-- r- r- ON r- r- ©0 ©0 00 (» 00 m 3 0 .2 0 w 4J Ov ON ON ON ON « ON ON d 0 5 1 & a M 00 2 3 N N 3 N 0 0 0 N 0 05 0 3 N 0 0 0 0 0 3 i “ « Sh 3 0 'p r-. i> r- ’d d r- r- f» r- t-- r^ s 0 OT £ £ £ £ On ON ON 4) ON n . £ On On ON ON On ^ . . 0) 0 d NM NM l-H »— 1 NM X nM NM 00 d li S N « d d d d .s ■4-^ .9 00 .9 _1 ^ w J 0 N 3 ,2 ..H M m ON m m b g.i ^ «5 g 'S Uii « P. -jii^ c-^ C^ Cl ■3 “P CN 'B CN 3 05 3 3 3 3 3 3 3 § O' s S s 3 3 d 3 p a 3 3 f-l 0 •iJ 3 3 3 a a i a cr a ^ C^ U > & s 3 < < < 3 < m cd +-> < < < < < 3 '0 0 ■0 -o V d 4) « _> ■p S d .y 3 0 d 0 4— c« 1 0 0 j q q ¥-1 ^ .d c« d -o 00 cd 0 0 S h." On 1 1 vd 1 1 d 1 s I 1 0 0 & s ^ s 0 “O 1 m 0) 3 3 4_ 3 ’■§ § d > 3 8 q q q 0 /-^ ’o '3 00 e D S d 0 « feO 2 05 3 p 0 EA IT) cn m 5 : cx-d u ^ S 3 •y > > e^ w ^ d I -J»- +- 4— 05 >, 0 « 3 +- -o d fT q c6 q cn q A q CN cn r4 0 VD cn q N q; ON q 3 vd P M {^ €N| CN| CN fN| CN »— 1 ¥—1 i-M CN 'f=*l S 0 P S « 3 0.4-, •¥ « 3 max 0 "o H a 0 s T) 3 S c^ 05 ci 2 U a X £ !Z1 K 4- M a: 4- a a 4~ +- •0 -g ^ g 8.M bO d cd 3 H U q 0 VO CN CN 0 On vd VO ON q 00 d VD d VI d d \D d c« }-( u cn cn »n csi add * fl -a a a fl o\ •I M 1 1 PE i> ^ M cn 00 o s « s M M « “ K oe3 I •2 o 60 « bD 3^- ^ S -p Is 8 u _ ^ '6 . o J « G ^ e ^ ^ .y bO^ < u < < ^ -a .2 a o ^ U P u S3 ^ ^ see OOP 5 u u u « ^ Q ^ s s s S£ o o\ VO M “ p jq g o 5 Q ^ w ei s •2 Q « S3 W 5^ 2 m u 60 8 5 3 ^ O •S ^ ? & e^ Co C O 65 2 .S •S’ -S ^ 60 1 s O 6, ^ 3 ^ .5 3 ^ g O g ^ k. .y 0, ^ a § 3 > S’ s « d 'd d 0 IT) VI cn 0 VN m « « bO 00 ON IT) VO 00 <—i •O W VI 3 fp. ’u 1-4 00 N S cn 00 0 3 a 0 3 1 0 « S 3 S 05 u n • 3 1 q 6, pi a 5 50 o S s I: S k'S i g ■2 o Co .2 U S 6, ffl ■c ^ ^ -S’ I. o »3 ^ p o ^ ^ s 6b ^ ^ 0. g 6b •S ’fc* o u CO SCHMALHOFER— CRAB SPIDER THERMAL ECOLOGY 477 rJ 00 r' 00 00 00 r- 00 l> 00 r-- 00 00 r- 00 4) os g Os Os os os o o Os os os os (D ’— ' T— i hJ hJ t-H P o C/D £ w 3 3 Dh o. o os 3 eu 3 Oh 3 &H 3 Oh 3 Ph 3 Ph o o r- r-- os os o Os fN| r- os •S d8 fN u .S .S a •S ■p ■p •S •S .a • S +-I +-> iH CN c« 3 5 k. o, (a Q 3 .2 -2 A^ A^ a, a, 3 I S ^ 5 3 -a a, a, a, a, e « 50 T3 ■p s u d w r- ^ IH « a o 0 1 s 5o K -a o Eo d «3 Q .2 a .2 §-|t s 2 taj * % P 50 (SJ a > o 4) 00 cn q q q q 00 in rP in c6 in 'It cn -{- +- 4- ■i— +- +~ +- +_ CM q q ’—1 q q q QO CM q q m o SO fO CM 00 ^ Os q OS cn Os fP OS CM CM CM fO CM cn CM CM CM CM CM CM o c- m oq oq q q q os d os '-H CM in —H cn cn cn CM CM t-H -—I T— 1 CM CM CM CM aH 1— 1 T3 ’d d d d d d d xT d d d d m 3 3 3 3 3 3 3 3 3 3 3 3 3 3 d d s' 4) C " ^ 3 a, o a -s s & 2 C^ -a 1^ H Cj M o s -2 P a A i-. a Clj Q S i H (as Theridion saxatile in Pulz 1987) Crustulina guttata Wider 1834 n*, 1 41.8 19.6 Almquist 1970 Dipoena inornata O. P.-Cambridge 1861 n*, 1 43.2 15.5 Almquist 1970 Zoridae Zora spinimana Sundevall 1833 d, 1 41.8 19.5 Almquist 1970 478 cd 1-1 « e o 0) 301 25 20 15 ab 10 A'J-^ .<,# / / A^'■' />* y ^o- ^ 4' ' -c^' Figure 2, — Preferred temperatures of Misumen- ops asperatus, Misumenoides formosipes, noctur- nally-active spiders and diumally=active spiders. Data for nocturnal and diurnal spiders were taken from Table 2; winter-active species were not in- cluded in the analysis. Values with different letters are significantly different using post-hoc Mann- Whitney U tests with adjusted a = 0.009. Error bars represent one SE. M. formosipes Figure 3. — Linear regression of preferred tem- perature against CTj^a^- The regression was based on literature data presented in Table 2. Regression equation: y = 0.275x + 36.26. Although not in- cluded in the regression, for comparative purposes data points for Misumenops asperatus and Misu- menoides formosipes were included in the graph. Squares (■) represent other spiders species, circles (•) represent M. asperatus and M. formosipes. THE JOUR2^ AL OF ARACHNOLOGY CT max CT min Figure 4. — The relationship between average dai- ly high and low temperatures and the preferred tem- perature ranges (PTR’s) of Misumenops asperatus and Misumenoides formosipes. Periods of adult ac- tivity are demarcated with vertical bars. (up- per dashed line), CT^un (lower dashed line), and zone of thermal discomfort (TD, dotted lines) are indicated for each spider species. The bounds of a species’ PTR (dot-dashed lines) were calculated as one standard deviation around mean T^. bees (Schmalhofer 1996). Presumably many of the other bees used by these spiders, such as anthophorids and megacMlids, display tem- perature-flight relationships similar to those shown by honeybees and bumblebees. The broad temperature tolerances shown by M. as- peratus and M. formosipes allow these spiders to increase their foraging time, both daily and seasonally. Coupled with their ability to hunt equally well over a wide range of tempera- tures (Schmalhofer 1996), and their low ther- mal preferences, broad thermal tolerances benefit these spiders by affording them the op- SCHMALHOFER— CRAB SPIDER THERMAL ECOLOGY 479 portunity to hunt prey that is itself active over a wide range of environmental temperatures. Assuming that an ectotherm benefits by maintaining body temperature within some preferred range of temperature (Hertz et al. 1993), comparing preferred temperature rang- es (PTR’s) of M. asperatus and M. formosipes with the average range of temperatures nor- mally experienced by these spiders allows one to make predictions concerning the likelihood that the spiders will experience thermal stress (i.e., unfavorable temperatures that might lim- it activity or impair performance). Data indi- cate that M. asperatus experiences high ther- mal stress (ambient temperature > PTR) more frequently than low thermal stress (ambient temperature < PTR), while M. formosipes ex- periences low thermal stress more frequently than high thermal stress. In neither case did average daily low temperature fall below spi- der Thus, since ambient temperature did not fall to levels that would inhibit spider movement, M. asperatus and M. formosipes could alleviate low thermal stress by engaging in behaviors designed to elevate body tem- perature (e.g., basking in the sun). Normal hunting behavior (i.e., sitting near a flower’s nectaries and/or anthers, provided the position was exposed to the sun) would serve to achieve this result. High thermal stress appears to be more of a concern for these thomisids since ambient temperature typically falls within or above a spider’s preferred range, and even when am- bient temperature falls within the preferred range, floral-surface temperatures, and thus spider body temperatures, may be much high- er, potentially approaching Most diur- nally-active spiders avoid high, stressful tem- peratures by some behavioral mechanism (Pulz 1987). Ground- or vegetation-dwelling cursorial (non-web-building) spiders can move to shaded areas under twigs, leaves, stones, etc., while web-building spiders may have a shaded retreat associated with the web. In both cases, these spiders still have access to prey when in shade and can continue to hunt. This option of behavioral thermoregu- lation (shuttling between sun and shade) may not be available to flower-dwelling thonusids if the spiders are to maintain access to prey. Because M. asperatus and M, formosipes do not strike at prey unless it approaches within a few millimeters of the spider’s chelicerae (pers. obs.), spiders must remain near the an- thers and/or nectaries in order to have access to prey. In the open held habitats where M. asperatus and M. formosipes are typically found, the upper surfaces of flowers, where the anthers and nectaries are located, are gen- erally sun-exposed. Thus flower-dwelling thomisids may be faced with a trade-off be- tween maintaining access to prey and avoid- ing high thermal stress. ACKNOWLEDGMENTS T. Casey, R Morin, M. May, D. Morse, P. Harris, C. Kaunzinger, J. McGrady-Steed, J. Fox, M. Cole, W. Schmalhofer, Y. Lubin, and an anonymous reviewer gave helpful critiques of earlier drafts of the manuscript. N. Platnick provided information concerning currently ac- cepted species names and authorities. I thank W., L., and V. Schmalhofer for their encour- agement and moral support. Partial support was provided by a National Science Founda- tion Dissertation Improvement Grant (DEB 93-11203), and grants from the Anne B. and James H. Leathern Scholarship Fund and the William L. Hutcheson Memorial Forest Re- search Center. LITERATURE CITED Aitchison, C.W, 1978. Spiders active under snow in southern Canada. Symp. Zool. Soc., London, 42:139-148. Aitchison, C.W. 1984. Low temperature feeding by winter-active spiders. J. ArachnoL, 12:297-305. Almquist, S. 1970. Thermal tolerances and pref- erences of some dune-living spiders. Oikos, 21: 230-236. Anderson, J.F. 1970. Metabolic rates of spiders. Comp. Biochem. Physiol., 33:51-72. Bayram, A. & M.L. Luff. 1993. Cold-hardiness of wolf-spiders (Lycosidae, Araneae) with particu- lar reference to Pardosa pullata (Clerck). J. Therm. Biol., 18:263-268. Carrel, J.E. 1978. Behavioral thermoregulation during winter in an orb-weaving spider. Symp. Zool. Soc., London, 42:41-50. Dodson, G.N. & M.W. Beck. 1993. Pre-copulatory guarding of penultimate females by males crab spiders, Misumenoides formosipes. Anim. Be- hav., 46:951-959. Duman, J.G. 1979. Subzero temperature tolerance in spiders: Role of thermal hysteresis factors. J. Comp. Physiol. (B), 131:347-352. Foelix, R. 1996. Biology of Spiders. Harvard Univ. Press, Cambridge. 330 pp. Ford, M.J. 1978. Locomotory activity and the pre- dation strategy of the wolf-spider Pardosa amen- 480 THE JOURNAL OF ARACHNOLOGY tata (Clerck) (Lycosidae). Anim. Behav., 26:31- 35. Hagstrum, D.W. 1970. Ecological energetics of the spider Tarentula kochi (Araneae: Lycosidae). Ann. Entomol. Soc. America, 63:1297-1304. Heinrich, B. 1993. The Hot-Blooded Insects: Strat- egies and Mechanisms of Thermoregulation. Harvard Univ. Press, Cambridge. 601 pp. Henschel, J.R., D. Ward & Y. Lubin. 1992. The importance of thermal factors for nest-site selec- tion, web construction and behaviour of Stego- dyphus lineatus (Araneae: Eresidae) in the Negev desert. J. Therm.Biol., 17:97-106. Hertz, P.E., R.B. Huey & R.D. Stevenson. 1993. Evaluating temperature regulation by field-active ectotherms: the fallacy of the inappropriate ques- tion. American Nat., 142:796-818. Humphreys, W.E 1974. Behavioral thermoregula- tion in a wolf spider. Nature, 251:502-503. Humphreys, W.E 1987. Behavioral temperature regulation. Pp. 56-65. In Ecophysiology of Spi- ders. (W. Nentwig, ed.). Springer- Verlag, Berlin. Kirchner, W. 1973. Ecological aspects of cold re- sistance in spiders: A comparative study. Pp. 271-279. In Effects of Temperature on Ectothermic Organisms. (W Wieser, ed.). Springer- Verlag, Berlin. Kirchner, W 1987. Behavioral and physiological adaptations to cold. Pp. 66-77. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin. Krakauer, T 1972. Thermal response of the orb- weaving spider, Nephila clavipes (Araneae: Ar- giopidae). American Mid. Nat., 88:245-250. Li, D. & R.R. Jackson. 1996. How temperature affects development and reproduction in spiders: A review. J. Therm. Biol., 21:245-274. Lockley, T.C, O.P. Young & J.L. Hayes. 1989. Nocturnal predation by Misumena vatia (Ara- neae, Thomisidae). J. Arachnol., 17:249-251. Lubin, YD. & J.R. Henschel. 1990. Foraging at the thermal limit: burrowing spiders (Seothyra, Eresidae) in the Namib desert dunes. Oecologia, 84:461-467. May, M. 1985. Thermoregulation. Pp. 507-552. In Comprehensive Insect Physiology, Biochemistry, and Pharmacology. Vol. 4. (G.A. Kergut & L. Gilbert, eds.). Pergamon Press, New York. Morse, D.H. 1979. Prey capture by the crab spider Misumena calycina (Araneae: Thomisidae). Oecologia, 39:309-319. Moeur, J.E. & C.H. Eriksen. 1972. Metabolic re- sponses to temperature of a desert spider, Lycosa {Pardosa) carolinensis (Lycosidae). Physiol. ZooL, 45:290-301. Moulder, B.C. & D.E. Reichle. 1972, Significance of spider predation in the energy dynamics of forest-floor arthropod communities. Ecol. Mon- ogr., 42:473-498. Pollard, S.D., M.W Beck & G.N. Dodson. 1995. Why do male crab spiders drink nectar? Anim. Behav., 49:1443-1448. Pulz, R. 1987. Thermal and water relations. Pp. 26- 55. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin. Riechert, S.E. 1974. Thoughts on the ecological significance of spiders. Bioscience, 24:352-356. Riechert, S.E. & C.R. Tracy. 1975. Thermal bal- ance and prey availability: Bases for a model re- lating web-site characteristics and spider repro- ductive success. Ecology, 56:265-284. Roberts, M.J. 1995. Spiders of Britain and North- ern Europe. Harper Collins Publishers, London. 383 pp. 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 posture and coloration. Zool. J. Linn. Soc., 64:87-102. Schaefer, M. 1977. Winter ecology of spiders (Ar- aneida). Zeit. Ang. Entomol., 83:113-134. Schmalhofer, VR. 1996. The Effects of Biotic and Abiotic Factors on Predator-Prey Interactions in Old-Field Flower-Head Communities. Ph. D. Thesis, Rutgers University, New Brunswick. 178 pp. Sevacherian, V. & D.C. Lowrie. 1972. Preferred temperatures of two species of lycosid spiders, Pardosa sierra and P. ramulosa. Ann. Entomol. Soc. America, 65:111-114. Seymour, R.S. & A. Vinegar. 1973. Thermal rela- tions, water loss and oxygen consumption of a North American tarantula. Comp. Biochem. Phy- siol., 44A:83-96. Suter, R.B. 1981. Behavioral thermoregulation: So- lar orientation in Frontinella communis (Liny- phiidae), a 6-mg spider. Behav. Ecol. SociobioL, 8:77-81. Turner, J.S., J.R. Henschel & YD. Lubin. 1993. Thermal constraints on prey-capture behavior of a burrowing spider in a hot environment. Behav. Ecol. SociobioL, 33:35-43. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge Univ. Press, Cambridge. 328 pp. Manuscript received 10 January 1998, revised I June 1998. 1999. The Journal of Arachnology 27:481-488 PARAPHYLY OF THE ENOPLOGNATHA OVATA GROUP (ARANEAE, THERIDIIDAE) BASED ON DNA SEQUENCES A.-M. Tan‘, R.G. Gillespie' and G.S. Oxford^: ^Center for Conservation Research and Training, University of Hawaii, 3050 Maile Way, Gilmore 409, Honolulu, Hawaii 96822, USA; ^Department of Biology, University of York, RO. Box 373, York YOl 5YW, UK ABSTRACT: Five species of Enoplognatha Pavesi 1880 were recently recognized as a monophyletic Enoplognatha ovata group based on morphological data. We analyzed the E. ovata clade for monophyly using four species in the E. ovata group {E. ovata (Clerck 1757), E. latimana Hippa & Oksala 1982, E. margarita Yaginuma 1964 and E. afrodite Hippa & Oksala 1983) and three other closely related taxa {E. japonica Bosenberg & Strand 1906, E. thoracica (Hahn 1833), and E. intrepida Sprensen 1898). Two species of the presumed sister genus (Steatoda Sundevall 1833) were employed as outgroups. The results indicate that the ovata clade” is not monophyletic. The genus Enoplognatha Pavesi 1880 is characterized by the presence of a large col- ulus, a plesiomorphic character for the family; and accordingly, the genus is generally con- sidered one of the more primitive groups in the Theridiidae. The spiders are medium-to- large sized with a subspherical abdomen. Fe- males have a tooth on the posterior margin of the chelicerae; males usually have enlarged chelicerae, with enlarged teeth on the poste- rior margin, and have the paracymbium on the margin of the cymbium. The genus is very close to Steatoda Sundevall 1833, medium-to- large sized spiders, again characterized by a very large colulus (Levi 1962; Levy & Amitai 1981). The chelicerae are often enlarged in males, and have one or more teeth on the an- terior margin, none on the posterior margin. Enoplognatha is well known because of the striking color and pattern polymorphism ex- hibited by representative species in the genus, which has been most intensively studied in E. ovata (Clerck 1757). Three distinct morphs have been described in E. ovata (Locket & Millidge 1951; Hippa & Oksala 1979; Oxford 1976): lineata (all yellow), redimita (yellow with two dorsolateral carmine stripes on the abdomen), and ovata (yellow with a solid shield of carmine on the dorsal surface of the abdomen). The color pattern variation in E. ovata is genetically determined, and has been the subject of numerous studies on the genet- ics and evolution of the color polymorphism (Hippa & Oksala 1979, 1981; Oxford 1983, 1985, 1989, 1991, 1992; Oxford & Reillo 1993; Reillo & Wise 1988a, b). Consistent with most invertebrate color polymorphisms (Haldane 1939) the dominance hierarchy of the expression of morphs in E. ovata follows the inverse of morph frequencies in nature, i.e., the least dominant (or most recessive) al- lele is most frequent; the most dominant is the rarest. For the mode of inheritance of the polymorphism in E. ovata, Oxford (1983) has proposed a two locus model: one locus is con- cerned with pattern and color, the other with the regulation of this color locus during de- velopment. When red-pigmented alleles are linked to the late developing allele, the color morphs are sex-limited: males are lineata no matter which allele they carry. Enoplognatha latimana Hippa & Oksala 1982 shares color, regulatory, and black spotting polymorphisms with E. ovata (Oxford 1992), although E. la- timana lacks the ovata color morph. In the 1980s Hippa & Oksala (1982) erect- ed the E. ovata group to include E. ovata sen- su stricto, E. latimana, and E. penelope Hippa & Oksala 1982. Members of the group share the following characters: trichobothrium on the first metatarsus subapical; elongated, sclerotized and subtubular tip of conductor in male palp; female vulva with massive copu- latory pockets and abdomen with sharply de- limited dorsolateral black spots (Hippa & Oksala 1982). Further examination of material 481 482 THE JOURNAL OF ARACHNOLOGY from Europe and Japan added another two species to the E. ovata group (Hippa & Oksala 1983): E. afrodite Hippa & Oksala 1983 and E. margarita Yaginuma 1964. E. margarita shares the subapical trichobothria and sclero- tized subtubular tip of the conductor with E. ovata, E. latimana and E. penelope. However, it lacks the massive copulatory pockets. Con- sidering all of these characters as synapomor- phies, Hippa & Oksala (1983) hypothesized that E. margarita was the closest sister to {E. ovata + E. latimana + E. penelope). Eno- plognatha afrodite has a similar body shape, ground color and spotting pattern to (E. ovata + E. latimana + E. margarita) but lacks these synapomorphies. Accordingly, Hippa & Oks- ala considered E. afrodite as the most ances- tral species in the group. More recently, Oxford & Reillo (1994) questioned the phylogeny of the E. ovata group proposed by Hippa & Oksala. Their concern arose because E. ovata, E. latimana, E. penelope and E. afrodite all have European distributions (although the former two have been introduced into North America). All oc- cur in the Mediterranean region; but only E. latimana and E. ovata occur further north, with E. ovata alone extending well into north- ern Europe. Based on this distributional infor- mation, Oxford & Reillo hypothesized a pos- sible Mediterranean origin of the E. ovata group, suggesting that the Asian E. margarita may have been phylogenetically misplaced by Hippa & Oksala. Indeed, the phylogeny pre- sented by Hippa & Oksala was open to criti- cism because of the lack of a suitable out- group for character polarization, few (only nine) characters used, and because there was no quantitative assessment of phylogeny. In the current study we examined four spe- cies in the E. ovata group, and three other species of Enoplognatha: E. japonica Bosen- berg & Strand 1906 from Japan, E. thoracica (Hahn 1833) from England, and E. intrepida Sprensen 1898 from North America. As out- groups in the analysis we used two species of Steatoda: S. grossa (C.L. Koch 1838) and S. bipunctata (Linnaeus 1758). We examined the pattern of sequence evolution in the E. ovata group to ascertain the monophyly of the clade. In this way we can evaluate the hypothesis that the Mediterranean served as the center of origin for the group as suggested by Oxford & Reillo (1994). METHODS Spiders sequenced. — Enoplognatha: E. ovata, two individuals from two localities: Grimes Graves, Norfolk, U.K., collected by G.S. Oxford, June 1991; and Berceto, Italy, collected by G.S. Oxford & RR. Reillo, Au- gust 1991. E. latimana, one individual: Grimes Graves, Norfolk, U.K., collected by G.S. Oxford, June 1991. E. afrodite, one in- dividual: near Carcassonne, S. France, col- lected by S. Peet, July 1988. E. margarita, one individual: Nukabira, Kamishihoro-cho, Hok- kaido, Japan, collected by M. Matsuda, Au- gust 1992. Other Enoplognatha species ex- amined: E. japonica, one individual: Hokkaido, Japan, collected by M. Matsuda, July 1989; E. thoracica, one individual: Flat- ford Mill, Suffolk, U.K., collected by C.J. Smith, May 1978; E. intrepida, one individ- ual: Third Hill Mountain, Berkeley County, West Virginia, USA, collected by P.J. Martin- at. May 1986 (det. D.T Jennings, deposited in Smithsonian, Museum of Natural History). We also extracted DNA from E. penelope, one individual: Sami, Kefallinia, Greece, collected by J. Murphy, May 1987. However, we were not successful in amplifying the product. Out- groups: We used two species of Steatoda as the outgroup: Steatoda grossa: Molokai, Ha- waii, collected by A.-M. Tan & G.S. Oxford October 1993 and S. bipunctata: Yorkshire, U.K., collected by G.S. Oxford, January 1994. Voucher specimens for all species used are at the Center for Conservation Research and Training, University of Hawaii. DNA extraction and sequencing. — DNA samples were prepared by the conventional SDS-NaCl-Ethanol method (Medrano et al. 1990; Tan & Orrego 1992). Tissues from the legs or prosoma were placed in a 1.5 ml tube and ground with a pipette tip. After adding 15 p.1 of proteinase K, the tissues were incubated at 55 °C overnight. Proteins were removed by salt precipitation. DNA was precipitated, washed in alcohol and preserved in IX TE buffer (pH 8.0). For both double and single stranded PCR amplification we used the following primers (Table 1): E and B2 for the less variable re- gion of the 18S sequence; B and P for the more variable region of the 18S sequence; A and B2 for the 16S sequence. PCR amplifi- cation of double-stranded products was per- TAN ET AL.— PARAPHYLY OF ENOPLOGNATHA OVATA GROUP 483 Table 1. — Primers used. Position obtained refers to Drosophila (Clary & Wolstenholme 1985). Gene primer Primer sequence in Drosophila Position obtained # Base pairs Reference 18S E 18S B2 CTGGTTGATCCTGCCAGTAG GCTGGCACCAGACTTGCCCTCC 24-553 529 modified from Hillis & Dixon 1991 18S B 18S P TTCCAGCTCCAATAGCGTAT GTCTTGCGACGGTCCAAGA 606-916 325 W.C. Wheeler & C. Hayashi, pers. comm. 16S A 16S B2 CGCCTGTTTATCAAAAACAT CTCCGGTTTGAACTCAGATCA 12864-13417 450 S.R. Palumbi & T. Hsiao, pers. comm. formed in 12.5 jx! volume with 38 cycles us- ing Thermus aquaticus DNA polymerase (Saiki et al. 1985). Amplification was done with the following profile: 93 °C, 50 °C and 72 °C each for 30 seconds. Single strand prod- ucts were prepared by asymmetric PCR (Gyl- lensten & Erlich 1988) with 1:50 primer ratios in 50 p,l volumes and the same reaction pro- files as above. The products were assessed by mini-gel electrophoresis using 5 |xl aliquots, and washed in sterilized distilled water with three cycles of dialysis using Millipore MC 30 (Amicon Corp.). Dideoxy chain termina- tion sequencing (Sanger et al. 1977) was per- formed using the US Biochemicals Sequenase version 2.0 kit and ^^S labeled dATP. Negative controls were used in all PCR amplifications to make sure the sequences were not from contaminated sources. Sequences were con- firmed by resequencing the same strand from another PCR product. Phylogenetic analysis. — Ribosomal se- quences were initially aligned using the pro- gram SeqEd 1.0.3 (Applied Biosystems 1995), after which alignment of multiple sequences was optimized in CLUSTAL W 1.4 (Higgins & Sharp 1988) in SeqPup 0.6 (Gilbert 1996). The entire first sequence is optimally aligned with the second entire sequence, with mis- matches, gaps and insertions penalized equal- ly, and with an additional gap length penalty for each residue in the insertion. Subsequent detailed alignment was by eye using the sec- ondary structures (Kjer et al. 1994). The 18S sequences were aligned against the secondary structure of Eurypelma californica to match multiple sequences against conserved regions (Hendriks et al. 1988). The 16S sequences were aligned against Drosophila yakuba (Clary & Wolstenholme 1985), using the sec- ondary structure of the region. Sequences were first analyzed using Maximum Likeli- hood (ML) in PHYLIP (version 3.5c, Felsen- stein 1993), using a generalized Jukes & Can- tor (1969) model to allow for unequal base frequencies (Felsenstein 1981) as well as dif- ferent rates of transitions and trans versions. Sequences were also analyzed by Maximum Parsimony (MP) in PAUP (version 3.1.1, Swofford 1993). In both analyses gaps were treated as nrussing data. Bootstrap analyses (Felsenstein 1985) were used to estimate the statistical confidence of the different nodes in the trees. RESULTS The aligned sequences of the 18S region (Fig. 1) and 16S region (Fig. 2) are shown for each species (the two specimens of E. ovata were identical in sequence). Except for E. thoracica (18S only) and E. intrepida (16S only) we obtained 18S and 16S sequence for all species used. The data were first analyzed separately to determine the degree of congru- ence. The 18S sequences showed little bias in base composition, and no evidence for a tran- sition: transversion (TS:TV) bias. The ML tree (using a TS:TV ratio of 1:1) was similar to the MP tree (using a branch- and-bound search) (Fig. 3A): {E. thoracica + E. marga- rita) and {E. latimana + E. ovata) were both discrete clades, and E. japonica fell outside all other species of Enoplognatha. The only difference between the analyses was that E. afrodite was placed with {E. thoracica + E. margarita) in the ML tree, while its position relative to {E. thoracica + E. margarita) and (F. latimana + E. ovata) was unresolved in the MP tree. Constraining E. ovata, E. lati- mana, E. margarita and E. afrodite to be monophyletic increased the length of the MP tree by two steps. We tested the monophyly of E. ovata, E. latimana, E. margarita and E. afrodite by calculating likelihood values (Fel- 484 THE JOURNAL OF ARACHNOLOGY E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda h. Steatoda g. E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda b. Steatoda g. E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda b. Steatoda g. E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda b. Steatoda g. E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda b. Steatoda g. E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda b. Steatoda g. E. ovata E. latimana E. afrodite E. japonica E. margarita E. thoracica Steatoda b. Steatoda g. XXXXAAAGATTAAGCCATGCATGTCTAAGTACATGCCGTATTAAGGCGAAACCGCGAATGGCrCATTAMTCAGTTATGGTTCCTTAGATCGTACCTTACTACTTGGATAACTGTGGCAATTCT xxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx TCTC AA G XXXXXXXXXXXXXXXX G AGAGCTAATACATGCAGCAGAGCTCCGACCTTr*GGGA(XAGCGCTTrTATTAGACCAATACCAATCGGGCCTCGTGTCCGTTC'IGTGTGGTGACTCTGTATAACTTTGGGCTGATCGCACGQG * A. . .A T G A. . .A A G A. . .G T G A. . .A A C A A G TCA CAC A A.T. . .G TCA A A CTCGTCCCGGCGACGTATCTTTCAAGTGTCTGCCTTATCAACTGTCGATGGTATGTTACGCGCCTACCATGGTCGTAACGGGTAACGGGGAATCAGGGTTCGATTCCaSAGAGGGAGCCTGAGA • G. AACGGCTACCACATCCAAGGAAGGCAGCAGGCGCGCAAATTACCCACTCCAGAACGGGGAGGTAGTGACGAAAAATAACAATACGGGACTCT'ITTGAGACCCCGTAATTGGAATGAGTACACTC G G XX G.C C G.C C XXXXXXXX G G TAAATCCTTTAAXX / / / / / XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXTCGTCCTCCTACCGGTGGTTACTGCCCGCGCTGAACGAATCAGCCGGTTTCCTT CX/////XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX xxxx/ / // /XXXXXXXXXTGCGGTTAAAAAGCTCGTAGTTGGATCTCAGTTCCAGCCGGG . G XXXXXXXXXXXXXX / / / / /XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXATCTCAGTTCCAGCTGGG A C T.T A. CAAC XXXXXXXXXX/////XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXX/////XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX CA/ / / / /XXXXXXXXXXXXXXXXXXXXXXXTCGTAGTTGGATCTCAGTTGCAGAGGGACG C T . TTTT *TC TC . XXXXXXXXXX / / / / /TAAAGTTGTTGCGGTTAAAAAGCTCGTAGTTGGATCTCAGTTGCAGAGGGACG C T . TTTT TCA TC . T3ATGATCTTCATCGGTTGTCTTGGGTGACCGGCA'»GTTTACTTTGAAAAAATTAGAGTGCTCAAAGCATACGTGA*CGCATGAATAATGGTGCATGGAATAATGGAATAGGACCTCGGTTCTA xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx T *T C. . .aG.-*-. . .AT. .X C, A G * A A G* A C xxxxxxxxxxxx xxxxxxxxxxxxxxxxxx C A G » TTTTGTTGGTTTTCGGAACACGAGGTAATGATTAAGAGGGACAXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX GACGTGGTCATTCGTACTGCGACGCTAGAGGT3AAATTGTTGG xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx A A xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx A A GCCGGGGGCATTCGTACTGCGACGCTAGAGGTGAAATXXXXXX Figure 1. — Comparison of nuclear 18S ribosomal DNA sequences from 6 species of Enoplognatha and Steatoda bipunctata and S. grossa. Dots represent positions that are identical in sequence to the top sequence; asterisks represent gaps in the sequence required to maximize alignment; crosses indicate no data for a region. The sequence begins at position 24 in Drosophila and ends at position 916. The area marked by ///// indicates the end of the more conserved region of the 18S sequence (position 553 in Drosophila) and the beginning of the more variable region (position 606 in Drosophila). sensteie 1988) for phylogenies that forced these taxa to be monophyletic: PHYLIP was used to perform a statistical test of each of these trees against the one with highest like- lihood. This test uses the mean and variance of log-likelihood differences between trees, taken across sites (Kishino & Hasegawa 1989); trees are considered significantly dif- ferent if their means differ by more than 1 .96 standard deviations. The log likelihood value for the best tree was -1574.9, and was not significantly higher than the value obtained when E. ovata, E. latimana, E. margarita and E. afrodite were constrained to be monophy- letic (log likelihood —1577.8). The 16S sequences show a heavy AT bias, and accordingly most of the changes were A<~>T trans versions. The ML analysis was based on a model which uses the empirical frequencies of the bases observed in the input sequences, and thus accommodates biases in AT richness. Using TS:TV ratios of 1:1 and 2:1 we obtained a tree which was similar to that from MP analysis (Fig. 3B): {E. latimana TAN ET AL.— PARAPHYLY OF ENOPLOGNATHA OVATA GROUP 485 E. ovata E. latlmana XACCTGCTCAATGAAATT*TAAATAGCCGCAGTTATTTCTCTGTGCTMGGTAGCATAATCATTTGTTTrCT*AATrAA^CTAGAAT*GAAAGGTT*AAACATTTTAATT E. afrodite XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX. . .X*. ★ E. japonic a xxxxxxxx * . ATC . ★ E. margarita xxxxxxxx * E. intrepida xxxxxxxxxxxxxxxx. .*..**. ** * + atc ■k Steatoda b. XX. . .*.G .T. .**C. . . G.***. .*. *. . .* A * T . . . . Steatoda g. XXXXXXT.TC. . .*.TACG. .T.C .TAGCTC.A.***. .* E. ovata E. latimana TTTTTATTTTTAAAT*AAAAATTrAAATT*ATATTAAATGTAA*AMT*ACATTrATTTTTTAAAMGACGACAAGACCCTATCGAACTTAACTTTT*GTTTAGCTGGGGCA E. afrodite ....A.*. *A T.TTCT. . . a * a E. japonica A. . . . . .AAGAA. T T a * E. margarita , .A *.TT A E. intrepida ... .A • C. *.CTT.A.C . .ACAAA .... G * t' Steatoda b. *C. .* . . .TT.C*. . .TTA. . Steatoda g. CC CC.T. .A*GCTCC*.A E. ovata E. latimana GCTAATTAATTAA*AA*TTTTAATA**TTAATTrAT*A*ACAAA*T*GATCC*AATAAAATT*GATTAATT*AATCAAGTTACCGTAGGGATAACAGCGTAATAiTCTTTAA E. afrodite . .A.T. . . . . . .G * T * 'tt E. japonica . . .G .A*. ATT TTT A.*..*..*.., . .T. . T * a -T-r E. margarita * r * E. intrepida .A*.CTTA. .TC*.***. . .A.*. , .T * Tr Steatoda b. . .A.T .*T*. . . . . .TAA.*. .T. *G. . . . .CCA.C. * ..... .G * T Steatoda g. . .A . .A.**T.C TTTG.TC*A.*CT.TT.C. .AAA. .C*. , , . CTCAATACCA ... *G ... * r E. ovata E. latimana AAGC*TCTrATTTAAAAGAAAATTTGCGACCTCGATG'ITGAATTAATTTT*CCTATC*TAAT* * ** **GCAATAATTAG*AAAAGXXXXXXXXXXX E. afrodite . . .A* CXXXXXXXXXXXXXXXXXXXX] sfVYVW' ^"VlfVYYVYVVWW E. japonica . . .*T ir fp yyyvyyy' \AAAAAi \AAAAAAaAAAAAA E. margarita .*. .C . 1 • . . . v_ , AAAAAAA^ \AAAAAi ■CXXXXXXXXXXXXX . *.GXXXXXXXXXX VYYYYYYVYW E. intrepida Steatoda b. T. . T******C*AAT *X. . CATT* ^pTV AWVWW Steatoda g. T. ..*... . . .AG. . .TA.GGGA. . . *AT* . . . . G i AAaAAAA AA . . .GTAGTCTGATC Figure 2. — Comparison of mitochondrial 16S ribosomal DNA sequences from six species of Enoplogna- tha, and Steatoda bipunctata and S. grossa. Terminology as in Figure 1. A. B. Figure 3. — Phylogeny of representatives of the genus Enoplognatha based on Maximum Likelihood using: (A.) 18S sequences. All branches are significant based on the approximation of the Likelihood Ratio Test (LRT, indicated by * in PHYLIP, Felsenstein 1993). Parsimony analysis gave three trees with similar topologies, but with less resolution in the consensus: branches that were not supported by bootstrap values > 50% are indicated as dashed lines; for branches that were supported, bootstrap values are given above nodes. Tree length 65, Cl 0.923. (B.) 16S sequences. All branches are significant (approximate LRT, Felsenstein 1993). Parsimony analysis gave a single tree with similar topology (see text). Tree length 230, Cl 0.798. The off-center positions of E. thoracica and E. intrepida indicate only 18S and only 16S sequence data obtained respectively for these two species. 486 THE JOURNAL OF ARACHNOLOGY 100% 83% - E. intrepida - E.japonica _ E. afrodite ~ E. margarita - E. thoracica 100%' E. latimana E. ovata Steatoda grossa Steatoda bipunctata Figure 4. — Phylogeny of representatives of the genus Enoplognatha based on Maximum Likeli- hood using the combined data set of 16S and 18S sequences. All branches are significant (approxi- mate LRT, Felsenstein 1993). Parsimony analysis gave a similar topology but with less resolution: branches that were not supported by Maximum Par- simony are indicated as dashed lines; for branches that were supported, bootstrap values are given above nodes. + E. ovata) and {E. japonica, E. intrepida and E. afrodite) formed discrete clades. The pri- mary difference between the analyses was that E. margarita was placed with {E. japonica, E. intrepida and E. afrodite) on the ML tree, but with {E. latimana + E. ovata) on the MP tree. Constraining E. ovata, E. latimana, E. mar- garita and E. afrodite to be monophyletic in- creased the length of the MP tree by six steps and resulted in a significantly lower log like- lihood value for the ML tree (—1491.8 for the best tree, -1518.1 for the constrained tree). Because the results from the two data sets were largely in agreement the data sets were combined and analyzed together. The resulting ML tree differed from the MP tree only in the degree of resolution it provided (Fig. 4). In all analyses E. ovata fell with E. latimana, E. in- trepida with E. japonica (and in most cases with E. afrodite), E. margarita with E. thor- acica, The E. ovata + E. latimana clade fell outside all others. We concluded that E, ovata, E. latimana, E. margarita and E. afrodite are not monophyletic, and again tested the ro- bustness of these conclusions. Constraining E. ovata, E. latimana, E. margarita and E. af- rodite to be monophyletic increased the length of the MP tree by three steps and gave a sig- nificantly lower log likelihood value for the ML tree (—3244.9 for the best tree, —3290.5 for the constrained tree). DISCUSSION The species E. latimana, E. penelope, E. af- rodite, and E. margarita are similar in gross morphology to the well- studied E. ovata, and this similarity appears to be the basis for grouping these species into what has been considered to be a monophyletic clade (Hippa & Oksala 1983). The phylogenetic analysis presented here based on both the 16S and 18S sequences does not support monophyly of the ovata group” as described by Hippa & Oksala (1983). The E. latimana + E. ovata clade is strong- ly supported, and is consistent with evidence from color polymorphism: E. ovata and E, la- timana share color, regulatory, and black spot- ting polymorphisms (Oxford 1992), although the latter species lacks the ovata color morph. These genetic traits suggest a recent common ancestor for this species pair. Color polymor- phism has never been reported in any other species in the ovata group”. However, the 18S and 16S data sets individually and com- bined consistently place E. afrodite and E. margarita outside the E, latimana + E. ovata clade, more closely associated with E. japon- ica and E. intrepida, and E. thoracica respec- tively. We have no molecular sequence data from E. penelope, and therefore cannot eval- uate its position relative to others in the “F". ovata group”. The results do not refute the Mediterranean center of origin hypothesis of Oxford & Reillo (1994), although the lack of monophyly of the group indicated by the current results de- mands a considerably larger representation from the genus be surveyed before their origin can be identified with any degree of certainty. ACKNOWLEDGMENTS The work reported here was supported in Hawaii by National Science Foundation Grant DEB 9207753 to R.G.G. and G.S.O., and in York by Natural Environment Research Coun- cil Grant GR9/1503 to G.S.O. For allowing destructive use of certain specimens, we would like to thank Jonathan Coddington and Scott Larcher (Smithsonian), Charles Gris- wold (California Academy), Herb Levi (MCZ, Harvard) and Norman Platnick (American Museum of Natural History). We thank M. Matsuda, J. Murphy, and the late C.J. Smith for providing us with specimens. TAN ET AL.— PARAPHYLY OF ENOPLOGNATHA OVATA GROUP 487 LITERATURE CITED Clary, D.O. & D.R. Wolstenholme. 1985. The mi- tochondrial DNA molecule of Drosophila yaku- ba: Nucleotide sequence, gene organization and genetic code. J. Mol. EvoL, 22:252-271. Felsenstein, J. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. EvoL, 17:368-376. Felsenstein, J. 1985. Confidence limits on phylog- enies: An approach using the bootstrap. Evolu- tion, 39:783-791. Felsenstein, J. 1988. Phytogenies from molecular sequences: Inference and reliability. Ann. Rev. Gem, 22:521-565. Felsenstein, J. 1993. PHYLIP. Phylogenetic Infer- encence Package. Version 3.5c. University of Washington, Seattle. Gilbert, D.G. 1996. SeqPup, version 0.6. , Biology Dept., Indiana University, Bloomington, Indiana 47405. Gyllensten, V. & H. Erlich. 1988. Generation of single- stranded DNA by the polymerase chain reaction and its applications to direct sequencing of the HLA DQa locus. Proc. Nat. Acad. Sci. USA, 85:7652-7656. Haldane, J.B.S. 1939. The theory of the evolution of dominance. J. Genet., 37:365-374. Hendriks, L., C. Van Broeckhoven, A. Vandenber- ghe, Y. Van De Peer & R. De Wachter. 1988. Primary and secondary structure of the 18S ri- bosomal RNA of the bird spider Eurypelma cal- ifornica and evolutionary relationships among eukaryotic phyla. European J. Biochem., 177: 15-20. Higgins, D.G., & P.M. Sharp. 1988. CLUSTAL: A package for performing multiple sequence align- ment on a microcomputer. Gene, 73:237-244. Hillis, D.M. & M.T. Dixon. 1991. Ribosomal DNA: Molecular evolution and phylogenetic in- ference. Quart. Rev. Biol., 66:411-453. Hippa, H. & I. Oksala. 1979. Colour polymor- phism of Enoplognatha ovata (Clerck) (Araneae, Theridiidae) in western Europe. Hereditas, 90: 203-212. Hippa, H. & I. Oksala. 1981. Polymorphism and reproductive strategies of Enoplognatha ovata (Clerck) (Araneae, Theridiidae) in northern Eu- rope. Ann. Zool. Fennici, 18:179-190. Hippa, H. & I. Oksala. 1982. Definition and revi- sion of the Enoplognatha ovata (Clerck) group (Araneae: Theridiidae). Entomol. Scandinavica, 13:213-222. Hippa, H. & I. Oksala. 1983. Cladogenesis of the Enoplognatha ovata group (Araneae, Theridi- idae), with description of a new Mediterranean species. Ann. Ent. Fennici, 49:71-74. Jukes, T.H. & C.R. Cantor. 1969. Evolution of pro- tein molecules. Pp. 21-132, In Mammalian Pro- tein Metabolism. (H.N. Munro, ed.). Academic Press, New York. 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. J. Mol. EvoL, 29:170-179. Kjer, K.M., G.D. Baldridge & A.M. Fallon. 1994. Mosquito large subunit ribosomal RNA: Simul- taneous alignment of primary and secondary structure. Biochim. Biophys. Acta, 1217:147- 155. Levi, H.W. 1962. The spider genera Steatoda and Enoplognatha in America (Araneae, Theridi- idae). Psyche, 69:11-36. Levy, G. & P. Amitai. 1981. The spider genus En- oplognatha (Araneae: Theridiidae) in Israel. Zool. J. Linn. Soc., 72:43-67. Locket, G.H. & A.F. Millidge. 1951. British Spi- ders, Vol. 1. Ray Soc., London. Medrano, J.F., E. Aasen & L. Sharrow. 1990. DNA extraction from nucleated red blood cells. Bio- techniques, 8:43. Oxford, G.S. 1976. The colour polymorphism in Enoplognatha ovatum (Clerck) (Araneae: Theri- diidae)— temporal stability and spatial variabili- ty. Heredity, 36:369-381. Oxford, G.S. 1983. Genetics of colour and its reg- ulation during development in the spider Eno- plognatha ovatum (Clerck) (Araneae: Theridi- idae). Heredity, 51:621-634. Oxford, G.S. 1985. Geographical distribution of phenotypes regulating pigmentation in the spider Enoplognatha ovata (Clerck) (Araneae: Theridi- idae). Heredity, 55:37-45. Oxford, G.S. 1989. Genetics and distribution of black spotting in Enoplognatha ovata (Araneae: Theridiidae), and the role of intermittent drift in population differentiation. Biol. J. Linn. Soc., 36: 111-128. Oxford, G.S. 1991. Visible morph-frequency var- iation in allopatric and sympatric populations of two species of Enoplognatha (Araneae: Theri- diidae). Heredity, 67:317-324. Oxford, G.S. 1992. Enoplognatha ovata and E. la- timana: a comparison of their phenologies and genetics in Norfolk populations. Bull. British Ar- achnol. Soc., 9:13-18. Oxford, G.S. & P.R. Reillo. 1993. Trans-continen- tal visible morph-frequency variation at homol- ogous loci in two species of spider, Enoplogna- tha ovata s.s. & E. latimana. Biol. J. Linn. Soc., 50:235-253. Oxford, G.S. & RR. Reillo. 1994. The world dis- tributions of species within the Enoplognatha ovata group (Araneae: Theridiidae): Implications for their evolution and for previous research. Bull. British Arachnol. Soc., 9:226-232. Reillo, P.R. & D.H. Wise. 1988a. An experimental 488 THE JOURNAL OF ARACHNOLOGY evaluation of selection on color morphs of the polymorphic spider Enoplognatha ovata (Ara= neae: Theridiidae). Evolution, 42:1172-1189, Reillo, RR. & D.H. Wise. 1988b. Genetics of color expression in the spider Enoplognatha ovata (Araneae: Theridiidae) from coastal Maine. American Midi, Nat., 119:318-326. Saiki, R.K., S, Scharf, F. Faloona, K.B. Mullis, G.T Horn, H.A. Erlich & N. Amheim. 1985. Enzy= matic amplification of B-globingenomic se- quences and restriction site analysis for diagnosis of sickle cell anemia. Science, 230:1350-1354. Sanger, F., S. Nicklen & A.R. Coulsen. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA, 74:5463-5467. Swofford, D.L. 1993. PAUP: Phylogenetic analysis using parsimony, version 3.1.1. Smithsonian In- stitution, Washington, DC. Tan, A.-M. & C. Onego. 1992. DNA amplification from museum collections of extracts originally intended for allozyme analysis. Mol. EcoL, 1:95- 97. Manuscript received 10 February 1996, revised 1 October 1998. 1999. The Journal of Arachnology 27:489-496 CHEAP TRANSPORT FOR FISHING SPIDERS (ARANEAE, PISAURIDAE): THE PHYSICS OF SAILING ON THE WATER SURFACE Robert B. Suter: Department of Biology, Vassar College, Poughkeepsie, New York 12604 USA ABSTRACT. Many pisaurid spiders inhabit the edges of bodies of fresh water and actively propel themselves across the water surface using both rowing and galloping gaits. They also sail across the water, taking advantage of the wind and their nearly frictionless interaction with the water surface. The physical interactions of Dolomedes triton (Walckenaer 1837) (Araneae, Pisauridae) with moving air, in a wind tunnel in which the floor was water, formed the core of the present investigation. Spiders in an elevated (sailing) posture were subjected to greater drag forces attributable to air motion than were spiders in a prone (non-sailing) posture and therefore were transported substantially faster than prone spiders. In the context of transport velocity, the benefit of adopting an elevated posture was substantially greater (relative to mass) for small spiders than for large ones, although even under the relatively steady flow conditions of the wind tunnel the velocities of the small spiders in the elevated posture were more variable than either small prone spiders or large spiders. The efficacy of adopting an elevated posture was a consequence of the steep air velocity gradient that existed above the surface of the water in the wind tunnel and that also exists above any pond over which the air is moving. Taken as a whole, the data indicate that sailing is a remarkably cheap form of transportation for Dolomedes, but that, at least at the edges of large bodies of water, it involves risks because it is directionally uncontrolled. Locomotion by spiders includes ordinary terrestrial modes such as walking, running, and jumping, and quite unusual modes of air- borne and aquatic locomotion. Although the eight-legged stepping patterns of spiders on land obviously differ in detail from those of insects (Cocatre-Zilgien & Delcomyn 1993), the biomechanics of terrestrial locomotion by insects and by spiders probably differ little be- cause they share size (Price 1984; Pennycuick 1992), exoskeletal architecture, and important aspects of the nervous system (Osorio et al. 1995, 1997). Consequently, the rich literature on the biomechanics of terrestrial locomotion in insects (e.g.. Full & Tu 1990; Full et al. 1995) contributes substantially to our under- standing of terrestrial locomotion in spiders. The same cannot be said of the biomechanics of aerial dispersal via ballooning and of lo- comotion on the water surface, forms of spi- der locomotion that are shared by only a few insects (ballooning: McManus & Mason 1983; Cox & Potter 1986; aquatic locomotion: Andersen 1976). For these unusual modes of locomotion, most of our knowledge of the physics and biomechanics comes from the lit- erature on arachnids (ballooning: Humphrey 1987; Suter 1991, 1992; aquatic locomotion: Suter et al. 1997; Suter 1999a). Fishing spiders, Dolomedes triton (Walck- enaer 1837) (Pisauridae), the subjects of this paper, actively propel themselves across the water surface using two distinct gaits: rowing propels the spiders horizontally at velocities < 0.27 m/sec (McAlister 1959; Shultz 1987; Suter et al. 1997) and galloping, used by these spiders during some kinds of prey-capture (Gorb & Barth 1994) and during escape from predators (Suter unpubl. data), propels the spi- ders at horizontal velocities up to 0.75 m/sec (Suter 1999a). In both rowing and galloping, the spider accelerates forward by rapidly mov- ing its propulsive legs backwards, transferring momentum to the water through the genera- tion of drag. These active aquatic gaits are dis- tinct from the alternating tetrapod locomotion used by the spiders on solid substrate (Barnes & Barth 1991; Shultz 1987). Fishing spiders also move across the water surface propelled by air movements. Deshefy (1981) reported on one distinctive form of this “sailing” in which the spider extends and el- 489 490 THE JOURNAL OF ARACHNOLOGY evates its most anterior pair of legs, taking advantage of the increased wind speed 2-3 cm above the water surface. I have observed a second distinctive form of sailing in pisaurid spiders, in which the spider extends and de- presses all of its legs, thereby raising its body well above the water surface and allowing the body and proximal leg segments to interact with more rapidly moving air currents. What follows are analyses of (1) the wind velocity gradient in the boundary layer above a pond’s surface, (2) the drag forces acting on elevated vs. prone spiders, and (3) the velocity changes that result from modification of posture during sailing. METHODS Pond measurements.— I used a hot-wire anemometer (Thermonetics Corporation mod- el HWA-103) to measure wind speed just above the surface of a pond at The Rockefel- ler University Field Research Center, Mill- brook, Dutchess County, New York. At the time and location of the data collection, the water surface was upwind of the pond’s usual shore but was separated from the shore by about 4 m of mud flat. I arranged the ane- mometer assembly (below) so that its sensor pointed upwind (away from the shore) and was over the water 0.5 m from the edge of the mud flat. As a result of the location and orientation of the sensor, it measured the speed of air that had traveled at least 120 m across the pond’s surface unimpeded by struc- tures other than the surface of the water itself. The height of the sensor above the water sur- face was controlled by a motorized cam which, as it rotated, raised and lowered the 0.5 m boom to which the sensor was attached. The resulting motion of the sensor tip was vertical (in space) and sinusoidal (in time), with a period of 4.3 sec, an excursion from 0.5 to 8.9 cm, and a maximum velocity of 0.06 m/sec (a small fraction of the recorded air velocities). Analog signals from the anemometer were digitized at 10 Hz by an analog-to-digital (A/ D) converter (Vernier Software Co., model ULI 5.0) under the control of data logging software (Vernier Software Co., Logger 3.04) running on a laptop computer (Apple Corpo- ration, model PowerBook 5300c). Power for the computer was supplied by its inboard bat- tery and power for the motorized cam and A/ D converter was supplied by a 12 V lead- acid battery. The analysis of air speed as a function of distance to the water surface was complicated by the high variability in wind speed, presum- ably due to turbulence, at any single height. Although the mainstream velocity (sensu Den- ny 1993) was unknown for the data collected at the pond, the air’s velocity in the boundary layer above the water must decrease to zero as height approaches zero (Denny 1993). Ac- cordingly, I used logarithmic curve fits to characterize the relationship between height and velocity both for the pond data and for the wind tunnel data (below). Spiders. — The adult Dolomedes triton (Ar- aneae, Pisauridae) used in these experiments were collected from small ponds in Mississip- pi, and the juvenile was the progeny of one of the field-caught females. All were held in my laboratory under conditions described elsewhere (Suter et al. 1997). In the experiments described below, I in- vestigated the motion of, and the forces acting upon, killed and dried spiders of two sizes. A third-instar juvenile that had been in the lab since hatching (wet mass 0.013 g, 0.126 mN) and two adult males of approximately the same size (0.186 g, 1.82 mN; 0.243 g, 2.38 mN) were anaesthetized with CO2 and killed by freezing. After post-mortem thawing, the spiders were immobilized in a prone posture (the juvenile and one adult) or an elevated posture (the second adult) and allowed to air dry for several weeks. During the weeks of experimentation, the postures of the adult spi- ders (Fig. 1) remained unchanged. During measurements of horizontal velocity, however, the posture of the juvenile was changed from prone to elevated to make possible direct com- parisons of the same spider in two postures: to accomplish the posture change, I softened the spider’s most proximal leg joints by moist- ening them, and then repositioned the limbs and air dried the spider for several days. The weights of the dried adult spiders (el- evated, 0.736 mN; prone, 0.959 mN) were matched more closely by fastening with epoxy a small, flat coil of nichrome wire (0.221 mN) to the dorsal surface of the cephalothorax of the spider in the elevated posture. The adult spiders then both had weights of 0.96 mN. Temporary weight modifications were accom- plished by hanging the same short length of SUTER—PHYSICS OF SAILING IN DOLOMEDES 491 Figure 1. — Adult spider postures as digitized from photographs of the large dried spiders used in the experiments reported here. For comparison, the small spider in the prone and elevated postures had heights of 2.1 mm and 4.4 mm respectively (filled bars at left). A spider in the prone posture (left) is exposed to lower horizontal air velocities than is a spider in the elevated posture (right). The three curves are the same as those shown in Figure 5, but with the axes reversed. 30-gauge copper wire over the cephalothorax of each of the adult spiders. All experiments with the dried spiders were conducted at lab- oratory temperatures between 20-23 °C, Wind tunnel measurements.— The hori- zontal wind tunnel used in this study had an experimental chamber measuring 20 X 20 X 87 cm (length). The floor of the chamber had a 0.6 cm deep cavity beginning 36 cm down- wind from the air inlet and having horizontal dimensions of 18 X 40 cm. When the cavity was filled with water, the surface of the water and the surface of the remainder of the floor of the experimental chamber formed an un- broken, flat surface. Both ends of the experi- m.ental chamber were fitted with 2 cm thick furnace filter fiber to suppress turbulence. With the exception of the filter, the upwind end of the tunnel was open to room air. Ten cm beyond the filter at the downwind end of the tunnel the air entered a 10.1 cm (diameter) polyvinyl chloride (PVC) pipe on the end of which was mounted a small fan oriented so that it pulled air through the tunnel. Although valving in the pipe leading to the fan allowed me to control air speed in the experimental chamber, I conducted all tests at a nominal air speed of 0.65 m/sec (measured at the center of the air stream, 52 cm from the upwind end of the experimental chamber). I used a hot-wire anemometer (Thermonet- ics Corporation, model ITWA-103) to monitor air velocity in the chamber and to measure the airspeed profile as a function of the distance from the water surface. The 5 mm (diameter) anemometer probe was inserted through a 6 mm (diameter) hole in the top of the experi- mental chamber, 52 cm from its upwind end. A micromanipulator, mounted on the outside of the chamber, facilitated adjustment of the position of the anemometer’s sensor relative to the water surface. At the beginning of a sailing trial, a dried spider was placed gently on the water surface approximately 42 cm from the upwind end of the chamber (6 cm from the upwind edge of the water surface) and at the side-to-side cen- ter of the water surface. The spider was held at that location by a pair of eichrome wires assembled in an inverted “V” and attached to a probe that could be raised several cm, re- leasing the spider. The spider was released only after the tunnel fan was turned on and the chamber had reached a constant nominal velocity of 0.65 m/sec. To measure the sailing velocity of a dried spider, I recorded its lo- cation in the horizontal plane, beginning when the spider crossed a line 2.5 cm downwind from its release site, using an SVHS video camera (Panasonic model AG-455) and Image (National Institutes of Health software, ver- sion 1.55 f) as a frame grabber and image dig- itizer (at a rate of 20 frames per second). I digitized the location of the center of the spi- der’s cephalothorax using tools resident in Im- age, and then calculated velocity as the dis- tance moved divided by the frame interval. I used a horizontal balance (Suter et al. 1997) to measure horizontal drag forces on the adult dried spiders in the wind tunnel. The balance employed an electronic clinometer (Applied Geomechanics Inc. model 900 Bi- axial Clinometer) with a resolution of 0.01° (1.75 X 10 rad) to measure small angular displacements that were directly proportional to the horizontal force applied to the spider. 492 THE JOURNAL OF ARACHNOLOGY Time (s) Figure 2. — The horizontal wind speed (solid line) within the 10 cm deep layer of air just above the surface of a pond varied significantly with height of the sensor (dashed line) in each case although the effect was small (0.04 < < 0.12, P < 0.05). The residual variability is assumed to be a conse- quence of the turbulence in the air. The end of the vertical arm of the balance consisted of a pair of nichrome wires assem- bled in an inverted “V” which immobilized the spiders without applying any vertical force to them. Clinometer output (in volts) was dig- itized by an A/D converter (National Instru- ments Corporation, model NB-MIO-16L) driven by a Lab View 3 program (National In- struments) on an Apple microcomputer (Pow- er Macintosh 7100/80AV). RESULTS Pond and wind tunnel air movement, — Velocity measurements of the air within 10 cm of the surface of the pond (Vp) indicated that the air was turbulent, with large fluctua- tions in velocity at several scales (Fig. 2). Only a small amount of the variability in Vp could be attributed to the changes in the po- sition of the anemometer’s sensor (0.04 < < 0.12), although the small effect of the Figure 3. — During a period of relatively constant average wind velocity (23 sec to 50 sec in top graph of Figure 2), velocity of the air just above the pond’s surface (A, Up) varied approximately as the logio of height (B, assuming, as is required by boundary layer physics, that Up at the surface must be zero; Up = 0.16 log,o height + 0.83, r’ = 0.225, n = 267, P = 0.0001). height of the sensor was significant {P < 0.05 in each sample in Fig. 2). For a subsample of velocity data in which large-scale fluctuations in Up were relatively small, the influence of sensor height was more prominent and Up var- ied approximately as the logio of height (Fig. 3). The logarithmic relationship between Up and height can be seen most clearly during short segments of the pond data (Fig. 2) which include only a half cycle or a full cycle of the sensor and therefore do not conflate velocities that are widely separated in time (Fig. 4), and is entirely in accord with boundary layer the- ory (Schlichting 1979). In the wind tunnel, constant fan velocity and suppression of some of the turbulence made the influence of height on wind tunnel velocity (U^) much more detectable: logarith- mic curve fits on three data sets indicated that sensor height explained more than 90% of the variation in Up (Fig. 5; 0.91 < r- < 0.97). Sailing velocity and drag forces in the wind tunnel. — Images of dried spiders sailing downwind in the wind tunnel indicated that, SUTER— PHYSICS OF SAILING IN DOLOMEDES 493 Height (cm) Figure 4.— The relationship between Vp and height above the water is most clearly visible during the brief periods of time (from Fig. 2) which include one full cycle of the sensor and therefore do not conflate velocities that are widely separated in time. In the examples shown here, logjo curve fits worked well: 0.86 < ri < 0.98. by the time each had reached the line where video digitizing began, it was moving at rel- atively constant velocity: thus it had reached the terminal velocity (V^) at which the forces propelling it (air-induced drag) were in bal- ance with the forces resisting its motion (wa- ter-induced drag). Under wind tunnel condi- tions, a large spider (0.42 g, wet weight) dried in an elevated posture always had greater ve- locities (Vt) than a similar-sized spider that had been dried in a prone posture. At the weights at which the effect of the two postures could be compared directly, being elevated conferred about a two-fold velocity advantage over being prone, but a much greater (3.7- foM) advantage accrued to the elevated form of the very small spider (0.013 g, wet weight) (Fig. 6). A stepwise multiple regression of ve- locity on height (a function of posture) and weight for the large spider yielded a highly significant relationship (P < 0.01, F ~ 106.96, adjusted F = 0.876) in which velocity varied directly with height and inversely with weight. For the small spider, the difference be- tween V^ for elevated and prone postures was highly significant (Mann- Whitney U test, Z ~ “2.646, P = 0.008) and the variability in Vj for the elevated small spider was much greater than that for any other group (Fig. 6), The very high variability in for the elevated small spider is probably attributable to small eddies in the air stream close to the water sur- face that can influence very small objects but are averaged out when interacting with the much greater leg span of the larger spiders. To test for effects of horizontal orientation relative to the direction of air movement in the wind tunnel, I released dried large spiders at different horizontal orientations, with the expectation that any relationship would be ap- proximately sinusoidal. The orientation of the large dried spiders did not have a significant influence on Vj for either the elevated or the prone spiders (Fig. 7). In direct measurements of the drag force exerted on spiders by moving air, drag on a spider in the elevated posture was significant- ly higher than drag on the prone spider (0.055 ± 0.002 mN vs. 0.0'29 ± 0.002 mN; t = 31.4, P < 0.0001) as expected from the results of dynamic tests (Fig. 6). The L92-fold differ- ence between the mean values fits well with 494 THE JOURNAL OF ARACHNOLOGY Height (mm) m E >. >4— if ■q _o 0 > 0 •«—> C o N O X Figure 5. — In the wind tunnel, air velocity (V^) varied with the logio of height above the surface of the water. For the three runs shown here, the equa- tions were = 0.36 logjo height + 0.28, = 0.987, = 0.39 logio height + 0.28, > 0.99, and = 0.28 log,o height + 0.39, r2 = 0.920. the 2.12“fold differences between measure- ments of Vi for the identical elevated and prone spiders at their lightest weight (Fig. 6). DISCUSSION A hshing spider, lying prone on the surface of a pond and not anchored to floating vege- tation or debris, will move passively across the water at a rate influenced by the spider’s mass and the velocity of wind over the pond (Fig. 6). The actual air velocity to which the spider is exposed, however, is strongly influ- enced by the location of the spider’s body parts relative to the water surface (Figs. 4, 5). This connection between elevation and air ve- locity means that the velocity at which the spi- der can travel under the influence of air move- ments is closely tied to the spider’s posture (Figs. 1, 6). Quasi-passive locomotion, like balloon- ing.— Although the propulsive forces in- volved in sailing are environmental rather than physiological, the importance of posture and the stereotyped performance of the pos- tures employed in sailing (above and Deshefy 1981) indicate that this form of locomotion is not purely passive. On the other hand, because wind direction is not controlled by the spider and because the keel-less spider apparently has no control of its own direction relative to the wind (as do humans in sailboats), the sail- ing spider cannot influence its destination. In that regard, this quasi-passive form of loco- Vertical Force (mN) Figure 6. — During wind tunnel measurements of the horizontal terminal velocities (U) of sailing spi- ders, U varied with spider size as well as with mass and posture. For the large spiders (wet weight = 1.82 and 2,38 mN), the elevated posture (•) resulted in a doubling of U relative to U for the prone pos- ture (o) (dry weight = 0.96 mN - elevated: 0.088 ± 0.007 m/sec vs. prone: 0.042 ± 0.006 m/sec, ra- tio = 2.12; dry weight = 1.42 mN - elevated: 0.062 ± 0.004 m/sec vs. prone: 0.024 ± 0.005 m/sec, ra- tio = 2.51). For the small spider (wet weight = 0.013 g), the elevated posture (A) resulted in a 3.7- fold increase in U (based on comparison of medi- ans, designated with dashes) relative to U for the prone posture (+). See text for statistical analyses. Exponential curves (because V, must approach zero as vertical force becomes very large) fitted to the data for the large spiders were 0.187 F70242 vated, r- - 0.873) and 0.104 (prone, F = 0.766) where F^ is the vertical force. The velocities shown here, for all but the heaviest prone spider, overestimate U achievable during sailing by live spiders because the dried spiders were lighter and therefore created shallower dimples and less water- generated drag (Suter 1997). motion resembles ballooning in which, once airborne, the spider has no control over its horizontal direction and therefore little control over its destination, and in which posture is crucial (Suter 1992). Two other parallels between sailing and bal- looning are worth noting: neither requires sub- stantial muscular input, and in both forms of locomotion, smaller individuals have substan- tial advantages. The energetic cost of whole- body sailing (cf. Deshefy 1981) can be esti- mated as the work needed for the spider to raise its body to an effective sailing height — for the larger spiders used as models in this SUTER— PHYSICS OF SAILING IN DOLOMEDES 495 Figure 7.-— Sinusoidal curve fits on the velocities of large spiders in elevated (upper) and prone (low- er) postures, after release at different horizontal ori- entations relative to the direction of air motion in the wind tunnel, revealed that orientation had no significant effect on Vj. Each datum consists of mean ± SD velocities and angles (relative to 0°, facing directly upwind) for one spider's motion fol- lowing a single release: variation in V^ reflects mea- sured changes in velocity between digitized frames from videotaped records; variation in angle is a consequence of the slow rotation of the released spider during its movement downwind. For the el- evated posture, Ft == 0.005 sin (angle) + 0.094 (ri = 0.119, n ~ 19, P > 0.05), and for the prone posture. Ft = 0.013 sin (angle) + 0.054 (P - 0.144, « - 17, F > 0.05). study (live weight — 2.1 mN), the work re- quired to raise the center of mass from 4-12 mm above the water is about L68 X 10“^ joules, an amount of work that needs to be done only once during a sailing episode. In comparison, for the smaller of the spiders used as models in this study (live weight = 0.126 mN), the cost of elevating the center of mass 23 mm is about 2.90 X 10“^ joules. The smaller spider's velocity in the elevated pos- ture is approximately the same as that of the larger spider (Fig. 6), but the cost of attaining that posture for the larger spider is 58 times as great. Thus, the efficiency of sailing is far greater for very small spiders, but for any spi- der the cost is very small: consumption of a single fruit fly (Golley 1961) would provide sufficient energy to elevate the larger spider thousands of times! Sailing is, like ballooning (Suter 1999b), a remarkably cheap form of transport. Risk assessment.— Fishing spiders are at- tacked from below by fish (G. Miller pers, commun.) and are likely also to suffer pre- dation from above by aeurans and birds. Be- cause their predators’ feature detectors un- doubtedly respond to specific cues (e.g., shape, size, motion: Lettvin et al. 1959; Ewert 1974; Ewert et al. 1983) the suppression of any of these cues can result in a reduction in the probability of eliciting predation. In this context, sailing can be viewed as an incon- spicuous form of locomotion that offers a rel- ative reduction in predation risk through the suppression of visual cues: propulsive mo- tions of the legs relative to the body are ab- sent, and surface waves caused by rowing and galloping locomotion (Suter et al. 1997; Suter 1999a) are not generated at the velocities achieved during sailing (Fig. 6; Denny 1993; Vogel 1994). Another kind of risk, that associated with motion whose direction is not controlled by the spider, should rise with elevated uncer- tainty about the ecological suitability of the destination. In the ponds usually inhabited by D. triton, this risk is minimal because the ul- timate destination of a sailing spider will al- ways be an edge of the pond, a location that may be unfamiliar to the spider but that is apt to be as ecologically suitable as the spider’s original location. In contrast, species of Do- lomedes that inhabit the shores of islands in the Great Lakes of North America cannot be assured of a benign destination if the wind is offshore: sailing away from a shore can take the spider into open waters where both food and cover are unavailable and where the di- rection of the nearest shoreline is undetect- able. Function.— “For fishing spiders living at the edges of ponds, sailing is both energetically cheap and relatively safe (above). But this form of locomotion has only rarely been ob- served in nature and its function in the context of more controlled modes of locomotion on the water surface (Suter et al. 1997; Suter 1999a) remains unclear. Because sailing is most efficient for smaller spiders (Fig. 6) and because heavy D. triton are unable to elevate their bodies into rapidly moving air without exceeding the ability of the water’s surface tension to support them (Suter 1999a), the pri- mary function of sailing may be to facilitate 496 THE JOURNAL OF ARACHNOLOGY dispersal in young spiders, a hypothesis that is yet to be tested. ACKNOWLEDGMENTS I thank Fernando Nottebohm for permission to use ponds at The Rockefeller University Field Research Center, Gary Lovett at the In- stitute of Ecosystem Studies for educating me about the dynamics of turbulent atmospheres, and Edgar Leighton at the University of Mis- sissippi for providing the spiders. The study was supported in part by funds provided by Vassar College through the Class of ’42 Fac- ulty Research Fund. LITERATURE CITED Andersen, N.M. 1976. A comparative study of lo- comotion on the water surface in semiaquatic bugs (Insects, Hemiptera, Gerromorpha). Densk. Meddr. Dansk Naturh. Foren., 139:337-396. Barnes, W.J.R & EG. Barth. 1991. Sensory control of locomotor mode in semi-aquatic spiders. In Locomotor Neural Mechanisms in Arthropods and Vertebrates. (D.M. Armstrong & B.M.H. Bush, eds.) Manchester Press, Manchester, 340 pp. Cocatre-Zilgien, J.H. & F. Delcomyn. 1993. A new method for depicting animal step patterns. Anim. Behav., 45:820-824. Cox, D.L. & D.A. Potter. 1986. Aerial dispersal behavior of larval bag worms, Thyridopteryx ephemeraeformis (Lepidoptera: Psychidae). Ca- nadian EntomoL, 118:525-536. Denny, M.W. 1993. Air and Water: The Biology and Physics of Life’s Media. Princeton Univ. Press, Princeton, 341 pp. Deshefy, G.S. 1981. ‘Sailing’ behaviour in the fishing spider, Dolomedes triton (Walckenaer). Anim. Behav., 29:965-966. Ewert, J.-P. 1974. The neural basis of visually guided behavior. Sci. American, 230:34-42. Ewert, J.-R, R.R. Capranica & D.J. Ingle (eds.). 1983. Advances in Vertebrate Neuroethology. Plenum Press, New York, 1238 pp. Full, R.J. & M.S. Tu. 1990. Mechanics of a rapid running insect: two-, four- and six-legged loco- motion. J. Exp. Biol., 156:215-231. Full, R.J., A. Yamauchi & D.L. Jindrich. 1995. Maximum single leg force production: Cock- roaches righting on photoelastic gelatin. J. Exp. Biol., 198:2441-2452. Golley, FB. 1961. Energy values of ecological ma- terials. Ecology, 42:581-584. Gorb, S.N. & EG. Barth. 1994. Locomotor behav- ior during prey-capture of a fishing spider, Do- lomedes plantarius (Araneae: Araneidae): Gal- loping and stopping. J. ArachnoL, 22:89-93. Humphrey, J.A.C. 1987. Fluid mechanical con- straints on spider ballooning. Oecologia, 73:469- 477. Lettvin, J.Y, H.R. Maturana, W.S. McCulloch & W.H. Pitts. 1959. What the frog’s eye tells the frog’s brain. Proc. Instit. Radio Engrs., 47:1940- 1951. McAlister, W.H. 1959. The diving and surface- walking behaviour of Dolomedes triton sexpunc- tatus (Araneida; Pisauridae). Anim. Behav., 8: 109-111. McManus, M.L. & C.J. Mason. 1983. Determina- tion of the settling velocity and its significance to larval dispersal of the gypsy moth (Lepidop- tera: Lymantridae). Environ. EntomoL, 12:270- 272. Osorio, D., M. Averof & J.P. Bacon. 1995. Arthro- pod evolution: Great brains, beautiful bodies. Trends Ecol. Evol., 10:449-454. Osorio, D., J.P. Bacon & P.M. Whitington. 1997. The evolution of arthropod nervous systems. American Sci., 85:244-253. Pennycuick, C.J. 1992. Newton Rules Biology: A Physical Approach to Biological Problems. Ox- ford Univ. Press, Oxford. Ill pp. Price, RW 1984. Insect Ecology (2nd ed.). Wiley- Interscience, New York. 607 pp. Schlichting, H.T 1979. Boundary Layer Theory (7th ed.). McGraw-Hill, New York. Shultz, J.W 1987. Walking and surface film loco- motion in terrestrial and semi-aquatic spiders. J. Exp. Biol., 128:427-444. Suter, R.B. 1991. Ballooning in spiders: Results of wind tunnel experiments. Ethol. Ecol. & EvoL, 3:13-25. Suter, R.B. 1992. Ballooning: Data from spiders in freefall indicate the importance of posture. J. A- rachnol., 20:107-113. Suter, R.B. 1999a. Walking on water. American Sci., 87:154-159. Suter, R. B. 1999b. An aerial lottery: The physics of ballooning in a chaotic atmosphere. J. Arach- noL, 27(l):281-293. Suter, R.B., O. Rosenberg, S. Loeb, H. Wildman & J.H. Long, Jr. 1997. Locomotion on the water surface: Propulsive mechanisms of the fisher spi- der Dolomedes triton. J. Exp. Biol. 200:2523- 2538. Vogel, S. 1994. Life in Moving Fluids (2nd ed.). Princeton: Princeton Univ. Press. 467 pp. Manuscript received 8 June 1998, revised 11 Sep- tember 1998. 1999. The Journal of Arachnology 27:497-502 NOTES ON THE SOCIAL STRUCTURE, LIFE CYCLE, AND BEHAVIOR OF ANELOSIMUS RUPUNUNI Leticia Aviles: Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona 85721 USA Patricio Salazar: Departamento de Biologia, Pontificia Universidad Catolica del Ecuador, Quito, Ecuador ABSTRACT^ Observations on the colony structure, life cycle, and behavior of Anelosimus rupununi in eastern Ecuador point to a level of social organization similar to that of Anelosimus eximus and Anelosimus domingo, confirming its status as a non-territorial, permanently-social species. Anelosimus rupununi colony members were seen to cooperate in prey capture and transport, to feed communally, and to take turns in tending the egg sacs. Sex ratios were also highly female-biased. There were, however, some interesting differences with these other species. Anelosimus rupununi egg sacs were grouped as part of maternal care efforts, with grouped sacs being more likely to be tended than ungrouped sacs. Males and females ap- parently matured at the same instar, males appeared shorter-lived than females, and individuals within the nests were clearly synchronized with each other in the stage of their life cycle. Also, as would be expected from its notably smaller body size, A. rupununi'^ life cycle appeared shorter than that of A. eximius. Understanding the evolution of animal so- cial systems often requires exploring within a comparative framework both the environmen- tal conditions that may have selected for so- cial living and the suite of traits that in par- ticular phylogenetic lineages may have facilitated or hindered the transition from one level of social organization to another (Crespi & Choe 1997). Spiders appear ideal for this exploration because they have given rise to several independent derivations of complex social behavior involving cooperation in nest building, prey capture, feeding, and brood care (for a recent review, see Aviles 1997). Additionally, the genera that contain these co- operative species— also known as “non-terri- torial permanent social” or “quasisocial” — contain species with other levels of social organization. The genus Anelosimus Simon 1891 (Araneae, Theridiidae) in America, in particular, includes at least four non-territorial permanent-social species, which are mostly tropical, and several periodic-social or solitary species that inhabit both tropical and temper- ate areas of the New World (Levi 1956, 1963, 1972; but see Furey 1998). Among these, only the permanent-social Anelosimus eximius Keyserling 1884 has been relatively well stud- ied (see references cited in Aviles 1997). Oth- er species have received comparatively little attention (see Brach 1977; Fowler & Levi 1979; Nentwig & Christenson 1986; Smith 1987; Rypstra & Tirey 1989; Aviles & Mad- dison 1991; Furey 1998; Aviles & Gelsey 1998). Here we present observations on the col- ony structure, life cycle and behavior of Ane- losimus rupununi Levi 1956 in eastern Ec- uador. This species had been previously reported from Trinidad, British Guiana, northwestern Peru, Brazil and Paraguay (Levi 1963), although the Paraguayan specimens apparently correspond to misidentified A. lor- enzo Levi 1979 (Fowler & Levi 1979). No previous records existed from Ecuador where this study was conducted. The only published information on A. rupununi was a photograph of a nest (Levi 1972) and the suggestion that the species is “probably quasisocial” be- cause it forms extensive colonies (Fowler & Levi 1979). Fowler & Levi (1979) noted that the apparently closely related A. lorenzo forms perennial colonies that may contain hundreds of individuals that cooperate in prey capture, feeding and brood care. Here we confirm that A. rupununi has a level of social organization comparable to that of the permanent-social A. eximius and A. domingo Levi 1963. We note, however, some interest- ing differences between A. rupununi and these two species. In particular, in A. rupu- nuni individuals within the nests are relative- 497 498 THE JOURNAL OF ARACHNOLOGY ly well synchronized in the stage of their life cycle and group their egg sacs as part of ma- ternal care efforts. We discovered colonies of A. rupununi at two sites in eastern Ecuador, the Yasuni Na- tional Park (YNP) and the Jatun Sacha Bio- logical Station (JSBS). The YNP (including the adjacent Waorani Reserve) comprises 1,662,000 hectares of primary rainforest. We visited the area near the confluence between the Tiputini and Tivacuno rivers (0°41'S, 76°24^W, 210-250 m elevation) where the Estacion Cientifica Yasuni (ECY, Pontificia Universidad Catolica del Ecuador) is located. The JSBS is located on the southern banks of the Upper Napo River (r4'S, 77°36'W, 450 m) and comprises 2000 hectares of most- ly primary forest surrounded by farms. At both sites we searched for colonies of Ane- losimus spp. both within the forest and along the forest edge. At the YNP (19-23 July 1997) we inspected 4 km inside the forest, 10 km along the Tiputini river, and 7 km along the ECY-Tibacuno road. At the JSBS (9-17 July and 3-9 August 1997 and April 1998) , we inspected 6 km inside the forest, 55 km along the Arajuno, Napo, and Huam- buno rivers, and 20 km along the Tena-Ahu- ano road. Additionally, the colonies located at the JSBS in mid- July 1997 were monitored bi-weekly until mid-February 1998 or until their extinction. Web architecture. — We located eight nests of A. rupununi, all of them in forest edge or open and disturbed areas. A nest at the YNP occurred on the crown of a tree that hung over the Tiputini River. The seven nests located at the JSBS occurred on trees or bushes in farms adjacent to the preserve. We could not locate any nests in the forest interior where, in con- trast, we located numerous A. domingo and A. eximius nests. The nests of A. rupununi dif- fered from those of A. eximius in several re- spects: they were made of silk of a whiter and lighter appearance, contained almost no dry leaves and did not have a definite top-bottom polarity. In fact, rather than being basket- shaped, with a basal sheet and extensive silk lines extending upwards, the nests of A. ru- pununi enveloped pieces of vegetation on all sides and had only short if any lines extending upwards. This architecture may result from the location of the nests in areas with no other vegetation above. This architecture, combined with the spatial distribution of the nests, sug- gests that A. rupununi is a forest edge or can- opy species. Individual instars. — Spiders of all instars were of a generally dark brown or black col- oration that obscured the dorsal abdominal pattern characteristic of Anelosimus. Based on the state of the genitalia, general size, and body proportions, we classified the later-instar spiders into “juveniles,” “subadults” and “adults.” Measurements of the tibia plus the patella of leg I yield a multimodal distribution that supports this a priori classification of in- stars (Fig. 1). Based on these data, it appears that after the last undifferentiated juvenile in- star, both males and females have only one subadult instar before acquiring sexual matu- rity. This situation differs from that found in A. eximius and A. domingo where females are significantly larger than males as a result of having one additional subadult instar before maturing (Aviles 1986; unpubl. data). Egg sacs were subspherical in shape (2.00 X 2.00 X 2.75 mm, n = 1), of an off-white colora- tion, and tended to be bundled in groups of up to eight sacs (see below). Possibly as a camouflage mechanism, the sac bundles con- tained debris attached to their surface that gave them a flower like appearance. Egg sacs contained from 8-13 eggs (mean ± SE: 10.6 ± 0.6 eggs; n = 13). No parasitoids were pre- sent inside the sacs examined. Colony age structure and life cycle. — The four nests whose contents we inspected con- tained from a single adult female to close to 3000 spiders. The age distribution within the two largest nests (Table 1) suggested a definite synchronization in life cycle stage among in- dividuals within a colony. When seen in July 1997, the YNP colony contained mostly adult females and egg sacs, while the JS 1 colony contained primarily subadult females and no egg sacs. The synchronization in life cycle stages within the JS 1 colony continued throughout the period it was observed (Fig. 2). Six weeks after it was first recorded, the spiders in this colony had matured and laid their eggs. By week 10, the first spiderlings had emerged from the sacs, and by week 16 some of these spiderlings had reached the adult instar. In the mean time, the maternal females had de- creased in number and were apparently all gone before their offspring reached the sub- AVILES & SALAZAR— NOTES ON ANELOSIMUS RUPUNUNI 499 0.3 C o •c o Q. O 2 - 1 - Males n 0.2- Juveniles and females 0.1 LI 0.3 0.5 0.7 0.9 Tibia + patella length (mm) ■ juveniles □ subadults I '5 4 h3 2 hi h7 6 h5 4 3 -2 1.1 adults c o O Figure 1 . — Tibia + patella (leg pair I) measurements of late-instar Anelosimus rupununi spiders sepa- rated a priori into “juveniles,” “subadults,” and “adults” based on the state of their genitalia, size, and body proportions. Mean total body length (± SE) for the different instars and sexes are as follows: late- instar juveniles = 1.39 ± 0.03 mm; subadult females = 1.76 ± 0.02 mm; subadult males = 1.59 ± 0,05 mm; adult females = 1.81 ± 0.01 mm; adult males = 1.75 ± 0.05 mm. adult instar. No new sacs appeared in the col- ony between weeks 14-18. Males appeared much shorter-lived than females, as no adult males were noted in this colony during the periods of egg sac and offspring development. Consistent with this observation, almost no males were present in the YNP colony when it contained primarily adult females and egg sacs (Table 1). A shorter male lifespan in A. rupununi contrasts with the situation in A. eximius where males and females have adult lives of comparable length (Aviles 1986). The syn- chronization of life cycle stages within A. ru- pununi colonies also contrasts with the situ- ation in A. eximius and A. domingo, where, although separate, the generations within the colonies are less clearly distinct (Aviles pers. obs.). Strong internal synchronization of life cycle stages has also been described for other permanent- social species such as Achaeara- nea wau Levi, Lubin & Robinson 1982 (Lu- bin & Robinson 1982), Stegodyphus dumi- cola 1898 (Seibt and Wickler 1988), and Aebutina binotata Simon 1892 (Aviles in press). During the six months following its discov- ery, the spiders at the JS 1 colony completed one and a half generation cycles (Fig. 2). Ma- ture spiders of the offspring generation re- mained at the original site and in early De- cember—four months after the onset of the prior egg-laying cycle — started to lay their own eggs. Anelosimus rupununi, therefore. Table 1. — Size and inhabitants of Anelosimus rupununi nests when first seen in July 1997 at the Yasuni National Park (YNP) and the Jatun Sacha Biological Station (JS). Females Males Colony (cm) % scored Sacs Juv. Subad. Adult Subad. Adult YNP 88 X 45 X 12 100 244 14 5 229 0 4 JS 1 130 X 95 X 80 10 0 25 224 27 10 12 JS 2.1 100 10 present 0 7 0 0 JS 3 100 0 0 0 1 0 0 500 THE JOURNAL OF ARACHNOLOGY Egg sacs Juveniles Subadults Adult females Adult males Egg sacs Juveniles Subadults Adult females Adult males I 1 1 1 1 1 1 1 0 4 8 12 16 20 24 28 Weeks since nest discovety Figure 2. — Idealized life cycle of an Anelosimus rupununi colony based on bi-weekly inspections of a nest discovered at the Jatun Sacha Biological Station in mid-July 1997. Lines mark the presence of individuals of particular life cycle stages within the colony. may complete three generations a year, in con- trast with A. eximius that completes only be- tween 2.4-2. 