(ISSN 0161-8202) 6>L ^vVThe Journal of ARACHNOLOGY OFFICIAL ORGAN OF THE AMERICAN ARACH NOLOG 1C, SONIA/v TY 6VIV1I U LIBR^S^ VOLUME 35 2007 NUMBER 2 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF : James E. Carrel, University of Missouri-Columbia MANAGING EDITOR '. Paula Cushing, Denver Museum of Nature & Science SUBJECT EDITORS’. Ecology — Soren Toft, University of Aarhus; Systematics — Mark Harvey, Western Australian Museum; Behavior — Gail Stratton, University of Mississippi; Morphology and Physiology — Jeffrey Shultz, University of Maryland EDITORIAL BOARD'. Alan Cady, Miami University (Ohio); Jonathan Coddington, Smithsonian Institution; William Eberhard, Universidad de Costa Rica; Rosemary Gil- lespie, University of California, Berkeley; Charles Griswold, California Academy of Sci- ences; Marshal Hedin, San Diego State University; Herbert Levi, Harvard University; Brent Opell, Virginia Polytechnic Institute & State University; Norman Platnick, Ameri- can Museum of Natural History; Ann Rypstra, Miami University (Ohio); Paul Selden, University of Manchester (U.K.); Matthias Schaefer, Universitset Goettingen (Germany); William Shear, Hampden- Sydney College; Petra Sierwald, Field Museum; I-Min Tso, Tunghai University (Taiwan). The Journal of Arachnology (ISSN 0161-8202), a publication devoted to the study of Arachnida, is published three times each year by The American Arachnological Society. Memberships (yearly): Membership is open to all those interested in Arachnida. Subscrip- tions to The Journal of Arachnology and American Arachnology (the newsletter), and an- nual meeting notices, are included with membership in the Society. Regular, $40; Students, $25; Institutional, $125 . Inquiries should be directed to the Membership Secretary (see below). Back Issues: Patricia Miller, P.O. Box 5354, Northwest Mississippi Community College, Senatobia, Mississippi 38668 USA. Telephone: (601) 562-3382. Undelivered Is- sues: Allen Press, Inc., 810 E. 10th Street, P.O. Box 368, Lawrence, Kansas 66044 USA. THE AMERICAN ARACHNOLOGICAL SOCIETY PRESIDENT : Elizabeth Jakob (2005-2007), Department of Psychology, University of Massachusetts, Amherst, MA 01003 USA. PRESIDENT-ELECT’. Paula Cushing (2005-2007), Denver Museum of Nature & Sci- ence, Denver, CO 80205 USA. MEMBERSHIP SECRETARY’. Jeffrey W. Shultz (appointed), Department of Entomology, University of Maryland, College Park, MD 20742 USA. TREASURER’. Karen Cangialosi, Department of Biology, Keene State College, Keene, NH 03435-2001 USA. SECRETARY’. Alan Cady, Dept, of Zoology, Miami University, Middletown, Ohio 45042 USA. ARCHIVIST. Lenny Vincent, Fullerton College, Fullerton, California 92634 USA. DIRECTORS’. Gary Miller (2005-2007), Christopher Buddie (2005-2007), Jason Bond (2007-2009). PAST DIRECTOR AND PARLIAMENTARIAN. H. Don Cameron (appointed), Ann Arbor, Michigan 48105 USA. HONORARY MEMBERS’. C.D. Dondale, H.W. Levi, A.F. Millidge. Cover photo: Female Satrapanus grayi (Pseudoscorpiones, Chemetidae) endemic to Lord Howe Island, Australia. Photo by Mark Harvey. Publication date: 1 1 September 2007 © This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 2007. The Journal of Arachnology 35:215-226 GEOGRAPHICAL DISTRIBUTION OF TWO SPECIES OF MESOBUTHUS (SCORPIONES, BUTHIDAE) IN CHINA: INSIGHTS FROM SYSTEMATIC FIELD SURVEYS AND PREDICTIVE MODELS Cheng-Mln Shi,1’2 Zu-Shi Huang,1’2 Lei Wang,1’2 Li-Jun He,1’2 Yue-Pleg iiuaJ-2 Liang Leng1’2 and De-Xing Zhang13: 1 State Key Laboratory of Integrated Management of Pest Insects and Rodents, Institute of Zoology, Chinese Academy of Sciences, Beijing 100080, China; 2Graduate University of the Chinese Academy of Sciences, Beijing 100049, China ABSTRACT, Although Mesobuthus scorpions in China have become endangered in recent years, they are largely underinvestigated. Even the baseline data on their distributions are lacking. Here the geograph- ical distributions of two Mesobuthus scorpions in China are provided through a combined study of sys- tematic field surveys and GIS-based ecological niche modeling using 227 surveyed point occurrence data across an area of ca. 2800 X 1700 km2 and validated historical records. Mesobuthus martensii (Karsch 1879) appears to be restricted to latitude south of 43°N and the north side of the Yangtze River, bordered by the Helan Mountains and the Tengger and Mo Us sand desert in the west and limited by the sea in the east. Mesobuthus eupeus (C.L. Koch 1839) reaches the east side of the Helan Mountains and the west edge of the Loess Plateau, extending westward along the northern slope of the Qilian Mountains and ultimately penetrating to the northern part of the Junggar Basin. The former is mainly found in semi- humid and humid regions while the latter is an arid and semi-arid dweller. The two species show a parapatric distribution on the whole with a contact zone formed at the boundary of their ranges across the big turning of the Yellow River in the central-western part of Inner Mongolia, Ningxia and the middle part of the Gansu Province. This pattern of distribution is shaped both by the fundamental ecological niche constraint of the species and possibly by the biological interactions between the two species. Some diagnostic features for the two species are also provided for quick identification. Keywords : Mesobuthus martensii , Mesobuthus eupeus , ecological niche modeling, contact zone Formal description of the scorpion fauna of China began in 1 840’s (Gervais 1844). This was followed by occasional reports of new species, records, or amendments throughout the last two centuries mostly by non-Chinese scholars (Simon 1880; Karsch 1881; Pocock 1889; Kraepelin 1899, 1901; Birula 1898, 1904, 1911, 1917, 1925; Kishida 1939; Va- chon 1952; see Shi & Zhang 2005 for review). Wu (1936) was the first Chinese scholar who studied the scorpion fauna of China; since then there have been few additional reports in this field. In fact, scorpion taxonomy and bio- geography still remained a virgin field in Chi- na until very recently. This is rather surprising given that scorpions have long been used in traditional Chinese medicine (e.g., it was ai- 3 Corresponding author. E-mail: dxzhang@ioz. ac.cn - •• I ready well described in the medical code Peaceful Holy Benevolent Prescriptions com- piled between 978-992 AD in the Song Dy- nasty). Recently, Zhu et al. (2004) compiled a checklist of the scorpions In China, and Shi & Zhang (2005) reviewed scorpion system- atics with special emphasis on the buthids in the region. The species diversity of Chinese scorpions seems to be rather low in compari- son with scorpion faunas of other regions of the world. The nineteen species and subspe- cies reported in China as listed in Zhu et al. (2004) belong to 9 genera and 5 families. For comparison, 16 species belonging to 9 genera and 3 families were recorded from Israel and Sinai (Levy & Amitai 1980), and more than 60 species (11 genera and 4 families) in Baja California, Mexico and adjacent islands (Wil- liams 1980). Considering the large size of 215 216 THE JOURNAL OF ARACHNOLOGY China and the ecological diversity of its dif- ferent regions, it is quite possible that the scorpion fauna in China has been largely un- derestimated (Lourengo et ah 2005a). This suspicion was justified by the recent discovery of more than 10 new species from Tibet, Yun- nan, and Hainan Province (Qi et al. 2005; Lourengo et al. 2005a, 2005b). Therefore, ad- ditional new species can be expected, espe- cially in the arid areas. Among the 19 scorpion species and sub- species listed in Zhu et al. (2004), the Meso- buthus scorpions are the most well known species in China. The genus Mesobuthus Va- chon 1950, with some 15 species reported so far (Fet & Lowe 2000; Gantenbein et al. 2000; Lourengo et al. 2005a), is widespread in the Palaearctic Region from Turkey to Korea. In fact, Mesobuthus species are the most com- mon and abundant scorpions in a variety of arid habitats, from sand deserts to high moun- tains, with the centers of diversity in Central Asia and Iran (Fet et al. 2000; Gantenbein et al. 2003). Historically, three species had been reported in China, viz.: M. martensii (Karsch 1879) (Simon 1880; Karsch 1881; Pocock 1889; Birula 1911; Kishida 1939; Song et al. 1982; Qi et al. 2004), M. eupeus (C.L. Koch 1839) (Birula 1911, 1925; Schenkel 1936; Fet 1989, 1994), and M. caucasicus (Nordmann 1840) sensu Fet 1998 and Gantenbein et al. 2003 (Fet 1989, 1994). A fourth species, M. songi Lourengo et al. 2005 was just described from Tibet (Lourengo et al. 2005a). Of the aforementioned scorpions, M. martensii , com- monly called the “Chinese scorpion”, was the most studied and the most popularly used spe- cies in traditional Chinese medicine. Over the past decade, more than 70 different peptides, toxins, or homologues have been isolated from venom of this species (Goudet et al. 2002). Even so, its taxonomic status is still blurred due to dubious synonymization (Karsch 1881) and misleading discrimination where different type specimens were exam- ined (Pocock 1889). Another species, the mot- tled scorpion, M. eupeus is widely distributed with a dozen or so subspecies being recog- nized mainly based on coloration and mor- phometric characteristics (Fet 1994). Two sub- species, M. eupeus mongolicus (Birula 1911) and M. eupeus thersites (C.L. Koch 1839), oc- cur in northwest China (Birula 1911, 1925; Fet 1989, 1994; Zhu et al. 2004; Shi & Zhang 2005). This species has also been used in Chi- nese medicine in recent years (CMS & DXZ pers. obs.). Both species are now threatened by human overexploitation and habitat loss. Fortunately, M. martensii has now been listed in the China Species Red List and is consid- ered vulnerable (Wang & Xie 2005), a timely action for protecting this species in China. However, the lack of some basic knowledge about these species greatly impairs conserva- tion efforts. For example, distribution records of M. martensii often appear to be a broad- brush description such as “from Inner Mon- golia to Korean Peninsula.” The same treat- ment was given to M. eupeus in China but with an even vaguer outline. Thus, up-to-date baseline data on the geographical distribution of these species, which are fundamental for conservation planning (Ferrier et al. 2002; Funk & Richardson 2002; Rushton et al. 2004; Elith et al. 2006), are critically needed. During the last 6 years, we have conducted a systematic sample collection and an exten- sive field survey of the geographical distri- bution of two mesobuthid scorpions in China. This allows us to draw a detailed picture of the biogeographical distribution of these spe- cies and predict their potential distribution ranges using some ecological modeling meth- ods, such as the Genetic Algorithm for Rule- set Prediction (GARP, Stockwell & Nobel 1992), in a geographic information system (GIS) environment. Here we report the results of our combined study with emphasis on the field survey data, which are unique in scale and density in the study of the two Mesobu- thus species. METHODS Occurrence data and field survey. — The literature dealing with M. martensii and M. eupeus was consulted to extract distributional records, which served as the start point for our field survey. Extensive geographical surveys were carried out from May to September dur- ing 2001-2006, using the following strategies: first, the places with historical records were visited; then we extrapolated our survey out- ward from around the areas where scorpions were successively collected, according to the similarity of the topography and climatic pa- rameters in 50-100 km intervals; finally, we explored the areas of different landscape ad- jacent to the outermost sample localities. The SHI ET AL.— GEOGRAPHICAL DISTRIBUTION OF MESOBUTHID SCORPIONS 217 above survey was complemented by irregular collections from the putative distributional ar- eas. Scorpion collection was carried out either by a stone-rolling method in daytime or UV- light collection at eight (Williams 1968). The sample localities were positioned using a GPS receiver (Garmin International), with their longitude, latitude, and elevation recorded. The animals obtained were either deposited in 99.1% ethanol or brought back to the labora- tory alive. A subset of specimens was pre- served in 75% alcohol for subsequent mor- phological analysis under a Nikon SMZ1500 stereomicroscope. All the materials are depos- ited at the Laboratory of Molecular Ecology and Evolution, Institute of Zoology (MEE- IOZ), Chinese Academy of Sciences, Beijing. Modeling the species ? potential distribu- lion. — A species’ presence in a place can of- ten be assured when it was successfully sam- pled there, but its absence cannot be deduced simply from its not being collected in an area. Therefore, a field survey alone is not enough for delimiting a species’ distribution range. Modem computational and GIS technologies provide opportunities to use species’ pres- ence-only locality data to predict their poten- tial distributions. All the point occurrence data collected by field surveys and literature re- views were geo-referenced to the nearest 0.1° of latitude and longitude and used for ecolog- ical niche modeling. Ecological niche models use known occurrence point locations for a species and values for environmental variables (i.e., temperature and precipitation) at those point locations to generate an approximation of the fundamental ecological niche for that species. This ecological niche can then be pro- jected onto a map of the study area in order to predict where that species might occur. Based on ecological niche models, species’ distributions were modeled using the Genetic Algorithm for Rule- set Prediction (GARP) (Stockwell & Noble 1992). The occurrence points are divided evenly into training and test data sets. GARP models based on presence- only data, and absences are included in the modeling exercise via sampling of pseudo-ab- sence points from the set of pixels where the species has not been detected. A method is chosen from a set of possibilities (e.g., logistic regression, bioclimatic rules) and then applied to the training data to develop or evolve a rule through an iterative process of rale selection, evaluation, testing, and incorporation or rejec- tion (Peterson et al. 2006). Predictive accuracy is evaluated based on the testing data. The change in predictive accuracy between itera- tions was used as the criterion to evaluate whether particular rules should be incorporat- ed into the model. The algorithm runs 1000 iterations or until convergence. GARP modeling was carried out on Desktop- GARP (http//www.lifemapper.org/desktopgarp), which offers much improved flexibility in choice of predictive environmental data layers. The 14 environmental data layers, in the form of raster grids with 0.1° resolutions, were ob- tained from the website http//www.lifemapper. org/desktopgarp. An environmental layer jack- knifing procedure is performed to determine what environmental factors are more significant or important than others for our species (Peter- son & Cohoon 1999). Based on the evaluation of commission and omission errors as well as accuracy values, the environmental layers are incorporated or discarded in subsequent mod- eling. To optimize model performance, we devel- oped 100 replicate models of ecological niche for each species. The 10 best models were se- lected from the replicate models according to the procedure developed by Anderson et al. (2003) using “Best Subset Selection Parame- ters” option with the “extrinsic” omission limit set to 10%. Model quality was tested via the testing data. Chi-squared tests were em- ployed to determine whether test points fall into regions of predicted occurrence more of- ten than expected by chance (Peterson 2001; Anderson et al. 2002a, b). Finally, these 10 models were imported into ARCVIEW GIS 3.2 (Environmental Systems Research Insti- tute, Inc. 1999) and overlaid to create one consensus predictive range map. Grid cells that were predicted as present by at least 9 of these models were extracted to give an opti- mal model for M. martensii and M. eupeus , respectively. Sympatric zone was obtained by superimposing the optimal models of the two species in ARCVIEW GIS 3.2 with spatial an- alyst extension. RESULTS The scorpions.— The adults of the two spe- cies, M. martensii and M. eupeus , can be eas- ily recognized based on size and coloration, 218 THE JOURNAL OF ARACHNOLOGY and the contrasts are especially obvious when they are observed together (for example, when they occurred at the same locale or were in the same collecting jars). Generally, M. mar- tens ii is larger, with carapace and mesosomal tergites yellowish-brown to blackish-brown. Metasomal segments I— IV are yellowish, while metasomal segment V has conspicuous darkish-brown to blackish spots, especially on the ventral and lateral surfaces. Mesobuthus eupeus is smaller, with the whole body yellow to yellowish-brown and with the coloration relatively uniform on the carapace, mesoso- mal tergites, and metasomal segments. Meta- somal V has only slightly brownish pigmen- tation. Mesosomal tergites often have irregular longitudinal blackish-brown stripes. In the laboratory, the two species can be easily distinguished by the following charac- ters: in M. martensii , the ventromedian cari- nae of the metasomal segment II and III are crenulate and the granules are evenly devel- oped; the ventrolateral carinae of the meta- somal segment V are serratocrenulate and the granules are evenly developed or slightly in- creasing in size posteriorly; the pedipalp chela is slender with Cl/Cw = 4.45 ± 0.23 (Cl = chela length, Cw = chela width, mean ± SD, n = 65) for the female and 3.71 ± 0.24 (n = 43) for the male, and both the fixed finger and the movable finger having 12-13 rows of oblique granules. In M. eupeus , the ventro- median carinae of the metasomal segment II and III are serratocrenulate with the granules increasing in size posteriorly; the ventrolateral carinae of the metasomal segment V are cren- ulate, and the granules are irregularly increas- ing posteriorly with 1-3 of them significantly enlarged and lobate; the pedipalp chela is strong with Cl/Cw = 3.47 ± 0.25 ( n = 52) for the female and 3.23 ± 0.27 ( n = 26) for the male, the fixed finger and the movable fin- ger having 10 and 11 rows of oblique gran- ules, respectively. Occurrence and distribution. — Our sur- vey covers 16 provincial administrative re- gions and we succeeded in collecting scorpi- ons from 211 sites belonging to 174 counties across an area of ca. 2800 X 1700 km2. In- formation about the voucher specimens’ exact localities is available from authors on request for scientific purposes only. We reserve the right not to release these data for public re- view for conservation consideration. These two species have been overexploited for com- mercial purposes and these activities have be- come rampant in recent years (CMS & DXZ pers. obs.). Releasing our data will make the situation even worse. However, below we pre- sent broad information about the two species’ ranges. Mesobuthus martensii (Karsch 1879): his- torical records of the distribution of M. mar- tensii are rather cursory. Many literature sources gave some very broad descriptions, such as Northwest and/or North of China or just provincial names. We were only able to specify 12 localities to recognizable adminis- trative regions (Table 1), which are shown by triangles in Fig. 1. We have collected M. martensii from 175 sites belonging to 15 provincial regions: An- hui (2 sites), Beijing (3), Gansu (16), Hebei (17), Henan (16), Hubei (2), Inner Mongolia (4), Jiangsu (1), Liaoning (10), Ningxia (9), Qinghai (1), Shandong (46), Shaanxi (27), Shanxi (20), and Tianjin (1) (squares in Fig. 1). The northernmost locality where we have collected this scorpion is Beipiao (41.83°N, 120.61°E), Liaoning Province, and the west- ernmost sample site is Guide (36.0 1 °N, 101.40°E), Qinghai Province. We successively collected specimens on eight islands of Mia- odao Archipelago and from Wafangdian (39.37°N, 121.50°E) of Liaodong Peninsula. The elevations of the sample sites range from < 10 m (Shandong and Liaodong Peninsula) to more than 2300 m (Qinghai Province) above sea level. Almost all the sites lie in the mountain areas, great or small. Thus this spe- cies is a typical “mountaineer,” and shows strong lithophilism in the plain regions. On the Loess Plateau the species may be found in the crevices and some burrows of other in- vertebrates or small mammals and reptiles. On three occasions we found this species in hu- man dwellings. Mesobuthus eupeus (C.L. Koch 1839): doc- umented localities of this species in China are very few (Table 1). Two subspecies had been recorded. The subspecies M. eupeus mongo- licus was described from Gansu, Gobi-Altai and Alashan regions by Birula in 1911, and later reported from Tianshan (Tien-shan) Mountains near Urumchi (Schenkel 1936), Xinjiang Uygur Autonomous Region. Birula (1911, 1925) had recorded that this subspecies was found in Lanzhou, Gansu Province, but SHI ET AL . — -GEOGRAPHICAL DISTRIBUTION OF MESOBUTHID SCORPIONS 219 Table 1. — -Historical occurrences of Mesobuthus martensii and M. eupeus in China, Species Locality Lat. °N Long. °E References M martensii Beijing (Pekin) 39.9 116.4 Simon 1880, Karsch 1881, Pocock 1889 Yantai (Tchefou) Shandong 37.5 121.4 Simon 1880, Pocock 1889, Wu 1936 Tianjin. (Tientsin) 39.1 117.2 Karsch 1881 Alashan, Inner Mongolia 38.8 105.7 Birula 1911 Wulingshan, Hebei 40.6 117.4 Kishida 1939 Chaoyang, Liaoning 41.5 120.4 KIshida 1939 Lingyuan, Liaoning 41.2 119.3 Kishida 1939 Chengde, Hebei 40.7 118.1 Kishida 1939 Kaifeng, Henan 34.7 114.4 Wu 1936 Xuanhua, Hebei 40.6 115.0 Wu 1936 Donghai, Jiangsu 34.5 118.7 Wu 1936 Changshandao, Shandong 37.9 120.7 Wu 1936 M. eupeus Gobi-altai, Mongolia 42.5 104.0 Birula 1911 Alashan, Inner Mongolia 40.5 103.0 Birula 1911 Lanzhou, Gansu (Lantsho-fu, Kan-su) 36.0 103.7 Birula 1911, 1925 Tianshan, Xinjiang (TIen-shan Urimchi) 43.8 87.6 Schenkel 1934 we failed to find this species there or in ad - jacent areas whereas M. martensii were found instead (Lanzhou: 36.12°N, 103.72°E; Yu- zhong: 36.35°N, 104.28°E; and Gaolan: 36.33°N, 103.9G°E). Another subspecies, M. eupeus thersites , was also observed in north- west China (Fet 1989, 1994), Here we just Identified specimens to species, with no effort to further identify them to subspecies. We collected M. eupeus from 36 sites (cir- cles in Fig. 2), spanning Gansu (12 sites), Ni- ngxia (12), Xinjiang (1) and Inner Mongolia (11). The southernmost site is Jingyuan (36.50°N, 104.60°E; Gansu), and the eastern- most one is Urad Qianqi (40.67°N, 108.73°E; Inner Mongolia). The majority of sites lie in the Gobi desert and bald mountains. There are three exceptions: one site is in a deserted hu- man dwelling, one site lies in the interdunal zone of sand desert, and the third is in a de- graded pasture. Contact Zone; In 8 localities (bisected cir- cles in Figs. 1—3) we collected M. martensii and M. eupeus at the same site: Jingyuan (36.5°N, 104. 6°E) and Jingtai (37.2°N, 104 3°E) of Gansu province; Qingtongxia (37.7°N, 106. 0°E), Siyanjing (37.7°N, 105.7°E) and Shikong (37.7°N, 105.5°E) of Niegxia Autonomous Region; and Alaska n Zuoqi (37.8°N, 105. 5°E), Urad Qianqi (40.7°N, 108. 7°E) and Urad Zhongqi (41.3°N, 108.6°E) of Inner Mongolia. All these sites are near or on the lithe-mountains. Prediction of potential distribution and sympatrle zone. — In total, 227 occurrence points, 187 for M. martensii and 40 for M. eupeus , were used for modeling the potential distributions. Of the 14 environmental layers six were excluded from modeling because of their high omission and commission errors and low predictive accuracy. The eight envi- ronmental layers used in the predictive mod- eling are annual means of frost days, solar ra- diation, precipitation, minimum temperature, mean temperature, maximum temperature, water vapor pressure and wet days. The chi- square test yielded significant results for all the models produced (P < 10-34 for M. mar- tensii and P < 1()"5 for M. eupeus ). Projections of models onto maps permit vi- sualization of the ecologically potential distri- bution ranges of the two species. Thus, M. martensii appears restricted to the latitude south of 43°N and to the north side of the Yangtze River, bordered by the Helan Moun- tains in the west and limited by the Yellow Sea and the East China Sea in the east (Fig. 1); an area covering the North China Plain in the east, the Loess Plateau in the west, and the Liaohe Plain in the north. The occurrence points for M. eupeus are concentrated In a relatively small sub- area of 220 THE JOURNAL OF ARACHNOLOGY Kazakhstan Mongolia jp/rgyzst; Korea China Korea Burma Sanglaaesh Thailand filippmes rambo< Mongolia £? Liaoning Inner Mongolia Sharixt 100 200km Figure 1. — Ecological niche-based prediction of potential distribution of M. martensii. Squares, present data; triangles, historical records. Bisected circles show the sites where M. martensii and M. eupeus co- occurred. SHI ET AL.— GEOGRAPHICAL DISTRIBUTION OF MESOBUTHID SCORPIONS 221 Figure 2. — Ecological niche-based prediction of potential distribution of M. eupeus. Circles, present data; triangles, historical records. Bisected circles show the sites where M. eupeus and M. martensii co- occurred. its range, being less optimal for ecological niche modeling. A circular potential distribu- tion in the region 36-48°N and 94-120°E (Fig. 2) was predicted, plus some patchy ap- pearance in Xinjiang, west Mongolia, and Ka- zakhstan and Kyrgyzstan. This is in accor- dance with historical records of this species there. Outside China, this species ranges from central Anatolia through Caucasus and Turkey to Mongolia and has been found in Afghani- stan, Armenia, Azerbaijhan, Georgia, Iran, Iraq, Kazakhstan, Kyrgyzstan, Pakistan, Rus- sia (Astrakhan region), Tajikistan, Turkmen- istan, and Uzbekistan (Fet 1989, 1994; Kara- tas & Karatas 2001, 2003). Superimposition of the two species models reveal limited areas of potential sympatry (Fig. 3). The potential sympatric zone lies across the big turning of the Yellow River in central-western Inner Mongolia, Ningxia and the middle part of the Gansu Province. Alto- gether, 24 sites lie in the potential sympatric zone. Ten of those represent collections of M. eupeus and six of those M. martensii. The re- maining eight sites where two species co-oc- curred fell exactly into the predicted sympat- ric zone. DISCUSSION Our data indicate that M. martensii and M. eupeus show a parapatric distribution with a contact zone formed at the boundary of their ranges. Mesobuthus martensii mainly occurs in the humid and semi-humid regions where the mean annual rainfall exceeds 300 mm. Its range spans the so-called second and third geographical cascades of mainland China, with the average altitude descending eastward from more than 2500 m to less than 10 m above sea level, and includes the littoral zone (Remy & Feroy 1933; Millot & Vachon 1949) and some islands of Bohai Sea. The northern limit of the Chinese scorpion might not ex- ceed 43°N. This is in accordance with Kishida (1939), who recorded that there was no scor- pion in Ongniud Qi (42.9°N, 119.0°E). The 222 THE JOURNAL OF ARACJJNOLOGY Figure 3. — Ecological niche-based prediction of potential range of the contact zone between M. mcir- tensii and M. eupeus. Squares, M. martensii ; circles, M. eupeus. Bisected circles show the sympatric sites. Tengger and Mo Us sand desert constitute the northwestern distributional boundary for M. martensii. In contrast, M. eupeus is an arid dweller living in the arid and semi-arid environment, as recorded in the literature. All our sampling sites are in or near the desert regions, ranging from the central western area of Inner Mon- golia to western Gansu Province, and pro- ceeding further west- and northward to the Al- tai region, Xinjiang. Although our data about M. eupeus are relatively restricted, our models indicated the potential distribution of M. eu- peus, especially in China, ranges from the east side of the Helan Mountains, expanding west- ward along the northern slope of the Qilian Mountains, ultimately penetrating to the northern part of the Junggar Basin. We think the sample sites presented here largely cover the whole range of its distribution in China. This species is abundant in the central-western part of Inner Mongolia and Gansu and the west part of Ningxia where the average rain- fall is less than 300 mm or even less than 1 00 mm in some areas. Distribution patterns of species are shaped by a number of factors, including barriers to dispersal, physical and biological factors that make particular regions of habitat unsuitable for viability and/or reproduction, etc. (Burton 1998). Potential distribution predicted by GARP ecological niche modeling relied on the fundamental niche (Soberon & Peterson 2005). The actual geographic distribution is a modification of the fundamental niche because it is defined by the complex interaction of the realized environment (as well as some biolog- ical and historical realities) and the funda- mental niche (Brown et al. 1996; Patterson 1999; Peterson et al. 1999; Anderson et al. 2002a, b; Anderson & Martlnez-Meyer 2004). This may be why some regions in Central Asia are predicted to be suitable for M. mar- tensiii, but it is unlikely that this species could have occurred there in reality. There is no re- port of M. martensii from this region and it cannot escape the attention of the scorpiolo- gists’ investigation given that it is a “hotspot” region of Mesobuthus. Other explanations also exist for its absence: 1) the populations there have gone locally extinct and 2) M. martensii has failed to disperse to this region because the existence of M. eupeus acts as an effective SHI ET AL.