ie Shi gee Tae ficeniberiitie tt Lan ik civ? Eeeenerat

se WN tai Tecan


pee erie

Sopot ease


Museum of

Comparative Zoology

The Systematics of Neotropical Orb-weaving Spiders in the Genus Metepeira (Araneae: Araneidae

VOLUME 157, NUMBER 1 8 JUNE 2001

(US ISSN 0027-4100)


Breviora 1952—


Memoirs 1865-1938

Jounsonia, Department of Mollusks, 1941-1974 OCCASIONAL PAPERS ON Mo.Luusks, 1945—

SPECIAL PUBLICATIONS. 1. Whittington, H. B., and W. D. I. Rolfe (eds.), 1963 Phylogeny and Evolution of Crustacea. 192 pp.

2. Turner, R. D., 1966. A Survey and illustrated Catalogue of the Tere- dinidea (Mollusca: Bivalvia). 265 pp.

3. Sprinkle, J., 1973. Morphology and Evolution of Blastozoan Echino- derms. 284 pp.

4. Eaton, R. J., 1974. A Flora of Concord from Thoreau’s Time to the Present Day. 236 pp.

5. Rhodin, A. G. J., and K. Miyata (eds.), 1983. Advances in Herpetology and Evolutionary Biology: Essays in Honor of Ernest E. Williams. 725 pp.

6. Angelo, R., 1990. Concord Area Trees and Shrubs. 118 pp.

Other Publications. Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprinted 1964.

Brues, C.T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. (Bulletin of the M. C. Z., Vol. 108.) Reprinted 1971.

Creighton, W. S., 1950. The Ants of North America. Reprinted 1966.

Lyman, C. P., and A. R. Dawe (eds.), 1960. Proceedings of the First In- ternational Symposium on Natural Mammalian Hibernation. (Bulletin of the M. C. Z., Vol. 124.)

Orinthological Gazetteers of the Neotropics (1975-). Peter’s Check-list of Birds of the World, vols. 1-16.

Proceedings of the New England Zoological Club 1899-1947. (Complete sets only.)

Proceedings of the Boston Society of Natural History. Price list and catalog of MCZ publications may be obtained from Publica-

tions Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.

This publication has been printed on acid-free permanent paper stock.

© The President and Fellows of Harvard College 2001.



CONTENTS FANS trac tees See LEER reed PA ae LY, 1 Imtrocductionie 2... 2e5 ao ae ek ee 2, Acknowledgments) #2 esses en ei ee neat ae 2} Materials and Methods __..- 3 Collections Examined __ 3 Locality Data Storage and Manipulation ___ 4 Examination and Illustration 4 Metepeira F. O. P.-Cambridge ~............--.-.-.-- 5 Key stoplemale Metepeind, yo sss. se seen ee 12 Key ston MialesMetenet ia te saws. se ee 2 lef IMICLenetnauoxts Group les saree mele. ewes Is) 1. Metepeira datona Chamberlin and Ivie 20 2. Metepeira desenderi Baert ___........ ail 3. Metepeira grandiosa grandiosa Chamberlin and Ivie 23 4. Metepeira grandiosa alpina Chamberlin and Ivie _ 24 Metenevrawicilasc: Group ese elo cian 26 5. Metepeira cajabamba New Species _.. 26 6. Metepeira glomerabilis (Keyserling) .. 28 7. Metepeira vigilax (Keyserling) —_... 30 8. Metepeira rectangula (Nicolet) 32 Metepeira labyrinthea Group ........-.-------------------- 33 9. Metepeira spinipes F. O. P.-Cambridge . 34 10. Metepeira lacandon New Species Oil Metepeira nigriventris Group —.........--.--------------- 38 11. Metepeira nigriventris (Taczanowski) 38 12. Metepeira tarapaca New Species _..... 40 13. Metepeira calamuchita New Species. 42 14. Metepeira galatheae (Thorell) 43 15. Metepeira karkii (Tullgren) 46 iMeteneinascompsa Croup =: at en eee 47 16. Metepeira compsa (Chamberlin) —__ 48 17. Metepeira roraima New Species —-...... D3 18. Metepeira gressa (Keyserling) 54 Metepeira incrassata Group —.2--- 56 19. Metepeira maya New Species __.......... 56 20. Metepeira inca New Species -............... 58

‘Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts 02138. Current address: Institute of Evolutionary and Ecological Sci- ences, Leiden University, 2311 GP Leiden, The Netherlands;

