(ISSN 0161-8202)

Journal of
ARACHNOLOGY

PUBLISHED BY THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 36

2008

NUMBER 3

THE JOURNAL OF ARACHNOLOG Y

EDITOR-IN-CHIEF: James E. Carrel, University of Missouri-Columbia

MANAGING EDITOR: Paula Cushing, Denver Museum of Nature & Science

SUBJECT EDITORS: Ecology — Suren Toft, University of Aarhus; Systematics — Mark Harvey, Western Austra-
lian Museum and Ingi Agnarsson, University of Akron; Behavior — Gail Stratton, University of Mississippi; Mor-
phology 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 Gillespie, University of California, Berkeley; Charles
Griswold, California Academy of Sciences; Marshal Hedin, San Diego State University; Herbert Levi, Harvard
University; Brent Opeil, Virginia Polytechnic Institute & State University; Norman Platnick, American Museum
of Natural History; Ann Rypstra, Miami University (Ohio); Paul Selden, University of Kansas; Matthias Schae-
fer, Universitet Goettingen (Germany); William Shear, Hampden-Sydney College; Petra Sierwald, Field Mu-
seum; 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. Subscriptions to The Journal of Arachnology and American Arachnology (the
newsletter), and annual meeting notices, are Included with membership in the Society. Regular, $55; Students, $30;
Institutional, $125 . Inquiries should be directed to the Membership Secretary (see below). Back Issues: Patricia
Miller, P.O. Box 5354, Northwest Mississippi Community College, Senatobia, Mississippi 38668 USA. Telephone:
(601) 562-3382. Undelivered Issues: Allen Press, Inc., 810 E. 10th Street, P.O. Box 368, Lawrence, Kansas 66044
USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: Paula Cushing (2007-2009), Zoology Department, Denver Museum of Nature & Science, Denver,
CO 80205-5798 USA.

PRESIDENT-ELECT: Rosemary Gillespie (2007-2009), Environmental Science, Policy & Management, Division of
Organisms and Environment, University of California, Berkeley, CA 94720-3114 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: Elizabeth Jakob (2007-2009), Jason Bond (2006-2008), Greta Binford (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: Clitaetra irenae female (Araneare, Nephilidae) from South Africa. Photo by Matjaz Kuntner.

Publication date: 7 November 2008

©This paper meets the requirements of ANSI/NISO Z39.48-1392 (Permanence of Paper).

2008. The Journal of Arachnology 36:487^-90

A redescription of Varacosa apothetica (Wallace) (Araneae, Lyeosldae)

Jamfn M. Dreyer: Department of Biology, Hope College Holland, Michigan 49423, USA. E-mail: dreyerj@hope.edu

Abstract. Lycosa apothetica Wallace 1947 is redescribed as a member of the genus Varacosa Chamberlin & Ivie 1942
based on genitalic morphology. The species is freshly illustrated, and information is provided as^
interesting characteristics.

Keywords: Lycosa, Trochosa, Florida spiders

Like many newly described members of the family
Lycosidae Sundevall 1833 from the mid-2Glh century, Lycosa
apothetica Wallace 1947 was originally described in the genus
Lycosa Latreille 1804. Roewer (1955) placed the species in
Varacosa Chamberlin & Ivie 1 942 when he elevated this genus
from its subgeneric status within Trochosa C.L. Koch 1847,
but his reasons for doing so were not made clear. Varacosa
apothetica was not part of Brady’s (1980) Trochosa aver a
group nor was it treated as a member of Varacosa by Jimenez
& Dondale (1988). Platnick (2008) therefore placed V
apothetica within Trochosa along with the rest of Roewer’s
Varacosa not included by Jimenez & Dondale (1988). Based
on Wallace’s illustrations (1947), it seemed likely that the
species belonged to Varacosa. Wallace (1947) examined a total
of 57 specimens from the Southeastern United States (8<?c?,
492?) in his original description. Having examined most of the
type material, I here support Roewer’s (1955) combination
Varacosa apothetica by showing that the species bears a
prominent synapomorphy of Varacosa ; i.e., the conspicuous
anterior curvature of the transverse piece of the epigynum.

METHODS

Descriptions and drawings are based on specimens viewed
in 70-75% ethanol under direct illumination. The epigynum
was removed and cleared with clove oil, and the spermathecae
were illustrated within this liquid. For clarity, palpal setae
were omitted. Measurements reported here are those found in
Wallace’s (1947) original description of the species. Figure 8
reflects the collection localities of the specimens examined here
as well as those included in Wallace (1947) that could not be
located.

Abbreviations. — Male palpal structures: palea region (pr),
tegular lobe (tl), embolus (emb), terminal apophysis (ta),
median apophysis (ma), tegulum (tg). Female genitalic
structures: head of spermathecae (hs), stalk of spermathecae
(ss), fertilization ducts (fd). Body Dimensions: carapace width
(CW), carapace length (CL), Eyes: posterior ocular quadran-
gle width (POQW), posterior ocular quadrangle length
(POQL), posterior median eye width (PMEW), posterior
lateral eye width (PLEW), clypeus height (CH). Palpal
segments: palpal femur (PF), palpal patella (PP), palpal tibia
(PT), palpal cymbium 3 (PC), palpal tarsus & claw 9 (PTC).
Legs: femur (FI— 4), patella (P 1—4), tibia (Til-4), metatarsus
(Ml-4), tarsus (Tl-4). Collections: American Museum of
Natural History, New York (AMNH); Florida State Collec-
tion of Arthropods, Gainesville (FSCA).

TAXONOMY

Family Lycosidae Sundevall 1833
Genus Varacosa Chamberlin & Ivie 1942

Trochosa C.L. Koch 1848:95, in part. Brady 1980:168, in
part. Platnick 2008, in part.

Trochosa ( Varacosa ) Chamberlin and Ivie 1942:36.

Varacosa: Roewer 1955:304. (Raised to generic status).
Jimenez & Dondale 1988:172.

Type species. — Trochosa avara Keyserling 1877, by original
designation.

Varacosa apothetica (Wallace 1947)

Figures 1-7

Lycosa apothetica Wallace 1947:33, figs. 1, 2.

Varacosa apothetica Roewer 1955:306.

Type material examined. — Varacosa apothetica: allotype c?
USA: Florida: Alachua County: 29°39'N, 082° 1 9' W, 26
October 1937, H.K. Wallace (AMNH); Lycosa apothetica:
holotype 2 USA: Florida: Alachua County: Station 7B,
29°39'N, 082°19'W, 26 October 1937, H.K. Wallace
(AMNH). Paratypes: U.S.A.: Florida: 29 Alachua County,
Station 1, 29°39'N, Q82°19'W, 19 January 1937, H.K. Wallace
(FSCA); 1? same location, 29°39'N, 082°19'W, 30 January
1937 (FSCA); 29 Alachua County, Station IB, 29°39'N,
082°19'W, 30 January 1937, H.K. Wallace (FSCA); 59
Alachua County, Station 1 vicinity, 29°39'N, 082°19'W, 30
January 1937 (FSCA); 29 same location, 29°39'N, 082°19'W,
19 April 1937 (FSCA); 1? same location, 29°39'N, 082°19'W,
22 January 1939, H.K. Wallace (FSCA); 29 Alachua County,
Station 2 vicinity, 29°39'N, 082°19'W, 30 January 1937, H.K.
Wallace (FSCA); 22 Alachua County: Station 7A, 29°39'N,
082°19'W, 7 February 1937 (FSCA); 29 Alachua County,
Station 7B, 29°39'N, 082°19'W, 20 January 1937 (FSCA);
32 same location, 29°39'N, 082°19'W, 8 February 1937
(FSCA); I<?2? same location, 29°39'N, 082°19'W, 26 May
1937 (FSCA); 12 Alachua County: Gainesville, 29°39'N,
082°19'W, 18 March 1938, W.J. Gertsch (FSCA); 1<J Alachua
County: pine fiatwoods, 29°39'N, 082°19'W, 2 July 1938, C.
Benton (FSCA); 1? Alachua County, 29°39'N, 082°19'W, 11
April 1933, H.K. Wallace (FSCA); 42 Alachua County,
29°39'N, 082°19'W, 15 February 1938, H.K. Wallace (FSCA);
39 Leon County, 30°26'N, 084° 1 6' W, 16 March 1936, H.K.
Wallace (FSCA). Georgia: 19 Turner County, 31°42'N,
083°39'W, 6 May 1937, H.K. Wallace (FSCA); No Locality:

487

488

THE JOURNAL OF ARACHNOLOGY

Figures 1, 2. — Dorsal view of carapace and abdomen: 1. Varacosa
apothetica, allotype male from Florida, USA; 2. Lycosa apothetica
holotype female from Florida, USA. Scale = 1 mm

8?, H.K. Wallace (FSCA). Mississippi : 1<329 Hancock County,
42 mi East of New Orleans, 30°18'N, 089°20'W, 15 July 1939,
H.K. Wallace (FSCA).

Other material examined. — USA: Alabama : 2$ Baldwin
County: Foley, 30°24'N, 087°41'W, 25 June 1912 (FSCA).
Florida: 19 Alachua County, 29°39'N, 082°19'W, 26 January
1958, H.V. Weems Jr. (FSCA).

Etymology. — Wallace did not comment on etymology.
However, apothetica is presumably derived from a Greek
word meaning storehouse, and could refer to the Devil’s Mill
Hopper near the type collection locality (Dondale, pers.
comm. 2007).

Diagnosis. — Males of V apothetica are distinguished from
other members of this genus by the large palea, which lacks
sclerotization, a transparent terminal apophysis superior to
the tegular lobe, and the relatively small tegular apophysis
(Figs. 3, 4). Females are separated from their congeners by the
anteriorly widened curved portion of the transverse piece of
the median septum, each being about one-third the width of
the entire epigynum (Figs. 5, 7).

Description. -Male: Chelicerae: light tan; without strong
boss; three promarginal and retromarginal teeth. Carapace
(Fig. 1): light yellow background with medium brown
markings; median light stripe extending from between eyes
to rear of carapace containing two darkened areas posterior to
PLEs; submarginal light bands; wavy dark margins; few dark
brown radial lines running from median light area to
submarginal light area; darkened region surrounding ALEs
and PER; AER slightly procurved; eyes and anterior portion

Figures 5-7. — 5. Ventral aspect of epigynum, Lycosa apothetica ,
holotype female from Alachua County, Florida, USA. 6,7. Varacosa
apothetica paratype female from Alachua County, Florida, USA: 6.
Internal genitalia; 7. Epigynum of same, hs = head of spermathecae;
ss = stalk of spermathecae; fd = fertilization ducts. Scale = 0.5 mm

Pr

tl

emb

ta

tga

tg

Figures 3, 4. — Varacosa apothetica allotype male from Alachua
County, Florida, USA: 3. Ventral aspect of palp; 4. Retrolateral aspect of
same, pr = palea region; tl = tegular lobe; emb = embolus; ta = terminal
apophysis; tga = tegular apophysis; tg = tegulum. Scale = 1 mm

of carapace lightly iridescent. Dorsum of abdomen: light
yellow background, uniformly mottled medium grey; heart
mark may be present, if so not strongly outlined; weak
chevrons if any; darker anteriorly. Legs: femora annulate;
ventral surface of femur iridescent, perhaps even the entire leg;
leg formula IV:I:IEin. Endites and labium: light yellow
overall, darker proximally; labium nearly square. Sternum:
yellow; bare but for disparate long setae. Venter: yellow; low
numbers of small dark spots near the margins. Pedipalpus
(Figs. 3, 4): large, wrinkled palea; tegular apophysis with
strong scleritization on dorsal tip; slightly curved embolus
lying in distally elongated tegular lobe; transparent terminal

5

DREYER — VARACOSA APOTHETICA

489

Figure 8. — Map of Varacosa apothetica collection locality records.
Southeastern United States. Scale = 150 mi/240 km

apophysis protruding from palea over the conductor and
embolus when viewed dorsally; tegulum devoid of major
topography or scleritization; no macrosetae on tip.

Female: similar to male. Carapace (Fig. 2); Epigynum
(Figs. 5, 7): Thin median septum; thin transverse piece, except
for heavily thickened anterior lobes, each nearly one-third the
width of the epigynum; deep excavations along margin of
median septum and transverse piece; median septum and
transverse piece sclerotized; hoods triangular; whole of
epigynum nearly circular. Spermathecae (Fig. 6): Large
spermathecae extending anteriorally from lateral position;
stalk of spermathecae angled toward top of median septum
before bending > 90° and widening into head of spermathe-
cae; fertilization ducts appear suspended above darkened
structure posterior to spermathecae when viewed from within.

Measurements. — Wallace's original measurements ( 1947) for
both the male and female are again reported here in Table 1.

Distribution and habitat preferences. — Wallace’s records
(1947) indicate that this species is found only in the
southeastern USA, from Florida and Georgia west to
Mississippi (Figure 8). Most specimens have been collected
in Gainesville, Florida. Wallace (1947) reports that “Males
have been collected only in October, November, December,
and February while females have not been taken after May
until October.” It is “secretive,” “usually stays close by, or in,
the mouth of it's [sic] retreat” and is “usually found in moist
situations in pine flatwoods (pond margins, cypress bay
margins, etc.), but may be found occasionally in other
situation [sic]”( Wallace 1947).

Remarks. — Wallace (1947) lists 57 paratypes in his original
description. I examined the majority of those, and three
additional specimens collected later. This species bears a
synapomorphy with Varacosa : the conspicuous anterior
curvatures of the transverse piece. 1 therefore support
Roewer’s combination: Varacosa apothetica (Wallace 1947),
contra Platnick (2008). In V. apothetica these structures are
much wider than those of other Varacosa. Of its congeners, the
V. apothetica epigynum most closely resemble V. gosiuta
(Chamberlin 1908) and V. shenandoa (Chamberlin & Ivie
1942) (Brady 1980). The palp of the male differs from most

Table 1. — Features of V. apothetica (Wallace), taken from Wallace
(1947). All measurements in millimeters. See text for abbreviations.

Dimension

V apothetica <5

V. apothetica ?

CW

2.5

2.9

CL

3.5

4.0

POQW

1.0

1.4

POQL

0.8

1.0

PMEW

0.4

0.4

PLEW

0.3

0.3

CH

0.6

0.6

PF

1.3

1.4

PP

0.6

0.7

PT

0.6

0.8

PC/PTC

1.0

1.2

FI

2.7

3.9

PI

1.4

1.5

T1

2.3

2.2

Ml

2.3

2.0

T1

1.5

1.5

Total 1

10.2

11.1

F2

2.4

2.7

P2

1.3

1.5

T2

2.0

2.0

M2

2.1

2.0

T2

1.5

1.5

Total 2

9.3

9.7

F3

2.4

2.6

P3

1.2

1.3

T3

1.7

1.8

M3

2.3

2.2

T3

1.3

1.5

Total 3

8.9

9.4

F4

3.1

3.3

P4

1.2

1.6

T4

2.5

2.8

M4

3.3

3.7

T4

1.8

1.9

Total 4

11.9

13.3

other Varacosa, featuring a large palea and relatively small
median apophysis similar to that of V. hoffmannae Jimenez &
Dondale 1988.

One notable feature of this species is the iridescent quality
noted on the body of the males. Male specimens exhibited
varying levels of iridescence over their bodies, but each was
found to have femora that bore this quality. It is not clear if
this is an artifact of the long term preservation of these
specimens or a true characteristic of the species.

ACKNOWLEDGMENTS

I wish to thank the following individuals for their
contribution to this work Allen R. Brady and Charles D.
Dondale for their identification of this project and helpful
encouragement; Norman I. Platnick (AMNH) and G.B.
Edwards (FSCA) for the loan of specimens; Lou Sorkin
(AMNH) for his assistance with material; C.D. Dondale and
Thomas L. Bultman for reading an early version of the
manuscript; Dr. Volker Framenau, an anonymous reviewer,
and Dr. Ingi Agnarsson for their helpful critiques during the
review process; Sarah C. Crews for her suggestions in relation
to the illustrations; and the Hope College Department of

490

THE JOURNAL OF ARACHNOLOGY

Biology, particularly the Chair T.L. Buitman, for support of
this research.

LITERATURE CITED

Brady, A.R. 1980. Nearctic species of the wolf spider genus Trochosa
(Araneae: Lycosidae). Psyche 86:167-212.

Chamberlin, R.V. & W. Ivie. 1942. A hundred new species of
American spiders. Bulletin of the University of Utah 32:1-117.
Jimenez, M.L. & C.D. Dondale. 1988. Descripcion de una nueva
especie del genero Varacosa de Mexico. Journal of Arachnology
15:171-175.

Platnick, NT. 2008. The World Spider Catalog, Version 8.0. The
American Museum of Natural History, New York. Online at
http://research.amnh.org/entomology/spiders/catalog/. Accessed 1 6
January 2008.

Roewer, C.F. 1955. Katalog der Araneen von 1758 bis 1940, bzw.
1954. Volume 2. Institut Royal des Sciences Naturelles de Belgique,
Bruxelles. 923 pp.

Wallace, H.K. 1947. A new wolf spider from Florida, with notes on
other species. Florida Entomologist 30:33-42.

Manuscript received 25 November 2007, revised 10 February 2008.

2008. The Journal of Arachnology 36:491-501

Description of Zabius gaucho (Scorpiones, Buthidae), a new species from southern Brazil,

with an update about the generic diagnosis

Luis E. Acosta: CONICET - Catedra de Diversidad Animal I, Facultad de Ciencias Exactas, Fisicas y Naturales,
Universidad Nacional de Cordoba, Avenida Velez Sarsfield 299, X5000JJC Cordoba, Argentina. E-mail:
lacosta@com.uncor.edu

Denise M. Candido: Laboratorio de Artropodes, Instituto Butantan, AvenidaVital Brasil, 1500, Butanta, 05503-900,
Sao Paulo, Brazil

Erica H. Buckup: Museu de Ciencias Naturais, Fundaqao Zoobotanica, Rua Dr. Salvador Franqa, 1427, 90690-000,
Porto Alegre, Brazil

Antonio D. Brescovit: Laboratorio de Artropodes, Instituto Butantan, AvenidaVital Brasil, 1500, Butanta, 05503-900,
Sao Paulo, Brazil. E-mail: adbresc@terra.com.br

Abstract. This paper provides the description of a new species in the genus Zabius Thorell (Scorpiones, Buthidae), Z. gaucho
n. sp., from four localities in the State of Rio Grande do Sul, Brazil. It differs from Zabius fuscus (Thorell 1 877) and Z. birabeni
Mello-Leitao 1938 in details of the telson shape, the longitudinal carinae on mesosomal tergites II-VI, and the number of
pectinal teeth. The genus was hitherto known only from Argentina, Z. fuscus being a frequent inhabitant of the central Sierras;

Z. birabeni , in turn, is probably a rare and non-orophilous scorpion, collected in scattered localities on the monte/chaco ecotone
and in northern Patagonia. The presence of a species of Zabius in southern Brazil lends additional support to the generalized
distributional track known as “peripampasic track,” which zoogeographically links the central Sierras Pampeanas with ancient
mountains in the southern province of Buenos Aires, southeastern Uruguay and southern Brazil.

Keywords: Scorpions, Neotropics, Argentina, taxonomy, new records

Resumo. Neste trabalho descrevemos uma especie nova do genero Zabius Thorell (Scorpiones, Buthidae), Z. gaucho n.sp.,
procendente de quatro localidades do Rio Grande do Sul, Brasil. Distingue-se de Zabius fuscus (Thorell 1877) e Z. birabeni
Mello-Leitao 1938 por detalhes da morfologia do telson, das cristas longitudinais dos tergitos do mesossoma, e pelo
numero de dentes pectineos. O genero era conhecido ate o momento so para a Argentina, sendo Z. fuscus urn escorpiao
muito freqiiente na regiao serrana central; Z. birabeni , no entanto, e aparentemente uma especie nao orofila e rara, coletada
em localidades dispersas no ecotono monte/chaco (oeste do pais) e no norte da Patagonia. A presem;a de uma especie de
Zabius no sul do Brasil representa um apoio adicional ao padrao generalizado de distribui^ao denominado de “track
peripampasico”, que vincula zoogeograficamente as Sierras Pampeanas com sistemas orograficos antigos do sul da
Provincia de Buenos Aires, do sudeste do Uruguai e do sul do Brasil.

The small buthid genus Zabius Thorell 1894 is restricted to
Argentina and previously included only two nominal species:
Z. fuscus (Thorell 1877) and Z. birabeni Mello-Leitao 1938.
The former species is a very common scorpion occurring in
orographic systems in central Argentina, while the latter seems
to be a rare species, reported from scattered rockless localities
in western and northern Patagonia (Abalos 1953; Maury 1979;
Acosta 1989, 1993, 1996; Acosta & Rosso de Ferradas 1996;
Mattoni & Acosta 1997; Acosta & Maury 1998; Ojanguren
Affilastro 2005). Therefore, the discovery of several specimens
of a hitherto undescribed species of Zabius in the State of Rio
Grande do Sul, Brazil, represents a remarkable novelty. This
new species is described below as Zabius gaucho n. sp. In this
paper we also provide new records for Z. fuscus and Z.
birabeni and discuss some doubtful reports of the former.
Abbreviated synonymies are given to include a few references
overlooked by or published after Fet & Lowe (2000). Since the
generic diagnosis of Zabius available in the literature is brief
(e.g., Mello-Leitao 1945; Abalos 1953), we provide a more
complete version, both so as to cover the character states
peculiar to Z. gaucho n. sp. and to include several characters
introduced by recent taxonomists. Zabius is the southernmost

buthid genus occurring in South America and also worldwide.
The Neotropical region contains relatively few genera of that
family, although one of them, Tityus C.L. Koch 1836, is the
most speciose in the order (Fet & Lowe 2000). The presence of
a member of Zabius in the state of Rio Grande do Sul has
interesting biogeographic implications since it adds further
evidence supporting the extension of the generalized distribu-
tional track known as the “peripampasic track” (Acosta 1989,
1993) into southern Brazil as briefly discussed below.

METHODS

Descriptions and line drawings were made using a Leica MSS
stereomicroscope equipped with drawing tube. Measurements
were taken with a graduated ocular and followed guidelines of
Stahnke (1970). Photographs were made using a Canon 400D
XTI with a 100 mm Macro-Canon lens. Descriptive terms and
abbreviations are as follows: carapacial carinae (Stahnke 1970):
Am, anterior median; Cm. central median; Pm, posterior median;
Cl, central lateral. Carapacial furrows: based on Stahnke
(1970), not abbreviated. Mesosomal carinae (adapted from
Vachon 1952): Md, median; Sm, submedian; Sl, sublateral.
Carinae of metasomal segments (Francke 1977). Segments I-

491

492

THE JOURNAL OF ARACHNOLOGY

IV: Di . dorsal lateral; Lsm, lateral supramedian; Lim. lateral
inframedian; Vl, ventral lateral; Vsm. ventral submedian.
Segment V: Dl. dorsal lateral; Lm. lateral median; Vl. ventral
lateral; Vsm. ventral submedian; Vm. ventral median. Chelal
carinae (Soleglad & Sissom 2001; partially also Vachon 1952):
Dl, digital; D3, dorsal secondary; D4. dorsal marginal; D5,
dorsal internal; I, internomedian; V3, ventrointernal; Vl,
ventroexternal; E, external secondary; Va, ventral accessory.
Carinae on pedipalp femur and patella are given topological
terms, to replace the use of “internal” and “external” (actually,
structures inside or outside the tegument); correspondences
with Stahnke’s (1970) nomenclature in brackets: prodorsal
(instead of dorso-interior), retrodorsal (dorso-exterior), pro-
ventral (ventro-interior), retroventral (ventro-exterior), dorsal
median (dorso-median), ventral median (ventro-median), retro-
lateral median (exterior-median), prolateral median (not
mentioned in Stahnke 1970).

Acronyms of repositories. — CDA: Catedra de Diversidad
Animal I, Facultad de Ciencias Exactas, Fisicas y Naturales,
Universidad Nacional de Cordoba, Cordoba, Argentina;
IBSP: Institute Butantan, Sao Paulo, Brazil; LEA: Collection
of Luis E. Acosta, Cordoba, Argentina; MACN: Museo
Argentine de Ciencias Naturales “Bernardino Rivadavia”,
Buenos Aires, Argentina; MCN: Museu de Ciencias Naturais,
Fundagao Zoobotanica do Rio Grande do Sul, Porto Alegre,
Brazil; MNRJ: Museu Nacional, Universidade Federal do Rio
de Janeiro, Rio de Janeiro, Brazil; ZMH: Zoologisches
Museum, Hamburg, Germany.

Additional material examined. Tityus argent inns: ARGEN-
TINA: Jujuy: Parque Nacional Calilegua, Toma de Agua
(1340 m), 23°41'S, 64°52'W, 25 February 1997 (L. Acosta), 1 <J
(LEA). Tityus bahiensis: BRAZIL: Mato Grosso do Sul:
Brasilandia, Fazenda Banna, 21°15'S, 52°03'W, 20 February
1993 (A.F. Beck, J.O. Silva), 1 <3, I 9 (LEA). Tityus confluens:
ARGENTINA: Cordoba: Chancani, 31°22'S, 65°29'W, 19
December 1987 (L. Acosta, F. Pereyra), 4 9 (CDA 000.506).
Tityus fasciolatus: BRAZIL: Distrito Federal Brasilia, 15°48'S,
47°55'W, April 1992 (M. Knox Brito), 1 3, 1 9 (LEA). Tityus
serrulatus: BRAZIL: Distrito Federal Brasilia, 15 48'S,
47°55'W, April 1992 (M. Knox Brito), 1 9 (LEA). Tityus
trivittatus: ARGENTINA: Santa Fe: Santa Fe, 31°38'S,
60°42'W, January 1985 (C. Espindola), 1 9 (CDA 000.533).
Tityus uruguayensis: URUGUAY : Canelones: 1 <3, Solymar
Norte, 34°49'S, 55°56'W, 18 February 1999, C.A. Toscano-
Gadea (LEA). Centruroides sculpturatus: USA: Arizona: 5
specimens. Alamo Dam State Park, Yuma Co., 34°14'N
1 13°35'W, 1 May 1970, M.A. Cazier (LEA). Rhopalurus
agamemnon : BRAZIL: Goids: 3 specimens, Campinagu, Serra
da Mesa, ca. 13°51'S, 48°33'W, 18 February-2 March 1996,
Silvestre, Brandao, Yamamoto (LEA). Rhopalurus rochae :
BRAZIL: Bahia: 2 specimens, Caatinga do Moura, 10°59'S,
40°45'W, 24-29 January 1980, P.E. Vanzolini (LEA).

SYSTEMATICS

Family Buthidae C.L. Koch 1837
Genus Zabius Thorell 1894

Zabius Thorell 1894:279.

Type species. — Isometrus fuscus Thorell 1877 by original
designation.

Diagnosis. — Small to medium-sized scorpions (up to 65 mm).
General color yellowish to dark ferrugineous, tegumentary
carinae usually darker. Anterior margin of carapace with a
notch; ocular mound slightly displaced anteromedially. Most
carapacial carinae not easily recognizable, normally obscured
by the irregular granulation: Am and Cm feeble or absent. Pm
fairly or well developed between the slight transverse posterior
furrow and the posterior border. Sternum narrowed elongated
pentagonal. Type 1 (nomenclature after Soleglad & Fet 2003):
apex reduced and slightly depressed, bordered posteriorly by
well defined lobes (resembling the anterior part of the lateral
lobes of Type 2 and giving the sternum the appearance of being
anteriorly bifid); posterior depression very deep, not surround-
ed by lobes. Mesosomal tergites I-VI with three distinct
longitudinal carinae (one Md, two Sm), tergite VII with five
carinae. Sternite V with two Sm and two Sl carinae. Metasoma
slender. Metasomal segments I-III bearing ten longitudinal
carinae (Lim complete in segment I, but much weaker or almost
vanishing in segments II and III); segment IV with eight carinae
(Lim lacking); segment V with five carinae (complete), plus Vsm
carinae limited to the proximal half. Telson without subaculear
apophysis; a blunt small protuberance is present instead; vesicle
globose (Z. fuscus, Z. birabeni) or more elongated (Z. gaucho n.
sp.), aculeus relatively short. Pedipalp chela dilated, with a lobe
at the base of the movable finger, corresponding to a notch in
the fixed finger (both dilatation and lobation more accentuated
in males). Nine chelal carinae (Figs. 10-12, 14), eight being
complete, plus a short basal Va between E and Vl, separating
trichobothria Eb\ and Eb2 (Abalos 1953: 350 named this
shortened Va the “dorsal accessory”); angle D3:D4:D5 = 90°.
Chelal carinae are conspicuous and well defined, except for V3,
the granulation (or crenulation) of which is present only in the
basal third, the rest being a smooth tegumentary border.
Denticular margin of the movable finger with 9-12 oblique
rows, with outer and inner accessory denticles, but without
supernumerary denticles. Trichobothriotaxy: subtractive neo-
bothriotaxy, type A, group a (Vachon 1974, 1975); femur with
10 trichobothria (d2 missing), patella with 13 (d2 very small and
difficult to see), chela with 15. Pectines with fulcra. Legs
without tibial spur; prolateral basitarsal spur on legs III and IV
bifid.

Comments on chelal carination. — The number and nomencla-
ture of carinae on the chela received some attention in the
literature. Vachon (1952) recognized a maximum of seven chelal
carinae, while Stockwell (1989) proposed a pattern of eight basic
carinae. More recently Prendini (2000), followed by Soleglad &
Sissom (2001), identified a derived pattern of 10 carinae in most
non-buthids, whereas buthids (represented as out-group in their
analyses only by the genus Centruroides Marx 1890) were
characterized as having an eight-carinate condition, presumably
basal for the order. There are many genera of Buthidae without or
with very obsolete chelal carinae, so that a homology assessment is
difficult. Others (e.g., Zabius and examples mentioned below)
have, in contrast, well developed carinae suggesting that a 9-
carinae pattern might be the basic condition for this family. In
their character state statement Prendini (2000) and Soleglad &
Sissom (2001) overlooked an “accessory” carina (Va), feeble in
some Centruroides but well developed in several other buthids. In
Zabius, carination is strong, and nine carinae can clearly be
identified (Figs. 10-12, 14) as described by Abalos (1953) and

ACOSTA ET AL. — ZA BIUS GAUCHO (BUTHIDAE) FROM BRAZIL

493

Ojanguren Affilastro (2005). Assuming that D2 (subdigital) and
V2 (ventral median) are lacking, and using D1 (digital) and VI
(ventral external) as “landmarks” an additional short though well
defined carina (Va, ventral accessory: Vachon 1952) is observed
between VI and E. Trichobothrium Eby is placed in the
intercarinal sector Dl-E, Eb2 between E and Va, and Eb\ between
Va and VI. The same condition (short, well defined Va, and the
described arrangement of Eh trichobothria) was verified in Tityus
argentinus Borelli 1899, T. bahiensis (Perty 1833), T. confluens
Borelli 1899, T. fasciolatus Pessoa 1935, T. serrulatus Lutz & Mello
1922, T. trivittatus Kraepelin 1898 and T. urugmyensis Borelli
1901 (L. Acosta pers. obs.). A rapid survey of the literature
revealed a similar arrangement in at least 20 additional species of
Tityus as well as in members of Alayotityus Annas 1973, Tityopsis
Armas 1974, and Microtityus Kjellesvig-Waering 1966 (Annas
1984; Lourengo & Vachon 1996). The development of Va appears
to be somewhat variable in Centruroides : in C. margaritatus
(Gervais 1841), C. exsul (Meise 1934), and C. testaceus (DeGeer
1778) it is short and granular, while in C. gracilis (Latreille 1804) it
is obsolete (Sissom & Francke 1983; Sissom & Lourengo 1987,
designated as “external secondary” therein). In Centruroides
exilicauda (Wood 1863), the position of Va is occupied by a V-
shaped rugulose sculpturation, not well defined, which leaves Eb\
and Eb2 to each side. Va is rudimentary (as an irregular row of
small granules) in Rhopalurus rochae Borelli 1910 and R.
agamemnon C.L. Koch 1839 (L. Acosta pers. obs.).

In Old World buthids, when present, Va appears to have a
full extension. In two Hottentotta species illustrated in
Lamoral (1979, figs. 45 and 66, sub Buthotus ), a diagonal Va
carina runs between Eb2 and Eb2, extending along the entire
palm to the articular condyle. In a generalized drawing,
Vachon (1952:62, fig. 69) shows a complete Va. too. In ventral
view (figs. 46 and 67 of Lamoral 1979), a longitudinal carina is
observed from the same condyle up to the palm base in the
position typical to VI. Further Old World buthids with a
complete Va include, e.g., Butbeolus thalassinus Simon 1882
(smooth to finely granular), Compsobuthus acute car inatus
(Simon 1882) (faint, smooth ridge), C. brevimanus (Werner
1936) (faint, granular ridge), and Amkoctonus bicolor Ehren-
berg 1828 (strong, smooth) (Sissom 1994; M. Soleglad, in lift.),
whilst the illustration of Compsobuthus vachoni Sissom 1994
suggests a short Va (Sissom 1994:19). Although examples are
not intended to give an exhaustive revision, it seems clear that
the full Va carina of Old World representatives and the
shortened version in Zabius, Tityus, Centruroides, Tityopsis,
and Alayotityus likely represent two states of the same feature.
If homologies, as currently assessed, are correct (especially
concerning the identification of VI), Va might prove to be
unique for buthids, not found elsewhere in the order.

Included species. — Zabius fuscus, Zabius birabeni, and
Zabius gaucho new species.

Distribution. — South America: Central Argentina to northern
Patagonia; southern Brazil (Fig. 1 5); Paraguay? The presence of
a Zabius species in southern Brazil adds evidence to the so called
“peripampasic track” (Acosta 1989, 1993). This generalized
track was identified through the approximate congruence of the
distributions of several scorpion and opilionid taxa, primarily
connecting the Pampean Sierras in central Argentina (provinces
of Cordoba and San Luis), with low systems in the province of
Buenos Aires (Tandilia and Ventania), and in southeastern

Uruguay. Additional extensions of the track were suggested
both to sub Andean chains in northwestern Argentina, and to
southern Brazil (Acosta 1989, 1993). Scorpion taxa which
support this distributional pattern are: ( 1 ) the Bothriurus
prospicuus species group, ranging from northwestern Argentina
to Uruguay (Acosta & Peretti 1998), with a hitherto unnamed
species in southern Brazil (C. Mattoni pers. comm.), (2) the
Bothriurus flavidus species group, with one widespread species
occurring in Cordoba, San Luis, La Pampa and Buenos Aires,
and one undescribed species in Uruguay (L. Acosta & C.
Toscano-Gadea unpubl.), (3) the Urophonius brachycentrus
species group, with one species widely distributed in central
Argentina, and a second in the province of Buenos Aires,
Uruguay, and southern Brazil (Maury 1977). The recent
discovery of new species either in Uruguay or Brazil (Z. gaucho
and the above mentioned Bothriurus) was indeed at least
predictable based on the track pattern. While the track is
presumed to link old mountains, most scorpions in it are no
longer strictly orophilous, although their distributions appear to
be primarily associated with those ranges. Harvestmen of the
peripampasic track include: (a) representatives of the genus
Ceratomontia Roewer 1914 (Triaenonychidae), with one species
in the central sierras + Ventania, another one in Tandilia +
Ventania + Uruguayan ranges + southern Brazil, and a third one
in southern Brazil (Maury & Roig Alsina 1985; Maury 2000),
and (b) genus Neopucroliella Roewer 1931 (Gonyleptidae), with
several species in the central sierras and one in Ventania (this is
the only taxon not yet recorded in Uruguay or Brazil; Acosta
1990).

KEY TO THE SPECIES OF ZABIUS

1. Mesosomal tergites II- VI: Sm carinae extended posteri-

orly in a noticeable projection (at least with the length
equivalent to three granules; Fig. 13); anterior margin of
carapace with a moderate notch (Figs. 3, 4); pectines with
7 10 teeth (female, male unknown); pectinal basal plate
subrectangular (Fig. 7); telson only slightly globose,
somewhat elongated; subaculear tuberosity weakly devel-
oped (Fig. 5) Z. gaucho

Mesosomal tergites 1 1— VI: Sm carinae not extended
posteriorly in a projection, but end at the tergite margin

or with at most a single granule prominent; front border
of carapace with an accentuated notch (Fig. 8); pectines
with 11-15 teeth (male and female); pectinal basal plate
divided in two lobes by a noticeable notch on the anterior
border (Fig. 9); telson markedly globose, subaculear
tuberosity small but evident 2

2. Total length up to 65 mm. General color dark ferrugi-
nous to light hazel. Stigmata elliptical; metasomal
segment V: Vl carinae uniformly granular, ranging from
a row of small granules to small blunt tubercles, none of
which is noticeably larger than the rest; denticular margin
of the movable finger of the pedipalp chela with 12

oblique granular rows Z. fuscus

Total length up to 45 mm; general color straw yellow;
stigmata rounded; metasomal segment V: Vl carinae with 2

or 3 large, tall tubercles or small apophyses in the middle,
which are clearly distinct from the remaining granules in the
row; denticular margin of the movable finger of the pedipalp
chela with 9 or 10 oblique granular rows Z. birabeni

494

THE JOURNAL OF ARACHNOLOGY

Figures 1, 2. — Zabius gaucho new species, female paratype (Riozinho, IBSP). 1, Dorsal aspect. 2. Ventral aspect. Scale line: 5 mm.

Zabius gaucho new species
Figs. 1-7, 10-13

Type material. — BRAZIL: Rio Grande do Sul: holotype
female, Taquara, Morro do Itacolomi (29°38'S, 50°46'W), 16
July 1992, L. Frug (MCN 553). Paratypes: BRAZIL: Rio
Grande do Sul: 1 female. Nova Petropolis (29°21'S, 51°08'W),
4 August 1993, E.E. Ely (MCN 580); 1 female, Riozinho
(29°38'S, 50°27'W), 21 January 2002, A.D. Brescovit (CDA
000.254); 1 female, same locality, 17 December 2001, I.
Bernard (IBSP 3464); 1 female, “Taguara del Mondo nuovo”
[actually Taquara], Mus. StraBburg, c. 1893 (c. = exchanged),
det. [Zabius] fuscus” [stricken out and

corrected on the label] (ZMH).

Other materia! examined. — BRAZIL: Rio Grande do Sul: 1
9, Tenente Portela (27°22'S, 53°46'W), 17 December 2002, M.
Dorneles (MCN 720, damaged).

Etymology. — The species name is derived from the Portu-
guese adjective “gaucho,” treated here as indeclinable; it is
popularly applied to the inhabitants of the Brazilian state of
Rio Grande do Sul, where the new species was collected.

Diagnosis. — Mesosomal tergites II— VI with 8m carinae
extended in a short but conspicuous posterior projection.
Anterior margin of carapace with a moderate notch. Pectines
with 7-10 teeth; basal plate subrectangular. Telson only
slightly globose, somewhat elongated; subaculear tuberosity
little developed. Because of their similar size and coloration
(hazel to dark ferrugineous), the overall appearance of Z.
gaucho n. sp. resembles that of Z. fuscus (Z. birabeni is readily
separated because of its small size, its lighter coloration, and
the distinctive small apophyses on Vl carinae of metasomal
segment V). A careful examination reveals, however, that Z.

fuscus shares more diagnostic characters with Z. birabeni than
with Z. gaucho n. sp. In both Z. fuscus and Z. birabeni the
carinae of mesosomal tergites do not show marked projections
and the telson is more markedly globular (cf. Ojanguren
Affilastro 2005). They also have higher pectinal teeth counts,
and the pectinal basal plate has a distinctive anterior notch as
if composed of two halves fused medially (Fig. 9). Specimens
of Z. gaucho n. sp. show a faint and irregular reticulate
pigment, especially on the mesosomal tergites, which is not
present in the congeners. Granules of some carinae or granular
areas (e.g., medial border of pedipalp femur; Dl carinae all
along the metasoma) are taller and more acute in Z. gaucho n.
sp. Carapace carinae are poorly defined in all three species,
even least defined in the new species.

Description, — Male unknown. Females: total body length
41.7-55.3 mm. Detailed measurements of holotype, together
with summary values of the available sample (mean,
maximum, minimum): Table 1.

Color: General color reddish hazel, with granules and
pedipalp fingers darker; legs and ventral side (coxae, sternal
region, pectines, steraites I— III) lighter, sternites IV and V
gradually darkening to the general coloration; chelicerae
yellowish, with reticulate pigment dorsaliy on the cheliceral
hand behind the fingers. Carapace and especially mesosomal
tergites with very faint reticulate pigmentation.

Morphology: Carapace anterior margin with a V-shaped but
not accentuated median notch, bordered by a granular row.
Surface of carapace covered by coarse granulation. Median
anterior and median central furrows shallow, the former
wider; median posterior and postmarginal furrows forming a
deep triangular depression. Lateral posterior furrow deep, the

ACOSTA ET AL. — ZABIUS GAUCHO (BUTHIDAE) FROM BRAZIL

495

Figures 3-9. — Zabius species. 3-7. Z. gaucho new species. 3. Carapace, front half. 4. Outline of carapace. 5. Telson and metasomal segment V,
lateral view. 6. Metasomal segment V, ventral view. 7. Sternum, genital operculi, pectinal plate and pectines. 8, 9. Z. fuscus (Thorell). 8. Outline
of carapace. 9. Sternum, genital operculi, pectinal plate and pectines. Scales: 1 mm (Figs. 4 and 8 not at scale).

remaining ones just insinuated by a slight depression and/or
lack of granulation. Most carapacial carinae unrecognizable,
much less defined than in congeners, Am and Cm almost
absent; short superciliary crest as a row of pearl-like, dark
conspicuous granules, leaving a well defined interocular
groove in between; Cl represented by an oblique row of small
granules (most granulation on the area is oriented a similar
way); Pm fairly defined along the longitudinal tegumentary
ridge each side of the posterior marginal furrow, up to the
posterior transverse furrow. Posterior margin with a row of
granules; posterior end of Pm terminates with a pointed

granule on the carapace margin, aligned with the Sm carinae of
tergites (see below).

Mesosomal tergites: finely granular on the pretergites and
the anterior half of tergites, the posterior half with coarse
granulation, as on carapace. Tergites I-VI with three aligned
longitudinal carinae (one Md, two Sm). Md very short on
tergite I, becoming more developed posteriorly (on tergite II it
is extended up to one third of the tergal plate, III and IV up to
the half, and almost complete on V and VI); in all cases ending
in a blunt granule. Sm carinae shorter than Md; they are feeble
on tergite I, but on tergites II— VI ending in an acute granule of

496

THE JOURNAL OF ARACHNOLOGY

Figures 10-14. — Zubins species. 10-13. Z. gaucho new species. 10-12. Right pedipalp chela: 10. Lateral view, 1 1 . Ventral view, 12. Mesal view.
13. Mesosomal tergite V, dorsal view. 14. Z. fuscus (Thorell), schematic transverse section of the right pedipalp chela (basal third), showing all
nine carinae. Abbreviations: Dl, digital, D3, dorsal secondary, D4, dorsal marginal, D5, dorsal internal, I, internomedian, V3, ventrointernal,
VI, ventroexternal, E, external secondary, VA, ventral accessory. Scales: 1 mm (Fig. 14 not at scale).

increasing size, almost resembling a tiny apophysis over the
tergite’s edge. Tergite VII with Sl and Sm carinae complete,
Md restricted to a very short, basal portion. Sternites I and II
almost smooth, from sternite III onwards, coarsely granular
on the lateral border, and Sl carinae insinuated; these are
more defined on sternite IV; sternite V granular, Sm and Sl
carinae almost complete (they do not reach the anterior
border). Stigmata oval. Number of pectinal teeth (n = 1 1): 7
teeth (2 pectines), 8 (8), 10 (1); holotype with 10-[damaged],
Metasoma. Segments I-IV: Vsm, Vl, Lsm, and Dl carinae
complete, the latter with large and acute granules (especially
posteriorly); Lim complete on segment I, it is very weak on
segments II and III, and is absent on segment IV. Segment V:
Vm and Vl complete, Vsm less conspicuous and restricted to
the basal half; Lm is represented by few unordered granules,
while Dl are complete but little evident, formed by blunt

granules. Telson slightly globose, with almost no trace of
subaculear tubercle; ventral surface rugulose. Pedipalps.
Medial surface of trochanters coarsely granular, granules are
especially tall and acute proventrally. Pro- and retrodorsa! and
proventral carinae of femur complete, regularly granular;
prolateral median carina of tall conical, unequal granules or
small apophyses; retrolateral median carina homogeneously
crenulate, ending in an acute conical apophysis near the base;
retroventral carina not defined, replaced by a flattened area
with subtle granulation. Patella with pro- and retrodorsal, and
retroventral carinae regularly granular; proventral carina
crenulate, ending in a large conic protuberance (absent on
the basal third of the segment, where a smooth concavity
exists); dorsal median and retrolateral median carinae feeble,
ventral median very weak; prolateral surface smooth, except
for a large conical protuberance, aligned with the proventral

ACOSTA ET AL. — ZABIUS GAUCHO (BUTHIDAE) FROM BRAZIL

497

Table 1. — Detailed measurements (mm) of the holotype female of
Zabius gaucho n. sp. (MCN 553), and summary values for the
available sample (;; = 5; ZMH not included).

Character

Holotype

Mean (max-min)

Body length

55.3

46.5 (55.3-41.7)

Carapace length

6.6

5.7 (6.6-5. 3)

anterior width

3.1

3.6 (4. 1-3.0)

posterior width

6.3

6.0 (6.5-5. 6)

Mesosoma length

16.6

12.5 (16.6-10.4)

Metasoma length

26.4

23.2 (26.4-20.2)

Metasomal segment I length

4.5

3.7 (4.5-3. 1)

width

2.8

2.4 (2. 8-2.0)

Metasomal segment II length

5.1

4.5 (5. 1-3.8)

width

2.5

2.1 (2.5-1. 8)

Metasomal segment III length

5.3

4.6 (5. 3-4.0)

width

2.3

2.0 (2.3-1. 6)

Metasomal segment IV length

5.2

4.8 (5. 2-4.3)

width

2.2

1.9 (2.2-1. 7)

Metasomal segment V length

6.3

5.5 (6. 3-5.0)

width

2.0

1.8 (2.0-1. 7)

height

2.3

2.0 (2.3-1. 8)

Telson length

5.7

5.1 (5.7-L7)

width

2.1

1.8 (2. 1-1.6)

height

2.1

1.9 (2.1-1. 8)

Aculeus length

2.4

1.9 (2.4-1. 7)

Pedipalp total length

24.4

22.1 (24.4-20.8)

Pedipalp femur length

6.3

5.7 (6.3-5. 1)

width

2.1

1.8 (2. 1-1.5)

Pedipalp patella length

6.7

5.9 (6.7-5. 5)

width

2.3

2.2 (2.3-2. 1)

Pedipalp chela length

11.4

10.6 (11.4-10.1)

width

4.3

4.0 (4. 3-3.0)

height

3.5

3.0 (3.6-2. 1)

Movable finger length

7.0

6.5 (7.0-6.2)

one. Chelal carinae Dl, D3, and D4 long, attaining the finger
dorsum, with blunt granules on the hand but smooth on the
finger; D5 and I crenulate, the area in between as a
conspicuous longitudinal concavity; VI well defined, with
conspicuous granulation in all its length; V3 only granular at
its base, the rest is at most a smooth weak tegumentary
border; E extended up to the vicinity of the small Esb, it is
crenulated basally, the rest with normal granulation; Va
carina between E and VI crenulated, limited to the basal
fourth of the manus. Trichobothriotaxy as for the genus.
Denticular margin of the movable finger with 1 1 or 12 oblique
slightly imbricate rows of granules, the most basal continued
longitudinally along the edge, 11 or 12 inner and outer
accessory denticles, without supernumerary granules.

Distribution. —This species has only been recorded from the
state of Rio Grande do Sul, Brazil. Nothing is known about
the natural history of this species. The two specimens from
Riozinho were collected in urban areas, despite the fact that
this city is the third best protected Atlantic Forest area, with
90% of the native forest preserved (SOS Mata Atlantica,
2004). Both specimens were collected in the same place, a
peripheral district divided into several plots that has been
inhabited for more than 10 years. The first specimen (CDA
000.254) was found inside a house, the second one (MCN 705)
in the backyard between the house and the margin of a narrow
stream in a grassy area with brick piles, stones, and wood. It is
surprising that such a large and conspicuous species remained

undetected until now. Its presence in ancient collections (e.g.,
the specimen of ZMH), and the inclusion of Zabius in an old
key of Brazilian scorpions (Ihering 1915) suggests that earlier
authors knew or suspected the existence of this genus in
southern Brazil, although it was never explicitly stated.

Zabius fuscus (Thorell 1877)

Figs. 8, 9, 14

Isometrus fuscus Thorell 1 877: 1 40.

Zabius fuscus: Thorell 1893:372: Ringuelet 1953:280; Abalos
1953:349, 350, figs. 1-22; Maury 1973:353, 363; Bucher &
Abalos 1979:401, fig. 45; Gualdoni et al. 1986:87; Acosta
1989:122, figs 88-90; Corronca & Peralta 1995:121; Mat-
toni & Acosta 1997:72, 77; Fet & Lowe 2000:279 [complete
reference list therein]; Ojanguren Affilastro 2005:103
[redescription, illustrations].

[Zabius] fuscus: Kraepelin 1895:8, 18 [mentioned as "Tityus
fuscus” but accepting Thorell's combination].

Type material. ARGENTINA: holotype, “Cordova”,
“Weijenbergh ded.” deposited in Naturhistoriska Riksmuseet
Stockholm (Coll. Thorell 43/22).

New records. ARGENTINA: Cordoba: 1 9, Cerro Colo-
rado, 30°06'S, 63 56' W, 21 February 1987, L. Acosta, A.
Peretti (CDA 000.208); 2 3, between Cerro Colorado and
Caminiaga, ca. 30 04'S, 63°59'W, 21 February 1987, L. Acosta
(CDA 000.216); 1 juvenile. Dean Funes, 30°29'S, 64°16'W, 24
June 1986, L. Acosta (CDA 000.214); 1 9 with young on the
dorsum, “La Industrial Salinera”, between Lucio V. Mansilla
and San Jose de las Salinas (under railway sleeper, border of
saline), 29°54'S, 64°40'W, 19 November 1998, C. Mattoni, A.
Peretti (CDA 000.221); 1 juvenile, between Canada de Rio
Pinto and Todos los Santos, ca. 30 47'S, 64°17'W, 22
November 1986, L. Acosta (CDA 000.206); 2 3, Todos los
Santos, 6 km to Ongamira, ca. 30 47'S, 64 ’23 'W, 26 December
1987, L. Acosta, A. Peretti (CDA 000.209); 3 juveniles,
Candonga, 31°05'S, 64 20'W, July 1957, E. Gibson (CDA
000.224); 2 juveniles, same collection data except July 1958
(CDA 000.225); 1 juvenile. Villa Colanchanga, Dique La
Quebrada, 31 "09'S, 64°21'W, 15 September 1985, L. Acosta
(CDA 000.204); 1 3, 1 9, 2 juveniles, Cerro Uritorco, 30°51'S,
64°29'W, 2 May 1987, L. Acosta (CDA 000.218); 3 juveniles,
La Cumbre, road to Cuchi-Corral. ca. 30°59'S, 64 34' W,
March 1986, M. Cabrera (CDA 000.210); 1 3, Tiu Mayo, 2 km
to La Cumbre, 30°58'S, 64 26'W. 3 May 1987, L. Acosta (CDA
000.220); 1 3, 1 juvenile, Vaquerias, 31°07'S, 64°27'W, 29
December 1985, L. Acosta (CDA 000.211); 1 juvenile, Cerro
Pan de Azucar, 31°14'S, 64 ’25'W. 15 April 1990, L. Acosta, A.
Peretti (CDA 000.212); 1 juvenile. Villa San Jose, Mallin,
31°18'S, 64 34'W, 23 August 1986, L. Acosta, R. Pizzi (CDA
000.219); 1 3, 2 juveniles, Tanti, Cerro Blanco, 31°21'S,
64°36'W, July 1968, R. Martori (CDA 000.230); 2 juveniles.
Villa del Lago, 31°24'S, 64°30'W, 10 May 1981, L. Acosta
(CDA 000.229); 1 juvenile, San Antonio de Arredondo (in a
house, upstairs), 31°29'S, 64°32'W, 27 January 1988. H. Getar
(CDA 000.228); 1 9, Cuesta Blanca, 31°29'S, 64°35'W, January
1989, C. Bone (CDA 000.233); 1 9, El Diquecito, 31°22'S,
64°23'W, October 1970, Brigado, Puentes (CDA 000.226); 1 3,
Villa Diquecito (“Las Bateas”), 31°21'S, 64°21'W, 24 October
1980, L. Acosta (CDA 000.202); 1 9, Cordoba (center of the

498

THE JOURNAL OF ARACHNOLOGY

city), 31°24'S, 64°11'W, 28 November 2000, N. Fernandez
(CDA 000.244); 1 juvenile, San Clemente, 31°43'S, 64°38'W, 23
September 1986, L. Acosta (CDA 000.222); 2 ?, Atos Pampa,
31°59'S, 64342'W, 16 December 1967, no collector (CDA
000.223); 1 $, between Atos Pampa and Yacanto de Calamu-
chita, ca. 32°01'S, 64°42'W, 26-27 December 1987, P. Boero de
Cabrera (CDA 000.203); 1 juvenile. Las Albahacas, 32°54'S,
64°50'W, 15 November 1981, M. Gualdoni (CDA 000.231); 1 <J,

1 ?, 3 km N of Achiras, 33°10'S, 65°00'W, 1 March 2000, L.
Acosta, A. Peretti (CDA 000.026); 1 S, 1 juvenile, Quebrada de
la Mermela, Chancani, 31°24'S, 65°25'W, 20 December 1987, L.
Acosta, F. Pereyra (CDA 000.236); 2 juveniles, Nina Paula (near
Mina Clavero), 31°45'S, 64°56'W, 1 1 March 1981, M. Gualdoni
(CDA 000.207). San Luis: 1 ?, 2 juveniles. Sauce de los
Chorrillos (5 km E San Luis), 33°17'S, 66°15'W, 15 July 1970,
R.D. Sage (CDA 000.197); 1 <?, 1 juvenile, 5 km road from San
Francisco del Monte de Oro to Carolina, 32°40'S, 66°08'W, 12
March 1994, L. Acosta, C. Mattoni (CDA 000.199); 3 juveniles,
Suyuque Nuevo, 33°08'S, 66°15'W, 14 March 1994, L. Acosta
(CDA 000.200). La Rioja: 1 <3, 1 juvenile, Santa Lucia, near
Chamical (620-660 m), 30°29'S, 66°20'W, 6 December 1994, L.
Acosta, C. Mattoni (CDA 000.085).

Distribution. -Zabius fuscus appears to be an almost strictly
orophilous scorpion (Acosta 1989; Acosta & Rosso de
Ferradas 1996) the range of which coincides with the Pampean
Sierras of San Luis, Cordoba, La Rioja, Catamarca and
Santiago del Estero in central Argentina (Fig. 15; Maury 1979;
Acosta & Maury 1998; Ojanguren Affilastro 2005). The type
locality is “Cordoba” and most likely refers to the city.
Despite the statement of Teruel (2003: 147) about the
“sporadic presence” of Z. fuscus in collections, this species is
very common, and good series are available in Argentinean
institutions. Many localities from Cordoba and San Luis were
cited by Mello-Leitao (1934), Abalos (1953), and Ojanguren
Affilastro (2005). Mattoni & Acosta (1997) added some
localities in the insular sierras of the province of La Rioja
(Loma Larga, between Malanzan and Loma Larga, La
Calera, Oita, Dique de Oita, Quebrada del Padre), which
might represent a disjunct area for the species. Specimens of Z.
fuscus are readily caught under mid-sized or large stones. In
summer, collecting with UV light suggests that many
individuals hide in deep crevices of the rock walls and in
“pircas” (a kind of fence made with piled stones), though not
in high densities. Females are normally seen at the opening of
their retreats, with only the pedipalp fingers outside,
presumably to detect the presence of prey; males are more
active, and more frequently are found clinging or walking on
the wall. Zabius fuscus is commonly collected together with
Bothriurus cordubensis Acosta 1995 and, like this species, its
range excludes the higher montane belt of the Sierras Grandes
(above 1800-1900 m). Aside from several findings of Z. fuscus
in houses in small villages in the sierras, where they are
thought to enter accidentally, some specimens were sporadi-
cally collected in the highly urbanized city of Cordoba (see
record above), most probably because of being transported
together with wood or rocks. One female with young on the
dorsum was caught in an unexpected site, on the very border
of the extensive saline area called “Salinas Grandes” (A.
Peretti & C. Mattoni, pers. comm.). The specimen was close to
an almost abandoned old salt factory under a railway sleeper

lying on a salty surface. This finding is remarkable since
the site consists of a plain with only small halophilous
shrubs, is surrounded by Chaco vegetation, and lies around
40 km from the nearest rocky area. Without providing details,
Kraepelin (1899) recorded Z. fuscus from Paraguay, which
was considered rather unlikely by Maury (1984) and
Lourenqo (1994). Kraepelin's (1899) material is stored in
ZMH (“77/rnv (P-hajkts) [Zabius] fuscus (Thor.), Burmeister d.
1890, Paraguay” [genus name stricken out and corrected on
the label], examined). Since the specimen consists of a badly
preserved juvenile and comes from a very old collection,
doubts remain about the accuracy of the identification and the
locality, so we agree that a further confirmation is needed. The
species was also cited from the province of Jujuy (northern
Argentina) by Mello-Leitao (1934) — a specimen collected by
Salvador Mazza but no longer kept in any repository; Maury
(1979) did not include this doubtful locality in his map.
References of Z. fuscus from the province of Tucuman
(Abalos 1953) were accepted by Maury (1979) and Corronca
& Peralta (1995) but deemed to need confirmation by Mattoni
& Acosta (1997) and Ojanguren Affilastro (2005). According
to Teruel (2003), a specimen of Zabius was collected in Tucu-
man though the species identity remains an open question. Be
this record assigned to Z. fuscus or not, the genus seems not as
frequent in Tucuman as in other parts of its range.

Zabius birabeni Mello-Leitao 1938
Zabius birabeni Mello-Leitao 1938:84, fig. 1; Abalos 1953:353,
figs. 23-27 (in part?); Fet & Lowe 2000:279 [complete
reference list therein]; Ojanguren Affilastro 2005:105
[redescription, illustrations].

Type material. -ARGENTINA: Rio Negro : holotype male,
Valcheta (40°41'S, 66°09'W), M. Biraben (Museo de La Plata
18060). According to Mello-Leitao (1938) a paratype from
“Gaviotas, province of La Pampa” should exist at MNRJ, but
it is not included on the type list provided by Kury &
Nogueira (1999) and is probably lost.

New record. -ARGENTINA: La Rioja : 1 juvenile, Bajo
Hondo (31°41'S, 66°00'W), 13 November 1982, E. Maury
(MACN).

Distribution.- Recorded from the Argentinean provinces of
Rio Negro, La Pampa. Buenos Aires, Entre Rios, La Rioja,
San Juan, San Luis and Cordoba (Abalos 1953; Maury 1973;
Acosta 1996; Ojanguren Affilastro 2005). With the exception
of a few references from the sierras in southern province of
Buenos Aires (Abalos 1953; Maury 1973; Ojanguren Affilastro
2005), whose accuracy may in some cases need revision
(Acosta 1996), Z. birabeni is characteristic of rockless areas
(Fig. 15). Although the records are somewhat scattered, most
localities embrace an ample arc that roughly corresponds to an
ecotonal area between Monte and Chaco+Espinal (Acosta
1996). In contrast to this wide range, the species appears to be
quite rare (Acosta 1996; Ojanguren Affilastro 2005). The
locality “50 km Sierra Grande”, province of Rio Negro
(Ojanguren Affilastro 2005), represents the southernmost
record for the entire family Buthidae.

ACKNOWLEDGMENTS

We are indebted to Dr. Hieronymus Dastych (ZMH) and
the late Dr. Emilio Maury (MACN) for enabling LEA to

ACOSTA ET AL. — ZABIUS GAUCHO (BUTHIDAE) FROM BRAZIL

499

Figure 15. — Studied localities of Zabius gaucho new species (black stars), Z. fuscus (light circles), and Z. birabeni (black circles). Solid thick
lines: country boundaries; thin lines: rivers; dashed lines: province or state boundaries. Abbreviations: Ju: province of Jujuy, Tuc: province of
Tucuman, SE: province of Santiago del Estero, LR: province of La Rioja, ER: province of Entre Rios, Cor: province of Cordoba, SL: province
of San Luis, LP: province of La Pampa, BA: province of Buenos Aires, RN: province of Rio Negro, RS: state of Rio Grande do Sul. Inset:
location of the represented area.

examine several specimens from their collections. The senior
author also thanks Camilo Mattoni and Alfredo Peretti for
providing information on the capture site of Zabius fuscus in
the Salinas Grandes area, and Michael Soleglad for fruitful
discussion on the nomenclature of chelal carinae. The latter,
Victor Fet, and two referees (Lorenzo Prendini, Erich
Volschenk) made many useful suggestions to improve the
manuscript. The collaborative work was facilitated through a
travel grant to LEA ( Programa de Centres Associados de P6s-
graduagao Brasil I Argentina, CAPES I SPU - EJniversidade
Estadual de Campinas / Universidad Nacional de Cordoba).
Additional support was provided by Projeto Biota/Fapesp (#
99/05446-8 to ADB and DMC) and CNPq to ADB (grant #

301776/2004-0). LEA is a researcher of the Consejo Nacional
de Investigaciones Cientificas y Tecnicas, Argentina.

LITERATURE CITED

Abalos, J.W. 1953. El genero Zabius Thorell. 1894 (Buthidae,
Scorpiones). Anales del Institute de Medicina Regional 3:349-
356.

Acosta, L.E. 1989. La fauna de escorpiones y opiliones (Arachnida)
de la provincia de Cordoba. Ph. D. Thesis, Facultad de Ciencias
Exactas, Fisicas y Naturales, Universidad Nacional de Cordoba.
333 pp.

Acosta, L.E. 1990. Redescripcion de Neopucroliella mesembrina
Ringuelet, 1961 nov. stat. (Opiliones, Gonyleptidae, Pachylinae).
Revista de la Sociedad Entomologica Argentina 48:97-105.

500

THE JOURNAL OF ARACHNOLOGY

Acosta, L.E. 1993. Escorpiones y opiliones de la provincia de
Cordoba (Argentina): diversidad y zoogeografia. Bulletin de la
Societe Neuchateloise des Sciences Naturelles (XII Ie Colloque
europeen d' Arachnologie) 116:1 1-17.

Acosta, L.E. 1996. Nuevas localidades de escorpiones argentinos.
Revista de la Sociedad Entomologica Argentina 55:121-125.

Acosta, L.E. & E.A. Maury. 1998. Scorpiones. Pp. 545-559. In
Biodiversidad de Artropodos Argentinos. Una perspectiva biotax-
onomica. (J.J. Morrone & S. Coscaron, dirs.). Ediciones Sur, La
Plata, Argentina.

Acosta, L.E. & A.V. Peretti. 1998. Complemento a la description de
Bothriurus cordubensis (Scorpiones, Bothriuridae), con anotaciones
sobre patrones evolutivos del genero en Argentina. Revue
Arachnologique 12:95-108.

Acosta, L.E. & B. Rosso de Ferradas. 1996. Aracnidos de la Provincia
de Cordoba. Pp. 71-99. In Biodiversidad de la Provincia de
Cordoba, Fauna 1. (I.E. di Tada & E.H. Bucher, eds.).
Universidad Nacional de Rio Cuarto, Rio Cuarto, Argentina.

Armas, L.F. de. 1984. Escorpiones del archipielago Cubano. VII.
Adiciones y enmiendas (Scorpiones: Buthidae, Diplocentridae).
Poeyana 275:1-37.

Bucher. E.H. & J.W. Abalos. 1979. Fauna. Pp. 369-434. In Geografia
Fisica de la Provincia de Cordoba. (J. Vazquez, R. Miatello & M.
Roque, dirs.). Ed. Boldt, Buenos Aires.

Corronca, J.A. & M.A. Peralta. 1995. Lista de tipos y material de
escorpiones (Arachnida: Scorpionida), de la coleccion de la
Fundacion Miguel Lillo (Tucuman, Argentina). Acta Zoologica
Lilloana 43:117-123.

Fet, V. & G. Lowe. 2000. Family Buthidae C.L. Koch, 1837.
Pp. 54-286. In Catalog of the Scorpions of the World (1758-1998).
(V. Fet. W.D. Sissom, G. Lowe & M.E. Braunwalder, eds.). The
New York Entomological Society, New York.

Francke, O.F. 1977. Scorpions of the genus Diplocentrus from
Oaxaca, Mexico (Scorpionida. Diplocentridae). Journal of Arach-
nology 4:145-200.

Gualdoni, C.M., M. del C. Corigliano & A.M. Oberto. 1986.
Caracteristicas biometricas del crecimiento y alometria en Zabius
fuscus (Thorell, 1877) (Scorpiones, Buthidae). Revista de la
Universidad Nacional de Rio Cuarto 6:87-96.

Ihering, R. von. 1915. Os escorpioes do Brazil meridional. Annaes
Paulistas de Medicina e Cirurgia 5(2-3-4):49 57.

Kraepelin, K. 1895. Nachtrag zu Theil I der Revision der Scorpione.
Beiheft, Jahrbuch der Hamburgischen Wissenschaftlichen Anstal-
ten 12:73-96.

Kraepelin, K. 1899. Scorpiones und Pedipalpi. Das Tierreich 8:
1-265, Verlag von R. Friedlander und Sohn, Berlin.

Kury, A.B. & A.L.C. Nogueira. 1999. Annotated check list of type
specimens of Arachnida in the Museu Nacional - Rio de Janeiro. I.
Scorpiones. Pseudoscorpiones and Solifugae. Publicagoes Avulsas
do Museu Nacional. Rio de Janeiro 77:1-19.

Lamoral, B.H. 1979. The scorpions of Namibia (Arachnida:
Scorpionida). Annals of the Natal Museum 23:497-784.

Lourenyo, W.R. 1994. Synopsis de la faune des Scorpions (Chelicer-
ata) du Paraguay. Revue Suisse de Zoologie 101:677-684.

Louremjo, W.R. & M. Vachon. 1996. Complements a la phylogenie et
a la biogeographie des genres Alayotityus Armas et Tityopsis
Armas (Scorpiones, Buthidae). Biogeographica 72:33-39.

Mattoni, C.l. & L.E. Acosta. 1997. Scorpions of the insular Sierras in
the Llanos District (Province of La Rioja, Argentina) and their
zoogeographical links. Biogeographica 73:67-80.

Maury, E.A. 1973. Los escorpiones de los sistemas serranos de la
provincia de Buenos Aires. Physis, Seccion C 32:351-371.

Maury, E.A. 1977. Comentarios sobre dos especies de escorpiones del
genero Urophonius (Bothriuridae). Revista del Museo Argentine de
Ciencias Naturales "Bernardino Rivadavia”, Entomologia 5:143-
160.

Maury, E.A. 1979. Apuntes para una zoogeografia de la escorpio-
fauna argentina. Acta Zoologica Lilloana 35:703-719.

Maury, E.A. 1984. Lista de los escorpiones conocidos del Paraguay
(Scorpiones: Buthidae, Bothriuridae). Neotropica 30:215-217.

Maury, E.A. 2000 (1999). Triaenonychidae sudamericanos. VII. El
genero Ceratomontia en el Brasil (Opiliones: Laniatores). Revista
de la Sociedad Entomologica Argentina 58:33-36.

Maury, E.A. & A.H. Roig Alsina. 1985. Triaenonychidae sudamer-
icanos. I. El genero Ceratomontia Roewer 1915 (Opiliones,
Laniatores). Historia Natural 5:77-92.

Mello-Leitao, C. de. 1934. Estudo monografico dos escorpiSes da
Republica Argentina. Octava Reunion de la Sociedad Argentina de
Patologia Regional del Norte, Pp. 1-97, Buenos Aires.

Mello-Leitao, C. de. 1938. Notas sobre alacranes argentinos. Notas
del Museo de La Plata, Zoologia 3:83-95.

Mello-Leitao, C. de. 1945. Escorpioes sul-americanos. Arquivos do
Museu Nacional 40:1-468.

Ojanguren Affilastro, A. A. 2005. Estudio monografico de los
escorpiones de la Republica Argentina. Revista Iberica de
Aracnologia 11:7 5-24 1 .

Prendini, L. 2000. Phylogeny and classification of the superfamily
Scorpionoidea Latreille 1802 (Chelicerata, Scorpiones): an exem-
plar approach. Cladistics 16:1-78.

Ringuelet, R.A. 1953. Geonenria de los escorpiones en la Argentina y
las divisiones zoogeograficas basadas en su distribution. Revista
del Museo de La Plata, Zoologia 6:277-284.

Sissom, W.D. 1994. Descriptions of new and poorly known scorpions
of Yemen (Scorpiones: Buthidae, Diplocentridae, Scorpionidae).
Fauna of Saudi Arabia 14:3-39.

Sissom, W.D. & O.F. Francke. 1983. Redescription of Centruroides
testaceus (DeGeer) and description of a new species from the
Lesser Antilles (Scorpiones: Buthidae). Occasional Papers, The
Museum, Texas Tech University 88:1-13.

Sissom, W.D. & W.R. Lourenpo. 1987. The genus Centruroides in
South America (Scorpiones. Buthidae). Journal of Arachnology
15:11-28.

Soleglad, M.E. & V. Fet. 2003. The scorpion sternum: structure and
phylogeny (Scorpiones: Orthosterni). Euscorpius, Occasional
Publications in Scorpiology 5:1-34.

Soleglad, M.E. & W.D. Sissom. 2001. Phylogeny of the family
Euscorpiidae Laurie, 1896: a major revision. Pp. 25-111. In
Scorpions 2001. In Memoriam Gary A. Polis. (V. Fet. & Selden,
P.A., eds.). British Arachnological Society, Burnham Beeches,
Buckinghamshire, UK.

SOS Mata Atlantica. 2004. Atlas dos Municipios da Mata Atlantica.
Online at http: // www.sosmatatlantica.org.brndex.php?section=
project&action=AtlasMunicipio.

Stahnke, H.L. 1970. Scorpion nomenclature and mensuration.
Entomological News 81:297-316.

Stockwell, S.A. 1989. Revision of the phylogeny and higher
classification of Scorpions (Chelicerata). Ph.D. Thesis, University
of California at Berkeley. 413 pp.

Teruel, R. 2003 (2002). Confirmation de la presencia del genero
Zabius Thorell 1894 (Scorpiones: Buthidae) en la provincia de
Tucuman, Argentina. Revista Iberica de Aracnologia 6:147-148.

Thorell, T. 1877. Etudes scorpiologiques. Atti Societa Italiana Scienze
Naturali 19:75-272.

Thorell, T. 1894. Scorpiones exotici R. Musei Historiae Naturalis
Florentini. Bolletino della Societa Entomologica Italiana 25:356—
387.

Vachon, M. 1952. Etudes sur les Scorpions. Institut Pasteur d’Algerie,
Alger. 482 pp.

Vachon, M. 1974. Etude sur les caracteres utilises pour classer les
families et les genres de Scorpions (Arachnides). 1. La trichobo-
thriotaxie en Arachnologie. Sigles trichobothriaux et types de
trichobothriotaxie chez les Scorpions. Bulletin du Museum

ACOSTA ET AL ,—ZABIUS GAUCHO (BUTHIDAE) FROM BRAZIL

501

National d’Histoire Naturelle, 3e serie n° 140 (Zoologie 104):
857-958.

Vachon, M. 1975. Sur Fertilisation de la trichobothriotaxie du bras des
pedipalpes des Scorpions (Arachnides) dans le classement des

genres de la famille des Buthidae Simon. Comptes Rendus des
Seances de FAcademie des Sciences, Paris, ser. D, 281:1597-1599.

Manuscript received 7 May 2007, revised 4 April 2008.

2008. The Journal of Arachnology 36:502-512

Revision of the spider genus Taira (Araneae, Amaurobiidae, Amaurobiinae)

Zhi-Sheng Zhang: College of Life Sciences, Southwest University, Chongqing, 400715, China

Ming-Sheng Zhu1 and Da-Xiang Song: College of Life Sciences, Hebei University, Baoding, Hebei 071002, China

Abstract. The cribellate amaurobiid genus Taira, which is suggested to be most similar to the genus Amaurobius and some
other related genera, is revised based on genital characters. Eight species, one from Japan and seven from China, are
included. Five new species from China are described: T. cangshan, T. concava, T. latilabiata, T. obtusa, and T. sulciformis.
Titanoeca decorata Yin & Bao 2001 is newly transferred to this genus and its male is described for the first time. Taira
lunaris Wang & Ran 2004 is newly synonymized with T. liboensis Zhu, Chen & Zhang 2004.

Keywords: Genital structures, new species, China, Japan, synonomy

The amaurobiid genus Taira Lehtinen 1967, with the type
species T. flavidor salis (Yaginuma 1964), is restricted to Japan
and South China. Only three species have been previously
included in the genus, the type species, T. liboensis Zhu, Chen
& Zhang 2004 and T. lunaris Wang & Ran 2004, both from
Libo County, Guizhou Province of China (Zhu et al. 2004;
Wang & Ran 2004).

Because the phytogeny of the family Amaurobiidae is
poorly understood, the relationships of Taira with other
amaurobiid genera are ambiguous. The inclusion of Taira and
Tamgrinia Lehtinen 1967 in the same tribe Tairini of the
subfamily Amaurobiinae by Lehtinen (1967) was questioned
by Wang (2000). After comparing the ultrastructures of Taira
and Tamgrinia with other amaurobiid genera [ Arctobius
Lehtinen 1967 (Arctobiinae), Amaurobius C.L. Koch 1337
(Amaurobiinae), Callobius Chamberlin 1947 (Amaurobiinae),
Pimus Chamberlin 1947 (Amaurobiinae), Coelotes Blackwall
134! (Coelotinae), and Rubrius Simon 1887 (Macrobuninae)],
Wang (2000) found that Taira is closer to Ambaurobius and
allied genera ( Callobius and Pimus) than Tamgrinia. The
ultrastructural characters shared by Taira and Amaurobius
and related genera that differ from those of Tamgrinia include:
the striped small hood of the trichobothria, the presence of
“amaurobiid ALS spigots” (named as “modified spigots” by
Griswold et al. 2005) and the presence of paracribellar spigots
on both the posterior lateral and posterior median spinnerets,
while Tamgrinia have smooth small trichobothrial hoods and
none of the spigots mentioned above (Wang & Ran 2004).

In this paper, the genus Taira is revised with particular
attention to the genital structures. Eight species are recognized
including five new species from China. The species Titanoeca
decorata Yin & Bao 200 1 is newly placed in this genus and its
male is described for the first time. Taira lunaris is
synonymized with T. liboensis.

METHODS

All specimens are preserved in 75% ethanol and were
examined, illustrated, and measured using a Tech XTL-II
stereomicroscope equipped with an Abbe drawing device. Eye
sizes are measured as the maximum diameter from either
above or in front. Leg measurements are shown as: total

1 Corresponding author. E-mail: mingshengzhu@263.net

length (femur, patella and tibia, metatarsus, tarsus). All
measurements are in millimeters.

The distribution map was generated using GIS ArcView
software and the .dbf files of the studied species are
downloadable from http://www.amaurobiidae.com, which is
published and maintained by Xin-Ping Wang.

The following abbreviations are used: Museum of Hebei
University, Baoding, China (MHBU); the Collection of the
Arachnological Society of Japan (CASI) at Otemon Gakuin
University, Osaka; Hunan Normal University, Changsha,
China (HNUC).

Abbreviations used in this study are: ALE = anterior lateral
eye; AME = anterior median eye; MOA = median ocular area;
PLE = posterior lateral eye; PME = posterior median eye;
DTA = dorsal tibial apophysis; RTA = retrolaterial tibial
apophysis.

TAXONOMY

Family Amaurobiidae Thorell 1870
Subfamily Amaurobiinae Thorell 1870
Taira Lehtinen 1967
Taira Lehtinen 1967:266, 340.

Type species. — Amaurobius flavidor salis Yaginuma 1964, by
original designation.

Diagnosis. — This genus is similar to Amaurobius , but can be
distinguished from the latter by the small RTA, the big
interior branch and the relatively small exterior branch of the
DTA (versus the small interior and big exterior branches of
the DTA), the absence of an embolic basal process (except T.
liboensis ) (versus a dorsally pointed process, also indicated by
Thaler & Knaflach (2000:341, figs. 19-22)), the big double
membranous conductor (versus a single membrane), the
scimitar-shaped median apophysis with the inflated base, the
big, anteriorly pointed tegular process of the male palp, the
wide epigynal teeth, and the retrolaterally curved fertilization
ducts of the female epigynum (Figs. 1-5).