— GEOGRAPHICAL DISTRIBUTION OF MESOBUTHID SCORPIONS 223 biological barrier preventing M. martensii from dispersing further west. Biological interaction as a constraint for species distribution is possible in the case of M. eupeus. Ecological models predict that M. eupeus can survive further south and east in the middle part of Inner Mongolia, but this species is restricted to the northwest possibly due to the existence of M. martensii. We sug- gest that In the central western region of Inner Mongolia, NIngxia and the middle part of Gansu the two species mutually limit the range expansion of their counterparts, proba- bly due to the competition for food and other resources (such as shelters). The absence of M. martensii in the area north of mid-Liaoning may be the result of population extinction due to some undefined reasons in relatively recent time. This is in accordance with our field survey at Tiding (42.3°N, 123.8°E) where we failed to spot a scorpion. However, several elderly residents ascertained the occurrence of the scorpion there when they knocked down old houses be- fore the 1970’s, but they have not seen any since the 1980’s. The abuse of insecticides and other poisonous chemicals may be an expla- nation for Its disappearance to some extent, but it cannot account for all. Scorpions are nocturnal obligatory predators but with poor optical sensitivity, detecting their prey through substrate vibration using the tarsal sensilla (Brownell 1977; Brownell & Farley 1979a, b, c; Brownell & Hemmen 2001; Foelix & Scha- bronath 1983) and/or via air borne vibrations by the trichobothria on the pedipalps (Hjelle 1990). These biological features effectively reduced their exposure to poison because the scorpions will not prey on the poisoned dead creatures that produce no vibration. Actually, M. martensii is abundant under the fence stones of farmland in Haiyang (36.83°N, 12L03°E), Shandong Province. Some other anthropogenic activities, such as house re- building and large-scale soil cultivating, which cause habitat loss and recent exception- al climate change may serve as alternative ex- planations for the absence of this species in Tieling. This deserves further investigation. The two Mesohuthus scorpions were found on either side of the Yellow River. The ma- jority of sympatric sites (7 of 8) lie on the outer (western and northern) bank of the river (Fig. 3). This observation suggests that the Yellow River does not constitute an effective barrier for the two species and the present pat- terns of distribution are the results of histori- cal processes, an interesting topic meriting further investigation by phylogeographical ap- proaches. Our results cast further doubt about the identity of M. martensii (Kishida 1939; Qi et al. 2004). The type specimens upon which M. confucius was described by Simon (1880) were collected from Yantai (Tchefou) and Bei- jing (Pekin). With no convincing evidence, M. confucius was then synonymized with M. martensii by Karsch (1881) when he exam- ined specimens collected from Beijing and Tianjin. The holotype locality of M. martensii is presumably Singapore (Karsch 1879). How- ever, none of our predicting models showed that Singapore is a suitable area for the sur- vival of M. martensii. There are two possible interpretations: 1) the type specimen of Karsch (1879) represents a different taxon from that found in Singapore but the name (then Buthus martensii ) was wrongly given to the Chinese scorpion, i.e., M. martensii and M. confucius are two different species; 2) the type specimen of Karsch (1879) is of Chinese origin. Karsch (1879) recorded “Exemplum singulum typicum in Mus Berol. asservatum a Prof, de Martens in Singapore collectum”; there was the possibility that the specimen was initially transported by Chinese immi- grants from China to Singapore possibly for medical use. It is a pity that neither Qi et al. (2004) nor we could have examined that spec- imen; thus this issue remains to be clarified In the future. ACKNOWLEDGMENTS We would like to thank the following friends, schoolmates and colleagues for their assistance with various aspects of field survey: Jia Chen, Hoegqiang Dong, Yujun Dong, Tinggui Feng, Ruidong Han, Ruicai Huo, Ya- jie Ji, Jim I m Kang, Jun Lei, Jiyuan Li, Huatao Liu, Fei Lu, Yudi Liu, Chengjun Lti, JIancang Ma, Conghai Tang, Lijuan Tang, Duohong Wang, Junpieg Wang, Ronghua Wang, En Wu, Yuchue Wu, Yongfeng Xu, Ruisheng Yang, Bowel Zhang, Deyi Zhang, Rui Zhang, X .ni- che ng Zhao, Guangjian Zhu. We thank Pro- fessors Hongzhaeg Zhou and Aiping Liang for providing specimens from Inner Mongolia and Tibet. We are grateful to Professors Dax- 224 THE JOURNAL OF ARACHNOLOGY iang Song and Suigong Yin for providing valuable information about scorpion research- es in China, Professors Naxin Bei and Yu- anhua Wu for accommodation during our field survey and Professor Mingsheng Zhu for help with some publications. We thank Dr. Spren Toft, Dr. Victor Fet and an anonymous referee for their constructive comments on an earlier version of the manuscript and Dr. Fet, in par- ticular, for linguistic improvement. We are in debt to Paula Cushing, the journal’s managing editor, for the thorough linguistic and forrnat- and-style corrections on the manuscript. This research was supported by the Natural Science Foundation of China (NSFC grant no. 30570246), the CAS Knowledge Innovation Program (grant no. KZCX2-YW-428) and the CAS “Bai Ren Ji Hua” Professorship. LITERATURE CITED Anderson, R.P., M.P. Gomez-Laverde & A.T. Peter- son. 2002a. Geographical distributions of spiny pocket mice in South America: insights from predictive models. Global Ecology and Bioge- ography 11:131-141. Anderson, R.P., A.T. Peterson & M. Gomez-Lav- erde. 2002b. Using niche-based GIS modeling to test geographic predictions of competitive exclu- sion and competitive release in South American pocket mice. Oikos 98:3-16. Anderson, R.P., D. Lew & A.T. Peterson. 2003. Evaluating predictive models of species’ distri- butions: criteria for selecting optimal models. Ecological Modeling 162:211-232. Anderson, R.P. & E. Martfnez-Meyer. 2004. Mod- eling species’ geographic distributions for pre- liminary conservation assessments: an imple- mentation with the spiny pocket mice (. Herteromys ) of Ecuador. Biological Conserva- tion 116:167-179. Birula, A. A. 1898. Miscellancea scorpiologica. III. Zur Synonymic der russischen Skorpione (Fort- setzung). Annuaire du Musee Zoologique de 1’ Academic Imperiale des science de St. Peters- bourg 3:295-299. Birula, A. A. 1904. Miscellanea scorpiologica. VI. Ueber einige Buthus- Arten Centrasiens nebst i hir- er geographischen Verbreitung. Annuaire du Mu- see Zoologique de 1’ Academic Imperiale des Sci- ences de St.-Petersbourg 9:20-27. Birula, A. A. 1911. Arachnologische Beitrage. I. Zur Scorpionen- und Solifugen-Fauna des Chine- sischen Reiches. Revue Russe d’Entomologie 11(2): 195-199. Birula, A. A. 1917. (Byalynitskii-Birulya A. A.) Ar- throgastric Arachnids of Caucasia. Part 1. Scor- pions. Annals of the Caucasian Museum. Series A, No. 5. Translated from Russian. Israel Pro- gram for Scientific Translation, Jerusalem 1964. 170 pp. Birula, A. A. 1925. Skorpiologische Beitrage. 10. Zur geographischen Verbreitung zweier weitver- breiteter Skorpionen- Arten Palaearcticums. Zool- ogischer Anzeiger 63(l-2):93-96. Brown, J.H., G.C. Stevens & D.M. Kaufman. 1996. The geographic range: size, shape, boundaries, and internal structure. Annual Review of Ecolo- gy and Systematics 27:597-623. Brownell, P.H. 1977. Compressional and surface waves in sand: used by desert scorpions to locate prey. Science 197:479-482. Brownell, P.H. & R.D. Farley. 1979a. Prey-local- izing behaviour of the nocturnal desert scorpion, Paruroctonus mesaensis : orientation to substrate vibrations. Animal Behaviour 27:185-193. Brownell, P.H. & R.D. Farley. 1979b. Detection of vibrations in sand by tarsal sense organs of the nocturnal scorpion, Paruroctonus mesaensis. Journal of Comparative Physiology A 131:23— 30. Brownell, P.H. & R.D. Farley. 1979c. Orientation to vibrations in sand by the nocturnal scorpion Paruroctonus mesaensis : mechanism of target localization. Journal of Comparative Physiology A 131:31-38. Brownell, P.H. & J.L. van Hemmen. 2001. Vibra- tion sensitivity and a computational theory for prey-localizing behavior in sand scorpions. American Zoologist 41:1229-1240. Burton, R. 1998. Intraspecies phylogeography across the Point Conception biogeographic boundary. Evolution 52:734-745. Elith, J., C.H. Graham, R.P. Anderson, M. Dudik, S. Ferrier, A. Guisan, R.J. Hijmans, F. Huett- mann, J.R. Leathwick, A. Lehmann, J. Li, L.G. Lohmann, B.A. Loiselle, G. Manion, C. Moritz, M. Nakamura, Y. Nakazawa, J.M. Overton, A.T Peterson, S.J. Phillips & N.E. Zimmermann. 2006. Novel methods improve prediction of spe- cies’ distribution from occurrence data. Ecogra- phy 29:129-151. Ferrier, S., G. Watson, J. Pearce & M. Drielsma. 2002. Extended statistical approaches to model- ing spatial pattern in biodiversity in northeast New South Wales. I. Species-level modeling. Biodiversity and Conservation 11:2275-2307. Fet, V. 1989. A catalogue of scorpions (Chelicerata: Scorpiones) of the USSR. Rivista del Museo Civico di Scienze Naturali “Enrico Caffi” (Ber- gamo) 13(1 988):73 — 17 1 . Fet, V. 1994. Fauna and zoogeography of scorpions (Arachnida: Scorpiones) in Turkmenistan. Pp. 525-534. In Biogeography and Ecology of Turk- menistan (V. Fet & K.I. Atamuradov, eds.). Klu- wer Academic Publishers, Dordrecht, Nether- lands. SHI ET AL. — GEOGRAPHICAL DISTRIBUTION OF MESOBUTHID SCORPIONS 225 Fet, V. 1998. Case 3026. Androctonus caucasicus Nordmann, 1840 (currently Mesobuthus caucas- icus; Arachnida, Scorpiones): proposed conser- vation of the specific name. Bulletin of Zoolog- ical Nomenclature 55:14-16. Fet, Y, B.E. Hendrixson, W.D. Sissom & G. Levy. 2000. First record for the genus Mesobuthus Va- cfaon, 1950 in Israel: Mesobuthus nigrocinctus (Ehrenberg, 1828), comb. n. (Scorpiones: Butin idae) from Mt. Hermon. Israel Journal of Zool- ogy 46:287-295. Fet, V. & G. Lowe. 2000. Family Buthidae C.L. Koch, 1837. Pp. 54-286. In Catalog of the Scor- pions of the World (1758-1998) (V. Fet, W.D. Sissom, G. Lowe & M.E. Braunwalder, eds.). The New York Entomological Society, New York. Foelix, R.F. & J. Schabronath. 1983. The fine struc- ture of scorpion sensory organs. I. Tarsal sensilla. Bulletin of the British Araehnological Society 6: 53-67. Funk, Y. & K. Richardson. 2002. Systematic data in biodiversity studies: use it or lose it. System- atic Biology 51:303-316. Ganbentein, B , C. Kropf, C.R. Largiader & A. Scholl. 2000. Molecular and morphological evi- dence for the presence of a new Buthid taxon (Scorpiones: Buthidae) on the Island of Cyprus. Revue Suisse de Zoologie 107:213-232. Ganbentein, B., V. Fet & A.V. Gromov. 2003. The first DNA phylogeny of four species of Meso- buthus (Scorpiones: Buthidae) from Eurasia. Journal of Arachnology 31:412-420. Gervais, R 1844. Remarques sur la famille des scor- pions. Archives du Museum d’Histoire Naturelle, Paris 4:201-240. Goudet, C., C-W. Chi & J. Tytgat. 2002. An over- view of toxins and genes from the venom of the Asian scorpion Buthus martensii Karsch. Toxi- con 40:1239-1258. Hjelle, H.T. 1990. Anatomy and morphology. Pp. 9-63. In The Biology of Scorpions (G.A. Polls, ed.) Stanford University Press, Stanford, Cali- fornia. Karatas, A. & A. Karatas. 2001. First record of Me- sobuthus eupeus (C.L. Koch, 1893) from central Anatolia (Scorpiones: Buthidae). Pp. 297-299. In Scorpions 2001. In Memoriam Gary A. Polls (Y. Fet & P.A. Selden, eds.). British Araehnological Society, Burnham Beeches, Buckinghamshire, UK. Karatas, A. & A. Karatas. 2003. Mesobuthus eupeus (C.L. Koch, 1893) (Scorpiones: Buthidae) in Turkey. Euscorpius 7:1-5. Karsch, F. 1879. Scorpionologische Beitrage II. Mitteilunger der Miinchener Entomologische, Ver. 3(2):97— 136. Karsch, F. 1881. Chinesische Myriopoden und Ar- achnoiden. Berliner Entomologische Zeitschrift 25:219-220. Kishida, K. 1939. Arachnida of Jefaol, Scorpiones. Pp. 1-66. In Report of the First Scientific Ex- pedition to Manchoukuo. Sect. 5, div.l. Kraepelin, K. 1899. Scorpiones und Pedipalpi. Das Tierreich Deutslands 8:1-265. Kraepelin, K. 1901. Catalogue des Scorpions des collections du Museum d’Histoire Naturelle de Paris. Bulletin du Museum d’Histoire Naturelle, Paris 7(6): 265-274. Levy, G. & P. Amitai. 1980. Fauna Palaestina Arachnida I: Scorpiones. The Israel Academy of Sciences and Humanities, Jerusalem. 130 pp. Lourengo, W., J.-X. Qi & M.-S. Zhu. 2005a. De- scription of two new species of scorpions from China (Tibet) belonging to the genera Mesobu- thus Vachon (Buthidae) and Heterometrus Eh- renberg (Scorpionidae). Zootaxa 98:1-16. Lourengo, W., J.-X. Qi & M.-S. Zhu. 2005b. De- scription of a new species of Isometrus Ehren- berg 1828 (Scorpiones, Buthidae) from the island of Hainan, China. Boletin Sociedad Entomolo- gica Aragonesa 36:57-63. Millot, I. & M. Vachon. 1949. Ordre des Scorpions. Traite de Zoologie 6:386-436. Patterson, B.D. 1999. Contingency and determinism in mammalian biogeography: the role of history. Journal of Mammalogy 80:345-360. Peterson, A.T. 2001. Predicting species’ geographic distributions based on ecological niche modeling. The Condor 103:599-605. Peterson, A.T & K.P. Cohoon. 1999. Sensitivity of distributional prediction algorithms to geograph- ic data completeness. Ecological Modeling 117: 159-164. Peterson, A.T, J. Soberon & V. Sanchez-Cordero. 1999. Conservatism of ecological niches In evo- lutionary time. Science 285:1265-1267. Peterson, A.T, V. Sanchez-Cordero, E. Martinez- Meyer & A.G. Navarro-SIgtienza. 2006. Track- ing population extirpations via melding ecologi- cal niche modeling with land-cover information. Ecological Modeling 195:229-236. Pocock, R.I. 1889. Notes on some Buthidae, new and old. The Annals and Magazine of Natural History 3(6):334-351. Qi, J.-X., M.-S. Zhu & W. Lourengo. 2004. Rede- scription of Mesobuthus martensii martensii (Karsch, 1879) (Scorpiones: Buthidae) from Chi- na. Revista Iberica de Aracnologia 10:137-144. Qi, J.-X., M.-S. Zhu & W. Lourengo. 2005. Eight new species of the genera Scorpiops Peters, Eus- corpiops Vachon, and Chaerilus Simon (Scorpi- ones: Euscorpiidae, Chaerilidae) from Tibet and Yunnan, China. Euscorpius 32:1-40. Remy, P. & P. Leroy. 1933. Presence de Scorpions dans la zone subterrestre du littoral marin. Bui- 226 THE JOURNAL OF ARACHNOLOGY letin mensuel Societe Linneenne de Lyon 2(3): 3-42. Rushton, S.R, S.J. Ormerod & G. Kerby. 2004. New paradigms for modeling species distribu- tions? Journal of Applied Ecology 41:193-200. Schenkel, E. 1936. Schwedisch-chinesische wissen- schaftliche Expedition nach den nordwestlichen Provinzen Chinas. B. Scorpiones. Arkiv for Zoologi 27A(20):5. Shi, C.-M. & D.-X. Zhang. 2005. A review of the systematic research on buthid scorpions (Scor- piones: Buthidae). Acta Zootaxonomica Sinica 30:470-477. Simon, E. 1880. Arachnides recueills aux environs de Pekin. Annales Societe Entomologique de France, 5 serie, Tome Dixie, 97-128. Soberon, J. & A.T. Peterson. 2005. Interpretation of models of fundamental ecological niche and spe- cies’ distributional areas. Biodiversity Informat- ics 2:1-10. Song, D.-X., X.-X. Lii ;& J.-W. Shang. 1982. Mor- phology and biology of Buthus martensii Karsch, 1879. Bulletin of Biology 26:22-25. Stock well, D.R.B. & I.R. Nobel. 1992. Induction of sets of rules from animal distribution data: A ro- bust and informative method of data analysis. Mathematics and Computers in Simulation 33: 385-390. Vachon, M. 1952. Etudes sur les Scorpions. Algiers. Institut Pasteur d’Algerie. 482 pp. Wang, S. & Y. Xie., eds. 2005. China Species Red List. Vol. III. Invertebrates. Higher Education Press, Beijing. Pp. 543-544. Williams, S.C. 1968. Methods of sampling scorpion populations. Proceedings of the California Acad- emy of Sciences, Series 4 36:221-230. Williams, S.C. 1980. Scorpions of Baja California, Mexico and adjacent islands. Occasional Papers of the California Academy of Sciences 153:1- 127. Wu, H.W. 1936. A review of the scorpions and whip-scorpions of China. Sinensia 7:113-127. Zhu, M.-S., J.-X. Qi & D.-X. Song. 2004. A check- list of scorpions from China (Arachnida: Scor- piones). Acta Arachnologica Sinica 13:111-118. Manuscript received 27 April 2006, revised 31 Jan- uary 2007. 2007. The Journal of Arachnology 35:227-237 CYTOGENETICS IN THREE SPECIES OF POLYBETES SIMON 1897 FROM ARGENTINA (ARANEAE, SPARASSIDAE) I. KARYOTYPE AND CHROMOSOME BANDING PATTERN Sergio Gustavo Rodnguez-Gil: Laboratorio de Citogenetica y Evolucion, Departamento de Ecologfa, Genetica y Evolucion, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. Intendente Giiiraldes y Costanera Norte, C1428EHA. Ciudad Autonoma de Buenos Aires, Argentina. E-mail: rodrigil@bg. fcen.uba.ar Maria Susana Merani: Centro de Investigaciones en Reproduccion, Facultad de Medicina, Universidad de Buenos Aires, Paraguay 2155, C1121ABG, Ciudad Autonoma de Buenos Aires, Argentina. Cristina Luisa Scioscia: Division Aracnologia, Museo Argentine de Ciencias Naturales “Bernardino Rivadavia”, Av. Angel Gallardo 470, C1405DJR, Ciudad Autonoma de Buenos Aires, Argentina. Liliana Maria Mola: Laboratorio de Citogenetica y Evolucion, Departamento de Ecologfa, Genetica y Evolucion, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. Intendente Giiiraldes y Costanera Norte, C1428EHA. Ciudad Autonoma de Buenos Aires, Argentina. ABSTRACT. Species of Polybetes are known exclusively from South America. Currently there are 13 described species, 9 occurring in Argentina. Cytogenetic studies in spiders are scarce; the cytogenetics of only about 1% of nearly 39,500 described species are known. Within the Sparassidae, 38 species out of 1,009 (< 4%) have been cytogenetically analyzed; the most frequent chromosome number is 2 n = 43/46 (male/female), n = 20 + X,X2X3, present in almost half of the species studied. Female diploid chromosome number is only known for four species: Heteropoda venatoria (Linnaeus 1767) (2 n — 44); Pediana regina (L. Koch 1875), Isopeda sp. and Olios sp. (2 n = 46). Within the genus Polybetes, only P. pythagoricus (Holmberg 1875) had been previuosly cytogenetically analyzed. In the present work, the karyotype, het- erochromatin content and distribution, and silver stained nucleolus-organizer regions of P. pythagoricus, P. rapidus (Keyserling 1880) and P. punctulatus Mello-Leitao 1944 are described and compared. In P. pythagoricus the identification of the chromosome pairs by means of G-banding is also performed. Females of the three species show a chromosome complement of 44 telocentric chromosomes, with a similar karyotype. Males of P. pythagoricus show 42 telocentric chromosomes, the two sex chromosomes being the largest and of different size. In the three species, two pairs of telomeric NORs and small pericentrom- eric positive C -bands in all chromosomes were detected. This C-banding pattern seems to be characteristic of spiders. Comparative analysis of chromosome complements in Sparassidae indicates that 2 n = 42/44 (XIX20/X1X1X2X2) (male/female) may represent the ancestral karyotype for Polybetes. Keywords: Chromosome number, telocentric chromosomes, heterochromatin, nucleolus-organizing regions Species of Polybetes are known exclusively from South America. To date there are thirteen described species, nine of them occurring in Argentina (Platnick 2006): P. germaini Simon 1897, P. martius (Nicolet 1849), P. obnuptus Simon 1896, P. pallidus Mello-Leitao 1941, P. punctulatus Mello-Leitao 1944, P. pytha- goricus (Holmberg 1875), P. quadrifoveatus (Jarvi 1914), P. rapidus (Keyserling 1880), and P. trifoveatus (Jarvi 1914). In nature, they are found under the bark of trees (e.g., P. py- thagoricus is common under the bark of Eu- caliptus), in the branches of trees (P. rapidus ), and others are found in grasses such as Cor- taderia spp. ( P . punctulatus ). Polybetes pytha- goricus and P. rapidus are also common in 227 228 THE JOURNAL OF ARACHNOLOGY cities where they are found in parks, gardens or even in the roofs of buildings. They are nocturnal and sometimes enter houses on stormy days. Despite their usual aggressive- ness, they possess venom of low toxicity that causes little local injury and is harmless to humans (Scioscia, personal observations). Cytogenetic studies in spiders are scarce, with only approximately 1% of nearly 39,500 described species determined. Within the Spar- assidae, 36 species out of 1,009 (< 4%) had been previously cytogenetically analyzed; male diploid chromosome numbers range from 2 1 to 44; female diploid chromosome number is only known for four species: Heteropoda venatoria (Linnaeus 1767) (2 n = 44); Pediana regina (L. Koch 1875), Isopeda sp. and Olios sp. (2 n ~ 46). The most frequent sex chromosome deter- mination system is X , X2X30/X , X , X2X2X3X3 (male/female) (Hackman 1948; Suzuki 1950; Suzuki & Okada 1950; Bole-Gowda 1952; Su- zuki 1952; Mittal 1961; Diaz & Saez 1966a, b; Mittal 1966; Benavente & Wettstein 1978; Olivera 1978; Datta & Chatterjee 1983; Rowell 1985; Srivastava & Shukla 1986; Parida & Sharma 1986, 1987; Rowell 1991a, b; Hancock & Rowell 1995; Platnick 2006). Within the ge- nus Polybetes, only P. pythagoricus had been previously cytogenetically analyzed (Diaz & Saez 1966a, b; Benavente & Wettstein 1978; Olivera 1978). There are only a few studies in spiders that characterize banding patterns of chromo- somes; the distribution of C heterochromatin is known in eleven Sparassidae, eight Aranei- dae, five Lycosidae, four Tetragnathidae, two Nephilidae, two Sicariidae, one Scytodidae, and one Salticidae species, and G-banding was performed only in Lycosa thorelli (Key- serling 1877) (Lycosidae) and P. pythagoricus (Brum-Zorrilla & Cazenave 1974; Olivera 1978; Brum-Zorrilla & Postiglioni 1980; Rowell 1985; Datta & Chatterjee 1988; Row- ell 1991b; Gorlova et al. 1997; Silva et al. 2002; De Araujo et al. 2005a, b). In the present work, the karyotype, hetero- chromatin content and distribution, and silver stained nucleolus-organizer regions (NORs) of Polybetes pythagoricus , P. rapidus and P. punctulatus are described and compared. Fur- thermore, in P. pythagoricus the identification of the chromosome pairs by means of G-band- ing was performed. METHODS Adult females and males were collected in the field and were bred at the Arachnology Division of the Museo Argentino de Ciencias Naturales “Bernardino Rivadavia” (MACN). Voucher specimens of all species in this study have been deposited in the National Collec- tion of Arachnology (MACN-Ar, Museo Ar- gentino de Ciencias Naturales “Bernardino Rivadavia”, Cristina Scioscia): Polybetes py- thagoricus, five females and three males from Buenos Aires City, 34°35'15"S, 58°40'21"W, and Buenos Aires Province (Los Polvorines, 34°3Q'00S, 58°41 '00"W; Villa Madero, 34°42'00"S, 58°30'00"W; San Isidro, 34°28'15"S, 58°31'43"W; and Martin Garcia Island Natural Preserve, 34°1 1 ' 1 5"S, 58°16'52"W); P. rapidus , six females from Buenos Aires Province (Bella Vista, 35°14'00"S, 59°53'00"W; Merlo, 34°40'12"S, 58°45'10"W; Villa Madero and Martin Garcia Island Natural Preserve); P. punctulatus , one female from Martin Garcia Island Natural Pre- serve and two immature females born in the lab. For cytogenetic preparations, the specimens were cooled; and injected with 0.1 ml of 0.01% colchicine solution. After 1.25 h, sev- eral drops of hemolymph were removed from the coxal joints, and gonads together with some digestive tissues were dissected. Each sample was dispersed in 2 ml of hypotonic solution (KC1 0.56%) for 15 min, centrifuged at 800 rpm for 5 min, and fixed in 1 ml of 3 : 1 (methanol : acetic acid). The cell suspen- sion was dropped onto clean slides, air-dried and stained with Giemsa 3% for chromosome counts and karyotyping. C-band preparations were made following Sumner (1972) with some modifications: 0.2 N HC1 for 1 h; saturated solution of Ba(OH)2 for 30 s — 1 min; 2 X SSC at 60° C for 1 h. Slides were air-dried and stained with 3% Gi- emsa. G-bands preparations were made as fol- lows: PBS solution for 15 min; 0.1% trypsin for 45 s-1 min. Slides were air-dried and stained with 3% Giemsa. NOR-banding was performed following Howell & Black (1980). Eight to fifteen well-spread mitotic meta- phases were measured to determine the kar- yotype of each species. Chromosome mea- surements were made using the computer RODRIGUEZ-GIL ET AL.— KARYOTYPE AND CHROMOSOME BANDING 229 Figures 1-4. — Mitotic metaphases in spider chromosomes. 1. Polybetes rapidus female (2 n = 44); 2. P. punctulatus female (2 n = 44); 3. P. pythagoricus female (2 n = 44); 4. P. pythagoricus male (2 n = 42). Scale =10 |xm. application Micromeasure version 3.3 (Reeves & Tear 2000). The total haploid complement length (TCL) in females was calculated by adding the mean value of each chromosome pair (in arbitrary units). In males of P. pytha- goricus, the relative length of all chromo- somes was analyzed to identify the two chro- mosomes that have no homologues (sex chromosomes), and TCL was afterwards cal- culated. The idiogram of each species was drawn on the basis of the relative percentage of each chromosome pair length to the TCL. Chromosome measurements were also made using a vernier calliper to estimate TCL in microns. RESULTS Chromosome complement. — Females of P. rapidus, P. punctulatus and P. pythagori- cus show a chromosome complement of 44 telocentric chromosomes, and 42 telocentric chromosomes in males of P. pythagoricus. The sex chromosomes cannot be distinguished by their differential pycnosis in somatic meta- phases of males and females (Figs. 1-4). The total haploid complement length (TCL) is similar in the three species: 67.29 ± 4.91 |xm in P. rapidus, 66.28 ± 6.91 |Jim in P. py- thagoricus and 63.70 ± 1.52 pan in P. punc- tulatus. Females of the three species show a similar karyotype: there are three large pairs of dif- ferently-sized chromosomes that can be distin- guished; the rest of the chromosomal comple- ment gradually decreases in size, except for the last pair that is slightly smaller. The largest chromosome pair shows slight size differences in the three species (P. pythagoricus, 7.00%; P. punctulatus, 6.80%; and P. rapidus, 6.63%), while the second pair is longer in P. pythagoricus (6.45%) (P. punctulatus, 6.12%; and P. rapidus, 6.16%) (Figs. 5, 7-9). In 230 THE JOURNAL OF ARACHNOLOGY 8 i 7 8 & 5 Of 1 4 w a: 3 2 1 0 ■ P. rapidus OP.punotui^us OP. pythagoricus XI X2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 18 17 18 19 20 Chfomos ome pairs Figures 5, 6. — Comparative ideograms of spider chromosomes. 5. Females of Polybetes rapidus , P. punctulatus and P. pythagoricus', 6. Female and male of P. pythagoricus. males of P pythagoricus , the length analysis of all chromosomes of the complement makes it possible to determine that the two largest chromosomes of different sizes are the sex chromosomes X1X2 (Figs. 6, 10). Meiotic studies performed in males of P. punctulatus and P. rapidus (Rodriguez Gil 2006) demon- strated that in these species the sex chromo- somes are also the largest of the complement. C-banding and NORs silver staining. — In females of the three species, small positive C-bands in the pericentromeric region of all chromosomes were detected, except in the X2 pair of P. pythagoricus where they are more prominent (Figs. 11, 12). The number of chro- mosomes with telomeric nucleolus-organizer regions (NORs) silver stained varied from 1 to 4 in different cells of the three species (Figs. 13, 14). It was not possible to determine precisely the NOR pairs; one pair was medi- um sized and the other was among the smaller ones. RODRIGUEZ-GIL ET AL.— KARYOTYPE AND CHROMOSOME BANDING 231 7) Poiybetes rapidus II •# *• •• « « •• *• •• « *• •• *• •• •• *• •• '* « M M II I 2 3 4 5 6 7 * 9 10 It 12 13 14 IS 16 17 IK 19 2u Xl X2 8) Poiybetes punctulatus || II II II II »• •• t« •» •• •« •• •• «• •* •• •• •• •• #* || || 1 2 3 4 S 6 7 8 9 10 11 12 13 14 15 Ifi 17 IS 19 20 XI X2 9- JO) Poiybetes pythagoricus • « •• •• »• *• •' •• •» •• •• 11 11 •» •• •• •• •• •• •* M II II II II •« •« •§ Ml •• M •• •• •• •• •• •• •• •» •• •* •• •* | | I 2 3 4 5 6 7 $ 9 10 II 12 IS 14 15 10 17 18 V) 20 Xt X2 Figures 7-10. — Karyograms from spider cells depicted in Figures 1-4. 7. Poiybetes rapidus female; 8. P. punctulatus female; 9. P. pythagoricus female; 10. P. pythagoricus male. G-bandmg, — Despite applying the G banding technique to the three species, only P. pythagoricus females yielded reproducible results, making it possible to determine the chromosome pair’s Identification (Figs. 16, 17). In P. punctulatus , only a few dark bands, with little contrast, were obtained in all the chromosomes (Fig. 15); therefore, the chro- mosome pair’s identification was not possible. In P. rapidus no bands were obtained. DISCUSSION Chromosome complement and karyo- type.— Poiybetes punctulatus and P rapidus were here cytogenetically characterized for the first time. Poiybetes pythagoricus had been previously analyzed in Uruguayan pop- ulations; Benavente & Wettstein (1978) per- formed an ultrastructural study of sex chro- mosomes pairing In meiosis, and Diaz & Saez (1966a, b) reported 2 n — 42 and n — 20 + XjX2 in males. However, Olivera (1978), in a preliminary report of P pythagoricus , de- scribed contradictory data reporting a 2 n = 40/42 (male/female) with sex determination system X1X2/X1X1X2X2 at mitosis but in male meiosis described the presence of 10 chro- mosomes plus 2 Xs at one pole and 10 chro- mosomes at the other one in anaphase I. The three species of Poiybetes analyzed in this work have 2n = 44 = 40 + XlXlX2X2 in fe- males and 2n = 42 = 40 -f XjX2 in P. pytha- goricus males; they show very similar kar- yotype and total haploid complement length, with all the chromosomes telocentric. The sex chromosomes are the largest ones, the X, show slight size differences in the three spe- cies and X2 is longer in P. pythagoricus. The conservative karyotype present in the three species could be considered characteristic for the genus. Currently, cytogenetic studies in Sparassi- dae have been performed on 38 species from 17 genera (Table 1). Usually the chromosomes are telocentric and one of the sex chromo- somes is the largest of the complement. The predominant diploid numbers in this family are 2 n = 43, n = 20 + XlX2X3, 15 species; and 2n = 41, n = 19 + X,X2X3, 10 species. In other genera, besides Poiybetes , there seems to be karyotypic conservation as in Heteropoda in — 19 + X1X2X3, in 5 of 6 spe- cies studied) and Isopeda (n = 20 + X,X2X3, in the 4 species analyzed). Bole-Gowda (1952) stated that Heteropoda sexpunctata Si- mon 1885 has a derived karyotype, 2 n = 20 + X (male) with 19 metacentric (including the X chromosome) and two acrocentric auto- somes, on the basis of Robertsons law. On the other hand, in the genus Sparassus, there is variation not only in chromosome numbers (2 n = 22 to 44) but in the sex chromosome determination system as well (XjX2, XjX2X3, X!X2X3X4); although none of the entities was Identified at the species level, and it is possi- ble that the genus may be misidentihed in some of them (Parida & Sharma 1987). An- 232 THE JOURNAL OF ARACHNOLOGY # mj * ^ ' ' m % * 3- % .ufcJk * * , I > .** f w .. ,y ’ # <4L | 13 - ’# V«‘ vV I 15 4’ /V* 0 % ^ /# *1 #« •"» W&* *#•%! 1 >%# / #!*• i %• 14 16 ffif* m*m> HD s „ Y v Y ^ 5 I « l\^ <'* ^Sf- * ^ n «\s«/ Figures 11-16. — C -banding in spider chromosomes: 11. Polybetes punctulatus female; 12. P. pytha- goricus female (arrowheads point to prominent C-bands). NORs silver staining in spider chromosomes: 13. P. rapidus female; 14. P. pythagoricus female (arrows point to NOR regions). G-banding in spider chromosomes: 15. P. punctulatus female; 16. P. pythagoricus female. Scale = 10 jam. RODRIGUEZ-GIL ET AL.— KARYOTYPE AND CHROMOSOME BANDING 233 Figure 17. — G-banding karyogram and idiogram of Polybetes pythagoricus chromosomes (from cell depicted in Fig. 16). cestral populations of Delena cancerides Wal- ckenaer 1837 also show n = 20 + XjX2X3, but this species has a number of chromosomal races that differ by the presence of particular combinations of chromosomal fusions, either in homozygous or heterozygous condition (Rowell 1985, 1990, 1991a, b; Hancock & Rowell 1995). A reduced complement is also present in Micrommata virescens (Clerck 1757), but neither the chromosome number nor the sex chromosome complement is known with certainty (Hackman 1948). Heterochromatin characterization. — The three species of Polybetes analyzed here show only small pericentromeric heterochromatic bands in all the chromosomes with no differ- ential pycnosis of the sex chromosomes, al- though Olivera (1978) reported that “substan- tial” heterochromatic blocks were present in P. pythagoricus mitotic and meiotic chromo- somes. Polybetes pythagoricus X2 chromo- somes showed a larger C-band than in the oth- er two species, which may explain the differences in these chromosomes’ size. Since the pioneer characterization of Schi- zocosa malitiosa (Tullgren 1905) (Lycosidae) by Brum-Zorrilla & Cazenave (1974), few spider species have been analyzed with regard to the heterochromatin content and distribu- tion. Our results fit with previous data of most of the other spider species analyzed that have a small amount of pericentromeric heterochro- matin in the autosomes and sex chromosomes; this condition seems to be characteristic in spiders. In Loxosceles intermedia Mello-Lei- tao 1934 (Sicariidae) and Isopeda species, pericentromeric C-bands are more conspicu- ous (Brum-Zorrilla & Postiglioni 1980; Row- ell 1985, 1991b; Datta & Chatterjee 1988; Gorlova et al. 1997; Silva et al. 2002). In a few species, telomeric localization of hetero- chromatin (telomeric C-bands) has also been described in some chromosomes of the com- plement; these bands have usually appeared in a polymorphic condition (Rowell 1985; Datta & Chatterjee 1988; Rowell 1991b; De Araujo 2005a). In Nephilengys cruentata (Fabricius 1775) (Nephilidae) interstitial C-bands are present in some autosomes; the same occurs in some autosomes and the sex chromosomes of one unidentified species of Scytodes (De Araujo et al. 2005a, b). 