6 generations a year at this same latitude (Aviles 1986). Sex ratio.— We estimated the tertiary sex ratio in a colony that contained primarily sub- adult spiders and, thus, would not have been affected by the shorter life span of adult males (Colony IS 1, Table 1). We are assuming that the collected ^/lO fraction of this colony is rep- resentative of the whole, as spiders of all in- stars appeared homogeneously distributed throughout the nest. The sex ratio among the 273 subadult and adult spiders in this sample was 8% males (4.7-12.5%, 95% c.i.) (Table 1). This value is strikingly similar to the 9% and 8% males reported as the sex ratio among developing embryos in A. eximius and A. domingo, respectively (Aviles & Maddison 1991). Sacs per female.- — Given the synchroniza- tion in life cycle stages within A. rupununi nests, the ratio of sacs to females during the peak of the egg laying period may be a rea- sonable representation of the number of sacs produced per female. The two colonies cen- sused at this stage contained 1.04 and 1.40 sacs per female, respectively (Table 1). In con- trast, estimates of the egg sac production in A. eximius that take into account its more pro- tracted egg-laying period indicate that females in this species typically produce fewer than one sac per female (Aviles & Tufino 1998; see also Vollrath 1986). The fecundity of A. ru- pununi, therefore, may be higher than what its small adult female body size and small num- ber of eggs laid per sac would lead us to sus- pect. Behavior. — We conducted casual observa- tions on prey capture and feeding in the YNP colony after it was collected whole and brought intact to the field station (the colony enveloped a stiff piece of vegetation and, thus, maintained its original shape and structure). Following the artificial introduction of prey items in the nest, we recorded two cooperative prey capture events. In one event seven spi- ders participated in biting and subduing a cricket that was 4-”5 times larger than the in- dividual spiders. In another event three fe- males cooperated in moving a captured mem- bracid towards the nest’s interior. In both cases, communal feeding followed. Groups of communally feeding spiders were also ob- served in other colonies in the field. A behavior not previously observed in spe- cies in the genus Anelosimus consisted in the bundling of the egg sacs in groups. Also in the YNP colony (see above), we observed a female in the process of completing an egg sac and attaching it to a nearby pair of bun- dled sacs. The egg sac was initially suspended from the web as the spider crawled around it adding to its surface silk she pulled from her spinnerets. After the sac was completed, the spider detached it and brought it towards a pre-existing pair of sacs. After attaching the sac to the pair, the spider mounted guard by the new-formed trio. Out of the 244 sacs present in this colony AVILES & SALAZAR— NOTES ON ANELOSIMUS RUPUNUNI 501 Figure 3. — Frequency distribution of guarded and unguarded egg sac bundles present in the YNP nest five days after it was collected whole. when collected, 182 — or 75% —occurred in groups of 2-7 sacs (Fig. 3). The remaining sacs were single. When we started dissecting the colony in the laboratory five days after its collection, we noted that 31 out of the 234 females in the nest were involved in sac guarding. However, 105 of the 244 sacs were being guarded as grouped sacs were signifi- cantly more likely to be tended than un- grouped sacs (median number of sacs per bun- dle among guarded sacs == 3, among unguarded sacs = 1; Mann- Whitney U = 1975, P < 0.0001) (Fig. 3). Usually a single female mounted guard by each sac bundle, ap- parently relayed by other individuals. During a one hour observation period we noted that females repeatedly moved away from the sacs they were tending as a second female with whom they exchanged leg touches approached the area. Sac bundling was also observed in the Jatun Sacha colonies. The 10 sacs present in the JS 2.1 colony (Table 1), for instance, occurred in a group of eight and a group of two. ACKNOWLEDGMENTS We wish to thank the Department of Bi- ology of the Pontificia Universidad Catolica del Ecuador for making this collaboration possible and for logistic support at the YNP, the Fundacion Jatun Sacha for making their facilities at the Jatun Sacha Biological Sta- tion available for this study, the Institute Ecuatoriano de Areas Naturales y Vida Sil- vestre for issuing research and collecting per- mits, J. Aukema, G. Canas, F. Martinez, and P Tufino for assistance in the field, and Ei- leene Hebets for assistance in the laboratory. This project was funded with NSF grant DEB-9707474 to L.A. This paper is dedicated to the memory of Vincent Roth whose enthusiasm and love for arachnology inspired us all. Vince collected the samples of A. rupununi that pointed to the presence of this species in Ecuador. This study would not have been possible without him. LITERATURE CITED Aviles, L. 1986. Sex-ratio bias and possible group selection in the social spider Anelosimus eximius. American Nat., 128:1-12. Aviles, L. 1997. Causes and consequences of co- operation and permanent-sociality in spiders. Pp. 476-498, In The Evolution of Social Behavior in Insects and Arachnids. (J. Choe & B. Crespi, eds,). Cambridge Univ. Press, Cambridge, AviMs, L. In press. Nomadic behavior and colony life cycle in a cooperative spider. Biol. J. Linn. Soc. Aviles, L. & W.P. Maddison. 1991. When is the sex ratio biased in social spiders?: Chromosome studies of embryos and male meiosis in Anelo- simus species (Araneae, Theridiidae). J, Arach- nol., 19:126-135. Aviles, L. & G. Gelsey. 1998. Natal dispersal and demography of a subsocial Anelosimus species and its implications for the evolution of sociality in spiders. Canadian J. ZooL, 76:2137-2147. Aviles, L., & P. Tufino. 1998. Colony size and in- dividual fitness in the social spider Anelosimus eximius. American Nat., 152:403-418. Brach, V. 1977. Anelosimus studiosus (Araneae: Theridiidae) and the evolution of quasisociality in theridiid spiders. Evolution, 31:154^161. Crespi, B.J. & J.C. Choe. 1997. Explanation and evolution of social systems. Pp. 499-524, In The Evolution of Social Behavior in Insects and Arachnids. (J. Choe & B. Crespi, eds.). Cam- bridge Univ. Press, Cambridge. Fowler, H.G. & H.W. Levi. 1979. A new quasiso- cial Anelosimus spider (Araneae, Theridiidae) from Paraguay, Psyche, 86:1U17. Furey, R.E. 1998. Two cooperatively social popu- lations of the theridiid spider Anelosimus studio- sus in a temperate region. Anim. Behav., 55:727- 735. Levi, H.W. 1956. The spider genera Neottiura and Anelosimus in America (Araneae: Theridiidae). Trans. American Micros. Soc., 75:407-422. Levi, H.W. 1963. The American spiders of the ge- nus Anelosimus (Araneae, Theridiidae). Trans. American Micros. Soc., 82:30-48. Levi, H.W. 1972. Taxonomic-nomenclatural notes on misplaced theridiid spiders (Araneae: Theri- diidae), with observations on Anelosimus. Trans. American Micros. Soc., 91:533-538. 502 THE JOURNAL OF ARACHNOLOGY Lubin, YD. & M.H, Robinson, 1982. Dispersal by swarming in a social spider. Science, 216:319- 321. Nentwig, W. & TE. Christenson. 1986. Natural history of the non-solitary sheetweaving spider Anelosimus jucundus (Araneae: Theridiidae). Zool. J. Linn. Soc., 87:27-35. Rypstra, A.L. & R.S. Tirey. 1989. Observations on the social spider Anelosimus domingo (Araneae, Theridiidae), in Southwestern Peru. J. Arachnol., 17:368-371. Seibt, U. & W. Wickler. 1988. Bionomics and so- cial structure of “Family Spiders” of the genus Stegodyphus, with special references to the Af- rican species S. dumicola and S. mimosarum (Ar- aneida, Eresidae). Verb, naturwiss. Ver. Ham- burg, 30:255-303. Smith, D.R. 1987. Genetic variation in solitary and cooperative spiders of the genus Anelosimus (Ar- aneae: Theridiidae). Pp. 347-348, In Chemistry and Biology of Social Insects. (J. Eder & H. Rembold, eds.). Verlag J. Pepemy, Munich. Vollrath, E 1986. Eusociality and extraordinary sex ratios in the Anelosimus eximius (Araneae: Theridiidae). Behav. Ecol. SociobioL, 18:283- 287. Manuscript received 15 March 1998, revised 26 October 1998. 1999. The Journal of Arachnology 27:503-512 MOVEMENT OF THE MALE BROWN TARANTULA, APHONOPELMA HENTZI (ARANEAE, THERAPHOSIDAE), USING RADIO TELEMETRY Margaret E. Janowski-Bell’ and Norman V. Horner: Department of Biology, Midwestern State University, Wichita Falls, Texas 76308-2099 USA ABSTRACT. This study was designed to gain insight into the “migratory” life history component of the male brown tarantula, Aphonopelma hentzi (Girard 1854), and to determine if radio telemetry could successfully answer questions regarding the ecology of theraphosids. Tarantulas were equipped with radio transmitters and movement monitored using an antenna and radio receiver. Overall movement of males was in all directions and randomness could not be excluded as a factor. Individual males moved relatively large distances, up to 1300 m, and significant directedness was only found in three individuals. In addition, notes on habitat, ecology and behavior are presented. Many spiders disperse over large distances by ballooning, and this is well documented in the aranemorphs (Weyman 1993, 1995; Wey- man et al. 1995). Mygalomorph spiderlings have been observed ballooning over short dis- tances (Bristowe 1939; Coyle 1983, 1985; Coyle et al. 1985); and Coyle (1983, 1985) concluded that ctenizid spiderlings could trav- el significant distances if they launch from taller vegetation. However, immature tarantu- las (Theraphosidae) are not known to balloon. There is no mention in the literature of the large scale movement of male tarantulas. This study was designed to determine the large scale distance and direction traveled by the mature male tarantula, Aphonopelma hentzi (Girard 1854), using radio telemetry. Mature males leave their burrows to search for mates from June to December (Baerg 1928, 1958; Gertsch 1979; Minch 1979a). In- dividual males have been observed crossing highways, appear to be moving in the same direction, and resist being redirected (Baerg 1958, 1963). Baerg (1958) stated that rarely are the movements of hundreds of individuals reported. Magnusson (1985) witnessed a co- ordinated movement of 89 male Cyclosternum sp. in Brazil. Mass movements could occur when weather conditions are ideal for travel (Baerg 1958). In addition, Baerg (1958) pro- posed a general “migration” of tarantulas ' Current address: 105 Tucker Hall, Division of Bi- ological Sciences, University of Missouri-Colum- bia, Columbia, Missouri 65211 USA. throughout the southwestern United States, Mexico, and possibly Panama. Baerg (1958) also suggested that reduced inbreeding could result from males moving large distances in search of mates. Migration differs from dispersal in many respects. Dispersal typically refers to the movement of individuals in a population that results in an increase in the mean distance be- tween individuals (Andrewartha & Birch 1954; Southwood 1981; Dingle 1996). Ac- cording to Danthanarayana (1986) and Dingle (1996), migrants usually exhibit five basic be- havioral characteristics: 1) persistent move- ment, 2) undistracted by the presence of re- sources promoting growth and maintenance (Kennedy 1961), 3) “straightened out” move- ment, 4) distinct leaving (Southwood 1962) and arriving behaviors, 5) reallocation of en- ergy specifically to support movement. In the United States Aphonopelma can be found west of the Mississippi River to the Pa- cific Coast and north into Arkansas, Utah, and Nevada (Baerg 1928; Gertsch 1979; Roth 1993). They are typically found on hillsides covered in sparse vegetation and mixed with diverse desert growth (Baerg 1928, 1958; Gertsch 1979). Tarantulas are usually noctur- nal, but may be active from late afternoon into late morning when light levels are low (Baerg 1958; Comstock 1975; Minch 1978). Radio telemetry is an effective tool for col- lecting data on organisms that are difficult to follow, observe or relocate (Mech 1983). His- 503 504 THE JOURNAL OF ARACHNOLOGY torically, it has been used extensively to fol- low the movements of larger animals (Mech 1983). As technology decreased the size of transmitters, this technique has been increas- ingly applied to smaller invertebrates includ- ing crayfish (Covich 1977), crabs (Gherardi et al. 1987, 1988a, b, c; Gherardi & Vannini 1989; Fletcher et al. 1990), snails (Bailey 1989; Tomiyama & Nakane 1993), and insects (Hayashi & Nakane 1988, 1989; Riecken & Raths 1996). Radio telemetry may provide unique insights into the ecology and behavior of the larger arachnids. This technology is used to study the brown tarantula, A. hentzi, an ideal subject for radio telemetry due to its activity, abundance and size. METHODS Study site.— The study was conducted on the W.T Waggoner Estate, 19.4 km WSW of Electra, Wilbarger County, Texas (33°58'N, 99°08'W). This area is part of the Rolling Plains region of Texas, which is a subsection of the Great Plains region of the central Unit- ed States (Lewis 1962). It is characterized by rolling-to-rough topography broken by inter- mittent streams (Lewis 1962). Annual rainfall for the area is approximately 76.2 cm, with May and September being the wettest months (Lewis 1962). The dominant vegetation is scrub mesquite (Prosopis), goat bush (Caste- ia), prickly pear cactus (Opuntia), turkey cac- tus (Opuntia), little blue stem {Schizachyr- ium), mesquite grass (Bouteloua), and broom weed (Xauthocephalum). The southern portion of the study site is dissected by a paved farm- to-market road running east to west. The area is broken by occasional dirt or gravel main- tenance roads. Past and current land use at the site include oil production and cattle grazing. Transmitters and receiver.— All radio te- lemetry equipment was purchased from Wild- life Materials, Inc. (Carbondale, Illinois). A model TRX-IOOOS receiver was used with a folding three-element yagi directional anten- na. The frequency range used was 150.000- 150.999 Mhz. Each transmitter (SOPB-2011) had a different frequency thereby identifying individuals. Transmitters weighed approxi- mately 0.6-0. 8 g and were 9X5X4 mm. The flexible antenna, constructed of wire sim- ilar to guitar string, was 7.62 cm in length. To prevent possible chafing of the abdomen and to provide minimal physical contact the an- tenna was bent upward at a 45° angle. Procedure. — -Male tarantulas were equipped with transmitters from 2 September- 17 October 1994 and from 9-19 July 1995. They were captured in a clear plastic container on predominantly open ground. Individuals were examined to determine overall physical condition. Those lacking obvious physical ab- normalities and exhibiting activity were weighed (to the nearest 0. 1 g) using an Ohaus LS200 portable scale. The only exception was specimen #13-94, which was missing the third left leg. Males ranged from 2. 5-7. 5 g. They were anesthetized with carbon dioxide for 2 min or until docile. Tarantulas were then placed on a thick synthetic sponge with legs extended. Their legs were restrained by plac- ing a second sponge which had been cut to expose the cephalothorax and abdomen over the first sponge. String was used to hold the two sponge pieces in place. This restrained the spider and facilitated attachment of the trans- mitter. To assist in the attachment of the con- tact adhesive, the “hairs” (setae) were re- moved from an area on the carapace, posterior to the eyes, by gently rubbing the area with a pair of forceps. A small amount of waterproof contact adhesive was placed on the the cara- pace and on the transmitter. After 5 min the adhesive on the transmitter was pressed into the adhesive on the spider. This was allowed to set for 20-40 min before the tarantula was removed from the sponge and placed back into the capture container. Equipped tarantulas (Fig. 1) were released at the exact site where they were collected within 2 h of capture. Tarantulas monitored in the fall of 1994 were observed a minimum of three days per week, while those in 1995 were observed once everyday, weather permitting. Locations of spiders were marked and labeled using flag- ging tape on adjacent vegetation. Direction traveled since the last observation was deter- mined by compass. Readings were corrected to reflect true north. Approximate distance traveled between observations was obtained using a tape measure or by pacing. Seventeen Aphonopelma hentzi males were monitored in the Fall of 1994 for movement. Of these, seven individuals retained their transmitters for four or more days and were considered for data analysis. This yielded a total of 113 observations. Six additional males JANOWSKI-BELL & HORNER— MOVEMENT OF MALE TARANTULAS 505 Figure 1. — Aphonopelma hentzi male with attached radio transmitter. were monitored in July 1995. These individ- uals were checked for short-term movement once every 24 hours. Four of these individuals retained their transmitters for three or more days, and yielded 20 observations. Identification. — The tarantula population studied was identified as Aphonopelma hentzi. Representative specimens are on deposit in the American Museum of Natural History, New York. The study site lies outside the known distribution reported by Smith (1994) for three species in the region. As a result, a name could not be assigned to this spider us- ing Smith’s (1994) descriptions. The validity of Smith’s species are in question (Prentice 1997). Cokendolpher (pers. comm.) noted that Smith failed to take individual variation into account. Therefore, the old name is applied to the common tarantula of Texas and Oklahoma. Representative specimens from the area were confirmed as A. hentzi by Dr. Rick West, Research Associate, Royal B.C. Muse- um (West pers. comm.). Statistics. — The samples from each year were compared using the Mann Whitney La- test for unmatched pairs (Fowler & Cohen 1990). There was a difference between the samples when weight was considered {U — 0, P < 0.05). However, there was no difference between the samples when the rate {U = 10, P > 0.05) and inflection points per day (C/ — 14, P > 0.05) were considered. Based upon these data, the samples from both years were combined for statistical analysis except where indicated. RESULTS Movement.^ — Figures 2-6 illustrate the movement of males and give a brief descrip- tion of the habitat for each observation. The weight (x = 5.0 ± 1.5 g SE), total time ob- served, total path distance, rate (x = 53.8 ± 25.7 m/d SD), number of inflection (turning) points per day (x = 0.64 ± 0.19 SD), point- to-point distance (distance from the first ob- servation to the last observation), and point- to-point angles with 0° being north (angle from first observation to last observation) ((() = 253.6° ± 70.8° SD) for each male are pre- sented in Table 1. Male #12-94 traveled the farthest with regard to both point-to-point dis- tance and rate (Fig. 4, Table 1). The most in- flection points per day was exhibited by #5- 94 in 1994 (Fig. 6, Table 1) and #1-95 in 1995 506 THE JOUR^IAL OF ARACHNOLOGY Table 1= — Individual tarantula number, weight, days observed, total distance of path traveled, rate of travel per day, number of inflection points per day, distance from first observation to last observation, direction from first observation to last observation. Spider # Weight (g) Total time observed (d:h:iniii) Total path distance (m) Rate (ro/d) Inflection points per day Point-to-point distance (m) Point- to-point angle (degrees) F94 — 4:18:35 148.8 31.2 0.60 102.1 236 5-94 7.0 24:00:00 1320.2 55.0 LOO 677.3 327 9-94 6.6 4:17:26 281.4 59.6 0.60 208.2 141 10-94 6.2 7:17:14 412.8 53.5 0.75 264.2 114 12-94 7.5 18:10:27 1750.2 94.9 0.72 1360.4 305 13-94 5.3 4:15:48 123.9 26.7 0.60 116.9 228 18-94 6.6 13:22:26 364.2 26.1 0.50 272.2 313 1-95 3.6 9:22:09 815.6 82.2 0.80 324.6 98 2-95 — 3:00:34 211.1 69.9 0.66 188.3 215 3-95 2.6 2:22:50 222.2 75.8 0.66 176.3 243 6-95 4.2 3:22:55 69.2 17.7 0.25 14.1 6 (Fig. 5, Table 1). Male #6-95 traveled the least with respect to rate, inflection points per day and poinhto^point distance (Fig. 3, Table 1). There was no correlation between tarantula weight and rate (rg = 0.10, « = 9, F > 0.05). However, there was a marginally significant correlation between rate and inflection points per day (r^ = 0.68, « = 11, 0.02 < P < 0.05). Overall movement of individual males was in almost all directions (Table 1). The Ray- leigh test [uses the mean vector (r) to deter- mine directedness (Batschelet 1981)] could not exclude randomness as a factor in the point-to-point movement of the combined samples from both years (r = 0.23, « = 11, P = 0.5) or of the sample from 1994 (r = 0.31, w = 7, F = 0.55). However, the move- ment of male #^-94 (r = 0.46, n = 26, 0.001 < F < 0.004; 4) - 328.9° ± 59.7° SD) and male #12-94 (r - 0.62, w = 18, F < 0.001; c!> = 335.4° ± 49.7° SD) indicated directed- ness. The results from Rao’s spacing test [uses angular data to determine directedness (Bat- schelet 1981)] yielded similar results with one exception; #13-94 exhibited directedness (U - 188, « = 5, F < 0.05; ^ - 247,1° ± 34.1°). Performance of traesmitters*=-All trans- mitters, except one, were recovered in work- ing order at the end of the study. The trans- mitter attached to #5-94 was recovered completely wrapped in silk within the en- trance of a tarantula burrow. There were abra- sions and breaches on the epoxy coating of the transmitter and a very large tarantula with a taunt abdomen was observed within the bur- row. There was no correlation between rate (m/d) and the percentage of the transmitter weight to body weight (r^ = 0.104, w = 9, F > 0.05). Three transmitters were known to have been removed from spiders within bur- rows. These transmitters were tightly wrapped in silk, and recovered just inside the burrow or a few centimeters from the buixow en- trance. Microhabitat and refugia.—Males were found moving through a variety of habitats from relatively barren, rocky ground to areas of dense vegetation. Tarantulas were observed to be traveling easily through the grass many centimeters above the ground. One male was observed hanging from vegetation several centimeters above the ground. Most males were inactive during the day and remained in sheltered environments (e.g., scrub thickets). Scrub thickets were 1-3 m in height, 1-3 m in diameter and dominated by mesquite (Prosopis), Other plants included goat bush {Castela), prickly pear cacti {Opun- tid), turkey cacti (Opuntid) and an understory of dense grasses (Schizachyrium, Bouteloud). Several small mammal burrows, outcroppings, and large limestone rocks were also used for shelter during the day. One locale, character- ized by limestone slabs 0.5 m in diameter and interspersed in dense grass, attracted three males within two days yielding a total of five observations. The abundance of data obtained JANOWSKI BELL & HORNER— MOVEMENT OF MALE TARANTULAS 507 LEGEND ■ Capture site * Tarantula burrow =r Predominately open ground ▲ Mesquite thicket Long, dense grass / Slope Goat bush thicket Large limestone rocks Intermittant creek Washout • • • Rocky soil Mammal burrow Paved road “ Direction of movement NORTH A B ◄ !! * 1 Figures 2-4. — Movement of Aphonopelma hentzi males. 2. Male #18-94 had the lowest rate and number of inflection points per day in 1994; 3. Male #6-95 had the lowest rate and number of inflection points per day m 1995; 4. Male #12-94 had the highest rate in 1994 at this locale was not typical of data collected during the rest of this study. Seven males were found within burrows with individuals that were assumed to be ma- ture females. Males remained at these locales from 1-3 days before continuing movement. Female burrows were found by males within thickets, in dense grass and on open ground. 508 THE JOURNAL OF ARACHNOLOGY NORTH LEGEND ■ Capture site * Tarantula burrow rz Predominately open ground A Mesquite thicket Long, dense gross / Slope A Goat bush thicket Large limestone rocks Intermittent creek <; Washout —< Dead mesquite stump • • • Rocky soil i)0 Cactus thicket Crack in ground O Transmitter replaced % Mammal burrow ■ Direction of movement 6 Figures 5, 6. — Movement of Aphonopelma hentzi males. 5. Male #1-95 had the highest rate and number of inflection points in 1995; 6. Male #5-95 had the highest number of inflection points per day in 1994. JANOWSKI-BELL & HORNER— MOVEMENT OF MALE TARANTULAS 509 Males probably visited more female burrows than detected due to the cover provided by dense vegetation. These “locales” or “sites” were not closely inspected for fear of dis- turbing the males and biasing movement re- sults. Behavioral observations. — On 2 Septem- ber 1994 male #1-94 was observed mating with a female approximately 0.5 m from the entrance of the female’s burrow on barren soil. Upon approach the spiders separated and then the female quickly retreated into her burrow. The male was captured and outfitted with a transmitter. Male #5-94 was 45.7 cm from her burrow for at least 18:54 h. On two separate occasions males were ob- served moving in overlapping counter-clock- wise circles 60-90 cm in diameter. This is similar to observations made by Shillington & Verrel (1997). Observations towards the end of the sum- mer indicated most males continually lost body mass and ultimately possessed very small abdomens. On the morning of 3 October 1994 male #12-94 (Fig. 4) was found in the open, positioned vertically with his abdomen in the air. The last two pairs of legs were stroking the antenna of the transmitter. His ab- domen was very small. In the evening he was found dead, legs curled under the shrunken abdomen. Males #2-95 and #3-95 were also found dead. However, males #1-95 (Fig. 5) and 6-95 (Fig. 3) were recovered, transmitter removed, and released at the end of the 1995 study period. DISCUSSION The movement of male tarantulas has been a subject of speculation and interest to arach- nologists for many years (Smith 1994). Most of what is known regarding the ecology of male tarantulas has been obtained from stud- ies regarding the ecology of female and im- mature individuals within the proximity of their burrows (Minch 1978, 1979a, b, c; Kotz- man 1990; Shillington & Verrell 1997) or in the laboratory (Baerg 1938, 1963), with two known exceptions (Baerg 1958; Sanderson 1988 as cited by Smith 1994). This is the first study known to extensively document the movement of male tarantulas. Radio telemetry proved to be useful and en- abled us to obtain data that would have been difficult to acquire using other methods. The performance of the transmitters was excellent. The signal could be detected several hundred meters from the spider. In addition, the males could be located easily when in burrows and under rocks. It was assumed the transmitters would have a minimal effect upon the move- ment of individual tarantulas. Maneuverability in small spaces was a concern, but the antenna proved flexible enough to allow males to enter and remain in burrows. Tagged males were observed crawling around or past residing fe- males within the burrow. Attachment of the transmitter to the taran- tula was a problem. Many of the contact ad- hesives used did not maintain their bond. This resulted in several spiders losing their trans- mitters within a few days and was the limiting factor of this study. Organisms use a hierarchical set of cues to locate resources and exhibit behaviors appro- priate for each level: habitat, patch, individual resource. (Bell 1991). It has been shown (Baerg 1958; Minch 1979c; Shillington & Verrel 1997) that male tarantulas are able to detect local “cues” provided by females in the vicinity of their burrows. These have not been shown to be directional, but do elicit lo- cal search behavior described as “animated circular motion” (Shillington & Verrel 1997), and were observed in this study. Baerg (1928, 1958), Gabel (1972), and Kotzman (1990) have noted the clumped or patchy distribution of theraphosid burrows. Theraphosids are al- most blind and cannot see beyond 2.5-5 cm (Baerg 1958). As a result, it is unlikely taran- tulas use visual environmental cues when searching for burrow patches. Bell (1991) pro- posed several search strategies organisms may adopt when lacking environmental cues while searching for resource patches: random walk, straight line, systematic movement pattern (spiral or parallel movement), kinesthetic-in- put mapping or a combination of these strat- egies. The results of this study did not reveal a systematic movement pattern, and this may be because observations were not at the ap- propriate scale. Three large scale loops were observed among individuals considered for data analysis (Figs. 5, 6). These imply male tarantulas are conducting systematic searches to locate mates within “colonies.” The move- ment of searching males between “colonies” may be a combination of random walks and straight line movements as indicated by the 510 THE JOURNAL OF ARACHNOLOGY directedness expressed in only a few individ- uals. However, the lack of directedness may be a function of the time individuals were foL lowed. There was limited evidence to support the axiom that male tarantulas are migrating us- ing Dingle (1996) and Danthanarayana’s (1986) definition. Based upon their observa- tions of North American tarantulas, Baerg (1963) and Minch (1978) report that males travel farther during the mating period than any other period of the life cycle. There was some evidence that movement of individuals was directed. Later in the season males were observed with notably smaller abdomens in- dicating energy may be reallocated specifi- cally for movement. Male #12-94 (Fig. 4) moved 30% (537 m) of the total distance (1750 m) in 9% (1 day, 16h, 25 min) of the total time (18 days, lOh, 27 min). Baerg (1928) and Minch (1979c) noted that males die at the end of the mating season. Their observations noted that preceding death the abdomen becomes shrunken, ability to ex- tend legs is lost, and overall sluggish behav- ior occurs. There was little evidence for a synchronized, directional movement of all males sampled. As a result, no definitive con- clusions can be drawn regarding the charac- terization of the movement of male tarantulas as migratory. Further behavioral studies would serve to elucidate the characterization of this behavior. Early summer males were smaller in size and weight. As a group they were not signif- icantly different with respect to their rate and inflection points per day, and they behaved the same ecologically with regard to movement. Further study is needed to determine if the early males have overwintered as adults or simply molted earlier than the late summer brood. Males frequented scrub thickets, small mammal burrows and large limestone rocks throughout the study. These habitats probably provided protection from predators and al- lowed better thermoregulation during the day. In addition, the locale characterized by lime- stone rocks may have a high density of mature females not observed due to burrows being hidden by the rocks. Males were frequently found within taran- tula burrows in the presence of other individ- uals. These were presumed to be mature fe- males and males were courting and mating with them. Multiple matings with the same female are probable given the length of time males were in the presence of these females, which ranged from one to several days. Males also visited multiple females (Figs, 2, 4, 5, 6). The data reflect fewer matings than probably occurred due to sampling and lack of detec- tion within thickets. Baerg (1958) suspected mating occurs primarily inside the burrow. However, male #1-94 was observed mating with a female approximately 0.5 m from her burrow entrance. This study indicates radio telemetry is a valuable technique for studying the move- ment of male Aphonopelma hentzi. Males were observed moving large distances, up to 1300 m, over a significant period of time, up to 18 days, while searching for mates. Cur- rent research on the behavior and ecology of movement in tarantulas includes: evaluation of extensive movement, influence of different habitat types, effect of habitat fragmentation. ACKNOWLEDGMENTS We thank the WT. Waggoner Estate in Ver- non, Texas for their hospitality, providing free access to their land, and furnishing living quarters on the ranch. Funds for the purchase of the radio-telemetry equipment and other supplies for this project were provided by the Administration and Biology Department at Midwestern State University, for which we are most appreciative. We are indebted to Dr. Rick West, Royal B.C. Museum, for species confir- mation. Becky Janowski Technical Services (http:5si.com/bechtech/) donated time and ser- vices to produce figures. We are indebted to Pauline Janowski who generously furnished funds for extraneous expenses. David V. Bell and R.W. Whisnand II provided field assis- tance, for which we are grateful. Appreciation is expressed to James Carrel and James Cok- endolpher for early reviews of this manuscript and two anonymous reviewers from the Jour- nal whose suggestions were beneficial. This paper is dedicated to the memory of Mr. Glen Collier, Waggoner Ranch Game Warden, who acted as a liaison between Mid- western State University and the Waggoner Estate. He was an insightful, sagacious natu- ralist, photographer and friend. JANOWSKI-BELL & HORNER— MOVEMENT OF MALE TARANTULAS 511 LITERATURE CITED Andrewartha, H.G. & L.C. Birch. 1954. The Dis- tribution and Abundance of Animals. Univ. of Chicago Press, Chicago. Bailey, S.E. 1989. Foraging behavior of terrestrial gastropods: integrating field and laboratory stud- ies. J. Molluscan Stud., 55:263-272. Baerg, W.J. 1928. The life cycle and mating habits of the male tarantula. Q. Rev. Biol., 3:109-116. Baerg, W.J. 1938. Tarantula studies. J. New York Entomol. Soc., 46:31-43. Baerg, W.J. 1958. The Tarantula. Univ. Kansas Press. Lawrence, Kansas. Baerg, W.J. 1963. Tarantula life history records. J. New York Entomol. Soc., 71:233-238. Batschelet, E. 1981. Circular Statistics in Biology. Academic Press, New York. Bell, W, 1991. Searching Behavior: The Behavior- al Ecology of Finding Resources. Chapman and Hall, New York. Bristowe, W.S. 1939. The Comity of Spiders. Vol. 1. Ray Society. London. Comstock, J.H. 1975. The Spider Book. Cornell Univ. Press, Ithaca, New York. Covich, A. 1977. Shapes of foraging areas used by radio-monitored crayfish. American Zook, 17: abst. #205. Coyle, FA. 1983. Aerial dispersal by mygalo- morph spiderlings (Araneae, Mygalomorphae). J, ArachnoL, 11:283-286. Coyle, F.A. 1985. Ballooning behavior of Ummidia spiderlings (Araneae, Ctenizidae). J. ArachnoL, 13:137-138. Coyle, F.A., M.H. Greenstone, A.L. Hultsch & C.E. Morgan. 1985. Ballooning mygalomorphs: Es- timates of the masses of Sphodros and Ummidia ballooners (Araneae: Atypidae, Ctenizidae). J. ArachnoL, 13:291-296. Danthanarayana, W. 1986. Introductory chapter. Pp.1-10, In Insect Flight: Dispersal and Migra- tion. (W. Danthanarayana, ed.). Springer-Verlag, Berlin, Germany, Dingle, H. 1996. Migration: The Biology of Life on the Move. Oxford Univ. Press, New York. Fletcher, W.J., I.W. Brown & D.R. Fielder. 1990. Movement of coconut crabs, Birgus latro, in a rainforest habitat in Vanuatu, Pacific Sci., 44: 407-416. Fowler, J. & L. Cohen. 1990. Practical Statistics for Field Biology. Open Univ. Press, Philadel- phia, Gabel, J.R. 1972. Further observations of thera- phosid tarantula burrows. Pan-Pacific Entomol., 48:72-73. Gertsch, WJ. 1979. American Spiders. 2nd ed. Van Nostrand Reinhold Co., New York. Gherardi, E, F. Micheli & C. Nocchi. 1987. Ecoethology in river crabs: The use of radio-te- lemetry. Monit. Zool. Italiano, 21:189-190. Gherardi, E, G. Messana, A. Ugolini & M. Vannini. 1988a. Studies on the locomotor activity of the freshwater crab, Potomon fluviatile. Hydrobiol- ogia, 169:241-250. Gherardi, E, M.C. Nocchi & M. Vannini. 1988b. Slope orientation in freshwater crabs: A field study. Monit. ZooL Italiano, 22:63-76. Gherardi, E, E Tarducci & M. Vannini. 1988c. Lo- comotor activity in the freshwater crab Potamon fluviatile: The analysis of temporal patterns by radio-telemetry. Ethology, 77:300-316. Gherardi, F. & M, Vannini. 1989. Spatial behavior of the freshwater crab, Potomon fluviatile, a ra- diotelemetric study. Biol. Behav., 14:28-45. Hayashi, F. & M. Nakane. 1988. A radio-tracking study on the foraging movements of the dobson- fly larva, Protohermes grandis (Megaloptera: Corydalidae). Kontyu, 56:417-429. Hayashi, F, & M. Nakane. 1989. Radio tracking and activity monitoring of the dobsonfly larva, Protohermes grandis (Megaloptera: Corydali- dae). Oecologia, 78:468-472. Kennedy, J.S. 1961. A turning point in the study of insect migration. Nature, 189:785-791. Kotzman, M. 1990. Patterns of the Australian ta- rantula Selencosmia sterlingi (Araneae: Thera- phosidae) in an arid area. J. ArachnoL, 18:123- 130. Lewis, R.D. 1962. Texas Plants — A Checklist And Ecological Summary. Texas Agric. Exp. Stn., College Station, Texas. Magnusson, WE. 1985. Group movement by male mygalomorph spiders. Biotropica, 17:56. Mech, L.D. 1983. Handbook of Animal Radio- Tracking. Univ. Minnesota Press. Minneapolis. Minch, L.W. 1978. Daily activity patterns in the tarantula Aphonopelma chalcodes Chamberlin. Bull. British ArachnoL Soc., 4:231-237. Minch, L.W. 1979a. Annual activity patterns in the tarantula Aphonopelma chalcodes Chamberlin (Araneae: Theraphosidae). Nov. Arthrop., 1:1- 34. Minch, L.W. 1979b. Burrow entrance plugging be- havior in the tarantula Aphonopelma chalcodes Chamberlin (Araneae: Theraphosidae). Bull. British ArachnoL Soc., 4:414-415. Minch, L.W. 1979c. Reproductive behavior of the tarantula Aphonopelma chalcodes Chamberlin (Araneae: Theraphosidae). Bull. British Arach- noL Soc., 4:416-420. Prentice, T.R. 1997. Theraphosidae of the Mohave Desert west and north of the Colorado River (Ar- aneae, Mygalomorphae, Theraphosidae). J. Ar- achnoL, 25:137-176. Riecken, U. & U. Raths. 1996. Use of radio telem- etry for studying dispersal and habitat use of Carabus coriaceus L. Ann. Zool. Fennici, 33: 109-116. Roth, V.D. 1993. Spider Genera of North America. 512 THE JOURNAL OF ARACHNOLOGY 3rd ed. American Arachnological Society, Gainesville, Florida. Sanderson, M. 1988. Observations on Aphonopel- ma behlei in Northern Arizona. American Mu- seum of Natural History-New York (unpub- lished). Shillington, C. & R Verrell. 1997. Sexual strategies of a North American ‘tarantula’ (Araneae: Ther- aphosidae). Ethology, 103:588-598. Smith, A.M. 1994. Theraphosid Spiders of the New World. Vol. 2. Tarantulas of the U.S.A. and Mexico. Fitzgerald Publ., London, England. Southwood, T.R.E. 1962. Migration of terrestrial arthropods in relation to habitat. Biol. Rev., 37: 171-214. Southwood, TR.E. 1981. Ecological aspects of in- sect migration. Pp. 196-208, In Animal Migra- tion. (D.J. Aidely, ed.) Cambridge Univ. Press, Cambridge, England. Tomiyama, K. & M. Nakane. 1993. Dispersal pat- terns of the giant African snail, Achatina fulica (Ferussac) (Stylommatophora: Achatinidae), equipped with a radio-transmitter. J. Molluscan Studies, 59:315-322. Weyman, G.S. 1993. A review of the possible causative factors and significance of ballooning in spiders. Ethol. Ecol. EvoL, 5:279-291. Weyman, G.S. 1995. Laboratory studies of the fac- tors stimulating ballooning behavior by linyphiid spiders (Araneae, Linyphiidae). J. ArachnoL, 23: 75-84. Weyman, G.S., PC. Jepson & K.D. Sunderland. 1995. Do seasonal changes in numbers of aeri- ally dispersing spiders reflect population density on the ground or variation in balloning motiva- tion? Oecologia, 101:487-493. Manuscript received 1 May 1998, revised 10 Oc- tober 1998. 1999. The Journal of Arachnology 27:513-521 PHENOLOGY AND LIFE HISTORY OF THE DESERT SPIDER, DIGIJETIA MOJAVEA (ARANEAE, DIGUETIDAE) April M. Boulton and Gary A. Polis: Department of Environmental Science and Policy, University of California, Davis, California 95616 USA ABSTRACT. The desert spider, Diguetia mojavea Gertsch 1958, is a numerical dominant in many California deserts. We report data collected over a three-year period (1984-86) on reproduction, life history, phenology, microhabitat, prey, and dispersion for D. mojavea in the Coachella Valley, California. This is one of few studies to calculate life history table parameters for a desert arachnid. The average female laid 1065 eggs, while the net reproductive rate (RJ was 1.41; generation time (T) was calculated as 204.85 days. These spiders appear to fit a Type III survivorship curve. Density of D. mojavea was typical for a desert spider at 0.02 spiders/m^. Finally, our findings complement the only other study on D. mojavea (Nuessly & Goeden 1984). Spiders in the family Diguetidae Gertsch 1949 are primitive, six-eyed weavers con- tained in three genera, Pertica Simon 1903, Segestrioides Keyserling 1883 and Diguetia Simon 1895 (Platnick 1989). Diguetia, the dominant genus in this family, consists of spi- ders with elongate legs that weave character- istic funnel or net webs, or a combination thereof. Although they range widely from the southwestern United States to southern Mex- ico (Comstock 1948; Gertsch 1958; Lopez 1984) and parts of Argentina (Gerschman de Pikelin & Schiapelli 1962), few papers have focused on this family since its formal de- scription by Gertsch (1949). Gerschman de Pi- kelin & Schiapelli (1962) studied the web characteristics of D. catamarquensis (Mello- Leitao 1941) in Argentina, Eberhard (1967) investigated prey capture and wrapping be- havior in D. alboUneata Simon 1898, and Bentzien (1973) described behavior and repro- ductive biology of D. imperiosa Gertsch & Mulaik 1940. Diguetia canities McCook 1895, the most widespread species (Cazier & Mortenson 1962), is best-studied due both to its relative abundance and its commercial im- portance in insecticide development (e.g., Krapcho et al. 1995; Hughes et al. 1997). Diguetia mojavea Gertsch 1958 is distrib- uted throughout southern California and ad- jacent areas in Nevada (Gertsch 1958), and it also appears to be one of the numerically dominant spider species in some California desert areas (Polis 1991). However, only one paper (Nuessly & Goeden 1984) focuses on the biology and ecology of this species, and a few others mention D. mojavea briefly (e.g., Polis & McCormick 1986; Polis 1991). Here, we examine D. mojavea'^ phenology in more detail. We also report life history statistics used to calculate D. mojavea^ net reproduc- tive rate and rate of potential increase both because of its significant role in various desert ecosystems and its potential impact as an im- portant biological control agent (Nuessly & Goeden 1983). METHODS Relevant biology.— Several characteristics facilitate research on D. mojavea. First, pop- ulations are relatively dense (see Results). Second, the web of . adults is large (mean length: 33.8 cm; mean width: 24.0 cm; see also Nuessly & Goeden 1984) and quite vis- ible, especially in early morning or late after- noon at low sun angles. Third, the adult fe- male web includes a retreat containing eggs, thus facilitating studies on reproductive biol- ogy. Fourth, prey and diet are easily quantified because D. mojavea incorporates most prey into its web (Gertsch 1958). Study site. --Field studies were conducted within and adjacent to the Coachella Valley Reserve of southern California (Riverside County, California; 33°54'N, 166°37' W). The Reserve encompasses about 780 km^ and spans an elevational gradient from 320 m in the northwest to sea level in the southeast. Winters are mild; summers, hot and dry. Air temperature in July annually exceeds 40 °C 513 514 THE JOURNAL OF ARACHNOLOGY and temperatures greater than 50 °C occur (Edney et al. 1974; Polls 1988). It is a low elevation rain shadow desert, with annual rainfall at the University of California’s Deep Canyon Field Station averaging 116 mm, ranging from 34 mm in 1961 to 301 mm in 1976. Vegetation includes Atriplex caescens (saltbush), Salsola australis (Russian thistle), Larrea tridentata (creosote), Tamarix sp. (salt cedar), and annual plants and grasses. We sur- veyed web sites dispersed over an area of 7500 m^ divided into 300 quardats, each 5 X 5 m. Quadrats were marked with flags and surveyed at least every three weeks from early June to September in 1984-86. Webs and egg sacs were also collected in December 1984- 86 and 1997. Egg sac analysis. — Diguetia mojavea’s egg morphology and egg sac construction are sim- ilar to D. canities (Cazier & Mortenson 1962). Each sac is constructed beneath the previous one in a shingle-like fashion, which is then incorporated into the tube retreat (Gertsch 1979). To examine seasonal patterns in egg- laying in August, September, and December 1984-86, we randomly chose 31 retreats to examine the number of egg sacs. All females were usually absent in our December survey (see Life history results); thus no new egg sacs could be laid, and these data are then used to estimate average number of egg sacs laid in a female’s life. Egg sacs were dissected for egg counts. Stages within the sac were classified as either egg, embryo/deutovum, 1st instar, 2nd instar (based on cephalothorax length), or dead. Because maternal-guarding of the egg sacs seemed to play an important role in the life history of the adult females, we assumed each web contained only the res- ident female’s eggs. Here as throughout the paper, means are reported with their standard deviations. Spider web/microhabitat analysis. — For each spider in our quadrats, we recorded life stage (spiderling, adult) and sex of adults. Web characteristics were collected for the 137 webs in our plot in 1984. The volume of each web was calculated using height, width, and length. We measured retreat height and iden- tified the plant on which webs were placed. Entire retreats and egg sacs were randomly collected in 1985-86 outside our quadrats. These were preserved in alcohol. For these spiders, we recorded adult-spider mass, the number of egg sacs and the egg stage (see above) for each web. Prey analysis.— We exanfined diet by an- alyzing the prey from 111 retreats collected from 1984-86. Most prey are incorporated into the web, but some very large prey were discarded into the sheet or onto the ground. Prey items were easily separated from the web using a dilute bleach solution (Nuessly & Goeden 1984). A separate sub- sample (n = 26) of prey was taken from the sheet-web for analysis. Collected prey were identified to or- der and/or family. Dispersion, phenology and life history.— The 300 quadrats in our survey area of 7500 m^ were censused throughout the study to de- termine D. mojavea's density, phenology and dispersion. These data were used to calculate survivorship curves and a Greig-Smith block size analysis in 1984 (Pielou 1977). This par- ticular dispersion analysis determines if or- ganism-spacing is aggregated, regular, or ran- dom. Fecundity variables (Pianka 1978) were also calculated. Parameters (fraction of sur- viving spiders at age X) and (number of offspring produced by an average spider at age X) were used to calculate net reproductive rate (RJ and generation time (T). All formulae follow Pianka (1978). RESULTS Phenology. — The life cycle of D. mojavea encompasses approximately one year. Spider- lings emerge in late December through March. Females mature in May through June; the first egg sacs are produced in August through Sep- tember (Figs. 1, 2), 2-3 months after the last molt. Approximately 85% of all egg sacs are laid during this two month period. Adult males first appeared in July. From 1984-86, adult females senesced and died from October to mid-December. In 1997, a few adult fe- males were observed living in their webs as late as 20 December (7 out of 31). Males typ- ically died one to two months before females (Fig. 2). We observed males in 14% of the 137 webs collected in July-September. The average mass of the adult females was 36.8 ± 23.6 mg, with a mean carapace length of 3.8 ± 0.8 mm. From our 1997 sample, the mean carapace length of 1st instar spiderlings was 0.66 ± 0.03 mm. Egg sac analysis. — All egg cases appeared to be laid by late September. One to 13 total BOULTON & POLIS—NATURAL HISTORY OF DIGUETIA MOJAVEA 515 Date (moiith/day/year) Figure L — Oviposition phenology. Seasonal changes in percentage of webs of D. mojavea found with egg sacs {n = 96). egg cases were laid per female. All 3 1 retreats examined in our December collection con- tained egg sacs (Fig. 1); the number of egg sacs and the total number of eggs a female laid were significantly correlated (Fig. 3). Over the three-year period (1984-86), we col- lected 364 webs that contained 1083 intact egg cases. The average number of egg cases per female was 4.9 ± 2.7 (Fig. 4). The mean number of eggs per sac from 1984-86 was 217.4 ± 31.6, data taken from a subsample of 237 egg cases. Thus, the mean number of eggs laid per female was 1065.3 ± 381.2. Egg sacs were laid over a period of days as evidenced by personal observation, presence of multiple egg sacs, and staggered emergence of spiderlings. We examined the stage of de- velopment for the dissected sub-set {n = 237) of the egg sacs we collected. Each case con- tained only individuals in the same develop- mental phase. However, developmental stage did differ among egg cases within a particular female’s retreat. Of the 237 cases examined, 47% {n = 111) were classified as containing eggs; 19%, embryos (deutova); 17%, 1st in- star; 4%, 2nd iestar; and 13%, shriveled, dead eggs. The sequence of development followed the order in which the sacs were deposited: the uppermost egg sac located at the tip of the retreat always contained the most advanced stage, and sacs toward the retreat opening contained only eggs. Web/microhabitat analysis.— The first typical webs we noted were built in late May by spiders with a body length of 3-4 mm. As the summer progressed, webs became larger and were placed in progressively higher veg- etation. The mean size of a web in early sum- mer (10 June) was 28.3 cm in length (range: 10-50 cm) by 21.1 cm across (range: 10-43 cm). By mid-summer, mean size increased to 33.8 cm (20-60 cm) by 24.0 cm (12--40 cm). There was a significant correlation between spider weight and web volume (Fig. 5) and 516 THE JOURNAL OF ARACHNOLOGY Aug Sept Oct Nov Dec Jan Feb Mar Apr May June July MONTH Figure 2. — Phenology. Bars show months in which each stage occurred. Data collected from 1984- 1986. between shrub height and spider weight (Fig. 6) (analyses conducted on July/August data). Web location changed throughout the year with less desirable sites (i.e., dead bushes) supporting progressively fewer spiders. In ear- ly sununer, 68.6% of the webs were built in perennial bushes (39.8% in Atriplex and 28.8% in Salsola), while the remaining 31.4% were placed in both living and dead annuals. Larrea and Tamarix were only rarely used as web-sites possibly due to their thin, exposed and flexible branches. Webs persisting into Figure 3. — Egg production. The number of eggs laid per female was significantly correlated with the total number of egg sacs (y = — 1.73 T 193. 66x, R = 0.86, P < 0.01), number of retreats examined = 31. late summer remained only on larger Atriplex and Salsola as winds damaged and uprooted annual plants. Moreover, 43.8% of monitored webs were tom by wind and/or abandoned; unprotected web sites near the ground repre- sented 72.1% of these cases. Prey analysis. — From May through Octo- ber, spiders were observed feeding primarily in early morning or late afternoon, thus avoid- ing the mid-day heat. Prey capture was ob- served on several occasions. After prey were detected in the web, the resident would mn to the prey and immediately (< 10 seconds) im- mobilize it with a bite. Silk, although used to secure prey to the web, was not used for im- mobilization (see also Eberhard 1967). Prey remains were analyzed from 111 webs; a total of 6771 individual prey was identified to order and/or family (Table 1). On average, each web contained 61 ± 17.8 prey items; mean prey-size in retreats was 5.2 ± 0.9 mm. Homoptera (Cicadellids), small Hy- menoptera, and Coleoptera comprised 88.1% of D. mojavea's diet. A coleopteran egg pred- ator (Cleridae, Phyllobaenus discoideus) was occasionally caught. Five other spider species comprised 3.4% of the diet and occurred in 30% of examined retreats. Cannibalism was recorded three times. The sheet-web subsample {n = 26) yielded fewer (< 5 prey/sheet- web) but much larger prey (14.9 ± 6.0 mm; 131 prey analyzed). Only 3% of all prey items (mean = 1.9 ± 0.7) BOULTON & POLIS—NATURAL HISTORY OF DIGUETIA MOJAVEA 517 Number of Egg Sacs Figure 4. — Distribution of egg sacs in retreats. Egg sac number per retreat for data from 1984-1986 (total number of egg sacs examined = 364). were dropped on the sheet^web by the spider. Although we did not measure biomass, large prey certainly represented more than 3% of total prey biomass. The largest prey items were a mantid (25 mm), mud dauber wasp (25 mm), robber fly (24 mm), and cicadid (24 mm). Grasshoppers were the most common y - 0.609 + 0.036X, R = 0.55, p < 0.001 Figure 5. — Web volume as a function of spider mass. Spider weight was significantly correlated with web volume (y = 0.609 + 0.036x, R = 0.55, P < 0.001) for all three years combined (1984- 1986). larger prey in sheet^webs but constituted less than 1% of D. moJavea\ total diet. Mortality and survivorship.- — Several predators were observed in diguetid webs. Clerid beetle larvae (P. discoideus), reported egg predators (Cazier & Mortenson 1962), emerged in the lab from about one- seventh of the webs analyzed {n ~ 137). Salticids were observed eating both diguetid eggs and adults; Figure 6. — Shrub height as a function of spider mass. Spider weight was significantly correlated with shrub height (y = 52.889 + 0.27 lx, R = 0,24, P < 0,01). Data are for all three years combined (1984-1986). 518 THE JOURNAL OF ARACHNOLOGY Table 1. — Prey of Dignetia mojavea. Italics indicate prey orders with families listed beneath when possible. Numbers in parentheses indicate total for the group. Taxa % of diet % Occurrence among retreats Arachnida (3.4) 30 Diguetidae 1.0 Mimetidae 0.8 Oxypidae 0.2 Salticidae 1.4 Hemiptera (e.g., Pentatomidae) (2.2) 16.2 Homoptera (36.7) Cicadellidae 33.9 80.2 Cicadidae 0.1 Isoptera (<0.01) 0.05 Orthoptera (e.g., Acrididae, Mantidae) (0.5) 4.5 Coleoptera (17.9) 61.3 Tenebrionidae 16.9 Cleridae 0.01 Diptera (e.g., Asilidae) (3.4) 23.4 Hymenoptera (33.5) Pompillidae 27.3 Apidae 3.2 Formicidae 2.9 Sphecidae 0.1 79.3 Lepidoptera (e.g., Coleophoridae) (2.2) 14.4 Habronattus tranquillus and Metaphidippus manni (G. & E. Peckham) appeared to be the most frequent predators of D. mojavea. Para- sitism was not observed in this study. Figure 7 shows average spider density through time summed over all quadrats in 1984. Adult density decreased in an almost linear fashion throughout the summer from July through September 1984, while egg pro- duction increased throughout each summer. If we assume that egg production in 1983 was similar in our plot to that in 1984, only 137 females in 7500 m^ out of approximately 123,600 eggs survived to adulthood (< 0.01%). This represents a Type III survivor- ship curve (Pianka 1978). Finally, adult den- sity decreased from 0.02 spiders/m^ (137 spi- ders/7500 m^) in July to 0.003 spiders/m^ (19 spiders/m^) in September. Dispersion. — A Greig-Smith block size analysis of dispersion (Fig. 8) indicated that these spiders were not randomly dispersed in our study plots. Significant aggregations ap- peared at block sizes of 1 and 8 m. Life history. — -We calculated life history statistics using the data on number of eggs laid per female and spiderling/adult emer- gence/survival. Table 2 summarizes fecundity variables (Pianka 1978). The empirical net re- productive rate, R^, (Pianka 1978) was 1.41, which was easily calculated since all spiders remaining at the end of the season were fe- male. Generation time (J) was calculated as 204.85 days. The average number of eggs laid per female (1065.3) was applied to all adult females with egg cases, making (the num- ber of offspring produced by an average or- ganism at age X) equal to 1.0 (or 100%) for August and September age values (i.e., egg cases were present in August through Octo- ber; Fig. 2). For the July age class, was 0.0 because no egg cases were observed before BOULTON & POLIS—NATURAL HISTORY OF DIGUETIA MOJAVEA 519 Figure 7. — Survivorship. Spider density decreases as time of season increases while egg number (XlOOO) increases throughout the season. Data are from 1984. August. The maximum rate of natural in- crease, (In RJT) was calculated as 1.68 X 10-Yd. DISCUSSION There are a number of unique morpholog- ical and ecological characteristics exhibited by D. mojavea. Phenology. — Recall that we observed Block Size (Log Square Meters) Figure 8. — Dispersion analysis. Greig-Smith block size analysis shows statistically significant aggregations at 1 and 8 m blocks. Data are from 1984. hatching of spiderlings in the field from De- cember-March, adult females from June-De- cember, adult males from July-October, and egg-laying from August-October. The univol- tine and semelparous qualities of D. mojavea are typical of other diguetids (e.g., Bentzien 1973) and other spiders, in general (Foelix 1996). Moreover, a similar 1-year, 1-egg sac pattern is found in many desert arthropods, which may be due to the costliness of egg production and longevity in desert ecosystems (Polis 1991). Reproduction and life history. — Our ob- servations indicated that egg production in D. mojavea is slightly lower than that reported by Nuessly & Goeden (1984) (an average of 6.4 egg sacs/nest with a mean of 176 eggs/sac for a total of 1126 eggs/web). Diguetia mo- javea" s egg production fits well within the range of eggs produced by other spiders (Foe- lix 1996). Staggered emergence is seen in spiders that have multiple egg sacs. This tactic may pro- vide insurance against synchronous emer- gence during unfavorable conditions in harsh environments such as the desert. Thus, D. mo- javea lessens the risk of failure through vari- able hatching times. Such a strategy is also seen in certain annual weed species and var- 520 THE JOURNAL OF ARACHNOLOGY Table 2. — Life history parameters calculated from the data collected throughout the study: = fraction of spiders surviving at age X\ = the number of offspring produced by an average spider at age X\ = fecundity schedule; = net reproductive rate; T — generation time (days). Formulae are based on Pianka (1978). Age (X) in days h m. Ijn, Xl^, 101 1.0 0.0 0.0 0.0 138 0.7 1.0 0.7 96.6 143 0.57 1.0 0.57 81.51 191 0.14 1.0 0.14 26.74 Total R, = 1.41 T = 204.85 ious other opportunistic species in the desert environment (Polls 1991). Staggered emer- gence is observed in most spiders that have multiple egg sacs (e.g., Bristowe 1958; Jack- son 1978); it is probably an adaptation to high parasitoid pressure and/or harsh abiotic con- ditions. The net reproductive rate indicates that this population could potentially increase 1.41 times per generation. The maximum rate of increase is one of the lowest ever reported (e.g., Pianka 1970) even when compared to other arachnids (e.g., scorpions from Polls & Farley 1980). This finding may reflect the high rate of egg mortality and low survivor- ship of females to first age of reproduction. Mortality. — Our sample produced a Type III survivorship curve. Many desert inhabi- tants have high mortality rates in the early stages of their life history (Polls & Yamashita 1991), which may be caused by a variety of desert stresses. We observed a number of predators preying on diguetid adults and eggs. Nuessly & Goeden (1984), on the other hand, observed molting only as a cause of mortality in the field. Prey analysis. — Spiders, on the whole, are usually characterized as generalist predators (Foelix 1996). Our results reinforce this gen- eralization. Diguetia mojavea consumed prey from 10 orders. Homoptera, Hymenoptera, and Coleoptera made up most (more than 88.7%) of D. mojavea s diet. This finding con- trasts Nuessly & Goeden’s (1984) report, who noted that the introduced biological control agent, a coleophorid moth, accounted for nearly 70% of D. mojavea'?, diet in Indio, Cal- ifornia; less than 3% of the diet in our study consisted of this moth. This discrepancy is probably to due to the fact that Indio has had several introductions of this moth for control of Salsola. Our site remains relatively undis- turbed from this introduction. Our analysis of prey contents in their sheet- webs is a first for diguetids. These prey had more biomass than those prey incorporated into the retreat but constituted a numerically minute (3%) amount of the total diet. This may be due to the inherent difficulty of cap- turing and/or handling larger prey. Previous findings. — Nuessly & Goeden’s (1984) paper is the only other study to ex- amine the natural history of D. mojavea. They noted the following characteristics: a one-year life cycle; a diet consisting largely of coleo- phorids and cicadellids; a significant positive correlation between number of prey and egg number; and observed mortality due only to molting with no direct evidence of predation. They did not calculate life history parameters of reproduction and survivorship. Several dif- ferences existed between our study and that of Nuessly & Goeden’s (1984). Their study was conducted at Indio, California for six months, a recently cultivated area populated by an in- vasive weed community. Our study was much longer (3.5 years) and was conducted on the floor of the Coachella Valley, which is a nat- ural, undisturbed area. ACKNOWLEDGMENTS This work would have been impossible without the hard work of Kenneth H. Sculteu- re and Sharon McCormick. We thank each very much for the companionship and help during many long hours in the field and lab. Financial assistance was provided by the Na- tional Science Foundation and Vanderbilt Uni- versity’s Natural Science Committee and Uni- versity Research Council. BOULTON & POLIS— NATURAL HISTORY OF DIGUETIA MOJAVEA 521 LITERATURE CITED Bentzien, M.M. 1973. Biology of the spider Di- guetia imperiosa (Araneida: Diguetidae). Pan- Pacific EntomoL, 49:110-123. Bristowe, W.S. 1958. The World of Spiders. Col- lins, London. Cazier, M.A. & M.A. Mortenson. 1962. Analysis of the habitat, web design, cocoon and egg sacs of the tube weaving spider Diguetia canities McCook (Araneae, Diguetidae). Bull. Southern California Acad. Sci., 61:65-88. Comstock, J.H. 1948. The Spider Book. Cornell Univ. Press, Ithaca, New York. Eberhard, W, 1967. Attack behavior of diguetid spiders and the origin of prey wrapping in spi- ders. Psyche, 74:173-181. Edney, E.B., S. Haynes, & D. Gibo. 1974. Distri- bution and activity of the desert cockroach Ar- enivaga investigata (Polyphagidae) in relation to microclimate. Ecology, 55:420-427. Foelix, R.E 1996. Biology of Spiders, 2nd ed. Ox- ford Univ. Press, Oxford. Gerschman de Pikelin, B. & R.D. Schiapelli. 1962. La familia Diguetidae (Araneae) en la Argentina. Physis, 23:205-208. Gertsch, W.J. 1949. American Spiders. D. Van Nostrand Company, New York. Gertsch, W.J. 1958. The spider family Diguetidae. American Mus. Nov., No. 904, Pp. 1-24. Gertsch, W.J. 1979. American Spiders. 2nd ed. D. Van Nostrand Company, New York. Hughes, P.R., H.A. Wood, J.P. Breen, S.E Simpson, A.J. Duggan, & J.A. Dybas. 1997. Enhanced bioactivity of recombinant baculoviruses ex- pressing insect-specific spider toxins in Lepidop- teran crop pests. J. Invert. Pathol., 69:112-118. Jackson, R.R. 1978. Life history of Phidippus johnsoni (Araneae, Salticidae). J. ArachnoL, 6: 1-29. Krapcho, K.J., R.M. Krai, Jr., B.C. Vanwagenen, K.G. Eppler & T.K. Morgan. 1995. Character- ization and cloning of insecticidal peptides from the primitive weaving spider Diguetia canities. Insect Biochem. Mol. Biol., 25:991-1000. Lopez, A. 1984. Some observations on the internal anatomy of Diguetia canities (McCook, 1890) (Araneae, Diguetidae). J. ArachnoL, 11:377-384. Nuessly, G.S. & R.D. Goeden. 1983. Spider pre- dation on Coleophora parthenica (Lepidoptera: Coleophoridae), a moth imported for the biolog- ical control of Russian thistle. Environ. Ento- mol., 12:1433-1438. Nuessly, G.S. & R.D. Goeden. 1984. Aspects of the biology and ecology of Diguetia mojavea Gertsch (Araneae, Diguetidae). J. ArachnoL, 12: 75-85. Pianka, E.R. 1970. On r- and K-selection. Ameri- can Nat., 104:592-597. Pianka, E.R. 1978. Evolutionary Ecology. 2nd ed. Harper & Row Publishers, New York. Pielou, E.C. 1977. Mathematical Ecology. 2nd ed. Wiley, New York. Platnick, N.I. 1989. A revision of the spider genus Segestrioides (Araneae, Diguetidae). American Mus. Nov., No. 2940, Pp. 1-9. Polls, G.A. 1988. Trophic and behavioral responses of desert scorpions to harsh environmental peri- ods. J. Arid Environ., 14:123-134. Polls, G.A., ed. 1991. The Ecology of Desert Com- munities. Univ. Arizona Press, Tucson, Arizona. Polls, G.A. & R.D. Farley. 1980. Population biol- ogy of a desert scorpion: survivorship, micro- habitat, and the evolution of life history strategy. Ecology, 61:620-629. Polls, G.A. & S.J. McCormick. 1986. Scorpions, spiders, and solpugids: Predation and competi- tion among distantly related taxa. Oecologia, 7 1 : 111-116. Polls, G.A. & T. Yamashita. 1991. The ecology and importance of predaceous arthropods in desert communities. Pp. 180-222, In The Ecology of Desert Communities. (G.A. Polls, ed.). Univ. Ar- izona Press, Tucson, Arizona. Manuscript received 30 April 1998, revised 30 Oc- tober 1998. 1999. The Journal of Arachnology 27:522-530 HOST SPECIFICITY AND DISTRIBUTION OF THE KLEPTOBIOTIC SPIDER ARGYRODES ANTIPODIANUS (ARANEAE, THERIDIIDAE) ON ORB WEBS IN QUEENSLAND, AUSTRALIA Paul Grostal and David Evans Walter: Department of Entomology, The University of Queensland, St. Lucia, Queensland 4072, Australia ABSTRACT. We investigated host specificity, the effects of host size, and the effects of the size, structure and occupancy of host webs on the abundance of the kleptobiotic spider Argyrodes antipodianus O.R -Cambridge 1880. The kleptobiont is not host specific, but does prefer orb webs that are surrounded by a scaffold of threads (barrier- web). Across all hosts, host size had little effect on the abundance of the kleptobiont, while host density and the presence of other species of Argyrodes on webs had no effect. Web diameter, although not strongly related to the abundance of A. antipodianus in the field, limited kleptobiont numbers in greenhouse experiments. On webs of the Golden Orb Spider, Nephila plumipes (Latreille 1804), numbers of A. antipodianus were not affected by size of the scaffold or by aggregation of host webs. However, presence of host males was associated with a significantly higher abundance of A. antipodianus, suggesting that these kleptoparasites may take advantage of distracted females and impose a cost on mating in N. plumipes. Many spiders of the genus Argyrodes Si- mon 1864 live in close association with web- building spiders, and remove and feed on prey items captured in the webs of their hosts. These small web visitors are referred to as “kleptoparasites’" or “kleptobionts” (Vollrath 1984, 1987; Elgar 1993). Although observa- tions of their unusual foraging behavior are relatively common, little is known of the mechanisms that influence the infestation lev- els of kleptobiotic Argyrodes on host webs. Abundance and diversity of Argyrodes on webs may vary considerably among and with- in host species (Kaston 1965; Levi 1978, 1985; Whitehouse 1988; Elgar 1989), with up to 46 individuals and up to 3 species found on a single web (Exliee & Levi 1962; Vollrath 1981). The abundance of these kleptobionts may be influenced by a range of factors such as prey availability, weather, host behavior or web characteristics (Robinson & Robinson 1973; Smith-Trail 1981; Vollrath 1984; Larch- er & Wise 1985; Vollrath 1987; Whitehouse 1988; Elgar 1989; Cangialosi 1990a, b; Whitehouse & Jackson 1993; Elgar 1993). In- fluential web characteristics could include size, architecture (e.g., relative size of web scaffold), abundance or aggregation of host webs (Whitehouse 1988; Elgar 1993). Inter- actions with the host or other web “visitors,” such as host males or other kleptobionts also could influence web colonization (Vollrath 1984, 1987; Grostal & Walter 1997). Argyrodes antipodianus O.P.-Cambridge 1880 is an abundant kleptobiont on the webs of orb weaving spiders in southeast Queens- land This spider is a relatively small-bodied species {ca. 3 mm long), that is easily recog- nized in Australia by its conical, bright silver abdomen (Grostal, in press). Argyrodes anti- podianus is associated with at least ten host species that build four different types of web (orb, funnel, tangle and space), but in New Zealand the kleptobiont is most common on the non-cribellate, sticky orb webs of Erio- phora pustulosa (Walckenaer 1841) (White- house 1988; Elgar 1993). Consequently, Whitehouse (1988) refers to A. antipodianus as a host specialist (sensu Vollrath 1984). In this paper we used field surveys to ex- amine the host range of A. antipodianus, and to investigate how the abundance of this spi- der is influenced by the architecture, size and relative abundance of host webs, and presence of other species of Argyrodes on these webs. We thee examined A. antipodianus on webs of one of its common hosts, the Golden Orb 522 GROSTAL & WALTER— HOST SPECIFICITY OF ARGYRODES 523 1 2 3 Figures 1-3. — ^Three types of orb web sampled during four surveys in eastern Queensland: 1. Orb only (e.g., Eriophora transmarina), frontal aspect; 2. Orb and barrier (e.g., Nephila plumipes), fronto-lateral aspect; 3. Orb and tangle (e.g., Cyrtophora moluccensis), lateral aspect. Spider, Nephila plumipes (Latreille 1804) and determined the influence of the relative size of barrier-web, web aggregation and the num- ber of host males on kleptobiont numbers. Fi- nally, we used greenhouse experiments to es- tablish the effect of orb size of N. plumipes on the retention of A. antipodianus on webs. We predicted that the kleptobionts would be positively associated with web size and web aggregation, but negatively associated with numbers of other kleptobionts and of host males. METHODS Host range and abundance of A. antipo- dianus.-— We conducted four surveys during 1995 in eastern Queensland: two surveys in the south east (Pinkenba 27°25'S, 153°07^E and Everton Park, Brisbane, 27°25'S, 152°59'E), one on the central coast (Yeppoon, 23°07'S, 150°44'E) and one in the far north (Cairns 16°53'S, 145°45'E). The Everton Park site (area = 2500 m^) was surveyed during October and was dominated by a semi-closed dry sclerophyll forest. Pinkenba (area = 16,000 mQ, surveyed in August, consisted of an open stand of casuarina. The site at Yep- poon (area = 12,000 m^) was censussed in May and consisted of an open palm forest, while the one in Cairns (area = 3,920 m^) was a closed rainforest thicket, and was examined in August. The month and location of the sur- veys depended on the opportunity to visit the sites. We searched each site for orb webs that were located up to 200 cm above ground level and were over 9 cm in diameter. Spiders that constructed smaller webs were often juvenile and thus difficult to identify. For each web we collected the following data: species of the web builder, the spider’s body length (cepha- lothorax and abdomen, measured with a clear ruler to nearest mm), the diameter of the orb (to nearest cm), and the number and species of Argyrodes on the web. The webs were di- vided into three categories based on their ar- chitecture: orb only, orb with barrier, and orb with tangle (Fig. 1). A barrier is a three-di- mensional scaffold of non-sticky threads in front of and behind an orb (Fig. 2). A tangle consisted of a dense tent-like scaffold (Fig. 3) that extended above and below the orb. Tan- gles were more complex and larger relative to orb size, than barriers. All Argyrodes species were collected and preserved in 80% ethanol for later identification. A sample collection of the spiders was deposited with the Queensland Museum (Brisbane, Australia). Abundance of A, antipodianus on webs of A. plumipes. — We conducted two additional surveys on separate plots at Pinkenba (one during April, the other during May 1995). In the first survey we sampled the webs of adult N. plumipes only, and in the second survey we examined webs of all stages of N. plumi- pes. The plots were adjacent to the one pre- viously sampled for a range of different hosts 524 THE JOURNAL OF ARACHNOLOGY (see above) and consisted of an open stand of casuarina. For both surveys, we searched each site for webs of N. plumipes up to a height of 2.5 m, using a stepladder for webs above 2 m. All data were collected as in the previous sur- veys, except that we used carapace width to measure host size (a more precise measure at the intraspecific level; Higgins & Rankin 1996). Additionally, we recorded the presence of visiting host males on webs and we qualita- tively assigned the webs into five categories, based the complexity of the barrier (0 = no barrier; 4 = most complex barrier). We cate- gorized barrier complexity by visually com- paring the size of the orb relative to the vol- ume occupied by the barrier threads and their density (no. of threads/volume). To check the accuracy of this estimation, we collected 10 clean webs for each of the categories 1 to 4 (webs with barriers present). We used dis- secting scissors to separate orbs from barriers during collection. Then, for each web, we cleaned the silk from any debris and separate- ly weighed orbs and barriers with an electron- ic balance in the laboratory. We used these results to calculate the mean ratio (± SE) of barrier weight : orb weight for each category and to check if the categories are discrete (i.e., if the means significantly differ). Finally, for the survey of adult N. plumipes we recorded whether host webs were aggre- gated or not. A web was ranked as aggregated if its threads overlapped or interlocked with those of another web (Elgar 1989). Aggrega- tions containing webs of immature hosts were excluded from the sample. Retention of A. antipodianus on webs of N. plumipes. — The experiments were con- ducted in a ventilated greenhouse (Brisbane, September 1995). We used female N. plumi- pes of two age groups: juveniles (10-11 mm long) and adults (27-32 mm long), but only adult females of A. antipodianus. The spiders were housed in large cages (170 X 170 X 170 cm) which were covered with a fine plastic mesh. Four wooden racks, composed of a cen- tral rod (165 cm high) with four arms, were placed in the comers of each cage to provide support for webs spun by host spiders (Grostal & Walter 1997). Eight cages were used for each experiment, which was repeated six times over 18 days. One N. plumipes was placed in each cage 48 hours before each trial and allowed to spin a web. Four adult and four juvenile N. plumipes were used for each experiment. First, we ran- domly removed four hosts (two juveniles and two adults) from their webs. Care was taken not to damage the web while removing the spiders. Thus, each test consisted of four webs of adult N. plumipes: two with hosts included and two with hosts removed, and four webs of juveniles: two with hosts present, and two with hosts removed. Ten A. antipodianus were then placed on each web. After 24 h we re- corded the number of A. antipodianus that re- mained on the webs. Statistical analysis. — For surveys of host range and abundance of A. antipodianus, we used linear regression to estimate the relation- ship of the number of A. antipodianus per web with: 1) host body length; and 2) diameter of host web. The effect of presence of other Ar- gyrodes species (+/— ) on webs on the mean number of A. antipodianus per web was ana- lyzed using single-factor ANOVA. Prior to the analysis, data were log-transformed for nor- mality. Data from all four sites were pooled for the above analyses. Finally, we calculated the mean number of A. antipodianus per web for each host species, on every site {n = 27). Then, we regressed these means against the density of the corresponding host species at a given site (no. individuals/ 10,000 m^, see Ta- ble 1). We examined all regression data with scatterplots to check for non-linear relation- ships. For surveys of N. plumipes, we regressed the number of A. antipodianus per web against width of host carapace and diameter of host web. The effects of: 1) aggregation of host webs (+/—); 2) presence of male N. plu- mipes (+/—); and 3) the rank of web barrier (0, 1, 2, 3, 4) on the abundance of A. anti- podianus (number per web) were analyzed separately with single-factor ANOVA. For the greenhouse experiments, we compared the numbers of A. antipodianus retained on webs that were spun by juvenile and adult N. plu- mipes, with and without the hosts, using a two-way ANOVA. All data were normalized by log-transformation before analysis. RESULTS Host range and abundance of A. antipo- dianus.— A total of 744 webs was examined GROSTAL & WALTER— HOST SPECHTCITY OF ARGYRODES 525 Table 1. — Average body length (mm) of host spiders (cephalothorax + abdomen), density (no./10,000 m^) of host webs sampled and the average number of Argyrodes antipodianus on three types of host web (orb only, orb and barrier, orb and tangle) at four sites in coastal Queensland: Everton Park, Pinkenba (both in south-east), Yeppoon (central-east) and Cairns (far north). Values are totals or means ± standard errors. Site/Web type/Host Host length (No. webs/10,000 m^) A. antipodianus per web Everton Park Orb only Araneus dimidiatus 7.7 ± 1.0 (156) 0.2 ± 0.1 Argiope sp. 6.9 ± 2.3 (16) 0 Eriophora transmarina 12.7 ± 2.1 (24) 0.3 ± 0.2 Leucauge sp. 6.7 ± 1.4 (36) 0 Orb & Barrier Nephila plumipes 12.9 ± 4.1 (292) 2.7 ± 0.3 Orb & Tangle Cyrtophora hirta L. Koch 1872 6.0 (4) 0 Cyrtophora moluccensis (Doleschall 1857) 10.9 ± 4.1 (84) 0.2 ± 0.2 Pinkenba Orb only Araneus eburnus 4.0 ± 0.7 (1) 0 Argiope sp. 5.5 (1) 1.0 Eriophora transmarina 4.9 ± 0.9 (68) 0.02 ± 0.01 Leucauge sp. 5.9 ± 1.2 (22) 0 Orb & Barrier Nephila plumipes 10.7 ± 3.6 (116) 5.9 ± 0.3 Orb & Tangle Cyrtophora hirta 4.0 ± 2.1 (3) 2.0 ± 1.7 Cyrtophora moluccensis 8.3 ± 4.1 (5) 6.0 ± 1.7 Yeppoon Orb only Araneus dimidiatus 5.4 ± 1.1 (36) 0.1 ± 0.1 Gasteracantha sp. 4.3 ± 0.9 (10) 0 Orb & Barrier Nephila pilipes (Fabricius 1793) 21.3 ± 11.5 (7) 0.6 ± 0.4 Nephila plumipes 16.5 (1) 0 Orb & Tangle Cyrtophora sp. 6.3 ± 1.2 (8) 0 Cairns Orb only Araneus dimidiatus 4.9 ± 0.8 (311) 0.2 ± 0.1 Argiope sp. 10.7 ± 2.9 (33) 0.3 ± 0.2 Eriophora transmarina 4.4 ± 0.5 (13) 0 Gasteracantha sp. 5.9 ± 1.8 (20) 0 Leucauge sp. 4.8 ± 1.2 (20) 0 Orb & Barrier Nephila pilipes 12.8 ± 10.9 (20) 0.4 ± 0.3 Nephiiengys sp. 5.6 ± 1.8 (13) 0 Orb & Tangle Cyrtophora sp. o b 1+ b 3.5 ± 1.5 526 THE JOURNAL OF ARACHNOLOGY Table 2. — Average number of individuals of each species of Argyrodes per web (± standard error) at Everton park (south-east Queensland), Pinkenba (south-east Queensland), Yeppoon (central-east Queens- land) and Cairns (far north Queensland). Species Everton Park Pinkenba Yeppoon Cairns A. antipodianus 1.40 ± 0.19 3.36 ± 0.24 0.12 ± 0.05 0.19 ± 0.05 A. rainbowi 0.13 ± 0.04 0.03 ± 0.01 — — A. species 1 0.22 ± 0.07 0.03 ± 0.01 0.01 ± 0.01 — A. fissifrons 0.09 ± 0.04 — — — A. miniaceus — — 0.22 ± 0.12 0.25 ± 0.08 A. kulczynski — — 0.24 ± 0.08 0.01 ± 0.01 in the four surveys, and A. antipodianus was associated with eight of the 12 host species sampled (Table 1). Only webs of Araneus eburnus (Keyserling 1886), Gasteracantha sp., Leucauge sp. and Nephilengys sp. had no A. antipodianus; however, for some of these hosts (e.g., Nephilengys sp.) very few webs were found (Table 1). Webs of Araneus dim- idiatus (L. Koch 1871), present at every site except Pinkenba, consistently had low num- bers of A. antipodianus, in spite of the high abundance of this host species (Table 1). Sim- ilarly, webs of Eriophora transmarina (Key- serling 1865), although relatively common in the southeastern Queensland sites, had very few A. antipodianus (Table 1). There was no apparent linear relationship between density of hosts (no. per 10,000 m^) and the mean number of A. antipodianus per web across all four sites {R^ = 0.02; = 0.53, P - 0.47), although spiders belonging to Nephila spp. and Cyrtophora spp., were clearly the pre- ferred hosts (Table 1). At Everton Park and Pinkenba, A. antipo- dianus was over six times more abundant than any other species of Argyrodes; however, in tropical Queensland (Cairns and Yeppoon) other species of Argyrodes were more abun- dant (Table 2). In Cairns, A. miniaceus (Do- leschall 1857) was more numerous, while Yeppoon was dominated by A. miniaceus and A. kulczynski (Roewer 1942). Three additional species of Argyrodes (A. fissifrons O.P.