Description. — Carapace pear-shaped, with many black
hairs. Cephalic area more or less elevated. Fovea longitudinal.
Cervical groove and radial furrow distinct. Anterior eye row
slightly retrocurved and posterior one procurved. MOA
trapeziform, longer than wide or as long as wide. Chilum
undivided. Chelicerae with distinct lateral condyles, 3 or 4

502

ZHANG ET AL.— REVISION OF TAIRA

503

Figures 1-5. — Amaurobius fenestralis (Strom 1768). Males and females from Denmark and Sweden (loaned from Zoological Museum,
University of Copenhagen, Denmark and Naturhistorska Riksmuseum, Stockholm, Sweden) 1. Female, epigynum. 2. Same, vulva. 3. Feft palp
of the male, prolateral view. 4. Same, ventral view. 5. Same, retrolateral view. Abbreviations: a = atrium; c = conductor; cd = copulatory duct; e
= embolus; ebp = embolic basal process; eDTA = exterior branch of dorsal tibial apophysis; eh = epigynal hood; fd = fertilization duct; iDTA
= interior branch of dorsal tibial apophysis; It = laterial teeth; ma = median apophysis; ml = median lobe; ta = tegular apophysis; s =
spermatheca; RTA = retrolateral tibial apophysis.

promarginal and 3 or 4 retromarginal teeth. The second
promarginal tooth largest. Legs slightly slender (especially
those of males), with 3 claws on the tarsi, many spines on
femora, tibiae, and metatarsi. Metatarsus IV with a single row
of setae making up the calamistrum (about half the length of
the metatarsus IV), calamistrum of male more or less reduced.
Leg formula: 1423 or 1243. Female palp with two claws on the
end of tarsi, one of the paired claws lost. Abdomen oval, with
some irregular black markings. Cribellum of female obviously
divided and narrower than the width of anterior lateral
spinnerets. Male cribellum somewhat reduced. Male palpal
tibia with a strong RTA and a branched DTA. The interior
branch of DTA well developed, sometimes with a groove on
the prolateral view of left palp. And the exterior branch of it
small, often depressed prolaterally. The cymbium modified
retrolaterally, with a slightly narrow base and a blunt process
pointing retrolaterally. Embolus wide, short, and flat, with
slightly thin apex. Conductor membranous, doubled. Median
apophysis scimitar-shaped, with an inflated base, originating
on a membranous area of tegulum posteriorly. Tegulum with
one or two big, anteriorly pointed processes. Female epigynum
with two small hoods, a wide median lobe, and a pair of

epigynal teeth (the projections of epigynal cuticle) originating
on the outside of median lobe. A pair of pores can be definitely
located on the dorsal view of epigynum. Copulatory ducts
lying between spermathecae and connected with each other.
Spermathecae ball-like. Fertilization ducts originating behind
the spermathecae and curved outwards.

Biology. — Species of this genus can be found under stones,
tree bark, and caves or in rock crevices. The web is small, with
one or more retreat(s) on the web mesally.

Distribution. Members of the genus Taira are restricted to
East Asia, including South China and Japan (Fig. 41).

Relationships with other amaurobiine genera. — In order to
evaluate the relationships of Taira and other amaurobiine
genera, some specimens of Amaurobius, Callobius, and
Tamgrinia were examined, as well as some literature sources
(e.g., Lehtinen 1967; Leech 1972; Wang 2000).

We support the conclusion of Wang (2000) that the genus
Taira is phylogenetically far from Tamgrinia and more similar
to Amaurobius , Callobius , and Pimus. Wang & Ran (2004)
indicated that the genus Taira can be distinguished from
Tamgrinia at least by the following characters: the undivided
chilum, the transversely striped cephalothorax, the large

504

THE JOURNAL OF ARACHNOLOGY

trichobothrial hoods, the presence of “modified spigots,” and
the paracribellar spigots. As far as genitai characters are
concerned, the differences between Taira , Amaurobius , Callo-
bius , and Pimus with Tamgrinia include: the absence of an
embolic supporter and tegular membranous apophysis, the
unsclerotized conductor of the male palp, the epigynal atrial split,
the absence of a septal sclerite, the connected copulatory ducts,
and the nearly ball-like spermathecae of the female epigynum.

The features shared by Taira, Amaurobius , Callobius, and
Pimus are: the striped small hood of the trichobothria, the
presence of “modified spigots,” and the presence of para-
cribellar spigots on both posterior lateral spinnerets and

posterior median spinnerets (Wang & Ran 2004). Another
structure shared by these genera is the undivided chilum.

After comparing the genital structures of Taira with
Amaurobius, Callobius, and Pimus, we regard Taira as most
similar to Amaurobius because of the following shared
characters: the presence of RTA and the branched DTA, the
short embolus, the membranous conductor, the posteriorly
originated median apophysis, the presence of a tegular process
on the male palp, the wide median lobe, the presence of
epigynal teeth, the nearly spherical shaped spermathecae
and the positions of the copulatory ducts of the female
epigynum.

KEY TO SPECIES OF THE GENUS TAIRA

1. Female 2

Male (those of T. obtusa and T. latilabiata unknown) 9

2. Median septum with width less than 2 times length (Figs. 7, 36) 3

Median septum with width at least 2 times length (Figs. 12, 17, 22, 27, 29, 34) 4

3. Median septum with the widest mid-part (Fig. 36) T. sulciforis

Median septum with the widest anterior edge (Fig. 7) T. flavidorsalis

4. Median septum with width about 4 times length (Fig. 12) T. cangshan

Median septum with width less than 3 times length (Figs. 17, 22, 27, 29, 34) 5

5. Median septum small, slightly pointed posteriorly (Fig. 17) T. concava

Median septum large, not pointed posteriorly (Figs. 22, 27, 29, 34) 6

6. Spermathecae small, widely separated (Fig. 28) T. latilabiata

Spermathecae big, narrowly separated (Figs. 23, 30, 35) 7

7. Epigynal teeth with wide posterior margin (Fig. 34) T. obtusa

Epigynal teeth with a tip end (Figs. 22, 29) 8

8. Spermathecae with sharp anterior margin (Fig. 23) T. decorata

Spermathecae with blunt anterior margin (Fig. 30) T. liboensis

9. DTA far away from tibia (Figs. 11, 21, 26, 33, 40) 10

DTA close to tibia (Fig. 16) T. cangshan

10. DTA with sharp exterior branch and normal interior one (Fig. 31) T. liboensis

DTA not with sharp exterior branch, with grooved interior one (Figs. 9, 19, 24, 38) 11

1 l.DTA with interior branch as long as wide (Fig. 38) T. sulciformis

DTA with interior branch much longer than wide (Figs. 9, 19, 24) 12

12. RTA big (Figs. 20, 21) T. concava

RTA small (Figs. 10, 25) 13

13. Embolic end wide (Fig. 25) T. decorata

Embolic end narrow (Fig. 10) T. flavidorsalis

Taira flavidorsalis (Yaginuma 1964)

Figs. 6-11, 41

Amaurobius flavidorsalis Yaginuma 1964:20, figs. 1-5; Yagi-
numa 1971:128, figs. 107.1-3; Yaginuma 1986:9, fig. 7.3;
Chikuni 1989:21, fig. 1.

Taira flavidorsalis (Yaginuma): Lehtinen 1967:266, figs. 204, 207.

Material examined. —Holotype: JAPAN: Hiroshima Prefec-
ture: male, Mt. Azuma (35°00'N, 133°00'E), 15 August 1963,
Y. Morinaga (CASJ).

Paratype: JAPAN: Tottori Prefecture: 1 female, Mt. Daisen
(35°12'N, 133°21'E), July 1937, M. Tanaka (CASJ).

Diagnosis. -This species is similar to T. concava (Figs. 17-
21), but can be distinguished from the latter by the small and
ventrally originated RTA, the narrow furrow of interior
branch of DTA, the less depressed exterior branch of DTA,
the thin apex of embolus, the small conductor, the slightly
inflated base of median apophysis, and the sunken apex of
tegular process of male palp. Also, the same relative shape of

the median lobe, the mesally situated copulatory ducts, and
the nearly elliptical spermathecae of the female epigynum.

Redescription. — Cheliceral promargin with three teeth,
retromargin with four.

Male palp (Figs. 9-11) with a small and ventrally originated
RTA and a big DTA. The interior branch of DTA long,
anterodorsally pointed, with a narrow furrow. The exterior
branch of DTA small, depressed prolaterally and pointed dorsally.
The base of cymbium narrow and modified retrolaterally.
Embolus originated prolaterally, short, flat and slightly curved,
with a thin apex. Conductor membranous and doubled. Median
apophysis situated posteriorly, with an acute tip and inflated base.
Tegular process close to the base of embolus, with a sunken apex.

Female epigynum (Figs. 7-8) with a big median lobe. Epigynal
teeth originated retrolaterally, with blunt apex. Copulatory ducts
small, between the spermathecae. Spermathecae nearly elliptical.

For further detail, see Yaginuma (1964).

Distribution. —This species has been found in Japan
(Hiroshima, Tottori) (Fig. 41).

ZHANG ET AL. REVISION OF TAIRA

505

iDTA

eDTA

Figures 6-1 1. — Taira flavidorsalis (Yaginuma 1964). 6. Male body, dorsal view. 7. Female, epigynum. 8. Same, vulva. 9. Left palp of the male,
prolateral view. 10. Same, ventral view. 11. Same, retrolateral view.

Remarks. — Judging from the publication (Wang & Ran
2004: 31, figs. 5-6), the specimen illustrated as T. flavidorsalis
is misidentified. Further examination of this specimen is badly
needed in order to determine its species status.

Taira cangshan new species
Figs. 12-16, 41

Material examined. — Holotvpe: CHINA: Yunnan : male,
Dali, Mt. Cangshan (25°58'N, 99°52'E), Yudailu, elev.
2400 m, 1 January 2004, Zi-Zhong Yang (MHBU).

Paratypes: CHINA: Yunnan : two females, the same location
as holotype, 24 October 2004, Hai-bo Pu and Zi-Zhong Yang
(MHBU).

Other material examined: 7 ?, Lushui, Pianma (26°01'N,
98°37'E), elev. 2200 m, Yunnan, 9 May 2004, Zhi-Sheng
Zhang and Zi-Zhong Yang (MHBU).

Etymology.-— The specific name refers to the type locality
and is a noun in apposition.

Diagnosis. — The male palpal tibial apophyses of the new
species are different from other Taira species: RTA small,
DTA located distally. But the palpal organ and epigynum are
similar to the others. It can be distinguished from the type
species, T. flavidorsalis (Figs. 7-11) by the blunt apex of
embolus, the acute tip of median apophysis, the bifurcated
tegular apophysis of male palpal bulb; the epigynal median
lobe that is four times wider than long and procurved, the wide
base of the epigynal teeth, the slender copulatory ducts and the
small spermathecae.

Description. — Male (holotype): Total length 6.43: prosoma
3.16 long, 2.24 wide; opisthosoma 3.57 long, 2.50 wide.
Prosoma yellow, with yellow brown anterior and retrolateral

areas. Eye sizes and interdistances: AME 0.15, ALE 0.15,
PME 0.13, PLE 0.15; AME-AME 0.05, AME-ALE 0.10,
PME-PME 0.20, PME-PLE 0.25, ALE-PLE 0.08. MOA 0.50
long, front width 0.33, base width 0.45. Clypeus height 0.25.
Chelicerae brown, with 4 promarginal and 3 retromarginal
teeth. Labium brown. Endites yellow brown. Sternum deep
yellow. Leg measurements: I 14.28 (4.59, 4.79, 3.37, 1.53), II
9.49 (2.86, 3.37, 2.14, 1.12), III 7.86 (2.45, 2.65, 1.84, 0.92), IV
9.58 (2.75, 3.26, 2.45, 1.12). Leg formula: 1423. Opisthosoma
gray yellow, with indistinct 3 chevron-like markings dorsally.
Palp (Figs. 14-16) with a small RTA, located near the base of
tibia. DTA relatively small, close to the anterior part of tibia
dorsally. Interior branch short and strong, with a bifurcated
apex. Exterior branch small, with an acute tip, pointing
anterodorsally. The anterior part of embolus relatively thin.
Conductor membranous and doubled. Median apophysis with
a bulbous base and arcuate anterior part. Tegulum bifurcated,
pointing antero-retrolaterally.

Female: Total length 6.83-8.77. A female (one of the
paratypes) total length 8.06: prosoma 3.57 long, 2.55 wide;
opisthosoma 4.79 long, 3.26 wide. Eye sizes and interdistances:
AME 0.15, ALE 0.1 8^ PME 0.13, PLE 0.15; AME-AME 0.08,
AME-ALE 0.20, PME-PME 0.28, PME-PLE 0.38, ALE-PLE
0.10. MOA 0.55 long, front width 0.38, back width 0.53. Clypeus
height 0.33. Leg measurements: 1 10.51 (2.96, 3.67, 2.45, 1.43), II
8.47 (2.45, 3.06, 1.84, 1.12), III 7.24 (2.24, 2.45, 1.63, 0.92), IV
9.07 (2.65, 3.16, 2.24, 1.02). Leg formula and other characters as
the male holotype. Epigynum (Figs. 12, 13) with a wide median
lobe, 4 times wider than long and curved posteriorly. Epigynal
teeth with wide base. Copulatory ducts slender, between
spermathecae. Spermethecae ball-like, widely separated.

506

THE JOURNAL OF ARACHNOLOGY

Figures 12-16. — Taira cangshan new species. 12. Female, epigynum. 13. Same, vulva. 14. Left palp of the male, prolateral view. 15. Same,
ventral view. 16. Same, retrolateral view. Scale lines: 0.2 mm.

Distribution. — Taira cangshan is found in China (Yunnan)
(Fig. 41).

Habitat. — Specimens have been found under stones, with a
small irregular web.

Taira concava new species
Figs. 17-21, 41

Material examined.- - Holotype: CHINA: Sichuan: male,
Mt. Emei (29 32'N, 103°19'E), elev. ~ 1600 m, 16 September
2004, matured at the end of October 2004, Zhi-Sheng Zhang
(MHBU).

Paratype: CHINA: Sichuan: 1 female, collected with
holotype (MHBU).

Etymology. -The specific name is Latin, meaning depressed
and refers to the depressed exterior branch of DTA.

Diagnosis. -The new species is similar to T. sulciformis
(Figs. 36-40), but can be distinguished from the latter by the
narrow interior branch of DTA, obviously depressed exterior
branch of DTA, the relatively thin and long median apophysis,
the blunt tegular apophysis of male palp; the wider than long
median lobe and the short copulatory ducts of female epigynum.

Description. -Male holotype: Total length: 4.74, prosoma
2.19 long, 1.58 wide; opisthosoma 2.70 long, 1.79 wide.
Prosoma yellow, with many tiny black markings. Eye sizes and

interdistances: AME 0.08, ALE 0.13, PME 0.10, PLE 0.10.
AME-AME 0.10, AME-ALE 0.10, PME-PME 0.13, PME-
PLE 0.20, ALE-PLE 0.05. MOA 0.33 long, front width 0.25,
back width 0.33. Clypeus height 0.20.

Chelicerae yellow brown, with 4 promarginal and 3 retro-
marginal teeth. Endites deep yellow, labium dark yellow.
Sternum dark yellow. Leg measurements: I 10.51 (2.75, 3.67,
2.81, 1.28), II 7.85 (2.24, 2.65, 1.99, 0.97), III 6.23 (1.84, 2.04,
1.58, 0.77), IV 7.55 (2.19, 2.45, 1.99, 0.92). Leg formula: 1243.
Opisthosoma grey brown, with a pair of sigilla and indistinct 3
chevron-like markings. Palpal tibia with a big, ball-like RTA
and wide DTA. The interior branch of DTA long and thin,
with a furrow and the exterior branch of it wide, with an
obvious pit. The base of cymbium narrow and modified.
Embolus with thin apex. Conductor big, membranous and
doubled. Median apophysis slender, with an inflated base.
Tegulum with a blunt process (Figs. 19-21).

Female paratype: Total length 5.30: prosoma 2.30 long, 1.58
wide; opisthosoma 3.16 long, 2.30 wide. Eye area too badly
damaged to be measured. Leg measurements: I 6.07 (1.73,
2.14, 1.38, 0.82), II 4.84 (1.43, 1.68, 1.07, 0.66), III 4.18 (1.27,
1.43, 0.92, 0.56), IV 5.25 (1.58, 1.84, 1.17, 0.66). Leg formula:
1423. Epigynum with a small median lobe that is far wider
than long. Epigynal teeth widely separated, with blunt apex.

ZHANG ET AL. -REVISION OF TAIRA

507

Figures 17-21. — Taira concava new species. 17. Female, epigynum. 18. Same, vulva. 19. Left palp of the male, prolateral view. 20. Same,
ventral view. 21. Same, retrolateral view. Scale lines: 0.2 mm.

Copulatory ducts thin and short. Spermathecae ball-like
■(Figs. 17, 18).

Distribution. — This species is found in China (Sichuan)
(Fig. 41).

Habitat. — Specimens have been found within crevices of
cliffs or under rocks.

Taira decor at a (Yin & Bao 2001) new combination
Figs. 22-26, 41

Titanoeca decorata Yin & Bao 2001:60, figs. 2a-e.

Type specimens. — Holotype: CHINA: Hunan: female, Mt.
Mangshan, Yizhang County, (24°59'N, 1 12°50'E), 18 May
1999, Jian-Hui Yang (HNUC) (not examined).

Material examined. — CHINA: Fujian: 1 <$, 3 9, Moshikeng,
Mt. Wuyi (26°54'N, 1 16°42'E), 20 May 2004, Feng Zhang
(MHBU); 1 d, 8 ?, between Guadun and Tongmu, Mt. Wuyi,
24 May 2004, Feng Zhang (MHBU); 6 9, Guadun, Mt. Wuyi,
23 May 2004, Feng Zhang (MHBU); 3 ?, Mt. Wuyi, 10 July
1986, Ming-Sheng Zhu (MHBU).

Diagnosis. — This species is similar to T. flavidor salis
(Figs. 7-11), but can be distinguished from the latter by the
narrow sulciform interior branch of DTA, the slightly
bifurcated embolus, the prolaterally outstanding median part
of median apophysis, the blunt tegulum of male palp; the wide
median lobe of epigynum, the small epigynal teeth, the

epigynal hood far away from the epigynal teeth and the
slightly peaked spermathecae anteriorly.

Description. -Male: Total length: 4.18^4.44. One male total
length 4.44: prosoma 2.40 long, 1.79 wide; opisthosoma 2.14
long, 1.53 wide. Prosoma yellow, with dark yellow Cephalic area.
Eye sizes and interdistances: AME 0.10, ALE 0.15, PME 0.13,
PLE 0. 1 5; AME-AME 0.08, AME-ALE 0. 1 0, PME-PME 0. 1 5,
PME-PLE 0.20, ALE-PLE 0.08. MOA 0.43 long, front width
0.28, back width 0.40. Clypeus height 0.13. Chelicerae yellow
brown, with 4 promarginal and 3 retromarginal teeth. Labium
and endites deep yellow. Sternum yellow, with deep yellow lateral
margin. Leg measurements: I 1 1.63 (3.1 1, 4.08, 3.11, 1.33), II 7.75
(2.24, 2.70, 1.99, 0.82), III 6.22 (1.89, 2.04, 1.58, 0.71), IV 7.70
(2.19, 2.60, 2.09, 0.82). Leg formula: 1243. Opisthosoma
yellowish, with many white hairs and many irregular markings,
a pair of yellow longitudinal markings and indistinct 4 chevron-
like markings. Cardiac markings red. Palpal tibia with a small
RTA, located on the anterior part of tibia. The interior branch of
DTA with a narrow groove and the exterior branch depressed in
prolateral view. Embolus bifurcated anteriorly. Conductor
membranous and doubled. Median apophysis inflated in the
middle part. Tegular process blunt (Figs. 24-26).

Female: Epigynum with wide median lobe. Epigynal teeth
close to the lateral margins of median lobe. Copulatory ducts
thin and short. Spermathecae peaked anteriorly. Fertilization

508

THE JOURNAL OF ARACHNOLOGY

Figures 22-26. — Taira decorata (Yin & Bao 2001). 22. Female, epigynum. 23. Same, vulva. 24. Left palp of the male, prolateral view. 25.
Same, ventral view. 26. Same, retrolateral view. Scale lines: 0.2 mm.

ducts originating on the spermathecae posteriorly and curved
retrolaterally (Figs. 22, 23).

For further details, see Yin & Bao (2001).

Distribution. -Taira decorata has been found in China
(Zhejiang, Fujian and Hunan) (Fig. 41).

Remarks. — This species was first described by Yin & Bao
(2001) based on female specimens collected from Mt. Mangshan
of Hunan Province and Mt. Fengyuan (28°06'N, 119°06'E) of
Zhejiang Province and placed in the genus Titanoeca of the
family Titanoecidae. After comparing the original figures with
the specimens we collected from Mt. Wuyi of Fujian, we have
determined that they are the same species though the type
specimens are unavailable. Moreover, Mt. Wuyi is situated
between Mt. Mangshan and Mt. Fengyuan.

Taira latilabiata new species
Figs. 27, 28, 41

Material examined. — Holotype: CHINA: Guizhou: female,
Weining, Chaohai National Natural Reserve (26°51'N,
104°15'E), 2 July 2005, Hui-Ming Chen (MHBU).

Paratvpes: CHINA: Guizhou: 6 females, collected with
holotype (MHBU).

Etymology. — The specific name, derived from Latin adjec-
tive “ latilabiatus refers to the wide and labiate median lobe
of the female epigynum.

Diagnosis. — The new species is similar to T. liboensis
(Figs. 29, 30), but differs from the latter by the wide epigynal
median lobe, the long copulatory ducts and the relatively small
spermathecae.

Description. — Female: total length 7.04—7.75. Holotype
total length 7.75: prosoma 3.47 long, 2.35 wide; opisthosoma
4.39 long, 3.06 wide. Prosoma brown, with many black hairs
and a “V” shaped yellow marking in front of the fovea and
behind MOA. Eye sizes and interdistances: AME 0.15, ALE
0.18, PME 0.15, PLE 0.18; AME-AME 0.08, AME-ALE
0.13, PME-PME 0.20, PME-PLE 0.30, ALE-PLE 0.13.
MOA 0.55 long, front width 0.35, back width 0.50. Clypeus
height 0.20. Chelicerae brown, with 4 promarginal and 3 or 4
retromarginal teeth. Labium yellow brown. Sternum deep
yellow. Legs yellow, with 2 blackish annular markings on

ZHANG ET AL. REVISION OF TAIRA

509

Figures 27, 28. Taira latilabiata new species. 27. Female, epigynum. 28. Same, vulva. Scale lines: 0.2 mm.

femora. Leg measurements: I 9.79 (2.75, 3.47, 2.14, 1.43), II
7.96 (2.35, 2.86, 1.63, 1.12), III 7.14 (2.24, 2.45, 1.53, 0.92), IV
9.07 (2.65, 3.16, 2.14, 1.12). Leg formula: 1423. Gpisthosoma
grey black, with yellowish markings and indistinct 3 chevron-
like markings dorsally. Epigynum with a wide labiate median
lobe and a pair of epigyna! teeth. Two pairs of epigynal hoods
situated in front of epigynal teeth. Copulatory ducts located
between spermathecae and connected to each other. Sper-
mathecae ball-like (Figs. 27, 28).

Male: unknown.

Distribution. — Taira latilabiata is known only from Guiz-
hou province in China (Fig. 41).

Taira liboensis Zhu, Chen & Zhang 2004
Figs. 29-33, 41

Taira liboensis Zhu, Chen & Zhang 2004 (January):61, figs.

1 A-F.

Taira lunaris Wang & Ran 2004 (March):31, figs. 1-4 (female

holotype and paratype from Yueliang Cave, Libo, Guizhou,

China, deposited in the Institute of Zoology, Beijing, China,

not examined). New synonymy.

Material examined. - Holotype : CHINA: Guizhou : female,
Shuijiang Cave, Libo (25°24'N, 107°52'E), 8 July 2001, Hui-
Ming Chen (MHBU).

Paratypes: CHINA: Guizhou : 3 males, 2 females, same data
as holotype (MHBU).

Other material: CHINA: Guizhou : 1 ?, Shuijiang Cave,
September 1998, Hui-Ming Chen (MHBU); 3 <5, 1 9, Heshang
Cave, Guiyang (26°35'N, 106°42'E), 13 May 1998, Hui-Ming
Chen (MHBU); Sichuan : 1 <J, 2 9, Mt. Emei (29°32'N,
103°19'E), elev. ~ 800 m, 16 September 2004, Zhi-Sheng
Zhang (MHBU).

Diagnosis. — This species can be easily distinguished from
the other Taira species by the indistinct RTA, the peaky
interior branch of the DTA, the undepressed exterior branch
of the DTA, the presence of the embolic process and two
tegular processes of the male palp. It differs from T.
flavidorsalis (Yaginuma 1964) (Figs. 7-11) by the wide apex
of the male palpal embolus, the wide median lobe, and the
long distances between the epigynal teeth and median lobe.

Redescription. — Cheliceral promargin with four teeth, retro-
margin with three.

Male: palp (Figs. 31-33) with indistinct RTA. Interior
branch of DTA peaky, without groove, a small process located
near the base of DTA in the prolateral view. Exterior branch
of DTA small. Embolus with a basal process. Conductor
membranous and doubled. Median apophysis thin, relatively
long, arcuate curved, with an acute tip and an inflated base.

Tegulum with two processes near the apex of median
apophysis.

Female: epigynum (Figs. 29, 30) with a brownish labiate
median lobe. Epigynal teeth with wide base and slightly peaky
tip. Two pairs of hoods situated in front of the epigynal teeth.
Copulatory ducts short, connected to each other. Spermathe-
cae nearly elliptical.

For further details, see Zhu et al. (2004).

Distribution. — Taira liboensis has been found in China
(Guizhou, Sichuan) (Fig. 41).

Remarks. — Taira liboensis and T. lunaris were both pub-
lished in 2004 in different journals. The former was published
in January 2004 (indicated in the page header of the printed
paper), while the latter, published in the Bulletin of the British
Arachnological Society was published in March 2004 (personal
communication with Ian Dawson, the Secretary of the British
Arachnological Society). Therefore, the latter is a junior
synonym.

Taira obtusa new species
Figs. 34, 35, 41

Material examined. — Holotype: CHINA: Hubei : female,
Shennongjia National Nature Reserve (31°28'N, 110°22'E),
Yazikou, 4 September 2004, Zhi-Sheng Zhang & Hui-Ming
Chen (MHBU).

Paratypes: 11 females, collected with holotype (MHBU);
Guizhou : 2 females, Mt. Fanjin (27°55'N, 108°41'E), Huix-
iangpin to Yu’ao, 2 August 2001, Jun-Xia Zhang & Zhi-Sheng
Zhang (MHBU).

Etymology.- -The specific name comes from the Latin word
“ obtusus,” meaning “blunt,” referring to the blunt end of the
epigynal teeth.

Diagnosis. -The female of the new species is similar to that
of T. decorata (Yin & Bao 2001) (Figs. 22, 23), but can be
distinguished from the latter by the anteriorly and laterally
extended median lobe of epigynum, the big blunt epigynal
teeth, the wide copulatory ducts and the shape of the
spermathecae.

Description. -Female: total length 6.50-7.50. Holotype
total length 6.83: prosoma 3.06 long, 1.94 wide; opisthosoma
4.08 long, 2.96 wide. Prosoma yellow brown with dark brown
lateral margins. Eye sizes and interdistances: AME 0.10, ALE
0.15, PME 0.13, PLE 0.15; AME-AME 0.08, AME-ALE
0.18, PME-PME 0.20, PME-PLE 0.25, ALE-PLE 0.08.
MOA 0.48 long, front width 0.30, back width 0.43. Clypeus
height 0.20. Chelicerae brown, with 4 promarginal and 3
retromarginal teeth. Endites dark yellow, labium brown.
Sternum yellow, with yellowish margin. Leg measurements: I

510

THE JOURNAL OF ARACHNOLOGY

Figures 29-33. — Taira liboensis Zhu, Chen & Zhang 2004. 29. Female, epigynum. 30. Same, vulva. 31. Left palp of the male, prolateral view.
32. Same, ventral view. 33. Same, retrolateral view. Scale lines: 0.2 mm.

8.47 (2.35, 2.96, 2.04, 1.12), II 6.73 (2.04, 2.24, 1.53, 0.92), III
5.51 (1.73, 1.84, 1.33,0.61), IV 7. 14 (2.04, 2.55, 1.73,0.82). Leg
formula: 1423. Opisthosoma yellowish, with many irregular
black markings, a pair of cardiac markings and four indistinct
chevron-like markings. Epigynum (Figs. 34-35) with antero-
laterally extended trapeziform median lobe. Epigynal teeth with
narrow base and wide body. Copulatory ducts located between
spermathecae posteriorly. Spermathecae narrow anteriorly.

Male: unknown.

Distribution. Taira obtusa is found in China (Hubei,
Guizhou) (Fig. 41).

Habitat. — Specimens have been found within crevices of
rocks.

Taira sulciformis new species
Figs. 36^10, 41

Material examined. — Holotype: CHINA: Fujian : male, Mt.
Wuyi (26°54'N, 116°42'E), Tongmu, 16 May 1985, Ming-
Sheng Zhu (MHBU).

Paratypes: CHINA: Fujian : 1 male, 4 females, collected with
holotype (MHBU); 1 female, Mt. Wuyi, Tongmu, 19 July
2003, Chao Zhang (MHBU).

Figures 34, 35. — Taira obtusa new species. 34. Female, epigynum. 35. Same, vulva. Scale lines: 0.2 mm.

ZHANG ET AL. REVISION OF TAIRA

511

Figures 36M0. — Taira sulciformis new species. 36. Female, epigynum. 37. Same, vulva. 38. Left palp of the male, prolateral view. 39. Same,
ventral view. 40. Same, retrolateral view. Scale lines: 0.2 mm.

Etymology. — The specific name is a combination of “ sulc
and “-for mis” The former means “groove,'’ and the latter
means “with ... shape.” The combination refers to the obvious
wide and big groove of the interior branch of DTA.

Diagnosis. — The new species is similar to T. concava
(Figs. 17-21), but can be distinguished from the latter by the
wide interior branch of DTA, indistinct depression of exterior
branch of DTA prolaterally, the ventraily curved apex of
embolus, the short median apophysis, the wide and flat tegular
apophysis of the male palp; the longer median lobe, the wide
and anteriorly curved copulatory ducts of female epigynum.

Description. — Male: total length 6.02-6.22. Holotype total
length 6.22: prosoma 3.26 long, 2.24 wide; opisthosoma 3.26
long, 2.14 wide. Prosoma deep yellow. Eye sizes and
interdistances: AME 0.15, ALE 0.18, PME 0.15, PLE 0.15.
AME-AME 0.13, AME-ALE 0.15, PME-PME 0.23, PME-
PLE 0.33, ALE-PLE 0.08. MOA 0.48 long, front width 0.38,
back width 0.50. Clypeus height 0.23. Chelicerae red brown,
with 4 promarginal and three retromarginal teeth. Endites and
labium yellow. Sternum yellowish. Leg measurements: I 16.52
(4.28, 5.51, 4.69, 2.04), II 1 1.94 (3.37, 4.08, 3.06, 1.43), III 9.28
(2.75, 2.96, 2.45, 1.12), IV 11.72 (3.26, 3.98, 3.26, 1.22). Leg

formula: 1243. Opisthosoma yellowish, with many black
irregular markings, a pair of yellow longitudinal markings
and four indistinct yellowish chevron-like markings. Palpal
tibia with a dorso-lateral RTA. Interior branch of DTA wide,
with an obvious groove prolaterally. Exterior branch of DTA
small, blunt, and slightly depressed prolaterally. Apex of
embolus thin and curved ventraily. Conductor membranous
and doubled. Median apophysis relatively short, with an acute
tip and an inflated base. Apex of tegular apophysis wide and
flat (Figs. 38-40).

Female: total length 8.87-1 1.53. A female (one of paratypes)
total length 1 1.53: prosoma 5.30 long, 3.67 wide; opisthosoma
6.53 long, 4.39 wide. Eye sizes and interdistances: AME 0.23,
ALE 0.23, PME 0.18,' PLE 0.20. AME-AME 0.18, AME-
ALE 0.35, PME-PME 0.43, PME-PLE 0.63, ALE-PLE 0.13.
MOA 0.83 long, front width 0.60, back width 0.78. Clypeus
height 0.40. Leg measurements: I 13.88 (4.08, 4.49, 3.47, 1.84),
II 11.83 (3.57, 4.08, 2.75, 1.43), III 10.10 (3.06, 3.57, 2.35,
1.12), IV 12.55 (3.57, 4.49, 3.16, 1.33). Leg formula: 1423.
Epigynum (Figs. 36, 37) with the long median lobe and a pair
of blunt epigynal teeth. Copulory ducts curved anteriorly.
Spermathecae ball-like.

512

THE JOURNAL OF ARACHNOLOGY

Figure 41. — Distribution of eight Taira species.

Distribution. — Taira sulciformis has been found in China
(Fujian) (Fig. 41 ).

ACKNOWLEDGMENTS

We are grateful to Takahide Kamura (CASJ) for the loan
of type specimens, Nikolaj Scharff (Zoological Museum,
University of Copenhagen, Denmark) and Torbjorn Kro-
nestedt (Naturhistorska Riksmuseum, Stockholm, Sweden)
for the loan of valuable specimens. Thanks also due to our
colleagues of Flebei University (Feng Zhang, Hui-Ming
Chen, Jun-Xia Zhang, and Chao Zhang) and Zi-Zhong Yang
(Dali University, China) for collecting valuable specimens.
Xinping Wang (University of Florida, Gainesville, Florida,
USA) and Hirotsugu Ono (National Science Museum,
Tokyo) read the manuscript. This study was supported by
the National Natural Science Foundation of China (NSFC
No. 30499341, to Ming-Sheng Zhu) and Chinese Special
Basic Scientific Research Project (2006FY120100, to Zhi-
Sheng Zhang).

LITERATURE CITED

Blackwall, J. 1841. The difference in the number of eyes with which
spiders are provided proposed as the basis of their distribution into
tribes; with descriptions of newly discovered species and the
characters of a new family and three new genera of spiders.
Transactions of the Linnean Society of London 18:601-670.
Chamberlin, R.V. 1947. A summary of the known North American
Amaurobiidae. Bulletin of the University of Utah, Biological
Sciences 38(8): 1-3 1 .

Chikuni, Y. 1989. Pictorial Encyclopedia of Spiders in Japan. Kaisei-
sha Publishing Company, Tokyo. 310 pp.

Griswold, C.E., M.J. Ramirez, J.A. Coddington & N.I. Platnick.
2005. Atlas of phylogenetic data for entelegyne spiders (Araneae:
Araneomorphae: Entelegynae) with comments on their phylogeny.
Proceeding of the California Academy of Sciences, 4th Series.
Volume 56. Supplement II, 1-324.

Koch. C.L. 1837. Ubersicht des Arachnidensystems. Niirnberg, Heft
1, 39 pp.

Leech, R.E. 1972. A revision of the Nearctic Amaurobiidae
(Arachnida: Araneida). Memoirs of the Entomological Society of
Canada 84:1-182.

Lehtinen, P.T. 1967. Classification of the cribeilate spiders and some
allied families, with notes on the evolution of the suborder
Araneomorpha. Annales Zoologici Fennici 4:199-468.

Simon, E. 1887. Araclmides. In Mission scientifique du Cap Horn,
1882-1883. Paris: Tome 6-Zoologie, E1-E42.

Thaler, K. & B. Knofiach. 2000. A superspecies in the genus
Amaurobius on Crete, and additional records from Greece (Araneae:
Amaurobiidae). Pp. 337-344. In European Arachnology 2000:
Proceedings of the 19th European Colloquium of Arachnology. (S.
Toft & N. Scharff, eds.). Aarhus University Press, Aarhus, Denmark.

Thorell, T. 1870. On European spiders. Nova Acta Regiae Societatis
Scientarium. Upsaliensis 7:109-242.

Wang, X.P. 2000. A revision of the genus Tamgrinia (Araneae:
Amaurobiidae), with notes on amaurobiid spinnerets, tracheae and
trichobothria. Invertebrate Taxonomy 14:449^164.

Wang, X.P. & J.C. Ran. 2004. On the spider genus Taira (Araneae,
Amaurobiidae). Bulletin of the British Arachnological Society
13:31-32.

Yaginuma, T. 1964. A new spider of Amaurobius (= Ciniflo ) from
Hiroshima Prefecture, Japan. Miscellaneous Report of the Hiwa
Museum for Natural History 7:20-23.

Yaginuma, T. 1971. Spiders of Japan in Color (Enlarged and revised
edition). Hoikusha, Osaka. 197 pp.

Yaginuma, T. 1986. Spiders of Japan in Color (New edition).
Hoikusha Publishing Company, Osaka. 305 pp.

Yin, C.M. & Y.H. Bao. 2001. Two new species of the family Titanoecidae
from Hunan Province (Arachnida: Araneae). Journal of Changde
Teachers University (Natural Science Edition) 1 3(3):58— 61 .

Zhu, M.S., H.M. Chen & Z.S. Zhang. 2004. A new species of the
genus Taira from China (Araneae: Amaurobiidae: Amaurobiinae).
Journal of Hebei University (Natural Science Edition) 24:61-64.

Manuscript received II July 2007, revised 18 February 2008.

2008. The Journal of Arachnology 36:513-517

Subtle pedipalp dimorphism: a reliable method for sexing Juvenile spiders

Nagissa Mahmoudi, Maria Modanu, Yoni Brandt1, and Maydianne C. B. Andrade2: Integrative Behaviour & Neuroscience
Group, Department of Biological Sciences, University of Toronto Scarborough, 1265 Military Trail, Scarborough,
Ontario MIC 1A4, Canada.

Abstract. Quantifying primary sex ratios is necessary for studies in a wide range of areas including adaptive sex ratio
modification, population demography, and sex-biased developmental mortality. Adult and penultimate male spiders are
easy to sex, due to the great thickening of the male pedipalps, which are used for delivering sperm to the female
reproductive tract. However, in many spider species, males and females are apparently monomorphic at hatching, are
difficult to rear, and cannibalize their siblings, making assessment of primary sex ratios problematic. One technique for
sexing spiders is karyotyping, but this can be challenging and time-consuming, particularly for species with high fecundity,
and often requires destructive sampling. Here we report that, although apparently monomorphic, early-instar juveniles of
two species of black widow spiders (Latrodectus hasselti Thorell 1870 and Latrodectus Hesperus Chamberlin & Ivie 1935)
can be sexed reliably. Palp width measurements are significantly different for males and females at the 3rd instar, with the
palpi of juvenile females thinner than those of males. Moreover, sex identification with 89-100% accuracy can be achieved
by an experienced observer visually inspecting the palpi of 3rd instar spiderlings under a dissecting microscope. Our results
suggest that minimal investment in a pilot study can yield an accurate method for sexing juvenile spiders in the laboratory
or field. The suitability of this method should be examined in other species with apparently monomorphic spiderlings,
particularly those in which adult males have significantly enlarged palpi.

Keywords: Latrodectus , black widow spiders, monomorphic spiders, sexual dimorphism, sex identification

Assessing primary sex ratios, or sex ratios of juveniles
observed in the field, is desirable for study in a variety of areas
(e.g., adaptive sex ratio modification, population phenology,
population dynamics and development of sex-specific growth
patterns or behavior). For example, adaptive sex ratio
modification has been demonstrated in a variety of species
(Austad & Sunquist 1986; Emlen 1997; Nager et al. 1999;
Cameron et al. 1999; West & Sheldon 2002; West et al. 2002),
including social spiders (e.g., Aviles & Maddison 1991).
However in non-social species, there is continued speculation
about the possible importance of sex ratio modification since
this has only been demonstrated in one non-social species,
Pityohyphantes phrygianus (C.L. Koch 1836) (Gunnarsson &
Andersson 1992; Uhl & Gunnarsson 2001).

Primary sex ratios (ratio of maleTemale at hatching) are
difficult to determine in many spiders in which males and
females are apparently monomorphic at hatching (Foelix
1996). Karyotyping has been used to determine sex ratios of
fertilized eggs (Aviles & Maddison 1991), and this eliminates
the need to rear spiders to later instars. However, this method
does not work for all species (Watt, Hasenkampf & Andrade,
unpublished), may not be practical for field studies, requires
destructive sampling of eggs, and may be prohibitively time-
consuming for species with high fecundity. Rearing spiderlings
to advanced instars or adulthood to assess sex is problematic
because there may be variable success in rearing juveniles due
to challenging diet requirements and sibling cannibalism. Even
in species that show pronounced sexual dimorphism as adults,
juveniles may be apparently monomorphic to the naked eye
until later instars when male and female body shape or size
diverge significantly and the male's palpi thicken dramatically

'Current address: Department of Biology, University of South
Dakota, Vermillion, South Dakota 57069, USA.

2 Corresponding author. E-mail: mandrade@utsc.utoronto.ca

in the penultimate instar (Kaston 1970; Mahmoudi, Jovovic,
Andrade & Brandt, unpublished). Nonetheless, a number of
authors have noted that the pedipalps of males thicken
gradually over several instars prior to the penultimate instar
(e.g., Latrodectus - Bhatnagar & Rempel 1962; Cyrtophora -
Berry 1987).

Our study describes a method of assessing spiderling sex
using the relative size of the palpi in early-instar spiders of two
species of black widows (Latrodectus hasselti Thorell 1870 and
Latrodectus hesperus Chamberlin & Ivie 1935). In this genus,
males and females appear to be monomorphic at hatching,
although they show extreme reversed sexual size dimorphism
at adulthood (Kaston 1970). Initial studies of development in
L. hesperus by one of us ( Y. Brandt) suggested there might be
differences in the dimensions of pedipalps of males and
females very early in development. The objective of this study
was to determine whether early sex differences in palp
dimensions could provide a simple, non-destructive, morphol-
ogy-based method for sexing spiderlings. We examined early-
instar spiderlings of both species and predicted their sex based
on visual inspection and measurement of pedipalp width. We
then reared spiderlings to adulthood and report this method
yields 94% accuracy or better in determining the sex of 3rd
instar spiderlings.

METHODS

Spiders for this study were acquired from outbred
populations of redback spiders (Latrodectus hasselti) and
western black widow spiders (L. hesperus), that originated
from field-mated, adult females collected in New South Wales,
Australia (2002) and San Diego, California (2007), respective-
ly. Voucher specimens have been deposited in the Entomology
collection of the Royal Ontario Museum (L. hasselti:
ROMEnt 112065 through 112067; L. hesperus: ROMEnt
112068 through 112070). Egg sacs (L. hasselti, n = 4 sacs; L.

513

514

THE JOURNAL OF ARACHNOLOGY

A B

Figure 1 . -Digital photographs of dorsal view of portion of the cephalothorax and pedipalps of spiderlings of L. hasselti (4th instar female [A]
and male [B] ) and L. hesperus (3rd instar female [C] and male [D] ) showing location of measurement of palpal tibia width (black line). Scale bars
are 0.50 mm.

hesperus , n = 2 sacs) were removed from females’ cages shortly
after production. Each egg sac was cut open, 40-75 unhatched
eggs were randomly chosen and each egg was transferred into
an individual clear plastic cage measuring 2.3 X 3.0 x 2.3 cm
(Arnac Plastic Products Ltd). Eggs and spiders were kept in a
controlled-environment room at 25 ± 5° C, 12 hr light : 12 hr
dark cycle throughout the study.

Cages were examined every other day, and hatch dates and
molt dates were noted. Normally, 1st instar spiderlings remain
inside the egg sac and survive on yolk reserves until they molt,
emerge from the sac, and begin to take prey (Foelix 1996).
Beginning at the 2nd instar, the spiders were fed 1-2 small,
wingless fruit Hies (Drosophila melanogaster) three times each
week. The number of fruit flies each spider received was
gradually increased as the spider developed. After each molt,
approximately 2 more fruit flies were added to each feeding. By
the 4th instar, both males and females began receiving larger,
heavy-bodied fruit Hies (D. hydei) three times each week.

Preliminary work suggested that pedipalps of juvenile males
might be wider than those of juvenile females, particularly at

the palpal tibia (Jovovic, Brandt & Andrade, unpublished;
e.g., Fig. 1A, B). To test whether the appearance of the palps
reliably indicated sex differences, an experienced observer
examined the pedipalps of each spiderling under a dissecting
microsope or examined photographs of the palpi and
predicted its sex based on overall pedipalp girth (L. hasselti'.
N. Mahmoudi; L. hesperus'. M. Modanu). To determine the
accuracy of these predictions, all spiders were reared until the
5th (L. hasselti) or 6th instar ( L . hesperus ), at which time males
and females can be reliably distinguished by the development
of notable swellings of the pedipalp in penultimate instar
males (Bhatnagar & Rempel 1962; Kaston 1970).

For the visual examination predictions, intact spiderlings
were placed under a dissecting microscope (Zeiss Stemi 2000C)
and examined, or photographs were taken of their palps (see
below). Palps were scored categorically as thick (male-like) or
thin (female-like). For L. hasselti , sex was determined from 1
(n = 30) or 3 (n = 58) independent viewings of individuals at
the 3rd instar. For L. hesperus, there were 3 independent
viewings of individuals at the 2nd instar, then 3 independent

MAHMOUDI ET AL.— RELIABLE METHOD FOR SEXING SPIDERLINGS

515

Table 1. — Accuracy of sex identification based on visual examination of palpal tibia width of spiderlings for two species of Latrodectus
spiders. *Nuil expectation = 50% accuracy; § two-tailed test, df = 1 with Yates’ correction.

Species Instar # Correct # Incorrect % accuracy y2* T§

L. hesperus 2nd 83 52 61.5% 3.143 0.0762

L. hesperus 3rd 94 6 94.0% 45.858 < 0.0001

L. hasselti 3rd 78 10 88.6% 29.093 < 0.0001

viewings of photographs of the spider’s palps at the 3rd instar.
Spiderlings were predicted to be males or females as a function
of the majority of the independent assessments.

We determined whether palp width was sexually dimorphic
in juveniles by comparing palpal tibia widths. Palps were
measured in photographs taken at the 3rd instar ( L . hesperus)
or 4th instar (L. hasselti) using a high-resolution digital camera
(Nikon DXM 1200) attached to a dissecting microscope. Each
spider was anaesthetized by brief exposure to CCA and then
laid fiat on the abdomen. A small sheet of paper was placed
under their pedipalps and this was briefly elevated to extend
the pedipalps horizontally for photographing. The width of
the tibia at the point of connection to the tarsus was measured
for each palp using Image Tool (version 3.0, Fig. 1). In L.
hasselti, maximum prosoma width of a subset of 4th instar
spiderlings was also measured and relative palp width (palp
width / prosoma width) calculated as this relative measure
could allow more accurate predictions.

Analysis. — For each species and instar, we compared
accuracy of the categorical assignment method to an expected
null of 50% correct using yf analysis (with Yates’ correction).
We examined whether or not absolute or relative palp width is
dimorphic using /-tests (with unequal variances if Levene’s test
showed a significant difference in variance). We also
determined the extent to which absolute or relative palp
dimensions reliably predict sex using a logistic regression
model with confirmed sex as the dependent variable and rel-
ative or absolute palp width as the predictor, /-tests and logis-
tic analyses were completed in SPSS (version 13.0) and y2 in
GraphPad (http://www.graphpad.com/quickcalcs/contingency 1 .
cfm). Sample sizes vary because in some cases isolated eggs did
not hatch, or photographs of spiderlings were of insufficient
quality to accurately measure body dimensions.

RESULTS

Visual examination method. — Visual examination of palps
yielded accurate sex identification in a high proportion of
cases (Table 1), with no difference in accuracy between the
two Latrodectus species (y2 = 1.109 , df = 1 , P — 0.2923).
When palps were visually examined at the 3rd instar,
approximately 89% (n = 88) of the predictions made for L.
hasselti and 94% (n = 100) of the predictions for L. hesperus
were accurate (Table 1). Accuracy decreased significantly and
prediction was no better than chance if spiderlmg sex was
predicted based on direct examination of spiderlings at the 2nd
instar (L. hesperus, 62% accuracy, Table 1).

Since the highest accuracy was achieved through visual
inspection at the 3rd instar, we focused on understanding
sources of error in these predictions. Prediction errors (n — 6)
for L. hesperus spiders categorized at the 3rd instar were all
cases of spiderlings initially predicted to be female that were in
fact male. Most errors (5/6) occurred for spiderlings in which

the 3 independent predictions were in disagreement. Although
in 16 such equivocal cases, the majority categorization was
correct, in 24% of these cases (5/21), 2/3 viewings suggested
“female-like” palps, but spiders were in fact male. The one
additional error occurred when 3/3 viewings suggested
“female-like” palps (but the spider was male). We examined
our error rate if we excluded from the data set all equivocal
cases where the majority categorization was “female-like” ( n

- 7 cases, in two of which this was the correct categorization).
In this reduced data set, accuracy increases to 99% (92/93) for
sexing by visual examination in L. hesperus (although this is
not a significant improvement in accuracy, y~ = 2.083, df — 1 ,
P = 0. 1490).

Similarly, for 3rd instar L. hasselti, 90% (9/10) of prediction
errors occurred when individuals considered to have “female-
like” palps as spiderlings were actually male. For those L.
hasselti spiderlings that were examined 3 times (n = 58), there
were a total of 9 prediction errors (84% correct). In these data,
there were 24 equivocal cases (1 of the 3 assessments
inconsistent with the others), and in 8 (33%) of these cases,
predictions were incorrect. We again examined our error rate
if we excluded from the data set all equivocal cases where the
majority categorization was “female-like” (n = 12 cases, in 5
of which this was the correct categorization). In this reduced
data set for L. hasselti, accuracy of the visual examination
method increases to 96% (44/46 correct, not a significant
improvement in accuracy: y~ ~ 2.306, df = !, P = 0.1289).

Finally, for 3ld instar spiders, we examined the level of
accuracy that could be achieved if only a single visual
examination was used to identify sex, as this may be required
in some field studies. For L. hasselti spiderlings examined
independently 3 times (n = 58), accuracy was 76% at the first
and second assessments, and increased to 86% in the third
assessment. Accuracy was 97% for an additional 30 L. hasselti
spiderlings that were assessed only once. For L. hesperus (n =
100), accuracy increased from 79% at the first examination, to
94% at the second and finally 98% at the third examination. In
both species, accuracy of identification increased with each
subsequent (blind) attempt at categorization. As was the case
for categorizations based on the consensus of 3 rankings, most
of the errors were spiders predicted to be “female-like” that
were actually male (29/36 errors for L. hasselti, 27/29 errors for
L. hesperus).

Pedipalp measurement method. — Absolute pedipalp width at
the tibia-tarsal joint was sexually dimorphic in spiderlings of
both species (Table 2, Fig. 2). Males had significantly wider
pedipalps than females at the 3rd instar in L. hesperus (/ =

— 15.847, df = 82.504, P < 0.001, Fig. 2A) and at the 4th instar
in L. hasselti (/ = 14.00, df = 23, P < 0.001, Fig. 2B). For L.
hesperus, absolute pedipalp width accurately predicts the sex
of 94.6% of 3,d instar spiderlings (logistic model y2= 79.04, P
< 0.001). Similarly, absolute palp width accurately predicts

516

THE JOURNAL OF ARACHNOLOGY

Table 2. — Mean absolute and relative palpal tibia dimensions ± S.D. for juvenile males and females of two species of Latrodectus spiders.
*palpal tibia width/ prosoma width.

Absolute width (mm)

Relative width*

Species

Instar

Male (/?)

Female ( n )

Male («)

Female (n)

L. hesperus

L. hasselti

yd

4^

0.128 ± 0.013 (68)
0.180 ± 0.006 (6)

0.097 ± 0.006 (24)
0.111 ± 0.002 (19)

0.225 ± 0.014 (6)

0.119 ± 0.003 (15)

the sex of 100% of 4l1' instar L. hasselti (n = 25, logistic model
X2= 27.554, df = 1, P < 0.001).

Relative palp width (palp width/prosoma width, n = 25 L.
hasselti spiderlings) was also sexually dimorphic (/ = 7.438, df
= 5.571, P < 0.001, Table 2, Fig. 2C), and yielded 100%
accuracy in predicting sex of 4th instar L. hasselti spiderlings (/;
= 25, model ;/2= 27.554, df = 1, P < 0.001).

DISCUSSION

We have shown that, in Latrodectus hasselti and Latrodectus
hesperus juveniles, the palpal tibia is significantly wider in
males than in females, and this allows accurate identification
of sex using a simple visual examination of the spiderlings. In
both species, accuracy of over 89% can be attained, and this
may increase to 99% if the relatively rare individuals (7/100 for
L. hesperus, 12/58 for L. hasselti) for which equivocal
categorization predicts “female-like” palps are excluded (i.e.,
3 independent predictions not in agreement, and the majority
of predictions suggest “female”). In L. hesperus, categoriza-
tion of spiderlings based on visual examination of palps was
incorrect in only 6/100 (6%) of cases. This relatively low error
rate could be accommodated statistically in studies of
population dynamics or sex ratio, particularly since the
polarity of these errors was always males mis-categorized as
females. There are a number of alternatives that could increase
the accuracy of sexing. First, individuals with equivocal
predictions may be reared through one or two additional
molts until swelling of male palpi is more readily observable,
thus providing highly accurate data with a much-reduced need
for labor-intensive rearing than is possible without this
technique. Second, repeated scoring of the same individuals
may increase reliability. Third, the person scoring the spiders
could examine a large number of spiders before attempting
categorization. In our study, accuracy of the visual identifi-
cation method increased with each subsequent attempt at
categorization within each species, (i.e., from 76% to 97% in
L. hasselti and from 79% to 98% in L. hesperus). Presumably,
this was due to the increased experience of each investigator
with each independent examination of the specimens (note
that each species was scored by a different researcher, see
methods). Fourth, since most of the ambiguous individuals are
cryptic males, these individuals may be scored as males,
regardless of whether they were scored more frequently as
male-like or as female-like, with a relatively small increase in
error rate.

Although the polarity of errors for L. hasselti was similarly
consistent, the proportion of cases with equivocal categoriza-
tion was far higher (41% of 58 cases). This would be more
difficult to accommodate statistically, and would require
labor-intensive rearing of almost half of the spiders examined
for direct assessment. The higher proportion of equivocal

cases for L. hasselti may arise because the use of direct
observation of individuals is inferior to examination of
photographs for assessment of sex. It is likely that the
difference in the number of equivocal cases arises from
observer error driven by the method of observation because,
if anything, palpi are more dimorphic in L. hasselti than L.
hesperus (Figs. 1, 2), but our predictions were more accurate
for L. hesperus (Table 1). We note however, that the accuracy
of categorization derived from a single examination can be
very high after the observer has experience using this
technique (i.e., for L. hasselti, accuracy = 97% for n = 30
spiderlings examined only once).

Our logistic analyses showed that a regression model using
measurements of absolute or relative pedipalp width as a
predictor could similarly yield up to 100% accuracy in
assessing spiderling sex. While this method did not signifi-
cantly improve accuracy over the visual categorization method
it may widen the range of instars and situations in which this
technique may be applied.

The practical application of this morphology-based tech-
nique to other species will depend on whether or not similar
early sexual dimorphism in palp width exists. This may be
particularly useful in species that are monomorphic to the
naked eye at hatching but in which the male’s palpi are
notably enlarged at adulthood, as in the genus Latrodectus. If
palp dimorphism generally develops gradually over several
instars, then perhaps the best candidates for sexing spiders
visually at an early instar are those species in which males
attain sexual maturity in few instars. Applying this method to
other species would critically require pilot studies in which
individuals are reared until males and females are clearly
distinguishable to confirm the applicability of this technique.
Early indications that such studies may be fruitful would
include significant bimodal clustering of absolute or relative
pedipalp width measurements in early-instar spiderlings (e.g.,
Fig. 2). Such pilot studies should also identify the earliest
instars at which this technique would work. In our data set for
example, predictions were no better than chance for 2nd instar
spiderlings, but highly accurate for spiderlings one instar later
(Table 1).

Once pilot studies are complete, the more simple visual
examination method can potentially be used for a range of
studies. In addition to non-destructive sampling, one advan-
tage of this method is that an experienced observer should be
able to classify the sex of juvenile spiders without the
disturbance involved in collecting or measuring body parts.
Clearly, this method should be very useful for laboratory
studies. In addition, for larger species it may be possible to use
macro-photography, a portable microscope, or visual inspec-
tion using a hand lens in the field to determine sex. One
limitation of the visual examination method is that it requires

MAHMOUDI ET AL. RELIABLE METHOD FOR SEXING SPIDERLINGS

517

A

B

Palpal Tibia width (mm)

Palpal tibia width / prosoma width

Figure 2- Distribution of absolute dimensions of palpal tibia of male
(cross-hatched bars) and female (black bars) spiderlings of (A) L. hesperus
(3rd instar), (B) L. hasselti (4th instar) and (C) distribution of relative palp
dimensions (palpal tibia width/ prosoma width) for 4th instar L. hasselti.

knowledge of spider instar, or at least that the spiders under
consideration are in the same instar. In the field this may pose
difficulties for multivoltine species unless populations are part
of long-term monitoring studies. This problem will likely be
greatly reduced if measures of relative pedipalp width are
used, but confirming this requires additional study.

ACKNOWLEDGMENTS

We thank undergraduate assistants in the Andrade lab for
help rearing spiderlings, particularly Kuhan Perampaladas
and Sen Sivalinghem. We thank Uros Jovovic for contribu-
tions to a pilot study and Matthias Foellmer for helpful
comments on the manuscript. Funding was provided by the
Department of Biological Sciences (UTSC), grants to MCBA
from the Natural Sciences and Engineering Research Council
of Canada (NSERC Discovery Grant), Canadian Foundation
for Innovation, and Ontario Innovation Trust. During
completion of this work, MCBA was partially supported by
a University Faculty Award (NSERC), YB was supported by
a Premier’s Research Excellence Award to MCBA, and MM
was supported by an NSERC CGS M fellowship.

LITERATURE CITED

Austad, S.N. & M.E. Sunquist. 1986. Sex-ratio manipulation in the
common opossum. Nature 324:58-60.

Aviles, L. & W. Maddison. 1991. When is the sex-ratio biased in
social spiders - chromosome studies of embryos and male meiosis
in Anelosimus species (Araneae, Theridiidae). Journal of Arachnol-
ogy 19:126-135.

Berry, J.W. 1987. Notes on the life history and behavior of the
communal spider Cyrtophora moluccensis (Doleschall) (Araneae,
Araneidae) in Yap, Caroline Islands. Journal of Arachnology
15:309-319.

Bhatnagar, R.D.S. & J.G. Rempel. 1962. The structure, function, and
postembryonic development of the male and female copulatory
organs of the black widow spider Latrodectus curacaviensis
(Muller). Canadian Journal of Zoology 40:465-510.

Cameron, E.Z., W.L. Linklater, K.J. Stafford & C.J. Veltman. 1999.
Birth sex ratios relate to mare condition at conception in
Kaimanawa horses. Behavioral Ecology 10:472^175.

Emlen, S.T. 1997. When mothers prefer daughters over sons. Trends
in Ecology and Evolution 12:291 292.

Foelix, R.F. 1996. The Biology of Spiders, Second edition. Harvard
University Press, Cambridge, Massachusetts. 330 pp.

Gunnarsson, B. & A. Andersson. 1992. Skewed primary sex-ratio in the
solitary spider Pityohyphantes phyrgianus. Evolution 46:841-845.
Kaston, B.J. 1970. Comparative biology of American black widow
spiders. San Diego Society of Natural History, Transactions 1 6:33—
82.

Nager, R.G., P. Monaghan, R. Griffiths, D.C. Houston & R.
Dawson. 1999. Experimental demonstration that offspring sex
ratio varies with maternal condition. Proceedings of the National
Academy of Sciences USA 96:570-573.

Uhl. G. & B. Gunnarsson. 2001. Female genitalia in Pityohyphantes
phrygianus, a spider with a skewed sex ratio. Journal of Zoology,
London 255:367-376.

West, S.A. & B.C. Sheldon. 2002. Constraints in the evolution of sex
ratio adjustment. Science 295:1685-1688.

West, S.A., S.E. Reece & B.C. Sheldon. 2002. Sex ratios. Heredity
88:1 17-124.

Manuscript received 26 October 2007, revised 11 April 2008.

2008. The Journal of Arachnology 36:518-526

Activity pattern of the Neotropical harvestman Neosadocus maximus (Opiliones, Gonyleptidae): sexual

and temporal variations

Francini Osses and Tais M. Nazareth: Programa de Pos-graduaqao em Ecologia e Conservagao de Recursos Naturals,
Universidade Federal de Uberlandia, CP 593, 38400-902 Uberlandia, MG, Brazil

Glaueo Machado1: Departamento de Ecologia, Institute de Biociencias, Rua do Matao, Travessa 14, no 321, 05508-900,
Sao Paulo, SP, Brazil

Abstract. We investigated the activity pattern of males and females of the neotropical harvestman Neosadocus maximus
Giltay 1928 focusing on behavioral variations between day and night and also between summer and winter. Our study also
proposes a new method for quantifying arthropod behavior in captivity based on totally random samplings, which
minimizes problems of pseudoreplication, so common in studies of behavioral repertoires. Eighteen individuals (nine males
and nine females) collected at the Parque Estadual Intervales, Sao Paulo state, Brazil, were maintained in the laboratory
from June 2003 to February 2004 for qualitative and quantitative observations. Thirty-four behavioral acts grouped in five
behavioral categories were recorded, with “resting” (53.1%) and “exploration” (30.8%) being the most frequent. The
behavioral repertoire of males (32 acts) was more diversified than that of females (29 acts). Moreover, there was a
significant effect of sex (male X female) on the frequency of the behavioral categories: females expended more time resting
than did males, whereas males expended more time in exploration and grooming activities than did females. There was also
a significant effect of time of day (day X night) and of season (winter X summer) on the frequency of the behavioral
categories: individuals were most active at night and during summer. Finally, temperature, but not humidity, had a positive
effect on the activity of the individuals. In conclusion, the activity of N. maximus has sexual, daily, and seasonal variations,
which are both quantitative and qualitative.

Keywords: Behavioral sampling, behavioral repertory, ethogram, rhythm, seasonality, Zeitgeber

Arachnids are generally considered to be primarily noctur-
nal, although there are exceptions among representatives of
the orders Araneae, Solifugae, and Opiliones (Cloudsley-
Thompson 1978; Foelix 1996; Punzo 1998; Hoenen &
Gnaspini 1999). However, the activity patterns of few
arachnids have been quantitatively addressed so that studies
on spiders roughly divide the species into day- or night-active
(e.g., Marc 1990; Alderweireldt 1994). Additionally, quantita-
tive data on sexual and temporal variations in the activity
patterns of arachnids have rarely been reported. Among
scorpions, for instance, the seasonal surface activity of mature
males differs markedly from that of females in many species
(Warburg & Polis 1990). At least for some North American
scorpions, females are more active in spring and fall, which is
coincident with the peak of insect abundance, whereas males
are more active on the surface during midsummer, corre-
sponding to the period of greatest abundance of virgin females
(Polis & Farley 1 979;., b; Polis 1980).

Most harvestmen species seem to concentrate their activity
at night, when temperature decreases and relative air humidity
increases (Todd 1949). Certain species, however, are more
active during daytime, and others may be found active
throughout the day (Bishop 1950; Pfeifer 1956; Williams
1962; Hoenen & Gnaspini 1999). The literature on rhythms is
scarce for harvestmen, but includes some information on
sexual differences in the activity patterns. Gnaspini (1996), for
instance, described that ovigerous females of the Brazilian
cavernicolous harvestman Goniosoma spelaeum Mello-Leitao
1922 (Gonyleptidae, Goniosomatinae) leave the cave to forage
more frequently than males. Recently, the behavioral reper-

1 Corresponding author. E-mail: glaucom@ib.usp.br

tory of males and females has been compared quantitatively in
two gonyleptid harvestmen from Brazil, namely Discocyrtus
oliverioi Soares 1945 (Pachylinae) and Mischonyx cuspidatus
Roewer 1913 (Gonyleptinae). In both species females fed more
frequently than males, whereas males were involved in social
interactions more frequently than females (Elpino-Campos et
al. 2001; Pereira et al. 2004). Although these two studies
describe the daily activity pattern of the harvestmen, they do
not provide data on seasonal variation in the frequency of the
behavioral categories because both were done only during the
summer (December to February).

The great majority of the information regarding seasonal
variations in harvestmen is related to the phenology of the
species (e.g., Phillipson 1959; Friebe & Adis 1983; Martens
1993; Tsurusaki 2003). These phenological studies provide
data on the occurrence of different instars throughout the
year, but generally do not deal with seasonal differences in
behavior. However, examples of seasonal changes in behavior
may be found in the neotropical harvestmen Acutisoma
longipes Roewer 1913 and Goniosoma albiscriptum Mello-
Leitao 1932 (Gonyleptidae, Goniosomatinae), which form
aggregations inside caves only during the dry and cold season
(Machado et al. 2000; Willemart & Gnaspini 2004a) and
reproduce only during the wet and warm season (Machado &
Oliveira 1998; Willemart & Gnaspini 2004b). The most
comprehensive study on seasonal changes in harvestman
behavior has been conducted by Gnaspini et al. (2003) with
G. spelaeum. The authors showed that exits of individuals
from the cave were related to sunset time, occurring earlier in
winter (June/July) than in summer (December/January).
However, the advancement in the time during which
individuals leave the cave in summer does not exactly follow

518

OSSES ET AL.— ACTIVITY PATTERN OF NEOSADOCUS MAXIMUS

519

the advancement in the time of sunset in comparison to winter
data, suggesting a change in the phase angle.

Ethograms are important starting points for ethological
research and for understanding the biology and ecology of a
wide range of animals (Lehner 1996). This methodology
allows qualitative and quantitative comparisons between the
behavioral repertories of different species, as well as individ-
uals of the same species belonging to different sexes. In this
study we used such an approach to compare the behavioral
repertory of males and females of the neotropical harvestman
Neosadocus maximus Giltay 1928 (Gonyleptidae, Gonylepti-
nae), a species commonly found in forest fragments of the
Atlantic Forest in southeastern Brazil Additionally, we
investigated for the first time whether the activity pattern of
males and females varies between day and night, and also
between summer and winter. Our study also proposes a new
method for quantifying arthropod behavior in captivity based
on totally random samplings, which minimizes problems of
pseudoreplication, so common in studies of behavioral reper-
tories.

METHODS

Taxonomic note. — The genus Neosadocus is defined by a
cluster of tubercles on the dorso-basal portion of femur IV of
males (A.B. Kury pers. comm.), and includes species separated
by subtle variations (Kury 2003). Misidentifications are
commonplace in the literature and the species treated in this
study as Neosadocus maximus has been previously named N.
variabilis by Gnaspini (1996), Machado & Pizo (2000), and
Willemart (2002) or Neosadocus sp. (Willemart et al. 2007).
Two other studies (Willemart & Gnaspini 2003; Castanho &
Pinto-da-Kocha 2005), however, properly identified the
species as N. maximus. The species studied by Machado &
Vital (2001) and identified as Neosadocus aff. variabilis is
probably closely related to N. maximus, but it is a distinct
taxon. Voucher specimens are deposited in the Museu de
Zoologia da Universidade de Sao Paulo (MZSP, Brazil).

Collection and rearing. — Individuals of N. maximus were
collected in a dense Atlantic Forest fragment at the Parque
Estadual Intervales (24°14'S; 48°04'W; 800 m elev.), close to
the municipality of Kibeirao Grande, southern Sao Paulo
state, Brazil. The annual rainfall in the region ranges from
2,000 to 3,000 mm and the mean annual temperature ranges
from 12° C to 20° C. The climate is seasonal with two well-
defined seasons. The dry and cold season (hereafter called
winter) lasts from April to September, has a mean monthly
rainfall of 139 mm, mean monthly air temperature of 17.4° C,
and mean monthly air humidity of 79% (data taken in loco
during 5 yr, 1998-2002). The rainy and warm season
(hereafter called summer) lasts from October to March, has
a mean monthly rainfall of 306 mm, mean monthly air
temperature of 20.9° C, and mean monthly air humidity of
83% (data taken in loco during 5 yr, 1998-2002).

Behavioral observations were done in the laboratory based
on 18 specimens: nine females and nine males. Each individual
received a number from 1 to 18 marked on their dorsal scute
with synthetic enamel paint for model airplanes (cf. Elpino-
Campos et al. 2001). Individuals were maintained in a
communal terrarium (40 X 90 cm base, 20 cm height)
containing soil, two individuals of an aroid plant, and pieces

of rotting logs where they could rest and find protection from
light. They were fed pieces of dead cockroaches three times a
week and always received additional food before the
behavioral samplings. From April to September the conditions
in the laboratory were (mean ± SD): temperature of 22.1 ±
2.0° C, humidity of 75.6 ± 7.1%, and photoperiod of
12L:12B. Between October and March we changed the
laboratory conditions in order to simulate natural climatic
oscillations in nature: temperature was 25.5 ± 1.2° C,
humidity was 82.0 ± 5.4%, and photoperiod was 1 3L: 1 ID.
Although temperature variation in the laboratory was not the
same as In the field, the magnitude of the difference between
winter and summer was nearly the same, and the air humidity
was close to that measured in the study site. Moreover, the
laboratory photoperiod, which is the most important Zeitge-
ber for many organisms (including arthropods, see Aschoff
1960), reproduced the same seasonal variation observed in
nature from winter to summer. Thus, we believe that the
results obtained in captivity can be generalized to harvestmen
under field conditions.

Behavioral sampling.- The ethogram of N. maximus was
derived from preliminary 18 h of qualitative ad libitum
observations of captive individuals (Table 1). The behavioral
acts recorded during this phase were classified into five major
groups of activities or “behavioral categories”: feeding,
exploration, self-grooming, resting, and social interactions
(Table 2). Quantitative behavioral observations were done in
two distinct seasons: winter samples (n = 24) were gathered
from June to August 2003, and summer samples (n = 24) were
gathered from December 2003 to February 2004. In this study
we used a new sampling procedure that was designed to
minimize the problem of pseudoreplication (see Discussion).
According to our protocol, each sample consisted of 1 h of
observation randomly selected from midnight to 23:00. After a
given 1 h period was selected it was excluded from the next
assortments. No sample was done with less than 24 h of
interval from the previous sample. During each sample, the
behavioral acts were recorded at 1 min intervals (“fixed-
interval time point sampling” sensu Martin & Bateson 1993).
Only one individual was recorded at each 1 min interval, and
its identity was based on a list of random numbers ranging
from 1 to 18, in which each number corresponded to the mark
of one individual. At the end of each sample, 60 behavioral
acts were always recorded (one per minute), and each
individual was scanned from one to nine times (median of
three times, n = 48 sections). When a selected individual was
not visible in the terrarium, we searched for it by gently raising
the pieces of rotting logs. A flashlight with red filter was used
for the nocturnal observations (cf. Elpino-Campos et al. 2001).

Statistical analyses. — The data were classified into two
periods: winter and summer. Within each of these two
categories, they were also classified in day and night samples.
Since the photoperiod differed between the two seasons, the
number of day samples was 12 in winter and 13 in summer.
Finally, the data were classified into males (n = 9) and females
(n = 9). A log-linear analysis was used to test for the effects of
the period of the day, season, and sex on the frequency of the
behavioral categories. This analysis is based on fitting a
hierarchical log-linear model to the cell frequencies so that it is
possible to restate the problem of analyzing multi-way

520

THE JOURNAL OF ARACHNOLOGY

Table 1. — Behavioral acts of the harvestman Neosadocus maximus based on 18 h of ad libitum observations of nine males and nine females
maintained under laboratory conditions. Behavioral acts marked with an asterisk were modified from Elpino-Campo et al. (2001).

Behavioral acts

Description

Feeding*

Walkabout carrying food
Drinking water
Walkabout - legs II up*

Walkabout - legs II touching the substrate

Motionless - legs II down

Motionless - legs II up

Motionless - legs II back

Motionless - legs II touching the substrate*

Cleaning the pedipalps*

Leg threading*

Cleaning the venter, dorsum, legs III or IV
using legs II

Resting*

Aggressive posture
Mutual leg tapping

Intertwining legs IV

Chasing other individual
Fighting for food*

Attacking with the pedipalps

Touching other individuals*

To manipulate organic matter using the chelicerae and pedipalps, inserting it into the mouth.
This behavior may be accomplished with the body standing on the legs or with the venter in
contact with the substrate. When two or more individuals manipulate the same piece of food at
the same time, this behavior is classified as “sharing food.”

To walk carrying food with the chelicerae and pedipalps.

To put the mouthparts in direct contact with the water.

To walk with legs II extended sideways or towards the front without touching the substrate.

To walk with legs II extended sideways or towards the front touching the substrate.

To stay motionless with legs II extended sideways downward without touching the substrate.

To stay motionless with legs II extended sideways upward (Fig. 2).

To stay motionless with legs II extended back and upward.

To stay motionless with legs II touching the substrate.

To pass the tip of the pedipalps through the chelicerae and mouthparts.

Process of cleaning in which each leg is passed through the chelicerae and mouthparts until the
end of tarsus is reached.

Process of indirect cleaning in which legs II are passed on some parts of the body (venter,
dorsum, legs III or IV) and are then cleaned, passing them through the chelicerae and
mouthparts until the end of tarsus is reached.

To remain motionless with legs retracted over the body and with the venter in contact with the
substrate (Fig. 1). While resting individuals may be: (1) isolated, at least 3 cm from each other
or (2) in groups, at least three individuals at 0-2 cm apart from each other with legs
overlapping. Individuals may rest in protected (under rotting logs) or unprotected sites (open
areas of the terrarium).

To remain motionless with the pedipalps raised toward another individual.

Occurs when two individuals walk around each other touching themselves with legs I and II. This
is a preliminary step of an intra-sexual fight accomplished exclusively by males.

After mutual leg tapping, the individuals turn their backs to each other and intertwine legs IV. In
this position, each male tries to capsize each other promoting a sudden upward movement in
which a male brings femur IV close to the body and pinches legs IV of the opponent.

Occurs when a male (generally the winner of the fight) runs after his opponent.

Any aggression toward another individual possessing food. It involves attack with pedipalps and
touching with the first two pairs of legs.

To approach another individual not possessing food and aggressively attack it with the
pedipalps.

Physical contact between two individuals established using legs I and/or II, without aggressive
posture (see above) by either individual.

frequency tables in terms that are very similar to ANOVA.
According to this model, the natural logarithm of the expected
cell frequency is written as an additive (linear) function of
effects, including first-order effects and higher-order interac-
tions among the categorical variables (Christensen 1997). This
statistical procedure is more than just an alternative form of
the chi-square test, which is sensitive to the most common
violation of their assumptions, i.e., lack of independence
between observations (Kramer & Sdimidhammer 1992). Its
strength lies in that it can be extended to quite complicated
contingency tables involving several variables.

A multiple regression was used to test the relationship
between temperature and humidity (independent variables)
and number of behavioral acts accomplished per hour
(dependent variable). Regarding this analysis, it is important
to stress the following: (a) there is no significant correlation
between the independent variables (r = 0.1 39; i3 = 0.346), so
that there is no problem of multicollinearity in the multiple
regression; (b) the number of behavioral acts does not include
the category “resting” since we were interested in testing the
influence of temperature and humidity only on the activity of
the individuals; (c) data from both summer and winter were

used, comprising 48 samples. A two-factor ANOYA was
performed to test for differences in the number of behavioral
acts per hour (excluding “resting”) between day and night in
both seasons.

RESULTS

General patterns. — The behavioral repertory of captive
individuals of Neosadocus maximus comprises 34 behavioral
acts (Table 1). No behavioral act directly related to repro-
duction, such as courtship, copulation, egg-laying, or parental
care has been recorded in captivity. During the 48 h of
quantitative observations, four acts of the behavioral reper-
tory were not recorded, all belonging to the category “social
interactions” (numbers 27-29 and 31 in Table 2). The most
common behavioral categories were “resting” and “explor-
ing,” which accounted for 53.1% and 30.8% of the total of
observations, respectively (Table 2). Most of the “resting”
observations were of individuals grouped underneath rotting
logs (Fig. 1), whereas most of the “exploring” observations
were of motionless individuals with their second pair of
sensorial legs upwards (Table 2; Fig. 2). The great majority of
the individuals fed alone, and food sharing was recorded in

OSSES ET AL.— ACTIVITY PATTERN OF NEOSADOCUS MAXIMUS

521

Table 2. — Frequency of each behavioral act for captive adults of the harvestman Neosadocus maximus in two seasons (winter and summer). In
parentheses is the total number of behavioral acts observed for the individuals in each season.