234 THE JOURNAL OF ARACHNOLOGY Table 1. — Karyotype characteristics and collecting locality of the Sparassidae species cytogenetically analyzed (f = female). Species 2 n n (male) Locality References Bhutaniella sikkimensis 42 19 + X,X2X3X4 India Datta & Chatterjee 1983 (Gravely 1931) Delena cancerides Wal- 43 20 + X,X2X3 Australia Rowell 1985, 1991a, b; ckenaer 1837 (ancestral Hancock & Rowell 1995 karyotype) Delena sp. 43 20 + X,X2X3 McIntosh in Suzuki 1952 Heteropoda leprosa Si- 41 19 + XjX2X3 India Datta & Chatterjee 1983 mon 1884 Heteropoda phasma Si- 41 19 + X,X2X3 India Srivastava & Shukla 1986 mon 1897 Heteropoda procera (L. 41 19 + X!X2X3 Australia Rowell 1985 Koch 1867) Heteropoda sexpunctata 21 10 + X India Bole-Gowda 1952 Simon 1885 Heteropoda venatoria 41-44 f 19 + X,X2X3 India, Japan Suzuki & Okada 1950; (Linnaeus 1767) Bole-Gowda 1952; Sri- vastava & Shukla 1986 Heteropoda sp. nov. 41 19 + XlX2X3 Australia Rowell 1985 Holconia immanis (L. 43 20 + X1X2X3 Australia Rowell 1991a, b (sub Iso- Koch 1867) poda ) Is ope da vasta (L. Koch 43 20 + XiX2X3 Australia Rowell 1991b (sub Isopoda 1867) vaster (sic)) Isopeda villosa L. Koch 43 20 + XiX2X3 Australia Rowell 1991a, b (sub Iso- 1875 poda) Isopeda sp. 43-46 f 20 + XjX2X3 Australia Rowell 1985 (sub Isopoda ) Isopeda sp. nov. 43 20 + XtX2X3 Australia Rowell 1991b (sub Isopo- da) Isopedella leal Hogg 43 20 + X,X2X3 Australia Rowell 1991b (sub Isopoda 1903 tepperi Hogg) Micrommata virescens ±35 16 + XjX2X3 (?) Finland Hackman 1948 [sub Mi- (Clerck 1757) crommata viridissima (De Geer)] Neosparassus diana (L. 43 20 + XjX2X3 Australia Rowell 1991b (sub Olios ) Koch 1875) Olios lamarcki (Latreille 42 20 + x,x2 India Bole-Gowda 1952 1806) Olios sp. 1 43 20 + X!X2X3 McIntosh in Suzuki 1952 Olios sp. 2 43-46 f 20 + XtX2X3 Australia Rowell 1985 Parapalystes whiteae (Po- 43 20 + X,X2X3 India Mittal 1961, 1966 (sub Pa- cock 1902) lystes) Pediana regina (L. Koch 43-46 f 20 + XtX2X3 Australia Rowell 1985, 1991b 1875) Pediana sp. nov. 43 20 + XjX2X3 Australia Rowell 1991b Polybetes punctulatus 44 f 20 + x,x2 Argentina this work; Rodriguez Gil Mello-Leitao 1944 2006 Polybetes pythagoricus 42-44 f 20 + x,x2 Uruguay Diaz & Saez 1966a, b (sub (Holmberg 1875) P. pitagorica (sic)) Argentina this work; Rodriguez Gil 2006 40-42 f Uruguay Olivera 1978 (sub P. pitha- gorius (sic)) Polybetes rapidus (Key- 44 f 20 + x,x2 Argentina this work; Rodriguez Gil serling 1880) 2006 RODRIGUEZ-GIL et al.— karyotype and chromosome banding 235 Table 1. — Continued. Species 2 n n (male) Locality References Pseudopoda prompta (O. 41 19 + X,X2X3 India Srivastava & Shukla 1986 P.-Cambridge 1885) (sub Heteropoda ) Sinopoda forcipata 41 19 + XjX2X3 Japan Suzuki 1952 (sub Hetero- (Karsch 1881) poda) Sparassus sp. 1 44 21 + X,X2 India Parida & Sharma 1987 Sparassus sp. 2 42 20 + XjX2 India Parida & Sharma 1987 Sparassus sp. 3 41 19 + X,X2X3 India Parida & Sharma 1986, 1987 Sparassus sp. 4 41 19 + X^xJ India Parida & Sharma 1987 Sparassus sp. 5 22 10 + x,x2 India Parida & Sharma 1987 Sparassus sp. 6 42 20 + XjX2 India Datta & Chatterjee 1983 (sub Parassus sp. 1) Sparassus sp. 7 44 20 + X!X2X3X4 India Datta & Chatterjee 1983 (sub Parassus sp. 2) Sparassus sp. 8 42 20 + X,X2 India Datta & Chatterjee 1983 Spariolenus tigris Simon 41 19 + XjX2Xj India Bole-Gowda 1952 1880 Thelcticopis severa (L. 43 Possibly XlX2X3 Japan Suzuki 1950, 1952 (sub Koch 1875) Thelticopis (sic)) In spermatogonial mitosis, after C-banding, sex chromosomes have shown two different patterns. In three species of Lycosidae and in Delena cancerides the sex chromosomes were more darkly stained than the autosomes, while there was no difference in the sex chromo- somes and autosomes in Isopeda and Aranei- dae species. In the three Polybetes species pre- sented here and in four araneids, there is also no difference in the sex chromosomes and au- tosomes in female somatic and gonial cells. In one species, Schizocosa malitiosa, only one X chromosome was notable in that it exhibited complete heterochromatinization (Brum-Zor- rilla & Cazenave 1974; Brum-Zorrilla & Pos- tiglioni 1980; Rowell 1985; Datta & Chatter jee 1988; Rowell 1991b). NORs silver staining and G-banding. — The variation in the number of chromosomes per cell bearing nucleolus-organizer regions observed in Polybetes species is common in the Ag-technique. It is characteristic of silver staining that not all the NORs are usually sil- ver stained in species with multiple NORs, but only those transcriptionally active during the preceding interphase (Sumner 2003). It can be concluded that two chromosomal pairs with telomeric NORs are present in the three spe- cies here analyzed. Although the identification of the NOR pairs should be regarded as ten- tative, it seems possible that they correspond to the same pairs in the three species. Only a pair of NORs was detected at early somatic stages of P. pythagoricus Uruguayan speci- mens (Olivera 1978). G-banding allows the precise identification of homologues and facilitates karyotypic com- parisons between related species. Although good quality G-bands can be produced in rep- tiles, birds, mammals, in some fishes and am- phibians, and in a few plants, this method does not yield consistent results in inverte- brate chromosomes and only in a few species of insects have well-defined G-bands been ob- tained. The difficulty in obtaining good qual- ity G-bands in invertebrates may reflect dif- ferences in mitotic chromosome substructure, e.g., tight compaction of the chromatin com- pared to vertebrates (Lorite et al. 1996; Appels et al. 1998; Baldanza et al. 1999; Sumner 2003). In spiders, G-banding had been per- formed in three species of Lycosidae (but only Lycosa thorelli showed consistent G-banding), and in P. pythagoricus from Uruguay (where only a few pairs could be identified) (Olivera 1978; Brum-Zorrilla & Postiglioni 1980). In the present work, the identification of all chro- mosome pairs was possible in P. pythagori- cus. Taking into account that the pattern of pachytene chromomeres resembles that of G-bands on the same chromosome (Sumner 2003), it would be interesting to perform a 236 THE JOURNAL OF ARACHNOLOGY comparative analysis of the chromomere pat- tern of the three species in order to know if the scarcity and absence of G-bands in P. punctulatus and P. rapidus respectively is due to structural differences or to technical pro- cedures. The three species of Polybetes here ana- lyzed are easily distinguished by morpholog- ical characters, but they are very conservative karyotypically. This fact could be useful in future for the delimitation of genera in a sys- tematic revision of the family. ACKNOWLEDGMENTS The present study was supported by grants from the National University of Buenos Aires (UBA) (Ex 317 to Drs. L. Poggio and L. Mola) and from the National Council of Sci- entific and Technological Research (CONI- CET) (PIP 02296 and 05927 to Drs. L. Poggio and L. Mola and PIP 02202 and 05654 to Drs. A. Gonzalez and C. Scioscia). The authors wish to thank to Mr. Hernan Dinapoli for tech- nical assistance, to Lie. Pablo Rebagliati, Lie. Mariana Lopez and the student Luis Piacentini for collecting some of the specimens, and to Dr. Marfa Ines Pigozzi for critical reading of the manuscript. The authors wish to dedicate this paper to the memory of Dr. Carlos A. Naranjo. LITERATURE CITED Appels, R., R. Morris, B.S. Gill & C.E. May. 1998. Chromosome Biology. Kluwer Academic Pub- lishers, Boston, Massachusetts. 401 pp. Baldanza, F., L. Gaudio & G. Viggiani. 1999. Cy- totaxonomic studies of Encarsia Forster (Hy- menoptera: Aphelinidae). Bulletin of Entomolog- ical Research 89:209-215. Benavente, R. & R. Wettstein. 1978. Ultrastructural cytogenetics of the sex determination mecha- nisms of araneids. Revista de Microscopfa Elec- tronica 5:320-321. Bole-Gowda, B.N. 1952. Studies on the chromo- somes and the sex-determining system in four hunting spiders (Sparassidae). Proceedings of the Zoological Society, Bengal 5:51-70. Brum-Zorrilla, N. & A. Postiglioni. 1980. Kariol- ogical studies on Uruguayan spiders. I. Banding pattern in chromosomes of Lycosa species (Ar- aneae-Lycosidae). Genetica 54:149-153. Brum-Zorrilla, N. & A.M. Cazenave. 1974. Hetero- chromatin localization in the chromosomes of Lycosa malitiosa (Arachnida). Experientia 30: 94-95. Datta, S.N. & K. Chatterjee. 1983. Chromosome number and sex-determining system in fifty-two species of spiders for north-east India. Chromo- some Information Service 35:6-8. Datta, S.N. & K. Chatterjee. 1988. Chromosome and sex determination in 13 araneid spiders of North-Eastern India. Genetica 76:91-99. De Araujo, D., D.M. Celia & A.D. Brescovit. 2005a. Cytogenetic analysis of the neotropical spider Nephilengys cruentata (Araneomorphae, Tetragnathidae): Standard staining, Nors, C- Bands and base-specific fluorochromes. Brazilian Journal of Biology 65:193-202. De Araujo, D., D.M. Celia & A.D. Brescovit. 2005b. Variabilidade cromossomica em tres es- pecies de Scytodes Latreille, 1804 (Araneae, Haplogynae, Scytodidae). Primer Congreso La- tinoamericano de Aracnologfa (Minas, Uruguay) Actas: 107. Diaz, M. & F.A. Saez. 1966a. Karyotypes of South American Araneida. Memorias do Instituto Bu- tantan Commemorativo 33:153-154. Diaz, M. & F.A. Saez. 1966b. Investigaciones ci- togeneticas sobre algunas especies de araneidos uruguayos. Anales (II) Congreso Latinoameri- cano de Zoologfa (San Pablo):3-9. Gorlova, O.Y., J.P. Gorlov, E. Nevo & D.V. Logu- nov. 1997. Cytogenetic studies on seventeen spi- der species from Israel. Bulletin of the British Arachnological Society 10:249-252. Hackman, W. 1948. Chromosomenstudien an Ara- neen mit besonderer Beriicksichtigung der Gech- lechtschromosomen. Acta Zoologica Fennica 54: 1-101. Hancock, A.J. & D.M. Rowell. 1995. A chromo- somal hybrid zone in the Australian huntsman spider, Delena cance rides (Araneae: Sparassi- dae). Evidence for a hybrid zone near Canberra, Australia. Australian Journal of Zoology 43:173- 180. Howell, W.M. & D.A. Black. 1980. Controlled sil- ver staining of nucleolus organizer regions with a protective colloidal developer: a 1-step method. Experientia 36:1014-1015. Lorite, R, E. Chica & T. Palomeque. 1996. G-band- ing and chromosome condensation in the ant, Tapinoma nigerrimum. Chromosome Research 4:77-79. Mittal, O.R 1961. Chromosome number and sex mechanism in twenty-one species of the Indian spiders. Research Bulletin (N. S.) of the Panjab University 12:271-273. Mittal, O.P. 1966. Karyological studies on Indian spiders. IV. Chromosomes in relation to taxono- my in Eusprassidae, Selenopidae and Thomisi- dae. Genetica 37:225-234. Olivera, G. 1978. Estudio citogenetico en Polybetes pythagoricus (Araneomorphae-Sparassidae). XX. Semana Univisitaria Gaucha de Debates Biolo- gicos, Porto Allegre 24(Abstracts):25 1-253. Parida, B.B. & G.P. Sharma. 1986. Karyotype and RODRIGUEZ-GIL ET AL.— KARYOTYPE AND CHROMOSOME BANDING 237 spermatogonia in an Indian hunting spider, Spar- assus sp. (Sparassidae: Arachnida) with multiple sex chromosomes. Chromosome Information Service 40:28-30. Parida, B.B. & G.P. Sharma. 1987. Chromosome number, sex mechanism and genome size in twenty seven species of Indian spiders. Chro- mosome Information Service 43:11-13. Platnick, NX 2006. The World Spider Catalog, ver- sion 7.0. American Museum of Natural History, New York. Online at http://research.amnh.org/ entomology/spiders/catalog/index. html. Reeves, A. & J. Tear. 2000. Micromeasure for Win- dows. Version 3.3 free program distributed by the author from http://www.colostate.edu/Depts/ Biology/Micromeasure. Rodrfguez-Gil, S.G. 2006. Citogenetica de aranas sinantropicas de Argentina (Arachnida: Ara- neae). Ph. D. Thesis. Facultad de Ciencias Ex- actas y Naturales, Universidad de Buenos Aires. 155 pp. Rowell, D.M. 1985. Complex sex-linked fusion het- erozygosity in the Australian huntsman spider Delena cancerides (Araneae: Sparassidae), Chro- mosoma (Berlin) 93:169-176. Rowell, D.M. 1990. Fixed fusion heterozygosity in Delena cancerides Walck. (Araneae: Sparassi- dae): an alternative to speciation by monobran - chial fusion. Genetica 80:139-157. Rowell, D.M. 1991a. Chromosomal fusion and mei- otic behaviour in Delena cancerides (Araneae: Sparassidae). I. Pairing behaviour and X-chro- mosome segregation. Genome 34:561-566. Rowell, D.M. 1991b. Chromosomal fusion and meiotic behaviour in Delena cancerides (Ara- neae: Sparassidae). II. Chiasma position and its implications for speciation. Genome 34:567-573. Silva, R.W., D.D.R. Klisiowicz, D.M. Celia, O.C. Mangili & I.J. Sbalqueiro. 2002. Differential dis- tribution of constitutive heterochromatin in two species of brown spider: Loxosceles intermedia and L. laeta (Araneae, Sicariidae), from the met- ropolitan region of Curitiba, PR (Brazil). Acta Biologica Paranaense, Curitiba 31:123-136. Srivastava, S.C. & S. Shukla. 1986. Chromosome number and sex determining mechanism in forty- seven species of Indian spiders. Chromosome In- formation Service 41:23-26. Sumner, A.T. 1972. A simple technique for dem- onstrating centromeric heterochromatin. Experi- mental Cell Research 75:304-305. Sumner, A.T. 2003. Chromosomes: Organization and Function. Blackwell Publishing, Oxford, UK. 287 pp. Suzuki, S. 1950. Sex determining mechanism and karyotypes in spiders. Zoological Magazine (To- kyo) 59:31-32. Suzuki, S. 1952. Cytological Studies in spiders II. Chromosomal investigation in the twenty two species of spiders belonging to the four families, Clubionidae, Sparassidae, Thomisidae, and Ox- yopidae, which constitute Clubionoidea, with special reference to sex chromosomes. Journal of Science of the Hiroshima University, Series B, Division. 1 13:1-52. Suzuki, S. & A. Okada. 1950. A study on the chro- mosomes of a spider, Heteropoda venatoria, with special reference to X1-, X2- and X3- chromo- somes. Journal of Science of the Hiroshima Uni- versity, Series B, Division 1. 11:1-47. Manuscript received 2 August 2005, revised 19 January 2007. 2007. The Journal of Arachnology 35:238-277 A REVIEW OF SOME AUSTRALASIAN CHERNETIDAE: SUNDOCHERNES , TROGLOCHERNES AND A NEW GENUS (PSEUDOSCORPIONES) Mark S. Harvey: Department of Terrestrial Invertebrates, Western Australian Museum, Locked Bag 49, Welshpool DC, Western Australia 6986, Australia. E-mail: mark.harvey@museum.wa.gov.au Erich S. Volschenk: Department of Zoology, James Cook University, Townsville, Queensland 4811, Australia. Current address: Department of Terrestrial Invertebrates, Western Australian Museum, Locked Bag 49, Welshpool DC, Western Australia 6986, Australia ABSTRACT. A systematic review of some Australasian species previously allocated to the chernetid genus Sundochernes Beier 1932 reveals numerous discrepancies from the type species, S. modiglianii (Ellingsen 1911). Three of these species are removed to the genus Troglochernes Beier 1969, previously known from only a single troglobitic species, and a fourth is removed to a new genus. Troglochernes contains six species: the type species T. imitans Beier 1969 from caves on the Nullarbor Plain, Western Australia; three species newly transferred from Sundochernes, T. guanophilus (Beier 1967) new combi- nation, from Fig Tree Cave, New South Wales, T. dewae (Beier 1967) new combination, from bird nests in New South Wales, Queensland and Western Australia, T. novaeguineae (Beier 1965) new combination, from central Papua New Guinea; and two new species, T. cruciatus Volschenk, new species from Rope Ladder Cave, North East Queensland and T. omorgus Harvey & Volschenk, new species from a beetle in Queensland. The Lord Howe Island endemic pseudoscorpion Sundochernes grayi Beier 1975 is transferred to a new genus, Satrapanus Harvey & Volschenk, as it lacks the diagnostic features of Sundochernes. Problems with the generic allocation of species currently placed within Sundochernes are discussed and the female genitalia of Nesochernes gracilis Beier 1932 and Paraustrochernes victorianus Beier 1966 are illustrated for the first time. Troglochernes imitans is one of the most highly modified troglobitic members of the Chernetidae, displaying extremely elongate pedipalps and legs suggesting an extended period of isolation from ancestral epigean populations. The remaining cave-dwelling species, T. cruciatus and T. guanophilus, are less modified and show fewer morphological modifications which may suggest more recent colonization of the cave environments. Keywords: Taxonomy, morphology, new species, Nesochernes, caves, bird nests The Australasian chernetid fauna is mod- erately well developed, with representatives of 36 genera currently described. Few genera, however, contain more than a handful of spe- cies, and many are inadequately defined with crucial details, such as the internal female genitalia, often unknown. The genus Sundoch- ernes Beier 1932 is one of the larger genera of the region with 10 named species ranging from tropical south-eastern Asia to temperate southern Australia. A further species, doubt- fully referred to the genus, was named from Brazil (Beier 1974a). The type species Chel- ifer modiglianii Ellingsen 1911 was originally named from specimens collected in Sumatra (Ellingsen 1911), and later recorded from Ma- laysia (Beier 1967a). It was briefly rede- scribed, and illustrated for the first time, by Beier (1932). Our examination of the syntypes revealed that some species attributed to Sun- dochernes were clearly not congeneric with S. modiglianii. The study presented here dem- onstrates that four species originally attributed to Sundochernes lack the diagnostic features of that genus, and alternative taxonomic ar- rangements should be sought. Through the study of a range of specimens of Trogloch- ernes imitans Beier 1969, we have been able to ascertain that three of these species, along with two other newly described Australian species, can be attributed to Troglochernes Beier 1969 rather than Sundochernes. We 238 HARVEY & VOLSCHENK— AUSTRALASIAN CHERNETIDAE 239 have also been able to demonstrate that Sun - dochernes grayi Beier 1976 belongs to a sep- arate genus, here named Satrapanus , largely based upon the distinctive morphology of the female internal genitalia. The morphology of the spermathecae, which has been shown to be of significant value at the generic level within the Chernetidae (e.g., Vachon 1938; Muchmore 1974, 1975; Mahnert 1978), pro- vides support for the recognition of distinct genera. METHODS The specimens examined for this study are lodged in the following institutions: Austra- lian Museum, Sydney, Australia (AM); Amer- ican Museum of Natural History, New York, USA (AMNH); Australian National Insect Collection, Canberra, Australia (ANIC); Mu- seum of Natural History, London, UK (BMNH); Bishop Museum, Honolulu, Hawaii, USA (BPBM); Field Museum of Natural His- tory, Chicago, Illinois, USA (FMNH); Museo Civico di Storia Naturale “Giacomo Don a A Genova, Italy (MCSNG); Museum d'Histoire Naturelle, Geneve, Switzerland (MHNG); Museum National d'Histoire Naturelle, Paris (MNHN); Museum of Tropical Queensland, Townsville, Australia (MTQ); Naturhisto- risches Museum, Wien, Austria (NHMW); Museum Victoria, Melbourne, Australia (NMV); Queensland Museum, Brisbane, Aus- tralia (QM); South Australian Museum, Adelaide, Australia (SAM); United States National Museum, Smithsonian Institution, Washington, DC, USA (USNM); Western Australian Museum, Perth, Australia (WAM); and Urn vers itets Lund, Lund, Sweden (ZMLU). Some specimens of Satrapanus gra- yi were collected by staff from the Australian Museum's Centre for Biodiversity and Con- servation Research, abbreviated to CBCR. In addition to the specimens detailed below, we also examined the internal genitalia of fe- males of two Australasian chemetids. Neso- chernes gracilis norfolkensis Beier 1976: 2 males, 2 females, 1 tritonymph, 1 deuto- nymph, 1 protonymph from Filmy Fern Walk, 29°0LS, 167°57'E, Norfolk Island National Park, Norfolk Island, Australia, 30 November 1984, litter under Araucaria heterophylla , T.A. Weir (WAM T68620). Paraustrochernes victorianus Beier 1966: 1 female from Cum- berland Falls, Victoria, Australia, 37°34'S, 145°53'E, 27 May 1991, under bark of Eu- calyptus regnans , M.S. Harvey & M.E. Blos- felds (WAM T66536). Terminology and mensuration mostly fol- lows Chamberlin (1931), with the exception of the nomenclature of the pedipalps and legs, and with some minor modifications to the ter- minology of the trichobothria (Harvey 1992b). The specimens were studied using three techniques. Temporary slide mounts were prepared by immersion of specimens in concentrated lactic acid at room temperature for several days, and mounting them on mi- croscope slides with 10 or 12 mm coverslips supported by small sections of 0.25, 0.35 or 0.50 mm diameter nylon fishing line or with small slivers of broken coverslips. After study, the specimens were returned to 75% ethanol with the dissected portions placed in 12 x 3 mm glass genitalia microvials (BioQuip Prod- ucts, Inc.). Permanent slide mounts were pre- pared by removing the pedipalps, the chelic- era, left leg I, and left leg IV from specimens with the use of eye-scissors or small needles, and clearing overnight with 10% potassium hydroxide at room temperature. The speci- mens were then washed in several rinses of water and 5% acetic acid (to neutralise the potassium hydroxide), and dehydrated through a graded ethanol series. They were then transferred to Euparal Essence overnight at room temperature, prior to mounting in Eu- paral on microscope slides using 10 or 12 mm coverslips supported by small sections of 0.25, 0.35 or 0.50 mm diameter nylon fishing line. All specimens were studied using an Olympus BH -2 compound microscope and il- lustrated with the aid of a drawing tube. Mea- surements were taken at the highest possible magnification using an ocular graticule. The maps were produced with the computer program ArcVIew 3.2 after the relevant local- ity data were stored in an Access database. Coordinates were obtained from various sources, including the Geoscience Australia Place Names Search website (http://www.ga. gov.au/maps/names) and the GeoNet Names Server (http://earth-info.nga.mil/gns/html/) produced by the National Geospatial-Intelli- gence Agency. Recently collected specimens were usually provided with GPS coordinates taken at the collecting site. Coordinates ob- tained from indirect sources (such as gazet- teers) are listed below within parentheses. 240 THE JOURNAL OF ARACHNOLOGY FAMILY CHERNETIDAE MENGE 1855 SUBFAMILY CHERNETINAE MENGE 1855 Genus Sundochernes Beier 1932 Sundochernes Beier 1932:162; Beier 1933:531; Beier 1976:225; Harvey 1991:635. Type species.- — Chelifer modiglianii Eh lingsee 1911, by original designation. Type material.- — - Sundochernes modigli- anii (Ellingsen 1911): Syntypes: 1 male, 7 fe- males, 1 nymph, Sirarnbas (as Si-Rambe), Su- matra, Indonesia [G°49'N, 99°32'E], no date, E. Modigliani (MCSNG), examined. Sundochernes australiensis Beier 1954b: Holotype female, Denmark, near mouth of Denmark River, Western Australia, Australia [34°58'N, 117°22'E], karri [Eucalyptus diver- sicolor\ forest, 26 January 1952, T. Gislen (ZMLU), not examined. Sundochernes brasiliensis Beier 1974a: Ho- lotype male. Nova Teutonia, Santa Catarina, Brazil [27°03'S, 52°24'W], 300-500 m, F. Plaumaee (MHNG), examined. Sundochernes dubius Beier 1954b: Holo- type female, Augusta, Western Australia, Aus- tralia [34°19'S, 1 15°09'E], 12 December 1952, T. Gislen (ZMLU), not examined. Sundochernes gressitti Beier 1957: Holo- type female, Ngaremeskang, Babelthuap, Pa- lau [07°3LN, 134°33'E], 30 m, 21 December 1952, J. L. Gressitt (USNM 2262), examined. Paratype: 1 female, Ngercheu Islands [as Gar- akayo Island], “Pelew” Islands [= Palau] [07°05'N, 134°16'E], 8 August 1945, H. S. Dybas (FMNH), examined. Sundochernes malayanus Beier 1963: Ho- lotype male, Raetau Panjang, 5 mi N of Klang, Selangor, Malayasia [03°25'N, 101°28;E], from nest of Olive Bulbul, Mi - croscelis olivacea , 28 June 1961 (BPBM), ex- amined. Paratype: 1 tritonymph, same data as holotype, from nest of Yellow-vented Bulbul, Pycnonotus goiavier , 7 June 1961 (BPBM), examined. Sundochernes queenslandicus Beier 1975: Holotype male, Marburg, Queensland, Austra- lia [27°34;S, 152°35'E], litter and soil, 16 May 1966, K. E. Lee (SAM N197761), examined. Diagnosis. — Sundochernes differs from all other chemetid genera by the following com- bination of characters: flagellum with 3 blades; spermathecae with 2 thickened tubes with rounded terminal bulbs; legs without tac- tile setae; 1 pair of eyespots present; vestitural setae generally small, dentate and clavate. Remarks.— =Beier (1932, 1933) recorded three blades in the cheliceral flagellum in Chelifer modiglianii , and accordingly placed his new genus Sundochernes in the tribe Cher- netini which was based primarily upon the number of cheliceral blades. Descriptions of species subsequently attributed to Sundoch- ernes have either reiterated the possession of three cheliceral blades or have omitted flagel- la) blade counts. MSH examined the syntypes of C modi- glianii during 1986 while identifying Indo- nesian specimens of pseudoscorpioes (Harvey 1988), and made observations on the chelic- erae and female genitalia. Drawings made at the time have been subsequently mislaid, and the syntypes have not been available to us again, precluding the provision of illustrations of genetically important features. Neverthe- less, notes made at the time of study indicate that the cheliceral flagellum consists of 3 blades, as stated by Beier (1932, 1933), and that the female geeitalic region consists of spermathecae with 2 thickened tubes with rounded terminal bulbs. Examination of the type or other material of most Sundochernes species (listed above) indicates that although some species possess three blades (e.g., S. modiglianii , S. austral- iensis■, S. dubius , S. brasiliensis , S. queenslan- dicus, S. gressitti , and S. malayanus ), others possess four blades. As noted above, this basic distinction in flagellar blade number has long been used to separate chemetid taxa, com- mencing with Beier (1932, 1933) who used flagellar number to diagnose the tribes Cher- netini and Hesperochemetini within the Cher- netinae. Although these tribes are no longer recognized, the number of flagellar setae is still given considerable significance in cher- eetid taxonomy (e.g., Muchmore 1974). Aside from flagellar blade number, the species with four blades have been found by us to possess fundamentally different spermathecal mor- phology from that found in S . modiglianii which is afforded high value in chemetid tax- onomy. Our study suggests that these four species can be referred to two different genera with S. guanophilus , S, dewae , and S. novae- guineae placed in Troglochemes , and S. grayi in a new genes, here named Satrapanus . This makes Sundochernes a slightly more coherent HARVEY & VOLSCHENK— AUSTRALASIAN CHERNETIDAE 241 genus but the situation is further complicated by species such as S. queenslandicus which Is clearly distinct from both S. modiglianii , Troglochernes and Satrapanus , suggesting that a further new genus will be required to accommodate it. The systematic position of the remaining species of Sundochernes can only be determined when detailed examina- tion of each species Is completed, with partic- ular reference to spermathecal morphology as highlighted for other chemetids by Vachon (1938), Muchmore (1974, 1975) and Mahnert (1978). Genus Troglochernes Beier 1969 Troglochernes Beier 1969:185; Harvey 1991:638. Type species.— -Troglochernes imitans Beier 1969, by original designation. Diagnosis.