21. Metepeira gosoga Chamberlin and Ivie

Atcha Cadre Some iligteeetye tale SON ceed Siete teeta Te 59 22. Metepeira olmec New SESS 60. 23. Metepeira comanche Levi ____------- 62 24. Metepeira pimungan New Species ..... 62 25. Metepeira triangularis (Franganillo) _ 63 26. Metepeira arizonica Chamberlin and [vie tir ee aie Peasy sehen as 6 Ahead Bry) 66 27. Metepeira atascadero New Species ~.. 67 28. Metepeira incrassata F. O. P.- Cambridge yrs ene. pales ee ee 68 Metepeira ventura Group 20.0222 71 29. Metepeira ventura Chamberlin and vi See Renee el ET ae eee re: ee Bk 71 30. Metepeira revillagigedo New Species 73 31. Metepeira celestun New Species __... 74 32. Metepeira uncata F. O. P.-Cambridge _. 76 33. Metepeira crassipes Chamberlin and [hvalevnael Scie oe RM Den OLB ayiee elem ccd RELY a 34. Metepeira chilapae Chamberlin and vac jos epee ae Wee ae ee eee ee 78 Metepeira minima Group ~....--..--.------------------------- 80 35. Metepeira petatlan New Species —...... 80 36. Metepeira minima Gertsch _............-.- 82, 37. Metepeira pacifica New Species 84 38. Metepeira jamaicensis Archer 86 [Eaten abinen (Cue) 88 Ds Gl x es ee sae RE ree | Eu 91 ABSTRACT. Of the 39 species and three subspecies

of the orb-weaver genus Metepeira in the Americas, 36 species and two subspecies are known to occur outside of the U.S. and Canada. Yet, despite their conspicuous webs, diurnal foraging, and _ relatively common presence, the taxonomy of Metepeira is poorly understood, probably because the genitalia are small and difficult to distinguish. In fact, many names for species south of the U.S. were, at some time, in- correctly synonymized with the name Metepeira la- byrinthea. In this paper, 14 new species are named (Metepeira atascadero, M. cajabamba, M. calamuchi- ta, M. celestun, M. inca, M. lacandon, M. maya, M. olmec, M. pacifica, M. petatlan, M. pimungan, M. re- villagigedo, M. roraima, M. tarapaca); 11 new junior

Bull. Mus. Comp. Zool., 157(1): 1-92, June, 2001 il

2 Bulletin Museum of Comparative Zoology, Vol. 157, No. 1

synonyms are reported (M. acostai, M. bani, M. dom- inicana, M. grinnelli, M. latigyna, M. perezi, M. san- ta, M. salei, M. seditiosa, M. vaurieorum, M. virgi- nensis); five cases of erroneously synonymized names are reversed; 22 species and two subspecies are re- described (M. arizonica, M. triangularis, M. chilapae, M. comanche, M. compsa, M. crassipes, M. datona, M. desenderi, M. galatheae, M. glomerabilis, M. go- soga, M. grandiosa alpina, M. grandiosa grandiosa, M. gressa, M. incrassata, M. jamaicensis, M. karkii, M. minima, M. nigriventris, M. rectangula, M. spi- nipes, M. uncata, M. ventura, M. vigilax); and a key to all Metepeira species is presented. In addition, sev- eral ecological and life history observations are re- ported for various species.


The absence of a comprehensive revi- sion of Neotropical Metepeira has left the taxonomy of this group in shambles. Over the years, a fair number of species have been named, particularly by A. F. Archer, R. V. Chamberlin, and W. Ivie. However, these efforts have been sporadic and, for the most part, scant. For example, the de- scription of Metepeira dominicana (Ar- cher, 1965) provides little information oth- er than “form typical of Metepeira in all respects,” a few measurements, and two unrecognizable figures. Even when species are properly described they have far less taxonomic value when published alone, in the absence of a full comparative revision.

The poor understanding of Metepeira taxonomy has persisted despite great eco- logical and behavioral interest in this ge- nus. Indeed, many species are obligate or facultative social species and offer excel- lent models for investigating genetic and environmental factors that influence colo- ny formation (e.g., Uetz and Cangialosi, 1986; Uetz et al., 1987). The monumental work carried out over many years by G. W. Uetz has made great strides in our under- standing of gregarious social behavior in spiders and in risk-sensitive foraging the- ory in general (e.g., Uetz, 1996). Still, in the absence of solid taxonomic literature, behavioral ecologists have been forced to apply informal names to their study ani- mals (e.g., Metepeira “atascadero” in Uetz [1989] or Metepeira “Species A” in Viera

[1989]), but this practice can lead to trou- ble. In one case, the behavior of several different species was initially studied un- der the false assumption that they all be- longed to the same species (e.g., Uetz et al., 1982). Clearly, a strong taxonomic foundation is important for further biolog- ical work.

Ultimately, the relatively small, indis- tinct genitalia and the relatively homoge- neous abdominal patterns are to blame for the weakness in our knowledge of Mete- peira taxonomy. Many of these species are undoubtedly hard to distinguish, and this fact has surely intimidated arachnologists from taking on the painful task of revising the group. In the absence of good distin- guishing characteristics, the catalogs of Bonnet (1957) and Roewer (1942) synon- ymized the names of many Neotropical species with the name Metepeira labyrin- thea. Levis (1977) revision of Nearctic species observes that M. labyrinthea is ac- tually limited to the eastern United States. One task in this revision consists of reas- serting the names of species that were im- properly synonymized and clarifying the diagnostic characters that are needed to identify them.