Frequency of the behavioral acts (%)

Males (n = 9)

Females (n = 9)

Total (n — 18)

Behavioral acts

Winter (697)

Summer (661)

Winter (743)

Summer (779)

Winter (1440) Summer (1440)

Feeding

3.73

14.54

3.23

13.99

3.48

14.23

1 . With the body erect

3.01

11.5

2.15

9.50

2.57

10.42

2. With the body in contact with the

0.29

1.83

0.82

2.43

0.56

2.15

substrate

3. Walkabout carrying the food

0.14

0.61

0.13

1.16

0.14

0.90

4. Food sharing

0

0.15

0

0.51

0

0.35

5. Drinking water

0.29

0.45

0.13

0.39

0.21

0.41

Exploration

34.85

39.94

21.27

28.50

27.85

33.75

6. Walkabout - legs II up

4.3

3.78

2.16

2.18

3.19

2.92

7. Walkabout - legs II touching the substrate

5.16

4.69

4.98

3.34

5.07

3.96

8. Motionless - legs II down

2.73

11.04

2.69

7.32

2.71

9.03

9. Motionless - legs II up

14.20

12.56

7.27

11.04

10.63

11.73

10. Motionless - legs II back

5.45

5.60

1.48

2.95

3.40

4.17

11. Motionless - legs II touching the

3.01

2.27

2.69

1.67

2.85

1.94

substrate

Self-Grooming

5.90

7.10

3.22

5.26

4.51

6.12

12. Cleaning the pedipalps

2.44

3.93

1.48

2.95

1.94

3.40

13. Leg threading - legs I

1.00

1.66

0.54

1.28

0.76

1.46

14. Leg threading - legs II

1.59

0.45

0.94

0.77

1.25

0.63

15. Leg threading - legs III

0.29

0.61

0.13

0.26

0.21

0.42

16. Cleaning legs III with legs II

0

0.15

0

0

0

0.07

17. Leg threading - legs IV

0.44

0.15

0

0

0.21

0.07

18. Cleaning legs IV with legs II

0.14

0

0

0

0.07

0

19. Cleaning the venter with legs II

0

0.15

0

0

0

0.07

20. Cleaning the dorsum with legs II

0

0

0.13

0

0.07

0

Resting

53.37

36.91

69.86

50.45

61.87

44.22

21. Resting alone protected

10.90

3.63

8.34

10.53

9.58

7.35

22. Resting alone unprotected

5.16

6.20

3.77

0.90

4.44

3.33

23. Resting in group protected

37.02

27.08

57.07

38.76

47.36

33.40

24. Resting in group unprotected

0.29

0

0.68

0.26

0.49

0.14

Social interactions

2.15

1.51

2.42

1.80

2.29

1.68

25. Aggressive posture

0

0

0

0.13

0

0.07

26. Mutual leg tapping

0.14

0

-

-

0.07

0

27. Intertwining legs IV

0

0

-

-

0

0

28. Chasing another individual

0

0

-

-

0

0

29. Male vs. female fight

0

0

0

0

0

0

30. Fighting for food

0

0.45

0.13

0.13

0.07

0.28

31. Attacking with the pedipalps

‘0

0

0

0

0

0

32. Male touching male

1.29

1.06

-

-

0.69

0.49

33. Female touching female

-

-

0.81

0.77

0.35

0.42

34. Female touching male, and vice versa

0.72

0

1.48

0.77

1.11

0.42

only 0.3% of the observations (Table 2). Probably as a
means to avoid food sharing or even food plundering,
some individuals carried the food away before starting
consumption. Grooming behavior occurred when the har-
vestmen were moving around the terrarium and, more
frequently, after feeding (when both legs and pedipalps were
cleaned) and social interactions, such as fights and mutual leg
tapping.

Sexual variation, — The behavioral repertoires of males and
females of N. maximus differed both qualitatively and
quantitatively. The behavioral acts number 26-29 were
accomplished exclusively by males and are related to a ritual
of intra-sexual aggression (Table 1). Five fights between males
were observed in the qualitative observations, but during the

quantitative samplings this behavior was not recorded. A fight
starts with two males walking around each other, touching the
opponent with the first and the second pairs of legs. After
about three to five complete turns the males stop and, if the
fight escalates, they turn their backs to each other and
intertwine their fourth pairs of heavily-armed legs. In this
position, each male moves the femur IV pinching the
opponent’s leg between the apophysis and spines of their
own femur and coxa IV. The fight ends when one of the
contenders leaves the place.

The log-linear analysis showed that there was a significant
effect of sex (male vs female) on the frequency of the
behavioral categories (Table 3). Females were seen resting
and interacting with conspecifics more frequently than males

522

THE JOURNAL OF ARACHNOLOGY

Figures 1, 2.- -Adult Neosadocus maximus. 1. Female of N.
maximus resting under a rotting log in captivity. Note that the legs
are retracted over the body and the venter is in contact with the
substrate. 2. Male of N. maximus walking on the vegetation in the
field. Note the second pair of sensory legs extended sideways and
touching the substrate (photos: B.A. Buzatto).

(Fig. 3). Males, on the other hand, were more active, exploring
and grooming more than females (Fig. 3). Individuals of both
sexes fed with a similar frequency (Fig. 3). Aggressiveness
among males, however, may explain why females rested

Table 3.- Results of the log-linear analysis used to test for the
effects of time (day vs night), season (summer vs winter), and sex
(male vs female) on the frequency of behavioral categories accom-
plished by captive individuals of the harvestman Neosadocus maximus.

Model

Chi-square

df

P

Season X Time X Sex

1.587

4

0.811

Season X Time

8.175

8

0.417

Time x Sex

0.803

4

0.938

Season X Sex

8.498

12

0.745

Season (summer vs winter)

160.960

8

< 0.001

Time (night vs day)

33.929

8

< 0.001

Sex (male vs female)

76.722

8

< 0.001

153 887 255 57 1528

Grooming Exploration Feeding Social Resting

interactons

Behavioral categories

Figure 3. — Comparison of the relative frequency of occurrence of
the behavioral categories observed for males (black bars) and females
(white bars) of the harvestman Neosadocus maximus. The numbers
above the bars indicate the total number of records in each
behavioral category.

aggregated more frequently than males, both in winter and
summer (Table 2).

Temporal variations. — The log-linear analysis showed that
there was a significant effect of the period of time (day vs
night) on the frequency of the behavioral categories (Table 3).
Individuals of N. maximus remained inactive during most of
the daylight hours, when they were resting underneath rotting
logs (Figs. 4, 5). In winter, individuals became active at about
17:00 h, i.e., one hour before dusk (Fig. 4). In summer,
activity began at about 19:00 h, just after dusk (Fig. 5). Night
activities included exploring the environment by walking
around the terrarium and feeding on pieces of dead
cockroaches (Figs. 4, 5). Male-male fights were also restricted
to the nocturnal period. In winter, activities started to decrease
at about 02:00 h, whereas in summer activities started to
decrease only at about 05:00 h, when most of the individuals
were seen resting or moving to the shelters underneath rotting
logs (Figs. 4, 5).

The log-linear analysis also showed that there was an effect
of season (winter vs summer) on the frequency of the
behavioral categories (Table 3). Even though resting and
exploring were the most common activities in both seasons,
there was an increase in the frequency of resting from summer
to winter, and a decrease in the frequency of exploration from
summer to winter (Table 2). Additionally, there was a marked
reduction in the frequency of feeding from one season to the
other: during winter, in nearly 4% of the observations
individuals were feeding versus 14% during summer (Table 2).
The frequency of the categories self-grooming and social
interactions were similar between seasons (Table 2). It is
worthy to note that the interactions between sex and season as
well as time and season have no significant effect on the
frequency of the behavioral categories (Table 3).

A positive relationship between the number of behavioral
acts and the temperature was detected, but this relationship
was not significant for humidity (Table 4). The mean number
of behavioral acts accomplished per hour during the night was
greater than during the day in both seasons (F{\_ 44) = 99.9 1 5;

OSSES ET AL.— ACTIVITY PATTERN OF NEOSADOCUS MAXIMUS

523

Figures 4, 5. — Daily activity schedule of the harvestman Neosa-
docus maximus in (4) winter and (5) summer. The moon and the sun
indicate dusk and dawn, respectively.

P < 0.001; Fig. 6). Additionally, when comparing seasons, the
mean number of behavioral acts accomplished per hour was
greater in summer than in winter (F(i, 44) = 25.368; P < 0.001;
Fig. 6).

DISCUSSION

Comparisons with other species. — Harvestmen are excellent
model organisms for behavioral studies because individuals of
many species are abundant in the field and are easily
maintained in captivity, where they behave in a similar way
to that observed in the field (e.g., Capocasale & Bruno-Trezza
1964; Elpino-Campos et al. 2001; Willemart 2001; Pereira et
al. 2004), facilitating quantitative and qualitative comparisons
between species. In this study, we used the neotropical
harvestman Neosadocus maximus in order to investigate sexual
and temporal variations in the behavior of captive individuals.
The data obtained here may also be compared with those
previously obtained for other harvestman species reared under
similar conditions. The number of behavioral acts recorded for

Table 4. — Multiple regression between the number of behavioral
acts (excluding resting) accomplished by captive individuals of the
harvestman Neosadocus maximus and two independent variables:
temperature and humidity (F(2, 45> = 4.80; P = 0.013).

Independent variable

b

F(l, 44)

P

Humidity

0.066

1.174

0.284

Temperature

0.485

7.425

0.009

S.

S

©

60 -

ts

50 -

”5

.2

40 -

>

©

30 -

©

20 -

a

S

3

10 -

Z

0 -

b

q Day Night ^ ^ Day Night

Summer Winter

J

Figure 6. — Comparison of the number of behavioral acts (exclud-
ing resting) accomplished per hour by individuals of the harvestman
Neosadocus maximus during day and night in two seasons (summer
and winter). The horizontal lines represent the mean, the boxes
represent the standard deviation, and the vertical lines represent the
range (minimum and maximum values in each sample). Different
letters above the box-plots indicate significant difference with P
< 0.05.

N. maximus is greater than those recorded for the harvestmen
Discocyrtus oliverioi (25 acts, Elpino-Campos et al. 2001) and
Mischonyx cuspidatus (20 acts, Pereira et al. 2004). This
difference may be attributed to two main factors: (1) some
behavioral acts described in the previous studies were split into
two or more different acts in our study (e.g., acts 8-1 1); (2) N.
maximus , in fact, presents a number of distinctive behavioral
acts that were not recorded before, such as those related to
self-grooming (number 16 and 18-20) and social interactions
(number 26-28). There was also a marked difference among
the three harvestman species in the frequency of the behavioral
categories exploration and resting. In D. oliverioi and M.
cuspidatus , exploration was the most common behavioral
category, comprising nearly 70% of the observations, whereas
in N. maximus this category accounted for 28% in winter and
34% in summer. Resting, which ranged from 1 1 to 17% in D.
oliverioi and M. cuspidatus, was the most frequent behavioral
category for N. maximus comprising 44 to 62% of the
observations. These discrepancies could be related to differ-
ences in the sampling methods, since Elpino-Campos et al.
(2001) and Pereira et al. (2004) concentrated the quantitative
observations at night (when the individuals were more active),
whereas in the present study we scattered the behavioral
samplings equally throughout the whole day. In fact, if we
analyze the data obtained only in the night samples for N.
maximus, the frequency of resting drops to 22% and of
exploration increases to 52%. Therefore, we believe that the
data presented in this study are a more realistic scenario of
time allocation in harvestmen during the whole day.

Contrary to D. oliverioi and M. cuspidatus , individuals of N.
maximus did not copulate or lay eggs in captivity, so that the
behavioral category reproduction, which comprised one to
seven behavioral acts and accounted for 0.1 to 1.8% of the
observations in the previous studies, was not included here.
Field observations indicate that N. maximus oviposits on the
undersurface of leaves and eggs are attended by females (G.
Machado unpubl. data). It is possible that the lack of

524

THE JOURNAL OF ARACHNOLOGY

appropriate oviposition sites in the terrarium have constrained
the reproductive activity of the females. Males, on the other
hand, accomplished certain behavioral acts, such as intra-
sexual aggressions, that are possibly related to territorial
defense and reproduction. Male-male fights have already been
described for several harvestman species (e.g., Pabst 1953;
Parisot 1962; Edgar 1971; Mora 1990; Macias-Ordonez 2000;
Willemart et al. 2006), but this is the first record of this kind of
agonistic behavior among gonyleptids. The use of legs IV,
which bear large spines and tubercles, shed light on the
behavioral roles of the leg armature in N. maximus and other
gonyleptid harvestmen as well. Until now, the only function
attributed to the armature of legs IV in males was to deliver a
nipping upon manipulation, which has been interpreted as a
defensive behavior against potential predators (Bristowe 1925;
Capocasale & Bruno-Trezza 1964; Gnaspini & Cavalheiro
1998; Machado 2002). A more detailed description of the
fights and the functional morphology of legs IV in N. maximus
will be described elsewhere (Willemart et al. unpubl. data).

Sexual and temporal variation. Males and females of N.
maximus differed in the relative frequency of the behavioral
categories, with males exploring more frequently than females.
This result contrasts with those obtained for D. oliverioi and
M. cuspidatus , in which females fed and explored more
frequently than males (Elpino-Campos et al. 2001; Pereira et
al. 2004). Since both D. oliverioi and M. cuspidatus reproduced
in captivity, the higher frequency of feeding activities in
females compared to males was attributed to the accumulation
of energy for egg production and maturation (Pereira et al.
2004). The higher frequency of exploration by males compared
to females in N. maximus may be explained by at least two
non-exclusive factors: (1) in this species males seem to be
territorial (G. Machado unpubl. data) and thus need to invest
time patrolling and exploring their territories; (2) females did
not oviposit in captivity and, thus, were not continuously
producing eggs, which would reduce the demand for food
resources and consequently decrease the frequency of activities
related to foraging. Since self-grooming occurs more frequent-
ly when individuals are moving around and after feeding
(Pereira et al. 2004), the fact that males explored more than
females may account for the higher frequency of grooming
compared to females.

The daily activity of the individuals of N. maximus in
captivity was predominantly nocturnal, and light seems to be
the most important Zeitgeber promoting synchronization of
the activity rhythm. This observation is congruent with data
previously obtained in the field, where 100% of the feeding
observations in this species occurred at night (Machado &
Pizo 2001). The activity pattern, however, presents seasonal
variations with a clear change in the phase angle between
activity and sunset/sunrise hours from winter to summer.
During winter, individuals left the shelters earlier, one hour
before dusk. The peak of activity occurred nearly at 20:00 h,
and at 02:00 h there was a marked decrease in the frequency of
behavioral acts not related to resting. In contrast, during
summer, individuals left the shelter only after dusk, remained
active throughout the night, and returned to the shelter at
about 05:00 h, nearly one hour before the onset of light. In
some aspects, this pattern of activity is similar to that
described for the cavernicolous harvestman Goniosoma

spelaeum, in which the individuals also left the cave earlier
in winter when compared to summer (Gnaspini et al. 2003).
However, contrary to the present study, individuals of G.
spelaeum returned later to the shelter during winter. The
authors attributed this change in the phase angle to the time
available to forage outside the cave, which is shorter during
summer. It is possible that captive individuals of N. maximus
can find food faster than individuals of G. spelaeum in the wild
and thus can return earlier to their shelters. Another
possibility is that different species respond differently to
seasonal variations in biotic and abiotic factors. We suggest
that future studies investigate how hunger and other
physiological constraints modulate long-term changes in the
biological rhythms of harvestmen.

In our study we demonstrate that temperature, but not
humidity, has a positive relationship with the activity of
captive individuals of N. maximus, which may explain why the
individuals were more active during summer. A similar result
has been reported by Capocasale & Bruno-Trezza (1964), who
reared individuals of the gonyleptid Acanthopachylus aculeatus
(Kirby 1818) in the laboratory and showed that foraging
activity seems to be directly related to temperature. However,
the predominant nocturnal activity of the species in the field
and in the laboratory should not be attributed to temperature
since the latter decreases at night. We believe that the decrease
in light intensity at dusk, rather than the decrease in
temperature, controls the beginning of activity in N. maximus.
After darkness, however, air temperature can be an important
abiotic factor determining harvestmen activity; the warmer the
night the more active are the individuals. Additionally, it is
important to stress the role of phylogeny in determining the
activity pattern of harvestmen. N. maximus belongs to a clade
in which most species are mainly nocturnal. Predominant
diurnal activity among gonyleptids seems to be restricted to
the clade composed of the subfamilies Progonyleptoidellinae +
Caelopyginae (Hoenen & Gnaspini 1999).

Methodological approach. — The density of harvestmen used
in our terrarium (18 individuals in 0.36 nr) is certainly much
higher than the density in the wild, which is no more than 0.04
individuals/nr (G. Machado unpub. data). Crowding may
increase the frequency of some behavioral acts, mainly those
grouped in the behavioral category “social interactions”.
However, for the great majority of the behavioral acts (including
those grouped in the behavioral categories “feeding,” “explo-
ration,” “self-grooming,” and “resting”) the density of individ-
uals in captivity probably had no evident effect. Moreover, since
these last four behavioral categories comprise nearly 98% of the
behavioral acts in both seasons, we believe that the results
obtained under captive conditions for N. maximus provide a
realistic scenario of time allocation throughout the day and also
in different seasons. The lack of appropriate oviposition sites in
our terrarium probably inhibited some behavioral acts, mainly
those related to reproduction. We acknowledge this drawback
of our laboratory work and recommend that future studies try
to reproduce as good as possible the oviposition sites of the
study species in the rearing terrarium.

Despite some minor problems mentioned above, harvest-
men are very convenient animals to keep in captivity since
many species are relatively easy to maintain and may live for
several months or even years (Willemart 2007). In our case, to

OSSES ET AL. — ACTIVITY PATTERN OF NEOSADOCUS MAXIMUS

525

study animals in the laboratory provided the opportunity to
quantify the behavior of a great number of individuals in each
sampling section and also made it possible to compare the
behavioral schedule of the very same subjects in two seasons,
which would be very difficult in the wild. Simulating natural
variations in climatic conditions, we showed that there were
differences in the frequency of five behavioral categories
recorded for N. maximus, both at the scale of the day and of
the year (winter vs summer). This temporal variation may be
endogenously regulated and/or dependent on the environmen-
tal variables we manipulated, but we can not actually
differentiate between these two possibilities.

In this study we also proposed a new sampling procedure
for quantitative ethograms that minimizes problems of
pseudoreplication. Traditionally, behavioral samplings in
quantitative ethograms are accomplished at regular intervals
and all subjects are recorded at each scan. Using this
procedure, one will face at least two situations that violate
the assumption of independent observations required by many
statistical tests, especially the frequently used chi-square (see
discussion in Kramer & Schmidhammer 1992): (1) one
behavioral act may influence the chance of another behavioral
act being accomplished in sequence; (2) if the individual A is
interacting with the individual B, this behavioral act will be
counted twice because the interaction is reciprocal. Yet the
“traditional” ethograms have an advantage: the amount of
information in each sampling section is high; more specifically
n X s, where n is the number of individuals and s is the number
of scans. Using the method proposed here, the amount of data
in each sampling section is only s because only one individual
is scanned at a time. Sampling one individual per scan,
however, is exactly the solution for the situation (2) above.
Additionally, the median number of times the same individual
was scanned per sampling section along the entire period of
our study was three. These repeated samplings on the same
individuals were spaced out in time, which attenuates the
problem exposed in the situation (1). Finally, we also took
care of spreading the sampling sections along the time, spacing
them with intervals of at least 24 h in order to attenuate the
possible influence the activities accomplished in one day could
have on activities accomplished in the following day. This is
not a standard procedure in “traditional” ethograms, which
concentrate the samplings in the periods of more activity or in
fixed times of the day.

We are aware that sampling the same individuals (in our
case 18 harvestmen) is, per se, a source of pseudoreplication,
but changing each animal for another one after it was sampled
does not seem a reasonable procedure in behavioral studies
and would demand a huge quantity of animals. Anyone
interested in controlling potential differences among individ-
uals may include each one of them as additional factors in the
log-linear analysis. Here we avoided this approach because
our main goal was to detect general patterns.

Conclusions. — In conclusion, we demonstrated that the
activity pattern of the neotropical harvestman N. maximus,
determined here by the frequency of five behavioral categories,
shows sexual and temporal variations. These variations are
both quantitative and qualitative since some behavioral acts
are restricted to one sex or period of the day. The sampling
protocol proposed here should be used in future studies

dealing with behavioral repertoires and ethograms because it
minimizes the problems of pseudoreplication and provides a
more realistic view of the allocation of time and energy for
different activities. The great advantage of this method is that
it provides a suitable sampling design that generates indepen-
dent data, instead of trying to correct the problem of non-
independence a posteriori using complicated statistical proce-
dures. Our protocol should be useful as a standard method in
behavioral samplings not only for harvestman, but for the
study of any arthropod reared in captivity.

ACKNOWLEDGMENTS

We are grateful to Gustavo S. Requena and Bruno A.
Buzatto for collecting some individuals used in this study, to
Adriano B. Kury for helping with the identification of the
studied species, to Paulo De Marco Jr. and Rafael Lourenqo
for helping with the statistical analyses, and to Pedro
Gnaspini, Sonia Hoenen, Rodrigo H. Willemart, Bruno A.
Buzatto, Paulo Enrique C. Peixoto, Gail Stratton, and two
anonymous reviewers for insightful comments on an early
draft of the manuscript. The authors are supported by grants
from CAPES (FO and TMN) and Fundaqao de Amparo a
Pesquisa do Estado de Sao Paulo (GM 02/00381-0).

LITERATURE CITED

Alderweireldt, M. 1994. Day/night activity rhythms of spiders
occurring in crop-rotated fields. European Journal of Soil Biology
30:55-61.

Aschoff, J. 1960. Exogenous and endogenous components in
circadian rhythms. Pp. 11-28. In Biological Clocks. Cold Spring
Harbor Symposia on Quantitative Biology, Volume 25. Long
Island Biological Association. New York.

Bishop, S.C. 1950. The life of a harvestman. Nature Magazine
43:264-267, 276.

Bristowe, W.S. 1925. Notes on the habits of insects and spiders in
Brazil. Transactions of the Royal Entomological Society of
London 1924:475-504.

Capocasale, R. & L. Bruno-Trezza. 1964. Biologia de Acanthopachy-
lus aculeatus (Kirby, 1819), (Opiliones; Pachylinae). Revista de la
Sociedad Uruguaya de Entomologla 6:9-32.

Castanho, L.M. & R. Pinto-da-Rocha. 2005. Harvestmen (Opiliones:
Gonyleptidae) predating on treefrogs (Anura: Hylidae). Revista
Iberica de Aracnologia 1 1:43-45.

Cloudsley-Thompson, J.L. 1978. Biological clocks in Arachnida.

Bulletin of the British Arachnological Society 4:184-191.
Christensen. R. 1997. Log-Linear Models and Logistic Regression.
Second edition. Springer-Verlag Inc. New York, New York.
508 pp.

Edgar, A.L. 1971. Studies on the biology and ecology of Michigan
Phalangida (Opiliones). Miscellaneous Publications, Museum of
Zoology, University of Michigan, Ann Arbor 144:1-64.
Elpino-Campos, A., W. Pereira, K. Del-Claro & G. Machado. 2001.
Behavioural repertory and notes on natural history of the
neotropical harvestman Discocyrtus oliverioi (Opiliones: Gonylep-
tidae). Bulletin of the British Arachnological Society 12:144-150.
Foelix, R.F. 1996. Biology of Spiders. Second edition. Oxford
University Press, New York. 330 pp.

Friebe, B. & J. Adis. 1983. Entwicklungzyklen von Opiliones
(Arachnida) im Schwarzwasser-Uberschwemmungswald (Igapo)
des Rio Taruma Mirim (Zentralamazonien, Brasilien). Amazoni-
ana 8:101-1 10.

Gnaspini, P. 1996. Population ecology of Goniosonia spelaeum, a
cavernicolous harvestman from southeastern Brazil (Arachnida:
Opiliones: Gonyleptidae). Journal of Zoology 28:137-151.

526

THE JOURNAL OF ARACHNOLOGY

Gnaspini, P. & A.J. Cavallieiro. 1998. Chemical and behavioral
defenses of a Neotropical cavernicolous harvestman Goniosoma
spelaeum (Opiliones, Laniatores, Gonyleptidae). Journal of Ara-
chnology 26:81-90.

Gnaspini, P., F.H. Santos & S. Hoenen. 2003. The occurrence of
different phase angles between contrasting seasons in the activity
patterns of the cave harvestman Goniosoma spelaeum (Arachnida,
Opiliones). Biological Rhythm Research 34:31-49.

Hoenen, S. & P. Gnaspini. 1999. Activity rhythms and behavioral
characterization of two epigean and one cavernicolous harvestmen
(Arachnida, Opiliones, Gonyleptidae). Journal of Arachnology
27:159-164.

Kramer, M. & J. Schmidhammer. 1992. The chi-squared statistic in
ethology: use and misuse. Animal Behaviour 44:833-841.

Kury, A.B. 2003. Annotated catalogue of the Laniatores of the New
World (Arachnida, Opiliones). Revista Iberica de Aracnologia,
vol. especial monografico 1:1-337.

Lehner, P.N. 1996. Handbook of Ethological Methods. Garland
STPM Press, New York. 672 pp.

Machado, G. 2002. Maternal care, defensive behavior, and sociality
in Neotropical Goniosoma harvestmen (Arachnida, Opiliones).
Insectes Sociaux 49:1-6.

Machado, G. & P.S. Oliveira. 1998. Reproductive biology of the
Neotropical harvestman Goniosoma longipes (Arachnida, Opi-
liones: Gonyleptidae): mating and oviposition behaviour, brood
mortality, and parental care. Journal of Zoology 246:359-367.

Machado, G. & M.A. Pizo. 2000. The use of fruits by the Neotropical
harvestman Neosadocus variabilis (Opiliones, Laniatores, Gony-
leptidae). Journal of Arachnology 28:357-360.

Machado, G., R.L.G. Raimundo & P.S. Oliveira. 2000. Daily activity
schedule, gregariousness, and defensive behavior in the Neotrop-
ical harvestman Goniosoma longipes (Arachnida: Opiliones: Gony-
leptidae). Journal of Natural History 34:587-596.

Machado, G. & D.M. Vital. 2001. On the occurrence of epizoic algae
and liverworts on the harvestmen Neosadocus aff. variabilis
(Opiliones: Gonyleptidae). Biotropica 33:535-538.

Macias-Ordonez, R. 2000. Touchy harvestmen. Natural History
109:58-61.

Marc, P. 1990. Nycthemeral activity rhythm of adult Clubiona
corticalis (Walckenaer, 1802) (Araneae, Clubionidae). Acta Zool-
ogica Fennica 190:279-285.

Martens, J. 1993. Bodenlebende Arthropoda im zentralen Himalaya:
Bestansaufnahme, Wege zur Viefalt und Okologische Nischen.
Pp. 23-250. In Neue Forschungen im Himalaya. (U. Scheinfurth,
ed.). Erdkundliches Wissen. Volume 112. Franz Steiner Verlag,
Stuttgart.

Martin, P. & P. Bateson. 1993. Measuring Behaviour: an Introduc-
tory Guide. Second edition. Cambridge University Press, New
York. 222 pp.

Mora, G. 1990. Parental care in a Neotropical harvestman,
Zygopachylus albomarginis (Arachnida, Opiliones: Gonyleptidae).
Animal Behaviour 39:582-593.

Pabst, W. 1953. Zur Biologie der mitteleuropaischen Troguliden.
Zoologische Jahrbucher, Abteilung fiir Systematik, Okologie und
Geographic der Tiere 82:1-156.

Parisot, C. 1962. Etude de quelques opilions de Lorraine. Vie et
Millieu 13:179-197.

Pereira, W., A. Elpino-Campos, K. Del-Claro & G. Machado. 2004.
Behavioral repertory of the Neotropical harvestman Ilhaia
cuspidata (Opiliones, Gonyleptidae). Journal of Arachnology
32:22-30.

Pfeifer, H. 1956. Zur Okologie und larvalsysteinatik der Weber-
knechte. Mitteilungen aus dem Zoologischen Museum in Berlin
32:59-104.

Phillipson, J. 1959. The seasonal occurrence, life histories and
fecundity of harvest-spiders (Phalangida, Arachnida) in neighbor-
hood of Durham City. Entomologist's Monthly Magazine
95:134-138.

Polis, G.A. 1980. Seasonal patterns and age-specific variation in the
surface activity of a population of desert scorpions in relation to
environmental factors. Journal of Animal Ecology 49:1-18.

Polis, G.A. & R.D. Farley. 1979a. Behavior and ecology of mating in
the cannibalistic scorpion, Paruroctonus mesaensis Stahnke (Scor-
pionida: Vaejovidae). Journal of Arachnology 7:3-46.

Polis, G.A. & R.D. Farley. 1979b. Characteristics and environmental
determinants of natality, growth and maturity in a natural
population of the desert scorpion Paruroctonus mesaensis (Scor-
pionida: Vaejovidae). Journal of Zoology 187:517-542.

Punzo, F. 1998. The Biology of Camel Spiders (Arachnida,
Solifugae). Kluwer Academic Publishers, Dordrecht, The Nether-
lands. 301 pp.

Todd, V. 1949. The habits and ecology of the British harvestmen
(Arachnida, Opiliones), with special reference to those of the
Oxford District. Journal of Animal Ecology 18:209-229.

Tsurusaki. N. 2003. Phenology and biology of harvestmen in and near
Sapporo, Hokkaido, Japan, with some taxonomical notes on
Nelima suzukii n.sp. and allies (Arachnida: Opiliones). Acta
Arachnologica 52:5-24.

Warburg, M.R. & G.A. Polis. 1990. Behavioral responses, rhythms,
and activity patterns. Pp. 224-246. In The Biology of Scorpions.
(G.A. Polis, ed.). Stanford University Press, Stanford, California.

Willemart, R.H. 2001. Egg covering behavior of the neotropical
harvestman Promitobates ornatus (Opiliones, Gonyleptidae). Jour-
nal of Arachnology 28:249-252.

Willemart, R.H. 2002. Cases of intra- and inter-specific food
competition among Brazilian harvestmen in captivity (Opiliones,
Laniatores, Gonyleptidae). Revue Arachnologique 14:49-58.

Willemart, R.H. 2007. Rearing and maintenance of harvestmen in
captivity. Pp. 520-524. In Harvestmen: The Biology of Opiliones.
(R. Pinto-da-Rocha, G. Machado & G. Giribet, eds.). Harvard
University Press, Cambridge, Massachusetts.

Willemart. R.H., M.C. Chelini, R. de Andrade & P. Gnaspini. 2007.
An ethological approach to a SEM survey on sensory structures
and tegumental gland openings of two neotropical harvestmen
(Arachnida, Opiliones, Gonyleptidae). Italian Journal of Zoology
74:39-54.

Willemart, R.H.. J.P. Farine, A.V. Peretti & P. Gnaspini. 2006.
Behavioral roles of the sexually dimorphic structures in the male
harvestman Phalangium opilio (Opiliones. Phalangiidae). Canadian
Journal of Zoology 84:1763-1774.

Willemart, R.H. & P. Gnaspini. 2003. Comparative density of hair
sensilla on the legs of cavernicolous and epigean harvestmen
(Arachnida: Opiliones). Zoologisher Anzeiger 242:353-366.

Willemart, R.H. & P. Gnaspini. 2004a. Spatial distribution,
displacement, gregariousness and defensive behavior in the
Brazilian cave harvestman Goniosoma albiscriptum (Arachnida,
Opiliones, Laniatores). Animal Biology 54:221 235.

Willemart, R.H. & P. Gnaspini. 2004b. Breeding biology of the
cavernicolous harvestman Goniosoma albiscriptum (Arachnida,
Opiliones, Laniatores): sites of oviposition, egg-batches, charac-
teristics and subsocial behaviour. Invertebrate Reproduction and
Development 45:15-28.

Williams, G.C. 1962. Seasonal and diurnal activity of harvestmen
(Phalangida) and spiders (Araneida) in contrasted habitats.
Journal of Animal Ecology 31:23-42.

Manuscript received 29 August 2006. revised 26 February 2008.

2008. The Journal of Arachnology 36:527-532

First male sperm precedence in multiply-mated females of the cooperative spider Anelosimus studiosus

(Araneae, Theridiidae)

Thomas C. Jones: Department of Biological Sciences, East Tennessee State University, Johnson City, Tennessee 37614,
USA. E-mail: jonestc@etsu.edu

Patricia G. Parker: Department of Biology, University of Missouri at St. Louis, St. Louis, Missouri 63121, USA

Abstract. Patterns of sperm usage in multiply-mated females have profound fitness consequences for males, and create
strong selective pressure on male behavior. In the cooperative theridiid spider Anelosimus studiosus Hentz 1850 adult males
are tolerated in females’ webs, and females have been observed to mate multiply with different males. In this experiment,
virgin females were mated with two different males on consecutive days under controlled conditions to determine paternity
patterns and behavioral responses of males to non-virgin females. The paternity of broods was analyzed using randomly
amplified polymorphic DNA (RAPDs). Fifteen broods were analyzed and complete first male sperm precedence was
found. Mating behavior differed between first and second males with the first males attempting fewer intromissions, but
having a longer total time of intromission. This suggests that the second males are either prevented from normal
copulation, or are reacting to the different condition of the females. The sperm precedence pattern is discussed with respect
to its ramifications for male behavior, juvenile inclusive fitness, and the evolution of cooperative behavior.

Keywords: Social spiders, mating behavior. RAPDs, sexual selection

When females mate multiply, sperm precedence patterns can
affect the fitness of all individuals involved. For males who
have both mated with the same female, the fitness conse-
quences are clear and directly related to the proportion of her
brood they have fertilized (Trivers 1972). Because of this,
males often compete with each other for access to females, or
to be chosen as mates by females (Andersson 1994). Males
may also compete for fertilizations after copulation through
such avenues as mate guarding or copulatory plugs, the
necessity or efficacy of which is affected by sperm precedence
(Parker 1984). Patterns of sperm precedence may affect the
female’s fitness by influencing the genetic variability of her
brood, or the proportions of her brood fertilized by males of
differing quality. There is also accumulating evidence of
females manipulating fertilization patterns of their broods in
response to male attributes (Eberhard 1996). Examples of this
in spiders include selective sperm storage in response to
copulation duration (Bukowski & Christenson 1997a & b),
and the fact that, with paired spermathecae, spiders may be
pre-adapted for paternity manipulation (Snow & Andrade
2005).

Patterns of sperm precedence will also affect the composi-
tion of full and half-sibs within broods of polyandrous
females. In social species, the relatedness among brood-mates
can have profound effects on their fitness (Hamilton 1964).
The relatedness among group members is therefore important
for a complete understanding of the selective costs and
benefits to group living (for review see Caraco & Giraldeau
1991).

Mating systems are particularly important to the evolution
of sociality in spiders. Social spider colonies are generally
inbred (Riechert & Roeloffs 1993; Johannesen et al. 2002). In
fact, genetic analysis of Anelosimus eximius Keyserling 1884
colonies suggests that there is no gene Bow at all among
colonies (Smith & Hagen 1996). It is likely that the cooperative
behaviors and female-biased sex ratios of cooperatively social
spider species are maintained by interdemic selection, fostered

by the high levels of relatedness among colony members
(Aviles 1997). Isolated local populations of asocial or
subsocial spiders will become inbred through genetic drift,
which could then promote the evolution of cooperative
behaviors. The rate at which a population loses genetic
diversity (i.e., the effective population size) is affected by its
mating system in that monogamous populations lose diversity
faster than promiscuous populations (Parker & Waite 1997).
This inbreeding is likely an important factor in the evolution-
ary transition from subsociality to permanent sociality in
spiders (Bilde et al. 2005).

Anelosimus studiosus Hentz 1850 is a relatively small (about
8 mm long) theridiid spider which ranges from Argentina to
New England (Agnarsson 2006; Agnarsson et al. 2007). This
species is common in the southeastern USA and can be found
in extremely high densities along waterways (Jones et al. 2007).
This species is described as subsocial (Wilson 1971) and,
specifically, “prolonged subsocial” (Rayor & Taylor 2006), in
that juveniles and adult males are tolerated in an adult
female’s web, but other adult females usually are not (Brach
1977; but see Furey 1998; Jones et al. 2007). Previous
experiments have demonstrated that, under controlled labo-
ratory conditions, colony prey capture increases with the age
and number of juveniles in the colony (though resources per
individual decline with colony size), and variation in prey mass
decreases with the number of juveniles present (Jones & Parker
2000). Females can produce up to at least three broods over
their lives usually with several weeks between broods (Jones,
unpubl. data). It has also been shown that in semi-natural
conditions, delayed juvenile dispersal benefits juvenile survi-
vorship and development as well as the mother’s ability to
produce future broods (Jones & Parker 2002). While it is clear
that individual juveniles are better off in their natal group than
on their own, the exact relationship of individual fitness to
group size is not yet known. Whatever this relationship, a
juvenile’s fitness is likely to be affected by its relatedness to its
brood-mates. We have observed that females will mate

527

528

THE JOURNAL OF ARACHNOLOGY

multiply if presented with more than one male, so broods may
be composed of either full-sibs, or a mixture of full and half-
sibs.

Members of the family Theridiidae are “entelegyne”
spiders, in that the female reproductive tract has a conduit
morphology, with sperm leaving the storage organ to fertilize
eggs through a different opening than that into which they
were deposited (Foelix 1996). It has been suggested that this
morphology will put the first mate’s sperm closest to the point
of fertilization, and thus lead to first mate sperm precedence
(Austad 1984). However, studies of sperm precedence in
entelegyne spiders have yielded estimates of proportions of
first male from 0.95 to only 0.37 (reviewed by Elgar 1998). In
the theridiid Australian redback spider the mean first male
precedence was 0.44, but ranged from 0.0 to 1.0 (Andrade
1996). In this study we seek to determine sperm precedence in
A. studiosus by sequentially pairing females with two males,
recording mating behavior, and using the animals’ DNA and
RAPD analysis to determine parentage.

METHODS

Collection and Rearing. — Anelosimus studiosus colonies
containing juveniles were collected in southern Louisiana in
September of 1999 from bayous in Tickfaw State Park
(30°22'N, 90°37'W) and Fontainebleau State Park (30°20'N,
90 02' W). The bayous were accessed by pirogue and the webs
were mostly collected from low-hanging cypress branches
(voucher specimens are deposited in the Museum of Biological
Diversity at The Ohio State University, Columbus, Ohio). We
maintained the colonies in plastic containers (18 X 14 x 6 cm)
that were laced with sewing thread to provide substrate for
web building. The colonies were kept at buffered ambient
temperatures (20-28 C) under natural light conditions, were
fed Drosophila melanogaster and D. virilis ad libidum, and
misted with distilled water three times a week. We kept the
colonies in a greenhouse under natural lighting conditions. In
order to ensure the virginity of experimental animals, as the
juveniles approached maturity, females were isolated in new
containers, and the penultimate males were grouped together
(by natal colony) in another set of containers.

Experimental Matings. — Females that had undergone their
final molt in isolation were selected for this experiment.
Twenty-five females from eight different colonies, but no more
than four females from any one colony, were used. We
released males from two different colonies (other than the
colony of their prospective mate) into a female’s container one
at a time, on consecutive days. Matings were videotaped with
a Sony digital 8 camera for behavioral analysis. The resolution
of the video did not allow fine details of palpal insertions, such
as the extension of the embolus, to be observed. We estimated
copulation as periods when the male’s palps were resting
stationary against the female’s genital opening. The onset of
copulation was recorded as when the palp would ease against
the genital opening and stop, and the end of copulation was
recorded when the male pushed against the female and broke
loose with a conspicuous jerk. We quantified the number and
timing of copulations and copulation attempts. After mating,
the males were frozen for later DNA extraction. Mating
usually commenced within a few minutes after the introduc-
tion of the males. We removed the males after they had broken

copulation, moved off from the female, and for ten minutes
showed no further attempt at copulation. On four occasions
the males had not attempted copulation after 15 min. These
males were removed from the container and the process was
restarted with new males. Abstinent males were not reused.
After the females had mated the second time, we returned
them to the greenhouse rearing conditions. Females that
produced broods were allowed to rear them through the third
instar, after which the female and juveniles were frozen for
DNA extraction.

DNA Extraction and PCR. — To the 1.5 ml tubes containing
the frozen spiders, 200 pi of CTAB and 1.5 pi Proteinase K
(100 pg/ml) were added. The spiders were thoroughly ground
with a pestle in the tube and incubated at 60° C for 1 h. One
extraction with 100 pi of phenol and 100 pi of CIA (24:1
chloroforrmisoamyl alcohol), and one extraction with 200 pi
of CIA were performed. The samples were ethanol precipitat-
ed, resuspended in 100 pi of distilled water, and stored at 4° C.
The concentration of the samples was estimated by gel
comparison with concentration standards.

Randomly amplified polymorphic DNA (RAPDs) uses
single relatively non-specific ten base pair primers (synthesized
at OSU) to amplify regions of the genome that contain
complementary primer annealing sites. The regions that are
amplified are arbitrary but heritable, and therefore, useful
(Williams et al. 1990). Under similarly controlled mating
conditions, RAPDs were used to assign sperm precedence in a
beetle (Carbone & Rivera 2003). The reactants for an
individual 14 pi reaction consisted of: 9.9 pi, UV irradiated
distilled water; 1.5 pi of 1 pM dNTPs; 1.5 pi reaction buffer
(10 mM Tris HC1, pH 8.3; 50 mM KC1; 2 mM MgCl); 0.8 pi of
10 pM primer; 0.1 pi Taq DNA polymerase (5 U/pl); 1.2 pi
template DNA (approx. 25 ng/pl). The reactions were run
through four initial “touch down” cycles (94° C for 1 min;
35° C for 1 min; 0.3 slope to 72° C for 2 min), and then 32
amplification cycles (94° C for 10 s; 35° C for 30 s; 12° C for
30 s). The finished reactions were held at 4°C until they were
visualized. For visualization, the amplified products were run
out on a 1.2% agarose gel (80-120 V), stained with ethidium
bromide, then visualized and photographed under UV light.

Paternity analysis. — RAPDs are dominant markers, and
band presence/absence is particularly sensitive to reaction
conditions because of the short length of the primers.
Therefore, repeatability of RAPD markers has been prob-
lematic, making them not as robust in parentage analyses as
some other molecular techniques (e.g., microsatellites or
mullilocus minisatellite DNA fingerprinting; for review see
Parker et al. 1998). In this experiment, however, RAPDs
were useful to assess paternity of broods because the pool of
potential fathers is limited and known, and because
repeatability was confirmed. A unique bands analysis was
used to assign the father of each brood member. On the gels,
the mother and two potential fathers were run as triads
twice, flanking the offspring lanes. Bands that were observed
in the lanes of one of the males, but not in the lanes of the
other male or mother, were scored for their presence in the
offspring lanes. Multiple primers were screened for the
families until a total of at least two diagnostic bands were
found for each juvenile. Such an analysis is simple and
robust since no inference is made from band absence, and the

JONES & PARKER SPERM PRECEDENCE IN A COOPERATIVE SPIDER

529

repeatability of each diagnostic band is confirmed by
amplifying the triads of adults twice, and running them on
the flanking lanes on both sides of the gels.

RESULTS

Mating behavior. — When the males were placed in the
containers near the females, they would typically remain
motionless for up to 1 min. They would then begin to move
around in the web while rapidly drumming their first pair of
legs on the silk. The movement of the males appeared
undirected until the females moved within the web, at which
point the males would begin to move toward the females while
still drumming. As the males approached, the females would
typically bounce in the web apparently signaling sexual
receptivity because the males would move more quickly
toward them afterwards. The males continued drumming even
as they made contact with the females. The males would orient
themselves to face the same direction as the female, with their
ventral surfaces adjacent, but with no consistent absolute
orientation. As the males moved into position, the amplitude
of their drumming eased to a stop, which was taken to be the
onset of copulation. After copulation the spiders separated
with a conspicuous jerk, followed by the males moving a short
distance from the females (1-2 cm). If only one copulation had
taken place, the males would resume drumming and repeat the
courtship, but would typically move in more quickly and
insert on the other side. The females in this experiment, in all
cases, appeared receptive to both males. Also, no occurrences
or apparent attempts of sexual cannibalism were observed.

As measured by the number of copulation onsets and
breaks, first males had fewer copulations (mean 2.2, range 2-
4) than second males (mean 3.7, range 2-12; Mann-Whitney U
= 489, P < 0.001, Fig. 1). Only three of the 25 first males had
more than two copulations, and in those cases there were one
or two short copulations followed by two long ones. The total
time spent in copula was longer for first males (mean 44.3 min,
range 34.7-52.4) than for second males (mean 15.9 min, range
1.5-42.1; t = 11.3, P < 0.001, Fig. 1). In six cases the second
males had more than three copulations, but their total time of
copulation (mean = 11.0 min) was significantly shorter than
the second males that had three or fewer copulations (mean =
19.6; t = 1.9, P = 0.04). Considering individual females, their
first mate’s total time of copulation was not related to their
second mate’s total time (Fig. 2). The duration of the first
male’s first copulation was a strong predictor of the duration
of his second copulation (in cases where there were more than
two copulations, the initial apparent “false starts’’ [copula-
tions lasting less than one minute] were excluded: Fig. 3).

Paternity analysis. — Of the 25 females in the mating
experiment, 22 produced egg cases, of which 17 had juveniles
emerge (which was considerably lower than the mean of 36
juveniles observed in nature; Brach 1977). The average
number of juveniles per family was 11.6 (range 4-21). Two
of the families were unusable because one of the males (in each
family) could not be amplified by PCR. Of the fifteen
remaining families, all 168 of the juveniles were assigned to
the first male. While the assignments were based on the
presence of at least two of the first male’s unique bands, the
lack of the second male’s unique bands in juvenile lanes
further confirmed the assignments.

Male Male Male Male

Figure 1. — Comparison of the mean number of copulations and
total duration of copulation between first and second males in
Anelosimus studiosus. Reported are means with standard error bars.

DISCUSSION

Using a direct DNA-based analysis of parentage, we found
complete first male sperm precedence in A. studiosus. We also
found significant differences in the mating behavior of first
and second males. The patterns of sperm precedence, the
increased frequency of starting and stopping copulation with
less actual time spent in copula of second males, suggests that
the second males could not successfully copulate.

There is evidence from other spider species that second
males can be prevented from successfully copulating. Copu-
lation “plugs” made by a hardening of seminal fluids in the
female reproductive tract or by the tip of the male’s
intromittent organ breaking off, have been reported for
several spider species Phidippus johnsoni Peckham & Peckham
1983 (Jackson 1980), Agelena limbata Thorell 1897 (Matsu-
moto 1993; Yoward & Oxford 1996; Schneider et al. 2005).
However, the efficacy of these plugs in preventing subsequent

T otal Time of F irst Male Copulation

Figure 2. — Plot of a female’s total time of copulation with the first
male versus her total time of copulation with the second male in
Anelosimus studiosus.

530

THE JOURNAL OF ARACHNOLOGY

q 15 20 25 30

Duration of First Male’s First
Copulation (Min)

Figure 3. — Plot of the duration of the first male’s first copulation
versus the total time of his second copulation in Anelosimus studiosus.

fertilizations is mixed among taxa, and they may instead
function to increase the sperm retention and survival
(reviewed in Huber 2005). In another theridiid spider,
fertilization patterns are affected by where the broken organ
tip rests in the female reproductive tract (Snow et al. 2006).
Across spider taxa, male genital mutilation appears to be an
indicator of strong selection on paternity protection, being
correlated with the evolution of male sacrifice behavior and
size dimorphism (Miller 2007). We did not determine in this
study whether or not the second male was in any way
prevented from copulation. The fact that four of the second
males’ copulatory behavior was similar to the first males’
(having two intromissions totaling > 30 min), suggests that
copulatory plugs are not ubiquitous, or at least not completely
effective. However, we have observed a second male repeat-
edly moving in and breaking from a female, apparently
struggling to insert his palps.

Interestingly, first male total copulation time was only 16%
that observed in a congener (Klein et al. 2005). The fact that
second males spent less total time in copulation contradicts
previous findings in an araneid spider (Micrathena gracilis
Walckenaer 1805) in which second males copulated over twice
as long as the first male (Bukowski & Christenson 1997b), but
similar to patterns found in a tetragnathid spider ( Tetragnatha
versicolor Walckenaer 1842; Danielson-Francois & Bukowski
2004). In the latter case sperm release was equal between first
and second males. A similar experiment in which males were
introduced to females that had only been mated on one side of
their tract suggested that males are responding specifically to
the reproductive status of the female’s reproductive tract,
rather than the female’s behavior or overall condition
(Bukowski et al. 2001). Second males also copulated for
shorter periods in a cellar spider, P/iolcus phalagioides Fuesslin
1775 (Schafer & Uhl 2002). In this case, however, the second
males had a higher proportion of reproductive success,
apparently as a result of their ability to remove the first
male’s sperm.

Since second males generally attempted copulation, the
differences in their mating behavior seems most likely because
they are prevented from normal copulation. The possibility

remains however, that they could be altering their behavior in
response to the previously mated condition of the female.
There are examples of non-virgin spiders being less attractive
to males. The presence of sex pheromones has been
documented across a wide range of spider species and, in
some cases, these pheromones are volatile (Shultz & Toft 1993;
Miyashita & Hayashi 1996; Rovner 1996; Costa et al. 1997;
Searcy et al. 1999), and in other cases are contact based
(Trabalon et al. 1997, 1998). In one linyphiid spider (Nereine
litigiosa Keys 1886) pheromones are incorporated in the
female's web, and males destroy the web prior to mating,
reducing the probability that a second male will find her
(Watson 1986). Whether pheromones exist in this species is
not known, but pheromone-like compounds have been
extracted from the cuticle of its congener A. eximius (Bagneres
et al. 1997). It is possible that, even though the female remains
sexually receptive after mating, her production of pheromones
decreases, thus making her less attractive to second males. The
speed with which the first and second males begin drumming
and searching for the female might give insight into
pheromone levels, but such a measure would be confounded
by the introduction of the males, which was not standardized
in terms of their distance from the female.

This study found complete first male sperm precedence,
accentuating the question why an A. studiosus female should
mate multiply at all. There are many potential costs to mating
such as loss of foraging opportunities, increased predation
risk, and disease transmission (reviewed in Lewis 1987). Male
spiders provide no parental care, and it is unlikely that any
substances that males transfer along with sperm provide direct
benefits to the females as has been observed in some insects
(Gwynne 1984; Boggs 1990). With linyphiid spiders the male
cohabitats in the female’s web. eating prey, until he has mated
with her; females apparently mate with these males to induce
them to leave (Watson 1993). However, there is evidence from
the same species that multiple mating has indirect benefits in
terms of the size and growth rate of juveniles (Watson 1998).
In the pisaurid species Pisaura mirabilis (Clerk 1757), in which
females mate with multiple males, males present nuptial prey
items (Drengsgaard & Toft 1999). Mating multiply allows the
possibility of cryptic female choice in which she chooses the
sperm of the male she prefers (Eberhard 1996). Again, given
complete first male precedence, this seems unlikely to be
occurring with this species. Perhaps the simplest benefit to a
female from multiple mating would be to ensure that all her
eggs get fertilized. It may also be that there is no selective
benefit to mating with multiple males, and that A. studiosus
females simply remain receptive from maturity until their
abdomens are distended with eggs regardless of the number of
times they have mated.

Previous work on this species has demonstrated that by
delaying dispersal and remaining part of their natal colony,
juveniles enhance their survival and development (Jones &
Parker 2002). This can be extrapolated to suggest that
juveniles’ direct fitness benefits from delayed dispersal. This
study finds complete first male sperm precedence within the
broods of doubly-mated A. studiosus females. This suggests
that if an individual juvenile’s presence in the colony
contributes to the survivorship of its brood-mates, its indirect
fitness (sensu Hamilton 1964) would be maximized because

JONES & PARKER— SPERM PRECEDENCE IN A COOPERATIVE SPIDER

531

they are all full-sibs. In this species it is now documented that
there are colonies that contain multiple adult females in North
America (Furey 1998), the incidence of which increases with
latitude (Jones et al. 2007). If these colonies develop by- non-
dispersal of juveniles, fertilization patterns could have
profound effects on the genetic structure of these large
colonies.

Finding complete first-male sperm precedence may be
surprising, but these results should be taken with some
caution given the highly controlled conditions. Factors such
as the number and timing of matings, which may influence
precedence (Eberhard 1996), were held constant. Currently
studies are underway exploring relatedness within and among
natural colonies using microsatellite loci.

ACKNOWLEDGMENTS

We wish to thank the Department of Evolution, Ecology,
and Qrganismal Biology at The Ohio State University. Special
thanks to George Keeney and Lisa Wallace for technical
assistance. We also thank T. Grubb, E. Marschall, G. Uetz,
and members of the Parker lab for useful discussion. Thanks
also to G. Stratton and two anonymous reviewers for their
insightful comments.

LITERATURE CITED

Agnarsson, I. 2006. Revision of the new world eximius lineage of
Anelosimus and a phylogenetic analysis using worldwide exem-
plars. Zoological Journal of the Linnean Society 146:453-593.
Agnarsson, L, W.P. Maddison & L. Aviles. 2007. The phylogeny of
the social Anelosimus spiders (Araneae: Theridiidae) inferred from
six molecular loci and morphology. Molecular Phylogenics and
Evolution 43:833-851.

Andersson, M. 1994. Sexual Selection. Princeton University Press,
Princeton, New Jersey. 624 pp.

Andrade, M.C.B. 1996. Sexual selection for male sacrifice in the
Australian redback spider. Science 271:70-72.

Austad, S.N. 1984. Evolution of sperm priority patterns in spiders.
Pp. 233-249. In Sperm Competition and the Evolution of Animal
Mating Systems. (R.L. Smith, ed.). Academic Press, London.
Aviles, L. 1997. Causes and consequences of cooperation and
permanent sociality in spiders. Pp. 476-498. In The Evolution of
Social Behavior in Insects and Arachnids. (J.C. Choe & B.J.
Crespi, eds.). Cambridge University Press, Cambridge, UK.
Bagneres, A.G., M. Trabalon, G.J. Blomquist & S. Schulz. 1997.
Waxes of the social spider Anelosimus eximius (Araneae, Theridi-
idae): Abundance of novel n-propyl esters of long-chain methyl-
branched fatty acids. Archives of Insect Biochemistry and
Physiology 36:295-314.

Bilde, T., Y. Lubin, D. Smith, J. Schneider & A.A. Maklakov. 2005.
Transition to social inbred mating systems: role of inbreeding
tolerance in a subsocial predecessor. Evolution 59:160-174.

Boggs, C.L. 1990. A general model of the role of male-donated
nutrients in female insects’ reproduction. American Naturalist
136:598-617.

Brach, V. 1977. Anelosimus studiosus (Araneae: Theridiidae) and the
evolution of quasisociality in theridiid spiders. Evolution 31:
154-161.

Bukowski, T.C. & T.E. Christenson. 1997a. Natural history and
copulatory behavior of the spiny orbweaving spider Micrathena
gracilis (Araneae, Araneidae). Journal of Arachnology 25:307-320.
Bukowski, T.C. & T.E. Christenson. 1997b. Determinants of sperm
release and storage in a spiny orbweaving spider. Animal
Behaviour 53:381-395.

Bukowski, T.C., C.D. Lynn & T.E. Christenson. 2001. Copulation
and sperm release in Gastracantha canceriformis (Araneneae:
Arandeidae): differential male behaviour based on female mating
history. Animal Behaviour 62:887-895.

Caraco, T. & L.-A. Giraldeau. 1991. Social foraging: producing and
scrounging in a stochastic environment. Journal of Theoretical
Biology 153:559-589.

Costa, F.G., C. Viera & G. Francescoli. 1997. Male sexual behavior
elicited by a hybrid pheromone: a comparative study on Lycosa
thorelli, L. carbonelli, and their progeny (Araneae, Lycosidae).
Canadian Journal of Zoology 75:1845-1856.

Oanielson-Francois, A.M. & T.C. Bukowski. 2004. Female mating
history influences copulation behavior but not sperm release in the
orb-weaving spider Tetragnatha versicolor (Araneae, Tetragnathi-
dae). Journal of Insect Behavior 18:131-148.

Drengsgaard, I.L. & §. Toft. 1999. Sperm competition in a nuptial
feeding spider, Pisaura mirabilis. Behaviour 136:877-897.

Eberhard, W.G. 1996. Female Control: Sexual Selection by Cryptic
Female Choice. Princeton University Press, Princeton, New Jersey.
472 pp.

Elgar, M.A. 1998. Sexual selection and sperm competition in
arachnids. Pp. 307-337. In Sperm Competition and Sexual
Selection. (T.R. Birkhead & A.P. Moller, eds.). Academic Press,
London.

Foelix, R.F. 1996. Biology of Spiders, Second edition. Oxford
University Press, New York. 330 pp.

Furey, R.E. 1998. Two cooperatively social populations of the
theridiid spider Anelosimus studiosus in a temperate region. Animal
Behaviour 55:727-735.

Gwynne, D.T. 1984. Courtship feeding increases female reproductive
success in bush crickets. Nature 307:361-363.

Hamilton, W.D. 1964. The genetical evolution of social behavior I &
II. Journal of Theoretical Biology 7:1-52.

Huber. B.A. 2005. Sexual selection research on spiders: progress and
biases. Biological Reviews 80:363-385.

Jackson, R.R. 1980. The mating system of Phidippus johnsoni
(Araneae, Salticidae): II. Sperm competition and the function of
copulation. Journal of Arachnology 8:217-240.

Johannesen, J., A. Hennig, B. Dommermuth & J.M. Schmeider. 2002.
Mitochondrial DNA distributions indicate colony propagation
by single matri-Iineages in the social spider Stegodyphus dumi-
cula (Eresidae). Biological Journal of the Linnean Society
76:591-600.

Jones, T.C. & P.G. Parker. 2000. Costs and benefits of foraging
associated with delayed dispersal in the spider Anelosimus studiosus
(Araneae: Theridiidae). Journal of Arachnology 28:61-69.

Jones, T.C. & P.G. Parker. 2002. Delayed juvenile dispersal benefits
both mother and offspring in the cooperative spider Anelosimus
studiosus (Araneae: Theridiidae). Behavioral Ecology 13:142-148.

Jones, T.C., S.E. Riechert, S.E. Dalrymple & P.G. Parker. 2007.
Fostering model explains variation in levels of sociality in a spider
system. Animal Behaviour 73:195-204.

Klein, B.A., T.C. Bukowski & L. Aviles. 2005. Male residency and
mating behavior in a subsocial spider. Journal of Arachnology
33:703-710.

Lewis, W.M. Jr. 1987. The cost of sex. Pp. 33-57. In The Evolution of
Sex and Its Consequences. (S.C. Steams, ed.). Birkhuser-Verlag,
Basel.

Matsumoto, T. 1993. The effect of the copulatory plug in the funnel-
web spider Agelena limbata (Araneae: Agelenidae). Journal of
Arachnology 21:55-59.

Miller, J.A. 2007. Repeated evolution of male sacrifice behavior in
spiders correlated with genital mutilation. Evolution 61:1301-1315.

Miyashita, T. & H. Hayashi. 1996. Volatile chemical cue elicits
mating behavior of cohabiting males of Nephila clavata (Araneae,
Tetragnathidae). Journal of Arachnology 24:9-15.

532

THE JOURNAL OF ARACHNOLOGY

Parker, G.A. 1984. Sperm competition and the evolution of animal
mating strategies. Pp. 1-60. In Sperm Competition and the
Evolution of Animal Mating Systems. (R.L. Smith, ed.). Academic
Press, London.

Parker, P.G., A. A. Snow, M.D. Schug, G.C. Booten & P.A. Fuerst.
1998. What molecules can tell us about populations: choosing and
using a molecular marker. Ecology 79:361-382.

Parker, P.G. & T.A. Waite. 1997. Mating systems, effective
population size and conservation of natural populations.
Pp. 243-261. In Behavioral Approaches to Conservation in the
Wild. (J. Clemmons & R. Buckholz, eds.). Cambridge University
Press, Cambridge, UK.

Rayor, L.S. & L.A. Taylor. 2006. Social behavior in anrblypygids,
and reassessment of arachnid social patterns. Journal of Arachnol-
ogy 34:399-42 1 .

Riechert, S.E. & R.M. Roeloffs. 1993. Evidence for and consequences
of inbreeding in the cooperative spiders. Pp. 283-303. In The
Natural History of Inbreeding and Outbreeding. (N.W. Thornhill,
ed.). University of Chicago Press, Chicago.

Rovner, J.S. 1996. Conspecific interactions in the lycosid spider
Rabidosa rabida : the roles of different senses. Journal of
Arachnology 24:16-23.

Searcy, L.E., A.L. Rypstra & M.H. Persons. 1999. Airborne chemical
communication in the wolf spider Pardosa milvina. Journal of
Chemical Ecology 25:2527-2533.

Schafer, M.A. & G. Uhl. 2002. Determinants of paternity success in
the spider Pholcus phalangioides (Pholcidae: Araneae): the role of
males and female mating behaviour. Behavioral Ecology and
Sociobiology 51:368-377.

Schneider, J.M., L. Fromhage & G. Uhl. 2005. Copulation patterns in
the golden orb-web spider spider Nephila madagascariensis.
Journal of Ethology 23:51-55.

Shultz, Z. & S. Toft. 1993. Identification of a sex pheromone from a
spider. Science 260:1635-1637.

Smith, D.R. & R.H. Hagen. 1996. Population structure and
interdemic selection in the cooperative spider Anelosimus eximius.
Journal of Evolutionary Biology 9:589-608.

Snow, L.S.E., A. Abdel-Mesih & M.C.B. Andrade. 2006. Broken
copulatory organs are low-cost adaptations to sperm competition
in redback spiders. Ethology 112:379-389.

Snow, L.S.E. & M.C.B. Andrade. 2005. Multiple sperm storage
organs facilitate female control of paternity. Proceedings of the
Royal Society B-Biological Sciences 272:1 139-1 144.

Trabalon, M., A.G. Bagneres & C. Roland. 1997. Contact sex signals
in two sympatric spider species, Tegenaria domestica and Tegenaria
pagana. Journal of Chemical Ecology 23:747-758.

Trabalon, M„ G. Pourie & N. Hartmann. 1998. Relationships among
cannibalism, contact signals, ovarian development and ecdysteroid
levels in Tegenaria atrica (Araneae, Agelenidae). Insect Biochem-
istry and Molecular Biology 28:751-758.

Trivers, R.L. 1972. Parental investment and sexual selection.
Pp. 136-179. In Sexual Selection and the Descent of Man, 1871-
1971. (B. Campbell, ed.). Aldine-Atherton Publishing Company,
Chicago.

Watson, P.J. 1986. Transmission of female sex pheromone by males
in the spider Linyphia litigiosa Keyserling (Linyphiidae). Science
233:219-221.

Watson, P.J. 1993. Foraging advantage of polyandry for female sierra
dome spiders (Linyphia litigiosa , Linyphiidae) and assessment of
alternative direct benefit hypotheses. American Naturalist
141:440-465.

Watson, P.J. 1998. Multi-mating and female choice increase offspring
size and growth in the spider Neriene litigiosa (Linyphiidae).
Animal Behaviour 55:387-403.

Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski & S.V.
Tingey. 1990. DNA polymorphisms amplified by arbitrary primers
are useful as genetic-markers. Nucleic Acids Research
18:6531-6535.

Wilson. E.O. 1971. The Insect Societies. Belknap Press of Harvard
University Press, Cambridge, Massachusetts. 548 pp.

Yoward. P. & G. Oxford. 1996. Single palp usage during copulation
in spiders. Newsletter of the British Arachnological Society 77:8-9.

Manuscript received 17 December 2006, revised 6 March 2008.

2008. The Journal of Arachnology 36:533-537

Frequency and consequences of damage to male copulatory organs in a widow spider

Michal Segoli: Department of Life Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel. E-mail:
msegoli@bgu.ac.il

Yael Lubin: Mitrani Department of Desert Ecology, Jacob Blaustein Institute for Desert Research, Ben-Gurion
University of the Negev, Sede Boqer Campus, Israel

Ally R. Harari: Department of Entomology, Agricultural Research Organization, The Volcani Center, Bet Dagan, Israel,
and Department of Life Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel

Abstract. Copulatory organ breakage, in which a portion of the male’s genitalia breaks off and remains in or attached to
the female’s genitalia may represent a male strategy of high investment in a single mating. Such a strategy is expected when
mating opportunities for males are limited and competition for females is high. We studied costs and benefits for males as a
consequence of male organ breakage in the white widow spider (Latrodectus pallidus O. Pickard-Cambridge 1872). In order
to estimate the frequency and consequences of such damage we provided each male with four virgin females simultaneously
in an outdoors enclosure. We recorded male mating success and loss of the tip of the embolus (the male intromittent organ)
inside the female’s genitalia for each male. In order to test the effect of the broken tip as a mating plug, we collected females
from natural populations and observed the location of embolus tips inside their genitalia. We found that damage to the
male organ was frequent but did not necessarily result in male sterility. From the field data, we found that the likelihood of
a second embolus tip entering the spermatheca is significantly lower than that of the first tip, suggesting the possibility that
the tip functions as a partial mating plug.

Keywords: Male mating strategy, embolus tip, Latrodectus pallidus

The reproductive success of a male is usually a function of
the number of females he inseminates, especially when males
produce numerous gametes and when little time and energy is
spent on care of offspring. Under these circumstances each
male is expected to mate with many females (Darwin 1871;
Bateman 1948; Trivers 1972; Andersson 1994). Nevertheless,
in some cases males invest highly in a single female. This
investment may increase the male’s reproductive success in the
current mating but dramatically reduce the number of matings
that he can potentially achieve. Such a strategy can be
promoted by evolutionary processes if the probability of
encountering and mating with an additional receptive female
is sufficiently low (Parker 1979; Buskirk et al. 1984; Elgar 1992;
Simmons et al. 1992; Parker 1998; Andrade 2003) and if males
strongly compete for females (Thornhill 1980; Fromhage et al.
2005).

Copulatory organ breakage, in which a portion of the
male’s genitalia breaks off during copulation is relatively
common in spiders (Wiehle 1967; Breene & Sweet 1985; Foelix
1996; Schneider et al. 2001; Miller 2007). Broken organs inside
the female’s genitalia may function as a mating plug to prevent
fertilization by later arriving males, but it may also reduce the
probability of the male fertilizing additional females (Foelix
1996). Thus, this trait may represent a male strategy of high
investment in a single mating. To date, few studies have
quantified the costs and benefits of male organ breakage. In
the spider Nephila funestrata Thorell 1859, for example, males
often damage both of their paired mating organs while
copulating with a single virgin female (Fromhage & Schneider
2005). The occurrence of a male organ part inside the female’s
genitalia was shown to reduce the number of copulatory
insertions by a second male (Fromhage & Schneider 2006).
Similarly, in Argiope bruennichi Scopoli 1772, males can use

each copulatory organ once, and insertions into a previously
used insemination duct were significantly shorter when the
previous male had left parts of his genitalia inside the
insemination duct (Nessler et al. 2007).

In widow spiders ( Latrodectus ), the tip of one or both of the
male’s intromittent organs (emboli) often breaks-off during
copulation to be left inside the female’s genitalia (Levi 1959;
Bhatnagar & Rempel 1962; Wiehle 1967; Kaston 1970;
Berendonck & Greven 2002; Segoli et al. 2006). Males without
embolus tips were assumed to be functionally sterile ( Bhatna-
gar & Rempel 1962; Foelix 1996), but there is evidence that
this is not always the case, as shown in L. mactans Fabricius
1775 (Breene & Sweet 1985) and L. hasselti Thorell 1870
(Snow et al. 2006). It was suggested that a tip inside the
female’s spermatheca functions as a mating plug (Foelix 1996;
Berendonck & Greven 2002); however, this was demonstrated
only in L. hasselti. In this species first male sperm precedence
was found when two males inseminated a single genital pore
(Snow & Andrade 2005) and when the first tip was deposited
in the entrance of the spermatheca (Snow et al. 2006). In
several other Latrodectus species, however, more than one tip
can be found inside the female’s spermathecae (Uhl 2002),
suggesting that the embolus tip is not totally effective as a
mating plug.

In this study we investigated two aspects of the adaptive
value and costs of damage to the male organ in the white
widow spider, Latrodectus pallidus O. Pickard-Cambridge
1872: 1) future inseminating opportunities for males who have
broken emboli (male sterility hypothesis) and 2) the risk of
sperm competition (mating plug hypothesis). Males of this
species suffer high extrinsic mortality and normally do not
encounter more than one female in natural conditions, while
females are often polyandrous (Segoli et al. 2006). Thus, males

533

534

THE JOURNAL OF ARACHNOLOGY

that encounter a virgin female would benefit from blocking the
spermatheca of their mate and thereby reducing or preventing
access to future rivals. There is evidence that L. pallidus males
invest highly in each mating: they cohabit in females’ webs
longer than required for mating (Segoli et al. 2006), engage in
an energetically demanding courtship display (M.S. personal
observations), and are sometimes cannibalized by the female
(Segoli et al. 2006). Thus, breakage of the male organ may be
an integral part of the male mating strategy in this species even
at the cost of a limited fertilization success in the future.

In order to estimate the frequency and consequences of male
organ breakage we asked the following questions: 1) How
frequently do embolus tips break? 2) Does the loss of embolus
tips prevent the male from remating? and 3) does the presence
of a tip inside the female’s spermatheca reduce the probability
of another tip entering the spermatheca? In order to answer
the first two questions we conducted an experiment in which
we provided 21 males each with four virgin females
simultaneously in outdoor enclosures. Thus, each male had
the opportunity to possibly mate with four females. For each
male, we recorded fertilization success and the loss of embolus
tips by recording the successful production of viable egg sacs
in females and by recording the presence of the embolus tips in
the genitalia of the females. In order to answer the third
question, we collected females from natural populations and
recorded the location of male tips inside their genitalia.

METHODS

The white widow spider (L. pallidus ) is common in the
Negev desert of Israel (Levy 1998). We collected males and
females from the Sede Teman area (31°17'N, 34°43'E) and
Sayeret Shaked Park (31°16'N, 34°38'E) (northern Negev,
Israel) in April 2003. Voucher specimens were deposited in the
National Collection of Arachnids, Hebrew University of
Jerusalem. Spiders were collected as juveniles or sub-adults
and reared to maturity under lab conditions on the Sede Boqer
Campus of Ben-Gurion University. Males were kept in plastic
cups (200 cc) and fed weekly with Drosophila. Females were
kept in terraria (10 X 20 X 15 cm) containing small dry
shrubs, on which they constructed their webs. Females were
fed weekly with flour beetles (larvae of Tenebrio molitor),
crickets (Ache t a domestica), grasshoppers (Schistocerca sp.)
and houseflies ( Musca domestica).

Enclosure experiment. — We placed four adult virgin females
with their webs in a square outdoor arena (135 cm length X
135 cm width X 50 cm height), one in each corner. Each arena
was constructed from a wooden base and frame with plastic
sheets as walls and a removable mesh cover. Females were not
fed during the trial. Once the females repaired their webs
(~1 day), we placed one adult naive male in the center of each
arena (n — 21 replicates). The location and activity of the male
(no movement, courting, in mating position) were recorded
three times a day until it died. The number of daily
observations was determined from preliminary observations,
which indicated that males stay at least one day with each
female. Females were measured and weighed at the end of
each trial. They were kept until they produced seven egg sacs
or until two months passed without laying eggs. Females that
produced fertile egg sacs were assumed to have mated. This
assumption is valid because mated females kept with adequate

food rarely fail to produce fertile egg sacs (M.S. personal
observations). The reproductive success of females was
measured by the total number of eggs and by the number of
fertile eggs (eggs that hatched) from the first five egg sacs. We
used data from the first five egg sacs since most females lay 1-
3 egg sacs in the field and five egg sacs was the maximum
observed in nature (M.S. personal observations). Post mortem
we checked females’ spermathecae for the presence of embolus
tips. Females have paired copulatory ducts, each leading to a
spermatheca, and males have paired intromittent organs
(emboli). During mating, the male inserts one embolus at a
time into one of the female’s genital openings. Thus, a male
may leave none, one, or both tips (one in each side) in the
genitalia of a female. We obtained complete data on male
embolus tips in the genital tracts of all four females from 15
trials. Spermathecae were examined by placing them in a 5%
KOI I solution; after a week the tissue became transparent and
the embolus tips were visible under a dissecting microscope
(Berendonck & Greven 2002). For each trial we determined
the order and number of females the male visited, which of the
females he inseminated, the reproductive success of each
female and finally, which females possessed embolus tips
inside the spermathecae.

Females collected from the field. — We collected 216 adult
females from their webs at three locations in Israel: Goral
Hills, near Lehavim (31°22'N, 34°49'E, n = 192), Kfar
Edomim (31°49'N, 35°19'E, n = 13), and Sayeret Shaked
(31°16'N, 34°38'E, n = 11) from March 1998 till September
1999. Egg sacs, if present, were left unharmed in the web. We
dissected the females, removed their spermathecae and
copulatory ducts, placed them in a 5% KOH solution and
examined them for the presence of embolus tips as above.
Since the number of tips in the right and left genitalia were
correlated (Spearman rank correlation, n = 216, Rs = 0.632, P
< 0.01), we considered only the right spermatheca and
copulatory duct of each female, thereby avoiding pseudorep-
lication.

We compared cases in which one male tip was found in the
female genitalia (genital duct + spermatheca) to cases in which
two tips were found. We estimated the probability of a first
embolus tip to enter the spermatheca as the percentage of
females with an embolus tip located inside the spermatheca
out of the total number of females with one embolus tip found
in their right genitalia. We compared this with the probability
of a second tip entering the spermatheca: the percentage of
females with two tips inside the spermatheca, out of the total
number of females with two tips in their right genitalia. We
expected that if the first tip prevents the second tip from
entering the spermatheca, the probability of finding a second
tip inside the spermatheca would be lower than for the first tip.

RESULTS

Enclosure experiment, — After placement in the arena with
the four females, males started courting one of the females.
Courtship included the following behaviors: adding silk to the
female’s threads, vibrating the web and cutting sections from
the web. Mating was difficult to observe since it took place
inside the female’s retreat. On the following days males were
observed courting or standing motionless on the web, either
inside or outside the retreat. The median time from the

SEGOLI ET AL.— DAMAGE TO MALE CQPULATORY ORGAN

535

Number of mated females

Figure i. — Number of males that mated with 0-3 females (n = 19
trials). Each male was provided with four virgin females simulta-
neously in an outside enclosure.

introduction of the male into the arena with females until the
death of the male was 5 days (range 1-23 days, n = 21). One
male escaped from his arena and entered another arena. He
was returned to his arena after visiting one female in the
adjacent arena. We excluded the two males from these two
arenas from analyses of the number of mated females.

Three males out of 19 did not inseminate any female, most
of the males inseminated one or two females and two males
inseminated three females (Fig. 1). The proportion of insem-
inated females was higher among females that were visited first
(77%) than among females visited later (41%) (Fisher’s exact
test, n = 22 for first females and n = 29 for females visited
later, P — 0.02).

Three mated females died during the experiment and the
remaining mated females produced seven egg sacs before the
end of the experiment. The mean number of eggs per egg sac
was 130 ± 30 (± SD, n = 30 females; averages of eggs per sac
for each female were averaged over all females) and the
number of hatched eggs was 103 ± 40. The total number of
eggs that were produced by mated females in the first five sacs
was not influenced by the number of embolus tips inside their
spermathecae nor by mass, size, or age of females (GLM
stepwise backward model, n = 25, P > 0.1 for all). The results
were still not significant when considering hatched eggs only.
Thus, there were no differences among the females in their
reproductive success.

Six out of 21 males (29%) were cannibalized by females.
Cannibalism was observed directly or could be inferred from
the transparent body of the dead male found on the female’s
web. In five out of six cases the cannibalistic female did not
produce egg sacs, indicating that cannibalism occurred before
copulation, or that the female did not use the male’s sperm for
fertilization.

Data on the presence of male embolus tips in the genitalia of
females and fertilization success are presented in Table 1.
Three males did not lose any embolus tip with the first female
they visited; nevertheless, one of these fertilized the female.
Four males lost one embolus tip in the first mating. Two of
these fertilized the first female only and the other two fertilized
one and two additional females. Eight males lost both embolus

Table 1. — Embolus tips inside spermathecae and fertilization of
1st, 2nd, and 3rd females visited by 15 males in the arena experiment.
Numbers in columns represent the number of tips found in the female
spermathecae. Shaded cells indicate that the female produced fertile
egg sacs.

# males

First

Second

Third

2

0

0

0

1

0

0

0

2

1

0

0

4

2

0

0

3

2

0

0

1

2

0

0

1

1

1

0

1

1

1

0

tips in the first mating. Four of these fertilized the first female
only, three fertilized an additional female, and one fertilized
two additional females.

Females collected from the field. — Forty-eight out of 216
females contained no embolus tip inside their right genitalia.
In 86 cases out of the 95 females that contained one tip in their
genitalia, the tip was placed inside the spermatheca, and in
nine cases the tip was located in the genital duct. Thus, we
estimated the probability for the first tip in the female’s
genitalia to enter the spermatheca as 0.9. We found 58 females
with two embolus tips inside their right genitalia. Out of these,
in three cases both tips were located in the genital duct, in 28
the two tips were placed inside the spermatheca, and in 27 one
tip was placed inside the spermatheca and the second was In
the genital duct. Thus, we estimated the probability of a
second tip to enter the spermatheca as 0.5. The likelihood of a
first tip to enter the spermatheca was significantly greater than
the likelihood of a second tip to enter the spermatheca (Fig. 2,
Fisher’s exact test, P < 0.0001). Additionally, fourteen females
contained three tips in their right genitalia. Of these, one
female had no tips inside the right spermatheca, 8 had one tip,
two had two tips and three had three tips. Finally, one female
contained five embolus tips, four of which were found inside
the spermatheca.