— Troglochernes differs from all other chemetid genera by the following com- bination of characters: flagellum with 4 blades, or possibly 3 blades in one species; spermathecae with 2 thickened and slightly curved tubes fused basally; legs without tac- tile setae; carapace unicolored and with two transverse furrows; eyes or eyespots absent; vestitural setae generally small, dentate and clavate. Description, — Adults: Vestitural setae mostly short, slightly curved, and dentate. Pedipalps: with most surfaces finely to heavily granulate. Fixed finger with 8 tricho- bothria, movable chelal finger with 4 tricho- bothria; esb closer to eb than to est; isb ap- proximately midway between it and ist; it situated in distal third of fixed finger; sb closer to b than to st. Marginal teeth of chela all closely spaced; both chelal fingers with exter- nal and internal rows of accessory teeth. Ven- om apparatus present in movable finger with nodus ramosus terminating midway between t and st, or adjacent to st. Chelicera: with 6 or 7 setae on hand; Is and is acuminate, sbs, bs\ bs" and bsm (when pres- ent) dentate, es either acuminate (most spe- cies) or dentate (T. omorgus ); movable finger with 1 acuminate seta (gs); with 2 dorsal and 1 ventral lyrifissures; lamina exterior present; movable finger with 1 dorsal tooth; galea long and slender with 5-6 rami; flagellum com- posed of 4 blades, or possibly 3 blades in one species; two anterior blades dentate along dis- tal anterior half, two shorter blades smooth, or in case of species with 3 blades, anterior blade dentate and others smooth. Cephalothorax: carapace with eyes or eye- spots absent; unicolored; with two transverse furrows; posterior margin straight or nearly so. Median maxillary lyrifissure present and sub-medially situated; posterior maxillary lyr- ifissure present. Abdomen: tergites and stemites generally divided. Pleural membrane wrinkled striate for entire length, without setae, but females of two species with setae. Each stigmatic sclerite with 1 or more setae. Spiracles simple, with spiraeular helix. Genitalia: male genitalia of typical cheme- tid form; female spermathecae with 2 thick- ened and curved tubes fused basally. Legs: legs I and II with an oblique junction between femur and patella; legs III and IV without tactile setae on tibiae or tarsi; meta- tarsus and tarsus fused into single segment (tarsus); tarsi with single raised slit sensillum; subterminal tarsal seta curved and acuminate; tarsal claws simple; arolium slightly shorter than claws. Nymphs : Much like adults, but trichoboth- rial patterns as follows: tritonymph with 7 on fixed finger and 3 on movable finger; deuto- nymph with 6 on fixed finger and 2 on mov- able finger; and protonymph with 3 on fixed finger and 1 on movable finger. Chelicera of protoeymph lacking seta gs. Tarsi of proto- nymph without single raised slit sensillum. Remarks. — Beier (1969) proposed the new genus Troglochernes for T. imitans , a highly troglomorphic species from caves on the Null- arbor Plain, Western Australia. He treated the genus as a member of the Hesperochernetini as it possessed four fiagellal blades, and dis- tinguished it from most other genera by the elongate pedipalps and legs, and lack of a tac- tile seta on the posterior tarsi. Our examina- tion of the female spermathecae of T. imitans (Fig. 14) reveals a form unlike that docu- mented for any other Australasian chemetid genus and we consider the genus distinct from all other previously described chemetid gen- era on the basis of this character. Other Aus- tralasian species with extremely similar sper- mathecae have been detected, and despite the lack of extreme troglomorphisms, we include them here in Troglochernes and extend the di- agnosis of the genus to include less highly troglomorphic species than T. imitans . Three of these species which are here transferred to Troglochernes were previously placed in the 242 THE JOURNAL OF ARACHNOLOGY Figures 1, 2. — Spermathecae, ventral: 1. Nesochernes gracilis norfolkensis Beier, female from Norfolk Island, Australia (WAM T68620); 2. Paraustrochernes victorianus Beier, female from Cumberland Falls, Victoria, Australia (WAM T66536). Scale lines = 0.1 mm. genus Sundochernes by Beier (1965, 1967b), but we cannot agree with this placement as they are clearly not congeneric with the type species, S. modiglianii (Ellingsen). Some oth- er species currently included within Sundoch- ernes are not congeneric with S. modiglianii , and we discuss these problems in more detail under that genus. In addition, we have found that all species here attributed to Troglocher - nes lack eye-spots, a feature that further dis- tinguishes it from Sundochernes. Species of Troglochernes differ from the other Australasian chemetid genera with four blades in the flagellum as follows: from Aus- trochernes Beier 1932 by the lack of tactile setae on tarsus IV [present in Austrochernes, see With (1905) and Beier (1932)]; from Par- austrochernes Beier 1966 by the unicolored carapace [bicolored metazone in Paraustroch- ernes; see Beier (1966)], and the presence of only a single pair of spermathecae in the fe- male genitalia [two pairs of spermathecae in P. victorianus (Fig. 2)]; and from Maracher- nes Harvey 1992a by the general shape of the chelal hand (which is not much wider than the base of the fingers in Marachernes ) and the lack of an intemobasal mound bearing acces- sory teeth on the male movable chelal finger (Harvey 1992a, 1994). It differs from Satra- panus by the lack of eye-spots, which are pre- sent in Satrapanus , by the morphology of the female genitalia in which the spermathecae are usually lightly curved, and by the color of the carapace which is uniformly unicolored In Troglochernes , but is distinctly bicolored in Satrapanus , with the metazone paler than the remaining carapace. Apart from these Australian genera, only 12 other genera, mostly from the northern hemi- sphere, have been reported as lacking tactile setae on the posterior tarsi and possessing four blades in the flagellum [Muchmore (1974) re- ported that species of Chernes mostly have a four-bladed flagellum, but Dr. V. Mahnert (in litt.) informs me that three-bladed specimens are equally abundant]. Troglochernes differs from these genera as follows: The spermathecal morphology of two thick- ened and curved tubes that are fused basally segregates Troglochernes from Chelodamus R.V. Chamberlin 1925 (from Central Ameri- ca), Chernes Menge 1855 (from Europe, North Africa, Asia and North America), Hes- perochernes Chamberlin 1924 (from North America and Japan), Chelanops Gervais 1849 (from South America), Semeiochernes Beier 1932 (from South America), and Illinichernes Hoff 1949 (from North America), which all possess long, slender spermathecae (Benedict & Malcolm 1982; Chamberlin 1952; Mahnert 1978, 1987; Muchmore 1974, 1975, 1984, 1999), and from Gigantochernes Beier 1932 (from South America) and Cocinachernes Hentschel & Muchmore 1989 (from Mexico), which have four ( Cocinachernes ) or appar- ently five ( Gigantochernes ) short spermathe- cal tubes (Hentschel & Muchmore 1989; Vi- tali-di Castri 1972). The spermathecal morphology of Atheroch - ernes Beier 1954a (from Venezuela), Eume - cochernes Beier 1932 (from Hawaii) and Ne- sochernes Beier 1932 (from New Zealand and Norfolk Island) are unknown but each can be readily separated from Troglochernes. Ather - ochernes differs by the presence of 5 setae on the cheliceral hand (6 or more setae in Trog- lochernes), and by the presence of accessory teeth only on the movable chelal finger (ac- cessory teeth on both chelal fingers in Trog- lochernes) (Beier 1954a). Eumecochernes has HARVEY & VOLSCHENK— AUSTRALASIAN CHERNETIDAE 243 trichobothrium isb situated b as al ly to est (Beier 1932), whereas it is situated opposite or slightly distal to est in Troglochernes . Specimens of Nesochernes gracilis norfolk- ensis have spermathecae with three pairs of ducts each distally with small pores (Fig. 1); this arrangement is quite different to that of Troglochernes. While most species of Ceriochernes Beier 1937, including the type species, C. detritus Beier 1937 from the Philippines, C. foliaceo- setosus Beier 1974a from Brazil and C. ves- titus Beier 1974b from Nepal and Pakistan, possess three flagellar blades (Beier 1937, 1974b, 1974a; Dashdamirov 2005), the Bra- zilian species C. amazonicus Mahnert 1985 possesses four blades (Mahnert 1985). The number of flagellar blades has not been re- ported for the remaining species currently in- cluded in the genus — C besucheti Beier 1973 from Sri Lanka, C. nepalensis Beier 1974b and C. martensi Beier 1974b from Nepal, and C brasiliensis Beier 1974a from Brazil (Beier 1973, 1974b, 1974a). Lack of knowledge of the morphology of the spermathecae for most species of Ceriochernes is severely hampering our understanding of this widespread and un- doubtedly paraphyletic genus (Dashdamirov 2005). The sole member of Ceriochernes that has four flagellar blades and lacks tactile setae on the posterior tarsi, C. amazonicus , has highly unusual spermathecae in which there are numerous spermathecal bulbs, each cir- cular on long thin stalks, leading from a cen- tral atrium (Mahnert 1985). Ecology. — Although habitat preferences are unknown for T. novaeguineae, the remain- ing five species of Troglochernes occur in caves or are intimately associated with other animals. Troglochernes guanophilus, T. cm - ciatus and the highly troglomorphic T. imitans are known from caves where they inhabit gua- no deposits or reside under nearby rocks or leaf litter lying on the floor of the cave. Trog- lochernes dewae has been collected solely from bird nests, including that of the Galah ( Cacatua roseicapilla ), Sulphur-Crested Cockatoo (C. galerita ), Carnaby’s Cockatoo ( Calyptorhynchus latirostris ) and Rufous Treecreeper {Climacteris rufa ). The sole spec- imen of T. omorgus found attached phoreti- cally to the beetle Omorgus costatus (Trogi- dae), individuals of which are known to occur in caves where they live and breed in bat gua- no (Scholtz 1986). KEY TO SPECIES OF TROGLOCHERNES 1. Large species with long, slender pedipalps, e.g., chela (with pedicel) 2.036-2.408 (d), 1.992-2.528 (?) mm long and 4.98-5.39 (d), 4.58-5.49 (?) times longer than wide .... Troglochernes imitans Small species with short, robust pedipalps, e.g., chela (with pedicel) 1.00-1.34 (d), 1.05- 1.61 (?) mm long and 2.50-3.00 (d), 2.45-2.99 (?) times longer than wide ......... 2 2. Posterior margin of carapace with 25 setae; tergites generally with more than 30 setae . . . ......................................................... Troglochernes omorgus Posterior margin of carapace with less than 20 setae; tergites generally with less than 30 setae 3 3. Cheliceral seta es dentate; posterior margin of carapace with 16-20 setae Troglochernes dewae Cheliceral seta es acuminate; posterior margin of carapace with 8—16 setae ........... 4 4. Posterior margin of carapace with 8 setae .................. Troglochernes novaeguineae Posterior margin of carapace with 10 or more setae . 5 5. Posterior margin of carapace with 10-12 setae Troglochernes guanophilus Posterior margin of carapace with 14-16 setae Troglochernes cruciatus Troglochernes imitans Beier 1969 Figs. 3, 7-14, 71 Troglochernes imitans Beier 1969:185-187, fig. 1; Richards 1971:19, 24, 25, 27, 28, 30, 43; Beier 1975:203; Harvey 1981:247; Harvey 1985:136; Harvey 1991:638; Moulds 2004:12. Type material examined.— -AUSTRALIA : Western Australia: Holotype male, Dingo Cave [6N-160], Nullarbor Plain, Western Aus- tralia, Australia [31°51'S, 126°44'E], near en- trance, 28 October 1968, J. Lowry (SAM N 1980 192). Allotype female, same data as ho- 244 THE JOURNAL OF ARACHNOLOGY Figures 3-6. — 3. Troglochernes imitans Beier, male from Scudd Cave, Western Australia (WAM 98/ 1508); 4. Troglochernes dewae (Beier), female from Gingin Shire, Western Australia (WAM T48341); 5. Troglochernes cruciatus, sp. nov., female paratype from Rope Ladder Cave, Queensland (WAM T68621); 6. Satrapanus grayi (Beier), female from Lord Howe Island (AM). lotype (SAM N 1980 193). Paratypes: 2 fe- males, same data as holotype (NHMW). Other material examined. — AUSTRA- LIA: Western Australia: 1 $, Murra-El-Elev- yn Cave [6N-47] [32°02'S, 126°02'E], on dry guano, 21 April 1973, K. Williamson (WAM 74/361); 1 9, Murra-ELElevyn Cave [6N-47] [32°02'S, 126°02'E], under mineral crusts, 21 April 1973, P.J. Bridge (WAM 74/362); 1 5 yr) were found to occupy the same bur- rows where spiders had been returned after tissue sampling. In our opinion, the presence of these large spiders probably indicates the long-term survival of wild caught specimens following autospasy (for DNA), or at the least, re-colonization of empty burrows. Survey of PCR primers for Brachypel- ma.— We generated —2200 bp of DNA se- quence from 22 B. vagans , for three genetic regions (except XTS-2 in Las Cuevas 10 and 11). For IrRNA-NDl, the most consistent PCR amplification was using primers LR-N- 13398 and NI-J- 12261 to give a 940 bp frag- ment. However, other primer combinations were quite effective in amplifying a shorter fragment of the same region, like LR-N- 12945 with NI-J- 12261 which readily yielded —650 bp. Using the same thermal profile (on re- quest), LR-N- 13398 and NI-J- 12581 amplified a similar size fragment (—650 bp) with far less consistency and lower product yield than oth- er primer combinations. Primers Hbl6S with HbNDl failed to give consistent amplifica- tions in B. vagans using the thermal profile for Salticidae (Masta 2000b) or with profile variations. For the COl, the primer Cl-J-1751 worked best with CI-N-2776 to give a 960 bp fragment, but Cl-J-1751 also worked well with CI-N-2568 yielding about 850 bp. Here, consistency was increased by adding 5 low stringency cycles of 95° (30 s), 50° (30 s), 72° (60 s) prior to 35 cycles as for CI-J-1751 with CI-N-2776. The same modified profile can also be used to amplify COl from other spe- cies of Brachypelma and more distantly relat- ed Theraphosidae (Longhorn 2001, unpub- lished data). Primer CI-J-1718 with CI-N-2776 failed to give strong amplifications for B. vagans under a variety of thermal con- ditions. Characterization of the IrRNA-NDl. — The 940 bp fragment of IrRNA-NDl from B. vagans gave the closest nucleotide match (by BlastN) to published mitochondrial sequences from jumping spiders (Araneae, Salticidae). Nucleotide identity was higher between B. va- gans and several Salticidae (Top match 6e=44 to AY477266; Hedin & Maddison 2001, 2003) than another tarantula (6e-41) Haplo- pelma huwenum (Wang et al. 1993; Qiu et al. 2005) [In Genbank as Ornithoctonus huwena NC-005925, but recently transferred to Hap- lopelma (Araneae, Theraphosidae)]. However, measures of similarity alone often give mis- leading views of taxon affinities, while infor- mative characters can be more useful. That said, similarity is often useful to identify func- LONGHORN ET AL.— MOLECULAR MARKERS FOR TARANTULAS 283 U D arm AG G A A C G U 1 M auugc a u C/U U c u U -A U-A U u A -US U-A Acceptor arm AUU Au U c c u/c A-U U-A A-U G-C .A-U. A A U>— A (U AG) AU D arm C/U- c u u u A- u u A A U U u u u AG A A C G ! I I XjUGC u T arm A A UUA Mill U U A A U U U Anticodon arm A - U A U-A A-U G-C A — U . A A U^— A (UAg; ^A U u A Uau/c AUU Fig. 2. — -Putative secondary structures for tRNA leucine in the mygalomorph spider B. vagans. Left: Structure fitted to the truncated model from the araneomorph spider Habronattus (Masta & Boore 2004). Right: Fitted to yield a conventional “cloverleaf” structure with a functional TifiC arm. tionally important domains. Four regions of high identity were detected between IrRNA- ND1 of B. vagans and other spiders. The larg- est high identity segment (about 203 nucleo- tides) corresponded to the peptidyl transferase centre of the IrRNA, identified using the sec- ondary structure map of the salticid Habron- attus oregonensis (Peckham & Peckham 1888) (Masta 2000b). The next largest identity region (30/33 bp) was the DHU and anticodon arms of tRNALEU(CUN). With the exception of two sites, the B. vagans tRNA sequences were identical across all 22 individuals, and had a higher AT composition (75.6%) than IrRNA (—68.5%). The tRNALEU(CUN) in H. oregonen- sis folds to an unusual truncated structure, lacking the TijiC arm (Masta 2000b; Masta & Boore 2004). According to this truncated model, the TiJjC arm is substituted by a TV- replacement loop. It is possible to fit a clo- verleaf structure to this tRNA in H. oregone- nesis, but this model allows less complementary bases in the acceptor arm than the truncated model and more problematic overlap with ND1 (Masta 2000b; Masta & Boore 2004). The truncated model fits reason- ably well to tRNALEU(CUN) from B. vagans , with strong intra-molecular pairing (high de- gree of complementary) in the DHU-arm (Fig. 2, left). In the anticodon arm, the UAG anti- codon was found as expected, supported by five paired stem bases as H. oregonenesis. The cloverleaf structure fits well to the B. vagans data (Fig. 2, right), revealing the possibility of a conventional Tij;C arm. For our data, the clo- verleaf model also requires greater mismatch in the acceptor stem than the truncated model plus more overlap with ND1 (Longhorn 2001). The start codon of Brachypelma ND1 appears to be an atypical ATT (AUU), as in other spiders (Masta 2000b). To accept the cloverleaf structure as the preferred model for B. vagans requires the first seven codons (18 b.p.) of ND1 to also function as part of the tRNALEU(CUN). If allowed, our data suggest that the canonical cloverleaf model provides a bet- ter fit to Brachypelma tRNALEU(CUN) than the truncated model, which fitted best for H. or- egonenesis (Masta 2000b). Characterization of ITS-2. — The 358 bp ITS-2 from B. vagans gave the closest nucle- otide (BlastN) match to Aphonopelma hentzi 284 THE JOURNAL OF ARACHNOLOGY (Girard 1852) (AY2 10803, Mallatt et al. 2004). This is another tarantula closely allied to the Brachypelma , both in the subfamily Theraphosinae (Smith 1994). A single region of high identity (e-129) was found between these two genera (92%; 338/367). Across dif- ferent B. vagans, much of the fragment at the 5.8S rRNA end is invariant (up to 120 bp here; 162 bp in A. hentzi). The small segment of 28S rRNA included was also invariant across B. vagans , with only two nucleotide differences from A. hentzi . In the actual ITS- 2 spacer, there were size differences between A. hentzi (203 bp) and B. vagans (193 bp) and nineteen inter-specific nucleotide differences, most at the 3'. The available ITS-2 from other spiders matched B. vagans with much less identity, the next most significant (2e-50) was Orsonwelles spp. Hormiga 2002 (Araneae, Linyphiidae) (Hormiga et al. 2003). Sequenc- es from other araneomorphs matched less well again, like Theridiidae (S.W. A’Hara, unpub- lished), Nesticidae (Hedin 1997b), Linyphi- idae (Hormiga et al. 2003) and Salticidae (Ar- nedo & Gillespie 2006), many of which had a similar size (150-250 bp) to our ITS-2 (193 bp), considerably smaller than other arthro- pods. Characterization of COl. — The 960 bp fragment of B. vagans COl gave the closest nucleotide (BlastN) match to Promyrmekia- phila sp. Schenkel 1950 (Araneae, Mygalo- morphae, Cyrtaucheniidae: AY621508; Bond 2004) with 84% identity (2e-151, 623/740 sites). Similarly, COl from other rnygalo- morphs also matched B. vagans with a high identity. For example, COl from Sphodros abbotti Walckenaer 1835 (Atypidae: AF303528; Hedin 2001) matched with 83% nucleotide identity (559/666 sites), while Apo- mastus schlingeri Bond & Opel! 2002 (Cyr- taucheniidae: e.g., DQ388588; Bond et al. 2006), or Antrodiaetus unicolor (Hentz 1842) (Antrodiaetidae: e.g., AY896899; Hendrixson & Bond 2005) matched with higher identity (up to 85%) but over a shorter length (up to 534/622 sites). The most significant match (6e-124) to another member of the family Theraphosidae was with H. huwenum (Qiu et al. 2005; as O. huwena). Surprisingly, the lev- el of COl identity between these tarantulas was almost identical to levels seen between B. vagans and other families of mygalomorph spiders (83%; 518/616 sites, versus above). This was surprising as Brachypelma and Hap- lopelma only currently warrant different sub- families (both Theraphosidae, Theraphosinae and Ornithoctoninae, respectively). Recently, a few COl sequences from other Brachypelma species have been published (Petersen et al. 2006). Each of these is quite short, only averaging around 300 bp, and se- verely limits the number of sites for compar- ison. The highest identity was between our B. vagans and B. albopilosum up to 97%; (DQ224243, 141/144 sites), followed by B. angustum (to 95%, DQ224245, 137/144). This result supports a close phylogenetic position of B. vagans with these Mesoamerican spe- cies, compared to other Brachypelma from the Pacific coast of Mexico, like B. smithi (as in Longhorn 2001; see also Petersen et al. 2006). Translated (BlastX) searches showed that B. vagans sequences also displayed greatest ami- no acid similarity with H. huwenum (85%; 268/312 sites). However, the next most sig- nificant matches were from araneomorph spi- ders of the Salticidae, probably due to con- vergence. Overall, the B. vagans COl showed similar nucleotide composition to other spi- ders except Heptathela (Table 3), which sug- gested that shared biases in composition were not a factor that led protein searches to iden- tify high similarity with between salticids and B. vagans or the unexpectedly high genetic divergence between Brachypelma and Hap- lopelma COl sequences. Population structure in Brachypelma va- gans.— It was possible to join all 22 B. vagans sequences by parsimonious connections into a haplotype network for each fragment (Fig. 3), each identical with gaps were coded as miss- ing or 5th state. The two mitochondrial frag- ments (IrRNA-ND 1 and COl) gave similar structure among individuals and separated the two populations (Pooks Hill and Las Cuevas). The IrRNA-ND 1 (Fig. 3, top) revealed slightly more divergence between populations than COl (Fig. 3, middle) though both regions are size equivalent (—950 bp). Both mitochondri- al fragments also revealed more unique hap- lotypes in the Las Cuevas population than at Pooks Hill, and both separated the outgroup B. angustum. These results contrast with the smaller ITS-2 (Fig. 3, bottom), which could not distinguish the two B. vagans populations, nor showed sufficient nucleotide differences to separate the outgroup. Across all fragments, LONGHORN ET AL.— MOLECULAR MARKERS FOR TARANTULAS 285 Table 1. — Specimens with GenBank accession numbers. Samples of B. vagans are listed by their col- lection location, either from P = Pooks Hill or C = Las Cuevas. Sample (all B. vagans unless indicated) Carapace Length (cm) Fragment sequenced ITS-2 COl IrRNA-NDl PI 1.95 AI585053 AJ584615 AJ585387 P2 2.20 AJ585054 AJ584616 AJ585388 P3 2.00 AJ585055 AJ584617 AJ585389 P4 2.50 AJ585056 AJ584618 AJ585390 P5 3.00 AJ585057 AJ584619 AJ585391 P6 2.70 AJ585058 AJ584620 AJ585392 P7 2.20 AJ585059 A J5 84621 AJ585393 P8 2.85 AJ585060 A J5 84622 AJ585394 C9 2.20 AJ585061 A J5 84623 AJ585395 CIO 3.00 No Data A J5 84624 AJ585396 Cll 3.00 No Data A J5 84625 AJ585397 C12 1.80 AJ585062 A J5 84626 AJ585398 C13 1.50 AJ585063 A J5 84627 AJ585399 C14 1.90 AJ585064 AJ584628 AJ5 85400 C15 1.75 AJ585065 A J5 84629 AJ5 85401 C16 3.10 AJ585066 AJ584630 A J5 85402 C17 2.50 AJ585067 A J5 84631 AJ585403 C18 1.40 AJ585068 AJ584632 AJ585404 C19 1.70 AJ585069 AJ584633 AJ585405 C20 1.40 AJ585070 A J5 84634 AJ585406 C21 3.00 AJ585071 AJ584635 A J5 85407 C22 2.40 AJ585072 AJ584636 A J5 85408 B. angustum 1.75 AJ585073 AJ584637 AJ5 85409 there were thirty-six polymorphic nucleotides, twenty-four of which were parsimony infor- mative (IrRNA-NDl = 11 sites [1.14%] > COl = 10 sites [1.06%] > ITS-2 = 3 sites [0.84%]). Details of nucleotide polymorphism within and between populations of B. vagans are given in Tables 4 and 5. In both popula- tions, the number of segregating nucleotides per sequence (S) was similar (except with Table 2. — Mean nucleotide composition of ge- netic fragments of B. vagans. Length Fragment (bp) Ade- nine (A) Cyto- sine (C) Gua- nine (G) A + T (%) COl 963 0.206 0.121 0.229 65.0 IrRNA rRNA 498 0.349 0.156 0.159 68.5 tRNA(LEU) 53 0.377 0.131 0.113 75.6 ND1 396 0.331 0.174 0.110 71.5 5.8S rRNA 119 0.197 0.267 0.309 42.4 ITS-2 193 0.187 0.294 0.296 41.0 28S rRNA 45 0.330 0.252 0.209 53.9 IrRNA-NDl). The Pooks Hill population con- sistently had a greater number of average se- quence differences (k) among individuals than Las Cuevas, despite fewer haplotypes overall. Within each population, our results suggest that differences in directional selection are not detectable (neither population deviates signif- icantly from neutrality) and that demographic conditions are relatively stable (non-signifi- cant Tajima’s D statistic). For both the COl and IrRNA-NDl frag- ments, the largest proportion of sequence di- versity was attributable to differences between populations. Estimates of gene flow and pop- ulation structure were difficult to determine with ITS-2. This was surprising, as this region is often large and variable in arthropods, and hence widely used for phylogeographic stud- ies. Estimates of between population gene flow between using Fsx values were much lower from nuclear ITS-2 than from either mi- tochondrial fragment, reflecting its short length and relative invariance. Overall, results 286 THE JOURNAL OF ARACHNOLOGY Table 3. — Nucleotide composition of COl across exemplar Araneae. * = Complete. [Infraorder] Family Genus Accession % A % C % G Length A + T [Aran.] Agelenidae Tegenaria AY138836 22.7 15.3 19.8 450 64.9 [Aran.] Araneidae Argiope AY731171 27.5 12.1 18.8 1536* 69.1 [Aran.] Desidae Badumna AF2 18280 25.4 11.2 20.7 552 68.1 [Aran.] Eresidae Stegodyphus AY6118Q5 26.6 11.1 19.7 1000 69.2 [Aran.] Linyphiidae Frontinella DQ029220 23.8 13.0 19.5 954 67.5 [Aran.] Lycosidae Rahidosa DQ029232 23.9 11.6 20.4 942 68.0 [Aran.] Oecobiidae JJroctea DQ973166 22.8 13.2 21.2 964 65.7 [Aran.] Salticidae Habronattus NC005942 26.8 11.4 18.7 1542* 69.9 [Aran.] Thomisidae Xysticus AY297423 24.8 12.7 17.9 1047 69.4 [Aran.] Tetragnathidae Nephila NC008063 27.1 11.7 18.7 1536 69.7 [Aran.] Dysderidae Dysdera AF244321 21.2 14.0 22.9 471 63.1 [Aran.] Hypochilidae Hypochilus AF303527 23.3 14.0 22.4 1536 63.6 [My gal.] Antrodiaetidae Antrodiaetus AY297423 23.7 11.4 21.0 1008 67.6 [My gal.] Atypidae Sphodros AF303528 21.3 11.9 21.2 1047 66.9 [My gal.] Cyrtaucheniidae Apomastus DQ389886 25.7 11.6 19.0 810 69.4 [Mygal] Hexathelidae Atrax A AL 11676 25.4 12.9 19.6 658 67.5 [Mygal.] Theraphosidae Omithoctonus NC-005925 24.7 12.2 21.4 1536* 66.4 [Mygal.] Theraphosidae Brachypelma AJ584636 20.7 12.2 22.9 963 64.9 [Suborder Mesothelae] Heptath- elidae Heptathela NC005924 28.2 17.4 15.1 1533* 67.5 Average (all Araneae) 24.5 12.7 20.0 1057 67.3 Average (only Araneomorphae) 24.6 12.5 20.1 1090 67.4 Average (only Mygalomorphae) 24.2 12.8 20.0 1079 67.1 indicated that the Pooks Hill population is more structured than at Las Cuevas, even though fewer individuals were sampled at Pooks Hill. An equally plausible explanation is that the Pooks Hill population is older (un- der neutrality). Estimates of the effective numbers of migrating individuals per genera- tion (Nm) were surprisingly low from both mitochondrial fragments, almost to a level where it is difficult to distinguish between low and non-existent levels of gene flow among populations, suggesting a high degree of mi- tochondrial sequence isolation. DISCUSSION Non-lethal DNA sampling by autospa- sy.— -The evaluation of non-lethal DNA sam- pling techniques is important in any genetic studies where the goal is conservation. Here, we induced limb autospasy from fossorial ta- rantulas for genetic material. We refer to this induced response as autospasy rather than au- totomy, which has been incorrectly used else- where and strictly applies to a reflex action alone (after Pieron 1907; Wood 1926; Roth & Roth 1984). In general, the ability to cast limbs is found in a wide variety of arthropods. In most spiders, limb separation involves rup- ture at the coxa-trochanter boundary, achieved by snapping the coxa upwards while the femur is kept static (Bauer 1972). This is followed by muscle contraction around the internal margin of the coxa to close the wound and minimize hemolymph loss (after Wood 1926; Bauer 1972). In some cases, autospasy can oc- cur at the patella- tibial joints in certain long- legged Linyphiidae and perhaps Filistatidae (Roth & Roth 1984). A third type of autospasy has been described between the femur and pa- tella, at the patellar cleavage plane, but only been in two genera of Agelenidae (Roth 1981). Autospasy can be easily induced in other arachnids, especially Opiliones, but not in Scorpiones (Wood 1926) or some primitive spiders (Roth 1981). Therefore, while limb re- moval is probably a useful method to obtain non-lethal DNA samples from most spiders, it is not universally applicable for all arachnids. To reduce trauma to B. vagans during tissue sampling, we considered C02 anesthetization, as used to attach radio-transmitters (Janowski- Bell & Horner 1999) or insert transponders (Reichlieg & Tabaka 2001). However, because autospasy is partly voluntary, we suggest that LONGHORN ET AL.— MOLECULAR MARKERS FOR TARANTULAS 287 Figure 3. — Haplotype networks for the two B. vagans populations (top = from IrRNA-NDl; middle = COl; bottom = ITS-2). In all cases gaps = missing data. Asterisked boxes notify ancestral sequences. Black lines represent single mutational steps, and small squares lost/un-sampled haplotypes. P = Pooks Hill population of B. vagans , C = Las Cuevas population of B. vagans , and ANG = B. angustum. the method would be damaging to anesthe- tized specimens and result in excessive fluid loss (as shown by Bonnet 1930). Without anesthetization, one captive tarantula did show more hemolymph loss than other similar sized spiders after autospasy, either in captive or field populations. Even after the application of artificial coagulants (corn starch and nail Table 4. — DNA polymorphism within populations. Gene Population Haplo- types (S) Nucleotide differences (k) -gene e (S) -gene Tajimas’ D Significance COI Las Cuevas 4 6 1.626 1.887 -0.495 > 0.1 = N/s Pooks Hill 3 6 2.428 2.314 0.231 > 0.1 = N/s IrRNA-NDl Las Cuevas 4 3 0.780 0.943 -0.529 > 0.1 = N/s Pooks Hill 3 7 2.393 2.699 -0.541 > 0.1 = N/s ITS-2 Las Cuevas 2 1 0.303 0.331 -0.195 > 0.1 = N/s Pooks Hill 2 1 0.250 0.385 -1.055 > 0.1 = N/s 288 THE JOURNAL OF ARACHNOLOGY Table 5. — DNA polymorphism between populations. Gene Fixed nucleotide differences Avg. nucleotide differences (K) between populations Fst Nm COI 3 7.679 0.736 0.18 IrRNA-NDl 6 9.786 0.838 0.10 ITS -2 0 0.519 0.520 4.56 hardener) the wound re-opened when this spi- der moved, though fluid-loss finally ceased when it was artificially restrained. Closer ex- amination showed abdominal darkening char- acteristic of the pre-molt phase, and the spider successfully molted two weeks later. The molt occurred without any regeneration of the lost leg, which was not observed until a subse- quent molt (8 mo later). As tarantulas are un- willing to accept prey items during their molt- ing period, linking forced autospasy with prey-lure collection techniques may lower the probability of capturing pre-molt individuals in field studies, and prevent this potential problem. However, if needed, tarantulas can willingly cast limbs while in pre-molt, though this may be the most vulnerable time for limb removal. Alternative strategies for genetic studies of Brachypelma. — DNA can be easily ex- tracted from exuviae (Peterson et al. 2007), which provides an alternative non-lethal tissue source than limb removal for arthropods. While DNA yields are low, it is possible to amplify high-copy fragments (like IrRNA). However, current attempts to use tarantula ex- uviae have either failed to sequence large fragments or been confounded by fungal con- taminants (Longhorn 2001; Peterson et al. 2007). With more taxon specific primers, ex- uviae will doubtless provide a useful, albeit challenging tissue source for DNA samples, especially for captive spiders. That said, ex- uviae are little used in genetic studies of nat- ural populations, mainly due to DNA degra- dation and contamination. Furthermore, most adult tarantulas molt yearly (or less), which is unlikely to coincide with a short research pe- riod and these spiders often keep old exuviae deep inside their burrows, making it difficult for the investigator to access without undue damage. The most valuable use for exuviae may be for genetic investigations of tarantulas seized during wildlife enforcement (Peterson et al. 2007), although investigators must again wait for a molt, probably an intolerable delay. Regardless of approach, it is critical to pro- mote non-lethal DNA sources in conserva- tion-focused studies, and build a resource of characterized genetic data. For the genus Bra - chypelma , a DNA reference collection may be extremely useful to wildlife-trade enforcement and conservation efforts. At best, such a ge- netic resource would be based on samples with clear provenance and a supporting mor- phological determination. Our genetic samples of B. vagans fulfil these criteria for this spe- cies and provide a useful reference for com- parison with similar spiders of uncertain his- tory or collection location. Population structure in Brachypelma . — Due to the small number of genetic fragments sampled, it is unknown whether the discrep- ancies in Fst between the three genetic regions can be attributed to differences between nu- clear and mitochondrial gene flow in general. However, the two-mitochondrial fragments did not suggest differential selection pres- sures. Reasonable levels of polymorphism di- versity in these two fragments allowed gene flow estimates (Nm) to be calculated with more confidence than from the less variable nuclear ITS- 2. However, the validity of gene flow estimates based on indirect measures such as Fst is unclear. Slatkin (1989) sug- gested that when samples are large (around 10 or more individuals, as here), values for Nm are robust estimators of gene flow. More re- cently, it has been proposed that measures of genetic variability based on FST may be valu- able, but their transformation into quantitative estimates of gene flow may be unnecessary at best, and misleading at worst (Whitlock & McCauley 1999). Theoretical concerns aside, estimates of gene flow are important to un- derstand species integrity. In general, gene flow acts to counter the effects of isolation. Broadly, gene flow influences the persistence LONGHORN ET AL.