This paper is part of my Ph.D. thesis for the Department of Organismic and Evo- lutionary Biology, Harvard University. Iam indebted to many people for their help, assistance, and encouragement in this pro- ject. I am especially thankful for the ded- ication and support of my advisors, Her- bert W. Levi and Edward O. Wilson. I am grateful that my colleagues in the Depart- ment of Invertebrate Zoology provided such a pleasant place to work: Edward Cutler, Ardis Johnston, Laura Leibensper- ger, Damhnait McHugh, Diana Sherry, Van Wallach, and Dee Woessner, among others.

Field collecting and new specimen ac- quisitions were made possible with the help of Gita Bodner, Fundacion Capacitar, Tim Coonan (CINP), Fred Coyle, Dawn

Fitzpatrick, Germania Jacome, Ant6nia Monteiro, Tila Perez, George Putnam, Linda Rayor, Grace Smith (NAWF), and George Uetz. I am particularly indebted to George Uetz for his assistance and corre- spondence.

I am thankful for the comments by those who read this paper—especially to the members on my thesis committee: H. W. Levi, N. E. Pierce, and E. O. Wilson. I am also indebted to Kathy Horton for her help in formatting and preparing the manuscript and to the Colles Fund for de- fraying the costs of publication. Curators at various institutions who lent me speci- mens are listed in the Materials and Meth- ods section. I cannot overstress the value of museum collections and expert curators, without which research in taxonomy would not be possible. Museum collections are the most important tools available for un-

derstanding biodiversity. MATERIALS AND METHODS

Collections Examined. The taxonomic revision was carried out on specimens bor- rowed from the following collections. The abbreviations correspond to those listed with each record after every species de- scription. I am grateful to the museums, curators, and staff that graciously loaned the material.

ADC A. Dean, Texas A&M University, College Station, Texas, United States

American Museum of Natural History, New York, United States; N. Platnick, L. Sorkin Natural History Museum, Lon- don, England; P. Hillyard California Academy of Sciences, San Francisco, California, Unit- ed States; C. Griswold

Carlos Valderrama A.; Bogota, Colombia

Florida State Collection of Ar- thropods, Gainesville, Florida, United States; G. B. Edwards Institut Royal des Sciences Na-






















turelles de Belgique, Brussels, Belgium; L. Baert

J. A. Kochalka, Ciudad Univer- sitaria, Paraguay

J. Carico, Lynchburg, Virginia, United States

J. Maes, Le6n, Nicaragua Museo Argentino de Ciencias Naturales, Buenos Aires, Argen- tina; E. A. Maury, C. L. Scioscia Museu de Ciéncias Naturais, Fundagao Zoobotanica do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; E. H. Buckup, M. A. L. Marques Museum of Comparative Zool- ogy, Harvard University, Cam- bridge, Massachusetts, United States: H. W. Levi

Museo Ecuatoriano de Ciencias Naturales, Quito, Ecuador; Ger- mania Eistévez Jacome

M. E. Galiano, Buenos Aires, Argentina

Maria Luisa Jiménez, Centro de Investigaciones Bioldgicas del Noroeste, La Paz, Mexico Museo de Universidad Nacional, La Plata, Argentina; R. F. Arro- zpide, C. Sutton

Museu Nacional, Rio de Janeiro, Brazil; A. Timotheo da Costa Museo Nacional de Historia Natural, Santo Domingo, Re- publica Dominicana; Félix Del Monte

Museo de Historia Natural, Universidad Nacional Mayor de San Marcos, Lima, Peru; D. Silva Museu de Zoologia, Universida- de de Sio Paulo, Sao Paulo, SP, Brazil; P. Vanzolini, J. L. Leme Museo Zoologico de “La Spe- cola” Universita di Firenze, Florence, Italy; S. Whitman Naturhistoriska Riksmuseet, Stockholm, Sweden; T. Krones- tedt

Polska Akademia Nauk, Warsza-

4 Bulletin Museum of Comparative Zoology, Vol. 157, No. 1

wa, Poland; J. Prészynski, A. Slo- jewska, W. B. Jedryczkowski

R. E. Leech, Edmonton, Alber- ta, Canada

Forschungsinstitut Sencken- berg, Frankfurt am Main, Ger- many; M. Grasshoff

Susan Riechert, Knoxville, Ten- nessee, United States

National Museum of Natural History, Smithsonian Institution, Washington, D.C., United States; J. Coddington, S. F. Larcher

Zoologisches Museum der Humboldt Universitat, Berlin, Germany; M. Moritz

Zoologisk Museum, Copenha- gen, Denmark; H. Enghoff, N. Scharff

Zoologische Staatssammlung, Munich, Germany








Locality Data Storage and Manipula- tion. Locality data from each collection vial were entered into a database designed us- ing Claris FileMaker Pro™. Geographic coordinates were added to locality data that lacked them using maps, USBGN gaz- etteers, and on-line databases (http://164. 214.2.59/ens/html/ and __ http://mapping. Occasionally locality information was illegible or unknown or one of several homonymous sites. In such cases a reasonable, educated guess was made and a “[?]” designation was append- ed to the locality. In some cases the itin- erary of a collector was reconstructed from other known records, and the ambiguous locality was assigned a coordinate halfway between the previous and following known collection sites. The locality database worked in concert with the mapping pro- gram Atlas Pro™ to generate thematic maps on the fly. These maps helped in the process of delimiting species and discov- ering cryptic species.