Total # of tips in right genitalia (spermatheca + dye!)

Figure 2. — Embolus tips found in the spermatheca alone and in the
entire genitalia: percentage of females with no embolus tips (white
section), one embolus tip (gray section) or two embolus tips (black
section) inside their right spermatheca, out of field-collected females
with either one (n = 95) or two tips (n = 58) in their right genitalia
(spermatheca + genital duct).

536

THE JOURNAL OF ARACHNOLOGY

DISCUSSION

In this study we estimated the frequency and consequences
of damage to the male copulatory organ in the white widow
spider L. pallidus. We found that damage to the male organ
was frequent but did not necessarily result in male sterility. We
showed that the occurrence of a male’s embolus tip inside the
female’s spermatheca functions as a partial mating plug: it
probably obstructs but does not always prevent the entrance
of an additional tip into the spermatheca.

In contrast to our study, male sterility following damage to
the male copulatory organ has been demonstrated in several
spider species. For example, in Argiope keyserlingi Karsch
1878, experimental removal of one copulatory organ prevent-
ed males from copulating with more than one female,
suggesting that males can use each of their paired organs
only once (Herberstein et al. 2005). In Nephila funestrata, 95%
of the males mating with a virgin female had a damaged organ
that probably prevented them from remating (Fromhage &
Schneider 2005). In widow spiders the loss of an embolus tip
inside the female genital tract was previously assumed to result
in functional sterility of the male (Bhatnagar & Rempel 1962).
Breene & Sweet (1985), however, found that some males of a
congener (L. mactans) were able to successfully inseminate
three females suggesting that males either do not always lose
their tips or that they can inseminate in spite of embolus
breakage. In L. hasselti , males are normally sterile after
mating (Andrade & Bant a 2002), but when tips were cut
experimentally males were able to inseminate additional
females (Snow et al. 2006). Thus, the loss of embolus tips
alone cannot be responsible for the post-mating sterility in L.
hasselti. In our study we found that at least one male mated
and inseminated a female without losing any embolus tip and
five males inseminated one or two females after losing both
tips in previous matings. Thus, we suggest that the loss of
embolus tips in L. pallidus is common, but does not prevent
the male from fertilizing additional females.

Although damage to the copulatory organ in L. pallidus was
not an absolute constraint on the male’s reproductive success,
only a few males (2 out of 19) inseminated more than two
females. This suggests that insemination with a broken
embolus is mechanically difficult and is less likely to be
successful than insemination with an intact embolus. Addi-
tionally, in the absence of tips, males may have difficulties
filling their emboli with sperm (sperm induction) and therefore
low fertilization success may result from sperm depletion
rather than an inability to transfer sperm (Snow et al. 2006).
However, it is not yet known whether white widow males refill
their emboli between mating attempts. Finally, insemination
with a broken embolus may be especially difficult when
mating with an already mated female with a plugged
spermatheca. If so, embolus breakage may still carry a cost
for males in mating systems where sperm competition exists.

Although males were not competing for females in this
experiment, there is evidence that embolus breakage may give
the males an advantage in sperm competition. Most of the
males that lost both tips (8 out of 10) left them in each of the
two spermathecae of the first female that they mated,
indicating that they had mated with her twice. However, there
was no difference in the reproductive success of females with
one or two tips in their spermathecae. A similar result was

obtained in a study of L. hasselti where repeated mating did
not increase the probability of successful fertilization nor the
number of offspring produced in successful matings (Andrade
& Santa 2002). We suggest that males leave both tips in order
to protect both of the female’s spermathecae from future
insemination by rival males.

The analysis of spermathecae from females collected in the
field further supports the view that the broken embolus
functions as a partial mating plug. The probability of a first
embolus tip entering the spermathecae was significantly
higher (90%) than that of the second tip (50%). It is also
possible that a second tip replaced the first, but this is unlikely
considering the narrow entrance to the spermatheca (Beren-
donck & Greven 2002). However, the results also suggest that
the tip is not totally effective as a plug: in half of cases a
second tip did enter the spermatheca, and in four cases more
than two tips entered the spermatheca. In contrast, in a study
of L. hasselti it was shown that in —90% of the cases where
two males inseminated the same genital pore, the second tip
did not enter the spermatheca resulting in a first male sperm
priority (Snow et al. 2006). Although it is difficult to compare
the results of this controlled experiment with our field data, it
implies that the plug in L. hasselti is more efficient than in L.
pallidus. From an evolutionary point of view, the differences
in the efficiency of the plug between species may reflect an
arm-race between males and females over control of paternity.
In this light it would be interesting to compare the efficiency
of the plug in different Latrodectus species in relation to
mating opportunities, effective sex ratio, and sexual canni-
balism.

In contrast to embolus tip breakage, sexual cannibalism
does not seem to be an integral part of the male mating
strategy in L. pallidus. In L. hasselti , males initiate cannibalism
by placing their abdomen in front of the female’s mouthparts
during copulation: cannibalized males copulate longer and
cannibalistic females are less likely to remate (Andrade 1996).
Similar sacrificial behavior was also observed in L. geome-
tricus C.L. Koch 1841 (Segoli et al. 2008). In L. pallidus ,
however, most of the cannibalized males (5 out of 6) did not
fertilize the cannibalistic female and thus could not benefit
from cannibalism. This illustrates the distinction between male
sacrifice behavior as an adaptive strategy and cannibalism as
an unavoidable consequence of mating with a dangerous
partner.

In conclusion, damage to male copulatory organs is
consistent with a male strategy of high investment in a single
female. Embolus damage does not necessarily result in male
sterility and may provide some paternity advantage over
subsequent males. This benefit will be expressed only when
females mate multiply and when mating opportunities are
limited for males, as is the case in the mating system of white
widow spiders (Segoli et al. 2006).

ACKNOWLEDGMENTS

We thank Iris Musli, Efrat Gavish, Alexei Maklakov,
Dinesh Rao, Yael Teleman, Moran Segoli, Tamar Keasar, and
Jutta Schneider for discussions and assistance, and two
anonymous referees for valuable comments on the manuscript.
This is publication no. 613 of the Mitrani Department of
Desert Ecology.

SEGOLI ET AL.— DAMAGE TO MALE COPULATORY ORGAN

537

LITERATURE CITED

Andersson, M. 1994. Sexual Selection. Princeton University Press,
Princeton, New Jersey. 559 pp.

Andrade, M.C.B. 1996. Sexual selection for male sacrifice in the
Australian redback spider. Science 271:70-72.

Andrade, M.C.B. 2003. Risky mate search and male self-sacrifice in
redback spiders. Behavioral Ecology 14:531-538.

Andrade, M.C.B. & E.M. Banta. 2002. Value of male remating and
functional sterility in redback spiders. Animal Behaviour 63:857—
870.

Bateman, A.J. 1948. Intra-sexual selection in Drosophila. Heredity
2:349-368.

Berendonck, B. & H. Greven. 2002. Morphology of female and male
genitalia in Latrodectus revivensis with regard to sperm priority
patterns. Pp. 157-167. In European Arachnology 2000. (S. Toft &
N. Scharff, eds.). Aarhus University Press, Aarhus, Denmark.
358 pp.

Bhatnagar, R.D.S. & J.G. Rempel. 1962. The structure, function, and
postembryonic development of the male and female copulatory
organs of the black widow spider Latrodectus curacaviensis
(Muller). Canadian Journal of Zoology 40:465-510.

Breene, R.G. & M.H. Sweet. 1985. Evidence of Insemination of
multiple females by the male black-widow spider, Latrodectus
mactans (Araneae, Theridiidae). Journal of Arachnology 13:331—
335.

Buskirk, R.E., C. Frohlich & K.G. Ross. 1984. The natural-selection
of sexual cannibalism. American Naturalist 123:612-625.

Darwin, C.R. 1871. The Descent of Man and Selection in Relation to
Sex. The Modern Library, Random House, New York. 743 pp.

Elgar, M.A. 1992. Sexual cannibalism in spiders and other
invertebrates. Pp. 128-155. hi Cannibalism: Ecology and Evolu-
tion Among Diverse Taxa. (M. Elgar & B. Crespi, eds.). Oxford
University Press, Oxford, UK.

Foelix, R. 1996. Biology of Spiders. Second edition. Oxford
University Press, New York. 330 pp.

Forster, L.M. 1992. The stereotyped behavior of sexual cannibalism
in Latrodectus hasselti Thorell (Araneae, Theridiidae), the Austra-
lian redback spider. Australian Journal of Zoology 40:1-11.

Fromhage, L., M.A. Elgar & J.M. Schneider. 2005. Faithful without
care: the evolution of monogyny. Evolution 59:1400-1405.

Fromhage, L. & J.M. Schneider. 2005. Safer sex with feeding females:
sexual conflict in a cannibalistic spider. Behavioral Ecology 16:
377-382.

Fromhage, L. & J.M. Schneider. 2006. Emasculation to plug up
females: the significance of pedipalp damage in Nephila fenestrata.
Behavioral Ecology 17:353-357.

Herberstein, M.E., K.L. Barry, M.A. Turoczy, E. Wills, C. Youssef &
M.A. Elgar. 2005. Post-copulation mate guarding in the sexually
cannibalistic St Andrew’s Cross spider (Araneae Araneidae).
Ethology Ecology & Evolution 17:17-26.

Kaston, B.J. 1970. Comparative biology of American black widow
spiders. Transcriptions of the San Diego Society of Natural
History 16:33-82.

Levi, H.W. 1959. The spider genus Latrodectus (Araneae, Theridi-
idae). Transactions of the American Microscopical Soceity 78:
7-43.

Levy, G. 1998. Fauna Palestina Arachnida III: Araneae: Theridiidae.
Academy of Sciences and Humanities, Jerusalem, Israel. 264 pp.

Miller, J.A. 2007. Repeated evolution of male sacrifice behavior in
spiders correlated with genital mutilation. Evolution 61:1301-1315.

Nessler, S.H., G. Uhl & J.M. Schneider. 2007. Genital damage in the
orb-web spider Argiope bruennichi (Araneae: Araneidae) increases
paternity success. Behavioral Ecology 18:174-181.

Parker, G.A. 1979. Sexual Selection and Sexual Conflict. Pp. 123-166.
In Sexual Selection and Reproductive Competition in Insects.
(M.S. Blum & N.A. Blum, eds.). Academic Press, New York.

Parker, G.A. 1998. Sperm competition and the evolution of
ejaculates: towards a theory base. Pp. 3-54. In Sperm Competition
and Sexual Selection. (T.R. Birkhead & A.P. Moller, eds.).
Academic Press, London.

Schneider, J.M., M.L. Thomas & M.A. Elgar. 2001. Ectomised
conductors in the golden orb-web spider, Nephila plumipes
(Araneoidea): a male adaptation to sexual conflict? Behavioral
Ecology and Sociobiology 49:410-415.

Segoli, M., R. Arieli, P. Sierwald, A.R. Harari & Y. Lubin. 2008.
Sexual cannibalism in the brown widow spider (Latrodectus
geometricus). Ethology 1 14:279-286.

Segoli, M„ A.R. Harari & Y. Lubin. 2006. Limited mating
opportunities and male monogamy: a field study of white widow
spiders, Latrodectus pallidus (Theridiidae). Animal Behaviour 72:
635-642.

Simmons, L.W., R.J. Teale, M. Maier, R.J. Standish, W.J. Bailey &
P.C. Withers. 1992. Some costs of reproduction for male bush-
crickets, Requena verticals (Orthoptera, Tettigoniidae): allocating
resources to mate attraction and nuptial feeding. Behavioral
Ecology and Sociobiology 31:57-62.

Snow, L.S.E., A. Abdel-Mesih & M.C.B. Andrade. 2006. Broken
copulatory organs are low-cost adaptations to sperm competition
in redback spiders. Ethology 112:379-389.

Snow, L.S.E. & M.C.B. Andrade. 2005. Multiple sperm storage
organs facilitate female control of paternity. Proceedings of the
Royal Society B-Biological Sciences 272:1139-1144.

Thornhill, R. 1980. Sexual selection within mating swarms of the
lovebug, Plecia nearctica (Diptera, Bibionidae). Animal Behaviour
28:405-412.

Trivers, R.L. 1972. Parental investment and sexual selection.
Pp. 136-179. In Sexual Selection and the Descent of Man, 1871-
1971. (B.G. Campbell, ed.). Aldine-Atherson, Chicago.

Uhl, G. 2002. Female genital morphology and sperm priority patterns
in spiders (Araneae). Pp. 145-156. In European Arachnology 2000.
(S. Toft & N. Scharff, eds.). Aarhus University Press, Aarhus,
Denmark.

Wiehle, H. 1967. Steckengebliebene Emboli in den Vulven von
Spinnen (Aracfa., Araneae). Senckenbergiana Biologica 48:197-
202.

Manuscript received 13 May 2007, revised 30 April 2008.

2008. The Journal of Arachnology 36:538-544

Molting interferes with web decorating behavior in Avgiope keyserlingi (Araneae, Araneidae)

Andre Walter1, Mark A. Elgar2, P. Bliss1 and Robin F. A. Moritz1: 'Institut fur Biologie, Martin-Luther-Universitat
Halle-Wittenberg, Hoher Weg 4, D-06099 Halle (Saale), Germany. E-mail: andre.walter@zoologie.uni-halle.de;
department of Zoology, University of Melbourne, Melbourne, Victoria 3010, Australia

Abstract. Various orb weaving spiders decorate their webs with extra silk structures. In the araneid genus Avgiope, these
web decorations consist of flimsy aciniform silk threads arranged in zig zag shaped bands. The adaptive value of these
structures is still unclear and controversy over a suite of possible functional explanations persists: the high variation of web
decoration adds further uncertainty. Web decorations can differ in shape, size, and frequency across species and even
within species. Physiological processes may influence individual variation in web decorating behavior. Molting events are
major physiological transitions combined with fundamental alterations of the metabolic state of the spiders. For gaining
new insights into possible proximate mechanisms driving web decorating behavior, we observed subadult Avgiope
keyserlingi Karsch 1878 females in the laboratory and registered the individual variation of web decorations associated with
the maturity molt under laboratory conditions.

We found substantial individual variation of web decorations of A. keyserlingi. The most striking result was that
subadult spiders built dramatically oversized decorations prior to the last molt. Since aciniform silk is used for both
constructing web decorations and immobilizing prey we suggest that these extensive decorations might provide a store for
the swift replenishment of aciniform silk after the molt. High silk recycling rates make temporary outsourcing less costly
and facilitate a rapid resumption of prey capture following lost foraging opportunities during the molting phase. Thus, we
argue that the solution of the riddle of web decorations might reside in the physiology of molting spiders.

Keywords: Orb-web spiders, web decorations, maturity molt, gland regulation

Web decorating is a characteristic behavior of various orb
weaving spiders (Robinson & Robinson 1973; Edmunds 1986;
Bruce 2006), yet the possible functional explanations remain
controversial despite extensive investigations (Herberstein et
al. 2000; Eberhard 2003; Bruce 2006). “Web decorations”
(first mentioned as such by McCook 1889, but also called
“stabilimenta” by Simon 1895 and many modern authors) in
the araneid genus Avgiope consist of numerous flimsy acini-
form silk threads (Peters 1993), mostly arranged in zig zag
shaped bands (Bruce 2006). Although web decorations of
Avgiope are considered as prey attractants by some (Craig &
Bernard 1990; Tso 1996; Bruce et al. 2001; Li 2005; Cheng &
Tso 2007), this view is not unanimous and alternative
functional explanations include anti-predator devices (Ewer
1972; Schoener & Spiller 1992; Blackledge & Wenzel 2001);
advertisement for web protection (Eisner & Nowicki 1983;
Kerr 1993; Blackledge & Wenzel 1999); thermoregulation
(Humphreys 1992); mechanical support [Robinson & Robin-
son 1970; see also Watanabe 2000 for Octonoba sybotides
(Bosenberg & Strand, 1906), Uloboridae]; and acting as a
molting platform (Robinson & Robinson 1973, 1978). In the
rapidly growing literature on this topic, tests for non-visual
functions are clearly underrepresented (Bruce 2006). In
particular the potential relationship between physiological
processes and web decorating behavior has been addressed in
only a very few studies (e.g., Peters 1993; Tso 2004; Walter et
al. 2008a).

Typically, decorating behavior in species of Avgiope is
highly variable (Bruce & Herberstein 2005) and web decora-
tions can differ in shape (number and arrangement of zig zag
bands), size and frequency (Lubin 1975; Edmunds 1986;
Nentwig & Heimer 1987; Schoener & Spiller 1992). One
problem for determining the adaptive value of web decora-
tions stems from this high variation (Robinson & Robinson

1974), which occurs across species and within species at both
the population and individual level (Herberstein et al. 2000;
Starks 2002; Bruce & Herberstein 2005; Rao et al. 2007). Most
studies explore the adaptive significance of these structures
(e.g., Blackledge 1998; Craig et al. 2001), although phyloge-
netic analyses of web decoration patterns suggests that
interspecific variance shows weak homologies at best and
yields phylogenetically feeble signals (Herberstein et al. 2000;
see also Scharff & Coddington 1997). We agree with Eberhard
(2003) that an accumulation of single “experiments per se ...
are no guarantee of reliable conclusions.” Thus, understand-
ing the intra-individual variance of web decorations in detail is
necessary before embarking on the interpretation of web
decorating behavior in general.

The production of web decoration is governed by an
enhanced activity of the silk glands and hence physiological
processes are expected to impact web decorating behavior
(Tso 2004; Walter et al. 2008a). The major physiological
transitions in the life history of spiders are the repeated
molting events. Molting requires a drastic change of anabolic
and metabolic biochemical pathways requiring fundamental
alterations of the physiological state of the animal. Apart from
hormonal changes (Bonaric 1987; Foelix 1996; Craig 2003),
molts are particularly vulnerable events in the life of spiders, in
terms of both increased physiological stress (Pulz 1987;
Vollrath 1987a) and increased risk of predation (Tolbert
1975; Tanaka 1984; Vollrath 1987b; Baba & Miyashita 2006).
It would, therefore, be surprising if web decoration behavior
was not affected by molting. Indeed several studies suggest
that molting might have profound effects on the web
decorating activity of Avgiope (Robinson & Robinson 1970,
1973; Edmunds 1986; Nentwig & Heimer 1987). Yet if we
observe consistent changes in the patterns of decoration
behavior associated with the molting process, this might

538

WALTER ET AL.— WEB DECORATING BEHAVIOR IN ARGIOPE KEYSERLINGI

539

provide insights into the proximate mechanisms driving web
decoration and their potential adaptive value.

METHODS

Study species and experimental design. — We chose the St.
Andrew’s cross spider, Argiope keyserlingi Karsch 1878, to
study the variation in web decoration under highly controlled
laboratory conditions. This orb web spider is distributed along
the east coast of Australia (northern Queensland to Victoria in
the south), building their webs between branches and leaves of
bushes, e.g., in parks and gardens. Argiope keyserlingi is a well
studied species in terms of its natural history (Rao et al. 2007),
its sexual cannibalism (Elgar et al. 2000; Herberstein et al. 2005)
and its web decorating behavior (Herberstein 2000; Bruce et al.
2001, 2005; Herberstein & Fleisch 2003). St. Andrew’s cross
spiders typically build cruciate web decorations consisting of up
to four zig zag bands forming a large “X” in the orb web (Rao
et al. 2007). This allows an unambiguous interpretation of
deviations from the “complete cross.”

We collected 55 subadult female spiders in Ku-ring-gai
Bicentennial Park (West Pymble/Sydney, Australia) and trans-
ferred them individually to Perspex frames (58 cm X 58 cm X
15 cm) in the laboratory, where they were kept under natural
light conditions. Every other day, each spider was fed with one
blowfly ( Lucilia spp.). At this life stage the spiders are still of a
similar size as the blowflies; thus, it has turned out in preliminary
observations that this feeding regime is sufficient to keep spiders
“well-fed.” At this same time, each web was moistened with five
shots from a water spray. Given that spiders typically build a
new web each day, we recorded daily the number of decoration
bands (shape) and decoration size to assess the variation of web
decorating behavior within a total observation period of
30 days. We estimated the size of the web decoration by
computing a trapezium area similar to Tso (1999): (a+c)/2 X h (a
and c = upper and lower width of zig zag bands, h = height of
zig zag bands, see Fig. 1 ). Additionally, we quantified the size of
all newly built webs following Herberstein & Tso (2000): (dv/2)
X (dh/2) X n (dv = vertical and dh = horizontal diameter of the
capture area, see Fig. 1) and measured the spider body size
(length from clypeus to the end of the opisthosoma). Voucher
specimens were deposited in the Entomological Collection of the
Martin-Luther-University Halle-Wittenberg (Zoological Insti-
tute), Germany (identification number 2569).

Statistical analyses. — We used STATISTICA® (version 6.0)
for all statistical analyses including the paired /-test to evaluate
differences in the sizes of decorated and undecorated webs.
Chi square-tests and /-tests were used to detect differences in
the proportion of decorated web parts and constructed
decoration patterns. Web and web decoration sizes prior,
during, and after molting events were analyzed with an
ANOVA. Pearson-correlations were computed between web
size and decoration size.

RESULTS

Web decorating frequency. — All females could be observed
over the whole 30 day period. Forty-six of the 55 subadult A.
keyserlingi molted to maturity within this time. The spiders
constructed new webs every second day (mean 2.29 ± 0.07 SE
day). Typically, the new web decorations were built together
with new webs, and therefore the decorating activity mostly

Figure 1 . — Web and decoration measurements from the webs of A.
keyserlingi-. Left: determination of the size of capture areas (including
hub region), dh = horizontal diameter, dv = vertical diameter; Right:
determination of decoration band sizes, a = upper width of the band, c
= lower width of the band, h = length of the band (lettering after
trapezium formula).

followed an equal rhythm (a mean value of every 2.37 ± 0.37
SE day). The few exceptions were all in the context of molting
events (see below). However, 233 (37.5%) of all newly built
webs ( n = 622) did not contain a web decoration. Many
spiders occasionally failed to decorate their webs, but only Five
animals (9.1%) never built a web decoration at all during the
observation period.

Web size. — The spiders more than doubled the catching
area of their webs within the 30 day observation period. The
mean size of the first web we measured was 635.30 ± 44.46 SE
cm2 (n — 55) and the mean size of the last measured web was
1630.61 ± 21.99 SE cm2 (n = 55). Over the whole observation
period, undecorated webs were significantly larger than
decorated webs, ranging from 625.21 ± 56.88 SE cm- to
646.54 ± 70.57 SE cm2 at the beginning of to the experiment
to between 1563.21 ± 42.56 SE cm2 and 1700.50 ± 37.42 SE
cm2 at the end of the period (paired t-test: / = 2.1 1, P < 0.05).
However, the mean decoration size did not significantly
change over time, and ranged from 55.17 ± 72.72 SE mm"
(n = 55) at the beginning to 46.25 ± 54.06 SE mm2 (n = 55) at
the end of the observation period (Pearson, r — 0.04, P —
0.29). We found a significant positive correlation between
spider size and web area (Pearson, r~ = 0.31, P < 0.01; n =
621). In contrast, we found no significant correlation between
spider size and web decoration size. Consequently, the size of
the decorated web area in relation to the total web decreased
over time.

Variation of web decorating behavior. — The variation in web
decoration shapes was very high and the “typical” cruciate
type was rarely constructed (Fig. 2); females of A. keyserlingi
most often constructed single arm decorations (n = 47 spiders
in 65.13%, n = 282 observations), and decorations with two (/?
= 31 spiders in 24.48%, n — 106 observations), three (n — 10
spiders in 5.54%, n = 24 observations), or four (n = 15 spiders
in 4.85%, n = 21 observations) zig zag bands were less
frequent. In all partial cross shapes (one to three arms), the
bands were added to the lower web half significantly more
often (85.2% vs. 14.8%, n = 50; X2 = 9.12, P < 0.01). There

540

THE JOURNAL OF ARACHNOLOGY

Number of web decoration bands
(decorations, n=433)

Figure 2. — The variation in web decoration patterns of A.
keyserlingi females under laboratory conditions. Partial cross shapes
(one to three decoration bands = number 1-3 in the diagram) are
more frequent than the typical cruciate shape (number 4).

was also strong intra-individual variance; most spiders
(65.45%, n — 36) altered the web decoration pattern up to
four times over the observation period: Thirteen spiders
(23.64%) altered their web decoration pattern once; nine
(16.36%) altered the pattern twice; six individuals (10.51%)
altered it three times; and three spiders (5.46%) altered the
decoration four times. Only 19 spiders (with 34.55% signifi-
cantly less, X2 = 9.55, P < 0.01) constructed the same number
of arms within the observation period and five individuals
(9.09%) built no web decoration at all. These latter spiders
also had a significantly lower web decorating frequency (new
decoration every 3.7 ± 4 SE days) than individuals that
constructed more variable shapes over time (new decoration
every 1.5 ± 0.9 SE days, n = 31; t-test: t = 2.91, P < 0.01).

Web decorating behavior in the context of molting events.
Within the 30 day observation period 46 of 55 subadult
spiders molted to maturity. Spiders suspended the two-day
web building rhythm a few days before molting, and on
average 3.3 ± 1.6 SE days elapsed between the ‘last” web
building and the start of the molt. The mean interval between
constructing the “last” web decoration prior to the final molt
into sexual maturity (2.8 ± 1.5 SE days) was also longer than
the mean decorating interval at other times (every 2.37 ± 0.37
SE day, see above). The molting events coincided with an
increase in overall web size: the web size had increased by 19%
(mean +260 cm2) in the ten day period after the molt (from
1080.37 ± 42.9 SE cm2, n = 101 prior to the molt to 1340.36 ±
25.56 SE cm2, n = 205; paired t-test: t = —4.88, P < 0.01). Ten
spiders (22%) added a new web decoration to an old web prior
to the molt. The change in web decorating and web building
frequency was exclusively observed in combination with
molting events, and the most conspicuous change was the
dramatic increase in the web decoration size (Fig. 3) during
the pre-molting phase (last subadult webs). The size of the

“regular” web decorations, both in penultimate webs before
and in the first webs after the molt, were significantly smaller
(68.78 ± 10.45 SE mm2, n = 43 vs. 58.39 ± 9.24 SE mm2, n =
45) than those constructed directly in the last web before
molting (21 1.74 ± 35.94 SE mm2, n = 46; ANOVA: F = 14.36,
P < 0.01).

The “supersized” decorations of the molting webs were
characterized by a partial loss of the typical zig zag look
(Fig. 3, right). Moreover, these peculiar decoration bands
overlapped in the hub region of the web, which was never
observed in intermolt webs. Finally, only one individual
molted in a web without a web decoration.

All in all, individuals of A. keyserlingi reduced their web
building frequency (Fig. 4A) and increased the size of their
web decorations prior to their final molt to sexual maturity
(Fig. 4B).

DISCUSSION

Although individuals of A. keyserlingi usually build cruciate
web decorations (Rao et al. 2007) consisting of up to four zig
zag-shaped silk bands (Bruce 2006), we observed substantial
individual variation in web construction and decorating
behavior in A. keyserlingi in our study. Web size strongly
correlated with the spider’s size and larger females built larger
webs. Moreover, we could confirm previous reports by
Hauber (1998) and Craig et al. (2001 ) on a negative correlation
between web size and decoration size. Undecorated Argiope
webs were larger than decorated ones. Since we kept the
feeding regime constant, this might indicate a tradeoff between
web size and decoration as suggested by Craig et al. (2001).

Although web size was positively correlated with spider
size, larger spiders did not build larger web decorations.
Consequently, the relative decoration area of the web
decreased over time, which may reflect previous reports of
reduced web decorating behavior in later adult stages of
Argiope spiders (Peters 1953; Edmunds 1986; Nentwig &
Heimer 1987). The intra-individual variation in the shape of
the decoration was remarkably high. Very few spiders
consistently built only one particular pattern. An explanation
for this may be given by the results of Craig et al. (2001) on
Argiope argentata (Fabricius 1775). They argue that individ-
ual decoration patterns have a genetic component and any
variation represents the influence of ecological conditions.
Most spiders in our study alternated the web decoration type,
some individuals up to four times. Although this high
variation may have been affected by the laboratory condi-
tions it has also been observed in many other Argiope species
(e.g., Blackledge 1998 in A. aurantia Lucas 1833 and A.
trifasciata (Forskal 1775); Hauber 1998 in A. appensa
(Walckenaer 1842); Seah & Li 2002 in A. versicolor
(Doleschall 1859); Bruce & Herberstein 2005 in A. picta L.
Koch 1871 and A. aetherea (Walckenaer 1842)).

Argiope keyserlingi females in our study regularly rebuilt
their orb webs every second day, and the web decorating
frequency followed this rhythm. The only exceptions occurred
on those days leading up to the commencement of the final
molt to sexual maturity. During this time, some spiders added
web decorations to their old webs. Typically, Argiope spiders
do not rebuild orb webs several days before they molt to
maturity (Robinson & Robinson 1978; Nentwig & Heimer

WALTER ET AL.- WEB DECORATING BEHAVIOR IN ARGIOPE KEYSERLINGI

541

Figure 3. — The “regular” web decoration (left) and the “supersized” web decoration (right) of A. keyserlingi.

1987; Eberhard 1990). Robinson & Robinson (1973) suggest
that a tradeoff between silk production and the biosynthetic
efforts in preparation of the molt provides an adaptive
explanation for this phenomenon. However, the frequency of
web decorating prior to molting did not decline, despite the
reduction in web building, because some spiders added new
decorations to already existing, old, and dilapidated webs.

Indeed, the dramatically oversized decorations that spiders
built prior to the molt (Fig. 4) were the most conspicuous
difference to the intermolt webs of subadult and the webs of
adult individuals. The phenomenon that spiders build more
frequent and/or more perfect web decorations prior to the
molt was already observed in A. argentata and A. savignyi Levi
1968 in the laboratory (Nentwig & Heimer 1987). Moreover,
Edmunds (1986) noticed larger and denser decorations prior
to moltings in a wild population of A. Jlavipalpis (Lucas 1858).
These anecdotal reports, however, have never been empirically
quantified. In our study we could show that web decorations
in A. keyserlingi were three times larger shortly before the
maturation molt and did not correspond with the individual
variation in decoration shape. Decoration size decreased to the
intermolt level immediately after the molt. Consequently, very
large decorations were thus directly linked to the molting
procedure.

Do our findings contribute to resolving the controversy
over the adaptive significance of web decorations (see Bruce
2006)? Web decorations have been discussed in a variety of

contexts, including in the context of prey attraction (Herber-
stein 2000; Herberstein & Fleisch 2003; Li 2005). Although
we cannot exclude (his explanation for decorations in regular
webs, the observed increase in web decorating activity in A.
keyserlingi prior to the molt is not predicted by this
hypothesis. Spiders decrease their foraging efforts during
the pre-molt phases (Higgins 1990), presumably because there
is little opportunity to consume food during molting.
Nevertheless, web decorations may provide particular me-
chanical support for orb webs (Simon 1895) during the
molting phase (Robinson & Robinson 1970, 1973, 1978;
Nentwig & Heimer 1987). Higgins (1990) argued that the web
decorations of Nephila clavipes (Linnaeus 1767) (Nephilidae)
help prevent the spiders contacting the sticky spiral, which
could interrupt the moiling procedure by hindering individ-
uals from freeing themselves from the old exoskeleton. Since
molting events are generally vulnerable phases in the life of a
spider (Robinson & Robinson 1973; Baba & Miyashita 2006)
the potentially protective properties of web decorations may
be relevant in preserving the integrity of the web during the
molt (Horton 1980; Eisner & Nowicki 1983; Kerr 1993).
Additionally, the potential protection against predators
(Eberhard 1990; Schoener & Spiller 1992; Blackledge &
Wenzel 2001) would also predict an increase in decorating
investment because spiders are especially vulnerable to
predators during the molt or shortly afterwards (Tanaka
1984; Baba & Miyashita 2006).

542

THE JOURNAL OF ARACHNOLOGY

O
«D i-
*“ II

c r

0 3

|f

Q. O

.£ <5
> £
•S s

it

fl

Web decoration size

hk. t

v hJ

\ V

* J A

\ A

-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10

Day

Web building frequency

Figure 4. — A. The size of the web decoration of A. keyserlingi
dramatically increases prior to the maturation molt and then returns
to the level prior to the molting event; B. Web building frequency
decreases prior to the last molt; dotted line: day of molting (= day 0).

Shortly after a molt spiders are vulnerable to desiccation
due to the slowly sclerotizing exoskeleton. In this phase it must
be particularly important to balance the hygric status through
water ingestions. In this context (large) web decorations might
be practical tools: some Argiope spiders directly ingest water
from parts of their web decorations (Olive 1980; Walter et al.
2008a).

Since Argiope spiders can also successfully molt on webs
without a web decoration (Nentwig 1986; own observations)
the adaptive effects of the decorations may play a subsidiary
role. Instead, the increase in web decoration investment may
proximately derive from direct physiological processes,
particularly since resource allocations directly influence
interactions between molting, silk composition and web
building behavior (Townley et al. 2006). Thus, it might be
necessary to “outsource” a certain amount of nutrients for
optimizing the molting procedure. Higgins & Rankin (2001)
showed that “well-fed” individuals of the orb weaving spider
N. clavipes more often suffer from molting failures when
exceeding a critical pre-molt mass. They concluded that this
might be the cost for the ability of rapid growth based on an
almost non-limited food intake in this species. This may also

be relevant for the rapid growth of Argiope spiders.
Outsourcing body mass in the form of silk proteins may
ensure an “optimal” molt-weight. In this context it is possible
that N. clavipes builds web decorations only shortly before a
molt (Higgins 1990). Conversely, a molt is always combined
with a loss of body mass (through the failure to consume
exuvia) (Hutchinson et al. 1997), and outsourcing silk proteins
may allow spiders to minimize nutrient waste.

The link between the increase of web decorating behavior
and moltings might also be explained by a requirement to
outsource specific, physiologically important compounds that
would be otherwise metabolized during the molting proce-
dure or the non-foraging days shortly before and after the
molt. Such allocation occurs for different compounds in
several spider species [e.g., choline, Higgins & Rankin 1999
for N. clavipes and Townley et al. 2006 for Argiope trifasciata
and A. aurantia; GABamide, Townley & Tillinghast 1988 for
Araneus cavaticus (Keyserling 1881)]. Perhaps the enlarged
decoration simply provides a storage area for the silk proteins
themselves. The aciniform decoration silk is also used for
immobilizing prey (Peters 1993; Tso 2004). Thus, web
decorating might be crucial for maintaining a certain level
of activity in the aciniform glands for an optimal perfor-
mance of Argiope' s typical “wrap attack” strategy of prey
capture (Olive 1980; Tso 2004; Walter et al. 2008b). After
molting, spiders must swiftly resume capturing prey to
compensate for lost foraging opportunities of the previous
days. For subsequent capture events, Argiope requires large
amounts of wrapping silk that has to be newly synthesized
after the molt. Since several types of silk glands are
remodeled during a molt, they may not be fully operative
in the days immediately after the molt (Townley et al. 2006).
If this is also true for the glandulae aciniformes , the extensive
web decorations may provide an ideal store of the crucial silk
components, allowing the swift replenishment of the acini-
form silk following molting. The highly efficient recycling of
web parts (Peakall 1971) thereby clearly reduces the costs of
silk production (Janetos 1982; Opell 1998) by reusing the
relevant amino acids.

To confirm the physiological background of our observa-
tions, further studies should concentrate on the impact of
different metabolic processes on the web decorating behavior
prior to moltings, with a focus on those spiders that
nonetheless molt without decorations. However, irrespective
of the actual ultimate adaptive mechanisms of web decora-
tions, it seems that these structures may play a more specific
role in the molting web than in the regular capture web in
Argiope. Given the large size of the molting decorations in
contrast to relatively small and highly variable decorations in
regular webs, it may well be that the clue to solving the riddle
of these structures lies in the physiology of the molting
spider.

ACKNOWLEDGMENTS

We thank two anonymous reviewers for their useful
comments. This work is supported, in part, by grants of the
Deutsche Forschungsgemeinschaft (PB) and the Ministry for
Science and Culture of the Federal State of Saxony-Anhalt
(AW). The experiments carried out in this study comply with
the current laws of Germany and Australia.

WALTER ET AL. — WEB DECORATING BEHAVIOR IN ARGIOPE KEYSERLINGI

543

LITERATURE CITED

Baba, Y. & T. Miyashita. 2006. Does individual internal state affect
the presence of a barrier web in Argiope bruennichi (Araneae:
Araneidae)? Journal of Ethology 24:75-78.

Blackledge, T.A. 1998. Stabilimentum variation and foraging success
in Argiope aurantia and Argiope trifasciata (Araneae: Araneidae).
Journal of Zoology 246:21-27.

Blackledge, T.A. & J.W. Wenzel. 1999. Do stabilimenta in orb
webs attract prey or defend spiders? Behavioral Ecology
10:372-376.

Blackledge, T.A. & J.W. Wenzel. 2001. Silk mediated defense by an
orb web spider against predatory mud-dauber wasps. Behaviour
138:155-171.

Bonaric, J.-C. 1987. Moulting hormones. Pp. 111-118. In Ecophys-
iology of Spiders. (W. Nentwig, ed.). Springer, Berlin.

Bruce, M.J. 2006. Silk decorations: controversy and consensus.
Journal of Zoology 269:89-97.

Bruce, M.J., A.M. Heiling & M.E. Herberstein. 2005. Spider signals:
are web decorations visible to birds and bees? Biology Letters
1:299-302.

Bruce, M.J. & M.E. Herberstein. 2005. Web decoration polymor-
phism in Argiope Audouin, 1826 (Araneidae) spiders: ontogenetic
and interspecific variation. Journal of Natural History 39:3833—
3845.

Bruce, M.J., M.E. Herberstein & M.A. Elgar. 2001. Signaling conflict
between prey and predator attraction. Journal of Evolutionary
Biology 14:786-794.

Cheng, R.-C. & I-M. Tso. 2007. Signaling by decorating webs: luring
prey or deterring predators. Behavioral Ecology 18:1085-1091.

Craig, C.L. 2003. Spiderwebs and Silk. Tracing Evolution from
Molecules to Genes to Phenotypes. Oxford University Press. New
York. 230 pp.

Craig, C.L. & G.D. Bernard. 1990. Insect attraction to ultraviolet-
reflecting spider webs and web decorations. Ecology 71:616-623.

Craig, C.L., S.G. Wolf, L.D. Davis, M.E. Hauber & J.L. Maas. 2001.
Signal polymorphism in the web-decorating spider Argiope
argentata is correlated with reduced survivorship and the presence
of stingless bees, its primary prey. Evolution 55:986-993.

Eberhard, W.G. 1990. Function and phylogeny of spider webs.
Annual Review of Ecology and Systematics 21:341-372.

Eberhard, W.G. 2003. Substitution of silk stabilimenta for egg sacs by
Allocyclosa bifurca (Araneae: Araneidae) suggests that silk
stabilimenta function as camouflage devices. Behaviour 140:847-
868.

Edmunds, J. 1986. The stabilimenta of Argiope flavipalpis and
Argiope trifasciata in West Africa, with a discussion of the function
of stabilimenta. Pp. 61-72. In Proceedings of the Ninth Interna-
tional Congress of Arachnology, Panama. (W.G. Eberhard, Y.D.
Lubin & B.C. Robinson, eds.). Smithsonian Institution Press,
Washington, D.C.

Eisner, T. & S. Nowicki. 1983. Spider web protection through visual
advertisement: role of the stabilimentum. Science 219:185-187.

Elgar, M.A., J.M. Schneider & M.E. Herberstein. 2000. Female
control of paternity in the sexually cannibalistic spider Argiope
keyserlingi. Proceedings of the Royal Society of London B
267:2439-2443.

Ewer, R. 1972. The devices in the web of the West African spider
Argiope flavipalpis. Journal of Natural History 6:159-167.

Foelix, R.F. 1996. Biology of Spiders. Second edition. Oxford
University Press, New York. 330 pp.

Hauber, M.E. 1998. Web decorations and alternative foraging tactics
of the spider Argiope appensa. Ethology, Ecology & Evolution
10:47-54.

Herberstein, M.E. 2000. Foraging behaviour in orb-web spiders
(Araneidae): do web decorations increase prey capture success in

Argiope keyserlingi Karsch, 1878? Australian Journal of Zoology
48:217-223.

Herberstein, M.E., C.L. Craig, J.A. Coddington & M.A. Elgar. 2000.
The functional significance of silk decorations of orb-web spiders:
a critical review of the empirical evidence. Biological Reviews
75:649-669.

Herberstein, M.E. & A.F. Fleisch. 2003. Effect of abiotic factors on
the foraging strategy of the orb-web spider Argiope keyserlingi
(Araneae: Araneidae). Animal Ecology 28:622-628.

Herberstein. M.E., A.C. Gaskett, J.M. Schneider, N.G.F. Vella &
M.A. Elgar. 2005. Limits to male copulation frequency: sexual
cannibalism and sterility in St Andrew’s Cross spiders (Araneae,
Araneidae). Ethology 111:1050-1061.

Herberstein, M.E. & I-M. Tso. 2000. Evaluation of formulae to
estimate the capture area and mesh height of orb webs
(Araneoidea, Araneae). Journal of Arachnology 28:180-184.

Higgins, L.E. 1990. Variation in foraging investment during the
intermolt interval and before egg-laying in the spider Nephila
clavipes (Araneae: Araneidae). Journal of Insect Behavior
3:77-783.

Higgins, L.E. & M.A. Rankin. 1999. Nutritional requirements for
web synthesis in the tetragnathid spider Nephila clavipes. Physio-
logical Entomology 24:263-270.

Higgins, L.E. & M.A. Rankin. 2001. Mortality risk of high rate of
weight gain in the spider Nephila clavipes. Functional Ecology
15:24-28.

Horton. C.C. 1980. A defense function for the stabilimenta of two orb
weaving spiders (Araneae, Araneidae). Psyche 87:13-20.

Humphreys. W.F. 1992. Stabilimenta as parasols: shade construc-
tion by Neogea sp. (Araneae: Araneidae: Argiopinae) and its
thermal behaviour. Bulletin of the British Arachnological Society
9:47-52.

Hutchinson, J.M.C., J.M. McNamara, A.I. Houston & F. Vollrath.
1997. Dyar’s rule and the investment principle: optimal moulting
strategies if feeding rate is size-dependent and growth is
discontinuous. Philosophical Transactions of the Royal Society B
352:113-138.

Janetos, A.C. 1982. Foraging tactics of two guilds of web-spinning
spiders. Behavioral Ecology and Sociobiology 10:19-27.

Kerr, A.M. 1993. Low frequency of stabilimenta in orb webs of
Argiope appensa (Araneae: Araneidae) from Guam: an indirect
effect of an introduced avian predator? Pacific Science
47:328-337.

Li, D. 2005. Spiders that decorate their webs at higher frequency
intercept more prey and grow faster. Proceedings of the Royal
Society of London B 272:1753-1757.

Lubin. Y.D. 1975. Stabilimenta and barrier webs in the orb webs
of Argiope argentata (Araneae, Araneidae) on Daphne and
Santa Cruz Islands, Galapagos. Journal of Arachnology
2:1 19-126.

McCook, H.C. 1889. American Spiders and Their Spinning Work.
Academy of Natural Sciences, Philadelphia. 373 pp.

Nentwig, W. 1986. Stabilimente in Radnetzen von Argiope- hrien
(Araneae: Araneidae): ein funktionsloses Relikt? Verhandlungen
der Deutschen Zoologischen Gesellschaft 79:182.

Nentwig, W. & S. Heimer. 1987. Ecological aspects of spider webs.
Pp. 211-225. In Ecophysiology of Spiders. (W. Nentwig, ed.).
Springer, Berlin.

Olive, C.W. 1980. Foraging specializations in orb-weaving spiders.
Ecology 61:1 133-1 144.

Opell. B.D. 1998. Economics of spider orb-webs: the benefits of
producing adhesive capture thread and of recycling silk. Func-
tional Ecology 12:613-624.

Peakall, D.B. 1971. Conservation of web proteins in the spider,
Araneus diadematus. Journal of Experimental Zoology
176:257-264.

544

THE JOURNAL OF ARACHNOLOGY

Peters, H.M. 1953. Beitriige zur vergleiclienden Ethologie und
Okologie tropischer Webespinnen. Zeitschrift fur Morphologie
und Okologie der Tiere 42:278-306.

Peters, H.M. 1993. Uber das Problem der Stabilimente in Spinnen-
netzen. Zoologische Jahrbucher, Abteilung fiir allgemeine Zoolo-
gie und Physiologie der Tiere 97:245-264.

Pulz, R. 1987. Thermal and water relations. Pp. 26-55. In
Ecophysiology of Spiders. (W. Nentwig, ed.). Springer, Berlin.

Rao, D., K. Cheng & M.E. Herberstein. 2007. A natural history of
web decorations in the St Andrew’s Cross spider ( Argiope
keyserlingi). Australian Journal of Zoology 55:9-14.

Robinson, B.C. & M.H. Robinson. 1974. The biology of some
Argiope species from New Guinea: predatory behaviour and
stabilimentum construction (Araneae: Araneidae). Zoological
Journal of the Linnean Society 54:145-159.

Robinson, B. & M.H. Robinson. 1978. Developmental studies of
Argiope argentata (Fabricius) and Argiope aemula (Walckenaer).
Symposia of the Zoological Society of London 42:31-40.

Robinson. M.H. & B. Robinson. 1970. The stabilimentum of the orb
web spider, Argiope argentata : an improbable defence against
predators. Canadian Entomologist 102:641-655.

Robinson, M.H. & B. Robinson. 1973. The stabilimenta of Nephila
clavipes and the origins of stabilimentum-building in araneids.
Psyche 80:277-288.

Scharff, N. & J.A. Coddington. 1997. A phylogenetic analysis of the
orb-weaving spider family Araneidae (Arachnida, Araneae).
Zoological Journal of the Linnean Society 120:355-434.

Schoener, T.W. & D.A. Spiller. 1992. Stabilimenta characteristics of
the spider Argiope argentata on small islands: support of the
predator-defense hypothesis. Behavioral Ecology and Sociobiology
31:309-318.

Seah, W.K.. & D. Li. 2002. Stabilimentum variations in Argiope
versicolor (Araneae: Araneidae) from Singapore. Journal of
Zoology 258:531-540.

Simon, E. 1895. Histoire Naturelle des Araignees, Volume 1, fascicle
4. Libraire Encyclopedique de Roret, Paris. Pp. 761-1084.

Starks. P.T. 2002. The adaptive significance of stabilimenta in orb-
webs: a hierarchical approach. Annales Zoologici Fennici
39:307-315.

Tanaka, K. 1984. Rate of predation by a kleptoparasitic spider,
Agyrodes fissifrons, upon a large host spider, Agelena limbata.
Journal of Arachnology 12:363-367.

Tolbert, W.W. 1975. Predator avoidance behaviors and web defensive
structures in the orb weavers Argiope aurantia and Argiope
trifasciata (Araneae, Araneidae). Psyche 82:29-52.

Townley, M.A. & E.K. Tillinghast. 1988. Orb web recycling in
Araneus cavaticus (Araneae, Araneidae) with an emphasis on the
adhesive spiral component, GABamide. Journal of Arachnology
12:303-319.

Townley, M.A., E.K. Tillinghast & C.D. Neefus. 2006. Changes in
composition of spider orb web sticky droplets with starvation and
web removal, and synthesis of sticky droplet compounds. Journal
of Experimental Biology 209:1463-1486.

Tso, I-M. 1996. Stabilimentum of the garden spider Argiope
trifasciata : a possible prey attractant. Animal Behaviour
52:183-191.

Tso, I-M. 1999. Behavioral response of Argiope trifasciata to recent
foraging gain: a manipulative study. American Midland Naturalist
141:238-246.

Tso, I-M. 2004. The effect of food and silk reserve manipulation on

decoration-building of Argiope aetheroides. Behaviour 141:603-
616.

Vollrath, F. 1987a. Growth, foraging and reproductive success.
Pp. 274-286. In Ecophysiology of Spiders. (W. Nentwig, ed.).
Springer, Berlin.

Vollrath, F. 1987b. Kleptobiosis in spiders. Pp. 357-370. In
Ecophysiology of Spiders. (W. Nentwig, ed.). Springer, Berlin.

Walter, A., P. Bliss, M.A. Elgar & R.F.A. Moritz. 2008a. Argiope
bruennichi shows a drinking-like behaviour in web hub decorations
(Araneae, Araneidae). Journal of Ethology. Doi 1 0. 1 007/s 1 0 1 64-
007-0077-5.

Walter, A., M.A. Elgar, P. Bliss & R.F.A. Moritz. 2008b. Wrap
attack activates web-decorating behavior in Argiope spiders.
Behavioral Ecology, doi: 10. 1093/beheco/arn030.

Watanabe, T. 2000. Web tuning of an orb-web spider, Octonoba
sybotides, regulates prey-catching behaviour. Proceedings of the
Royal Society of London B 267:565-569.

Manuscript received 31 August 2007, revised 21 April 2008.

2008. The Journal of Arachnology 36:545-551

Ontogenetic changes in web architecture and growth rate of Tengella radiata (Araneae, Tengellidae)

Gilbert Barrantes and Ruth Madrigal-Brenes1: Escuela de Biologia, Ciudad Universitaria Rodrigo Facio, Universidad de
Costa Rica, San Pedro, San Jose, Costa Rica

Abstract. In some spiders features of the webs of early instars may represent features of the ancestor's web. Some second
instar spiderlings (first instar outside of the egg sac) of Tengella radiata (Kulczynski 1909) construct a small sheet web
without any type of retreat. In subsequent instars, spiderlings construct webs that consist of a sheet with a small retreat that
opens near its center. Webs gradually change as spiderlings growth and webs of 7th instar spiders are indistinguishable from
those of adult females. Spiders only begin to include cribellate threads in their webs during the 7th instar. The growth of T.
radiata is slow during the first three instars, but spiders’ sizes increase steadily in the subsequent stages. Legs I of adult
males are longer than in females, indicating an allometric growth that occurred mainly during the last molt of males.

Keywords: Spiderlings’ webs, cribellate silk, cephalothorax and leg growth

Little is known about ontogenetic changes in most spiders’
webs, particularly in non-orb weavers (Eberhard 1990). The
first webs constructed by newly emerged spiderlings (second
instar spiderlings) in several spider families differ from the
webs of adults and they tend to represent less derived stages or
characters compared to the features of the webs of adult
spiders (Nielsen 1931; Eberhard 1977, 1985, 1986, 1990;
Robinson & Lubin 1979). Some of these differences might
also be related to underdevelopment of silk glands, as is the
case in Uloborus diversus Marx 1898 (Eberhard 1977). Webs
constructed by young spiderlings of this species lack cribellate
silk, and radial threads are more numerous than in webs of
adult spiders. Another possible difference in young spider-
lings’ webs may also be related with the type and size of prey
that spiderlings can handle (Lubin 1986).

Webs of young cribellate and non-cribellate spiderlings have
been studied in Orbiculariae orb-weaving spiders (Eberhard
1977, 1985, 1986). In some of these spiders in which
architecture of adult webs depart from a typical orbicular
web (Eberhard 1985, 1986), the webs of newly emerged
spiderlings are orbicular, indicating that their ancestors
probably possessed orbicular webs. A similar pattern is
showed by cribellate and non-cribellate non-orb weavers such
as the psechrid Fecenia sp. (Robinson & Lubin 1979) and some
theridiid species (Nielsen 1931; Szlep 1965).

In this study, we focus on the cribellate spider Tengella
radiata (Kulczynski 1909), a species of Tengellidae that builds
funnel webs and is restricted to Costa Rica (Eberhard et al.
1993). Its web is regularly inhabited by some symbiont spiders
(e.g., Philoponella sp., Mysmenopsis spp.) and plokiophilid
bugs (e.g, Lipokophila spp.) (Eberhard et al. 1993). The
structure and production of the cribellate threads of this spider
have been described and compared with threads of other
cribellate spiders (Eberhard 1988; Eberhard & Pereira 1993).
The courtship and copulation have also been reported and
compared with those of other spiders in related families
(Barrantes 2008), and Santana et al. (1990) investigated the
predation rate in the field and estimated the metabolic rate of
this spider. Other than this reported work, there have been no

‘Corresponding author. E-mail: ruthymad@gmail.com

published studies on the ontogenetic changes in the funnel,
criballete web. or on the growth of this spider.

If there are ontogenetic changes in webs, we expect there to be
transitional stages between the first webs and the adult webs. We
describe here the architecture of the adult web and changes in the
web architecture that occur between the immature stages of
Tengella radiata. We also describe the number of molts, growth,
and feeding behavior of spiderlings for this spider.

METHODS

We observed (/? > 30) and photographed (/; = 5) webs of
wild mature females of T. radiata in San Antonio, Escazu, San
Jose province (= SAE; 10°56'N, 84°08'W) and used this
information to describe the adult web. The egg stage period
and maternal behavior were described from a female raised
from eggs in captivity and maintained in a plastic box (30 x 18
X 1 1 cm) where she constructed her web.

In order to rear and study numerous spiderlings we
collected egg sacs from two mature female from two localities.
One was from an adult female in SAE and a second from a
female from La Selva Biological Station (= LSBS; 10’26'N,
83°59'W), Organization for Tropical Studies, Heredia prov-
ince, Costa Rica. Each egg sac was placed on a mass of dry
cotton inside a plastic container ( 13 X 13x6 cm) with a little
piece of humid cotton at a corner. All containers were
maintained indoors at a temperature of 1 8—20 C. From one
egg sac (SAE), 16 spiderlings were separated as they emerged
and each one was maintained individually in a plastic
container (13 X 13x6 cm); some of these spiderlings were
placed in a larger container (30 X 15 X 10 cm) when they
reached their seventh instar. Webs of some of these 16
individuals were photographed and their shed cuticles
measured at each stage (details described below) to describe
the ontogenetic changes in the web and the growth of this
spider. From a second egg sac (LSBS) we separated three
groups of 10 spiderlings and each group was maintained in a
container (13 X 13x6 cm). Feeding behavior was observed in
these groups of spiderlings. (Spiderlings from the second egg
sac were released to the wild as they molted the third time as
were those remaining spiderlings from the first egg sac). We
consider spiderlings that recently emerged from the egg sac as
second instar individuals (Foelix 1996).

545

546

THE JOURNAL OF ARACHNOLOGY

The plastic container of each of the 16 spiderlings kept
individually had four small rocks, whose base and height
varying from 2 to 3 cm. One rolled dry leaf, forming a cone of
ca 2 cm with an opening of ca 0.5 cm in diameter, was
attached to one of the rocks with masking tape; rocks were all
fastened to the bottom of the container. This “rock
landscape” was fashioned to offer spiderlings enough supports
to construct their webs, and the rolled leaf was offered as a
“natural tunnel.” Two to four webs constructed by spiderlings
from 3rd to 6th instars were lightly coated with cornstarch and
then photographed. The longest and widest sides of these,
approximately rectangular webs were measured. Spiderlings of
the photographed webs were collected and preserved in
alcohol to avoid possible effects of cornstarch in the
construction of subsequent webs. The web of the second
instar spiderlings was sketched using photos and observations
under the dissecting microscope; cornstarch adhered to
threads of these webs was insufficient to allow good contrast
photographs.

We measured the length of tibiae and femora of legs I and
IV, and the length and width of the cephalothorax in shed
cuticles of the different stages of those 16 spiders maintained
individually in containers, three adult males, and three adult
females to estimate the spiders’ growth between stages (2nd
instar to adult stage). Sample sizes were not the same for each
stage because some spiderlings were collected and some shed
cuticles were destroyed when we withdrew them from the web,
and tibiae in all second stage cuticles collapsed and were
impossible to measure. To calculate the percentage of growth
between subsequent stages, we subtracted the length of a
particular structure to the mean of that structure of the
previous stage (Mps), then divided this difference between Mps,

and multiplied this proportion by 100 (e.g., [FI2 1 — MFIi]/
MFI] * 100; FI2 1 - femur I from individual 1 of second instar,
MFIr mean of femur of all individuals of instar 1). We used a
digital camera (Nikon, Coolpix 4500) to photograph each
femur, tibia, and cephalothorax under a dissecting micro-
scope, and measured them using the software program Image
Tool v. 3.0. Four 3rd instar spiderlings were observed under
the dissecting microscope to check for the presence of the
cribellar plate and calamistrum.

During the 2nd and 3rd instars, spiderlings were offered a
Drosophila fly every other day, 4th instar spiderlings were
offered two Drosophila flies every other day, and spiderlings of
later instars were offered one blow fly (Calliphoridae) or a
moth every three days. All containers had a small piece of wet
cotton for the spiderlings to drink. Feeding behavior
observations were made on both solitary spiderlings and on
those maintained in groups. Additional behavioral observa-
tions of adults and spiderlings were obtained from adult
spiders raised from eggs in captivity and complemented with
the field information from SAE and LSBS. Voucher specimens
of spiders from all stages were deposited in the Museo de
Zoologia of the Universidad de Costa Rica, San Jose.

RESULTS

Adult web.— The adult web of T. radiata consists of a large,
more or less triangular, horizontal sheet with a dispersed
tangle above and beneath it (Fig. 1; Eberhard et al. 1993). It is
usually built near a large object such as rock or a tree trunk.
At the “interior,” most protected section (the apex) of the web
the spider constructs a tunnel that varies between 5 and 20 cm
long (it = 20). The spider rests at the mouth of the tunnel or
inside it during the day. At night (// = 10) the spider usually

Figure 1 . - Frontal view of an adult web of T. radiata (without cornstarch): a. Threads of the upper tangle, b. Tunnel at the “interior” section
of the sheet, c. Sheet.

BARRANTES & MADRIGAL-BRENES— WEB ARCHITECTURE IN TENGELLA RADI AT A

547

rests at the mouth of the tunnel, but is often observed
repairing the web or producing adhesive cribellate threads that
she lays on the sheet and upper tangle. Newly constructed
webs frequently lack cribellate threads, but this silk accumu-
lates on the sheet and upper tangle over time, until the web
collapses apparently due to the weigh of debris accumulated
on its sheet. Adult males (n = 3) did not build webs when
placed in a large plastic container. However, they killed and
fed on prey that walked nearby. Females (n — 5) placed in
similar containers built a complete web.

Duration of egg stage and maternal care. — One spider
produced two egg sacs in captivity, the first 32 and the second
48 days after copulating. Spiderlings emerged from the first
egg sac 58 days later; no spiderlings emerged from the second
egg sac. The egg sacs were attached to the roof of the tunnel,
as were five egg sacs observed in the field. In captivity, the
spider added some small pieces of prey cuticle to the external
surface of the egg sacs; egg sacs observed in the field were also
covered with detritus. The female spent most of her time

(except when capturing prey and feeding) hanging from the
tunnel with her legs and palps surrounding the egg sac. Nearly
the entire time, she contacted the egg sac with legs II and III
(less frequently with legs IV), and her palps. Occasionally, she
stood on the bottom of the tunnel, with one of her legs
(usually one leg III) raised to contact the egg sac. When the
spider produced the second egg sac, she concentrated her care
on the second sac. The spider died after living for 18 months
and 20 days (from emergence through adulthood).

Ontogenetic changes in webs. — During the second instar, 4
of the 16 spiderlings used the tunnels (rolled leaf provided) but
did not construct webs, 8 remained under a rock, and 4
constructed a web that consisted of a small, more or less
rectangular, sheet (Fig. 2). At first the web consisted only of
more or less horizontal threads that extended between rocks or
between rocks and the container wall. The horizontal threads
were about 1.5 cm above the container floor and other threads
connected them to the floor and to the container wall above.
These threads were part of the scaffolding that supported the

Figures 2-7. — Webs of early stages of T. radiata (with cornstarch): 2. Molting web of second stage spiderlings; 3. Type I web of third stage
spiderling; 4. Type II web of third stage spiderling; 5. Web of fourth stage spiderling; 6. Web of fifth stage spiderling; 7. Web of seventh stage
spiderling. White arrows in Figures 3, 4, 5, and 7 show the tunnel opening. The right tip of the short white line in Figure 5 shows the spiderling
inside the retreat.

548

THE JOURNAL OF ARACHNOLOGY

rest of the web. During the next days the spiderlings
constructed a dense sheet of very fine threads. The angle of
the sheet varied from nearly horizontal to about 15° (angles
were visually estimated). The complete web was constructed
over the first 2 to 4 days. When the web was finished, the
spiderling remained motionless, near the center, and on top of
the sheet until its next molt. These webs all lacked tunnels or
any other type of retreat. During this stage, which lasted
11.3 days (SD = 0.4) spiderlings did not feed; they did not
react to the presence of prey on their webs.

The third-stage spiderlings (// = 16) constructed two types
of webs. The most common type (13 out of 16) was a small
(range = 3-4 x 2-2.5 cm) dense, more or less horizontal sheet
with a resting place (retreat) that was constructed nearly
perpendicularly under the sheet (type 1) (Fig. 3). This retreat
was a small bag ofloose silk with the exit near the center of the
sheet (Fig. 3). The other web (type II) was also a flat sheet, but
with the bag-like retreat constructed on the sheet forming a
small roof, with a relatively dense tangle above the sheet
(Fig. 4). Threads of the tangle served as support for the much
finer threads of the sheet and retreat. All spiderlings in this
stage captured Drosophila flies dropped on the sheet.

The webs of fourth-stage spiderlings were in general larger
(range = 5-6 X 3-3.5 cm) than those of the previous stage.
Spiderlings expanded the sheet and retreat of the web
constructed in the previous stage, and constructed a tunnel
under the sheet. The closed end of the tunnel nearly reached
one border of the sheet (Fig. 5), and spiderlings rested deep
inside the tunnel (Fig. 5). The general shape of these webs was
similar to webs of the previous instar and type I and II webs
were still distinguishable. Webs of fifth and sixth stages were
similar in the general features but larger compared to webs of
the previous stage (Fig. 6).

Seventh-stage juveniles constructed a much denser tangle
above the level of the sheet. In this stage the tunnel of two webs (n
= 9) was constructed on the sheet for most of its length, with its
opening near the center of the sheet. The furthest end of the tunnel
curved down and under the sheet, similar to a type II design
( Fig. 7). The other seven webs were similar to the adult webs, with
the opening of the tunnel at one extreme of the sheet. This was the
first stage in which cribellate threads were observed on the sheet
and upper tangle of the web (Fig. 8), though the cribellar plate
and calamistrum were already present in 3rd instar spiderlings.
The web of older stages (in larger containers) was indistinguish-
able from webs of adult spiders.

Spiderling feeding behavior. Third and fourth instar
spiderlings attacked by rushing onto the sheet to the prey
and biting it. If prey was small, relative to the spiderling size,
the spiderling fed on the prey without releasing it. If prey was
large, it was released by the spiderling as soon as the prey’s
struggling stopped, returning to feed on it a few seconds later.
One fourth instar spiderling (in a group of 10) wrapped the
prey after biting it. The spiderling released the prey as it
stopped struggling and nearly immediately began to wrap it.
During wrapping, the spiderling turned in place while
attaching wrapping lines to the substrate (the sheet). The
wrapping movements were similar to those described for adult
spiders (Barrantes & Eberhard 2007). Another spiderling
approached a large blow fly (larger than a house fly) slowly.
The spiderling backed away as the fly struggled and then cut

Figure 8. — Cribellate threads (white arrow) from webs of seventh
stage spiderlings.

some sheet threads and walked, hanging under the sheet to the
prey and bit one of the fly’s legs through the sheet. Neither
wrapping nor attacks from under the sheet were observed in
any fifth instar individuals. Older juvenile spiders (sixth or
older instars) always moved on the upper surface of the sheet,
and often wrapped large prey (e.g., moths and large flies), with
movements similar to those of adult spiders (Barrantes &
Eberhard 2007). Large prey were carried inside the tunnel by
7th stage, subadult, and adult spiders, and their carcasses were
left inside the tunnel.

Additional behavioral observations. — Early instar spiders did
not molt at any consistent site in the web (n > 50). Flowever,
older stages (7th to pre-adult) molted inside the tunnel but they
carried the shed skin to the farthest extreme of the web (n = 9).

Number of molts and growth. — The males had eight molts
(» = 4) and the females nine (« = 4) to attain their adult
stages. The mean time from emergence from the egg to the
eighth molt was 186.7 days (± 10.5), when males molted to the
adult stage. Females lasted 42 days (± 7) more to their next
and last molt (Fig. 9).

The general pattern of growth of the cephalothorax (width and
length) and legs I and IV (tibia and femur) were similar. In the
three first stages these structures grew very little, but during the
following stages the size of the cephalothorax and legs increased
steadily (Figs. 10-12). In fact, the growth of all morphological
features was proportionally much larger between the third and
the fourth stage, than between other subsequent stages (Table 1 ).
It is also evident that the length of the cephalothorax increased
faster than its width ( Fig. 1 0), indicating an allometric growth of
the length relative to the width of the cephalothorax. In addition,
legs of adult males were also notably longer than those of adult
females, despite the additional molt of females.

DISCUSSION

The architecture of the webs of T. radiata changes as spiders
mature. The first web constructed by 2nd instar spiderlings,
which have recently emerged from the egg sac, apparently
serves as a molting place, since spiderlings in this stage did not
capture prey. In nature and in captivity the second instar
spiderlings construct a communal molting web inside the

BARRANTES & MADRIGAL-BRENES— WEB ARCHITECTURE IN TENGELLA RADIATA

549

9 55

50

45

40

35

m

m 30
Q

25

20

15

10

5

123456789

Stages

11

10

Molt

Figures 9-12. — Inter-molt time and growth of T. radiata : 9. Intermolt time (mean, standard error and standard deviation) from the first to the
ninth stage out of the egg sac; sample size above each stage mean; 10. Mean and standard deviation of the length (solid line) and width (dotted
line) of cephalothorax (cph); sample size above each mean, F and M indicate the values for adult females and adult males respectively; 1 I . Mean
and standard deviation of the length of femur (solid line) and tibia (dotted line) of leg I; 12. Mean and standard deviation of the length of femur
(solid line) and tibia (dotted line) of leg IV.

tunnel of their mother’s web, and only begin to abandon the
tunnel and the web after they molt to the third instar (GB
unpublished data).

Both of the two types of webs constructed by third through
sixth instar spiderlings had retreats that opened near the center of
the sheet (Figs. 3-7). If, as in other groups, the webs of early
instars represent less derived characters than those of the adult

web (Eberhard 1986, 1990), the ancestral web of Tengella might
have been a sheet with a tunnel retreat extending below its center.
However, comparative data of closely related species (Codding-
ton 2005) are required to test this hypothesis.

The lack of cribellate threads on the sheet of juvenile
spiders (2nd to 6th instar) may either represent an ancestral
condition, an undeveloped condition of the cribellum

Table 1. — Percentage of growth of the width and length of the cephalothorax (cph) and the femur (F) and tibia (T) of legs 1 and IV between
successive stages (first row). Sample size in parentheses beside or under the stage codes.

VII to VIII VIII to IX

I to II (5)

II to III (9)

III to IV (6)

IV to V (9)

V to VI (9) VI to VII (8)

(7)

(6)

VIII to 3 (3) IX to 9 (3)

Cph W

0.6 ± 9.9

19.9 ± 1.9

104.4 ± 11.7

30.7 ± 2.9

19.8 ± 4.3

35.3 ± 8.7

23.9 ± 5.8

18.6 ± 3.5

32.9 ± 4.7

30.5 ± 3.3

Cph L

0.1 ± 0.2

21.6 ± 3.3

106.4 ± 9.2

35.6 ± 4.4

20.1 ± 3.7

36.6 ± 9.7

26.5 ± 6.4

19.3 ± 3.4

30.5 ± 3.2

25.3 ± 2.6

FI

21.4 ± 5.2

41.0 ± 3.9

104.4 ± 14.6

46.5 ± 5.5

22.4 ± 9.8

42.0 ± 8.4

23.8 ± 8.7

23.7 ± 3.7

82.4 ± 5.9

15.8 ± 6.5

TbI

56.6 ± 5,6

114.5 ± 8.0

41.8 ± 7.3

24.4 ± 8.1

38.9 ± 5.9

30.6 ± 10.8

18.5 ± 4.5

84.8 ± 7.7

18.3 ± 7.1

FIV

5.9 ± 3.3

32.3 ± 4.0

110.9 ± 17.2

42.0 ± 5.0

26.2 ± 5.5

38.7 ± 6.0

26.3 ± 13,3

22.8 ± 5.3

67.8 ± 2.6

14.2 ± 1.4

TbIV

29.6 ± 5.5

109.5 ± 13.3

44.0 ± 4.2

25.1 ± 5.5

37.5 ± 8.3

26.3 ±11.1

23.1 ± 4.7

86.3 ± 4.3

20.2 ± 2.0

550

THE JOURNAL OF ARACHNOLOGY

apparatus, or both. The presence of the cribellum and
calamistrum in third instar spiderlings but the lack of
criballate silk on their webs, indicate that these structures
may not be functional in the early stages of T. radiata, as
occur in some uloborids (Eberhard 1977; Opell 1982), or that
the high demand of energy involved in drawing the cribellar
fibrils (Eberhard 1988) prevent young spiderlings from using
this type of silk.

The lack of cribellate silk in early stages and newly
constructed webs of adult T. radiata did not prevent spiders
from using their webs to capture prey, though cribellate silk
possibly restrain and reduce movements of prey on the web.
Spiderlings at early stages showed attack and wrapping
behaviors similar to those of adult spiders (Barrantes &
Eberhard 2007). The wrapping behavior of spiderlings only
differed from that of adults in that spiderlings did not hold the
prey while wrapping it as adults do (Barrantes & Eberhard
2007). In addition, one spiderling attacked a large prey from
under the sheet. This attack is apparently restricted to early
spiderling stages, as it has not been observed in either large
juveniles or adult spiders. The low frequency of occurrence of
this behavior is unclear since attacking a large prey from under
the sheet provides some protection to the spiderling as the
sheet restrains the struggling prey and reduces the force of its
movements (Robinson 1975; Lubin 1980).

In general the growth of T. radiata is slow, as expected from
its apparently extremely low metabolic rate (Santana et al.
1990). Growth was very slow during the earliest stages
(Figs. 10-12), but the relative increment in cephalothorax
and leg size was more than 100% between the third and fourth
stage. After the fourth stage, the cephalothorax and legs
increased steadily with each subsequent molt, though the
intermolt period tended to increase with each stage. This
possibly indicates the need for more energy and time for
growth and development of internal organs as spiders mature.
Length of legs 1 and IV is nearly the same until the fourth
instar. However, in the following stages, increments in the
length of leg I are larger than in leg IV, possibly reflecting the
different functions of these legs in young and adult spiders.
For example, the tactile function of legs I likely favor their
longer length (Foelix 1996).

The comparatively longer legs of adult males result from
the allometric growth that occurred mainly during the last
molt of males, as happens in wolf spiders (Framenau 2005).
Though experimental evidence is lacking, field observations
(e.g., males observed near or on females’ webs) indicate that
adult males abandon their webs to find receptive females, so
that longer legs may result in greater step size to bridge gaps.
If a more efficient search for females lead to a higher
reproductive success in males, it is likely that longer legs in
males evolved, at least partially, through indirect male-male
sexual competition (Anderson 1994; Framenau 2005).
Additionally, natural selection might have also favored
longer legs in males, if such a trait allows them to run faster
to escape from predators and from females during courtship,
and provides them a greater sensory range (Gertsch 1949;
Framenau 2005). Accordingly, the larger size of adult
females’ cephalothorax correlates with their larger body
which is related to their capacity to produce large numbers of
eggs (Gertsch 1949).

ACKNOWLEDGMENTS

We thank W.G. Eberhard, G. Stratton, and two anonymous
reviewers for providing valuable comments that greatly
improved this manuscript. We also thank J-L Weng for the
picture of cribellate threads. Research was supported by
Vicerrectoria de Investigation, Universidad de Costa Rica.

LITERATURE CITED

Anderson, M. 1994. Sexual Selection. Princeton University Press,
Princeton, New Jersey. 599 pp.

Barrantes, G. 2008. Courtship behavior and copulation in Tengella
radiata (Araneae, Tengellidae). Journal of Arachnology 36:606-608.
Barrantes, G. & W.G. Eberhard. 2007. The evolution of prey
wrapping behaviour in spiders. Journal of Natural History
41:1631-1658.

Coddington, J.A. 2005. Phylogeny and classification of spiders.
Pp. 18-24. In Spiders of North America: an identification manual.
(D. Ubick, P. Paquin, P.E. Cushing & V. Roth, eds.). American
Arachnological Society.

Eberhard, W.G. 1977. The webs of newly emerged Uloborus diversus
and of male Uloborus sp. (Araneae, Uloboridae). Journal of
Arachnology 4:201-206.

Eberhard, W.G. 1985. The “sawtoothed” orb of Eustala sp., with a
discussion of the ontogenetic patterns of change in web design in
spiders. Pyche 92:105-118.

Eberhard, W.G. 1986. Ontogenetic changes in the web of Epeirotypus sp.

(Araneae, Theridiosomatidae). Journal of Arachnology 14:125-128.
Eberhard, W.G. 1988. Combing and sticky attachment behaviour by
cribellate spiders and its taxonomic implications. Bulletin of the
British Arachnological Society 7:247-251.

Eberhard, W.G. 1990. Function and phylogeny of spider webs.

Annual Review of Ecology and Systematics 21:341-372.

Eberhard, W.G. & F. Pereira. 1993. Ultrastructure of cribellate silk of
nine species in eight families and possible taxonomic implications
(Araneae: Amaurobiidae, Deinopidae, Desidae, Dictynidae. Filis-
tatidae, Hypochilidae, Stiphidiidae, Tengellidae). Bulletin of the
British Arachnological Society 21:161-174.

Eberhard, W. G, N. Platnick & R.T. Schuh. 1993. Natural history
and systematics of arthropod symbionts (Araneae; Hemiptera;
Diptera) inhabiting webs of the spider Tengella radiata (Araneae,
Tengellidae). American Museum Novitates No. 3065. 17 pp.
Foelix, R.F. 1996. Biology of Spiders. Second edition. Oxford
University Press, New York. 330 pp.

Framenau, V.W. 2005. Gender specific differences in activity and
home range reflect morphological dimorphism in wolf spiders
(Aranea, Lycosidae). Journal of Arachnology 33:334-346.

Gertsch, W.J. 1949. American Spiders. D. Van Nostrand Company,
Inc., Princeton, New Jersey. 285 pp.

Lubin, Y.D. 1980. The predatory behavior of Cyrtophora (Araneae,
Araneidae). Journal of Arachnology 8:159-183.

Lubin, Y.D. 1986. Web building and prey capture in the Uloboridae.
Pp. 132-171. In Spiders: Webs, Behavior, and Evolution. (W.A.
Shear, ed.). Stanford University Press, Stanford, California.
Nielsen, E. 1931. The Biology of Spiders with Especial Reference to
the Danish Fauna. Levin & Munksgaard, Copenhagen, Denmark.
248 pp.

Opell, B.D. 1982. Post-hatching development and web production of
Hyptiotes cavatus (Hentz) (Araneae, Uloboridae). Journal of
Arachnology 10:185-191.

Robinson, M.H. 1975. The evolution of predatory behaviour in
araneid spiders. Pp. 292-312. In Function and Evolution in
Behaviour. (G. Baerends, C. Beer & A. Manning, eds.). Clarendon
Press, Oxford, UK.

Robinson, M.H. & Y.D. Lubin. 1979. Specialists and generalists: the
ecology and behavior of some web-building spiders from Papua

BARRANTES & MADRIGAL-BRENES— WEB ARCHITECTURE IN TENGELLA RADI AT A

551

New Guinea II. Psechrus argentatus and Fecenia sp. (Araneae:
Psechridae). Pacific Insects 21:133-164.

Santana, M., W.G. Eberhard, G. Bassey, K.N. Prestwich & R.D.
Briceno. 1990. Low predation rates in the field by the tropical
spider Tengella radiata (Araneae: Tengeilidae). Biotropica
22:305-309.

Szlep, R. 1965. The web-spinning process and web-structure of
Latrodectus tredecimguttatus, L. pallidus and L. revivensis.
Proceedings of the Zoological Society of London 145:75—
89.

Manuscript received 24 September 2007, revised 26 May 2008.

2008. The Journal of Arachnology 36:552-556

Does the microarchitecture of Mexican dry forest foliage influence spider distribution?