— MOLECULAR MARKERS FOR TARANTULAS 289 of local populations and facilitates the spread of adaptive traits among complex landscapes (Hanski & Gilpin 1997). Direct measures of migration may be the most preferable ap- proach to estimate gene flow, but such mea- sures have only been applied cursorily in ta- rantulas with implanted tags (Reichling & Tabaka 2001) or radio-transmitters (Janowski- Bell & Homer 1999). As well as the time- constraints and technological difficulties, di- rect measures of dispersal may not reflect the actual movement of genetic material. For gene flow to have occurred, the migrant must also reproduce effectively in the new location. This is probably not the case for the majority of male tarantulas, which can fall prey to pred- ators, harsh environmental factors, or even un- receptive females. Indirect estimation of gene flow in taran- tulas.— The estimation of gene flow in fos- sorial tarantulas is complicated by the influ- ence of sex-biased dispersal. In all species of Brachypelma , mature females display strong site fidelity while mature males show long- range dispersal. Due to these sex-specific as- pects, estimates of population subdivision from maternally inherited mitochondrial DNA can be over-estimates compared to bi-paren- tally inherited nuclear DNA (Gomez-Zurita & Vogler 2003). In fossorial tarantulas, direct measures of gene flow from male migration may provide the best estimates of long-range gene flow, but these overlook finer scale ef- fects of juvenile dispersal and colony forma- tion. The indirect estimation of gene flow from maternally inherited mitochondrial frag- ments may present the clearest picture of fine sub-structure and gene flow. We concede that inferences from mitochondrial markers are probably not suitable to track genetic admix- ture from long-range male dispersal, but sug- gest that finer scale genetic differences within populations are equally interesting. For our samples of B. vagans, results are consistent with the Pooks Hill population hav- ing more connectivity with neighboring pop- ulations than those at Las Cuevas. Field ob- servations agreed as the Pooks Hill population was from a semi-open grassland habitat indis- tinguishable from nearby regions where it was likely that B. vagans also occurred in high densities. In contrast, the Las Cuevas popu- lation was restricted to an isolated grass clear- ing enclosed by moist broadleaf forest where intensive searches failed to reveal further B. vagans burrows (but these have since been re- ported to exist at low densities). Overall, the geographic distance between the two B. vagans populations (about 50 km) reflects the entire range of other, more threat- ened species in the genus Brachypelma (Locht et al. 1999; West 2005). As a result, inferences from our localized geographic sampling of B. vagans may provide a useful baseline to com- pare against the overall genetic divergences from other species of Brachypelma. However, ecologies of B. vagans slightly differ from other species, and comparisons of divergences across taxa should be treated with caution since different Brachypelma species probably differ in their abilities to colonize new areas. Alternative sources of nuclear genes. — The nuclear ITS-2 was not suitable to resolve genetic sub- structure in B. vagans due to its short size and relative invariance. At the time of this study, the only known spider ITS-2 were from the araneomorph families Theridi- idae and Nesticidae at —150-250 bp (Hedin 1997b), reflecting the paucity of genetic re- gions amenable to study at that time. Aside from ITS-2 and neighboring rRNA, we ini- tially had few other choices of regions to se- lect, given a lack of nuclear data for any spi- ders, especially Theraphosidae. There has since been a gradual accumulation of PCR tar- geted nuclear sequences of spiders in public domain, plus several thousand expressed se- quence tags (ESTs) from the tarantula Acan- thoscurria gomesiana Mello-Leitao 1923 (Lorenzini et al. 2006). Together, these new data can provide a foundation for genetic stud- ies of theraphosid spiders. Several studies have suggested that nuclear gene introns are often the most suitable region for population level genetic studies. However, with EST data, the location of introns is often unknown, as most are derived from mature mRNA. Before additional nuclear markers can be derived from spider ESTs, gene-specific primers need to be designed and the location, size and var- iability of introns identified at the genomic re- gion of interest. Overall, it would be desirable to explore estimates of population sub-struc- ture in fossorial tarantulas using several nu- clear fragments. However, this study confirms that selected segments of the mitochondrial genome can alone provide valuable genetic 290 THE JOURNAL OF ARACHNOLOGY data for phylogeographic studies of these spi- ders. Of the three genetic regions, the mitochon- drial IrRNA-NDl was best suited for the char- acterization of population subdivision and ge- netic polymorphism in B. vagans. Similar conclusions arose from both IrRNA-NDl and CO 1 , which confirmed the suitability of both these mitochondrial markers for inferring pop- ulation sub-structure. The knowledge of ap- propriate molecular makers is vital to facilitate future genetic studies with tarantulas. Both mitochondrial fragments were able to distin- guish individuals of B. vagans from different collection sites, and both regions revealed a greater degree of population sub-structure than anticipated. Such information is critical for conservation, which is often focused around saving as much of the discrete genetic diversity as possible. Taken to the extreme, the spiders from different collection locations could be considered as different phylogenetic species, or at least, discrete geographical lin- eages, equally worth conservation efforts re- flecting their distinct identities. For the genus Brachypelma as a whole, habitat fragmentation continues to threaten the cohesion of natural populations, particularly for the species endemic to the Pacific coast of Mexico. Thankfully, the collection of spiders in the genus Brachypelma is now restricted by both national and international laws, including CITES. The next critical step to the conser- vation of these spiders will be to enhance un- derstanding of the genetic affinities and pop- ulation sub-structure of each threatened species. At best, future genetic studies of Bra- chypelma will ensure the future survival of viable populations, define species limits, and be used to conserve the maximum genetic di- versity of each discrete lineage in this high- profile genus. ACKNOWLEDGMENTS We thank the Forestry Department of Be- lize, and staff at the LCRS, particularly Chris Minty and Chapal Bol. We thank Vicki and Ray Snaddon for their hospitality at Pook’s Hill Lodge. This work was funded by a LCRS research grant, the American Tarantula Soci- ety, and British Airways “Conservation Scheme.” We are indebted to Steve Reichling for field advice, and to Andrew Smith for his faith that genetics may someday be useful in theraphosid systematics. We thank Susan Masta and Jesus Gomez-Zurita for valuable comments, Pierre Paquin for French transla- tion, and Paula Cushing and Gail Stratton for assistance with the manuscript. This work formed part of a Masters thesis (2001) of the first author from Imperial College, London, UK. LITERATURE CITED Anonymous. 1998. Checklist of CITES Species: A Reference to the Appendices to the Convention on International Trade in Endangered Species of Wild Fauna and Flora 1998. WCMC, Cambridge, UK. Arnedo, M.A. & R.G. Gillespie. 2006. Species di- versification patterns in the Polynesian jumping spider genus Havaika Proszynski, 2001 (Ara- neae, Salticidae). Molecular Phylogenetics and Evolution 41:472-495. Ayoub, N.A. & S.E. Riechert. 2004. Molecular ev- idence for Pleistocene glacial cycles driving di- versification of a North American desert spider, Agelenopsis aperta. Molecular Ecology 13: 3453-3465. Baerg, W.J. 1958. The Tarantula. University of Kansas Press, Lawrence, Kansas. 88 pp. Bauer, K.H. 1972. Funktionsmechanisms der auto- tomie bei Spinnen (Araneae) und morpholo- gischen Voraussetzungen. Zeitschrift fur Mor- phologic der Teire. 72:173-202. Bond, J.E., M.C. Hedin, M.G. Ramirez & B.D. Opell. 2001. Deep molecular divergence in the absence of morphological and ecological change in the Californian costal dune endemic trapdoor spider Aptostichus simus. Molecular Ecology 10: 899-910. Bond, J.E. 2004. Systematics of the Californian euctenizine spider genus Apomastus (Araneae: Mygalomorphae: Cyrtaucheniidae): the relation- ship between molecular and morphological tax- onomy. Invertebrate Systematics 18:361-376. Bond, J.E., D.A. Beamer, T. Lamb & M. Hedin. 2006. Combining genetic and geospatial analyses to infer population extinction in mygalomorph spiders endemic to the Los Angeles region. An- imal Conservation 9:145-157. Bonnet, P. 1930. La mue, 1’autotomie et la regen- eration chez les Araignees, avec une etude des Dolomedes d’ Europe. Bulletin de la Societe d’Histoire Naturelle de Toulouse 59:237-700. Brautigam, S.E. & M.H. Persons. 2003. The effect of limb loss on the courtship and mating behav- ior of the wolf spider Pardosa milvina (Araneae: Lycosidae). Journal of Insect Behavior 16:571- 587. Breene, R.G., III. 1998. The ATS Arthropod Med- ical Manual: Diagnosis and Treatment. The LONGHORN ET AL.— MOLECULAR MARKERS FOR TARANTULAS 291 American Tarantula Society, New Mexico. 38 pp. Clement, M., D. Posada & K.A. Crandall. 2000. TCS: a computer program to estimate gene ge- nealogies. Molecular Ecology 9:1657-1659. Cleva, S. 1998. Investigation probes a web of smuggling. Fish & Wildlife News Sept/Oct: 12- 13. Gomez-Zurita, J. & A.P. Vogler. 2003. Incongruent nuclear and mitochondrial phylogeographic pat- terns in the Timarcha goettingensis species com- plex (Coleoptera, Chrysomelidae). Journal of Evolutionary Biology 16:833-843. Hanski, LA. & M.E. Gilpin. 1997. Metapopulation Biology: Ecology, Genetics and Evolution. Ac- ademic Press, New York. 512 pp. Hedin, M.C. 1997a. Molecular phylogenetics at the population/species interface In cave spiders of the Southern Appalachians (Araneae: Nesticidae: Nesticus). Molecular Biology & Evolution 14: 309-324. Hedin, M.C. 1997b. Speciational history in a di- verse clade of habitat-specialised spiders (Aran- ae: Nesticidae: Nesticus ): inferences from geo- graphic based sampling. Evolution 51:1929- 1945. Hedin, M.C. 2001. Molecular insights into species phylogeny, biogeography, and morphological stasis in the ancient spider genus Hypochilus (Araneae; Hypochilidae). Molecular Phylogenet- ics & Evolution 18:238-251. Hedin, M.C. & W.P. Maddison. 2001. A combined molecular approach to phylogeny of the jumping spider subfamily Dendryphantinae (Araneae: Salticidae). Molecular Phylogenetics & Evolu- tion 18:386-403. Hedin, M.C. & W.P. Maddison. 2003. Phylogeny of Habronattus jumping spiders (Araneae: Saltici- dae), with consideration of genital and courtship evolution. Systematic Entomology 28:1-22. Hendrixson, B.E. & J.E. Bond. 2005. Testing spe- cies boundaries in the Antrodiaetus unicolor complex (Araneae: Mygalomorphae: Antrodiae- tidae): ‘Paraphyly’ and cryptic diversity. Molec- ular Phylogenetics & Evolution 36:405-416. Hormiga, G., M. Arnedo & R.G. Gillespie. 2003. Speciation on a conveyor belt: sequential colo- nization of the Hawaiian Islands by Orsonwelles spiders (Araneae, Linyphiidae). Systematic Bi- ology 52:70-88. Hudson, R.R., M. Slatkin & W.P. Maddison. 1992. Estimation of levels of gene flow from DNA se- quence data. Genetics 132:583-589. Janowski-Bell, M.E. & N.V. Horner. 1999. Move- ment of the male brown tarantula, Aphonopelma hentzi (Araneae, Theraphosidae) using radio te- lemetry. Journal of Arachnology 27:503-512. Johnson, S.A. & E.M. Jakob. 1999. Leg autotomy in a spider has minimal costs in competitive abil- ity and development. Animal Behaviour 57:957- 965. Joyce, D.A. & A.S. Pullin. 2003. Conservation im- plications of the distribution of genetic diversity at different scales: a case study using the marsh fritillary butterfly ( Euphydryas aurinia). Biolog- ical Conservation 114:453-461. Locht, A., M. Yanez & I. Vazquez. 1999. Distri- bution and natural history of Mexican species of Brachypelma and Brachypelmides (Theraphosi- dae, Theraphosinae) with morphological evi- dence for their synonymy. Journal of Arachnol- ogy 27:196-200. Longhorn, S.J. 2001. Genetic evaluation of CITES II protected spiders of the genus Brachypelma (Theraphosidae); including an outline of non-le- thal tissue sampling. M.Sc. Thesis. Imperial Col- lege, University of London, London, UK. Longhorn, S.J. 2002. Non-lethal DNA sampling from CITES II protected ‘Tarantula’ spiders of Belize. Las Cuevas Newsletter 9:8-9. Lorenzini, D.M., P.I. Silva, M.B. Soares, P. Arruda, J. Setubal & S. Daffre. 2006. Discovery of im- mune-related genes expressed In hemocytes of the tarantula spider Acanthoscurria gomesiana. Developmental & Comparative Immunology 30: 545-556. Lushai, G., W. Fjellsted, O. Marcovitch, K. Aa- gaard, T. Sherratt, J.A. Allen & N. Maclean. 2000. Application of molecular techniques to non-lethal tissue samples of endangered butterfly populations ( Parnassius apollo L.) in Norway for conservation management. Biological Con- servation 94:43-50. Mallatt, J.M., J.R. Garey & J.W. Schultz. 2004. Ec- dysozoan phylogeny and Baysian inference: first use of nearly complete 28S and 18S rRNA gene sequences to classify arthropods and their kin. Molecular Phylogenetics & Evolution 31:178- 191. Masta, S.E. 2000a. Phylogeography of the jumping spider Habronattus pugillis (Araneae: Saltici- dae): recent vicariance of sky island populations? Evolution 54:1699-1711. Masta, S.E. 2000b. Mitochondrial sequence evolu- tion in spiders: intraspecific variation in tRNAs lacking the Ti|/C arm. Molecular Biology & Evo- lution 17:1091-1100. Masta, S.E. & J.L. Boore. 2004. The complete mi- tochondrial genome sequence of the spider Ha- bronattus oregonensis reveals rearranged and extremely truncated tRNAs. Molecular Phylo- genetics & Evolution 21:893-902. M’Rabet, S.M., Y. Hemaut, R. Rojo & S. Calme. 2005. A not so natural history of the tarantula Brachypelma vagans: interaction with human ac- tivity. Journal of Natural History 39:2515-2523. Murphy, N.P., V.W. Framenau, S.C. Donnellan, M.S. Harvey, Y.-C. Park & A.D. Austin. 2006. 292 THE JOURNAL OF ARACHNOLOGY Phylogenetic reconstruction of the wolf spiders (Araneae: Lycosidae) using sequences from the 12S rRNA, 28S rRNA, and NADH1 genes: im- plications for classification, biogeography and the evolution of web building behavior. Molec- ular Phylogenetics and Evolution 38:583-602. Nicholas, M. 2002. Hunting the Mexican red ramp tarantula in Belize. Journal of the British Taran- tula Society 18:9-20. Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. 513 pp. Paquin, P. & M. Hedin. 2004. The power and perils of 'molecular taxonomy’: a case study of eyeless and endangered Cicurina (Araneae: Dictynidae) from Texas caves. Molecular Ecology 13:3239- 3255. Petersen, S.D., T. Mason, S. Akber, R. West, B. White & P. Mason. 2007. Species identification of tarantulas using exuviae for international wild- life law enforcement. Conservation Genetics 8: 497-502. Pieron, H. 1907. Autotomie et “Autospasie” Comte Rendu hebdomadaire Seances. Memoires Societe Biologic, Paris 63:425-427. Qiu, Y., D. Song, K. Zhou & H. Sun. 2005. The mitochondrial sequences of Heptathela hang ~ zhouensis and Ornithoctonus huwena reveal unique gene arrangements and atypical tRNAs. Journal of Molecular Evolution 60:57-71. Reichling, S.B. 1999. Nearest neighbour relation- ships among theraphosid spiders in Belize. The Southwestern Naturalist 44:518-521. Reichling, S.B. 2000. Group dispersal in juvenile Brachypelma vagans (Araneae, Theraphosidae). Journal of Arachnology 28:248-250. Reichling, S.B. 2003. Tarantulas of Belize. Kreiger Publishing, Malabar, Florida. 127 pp. Reichling, S.B. & C. Tabaka. 2001. A technique for individually identifying tarantulas using passive integrated transponders. Journal of Arachnology 29:117-118. Roth, Y.D. 1981. A new genus of spider (Angelen- idae) from California exhibiting a third type of leg autospasy. In Contributions to arachnid sys- tematics in honor of Willis J. Gertsch, on the occasion of his seventy-fifth birthday. Bulletin of the American Museum of Natural History 170: 101-105. Roth, V.D. & B.M. Roth. 1984. A review of appen- dotomy in spiders and other arachnids. Bulletin of the British Arachnological Society 6:137-146. Rozas, J. & R. Rozas. 1999. DnaSP version 3: an integrated program for molecular population ge- netics and molecular evolution analysis. Bioie- formatics 15:174-175. Shillington, C. & B. McEwen. 2006. Activity of juvenile tarantulas in and around the maternal burrow. Journal of Arachnology 34:26-265. Shillington, C. & P Verrell. 1997. Sexual strategies of a North American Tarantula’ (Araneae: Ther- aphosidae). Ethology 103:588-598. Simon, C., F. Frati, A. Beckenbaeh, B. Crespi, H. Liu & R Flook. 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene se- quences and a compilation of conserved poly- merase chain reaction primers. Annals of the En- tomological Society of America 87:651-701. Slatkin, M. 1989. Detecting small amounts of gene flow from phytogenies of alleles. Genetics 121: 609-612. Slatkin, M. & W.R Maddison. 1989. A cladistic measure of gene flow inferred from the phytog- enies of alleles. Genetics 123:603-613. Smith, A.M. 1994. Pp. 174-176. In Theraphosid Spiders of the New World, Volume 2. Tarantulas of the USA and Mexico. Fitzgerald Publishing, London, UK. Striffler, B. & A. Graminske. 2003. Brachypelma — die bunten Vogelspinnen Mexikos. DRACO 4: 52-61. Tajima, F. 1983. Evolutionary relationship of DNA sequences in finite populations. Genetics 105: 437-460. Tajima, F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymor- phism. Genetics 123:585-595. Templeton, A.R., E. Routman & C. Phillips. 1995. Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial DNA haplotypes in the Tiger Salamander, Ambrystoma tigrinum. Ge- netics 140:767-782. Watterson, G.A. 1975. On the number of segregat- ing sites in genetical models without recombi- nation. Theoretical Population Biology 7:256- 276. West, R.C. 2005. The Brachypelma of Mexico. Journal of the British Tarantula Society 20:1 OS- 119. Whitlock, M.C. & D.E. McCauley. 1999. Indirect measures of gene flow and migration: Fst # 1/ (4Nm+ 1). Heredity 82:117-125. Wood, HD. 1926. Autotomy in Arachnid a Journal of Morphology 42:143-192. Woodman, J.D., J.E. Ash & D.M. Rowell. 2006. Population structure in a saproxylic funnelweb spider (Hexathelidae: Hadronyche ) along a for- ested rainfall gradient. Journal of Zoology 268: 325-333. Wright, S. 1951. The genetical structure of popu- lations. Annals of Eugenics 15:323-354. Yanez, M. & G. Floater 2000. Spatial distribution and habitat preference of the endangered taran- tula, Brachypelma klaasi (Araneae: Theraphosi- dae) in Mexico. Biodiversity and Conservation 9:795-810. Manuscript received 10 June 2005, revised 25 April 2007 . 2007. The Journal of Arachnology 35:293-306 CHROMOSOMES OF CROSSOPRIZA LYONI (BLACKWALL 1867), INTRAINDIVIDUAL NUMERICAL CHROMOSOME VARIATION IN PHYSOCYCLUS GLOBOSUS (TACZANOWSKI 1874), AND THE DISTRIBUTION PATTERN OF NORs (ARANEOMORPHAE, HAPLOGYNAE, PHOLCIDAE) Rosangela Martins Oliveira,1 Aline Carolina de Jesus,1 Antonio Domingues Brescovit2 and Doralice Maria Celia1: 'Universidade Estadual Paulista-UNESP, Institute de Biociencias, Departamento de Biologia, Caixa Postal 199, CEP 13506-900, Rio Claro, State of Sao Paulo, Brazil. E-mail: characid@yahoo.com.br 2Instituto Butantan, Laboratorio de Artropodes Pe^onhentos, Av. Vital Brasil, n. 1500; CEP 05503-900, Sao Paulo, State of Sao Paulo, Brazil ABSTRACT. Pholcidae (Haplogynae) encompasses 967 described species, of which only 14 have been cytogenetic analyzed. Several chromosomal features have already been described including presence of meta- and sub-metacentric chromosomes and sex determination chromosome system (SDCS) of the X, XjX2Y, and XtX2 types, which contrast with the telo- and acrocentric chromosomes and SDCS of the XjX2 type typical of entelegyne spiders. To obtain further cytogenetic information for the family, we examined two pholcid species, Crossopriza lyoni (Blackwall 1867) and Physocyclus globosus (Taczanowski 1874) using both conventional staining and silver staining techniques. Crossopriza lyoni exhibited 2n = 23 = 22 + X in males and 2n = 24 = 22 + XX in females, while P. globosus showed 2n=15=14 + X and 4n = 30 = 28 + 2X, both in male adults, 2n = 16 = 14 + XX in female adults and embryos, and 2n = 15 = 14 + X in male embryos. Both species revealed predominately metacentric and submetacentric chromosomes and a SDCS of the X/XX type. The cytogenetic data obtained in this work and those already recorded for C. lyoni indicate interpopulational and intraspecific numerical chromosome variation, sug- gesting the presence of chromosomal races or cytotypes in this species. The intraindividual numerical chromosome variation observed in male adult specimens of P. globosus may be explained by the presence of cytoplasmatic bridges between germ cells. The use of the silver staining technique to reveal the nucleolar organizer region (NOR) showed that chromosome pairs 4 and 6 and the X chromosome in C. lyoni are telomeric NOR-bearers, and that the chromosome pair 2 in P. globosus possesses a proximal NOR in the long arm. Keywords: Chiasma, chromosome rearrangements, meiosis, syncytial, tetraploidy The family Pholcidae currently has 81 gen- era and 967 described species (Platnick 2007) and is included in the Haplogynae (Codding- ton & Levi 1991; Ramirez 2000), which is considered less morphologically derived than Entelegynae. Chromosome analyses within Pholcidae were initiated by Painter (1914) in Spermophora senoculata (Duges 1836) (as Spermophora meridionalis Hentz 1837). Due to the inefficient cytogenetic techniques of the time, Painter noted only that the chromosome complement consisted of small metacentric chromosomes and that the sex determination chromosome system (SDCS) was of the X,X2 type. Since then, considering the number of named species, little cytogenetic information on Pholcidae has been added to the literature. Cytogenetic data currently exist for 14 pholcid species (Table 1), which represent less than 2% of the total known. The Indian Crosso- priza lyoni (Blackwall 1867) has provided the greatest amount of chromosomal data that en- compassed several populations. Bole-Gowda (1958) described the presence of 2n = 27 = 1311 + X with metacentric autosomes and sex chromosomes; Sharma et al. (1959) recorded 2n = 24 = 1 III + XjX2 with metacentric au- tosomes and acrocentric sex chromosomes; Srivastava & Shukla (1986) observed 2n = 25 = 1211 + X but did not provide any infor- mation regarding chromosome morphology, and finally Parida & Sharma (1987) and Shar- 293 294 THE JOURNAL OF ARACHNOLOGY 3 ^ o co ON 43 & 2 X bJO 0 O o d toJ) 0 I'S i! cd s y, § >> O ,2 ft, o co &*a It, O' ®3 fl cd CM „ u, O <3 -p £> g 6 § 3 U 3 £ ■O £ jl 0 § *-*•! '3 s « * S J* § 1 11 f S £ m <8 o o '5 a c a aj w o 0 5 1 1 I II £ S 'O O 0 -JH N L •d e 0 1 43 O 0 . 5 £ o « >n fi U 3 f2 „ 0 Q R O c G S S G *R t g O G ■S& § § '“> t * U X & m o o CM 4D in o o cm 3 b3 «s B S3 & 3 d 0 0 0 0 “d •a •c •a O O O o cd 3 o o o .° .Q ’■0 ’a ’S. "J3 '.& o 5-i cd o H S p p d «l ^0 O o 3 Ph 0 £ 0 Z 3 3 Ph X DO X 3 X m < S X oo nr X ^ X « 3 X X X X X X X e + + + + + + + 3 CM in cn s CM 5 3 HH T—( ?— i '— 1 ?-H II II ii II ii II II CM i> a in cn CO CO en CM CM CM CM CM CM X X X + + K K CM t-h I> 00 '-l H II II II II in i> *n A T-* ’-H CM CM R O G .K? c S’ t- r £5 vo o <*> 00 PM g ^ ^ ' . ' . o o o o R R o g A An *1 3 8 s S •« g R R O G PM G >~»j *25 U a 0 X g 25 25 G CM HS On g oo 2° g 22 g 22 I S'" CM O o CM 00 ON 70 *-h On 3 W 00 ^~S cd ON 0 Sic /^N ° b sm A in 0 N 0 ON ^ dA JB* 3 P N 3 W) SC 'O 0 d c ^ g 0 P o H 43 © m X X DO O a, S o 0 X 0 a 0 K) •Q PQ &a R 3 ■|2 NO a o ° 'O S2 e G S «§ W ft. O S ^ «3 c 2 « “ S ed £ ^ X O Z, + + + + + u O + + T s A s s S H S s VO CM CM VO A o L CM GO CM CM CM ^_i f-H < CM ’— i R G »R £3 § *5 ■53 -R X X X X X + + in vo in CM CM r-t « & -S O 1 § 0 0 X X ON CM On tfl t; 0 o a G . CL 25 & G VO «3 ON Q £■ S Sg <->x 1 1 1-8 G bo £ £ X + ^ S X Sx 11 x ^x 3 Q * * * a a a ft. o . - ss ft, ft. G 0 ft. G O A ^ ft, ^ RS u 00 o ^ s G O R G 6*5 g - o VO OLIVEIRA ET AL.— CYTOGENETICS OF TWO PHOLCIDS 295 ma & Parida (1987) reported the presence of 2n = 23 — 1 1 II + X, also with no description of chromosome morphology. In Physocyclus Simon 1893, all cytogenetic described species (. Physocyclus californicus Chamberlin & Gertsch 1929, Physocyclus enaulus Crosby 1926, and Physocyclus sp.) occur in the Ne- arctic region (Cokendolpher 1989) and show a great karyotypic uniformity in relation to diploid number (2n — 15), metacentric chro- mosome morphology, and X/XX sex deter- mination chromosome system type. Recently, cytogenetic analyses were carried out in three pholcid species. Pholcus phalan- gioides (Fuesslin 1775) revealed 2n = 24 = 1 III + XiX2 in males with metacentric auto- some s and acrocentric sex chromosomes (Rodrfguez-Gil et al. 2002); Mesabolivar lu- teus (Keyserling 1891) showed 2n = 15 — 711 + X in males and 2n = 16 — 711 + XX in females with a metacentric chromosome mor- phology; in the male specimens of Micro- pholcus fauroti (Simon 1887), the diploid number was 2n = 17 = 811 + X, with the chromosomes being described as biarmed (Ar- aujo et al. 2005b). Existing karyotypic descriptions for pholcid species (Table 1) show that the diploid num- ber varies from 2n — 15 to 2n = 32, that the predominant chromosome morphology is metacentric and that the most frequent SDCS is of the X/XX type. In addition to this type of SDCS, some species of this family pre- sented X1X2/X1X1X2X2 and XlX2Y/XlXlX2X2 types in a decreasing succession of occur- rence. In the other 1 1 haplogyne families, specif- ically Diguetidae, Drymusidae, Dysderidae, Filistatidae, Leptonetidae, Ochyroceratidae, Plectreuridae, Scytodidae, Segestriidae, Sica- riidae, and Tetrablemmidae, from which 28 species have been cytogenetically character- ized (Hackman 1948; Suzuki 1954; Begak & Begak 1960; Diaz & Saez 1966a, 1966b; Be- navente & Wettstein 1980; Silva 1988; Tug- mon et al. 1990; Silva et al. 2002; Krai et al. 2006), the diploid number varies from 2n = 7 to 2n = 37, the predominant chromosome morphology is also metacentric, and the SDCS may be of the X (43%), XiX2 (26%), X,X2Y (24%) or XY (7%) types. The majority of cytogenetically analyzed araneomorph species belong to Entelegynae. Approximately 500 species of entelegyne spi- ders, representing nearly 30 families, have a diploid number varying between 2n = 14 and 2n = 52, a predominantly acrocentric chro- mosome morphology, and a SDCS of the XjX2 type in most species. The few cytoge- netically studied species of Pholcidae and oth- er haplogyne families exhibit karyotypic pe- culiarities, such as predominantly metacentric chromosome morphology and X/XX sex de- termination chromosome system, that con- trasts with those of Entelegynae. The cytoge- netic analysis of other haplogyne species will probably provide additional information that can be useful in establishing some strategies of karyotype differentiation among the species of this group and also between Haplogynae and Entelegynae. Considering the karyotypic peculiarities of Pholcidae, the present work aims to charac- terize the cytogenetics of two species of this family, Crossopriza lyoni and Physocyclus globosus (Taczanowski 1874). The diploid number, chromosome morphology, SDCS type, and behavior of chromosomes during meiosis were determined with conventional staining, and the distribution pattern of the ac- tive nucleolar organizer regions (NORs) in the chromosomes was established using silver staining. METHODS The chromosomal characterization in C. lyoni was performed through the analysis of 27 adult specimens (24 males and 3 females); in P. globosus, 10 adult specimens (7 males and 3 females) and 12 embryos were used. The individuals of both species were collected from natural populations in the city of Rio Claro (22°05'S, 47°30'W), State of Sao Paulo, Brazil. The adult specimens were deposited in the collection of the Butantan Institute, city of Sao Paulo, State of Sao Paulo, Brazil. All an- alyzed adult specimens were collected in Au- gust 2003 and P. globosus embryos were col- lected in January 2004. The gonadal and embryonic chromosome preparations were obtained using the tech- nique described by Webb et al. (1978), al- though a few were prepared from testicles not submerged in colchicine solution. Conven- tional staining was performed using a 3% Gi- emsa solution for 12 to 15 minutes. The NOR silver staining was carried out according to the 296 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — Gonadal mitotic metaphases from adult Crossopriza lyoni individuals. 1, 3. Convention- ally stained, male with 2n — 23 and female with 2n - 24. 2, 4. Same cells seen in 1 and 3, respectively, submitted to silver staining. Arrows indicate the NOR-bearing chromosomes. Scale bar =10 jim. methodology described by Howell & Black (1980). RESULTS Cytogenetics of Crossopriza lyoni. — Anal- ysis of 356 gonadal cells of C. lyoni using conventional staining revealed 2e = 23 chro- mosomes in the spermatogoeial metaphases (Fig. 1) and 2n = 24 in the oogonial meta- phases (Fig. 3), a meiotic formula of 2e = 1 III + X in the spermatocytes I (Fig. 5), and the occurrence of n = 11 or n = 12= 11 + X with metaceetric and submetaceetric chro- mosomes in the metaphases II of males (Figs. 6, 7). These data showed the occurrence of the X/XX sex determining system in C. lyoni. The conventionally stained mitotic meta- phases of C lyoni showed chromosomes with little morphological definition (Figs. 1, 3). In these cells, the chromosome elements of pair 1 and the X sex chromosome were always easily identified as being the largest elements OLIVEIRA ET AL. — CYTOGENETICS OF TWO PHOLCIDS 297 I -l Figures 5-7.-— Spermatocytes of adult Crossopri - za lyoni specimens in conventional staining. 5. Meiocyte In prophase I, with 2n = 1 III + X; note the cross or ring configuration of bivalents 1 and 2, evidencing the occurrence of one and two chias- mata, respectively; the arrows indicate the chro- mosomal elements of a bivalent with precocious separation. 