Elevation (in meters) was estimated for each locality that lacked this information. In some cases, elevation was estimated us-

ing contour maps, such as DMAAC ONC aeronautical maps; in most cases, elevation was estimated using NOAA data with an on-line database server (http://phylogeny. html).

The enhanced locality database was used to reveal ecological and life history traits. Seasonality of species was expressed by plotting a circular histogram showing the relative amount of collecting activi per 5-day interval (Figs. 300-337). While locality dates alone cannot control for the seasonal activity of human collectors, these data at least provide an estimate of spider seasonal abundance, if only approximate. Some sympatric species show incongruous seasonal abundance, which is at least some evidence that seasonality of spider collec- tors does not unduly overshadow the sea- sonality of the spiders themselves.

Examination and Illustration. Speci- mens were examined under 80% ethanol in a dish with light and dark sand grains for specimen support. Digital photographs of preserved specimens were taken through a Nikon SMZ-10 photomicro- scope using a Panasonic WV-CL320 CCD video camera, chosen for its high sensitiv- ity to light. Video images were captured using a QuickImage™24 digitizer and ed- ited on a Quadra 700 Macintosh® com- puter. The computer allows relatively in- expensive pictures to be printed rapidly on a 1,200 dpi Xanté™ Accel-a-Writer 8200 laser printer. Digital pictures were used to help sort out individuals to species, to cre- ate publishable pictures of gross dorsal and ventral markings, and to aid in the il- lustration of genitalia. As an aid in illustra- tion, the digital pictures functioned as a camera lucida because they assured accu- racy when drawing the proportions of gen- ital parts and sclerites. Usually a digital picture was laid over carbon paper and an outline of the genitalia was transferred to coquille board underneath. The illustra- tion continued on the coquille board using a Staedtler OmniChrom™ pencil and a drafting pen with India ink and then was scanned at 600 dpi on a LaCie Silverscan-

ner II™. The resulting digital image was edited in Adobe Photoshop™ and reduced in size to 1,200 dpi. The edited figures were finally arranged on plates using Can- vas®,

External genital structures were manip- ulated with pins to reveal hidden parts. The terminal division on the male palp is hinged, so it had to be pried open to see the embolus and embolic apophyses prop- erly. In females, mating plugs had to be removed from epigynal openings using pins. Sometimes the entire epigynum was partly cut from the body so as to see it from a posterior view.

Internal genital structures were studied by clearing them in clove oil and examin- ing them using an Olympus BH-2 com- pound microscope. Sketches were made directly on the computer in Canvas™ by aiming the camera lucida at the computer monitor. While internal genital structures helped in the process of delimiting spe- cies, they did not prove to be as useful as external genital structures in describing species; thus, these working sketches are not figured herein.

Measurements of the spiders were tak- en using a Leitz stereo dissecting micro- scope with a calibrated reticule. Sizes of leg articles, eyes, and carapace, were per- formed on one specimen of each sex, for each species. The respective localities of the candidate specimens were indicated in the descriptions. This study placed little reliance on spider leg measurements be- cause they are not usually very useful in spider taxonomy, and because Metepeira species are notorious for their variability in size (Levi, 1977; Piel, 1996).

All eye sizes were reported as a ratio of the posterior median eye diameters to the diameter of every other eye type. For ex- ample, in the case of “ratio of eye diame- ters: posterior medians and anterior me- dians 2.0, anterior laterals 0.5, posterior laterals 1.0,” the reader should interpret the anterior medians to be half the size of the posterior medians, and the anterior lat- erals to be twice the size of the posterior


medians. Eye separations were expressed in terms of their own diameters, or in terms of the anterior lateral eyes when be- tween eyes of different types. Oval eyes were measured as an average of the lon- gest and shortest lengths.

In parallel with the last revision of Me- tepeira (Levi, 1977), leg measurements were made on each article distal to the tro- chanter for the first leg and on the com- bined lengths of the patellae and tibiae for all remaining legs. Variation in total body size was provided as an average, minimum, and maximum of the total lengths from a number of mature specimens, usually cho- sen from a wide geographic spread.

Metepeira F. O. P.-Cambridge

Metepeira F. O. P.-Cambridge, 1903: 457. Type spe- cies by original designation M. spinipes F. O. P.- Cambridge 1903. The name is feminine.

Diagnostic Abstract. Web combines bar- rier or scaffolding structure surrounding a classic araneid orb with a retreat suspend- ed in air (Fig. 1). Like a raccoon with its facial colors reversed, the eye region is lighter than any other part of the carapace (Fig. 2). The venter has a wide median white line set on a black background that, with only some exception, extends anteri- orly on the sternum (Fig. 3). With one ex- ception, the total lengths of distal leg ar- ticles (metatarsus and tarsus) exceed that of the middle articles (patella and tibia). The median apophysis has two distinctive flagella (F in Fig. 5) and, in some species, an easily recognizable keel (K in Fig. 5). The dorsal abdominal markings (the foli- um) look like an inverted fleur-de-lis, al- lowing easy recognition of the genus in the field (Fig. 2).