Pablo Corcuera1, Maria Luisa Jimenez2 and Pedro Luis Vaiverde1: 'Universidad Autonoma Metropolitana-Iztapalapa,
Departamento de Biologia, Av. San Rafael Atlixco 186, Col. Vicentina, Iztapalapa, C.P. 09340, Mexico D.F.; 2Centro
de Investigaciones Biologicas del Noroeste, Apdo. Postal 128, La Paz, B.C.S. 23090, Mexico. E-mail:
pcmr@xan um . uam . mx

Abstract. Spider species diversity has been associated with vegetation structure and stratification but there are few studies
comparing the spider distribution in different shrubs and trees. In this study we analyzed the species distribution of the
spider community of 1 1 shrub and tree species in two different study sites in a Mexican tropical dry forest. We present
results from multivariate analyses that explain their distribution. A classification analysis based on spider abundances
separated one shrub, Croton ciliatoglanduliferus , from the rest of the plant species. This was explained by the presence of
large numbers of the oxyopid Peucetia viridans (Hentz 1832) on this plant. A second cluster segregated broad-leaved from
small-leaved, bipinnate species. This was mainly due to higher spider abundances in the latter type of plants. Four
vegetation variables were estimated and their influence on the species distribution was assessed by means of a principal
components and regression analysis. With the exception of P. viridans, all spiders were positively associated with number of
leaves and number of branchlets per 50 cm branch and negatively with foliage area.

Keywords; Community ecology, species abundances, plant structure

Habitat structure is an important factor that influences
diversity, abundance, and distribution of spider species (Lubin
1978; Hatley & MacMahon 1980; Evans 1997; Whitmore et al.
2002). The available evidence has been gathered from both
natural communities (e.g., Lubin 1978; Robinson 1981; Raizer
& Amaral 2001) and agricultural systems (Rypstra et al. 1999;
Samu et al. 1999). Habitat structure and complexity are
related to factors such as prey abundance, shelter against
enemies and suitable microclimatic conditions (Riechert &
Tracy 1975; Gunnarsson 1996; Halaj et al. 1998; Raizer &
Amaral 2001). Habitat preferences, however, can be highly
specific and species belonging to different guilds have
particular requirements according to their morphological,
physiological, and behavioral features (Turnbull 1973; Wise
1993).

Variation in plant height, foliage density, leaf surface area,
number of leaves and branchlets, and number and type of
inflorescences, can affect the abundance and distribution of
foliage-dwelling spiders (Hatley & MacMahon 1980; Evans
1997; Halaj et al. 1998; Uetz et al. 1999; Raizer & Amaral
2001; Corcuera et al. 2004; Heikkinen & MacMahon 2004;
Souza & Martins 2004, 2005). In this study, we evaluated the
influence of plant architecture on the spider community by
means of multivariate and regression analyses. We analyzed
the abundance of foliage spiders and four plant attributes of
1 1 of the most abundant trees and shrubs found in a tropical
dry forest in western Mexico.

Information on Mexican spiders is widely dispersed. After
extensive bibliographical research, Jimenez (1996) found 7,916
species. It is not known how many specimens were collected
from foliage since most studies were concerned with taxonomy
(Jimenez 1996). There have been a few reports on foliage
spiders on cacao and coffee plantations (Ibarra Nunez et al.
1995, 1997; Moreno-Molina et al. 2001; Pinkus-Rendon et al.
2006), but these studies concentrate on species richness and
diversity. Besides a diversity analysis (Corcuera et al. 2004), to
our knowledge nothing has been written on the distribution of
foliage spider communities in dry forests.

METHODS

Study sites. — Tropical dry forests cover 42% of the tropical
and subtropical land area on the planet (Murphy & Lugo
1986). The dominant plant species are strongly drought
deciduous (Mooney et ai. 1989). In Mexico, they are the
prevailing vegetation type along the west coast and cover ca.
12.4% of the country’s area (Arizmendi et al. 1990). Mexican
tropical dry forests are found between 0 and 1990 m elevation
(Rzedowski 1978). Mean annual temperature ranges from 20
to 29°C, and mean annual precipitation from 300 to 1800 mm
(Rzedowski 1978). Dry forests are strongly seasonal, with a
long dry season and intense rainy season (Rzedowski 1978;
Murphy & Lugo 1986).

The study sites are located in the Municipality of Villa
Corona in the state of Jalisco (20°20'N, 103°35'W). Altitude
above sea level is 1 640 m. Mean annual temperature was
20.3°C and mean annual precipitation from the last 15 years
was 826 mm. Most of the rain falls between mid-June to mid-
September and there are between 6 and 8 dry months each
year.

Plant variables. — Eleven trees and shrubs were sampled in
two sites (El Caracol and Charco Verde) to test the effect of
plant architecture on the distribution of foliage spiders:
Bursera schlechtendalii, B. bipinnata, Guazuma ulmifolia ,
Heliocarpus appendiculatus , Ipomoea woleottiana, Prosopis
juliflora. Mimosa galeotti, Lysiloma acapulcense , Croton
ciliatoglanduliferus , Acacia cymbispina, and Byrsonima sp.
(Table 1). These plant species are typical of Mexican dry and
thorn forests and were the most common shrubs and trees in
the study sites (Table 2). Details about plant cover estimation
are given elsewhere (Corcuera & Butterfield 1999; Corcuera &
Zavala-Hurtado 2006).

Foliage area, number of leaves and number of branchlets
(i.e., small branches) were determined for each plant species.
The sample unit was a 50 cm terminal branch from a limb
rising horizontally from the center of the plant (McCaffrey et
al. 1984). Foliage area was measured by drawing the contour

552

CORCUERA ET AL.— PLANT ARCHITECTURE AND SPIDER DISTRIBUTION

553

Table 1. — Mean (± SD, n = 10 per species) of plant height, foliage area, number of leaves and branchlets on a 50 cm terminal branch in = 3
per species) for 1 1 species in two dry forest sites, El Caracol and Charco Verde, in the Municipality of Villa Corona, Jalisco, Mexico. * = small-
leaved species.

Plant species

Code

Plant height (m)

Foliage area (cm2)

Number of leaves

Number of branchlets

Bursera schlechtendalii

Busc

3.4 (0.32)

850 (386.1)

70.0 (28.28)

22.5 (6.4)

Bursera bipinnata *

Bubi

3.4 (0.71)

541 (140.3)

2473.2 (2692.16)

22.0 (9.9)

Croton ciliatoglanduliferus

Crci

1.1 (0.32)

1026 (631.8)

28.0 (16.17)

9.0 (4.2)

Guazuma ulmifolia

Guul

4.8 (0.98)

3349 (989.7)

45.8 (19.30)

21.5 (0.7)

Acacia cymbispina *

Accy

3.3 (0.71)

120 (27.6)

10670.4 (2313.43)

10.5 (4.9)

Prosopis juliflora *

Prju

3.9 (1.25)

242 (105.1)

6696.0 (2136.01)

11.0 (0.7)

Byrsonima sp.

Bysp

2.7 (0.70)

1712 (1065.1)

64.5 (21.71)

15.0 (4.2)

Ipomoea wolcottiana

Ipwo

5.0 (0.72)

2556 (960.2)

12.2 (7.79)

11.0 (2.1)

Heliocarpus appendiculatus

Heap

5.3 (0.67)

2050 (975.0)

15.0 (7.53)

8.0 (1.4)

Lysiloma acapulcense *

Lysp

4.4 (0.47)

340 (105.9)

30240.0 (11671.35)

11.5 (3.5)

Mimosa galeotti *

Miga

2.7 (0.48)

320 (209.6)

16301.0 (5250.44)

11.5 (7.8)

of all leaves present on the branch on millimetric paper with
1 mm divisions. The procedure was repeated on three
branches for each species and the mean area (cm2) per branch
was calculated. Mean number of leaves, or leaflets for
bipinnate species, and branchlets per branch was obtained
from the three samples of each species. Plant height was
recorded in a sample of 10 individuals for each species. All
plant variables were averaged from measures from both sites.

Spiders. — Spiders were collected by branch beating (South-
wood 1978) in June, July, September, October, and November
1999, and January and April 2000. For each plant species, a
terminal branch was chosen and beaten 10 times with a cane
(trial samplings showed that more strokes did not dislodge
more specimens) (Southwood 1978; McCaffrey et al. 1984).
This procedure was repeated on 10 individuals of each plant
species in each of the seven visits to each site. The specimens
were collected in 60 cm diameter muslin covered trays. Two
persons collected the spiders from the canvas using tweezers
and manual aspirators. McCaffrey et al. (1984) found that this
technique efficiently sampled the arachnofauna of foliage
dwelling spiders. The number of individuals for each spider
species (11 plants X 7 dates) was added in order to execute the
analyses, and the results were log transformed to obtain a
normal distribution. The specimens were preserved in 70%
alcohol and identified later at the Centro de Investigaciones

Biologicas (CIBNOR) in La Paz, Baja California. Voucher
specimens have been deposited in the collection at the
Laboratorio de Ecologia Animal, UAM-Iztapalapa, Mexico
City.

Multivariate analyses, — We analyzed the spider community
similarities with a classification using an unweighted pair
group average method with percent similarity. A Principal
Components Analysis (PCA) was used to analyze the
distribution of spider species in relation to the plant species.
Regressions were used to assess the relationship between the
main PCA axes and the plant variables. The classification and
ordination analyses were carried out using the statistical
software MVSP 3.2 (Multivariate Statistical Package; Kovach
1999).

RESULTS

Plant variables. — Foliage area was greater for broad-leaved
trees. G. ulmifolia had the largest area (3349 cm-), followed by
I. wolcottiana (2556 cm2), and H. appendiculatus (2050 cm2).
Croton ciliatoglanduliferus and Byrsonima sp. had intermediate
foliage areas (1026 cm2 and 1712 cm2, respectively); both are
broad-leaved shrubs. All small-leaved species had much lower
foliage areas (Table 1).

Lysiloma acapulcense and M. galleotti, small-leaved species,
had the highest mean number of leaves per terminal branch

Table 2. — Plant cover percentage and spider abundance and richness of 1 1 trees and shrub species in two dry forest sites, El Caracol (C) and
Charco Verde (V), in the Municipality of Villa Corona, Jalisco, Mexico. * = small-leaved species.

Species

Plant cover (%)

Spider abundance

Spider richness

C

V

C

V

C

V

Bursera schlechtendalii

0.6

2.7

49

31

9

9

Bursera bipinnata *

2.9

3.3

68

84

11

11

Croton ciliatoglanduliferus

9.9

0.8

87

83

11

6

Guazuma ulmifolia

2.3

5.4

28

29

7

1 1

Acacia cymbispina *

31.1

23.2

128

108

12

9

Prosopis juliflora *

4.8

0.5

70

119

10

11

Byrsonima sp.

4.8

1.6

43

27

12

7

Ipomoea wolcottiana

16.0

6.0

35

35

8

8

Heliocarpus appendiculatus

1.9

7.9

32

45

10

6

Lysiloma acapulcense *

2.9

18.8

53

65

10

11

Mimosa galeotti *

11.6

4.9

74

47

1 1

10

Other plant species

11.1

24.8

—

—

—

—

Total

100

100

667

673

—

—

554

THE JOURNAL OF ARACHNOLOGY

r

4

— r~

36

— r~

84

CrciV

CrciC

MigaV *'

GuulV

BuscV

ByspV

HeapC

IpwoC

IpwoV

ByspC

GuulC

BubiV *

PrjuV ®

AccyV*

AccyC*

HeapV

PrjuC *

LyspV «

MigaC «■

BubiC *

LyspC *

BuscC .

20

52

68

100

Percent Similarity

Figure 1. — Classification of 1 1 plant species in two sites according to spider species abundance. Plant codes are the same as in Table 1. The
additional last capital letter represents the sample site (C = Caracol, V = Charco Verde). * = small-leaved species.

(30,240 and 16,301, respectively), while the broad leaved trees,
H. appendiculatus and I. wolcottiana, had the lowest number of
leaves (15 and 12.2, respectively) (Table 1). Mean number of
branchlets per terminal branch was higher for B. schlechten-
dalii (22.5), Bursera bipinnata (22), and G. ulmifolia (21.5),
while C. ciliatoglanduliferus (9) and H. appendiculatus (8) had
the lowest values (Table 1). Mean plant height among species
varied from 1.1m (C. ciliatoglanduliferus ) to 5.3 m (H.
appendiculatus) (Table 1).

The dominant species in both sites was A. cymbispina, a
shrub that grows in areas that have been altered by cattle and
goat grazing. In both sites, P. juliflora was the second most
abundant species. C. ciliatoglanduliferus is an invasive shrub
particularly common in one site (El Caracol). In this site M.
galeottii was also dominant, while L. acapulcense was common
in Charco Verde (Table 2).

Spider abundance and composition. — A total of 1340 adult
spiders belonging to 2 1 species were caught in the two sampled
sites (667 in El Caracol, and 673 in Charco Verde) (Table 2).
Species composition was similar in both sites. Isaloides cf
yollotl (Jimenez, 1992), Hamataliwa puta (O. Pickard-Cam-
bridge 1894) and Peucetia viridans (Hentz 1832) represented
73% of the total numbers caught in El Caracol and 69% in
Charco Verde. Four species were represented by only one
individual: Micrathena gracilis (Walckenaer 1805) and Mallos
sp. in El Caracol, and Euryopis calif arnica Banks 1904 and
Ocrepeira sp. in Charco Verde. The other species found were
Neoscona oaxacensis (Keyserling 1864), Euriophora edax
(Blackwall 1863), Wamba crispulum (Simon 1895), Theridion
sp., Mimetus puritans Chamberlin 1923, Tmarus ehecaltocatl
Jimenez 1992, Misumenoides sp., Modysticus cf. floridana
(Banks 1895), Apollophanes punctipes (O.P. -Cambridge 1891),
Philodromus albicans O. Pickard-Cambridge 1897, Oxyopes
bifidus F.O. Pickard-Cambridge 1902, Phidippus sp., Para-
marpissa pi" a tic a. (Peckham & Pekham 1888) and Metaphi-
dippus cf. apicalis F.O. Pickard-Cambridge 1901.

In both study sites, the dominant families were hunters, in
particular Oxyopidae with 49% and 43% (El Caracol and

Charco Verde, respectively), followed by Thomisidae (36%
and 38%). The family Salticidae was represented by 10% and
1 1% of the total catch. In spite of spiders being sampled from
the foliage, web weavers were only represented by 1.5% of the
total catch in El Caracol, and 5.1% in Charco Verde.

Spider species distribution, — A classification of the plants
based on the spider abundances resulted in three main clusters
at the 50% similarity level (Fig. 1). The first cluster separated
C. ciliatoglanduliferus (Crci) from all other plant species. The
second cluster included all the broad-leaved species with the
exception of M. galeotti (MigaV) from Charco Verde, while
the third cluster included all the small-leaved plants and two
broad-leaved trees, H. appendiculatus (HeapV) from Charco
Verde and Bursera schlechtendatii (BuscC) from El Caracol.
The first division was explained by the presence of P. viridans,
one of the most abundant species, which was found
almost exclusively on C. ciliatoglanduliferus . Most spiders
had higher abundances in small-leaved plants (Table 2), which
explains the separation between the second and third clusters
(Fig. 1).

The first two PCA axes based on spider abundances
accounted for 88% of the variance (58% and 30%, respective-
ly). Since some spider species had less than 5 individuals, only
15 out of 21 species were included in the analysis. Nine of these
species were common to both sites. The first axis of the
ordination (eigenvalue = 0.58) was negatively correlated with
plant height (r = —0.87, P < 0.001). The ordination along this
axis was determined by the large numbers of P. viridans on C.
ciliatoglanduliferus (both had the highest scores on the positive
side (Fig. 2). The second axis (eigenvalue = 0.55) was
negatively correlated with foliage area (r = —0.91, P <
0.001) and positively with number of leaves (r = 0.86, P <
0.001) and branchlets (r = 0.61, P < 0.05). The ordination
pattern along this axis clearly segregated all spider species and
small-leaved bipinnate plants from broad-leaved plant species
(Fig. 2). Spider species, as well as all small-leaved bipinnate
species had positive scores. These plants had a high numbers
of leaves and branchlets, and low foliage area (Table 1,

CORCUERA ET AL.- -PLANT ARCHITECTURE AND SPIDER DISTRIBUTION

555

0.6

lyoV

Axis I

HpuV

PhspV

Lysp

Accy

PpiC
HpuC

lylpuV

Prju

lyoC

FhspC

Thi?

MspC PptfMiga

ThspV

feuC jrehV
Mspv'

Bubi

-0.4

Bysp

Heap

Busc

Guui

ipwo

-0.2
Axis 2

PviC

PviV

0.8

Crci

Figure 2. — Principal Components Analysis of the spider species present in 1 1 dry forest plant species. Spiders are: Pvi = Peucetia viridans, Iyo
= Isaloides cf yollotl, Hpu = Hamataliwa puta , Mpu = Mimetus puritans , Phsp = Phiddipus sp., Ppi = Paramarpissa piratica , The = Truants
ehecatlocatl, Thsp = Theridon sp.. Misp = Misumenoides sp. In the codes for spider species names, the additional last capital letter represents the
sample site (C = Caracol, V = Charco Verde). Codes for plant names are in bold and are the same as in Table 1.

Fig. 2). Conversely, the plant species on the negative side
included those with high foliage area but low number of leaves
and branchlets (Table 1, Fig. 2). The ordination shows the
relationship of each spider species with the plants. For
instance, I. cf. yollotl , a common spider species was
particularly abundant on A. cymbispina and that is why these
species appear together in Fig. 2.

DISCUSSION

In spite of intensive sampling (a total of 1540 branches
during a seven month period), only 21 species were found
among 1340 individuals collected in the two sampled sites (677
in El Caracol, and 673 in Charco Verde). Rarefaction analyses
(P. Corcuera, unpublished results) showed that only two or
three additional foliage species are likely to be found in the
study area.

Small-leaved plants appear to be suitable sites for foliage
spiders. Evans (1997) found that social crab spiders preferred
Eucalyptus species with smaller leaves. Perhaps more important-
ly, and regardless of plant taxon, density of leaves per branch
(e.g., Gunnarsson 1990; Souza & Martins 2005) as well as
structural complexity are better predictors of spider diversity.
Branching or twig density, as well as leaf density have been found
to be strongly related with number of spiders, diversity, and
abundance of various functional groups (Hatley & MacMahon
1980; Halaj et al. 1998; Corcuera et al. 2004). These variables also
explained the spider distribution in this study.

A classification of the plant species (Fig. 1) separated most
small-leaved bipinnate trees and shrubs from most broad-
leaved species. The second axis of an ordination also
segregated the plants and gave high positive scores to all
spiders and small-leaved plants and negative to all broad-
leaved (Fig. 2). This axis was positively correlated with
number of leaves and branchlets and negatively with foliage
area. The first axis was correlated with plant height and was
explained by high numbers of Peucetia viridans , a very

common spider in the study sites and the only one that was
associated with the small shrub Croton ciliatoglanduliferus.

Causal explanations for habitat preferences of foliage
spiders have not been explored in depth but some hypothesis
have been suggested. For example, Peucetia species are known
to favor plants with glandular trichomes, presumably because
arthropods are trapped by these hairs and represent available
prey for the spider (Vasconcelos-Neto et al. 2006). Halaj et al.
(1998) suggested that plants with higher cover are easier to
locate and might provide more resources. This might explain
higher diversities of most spiders on the most common trees
and shrubs. Total number of individuals was significantly
correlated with plant cover in the two sites (r — 0.83, P <
0.005 for El Caracol and r = 0.67, P < 0.05 for Charco
Verde). This may explain why A. cymbispina , which had the
highest plant species cover in both sites, supported high
densities of most spiders (Table 2). In the same way, M.
galeotti , a small-leaved tree from Chaco Verde, was included
in the broad-leaved cluster in the classification (Fig. 1,
Table 2). This species had low densities of spiders probably
because of its low cover in this site. However, P. juliflora and
B. bipinnata , with high richness and species abundances, had
very small cover in one or both sites (Table 2). Plants with
higher cover might be easier to locate, but they do not
necessarily provide more resources. Other factors (i.e., branch
and leaf density) appear to be more important to understand
the distribution of foliage spiders.

Some plant attributes might provide suitable microclimatic
conditions. Riechert & Tracy (1975) suggested that certain
plants might be favored because of their ability to modify the
thermal environment. In hot climates with long drought
periods, preserving body temperature would be a most
important factor. Small-leaved plants, especially C. cymbispina
and P. juliflora could provide a cooler environment because
they either do not shed their leaves (as does P. juliflora) or
remain green during the early draught, which is when

556

THE JOURNAL OF ARACHNOLOGY

spiderlings start to disperse. This species also starts producing
leaves early, before the rains, when broad-leaved trees and
shrubs are still deciduous. Once settled on these plants, there
would be no reason to move to shrubs or trees where
conditions would be less favorable.

Besides resource availability and favorable physical condi-
tions, accessibility of refuges against predators plays an
important role in determining spider distribution. Gunnarsson
(1996) suggested that high leaf densities could provide shelter
from bird predation. This would not seem the case in our
study sites, since bird attacks tended to be more frequent in
small-leaved trees and shrubs (Corcuera 2001), where spiders
are more abundant.

Few studies have compared differences in the abundance of
spiders on foliage of different shrub and tree species (e.g., Halaj
et al. 1998; Raizer & Amaral 2001; Souza & Martins 2004).
Although some spider species were found in small numbers (< 5
individuals), and it is not possible to reach any conclusions
about their distribution, our results reveal that foliage spider
species were positively influenced by small-leaved trees and
shrubs with a high number of leaves and branches, and
negatively by broad-leaved plants with a high foliage area
among 1 1 plant species of the Mexican tropical dry forest.

LITERATURE CITED

Arizmendi, M.C., H. Berlanga, L.M. Marquez-Valdelamar, L.
Navarijo & J.E. Ornelas. 1990. Avifauna de la Region de Chamela,
Jalisco. Cuaderno No. 4, Instituto de Biologia. UNAM, Mexico.

62 pp.

Corcuera, P. 2001. Plant use and the abundance of four bird guilds in
a Mexican dry forest-oak woodland gradient in two contrasting
seasons. Huitzil 2:2-14 (http://www.huitzil.net/ci-02pdf.pdf).
Corcuera, P. & J.E.L. Butterfield. 1999. Bird communities of dry
forest and oak woodland of western Mexico. Ibis 141:240-255.
Corcuera, P., M.L. Jimenez & G. Lopez. 2004. Comparacion de la
diversidad de aranas asociadas al follaje en una selva baja
caducifolia de Jalisco. Contactos Octubre-Diciembre: 17-26.
Corcuera, P. & J.A. Zavala-Hurtado. 2006. The influence of
vegetation on bird distribution in dry forests and oak woodlands
of western Mexico. Revista de Biologia Tropical 54:657-672.
Evans, T.A. 1997. Distribution of social crab spiders in eucalypt
forests. Australian Journal of Ecology 22:107-1 1 1.

Gunnarsson, B. 1990. Vegetation structure and the abundant and size
distribution of spruce-living spiders. Journal of Animal Ecology
59:743-752.

Gunnarsson, B. 1996. Bird predation and vegetation structure
affecting spruce-living arthropods in a temperate forest. Journal
of Animal Ecology 65:389-397.

Halaj, J.. D.W. Ross & A.R. Moldenke. 1998. Habitat structure and
prey availability as predictors of the abundance and community
organization of spiders in western Oregon forest canopies. Journal
of Arachnology 26:203-220.

Hatley, C.L. & J.A. MacMahon. 1980. Spider community organiza-
tion: seasonal variation and the role of vegetation architecture.
Environmental Entomology 9:632-639.

Heikkinen, M.W. & J.A. MacMahon. 2004. Assemblages of spiders on
models of semi-arid shrubs. Journal of Arachnology 32:313-323.
Ibarra Nunez, G., A. Garcia & M. Moreno. 1995. La comunidad de
artropodos de dos cafetales con diferentes practicas agricolas
(organico y convencional): el caso de las Aranas. Resumenes XXX
Congreso Nacional de la Sociedad Mexicana de Entomologia,
Texcoco, Edomex, Mexico. Pp. 12-13.

Ibarra Nunez, G., A. Garcia & M. Moreno. 1997. Diversidad de
aranas tejedoras (Aranae: Araneidae, Tetragnathidae, Theridiidae)

en cafetales del Soconusco, Chiapas, Mexico. Folia Entomologica
Mexicana 102:11-20.

Jimenez, M.L. 1996. Araneae. Pp. 83-101. In Biodiversidad,
Taxonomia y Biogeografia de Artropodos de Mexico: Hacia una
Sintesis de su Conocimiento. (J. Llorente, A.N. Garcia-Aldrete &
S.E. Gonzalez, eds.). Instituto de Biologia, Universidad Nacional
Autonoma de Mexico, Mexico.

Kovach, W.L. 1999. MVSP: a multivariate statistical package for
Windows, ver. 3.1. Kovach Computing Services, Pentraeth, Wales,
UK. 133 pp.

Lubin, Y.D. 1978. Seasonal abundance and diversity of web-building
spiders in relation to habitat structure on Barro Colorado Island,
Panama. Journal of Arachnology 6:31-51.

McCaffrey, J.P., M.P. Parrella & R.L. Horsburgh. 1984. Evaluation
of the limb-beating method for estimating spider (Araneae)
populations on apple trees. Journal of Arachnology 11:363-368.

Mooney. H.A., S.H. Bullock & J.R. Ehleringer. 1989. Carbon isotope
ratios of plants of a tropical dry forest in Mexico. Functional
Ecology 3:137-142.

Moreno-Molina, E.B., G.Y. Ibarra Nunez & A. Garcia Ballinas. 2001.
Diversidad de aranas en follaje de cacao en el Soconusco, Chiapas,
Mexico. Memorias del XXXII Congreso Nacional de la Sociedad
Mexicana de Entomologia. Metepec, Puebla, Mexico. 17 pp.

Murphy, P.G. & A.L. Lugo. 1986. Ecology of tropical dry forest.
Annual Review of Ecology and Systematics 17:67-88.

Pinkus-Rendon, M.A., G. Ibarra-Nunez, V. Parra-Tabla, J.A.
Garcia-Ballinas & Y. Henaut. 2006. Spider diversity in coffee
plantations with different management in Southeast Mexico.
Journal of Arachnology 34:104-112.

Raizer, J. & M.E.C. Amaral. 2001. Does the structural complexity of
aquatic macrophytes explain the diversity of associated spider
assemblages? Journal of Arachnology 29:227-237.

Riechert, S.E. & C.R. Tracy. 1975. Thermal balance and prey
availability: bases for a model relating web-site characteristics to
spider reproductive success. Ecology 56:265-285.

Robinson, J.V. 1981 . The effect of architectural variation in habitat on
a spider community: an experimental field study. Ecology 62:73-80.

Rypstra, A.L., P.E. Carter, R.A. Balfour & S.D. Marshall. 1999.
Architectural features of agricultural habitats and their impact on
the spider inhabitants. Journal of Arachnology 27:371-377.

Rzedowski, J. 1 978. La Vegetacion de Mexico. Limusa, Mexico. 432 pp.

Samu, F., K.D. Sunderland & C. Szinetar. 1999. Scale-dependent
dispersal and distribution patterns of spiders in agricultural
systems: a review. Journal of Arachnology 27:325-332.

Southwood, T.R.E. 1978. Ecological Methods (2nd edition). Chap-
man and Hall, London. 524 pp.

Souza, A.L.T. De & R.P. Martins. 2004. Distribution of plant-
dwelling spiders: inflorescences versus vegetative branches. Austral
Ecology 29:342-349.

Souza, A.L.T. De & R.P. Martins. 2005. Foliage density of branches
and distribution of plant-dwelling spiders. Biotropica 37:416-420.

Turnbull, A.L. 1973. Ecology of the true spiders (Araneomorphae).
Annual Review of Entomology 18:305-348.

Uetz, G.W., J. Halaj & A.B. Cady. 1999. Guild structure of spiders in
major crops. Journal of Arachnology 27:270-280.

Vasconcelos-Neto, J., G.Q. Romero & A.J. Santos. 2006. Association
of spiders in the Genus Peucetia (Oxyopidae) with plants bearing
glandular hairs. Biotropica 39:221-226.

Whitmore, C., R. Slotow, T.E. Crouch & A.S. Dippenaar-Schoeman.
2002. Diversity of spiders (Araneae) in a savanna reserve. Northern
Province, South Africa. Journal of Arachnology 30:344-356.

Wise, E.H. 1993. Spiders in Ecological Webs. Cambridge University
Press, Cambridge, UK. 328 pp.

Manuscript received 29 January 2005, revised 29 January 2008.

2008. The Journal of Arachnology 36:557-564

Microhabitat preferences for the errant scorpion, Centruroides vittatus

(Scorpiones, Buthidae)

C. Neal McReynolds: Department of Biology and Chemistry, Texas A&M International University, Laredo, Texas
78041, USA. E-mail: nmcreynolds@tamiu.edu

Abstract. Vegetation as a preferred microhabitat for scorpions has rarely been considered despite many Buthidae (the
bark scorpions) being non-burrowing errant scorpions that are active on both the ground and vegetation. Microhabitats
can serve multiple functions for Centruroides vittatus (Say 1821), but a particular microhabitat can be preferred for a
certain function such as a refuge, foraging, or feeding.

Observations of microhabitat use by C. vittatus were performed in Laredo, Texas of the Tamaulipan Biotic Province.
Comparisons of microhabitat use by C. vittatus at different temperatures or precipitation levels were performed. Foraging
and feeding by C. vittatus among microhabitat classes were also compared. The observed use of vegetation by C. vittatus
during different seasons was compared to the expected use based on relative abundance of vegetation in the habitat.

Air temperature, but not precipitation, had a significant effect on microhabitat use by C. vittatus. Microhabitat had a
significant effect on foraging of C. vittatus with caterpillars comprising 34.6% of the prey items and half of the scorpions
feeding on caterpillars were in blackbrush ( Acacia rigidula). The lowest proportion of scorpions observed feeding was on
the ground (3.8%) and the highest in blackbrush (40.4%). The frequency of C. vittatus among vegetation classes was
significantly different compared to the relative abundance of plant species in the plot. Scorpions were observed on prickly
pear cactus (Opuntia engelmannii) and strawberry cactus ( Echinocereus enneacanthus) at a higher frequency than expected,
and scorpions were observed on guajillo (Acacia berlandieri) and tasajillo (Opuntia leptocaulis) at a lower frequency than
expected. The frequency of scorpions on blackbrush was higher than expected during the spring.

Vegetation is an important microhabitat for C. vittatus in south Texas. The results indicate the possibility that C. vittatus
in south Texas used various plant species to carry prey captured on the ground into vegetation to feed, used blackbrush to
forage for caterpillars, and used strawberry and prickly pear cacti as a possible refuge.

Keywords: Habitat selection, foraging, refuge, feeding

Scorpions utilize a diversity of habitats (Hadley & Williams
1968; Polis 1990). Studies of habitat selection by scorpions
have compared soil types for foraging (Polis & McCormick
1986a) or to build burrows (Polis & Farley 1980; Bradley &
Brody 1984; Bradley 1988; Smith 1998). The effects of
vegetation on scorpions has been considered in association
with soil types (Bradley 1988), fire (Smith & Morton 1990), or
as refuge from predation or cannibalism (Polis 1980a)
including scorpions fleeing under vegetation to avoid preda-
tors because of low light levels (Camp & Gaffin 1999).
However, vegetation as a preferred microhabitat for scorpions
has rarely been considered despite many Buthidae (the bark
scorpions) being non-burrowing errant scorpions that are
active on both the ground and vegetation (Hadley & Williams
1968; Polis 1990).

Centruroides vittatus (Say 1821) (Scorpiones; Buthidae), the
striped bark scorpion, has a wide distribution utilizing a
number of different habitats (Shelley & Sissom 1995). Studies
of habitat use by C. vittatus have already been done in the
desert of west Texas (Brown & O’Connell 2000; Brown et al.
2002), in the deciduous forest of Arkansas (Yamashita 2004),
and in the chaparral of south Texas (McReynolds 2004). This
study will consider if C. vittatus has microhabitat preferences
that can increase the fitness of the scorpion in the chaparral of
south Texas.

Centruroides vittatus can utilize microhabitats for a refuge,
foraging, or feeding. Many buthids will use vegetation or
debris as a refuge during the day (Polis 1990). Refuges such as
burrows can be used to avoid extreme temperatures during the
day for many species of scorpions (Hadley 1974), and rocks

and cracks in the ground can serve the same function for
Centruroides sculpturatus Ewing 1928 (Crawford & Krehoff
1975) and C. vittatus (Brown et al. 2002). Vegetation could
also be a refuge from extreme conditions such as high
temperatures during the day or low temperatures at night.
One possibility is that cacti such as Texas prickly pear cactus
(Opuntia engelmannii ), tasajillo (Opuntia leptocaulis) and
strawberry cactus (Echinocereus enneacanthus) can be a refuge
from these extreme conditions because the high water content
in the cacti could provide a buffer from temperature changes
due to the high specific heat of water.

Centruroides vittatus can climb into vegetation to forage or
to feed on prey captured on the ground. Brown & O’Connell
(2000) hypothesized that C. vittatus climbs into vegetation
because of predator avoidance or higher prey availability. If
feeding scorpions carry prey up vegetation to avoid ground
predators (such as Lycosidae and Solifugae), then any
vegetation can be used as a site to feed assuming that all
plant species provides the same protection from ground
predators. In addition, feeding scorpions carrying prey up
vegetation from the ground should mainly feed on prey that is
captured on the ground (Brown & O’Connell 2000). If
scorpions are foraging in vegetation, then scorpions can be
searching for prey found only in vegetation (e.g., lepidopteran
caterpillars) and should prefer vegetation with high availabil-
ity of these prey such as blackbrush ( Acacia rigidula) and
guajillo ( Acacia berlandieri). Furthermore, scorpions can
forage in the vegetation when prey availability in vegetation
is higher such as during periods of high precipitation (see Polis
1979, 1980b).

557

558

THE JOURNAL OF ARACHNOLOGY

Microhabitats can serve multiple functions for C. vittatus ,
but a particular microhabitat can be preferred for a certain
function. This study will consider how certain conditions can
affect microhabitat use by scorpions. Microhabitat use will be
compared in relation to temperature and precipitation for
possible shifts in activity among microhabitats. Microhabitat
use will be compared in relation to prey capture and feeding to
determine if scorpions are foraging in vegetation and/or
carrying prey from the ground to vegetation. The observed use
of trees, shrubs and cacti by C. vittatus during three time
periods will be compared to a census of plant species. These
comparisons are to determine if microhabitat selection was
random or C. vittatus shows a preference for microhabitats
during any seasonal periods.

METHODS

Study animal. — Centruroides vittatus has a wide distribution
with Laredo, Texas in the southern portion of the distribution
(Shelley & Sissom 1995). Centruroides vittatus is nocturnal
with refuges during the day in debris, beneath vegetation,
under bark, and in holes in the ground, but C. vittatus and
other bark scorpions rarely dig their own burrows (Polis
1990). Scorpions emerge from their refuge only occasionally to
forage (Polis 1980b; Bradley 1988; Warburg & Polis 1990).
Different sized scorpions can be observed throughout the year
with birth of C. vittatus between April and September and age
of maturity of 36 to 48 mo (Polis & Sissom 1990). On nights of
emergence, C. vittatus is active on the ground and/or in
vegetation. Courtship by C. vittatus has rarely been observed
and females carrying first instars observed only occasionally in
the field (pers. obs.). Voucher specimens of C. vittatus were
deposited in the invertebrate collection at Texas A&M
International University.

Habitat. — This study was done on the campus of Texas
A&M International University (27°35'N, 99°26'W), Laredo,
Texas. Laredo is in the Tamaulipan Biotic Province that is
characterized by low precipitation and high average temper-
atures (Blair 1950). The habitat of the research plots can be
described as thorny brush (Blair 1950) or chaparral. Vegeta-
tion in the plots included blackbrush ( Acacia rigidula ), guajillo
( Acacia berlandieri), honey mesquite (Prosopis glandulosa),
Texas prickly pear cactus ( Opuntia engelmannii ), tasajillo
( Opuntia leptocaulis ), strawberry cactus (Echinocereus ennea-
canthus ), cenizo (Leucophyllum frutescens), guayacan ( Guaia -
cum angustifolium ), leather stem (Jatropha dioica), lotebush
( Ziziphus obtusifolia), Spanish dagger ( Yucca treculeana ), and
other species. Three research plots of the campus were studied
from 14 September 2000 to 8 August 2002 over 20 nights in
2000, 67 in 2001, 15 in 2002. The study continued from 28
August 2002 to 12 May 2005 in the main research plot over 21
nights in 2002, 52 in 2003, 46 in 2004, and 26 in 2005. At the
start, three circular sites of 100 nr were placed in each plot to
include different vegetation. Additional sites were placed at
random in the main research plot, and these sites were first
searched on 1 November 2000. The sites in the other two plots
were abandoned 8 August 2002 because of construction
nearby and light pollution from streetlights on the campus.

Data collection. -Scorpions were observed at night by
locating the scorpion fluorescing under ultraviolet light (see
Sissom et al. 1990). Observed scorpions were active and either

out of or just emerging from their refuges. No data were
collected on scorpions in their refuges to avoid destruction of
the habitat. Scorpion data were collected after sunset between
19:30 Central Standard Time, USA (CST) at the earliest and
01:00 CST at the latest for an average of two hours per night
of observation. Sites were selected at random and searched
during a night of observation with a mode of three sites
searched per night. Data were collected on all scorpions
observed within or near the site. Data collected for each
scorpion included date and time of observation, species of
scorpion, microhabitat used, if prey was captured or not, and
prey taxa. Air temperature was collected each night using a
portable weather meter, Kestrel® 3000, from 16 July 2000 to
12 May 2005. Precipitation data were radar estimates for the
field site that were provided by the Center for Earth and
Environmental Studies, Texas A&M International University
from 1 June 2003. Total precipitation for the two weeks just
prior to the sample night was used for analysis because
precipitation for the prior two weeks showed a significant
effect on the prey availability in blackbrush (unpubl. data). All
months of a year were sampled, but scorpions were rarely
active during December and January. Scorpions can be active
during all other months especially when the temperature is
above 20° C during the night. Data collection occurred during
94 nights between January-April, 66 nights between May-
August, and 115 nights between September-December.

The microhabitat data were placed in different classes:
ground, grass, blackbrush, guajillo, prickly pear cactus,
tasajillo, strawberry cactus, and other vegetation for the
comparisons in this paper. If observed on soil, leaf litter, or a
rock, the scorpion was considered on the ground. Grasses
were not identified to species, but all other plants were
identified to species if possible. Other vegetation included
small trees that are rarely taller than 2 meters with the
exception of a few mesquites and perennial shrubs such as
cenizo, guayacan, leather stem, lotebush, and Spanish dagger.
Mesquite was included in other legumes instead of other
vegetation for the comparison of prey captured in different
microhabitats. Annuals were rare in the habitat except for
ephemeral wildflowers after heavy rains and scorpions were
rarely observed climbing these wildflowers. Prey capture
classes Included no prey captured, caterpillars (Lepidoptera
larvae), other insects (including adult Lepidoptera), and IGP
(intraguild prey including Scorpiones, Araneae, Solifugae, and
Chilopoda). Prey capture by scorpions can be observed as
scorpions digest externally, thus prey items can be observed in
pedipalps or chelicerae (Polis 1979).

Census of vegetation. — A census of plant species in the main
research plot was performed on four randomly selected sites
from the 12 sites in use during the summer of 2001 and the
spring of 2002. Each circular site had all trees, shrubs and cacti
within the 100 nr area identified to species and counted
(Table 1). Grasses and ephemeral wildflowers were not
sampled. The proportions of plant species in the four sites
can be used to predict the expected frequency of scorpions on
vegetation as if there was no preference in vegetation use
(Table 1). Only scorpions on live vegetation that were
included in the census were included in the comparisons (no
scorpions on the ground, grass, ephemeral wildflowers, or
dead vegetation). The observed vegetation use of C. vittatus in

McREYNOLDS— MICROHABITAT PREFERENCES FOR CENTRUROIDES VITTATUS

559

Table 1. — The number, proportion (%), and estimated density of plant species censused from four random sites in the main research area
during the summer of 2001 and spring of 2002. Each site was a circle with an area of 100 nr.

Species

Common name

Number

Proportion

(%)

Density

(#/ha)

Acacia rigiduia

Blackbrush

51

28.3

1275

Acacia berlandieri

Guajillo

33

18.3

825

Opuntia leptocaulis

Tasajillo

33

18.3

825

Opuntia engelmannii

Prickly pear cactus

12

6.7

300

Guaiacum august [folium

Guayacan

12

6.7

300

Echinocereus enneacanthus

Strawberry cactus

5

2.8

125

Jatropha dioica

Leather stem

5

2.8

125

Prosopis glandulosa

Honey mesquite

2

1.1

50

Other vegetation

27

15.0

675

Total

180

4500

the main research plot during three time periods was
compared to the expected vegetation use. The three time
periods were based on a previous analysis of seasonal
differences in microhabitat use (McReynolds 2004).

Data analyses. — Analysis of contingency tables (Model I)
for effects on microhabitat and foraging used the G-test of
independence (Sokal & Rohlf 1995). Planned comparisons
were performed on a significant association for the contin-
gency tables to test predictions on microhabitat preferences.
The first planned comparison was among ground and
vegetation. The second planned comparison was either to test
for differences among vegetation classes that were predicted to
be used for foraging ( Acacia spp.), for refuges (cacti) and other
vegetation or to test for differences in prey capture among
legumes ( Acacia spp. and Prosopis glandulosa) and other
vegetation. Other comparisons were performed to complete
the orthogonal comparisons. The replicated goodness of fit G-
test was used to compare the observed vegetation use by C.
vittatus for three time periods to the expected vegetation use
based on the census of vegetation (Sokal & Rohlf 1995).

RESULTS

Effects on microhabitat use. — Air temperature had a
significant effect on microhabitat use (Fig. 1). In planned
comparisons, ground classes were significantly different from
pooled vegetation classes, blackbrush classes were significantly
different from guajillo classes, and grass classes were
marginally significantly different from other vegetation classes
(Fig. 1, Table 2). However, there was no significant difference
among Acacia spp., cacti, or other vegetation classes (Fig. 1,
Table 2). The proportion of scorpions on vegetation was
highest at intermediate temperature class (20-25° C) and
lowest at high temperature class (> 30° C) (Fig. 1). Precipi-
tation for the two weeks prior to observations had no
significant effect on microhabitat use (Fig. 2).

Foraging.-- Microhabitat had a significant effect on forag-
ing of C. vittatus (Fig. 3). In planned comparisons, ground
was significantly different from all vegetation, and legumes
were significantly different from other vegetation (Fig. 3,
Table 3). Only a very small proportion of scorpions on the
ground had prey compared to scorpions in vegetation. The
scorpions in the legumes had a high proportion of caterpillars
and other insects as prey while the scorpions in other
vegetation and cacti had a high proportion of intraguild prey
(IGP). The lowest proportion of scorpions observed feeding (/?

= 104) was on the ground (3.8%) and the highest in
blackbrush (40.4%). Caterpillars were 34.6% of the prey items
for C. vittatus , and half of the scorpions observed feeding on
caterpillars were in blackbrush. Intraguild prey (IGP) were
17.3% of the prey items for C. vittatus with 9.6% Araneae,
3.8% Scorpiones, 1.9% Solifugae, and 1.9% Chilopoda.

Microhabitat preferences. — The proportion of C. vittatus on
vegetation was compared to the expected proportion for three
time periods (Fig. 4). The expected proportion assumed that
scorpions have no preference for vegetation, and the
distribution of scorpions on vegetation will be random relative
to the abundance of plant species in the research plot (see
Table 1 ). The proportion of scorpions on vegetation was
significantly different from expected for all three time periods
and the pooled data, and the three time periods were
significantly heterogeneous (Fig. 4, Table 4). Scorpions were
observed on prickly pear and strawberry cacti at a higher
frequency than expected for every time period and the pooled
data, and scorpions were observed on both guajillo and
tasajillo at a lower frequency than expected for every time
period and the pooled data. However, the frequency of
scorpions on blackbrush was higher than expected during the
January-April time period but lower than expected during the
May-August time period and only slightly higher than
expected during the September-November time class. The
heterogeneity between time periods was due to fluctuations in
the frequency of scorpions in the blackbrush and other
vegetation classes.

DISCUSSION

Comparisons of scorpions have often noted the lack of
activity on vegetation (Bradley 1988; Warburg & Polis 1990).
One explanation for this pattern is that adaptation to a
specialized habitat (sand) can reduce effectiveness in climbing
(see Fet et al. 1998). However, bark scorpions (Buthidae) are
known to be active on vegetation (Polis 1990). Microhabitat
use of Buthus occitanus (Amoreux 1789) includes juveniles on
bushes but not adults '(Skutelsky 1996). In a study by Hadley
& Williams (1968), Centruroides sculpturatus pursues prey up
and down vegetation and under rocks and is more active than
the other scorpion species. Centruroides vittatus in south Texas
(present study) uses vegetation at higher frequency than in
west Texas (Brown & O’Connell 2000) and in Arkansas
(Yamashita 2004). Vegetation is important microhabitat for
C. vittatus in Laredo, Texas with 54.1% on trees, shrubs or

560

THE JOURNAL OF ARACHNOLOGY

Figure 1 . The number of Centruroides vittatus using different microhabitats among temperature classes. The two Opimtia spp. classes, O.
leptocaulis (tasajillo) and O. englemanii (prickly pear cactus), were pooled for the statistical analysis. The frequency of microhabitat use was
significantly different among temperature classes (G = 74.79, P < 0.001, df = 18, n = 1252). See Table 2 for planned comparisons among
microhabitat classes.

cacti; 7.8% on grass and only 38.1% on the ground. In other
studies, C. vittatus were observed in trees (9.1%) and grass
(10.6%) in Arkansas (Yamashita 2004) and 26.4% climbing
vegetation in west Texas (Brown & O’Connell 2000).

Scorpion activity and thus microhabitat use can shift
because of environmental factors (i.e., temperature and/or
precipitation). The environmental factors can have a direct
effect on the scorpion activity or indirectly on prey availability
(Polis 1980a. 1988 but see Bradley 1988). Microhabitat use of
C. vittatus shifted to the ground at high nocturnal tempera-
tures. However, there was no support for the prediction that
microhabitat use would shift from refuges in cacti to foraging
in Acacia spp. with differences in temperature although the use
of blackbrush was high during intermediate temperatures
relative to guajillio. The high activity of scorpions on the

Table 2.- Planned comparisons among microhabitat classes of the
contingency table for microhabitat vs. temperature classes. The two
Opimtia spp. classes, O. leptocaulis (Tasajillo) and O. englemanii
(Prickly pear cactus), were pooled for the statistical analysis. NS =
not significant. See Fig. 1.

Planned comparisons

G

df

P

Ground vs. All vegetation

43.71

3

< 0.001

Cactus vs. Acacia spp. vs. Grass and
Other vegetation

8.71

6

NS

Opuntia spp. vs. Strawberry cactus

3.77

3

NS

Blackbrush vs. Guajillo

11.33

3

< 0.05

Grass vs. Other vegetation

7.27

3

0.1-0.05

Total

74.79

18

< 0.001

ground at high temperatures (> 30° C) during the night does
fit the pattern of high activity of C. vittatus on the ground
during July and August as previously reported (McReynolds
2004). This can indicate low prey availability in vegetation and
relatively higher prey availability on the ground during the
hottest period of the year. Microhabitat use of C. vittatus did
not shift with precipitation, and there was no evidence that
foraging in blackbrush increased with high precipitation as
predicted because of the observed increase in caterpillar
availability with high precipitation (unpubl. data). One
possible reason that activity and foraging behavior does not
change with precipitation (and prey availability) is the threat
of predation including cannibalism by adults on juveniles (see
Polis 1980a, 1980b). Only adult Paruroctonus mesaensis (now
Smeringus mesaensis (Stahnke 1957)) have a significant
positive correlation with prey availability while the other age
classes have a positive but not significant correlation (Polis
1980b).

Scorpions can utilize different microhabitats and in
particular different vegetation for feeding, foraging or refuge.
Scorpion species can feed where the prey was captured, can
carry prey to burrow (or other refuge) before feeding, or can
carry it into vegetation. For example, Parabuthus pallidus
Pocock 1895 carries prey back to the burrow but Parabuthus
leiosoma (Ehrenberg 1828) feeds where prey is captured (Rein
2003). Scorpions with prey on vegetation are usually
attributed to scorpions carrying prey from the ground into
vegetation to feed (Polis 1979; Brown & O’Connell 2000).
Only 3.8% of the scorpions with prey were on the ground in
south Texas but many prey of C. vittatus were usually

McREYNOLDS— MICROHABITAT PREFERENCES FOR CENTRUROIDES VITTATUS

561

220 -r
200-
180
160

140

I120

O

100

<D

X

| 80

z

60

40

20-
0 -

Strawberry cactus
Tasajillo

Prickly pear cactus
Guajillo
Blackbrush
Grass

Other vegetation
Ground

0-1

1-5

>5

Precipitation (cm)

Figure 2. — The number of Centruroid.es vittatus using different microhabitats among classes of total precipitation for the prior two weeks. The
frequency of microhabitat use was not significantly different among precipitation classes (G = 7.37, not significant, df — 14, n = 985).

Figure 3. — The number of Centruroides vittatus using different microhabitats among prey capture classes. The frequency of microhabitat use
was significantly different among prey capture classes (G = 112.17, P < 0.001, df = 12, n = 1890). See Table 3 for planned comparisons among
microhabitat classes.

562

THE JOURNAL OF ARACHNOLOGY

Table 3. — Planned comparisons among microhabitat classes of the
contingency table for microhabitat vs. prey capture classes. NS = not
significant. See Fig. 3.

Planned comparisons

G

df

P

Ground vs. All vegetation

83.77

3

< 0.001

Legumes vs. Cactus and Other vegetation

14.92

3

< 0.01

Blackbrush vs. Other legumes

6.53

3

NS

Cactus vs. Other vegetation

6.95

3

NS

Total

112.17

12

< 0.001

observed on the ground (e.g., many of the intraguild prey)
(pers. obs.). Intraguild prey (including cannibalism) were
17.3% of the prey in south Texas as compared to 27.91% of
spider prey and 9.30% cannibalism for C. vittatus in Arkansas
(Yamashita 2004). These data indicate that C. vittatus in south
Texas do carry prey captured on the ground into vegetation to
feed. One possible function of moving prey (including
intraguild prey) to vegetation is to avoid intraguild predation
(see Bradley & Brody 1984; Polis & McCormick 1986b, 1987).
Another possible function can be to avoid scavenging ants.
Ants were observed causing feeding scorpions to move (pers.
obs. by E. Lopez and C.N. McReynolds). Scorpions can also
move between vegetation. A scorpion was observed feeding on
a caterpillar on tasajillo but near a blackbrush where the
caterpillars are available (pers. obs.). Perhaps this is to avoid
ants or predators (e.g., other scorpions).

Foraging in vegetation has not been considered important
for scorpions (Polis 1979) except for errant scorpions
(McCormick & Polis 1990) and by juveniles (e.g., juvenile
Buthus occitanus ambush prey in vegetation but not adults
(Skutelsky 1996)). There was no evidence of foraging in
vegetation by C. vittatus in west Texas (Brown & O’Connell
2000). However, the diet of C. vittatus in Laredo, Texas
includes a number of items that have a high availability in
trees or shrubs. Caterpillars were an important prey item for
the scorpions in south Texas and most of the caterpillar taxa
captured by C. vittatus were available in blackbrush (unpubl.
data). Caterpillars are rarely reported as an important item in
the diet for scorpions (see McCormick & Polis 1990).
Caterpillars were only 1.4% of the diet for P. mesaensis (Polis
1979), and no caterpillars were reported for C. vittatus in west
Texas (Brown & O’Connell 2000) but 1 1 .6% of the prey for C.
vittatus in Arkansas (Yamashita 2004) and 34.6% in south
Texas (present study). It is predicted that scorpions foraging in
blackbrush (and other legumes) will increase when caterpillar
availability increases. However, scorpions prefer blackbrush
only in January-April and there is no shift to blackbrush with
high precipitation. If caterpillar availability is higher with the
blooming of blackbrush in March and April, then this can
explain why there is not an overall preference for blackbrush
but there is a higher proportion of prey captured during
March and April (McReynolds 2004) and the proportion of
scorpions on blackbrush is higher than expected during the
January-April time period.

January-April May-August September-November Pooled Expected

Figure 4. The proportion (%) of Centruroides vittatus on vegetation during each seasonal time period compared to the expected proportion
(%) on vegetation. The observed frequency of scorpions on vegetation from January to April (n = 174), from May to August (/; = 166), from
September to November (/? = 415) and for the pooled data {n = 755) were compared to the expected frequency based on relative abundance of
vegetation (see Table I) in the replicated goodness of fit test (see Table 4).

McREYNOLDS— MICROHABITAT PREFERENCES FOR CENTRUROIDES VITTATUS

563

Table 4. — Replicated goodness of fit test for microhabitat classes
comparing the observed frequency of scorpions on vegetation to the
expected frequency based on relative abundance of vegetation (see
Table 1) during three seasonal time periods. See Fig. 4.

Time Periods

G

df

P

January to April

113.74

5

< 0.001

May to August

130.43

5

< 0.001

September to November

293.61

5

< 0.001

Pooled

473.47

5

< 0.001

Heterogeneity

64.31

10

< 0.001

Total

537.77

15

< 0.001

Most scorpion species dig a burrow, but many buthids do
not dig their own burrow but use holes, space beneath rocks,
and openings under bark and below vegetation as diurnal
refuges (Polis 1990). Centruroides vittatus in west Texas have
patchy distribution under rocks as diurnal refuges (Brown et
al. 2002). Scorpions have been observed entering cracks or
holes in the ground in south Texas (pers. obs.), but rocks are
not available in the main research plot of this study. An
alternative refuge for C. vittatus in south Texas can be a cactus
because the high water content can provide a buffer from
temperature changes due to the high specific heat of water.
Scorpions have been observed going under the pads of a
prickly pear cactus or down the openings between the stems of
a strawberry cactus (pers. obs.). This can explain the higher
than expected frequency of C. vittatus on strawberry cactus
and prickly pear cactus in the research plot despite the low
frequency of scorpions feeding on cacti and the low
probability of foraging success because of the low prey
availability on cacti (pers. obs.). This is assuming that C.
vittatus is a central place forager (Orians & Pearson 1979) and
there is a high probability that the scorpion would be near its
refuge after emerging for the night (see Polis et al. 1985).
However, C. vittatus has a lower than expected frequency in
tasajillo. Perhaps tasajillo with thin stems and the more
treelike structure does not provide the refuge that the prickly
pear cactus or the strawberry cactus can provide. Sampling of
potential refuges during the day at different temperatures will
be required to establish diurnal refuge preferences of C.
vittatus in south Texas (see Brown et al. 2002).

The high frequency of scorpions with prey in vegetation and
low frequency on the ground indicates that C. vittatus carry
prey caught on the ground into vegetation to feed. Foraging
by C. vittatus especially in blackbrush for caterpillars has been
demonstrated. However, the prediction that foraging scorpi-
ons will show a preference for blackbrush and other legumes
was not supported. The frequency of scorpions in blackbrush
did not increase with higher precipitation as is predicted
because high precipitation increases availability of caterpillars
in blackbrush. The frequency of scorpions in blackbrush was
higher than expected in comparison to relative abundance of
plant species only in January through April. It is possible that
prey availability is higher early in the year especially March
and April, but there is no evidence yet supporting this
prediction. The results show that C. vittatus in south Texas
used strawberry and prickly pear cacti with a higher than
expected frequency in all time periods. However, the low
frequency of scorpions with prey on cacti suggests that this

preference for cacti was not for feeding or foraging, and
perhaps scorpions were utilizing the cacti for another function
such as a diurnal refuge.

ACKNOWLEDGMENTS

I thank my undergraduate research students from Fail 2000
to Spring 2005 for assistance in the field to collect the data for
the paper. They included Elaine Boteilo, Ezequiel Chapa,
Yareth Coutino, Ruperto Contreras III, Luis Ricardo de
Leon, Paula Flores, Nicolas Gallegos, Jose Hernandez,
Juvenal Herrera IV, Mike Herrera HI, Eduardo Lopez, Ruben
Mendez, Erica Morales, Rocio Moya, Metzli Ortega, Julianna
Quintanilla, Oscar Ramos, Amede Rubio, and Israel Salinas.
In particular, I thank Metzli Ortega for the census of plant
species and Eduardo Lopez for his observations on the feeding
behavior of scorpions and their interactions with ants. Thanks
to Gene McReynolds for his modifications of the ultraviolet
lights and to Michael Daniel for a discussion on the potential
of cactus as a refuge. I also thank Dan Mott, Julianna
Quintanilla, Oscar Ramos, Soren Toft, Tsunemi Yamashita,
Fernando Quintana, and two anonymous reviewers for
suggestions to improve the paper immeasurably. Financial
support was provided by grants from Texas A&M Interna-
tional University. The scorpion research at Texas A&M
International University was started at the suggestion of the
late Gary Polis.

LITERATURE CITED

Blair, W.F. 1950. The biotic provinces of Texas. Texas Journal of
Science 2:93-1 17.

Bradley, R.A. 1988. The influence of weather and biotic factors on the
behavior of the scorpion (. Paruroctonus utahensis). Journal of
Animal Ecology 57:533-551.

Bradley, R.A. & A.J. Brody. 1984. Relative abundance of three
vaejovid scorpions across a habitat gradient. Journal of Arachnol-
ogy 11:437-440.

Brown, C.A., J.M. Davis, D.J. O’Connell & D.R. Formanowiz, Jr.
2002. Surface density and nocturnal activity in a west Texas
assemblage of scorpions. Southwestern Naturalist 47:409-419.
Brown, C.A. & D.J. O’Connell. 2000. Plant climbing behavior in the
scorpion Centruroides vittatus. American Midland Naturalist
144:406-418.

Camp, E.A. & D.D. Gaffrn. 1999. Escape behavior mediated by
negative photoaxis in the scorpion Paruroctonus utahensis. Journal
of Arachnology 27:679-684.

Crawford, C.S. & R.C. Krehoff. 1975. Die! activity in sympatric
populations of the scorpions Centruroides scidpturatus (Buthidae)
and Diplocentrus spitzeri (Diplocentridae). journal of Arachnology
2:195-204.

Fet, V., G.A. Polis & W.D. Sissom. 1998. Life in sandy deserts: the
scorpion model. Journal of Arid Environments 39:609-622.
Hadley, N.F. 1974. Adaptationai biology of desert scorpions. Journal
of Arachnology 2:11-23.

Hadley, N.F. & S.C. Williams. 1968. Surface activities of some North
American scorpions in relation to feeding. Ecology 49:726-734.
McCormick, S.J. & G.A. Polis. 1990. Prey, predators, and parasites.
Pp. 294-320. In The Biology of Scorpions. (G.A. Polis, ed.).
Stanford University Press, Stanford, California.

McReynolds, C.N. 2004. Temporal patterns in microhabitat use for
the scorpion Centruroides vittatus (Scorpiones, Buthidae). Euscor-
pius 17:35-45.

Orians, G.H. & N.E. Pearson. 1979. On the theory of central place
foraging. Pp. 155-177. In Analysis of Ecological Systems. (D.J.

564

THE JOURNAL OF ARACHNOLOGY

Horn, R. Mitchell & G.R. Stair, eds.). Ohio State University Press,
Columbus, Ohio.

Polis, G.A. 1979. Prey and feeding phenology of the desert sand
scorpion Paruroctonus mesaensis (Scorpionidae: Vaejovidae).
Journal of Zoology (London) 188:333-346.

Polis, G.A. 1980a. The effect of cannibalism on the demography and
activity of a natural population of desert scorpions. Behavioral
Ecology and Sociobiology 7:23-35.

Polis, G.A. 1980b. Seasonal patterns and age-specific variation in the
surface activity of a population of desert scorpions in relation to
environmental factors. Journal of Animal Ecology 49:1-18.

Polis, G.A. 1988. Foraging and evolutionary responses of desert
scorpions to harsh environmental periods of food stress. Journal of
Arid Environments 14:123-134.

Polis, G.A. 1990. Ecology. Pp. 247-293. In The Biology of Scorpions.
(G.A. Polis, ed.). Stanford University Press, Stanford, California.

Polis, G.A. & R.D. Farley. 1980. Population biology of desert
scorpion: survivorship, microhabitat, and the evolution of life
history strategy. Ecology 61:620-629.

Polis, G.A. & S.J. McCormick. 1986a. Patterns of resource use and
age structure among species of desert scorpion. Journal of Animal
Ecology 55:59-73.

Polis, G.A. & S.J. McCormick. 1986b. Scorpions, spiders and
solpugids: predation and competition among distantly related
taxa. Oecologia 71:111-116.

Polis, G.A. & S.J. McCormick. 1987. Intraguild predation and
competition among desert scorpions. Ecology 68:332-343.

Polis, G.A., C.N. McReynolds & R.G. Ford. 1985. Home range
geometry of the desert scorpion Paruroctonus mesaensis. Oecologia
67:273-277.

Polis, G.A. & W.D. Sissom. 1990. Life history. Pp. 161-223. In The
Biology of Scorpions. (G.A. Polis, ed.). Stanford University Press,
Stanford, California.

Rein, J.O. 2003. Prey capture behavior in the East African scorpions
Parabuthus leiosoma (Ehrenberg, 1828) and P. pallidus Pocock,
1895 (Scorpiones: Buthidae). Euscorpius 6:1-8.

Shelley, R.M. & W.D. Sissom. 1995. Distributions of the scorpions
Centruroides vittatus (Say) and Centruroides hentzi (Banks) in the
United States and Mexico (Scorpiones, Buthidae). Journal of
Arachnology 23:100-1 10.

Sissom, W.D., G.A. Polis & D.D. Watt. 1990. Field and laboratory
methods. Pp. 445-461. In The Biology of Scorpions. (G.A. Polis,
ed.). Stanford University Press, Stanford, California.

Skutelsky, O. 1996. Predation risk and state-dependent foraging in
scorpions: effects of moonlight on foraging in the scorpion Buthus
occitanus. Animal Behaviour 52:49-57.

Smith, G.T. 1998. Density of the burrowing scorpion Urodacus
armatus (Scorpiones; Scorpionidae) in relation to vegetation types:
implications for population decline following agricultural clearing.
Pacific Conservation Biology 4:209-214.

Smith, G.T. & S.R. Morton. 1990. Responses by scorpions to fire-
initiated succession in arid Australian Spinifex grasslands. Journal
of Arachnology 18:241-244.

Sokal, R.R. & F.J. Rohlf. 1995. Biometry, 3rd edition. W.A. Freeman
and Company, New York. 887 pp.

Warburg, M.R. & G.A. Polis. 1990. Behavioral responses, rhy-
thms, and activity patterns. Pp. 224—246. In The Biology of
Scorpions. (G.A. Polis, ed.). Stanford University Press, Stanford,
California.

Yamashita, T. 2004. Surface activity, biomass, and phenology of the
striped scorpion, Centruroides vittatus (Buthidae) in Arkansas,
USA. Euscorpius 17:25-33.

Manuscript received 29 January 2007, revised 20 January 2008.

2008. The Journal of Arachnology 36:565-573

Observations on phenology and overwintering of spiders associated with apple and pear orchards in

south-central Washington

Eugene R. Miiiezky, David R. Horton and Carrol O, Calkins: Yakima Agricultural Research Laboratory, USDA-ARS,
5230 Konnowac Pass Road, Wapato, Washington 98951, USA. E-mail: Gene.Msliczky@ars.usda.gov

Abstract. Beating tray and sweep net samples from apple and pear orchards in south-central Washington State were used
to obtain information on life history and phenology of orchard-associated spiders. Cardboard shelters placed in the
orchards in the fall and collected during the winter yielded information on spider overwintering. Data were obtained for 43
species in 28 genera and 12 families. The eight most abundant species were Pelegrina aeneola (Curtis 1892), Meioneta
filhnorana (Chamberlin 1919), Oxyopes scalaris Hentz 1845, Theridion neomexicanum Banks 1901, Misumenops lepidus
(Thorell 1877), Xysticus cunctator Thorell 1877, Philodromus cespitum (Walckenaer 1802), and Sassacus papenhoei Peck ham
& Peckham 1895. Each was represented by more than 690 specimens. Salticidae, Philodromidae, and Linyphiidae were
represented by the largest number of species. Most species appear to have univoltine life cycles in the study area. Species
matured at different times during the season between spring and fall. Twenty-seven species utilized cardboard shelters for
overwintering, but some common spiders failed to do so and apparently use alternative locations. Some species
overwintered in a broad range of developmental stages, whereas other species overwintered in only one or two instars.

Keywords: Araneae, life history, seasonality

Spiders have been of interest as predators of arthropod pests
in orchards for at least 50 years. Chant (1956) found
differences between the spider faunas of insecticide treated
compared to untreated orchards and listed the natural (i.e.,
non-orchard) habitats in which many species occurred. Don-
dale (1956, 1958) recorded 77 species from apple orchards in
Nova Scotia, Canada, provided data on abundance, and
observed spiders feeding on apple pests. Similar studies were
conducted over the next 35 years: Dondale et al. (1979) and
Bostanian et al. (1984) in Canada; Specht & Dondale (1960),
Legner & Oatman (1964), and McCaffrey & Horsburgh (1980)
in the United States; Mansour et ah (1980a) in Israel; Dondale
(1966) in Australia; and Hukusima (1961) in Japan. The
efficacy of limb tapping for estimating spider populations in
apple orchards was investigated by McCaffrey et al. (1984).

Interest in the spider fauna of apple orchards has continued
in recent years as pest management programs that rely less
heavily on broad-spectrum insecticides are employed with
greater frequency, improving the prospects for significant
impact by natural enemies on pest control. Studies have
examined the tree canopy fauna (Olszak et al. 1992;
Wisniewska & Prokopy 1997; Brown et ah 2003; Pekar &
Kocourek 2004), the soil surface fauna (Bogya & Marko 1999;
Pekar 1999a; Epstein et al. 2000), and the fauna of the
herbaceous layer (Pekar 1999b; Bogya et al. 2000; Miiiezky et
al. 2000). Overwintering spiders (Bogya et ah 2000; Pekar
1999c; Horton et al. 2001) and the bark dwelling fauna (Bogya
et ah 1999b) have also received attention. Regional geographic
effects were found to be most important in determining the
composition of spider assemblages in orchards in different
areas (Bogya et ah 1999a).

The phenologies of numerous spider taxa have been
documented by field and laboratory studies (Merrett 1967,
1968; Aiken & Coyle 2000; Stiles & Coyle 2001). Annual,
biennial, and intermediate length life cycles have been
described (Toft 1976, 1978; Wise 1984), and in some cases a
species’ life cycle has been found to vary within its range.

probably influenced by local climatic conditions (Dondale
1961; Putman 1967). Schaefer (1977) distinguished five types
of life cycle among spiders from a north temperate region,
studied their adaptive strategies for surviving the winter, and
noted the importance of the cold season in synchronizing a
species’ life cycle.

Here we report on spiders that are found in apple and pear
orchards in Yakima County, Washington, an important fruit
growing area of the state. Life cycles of several common,
orchard-associated species are described, and seasonal distri-
bution data for adults and penultimate stage males of less
common species are presented. Information on species that
utilized artificial shelters for overwintering is also given.
Information of this kind may be useful in assessing the
potential contribution of spiders to orchard pest control and
in scheduling pesticide applications to minimize adverse
impact on spiders.

METHODS

Study orchards. — Sampling was conducted in 42 orchards
(apple and pear), all of which were located in Yakima County,
Washington. All orchards were within a radius of 46 km of the
city of Yakima. Some study orchards have been removed from
production since completion of the study. The following are
latitudes and longitudes for 8 orchards located at the periphery of
the study area: 46.4529°N, 120.2292°W; 46.5029°N, 1 20.1 667° W;
46.5835°N, 120.3501°W; 46.5185°N, 1 20.4235 AV; 46.7432°N,
120.7738°W; 46.6618°N, 120.7557°W; 46.47 15°N, 120.3834 W;
46.3106°N, 120.1236°W. All other orchards fell within the area
demarcated by the peripheral orchards. Insect pest management
in the study orchards ranged from conventional programs based
on synthetic, broad-spectrum insecticides (e.g., azinphos-methyl)
to state-certified organic programs in which use of synthetic
insecticides is prohibited. The codling moth Cydia pomonella
(Linnaeus 1758) is the key pest of apple and pear in this region.

The study was conducted from 1996 to 2001, but not all
orchards were sampled each year. Nine apple orchards were

565

566

THE JOURNAL OF ARACHNOLOGY

sampled in 1996. The following year three pear orchards were
sampled in addition to the nine apple blocks. The maximum
number of orchards sampled in one year was 19 in 1999.
Orchard size was 0.5 to 32 ha.

Sampling. — Arboreal spiders were sampled with a beating
tray, 0.45 nr in area (Bioquip Products, Gardena, CA). One
limb 1-2 m above ground on each of 25 trees (15 trees in 1999)
was struck three times with a heavy rubber hose to dislodge
spiders. Most specimens were promptly preserved in 70%
isopropyl alcohol, but selected specimens were saved and
reared (see below). Trees in all parts of an orchard were
sampled while walking a winding path. The sampling period
usually included April to October. Samples were collected
every 1-2 weeks in 1996-1998 and 2000 and monthly in 1999
and 2001. All specimens taken during this study are held at the
Yakima Agricultural Research Laboratory (USDA-ARS) in
Wapato, Washington, USA.

Sweep net sampling (net diameter = 38 cm) of the
understory vegetation was done in 1996 and 1997 in the same
orchards that were monitored with beat trays. An 180° swing
of the net constituted a sweep, and 25 sweeps per sample were
taken while walking a winding path so as to sample in all parts
of the orchard. The sampling periods were late June to late
October 1996 and mid-May to mid-October 1997. Samples
were taken every 1-2 weeks with longer intervals after an
orchard was mowed.

Spiders were also collected by hand on an irregular basis
when chanced upon or during an occasional more serious
search. Generally, one or a few specimens of interest were
collected. Immatures were reared to obtain a positive
identification if necessary. These data were used to supplement
beat tray and sweep net data.

Cardboard shelters of two types were used to collect
overwintering spiders. The first consisted of a bundle of ten,
12.5 cm X 17.5 cm sheets of cardboard (flute size ~ 4 mm X
5 mm) tied to the lowest crotch of a tree. The second was a
7.6 cm wide strip of cardboard wrapped once around the
trunk of a tree 0. 5-1.0 m above ground. Shelters were set out
in September and October, retrieved in December or January,
and stored in a cold room until processed. The number of
shelters set out and the number of orchards sampled varied
from year to year.

Incidental to a 1998 study of apple bin use by overwintering
codling moth larvae, spiders that had also overwintered in the
bins were collected. Bins were removed from the field after
harvest, held in a cold room at 0.6°-1.7° C until late January,
and then placed in a greenhouse at 2 1 °-24° C. Spiders were
collected as they emerged from the bins.

Sample processing. —Selected specimens were reared for
positive identification. They were held in 35 ml plastic cups
and provided water and prey of appropriate size weekly. Field
captured Lygus sp. (Hemiptera: Miridae) and laboratory
reared Drosophila sp. (Diptera: Drosophilidae) were readily
consumed by most species. Once a familiarity with the local
fauna was acquired it was possible to identify a majority of
immatures to species. Specimens of definite developmental
stage (e.g., penultimate female, antepenultimate male, etc.)
mentioned in Results in all cases refer to reared individuals.

Immatures were sorted into small, medium, and large size
classes on a species-by-species basis. Since many spiders pass

through five to seven nymphal instars, this roughly corre-
sponds to instars one, two, and three for small, four and five
for medium, and six and seven for large immatures. Although
somewhat arbitrary, this allowed an estimate of the age
distribution of immatures through the season. Penultimate
males were readily distinguished by their enlarged pedipalps.
Theridion , Erigone, and Meioneta, because of their small size,
were sorted into a single class of immatures, in addition to
penultimate males and adults; antepenultimate male Erigone
and Meioneta were distinguishable based on a slight enlarge-
ment of the pedipalps.

Data presentation. — Data from beat tray and sweep net
collections were combined to obtain monthly totals. These
data are presented graphically for eight abundant species to
show the proportion of different developmental stages in each
month’s collection. Note that a small number of beat tray
samples (17) were taken during the first week of November.
These data were pooled with the extensive October data for
the graphs. Few specimens were taken in November because
leaf fall was well underway and spiders had begun a general
movement out of the trees. Information for less abundant
species is summarized in the Tables which also include data
from hand collections when available. Overwintering data
from all years were combined.

RESULTS

Salticidae. — Adults of both sexes of Pelegrina aeneola
(Curtis 1892) were abundant in April, May, and June
(Fig. 1). Thereafter males were rare although females were
present through October. Females with egg sacs were most
common during May and June, but an egg sac containing
undispersed first instar nymphs was collected as late as 8
September 2000. A female collected with her egg sac on 3 June
1997 produced a second clutch of eggs in the laboratory. Small
immatures were present in all months but comprised over 85%
of the population in July. Penultimate males and large
nymphs, which included penultimate females, were most
abundant in October. Large immatures and penultimate males
were the principal overwintering stages although small and
medium-sized immatures were well represented (Table 1).
Penultimate males were uncommon in the trees during April,
but five adults hand-collected in the litter in early April 1997
may indicate that many penultimate males undergo their final
molt in this location and then move up into the trees. Pelegrina
aeneola appears to have an annual life cycle in the study area.

The phenology of Sassacus papenhoei Peckham & Peckham
1895 was similar to that of P. aeneola but lagged about a
month behind (Fig. 1). Adults were most abundant during
June, and small immatures dominated the population during
August. The few individuals found in overwintering refuges
were small and medium-sized immatures (Table 1) as were the
few specimens taken in beating tray samples during April. An
annual life cycle in the study area is indicated for S. papenhoei.

Four species of large jumping spiders in the genus Phidippus
occurred in the orchards. Collection data, summarized in
Table 2, indicates annual life cycles for all four but with
different maturation times. The first two or three nymphal
instars were very similar in appearance and could not reliably
be sorted to species. Phidippus comatus Peckham & Peckham
1901 matured during the summer. A female guarding her egg

MILICZKY ET AL.— ORCHARD SPIDERS

567

Pelegrina aeneola

Xysticus cunctator

100%

204 210 1341 1646 926 1036

Sassacus papenhoei

37 279 199 105

Misumenops lepidus

14 553 340 135

m

■w;

Oxyopes scalaris

Philodromus cespitum

F ESM ^sM □ Lg « Med QSm

F SM ^sM QLg SMed OSm

Figure I. - -Percentage of adults and different sized immatures of six common spider species found by month in combined beating tray and
sweep net collections. F = female; M = male; sM = penultimate male; Lg = large immature; Med = medium sized immature; Sm = small
immature. Number at the top of the column is the total number of specimens taken for the month.

sac was found among a group of ripening pears on 19 August
1998, and a female with a hatched egg mass was found in a
codling moth pheromone trap on 29 October 1996. Three
small immatures collected on 26 September 2003 yielded a
recognizable P. comatus after two molts, a male after five
molts, and a penultimate female after six molts. Phidippus
clarus Keyserling 1885 also appears to mature during the
summer (Table 2). Phidippus audax (Hentz 1845) appears to
mature in the spring. Specimens from overwintering shelters
(Table 1) were mostly medium and large immatures and. of
the 42 individuals that overwintered in apple bins, 26 were
penultimate females and 12 were penultimate males. Phidippus
johnsoni (Peckham & Peckham 1883) also appears to mature in
the spring. Overwintering and seasonal collection data for
other Salticidae is summarized in Tables 1 and 2 respectively.

Oxyopidae. — Oxyopes scalaris Hentz 1845 was infrequently
collected from March to June, but its numbers increased

substantially in July and August. Medium-sized immatures
made up 50% of the population in September and 73% in
October, but large immatures were present in only small
numbers both months. Miscellaneous collections included 2
females, 2 penultimate females, and 1 antepenultimate female
on 26 May 2000; 1 female and 1 penultimate female on 30 May
2000; and 1 penultimate male on 1 June 2000. Oxyopes scalaris
appears to mature in the spring and early summer and to be
univoltine in the study area.

Thomisidae. — Immature Xysticus cunctator Thorell 1877
were commonly taken in the beat tray samples. Adults,
however, were rare in the trees (one female out of 609
specimens) but were more frequently swept from the
understory vegetation (five males and four females out of
288 specimens). Adults were most numerous in May according
to sweep net and beat tray collections, but hand collected
adults of both sexes were taken in April. New generation

568

THE JOURNAL OF ARACHNOLOGY

Table 1. — Summary of spider overwintering stages: sm., med., Ig. = small, medium, and large immatures (large immatures included
antepenultimates of both sexes and penultimate females); sM = penultimate males; M = adult male; F = adult female. Additional observations:
one med. P. cespitum molted three times to an adult female; the sm. P. insperatus molted four times to a male and seven of the med.’s molted
three times each yielding five males and two females; one sM P. alascensis was reared to the adult stage.

Spider species

sm.