-Cam- bridge 1869, A. rainbowi (Roewer 1942) and Argyrodes sp. 1) were also collected. In the presence of other species of Argyrodes, the abundance of A. antipodianus (2.6 ± 0.5 spi- ders per web) was somewhat higher than that on webs with no congeners, although the dif- ference was not significant (1.8 ± 0.1 spiders per web; ANOVA: = 2.83, P = 0.093). Both body length and orb diameter of host spiders showed a positive linear relationship with the numbers of A. antipodianus. When data from all surveys were pooled, host length accounted for 28% of the variance in A. an- tipodianus numbers (Fj^ 742 ^ 286.5, P < 0.0001). Orb diameter seemed to impose an upper limit on the numbers of the kleptobiont (Fig. 4: broken line), although the two vari- ables were not strongly related {R} ~ 0.13; F, 745 = 107.3, P < 0.0001). Orb diameter also showed a positive relationship with the body length of hosts {R^ = 0.60; F, 742 == 1118.6, P < 0.0001). Species of Nephila and Nephilengys con- struct webs that consist of a vertical orb and a non-viscid barrier (Fig. 2). Cyrtophora spp. make non-viscid, horizontal orbs with an ex- tensive tangle (Levi 1978; Shear 1994; Fig. 3). Spiders that construct webs consisting almost exclusively of a catching orb with little or no barrier include Araneus dimidiatus, A. ebur- nus, Argiope sp., Eriophora transmarina, Gasteracantha sp. and Leucauge sp.. Architecture of host webs (Fig. 1) influ- enced the abundance of A. antipodianus (AN- OVA: Fi,745 = 217.04, P < 0.0001). When all data were pooled, webs containing orbs with barriers had the highest numbers of the klep- tobiont (mean of 4.6 ± 0.3/web). Generally, orbs with a tangle had intermediate numbers of A. antipodianus (1.5 ± 0.5/web) and webs that consisted only of orbs had the lowest numbers (0.09 ± 0.02/web). In Cairns webs with orb and tangle had the most A. antipo- dianus, although these results applied only to two individuals of an unidentified species of Cyrtophora. Abundance of A. antipodianus on webs of N. plumipes, — At Pinkenba we examined a total of 299 webs in the survey of all stages GROSTAL & WALTER— HOST SPECIHCITY OF ARGYRODES 527 Figure 4. — Numbers of Argyrodes antipodianus per web against orb diameter (cm) of host webs at Everton Park, Pinkenba, Yeppoon and Cairns. of N. plumipes and 213 webs in the survey of adult N. plumipes. There was an average of 5.7 ± 0.3 A. antipodianus per web in the for- mer and 2.3 ± 0.1 in the latter census. Num- bers of A. antipodianus were positively related with the width of host carapace, although, as in the across-species comparison, this rela- tionship was not strong for all host stages {R? = 0.27; Fi 297 "" 109.7, P < 0.0001) or adult hosts {R^ - 0.15; = 36.18, P < 0.0001). Orb diameter of N. plumipes showed a similar pattern of relation with the abundance of A. antipodianus (all hosts R} = 0.22, Fj 297 85.95, P < 0.0001; adult hosts ^2 = 0.12, Fi,2ii 28.44, P < 0.0001). Our visual estimation of the complexity of the barrier (categories 0 to 4) was sufficiently accurate, since the ratios of orb weight/barrier weight (± SE, n == 10) differed between cat- egories (category 1 ^0.1 ± 0.02; 2 - 0.3 ± 0.1; 3 — 1.1 ± 0.4; 4 — 1.9 ± 0.5). Never- theless, barrier complexity did not have an ef- fect on A. antipodianus in either survey (AN- OVA, all webs F4, 239 = 1-44, P = 0.221; adults only F3 209 ^ 0.36, P = 0.786). Aggregation (±, recorded only for adult webs) did not affect the numbers of A. anti- podianus (ANOVA, F,,2ii = 2.13, P = 0.146). However, abundance of these kleptobionts was over 65% higher on webs that had male N. plumipes. This result was highly significant for the survey of all stages of N. plumipes, with 9.0 ± 1.0 A. antipodianus on webs with males '{n = 24), and 5.5 ± 0.3 kleptobionts on webs without males {n = 275; ANOVA: ^1,297 ” 15.78, P < 0.001) and for the census of adult hosts (Fig. 5; ANOVA: Fj 211 = 7.96, P - 0.005). Retention of A. antipodianus on webs of N. plumipes. — On average, after 24 hours, large webs (32 ± 4 cm diameter) built by adult N. plumipes retained over 85% more A. antipodianus than small webs (18 ± 3 cm di- ameter), built by juvenile hosts (ANOVA, P < 0.0001; Fig. 6, Table 3). However, the pres- ence of hosts on webs was of no consequence to the kleptobiont (ANOVA, P = 0.895; Table 3). When juvenile hosts were excluded, six of the twelve webs were destroyed or damaged by more than 30% by A. antipodianus (pers. obs.), and were not included in the analysis. Webs of adult N. plumipes did not differ in shape or architecture from those built by the juveniles. DISCUSSION Kleptobiotic Argyrodes may be found on a range of webs (Kaston 1965; Elgar 1993), al- though they are likely to be more abundant on webs that are easy to forage on, supply suf- ficient food and provide ample refuge. White- house (1988) found that in New Zealand A. antipodianus specialized on a single host spe- cies, Eriophora pustulosa, in whose webs it foraged most efficiently. We have unpublished data that is consistent with Whitehouse Table 3.- — ^Two-way ANOVA for the effect of web size and presence of N. plumipes on the numbers of A. antipodianus retained on webs after 24 hours in the greenhouse (initial number of kleptobionts per web = 10). Category df MS F P Web size 1 0.903 51.00 <0.0001 Host presence 1 0.0003 0.02 0.895 Web size * host pres. 1 0.021 1.19 0.282 Residual 38 0.02 528 THE JOURNAL OF ARACHNOLOGY MALES + MALES - Figure 5. — Average number of Argydoes anti- podianus per web on webs of adult Nephila plu- mipes females that included (males L), or did not include male hosts (males — ). (1988), i.e., other webs such as tangle webs constructed by theridiids (e.g., Latrodectus spp.) or space webs made by amaurobids (e.g., Badumna spp.) were rarely colonized by A. antipodianus (RG., pers. obs.). However, we also found that the kleptobiont has a broad host distribution, perhaps because our sam- pling areas had a higher diversity of web spi- ders than those examined by Whitehouse (El- gar 1993). Additionally, we found some evidence of web specificity by this klepto- biont, as it was found primarily on orb webs that included a scaffold (barrier or tangle): those of Nephila and Cyrtophora species. Elgar (1993) pointed out that host specific- ity is likely to vary continuously and can be influenced by the abundance and diversity of hosts. Our data show that relative abundance of webs of each host species was not signifi- cantly correlated with the abundance of A. an- tipodianus. However, availability of hosts probably does affect host choice by the klep- tobiont. For example, Eriophora pustulosa were the preferred hosts in New Zealand dur- ing summer (Whitehouse 1988), but these were the only orb weavers present in the study site. On the other hand, in the presence of more complex orb webs (with scaffold) in Queensland, orb weavers that construct simple orb webs similar to E. pustulosa had few or no A. antipodianus. If kleptobiotic Argyrodes have a negative effect on their hosts, then web characteristics that favor them will carry a disadvantage to □ Occupied webs, adult host Ei3 Unoccupied web, adult host B Occupied webs, juvenile host Figure 6. — ^Effects of web size and host presence on the retention of Argyrodes antipodianus (number remaining after 24 hours) on webs of Nephila plu- mipes in the greenhouse. Data presented as means + standard error. the web owner and may be under conflicting selective pressures (e.g., larger webs might catch more food, but may also increase the kleptobiont load). Elgar (1989) found that the intensity of infestation of Nephila edulis (La- billardiere 1799) webs by A. antipodianus was correlated with host size. Our data show a positive correlation between the number of these kleptobionts and both host size and orb diameter, although little of the variance is ex- plained (13-28%). This may be because host size or orb diameter is not always clear indi- cators of web size for spiders that construct webs of varying architecture. For instance, while orbs built by Cyrtophora (orb & tangle) are small, those of Nephila (orb & barrier) are large relative to total web space (Figs. 2, 3). Further, our results could have been confound- ed by survey site, season and species of host, which were all pooled. However, the correla- tion did not improve when we controlled for variation in web architecture, site and host species by using only N. plumipes. Neverthe- less, orb size may impose an upper limit on the numbers of A. antipodianus on host webs: large orbs may accommodate few or many A. antipodianus, but small orbs contain only few kleptobionts. This was supported by our data GROSTAL & WALTER— HOST SPEdHCITY OF ARGYRODES 529 from the greenhouse, which showed that in- dependent of host presence, small webs (ju- venile N. piumipes) retain fewer A. antipodi- anus than large webs (adult hosts). Although we examined 512 webs in our surveys of N. piumipes, we did not find an obvious effect of web architecture or aggre- gation on the numbers of A. antipodianus; and perhaps the distribution of this kleptobiont is more random than previously hypothesized (Elgar 1989). However, apart from the struc- tural characteristics of webs that we measured, several other factors may directly influence the number of A. antipodianus on webs. These could include the web tenacity of hosts (Levi 1978), food abundance and quality, host be- havior and environmental factors, all of which ought to be examined in future studies. Contrary to our hypothesis that other web visitors might have a damping effect on num- bers of A. antipodianus, the presence of other Argyrodes had no significant effect. Also, sur- prisingly, male hosts were associated with greatly elevated numbers of this kleptobiont, as there were two-thirds more A. antipodianus on webs of female N. piumipes colonized by males, than on webs with no males. We offer two alternative hypotheses to explain this un- expected result. First, both males and klepto- bionts may be responding to the same factors, e.g., food availability, position in wind corri- dors or insolation. Also, pheromones emitted by female hosts can be perceived not only by the males, but perhaps also by the klepto- bionts, consequently facilitating web location. Second, the activity of Nephila males (includ- ing feeding and mating attempts) may be ben- eficial to A. antipodianus through disturbance of the web and distraction of the female host. Thus, with males present, A. antipodianus would face lower levels of aggressive re- sponse by the web owner, and perhaps have a higher foraging success, thus remaining on the web longer. Possibly, A. antipodianus engages in “smokescreening” behavior (Wilcox et al. 1996) by increasing its feeding while female hosts are distracted. If mating attempts of male M piumipes cause higher infestation lev- els of kleptobionts, then reproduction of this host may come at a previously unnoticed cost ACKNOWLEDGMENTS We would like to thank the Federal Airports Corporation for permission to collect on their lands. We also thank Karen Caegialosi, Lin- den Higgins, Liza Miller, Heather Proctor and Owen Seeman for suggestions that improved the manuscript The research was funded in part by the Australian Postgraduate Award and by the Department of Entomology, Uni- versity of Queensland. LITERATURE CITED Christenson, T.E. & K.C. Goist. 1979. Costs and benefits of male-male competition in the orb weaving spider, Nephila clavipes. Behav. EcoL Sociobiol., 5:87-92. Cangialosi, K.R. 1990a. Kleptoparasitism in colo- nies of the social spider Anelosimus eximius (Ar- aneae: Theridiidae). Acta Zool. Fennici, 190:51- 54. Cangialosi, K.R. 1990b. Social spider defense against kleptoparasitism. Behav. Ecol. Socio- biol., 27:49-54. Elgar, M.A. 1989. Kleptoparasitism: A cost of ag- gregating for an orb-weaving spider. Anim. Be- hav., 37(6): 1052-1055. Elgar, M.A. 1993. Inter-specific associations in- volving spiders: Kleptoparasitism, mimicry and mutualism. Mem. Queensland Mus., 33(2):411- 430. Grostal, P. in press. Five species of kleptobiotic Argyrodes Simon (Araneae: Theridiidae) from eastern Australia: Systematics and ecology. Mem. Queensland Mus. Grostal, P. & D.E. Walter. 1997. Kleptoparasites or commensals? Effects of Argyrodes antipodianus (Araneae: Theridiidae) on Nephila piumipes (Ar- aneae: Tetragnathidae). Oecologia, 111:570-574. Higgins, L.E. & M.A. Rankin. 1996. Different pathways in arthropod postembryonic develop- ment. Evolution, 50(2):573-82. Kaston, B.J. 1965. Some little known aspects of spider behavior. American Midi. Nat., 73(2): 336-356. Larcher, S.F. & D.H. Wise. 1985. Experimental studies of the interactions between a web-invad- ing spider and two host species. J. ArachnoL, 13: 43-59. Levi, H.W. 1978. Orb- webs and phylogeny of orb- weavers. Symp. Zool. Soc. London, 42:1-15, Levy, G. 1985. Spiders of the genera Episinus, Ar- gyrodes and Coscinida from Israel, with addi- tional notes on Theridion (Araneae: Theridiidae). J. Zool. (A), 207:87-123. Lubin, Y.D. 1974. Adaptive advantages and the evolution of colony formation in Cyrtophora (Araneae: Araneidae). Zool. J. Linn. Soc., 54: 321-339. Robinson, M.H. & B. Robinson. 1973. Ecology and behaviour of the giant wood spider Nephila maculata (Fabricius) in New Guinea. Smiths. Contr. ZooL, 149:1-76. 530 THE JOURNAL OF ARACHNOLOGY Rypstra, A.L 1985 Aggregations of Nephila cla- vipes (L.) (Araneae, Araneidae) in relation to prey availability. J. ArachnoL, 13:71-78. Shear, W.A. 1994. Untangling the evolution of the web. American Scientist, 82:256-266. Smith-Trail, D. 1981. Predation by Argyrodes (Theridiidae) on solitary and communal spiders. Psyche, 8:349-355. Vollrath, F. 1981. Energetic considerations of a spi- der parasite-spider host system. Rev. ArachnoL, 3(2):37-44. Vollrath, F. 1984, Kleptobiotic interactions in in- vertebrates. Pp. 61-94, In Producers and Scroungers: Strategies of Exploitation and Para- sitism. (C.J. Barnard, ed.). Chapman and Hall, New York. Vollrath, F. 1987, Kleptobiosis in spiders. Pp. 274- 286, In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer Verlag, Berlin. Whitehouse, M.E.A. 1988. Factors influencing specificity and choice of host in Argyrodes an- tipodiana (Araneae, Theridiidae). J, ArachnoL, 16:349-355. Whitehouse, M.E.A. & R.R. Jackson. 1993. Group structure and time budgets of Argyrodes antipo- diana (Araneae, Theridiidae), a kleptoparasitic spider from New Zealand. New Zealand J. ZooL, 21:253-268. Wilcox, R.S., R.R. Jackson & K. Gentile. 1996. Spiderweb smokescreens: spider trickster uses background noise to mask stalking movements. Anim. Behav., 5 1(2):3 13-326. Manuscript received 14 February 1997, revised 1 October 1998. 1999. The Journal of Arachnology 27:531-538 ABUNDANCE OF SPIDERS AND INSECT PREDATORS ON GRAPES IN CENTRAL CALIFORNIA Michael J. Costello^: University of California Cooperative Extension, 1720 South Maple Ave., Fresno, California 93702 USA Kent M* Daane: Center for Biological Control, Division of Insect Biology, Department of Environmental Science, Policy and Management, University of California, Berkeley, California '94720 USA ABSTRACT We compared the abundance of spiders and predaceous insects in five central California vineyards. Spiders constituted 98.1% of all predators collected. More than 90% of all spiders collected were from eight species of spiders, representing six families. Two theridiids (Theridion diiutum and T. melanurum) were the most abundant, followed by a miturgid (Cheiracanthium inclusum) and an agelinid {Hololena nedrd). Predaceous insects comprised 1.6% of all predators collected, and were represented by six genera in five families. Nabis americoferis (Heteroptera, Nabidae) was the most common predaceous insect, with its densities highest late in the growing season. Chrysoperla carnea, Chrysoperla comanche and Chrysopa oculata (Neuroptera, Chrysopidae) and Hippodamia convergens (Coleoptera, Coccinellidae) were most abundant early in the season. The donunance of spiders may be due to their more stable position in the vineyard predator community compared to predaceous insects. We also suggest that the low per- centage of predaceous insects (e.g., lacewings) may reflect the lack of preferred prey (e.g., aphids) on grapevines. Spiders are important predators in agroe- cosystems (reviews in Nyfeller & Benz 1987; Nyfeller et aL 1994). Many researchers have provided descriptions of spider species abun- dance or composition in a variety of agroe- cosystems (e.g.. Bishop 1980; Dean et al. 1982; Agnew & Smith 1989; Bardwell & Av- erill 1997; Wisniewska & Prokopy 1997). Other researchers have provided qualitative observations on the abundance of spiders (Carroll & Hoyt 1984) or recorded spider pre- dation events (Reichert & Bishop 1990; Ny- feller et al. 1992). However, it is less common for researchers to compare spider abundance to that of predaceous insects. Those studies that have analyzed the relative abundance of all predaceous arthropods vary considerably in the presentation of the data. For example, MacLellan (1973) reported on predaceous ar- thropods collected on apples in southeastern Australia, presenting numbers of spiders col- lected by size and numbers of predaceous in- sects collected by family. Plagens (1983) re- * Current address: Costello Agricultural Research & Consulting, P.O. Box 165, Tollhouse, California 93667 USA. ported population densities of the most abundant spiders {Misumenops spp.) found on Arizona cotton, presenting predaceous insects as overall percentages but not itemizing for different taxonomic groups. In these publica- tions, the amount of detail presented reflects the focus of the research, depending in part upon the breadth of the predator taxon being studied. More commonly, researchers present more detailed descriptions of the predaceous insect fauna, while spiders are grouped to- gether and data presented as an overall mean, numerical rank or percentage of the number collected (e.g., Roach 1980; Knutson & Gil- strap 1989; Royer & Walgenbach 1991; Bra- man & Pendley 1993). Few studies have pro- vided equivalent comparisons of spiders and predaceous insects at the genus or species lev- el (but see Breene et al. 1989). In vineyards, several researchers have cat- aloged the abundance of predaceous arthro- pods on grapevines. In southern Germany, Buchholz & Schnift (1994) presented num- bers of predaceous insects by family, identi- fying salticids to species and thomisids to ge- nus, but leaving most spiders unidentified. In California vineyards, spider species composi- 531 532 THE JOURNAL OF ARACHNOLOGY tion, relative abundance and seasonal occur- rence were described by Costello & Daane (1995) and Roltsch et al. (1998), but neither study included data on predaceous insects. Here, we present data that compare the rela- tive abundance of spiders to predaceous in- sects on grapevines in California’s central val- ley. METHODS Study sites. — The data presented are from five central valley vineyards that were sam- pled from 1995-1997. Grapevine cultivar and cultural practices varied among the sites. In 1995, three vineyards in Fresno County were sampled: a raisin vineyard {cultivar “Thomp- son Seedless” near Del Rey, California) a ta- ble grape vineyard {cultivar “Ruby Seedless” near Reedley, California) and a juice vineyard {cv “Thompson Seedless” near Parlier, Cali- fornia). In 1995 and 1996, a winegrape vine- yard in San Joaquin County (cv “Cabernet Sauvignon” near Woodbridge, California) and, in 1996 and 1997, a juice vineyard in Madera County were sampled (cv “Thompson Seedless” near Ripperdan, California). These sites were part of studies designed to deter- mine the impact of cover crops on vineyard insect pests and their natural enemies (see Costello & Daane 1998b; Daane & Costello 1998). All of the study sites were bordered by cultivated vineyards or orchards. In each year, all vineyards received multiple applications of sulfur for control of powdery mildew, Uncinula necator Burrill, and one or two applications of cryolite (sodium alumi- nofluoride) for control of omnivorous leaf- roller, Platynota stultana Walshingham 1884 (Lepidoptera, Tortricidae), and grapeleaf fold- er, Desmia funeralis (Hiibner 1796) (Lepidop- tera, Pyralidae). Sampling. — Costello & Daane (1997) pro- vide a detailed description of sampling meth- ods. In brief, the Del Rey, Ripperdan, Parlier and Woodbridge vineyards were sampled by shaking a 0.89 m^ section of vine foliage into a funnel shaped collector, and the Reedley vineyard was sampled by shaking the foliage of two grapevines onto a drop cloth and col- lecting all predators with small battery-pow- ered vacuums. Samples were taken monthly, from May to September, except for the Rip- perdan vineyard in 1996, which was sampled from July to September. On each sampling date, samples were taken between 0700-1200 h PDT. Samples from the replicated cover crop studies were pooled across treatments and sample dates. A total of 100 samples was taken from the Reedley vineyard, 180 from the Del Rey vineyard and 120 from the Parlier vineyard (one season each). A total of 243 samples was taken from the Ripperdan vine- yard and 360 from the Woodbridge vineyard (two seasons each). Voucher specimens were deposited at the Essig Museum at the Univer- sity of California at Berkeley. For each vineyard and sampling method, means were transformed to numbers of pred- ators per vine. Seasonal abundance of spiders and predaceous insects were plotted against cumulative degree days above 10 °C (the low- er developmental threshold for grapevines) from 1 January, for each sample year. RESULTS We collected a total of 13,348 spiders (2781 at Del Rey, 6468 at Woodbridge, 1273 at Rip- perdan, 679 at Parlier and 2147 at Reedley) and 219 predaceous insects (36 at Del Rey, 122 at Woodbridge, 6 at Ripperdan, 43 at Par- lier and 12 at Reedley). Over all sites, spiders constituted 98.1% of all predators collected, whereas the insect predators comprised just 1.6% of total predators. At individual sites, spiders comprised at least 94% of predators collected, with the highest percentage at Rip- perdan (99.5%) and the lowest at Parlier (94.0%) (Table 1). Predaceous insects com- prised 6.0% or less of all predators at each site, the highest percentage found at Parlier (5.9%) and the lowest at Ripperdan and Reed- ley (0.5%) (Table 1). The only other arthropod predator collected was Anystis agilis (Banks 1915) (Acari, Anystidae), a predaceous mite that feeds on insects as opposed to spider mites. Only 17 Anystis agilis were collected, all at the Reedley site, comprising 1.5% of the predators collected there. Spiders. — Eight species from six families constituted >90% of all spiders collected. By family, these were: (1) Miturgidae: Cheira- canthium inclusum (Hentz 1847); (2) Corin- nidae: Trachelas pacificus (Chamberlin & I vie 1935); (3) Theridiidae: Theridion dilutum Levi 1957 and Theridion melanurum Hahn 1831; (4) Oxyopidae: Oxyopes scalaris Hentz 1845 and Oxyopes salticus Hentz 1845; (5) Agelinidae: Hololena nedra Chamberlin & COSTELLO & DAANE— SPIDERS AND INSECT PREDATORS IN VINEYARDS \o fO 04 0 O' O' cn 04 00 in in CO Y_ 0- 04 (U VO '—1 04 q ON 00 q q q in in q CN O) a .3 04 ON d in in 0 d d cb d d 0 d in 04 04 ON ^ * /-N /— N /-N ^ a w a\ 04 On On On 0 04 NO NO ON in O' NO CO ■O CO •g 00 +1 r- d \Q 0 NO d <— ( d 0 in d d 04 d 04 d O) 0 d 0 d 04 d 0 0 d d 0 d CO d R o’ d \D 04 O' -H On NO O' 0 04 CN CO ON On NO 04 T3 ^ Dh cd 04 a\ 04 q 04 00 q q ON 00 0 0 0 q 0 fb d ^ >— 1 d 04 0 d d d d d 0 d o^ \D m in O' O' -H O' NO 0 (N On ON m NO 1 .2 On q q 00 q q 04 NO O) q CO 0 0 0 m »n k. (N d 04 0^ 04 d d d d d d 0 0 0 d d o^ ^ On "O m u ^ pq ^-N X<-»S /— N s 0 C/D r-- \o O' ON 0 ON m 0 00 04 y-H y-H 0 04 a c ^ w \D 0 0 d d d 04 0 d d 0 d d 0 d 0 d 0 d 0 d d 0 d Is (U (U § (N m cn in fO ^ in 'Tf CO O' CO Y— 1 in 0 NO D m q 00 eo 00 04 q 0- 0 0 0 0 0 & ce ‘u 0 s r-s /Y~S w r- 0 in O' CO NO 04 NO NO On 04 CO C ^ >> 00 cn q 04 ^ m q 0 0 0 0 '-H • 1— 1 ^ 0 + 1 d d q d d d d cb d d d d d 0!^ 0 _i 0 NO 0 in On O' CO o- O' CO _( CO 04 ON . S 0 0 On in 00 in 04 00 ON CO 0^4 ON ON O' 0 "? 04 d- q q 04 q q CO q ^ 0 0 0 04 04 d fb 0 ON cb O'^ 0 d d d d d d d 04 -S s in 04 On « s r~s ^-s /-“s ^-s t+H ^ D W) 'O W in NO On q m 00 ON 04 —1 04 04 04 0 S 00 04 O' 0 0 04 q 0 0 in 0 0 0 0 1-H + 1 04 d d d d d d 04 d d d d d d d d T3 d \D O' 0 in q NO ON m 04 NO -H G V, 0 d r-; 04 04 in 00 q q ’-H in 04 0 0 0 0 D s 0 }-l 1 (U 06 04 d d d 0 ON 0 d d d d d d d Y-( a § S s 0 0 04 in CO 00 00 00 NO 00 O' 0 rs 04 0 in CNl in cn in 0 0 0 0 'S d 04 0^ d ^ 0 d 04 ON 0 d d d d d 0 0 d "a ^ a. 2 fO 04 04 ON g :s ^-N ,Y-S /—S a S NO NO NO CO NO 0 ’X3 0 fl w ^«~s /•'s /— s t— ( ?-H 1—1 04 '-H s a cd c/3 \D 00 00 On 00 0 0 0 0 0 T3 + 1 q cn 04 0 0 q d d d d d d Ih r— ( d d d d d d T— 1 B a d W w w w w w- W O' O' O' ^ O' 'S cd a § m ON m 04 04 On 04 r— 1 Y— 1 y-H CO ^ 0 a 0 ^ c D •’“' S (U 00 K q q q in ON d 0 d cn d q 04 0 0 d 0 d 0 d 0 0 d d 0 0 d -d - ? r 2 1. 0 a S -C CI( e to « d d •2 k. Co <3 M W d ^ i Oh 0 i-i 0 g s Cj s _-5 ‘2 >3 &-1 > s 0 S !... 1 3 0 Table 1. — Me for each site. Su Vh 0 d *73 0L| Araneae Theridion spp. Cheiracanthiui inclusum Trachelas paci Hololena nedr Oxyopes spp.^ Metaphidippus Erigone dento, Other spiders Spider total Acari Anystis agilis Insecta Hippodamia a Chrysopidae Nabis americq Orius spp. Geocoris spp. Zelus renardii Tenodera arid sinensis u 0 cd T3 « i-i a '0 0 fi hH 533 534 THE JOURNAL OF ARACHNOLOGY Ivie 1942; and (6) Salticidae: Metaphidippus vitis (Cockerell 1895). Overall spider abundance varied among sites, ranging from a high of 49.1 spiders per vine (Woodbridge site) to a lov^ of 10.7 spi- ders per vine (Reedley site) (Table 1). Species composition also varied among sites and may have contributed to differences in spider abun- dance. For example, overall spider abundance was highest at the Del Rey and Woodbridge sites, where the dominant spiders were the small, web-building theridiids, T. dilutum and T. melanurum. In contrast, overall spider abundance was more than 50% lower at the other sites, where larger spiders, such as the nocturnal hunters C. inclusum and T. pacifi- cus, dominated the spider community (Table 1). There were also differences in spider sea- sonal abundance (Fig. 1). Theridion spp. was the most abundant spider group, with the highest overall spider density in both the ear- ly-season (—17 per vine) and late-season (—34 per vine) samples, but equivalent with C. in- clusum in mid-season samples (—7.5 per vine). Cheiracanthium inclusum was the next most abundant spider, with densities relatively low early in the season (—2 per vine) and peaking late in the season (—18 per vine). The agelinid, Hololena nedra, maintained a rela- tively steady population density of —4.7 spi- ders per vine throughout the season. The sea- sonal abundance patterns reported here are consistent with those reported in Costello & Daane (1995). Insects. — Predaceous insects collected in- clude Hippodamia convergens Guerin-Mene- ville 1842 (Coleoptera, Coccinellidae); Chry- soperla comanche Banks 1938, Chrysoperla carnea (Stephens 1836), and Chrysopa ocu- lata Say 1839 (Neuroptera, Chrysopidae); Na- bis americoferus Carayon 1961 (Heteroptera, Nabidae); Orius spp. (Heteroptera, Anthocor- idae); Geocoris spp. (Heteroptera, Lygaeidae); Zelus renardii Kolenati 1856 (Heteroptera, Reduviidae); and Tenodera aridifolia sinensis Saussure 1871 (Mantodea, Mantidae). Overall, predaceous insect density was low- est at the Reedley and Ripperdan sites, with seasonal means of 0.06 and 0.10 predators per vine, respectively, and most abundant at the Woodbridge and Parlier sites, averaging 1.1 and 1.3 predators per vine, respectively (Table 1). There were also differences among sites in Trachelas pacificus 5 1 ^ 0 5 1 Oxyopesspp. 500 May 1000 June July 1 500 2000 2500 August September Mean Cumulative Degree Days (D 10) Figure 1 . — Mean seasonal abundance of the most abundant spider species on grapevines, plotted against cumulative seasonal degree days above 10 °C (since January 1), all vineyards and years com- bined. species composition. At the Woodbridge site, the most abundant insect predators were the chrysopids (0.5 per vine), whereas at the Par- lier site, N. americoferus was most frequently collected (0.8 per vine) (Table 1). Predaceous insect seasonal patterns show that N. americoferus was the most abundant insect predator overall (Fig. 2). Its population rose from near zero in early-season samples to —0.6 per vine in late-season samples. Chry- sopids were the most abundant predaceous in- sects in early-season samples, with densities of —0.6 per vine, but thereafter were quite rare (Fig. 2). Coccinellidae were also relatively abundant in early-season samples (0.35 per vine at the first sampling period) and their density also steadily dropped in later samples. COSTELLO & DAANE—SPIDERS AND INSECT PREDATORS IN VINEYARDS 535 1.0-1 0.5- 0.0 I.On 0.5 0.0 Orius spp. Geocoris spp. 500 1000 1500 2000 2500 May June July August September o Mean Cumulative Degree Days (D 10) Figure 2. — Mean seasonal abundance of the most abundant predaceous insect groups on grapevines, plotted against cumulative seasonal degree days above 10 °C (since January 1), all vineyards and years combined. Orius spp. and Geocoris spp. were not col- lected until the third sampling period (mid- summer), and peaked in late-season samples at 0. 1 1 and 0.06 per vine, respectively. DISCUSSION These results show that spiders overwhelm- ingly outnumber predaceous insects on grape- vines in California’s central valley. The expla- nation for this may partly lie in the type and abundance of prey species: the low number of predaceous insects may reflect the lack of pre- ferred prey on grapevines. At all of our study sites, the most abundant insects on grape fo- liage are various Diptera, which are most abundant in the spring and early summer, and the leafhoppers Erythroneura elegantula Os- bom 1928 and E. variabilis Beamer 1929 (Homoptera, Cicadellidae). Erythroneura spp. have three generations in the central valley, with nymphal peaks occuring in late May, mid-July and early September. Leafhopper densities which reach 10-15 nymphs per leaf may require insecticide treatment to prevent economic damage. In comparison, there were low densities of other potential arthropod prey, such as lepidopteran larvae {Platynota stultana and Desmia funeralis), mealybugs {Pseudococcus maritimus Ehrhom 1900) and spider mites {Tetranychus pacificus McGregor 1919 and Eotetranychus willametti [Mc- Gregor 1917]). Prey such as aphids and white- flies are only occasionally found on grape- vines, and at relatively low densities. Insect predators such as coccinellids and chrysopids will feed on a variety of soft bod- ied insects, including Erythroneura spp.; how- ever, they are better known as predators of aphids and mealybugs (Daane et al. 1998). The lack of preferred prey likely affects the dispersal habits of adult coccinellids and chry- sopids, and their density on grapevines. For example, migration of Hippodamia conver- gens from overwintering sites in the Sierra Nevada foothills to the San Joaquin Valley is arrested when adult beetles find aphids and their honeydew (Hagen 1962). Similarly, Chrysopa carnea responds to aphid honeydew (Hagen 1950). It is well known that cover crops such as vetches and cereals support high populations of aphids (Bugg et al. 1991), and we suspect that the relatively high early sea- son populations of H. convergens and chry- sopids we found on the grapevines were due to the presence of aphids on cover crops and weeds in and around the study vineyards at that time. The decline of these predators, dur- ing the season, followed the decline of their preferred prey on the cover crops. Although spiders are polyphagous, we found differences among vineyard species in prey preference. For example, Metaphidippus vitis does not feed on leafhoppers in the lab- oratory; and, in this study, its numbers were relatively low compared with other spider spe- cies. In contrast, field observations suggest that Theridion spp. feed primarily on leafhop- pers, with high populations of Theridion pos- itively correlated with high leafhopper densi- ties (Costello & Daane 1995). In this study, Theridion spp. reached the highest density of any spider group. Theridion spp. numbers were highest at the Woodbridge and Del Rey sites, where there were also high population levels of leafhoppers (Daane & Costello 536 THE JOURNAL OF ARACHNOLOGY 1998). Theridion dilutum and T. melanurum are small (adults are —0.5 cm), have low food requirements, occupy very little territory com- pared to larger spiders such as Cheiracan- thium inclusum and Hololena nedra, and Theridion spp. populations increase consider- ably from mid- to late-summer. Therefore, Theridion spp. densities may be highest be- cause they readily feed on leafhopper nymphs and because grapevines can support more of these spiders per given area compared with other spider species. That nabids increased over the course of the season may reflect their ability to use leafhop- pers as food. Nabids are good predators of leafhoppers (Martinez & Pienkowski 1982; Flinn et al. 1985). Other insect predators, such as Orius spp., prefer thrips and spider mites. Geocoris spp. feed on lepidopteran and he- mipteran eggs and nymphs, spider mites, aphids and whiteflies (Hagler & Cohen 1991). The low densities of these prey items on vines may explain the low density of Orius and Geocoris species we found. Spiders may also comprise the majority of the predator community because most species overwinter in the vineyard and are therefore permanent residents. They are a more stable part of the predator community than insect predators because of their broader diet breadth and their ability to subsist for long periods of time without food. Insect predators such as Hippodamia convergens and chrysopids are more migratory, and often follow migratory pest populations. All but one of the spiders mentioned in this study have been found over- wintering in cardboard bands placed around the vine trunks, the exception being Erigone dentosa (M.J. Costello & K.M. Daane unpubl. data). None of the predaceous insects has been found overwintering on the vines. That E. dentosa was not found overwintering in vine- yards and was only found in the early part of the growing season, suggests that it is more migratory than the other spider species, prob- ably ballooning into vineyards in the spring and leaving for other habitats during the sum- mer. Finally, the sampling methods used will af- fect the kinds and numbers of predators col- lected. Costello & Daane (1997) compared the D-vac to foliage beating in vineyards, and found that spider density was underestimated by 87% with the D-vac, and overestimated by 35% with the funnel shake method. The D- vac also biased samples toward smaller and more mobile spiders compared to beating or shaking methods. In addition, foliage shaking methods do not collect flying predators. This is most important for the tiger fly, Coenosia humilis Meigen 1826 (Diptera, Muscidae), which can be quite common in San Joaquin Valley vineyards. The adult captures its prey on the wing and has been observed feeding on leafhopper adults (immature Coenosia feed on earthworms in the soil and, therefore, are not collected). We have collected this fly with the D-vac and have usually found the mean density to be less than 5 per vine (unpubl. data). In addition, very small predators such as A. agilis may never be sampled with the D-vac, and are probably more efficiently sam- pled with the drop cloth method than the fun- nel method. This may partly explain why an additional small insect predator, Leptothrips mali, was observed at the Woodbridge site but was never found in the samples. This is the first report that spiders comprise such a high percentage of a predator com- munity in vineyards. The great number of spi- ders in comparison to other predators reveal, empirically, why so much research has fo- cused on spiders as vineyard predators (Zalom et al. 1993; Costello & Daane 1998; Roltsch et al. 1998). These results suggest that pre- daceous insects play a minor role in suppress- ing insect pest populations in California vine- yards. We note that leafhoppers were the primary prey species in our study sites. In vineyards with high mealybug or lepidopteran populations, the natural densities of preda- ceous insects may be higher. More work is needed in determining the role of spiders on economically important vineyard insects such as leafhoppers and the lepidopteran complex. We are currently working on the development of immunochemical assays to estimate prey consumption by vineyard spiders. ACKNOWLEDGMENTS We are grateful to the Robert Mondavi Winery, Fred Smeds, Kenneth Chooljian, Paul Wulf, and the UC Kearney Agricultural Cen- ter for the use of their vineyards. Funding was provided by the California Table Grape Com- mission, Lodi-Woodbridge Winegrape Com- mission, UC Statewide IPM Project, UC Sus- tainable Agriculture Research and Education COSTELLO & DAANE— SPIDERS AND INSECT PREDATORS IN VINEYARDS 537 Program and the USD A National Research Initiative Competitive Grants Program. Thanks to Eric Davidian, Glenn Yokota, Ross Jones, Kimberly Miyasaki and Jose Cantu for field and laboratory assistance, and to Drs. Robert Bugg, Mark Mayse and James Hagler for reviewing an earlier version of the manu- script. LITERATURE CITED Agnew, C.W. & J.W. Smith, Jr. 1989. Ecology of spiders (Araneae) in a peanut agroecosystem. En- viron. EntomoL, 18:30-42. Bardwell, C.T. & A.L. Averill. 1997. Spiders and their prey in Massachusetts cranberry bogs. J. ArachnoL, 25:31-41. Bishop, A.L. 1980. The composition and abun- dance of the spider fauna in south-east Queens- land cotton. Australian J. Zool., 28:699-708. Breene, R.G., W.L. Sterling & D.A. Dean. 1989. Predators of the cotton fleahopper on cotton. Southwestern EntomoL, 14: 159-166. Buchholz, U. & G. Schruft. 1994. Rauberische Ar- thropoden auf Bliiten und Friichten der Weinrebe {Vitis vinifera L.) als Antagonisten des einbin- digen Traubenwicklers (Eupoecilia ambiguella Hbn.) (Lep., Cochylidae). J. App. EntomoL, 118: 31-37. Bugg, R.L., J.D. Dutcher & P. McNeill. 1991. Cool-season cover crops in the pecan orchard un- derstory: Effects on Coccinellidae (Coleoptera) and pecan aphids (Homoptera: Aphididae). Biol. Cont., 1:8-15. Carroll, D.P. & S.C. Hoyt. 1984. Natural enemies and their effects on apple aphid Aphis pomi DeGeer (Homptera: Aphididae), colonies on young apple trees in central Washington. Envi- ron. EntomoL, 13:469-481. Costello, M.J. & K.M. Daane. 1995. Spider species composition and seasonal abundance in San Joa- quin Valley grape vineyards. Environ. EntomoL, 24:823-831. Costello, M.J. & K.M. Daane. 1997. Comparison of sampling methods used to estimate spider (Ar- aneae) species abundance and composition in grape vineyards. Environ. EntomoL, 26:142-149. Costello, M.J. & K.M. Daane. 1998a. Influence of ground cover on spiders in a table grape vine- yard. EcoL EntomoL 23:33-40. Costello, M.J. & K.M. Daane. 1998b. Role of cov- er crops on pest and beneficial arthropods in vineyards. Pp. 93-106. In Cover Cropping in Vineyards: A Growers Handbook (C.A. Ingels, R.L. Bugg, G. McGourty & L.P. Christensen, eds.). Univ. of California Division of Agriculture and Natural Resources Publication No. 3338, Oakland, California. Daane, K.M. & M.J. Costello. 1998. Can cover crops reduce leafhopper abundance in vineyards? California AgricuL, 52(5):27-33. Daane, K.M., K.S. Hagen & N.J. Mills. 1998. Pre- daceous insects for insect and mite management. Pp. 62-1 14. In Mass-reared natural enemies: Ap- plication, regulation, and needs (R.L. Ridgway, M.P. Hoffmann, M.N. Inscoe & C.S. Glenister, eds.). Thomas Say PubL, EntomoL Soc. Ameri- ca. Dean, D.A., W.L. Sterling & N.V Homer. 1982. Spiders in eastern Texas cotton fields. J. Arach- noL, 10:251-260. Flinn, PW, A.A. Hower & R.A. Taylor. 1985. Preference of Reduviolus americoferus (Hemip- tera: Nabidae) for potato leafhopper nymphs and pea aphids. Canadian EntomoL, 113:365-369. Hagen, K.S. 1950. Fecundity of Chrysopa calif ar- nica as affected by synthetic food. J. Econ. En- tomoL, 43:101-104. Hagen, K.S. 1962. Biology and ecology of preda- ceous Coccinellidae. Ann. Rev. EntomoL, 7: 289-326. Hagler, J.R. & A.C. Cohen. 1991. Prey selection by in vitro- and field-reared Geocoris punctipes. EntomoL Exp. AppL, 59:201-205. Knutson, A.E. & EE. Gilstrap. 1989. Direct eval- uation of natural enemies of the southwestern com borer (Lepidoptera: Pyralidae) in Texas com. Environ. EntomoL, 18:732-739. Martinez, D.G. & R.L. Pienkowski. 1982. Labo- ratory studies on insect predators of potato leaf- hopper eggs, nymphs, and adults. Environ. En- tomoL, 11:361-362. MacLellan, C.R. 1973. Natural enemies of the light brown apple moth, Epiphyas postvittana, in the Australian capital territory. Canadian EntomoL, 105:681-700. Nyfeller, M. & G. Benz. 1987. Spiders in natural pest control: A review. J. AppL EntomoL, 103: 321-339. Nyfeller, M., W.L. Sterling & D.A. Dean. 1992. Impact of the striped lynx spider (Araneae: Oxy- opidae) and other natural enemies on the cotton fleahopper (Hemiptera: Miridae) in Texas cotton. Environ. EntomoL, 21:1178-1188. Nyfeller, M., W.L. Sterling & D.A. Dean. 1994. Insectivorous activities of spiders in United States field crops. J. AppL EntomoL, 118:113- 128. Plagens, M.J. 1983. Populations of Misumenops (Araneidae: Thomisidae) in two Arizona cotton fields. Environ. EntomoL, 12:572-575. Reichert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: Spiders in garden test systems. Ecology, 71:1441-1450. Roach, S.H. 1980. Arthropod predators on cotton, corn, tobacco, and soybeans in South Carolina. J. Georgia EntomoL Soc., 15:131-138. Roltsch, W.R., R. Hanna, H. Shorey, M. Mayse & 538 THE JOURNAL OF ARACHNOLOGY F. Zalom. 1998. Spiders and vineyard habitat re- lationships in central California. Pp. 311-338. In Enhancing Biological Control: Habitat Manage- ment to Promote Natural Enemies of Agricultural Pests (C.H. Pickett & R.L. Bugg, eds.). Univ. of California Press, Berkeley, California. Royer, TA. & D.D. Walgenbach. 1991. Predaceous arthropods of cultivated sunflower in eastern South Dakota. J. Kansas Entomol. Soc., 64:112- 116. Wisniewska, J. & R.J. Prokopy. 1997. Pesticide ef- fect on faunal composition, abundance, and body length of spiders (Araneae) in apple orchards. Environ. Entomol., 26:763-776. Zalom, EG., R. Hanna, C. Elmore & P. Christensen. 1993. A cover crop for vineyard pest, weed, and nutrition management. 1992-93 Research Report for California Table Grapes, California Table Grape Commission, Fresno. Manuscript received 6 January 1998, revised 21 July 1998. 1999. The Journal of Arachnology 27:539-541 RESEARCH NOTE UNUSUAL PHENOTYPE SUGGESTS ROLE FOR HOMEOTIC GENES IN ARACHNID DEVELOPMENT Studies of segmental mutations in Dro- sophila melanogaster have led to the discov- ery of several classes of regulatory genes im- portant in determining body pattern (Carroll 1995). These regulatory genes are also known as transcriptional factors because their proteins bind to another gene’s control re- gions, or promoters, allowing for controlled expression or repression. For example, at- tachment of maternal effect gene transcripts to specific areas of an ovum in Drosophila, initiates the anterior-posterior (AP) body axis (Melton 1991). Diffusion of maternal effect proteins from opposing ovum poles creates a dual concentration gradient that differentially activates or represses additional classes of transcriptional factors. This cascade of reg- ulatory gene expression determines position- al information, indicating body polarity and regional specificity. Regional specificity within body plans is largely determined by a class of transcription- al factors known as homeotic genes. Specific concentrations of maternal effect proteins turn on different homeotic genes controlling the identity of segments along the AP axis of an arthropod’s body. Remarkably, homeotic genes are arranged on chromosomes in linear clusters corresponding to their exact sequence of expression. Those located toward the left end of the complex are expressed in posterior parts of the body, while those to the right are expressed toward the anterior parts (Kenyon 1994). Bithorax (BX-C) and Antenapedia (ANT-C) are the two known complexes of homeotic genes. ANT-C in Drosophila controls the identity of appendages (Carroll 1995). A mu- tant ANT-C gene causes flies to grow a leg in their antennae socket. BX-C in Drosophila controls the morphology of the posterior tho- rax and abdomen. A mutant BX-C gene re- sults in two rear-thoraxes instead of one front and one rear thorax, causing flies to have two pairs of wings instead of one pair of wings and one pair of halteres. In organisms other than Drosophila, genetic mechanisms under- lying body plan development are less well un- derstood and have only recently been inves- tigated in arachnids (Damen et al. 1998; Telford & Thomas 1998). Recently, I collected an immature Misu~ menops sp. (Thomisidae) on the Hawaiian Is- land of Maui which showed a dramatic seg- mental mutation. It was collected from wet forest in the Nature Conservancy’s Waikamoi preserve on east Maui. This individual closely resembled Misumenops anguliventris Simon 1900, one of 17 described species of Misu- menops endemic to the Hawaiian Islands. Thomisids are one of the few spider families containing genera that are known to be excep- tionally diverse in the Hawaiian archipelago (Gillespie 1994). After closer examination of this individual under a light microscope, I noticed a second set of eight eyes on its abdomen. These “ab- dominal eyes” displayed the exact pattern of the eight “normal” eyes on the cephalothorax. In addition, the dorsal aspect of the abdomen displayed the same type and pattern of setae also found on the carapace of the cephalotho- rax. Despite these dramatic morphological ab- errations, the posterior aspect of the abdomen resembled a “normal” abdomen. At the time of collection, the spider was a third or fourth instar juvenile. It lived for two months and appeared typical in behavior. After death, the spider was preserved in 70% alcohol and was then prepared and examined using a Hitachi S-800 scanning electron microscope (see Figs. 1,2). The specimen has been deposited in the Bishop Museum entomological collections. Kaston (1982) summarized accounts of oc- 539 540 THE JOURNAL OF ARACHNOLOGY Figure 1. — Scanning electron micrograph of anomalous Misumenops sp. thomisid, dorso-lateral view. ular anomalies in spiders. Of the nine cases he described, only two involved spiders gain- ing eyes. These specimens, having 14 and 16 eyes, were explained as a result of embryonic duplication of a head region. It is premature to suggest that these phenotypes were a result of a mutation in a major regulatory gene be- cause there are no accompanying figures showing where these eyes were located. If ad- ditional eyes were located on a segment other than the cephalic, this confusion in segment identity would strongly suggest abnormal ho- meotic gene expression. The “abdominal” eyes and duplicated se- tation pattern shown in the spider I collected may be explained by several different hypoth- eses. First, a mutation in a homeotic gene of the bithorax complex may account for the ab- normal phenotype. However, homeotic muta- tions are expressed in individual segments and because spider abdomens are composed of several segments, generation of this pheno- type would require independent mutations in all segments. It is more likely that this mutant was produced because the wrong set of ho- meotic genes was turned on, while the correct set was not. This could be brought about by several mechanisms. A mutation in a regula- tory gene determining body polarity might Figure 2. — Illustration of anomolous Misumen- ops sp. thomisid. PME = posterior median eyes, CP = cephalothorax, AD = abdomen. turn on an entire group of regionally inappro- priate homeotic genes. Alternatively, dupli- cation of anterior structures might arise if ma- ternal nurse cells placed transcripts activating anterior development at the anterior end of the ovum as well as the where the abdomen would normally arise. The mutant phenotype could also be created through a chromosomal aberration produced during gametogenesis. If the linear cluster of homeotic genes is dupli- cated at the region corresponding to the ceph- alothorax, this might result in the development of two cephalothoraxes. Evolution of homeotic genes may explain the immense diversity of body forms seen among arthropods (Kenyon 1994; Carroll 1995). Small mutations in these highly con- served genes result in macro-mutations, pro- viding an evolutionary mechanism for gener- ating novel phenotypes. Homeotic genes have also been identified in cnidarians, nematodes and annelids. Most recently, homeotic genes have been identified in a spider (Damen et al. 1998) and mite (Telford & Thomas 1998). Comparative investigation of homeotic gene expression will undoubtedly play an important role in understanding the evolution of arach- nid morphology. I thank Tina Carvalho, Marilyn Dunlap, Steve Robinow, Randy Haley, Rosemary Gil- lespie and Cherry Ann Rivera for assistance in preparing this note and the Nature Conser- vancy for access to the collecting site. garb— UNUSUAL ARACHNID PHENOTYPE 541 LITERATURE CITED Carroll, S.B. 1995. Homeotic genes and the evo- lution of arthropods and chordates. Nature, 376: 479-485. 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. Proc. Natl. Acad. Sci., 95:10665-10670. Gillespie, R.G. 1994. Hawaiian spiders of the ge- nus Tetragnatha. III. Tetragnatha acuta clade. J. ArachnoL, 22:161-168. Kaston, B.J. 1982. Additional ocular anomalies in spiders. J. Arachnol., 10:279-281. Kenyon, C. 1994. If birds can fly, why can’t we? Homeotic genes and evolution. Cell, 78:175- 180. Lewis, E.B. 1978. A gene complex controlling seg- mentation in Drosophila. Nature, 276:565-570. Melton, D.A, 1991. Pattern formation during ani- mal formation. Science, 252:234-241. Telford, M.J. & R.H. Thomas. 1998. Expression of homeobox genes shows chelicerate arthropods retain their deutocerbral segment, Proc. Natl. Acad. Sci., 95:10671-10675. Jessica E. Garb: Department of Zoology and Center for Conservation Research and Training, University of Hawaii, Edmundson Hall, Honolulu, Hawaii 96822 USA Manuscript received 30 May 1997, revised 9 No- vember 1998. 1999. The Journal of Arachnology 27:542-546 RESEARCH NOTE NOMENCLATURE OF THE ORB-WEB In science, we have to know what we are talking about when we say something. Uefor- tunately, in the literature about orb-webs and orb-web construction, different terms are used for the same part of the web and— even worse— the same term is used by different au- thors for different parts of the web. The pre- sent note tries to improve the situation by pro- posing a nomenclature for the different parts of the orb-web (Fig. 1). At the same time, an overview of the terms used by various other authors is given in English, German and French. Anchor threads and frame threads.— The web is supported by anchor threads attached to the supporting structure at anchor points. The thread along the outside of the web is called the frame. The primary frame (thread) is attached on both ends to anchor threads and forms the outermost outline of the web. A sec- ondary frame (thread) is attached to two pri- mary frame threads that form a comer with each other (Mayer 1952). The point where two primary frame threads connect to an an- chor thread is called a frame point. The (pri- mary) frame thread along the top of the web is also called bridge thread— not to be con- fused with the bridging line, the thread the spider lets float in the breeze to cross open gaps (Peters 1989). Radii.— The threads running more or less straight from the center of the web to the cir- cumference are called radii. There are differ- ent kinds of radii, but not all authors make the same distinction among them. In the normal orb-web, as exemplified by that of Araneus diadematus, I propose to distinguish between proto-radii, primary radii and secondary radii. Some species additionally have subsidiary ra- dii and others have accessory radii. Proto-radii only exist during the early stag- es of web construction. They are threads be- tween the proto-hub (hence the name for the proto-radii) and the supporting structure. Dur- ing the web building process, proto-radii are usually converted into (and partially replaced by) anchor threads (Eberhard 1990; Zschokke 1996). Primary radii are radii that are con- structed simultaneously with a frame thread (primary or secondary); secondary radii are constructed without building a frame thread at the same time. When looking at a finished web, primary and secondary radii cannot readily be distinguished, although Wirth (1988) has been able to tell the difference by studying the fine structure of the connection between the radius and the frame thread. In Fig. 1, the distinction was based on the re- cording of the construction of the web. If the distinction between primary and secondary ra- dii is not possible or not necessary, both can simply be named Tadius.’ Some spiders build subsidiary radii. Sub- sidiary radii are radii that do not start at the hub but somewhere further out: they are either attached to another radius (Cyrtophora sp.) or to the auxiliary spiral (Nephiia sp.). Finally, some symphytognathoid spiders build acces- sory radii. Accessory radii are made after sticky spiral construction but before hub mod- ification (Coddington 1986b; pers. comm.). They are distinct from normal radii by not be- ing connected to the sticky spiral. In the past, there has been a certain con- fusion about the use of the terms ‘primary ra- dius/ ‘secondary radius' and ‘tertiary radius’. ‘Primary radius’ has been used to designate what I call a proto-radius (Tilquin 1942; Cod- diegton 1986a), to designate a radius con- structed together with a primary frame thread (Mayer 1952), or— as proposed here— to des- ignate a radius constructed simultaneously with a (primary or secondary) frame thread (Peters 1937b; Wirth 1988). The term ‘sec- ondary radius’ has been used by several au- thors (Tilquin 1942; Mayer 1952; Savory 542 ZSCHOKKE— ORB-WEB NOMENCLATURE 543 bridge thread hub spiral sticky spiral ; secondai7 radii frame points hub secondary frame primary radii ~ f4(remains of) “'F auxiliary spiral primary frame anchor points anchor threads Figure 1.— Web of Araneus diadematus with the terms for its parts. See text for names of structures not found in the web of Araneus diadematus (e.g., retreat, stabilimentum). 1952) to designate a radius built at the same time as a (secondary) frame thread. Many au- thors (Peters 1937b; Wirth 1988; Vollrath 1992) have — as proposed here— -used it to de- scribe a radius built without a frame thread and some other authors (e.g.. Shear 1986) have used it to describe what I call a subsid- iary radius. The term ‘tertiary radius’ has been used to describe what I call secondary radius (Tilquin 1942; Mayer 1952; Vollrath 1992) or to describe what I call subsidiary radii (Eber- hard 1972). Jackson (1973) has devised an entirely dif- ferent terminology to distinguish between dif- ferent radii. His distinction is not based on the mode of construction but on the position in the web, allowing the identification of each radius. Spirals. — Spirals are the distinctive feature of orb-webs. Most orb-webs contain two ma- jor spirals: a spiral made out of sticky silk, the sticky spiral, and a spiral made out of non- sticky silk, the auxiliary spiral. The auxiliary spiral is removed by most spiders during con- struction of the sticky spiral and all that is left are small balls of silk along the radii (c/ Fig. 1). There are, however, webs of a few spiders (e.g., Nephila sp. and in parts of the web of Scoloderus sp.) where the auxiliary spiral re- mains in the completed web. Accordingly, some authors (e.g.. Stem & Kullmann 1975) distinguish between a structural spiral (‘Festi- 544 THE JOURNAL OF ARACHNOLOGY Table 1 . — List of proposed terms and terms used by other authors to designate certain parts of a spider’s web. In English, some authors used the word ‘line’ or ‘strand’ in place of ‘thread’. Other terms Proposed term in English In German In French anchor points mooring point Anheftungspunkte Verankerung spunkte anchor thread frame thread Ankerfaden fil d’ attache guy thread Haltefaden mooring thread Spannseil Tragseil Verankerungsfaden primary frame foundation Begrenzung Rahmen cadre secondary frame auxiliary frame Sekundarer Rahmen cadre secondaire cord Rahmen 2. Ordnung inner frame radial Y- structure section thread Y-frame Hilfsrahmen frame point bridge thread frame Y-structure Briickenfaden cable suspenseur fil suspenseur radius radial Radialfaden diametres radial thread Radialspeiche rayon radiating thread Radius spoke Speiche ray Stiitzfaden proto-radius primary radius Ausgangsstrahlen Grundstrahlen primary radius secondary radius secondary radius tertiary radius subsidiary radius secondary radius tertiary radius proto-hub hub rudimentary hub Nabe Warte moyeu sticky spiral capture spiral Fangspirale fil spiralaire catching spiral Klebfaden spire captrice ensnaring spiral Klebspirale spirale definitive outer spiral spirale exteme permanent spiral viscid spiral spirale gluante auxiliary spiral nonsticky spiral Festigungsspirale spirale auxiliare preliminary spiral Geriistspirale spirale provisoire primary spiral provisional spiral scaffolding spiral structural spiral temporary spiral Hilfsspirale spirale seche hub spiral inner spiral strengthening spiral Befestigungsspirale spirale interne U-tum loop Umkehrpunkt coudes en epingle a cheveux (point of) reversal Umkehrstellen grecques reverse switchback turnback turning point retour ZSCHOKKE— ORB=WEB NOMENCLATURE 545 Table 1. — Continued. Proposed term Other terms in English In German In French signal thread guide line to retreat Signalfaden fil avertisseur fil d’avertissement retreat hiding place Schlupfwinkel Versteck Warte demeure retraite refuge stabiiimentum Stabiliment stabiiimentum in hub: revetement gungsspirale’) which is left in the finished web as opposed to a temporary spiral which is removed during construction of the sticky spiral. Curiously, the term temporary spiral has also been used to describe the auxiliary spiral in webs where the auxiliary spiral is permanent (Eberhard 1975). The hub spiral is the innermost part of the auxiliary spiral which is not removed during construction of the sticky spiral. Other structures.— "In addition to frame, radii and spirals, some orb-webs contain ad- ditional structures. Some spiders do not sit on the hub when waiting for prey, but rather they sit hidden in the retreat (e.g., Zygielia x-no- tata). The signal thread connects the hub of the web with the retreat, allowing the spider sitting in the retreat to detect any vibrations occurring in the web and to dash to the center of the web without being slowed by the sticky spiral. Stabiiimentum is the name for a variety of additional silk structures on the orb web (for an overview of the different kinds of sta- bilimenta, see Foelix 1996). Terminology used by other authors and in other languages.-^Table 1 gives an over- view of the terms proposed in this note (first column) and the terms used by various other authors in English, German and French. The names are also given in German and French because many of the classic papers on spiders web were written in German (e.g., Wiehle 1927; Peters 1937a, 1937b; Mayer 1952) or in French (e.g., Tilquin 1942; Le Guelte 1964). I am grateful to Fritz Volirath for letting me use his library and to Jonathan Coddington for suggestions concerning this note. I have com- piled most of the data presented here during my dissertational work financed by a grant of the Swiss National Science foundation to Fritz Volirath. LITERATURE CITED Coddington, J, 1986a. The monophyletic origin of the orb web. Pp. 319-363, In Spiders: Webs, Be- havior and Evolution. (W.A. Shear, ed.). Stanford Univ. Press, Stanford. Coddington, J.A. 1986b. The genera of the spider family Theridiosomatidae. Smithson. Contrib. ZooL, 422:1-96. Eberhard, W.G. 1972. The web of Uloborus div- ersus (Araneae: Uloboridae). J. ZooL, 166:417- 465. Eberhard, W.G. 1975. The 'inverted ladder’ orb web of Scoloderus sp. and the intermediate orb of Eustala (?) sp. Araneae: Araneidae. J. Nat. Hist., 9:93-106. Eberhard, W.G. 1990. Early stages of orb construc- tion by Philoponella vicina, Leucauge mariana, and Nephila ciavipes (Araneae, Uloboridae and Tetragnathidae), and their phylogenetic implica- tions. J. ArachnoL, 18:205-234. Foelix, R.E 1996. Biology of Spiders. Oxford Univ. Press, Oxford. Jackson, R.R. 1973. Nomenclature for orb web thread connections. Bull. British Arachnol. Soc., 2:125-126. Le Guelte, L. 1964. Remarques sur la construction de la toile de Faraignee Zilla x notata. Bull. Soc. Sci. Bretagne, 39:83-91. Mayer, G. 1952. Untersuchungen iiber die Herstel- lung und Straktur des Radnetzes von Aranea dia- dema und Ziiia x-notata mit besonderer Beriick- sichtigung des Unterschiedes von Jugend- und Altersnetzen. Z. TierpsychoL, 9:337-362. Peters, H.M. 1937a. Studien am Netz der Kreuz- spinne (Aranea diadema). L Die Grandstraktur des Netzes und Beziehungen zum Bauplan des Spinnenkorpers. Z. Morph. Okol. Tiere, 32:613- 649. Peters, H.M. 1937b. Studien am Netz der Kreuz- spinne (Aranea diadema). IL liber die Herstel- lung des Rahmens, der RadialfMen und der Hilfsspirale. Z. Morph. Okol. Tiere, 33:128-150. Peters, H.M. 1989. On the structure and glandular origin of bridging lines used by spider for mov- 546 THE JOURNAL OF ARACHNOLOGY ing to distant places. Acta ZooL Fennica, 190: 309-314. Savory, T.H. 1952. The Spider's Web. Frederick Wame, London, Shear, W. A. 1986. The Evolution of Web-Building Behavior in Spiders: A Third Generation of Hy- potheses. Pp, 364-400, In Spiders: Webs, Be- havior and Evolution. (WA. Shear, ed.). Stanford Univ. Press, Stanford. Stem, H. & E. Kullmann. 1975. Leben am seidenen Faden. Die ratselvolie Welt der Spinnen, C. Ber- telsmann Verlag, Miinchen. Tilquin, A. 1942. La toile geometrique des araig- nees. Presses Univ. France, Paris. Vollrath, F. 1992. Analysis and interpretation of orb spider exploration and web-building behav- ior. Adv. Stud. Behav., 21:147-199. Wiehle, H. 1927. Beitrage .zur Kenntnis des Rad- netzbaues der Epeiriden, Tetragnathiden und Uloboriden. Z. Morph. Okol. Tiere, -8:468-537. Wirth, E. 1988. Sensorische und mechanische Gmndlagen des Netzbauverhaltens. Ph.D. thesis, Goethe Universitat Frankfurt, am Main. Zschokke, S. 1996. Early stages of web construc- tion in Araneus diadematus Clerck. Rev. .Suisse ZooL, hors serie 2:709-720. Samuel Zschokke: Dept, of Integrative Bi- ology, Section of Conservation Biology (NLU), University of Basel, .St. Johanes- Vorstadt 10, CH-4056 Basel, Switzerland Manuscript received 14 April 1998, revised 20 Oc- tober 1998. 1999. The Journal of Arachnology 27:547-549 RESEARCH NOTE ON SOFANAPIS ANTILLANCA (ARANEAE, ANAPIDAE) AS A KLEPTOPARASITE OF AUSTROCHILINE SPIDERS (ARANEAE, AUSTROCHILIDAE) Kleptoparasitic habits are well known in certain spiders, notably some mysmenids and members of the theridiid genus Argyrodes Si- mon 1864 (Elgar 1993). Members of the Die- tynidae, Heteropodidae, Oonopidae, Saltici^ dae, and Symphytognathidae have also been recorded as kleptoparasites of web~building spiders (Elgar 1993, table 1). We present here the first evidence of kleptoparasitism in the Anapidae, as well as the first report of a klep- toparasite associated with the primitive and relictual spider subfamily Austrochilinae. Austrochilines comprise two genera, Aus~ trochilus Gertsch & Zapfe 1955 and Thaida Karsch 1880, restricted to the temperate for- ests of Chile and adjacent Argentina. They build conspicuous, large (about 50-120 cm long), horizontal, aerial webs, consisting of a single layer of threads forming an irregular net (Forster et al. 1987; Zapfe 1955). The cribel- late, whitish threads make the web easily vis- ible. The horizontal net gradually bends into a concavity, forming a funnel that goes far back into log cracks, tree roots, or rocks, end- ing in a retreat where the spider rests during the day. At night, the spider hangs under its web and can be seen combing cribellate silk, or feeding. Large prey items, up to the spider’s body size, are wrapped with silk before being eaten. The retreats of adult females often contain much silk and several egg cases (Forster et al. 1987). The web is repaired when damaged; and the newly constructed patches, with bluish cribellate threads, are clearly discernible from the old portions, with powdered threads. Silk accumulation and web repair indicate that the web is persistent, a frequent characteristic of the hosts of kleptoparasites (Elgar 1993). Sofanapis antillanca Platnick & Forster 1989 are very small spiders that were previ- ously collected either by pyrethrin-fogging in logs, or in Berlese samples of leaf litter and moss (Platnick & Forster 1989). During recent field work, we found some specimens on aus- trochiline webs, and after a systematic exam- ination of webs, found evidence of kleptopar- asitic behavior in these anapids. During the day, individuals of S. antillanca were collect- ed on austrochilid webs (of both Austrochilus and Thaida species), hanging from threads, rather deep in the mouth of the funnel, but still visible with a headlamp. At night, when the host is on its web, the anapids were mostly concentrated around the opening of the funnel, closer to the horizontal web. They hang from the host web’s threads (Fig. 1), or more often, from an irregular mesh made of extremely fine threads. That mesh is presumably constructed by the Sofanapis, as it has not been found in non-infested webs. A web of an adult female of Thaida peculiaris Karsch 1880 observed at Aguas Calientes in the Parque Nacional Puy- ehue, Osomo, Region X (Los Lagos), Chile, was found to host several specimens of S. an- tillanca, and in that web the thin (anapid) net was particularly dense. No orbwebs were found on any host web, nor was any insect found caught in the presumed anapid silk. Consequently, it seems that antillanca does not construct its own web for prey capture. The type specimens of S. antillanca, collected by pyrethrin fogging inside a rotten tree trunk (Platnick & Forster 1989), may actually have come from an austrochilid web. A few hosts were observed while they were feeding. We found several S. antillanca walk- ing on the prey, some of them around the host’s mouth (Figs. 3-4). In those cases, al- though visibility was far from ideal, there was no evidence that the anapids were feeding from the fluids exposed by the chewing of the austrochiline. In another situation, where a host was feeding on a tipulid crane-fly, an in- 547 548 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — Sofanapis antillanca on webs of austrochilines. 1, Female walking on host web, from Chepu; 2, Austrochiline and S. antillanca feeding on a tipulid, from Contulmo (arrow points to the anapid feeding on the tipulid’s leg); 3, Austrochiline and S. antillanca feeding on a beetle, from Contulmo (arrows point to the four anapids); 4, Same, closer view. dividual of S. antillanca was seen feeding di- rectly from the insect’s leg (Fig. 2). On an- other occasion, a specimen of S. antillanca was observed feeding alone on a small mos- quito caught in an araneid web (also in the Parque Nacional Puyehue). This observation suggests that the anapid might have some abil- ity to locate prey in the host’s web, indepen- dent of the movements of the host, and that S. antillanca is not adapted to kleptoparasi- tism to the extreme condition of total depen- dence on the host (as apparently occurs with Curimagua bayano Forster & Platnick 1977, a symphytognathid with reduced mouthparts, Vollrath 1978). Moreover, the presence of oc- casional individuals of S. antillanca on webs of araneids and hahniids (in the Monumento Natural Contulmo in Arauco, Region VIII, Chile), as well as in Berlese samples of leaf litter and moss, indicates that these spiders are not obligately associated with austrochilines. However, the particularly high density of S. RAMIREZ & PLATNICK— A KLEPTOPARASITE ANAPID 549 antillanca collected on austrochiline webs suggests a special association with these hosts. Austrochilines are common in extremely to moderately moist forests of central and south- em Chile and adjacent Argentina. However, the kleptoparasitic anapids were found only in the most humid localities. In addition to the Puyehue and Contulmo localities noted above, Sofanapis have been taken from austrochiline webs at the following localities in Chile: Cal- eta La Arena in Llanquihue, and 15 km S of Chepu in Chiloe (both in Region X). Several intense but fruitless searches were performed in less humid localities, in both Chile and Ar- gentina. Material examined. — (Most austrochilines were either juveniles or were not collected together with its kleptoparasites): CHILE: Region IX: Cautm, Monumento Natural Contulmo, elev. 340 m, 38°0rS, 73HrW, 13 February 1992 (N. Platnick, P. Goloboff, M. Ramirez) 49 S. antillanca (Figs. 3, 4, AMNH); 18 November 1993 (N. Platnick, K. Catley, M. Ramirez, T. Allen) 1 d 5. antillanca (Fig. 2, AMNH). Region X (Los Lagos): Osomo, P. Nac. Puyehue, Aguas Calientes, 12 February 1992 (Plat- nick, Goloboff, Ramirez) many S 9 S. antillanca on a web of a female Thaida peculiaris Karsch (AMNH). Llanquihue, Caleta La Arena, 30 January 1991 (M. Ramirez) 2(33 9 2juv S. antillanca on a web of a subadult S austrochiline (MACN); same data, 29 1 juv on uncollected austrochiline’s web. Chiloe, Chepu, 15 km S de Chepu, 3 February 1991 (M. Ramirez) 1369 ljuv, on uncollected austro- chiline’s web (Fig. 1, MACN). LITERATURE CITED Elgar, M.A. 1993. Inter-specific associations in- volving spiders: kleptoparasitism, mimicry and mutualism. Mem. Queensland Mus., 33:411- 430. Forster, R.R., N.I. Platnick & M.R. Gray. 1987. A review of the spider superfamilies Hypochiloidea and Austrochiloidea (Araneae, Araneomorphae). Bull. American Mus. Nat. Hist., 185:1-116. Vollrath, F. 1978. A close association between two spiders: Curimagua bayano synecious on a Di- plura species. Psyche, 85:347-353. Zapfe, H.C. 1955. Filogenia y funcion en Austro- chilus manni Gertsch y Zapfe (Araneae-Hypo- chilidae). Trab. Lab. Zool. Univ. Chile, 2:1-53. Martin J. Ramirez: Graduate Fellow, Lab. de Artropodos, Dept, de Biologia, Univer- sidad de Buenos Aires, Pabellon II Ciudad Universitaria, 1428 Buenos Aires, Argenti- na; Adscript, Museo Argentino de Ciencias Naturales “Bernardino Rivadavia.” Norman I. Platnick: Curator, Department of Entomology, American Museum of Nat- ural History, Central Park West at 79th Street, New York, New York 10024-5192 USA. Manuscript received 15 December 1997, revised 27 April 1998. 1999. The Journal of Arachnology 27:550-552 RESEARCH NOTE CARBOHYDRATE ANALYSIS IN SPIDER HEMOLYMPH OF SELECTED LYCOSID AND ARANEID SPIDERS (ARANEAE: LYCOSIDAE AND ARANEIDAE) The literature gives incomplete and con- flicting information concerning the carbohy- drate composition of hemolymph in the Ara- neae. Various analyses of hemolymph components have been reported on several spiders including: the spider Cupiennius salei (Keyserling 1877) by Loewe et al. (1970); the theraphosid Aphonopelma hentzi (Girard 1854) by Stewart & Martin (1970); the large orb weaver Nephila madagascarensis (Vinson 1863) by Rakatovao & Ratsimamanga (1975); the spider Hexathele hochstetteri Ausserer 1871 by Bedford (1977); the araneids Araneus gemma (McCook 1888) and Argiope trifascia- ta (Forskal 1775) by Cohen (1980); selected species of the family Lycosidae by Punzo (1982); sparassids, pisaurids, and amaurobids by Punzo (1983); the theraphosid Eurypelma californicum Ausserer 1871 by Schartau & Leidescher (1983); hemolymph inorganic ions in 15 spiders and six scorpion species by Burton (1984). Punzo (1989) studied four spe- cies of mygalomorphs Bothriocyrtum califor- nicum (O.P Cambridge 1874), Aphonopelma echinum (Chamberlin 1940), Euagrus comsto- cki Gertsch 1935, and Atypus bicolor Simon 1836. With regard to the identity of carbohydrates present in the hemolymph, Rakatovao & Rat- simamanga (1975) and Bedford (1977) iden- tified trehalose and glucose, while Schartau & Leidescher (1983) identified only glucose. Loewe et al. (1970) and Stewart & Martin (1970) hypothesized the presence of trehalose because their data indicated sugars other than glucose present. The studies by Cohen (1980) and Punzo (1982, 1983, 1989) quantified car- bohydrate concentration differences among species but did not identify the carbohy- drate(s) present. The studies by Loewe et al. (1970) and Schartau & Leidescher (1983) re- ported the presence of circulating glycopro- teins, while the other studies did not test for these carbohydrates. The findings of Loewe et al. (1970), Stewart & Martin (1970), and Bed- ford (1977) indicate individual variation with- in conspecifics, while the other studies report total carbohydrate concentration within a giv- en species. In an attempt to clarify the carbohydrate composition of spider hemolymph, this study was conducted using Argiope aurantia Lucus 1833, Hogna carolinensis (Walckenaer 1805), Arctosa littoralis (Hentz 1844), and Rabidosa rabida (Walckenaer 1837). Spiders used in this study were collected in north-central Tex- as. Species were selected based on availabil- ity, size, and volume of hemolymph recovered from each individual. Voucher specimens of species used in the study are on deposit at the American Museum of Natural History (AMNH), New York. The lycosids were placed in metal cans (10 cm X 14 cm) which contained a small amount of soil, offered crickets and water, and were tested within 48 h of collection. Since A. aurantia is an orb weaver, the specimens were removed from the webs, brought to the laboratory, and tested. The spiders were anesthetized with carbon di- oxide and placed on a surgical restraint fol- lowing Randal (1980). The legs were severed at mid-femur and hemolymph was collected with capillary tubes, which were stored in 400p.l microfuge tubes. The tubes were la- beled and centrifuged at 12,000 X g for 10- 12 min. The resultant cell-free hemolymph was refrigerated at approximately 3 °C and an- alyzed within 72 h. Assays included total anthrone reaction to carbohydrates in untreated hemolymph and anthrone reaction after protein precipitation (Dubois et al. 1956) and specific enzyme di- gestion with glucose oxidase (Fleming & Peg- ler 1963) with and without trehalase digestion. 550 BARRON ET AL.— ANALYSIS OF SPIDER HEMOLYMPH 551 A variation on the phenol^ sulfuric acid test, using an addition of trichloroacetic acid (TCA) to bring the sample to a total of 5% TCA, was used to test for free carbohydrate. This new sample at 5% TCA was centrifuged 8-10 min at 12,000 X g, and 200|jl1 of the resultant clear supernatant was retained for to- tal carbohydrate analysis. Solutions of known concentration were used to plot standard curves from the resultant absorbance data at 470 nm in order to quantify total carbohydrates. Each group of spider he- molymph was analyzed against standards pre- pared at the same time. There was a range of variation among individuals for total carbo- hydrate concentration. As an example, R. ra- bida exhibited a range from 3.54|jLg to 45.8pg of carbohydrates per 10|rl of hemolymph. The variation in the amount of total carbohydrate present was a consistent feature of the data (C.V. = 44). Total glucose was determined from com- parisons of glucose oxidase activity of solu- tions of known concentration. Considerable individual variation occurred in the amount of glucose present in 5 pi hemolymph samples. Absorbance data for each control group var- ied. With regard to different sets of individu- als, the data were qualitatively based on var- iance of these standard absorbencies. The amount of glucose present after diges- tion with trehalase was determined for A. au- rantia. The total amount of glucose present increased following treatment of the hemo- lymph with trehalse. Based on a two point ANOVA, glucose present after trehalase di- gestion did not differ from the total amount of glucose present before treatment (P > 0.2). For example, A. aurantia hemolymph con- tained an average of 5.77 ±2.12pg of glucose per 5 pi hemolymph without trehalase treat- ment and an average of 7.89 ±3.1 Ipg glucose per 5 pi hemolymph after trehalase treatment. Total carbohydrate analysis after protein precipitation with TCA was performed on 16 A. aurantia and two R, rabida. The presence of TCA altered the color yield for the glucose standard solutions. Thus, standards were treat- ed the same way as the hemolymph samples and analyzed simultaneously (TCA and con- trol samples) to correct for this effect. In com- paring the total carbohydrates with and with- out TCA treatment, there was a two-fold decrease in the amount of carbohydrates pre- sent in all but one of the 18 specimens tested. Free carbohydrates accounted for less than 50% of the carbohydrates present in the ma- jority of the samples. In summary, glucose is the only detectable carbohydrate present in the hemolymph. The differences between total carbohydrate and to- tal glucose are small and may be accounted for by the change in color yield of the stan- dards from test to test. Thus, the absorbance data sets are not readily comparable due to the shift in relative color yield. The increase in the amount of glucose present after trehalase treatment is not significant. Data indicate a high degree of individual variation among in- dividuals in. the concentration of glucose in the hemolymph. These varying concentrations may reflect the physiological conditions of the animal and the environmental stresses that are placed upon it (Clarke 1979). The data sup- port the observation noted by Loewe et al. (1970) and Schartau & Leidescher (1983) that glucose exists partially in glycoproteins. Free glucose accounts for less than 50% of the total glucose present in this study. More work is needed to describe the existence of a hemo- lymph glycoprotein in the Araneae. The ex- tensive physiological work done on the class Insecta should serve as a model for investi- gations in other classes of arthropods. LITERATURE CITED Bedford, JJ. 1977. The carbohydrate levels of in- sect haemolymph. Comp. Biocfiem. Physiol., 57 A: 83-86. Burton, R.E 1984. Hemolymph composition in spi- ders and scorpions. Comp. Biochem. Physiol., 78A:613-616. Clarke, K.U. 1979. Visceral anatomy and arthro- pod phytogeny. Pp. 467--550. In Arthropod Phy- togeny. (A. Gupta, ed.). Van Nostrand Reinhold Co., New York. Cohen, A.C. 1980. Hemolymph chemistry of two species of araiieid spiders. Comp. Biochem. Phy- siol., 66A:715^717. Dubois, M., K.A. Gilles, J.K. Hamilton, P.A. Rebers & F. Smith. 1956. Colorimetric methods for de- temiination of sugars and related substances. Anal. Chem., 28:350-356. Fleming, ID. & H.F. Pegler. 1963. The determi- nation of glucose in the presence of maltose and isomaltose by a stable, specific enzymic reagent. Analyst., 88:967-968. Loewe, R., B. Linzen & W. von Stackelberg. 1970. Die gelosten Stoffe in der Haemolymphe einer 552 THE JOURNAL OF ARACHNOLOGY Spinne, Cupiennius salei Keyserling. Z. Vergl. Physiologic., 66:27-34. Punzo, F. 1982. Hemolymph chemistry of lycosid spiders. Comp. Biochem. Physiol., 71B:703- 707. Punzo, F 1983. Hemolymph chemistry of the spi- ders Heteropoda venatoria (Sparassidae), Pi- saurina mira (Pisauridae) and Amaurobius ben- netti (Amaurobidae). Comp. Biochem. Physiol., 75A:647-652. Punzo, F 1989. Composition of the hemolylmph of mygalomorph spiders (Orthognatha). Comp. Biochem. Physiol., 93A:757-760. Rakotovao, L.H. & A.R. Ratsimamanga. 1975. Physiolgie des insectes - Les constituants gluci- diques de Nephila madagascariensis femelle ad- ulte. C.R. Acad. Sci. Paris, 280:185-188. Randall, J.B. 1980. Surgical restraint apparatus for living spiders. J. Arachnol., 10:91. Schartau, W. & T. Leidescher. 1983, Composition of the hemolymph of the tarantula Eurypelma californicum. J. Comp. Physiol., 152:73-77. Stewart, D.M. & A.W. Martin. 1970. Blood and fluid balance of the common tarantula Dugie- siella hentzi. Z. Vergl. Physiologic, 70:223-246. P.D. Barron* and N.V. Horner: Depart- ment of Biology; and R.L. Cate: Depart- ment of Chemistry: Midwestern State Uni- versity, Wichita Falls, Texas 76308-2099 USA ' Current address: San Jacinto College Central, Pasadena, Texas 77501-2007 USA. INSTRUCTIONS TO AUTHORS (revised October 1996) Manuscripts are preferred in English but may be ac- cepted in Spanish, French or Portuguese subject to availability of appropriate reviewers. Authors whose pri- mary language is not English may consult the Associate Editor for assistance in obtaining help with English manuscript preparation. All manuscripts should be pre- pared 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 1500 words should be prepared as Feature Articles, shorter papers as Research Notes. Send four identical copies of the typed material together with photocopies of illustra- tions to the Managing Editor. Do not send original handmade illustrations until the manuscript has been ac- cepted. Mail to: Petra Sierwald, Managing Editor; Division of In- sects, Dept, of Zoology, Field Museum, Roosevelt Road at Lakeshore Drive, Chicago, IL 60605 USA [Telephone: (312) 922-9410, ext. 841; FAX: (312) 663-5397; E-mail: SIERWALD@FMNH.ORG] Correspondence relating to the initial submission of a manuscript, as well as the review process, should be directed to the Associate Editor. Correspondence re- lating all other matters should be directed to the Editor. After the manuscript has been accepted, the author will be asked to submit the manuscript on a computer disc. It must be in a widely used program (preferably in MS DOS WordPerfect). Indicate clearly on the computer disc the word processing program used. FEATURE ARTICLES Title page. — The title page will include the com- plete name, address, and telephone number of the au- thor 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 capital letters should be placed at the beginning of the first paragraph set off by a period. In articles written in English, a second abstract in an acceptable language may be included pertinent to the nationality of the author(s) or geo- graphic region(s) emphasized in the article. A second abstract in English must be included in articles not written in the latter language. Text. — Double-space text, tables, legends, etc. throughout. Three levels of heads are used. The first level (METHODS, RESULTS, etc.) is typed in capitals and on a separate line. The second level head 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 sepa- rated 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 regardless of language of the text. 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). Literature cited section. — Use the following style: Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid composition of major ampullate gland silk (dragline) of Nephila clavipes (Araneae, Tetragnath- idae). J. Arachnol., 18:297-306. Krafft, B. 1982. The significance and complexity of communication in spiders. Pp. 15-66, In Spider Communications: Mechanisms and Ecological Sig- nificance. (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. Running head. — The author surname(s) and an ab- breviated 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 con- tact the Editor. Tables. — Each table, with the legend above, should be placed on a separate manuscript page. Only hori- zontal lines (usually three) should be included. Use no footnotes; instead, include all information in the leg- end. Make notations in the text margins to indicate the preferred location of tables in the printed text. Illustrations. — Address all questions concerning il- lustrations to: James W. Berry, Editor-In-Chief; Dept, of Biolog- ical Sciences, Butler University, Indianapolis, Indi- ana 46208 USA [Telephone (317) 940-9344; FAX (317) 940-9519; E-mail: jwberry@thomas.butler.edu] All art work must be camera-ready for reproduction. In line drawings, pay particular attention to width of lines and size of lettering when reductions are to be made by the printer. Multiple photos assembled on a single plate should be mounted with only a minimum of space separating them. In the case of multiple illus- trations 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. The author should indicate whether the illustrations should be one column or two columns in width. The overall dimen- sions of original artwork should be no more than 1 1 inches (28 cm) X 14 inches (36 cm). Photocopies in review manuscripts should be reduced to the exact measurements that the author wants to appear in the final publication. Larger drawings present greater dif- ficulty in shipping and greater risks of damage for which the JOA assumes no responsibility. Make no- tations in the text margins to indicate the preferred position of illustrations in the printed text. 