Description and Diagnosis. For field ecologists, the most obvious and distinctive feature of Metepeira is the combination of orb and barrier web (Fig. 1). The barrier web forms scaffolding around an almost vertical orb and supports the spider's re- treat, which is thus suspended away from any substrate.

In contrast to most araneids, the cara-

6 Bulletin Museum of Comparative Zoology, Vol. 157, No. 1

pace of Metepeira is lightest in the eye re- gion. However, this distinctive feature varies within the genus: in the case of M. rectangula (Nicolet, 1849), the lighter re- gion takes up almost half the carapace (Fig. 65); in the case of Metepeira F. O. P.-Cambridge, 1903, the lighter region is usually limited to the anterior edges of the carapace (Fig. 2). White, downy hairs of- ten cover the carapace but are especially white and conspicuous on the lighter parts of the carapace outside the eye region. In some species, such as M. spinipes, these hairs make the carapace look gray or sil- very when the spider is alive, but dark brown when the spider is in ethanol.

The eyes of Metepeira are not particu- larly unusual. Eye separations relative to eye diameters increase with spider size: larger spiders tend to have relatively great- er eye separations. In either sex, the pos- terior median eyes are between 1.1 and 1.7 times the size of the anterior medians, and the separation between posterior me- dian eyes is between 0.4 and 0.7 of that between anterior median eyes. The sepa- ration between the anterior median eyes and the anterior lateral eyes is between 1 and 3.7 times the size of anterior median eyes in males and between one and five times the size in females. The diameter of the anterior median eyes exceeds the height of the clypeus.

The shape of the female abdomen rang- es from wider than long and rhomboid (M. datona, Fig. 12) to roundish (e.g., M. de- senderi, Fig. 20; M. rectangula, Fig. 65), to longer than wide and oval (e.g., M. inca, Fig. 169). The dorsal folium has a recog- nizable white fleur-de-lis pattern, usually on a dark background, its edges shaped by a wavy, zig-zag white outline (Fig. 2). The dorsum of live spiders is often more red- dish—a pigment that rapidly dissolves in alcohol.

Somewhat less common among other araneids is the median white line on the venter of the abdomen (Fig. 3), which is present (though shortened) even in the most darkly pigmented species. However,

unique among araneid genera is the com- bination of median white line on the ven- ter and median white line on a black or brown sternum. Some Metepeira species lack a complete white line on the sternum, but even those, such as M. datona, that usually have an entirely black sternum nonetheless show hints of white markings in some specimens. Characteristics found in the carapace, abdomen, and sternum of Metepeira are also found in Araneus koepckeorum Levi, but this last species lacks the white line on the venter.

With the exception of M. datona, and in some cases, M. desenderi, all Metepeira species have a combined metatarsus and tarsus that is longer than the combined pa- tella and tibia. This feature is unusual among araneids and is not found in Kaira O. P.-Cambridge or other likely relatives to Metepeira (Levi, 1977; Piel and Nutt, 1997).

In most species the leg articles are ringed, usually with brownish black on the distal and dorsal surfaces of each article, except for the patellae and tarsi which are usually entirely dark. In mainly tropical and high-altitude species, the coxae are mostly black (e.g., Fig. 75), but in desert/ mesquite species they appear yellowish white (e.g., Fig. 28).

Unlike many other araneids—and _ per- haps because of the small male size—the coxa on leg I of male Metepeira lacks the hook and corresponding groove typically found on femur II. In addition, males lack a tooth on the lateral side of the endite, and they lack a basal tooth on the palpal femur. The phylogenetic analysis of Scharff and Coddington (1997) incorrectly codes Metepeira as having a tooth on the endite. However, had the authors coded this character as absent, they would have decreased the length of their preferred tree because the nearest relatives hypoth- esized for Metepeira (Kaira, Zygiella, and Singa) also lack this tooth.

Macrosetae usually concentrate on arti- cles that contact other spiders during mat- ing or grappling. In contrast to most gen-


Barrier Web White Dorsal Light Eye Region / Fleur-de-Lis Pattern

Median White Line on Sternum

See \\\ , * if _ < We HUAI ROT SES Sa \ Median White it BS SVANS TS 7 D 3 Line on Black Venter Orb Web nigriventris incrassata minima labyrinthea compsa ventura

Eee " —— Flagellae

Dy Lae \Y7 ‘on Stalk (+) : DEA (-)

K on MA (-)

F, Teeth on Face of K (+) DEA (+) | North America

SC fEnlarged TA (+) : ; [) South America



Figure 1. Web of immature Metepeira grandiosa alpina from Chihuahua, Mexico.

Figure 2. Dorsum of adult female Metepeira crassipes.

Figure 3. Venter of adult female Metepeira tarapaca new species.