Overwintering stage

med. Ig. sM

M F

Salticidae

Pelegrina aeneola (Curtis 1892)

39

87

227

192

1 1

Sassacus papenhoei Peckham & Peckham 1 895

5

5

Sassacus vitis (Cockerell 1894)

4

3

1

Phidippus audax (Hentz 1845)

1

10

12

4

Phanias watonus (Chamberlin & Ivie 1941)

21

49

23

4

10 28

Salticus scenicus (Clerck 1757)

5

24

34

46

1

Oxyopidae

Oxyopes scalaris Hentz 1 845

2

Thomisidae

Xysticus cunctator Thorell 1877

1

Misumenops lepidus (Thorell 1877)

29

48

5

34

Philodromidae

Philodromus cespitum (Walckenaer 1802)

310

83

10

1

Philodromus insperatus Schick 1965

1

9

Philodromus californicus Keyserling 1884

9

129

1

Philodromus rufus Walckenaer 1826

25

269

7

Philodromus speciosus Gertsch 1 934

2

10

1

3 5

Philodromus alascensis Keyserling 1884

1

2

Tibellus oblongus (Walckenaer 1802)

27

16

4

Ebo pepinensis Gertsch 1933

3

3

2

Clubionidae

Cheiracanthium mildei L. Koch 1864

104

210

37

40

Cheiracanthium inclusum Hentz 1847

1

1

Corinnfdae

Phrurotimpus borealis Emerton 1911

14

203

140

Anyphaenidae

Anyphaena pacifica (Banks 1896)

2

6

6

Theridlidae

Steatoda hespera Chamberlin & Ivie 1933

15

31

4

1

3 2

Linyphiidae

Meioneta fillmorana (Chamberlin 1919)

1

Erigone spp.

1

Pityohyphantes minidoka Chamberlin & Ivie 1943

11

19

3

1

1

Tetragnathidae

Tetragnatha laboriosa Hentz 1850

6

1

1

Dictynidae

Dictyna coloradensis Chamberlin 1919

35

21

1

spiderlings appeared in May and remained abundant through

a few weeks later than in X. cunctator

. Penultimate males first

August (Fig. 1). Antepenultimate males, as indicated by

appeared in August

and

were abundant in

September,

rearing, were present as early as July. Penultimate females

October, and the following April

Penultimate

males and

(reared individuals) and penultimate males (Fig. 1) were
present by September. Larger immatures, including penulti-
mates of both sexes, would therefore be the primary
overwintering stages. This species appears to have an annual
life cycle in the study area.

The phenology of Misumenops lepidus (Thorell 1877) was
similar to that of X. cunctator although new generation
spiderlings did not appear in large numbers until July (Fig. 1),

other immature stages overwintered (Table 1). Misumenops
lepidus appeared to have an annual life cycle.

Philodromidae. — The phenology of Philodromus cespitum
(Walckenaer 1802) (Fig. 1) was similar to that of X. cunctator
and M. lepidus but was shifted later into the season. Adult
females were most common in June and July, and small
spiderlings, which could be found in all months, were most
abundant from August to October. Five females with egg sacs

MILICZKY ET AL.— ORCHARD SPIDERS

569

Table 2. — Summary by month of beating tray, sweep net, and hand collections of less common spider species: numbers of penultimate males,
males, and females, respectively, are given. Additional observations: Penultimate female E. militaris were collected in April. May, September, and
October; two antepenultimate female X. gulosus collected in July, one female collected in December; one penultimate female P. insperatus
collected in June, one female collected in June laid three egg clutches in the lab; one antepenultimate female P. californicus collected in April, one
antepenultimate female and 26 other large immatures collected in October; three antepenultimate male A. trifasciata collected in July and one in
August, 2 penultimate females collected in August.

Spider species

April

May

June

July

Aug

Sept

Oct

Salticidae

Eris militaris (Hentz 1845)

0-2-0

2-2-11

0-0-4

1-0-4

3-0-1

0-1-0

1-3-0

Phanias watonus (Chamberlin & Ivie 1941)

0-0-2

Phidippus audax (Hentz 1845)

0-1-0

0-0-1

0-1-1

0-0-1

10-0-1

Phidippus darns Keyserling 1885

1-7-11

0-1-1

0-0-3

0-0-1

Phidippus comat us Peckham & Peckham 1901

1-0-0

9-1-1

1-11-8

0-4-4

0-1-0

0-0-1

Phidippus johnsoni Peckham & Peckham 1883

Sasscicus vitis (Cockerell 1894)

Salticus scenicus (Clerck 1757)

0-3-2

1-0-0

0-1-1

1-0-0

0-2-0

1-0-0

2-1-0

Thomisidae

Xysticus gulosus Keyserling 1880

1-0-0

0-0-1

0-1-3

Philodromidae

Philodromus insperatus Schick 1965

Philodromus californicus Keyserling 1884

2-0-0

0-2-2

0-0-4

3-0-0

Philodromus rufus Walckenaer 1826

0-2-0

Tibellus oblongus (Walckenaer 1802)

0-1-2

0-0-1

1-9-9

1-1-7

0-0-1

Tibellus asiaticus Kulczyn’ski 1908

0-0-1

Clubionidae

Cheiraccmthium inclusion Hentz 1847

2-1-0

0-0-1

Corinnidae

Castianeira longipalpa Hentz 1847

1-0-0

Anyphaenidae

Anyphaena pacifica (Banks 1896)

Linvphiidae

0-0-1

0-0-1

3-0-0

Spirembolus mundus Chamberlin & Ivie 1933

7-0-0

14-0-0

1 1-0-0

1-1-0

0-1-0

Tenuiphantes tenuis (Black wall 1852)

0-0-1

0-2-2

0-0-1

0-0-2

0-1-1

0-1-0

Collinsia ksenius (Crosby & Bishop 1928)

0-0-1

0-5-3

0-3-1

0-1-3

Walckenaeria subspiralis Millidge 1983

Pityohyphantes minicloka Chamberlin & Ivie 1943

0-1-0

0-5-0

0-1-0

0-1-0

0-2-4

Tetragnathidae

Tetragnatha laboriosa Hentz 1850

5-4-2

0-2-2

4-14-11

1-0-3

1-0-2

Tetragnatha versicolor Walckenaer 1841

Araneidae

0-2-0

0-1-1

Argiope trifasciata (Forsskal 1775)

1-0-0

3-1-0

0-0-2

were collected in July, and a female with an egg sac containing

summarized in

Table 3. All

sizes of C.

mildei

immatures

undispersed spiderlings was found in September. Small and
medium sized immatures most commonly overwintered
(Table 1), and an annual life cycle in the study area is
indicated.

Tibellus oblongus (Walckenaer 1802) occurred in many
orchards. While primarily an inhabitant of the understory
vegetation (212 of 256 specimens), individuals were occasion-
ally found in the trees. It overwintered as immatures of various
sizes (Table 1) and is probably univoltine in the study area.
Overwintering and seasonal collection data for less common
species in the Philodromidae are given in Tables 1 and 2
respectively.

Clubionidae. — Cheiraccmthium mildei L. Koch 1864 oc-
curred in many study orchards and appears to have an annual
life cycle in the study area. Collection data for the species are

overwintered including penultimate males and females (Ta-
ble 1). Cheiraccmthium inclusion Hentz 1847 was collected in
only one of our study orchards but appears to have a
phenology similar to that of C. mildei. Small immatures were
most abundant in July (15 specimens) and August (20),
medium-sized immatures were most abundant in September
(19) and October (17), and large immatures were most
abundant in October (13).

Theridiidae. — Theridion neomexicanum Banks 1901 was a
common, primarily arboreal species that is clearly univoltine
in the study area (Fig. 2). Males were first noted in May, but
both sexes were most abundant in June and July. A male and a
penultimate female were found together in a web on an apple
leaf on 1 1 June 1998, females with egg sacs were collected on
apple leaves on 10 July 1998 and 27 July 1999, and a female

570

THE JOURNAL OF ARACHNOLOGY

Table 3. — Seasonal occurrence of Cheiracanthium mildei based on
combined beat tray, sweep net, and hand collections. Spider stages:
small, medium, large = small, medium, and large immatures (large
immatures included antepenultimate nymphs of both sexes and
penultimate females); sM = penultimate males. Additional observa-
tions: two of the July and one of the September females were guarding
egg sacs.

Month of
collection

Developmental stage

small

medium

large

sM Male

Female

May

1

1 1

June

1

1

1

July

28

5

3

1

3

August

29

28

2

September

18

29

14

5

3

October

19

44

23

10

with a vacated egg sac was found on 6 August 1999.
Penultimate males were not observed by October, and
overwintering must therefore occur as immatures smaller than
penultimates in undetermined locations.

Linyphiidae. — Meioneta fillmorana (Chamberlin 1919) is
also primarily an arboreal spider. Its numbers peaked in
May, after which there was a four month decline and then a
marked rebound in October. Such a late season increase was
unusual for spiders in this study especially since the October
population of M. fillmorana consisted entirely of adults
(Fig. 2). Six years of beat tray samples all showed a similar
pattern, however. The species appears to be univoltine.
Although we obtained virtually no overwintering data for
M. fillmorana , it seems reasonable to infer, given the
preponderance of females in the October collections, that
overwintering probably takes place in the egg stage.

At least two species of Erigone occurred in the orchards:
Erigone dentosa O. Pickard-Canrbridge 1894 and Erigone
aletris Crosby & Bishop 1928. We were unable to separate the
two species with certainty. Adults of both sexes were present
from at least May to October and were most abundant in
October and June (Table 4). Although Erigone were well
represented in the trees, sweep samples for 1996 and 1997
yielded a majority (65%) of the specimens that were collected.
Erigone is also common on the ground based on pitfall trap
collections (Miliczky et al. 2000), and thus has a broad
distribution among habitats within the orchard. Tables 1 and
2 have data for less common Linyphiidae.

Other families. — Data for less commonly collected spiders
in other families is summarized in Tables 1 and 2. Included in
this group: Anyphaena pacifica (Banks 1896) (Anyphaenidae)
was occasionally found in the trees and immatures utilized
overwintering shelters; Tetragnatha laboriosa Hentz 1850
(Tetragnathidae) was commonly swept from understory
vegetation, and immatures were taken with some regularity
in late season beat trays; Dictyna coloradensis Chamberlin
1919 (Dictynidae) was common only at the USDA research
farm where it constructed webs on apple and pear leaves and
tall weeds in adjacent uncultivated ground.

DISCUSSION

Many temperate zone spiders have a single generation per
year (Gertsch 1979; Foelix 1996), and this appeared to be true

of Washington species for which sufficient data were obtained.
Within the broad latitudinal range of the temperate zone,
however, factors that may influence spider development vary
widely, and other life history patterns have been documented.
About half of the 52 species studied in Denmark by Toft
(1976, 1978) were biennial, and Almquist (1969) found a
similar proportion of biennial species among 20 studied in
Sweden. Almquist (1969) also observed that in Sweden the life
cycles of some species were twice as long as the life cycles of
the same or related species in southwestern Europe. Similarly,
Philodromus cespitum is biennial in Nova Scotia (Dondale
1961) but univoltine farther to the south in Ontario (Putman
1967).

In North America and Denmark maturation times among
univoltine spiders form a continuum from the spring to the fall
(Toft 1976; Gertsch 1979). Some litter inhabiting species of
Linyphiidae mature and reproduce even during the winter
months (Duffey 1956; Schaefer 1977). Washington species for
which we acquired sufficient data appeared to have well-
defined periods of reproduction (stenochrony) and could be
classified as stenochronous with reproduction in spring and
summer (Schaefer 1977).

We noted that at a given time during the season, and also in
the overwintering shelters, some species were represented by
several developmental stages, whereas others were represented
by only a few. Small, medium, and large immature Pelegrina
aeneola could be found during much of the season (Fig. 1) and
all these stages also overwintered (Table 1). Cheiracanthium
mildei showed a similarly broad range of overwintering stages,
while Phanias watonus was even more extreme as adults of
both sexes also commonly overwintered (Table 1). In contrast,
small immature Philodromus cespitum dominated the orchard
collections from August to October (Fig. 1) and was the
principal overwintering stage (Table 1). Philodromus calif ami-
cus and Dictyna coloradensis spent the winter primarily as
large immatures and/or penultimate males and penultimate
females (Table 1). Factors tending to increase the length of
time during the season when a given developmental stage is
present include long-lived females that remain with an egg sac
until the young disperse and produce more than one dutch of
eggs. Egg sac guarding by P. aeneola was observed in the field,
egg sacs were found as late as September, and females are
capable of producing a second clutch of eggs. Pelegrina
galatea (Walckenaer 1837) also guarded its eggs and produced
multiple clutches in the laboratory (Horner & Starks 1972).
We observed female C. mildei guarding eggs in the field and
multiple clutches of eggs were produced under laboratory
conditions (Mansour et al. 1980b).

A number of the more common orchard spiders were poorly
or not at all represented in overwintering shelters and
presumably seek alternative sites inside or outside the orchard.
Pekar (1999c) noted a similar phenomenon. Oxyopes scalaris,
the only member of the family that occurs in Washington
(Crawford 1988), presented an interesting case. Large numbers
of small immature O. scalaris appeared in the orchards rather
abruptly in July, and size increase in the population was
observed as the season progressed. Large immatures and
adults were rare in orchard collections (Fig. 1), however, and
only two immatures were found in overwintering shelters
(Table 1). The fate of the medium sized immatures that are so

MILICZKY ET AL.— ORCHARD SPIDERS

571

Theridion neomexicanum Meioneta fillmorana

Figure 2. — Percentage of adults and immatures of two common spider species found by month in combined beating tray and sweep net
collections. F = female; M = male; sM = penultimate male; ssM = antepenultimate male (M. fillmorana only); 1mm - all other immature stages.
Number at the top of the column is the total number of specimens taken for the month.

common in the orchards in October remains to be determined.
The opposite situation was noted in Steatoda hespera which
utilized overwintering shelters but was not taken by beating
tray or sweep net. Like most members of the genus, S. hespera
webs are probably situated in rock and bark crevices and in
cavities near the ground (Levi 1957) and would not have been
sampled. Horton et al. (2001) changed a portion of their
cardboard bands on pear and apple trees weekly from August
to December. They determined that several species utilized the
bands as temporary refuges during the autumn but overwin-
tered elsewhere.

Despite their abundance in terrestrial habitats and their
exclusively predatory habits, Debach & Rosen (1991) noted a
general neglect of spiders as potential biological control agents
and attributed this, in part, to their generalist predatory
habits. Other authors, noting the diversity of prey capture
strategies and microhabitat exploitation patterns of spiders,
have emphasized the contribution of the spider community as
a whole to insect control in agroecosystems (Reichert &
Lockley 1984; Marc & Canard 1997). Interest in the
composition of spider faunas in orchards began over 50 years
ago and appears to have increased given the number of studies
conducted in recent years. However, studies attempting to
evaluate the importance of spider predation on orchard pests
are few (e.g., MacLellan 1973; Mansour et al. 1980a; Amalin
et al. 2001; Miliczky & Calkins 2002).

Table 4. — Summary by month of combined beating tray and sweep
net collections of Erigone spp. Spider stages: ssM = antepenultimate
male; sM = penultimate male.

Month of

Developmental stage

collection

ssM

sM

Male

Female

April

4

1

May

40

28

4

6

June

19

59

71

53

July

49

68

23

14

August

24

35

23

14

September

43

50

34

22

October

2

32

192

197

During this study we observed most of the common orchard
spiders feeding on pests. Pelegrina aeneola , O. scalaris, P.
cespitum , X. cunctator , and M. lepidus all used a variety of
smaller pest species as prey, including leafhoppers, leafminers,
aphids, thrips, and mites. The webs of M. fillmorana snared
aphids and thrips as well as tiny flies and parasitoid wasps.
The large salticid Phidippus clams took prey up to the size of
an adult earwig. Some of the orchard spiders or a close relative
may be important predators in agroecosystems more general-
ly. Pelegrina aeneola and other members of the genus may be
important in biological control because they are often
abundant and are known to feed on pest insects (Horner
1972; Jennings & Houseweart 1978; Mason & Paul 1988).
Oxyopes salticus Hentz 1845, a close relative of 0. scalaris, is a
dominant predator in row crops in the United States and an
important predator of pest insects (Young & Lockley 1985).
Cheiracanthium mildei was described by Wise (1993) as a
potentially important biological control agent in a number of
agroecosystems. Mansour et al. (1980a) determined that C.
mildei was the most effective spider predator of a lepidopteran
pest of apples in Israel. Miliczky & Calkins (2002) rated it as
having the greatest potential as a predator of pest leafrollers in
Washington orchards out of 1 1 species tested.

The role of spiders in orchard pest control is of considerable
interest given the current trend toward reduced use of broad
spectrum insecticides, the large numbers of spiders often
observed when pesticide use is decreased or eliminated, and
the great diversity among orchard-inhabiting spiders in size,
behavior, and prey-capture strategies. All of these factors
suggest that spiders should have substantial potential for
contributing to orchard pest control. Future studies should
further document the importance of this interesting but often
overlooked group of beneficial organisms in controlling pest
species in orchards and other agricultural systems.

ACKNOWLEDGMENTS

Funding for this research was provided by an Initiative for
Future Agriculture and Food Systems (IFAFS) grant and by
grants from the Washington Tree Fruit Research Commis-
sion. Rod Crawford (The Burke Museum, University of
Washington, Seattle), Andrew Moldenke (Oregon State
University, Corvallis), and G.B. Edwards (Florida State

572

THE JOURNAL OF ARACHNOLOGY

Collection of Arthropods, Gainesville) assisted with spider
identifications. Rod Crawford, Lerry Lacey (USDA - Yakima
Agricultural Research Laboratory), and two anonymous
reviewers read earlier drafts of this paper and made helpful
comments for its improvement. Merilee Bayer assisted with
orchard sampling and with sorting and rearing of spiders. We
are especially grateful to the many Yakima County apple and
pear growers who allowed ready access to their orchards for
sampling and observation.

LITERATURE CITED

Aiken, M. & F.A. Coyle. 2000. Habitat distribution, life history and
behavior of Tetragnatha spider species in the Great Smoky
Mountains National Park. Journal of Arachnology 28:97-106.
Almquist, S. 1969. Seasonal growth of some dune-living spiders.
Oikos 20:392-408.

Amalin, D.M., J.E. Pena & R. McSorley. 2001. Predation by hunting
spiders on citrus leafminer, Phyllocnistis citrella Stainton (Lepi-
doptera: Gracillariidae). Journal of Entomological Science
36:199-207.

Bogya, S. & V. Marko. 1999. Effect of pest management systems on
ground-dwelling spider assemblages in an apple orchard in
Hungary. Agriculture, Ecosystems & Environment 73:7-18.

Bogya, S., V. Marko & Cs. Szinetar. 1999a. Comparison of pome
fruit orchard inhabiting spider assemblages at different geograph-
ical scales. Agricultural and Forest Entomology 1:261-269.

Bogya, S., V. Marko & Cs. Szinetar. 2000. Effect of pest management
systems on foliage- and grass-dwelling spider communities in an
apple orchard in Hungary. International Journal of Pest Manage-
ment 46:241-250.

Bogya, S., Cs. Szinetar & V. Marko. 1999b. Species composition of
spider (Araneae) assemblages in apple and pear orchards in the
Carpathian Basin. Acta Phytopathologica et Entomologica Hun-
garica 34:99- 121.

Bostanian, N.J., C.D. Dondale, M.R. Binns & D. Pitre. 1984. Effects
of pesticide use on spiders (Araneae) in Quebec apple orchards.
Canadian Entomologist 116:663-675.

Brown, M.W., J.J. Schmitt & B.J. Abraham. 2003. Seasonal and
diurnal dynamics of spiders (Araneae) in West Virginia orchards
and the effect of orchard management on spider communities.
Environmental Entomology 32:830-839.

Chant, D.A. 1956. Predacious spiders in orchards in south-eastern
England. Journal of Horticultural Science 31:35—46.

Crawford, R.L. 1988. An annotated checklist of the spiders of
Washington. Burke Museum Contributions in Anthropology and
Natural History No. 5. 48 pp.

Debach, P. & D. Rosen. 1991. Biological Control by Natural
Enemies. Second edition. Cambridge University Press, Cambridge,
UK. 456 pp.

Dondale, C.D. 1956. Annotated list of spiders (Araneae) from apple
trees in Nova Scotia. Canadian Entomologist 88:697-700.
Dondale, C.D. 1958. Note on population densities of spiders
(Araneae) in Nova Scotia apple orchards. Canadian Entomologist
90:111-113.

Dondale, C.D. 1961. Life histories of some common spiders from
trees and shrubs in Nova Scotia. Canadian Journal of Zoology
39:777-787.

Dondale, C.D. 1966. The spider fauna (Araneida) of deciduous
orchards in the Australian Capital Territory. Australian Journal of
Zoology 14:1157-1192.

Dondale, C.D., B. Parent & D. Pitre. 1979. A 6-year study of spiders
in a Quebec apple orchard. Canadian Entomologist 111:377-
380.

Duffey, E. 1956. Aerial dispersal in a known spider population.
Journal of Animal Ecology 25:85-111.

Epstein, D.L., R.S. Zack, J.F. Brunner, L. Gut & J.J. Brown. 2000.
Effects of broad-spectrum insecticides on epigeal arthropod
biodiversity in Pacific Northwest apple orchards. Environmental
Entomology 29:340-348.

Foelix, R.F. 1996. Biology of Spiders. Second edition. Oxford
University Press, New York. 330 pp.

Gertsch, W.J. 1979. American Spiders. Second edition. Van Nostrand
Reinhold Company, Toronto. 274 pp.

Horner, N.V. 1972. Metaphidippus galathea as a possible biological
control agent. Journal of the Kansas Entomological Society
45:324-327.

Horner, N.V. & K.J. Starks. 1972. Bionomics of the jumping spider

Metphidippus galathea. Annals of the Entomological Society of
America 65:602-607.

Horton, D.R., E.R. Miliczky. D.A. Broers, R.R. Lewis & C.O.
Calkins. 2001. Numbers, diversity, and phenology of spiders
(Araneae) overwintering in cardboard bands placed in pear and
apple orchards of central Washington. Annals of the Entomolog-
ical Society of America 94:405-414.

Hukusima, S. 1961. Studies on the insect association in crop field.
XXI. Notes on spiders in apple orchards. Japanese Journal of
Applied Entomology and Zoology 5:270-272.

Jennings, D.T. & M.W. Houseweart. 1978. Spider preys on spruce
budworm egg mass. Entomological News 89:183-186.

Legner, E.F. & E.R. Oatman. 1964. Spiders on apple in Wisconsin
and their abundance in a natural and two artificial environments.
Canadian Entomologist 96:1202-1207.

Levi, H.W. 1957. The spider genera Crustulina and Steatoda in North
America, Central America, and the West Indies (Araneae,
Theridiidae). Bulletin of the Museum of Comparative Zoology
1 17:365-424.

MacLellan, C.R. 1973. Natural enemies of the light brown apple
moth, Epiphyas postvittana, in the Australian Capital Territory.
Canadian Entomologist 105:681-700.

Mansour, F., D. Rosen & A. Shulov. 1980a. A survey of spider
populations (Araneae) in sprayed and unsprayed apple orchards in
Israel and their ability to feed on larvae of Spodoptera littoralis
(Boisd.). Acta Oecologica/Oecologica Applicata 1:189-197.

Mansour, F., D. Rosen & A. Shulov. 1980b. Biology of the spider
Chiracanthium mildei (Arachnida: Clubionidae). Entomophaga
25:237-248.

Marc, P. & A. Canard. 1997. Maintaining spider biodiversity in
agroecosystems as a tool in pest control. Agriculture, Ecosystems
& Environment 62:229-235.

Mason, R.R. & H.G. Paul. 1988. Predation on larvae of Douglas-fir
tussock moth. Orgyia pseudotsugata (Lepidoptera: Lymantriidae),
by Metaphidippus aeneolus (Araneae: Saliticidae). Pan-Pacific
Entomologist 64:258-260.

McCaffrey, J.P. & R.L. Horsburgh. 1980. The spider fauna of apple
trees in central Virginia. Environmental Entomology 9:247-252.

McCaffrey, J.P., M.P. Parrella & R.L. Horsburgh. 1984. Evaluation
of the limb-beating sampling method for estimating spider
(Araneae) populations on apple trees. Journal of Arachnology
11:363-368.

Merrett, P. 1967. The phenology of spiders on heathland in
Dorset I. Families Atypidae, Dysderidae, Gnaphosidae, Clubioni-
dae, Thomisidae and Salticidae. Journal of Animal Ecology 36:
363-374.

Merrett, P. 1968. The phenology of spiders on heathland in Dorset.
Families Lycosidae, Pisauridae, Agelenidae, Mimetidae, Theridi-
idae, Tetragnathidae, Argiopidae. Journal of Zoology, London
156:239-256.

Miliczky, E.R. & C.O. Calkins. 2002. Spiders (Araneae) as potential
predators of leafroller larvae and egg masses (Lepidoptera:
Tortricidae) in central Washington apple and pear orchards.
Pan-Pacific Entomologist 78:140-150.

MILICZKY ET AL. ORCHARD SPIDERS

573

Miliczky, E.R., C.O. Calkins & D.R. Horton. 2000. Spider
abundance and diversity in apple orchards under three insect pest
management programmes in Washington State, U.S.A. Agricul-
tural and Forest Entomology 2:203-215.

Olszak, R.W., J. Luczak, E. Niemczyk & R.Z. Zajac. 1992. The spider
community associated with apple trees under different pressure of
pesticides. Ekologia Polska 40:265-286.

Pekar, S. 1999a. Side-effect of integrated pest management and
conventional spraying on the composition of epigeic spiders and
harvestmen in an apple orchard (Araneae, Opiliones). Journal of
Applied Entomology 123:115-120.

Pekar, S. 1999b. Effect of IPM practices and conventional spraying
on spider population dynamics in an apple orchard. Agriculture,
Ecosystems & Environment 73:155-166.

Pekar, S. 1999c. Some observations on overwintering spiders
(Araneae) in two contrasting orchards in the Czech Republic.
Agriculture, Ecosystems & Environment 73:205-210.

Pekar, S. & F. Kocourek. 2004. Spiders (Araneae) in the biological
and integrated pest management of apple in the Czech Republic,
journal of Applied Entomology 128:561-566.

Putman, W.L. 1967. Life histories and habits of two species of
Philodromus (Araneida: Thonrisidae) in Ontario. Canadian Ento-
mologist 99:622-631.

Reichert, S.E. & T. Lockley. 1984. Spiders as biological control
agents. Annual Review of Entomology 29:299-320.

Schaefer, M. 1977. Winter ecology of spiders (Araneida). Zeitschrift
fur angewandte Entomologie 83:1 13-134.

Specht, H.B. & C.D. Dondale. I960. Spider populations in New
Jersey apple orchards. Journal of Economic Entomology
53:810-814.

Stiles, G.J. & F.A. Coyle. 2001. Habitat distribution and life history
of species in the spider genera Theridion , Rugalliodes, and Want ha
in the Great Smoky Mountain National Park (Araneae, Theridi-
idae). Journal of Arachnology 29:396-412.

Toft, S. 1976. Life-histories of spiders in a Danish beech wood.
Natura Jutiandica 19:5-40.

Toft, S. 1978. Phenology of some Danish beech-wood spiders. Natura
Jutiandica 20:285-304.

Wise, D.H. 1984. Phenology and life history of the filmy dome spider
(Araneae: Linyphiidae) in two local Maryland populations. Psyche
91:267-288.

Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge University
Press, Cambridge, UK. 328 pp.

Wisniewska, J. & R.J. Prokopy. 1997. Pesticide effect on faunal
composition, abundance, and body length of spiders (Araneae) in
apple orchards. Environmental Entomology 26:763-776.

Young, O.P. & T.C. Lockley. 1985. The striped lynx spider, Oxyopes
salticus (Araneae: Oxyopidae), in agroecosystems. Entomophaga
30:329-346.

Manuscript received 9 May 2007, revised 21 February 2008.

2008. The Journal of Arachnology 36:574—582

Distribution of Geraeocormobius sylvarum (Opiliones, Gonyleptidae):

Range modeling based on bioclimatic variables

Luis E. Acosta: CONICET - Catedra de Diversidad Animal I, Facultad de Ciencias Exactas, Fisicas y Naturales,
Universidad Nacional de Cordoba, Av. Velez Sarsfield 299, X5000JJC Cordoba, Argentina. E-mail:
lacosta@com.uncor.edu

Abstract. The potential distribution of the harvestman Geraeocormobius sylvarum Holmberg 1887 (Opiliones,
Gonyleptidae, Gonyleptinae) from Argentina, Brazil and Paraguay, is modeled using the presence-only, GIS-based
method Bioclim. The model was run on 2.5 min resolution climate layers using 19 derived bioclimatic variables. The
bioclimatic profile of the species is described, and presumable limiting factors in each part of the range are discussed.
Modeled distribution of G. sylvarum shows a remarkable correspondence to the Alto Parana Atlantic forest ecoregion, with
a marginal presence around the Araucaria forests and in gallery/flood forests towards the Southwest. Results support the
650 km yungas-Mesopotamia disjunction, as previously proposed, and reveal that localities in northwestern Argentina
have extreme values concerning seasonality parameters with remarkably decreased rainfall in winter. Evidence suggesting
that the disjunct pattern may have been derived by antropic introduction is briefly discussed.

Keywords: Neotropics, bioclim, potential distribution, ecological niche modeling, environmental envelope

Harvestmen (Arachnida, Opiliones) are generally regarded
as a well suited taxon for biogeographic studies (Ringuelet
1959; Giribet & Kury 2007). Two key features make them
useful for those purposes: their low vagility and a close
dependence on environmental conditions, mainly humidity
(Acosta 2002; Pinto-da-Rocha et al. 2005; Machado et al.
2007). Distribution of most species is thus dependent on the
geographic continuity of suitable environments (Acosta 2002).
In addition, harvestman endemicity may be particularly
remarkable in some areas (Machado et al. 2007). Small-area
endemicity is striking in forested ranges like the Brazilian
Serra do Mar (Pinto-da-Rocha et al. 2005), and is also
important in the montane forests of northwestern Argentina,
the so called “yungas” region (Acosta 2002).

However, small-ranged endemics should not be seen as the
rule for harvestman distribution. Under certain conditions,
many species are known to spread over thousands of knr
(Curtis & Machado 2007) as long as the suitable environment is
not restricted by any geographic or ecological barrier. For
example, while many central European species appear geo-
graphically restricted, no less than 30 harvestmen in that region
extend over quite large ranges (Martens 1978). Broad ranged
species are also typical for most of the Argentinean Mesopo-
tamia, which is the humid and sub-humid region between the
Parana and Uruguay rivers (Acosta 2002). Its northernmost
portion, roughly matching the administrative province of
Misiones. bears subtropical forest physiognomy - actually a
part of the “Paranense Biogeographic Province” (Cabrera &
Willink 1973) that covers adjacent areas in Paraguay and Brazil
as well. The rest of Mesopotamia is a mosaic of shrubs, swamps,
grasslands and gallery forests (Hueck & Seibert 1972; Cabrera
& Willink 1973). In accordance with these differences, the
Mesopotamian opiliofauna has been split into two different,
though overlapping, sub-areas (Acosta 2002): the Misiones sub-
area and the Mesopotamian sensu strieto sub-area. Towards the
West, as precipitation decreases, these humid and sub-humid
realms give way to the semiarid Chaco, an effective distribu-
tional limit for Mesopotamian harvestmen (Acosta 2002).

Geraeocormobius sylvarum Holmberg 1887 (Opiliones,
Gonyleptidae, Gonyleptinae) is a large and conspicuous
harvestman known to inhabit subtropical forests in north-
eastern Argentina, southern Brazil and southeastern Paraguay
(Ringuelet 1959; Kury 2003; Acosta et al. 2007), being thus
characteristic of the Misiones sub-area (Acosta 2002). Acosta
et al. (2007) provided several new records for this species;
among them three localities in the province of Tucuman,
Argentina, revealing a disjunct presence in montane forests in
northwestern Argentina (NWA). These separate populations
lie about 650 km away from the westernmost record in the
core area, with the sub-xeric Chaco in between. As stressed by
Acosta et al. (2007), most records of G. sylvarum concentrate
in the Alto Parana Atlantic forests ecoregion (referred to as
“Paranense forests” below) and in some adjacent sectors of
the Araucaria moist forest ecoregions (nomenclature after
Olson et al. 2001). Argentinean captures outside the men-
tioned ecoregions (in provinces of Corrientes and Chaco)
seemingly are associated with gallery forests and/or seasonal
inundation sites (the Humid Chaco and Southern Cone
Mesopotamian savanna ecoregions of Olson et al. (2001)).
In turn, findings of G. sylvarum in NWA correspond to the
“tucumano-boliviano” forests, or “yungas” (Hueck & Seibert
1972; Acosta 2002), or the Southern Andean Yungas
ecoregion (Olson et al. 2001). The available localities show
some spatial bias, since they are concentrated in and around
the province of Misiones (a traditionally well sampled area;
Ringuelet 1959), while extensive areas in Paraguay are left
almost undocumented (cf. Fig. 3).

Interestingly, two further Mesopotamian harvestmen, Dis-
cocyrtus dilatatus Sorensen 1884 and D. prospicuus (Holmberg
1876) (Gonyleptidae, Pachylinae) have Mesopotamia-yungas
disjunct ranges as well (Acosta 1995, 2002). A very basic
question remains unanswered, however: Are Chacoan condi-
tions really inhospitable to Mesopotamian species or is our
record incomplete, leading us to wrongly assume this region to
be hostile for harvestmen? Nothing is known about the
climatic tolerances of D. dilatatus, D. prospicuus or G.

574

ACOSTA— DISTRIBUTION MODELING OF GERAEOCORMOBIUS (OPILIONES)

575

sylvarum as very little is known about harvestmen physiology
and tolerance to physical factors in general (Santos 2007). It is
generally accepted that harvestman distribution is governed by
climatic constraints, mainly humidity and temperature, but
papers addressing how actual climatic conditions affect a
given species are almost lacking (Curtis & Machado 2007;
Machado et al. 2007).

The occurrence records contain some useful clues. Point
records are the very basis for discovering a species range,
though the vast majority of Neotropical harvestmen have been
scarcely recorded, so that species with more than 15 records
are rare (cf. Kury 2003). In fact, most distributional patterns
are then intuitively extrapolated from the few records
available, emphasizing the need for gathering more data and
filling in the gaps. Alternatively, recent developments aimed to
model potential species ranges, using available records and
several types of environmental predictors, offer innovative
methods to help detect areas where presence of a given species
may be expected though still not documented (Guisan &
Zimmermann 2000; Guisan & Thuiller 2005; Hernandez et al.
2006; Peterson 2006). Using a Geographic Information System
(GIS) and appropriate climate information, the envelope
approach infers the species’ bioclimatic profile, a species-
specific feature (actually, a subset of the fundamental
ecological niche) that gives us a rough insight into their
presumed tolerance limits, at least with respect to the variables
used to build the model. It is then possible to project this
profile onto the geographic space, to identify all areas with
similar climatic conditions, i.e., where the species would meet a
potentially suitable environment (Guisan & Zimmermann
2000; Hijmans & Graham 2006; Hernandez et al. 2006; Pearce
& Boyce 2006). This paper is intended to characterize the
bioclimatic profile of G. sylvarum and to model its potential
distribution using the bioclim algorithm, taking advantage of
the number of records now available for this species (close to
the minimum sample size for the method to attain an
acceptable accuracy; Hernandez et al. 2006). Thereby it is
also intended to test if the disjunct Mesopotamian-yungas
pattern is supported by the model and to identify areas in
which future sampling efforts would be productive in filling in
distribution gaps. Results are matched to major biomes and
habitat types in order to get a more comprehensive
understanding, though still preliminary, of the ecological
requirements and potential range of this wide-ranged harvest-
man.

METHODS

Data acquisition. -All published references for G. sylvarum
were taken into account (Holmberg 1887; Ringuelet 1959;
Soares & Soares 1985; Kury 2003; Acosta et al. 2007).
Historical records were inspected for taxonomic reliability (cf.
Acosta et al. 2007 for localities excluded), while three localities
from the province of Misiones (published by Ringuelet 1959)
remained unrecognizable and were set aside: “Campamento
Yacu-Poi, near Puerto Bemberg”; “60 km Puerto Iguazu”
(which direction?), and “Pasarela Rio Uruguay.” The full
dataset consisted of 48 unique point localities; during the
analysis, duplicate records from the same gridcell were
removed by the software, resulting in 46 effective records.
Localities were identified and geo-referenced using printed

road maps and digital gazetteers (mainly NGA GEOnet
Names Server [GNS], United States Board on Geographic
Names; Google Earth ®), all cross-checked for final accuracy.
The label data or collector’s information was used to
determine location as precisely as possible. Published records
typically refer to a locality, which may represent not more
than the nearest reference to the actual collecting site; such an
imprecision in coordinates (impossible to measure) is deemed
not to affect the results considering both the regional scale and
coarse approach used.

Climate layers. — The model was run on Worldclim 1.4.
(Hijmans et al. 2005a), a set of global climate layers,
containing extrapolated monthly data for the 1950-2000
period on maximum, minimum and mean temperature, and
precipitation. The 2.5 min resolution (i.e., approximately 4.5 x
4.5 km gridcell) was selected. Information contained in
Worldclim (climate data, together with a digital elevation
model) is used by the software to derive the 19 bioclimatic
variables available for modeling, listed in Table 1. The
abbreviation “be” followed by a number is used below to
identify each of these bioclimatic variables.

Modeling method. -Modeling was performed through the
presence-only method Bioclim, as implemented in Diva-Gis
5.4 (Hijmans et al. 2005b). Bioclim is a frequency distribution-
based algorithm, which calculates the envelope that bounds
the bioclimatic preferences of the species (Fischer et al. 2001;
Walther et al. 2004; Hernandez et al. 2006). Values of each
derived bioclimatic variable are extracted from all localities
and arranged in a cumulative frequency distribution (cf.
Fig. 2). The envelope is defined as a multi-dimensional hyper-
box (each variable representing a dimension) delimiting, at a
given percentile, the climatic conditions in the occurrence
localities (Guisan & Zimmermann 2000). This set of values
constitutes the bioclimatic profile of the species (Fischer et al.
2001), as summarized for the target species in Table 2. In the
potential distribution maps (Figs. 1, 4-5), gridcells are scored
as suitable (if within the envelope; i.e., the presence of the
species can be expected) or unsuitable (outside the envelope).
User-defined percentiles were set to define the extent of the
envelope (as cut-off) or to rank the gridcells suitability
(Fischer et al. 2001; Hijmans et al. 2005b; Hernandez et al.
2006).

Evaluation. -The accuracy of the predictive range generated
by the model was assessed by calculating the AUC (area under
curve) in a receiver operating characteristic (ROC) plot, and
the maximum Kappa (rnax-K) value, both analyses made in
Diva-Gis (Hijmans et al. 2005b). In this study, 70% of the
original points were randomly resampled as training data to
perform 20 repetitions of the model. Test data included
pseudo-absence points selected at random from the back-
ground. ROC/max-K were calculated against the grids stack
(0-100 percentile) of the 20 range polygons that were modeled
using the training points. AUC values over 0.8 are deemed to
reflect a “good” model performance; above 0.9 the accuracy is
considered “high” (Luoto et al. 2005). In turn, max-K over 0.4
are deemed to be “good” and “excellent” if above 0.75
(Randin et al. 2006).

Input variants. — A separate run was performed with records
from NWA removed to verify if the species is still predicted in
that area. To identify the factors limiting the distribution, the

576

THE JOURNAL OF ARACHNOLOGY

Table 1. — Median, minimum and maximum values, and range for all 19 biocliinatic variables in the envelope of Geraeocormobius sylvarum.
Absolute temperature values are in degrees Celsius (° C), precipitation in mm. Numbers preceding each variable are referred to in the text and
Table 2. Main differences between profiles with and without yungas records are emphasized in bold.

Biocliinatic variables

All records (n = 46)

Yungas removed (n = 44)

Median

Min-max (range)

Median

Min-max (range)

( 1 ) Annual mean temperature

20.39

16.46-23.12 (6.66)

20.62

16.46-23.12 (6.66)

(2) Mean monthly T° range

12.43

10.65-13.83 (3.18)

12.43

10.65-13.83 (3.18)

(3) Isothermality (2/7 x 100)

56.08

47.25-65.00 (17.75)

56.21

49.48-65.00 (15.52)

(4) T seasonality (STD x 100)

368.87

252.92-461.59 (208.67)

366.90

252.92-430.76 (177.84)

(5) Max T° of warmest month

31.65

27.20-34.20 (7.00)

31.70

27.20-34.20 (7.00)

(6) Min T° of coldest month

9.25

4.80-12.50 (7.70)

9.40

4.80-12.50 (7.70)

(7) T° annual range (5-6)

22.50

19.00-25.40 (6.40)

22.45

19.00-24.30 (5.30)

(8) Mean T° wettest quarter

21.52

16.20-25.23 (9.03)

21.36

16.20-25.23 (9.03)

(9) Mean T° driest quarter

16.84

13.40-19.22 (5.82)

16.92

14.03-19.22 (5.18)

(10) Mean T° warmest quarter

24.76

20.33-27.58 (7.25)

24.89

20.33-27.58 (7.25)

(11) Mean T° coldest quarter

16.01

12.35-18.50 (6.15)

16.03

12.50-18.50 (6.00)

(12) Annual precipitation

1723.5

934-2235 (1301)

1730.5

1250-2235 (985)

(13) Precipitation wettest month

188.5

157-245 (88)

186

157-245 (88)

(14) Precipitation driest month

99.5

11-150 (139)

101

42-150 (108)

(15) Precipitation seasonality (CV)

20.34

9.42-83.74 (74.32)

19.94

9.42-41.88 (32.46)

(16) Precipitation wettest quarter

495.5

432-625 (193)

495.5

432-625 (193)

(17) Precipitation driest quarter

344.5

40-496 (456)

349.5

140-496 (356)

(18) Precipitation warmest quarter

457

354-625 (271)

455

354-625 (271)

(19) Precipitation coldest quarter

360

40-540 (500)

362

140-540 (400)

model (full dataset) was alternatively run for each variable
separately and combining some of them so as to visually
inspect the effects on the resulting predicted range.

RESULTS

Biocliinatic profile.-- Table 1 summarizes values relevant to
the biocliinatic tolerance range of G. sylvarum. Separate
profiles with and without the Tucuman localities are given
from which it is clear that, for some variables, conditions in
the yungas suggest differences from the core area (cf. Table 2
also). Sites from Tucuman represent both the westernmost
records and the highest elevation. In those localities precip-
itation (be 12) is the lowest (Fig. 2B). with rainfall decreasing
substantially during the winter (Fig. 2D). There, seasonality is
higher than in sites of the core area (Figs. 2A, C); in
particular, precipitation seasonality (be 15) is strongly skewed
to low values, with both Tucuman records clearly separated by
a decided gap from the rest (Fig. 2C). When these localities of
NWA are set aside, Rio Tragadero and 10 km Puerto
Antequera (close to each other) become the species’ western-
most edge and come to hold many of the biocliinatic extremes
related to seasonality and winter decrease of precipitation
(Table 2 and Fig. 3). However, geographically extreme
localities did not necessarily hold the highest or lowest
biocliinatic values in all cases (Table 2). Other localities with
many biocliinatic variables showing extreme values are
Clevelandia (Santa Catarina, Brazil) and Asuncion (Para-
guay). Clevelandia has nine end values indicating, in general, a
cooler and more humid climate than the rest; in turn,
Asuncion stands as the warmest site of the species range,
with highest values for six variables, all temperature-related.
With respect to the resulting envelope, a permissive percentile
cut-off of 0.005 allowed 33 out of 46 observations (71.7%) to
be included within all possible 171 bidimensional envelopes
(the remaining 13 observations being outsiders in at least one

bidimensional envelope). With the default percentile of 0.025,
localities within the species envelope are reduced to 21 out of
46 observations (45.7%).

Potential range. — The predicted range of G. sylvarum under
the model parameters is displayed in Fig. 1. To a great extent,
the predicted core area of G. sylvarum roughly matches the
Paranense forests ecoregion (Fig. 3). In eastern Paraguay,
where records are almost lacking, the prediction partially
redraws the boundaries of this ecoregion with the contiguous
Humid Chaco (Fig. 3). On the Brazilian side, the predicted
range seemingly enters the Araucaria forests only marginally,
i.e., in areas surrounded by complex eastward Paranense
projections that follow large rivers. A large portion of the
Araucaria ecoregion is scored as unsuitable; moreover, it is to
be noted that Clevelandia, one of the localities with many
biocliinatic extreme values (Table 2), is placed near this
presumable distributional limit (Fig. 3). In Brazil the modeled
range reaches up to the southern states of Mato Grosso do Sul
and Sao Paulo in the North (though with no records so far),
and weakly up to the western slopes of the Serra do Mar in the
East (Fig. 1). Geraeocormobius sylvarum was not hitherto
collected in the eastern slopes of the Serra do Mar; probably
replaced there by its congener Geraeocormobius rohri (Mello-
Leitao 1933) (A.B. Kury, pers. comm.). On the other side,
occurrences of G. sylvarum in the northern province of
Corrientes and eastern Chaco give support to the potential
areas southwest of the paranense forests. Subtropical vegeta-
tion partially extends into northern Corrientes, as well as
along flood and gallery forests along the Parana and Paraguay
rivers (Hueck & Seibert 1972), and this seems to provide
suitable conditions for G. sylvarum some hundreds of
kilometers away from the main range.

With the full dataset, the biocliinatic model predicts the
presence of G. sylvarum in the yungas but not across the sub-
xeric Chaco (Fig. 1); thus, the presumed disjunction is

ACOSTA— DISTRIBUTION MODELING OF GERAEOCORMOBIUS (OPILIONES)

577

C/5 ”0

53 5

G dJ

3 •§
c g
w '5b

C/5 C

o

•r o

CJ g
X >
w dJ

O X
*55 4J

,p e

x "p
S o

c .5

G O
w Oh

C/5 (D

rt ^

g .a

.G c/5

G .G
o —
■G G

'S °
a o

.O -o

2 "O

03 £

^ 03

a

g Q

Q>

b/j

P T3

3 «

dJ o

"§ s

Z, dJ

•— S-H

G P
o 3

<3

w c

G kG
c^J g
0) G
*0 o

.G H

° j s

S5J P

*03 ^ c+h

iG C/5

G <d -O

CJ ~ ^

0 -2 o

_J ,« O

1 H Si

l Ci„

<N •-

X C/5

o o u

x S x

c3 t3 c3

53 h
x g

p p p p

o o o o
o o o o

P) PI PI PI

§ £ £ £ §

p p p

o

o> S

^ g g g on r: (

G G G G G — 1 G

o (U 1) l) (U <U

c3 c3 t3 (3 c3

JJ dJ

o3

oo 5^

22 <£

b/j w

o O CJ o

< < < <

O c r?

G

. G O .G

< a* X & <

G

b/j ca)

G O

53 x; ^

_o dj aj

_dJ

<u

r- G
G bJj

o .£

X cP

dJ 1)

s s

bX) %
G ^
• G CJ

ed <

o o
o o
pi pi

a> <d

03 03

< <

p p

o o

o o

PI PI

. o o

00007:
Os OS ^

o3

dJ

dj

PI

>, jj

G 03

c/5 G G ^

O > > O

CJ 03 03 CJ

< H H <

I ^ I I I

_ os
O °°'

— r~

‘too'5 I (f| I | „ | Os Os

PI ^ X

1 4 so , — 1

in r— '

: pT

os

os

os

p ; Xh
X

X

pf ^ 2

^ rn w

Os" w

p O Pj Pi P- P-

00 O 00 cp cn s©

o o o m 00

00 00 00 cn iri

SO so SO P~ P- QC

PI P) PI PI PI PI

I I I I I I

- „ ^ o ^

■ !b S2 CJ S 00 1 1 1 rt,2 1 1 bi . r 1

so ^ in

pi

n

p- o p- p- p- 0 0

so O X X X o o

X O X X ^ o o

m 00 o cp os cn 'rt

p so p sd in n so

pi pi pi pi pi pi pi

I I I I I I I

O m m o O O p

o cn m o o o x

o m »n »n o in 00

m m — o pi *n in

cn in Tf p p^ p p

pi pi pi pi pi pi pi

I I I I I I I

m

p

cn

p

o

m

o

cn

cn

p

p

cn

p

p

p

p

p

o

o

p

p

QC

pi

X

o

cn

o

cn

cn

X

X

cn

X

X

X

X

o

o

p

X

i n

X

cn

X

o

X

o

cn

cn

X

p

X

X

X

X

pi

o

in

p

X

en

cn

cn

X

X

x

SO

o

X

X

X

so

cn

X

p

p

0©

X

pi

in

p

it]

in

in

0©

oo

p

«n

in

X

X

X

p

pi

pi

o

as

pi

cn

pi

pi

X

X

X

ITi

in

in

»n

X

in

•n

in

•n

«n

in

in

>n

>n

»n

«n

<n

I I I I I

C/5

I I I I

3 g

SPo

I:.

c3 CJ

— ’>

g on
2 o

<D

a- 'C?

2 5

< X

O '03

O

X

X

X

aj

U U X'
c o 2

QJ

U -O
-S G

aj O

CO Qi

3 G-

H S

53 o
« .y

111
b/j c oj
G a3 G
X C/5 Oh

£0

G

G
X
o O
t: g
aj G

P C/5

CL. <

H

G

X

G

G

CJ

O

H

G

X

G

OJ

N

G

X

I I I I I I

o

cu

G

G

X

G

O

X <0

G -G G

G .d ^

'G ‘53 H
G O _

u &: u a u z

S3 5

> I
«S c2

tx U

G G G
C^0 C/0 00

G -0

G G

vG X

G G

H H

G

VG

S o

G CJ

C/5 C/5 C/5 C/5 C/5

CJ b CJ 1)

aj

G g

O H

C/5 C/5 C/5 C/5

G

t=j 5 s u 'I % i i i £ £

G

G

- - u „ „ -

g c g a a a

G G *-*

1 G O O O

G G G X X X

0- cu 00 a! oC oi

GGGGGGGGGGG

.= .5 .5 .5 .= .2 .2 .5 .£ .2 .2 3

<D QJ OJ <U <D OJ (L) dJ dJ dJ dJ o3J'n'n'n'n'n'n'n'n'n

b/j b/j b/j bij b/j bij b/j bLj Sfj bXj bJj ^ 3 G G G G G G G G

5— ' ?— • 5— ' *— < *— • *— < *— < S-h 5— « S— i J— ( S-h *— < S-h J-h S— i J— i V— > J— « S— «

<<<<<<<<<<<d.oacQcacQcncaQ2CQCQ

578

THE JOURNAL OF ARACHNOLOGY

Figure 1 . -Predicted range (0-100 percentile) of Geraeocormobius sylvarum Holmberg, as modeled with Bioclim, using the full data set and all
19 bioclimatic variables. Provinces in Argentina: T, Tucuman (detailed map in Fig. 4); Fo, Formosa; Ch, Chaco; SF, Santa Fe, Co, Corrientes;
M, Misiones. States in Brazil: RS. Rio Grande do Sul, SC, Santa Catarina, Pr, Parana, SP, Sao Paulo, MGS, Matto Grosso do Sul.

supported. However, the probability assigned by the model for
its occurrence in Tucuman is very low. Even in a 0-100
percentile envelope, suitable gridcells are few (just four), and if
ranked, they fit in the lowest percentile levels (Fig. 4A). With a

0.025 cut-off, no suitable gridcell remains in NWA. As
expected by these results, modeling with Tucuman records
removed did not predict the species there either. Otherwise,
modeling the range without these NWA records had minimal

O 0.6

E

o 0.2

1 1 1 1 1 1 1 r 7 '

t

1.0

1 1 1 1 1 1 . 1 1 ;*

S’ 0 8

c

0

Z3

CT _ _

/

j

0 0.6

0

>

i

♦♦♦♦*

*♦♦♦

J

r

L

£0 04

13

E

ZJ

1

♦*

J

/

♦

£ i i i i i i i

O 0.2

0.0

‘ 1

L i_i 1 * 1 , l ,

O 10 20 30 40 50 60 70 80 90

bc15: precipitation seasonality

D

O 100 200 300 400

bc17: precipitation driest quarter

Figure 2. — Bioclimatic profile of Geraeocormobius sylvarum, selected variables: full data set plotted for cumulative relative frequency; arrows
indicate the position of the two localities from province of Tucuman (NWA). A: bc4 - temperature seasonality (standard deviation x 100); B:
be 12 annual precipitation (mm); C: bcl5 - precipitation seasonality (coefficient of variation); D: bcl7 - precipitation of the driest quarter (mm).

ACOSTA- - DISTRIBUTION MODELING OF GERAEOCORMOBIUS (OPILIONES)

579

Figure 3. — Correspondence of the modeled range of Geraeocormobius sylvarum (grey area, 0-100 percentiles) and the Alto Parana Atlantic
forests ecoregion (squared pattern fill); the overlap is shown as a lighter grey squared area. AMF: Araucaria moist forests (bordered by a thin
line), where G. sylvarum is predicted to occur marginally. Dots: all locality records for the species. References of selected localities: 1 : Asuncion; 2:
Rio Tragadero; 3: Clevelandia; 4: Caviuna; 5: Paranavai.

effects in the core area: just a small clipping occurs, reducing
the range on marginal sides, like the Corrientes-Chaco
portions, and its northernmost extension in Brazil.

Limiting factors. — Models of G. sylvarum run separately,
with either temperature ( be 1 — 11) or precipitation (be 1 2—19)
variables, uncover the relative contribution of these variables
to the predicted range shape. Prediction based on temperature
variables bcl-bcll (Figs. 4B, 5A) more closely recovers the
model obtained by all 19 variables together, with a slight
“permissiveness” in the southernmost border (province of
Corrientes) and especially in NWA (patchy suitable areas
appear along the yungas in Tucuman, Jujuy, Salta, and
southern Bolivia). Precipitation variables bcl2-bcl9 (Figs. 4C,

5B) clip most of these areas away from the model, but, in turn,
extend suitable conditions considerably eastwards and north-
wards. In fact, many precipitation variables (when run alone)
predict G. sylvarum in the yungas, but almost all prevent it
entering the Chaco, some being more rigorous (be 15, be 16,
be 18), others more “permissive” (bcl2, be 14, bcl7, bcl9); only
be 13 enables a somewhat continuous range. In contrast, only a
few temperature variables (bc5 and bc8) contribute to the
Chaco gap. Temperature variables are more related to range
restrictions in the North and the East. The northernmost
boundaries in Brazil, where the range meets the Cerrado and is
limited by high temperatures, are fairly well shaped by bcl,
bc4, bc6, bc9, and bell. As noted above, the modeled range

Figure 4. — Detail of the predictive modeling in province of Tucuman, Argentina. A: model with all 19 bioclimatic variables (small dots
indicate the occurrence records), B: model with only temperature variables (bcl— be 1 1), C: model with only precipitation variables (be 1 2 be 19).
Light grey: 0-25 percentile; dark grey: above 0.25.

580

THE JOURNAL OF ARACHNOLOGY

Figure 5. — Potential distribution of Geraeocormobius sylvarum modeled only with temperature or precipitation variables (light grey: 0-25 percentile;
dark grey: above 0.25). A: modeled range with temperature variables (be 1 —be 1 1), B: modeled range with precipitation variables (bcl 2— be 19).

seems to circumvent the Araucaria forest, entering this
ecoregion only marginally (Fig. 3). Several temperature
variables (bcl, bc5, bc6, bc8, bc9, bclO, bell) leave a defined
gap there. The unsuitability appears to be related to the high
elevation, which in turn causes the seemingly unsuitable cold
climate: Clevelandia, elev. 890 m, is the highest record of G.
sylvarum in Brazil and bears several extreme cold climate
values (Table 2), but many negative sites in the Araucaria area
are above 1 100 m.

Accuracy. — Modeling with the training samples (30% of the
original points randomly removed) consistently recovered the
main portions of the range predicted with all points. More than
half of the 20 repetitions matched in most parts of the core area,
and just marginal sectors overlapped in only 10 repetitions or
less. AUC values ranged from 0.8505 to 0.9402 (mean =
0.8935), and max-K from 0.7082 to 0.8832 (mean = 0.7900);
thus, individual and average AUC and max-K values resulted in
good to high model performances. These values are consistent

ACOSTA— DISTRIBUTION MODELING OF GERAEOCORMOBIUS (OPILIONES)

581

with performances of Bioclim obtained by Hernandez et al.
(2006) for similar sample sizes, and with comparative evalua-
tions made by Sangermano & Eastman (2007).

DISCUSSION

As stressed by Peterson (2006), for the vast majority of
species nothing more is known than a few “dots on maps” -
and certainly, this applies for most Neotropical harvestmen,
too. Ecological niche (or habitat suitability) modeling offers a
first step towards inferring the basic ecological dimensions
that are relevant to limit the species’ distribution. While
previous knowledge just imprecisely related G. sylvarum to
Paranense forests (and thereby assumed it was dependent on
humid conditions), the bioclimatic analysis provided, for the
first time, defined values to describe and discuss the species
profile. The envelope method proved to be well suited as an
initial approach when species records and biological knowl-
edge are still scarce (Pearce & Boyce 2006). Results were
convincing when matching the predicted range polygon to the
Paranense forests ecoregion {Fig. 3), a biogeographical area
that was previously associated with G. sylvarum based on a
“non analytical, expert-based” assessment (Acosta et al.
2007). The model accuracy as measured by AUC and max-K
was acceptable as well.

One of the main issues tackled in this paper, the
Mesopotamia-yimgas disjunction, received support from this
model; but at the same time, some new questions arose
concerning the presence of G. sylvarum in NWA. Despite the
availability of presence records, the species is only weakly
predicted there. This contrasts with other Mesopotamian
harvestmen (D. dilatatus and D. prospicuus among them) that,
in preliminary models, were predicted in the yungas, with or
without positive records in NWA (Acosta 2007). Both D.
dilatatus and D. prospicuus belong to the Mesopotamian sensu
stricto sub-area (Acosta 2002), so their climatic requirements
are not expected to be as humidity-dependent as for G.
sylvarum. Since Bioclim is deemed to over-predict a species’
range (Peterson 2001), these facts, together with the profile
values for some variables, suggest an important disparity of
the bioclimatic conditions of NW and NE Argentina
concerning preferences of G. sylvarum. However, one should
not lose sight of the importance of the resolution of the
climatic data. In general the resolution used (2.5 min) is
acceptable at a regional scale, but its coarseness may not
adequately reflect small-distance variations in montane areas,
like in Tucuman (records range from 500 to 1250 m elevation
within a distance of less than 3 km; i.e., less than the gridcell
size).

As for the causes of the disjunction, a satisfactory
explanation remains an open question. Acosta (1995, 2002)
suggested that the disjunct ranges of the two Discocyrtus
species may be a consequence of paleoclimatic cycles, with
associated expansion/retraction events affecting the humid
forests. A hypothetical vegetational “bridge” (as proposed by
Nores 1992 for birds) would have acted as a corridor, enabling
Mesopotamian harvestmen to expand their ranges up to the
yungas, to leave isolated populations there as climate turned
rigorous and the forests retracted. In the case of G. sylvarum ,
however, several elements may render this explanation less
likely. Records in the yungas are actually few, indicating a

quite limited distribution, and the bioclimatic model accord-
ingly predicts this species only weakly for the region. The
yungas is one of the best sampled areas in Argentina for
harvestmen (Acosta 2002), so it seems unlikely to find G.
sylvarum elsewhere in NWA. People from the species’ core
range are in general aware of G. sylvarum, probably because of
its large size, abundance and strong odor. These conspicuous
and smelly harvestmen are also locally well known in El Corte,
although people that have lived there for the last 30 years
suggest that the species “appeared” only in recent years (A. M.
Frias, pers. comm.). These observations, together with the
apparent tolerance of G. sylvarum to disturbed areas (Acosta
et al. 2007), may suggest that transportation by humans in
historic times might best account for the presence of this
Mesopotamian harvestman in Tucuman.

ACKNOWLEDGMENTS

I am indebted to Maria Laura Juarez, Gustavo Flores,
Dardo Marti, Adriano Kury, and Miryam Damborsky for
providing coordinates or details of the collecting sites of G.
sylvarum. Comments and suggestions given by Raquel M.
Gleiser, Soren Toft, Paula Cushing and an anonymous referee
greatly improved the manuscript. Financial support was given
by Consejo Nacional de Investigaciones Cientificas y Tecnicas
(CONICET: P.I.P. N° 6319) and SECyT-UNC. The author is
a researcher of CONICET.

LITERATURE CITED

Acosta, L.E. 1995. Nuevos hallazgos de Discocyrtus dilatatus en
Argentina, con notas sobre taxonomia, sinonimia y distribucion
(Opiliones, Gonyleptidae, Pachylinae). Revue Arachnologique 10:
207-217.

Acosta, L.E. 2002. Patrones zoogeograficos de los Opiliones
argentinos (Arachnida: Opiliones). Revista Iberica de Aracnologia
6:69-84.

Acosta, L.E. 2007. Distribution of harvestmen (Opiliones) in the
Argentinean Mesopotamia: a modeling approach based on
bioclimatic variables. 17th International Congress of Arachnology,
Sao Pedro, SP, Brazil. Abstracts, p. 49.

Acosta, L.E., A.B. Kury & M.L. Juarez. 2007. New records of
Geraeocormobius sylvarum (Arachnida, Opiliones, Gonyleptidae),
with a remarkable disjunction in northwestern Argentina. Boletin
de la Sociedad Entomologica Aragonesa 41:303-306.

Cabrera, A.L. & A. Willink. 1973. Biogeografia de America Latina.
Coleccion Monografias Cientificas O.E.A., serie Biologia, n° 13,
pp. I-VI, 1-117.

Curtis, D.J. & G. Machado. 2007. Ecology. Pp. 280-308. In
Harvestmen: The Biology of Opiliones. (R. Pinto-da-Rocha, G.
Machado & G. Giribet, eds.). Harvard University Press, Cam-
bridge, Massachusetts.

Fischer, J., D.B. Lindenmayer, H.A. Nix, J.L. Stein & J.A. Stein.
2001. Climate and animal distribution: a climatic analysis of the
Australian marsupial Trichosurus caninus. Journal of Biogeogra-
phy 28:293-304.

Giribet, G. & A.B. Kury. 2007. Phylogeny and biogeography.
Pp. 62-87. In Harvestmen: The Biology of Opiliones. (R. Pinto-
da-Rocha, G. Machado & G. Giribet, eds.). Harvard University
Press, Cambridge, Massachusetts.

Guisan, A. & W. Thuiller. 2005. Predicting species distribution:
offering more than simple habitat models. Ecology Letters 8:993-
1009.

Guisan, A. & N.E. Zimmermann. 2000. Predictive habitat distribu-
tion models in ecology. Ecological Modelling 135:147-186.

582

THE JOURNAL OF ARACHNOLOGY

Hernandez, P.A., C.H. Graham, L.L. Master & D.L. Albert. 2006.
Tire effect of sample size and species characteristics on perfor-
mance of different species distribution modeling methods. Eco-
graphy 29:773-785.

Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones & A. Jarvis.
2005a. Very high resolution interpolated climate surfaces for global
land areas. International Journal of Climatology 25:1965-1978.

Hijmans, R.J. & C.H. Graham. 2006. The ability of climate envelope
models to predict the effect of climate change on species
distributions. Global Change Biology 12:1-10.

Hijmans, R.J., L. Guarino, A. Jarvis, R. O’Brien, P. Mathur, C.
Bussink, M. Cruz, I. Barrantes & E. Rojas. 2005b. DIVA-GIS,
version 5.4.0. 1. Online at http://www.diva-gis.org/

Holmberg, E.L. 1887. Viaje a Misiones. Boletin de la Academia
Nacional de Ciencias, Cordoba 10:5-391.

Hueck, K. & P. Seibert. 1972. Vegetationskarte von Siidamerika.
Vegetationsnronographien der einzelnen GroBraumen, 2a: 1-71 +
map. G. Fischer, Stuttgart.

Kury, A.B. 2003. Annotated catalogue of the Laniatores of the New
World (Arachnida, Opiliones). Revista Iberica de Aracnologia,
Volumen especial monografico 1:5-337.

Luoto, M., J. Pdyry, R.K. Heikkinen & K. Saarinen. 2005.
Uncertainty of bioclimate envelope models based on the geo-
graphical distribution of species. Global Ecology and Biogeogra-
phy 14:575-584.

Machado, G., R. Pinto-da-Rocha & G. Giribet. 2007. What are
harvestmen? Pp. 1-13. In Harvestmen: The Biology of Opiliones.
(R. Pinto-da-Rocha, G. Machado & G. Giribet, eds.). Harvard
University Press, Cambridge, Massachusetts.

Martens, J. 1978. Weberknechte, Opiliones. Die Tierwelt Deutsch-
lands 64:1—464. Gustav Fischer, Jena

Nores, M. 1992. Bird speciation in subtropical South America in
relation to forest expansion and retraction. Auk 109:346-357.

Olson, D.M., E. Dinerstein, E.D. Wikramanayake, N.D. Burgess,
G.V.N. Powell, E.C. Underwood, J.A. D’Amico, H.E. Strand, J.C.
Morrison, C.J. Loucks, T.F. Allnutt, J.F. Lamoreux, T.H.
Ricketts, I. Itoua, W.W. Wettengel, Y. Kura, P. Hedao & K.
Kassem. 2001. Terrestrial ecoregions of the world: a new map of
life on Earth. BioScience 51:933-938.

Pearce, J.L. & M.S. Boyce. 2006. Modelling distribution and
abundance with presence-only data. Journal of Applied Ecology
43:405-^412.

Peterson, A.T. 2001. Predicting species’ geographic distributions
based on ecological niche modeling. Condor 103:599-605.

Peterson, A.T. 2006. Uses and requirements of ecological niche
models and related distributional models. Biodiversity Informatics
3:59-72.

Pinto-da-Rocha, R., M.B. da Silva & C. Bragagnolo. 2005. Faunistic
similarity and historic biogeography of the harvestmen of southern
and southeastern Atlantic Rain Forest of Brazil. Journal of
Arachnology 33:290-299.

Randin, C.F., T. Dirnbock, S. Dullinger, N.E. Zimmermann, M.
Zappa & A. Guisan. 2006. Are niche-based species distribution
models transferable in space? Journal of Biogeography 33:1689-
1703.

Ringuelet, R.A. 1959. Los aracnidos argentinos del Orden Opiliones.
Revista del Museo Argentine de Ciencias Naturales, Ciencias
Zoologicas 5:127-439.

Sangermano, F. & R. Eastman. 2007. Linking GIS and Ecology - The
use of Mahalanobis typicalities to model species distribution.
Pp. 1-13. In Memorias XI Conferencia Iberoamericana de
Sistemas de Informacion Geografica. (G.D. Buzai, ed.). Buenos
Aires.

Santos, F.H. 2007. Ecophysiology. Pp. 473-488. In Harvestmen: The
Biology of Opiliones. (R. Pinto-da-Rocha, G. Machado & G.
Giribet, eds.). Harvard University Press, Cambridge, Massachu-
setts.

Soares, B.A.M. & H.E.M. Soares. 1945. Alguns opilides do Museu
Nacional do Rio de Janeiro. Papeis do Departainento de Zoologia
do Estado de Sao Paulo 5:221-226.

Soares, H.E.M. & B.A.M. Soares. 1985. Contribution a 1’etude des
opilions (Opiliones: Cosmetidae, Phalangodidae, Gonyleptidae) du
Paraguay. Revue Suisse de Zoologie 92:3-18.

Tavares, M.L.R. 1980. Novas ocorrencias de opiiioes no Rio Grande
do Sul e descrigao da femea de Melloleitaniana riodariensis Soares
& Soares, 1945 (Arachnida-Opiliones-Gonyleptidae). Iheringia,
Serie Zoologia 55:155-159.

Walther, B.A., M.S. Wisz & C. Rahbek. 2004. Known and predicted
African winter distributions and habitat use of the endangered
Basra reed warbler ( Acrocephalus griseldis) and the near-threatened
cinereous bunting ( Emberiza cineracea). Journal of Ornithology
145:287-299.

Manuscript received 1 June 2007, revised 18 February 2008.

2008. The Journal of Arachnology 36:583-594

Ecology and web allometry of Clitaetra irenae , an arboricolous African orb-weaving spider

(Araneae, Araneoidea, Nephilidae)

Matjaz Kuntner1’2’5, Charles R. Haddad1, Gregor Aljancic4 and Andrej Blejec4: 'Institute of Biology, Scientific Research
Centre of the Slovenian Academy of Sciences and Arts, Novi trg 2, SI- 1001 Ljubljana, Slovenia; department of
Entomology, National Museum of Natural History, NHB-105, Smithsonian Institution, P.O. Box 37012, Washington,
DC 20013-7012, USA; department of Zoology & Entomology, University of the Free State, P.O. Box 339,
Bloemfontein 9300, South Africa; department of Biology, University of Ljubljana, P.O. Box 2995, SI- 1001 Ljubljana,
Slovenia

Abstract. Analysis of ecological data of the arboricolous nephilid spider Clitaetra irenae Kuntner 2006, endemic to
Maputaland forests, South Africa, indicates the species’ dependence on this highly threatened habitat. We tested C. irenae
habitat dependence via GIS analysis by plotting the known distribution against southern African ecoregions. In the
southern part of its range, C. irenae inhabits almost exclusively one ecoregion. the Maputaland coastal plain forests; but
further north, in tropical southern Africa, it continues inland into Malawi’s woodlands. We test and refute the hypotheses
that C. irenae inhabits exclusively mature trees, trees of a particular species, trees with a smooth bark, tree habitats at
certain height above ground, and only closed canopy forest stands. The ecological niche of C. irenae is flexible as long as
suitable trees under at least partially closed canopy are available. We quantify the C. irenae ontogenetic web changes from
orb to ladder and the simultaneous hub displacement towards the top frame. Such web allometry allows the web to increase
vertically but not horizontally, which enables the spider to remain on the same tree throughout its development and thus
the ladder web architecture is an adaptation to an arboricolous life style. Adult hub displacement, common in spiders with
vertical webs, is explained by gravity. Clitaetra irenae web orientation on trees correlates with forest closure, and might
indicate the Maputaland forest quality. We argue for utilization of the ecology of arboricolous nephilid orb-weaving
spiders (Clitaetra and Herennia ) in systematic conservation assessments in the Old World tropics.

Keywords: Behavioral ecology, evolution, conservation. Maputaland, South Africa, Herennia

Opsomming. Analise van ekologiese data van die boombewonende nephilid spinnekop Clitaetra irenae Kuntner 2006,
endemies tot Maputaland woude in Suid-Afrika, dui die spesie se afhanklikheid van hierdie hoogs bedreigde habitat aan.

Ons het C. irenae habitat afhanklikheid via GIS analiese getoets deur die bekende verspreiding teen die suider-Afrikaanse
ekologiese streke aan te teken. In die suidelike deel van die verspreiding kom C. irenae in slegs een ekostreek, die
Maputaland kusvlakte woude, voor, maar verder noord in die verspreiding kom dit verder in die binneland, in Malawiese
bosveld. voor. Ons toets en verwerp die hipotesisse dat C. irenae slegs volgroeide borne bewoon, borne van ‘n spesifieke
spesie, borne met gladde bas, boomhabitatte op ‘n sekere hoogte van die grond af, en slegs toe kroonbedekking woude. Die
ekologiese nis van C. irenae is aanpasbaar solank daar geskikte borne onder ten minste gedeeltelike toe kroonbedekking
beskikbaar is. Ons kwantifiseer C. irenae ontogenetiese webveranderings van ‘n wawiel- na ‘n leerweb en die gelyktydige
verskuiwing van die kern na die bokant van die raam. Sulke web allometrie gee die spinnekop die vermoe onr die web
vertikaal te vergroot sender horisontale veranderings, wat die spinnekop in staat stel om op dieselfde boom te bly regdeur
sy ontwikkeling. Ons sien die leerweb dus as ‘n aanpassing tot boomlewende gewoontes. Volwasse kern verskuiwing,
algemeen in spinnekoppe met vertikale webbe, word deur gravitasie verduidelik. C. irenae se web orientasie op borne
korreleer met kroonbedekking, en kan ‘n aanduider wees van Maputaland woud kwaliteit. Ons stel voor dat nephilid
wawiel-web spinnekoppe (Clitaetra en Herennia) se ekologie gebruik word in sistematiese bewarings assesserings in die Oil
Wereld tropiese gebiede.

Identifying areas of high conservation value (systematic
conservation assessment sensu Knight et al. 2006) is one of the
main priorities of modern ecology and conservation biology.
Traditionally, conservation assessments have been largely
based on vertebrate biodiversity data, while data on more
diverse groups such as arthropods remain poorly utilized. This
is particularly true for species-rich tropical faunas, where the
natural history and ecology of most arthropod species
continue to be unstudied. Among spiders, a clade with nearly
40,000 described species (Platnick 2007) and many more
projected (Coddington & Levi 1991; Coddington 2005),
potential indicator species are rarely identified. Most recent
studies have investigated the impacts of pollution, disturbance,

5 Corresponding author. E-mail: kuntner@gmail.com

and habitat classification on spiders at the community level,
while species-level studies have focused on effects on predation
ecology and heavy metal assimilation (Marc et al. 1999).
However, these studies primarily dealt with the European and
American faunas, while examples from the Afrotropical region
are unknown. Here, we investigate the ecology of the recently
described Clitaetra irenae Kuntner 2006, an arboricolous
nephilid spider endemic to southern African forests, and use
the newly acquired behavioral and ecological data to test its
dependence on a highly threatened habitat - the Maputaland
coastal plain forests.

Maputaland is an ecological-geographical entity comprising
the coastal plain of north-eastern parts of KwaZulu-Natal
(South Africa) from Richards Bay (28°48'S, 32°05'E) in the
south to Xai-Xai in southern Mozambique (25°02'S, 34°25'E)

583

584

THE JOURNAL OF ARACHNOLOGY

in the north, and extending inland along the Lebombo
Mountain range into eastern Swaziland (Olson et al. 2001;
Van Wyk & Smith 2001). Due to the high levels of plant and
vertebrate endemism (Van Wyk 1994, 1996; Van Wyk & Smith
2001) and the patchy distribution of unique habitats such as
sand forest, Maputaland is of prime conservation importance.
Furthermore, the composition of certain faunal and floral
assemblages of different sand forest patches are significantly
different (e.g., Kirkwood & Midgley 1999; Matthews et al.
1999, 2001; Van Rensburg et al. 1999, 2000), which supports
the conservation of this particular habitat and its accompa-
nying biological diversity throughout the region. Maputaland
appears among six out of nine global biodiversity conservation
priority areas (Brooks et al. 2006), being recovered within the
crisis ecoregions, the biodiversity hot spots, the endemic bird
areas, the centers of plant diversity, and the global 200
ecoregions, as well as within a megadiverse country. South
Africa (for details see Brooks et al. 2006, and references
therein).

The Old World tropical nephilid spider genus Clitaetra
Simon 1889 is important in araneoid systematics because of its
phylogenetic position as the sister taxon to all other nephilids
(Kuntner 2005, 2006); these, in turn, appear to be the basal
araneoid lineage (Kuntner et al. 2008). While the large-bodied
nephilids ( Nephila , Nephilengys ) are well studied (Kuntner
2007) and recognized as ecologically important in most
(sub)tropical ecosystems, Clitaetra ecology has largely been
unknown (but, see Kuntner 2006). Of the six known Clitaetra
species only C. irenae (Fig. 1) biology has been studied in
some detail — Kuntner (2006) focused on its web-building
behavior and general natural history. Clitaetra irenae was
found in South Africa’s subtropical coastal dune and sand
forests of northern KwaZulu-Natal, with an outlying museum
record from Malawi (Kuntner 2006: fig. 26). As in the
Australasian nephilid spider genus Herennia Thorell 1877 (see
Kuntner 2005), Clitaetra species are obligate tree dwellers
building orb-webs against tree bark (Figs. 1, 2), with a few
centimeters clearance between the planar web and the
substrate. As in Herennia (Kuntner 2005), Nephilengys Koch
1872 (Japyassu & Ades 1998; Kuntner 2007), and Nephila
Leach 1815 (Bleher 2000), juvenile Clitaetra webs resemble
standard araneoid orb-webs (Figs. 2D, E). However, the webs
of adult female C. irenae are highly modified ladder webs
(Figs. 2A, C), defined as two dimensional orb-webs with
parallel side frames and zig-zag sticky lines substituted for
spirals (Kuntner et al. 2008:fig. 23). The species’ apparent
dependence on a particular habitat (Maputaland forests) and
microhabitat (mature trees), as indicated by historical
distribution records and the distribution of these particular
forests, prompted our detailed investigation into its behavioral
ecology emphasizing its ontogenetic web architecture shifts.

We investigate C. irenae web allometry by quantifying the
developmental modification of the orb web into a ladder, pose
a behavioral explanation for such an allometric shift, and
deduce its ecological implications. Through explicit hypothe-
ses (below) we examined whether or not the web modification
in C. irenae correlates with a particular habitat type, with
climatic conditions and/or altitude, and whether the species
depends on mature Maputaland forests for sustainable
populations. A species that is a narrow endemic of southern

Africa and 'a narrow habitat specialist (an obligate tree

dweller) in a threatened coastal plain habitat is potentially at

risk of habitat loss, and could be added to the list of species

used as indicators of habitat quality.