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, figure legends, footnotes, ta- bles with legends, figures. Page charges, proofs and reprints. — There are no page charges, but authors will be charged for changes made in the proof pages. Authors will receive a reprint order form along with. their page proofs. Reprints will be billed at the printer’s current schedule of costs. RESEARCH NOTES Instructions above pertaining to feature articles ap- ply also to research notes, except that abstracts and most headings are not used and the author’s name and address follow the Literature Cited section. CONTENTS The Journal of Arachnology Volume 27 Feature Articles Number 2 Fossil araneomorph spiders from the Triassic of South Africa and Virginia by P.A. Selden, J.M. Anderson, H.M. Anderson & N.C. Fraser 401 New species and cladistic reanalysis of the spider genus Monapia (Araneae, Anyphaenidae, Amaurobioidinae) by Martin j. Ramirez 415 The females of Anelosimus dubiosus and Anelosimus jabaquara (Araneae, Theridiidae) by M. de Oliveira Gonzaga & A.J. dos Santos 432 Revision of the groenlandica subgroup of the genus Pardosa (Araneae, Lycosidae) by C.D. Dondale 435 A revision of Central African Trabea (Araneae, Lycosidae) with the description of two new species from Malawi and a redescription of T. puree Hi by M. Alderweireldt 449 The identity of Pachyloides tucumanus n. comb, (ex Bosqia), with a proposal of generic synonymy and the new name Pachyloides yungarum (Opiliones, Gonyleptidae, Pachylinae) by L.E. Acosta ... 458 New synonyms in the genera Discocyrtus and Pachyloides (Opiliones, Gonyleptidae, Pachylinae) by L.E. Acosta 465 Thermal tolerances and preferences of the crab spiders Misumenops asperatus and Misumenoides formosipes (Araneae, Thomisidae) by V.R. Schmalhofer 470 Paraphyly of the Enoplognatha ovata group (Araneae, Theridiidae) based on DNA sequences by A.-M. Tan, R.G. Gillespie & G.S. Oxford 481 Cheap transport for fishing spiders (Araneae, Pisauridae): The physics of sailing on the water surface by R.B. Suter 489 Notes on the social structure, life cycle, and behavior of Anelosimus rupununi by L. Aviles & P. Salazar 497 Movement of the male brown tarantula, Aphonopelma hentzi (Araneae, Theraphosidae), using radio telemetry by M.E. Janowski-Bell & N.V. Horner 503 Phenology and life history of the desert spider, Diguetia mojavea (Araneae, Diguetidae) by A.M. Boulton & G.A. Polls 513 Host specificity and distribution of the kleptobiotic spider Argyrodes antipodianus (Araneae, Theridiidae) on orb webs in Queensland, Australia by P. Grostal & D.E. Walter 522 Abundance of spiders and insect predators on grapes in central California by M.J. Costello & K.M. Daane 531 Research Notes Unusual phenotype suggests role for homeotic genes in arachnid development by J.E. Garb 539 Nomenclature of the orb-web by S. Zschokke 542 On Sofanapis antillanca (Araneae, Anapidae) as a kleptoparasite of austrochiline spiders (Araneae, Austrochilidae) by M.J. Ramirez & N.I. Platnick 547 Carbohydrate analysis in spider hemolymph of selected lycosid and araneid spiders (Araneae: Lycosidae and Araneidae) by P.D. Barron, N.V. Horner & R.L. Cate 550 yiv_ /^(353 The Journal of ARACHNOLOGY- OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY VOLUME 27 1999 NUMBER 3 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF : James W. Berry, Butler University MANAGING EDITOR: Petra Sierwald, Field Museum ASSOCIATE EDITORS: Matthew Greenstone, USDA; Robert Suter, Vassar College EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E. Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C. Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ; C. D. Dondale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galia- no, Mus. Argentine de Ciencias Naturales; C. Griswold, Calif. Acad. Sci.; N. V. Horner, Midwestern State Univ; D. T. Jennings, Garland, Maine; V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ; N. I. Platnick, American Mus. Natural Hist.; G. A. Polis, Vanderbilt Univ; S. E. Riechert, Univ. Tennessee; A. L. Rypstra, Miami Univ, Ohio; M. H. Robinson, US. National Zool. Park; W. A. Shear, Hampden- Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa Rica. 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, $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 38677 USA. BUSINESS MANAGER: Robert Suter, Dept, of Biology, Vassar College, Pough- keepsie, New York 12601 USA. SECRETARY: Alan Cady, Dept, of Zoology, Miami Univ, Middletown, Ohio 45042 USA. ARCHIVIST: Lenny Vincent, Fullerton College, Fullerton, California 92634. DIRECTORS: David Wise (1998-2000), Paula Cushing (1999-2001), Ann Rypstra (1999-2001). HONORARY MEMBERS: C. D. Dondale, H. W. Levi, A. F. Millidge, W. Whit- comb. Cover photo: A green lynx spider {Peucetia viridens) holding a wasp by the antennae. As the spi- der held the wasp, two more wasps arrived and attempted to mate with the wasp. (Photo by Gail Stratton) Publication date: 21 December 1999 @ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 1999. The Journal of Arachnology 27:553-603 REVISION AND CLADISTIC ANALYSIS OF THE ERIGONINE SPIDER GENUS SISICOTTUS (ARANEAE, LINYPHIIDAE, ERIGONINAE) Jeremy A. Miller: Department of Biological Sciences, The George Washington University, Washington, D.C., 20052, and Department of Entomology, National Museum of Natural History, NHB-105, Smithsonian Institution, Washington, D.C. 20560 USA ABSTRACT. The erigonine spider genus Sisicottus is revised for the first time. Cladistic analysis of Sisicottus suggests the following hypothesis of interspecific relationships: ((5. montigenus, S. quoylei) {S. panopeus (S. montanus {S. crossoclavis {S. cynthiae (S. orites {S. nesides, S. aenigmaticus))))))). The monophyly of the genus is unambiguously supported by six putative synapomorphies: a terminal embolic hook, a suprategular membrane projecting apically from the distal suprategular apophysis, copulatory ducts that originate on the ectal side of the spermathecae, imbricated stridulatory striae, the presence of two dorsal macrosetae on tibia III, and the absence of a trichobothrium on metatarsus IV. Evidence for the monophyly of each Sisicottus species is discussed. A taxonomic key, diagnoses, descriptions, quantitative character values, illustrations, locality records, natural history information, and distribution maps are pre- sented for the nine recognized species. Five new species are described: S. quoylei, S. panopeus, S. crossoclavis, S. cynthiae, and S. aenigmaticus. Typhochrestus uintanus (NEW COMBINATION) is for- mally transferred out of Sisicottus. The genus Sisicottus Bishop & Crosby 1938 (Linyphiidae) is a lineage of small to medium- size erigonine spiders made up of nine known species. Sisicottus species usually live in moss and litter in conifer forests where they presumably build small prey capture webs. The genus is known from North Amer- ica north of Mexico and from the Kuril Is- lands between Japan and the Kamchatka Pen- insula. It is most diverse in the northwestern United States and southwestern Canada. By 1995, Sisicottus was one of 534 valid linyphiid genera (Platnick 1997). However, it is one of few linyphiid genera defined explic- itly by putative synapomorphies. At best, most linyphiid genera seem to be delimited to pre- serve homogeneity among members. The quality of systematics as an information stor- age and retrieval system is undermined when genera are circumscribed without due consid- eration of evidence in support of monophyly. The original diagnosis of Sisicottus was in- adequate to prevent Sisicottus from serving as a polyphyletic wastebasket. Species once placed in Sisicottus are currently placed in four different genera. The erroneous place- ment of some species in Sisicottus appears to have been based on understandable misinter- pretations of homology. In other cases, place- ment in Sisicottus seems inexplicable. Some morphological features in Sisicottus exhibit a range of evolutionary plasticity. The distal suprategular apophysis of the male pal- pus is recognizably different in every Sisicot- tus species. The shape of the male palpal tibia and the dorsal plate of the female epigynum exhibit slightly less interspecific variation. Other characters, such as the form of the male paracymbium, the marginal suprategular apophysis of the palpus, the embolic division, and the path of the female copulatory ducts are nearly invariant within the genus. If one gives primacy to such characters, Sisicottus does in fact comprise a homogeneous group of species. But relying on intuition to predict which characters will be stable and which will be homoplastic is unscientific. Shared derived similarity (synapomorphy) is the evidence upon which monophyletic groups are recog- nized. Synapomorphies are best discovered by incorporating as much comparative data as possible into a cladistic analysis. The task of bringing phylogenetic order to the chaos that is linyphiid systematics is a monumental one, 553 554 THE JOURNAL OF ARACHNOLOGY but its reward will be a hierarchical structure based on repeatable methods and explicit character evidence. Currently, even most pro- fessional spider systematists can identify lin- yphiids only with difficulty or not at all. An active dialog on the comparative morphology of linyphiids and the history of character states will lead to a phylogenetically based and usable taxonomy. An improved taxonomy will facilitate the communication of ideas and findings concerning this diverse and important spider family. Such communication will not be limited to professional spider systematists, but will include ecologists, biogeographers, conservationists, and amateur taxonomists. TAXONOMIC HISTORY Bishop & Crosby (1938) established Sisi- cottus to accommodate Tmeticus montanus Emerton 1882 and a new species, Sisicottus montigenus Bishop & Crosby 1938. They di- agnosed the genus based on characteristics of the male palpus. According to the original de- scription, the two founding species shared similar dorsomesal tibial apophyses, a lamella character! Stic a (misidentified as a radical tail- piece) described as “bulb-like,” and an open- coiled embolus making one turn about the dis- tal end of the bulb (Bishop & Crosby 1938: 57). Bishop & Crosby synonymized Erigone collina Marx 1890, Grammonota orites Chamberlin 1919, Oedothorax nesides Cham- berlin 1921, and Oedothorax pidacitis Crosby & Bishop 1927 under S. montanus. Although Bishop & Crosby recognized that a “larger and usually somewhat paler” (Bishop & Cros- by 1938:59) form existed sympatrically in the west with the smaller form of S. montanus typical of eastern populations, they were un- able to come up with a reliable way of sepa- rating the two morphs. The western form al- luded to is undoubtedly S. orites and/or S. nesides. Both of these species are common and widespread in the west and are larger than S. montanus. However, the observation that these western species are typically paler than S. montanus is erroneous. Working near the end of the 19^^ century, Marx objected to the trend begun by his con- temporaries Menge, Emerton, and Simon, of splitting Erigone Audouin 1826 into the many smaller genera that now comprise the Erigon- inae. Instead of transferring previously de- scribed species into Erigone, Marx’s (1890) catalog features several replacement names for valid species. One example of this was E. collina, which was meant to replace T. mon- tanus. Chamberlin & Ivie (1933) synonymized Oedothorax pidacitis under Grammonota or- ites and transferred the species to Oedothorax Bertkau 1883, making it congeneric with the very similar O. nesides. Shortly after the es- tablishment of Sisicottus as a new genus, Chamberlin & Ivie (1939) rejected the broad definition of S. montanus and re-elevated S. orites and S. nesides to species status. Al- though they wrote no justification for their de- cision, they did illustrate the dorsal view of the male palpal tibiae of S. orites, S. nesides, and montanus. Chamberlin & Ivie also placed two new species in Sisicottus: S. uin- tanus Chamberlin & Ivie 1939 and S. cor- nuella Chamberlin & Ivie 1939. Holm (1967) suggested that S. uintanus be transferred to Typhochrestus Simon 1884 based on a comparison with T. pygmaeus (Sprensen 1898), which is not the type species of the genus. He commented on the superficial similarity of an embolus coiled around a straight apophysis shared by S. uintanus, S. montanus, and T. pygmaeus, but realized that the apophysis of S. uintanus and T. pygmaeus arises from the embolic division whereas the apophysis in Sisicottus arises from the supra- tegulum or “median apophysis” in Holm’s terminology. Although this change has been adopted by some authors (Buckle et al. 1994), Holm’s statement was too tentative to meet Platnick’s (1989) criteria for formal transfers. Sisicottus cornuella was transferred to WaU ckenaeria Blackwall 1833 by Millidge (1983) apparently based on characteristics given in his definition of Walckenaeria (e.g., sternum longer than wide, distinctly sclerotized pedi- cel, fourth metatarsal trichobothrium, strongly pectinate tarsal claws) and on the presence of a short horn on the male carapace and details of the male and female genitalia typical of the minuta group of Walckenaeria species. Sisicottus atypicus Chamberlin & Ivie 1944 was described from the male only. Sciastes ogeechee Chamberlin & Ivie 1944 was de- scribed from a single female in the same pa- per. These two species were later found to be conspecific and were synonymized under Souessoula parva (Banks 1899) by Ivie MILLER— REVISION OF SISICOTTUS 555 (1967). This species lacks nearly all of the characteristics that distinguish Sisicottus from other erigonines including a coiled embolus. Sisicottus hibernus Barrows 1945 is a very unusual species that was inexplicably de- scribed as a Sisicottus. It shares none of the synapomorphies that define Sisicottus and was transferred to Carorita Duffey & Merrett 1963 by Zujko-Miller (1999) based on the re- sults of a phylogenetic study. METHODS Abbreviations for anatomical structures and quantitative characters are listed in Table 1. Abbreviations for specimen collections are found in the acknowledgments. Boundaries for quantitative characters are illustrated in Figs. 1, 2, 5-7, and 22-24. All measurements are in mm. Light microscopy. — Measurements were performed using a Leitz binocular dissecting scope with greenough objectives and an eye- piece micrometer scale in 20 X oculars. Five specimens were remeasured five times for each character during this study. This sam- pling indicated that the measurements are ac- curate to one micrometer unit for both powers of magnification used. Carapace length was measured at 80 X and one micrometer unit had a value of 0.024 mm. All other measurements were made at 200 X and one micrometer unit had a value of 0.0095 mm. Illustrations of ex- ternal structures were drawn using a 20 X 20 ocular grid at 200 X. For observation of inter- nal structures, specimens were cleared in methyl salicylate (Holm 1979) and illustrated using an Olympus BH-2 compound micro- scope at 400 X fitted with a camera lucida. Cleared specimens were positioned for illus- tration using the method described by Cod- dington (1983). Male palpus drawings and scanning electron micrographs are from the left appendage unless otherwise indicated. Specimens in which tracheal structures were to be viewed had windows cut in the dorsal integument of the carapace and abdomen. Specimens were digested in dilute sodium hy- pochlorite (household bleach) at room tem- perature for several hours until all the non- chitinous parts had dissolved (Millidge 1984a). Chlorazol black was used to stain the tracheae. Illustrations were made using a Wild M-20 compound microscope fitted with a camera lucida. Electron microscopy. — Scanning electron microscopy (SEM) was conducted using a Jeol 35U at Clemson University and an Am- ray 1810 at the National Museum of Natural History (Smithsonian Institution). Male and female specimens representing most Sisicottus species were examined using SEM. I was un- able to examine Sisicottus aenigmaticus new species or males of S. quoylei new species. Male and female specimens of Typhochrestus uintanus and T. digitatus (O. Pickard-Cam- bridge 1872) were also examined. This last species was represented by specimens pre- pared by G. Hormiga according to methods described in Hormiga (in press). I used SEM to observe male and female genitalia and spin- neret spigot morphology. Spinneret spigots were identified using Coddington (1989). Ab- domens and some genitalia were prepared for SEM by taking them through a rehydration series, placing them in a buffered 2.5% glu- taraldehyde solution for 48 hours, then dehy- drating to 100% ethanol. Specimens were then ultrasonicated for up to one minute. Ethanol was then removed either by critical point dry- ing in a Seevac CPD-100 or preparation in hexamethyldisilazane for five minutes (Poly- sciences, Inc., CAT #0629). Some genitalia were simply ultrasonicated, dehydrated in 100% ethanol, placed in hexamethyldisila- zane, air dried, and mounted. Quantitative characters. — Quantitative characters were selected on the basis of their estimated potential utility in distinguishing species and groups of species. Quantitative character values for samples of each species (Tables 2, 3) and for the type specimens alone (Table 4) are an important part of each de- scription. Descriptions. — Species descriptions draw from as many individuals over as wide a geo- graphic range as possible. This approach was chosen in order to account for as much intra- specific variation as possible. Some illustra- tions required the examination of multiple specimens to achieve a clear interpretation of the anatomy. Because of the relative difficulty involved in distinguishing some species, the diagnostic section of each species description has been designed to convey all available in- formation that might be relevant to accurate species identification. Descriptions highlight the unique characteristics of each species and Table 1. — Anatomical and quantitative abbreviations used in text, figures, and tables. For morphometric characters, maximum lengths were recorded unless otherwise specified, # indicates a quantitative character. 556 THE JOURNAL OF ARACHNOLOGY W 00 H ^ ^ 9 ^ QfcCLiHH SohC/:iPhDhPh ooooooooooooHHHH>>> §• O T3 N ^ CH- O ^ O 1/3 3|r 1-1 c/3 O O to to T3 X} a a Tl) a> o ^ o CH- a ^ § iH 1-1 O c3 S I c o a o C/D o § -o S O O P O o P O u O C/D cd (D P o CH- ^ - ■> w p P TJ o ^ ’O a (/) S -u « p ^ o ^ S S .2 Dh X) P -73 O P 2 -S p p S ^ Os o O X O (U CH- 03 CH W3 r* o § 'd p CH § :3 S .22 S a -p n Cis -g ^ h-H P P O c/) 0+ ’X ^ ^ o .2 p 3 ~ Ui "^2 g ^ P D o E a ^ XS o a ^ ^ Oh Hi O o o X o "d a a ^ p d) H liO Oh Oh .2 p - _N W O a o "dog flj H O a &x)x ^ 5 p X -Ol X O E ‘h p p 2 05 d Q ^dHeuaHOHaHC/5 <<< 0.12 mm; Fig. 94, FTP, PTA) orites (Chamberlin 1919) 5 (1). Distal suprategular apophysis extends to ventral midline of palpal bulb (Fig. 19). Palpal tibial apophysis very short (palpal tibial apophysis length < 0.04 mm; Fig. 20, PTA). Eastern North America 6 Distal suprategular apophysis extends about half way down palpal bulb (Fig. 61). Palpal tibial apophysis longer (palpal tibial apophysis length > 0.04 mm; Fig. 62, PTA). Asia, eastern or western North America 7 6 (5). Distal suprategular apophysis membranous; inner margin strongly convex; terminal margin ser- rated (Figs. 19, 26). North Carolina and Tennessee montigenus Bishop & Crosby 1938 Distal suprategular apophysis moderately sclerotized; inner margin sinuous; terminus a sharp apex (Fig. 34). Northeastern United States, southeastern Canada .......... quoylei new species 7 (5). Palpal tibial apophysis short (palpal tibial apophysis length < 0.08 mm); ectal tibial process present (Fig. 62, FTP). Distal suprategular apophysis extends ventrally beyond level of supra- tegular membrane (Figs. 47, 61, DSA). Less than 6 macrosetae in cluster on ectal side of palpal tibia. Widespread in North America montanus (Emerton 1882) Palpal tibial apophysis long (palpal tibial apophysis length > 0.08 mm); ectal tibial process absent (Fig. 41), Distal suprategular apophysis does not extend ventrally beyond level of su- prategular membrane (Figs. 40, 45, DSA). More than 7 macrosetae in cluster on ectal side of palpal tibia. Western North America and Asia panopeus new species Females 1 . Ventral plate of epigynum with deep invagination (ventral plate invagination depth < 0.04 mm; Fig. 96, VP) .................... 2 Ventral plate of epigynum with invagination shallow or absent (ventral plate invagination depth < 0.04 mm; Figs. 63, 76, VP) ....... 6 2 (1). Posterior face of dorsal plate narrow (width of posterior face of dorsal plate < 0.12 mm), subrectangular with convex lateral margins (Fig. 37, DPP). Dorsal fold of dorsal plate mem- branous. Lateral margin of epigynal capsule straight (Fig. 38, CDC). Carapace small (carapace length < 0.80 mm). Eastern North America . 3 Posterior face of dorsal plate wide (width of posterior face of dorsal plate >0.12 mm), trian- gular or trapezoidal with straight to concave lateral margins (Fig. 97, DPP). Dorsal fold of dorsal plate sclerotized, trapezoidal, widest posteriorly. Lateral margin of epigynal capsule strongly bowed (Fig. 98, CDC). Carapace large (carapace length > 0.86 mm). Western North America 4 3 (2). Ventral plate invagination wide and very deep (ventral plate invagination width > 0.06 mm; ventral plate invagination depth > 0.09 mm; Fig. 22, VPIW, VPID) ............................................... montigenus Bishop & Crosby 1938 Ventral plate invagination narrow and only moderately deep (ventral plate invagination width < 0.05 mm, ventral plate invagination depth < 0.09 mm; Fig. 36) . quoylei new species 4 (2). Ventral plate invagination wide (ventral plate invagination width ca. 0.06 mm). Posterior face of dorsal plate trapezoidal (Fig. 107, DPP) . . aenigmaticus new species Ventral plate invagination narrow (ventral plate invagination width < 0.05 mm). Posterior face of dorsal plate triangular (Figs. 97, 105, DPP) .................................... 5 (4). Posterior face of dorsal plate with lateral margins never more than slightly concave (Fig. 97, 5 MILLER— REVISION OF SISICOTTUS 569 DPP). Epigynum long (epigynum length = 0.152-0.209 mm). Copulatory duct capsule wide (copulatory duct capsule width = 0.143-0.238 mm) orites (Chamberlin 1919) Posterior face of dorsal plate with concave lateral margins (Fig. 105, DPP). Epigynum short (epigynum length = 0.133-0.181 mm). Copulatory duct capsule narrow (copulatory duct cap- sule width = 0.100-0.176 mm) nesides (Chamberlin 1921) 6 (1). Dorsal fold of dorsal plate membranous (Fig. 65, DF) ............................... 7 Dorsal fold of dorsal plate sclerotized, trapezoidal, widest posteriorly (Fig. 71, DF) ......... 8 7 (6). Ventral plate enfolded forming broad groove (Figs. 53, 54, G). Posterior face of dorsal plate triangular with sharply pointed ventral margin (Fig. 64, DPP). Eastern or western North America montanus (Emerton 1882) Ventral plate without groove (Figs. 51, 52). Posterior face of dorsal plate subrectangular with flat ventral margin (Fig. 44, DPP). Western North America and Asia ..... panopeus new species 8 (6). Ventral margin of posterior face of dorsal plate dorsal to ventral extent of spermathecae (Fig. 77, DPP). Posterior margin of copulatory duct capsule oriented anteriorly (Fig. 78, CDC). Oregon cynthiae new species Ventral margin of posterior face of dorsal plate about level with ventral extent of spermathecae (Fig. 70, DPP). Posterior margin of copulatory duct capsule oriented posteriorly (Fig. 71, CDC). Western North America crossoclavis new species Sisicottus montigenus Bishop & Crosby 1938 Figs. 17-28, 31, 32 Sisicottus montigenus, in part: Bishop & Crosby 1938: 60-61, figs. 10-11 [3,9]. Roewer 1942: 650. Bonnet 1958: 4066. 3349 syntypes from UNITED STATES: North Carolina, Yancey County, Mt. Mitchell, 12 October 1923, in AMNH, examined. Diagnosis. — Males of S. montigenus are distinguished from those of all other Sisicottus species except its sister species, S. quoylei, by the form of the palpal tibial apophysis, tvhich is quite short with a flat distal margin and no ectal tibial process (Fig. 20, characters 16, 17). The form of the distal suprategular apophysis is unique among Sisicottus, being membranous and quite transparent, extending to near the ventral midline of the palpal bulb with variable amounts of serration along the margin, especially in the near terminus (Figs. 25-27, characters 3, 4). This characteristic is the most reliable way to distinguish this spe- cies from all other Sisicottus species including S. quoylei. Females of S. montigenus have a ventral plate invagination that is almost always deep- er than any other Sisicottus species (Fig. 22). They can be distinguished from other species with a deep ventral plate invagination, except S. quoylei, by the posterior face of the dorsal plate, which is subrectangular (Fig. 23, char- acter 21) and the membranous dorsal fold of the dorsal plate (Fig. 24, character 24). The ventral plate invagination is wider than any species except S. aenigmaticus (Table 3). Sis- icottus montigenus can be distinguished from all species except S. quoylei by the form of the anterior margin of the capsule which is nearly flat rather then formed into two convex lateral lobes (Fig. 24, character 30). Depth and width of the ventral plate invagination reliably separate S. montigenus from its sister species, S. quoylei (Fig. 31). Description.— Small (carapace length = 0.67-0.80 mm); coloration much darker than in other Sisicottus species. Distal suprategular apophysis of male palpus membranous, long, extends to near ventral midline of palpal bulb; with serrated outside and terminal margins (Figs. 25-27). Palpal tibia short; palpal tibial apophysis short with flat distal margin and no ectal tibial process (Fig. 20); sparse cluster of macrosetae (3-7) on ectal side of palpal tibia. Females with deep and wide invagination of ventral plate of epigynum (Figs. 22, 31). Pos- terior face of dorsal plate generally rectangu- lar, usually slightly taller than wide with a flat or slightly convex ventral margin (Fig. 23). Dorsal fold of dorsal plate membranous. Lat- eral margins of copulatory duct capsule in dorsal view sinuous with posterior tips of cap- sule oriented posteriorly; anterior margin of capsule nearly flat; fertilization ducts sinuous (Fig. 24). See Tables 2-4. Natural history. — Sisicottus montigenus has been collected exclusively in high eleva- tion spruce and Fraser fir forests of the south- ern Appalachian mountains. Over the last de- cade, much of this habitat has been severely 570 THE JOURNAL OF ARACHNOLOGY Figures 17-24. — Sisicottus montigenus with limits of some quantitative characters. 17-20, Palpus of male from Mt. Mitchell, North Carolina. 17, Mesal view; 18, Ventral view; 19, Fetal view; 20, Palpal tibia, dorsal view; 21, Suprategulum separated from palpus of male from Clingman’s Dome, North Car- olina, ectal view; 22, 23, Epigynum of female from Mt. Mitchell, North Carohna. 22, Ventral view; 23, Posterior view. 24, Cleared epigynum from Clingman’s Dome, North Carolina, dorsal view. Scales: Figs. 21, 24 = 0.05 mm; other figures = 0.1 mm. altered. The extensive damage to these forests has been blamed principally on the balsam wooly adelgid, Adelges piceae Ratzeburg (Homoptera: Adelgidae), an introduced Eu- ropean pest of firs. Sisicottus montigenus is usually associated with ground microhabitats, particularly mosses, but has also been taken on beating sheets. In one collection from Clingman’s Dome, North Carolina in 1977, S. montigenus was the most abundant spider spe- cies. However, my repeated collecting efforts at this now much altered locality (Wheeler & McHugh 1994; F. Coyle pers. comm.) have failed to produce additional specimens. This apparent rapid change in population size in response to environmental degradation sug- gests that this species may be worthy of con- sideration for federal protection as a threat- ened or endangered species. Distribution. — North Carolina and Tennes- see; restricted to high elevation spruce-fir for- ests (Fig. 32). A record from Michigan is al- most certainly erroneous (Drew 1967), although this site is not very far outside the currently recognized range of S. quoylei. Any Sisicottus that Drew may have collected ap- pear to be lost; no Sisicottus specimens can be found in the Michigan State University En- tomology Museum (R.J. Snider pers. comm.). Material examined. — UNITED STATES: North Carolina: Mitchell County., Roan High Bluff, 6200 feet, in moss from rocks in spruce-fir forest, 17 November 1978, 1(559 (F. Coyle & D. Pittillo, FAC); Swain County, Clingman’s Dome, 6600 feet, in moss from spruce-fir forest floor, 6 November 1977, 95269 (F. Coyle, FAC); Mt. Mitchell, 6600 feet, in moss from spruce-fir forest floor, 20 October 1977, 3512 9 (F. Coyle, FAC), Mt. Mitchell, 6600 feet, in moss from rock ledges in spruce-fir forest, 20 October 1977, 7 9 (F. Coyle, FAC). Tennessee: Sevier County, GSMNP, Mt. LeConte, 100 m below spring along Trillium Gap MILLER— REVISION OF SISICOTTUS 571 Figures 25-30. — Scanning electron micrographs of Sisicottus montigenus and S. quoylei. 25-28, S. montigenus from Mt. Mitchell, North Carolina. 25, Palpus, ectal view; 26, Terminus of distal suprategular apophysis; 27, Palpus, ventral view; 28, Epigynum, ventral view. 29, 30, Epigynum of female S. quoylei from Mt. MacIntyre, New York. 29, Ventral; 30, Posterior. 572 THE JOURNAL OF ARACHNOLOGY Tr., UTM: N394833 E27913, 6300 feet, beating 14: 20-15:20 in 25 yr-old fir forest, 19 July 1995, 1(33$ (Coyle, Williams & Carbiener, GSMNP); Mt. LeConte, 35°37'N, 83°27'W, 5c310$ (AMNH). Sisicottus quoylei new species Figs. 29-38 Sisicottus montigenus, in part: Bishop & Crosby 1938: 60-61, fig. 9 [6]. Roewer 1942: 650. Bon- net 1958: 4066. Types. — Male holotype with one female paratype from CANADA: Newfoundland, King’s Point, beating black spruce, 19 August 1984, L. Hollett, deposited in CNC. Etymology. — Named for the protagonist in E. Annie Proulx’s Pulitzer Prize winning nov= el. The Shipping News, which is set within the range of this species. The name is also a hom- onym of the patronymic that would result from a species named for Dr. Frederick A. Coyle, my master’s thesis advisor. Diagnosis. — Males of S. quoylei are distin- guished from those of all other Sisicottus spe- cies except its sister species, S. montigenus, by the form of the palpal tibial apophysis, which is quite short with a flat distal margin and no ectal tibial process (Fig. 35, characters 16, 17). It is separated from S. montigenus males by a moderately sclerotized distal su- prategular apophysis (membranous in S. mon- tigenus) which tapers to a single sharp ter- minal apex on its inside margin (Fig. 34, character 3). Females of S. quoylei are distinguished from all other Sisicottus species except S. montigenus by the form of the anterior margin of the capsule which is nearly flat rather than formed into two convex lateral lobes (Fig. 38, character 30). They are separated from S. montigenus by the form of the ventral plate invagination which is wider and deeper in 5. montigenus (Figs. 31, 38). Females of S. quoy- lei have a ventral plate invagination that is deeper than that of S. montanus, S. panopeus, S. crossoclavis, or S. cynthiae (character 19). They can be distinguished from other species with a deep ventral plate invagination, except S. montigenus, by having a dorsal plate with a subrectangular rather then triangular or trap- ezoidal posterior face (Fig. 37, character 21), and by the dorsal fold of the dorsal plate, which is membranous rather then sclerotized (Fig. 38, characer 24). Figure 31. — Scattergram of ventral plate invagi- nation depth plotted against ventral plate invagina- tion width for females of Sisicottus montigenus (•) and S. quoylei (A). Description. — Small (carapace length 0.71-0.82 mm); coloration typical (see de- scription section for Sisicottus). Distal supra- tegular apophysis of male palpus moderately sclerotized, long, extends to near ventral mid- line of palpal bulb; terminus pointed (Fig. 34). Palpal tibia short; palpal tibial apophysis short with flat distal margin without ectal tibial pro- cess (Fig. 35); sparse cluster of macrosetae (3-6) on ectal side of palpal tibia. Females with deep and narrow invagination of ventral plate of epigynum (Fig. 36). Posterior face of dorsal plate generally rectangular, usually slightly taller than wide with convex ventral margin (Fig. 37). Dorsal fold of dorsal plate membranous. Lateral margins of copulatory duct capsule in dorsal view sinuous with pos- terior tips of capsule oriented posteriorly; an- terior margin of capsule nearly flat; fertiliza- tion ducts sinuous (Fig. 38). See Tables 2-4. Natural history. — Locality labels indicate that this species is associated with conifer for- ests, especially balsam fir. It is sympatric with S. montanus over at least part of its range. Distribution.— New Brunswick, New- foundland, Nova Scotia, and New York (Fig. 32). Material examined. — CANADA^ New Bruns- wick: Green River, 30 mi N. Edmundston, balsam fir, 6 June 1959, 131? (TR. Renoult, CNC), bal- sam fir, 30 June 1965, 13 (TR. Renoult, CNC); Kedgwick River, balsam fir, 27 June 1966, 1 ? (TR. Renoult, CNC). Newfoundland: Bottom Brook, beating balsam fir, 15 August 1984, 13 (L. Hollett, CNC); Gallants, fir foliage, 23 June 1982, 13 (K.P. Lim, CNC); Hampden, beating balsam fir, 14 June MILLER— REVISION OF SISICOTTUS 573 Figure 32. — Eastern United States and southeastern Canada, showing distribution of Sisicottus monti- genus (★) and S. quoylei (A). 1977, 19 (CNC); Noel Pauls Brook, ex Abies bal- samea, 26 June 1977, 1 9 (L. Hollett, CNC), ex Abi- es balsamea, 8 August 1977, 1(3 (L. Hollett, CNC), Abies balsamea, July 1984, 1 9 (L. Hollett, CNC); Paddys Brook, 10 mi. W St. Johns, November 1982-April 1983, 1 9 (D.W. Langer, CNC); Portland Creek, June 1974, 19 (Heinrich, CNC); Steady Brook, 48°57'N, 57°50'W, beating balsam fir, 17 August 1984, 1 9 (L. Hollett, CNC); St. Fintans, 30 June 1942, 1 9 (E.J. Gillan, CNC). Nova Scotia: Figures 33-38. — Sisicottus quoylei from King’s Point, Newfoundland. 33, Palpus, ventral view; 34, Palpus, ectal view; 35, Palpal tibia, dorsal view; 36, Epigynum, ventral view; 37, Epigynum, posterior view; 38, Cleared epigynum, dorsal view. Scales: Fig. 38 = 0.05 mm; all others = 0,1 mm. 574 THE JOURNAL OF ARACHNOLOGY Figures 39-44. — Sisicottus panopeus from Mt. Rainier, Washington. 39-42, Male palpus. 39, Ventral view; 40, Fetal view; 41, Palpal tibia, dorsal view; 42, Suprategulum separated from palpus, ectal view. 43, 44, Epigynum. 43, Ventral view; 44, Posterior view. Scales: Fig. 42 = 0.05 mm; all others = 0.1 mm. Cape Breton Highlands National Park, North Mountain, 46°48'N, 60°4rw , ex fen-pans, 8 June 1983, 19 (H. Goulet, CNC). UNITED STATES: New York: Lake Tear, Mt. Marcy, 4 September 1922, IS (Bishop, AMNH); Mt. MacIntyre, 1 July 1923, IS 19 (Crosby, AMNH). Sisicottus panopeus new species Figs. 39-46, 51, 52, 57, 58 Sisicottus montanus: Lowrie & Gertsch 1955; 6 (misidentification). Holm 1960: 124 (misidentifi- cation). Bragg & Leech 1972: 69 (misidentifica- tion). Crawford 1988; 15 (in part). Crawford & Edwards 1988: 437, figs. 21-22 [9] (misidenti- fication). Platnick 1993: 351. Types. — Male holotype from UNITED STATES: Washington, Mt. Rainier National Park, Paradise, 46°48'N, 12r44'W, 12 Sep^ tember 1965, J. & W. Me, deposited in AMNH. Etymology. — Derived from the monotypic mollusc genus Panope; P. generosa, the geo- duck clam, is the mascot of my alma mater, The Evergreen State College. Diagnosis. — Males of S. panopeus share with S. montanus and S. crossoclavis a distal suprategular apophysis that extends about half way down the ectal side of the palpal bulb (Fig. 40, character 4); in all other Sisicottus species, the distal suprategular apophysis ex- tends to near the ventral midline of the papal bulb. They are distinguished from S. montan- us by their longer palpal tibia and palpal tibial apophysis (Figs. 41, 57), by the lack of an ectal tibial process (Fig. 41, character 17), by the presence of more macrosetae in their ectal tibial cluster (7-1 1 in S. panopeus, 2-6 in S. montanus), and by the form of the distal su- prategular apophysis which projects ventrally past the level of the suprategular membrane in S. montanus (Figs. 47, 61) but stops near the level of the suprategular membrane in S. panopeus (Figs. 40, 45). Sisicottus crosso- clavis can be distinguished from both of these species by its heavily sclerotized distal supra- tegular apophysis with a serrated terminal margin (Fig. 67, character 3); S. panopeus and S. montanus have a moderately sclerotized distal suprategular apophysis with a rounded terminal margin (Fig. 40). MILLER— REVISION OF SISICOTTUS 575 Figures 45-50. — Scanning electron micrographs of Sisicottus palpi. 45, 46, S. panopeus from Mt. Rain- ier, Washington. 45, Ectal view; 46, Ventral view. 47, 48, S. montanus from Mt. Mansfield, Vermont. 47, Ectal view; 48, Ventral view. 49, 50, S. crossoclavis from Rabbit Creek, Washington. 49, Ectal view detailing distal suprategular apophysis; 50, Ventral view. 576 THE JOURNAL OF ARACHNOLOGY Figures 51-56. — Scanning electron micrographs of Sisicottus epigyna. 51, 52, S. panopeus from Lake Louise, Alberta. 51, Ventral view; 52, Posterior view. 53, 54, S. montanus from Piscataquis County, Maine. 53, Ventral view; 54, Posterior view. 55, 56, S. crossoclavis from Deemer Creek, Washington. 55, Ventral view; 56, Posterior view. MILLER— REVISION OF SISICOTTUS 577 Figure 57. — Scattergram of palpal tibial apoph- ysis length plotted against palpal tibia length for males of Sisicottus panopeus (•), S. montanus (A), and S. crossoclavis (X). Females of S. panopeus, like those of S. montanus, S. crossoclavis, and S. cynthiae, have a shallow ventral plate invagination (Fig. 43, character 19); this separates them from fe- males of S. montigenus, S. quoylei, S. orites, S. nesides, and S. aenigmaticus . Sisicottus panopeus is unique among species with a shallow ventral plate invagination in having a dorsal plate with a posterior face that is sub- rectangular with a flat ventral margin (Fig. 44, character 21). Sisicottus panopeus can also be distinguished from S. crossoclavis, S. cyn- thiae, S. orites, S. nesides, and S. aenigmati- cus by the form of the dorsal fold of the dorsal plate which is membranous in S. panopeus (cf. Fig. 65, character 24) instead of sclerotized (Fig. 71). Description. — Medium-sized (carapace length 0.67-0.96 mm); coloration typical (see description section for Sisicottus). Distal suprategular apophysis of male palpus mod- erately sclerotized, of moderate length, ex- tends about half way down ectal side of palpal bulb; rounded on inside margin; terminal mar- gin about level with suprategular membrane (Figs. 40, 45). Palpal tibia moderately long with a long apophysis; ectal tibial process ab- sent (Fig. 41); dense cluster of macrosetae (7- 11) on ectal side of palpal tibia. Females with ventral plate invagination shallow to absent (Figs. 43, 51). Posterior face of dorsal plate subrectangular with flat ventral margin (Figs. 44, 52). Dorsal fold of dorsal plate membra- nous. Lateral margins of copulatory duct cap- sule in dorsal view sinuous with posterior tips of capsule oriented posteriorly; anterior mar- gin of capsule formed into two convex lateral lobes; fertilization ducts sinuous {cf. Fig. 65). Internal structure of epigynum virtually iden- tical to that of S. montanus (Fig. 65). See Ta- bles 2-4. Natural history. — Sisicottus panopeus and S. nesides were found in the same vial or iden- tically labeled vials several times during this study. Crawford & Edwards (1988) observed that in Washington these two species occupy distinct ecological niches, with S. panopeus apparently restricted to alpine and subalpine habitats and S. nesides more common at lower elevations. These two species are syntopic, however, where their ecological ranges over- lap at or just below the tree line. Sisicottus panopeus has also been collected syntopically Figure 58. — North America with inset of the Kuril Islands, showing distribution of Sisicottus panopeus (★) and S. montanus (A). 578 THE JOURNAL OF ARACHNOLOGY with S. montanus, S. crossoclavis, and S. or- ites. Collection labels indicate that this species may be found in moss, bogs, meadows, and forests, especially those with a conifer com- ponent. However, Sisicottus panopeus from the Kuril Islands occur in an ecological niche that is unusual for Sisicottus since many of the records come from islands that are completely unforested and occasionally devoid of even shrub vegetation (R. Crawford pers. comm.). Distribution.-— Wyoming, Montana, Wash- ington, Alberta, British Columbia and Alaska. Recent collections of this species from the Kuril Islands (between Japan and the Kam- chatka Peninsula) make it the only Sisicottus species known from Asia (Fig. 58). Further collections in Asia may well yield additional records of S. panopeus. Material examined. — CANADA: Alberta: Bow Lake, Banff NatT Park, elev. 6400 feet, moss and willow litter nr lake, 9 August 1973, 1 $ (E.E. Lind- quist, CNC); Bow Pass, 64 mi. NW Banff, Berlese spruce duff, 12 October 1953, 1629 (O. Peck, CNC); Cathedral Prov. Park, Pyramid L. Trail, 5-9 July 1986, 1619 (S.G. Cannings, CNC); Mt. Edith, Cavill Lodge, 52°4rN, 118°09'W, 24 August 1965, 1 $ (J. & W. Ivie, AMNH); Highwood Pass, 35 mi. S Kananaskis, ex spruce-larch litter, 28 July 1970, 1 9 (E.E. Lindquist, CNC); Kananaskis P.P., Mt. In- defatigable, lichens on stones in lodgepole pine for- est at base of mountain, 20 July 1983, 2619 (V. Behan, CNC); Lake Louise, 4 August 1927, 26 (Crosby, AMNH); Marmot Cr., 13 mi. SW Kan- anaskis F.E.S., 6000 feet, ex damp moss cover on forest floor, 20 August 1970, 1 (3 1 9 (E.E. Lindquist, CNC); Cameron Lake, Waterton Lakes NatT Park, interception trap, 4-17 July 1980, 2(349 (H.J. Tes- key, CNC), Crandell L. trail, mossy area on N. slope in mixed woods, 13 June 1980, 3 9 (I.M. Smith, CNC), Rowe Lake Trail, 6300 feet, sifting moss, edge of stream, 7 June 1980, 5 9 (J.M. Camp- bell, CNC). British Columbia: Manning Prov. Pk., Skyline Trail, 1768 m, Phyllodose, Kalmia, moss, lichen, 12 July 1986, 3 9 (V. Behan, CNC); Nitinat, Heather Mtn., V.I., ca. 3600 feet, moss on seepage slope, 14 July 1979, 3 9 (I.M. Smith, CNC), moss at cold seepage area, 27 July 1979, 49 (I.M. Smith, CNC); Strathcona Park, Vancouver Is., Cream Lake, 1260 m, moss litter, 18 August 1988, 2 9 (C. Guppy, CNC); Yoho Glacier, 5 August 1914, 1319 (Emerton, AMNH); Yoho Glacier Camp, 5 August 1914, 637 9 (Emerton, MCZ). RUSSIA: Kuril Is- lands: Ekarma Island, E side Cape Shpileroi, 4 m, 48.958°N, 153.920°E, in grass litter of beach mead- ow, 10 August 1996, 136 9 (TW. Pietsch, UWBM); Kharimotan Island, Inland from Severgi- na Bay, 15 m, 49.159°N, 154.478°E, ex sphagnum and other mosses around dry interdune wetlands, 8 August 1996, 5 9 (R. Crawford, UWBM); Matua Island, army base, east end of island, 25 m, 48.068°N, 153.257°E, ex alder and grass litter in thickets of Alnus maximowiczU, 14 August 1996, 135 9 (R. Crawford, UWBM); Ohirinkotan Island, E side Cape Ptichy, NW comer of island, 15 m, 48.986°N, 153.472°E, litter of tall grass meadow on steep slope -treeless, 10 August 1996, 5 9 (TW. Pietsch, UWBM); Onekotan Island, slope above Tmndi River, 90 m, 49.280°N, 154.749°E, ex alder thicket litter in coastal slope meadow, 9 August 1996, 2329 (R.L. Crawford, T Pearce, UWBM); Onekotan Island, 2 km S of Cape Subbotyna -river valley, 49.396°N, 154.646°E, ex herbaceous litter in river valley, 5 August 1996, 29, (T. Pearce, UWBM); Paramushir Island, SW shore, Sholikhoua Bay, 50°22'N, 155°37'E, 13-25 August 1996, 13 (Y. Marusik, UWBM); Shiashkotan Island, Zakat- naya Bay, 20 m, 48.778°N, 54.036°E, ravine in coastal slope meadow ex litter Alnus maximowiczii, Sorbus samburifolia, 11 August 1996, 5317 9 (R. Crawford, UWBM); Ushishir Island, Kratemaya Bay (central peninsula), 5-20 m, 47.5 10°N, 152.8 15°E, ex litter of Petasites patch in north ex- posed steep grass meadow, 20 August 1995, 233 9 (Y Mamsik, UWBM). UNITED STATES: Alaska: Aleutian IsL, Umnak, Fox Islands, July 1958, 435 9 (C. Lindroth, MCZ); Lituya Bay, Glacier Bay National Monument, Mt. Blunt, subalpine, 58.630°N, 137.493°W, 2100 feet, sifted from moss in shmbland, 9 August 1979, 1319 (D.H. Mann, UWBM). Montana: Glacier National Park, Swift- current Mountain, 7500 feet, 19 August 1953, 19 (Levi, MCZ); Washington: Clallam County, Olym- pic National Park, Waterhole Camp, 4975 feet, 47.944°N, 123.425°W, pitfalls in spring meadow, 30 July-8 August 1986, 11339 (R. Crawford, UWBM); Olympic NatT Park, Obstmction Peak, 5900-6000 feet, 3 August 1973, 3349 (A. Sme- tana, AMNH); King County, Source Lake 3760- 3840 feet, 47.455°N, 121.45rw, under rock up slope from lake adjacent to snowfield, 2 August 1986, 13 (R. Crawford, UWBM); Okanogan Coun- ty, Cold Spr Camp, 1850 m, 48.938°N, 119.789°W, ex rotten log, 3 August 1985, 1319 (R. Crawford, UWBM); Pend Oreille County, Deemer Creek, 4600 feet, 48.93 1°N, 1 17.089°W, sifted from willow litter in bog, 13 June 1986, 19 (R. Crawford, UWBM); Pierce County, Mt. Rainier National Park, Golden Gate 6400 feet, 46.799°N, 121.722°W, 2 pit- falls -heather and sedge meadow, 25 August-6 Sep- tember 1975, 19 (D.H. Mann, UWBM); Paradise Camp, Mt. Rainier, 19 August 1927, 2369 (Cros- by, 1927); Paradise, Mt. Rainier National Park, 46°48'N, 12U44'W, 12 September 1965, 10313 9 (J. & W Ivie, AMNH); Pierce County, Bearhead Mtn., 6000-6089 feet, 47.023°N, 121.814°W, ex heather and under rock, 15 August 1982, 5 9 (R. MILLER— REVISION OF SISICOTTUS 579 Figures 59-65. — Sisicottus montanus. 59-62, Palpus of male from Mt. Washington, New Hampshire. 59, Mesal view; 60, Ventral view; 61, Ectal view; 62, Palpal tibia, dorsal view. 63, 64, Epigynum of female from Mt. Washington, New Hampshire. 63, Ventral view; 64, Posterior view. 65, Cleared epigynum of female from near Soubunge Mountain, Pascataquis County, Maine, dorsal view. Scales: Fig. 65 =.05 mm; other figures = 0.1 mm. Crawford, UWBM); Skagit County, Coney Pass, 3400 feet, 48.329-331°N, 12L736°W, swept -sub- alpine meadow, 27 July 1980, Id, (R. Crawford, UWBM); Skagit County, Dock Butte, 5000 feet, 48.640°N, 122.803°W, under rocks and wood, 13 September 1986, 6dll$ (R. Crawford, UWBM); Snohomish County, Box Mtn. Lake, 5050 feet, 48.223°N, 12L121-3°W, under wood on sand/mud shore, 5 August 1989, 2$ (R. Crawford, UWBM). Wyoming: Grand Canyon, Yellowstone Park, 30 August 1927, 26 (Crosby, AMNH); Teton Park, Holly Lake, 9400 feet, 43°N, 110°W, 10 August 1950, 165 9 (D.C. Lowrie, AMNH); Togwatee Pass, 10,000 feet, 43°N, 110°W, 8 August 1950, 1 9 (D.C. Lowrie, AMNH). Sisicottus montanus (Emerton 1882) Figs. 47, 48, 53, 54, 57-65 Tmeticus montanus Emerton 1882: 55, fig. pi. xvi, fig. 3 [6,9]. Male lectotype from UNITED STATES: New Hampshire, Mt. Washington, 13 June 1877, J.H. Emerton, in MCZ, examined. Erigone collina Marx 1890: 533, 538, 593 {nomen novum). Synonymy by Bishop & Crosby 1938. 580 THE JOURNAL OF ARACHNOLOGY Oedothorax montanus: Crosby 1905: 312; Petrunk- evitch 1911: 264. Gongylidium montanus: Emerton 1920: 315. Sisicottus montanus: Bishop & Crosby 1938: 57— 60, figs. 6-8 [6 91 Chamberlin & Ivie 1939, fig. 40 [61 Roewer 1942: 650. Bonnet 1958: 4065. Holm 1967: 61. Kaston 1981: 208-209, figs. 653-657 [6,9]. West et al. 1984: 87. Koponen 1987: 281-283, 285. Crawford 1988: 15 (in part). Jennings et al. 1988: 61, 63. Aitchison-Benell & Dondale 1990: 224. Dondale et al. 1997: 89. Plat- nick 1993: 351 (after Crawford & Edwards 1988, misidentification); 1997: 427. Diagnosis. — Males of S. montanus are dis- tinguished from those of all other Sisicottus species by the form of the palpal tibia; the palpal tibial apophysis is tapered, ectally curved, and longer than that of S. montigenus and S. quoylei and the palpal tibia is shorter than that of S. panopeus, S. crossoclavis, S. cynthiae, S. orites and S. nesides (Figs. 57, 62, character 16). Sisicottus montanus males share with S. panopeus and S. crossoclavis a distal suprategular apophysis that extends about half way down the ectal side of the palpal bulb (Fig. 61, character 4); all other Sisicottus spe- cies have a distal suprategular apophysis that extends to near the ventral midline of the pal- pal bulb. Sisicottus montanus is distinguished from S. panopeus and S. crossoclavis by the presence of an ectal tibial process in S. mon- tanus (Fig. 62, character 17). Also, the form of the distal suprategular apophysis is very different in these species. Sisicottus crosso- clavis has a heavily sclerotized distal supra- tegular apophysis with a serrated terminal margin (Fig. 67, character 3). Sisicottus mon- tanus and S. panopeus both have a moderately sclerotized distal suprategular apophysis (character 3) but in S. montanus, the distal su- prategular apophysis projects ventrally past the level of the suprategular membrane (Fig. 61) while in S. panopeus, the distal suprate- gular apophysis projects at most only slightly beyond the level of the suprategular mem- brane (Fig. 40). Sisicottus panopeus and S. crossoclavis also have more macrosetae in their ectal tibial cluster (7-11 in S. panopeus', 7—9 in S. crossoclavis) than does S. montanus (2-6). Females of S. montanus, like those of S. panopeus, S. crossoclavis, and S. cynthiae, have a shallow ventral plate invagination (Fig. 63, character 19); this separates them from fe- males of S. montigenus, S. quoylei, S. aenig- maticus, S. orites, and S. nesides. Sisicottus montanus is unique among species with a shallow ventral plate invagination in having a groove formed by the enfolding of the ventral plate (Figs. 53, 54). Sisicottus montanus can also be distinguished from S. panopeus by the form of the posterior face of the dorsal plate which is pointed ventrally in S. montanus (Fig. 64, character 21) and is flat ventrally in S. panopeus (Fig. 44). Sisicottus montanus can also be distinguished from S. crossoclavis, S. cynthiae, S. orites, S. nesides and S. aenig- maticus by the form of the dorsal fold of the dorsal plate, which is membranous in S. mon- tanus (Fig. 65, character 24) instead of scler- otized (Fig. 71). Description. — Medium-sized (carapace length = 0.67-0.96 mm); coloration typical (see description section for Sisicottus). Distal suprategular apophysis of male palpus mod- erately sclerotized, of moderate length, ex- tends about half way down ectal side of palpal bulb; rounded on inside margin; terminal mar- gin ventral to level of suprategular membrane (Figs. 47, 61). Palpal tibia short with moder- ately short palpal tibial apophysis (Fig. 62); small ectal tibial process present; sparse clus- ter of macrosetae (2-6) on ectal side of palpal tibia. Females with ventral plate invagination shallow to absent (Figs. 53, 63). Posterior face of dorsal plate triangular, widest near its dor- sal margin with sharply rounded or pointed ventral apex (Fig. 64). Ventral plate enfolded forming a groove (Figs. 53, 54). Lateral mar- gins of copulatory duct capsule in dorsal view sinuous with posterior tips of capsule oriented posteriorly; anterior margin of capsule formed into two convex lateral lobes; fertilization ducts sinuous (Fig. 65). See Tables 2-4. Natural history. — Aitchison-Benell & Dondale (1990) have reported that S. montan- us in Manitoba may be found in boreal forest, bogs, ditches, mixed woods, leaf litter, moss and grass. Jennings et al. (1988) found that S. montanus in Maine preferred uncut spruce-fir forest habitats over clearcut strips. In a study of spiders living in ground habitats across an ecological/elevational gradient in Quebec, Ko- ponen (1987) found S. montanus to be the dominant species in his collections from bal- sam fir forest at 850 m elevation. In this study, S. montanus was more rarely found in three other habitats: a mixed deciduous forest site MILLER— REVISION OF SISICOTTUS 581 at 580 m elevation, a windy high elevation (920 m) scrub forest habitat near the tree line, and a somewhat sheltered summit below the tree line (870 m) with short birch and spruce trees. Collection labels record S. montanus from elevations of sea level to above tree line. This species is often associated with moss, bogs, and forest litter, especially from forests with a conifer component. It has been col- lected syntopically with S. panopeus, S. orites, and S. nesides and is also sympatric with S. quoylei. Distribution. — Canada, New England, Alaska, Washington, and the Rocky Moun- tains region south to Arizona (Fig. 58). Material examined. — CANADA: Alberta: Lake Louise, 4 August 1927, 3 (5169 (Crosby, AMNH); Sulfur Mt., Banff, 2 August 1927, IcJ (Crosby, AMNH); Lodgepole Pine Cpgd area, 1 mi. S Elk- water, Cypress Hills Prov. Pk., ex moist herbal mate substrate by seepage, 20-27 July 1978, 19, (E.E. Lindquist, CNC); House R. at Little Smoky River, 55°27'N, 117°10'W, 6 September 1968, ld3 9, (W. Ivie, AMNH); Jasper, 18 August 1914, lc3 (Emer- ton, MCZ); White Court, 54°08'N, 115°4rw, 6 September 1968, 3c37 9 (W. Ivie, AMNH). British Columbia: Metlakatla, 1(319 (Emerton, AMNH); Summit Lake, pitfall in moss above tree line, 1 June-8 July 1981, 2c319 (Dondale, CNC). Mani- toba: R.M.N. Pk., Clear Lake, pan trap, beaver meadow, 8 June 1979, 1(319 (S.J. Miller, CNC); Lake Audy, Riding Mtn. Natd Park, sifting grass and moss, 28 August 1979, 8(3 17 9 (J. & M. Red- ner, CNC); nr. Wasagaming Riding Mtn. Nat’l P, in boggy area, 23 August 1979, 19 (J. & M. Redner, CNC). New Brunswick: Kouchibouguac N.P., forest edge on beach, 8 June 1977, 1(3 (S.J. Miller, CNC). Newfoundland: Comer Brook Lake, sifted from moss, 13 July 1984, 2(319 (L. Hollett, CNC). Northwest Territories: Mackenzie, Alexandra Falls, Hay River, 60°30'N, 166°17'W, 16 August 1965, 1 9 , (J. & W. Ivie, AMNH); Lac Maunoir, pitfall trap, 19-27 July 1969, 2c349 (G.E. Shewell, CNC). Nova Scotia: Cape Breton Highlands National Park, MacKenzie Mtn., 300 m, ex malaise trough, 28 June-7 July 1983, 2(319 (J.R. Vockeroth, CNC); Cape Breton Highlands National Park, 46°48'N, 60°4LW, 400 m, fen-pans, 8 June 1983, 1(3 (H. Goulet, CNC). Ontario: Bondi Village, Kuskoka District, moss in fir woods, 28 August 1975, lc329 (D. Maddison, CNC); English River (settlement), 49°13'N, 90°58'W, 24 July 1965, 29 (J. & W. Ivie, AMNH); Goward, 47°03'N, 79°55'W, 20 August 1952, 19 (C. Goodnight, AMNH); Nipigon, 48°N, 88°W, 12 August 1948, 1 9 (Gertsch & Kurata, AMNH). Quebec: Lac Cornu, Cte de Terrebonne, 3-4 September 1989, 19 (R. Hutchinson, CNC); Parc des Grandes Jardins, Mont du Lac des Cygnes, 2 June-14 September 1985, 70(345 9 (S. Koponen, UTZM); St. Methode, nr. Lac St. Jean, litter, river bank, 13 July 1982, 1(3 (C. Dondale & J. Redner, CNC); Sherbrooke, sifting litter under trees, 20 September 1972, 2(32 9 (Dondale & Redner, CNC). Saskatchewan: Prince Albert, 24 August 1914, 1(3 (Emerton, MCZ). Yukon: Alaska Hiway, Milepost 700, 60°05'N, 130°25'W, 2 September 1968, lc3 (W. Ivie, AMNH); Dempster Hwy., km 220N, Tam- arack Bog, ex Alnus crispa, Picea mariana litter, 26 June 1987, 19 (V Behan, CNC); Kluane Lake, Kluane Nat’l Park, litter, 6 July 1981, 2(329 (C.D. Dondale, CNC); North Fork Pass, 64°33'N, 138°15'W, sifting litter, 20 June 1981, 5 9 (C.D. Dondale, CNC). UNITED STATES: Alaska: Cha- tanika River Roadside Park, 65°08'N, 147°30'W, 17 August 1968, 1(33 9 (W. Ivie, CNC); Trail to Den- ver Glacier, Skagway, 25 June 1936, 1(3 (Crosby, AMNH); Matanuska, 61°32'N, 149° 1 2' W, Septem- ber 1944, 1 9 (Chamberlin, AMNH), October 1943, 1(3 (J.C. Chamberlin, AMNH); Primrose Camp, 18 mi. N. of Seward, 60°20'N, 149°20'W, 24 August 1968, 15(323 9 (W. Ivie, AMNH). Arizona: Kaibab For., 36°30'N, 112°30'W, 4 September 1931, 4(349 (R.V Chamberlin, AMNH). Maine: Pascataquis County: near Soubunge Mtn., pitfall coll., dense and stripcut spmce-fir forest. May, June, and July, 1977 and 1978, specimens in many vials, (D.T. Jen- nings, M.W. Houseweart, AMNH, CAS, CNC, USNM); Van Buren, 15 July 1914, 1(3 (Emerton, MCZ). Massachusetts: Berkshire County, Mt. Grey lock, 3400 feet, decid. litter, 15 October 1990, 2(319 (R.L. Edwards, USNM). Montana: Gird Creek, Ravalli County, 26 August 1934, 1 9 (W.L. Jellison, AMNH). New Hampshire: Coos County, Mt. Washington toll road, 0.3 mi. below halfway hse., 1100 m, Berl. litter, spmce-fir-birch forest, 15 October 1978, 1(3 (A. Newton, M. Thayer, MCZ); Mt. Washington, 13 June 1877, 3 9 paralectotypes, (J.H. Emerton, MCZ). New York: Catskill Mtn. Pk., North Mtn. Trail, ex litter under balsam fir, 1 Au- gust 1985, 19 (V. Behan, CNC); Mt. MacIntyre, Essex County, 1 July 1923, 19 (AMNH). Utah: Mirror Lake, Uintah Mountains, 40°43'N, 110°53'W, 28 July 1936, 4(349 (W. Ivie, AMNH); west side Utah Lake, heron rookery, 27 May 1934, 49 (Ivie, AMNH). Vermont: Camels Hump, 7 Sep- tember 1908, 2(3 (Emerton, MCZ); Mt. Mansfield, pitfall B fir forest, 26 May- 15 June 1982, 85(3309 (C. Dondale, J. Redner, CNC); Stratton Mtn., 3000 feet, 4 July 1913, 1(3 (Emerton, MCZ). Washing- ton: Okanogan County, Cold Spring Camp, 1850 m, ex rotten log, 48.938°N, 119.789°W, 3 August 1985, 1(3 (R. Crawford, UWBM). Wyoming: Bay Bridge, Yellowstone Lake, 11 August 1940, 8(399 (W. Ivie, AMNH); Grand Canyon, Yellowstone Park, 30 August 1927, 16 (Crosby, AMNH). 582 THE JOURNAL OF ARACHNOLOGY Figures 66-71. — Sisicottus crossoclavis. 66-68, Palpus of holotype from Hayden Lake, Idaho. 66, Ventral view; 67, Fetal view; 68, Palpal tibia, dorsal view. 69, 70, Epigynum of female from Deemer Creek, Washington. 69, Ventral view; 70, Posterior view. 71, Cleared epigynum of female from Rabbit Creek, Washington, dorsal view. Scales: Fig. 71 = 0.05 mm; other figures = 0.1 mm. Sisicottus crossoclavis new species Figs. 49, 50, 55-57, 66-72 Sisicottus sp. #1: Crawford 1988: 15. Types. — Male holotype with female para- Figure 72. — Northwestern United States and southwestern Canada, showing distribution of Sisi- cottus crossoclavis (X), S. cynthiae (•), and S. aenigmaticus (A). type from UNITED STATES: Idaho, Harrison Cr., E side Hayden Lake, 47°N, 116°W, 25 July 1959, EC. Raney, deposited in AMNH. Etymology. — Formed form the Greek word krossos, meaning tasseled, and clavis, a syn- onym for the suprategulum used by FO. Pick- ard-Cambridge (H.D. Cameron pers. comm.). Diagnosis. — The distal suprategular apoph- ysis in males of S. crossoclavis is widened distally with a serrated terminal margin and is unique among Sisicottus (Fig. 67). Males of S. crossoclavis share with S. cynthiae, S. an- tes, and S. nesides a heavily sclerotized distal suprategular apophysis (character 3); all other Sisicottus species have either a moderately sclerotized or membranous distal suprategular apophysis. They share with S. panopeus and S. montanus a distal suprategular apophysis that extends about half way down the ectal MILLER— REVISION OF SISICOTTUS 583 side of the palpal bulb (Fig. 67, character 4); all other Sisicottus species have a distal su- prategular apophysis that extends to near the ventral margin. They are distinguished from all other Sisicottus species except S. panopeus by the lack of an ectal tibial process (Fig. 68, character 17). Females of S. crossoclavis, like those of S. panopeus, S. montanus, and S. cynthiae, have a shallow ventral plate invagination (Figs. 69, character 19); this separates them from fe- males of S. montigenus, S. quoylei, S. orites, S. nesides, and S. aenigmaticus. Sisicottus crossoclavis is distinguished from S. pano- peus, S. montanus, S. montigenus, and S. quoylei by the form of the dorsal fold of the dorsal plate which is sclerotized rather then membranous (character 24). Sisicottus cros- soclavis is distinguished from S. cynthiae by the unusual form of the dorsal plate in S. cyn- thiae. In S. cynthiae, the posterior face of the dorsal plate has a ventral margin that is dorsal to the ventral extent of the spermathecae (Fig. 77); in all other Sisicottus species, including S. crossoclavis, the ventral margin of the pos- terior face of the dorsal plate is at about the level of the ventral extent of the spermathecae (Fig. 70). Description. — Medium-sized (carapace length ^ 0.81-0.97 mm); coloration typical (see description section for Sisicottus). Distal suprategular apophysis of palpus heavily sclerotized, of moderate length, extends about half way down ectal side of palpal bulb; wid- ened distally with serrated terminal margin (Fig. 67). Palpal tibia long with long, gradu- ally curving palpal tibial apophysis; ectal tib- ial process absent (Fig. 75); dense cluster of macrosetae (7-9) on ectal side of palpal tibia. Females with ventral plate invagination shal- low to absent (Figs. 55, 69). Posterior face of dorsal plate triangular with broadly rounded ventral margin (Figs. 56, 70). Dorsal fold of dorsal plate sclerotized (Fig. 71). Lateral mar- gins of copulatory duct capsule in dorsal view sinuous with posterior tips of capsule oriented posteriorly; anterior margin of capsule formed into two convex lateral lobes; fertilization ducts spiral (Fig. 71). See Tables 2-4. Variation. — Of three known male speci- mens, two have a very long palpal tibial apophysis and palpal tibia, but a third speci- men from Rabbit Creek, Washington has a palpal tibial apophysis and palpal tibia of more moderate length (Fig. 57). Despite the difference in size, the tibias are similar in shape. Furthermore, all three specimens share virtually identical palpal bulbs and were found associated with indistinguishable females. I have concluded that both tibial morphotypes belong to a single species. The short tibia in the Rabbit Creek specimen may have been due unfavorable conditions during develop- ment or to some genetic condition. Natural history. — Sisicottus crossoclavis has been collected from moss, rotting logs, and forest litter. Collection labels suggest that this species may have an affinity for relatively wet microhabitats. Distribution. — Alberta, Idaho, and Wash- ington (Fig. 72). Material examined. — CANADA: Alberta: Wa- terton Lakes N.R, moss and litter on damp seepage rock face, 26 June 1980, 1 9 (I.M. Smith, CNC); UNITED STATES: Washington: Ferry County, S Fk. Boulder Cr., 2560 feet, 48.756°N, 118.249°W, ex cottonwood-cedar litter, 15 July 1989, 19 (R. Crawford, UWBM). Ferry County, Rabbit Creek, 3500 feet, 48.539°N, 118.605°W, alder litter; ex moss, logs, and ground in forest, 18 July 1989, 1(359 (R. Crawford, UWBM). Pend Oreille Coun- ty, Deemer Creek, 4600 feet, 48.93 TN, 117.089°W, under rocks and logs; from pitfalls; in soggy moss at stream edge; web in soil depression; rotting log, 11-14 July 1986, 1(369 (R. Crawford, UWBM); Spokane County, Mt. Spokane, 5000 feet, ex moss, herbs, nr. edge seepage area, 28 August 1985, 1 9 (C.C. Lindquist, CNC). Sisicottus cynthiae new species Figs. 72-80, 85, 86 Types. — Male holotype from UNITED STATES: Oregon, Benton County, Mary’s Peak, 44°N, 123°W, 29 September 1960, J.D. Lattin, deposited in AMNH. Etymology. — Named for my friend, Cyn- thia Zujko-Miller, whose support during the course of this project contributed substantially to its completion. Diagnosis.— Males of S. cynthiae share with S. crossoclavis, S. orites, and S. nesides a heavily sclerotized distal suprategular apophysis (character 3); all other Sisicottus species have either a moderately sclerotized or membranous distal suprategular apophysis. They are distinguished from S. crossoclavis and S. orites by their much shorter palpal tib- ial apophysis (Fig. 75) and from S. nesides by 584 THE JOURNAL OF ARACHNOLOGY Figures 73-78. — Sisicottus cynthiae. 73-75, Palpus of holotype from Mary’s Peak, Oregon. 73, Ventral view; 74, Fetal view; 75, Palpal tibia, dorsal view. 76, 77, Epigynum of female from Mary’s Peak, Oregon. 76, Ventral view; 77, Posterior view. 78, Cleared epigynum of female from Grass Mountain, Oregon, dorsal view. Scales: Fig. 78 = 0.05 mm; other figures = 0.1 mm. the terminus of the distal suprategular apoph- ysis which has a broad, rippled ventral margin in S. cynthiae (Fig. 73). The corresponding region in S. nesides terminates in two apical points and is not widened distally (Fig. 102). The form of the ectal tibial process is unique in S. cynthiae being generally larger and more ectally placed than in other Sisicottus species (Fig. 75, character 17). Females of S. cynthiae are distinguished from those of all other Sisicottus species by the form of the posterior face of the dorsal plate. In S. cynthiae, the posterior face of the dorsal plate has a ventral margin that is dorsal to the ventral extent of the spermathecae (Fig. 77); in all other Sisicottus species, the ventral margin of the posterior face of the dorsal plate is at about the level of the ventral extent of the spermathecae (Fig. 70). Sisicottus cyn- thiae, like S. panopeus, S. montanus, and S. crossoclavis, have a shallow ventral plate in- vagination (Figs. 76, character 19); this sep- arates them from females of S. montigenus, S. quoylei, S. orites, S. nesides, and S. aenig- maticus. Sisicottus cynthiae is distinguished from S. panopeus, S. montanus, and S. cros- soclavis by the orientation of the posterior part of the copulatory duct capsule (character 32). In S. cynthiae, the posterior tips of the capsule are oriented mesally (Fig. 78). In S. panopeus, S. montanus, and S. crossoclavis, the posterior tips of the capsule are oriented posteriorly (Fig. 65). Sisicottus cynthiae is distinguished from S. panopeus, S. montanus, S. montigen- us, and S. quoylei by the form of the dorsal fold of the dorsal plate which is sclerotized MILLER— REVISION OF SISICOTTUS 585 Figures 79-84. — Scanning electron micrographs of Sisicottus palpi. 79, 80, S. cynthiae from Mary’s Peak, Oregon. 79, Ectal view; 80, Ventral view. 81, 82, S. orites from Mirror Lake, Utah. 81, Ectal view; 82, Ventral view. 83, 84, S. nesides from Primrose Camp, Alaska. 83, Ectoventral view; 84, Ventral view detailing distal suprategular apophysis. 586 THE JOURNAL OF ARACHNOLOGY Figures 85-90. — Scanning electron micrographs of Sisicottus epigyna. 85, 86, S. cynthiae from Charles- ton, Oregon. 85, Ventral view; 86, Posterior view. 87, 88, S. orites from Smith and Morehouse Canyon, Utah. 87, Ventral view; 88, Posterior view. 89, 90, S. nesides from Change Creek, King County, Wash- ington. 89, Ventral view; 90, Posterior view. MILLER— REVISION OF SISICOTTUS 587 rather then membranous (Fig. 78, character 24). Description. — Large (carapace length = 0.88-1.10 mm); coloration typical (see de- scription section for Sisicottus). Distal supra- tegular apophysis of male palpus heavily sclerotized, long, extends to near ventral mid- line of palpal bulb; slightly widened near ter- minal margin, which has rippled appearance (Figs. 73, 80). Palpal tibia long with medium sized apophysis; ectal tibial process pro- nounced (Fig. 75); moderately dense cluster of macrosetae (6-8) on ectal side of palpal tibia. Female with shallow ventral plate in- vagination (Figs. 76, 85). Posterior face of dorsal plate triangular with broadly rounded ventral margin located dorsal to ventral extent of spermathecae (Fig. 77). Lateral margins of copulatory duct capsule in dorsal view sinu- ous with tips of capsule oriented mesally to- ward each other; anterior margin of capsule formed into two convex lateral lobes; fertil- ization ducts looped (Fig. 78). See Tables 2-4. Natural history. — One collection label states that S. cynthiae has been collected from moss and another indicates that a female spec- imen was found in the stomach of a salaman- der. Sisicottus cynthiae is syntopic with S. Re- sides. Distribution. — Oregon (Fig. 72). Material examined. — UNITED STATES: Oregon: Benton County, Mary’s Peak, 44°N, 123°W, 29 September 1960, 4319 (J.D. Lattin, AMNH); Benton County, Grass Mountain, 44°N, 123°W, 30 October 1960, 4349 (J.D. Lattin, AMNH); Benton County, McGlynn Dr. ravine, moss on road bank, 23 January 1977, 13 (L. Rus- sell, CNC); Charleston, 43°20'N, 124°20'W, 7 Au- gust 1947, 135 9 (I.M. Newell, AMNH), July 1947, 1329 (I.M. Newell, AMNH); Lane County, Klickitat Mtn., N side, 23 January 1977, 29 (L. Russell, CNC); 10 mi. N of Philomath {ex newt), 44°40'N, 123°22'W, about 1950, 19 (R. Freiburg, AMNH). Sisicottus orites (Chamberlin 1919) Figs. 81, 82, 87, 88, 91-101 Grammonota orites Chamberlin 1919: 249 [3,9]. Male holotype from UNITED STATES: Utah, Chalk Creek, Chamberlin, in MCZ, examined. Oedothorax pidacitis Crosby & Bishop 1927: 151 [3]. Male holotype from UNITED STATES: Col- orado, Larimer County, Pingree Park, Stormy Peaks, 10,000 feet, 20 August 1924, Crosby, in AMNH, examined. Synonymy by Chamberlin & Ivie 1933. Oedothorax orites: Chamberlin & Ivie 1933: 22 [9]. Sisicottus montanus, in part: Bishop & Crosby 1938: 57-60, fig. 5 [3]. Sisicottus orites: Chamberlin & Ivie 1939: fig. 38 [3]. Platnick 1993: 351. Diagnosis. — Males of S. orites share with S. crossoclavis, S. cynthiae, and S. Resides a heavily sclerotized distal suprategular apoph- ysis (character 3); all other Sisicottus species have either a moderately sclerotized or mem- branous distal suprategular apophysis. They are distinguished from S. cynthiae by their longer palpal tibial apophysis (Fig. 94). They are distinguished from S. crossoclavis by the presence of an ectal tibial process (Fig. 94, character 17), a long distal suprategular apophysis that extends to near the ventral mid- line of the palpal bulb, and a smooth terminal margin of the distal suprategular apophysis (Fig. 92, character 4). They are distinguished from S. Resides by the terminus of the distal suprategular apophysis which is rounded, of- ten with a shallow central concavity (Fig. 92); in S. Resides, the terminus is bifurcated with the inside lobe coming to a sharp apex on or outside of the median line of the distal supra- tegular apophysis and the outside lobe coming to its apex on the outer margin (Fig. 102). Dimensions of the palpal tibia (Fig. 99) and the number of macrosetae in the ectal tibial cluster (8-13 in S. orites; 7-10 in S. Resides) may also be useful for distinguishing S. orites from S. Resides. Females of S. orites, S. Resides, and S. aenigmaticus differ from those of all other Sisicottus species by the form of the copula- tory duct capsule in dorsal view which has strongly bowed lateral margins (Fig. 98, char- acter 31); in all other Sisicottus species, the lateral margins are sinuous to moderately bowed. Unlike S. panopeus, S. montanus, S. crossoclavis, and S. cynthiae, these species have a deep ventral plate invagination (Fig. 96, character 19). Sisicottus orites is distin- guished from S. aenigmaticus by the width of the ventral plate invagination and by the form of the posterior face of the dorsal plate which is trapezoidal with concave sides in S. aenig- maticus (Fig. 107) and triangular with nearly straight sides in S. orites (Fig. 97, characters 21, 22). Females of S. orites are difficult to 588 THE JOURNAL OF ARACHNOLOGY Figures 91-98. — Sisicottus orites. 91-94, Palpus of male from Mirror Lake, Utah. 91, Mesal view; 92, Ventral view; 93, Fetal view; 94, Palpal tibia, dorsal view. 95, Suprategulum separated from palpus of male from Smith and Morehouse Canyon, Utah, ectal view. 96, 97, Epigynum of female from Mirror Lake, Utah. 96, Ventral view; 97, Posterior view. 98, Cleared epigynum of female from Smith and More- house Canyon, Utah, dorsal view. Scales: Figs. 95, 98 = 0.05 mm; other figures = 0.1 mm. distinguish from those of S. nesides. Sisicottus nesides, like S. aenigmaticus, have a dorsal plate with concave sides on its posterior face (character 22). Also, epigynum length and copulatory duct capsule width are both usu- ally larger in S. orites than in either S. nesides or S. aenigmaticus (Fig. 100). Description. — Large (carapace length = 0.88-0.20 mm); coloration typical (see de- scription section for Sisicottus). Distal supra- tegular apophysis of male palpus heavily sclerotized, long, extends to near ventral mid- line of palpal bulb; terminal margin rounded or with shallow concave invagination (Figs. 82, 92). Palpal tibia long with long apophysis; ectal tibial process present (Fig. 94); very MILLER— REVISION OF SISICOTTUS 589 Figures 99-100. — Scattergrams of morphometric characters for males and females of Sisicottus orites (X) and S. nesides (□) and female of S. aenigma- ticus (A). 99, Palpal tibial apophysis length plotted against palpal tibia length in males; 100, Epigynum length plotted against copulatory duct capsule width in females. dense cluster of macrosetae (8-13) on ectal side of palpal tibia. Females with deep ventral plate invagination (Figs. 87, 96). Posterior face of dorsal plate triangular with nearly straight sides (Figs. 88, 97). Dorsal fold of dorsal plate sclerotized (Fig. 98). Lateral mar- gins of copulatory duct capsule in dorsal view strongly bowed with tips of capsule oriented mesally toward each other; anterior margin of capsule formed into two convex lateral lobes; fertilization ducts looped (Fig. 98). See Table 2-4. Natural history. — Collection labels indi- cate that this species is associated with wet moss and similar microhabitats. It has been collected syntopically with S. montanus, S. panopeus, and S. nesides. Distribution. — From California, Utah, and New Mexico north to Washington and Alberta (Fig. 101). Material examined. — CANADA; Alberta: Wa- terton Lakes Nat’l Park, Cameron Lake, 5300-5500 feet, 9-19 June 1980, 8 c? 2 9 (J.M. Campbell, CNC), Lower Bertha Falls, sifted moss, 10 June 1980, 1 9 (J.M. Campbell, CNC); Lake Louise, 4 August 1927, 1(349 (Crosby, AMNH); Mt. Edith, Cavill Lodge, 52°41'N, 118°09'W, 24 August 1965, 109 (J. & W. Ivie, AMNH); Whitemud Creek, Ed- monton, soil sample, 8 May 1959, 3 9 (L.K. Smith, CNC). UNITED STATES; California: Laguna Lake, Laguna Canyon, 33°36'N, 117°45'W, 6 July 1934, 19 (Ivie & Rasmussen, AMNH). Colorado: Berthoud Pass, 39°58'N, 105°48'W, 24 August 1935, 1349 (Ivie, AMNH); Cameron Pass, 11,000 feet, 40°3LN, 105°52'W, 3 August 1946, 19 (C.C. Hoff, AMNH); Miguel County, Trout Lk., NE of Lizard Head Pass, San Juan Mtns., 3100 m, mud flats, moist sedges, 20 July 1959, 1 9 (H.W. Levi, MCZ); Pikes Peak, 11,600 feet, 38°52'N, 105°5W, 22 July 1940, 3 9 (Ivie, AMNH). Idaho: 2 miles south of Tamarack, 44°56'N, 116°23'W, 17 October 1944, 2329 (Ivie, AMNH); Targhee Pass, 44°38'N, 111°18'W, 30 June 1962, 29 (Ivie, AMNH); Wil- low Flat camp, Franklin County, 42°N, 111°W, 5 July 1952, 3 9 (B. Malkin, AMNH). Nevada: Ruby Valley, 40°15'N, 115°25'W, September 1937, 235 9 (Chamberlin, AMNH); White Pine County, Schell Mtns., Timber Cr. Cpgd. area, 30 mi. NE Ely, 8700 feet, from wet moss and substrate, 27 June 1989, 1 9 (E.E. Lindquist, CNC); White Pine County, Schell Mtns., 15 mi SE Ely, 7500 feet, wet moss- stream edge, 26 June 1989, 1 9 (E.E. Lindquist, CNC). New Mexico: Panchuela Cpgd, 18 mi. N Pe- cos, 8400 feet, ex liverwort carpet on moist rock wall, 28 August 1973, 13 (E.E. Lindquist, CNC). Oregon: Douglas County, Diamond Lake, 43°10'N, 122°08'W, 7 September 1949, 19 (V. Roth, AMNH); Grant County, Strawberry Creek Falls, 6800 feet, ex moss on rocks in falls spray zone, 23 July 1985, 1 9 (E.E. Lindquist, CNC); Grant Coun- ty, Strawberry Lake area, 6400 feet, ex moss, herbs, rotting wood in seepage, 23 August 1985, 13 (E.E. Lindquist, CNC); Langdon Lake, Blue Mts, 13 Sep- tember 1949, 49 (V. Roth, AMNH). Utah: nr Alta, Little Cottonwood Canyon, under stones in creek, 25 June 1985, 29 (C. Dondale & J. Redner, CNC); City Creek Can., Rotary Park, 46°48'N, 111°46'W, 11 September 1942, 8389 (Ivie, AMNH), 16 Sep- tember 1942, 4349 (Ivie, AMNH); Logan Canyon, 49 (Chamberlin, MCZ); Mirror Lake, Uintah Mountains, 40°43'N, 110°53'W, 22 September 1932, 4349 (Ivie, AMNH), 28 July 1936, 4349 (Ivie, AMNH), 18 August 1942, 4349 (Ivie, AMNH); Smith and Morehouse Canyon, 40°47'N, 110°6'W, 7 October 1932, 4349 (W. Ivie, USNM); So. Fork Raft Riv., 8 mi. So. Lynn, 41°53'N, 590 THE JOURNAL OF ARACHNOLOGY 113°45'W, 6 September 1932, 5c? 109 (Chamberlin & Ivie, AMNH); Vicinity of Salt Lake City, quad 40°N, lirw, misc. 1928-1936, 4(312 9 (AMNH). Washington: Seattle, 47°35'N, 122°20'W, May 1952, 1(319 (Borys Malkin, AMNH). Wyoming: Canyon east of Bedford, 42°50'N, 110°50'W, 27 June 1962, 29 (Ivie, AMNH); Centennial, Wyo- ming University, 9500 feet, under log, 17 August 1936, 1(3 (AMNH). Sisicottus nesides (Chamberlin 1921) Figs. 83, 84, 89, 90, 99-105 Oedothorax nesides Chamberlin 1921; 36, plate III, figs. 1-2 [(3]. Male holotype from UNITED STATES: Alaska, St. Paul Island, 1910, H. Heath, in MCZ, examined. Sisicottus montanus: Bishop & Crosby 1938: 57- 60, fig. 4 [(3] (in part). Bragg & Leech 1972: 69 (misidentification). Sisicottus nesides: Chamberlin & Ivie 1939: fig. 39 [c3]. Crawford & Edwards 1988: 437; figs. 23- 24 [9]. Platnick 1993: 351. Dondale et al. 1997: 89. Sisicottus montanus nesides: Holm 1960: 124. El- evated by Crawford & Edwards 1988. Sisicottus orites: West et al. 1984: 87 (misidentifi- cation). Diagnosis. — Males of S. nesides share with S. crossoclavis, S. cynthiae, and S. orites a heavily sclerotized distal suprategular apoph- ysis (character 3); all other Sisicottus species have either a moderately sclerotized or mem- branous distal suprategular apophysis. They are distinguished from S. crossoclavis by the presence of an ectal tibial process (Fig. 103, character 17) and a long distal suprategular apophysis that extends to near the ventral mid- line of the palpal bulb and lacks a serrated terminal margin (Fig. 102, character 4); the distal suprategular apophysis in S. crossoclav- is extends only about half way down the ectal face of the palpal bulb and has a serrated ter- minal margin (Fig. 67). Sisicottus nesides can be distinguished from other Sisicottus species with a long, heavily sclerotized distal supra- tegular apophysis by the terminus of the distal suprategular apophysis which is bifurcated with the inside apex coming to a sharp point on or outside of the median line and the out- side apex coming to a point on the outer mar- gin; this condition is unique among Sisicottus (Figs. 84, 102). The distal suprategular apoph- ysis in S. cynthiae has a rippled terminal mar- gin (Fig. 73); the distal suprategular apophysis in S. orites has a rounded terminal margin, often with a shallow central concavity (Fig. 92). Dimensions of the palpal tibia (Fig. 99) and the number of macrosetae in the ectal tib- MILLER— REVISION OF SISICOTTUS 591 ial cluster (7- 10 in S. nesides; 8-13 in S. or- ites) may also be useful for distinguishing S. nesides from S, o rites. Females of S. nesides, S. orites, and S. aenigmaticus differ from those of all other Sisicottus species by the form of the copula- tory duct capsule in dorsal view which has strongly bowed lateral margins {cf. Fig. 98, character 31); in all other species, the lateral margins are sinuous to moderately bowed. Unlike S. panopeus, S. montanus, S. crosso- clavis, and S. cynthiae, these species have a deep ventral plate invagination (Fig. 104, character 19). Sisicottus nesides is distin- guished from S. aenigmaticus by the width of the ventral plate invagination and by the form of the posterior face of the dorsal plate which is trapezoidal in S. aenigmaticus (Fig. 107, character 21) and triangular in S. nesides (Fig. 105). Females of S. nesides are difficult to dis- tinguish from those of S. orites. Sisicottus ne- sides have a dorsal plate with concave sides on the posterior face (Fig. 105, character 22) while the dorsal plate of S. orites has nearly straight sides on the posterior face (Fig. 97). Also, epigynum length and copulatory duct capsule width are both usually greater in S. orites than in either S. nesides or S. aenig- maticus (Fig. 100). Description. — ^Large (carapace length = 0.88-1.20 mm); coloration typical (see de- scription section for Sisicottus). Distal supra- tegular apophysis heavily sclerotized, long, extends to near ventral midline; terminus bi- furcated with longer inside apex confing to sharp point on or outside median line and the shorter outside apex coming to point on the outside margin (Figs. 84, 102). Palpal tibia moderately long with moderately long palpal tibial apophysis; ectal tibial process present (Fig. 103, character 17); dense cluster of ma- crosetae (7-9) on ectal side of palpal tibia. Females with deep ventral plate invagination (Figs. 89, 104). Posterior face of dorsal plate triangular with concave sides (Figs. 90, 105). Dorsal fold of dorsal plate sclerotized (cf. Fig. 98). Lateral margins of copulatory duct cap- sule in dorsal view strongly bowed with tips of capsule oriented mesally toward each other; fertilization ducts looped (cf Fig. 98). Aside from some quantitative differences (Fig. 100), internal structure of epigynum virtually iden- tical to that of S. orites (Fig. 98). See Tables 2-4. Natural history. — This species has often been collected in moss and litter microhabitats in forests. In Washington, this species is par- tially separated from S. panopeus by ecolog- ical/elevational constraints (Crawford & Ed- wards 1988; also see the Natural History section in S. panopeus). Sisicottus nesides oc- curs syntopically with S. montanus, S. cyn- thiae, S. aenigmaticus, S. crossoclavis, and S. orites. Distribution.