Figure 4. Hypothetical phylogenetic relationships among Metepeira species groups. Shaded branches indicate species groups that live in South America; open branches indicate species groups that live in North America, Central America, and the Caribbean. Abbreviations: DEA, distal embolic apophysis; K, keel of median apophysis; TA, terminal apophysis; (+), character state gain; (—), character state loss.

Figures 5, 6. Male palpus. 5, mesal view, Metepeira compsa. 6, ventral view of distal embolic division, Metepeira labyrinthea (Hentz).

Abbreviations: BEA, basal embolic apophysis; C, conductor; DEA, distal embolic apophysis; E, embolus; F, flagellum on median apophysis; K, keel of median apophysis; MA, median apophysis; TA, terminal apophysis; TD, terminal division.

Figures 7-13. Metepeira datona Chamberlin and Ivie (sp. 1; 17°53’N, 76°19’W). 7, male palpus, mesal. 8, epigynum, posterior. 9, epigynum, ventral. 10, male, dorsal. 11, male, ventral. 12, female, dorsal. 13, female, ventral.

Scale bar: dorsum and venter figures 1.0 mm.

8 Bulletin Museum of Comparative Zoology, Vol. 157, No. 1

era related to Araneus, Metepeira has con- centrated macrosetae on femur I instead of tibia II (Scharff and Coddington, 1997). Female Metepeira have between two and five macrosetae on the anterior side of the femur, and between zero and seven on the anteroventral side. Males typically have more setae than their conspecific females: four to nine on the anterior side and two to nine on the anteroventral side. Variation in the number of macrosetae appears to correlate with body size. In most species, the male palpal tibia and patella each have two strong macrosetae (Levi, 1977, fig. 8).

Compared with other araneid genera, Metepeira have rather small and similar genitalia, which on the one hand makes the genus easy to recognize, but on the other hand makes species tough to iden- tify. The small epigynum is fleshy, variable in shape, and weakly sclerotized. Unlike Araneus, Metepeira’s scape never has a pocket but always ends with a pointed tip (e.g., Fig. 31). The cleared epigynum— and in many cases the uncleared epigyn- um—reveals a pair of sclerotized spherical structures where the embolus is inserted, as well as ducts to pass semen to the larger, spherical seminal receptacles. In some species, these spherical structures are wide apart (e.g., Figs. 16, 17), in others they are tubular (e.g., Figs. 39, 40), but in many, they are closer together (e.g., Figs. 93, 94). Frequently the deeper, large seminal re- ceptacles can be seen through uncleared tissue (e.g., Figs. 201, 295).

The male palp is more distinctive. In particular, the median apophysis (MA in Fig. 5), while not always a good character for separating closely related species, is ex- cellent when it comes to identifying the genus. Two flagellae (F in Fig. 5) grace- fully curve off the base of the median apophysis, and in some species, a toothed or smooth keel (K in Fig. 5) extends in the opposite direction. This design is also seen in Kaira, Aculepeira, and Amazonepeira, but none of these have flagellae that ap- pear so integral to the base structure.

The terminal division on the Metepeira

palp is very similar in almost all species. | When this structure is pulled up, a basal embolic apophysis (also known as an em- bolar lamella) can be seen in the shape of a club or spatula (E in Figs. 5, 6). Some- times a distal embolic apophysis can be seen if it is not hidden from view by an overhanging terminal apophysis. When the terminal apophysis is large and_sclero- tized—which is the case in all but the Me- tepeira foxi species group—it has a rec- ognizable toothed notch, like the mouth on a wrench (Fig. 6). Virgin males have a cap on the embolus that remains in the epigynum after mating and presumably serves as a barrier to subsequent mating (Levi, 1977). The shape of the embolus cap varies from tiny (e.g., Fig. 178) to short but wide (e.g., Fig. 199) to large and winged (Fig. 46). Finally, the terminal di- vision lacks a stipes—a sclerite between the radix and the embolus that is frequent- ly found in other genera related to Ara- neus (Scharff and Coddington, 1997).

Natural History. All Metepeira species build a unique web that combines an orb with a barrier web (Levi, 1977; Lopez, 1993). As with Cyrtophora Simon or Me- cynogea Simon (Levi, 1997), the retreat of Metepeira hangs in the air, away from sub- strate, and is suspended by a scaffolding structure created by the barrier web (Fig. 1). The spider detects vibrations in the web and gains quick access to the hub us- ing a signal line that runs from the retreat to the center of the orb (Fig. 1). Tan col- ored egg sacs are strung together, usually above the retreat, and the most recently laid eggs are nearest to the spider. In some species the egg sacs and retreat are deco- rated with insect parts (e.g., M. spinipes); in other species they are carefully wrapped by leaves and woven together (e.g., M. da- tona). Unlike the webs of Cyrtophora and Mecynogea, the orb web of Metepeira is oriented vertically, and the number of radii and sticky spirals are more typical of other araneines.