Specifically, we tested the following hypotheses:

(1) Endemism: Kuntner (2006) suggested that C. irenae
was endemic to northern KwaZulu-Natal. However,
such an hypothesis was defined geographically, not
ecologically, and failed to explain the single outlying
record from Malawi (Kuntner 2006). We tested
whether C. irenae is endemic to the wider Maputa-
land (including Mozambique and Swaziland) by
using precisely defined ecoregions (Olson et al. 2001).

(2) Habitat preference : Most material examined by
Kuntner (2006) was collected in sand forests along
the Maputaland coastal plain. We investigated
Kuntner’s (2006) expectation that C. irenae is
confined to mature sand forests. During our survey
we noted the habitat type and canopy closure.

(3) Tree preference : To test the spiders’ dependence on a
particular tree species we attempted to identify all
trees where spiders were found. We predicted the
spiders would prefer a particular tree species as
vaguely suggested by Kuntner (2006).

(4) Tree size: If the species was confined to mature
forests, we predicted that larger spiders would prefer
larger trees, and that web size would correlate with
spider size.

(5) Substrate (bark) structure: To test if the spiders
utilize certain types of substrate, we classified bark
into three categories (smooth, medium, or rough).
We expected that trees with smooth bark would be
more likely to host these spiders as the uneven surface
of rough bark may hinder two-dimensional web
construction.

(6) Website orientation: The spider’s website orientation
on a tree might differ between closed canopy forests
and partially open canopy tree stands (woodland/
thicket) due to sunlight penetration. Assuming that
the species is indeed native to closed canopy forests
and that inhabiting other types of tree stands is an
artifact of habitat fragmentation, we expected that we
would find no web orientation preference in closed
canopy stands, where continuous direct sunlight is an
exception. In contrast, in more open tree stands
where webs could potentially be exposed to direct
sunlight, web orientation should be away from the
sun (to the south).

(7) Distance from ground: Finally, we tested for differ-
ences in web height above ground among the
spider instars (size classes). We predicted that spiders
would preferentially occupy lower positions on a tree
as they age because larger tree trunk circumference
closer to the ground would be more suitable for larger
webs.

KUNTNER ET AL. ECOLOGY OF CL1TAETRA IRENAE

585

Figure 1. — Clitaetra irenae from South Africa (A-C, Fanies Island, 2001; D, Ndumo Game Reserve, 2006): A. Female holotype (see Kuntner
2006) at the hub of her web; B. Female feeding at the hub of her web; note the egg-cases above the hub camouflaged with prey remains; a male
was present in her web (not shown); C. Paratype male (see Kuntner 2006) at the hub of his orb web, feeding on dipteran prey; D. Subadult female
at web hub, feeding on cricket. Scale bars = ! .0 mm.

(8) Web allometry: Clitaetra irenae web architecture
shifts during spider ontogeny from orb to ladder
web. Thus, we predicted that the ladder index and the
hub displacement index (both defined below) would
increase with spider size.

METHODS

Field study. — We visited six reserves (Fig. 3) in KwaZulu-
Natal (South Africa) between 9-30 April 2006, the period
when adult Clitaetra irenae are known to occur (Kuntner
2006). These reserves (Ndumo Game Reserve, Tembe Ele-

586

THE JOURNAL OF ARACHNOLOGY

Figure 2. — Clitaetra irenae web architecture: A-C. Adult female webs; A. Frontal view, note ladder shape of web with parallel side frames (F),
“sticky spiral’’ (SS) not spiralling, with silk enforced hub (H) displaced towards top frame; B. same web, lateral view; C. Female (arrow) in web,
with first instar offspring in upper web; above hub is empty egg sac. D. Juvenile web (second instar); note typical orb-web architecture with bent
side frames (F), round sticky spiral (SS), and upper radii as long as lower ones. E. Another second instar web. with parameters measured (a =
web width, b = web height, c = top frame to hub). Photographs by M. Kuntner taken in Ndumo Game Reserve, South Africa, 2006; web
measurements (cm) in A, B: a = 8, b = 29, c = 10.5; D: a = 5, b = 11, c = 5.5.

phant Park, Kosi Bay Nature Reserve, Sodwana Bay National
Park, Hluhluwe-Imfolozi National Park, and the Greater St.
Lucia Wetlands Park) slightly surpass the previously known
C. irenae geographic range, except for the Malawi datum
(Kuntner 2006), and therefore test the species’ geographical
limits in Maputaland. Within the reserves we searched for the
easily recognizable species (females, see Fig. 1A; males, see

Fig. 1C; for identification details see Kuntner 2006) at various
localities, focusing on all available habitat types.

Where more than a single C. irenae spider was found, we
measured the following ecological and behavioral parameters
on all webs: date of collection; locality; site; habitat; latitude
and longitude; host tree species (if known); tree bark structure
categorized as smooth (as exemplified by Celtis africana ),

KUNTNER ET AL. — ECOLOGY OF CLITAETRA IRENAE

587

KOSI BAY

; Maputaland coastal forest mosaic
KwaZulu-Cape coastal forest mosaic
[ J| Southern Africa mangroves

Zambezian and Mopane woodlands
Drakensberg montane grasslands, woodlands and forests
Maputaland-Pondoland bushland and thickets
other ecoregions

HLUHLUWE

A ^ SODWANA BAY

ST. LUCIA

☆

(^) not found
• 1958

A 1991, 1998

O 2001
2006

25

50

Figure 3. — Clitaetm irenae, currently known distribution plotted against southern African ecoregions. The outlying 1958 Malawi record (see
Kuntner 2006) falls into “Zambezian and Mopane woodlands," which stretches south adjacent to “Maputaland coastal forest,” the species'
prime ecoregion. The 1991-2001 data are from Kuntner (2006), new Field records (2006) are from this study.

medium (as in Balanites maughamii), or rough (as in Acacia
nigrescens ); tree trunk circumference at the level of spider web
hubs; canopy coverage categorized as closed, partially open,
or open; stage or instar number; web width (Fig. 2E), web
height (Fig. 2E), distance from web hub to top frame
(Figs. 2D, E) and web hub height above ground; website
orientation (8 directions: N, NE, E, SE, etc.), and possible
further comments.

The First six (autecological) parameters document locality,
habitat, and test hypothesis 2 (habitat preference). The next
four parameters, i.e., host tree species, bark structure, trunk
circumference and canopy coverage, are microhabitat data
and test hypotheses 2-5. Spiders were assigned to seven
categorical size classes (stages) corresponding to ontogenetic
instar numbers (size correlates with age and instar number).
Instar numbers were estimated from relative spider size, falling
into fairly discrete size classes, assuming seven post-egg sac
instars during female ontogeny: adult females were scored as
7, newly hatched animals 1, and others fell in between up to 6

(penultimate). Instars 2-7 build their own webs, but first instar
spiderlings remain in the mother’s web (Fig. 2C). No
developmental study has been done on Clitaetra. The
assumption of seven instars in C. irenae ontogeny follows
from data on other nephilids; e.g., the giant females Nephila
pilipes (Fabricius 1793) go through about 10 juvenile instars
while the small males go through about 4, the number of molts
depending on the season and food availability (Higgins 2002).
Three web parameters, namely web width, web height, and
distance from hub to top frame, quantified developmental
shifts in webs (see below). The last two parameters, distance
from hub to ground and website orientation, indicate
microhabitat preference and test hypotheses 6 and 7.

In order to increase visual contrast, webs were dusted with
cornstarch for measurement and photography. All measure-
ments were taken with a tape measure and are reported in
centimeters, unless noted otherwise. Web orientation was
taken with a compass. Web architecture abbreviations are: H
= hub (Figs. 2A, B, D, E); F = frame (Figs. 2A. D); RA =

588

THE JOURNAL OF ARACHNOLOGY

radius (Figs. 2A, D); SS = sticky spiral (Figs. 2A, D); a = web
width (Fig. 2E); b = web height (Fig. 2E); c = upper web
height = distance top frame to hub (Fig. 2E).

Web allometry. — We derived two ratios quantifying web
allometry. The ladder index is the relative web height, defined
as the ratio of web height to web width, and quantifies the
transition from orb to ladder web during ontogeny. Similar to
web shape sensu Zschokke (1993), the ladder index, as used
here, differs in values increasing with extreme architecture.
Hub displacement is lower orb/total web height using the
formula (b-c)/b, where b = web height and c = top to hub
(Fig. 2E). Hub displacement index is similar to web asymme-
try indices of Masters & Moffat (1983), Rhisiart & Vollrath
(1994), and Kuntner et al. (2008), but its values increase rather
than decrease with the hub being eccentric towards the top
web frame.

Statistical analyses. — We explored the relationships between
the spider size (instar) and the following variables: 1) web
height, 2) web width, 3) distance of web from the ground, 4)
tree circumference, 5) ladder index, and 6) hub displacement
index. Web parameters were plotted as box-plots (Tukey
1977). Due to the low number of webs in stages 4 and 5,
individual points were plotted for these stages. Differences in
medians between stages 2 and 7 were tested using the Mann-
Whitney U- test. Freely available statistical environment R (R
Development Core Team 2006) was used for plotting and for
significance tests. Web orientation on trees (eight main
directions) was interpreted through circular statistics using
the Rayleigh test (Fisher 1993).

GIS analysis. — In order to test hypothesis 1 (endemism), we
analyzed all available C. irenae locality data (Kuntner 2006;
this study) using GIS. Our analysis builds on the digital base
map by Bletter et al. (2004), which was obtained from the New
York Botanical Garden (http://www.nybg.org/bsci/digital_
maps/), with permission from the authors. The base map,
which derives from the Environmental Systems Research
Institute (ESRI) data sets ArcWorld®, ArcAtlas®,and Digital
Chart of the World® (http://www.esri.com/), was built
specifically for the Neotropics (Bletter et al. 2004), but also
contains detailed data on the World’s terrestrial ecoregions
(Olson et al. 2001), courtesy of the World Wildlife Fund
(http://www.wwf.org/). Olson et al. (2001) recognize 867
terrestrial ecoregions worldwide, which is a much improved
resolution compared with previous attempts to classify
terrestrial biotas. The map delimits six ecoregions within the
geographical limits relevant to this study (Fig. 3). The C.
irenae locality data (1958; 1991-2001 data are from Kuntner
(2006); new field records (2006) are from this study) were
superimposed on the southern African part of the base map
using ArcView® GIS.

RESULTS

Clitaetra irenae occurred in all reserves (Fig. 3) except
Hluhluwe-Imfolozi National Park, which falls into the
Drakensberg montane grasslands, woodlands, and forests
ecoregion (Fig. 3). In Kosi Bay and Sodwana Bay the spiders
were present but their abundances per trees were too low to be
measured. In some reserves, e.g., Ndumo and Tembe, the
species was abundant. The outlying record of a single C. irenae
male from Malawi (1958 datum from Kuntner 2006) falls on

the ecoregion “Zambezian and Mopane woodlands,” which
stretches south adjacent to “Maputaland coastal forest”
(Fig. 3), the species’ main ecoregion.

Figure 4 summarizes the data obtained in Ndumo Game
Reserve, Tembe Elephant Park and the two localities in the
Greater St. Lucia Wetlands Park. The full data set with exact
geographical coordinates for each habitat within the reserves is
published online at www.nephilidae.com. In total, we investi-
gated 166 spiders and their webs, the majority (N = 1 18, 71%)
being second instars; the numbers of other instars investigated
were much lower (Fig. 4A). Habitat preference characterizes
the four forest types where the webs were measured (Fig. 4B).
We characterized the habitat at Nyamiti Hide (Ndumo Game
Reserve), where the majority of the aggregated second instar
webs were found, as “riverine bush” (“deciduous orthophyll
scrub with trees” of De Moor et al. (1977); “subtropical bush”
of Haddad et al. (2006)), since the tree stands had a partially
open canopy, with the terrain sloping gradually towards the
Nyamiti Pan. Fig. 4C show that the most data (133 individuals)
were taken in forest stands with only partially open canopy
(riverine bush at Nyamiti Hide was all partially open habitat),
and that all sand forest patches examined (Ndumo, Tembe, St.
Lucia) were closed canopy (33 data points). There was little
overlap of tree species (if known) between localities. However,
sand forest trees harboring spider webs had a smooth and
medium bark texture, while those in riverine bush had rough or
medium bark (Fig. 4D).

Figures 5A-F show the. relationships of interval and ratio
data by instar, and report the results of the non-parametric
comparisons (Mann-Whitney U test) of instar 2 (« = 118) and
7 (n — 18). Web height and web width increased significantly
with spider size (Figs. 5A, B , P < 0.001). These two
relationships support our assumption that spider web size
correlates with age (instar). No significant differences were
found in the web distance from the ground (Fig. 5C, P - 0.39)
among the spiders of different size. Contrary to our
prediction, larger spiders did not prefer larger trees (Fig. 5D,
P = 0.15). However, as predicted, both measures of web
allometry, i.e., the ladder index (Fig. 5E, P < 0.001) and the
hub displacement index (Fig. 5F, P < 0.001), increased with
spider size.

Figure 6 plots web orientation frequencies in eight catego-
ries (north, northeast, east, etc.) using circular statistics (Fisher
1993). Pooling all data (Fig. 6A) shows that C. irenae webs
were randomly distributed on all sides of trees. However, the
closed canopy data (Fig. 6B) reveal a significant (Rayleigh
test, P < 0.05) preference for the northern side of trees, and
the partially open canopy data (Fig. 6C) plots as significantly
bimodal (Rayleigh test, P < 0.05), the spiders preferring
southern and eastern faces of trees. Similarly, we found that
the webs on smooth bark (only found in closed canopy sand
forests), showed a significant preference towards the north.
However, the webs on medium bark (closed and partially open
forests), and on rough bark (only in partially open canopy
forest), showed random distributions. The northern orienta-
tion of webs under closed canopy is related to canopy closure,
not bark type, as both subsets of webs, the smooth bark webs
and medium bark webs showed a significant preference for the
northern side of trees (Rayleigh test, P < 0.05 and P < 0.1,
respectively).

KUNTNER ET AL. — ECOLOGY OF CLITAETRA I RENA E

589

Figure 4. — Field data summary (category, n, percentage): A. Individuals by instar; B. Habitat; C. Canopy coverage; D. Bark structure.

DISCUSSION

The disproportionately high numbers of small juveniles
(instar 2) measured compared to other instars (Fig. 4A) may
be explained by the fact that second instars tend to aggregate
around the mother’s web and are thus more easily located than
instars 3-6. One of the alternative explanations, the higher
mortality of larger juveniles, is contradicted by the fact that 18
females versus fewer 4-6th instar juveniles were collected
(Fig. 4A). Seasonality seems the best explanation: as our study
focused on the known adult C. irenae phenology, our April
field work only sampled a part of the species’ life cycle.

Distribution and phenology. — The ecoregion with the ma-
jority of C. irenae records (Fig. 3) is the Maputaland coastal
forest mosaic, but some records also fall into the adjacent
ecoregions, including the Southern Africa mangroves and the
KwaZulu-Cape coastal forest mosaic. The localities we visited
that fall into these latter two ecoregions were no different with
regards to forest structure and tree and bark microhabitat
from the ones falling into Maputaland, and thus we view these
records as continuous with “larger Maputaland.” At this
resolution, the borders between these ecoregions are fairly
arbitrary. It should be noted, however, that two of the three
South African localities visited where C. irenae was not found
are “Drakensberg montane grasslands, woodlands and
forests” (Fig. 3), whose forest microhabitat structure is quite
different from Maputaland. These habitats are further inland
and at higher altitudes than C. irenae typically inhabits,
perhaps indicating particular microclimatic, altitudinal and

habitat preferences. We predict that in the south the species
inhabits Maputaland coastal forests into Mozambique, but
that further north in tropical southern Africa, it continues
inland into the adjacent ecoregion (the Zambezi an and
Mopane woodlands) as far north as Malawi. Although no
specimen records currently exist from Mozambique and
Swaziland, we predict that the species occurs there. Our data
support the endemism hypothesis by 1) showing the continuity
of the Maputaland forest mosaic ecoregion into Mozambique,
and by 2) showing the ecoregion Zambezian and Mopane
woodlands’ adjacency to Maputaland. Where found, the C.
irenae adult abundances were greatest during the beginning of
our study (11-13 April 2006) while towards the end (28-29
April 2006) no further adults were found.

Habitat preference. — No C. irenae webs were measured in
open canopy stands (Fig. 4C), which is not to be interpreted to
mean that the spiders never occur there. In Maputaland C.
irenae inhabits most tree habitats, including lone trees in semi-
open canopy areas, and even synanthropic vertical surfaces,
but preferentially occupies partially open and closed canopy
stands. However, three of the four habitats where spiders and
their webs were investigated in detail are forests (Fig. 4B),
which is consistent with our assumption that forests and not
other types of tree stands are the species’ prime habitat.
Evidently, the species is not confined to sand forests, refuting
our hypothesis 2. We also conclude that C. irenae does not
exclusively inhabit closed canopy forests (Fig. 4C), and is
much more common in partially open stands (see below).

590

THE JOURNAL OF ARACHNOLOGY

A Web Height

p < 0.001

E

o

JQ

0

5

B Web Width

CM

CO

'sf

o

4 5

Instar

p < 0.001

6 7

3

Tree Size

E

CD

O

C

CD

CD

E

0

1

o

0

0

o

o

CO

o

o

CM

o

o

p= 0.148

6 7

Instar

Instar

E Ladder Index

p < 0.001

F Hub Displacement

p< 0.001

Figure 5. — Clitaetra irenae web parameters by spider size (instar number): A. Web height; B. Web width; C. Distance from ground; D. Tree
size preference; E. Ladder index; F. Hub displacement. Differences between stages 2 and 7 were tested (Mann- Whitney U test), shown as P values.

Many trees with C. irenae webs (n = 71) could not be
identified to species, but evidently C. irenae do not associate
with one or a few particular tree taxa, nor with trees of a
particular bark structure, refuting the hypotheses 3 and 5. This
is reflected in the diverse habitats in which the species was
encountered, each of which contains unique plant assem-
blages.

Our predictions that larger spiders chose larger, matured
trees, and lower portions of trees, ignored web allometry and
were based on the assumption that spider web size increases

with age. While the measured webs indeed increased with age
(Figs. 5 A, B) the fact that a variety of tree trunk sizes were
utilized by each instar (Fig. 5D) refutes our hypothesis 4.
Similarly, spiders of a certain age showed no preference for
website height above ground (Fig. 5C) refuting the hypothesis
7. This would indicate a fair degree of tolerance to
microclimatic variation with web height above ground
(varying between 15.5 cm and 337.0 cm), which are expected
to differ substantially from ground to canopy height.
Dispersal of second instars and subsequent web construction

KUNTNER ET AL.— ECOLOGY OF CLITAETRA IRENAE

B Closed Canopy Web Orientation

S

C Partially Open Canopy Web Orientation

N

Figure 6. — Clitaetra irenae web orientation using circular statistics
(Fisher 1993): A. All data (no orientation preference detected); B.
Closed canopy forest (significantly north); C. Partially open canopy
forest (significantly bimodal, south or east). Light versus dark grey
graph colors represent statistically insignificant versus significant
(Rayleigh test, P < 0.05) orientation.

appears to be in the immediate vicinity, and on the same tree,
as the mother’s web.

Web allometry. — The two indices quantifying the ontoge-
netic web changes from orb to ladder and the simultaneous
hub displacement towards the top frame (Figs. 5E, F) both
increase with spider size, which supports the prediction of

591

hypothesis 8. Such web allometry explains the lack of
correlation between spider age and tree size. The ladder index,
which is the relative web height, increases significantly with
age, which is consistent with our initial observations of
modified adult webs {Fig. 2). To a growing spider and its web
the limiting factor on a tree of a given size is the horizontal
website availability (tree circumference). The observed onto-
genetic web allometry allows the growing spider’s web to
increase vertically, while at the same time the corresponding
horizontal web increase is not required, allowing the spider to
remain on the same tree. The smallest tree on which an adult
female web was constructed (tree circumference 42.0 cm, web
width 6.5 cm) would suggest that the curvature of the tree
trunk may also play a role in the web site selection, making
available points of attachment on the bottom and top of the
web but eliminating the restriction of horizontal attachment
points on the side of the web. We see the ladder web as an
adaptation to an arboricolous life style because it eliminates
the need of a potentially costly and dangerous walk from one
(smaller) tree to another (larger) tree. Ants, which are
abundant on trees where C. irenae occur, may be important
predators (Kuntner 2006). However, ants are not known to
invade webs. Thus, orb-web spiders may be safer from
predators on (or close to) their webs than walking about.

The webs of all adult nephilid spiders have displaced hubs,
mostly towards the top (for rare horizontal displacement in
Nephilengys , see Kuntner 2007). Hubs are often displaced
above the web center in other araneoid spiders with vertical
webs (see Kuntner et al. 2008), notably in araneids. The logical
explanation for an ontogenetic shift from a central hub in
small juveniles towards the eccentricity seen in larger, heavier
spiders is gravity. Masters & Moffat (1983) demonstrated that
predation success of the araneid spider Larinioides sclopetarius
(Clerck 1757) improves in webs with hubs displaced above the
web center, as the time to reach the prey upwards and
downwards is thus optimized.

Ladder webs on tree trunks, present in the extant species of
the basal Cliiaetra-Herennia grade, were the ancestral life style
of the pan tropical clade Nephilidae and reversed to aerial orb
webs in the ancestor of Nephila (Kuntner et al. 2008). All
nephilids, C. irenae included, retain the non-sticky spiral
(NSS) in their finished web unlike most other orbweavers
(Kuntner 2006; Kuntner et al. 2008). NSS or auxiliary spiral
functions as a guide during the spider’s sticky spiral
construction (Zschokke 1993). Unlike in Nephila and Nephi-
lengys, NSS in Clitaetra webs is difficult to discern, perhaps
because these spirals are thin and may get stuck with the
narrowly meshed sticky ones. Kuntner (2006) determined the
NSS presence by observing the spiders build at night (and not
cutting the NSS when laying the sticky spiral), and the same
could not be determined by photographs of finished webs. The
retention of the NSS in nephilids has been suggested to have
evolved in response to female gigantism (Hormiga et al. 1995).
However, NSS seems to have been present in the nephilid
ancestor, where sexual dimorphism was only moderate and
not extreme (Kuntner et al. 2008). Perhaps the retention of
NSS was originally related to ladder web architecture, and its
presence in derived nephilids represents evolutionary time lag.
The convergent presence of NSS in Scoloderus webs (below)
may also be related to ladder web architecture.

592

THE JOURNAL OF ARACHNOLOGY

While ladder web architecture is homologous in Clitaetra
and Herennici (Kuntner et al. 2008), it is clearly not related to,
and differs functionally from, other known araneoid ladder
webs. For example, neotropical Scoloderus builds an extreme
ladder and, as in nephilids, retains the NSS (Eberhard 1975:
fig. 2), but this web is aerial, not arboricolous, and its hub is
displaced to the lower, not upper frame of the web, making the
spirals above the hub resemble a ladder. The hub of the
extreme ladder web of an unidentified araneoid from New
Guinea described by Robinson & Robinson (1972), however,
is displaced up as in nephilids, but that web is aerial and
apparently lacks the NSS. Furthermore, while nephilid ladder
webs are permanent structures (Kuntner 2005, 2006; Kuntner
et al. 2008), the above webs are taken down daily and rebuilt
every night (Robinson & Robinson 1972; Eberhard 1975).
Ladder webs apparently evolved convergently in araneids and
nephilids, perhaps in order to exploit new websites (trees) or
food resources. The first may be particularly true for
nephilids: while no study has focused on the prey of Clitaetra
and Herennia these spiders exploit tree trunks as websites and
Herennia even evolved a unique web design using pseudoradii
(Kuntner 2005; Kuntner et al. 2008). Eberhard (1975) pointed
out that a ladder web architecture allows for a more constant
mesh size than any circular orbweb, which may be related to
specialization for certain prey types. Aerial ladder webs have
been suggested to represent convergent adaptation for
ensnaring moths (Eberhard 1975; Stowe 1978). However, prey
capture in tropical spiders is often anecdotal. While Stowe
(1978) showed that 68% of 212 prey items of Scoloderus were
moths, Robinson & Robinson (1972) observed a single prey
item, a moth, attracted into the ladder web of their unknown
spider by a light source. Although the focus of our study was
not a quantification of C. irenae prey, the few sporadically
observed insect prey items in their webs (Fig. 1A-D:
Orthoptera, Diptera, Lepidoptera (Pyralidae), Homoptera)
suggest that the species is an opportunistic predator not
particularly specialized on moths.

Web orientation. — Superficially, C. irenae webs appear
randomly oriented (Fig. 6A). However, preferential spider
web orientation occurs in both closed and partially open
canopy forests. In closed canopy forests (Fig. 6B) the spiders
show a preference for the northern side of the tree, refuting the
first part of the web orientation hypothesis (that no
orientation preference would be shown in closed canopy
forests). Two types of bark that were found under closed
canopy (smooth, medium) both show a significant orientation
to the north, which indicates that northern orientation of webs
under closed canopy is related to canopy closure, not bark
type. Conversely, in partially open canopy forests (Fig. 6C)
the spiders preferentially chose southern and eastern faces
of trees, which is supportive of the second part of our hypo-
thesis 6.

Web orientation probably affects web microclimate. Pro-
longed direct sun exposure could harm the spider or its web, or
may affect prey availability and/or predation pressure. In the
southern hemisphere, the predicted preference for the south-
ern, shady side of trees makes sense in a partially open forest.
It is somewhat surprising, however, that the spiders inhabiting
closed canopy forests seem to prefer a northern (perhaps
warmer) orientation, where we predicted randomness, though

this result does not, per se, refute hypothesis 6. Since we
mainly scored smooth barked trees in sand forests (all closed
canopy) it is not surprising that the web orientation patterns
on smooth bark trees resemble those of closed canopy forests
(Fig. 6B). The preference for the southern and eastern sides of
trees in partially open canopy habitats (Fig. 6C) would suggest
an aversion for direct sunlight and a preference for darker
sides of trees.

Implications for Maputaland ecology. — Previous Maputa-
land studies have focused on organisms such as dung beetles
and birds to assess community heterogeneity and the impacts
of habitat destruction and regeneration on faunal and floral
components (Van Rensburg et al. 1999, 2000; Davis et al.
2002; Wassenaar et al. 2005). However, Maputaland arachnids
are highly diverse, with 457 species recorded from the Ndumo
Game Reserve, 10.1 12-ha in size (Haddad et al. 2006). Given
such diversity it is likely that several species are endemic to
Maputaland and could be used as indicators of changes and
disturbances to habitats, including sand forest.

Spiders are generally more sensitive than other arthropod
groups to the vegetative structural conditions in a habitat,
particularly web-builders (e.g., Marc et al. 1999; Stiles & Coyle
2001; Finch 2005). In monitoring or habitat evaluation using
spiders as bio-indicators, the absence of species typical for a
habitat (stenotypic species) is often indicative of low habitat
quality, including vegetation structure (Bonte & Maelfait
2001). Thus, the identification of a sensitive species in the
Maputaland fauna could provide an indication of the current
condition of forest patches and the extent of disturbance to
which the local fauna is presently being exposed. Once
identified, indicator spider species can also be used for long-
term monitoring of landscapes (Bonte et al. 2002), which will
enable conservationists to assess changes in the condition of
these forests over time. A particularly important current
conservation issue in South Africa is the impact of heavy
utilization of sand forests by elephants in the Tembe Elephant
Park, Maputaland. This is thought to lead to the irreversible
opening up of the sand forest into a structure comparable to
mixed woodland (W.S. Matthews pers. comm.), which is likely
to have a strongly negative long-term effect on the plant and
animal communities, and on the diversity of the sand forest.
Additionally, the growing rural human population in Maputa-
land is putting increasing pressure on sand forest patches
outside conservancies (Kirkwood & Midgley 1999).

In Africa, particularly, forests are opened up by big game,
and such forest gaps are not necessarily indicative of unnatural
disturbance. Indeed, in the reserve where elephants occur
(Tembe), partially open canopy patches seem to be continuous
with closed canopy sand forests. Thus, our finding of most
individuals of C. irenae in partially open canopy forest stands
does not refute our hypothesis about C. irenae dependence on
the undisturbed Maputaland forest habitats. This could be an
artifact of our under-sampling of closed canopy locations.
Assuming that indigenous, undisturbed Maputaland forests
are closed canopy, and thus the original preference of C. irenae
spiders in a quality habitat is towards the north, scoring web
orientation might be indicative of forest quality/disturbance.
The easy C. irenae identification (Fig. 1, see Kuntner 2006)
warrants further investigation as to whether it might be a
suitable bioindicator.

KUNTNER ET AL. ECOLOGY OF CLITAETRA IRENAE

Conclusions. — The data at hand suggest the ecological and
behavioral dependence of C. irenae on the threatened
Maputaland forests. The wider Maputaland endemism hy-
pothesis receives support, but the hypotheses that C. irenae
inhabits exclusively sand forests, mature trees, trees of a
particular species, trees with a smooth bark, tree habitats at
certain height above ground, and only closed canopy forest
stands, are refuted. Evidently the species’ ecological niche is
flexible to an extent but requires suitable tree habitat under at
least partially closed canopy. However, the web orientation on
trees appears to be indicative of closed versus partially open
canopy forest.

Conservationists may benefit from utilizing the available
arthropod data in assessing the quality of tropical forests. The
ecology of obligate arboricolous orb-weaving spiders (like the
nephilids Clitaetra and Herennia), seems especially well suited
for systematic conservation assessments in the (sub)tropics
because they range from western Africa (Kuntner 2006)
through South and Southeast Asia into Australasia (Kuntner

2005) where some species are narrow endemics and others
appear to be widespread and invasive (Kuntner 2005, 2006).
The "Africa + Asia + Australasia” tropical belt matches the
maps of global biodiversity conservation priorities (Brooks et
al. 2006), but also lies precisely in a zone of high population
pressure and low human development index (see Jha & Bawa

2006) , a detrimental combination of factors associated with
heavy deforestation.

ACKNOWLEDGMENTS

We thank Jonathan Coddington, Ingi Agnarsson, Jeremy
Miller, Wayne Matthews, Tatjana Celik, and Simona Kralj-
Fiser for discussions and detailed comments on an early draft,
Marjan Jarnjak for producing the species distribution map,
Douglas C. Daly (New York Botanical Garden, Bronx) for
kindly authorizing the use of the base map of Bletter et al. as
our GIS source, and Ansie Dippenaar-Schoeman, Xander
Combrink, Sharron Hughes, and Irena Kuntner for help,
assistance, or advice. The comments of S. Toft, S. Zschokke
and an anonymous reviewer further improved our paper. This
research, made possible through a Ezemvelo KZN Wildlife
collecting and export permit no. 1010/2006, was supported by
the Slovenian Research Agency (grant Z 1-7082-06 18 to MK)
and the EU 6th Framework Programme (Marie Curie grant
MIRG-CT-2Q05-036536 to MK).

LITERATURE CITED

ap Rhisiart, A. & F. Vollrath. 1994. Design features of the orb web of
the spider, Araneus diadematus. Behavioral Ecology 5:280-287.
Bleher, B. 2000. Development of web-building and spinning
apparatus in the early ontogeny of Nephihi madagascariensis
(Vinson, 1863) (Araneae: Tetragnathidae). Bulletin of the British
Arachnological Society 11:275-283.

Bletter, N., J. Janovec, B. Brosi & D.C. Daly. 2004. A digital base
map for studying the Neotropical Bora. Taxon 53:469-477.

Bonte, D., L. Baert & J.-P. Maelfait. 2002. Spider assemblage
structure and stability in a heterogeneous coastal dune system
(Belgium). Journal of Arachnology 30:331-343.

Bonte, D. & J.-P. Maelfait. 2001. Life history, habitat use and
dispersal of a dune wolf spider (Pardosa monticola [Clerck, 1757]
Lycosidae, Araneae) in the Flemish coastal dunes (Belgium).
Belgian Journal of Zoology 131:145-157.

593

Brooks, T.M., R.A. Mittermeier, G.A.B. da Fonseca, J. Gerlach. M.
Hoffmann, J.F. Lamoreux, C.G. Mittermeier, J.D. Pilgrim &
A.S.L. Rodrigues. 2006. Global biodiversity conservation priori-
ties. Science 313:58-61.

Coddington, J.A. 2005. Phylogeny and classification of spiders.
Pp. 18-24. In Spiders of North America: an Identification Manual.
(D. Ubick, P. Paquin, P.E. Cushing & V. Roth, eds.). American
Arachnological Society.

Coddington, J.A. & H.W. Levi. 1991. Systematics and evolution of
spiders (Araneae). Annual Review of Ecology and Systematics
22:565-592.

Davis, A.L.V., R.J. Van Aarde, C.H. Scholtz & J.H. Delport. 2002.
Increasing representation of localized dung beetles across a
chronosequence of regenerating vegetation and natural dune forest
in South Africa. Global Ecology and Biogeography 11:191-209.

De Moor, P.P., E. Pooley, G. Neville & J. Barichievy. 1977. The
vegetation of Ndumo Game Reserve, Natal: a quantitative
physiognomic survey. Annals of the Natal Museum 23:239-272.

Eberhard, W.G. 1975. The ‘inverted ladder’ orb web of Scoloderus sp.
and the intermediate orb of Eustala (?) sp. Araneae: Araneidae.
Journal of Natural History 9:93-106.

Finch, O.-D. 2005. Evaluation of mature conifer plantations of
secondary habitat for epigeic forest arthropods (Coleo-
ptera: Carabidae; Araneae). Forest Ecology and Management 204:
21-34.

Fisher, N.I. 1993. Statistical Analysis of Circular data. Cambridge
University Press, Cambridge, UK. 298 pp.

Haddad, C.R., A.S. Dippenaar-Schoeman & W. Wesolowska. 2006.
A checklist of the non-acarine arachnids (Chelicerata: Arachnida)
of the Ndumo Game Reserve, Maputaland, South Africa. Koedoe
49:1-22.

Higgins, L.E. 2002. Female gigantism in a New Guinea population of
the spider Nephila maculata. Oikos 99:377-385.

Hormiga, G., W.G. Eberhard & J.A. Coddington. 1995. Web-
construction behavior in Australian Phonognatha and the phylog-
eny of nephiline and tetragnathid spiders (Araneae: Tetragnathi-
dae). Australian Journal of Zoology 43:313-364.

Japyassu, H.F. & C. Ades. 1998. From complete orb to semi-orb
webs: developmental transitions in the web of Nephilengys
cruentata (Araneae: Tetragnathidae). Behaviour 135:931-956.

Jha, S. & K.S. Bawa. 2006. Population growth, human development,
and deforestation in biodiversity hotspots. Conservation Biology
20:906-912.

Kirkwood, D. & J.J. Midgley. 1999. The floristics of sand forest in
northern KwaZulu-Natal, South Africa. Bothalia 29:293-304.

Knight, A.T., A. Driver, R.M. Cowling, K. Maze, P.G. Desmet, A.T.
Lombard, M. Rouget, M.A. Botha, A.F. Boshoff, J.G. Castley,
P.S. Goodman, K. Mackinnon, S.M. Pierce, R. Sims-Castley, W.I.
Stewart & A. Von Hase. 2006. Designing systematic conservation
assessments that promote effective implementation: best practice
from South Africa. Conservation Biology 20:739-750.

Kuntner, M. 2005. A revision of Herennia (Araneae: Nephilidae:
Nephilinae), the Australasian ‘coin spiders.' Invertebrate System-
atics 19:391-436.

Kuntner, M. 2006. Phylogenetic systematics of the Gondwanan
nephilid spider lineage Clitaetrinae (Araneae, Nephilidae). Zoolo-
gica Scripta 35:19-62.

Kuntner, M. 2007. A monograph of Nephilengys, the pantropical
‘hermit spiders’ (Araneae, Nephilidae, Nephilinae). Systematic
Entomology 32:95-135.

Kuntner, M„ J.A. Coddington & G. Hormiga. 2008. Phylogeny of
extant nephilid orb-weaving spiders (Araneae, Nephilidae): testing
morphological and ethological homologies. Cladistics 24:147-217.

Marc, P., A. Canard & F. Ysnel. 1999. Spiders (Araneae) useful for
pest limitation and bioindication. Agriculture, Ecosystems and
Environment 74:229-273.

594

THE JOURNAL OF ARACHNOLOGY

Masters, W.M. & A.J.M. Moffat. 1983. A functional explanation of
top-bottom asymmetry in vertical orbwebs. Animal Behaviour
31:1043-1046.

Matthews, W.S., A.E. Van Wyk & N. Van Rooyen. 1999. Vegetation
of the Sileza Nature Reserve and neighbouring areas, South Africa,
and its importance in conserving the woody grasslands of the
Maputaland Centre of Endemism. Bothalia 29:151-167.

Matthews, W.S.. A.E. Van Wyk, N. Van Rooyen & G.A. Botha.
2001. Vegetation of the Ternbe Elephant Park, Maputaland, South
Africa. South African Journal of Botany 67:573-594.

Olson, D.M., E. Dinerstein, E.D. Wikramanayake, N.D. Burgess,
G.V.N. Powell, E.C. Underwood. J.A. D’Amico, I. Itoua, H.E.
Strand, J.C. Morrison, C.J. Loucks, T.F. Allnutt, T.H. Ricketts,
Y. Kura, J.F. Lamoreux, W.W. Wettengel, P. Hedao & K.R.
Kassem. 2001. Terrestrial ecoregions of the World: a new map of
life on Earth. BioScience 51:933-938.

Platnick, N.I. 2007. The World Spider Catalog, Version 8.0.
American Museum of Natural History, New York. Online at
http://research.amnh.org/entomology/spiders/catalog/.

R Development Core Team. 2006. R: A language and environment
for statistical computing. Version 2.4.0 R Foundation for
Statistical Computing, Vienna, Austria. Online at http://www.
R-project.org.

Robinson, M.H. & B. Robinson. 1972. The structure, possible
function and origin of the remarkable ladder-web built by a New-
Guinea orbweb spider (Araneae: Araneidae). Journal of Natural
History 6:687-694.

Stiles, G.J. & F.A. Coyle. 2001. Habitat distribution and life history
of species in the spider genera Theridion , Rugathoides , and Wamba
in the Great Smoky Mountains National Park (Araneae,
Theridiidae). Journal of Arachnology 29:396^112.

Stowe, M.K. 1978. Observations of two nocturnal orbweavers that
build specialized webs: Scoloderus cordatus and Wixia ectypa
(Araneae: Araneidae). Journal of Arachnology 6:141-146.

Tukey, J.W. 1977. Exploratory Data Analysis. Addison-Wesley
Publishing, Reading, Massachusetts. 688 pp.

Van Rensburg, B.J., S.L. Chown, A.S. Van Jaarsveld & M.A.
McGeogh. 2000. Spatial variation and biogegraphy of sand forest
avian assemblages in South Africa. Journal of Biogeography
27:1385-1401.

Van Rensburg, B.J., M.A. McGeogh, S.L. Chown & A.S. Van
Jaarsveld. 1999. Conservation of heterogeneity among dung beetles
in the Maputaland Centre of Endemism, South Africa. Biological
Conservation 88:145-153.

Van Wyk, A.E. 1994. Maputaland-Pondoland region. Pp. 227-235. In
Centres of Plant Biversity: A Guide and Strategy for Their
Conservation. (S.D. Davis, V.H. Heywood & A.C. Hamilton,
eds.). Oxford University Press. Oxford, UK.

Van Wyk, A.E. 1996. Biodiversity of the Maputaland Centre.
Pp. 198-207. In The Biodiversity in African Savannahs. (L.J.G.
Van der Maesen, X.M. Van der Burgt & J.M. Van Medenbach de
Rooy, eds.). Kluwer Academic Publishers, Dordrecht.

Van Wyk. A.E. & G.F. Smith. 2001 . Regions of floristic endemism in
Southern Africa: a review with emphasis on succulents. Urndaus
Press, Hatfield, South Africa. 160 pp.

Wassenaar, T.D., R.J. Van Aarde, S.L. Pimm & S.M. Ferreira. 2005.
Community convergence in disturbed subtropical dune forests.
Ecology 86:655-666.

Zschokke, S. 1993. The influence of the auxiliary spiral on the capture
spiral in Araneus diadematus Clerck (Araneidae). Bulletin of the
British Arachnological Society 9:169-173.

Manuscript received 14 July 2007, revised 28 January 2008.

2008. The Journal of Arachnology 36:595-599

Differential survival of Geolycosa xera archboldi and G. hubhelli (Araneae, Lycosidae) after fire in

Florida scrub

James E. Carrel1: Division of Biological Sciences, 209 Tucker Hall, University of Missouri-Columbia, Columbia,
Missouri 6521 1-7400, USA. E-mail: carrelj@missouri.edu

Abstract. A replicated pre- and post-burn study of survival of small and large Geolycosa xera archboldi McCrone 1963
and G. hubhelli Wallace 1942 in Florida scrub was conducted. These two syntopic species were chosen because G. x.
archboldi prefers large gaps of barren sand in the scrub matrix, sites with little fuel for fires, whereas G. hubhelli strongly
favors small gaps having some leaf litter, sites with modest or high fuel-loads. On the basis of these species-specific
differences in microsite characteristics, I hypothesized that G. x. archboldi would be very fire tolerant but that G. hubbelli
would be fire intolerant. I established two size classes for the Geolycosa : small spiders had 3-5 mm diameter X 5-9 cm deep
burrows; large spiders had > 6 mm diameter X 10-17 cm deep burrows. Burrows of 25 spiders in each species X size class
were marked before a burn in seven burn units ( = fire management areas) and survival or mortality of each occupant was
ascertained over the course of 5 days post-burn. Thus, the experimental design was 2 species X 2 size classes X 7 burn units
X 25 replicates/burn unit (n = 700 spiders total). Survivorship was very high in small and large G. x. archboldi and in large
G. hubbelli (93-96%), but it was low in small G. hubbelli (35%). Temperature recordings suggest mortality in small G.
hubbelli was caused by high temperatures at depths of 5-10 cm during intense, but brief burns that characterize fires in
Florida scrub. In contrast, large G. hubbelli had burrows sufficiently deep so that most of them did not experience lethal
temperatures during bums.

Keywords: Burrowing wolf spiders, endemism. Lake Wales Ridge, body size, fire ecology

Florida scrub is a fire-prone ecosystem confined to ancient
sand ridges in the peninsular part of the state. This ecosystem
also supports biotic communities that comprise a globally
important, imperiled center of endemism (Deyr up 1989;
Deyrup & Eisner 1993; Dobson et al. 1997; Menges 1999;
Marshall et al. 2000; Estill & Cruzan 2001; Weekley et al.
2008). Presumably, as part of the suite of characters needed to
survive in scrub, endemic species have evolved adaptations to
frequent landscape-level burns that rapidly consume the leaf
litter and standing vegetation. For example, the dominant
woody shrubs have most of their biomass below ground, so
they survive and quickly regenerate the shrub matrix by
sprouting. In contrast, most endemic herbs are killed by fire
and post-burn increases in abundance are due to seedling
recruitment (Weekley & Menges 2003, and references therein).
Scrub animals have three common methods for coping with
fire at a landscape scale. On the one hand, some such as sand
skinks ( Plestiodon reynoldsi), gopher tortoises ( Gopherus
polyphemus), and flightless pygmy mole crickets ( Neotridacty -
lus archboldi), persist in place by exploiting a subterranean life
style in the sandy soils (Robbins & Myers 1992; Deyrup 2005).
On the other hand, the Florida scrub jay ( Aphelocoma
coerulescens) and other highly dispersive animals flee the
oncoming flames on wing or foot and settle in unburned scrub
(Robbins & Myers 1992). A third approach, one used by
weak-flying insects and arboreal spiders, such as the red
widow spider ( Latrodectus bishopi Kaston 1938), is to
experience high mortality during a bum and to recolonize
subsequently from nearby, unburned refugia (Deyrup &
Eisner 1996; Carrel 2001, 2008).

Two species of rare burrowing wolf spiders, Geolycosa xera
archboldi McCrone 1963 and G. hubbelli Wallace 1942, are

'Current address: Archbold Biological Station, 123 Main Drive,
Venus, Florida 33960, USA.

endemic to oak scrub on the Lake Wales Ridge in the middle
of peninsular Florida (Marshall et al. 2000). Because the
spiders spend most of their lives below ground in tubular
burrows they construct in the sand, I expected that they might
be fire tolerant, similar to other subterranean animals. But
knowing that small individuals build much shallower burrows
than larger, older individuals (Table 1 and Figure 1), I
hypothesized that survival of a burn in Geolycosa might be
size dependent because smaller spiders build more shallow
burrows than larger spiders and. as a result, small spiders
could be more exposed to lethal temperatures that penetrate
the upper layer of soil when scrub is burned. In addition,
because G. x. archboldi prefers large (> 1 nr), barren gaps of
sand and does not decorate its burrow entrance with a turret,
whereas G. hubbelli favors small gaps (~ 0.1 nr) in the
shrubby matrix having leaf litter from which it obligatorily
builds a turret (Carrel 2003a, b), I also hypothesized that the
latter species might be more likely to perish in a fire. To test
these ideas, I conducted a pre- and post-fire study of survival
(or mortality) of individual G. x. archboldi and G. hubbelli in
two size classes (small and large individuals, Tables 1 and 2)
over the course of several burn events in Florida scrub. I also
collected ambient temperate data in Geolycosa burrows and on
the soil surface during a fire. To my knowledge this is the first
replicated, quantitative study of survivorship in any spider
exposed to burning of its habitat, and it may be one of the few
such studies with any terrestrial arthropod to date (Warren et
al. 1987; Whelan 1995; Siemann et al. 1997; Swengel 2001).

METHODS

Study site. — I conducted a pre- and post-fire study of
Geolycosa survival in flat, oak scrub at the 2101 ha Archbold
Biological Station, in southern Highlands County, Florida
(elev. 36-46 m, 27°H'N, 81°21'W). The work was performed
in the most extensive vegetative association, called scrubby

595

596

THE JOURNAL OF ARACHNOLOGY

Table 1.- -Depth and volume of burrows constructed by small and
large Geolycosa spiders. Typical data were calculated using best-fit
regression equations published by Carrel (2003a).

Burrow size class

(diameter, mm)

Burrow

Small

Large

Spider species

properties

(3-5)

(6-15)

G. xera archboldi

Depth (cm)

4. 6-8. 2

10.0-15.4

Volume (cc)

0.4-2. 3

4.5-19.7

G. hubbelli

Depth (cm)

5.6-9. 1

10.3-16.6

Volume (cc)

0.7-2. 8

4.5-49.0

flatwoods, which has fire-resistant slash pines (Pinus elliotti)
scattered in a dense matrix dominated by low-growing
shrubby oaks ( Quercus inopina , Q. chapmanii, and Q.
geminate/), palmettos ( Serenoa repens and Saba I etonia,
Arecaceae), and shrubby lyonias ( Lyonia ferruginea , L.
fruticosa, and L. lucida, Ericaceae) (Abrahamson et al.
1984). For management purposes, the scrub at Archbold is
organized into a series of 187 burn units and a detailed history
of burning in each unit is available (Main & Menges 1997;
unpublished Archbold records). 1 was able to work in seven
units, ranging in size from 4.6 to 66.5 ha, two of which were
burned in February 2001, one in October 2002, two in July
2007, and two in August 2007. Voucher specimens of both
Geolycosa species were deposited in the Invertebrate Collec-
tion at Archbold.

Experimental design. —I haphazardly located 25 small (3-
5 mm diam.) and 25 large (6-15 mm diam.) burrows of both

Geolycosa xera archboldi

Geolycosa hubbelli

Figure 1. Silhouettes of small and large burrows of Geolycosa
xera archboldi and G. Iiubbelli prepared from representative plaster
casts (Carrel 2003a). Note interspecific difference in the architecture
of burrow bases.

Table 2. — Attributes of two Geolycosa species placed into two size
classes (small and large) based on diameter of their burrow openings,
for study of survivorship after fire. Typical data were calculated using
best-fit regression equations (Carrel 2003a). Sample size in this study
(«) for each species X size class is also given.

Spider
size class

G. x. archboldi
(turret absent)

G. hubbelli
(turret present)

Small

Burrow diameter (mm)

3-5

3-5

Carapace width (mm)

1. 4-2.3

1. 5-2.1

Body mass (mg)

8-30

8-17

Sample size (n)

175

175

Large

Burrow diameter (mm)

6-10

6-15

Carapace width (mm)

2. 7-4.3

2.4-5. 1

Body mass (mg)

40-230

25-600

Sample size (n)

175

175

Geolycosa species by visually searching in seven different
burn units 1-2 days before each was burned. Burrows were
> 10 m from the perimeter of a burn unit to avoid edge
effects, particularly kerosene-induced flames from drip
torches used to ignite the leaf litter and vegetation. In
previous studies (Carrel 2003a) I showed that the persistently
open, circular burrows render these spiders very detectable:
by conducting a rapid, but thorough visual search of an area
(10-100 nr), one typically locates 90-95% of individuals
actually present. Furthermore, the presence or absence of a
turret constructed from leaves and debris, held in place with
silk around the burrow opening, is a reliable tool for telling
the species apart (Carrel 2003a). In addition, burrow
diameter, as measured with calipers to the nearest 0.1 mm,
is a highly reliable surrogate for the size of the occupying
spider as well as the depth and volume of its burrow
(Tables 1 and 2).

Before a burn, I marked the location of each spider burrow
(n = 700 total) by placing two thin metal stakes vertically in
the sand —10 cm on opposite sides of the burrow entrance.
Following a burn, I revisited each burrow for 5 consecutive
days and determined if the resident spider was alive. I used
four criteria for survivorship: sighting of a spider sitting near
the top of its burrow; luring a spider from the burrow by the
presence of insect prey that I tethered on a thread near the
entrance; restoration of a damaged burrow entrance or turret;
and placement of newly excavated sand on the ground near a
burrow. If all these criteria were negative, on the fifth or sixth
day post-fire I carefully excavated a spider’s burrow looking
for its body. In so doing I could confirm that the burrow was
occupied by a spider and, based on the soft, decomposing
condition of a corpse, that the resident individual perished
during or shortly after the blaze.

Air and soil temperature measurements.- I used Hobo U-12
digital dataloggers (Onset Computer Corporation, Pocasset,
Massachusetts) fitted with Type K thermocouples to record
air and soil temperatures in the scrub, following the
established protocols of Wally et al. (2006). After calibrating
each machine, I programmed the dataloggers in the laboratory
to record one reading per second and to output maximum
temperatures at 1 min intervals prior to deployment in the
scrub. I obtained two sets of temperature data: maximum
daily temperatures inside G. x. archboldi burrows and nearby

CARREL— SURVIVAL OL GEOLYCOSA AFTER FIRE

597

in undisturbed soil on hot, sunny days; and the intensity and
duration of fire at point sites on the soil surface in oak scrub in
order to gain a better perspective of the thermal dynamics
experienced by subterranean spiders.

The first set of temperature recordings was designed to
determine whether the open burrows of small and large G. x.
archboldi under typical summer daytime conditions were
significantly warmer than intact soil at comparable depths in
the scrub. I chose to study only G. x. archboldi because this
species occurs predominantly in large, barren gaps of
unshaded sand where solar heating is the most intense in
scrub. In contrast, G. hubbelli is typically found in small gaps
with leaf litter on the sand, so its burrows are insulated from
solar heating both by the leaf litter and by shade cast by the
surrounding shrub matrix. Thus, my reasoning was that if
maximum daytime temperatures in open G. x. archboldi
burrows were comparable to those in undisturbed soil at
comparable depths, then a similar burrow/soil equivalency
probably would hold for G. hubbelli (even though the maxima
obviously would be smaller). (Subsequent measurements
showed this relationship was valid, JEC unpubl. data.) Over
the course of 3 weeks in late August-early September 2007 I
simultaneously set up ten replicate sets for 1 day each with
thermocouples in five different positions: at 0, 5 cm, and
10 cm depth in intact sand and at the bottom of small (3-
5 mm diam. X 3. 5-5. 2 cm depth) and large (6-12 mm diam. X
10.5-14.3 cm depth) G. x. archboldi burrows after the resident
spiders were removed. Maximum daily air temperatures at
1.5 m above ground were also obtained at the official
Archbold weather station on the days that soil temperatures
were recorded.

Secondly, in an attempt to characterize the intensity and
duration of a fire in oak scrub, I acquired data on soil surface
temperatures during a “category 3” burn in August 2007 from
the plant ecology group at Archbold. (“Category 3”, the
highest intensity in the classification scheme used by Archbold
staff, means that most surface litter was consumed, all leaves
of palmettos and shrubs 0-2 m elevation were completely
consumed, and small twigs on shrubs were consumed in a
blaze.) Following their published protocol (Wally et al. 2006),
many thermocouples attached to datalogggers were placed in
contact with the soil surface at a variety of locations to record
soil surface temperatures during a burn event. Using data
from ten dataloggers in sites that experienced heavy burns. I
normalized the temporal records so that the peak maximum
temperatures all occurred at the 10 min mark, so that there
would be several min of pre-burn data as well as > 30 min
post-maximum peak data. By definition, ignition threshold is
> 60° C and cessation of fire is set at < 60° C; the 60° C
benchmark is used because it corresponds to the temperature
at which plant cell death occurs (Wally et al. 2006, and
references therein).

Statistical analyses. — I used the General Linear Models
program of SPSS to perform ANOVA to evaluate the
significance of variables in the sets of data on spider survival
(SPSS 2005). The Levene test statistic was calculated to
confirm that the variance did not differ significantly between
the groups (P > 0.05). Differences in post-burn survivorship
of spiders were analyzed by Chi square tests with Yates
correction for small sample size (X2C, Simpson et al. 1960).

Table 3. — Effect of burn event, species identity and body size of
spiders (as measured by burrow diameter) on post-burn survival of
two Geolycosa species in Florida scrub.

Source of

variation

df

MS

F

P

Burn event

6

0.178

1.966

0.068

Species

1

14.573

160.9

< 0.001

Size of spider

1

18.241

201.4

< 0.001

Species X size

1

15.156

167.3

< 0.001

Error

672

0.091

I calculated the average (mean ± SE, n — 10) and range of
maximum daily temperatures at all five locations in soil and in
the air. 1 used the General Linear Models program of SPSS to
perform univariate ANOVA to evaluate the significance of
location in data on soil temperatures. The Levene test statistic
was calculated to confirm that the variance did not differ
significantly between the groups ( P > 0.05). Subsequently I
performed two post hoc multiple range tests (Student-New-
man-Keuls (SNK) and Tukey HSD) to determine in a pair-
wise fashion which locations had significantly different
temperatures (P set at 0.05) (SPSS 2005).

After normalizing the soil surface temperature data during
one burn event so that temperatures peaked at all locations {n
= 10) in the 10th minute, I calculated the minimum, mean,
and maximum temperature minute by minute for 30 min.

RESULTS

Post-burn survival of Geolycosa species. — Spider species,
spider size, and spider species X spider size interaction were all
highly significant variables determining the post-burn survival
of Geolycosa species (Table 3). This meant that there was a
complex interaction between spider species identity and spider
size that required further analysis. Fortunately, as there were
no significant differences among the seven burns according to
the AVOVA results (Table 3), I was able to combine the data
and delete “burn event” as a variable, which greatly simplified
further analyses. As shown in Table 4, few small G. hubbelli
(35.4%) survived the burns. In contrast, I found almost all
large G. hubbelli (94.5%) and almost all G. xera archboldi
regardless of size (small = 93.1%, large = 96.0%) were alive
5 days post-burn in the scrub. The intraspecific, size-depen-
dent difference in survivorship for G. hubbelli was highly
significant (X2C = 133.49, df — 1, P < 0.0001).

Maximum daily temperatures in G. x. archboldi burrows.
On ten sunny days in late summer 2007, maximum air
temperatures at the Archbold weather station were hot,
averaging 34.6 ± 0.3° C (mean ± SE, range 33.3-36.1 C),

Table 4. — Survivorship of Geolycosa spiders as a function of
burrow/body size and species identification. Results of statistical
analyses (Chi square test with Yates correction for small sample size,
X2C) for intraspecific size-based differences in survival are given.

% Surviving

Species

Small
(» = 175)

Large
(n = 175)

X2 c

P

G. xera archboldi

93.1

96.0

0.89

NS

G. hubbelli

35.4

94.5

133.49

< 0.0001

598

THE JOURNAL OF ARACHNOLOGY

Time (min)

Figure 2.- -Intensity and duration of a fire at a point source in
Florida scrub. Note rapid onset and rise to peak temperature at soil
surface, followed by somewhat less rapid decline. See methods
for details.

but the maximum daily temperature on the surface of fully
exposed sand in scrub was much greater: nearly 16° C hotter
(51.2 ± 0.9, 47.8-56.4 C). Univariate ANOVA showed there
was a significant difference among the soil temperature data
by location (F445 = 3.599, P = 0.013). Post hoc analysis
revealed that, despite intense solar heating, the maximum
daily temperatures at the bottoms of spider burrows and
down in undisturbed soil remained significantly lower than at
the surface (SNK and Tukey HSD tests, P < 0.05). Small,
shallow spider burrows got as warm as soil at 5 cm depth
(burrow: 38.5 ± 0.4, 36.7-40.3° C; soil: 37.8 ± 0.5, 35.7-
40.0° C; P = 0.44). In addition, large, deep spider burrows
stayed even cooler (P < 0.05) than shallow ones during the
day and their maximum daily temperatures were the
equivalent to those in soil at 10 cm depth (burrow: 33.3 ±
0.4, 31.0-36.0° C; soil: 34.1 ± 0.4, 32.7-36.0° C; P = 0.28).
Hence, despite the fact that the spiders’ burrows remained
constantly open, the most extreme thermal climate experi-
enced by resident animals if they were deep in the burrows
would be virtually the same as if they were buried in
undisturbed soil at a comparable depth, far less than that at
the burrow entrance.

Soil surface temperature during a burn. The time course of
a burn in the scrub at any point in the burn unit was
remarkably rapid. As shown in Fig. 2, the fire went from
ignition temperature (60° C by definition) to peak maximum
soil temperature (609-846 C) in < 2 min, then it declined to
~ 60 C in another 7 min. Hence, from the perspective of a
Geolycosa hiding in its burrow, the fire lasted < 10 min.

DISCUSSION

Mortality in G. hubbelli. The results were generally in
agreement with my initial hypotheses with one exception: the
post-burn survivorship of large G. hubbelli was much greater
than expected. In fact, to my surprise, it matched that for
small or large G. x. archboldi (93-96%). 1 suspect burrow
architecture makes large G. hubbelli very fire tolerant. As G.
hubbelli grow they construct burrows that are not only wider
in diameter and deeper, but also they excavate increasingly
large, ovoid chambers at the bottoms (Table 1 and Fig. 1).

Such bulbous refugia > 10 cm below the surface evidently
protect large G. hubbelli from the brief but intense fires in the
leaf litter and shrubbery above them, probably because the
intense heat fails to penetrate to this depth.

I think the cause of mortality in small G. hubbelli is not
fire-induced asphyxiation. Under natural conditions in sandy
Florida soils, extensive measurements of prevailing gases in
— 16 cm deep burrows occupied by a closely related
burrowing wolf spider, G. micanopy Wallace 1942, showed
no significant increases in C02 or decreases in 02 concen-
trations relative to ambient atmospheric values (Anderson &
Ultsch 1997). Thus, during a fire in Florida scrub, I doubt
there would be extensive depletion of oxygen down in the
spiders' porous burrows. Moreover, detailed physiological
studies by Prestwich (1983a, b; 1988a, b) have demonstrated
that active Florida spiders rely mostly on anaerobic
metabolism because nearly all of their tissue phosphagen is
quickly (within 10-15 s) depleted after onset of activity.
Hence, a 10-min period of hypoxia during a fire in Florida
scrub should, at best, present Geolycosa spiders at rest in
their burrows only with a mild respiratory challenge.

I suspect the primary cause of fire-induced mortality in
small G. hubbelli is high temperature in surficial soils and
burrows. Field measurements show that soil temperatures at
2-3 cm depth rise to 80° C during intense fires in scrub, and
at depths —5 cm the temperature may reach 65° C when
fuel-loads are modest (< 0.6 kg dry leaves and stems on the
ground/m2) (Hierro & Menges 2002; Alexis et al. 2007).
However, if the fuel-load on the ground in Florida scrub is
high (~1 kg/m"), as often is the case near burrows of G.
hubbelli , then maximum soil temperatures at 5 cm depth
during a burn are very hot (88 ± 9° C) (Hierro & Menges
2002). Several other studies have reported similar relation-
ships between fuel load and soil temperature profiles
(Whelan 1995). Hence, the relatively shallow burrows of
small G. hubbelli probably reach temperatures that exceed
the upper lethal temperatures of spiders, which range from
45 to 55° C for most species (Pulz 1987; Hanna & Cobb
2007).

Assessment of fire effects on Geolycosa populations. — The
strengths of this study are: 1. burn events were true
replicates spanning 7 months of the calendar year; 2. pre-
and post-burn sampling of many (u = 700) individual
spiders was conducted; 3. sampling was size-based and
quantitative. These attributes set it apart from almost all
other previous studies that suffer from no replication or
pseudoreplication and from nonquantitative or semiquanti-
tative sampling methods (Warren et al. 1987; Siemann et al.
1997; Swengei 2001; van Mantgem et al. 2001; Hanula &
Wade 2003). However, as explicitly pointed out by Whelan
(1995), this study did not involve censuses of burrowing
wolf spider populations before and after fires at randomly
chosen sites. Thus, I cannot make any conclusions about
whether fire has a significant impact on Geolycosa popula-
tions in Florida scrub. But the data in this study suggest fire
probably is not at all deleterious to populations of G. x.
archboldi and it may have only a weak negative effect in the
short-term on G. hubbelli populations. Long-term studies
still in progress will address this subject (JEC, unpubl.
data).

CARREL-SURVIVAL OF GEOLYCOSA AFTER FIRE

599

ACKNOWLEDGMENTS

I thank the Archbold Biological Station and its staff,
especially H. Swain, M. Deyrup, E. Menges, C. Weekley, F.
Lohrer, and K. Main, for providing long-term technical,
financial, and intellectual support of the highest caliber.
Additional support came from a Development Gift Fund at
the University of Missouri. Finally, I deeply appreciate the
continuing encouragement for extended field studies provided
by Jan Weaver and other members of my family.

LITERATURE CITED

Abrahamson, W.G., A.F. Johnson, J.N. Layne & P.A. Peroni. 1984.
Vegetation of the Archbold Biological Station, Florida: an
example of the southern Lake Wales Ridge. Florida Scientist
47:209-250.

Alexis, M.A., D.P. Rasse, C. Rumpel, G. Bardoux, N. Pechot, P.
Schmalzer, B. Drake & A. Mariotti. 2007. Fire impact on C and N
losses and charcoal production in a scrub oak ecosystem.
Biogeochemistry 82:201-216.

Anderson, J.F. & G.R. Ultsch. 1987. Respiratory gas concentrations
in the microhabitats of some Florida arthropods. Comparative
Biochemistry and Physiology 88A:585-588.

Carrel, J.E. 2001. Population dynamics of the red widow spider
(Araneae: Theridiidae). Florida Entomologist 84:385-390.

Carrel, J.E. 2003a. Ecology of two burrowing wolf spiders (Araneae:
Lycosidae) syntopic in Florida scrub: burrow/body size relation-
ships and habitat preferences. Journal of the Kansas Entomolog-
ical Society 76:16-30.

Carrel, J.E. 2003b. Burrowing wolf spiders, Geolycosa spp.
(Araneae: Lycosidae): gap specialists in fire-maintained Florida
scrub. Journal of the Kansas Entomological Society 76:557-
566.

Carrel, J.E. 2008. The effect of season of fire on density of female
garden orbweavers (Araneae: Araneidae: Argiope) in Florida
scrub. Florida Entomologist 91:332-334.

Deyrup, M.A. 1989. Arthropods endemic to Florida scrub. Florida
Scientist 52:254-270.

Deyrup, M. 2005. A new species of flightless pygmy mole cricket from
a Florida sand ridge (Orthoptera: Tridactylidae). Florida Ento-
mologist 88:141-145.

Deyrup, M. & T. Eisner. 1993. Last stand in the sand. Natural
History 102(1 2):42-47.

Deyrup, M. & T. Eisner. 1996. Description and natural history of a
new pygmy mole cricket from relict xeric uplands of Florida
(Orthoptera: Tridactylidae). Memoirs of the Entomological
Society of Washington 17:59-67.

Dobson, A.P., J.P. Rodriguez, W.M. Roberts & D.S. Wilcove. 1997.
Geographic distribution of endangered species in the United
States. Science 275:550-553.

Estill, J.C. & M.B. Cruzan. 2001. Phytogeography of rare plant
species endemic to the southeastern United States. Castanea
66:3-23.

Hanna, C.J. & V.A. Cobb. 2007. Critical thermal maximum of the
green lynx spider, Peucetia viridans (Araneae, Oxyopidae). Journal
of Arachnology 35:193-196.

Hanula, J.L. & D.D. Wade. 2003. Influence of long-term dormant-
season burning and fire exclusion on ground-dwelling arthropod
populations in longleaf pine flatwoods ecosystems. Forest Ecology
and Management 175:163-184.

Hierro, J.L. & E.S. Menges. 2002. Fire intensity and shrub
regeneration in palmetto-dominated flatwoods of central Florida.
Florida Scientist 65:51-61.

Main, K.N. & E.S. Menges. 1997. Archbold Biological Station fire
management plan. Land Management Publication 97-1. 104 pp.

Marshall, S.D., W.R. Hoeh & M.A. Deyrup. 2000. Biogeography and
conservation biology of Florida’s Geolycosa wolf spiders: threat-
ened species in endangered ecosystems. Journal of Insect Conser-
vation 4:1 1-21.

Menges, E.S. 1999. Ecology and conservation of Florida scrub.
Pp. 7-22. In Savannas, Barrens and Rock Outcrop Plant
Communities of North America. (R.A. Anderson, J.S. Fralish &
J.M. Baskin, eds.). Cambridge University Press, New York.

Prestwich, K.N. 1983a. Anaerobic metabolism in spiders. Physiolog-
ical Zoology 56:1 12-121.

Prestwich, K.N. 1983b. The roles of aerobic and anaerobic
metabolism in active spiders. Physiological Zoology 56:122-132.

Prestwich, K.N. 1988a. The constraints on maximal activity in
spiders. I. Evidence against the fluid insufficiency hypothesis.
Journal of Comparative Physiology B 158:437-447.

Prestwich, K.N. 1988b. The constraints on maximal activity in
spiders. II. Limitations imposed by phosphagen depletion and
anaerobic metabolism. Journal of Comparative Physiology B
158:449-456.

Pulz, R. 1987. Thermal and water relations. Pp. 26-55. In
Ecophysiology of Spiders. (W. Nentwig, ed.). Springer-Verlag,
Berlin.

Robbins, L.E. & R.L. Myers. 1992. Seasonal effects of prescribed
burning in Florida: a review. Tall Timbers Research Inc.,
Tallahassee, Florida. Miscellaneous Publication Number
8. 96 pp.

Siemann, E., J. Haarstad & D. Tilman. 1997. Short-term and long-
term effects of burning on oak savanna arthropods. American
Midland Naturalist 137:349-361.

Simpson, G.G., A. Roe & R.C. Lewontin. 1960. Quantitative
Zoology. Revised edition. Harcourt, Brace & World. New York.
440 pp.

SPSS. 2005. SPSS for Macintosh, Release 11.0.4. SPSS Incorporated,
Chicago, Illinois.

Swengel, A.B. 2001. A literature review of insect response to fire,
compared to other conservation managements of open habitat.
Biodiversity and Conservation 10:1 141-1 169.

van Mantgem, P., M. Schwartz & M.-B. Keifer. 2001. Monitoring fire
effects for managed burns and wildfires: coming to terms with
pseudoreplication. Natural Areas Journal 21:266-273.

Wally, A.L., E.S. Menges & C.W. Weekley. 2006. Comparison of
three devices for estimating fire temperatures in ecological studies.
Applied Vegetation Science 9:97-108.

Warren, S.D., C.J. Scifres & P.D. Teel. 1987. Response of grassland
arthropods to burning: a review. Agriculture, Ecosystems and
Environment 19:105-130.

Weekley, C.W. & E.S. Menges. 2003. Species and vegetation
responses to prescribed fire in a long-unburned, endemic-rich
Lake Wales Ridge scrub. Journal of the Torrey Botanical Society
130:265-282.

Weekley, C.W., E.S. Menges & R.L. Pickert. 2008. An ecological map
of Florida’s Lake Wales Ridge: a new boundary delineation and an.
assessment of post-Columbian habitat loss. Florida Scientist
71:45-64.

Whelan, R.J. 1995. The Ecology of Fire. Cambridge University Press,
Cambridge, UK. 346 pp.

Manuscript received 22 January 2008, revised 2 May 2008.

2008. The Journal of Arachnology 36:600

BOOK REVIEW

Harvestmen: the Biology of Opiliones. Edited by Ricardo Pinto-da-Rocha, Glauco Machado and
Gonzalo Giribet. 2007. Harvard University Press, Cambridge, Massachusetts. 597 pp. ISBN 1 3:978-0-
674-02343-7. US$125.

Except for some of the very small orders, nearly all
arachnids have now received at least a first book on “The
Biology of...” It is a curious question why it took so long for
such a book to appear on the third most diverse order,
Opiliones. But here it is at last, and it proves to be well worth
the wait. Twenty-five authors have contributed to 15 chapters
which summarize virtually everything that is known about
harvestmen up to 2006. The organization of the book follows
the general pattern: there are chapters on morphology,
phylogeny and biogeography, systematics and paleontology,
ecology, feeding, enemies and defense, reproduction, develop-
ment, and social behavior. Each of the chapters is meticu-
lously researched and, rather than simply recounting what is in
the literature, the authors have synthesized and analyzed what
they found. The result is that each chapter is an original review
article that in itself is a significant contribution. Henceforward
it will not be possible to write about or research harvestmen
without referring to this book. And interestingly, the
preponderance of South Americans among the chapter
authors signals an important shift: the center of research on
this group of arthropods has moved south, perhaps propelled
by the enormous diversity in the group to be found in tropical
America.

A number of the chapters were of particular interest to this
reviewer, especially the one on systematics by Pinto-da-Rocha
and Giribet (it is also worth noting that one or the other of the
editors has co-authored nine of the fifteen chapters in the
book). This chapter is extraordinarily complete. Not only are
keys to taxa included, but each described family is discussed in
detail and abundantly illustrated. Keys are useful especially in
the suborder Laniatores, where much reshuffling of families
has taken place in the last decade. Reference is frequently
made in these discussions to advances in our understanding of
harvestmen systematics using phylogenetic data based on new
molecular evidence. Areas requiring attention, such as the
possibly paraphyletic family Ceratolasmatidae, are clearly
pointed out. An overview of the current state of classification
is given in a four-page table, covering the subfamily level,
which also gives the numbers of genera and species currently
included in each. Readers with long memories will recall that
Ernst Mayr, in a text on taxonomy, used Opiliones as an
example of an “over-split” group with, on average, less than
two species per genus. The problem still exists in some places;

the subfamily Tricommatinae has 51 species in 29 genera! Only
one small quibble with this chapter — some of the many
scanning electron micrographs used for illustrations are
printed too small.

The chapter on defense mechanisms is another gem,
especially the section on chemical defenses, a signature feature
of harvestmen biology. Here again, a useful chart puts in one
place all the molecules and the species that produce them
(except for the 22 gonyleptids studied by Hara et al. [2005]),
and a quick perusal of that chart points the way for future
work. Why, for example, does the single phalangiid studied so
far (Phalangium opilio L. 1761) produce naphthoquinones,
while the supposedly closely related sclerosomatine Leiobunum
species produce long-chain alcohols and ketones? Research in
the ecological chemistry of harvestmen has so far been focused
on gonyleptomorph Laniatores, all of which produce a
mixture of benzoquinones (at least 37 different molecules),
while the chemistry of the defensive secretion is not known for
even a single member of the Dyspnoi and is known for only
one travunioid, Sclerobunus nondimorphicus Briggs 1971.
Clearly this is an area of research where discoveries are
waiting to be made, and one which I am currently exploring
with a chemist colleague.

Finally, it was fun to read the table on pp. 2-3, which lists
vernacular names for harvestmen from more than 30
countries. Predominant are names that refer to harvest time,
the long legs of the most obvious members of the order, and
their perceived similarity to spiders. We also learn that it is
only in Finnish in which the name for a species of Opiliones,
lukki, means exactly that.

This is an important and excellent book which should be in
every arachnologist’s library, and which will be indispensable
for university and departmental libraries.

LITERATURE CITED

Hara, M.R., A.J. Cavalhiero, P. Gnaspini & D.Y.A.C. Santos. 2005.

A comparative analysis of the chemical nature of defensive

secretions of Gonyleptidae (Arachnida: Opiliones: Laniatores).

Biochemical Systematics and Ecology 33:1210-1225.

William A. Shear: Biology Department, Hampden-Sydney
College, Hampden-Sydney, Virginia 23943 USA. E-mail:
wshear@hsc.edu

600

2008. The Journal of Arachnology 36:601-603

SHORT COMMUNICATION

A new species of Xysticus (Araneae, Thomisidae) from Alberta, Canada

Charles D. Dondale: Biodiversity, Research Branch, Agriculture & Agri-Food Canada, 960 Carling Avenue, Ottawa,
Ontario K1A 0C6, Canada. E-mail: cjdondale@horizontech.ca

Abstract. A new species of the crab-spider genus Xysticus (Thomisidae), X. albertensis, is described from northern
Alberta, Canada. Specimens are compared with those of three species that closely resemble them and live in the same
geographical region, namely, X. chippewa Gertsch 1953, X. canadensis Gertsch 1934, and X. britcheri Gertsch 1934.

Keywords: Taxonomy, crab spider, western Canada

The crab-spider genus Xysticus C.L. Koch 1835 was established for
the widespread Palearctic species X. audax (Schrank 1803), and is
currently regarded as a major world entity comprising more than 300
species (Platnick 2007). North America is home to nearly 70 species, a
possible six of which are regarded as Holarctic (Dondale 2005;
Dondale et al. 2006). The purpose of the present contribution is to
describe a new species from northern Alberta. Canada, individuals of

which closely resemble those of three species that are found in the
same region, namely, X. chippewa Gertsch 1953, X. canadensis
Gertsch 1934. and X. britcheri Gertsch 1934.

Specimens are lodged in the following institutions: Canadian
National Collection of Insects & Arachnids, Ottawa, Ontario,
Canada (CNC); Strickland Entomological Museum, University of
Alberta, Edmonton, Alberta, Canada (SEM).

Figures 1-4. — Xysticus albertensis new species: 1. Palpus of holotype, ventral view; 2. Tegular apophyses of same, retrolateral view; 3.
Epigynum of paratype, ventral view; 4. Spermathecae and copulatory tubes of same, dorsal view. ATS , atrial sclerite; BTA , basal tegular
apophysis; CT, copulatory tube; DTA, distal tegular apophysis; E , embolus; RTA , retrolateral tibial apophysis; VTA, ventral tibial apophysis.
Scale bar for Figures 1, 3, 4 = 0.20 mm, for Figure 2 = 0.08 mm.

601

Figures 5-13. — 5-7. Tegular apophyses of male palpi, retrolateral view: 5. Xysticus chippewa Gertsch; 6. X. canadensis Gertsch ; 7. X. britcheri
Gertsch. 8-10. Atrial sclerites of female epigyna, ventral view: 8. X. chippewa; 9. X. canadensis; 10. X. britcheri. 11-13. Spermathecae and
copulatory tubes, dorsal view: 1 1 . X. chippewa; 12. X. canadensis; 13. X. britcheri. CT , copulatory tubes. Scale bar for Figures 5-7 = 0.08 mm, for
Figures 8-10 = 0.04 mm, for Figures 1 1-13 = 0.20 mm.

Family Thomisidae Sundevall 1833
Genus Xysticus C.L. Koch 1835

Type species. — Aranea audax Schrank, 1803, original designation

Xysticus albert ensis new species
Figs. 1-4

“ Xysticus sp. 1 ”: Nordstrom & Buckle 2004:9.

Type specimens. — Holotype male, paratype male, and paratype
female from the margin of a small lake unofficially named “Esker
Lake” by the collector, situated between Colin Lake (59°34'N,
1 1 0 08 ' W ) and Woodman Lake, in Colin-Cornwall Wildland Park,
Alberta, Canada, 6-9 July 2002, Ted Johnson (CNC). Paratypes: 1
male, 1 female, with same data (CNC); 2 males, with same data
(SEM).

Etymology. -The name of the new species is derived from that of
the Canadian province in which the type-series was collected.

