X
(/)

i 1*^ ' ' I I I | / 2

"^SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillMs‘"’s3 I d VH 8 H^LI B RAR I ES^^’sMITHSON

^ XV 2 ^ Z \ ^ ^ ^ 2

oc

K C

OQ

_ -=• 2

NviNOSHilws S3 lava an libraries Smithsonian institution NoiiniiiSNi NViNOSHii
z 'Z ^ ^ 5

O rr, 2

■n ^ m ''<2L£i>^ 2 rn ^ Xi2ili2!^ rn

'"’smITHSONIAN“|NSTITUTION NOIinillSNl~NVINOSHilWS S3 I d Vd 8 H'lI B R AR I ES^SMITHSON
■n z (/> z w H .... t/)

5 ^ ., ,..y H I ^ ^

I W 8

> i > 'W^ 2 5 2

NviNOSHiiws^^’sa I dvaa n libraries Smithsonian institution NoiiniiisNf nvinoshiii!

C/) ^ ^ ^ X

nj v<s?v^ uj

^SMITHSONIAN'’lNSTITUTION NOliniliSNl'^NVINOSHillMS S3iavaan LI B RAR I ES^ SMITHSON!

^ I S "^ " = °

>

30

^ ' m ^ rn ^ Xjvas>a2^ rn X^Ss®

E <n — CO ■ _ (/) _

NviNosHiiiNS S3 1 a vd an libraries Smithsonian institution NoiiniiiSNi nvinoshiii

Z (/> z , w ^ ^ z • w z

2 ■ > ''^CivA^ 2 I 5

'SMITHSONIAN~INSTITUTION‘"’NOIiniliSNI_NVINOSHiiyMs‘"’s3 I d"vd 8 ll^LI B R AR I ES^^’sMITHSONIW

' — ^ z \ </» X '«

x^^flii^ ijJ x^iATItT^ 111 xrTtrlTiT;^ ^

o

Z

£/)

O

z

NVINOSHillNS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHII^!

-w x.yiiii^ m ^^ccfoiix' " m ^ XiOiixs^x' rn

LO V. Z (/> _ O) E {/)

SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3ldVdan LIBRARIES SMITHSONIf
CO z . (O z ^ w z ,..■. w

^ €f .viA , S .< 2 ,< 2

„ >

: y) ‘ z CO 2 CO

NviNosHims S3idvdan libraries Smithsonian institution NoiiniiiSNi nvinoshiii

CO CO ^ CO

^ (lit' rf 1^^ =i

n i

Ki iris rill®

riTUTION

X
(/)

I s I*

‘^NOlinillSNI NVINOSHillMS*^'"

s >

to — Z

BR ARIES SMITHSONIAN INSTITUTION
to

Ui

lavuan librar

ES SMITHSONIAN

X
to
O

z
z

INSTITUTION NOliniliSNI

NviNOSHiii/\is"S3 1 a va 0 n

JITUTION

_ to

NOliniliSNI NViNosHims S3 lava an

_ Ui

LIBRARIES

I avaa n

_ >

to 2

ES SMITHSONIAN
00 2

INSTITUTION NOliniliSNI
to

oj ZJ

SMITHSONIAN INSTITUTION
S I -•*

2 io

NviNOSHiiws S3 1 a va an

2 vv. .

a:

to

ri

f

1

s

■i

\

‘t

'I

't

1

$

i

The Journal of
ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 19

1991

NUMBER 1

THE JOURNAL OF ARACHNOLOGY

EDITOR: James E. Carico, Lynchburg College

ASSOCIATE EDITOR: Gary L, Miller, The University of Mississippi

EDITORIAL BOARD: J. E. Carrel, Univ. Missouri; J. A. Coddington, Na-
tional Mus. Natural Hist.; J. C. Cokendolpher, Lubbock, Texas; F. A. Coyle,
Western Carolina Univ.; C. D. Dondale, Agriculture Canada; W. G. Eberhard,
Univ. Costa Rica; M. E. Galiano, Mus. Argentine de Ciencias Naturales; M. H.
Greenstone, BCIRL, Columbia, Missouri; N. V. Horner, Midwestern State Univ.;
D. T. Jennings, Garland, Maine; V. F. Lee, California Acad. Sci.; H. W. Levi,
Harvard Univ.; E. A. Maury, Mus. Argentine de Ciencias Naturales; N. 1. Plat-
nick, American Mus. Natural Hist.; G. A. Polis, Vanderbilt Univ.; S. E. Riechert,
Univ. Tennesse; A. L. Rypstra, Miami Univ., Ohio; M. H. Robinson, U.S. Na-
tional Zool. Park; W. A. Shear, Hampden-Sydney Coll.; G. W. Uetz, Univ. Cin-
cinnati; C. E. Valerio, Univ. Costa Rica.

The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the
study of Arachnida, is published three times each year by The American Arach-
nological Society. Memberships (yearly): Membership is open to all those inter-
ested in Arachnida. Subscriptions to The Journal of Arachnology and American
Arachnology (the newsletter), and annual meeting notices, are included with mem-
bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or
$90 (all other countries). Inquiries should be directed to the Membership Secretary
(see below). Back Issues: Susan E. Riechert, Department of Zoology, Univ. of
Tennessee, Knoxville, TN 37916 USA. Undelivered Issues: Allen Press, Inc.,
1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: George W. Uetz (1989-1991), Department ofBiological Sciences,
University of Cincinnati, Ohio 45221, USA.

PRESIDENT-ELECT: Allen R. Brady ( 1989-199 1 ), Biology Department, Hope
College, Holland, Michigan 49423 USA.

MEMBERSHIP SECRETAR Y: Norman 1. Platnick (appointed), American Mu-
seum of Natural History, Central Park West at 79th St., New York, New York
10024 USA.

TREASURER: Gail E. Stratton (1989-1991), Department of Biology, Albion
College, Albion, Michigan 49224 USA.

SECRETARY: James W. Berry (1 989-199 1), Department ofBiological Sciences,
Butler University, Indianapolis, Indiana 46208 USA.

ARCHIVIST: Vincent D. Roth, Box 136, Portal, Arizona 85632 USA.

DIRECTORS: Petra Sierwald (1989-1991), William A. Shear (1989-1991),
Matthew H. Greenstone (1990-1992).

HONORARY MEMBERS: C. D. Dondale, W. J. Gertsch, H. Homann, H. W.
Levi, A. F. Millidge, M. Vachon, T. Yaginuma.

Cover illustration: Female Phidippus mystaceus, (Araneae, Salticidae), Eastern USA, by G. B. Edwards.

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.

1991. The Journal of Arachnology 19:1-3

ON SOUTH AMERICAN TEMINIUS (ARANEAE, MITURGIDAE)

Norman I. Platnick: Department of Entomology, American Museum of Natural History,
Central Park West at 79th Street, New York, New York 10024 USA

Martin J. Eamirez: Museo Argentine de Ciencias Naturales, Av. Angel Gallardo 470,
Buenos Aires 1405 Argentina

Abstract. The Paraguayan species Teminius agalenoides (Badcock), originally described
from juveniles only, is removed from the synonymy of T. insularis (Lucas) and considered
valid. The Argentine species Philisca filixnotaia Mello-Leitao is transferred to Teminius and
placed as a junior synonym of T. agalenoides-, males of that species are described for the
first time. Syrisca fasciata Schenkel, from Venezuela, is newly synonymized with Teminius
hirsutus (Petrunkevitch).

Because the American spiders of the genus
Teminius have a complex nomenclatural history,
at the specific, generic, and familial levels, nu-
merous synonyms were recognized in a recent
review (Platnick & Shadab 1989). Since the pub-
lication of that review, additional information
on South American members of the genus has
come to light, including Paraguayan and Argen-
tine collections (some of which were previously
misplaced in the genus Philisca) that include a
fourth valid species of Teminius. We also note
here some errors and omissions in that review.

In addition to material in the collections of the
American Museum of Natural History (AMNH)
and the Museo Argentino de Ciencias Naturales
(MACN), types and other specimens were kindly
made available by P. Hillyard of the Natural
History Museum, London (BMNH), J. A. Ko-
chalka of the Inventario Biologico Nacional, San
Lorenzo, Paraguay (IBNP), L. L. Baert of the
Institut Royal des Sciences Naturelles de Bel-
gique, Brussels (IRSN), R. Arrozpide of the Mu-
seo de La Plata (MLP), and C. Stocker of the
Naturhistorisches Museum, Basel (NMB). We
thank M. U. Shadab (AMNH) for help with il-
lustrations.

Teminius hirsutus (Petrunkevitch)

Although listed as a new combination by Plat-
nick & Shadab (1989:9), this name was in fact
used earlier (in the family Gnaphosidae) by Kraus
(1955:40, figs. 114-115), who erroneously con-
sidered Petrunkevitch’s original males and fe-
males not to be conspecific, but who did point
out that his own females resemble those of Syr-

isca pulchra Petrunkevitch (which was synony-
mized with T. hirsutus by Platnick & Shadab).

Through a clerical error, an additional syn-
onym was unfortunately omitted from the re-
view. The female holotype of Syrisca fasciata
Schenkel (1953:45, fig. 39), taken in El Pozon,
Acosta, Falcon, Venezuela (NMB), has been ex-
amined. It has the distinctive, medially arched
spermathecae of T. hirsutus, and Schenkel’s name
is here placed as a junior synonym of T. hirsutus
(NEW SYNONYMY).

Teminius insularis (Lucas)

This widespread species occurs in Florida, the
Greater Antilles, and throughout tropical South
America. Examination of the available Argen-
tine material indicates that it is probably con-
fined to the extreme northern and eastern por-
tions of that country; the confirmed Argentine
and Paraguayan records are listed here.

Specimens examined.— ARGENTINA: Chaco: El
Pintado, Sept. 1959 (A. Bachmann), 1 male (MACN);
Gancedo (M. Biraben), 1 female (MLP). Corrientes:
San Cosme, 1937 (J. Wurth), 1 female (MACN); Solan
(M. Biraben), 1 female (MLP). EntreRws: Concepcion
del Uruguay, Nov. 1963 (E. A. Maury), 1 female
(MACN); Parque Nacional El Palmar (M. Biraben), 1
female (MLP), Feb. 1981 (P. A. Goloboff), 1 male
(MACN), Nov. 1988 (M. E. Galiano), 1 female
(MACN). Jujuy: Fraile Pintado, Oct. 1967 (E. A. Mau-
ry), 1 male, 1 female (MACN); San Salvador de Jujuy,
17 Jan. 1966 (E. A. Maury), 5 females (MACN). Mi-
siones: Candelaria, Dec. 1943 (J. M. Viana), 1 female
(MACN); Eldorado, 1 Sept.-15 Nov. 1964 (A. Ko-
vacs), 2 females (AMNH); San Javier, Dec. 1948 (M.
Biraben), 1 female (MACN). PARAGUAY: Central:

2

THE JOURNAL OF ARACHNOLOGY

Figures 1-4.— Teminius agalenoides (Badcock): 1, left male palp, ventral view; 2, same, retrolateral view; 3,
epigynum, ventral view; 4, same, dorsal view.

Asuncion, 26 Mar. 1985 (J. A. Kochalka), 1 female
(IBNP); Itagua, 27 Sept. 1985 (J. A. Kochalka), 1 fe-
male (IBNP); San Lorenzo, 29 Dec. 1985-July 1986
(J. A. Kochalka), 3 males (IBNP). Concepcion: no spe-
cific locality, Nov.- 15 Dec. 1988 (J. A. Kochalka), 3
males (IBNP). Presidente Hayes: Ruta Trans-Chaco,
km. 20 to 130, 25 July 1981 (J. A. Kochalka), 1 female
(IBNP).

Teminius agalenoides (Badcock),

new combination

Figs. 1-4

Syrisca agalenoides Badcock, 1932:32 (two juvenile
syntypes from Nanahua, Presidente Hayes, Para-
guay, in BMNH, examined).

Philisca filixnotata Mello-Leitao, 1938:1 15 (juvenile
holotype from Monte Veloz, Buenos Aires, Argen-
tina, in MLP, examined); Mello-Leitao, 1941:176,
fig. 66. NEW SYNONYMY.

Teminius insularis: Platnick & Shadab, 1989:4 (in part).