In some species, such as M. pimungan (personal observation) and, to a lesser de-

gree, M. incrassata (G. Uetz, personal communication), juveniles and adults without eggs will live on webs lacking a suspended retreat. Instead, the spider sits on a white disk-shaped stabilimentum in the center of the hub. Of 110 M. pimun- gan specimens observed on San Miguel Is- land, about 40% occupied webs of this type. In two cases the disk stabilimenta were partly separated from the hub by barrier web lines and were further bent over to form a partly covered protective retreat for the spider. This observation makes it possible to imagine that the disk stabilimentum seen in M. pimungan re- sults from the fusion of the suspended re- treat with the hub.

When food supplies are plentiful, spi- ders of all kinds show an increased toler- ance for one another and an increased ten- dency to aggregate (e.g., Gillespie, 1987; Rypstra, 1986). The suspended retreats and barrier webs of Metepeira, Cyrtopho- ra, and Mecynogea may actually further fa- cilitate in the formation of aggregations by easing dependency on substrate availabili- ty and by providing a common support sys- tem (Burgess and Witt, 1976; Uetz, 1986). In any case, colony formation is known to occur in: all three genera (e.g., Rypstra, 1979), but especially in Metepeira. Small colonial aggregations of two to 10 individ- uals occur in M. datona (Spiller and Schoener, 1989), M. minima (personal ob- servation), M. glomerabilis (R. Baptista, personal communication), and M. atascad- ero new species (e.g., Uetz and Hodge, 1990). Medium-size colonies of 10 to 30 individuals occur in M. pimungan (person- al observation), M. gressa (Viera and Cos- ta, 1988), M. nigriventris (L. Rayor, per- sonal communication), M. tarapaca (VN. Roth, locality label), and M. spinipes (e.g., Uetz, 1988a). Large colonies, sometimes in the thousands of individuals, commonly occur in M. incrassata (e.g., Uetz and Hodge, 1990). Near rivers and in other lush habitats, M. tarapaca colonies can reach 200 individuals (M. Roy, personal communication). These cases of social be-


havior, broadly spread across seven differ- ent species groups, may mean that aggre- gation is a frequently lost and relatively old trait, or it may mean that species are prone to converge and evolve the same behavior independently.

Either way, much research has focused on elucidating the selective forces behind colonial behavior in Metepeira. In partic- ular, Uetz (1988a,b, 1996) has provided strong support for the hypothesis that Me- tepeira forage using a risk-sensitive strat- egy. He suggests that spiders in abundant habitats seek to minimize individual vari- ance in prey capture by aggregating in col- onies, whereas spiders in poor habitats seek to maximize variance by living soli- tarily—perhaps in a risky attempt to find areas of local prey maxima. The diversity of social tendencies among species is therefore commensurate of the diversity of ecological habitats that they inhabit.

Indeed, Metepeira species thrive in a wide array of habitats, though often they are quite harsh. These include wet, mon- tane cloud forests in Mexico and Panama (M. incrassata, M. olmec); tropical and wet agricultural areas (M. uncata, M. vigilax, M. glomerabilis, M. roraima); high-eleva- tion pine forests (M. lacandon, M. nigri- ventris, M. grandiosa alpina); Canadian bogs (M. grandiosa palustris); deciduous forests in the eastern U.S. (M. labyrin- thea); Caribbean coastal shrubbery (M. da- tona, M. minima, M. triangularis, M. ja- maicensis, M. maya, M. celestun); Mexican mesquite grasslands (M. atascadero, M. chilapae); Patagonian dunes and scrub (M galatheae) and pampas grass (M. karkii); dry Californian buckwheat and sage (M. crassipes, M. ventura, M. foxi, M. grandio- sa grandiosa); and arid and semiarid de- serts (M. arizonica, M. inca, M. ventura, M. crassipes). Although some species (e.g., M. galatheae, M. spinipes, M. compsa) cov- er vast geographic areas and live in many different habitats, many species are more biogeographically restricted. In fact, sev- eral species follow narrow ecological zones that decrease in elevation with distance

10 Bulletin Museum of Comparative Zoology, Vol. 157, No. 1

from the equator (e.g., M. rectangula, M. vigilax, M. cajabamba, Fig. 36; M. arizon- CG, See 2i3)):

Close cohabitation with different inter- and intrageneric species is not uncommon. Colonies of M. incrassata are known to contact webs of Nephila clavipes Linnaeus (Hodge and Uetz, 1996) and Mecynogea ocosingo and Gasteracantha cancriformis (personal observation). Often M. crassipes, M. ventura, M. foxi, and M. grandiosa grandiosa are collected together (Levi, 1977), as are M. minima and M. celestun (personal observation). Species that have been collected from identical localities, though not necessarily at the same time, include: M. chilapae and M. spinipes; M. chilapae and M. atascadero; M. karkii and M. galatheae; M. calamuchita new species, M. gressa, and M. galatheae; M. rectan- gula, M. calamuchita, and M. galatheae; M. compsa and M. gressa; M. vigilax and M. compsa; M. glomerabilis and M. vigilax; M. compsa and M. glomerabilis; M. comp- sa and M. nigriventris; M. conypsa and M. inca; M. datona and M. jamaicensis; and M. datona and M. triangularis.