- — From Alaska, Alberta, Brit- ish Columbia, Oregon, Washington, and the Yukon (Fig. 101). A published record from Nebraska is almost certainly erroneous (Rapp 1980 pers. comm.). A male and a female spec- imen from Lincoln in the UNSRC identified by M.H. Muma as S. nesides are in fact a species of Walckenaeria (possibly W. maesta Millidge 1983) and an unidentified hetero- specific female. Material examined. — CANADA: Alberta: Cameron Lake, Waterton Lakes Nat’l Park, 5300 feet, interception trap, 9-28 June 1980, 13 (J.M. Campbell, CNC). British Columbia: Albert Bay, 50°34'N, 126°58'W, 21 June 1936, 19 (Crosby & Bishop, AMNH); Lake Cowichan, moss on log, 8 July 1976, 1329 (I.M. Smith, CNC); Grahm Is- land, Masset, 1944, 13 (M.C. Clark, MCZ); John- son Bay, Babine Lake, leaf litter, 4 July 1987, 23109 (R. West, CNC); Malahat, Coldstream Prov. Park, V.I., moss on rock at spring run, 1 1 July 1979, 3313 9 (1. Smith, CNC); Manning Prov. Park, Pitfall in rhododendron flat, 20 June-3 July 1979, 19 (Dondale, CNC); Metlakatla, 1329 (Emerton, AMNH); Prince Rupert, 54°09'N, 130°20'W, 22 June 1936, 19 (C.R. Crosby, AMNH); Queen Charlotte Is., Louise Is., Skedans, wet moss at seepage spots on old rd, 8 August 1983, 1 3 (J.M. Campbell, CNC); 7.0 km NW Q.C. City, 4-15 August 1983, flight intercept trap, 23 (J.M. Campbell, CNC); Wap Lake, Revelstoke, pit- fall -rocky bank, July 1985, 1 9 (M.E. Martin, CNC); Wellington, V.I., 5 October 1949, 43109 (R. Guppy, AMNH). Yukon: Kathleen Lake, Kulane Nat’l Park, litter and stones, 12-15 June 1981, 29 (C.D. Dondale, CNC). UNITED STATES: Alaska: Aleutian IsL, Adak IsL, Andreanof Isl., 26-29 July 1958, 1 9 (C.H. Lindroth, MCZ); Umnak, Fox Is- lands, July 1958, 19 (C. Lindroth, MCZ); Unalas- ka. Fox Islands, Mt. Makushin, 11-14 July 1958, 1349 (C. Lindroth, MCZ); Trail to Denver Glacier, Skagway, 25 June 1936, 138 9 (Crosby, AMNH); Haines, Quad. 59°N, 135°W, 20-25 August 1945, 1 9 (Chamberlin, AMNH); Juneau, litter and stones, 8-10 June 1981, 29 (C.D. Dondale, CNC), 58°N, 134°W, 28-29 April 1945, 19 (Chamberlin, 592 THE JOURNAL OF ARACHNOLOGY Table 2. — Quantitative character values for adult males of Sisicottus species. Range, mean, standard deviation, and sample size are given for all measurements (in mm). For number of macrosetae in tibial cluster, mode is given in parenthesis. S. montigenus S. quoylei S. panopeus Carapace (length) 0.70-0.80 0.75-0.82 0.67-0.96 0.74 ± 0.03 0.78 ± 0.03 0.87 ± 0.06 n — 20 n = 6 n = 20 Metatarsus I (length) 0.40-0.49 0.39-0.47 0.50-0.58 0.44 ± 0.03 0.43 ± 0.03 0.54 ± 0.02 n — \9 n = 6 n = \9 Tml 0.47-0.62 0.41-0.60 0.45-0.54 0.56 ± 0.04 0.54 ± 0.07 0.50-0.02 n — \9 n = 6 n = \9 Palpal tibial apophysis (length) 0.014-0.034 0.019-0.038 0.081-0.124 0.023 ± 0.007 0.031 ± 0.008 0.102 ± 0.013 n = 22 n = 6 n = 20 Palpal tibia (length) 0.138-0.176 0.147-0.171 0.238-0.295 0.157 ± 0.011 0.163 ± 0.008 0.265 ± 0.016 n ^ 22 n = 6 n = 20 Palpal tibia (width) 0.128-0.166 0.138-0.162 0.181-0.223 0.150 ± 0.009 0.151 ± 0.008 0.202 ± 0.014 n = 22 n ~ 6 n = 20 Paracymbium (length) 0.095-0.162 0.109-0.119 0.133-0.170 0.116 ± 0.016 0.115 ± 0.004 0.150 ± 0.008 n = 20 n = 6 n = 20 Paracymbium (width) 0.109-0.133 0.119-0.133 0.138-0.162 0.120 ± 0.006 0.124 ± 0.006 0.152 ± 0.007 n = 20 n = 6 n = 20 Lamella characteristica (length) 0.090-0.124 0.109-0.119 0.124-0.156 0.107 ± 0.010 0.112 ± 0.004 0.138 ± 0.008 n = 20 n = 6 n — 20 Number of macrosetae in tibial cluster 3-7 (6) 3-6 (5) 7-11 (8) 5.35 ± 1.14 4.67 ± 1.03 8.55 ± 0.94 n = 20 n = 6 n = 20 AMNH); Primrose Camp, 18 mi. N. of Seward, 60°20'N, 149°20'W, 24 August 1968, \36229 (W. Ivie, AMNH); 5 mi. S rapids on Richardson Hiway, 26 June 1945, \S (J.C. Chamberlin, AMNH); Skagway, 59°28'N, 135°15'W, 24 June 1939, 2 9 (Crosby, AMNH). Oregon: Benton County, Mary’s Peak, sod in alder clearing, 28 November 1976, 1 9 (L. Russell, CNC); Latourell Falls, 45°N, 122°W, 4 August 1929, 1(349 (Chamberlin, AMNH); Proxy Falls, Hwy 242, 28 mi. SW Sisters, 3000 feet, ex wet moss and debris, depression by waterfall, 19 August 1985, 1 9 (E.E. Lindquist, CNC); Tillamook County, moss near stream, 19 December 1976, 29 (L. Russell, CNC). Washington: Chelan County, Nason Creek, 47.783°N, 120.874°W, 2280 feet, un- der rocks by stream, 6 June 1992, Id (Rick Sugg, UWBM); Clallam County, Pillar Pt (W side), 48.220°N, 124.125-130°W, 0-100 feet, moss on forest floor, 22 May 1987, 2S (R. Crawford, UWBM); Emmons Trail, Tainier Park, 46°54'N, 12L39'W, 6 July 1938, 29 (Ivie, AMNH); Friday Harbor, 1924, 1 9 (AMNH); Thurston County, The Evergreen State College, wetland south of Ever- green Parkway, 47°4'10"N, 122°57'39"W, elev. 50 m, ex moss on downed Salix tree, 16 July 1993, 1 9 (J. Miller, author’s personal collection). Woodland SE of Comer of Overhulse Place and Driftwood Road, 47.026°N, 122.962°W, elev. 50 m, 26 October 1992, 19 (J. Miller, author’s personal collection); Jefferson County, Olympic National Park, Hoh Riv., 510 feet, 47.846°N, 123.960°W, moss in Acer macrophyllum, 29 January 1983, 1 9 (J. Lon- gino, UWBM); Lewis County, Lewis & Clark S.P., 46.519°N, 122.815°W, 380 feet, leaf litter, 29 Oc- tober 1988, Id 19 (R. Crawford, UWBM); King County, E. of Change Creek, 1250 feet, 47°44'N, 121°66'W, wet moss and litter in and near seepage, 7 July 1996, 3d49 (J. & C. Zujko-Miller, J. & H. Miller, author’s personal collection); King County, W of Change Cr., 1280 feet, 47.439°N, 12L633°W, 9 April 1989, 2d69 (R. Crawford, UWBM); King County, Happy Valley bog, 47.640°N, 122.017°W, MILLER— REVISION OF SISICOTTUS 593 Table 2. — Extended. S. montanus S. crossoclavis S. cynthiae S. orites S. nesides 0.71-0.93 0.83-0.97 1.00-1.10 0.93-1.20 0.99-1.24 0.86 ± 0.04 0.88 ± 0.08 1.04 ± 0.04 1.07 ± 0.07 1.10 ± 0.08 n = 56 n = 3 n = 9 n = 31 n = 22 0.45-0.58 0.51-0.66 0.67-0.74 0.66-0.81 0.71-0.83 0.52 ± 0.03 0.58 ± 0.07 0.71 ± 0.02 0.73 ± 0.05 0.77 ± 0.03 n = 53 n — 3 n = 9 n = 31 n = \9 0.45-0.65 0.49-0.56 0.49-0.58 0.42-0.56 0.47-0.60 0.58 ± 0.04 0.52 ± 0.04 0.54 ± 0.03 0.50 ± 0.03 0.54 ± 0.04 n = 53 n = 3 n = 9 n = 31 n = \9 0.043-0.076 0.124-0.171 0.067-0.090 0.105-0.162 0.081-0.114 0.058 ± 0.007 0.155 ± 0.027 0.079 ± 0.009 0.126 ± 0.014 0.096 ± 0.011 n = 51 n = 3 n = 9 n = 39 n = 21 0.162-0.228 0.261-0.333 0.257-0.318 0.266-0.337 0.242-0.323 0.193 ± 0.012 0.306 ± 0.039 0.295 ± 0.023 0.306 ± 0.020 0.289 ± 0.021 n = 51 n = 3 n = 9 n ^ 31 n = 22 0.152-0.185 0.209-0.228 0.176-0.219 0.185-0.261 0.162-0.233 0.172 ± 0.009 0.217 ± 0.010 0.200 ± 0.015 0.229 ± 0.019 0.120 ± 0.016 « = 57 n = 3 n = 9 n = 31 n = 22 0.105-0.152 0.143-0.166 0.124-0.171 0.152-0.195 0.157-0.195 0.130 ± 0.009 0.157 ± 0.013 0.155 ± 0.015 0.175 ± 0.011 0.174 ± 0.010 n = 51 n = 3 n ^ 9 n ^ 31 n = 22 0.119-0.152 0.143-0.147 0.152-0.176 0.133-0.190 0.143-0.181 0.136 ± 0.006 0.146 ± 0.003 0.165 ± 0.009 0.178 ± 0.011 0.166 ± 0.008 n ^ 51 n = 3 n = 9 n = 31 n = 22 0.100-0.133 0.128-0.162 0.133-0.143 0.143-0.176 0.124-0.157 0.121 ± 0.007 0.143 ± 0.017 0.138 ± 0.004 0.159 ± 0.009 0.145 ± 0.009 n = 51 n — 3 n = 9 n = 31 n = 22 2-6 (4) 7-9 (9) 6-8 (7) 8-13 (9) 7-10 (7) 4.16 ± 0.80 8.33 ± 1.16 7.22 ± 0.67 9.38 ± 1.11 7.68 ± 0.72 n ^ 51 II n ^ 9 n ^ 31 n = 22 125 feet, sifted in deciduous litter, 11 October 1980, 16 (R. Crawford, UWBM); Snohomish County, Lake Twenty two Research Natural Area, trail head, old growth forest, sifted from moss on log, 25 June 1993, 4$ (J. Miller, author’s personal collection); Mt. Rainier Nat. Park, Nisqually J^ver, 3900 feet, 8 August 1973, 1(3119 (A. Smetana, CNC); Mt. Rainier Nat. Park, North Puyallup River, 3700 feet, 10 August 1973, 7 9 (A. Smetana, CNC); Paradise, Rainier Nat’l Park, 46°48'N, 12r44'W, 12 Septem- ber 1965, 3(3109 (J. & W. Me, AMNH); Pend Oreille County, Deemer Creek, 48.93 1°N 117.089°W, 4600 feet, sifted from soggy moss at stream edge, 13 June 1986, 1(3 (R. Crawford, UWBM); 4 mi. N Silver Fir Cmpg., Mt Baker, 4000 feet, 16 August 1975, 19 (J.M. Campbell, CNC); Skagit County, E of Swede Cr., 48.562-65°N, 122.216°W, 350 feet, ex mixed leaf litter, 19 March 1988, 19 (R. Crawford, UWBM); 10 miles north of Vancouver, 45°45’N, 122°38'W, 10 September 1935, 1 9 (Chamberlin & Ivie, AMNH); Whatcom County, Blue Lake Trail, 48.652°N, 121.786°W, 5000 feet, ex rotten log, 13 September 1986, 19 (R. Crawford, UWBM); Yakima County, Bear Creek Mtn Trail, 46.552°N, 121.315°W, 6160 feet, 6 -ex wet moss by stream, 9 -ex rotten log, 4 Sep- tember 1986, 1(319 (R. Crawford, UWBM). Sisicottus aenigmaticus new species Figs. 72, 100, 106-108 Sisicottus orites: Crawford 1988: 15 (misidentifi- cation). Crawford & Edwards 1988: 437; figs. 25-26 [9] (misidentification). Types.— Female holotype from UNITED STATES: Washington, King County, W. of Change Cr. 1280 feet, 47.439°N, 121.663°W, ex moss in & nr spring, 22 August 1981, R.E. Nelson, deposited in UWBM. Etymology. — The specific name is a Latin adjective meaning enigmatic. 594 THE JOURNAL OF ARACHNOLOGY Table 3. — Quantitative character values for adult females of Sisicottus species. Sisicottus aenigmaticus is the female holotype alone. For all other species, range, mean, standard deviation, and sample size are given (in mm). Ventral plate invagination width is poorly defined in S. panopeus, S. montanus, and S. crossoclavis and was not recorded for these species. S. montigenus S. quoylei S. panopeus S. montanus Carapace (length) 0.67-0.80 0.74 ± 0.03 0.71-0.80 0.76 ± 0.02 Metatarsus I (length) Tml Epigynum (length) Copulatory duct capsule (width) Dorsal plate posterior face (width) Dorsal plate posterior face (height) Ventral plate invagination (depth) Ventral plate invagination (width) n = 33 n = 12 0.31-0.46 0.35-0.42 0.42 ± 0.03 0.39 ± 0.02 n — 33 n = 10 0.48-0.77 0.52-0.63 0.60 ± 0.05 0.58 ± 0.04 n = 33 n ^ 10 0.124-0.176 0.109-0.128 0.149 ± 0.012 0.130 ± 0.010 n = 33 n = 12 0.124-0.176 0.114-0.152 0.149 ± 0.012 0.130 ± 0.010 n = 33 n = 12 0.081-0.119 0.076-0.105 0.099 ± 0.010 0.089 ± 0.008 n ^ 33 n — 12 0.105-0.138 0.081-0.124 0.118 ± 0.009 0.112 ± 0.011 n ^ 33 n ^ 12 0.090-0.119 0.071-0.081 0.095 ± 0.008 0.075 ± 0.004 n = 29 n = 12 0.062-0.109 0.014-0.043 0.080 ± 0.012 0.032 ± 0.009 n = 33 n = 12 0.82-0.95 0.72-0.88 0.88 ± 0.04 0.81 0.04 n = 21 n = 57 0.45-0.56 0.39-0.52 0.50 ± 0.03 0.45 0.03 n = 21 n 57 0.45-0.73 0.50-0.75 0.52 ± 0.06 0.59 -+• 0.04 n = 21 n 57 0.114-0.162 0.095-0.166 0.139 ± 0.012 0.148 -f- 0.012 n = 22 n = 59 0.105-0.162 0.109-0.176 0.133 ± 0.014 0.138 4- 0.015 n = 22 n = 59 0.076-0.100 0.067-0.114 0.085 ± 0.008 0.087 ± 0.011 n = 22 n = 59 0.081-0.109 0.071-0.114 0.096 ± 0.009 0.088 4- 0.008 n - 22 n 59 0-0.024 0.005-0.033 0.010 ± 0.005 0.019 4- 0.006 n = 22 n = 59 Diagnosis. — Males are unknown, but are probably similar to S. orites and S. nesides. Females of S. aenigmaticus, S. orites, and S. nesides differ from those of all other Sisicottus species by the form of the copulatory duct capsule in dorsal view which has strongly bowed lateral margins (Fig. 108, character 31); in all other species, the lateral margins are sinuous to moderately bowed. They are distinguished from all other Sisicottus species except S. montigenus by their very wide ven- tral plate invagination (Fig. 106). The dorsal plate with a trapezoidal posterior face is unique in Sisicottus (Fig. 107, character 21). They also differ from all other Sisicottus spe- cies in having smaller, more widely spaced spermathecae with very narrow margins and copulatory ducts with a relatively narrow proximal part (Fig. 108). Description. — Large (carapace length == 0.91 mm); single known specimen lighter than nor- mal (see remarks below). Ventral plate invagi- nation deep and wide (Fig. 106). Posterior face of dorsal plate trapezoidal, flat ventrally, widest dorsally with concave sides (Fig. 107). Dorsal fold of dorsal plate sclerotized (Fig. 108). Lat- eral margins of copulatory duct capsule in dor- sal view strongly bowed with the tips of the capsule oriented mesally toward each other; spermathecae relatively small with very narrow margins; copulatory ducts relatively narrow and sinuous; anterior margin of capsule formed into two convex lateral lobes; fertilization ducts looped (Fig. 108). See Table 3. Remarks. — This species is known from a single female specimen. It is lightly sclero- tized in a way that is characteristic of speci- mens that have only recently molted to the adult instar. This specimen may be a mutant individual of S. nesides, which has also been collected at the Change Creek site, but I think it more likely that it is a member of a distinct MILLER— REVISION OF SISICOTTUS 595 Table 3. — Extended. S. crossoclavis S. cynthiae S. orites S. nesides S. aenigmaticus 0.81-0.90 0.88-1.06 0.88-1.13 0.86-1.12 0.91 0.85 ± 0.04 0.98 ± 0.05 LOO ± 0.06 0.97 ± 0.07 « = 13 « = 15 n = 12 n = 45 0.43-0.53 0.60-0.74 0.55-0.78 0.58-0.71 0.63 0.5 ± 0.03 0-.67 ± 0.04 0.65 ± 0.05 0.65 ± 0.04 « = 13 n = U n = 1\ n = 39 0.44-0.54 0.45-0.60 0.34-0.62 0.48-0.63 0.55 0.50 ± 0.03 0.56 ± 0.05 0.51 ± 0.04 0.54 ± 0.03 « = 12 n — 14 n = 11 n = 39 0.138-0.166 0.128-0.162 0.152-0.209 0.133-0.181 0.14 0.156 ± 0.009 0.144 ± 0.010 0.179 ± 0.013 0.150 ± 0.010 « = 13 n = 15 n = 13 « - 50 0.105-0.147 0.128-0.162 0.143-0.238 0.100-0.176 0.12 0.130 ± 0.013 0.144 ± 0.010 0.187 ± 0.023 0.145 ± 0.016 « = 13 n - 15 n = 13 n = 50 0.105-0.143 0.109-0.152 0.124-0.204 0.143-0.228 0.21 0.126 ± 0.014 0.135 ± 0.012 0.165 ± 0.018 0.181 ± 0.019 n = n n = \5 n = 13 n = 50 0.100-0.124 0.086-0.133 0.109-0.181 0.105-0.171 0.152 0.112 ± 0.007 0.108 ± 0.014 0.152 ± 0.016 0.147 ± 0.014 w = 13 n = \5 n = 13 n = 50 0-0.019 0.014-0.038 0.052-0.090 0.043-0.086 0.062 0.011 ± 0.005 0.023 ± 0.007 0.074 ± 0.009 0.058 ± 0.009 n = U n = \5 n = 60 n = 49 0.014-0.043 0.005-0.043 0.005-0.043 0.067 0.023 ± 0.007 0.022 ± 0.009 0.020 ± 0.008 n= 15 n = 13 n = 49 species. Since syntopy is common for Sisicot- tus species, I do not regard the fact that S. ne sides have been collected at the type local- ity of S. aenigmaticus as evidence that the lat- er is an aberrant form of the former. Of course, the collection of more specimens will be need- ed in order to test this hypothesis. The Change Creek collection site has yielded a number of other rare and unique spiders including un- described species of the linyphiid genera Hal- Table 4. — Quantitative character values for male holotype specimens of Sisicottus species. Data for the female holotype of S. aenigmaticus are given in Table 3. All measurements in mm. S. mont- igenus 5. quoylei S. pan- opeus S. mon~ tanus S. cross- S. oclavis cynthiae S. orites 5. nesides Carapace length 0.75 0.75 0.88 0.89 0.97 1.00 0.97 LOO Metatarsus I length 0.39 0.52 0.48 0.66 0.68 0.68 Tml 0.41 0.45 0.64 0.56 0.54 0.54 Palpal tibial apophysis length 0.02 0.04 0.09 0.06 0.17 0.08 0.12 0.09 Palpal tibia length 0.16 0.15 0.23 0.17 0.32 0.3 0.29 0.28 Palpal tibia width 0.15 0.15 0.2 0.19 0.23 0.2 0.19 0.16 Paracymbium length 0.11 0.11 0.16 0.13 0.17 0.17 0.18 0.17 Paracymbium width 0.12 0.12 0.15 0.13 0.15 0.16 0.15 0.14 Lamella length 0.14 0.11 0.14 0.12 0.16 0.14 0.14 0.13 Macrosetae in tibial cluster 3 6 7 4 7 7 10 7 596 THE JOURNAL OF ARACHNOLOGY Figures 102-108. — Sisicottus nesides and S. aenigmaticus. 102-105, S. nesides from Primrose Camp, Alaska. 102, Palpus, ventral view; 103, Palpal tibia, dorsal view; 104, Epigynum, ventral view; 105, Epigynum, posterior view. 106-108, Epigynum of S. aenigmaticus holotype from Change Creek, Wash- ington. 106, Ventral view; 107, Posterior view; 108, Cleared, dorsal view. Scales: Fig. 108 = 0.05 mm; other figures = 0.1 mm. orates Hull 1911 and Eulaira Chamberlin & I vie 1933, and something that may belong to the dictynid genus Saltonia Chamberlin & Ivie 1942 (R. Crawford pers. comm.). Natural history. — The single specimen was collected from moss in and near a spring. Distribution. — Known only from the type locality (Fig. 72). ACKNOWLEDGMENTS The following persons and institutions kindly loaned specimens for this study: R. Bradley, Ohio State University Museum of Biodiversity (OSU); J.A. Coddington, Nation- al Museum of Natural History, Smithsonian Institution (USNM); F.A. Coyle, personal col- lection (FAC) and Western Carolina Univer- sity Great Smoky Mountains National Park Collection (GSMNP); R.L. Crawford, Thomas Burke Memorial Washington State Museum (UWBM); C.D. Dondale & J.H. Redner, Ca- nadian National Collection (CNC); K. Eskov, Paleontological Institute, Russian Academy of Sciences (RAS); C.E. Griswold, California Academy of Sciences (CAS); S. Koponen, Zoological Museum, University of Turku (UTZM); H.W Levi, Museum of Comparative Zoology, Harvard University (MCZ); N.I. Platnick, American Museum of Natural His- tory (AMNH); B.C. Ratcliffe, University of Nebraska Systematics Research Collection (UNSRC); N. Scharff, Zoological Museum, University of Copenhagen (ZMUC). Y. Ma- rusik and S. Koponen facilitated the loan of specimens from Russia. Asian specimens bor- rowed from UWBM and RAS were supplied by the International Kuril Islands Project, sup- ported in part by the International Programs Division and the Biological Sciences Direc- torate (Biotic Surveys and Inventories Pro- gram) of the U.S. National Science Founda- MILLER— REVISION OF SISICOTTUS 597 Figures 109-114= — Scanning electron micrographs of Sisicottus and Typhochrestus. 109, Tarsal claw of female Sisicottus panopeus from Mt. Rainier, Washington. 110-112, Typhochrestus uintanus from Mirror Lake, Utah. 110, Epigynum, ventral view; 111, Palpus, apical view; 112, Palpus, ectal view. 113, 114, Typhochrestus digitatus from Whiteford Burrows, England. 113, Palpus, ventral view; 114, Palpus, ectal view. 598 THE JOURNAL OF ARACHNOLOGY Table 5. — Data matrix, characters and states. Number of steps (St), consistency index (Cl), retention index (RI), and rescaled consistency index (RC) are from the preferred most parsimonious tree. = unknown; “ — ” = not applicable. Characters and states Is^ landi- ana prin- ceps Diplo- cent- ria biden- tata pho- chres- tus digip atus Ty~ pho- chres- tus uin- tanus Erb gone psych- ro~ phila Tme- ticus tolli Wab cken- aeria di- recta Gona- tium ru- bens Male palpus 1. embolus: short; long 0 0 1 1 0 0 1 1 2. terminal embolic hook: abs; pres 0 0 0 0 0 0 0 0 3. DSA sclerotization: mem; light; heavy 1 1 0 1 2 1 1 1 4. DSA: short; long 0 0 0 0 1 0 0 0 5. STM: absent; present 0 0 0 0 0 0 0 1 6. MSA: absent; present 0 1 1 1 0 0 0 0 7. TP; present; absent 0 0 0 0 0 0 0 0 8. TP shape; straight; spiral; ectal; anterior 2 0 1 1 3 0 1 0 9. LC: absent; present 0 0 0 0 0 0 0 1 10. ARP: absent; present 1 1 1 1 1 1 0 0 11. ARP shape; short; long and spiral 0 0 1 1 0 0 — — 12. PT papillae: absent; present 1 1 1 1 0 0 0 — 13, TS: absent; present 0 0 0 0 0 1 0 1 14. P, ventral view; narrow; wide 0 0 0 0 0 0 0 0 15. CE: small; large 1 0 0 0 1 0 0 1 16. TA length: short; long 0 1 0 0 1 1 0 0 17. ETP: strong; weak or absent 1 0 0 1 0 0 0 0 18. patella apophysis: absent; present 0 0 0 0 1 1 0 0 Female genitalia 19. VP: beyond EF; shallow inv; deep inv 0 0 2 2 0 2 2 2 20, median VP: convex; concave 0 0 0 0 1 0 1 1 21. DPP: rect; invert tri; tri; trapezoid 1 0 0 0 1 0 0 2 22. sides DPP: convex; concave 0 0 0 0 0 0 0 0 23. vent mar DPP: concave; convex 1 0 0 0 1 1 1 1 24. DF sclerotization: light; heavy 0 0 0 0 0 0 0 0 25. CO: small; large 1 0 0 1 0 0 1 1 26. CD origin: ectal; mesal 1 0 1 1 1 1 1 1 27. CD anterior proj: absent; present 0 0 1 1 1 0 0 1 28. CD encapsulation: absent; present 0 0 1 0 1 1 1 0 29. CDC: partial; complete — — 1 — 1 0 1 — 30. ant lat CDC: concave; straight; convex — — 0 — 0 — 1 • — 31. lat mar CDC: curved; bowed — — 0 — 0 — 0 — 32. post CDC orientation: post; mesal ■ — — 0 — 0 0 0 — 33. FD origin: posterior; mesal 1 0 1 1 1 1 1 1 34. FD shape: sinuous; spiral 0 1 0 0 0 0 0 1 Somatic morphology 35, cephalic region: not raised; raised 0 0 0 0 1 0 0 ■ 1 36. post PME lobe: absent; present 0 0 1 0 0 0 0 0 37. cuticular pores: absent; present 0 0 1 ? 0 0 1 0 38. cheliceral file: ridged; scaly; imb 2 2 2 ? 2 1 0 1 39. dorsal spur: absent; present 0 0 0 0 0 1 0 0 40. tibia III m=setae: two; one 0 1 0 0 0 1 1 1 41. Tm IV: absent; present 0 0 0 0 1 1 1 1 MILLER^REVISION OF SISICOTTUS 599 Table 5. — Extended. Gon- gyli- dium ruf- ipes Oedo- tho- rax gibo- sus Sisi- cottus monti- genus Sisi- cottus quoy- lei Sis- cottus pano- peus Sis- cottus mont- anus Sis- cottus cross- ocla- vis Sisi- cottus cyn- thiae Sisi- cottus orites Sisi- cottus ne si- des Sisi- cottus aenig- mat- icus St Cl RI RC 1 1 1 1 1 1 1 1 ? 2 0.50 0.67 0.33 0 0 1 1 1 1 1 1 1 1 ? 1 1.00 1.00 1.00 1 1 0 1 1 1 2 2 2 2 ? 4 0.50 0.60 0.30 1 0 1 1 0 0 0 1 1 1 ? 4 0.25 0.50 0.13 0 0 1 1 1 1 1 1 1 1 ? 2 0.50 0.88 0.44 0 1 1 1 1 1 1 1 1 1 ? 2 0.50 0.80 0.40 1 1 1 1 1 1 1 1 1 1 ? 1 1.00 1.00 1.00 7 4 0.75 0.50 0.38 1 1 1 1 1 1 1 1 1 1 ? 1 1.00 1.00 1.00 0 0 0 0 0 0 0 0 0 0 7 1 1.00 1.00 1.00 7 1 1.00 1.00 1.00 1 1 1 1 1 1 1 1 1 1 7 2 0.50 0.50 0.25 1 1 1 1 1 1 1 1 1 1 7 2 0.50 0.80 0.40 1 0 1 1 1 1 1 1 1 1 7 2 0.50 0.88 0.44 1 1 1 1 1 1 1 1 1 1 7 3 0.33 0.50 0.17 0 0 1 1 0 0 0 0 0 0 7 4 0.25 0.25 0.06 0 0 1 1 1 0 1 0 0 0 7 5 0.20 0.20 0.04 0 0 0 0 0 0 0 0 0 0 7 2 0.50 0 0 2 2 2 2 1 1 1 1 2 2 2 4 0.50 0.60 0.30 0 1 1 1 1 1 1 1 1 1 1 3 0.33 0.60 0.20 2 0 0 0 0 1 1 1 1 1 3 6 0.50 0.57 0.29 0 0 0 0 0 0 0 0 0 1 1 1 1.00 1.00 1.00 1 1 1 1 1 1 1 1 1 1 1 1 1.00 1.00 1.00 0 0 0 0 0 0 1 1 1 1 1 1 1.00 1.00 LOO 1 0 0 0 0 0 0 0 0 0 0 4 0.25 0.25 0.06 1 1 0 0 0 0 0 0 0 0 0 2 0.50 0.88 0.44 0 1 1 1 1 1 1 1 1 1 1 4 0.25 0.25 0.06 1 1 1 1 1 1 1 1 1 1 1 3 0.33 0.33 0.11 1 0 1 1 1 1 1 1 1 1 1 2 0.50 0 0 1 — 1 1 2 2 2 2 2 2 2 2 1.00 1.00 1.00 0 — 0 0 0 0 0 0 1 1 1 1 1.00 1.00 LOO 0 0 0 0 0 0 0 1 1 1 1 1 1.00 1.00 1.00 1 0 1 1 1 1 1 1 1 1 1 2 0.50 0 0 0 0 0 0 0 0 1 1 1 1 1 3 0.33 0.67 0.22 0 0 0 0 0 0 0 0 0 0 7 2 0.50 0 0 0 1 0 0 0 0 0 0 0 0 7 2 0.50 0 0 0 1 0 0 0 0 0 0 0 0 7 3 0.33 0 0 1 1 2 2 2 2 2 2 2 2 2 3 0.67 0.67 0.44 1 1 0 0 0 0 0 0 0 0 7 3 0.33 0 0 1 1 0 0 0 0 0 0 0 0 0 3 0.33 0.60 0.20 1 1 0 0 0 0 0 0 0 0 0 2 0.50 0.80 0.40 600 THE JOURNAL OF ARACHNOLOGY Figure 1 15= — Preferred most parsimonious cladogram of Sisicottus species based on successive character weighting and implied weights. Underlined numbers indicate unambiguous character change optimizations; the remaining characters were optimized to favor reversal over parallel evolution (Farris optimization) unless explicitly justified in the text. Bremer support values appear as boxed numbers to the right of each applicable intemode. The circled numbers at each node identify clades discussed in the text. MILLER— REVISION OF SISICOTTUS 601 tion. Grant Number DKB-9505031 (Theodore W. Pietsch, principal investigator). Significant financial support for this study was provided by a generous grant from the AMNH Theodore Roosevelt Memorial Fund. Equipment and financial support was provided by an NSF-PEET grant to G. Hormiga and J. Coddington (DEB-9712353). The George Washington University provided support in the form of a Weintraub fellowship. Western Carolina University, the National Museum of Natural History (Smithsonian Institution), and the George Washington University all provid- ed valuable lab space, materials, and equip- ment. FA. Coyle, G. Hormiga, J. Coddington, N. Scharff, P. van Helsdingen, K. Thaler, M. Saaristo, P. Bjorn, D. Southard, R.C. Bruce, R.H. Lumb, S. Rundle, J. West, and an anon- ymous reviewer all made helpful comments on various drafts of the manuscript. F.A. Coyle was an excellent mentor during much of this project providing copious advice, guidance, and support. G. Hormiga and J. Coddington served as advisors during the later stages of this research and provided extremely valuable assistance, advice, and resources which sig- nificantly improved the value of this product. G. Hormiga gave me unrestricted access to unpublished data. J.H. Redner sent me illus- trations of two of the new species described in this paper. R.L. Crawford shared his knowl- edge of Sisicottus taxonomy and ecology. H. D. Cameron made sure that my proposed species names were in agreement with the rules of zoological nomenclature and kindly shared some of his extensive knowledge about the history of arachnological terms and no- menclature. Y. Marusik searched for addition- al records of Sisicottus in Asia. W. Rapp and R.J. Snider corresponded with me about du- bious published records of Sisicottus species from Nebraska and Michigan, respectively. JoAn Hudson of the Clemson University elec- tron microscope facility was extremely gen- erous with her time, guidance, and materials. G. Venable donated expertise and materials that improved the quality of the SEM plates. LITERATURE CITED Aitchison-Benell, C.W. & C.D. Dondale. 1990. A checklist of Manitoba spiders (Araneae) with notes on geographic relationships. Naturaliste Canada (Rev. Ecol. Syst.), 117:215-237. Barrows, W.M. 1945. New spiders from the Great Smoky Mountains National Park. Ann. Entomol. Soc. America, 38:70-76. Bishop, S.C. & C.R. Crosby. 1938. Studies in American spiders: Miscellaneous genera of Eri- goneae, part II. J. New York Entomol. Soc., 46: 55-107. Blest, A.D. 1976. The tracheal arrangement and the classification of linyphiid spiders. J. ZooL, London, 180:185-194. Bonnet, P. 1958. Bibliographia Araneoram. Tome II, 4*"® partie; N-S. Douladoure, Toulouse, 3027- 4230 pp. Bragg, P.D. & R.E. Leech. 1972. Additional re- cords of spiders (Araneida) and harvestmen (Phalangida) for British Columbia. J. Entomol. Soc. British Columbia, 69:67-71. Bremer, K. 1988. The limits of amino acid se- quence data in angiosperm phylogenetic recon- struction. Evolution, 42:795-803. Brignoli, P.M. 1983. A Catalog of the Araneae De- scribed Between 1940 and 1981. Manchester Univ. Press, Manchester, 755 pp. Buckle, D.J., D. Carroll, R.L. Crawford & V.D. Roth. 1994. Linyphiidae of America North of Mexico: Checklists, Synonymy, and Literature. Version 2.1. Unpubl. document, send inquiries to D.J. Buckle, 620 Albert Ave., Saskatoon, Sas- katchewan S7N 1G7, Canada. Carpenter, J.M. 1988. Choosing among multiple equally parsimonious cladograms. Cladistics, 4: 291-296. Chamberlin, R.V. 1919. New western spiders. Ann. Entomol. Soc. America, 12:239-260. Chamberlin, R.V. 1921. Linyphiidae of St. Paul Is- land, Alaska. J. New York Entomol. Soc., 29:35- 42. Chamberlin, R.V. & W. Me. 1933. Spiders of the Raft River Mountains of Utah. Bull. Univ. Utah (Biol. Series), 23:1-79. Chamberlin, R.V. & W. Me. 1939. Studies on North American spiders of the family Micry- phantidae. Congr. Int. Entomol. 7, Berlin, Verb., 1:56-73. Chamberlin, R.V. & W. Ivie. 1944. Spiders of the Georgia region of North America. Bull. Univ. Utah (Biol. Series), 35:1-267. Coddington, J.A. 1983. A temporary slide mount allowing precise manipulation of small struc- tures. Verb. Naturwiss. Ver. Hamburg, 26:291- 292. Coddington, J.A. 1989. Spinneret silk spigot mor- phology: evidence for the monophyly of orb- weaving spiders, Cyrtophorinae (Araneae), and the group Theridiidae plus Nesticidae. J. Arach- noL, 17:71-95. Coddington, J.A. 1990. Ontogeny and homology in the male palpus of orb-weaving spiders and their relatives, with comments on phylogeny (Ar- 602 THE JOURNAL OF ARACHNOLOGY aneoclada: Araneoidea, Deinopoidea). Smithson- ian Contrib. ZooL, 496:1-52. Coddington, J.A. & H.W. Levi, 1991. Systematics and evolution of spiders (Araneae). Ann. Rev. Ecol. Syst., 22:565-592. Coddington, J.A. & N. Scharff. 1994. Problems with zero-length branches. Cladistics, 10:415- 423. Coddington, J.A. & N.I. Scharff. 1996. Problems with “soft” polytomies. Cladistics, 12:139-146. Crawford, R.L. 1988. An annotated checklist of the spiders of Washington. Burke Mus. Contrib. An- thropol. Nat. Hist., 5:1-48. Crawford, R.L. & J.S. Edwards. 1988. Alpine spi- ders and harvestmen of Mount Rainier, Washing- ton, U.S.A.: taxonomy and bionomics. Canadian J. Zook, 67:430-446. Crosby, C.R. 1905. A catalog of the Erigoninae of North America, with notes and descriptions of new species. Proc. Acad. Nat. Sci. Philadelphia, 57:301-343. Crosby, C.R. & S.C. Bishop. 1927. New species of Erigoneae and Theridiidae. J. New York En- tomol. Soc., 35:147-157. de Queiroz, K. & M.J. Donoghue. 1990. Phylo- genetic systematics and species revisited. Cladis- tics, 6:83-90. Dondale, C.D., 1990. Litter Araneae (Araneida). Pp. 477-502, In Soil Biology Guide (D.L. Din- dal, ed.). John Wiley & Sons, New York. Dondale, C.D., J.H. Redner & Y.M. Marusik. 1997. Spiders (Araneae) of the Yukon. Pp. 73-113, In Insects of the Yukon. (H.V. Danks & J.A. Downes, eds). Biol. Surv. of Canada (Terrestrial Arthropods), Ottawa. Donoghue, M.J. 1985. A critique of the biological species concept and recommendations for a phy- logenetic alternative. Bryologist, 88:172-181. Drew, L.C. 1967. Spiders of Beaver Island, Mich- igan. Publ. Mus. Michigan State Univ., Biol. Sen, 3:153-207. Emerton, J.H. 1882. New England spiders of the family Theridiidae, Trans. Connecticut Acad. Arts Sci., 6:1-86. Emerton, J.H. 1920. Catalog of the spiders of Can- ada known to the year 1919. Trans, Roy. Cana- dian Inst., 12:309-338. Farris, J.S. 1969. A successive approximations ap- proach to character weighting. Syst. ZooL, 18: 374-385. Farris, J.S. 1988. Hennig86, version 1.5. Program and documentation. Computer program distrib- uted by D. Lipscomb, Dept, of Biological Sci- ences, The George Washington University, Washington D.C., 20052. Goloboff, PA. 1993a. NONA. Noname (a bastard son of Pee- Wee), version 1.6 (32 bit version). Program and documentation. Computer program distributed by J.M. Carpenter, Dept, of Entomol- ogy, American Museum of Natural History, New York. Goloboff, PA. 1993b. Pee- Wee. (P)arsimony and (I)mplied (WE)ights, version 2.6 (32 bit version). Program and documentation. Computer program distributed by J.M. Carpenter, Dept, of Entomol- ogy, American Museum of Natural History, New York. Goloboff, P. A. 1993c. Estimating character weights during tree search. Cladistics, 9:83-91. Goloboff, P.A. 1995. PHASE (PH)ylogenetic (A)nalysis for (S)ankovian (T)ransformations, version 1.1 (32 bit version). Program and docu- mentation. Computer program distributed by J.M. Carpenter, Dept, of Entomology, American Museum of Natural History, New York. Griswold, C.E., J.A. Coddington, G. Hormiga & N. Scharff. 1998. Phylogeny of the orb web build- ing spiders (Araneomorphae, Orbiculariae: Dei- nopoidea, Araneoidea). Zool. J. Linnean Soc., 123:1-99. Hennig, W 1966. Phylogenetic Systematics. (D.D. Davis & R. Zangerl, trans.). Univ. of Illinois Press, Urbana. 263 pp. Holm, A. 1960. On a collection of spiders from Alaska. Zool. Bidr. Uppsala, 33:109-134. Holm, A. 1967. Spiders (Araneae) from West Greenland. Medd. Gr0nland, 184:1-99. Holm, A. 1979. A taxonomic study of European and east African species of the genera Pelecopsis and Trichopterna (Araneae, Linyphiidae), with descriptions of a new genus and two new species of Pelecopsis from Kenya. Zool. Scripta, 8:255- 278. Hormiga, G. 1993. Implications of the phylogeny of Pimoidae for the systematics of linyphiid spi- ders (Araneae, Araneoidea, Linyphiidae). Mem. Queensland Mus., 33:533-542. Hormiga, G. 1994a. Cladistics and the comparative morphology of linyphiid spiders and their rela- tives (Araneae, Araneoidea, Linyphiidae). Zool. J. Linnean Soc., 111:1-71. Hormiga, G. 1994b. A revision and cladistic anal- ysis of the spider family Pimoidae (Araneoidea: Araneae). Smithsonian Contrib. Zool., 549:1- 104. Hormiga, G. in press. Higher level phylogenetics of erigonine spiders (Araneae, Linyphiidae, Eri- goninae). Smithsonian Contributions to Zoology. Hormiga, G., W.G. Eberhard & J.A. Coddington. 1995. Web-construction behavior in Australian Phonognatha and the phylogeny of nephaline and tetragnathid spiders (Araneae: Tetragnathi- dae). Australian J. Zook, 43:313-364. International Commission on Zoolological Nomen- clature. 1985. International Code of Zoolologi- cal Nomenclature. 3rd ed. Intern. Trust for Zook Nomenclature, London. 338 pp. MILLER— REVISION OF SISICOTTUS 603 Ivie, W. 1967. Some synonyms in American spi- ders. J. New York EntomoL Soc., 75:126-131. Jennings, D.T, M.W. Houseweart, C.D. Dondale & J.H. Redner, 1988. Spiders (Araneae) associated with strip-clearcut and dense spmce-fir forests of Maine. J. ArachnoL, 16:55-70. Kaston, BJ. 1981. Spiders of Connecticut. Rev. ed. State Geol. Nat. Hist. Surv. Connecticut Bull., 70:1-1020. Koponen, S. 1987. Communities of ground-living spiders in six habitats on a mountain in Quebec, Canada. Holarct. EcoL, 10:278-285. Lowrie, D.C. & W.J. Gertsch. 1955. A list of the spiders of the Grand Teton Park Area, with de^ scriptions of some new North American spiders. American Mus. Nov., 1736:1-29. Maddison, W.P. & D.R. Maddison. 1992. Mac- Clade: Analysis of phylogeny and character evo- lution, version 3.0. Sinauer Assoc., Sunderland, Massachusetts. Marx, G. 1890. Catalog of the described Araneae of temperate North America. Proc. United States Nat. Mus., 12:497-594. Merrett, P. 1963. The palpus of male spiders of the family Linyphiidae. Proc. Zool. Soc. London, 140:347-467. Millidge, A.F. 1983. The erigonine spiders of North America. Part 6. The genus Walckenaeria Blackwall (Araneae, Linyphiidae). J. ArachnoL, 11:105-200. Millidge, A.F. 1984a. The taxonomy of the Liny- phiidae, based chiefly on the epigynal and tra- cheal characters (Araneae: Linyphiidae). Bull. British ArachnoL Soc., 6:229-267. Millidge, A.F. 1984b. The erigonine spiders of North America. Part 7. Miscellaneous genera (Araneae, Linyphiidae). J. ArachnoL, 12:121- 169. Nelson, G. 1989. Cladistics and evolutionary mod- els. Cladistics, 5:275-289. Nixon, K.C. & Q.D. Wheeler. 1990. An amplifi- cation of the phylogenetic species concept. Cla- distics, 6:211-223. Petrunkevitch, A. 1911. A synonymic index-cata- log of spiders of North, Central and South Amer- ica with all adjacent islands, Greenland, Ber- muda, West Indies, Terra Del Fuego, Galapagos, etc. Bull. American Mus. Nat. Hist., 29:1-791. Platnick, N.I. 1989. Advances in Spider Taxonomy 1981-1987. Manchester Univ. Press, Manchester. 637 pp. Platnick, N.I. 1993. Advances in Spider Taxonomy 1988-1991. New York EntomoL Soc. 846 pp. Platnick, N.I. 1997. Advances in Spider Taxonomy 1992-1995. New York EntomoL Soc. 976 pp. Rapp, W.E 1980. A catalog of spiders of Nebraska. Nov. Arthropodae, 1:1-39. Roewer, C.F 1942. Katalog der Araneae von 1758 bis 1940. Volume 1. Natura, Bremen. 1040 pp. Scharff, N. & J.A. Coddington. 1997. A phyloge- netic analysis of the orb-weaving spider family Araneidae (Arachnida, Araneae). Zool. J. Lin- nean Soc., 120:355-434. Swofford, D.L. 1993. PAUP. (P)hylogenetic (A)nalysis (U)sing (P)arsimony, version 3.1. Il- linois Nat. Hist. Survey, Champaign, Illinois. West, R., C.D. Dondale & R.A. Ring. 1984. A re- vised checklist of the spiders (Araneae) of British Colombia. J. EntomoL Soc. British Columbia, 81:80-98. Wheeler, Q.D. & J.V McHugh. 1994. A new southern Appalachian species, Dasycerus bicolor (Coleoptera: Staphylinidae: Dasycerinae), from declining endemic fir forests. The Coleopts. Bull., 48:265-271. Wheeler, Q.D. & K.C. Nixon. 1990. Another way of looking at the species problem: a reply to de Queiroz and Donoghue. Cladistics, 6:77-81. Wiehle, H. 1960. Spinnentiere oder Arachnoidea (Araneae). XI: Micryphantidae B Zwergspinnen. Tierwelt Deutschlands, 47:1-620. Zujko-Miller, J. 1999. On the phylogenetic rela- tionships of Sisicottus hibernus (Araneae, Liny- phiidae, Erigoninae). J. Arachnol, 27:44-52. Manuscript received 2 July 1997, revised 18 De- cember 1998. Note added in proof : Dondale (1990) included two illustrations labeled "'Sisicottus sp.” in his key to litter-inhabiting spiders in North America. Although there is no documentation of what specimens the illustrations were based on, his figure 17.142 is probably the palpal tibia of S. quoylei; his figure 17.157 is probably the epigynum of S. montanus. 1999. The Journal of Arachnology 27:604-662 RADIATION OF THE GENUS DYSDERA (ARANEAE, DYSDERIDAE) IN THE CANARY ISLANDS: THE ISLAND OF TENERIFE Miquel A. Arnedo* and Carles Ribera; Departament de Biologia Animal, Universitat de Barcelona, Avinguda Diagonal 645, E-08028 Barcelona, Spain ABSTRACT. An overwhelming number of endemic species belonging to the spider genus Dysdera have been reported from the oceanic archipelago of the Canary Islands. A complete taxonomic revision is currently being performed in order to assess the extent of this species’ radiation, as well as to supply enough data to place it in a phylogenetic framework. The present article is devoted to the Dysdera species inhabiting the island of Tenerife. A total of 22 species is recognized in Tenerife, including the cosmopolitan Dysdera crocota C.L. Koch 1839. Two new species are described: Dysdera guayota new species and Dysdera hernandezi new species. Ten new synonymies are reported: D. moquinalensis Wunderlich 1991 and D. vilaflorensis Wunderlich 1991 = D. brevispina Wunderlich 1991; D. medinae Wunderlich 1991 = D. cribeilata Simon 1883; D. inaequuscapillata Wunderlich 1991 — D. crocota; D. pergrada Wunderlich 1991, D. pseudopergrada Wunderlich 1991, D. tabaibaensis Wunderlich 1991, D. teideensis Wunderlich 1991 and D. tenerijfensis Strand 1908 = D. macra Simon 1883; D. obscuripes Wunderlich 1991 = D. propinqua Ribera, Ferrandez & Blasco 1985. Sixteen species are redescribed: D. ambulotenta Ribera, Ferrandez & Blasco 1985; D. brevisetae Wunderlich 1991, D. brevispina Wunderlich 1991; D. chioensis Wunderlich 1991; D. cribeilata Simon 1883; D. curvisetae Wunderlich 1987; D. esquiveli Ribera & Blasco 1986; D. gibbifera Wunderlich 1991; D. gollumi Ribera & Amedo 1994; D. labradaensis Wunderlich 1991; D. macra Simon 1883; D. minutissima Wunderlich 1991; D. montanetensis Wunderlich 1991; D. propinqua Ribera, Ferrandez & Blasco 1985; D. unguimmanis Ribera, Ferrandez & Blasco 1985 and D. volcania Ribera, Ferrandez & Blasco 1985. The females of four species: D. brevisetae, D. brevispina, D. minutissima and D. montanetensis are described for the first time. Females formerly assigned to both D. gibbifera and D. volcania are considered to be incorrect identifications. A neotype is designated for D, macra. The presence of D. rugichelis Simon 1907 in Tenerife is considered to be doubtful. Ecological and distributional patterns of the species are discussed. Species of the spider genus Dysdera La- treille 1 804 are usually found in slightly damp but warm ground habitats. They are nocturnal wandering hunters that spend daytime in silk= en cocoons under stones, logs or bark (Rob= erts 1995; pers. obs.). Dysdera specimens are not unusual in caves, which can be considered as an expansion of their typical habitats; and several cases of troglomorphic species have been reported (Ribera 1983, 1993; Ribera et al. 1986). This species-rich genus includes about 200 species with a circum-Mediterra- nean distribution, with the single exception of the anthropophilous cosmopolitan D. crocota C.L. Koch 1839. The so-called Macaronesian archipelagos (Fig. 1) represent the western- most limit of Dysdera's range. One of these ' Division of Insect Biology, ESPM, University of Califomia-Berkeley, 201 Wellman Hall, Berke- ley, California 94720-3112 USA volcanic archipelagos, the Canary Islands, harbors about 50 endemic species (Simon 1883, 1907; Strand 1908; Schmidt 1973; Ri- bera et al. 1985; Ribera & Blasco 1986; Wun- derlich 1987, 1991; Ribera & Amedo 1994; Arnedo et al. 1996; Amedo & Ribera 1997), which represent about one quarter of the de- scribed species in the genus to date. This fig- ure is even more remarkable when compared with the number of endemics in the remaining archipelagos: one from the Azores (undescri- bed species), five from Madeira (Denis 1962; Wunderlich 1994) and one from Cape Verde (Berland 1936). In addition, seven of these species were troglobites with morphological adaptations to the hypogean environment. Nevertheless, this overwhelming number of Dysdera species held by the Canaries could suggest a taxonomic artifact instead of a true species radiation. A deeper look into Canarian Dysdera taxonomy revealed some instances 604 ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 605 Figures 1-3. — Maps 1-3. 1, Macaronesian archipelagos; 2, the Canary Islands; 3, Tenerife, with its oldest areas encircled. that could, at least, call into question this amazing number of endemics. On the one hand, 22 out of the 50 recognized species were described from only one of the sexes, 19 of which were known from a single specimen. Moreover, some of the species lacked infor- mation regarding their locality, type material was lost, or both. Finally, 27 species were de- scribed in a single publication together with 106 new species from the Macaronesia (Wun- derlich 1991). On the other hand, most of the published descriptions of the Canarian Dys- dera were vague enough to correspond to more than one species, or failed to supply the necessary information for the study of such interesting spiders in a phylogenetic frame- work. With the aim of confirming the existence of this radiation, completing the species descrip- tions as well as their geographical distribu- tions and, finally, offering enough data to per- form a phylogenetic analysis of the group, a major revisionary work on Canarian Dysdera is currently being developed (Ribera & Ar- nedo 1994; Amedo & Ribera 1996; Amedo et al. 1996; Amedo & Ribera 1997). The present article is devoted to the taxonomic revision of the genus on the Island of Tenerife. The Canary Islands lie in the Atlantic Ocean 100 km from the north-western coast of Africa (Fig. 2). The different volcanic ep- isodes that formed the archipelago are prob- ably the result of a propagating fracture orig- inated in the Atlas formation during the Alpine orogeny, about 25 Mya ago (Anguita & Heman 1975). This model would explain both the reduction of the age of the islands from east to west and the continuation of ac- tive volcanism in the older islands. The ap- proximate ages of the subaerial parts of the islands, as recovered using the K-Ar tech- nique, range from about 22 Mya to less than 1 Mya. The estimated geological age for each island is: Fuerteventura 20-22 Mya, Lanzarote 15-19 Mya, Gran Canaria 14-16 Mya, Ten- erife 11.6-14 Mya, La Gomera 10-12 Mya, La Palma 1.6-2 Mya and El Hierro 0.8-1 Mya (Cantagrel et al. 1984, Mitchell-Thome 1985, Ancochea et al. 1990, Coello et al. 1992). The island of Tenerife is located roughly at the center of the line drawn through the archipel- ago. Tenerife is both the biggest (2058 km^) and the highest (3717 m) island in the archi- pelago. Elevation together with trade winds play important ecological roles on oceanic islands, especially at tropical and subtropical latitudes. They are both responsible for the presence and distribution of the different ecological zones. In the particular case of the Canaries, the joint effect of the hunfid and cool NE trade winds, between altitudes of 400-1200 m, and the dry 606 THE JOURNAL OF ARACHNOLOGY trade winds from the NW, above 2000 m, cause a temperature inversion. In this area, a nearly permanent cloud belt is formed. Con- sequently, strong ecological segregation is ob- served between northern, more humid, and southern, dryer slopes. Five major ecological zones can be recognized on northern slopes of the islands. The first, from the seashore to up to 250 m, is characterized by the presence of dry-arid subtropical shrubs. The second, from 250-600 m, features humid to semi-arid trop- ical shrubs and woods. The third, from 600- 1000 m, is covered by the cloud belt and fea- tures a typical subtropical wood, the so-called laurel forest. In the fourth, from 1000-2000 m, an endemic pine forest occurs. Finally, dry subalpine shrub is present from 2000 m to the top. Southern slopes lack a laurel forest zone and transition between sub-arid shrubs and the pine forest takes place at higher elevation. Apart from these climatic -related ecosys- tems, an additional ecological zone is present in volcanic islands: the hypogean environ- ment. The subterranean environment in the Canaries is represented by both lava tubes and the MSS (mesocavernous shallow stratum) (Juberthie et al. 1980, 1981; Oromf et al. 1986; Medina 1991). Due to their short life- span, lava tubes are found only in areas of the islands with a relatively recent pahoehoe-like basaltic volcanism. This explains the lack of tubes in the islands of La Gomera and their scarcity in Gran Canaria and most of Fuerte- ventura. However, even in the absence of caves, a very rich underground environment, in the form of shallow, intermediate-sized, in- terconnected voids, is present in all the is- lands. Before the present study 29 endemic spe- cies of Dysdera had been reported from Ten- erife, by far the most species-rich island in the archipelago. These species were: D. ambulo- tenta Ribera et al. 1985 (d,?; one locality); D. brevisetae Wunderlich 1991 (d, single specimen); D. brevispina Wunderlich 1991 (d, single specimen); D. chioensis Wunder- lich 1991 ($, one locality); D. cribellata Si- mon 1883 (<3,$); D. curvisetae Wunderlich 1987 (d, single specimen); D. esquiveli Ri- bera & Blasco 1986 (d,?); D. gibbifera Wun- derlich 1991 (d,$); D. gollumi Ribera & Ar- nedo 1994 (9, one locality); D. iguanensis Wunderlich 1987 (d,?);/). inaequuscapillata Wunderlich 1991 (<3,?); D. insulana Simon 1883 (6,9; one locality); D. labradaensis Wunderlich 1991 (9, one locality); D. levipes Wunderlich 1987 (6,9); D. medinae Wunder- lich 1991 ((3,9); D. minutissima Wunderlich 1991 (<3, single specimen); D. montanetensis Wunderlich 1991 (d, single specimen); D. moquinalensis Wunderlich 1991 ((3, single specimen); Z). obscuripes Wunderlich 1991 (<3,9); D. pergrada Wunderlich 1991 ((3,9; one locality); D. propinqua Ribera et al. 1985 (c3, single specimen); D. pseudopergrada Wunderlich 1991(d,9); D. rugichelis Simon 1907 ((3, single specimen in Tenerife); D. ta- baibaensis Wunderlich 1991 ((3, single spec- imen); D. teideensis Wunderlich 1991 (c3,9); D. tenerijfensis Strand 1908 (9; single spec- imen, lost); D. unguimmanis Ribera et al. 1985 (9, single specimen); D. vilaflorensis Wunderlich 1991 (d, single specimen) and D. volcania Ribera et al. 1985 (<3,9; one locali- ty) (Bosenberg 1895; Strand 1908; Denis 1941, 1953; Schmidt 1975; Ribera et al. 1985; Wunderlich 1987, 1991; Ribera & Amedo 1994; Amedo et al. 1996; Amedo & Ribera 1997). Six of these species displayed morpho- logical adaptations to the hypogean environ- ment and were considered to be tme troglob- ites. With a single exception {Dysdera ratonensis Wunderlich 1991 from La Palma), the lava tubes of Tenerife hold all troglo- morphic Dysdera documented so far in the Canaries. METHODS The current study was based on the adop- tion of the so-called ‘diagnosability’ (Baum 1992) phylogenetic species concept (Nixon & Wheeler 1990, 1992; Wheeler & Nixon 1990, Davis & Nixon 1992). Species are recognized as the most exclusive set of populations that display a unique combination of character- states, when semaphoronts are compared (Da- vis & Nixon 1992). This concept was selected because it is easily applicable in practice, it avoids any reference to processes, and is fully compatible with a phylogenetic framework. However, this definition is not free of theo- retical problems (Frost & Kluge 1994) and has been considered to be excessively restrictive. In addition, in the present approximation, only morphological characters were taken into ac- count, which has probably resulted in an un- derestimation of the total number of species. Additional studies considering molecular, eco- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 607 logical or behavioral characters would be nec- essary in order to recover the total amount of diversity of the genus. The first stage in the assessment of the taX“ onomic status of the Tenerifean species was to gather a large number of specimens (350), which were made available from scientific in- stitutions, various personal collections, and three collection expeditions to the island by the authors. The following colleagues and mu- seum kindly supplied material for the present study: Dr. E. Enghoff from the Zoologisk Mu- seum of Copenhagen (ZMK), R. Garcia ‘Felo’ (S/C de la Palma, Canary Islands) (RG), F. Gasparo (Trieste, Italy) (FG), Mr. RD. Hill- yard from the Natural History Museum of London (BMNH), Dr. M. Grasshoff from the Forschungsinstitut und Naturmuseum Senck- enberg (SMF), Dr. R Oromf from the Univ- ersidad de La Laguna (UL), Dr. G. Ortega from the Museo de Ciencias Naturales de San- ta Cruz de Tenerife (MCNT), Dr. C. Rollard from the Museum National d’Histoire Natu- relle de Paris (MNHN) and J. Wunderlich (Straubenhardt, Germany) (JW). The material provided by the authors’ expeditions is stored at the collection of Arachnids of the Univer- sity of Barcelona, Spain (UB). Characters were investigated under a Wild Heerbrugg (12~100X) dissecting microscope. Female vulva (= endogyne, Mcheidze 1972) was removed and muscle tissues digested us- ing a KOH (35%) solution before observation. Male bulbi and spinnerets were removed, cleaned by means of ultrasound and examined using a HITACHI S~2300 scanning electron microscope at 10-15 Kv. All measurements are in millimeters. Somatic morphology mea- surements were taken using an ocular micro- meter in the dissecting microscope. Characters examined together with their diagnostic reso- lution have been discussed elsewhere (Amedo et al. 1996). All characters were recorded in DELTA format (Dallwitz 1980; Dallwitz et al. 1993). Terminology. — Leg spination was record- ed for each segment using the format of Ar- nedo et al. (1996). Only spines present on fe- morae, patellae and tibiae were considered. In femora, spines are usually arranged in one or two (anterior and posterior) rows parallel to the segment. The number of rows and spines per row were recorded (In Tables 2-16, the number of spines in each row were separated by a slash). In patellae, the number of spines and their position (ventral or dorsal) was cod- ed. Tibiae usually show the most complex spi- nation pattern. For each tibia, the number of spines was recorded from four zones (here- after referred as ‘bands’): proximal, medial- proximal, medial-distal and distal. For coding purposes, the number of spines on the frontal, medial and posterior regions were separated by points. In the descriptions (intra-individual variation), hyphens separate the number of spines on each side of the body if different. In the tables (intraspecific variation) hyphens separate the minimum and maximum number of spines observed in any specimen. Between 6-10 individuals were examined from each species whenever possible. Structures of male bulb and female vulva were mostly named after Deeleman-Reinhold & Deeleman (1988), with the addition of sev- eral features particular to Canarian Dysdera (Amedo et al. 1996; Amedo & Ribera 1997). Nevertheless, some new characters from the female vulva have been added or have been redefined and deserve further considerations. The vulva of the genus Dysdera is divided into two major diverticles: an anterior diver- tide and a posterior one. They are separated by the epigastric furrow at the ventral side, and the oviduct opening at the dorsal one. The posterior diverticle is mostly membranous, with the single exception of the transversal bar. This is located at the anterior dorsal side and holds a frontal projection (‘bursal valve’, V) that closes the oviduct openings. On the other hand, most of the anterior diverticle is usually sclerotized. The most conspicuous stmcture is a T-shaped spermatheca (S) locat- ed at the ventral side of the most anterior mar- gin. The medial lateral margins of the anterior diverticle are invaginated forming two differ- ent pouches: a dorsal pouch, which corre- sponds to the so-called ‘dorsal arch’ (DA) (Deeleman-Reinhold & Deeleman 1988) and a ventral one, hereafter referred as ‘ventral arch’ (VA). The DA is usually completely sclerotized. The dorsal side of the DA locks the V and is called the ‘dorsal fold’ (DF) (Ar- nedo et al. 1996). The fold that separates both diverticles is named the ‘major fold’ (MF), to differentiate it from several additional folds that are sometimes found on the DA lateral borders. The development and sclerotization degree of the MF are highly variable. In some 608 THE JOURNAL OF ARACHNOLOGY Table 1. — Abbreviations used in text and figures (Figs. 4-12). Female genitalia DA = dorsal arch DF = dorsal fold MF = major fold S = spermatheca TB = transversal bar V = bursal valve VA = ventral arch AVD = additional ventral diverticle Eyes AME = anterior medial eyes PME = posterior medial eyes PLE = posterior lateral eyes Cheliceral teeth B = basal tooth M = medial tooth D = distal tooth Male copulatory bulb T = tegulum DD = distal division IS = internal sclerite ES = external sclerite DH = distal haematodoca C = crest AC = additional crest LF = lateral fold over L, between internal and exter- nal sclerites L = lateral sheet AL = additional lateral sheet at back internal border P = posterior apophysis Spinnerets ALS = anterior lateral spinnerets PMS — posterior medial spinnerets PLS = posterior lateral spinnerets MS — major amputate gland spigot PS = polar pyriform gland spigot continental Dysdera species, the MF almost entirely isolates the DA from the VA. The margins of the MF may be markedly separated from each other or stuck together forming an internal rim. The VA exhibits a wide range of sclerotization levels, from mostly sclerotized to completely membranous. In most of the Canarian representatives, an additional ventral diverticle (AVD) in the VA has been ob- served. The AVD is recognized by an internal rim ventral to the MF and by its own external sclerotization, usually tooth shaped. Spinner- ets and their associated spigot glands were as- signed after Platnick et al. (1991). See Table 1 for a complete list of abbreviations. FAMILY DYSDERIDAE Genus Dysdera Latreille 1804 Note: An excellent and thorough diagnosis and description of the genus Dysdera can be found in Deeleman-Reinhold & Deeleman (1988). Dysdera ambulotenta Ribera, Ferrandez & Blasco 1985 Figs. 13-24, Table 2 Dysdera ambulotenta Ribera, Ferrandez & Blasco 1985: 54-57, fig. 1 [3,$]. Holotype male and paratype female (allotype) from Cueva del Vien- to-Sobrado, El Amparo, Icod de los Vinos, Ten- erife, Canary Islands; 14 May 1981, J.L. Martin leg.; 6 (T-CS-17), 9 (T-CS-18). Stored at UL. Examined. Wunderlich 1991: 284-287. Diagnosis. — Dysdera ambulotenta can be distinguished from all other Canarian Dysdera species, except D. labradaensis, by its large size (carapace > 6.5 mm) and remarkable eye reduction. It differs from D. labradaensis (male unknown) by complete absence of both the posterior lateral (PLE) and posterior me- Table 2. — Intraspecific spination variability of Dysdera ambulotenta. Proximal Medial-proximal Medial-distal Distal Tibia 3 dorsal 1.0-2. 1 1. 0-2.0- 1 0 1.0.1 Tibia 4 dorsal 0-1. 0.0-1 1.0-3. 1 0-1.0-3.0-1 1.0.1 Tibia 3 ventral 1. 1-3.1 Ll-2.1 0 1. 0.0-1 Tibia 4 ventral 1. 1-2.1 Ll-2.1 0-2.0-2.1-2 0-1. 0-2.1 Number of rows Number of spines Femur 3 dorsal 2 0-3/0- 1 Femur 4 dorsal 2 0-4/0-4 ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 609 Figures 4-12. — Diagrams showing the characters included in the abbreviations list (Table 1). 4, Cara- pace frontal region, dorsal view; 5, Left chelicera, ventral view; 6, Right bulb, frontal view; 7, Right bulb distal tip, frontal view; 8, Bulb, external view; 9, Bulb, posterior view; 10, Vulva, ventral view; 11, Vulva, dorsal view; 12, Vulva, transverse section. dial eyes (PME), lack of spines on legs 1 and 2 and absence of an additional ventral diver- tide (AVD) in the vulva (Fig. 17). In both sexes absence of a polar spigot (PS) in the anterior lateral spinnerets (ALS) is unique to this species (Fig. 23). Description. — Holotype male: Figs. 13-15, 19-22. Carapace (Fig. 13) 7.35 mm long; maximum width 5.6 mm; minimum width 3.78 mm. Reddish-orange, frontally darker, becoming lighter towards back; slightly fo- veate at borders, slightly wrinkled with a fine- textured granular surface primarily at the an- terior end. Frontal border roughly straight, from 1/2 to ^5 carapace length; anterior lateral borders convergent (backwards long, parallel); rounded at maximum dorsal width point, back lateral borders rounded; back margin wide. straight PMF, PLF lost; AMF markedly re- duced; AMF diameter 0.09 mm; AMF sepa- ration 0.936 mm. Labium trapezoid- shaped, base wider than distal part; longer than wide at base; semicircular groove at tip. Sternum dark orange, frontally darker, becoming ligh- ter towards back; very slightly wrinkled, mainly between legs, frontal border; uniform- ly covered in slender black hairs. Chelicerae (Fig. 14) 4.41 mm long, about % of carapace length in dorsal view; fang me- dium-sized, 2.8 mm; basal segment dorsal side completely covered with piligerous gran- ulations (small, densely), ventral side smooth. Chelicera inner groove medium-size, about ^5 cheliceral length; armed with three teeth and lamina at base; D > B > M (large teeth; D, B not very different); D round, located rough- 610 THE’ JOURNAL OF ARACHNOLOGY Figures 13-18. — Dysdera ambulotenta. 13, Car- apace, dorsal; 14, Left chelicera, ventral; 15, Left male bulb, external; 16, Vulva, dorsal; 17 Vulva, ventral; 18, Vulva, lateral. Scale bars in mm. ly at center of groove; B close to basal lamina; M at middle of B and D. Legs orange. Lengths of male described above: fel 6.6 mm (all mea- surements in mm); pal 4.3; til 6.3; mel 5.4; tal 1.2; total 23.8; fe2 6.3; pa2 4; ti2 6.2; me2 5.2; ta2 1.2; total 22.9; fe3 5.2; pa3 2.9; ti3 3.8; me3 4.5; ta3 1.3; total 17.7; fe4 6.3; pa4 3.4; ti4 5.3; me4 6; ta4 1.4; total 22.4; relative length: 1 > 2 > 4 > 3; palp: fe 3.5; pa 2; ti 2; ta 1.9; total 9.4. Spination: palp, legl, leg2 spineless. Fe3 dorsal spines in two rows: an- terior 3 (distal); posterior 1 (proximal); ti3 dorsal spines arranged in three bands: proxi- mal 1.2.1; medial-proximal 1. 1-2.1; distal 1.0.1; ti3 ventral spines arranged in three bands: proximal 1 -0.3-2. 1; medial-proximal 1. 3-2.1; distal 1.0.1; with two terminal spines. Fe4 dorsal spines in two rows: anterior 4-2; posterior 3-2; ti4 dorsal spines arranged in four bands: proximal 1-0.0. 1-0; medial-prox- imal 1.2.1; medial-distal 1.3-2. 1; distal 1.0.1; ti4 ventral spines arranged in four bands: proximal 1.2- 1.1; medial-proximal 1.2.1; me- dial-distal 1 .2- 1 . 1 ; distal 1 . 1 -0. 1 ; with two ter- minal spines. Dorsal side of frontal legs with a fine-textured piligerous granular surface; ventral side of palp smooth; posterior legs densely covered with short hairs. Claws with more than 15 teeth, slender, length twice claw width. Abdomen 10.5 mm long; whitish; cy- lindrical. Abdominal dorsal hairs 0.027 mm long (short); medium-sized, roughly straight, not compressed, blunt, tip enlarged; uniform- ly, thickly distributed. Male bulb (Fig. 15): T as long as DD; ex- ternal distal border straight; internal sloped backwards. DD bent about 45° in lateral view; internal distal border not expanded. ES wider, more sclerotized than IS; IS continuous to tip. DD tip (Figs. 19-=21) straight in lateral view. C present, long; distal end on DD internal tip; well-developed; located far from DD distal tip; proximal border continuously decreasing; distal border markedly sloped, upper tip not projected, pointed; external side hollowed. AC absent. LF present; distally not projected; poorly developed. L well-developed; external border not sclerotized, laterally slightly fold- ed, distal border divergent, continuous. AL present, well-developed; proximal border in posterior view fused with DH. P (Fig. 22) fused to T; markedly sloped on its proximal part, perpendicular on distal; lateral length as long as or longer than T width; ridge present, perpendicular to T, not expanded; upper mar- gin smooth; not distally projected; back mar- gin not folded. Par atype female: Figs. 16”18, 23, 24. All characters as in male except: Carapace 7.21 mm long; maximum width 5.6 mm; minimum width 3.64 mm. AME separation 1.16 mm. Chelicerae 4 mm long; fang 3.22 mm. Legs orange. Lengths of female described above: fel 6.6 mm (all measurements in mm); pal 4.3; til 6.2; mel 5.4; tal 1.3; total 23.8; fe2 6.1; pa2 4.1; ti2 6.1; me2 5.4; ta2 1.3; total 23; fe3 5.2; pa3 3; ti3 3.9; me3 4.8; ta3 1.2; total 18.1; fe4 6.5; pa4 3.5; ti4 5.2; me4 6.3; ta4 1.5; total 23; relative length 1 > 2 = 4> 3; palp: fe 4; pa 2; ti 1.9; ta 2.5; total 10.4. Spination: palp, legl, leg2 spineless. Fe3 dor- sal spines in two rows: anterior 2; posterior 1; ti3 dorsal spines arranged in three bands: proximal 1.0.1; medial-proximal LO.l; distal 1.0.1; ti3 ventral spines arranged in three bands: proximal 2- 1.2,2- 1; medial-proximal 1.3- 1.0-1; distal LO.l; with two ternunal spines. Fe4 dorsal spines in two rows: anterior 1; posterior 2; ti4 dorsal spines arranged in four bands: proximal 1-0.0. 1-0; medial-prox- imal 1.2-0, 1; medial-distal 1.3- 1.1; distal ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 611 Figures 19-24. — Dysdera ambulotenta, right male bulb and female spinnerets. 19, DD frontal; 20, DD external; 21, DD internal; 22, P internal view; 23, Right ALS; 24, Right PLS. 1.0.0; ti4 ventral spines arranged in four bands: proximal 1.1.1; medial-proximal 1.1- 2.1; medial-distal 0-1. 1-0.1; distal 0-1. 1-0.1; with two terminal spines. Abdomen 7 mm long. Abdominal dorsal hairs 0.036 mm long (short); medium thick- ness, roughly straight, not compressed, blunt, tip enlarged; uniformly, thickly distributed. Vulva (Figs. 16-18) rectangle-like in dorsal view, frontally rounded; slightly wider than long; DF wide. MF well-developed; markedly sclerotized along its extent. VA frontal region 612 THE JOURNAL OF ARACHNOLOGY completely sclerotized; posterior region scler- otized except for internal area. AVD absent. S attachment projected under VA; arms as long as DA, slightly curved; tips dorsally project- ed; neck as wide as arms. TB usual shape. ALS (Figs. 23-24) without PS; remaining pir- iform spigots no more external than MS, ar- ranged in three rows; 18 piriform gland spig- ots; PMS, PLS with 5-10 aciniform gland spigots. Intraspecific variation. — Male cephalo- thorax ranges in length from 7.00-7.35 mm, female from 6.51-7.00 mm. Sometimes cara- pace lateral margin angled at maximum witdth point. AME reduction variable, from tiny bright spots to absent. Chelicera relative length from 0.43-0.48. D as large as or slight- ly larger than B. One male from Los Roques with M distinctly closer to D. In general, che- liceral teeth are large. Spination variability in Table 2. Additional material examined. — TENERIFE: El Sauzal: Cueva de Labrada-Mechas, 13 March 1982, 1(3 subadult (J.L. Martin, num. 2520 UL). Icod de los Vinos: Cueva de Felipe Reventon, 17 March 1984, some remains (J.J. Hernandez, num. 2534 UL). Cueva del Viento-Sobrado, 10 Deeember 1982, ljuv. (J.L. Martin, num. 2517 UL); 2 Novem- ber 1991, ljuv. (J.L. Martin, num. 2518 UL). La Orotava: Cueva del Bucio, 27 November 1984, some remains (J.L. Martin & A. Machado, num. 2794 UL); 4 March 1985, ljuv. (J.L. Martin & A. Machado, num. 2532 UL). 1 April 1991, 1 9 (Lucas & Rando, num. 2511 UL). Cueva de los Roques, 11 August 1986, 13 (J.L. Martin, num. 2512 UL); 27 October 1991, ? (one chelicera) (C. Ribera, num. 2568 UL); 25 September 1996, 1 9 (P. Oromi, num. 3184 UB). Distribution. — Tenerifean endemic. Exclu- sively known from lava tubes. It is the most widespread of troglomorphic Dysdera. Dysdera brevisetae Wunderlich 1991 Figs. 25-36 Dysdera brevisetae Wunderlich 1991: 289-290, fig. 14-16 [3]. Holotype male from Monte de las Mercedes, La Laguna, Tenerife, Canary Islands; in II, M. Knosel leg.; num. 37166. Stored at SMF. Examined. -Wunderlich 1991: 284-287. Diagnosis. — Dysdera brevisetae is distin- guished from any other markedly foveate spe- cies by its wider carapace frontal border and spineless femora and tibiae. Males differ from the morphologically similar D. macra by a poorly developed but complete bulb lateral sheet (L) (Fig. 31), and both males and fe- males have anterior medial eyes separated by less than Ys of its diameter from each other, longer chelicera inner groove and cheliceral distal tooth (D) markedly larger than basal one (B) (Fig. 26). Description.— Holotype male: Figs. 25-27, 31-34. Carapace (Fig. 25) 3.62 mm long; maximum width 2.51 nun; minimum width 1.84 mm. Dark red, darkened at borders; heavily wrinkled, foveate, with a fine-textured granular surface. Frontal border roughly round, from ¥2-% carapace length; anterior lat- eral borders parallel or slightly divergent; rounded at maximum dorsal width point, back lateral borders rounded; back margin narrow, straight; slightly stepped in lateral view. AME diameter 0.21 mm; PLE 0.21 mm; PME 0.16 mm; AME on edge of frontal border, separat- ed one from another about 1 of diameter, close to PLE; PME very close to each other, less than Va PME diameter from PLE. Labium trapezoid- shaped, base wider than distal part; as long as wide at base (triangle-like); semi- circular groove at tip. Sternum brownish-red, darkened on borders; heavily wrinkled; uni- formly covered in slender black hairs. Chelicerae (Fig. 26) 1.58 mm long, about % of carapace length in dorsal view; fang long, 1.35 mm; basal segment dorsal, ventral side completely covered with piligerous granula- tions. Chelicera inner groove long, about V2 cheliceral length; armed with three teeth and lamina at base; D > B = M (B, M small); D trapezoid, located near segment tip; B close to basal lamina; M close to D. Legs orange. Lengths of male described above: fel 2.4 mm (all measurements in nun); pal 1.44; til 2.14; mel 2; tal 0.51; total 8.49; fe2 2.23; pa2 1.35; ti2 1.91; me2 1.98; ta2 0.51; total 7.98; fe3 1.79; pa3 1.07; ti3 1.26; me3 1.68; ta3 0.51; total 6.31; fe4 2.19; pa4 1.12; ti4 1.82; me4 2; ta4 0.53; total 7.66; relative length: 1 > 2 > 4 > 3; palp: fe 1.4; pa 0.74; ti 0.74; ta 0.74; total 3.62. Spination: spineless. Dorsal side of frontal legs covered with hairs, lacking a gran- ular surface; ventral side of palp smooth; long, spine-like hairs on posterior ti, fe. Claws with 10-14 teeth, length twice claw width. Abdomen 3.73 nun long; whitish; cylindri- cal. Abdominal dorsal hairs 0.045 mm long; medium thickness, roughly straight, not com- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 613 Eigures 25-30. — Dysdera brevisetae. 25, Cara- pace, dorsal; 26, Left chelicera, ventral; 27, Right male bulb, external; 28, Vulva, dorsal; 29, Vulva, ventral; 30, Vulva, lateral. Scale bars in mm. pressed, blunt, tip enlarged; uniformly, thickly distributed. Male bulb (Fig. 27): T slightly smaller than DD; external, internal distal border sloped backwards. DD slightly bent in lateral view, clearly less than 45°; internal distal border not expanded. IS, ES equally developed; IS trun- cated at DD middle part; ES bend markedly sclerotized. DD tip (Figs. 31-33) straight in lateral view. C present, short; distal end on DD internal tip; well-developed; located close to DD distal tip; proximal border sharply de- creasing; distal border rounded, hardly stepped, upper tip not projected, rounded; ex- ternal side hollowed. AC present. LF absent. L poorly developed; external border not scler- otized, laterally slightly folded; distal border divergent, not continuous; upper sheet strong- ly folded at middle. AL present, very poorly developed; proximal border in posterior view fused with DH. P (Fig. 34) fused to T; per- pendicular to T in lateral view; lateral length from V3-% of T width; ridge present, perpen- dicular to T, not expanded; upper margin smooth; not distally projected; back margin slightly folded towards internal side. Female: (from El Moquinal, La Orotava, Tenerife; num. 2935, UB) Figs. 28-30, 32, 33. All characters as in male except: Carapace 3.66 mm long; maximum width 2.65 mm; minimum width 1 .98 mm. AME diameter 0.23 nun; PLE 0.2 mm; PME 0.16 mm. Sternum dark red. Chelicerae 1.72 nun long; fang 1.4 mm; basal segment proximal dorsal, ventral side scantly covered with piligerous granula- tions. Legs orange. Lengths of female de- scribed above: fel 2.42 mm (all measurements in mm); pal 1.54; til 2.1; mel 2.05; tal 0.53; total 8.64; fe2 2.25; pa2 1.44; ti2 1.91; me2 2; ta2 0.53; total 8.13; fe3 1.91; pa3 1.26; ti3 1.38; me3 1.68; ta3 0.51; total 6.56; fe4 2.28; pa4 1.26; ti4 1.91; me4 2.1; ta4 0.53; total 8.08; relative length 1 > 2 > 4 > 3; palp: fe 1.4; pa 0.79; ti 0.56; ta 0.88; total 3.63. Abdomen 4.66 nun long; whitish; cylindri- cal. Abdominal dorsal hairs 0.16-0.18 mm long; thin, curved, compressed, pointed; uni- formly, thickly distributed. Vulva (Figs. 28- 30) arch-like in dorsal view, frontally round- ed; slightly wider than long; DF wide. MF poorly developed. VA frontal region com- pletely sclerotized; posterior region sclero- tized at anterior area. AVD hardly visible. S attachment not projected under VA; arms as long as DA, slightly curved; tips dorsally pro- jected; neck as wide as arms. TB usual shape. ALS (Figs. 32-33) with PS; remaining piri- form spigots more external than MS, arranged in two rows; 7+1 piriform gland spigots; PMS, PLS with 5-10 aciniform gland spigots. Intraspecific variation. — Male cephalo- thorax ranges in length from 3.62-3.54 mm, female from 3.40-3.66 mm. PLE-PME from V^-% diameter. Sternum ornamentation some- what reduced. B may be larger than M. M sometimes closer to B. Additional material examined. — TENERIFE: La Laguna: Cocomoto, ? February 1989, Id (C. Deniz, num. 2680 UL). El Moquinal, under a bark of Erica scoparia, 18 October 1994, 19 (P. Oromi, num. 4001 UB). Monte de las Mercedes, 30 Janu- ary 1989, Id (H. Enghoff, num, 2640 ZMK). Los Silos: Monte del Agua, 14 March 1987, ljuv. (H. Enghoff, num. 2669 ZMK); 1 February 1988, 19 (JJ. Naranjo, num. 2598 UB); 3 March 1989, 1 9 614 THE JOURNAL OF ARACHNOLOGY Figures 31-36. — Dysdera brevisetae, right male bulb and female spinnerets. 31, DD frontal; 32, DD external; 33, DD posterior; 34, P external; 35, Right ALS; 36, Right PLS. (H. Enghoff & M. Baez, num. 2646 ZMK); 3 March 1989, 1 ? (H. Enghoff, num. 2659 ZMK); 18 February 1996, ljuv. (Amedo & Oromi, num. 3118 UB). Santa Cruz de Tenerife: Bailadero, ? November 1993, 1 9 (Arnedo & Ribera, num. 4784 (T21) UB). Cabezo de Tejo, 26 February 1996, 1 $ (Oromi & Emerson, num. 3128 UB). Casas de la Cumbre, 23 February 1996, ljuv. (Oromi & Em- erson, num. 3127 UB). Cruz del Carmen, 12 May 1996, 16 (M. Naranjo, num. 3145 UB). Vueltas de Taganana, 20 February 1984, 16 (Garcia Alayon, num. 2687 UL); May 1995, 16 (P. Oromi, num. ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 615 Table 3. — Intraspecific spination variability of Dysdera brevispina. Proximal Medial- proximal Medial- distal Distal Tibia 3 dorsal 0-1. 0-2.1 0 0 1.0.0 Tibia 4 dorsal Ll-2.1 0 0 1.0-Ll Tibia 3 ventral 0-l.l-.2.0~l 0 0 1.0.0 Tibia 4 ventral 12-.3.0-1 0 0 1.0.0- 1 Number of rows Number of spines Femur 3 dorsal 0 0 Femur 4 dorsal 1 0-1 4181 UB). 29 November 1996, 1 ? (R Oromi, num. 3189 UB). Tegueste: Pedro Alvarez, 19 February 1997, 1(3 (B. Emerson, num. 3204 UB; 20 Febru- ary 1997, 1(3, (P. Oromi, num. 3205 UB). Distribution.— Tenerifean endemic. Abun- dant species known from several localities re- stricted to Anaga and Teno massifs. Comments.— Originally known from a sin- gle male specimen, this species has been ex- tensively collected in Tenerife. Dysdera brevispina Wunderlich 1991 Figs. 37-48, Table 3 Dysdera brevispina Wunderlich 1991: 289-290, figs. 17-19 [c3]. Holotype male from Cueva de Felipe Reventon, Icod de los Vinos, Tenerife, Ca- nary Islands; 23 March 1984, A. Machado leg.; num. T-FR-97. Stored at UL. Examined. -Wun- derlich 1991: 284-287. D. moquinalensis Wunderlich 1991: 301, figs. 65- 68 [(3]. El Moquinal, La Laguna, Tenerife, Ca- nary Islands; lc3; 20/4/90, P. Oromi leg.; Stored at UL. Examined. New synonymy. D. vilaflorensis Wunderlich 1991: 310-311, figs. 124-125 [c3]. MSS-6, Barranco del Chorrillo, Vi- laflor, Tenerife, Canary Islands; lc3; 15 May 1990, A.L. Medina leg.; Stored at UL. Examined. New synonymy. Diagnosis.— brevispina is distin- guished from most of the Canarian Dysdera species by its smooth carapace and by the che- liceral basal tooth (B) being the largest (Fig. 38). Males and females differ from the eastern Canary Islands endemics by lack of thick and lanceolate abdominal hairs, from D. chioensis and D. unguimmanis by absence of eye re- duction and from from D. guayota new spe- cies by absence of spines on frontal legs. Males distinguished from D. insulana by lack- ing medial fold in bulb lateral sheet (L) (Fig. 43), and both males and females by presence of cheliceral granulation and distinctive spi- nation pattern (Table 3). Description.— 7/o/ofyp^ male: Figs. 37-39, 43-47. Carapace (Fig. 37) 3.63 mm long; maximum width 2.7 mm; minimum width 1.68 mm. Brownish-orange, darkened at bor- ders; slightly foveate at borders, slightly wrin- kled and a black fine-textured granular surface mostly at the anterior end. Frontal border roughly round, from V2-% carapace length; an- terior lateral borders slightly convergent; rounded at maximum dorsal width point, back lateral borders straight; back margin wide, straight. AME diameter 0.16 mm; PLE 0.14 nun; PME 0.12 nun; AME on edge of frontal border, separated one from another about 1 di- ameter or more, close to PLE; PME very close to each other, about Vi PME diameter from PLE. Labium trapezoid-shaped, base wider than distal part; longer than wide at base; semicircular groove at tip. Sternum brownish- orange, darkened on borders; very slightly wrinkled, mainly between legs, frontal border; uniformly covered in slender black hairs. Chelicerae (Fig. 38) 1.56 nun long, about % of carapace length in dorsal view; fang long, 1.35 mm; basal segment proximal dorsal side scantly covered with piligerous granulations. Chelicera inner groove medium-size, about % cheliceral length; armed with three teeth, lam- ina at base; additional tooth on left chelicera; B > D = M (D, M markedly small); D tri- angular, located roughly at center of groove; B close to basal lamina; M close to B. Legs pale yellow. Lengths of male described above; fel 3.45 nun (all measurements in mm); pal 2.19; til 3.12; mel 2.93; tal 0.6; total 12.29; fe2 3.03; pa2 1.96; ti2 2.79; me2 2.7; ta2 0.6; total 11.08; fe3 3.12; pa3 1.4; ti3 1.72; me3 2.23; ta3 0.56; total 9.03; fe4 3.68; pa4 1.86; 616 THE JOURNAL OF ARACHNOLOGY Figures 37-42. — Dysdera brevispina. 37, Cara- pace, dorsal; 38, Left chelicera, ventral; 39, Right male bulb, external; 40, Vulva, dorsal; 41 Vulva, ventral; 42, Vulva, lateral. Scale bars in mm. ti4 2.65; me4 3.26; ta4 0.65; total 12.1; rela- tive length: 1 > 4 > 2 > 3; palp: fe 1.72; pa 0.84; ti 0.93; ta 0.93; total 4.42. Spination: palp, legl, leg2 spineless. Fe3 dorsal spine- less; ti3 dorsal spines arranged in two bands: proximal 1.1 -0.1; distal 1.0.0; ti3 ventral spines arranged in two bands: proximal 2-1.3- 0.2-1; distal 1.0- 1.1; with two terminal spines. Fe4 dorsal spineless; ti4 dorsal spines ar- ranged in two bands: proximal 1.1.0; distal 1.0.0; ti4 ventral spines arranged in two bands: proximal 1. 3-2.0; distal 1.0.0; with two terminal spines. Dorsal side of frontal legs smooth; ventral side of palp covered with hairs, lacking small a granular surface. Claws with 10-14 teeth; hardly larger than claw width. Abdomen 4.52 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.018-0.027 mm long (smalls); medium thickness, roughly straight, not compressed, blunt, tip not en- larged; uniformly, scantly distributed. Male bulb (Fig. 39): T slightly smaller than DD; external, internal distal border sloped backwards. DD slightly bent in lateral view, clearly less than 45°; internal distal border not expanded. ES more sclerotized than IS; IS truncated at DD middle part. DD tip (Figs. 43—45) straight in lateral view. C present, short; distal end on DD internal tip; well-de- veloped; located close to DD distal tip; prox- imal border sharply decreasing; distal border stepped, upper tip not projected, pointed; ex- ternal side hollowed. AC present. LF absent. L well-developed; external border not sclero- tized, laterally slightly folded; distal border di- vergent, continuous. AL present, very poorly developed; proximal border in posterior view toothed. P (Fig. 46) fused to T; perpendicular to T in lateral view; lateral length from of T width; ridge present, perpendicular to T, not expanded; upper margin slightly toothed, mainly on external side, on its distal part, very few teeth; not distally projected; back margin not folded. Female: (from Cueva Felipe Reventon, Icod de los Vinos, Tenerife; num. 2744, UL). Figs. 40-42, 47, 48. All characters as in male except: Carapace 2.98 mm long; maximum width 2.37 mm; minimum width 1.44 mm. Brownish-orange, uniformly distributed. AME diameter 0.12 mm; PLE 0.11 mm; PME 0.09 mm; PME about ^5 PME diameter from PLE. Chelicerae 1.3 mm long; fang 1.21 mm Leg lengths of female described above: fel 3.45 mm (all measurements in mm); pal 1.72; til 2.33; mel 2.23; tal 0.51; total 10.24; fe2 2.42; pa2 1.68; ti2 2.19; me2 2.1; ta2 0.56; total 8.95; fe3 1.96; pa3 1.12; ti3 1.44; me3 1.86; ta3 0.51; total 6.89; fe4 2.79; pa4 1.49; ti4 2.33; me4 2.7; ta4 0.6; total 9.91; relative length 1 > 4 > 2 > 3; palp: fe 1.4; pa 0.6; ti 0.6; ta 0.93; total 3.53. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0- 1.1; distal 1.0.0; ti3 ventral spines arranged in two bands: proximal 2- 1. 3-0.2- 1; distal 1.1- 0,1; with two terminal spines. Fe4 dorsal spineless; ti4 dorsal spines arranged in two bands: proximal 0.1.0; distal 1.0.0; ti4 ventral spines arranged in two bands: proximal 1.3- 2.0; distal 1.0.0; with two terminal spines. Abdomen 4.19 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.072-0.09 mm long; medium thickness, curved, compressed, blunt, tip not enlarged; uniformly, thickly dis- tributed. Vulva (Figs. 40-42) arch-like in dor- sal view, frontally pointed; as wide as long; DF wide. MF poorly developed. VA frontal region completely sclerotized; posterior re- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 617 Figures 43-48. — Dysdera brevispina, right male bulb and female spinnerets. 43, DD frontal; 44, DD external; 45, DD posterior; 46, P external; 47, Right ALS; 48, Right PLS. gion sclerotized at anterior area; small scale on back border internal part. AVD hardly vis- ible. S attachment not projected under VA; arms as long as DA, straight; tips dorsally pro- jected; neck as wide as arms. TB usual shape. ALS (Figs. 47, 48) with PS; remaining piri- form spigots more external than MS, arranged in three rows; 9+1 piriform gland spigots; PMS, PLS with 10-15 aciniform gland spig- ots. Intraspecific variation. — Male cephalo- thorax ranges in length from 2.37-3.63 mm. 618 THE JOURNAL OF ARACHNOLOGY Table 4. — Intraspecific spination variability of Dysdera chioensis. Proximal Medial-proximal Medial-distal Distal Tibia palp dorsal 0 1.0.0 0 0 Tibia 3 dorsal O-LO-2.0-1 0-1. 0-2.0- 1 0.0- 1.0 1.0-1. 1 Tibia 4 dorsal 0.0-2.0-1 1.0-3. 1 0 1.0-2. 1 Tibia 1 ventral 0.0- 1.0 0.0-2.0 0 0 Tibia 2 ventral 0 0.1-2.0 0 0 Tibia 3 ventral 0.0-2.0-1 1.2.0-1 0 1.0- 1.0-1 Tibia 4 ventral 0-1. 2.0-1 O-Ll-2.1 0-1. 0-2.0 O-l.O-l.l Number of spines Patella palp dorsal 1-2 Patella 3 ventral 0-3 Patella 4 ventral 0-2 Number of rows Number of spines Femur 1 frontal 1 1-2 Femur 2 frontal 1 2-3 Femur 3 dorsal 2 2-5/1-2 Femur 4 dorsal 2 1 -2/5-9 female from 2.84”3.47 mm. Carapace frontal lateral borders parallel. Specimens from caves may show certain reduction in eye size. PLE- PME from Vy-% diameter, AME separation from ^5-1 diameter apart. Occasionally, D as large as B. In general, cheliceral teeth are small; S arms slightly curved. Spination var- iability in Table 3. Additional material examined. — TENERIFE i Icod de los Vinos: Cueva Felipe Reventon, ? May 1994, 1 9 (M. Arechavaleta, num. 2798 UL); ? May 1994, 1 9 (M. Arechavaleta, num. 2806 UL). Cue- va del Viento-Sobrado, 14 April 1983, 1 9 (J.L. Martin, num. 2521 UL). Santa Cruz de Tenerife: El Bailadero, 27 November 1993, 19 (M.A. Amedo & C. Ribera, num. 2588 UB). Santa Ursula: Bco. del Pino, (8411 T/C -T), 21-28 July 1985, 1 9 (J.M. Peraza, num. 2616 MCNT). Distribution,- — ^Tenerifean endemic. Known from several localities spread through the is- land’s northern slopes (including Anaga). One single locality on south-western slopes (Vilaf- lor), collected in MSS trap. Unknown from Teno massif. Comments. — After examination of the D. moquinalensis holotype, no distinctive mor- phological difference from D. brevispina, apart from the highly polymorphic carapace color, could be found. The only difference of D. vilafiorensis holotype from D. brevispina is an overall smaller size. Dysdera chioensis Wunderlich 1991 Figs. 49^53, 73, 74, Table 4 Dysdera chioensis Wunderlich 1991: 291, figs. 21- 23 [9]. Holotype female from Cueva Grande del Chio, Guia de Isora, Tenerife, Canary Islands; 29 June 1985, G.LE.T leg.; num. T-GC-5; Stored at UL. Examined. -Wunderlich 1991: 284-287. Diagnosis.— chioensis can be dis- tinguished from all other Dysdera species, ex- cept D. labradaensis, by distinct reduction of eye size and presence of spines on frontal legs. Its smaller size and spinated pedipalps distinguish this species from D. labradaensis. It differs from the morphologically similar species D. guayota new species by having re- markably reduced eyes and presence of spines on palps, and in females by presence of tooth- like expansions in vulva ventral arch (VA) (Figs. 52, 53). 'Description.— Holotype female: Figs. 49- 53, 73-74. Carapace (Fig. 49) 3.73 mm long; maximum width 3.03 mm; minimum width 2 mm. Reddish-orange, frontally darker, becom- ing lighter towards back; smooth with some black granular material mainly at front; hairy, covered with black hairs mainly at lateral and back borders. Frontal border roughly round, from V2~% carapace length; anterior lateral borders slightly convergent; sharpened at maximum dorsal width point, back lateral bor- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 619 Figures 49-53. — Dysdera chioensis. 49, Cara- pace, dorsal; 50, Left chelicera, ventral; 51, Vulva, dorsal; 52 Vulva, ventral; 53, Vulva, lateral. Scale bars in nun. ders rounded; back margin wide, slightly bi- lobed. Eyes markedly reduced in size. AME diameter 0.09 mm; PLE 0.05 mm; PME 0.03 mm; AME separation 1.02 mm; AME-PLE separation 0.05 mm; PLE-PME separation 0.14 mm; PME separation 0.16 mm. Labium trapezoid-shaped, base wider than distal part; longer than wide at base; semicircular groove at tip. Sternum yellowish-orange, darkened on borders; smooth; uniformly covered in slender black hairs. Chelicerae (Fig. 50) 1.77 mm long, about V3 of carapace length in dorsal view; fang me- dium-sized, 1.16 mm; basal segment dorsal side completely covered with piligerous gran- ulations (distally scarce), ventral side smooth. Chelicera inner groove short, about Vs chelic- eral length; armed with three teeth and lamina at base; B > D — M; D round, located roughly at center of groove; B close to basal lamina; M close to B. Legs yellow. Lengths of female described above: fel 2.7 mm (all measure- ments in mm); pal 1.68; til 2.14; mel 1.91; tal 0.6; total 9.03; fe2 2.37; pa2 1.63; ti2 2.07; me2 1.96; ta2 0.6; total 8.63; fe3 2.16; pa3 1.21; ti3 1.68; me3 2.1; ta3 0.65; total 7.8; fe4 2.56; pa4 1.4; ti4 2.1; me4 2.51; ta4 0.65; total 9.22; relative length 4 > 1 > 2 > 3; palp: fe 1.9; pa 0.74; ti 0.65; ta 0.93; total 4.22. Spi- nation: Palp pa 1, palp ti 1 medial internal. Fel 1 distal, anterior margin. Fe2: 3-2 distal, anterior margin. Fe3 dorsal spines in two rows: anterior 5; posterior 1 (distal); pa3 spineless; ti3 dorsal spines arranged in two bands: proximal 1.2.1; distal 1.0.1.; ti3 ventral spines arranged in two bands: medial-proxi- mal 1.2.1; distal 1. 0.0-1; with two terminal spines. Fe4 dorsal spines in two rows: anterior 1; posterior 7-5; pa4 1 ventral medial; ti4 dor- sal spines arranged in three bands: proximal 0.0.0- 1; medial-proximal 1.1.1; distal 1.0.1; ti4 ventral spines arranged in three bands: proximal 0.2.0; medial-proximal 1.1.1; medi- al-distal 0; distal 1.1.1; with two terminal spines. Dorsal side of frontal legs, ventral side of palps covered with hairs, lacking a granular surface. Claws with 10-14 teeth; hardly larger than claw width. Abdomen 5.12 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.216-0.234 mm long; thin, curved, compressed, pointed; uni- formly, thickly distributed. Vulva (Figs. 51- 53) arch-like in dorsal view, frontally pointed; slightly wider than long; DF wide. ME well-developed. VA frontal region completely sclerotized; posterior region sclerotized at an- terior area; tooth-shaped expansion from in- ternal back border; not joined to lateral scler- otization, slightly shorter than DA lateral margins. AVD clearly recognizable. S attach- ment projected under VA; arms are shorter than DA, straight; tips not projected; neck as wide as arms. TB usual shape. ALS (Figs. 73, 74) with PS; remaining piriform spigots more external than MS, arranged in two rows; 9 + 1 piriform gland spigots; PMS, PLS with 5- 10 aciniform gland spigots. Male: Unknown. Intraspecific variation. — Female cephalo- thorax ranges in length from 3.36-5.6 mm. D at center of chelicera groove or at tip. Speci- mens from Los Roques, S arms longer. Spi- nation variability in Table 4. Additional material examined. — TENERIFE: Gma de Isora: Cueva Grande del CMo, 28 January 1993, ljuv. (P. Oromi, num. 2545 UL); 21 October 1994, 1 ? (Arnedo & Ribera, num. 4821 (T32) UB). 