Diagnosis.— Males can be distinguished from
those of the other known species of Teminius by
the greatly widened median tegular apophysis of
the palp (Fig. 1), females by the posteriorly wid-
ened and elevated anterior epigynal extension
and the w-shaped median epigynal plate (Fig. 3).
Unfortunately, the types of both specific names
are juveniles (the holotype of P. filixnotata, listed

by Mello-Leitao, 1 938, as a female, is not adult).
As those types do appear to belong to Teminius,
and were collected within the range of this spe-
cies, it seems best to consider Badcock’s and Mel-
lo-Leitao’s names synonymous, particularly as
the female figured (but not described) by Mello-
Leitao (1941) belongs to this species as well.
Measurements are in mm.

Male (Guayapa).— Total length 9.08. Cara-
pace 4.24 long, 3.26 wide. Femur II 3.38 long.
Eye sizes and interdistances: AME 0.15, ALE
0.15, PME 0.13, PLE 0.13; AME-AME 0.09,
AME- ALE 0.06, PME-PME 0.16, PME-PLE
0.21, ALE-PLE 0.09; MOQ length 0.43, front
width 0.38, back width 0.4 1 . Leg spination (only
surfaces bearing spines listed): femora: I dl-1-0,
pO-0-2; II dl-1-0, pO-1-2; III dl-1-1, pl-1-2, rO-
1-2; IV dl-1-1, pO-1-2, rO-0-2; tibiae: I pi -0-1,
v2-2-lp; II pO-0-1, vlr-2-lp; III, IV dl-1-0, pl-
1-1, v2-2-2, rO-1-1; metatarsi: I, II v2-0-0; III
pl-2-1, V2-0-2, rl-2-1; IV p2-2-2, v2-2-2, r2-2-
2. Median tegular apophysis wide, complexly
folded (Fig. 1); retrolateral tibial apophysis with
dorsally directed hook (Fig. 2).

Female (Guayapa). — Total length 12.1 1. Car-
apace 5.03 long, 4.09 wide. Femur II 3.45 long.
Eye sizes and interdistances: AME 0.16, ALE

PLATNICK & RAMiREZ-TEMINIUS (MITURGIDAE)

3

0.18, PME 0.13, PLE 0.17; AME-AME 0.16,
AME-ALE 0.10, PME-PME 0.24, PME-PLE
0.32, ALE-PLE 0.16; MOQ length 0.54, front
width 0.48, back width 0.50. Leg spination: fem-
ora: I, II dl-1-0, pO-0-1; III dl-1-1, pl-1-2, rO-

1- 1; IV dl-1-1, pO-1-1, rO-0-1; tibiae: I, II vlr-
Ir-O; III pl-1-1, V2-2-2, rO-1-1; IV pl-1-1, v2-

2- 2, rl-1-1; metatarsi: I, II vlr-0-0; III pi -2-1,
V2-0-2, r 1-2-1; IV p2-2-l, v2-2-2, r2-2-2. An-
terior epigynal extension elevated, widened pos-
teriorly (Fig. 3); spermathecal ducts coiled (Fig.
4).

Distribution.— Known only from Argentina
and western Paraguay; probably allopatric with
T. insularis, occurring to the south and west of
that species.

Specimens examined.— ARGENTINA; Buenos Aires:
Zelaya, Dec. 1938 (Heffer), 1 female (MACN). Chaco:
Pmedo(M. Biraben), 1 female (MLP). La Rioja: Guay-
apa, Patquia, 23 Jan. 1962 (L. Yiroff), 1 male (MACN),
Oct. 1965 (E. A. Maury), 1 male, 4 females (MACN),
Mar. 1968, 1 female (MACN); Tinogasta, 1947 (J.
Cranwell), 1 female (MACN). Salta: Rosario de la
Frontera (M. Biraben), 2 females (MLP); Salta, 1 fe-
male (AMNH). Santa Fe: Calchaqui (M. Biraben), 1
female (MLP); Constanza (M. Biraben), 1 female (MLP);
Departamento de Vera, Sept. 1945 (A. Giai, J. Cran-
well), 4 females (MACN). Santiago del Estero: Colonia
Dora (J. W. Abalos), 1 female (MLP); Girardet, Nov.
1939 (M. Biraben), 1 female (MLP); Nasal6, Nov. 1939
(M. Biraben), 1 female (MLP); Quimili, Nov. 1939 (M.

Biraben), 1 male, 1 female (MLP); Rumi Punco (M.
Biraben), 2 females (MLP). PARAGUAY: Chaco: Cer-
ro Le6n, Parque Nacional Defensores del Chaco, 18-
27 Nov. 1984 (J. A. Kochalka), 1 male (IBNP); Mad-
rej6n, Parque Nacional Defensores del Chaco, 5-17
Dec. 1981 (J. A. Kochalka), 1 male, 1 female (IBNP);
Misidn Cue, Tribu Nueva, 1-6 Sept. 1982 (J. A. Ko-
chalka), 1 female (IBNP). Nueva Asuncion: Estancia
La Madelon, elev. 320 m, 21 May 1984 (L. L. Baert,
J.-P. Maelfait, 1 female (IRSN). Presidente Hayes: 25
Laguas, 11-12 July 1983 (J. A. Kochalka), 1 female
(IBNP).

LITERATURE CITED

Badcock, H. D. 1932. Arachnida from the Paraguayan
Chaco. J. Linnean Soc. London (Zool.) 38:1-48.
Kraus, O. 1955. Spinnen aus El Salvador (Arachno-
idea, Araneae). Abh. Senckenb. Naturforsch. Ges.,
493:1-112.

Mello-Ixitao, C. F. 1938. Algunas aranas nuevas de la

Argentina. Rev. Mus. La Plata (N.S., Zool.), 1:89-
118.

Mello-Leitao, C. F. 1941. Las aranas de Cdrdoba, La
Rioja, Catamarca, Tucuman, Salta y Jujuy colec-
tadas por los Profesores Biraben. Rev. Mus. La Plata
(N.S., Zool.), 2:99-198.

Platnick, N. 1. & M. U. Shadab. 1989. A review of the
spider genus Teminius (Araneae, Miturgidae). Amer.
Mus. Novitates, 2963:1-12.

Schenkel, E. 1953. Bericht iiber einige Spinnentiere aus
Venezuela. Verb. Naturf Ges. Basel, 64:1-57.

Manuscript received April 1 990, revised May 1 990.

1991. The Journal of Arachnology 19:4-28

SYSTEMATIC STUDIES ON THE NITIDULUS GROUP
OF THE GENUS VAEJOVIS, WITH DESCRIPTIONS OF
SEVEN NEW SPECIES (SCORPIONES, VAEJOVIDAE)

W. David Sissom: Department of Biology, Elon College, Elon College, North Carolina
27244 USA

Abstract. New diagnostic characters for the Vaejovis nitidulus group are given, and seven
new species from Mexico are described: V. curvidigitus from Guerrero and Morelos; V.
kochi from Hidalgo and Mexico; V. mitchelli from San Luis Potosi and Queretaro; V.
platnicki from San Luis Potosi and Tamaulipas; V. pococki from San Luis Potosi and
Queretaro; V. rubrimanus from Nuevo Leon; and V. solegladi from Oaxaca and Puebla.

New records are given for V. nitidulus Koch, V. nigrescens Pocock, V. intermedins Borelli,

V. decipiens Hoffmann, and V. peninsularis Williams. Finally, hemispermatophores are
described and illustrated for seven species, and preliminary observations on their potential
usefulness in species and species group taxonomy are presented.

The nitidulus group is a significant and diverse
element of the genus Vaejovis in mainland Mex-
ico. The group was established by Sissom and
Francke (1985), although the precedent for its
recognition was set by Hoffmann (1931) in his
key to species of Vaejovis from Mexico. Sissom
and Francke (1985) assigned eleven taxa to this
group, including some species previously as-
signed to other groups. The present treatment
identifies some new important characters shared
by members of the group, provides description
of seven new species, and indicates new records
for several previously known species.

The species of the nitidulus group occur mainly
in the mountainous regions from southwestern
United States to central Oaxaca, Mexico. They
are generally found on moderate to steep slopes
characterized by arid to semiarid vegetation, al-
though some may be found in slopes dominated
by pines at higher elevations. There, they may
seek refuge in cracks and crevices of near- vertical
cliffs or deep in the layers of rock on the surface
of the slopes during the daylight hours. They
emerge at night shortly after the onset of dark-
ness, but remain on the surface for brief periods.
These habits make them difficult to collect by
rock-rolling techniques, which is perhaps the
main reason they are poorly represented in mu-
seum collections. Some species may also be lo-
cally rare or exhibit such sporadic surface oc-
currence that they appear to be rare.

A number of characters shared by members of
the group have been further studied since the

diagnosis for the group first appeared (Sissom &
Francke 1985). These characters, which are also
diagnostic, are as follows: (1) the carapace is ob-
tusely emarginate, with a distinct anteromedian
notch; (2) the genital operculi of the female pos-
sess a membranous longitudinal connection on
the anterior two-thirds to four-fifths; (3) the pec-
tinal teeth of the female are all subequal in size;
(4) the cheliceral movable finger bears a well de-
veloped serrula on the ventrodistal aspect; (5)
the ventral spinule row of tarsomere II of the legs
is flanked distally by a single pair of spines; (6)
the male hemispermatophore bears a two-
pronged hook along the ental (= medial) margin
of the distal lamina (this hook is positioned on
the dorsal face of the blade); (7) the capsular
region of the hemispermatophore is well devel-
oped, with median, basal, inner, and outer lobes
present (sometimes accessory lobes are present
as well); and (8) the ental process of the inner
hemispermatophoric lobe has a smooth margin
(i.e., it does not bear a row of booklets).

METHODS

Terminology for general morphology follows
that of Stahnke (1970) with the following excep-
tions: terminology for metasomal and pedipalpal
carinae is after Francke (1977) and trichobothrial
nomenclature is after Vachon (1974), except that
the fourth pedipalpal segment is considered the
patella, rather than the tibia, to be consistent
with Stahnke’s terminology.

Hemispermatophores are described and illus-

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

5

trated for seven species for which there was suf-
ficient male material available to allow dissec-
tion of one or more specimens. On each specimen,
the right hemispermatophore is dissected from
the body as described by Lamoral (1979:522,
fig. 74), except that upon removal, the paraxial
organ is retained intact. In lieu of dissecting the
tissues of the paraxial organ away from the hem-
ispermatophore, the entire structure is cleared
and viewed in clove oil (Sissom et al. 1990). An
advantage gained by this technique is that the
hemispermatophore will not be accidentally
damaged by further dissection. This is consid-
ered important in cases where only one or a few
male specimens are known and their hemisper-
matophores, consequently, are not expendable.
A discussion of hemispermatophore morphology
is provided in a separate section following the
species treatments; the terminology applied to
hemispermatophore structures, modified slightly
from Lamoral (1979) and Francke (1979), is out-
lined there.

Vmj&vis cmrvidigitms, new species
Figs. 1-10; 71, 72

Type data. — Holotype male from Taxco,
Guerrero, Mexico (18“33’N:99°36’W), 3 May
1963 (W. J. Gertsch and W. Ivie). Deposited in
the American Museum of Natural History, New
York,

Etymology. —The specific name is derived from
the Latin words curvus, meaning curved or bent,
and digitus, meaning finger, which describes the
distinct scalloping of the male chela fingers.

Distribution.— • Fflejovis curvidigitus is known
from several localities in the state of Morelos
and from the Taxco area in northern Guerrero,
Mexico.

Diagnosis.— Adults 30-40 mm in length. Base
color orange brown with distinct dusky mark-
ings. Stemite VII with lateral keels weak to mod-
erate, granular. Pectinal tooth count 18-21 in
males, 16-18 in females. Metasomal segment III
length/width 0.83-0.93 {N= 1 1); V iength/width
1.67-1.85 (iV= 1 1). Ventrolateral metasomal ca-
rinae moderate, granular to finely crenulate; ven-
tral submedian carinae obsolete on I-III, often
present, but faint on III and IV, Pedipaip patella
with 2 esb trichobothria; dorsoextemal caiina of
patella moderate, weakly crenulate. Pedipaip
chela with dorsal marginal and dorsointemal ca-
rinae weak, granular; keels of outer palm essen-
tially obsolete. Chela fixed finger with six sub-
rows of denticles; movable finger with six subrows

and seven inner accessory granules; fingers of
male distinctly scalloped. Chela length/width
3.40-3.75 in males (N = 4), 3.73-4.20 in females
{N = 7); fixed finger length/carapace length 0.74-
0.86 {N = 1 1); femur length/carapace length 0.88-
0.95 (V= 11).