Despite the wide biogeographic ranges of M. compsa (Puerto Rico and south to Argentina, Map 8) and M. datona (His- paniola and north to Florida, Map 1), they nonetheless come geographically close to one another but do not overlap. It is hard to imagine that the hurricanes that fre- quently pass through the Caribbean, as well as the homogeneous island environ- ments, would not gradually cause these two species distributions to overlap. Per- haps these abrupt, disjunct distributions are a rare example of competitive exclu- sion in Metepeira, which in other species is not thought to be an important factor (Wise, 1983).

Sphecid wasps are predators on Mete- peira. Locality labels indicated that M. pa- cifica has been found in the nests of Try- pargilum nitidum, T. tenoctitlan, and _T. bensoni. Jiménez and Tejas (1994) report that M. crassipes is the most frequent prey item in the nests of Trypargilum triden-

tatum. Colonial spiders, such as M. incras- sata, are especially vulnerable to wasps, other spiders, sarcophagid flies (e.g., Ar achnidomyia lindae, A. rayorae), and hummingbirds (Hieber and Uetz, 1990; Lopez, 1989; Rayor and Uetz, 1990).

Species Groups. Nearctic Metepeira were divided into two species groups: the M. labyrinthea group and the M. foxi group, based on the pattern on the ster- num and the shape of the median apoph- ysis (Levi, 1977). Baert (1987) questioned the taxonomic usefulness of the M. foxi species group (M. foxi, M. grandiosa, M. datona) because he found that M. desen- deri has both a keel on the median apoph- ysis and a white sternal line (Figs. 15, 21)—a combination that is incompatible by Levi's scheme. Nonetheless, the geni- talia of M. desenderi closely ally this spe- cies with the M. foxi group, so I am re- defining the M. foxi group based on purely genitalic characters. This is likely to be a basal, paraphyletic group (Fig. 4) (Piel and Nutt, 1997).

Seven additional species groups are dis- tal to the M. foxi. These remaining species are united by sharing a large terminal apophysis that is sclerotized and _ usually studded with teeth or denticles. The M. vigilax group (M. vigilax, M. cajabamba, M. glomerabilis, M. rectangula) are united by large emboli with long scooplike basal embolic apophyses (Fig. 60). Unlike the remaining species, the terminal apophysis in this group—albeit large—does not ac- tually overhang or hide the embolus. In addition to an overhanging terminal apophysis, the remaining taxa are also united by a distal embolic apophysis that either protrudes (Fig. 76), curves off (Fig. 185), or is secondarily lost (Fig. 264). The M. labyrinthea group (M. labyrinthea, M. lacandon, M. spinipes) share a toothless, smooth keel on the median apophysis (Figs. 67, 69).

The M. nigriventris group and the M. compsa group together share a median apophysis with teeth on the face of the keel (Figs. 92, 149). The M. incrassata

“al e datona S = desenderi


Map 3


\ = grandiosa alpina ¢

e grandiosa grandiosa ~~ Se

@ vigilax

| + rectangula


\e ns j My f Se Map 4 i 2

Maps 1, 2. Metepeira foxi species group. 1, M. datona, M. desenderi. 2, M. grandiosa grandiosa, M. grandiosa alpina. Maps 3, 4. Metepeira vigilax species group. 3, M. glomerabilis, M. cajabamba. 4, M. vigilax, M. rectangula.

group, the M. ventura group, and the M. minima group all lack a keel on the me- dian apophysis (Figs. 164, 222, 293). How- ever, both the M. compsa group and the M. incrassata group have epigyna with similar oval or round sclerotized rims (Figs. 151, 166), so it is likely that these are paraphyletic and consist of species leading up to a major North American (without a keel) and South American (with a keel) phylogenetic split (Fig. 4).

The South American branch includes the M. compsa group (M. compsa, M. ro- raima, M. gressa) and, more distally, the M. nigriventris group (M. nigriventris, M. tarapaca, M. calamuchita, M. galatheae, M. karkii). This latter group is united by a

distinctive and derived scape, which pro- jects out and down, creating a noticeable arch and overhang (Fig. 86).

The remaining species all lack a keel on the median apophysis, and with one ex- ception (M. inca), they live exclusively in North America (from the Caribbean and Panama to Nevada). The M. incrassata group (M. gosoga, M. maya, M. inca, M. comanche, M. olmec, M. atascadero, M. ar- izonica, M. incrassata, M. triangularis, M. pimungan) are very likely paraphyletic. The epigynum on each species seems au- tapomorphic and difficult to unite with any others. Some species (M. gosoga, M. maya, M. inca) have a pointed or projecting distal embolic apophysis (Fig. 171). Others have

1 Bulletin Museum of Comparative Zoology, Vol. 157, No. 1

a distal embolic apophysis that curves off sharply but does not project forward (Fig. 185). Finally, others have a distal embolic apophysis that curves off gently, almost to the point of hiding the existence of an apophysis (Fig. 206).

The M. minima group (M. jamaicensis, M. minima, M. pacifica, M. petatlan) and the M. ventura group (M. uncata, M. ven- tura, M. celestun, M. chilapae, M. revilla- gigedo new species, M. crassipes) are