Diagnosis. — Individuals of X. albertensis new species closely
resemble those of X. chippewa. but also bear some resemblance to
those of X. canadensis and X. britcheri. the last three of which are
currently treated as Holarctic. All four are characterized by the
possession of a smoothly curved distal tegular apophysis that is
neither angulate nor “heeled” (Figs. 1, 2, 5-7). Males of X.
albertensis are distinguished from those of the other three species
by the basally stout and abruptly hooked distal tegular apophysis
(compare Fig. 2 with Figs. 5-7). Also, the two tegular apophyses in
male X. albertensis are narrowly separated, whereas these structures
in the other three species are more separate. Females of the four
aforementioned species are characterized by possession of posteri-
orly converging atrial sclerites (Figs. 3, 8-10). Individuals of X.

albertensis differ from those of X. chippewa and X. canadensis by the
possession of slender copulatory tubes (compare Fig. 4 with
Figs. 11. 12) and from those of X. britcheri by thicker copulatory
tubes (compare Fig. 4 with Fig. 13). The atrial sclerites of X.
albertensis are only moderately separated posteriorly (Fig. 3),
whereas those of female X. canadensis are well separated (Fig. 9),
and those of X. britcheri are touching (Fig. 10). Individuals of both
sexes of X. albertensis are predominantly dark brown (much as in X.
britcheri), whereas those of X. chippewa and of X. canadensis are
much paler.

Description. — Holotype male: (Figs. 1, 2). Total length 3.98 mm;
carapace 2.49 mm long, 2.16 mm wide. Carapace brown on yellowish
background. Legs similar to carapace in color, paler distally; femur I
with 8 erect macrosetae on prolateral surface. Sternum with many
small round reddish brown conjoined spots. Abdomen dorsally with
extensive dark brown areas; venter pale, with scattered reddish spots.
Palpal tibia with stout tapered retrolateral apophysis and thick
angular ventral apophysis; embolus moderately thick basally, slender
distally, with free part separated from tegulum by distinct broad space
distal to tegular apophyses; distal tegular apophysis thick at base,
abruptly curved at tip, tapered to sharp point; basal tegular apophysis
short, curved.

Paratype female: (Figs. 3, 4). Total length 5.31 mm; carapace
2.41 mm long, 2.16 mm wide. Color much as in male, but somewhat
paler, with eye area and median area of carapace creamy white;
abdomen with only a few reddish or dark brown spots. Leg
macrosetae as in male. Epigynum with shallow atrium; atrial sclerites
converging posteriorly, moderately separated at posterior end
(Fig. 3). Copulatory tubes approximately two-thirds as long as
spermathecae; spermathecae shallowly lobed (Fig. 4).

DONDALE— A NEW SPECIES OF XYSTICUS

603

Variation. — Males, n = 3: Total length 3.82-4.32 mm; carapace
2.49-2.57 mm long, 1.99-2.22 mm wide.

Distribution. — Known only from the type locality.

ACKNOWLEDGMENTS

Donald J. Buckle first recognized Xysticus albertensis as new to
science and kindly passed the specimens to the author for description.

LITERATURE CITED

Dondale, C.D. 2005. Thomisidae. Pp. 246-247. In Spiders of North
America: an identification manual. (D. Ubick, P. Paquin, P.E.
Cushing & V. Roth, eds.). American Arachnological Society.

Dondale, C.D., T. Kronestedt & D.J. Buckle. 2006. Confirmation
of the presence of Xysticus chippewa in Europe (Araneae,
Thomisidae). Bulletin of the British Arachnological Society
13:361-369.

Nordstrom, W. & D. Buckle. 2004. Spider records from Colin-
Cornwall Lakes Provincial Park. Alberta Ministry of Community
Development, Edmonton, Alberta, Canada. 31 pp.

Platnick, N.I. 2007. The World Spider Catalog, Version 8.0.
American Museum of Natural History, New York. Online
at http://research.amnh.org/entomology/spiders/catalog/index.
html.

Manuscript received 7 February 2007, revised 12 February 2008.

2008. The Journal of Arachnology 36:604-605

SHORT COMMUNICATION

An easy method for handling the genus Phoneutria (Araneae, Ctenidae) for venom extraction

Leandro F. Garcia: Departamento de Biologia, Faculdade de Filosofia, Ciencias e Letras de Ribeirao Preto -

Universidade de Sao Paulo, USP, Ribeirao Preto, Sao Paulo, Brazil. CEP 14040-900, Brazil. E-mail: lfgarc@gmail.com

Luiz Henrique A. Pedrosa: Departamento de Bioquimica, Faculdade de Medicina de Ribeirao Preto - Universidade de
Sao Paulo, USP, Ribeirao Preto, Sao Paulo, Brazil. CEP 14040-900, Brazil

Denise R. B. Rosada: Departamento de Biologia, Centro Universitario Claretiano. Batatais, Sao Paulo, Brazil.

CEP 14300-000, Brazil

Abstract. This paper describes an easy, cheap, and safe method of capturing and handling the medically important spider
Phoneutria for venom extraction. The method does not injure or kill the spider and allows the extraction of pure venom.

Keywords: Armed spider, venomous spiders, vivarium

The venomous ctenid spiders of the genus Phoneutria (P. nigriventer
Keyserling 1891) (Figure la) are medically important due to their
aggressiveness (Fig. la), their great speed, and the dangerous effects
of their venom (Bticherl 1953a). Consequently, researchers are
interested in obtaining pure venom from these spiders so that the
medical significance of its components can be explored. Traditionally,
spiders have been immobilized using anesthesia (CGj, chloroform,
ether, or other substances) or with cold temperatures, but these
techniques can cause mortality or alter the behavior and physiology
of the animal (Harris et al. 1965; Randall 1982), with immediate or
latent harmful effects (Nicolas & Sillans 1989). Furthermore, venom
extraction following these handling methods has often entailed the
excision and maceration of the entire venom gland, a procedure that
produces impure venom (Biicherl 1953a). Biicherl (1953a) describes a
technique that avoids these problems. It consists of irritating the
spider so that it attacks and envenomates a device (two pipettes
connected with an elastic surgical tube) from which the venom can be
extracted. This may be the safest method for the extractor and does
not cause the animal’s death, but the process is laborious and does not
produce a sufficient amount of venom.

Here we propose an easy, cheap method for capturing and handling
Phoneutria for venom extraction that is safe for both the extractor
and the spider. The handling device consists of a transparent or semi-
transparent 2-liter plastic (PET) bottle (i.e., empty soft drink bottle)
that has been cut transversely across the middle. Only the top half of
the bottle is retained for use as a handling chamber (Fig. lb). An 8-
cm longitudinal slit is cut into the side of the handling chamber. Then,
while holding onto the top of the bottle mouth (with lid), the chamber
is placed over the spider, imprisoning it. At this point, the spider
typically attempts to climb upward toward the neck of the bottle.
Filter paper should be used as a floor to capture any venom released
by the spider; the venom can be recovered later by washing the filter
paper with an organic solvent (e.g., acetonitrile). The spider is forced
away from the bottle wall or lid and onto the filter paper by gently
tapping the chamber against the extraction bench. Once the spider is
on the filter paper, a glass stirring rod is inserted through the

longitudinal slit and pushed down onto the spider in the area between
the cephalothorax and the abdomen. This presses the spider onto the
filter paper and immobilizes it (Figure lc).

While continuing to press the animal down with the glass rod, the
handling chamber is removed and the sides of the animal’s
cephalothorax are grasped between the index finger and thumb
(Figure Id). Once the spider is firmly grasped, the glass rod can be
removed and the animal carried away for the venom extraction
(Figure le). Venom extraction consists of placing electrodes against
the region of the cephalothorax lying above the venom glands and
then stimulating the spider with 6 V (Biicherl 1953b). The venom is
collected from the chelicera in a small capillary tube (Smith & Micks
1968; Morris & Russell 1975) or on a glass plate.

LITERATURE CITED

Biicherl. W. 1953a. Novo processo de obtengao de veneno seco, puro,
de Phoneutria nigriventer (Keyserling, 1891) e titulagao da LD50 em
camundongos. Memorias do Instituto Butantan 25:153-174.
Biicherl, W. 1953b. Escorpioes e escorpionismo no Brasil. I.
Manutengao dos escorpioes em viveiros e extragao de veneno.
Memorias do Instituto Butantan 25:53-82.

Harris. R.L., R.A. Hoffman & E.D. Frazar. 1965. Chilling vs. other
methods of immobilizing Hies. Journal of Economic Entomology
58:379-380.

Morris, J.J. & R.L. Russell. 1975. The venom of the brown recluse
spider Loxosceles reclusa : composition, properties and an im-
proved method of procurement. Federation Proceedings 34:225.
Nicolas, G. & D. Sillans. 1989. Immediate and latent effects of carbon
dioxide on insects. Annual Review of Entomology 34:97-116.
Randall, J.B. 1982. Surgical restraint apparatus for living spiders.
Journal of Arachnology 10:91.

Smith, C.W. & D.W. Micks. 1968. A comparative study of the venom
and other compounds of three species of Loxosceles. American
Journal of Tropical Medicine and Hygiene 17:651-656.

Manuscript received 28 October 2007, revised 17 April 2008.

604

GARCIA ET AL. — METHOD FOR HANDLING PHONEUTRIA

605

Figure I . -The step-by-step handling method for venom extraction of Phoneutria nigriventer. a. The spider in an aggressive position; b. Left
side of an empty 2-liter plastic bottle and right side of a bottle with a transverse cut and a glass stick inserted in the perpendicular cut; c. The
trapped animal in the bottle being immobilized with the glass stick; d. Grasping the spider with fingers; e. Spider ready for venom extraction.

2008. The Journal of Arachnology 36:606-608

SHORT COMMUNICATION

Courtship behavior and copulation in Tengella radiata (Araneae, Tengellidae)

Gilbert Barrantes: Escuela de Biologi'a, Ciudad Universitaria Rodrigo Facio, Universidad de Costa Rica, San Pedro, San
Jose, Costa Rica. E-mail: gilbert.barrantes@gmail.com

Abstract. The first description of the courtship behavior and copulation is provided for Tengella radiata (Kulczynski
1909). The male courts the female by rocking his body and vibrating his abdomen. These behaviors seem to induce the
female to move out from her retreat onto the sheet and incline her body to facilitate intromission. The female has an active
role during the courtship: strumming the tunnel and sheet threads, apparently inducing the male to increase the frequency
and intensity of his courtship. Palpal insertion is extremely short. The female terminates the copulation by lunging at the
male.

Keywords: Courting, sexual selection, funnel web spider

Tengella radiata (Kulczynski 1909) has a wide distribution in Costa
Rica where it inhabits mature and secondary wet forests and coffee
plantations from 50 to 1500 m elev.( Wolff 1977; Santana et al. 1990,
pers. obs.), but it is, nevertheless, unknown outside of this small
country. Its web consists of a large, horizontal sheet with an upper
tangle that contains some cribellate threads and a tunnel at the
“interior” section of the sheet (Santana et al. 1990; Eberhard et al.
1993; Eberhard & Pereira 1993). Spiders rest near the tunnel opening
during the day.

The sexual biology of this spider is completely unknown. In nature
males occasionally co-inhabit webs with adult females (W.G.
Eberhard pers. comm), and I have occasionally observed males near
or on the sheet of possibly adult females. Here I describe for the first
time the courtship behavior and copulation of T. radiata and compare
these behaviors to those of some species within families of the
Tengella'' s sister groups lycosoids and agelenoids (Coddington 2005).
The family is of interest because it is a cribellate member of the
Lycosoidea.

Courtship behavior and copulation of two pairs of T. radiata were
filmed using a digital video camera Sony DCR-VX 1000 (30 frames/s).
Both females were virgins raised from eggs in captivity and
maintained in plastic boxes (30 X 18 X 1 1 cm) where they constructed
their webs. One female was paired with one male that was also raised
in captivity from different parents. The second female was paired with
an adult male collected in the field. Male pre-copulatory and
copulatory courtship behavior and copulation are defined as in
Eberhard & Huber (1998). Male courtship refers to those behaviors
that induce the female to respond in a way that favors the male's
reproduction (Eberhard 1996). Copulation consists of all genitalic
contact between a particular male-female pair, including the insertion
of the embolus into the epigynal opening. It finishes when the pair
separates from the copulatory position. Drawings were traced from
video images. Spiders and egg sacs were collected near San Jose,
Costa Rica; voucher specimens were deposited at the Museo de
Zoologia, Escuela de Biologia, Universidad de Costa Rica.

The first reaction of the male when placed on the female’s web was
to walk on the sheet, more or less randomly at first and then toward
the tunnel opening where the female rested. Following this movement,
courtship and copulation can be roughly divided into three
consecutive phases seen in both pairs: courtship by male, while the
female is in the tunnel, with the result that the female moves out of the
web tunnel; male courtship once the female is out on the sheet,
presumably to induce the female to adopt the copulatory position;
and copulation. The female responded to male courtship by either
sending vibratory signals through the web threads, launching an

apparent attack toward him, or adopting the copulatory position
(described below). In total, the courtship behavior and copulations of
T. radiata lasted 57 min in one pair (5 copulations) and nearly 90 min
(9 copulations) in the second pair. In both cases the female eventually
expelled the male from the web.

When the male walked directly toward the tunnel opening, he
stopped suddenly as the female began to strum the threads of the
sheet with more or less alternate movements of her palps (Fig. 1).
Strumming occurred in bouts of up to 20. During a strumming
movement the female first extended both palps anteriorly and then
flexed first one and then the other posteriorly, snagging some threads
of the sheet with the tip of the palps. These sheet threads were pulled
upward (visible in some video records) until they snapped free,
possibly due to the tension. The palpal movements gave the
impression of scratching the sheet surface rather than twanging
particular threads.

The female strumming movements apparently induced the male to
stop, at least momentarily. The male then responded by rocking his
body vigorously in antero-posterior direction (Fig. 2). He stood with
all his legs on the sheet and shook the sheet visibly as he moved
(occasionally legs I were lifted while rocking). The rocking movement
was produced primarily by the antero-posterior movement of the
male’s body rather than by bending his legs and occurred in bouts of
3-5 rocking movements ( n = 12). The male also frequently vibrated
his abdomen once or twice just before a bout of rocking movements (5
out of 8 bouts in which illumination and angle were favorable). The
female frequently stopped strumming the sheet (5 out of 7 sequences
where both male and female were in focus); nearly immediately the
male began to rock. If the female remained motionless after a rocking
bout, the male often advanced a few millimeters toward her (2 of 8
instances). However, in most cases the female began again to strum
the sheet as the male approached her, inducing him to stop and
resume courtship. It seemed that in these cases the male rocked his
body more vigorously before moving again toward the female that
remained inside but near the tunnel opening. In one instance the
female moved slowly out of the tunnel and stopped and strummed the
sheet before continuing toward the male. The male rocked his body
and advanced toward the female but she moved back a few
millimeters and then darted at him in an attacking position (first
legs slightly raised and directed forward and chelicerae spread). The
male moved rapidly backward nearly 4 cm then began to rock his
body again while the female returned to the tunnel.

In both pairs, after several separate bouts of rocking by the male
(12 in one pair and 17 in the other), the female moved out of the
tunnel, ceased strumming the sheet, and allowed the male to approach

606

BARRANTES— SEXUAL BIOLOGY OF TENGELLA RADIATA

Figure 1. — Strumming movements of the female’s palps on the
sheet during courtship. Arrows show the sequence of the palps’
movements: dots- initial position, dashed- subsequent position, solid-
final position.

her. The male moved over the female’s body so that they faced
opposite directions, and as he did so she inclined her body laterally
(Fig. 3), with her epigynum exposed (copulatory position). From this
position the male repeatedly contacted the female’s epigynum with
one palp (Figs. 3, 4), while his other palp was held in front of his
body. The male’s extended palp moved rapidly to touch the epigynum
briefly and then withdrew (mean duration of extend-contact-
withdraw cycle = 0.11 ± 0.02 s , n = 37). These movements
apparently correspond to “flubs” observed in other species (Watson
1991; Stratton et al. 1996; Huber 1998; Eberhard & Huber 1998). On
two occasions where the complete sequences were observed, the male
made 17 and 29 flubs before the palp engaged with the epigynum and
the haematodocha was finally inflated. The insertions with haema-
todocha expansion lasted 0.38 s (± 0.16, n = 5). The haematodocha
remained inflated during the entire insertion (n = 2). On three
occasions (2 in one pair and 1 in the other, where the angle and focus
were appropriate), the male was observed to move his abdomen up
and down in possible copulatory courtship movements. These
movements were slower than the pre-copulatory courtship vibratory

Figure 2. — Two partial sequences (a and b) of male-female
courtship. Black boxes- female strumming movements. Empty boxes-
male rocking movements. Dashed boxes- male moving toward the
female.

607

Figures 3, 4. — Position of male and female previous to and during
insertion. 3. The female lays on her right side as the male walks over
her body. 4. Male inserting his left pedipalp (embolus) in the left
opening of the female’s epigynum.

movements of the male’s abdomen, and were similar to the abdomen
bobbing movement of Leucauge (Eberhard & Huber 1998). The pre-
copulatory and copulatory courtship behaviors of a male after
subsequent copulations on opposite sides were similar (e.g., flubs: 19-
15-16; three subsequent copulations). The extremely short insertions
were successful as one of the females produced fertile eggs. The
insertions seemed to be ipsilateral (by the position of the female’s
epigynum and male’s palp), although 1 could not be completely
certain due to the dark color of the epigynum and male’s palp.

The female terminated copulation when she began to move her legs
to stand on the sheet after a single successful insertion of the male’s
palp (n = 5) or after several unsuccessful insertion attempts of the
male’s palp (n = 4) (I could not differentiate unsuccessful insertion
attempts from flubs). In one case the male remained over the female
and she darted toward him in an attacking position. The female’s
attack provoked an extremely rapid backward movement by the male
that positioned him at 3 or 4 cm from her. After the female had ended
the copulation, the male began a new approach with a sequence of
pre-copulatory courtship behaviors. The courting male stopped
frequently to pass his palps and sometimes his legs through his
mouthparts before approaching her again; occasionally the male
rubbed his palps against the sheet after grooming his palps with his
chelicerae. During each new male approach, the female inclined her
body toward the opposite side as the male walked over her and he
immediately began to contact her epigynum with his other palp.
Successive inclinations and insertions were on opposite sides (n = 12)
except when the male failed to insert his palp in the previous attempt.
In such cases the next male approach occurred on the same side of the
female side that he had approached previously. The female was
apparently responsible for the alternation of sides in subsequent
copulations; it was clear on two occasions that she began to incline
her body before the male could contact her.

Having copulated, the female did not necessarily accept the male
the next time he approached her. On several occasions (10 in one pair
and 4 in the other) the male stopped his approach as she began to
strum the sheet, and he restarted his rocking courtship behavior.
Neither male charged his palps with sperm during the courtship,
indicating that males charged their palps before encountering a
female.

The male’s pre-copulatory rocking and abdominal bobbing
movements during copulation may reduce the female’s aggression
and induce her to cooperate and use his sperm to fertilize her eggs
(Eberhard 1996; Stratton et al. 1996; Eberhard & Huber 1998; Peretti
et al. 2006). It is possible that these movements inform the female of
the male’s quality. For example, one of the females of this study
immediately lunged at and expelled from the web a small adult male
(ca. 15% shorter than the males studied) that 1 had previously placed
on her web.

608

THE JOURNAL OF ARACHNOLOGY

The strumming behavior of the female, her ability to expel males
with her attacks, and assumption of a distinctive acceptance posture,
all clearly show her active role in mating (Peretti et al. 2006). This
behavior possibly serves the female as a criterion for male selection as
it induces the male to restart, and in some cases, seemingly to intensify
his courtship behavior after he detects a female strumming. As in
many other spiders, there was no indication of males being able to
force females to cooperate (Huber 1996; Eberhard & Huber 1998).

It is possible to compare some aspects of the courtship behavior of
T. radiata with that of species of other related families: Agelenidae,
Lycosidae and Pisauridae (Coddington 2005). In all these families,
including T. radiata , males mount females facing in the opposite
direction (Nielsen 1932; Miller & Miller 1987; Hebets et al. 1996;
Stratton et al. 1996; Huber 1998; but see Bruce & Carico 1988).
However in T. radiata the male’s courtship induces the female to
incline her body to expose her epigynum, and thus his ventral surface
touches (or nearly so) the ventral surface of the female, while in wolf
spiders the male’s ventral surface touches the dorsal surface of the
female and the male’s pedipalp reaches the female’s epigynum around
the side of her abdomen. In at least some lycosids (Stratton et al.
1996) and in several agelenids (Huber 1998) males also flub (“scrape”
in Stratton et al. 1996) repeatedly prior to insertions. Copulations
involve ipsilateral palpal insertion on alternating sides in all families
(Stratton et al. 1996). Alternating insertions with only a single
expansion of the hematodocha also occurred in Rabidosa spp.
(Lycosinae) (Stratton et al. 1996). Female attack behavior similar to
that of T. radiata occurs in one lycosid (Miller & Miller 1987) and one
pisaurid (Arnqvist 1992). Many of these behaviors are possibly
homologous with those of Tengellidae, but information about many
more species is required to make stronger arguments regarding the
evolution of courtship behavior in Tengella and related families.
Particularly, the information on courtship behavior of agelenoids, the
sister group of Tengellidae and lycosoids, is very important to trace
the evolution of the courtship behavior in these groups of spiders.

I thank William G. Eberhard, Gail Stratton and two anonymous
reviewers for many helpful comments on the manuscript and the
Universidad de Costa Rica for financial support.

LITERATURE CITED

Arnqvist, G. 1992. Courtship behavior and sexual cannibalism in the
semi-aquatic fishing spider Dolomedes fimbriatus (Clerck) (Ara-
neae: Pisauridae). Journal of Arachnology 20:222-226.

Bruce, J.A. & J.E. Carico. 1988. Silk use during mating in Pisaurina
mira (Walkenaer) (Araneae, Pisauridae). Journal of Arachnology
16:1-4.

Coddington, J.A. 2005. Phylogeny and classification of spiders.
Pp. 18-24. la Spiders of North America: an identification manual.

(D. Ubick, P. Paquin, P.E. Cushing & V. Roth, eds.). American
Arachnoiogical Society.

Eberhard, W.G. 1996. Female Control: Sexual Selection by Cryptic
Female Choice. Princeton University Press, Princeton, New Jersey.
501 pp.

Eberhard, W.G. & B.A. Huber. 1998. Courtship, copulation, and
sperm transfer in Leucauge mariana (Araneae, Tetragnathidae)
with implications for higher classification. Journal of Arachnology
26:342-368.

Eberhard, W.G. & F. Pereira. 1993. Ultrastructure of cribellate silk of
nine species in eight families and possible taxonomic implications
(Araneae: Amaurobiidae, Deinopidae, Desidae, Dictynidae, Filis-
tatidae, Hypochilidae, Stiphidiidae, Tengellidae). Journal of
Arachnology 21:161-174.

Eberhard, W.G., N.I. Platnick & R.T. Schuh. 1993. Natural history
and systematics of arthropod symbionts (Araneae; Hemiptera;
Diptera) inhabiting webs of the spider Tengella radiata (Araneae,
Tengellidae). American Museum Novitates No. 3065:1-17.

Hebets, E.A., G.E. Stratton & G.L. Miller. 1996. Habitat and
courtship behavior of the wolf spider Schizocosa retrorsa (Banks)
(Araneae, Lycosidae). Journal of Arachnology 24:141-147.

Huber, B.A. 1998. Spider reproductive behaviour: a review of
Gerhardt’s work from 1911-1933, with implications for sexual
selection. Bulletin of the British Arachnoiogical Society 11:8-91.

Miller, G.L. & P.R. Miller. 1987. Life cycle and courtship behavior of
the borrowing wolf spider Geolycosa turricula (Treat) (Lycosidae).
Journal of Arachnology 15:385-394.

Nielsen, E. 1932. The Biology of Spiders: With Especial Reference to
the Danish Fauna. Volume 1. Levin & Munksgaard, Copenhagen.
248 pp.

Peretti, A., W.G. Eberhard & R.D. Briceno. 2006. Copulatory
dialogue: females sing during copulation to influence male genitalic
movements. Animal Behaviour 72:413-421.

Santana, M., W.G. Eberhard, G. Bassey, K.N. Prestwich & R.D.
Briceno. 1990. Low predation rates in the field by the tropical
spider Tengella radiata (Araneae: Tengellidae). Biotropica 22:305-
309.

Stratton, G.E., E.A. Hebets, P.R. Miller & G.L. Miller. 1996. Pattern
and duration of copulation in wolf spiders (Araneae, Lycosidae).
Journal of Arachnology 24:186-200.

Watson, P.J. 1991. Multiple paternity as genetic bet-hedging in female
sierra dome spiders, Linyphia litigiosa (Lynyphiidae). Animal
Behaviour 41:343-360.

Wolff, R.J. 1977. The cribellate genus Tengella (Araneae: Tengelli-
dae?). Journal of Arachnology 5:139-144.

Manuscript received 6 March 2007, revised 11 April 2008.

2008. The Journal of Arachnology 36:609-611

SHORT COMMUNICATION

Snatching prey from the mandibles of ants, a feeding tactic adopted by East African jumping spiders

Robert R. Jackson: International Centre for Insect Physiology and Ecology, PO Box 30772, Nairobi, Kenya & School of
Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch, New Zealand

Kathryn Salm: School of Biological Sciences, University of Canterbury, Private Bag 4800 Christchurch, New Zealand

Simon D, Pollard': Canterbury Museum, Rolleston Avenue, Christchurch 8013, New Zealand

Abstract. Instances are documented of salticids robbing ants by adopting a specialized behavior pattern, “snatching.” The
salticid positioned itself beside an ant column on the wall of a building, repeatedly fixating its gaze on different individual
ants in the column and maintaining fixation on the ant by turning its body while the ant walked by. When close to an ant
that was carrying prey, the salticid maneuvered about so that it was head on, grabbed hold of the prey using its chelicerae,
and then rapidly pulled the prey out of the ant’s mandibles. Having secured the prey, the salticid moved away from the ant
column to feed. All observations were made at Mbita Point, by the shore of Lake Victoria in western Kenya. The salticids
were three species of Menemerus (Simon 1868): M. bivittatus (Dufour 1831 ), M. congoensis Lessert 1927 and an undescribed
species, Menemerus sp. n. The ant species were from the genera Crematogaster (Lund 1831) and Camponotus (Mayr 1861 ).

In all instances, the salticid was 2-6 mm in body length (juveniles of all three Menemerus species and adults of Menemerus
sp. n). Prey items taken from ants were, in most instances, “lake flies” (adults of Chaoboridae and Chironomidae).

Keywords: Salticidae, ants, predation, stealing prey, Chironomidae, Chaoboridae

In the tropics, ants (Formicidae) are the dominant insects
(Holldobler & Wilson 1990) and jumping spiders (Salticidae) are
the dominant spiders (Coddington & Levi 1991), but we are only
beginning to understand how salticids and ants interact (Nelson &
Jackson 2005, 2006a, b; Nelson et al. 2006). Salticids are unique
among spiders because of their complex eyes (Land 1969; Blest et al.
1990), exceptionally acute vision (Land & Nilsson 2002) and intricate
vision-guided predatory behavior (Jackson & Pollard 1996; Harland
& Jackson 2004). Most species in this large family (about 5,000
described species, Platnick 2008) appear to be active hunters that prey
primarily on a variety of insects, but typically they do not prey on
ants. It may not be surprising that many salticid species can detect
ants by sight and then avoid coming close to them (Nelson & Jackson
2006c), particularly considering the formidable defences shown by
ants (Blum 1981; Holldobler & Wilson 1990), including powerful
mandibles, poison-injecting stings and formic-acid sprays, and the
fact that ants are sometimes predators of salticids (Nelson et al. 2004).

Yet there is a large minority of salticids (the “myrmecophagic
species”) that selects ants as preferred prey (Li & Jackson 1996; Clark
et al. 2000; Jackson & Li 2001; Huseynov et al. 2005) and one salticid
species, Cosmophasis bitaeniata (Keyserling 1882), is known to
combine chemical ant mimicry with myrmecophagy (Allan & Elgar
2001) (i.e., by mimicking the cuticular hydrocarbons of the Australian
weaver ant, Oecophylla smaragdina (Fabricius 1775), C. bitaeniata
gains entry to the weaver ant’s nest and feeds unmolested on the ant’s
larvae). Here we revisit a different style of exploiting ants — robbing
ants of objects they carry in their mandibles. This was First described
by Bhattacharya (1936) who observed juveniles of Menemerus
bivattatus (Dufour 1831) (formerly Marpissa melanognathus ) in India
grabbing food out of the mandibles of Fire ants, Solenopsis geminata
(Fabricius 1804). Our own observations show that this tactic, which
we will call “snatching from ants” or just “snatching,” for short, is
unique neither to India nor to M. bivittatus. The baseline information
we provide here is a step toward later quantitative and experimental
research concerned with this poorly understood foraging method.

1 Corresponding author. E-mail: simon. pollard@canterbury.ac.nz

METHODS

Menemerus is a large, well-deFined genus, probably with many of
the African species yet to be described (Wesolowska 1999). Our
observations were on M. bivittatus (Dufour 1831), M. congoensis
(Lessert 1927) and a new, undescribed species, Menemerus sp. n. all
three of which are common in East Africa (see Jackson 1986, 1999).
Typical body lengths of adult females of each species are: M.
bivittatus, 10 mm; M. congoensis, 7 mm; Menemerus sp. n. 5 mm.
Voucher specimens of all species from this study (salticids, ants, and
prey) have been deposited with the Florida State Collection of
Arthropods in Gainesville and the National Museums of Kenya in
Nairobi.

Our study site was by the shore of Lake Victoria in western Kenya
(Mbita Point, the Thomas Odhiambo Campus of the International
Centre for Insect Physiology and Ecology). Mbita Point is 1200 m
above sea level (0°25'S-0°30'S, 34 °10'E-35°15'E) and has a mean
annual temperature of 21° C. In this habitat, midges (Diptera:
Chironomidae & Chaoboridae), known locally as “lake flies,” are
exceedingly abundant (Beadle 1981), often covering the walls of
buildings. As midges have notoriously short life spans, lake-fly
swarms quickly turn into enormous numbers of lake-fly corpses which
are routinely scavenged by ants.

We opportunistically observed Menemerus and other salticids when
they were seen in the vicinity of ants on building walls. Whenever we
saw a salticid persistently orienting toward an ant (identified to genus
only), we continued observation for 30-60 min or until the salticid
secured the prey. First we made about 30 preliminary observations of
Menemerus sp. n. snatching lake flies from an unidentified species of
Crematogaster (Lund 1831), but with no attempt made to identify the
lake fly to family and no records kept concerning the salticid other
than the species to which it belonged. We videotaped 10 of these
preliminary observations for more detailed information about
behavior.

This was followed by observations (n = 98) that were more
standardized with respect to the information we recorded. After each
of these observations, we collected the salticid, the ant and the “prey”
(i.e., object snatched from an ant). Salticids were identified to species,

609

610

THE JOURNAL OF ARACHNOLOGY

ants to genus and lake flies to family. We also recorded whether the
salticid was a juvenile or an adult, and we recorded whether adults
were male or female. Earlier convention (Jackson and Li 2001) is
adopted for indicating frequencies of occurrence: “usually,” “often,”
“typically,” and “typical” indicate ca 80% or more.

RESULTS

Snatching from ants. —The three Menemerus species, as well as
Evarcha culicivora (Wesolowska & Jackson 2003), Harmochirus
brachiatus (Thorell 1877), Hasarius adansoni (Audouin 1826) and
unidentified species of Hyllus (Koch 1846), Natta (Karsch 1879),
Myrmarachne (MacLeay 1839), Plexippus (C.L. Koch 1846), and
Thyene (Simon 1885), were common on the walls of buildings, with
many other salticids present in smaller numbers. However, only the
three Menemerus species were observed snatching from ants.

In the records of salticids snatching front ants (n = 98), salticid
body length varied from 2 mm ( n = 1, 1%, Menemerus sp. n.) to
6 mm (n = 1, 1%, M. bivittatus), with 6 (6%) being 3 mm, 27 (28%)
being 4 mm and the majority (n = 63, 64%) being 5 mm. For the
majority records of snatching from ants (n = 98), the salticid was
Menemerus sp. n. (78, 80%), with 58 (74%) of these 78 records coming
from adult females, 6 (8%) coming from adult males, and 14 (18%)
coming from juveniles. For all records for Menemerus congoensis (n =
10, 10% of 98) and M. bivittatus ( n = 10, 10%), the salticid was a
juvenile. The ants were undetermined species from the genus
Crematogaster (n = 72, 73%) and Camponotus (Mayr 1861) (n =
26, 27%).

Observed snatching sequences took place either in the morning
(07:00-1 1:00 hours, n = 75) or in the late afternoon (17:00-
19:00 hours, n = 23). The objects snatched from ants were usually
(90, 92% of 98) dead lake flies (body lengths: 5 to 10 mm)
(chironomid, n = 77, 79% of 98) and chaoborid (n = 13, 13%).
Besides lake flies, adult females of Menemerus sp. n. (body length
5 mm) were also observed snatching an ant egg (n = 1), a dead mayfly
(Ephemeroptera, n = 1), a dead Crematogaster worker (n = 3) and
what appeared to be plant material (/; = 3), with all of these objects
being comparable to lake Hies in size. Menemerus sp. n subsequently
ate the mayfly and the ant egg, but released and moved away from the
plant material and the dead ant a few seconds after contact. There
were also five instances in which Menemerus sp. n. (3 adult females
and 2 juveniles) snatched a dead lake fly (not identified to family)
from an ant and then, a few seconds later, released the lake fly and
walked away. In all other instances, the salticid ate the lake fly it
snatched from an ant.

Behavioral sequences were similar irrespective of the different prey,
ant genus, Menemerus species and, for Menemerus sp. n., whether the
salticid was a juvenile, an adult male, or an adult female. Five
behavioral stages were discerned: tracking, intercepting, attacking,
retreating, and feeding.

Tracking: a salticid positioned itself beside an ant column on the
wall of a building, repeatedly fixating (i.e., aligning the gaze of the
corneal lenses of its anterior-medial eyes) on different individual ants
active in the column and maintaining fixation on each ant for 5 s or
longer by continually turning its body while the ant walked by. The
ant being tracked was usually carrying an object in its mandibles. The
salticid usually remained 50-100 mm from the ant while tracking and
stepped out of the way whenever an ant turned and moved in its
direction.

Intercepting: a spider that had been tracking suddenly began
stepping about and maneuvering into position in front of the ant,
effectively blocking the ant’s forward progress. This usually happened
only a few seconds after the salticid was first seen tracking, as any
longer delay usually resulted in the ant moving far away from the
salticid. When intercepting, Menemerus usually took a veering path
and approached the ant column 20-45° off from straight ahead of the
targeted ant's forward trajectory. When Menemerus stepped in front

of the ant, the ant either stopped momentarily before moving off in a
different direction or it just slowed down and veered to the side, with
Menemerus continuing to maneuver itself in front of the active ant.

Attacking: during one of an ant’s momentary pauses when being
intercepted or else while the ant was attempting to step out of the
way, a spider suddenly extended its rear legs, moved its body 1-2 mm
forward, brought its chelicerae into contact with an object in the ant’s
mandibles and then immediately stepped a few millimeters backwards
or to the side, pulling the object out of the ant’s mandibles. In all
instances, the ant released the object when the salticid pulled away.

Retreating: after extracting an object from the ant’s mandibles,
the spider turned and rapidly walked away, usually not stopping until
about 100-200 mm from the ant column.

Feeding: the spider settled, usually in a space between bricks or in
some other secluded location on the wall, and then proceeded to feed
for 1-10 min. After feeding, the spider dropped the prey and walked
away, after which it often returned to the ant column and stole
another lake fly from the ants. As many as four lake flies were
sometimes stolen in succession.

There were about 10 instances each for M. congoensis and M.
bivittatus , and more than 40 for Menemerus sp. n., in which we
observed a salticid briefly tracking an ant that had empty mandibles,
but we never saw a salticid intercept these ants. There were also about
30 instances in which Menemerus sp. n. briefly tracked, but then failed
to intercept, as well as 9 instances of seeing a salticid track and then
intercept an ant that was carrying an object other than a lake fly, but
then move away without attacking (Menemerus sp. n., 5 ants carrying
plant material and 2 carrying a dead conspecific ant worker; M.
congoensis , 2 carrying dead conspecifics).

DISCUSSION

Bhattacharya (1936) provided minima! descriptive detail of
snatching behavior. He did not indicate how many times he observed
M. bivittatus snatching from ants and he referred to the objects M.
bivittatus stole as simply “food and eggs” (ant, spider, and object sizes
not indicated). Yet his observations on M. bivittatus in India appear
to have been similar to ours: tracking, intercepting, attacking (pulling
the prey out of the ant’s mandibles) and retreating with the prey
before feeding.

Bhattacharya (1936) also observed M. bivittatus adults, but not
juveniles, stalking, capturing, and feeding on house Hies, Musca
domestica (Linnaeus 1758) and he suggested that snatching prey from
ants might be the primary foraging tactic of M. biviattatus juveniles.
We hesitate to suggest that this is the primary tactic used by any of the
active stages of any of the three Menemerus species we studied
because we observed all stages of each of the three Menemerus species
frequently capture and eat free (i.e., not in the mandibles of ants)
living prey by practicing the stalk-and-leap routines that appear to be
typical of many salticid species (Forster 1982; Richman & Jackson
1992; Jackson & Pollard 1996).

Our observations suggest instead that snatching from ants is an
alternative foraging tactic sometimes adopted by small individuals
(i.e., individuals no more than 6 mm in body length) of Menemerus.
For the smallest of the three Menemerus species we studied (i.e.,
Menemerus sp. n.), this included adults of both sexes as well as
juveniles. However, for M. bivittatus and M. congoensis , this included
only juveniles.

Despite many salticid species being abundant at Mbita Point, the
only salticids we saw snatching prey from ants were the three
Menemerus species. These observations from East Africa, together
with Bhattacharya's (1936) records from India, suggest that snatching
from ants may be widespread among species of Menemerus living in
ant-rich habitats. Further work is needed for determining whether
this tactic is special to the genus Menemerus and for clarifying the
selection pressures that might have favored the evolution of snatching
behavior.

JACKSON ET AL— SPIDERS SNATCH PREY FROM ANTS

It is difficult to envisage a salticid needing an ant’s help
overpowering inoffensive, soft-bodied lake flies and seeing building
walls on the shore of Lake Victoria covered by lake flies does not
suggest that finding lake flies is a pressing problem for which a salticid
might need an ant’s assistance. We may be tempted by an image of
salticids grazing on clumps of lake flies, rather like antelopes grazing
on clumps of grass, but choosing and capturing a living lake fly may
be far from effortless for a salticid. Time considerations may be
important. Success for Menemerus during stalk-leap sequences may
often depends on slowly moving close enough to gauge an accurate
leap, with a targeted prey potentially nullifying the salticid’s efforts by
flying away. Stalking sequences typically take several minutes,
compared with the few seconds needed to intercept an ant.

Decision making is another potential problem for a salticid. The
image of unlimited lake-fly prey changes somewhat upon close
examination. Many of the lake flies covering building walls are in fact
already dead, but stray silk lines left by spiders hold dead lake flies in
place in lifelike postures on the wall. A light breeze often makes the
dead lake flies twitch and jiggle about. Menemerus and other salticids
were often seen stalking these dead flies, leaping on them when close
and then almost immediately releasing them, but there were only five
instances in which we observed Menemerus sp. n immediately release
and move away from a dead lake fly it had snatched from an ant.
Perhaps one of the primary advantages of stealing from ants is that
the salticid can rely on the ant to select lake (lies that are still fresh
enough to be palatable.

ACKNOWLDEGM ENTS

We are especially grateful to Hans Herren, Christian Borgeimeister,
Bart Knols, and Charles Mwenda at the International Centre for
Insect Physiology and Ecology for the numerous ways in which they
supported the research. Godfrey Sune, Stephen Alluoch, Silas Ouko
Orima, and Jane Atieno provided invaluable technical assistance. We
gratefully acknowledge support from the National Geographic
Society (RRJ), the Royal Society of New Zealand (Marsden Fund;
RRJ, SDP) and James Cook Fellowship (RRJ). For taxonomic
assistance, we thank Wanda Wesolowska, G.B. Edwards, Roy
Snelling and Arthur Harrison.

LITERATURE CITED

Allan, R.A. & M.A. Elgar. 2001. Exploitation of the green tree ant,
Oecophylla smaragdina, by the salticid spider Cosmophasis
bitaeniata. Australian Journal of Zoology 49:129-137.

Beadle, L.C. 1981. The Inland Waters of Tropical Africa: An
Introduction to Tropical Limnology. 2nd edition. Longman, New
York. 457 pp.

Bhattacharya, G.C. 1936. Observations of some peculiar habits of the
spider ( Marpissa melanognathus). Journal of the Bombay Natural
History Society 39:142-144.

Blest, A.D., D.C. O’Carroll & M. Carter. 1990. Comparative
ultrastructure of Layer 1 receptor mosaics in principal eyes of
jumping spiders: the evolution of regular arrays of light guides.
Cell Tissue Research 262:445-460.

Blum, M.S. 1981. Chemical Defences of Arthropods. Academic Press,
New York. 562 pp.

Clark, R.J., R.R. Jackson & B. Cutler. 2000. Chemical cues from ants
influence predatory behavior in Habrocestum pulex (Hentz), an
ant-eating jumping spider (Araneae, Salticidae). Journal of
Arachnology 28:299-341.

Coddington, J.A. & H.W. Levi. 1991. Systematics and evolution of
spiders (Araneae). Annual Review of Ecology and Systematics
22:565-592.

611

Forster, L.M. 1982. Vision and prey-catching strategies in jumping
spiders. American Scientist 70:165-175.

Harland, D.P. & R.R. Jackson. 2004. Portia perceptions: the Umwelt
of an araneophagic jumping spider. Pp. 5-40. In Complex Worlds
from Simpler Nervous Systems. (F.R. Prete, ed.). MIT Press,
Cambridge, Massachusetts.

Holldobler, B. & E.O. Wilson. 1990. The Ants. Belknap Press,
Harvard University, Cambridge, Massachusetts. 746 pp.

Huseynov, E.F., F.R. Cross & R.R. Jackson. 2005. Natural diet and
prey-choice behaviour of Aelurillus muganicus (Araneae: Salt-
icidae), a myrmecophagic jumping spider from Azerbaijan. Journal
of Zoology, London 267:159-165.

Jackson, R.R. 1986. Communal jumping spiders (Araneae, Salticidae)
from Kenya. New Zealand Journal of Zoology 13:13-26.

Jackson, R.R. 1999. Spider cities of Africa. New Zealand Science
Monthly 10:10-11.

Jackson, R.R. & D. Li. 2001. Prey-capture techniques and prey
preferences of Zenodorus durvillei, Z. metallescens and Z.
orbiculata, tropical ant-eating jumping spiders (Araneae: Salt-
icidae) from Australia. New Zealand Journal of Zoology
28:299-341.

Jackson, R.R. & S.D. Pollard. 1996. Predatory behavior of jumping
spiders. Annual Review of Entomology 41:287-308.

Land, M.F. 1969. Structure of the retinae of the principal eyes of
jumping spiders (Salticidae: Dendryphantinae) in relation to visual
optics. Journal of Experimental Biology 51:443-470.

Land, M.F. & D.E. Nilsson. 2002. Animal Eyes. Oxford University
Press, Oxford, UK. 221 pp.

Li, D. & R.R. Jackson. 1996. Prey-specific capture behaviour and
prey preferences of myrmecophagic and araneophagic jumping
spiders (Araneae: Salticidae). Revue Suisse Zoologie, hors serie,
423-436.

Nelson, X.J. & R.R. Jackson. 2005. Living with the enemy, jumping
spiders that mimic weaver ants. Journal of Arachnology
33:813-819.

Nelson, X.J. & R.R. Jackson. 2006a. Compound mimicry and trading
predators by the males of sexually dimorphic Batesian mimics.
Proceedings of the Royal Society of London (B) 273:367-372.

Nelson, X.J., D. Li & R.R. Jackson. 2006b. Out of the frying pan and
into the fire: a novel trade-off for Batesian mimics. Ethology
1 12:270-277.

Nelson, X.J. & R.R. Jackson. 2006c. Vision-based innate aversion to
ants and ant mimics. Behavioral Ecology 17:676-681.

Nelson, X.J., R.R. Jackson, G.B. Edwards & A.T. Barrion. 2004.
Predation by ants on jumping spiders (Araneae: Salticidae) in the
Philippines. New Zealand Journal of Zoology 31:45-56.

Nelson, X.J., R.R. Jackson & D. Li. 2006. Conditional use of honest
signalling by a Batesian mimic. Behavioral Ecology 17:575-580.

Platnick, N.I. 2008. The World Spider Catalogue. Version 7.0.
American Museum of Natural History, New York. Online at
http://research.amnh.-org/entomology/spiders/catalog81-87/index.
html.

Richman, D. & R.R. Jackson. 1992. A review of the ethology of
jumping spiders (Araneae, Salticidae). Bulletin of the British
Arachnological Society 9:33-37.

Wesolowska, W. 1999. A revision of the spider genus Menemerus in
Africa (Araneae: Salticidae). Genus 10:251-353.

Manuscript received 8 August 2007, revised 21 April 2008.

2008. The Journal of Arachnology 36:612-614

SHORT COMMUNICATION

Excretion behavior of adult female crab spiders Misumena vatia (Araneae, Thomisidae)

Douglass H. Morse: Department of Ecology & Evolutionary Biology, Box G-W, Brown Elniversity, Providence,
Rhode Island 02912, USA. E-mail: d_morse@brown.edu

Abstract. Excreta potentially provide parasites or predators with information about the presence of hosts or prey; hence,
vulnerable individuals experience strong selection to minimize danger from this source. Alternatively or additionally,
excreta could alert potential prey to a spider’s presence. Adult female crab spiders Misumena vatia (Clerck 1757) exhibited a
strong reluctance to excrete when retained under tightly confined conditions. Only 5% of regularly fed individuals (1% of
total observations) excreted over observation periods of as many as 50 days while confined in 7-dram vials (5 cm high,

3 cm diameter). Individuals retained large amounts of excreta during this time. However, when released upon vegetation
over two-thirds of them excreted within 5 min, after moving to the distal end of a leaf or petal such that the excreta fell
below them onto lower vegetation or the substrate. In the field they showed little tendency to excrete close to their hunting
sites. The ability to retain excreta should serve this relatively sedentary species well in situations where it suffers high rates
of attack or may reveal its presence to potential prey.

Keywords: Defecation, parasite avoidance, predator avoidance, retention behavior

Many animals may experience risks in voiding excretory material,
which likely enhance the possibility of alerting predators and parasites
to their presence and increase the danger of disease. Alternatively or
additionally, the presence of excreta could forewarn potential prey to
a predator’s presence. Considering the importance of these factors,
surprisingly little attention has focused on studies examining the
responses of predatory invertebrates to host or prey waste products
(Weiss 2006). In particular, workers have written little about
excretion in spiders (Curtis & Carrel 2000) or other arachnids (Sato
et al. 2003; Sato & Saito 2006). In fact, Curtis & Carrel (2000)
believed their paper on excretion behavior by garden spiders Argiope
aurantia Lucas 1833 (Araneidae) (referred to as defecation behavior
by the authors) to be the first explicit study of its sort on a spider,
although Tietjen (1980) reported on the nonrandom distribution of
excreta under laboratory conditions in Mallos gregalis (Simon 1909)
(Dictynidae). Throughout this paper I use the term “excretion”
(excreta, excrete, etc.) to identify the materials passing through a
spider’s anus, since the majority of this material consists of the
products of post-assimilatory metabolic processes from the Malpi-
ghian tubules, rather than undigested matter.

In response to the risk of this material providing cues to their
presence, potential prey or hosts might develop behavioral responses
that minimize this threat, such as excreting away from their normal
activities or retaining excreta indefinitely until they can safely void
them. Taxa that show high fidelity to a site should experience
particularly strong pressure to develop such tactics, as demonstrated
by Weiss (2003, 2006) for caterpillars.

Crab spiders Misumena vatia (Clerck 1757) (Thomisidae) typically
occupy flowers as sit-and-wait predators of insects and in the process
may often remain at hunting sites for several days at a time. Excretion
in these areas could attract a large number of predators and parasites
through either visual or olfactory cues provided by this material.
Although various vertebrates are usually considered the most
common predators of spiders, they only infrequently prey on
Misumena in coastal Maine, where I conducted this study (Morse
1985, 2007). More important are other invertebrates, especially
spiders, predatory wasps, and parasitoid wasps and flies (Morse
1988a, 1988b), some of which likely respond to olfactory cues.

Here I characterize the excretion behavior of adult female
Misumena retained for extended periods under confined conditions
and then provided with sites that allowed them to dispose of their

excreta some distance away from their hunting sites. I quantified the
spiders’ frequency and size of excretion when confined to small vials
and immediately following their release onto vegetation, both flowers
in the laboratory and leafy vegetation in the field. I also report
observations on the excretion patterns of free-ranging adult females in
the field. In combination, these results allow me to test whether these
spiders discriminate among potential excretion sites, whether the
spiders’ ability to separate themselves from their excreta affects which
sites they use for this purpose, and whether the site affects the size of
the excretion.

METHODS

Adult female Misumena are medium-sized spiders that molt into the
adult stage at 35-60 mg and may exceed 400 mg when they lay their
single egg mass. These spiders have two large, raptorial anterior pairs
of legs and two much smaller posterior pairs. They can change their
color between white and yellow, and most have a prominent pair of
deep red dorsolateral abdominal stripes.

Misumena frequent old fields and roadsides in my study area
(South Bristol, Lincoln County, Maine, USA), where they hunt on
flowers for large prey. I collected 72 adult females from these sites in
June and July for studies unrelated to this one. The design of that
work dictated in part the types of observations that I could make for
this study. I kept the spiders in 7-dram vials (5 cm tall, 3 cm diameter)
at ambient temperature and light regimes and fed them a moth
(Noctuidae, Geometridae) or large fly (Syrphidae, Muscidae) every
other day. Adult female spiders grew rapidly on this diet and did not
require supplementary liquids. In the process I retained individuals
for several days to well over one month. Retention times of the
spiders varied in accordance with their mass upon capture and how
rapidly they gained mass up to the point of egg laying. I recorded
excretions when feeding the spiders and cleaning their vials after
feeding or excretion. I did not start recording retention times of
excreta by the spiders until they had been in the vials for two days to
ensure that all individuals were in a similar hunger state. These
spiders will usually take a large prey item every other day (Morse &
Fritz 1982). Numbers of drops of excreta were counted whenever
possible.

For the laboratory observations, I released female Misumena from
their vials onto a flower, either an oxeye daisy Chrysanthemum
leucanthemum or black-eyed Susan Rudebeckia hirta, and observed

612

MORSE— CRAB SPIDER EXCRETION

613

their behavior. I then counted all excretions produced within the next
five minutes. Previous observations had demonstrated that the
females often excreted immediately after release on a fiower,
particularly if 1 had retained them for several days before release
(D.H. Morse, unpub. obs.).

I placed another group of previously confined adult females in the
field on young, non-flowering milkweed Asclepias syriaca plants, sites
previously recognized as favored nesting places (Morse 1985). Upon
releasing spiders onto the plants, I observed these individuals for five
minutes to determine whether they would excrete during that period,
since earlier unrelated observations had established that they often
excreted within this time.

I also tested the frequency with which free-ranging adult female
Misumena excreted in conspicuous hunting sites on wild marjoram
Origanum vulgare over periods as long as 17 days. These marjoram
stems averaged 0.5 m in height and grew densely at the test site, which
contained several hundred flowering stems. They bore terminal
rounded panicles composed of multiple small pinkish-purple flowers.
Immediately below their inflorescences marjoram stems bear dense
ovate leaves, such that the hunting sites in marjoram occur in the top
of a dense canopy of flowers and leaves. This situation made it
difficult for spiders to excrete from their hunting sites without soiling
nearby vegetation.

RESULTS

Characteristics of behavior and excreta. -Misumena exhibited
distinctive excretion behavior: individuals moved to the tip of a petal
or leaf, raised themselves on their two pairs of large forelimbs, the two
smaller pairs of posterior limbs usually not contacting the substrate at
this time, and then released varying numbers of drops of a whitish
liquid with dark brown flecks that quickly dried in the air to a dirty
light brown. Upon release onto the flowers the spiders typically
commenced excretion behavior quickly, often within the first
30 seconds. In the laboratory, excretions made from the tips of
flower petals fell onto the substrate below; in the field they most often
completely cleared the plant in question, landing on the grass in the
substrate. When voided from milkweed leaves in the field, the excreta
most often landed on another leaf of the plant below the excretion
site. If permitted to climb to the rim of their vial the spiders readily
excreted from there as well, exhibiting the same behavior as seen on
the petals and leaves (D.H. Morse, pers. observ.).

During large excretions in which the spiders voided many drops,
they released most of these drops in a nearly constant stream, so that
my counts of these drops were approximate. These excretions
averaged 5.0 ± 0.97% (± SE, n = 6) of the previous body mass
(D.H. Morse, unpub. data). The spiders distinctly spaced these drops
in smaller excretions.

Tendency to excrete. — I recorded only four excretions in the vials
during the feeding and cleaning sessions that took place every second
or third day. These involved 72 spiders and 335 observations of
spiders at these sessions, with spiders present for 115 such sessions
(5.6% of the spiders and 1.2% of the total observations showed
excretion). I did not retain most individuals long enough to obtain
probable maximum retention times of excreta, but individuals
regularly refrained from excreting for up to one month or more,
with a maximum of 47 days. Unfortunately I failed to record which
individuals excreted, but even if one assumes that the four longest-
remaining individuals (47, 39, 39, 35 days) excreted, thereby
accounting for the four excretions recorded during this period, seven
individuals retained their excreta for over 30 days [34 (2), 33 (4), 32
(1)]. Thus, individuals could routinely retain their excreta for long
periods.

In laboratory observations, 29 (67.4%) individuals excreted during
the five-minute period after release from the vial, and 14 failed to
excrete at this time. This result differed highly significantly from the
number expected from the spiders’ behavior in the vials, which would

predict zero or one excretion (G = 40.51, df = I , P < 0.001 in a G-test
for goodness of fit). None of these individuals excreted in subsequent
minutes. When released on the milkweed plants, 28 (73.7%)
individuals excreted within five minutes, and 10 did not excrete, a
highly significant difference, using the same rationale as the previous
test ( G = 55.39, df = 1, P < 0.001, same test).

Size of excretions. Excretions, measured as drops of liquid,
differed widely in volume, probably a consequence of how long
individuals had retained this material. In the laboratory sessions
excretions averaged (± SE) 6.6 ± 1.50 drops, range = 1 to 26 (n = 22
observations); on the milkweeds they averaged 9.0 ± 1.62 drops,
range = 1 to 24 (n = 18 observations). Excretions at release in the
field significantly exceeded those during laboratory sessions ( U =
130.5, P < 0.03 in a one-tailed Mann-Whitney U test). On average,
retention of excreta at release should exceed those recorded in the
laboratory.

The size of excretions at release was positively related to the time
that spiders had retained this material (R2 = 0.41, n = 17. P < 0.01).
In contrast, no relationship occurred between the size of excretions of
individuals in the laboratory sessions and their retention times (R2 =
< 0.01, n = 15, P > 0.9, same test).

Behavior in the field. — During six censuses run every third day on
marjoram, I made 45 observations of the free-ranging spiders out of a
possible 120 (number of spiders released X number of counts). I
observed a maximum of 1 1 spiders during a census, although
recording 18 of them during one census or another (mean ± SE =
7.5 ± 0.96 individuals). As a result, these spiders might have spent as
much as 62.5% (75 of 120 possible observations) of their time away
from hunting sites, and a minimum of 27.5%, based on unrecorded
individuals later found in the flowers (33 instances). These absences
would provide ample opportunity for the spiders to excrete unnoticed.
In the process of this census 1 failed to find a single excretion in the
vicinity of a hunting site.

DISCUSSION

Adult female Misumena often retain their excreta for long periods
under experimental conditions, a trait that should help to facilitate
their relatively sedentary behavior. Since these spiders often remain
for several days at a time on a superior hunting site (Morse & Fritz
1982), they should experience strong selection to void their excreta
carefully. Seldom if ever did the location of such an individual become
conspicuous in the field (at least to the human eye) as a result of their
disposition of excreta (D.H. Morse, unpub. observ.), including the
explicit observations reported here. Spiders possess a large stercoral
pocket (cloacal chamber) with a muscular sphincter that allows them
to store large amounts of excreta (Seitz 1987).

The clear difference in relationship between size of excretion and
time confined accords with the spiders excreting more regularly in the
field than under confined laboratory conditions. Under these
circumstances the individuals tested upon release in the field would
have gone longer without excreting than those measured earlier in the
laboratory, and hence, since fed regularly, would have accumulated
significantly more excreta. Less likely, they might simply void less
prior to the experiments, though I have no basis to support this
alternative. Curtis & Carrel (2000) reported that garden spiders fed
mealworms excreted over twice a day under otherwise normal field
conditions.

The spiders hunting on marjoram would have had little opportu-
nity to excrete without soiling nearby vegetation if they had remained
on their hunting sites. They might excrete low in the vegetation
without my detecting them. Such behavior would match other
observations suggesting that the spiders excrete more readily at sites
where the excreta fall far below their hunting areas than where the
excreta would fall in their midst (D.H. Morse, unpub. observ.).

Dropping excreta from their immediate vicinity to the vegetation
beneath them should make the spiders more conspicuous to other

614

THE JOURNAL OF ARACHNOLOGY

animals on or near the substrate than to those in the canopy or above
it. However, dropping their excreta away from their canopy-level
hunting sites should make the spiders less vulnerable to most winged
attackers, probably the most important threats to spiders in the leafy
canopy. When they venture down onto the grassy substrate they
expose themselves to attacks from such predators as meadow voles
Microtus pennsylvanicus and garter snakes Thamnophis sirtalis (Morse
1985). In the canopy the egg predator Trychosis cyperia (Ichneumo-
nidae) is their most important threat (Morse 1988b). In all of my
observations on Misumena , I have never seen them preyed upon by
birds (Morse 2007); however, the spider wasp Dipogon sayi
(Pompilidae), rare in the study areas, takes a very occasional small
adult female (two observations: D.H. Morse, unpub. data), and large
sphecid wasps (Sphecidae), also uncommon in the study areas, are
other potential predators, especially of penultimates (Morse 2007).

1 have little information on the role of disease or internal parasites,
factors that should also favor careful disposition of excreta. I have
twice reared horsehair worms (Gordioida) from adult females (D.H.
Morse, unpub. data), but these events are relatively rare, since the two
records come from a sample of several thousand females collected in
the field as adults or penultimates. Tietjen (1980) reported little sign
of bacterial or fungal growth about excreta of Mallos gregalis.
Behavioral traits, however, may play an important role in controlling
levels of parasitism of other animals (e. g., Hart 1992; Ezenwa 2004).

Other sparse information on the excretion behavior of spiders
suggests that they minimize the apparency of their excreta at their
norma! hunting level in the vegetation. Curtis and Carrel (2000) noted
that the garden spider often leaves its web to excrete, and that it
generally does so at night; however, it excretes under its web,
consistent with their impression that its major predators are birds and
predatory wasps. Mallos gregalis concentrated its excreta in parts of
an experimental enclosure that it used least frequently (Tietjen 1980).
Bonnet (1930) reported that the fishing spider Dolomedes fimbriatus
(Pisauridae) forcibly cast its excreta out from as far as 3-4 cm from its
body. It also frequently excreted before jumping into the water, which
could divert a would-be predator, as suggested by Seitz (1987).
Although typically presented in the context of predator avoidance,
some results may equally well minimize apparency of the spiders to
prospective prey, as described by Brown et al. (1995) for pike-minnow
interactions. I am unaware of any instances in which excretion
patterns of spiders can be unequivocally attributed to minimizing
apparency to prey, though this relationship might obtain in many
instances, perhaps simultaneously with predator or parasite avoid-
ance. Taken together, these observations, in combination with those
of Misumena reported here, all suggest that distinctive patterns of
excretion behavior may be widespread, but frequently ignored, among
spiders.

ACKNOWLEDGMENTS

I thank K. J. Eckelbarger, T. E. Miller, L. Healy, and other staff
members of the Darling Marine Center of the University of Maine for
facilitating work on the premises; M. Weiss and an anonymous

reviewer for comments on the manuscript; and J. Rovner for
enlightening commentary on the appropriate use of the words
“excretion” and “defecation” in spider biology. Voucher specimens
from this study were deposited in the American Museum of Natural
History, New York.

LITERATURE CITED

Bonnet, P. 1930. La mue, 1’autotomie et la regeneration chez les
araignees, avec une etude des Dolomedes d’Europe. Bulletin de la
Societe d’histoire naturelle de Toulouse 59:237-700.

Brown, G.E., D.O. Chivers & RJ.F. Smith. 1995. Localized
defecation by pike: a response to labeling by cyprinid alarm
pheromone. Behavioral Ecology and Sociobiology 36:105-110.
Curtis, J.T. & J.E. Carrel. 2000. Defaecation behaviour of Argiope
aurantia (Araneae: Araneidae). Bulletin of the British Arachno-
logical Society 11:339-342.

Ezenwa, V.O. 2004. Selective defecation and selective foraging:

antiparasite behavior in wild ungulates? Ethology 110:851-862.
Hart, B.L. 1992. Behavioral adaptations to parasitism: an ethological
approach. Journal of Parasitology 78:256-265.

Morse, D.H. 1985. Nests and nest-site selection of the crab spider
Misumena vatia (Araneae, Thomisidae) on milkweed. Journal of
Arachnology 13:383-390.

Morse, D.H. 1988a. Relationship between crab spider Misumena vatia
nesting success and earlier patch-choice decisions. Ecology
69:1970-1973.

Morse, D.H. 1988b. Interactions between the crab spider Misumena
vatia (Clerck) (Araneae) and its ichneumonid egg predator
Trychosis cyperia Townes (Hymenoptera). Journal of Arachnology
16:132-135.

Morse, D.H. 2007. Predator upon a Flower. Harvard University
Press, Cambridge, Massachusetts. 392 pp.

Morse, D.H. & R.S. Fritz. 1982. Experimental and observational
studies of patch-choice at different scales by the crab spider
Misumena vatia. Ecology 63:172-182.

Sato, Y. & Y. Saito. 2006. Nest sanitation in social spider mites:
interspecific differences in defecation behavior. Ethology
112:664-669.

Sato, Y„ Y. Saito & T. Sakagami. 2003. Rules for nest sanitation in a
social spider mite, Schizotetranychus miscanthi Saito (Acari:
Tetranychidae). Ethology 109:713-724.

Seitz. K.-A. 1987. Excretory organs. Pp. 239-248. In Ecophysiology
of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin.

Tietjen, W.J. 1980. Sanitary behavior by the social spider Mallos
gregalis (Dictynidae): distribution of excreta as related to web
density and animal movements. Psyche 87:59-73.

Weiss, M.R. 2003. Good housekeeping: why do shelter-dwelling
caterpillars fling their frass? Ecology Letters 6:361-370.

Weiss, M.R. 2006. Defecation behavior and ecology of insects.
Annual Review of Entomology 51:635-661.

Manuscript received 7 December 2007, revised 26 May 2008.

2008. The Journal of Arachnology 36:615-616

INSTRUCTIONS TO AUTHORS SUBMITTING ARTICLES TO THE JOURNAL

OF ARACHNOLOGY

(instructions last revised September 2008)

General: Manuscripts must be in English and should be
prepared in general accordance with the current edition of the
Council of Biological Editors Style Manual unless instructed
otherwise below. Use the active voice throughout. Authors
should consult a recent issue of the Journal of Arachnology for
additional points of style. Manuscripts longer than three
printed journal pages (12 or more double-spaced manuscript
pages) should be prepared as Feature Articles, shorter papers
as Short Communications. One invited Review Article will be
published in the first issue of each year at the discretion of the
editors. Suggestions for review articles may be sent to the
Managing Editor. Unsolicited review articles are also wel-
comed. All review articles will be subject to the same review
process as other submissions.

Submission: Submissions must be sent electronically in
Microsoft Word (or Word compatible) format. Large plates
and figures may be sent as pdf or jpg files. Send submissions to
the Managing Editor of the Journal of Arachnology. Douglass
H. Morse, Managing Editor, Hermon Carey Bumpus Professor
of Biology Emeritus, Department of Ecology & Evolutionary
Biology, Box G-W, Brown University, Providence, RI 02912
USA [Telephone: 401-863-3152; Fax: 401-863-2166; E-mail:
d_morse@brown.edu]

The Managing Editor will acknowledge receipt of the
manuscript, assign it a manuscript number and forward it to
an Associate Editor for the review process. Correspondence
relating to manuscripts should be directed to the Associate
Editor and should include the manuscript number. If the
manuscript is accepted, the author will be asked to submit the
final copy electronically to the Associate Editor. Submission
of final illustrations is detailed below. Authors are expected to
return revisions promptly. Revised manuscripts that are not
returned in a reasonable time period (no longer than six
months for minor revisions and one year for major revisions)
will be considered new submissions.

Voucher Specimens: Specimens of species used in your
research should be deposited in a recognized scientific
institution. All type material must be deposited in a recognized
collection/institution.

FEATURE ARTICLES

Title page. — The title page includes the complete name,
address, and telephone number of the corresponding author; a
FAX number and electronic mail address if available; the title
in upper and lower case, with no more than 65 characters and
spaces per line in the title; each author’s name and address;
and the running head.

Running head. — The author’s surname(s) and an abbreviat-
ed title should be typed in all capital letters and must not
exceed 60 characters and spaces. The running head should be
placed near the top of the title page.

Abstract. — The heading should be placed at the begin-
ning of the first paragraph set off by a period. A second

abstract, in a language appropriate to the nationality of
the author(s) or geographic region(s) emphasized, may be
included.

Keywords. — Give 3-5 appropriate keywords or phrases
following the abstract. Keywords should not duplicate words
in the title.

Text.- Double-space text, tables, legends, etc. throughout.
Three levels of heads are used.

• The first level (METHODS, RESULTS, etc.) is typed in
capitals and centered on a separate line.

• The second level head begins a paragraph with an in-
dent and is separated from the text by a period and a
dash.

• The third level may or may not begin a paragraph but is
italicized and separated from the text by a colon.

Use only the metric system unless quoting text or
referencing collection data. If English measurements are used
when referencing collection data, then metric equivalents
should also be included parenthetically. All decimal fractions
are indicated by a period (e.g., -0.123). Include geographic
coordinates for collecting locales if possible.

Citation of references in the text: Cite only papers already
published or in press. Include within parentheses the surname
of the author followed by the date of publication. A comma
separates multiple citations by the same author(s) and a
semicolon separates citations by different authors, e.g., (Smith
1970), (Jones 1988; Smith 1993), (Smith & Jones 1986, 1987;
Jones et al. 1989). Include a letter of permission from any
person who is cited as providing unpublished data in the form
of a personal communication.

Citation of taxa in the text: Include the complete taxonomic
citation for each arachnid taxon when it first appears in the
manuscript. For Araneae, this information can be found
online at http://research.amnh.org/entomology/spiders/cata-
log/INTR02.html. For example, Araneus diadematus Clerck
1757. Citations for scorpions can be found in the Catalog of
the Scorpions of the World (1758-1998) by V. Fet, W.D.
Sissom, G. Lowe & M.E. Braunwalder. Citations for
pseudoscorpions can be found in the Catalogue of the
Pseudoscorpionida by M.S. Harvey. Citations for some species
of Opiliones can be found in the Annotated Catalogue of the
Laniatores of the New World (Arachnida, Opiliones) by A.B.
Kury. Citations for other arachnid orders can be found in
Catalogue of the Smaller Arachnid Orders of the World by
Mark S. Harvey.

Literature Cited section. — Use the following style and
formatting exactly as illustrated; include the full unabbre-
viated journal title. Personal web pages should not be included
in Literature Cited. These can be cited within the text as (John
Doe, pers. website) without the URL. Institutional websites
may be included in Literature Cited.

615

616

THE JOURNAL OF ARACHNOLOGY

Carico, J.E. 1993. Trechaleidae: a “new” American spider
family. Pp. 305. In Proceedings of the Ninth International
Congress of Arachnology, Panama 1983. (W.G. Eberhard,
Y.D. Lubin & B.C. Robinson, eds.). Smithsonian Institu-
tion Press, Washington, D.C.

Huber, B.A. & W.G. Eberhard. 1997. Courtship, copulation,
and genital mechanics in Physocyclus globosus (Araneae,
Pholcidae). Canadian Journal of Zoology 74:905-918.
Krafft, B. 1982. The significance and complexity of commu-
nication in spiders. Pp. 15-66. In Spider Communications:
Mechanisms and Ecological Significance. (P.N. Witt & .I.S.
Rovner, eds.). Princeton University Press, Princeton, New
Jersey.

Platnick, N.I. 2006. The World Spider Catalog, Version 7.0.
American Museum of Natural History, New York. On-
line at http://research.amnh.org/entomology/spiders/catalog/
INTR01.html

Roewer, C.F. 1954. Katalog der Araneae, Volume 2a. Institut
Royal des Sciences Naturelles de Belgique, Bruxelles. 923 pp.

Footnotes. — Footnotes are permitted only on the first
printed page to indicate current address or other information
concerning the author. All footnotes are placed together on a
separate manuscript page. Tables and figures may not have
footnotes.

Taxonomic articles. — Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Subject
Editor for Taxonomy and Systematics. Papers containing
original descriptions of focal arachnid taxa should be listed in
the Literature Cited section.

Tables. — Each table, with the legend above, should be
placed on a separate manuscript page. Only horizontal lines
(no more than three) should be included. Tables may not have
footnotes; instead, include all information in the legend.

Illustrations. Original illustrations should be sent electron-
ically when the manuscript is submitted, preferably in tiff or
jpeg format. Distribution maps should be considered figures
and numbered consecutively with other figures. (Authors
wishing to submit figures as hard copies should contact the
Editor-in-Chief for specifications.) At the submission and
review stages, the resolution standards may be low as long as
editors and reviewers can view figures effectively. Final
illustrations must be submitted to the Editor-in-Chief,
typically by e-mail or on a CD, to ensure that the electronic
versions meet publication standards and that they match the
printed copy. All figures should be at least 4 inches wide, but
no larger than a sheet of letter-size paper with 1-inch margins
all around. The resolution should be at least 300 dpi (or ppi)
for halftone or color figures and 1200 dpi for line drawings. A
guide to the Digital Art Specs for Allen Press is posted online
at: http://www2.allenpress.com/allen_press/apguides/Digital_
Art_Spec.pdf. To determine if your electronic figures adhere
to the Allen Press specifications, you can also go to http://
verifig.allenpress.com, type in the password, allenpresscmy,
and follow the instructions. Color plates can be printed, but
the author must assume the full cost paid in advance, currently

about $1100 US per color plate. Alternatively, an author can
opt to have a figure printed in black and white, but for $30/
page have that figure appear in color in the online version of
the journal. Most figures will be reduced to single-column
width (9 cm, 3.5 inches), but large plates can be printed up to
two-columns width (18 cm, 7 inches).

Address all questions concerning illustrations to the Editor-
in-Chief of the Journal of Arachnology. James E. Carrel,
Editor-In-Chief, Division of Biological Sciences, 209 Tucker
Hall, University of Missouri-Columbia, Columbia, MO 65211-
7400, USA [Telephone: 573-882-3037; FAX: 573-882-0123; E-
mail: carrelj@missouri.edu]

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Each plate
must have only one legend, as indicated below:

Figures 1-4. A-us x-us , male from Timbuktu. 1, Left leg; 2,
Right chelicera; 3, Dorsal aspect of genitalia; 4, Ventral aspect
of abdomen. Scale = 1.0 mm.

The following alternate Figure numbering is also accept-
able:

Figures la-e. A-us x-us , male from Timbuktu, a. Left leg;
b. Right chelicerae; c. Dorsal aspect of genitalia; d. Ventral
aspect of abdomen. Scale = 1.0 mm.

Assemble manuscript. The manuscript should appear in
separate sections or pages in the following sequence; title page,
abstract, text, footnotes, tables with legends, figure legends,
figures. If possible, send entire manuscript, including figures,
as one Microsoft Word document. If figures or plates are
large, please separate them from the text and send them as a
pdf or jpg file.

Page charges, proofs and reprints. — Page charges are
voluntary, but non-members of AAS are strongly encouraged
to pay in full or in part for their article ($75 / journal page).
The author will be charged for changes made in the proof
pages. Hard copy or pdf reprints are available only from Allen
Press and should be ordered when the author receives the
proof pages. Allen Press will not accept reprint orders after the
paper is published. The Journal of Arachnology also is
available through www.bioone.org and www.jstor.org. There-
fore, you can download the PDF version of your article from
one of these sites if you or your institution is a member. PDFs
of articles older than one year will be made freely available
from the AAS website.

SHORT COMMUNICATIONS

Short Communications are usually limited to three journal
pages, including tables and figures (11 or fewer double-spaced
manuscript pages including Literature Cited; no more than 2
small figures or tables). Internal headings (METHODS,
RESULTS, etc.) are omitted. Short communications must
include an abstract and keywords.

COVER ARTWORK

Authors are encouraged to send quality photographs
(preferably in color) to the editor-in-chief to be considered
for use on the cover. Images should be at least 300 dpi.

CONTENTS

Journal of Arachnology

Volume 36 Featured Articles Number 3

A redescription of Varacosa apothetica (Wallace) (Araneae, Lycosidae) by Jamin M. Dreyer 487

Description of Zabius gaucho (Scorpiones, Buthidae), a new species from southern Brazil, with an update of the

generic diagnosis by Luis E. Acosta, Denise M. Candido, Erica H. Backup & Antonio D. Brescovit 491

Revision of the spider genus Taira (Araneae, Amaurobiidae, Amaurobiinae) by Zhi-Sheng Zhang, Ming-Sheng

Zhu & Da-Xiang Song 502

Subtle pedipalp dimorphism: a reliable method for sexing juvenile spiders by Nagissa Mahmoudi, Maria Modanu,

Yoni Brandt & Maydianne C.B. Andrade 513

Activity pattern of the Neotropical harvestman Neosadocus maximus (Opiliones, Gonyleptidae): sexual and temporal

variations by Francini Osses,Tais M. Nazareth & Glauco Machado 518

First male sperm precedence in multiply-mated females of the cooperative spider Anelosimus studiosus (Araneae,

Theridiidae) by Thomas C. Jones & Patricia G. Parker 527

Frequency and consequences of damage to male copulatory organs in a widow spider by Michal Segoli, Yael Lubin

& Ally R. Harari 533

Molting interferes with web decorating behavior in Argiope keyserlingi (Araneae, Araneidae) by Andre Walter,

Mark A. Elgar, P. Bliss & Robin F. A. Moritz 538

Ontogenetic changes in web architecture and growth rate of Tengella radiata (Araneae, Tengellidae) by Gilbert

Barrantes & Ruth Madrigal-Brenes 545

Does the microarchitecture of Mexican dry forest foliage influence spider distribution? by Pablo Corcuera, Maria

Luisa Jimenez & Pedro Luis Valverde 552

Microhabitat preferences for the errant scorpion, Centruroides vittatus (Scorpiones, Buthidae) by C. Neal

McReynolds 557

Observations on phenology and overwintering of spiders associated with apple and pear orchards in south-central

Washington by Eugene R. Miliczky, David R. Horton & Carrol O. Calkins 565

Distribution of Geraeocormobius sylvarum (Opiliones, Gonyleptidae): range modeling based on bioclimatic variables

by Luis E. Acosta 574

Ecology and web allometry of Clitaetra irenae, an arboricolous African orb-weaving spider (Araneae, Araneoidea,

Nephilidae) by Matjaz Kuntner, Charles R. Haddad, Gregor Aljancic & Andrej Blejec 583

Differential survival of Geolycosa xera archboldi and G. hubbelli (Araneae, Lycosidae) after fire in Florida scrub

by James E. Carrel 595

Book Review

Harvestmen: the Biology of Opiliones. Edited by Ricardo Pinto-da-Rocha, Glauco Machado and Gonzalo Giribet.

2007. Harvard University Press, Cambridge, Massachusetts. 597 pp. ISBN 13:978-0-674-02343-7. US$125.
by William A. Shear 600

Short Communications

A new species of Xysticus (Araneae, Thomisidae) from Alberta, Canada by Charles D. Dondale 601

An easy method for handling the genus Phoneutria (Araneae, Ctenidae) for venom extraction by Leandro F. Garcia,

Luiz Henrique A. Pedrosa & Denise R. B. Rosada 604

Courtship behavior and copulation in Tengella radiata (Araneae, Tengellidae) by Gilbert Barrantes 606

Snatching prey from the mandibles of ants, a feeding tactic adopted by East African jumping spiders by Robert R.

Jackson, Kathryn Salm & Simon D. Pollard 609

Excretion behavior of adult female crab spiders Misumena vatia (Araneae, Thomisidae) by Douglass H. Morse. ... 612

INSTRUCTIONS TO AUTHORS 615

USERNAME: emerton08 PASSWORD: therid08

SMITHSONIAN INSTITUTION LIBRARIES

3 9088 01445 0076