6, 7. Metaphases II, with n = 1 1 and n = 12 = 11 + X, respectively, showing the presence of metacentric and submetacentric chromosomes. Scale bar =10 pm, of the complement; the other chromosomes represented a series of gradually decreasing size. Additionally, the X chromosome was positively heteropycnotic in most gonadal metaphases and spermatogoeial anaphases. The silver-stained spermatogonial meta- phases exhibited 5 telomeric NORs occupying the short arm of the chromosome elements of pair 4, the long arm of the chromosome ele- ments of pair 6, and one of the arms of the metacentric X chromosome (Fig. 2). The oo- gonial metaphases showed only two telomeric NORs in the long arm of the pair 6 chromo- somes (Fig. 4), reflecting an in ter sexual het- erogeneity in the activity of these regions. In the diploteee and diakinesis of the male C. lyoni specimens, the autosomal bivalents showed a regular meiotic behavior of pairing and staining, and the X chromosome appeared as an extremely large and isopycnotic univa- lent (Fig. 5). In these cells, most of the biva- lents possessed an interstitial chiasma with a cross configuration, with the exception of bi- valents 1 and 2 that showed two chiasmata, assuming a ring configuration. In some of these cells, the smallest bivalent showed a precocious separation, but exhibited a regular segregation in the subsequent meiotic phases (Fig. 5). The metaphases II of male C, lyoni , with n 11 or n = 12 = 11 + X, confirmed the regular reductional segregation of all the chro- mosomes in the preceding anaphase I and the meta- and submetacentric morphology of the chromosomes (Figs. 6, 7). In the rnetaphases II with n = 12, the X chromosome was iden- tified through its large size and positive het- eropycnosis. Silver- stained spermatocytes I and II did not show NOR markings on the chromo- somes. However, early prophasic nuclei I from male specimens exhibited a strongly stained nucleolus. Conventionally stained interphasic nuclei of males and females showed one and two pos- itive heteropycnotic chromatinic blocks, re- spectively, which probably corresponded to a sexual chromatin (Fig. 8). The silver staining of the interphasic nuclei of males resulted in the marking of three nucleoli (Fig. 9), corrob- orating the results obtained regarding the number of A g~ NOR- bearing chromosomes in the spermatogonial metaphases, i.e., pairs 4 and 6 and the X chromosome. In the inter- 298 THE JOURNAL OF ARACHNOLOGY Figures 8, 9. — Interphasic nuclei of male Cros- sop riza lyoni. 8. Conventionally stained, showing a conspicuous heteropycnotic-positive chromatinic block (arrow). 9. Silver-stained, evidencing three nucleoli (N). Scale bar = 10 fxm. phasic nuclei of female specimens, only one strongly stained nucleolus was observed, con- firming the number of active NORs verified in the oogonial metaphases. Cytogenetics of Physocyclus globosus. — The testicular cells of the 7 analyzed adult specimens of P. globosus showed intraindi- vidual variation in the number of chromo- somes, i.e., of the 208 metaphases obtained from these individuals, 125 exhibited 15 = 14 + X chromosomes (Figs. 10, 12), and 83 showed 30 = 28 + 2X chromosomes (Figs. 11, 14). Of approximately 30 spermatogonial metaphases obtained from each individual, about 60% of cells showed 15 chromosomes and about 40% possessed 30 chromosomes. The analysis of 136 oogonial and embryonic metaphases showed the occurrence of 16 = 14 + XX chromosomes (Fig. 16) in 9 females (3 adults and 6 embryos) and 15 = 14 + X chromosomes in 6 male embryos. u a it ii 12 3 4 it it tt t 5 6 7 X wo WO WU WO 12 3 4 Sit! lift SUt tl 5 6 7 X I I Figures 10, 1 1 . — Male Physocyclus globosus kar- yotypes submitted to conventional staining. 10. 2n = 15 chromosomes. 11. 4n = 30 chromosomes. Scale bar =10 [xm. The spermatocytes of the adult specimens of P. globosus also exhibited intraindividual variation in the number of chromosomes. Spermatocytes I showed the meiotic formula 7 II + X (Fig. 18) or 1411 + 2X (Fig. 19) in prophase I and metaphase I; spermatocytes II exhibited 7 chromosomes (Fig. 20) or 8 — 7 + X chromosomes (Fig. 21) in the metaphases II, indicating that these came from the sper- matocytes I with 711 + X. Or they possessed 15 = 14 + X chromosomes (Fig. 22), indi- cating that they originated from spermatocytes I with 1411 + 2X. Considering the chromosome numbers ob- tained in the testicular, ovarian and embryonic cells of P. globosus , the diploid number and the chromosomal sex determination system were established: 2n = 15 = 14 + X — 711 + X in males and 2n = 16 = 14 + XX = 711 + XX in females. The testicular cells with 30 = 28 + 2X = 1411 + 2X were interpreted as tetraploids, indicating an intraindividual nu- merical chromosome variation in the male specimens of adult P. globosus. The gonadal and embryonic somatic meta- phases revealed that pairs 1,2,4 and 6 of the P. globosus karyotype are submetacentric, and OLIVEIRA ET AL.— CYTOGENETICS OF TWO PHOLCIDS 299 I 1 Figures 12-17. — Gonadal mitotic metaphases of Physocyclus globosus. 12, 14, 16. Conventionally stained with 2n = 15 (male), 4n = 30 (male tetraploid cell), and 2n = 16 (female), respectively. 13, 15, 17. Same cells seen in 12, 14 and 16, respectively, stained with silver nitrate, showing the NORs in the chromosomes of pair 2 (arrows). Scale bar =10 fxm. that pairs 3, 5 and 7 and the X chromosome are metacentric (Fig. 10). The autosomal pairs could be arranged in a series of gradually de- creasing size and the X chromosome was a size intermediate between chromosome pairs 1 and 2. Only mitotic metaphases of males and fe- males were subjected to silver staining, which revealed a NOR in the proximal region of the long arm of the second pair of chromosomes (Figs. 13, 15, 17). The metaphases with 2n | 15 and those with 2n = 16 exhibited a max- imum number of two NOR-bearieg chromo- somes, while the metaphases with 30 chro- mosomes presented a maximum number of four NOR-bearing chromosomes; in these 300 THE JOURNAL OF ARACHNOLOGY I— — I I— — t Figures 18-22. — Conventionally stained meiotic cells of male Physocyclus globosus. 18, 19. Diplotenes exhibiting 2n = 711 + X and 4n = 1411 + 2X, respectively. 20, 21, 22. Metaphases II with 7 chromosomes, 8 = 7 + X chromosomes and 15 = 14 + X chromosomes, respectively. Scale bar =10 [xm.. metaphases, the NORs were always marked in the chromosomes of pair 2. In the P. globosus testicular chromosome preparations, meiocytes with 15 or 30 chro- mosomes were found in all the phases of mei- osis I. In the subphases prophase I and meta- phases I, autosomal bivalents and the X univalent always occurred, even in the cells with 30 = 1411 + 2X (Figs. 18, 19). In the diplotene cells, the occurrence of an intersti- tial chiasma was observed in three bivalents, in the cells with 15 = 711 + X, and in six bivalents, in the cells with 30 = 1411 + 2X. Metaphases II showed chromosome numbers that confirmed the reductional segregation of the chromosomes during the preceding ana- phase I, including the asynaptic X chromo- somes present in the cells with 30 = 1411 + 2X. The silver- stained testicular meiocytes pro- vided no information on NORs or nucleolar material, with the exception of pachytene cells that exhibited a single strongly stained block of nucleolar material. Conventionally stained testicular and ovar- ian interphasic nuclei revealed the presence of one or two large heteropycnotic-positive blocks, respectively, which are probably re- lated to the sex chromatin (Figs. 23, 24). The occurrence of two chromatinic blocks in a few OLIVEIRA ET AL. — CYTOGENETICS OF TWO PHOLCIDS 301 Figures 23-25. — -Testicular interphasic nuclei of Physocyclus globosus. 23, 24. Conventionally stained, emphasizing one and two conspicuous faet- eropycnotic-positive chromatinic blocks, respec- tively. 25. Same nucleus seen in 23 submitted to silver staining, evidencing one nucleolus (N). Scale bar =10 pan testicular interphasic nuclei (Fig. 24) suggest- ed the presence of cells with two X chromo- somes in males. Silver-stained interphasic nu- clei exhibited a single marked nucleolus (Fig. 25). DISCUSSION The cytogenetical data obtained from C lyoni and P. globosus in relation to the meta- centric and submetacentric morphology of all the chromosomes of the complement and the presence of an X/XX sex determination sys- tem are similar to those described for related species belonging to the Nearctic, Neotropi- cal, Oriental and Palearctic regions (Bole- Gowda 1958; Srivastava & Shukla 1986; Far- ida & Sharma 1987; Sharma & Panda 1987; Cokendolpher 1989; Araujo et al. 2005b; Krai et al. 2006). However, the studied species showed some particularities regarding chro- mosome number when compared with related species described in the literature. The C lyoni specimens analyzed in this work showed a diploid number (2n = 23 = 22 + X) similar to the one found by Parida & Sharma (1987) and Sharma & Parida (1987) in specimens from the same species from two different Indian populations. Nev- ertheless, this diploid number differs from those described by Bole-Gowda (1958) - 2n = 27 = 26 + X, Sharma et al. (1959) - 2n = 24 = 22 + XjX* and Srivastava & Shukla (1986) - 2e = 25 — 24 + X, for individuals belonging to other, more geographically dis- tant Indian populations. Chromosome rearrangements of the centric fission, followed or not by pericentric inver- sion, and/or centric fusion types are suggested in order to explain the interpopulational and intraspecific numerical variation found in C. lyoni. The presence of predominantly meta- ceetric or submetacentric autosomes and of a SDCS of the XjX2 type, with acrocentric X chromosomes, or of the X type, with a rneta • centric X chromosome, corroborate such mechanisms of karyotypic differentiation. Chromosomal variations of the diploid number have already been described for some species of entelegyne spiders, such as Ageleno limbata Thorell 1897 (Agelenidae), Delena cancerides Walckenaer 1837 (Sparassidae) and Evarcha hoyi (Peckham & Peekhaml883) [as Pellenes hoyi (Peckham & Peckham 1909)] (Salticidae). In each one of these spe- cies, karyotypes belonging to different popu- lations were characterized as chromosomal races that appeared to have originated mainly by centric or tandem fusion, involving only autosomes or autosomes and sex chromo- somes (M addi son 1982; Rowell 1985, 1990, 1991; Tsurusaki et al. 1993; Hancock & Row- ell 1995). Likewise, the different karyotypes present in C. lyoni could represent chromo- somal races or cytotypes. There are three works that focus on the phylogeny of pholcid spiders (Huber 2000; Bmvo-Madaric et al. 2005; Astrin et al. 2006) and could be used to hypothesize the origin 302 THE JOURNAL OF ARACHNOLOGY of the C. lyoni cytotypes. However, the study of Astrin et al. (2006) did not include Sper- mophora senoculata, which Bruvo-Madaric et al. (2005) considered basal to all pholcines and part of Holocneminae. Due to its type of SDCS, this species was also considered to be the most basal of all pholcid species already analyzed from the cytogenetic point of view (Krai et al. 2006). The phytogeny proposed by Huber (2000) was based on morphological characters and that of Bruvo-Madaric et al. (2005) was made using both morphological and molecular data. Considering the phylogeny described by Bruvo-Mararic et al. (2005), S. senoculata , with 2n = 25 = 11 II + X,X2Y (Krai et al. 2006), is basal in relation to C. lyoni. The XtX2Y system of basal pholcid species could give rise to an XlX2 system, such as that reg- istered for one C. lyoni Oriental population with 2n = 24 = 1 III + X,X2 (Sharma et al. 1959), by gradual heterochromatinization and erosion of the Y sex chromosome. Taking into account this process of SDCS differentiation, the 2n = 24 = 1 III + X1X2 could represent the basic karyotype of C. lyoni. The other dip- loid numbers obtained for this species could be derived from this basic number. The pro- cess of Y sex chromosome heterochromatin- ization and erosion have been detected in many groups of arthropods and considered a usual event involved in SDCS evolution (White 1973; Smith & Virkki 1978; Steine- manm & Steinemanm 1998). On the other hand, the possibility of conversion of an X,X2Y system into an X system, as postulated by Krai et al. (2006) for some Haplogynae species, can not be excluded, especially if we consider the 2n = 23 = 1 III + X of Holoc- nemus caudatus (Krai et al. 2006), which rep- resents a genus closely related to Crossopriza (Huber 2000; Bruvo-Madaric et al. 2005). If so, the C. lyoni populations with 2n = 23 = 1111 + X could be ancestral. The proposal that 2n = 24 originated the chromosomal races in C. lyoni suggests kar- yotypic evolution by raising or lowering the diploid number of chromosomes and an origin of the X,X2 sex determination chromosome system from the X,X2Y system. The proposal that 2n = 23 is the ancestral condition re- quires an increase in the chromosome number and origin of an X system from an XjX2Ysystem. These propositions differ from those elaborated by other researchers, such as Suzuki (1951, 1954), Postiglioni & Brum-Zor- rila (1981), Maddison (1982), Rowell (1985, 1990) and Datta & Chatterjee (1988), in order to explain the karyotypic differentiation of most spider species. Considering that the highest chromosome numbers occur in some less morphologically derived spider species (Mesothelae and My- galomorphae) and that the acrocentric chro- mosome morphology and X1X2 sex determi- nation chromosome system are the most frequent among Araneae, these researchers have suggested that the above-mentioned kar- yotypic characteristics would be ancestral to Araneae. Lower chromosome numbers and other types of chromosome morphology and sex determination systems, particularly of the X and XjX2X3Y types, would be derived mainly from centric fusions followed or not by pericentric inversions, or by tandem fu- sions. Alternatively, some of these researchers postulated that the existence of an X sex de- termination chromosome system as an ances- tral condition can not be ruled out and X chro- mosome centric fission from species with a metacentric X chromosome could give rise to an XjX2 system. Nevertheless, cladistic analyses have indi- cated that Mesothelae, Mygalomorphae and Araneomorphae (Haplogynae and Entelegy- nae) have undergone independent processes of morphological differentiation. Independent processes also seem to have promoted the di- versification between haplogyne and entele- gyne spiders (Platnick et al. 1991; Griswold et al. 1999; Ramirez 2000). These data raise the possibility of an independent karyotypic differentiation in the Mesothelae, Mygalo- morphae, Haplogynae and Entelegynae spi- ders, i.e., the karyotypic differentiation of ex- tant spiders may occur by a raise or lowering of the basic chromosome number. Unfortunately, we do not have enough ev- idence to determine the process of karyotypic differentiation among the C. lyoni cytotypes. Cytogenetical analysis of other Crossopriza species will certainly provide additional infor- mation that will allow a more secure estab- lishment of the characteristics of the basic kar- yotype of the species of this genus, as well as the mechanisms involved in the origin of the chromosomal races or the cytotypes of C. lyoni . OLIVEIRA ET AL.— CYTOGENETICS OF TWO PHOLCIDS 303 Considering that C. lyoni is a species with a wide geographic distribution, the karyotype diversity recorded for this species is not sur- prising and the occurrence of other cytotypes can not be excluded. Therefore, it is not pos- sible to disregard the hypothesis that C. lyoni represents a species complex. Cytogenetic analyses may be useful in understanding the taxonomy of this species, that is, if distinct cytotypes are sympatric and there are not hy- brid karyotypes. In a few animal groups whose species are very morphologically sim- ilar, cytogenetic studies coupled with morpho- logical analyses have promoted the discovery of new species (Silva & Yonenaga-Yassuda 1998; Bertollo et al. 2000). In the P. glohosus sample analyzed, the number of chromosomes found in females, 2n = 16 = 14 + XX, and male embryos, 2n - 15 = 14 + X, is coincident with those de- scribed by Cokendolpher (1989) for other Physocyclus species, namely P. californicus , P. enaulus and Physocyclus sp. However, an intraindividual variation in the number of chromosomes, i.e,, 2n = 15 = 14 + X and 4n = 30 = 28 + 2X, was observed in the testic- ular cells of the adult specimens of P. glo- hosus. The presence of tetraploid cells in the male germ line of P. globosus is probably related to the occurrence of cytoplasmatic bridges be- tween cells of the same cyst. These bridges form a syncytium and promote synchroniza- tion in cell division and cell differentiation (Alberti & Weinmann 1985; Alberts et al. 2002; Michalik et al. 2003). Due to some pe- culiarities of these cytoplasmatic bridges, cell couples from a single cyst remained connect- ed during chromosome preparation, leading to the formation of cells that were apparently tet- raploid, and of interphasic nuclei with two sexual chromatinic blocks. In fact, the chro- mosomes of the resulting tetraploid cells ex- hibited the same degree of condensation and meiotic behavior. In the pholcid Mesabolivar luteus, some diplotene cells appeared in pairs, and the authors also suggested that the orga- nization of the testicular cells was responsible for this apparent tetraploidy (Araujo et al. 2005b). Cokendolpher & Brown (1985) also veri- fied the presence of a few polyploid cells in Physocyclus sp., which was attributed to the cell treatment with a colchicine solution. In P. globosus, the numerical chromosome varia- tion was certainly not due to the cell treatment with the colchicine solution, because this var- iation was also observed in preparations where the cells were not subjected to this so- lution. The meiotic testicular cells of C. lyoni and P. globosus showed that the autosomal biva- lents and the univalent X chromosome exhib- ited a regular behavior similar to those de- scribed by Suzuki (1954), Bole-Gowda (1958), Cokendolpher (1989), Tugmon et al. (1990), Gorlov et al. (1995), Gorlova et al. (1997), Shyh-Hwang (1999) and Rodriguez- Gil et al. (2002) for most Araneae in terms of condensation, synapsis, chiasma number and chromosome segregation. The occurrence of the nucleolus or NORs associated with specific chromosomes has been described in some spiders based on ul- trastructural analysis of bisected testicular cells using transmission electron microscopy (Benavente & Wettstein 1980; Wise 1983) or on analysis of silver impregnated gonadal and embryonic metaphases using light microscopy (Araujo et al. 2005a; Krai et al. 2006). In these analyses, the nucleolar material was associat- ed with the X chromosome in some haplogyne spiders, such as Dysdera crocata Koch 1838 (Dysderidae), with 2n = 11 = 10 + X (Be- navente & Wettstein 1980), Ochyrocera sp. Simon 1891 (Ochyroceratidae), with 2n = 13 = 12 + X (Krai et al. 2006), Scytodes thor- acica (Latreille 1802) (Scytodidae), with 2n = 19 = 18 + X (Krai et al. 2006), and Mon- oblemma muchmorei Shear 1978 (Tetrablem- midae), with 2n = 23 = 22 + X (Krai et al. 2006), and with autosomes of some entele- gyne species, such as two autosomal bivalents of Allocosa georgicola (Walckenaer 1837) (as Lycosa georgicola ) (Lycosidae, Entelegynae), with 2n = 28 = 26 + XtX2 (Wise 1983), and three autosomal pairs of Nephilengys cruen- tata (Fabricius 1775) (Nephilidae), with 2n = 22 + X1X1X2X2 (Araujo et al. 2005a). In C. lyoni, five NORs were found occu- pying the telomeric regions of the pair 4, pair 6 and X chromosomes, while in P. globosus, two NORs were found in the interstitial region of the pair 2 chromosomes. The presence of a NOR in the X chromosome of C. lyoni, D. crocata, Ochyrocera sp., Scytodes thoracica, and Monoblemma muchmorei, all bearing an X/XX sex determination system type, sug- 304 THE JOURNAL OF ARACHNOLOGY gests that the X chromosome can represent one of the elements that constitutes the basic NOR pattern in “Higher Haplogynes’X^/ww Coddington & Levi 1991). The absence of a NOR marking in the X chromosome of P. glo - bosus is possibly due to chromosome rear- rangements or differential activation of this region. Considering the low number of species whose chromosomes have been subjected to silver staining, it is not yet possible to deter- mine a quantitative pattern of active NORs in spiders. The interpopulatioeal and intraindividual numerical variations respectively found in C. lyoni and P. globosus have different origins and meanings in the two species. In C. lyoni , the interpopulational variation shows that structural chromosome rearrangements are acting upon the karyotypic evolution of this species. On the other hand, the apparent tet- raploidy of the spermatogoeial metaphases in P. globosus is a result of the tissue organiza- tion of the spermatogonia. Thus, variation in P globosus seems to have no relationship to chromosomal evolution of this species, but re- flects a mechanisms which guarantees that great quantities of spermatozoids are simul- taneously differentiated at the moment of re- production. ACKNOWLEDGMENTS We thank D. Araujo and M.C. Schneider from Ueiversidade Estadual Paulista (UNESP), Departamento de Biologia, Rio Claro, State of Sao Paulo, Brazil, and two anonymous reviewers for comments, sugges- tions, and critical review of the manuscript. LITERATURE CITED Alberti, G. & C. Weinmann. 1985. Fine structure of spermatozoa of some labidognath spiders (Fills- tatidae, Segestriidae, Dysderidae, Oonopidae, Scytodidae, Pholcidae; Araneae; Arachnida) with remarks of spermiogenesis. Journal of Morphol- ogy 185:1-35. Alberts, B„, A. Johnson, I. Lewis, M. Raff, K. Rob- erts & P. Walter. 2002. Molecular Biology of the Cell. Garland Science, New York. 1616 pp. Araujo, D., D.M. Celia & A.D. Brescovit. 2005a. Cytogenetic analysis of the neotropical spider Nephilengys cruentata (Araneomorphae, Tetrag- nathidae): standard staining, NORs, C-bands and base-specific fluorochromes. Brazilian Journal of Biology 65:193-202. Araujo, D., A.D. Brescovit, C.A. Rheims & D.M. Celia. 2005b. Chromosomal data of two pholcids (Araneae, Haplogynae): a new diploid number and the first cytogenetical record for the new word clade. Journal of Arachnoiogy 33:591-596. Astrin, J.J., B.A. Huber, M. Bernhard & C.F.C. Kliitsch. 2006. Molecular taxonomy in pholcid spiders (Pholcidae, Araneae): evaluation of spe- cies identification methods using COl and 16S rRMA. Zoologica Scripta 35:441-457. Bertollo, L.A.C., G.G. Bom, J.A. Dergam, A.S. Fenocchio & O. Moreira-Filho. 2000. A biodi- versity in the neotropical Erythrinidae fish. Ho - plias malabaricus. Karyotypic survey, geograph- ic distribution of cytotypes and cytotaxonomic considerations. Chromosome Research 8:603- 613. Begak, W. & M.L. Bee ah. 1960. Constituigao cro- mossomica de duas especies de aranhas do ge- nero Loxosceles. Revista Brasileira de Biologia 20:425-427. Benavente, R. & R. Wettstein. 1980. Ultrastructural characterization of the sex chromosomes during spermatogenesis of spiders having holocentric chromosomes and a long diffuse stage. Chro- mosoma 77:69-81. Bole-Gowda, B.N. 1958. A study of the chromo- somes during meiosis in twenty-two species of Indian spiders. Proceedings the Zoological So- ciety of Bengal 11:69-108. Bruvo-Madaric, B., B.A. Huber, A. Steinacher & G. Pass. 2005. Phylogeny of pholcid spiders (Ara- neae: Pholcidae): combined analysis using mor- phology and molecules. Molecular Phylogenetics and Evolution 37:661-673. Coddington, J.A. & H.W. Levi. 1991. Systematics and evolution of spiders (Araneae). Annual Re- view of Ecology and Systematics 22:565—592. Cokendolpher, J.C. 1989. Karyotypes of three spi- der species (Araneae: Pholcidae: Physocyclus ). Journal of the New York Entomological Society 97:475-478. Cokendolpher, J.C. & J.D. Brown. 1985. Air-dry method for studying chromosomes of insects and arachnids. Entomological News 96:114-118. Datta, S.N. & K. Chatterjee. 1988. Chromosomes and sex determination in 13 araneid spiders of North-Eastern India. Genetica 76:91-99. Diaz, M.O. & F.A. Saez. 1966a. Investigaciones ci- togeneticas sobre algunas species de araneidos Uraguayos. Congress© Latino Americano de Zoologia, Anais 2:3-9. Diaz, M.O. & F.A. Saez. 1966b. Karyotypes of South- American Araneida. Memorias do Insti- tute Commemorativo 33:153—154. Gorlov, I.P., O.Y. Gorlova & D.V. Logunov. 1995. Cytogenetic studies in Siberian spiders. Heredi- tas 122:211-220. Gorlova, O.Y., I.P. Gorlov, E. Nevo & D.V. Logu- nov. 1997. Cytogenetic studies on seventeen spi- OLIVEIRA ET AL.— CYTOGENETICS OF TWO PHOLCIDS 305 der species from Israel. Bulletin of the British Arachnological Society 10:249-252. Griswold, C.E., J.A. Coddington & N.I. Platnick. 1999. Towards a phylogeny of Entelegyne spi- ders (Araneae, Araneomorphae, Entelegynae). Journal of Arachnology 27:53-63. Hackman, W. 1948. Chromosomenstudien an Ara- neen mit besonderer Berucksichtigung der Ges- chlechtschromosomen. Acta Zoologica Fennica 54:1-101. Hancock, A.J. & D.M. Rowell. 1995. A chromo- somal hybrid zone in Australian huntsman spider, Delena cancerides (Araneae: Sparassidae). Evi- dence for a hybrid zone near Canberra, Australia. Australia Journal of Zoology 43:173-180. Howell, W.M. & D.A. Black. 1980. Controlled sil- ver staining of nucleolus organizer regions with protective colloidal developer: a 1-step method. Experientia 36:1014-1015. Huber, B.A. 2000. New World pholcid species (Ar- aneae: Pholcidae): a revision at generic level. Bulletin of the American Museum of Natural History 254:1-347. Krai, J., J. Musilova, F. St’ahlavsky, M. Rezac, Z. Akan, R.L. Edwards, F.A. Coyle & C.R. Almerje. 2006. Evolution of the karyotype and sex chro- mosome systems in basal clades of araneomorph spiders (Araneae: Araneomorphae). Chromo- some Research 14:859-880. Maddison, W.P. 1982. XXXY sex chromosomes in males of the jumping spiders genus Pellenes (Ar- aneae: Salticidae). Chromosoma 85:23-37. Michalik, R, M.R. Gray & G. Alberti. 2003. Ultra- structural observations of spermatozoa and sper- miogenesis in Wandella orana Gray, 1994 (Ar- aneae: Filistatidae) with notes on their phylogenetic implications. Tissue & Cell 35: 325-337. Painter, T.S. 1914. Espermatogenesis in spiders. Zoologische Jahrbiicher Abteilung fur Anatomie und Ontogenie der Tiere 38:509-576. Parida, B.B. & N.N. Sharma. 1987. Chromosome number, sex mechanism and genome size in 27 species of Indian spiders. Chromosome Infor- mation Service 43:11-13. Platnick, N.I., J.A. Coddington, R.R. Forster & C.E. Griswold. 1991. Spinneret morphology and the phylogeny of haplogyne spiders (Araneae, Ara- neomorphae). American Museum Novitates 3016:1-73. Platnick, N.I. 2007. The World Spider Catalog, Ver- sion 7.5. American Museum of Natural History, New York. Online at: http://research.amnh.org/ entomology/ spiders/catalog/index . html . Postiglioni, A. & N. Brum-Zorrilla. 1981. Karyo- logical studies on Uruguayan spiders. II. Sex chromosomes in spiders of the genus Lycosa (Araneae: Lycosidae). Genetica 56:47-53. Ramirez, M.J. 2000. Respiratory system morphol- ogy and the phylogeny of haplogyne spiders (Ar- aneae, Araneomorphae). Journal of Arachnology 28:149-157. Rodrfguez-Gil, S.G., L.M. Mola, A.G. Papeschi & C.L. Scioscia. 2002. Cytogenetic heterogeneity in common haplogyne spiders from Argentina (Arachnida, Araneae). Journal of Arachnology 30:47-56. Rowell, D.M. 1985. Complex sex-linked fusion het- erozygosity in the Australian huntsman spider Delena cancerides (Araneae: Sparassidae). Chro- mosoma 93:169-176. Rowell, D.M. 1990. Fixed fusion heterozygosity in Delena cancerides Walck. (Araneae: Sparassi- dae): an alternative to speciation by monobra- chial fusion. Genetica 80:139-157. Rowell, D.M. 1991. Chromosomal fusion and mei- otic behaviour in Delena cancerides (Araneae: Sparassidae). I. Chromosome pairing and X- chromosome segregation. Genome 34:561-566. Sharma, G.P., B.L. Gupta & R. Parshad. 1959. Cy- tological studies on the Indian spiders. III. An analysis of the chromosomes in the male germ cells of the spider Crossopriza lyoni (Blackwall), Fam. Pholcidae. Research Bulletin of the Panjab University 10:49-53. Sharma, G.P. & B.B. Parida. 1987. Study of chro- mosome in spiders from Orissa. Pranikee 8:71- 76. Shyh-Hwang, C. 1999. Cytological studies on six species of spiders from Taiwan (Araneae: Ther- idiidae, Psechridae, Uloboridae, Oxyopidae, and Ctenidae). Zoological Studies 38:423-434. Silva, D. 1988. Estudio cariotfpico de Loxosceles laeta (Araneae: Loxoscelidae). Re vista Peruana de Entomologia 31:9-12. Silva, M.J. & Y. Yonenaga-Yassuda. 1998. Karyo- type and chromosomal polymorphism of an un- described Akodon from Central Brazil, a species with the lowest known diploid chromosome number in rodents. Cytogenetics and Cell Ge- netics 81:46-50. Silva, R.W., D.R. Klisiowicz, D.M. Celia, O C Mangili & I.J. Sbalqueiro. 2002. Differential dis- tribution of constitutive heterochromatin in two species of brown spider: Loxosceles intermedia and Loxosceles laeta (Araneae, Sicariidae), from the metropolitan region of Curitiba, PR (Brazil). Acta Biologica Paranaense 31:123-136. Smith, S.G. & N. Virkki. 1978. Animal Cytogenet- ics. 3. Insecta. 5. Coleoptera. Gebriider Borntrae- ger, Berlin. Srivastava, M.D.L. & S. Shukla. 1986. Chromo- some number and sex determing mechanism in forty seven species of Indian spiders. Chromo- some Information Service 41:23-26. Steinemanm, M. & S. Steinemanm. 1998. Enigma of Y chromosome degeneration: Neo-Y and Neo- X chromosomes of Drosophila miranda a model 306 THE JOURNAL OF ARACHNOLOGY for sex chromosome evolution. Genetica 102/ 103:409-420. Suzuki, S. 1951. Cytological studies in spiders. I. A comparative study of the chromosomes in the family Argiopidae. Journal of Science of the Hi- roshima University (ELI) 12:67-98. Suzuki, S. 1954. Cytological studies in spiders. III. Studies on the chromosomes of fifty-seven spe- cies of spiders belonging to seventeen families, with general considerations on chromosomal evolution. Journal of Science of the Hiroshima University (B-l) 15:23-136. Tsurusaki N., Y. Ihara & T. Arita. 1993. Chromo- somes of the funnel-web spider Agelena limbata (Araneae: Agelenidae). Acta Arachnologica 42: 43-46. Tugmon, C.R., J.D. Brown & N.Y. Homer. 1990. Karyotypes of seventeen USA spiders species (Araneae, Araneidae, Gnaphosidae, Loxosceli- dae, Lycosidae, Oxyopidae, Philodromidae, Sal- ticidae and Theridiidae). Journal of Arachnology 18:41-48. Wang, X.-Z., S.-J. Cui, Z.-L. Yang, J AP. Wang & 1 Y.-J. Wang. 1997. On karyotype of the Pholcus qffinis (Araneae: Pholcidae). Acta Arachnologica Sinica 6:19-22. Webb, G.C., MJ.D. White, N. Contreras & J. Che- ney. 1978. Cytogenetics of the parthenogenetic grasshopper Warramaba (formerly Moraba) vir- go and its bisexual relatives. IV. Chromosome banding studies. Chromosoma 67:309-339. White, MJ.D. 1973. Animal Cytology and Evolu- tion. 3rd Edition. Willian Clowes & Sons, Lon- don. 961 pp. Wise, D. 1983. An electron microscope study of the | karyotypes of two wolf spiders. Canadian Jour- nal of Genetics and Cytology 25:161-168. Manuscript received 13 March 2006, revised 5 April 2007 . 