Description. —Based on adults; parenthetical
statements refer to females. Measurements are
given in Table 1.

Coloration: Base color of carapace, tergites, and
metasoma rich to dark orange brown with dis-
tinct dusky markings. Tergites with two dark
submedian stripes and pale orange center stripe.
Distal metasomal segments darker than preced-
ing ones; telson lighter orange brown. Pedipalps
yellow brown to light orange brown, contrast-
ingly lighter than body. Legs light yellow brown
or orange brown.

Prosoma: Anterior margin of carapace ob-
tusely emarginate; median notch weak, shallow.
Entire carapacial surface finely granular.

Mesosoma: Median carina on tergite I obso-
lete; on II-VI weak, granular; tergite VII penta-
carinate, with median carina weak, granular and
lateral carinae moderate, serrate. Pectinal teeth
numbering 18-21 in males, 16-18 in females.
Stemites III-VI smooth, sparsely setose; stemite
VII with pair of moderate (weak), granular lateral
carinae.

Metasoma: Segments I-IV: Dorsolateral cari-
nae strong, crenulate to serrate; distalmost den-
ticles on I-IV distinctly enlarged, spinoid. Lateral
supramedian carinae on I-III strong, crenulate;
on IV moderate, granular; distalmost denticles
on I-III enlarged, spinoid and on IV flared. Lat-
eral inframedian carinae on I complete, strong,
crenulate; on II-III present on posterior one-third,
strong, crenulate; on IV absent. Ventrolateral ca-
rinae on I-IV moderate, granular to finely cren-
ulate. Ventral submedian carinae obsolete on I-
III, sometimes faint on IV, granular. Dorsal and
lateral intercarinal spaces with scattered granu-
lation. Segment V (Fig. i): Dorsolateral carinae
moderate, crenulate on proximal one-third,
granular posteriorly. Lateromedian carinae pres-
ent on anterior three-fourths, granular. Ventro-
lateral and ventromedian carinae moderate,
crenulate to serrate. Intercarinal spaces with scat-
tered coarse granulation.

Telson: (Fig. 1). Ventral surface of vesicle with
subtle irregular punctations and granulation;
vesicle with 12-16 pairs setae.

Pedipaip: Trichobothrial pattern of pedipalps
(Figs. 2-9) Type C, orthobothriotaxic (Vachon

6

THE JOURNAL OF ARACHNOLOGY

Figures l-lO.—Morphoiogy of Vaejovis curvidigitus, new species, from southern Mexico: i, lateral aspect of
metasomal segments IV and V, and telson; 2, dorsal aspect of pedipalp femur; 3, dorsal aspect of pedipalp
patella; 4, external aspect of pedipalp patella; 5, ventral aspect of pedipalp patella; 6, dorsal aspect of pedipalp
chela; 7, external aspect of pedipalp chela; 8, ventral aspect of pedipalp chela; 9, dentition pattern on fixed finger
of pedipalp chela; 1 0, dentition pattern on movable finger of pedipalp chela.

1974). Femur (Fig. 2): carinae granulose; internal
face with about 12-15 medium-sized granules;
dorsal face with scattered coarse granulation. Pa-
tella (Figs. 3-5): dorsointemal and ventrointemal
carinae strong, crenulate; dorsoextemal and ven-
troextemal carinae moderate, crenulate; inner face
with moderate basal tubercle and an oblique keel
of 8-10 large granules; dorsal face without con-

spicuous granulation; external face moderately
coarsely granular. Chela (Figs. 6-10) with dorsal
marginal, dorsointemal, ventrointemal, and
ventroextemal carinae weak, granular; digital Ca-
rina vestigial, smooth on distal portion of manus;
other carinae obsolete. Dentate margin of chela
fixed finger with primary denticle row divided
into six subrows by five enlarged denticles; six

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

7

Table L— Measurements in mm and pectinal tooth counts of new species of Vaejovis: V. curvidigitus, V. kochi,
V. mitchelli, and V. pococki.

V. curvidigitus V. kochi V. mitchelli V. pococki

Holotype

Paratype

Holotype

Paratype

Holotype

Paratype

Holotype

Paratype

male

female

male

female

male

female

female

male

Total length

32.2

32.8

48.4

46.9

56.3

68.5

52.5

40.6

Carapace length

4.2

4.4

5.7

5.8

6.7

8.5

6.5

5.0

Mesosoma length

9.4

10.8

14.3

15.1

15.8

21.7

16.9

12.1

Metasoma length

14.2

13.4

21.7

19.3

26.4

29.6

22.5

18.2

I length

1.8

1.7

2.8

2.4

3.4

3.9

3.0

2.4

I width

2.5

2.5

3.8

3.7

3.6

4.6

3.8

3.0

II length

2.1

2.0

3.3

2.9

4.1

4.6

3.4

2.8

II width

2.6

2.6

3.9

3.5

3.6

4.5

3.6

2.9

III length

2.4

2.2

3.5

3.2

4.4

4.9

3.8

3.1

III width

2.6

2.5

3.9

3.6

3.4

4.4

3.6

2.9

IV length

3.4

3.1

5.0

4.5

6.2

6.7

5.1

4.3

IV width

2.6

2.5

3.9

3.6

3.4

4.2

3.5

2.9

V length

4.5

4.4

7.1

6.3

8.3

9.5

7.2

5.6

V width

2.5

2.5

3.7

3.6

3.4

4.0

3.5

2.8

Telson length

4.4

4.2

6.7

6.7

7.4

8.7

6.6

5.3

Vesicle length

2.3

2.6

4.5

4.2

4.8

5.8

4.0

3.4

Vesicle width

1.8

1.7

2.8

2.6

2.4

3.0

2.8

2.2

Vesicle depth

1.3

1.1

2.2

2.0

2.0

2.7

2.0

1.6

Aculeus length

2.1

1.6

2.2

2.5

2.5

2.9

2.6

1.9

Pedipalp length

14.2

14.7

21.1

20.0

25.9

31.3

22.8

18.0

Femur length

3.8

3.9

5.7

5.3

7.0

8.4

6.0

4.8

Femur width

1.1

1.2

1.6

1.5

1.6

2.2

1.7

1.4

Patella length

4.0

4.1

5.9

5.7

7.4

9.0

6.3

5.0

Patella width

1.2

1.3

1.8

1.8

1.8

2.4

2.0

1.6

Chela length

6.4

6.7

9.5

9.0

11.5

13.9

10.5

8.1

Chela width

1.7

1.6

2.6

2.3

2.2

2.4

2.2

2.2

Chela depth

1.9

1.8

3.0

2.6

2.7

2.8

2.5

2.4

Mov. fing. length

4.0

4.3

5.9

5.8

7.5

9.2

7.0

5.1

Fix. fing. length

3.1

3.5

4.7

4.7

6.4

7.8

5.7

4.0

Pectinal teeth (1/r)

19-19

16-17

22-22

20-20

28-28

25-26

21-21

21-20

inner accessory granules (Fig. 9). Dentate margin
of chela movable finger with primary denticle
row divided into six subrows by five enlarged
denticles; seven inner accessory granules (Fig.
1 0). Fingers of male with distinct scalloping (Fig.
7).

Hemispermatophore: (Figs. 71, 72). Distal
lamina relatively short, slender (laminar length/
width = 6.63, = 1), and slightly curved; inner

lobe of capsule long, tapered; median and basal
lobes small, rounded.

Variation.— Variation in pectinal tooth counts
is as follows: in males, one comb with 1 8 teeth,
four with 19, two with 20, and one with 21; in
females, six combs with 16 teeth, three with 17,
and three with 1 8 (the teeth of one female could
not be counted). Morphometric variation is sum-
marized in the diagnosis. Some of the specimens

from Morelos are darker brown than the major-
ity of the specimens examined. This is almost
certainly due to strong discoloration from poor
preservation.

Comparisons. — Vaejovis curvidigitus is similar
to V. kochi and V. nigrescens. From V. kochi it
may be easily distinguished by its smaller body
size, lower pectinal tooth counts, and the pos-
session of two, rather than three, patellar esb
trichobothria. Further, the ventrolateral and
ventral submedian carinae of metasomal seg-
ments I-IV are considerably more reduced in V.
kochi; the keels of segment V are likewise re-
duced in that species.

From V. nigrescens, V. curvidigitus may be
distinguished by its (1) more prominent scallop
in the male chela fingers; (2) its smaller body
size; (3) more robust pedipalp chelae; (4) its pro-

8

THE JOURNAL OF ARACHNOLOGY

portionately shorter, thicker metasomal seg-
ments, with segment III wider than long; (5) the
presence of crenulate, rather than smooth, ven-
trolateral carinae on the metasoma; (6) the fre-
quent presence of faint ventral submedian cari-
nae on metasomal segments III and IV; and (7)
the presence of a distinct pattern of dusky mark-
ings on the tergites of the adult.

Specimens examined.— MEXICO: Guerrero: Taxco,
18°33’N:99°36’W, 3 May 1963 (W. J. Gertsch and W.
Ivie), 1 male holotype (AMNH), October 1945 (L.
Isaacs), 1 female paratype (AMNH), 22 August 1976
(E. S. Ross), 1 female paratype (CAS). Morelos: Cuer-
navaca, September 1946 (H. Field), 3 male paratypes
(FMNH), November 1949 (N. L. H. Krauss), 1 female
paratype (USNM), Tepoztlan, 1946-1947 (H. Field),
4 female paratypes (FMNH).

Vaejovis kochi, new species
Figs. 1 1-20

Vejovis nitidulus nigrescens, Hoffmann 1937:204 (?).
Vaejovis nitidulus nigrescens, Diaz Najera 1964:24;

1975:25.

Type data. — Holotype male (RS-4036) from
Progreso, Hidalgo, Mexico, 5 July 1963 (L. Maz-
zotti). Deposited in the Museum National d’His-
toire Naturelle, Paris.

Etymology.— The specific name is a patronym
honoring Carl Ludwig Koch for his contributions
to arachnology in the 1 800’s.

Distribution.— Known from southeastern Hi-
dalgo and northeastern Distrito Federal, Mexico.

Diagnosis.— Adults 45-50 mm in length. Base
color brown, lacking conspicuous dusky mark-
ings. Stemite VII with lateral keels weak, smooth.
Pectinal tooth count 22 in males, 19-21 in fe-
males. Metasomal segment III length/width 0.89-
0.92 {N = 3); V length/width 1.7- 1.9 {N = 3).
Ventrolateral carinae on I-III weak, smooth; on
IV essentially obsolete. Ventral submedian ca-
rinae on I-IV obsolete. Keels of segment V ob-
solete to weak. Pedipalps: patella with 3 esb tri-
chobothria; fixed finger of chela with primary
denticle row divided into six subrows; movable
finger with six subrows and seven inner accessory
granules. Male chela palm somewhat rounded,
external and dorsal carinae obsolete; fingers dis-
tinctly scalloped. Chela length/width ratio 3.52-
3.65 in males (N = 2), 3.91 in female (N = 1);
fixed finger length/carapace length 0.81-0.85 {N
= 3); femur length/carapace length 0.91-1.00 {N
= 3).

Description.— Based on adults; parenthetical

statements refer to females. Measurements ap-
pear in Table 1.

Coloration: Carapace and tergites brown, lack-
ing noticeable dusky markings. Metasomal seg-
ments I-III orange brown; IV-V more reddish
brown, especially laterally and ventrally. Telson
reddish brown with dark brown aculeus. Pedi-
palp femur yellow brown; patella yellow brown
basally, orange brown distally. Chela more or less
uniformly reddish to orange brown, slightly
darker at base of fingers. Legs yellowish or light
yellow brown.

Prosoma: Anterior margin of carapace ob-
tusely emarginate, median notch weak. Inter-
ocular area smooth; lateral and posterior portions
of carapace densely, coarsely granular.