620 THE JOURNAL OF ARACHNOLOGY Table 5. — Intraspecific spination variability of Dysdera cribellata. Proximal Medial- proximal Medial- distal Distal Tibia 3 dorsal 1.0.1 0-1. 0.0 0 1.0.0 Tibia 4 dorsal 1.0.1 0 0 0-1. 0.0-1 Tibia 3 ventral 0-1. 0-1.0 0 0 0-1. 0.0 Tibia 4 ventral 0-1. 1.0 0-1. 0-1.0 0 0-1. 0.0 Number of rows Number of spines Femur 3 dorsal 0 0 Femur 4 dorsal 0 0 La Orotava: Cueva de Los Roques, 11 August 1986, ljuv. (J.L. Martin, num. 2536 UL); Novem- ber 1995, 19 (R Oromf, num. 2965 UB); 27 Oc- tober 1991, 19 (C. Ribera, num. 2566 UB); No- vember 1995, 1 9 (P. Oromf, num. 2967 UB). Distribution. — Tenerifean endemic. Known from two lava tubes located on dry, south- western slopes. Comments. — Even though several new fe- male specimens have been collected and a new locality has been found for this species, the male remains unknown. Dysdera cribellata Simon 1883 Figs. 54-65, Table 5 Dysdera cribellata Simon 1883 (nec. Simon 1907: 258-259, fig. 257 [c?]; incorrect identification): 294-295, fig. 17 [d]. Type male lost. Type fe- male from Canary Islands, unknown locality; un- known data, M. Verneau leg.; num. B-536; Stored at MNHN. Examined. -Bdsenberg 1895: 7. -Rei- moser 1919. -Denis 1941: 108. -Schmidt 1973: 360-361. -Amedo et al. 1996: 243. D. medinae Wunderlich 1991: 299, figs. 57-60 [(3,9]. Holotype 6 and paratype 9 from Monte de las Mercedes, La Laguna, Tenerife, Canary Is- lands; in II, M. Knosel leg.; Stored at SML Ho- lotype not examined, paratypes examined. New synonymy. D. volcania Ribera, Ferrandez & Blasco 1985: 59- 61, figs. 3E-F [9] (9, non (3). Paratype female (allotype) from Cueva de Felipe Reventon, Icod de los Vinos, Tenerife, Canary Islands; 3 March 1984, P. Oromf leg.; num. T-FR-107. Stored at UL. Examined. Incorrect identification. Diagnosis. — Dysdera cribellata can be dis- tinguished from all other Canarian species by its markedly foveate carapace, cheliceral basal tooth (B) being the largest (Fig. 55) and a lack of cheliceral granulation. It can be distin- guished from the similar D. insulana by its foveate carapace and distinctive spination pat- tern (Table 5). Description. — Male: (from Sima de la Ro- bada, Santa Cruz de Tenerife, Tenerife; num. 2552, UL). Figs. 54-56, 60-63. Carapace (Fig. 54) 3.22 mm long; maximum width 2.75 mm; minimum width 1.77 mm. Dark red, darkened at borders; heavily foveate, covered with circular depressions, some black granular material mainly at front. Frontal border rough- ly round, from \V2-% carapace length; anterior lateral borders slightly divergent; rounded at maximum dorsal width point, back lateral bor- ders rounded; back margin wide, straight. AME diameter 0.23 mm; PLE 0.21 mm; PME 0.16 mm; AME on edge of frontal border, sep- arated one from another about of diameter, close to PLE; PME very close to each other, about V3 PME diameter from PLE. Labium trapezoid-shaped, base wider than distal part; longer than wide at base; semicircular groove at tip. Sternum dark red, darkened on borders; heavily wrinkled; covered in hairs mainly on margin. Chelicerae (Fig. 55) 1.35 mm long, about V3 of carapace length in dorsal view; fang me- dium-sized, 1.16 mm; basal segment smooth, with no granulations. Chelicera inner groove long, about Vi cheliceral length; armed with three teeth and lamina at base; B > D > M (not very different, small); D trapezoid, locat- ed roughly at center of groove; B close to bas- al lamina; M close to B. Legs orange. Lengths of male described above: fel 2.79 mm (all measurements in mm); pal 1.86; til 2.61; mel 2.51; tal 0.56; total 10.33; fe2 2.51; pa2 1.63; ti2 2.33; me2 2.28; ta2 0.56; total 9.75; fe3 2; pa3 1.12; ti3 1.49; me3 1.86; ta3 0.51; total 6.98; fe4 2.75; pa4 1.49; ti4 2.33; me4 2.7; ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 621 56 Figures 54-59. — Dysdera cribellata. 54, Cara- pace, dorsal; 55, Right chelicera, ventral; 56, Left male bulb, external; 57, Vulva, dorsal; 58, Vulva, ventral; 59, Vulva, lateral. Scale bars in mm. ta4 0.6; total 9.87; relative length: 1 > 4 > 2 > 3; palp: fe 1.4; pa 0.7; ti 0.7. Spination: palp, legl, leg2 spineless. Fe3 dorsal spine- less; ti3 dorsal spines arranged in three bands: proximal 1.0.1; medial-proximal 0-1. 0.0; dis- tal 1.0.0; ti3 ventral spines arranged in two bands: proximal 1.1.1; distal 1.0.0; with two terminal spines. Fe4 dorsal spineless; ti4 dor- sal spines arranged in two bands; proximal 1.0.1; distal 0.0.1; ti4 ventral spines arranged in two bands: proximal 1.1.0; distal 1.0.0; with two terminal spines. Dorsal side of fron- tal legs smooth; ventral side of palp covered with hairs, lacking a granular surface. Claws with 10-14 teeth, length twice claw width. Abdomen 3.68 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.045-0.054 nun long; thin, curved, compressed, pointed; uni- formly, thickly distributed. Male bulb (Fig. 56): T slightly smaller than DD; external, internal distal border sloped backwards. DD slightly bent in lateral view, clearly less than 45°; internal distal border not expanded. ES more sclerotized than IS; IS truncated at DD middle part. DD tip (Figs. 60-62) straight in lateral view. C present, short; distal end on DD internal tip; well-de- veloped; located close to DD distal tip; prox- imal border sharply decreasing; distal border stepped, upper tip not projected, sloped; ex- ternal side hollowed. AC present. LF absent. L well-developed; external border not sclero- tized, laterally slightly folded; distal border di- vergent, not continuous; upper sheet strongly folded at middle. AL present, very poorly de- veloped; proximal border in posterior view toothed. P (Fig. 63) fused to T; perpendicular to T in lateral view; lateral length from %-V2 of T width; ridge present, perpendicular to T, not expanded; upper margin slightly toothed, mainly on external side, on its distal part, few teeth; not distally projected; back margin slightly folded towards internal side. Lectotype female: Figs. 57-59, 64, 65. Car- apace 3.82 mm long; maximum width 3.08 mm; minimum width 1.82 mm. Brownish-or- ange. Anterior lateral borders parallel; round- ed at maximum dorsal width point, back lat- eral borders straight. AME diameter 0.23 mm; PLE 0.21 mm; PME 0.16 mm; AME separated one from another about % of diam- eter; PME about % diameter from PLE. Ster- num orange, uniformly distributed; wrinkled. Chelicerae 1.67 mm long; fang 1.3 mm. Legs yellow. Lengths of female described above: fel 2.98 mm (all measurements in mm); pal 2.05; til 2.52; mel 2.47; tal 0.56; total 10.58; fe2 2.66; pa2 1.86; ti2 2.33; me2 2.28; ta2 0.56; total 9.69; fe3 2.28; pa3 1.3; ti3 1.54; me3 1.91; ta3 0.56; total 7.59; fe4 3.03; pa4 1.63; ti4 2.33; me4 2.84; ta4 0.65; total 10.48; relative length 1 > 4 > 2 > 3; palp: fe 1.49; pa 0.83; ti 0.74; ta 0.93; total 3.99. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines ar- ranged in two bands: proximal 1.0.1; distal 1.0. 0; ti3 ventral spines arranged in two bands: proximal 0.1.0; distal 1.0.0; with two terminal spines. Fe4 dorsal spineless; ti4 dor- sal spines arranged in one band: proximal 1.0. 1; ti4 ventral spines arranged in two bands: proximal 1.1.0; distal 1.0.0; with two terminal spines. Abdomen 4.84 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.054 mm long; thin, curved, compressed, pointed; uniformly, thickly distributed. Vulva (Figs. 57-59) arch- like in dorsal view, frontally pointed; slightly 622 THE JOURNAL OF ARACHNOLOGY Figures 60-65. — Dysdera cribellata, right male bulb and female spinnerets. 60, DD frontal; 61, DD external; 62, DD posterior; 63, P external; 64, Right ALS; 65, Right PLS. wider than long; DF wide. MF poorly deveL oped. VA frontal region completely sclero- tized; posterior region sclerotized at anterior area. AVD hardly visible. S attachment not projected under VA; arms are slightly shorter than DA, straight; tips not projected; neck as wide as arms. TB usual shape. ALS (Figs. 64, 65) with PS; remaining piriform spigots more external than MS, arranged in two rows; 11 + 1 piriform gland spigots; PMS, PLS with 5-10 aciniform gland spigots. Intraspecihc variation. — Male cephalo- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 623 thorax ranges in length from 3.62-4.66 mm, female from 3.82-3.96 mm. Carapace frontal lateral margins usually parallel. AME sepa- ration from diameter. PLE-PME from Va- Vi diameter. Sternum ornamentation somewhat reduced. Spination variability in Table 5. Additional material examined. — TENERIFE: ?: ?, 21 December 1940, 1 9 (J. Denis, num. BMNH 1940.12.21.15 BMNH). La Laguna: El Moquinal, 28 November 1993, 1 9 (Amedo & Ribera, num. 4794 (T29) UB); 28 November 1993, 19 (Amedo & Ribera, num. 4817 (T20) UB). Santa Cruz de Tenerife: Cmz del Carmen, 12 May 1996, 2S (M. Naranjo, num. 3148 UB); 12 May 1996, 19 (M. Naranjo, num. 3149 UB); 12 May 1996, Id, (M. Naranjo, num. 3150 UB); 12 May 1996, 19 (M. Naranjo, num. 3151 UB). Monte de las Mercedes, 24 May 1996, Id (P. Oromi, num. 3162 UB); 24 May 1996, Id (P. Oromi, num. 3163 UB). Sima de la Robada, 13 Febmary 1992, 19 (P. Ororm, num. 2514 UL). Taganana, 20 Febmary 1989, 1 9 (Garcia Alayon, num. 2599 UL). Dysdera medinae: TEN- ERIFE: Santa Cruz de Tenerife: Monte Aguirre, 4 June 1986, Id paratype (C.G. Campos, num. 2741 UL). Distribution. — ^Tenerifean endemic. Known from several localities spread through the northern slope of the island, including the An= aga and Teno massifs. Comments. — The original male material of this species seems to have been lost. However, the female used in the original description was available for the present study. The original description of D. cribellata (Simon 1883), as well as the remaining Dys~ dera species described in that work, lacked any reference to the locality. In a subsequent paper (Simon 1907) the original locations were assigned using new labelled material. Thus, D. cribellata was thought to be present in La Palma. However, after examination of the drawings of the male bulb drawings from both the description and the redescription, they were actually considered to belong to dif- ferent species (Amedo et al. 1996). Therefore, the report of this species in La Palma was due to a incorrect identification. The presence of D. cribellata in Tenerife has been documented previously (Bosenberg 1895; Denis 1941). Examination of the D. medinae male para- type and the D. volcania female allotype did not show any diagnostic character with regard to D. cribellata. In both cases misidentifica- tion was probably due to unavailability of D. cribellata type specimens. Dysdera crocota C.L. Koch 1839 Dysdera crocota C.L. Koch 1839: 81. -Schmidt 1973: 360-361. -Wunderlich 1991: 284-286, 292-293, figs. 28-31 [d,9]. -Amedo et al. 1996: 252-253. -Amedo & Ribera 1997. Dysdera inaequuscapillata Wunderlich 1991: 295, figs. 42-46 [d, 9 ]. Holotype male from Punta Hi- dalgo, La Laguna, Tenerife, Canary Islands; 14 Febmary 1986, R. Wiss leg.; num. 3934; Stored at UL. Examined. New synonymy. Diagnosis. — -Dysdera crocota can be dis- tinguished from all other Canarian species, ex- cept D. lancerotensis Simon 1907, by cara- pace with a bilobed posterior margin in both sexes, in males by having bulb posterior apophysis (P) not fused to tegulum (T) and a strongly sclerotized apophysis in frontal-distal tip of distal division (DD), and in females by an unsclerotized frontal part of vulva ventral arch (VA). Males and females differ from D. lancerotensis by lack of spines on frontal legs (although not always so), males by shape of the bulb frontal distal division tip and poste- rior apophysis with only two ridges, and fe- males by rectangular shape of vulva dorsal arch (DA) (in dorsal view) and projection of ventral under dorsal arch (in dorsal view). Material examined. — TENERIFE: ?: ?, 18/4/ 84, 1(329 (N.P. Ashmole, num. 2715 UL). Adeje: Playa Paraiso, 10-50 m, 24-30 December 1994, 1 9 (F Gasparo, FG). Buenavista: Teno Alto,? March 1994, 19 (Oromi, num. 2937 UB); ljuv., (Oromi, num. 4814 (T6) UB); 13, (Oromi, num. 4819 (T2) UB); 3 9 (R Oromi, num. 4823-5 UB). El Rosario: Tabaiba, MSS-1, 9 October 1990, 19 (A.L. Medina, num. 2774 UL). El Sauzal: Around Cueva Labrada,? November 1993, 5 9 (Amedo & Ribera, num. 4807-11 UB); 13 (Amedo & Ribera, num. 4832 (T46) UB; ljuv. (Amedo & Ribera, num. 4834 (T49) UB). El Tanque: El Tanque, 550m, 26 December 1994, 1 9 (F. Gasparo, FG). Icod de los Vinos: Altos de El Sobrado, 15 March 1995, 13, (G. Ortega, MCNT). Icod, 21 December 1982, 19 (P. Morales, num. 2768 UL). Garachico: La Montaneta, 18 Febmary 1996, 83493 juv. (Ar- nedo & Oroim, num. 3106-17 UB). La Laguna: Co- comoto,? Febmary 1989, 1 9 (C. Deniz, num. 2596 UL). El Moquinal, 28 November 1993, 1 9 (Amedo & Ribera, num. 4812 (T5) UB); 23 January 1997, 1 9 (P. Oromi, num. 3197 UB). La Laguna, 12 Feb- mary 1987, 13 (C.G. Campos, num. 2739 at UL); 28 December 1987, 13 (C.G. Campos, num. 2688 UL); ? November 1988, ljuv. (C. Deniz, num. 2597 624 THE JOURNAL OF ARACHNOLOGY UL). Las Mercedes, 24 November 1982, Ic? (A. Santaella, num. 2777 UL); 25 October 1984, IS (C.G. Campos, num. 2740 UL). Los Rodeos, 1 <5,3 ? (R.G. Becerra, num. 2582 RG). Mesa Mota, 4 June 1983, 19 (R. Vonk, num. 2773 UL). San Diego, 24 November 1982, 1 9 (E. Cavero, num. 2767 UL). La Matanza de Acentejo: La Matanza, 900 m, 2 June 1996, 19 (M. Naranjo, num. 3172 UB); 1(5 1 9, 4juv., (M. Naranjo, 3187 UB). La Or- otava: Around Cueva del Bucio, 21 October 1994, 1 S (Arnedo, Ribera & Serra, num. 4003 UB). Aguamansa, La Caldera, 21 October 1994, 25^19 (Amedo, Ribera & Serra, num. 4004-6 UB). La Victoria de Acentejo: Las Lagunetas, 4 February 1989, 19 (O. Torres, num. 2690 UL); 30 October 1994, 19 (R Oromi, num. 4002 UB); 25 April 1995, 19 (P. Oronu, num. 3182 at UB), 1 May 1995, 19 (Oronu, num. 4175 (134) UB). Los Re- alejos: Los Realejos, 25 February 1983, 1 9 (A. Fox, num. 2769 UL). Los Silos: Erjos, 15 April 1973, 1 9 (J.M. Fernandez, num. 2503a UB). Santa Cruz de Tenerife: Cruz del Carmen, 12 May 1996, ljuv. (M. Naranjo, num. 3144 UB); 25 January 1997, 19 (P. Oromi, num. 3193 UB). Parque de Anaga, 6 February 1988, 1 9 (R Suarez, num. 2689 UB). Santa Ursula: Monte de Santa Ursula, 13 De- cember 1996, 19, (P. Oronu, num. 3211 UB). San- tiago del Teide: Los Gigantes, 28 March 1994, li5 (P. Oromi, num. 2816 UB). D. inaequuscapillata: TENERIFE: La Laguna: Punta Hidalgo, 14 De- cember 1986, 1 paratype, (R. Wiss, num. 2623 SMF); 14 December 1986, lljuv. (R. Wiss, num. 2731 UL); 23 December 1986, 19 (C.G. Campos, num. 2729 UL); 23 December 1986, 26 (C.G. Campos, num. 2738 UL). Las Mercedes, ? June 1984, 1(5, (S. Morales, num. 2730 UL). Distribution. — Cosmopolitan species, spread all over the world, probably due to hu- man introduction. Comments. — The presence of D. crocota has been documented in all the islands of the archipelago, with the exception of Fuerteven- tura and Lanzarote. In the Canaries, D. cro- cota is always found in habitats disturbed by human activities. It may suggest that this spe- cies has recently been introduced in the ar- chipelago by man. After examination of several types of D. in- aequuscapillata Wunderlich 1991, they were considered to belong to D. crocota. This mis- identification is extraordinarily surprising. The original author was aware of the presence of D. crocota in the Canaries and even, in the same work, mentioned and drew several char- acters of D. crocota. However, he described D. inaequuscapillata as a different species on the basis of the ‘uniqueness' of its male bulb in the Canaries. Dysdera curvisetae Wunderlich 1987 Figs. 66-72 Dysdera curvisetae Wunderlich 1987; 291, figs. 12- 17 [(5]. Holotype male from small cave at the North coast of San Marcos, Icod de los Vinos, Tenerife, Canary Islands; in VIII, J. Wunderlich leg.; Stored at SMF. Examined. -Wunderlich 1991: 284-287. Diagnosis. — Dysdera curvisetae can be dis- tinguished from all other Dysdera species, ex- cept D. ratonensis and D. verneaui Simon 1883, by its wide frontal border and diamond- shaped carapace (Fig. 66). It differs from D. ratonensis and D. verneaui by its poorly-spi- nated legs and shape and size of male abdom- inal dorsal hairs. Description. — Holotype male: Figs. 66-72. Carapace (Fig. 66) 5.42 mm long; maximum width 4.2 mm; minimum width 2.94 mm. Dark red, frontally darker, becoming lighter towards back; slightly foveate at borders, slightly wrinkled with small black fine-tex- tured granular material mainly at front. Fron- tal border roughly straight, about ^5 carapace length; anterior lateral borders convergent; sharpened at maximum dorsal width point, back lateral borders straight; back margin wide, straight; transversal suture on dorsal medial posterior surface. AME diameter 0.32 mm; PLE 0.27 mm; PME 0.23 mm; AME on edge of frontal border, separated one from an- other about % of diameter, close to PLE; PME very close to each other, less than Va PME di- ameter from PLE. Labium trapezoid- shaped, base wider than distal part; longer than wide at base; semicircular groove at tip. Sternum dark red, darkened on borders; very slightly wrinkled, mainly between legs and frontal border; uniformly covered in slender black hairs. Chelicerae (Fig. 67) 3.08 mm long, about % of carapace length in dorsal view; fang long, 2.1 mm; basal segment dorsal, ventral side completely covered with piligerous granula- tions (small, dense). Chelicera inner groove medium-size, about V5 cheliceral length; armed with three teeth and lamina at base; D > B > M (D, B very similar, all large); D trapezoid, located roughly at center of groove; B close to basal lamina; M at middle of B and D. Legs dark orange-colored. Lengths of male ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 625 Figures 66-68. — Dysdera curvisetae. 66, Cara- pace, dorsal; 67, Left chelicera, ventral; 68, Right male bulb, external. Scale bars in mm. described above: fel 5.6 mm (all measure- ments in mm); pal 3.64; til 5.6; mel 5.6; tal 0.91; total 21.35; fe2 4.41; pa2 3.01; ti2 4.48; me2 2.68; ta2 0.84; total 15.42; fe3 3.64; pa3 1.75; ti3 2.94; me3 3.78; ta3 0.84; total 12.95; fe4 4.4; pa4 2.52; ti4 4.13; me4 5.25; ta4 0.98; total 17.28; relative length: 1 > 4 > 2 > 3; palp: fe 2.8; pa 1.54; ti 1.47; ta 1.26; total 7.07. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0.1; distal 1 -0.0.0; ti3 ventral spines arranged in two bands: proxi- mal 0-1. 1.0; distal 1.0.0; with two terminal spines. Fe4 dorsal spines in two rows: anterior 1; posterior 1; ti4 dorsal spines arranged in three bands: proximal 0.0-1. 1; medial-proxi- mal 1.1.1; distal 1.0.0; ti4 ventral spines ar- ranged in three bands: proximal 1.1.0; medial- proximal 0.0- 1.0-1; distal 1.0.1; with two terminal spines. Dorsal side of frontal legs, ventral side of palp with a fine-textured, gran- ular, piligerous surface. Claws with 10-14 teeth; hardly larger than claw width. Abdomen 7.7 mm long; cream-colored; cy- lindrical. Anterior abdominal dorsal hairs 0.126 mm long (large); medium thickness, curved, compressed, blunt, tip not enlarged; uniformly, thickly distributed. Posterior ab- dominal dorsal hairs 0.036-0.054 mm long; thick, curved, not compressed, tip enlarged. distally acuminated; uniformly, thickly dis- tributed. Spinneret gland spigot data not avail- able. Male bulb (Fig. 68): T slightly smaller than DD; external, internal distal border sloped backwards. DD slightly bent in lateral view, clearly less than 45°; internal distal border not expanded. IS and ES equally developed; IS truncated at DD middle part. DD tip (Figs. 69-71) straight in lateral view. C present, short; distal end on DD internal tip; well-de- veloped; located close to DD distal tip; prox- imal border sharply decreasing; distal border stepped, upper tip not projected, pointed; ex- ternal side hollowed. AC present. LF absent. L well-developed; external border not sclero- tized, laterally slightly folded; distal border di- vergent, most external part perpendicular, con- tinuous. AL present, very poorly developed; proximal border in posterior view fused with DH. P (Fig. 72) fused to T; perpendicular to T in lateral view; lateral length from of T width; ridge present, perpendicular to T, not expanded; upper margin smooth; not distally projected; back margin not folded. Female: Unknown. Intraspecific variation. — Unknown. Distribution. — ^Tenerifean endemic. Known from a single locality on the island’s northern slope. Dysdera esquiveli Ribera & Blasco 1986 Figs. 75-86, Table 6 Dysdera esquiveli Ribera & Blasco 1986: 42-44, fig. lA-F [3,?]. Holotype male and paratype fe- male from Cueva del Viento-Sobrado, Icod de los Vinos, Tenerife, Canary Islands; 23 March 1983, J.L. Martin leg.; 6 num.T-CV-1 18, num. T-CV- 119; Stored at UL. Examined. -Wunderlich 1991: 284-287. Diagnosis. — Dysdera esquiveli can be dis- tinguished from all other Dysdera species, ex- cept D. hernandezi new species and D. gol- lumi, by its small size (carapace < 2.5 mm) and complete eye reduction. It differs from D. gollumi (male unknown) by its smooth cara- pace. It differs from the very similar D. her- nandezi new species (male unknown) by pres- ence of cheliceral granulation, fang shape, and distinct spination pattern (Table 6). Description. — Holotype male: Figs. 75-77, 81-84. Carapace (Fig. 75) 1.96 mm long; maximum width 1.46 mm; minimum width 0.88 mm. Brownish-orange, uniformly distrib- 626 THE JOURNAL OF ARACHNOLOGY Figures 69-74. — 69-72, Dysdera curvisetae, right male bulb. 69, DD frontal; 70, DD external; 71, DD posterior; 72, P external. 73-74. Dysdera chioensis, spinnerets. 73, Right ALS; 74, Right PLS. uted; slightly foveate at borders, wrinkled at middle, covered with tiny granulations. Fron- tal border roughly round, markedly smaller than Vz carapace length; anterior lateral bor- ders slightly divergent, or parallel; rounded at maximum dorsal width point, back lateral bor- ders straight; back margin narrow, straight. Eyeless. Labium trapezoid-shaped, base wider than distal part; as long as wide at base (tri- angle-like); semicircular groove at tip. Ster- num orange, uniformly distributed; wrinkled; covered in hairs mainly on margin. ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 627 Table 6. — Intraspecific spination variability of Dysdera esquiveli. Proximal Medial-proximal Medial-distal Distal Tibia 3 dorsal 1.0.0- 1 0-1.0- 1.0 0 1.0.0- 1 Tibia 4 dorsal O-LO.O 0-1.0- 1.0-1 0 1. 0.0-1 Tibia 3 ventral 0-1.0- 1.0-1 0 0 0-1.0- 1.0 Tibia 4 ventral 0-2.0- 1.0 1.1.1 0-1.0- 1.0-1 0- 1.0.0- 1 Number of rows Number of spines Femur 3 dorsal 1 0-1 Femur 4 dorsal 2 0-1/1-3 Number of spines Patella 3 ventral 0-1 Patella 4 ventral 1-3 Chelicerae (Fig. 76) 0.67 mm long, about Va of carapace length in dorsal view; fang me- dium-sized, 0.51 mm; basal segment dorsal side completely covered with piligerous gran- ulations (distally scarce), ventral side smooth. Chelicera inner groove medium- size, about % cheliceral length; armed with three teeth and lamina at base; D M == B; D triangular, located roughly at center of groove; B close Figures 75-80. — Dysdera esquiveli. 75, Cara- pace, dorsal; 76, Left chelicera, ventral; 77, Left male bulb, external; 78, Vulva, dorsal; 79, Vulva, ventral; 80, Vulva, lateral. Scale bars in mm. to basal lamina; M close to B. Legs pale yel- low. Lengths of male described above: fel 1.67 mm (all measurements in mm); pal 1.06; til 1.39; mel 1.34; tal 0.46; total 5.92; fe2 1.52; pa2 0.89; ti2 1.32; me2 1.24; ta2 0.4; total 5.37; fe3 1.11; pa3 0.61; ti3 0.76; me3 1.01; ta3 0.28; total 3.77; fe4 1.52; pa4 0.78; ti4 1.14; me4 1.34; ta4 0.4; total 5.18; relative length: 1 > 2 > 4 > 3; palp: fe 0.86; pa 0.43; ti 0.43; ta 0.51; total 2.23. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; pa3 1-0 ventral; ti3 dorsal spines arranged in two bands: proximal 1.0.0; distal 1.0.0; ti3 ventral spines arranged in two bands: proximal 0.1- 0.0; distal 1.0.0; with two terminal spines. Fe4 dorsal spines in one row: 1; pa4 2-3 ventral; ti4 dorsal spines arranged in two bands: me- dial-proximal 1.0.1; distal 1.0.1; ti4 ventral spines arranged in four bands: proximal 1.0.0; medial-proximal 1.1.1; medial-distal 1.0- 1.1; distal 1.0.1; without terminal spines. Dorsal side of frontal legs smooth; ventral side of palp smooth; long, spine-like hairs on ventral posterior ti, fe. Claws with 10-14 teeth, length twice claw width. Abdomen 2.28 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.027 mm long; medium thickness, roughly straight, not com- pressed, blunt, tip enlarged; uniformly, scantly distributed. Male bulb (Fig. 77): T slightly smaller than DD; external distal border straight; internal sloped backwards. DD slightly bent in lateral view, clearly less than 45°; internal distal bor- der not expanded. IS and ES equally devel- oped; IS truncated at DD middle part. DD tip straight in lateral view. C present, short; distal 628 THE JOURNAL OF ARACHNOLOGY Figures 81-86. — Dysdera esquiveli, right male bulb and female spinnerets. 81, DD frontal; 82, DD external; 83, DD posterior; 84, P external; 85, Right ALS; 86, Right PLS. end on DD internal tip; welLdeveloped; lo- cated close to DD distal tip (Figs. 81-83); proximal border sharply decreasing; distal border stepped, upper tip not projected, round- ed; external side hollowed. AC present. LF absent. L poorly developed; external border not sclerotized, laterally slightly folded; distal border approximately parallel, not continuous, upper sheet slightly folded at middle (?). AL present, very poorly developed; proximal bor- der in posterior view fused with DH. P (Fig. 84) fused to T; perpendicular to T in lateral ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 629 Table 7. — Intraspecific spination variability of Dysdera gibbifera. Proximal Medial-proximal Medial-distal Distal Tibia 3 dorsal 1. 1-2.1 O-l.O.O 0 1. 0.0-1 Tibia 4 dorsal 0.0.1 1.0-1. 1 1.0- 1.0 1.0.1 Tibia 3 ventral O-l.O.O Ll.O-l 0 1.0- 1.0 Tibia 4 ventral 1. 1-2.1 1.0- 1.0-1 0.1.0 1.0.1 Number of rows Number of spines Femur 3 dorsal 0 0 Femur 4 dorsal 1 0-2 view; lateral length from % to as long as T width; ridge present, perpendicular to T, not expanded; upper margin smooth; distally slightly projected; back margin slightly folded towards internal side. Paratype female: Figs. 78-80, 85, 86. All characters as in male except: Carapace 2.14 mm long; maximum width 1.58 mm; mini- mum width 0.98 mm. Chelicerae 0.84 mm long; fang 0.6 mm. B > D = M (slightly). Leg lengths of female described above: fel 1.77 mm (all measure- ments in mm); pal 1.19; til 1.39; mel 1.26; tal 0.38; total 5.99; fe2 1.52; pa2 1.09; ti2 1.24; me2 1.14; ta2 0.38; total 5.37; fe3 1.19; pa3 0.66; ti3 1.21; me3 1.14; ta3 0.4; total 4.6; fe4 1.57; pa4 0.86; ti4 1.26; me4 1.52; ta4 0.4; total 5.61; relative length: 1 > 4 > 2 > 3; palp: fe 0.94; pa 0.38; ti 0.38; ta 0.51; total 2.21. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; pa3 1-0 ventral; ti3 dorsal spines arranged in two bands: proximal 1- 0.0.0; distal 1.0.0; ti3 ventral spines arranged in one band: distal 1.0.0; with two terminal spines. Fe4 dorsal spines in one row: 1; pa4 2-1 ventral; ti4 dorsal spines arranged in two bands: medial-proximal 1.0.1; distal 1.0.1; ti4 ventral spines arranged in four bands: proxi- mal 1.0.0; medial-proximal 1.1.1; medial-dis- tal 1.0.1; distal 1.0.1; without terminal spines. Abdomen 2.7 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.05 mm long; medium thickness, curved, compressed (?), blunt, tip not enlarged; uniformly, scantly dis- tributed. Vulva (Figs. 78-80) arch-like in dor- sal view, frontally pointed; slightly wider than long; DF wide. MF poorly developed. VA frontal region completely sclerotized; poste- rior region sclerotized at anterior area. AVD hardly visible. S attachment not projected un- der VA; arms as long as DA, slightly curved; tips not projected; neck as wide as arms. TB usual shape. ALS (Figs. 85, 86) with PS; re- maining piriform spigots more external than MS, arranged in one row; 3 + 1 piriform gland spigots; PMS, PLS with fewer than 5 aciniform gland spigots. Intraspecific variation. — Male cephalo- thorax ranges in length from 1.96-2.28 mm. Labrada specimen, carapace back margin slightly bilobulated. AME, PLE present, re- duced to tiny whitish spots. Sternum lacking ornamentation. Chelicera dorsal relative size from Va-% of carapace length. B larger than or equal to D, larger than or equal to M. In gen- eral, cheliceral teeth are small. M closer to B. P very slightly toothed at distal tip (1 or 2 teeth). Spination variability in Table 6. Additional material examined. — TENERIFE: El Sauzal: Cueva Labrada, 11 December 1984, 1(3 (J.J. Hernandez, num. 2529 UL). Icod de los Vinos: Cueva Felipe Reventon, 3 March 1984, 13 para- type (G.LE.T, num. 2526 UL); 20 June 1994, 13 (P. Oromi, num. 2801 UL); 22 April 1993, 1 9 (P. Oromf, num. 2548 UL); 18 May 1985, 13 (Her- nandez, Izquierdo & Medina, num. 2716 UL); ? May 1994, 13 (M. Arechavaleta, num. 2800 UL); ? May 1994, 1 9 (M. Arechavaleta, num. 2803 UL). Cueva del Viento-Sobrado, 23 March 1983, 13, ljuv. paratype (J.L. Martin, num. 2528 UL); 2 De- cember 1992, 13 (I. Izquierdo, num. 2549 UL); ? May 1994, 13 (Piquetas, num. 2804 UL). Distribution. — Tenerife endemic. Known from several lava tubes on the northern slope. Dysdera gibbifera Wunderlich 1991 Figs. 87-95, Table 7 Dysdera gibbifera Wunderlich 1991: 293-294, fig. 35, 36, 38, 39 [3] (3; non 9, incorrect identifi- cation). Holotype male from MSS-3 Monte del Agua, Los Silos, Tenerife, Canary Islands; 10 July 1988, A.L. Medina leg.; num.T-H3-124; 630 THE JOURNAL OF ARACHNOLOGY Stored at UL. Examined. -Wunderlich 1991:284- 287. -Amedo & Ribera 1997. Diagnosis* — -Dysdera gibbifera can be dis- tinguished from most of remaining Canarian Dysdera species by its large size (carapace 6.00 mm long). It differs from other large spe- cies such as D. ambulotenta, D. hirguan Ar~ nedo, Oromi & Ribera 1996, D. insulana, D. labradaensis and D. longa Wunderlich 1991, by its poorly spinated legs (Table 7). Description. — Holotype male: Figs, 87-92. Carapace (Fig. 87) 5.81 mm long; maximum width 4.97 mm; minimum width 3.22 mm. Dark brownish-red, frontally darker, becoming lighter towards back; smooth with some black fine-textured granular material mainly at front. Frontal border roughly triangular, from V2-% carapace length; anterior lateral borders slight- ly convergent; sharpened at maximum dorsal width point, back lateral borders straight; back margin wide, straight. AME diameter 0.23 mm; PLE 0.21 mm; PME 0.2 mm; AME slightly back from frontal border, separated one from another about Vi of diameter, far from PLE; PME about Vk of diameter apart, about % PME diameter from PLE. Labium trapezoid-shaped, base wider than distal part; longer than wide at base (rectangle-like); semicircular groove at tip. Sternum brownish- red, frontally darker, becoming lighter towards back; slightly wrinkled; covered in hairs mainly on margin. Chelicerae (Fig. 88) 2.94 mm long, about ^5 of carapace length in dorsal view; fang me- dium-sized, 2.1 mm; basal segment dorsal, ventral side completely covered with piliger- ous granulations (small, dense). Chelicera in- ner groove short, about Vs cheliceral length; armed with three teeth and lamina at base; D > B = M (large); D trapezoid, located near segment tip; B close to basal lamina; M close to B. Legs dark orange-colored. Lengths of male described above: fel 5.6 mm (all mea- surements in mm); pal 3.99; til 5.32; mel 4.9; tal 0.91; total 19.81; fe2 5.11; pa2 3.57; ti2 4.83; me2 4.76; ta2 0.91; total 19.18; fe3 4.41; pa3 2.59; ti3 3.5; me3 4.41; ta3 0.91; total 15.82; fe4 5.67; pa4 3.01; ti4 4.62; me4 6.02; ta4 1.12; total 20.44; relative length: 4 > 1 > 2 > 3; palp: fe 3.5; pa 1.68; ti 1.75; ta 1.47; total 8.4. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands; proximal 1.1.1; Figures 87-89. — Dysdera gibbifera. 87, Cara- pace, dorsal; 88, Left chelicera, ventral; 89, Right male bulb, external. Scale bars in mm. distal 1.0.0; ti3 ventral spines arranged in three bands: proximal 1 -0.0.0; medial-proxi- mal 1.1.0; distal 1.0.0; with two terminal spines. Fe4 dorsal spines in one row: 0-1; ti4 dorsal spines arranged in four bands: proximal 0.0.1; medial-proximal 1.1.1; medial-distal 1.0.0; distal 1.0.1; ti4 ventral spines arranged in three bands: proximal 1.1.1; medial-proxi- mal 1. 1-0.0; distal 1. 0-1.1; with two terminal spines. Dorsal side of frontal legs, ventral side of palp with a fine-textured, piligerous, gran- ular surface. Claws with more than 20 teeth, slender, length twice claw width. Abdomen 6.3 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.009=0.027 mm long (very small); medium thickness, roughly straight, not compressed, blunt, tip enlarged; uniformly, scantly distributed. ALS (Figs. 94, 95) with PS; remaining piriform spigots no more external than MS, arranged in three rows; 18 A 1 piriform gland spigots; PMS, PLS with more than 20 aciniform gland spig- ots. Male bulb: (Fig. 89). T slightly longer than DD, or T as long as DD; external, internal distal border sloped backwards. DD not bent, same T axis in lateral view; internal distal bor- der not expanded. ES wider, more sclerotized than IS; IS continuous to tip. DD tip (Figs. 90=92) straight in lateral view. C present, short; distal end on DD internal tip; well-de- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 631 Figures 90-95. — Dysdera gibbifera, right male bulb and spinnerets. 90, DD frontal; 91, DD external; 92, DD posterior; 93, P external; 94, Right ALS; 95, Right PLS. veloped; located far from DD distal tip; prox- imal border sharply decreasing; distal border sloping on its base, upper tip not projected, pointed; external side hollowed (slightly). AC absent. LF present; distally not projected; well-developed. L well-developed; external border sclerotized, laterally markedly folded backwards; distal border divergent, continu- ous. AL absent. P (Fig. 93) fused to T; mark- edly sloped on its proximal part, perpendicular on distal; lateral length as long as or longer than T width; ridge present, not sclerotized. 632 THE JOURNAL OF ARACHNOLOGY perpendicular to T, distinctly expanded, rounded; upper margin smooth; not distally projected; back margin not folded. Female: Unknown. Intraspecific variation, — Male cephalo^ thorax ranges in length from 5.81-7.00 mm. Carapace very slightly wrinkled. AME sepa- ration from of diameter. Sternum hardly wrinkled. P distally slightly toothed. Spination variability in Table 7. Additional material examined. — TENERIFE: Icod de los Vinos: Cueva de Felipe Reventon; 17 February 1985, 1*5 (J.J. Hernandez & A.L. Medina, num. 2709 UL). Los Silos: Monte del Agua, 6 July 1990, Id (C.G. Campos, num. 2779 UL). Distribution. — Tenerifean endemic. Known from two localities at westernmost part of the northern slope, including Teno. Comments. — The study of the female specimens assigned to this species in the orig- inal description (Wunderlich 1991) shows that they actually belonged to a different species, the already identified Tenerifean species D. in- sulana (Arnedo & Ribera 1997). Therefore, the female of this species is currently un- known. Dysdera gollumi Ribera & Arnedo 1994 Figs. 118-119 Dysdera gollumi Ribera & Arnedo 1994: 115-119, fig. 1-3 [9]. Holotype female from Cueva de Los Roques, La Orotava, Tenerife, Canary Islands; 27 October 1991, C. Ribera leg.; num. 2567; Stored at UB. Examined. Diagnosis. — Dysdera gollumi can be dis- tinguished from most of the Canarian Dysdera species by the eye reduction, spineless legs and small size (carapace < 2.00 mm long). It differs from the other small and troglomorphic species, D. esquiveli and D. hernandezi new species, by its markedly foveate carapace. Description. — Holotype female: Carapace 2.05 mm long; maximum width 1.49 mm; minimum width 0.79 mm. Dark reddish- brown, darkened at borders; heavily wrinkled, foveate, covered with small black ‘granules'. Frontal border roughly triangular, markedly smaller than Vi carapace length; anterior lat- eral borders divergent; rounded at maximum dorsal width point, back lateral borders round- ed; back margin projected. PME, PLE lost; AME markedly reduced (tiny bright spots); AME diameter 0.022 mm; AME separation 0.12 mm. Labium trapezoid- shaped, base wid- er than distal part; as long as wide at base (triangle-like); semicircular groove at tip. Sternum orange-brown, darkened on borders; heavily wrinkled; covered in hairs mainly on margin. Chelicerae 0.63 mm long, about Va of car- apace length in dorsal view; fang short, 0.44 mm; basal segment proximal dorsal side scantly covered with large piligerous granu- lations. Chelicera inner groove medium-size, about % cheliceral length; armed with three teeth and lamina at base; D ^ B > M (very slightly); D triangular, located roughly at cen- ter of groove; B close to basal lamina; M at middle of B and D. Legs bicolored, darker on proximal border, becoming lighter distally. Lengths of female described above: fel 2.1 mm (all measurements in mm); pal 1.03; til 1.96; mel 2; tal 0.51; total 7.6; fe2 1.72; pa2 1.03; ti2 1.68; me2 1.77; ta2 0.51; total 6.71; fe3 1.44; pa3 0.7; ti3 1.12; me3 1.49; ta3 0.42; total 5.17; fe4 1.91; pa4 0.98; ti4 1.58; me4 2.05; ta4 0.56; total 7.08; relative length 1 > 4 > 2 > 3; palp: fe 0.76; pa 0.36; ti 0.41; ta 0.61; total 2.14. Spination: spineless. Dorsal side of frontal legs smooth; ventral side of palp covered with hairs, lacking a granular surface. Claws with 8 teeth or less, robust, hardly larger than claw width. Abdomen 3.26 mm long; whitish; globular. Abdominal dorsal hairs 0.054 mm long; thin, curved, not compressed, blunt, tip not en- larged; uniformly, scantly distributed. Vulva arch-like in dorsal view, frontally rounded; slightly wider than long; DF wide. MF poorly developed. VA frontal region completely sclerotized; posterior region sclerotized at an- terior area. AVD hardly visible. S attachment prpjected under VA; arms slightly shorter than DA, slightly curved; tips not projected; neck as wide as arms. TB usual shape. ALS (Figs. 118, 119) with PS; remaining piriform spigots more external than MS, arranged in one row; 5 + 1 piriform gland spigots; PMS, PLS with fewer than 5 aciniform gland spigots. Male: Unknown. Intraspecific variation.^ — Female cephalo- thorax ranges in length from 1.82-2.05 mm. Cheliceral teeth small, B > M > D. Chelicera groove short. Vulva frontally pointed, in dor- sal view. As wide as long. Additional material examined. — TENERIFE: ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 633 Table 8. — Intraspecific spination variability of Dysdera guayota. Proximal Medial- proximal Medial distal Distal Tibia 3 dorsal 1. 1-2.1 0 0 1.0.1 Tibia 4 dorsal O.O.O-l 1.1.1 0 1.0.1 Tibia 3 ventral 1. 1-2.0 0 0 1.0- 1.0 Tibia 4 ventral 1.1.1 0-1. 1.0 0 0-1. 0.0-1 Number of rows Number of spines Femur 1 dorsal 1 1-2 Femur 2 dorsal 1 2-3 Femur 3 dorsal 2 1-10/0-9 Femur 4 dorsal 2 1 -4/4-7 La Orotava: Cueva de Los Roques, 28 December 1982, ljuv. (J.L. Martin, num. 2537 UL); ? Novem- ber 1995, 1 9 (P. Oromi, num. 2966 UB). Distribution. — Tenerifean endemic. Known from a single lava tube, located at dry, mid- dle-southern slope. Comments. — Drawings of carapace, che- licera and vulva of this species have been pub- lished elsewhere (Ribera & Amedo 1994). In the present article, SEM photographs of spin- nerets are provided for the first time. Dysdera guayota new species Figs. 96-107, Table 8 Types. — Holotype male from Las Cany- adas, La Orotava, Tenerife, Canary Islands; 22 October 1995, A. Camacho leg.; num. 3153. Stored at UB. Paratype female from Las Can- adas del Teide, close to crossroads to Vilaflor, Adeje, Tenerife, Canary Islands; 29 November 1993, Amedo & Fluhr leg.; num. 4826. Stored at UB. Etymology. — The name in apposition of this species means ‘deviT in the language of the ‘guanches,’ the ancient aboriginal inhabi- tants of Tenerife. Diagnosis. — Dysdera guayota new species can be distinguished from all other Canarian Dysdera species except D. chioensis, D. la- bradaensis and D. lancerotensis by having spines on anterior femora (legs 1, 2). Males differs from D. lancerotensis by lacking hook- like apophysis in frontal-distal tip of bulb dis- tal division (DD) and bulb posterior apohysis (P) fused to tegulum (T), and females by the sclerotization of frontal part of vulva ventral (VA). It differs from D. labradaensis and D. chioensis by not showing eye reduction. Description. — Holotype male: Figs. 96-98, 102-105. Carapace (Fig. 96) 3.63 mm long; maximum width 3.17 nun; nunimum width 2.1 mm. Brownish-orange, frontally darker, becoming lighter towards back; smooth with some small black ‘granules’ mainly at front; hairy, covered with black hairs mainly at lat- eral and back borders. Frontal border roughly straight, from carapace length; anterior lateral borders slightly convergent; sharpened at maximum dorsal width point, back lateral borders straight; back margin wide, straight. AME diameter 0.16 mm; PLE 0.12 mm; PME 0.11 mm; AME slightly back from frontal border, separated one from another about 1 di- ameter or more, close to PLE; PME about Va of diameter apart, about PME diameter from PLE. Labium trapezoid- shaped, base wider than distal part; longer than wide at base (rect- angle-like); semicircular groove at tip. Ster- num orange, uniformly distributed; very slightly wrinkled, mainly between legs, frontal border; uniformly covered in slender black hairs. Chelicerae (Fig. 97) 1.72 mm long, about Vs of carapace length in dorsal view; fang me- dium-sized, 1.12 mm; basal segment dorsal side completely covered with piligerous gran- ulations, ventral side smooth. Chelicera inner groove short, about Vs cheliceral length; armed with three teeth and lamina at base; D > B > M (all large, B broken?); D trapezoid, located near segment tip; B close to basal lamina; M close to B. Front legs dark orange, back legs yellow. Lengths of male described above: fel 2.98 mm (all measurements in mm); pal 1.82; til 2.61; mel 2.61; tal 0.6; total 10.62; fe2 2.65; pa2 1.68; ti2 2.56; me2 2.23; ta2 0.6; 634 THE JOURNAL OF ARACHNOLOGY Figures 96-101. — Dysdera guayota new species. 96, Carapace, dorsal; 97, Left chelicera, ventral; 98, Right male bulb, external; 99, Vulva, dorsal; 100, Vulva, ventral; 101, Vulva, lateral. Scale bars in mm. total 9.72; fe3 2.1; pa3 1.21; ti3 1.68; me3 2.05; ta3 0.56; total 7.6; fe4 2.7; pa4 1.44; ti4 2.14; me4 2.56; ta4 0.65; total 9.49; relative length: 1 > 2 > 4 > 3; palp: fe 1.79; pa 0.88; ti 0.7; ta 0.93; total 4.3. Spination: palp spine- less. Fel 3-2 distal, anterior margin. Fe2 2-3 distal, anterior margin. Fe3 dorsal spines in two rows: anterior 9-10; posterior 5-4; ti3 dor- sal spines arranged in two bands: proximal 1.1.1; distal 1.0.1; ti3 ventral spines arranged in two bands: proximal 1.1.0; distal 1.0.0; with one terminal spine on anterior margin. Fe4 dorsal spines in two rows: anterior 4-2; posterior 7-5; ti4 dorsal spines arranged in three bands: proximal 0.0.1; medial-proximal 1.1.1; distal 1.0.1; ti4 ventral spines arranged in three bands: proximal 1.1.1; medial-proxi- mal 0.1.0; distal 1.0.1; with two terminal spines. Dorsal side of frontal legs, ventral side of palp covered with hairs. Claws with 8 teeth or less; robust, hardly larger than claw width. Abdomen 3.59 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.09 mm long; thick, slightly curved, compressed, blunt, tip not enlarged; uniformly, thickly distributed. Male bulb (Fig. 98): T as long as DD; ex- ternal, internal distal border sloped back- wards. DD proximally bent about 45° in lat- eral view; internal distal border not expanded. IS and ES equally developed; IS truncated at DD middle part. DD tip (Figs. 102-104) sloped towards back in lateral view. C present, short; distal end on DD internal tip; well-de- veloped; located close to DD distal tip; prox- imal border sharply decreasing; distal border stepped, upper tip projected, pointed; external side hollowed. AC present. LF absent. L well- developed; external border not sclerotized, lat- erally slightly folded; distal border divergent, not continuous; upper sheet strongly fold at middle. AL present, well-developed; proximal border in posterior view toothed on its internal half-part. P (Fig. 105) fused to T; perpendic- ular to T in lateral view; lateral length from 1/2-% of T width; ridge present, perpendicular to T, not expanded; upper margin markedly toothed, along its extent, few teeth; not dis- tally projected; back margin not folded. Paratype female: Figs. 99-101, 106, 107. All characters as in male except: Carapace 3.36 mm long; maximum width 2.75 mm; minimum width 1.86 mm. Orange. AME di- ameter 0.18 mm; PLE 0.12 mm; PME 0.12 mm; PME >5 diameter from PLE. Sternum yellow, frontally darker, becoming lighter to- wards back; smooth. Chelicerae 1.75 mm long; fang 0.31 mm; basal segment proximal dorsal side scantly covered with piligerous granulations. B > D > M (B, D similar). Legs yellow. Lengths of female described above: fel 2.33 mm (all measurements in mm); pal 1.49; til 1.86; mel 1.4; tal 0.46; total 7.54; fe2 2.14; pa2 1.4; ti2 1.86; me2 1.58; ta2 0.46; total 7.44; fe3 1.72; pa3 1.02; ti3 1.35; me3 1.58; ta3 0.46; total 6.13; fe4 2.37; pa4 1.26; ti4 1.86; me4 2.1; ta4 0.56; total 8.15; relative length 4 > 1 > 2 > 3; palp: fe 1.49; pa 0.64; ti 0.51; ta 0.74; total 3.38. Spination: palp spineless. Fel: 2 distal, anterior margin. Fe2: 2-1 distal, ante- rior margin. Fe3 dorsal spines in one row: 1 (medial frontal); ti3 dorsal spines arranged in two bands: proximal 1.1.1; distal 1.0.1; ti3 ventral spines arranged in two bands: proxi- mal 1.1.0; distal 1.0.0; with two terminal spines. Fe4 dorsal spines in two rows: anterior 1; posterior 4; ti4 dorsal spines arranged in ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 635 Figures 102-107. — Dysdera guayota new species, right male bulb and female spinnerets. 102, DD frontal; 103, DD external; 104, DD posterior; 105, P internal; 106, Right AES; 107, Right PLS. two bands: medial-proximal 1.1.1; distal cal. Abdominal dorsal hairs 0.162 mm long; 1.0.1; ti4 ventral spines arranged in three medium thickness, curved, compressed, point- bands: proximal 1.1.1; medial-proximal 0.1.0; ed; uniformly, thickly distributed. Vulva distal l-O.O.O-l; with two terminal spines. (Figs. 99-101) arch-like in dorsal view, fron- Dorsal side of frontal legs smooth. tally rounded; slightly wider than long; DF Abdomen 3.59 mm long; whitish; cylindri- wide. MF poorly developed. VA frontal re- 636 THE JOURNAL OF ARACHNOLOGY Table 9. — Intraspecific spination variability of Dysdera hernandezi. Proximal Medial- proximal Medial- distal Distal Tibia 3 dorsal 0 0 0 1.0.0 Tibia 4 dorsal 0 0 0 0.0.1 Tibia 3 ventral 0 0 0 0 Tibia 4 ventral 1.0- 1.0 0 0 1.0.0 gion completely sclerotized; posterior region sclerotized at anterior area. AVD hardly visi- ble. S attachment projected under VA; arms as long as DA, slightly curved; tips dorsally projected; neck as wide as arms. TB usual shape. ALS (Figs. 106, 107) with PS; remain- ing piriform spigots more external than MS, arranged in two rows; 7+1 piriform gland spigots; PMS, PLS with 5-10 aciniform gland spigots. Intraspecific variation. — Female cephalo- thorax ranges in length from 3.15-3.36 mm. PLE-PME from y^-l diameter. Spination var- iability in Table 8. Paratypes. — TENERIFE: Adeje: Las Canadas, close to crossroads to Vilaflor; ljuv.; November 1993, Arnedo & Fluhr leg.; num. 4815 (TIO); Stored at UB. Roque del Conde; ljuv. paratype; 16 March 1996, P. Oromf leg.; num. 3170; Stored at UL. Arona: Los Cristi- anos; 1 9 paratype; 20 January 1996, Oromf leg.; num. 3094; Stored at UL. Distribution. — ^Tenerifean endemic. Briiown from several localities on dry, south-western slope. Dysdera hernandezi new species Figs. 108-112, 120, 121, Table 9 Types. — Holotype female from Cueva La- brada, El Sauzal, Tenerife; 11 December 1984, J.J. Hernandez leg.; num. 3214; Stored at UL. Etymology. — This species is dedicated to the late Juan Jose Hernandez Pacheco, enthu- siastic Canarian biospeleologist and collector of the only two known specimens of this spe- cies. Diagnosis. — Dysdera hernandezi new spe- cies can be distingushed from most of the Canarian Dysdera by its flat and enlarged che- liceral fang (Fig. 109). A very similar fang shape is also present in D. ramblae Arnedo, Oromi & Ribera 1996 from La Gomera, from which differs by the smooth carapace, smaller size and eye reduction. Description. — Holotype female: Figs. 108- 112, 120, 121. Carapace (Fig. 108) 1.91 mm long; maximum width 1.4 mm; minimum width 0.9 mm. Pale orange, uniformly distrib- uted; very slightly foveate at borders, wrin- kled at middle, covered with tiny granulations. Frontal border roughly round, markedly smaller than Vi carapace length; anterior lat- eral borders divergent; rounded at maximum dorsal width point, back lateral borders straight; back margin projected. PME lost; AME, PLE markedly reduced (bright tiny spots); AME diameter 0.018 mm; PLE 0.018 mm; AME separation 0.16 mm; AME-PLE separation 0.018 mm. Labium trapezoid- shaped, base wider than distal part; as long as wide at base (triangle-like); semicircular groove at tip. Sternum pale orange, uniformly distributed; slightly wrinkled; covered in hairs mainly on margin. Chelicerae (Fig. 109) 0.79 mm long, about Vs of carapace length in dorsal view; fang me- dium-sized, 0.65 mm; enlarged on middle part; basal segment smooth, with no granula- tions. Chelicera inner groove medium-size, about % cheliceral length; armed with three teeth and lamina at base; B > D > M (not very different, small); D triangular, located roughly at center of groove; B close to basal lamina; M close to B. Legs pale yellow. Lengths of female described above: fel 1.42 mm (all measurements in mm); pal 0.98; til 1.21; mel 1.21; tal 0.37; total 5.2; fe2 1.35; pa2 0.93; ti2 1.16; me2 1.16; ta2 0.35; total 4.95; fe3 1.02; pa3 0.61; ti3 0.84; me3 0.98; ta3 0.32; total 3.77; fe4 1.35; pa4 0.74; ti4 1.12; me4 1.26; ta4 0.37; total 4.47; relative length 1 > 2 > 4 > 3; palp: fe 0.79; pa 0.42; ti 0.37; ta 0.51; total 2.09. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in one band: distal 1.0.0; ti3 ventral spines spineless; with one terminal spine on anterior margin. Fe4 dorsal ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 637 Figures 108-112. — Dysdera hernandezi new species. 108, Carapace, dorsal; 109, Left chelicera, ventral; 110, Vulva, dorsal; 111, Vulva, ventral; 112, Vulva, lateral. Scale bars in mm. spineless; ti4 dorsal spines arranged in one band: distal 0.0.1; ti4 ventral spines arranged in two bands: proximal 1. 1-0.0; distal 1.0.0; with two terminal spines. Dorsal side of fron- tal legs smooth; ventral side of palp covered with hairs, lacking a granular surface; long, spine-like hairs on posterior ti, fe (mainly ven- tral). Claws with 10-14 teeth, length twice claw width. Abdomen 2.37 mm long; whitish; globular. Abdominal dorsal hairs 0.036 mm long; thin, curved, not compressed, blunt, tip not en- larged; uniformly, scantly distributed. Vulva (Figs. 110-112) DA arch-like in dorsal view, frontally rounded; slightly wider than long; DF wide. MF poorly developed. VA frontal region completely sclerotized; posterior re- gion sclerotized at anterior area. AVD hardly visible. S attachment not projected under VA; arms as long as DA, straight; tips not pro- jected; neck as wide as arms. TB usual shape. ALS (Figs. 120, 121) with PS; remaining pir- iform spigots more external than MS, ar- ranged in one row; 4+1 piriform gland spig- ots; PMS, PLS with 5-10 aciniform gland spigots. Male: Unknown. Intraspecific variation. — Female cephalo- thorax ranges in length from 1.91-2.14 mm. Carapace frontal width about Vi of its length. Sternum wrinkled. Spination variability in Ta- ble 9. Paratype. — TENERIFE: El Sauzal: Cue- va Labrada; 1 $ paratype; 22 November 1984, J.J. Hernandez leg.; num. 2585; Stored at UB. Distribution. — Tenerifean endemic. Known from a single lava tube, located on middle- northern slope. Dysdera iguanensis Wunderlich 1987 Dysdera iguanensis Wunderlich 1987: 57-58, Figs. 2-6 [61 -Wunderlich 1991: 294-295, fig. 41 [9]. -Wunderlich 1991: 284-287. -Amedo et al. 1996: 244, fig. IF [6]. -Amedo & Ribera 1997. Distribution. — Canarian endemic, known from Tenerife and a single location in Gran Canaria. In Tenerife it is an abundant species, spread through several localities on northern slope, including Anaga and Teno massifs. Comments. — A complete redescription of this species has been published elsewhere (Ar- nedo & Ribera 1997). Dysdera insulana Simon 1883 Dysdera insulana Simon 1883: 294-295, fig. 19 [6] (6, non 9). -Simon 1907: 257-258, fig. A [d]. -Strand 1911: 190. -Reimoser 1919. -Denis 1941: 108. -Denis 1953: 2. -Schmidt 1973: 360- 361. -Wunderlich 1991: 67, 296. -Amedo et al. 1996: 271-272. -Amedo & Ribera 1997. Distribution. — Canarian endemic, known from Tenerife and a single location in Gran Canaria. In Tenerife, known from several lo- calities restricted to Anaga and closer loca- tion, formerly occupied by low-elevation lau- rel forest. Comments. — A complete redescription of this species has been published elsewhere (Ar- nedo & Ribera 1997). Dysdera labradaensis Wunderlich 1991 Figs. 113-117, 122, 123, Table 10 D. labradaensis Wunderlich 1991: 296, figs. 47-49 [?]. Holotype female from Cueva Labrada, El Sauzal, Tenerife, Canary Islands; 12 September 1984, G.I.E.T. leg.; num. T-CL-59; Stored at UL. Examined. -Wunderlich 1991: 284-287. Diagnosis. — Dysdera labradaensis differs from similar and sympatric species D. ambu- 638 THE JOURNAL OF ARACHNOLOGY Table 10. — Intraspecific spination variability of Dysdera labradaensis. Proximal Medial-proximal Medial-distal Distal Tibia 3 dorsal O-LO- 1.0-1 O-LO-2.0-1 1.0- 1.0-1 l.O-l.l Tibia 4 dorsal L2.1-2 1.1-3. 1 1.0-1. 1 1.1.1 Tibia 3 ventral 1.1.1 1.1.1 0 1.1.1 Tibia 4 ventral 0-1. 0-2. 1-2 1.1.1 1.1.1 1.1.1 Number of spines Patella 3 ventral 0-1 Patella 4 ventral 0-1 Number of rows Number of spines Femur 1 frontal 1 3 Femur 2 frontal 1 1-2 Femur 3 dorsal 2 3-4(distal)/2-3(proximal) Femur 4 dorsal 2 6-8(distal)/2-3(proximal) lotenta by presence of six eyes, spinated an- terior femora and presence of a ridge on vulva ventral arch (VA). It can be distinguished from other species with a ridge on ventral arch (Grancanarian D. arabisenen Arnedo & Ri- bera 1997, D. tibicena Arnedo & Ribera 1997 and Tenerifean D. iguanensis, D. rnontaneten- Figures 113-117. — Dysdera labradaensis. 113, Carapace, dorsal; 114, Left chelicera, ventral; 115, Vulva, dorsal; 116, Vulva, ventral; 117, Vulva, lat- eral. Scale bars in mm. sis) by possessing a ridge longer than the ven- tral arch (Fig. 116). Description. — Holotype female: Figs. 113- 117, 122, 123. Carapace (Fig. 113) 8.33 mm long; maximum width 6.58 mm; minimum width 3.78 mm. Brownish-orange, frontally darker, becoming lighter towards back; smooth with some small black ‘granules’ mainly at front; hairy, covered with black hairs mainly at lateral and back borders. Fron- tal border roughly straight, from V2-% cara- pace length; anterior lateral borders slightly convergent; sharpened at maximum dorsal width point, back lateral borders straight; back margin wide, straight. Eyes markedly reduced in size; AME diameter 0.16 mm; PLE 0.14 mm; PME 0.12 mm; AME separation 0.52 mm; AME-PLE separation 0.07 mm; PLE- PME separation 0.2 mm; PME separation 0.09 mm. Labium trapezoid- shaped, base wider than distal part; longer than wide at base (rect- angle-like); semicircular groove at tip. Ster- num orange, frontally darker, becoming ligh- ter towards back; very slightly wrinkled, mainly between legs, frontal border; uniform- ly covered in slender black hairs. Chelicerae (Fig. 114) 3.22 mm long, about 35 of carapace length in dorsal view; fang me- dium-sized, 2.52 mm; basal segment dorsal, ventral side completely covered with piliger- ous granulations (small, dense). Chelicera in- ner groove short, about 35 cheliceral length; armed with three teeth and lamina at base; D = B > M (large, similar in size); D trapezoid. ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 639 located roughly at center of groove; B close to basal lamina; M close to B. Legs dark or-- ange-colored. Lengths of female described above: fel 7.7 mm (all measurements in mm); pal 4.9; til 7; mel 7.21; tal 1.12; total 27.93; fe2 7.28; pa2 4.69; ti2 6.72; me2 7; ta2 1.12; total 26.71; fe3 6.3; pa3 3.43; ti3 4.9; me3 6.51; ta3 1.19; total 22.33; fe4 8.4; pa4 4.2; ti4 7.14; me4 9.45; ta4 1.33; total 3.52; rela- tive length 4 > 1 > 2 > 3; palp: fe 3.92; pa 2.1; ti 2.03; ta 2.38; total 10.43. Spination: palp spineless. Fel: 3 distal, anterior margin. Fe2: 1-2 distal, anterior margin. Fe3 dorsal spines in two rows: anterior 4-3 (distal); pos- terior 2-3 (proximal); pa3 1-0 ventral; ti3 dor- sal spines arranged in four bands: proximal 1- 0.0.0; medial-proximal 1.2-1. 1; medial-distal 1.1. 0- 1; distal 1.0.1; ti3 ventral spines ar- ranged in three bands: proximal 1.1.1; medial- proximal 1.1.1; distal 1.1.1; with two terminal spines. Fe4 dorsal spines in two rows: anterior 5-6 (2 distal); posterior 2-3; ti4 dorsal spines arranged in four bands: proximal 1.2.1; me- dial-proximal 1.2.1; medial-distal 1.1.1; distal 1.1.1; ti4 ventral spines arranged in four bands: proximal 1.1. 2-1; medial-proximal 0- 1.0- 1. 1; medial-distal 1.1.1; distal 2-1. 1.1; with two terminal spines. Dorsal side of fron- tal legs, ventral side of palp slightly covered with a fine-textured piligerous granular sur- face. Claws with more than 20 teeth, length twice claw width. Abdomen 9.8 mm long; cream-colored; cylindrical. Abdominal dorsal hairs 0.036- 0.072 mm long (small, variable); medium thickness, roughly straight, not compressed, blunt, tip enlarged; uniformly, scantly dis- tributed. Vulva (Figs. 115-117) rectangle- like in dorsal view, frontally rounded; twice as wide as long; DF wide. MF well-devel- oped; markedly sclerotized along its extent. VA frontal region completely sclerotized; posterior region sclerotized except for most internal area; sclerotized ridge at ventral VA external margin, longer than VA, fused to VA along its extent, back ends bent to in- ternal side. AVD clearly recognizable. S at- tachment projected under VA; arms as long as DA, slightly curved; ends projected an- teriors; neck as wide as arms. TB usual shape. ALS (Figs. 122, 123) with PS; re- maining piriform spigots more external than MS, arranged in three rows; more than 20 piriform gland spigots; PMS, PLS with 10- 15 aciniform gland spigots. Male: Unknown. Intraspecific variation. — Female cephalo- thorax ranges in length from 7.00-8.33 mm. B larger than M. Spination variability in Table 10. Additional material examined. — TENERIFE: Icod de los Vinos: Cueva del Viento-Sobrado, 30 November 1980, ljuv. (J.L. Martin, num. 2522 UL); 5 April 1981, ljuv. (J.L. Martin, num. 2515 UL); 17 September 1990, ljuv. (J.J. Hernandez, num. 2746 UL); 17 September 1990, ljuv. (J.J. Hernandez, num. 2747 UL); ? May 1994, ljuv. (J. Sala, num. 2802 UL). La Orotava: Cueva del Bu- cio, 4 August 1985, ljuv. (Martin & Machado, num. 2743 UL). El Sauzal: Cueva Labrada, 21 March 1983, 19, some remains, (J.L. Martin, num. 2531 UL); 28 June 1986, 19 (P. Oromi, num. 2513 UL). Distribution. — ^Tenerifean endemic. Known from several lava tubes located on northern slope of the island. Dysdera levipes Wunderlich 1987 Dysdera levipes Wunderlich 1987: 59-60, fig. 19- 22 [6]. -Wunderlich 1991: 284-287. -Amedo et al. 1996: 258-261, figs. 14A-F, 15A-D, 16A-C [3,9]. -Amedo & Ribera 1997. Dysdera multipilosa Wunderlich 1991: 301-302, figs. 68-71 [9]. Distribution. — Canarian endemic, found in Tenerife, La Gomera and Gran Canaria. In Tenerife has been reported from two localities on the northern slope and a single locality on middle- southern slope. Comments. — A complete redescription of this species has been published elsewhere (Ar- nedo et al. 1996). D. levipes is the only en- demic species reported from three different is- lands: La Gomera, Tenerife and Gran Canaria (a single specimen). Dysdera macra Simon 1883 Figs. 124-136, Table 11 Dysdera macra Simon 1883: 295-296, fig. 18 [3] (3, non 9). Neotype male, by present designa- tion, from Monte de Santa Ursula, Santa Ursula, Tenerife, Canary Islands; 27 February 1997, P. Oromi leg.; num. 3206; Stored at UB. -Simon 1907: 256-267, 259-260; fig. 3, dorsal [3]. - Strand 1911: 189. -Reimoser 1919: 200. -Denis 1941: 108. -Schmidt 1973: 360-361. -Amedo et al. 1996: 272. D. tenerijfensis Strand 1908: 772 [9]. Holotype fe- 640 THE JOURNAL OF ARACHNOLOGY Figures 1 18-123.— Spinnerets. 118, 119, Dysdera gollumi. 118, Right ALS; 119, Right PLS. 120-121, Dysdera hernandezi new species. 120, Right ALS; 121, Right PLS. 122—123, Dysdera labradaensis . 122, Right ALS; 123, Right PLS. male from Aguamansa (Aqua Manza), La Oro- tava, Tenerife, Canary Islands; unknown data, un- known leg.. Probably lost. Not examined. -Wunderlich 1991: 283. New synonymy. D. pergrada Wunderlich 1991: 305-306, figs. 83- 91 [(3,9]. Holotype male from close to La Oro- tava. La Orotava, Tenerife, Canary Islands; in II, M. Knosel leg.; num. 37163; Stored at SMF Ex- amined. New synonymy. D. pseudopergrada Wunderlich 1991: 306, figs. 94-97 [c3,9]. Holotype male from Barranco del Infiemo, Adeje, Tenerife, Canary Islands; in II, ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 641 Table 1 1 . — Intraspecific spination variability of Dysdera macra (hardly distinguishable from spine-like hairs). Medial- Medial- Proximal proximal distal Distal Tibia 3 dorsal 0-1. 0.0 0 0 1.0.0 M. Knosel leg.; num, 37168; Stored at SME Ex- amined. New synonymy. D. tabaibaensis Wunderlich 1991: 308, figs. 103- 107 [<3]. Holotype male from Tabaiba, El Rosa- rio, Tenerife, Canary Islands; 25 April 1990, C.G. Campos leg.; num. 03863; Stored at UL. Exam- ined. New synonymy. D. teideensis 1991: 309-310, figs. 112- 118 [(3,9]. Holotype male from Retamar del Tei- de. La Orotava, Tenerife, Canary Islands; 21 April 1984, C.G. Campos leg.; num. 2772; Stored at UL. Examined. New synonymy. Diagnosis.— macra is distingush- ed from all other Canarian Dysdera species, except D. brevisetae and D. esquiveli, by a stepped carapace (Fig. 125) and spine-like hairs on legs. It differs from D. esquiveli by absence of eye reduction. It is distinguished from D. brevisetae in both sexes by possess- ing less granulation on chelicerae, shorter che- Figures 124-130. — Dysdera macra. 124, Cara- pace, dorsal; 125, Carapace, lateral; 126, Left che- licera, ventral; 127, Right male bulb, external; 128, Vulva, dorsal; 129, Vulva, ventral; 130, Vulva, lat- eral. Scale bars in mm. liceral inner groove and distal and basal che- liceral teeth similar in size, and males by a barely visible lateral sheet (L) on the bulbus (Fig. 131). Description. — Neotype male: Figs. 124- 126, 131-134. Carapace (Fig. 124) 3.63 mm long; maximum width 2.93 mm; minimum width 2.1 mm. Dark red, uniformly distribut- ed; slightly foveate at borders, wrinkled at middle, covered with a black,fine-textured, granular surface. Frontal border roughly round, about % carapace length; anterior lat- eral borders slightly convergent; rounded at maximum dorsal width point, back lateral bor- ders straight; back margin narrow, straight; stepped in lateral view (Fig. 125). AME di- ameter 0.16 mm; PLE 0.14 mm; PME 0.12 mm; AME on edge of frontal border, separat- ed one from another about 1 diameter or more, close to PLE; PME very close to each other, about % PME diameter from PLE. Labium trapezoid-shaped, base wider than distal part; as long as wide at base (triangle-like); semi- circular groove at tip. Sternum dark red, uni- formly distributed; slightly wrinkled; uniform- ly covered in slender black hairs. Chelicerae (Fig. 126) 1.91 mm long, about Vs of carapace length in dorsal view; fang me- dium-sized; 1.3 mm; basal segment smooth, with no granulations. Chelicera inner groove short, about Vs cheliceral length; armed with three teeth and lamina at base; D > B > M (large, D markedly larger); D trapezoid, lo- cated near segment tip; B close to basal lam- ina; M close to B. Legs dark orange-colored. Lengths of male described above: fel 2.42 mm (all measurements in mm); pal 1.63; til 2.19; mel 2; tal 0.46; total 8.7; fe2 2.19; pa2 1.49; ti2 1.96; me2 1.86; ta2 0.42; total 7.92; fe3 1.77; pa3 1.11; ti3 1.21; me3 1.81; ta3 0.46; total 6.36; fe4 2.37; pa4 1.3; ti4 1.72; me4 2.23; ta4 0.56; total 8.18; relative length: 1 > 4 > 2 > 3; palp: fe 1.68; pa 0.93; ti 0.74; ta 0.79; total 4.14. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal 642 THE JOURNAL OF ARACHNOLOGY spines arranged in one band: distal 1.0.0; ti3 ventral 1 terminal spines. Fe4 dorsal spineless; ti4 dorsal spineless; ti4 ventral 1 terminal spines. Dorsal side of frontal legs smooth; ventral side of palp covered with a fine-tex- tured, piligerous, granular surface; long, spine-like hairs on posterior ti, fe. Claws with 8 teeth or less, robust, length twice claw width. Abdomen 4.19 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.027 mm long (small); medium thickness, roughly straight, not compressed, blunt, tip enlarged; uniform- ly, thickly distributed. Male bulb (Fig. 127): T slightly smaller than DD; external distal border straight; inter- nal sloped backwards. DD slightly bent in lat- eral view, clearly less than 45°; internal distal border not expanded. ES more sclerotized than IS; IS truncated at DD middle part; ES bend markedly sclerotized. DD tip (Figs. 131-133) straight in lateral view. C present, short; distal end on DD internal tip; well-de- veloped; located close to DD distal tip; prox- imal border sharply decreasing; distal border rounded, hardly stepped, upper tip not pro- jected, rounded; external side hollowed. AC present. LF absent. L reduced to distal part; external end projected, pointed. AL present, very poorly developed; proximal border in posterior view fused with DH except for its most internal part. P (Fig. 134) fused to T; perpendicular to T in lateral view; lateral length about Va of T width; ridge present, per- pendicular to T, not expanded; upper margin markedly toothed, on its distal part, few teeth (4-6); distally slightly projected; back margin not folded. Female: (from Monte de Santa Ursula, S. Ursula, Tenerife; num. 3206, UB). Figs. 128- 130, 135, 136. All characters as in male ex- cept: Carapace 3.4 mm long; maximum width 2.75 mm; minimum width 2.05 mm. AME di- ameter 0.16 mm; PLE 0.13 mm; PME 0.12 mm; PME diameter from PLE. Sternum dark orange, uniformly distributed; very slightly wrinkled, mainly between legs, frontal border. Chelicerae 1.86 mm long; fang 1.26 mm. Leg lengths of female described above: fel 2.33 mm (all measurements in mm); pal 1.54; til 1.96; mel 1.86; tal 0.46; total 8.15; fe2 2.14; pa2 1.4; ti2 1.82; me2 1.77; ta2 0.42; total 7.55; fe3 1.77; pa3 1.02; ti3 1.21; me3 1.58; ta3 0.46; total 6.04; fe4 2.14; pa4 1.25; ti4 1.72; me4 1.96; ta4 0.51; total 7.58; rela- tive length 1 > 4 > 2 > 3; palp: fe 1.4; pa 0.7; ti 0.56; ta 0.74; total 3.4. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0.0; distal 1. 0.0-1; ti3 ventral 2 terminal spines. Fe4 dorsal spineless; ti4 dorsal spine- less; ti4 ventral 2 terminal spines. Abdomen 4.19 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.072-0.108 mm long; medium thickness, curved, compressed, blunt, tip enlarged; uniformly, thickly distrib- uted. Vulva (Figs. 128-130) arch-like in dor- sal view, frontally rounded; slightly wider than long; DF wide. MF poorly developed. VA frontal region completely sclerotized; pos- terior region sclerotized at anterior area. AVD hardly visible. S attachment not projected un- der VA; arms as long as DA, clearly curved; tips not projected; neck as wide as arms. TB usual shape. ALS (Figs. 135, 136) with PS; remaining piriform spigots more external than MS, arranged in one row; 4+1 piriform gland spigots; PMS, PLS with 5-10 aciniform gland spigots. Intraspecific variation. — ^Male cephalo- thorax ranges in length from 2.75-3.63 nun, female from 2.93-3.45 mm. Carapace frontal lateral borders slightly convergent or parallel. AME separation from %-l diameter PLE- PME separation from +3-% diameter. Sternum ornamentation variable, from smooth to slightly wrinkled. Chelicera realtive length from Basal segment lacking dorsal gran- ulations, reduced to basal portion, or at distal internal margin. Chelicera inner groove from V3-% its length. Fang relative size from D only slightly larger than or as large as B. P relative size from Va-%. Female abdominal dorsal hair blunt, enlarged at fontal part and becoming pointed, longer to back. Spination variation given in Table 1 1 . Additional material examined. — TENERIFE: Arafo: 3 km N of Arafo, 950 m; 28 December 1994, 26 (F. Gasparo, FG). Fuente del loco, 5 km NW of Arafo, 1930 m, 28 December 1994, 19 (F. Gasparo, FG). Adeje: Roque del Conde, 16 March 1996, 16 (Oromi, num. 3121 UB); 16 March 1996, 19 (Ororm, num. 3122 UB). Arico: Barranco del Rio, 14-21 April 1981, Id 19, (J.M. Peraza, num. 2612 MCNT); 16-23 November 1984, Id (J.M. Peraza, num. 2609 MCNT); 29 November 1993, 1 d (M.A. Amedo, num. 2576 UB), El Rosario: Ta- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 643 Figures 131-136. — Dysdera macra, right male bulb and female spinnerets. 131, DD frontal; 132, DD external; 133, DD posterior; 134, P external; 135, Right ALS; 136, Right PLS. baiba, MSS-2, 9 October 1990, ljuv. (A.L. Medina, num. 2775 UL). La Victoria de Acentejo: El Diab- lillo, 21 February 1997, 19, (P. Oromi, num. 3207 UB). Las Lagunetas, 28 January 1993, IS (P. Oromi, num. 2547 UL); 27 March 1995, 19 (P. Oromf, num. 4110 UB). Gm'a de Isora: Above CWo, 750 m, 28 December 1994, 2d 19, (F. Gas- paro, EG). Giiunar: Barranco del Agua, 14 January 1984, 19 (P. Oromi, num. 2681 UL); 17 January 1997, 19, (P. Oromi, num. 3194 UB). Barranco de Badajoz, 1900 m, 18 December 1996, 19 (P. Oromi, num. 3190 UB); 1800 m, 27 December 644 THE JOURNAL OF ARACHNOLOGY 1996, 1? (R Oromi, num. 3195 UB). La Orotava: Base del zig-zag, 17 October 1984, 2? (C.G. Cam- pos, num. 2697 UL). Close to the Refugio, 2200 m, in VI, IS 19 (C.G. Campos, num. 2627 SMF). Iz- aha, November 1994, 19, (Amedo, num. 4827 (T40) UB). La Rosa de Piedra, 25 February 1996, IS , (Oromi & Emerson, num. 3126 UB). Las Cana- das del Teide, ?, IcJ (A. Machado, num. 2808 UB); 17 May 1983, 19, (C.G. Campos, num. 2686 UL); 14 May 1993, 19 (P. Oromi, num. 2817 UB); 12 December 1993, ljuv. (Oromi, num. 4837 (T52) UL); 3 June 1995, IS (R Oromi, num. 2969 UB); 11 June 1995, 19 (P. Oroim, num. 2968 UB); 24 May 1996, 19 (N. Zurita, num. 3173 UB). Teide, 2700m, 21 April 1984, 1(3 (C.G. Campos, num. 2766 UL). Los Realejos: La Fortaleza, 1 July 1990, 19 (C.G. Campos, num. 2760 UL). 17 May 1996, ljuv. (N. Zurita, num. 3155 UB); 1 9 (A. Camacho, num. 3157 UB); ljuv. (N. Zurita, num. 3158 UB); 19 (A. Camacho, 3156 UB). Roque Peral, 9 No- vember 1983, IS (C.G. Campos, num. 2701 UL); 19 April 1984, lc3 (C.G. Campos, num. 2718 UL); 12 June 1984, 19 (C.G. Campos, num. 2693 UL); 18 June 1984, 2S (C.G. Campos, num. 2694 UL). Santa Ursula: Barranco del Pino, 15 November 1984, 19 (J.P Peraza, num. 2737 UL). Monte de Santa Ursula, 13 December 1996, 3(339 (P. Oromi, num. 3212 UB). Vilaflor: El Pinalito, 16-23 Feb- ruary 1985, 1(31 9, (J.M. Peraza, num. 2610 MCNT); 24-31 May 1985, lc3 (J.M. Peraza, num. 2611 MCNT). Madre del Agua, 15 March 1990, lc3 (C.G, Campos, num. 2717 UL). Dysdera teideensis: TENERIFE: La Orotava: Retamar del Teide, 21 April 1984, 1(3 paratype (C.G. Campos, num. 2624 SMF). Las Canadas del Teide, 18 October 1984, 1 9 paratype (C.G. Campos, num. 2719 UL). Teide, 3050 m, 21 April 1984, 1(3 paratype (C.G. Campos, num. 2703 UL). Distribution. — Tenerifean endemic. A widespread species, collected throughout the island with the exception of Anaga and Teno massifs. Comments. — The distribution of D. macra was unknown before the present study. Nei- ther the original description (Simon 1883) nor the redescription of the species (Simon 1907) made any reference to its locality. Moreover, the report of this species in La Gomera by Strand (1911) has been claimed to be wrong (Amedo et al. 1996). The only type material of this species that was available for studying was a juvenile, probably the one originally described by Si- mon (1883) as the female of D. macra. How- ever, in a subsequent article (Simon 1907) the same author transferred this specimen to D. croc Ota. Fortunately, in this particular case both the original description and later rede- scription allowed the identification of the specimens belonging to this species. Amedo et al. (1996) considered D. macra as a dis- tinctive species on the basis of a double- toothed P. However, reexamination under SEM of some specimens formerly determined as D. pergrada and D. teideensis showed the presence of this character. In their original descriptions (Wunderlich 1991), D. pergrada, D. pseudopergrada, D. tabaibaensis and D. teideensis were distin- guished by: size of abdominal dorsal hairs, distal structures of the bulb and curvatures of P. In addition, D. tabaibaensis displayed a shorter distance between AME and relatively larger M tooth. Examination of the type ma- terial of these species, together with the study of about 40 newly available specimens, showed that (1) most of the formerly listed characters are polymorphic within the popu- lations, (2) the suggested differences in the distal structures of the bulb simply do not ex- ist and (3) the only truly distinguishable char- acter, although present only in a single spec- imen, is the shorter AME distance of D. tabaibaensis, which seems to fit better that of D. brevisetae. However, both male genitalia and the remaining somatic characters of D. ta- baibaensis correspond to those exhibted by the rest of the mentioned species. Finally, be- cause all these species are compatible with the descriptions of D. macra and in order to avoid unnecessary proliferation of names, the pre- ferred option has been to synonymize all these species with D. macra. The Dysdera tenerijfensis holotype seems to have been lost. Strand’s original description is fully fitted by both D. brevisetae and D. macra. However, because the type locality (Aguamansa) is located into the distributional range of the second species, D. tenerijfensis is better considered as a synonym of D. macra. Dysdera minutissima Wunderlich 1991 Figs. 137-147, Table 12 Dysdera minutissima Wunderlich 1991: 299-300, fig. 61-62 [(3]. Holotype male from Aguamansa, La Orotava, Tenerife; 5 March 1987, H. Enghoff leg.; num. 2676; Stored at ZMK. Examined. - Wunderlich 1991: 284-287. Diagnosis. — Dysdera minutissima is distin- guished from most of the other Canarian Dys- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 645 Table 12. — Intraspecific spination variability of Dysdera minutissima. Proximal Medial- proximal Medial- distal Distal Tibia 3 dorsal 1.0.0 0 0 1.0.0 Tibia 4 dorsal O.O.O-l 0 0 O.O.O-l Tibia 3 ventral 0 0 0 0 Tibia 4 ventral 0.0- 1.0 0 0 0 dera species by its small size (carapace ~ 2.00 mm long) and markedly wrinkled cara- pace. In both sexes, it differs from the similar Tenerifean D. levipes and D. gollumi by hav- ing chelicerae with a granular surface, shorter cheliceral groove and spinated posterior legs. In males, the bulb distal division (DD) is lon- ger than the tegulum (T) (Fig. 139) and the lateral sheet (L) has a distinct shape (Fig. 142). Males are distinguished from Grancan- arian D. andamanae Arnedo & Ribera 1997 by absence of a lateral fold (LF) in bulb. In both sexes, it differs from D. paucispinosa Wunderlich 1991 from Gran Canaria and D. orahan Arnedo, Oromi & Ribera 1996 from La Gomera by possessing cheliceral granula- tion. Description. — Holotype male: Figs. 137-- 139, 142-145. Carapace (Fig. 137) 1.49 mm long; maximum width 1.14 mm; minimum Figures 137-141. — Dysdera minutissima. 137, Carapace, dorsal; 138, Left chelicera, ventral; 139, Left male bulb, external; 140, Vulva, dorsal; 141, Vulva, ventral. Scale bars in mm. width 0.74 mm. Dark red, darkened at bor- ders; heavily wrinkled, foveate, covered with tiny granulations; hairy, covered with black hairs mainly at lateral and back borders. Fron- tal border roughly triangular, markedly small- er than Vi carapace length; anterior lateral bor- ders divergent; rounded at maximum dorsal width point, back lateral borders straight; back margin narrow, straight. AME diameter 0.11 mm; PLE 0.09 mm; PME 0.09 mm; AME on edge of frontal border, separated one from an- other about Vi of diameter, close to PLE; PME very close to each other, about Vs PME di- ameter from PLE. Labium trapezoid-shaped, base wider than distal part; longer than wide at base (triangle-like); semicircular groove at tip. Sternum dark red, uniformly distributed; heavily wrinkled; uniformly covered in slen- der black hairs. Chelicerae (Fig. 138) 0.53 mm long, about Va of carapace length in dorsal view; fang short, 0.32 mm; basal segment dorsal, ventral side completely covered with large piligerous granulations. Chelicera inner groove short, about V3, cheliceral length; armed with three teeth and lamina at base; D = M = B; D trapezoid, located roughly at center of groove; B close to basal lamina; M close to B. Legs pale yellow, darkened frontal, proximally. Lengths of male described above: fel 1.26 mm (all measurements in mm); pal 0.77; til 1.01; mel 0.97; tal 0.35; total 4.36; fe2 1.13; pa2 0.77; ti2 0.9; me2 0.9; ta2 0.32; total 4.02; fe3 0.97; pa3 0.46; ti3 0.61; me3 0.83; ta3 0.3; total 3.17; fe4 1.22; pa4 0.63; ti4 0.99; me4 1.19; ta4 0.37; total 4.33; relative length: 1 = 4 > 2 > 3; palp: fe 0.79; pa 0.37; ti 0.39; ta 0.42; total 1.97. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0.0; distal 1.0.0; ti3 ventral spines spineless; with two terminal spines. Fe4 dorsal spineless; ti4 dorsal spines arranged in two bands: proximal 0.0.1; distal 0.0.1; ti4 ventral spines arranged 646 THE JOURNAL OF ARACHNOLOGY Figures 142-147. — Dysdera minutissima, right male bulb and female spinnerets. 142, DD frontal; 143, DD external; 144, DD posterior; 145, P external; 146, Right ALS; 147, Right PLS. in one band: proximal 0.0- 1.0; with two ter- minal spines. Dorsal side of frontal legs cov- ered with a fine-textured, piligerous, granular surface; ventral side of palp covered with hairs, lacking a granular surface. Claws with 10-14 teeth, slender, length twice claw width. Abdomen 1.77 mm long; whitish; globular. Abdominal dorsal hairs 0.036-0.045 mm long; thin, curved (?), compressed (?), point- ed; uniformly, thickly distributed. Male bulb: (Fig. 139). T slightly longer than DD; external, internal distal border ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 647 Table 13. — Intraspecific spination variability of Dysdera montanetensis. Proximal Medial-proximal Medial-distal Distal Tibia 3 dorsal LL2.0-1 0.1-2.0-1 0 1.0.0- 1 Tibia 4 dorsal O-l.O-Ll 1. 1-2.1 0 Ll-2.1 Tibia 3 ventral 1.2.0- 1 0 0 1.0- 1.0-1 Tibia 4 ventral 1.1.1 0-1.0- 1.0-1 0.0-2.0-1 LO-Ll Number of rows Number of spines Femur 3 dorsal 1 0-1 Femur 4 dorsal 1 1-4 sloped backwards. DD bent about 45° in lat- eral view; internal distal border not expanded. ES wider, more sclerotized than IS; IS contin- uous to tip (?). DD tip (Figs. 142-144) straight in lateral view. C present, short; distal end on DD internal tip; well-developed; lo- cated far from DD distal tip; proximal border continuously decreasing; distal border sloping in its base, upper tip projected, rounded; ex- ternal side smooth. AC absent. LF absent. L well-developed; external border not sclero- tized, not folded; distal border divergent, con- tinuous. AL absent. P (Fig. 145) fused to T; markedly sloped on its proximal part, perpen- dicular on distal; lateral length from of T width; ridge present, perpendicular to T, not expanded; upper margin markedly toothed, on its distal part, very few teeth (1-3); not dis- tally projected; back margin not folded. Female: (from Barranco del Pino, Santa Ur- sula, Tenerife; num. 2614, MCNT), Figs. 140-141, 146, 147. All characters as in male except: Carapace 1.68 mm long; maximum width 1.26 mm; minimum width 0.74 mm. Dark brownish-red. AME diameter 0.11 mm; PLE 0.09 mm; PME 0.08 mm; AME separat- ed one from another about % of diameter. Ster- num brownish-red; heavily wrinkled. Chelicerae 0.53 mm long; fang 0.37 mm. Leg lengths of female described above: fel 1.24 mm (all measurements in mm); pal 0.7; til 1.04; mel 0.99; tal 0.35; total 4.32; fe2 1.15; pa2 0.77; ti2 0.9; me2 95; ta2 0.32; total 4.09; fe3 1.01; pa3 0.51; ti3 0.65; me3 0.86; ta3 0.3; total 3.33; fe4 1.35; pa4 0.65; ti4 1.08; me4 1.28; ta4 0.37; total 4.73; relative length 4 > 1 > 2 > 3; palp: fe 0.65; pa 0.37; ti 0.37; ta 0.48; total 1.87. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0.0; distal 1.0.0; ti3 ventral spines spineless; with one terminal spine on anterior margin. Fe4 dorsal spineless; ti4 dorsal spines arranged in two bands: proximal 0.0.1; distal 0.0.1; ti4 ventral spines arranged in one band: proximal 0.1 -0.0; with two terminal spines. Abdomen 1.96 mm long; whitish; globular. Abdominal dorsal hairs 0.054-0.063 mm long; thin, curved, compressed, pointed; uni- formly, thickly distributed. Vulva (Figs. 140, 141) arch-like in dorsal view, frontally point- ed; as wide as long; DF wide. VA frontal re- gion completely sclerotized. S attachment pro- jected under VA; arms as long as DA, clearly curved; tips dorsally projected; neck as wide as arms. TB usual shape. ALS (Figs. 146, 147) with PS; remaining piriform spigots more external than MS, arranged in one row; 4+1 piriform gland spigots; PMS, PLS with 5-10 aciniform gland spigots. Intraspecific variation. — Male cephalo- thorax ranges in length from 1.49-1.54 mm. AME separation from diameter. PLE- PME from V3~% diameter apart. Cheliceral teeth very similar in size. D > B > M. Spi- nation variability in Table 12. Additional material examined. — TENERIFE: Santa Ursula: Barranco del Pino, 21-28 July 1985, IS (J.M. Peraza, num. 2615 MCNT). Distribution. — ^Tenerifean endemic. Known from two localities on middle-northern slope of the island. Comments. — Female specimens of this species were formerly unknown. Unfortunate- ly, during manipulation of the only available vulva it was lost. The character states reported for the vulva in the present work are based on preliminary drawings made before its loss. Dysdera montanetensis Wunderlich 1991 Figs. 148-159, Table 13 Dysdera montanetensis Wunderlich 1991: 300-301, fig. 63-64 [(3]. Holotype male from La Monta- 648 THE JOURNAL OF ARACHNOLOGY neta, MSS 6-9, Garachico, Tenerife; 26 April 1988, A.L. Medina leg.; num. T-64-17; Stored at UL. Examined. -Wunderlich 1991: 284-287. Diagnosis. — Dysdera montanetensis males are distinguished from most other Canarian species by a distinctly expanded, unscleroti- zed upper margin of posterior apophysis (P) (also present in D. gibbifera and D. volcania) (Fig. 150) in the male bulb. Females are dis- tinguishable by the presence of ridges on vul- va ventral arch (VA) (Figs. 152, 153) (also present in D. labradaensis and D. iguanensis). Dysdera montanetensis differs from D. gib- bifera by its smaller size, markedly spinated legs and, in males, by lack of lateral sheet (L) sclerotization (Fig. 155). Males and females are distinguished from D. volcania by pos- sessing smooth carapace and markedly spi- nated legs, from D. iguanensis by more distal location of cheliceral distal tooth (Fig. 149) and distinct spination pattern (Table 13) and from D. labradaensis by smaller size and lack of eye reduction. Description. — Holotype male: Figs. 148- 150, 154-157. Carapace (Fig. 148) 2.98 mm long; maximum width 2.35 mm; minimum width 1.37 mm. Brownish-orange, uniformly distributed; slightly foveate at borders, wrin- kled at middle, covered with tiny granulations. Frontal border roughly round, markedly smaller than Vi carapace length; anterior lat- eral borders slightly divergent or parallel; rounded at maximum dorsal width point, back lateral borders straight; back margin wide, straight. AME diameter 0.18 mm; PLE 0.18 mm; PME 0.12 mm; AME on edge of frontal border, separated one from another about Vi diameter, close to PLE; PME very close to each other, about V3 PME diameter from PLE. Labium trapezoid- shaped, base wider than distal part; longer than wide at base (rectan- gle-like); semicircular groove at tip. Sternum brownish-orange, frontally darker, becoming lighter towards back; wrinkled; covered in hairs mainly on margin. Chelicerae (Fig. 149) 1.21 mm long, about Vs of carapace length in dorsal view; fang short, 0.84 mm; basal segment dorsal, ventral side completely covered with large piligerous gran- ulations. Chelicera inner groove short, about V3 cheliceral length; armed with three teeth and lamina at base; D > B > M (large, not very different); D trapezoid, located near segment Figures 148-153. — Dysdera montanetensis. 148, Carapace, dorsal; 149, Right chelicera, ventral; 150, Right male bulb, internal; 151, Vulva, dorsal; 152, Vulva, ventral; 153, Vulva, lateral. Scale bars in mm. tip; B close to basal lamina; M at middle of B and D. Legs yellow, frontal slightly darker. Lengths of male described above: fel 2.75 mm (all measurements in mm); pal 1.72; til 2.42; mel 2.47; tal 0.74; total 10.1; fe2 2.7; pa2 1.63; ti2 2.28; me2 2.42; ta2 0.74; total 9.77; fe3 2.28; pa3 1.21; ti3 1.64; me3 2.16; ta3 0.7; total 7.98; fe4 3.17; pa4 1.54; ti4 2.47; me4 3.17; ta4 0.79; total 11.09; relative length: 4 > 1 > 2 > 3; palp: fe 1.35; pa 0.74; ti 0.93; ta 0.84; total 3.86. Spination: palp, legl, leg2 spineless. Fe3 dorsal spines in one row: 1; ti3 dorsal spines arranged in three bands: proximal 1.1.1; medial-proximal 0.2.0- 1; distal 1. 0.0-1; ti3 ventral spines arranged in two bands: prox- imal 1.2. 1-0; distal 1. 0.0-1; with two terminal spines. Fe4 dorsal spines in one row: 2-3; ti4 dorsal spines arranged in three bands: proximal 0-1. 1.1; medial-proximal 1. 1-2.1; distal 1.2.1; ti4 ventral spines arranged in four bands: prox- imal 1.1.1; medial-proximal 1-0. 1.0; medial- distal 0.1.0; distal 1.1.1; with two terminal spines. Dorsal side of frontal legs, ventral side of palp with a fine-textured, piligerous, gran- ular surface. Claws with more than 15 teeth, slender, length twice claw width. Abdomen 3.03 mm long; whitish; cylindri- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 649 Figures 154-159. — Dysdera montanetensis, right male bulb and female spinnerets. 154, DD frontal; 155, DD internal; 156, DD posterior; 157, P internal; 158, Right ALS; 159, Right PLS. cal. Abdominal dorsal hairs 0.072 mm long; thick, slightly curved, not compressed, blunt, tip not enlarged; uniformly, scantly distribut- ed. Male bulb (Fig. 150): T slightly longer than DD; external, internal distal border sloped backwards. DD bent about 45° in lateral view; internal distal border not expanded. IS and ES equally developed; IS continuous to tip (?). DD tip (Figs. 154 — 156) straight in lateral view. C present, short; distal end on DD in- ternal tip; poorly developed; located far from 650 THE JOURNAL OF ARACHNOLOGY DD distal tip; proximal border continuously decreasing; distal border sloping in its base, upper tip not projected, pointed; external side hollowed. AC absent. LF present; distally pro- jected; well-developed. L well-developed; ex- ternal border not sclerotized, laterally mark- edly folded; distal border divergent, continuous. AL absent. P (Fig. 157) fused to T; markedly sloped on its proximal part, per- pendicular on distal; lateral length as long as or longer than T width; ridge present, not sclerotized, perpendicular to T, distinctly ex- panded, rounded; upper margin smooth; not distally projected; back margin not folded. Female: (from Cueva Labrada, El Sauzal, Tenerife; num. 2519, UL; Figs. 151-153, 158, 159.) All characters as in male except: Cara- pace 3.35 mm long; maximum width 2.57 mm; minimum width 1.58 mm. Anterior lat- eral borders divergent. AME diameter 0.21 mm; PLE 0.2 mm; PME 0.14 mm. Chelicerae 1.35 mm long, fang 0.93 mm. Leg lengths of female described above: fel 2.93 mm (all measurements in mm); pal 1.91; til 2.56; mel 2.56; tal 0.74; total 10.7; fe2 2.89; pa2 1.77; ti2 2.42; me2 2.56; ta2 0.79; total 10.43; fe3 2.28; pa3 1.4; ti3 2.16; me3 2.33; ta3 0.79; total 8.89; fe4 3.54; pa4 1.77; ti4 2.76; me4 3.54; ta4 0.98; total 12.59; rel- ative length 4 > 1 > 2 > 3; palp: fe 1.54; pa 0.84; ti 0.74; ta 1.16; total 4.28. Spination: palp, legl, leg2 spineless. Fe3 dorsal spines in one row: 1; ti3 dorsal spines arranged in three bands: proximal 1.1. 1-0; medial-proximal 0.1- 2.0- 1; distal 1. 0.0-1; ti3 ventral spines ar- ranged in two bands: proximal 1.2.0; distal 1.0- 1.1-0; with two terminal spines. Fe4 dor- sal spines in one row: 1-3; ti4 dorsal spines arranged in three bands: proximal 0-1.0. 1; me- dial-proximal 1.2.1; distal 1.2.1; ti4 ventral spines arranged in three bands; proximal 1.1.1; medial-proximal 0; medial-distal 0.1- 2.0; distal 1.0.1; with two terminal spines. Abdomen 6.52 mm long; whitish; cylindri- cal. Abdominal dorsal hairs 0.09 mm long; medium thickness, slightly curved, not com- pressed, blunt, tip not enlarged; uniformly, scantly distributed. Vulva (Figs. 151-153) rectangle-like in dorsal view, frontally round- ed; slightly wider than long; DF wide, MF well-developed; sclerotized along its extent. VA frontal region completely sclerotized; pos- terior region sclerotized except for most in- ternal area; sclerotized ridge at ventral VA ex- ternal margin, as long as VA. AVD clearly recognizable. S attachment projected under VA; arms as long as DA, straight; tips dor- sally projected; neck as wide as arms. TB usu- al shape. ALS (Figs. 158, 159) with PS; re- maining piriform spigots no more external than MS, arranged in three rows; 10+1 pir- iform gland spigots; PMS, PLS with 5-10 aciniform gland spigots. letraspecific variation.-— Male cephalo- thorax ranges in length from 2.98“3.07 mm., female from 2.93=4.00 mm. AME separation from %=% of diameter. PLE-PME from Vs-Vi diameter apart. D markedly larger than or as large as B. Usually, teeth large, not markedly different. In some females (Teno, Labrada) ab- domen hairs are compressed and pointed. Vul- va as wide as long. Labrada female specimen #2516 shows carapace frontal lateral margins parallel, long. Carapace, sternum ornamenta- tion nearly smooth. Strong reduction in eye size. D at center of the chelicera groove. Re- duction in leg pigmentation, spination: ab- sence of fe spination and ti medial spination. Spination variation given in Table 13. Additional material examined. — TENERIFEi Ei Rosario: Las Raices, ? November 1993, 1 ? (Ar- nedo & Ribera, num. 4795 UB). La Orotava: Agua- mansa, MSS, 4 August 1985, 16 (J.L. Martin & A. Machado, num. 2580 UL). El Sauzal: Cueva La- brada, 4 November 1991, 19, (J.L. Martin, num. 2516 UB). Los Silos: Monte del Agua, 24 February 1997, 19 (N. Zurita, num. 3209 UB). Vilaflor: Fuente de Mesa, 9 March 1984, 19 (J.M. Peraza, num. 2770 UL). Distributioe.^Tenerifean endemic. Known from several localities spread throughout the northern slope excepting Anaga massif, and from a single locality at middle- southern slope. Commeets.=-“Former knowledge of this species was restricted to a single male speci- men. Dysdera propinqua Ribera, Ferrandez & Blasco 1985 Figs. 160=172, Table 14 Dysdera propinqua Ribera, Ferrandez & Blasco 1985: 61-63, fig. 4A-D [6]. Holotype male from Cueva Honda, Giiimar, Tenerife; 15 December 1982, J.L. Martin leg.; num. T-CH-14; Stored at UL. Examined. -Wunderlich 1991: 284-287. Ex- amined. D. nesiotes: Simon 1907: 260 (9, non 6). Simon ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 651 Table 14, — Intraspecific spination variability of Dysdera propinqua. Proximal Medial- proximal Medial- distal Distal Tibia 3 dorsal 1.0- 1.0-1 0 0 1.0.0 Tibia 4 dorsal 0- 1.0.1 0 0 0.0.1 Tibia 3 ventral 0.0- 1.0 0 0 0 Tibia 4 ventral 0.0- 1.0 0 0 0 Number of rows Number of spines Femur 3 dorsal 0 0 Femur 4 dorsal 2 0- 1/1-4 1883: 297 (D. insulana 9, non c3) 3 type females, unknown locality, Canary Islands; unknown data, M. Vemeau leg.; num. B=536; Stored at MNHN. Examined. Incorrect identification. D. obscuripes Wunderlich 1991: 302-303, figs. 72- 76 [c3,9]. Holotype male from pine forest close to La Orotava, La Orotava, Tenerife, Canary Is- lands; in IL, M. Knosel leg.; Stored at SMF. Ho- lotype not examined, paratypes examined. New synonymy. Diagnosis. — Dysdera propinqua is distin- guished, in both sexes, from most other Can- arian species by a combination of markedly Figures 160-165. — Dysdera propinqua. 160, Carapace, dorsal; 161, Left chelicera, ventral; 162, Right male bulb, external; 163, Vulva, dorsal; 164, Vulva, ventral; 165, Vulva, lateral. Scale bars in mm. foveate carapace, convergent anterior lateral carapace borders and poorly spinated posterior legs. Males and females differ from the sim- ilar and sympatric D. cribellata, by the basal cheliceral teeth not being the largest and pres- ence of cheliceral granulation. Dysdera pro- pinqua males differ from D. cribellata by lacking a fold in the lateral sheet (L) of the bulb, and in females by presence of tooth- shaped expansions in the vulval ventral arch (VA) (Figs. 164, 165). Description. — -Holotype male: Figs. 160- 162, 166-170. Carapace (Fig. 160) 4.1 mm long; maximum width 3.4 mm; minimum width 2.17 mm. Dark red, darkened at bor- ders; foveate at borders, slightly wrinkled at middle, with a black, fine-textured, granular surface; hairy, covered with black hairs main- ly at lateral and back borders. Frontal border roughly triangular, from V2-% carapace length; anterior lateral borders convergent; rounded at maximum dorsal width point, back lateral bor- ders straight; back margin wide, straight. AME diameter 0.27 mm; PLE 0.25 mm; PME 0.18 mm; AME on edge of frontal border, sep- arated one from another about % of diameter, close to PLE; PME very close to each other, about Vs PME diameter from PLE. Labium trapezoid-shaped, base wider than distal part; longer than wide at base (triangle-like); semi- circular groove at tip. Sternum dark red, dark- ened on borders; mostly wrinkled, except in middle part; uniformly covered in slender black hairs. Chelicerae (Fig. 161) 2.1 mm long, about % of carapace length in dorsal view; fang me- dium-sized, 1.47 mm; basal segment dorsal side completely covered with piligerous gran- ulations, ventral side smooth (spacing distally reduced). Chelicera inner groove short, about 652 THE JOURNAL OF ARACHNOLOGY Vs cheliceral length; armed with three teeth and lamina at base; D = B > M; D round, located roughly at center of groove; B close to basal lamina; M close to B. Legs orange. Lengths of male described above: fel 3.92 mm (all measurements in mm); pal 2.65; til 3.54; mel 3.41; tal 0.76; total 14.28; fe2 3.36; pa2 2.4; ti2 2.96; me2 3.16; ta2 0.76; total 12.64; fe3 2.6; pa3 1.49; ti3 1.89; me3 2.4; ta3 0.58; total 8.96; fe4 3.46; pa4 1.89; ti4 2.76; me4 3.21; ta4 0.83; total 12.15; relative length: 1 > 2 > 4 > 3; palp: fe 1.25; pa 1.06; ti 1.14; ta 1.08; total 4.53. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0.1; distal 1. 0.0-1; ti3 ventral 2 ternriinal spines. Fe4 dorsal spines in two rows: anterior 1-0; posterior 4-3; ti4 dorsal spines arranged in two bands: proximal 0.0.1; distal 0.0.1; ti4 ventral 2 terminal spines. Dorsal side of fron- tal legs with a piligerous, fine-textured, gran- ular surface; ventral side of palp covered with hairs, without a granular surface; very long hairs on back legs as well as on palps. Claws with 8 teeth or less, robust, hardly larger than claw width. Abdomen 6.02 mm long; cream-colored; cylindrical. Abdominal dorsal hairs 0.144 mm long; thick, roughly straight, compressed, lan- ceolate, frontally curved; uniformly, thickly distributed. Male bulb (Fig. 162): T slightly smaller than DD; external, internal distal border sloped backwards. DD slightly bent in lateral view, clearly less than 45°; internal distal bor- der not expanded. IS and ES equally devel- oped; IS truncated at DD middle part. DD tip (Figs. 166-169) straight in lateral view. C pre- sent, short; distal end on DD internal tip; well- developed; located close to DD distal tip; proximal border sharply decreasing; distal border stepped, upper tip not projected, round- ed; external side hollowed. L well-developed; external border not sclerotized, distally mark- edly folded; distal border divergent, discontin- uous but without a fold at the middle. AL pre- sent, hardly visible except for a small notch; proximal border in posterior view fused with DH except for its most internal part. P (Fig. 170) fused to T; perpendicular to T in lateral view; lateral length from of T width; ridge present, perpendicular to T, not expand- ed; upper margin markedly toothed, along its extent, few teeth; not distally projected; back margin not folded. Female: (from Barranco de Badajoz, Gui- mar, Tenerife; num. 3192, UB; Figs. 163-165, 171, 172.) All characters as in male except: Carapace 4.34 mm long; maximum width 3.57 mm; minimum width 2.52 mm. AME diame- ter 0.23 mm; PEE 0.21 mm; PME 0.16 mm; PME % diameter from PLE. Sternum brown- ish-red. Chelicerae 2.03 mm long; fang 1.33 mm. Leg lengths of female described above: fel 3.43 mm (all measurements in mm); pal 2.31; til 2.87; mel 2.87; tal 0.66; total 12.14; fe2 3.22; pa2 2.17; ti2 2.52; me2 2.73; ta2 0.63; total 11.27; fe3 2.45; pa3 1.47; ti3 1.82; me3 2.31; ta3 0.63; total 8.68; fe4 3.57; pa4 1.89; ti4 2.73; me4 2.15; ta4 0.77; total 12.11; rel- ative length 1 > 4 > 2 > 3; palp: fe 2.03; pa 1.05; ti 0.84; ta 1.12; total 5.04, Spination: palp, legl, leg2 spineless. Fe3 dorsal spine- less; ti3 dorsal spines arranged in two bands: proximal 1. 1-0.1; distal 1.0.0; ti3 ventral 2 ter- minal spines. Fe4 dorsal spines in two rows: anterior 1; posterior 3; ti4 dorsal spines ar- ranged in two bands: proximal 0.0.1; distal 0.0.1; ti4 ventral 2 terminal spines. Abdomen 6.23 mm long; cream-colored; cylindrical. Abdominal dorsal hairs 0.153 mm long; thick, roughly straight, compressed, lan- ceolate, frontally curved; uniformly, thickly distributed. Vulva (Figs. 163-165) arch-like in dorsal view, frontally pointed; slightly wider than long; DF wide. MF poorly developed. VA frontal region completely sclerotized; pos- terior region sclerotized at anterior area; tooth- shaped expansion from internal back border, not joined to lateral sclerotization, slightly shorter than DF lateral margins, markedly bent towards lateral area. AVD hardly visible. S attachment not projected under VA; arms as long as DA, clearly curved; tips dorsally pro- jected; neck as wide as arms. TB usual shape. ALS (Figs. 171-172) with PS; remaining pir- iform spigots more external than MS, ar- ranged in three rows; 9+1 piriform gland spigots; PMS, PLS with 10-15 aciniform gland spigots. Intraspecific variation.— Male cephalo- thorax ranges in length from 3.54-4.43 mm, female from 3.77-4,69 mm. Sternum orna- mentation variable, from hardly wrinkled be- tween coxae to completely wrinkled. Chelic- era lacking dorsal distal granulations or ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 653 Figures 166-172. — Dysdera propinqua, right male bulb and female spinnerets. 166, DD frontal, LF absent; 167, DD, frontal, LF present; 168, DD external; 169, DD posterior; 170, P external; 171, Right ALS; 172, Right PLS. somewhat reduced. P of T width. One spec- imen from Teno with well-developed LF (Fig. 167). Vulva frontally round. Ventral tooth- shaped sclerotization shorter, about Vs of DF lateral length. Spination variation given in Ta- ble 14. Additional material examined. — TENERIFE: ?: ?; 24 February 1984, Id (N.R Ashmole, num. 2728 UL). Fuente de Mesa, 9 March 1984, 2? (J.M. Peraza, num. 2770 UL). Adeje: Roque del Conde, 16 March 1996, Id (Oromi, num. 3123 UB); 19 (Oromi, num. 3124 UB). Arafo: Fuente 654 THE JOURNAL OF ARACHNOLOGY del Joco, 5km NW from Arafo, 1930 m, 28 Decem- ber 1994, 2$ (E Gasparo, EG). Arico: Barranco del Rio, 22 February 1984, ljuv. (R Ashmole, num. 2685 UL); 15-22 October 1984, 16 (J.M. Peraza, num. 2772 UL); 16-23 November 1984, 1619 (J.M. Peraza, num. 2721 UL); 17-26 September 1985, 1619 (J.M. Peraza, num. 2613 MCNT). Buenavista: Barranco de las Cuevas, 4 February 1989, 19 (H. Enghoff, num. 2642 ZMK). Casa Blanca close to W Buenavista, 4 February 1989, 1 9 (H. Enghoff, num. 2650 ZMK). El Rosario: Las Raices, November 1993, 1(3 (Amedo & Ribera, num. 4796 (T17) UB). Granadilla: Madre del Agua, November 1993, 1 9 (Amedo & Fluhr, num. 4797 (T22) UB); ljuv. (Amedo & Fluhr, num. 4798 (T42) UB); lc3 (Amedo & Fluhr, num. 4813 (T4) UB). Gmmar: Barranco de Badajoz, 18 December 1996, 49, (P. Oronu, num. 3210 UB). La Laguna: Bajamar, 10 September 1985, lc3 (J.M. Peraza, num. AR-202 MCNT). El Moquinal, 23 January 1997, 1639 (P. Oromf, num. 3196 UB). Monte de las Mercedes, 18 March 1990, 16 (C.G. Campos, num. 2726 UL). La Orotava: Base zig-zag, 17 Oc- tober 1984, 1(3 (C.G. Campos, num. 2696 UL). El Guanche close to Aguamansa, 5 March 1987, 1(3 (H. Enghoff, num. 2649 ZMK). Hierba Pajonera, 2050 m, 19 June 1984, 16 (C.G. Campos, num. 2724 UL). Izana, 13 March 1987, 19 (H. Enghoff, num. 2641 ZMK). Las Canadas del Teide, 19, (J. Wunderlich, num. 2629 JW); 1 March 1984, 16 (C.G. Campos, num. 2761 UL); 19 April 1984, 16 (C.G. Campos, num. 2763 UL); 2100 m, 18 June 1984, 19 (C.G. Campos, num. 2725 UL); 19 Sep- tember 1984, 1 9 (C.G. Campos, num. 2762 UL); 2050 m, 10 November 1984, 19, (C.G. Campos, num. 2722 UL); 29 June 1995, ljuv. (A. Camacho, num. 3159 UB); 2 May 1996, lc3 (N. Zurita, num. 3171 UB). Montana de Los Conejos, 2400 m, 18 June 1984, 1(3 (C.G. Campos, num. 2723 UL). Pico Viejo, 9 November 1983, 19 (C.G. Campos, num. 2695 UL). Retamar, 3050 m, 19 June 1984, ljuv. (C.G. Campos, num. 2765 UL). Teide, 2700 m, 19 June 1984, lc319 (C.G. Campos, num. 2713 UL). 3050 m, 18 September 1984, ljuv. (C.G. Campos, num. 2764 UL). Ucanca, 2100 m, 1 July 1990, 16 (C.G. Campos, num. 2745 UL). Santa Cruz de Ten- erife: Bailadero, November 1993, 16 (Amedo & Ribera, num. 4833 (T47) UB). Anaga, Cmz del Carmen, 12 May 1996, lc319 (M. Naranjo, num. 3146 UB; 1 9 (M. Naranjo, num. 3147 UB). Tagan- ana, 1 6 (Oromf, num. 2932 UB). Los Realejos: La Fortaleza, 25 Febmary 1983, lc3 (A. Fox, num. 2727 UL); 26 December 1984, \6 (C.G. Campos, num. 2699 UL. Pinar Roque Peral, 19 May 1984, 1(3 (C.G. Campos, num. 2692 UL); 18 October 1984, 1(3 (C.G. Campos, num. 2698 UL); 1(3, ljuv. (C.G. Campos, num. 2700 UL). Los Silos: Teno, Monte del Agua, 1 Febmary 1988, 1(31 9 (J.J. Nar- anjo, num. 2598 UL); 1(329 (P Oromf, num. 2683 UL); 1 March 1989, 29 (P. Oromf, num. 2684 UL); 30 November 1993, lc3 (M.A. Amedo, num. 3181 UB). Santa Ursula: Barranco del Pino, 15 Novem- ber 1984, 49 (J.M. Peraza, num. 2720 UL). Monte de Santa Ursula, 13 December 1996, 1 9 , (P. Oromf, num. 3191 UB). Vilaflor: El Pinalito, 20-27 De- cember 1984, 1(319, (J.M. Peraza, num. 2608 MCNT); 16-23 Febmary 1985, 19, (J.M. Peraza, num. 2607 MCNT); 1 (3 1 9 (J.M. Peraza, num. 2610 MCNT). Dysdera obscuripes: TENERIFE: Arico: Barranco del Rio, in I, 1(3 paratype, (Wunderlich, num. 2626 JW); 27 January 1985, 19 paratype, (Wunderlich, num. 2628 JW). La Laguna: El Mo- quinal, 20 April 1990, 1(3 (P. Oromf, num. 2620 SMF). La Orotava: Canadas del Teide, ?, Ic3 par- atype, (Wunderlich, num. 2622 JW). Distribution. — Tenerifean endemic. The most widespread species in Tenerife. It has been collected throughout the island, with the exception of the middle-northern slope. Comments. — Examination of several par- atypes of D. obscuripes showed that no di- agnostic feature exists when compared with D. propinqua holotype. The females used by Simon in the original description of D. insu- lana were also available for study. The author himself (Simon 1907) transferred these fe- males to D. nesiotes Simon 1907. The study of the specimens revealed that they were nei- ther D. insulana nor D. nesiotes, while they perfectly fit those characters of D. propinqua. The male type specimens of D. nesiotes are the only known specimens of this species so far. Dysdera unguimmanis Ribera, Ferrandez & Blasco 1985 Figs. 173-177, 187-189, Table 15 Dysdera unguimmanis Ribera, Ferrandez & Blasco 1985: 57-59, fig. 2A-E [9]. Holotype female from Cueva del Viento-Sobrado, Icod de los Vi- nos, Tenerife, Canary Islands; 10 February 1982, J.L. Martfn leg.; num. T-CV-121; Stored at UL. Examined. -Wunderlich 1991: 284-286. Diagnosis. — Dysdera unguimmanis is dis- tinguished from all other Dysdera species by absence of eyes, remarkable elongation of ap- pendages, reduction of body pigmentation and uniquely large tarsal claws (Fig. 187). Description. — Holotype female: Figs. 173- 177. Carapace (Fig. 173) 2.73 mm long; max- imum width 2.02 mm; minimum width 1.12 mm. Pale yellow, uniformly distributed; smooth with some small black granular tex- ture mainly at front; hairy, covered with black ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 655 Table 15. — Intraspecific spination variability of Dysdera unguimmanL Proximal Medial- proximal Medial- distal Distal Tibia 3 dorsal 0- 1.0.0 0 0 0-1. 0.0-1 Tibia 4 dorsal 1.0.1 0 0 1.0.1 Tibia 3 ventral 0 0 0 0-1. 0.0 Tibia 4 ventral 0 0 0 0-1. 0.0-1 hairs mainly at lateral and back borders. Fron- tal border roughly round, markedly smaller than V2 carapace length; anterior lateral bor- ders parallel, long; sharpened at maximum dorsal width point, back lateral borders straight; back margin wide, straight. Eyeless. Labium trapezoid- shaped, base wider than distal part; as long as wide at base (square- like); semicircular groove at tip. Sternum yel- low, uniformly distributed; smooth; covered in hairs mainly on margin. Chelicerae (Fig. 174) 1.23 mm long, about Vs of carapace length in dorsal view; fang me- dium-sized, 0.93 mm; basal segment proximal dorsal side scantly covered with piligerous granulations at internal margin. Chelicera in- ner groove medium-size, about % cheliceral 174 173 Figures 173-177. — Dysdera unguimmanis. 173, Carapace, dorsal; 174, Left chelicera, ventral; 175, Vulva, dorsal; 176, Vulva, ventral; 177, Vulva, lat- eral. Scale bars in mm. length; armed with three teeth and lamina at base; B > D > M (M, D small); D triangular, located roughly at center of groove; B close to basal lamina; M at middle of B and D. Legs whitish. Lengths of female described above: fel 3.79 mm (all measurements in mm); pal 1.64; til 3.41; mel 3.29; tal 1.01; total 13.14; fe2 3.84; pa2 1.64; ti2 3.54; me2 3.03; ta2 1.01; total 13.06; fe3 3.29; pa3 1.39; ti3 2.53; me3 3.03; ta3 1.01; total 11.25; fe4 3.92; pa4 1.52; ti4 3.11; me4 3.41; ta4 1.01; total 12.97; relative length: 1 > 2 > 4 > 3; palp: fe 1.14; pa 0.76; ti 0.88; ta 1.31; total 4.09. Spination: palp, legl, leg2 spineless. Fe3 dorsal spineless; ti3 dorsal spines arranged in two bands: proximal 1.0.0; distal 1.0.1; ti3 ventral 2 terminal spines. Fe4 dorsal spineless; ti4 dorsal spines arranged in two bands; proximal 1.0.1; distal 1.0.1; ti4 ventral 2 terminal spines. Dorsal side of frontal legs smooth; ventral side of palp smooth. Claws markedly bent; with 10-14 teeth, in two groups, slender, unusually long (Fig. 187). Abdomen 6.9 mm long; cream-colored; cy- lindrical. Abdominal dorsal hairs 0.072-0.09 mm long; thin, curved, compressed, pointed; uniformly, scantly distributed. Vulva (Figs. 175-177) rectangle-like (?); wider than long; DF wide. MF well-developed; sclerotized at frontal part. VA frontal region completely sclerotized; posterior region sclerotized at an- terior area; tooth-shaped expansion from back border; not joined to lateral sclerotization, about half of DF lateral margins. AVD clearlt recognizable. S attachment projected under VA; arms as long as DA, clearly curved; tips not projected; neck as wide as arms. TB usual shape. ALS (Figs. 188-189) with PS; remain- ing piriform spigots more external than MS, arranged in one row; 3+1 piriform gland spigots; PMS, PLS with fewer than 5 acini- form gland spigots. Male: Unknown. 656 THE JOURNAL OF ARACHNOLOGY Table 16. — Intraspecific spination variability of Dysdera volcania. Proximal Medial-proximal Medial-distal Distal Tibia 3 dorsal 1.0.1 0 0 1.0.0 Tibia 4 dorsal 0.0.1 1.0-1. 1 0 1.0.1 Tibia 3 ventral 1.1.0 0 0 1.0.0 Tibia 4 ventral 1.1.1 0 0 1.0.1 Intraspecific variation. — DA arch-like in dorsal view. Spination variability in Table 15. Additional material examined. — TENERIFE: Icod de los Vinos: Cueva de Felipe Reventon, 17 March 1984, ljuv. (J.J. Hernandez, num. 2584 UL); 10 September 1992, ljuv. (R Oromi, num. 2535 UB); November 1993, 1 9 (Amedo & Ribera, num. 4829 (T44) UB). Cueva del Viento-Sobrado, 10 September 1992, ljuv. (H. Enghoff, num. 2630 ZMK); 19 October 1994, ljuv. (Amedo & Ribera, num. 4822 (T33) UB); 9 June 1996, ljuv. (P. Oromi, num. 3174 UB. La Orotava: Cueva del Bu- cio, 30 October 1991, ljuv. (P. Oromi, num. 2540 UL). Distribution. — ^Tenerifean endemic. Known from several lava tubes located on the north- ern slope of the island. Dysdera volcania Ribera, Ferrandez & Blasco 1985 Figs. 178-186, Table 16 Dysdera volcania Ribera, Ferrandez & Blasco 1985: 59-61, fig. 3A-D [d] (cJ; non 9, incorrect identification). Holotype male from Cueva de Fe- lipe Reventon, Icod de los Vinos, Tenerife, Ca- nary Islands; 10 Febmary 1982, P. Oroim leg.; num. T-FR-106; Stored at UL. Examined. -Wun- derlich 1991: 284-287. Figures 178-180. — Dysdera volcania. 178, Car- apace, dorsal; 179, Left chelicera, ventral; 180, Left male bulb, external. Scale bars in mm. Diagnosis. — Dysdera volcania is distin- guished from other Canarian Dysdera, except D. montanetensis and D. gibbifera, by male bulb having distinctly expanded and unscler- otized upper margin of posterior apohysis (P) (Fig. 179). It differs from D. montanetensis and D. gibbifera by its markedly foveate car- apace. Description. — Holotype male: Figs. 178- 186. Carapace (Fig. 178) 3.5 mm long; max- imum width 2.5 mm; minimum width 1.44 mm. Dark red, darkened at borders; heavily wrinkled, foveate, covered with tiny granula- tions. Frontal border roughly round, markedly smaller than Vi carapace length; anterior lat- eral borders divergent; rounded at maximum dorsal width point, back lateral borders round- ed; back margin wide, straight. AME diameter 0.24 mm; PLE 0.21 mm; PME 0.15 mm; AME on edge of frontal border, separated one from another about Vi of diameter, close to PLE; PME about V4 of diameter apart, about Vs PME diameter from PLE. Labium trape- zoid-shaped, base wider than distal part; lon- ger than wide at base (rectangle-like); semi- circular groove at tip. Sternum dark red, uniformly distributed; wrinkled; covered in hairs mainly on margin. Chelicerae (Fig. 179) 1.30 mm long, about Vs of carapace length in dorsal view; fang short, 0.88 mm; basal segment dorsal, ventral side completely covered with large piligerous granulations. Chelicera inner groove short, about Vs cheliceral length; armed with three ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 657 teeth and lamina at base; D = Bmt M (or D slightly larger; D,B large); D trapezoid, locat- ed near segment tip; B close to basal lamina; M close to B. Legs orange. Lengths of male described above: fel 3.36 mm (all measure- ments in mm); pal 1.99; til 2.78; mel 2.78; tal 0.76; total 11.67; fe2 3.16; pa2 1.89; ti2 2.65; me2 2.78; ta2 0.76; total 11.24; fe3 2.65; pa3 1.26; ti3 1.89; me3 2.6; ta3 0.68; total 9.08; fe4 3.67; pa4 1.64; ti4 2.86; me4 3.59; ta4 0.83; total 12.59; relative length: 4 > 1 > 2 > 3; palp: fe 1.77; pa 0.93; ti 1.01; ta 1.06; total 4.77. Spination: palp, legl, leg2 spine- less. Fe3 dorsal spineless; ti3 dorsal spines ar- ranged in two bands; proximal 1.0.1; distal 1.0.0; ti3 ventral spines arranged in two bands: proximal 1.1.0; distal 1.0.0; with two terminal spines. Fe4 dorsal spineless; ti4 dor- sal spines arranged in three bands: proximal 0.0.0- 1; medial-proximal 1.1.1; distal 1.0.1; ti4 ventral spines arranged in two bands: prox- imal 1.1.1; distal 1.0.1; with two terminal spines. Dorsal side of front legs covered with a piligerous, fine-textured, granular surface; ventral side of palp scarcely covered with a piligerous, fine-textured granular surface. Claws with more than 20 teeth, slender, length twice claw width. Abdomen 4 mm long; whitish; cylindrical. Abdominal dorsal hairs 0.063 mm long; me- dium thickness, roughly straight, not com- pressed, blunt, tip not enlarged; uniformly, scantly distributed. Male bulb: (Fig. 180) T as long as DD; ex- ternal, internal distal border sloped back- wards. DD slightly bent in lateral view, rough- ly 45°; internal distal border not expanded. ES more sclerotized than IS; IS continuous to tip. DD tip (Figs. 181-183) straight in lateral view, C present, short; distal end on DD in- ternal tip; poorly developed; located far from DD distal tip; proximal border continuously decreasing; distal border sloping in its base, upper tip projected, pointed; external side smooth. AC absent. LF present; distally not projected; poorly developed. L well-devel- oped; external border not sclerotized, laterally markedly folded; distal border divergent, con- tinuous. AL present, hardly visible except for a small notch; proximal border in posterior view fused with DH. P (Fig. 184) fused to T; markedly sloped on its proximal part, perpen- dicular on distal; lateral length markedly lon- ger than T width; ridge present, not sclero- tized, perpendicular to T, distinctly expanded, rounded; upper margin smooth; not distally projected; back margin not folded. ALS (Figs. 185-186) with PS; remaining piriform spigots more external than MS, ar- ranged in two rows; 6+1 piriform gland spigots; PMS, PLS with 10-15 aciniform gland spigots. Female: Unknown. letraspecific variation. — Male cephalo- thorax ranges in length from 3.35-3.50 mm. AME separation from of diameter. PLE- PME from diameter apart. D markedly larger than B. Spination variability in Table 16. Additional material examined. — TENERIFE: Icod de los Vinos: Cueva de Felipe Reventon, 12 April 1986, 16 (A.L. Medina, num. 2714 UL). Distribution. — ^Tenerifean endemic. Known from a single locality, a lava tube located on middle-northern slope of the island. Comments.^Examination of the allotype female specimen used in the original descrip- tion of D. volcania, revealed that it actually corresponded to a D. cribellata female speci- men. Schmidt (1975) reported the presence of Dysdera rugichelis Simon 1907 in Tenerife, based on the study of a single male specimen. No other specimen belonging to this species has been documented afterwards, although this is a very abundant and widespread species in La Gomera and La Palma (Amedo et al. 1996). This record is considered to be doubt- ful, probably due to incorrect identification. This suggestion is further supported by the proved misidentification of another D. rugi- chelis specimen described by the same author as a new species (Schmidt 1981). DISCUSSION Even though a lot of new material has been available for the present study, several species remain poorly known. In 8 out of the 22 spe- cies discussed, one of the sexes is still un- known (D. chioensis, D. curvisetae, D. gib- bifera, D. gollumi, D. hernandezi, D. labradaensi, D. unguimmanis and D. volcan- ia). In addition, some species have been re- corded only once, or are known from a single locality. On the other hand, although several expeditions have been conducted with the main goal of collecting Dysdera specimens throughout the island, several island regions 658 THE JOURNAL OF ARACHNOLOGY Figures 181-189. — 181-186, Dysdera volcania, right male bulb and spinnerets. 181, DD frontal; 182, DD external; 183, DD posterior; 184, P internal; 185, Right ALS; 186, Right PMS (upper), PLS (lower). 187-189, Dysdera unguimmanis . 187 legl claws; 188 Right ALS; 189, Right PMS (upper), PLS (lower). remain undersampled or poorly known. In spite of incompleteness, present data permit limited discussion of ecological and distribu- tional patterns. As has been reported for other Canarian is- lands (Amedo et al. 1996; Arnedo & Ribera 1997), the level of insular endemism is ex- tremely high: 18 out of the 21 species docu- mented in Tenerife, roughly 85% of the spe- cies, are endemics. Three species are shared ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 659 with neighbor islands: two are found in Gran Canaria (D. iguanensis and D. insulana) and a third {D. levipes) is found in both La Go- mera and Gran Canaria. Distributional patterns are the result of both ecological factors and geological history. Hu= midity may be considered as the major eco- logical factor governing Dysdera distribution. Most of the species have been documented to occur on the northern slope of the islands from 400-1200 m, which is the most humid region of the island. This humid belt can be further extended to include localities on south- eastern-western slopes where the summit barely reaches 1200 m. Some locations, in spite of being on the dry southern slopes, are actually humid because they are close to near- ly permanent watercourses (Barranco del Rio) or correspond to the MSS (Barranco del Chor- rillo station). Nevertheless, there are some ex- ceptions to the rule, and some species have been reported to live in genuinely dry areas. The taxa D. macra and D. propinqua have widespread distributions that include humid northern locations as well as very dry places at the high-elevation environments from Las Canadas or at the southwestern slopes. The only species that has been documented exclu- sively from dry regions is D. guayota. Further investigation of possible physiological adap- tations of this species should be conducted in the future. The geological history of Tenerife is very complex, mainly because of the several vol- canic processes involved in its formation. Vul- canism in the Canaries, unlike other oceanic archipelagos (e.g., Hawaiian Islands), is a re- current process. Several studies, using K-Ar dating, have supplied a large number of data on the age determination of Tenerife (Anco- chea et al. 1990 and references herein). They provide a well-documented picture of the vol- canic evolution of the island. As recently as 2 Mya, Tenerife was split into three different is- lands that roughly corresponded to the pre- sent-day Anaga, Teno and Roque Conde mas- sifs (Fig. 3). These primitive islands originated in the late Miocene and after sev- eral volcanic pulses, volcanic activity ceased about 3. 5-4.5 Mya ago. Lava flows from a new volcanic cycle, about 1.9 Mya ago, con- nected the three massifs and formed Las Can- adas caldera. Volcanic activity in this area has been more or less continuous until historical ages. Finally, between 0.83 and 0.78 Mya the large ‘valleys’ of Giiimar and La Orotava were formed, probably due to a massive land- slide. Anaga, Teno and Roque massifs have been considered by several authors as refugial areas or sources of colonizers (Machado 1976; Cobolli Sbordoni et al. 1991; Oromi at al. 1991; Avanzati et al, 1994; Juan et al. 1996). This hypothesis is mainly based on (a) their original isolation (b) the absence of eruptions during the last 4,000,000 years and (c) an ex- tensive surface tranformation and habitat de- struction in the rest of the island, from 2 Mya ago until the present. There are many exam- ples of distributions from a wide array of taxa that apparently suit this suggested scenario. Tenerifean Dysdera provide additional cases which could fit the former hypothesis. Some species have been found exclusively in one of the massifs: D. insulana is known only from Anaga and closer localities, while D. gibbifera has been collected only in Teno massif and proximities. However, the remaining species have wider distributions. Whether these dis- tributions are the result of dispersal events from some of the mentioned massifs remains to be tested, especially by means of a phylo- geographic framework. Sympatry is another outstanding feature of Canarian Dysdera species in general, and Te- nerifean ones in particular. As many as four species have been collected in the same lo- cality. In our experience, it was not strange to find two specimens from different species un- der the same stone. More surprisingly, with a single exception (Cueva del CMo), all the lava tubes where troglobitic Dysdera species have been reported hold more than one species. Ob- viously, such a pattern can only be the result of strong ecological segregation. No close as- sociation between any Dysdera species and a particular plant community has been ob- served. In general, species distribution range over two or more different ecological zones. In addition, some species have been collected in areas where original forest has been dis- turbed by reforestation or introduction of alien plant species. The genus Dysdera has fre- quently been described as a specialist predator of woodlice (Cooke 1965), although a recent study on D. crocota prey-preference (Pollard et al, 1995) has shown that this species is bet- ter considered as a generalist predator. What- ever taxonomic prey-preference exists in Dys- 660 THE JOURNAL OF ARACHNOLOGY dera species, it is clear that this is strongly constrained from a morphological point of view by body and chelicera-fang size. Tenerife harbors both the largest (Z). labradaensis) and the smallest {D. minutissima) Dysdera species ever reported. In addition, there is a wide spectrum of chelicera-fang sizes and, in a less- er degree, of shapes. Experimental studies re- garding prey-preference segregation will con- stitute a promising field of investigation. Troglobitic species deserve further consider- ation. Seven Tenerifean species have been col- lected exclusively in lava tubes and show mor- phological evidence of adaptation to the hypogean environment. The cave-dweUing D. ra~ tonensis from La Pahna is the single case of trog- lomorphism in the Canaries outside Tenerife. Some of the ‘a priori’ troglomorphic characters held by these species include: eye reduction or loss, appendage elongation and depigmentation. However, these characters are unequally mani- fested by the different species. For instance, D. labradaensis and D. chioensis have eyes mark- edly reduced in size, all of them being present, while D. unguimmanis is completely eyeless. In general, the degree of troglomorphism in Can- arian species may be considered low, and in most cases it is restricted to eye reduction (Wunderhch 1993). In contrast, D. unguimmanis is one of the most troglomorphic taxa in the genus described to date. Apart from the absence of eyes, the no- ticeable leg elongation and nearly complete de- pigmentation, this species has an unusual devel- opment of the leg claws (unguis). This feature has been observed only in dysderid cave-dweU- ing species of the genera Stalita Schiodte 1847 and Folkia Kratochvie 1970 and has been con- sidered as a troglomorphic adaptation in coUem- bolans (Christiansen 1961) and cixiid planthop- pers (Howarth 1991). Even though hypogean Dysdera in Tenerife have been found exclusively in lava tubes, several observations suggest that more prob- ably Dysdera troglobites originated in, or are at least able to disperse through, the so-called mesocavernous shallow stratum (MSS) (Oromi et al. 1986). This hypothesis is sup- ported by the fact that (a) lava tubes have a geologically short time-span (between 0. 3-0.5 Mya), (b) two or more hypogean species usu- ally coexist in the same lava tube and (c) they have relatively wide distributional ranges, as shown by the distance between some of the lava tubes where they have been collected. Finally, considerations regarding morpho- logical affinities as well as inferences about speciation and adaptations to particular envi- ronments, i.e., troglomorphism, are avoided in this paper, since they are better discussed in the light of a cladogram for the species. ACKNOWLEDGMENTS This study would not have been possible without the use of material from museum and private collections, therefore, we would like to thank the following people for the loan of specimens: Dr. E. Enghoff (ZMK), R. Garcia Telo’, E Gasparo, Mr. P.D. Hillyard, Dr. M. Grasshoff, Dr. P. Oromf , Dr. G. Ortega, Dr. C. Rollard and Mr. J. Wunderlich. We are also grateful to people of the Serveis Cientifico Teenies of the Universitat de Barcelona for their help with the SEM work. We are deeply indebted to Pere Oromi for his help in field work and the generous sharing of his deep knowledge on Canarian fauna and flora. We would like to thank G. Giribet and P. Oromi for their useful comments and discussion. Ad- ditional thanks go to Andy Bohonak for help- ing with English corrections. The manuscript was greately improved by comments of J. Miller, P. Sierwald and J. Wunderlich. This re- search was supported by DGICYT PB93- 0811, Project 2192-PGC 94A of the Gener- alitat (Autonomous Government) of Catalonia and by both a FI grant from the Generalitat and a ‘Ajut per a la finalitzacio de la tesi doc- toral’ of the Universitat de Barcelona to M.A. LITERATURE CITED Ancochea, E., J.M. Fuster, E. Ibarrola, A. Cendrero, J. Coello, E Heman, J.M. Cantagrel & C. Ja- mond. 1990. Volcanic evolution of the island of Tenerife (Canary Islands) in the light of the new K-Ar data. J. Vol. Geotherm. Res., 44:231-249. Anguita, E & E Heman. 1975. A propagating frac- ture model versus a hot spot origin for the Canary Islands. Earth Planetary Sci. Lett., 27:11-19. Amedo, M.A., P. Oromf & C. Ribera. 1996. Ra- diation of the genus Dysdera (Araneae, Haplo- gynae, Dysderidae) in the Canary Islands: the Western Islands. Zool. Sen, 25:241-274. Amedo, M.A. & C. Ribera. 1996. Dysdera rato- nensis Wunderlich 1991 (Arachnida, Araneae) a troglomorphic species from La Palma, Canary Is- lands: Description of the male and redescription of the female. Rev. ArachnoL, 11:109-122. Amedo, M.A. & C. Ribera. 1997. Radiation of the genus Dysdera (Araneae, Haplogynae, Dysderi- ARNEDO & RIBERA— THE GENUS DYSDERA IN THE CANARY ISLANDS 661 dae) in the Canary Islands: The island of Gran Canaria. Zool. Scr., 26:205-243. Avanzati, A.M., M. Baratti & E Bernini. 1994. Mo- lecular and morphological differentiation between steganacarid mites (Acari: Oribatida) from the Ca- nary Islands. Biol. J. Linn. Soc., 52:325-340. Baum, D.A. 1992. Phylogenetic species concepts. TREE, 7(1): 1-2. Berland, L. 1936. Mission de M.A. Chevalier aux les du Cap Vert (1934). 1. Araignees. Rev. Fran- caise EntomoL, 3(l):67-88. Bosenberg, W. 1895. Beitrag zur Kenntnis der Ar- achniden-Fauna von Madeira und den Canarisch- en Inseln. Abh. Naturw. Ver. Hamburg, 13:1-13. Cantagrel, J.M., A. Cendrero, J.M. Fuster, E. Ibar- rola & C. Jamond. 1984. K-Ar chronology of the vulcanic eruptions in the Canarian archipel- ago: Island of La Gomera. Bull. Volcano!., 47(3): 597-609. Christiansen, K. 1961. Convergence and parallel- ism in cave Entomobryinae. Evolution, 15(3): 288-301. Cobolli Sbordoni, M., E. De Matthaeis, G. La Rosa, M. Mattocia & A. Vigna Taglianti. 1991. Bio- chemical differentiation and divergence time in the Canarian genus Eutrichopus (Coleoptera, Carabidae). Pp. 233-243, In Biogeographical As- pects of Insularity. Atti dei Convegni Lincei, 85. Coello, J., J.M. Cantagrel, F. Heman, J.M. Fuster, E. Ibarrola, E. Ancochea, C. Casquet, C. Jamond, J.R. Diaz de Teran & A. Cendrero. 1992. Evo- lution of the eastern volcanic ridge of the Canary Islands based on new K-Ar data. J. Vol. Geoth- erm. Res., 53:251-274. Cooke, J.A.L. 1965. A contribution to the biology of the British spiders belonging to the genus Dys- dera. Oikos, 16:20-25. Dallwitz, M.J. 1980. A general system for coding taxonomic descrip- tions. Taxon, 29:41-46. Dallwitz, M.J., T.A. Paine & E.J. Zurcher. 1993. User’s guide to the DELTA system: A general sytem for processing taxonomic descriptions, v. 4.01. CSIRO. Canberra. Davis, J.I. & K.C. Nixon. 1992. Populations, ge- netic variation, and the delimitation of phyloge- netic species. Syst. Biol., 41:421-435. Deeleman-Reinhold, C. & PR. Deeleman. 1988. Revision des Dysderinae. Tijdschr. EntomoL, 131:141-269. Denis, J. 1941. Les araignees des les Canaries. Ann. Soc. EntomoL France, 110:105-130. Denis, J. 1953. Araignees recueillies a Tenerife (les Canaries). Bull. Inst. R. Sci. Nat. Belgique., 29:1-8. Denis, J. 1962. Les araignees de LArchipel de Madere. Publ^ces Inst. Zool. Dr. Julyo Nobre, 79:9-118. Frost, D.R. & A.G. Kluge. 1994. A consideration of epistemology in systematic biology, with spe- cial reference to species. Cladistics, 10:259-294. Howarth, EG. 1991. Hawaiian cave faunas: Mac- roevolution on young islands. Pp. 285-295, In The Unity of Evolutionary Biology. (E.C. Dud- ley, ed.). Dioscorides Press, Portland, Oregon. Juan, C., K.M. Ibrahim, P. Oromi & G.M. Hewitt. 1996. Mitochondrial DNA sequence variation and phylogeography of Pimelia darkling beetles on the Island of Tenerife (Canary Islands). He- redity, 77:589-598. Juberthie, C., M. Bouillon & B. Delay. 1981. Sur r existence d’un milieu souterrain superficiel en zone calcaire. Mem. BiopeoL, 8:77-93. Juberthie, C., B. Delay & M. Bouillon. 1980. Sur I’existence d’un milieu souterrain superficiel en zone non calcaire. C.R. Acad. Sc., 290:49-52. Koch, C.L. 1839. Die Arachniden. Ftinfter Band. 136 pp. Niimberg. Latreille, PA. 1804. Tableau methodique des in- sectes. N. Die. Hist. Nat., 24:129-200. Machado, A. 1976. Introduction to a faunal study of Canary Islands’ laurisilva with special refer- ence to the ground beetles (Coleoptera, Carabi- dae). Pp. 347-412, In Biogeography and Ecolo- gy in the Canary Islands. (G. Kunkel, ed.). W. Junk PubL, The Hague, The Netherlands. Mcheidze, TS. 1972. Novije Vide paukov roda Harpactocrates (Dysderidae). Bull. Acad. Sci. Georgian SSR., 68:741-743. Medina, A.L. 1991. El medio subterraeneo superficial en las Islas Canarias: Caracterizacion y consideraciones sobre su fauna. [Ph.D. dissertation]. Univ. de La Laguna, Tenerife, Spain. Mitchell-Thome, R.C. 1985. Radiometric studies in Macaronesia. Bol. Mus. Mun. Funchal, 37(167):52-85. Nixon, K.C. & Q.D. Wheeler. 1990. An amplifi- cation of the phylogenetic species concept. Cla- distics, 6:211-224. Nixon, K.C. & Q.D. Wheeler. 1992. Extinction and the origin of species. Pp. 119-143, In Extinction and Phylogeny. (M.J. Novacek & Q.D. Wheeler, eds.). Columbia Univ. Press, New York. Oromf, P, J.L. Martin, A.L. Medina & L Izquierdo. 1991. The evolution of the hypogean fauna in the Canary Islands. Pp. 380-395. In The Unity of Evolutionary Biology. (E.C. Dudley, ed.). Dioscorides Press, Portland, Oregon. Oromf, P, A.L. Medina & M.L. Tejedor. 1986. On the existence of a superficial underground com- partment in the Canary Islands. Pp. 147-151, In Actas IX Congr. Intemat. Espeleol. Barcelona. Platnick, N.I., J.A. Coddington, R.R. Forster & C.E. Griswold. 1991. Spinneret morphology and the phylogeny of haplogynae spiders (Araneae, Ara- neomorpha). American Mus. Novit., 3016:1-73. Pollard, S.D., R.R. Jackson, A. Van Olphen & M.W Robertson. 1995. Does Dysdera crocata (Ara- neae, Dysderidae) prefer woodlice as prey?. Eth- oL Ecol. EvoL, 7:271-275. 662 THE JOURNAL OF ARACHNOLOGY Ribera, C. 1983. Araneidos de Marmecos L Publ. Dept. Zool. Barcelona, 9:73-76. Ribera, C. 1993. Dysdera caeca n. sp. y Harpactea stalitoides n. sp. (Araneae), dos nuevas especies cavemicolas de Mrruecos y Portugal. Rev. Ar- achnoL, 10(1): 1-17. Ribera, C. & M.A. Amedo. 1994. Description of Dysdera gollumi (Araneae, Haplogynae), a new troglobitic species from Tenerife, Canary Islands, with some comments on Canarian Dysdera. Mem. BiospeoL, 21:115-119. Ribera, C. & A. Blasco. 1986. Araneidos caver= nicolas de Canarias I. Vieraea, 16:41-48. Ribera, C., M.A. Ferrandez & A. Blasco. 1985. Ar- aneidos cavemicolas de Canarias II. Mem. Bios- peoL, 12:51-66. Ribera, C., M.A. Ferrandez & J.A. Perez. 1986. Los Dysderidae (Arachnida, Araneae) cavemi- colas de la Peninsula Iberica. Pp. 241-244, In Proc. 9th Intemat. Congr. Arachnol. Panama. Roberts, M.J. 1995. Collins Field Guide: Spiders of Britain & Northern Europe. Harper Collins Publishers, London. Schmidt, G. 1973. Zur Spinnen-Fauna von Gran Canaria. Zool. Beitr., 19:347-392. Schmidt, G. 1975. Spinnen von Teneriffa. Zool. Beitr., 27:501-515. Schmidt, G. 1981. Zur Spinnen-Fauna von La Go- mera. Zool. Beitr., 27:85-107. Simon, E. 1883. Etudes Arachnologiques XIV Me., materiaux pour servir a la faune arachnol- ogique des les de 1’ Ocean Atlantique. Ann. Soc. Entomol. France., 6:294-298. Simon, E. 1907. Etude sur les Araignees de la sous-section des Haplogynes. Ann. Soc. Ento- mol. Belgique, 51:246-264. Strand, E. 1908. Diagnosen neuer aussereuro- paischer Spinnen. Zool. Anz., 32:769-773. Wheeler, Q.D. & K.C. Nixon. 1990. Another way of looking at the species problem: A reply to De Queiroz and Donaghue. Cladistics, 6:77-81. Wunderlich, J. 1987. Die Spinnen der Kanarischen Inseln und Madeiras. Taxon. EcoL, 1:1-435. Wunderlich, J. 1991. Die Spinnen-Fauna der Mak- aronesischen Inseln. Beitr. Araneol., 1:1-619. Wunderlich, J. 1993. The Macaronesian cave- dwelling spider fauna (Arachnida, Araneae). Mem. Queensland Mus., 33(2): 181-186. Wunderlich, J. 1994. Zu Okologie, Biogeographie, Evolution und Taxonomie einiger Spinnen der Mak- aronesischen Inseln. Beitr. Araneol., 4:385-439. Manuscript received 1 November 1997, revised 27 April 1999. 1999. The Journal of Arachnology 27:663-671 ANTERIOR MEDIAN EYES OF LYCOSA TARENTULA (ARANEAE, LYCOSIDAE) DETECT POLARIZED LIGHT: BEHAVIORAL EXPERIMENTS AND ELECTRORETINOGRAPHIC ANALYSIS J. Ortega-Escobar: Faculty of Psychology, Univ. Autonoma, 28049 -Madrid, Spain A. Munoz-Cuevas: Laboratoire de Zoologie (Arthropodes), M.N.H.N.-C.N.R.S., 61 Rue Buffon, 75231 — Paris Cedex 05, France ABSTRACT. We studied solar light cues that may be used by Lycosa tarentula (Linnaeus 1758) (Ar- aneae, Lycosidae) for homing. Experiments performed under clear skies, under overcast skies, and under clear skies as seen through a plastic sheet (which changed the polarization from linear to elliptical), allowed us to discover which attributes of daylight were used by the spiders during orientation and homing. We found that patterns of linearly polarized light in the natural sky were sufficient to allow accurate homing by the spiders. The homing behavior of individuals having the anterior median eyes (AME) or all other eyes blinded allowed us to determine that AME were responsible for the reception of polarized light. Electroretinography of all eyes confirmed that only the anterior median eyes were differentially sensitive to the orientation of polarization in linearly polarized light. Arthropods are known to detect linearly po- larized light by means of structural speciali- zations of their eyes, whether compound or single ocellar. An orthogonal arrangement of rhabdoms in certain parts of the retina is the anatomical basis of polarized light perception (for review see Wehner 1989). This has been demonstrated in the compound eyes of some insects. A retinal specialization, called “POL area” (Wehner & Strasser 1985), is located in the marginal dorsal part of the compound eyes and oriented towards the zenith when the an- imal is walking. In other insects, for example in the water bug Notonecta glauca, the POL area occurs in the ventral rather than dorsal part of the eye; and this insect uses polarized light reflected from the water surface to detect the ponds into which it dives (Schwind 1984). It has been experimentally shown that the po- larization-sensitive POL area plays a role in celestial navigation in Cataglyphis spp. (Weh- ner 1982) and Apis mellifera (Wehner & Ros- sel 1985; Wehner & Strasser 1985). These Hy- menoptera were unable to orient themselves correctly when the POL area was masked. For spiders, Wehner (1992) reviews the processes and mechanisms by which they can return home: they can use idiothetic, tacto- chemical or visual information for homing. Linearly-polarized light is among the visual cues used by some spiders. In this group, an orthogonal disposition of rhabdoms in the ventral part of the retina of the anterior me- dian eyes (AME) has been found in the age- lenid Agelena gracilens C.L. Koch 1841 (see Schroer 1976) and the lycosid Lycosa taren- tula (Linnaeus 1758) (see Kovoor et al. 1993), and in the peripheral part of the retina of L. erythrognatha Lucas 1836 and L. thorelli (Keyserling 1876) (see Melamed & Trujillo- Cenoz 1966), while Baccetti & Bedini (1964) could not find a similar arrangement in the lycosid Arctosa variana C.L. Koch 1847. In A. variana, Magni et al. (1965) analyzed the electroretinographic responses of the different eyes to polarized light. They found that both AME and PME (posterior median eyes) are capable of analyzing the plane of polarized light. Behavioral studies on orientation by polar- ized light in spiders were initiated by Papi (1955a, b) in lycosids and by Corner (1958) in agelenids. Both investigators found that the only eyes involved in polarized-light detection were the AME (Comer & Claas 1985; Magni et al. 1964). Papi & Syrjamaki (1963) studied the orientation of an Arctic population of Ly- cosa fluviatilis{= Pardosa agricola (Thorell 663 664 THE JOURNAL OF ARACHNOLOGY 1856)), which exhibited a correct solar ori- entation throughout the day. In the present study, the ability of Lycosa tarentula to use celestial cues, mainly the po- larized light pattern of the sky, for homing was examined experimentally. The role of the anterior median eyes in the expression of this behavior was determined through eye-painting experiments, and polarization sensitivity was studied by electroretinography of all the eyes. METHODS Subjects. — Lycosa tarentula is a ground- living lycosid which constructs a burrow near- ly 15 cm in depth and 3 cm in diameter. As do the other members of the family, it has eight eyes arranged in three rows. The front row is composed of the anterior lateral eyes (ALE) and anterior median eyes (AME). The middle row comprises the posterior median eyes (PME); and the posterior one, the pos- terior lateral eyes (PLE). This spider is active during the day and at night. Experiments were performed on adult fe- males of L. tarentula collected at Canto Blan- co, 25 km from Madrid. They were captured as immature individuals and maintained in the laboratory under an artificial light/dark cycle of 12:12 h (light on at 0800 h local time) and at 25 ±2 °C. They were fed with mealflies (Calliphora vomitoria) and crickets (Acheta domestica). Spiders were used for experi- ments at least 5 days after their last molt. Voucher specimens of the adults have been deposited in the Museum National d’Histoire Naturelle (Paris, France). Behavioral experiments. — Adult females were transferred from their individual contain- ers to a terrarium measuring 60 X 30 X 35 cm placed on the roof of the Faculty building. This terrarium had a 15 cm deep substratum of soil; in the middle of one long side of the terrarium, an artificial burrow was built, sim- ilar to that which the spider digs in the field. Experimental procedure: After 5 days of habituation to the terrarium, experiments be- gan. Spiders were gently pushed forward in one of two paths running, right or left from the burrow, along half-the-length and the full width of the terrarium (Fig. 1). The orientation of the shortest return path was 30° NE for one direction and 300° NW for the other and a distance of 35 cm. When the spider arrived at the end of the path, it was placed into a trans- parent open glass container and transferred to the center of an open field 90 cm in diameter, and left in the center of it, and in a different direction from that of the burrow. The walls of the open field were 60 cm high and com- pletely white. The periphery was divided into 10° sectors to identify the direction followed by the spider. The position of the spider was recorded when it was at 40 cm from the cen- ter. Experiments were carried out under five conditions, in all cases with the sun obscured by an opaque screen (Table 1). Eyes were made non-functional by covering them with three coats of black paint (Pelikan Hobby Tempera #11). It was further ensured that the acrylic plas- tic (a Plexiglas® sheet with a polyethylene film) changed linearly-polarized light to ellip- tically-polarized light with maximum efficien- cy at a specific angle between the electric field of the light and the orientation of the sheet. This arrangement was accomplished by using a He/Ne laser (X = 632.8 nm) and introducing the Plexiglas sheet or the polyethylene film between crossed polarizers, together with a Soleil-Babinet compensator. We were able to compensate for the phase change introduced by the sheet and produce zero light at the de- tector. The transmission characteristics of this sheet, measured with a spectrophotometer (Hitachi U-2000), are shown in Fig. 2. Statistical analysis: The directions followed by the animals are shown as circular distri- butions, which were analyzed using circular statistics (Batschelet 1981), calculating the mean resultant vector for every distribution. Appendix I shows how we calculated the mean angle of the sample and the angular de- viation and it describes the Rayleigh and Mar- dia- Watson- Wheeler tests. The statistical evi- dence of directness was tested following the Rayleigh test. If directness was evident, the confidence interval for the mean angle was calculated to test whether the mean direction of the sample deviated significantly from the direction of the burrow. For each condition (e.g., overcast sky, blue sky, etc.) the Mardia- Watson-Wheeler test was used to test whether the two samples (animals that should orient towards 30° NE versus animals that should orient towards 300° NW) differed significantly from each other. Electroretinographic analysis. — Lycosa ORTEGA-ESCOBAR & MUNOZ CUEVAS— POLARIZED LIGHT DETECTION IN LYCOSA. 665 Figure 1. — Apparatus used to study homing in L. tarentula. Right, top view of terrarium in which the animal lived during the study; arrows indicate the possible outward paths. Left, dorsal view of the open field in which the animal was left after being taken from one of the comers opposite to the burrow. 0° was always oriented towards the north. The big arrow indicates the translation of the animal to the center of the open field (shown at half of its actual size in relation to the terrarium). tarentula females from Canto Blanco (Ma- drid, Spain) were maintained in the laborato- ry, in Paris, at 20 °C and under natural light- dark cycles (LD: 10/14). Animal weight was about 2 g. Diurnal electroretinograms were obtained as described in Carricaburu et al. (1990). For the recording of ERGs, the ani- mals were placed on a metallic plate used as the indifferent electrode. The different elec- trode was a thin wire, positioned on the cornea by means of a Prior micromanipulator. The electrodes were connected to a high input im- pedance solid state amplifier. The animal, the micromanipulator and the amplifier were en- closed in a Faraday cage. The output of the amplifier was connected to a cathode ray os- cillator (CRO), the ERGs were displayed on the screen and photographed. The light stimuli Table 1 . — Experimental conditions that were used in this study and cues available to the spiders in each one. Experimental conditions Cues available to the spider Releases under a clear sky All eyes functional Linearly polarized light pattern Light intensity gradient Unintended landmarks Releases under an overcast sky All eyes functional Unintended landmarks Releases under a clear sky filtered through plastic All eyes functional Elliptically polarized light pattern Light intensity gradient Unintended landmarks Releases under a clear sky Only anterior median eyes not functional Linearly polarized light pattern Light intensity gradient Unintended landmarks Releases under a clear sky Only anterior median eyes functional Linearly polarized light pattern Light intensity gradient Unintended landmarks 666 THE JOURNAL OF ARACHNOLOGY Wavelength (nm) Figure 2. — Transmission (%) characteristics of the components of the plastic sheet. were given by an electronic flash and were conveyed to the eyes by an optic fiber. A small device was placed between the tip of the fiber and the eyes making it possible to insert one of two polarizing sheets (Polaroid) and con^ sequently to stimulate the eye by linearly po- larized light, the plane of polarization being either vertical or horizontal. A light flash was delivered to adapt the eye to light and after a variable lapse of time named duration of dark adaptation, a second flash elicited the recorded ERG. The durations of dark adaptation were 1 s, 2 s, 5 s, 10 s, 20 s, 60 s, and 300 s. In arachnids, the full ERG is composed of two negative waves, (3 and 7, and a positive wave, 8 (Fig. 3), in contrast to insects, in which there is a first positive wave, a, prior to p, 7, and 8 (Fouchard & Carricaburu 1972). For each ERG, the amplitude (between the isoelectric line and the top of the p wave) and the latency (between the stimulus and the top of the p wave) were measured: both were found to be highly dependent on dark- adaptation and the hour of recording. Experiments were carried out on three an- imals for 24 hours. Only one spider provided a complete electroretinographic record series, without any trouble. Results shown in Fig. 6 relate to this animal. RESULTS Behavioral experiments. — Releases under a clear sky: As a control, prior to the release of spiders in the open field, the homing be- havior of each individual was observed in the terrarium in order to see if it could return home and was well-adapted to the burrow. Only those spiders well-adapted to the burrow (i.e., those immediately entering it upon con- tact with the first pair of legs) were used in the open-field experiments. Upon release. Figure 3. — Electroretinogram of an arachnid. Ab- breviations: ERG = electroretinogram; CPE = pho- toelectric cell response; A = Amplitude; L = La- tency. each spider remained motionless for several minutes and then began to walk following a linear path, sometimes after a turn to re-orient itself. When each spider had run 40 cm or more, it began to make nest-searching move- ments: legs I and II flexed towards the spider body. These movements have also been ob- served in the terrarium when the spider was returning to the burrow (the entrance of which had been blocked by soil) and do not occur in other situations. These movements, when seen in the open field, indicated that the spider was in search of its burrow entrance. Figure 4 A shows the distribution of movement directions by spiders after open-field releases under a clear sky. The spiders show a correct orien- tation towards the burrow direction (mean val- ue for 30° NE burrow direction sample: 34° ± 5°, r = 0.98, n = 5, Rayleigh test: P < 0.001; mean value for the 300° NW burrow direction sample: 290° ± 18°, r = 0.83, n = 21, Ray- leigh test: P < 0.001), although a bimodal dis- tribution was observed, with some spiders searching for the nest 180° away from it. There was a significant difference between the NE group and the NW group (Mardia- Watson- Wheeler test, P < 0.01). Releases under an overcast sky: Some spi- ders began by making a systematic search in the open field. This behavior consisted of cir- cular movements by the animals beginning at the release point and increasing in radius with time. From time to time the spider returned to the release point. Such spiders were not in- cluded in the analysis. Figure 4B shows that the orientation of the rest of the spiders under an overcast sky was random (mean value for ORTEGA-ESCOBAR & MUNOZ-CUEV AS— POLARIZED LIGHT DETECTION IN LYCOSA. 667 / a = 256° r = 0.42 n = 5 NS a = 270° r = 0.35 n=9 NS Figure 4. — Directions followed by individual spi- ders in the open field under different conditions. A = Releases under a clear sky; B = Releases under an overcast sky; C = Releases under the plastic sheet. Arrows outside of the circles point toward the two possible directions for returning to the nest, according to the outbound paths: 30° (thick arrow) and 300° (thin arrow). Filled circles correspond to home directions of animals that should orient to 300°, while open circles correspond to home direc- tions of animals that should orient to 30°. The thick arrow inside the circle is the mean vector direction for the sample of animals that should return to 30°; the thin arrow inside the circle is the mean vector direction for the sample of animals that should re- turn to 300°. Abbreviations: a — mean vector di- rection; r = length of the mean vector; n = sample size. 30° NE burrow direction sample: 256°, r == 0.42, n = 5, Rayleigh test: P = 0.440; mean value for 300° NW burrow direction sample: 45°, r — 0.14, n = 12, Rayleigh test: P = 0.797). There was no significant difference between the samples (Mardia- Watson- Wheel- er test, NS). Releases under a plastic sheet. As indicated in the Methods section, when sun light passes through the plastic sheet, its polarization changes from linear, which is its predominant characteristic (Waterman 1981), to elliptical, with a small modification of intensity. Spiders released in the open field in these conditions either exhibited the behavior of systematic search or headed in a random direction, as they did under an overcast sky. Only the latter individuals were considered in the analysis. Fig. 4C shows that (mean value for 30° NE burrow direction sample: 270°, r 0.35, n = 9, Rayleigh test: P — 0.342; mean value for 300° NW burrow direction sample: 332°, r = 0.16, w = 11, Rayleigh test: P = 0.710). There was no significant difference between the NE group and the NW group (Mardia- Watson- Wheeler test, P > 0.368). Releases with AMEs blinded: When the AMEs were blinded by black opaque paint, a non-directed distribution was observed (Fig. 5 A). In this case, spiders did not show the systematic search behavior described in Ex- periments B and C. The behavior of these an- imals was completely normal except that their orientation was not to the burrow (mean value for 30° NE burrow direction sample: 45°, r = 0.04, n = 6, Rayleigh test: P > 0.900; mean value for 300° NW burrow direction sample: 144°, r = 0.22, n = 10, Rayleigh test: P = 0.574). There was no significant difference between the NE group and the NW group samples (Mardia- Watson-Wheeler test, NS). Releases with PMEs, PLEs, and ALEs blinded: When all eyes, except the AMEs, were blinded (Fig. 5B), a clear orientation of the spiders to the burrow direction was ob- served (mean value for 30° NE burrow direc- tion sample: 31° ± 5°, r = 0.97, n = 5, Ray- leigh test: P < 0.001; mean value for 300° NW burrow direction sample: 298° ± 20°, r = 0.90, n = Rayleigh test: P < 0.001). There was a significant difference between the NE group and the NW group (Mardia- Watson- Wheeler test, P < 0.01). Electroretinographic analysis. — The elec- troretinographic responses of the AMEs were very different, depending on the plane of light polarization. During the photophase, vertical polarization resulted in a much higher ERG amplitude than did a horizontal one (Fig. 6 shows the change of amplitude with dark ad- aptation). A maximum was reached at 20 s of 668 THE JOURNAL OF ARACHNOLOGY a = 45° r = 0.04 n =6 NS a = 31° r = 0.97 n=5 P< 0.001 Figure 5. — Directions followed by individual spi- ders in the open field with different eyes blinded. A == Releases of animals with AMEs blinded. B = Releases of animals with PMEs, PLEs and ALEs blinded. Symbols as in Fig. 4. dark adaptation, and the amplitude did not change for adaptation up to 5 minutes. Laten- cies of ERG responses obtained for vertically (5 mn in daytime) or horizontally polarized light were similar, about 30 ms. ERGs of other eyes (ALE, PME and PEE) were not signifi^ cantly different, whether light was polarized or not. DISCUSSION Path integration is a route-based homing which allows the animals a straight return af- ter a more or less winding outward trip (Papi 1992). It is a process that allows an animal to deduce its position, in relation to a point of departure, from its own movement. To achieve this the animal has to measure two compo- nents of its outward journey: the direction and the distance. The first component can be mea- sured using external references to calculate the directions followed or using internal ref- erences such as centrally stored recordings of their own movements. Following Etienne et al. (1998), “The ability to “home” irrespec- tive of familiar references from the environ- ment remains the hallmark and safest opera- Figure 6. — Electroretinograms (ERG) under po- larized light. Diurnal amplitudes (mV) of AME ERGs, recorded for different times of dark adapta- tion. The 0 = the horizontal plane of polarization; the • = the vertical plane of polarization. tional criterion for dead reckoning [path integration] ”(p. 56). Our results show that L. tarentula is capable of homing by means of a mechanism which does not use familiar ref- erences and it is based on visual information as external reference to calculate home direc- tion. In experiments with Arctosa (Magni et al. 1964; Papi 1955a, b), the animals were typ- ically placed in the center of a cylinder and their escape directions were registered. They display so-called zonal orientation that gen- erally is not considered as homing (Papi 1992). In contrast, we have used a paradigm that has been used to test homing through path integration in arthropods and mammals (Etienne et al. 1998). In this paradigm, the subjects followed an L- shaped detour and they returned directly to the point of departure. In our study, L. tarentula also shows the ability to shortcut the outward path when returning. Generally, our spiders did not retrace the out- ward path. As with other spiders (Gomer & Claas 1985; Seyfarth et al. 1982), L. tarentula could use tacto-chemical information, visual, or idiothetic information to return home. Tac- to-chemical information is used by lycosid males to find females (reviewed by Tietjen & Rovner 1982), and virgin females of L. tar- entula leave silk threads placed several mil- limeters over the substratum (unpubl. data) that can be used by males to recognize fe- males’ sexual status and to find them. How- ever, in our study, this information was not available to the animal because the spider was placed in an open field that did not contain ORTEGA-ESCOBAR & MUNOZ-CUEV AS— POLARIZED LIGHT DETECTION IN LYCOSA. 669 this source of information. Idiothetic infor^ mation was not used by L. tarentula in hom^ ing because the spiders did not follow any particular direction under an overcast sky. If L. tarentula used idiothetic information, we should have observed successful homing be- havior under an overcast sky, as in the idioth- etic orientation of blind Cupiennius salei (Keyserling 1877) (see Seyfarth et al. 1982) towards a prey from which it has been chased. In this latter ctenid spider it has been argued that homing orientation could be based on non- visual cues, given its nocturnal habits. On the other hand in L. tarentula, a diurnal as well as nocturnal spider, visual and non- visual cues can be used for orientation. Which celestial cues are used by L. taren- tula to return home? The sun’s position is ex- cluded by the experimental conditions of our study. Other cues could be the skylight pat- terns of linearly “polarized light and the inten- sity gradient. L. tarentula may use these cues since it has a good orientation towards the burrow under a blue sky, while it becomes disoriented under an overcast sky. Under this condition, the spider lacks information about the sun’s position and the skylight polarization pattern. We obtained similar results when we used the plastic sheet, which did not signifi- cantly modify the intensity gradient but did cancel out the skylight pattern of linearly po- larized light. So, we think that the relevant cue for homing is the skylight polarization pattern. In Arctosa perita (Latreille 1799) Papi (1955b) also found that this cue was more im- portant for orientation than the light intensity gradient. How could one explain the presence of a bimodal distribution under a blue sky and with the PMEs, PLEs, and ALEs blinded? We sug- gest that the time of day at which observations resulting in orientation opposite to the burrow direction were made is the determining factor. These observations were carried out at the first hours of daylight (near sunrise) or when the sun was at its noon location. Under either of these conditions, a symmetrical distribution of the e-vector patterns occurs around the so- lar-antisolar meridian. So, the spiders tested at these hours could be confused by the sym- metry of the e-vector pattern. This would also be the reason for the absence of a bimodal distribution for the animals that should go to- wards 30° NE under a blue sky and with the PMEs, PLEs, and ALEs blinded. Although in both samples we are at the low limit of sample size, we think that the results are not invali- dated because we have an equal sample size for the animals that should orient towards 30° NE under an overcast sky or with the PMEs, PLEs, and ALEs blinded, and in the latter case the distribution is at random. This bimodal distribution has also been observed in the ce- lestial orientation of sandhoppers (Ugolini et al. 1993) in the morning and at sunset. The skylight polarization pattern is detected by the anterior median eyes of L. tarentula. In Arctosa variana, a wandering ripicolous ly- cosid species, Magni et al. (1964) have shown that polarized light detection is carried out by the AMEs and PMEs. The indirect eyes (PMEs, PLEs and ALEs) have no role in lin- early-polarized light detection by L. tarentula, since a completely random distribution is ob- served when the AMEs are blinded. This is in accordance with a morphological analysis of the AME retina (Kovoor et al. 1993) which has shown that, in ventral photoreceptors, suc- cessive lines of rhabdoms are oriented orthog- onally to each other; such an arrangement was not observed in any other eye type of L. tar- entula. From the analysis of the ERGs, it can be concluded that the anterior median eyes of L. tarentula perceive polarized light well, which is unlikely for the other eyes. This correlates perfectly with the results of behavioral exper- iments. The change in the direction of the po- larization plane produces a large change in the response (amplitude) of the ERG of the AMEs. During the photophase, Lycosa tar- entula being inside the burrow in an almost vertical position or outside the burrow with its body axis parallel to the ground, both vertical and horizontal planes of polarization will play a role in homing. Ongoing experiments will determine if a change in the direction of the polarization plane induces a change in the homing direction taken by the spiders as it has been shown for other spiders like Agelena la- byrinthica (Gomer & Claas 1985). APPENDIX I Suppose that ({>1, (|)2,. . .cj)„ are the directions taken by n animals to return home, then to calculate the mean angle and the angular de- viation of the sample we proceed as follows: 670 THE JOURNAL OF ARACHNOLOGY X = -(cos + cos ^>2 + • ' • + cos #n) y = -(sin 0 # = 180° + arctanr if x < 0. X To calculate the angular deviation we use 180^ s — IT ■V2(l - r). The length of the mean vector, r, is a mea- sure of the dispersion of the data and it is calculated with the following formula r ^ ^ With F2, sample size, and r, length of mean vector, the Rayleigh test gives us the critical level, P, that the parent population is random- ly distributed. Mardia-Watson-Wheeler test: The purpose of this test is to discover whether two inde- pendent samples of «i and W2 observations dif- fer significantly from each other. This test is also known as a uniform- score test because it uses only the order in which the observations of both samples are arranged. We pooled both samples and we ranked the elements of one sample, for example the smallest one. We cal- culate 8 = [3607(«i + «2)] where Wj and Hj are the sizes of both samples and we multiply each rank by 8, transforming each initial value to pj. The resultant vector of the first sample has the following components C, = X cos Pi, Si = 2 sin Pi and the length of the resultant vector is R, = VC^ + and as a test statistic we use B = R?. If n > 17, we use the quantity R2 = 2(n - 1)—^ nin2 which is approximately distributed as cM- squared with two degrees of freedom. ACKNOWLEDGMENTS We wish to thank P. Carricaburu for kindly obtaining the electroretinograms. We also wish to thank J.-M. Cabrera (Fac. Sciences, U.A.M.) for determining the polarization characteristics of the plastic sheet and F. Jaque (Fac. Sciences, U.A.M.) for determj.nieg the transmission characteristics of it. We acknowl- edge J.S. Rovner and an anonymous referee for reviewing the original manuscript and sug- gesting various changes. LITERATURE CITED Baccetti, B. & C. Bedini. 1964. Research on the structure and physiology of the eyes of a lycosid spider. I. -Microscopic and ultramicroscopic structure. Arch. Italiennes BioL, 102:97-122. Batschelet, E. 1981. Circular Statistic in Biology. Academic Press, London. 371 pp. Carricabura, R, A. Munoz-Cuevas & J. Ortega-Es- cobar. 1990. Electroretinography and circadian rhythm in Lycosa tarentula (Araneae, Lycosi- dae). Acta Zool. Fennica, 190:63-67. Etienne, A.S., J. Berlie, J. Georgakopoulos & R. Maurer. 1998. Role of dead reckoning in navi- gation, Pp. 54-68, In Spatial Representation in Animals. (S. Healy, ed.). Oxford Univ. Press, Oxford. Fouchard, R. & P. Carricabura. 1972. Analyse de Felectroretinogramme de ITnsecte. Vision Re- search, 12:1-15. Corner, R 1958. Die optische und kinasthetische Orientierung der Trichterspinne Agelena labyrin- thica, Z. Vergl. Physiol., 51:111-153. Gomer, P. & B. Claas. 1985. Homing behavior and orientation in the funnel-web spider, Agelena la- byrinthica Clerck. Pp. 275-297, In Neurobiology of Arachnids. (EG. Barth, ed.). Springer- Verlag, Berlin. Kovoor, J,, A. Munoz-Cuevas & J. Ortega-Escobar. 1993. Microanatomy of the anterior median' eye and its possible relation to polarized light recep- tion in Lycosa tarentula (Araneae, Lycosidae). Boll. ZooL, 60:367-375. Magni, E, F. Papi, H.E. Savely & R Tongiorgi. 1964. Research on the structure and physiology of the eyes of a lycosid spider. II. The role of different pairs of eyes in astronomical orienta- tion. Arch. Italiennes BioL, 102:123-136. Magni, E, F, Papi, H.E. Savely & R Tongiorgi. 1965. Research on the stracture and physiology of the eyes of a lycosid spider. III. Electroreti- nographic responses to polarized light. Arch. It- aliennes Biol., 103:146-158. Melamed, J. & O. Trajillo-Cenoz. 1966. The fine structure of the visual system of Lycosa (Ara- neae, Lycosidae). Zeit. Zellf., 74:12-31. Papi, E 1955a. Astronomische Orienterang bei der ORTEGA-ESCOBAR & MUNOZ-CUEVAS— POLARIZED LIGHT DETECTION IN LYCOSA. 671 Wolfspinne Arctosa perita. Z, Vergl. Physiol., 37:230-233. Papi, F. 1955b. Ricerche suU’orientamento di Arc- tosa perita (Latr.) (Araneae, Lycosidae). Pubbl. Staz. Zool. Napoli, 27:76-103. Papi, F. 1992. General aspects. Pp. 1-18, In Ani- mal Homing. (F. Papi, ed.). Chapman & Hall, London, New York. Papi, F. & J. Syrjamaki. 1963. The sun-orientation rhythm of wolf spiders at different latitudes. Arch. Italiennes Biol., 101:59-77. Schroer, W.-D. 1976. Polarisation sensitivity of rhabdomeric systems in the principal eyes of the funnel spider Agelena gracilens (Arachnida: Ar- aneae: Agelenidae). Entomol. Germanica, 3:88- 92. Schwind, R. 1984. Evidence for true polarization vision based on a two-channel analyser system in the eye of the water bug, Notonecta glauca. J. Comp. Physiol. A, 154:53-57. Seyfarth, E.-A., R. Hergenroder, H. Ebbes & EG. Barth. 1982. Idiothetic orientation of a wander- ing spider: compensation of detours and esti- mates of goal distance. Behav. Ecol. SociobioL, 11:139-148. Tietjen, W.J. & J.S. Rovner. 1982. Chemical com- munication in lycosids and other spiders. Pp. 249-280, In Spider Communication: Mecha- nisms and Ecological Significance. (P.N. Witt & J.S. Rovner, eds.). Princeton Univ. Press, Prince- ton, New Jersey. Ugolini, A., B. Laffort, C. Castellini & G. Beugnon. 1993. Celestial orientation and ultraviolet per- ception in Talitrus saltator. Ethol. Ecol. EvoL, 5:489-499. Waterman, T. 1981. Polarization sensitivity. Pp. 281-471, In Handbook of Sensory Physiology, vol. VII/6B. (H,-J. Autrum, ed.). Springer- Ver- lag, Berlin. Wehner, R. 1982. Himmelsnavigation bei Insekten. Neurophysiologie und Verhalten. Neujahrsbl. Naturf. Ges. Zurich, 184:1-132. Wehner, R. 1989. The hymenopteran skylight com- pass: Matched filtering and parallel coding. J. Exp. BioL, 146:63-85. Wehner, R. 1992. Arthropods. Pp. 45-144, In An- imal Homing. (F. Papi, ed.). Champan & Hall, London. Wehner, R. & S. Rossel. 1985. The bee’s celestial compass: A case study in behavioural neurobi- ology. Fort. Zook, 31:11-53. Wehner, R. & S. Strasser. 1985. The POL area of the honeybee’s eye: behavioural evidence. Phy- siol. Entomol., 10:337-349. Manuscript reveived 17 August 1998, revised 28 May 1999. 