2007. The Journal of Arachnology 35:307-312 A NOVEL TRAP TO CAPTURE BALLOONING SPIDERS Chris Woolley , C. F. George Thomas and Linda Hutchings: School of Biological Sciences, University of Plymouth, Drake Circus, Plymouth, Devon, PL4 8AA, UK. E-mail: c.woolley@plymouth.ac.uk Sara Goodacre and Godfrey M. Hewitt: School of Biological Sciences, University of East Anglia, Norwich, NR4 7TJ, UK Steve P, Brooks: Statistical Laboratory, Centre for Mathematical Sciences, Wilberforce Road, Cambridge, CB3 OWB, UK ABSTRACT. An unattended trap was designed to sample and retain spiders dispersing from agricultural grassland and crops. Traps comprised a removable bottle- trap fixed to the top of a vertical metal rod or “climbing-stick” that spiders climbed during normal pre-ballooning behavior. Bottle-traps caught over eight times - more spiders than sticks treated with insect trapping adhesive. Draping sticks with nets in- creased the effective area of the traps and increased the catch size threefold. On average, 9.1% of spiders were lost from traps during the daytime sampling period. No difference in average rate of loss of spiders from the bottle-traps was observed between night and daylight hours. The bottle-trap design is economical and simple to construct, erect and operate. Continuous sampling also allows multiple traps to be used simultaneously in various locations. Keywords; Aerial dispersal, sampling, bottle-trap, climbing- stick Aerial dispersal by ballooning is a key strat- egy in the life histories of many spiders, es- pecially pioneers of disturbed, patchy habitats exemplified by lieyphiids in agricultural land- scapes (Thomas et aL 2003a). Quantifying the dispersal power of these species is a necessary prerequisite for accurately modeling spatial population dynamics and developing success- ful sustainable management strategies. Vari- ous techniques that actively or passively in- tercept airborne spiders have been used to measure aspects of aerial dispersal. For ex- ample: the use of nets and sticky traps to mea- sure aerial density at one or more altitudes (Greenstone et aL 1987; Greenstone 1991; Thomas et aL 2003b); manual collection from fences, wire, or string to quantify numbers passing a point or line per unit time (Vugts & Van Wingerden 1976; Thomas et al. 2003b); or water traps to quantify deposition rates per unit area (Weyman et al. 1995; Thomas & Jepson 1999). These methods are either labor intensive, require operator attendance, cannot easily sample several locations at the same time, or may be cumbersome or expensive. An alternative sampling method exploits the climbing behavior normally exhibited by spiders as a precursor to ballooning (Black- wall 1827): spiders climb to a high point where a silk line can be produced above the surrounding vegetation and where suitable at- mospheric conditions for successful balloon- ing are likely to occur (Suter 1999). Sticks, canes or similar objects inserted into the ground, provide artificial platforms that stand higher than the surrounding vegetation. Spi- ders climbing and attempting to balloon from these can be observed, or caught and counted, to give a relative Indication of ballooning ac- tivity over a given period. Tborbek et al. (2002), in a validation of this technique, found that numbers of spiders observed climbing a 30 cm stick correlated well with numbers ob- tained from an aerial suction trap. Using a similar technique to sample several habitats over time, Duffey (1956) applied a tacky ad- hesive to the tops of canes to trap climbing spiders. However, the adhesive was adversely affected by hot, cold or wet weather and be- came clogged with winged insects during summer months. This paper describes and evaluates a novel 307 308 THE JOURNAL OF ARACHNOLOGY Figures 1-6. — Trap construction. 1. Two liter soft-drinks bottle. 2. Bottle bottom with the five rein- forcements removed. 3. Top removed and section below discarded. 4. Inverted top inserted into the re- maining section and secured with adhesive tape. 5. Screw cap glued underneath the central hub. 6. Finished trap with fine gauze fastened in place with a rubber band. design that develops the climbing-stick into a trap to allow continuous unattended sampling without the use of adhesive. Attached to the top of a climbing-stick is a “bottle-trap” op- erating on the lobster-pot principle. Climbing spiders are retained within the bottle-trap until it is removed or replaced. In the present paper we compare the trapping efficiencies of climb- ing sticks either with bottle-traps or with ad- hesive. The trap collects spiders climbing from the underlying vegetation before they first become airborne, and spiders already airborne arriving at the trap from sources upwind. In the present paper we do not differentiate between these two potential sources. However, we evaluate the effect of suspending a net skirt from the climbing-stick to increase the effective verti- cal and horizontal cross-sectional area of the trap. This increases both the source area of spiders emerging from the ground and the in- terception of airborne spiders. METHODS Trap construction. — The “lobster-pot” part comprising the bottle-trap was construct- ed from a standard straight-sided, clear plastic, 2-liter soft-drinks bottle (Fig. 1). The body of the trap was made by first removing, with a heated scalpel blade, the material between the five reinforcing moldings in the base (Fig. 2). The top section of the bottle was then re- moved, just below the shoulder, approximate- ly 9 cm from the top of the bottle opening (Fig. 3). A band approximately 7 cm deep was cut away from the main body and discarded. The removed top section was then inverted and fixed into the remaining base section of the bottle using adhesive tape (Fig. 4), ensur- ing no gaps remained between the two sec- tions. A 2 ml micro-tube screw-cap (Sar- stedt®, A.G. Sarstedt & Co, Niimbrecht, Germany) was glued with super-glue (Loc- tite®, Henkel, Dusseldorf, Germany) centrally beneath the now inverted base section and above the original bottle top opening forming the new base (Fig. 5). A 20 X 20 cm square of white voile gauze fabric was then fastened tightly over the five cut-away openings with a rubber band (Fig. 6). The cut-away openings covered in fine gauze voile material allowed vertical air flow, general ventilation and, when removed, the extraction of spiders from the trap. The climbing-stick was made from a 1.5 m length of 7.9 mm diameter aluminum rod. The surface was roughened with sandpaper to as- sist climbing spiders. WOOLLEY ET AL. — NOVEL DESIGN OF CLIMBING-STICK TRAP 309 Figures 7-11. — Trap construction. 7. Micro-tube. 8. Micro-tube with bottom removed, pushed over the end of the climbing-stick and glued in position. 9. Circular wire frame. 10. Netting pulled over pole with circular wire frame placed over netting. 1 1 . Finished trap with bottle-trap screwed on and net clipped to stick. An attachment for the bottle-trap was made using the body of the 2 ml micro-tube from which came the cap that had been glued to the bottle- trap. The bottom section of the main body was removed just above the taper (Fig. 7). A small amount of rapid drying epoxy res- in (Araldite®, Huntsman Advanced Materials, Everberg, Belgium) was applied to the inside of the tube, which was then placed over the end of the climbing- stick with the thread end uppermost and extending approximately 5 mm above the end (Fig. 8). The net was constructed from 2 cm mesh bird netting made from a natural-fibre twine. Sufficient material to form a small tent was draped over a 1.2 m wooden pole. A 3.14 m length of 2 mm fencing wire, formed into a 1 m diameter circle (Fig. 9) was placed over the netting and pole to weigh down the base of the net and keep it splayed out. The netting was pulled taut over the pole, arranged evenly around the frame, and its hem secured to the circular base with wire ties before cutting away excess material. (Fig. 10). Setting and operating the trap* — To set the trap, the climbing-stick was pushed verti- cally into the ground, and a bottle-trap placed over and screwed to the top of the stick. If a net was also used, this was first pulled up to form a cone and the climbing- stick placed through the apex before the stick was pushed into the ground. The net was then clipped to the stick using a small bulldog clip set at an angle to ensure the spiders continued climb- ing. The circular wire base was held down with wire pegs or stones. The bottle-trap was thee screwed to the top of the stick (Fig. 11). For continual sampling, bottle-traps were unscrewed and replaced with empty ones. For daily samples reported here, traps were typi- cally changed each evening after ballooning behavior had finished. Removed traps were placed in plastic bags in the field before re- turning to the lab. Spiders were extracted from traps by removing the voile gauze and shaking vigorously over a tray from which spiders were collected with an aspirator. Any spiders remaining in the trap were removed either with an aspirator or, if there was a lot of silk in the trap, with a small paint brash. Trap evaluation. — Experiments were per- formed with traps set along a transect in an 8 ha grass field on the estate farm at the Seale- Hayne Faculty, Newton Abbot, Devon, in the southwest of the UK. The temporary grass ley was approximately 150 mm tall at the time of sampling. The transect, orientated north-south, traversed the brow of a hill, the mid-section 310 THE JOURNAL OF ARACHNOLOGY Table 1 . — Total number of spiders caught per trap over an 1 1 day period from climbing-sticks with bottle-traps and climbing-sticks with adhesive. Trap number 1 2 3 4 5 6 7 8 9 10 Bottle-trap 18 17 14 46 78 107 131 75 53 25 Adhesive 1 6 8 9 6 5 16 7 8 0 being elevated relative to the extremities. An electric fence was used to protect the transect from disturbance by sheep and cattle that pe- riodically grazed the field. Three aspects of the trap were evaluated: catch size from climbing-sticks with bottle- traps compared with climbing-sticks with a polybutene-based insect trapping adhesive (Oecotak A5®, Oecos Ltd, Kimpton, Hert- fordshire, England) applied to the uppermost 15 cm of the stick; catch size from climbing- sticks and bottle-traps with and without nets; retention of spiders left in bottle-traps during the day and overnight. To compare catch size from climbing-sticks with either bottle-traps or adhesive, 10 traps of each design were set alternately at 10 m intervals. Bottle-traps were emptied on each of 11 successive days in March 2003; climb- ing-sticks with adhesive accumulated spiders over the same period. Climbing sticks with adhesive were checked periodically to ensure that the accumulation of trapped spiders or in- sects was not excessive and that there was am- ple exposed adhesive to maintain capture ef- ficiency. Total numbers caught per trap were recorded at the end of the sampling period. For catch size evaluations comparing climb- ing-sticks and bottle-traps with and without nets, 10 traps of each design were set alter- nately at 10 m intervals. Samples were taken and recorded daily over a 13 day period in March 2004. For the retention study, 10 climbing-sticks with bottle-traps were placed in the field as above. Numbers of spiders in each bottle-trap were recorded after 24 h at 17:00. Traps were then relocated to a tarmac substrate away from ground vegetation to minimize further ingress of spiders. Numbers of spiders remaining in the traps were again recorded at 09:00 and at 17:00 the following day. RESULTS Comparison between climbing-sticks with bottle-traps and climbing-sticks with adhesive. — For all traps, catch sizes were higher for climbing-sticks with bottle-traps than for climbing-sticks with adhesive (Table 1). Total catch size over the period for climb- ing-sticks with bottle-traps was 564 spiders and for climbing-sticks with adhesive, 66 spi- ders. Comparison between bottle-traps with and without nets. — Climbing-sticks with nets caught greater numbers of spiders than those without nets for 7 days out of the 13 day pe- riod (Table 2). Spiders were not recorded in any trap on 22, 23, 24, 28, and 29 March when high wind speeds suppressed ballooning activity. No differences were recorded on 26 March though catch size was very low with only 2 spiders recorded in all traps together. The total numbers of spiders caught by climb- ing-sticks with and without nets were 641 and 218 respectively. Retention of spiders in bottle-traps. — Of a total of 413 spiders in 10 bottle-traps re- corded at 17:00, 69 (15.3% ± 11.8%) had es- caped by 09:00 the following morning. A fur- ther 35 (9.1% ± 7.7%) escaped between 09: 00 and 17:00. The average loss over 24 h was 24.4% ± 16.6%. A significant linear regres- sion (adjusted R 2 = 63.6%, P = 0.004) be- tween initial numbers caught and numbers lost after 24 h indicated losses to be largely den- Table 2. — Daily totals of spiders caught for all traps with and without nets. Date 18/3 19/3 20/3 21/3 22/3 23/3 24/3 25/3 26/3 27/3 28/3 29/3 30/3 Nets 14 324 41 46 0 0 0 147 4 1 0 0 64 No nets 2 137 8 7 0 0 0 57 1 1 0 0 5 WOOLLEY ET AL. — NOVEL DESIGN OF CLIMBING STICK TRAP 311 sity independent. Mean rate of loss (± SE) from traps between 17:00 and 09:00 was 0.431 ± 0.141 spiders per hour and from 09: 00 to 17:00, 0.438 ± 0.148 spiders per hour. No significant difference in rate of loss was observed between night and day hours (F(U8) = 0.01, P = 0.976). DISCUSSION Climbing- sticks with bottle- traps are ex- tremely effective, cheap and easy to make and use. We estimate the cost of construction ma- terials to be less than $9 US per trap at current prices. Apart from the greater catch size, which, in total, was over eight times that of climbing sticks with adhesive, the bottle-traps also retain the advantage of easy replication and the ability to simultaneously sample dif- ferent habitats at large spatial and/or short temporal scales. The retention of live spiders means trapping agents such as adhesive or wa- ter and detergent are not required. Further- more, additional behavioral, ecological or ge- netic studies can be carried out on the trapped spiders if required. The addition of nets to climbing sticks with bottle traps increased catch size almost three fold. The trials reported here were conducted in short grass. However, in other trials con- ducted in taller crops, such as wheat, it was necessary to use 2.5 m climbing- sticks to raise the nets and bottle- traps above the crop in or- der to intercept airborne spiders. For compar- ative work sampling airborne spiders above crops of differing height, traps should be set at a constant height above the roughness length of the vegetation. Although losses from traps left operating for several consecutive days can be estimated, it is recommended that the traps are emptied daily, unless spiders are being collected only for laboratory studies. This avoids large amounts of silk accumulating inside the bot- tle-traps which makes separation of the spi- ders from the silk difficult and extraction much more time-consuming. Similarly, when large numbers of spiders were caught within a single day, we found traps were best emp- tied immediately after collection because of the quantity of silk produced if left overnight. We found traps were best removed in the evening after ballooning had finished. If traps cannot be changed until the morning, it should be carried out very early during summer months in order to prevent cross contamina- tion with the previous day's sample. If longer duration sampling is required and live spiders are not, a preserving fluid could be introduced into the bottom section of the bottle-trap. Spi- ders would fall into this, thereby reducing losses and minimizing any build-up of silk. A large variation in catch size was observed along the transect, particularly for the bottle- traps. This was possibly due to the greater trapping efficiency of the bottle-traps coupled with the undulating nature of the field, the greatest catch size being recorded at the high- est elevation. Linyphiids were by far the commonest spi- ders caught by the traps, being highest both in numbers and in occurrence throughout the year. Other spiders caught in lesser numbers belonged to the families Thomisidae and Ar- aneidae. Though immature thomisids were ob- served ballooning, adults of these families may have been present in traps as an accident of other behaviours such as rigging, locating shelter/feeding sites or web building. Care must therefore be taken before attributing dis- persal by ballooning to all spiders caught. The bottle-traps sometimes caught other in- sects including bush crickets, cantharid bee- tles, ephemeropterans, plecopterans, tipulids and various other dipterans. Some of this by- catch might prey on spiders but we did not see any evidence for this. Other potential loss- es are likely from predation among spiders but this was not quantified and is likely only if traps are left operating unchanged for longer periods. ACKNOWLEDGMENTS This work was funded through BBSKC grants D 14032, D20476 and D 14036. We would like to thank all the technical and farm staff at Seale-Hayne for their assistance in this work. LITERATURE CITED Blackwall, J. 1827. Observations and experiments, made with a view to ascertain the means by which the spiders that produce gossamer effect their aerial excursions. Transactions of the Lin- nean Society of London 15:449-459. Duffey, E. 1956. Aerial dispersal in a known spider population. Journal of Animal Ecology 25:85- 111. Greenstone, M.H. 1991. Aerial dispersal of arthro- pod natural enemies: altitudinal differences in 312 THE JOURNAL OF ARACHNOLOGY taxonomic distributions of dispersers. Pp. 104— 106. In Proceedings of the 10th Conference on Biometeorology and Aerobiology and the Spe- cial Session on Hydrometeorology. American Meteorological Society, Boston, Massachusetts. 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. Journal of Arachnology 15:163-170. Suter, R.B. 1999. An aerial lottery: the physics of ballooning in a chaotic atmosphere. Journal of Arachnology 27:281-293. Thomas, C.F.G., R.P. Blackshaw, L. Hutchings, C. Woolley, S. Goodacre, G.M. Hewitt, K. Ibrahim, S. Brooks & R. Harrington. 2003a. Modelling life-history/dispersal-strategy interactions to pre- dict and manage linyphiid spider diversity in ag- ricultural landscapes. Pp. 167-172. In Interna- tional Organization for Biological Control WPRS Bulletin, Volume 26. Landscape Manage- ment for Functional Biodiversity. (W.A.H. Ross- ing, H-M. PoeMing & G. Burgio, eds.). Univer- sity of Bologna, Italy. Thomas, C.F.G., P. Brain & P.C. Jepson. 2003b. Ae- rial activity of linyphiid spiders: modeling dis- persal distances from meteorology and behav- iour. Journal of Applied Ecology 40:912-927. Thomas, C.F.G. & P.C. Jepson. 1999. Differential aerial dispersal of linyphiid spiders from a grass and a cereal field. Journal of Arachnology 27: 294-300. Thorbek, R, CJ. Topping & K.D. Sunderland. 2002. Validation of a simple method for moni- toring aerial activity of spiders. Journal of Ar- achnology 30:57-64. 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., 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 5 June 2006, revised 27 Jan- uary 2007 . 2007. The Journal of Arachnology 35:313-317 A REVIEW OF THE WOLF SPIDER GENUS HIPPASELLA (ARANEAE, LYCOSIDAE, SOSIPPINAE) Eder S. S. Alvares1*2 and Aetoeto D* Brescovlt1: ^aboratorio de Artropodes, Instituto Butantan, Sao Paulo, Sao Paulo, Brazil 2Departamenfo de Zoologia, Institute de Biocieecias, Universidade de Sao Paulo, Sao Paulo, Brazil. E-mail: essalvares@yahoo.eom.br ABSTRACT. The monotypic genus Hippasella Mello-Leitao 1944 is revised, and its type-species H. nitida Mello-Leitao 1944 is considered a junior synonym of Tarentula guaquiensis Strand 1908, from Bolivia. Hippasella guaquiensis (Strand) comb. nov. is redescribed and the female genitalia are illustrated for the first time. This species now is recorded from Peru, Bolivia and Argentina. It appears to prefer vegetation near water. RESUMO. O genero monotipico Hippasella Mello-Leitao 1944 e revisado e sua espeeie-tipo H. nitida Mello-Leitao 1944 e considerada um sinonimo junior de Tarentula guaquiensis Strand 1908, da Bolivia. Hippasella guaquiensis (Strand) comb. nov. e redeserita e a genitalia da femea 6 ilustrada pela primeira vez. Esta especie e agora conhecida do Peru, Bolivia e da Argentina, onde parece preferir a vegetagao proxima a agua. Keywords* Neotropical, taxonomy, redescription The genus Hippasella was proposed by Me- llo-Leitao (1944) based on Hippasella nitida Mello-Leitao 1944, a species known only from a male specimen collected in La Plata, Argentina. The type-specimen of H. nitida is an adult male, but it is fragmented and in bad condition. Capocasale (1990) studied the type specimen of H, nitida and syeonymized Hip- pasella with Sosippus Simon 1888, based on the eye arrangement observed on the carapace fragments of the type specimen and on the absence of a palea and of a terminal apophysis in the male pedipalp. However, Sierwald (2000), refering to the figures of Capocasale (1990, figs. 12, 13), pointed out that the male pedipalp of H. nitida does not have a long finger-shaped apophysis (apophysis a in Sier- wald 2000: 136, fig. 7) shared by all Sosippus species. Moreover, the original size ratio of the eyes described by Mello-Leitao (1944: 343) for H. nitida does not match the ratio observed in the remaining Sosippus species. Based on these observations, Sierwald (2000) revalidated the genus Hippasella. For a long time the only known specimen of H. nitida was the fragmented type speci- men. Recently, after examining Lycosidae ma- terial housed in the Museu de Cieecias Na- turals, Porto Alegre, and in the Museo de Historia Natural San Marcos, Lima, we found some additional specimens of this species, in- cluding females. Moreover, after examining the types of Tarentula guaquiensis Strand 1908, housed in the Museum Wiesbaden, Wiesbaden, and kindly loaned by Dr. Michael Apel, we detected a new synonym for H ni- tida. In this paper, we present a more detailed redescription of this genus and the first illus- trations of the female genitalia. METHODS Descriptions and terminology follow Santos & Brescovit (2001). All measurements are in millimeters. The abbreviations used in the text are the following: ALE, anterior lateral eyes; AME, anterior mediae eyes; PLE, posterior lateral eyes; PME, posterior median eyes. The material examined are deposited in the following collections: IBSP, Instituto Butae- tan, Sao Paulo, Brazil; MHNSM, Museo de Historia Natural San Marcos, Lima, Peru; MCN, Museu de Ciencias Naturais, Fundagao Zoobotanica do Rio Grande do Sul, Porto Ale- gre, Brazil; MWNH, Museum Wiesbaden, Wiesbaden, Germany; SMF, Naturmuseum Senckeeberg, Frankfurt, Germany. 313 314 THE JOURNAL OF ARACHNOLOGY Figures 1-2. — Hippasella guaquiensis (Strand 1908), female from Huatajata, Bolivia: 1. Carapace, frontal view; 2. Carapace, lateral view. Scale bars = 2.00 mm. TAXONOMY Family Lycosidae Sundevall 1833 Subfamily Sosippinae Dondale 1986 Hippasella Mello-Leitao 1944 Hippasella Mello-Leitao 1944:342; Roewer 1955: 313; Roewer 1960:1002. Sosippus Simon: Capocasale 1990:140 (synonymy rejected by Sierwald 2000:138). Type species. — Hippasella nitida Mello- Leitao 1944, by original designation and mo- notypy. Diagnosis. — Males of Hippasella can be distinguished from males of other genera of Sosippinae by the tegular lobe in the pedipalp with a small and pointed lateral apophysis (Figs. 4, 5); a small and membranous median apophysis (Figs. 4, 5); and by a small lobe on the apical edge of the tegulum seen in the ven- tral view (Fig. 4). Females can be distin- guished by the large and flattened median sep- tum, and by spermathecae with a long and sigmoid curved stalk and with a small and not bilobate base (Fig. 8). Description. — Small lycosids (length 5.81- 8.40 mm). Carapace piriform, flattened dor- sally (Fig. 1). Eyes: anterior ocular row slight- ly procurve (Fig. 2); ocular quadrangle trap- ezoidal (Figs. 2, 3). Chelicerae strong; promargin with three teeth, the median bigger than lateral ones; retromargin with three big, equal and equidistant teeth. Sternum longer than wide, brownish, covered with grayish se- tae. Spinnerets: anterior lateral spinnerets con- ical, posterior median small, posterior lateral with distal segment not elongated. Male ped- ipalp: tibia cylindrical, 1.67 times longer than wide; cymbium piriform, without distal spines and with basal retrolateral edge dilated; te- gulum large, with spermatic ducts visible ven- trally, sigmoid; in ventral view (Fig. 4), the apical border bearing a small and transversal ly elongated projection at median region; retro- lateral margin of tegulum with a developed and ventrally pointed tegular lobe (Fig. 5). Median apophysis small, membranous and elongated, with distal end curved ventrally. Embolus with broad base, and distal area fi- liform and located below the base of median apophysis (Fig. 6). Terminal apophysis absent. Epigynum: median septum wider than long, flattened, with copulatory openings located anteriorly at its lateral borders (Fig. 7); epigy- nal plate and posterior half of median septum covered with small setae; atrium reduced, forming a narrow depression at lateral side of median septum. Internally (Fig. 8, 9) sper- mathecae hardly sclerotized, with small and not bilobate base, located dorsally at the stalk (Fig. 8); stalk elongated, curved in an “S” like shape. Head small, globular. Copulatory ducts large, flattened, and located medially to the base. Fertilization ducts small, membra- nous, located at posterior margin and curved dorsally (Fig. 9). Remarks.-— The placement of Hippasella in Sosippinae is based on the absence of a terminal apophysis and of a developed palea in the male pedipalp. Moreover, retrolaterally the base of the male pedipalpal cymbium of Hippasella is enlarged, as seen in species of Sosippus, Aglaoctenus Tullgren 1905 and Dia- pontia Keyserling 1876, and this character can be added to the diagnosis of Sosippinae. Hippasella guaquiensis (Strand 1908) comb. nov. Figs. 1-10 Tarentula guaquiensis Strand 1908:252. Lycosa guaquiensis (Strand): Petmnkevitch 1911: 559. Hippasella nitida Mello-Leitao 1944:343, fig. 32; ALVARES & BRESCOVIT— REVIEW OF HIPPASELLA 315 Figures 3-9. — Hippasella guaquiensis (Strand 1908), from Huatajata, Bolivia: 3. Female body, dorsal view; 4. Male pedipalp, ventral view; 5. Male pedipalp, retrolateral view; 6. Male pedipalp, antero-ventral view; 7. Female epigynum, ventral view; 8. Cleared female epigynum, ventral view; 9. Female epigynum, dorsal view. Abbreviations: BS = base of spermatheca, E = embolus, EP = epigynal plate, FD = fertil- ization duct, HS = head of spermatheca, MA — median apophysis, MS - median septum, S = stalk of spermatheca, TL = tegular lobe. Scale bar's: Figure 3 = 2.00 mm; Figures 4-9 = 0.25 mm. Roewer 1955:313; Sierwald 2000:138. New syn- onymy. Trochosa guaquiensis (Strand): Roewer 1955:301. Sosippus nitidus (Mello-Leitao): Capocasale 1990: 139, figs. 12, 13; Platnick 1993:508. Type specimens.— Tarentula guaquiensis : BOLIVIA: 1 male (without pedipalps), 1 fe- male syntype, Guaqui (not Peru) (16.5°S, 68.8°W), 1907, K. Seyd (MWNH #448); 1 pedipalp of the male syntype mounted on a microscope slide (SMF #13521-138). Hippasella nitida: ARGENTINA: male ho- lotype. La Plata, Buenos Aires (34.9°S, 57.9°W), M. Biraben (MLP #16035). Other material examined.— PERU: De partamento de Pasco: 2 $ , Pucayacu (10°39.2'S, 76°14.0'W), 8 May 2005, W. Pa- redes, D. Causso (MHNSM); Departamento de Cusco: 1 9 , Cusco (13°30.4'S, 71°59.0'W), June- July 1983, M. del Castillo (MHNSM); 2 9, Quebrada Jaluemoco Huayco (13°35.7' S, 71°58.5'W), 23 March 2005, W. Paredes (MHNSM); Departamento de Puno: 1 9, Arapa, border of Lake Titicaca (15°07.4'S, 316 THE JOURNAL OF ARACHNOLOGY Figure 10.— Hippasella guaquiensis (Strand 1908), known records. Peru: 1. Pucayacu; 2. Cusco; 3. Quebrada Jalunmoco Huayco; 4. Arapa; 5. Puno. Bolivia: 6. Huatajata; 7. Guaqui. Argentina: 8. La Plata. 70°06.3' W), November 1948, F. Blancas (MHNSM); 1 8 Puno (15°49.6'S, 70°01.3'W), November 1948, F. Blancas (MHNSM); BO- LIVIA: Departamento de La Paz: 1 8, 1 $, Huatajata, border of Lake Titicaca (16°06.0'S, 68°42.0'W), 8 August 1993, A.D. Brescovit & H. Hofer (MCN #23788); 1 9, same data (IBSP #66977). Diagnosis. — The same for the genus. Description. — Male (Huatajata, Bolivia MCN # 23788 ): Carapace brownish with gray setae and with 1 dorsal and 1 submarginal pale band, with dark radial bands; eyes sur- rounded by black areas. Chelicerae, sternum and labium brown; coxae and endites yellow- ish. Legs: dorsum of femora, patellae, and tib- iae yellowish with dark spots; metatarsi and tarsi light brown; venter of femora and tibiae yellowish with brown spots; patellae, metatar- si, and tarsi brown. Abdomen: dorsum with a dorsal longitudinal brown band delineated by black lines and bordered by a pale longitudi- nal band on each side; sides brown with 4-6 yellowish parallel longitudinal lines; venter yellowish with two median parallel black lines, extending from the epigastric furrow to the middle of abdomen. Spinnerets yellowish. Total length: 5.81; carapace length: 2.98; car- apace width: 2.13. Eye diameters: AME 0.14; ALE 0.12; PME 0.22; PLE 0.17. Eye inter- distances: AME- AME 0.08; AME- ALE 0.04; PME-PME 0.14; PME-PLE 0.16; PLE-PLE 0.59. Clypeus: 0.07 height. Leg I: femur 1.80/ patella 1.05/tibia 1.48/metatarsus 1.48/tarsus 0.92/total 6.73; leg II: 1.79/0.99/1.30/1.39/ 0.92/6.39; leg III: 1.76/0.82/1.22/1.61/0.89/ 6.30; leg IV: 2.13/1.07/1.74/2.40/1.17/8.51. Legs, spination: femur I: p0-0-2, dl-1-0, rO, II: p0-l-l, dl-1-0, rO; III: p0-l-l, dl-1-1, iO- 0-1; IV: pO (or pl-0-0, or pO-0-1), dl-1-1, rl- 0- 0; patellae I II: pO, rO; III: pi, rl; IV: pi, rl; tibia I: pl-1, dO, rO I v2- 0-2; II: pl-1, do- 0, r0-l v2-2-2 or vlr-2-2; III-IV: pl-1, d0-l, rl-1, v2-2-2; metatarsus I: p0-l-l, rO, v2-2-3; II: p0-l-l, r0-l-l, v2-2-3; III-IV: pl-1-1, rl- 1- 1, v2-2-3 (v3-2-3 in metatarsus IV). Tarsus and distal end of metatarsus of legs I and II weakly scopulate. Pedipalp (Figs. 4-6): see description in the genus. Female ( Huatajata , Bolivia MCN #23788): Coloration as in males except carapace with lateral pale bands almost marginal (Fig. 3), dorsum of abdomen with lateral pale bands darker, and venter of abdomen with a median longitudinal dark band bordered by a lateral pale band on each side. Total length: 8.18; car- apace length: 3.96; carapace width: 3.13. Eye diameters: AME 0.20; ALE 0.16; PME 0.29; PLE 0.25. Eye interdistances: AME-AME 0.10; AME-ALE 0.08; PME-PME 0.21; PME-PLE 0.26; PLE-PLE 0.89. Leg I: femur 2.55/patella 1.48/tibia 1. 79/metatarsus 1.84/ tarsus 1. 17/total 8.83; leg II: 2.37/1.45/1.66/ 1.79/1.12/8.39; leg III: 2.24/1.33/1.48/1.96/ I. 10/8.11; leg IV: 2.98/1.52/2.22/2.93/1.38/ II. 03. Leg spination as male, except: femur I: p0-0-l; IV: pO, rO; tibia I: pO, dO, rO, vO-lp-2; II: p0-0-l, dO, rO, v O-lr-2; III-IV: dO; metatarsus I: pO; II: pO-l-O or pO, vO. Epigy- num (Figs. 7-9): see generic description. Variation. — Nine females. Total length: 6.96-8.92; carapace length: 3.67-4.07; length of femur I: 2.16-2.55. Three males. Total length: 5.29-7.60; carapace length: 2.98-3.85; length of femur I: 1.80-2.55. Remarks. — The type locality of Tarentula guaquiensis was previously considered to be located in Peru. However, there is no locality called Guaqui in Peru, but there is a locality of the same name in Bolivia, near the edge of ALVARES & BRESCOVIT— REVIEW OF HIPPASELLA 317 Lake Titicaca, situated close to the border of Peru and Bolivia* Natural history.— Very little Is known about the habits of this species* As the spec- imens from Huatajata, Bolivia, and Arapa, Peru, were collected near the border of Lake Titicaca, we believe that this species lives in vegetation near water as some other South American Sosippinae. Distribution.— The only known records are from Peru, Bolivia and Argentina (Fig. 10). ACKNOWLEDGMENTS We would like to thank Adriana Davanzzo for revising the English version of the manu- script, Volker Frameeau, Petra Sierwald, Mark Harvey, and the editors of the Journal of Arachnology for comments and sugges- tions. We are grateful to the following curators supplying material for study, D. Silva (MHNSM), E.H. Buckup (MCN), M. Apel (MWNH), and P. Jager (SMF); we are partic- ularly grateful to M. Apel who kindly located and sent us the Strand types. Financial support was provided by FAPESP (n. 99/05446-8 and 02/11275-6). This study is part of the BIOTA/ FAPESP die Virtual Institute Program (www. biotasp.org.br) and of the Doctoral thesis in Zoology by the first author, developed in the Departamento de Zoologia, Institute de Bio- ciencias, Universidade de Sao Paulo, Sao Pau- lo, Brazil. LITERATURE CITED Capocasale, R.M. 1990. Las especies de la subfa- rm Ha Hippasinae de America del Sur (Araneae, Lycosidae). Journal of Arachnology 18:131-141. Mello-Leitao, C.E 1944. Aranas de la provincia de Buenos Aires. Revista del Museo La Plata (Nova Sene, Zoologia) 3:311-393. Petrankevitch, A. 1911. A synonymic index-cata- logue of spiders of North, Central and South America with all adjacent islands, Greenland, Bermuda, West Indies, Terra del Fuego, Gala- pagos, etc. Bulletin of the American Museum of Natural History 29:1-791. Platnick, NX 1993. Advances in Spider Taxonomy 1988-1991, with Synonymies and Transfer 1940-1988. New York Entomological Society, in association with the American Museum of Nat- ural History, New York. 846 pp. Roewer, C.E 1955. Katalog der Araneae von 1758 bis 1940, bzw. 1954. Bruxelles, 2:1-1751. Roewer, C.E 1960. Araneae Lycosaeformia II (Lycosidae) (Fortsetzung and Schluss). Explora- tion du Parc National de l’Upemba Mission G.E De Witte 55:519-1040. Santos, A.J. & A.D Brescovit. 