Mesosoma: Median carina on tergites I-III ob-
solete, on IV- VI weak, smooth; tergite VII pen-
tacarinate with median carina weak, smooth
(finely granular) and lateral carinae moderate to
strong, crenulate. Pectinal teeth numbering 22 in
males, 19-21 in females. Stemites III-VI smooth,
sparsely setose; VII with pair of weak, smooth
lateral keels.

Metasoma: Segments I-IV: Dorsolateral cari-
nae strong, crenulate (serrate); distalmost den-
ticles on I-IV enlarged, spinoid. Lateral supra-
median carinae on I strong, crenulate; on II
moderate, finely granular; on III-IV weak to
moderate, smooth (on I crenulate, on II-IV finely
granular); distalmost denticles on I-III enlarged,
spinoid; on IV widely flared. Lateral inframedian
carinae on I complete, strong, finely granular
(crenulate); on II-III present on distal one-fourth
of segment, moderate to strong, smooth (finely
granular); on IV absent. Ventrolateral carinae on
I-III weak, smooth; on IV essentially obsolete.
Ventral submedian carinae on I-IV obsolete.
Dorsal and lateral intercarinal spaces sparsely,
coarsely granular. Segment V (Fig. 1 1): Dorso-
lateral carinae weak, granular on anterior one-
third to one-half Lateromedian carinae obsolete.
Ventrolateral and ventromedian carinae weak,
smooth to finely granular. Intercarinal spaces
smooth, lustrous.

Telson: (Fig. 1 1). Ventral surface of vesicle
smooth, with about 12-15 pairs setae.

Pedipalp: Trichobothrial pattern (Figs. 12-19)
Type C, neobothriotaxic (Vachon 1974): patella
with 3 esb trichobothria (Fig. 14). Femur (Fig.
12): carinae strong, granulose; internal face with
8-10 large, pointed granules; dorsal face mod-
erately, coarsely granular. Patella (Figs. 13-15):
dorsointemal and ventrointemal carinae strong.

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

9

Figures 1 1-20. —Morphology of Vaejovis kochi, new species, from Hidalgo, Mexico: 11, lateral aspect of
metasomal segments IV and V, and telson; 12, dorsal aspect of pedipalp femur; 13, dorsal aspect of pedipalp
patella; 1 4, external aspect of pedipalp patella; 1 5, ventral aspect of pedipalp patella; 1 6, dorsal aspect of pedipalp
chela; 17, external aspect of pedipalp chela; 18, ventral aspect of pedipalp chela; 19, dentition pattern on fixed
finger of pedipalp chela; 20, dentition pattern on movable finger of pedipalp chela.

crenulate; dorsoexternal carina moderate,
smooth; ventroextemal carina moderate, gran-
ular. Inner face with oblique longitudinal carina
of 8 larger and 3-4 smaller granules; dorsal and
ventral faces smooth; external face with scattered
granulation. Chela (Figs. 1 6-20): Dorsal marginal
carina weak, finely granular; dorsointemal carina
weak, with a few larger, rounded granules; ven-

trointemal carina weak, granular; other carinae
essentially obsolete. Dentate margin of fixed fin-
ger with primary denticle row divided into six
subrows by five enlarged denticles; six inner ac-
cessory granules (Fig. 19). Dentate margin of
movable finger with primary denticle row divid-
ed into six subrows by five enlarged denticles;
seven inner accessory granules (Fig. 20). Fingers

10

THE JOURNAL OF ARACHNOLOGY

of male with distinct scalloping. Chela relatively
robust (Figs. 16-18).

Hemispermatophore: Not dissected due to
scarcity of material.

Variation.— Only two adult males, a single
adult female, and a number of late instar juvenile
females were available for study. Juveniles differ
from adults primarily in coloration and mor-
phometries: the body is yellowish brown with
distinct dusky markings; the distal metasomal
segments are more reddish brown, contrasting
with the rest of the body; the pedipalps are yel-
lowish; and the pedipalps are more slender. Pec-
tinal tooth count variation in all specimens is as
follows: in males, four combs with 22 teeth; in
females, four combs with 1 9 teeth, ten with 20,
and four with 21. Morphometric variation is
summarized in the diagnosis.

Comparisons. — Vaejovis kochi is most similar
to V. platnicki, V. nigrescens, and V. curvidigitus.
From V. platnicki, it can be easily distinguished
by its larger body size, uniform reddish brown
coloration, more robust pedipalp chelae, and
higher pectinal tooth counts (22 in males and 19-
21 in females, rather than 16 in males and 13-
1 5 in females).

Vaejovis kochi is easily distinguished from both
V. nigrescens and V. curvidigitus by the posses-
sion of three, rather than two, patellar esb tri-
chobothria. From V. nigrescens, it may be fur-
ther distinguished by the stronger scallop in the
male chela fingers and the more robust chela
palm. Metasomal segment V in V. kochi has weak
to obsolete keels; in V. nigrescens, the dorsal and
lateral keels are moderate, and the ventrolateral
and ventromedian keels are strong.

Specimens examined.— MEXICO: no locality data,
1 male paratype (RS-4086)(MNHN). Hidalgo: Progre-
so, 5 July 1963 (L. Mazzotti), 1 male holotype (RS-
4036)(MNHN), Cuautepec, 20 March 1964 (no col-
lector), 1 female, 1 juv. paratypes (RS-4287)(MNHN),
10 km NW Atotonilco El Grande, surface above Pu-
ente de Dios, 19 March 1981 (J. Reddell), 3 subadult
female paratypes (WDS). Distrito Federal: Teotihu-
acan, 14 November 1973 (E.-G. Burmeister), 3 sub-
adult female paratypes (WDS), San Juan Teotihuacan,
28 July 1947 (collector unknown), 1 subadult female
paratype (AMNH).

Vaejovis mitchelli, new species
Figs. 21-30; 79, 80

Type data. — Holotype male from 8 mi. W Jal-
pan, Queretaro, Mexico, 10 March 1977 (R. W.
Mitchell, et al.). Deposited in the American Mu-
seum of Natural History, New York.

Etymology.— The specific epithet is a patron-
ym honoring Robert W. Mitchell, collector of
most of the known specimens of this taxon, for
his outstanding contributions to arachnology and
Mexican cave biology.

Distribution. — Known from the type locality
and southeastern San Luis Potosi.

Diagnosis.— Adults 50-70 mm in length. Base
color dark orange brown to dark brown with faint
underlying dusky markings. Stemite VII with lat-
eral keels moderate, finely granular. Pectinal tooth
counts 27-28 in males, 25-26 in females. Me-
tasomal segment III length/width 1.28-1.41 in
males {N = 3), 1.10-1.13 in females {N = 3);
segment V length/width 2.45-2.58 in males {N
= 3), 2.26-2.36 in females {N = 3). Metasoma
with ventrolateral carinae moderate, smooth to
finely granular on I-IV; ventral submedian ca-
rinae on I-IV obsolete or discernible as faint,
smooth ridges. Pedipalp patella with 2 esb tri-
chobothria; fixed finger with primary denticle row
divided into seven subrows; movable finger with
eight subrows and eight inner accessory granules;
dorsal and external keels of palm reduced. Chela
length/width 4.90-5.81 in males {N = 3), 5.43-
5.92 in females {N = 3); fixed finger length/car-
apace length 0.91-1.08 (TV = 6); femur length/
carapace length 0.99-1.12 (TV = 6).

Description.— Based on adults; parenthetical
statements refer to females. Measurements are
given in Table 1.

Coloration: Carapace and tergites dark orange
brown to dark brown with underlying dusky pat-
tern. Metasoma: dark brown to dark orange
brown, either uniformly colored or with seg-
ments IV-V darker. Telson vesicle yellow orange
or reddish. Pedipalp femur and patella light or-
ange brown to brown, with variable underlying
dusky markings. Pedipalp chela manus light or-
ange brown; fingers brown basally, light yellow
brown distally. Legs yellow brown (males) to
brown (females) with dusky markings on prox-
imal segments, uniformly yellowish on tibiae and
tarsi.

Prosoma: Anterior margin of carapace ob-
tusely emarginate, median notch shallow. Inter-
ocular area finely granular with some scattered
coarse granules (smooth with few scattered coarse
granules); remainder of carapace densely (sparse-
ly), coarsely granular.

Mesosoma: Median carina on tergites I-II ob-
solete; on III faint; on IV- VI weak, granular. Ter-
gite VII pentacarinate, with median carina mod-
erate, granular; lateral carinae strong, crenulate
to serrate. Pectinal teeth numbering 27-28 (25-

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

11

Figures 2 1-30. —Morphology of Vaejovis mitchelli, new species, from Queretaro, Mexico: 21, lateral aspect
of metasomal segments IV and V, and telson; 22, dorsal aspect of pedipalp femur; 23, dorsal aspect of pedipalp
patella; 24, external aspect of pedipalp patella; 25, ventral aspect of pedipalp patella; 26, dorsal aspect of pedipalp
chela; 27, external aspect of pedipalp chela; 28, ventral aspect of pedipalp chela; 29, dentition pattern on fixed
finger of pedipalp chela; 30, dentition pattern on movable finger of pedipalp chela.

26). Stemites III-VI smooth, sparsely setose; VII
with pair of moderate, finely granular lateral ca-
rinae.

Metasoma: Dorsolateral carinae strong, cren-
ulate to serrate; distalmost denticles of dorsolat-
eral carinae on I-III enlarged, spinoid; on seg-
ment IV only slightly enlarged, spinoid. Lateral
supramedian carinae on I-II strong, crenulate;
on III-IV moderate, finely crenulate; distalmost
denticles on I-III slightly enlarged, spinoid and

on IV flared. Lateral inframedian carinae on I
complete, strong, crenulate; on II present on pos-
terior one-half, strong, crenulate; on III present
on posterior one-fourth, moderate, crenulate; on
IV absent. Ventrolateral carinae on I-IV mod-
erate, smooth to finely granular. Ventral sub-
median carinae on I-IV obsolete or present as
very faint, smooth ridges. Dorsal and lateral in-
tercarinal spaces with scattered coarse granules.
Segment V (Fig. 21); Dorsolateral carinae mod-

12

THE JOURNAL OF ARACHNOLOGY

erate, granular. Lateromedian carinae weak,
granular, present on anterior three-fourths. Ven-
trolateral and ventromedian carinae moderate,
crenulate to serrate. Intercarinal spaces with scat-
tered fine granulation.

Telson: (Fig. 21). Ventral surface with small,
irregularly spaced granules and punctations; about
25 pairs of setae.

Pedipalp: Trichobothrial pattern (Figs. 22-29)
Type C, orthobothriotaxic (Vachon 1974). Fe-
mur (Fig. 22) with carinae strong, granulose; in-
ternal face with 1 6-20 medium to large subcon-
ical granules; dorsal face densely, finely granular.
Patella (Figs. 23-25) dorsointemal, ventrointer-
nal, and ventroextemal carinae strong, crenulate;
dorsoextemal carina moderate, granular to weakly
crenulate. Internal face with moderate basal tu-
bercle and oblique longitudinal carina of 10-12
large, subconical granules. Dorsal face with scat-
tered fine granules; external face with moderately
dense coarse and fine granulation. Chela (Figs.
26-30). Dorsal marginal carina weak, granular.
Dorsal secondary carina faint, smooth. Digital
carina weak, smooth. External secondary carina
obsolete. Ventroextemal carina weak, smooth.
Ventromedian carina obsolete. Ventrointemal
carina weak, essentially smooth. Dorsointemal
carina moderate, with enlarged, sharp granules.
Dentate margin of fixed finger with primary den-
ticle row divided into seven subrows by six en-
larged granules; six inner accessory granules (Fig.
29). Dentate margin of movable finger with pri-
mary denticle row divided into eight subrows by
seven enlarged granules; eight inner accessory
granules (Fig. 30). Fingers of male with subtle
scalloping. Chela slender with fingers long and
tenuous (Figs. 26-28).

Hemispermatophore: (Figs. 79, 80). Distal
lamina relatively slender (laminar length/width
= 6.48, N = 1), with slight tapering; inner lobe
of capsule long, broad, slightly tapering distally;
median and basal lobes rounded.