1999. The Journal of Arachnology 27:672-674 RESEARCH NOTE A NEW SPECIES OF THE SPIDER GENUS ZELOTES (ARANEAE, GNAPHOSIDAE) FROM CALIFORNIA Recent sampling near Lake Skinner in southern California has produced the first known specimens of a small species of Zelotes that seems most closely related to Zelotes nan- nodes Chamberlin 1936, known only from southeastern Oregon, Nevada, and Utah (Plat- nick & Shadab 1983). We describe here this new species and provide some notes on its habitat and on other spiders taken in the same samples. Lake Skinner is surrounded by undisturbed Riversidian coastal sage scrub (Westman 1983) in the Southwestern Riverside County Multispecies Reserve. The new species was collected in June from pitfall traps set in two sampling plots, approximately 600 m and 450 m, respectively, from the lake’s northeast shore. The first site (on a north-facing, rela- tively steep slope), providing one male, has a very dense shrub cover, primarily of Artemisia californica Lesson (California sage) and sec- ondarily of Erigonum fasciculatum Bentham (California buckwheat) and Salvia mellifera E. Greene (black sage) with Salvia apiana Jepson (white sage) interspersed among the major shrub components. The composition ra- tio of the three major shrubs is approximately 4:2:1. The soil consists of decomposed granite of relatively fine particle size mixed with clay; large exposed rocks are absent. The substrate is essentially bare, with sparsely distributed annual Schismus grass or, in less exposed ar- eas, a thin lichen cover. Only near the bases of shrubs is any leaf litter present. The second site (on a south-facing, gentle slope), providing the female and a second male, is vegetated by sparsely distributed shrubs, primarily E. fasciculatum and second- arily A. californicum, with a few scattered S. mellifera shrubs. The composition ratio of the two major shrubs is approximately 2.5:1. The soil here also consists of decomposed granite but has a coarser particle size and a much low- er clay content, except in surface depressions where clay has been deposited by runoff. Be- tween the shrubs, the substrate is bare except for sparsely distributed Schismus grass and a few relatively large surface rocks. Leaf litter accumulates only beneath a few of the more closely grouped shrubs. Because of the hard ground surface, few burrows or surface open- ings of any kind were found; those that were noted opened in surface depressions or in the sandier sections of the plot. Three other gnaphosid species were col- lected in the June samples from both plots: Callilepis gosoga Chamberlin & Gertsch 1940, Cesonia classica Chamberlin 1924, and Drassyllus insularis (Banks 1900). The first site also provided Drassyllus fractus Cham- berlin 1936 and (the probably introduced) Ze- lotes nilicola (O. R-Cambridge 1874) in June samples, whereas the second site produced Gnaphosa californica Banks 1904, Micaria jeanae Gertsch 1942, and Zelotes monachus Chamberlin 1924. Over 70% of the Blabomma sanctum Chamberlin & Ivie 1937 specimens collected (from a possible 24 plots) were tak- en in December samples from the first site. The second site produced specimens of ap- parently undescribed species of Blabomma, Aptostichus (both in December samples), and Psilochorus (in June samples). The format of the description follows that of Platnick & Shadab (1983). We thank Mo- hammad Shadab of the American Museum of Natural History for help with the illustrations. Zelotes skinnerensis new species Figs. 1-4 Types. — Male holotype and female allo- type taken in pitfall traps at an elevation of ca. 470 m in a site 450 m from the NE shore- line of Lake Skinner, Southwestern Riverside County Multispecies Reserve, Riverside County, California (<3, 13-16 June, 1998; $ 6-10 June 1998, both by TR. Prentice), de- 672 PLATNICK & PRENTICE 673 Figures 1-4. — Zelotes skinnerensis new species. 1, Left male palp, ventral view; 2, Same, retrolateral view; 3, Epigynum, ventral view; 4, Same, dorsal view. posited in American Museum of Natural His- tory courtesy of Metropolitan Water District of Southern California and R. Redak (Dept, of Entomology, Univ. of California, Riverside). Etymology. — ^The specific name refers to the type locality. Diagnosis.-“This species, with its trans- verse embolus extending across the distal edge of the male palpal bulb, belongs to the laccus subgroup of the genus, and is likely to be confused only with what appears to be its sister species, Z. nannodes. Males have the distal edge of the terminal apophysis rounded (Fig. 1), whereas in Z. nannodes the prolateral edge of the terminal apophysis bears a sharply pointed projection (Platnick & Shadab 1983: figs. 231, 235). Females of the new species have a longer median epigynal plate (Fig. 3) than that found in females of Z. nannodes (Platnick & Shadab 1983: fig. 233); the single female available for study has what appear to be fragments of the male embolus extending from the epigynal openings and partially ob- scuring the median epigynal plate. Description. — Male: Total length 2.49, 2,65. Carapace 1.08, 1.14 long, 0.81, 0.82 wide. Femur II 0.64, 0.66 long. Eye sizes and interdistances: AME 0.03, ALE 0.05, PME 0.04, PLE 0.05; AME- AME 0.04, AME- ALE 0.01, PME-PME 0.05, PME-PLE 0.00, ALE- PLE 0.05; MOQ length 0.12, front width 0,10, back width 0.12. Embolus shorter than in Z. nannodes, without prolateral hump (Figs. 1, 2). Leg spination: tibia IV pi -0-1; metatarsi: III pO-2-2, V2-0-0, rO-1-2; IV pO-2-2. Female: Total length 3.30. Carapace 1.14 long, 0.88 wide. Femur II 0.67 long. Eye sizes and interdistances: AME 0.03, ALE 0.05, PME 0.05, PLE 0.05; AME-AME 0.04, AME- ALE 0.01, PME-PME 0.04, PME-PLE 0.04, ALE-PLE 0.04; MOQ length 0.1 1, front width 0.10, back width 0.14. Epigynum with trian- gular median plate longer than in Z. nannodes (Figs. 3, 4). Leg spination: tibiae: III vlr-2-2; IV pi -0-1; metatarsi: I vO-0-0; II vlr-0-0; III pO-2-2, V2-0-0, rO-1-2; IV pO-2-2. Other material examined. — One male tak- en in pitfall trap at an elevation of ca. 480 m in a site 600 m from the NE shoreline of Lake Skinner, Southwestern Riverside County Mul- tispecies Reserve, Riverside County, Califor- nia (6-10 June 1998, T.R. Prentice), deposited in University of California, Riverside. Distribution. — Known only from the type 674 THE JOURNAL OF ARACHNOLOGY locality in Riverside County, southern Cali- fornia. LITERATURE CITED Platnick, N.L & M.U. Shadab. 1983. A revision of the American spiders of the genus Zelotes. Bull. American Mus. Nat. Hist., 175:97-192. Westman, W.E. 1983. Xeric Mediterranean-type shrubland associations of Alta and Baja Califor- nia and the community/continuum debate. Ve- getatio, 52:3-19. Norman I. Platnick i Division of Inverte- brate Zoology, American Museum of Nat- ural History, Central Park West at 79th Street, New York, New York 10024 USA Thomas R. Prentice: Department of En- tomology, University of California, River- side, California 92521 USA Manuscript received 4 December 1998, revised 15 March 1999. 1999. The Journal of Arachnology 27:675-678 RESEARCH NOTE HARVESTMAN (OPILIONES, GONYLEPTIDAE) TAKES PREY FROM A SPIDER (ARANEAE, CTENIDAE) The diet of harvestmen has been addressed in several papers which showed that there is large variation in feeding habits (see Gnaspini 1996 for discussion). Several authors consid- ered opilionids in general as predators or pos- sibly as scavengers under natural conditions (Verhoeff 1900; Berland 1949; Bishop 1950; Todd 1950; Whiteley 1961; Juberthie 1964; Cannata 1988; Hillyard & Sankey 1989), as well as in the laboratory (Phillipson 1960; Briggs & Ubick 1981; Adams 1984; Acosta et al. 1995). Some studies done in both lab- oratory (Bishop 1950; Capocasale & Bruno- Trezza 1964; Edgar 1971; Anuradha & Par- thasarathy 1976; Holmberg et al. 1984; Gnaspini 1996) and in nature (Bristowe 1949; Cloudsley-Thomson 1958; Savory 1962) have reported that they accept both animal and plant matter. Therefore, harvestmen seem to be onmivorous, with a preference for animal matter (Gnaspini 1996). Herein we report for the first time a case of a harvestman taking prey from another animal. The harvestman and spider were observed in nature without touching or altering subjects, and were photographed by the senior author. The observations were made at Reserva Mu- nicipal da Mata de Santa Genebra, Campinas, state of Sao Paulo, Brazil (22°44'S, 47°06'W) on 26 September 1992 at about 2000 h. The temperature was about 25 °C, relative humid- ity was 60%, and the day was cloudy. Obser- vations were made using a diffuse red light, and pictures were taken using flash. Because of our previous personal observations, we do not believe that the flash photography influ- enced the behavior of both animals. The arachnids and prey were not collected. From the photographs, the harvestman was identified as a female Goniosoma cf. longipes (Roewer 1913) (Gonyleptidae, Goniosomati- nae) and the spider as Enoploctenus cyclo- thorax (Bertkau 1880) (Ctenidae). After the proper nomenclatural changes are made, this is probably the only species of Enoploctenus found in Brazil south of Rio de Janeiro (A.D. Brescovit pers. comm.). The moth could not be identified. The harvestman was first detected on a tree trunk about 20 cm from the spider. The spider was holding the prey (a moth partially wrapped in silk) with its chelicerae (Fig. 1). The harvestman slowly approached the spider, until it touched the spider with its first and second pairs of legs (Fig. 2). The harvestman touched the spider again three or four times. Meanwhile, the spider stood motionless. After 1-2 minutes, the harvestman suddenly moved over the spider, which dropped the prey and backed up about 3-4 cm. Because the har- vestman also moved back slightly (ca. 1-2 cm), the prey was now located between the two arachnids (Fig. 3). Just a few seconds lat- er, the harvestman started moving towards the prey, while the spider turned around and moved away (Fig. 4). The harvestman even- tually picked the prey up with its pedipalps, and remained at the spot handling the prey (maybe already eating it) for several minutes afterwards. No cases of prey theft or kleptoparasitism (regularly stealing food from other species) are known for harvestmen in nature. In cap- tivity, several species of North American har- vestmen have been observed to tug at and take “prey” (chopped pieces of mealworms) from conspecifics (J.C. Cokendolpher pers. comm.). In the wild, some Brazilian harvestmen have been observed eating prey taken from other animals: members of the genera Cosmetus and Metavononoides (Cosmetidae) have been ob- served taking prey from webs of Blechroscelis sp. spiders (Pholcidae) (A.B. Kury pers. comm.); a female Goniosoma inscriptum (Mello-Leitao 1922) (Gonyleptidae) was ob- served eating a homopteran prey wrapped in silk near a Thwaitesia sp. spider (Theridiidae) web, in Guapimirim, Rio de Janeiro (R. Pinto- 675 676 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — Sequence of the harvestman Goniosoma cf. longipes taking a moth from the spider Enoploctenus cyclothorax. 1. The spider was holding the prey with its chelicerae when the harvestman was first detected about 20 cm away; 2. Slowly approaching and then touching the spider with its legs; 3. After 1-2 minutes, the harvestman suddenly moved over the spider, and both backed up; 4. Just a few seconds later, the harvestman started moving towards the prey, while the spider turned around and moved away. Scale = ca. 10 mm. SABINO & GNASPINI— HARVESTMAN TAKES PREY 677 da-Rocha pers. comm.); and a female Gonio- soma longipes was observed carrying arthro- pod pieces (dipteran wings and orthopteran legs) wrapped in silk (G. Machado, pers. comm.). These facts may be interpreted either as evidence of prey theft or as the harvestmen having collected abandoned spider prey, per- haps even partially consumed prey. Because kleptoparasitism is defined as a regular pro- cedure of collecting prey from a given spe- cies, we do not consider any of these cases to be kleptoparasitism but rather consider them to be opportunistic theft. Kleptoparasitism is a relatively common practice for some species of at least five families of spiders (Coyle et al. 1991). In each of these cases, the klepto- parasitic spiders steal from other species of spiders. In addition, kleptoparasites also are generally much smaller than the host and do not confront the host when stealing the prey. In order to understand how/why this har- vestman succeeded in taking prey from so large a spider, it is important to look at their possible relationships. One of the main pred- ators of goniosomatine harvestmen is the ctenid spider Ctenus fasciatus Mello-Leitao 1943, which preys mainly on adult and sub- adult harvestmen (Pinto-da-Rocha 1993; Gnaspini 1996; G. Machado pers. comm.). Enoploctenus cyclothorax is common at the entrances of caves where Goniosomatinae har- vestmen occur (e.g., G. spelaeum (Mello-Lei- tao 1932) from the Ribeira Valley, Sao Paulo state -Trajano & Gnaspini-Netto 1991; Gnas- pini & Trajano 1994), but this spider has never been observed preying on the harvestmen (P.G. pers. obs.). Moreover, during a nocturnal observation (G. Machado, pers. comm.), an adult female E. cyclothorax quickly moved over an adult male G. longipes which was wandering about near the spider. Immediately after touching the harvestman with its first legs and palps, the spider backed up and stopped, while the harvestman kept on walk- ing in its path, showing no reaction to the spi- der. The same occurred between E. cyclothor- ax and G. spelaeum (EH. Santos pers. comm.). Possibly, the timidity of the species determined their relationships with harvest- men and the possibility of the spider having its prey stolen by a harvestman. Ctenus spp. are much more aggressive than Enoploctenus spp. (A.D. Brescovit pers. comm.). Other fac- tors involved could be the time since when the spider last fed as well as the age of the spider (which seemed to be a juvenile, based on size); i.e., juveniles may be less aggressive than adults, although no study on aggressive- ness of these spiders has been conducted (A.D. Brescovit pers. comm.). The size of the “thief” may also be important in this kind of behavior — Goniosomatinae are large heavy- bodied harvestman. Moreover, in the above cited cases observed by Kury and Pinto-da- Rocha, the harvestmen were always bigger (in body dimensions) than the spider, which could have prevented the spiders from protecting themselves from the theft. ACKNOWLEDGMENTS Dr. A.D. Brescovit (Instituto Butantan, Sao Paulo) identified the ctenid spider from the photographs. Dr. Brescovit, Dr. A.B. Kury (Museu Nacional, Rio de Janeiro), Dr. R. Pin- to-da-Rocha (Museu de Zoologia USP, Sao Paulo), G. Machado (Unicamp, Campinas), EH. Santos (IBUSP, Sao Paulo), and J.C. Cok- endolpher (Lubbock, Texas) improved the manuscript by providing information from their personal observations and making help- ful suggestions. L. Lopes (Unicamp, Campi- nas) helped during the field trip. The junior author has a research fellowship from CNPq (Conselho Nacional de Desenvolvimento Cientifico e Tecnologico). LITERATURE CITED Acosta, L.E., EE. Pereyra & R.A. Pizzi. 1995. Field observations on Pachyloidellus goliath (Opiliones, Gonyleptidae) in Pampa de Achala, province of Cordoba, Argentina. Bull. British Arachnol. Soc., 10(l);23-28. Adams, J. 1984. The habitat and feeding ecology of woodland harvestmen (Opiliones) in England. Oikos, 42:361-370. Anuradha, K. & M.D. Parthasarathy. 1976. Field studies on the ecology of Gagrellula saddlana Roewer (Palpatores, Opiliones, Arachnida) and its behaviour in the laboratory condition. Bull. Ethol. Soc. India, 1:68-71. Borland, L. 1949. Ordre des Opilions. Pp. 761- 793, In Traite de Zoologie. (PR Grasse, ed.). Maisson et Cie., Paris, vol. 6. Bishop, S.C. 1950. The life of a harvestman. Na- ture Magazine, 276:264-267. Briggs, T.S. & D. Ubick. 1981. Studies on cave harvestmen of the Central Sierra Nevada with descriptions of new species of Banksula. Proc. California Acad. Sci., 42:315-322. Bristowe, W.S. 1949. The distribution of harvest- 678 THE JOURNAL OF ARACHNOLOGY men (Phalangida) in Great Britain and Ireland, with notes on their names, enemies and food. J. Anim. EcoL, 18:100-114. Cannata, L. 1988. Observations on the Opiliones of the Cachot peat Log (Switzerland). Bull. Soc. Neuchatel. Sci. Nat., 111:67-70. Capocasale, R. & L. Bmno-Trezza. 1964. Biologia de Acanthopachylus aculeatus (Kirby, 1819) (Opiliones, Pachylinae). Rev. Soc. Umguaya En- tomol., 6:19-32. Cloudsley-Thompson, J.L. 1958. Spiders, Scorpi- ons, Centipedes and Mites. Pergamon Press, London. 227 p. Coyle, F.A., T.C. O’ Shields & D.G. Perlmutter. 1991. Observations on the behavior of the klep- toparasitic spider, Mysmenopsis furtiva (Araneae, Mysmenidae). J. ArachnoL, 19:62-66. Edgar, A.L. 1971. Studies on the biology and ecol- ogy of Michigan Phalangida (Opiliones). Misc. Publ. Mus. Zool. Univ. Michigan, 144:1-64. Gnaspini, P. 1996. Population ecology of Gonio- soma spelaeum, a cavernicolous harvestman from southeastern Brazil (Arachnida: Opiliones: Gonyleptidae). J. Zool., 239(3):417-435. Gnaspini, P & E. Trajano. 1994. Brazilian cave invertebrates, with a checklist of troglomorphic taxa. Revta Brasileira EntomoL, 38(3/4):549- 584. Hillyard, P.D. & J.H.P. Sankey. 1989. Harvestman. Synopses of the British Fauna, new series, 4:1- 119. Linnean Soc., London. Holmberg, R.G., N.PD. Angerilli & L.J. Lacasse. 1984. Overwintering aggregations of Leiobunum paessleri in caves and mines (Arachnida, Opili- ones). J. ArachnoL, 12:195-204. Juberthie, C. 1964. Recherches sur la biologie des Opilions. Ann. Speleol., 19:5-238. Phillipson, J. 1960. A contribution to the feeding biology of Mitopus morio (F.) (Phalangida). J. Anim. EcoL, 29:35-43. Pinto-da-Rocha, R. 1993. Invertebrados caverru- colas da por^ao meridional da Provmcia Espe- leologica do Vale do Ribeira, sul do Brasil. Revta Brasileira Zool., 10(2):229-255. Savory, T.H. 1962. Daddy longlegs. Sci. American, 207(4): 119-128. Todd, V. 1950. Prey of harvestmen (Arachnida, Opiliones). EntomoL Mon. Mag., 86:252-254. Trajano, E. & P. Gnaspini-Netto. 1991. Composi- gao da fauna cavemicola brasileira, com uma analise da distribui^ao dos taxons. Revta Brasi- leira Zool., 7(3):383-407. Verhoeff, C.W. 1900. Zur Biologie von Ischyrop- salis. Zool. Anz., 23:106-107. Whiteley, D.A. 1961. Unusual feeding habits of a harvestman (Opiliones). EntomoL Mon. Mag., 97:187. Jose Sabino: Museu de Historia Natural, In- stituto de Biologia, Universidade Estadual de Campinas, Caixa Postal 6109, 13083- 970, Campinas, SP, Brazil. Pedro Gnaspini: Departamento de Zoolo- gia, Instituto de Biociencias, Universidade de Sao Paulo, Caixa Postal 11461, 05422- 970, Sao Paulo, SP, Brazil. Manuscript received 15 December 1997, revised 25 November 1998. 1999. The Journal of Arachnology 27:679-684 RESEARCH NOTE ESCAPE BEHAVIOR MEDIATED BY NEGATIVE PHOTOTAXIS IN THE SCORPION PARUROCTONUS UTAHENSIS (SCORPIONES, VAEJOVIDAE) Desert grassland scorpions, Paruroctonus utahensis (Williams 1968), are nocturnal, sand-dwelling arachnids that inhabit relatively open, easily accessible sand dune systems. They maintain individual home burrows from which they emerge at night to hunt and to which they return. In general, scorpions can orient in low light levels (Fleissner 1977b) and may be less active during a full moon (Warburg & Polis 1990). Both positive and negative responses to light have been observed for a large number of arthropods, yet little is known about the effect of light on scorpion behavior. In an ear- ly study of vinegarooes (Uropygids), another group of arachnids, tests on photoreceptive behavior provided evidence of a negative pho- totactic response (Patten 1917, 1919). One of the first studies to examine this behavior in scorpions was by Abushama (1964), who used a simple two-choice behavioral test chamber with illumination directed from the side and found that the scorpion Leiurus quinquestria- tus Erenberg 1828 (Buthidae) exhibits a neg- ative phototactic response. Subsequent re- searchers (Jander 1965; Torres & Heatwole 1967; Zwicky 1970b) examined orientation behavior in several arthropods, including scor- pions, and observed negative phototaxis. Apart from these studies and the effect of pho- toreception on circadian activity patterns (Fleissner 1977a, 1977b, 1985, 1986), the in- fluence of light on scorpion behaviors, such as visually guided orientation, has received little attention. Most scorpions have two sets of eyes, a me- dial pair and a lateral set, on their dorsal pro- soma. The medial eyes are paired structures located on either side of a mid- sagittal plane through the carapace. The lateral eyes are lo- cated along the anterolateral margin of the carapace, can number from 0=5 (3 in P. utah- ensis) and usually, but not always, occur in equal numbers on the two sides (Hjelle 1990). Several studies have shown that the medial eyes have greater visual acuity and spatial dis- crimination, but lower absolute sensitivity, than the lateral eyes (Machan 1967, 1968; Fleissner 1974, 1977b). The average sensitiv- ity of 8.6 X 10“^ candles-m“^, suggested for the lateral eyes of Androctonus australis Lin- naeus 1758 (Buthidae), rivals the sensitivity of the moths of the genus Ephestia (Fleissner 1977b). We have observed that when P. utahensis are disturbed outside their burrows they often do not immediately return to their burrows, but rather ran toward nearby bushes. Given that scorpions orient in extremely low light levels, and that they possess highly sensitive eyes, we consider here whether these animals may be using visual information to guide such escape movements. In simulated natural habitats in the labora- tory, we have been able to elicit escape be- havior (a movement to the arena wall within 20 seconds of being dropped into the arena center) in P. utahensis. In this study we used a two-choice experimental apparatus, under both infrared and visible light conditions, and found that scorpions use visible light to escape toward dark regions. The animals used in this study were col- lected at eight in early 1997 from, sandy re- gions south of El Paso, Texas. Voucher spec- imens (voucher #626) of P. utahensis used in this study have been deposited (by EAC) at Texas A & M University, Department of En- tomology Insect Collection, College Station, Texas. Animals were measured, weighed, sexed and maintained in 3.8 liter glass jars containing 250 ml of sand collected from the animals’ natural habitat. Animals were main- tained and experiments conducted in the aei- 679 680 THE JOURNAL OF ARACHNOLOGY mal laboratory facilities at the University of Oklahoma under constant temperature and hu- midity (22 °C, RH 55-65%). The testing room was photoperiod reversed to allow for more convenient observations and video recording. The light-dark cycle was set as follows: dark 0840-1830 h and light 1830-0840 h. Animals were fed two small live crickets (Acheta do- mestica) per week and misted with water (20 ml/animal) twice a week. The testing chamber consisted of a round acrylic plastic (Plexiglas®) arena approxi- mately 15 cm tall and 76 cm in diameter. The floor of the arena was covered with 800 ml of autoclaved sand (20 minutes at 130 °C). Encircling the arena was a posterboard shield measuring 74 cm tall and 85 cm in diameter; the part of the shield surrounding one half of the arena (the “dark” side) was black and the part surrounding the other half of the arena (the “light” side) was white. The height of the shield was designed to limit the radius of a test animaTs aerial visual field. Test cham- bers were monitored by a low-light, infrared- sensitive camera (Panasonic CCTV camera, model #WV-BP314) mounted 1.83 m above the arena center. The camera was necessary to score an animal’s choice during infrared light (IR) trials (the animal was not visible to the experimenter) but not necessary in the white light (WL) trials in which just enough light was available for the experimenter to score the trial. The camera was also needed to produce video records for analysis. A tele- vision monitor, used to view IR experiments via the low-light camera, was positioned out- side the testing room. A voltage meter (Mi- cronta Digital Multimeter 22- 185 A) and broad-band solar cell (Radio Shack, silicon solar cell 2X4 cm, max ratings 0.55 V, 0.3 amp) were used to ensure that WL and IR intensities remained constant and that IR in- tensity remained at least lOX the WL inten- sity. The WL (Radio Shack Krypton mini- lamp, model #272-1150, 2.5 V, 0.430 amp) and IR (Ultrac IR-50FL, 50 W) light sources were centered above the arena at a height of 1.83 m and gave off an illumination level of 0.25 lux (measured using a Pasco scientific photometer, model 9152, calibrated to a 2700 °K tungsten filament lamp). Black felt was hung over the top of the apparatus to reduce overhead visual cues. To test the effect of visible light on the es- cape behavior of P. utahensis, 20 animals were randomly assigned to one of two groups, A and B (56" and 5$ in each group). Group A was exposed to white light (WL) first and infrared light (IR) second and group B was exposed to IR first and WL second. To elim- inate bias in the direction of placement, ani- mals were dropped into the arena center by means of a cylindrical tube (13 cm long and 5 cm in diameter). If an animal did not move within approximately one minute it was re- moved from the arena and re-dropped. Ani- mals were scored based on their first contact with the arena wall: ‘1’ for contact with the dark side, ‘0’ for contact with the light side. Once a movement was observed and scored, the scorpion was removed from the arena and returned to its jar and the substrate on the are- na floor was mixed to eliminate chemical cues that could be used by subsequent animals. Halfway through each light regime (after 25 drops) the sand substrate was removed and fresh substrate was spread on the arena floor. Before each drop, the arena and shield were rotated 45° independently (arena counter- clockwise, shield clockwise) so that no animal encountered the same arena/shield orientation twice. Each animal was tested five non-con- secutive times per light regime (approximately 20-30 minutes between each drop) and given at least three days off between the two light treatments. This set of tests was conducted over a two- week period in January 1998 be- tween the times of 0900-1700 h CST In order to have video records of move- ments, the above experiment was repeated un- der the same conditions except that IR was added to the WL. Groups were reduced to n = 5 (3$, 2(3) due to the limited number of healthy animals available from the first ex- perimental group. These tests were conducted over a two-week period in February-March of 1998, and daily video recording times were as in the first experiment. For both experiments, the behavioral scores of animals, based on their five drops, were summed and averaged. Scores for each treat- ment group were compared to a theoretical random score of 2.5 (no light-dark preference) using Wilcoxon’s Signed Rank Test. To verify that animals were responding to the light conditions of our experiment and not to other sensory cues, such as geomagnetic information or chemical trails, the video re- CAMP & GAFFIN— SCORPION PHOTOTAXIC ESCAPE RESPONSE 681 cords of the second experiment were reviewed and the animals’ initial wall contacts were plotted in three ways: relative to the shield, relative to the arena and relative to the room. These contacts were transcribed to transpar- encies and then to computer-generated plots for analyses. We calculated the mean vectors (Batschelet 1981) for each animal (n = 5, five trials) and the average vector for the test group (n = 10, 10 animals). Vectors were calculated in this manner to permit consideration of the data as independent measures. The Rayleigh test for randomness was used to determine the statistical significance of the mean vector for each group (Batschelet 1981). There was considerable variability in the behavior of animals in our experiments. Some animals moved quickly under both WL and IR light treatments, reaching the arena wall within 5 sec, while others took several min- utes to reach the wall. The distance traveled was also variable; some scorpions took a di- rect route to the dark or light side while others had a more circuitous route. Only 15% of the animals needed to be re-dropped before they initiated escape behavior. In experiment 1, twenty animals were scored as they made initial contact with the arena wall. The order in which animals were exposed to light (i.e., IR or WL first) did not affect scorpion behavior, and the results for the two groups of animals were therefore combined. The averaged scores for animals in each light treatment group were compared to a hypothetical random score of 2.5. Under WL, animals showed a significant preference for the dark region of the arena (x = 4.18 ± 0.28 SE, P < 0.01 j, whereas under IR, the animals did not exhibit a preference (x = 2.65 ± 0.42 SE). The second experiment utilized WL H- IR as the visible light source and IR served as the non-visible light source. The scores from this experiment were processed as in the first; again, there was a statistically significant pref- erence for the dark side under WL + IR (x = 3.70 ± 0.21 SE, P < 0.01) but not IR alone (x - 2.10 ± 0.43 SE). In the second experiment we recorded scorpion movements under both light treat- ments using the low-light, IR-sensitive cam- era. Using the video records we noted the an- imals’ initial wall contacts and plotted these in three different ways: relative to fixed points on the black and white shield, to fixed points in the arena, and to fixed points in the room. When plotted relative to a fixed point on the shield, the data reveal a distinct pref- erence toward the dark side under WL + IR (P < 0.05) but not under IR. In contrast, the initial wall contacts did not show any appar- ent distribution relative to fixed points in the arena or room (Fig. 1). Our study presents strong evidence that P. utahensis can use light to orient. When pro- vided with a choice between black and white sides of an arena, animals demonstrated a sig- nificant preference to move toward the dark region under WL but showed no significant preference for either region under IR. Researchers have previously established negative phototaxis (Angermann 1957; Abu- shama 1964; Jander 1965; Torres & Heatwole 1967) for several species of scorpion. Scor- pions are also known to have chemosensory organs which might contribute to orientation (Abushama 1964; Foelix & Schabronath 1983; Gaffin & Brownell 1992, 1997). We evaluated other cues that were potentially available for orientation by plotting initial wall contacts relative to fixed points in the arena and room. In both cases, initial wall contacts were not significantly grouped along the arena wall, suggesting that neither chem- ical cues from previous trials nor geomagnetic forces played a role in the scorpion’s orien- tation behavior in our study. The photoreceptors responsible for mediat- ing the orientation responses we have ob- served are unknown, but possible candidates include the medial and lateral eyes and meta- somatic (tail) photoreceptors. Previous re- searchers (Zwicky 1970b; Geethabali & Rao 1973) have shown that scorpions with their eyes masked could still orient to darkened re- gions of test chambers, thus suggesting that the tail photoreceptors are sufficient to guide this behavior. We found no indication of negative photo- taxis under IR illumination which is in line with physiological responses recorded by pre- vious researchers who examined spectral sen- sitivity in various species of scorpion (Ma- chan 1968; Geethabali & Rao 1973; Fleissner 1985; Zwicky 1968, 1970a). The medial and lateral eyes have been shown to have a peak response in the blue-green (450-500 nm) re- gion of the spectrum, with the lateral eyes 682 THE JOURNAL OF ARACHNOLOGY WL+ IR IR r = 0.61 r = 0.08 Figure 1. — Scorpion orientation behavior under white light plus infrared light (WL -I- IR) and infrared light (IR). Animals’ initial wall contacts have been plotted relative to the shield, the arena and the room. In each plot, the thin lines indicate the mean vector for each animal, while the thick arrow indicates the mean vector for each group. The light and dark halves of the shield, the arbitrary reference point within the arena (►) and the northerly direction (N) in the room are noted. Abbreviations: r = mean length of vector, 4) = mean angle of vector and P = probability based on Rayleigh tests for randomness. CAMP & GAFFIN— SCORPION PHOTOTAXIC ESCAPE RESPONSE 683 showing an additional peak in the long ultra- violet (300+ nm) region (Machan 1968). The metasomatic photoreceptors also show sensi- tivity in the same spectral range (300-500 nm) (Geethalbali & Rao 1973; Zwicky 1970b). Furthermore, it is interesting to note that scorpion cuticle fluoresces green under ultraviolet illumination peaking around 450- 500 nm (Fasel et al. 1997). Taken together, it is tempting to offer the hypothesis that scor- pion cuticle is acting as a light amplifier and that animal movements in our experiments may have been directed by the neural integra- tion of light intensity across the animaFs en- tire cuticular surface. In this study we have presented evidence of a distinctive, naturalistic behavior that could be used as an assay to resolve some long- standing questions concerning scorpion vi- sion. In particular, this assay could be used to determine which types of photoreceptors are required for orientation and the light intensity threshold necessary for the occurrence of this behavior. ACKNOWLEDGMENTS This work was supported in part by a grant from the Undergraduate Research Opportunity Program at the University of Oklahoma. We thank Drs. Joseph Bastian, Ari Berkowitz, Marielle Hoefnagels and David Sissom for valuable advice, technical support and con- sultation services. The experiments conducted complied with current laws governing animal care and usage in the United States. LITERATURE CITED Abushama, ET, 1964. On the behavior and sensory physiology of the scorpion Leiurus quinquestria- tus. Anim. Behav., 12:140-153. Angermann, H. 1957. Uber Verhalten, Spermato- phorenbildung und Sinnesphysiologie von Eus- corpius italicus Hbst. und verwandten Arten (Scorpiones, Chactidae). Zeitschrift fiir Tierpsy- chologie, 14:276-302. Batschelet, E. 1981. Circular Statistics in Biology. Academic Press, London. 371 pp. Fasel, A., P.A. Muller, R Suppan & E. Vauthey. 1997, Photoluminescence of the African scorpi- on Pandinus imperator. J. Photochem. Photobiol. B: Biol., 39:96-98. Fleissner, G. 1974. Circadiane Adaptation und Schirmpigmentverlagerung in den Sehzellen der Medianaugen von Androctonus australis L. (Buthidae, Scorpiones). J. Comp. Physiol., 91: 399-416. Fleissner, G, 1977a. Scorpion lateral eyes: Ex- tremely sensitive receptors of Zeitgeber stimuli. J. Comp. Physiol., 118:101-108. Fleissner, G. 1977b. The absolute sensitivity of the median and lateral eyes of the scorpion, Androc- tonus australis L. (Buthidae, Scorpiones). J. Comp. Physiol., 118:109-120. Fleissner, G. 1985. Intracellular recordings of light responses from spiking and nonspiking cells in the median and lateral eyes of the scorpion. Na- turw., 72:46-48. Fleissner, G. 1986, Die innere Uhr und der Licht- sinn von Skorpionen und Kafem; zur neurobiol- ogischen Analyse der Circadianen Uhr der Ar- thropoden. Naturw., 73:78-88. Foelix, R.F & J. Schabronath. 1983. The fine structure of scorpion sensory organs. I, Tarsal sensilla. Bull. British Arachnol. Soc., 6:53-67. Gaffin, D.D. & P.H. Brownell. 1992. Evidence of chemical signaling in the sand scorpion, Paru- roctonus mesaensis (Scorpionida: Vaejovidae). Ethology, 91:59-69. Gaffin, D.D. & P.H. Brownell. 1997. Response properties of chemosensory peg sensilla on the pectines of scorpions. J. Comp. Physiol. A, 181: 291-300. Geethabali & K.P. Rao. 1973. A metasomatic neu- ral photoreceptor in the scorpion. J. Exp. Biol., 58:189-196. Hjelle, J.T. 1990. Anatomy and morphology. Pp. 9-63, In The Biology of Scorpions. (Gary A. Polis, ed.). Stanford Univ. Press, Stanford, Cali- fornia. Jander, R. 1965. Die Phylogenie von Orienti- erungsmechanismen der Arthropoden. Verh. Deutschen Zool. Ges. Akademische Verlagsge- sellschaft Geest & Portig K.G., Leipzig. Machan, L. 1967. The effect of prolonged dark ad- aptation on sensitivity and the correlation of shielding pigment position in the median and lat- eral eyes of the scorpion. Comp. Biochem. Phy- siol., 26:365-368. Machan, L. 1968. Spectral sensitivity of scorpion eyes and the possible role of shielding pigment effect. J. Exp. BioL, 49:95-105. Patten, B.M. 1917. Reactions of the whip-tail scor- pion to light. J. Exp. Zool., 23:251-275. Patten, B.M. 1919. Photoreactions of partially blinded whip-tail scorpions. J. Gen. Physiol., 19: 435-458. Torres, F. & H. Heatwole. 1967. Orientation of some scorpions and tailless whip-scorpions. Zeitsch. Tierpsych., 24:546-557. Warburg, M.R. & G.A. Polis. 1990. Behavioral re- sponses, rhythms, and activity patterns. Pp. 224- 246, In The Biology of Scorpions. (Gary A. Po- lis, ed.). Stanford Univ. Press, Stanford, California. 684 THE JOURNAL OF ARACHNOLOGY Zwicky, K.T. 1968. A light response in the tail of Urodacus, a scorpion. Life Sciences, 7:257-262. Zwicky, K.T. 1970a. The spectral sensitivity of the tail of Urodacus, a scorpion. Experientia, 26:317. Zwicky, K.T. 1970b. Behavioral aspects of the ex- traocular light sense of Urodacus, a scorpion. Experientia, 26:747-748. Elizabeth A. Camp and Douglas D. Gaffin: Department of Zoology, University of Oklahoma, Norman, Oklahoma 73019 USA Manuscript received 6 October 1998, revised 18 March 1999. 1999. The Journal of Arachnology 27:685-688 RESEARCH NOTE NOTES ON CYCLOSA INSULANA (ARANEAE, ARANEIDAE) OF PAPUA NEW GUINEA The orb weaver Cyclosa insulana (Costa 1834) is one of several species within the fam- ilies Uloboridae and Araneidae that builds sta- bilimenta, conspicuous silk structures within the orb. Structures referred to as stabilimenta vary considerably among species, and their functions have been hotly debated. Stabili- menta may serve to attract prey (Craig & Ber- nard 1990) or camouflage the spider against predators (Edmunds 1986; Nentwig & Rogg 1988; Craig & Bernard 1990; Eberhard 1990; Schoener & Spiller 1992; Kerr 1993; Black- ledge 1998b) or startle predators (Schoener & Spiller 1992) (see also Nentwig & Ross (1988) who consider neither prey attraction nor camouflage as important). Stabilimenta may also prevent aerial insects and birds from damaging the web by advertizing the presence of the web (Ewer 1972; Horton 1980; Eisner & Nowicki 1983; Kerr 1993; Blackledge 1998a). Phylogenetic data suggest that stabi- limenta have evolved several times (Scharff & Coddington 1997), perhaps serving different functions. Cyclosa (Menge 1866) species exhibit both circular and linear stabilimenta (Edmunds 1986). Camouflaging (Marson 1947; Edmunds 1986; Neet 1990) and stabilizing (Neet 1990) functions have been suggested for the stabili- menta of some Cyclosa species, and Tso (1998) found support for the insect- attraction hypoth- esis in Cyclosa conica (Pallas 1772). Here we describe the characteristics of the linear stabi- limentum in a population of C. insulana in Papua New Guinea. We provide detailed mea- surements of web characteristics and a small manipulative test of the aposematic (i.e., “web advertisement”) function for stabilimenta. We also compare some unique observations of their mating behavior to those of a previous report (Robinson & Robinson 1980). Cyclosa insulana ranges from the Mediter- ranean to Australia. We carried out our field work in Cimbu Province, Papua New Guinea between 20-31 August 1995. Our field site was located at the Wara Sera Research Sta- tion, 10 miles northeast of the village of Haia and approximately 2286 m above sea level. We found C. insulana under the eaves of two buildings in areas measuring approximately 3 X 20 X 15 m and 2 X 6 X 6 m. We charac- terized C. insulana webs by measuring (1) web diameter from the two widest points (from the top of the orb to bottom), (2) sta- bilimentum length and by counting (3) the number of radii and sticky spirals of the webs of 39 adult females and juveniles. Adult males were seen only during courtship on the web of a single adult female. The radii and sticky spirals were numerous and tightly woven, so we counted them twice independently and av- eraged our findings for each web. Webs that were damaged to a point where we could not accurately assess diameter, number of radii and rings or stabilimentum length were not used in our analyses for that character. Web and stabilimentum. — A summary of web characteristics is found in Table 1 . Cyclosa insulana from Spain (Neet 1990) and Burma (Marson 1947) produce both linear and circular stabilimenta. In exposed (windy) sites, spiders produced significantly smaller webs with a greater number of circular stabilimenta than in unexposed (calm) sites (Neet 1990). In our study population, C insulana constructed per- manent linear stabilimenta, but no circular sta- bilimenta; and webs were relatively large (Ta- ble 1). Our study site was very well protected from wind, so the most parsimonious expla- nation is that local conditions have inhibited construction of circular stabilimenta. The linear stabilimentum of C insulana ex- tends through the hub at its midpoint. Like sev- eral spiders of the family Uloboridae (Lubin 1986), C insulana place egg sacs and debris, including plant material, exuviae and prey exo- skeletons, in the stabilimenta (Table 2). When sitting at the hub, and at the midpoint of the 685 686 THE JOURNAL OF ARACHNOLOGY Table 1. — Web characteristics. Diameter and stabilimentum length (in cm, mean ± SD) are compared to measurements of Cyclosa insulana webs in Balearics, Spain (Neet 1990). Neet (1990) Web characteristics This study Windy (n = 22) Calm {n = 20) Diameter 17.9 ± 5.1 (« - 37) 14.4 ± 4.47 11.7 ± 3.44 Radii 51 ± 13 {n = 35) Rings 53 ± 10 in = 31) Damage (%) 14.7 ± 16.0 Linear stabilimentum length 9.1 ± 31.6 {n = 39) 2.25 ± 1.87 2.25 ± 1.29 stabilimentum, the spider folds its legs against its body. This makes the spider appear (to our eyes) virtually indistinguishable from the sta- bilimentum. It seems hkely that the stabilimen- tum in this species serves to conceal the spider from aerial predators (Neet 1990) and perhaps potential prey. Our measurements show spiders to have dramatically longer stabilimenta than previously reported (Neet 1990). Linear stabi- limenta are permanent structures, independent of web renewal (Table 1; pers. obs.). So, under consistently calm conditions, where there would be no advantage in switching to circular stabilimenta, linear stabilimenta are expected to grow longer as prey items and exuviae are add- ed over time. Stabilimenta may serve to advertize the web Table 2. — Contents of stabilimenta (n = 10) ex- pressed in terms of the mean number of items found per stabilimentum and the mean proportion of sta- bilimentum length taken up by items found. Fungi and plant material were estimated only in terms of the proportion of stabilimentum length with fungi and/or plant material. Mean number ± SD Mean proportion of length ± SD Egg sacs 2.3 ± 2.2 0.24 ± 0.23 Exuviae 2.1 ± 2.0 0.1 ± 0.2 Fungi and/or plant material n/a 0.1 ± 2.1 Arthropod exoskeletons 4.0 ± 3.0 0.6 ± 0.2 Araneae 0.02 ± 0.63 n/a Homoptera 0.50 ± 0.85 n/a Coleoptera 3.0 ± 2.4 n/a Hemiptera 0.20 ± 0.63 n/a Diptera 0.20 ± 0.42 n/a Hymenoptera 0.80 ± 1.14 n/a to birds and large insects (Ewer 1972; Horton 1980; Eisner & Nowicki 1983; Kerr 1993), which might otherwise damage the web. In our study site, there was an abundance of vespid wasps capable of such damage. Stabilimenta that provided a visual contrast with background structures would be particularly effective as a conspicuous warning signal. Webs at our site were oriented perpendicularly to the ground and in between the wooden posts supporting the buildings, thereby possessing vertical or horizontal backgrounds (specific background for a given web could not be assigned since it is dependent on the direction of approach). We assessed stabilimentum orientation {n = 39) using a circular grid held behind the web, and leveled so that the spider was aligned with the grid origin. The grid was comprised of 16 sec- tions, each 15°, and we recorded the sections through which the stabilimentum passed. Some spiders built slightly nonlinear stabilimenta, but all tended toward a horizontal or vertical ori- entation (Fig. 1). This tendency might be re- lated to the consistent vertical/horizontal back- ground at this site. A comparison with C insulana webs characterized by more variable backgrounds would provide a test of this pos- sibihty. On the second day of our study, we noted that 2 of the 39 spiders in our study site changed the orientation of their stabilimenta from vertical to horizontal. We hypothesized that stabilimentum orientation was plastic, and that spiders would switch its orientation if the web incurred sufficient damage on the previ- ous day. To test this hypothesis, we chose 20 webs (10 vertical and 10 horizontal stabili- menta) and damaged them by cutting 1 or 2 guylines so that webs collapsed to approxi- mately % their original size. Examination of the webs on the following day revealed that McCLINTOCK & DODSON— CYCLOSA INSULANA OF PAPUA NEW GUINEA 687 Figure 1. — Stabilimentum orientation. Dashed lines indicate zones on the grid where stabilimenta were marked. Line thickness indicates the relative number of stabilimenta with a particular orientation. Note that several stabilimenta were not perfectly linear, but all tended toward a vertical or horizontal orientation on the web. none of the stabilimenta orientations had changed, though one was left unrepaired and two spiders had abandoned their location. Thus, it seems that C insulana does not alter stabilimentum orientation in response to web damage. We have no evidence to support the hypothesis that web orientation is manipulated to enhance the conspicuousness of the web against background vegetation. It is possible, however, that the amount of web damage in- flicted in our study was not sufficient to evoke this kind of response. A more detailed study involving more extensive and/or different kinds of damage (e.g., to the radii rather than guylines) may provide a better test. Courtship behavior. — Robinson & Rob- inson (1980) noted that up to 5 males may congregate on guylines of female’s webs where they may court females, rest, or fight and chase other males. The sequence of be- haviors exhibited by courting males typically involves (1) the attachment of a mating thread (2) plucking, bobbing, bouncing and high-in- tensity jerking on the mating thread, (3) and repeated approaches to the female. Female be- havior may include (1) no response, (2) ap- proaching the male, (3) plucking the web while facing the male and chasing the male. Contact between the male and female may lead to copulation or rejection of the male by the female (Robinson & Robinson 1980; pers. obs.). Our observations of courtship behav- iors, though brief, reveal a male-male com- petition strategy not previously reported in C insulana. On 25 August 1995, we observed courtship between 0945-1115 h, although courtship took place before and after our observations. Four males lined the periphery of a single fe- male’s web, each on separate guylines. Each male advanced along the guyline, apparently laying down silk. As males approached the hub, they plucked the silk vigorously. In re- sponse to plucking, the female oriented to the courting male. If the male continued courting, the female advanced toward him. Sometimes the male was chased off the web before con- tact. If contact occurred, the female struck at the male, the two grappled, and eventually the male fell off the guy line. Males grappled with the female between 1-10 seconds before fall- ing off the guyline. The entire courtship se- quence (plucking, approaching, retreating, grappling, striking, and falling) occurred re- peatedly. Once, the female seemed to assume a copulatory position for approximately 5 sec- onds while the male attempted copulation. Be- fore any palpal insertion was observed, how- ever, the female struck the male with her legs, knocking him off the guy line. Males occasionally plucked the web simul- taneously. On two separate occasions, we not- ed a male traveling around the web frame to the location of another male. After waiting for the other male to begin plucking, he cut a line of silk, effectively eliminating any direct vi- bratory transmissions between the courting male and the female. It is possible that cutting the guyline also reduced vibratory transmis- sion to other parts of the web (i.e., to the lo- cation of other males around the web). After cutting his competitor’s guyline, the male re- turned to his original position on another guy- line, and resumed courting. Throughout the day, damage of this kind reduced web size to approximately 40% of its original size. Silk- cutting behaviors have not been previously re- ported in C. insulana. Rovner (1968) noted severe reduction of the female web by court- ing females of Linyphia triangularis (Clerk). However, it is uncertain if this activity has the same function as in C. insulana. Male L. tri- angularis performed the silk-cutting whether or not competitor males were present. These males may have been reducing the potential for future suitor competition, or this action 688 THE JOURNAL OF ARACHNOLOGY may somehow increase the chances that a fe™ male L. triangularis will mate. In addition to this silk-cutting, male-male competition involved plucking silk lines lead- ing to the location of their competitor. On one occasion, an intruding (fifth) male was “chased” away when one of the males plucked the guyline on which he tried to enter the web. These observations are consistent with previous observations of male-male in- teractions (Robinson & Robinson 1980). It seems that females have the opportunity to detect signals generated from multiple males. Males apparently compete with other males for access to females by reducing the transmission medium between competing males and the female and perhaps by signal- ling their status to intruding males. It also seems possible that the opportunity for male choice exists; the quality of the female could be signaled by the mass of the stabilimeeta (proportional to the number of prey items and egg sacs present). ACKNOWLEDGMENTS We would hke to thank Peter Raven for spi- der identifications, Deb Wright and Andy Mack for building and sharing the Wara Sara research station, Ross Sinclair for guidance, the Wildlife Conservation Society and Ball State University for financial assistance, the Endler lab journal club for comments on the manuscript, and the people of Crater Mountain region for looking after us and conserving their land. Voucher specimens have been deposited with the Queensland Museum, Australia. LITERATURE CITED Blackledge, TA. 1998a. Signal conflict in spider webs driven by predators and prey. Proc. Roy. Soc. London B., 265:1991-1996. Blackledge, TA. 1998b. Stabilimentum variation and foraging success in Argiope aurantia and Ar- giope trifasciata (Araneae: Araneidae). J. ZooL, 246:21-27. Craig, C.L. & C.D. Bernard. 1990. Insect attraction to ultraviolet-reflecting spiders and web decora- tions. Ecology, 71:616-623. Eberhard, W.G. 1990. Function and phylogeny of spider webs. Ann. Rev. Ecol. Syst., 21:341-372. Edmunds,!. 1986. The stabilimenta of Argio/?ey?£i- vipalpis and Argiope trifasciata in West Africa, with a discussion of the function of stabilimenta. Pp. 61-72, In Proceedings of the Ninth Intern. Cong. ArachnoL, Panama 1983 (W.G. Eberhard, Y. Lubin & B.C. Robinson, eds.). Smithsonian Inst. Press, Washington, D.C. Eisner, T. & S. Nowicki. 1983. Spider web protec- tion through visual advertisement: role of the sta- bilimentum. Science, 219:185-187. Ewer, R.F. 1972. The devices in the web of the West African spider Argiope flavipalpis. J. Nat. Hist, 6:159-167. Horton, C.C. 1980. A defensive function for the stabilimentum of two orb-weaving spiders (Ar- aneae, Aranaeidae). Psyche, 87:12-20. Kerr, A.M. 1993. Low frequency of stabilimenta in orb webs of Argiope appensa (Araneae: Araneidae) from Guam: An indirect effect of an introduced avian predator? Pacific Sci., 47:328-337. Lubin, YD. 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, California. Marson, J.E. 1947. Some observations on the var- iations in the camouflage devices used by Cyclo- sa insulana (Costa), an Asiatic spider, in its web. Proc. Zool. Soc. London, 117:219-227. Neet, C.R. 1990. Function and structural variabil- ity of the stabilimenta of Cyclosa insulana (Cos- ta) (Araneae, Aranaeidae), Bull. British Arach- nol. Soc., 8:161-164. Nentwig, W. & H. Rogg. 1988. The cross stabili- mentum of Argiope argentata (Araneae: Aranei- dae) -nonfunctional or nonspecific stress reac- tion. Zool. Anz., 221:248-266. Robinson, M.H. & B. Robinson. 1980. Compara- tive studies of the courtship and mating behavior of tropical araneid spiders. Allen Press, Inc., Lawrence, Kansas. Scharff, N. & J.A. Coddington. 1997. A phyloge- netic analysis of the orb-weaving spider family Araneidae (Arachnida, Araneae). Zool. J. Linn. Soc., 120:355-434. Schoener, TW. & D.A. Spiller. 1992. Stabilimenta characteristics of the spider Argiope argentata on small islands -support of the predator-defense hy- pothesis. Behav. Ecol. SociobioL, 31:309-318. Tso, I.M. 1998. Stabilimentum-decorated webs spun by Cyclosa conica (Araneae, Araneidae) trapped more insects than undecorated webs. J. ArachnoL, 26:101-105. William J* McClintock: Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, California 93106 USA Gary N* Dodson: Department of Biology, Ball State University, Muncie, Indiana 47306 USA Manuscript received 3 October 1998, revised 17 March 1999. 1999. The Journal of Arachnology 27:689-691 RESEARCH NOTE THE EFFECT OF FEEDING HISTORY ON RETREAT CONSTRUCTION IN THE WOLF SPIDER HOGNA HELLUO (ARANEAE, LYCOSIDAE) A spider’s energetic state has been shown to influence a variety of behaviors. Hungry spiders are more likely to cannibalize one an- other (Rypstra 1983; 1986), modify their web construction (Henschel & Lubin 1997), and/ or may relocate more frequently than sated spiders (Turnbull 1964; Riechert & Tracy 1976; Olive 1982; Uetz 1992; Bradley 1993; McNett & Rypstra 1997). In many species, web site and/or microhabitat selection are also influenced by prey availability (reviewed in Wise 1993). Thus, hunger levels and prey availability influence the behavioral decisions made by spiders. However, not all studies re- port significant effects of hunger (e.g., Prov- encher & Riechert 1991). In this study, we investigate the effects of energetic state on re- treat construction in the wolf spider Hogna helluo (Walckenaer 1837)(Araneae, Lycosi- dae). Most wolf spiders are considered to be sit and wait predators which periodically change foraging site (Ford 1978; Stratton 1985). Sen- sory information from prey (Persons & Uetz 1996), as well as the recent consumption of prey, can increase patch residence time (Ford 1978; Wagner & Wise 1997). In two species of burrowing wolf spiders, Miller (1984) found that prey availability directly influenced burrow site selection. The wolf spider, H. helluo, lives on the soil surface of disturbed riparian areas and is com- mon in agricultural fields. Although it is a vagile hunter, females do construct burrows (Dondale & Redner 1990). In a previous study with this species, we found that hunger level influences locomotor activity (Walker et al. 1999). Hungry animals exhibit higher levels of activity than do satiated animals, which suggests that the time elapsed since last feed- ing may influence the degree of searching be- havior exhibited and patch residence time. Since this species facultatively constructs bur- rows, we hypothesized that energetic state also influences burrow construction in this species. Since burrow construction is a poten- tially energetically expensive endeavor (Mar- shall 1995), we predicted that adult female H. helluo maintained with access to high levels of food would be more likely to construct bur- rows than would spiders maintained at lower prey levels. To examine this question, 29 adult female H. helluo were randomly assigned to two treatments, high-fed {n = 14) and low-fed {n = 15). Spiders were fed crickets (Acheta dom- isticus) and were provided with water ad li- bitum. To standardize hunger, all spiders were fed to satiation then starved for one week pri- or to the experiment. To feed spiders to sati- ation, individuals were given 3“"4 crickets per day for several days. Spiders were considered sated when they refused to consume all the available prey items. Following standardiza- tion of hunger levels H. helluo were placed individually into 1.4 liter round containers containing 7”11 cm of moist peat moss sub- strate which had been smoothed to make the surface flat. Animals were then fed either one large (mean = 82.5 ±5.4 mg) or one small (mean 1 1.1 ± 0.62 mg) cricket once per week. These crickets were approximately 40% or 10 % of the body mass, respectively, of adult Hogna. Seven days later, the presence or ab- sence of a burrow was determined by visually inspecting the containers. Burrows were vi- sually conspicuous because of the presence of a large amount of silk and the disturbance of the soil which had been smooth prior to the introduction of the spider. To verify that the treatments had an effect on hunger, we estimated body condition on a random sample of eight animals per treatment (a body-size free measure of nutritional state 689 690 THE JOURNAL OF ARACHNOLOGY Table L — Mean carapace width and abdomen width (mm) for high and low-fed Hogna. Carapace width was not significantly different between high and low-fed spiders; however, abdomen width and body condition of starved spiders was significantly less than fed spiders. Trait Treatment Mean (S.E.) n Test statistic and P-value for comparing high and low-fed spiders Carapace width High-fed 11.54 (0.235) 8 t = 1.383, df= 14 Low-fed 12.00 (0.234) 8 P = 0.1884 Abdomen width High-fed 11.89 (0.252) 8 t = 3.878, df = 14 Low-fed 9.792 (0.481) 8 P = 0.0017 Body condition High-fed 12.127 (0.323) 8 = 29.65 Low-fed 9.559 (0.323) 8 P < 0.0001 Number of spiders High-fed 12 14 Fishers Exact Test with burrows Low-fed 6 15 P = 0.0209 or fatness, Jakob et al. 1996). Abdomen and carapace width were measured using an ocular micrometer on a Wild dissecting microscope. Since the data were normally distributed, body condition was estimated as the analysis of co- variance adjusted mean of abdomen width us- ing carapace width as the covariate. We used Fisher’s Exact test to test the hypothesis that high-fed spiders are more likely to burrow than low-fed spiders. We found no significant difference in car- apace width between high and low fed spiders (Table 1). However, high-fed spiders had wid- er abdomens than low-fed spiders. High-fed spiders also had significantly higher body con- dition than did low-fed spiders. These differ- ences suggest that high-fed spiders were in a much better nutritional state than were low- fed spiders. Significantly more high-fed spi- ders (85%) constructed burrows than did low- fed spiders (Table 1). Hunger level clearly influenced the proba- bility of burrow construction in Hogna helluo. Animals in the high-fed group were much more likely to construct burrows than were animals in the low-fed treatment group. Pre- vious studies have demonstrated that hunger influences locomotor activity in this species. Hungry individuals exhibit high levels of lo- comotor activity relative to sated individuals (Walker et al. 1999). Those data combined with data from this paper suggest that hunger can play an important role in the behavior of this species. Hunger does not seem to affect the behavior of spiders equally (see discussion in Prov- encher & Reichert 1991), Several studies have suggested that hunger is not an important fac- tor influencing spider behavior (Anderson 1974; Greenstone & Bennet 1978; Provencher & Riechert 1991; Walker et al. 1999). In par- ticular, Anderson (1974) found that Lycosa lenta Hentz 1844, another species of lycosid, seem to exhibit normal behavior over a 30- day starvation period. Also, we have found that hunger does not affect locomotor behav- ior in Pardosa milvina (Hentz 1844)(Walker et al. 1999). However, we have found that H. helluo is sensitive to recent levels of prey con- sumption both in activity levels (Walker et al. 1999) and in burrow construction (Table 1). Since burrows represent a considerable ener- getic investment as their construction requires not only the excavation of soil but also the deposition of silk, the fact that well-fed spi- ders were more likely to construct them is not surprising. Our data make it tempting to pre- dict that prey availability influences patch choice and residence time in this species. However, because our spiders were confined, we do not know whether well-fed H. helluo will build burrows wherever they are or if they are capable of connecting high prey capture with a particular site and using that informa- tion to decide whether to construct a burrow. ACKNOWLEDGMENTS This manuscript was improved by com- ments from Dr. Richard Bradley, an anony- mous reviewer, Rob Balfour and Michelle Grey. Lauren Searcy and Matt Thomann helped maintain spiders used in this experi- ment. Voucher specimens were deposited in Miami University’s Hefner Zoology Museum. This research was supported by the Miami University Departments of Zoology, Oxford WALKER ET AL.— FEEDING HISTORY AND RETREAT CONSTRUCTION 691 and Hamilton Campuses, National Science Foundation grant DEB 9527710 to ALR and SDM, and Miami University’s Summer SchoL ars Program. LITERATURE CITED Anderson, J.E 1974. Responses to starvation in the spiders Lycosa lenta Hentz and Filistata hiber- nalis (Hentz). Ecology, 55:576-585. Bradley R.A. 1993. The influence of prey avail- ability and habitat on activity patterns and abun- dance of Argiope keyseriingi (Araeeae: Aranei- dae). J. ArachnoL, 21:91-106. Dondale, C.D. & J.H. Redner. 1990. The insects and arachnids of Canada, Part 17: The wolf spi- ders, nursery web spiders, and lynx spiders of Canada and Alaska. Biosystematics Res. Inst., Ottawa, Canada. Ford, M. 1978. Locomotory behaviour and the pre- dation strategy of the wolf-spider Pardosa amen- tata (Clerck) (Lycosidae). Anim. Behav., 26: 3331-3335. Greenstone, M.H. & A.E Bennett. 1980. Foraging strategy and metabolic rate in spiders. Ecology, 81:1255-1259. Henschel, J.R. & Y.D. Lubin. 1997. A test of hab- itat selection at two spatial scales in a sit-and- wait predator: a web spider in the Namib Desert dunes. J. Anim. EcoL, 66:401-413. Jakob, E., S.D. Marshall & G.W. Uetz. 1996. Es- timating fitness: a comparison of body condition indices. Oikos, 77:61-67. Marshall, S.D. 1995. Natural history, activity pat- terns, and relocation rates of a burrowing wolf spider: Geolycosa xera archboldi (Araneae, Ly- cosidae). J. ArachnoL, 23:65-70. McNett, B.l. & A.L. Rypstra. 1997. Effects of prey supplementation of survival and web site tenac- ity of Argiope trifasciata (Araneae, Araneidae): A field experiment. J. Arachnoi., 25:352-360. Miller, G.L. 1984. The influence of microhabitat and prey availability on burrow establishment of young Geolycosa turricola (Treat) and G. mican- opy Wallace (Araneae: Lycosidae): A laboratory study. Psyche, 91:123-132. Olive, C.W. 1982. Behavioral response of a sit- and-wait predator to spatial variation in foraging gain. Ecology, 63:912-920. Persons, M.H. & G.W. Uetz. 1996. The influence of sensory information on patch residence time in wolf spiders (Araneae: Lycosidae). Anim. Be- hav., 51:1285-1293. Provencher, L. & S.E. Riechert. 1991. Short-term effects of hunger conditioning on spider behav- ior, predation, and gain of weight. Oikos, 62: 160-166. Riechert, S.E. & C.R. Tracy. 1975. Thermal bal- ance and prey availability: bases for a model re- lating web-site characteristics to spider reproduc- tive success. Ecology, 56:265-284. Rypstra, A.L. 1983. The importance of food and space in limiting web-spider densities: a test us- ing field enclosures. Oecologia, 59:312-319. Rypstra, A.L. 1986. High prey abundance and a reduction in cannibalism: the first step to soci- ality in spiders (Arachnida). J. ArachnoL, 14: 193-200. Stratton, G.E. 1985. Behavioral studies of wolf spi- ders: a review of recent research. Rev. ArachnoL, 6:57-70. Turnbull, A.L. 1964. The search for prey by a web- building spiders Achaearanea tepidariorum (C.L. Koch) (Araneae, Theridiidae) Canadian EntomoL, 96:568-579. Uetz, G.W. 1992. Foraging strategies of spiders. TREE, 7:155-159. Wagner, J.D. & D.H. Wise. 1997. Influence of prey availability and conspecifics on patch quality for a cannibalistic forager: laboratory experiments with the wolf spiders Schizocosa. Oecologia, 109:474-482. Walker, S.E., S.D. Marshall, A.L. Rypstra & D.H. Taylor. 1999. The effects of hunger on loco- motory behaviour in two species of wolf spider. Anim. Behav., 68:515-520. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge Univ. Press, Cambridge. Seae E. Walker and 'Samuel D. Marshall: Department of Zoology, Miami University, Oxford Ohio, 45056 USA Aun L. Rypstra; Department of Zoology, Miami University, 1601 Peck Blvd., Ham- ilton Ohio 45011 USA Manuscript received 17 August 1998, revised 29 January 1999. ^ Current address: James H. Barrow Field Sta- tion, Dept, of Biology, Hiram College, Hiram, Ohio 44234 USA. 1999. The Journal of Arachnology 27:692-696 RESEARCH NOTE SURVIVAL OF THE HUNTING SPIDER, HIBANA VELOX (ARANEAE, ANYPHAENIDAE), RAISED ON DIFFERENT ARTIFICIAL DIETS' Spiders occupy an important part of the overall predatory arthropod fauna in different terrestrial ecosystems (Riechert 1974; Riech- ert & Lockley 1984). They are also known to play an important role in the regulation of pest species in agriculture (Whitcomb et al. 1963; Dondale et al. 1979; Dean et al. 1982; Man- sour et al. 1982; Culin & Yeargan 1983; Man- sour et al. 1983; Graze & Grigarick 1989; Riechert & Bishop 1990; Barrion & Litsinger 1995). Baseline information on life history and biology is fundamental for ecological work and is also important to further investi- gate the potential of spiders as biological con- trol 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. Anoth- er reason is the lack of appropriate artificial diets. Since spiders are primarily carnivorous, they require behavioral cues from the prey to initiate attack and feeding (Riechert & Luczak 1982). This makes the rearing and mainte- nance of spiders in the laboratory a very la- borious task. Moreover, it appears that most spiders 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 complete nutritional require- ments for spider is necessary. Recently, it was reported that some species of wandering spi- ders are facultative nectar feeders (Taylor & 'This manuscript is Florida Agricultural Experiment Station Journal Series No. R-05 542. Foster 1996). This could explain the success of some previous works (Peck & Whitcomb 1968; Whitcomb 1967) on rearing spiders in the insectary using artificial diets. This finding of nectivorous feeding behavior inspired us to compare the survival of spiders under different artificial diets. For this study the hunting spider Hibana (= Aysha) velox (Becker 1879) (Any- phaenidae) was selected because it was found to be the dominant species in lime {Citrus au- rantifolia [Christm.] Swingle 1914) orchards in south Dade County, Florida. Also, its spider- lings were observed to feed on the larvae of citrus leafminer, Phyllocnistis citrella Stainton 1856. Voucher specimens of H. velox and P. citrella are deposited at the Division of Plant Industry (DPI), Gainesville, Florida. To test the effects of the different artificial diets on the survival of H. velox, egg sacs were collected in the field and brought into the laboratory and kept until the eggs hatched. The resulting offspring were used for the ex- periment. Each spiderling was maintained in a separate container as described by Peck & Whitcomb (1968) with some modifications to prevent cannibalism. Instead of glass tubing with both ends open, common laboratory glass vials (15 mm diameter X 60 mm long) were utilized. The open end was plugged with cotton through which a stemmed cotton swab had been inserted. The cotton swab inside the glass vial was saturated with the artificial diet by dipping. Three different artificial diets were included in the experiment: 30% sucrose solution (cane sugar, Publix Supermarket Inc., Lakeland, Florida); milk + yolk mixture (1 cup homogenized milk + 1 fresh chicken egg yolk), and soybean (non-dairy beverage). Wa- ter served as the control. Nutritional compo- sition of the milk + yolk and soybean diets 692 AMALIN ET AL.— DIFFERENTIAL ARTIFICIAL DIETS 693 Table 1. — Nutritional composition of milk + yolk and soybean diets based on the manufacturer’s nu- tritional analysis and given as amount per 100 ml of media. Nutrient composition (per 100 ml of media) Milk + yolk Soybean Total fat 3.0 g 13 g Saturated fat 13 g 0.0 g Cholesterol 98.0 mg 0.0 mg Sodium 83.0 mg 39,0 mg Total carbohydrates 6.1 g 11.0 g Sugars 5.2 g 6.5 g Protein 6.1 g 2.6 g Potassium 0.0 126.0 mg Vitamin A 350 lU 0.0 lU Thiamin (Bl) 0.0 0.05 mg Riboflavin 0.0 0.03 mg Niacin 0.0 0.52 mg Pantothenic acid 0.0 0.35 mg Pyridoxine hydrochloride 0.0 0.05 mg Folate 0.0 0.02 mg Vitamin C 0.52 mg 0.0 Vitamin D 44 lU 0.0 Biotin 0.0 2.6 (xg Calcium 139.0 mg 26.0 mg Iron 0.31 mg 0.31 mg Phosphorus 0.11 g 0.04 g Magnesium 0.0 17.4 mg Zinc 0.0 0.26 mg are provided (Table 1). Twenty spiderlings were included for each artificial diet. Two rep- lications in time were prepared and kept at 27 °C in different rearing chambers, one at 45% relative humidity (RH) and the other at 80% RH. The two RH conditions were chosen based on work by Peck & Whitcomb (1968) and Taylor & Foster (1996). At 45% RH, all the spiders on all of the diets died in less than 30 days from the start of the experiment. In contrast. Peck & Whit- comb (1968) reported best survival [42 days for Chiracanthium inclusum (Hentz 1847) and 90 days for Gladicosa (reported as Lycosa gu- losa (Walckenaer 1837)] when the spiders were kept at 45% RH. There was no signifi- cant difference for the age at death of H. velox on different diets. The mean age at death of the spiders kept at 45% RH was 13 days (range, 7=18) on soybean diet; 10 days (range, 8-11) on milk + yolk diet; 12 days (range, 7- 16) on 30% sucrose solution; and 11 days (range, 7-14) on water. Spiders kept at 80% RH survived longer, especially spiders on soy- bean and milk + yolk diets. In 30 days, the percent survival of spiders on soybean diet (82.5%) was significantly higher (P < 0.05) than on milk + yolk diet (46%). Spiders raised on 30% sucrose solution and water did not survive for the duration of the experiment. The mean age at death of spiders on sucrose solution was 14 days (range, 5-21); whereas for spiders on water, it was 1 1 days (range, 8- 12). On both soybean and milk + yolk diets, the first mortality occurred at 6 days after the start of the experiment. A drastic increase in mortality was observed on milk + yolk diet from day 6 to day 16 after the start of the experiment; the survival endpoint was reached at day 17 (Fig. 1). Mortality was less on soy- bean diet; the survival endpoint was at day 12 (Fig. 2). The single mortality at day 21 was due to fungal contamination. Although the percent survival on soybean diet was significantly higher (P < 0.05) than on milk + yolk diet, development of the spi- ders seemed to be delayed. This observation was based on percent molting, time of molt- ing, and carapace width of the spiders. Spiders raised on milk T yolk diet underwent two molts 30 days after the start of the experiment (Fig. 1). The earliest molt was 6 days after the % Survival 3 % Survival 694 THE JOURNAL OF ARACHNOLOGY 100 80 60 40 20 0 Days Figure 2. — Percent survival and molting of Hibana velox using soybean as artificial diet. % Molting a % Molting AMALIN ET AL.— DIFFERENTIAL ARTIFICIAL DIETS 695 start of the experiment; the mean age at first molt was 17 days (range, 6-30). The mean age at second molt was 25 days (range, 20- 30). On soybean diet, the molting of the spi- ders was late compared to the spiders on milk + yolk diet. The first molt was at 24 days and only 40% of the surviving spiders molted 30 days from the start of the experiment (Fig. 1). The average carapace width of spiders raised on milk + yolk diet was 0.70 mm (range, 0.50-0.85), whereas spiders on soybean diet had an average carapace width of 0.50 mm (range, 0.35-0.58). In general, the carapace width of spiders on milk + yolk diet was more than 25% greater than that of spiders on soy- bean diet. These findings suggest the impor- tance of supplying more complete nutritional requirements when rearing spiders using arti- ficial diets. The soybean diet is devoid of cho- lesterol (Table 1) which is the common source of sterol. It was reported that cholesterol is a precursor of ecdysone, the molting hormone (Foelix 1982). This may explain the delayed development of spiders on soybean diet. The milk + yolk diet has a high level of choles- terol, probably contributing to the normal pro- gress of spider development. Nevertheless, the high level of carbohydrates in the soybean diet (Table 1) suggests that carbohydrate could be an important component of the artificial diet for spiders. Carbohydrates are the major en- ergy source important for survival or longev- ity of any arthropod species (Singh 1984). In this experiment, we observed that the percent survival of spiders on soybean diet was almost twice that on milk + yolk diet. Furthermore, the drastic increase in mortality of spiders on milk + yolk diet in the first two weeks of rearing the spiders may be avoided if enough carbohydrate is available at that stage of development. From the result of this experiment, we can hypothesize that a com- bination of soybean and milk + yolk diets could provide more balanced nutritional re- quirements for wandering spiders. Thus, an experiment to assess the performance of com- bining soybean and milk + yolk diet on spider survival and development is underway. LITERATURE CITED Barrion, A.T & J.A. Litsinger. 1995. Riceland Spi- ders of South and Southeast Asia. CAB Inter- national, Wallingford, UK. 701 p. Culin, J.D. & K.V. Yeargan. 1983. Comparative study of spider communities in alfalfa and soy- bean ecosystems: Foliage-dwelling spiders. Ann. Entomol. Soc. America, 76:825-831. Dean, D.A., W.L. Sterling & N.V. Homer. 1982. Spiders in eastern Texas cotton fields. J. Arach- noL, 13:111-120. Dondale, C.D., B. Parent, & D. Pitre. 1979. A 6- year study of spiders (Araneae) in a Quebec ap- ple orchard. Canadian Entomol., 111:377-380. Foelix R. 1982. Biology of Spiders. Harvard Univ. Press, Cambridge. 306 p. Greenstone, M.H. 1979. Spider feeding behaviour optimizes dietary essential amino acid composi- tion. Nature, 282:501-503. Mansour, F., J.W. Ross, G.B. Edwards, W.H. Whit- comb & D. Richmae. 1982. Spiders of Florida citrus groves. Florida Entomol., 65:514-522. Mansour, EA., D.B. Richman & W.H. Whitcomb. 1983. Spider management in agroecosystems: Habitat manipulation. Environ. Manag., 7:43-49. Graze, M.J. & A. A. Grigarick. 1989. Biological control of aster leafhopper (Homoptera: Cicadel- lidae) and midges (Diptera: Chironomidae) by Pardosa ramulosa (Araneae: Lycosidae) in Cal- ifornia rice fields. J. Econ. Entomol., 82:745- 749. Peck, W.B. & W.H. Whitcomb. 1968. Feeding spi- ders on artificial diet. Entomol. News, 79:233- 236. Riechert, S.E. 1974. Thoughts on the ecological significance of spiders. BioScience, 24:352-356. Riechert, S.E. & J.Luczak. 1982. Spider foraging: Behavioral responses to prey. Pp. 353-384, In Bi- ology of Spider Communication: Mechanisms and Ecological Significance. (P.N. Witt & J. Rov- ner, eds.), Princeton Univ. Press, Princeton, New Jersey. 440 p. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: Spiders in garden test systems. Ecology, 71:1441-1450. Riechert, S.E. & T. Lockley. 1984. Spiders as bi- ological control agents. Ann. Rev. Entomol., 29: 299-320. 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.). Agric. Res. Serv. (Southern Region), U.S. Dept, of Ag- riculture, New Orleans. 306 p. Taylor, R.M. & W.A. Foster. 1996. Spider nectar- ivory. American Entomol., 82-86. Uetz, G.W., J. Bischoff & J. Rover. 1992. Survi- vorship of wolf spiders (Lycosidae) reared on different diets. J. ArachnoL, 20:207-221. Whitcomb, W.H. 1967. Wolf and lynx spider life histories. In Terminal report to National Science Foundation, Univ. of Arkansas. Div. Agric. Dept, of Entomology. Fayetteville, Arkansas. 142 p. Whitcomb, W.H., H. Exline & R.C. Hunter. 1963. 696 THE JOURNAL OF ARACHNOLOGY Spiders of the Arkansas cotton field. Ann. En- tomol. Soc. America, 56:653-660. D*M. Amalin^, J. Reiskind\ R. Mc- Sorley^and J. Pena^: ^xrechFAS, SW 280th St., Homestead, Florida 33031 USA; "'Dept, of Zoology, University of Florida, Gainesville, Florida 32611 USA; "^Bldg. 970, Hull Road, Dept, of Entomology and Nematology, University of Florida, Gaines- ville, Florida 32611 USA Manuscript received 1 April 1997, revised 27 No- vember 1998. INSTRUCTIONS TO AUTHORS (revised October 1999) Manuscripts are preferred in English but may be ac- cepted in Spanish, French or Portuguese subject to availability of appropriate reviewers. Authors whose pri- mary language is not English may consult the Associate Editor for assistance in obtaining help with English manuscript preparation. All manuscripts should be pre- pared 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 1500 words should be prepared as Feature Articles, shorter papers as Research Notes. Send four identical copies of the typed material together with photocopies of illustra- tions to the Managing Editor. Do not send original handmade illustrations until the manuscript has been ac- cepted. Mail to: Petra Sierwald, Managing Editor; Division of In- sects, Dept, of Zoology, The Field Museum of Natural History, 1400 South Lakeshore Drive, Chicago, IL 60605 USA [Telephone: (312) 665-7744; FAX: (312) 665-7754; E-mail: psierwald@finnh.org] The Managing Editor will either forward your man- uscript to one of the associate (subject) editors for the review process or initiate the review process directly. You will receive correspondence aclmowledging the receipt of your manuscript by the responsible associate or managing editor with the manuscript number of your manuscript. Please use this number in all corre- spondence regarding your manuscript. Correspondence regarding the review process should be directed to the respective associate or managing editor. Correspond- ence regarding all other matters should be directed to the Editor. After the manuscript has been accepted, the author will be asked to submit the manuscript on a computer disc. It must be in a widely-used program. Indicate clearly on the computer disc the word pro- cessing program used and the type of computer (Mac or PC). FEATURE ARTICLES Title page. — The title page will include the com- plete name, address, and telephone number of the au- thor 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 capital letters should be placed at the beginning of the first paragraph set off by a period. In articles written in English, a second abstract in an acceptable language may be included pertinent to the nationality of the author(s) or geo- graphic region(s) emphasized in the article. Text. — Double-space text, tables, legends, etc. throughout. Three levels of heads are used. The first level (METHODS, RESULTS, etc.) is typed in capitals and on a separate line. The second level head begins a paragraph with an indent and is separated from the text by a period and a dash. The third level may or may pot begin a paragraph but is italicized and sepa- rated 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 regardless of language of the text. 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). Literature cited section. — Use the following style: Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid composition of major ampullate gland silk (dragline) of Nephila clavipes (Araneae, Tetragnath- idae). J. Arachnol., 18:297-306. Krafft, B. 1982. The significance and complexity of communication in spiders. Pp. 15-66, In Spider Communications: Mechanisms and Ecological Sig- nificance. (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. Running head. — The author surname(s) and an ab- breviated 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 con- tact the Editor. Tables. — Each table, with the legend above, should be placed on a separate manuscript page. Only hori- zontal lines (usually three) should be included. Use no footnotes; instead, include all information in the leg- end. Make notations in the text margins to indicate the preferred location of tables in the printed text. Illustrations. — Address all questions concerning il- lustrations to: James W. Berry, Editor-In-Chief; Dept, of Biolog- ical Sciences, Butler University, Indianapolis, Indi- ana 46208 USA [Telephone (317) 940-9344; FAX (317) 940-9519; E-mail: jwberry@butler.edu] All art work must be camera-ready for reproduction. In line drawings, pay particular attention to width of lines and size of lettering when reductions are to be made by the printer. Multiple photos assembled on a single plate should be mounted with only a minimum of space separating them. In the case of multiple illus- trations 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. The author should indicate whether the illustrations should be one column or two columns in width. The overall dimen- sions of original artwork should be no more than 1 1 inches (28 cm) ;ts 14 inches (36 cm). Photocopies in review manuscripts should be reduced to the exact measurements that the author wants to appear in the final publication. Larger drawings present greater dif- ficulty in shipping and greater risks of damage for which the JOA assumes no responsibility. Make no- tations in the text margins to indicate the preferred position of illustrations in the printed text. 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, figure legends, footnotes, ta- bles with legends, figures. Page charges, proofs and reprints. — There are no page charges, but authors will be charged for changes made in the proof pages. Authors will receive a reprint order form along with their page proofs. Reprints will be billed by the Allen Press. RESEARCH NOTES Instructions above pertaining to feature articles ap- ply also to research notes, except that abstracts and most headings are not used and the author’s name and address follow the Literature Cited section. CONTENTS The Journal of Arachnology Volume 27 Feature Articles Number 3 Revision and Cladistic Analysis of the Erigonine Spider Genus Sisicottus (Araneae, Linyphiidae, Erigoninae) by Jeremy A. Miller 553 Radiation of the Genus Dysdera (Araneae, Dysderidae) in the Canary Islands: The Island of Tenerife by Miquel A. Arnedo & Carles Ribera 604 Anterior Median Eyes of Lycosa tarentula (Araneae, Lycosidae) Detect Polarized Light: Behavioral Experiments and Electroretinographic Analysis by J. Ortega-Escobar & A. Munoz-Cuevas 663 Research Notes A New Species of the Spider Genus Zelotes (Araneae, Gnaphosidae) from California by Norman I. Platnick & Thomas R. Prentice 672 Harvestman (Opiliones, Gonyleptidae) Takes Prey from a Spider (Araneae, Ctenidae) by Jose Sabino & Pedro Gnaspini 675 Escape Behavior Mediated by Negative Phototaxis in the Scorpion Paruroctonus utahensis (Scorpiones, Vaejovidae) by Elizabeth A. Camp & Douglas D. Gaffin 679 Notes on Cyclosa insulana (Araneae, Araneidae) of Papua New Guinea by William J. McClintock & Gary N. Dodson 685 The Effect of Feeding History on Retreat Construction in the Wolf Spider Hogna helluo (Araneae, Lycosidae) by Sean E. Walker, Samuel D. Marshall & Ann L. Rypstra 689 Survival of the Hunting Spider Hibana velox (Araneae, Anyphaenidae), Raised on Different Artificial Diets by D.M. Amalin, J. Reiskind, R. McSorley & J. Pena 692 ) fa 3 9088 01059 4356