2001. A revision of the South American spider genus Aglaoctenus Tullgren, 1905 (Araneae, Lycosidae, Sosippinae). Andrias 15:75-90. Sierwald, P. 2000. Description of the male of So- sippus placidus , with notes on the subfamily So- sippinae (Araneae, Lycosidae). Journal of Arach- nology 28:133-140. Strand, E. 1908. Exotisch araneologiscfaes.- — I. Amerikanische hauptsachlich in Peru, Bolivian und losemitetal in Californien gesammelte Spin- nen. — II. Spinnen aus Kamerun. — III. Ubersicht der bekanten Hysterocrates-Arten. — IV. Zur Kenntnis der Araneae rufipalpis (Luc). Jahrbii- cher des Nassauischen Verreins fur Naturkunde 61:223-295. Manuscript received 2 October 2006, revised 2 May 2007 . 2007. The Journal of Arachnology 35:318-324 A NEW SPECIES OF EUKOENENIA (PALPIGRADX, EUKOENENIIDAE) FROM MOROCCO Pablo Barranco and Jaime G. Mayoral: Departamento de Biologia Aplicada, Cite II-B, Universidad de Almeria, 04120 Almeria, Spain. E-mail: pbvega@ual.es ABSTRACT. The new species Eukoenenia maroccana is described from six specimens (two males, two females and two immatures) collected in Kef Aziza Cave, Morocco, and is distinguished from all other Eukoenenia species by the presence of thickened opisthosomal glandular setae in males on sternites IV- VI. The genitalia and chaetotaxy of both adult sexes show differences from other species of Eukoenenia and are discussed in this paper. RESUMEN. Se describe Eukoenenia maroccana a partir de seis ejemplares (dos machos, dos hembras y dos inmaduros) capturados en la gruta de Kef Aziza, Marruecos. Lo mas destacable y del todo singular de esta nueva especie es la particular presencia de setas glandulares esternales engrosadas del macho, la genitalia y resto de quetotaxia de ambos sexos. Keywords: Eukoenenia maroccana , taxonomy, North Africa, morphology Two species of Palpigradi have been pre- viously reported from Morocco (Harvey 2003). The endogenous species Eukoenenia mirahilis (Grassi & Calandruccio 1885) has been collected from a range of locations (Remy 1952a, 1956b, 1957), which were sum- marized by Harvey et al. (2006, fig. 2). A sec- ond endogenous species, Eukoenenia hanseni (Silvestri 1913), was recorded by Conde (1951) [see also Remy (1952a, 1957)]. Canals & Vinas (1960) captured a specimen within Kez Aziza (also named Kef Aziza) cave, which was identified as Koenenia sp., but this material has not been restudied and is now lost (Conde 1984, 1996). The study of a recent collection of several palpigrade specimens from Kef Aziza cave has revealed the pres- ence of a previously undescribed species. The specimens examined in this study are deposited in the National Museum of Natural Sciences, Madrid, Spain (MNCN) and the University of Almeria, Almeria, Spain (UAL). All measurements are expressed in microme- ters and were taken using an ocular micro- meter with a compound microscope. The fol- lowing abbreviations were utilized: L, total length of body (without flagellum which is lost in all specimens); B, length dorsal shield; P, pedipalpus; I and IV, legs I and IV; ti, tibia; btal, basitarsus 1; bta2, basitarsus 2; bta3, ba- sitarsus 3; bta4, basitarsus 4; tal, tarsus 1; ta2, tarsus 2; ta3, tarsus 3; a, width of basitarsus IV at level of seta r; er , distance between base of basitarsus IV and insertion of seta r; grt, length of tergal seta; gla , length of lateral seta; r, length of stiff seta; t/r, ratio between length of basitarsus IV and stiff seta length; tier , ratio between length of basitarsus IV and distance to insertion of stiff seta; gla! grt, ratio between lengths of lateral and tergal setae; B/bta, re- lation between lengths of prosomal shield and basitarsus IV; bta/ti, ratio between lengths of basitarsus IV and tibia IV. Setal nomenclature follows Conde (1974, 1971, 1984, 1988, 1989, 1992, 1993, 1994). TAXONOMY Family Eukoeneniidae Petrunkevitch 1955 Genus Eukoenenia Borner 1901 Koenenia Grassi & Calandruccio 1885:165 [junior primary homonym of Koenenia Beushausen 1884 (Mollusca: Bivalvia)]. Koenenia (. Eukoenenia ) Borner 1901:551. Type species. — Koenenia mirahilis Grassi & Calandruccio 1885, by monotypy. Remarks. — Eukoenenia includes 60 spe- cies and is the most diverse genus of Palpi- gradi (Harvey 2002, 2003; Mayoral & Bar- ranco 2002a). It is cosmopolitan with 25 species in Europe, 21 in Africa, 14 in Asia, 9 in America and 2 in Australia; some of them appear on different continents. Most species 318 BARRANCO & MAYORAL— NEW SPECIES OF EUKOENEN1A 319 Figures 1-4. — Eukoenenia maroccana new spe- cies: 1. Frontal organ, dorsal view; 2. Lateral organ, dorsal view; 3. Basitarsus 3-4 of leg I; 4. Basitarsus IV. Scale bars 100 fxm. Figures 5-8. — Eukoenenia maroccana new spe- cies: 5. Coxa I; 6. Coxa II; 7. Coxa III; 8. Coxa IV. Scale bar 100 |xm. Figures 9-10. — Eukoenenia maroccana new spe- cies: 9. Opisthosoma of male, ventral view; 10. Male genitalia, lateral view. Scale bars 100 fxm. are found in soil, but 27 are from caves. New species have been described recently (Mayoral & Barranco 2002b; Montano & Francke 2006). The distribution of some endogean spe- cies suggests human intervention (Savory 1974; Conde 1986; Harvey et al. 2006). The genus Eukoenenia is characterized by the absence of ventral sacs in opisthosomal sternites IV— VI, and segment IX is narrower than VIII, but larger than XI (Monniot 1966). Eukoenenia maroccana new species Figs. 1-13 Material examined. — MOROCCO: Errachidia: Holotype adult male, Kef Aziza cave, Tazougerte, Bouclenib [32°01'46"N, 03°47'17"W, 1040 m], July 1997, C. Hernando (MNCN 20.02/14845). Paratypes: MOROC- CO: Errachidia : 1 adult male, same locality and collector (UAL-Pp-022), 2 adult females, same locality and collector (UAL-Pp-023, MNCN 20.02/14846); 2 immature females, 320 THE JOURNAL OF ARACHNOLOGY Figure 1 1. — Eukoenenia maroccana new species: male genitalia. Scale bar 100 jxm. (type A), same locality and collector (UAL- Pp-024, MNCN 20.02/14847). Diagnosis. — This species differs from all other species of the genus by the combination of the presence of six lateral organs and the characteristic chaetotaxy and genitalia: males with 4 + 4 thickened secretory setae (a) and only one seta (s) on sternites IV-VI; different ventral opisthosomal chaetotaxy in both sexes; the presence of strongly developed fusules on very long dilated digitiform processes in male genitalia; and the shape of genitalia in fe- males. Description. — Male: Prosoma: frontal or- gan with 2 expanded granulate branches, blunt apically and each over 2.8 times longer than wide (Fig. 1). Lateral organ with 6 pointed blades, each 6 times longer than wide (Fig. 2). Dorsal shield with 10 + 10 short setae. Free segment of opisthosoma with 3 + 3 setae (th t2, t3), all of similar length. Chelicerae with 9 teeth on each side of chelicera and with 6 dor- sal and 1 ventral setae. Four deuto-tritosternal seta in linear arrangement. Chaetotaxy of cox- ae I— IV: 13, 13, 15 and 11 (Figs. 5-8). Basi- Figure 12. — Eukoenenia maroccana new species: opisthosoma of female, ventral view. Scale bar 100 pm. tarsus 3 of leg I slender, 4 times longer than broad, with 2 setae: stiff (r) and (grt) (Fig. 3), r shorter than the segment (120/95, tir = 1.26), inserted in proximal half and surpassing hind edge (32.5/112.5, s/er = 0.29). Basitar- sus IV with 7 setae (2 esd, 2 esp, gla, grt , and r) (Fig. 4), hta/ti 0.89. Stiff seta r 2.60 times shorter than tergal edge of article (t/r = 2.60) and inserted in its distal third ( tier = 1.66) very close to both grt. Both esd proximally inserted, followed by gla and grt , more or less at same level, all of them in proximal third. Opisthosoma: tergites II-VI with 3 + 3 dorsal setae, 2 pair of setae (r7, t3) between both slender seta (5), 2 + 2 on VII and VIII, only t seta present, and without s. Sternites II- III with 2 + 2 setae. Sternites IV-VI with 4 + 4 thickened setae in middle of opisthosoma (a}, a 2, a3, a4), outer ones (a4, 71) longer than others (ara3, 52.5). In addition, only one nor- mal seta (s) present on each side, as long as a4. Sternites with VII-VIII 2 + 2 setae (Fig. 9). Chaetotaxy of segments IX-XI each with 8 setae. Genitalia: 3 lobes present, with 44 setae; first lobe with deep medial indentation that separates two sides with a subtrapezoidal as- pect, 13 + 13 setae (including 2 + 2 fusules). Fusules inserted on a dilated digitiform base, external fusules longer, reaching past distal 321 BARRANCO & MAYORAL— NEW SPECIES OF EUKOENENIA Figure 13. — Eukoenenia maroccana new species: female genitalia. Scale bar 100 |xm. margin of second lobe. Outer 4 subapical se- tae extremely long, reaching past apex of third lobe. Second lobe with a large, single and pointed apical part, with 5 + 5 setae (a, b, c, c\ d). Third lobe of similar shape, but with 4 + 4 setae (x, y, z, w) (Figs. 10, 11). Female: generally similar to male but dif- fering as follows: Opisthosoma: 4 deuto-tritostemal seta in linear arrangement. Ventral setal formula: stemite III with 2 + 2 setae, sternites IV-VI with 3 + 3 setae (ah a2, a3) and a single 5 on each side; stemite VII with 2 + 2 setae (a7, a2) and one 5 on each side, sternite VIII with only 2 + 2, seta 5 absent (Fig. 12). Genitalia: first lobe with 11 + 11 setae in 322 THE JOURNAL OF ARACHNOLOGY Table 1. — Measurements (|im) of selected body parts of Eukoenenia maroccana. Body part Male 1, holotype Male 2 Female 1 Female 2 Immature L 1725 1604 1866 1115 936 B 584 605 697 605 — Pti 205 207 220 210 158 Pbtal 68 70 70 73 53 Pbta2 80 90 88 93 73 Ptal 43 50 45 43 38 Pta2 55 53 68 65 53 Pta3 80 90 98 95 55 Iti 227 — 247 238 190 Ibtal + 2 168 165 173 165 128 Ibta3 110 123 120 110 110 Ibta4 70 70 80 85 55 Ital 50 50 53 43 38 Ita2 53 55 58 58 48 Ita3 188 190 200 190 163 IVbta 208 212 212 215 168 IVti 233 225 242 235 185 IVtal 85 70 75 70 70 IVta2 125 120 125 133 105 a 23 — 25 33 23 er 125 130 145 143 103 grt 93 100 103 — 80 gla 90 88 103 95 78 r 80 85 80 83 65 t/r 2.60 2.49 2.65 2.59 2.58 tier 1.66 1.63 1.46 1.50 1.63 gla/grt 0.97 0.88 1.00 — 0.98 B/bta 2.81 2.85 3.10 2.81 — bta/ti 0.89 0.94 0.88 0.91 0.91 5 transverse rows, 4 sternal 2 + 2, 2 + 2, 2 + 2, 1 + 1 and distal 4 + 4, of which a7 and a2 of same length (25) which are shorter than a3 (42.5-48) and a4 (55-58). Second lobe with 3 + 3 setae (Fig. 7); 5 glandular orifices. Sper- mathecae elliptical (Fig. 13). Immature ( type A, immature female): Gen- erally similar to adults, but lateral organs with 4 blades; 3 deuto-tritosternal setae; and 8 che- liceral teeth. Basitarsus IV with 6 setae: 1 esp absent. Ventral and dorsal chaetotaxy as for female. One of the immatures was too dam- aged to be measured. Dimensions ( pm). — See Table 1. Etymology. — The specific name marocca- na refers to the country, Morocco, where the species was found. Remarks. — The prosoma of Eukoenenia maroccana bears six lateral organs, similar to that of E. depilata Remy 1960 (6), E. remyi Conde 1974 (4-6), E. spelaea (Peyerimhoff 1902) (5-6), E. hanseni (Silvestri 1913) (3- 6) and E. cf lyrifer Conde 1974 (6). Eukoe- nenia maroccana has 9 + 9 cheliceral teeth, while E. depilata , E. remyi , and E. spelaea have 8 + 8, E. hanseni has 7 + 7 and E. cf lyrifer also has 9 + 9. Eukoenenia depilata, E. remyi and E. cf lyrifer are only known from females. The ventral chaetotaxy of E. de- pilata and E. remyi has only one secretory seta (a) on sternites IV-VI, three s setae, which are slightly thickened, on each side in E. depilata (Remy 1960), only two s setae in E. remyi (Conde 1974). Although Conde (1974) did not describe the chaetotaxy of E. cf lyrifer, it is supposed that it is the same as that of E. lyrifer, which has two a and only one 5 setae. All of these species differ from the female of E. maroccana (with 1+3 + 3 + 1 and no thickened setae) and also the first lobe of the female genitalia is more rounded in these three species than in E. maroccana. The ventral chaetotaxy in males of E. ma- roccana have 4 + 4 a setae, which is similar BARRANCO & MAYORAL— NEW SPECIES OF EUKOENENIA 323 to that of E. spelaea, E. hanseni, and E. flo- renciae (which have 3 lateral organs), E. stri- natii (4 lateral organs), E. patrizzi (8-10 lat- eral organs), E. maros (4-5 lateral organs) and partially E. bouilloni (5 lateral organs), which has 5 + 5 secretory setae on sternite IV. All of these species have two 5 setae while E. ma- roccana has only one and none of them have the thickened secretory setae, which appears to be exclusively found in males of E. maroc - cana. These thickened setae are similar to those on the sternites V-VI in females of E. paulinae Conde 1994, which are present only on sternites IV and V (Conde 1994), and in E. angolensis (Remy 1956) and E. hesperia (Remy 1953) (Remy 1953, 1956a; Conde 1992, 1994). Other species with 5 lateral or- gans are E. pyrenaica and E. naxos, both with different sternal chaetotaxies, with two pair of a setae between a pair of s for the first one and 5 + 5 secretory setae in sternites V-VI between two s setae in E. naxos. The presence of fusules on dilated process- es is frequent in males of the genera Eukoe- nenia and Koeneniodes (Conde 1994). These processes can be only slightly developed, as in E. fossati Remy 1960, E. brignolii Conde 1979 and E. gasparoi Conde 1988 (Remy 1960; Conde 1979, 1988), moderately devel- oped, as in E. pauli Conde 1979, E. lawrencei Remy 1987, E. grassii (Hansen 1901) and E. janetscheki Conde 1993 (Conde 1979, 1981, 1993), or strongly developed, as in E. patrizii (Conde 1956) and E. maroccana. It is unusual to have different ventral opist- hosomal chaetotaxy in both sexes of the same species. This situation is also seen in E. ja- netscheki Conde 1993 where the female has an additional secretory seta on sternites IV- VI (Conde 1993). The mean value of B/bta in the four adults is 2.89, very far from the value of the caver- nicolous species, which should be lower than 2, but it is similar to the values for endoge- nous species (3-4) (Conde 1998). This situa- tion arises because the legs of E. maroccana are not elongated, although the specimens have the typical large size of cave dwelling species. ACKNOWLEDGMENTS We are most grateful to Carles Hernando for access to the material that he collected. LITERATURE CITED Berner, C. 1901. Zur ausseren Morphologic von Koenenia mirabilis Grassi. Zoologischer Anzei- ger 24:537-556. Canals, M. & R. Vinas. 1960. Atlas 68. Espeleoleg 8:325-332. Conde, B. 1974. Eukoenenia remyi n. sp., palpigra- de cavernicole d’ Herzegovina. Annales de Spe- leologie 29:53-56. Conde, B. 1979. Premiers Palpigrades du Gabon. Annales des Sciences Naturelles, Zoologie 1:57- 62. Conde, B. 1981. Donnees nouvelles sur Eukoenenia lawrencei Remy (Arachnides, Palpigrades). Re- vue Suisse de Zoologie 88:447-454. Conde, B. 1984. Les palpigrades: quelques aspects morpho-biologiques. Revue Arachnologique 5: 133-143. Conde, B. 1988. Nouveaux Palpigrades de Trieste, de Slovenie, de Malte, du Paraguay, de Thai'lan- de et de Borneo. Revue Suisse de Zoologie 95: 723-750. Conde, B. 1989. Palpigradi (Arachnida) de grottes d’Europe. Revue Suisse de Zoologie 96:823- 840. Conde, B. 1992. Palpigrades cavernicoles et endo- ges de Thailande et des Celebes (lere note). Re- vue Suisse de Zoologie 99:655-672. Conde, B. 1993. Description du male de deux es- peces de Palpigrades. Revue Suisse de Zoologie 100:279-287. Conde, B. 1994. Palpigrades cavernicoles et endo- ges de Thailande et de Celebes (2e note). Revue Suisse de Zoologie 101:233-263. Conde, B. 1996. Les Palpigrades, 1885-1995: ac- quisitions et lacunes. Revue Suisse de Zoologie, hors de serie 1:87-106. Conde, B. 1998. Palpigradida. Pp. 913-920. In En- cyclopaedia Biospeologica. Volume II. (C. Ju- berthie & V. Decu, eds.). Societe Internationale de Biospeologie, Moulis (C.N.R.S.), France and Bucarest (Academie Rumanie), Roumanei. Grassi, B. & S. Calandruccio 1885. Intorno ad un nuovo Aracnide Artrogastro (. Koenenia mirabils [sic]) che crediamo rappresentante d’un nuovo ordine (Microteliphonida). Naturalista Siciliano 4:127-133, 162-168. Harvey, M.S. 2002. The neglected cousins: what do we know about the smaller arachnid orders? Journal of Arachnology 30:357-372. Harvey, M.S. 2003. Catalogue of the Smaller Arachnid Orders of the World: Amblypygi, Uropygi, Schizomida, Palpigradi, Ricinulei and Solifugae. CSIRO Publishing, Melbourne, Aus- tralia. 400 pp. Harvey, M.S., F. Stahlavsky & PD. Theron. 2006. The distribution of Eukoenenia mirabilis (Palpi- gradi: Eukoeneniidae): a widespread tramp. Rec- 324 THE JOURNAL OF ARACHNOLOGY ords of the Western Australian Museum 23:199- 203. Mayoral, J.G. & R Barranco. 2002a. Palpfgrados: grandes desconocidos (Arachnida, Microthelyp- honida). Re vista Iberica de Aracnologfa 5:103- 110. Mayoral, J.G. & P. Barranco. 2002b. Descripcion de una nueva Eukoenenia Bomer, 1901 del sur- este Iberico (Arachnida, Palpigradi, Microthelyp- honida). Revista Iberica de Aracnologfa 6:129- 134. Monniot, F. 1966. Un palpigrade interstitiel: Lep- tokoenenia scurra, n. sp. Revue d’Ecologie et de Biologie du Sol 3:41-64. Montano, M.H. & O.F. Francke. 2006. Descripcion de una nueva especie de palpfgrado (Arachnida, Palpigradi) del genero Eukoenenia , colectado en la selva lacandona, en Mexico. Entomologfa Me- xicana 5:162-166. Remy, P. 1952a. Palpigrades du Maroc et de 1’ Al- geria occidental. Bulletin de la Societe des Sci- ences Naturelles de Maroc 30:159-163. Remy, P. 1952b. Palpigrades de Pile de la Reunion. Memories de lTnstitut Scientifique de Madagas- car, Serie A 7:69-79. Remy, P. 1953. Description d’un nouveau palpigra- de d’Afrique Occidentale Frangaise. Bulletin du Museum National d’Histoire Naturelle, Paris 25: 86-89. Remy, P. 1956a. Contribution a V etude de la micro- faune endogee de l’Afrique tropical: Palpigrades et Pauropodes. Revue de Zoologie et de Bota- nique Africaines 53:327-335. Remy, P. 1956b. Sur la microfaune endogee des jar- dins de Tanger (Symphyles, Pauropodes, Peni- cillates, Protoures, Palpigrades). Bulletin de la Societe des Sciences Naturelles de Maroc 35: 103-108. Remy, P. 1957. Nouvelles stations marocaines du Palpigrade Eukoenenia mirabilis Grassi. Bulletin de la Societe des Sciences Naturelles de Maroc 36:335-339. Remy, P. 1960. Palpigrades de Madagascar. II. Me- moires de lTnstitut Scientifique de Madagascar, Serie A 13:33-66. Savory, T. 1974. On the arachnid order Palpigradi. Journal of Arachnology 2:43-45. Manuscript received 3 February 2005, revised 7 March 2007. 2007. The Journal of Arachnology 35:325-333 REDESCRIPTION OF THE TYPE SPECIES OF CYNORTA (ARACHNIDA, OPILIONES, COSMETIDAE) Adriano Brilhante Kury: Departamento de Invertebrados, Museu Nacional, Quinta da Boa Vista, Sao Cristovao, 20.940-040, Rio de Janeiro - RJ - Brazil. E-mail: adrianok@gmail.com Osvaido Villarreal Manzanilla: Museo de Historia Natural La Salle, Fundacion La Salle de Ciencias Naturales, Apartado 1930, Caracas 1010- A, Venezuela. Cristiano Sampaio: Departamento de Invertebrados, Museu Nacional, Quinta da Boa Vista, Sao Cristovao, 20.940-040, Rio de Janeiro - RJ - Brazil. ABSTRACT. Cynorta conspersa (Perty 1833), the type species of Cynorta Koch 1839, is redescribed, based on abundant material from the lower Amazon basin, Brazil. A neotype is designated for this species and the species Cynorta mayi Mello-Leitao 1931 is herein considered a junior subjective synonym. Genital morphology of the species is described for the first time. An effort has been made to detect diagnostic characters for the genus Cynorta, which was used in many different senses in the past and includes a large number of unrelated Neotropical species. RESUMEN. Es redescrita Cynorta conspersa (Perty 1833), especie tipo del genero, con base en abundante material proveniente de la cuenca del bajo Amazonas de Brasil. Es designado un neotipo para esta especie y la especie Cynorta mayi Mello-Leitao 1931 es considerada como su sinonimo junior subjetivo. La morfologia genital es descrita por primera vez. Ha sido hecho un esfuerzo para detectar caracteres diag- nostics del genero Cynorta, el cual fue usado en el pasado con muchos significados diferentes, incluyendo un gran numero de especies neotropicales no relacionadas. Keywords: Neotropics, Brazil, taxonomy, new synonymy The family Cosmetidae Koch 1839, with more than 700 nominal species, is the second most diverse of Opiliones suborder Laniatores Thorell 1876 (Kury 2003). It is distributed in the Neotropics, with the greatest abundance in Central America and the Caribbean, stretching as far north as southern U.S.A. There are also many species in the Andean realm and the lowland Amazonian rainforest. The present state of cosmetid systematics is unsatisfactory, the genera being defined by a combination of area armature and tarsal counts. The high per- centage of monotypic genera in the faulty Roewerian system (e.g., Roewer 1923) has been counteracted by the recognition of large meaningless genera (Goodnight & Goodnight 1953), an equally ineffective approach to their taxonomy. Perty (1833) described the genus Cosmetus with many species of Cosmetidae from Brazil, among them Cosmetus conspersus Perty 1833 from “Brazil.” Koch (1839) was the first to narrow down the occurrence of the species from Para, creating the genus Cynorta to ac- commodate some of Perty ’s species, including C. conspersus , C. marginalis Banks 1909, C. posticata Banks 1909, C. dentipes F.O. Pic- kard-Cambridge 1904, C. geayi Roewer 1912, C. sulphurata Roewer 1912, C. sigillata Roe- wer 1912, C. flavoclathrata Simon 1879, C. vestita Roewer 1912, C. v -album Simon 1879, C. fraterna Banks 1909, C. albiornata Roe- wer 1912, C. scripta Simon 1879, C. calcar- basalis Roewer 1912, C. calcarapicalis Roe- wer 1912 and C. juncta (Gervais in Walckenaer 1844), all from localities in the Antilles, Brazil, Costa Rica, Cuba, Ecuador, French Guyana, Guatemala, Guyana, and Su- riname. Much later, Pickard-Cambridge (1904) designated Cosmetus conspersus as the type species of Cynorta. Mello-Leitao (1931) described Cynorta mayi from “Para,” but did not compare it with C. conspersa . The only literature records for Cosmetus conspersus, all 325 326 THE JOURNAL OF ARACHNOLOGY Figure 1. — Cynorta conspersa (Perty 1833), male neotype (MNRJ 6098) from Brazil, habitus: Dorsal view. Scale bar = 1 mm. in the Para state near the mouth of the Ama- zon River, are Cameta, at Rio Tocantins (S0rensen 1932), Belem and Tucurui (Kury 2003). Goodnight & Goodnight (1953), in an in- fluential paper, using the then dominant con- cept of considering only tarsal segmentation to define Opiliones genera, synonymized a great number of genera of Cosmetidae into only three: Vonones Simon 1879, Cynorta Koch 1839, and Paecilaema Koch 1839. Most of those synonymies were disclaimed by Kury (2003); but, even so, Cynorta is still the larg- est genus of Cosmetidae, with 154 species (22% of the diversity of the family) and is the type of the subfamily Cynortinae Mello-Lei- tao 1933, which is currently under the syn- onymy of Cosmetinae. The type material of C. conspersa is long lost (Roewer 1923), but we were able to ex- amine the four syntypes of C mayi in the Mu- seu Nacional, Universidade Federal do Rio de Janeiro, Brazil, which were compared with the descriptions and redescriptions in the litera- ture. As a result, we here designate a lectotype from the syntypes of C. mayi and a neotype KURY ET AL.— TYPE SPECIES OF CYNORTA 327 Figure 2. — Cynorta conspersa (Perty 1833), male neotype (MNRJ 6098) from Brazil, habitus: Lateral view. Scale bar = 1 mm. for C. conspersa , to stabilize the concept of the species and consider both nominal species to be synonyms. Abbreviations of depositories are: Museu Nacional, Universidade Federal do Rio de Ja- neiro, Brazil (MNRJ); Zoologische Staats- sammlung Miinchen, Germany (ZSMC). All measurements are in mm. Coordinates are in decimal degrees. SYSTEMATICS Family Cosmetidae Koch 1839 Genus Cynorta Koch 1839 Type species. — Cosmetus conspersus Perty 1833, by subsequent designation of Pickard- Cambridge (1904). Diagnosis. — Outline of the dorsal scutum of the beta type; chelicerae without strong sexual dimorphism, legs I— IV long, slender and unarmed, femur IV substraight; leg I with 6 to 7 tarsomeres; basitarsomeres of leg I of male much larger than distitarsomeres; tarsal claws of legs III-IV unpectinate; penis ventral plate subrectangular, as wide basally as dis- tally, with lateral borders parallel and distal border slightly concave and 3 + 4 lateral se- tae. Cynorta conspersa (Perty 1833) Figs. 1-10 Cosmetus conspersus Perty 1833:203. Cynorta conspersa (Perty): Koch 1839:21; Kury 2003:43. Poecilcema conspersum (Perty): Sprensen 1932: 336. Cynorta mayi Mello-Leitao 1931: 1 16, fig. 2; Mello- Leitao 1932:444, suppl, fig. 5. NEW SYNONY- MY. Type specimens. — Cosmetus conspersus : BRAZIL: male holotype, without further lo- cality data (ZSMC?), lost, not examined. BRAZIL: Para: male neotype (present des- ignation), Tucurui (3.6903°S, 49.7213°W), April 1981, A.C. Domingos (MNRJ 6098). BRAZIL: Para: Cynorta mayi : female lec- totype (present designation), 3 female paralec- totypes, without further locality data, E. May (MNRJ 1368). Other material examined. — BRAZIL: Para: 5 6, 11 9,1 juvenile, Belem (1.3904°S, 48.4490°W), 11 June 1974, W. Roth (MNRJ 6175); 12 8, 30 9, Belem, Clonal Garden (1.4300°S, 48.4564°W), insecticide blast in cacao tree, 14-15 December 1976, Hilton et al. (MNRJ 17641); 2 6, Belem, Utinga 328 THE JOURNAL OF ARACHNOLOGY Figure 3. — Ventral view of Cynorta conspersa (Perty 1833), male neotype (MNRJ 6098) from Brazil. Scale bars — 1 mm. (1.4558°S, 48. 5044° W), J.C. Carvalho (MNRJ 5050); 31 (3, 65 $, Tucurai (3.6903°S, 49.7213°W), April 1981, A.C. Domingos (MNRJ 4560); 6 6, 13 $, 4 juveniles, Tucurai (3.6903°S, 49.72 13°W), 20 April 1982, W. Roth (MNRJ 6318). Diagnosis*— Dorsal scutum pyriform with scutal areas obsolete, area I with one granule each side, III with a pair of spiniform large tubercles. Cheliceral sockets of carapace shah low, without laterofroetal projections. Chelic- eral bulla marginated laterally and posteriorly by a row of tubercles, ectal most developed. Basal tarsal segments I of the male slightly swollen. Femur and tibia IV much elongate, straight and unarmed. Tarsal counts: 6-7 (3), 12-16 (3), 8-9, 9-11. Tarsal claws III-IV un- pectinate. Penis: ventral plate with lateral bor- ders straight and parallel, distal border con- cave, uncleft; with fourth distal curved setae cylindrical and flattened distally and three me- dial lateral setae; glaes with a small ventro- distal projection, and dorsal process well de- veloped; stylus with veetro-distal mat covered with very small pointed granulations. Description of male neotype*— Measure- ments: dorsal scutum: carapace 1.45 long, 2.58 wide; abdominal scutum: 2.31 long, 3.26 wide; femora I— IV: 3.7, 9.3, 6.3, 10.1; tibiae I— IV: 2.6, 7.8, 3.2, 4.7. Body dorsal (Figs. I KURY ET AL.— TYPE SPECIES OF CYNORTA 329 Figures 4-5. — Cynorta conspersa (Perty 1833), male neotype (MNRJ 6098) from Brazil: 4. Left pedipalpus, mesal view; 5. Left pedipalpus ectal view. Scale bar = 1 mm. 2): dorsal scutum pyriform in dorsal view. Lateral border without granules or tubercles. Anterior margin with 2 sockets for the inser- tion of chelicerae, with 2 anterolateral projec- tions. Eye mound located anteriorly on the carapace, low, wide (about 30 % of total length [TL]), with 3 dorsal granules each side. With 4 mesotergal areas with dorsal minute setae; I with 1 granule each side, II and IV unarmed; III with 2 long spiniform projec- tions, straight with granules on its base. Pos- terior margin of dorsal scutum and free tergite I to III with a row of minute granules. Body ventral (Fig. 3): coxa I with a group of 6 an- terior tubercles, 1 medial row of 8-9 tuber- cles, 1 posterior with 6 granules and 4 distal tubercles; II with a group of 4 anterior gran- ules, 9 medial granules, 8-9 posterior granules and some small proximal granules between medial, posterior rows and 4 distal; III with a anterodistal row of 4 granules, a medial row of 6 granules, a posterior of 7 and 3 distal granules. Genital operculum with 2 lateropos- terior small projections, and few setae circu- larly distributed. Stigmatic area with setae ir- regularly distributed. Free sternites with a row of small setiferous granules each. Anal oper- culum with some small granules. Chelicera: basichelicerite with 1 ectal row of irregularly placed tubercles and 1 mesal row of tubercles (distal larger). Bulla slightly hypertelic, mov- able finger with 1 basal tooth and 6-7 small distal teeth. Pedipalpus (Figs. 4, 5): coxa with 1 distal tubercle and 1 small ventral granule. Trochanter with 2 ventral tubercles (mesal Figure 7. — Cynorta conspersa (Perty 1833), male neotype (MNRJ 6098) from Brazil: Left leg IV from trochanter to tibia, prolateral view. Scale bar = 1 mm. 330 THE JOURNAL OF ARACHNOLOGY Figures 8-10. — Cynorta conspersa (Perty 1833), male (MNRJ 4560) from Brazil, distal part of penis: 8. Lateral view; 9. Dorsal view; 10. Ventral view. Scale bars = 0.1 mm. larger). Femur compressed, with a row of ven- tral tubercles all along its length. Patella fo- liate, depressed, with small dorsal granules and 1 mesodistal tubercle. Tibia foliate, de- pressed with 3 dorsal rows of small granules. Tarsus short, with some dorsal granules and setae. Legs (Figs. 6, 7): coxa I with 2 anterior and 1 posterior tubercles; coxa II with 2 an- Figure 1 1 . — Lower Amazon basin, showing dis- tribution of Cynorta conspersa (black triangles) in the Brazilian State of Para. terior (dorsal larger) and 1 posterior fused with 1 of III; coxa III with 1 anterior fused with 1 of II; coxa IV with 3 dorsal tubercles, forming a common base. Trochanter I with 3 ventral tubercles; II with 2; III with 2 ventral and 2 retrolateral; IV with 2 retrolaterodistal granules and 1 prolaterodistal. Femora I— IV straight, with longitudinal row of very small granules and setae. Patella IV with 3 distal granules. Tibia IV slightly swollen distally. Metatarsus with 2 spiniform ventrodistal se- tae. Tarsi III and IV with 2 subparallel unpec- tinate claws, and tarsal process. Tarsal counts 7-6, 7-14, 9-9, 10-10. Distitarsi I-II with 3 articles each. Female: very similar to male. Small varia- tion in number of granules in rows of legs I-IV. Chelicerae slightly smaller. Variation: Range of tarsal counts and length femur-tibia I-IV are given in Tables 1 and 2 respectively. Remarks. — The type series of C. mayi con- sists of typical members of what we call C conspersa , and there are no differential char- acters in the description by Mello-Leitao KURY ET AL.— TYPE SPECIES OF CYNORTA 331 Table L — Tarsal counts of males and females of C. conspersa (MNR1 4560), Total number of individ- uals is given in parentheses. Number of tarsomeres Leg I (22) 6 10 <5, 10 $ 7 2 9 9 10 11 12 13 14 15 16 Leg II (32) Leg III (29) Leg IY (27) 5 (5, 8 2 8 63 8 2 1