Variation.— Variation in the pectinal tooth
counts of the adult specimens examined is as
follows: in males, two combs with 26 teeth and
four with 27; in females, four combs with 25
teeth and two with 26. In addition, pectinal tooth
counts of 20 neonates bom in the laboratory were
also counted. Because neonates cannot be ac-
curately sexed, the counts obtained include both
males and females. There were 1 comb with 24
teeth, 13 combs with 25 teeth, 13 with 26 teeth,
3 with 27 teeth, and 10 with 28 teeth. Because
pectinal tooth counts do not change after birth,

these counts provide a reasonable estimate of
variation in the pectinal tooth counts in this spe-
cies, which was not possible using only the six
adults. Morphometric variation is summarized
in the diagnosis.

Comparisons.— Faejovw mitchelU, V. nitidu-
lus, and V. pococki are the only three species in
the genus to have seven subrows on the pedipalp
chela fixed finger. Vaejovis mitchelU may be eas-
ily distinguished from V. nitidulus by its dark
coloration, by having eight subrows of denticles
on the movable finger and eight inner accessory
granules (not seven subrows and seven inner ac-
cessory granules), by having only two esb tricho-
bothria on the patella, and by differences in ped-
ipalp and metasomal morphometries.

The movable finger characteristic also serves
to distinguish V. mitchelU from V. pococki. In
addition, pectinal tooth counts in V. mitchelU
are distinctly higher than in V. pococki, and mor-
phometries of the pedipalp chela and metasoma
differ considerably between the two. The ven-
trolateral carinae of metasomal segments I-IV
are moderate in V. mitchelU, essentially obsolete
in V. pococki. Carinal development on metaso-
mal segment V is also stronger in V. mitchelU.

Comments.— The type series, collected by Dr.
Robert W. Mitchell and his 1977 Arachnology
class at Texas Tech University, was returned to
the laboratory alive after their capture. One of
the females was observed mating on 1 3 March
1977 and subsequently gave birth to 39 offspring
on 4 August 1977. Assuming the female had not
been previously inseminated in the field, the ges-
tation period was 144 days (= 4.5 months). A
second female gave birth to 36 young shortly
upon her return to the laboratory.

Specimens examined.— MEXICO: Queretaro: 8 mi.
W Jalpan, 10 March 1977 (R. W. Mitchell, et al.), 1
male holotype (AMNH), 1 male paratype, 2 female
paratypes (AMNH-OFF), 1 female paratype (WDS).
San Luis Potost Cueva de Cristian, 4 km E Xilitla, 4
January 1976 (A. Grubbs), 1 male paratype (WDS).

Vaejovis platnicki, new species
Figs. 31-40

Type data.— Holotype female from Guayla-
lejo, Tamaulipas, Mexico, 18 February 1973 (J.
P. Webb). Deposited in the American Museum
of Natural History.

Etymology.— The specific epithet is a patro-
nym honoring Dr. Norman I. Platnick, curator

SISSOM-SYSTEMATICS OF THE VAEJOVIS NITIDULUS GROUP

13

Figures 31-40.— Morphology of Vaejovis platnicki, new species, from Tamaulipas, Mexico: 31, lateral aspect
of metasomal segments IV and V, and telson; 32, dorsal aspect of pedipalp femur; 33, dorsal aspect of pedipalp
patella; 34, external aspect of pedipalp patella; 35, ventral aspect of pedipalp patella; 36, dorsal aspect of pedipalp
chela; 37, external aspect of pedipalp chela; 38, ventral aspect of pedipalp chela; 39, dentition pattern on fixed
finger of pedipalp chela; 40, dentition pattern on movable finger of pedipalp chela.

of Arachnida at the American Museum of Nat-
ural History, for his numerous contributions to
arachnid systematics.

Distribution. — Known only from southern T a-
maulipas and northeastern San Luis Potosi.

Diagnosis.— Adults 20-25 mm in length. Base
color yellow brown to orange brown; carapace,
tergites, and metasoma with strong variegated
pattern. Pectinal tooth count 13-15 in females;
stemite VII with carinae obsolete. Metasomal

14

THE JOURNAL OF ARACHNOLOGY

segments I-IV with ventrolateral and ventral
submedian carinae obsolete; segment V with
ventrolateral carinae weak, granular to crenulate
and ventromedian carina vestigial. Metasomal
segment III length/width 0.77-0.80 (N = 6); V
length/width 1.64-1.77 (N = 6). Pedipalp patella
with 3 esb trichobothria. Pedipalp chela fixed
finger with primary denticle row divided into six
subrows; movable finger with six subrows and
seven inner accessory granules. Chela manus with
dorsal marginal and dorsointemal carinae weak,
granular; other carinae obsolete. Chela length/
width 3.32-3.78 (N = 6); fixed finger length/car-
apace length 0.62-0.65 {N = 6); pedipalp femur
length/carapace length 0.73-0.79 {N = 6).

Description. — Based on adult females; mea-
surements appear in Table 2.

Coloration: Base color of body light brown to
orange brown; metasoma darker orange brown,
especially distal segments. Carapace, tergites, and
metasoma with strong variegated pattern; pedi-
palps and legs with less distinct dusky markings.

Prosoma: Anterior margin of carapace weakly
emarginate; median notch vestigial. Carapace
lustrous, sparsely granular.

Mesosoma: Median carina on tergite I obso-
lete, on II-VI weak, granular; tergite VII with
median carina weak, granular and lateral pairs
moderate, crenulate. Pectinal teeth numbering
13-15. Stemites III-VI smooth, sparsely setose,
with suboval stigmata; VII with carinae obsolete.

Metasoma: Segments I-IV; Dorsolateral cari-
nae moderate, crenulate; distalmost denticles en-
larged, spinoid. Lateral supramedian carinae on
I-III strong, crenulate; on IV moderate, granular;
distalmost denticles enlarged, spinoid on I-III,
flared on IV. Lateral inframedian carinae on I
complete, strong, crenulate; on II present on pos-
terior one-half, strong, crenulate; on III present
on posterior one-third, moderate, crenulate; on
IV absent. Ventrolateral carinae essentially ob-
solete (sometimes with a few small distal gran-
ules on I-II); ventral submedian carinae obsolete.
Intercarinal spaces smooth, lustrous. Segment V
(Fig. 31): Dorsolateral carinae moderate, serrate
proximally, granular distally. Lateromedian ca-
rinae obsolete. Ventrolateral carinae weak, gran-
ular to finely crenulate. Ventromedian carina
vestigial, present only on distal one-half, weak,
granular. Intercarinal spaces moderately granu-
lar.

Telson: (Fig. 3 1 ). Ventral aspect of vesicle with
few irregularly spaced punctations; midline with
a few small granules terminating in a subtle.

pointed subaculear tubercle; about 20 pairs of
setae.

Pedipalp: Trichobothrial pattern (Figs. 32-39)
Type C, neobothriotaxic (Vachon 1974); patella
with three esb trichobothria (Fig. 34). Femur (Fig.
32): dorsointemal and ventrointemal carinae
moderate, crenulate; dorsoextemal carina weak,
granular; ventroextemal carina essentially ob-
solete; inner face with about eight larger granules;
dorsal face moderately granular. Patella (Figs.
33-35): Dorsointemal carina moderate, smooth
to weakly crenulate; ventrointemal carina mod-
erate, crenulate; dorsoextemal and ventroexter-
nal carinae obsolete or faint, smooth; inner face
with vestigial basal tubercle and oblique longi-
tudinal carina of about 10 granules. Chela (Figs.
36-40): Dorsal marginal and dorsointemal ca-
rinae weak, granular; others obsolete. Fixed fin-
ger (Fig. 39) with primary row of denticles di-
vided into six subrows by five enlarged denticles;
six inner accessory granules. Movable finger (Fig.
40) with primary denticle row divided into six
subrows by five enlarged denticles; seven inner
accessory granules. Chela fingers ending in slight-
ly enlarged, blade-like terminal denticles which
overlap when chela closed.

Hemispermatophore: Not dissected; males not
available.

Variation.— Pectinal tooth counts in adults and
subadults varied as follows: in females, one comb
with 13 teeth, eight with 14, and seven with 15.
The first instar specimens from Guaylalejo had
pectinal tooth counts of 1 5- 1 5, 1 5-??, and 16-16.
Variation in morphometries is summarized in
the diagnosis.

Comparisons.— This species is quite distinct
from all other nitidulus group species with six
subrows of denticles on the chela fingers and three
esb trichobothria on the pedipalp patella. In ca-
rinal morphology of the metasoma, it is similar
to V. kochi\ for characters to distinguish these
species, see the “Comparisons” section under V.
kochi.

There is a strong superificial resemblance (small
size and variegated pattern) between V. platnicki
and V. bilineatus Pocock, which belongs to a
different species group. The possession of six
subrows of granules on the pedipalp chela fixed
finger, the basal position of chela trichobothria
ib and it, presence of pedipalpal carinae, and
other characters diagnostic for the nitidulus group
serve to distinguish these species.

Specimens examined.— MEXICO: San Luis Potosr.
El Tinieblo, March 1977 (R. W. Mitchell, et al.), 1

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

15

Table 2.— Measurements in mm and pectinal tooth counts of new species of Vaejovis: V. platnicki, V. rubriman-
us, and V. solegladi.

V. platnicki

V. rubrimanus

V. solegladi

Holotype

female

Holotype

male

Paratype

female

Holotype

male

Paratype

female

Total length

25.4

45.0

58.8

41.2

46.1

Carapace length

3.3

5.2

7.1

5.5

5.8

Mesosoma length

8.7

12.7

18.5

11.8

14.8

Metasoma length

10.2

21.1

25.6

18.5

19.7

I length/width

1. 3/2.2

2.9/2.1

3.4/3. 8

2.4/3. 3

2.6/3.5

II length/width

1.4/2. 1

3.6/2.6

3.9/3.6

2.8/3.3

3.0/3.4

III length/ width

1.6/2. 1

3. 7/2.5

4.2/3.4

3.0/3. 2

3. 2/3.4

IV length/width

2.4/2. 1

4.7/2.4

5.9/3.3

4.2/3. 1

4.5/3. 3

V length/width

3. 5/2.1

6.3/2.3

8.2/3. 2

6.1/3. 1

6.4/3.2

Telson length

3.2

6.0

7.6

5.4

5.8

Vesicle length/ width

2.1/1. 3

4.0/2.0

4.8/2.7

1.212.2

3. 8/2.3

Vesicle depth

0.9

1.6

2.3

1.6

1.8

Aculeus length

1.0

2.0

2.8

2.2

2.1

Pedipalp length

9.7

20.5

25.9

18.9

20.6

Femur length/width

2.5/1.0

5.4/1.4

6.9/1.9

5.2/1. 3

5.6/1. 5

Patella length/width

2.8/1. 1

5. 7/1. 7

7. 1/2.2

5.4/1. 3

5. 8/1.6

Chela length/width

4.4/1.2

9.4/2.5

11.9/2.6

8.3/1. 7

9.2/1.8

Chela depth

1.3

2.7

3.0

1.8

2.0

Movable finger length

2.7

6.0

8.0

5.6

6.2

Fixed finger length

2.2

5.0

6.8

4.8

5.2

Pectinal teeth (It/rt)

15-14

27-28

26-26

18-18

18-19

paratype female (WDS). Tamaulipas: Guaylalejo, 18
February 1973 (J. P. Webb), 1 holotype female, 1 para-
type female, 3 paratype first instars (AMNH), 1 para-
type female (WDS), Tampico, no date (Palmer), one
subadult paratype female (USNM), 1 7 mi. S Victoria,
28 December 1947 (no collector), 1 subadult paratype
female (AMNH), 25 km S Cd. Victoria (under rock),
7 January 1987 (J. A. Nilsson), one paratype female
(JAN), km 190, Highway 85, 18 February 1973 (C.
McConnell), 1 paratype female (WDS).

Vaejovis pococki, new species
(Figs. 41-50, 83, 84)

Vaejovis nitidulus nitidulus, Diaz Najera 1964:27; 1975:

30.

Type data. — Holotype female (RS-4288) from
Queretaro, Queretaro, Mexico, 5-23 August 1963
(collector unknown). Deposited in the Museum
National d’Histoire Naturelle, Paris.

Etymology.— The specific name is a patronym
honoring Reginald I. Pocock for his numerous
contributions to scorpion systematics at the turn
of the century.

Distribution.— Known from several localities
in southern San Luis Potosi and western Que-
retaro, Mexico.

Diagnosis.— Adults 40-65 mm in length. Base
color dark orange brown to reddish brown with
faint dusky markings. Stemite VII with lateral
keels faint to obsolete. Pectinal tooth counts 20-
21 in males, 19-21 in females. Metasomal seg-
ment III length/width 1 .0 1 - 1 . 1 2 (V = 9); V length/
width 2.00-2. \9 {N = 9). Metasoma with ven-
trolateral carinae on I-IV obsolete or weak,
smooth; ventral submedian carinae on I-IV ob-
solete. Metasomal segment V with keels weak to
obsolete. Pedipalp patella with 2 esb trichoboth-
ria; fixed finger with primary denticle row divid-
ed into seven subrows; movable finger with seven
subrows and seven inner accessory granules; dor-
sal and external keels of palm obsolete or faint.
Chela length/width 3.68-3.74 in males {N = 2),
4.38-4.86 in females {N= 7); pedipalp chela fixed
finger length/carapace length 0.80-0.88 {N = 9);
pedipalp femur length/carapace length 0.88-0.96
(V=9).

Description. — Based on adults; measurements
are given in Table 1.

Coloration: Carapace and tergites orange brown
to reddish brown with dusky underlying mark-
ings. Metasomal segments I-III orange brown
above, more reddish brown below; IV with ven-

16

THE JOURNAL OF ARACHNOLOGY

Figures 41-50.— Morphology of Vaejovis pococki, new species, from Queretaro, Mexico: 41, lateral aspect of
metasomal segments IV and V, and telson; 42, dorsal aspect of pedipalp femur; 43, dorsal aspect of pedipalp
patella; 44, external aspect of pedipalp patella; 45, ventral aspect of pedipalp patella; 46, dorsal aspect of pedipalp
chela; 47, external aspect of pedipalp chela; 48, ventral aspect of pedipalp chela; 49, dentition pattern on fixed
finger of pedipalp chela; 50, dentition pattern on movable finger of pedipalp chela.

tral and lateral faces reddish brown; V complete-
ly reddish brown. Telson vesicle orange red. Ped-
ipalps: femur yellow brown, patella more orange
brown. Chela darker orange brown to reddish
brown. Legs yellow brown with dusky markings.

Prosoma: Anterior margin of carapace ob-
tusely emarginate. Interocular area finely gran-
ular; remainder of carapace with moderately
dense, coarse granulation.

Mesosoma: Median carina on tergite I obso-
lete, on II-VI faint, smooth; tergite VII with me-
dian carina weak, granular; lateral pairs strong,
crenulate. Pectinal teeth numbering 20-21 in
males, 19-21 in females. Stemites III-VI smooth,
sparsely setose; VII with pair of faint to obsolete
lateral keels.

Metasoma: Segments I-IV: Dorsolateral cari-
nae strong, crenulate to serrate; distalmost den-

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

17

tides enlarged, spinoid. Lateral supramedian ca-
rinae on I-II strong, crenulate; on III strong, finely
crenulate; on IV weak, smooth to finely granular;
distalmost denticles on I-III enlarged, spinoid
and on IV flared. Lateral inframedian carinae on
I complete, strong, crenulate; on II-III present
on posterior one-third, granular; on IV absent.
Ventrolateral carinae obsolete or weak, smooth.
Ventral submedian carinae obsolete. Intercarinal
spaces essentially smooth, ventral faces with nu-
merous setae. Segment V (Fig. 41): Dorsolateral
carinae weak, finely granular; lateromedian ca-
rinae obsolete; ventrolateral and ventromedian
carinae weak, finely crenulate. Intercarinal spac-
es smooth.

Telson: (Fig. 41). Ventral surface of vesicle with
fine granulation, about 1 5 pairs of large setae.

Pedipalp: Trichobothrial pattern (Figs. 42-49)
Type C, orthobothriotaxic (Vachon 1974). Fe-
mur (Fig. 42): carinae strong, granulose; internal
face with about 8 larger granules and several
smaller ones; dorsal face with scattered fine gran-
ulation. Patella (Figs. 43-45) with dorsointemal
and ventrointemal carinae strong, crenulate;
dorsoextemal carina weak, smooth; ventroex-
temal carina weak, granular; internal face with
moderate basal tubercle and oblique longitudinal
carinae of 7-8 large granules. Dorsal face smooth
or finely granular; external face finely granular
(smooth). Chela (Figs. 46-50). Dorsointemal ca-
rinae weak, granular; all other keels obsolete or
very faint. Dentate margin of fixed finger with
primary row of denticles divided into seven sub-
rows by six enlarged granules; six inner accessory
granules (Fig. 49). Dentate margin of movable
finger divided into seven subrows by six larger
denticles; apical subrow with only one or two
granules; seven inner accessory granules (Fig. 50).
Scalloping subtle in male chela fingers.

Hemispermatophore: (Figs. 83, 84). Distal
laminar length/width = 6.25; lamina with dis-
tinct tapering toward distal end; inner lobe rel-
atively broad; basal lobe small, rounded; median
lobe larger, rounded.

Variation.— Only two adult males and seven
adult females were available for study. Variation
in pectinal tooth counts in these specimens is as
follows: in the males, one comb with 20 teeth
and three with 2 1 ; in females, two combs with
19 teeth, six with 20, and six with 21. Variation
in morphometries is summarized in the diag-
nosis.

Comparisons.— Faejovw pococki, in bearing
seven subrows on the chela fixed finger, is most

similar to V. nitidulus and V. mitchelU. For com-
parisons with V. mitchelti, consult the “Com-
parisons” section for that species. Vaejovis po-
cocki may be easily distinguished from V.
nitidulus by its dark reddish brown coloration,
its lower pectinal tooth counts, the possession of
only two esb trichobothria on the pedipalp pa-
tella, and extremely reduced carination of me-
tasoma V.

This species was apparently mistaken for V.
nigrescens and V. nitidulus by earlier authors
(Hoffmann 1931; Diaz Najera 1964, 1975). The
type specimens of V. pococki are apparently the
same ones examined by Diaz Najera (1964, 1975)
and referred by that author to V. nitidulus. Vae,-
Jovis pococki occurs on the eastern side of the
Sierra Madre Occidental and apparently is al-
lopatric with V. nigrescens (which occurs on the
western side of that mountain range). Although
V. pococki is quite similar to V. nigrescens in
coloration, it may be easily distinguished from
that species by the possession of seven subrows
of denticles on the pedipalp chela fingers.

Specimens examined.— MEXICO: Queretaro: Que-
retaro, 5-23 August 1963 (no collector data), 1 holotype
female, 3 paratype females (RS-4288)(MNHN), 8 km
NW Queretaro on border Guanajuato/Queretaro states,
January 1982 (S. A. Minton), 1 paratype male (MEB),
Queretaro (in house), Fall 1978 (S. A. Minton), 1 para-
type female (SAM), Queretaro (in house), early July
1988 (Mrs. M. Cervantes), 1 paratype male, 1 paratype
female (WDS). San Luis Potosv. 32 km S San Luis
Potosi (on vertical face of large boulder, UV light), 24
August 1984 (C. Myers, W. D. Sissom, L. Bom), 1
paratype female (WDS), Villa Hidalgo, 12 March 1977
(R. W. Mitchell), 1 juv. (AMNH-OFF), Alvarez, June-
September 1976 (W. W. Brown), 1 paratype female
(MCZ).

Vaejovis mbrimanus, new species
Figs. 51-60

Type data.— Holotype male from Gruta Sur
de San Bartolo, approximately 3 mi. S Santa Ca-
tarina, Nuevo Leon, Mexico, 3 December 1 966
(T. Raines). Deposited in the American Museum
of Natural History, New York.

Etymology.— The specific epithet is derived
from the Latin words "'ruber", meaning red, and
"manus", meaning hand, which describes the
coloration of the pedipalps in this species.

Distribution.— Known only from the type lo-
cality.

Diagnosis.— Adults 45-60 mm in length. Base
color of body yellow brown; pedipalp femur and

18

THE JOURNAL OF ARACHNOLOGY

Figures 51-60.— Morphology of Vaejovis rubrimanus, new species, from Nuevo Leon, Mexico: 51, lateral
aspect of metasomal segments IV and V, and telson; 52, dorsal aspect of pedipalp femur; 53, dorsal aspect of
pedipalp patella; 54, external aspect of pedipalp patella; 55, ventral aspect of pedipalp patella; 56, dorsal aspect
of pedipalp chela; 57, external aspect of pedipalp chela; 58, ventral aspect of pedipalp chela; 59, dentition pattern
on fixed finger of pedipalp chela; 60, dentition pattern on movable finger of pedipalp chela.

patella light yellow, chela orange red. Pectinal
tooth count 27-28 in males, 24-26 in females.
Metasomal segment III length/width 1 .48 in male,
1.23-1.24 in females; segment V length/width
2.74 in male, 2.60-2.61 in females. Metasoma
with ventral submedian carinae on I-IV present,
granular to finely crenulate; ventrolateral carinae
on I-IV moderate, finely crenulate; segment I
longer than wide in male, wider than long in

female; other segments distinctly longer than wide
in both sexes. Pedipalp patella with 3 esb tricho-
bothria; fixed finger with primary row of denti-
cles divided into six subrows; movable finger
with six subrows and seven inner accessory gran-
ules; dorsal and external keels of chela palm
smooth or finely granular. Scalloping of male
chela fingers moderate. Ratio of chela length/
width 3.76 in male, 4.62-4.67 in females; fixed

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

19

finger length/carapace length 0.96-0.97; femur
length/width 0.97-1.04.

Description.— Based on adults; parenthetical
statements refer to females. Measurements are
given in Table 2.

Coloration: Carapace and tergites yellow brown
without underlying dusky markings. Metasomal
segments I-IV yellow brown, IV slightly darker
on distal portion; V orange brown. Telson light
orange brown, aculeus dark reddish brown. Ped-
ipalp femur and patella uniformly yellow. Chela
palm yellowish proximally; inner surface yellow
orange; outer surface orange red. Fingers orange
brown basally, yellowish distally. Keels of ped-
ipalps and metasoma dark yellow brown. Legs
pale yellow.

Prosoma: Anterior margin of carapace ob-
tusely emarginate. Median ocular prominence
moderately raised above carapacial surface. In-
terocular area essentially smooth; remainder of
carapace with dense, fine granulation inter-
spersed with larger granules.

Mesosoma: Median carina on tergites I-III
weak, granular; on IV- VI moderate, granular;
tergite VII with median carina moderate, gran-
ular and lateral pairs strong, granulose. Pectinal
teeth numbering 27-28 in males, 24-26 in fe-
males. Stemites III-VI smooth, sparsely setose;
VII with pair of strong, crenulate lateral carinae.

Metasoma: Segments I-IV: Dorsolateral cari-
nae strong, crenulate; distalmost denticles slight-
ly enlarged on I-III, not enlarged on IV (Fig. 51).
Lateral supramedian carinae strong on I-IV,
crenulate on I-III, finely crenulate on IV; distal-
most denticles on I-III roughly equal in size to
preceding ones, on IV flared. Lateral inframedian
carinae on I strong, complete, irregularly cren-
ulate; on II present on posterior one-fourth,
strong, crenulate; on III present on posterior one-
fifth, strong, crenulate; on IV absent. Ventrolat-
eral carinae on I-IV strong, finely crenulate. Ven-
tral submedian carinae on I weak, finely granu-
lar; on II-IV moderate, finely granular to finely
crenulate. Dorsal and lateral intercarinal spaces
with few scattered coarse granules. Segment V
(Fig. 51): Dorsolateral carinae moderate, gran-
ular to crenulate. Lateromedian carinae mod-
erate, present on anterior three-fourths, irregu-
larly crenulate. Ventrolateral and ventromedian
carinae strong, crenulate. Dorsal and lateral sur-
faces with few scattered coarse granules.

Telson: (Fig. 51). Ventral surface with very
fine, irregular punctations and granulation, 10
pairs large setae.

Pedipalp: Trichobothrial pattern (Figs. 52-59)
Type C, neobothriotaxic; patella with 3 esb tri-
chobothria (Fig. 54). Femur (Fig. 52): carinae
strong, granulose; internal face with 7-8 larger,
pointed granules; dorsal face with sparse fine
granulation. Patella (Figs. 53-55): dorsointemal,
ventrointemai, dorsoextemal, and ventroexter-
nal carinae strong, granulose. Internal face with
moderate basal tubercle and oblique longitudinal
carinae of 6-8 large, subconical granules; dorsal
face finely granular. Chela (Figs. 56-60). Dorsal
marginal carina moderate, granular. Dorsal sec-
ondary and digital carinae weak, smooth to finely
granular. External secondary carina obsolete to
weak, finely granular. Ventroextemal carina
moderate, granular. Ventromedian carina ves-
tigial, weak around base of movable finger, gran-
ular. Ventrointemai carina weak to moderate,
granular. Dorsointemal carina strong, composed
of enlarged, sharp granules. Dentate margin of
fixed finger with primary denticle row divided
into six subrows by five enlarged granules; six
inner accessory granules (Fig. 59). Dentate mar-
gin of movable finger with primary denticle row
divided into six subrows by five enlarged gran-
ules; apical subrow consisting of a single granule;
seven inner accessory granules (Fig. 60). Fingers
of male with distinct scalloping.

Hemispermatophore: Not dissected due to
scarcity of material.

Variation.— The holotype male had a pectinal
tooth count of 26-27; the two paratype females
had counts of 24-24 and 26-26. Variation in
morphometries, based on the holotype male and
two adult paratype females, is summarized in the
diagnosis.

Comparisons.— Fhe/'ovri rubrimanus is most
similar to V. minckleyi Williams. From this spe-
cies it can be easily distinguished by the following
characters: (1) the outer keels of the pedipalp
chela palm are smooth to finely granular (not
granulose); (2) the entire outer chela surface is
reddish orange (not with the palm yellow and
fingers reddish brown); (3) the inferior margin of
the cheliceral fixed finger is smooth (not with
several denticles); (4) chela length/width ratio in
males is approximately 3.8 (not exceeding 4.6)
and in females 4. 6-4. 7 (not exceeding 6.0); and
(5) the metasoma is not as slender.

Comments.— The holotype male was taken
from a cave in Huasteca Canyon, near Monter-
rey, Nuevo Leon, but its occurrence in that hab-
itat is certainly accidental (Mr. J. R. Reddell,
pers. comm.). The two females were taken from

20

THE JOURNAL OF ARACHNOLOGY

steep slopes in the canyon with UV light. This
canyon is characterized by impressive vertical
walls reaching approximately 300 m in height,
at the base of which are talus slopes ranging to
about 60 degrees. Both specimens were taken
about 50-100 m from the base of such a slope.

Three other species were collected on the same
slope as the paratypes: Centruroides vittatus (Say),
Diplocentrus colwelH Sissom, and Vaejovis cras-
simanus Pocock. Of the species collected, the
diplocentrid was the most abundant and V. rub-
rimanus the least abundant. Vaejovis rubrimanus
was not found on the lower slopes, where most
of the other scorpions were taken.

Specimens examined.— MEXICO: Nuevo Leon:
Gruta Sur de San Bartolo, 3 December 1 966 (T. Raines),

I male holotype (AMNH), Canon de Huasteca, 3 mi.
S Santa Catarina, 22 May 1984 (W. D. Sissom, C. S.
Colwell), 2 female paratopotypes (WDS).

Vaejovis solegladi, new species
Figs. 61-70, 77, 78

Vejovis nitidulus nitidulus, Hoffmann 1931:371-372
(misidentification); 1939:318 (misidentification).
Vaejovis nitidulus nitidulus, Diaz Najera 1 975:29 (mis-
identification repeated).

Type data.— Holotype male from Cuicatlan,
Oaxaca, Mexico (no date or collector), C. C.
Hoffmann Collection. Deposited in the Ameri-
can Museum of Natural History, New York.

Etymology.— The specific epithet is a patro-
nym honoring Michael E. Soleglad for his con-
tributions to vaejovid systematics.

Distribution.— Known from the Tehuacan area
in Puebla and from northern and central Oaxaca,
Mexico.

Diagnosis. — Adults 38-55 mm in length. Base
color yellow, without dusky markings on cara-
pace and tergites. Stemite VII with lateral carinae
moderate, granular. Pectinal tooth count 18-22
in males, 1 8-20 in females. Metasomal segment
III length/width 0.95-1.00; segment V length/
width 1.90-2.02. Metasoma with inframedian
carinae present on distal two-thirds of segment

II and distal one-half of segment III; ventral sub-
median carinae on I-IV obsolete; ventrolateral
carinae on I-IV moderate, granular; ventral sur-
faces of metasoma and telson very hirsute. Ped-
ipalp patella with 2 esb trichobothria; chela ma-
nus with outer carinae greatly reduced, smooth
to finely granular or obsolete; fixed finger dentate
margin with six subrows of denticles; movable
finger with six subrows and seven inner accessory

granules. Chela very slender with elongate fin-
gers; chela fingers of male without scalloping.
Chela length/width ratio 4. 8-5. 3; chela length/
palm length ratio 2.75-3.0; fixed finger length/
carapace length 0.84-0.93; femur length/cara-
pace length 0.93-0.96.

Description.— Based on adults; measurements
of the holotype male and a paratype female are
given in Table 2.

Coloration: Carapace, tergites, and venter yel-
low to yellow brown, lacking underlying dusky
markings. Metasomal segments I-III yellowish;
IV yellowish above, orange brown with dusky
markings below; V yellow proximally, orange
brown to brown distally with dusky markings.
Telson vesicle orange brown with dusky mark-
ings on ventral surface. Pedipalps: femur and
patella yellowish, chela manus yellow to yellow
orange with fingers somewhat darker; carination
dark orange brown. Legs pale yellow.

Prosoma: Anterior margin of carapace ob-
tusely emarginate, median notch weak. Inter-
ocular area smooth or finely granular; remainder
of carapace densely, coarsely granular.

Mesosoma: Median carina on I-II obsolete, on
III-IV weak, granular; on V-VI moderate, gran-
ular. Tergite VII with median carina moderate,
granular; lateral pairs strong, granulose. Pectinal
tooth count 18-22 in males; 18-20 in females.
Stemites III-VI smooth, moderately setose; VII
with one pair moderate, granular lateral carinae.

Metasoma: Dorsolateral carinae strong, cren-
ulate to serrate; distalmost denticles on I-III
slightly enlarged, spinoid; on IV not noticeably
enlarged. Lateral supramedian carinae on I-III
strong, crenulate; on IV moderate, irregularly
granular; distalmost denticle on I-III enlarged,
spinoid and on IV flared. Lateral inframedian
carinae on I complete, strong, crenulate; on II
present on posterior two-thirds, moderate, cren-
ulate; on III present on posterior one-half, weak,
crenulate; on IV absent. Ventrolateral carinae
moderate, granular. Ventral submedian carinae
obsolete. Intercarinal spaces with scattered coarse
granulation; setae of ventral surface moderately
dense, not paired along ventral submedian ca-
rinae. Segment V (Fig. 61): Dorsolateral carinae
weak, granular on anterior one-half, smooth pos-
teriorly; lateromedian carinae essentially obso-
lete; ventrolateral and ventromedian carinae
weak, finely crenulate. Dorsal and lateral inter-
carinal spaces with scattered coarse granulation;
ventral surface smooth, moderately setose.

Telson: (Fig. 61). Ventral surface with irregular

Figures 61-70.— Morphology of Vaejovis solegladi, new species, from Oaxaca, Mexico: 61, lateral aspect of
metasomal segments IV and V, and telson; 62, dorsal aspect of pedipalp femur; 63, dorsal aspect of pedipalp
patella; 64, external aspect of pedipalp patella; 65, ventral aspect of pedipalp patella; 66, dorsal aspect of pedipalp
chela; 67, external aspect of pedipalp chela; 68, ventral aspect of pedipalp chela; 69, dentition pattern on fixed
finger of pedipalp chela; 70, dentition pattern on movable finger of pedipalp chela.

punctations interspersed with fine granules; ves-
icle of female moderately globose; ventral aspect
conspicuously hirsute, with 50 or more setae.

Pedipalp: Trichobothrial pattern (Figs. 62-69)
Type C, orthobothriotaxic (Vachon 1974). Fe-
mur (Fig. 62): carinae strong, granulose; internal
face with 7-9 larger granules and several smaller

ones; dorsal face essentially smooth. Patella (Figs.
63-65): dorsointemal and ventrointemal carinae
strong, granulose; dorsoextemal carina weak to
moderate, finely granular; ventroexteraal carina
moderate, granulose. Internal face with moder-
ate basal tubercle and oblique longitudinal carina
of 9-1 1 large, irregularly spaced granules; dorsal

22

THE JOURNAL OF ARACHNOLOGY

face more or less smooth. Chela (Figs. 66-70).
Dorsal marginal carina weak, granular. Dorsal
secondary and digital carinae weak, smooth. Ex-
ternal secondary carina obsolete. Ventroextemal
carina obsolete to weak, smooth. Ventromedian
carina obsolete. Ventrointemal carina weak,
granular. Dorsointemal carina moderate, com-
posed of enlarged, sharp granules. Dentate mar-
gin of fixed finger with primary denticle row di-
vided into six subrows by five enlarged granules;
six inner accessory granules (Fig. 69). Dentate
margin of movable finger with primary denticle
row divided into six subrows by five enlarged
granules; seven inner accessory granules (Fig. 70).
Fingers long and tenuous, ratio of movable finger
length/palm length 1. 8-2.1; fingers of male with-
out scalloping.

Hemispermatophore: (Figs. 77-78). Distal
lamina of average proportions (distal laminar
length/width = 6.09, N = \), not distinctly ta-
pered. Median lobe relatively large, rounded.

Variation.— Variation in pectinal tooth counts
is summarized as follows: in males, two combs
with 1 8 teeth, two with 20, one with 2 1 , and one
with 22; in females, two combs with 18 teeth,
two combs with 19, and two combs with 20.
Morphometric variation, based on the three adult
males and three adult females, is summarized in
the diagnosis.

Sexual differences, except in body size, are not
conspicuous. Keel structure and granulation,
which typically exhibit considerable sexual vari-
ation in this species group, are not noticeably
different in males and females of V. solegladi.
Morphometries are also quite similar, although
sample sizes do not permit statistical analysis.
Juveniles differ considerably from the adults in
coloration. Instead of yellow, their base colora-
tion is brownish with distinct underlying dusky
markings on all cuticular surfaces.

Comparisons. — Vaejovis solegladi is most sim-
ilar to V. intermedins and V. nigrescens. It can
be readily distinguished from V. intermedins by
the following characters: (1) males of V. solegladi
have pectinal tooth counts of 18-22 (not 21-26),
females 18-20 (not 19-24); (2) males lack scal-
loping on the pedipalp chela fingers; (3) the lat-
eral inframedian carinae on metasomal segments
II and III are more complete than in V. inter-
medins, extending two-thirds to one-half the
length of their respective segments; (4) the dis-
talmost denticles of the dorsolateral carinae of
the metasoma are not distinctly enlarged, as in
V. intermedins’, (5) the pedipalp chela fingers are

proportionately longer; and (6) the chela palm is
more slender. The hemispermatophore of V. so-
legladi has a proportionately stouter distal lam-
ina and a different configuration of dorsoectal
lobes than found in V. intermedins (cf Figs. 73
and 77).

Characters 2, 3, 4 and (with few exceptions) 6
also serve to distinguish V. solegladi from V.
nigrescens. In addition, the body color of V. so-
legladi is yellow, whereas that of V. nigrescens is
dark reddish brown, and the ventral surfaces of
the metasoma and telson are much more hirsute
than in V. nigrescens. The configuration of dor-
soectal hemispermatophoric lobes also serves to
distinguish the two species (cf. Figs. 75 and 77).

This species was previously mistaken for V.
nit idulns {Hoffmann 1931). Vaejovis solegladi is
readily distinguished from V. nitidulns by having
only six subrows of denticles (not seven) on the
pedipalp chela fixed finger; two patellar esb tri-
chobothria (not three); and 18-22 pectinal teeth
(not 24-28 in males and 21-27 in females).

Specimens examined.— MEXICO: Oaxaca’. Cuica-
tlan (no date or collector), 1 holotype male, 1 paratype
male, 1 paratype female (C. C. Hoffmann collection;
now AMNH), Cuicatlan (in house), 1931 (no collector),
1 juv. (AMNH), 5.8 mi. N Teotitl^, 31 July 1973 (L.
R. Erickson, M. E. Soleglad), 2 paratype females (MES,
cat. no. MX- 131), 30 mi. N Telixtlahuaca, 14 August
1 967 (J. Reddefi, J. Fish, T. Evans), 1 juv. male (AMNH).
Puebla; 6 km N Tehuac^, 22 August 1987 (J. Doyen),
1 male (UCB).

Vaejovis nitidulns C. L. Koch

Previous confusion surrounding the identity
of V. nitidnlus was discussed earlier (Sissom and
Francke 1 985), but new informaton is now avail-
able regarding some previous records. Hoff-
mann’s (1931) records of V. nitidulns in Oaxaca
were based on misidentifications; the taxon found
there represents a new species, V. solegladi, de-
scribed above. The two specimens from Etla,
Oaxaca listed by Bucherl (1959) as V. nitidulns
are juveniles of a species of Diplocentrns (Diplo-
centridae). Vaejovis nitidnlus seems to be re-
stricted to the Sierra Madre Oriental in Hidalgo
and northeastern Queretaro. Diaz Najera’s (1964,
1975) records for V. nitidulns in Guanajuato and
Queretaro (city), along the eastern side of the
Sierra Madre Occidental, are almost certainly
misidentifications. Two other species occur in
that area: V. nigrescens and V. pococki.

The hemispermatophore of V. nitidulns (Figs.
81-82) is characterized by having a relatively

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

23

Figures 71-78.— Right hemispermatophores of species of the Vaejovis nitidulus 71, 72, V. curvidigitus\
71, dorsal aspect; 72, ventral aspect; 73, 74, V. intermedius\ 73, dorsal aspect; 74, ventral aspect; 75, 76, V.
nigrescens; 75, dorsal aspect; 76, ventral aspect; 77, 78, V. solegtadi; 77, dorsal aspect; 78 ventral aspect
(composite drawing based on both left and right hemispermatophores). bl = basal lobe; BP = basal portion; c
= sperm canal; DL = distal lamina; dt = dorsal trough; ep = ental process of inner lobe; ebp = ectobasal process
of inner lobe; H = hooks; il = inner lobe; ml = median lobe; mtc = median transverse cleavage; ol = outer
lobe; tm = dorsal trough margin.

broad (distal lamina length/width 5.56-5.60, N
= 2), straight, and untapered lamina; the inner
lobe of the capsule is long and narrow; the basal
and median lobes are rounded.

New records.— MEXICO: Hidalgo-. Ixmiquilpan, July
1963 (collector unknown), 1 male, 1 female, 1 juv.
(MNHN, RS-4072), Jacala, 8-VIII-? (R. Haag), 1 fe-
male (MCZ), Zimapan, July 1963 (collector unknown),
1 male, 1 female (MNHN, RS-4091).

24

THE JOURNAL OF ARACHNOLOGY

Vaejovis nigrescens Pocock

The specimens referable to V. nigrescens that
I have been able to obtain are from the central
Mexican states of Aguascalientes, Distrito Fed-
eral, Guanajuato, Jalisco, Michoacan, and Za-
catecas. Hoffmann (1931, 1937) and Diaz Najera
(1964, 1975) cited records for V. nigrescens from
Hidalgo, Queretaro, and adjacent parts of San
Luis Potosi. These records are almost certainly
based on misidentifications. In western Quere-
taro and southern San Luis Potosi is a new spe-
cies (described above) that superficially resem-
bles V. nigrescens and was confused with it in
the past. In the southern portion of the Sierra
Madre Oriental in the state of Hidalgo, only two
nitidulus group species have been identified: V.
nitidulus and V. kochi. The latter superficially
resembles V. nigrescens in coloration, but differs
from it in morphometries, carination, and tri-
chobothrial pattern.

The hemispermatophore of V. nigrescens (Figs.
75-76) is relatively broad (distal lamina length/
width = 5.70, A = 1), slightly curved, and un-
tapered; the inner lobe of the capsule is long and
relatively broad; the basal and median lobes are
sharply rimmed.

New records.— MEXICO: Jalisco: Arandas (no date
or collector), one male (MNHN, RS-4290), 2 juvs.
(MNHN, RS-4286), Tepatitlan (no date or collector),
1 male, 1 female (MNHN, RS-4289). Michoacan: El
Sabino, April 1928 (H. Faber), 1 male, 1 female (ZMK),
UraapM, 15 July 1941 (Leavenworth) (on side of
house), 1 female (AMNH). Zacatecas: Valparaiso, 16
July 1963 (L. Mazzotti), 1 male (RS-4021)(MNHN).

Vaejovis intermedins Borelli

Vaejovis intermedins is known from south-
western Texas (Brewster, Crockett, Presidio,
Terrell, and Val Verde Counties) and the states
of Chihuahua, Coahuila, Durango, and Nuevo
Leon in Mexico. Attempts to confirm some ear-
lier records of Hoffmann (193 1) and Diaz Najera
( 1 964, 1975) have met with partial success. Hoff-
mann’s specimens from the Sierra de Guadelupe,
Distrito Federal could not be located, although
they are presumably deposited in the Institute
de Biologia, Mexico City. This record was cer-
tainly based on a misidentification, and the spec-
imens may be referable to V. nigrescens. I have
examined a specimen determined by Diaz Na-
jera as “ Vaejovis nitidulus intermedins" from San
Juan de los Lagos, Jalisco (MNHN, RS-4291)(see
Diaz Najera 1964: 25, 1975:26). This specimen

is not referable to V. intermedins, but rather to
V. cristimanus Pocock {intrepidus group). The
specimens from Ixmiquilpan, Hidalgo (Diaz Na-
jera 1964: 24, 1975: 25) were not located; how-
ever, I have seen only specimens of V. nitidulus
from that locality, and this record is almost cer-
tainly based on that species.

The hemispermatophore of V. intermedins
(Figs. 73-74) bears a long, slender distal lamina
(lamina length/width = 7.05-7.39, V = 3) and
trunk. The lamina is essentially straight and no-
ticeably tapered. The inner lobe of the capsule is
broader at its base; the basal and median lobes
are sharply rimmed.

New records. — MEXICO: Chihuahua: Clarines Mine,
5 mi. NW Santa Barbara (2072 m), 8 February 1947
(G. M. Bradt), 1 juv. female (AMNH). Coahuila: 15
mi. E Cuatro Cienegas de Carranza, 22 July 1972 (E.
A. Liner, R. M. Johnson, A. H. Chaney), 1 female
(FSCA). Nuevo Leon: Bustamente Canyon, Busta-
mente, 26 November 1986 (A. G. Grubbs), 2 females
(TMM), in the mountains 2 mi. NE Villa de Garcia,
19 August 1984 (W. D. Sissom, C. Myers, L. Bom), 6
males, 2 females (WDS); 9 mi. NNW, 2 mi. N Mina,
15 July 1975 (E. A. Liner, et al.), 1 female, 1 subadult
female (FSCA). U.S.A.: Texas: Brewster Co., Alpine,
22 April 1964 (J. F. Scudday), 1 female (CAS), off
Texas Farm Road 1 70, 9 mi. W Junction of State High-
way 118 (5 mi. W Terlingua), 19 May 1989 (R. N.
Henson), 1 female (RNH), Nugent Peak, Big Bend Na-
tional Park, June 1986 (S. Stockwell), 1 juv. (WDS);
road to Pine Canyon, 24 May 1987 (R. Henson), 1
female (1450 m), 1 juv. (1115 m)(RNH); Jeff Davis
Co., Davis Mountains, 27 June 1990 (W. Vandeven-
der), 1 female (RNH); Presidio Co., 2 mi. W Lajitas,
30 May 1970 (W. Seifert); Val Verde Co., NW side of
Amistad Reservoir (on road cut), 3 mi. E Del Rio, 24
May 1983 (W. D. Sissom, C. S. Colwell, N. McRey-
nolds), 1 female (CAS).

Vaejovis decipiens Hoffmann

Two early instar juvenile specimens, whose
locality data are presented below, are tentatively
referred to this species. They bear the appropri-
ate pedipalp chela dentition, patella external tri-
chobothrial pattern, pectinal tooth counts, and
metasomal carinal morphology; furthermore, in
coloration they are very similar to juveniles of
V. decipiens examined previously (Sissom and
Francke 1985).

New records.— MEXICO: Sonora: Sierra de Ala-
mos, 15-30 Jan 1968 (V. Roth), ljuv. female (AMNH),
Rancho Los Banos (30°30’N:1 10“40’W), 9 May 1966
(V. Roth), 1 juv. female (AMNH).

SISSOM=-SYSTEMATICS OF THE VAEJOVIS NITIDULUS GROUP

25

Figures 79-84.— Right hemispermatophores of species of the Vaejovis nitidulus group: 79, 80, V. mitchelli;
79, dorsal aspect; 80, ventral aspect; 81, 82, V. nitidulus; 81, dorsal aspect; 82, ventral aspect; 83, 84, V. pococki;
83, dorsal aspect; 84, ventral aspect. To identify structures refer to labels of preceding plate.

Vaejovis peninsularis Williams

Williams (1980) placed V. peninsularis in the
wupatkiensis group of Vaejovis, but it was sub-
sequently referred to the Vaejovis nitidulus group
(Sissom and Francke 1985). Williams and Berke
(1986), who resurrected the genus Serradigitus
Stahnke for certain species of the V. wupatkiensis
group, chose to retain V. peninusularis in Vae-

jovis, thus agreeing with Sissom and Francke
(1985). In particular, the possession of six sub-
rows of denticles on the chela fingers, the basal
position of triohobothria ib and it, and the pos-
session of three esb trichobothria on the pedi-
palpal patella are indicators of affinities with the
nitidulus group.

A specimen in the American Museum of Nat-
ural History labeled as a juvenile paratype by

26

THE JOURNAL OF ARACHNOLOGY

Williams is not V. peninsularis, but is referable
to Serradigitus gigantaensis (Williams). Seven
other juveniles of V. peninsularis collected by
Vince Roth at Mission San Ignacio on the same
date as the paratypes were also located in the
AMNH.

New records.— MEXICO: Baja California Sur: in
rockslide ca. 1.5 mi. from Pie de la Cueta Ranch on
trail to Guajademi (S of El Portrero), 23 Oct 1972 (D.
B. Richman, R. Reeder, P. D. Eliscu), 1 male, 1 female
(FSCA), under rock near Pie de la Cueta (between El
Portrero and Guajademi), 22 Oct 1972 (D. B. Rich-
man, et al.).

THE UTILIZATION OF HEMISPERMATO-

PHORE MORPHOLOGY IN VAEJOVID
SYSTEMATICS

The vaejovid spermatophore (and, conse-
quently, hemispermatophore), like that found in
most scorpion families, is lamelliform (Francke
1979; Lamoral 1979). Lamelliform hemisper-
matophores are characterized by the possession
of a basal trunk area and a distal blade-like struc-
ture referred to as the distal lamina (e.g., see Fig.
71). Near the junction of the trunk and distal
lamina on the ventral and ental (or medial) sur-
faces may be a system of lobes and processes
often referred to as the capsule (Francke 1979),
although there is considerable variation in the
complexity of this region. The terminology for
these lobes was reviewed by Lamoral (1979), and
I have attempted to apply his nomenclature to
vaejovid hemispermatophores. Vaejovid hem-
ispermatophores differ considerably from those
of scorpionids in the structure of the capsular
region, so the present interpretation of homol-
ogies should be considered tentative until com-
parative phylogenetic studies, which are in prog-
ress, can be completed. Illustrations of
hemispermatophores of seven species of the Vae-
jovis nitidulus group are presented together here
(Figs. 71-84) to facilitate comparisons.

There is considerable variation among vae-
jovids in the morphology of the hemisperma-
tophore, and this variation should prove ex-
tremely useful to systematics at the generic,
species group, and specific levels. The morphol-
ogy of the distal lamina differs considerably
among the different vaejovid groups. The rela-
tive length and slenderness of the blade, as well
as the degree of tapering towards the distal end,
differs among vaejovid species. For this study, I
have indicated relative slenderness of the blade
as the ratio between distal laminar length (mea-

sured from the base of the dorsal trough to the
tip of the lamina) and laminar width at mid-
length. Although based on the small sample sizes
available here, this ratio appears to be relatively
constant.

In vaejovids, the ental (= medial) margin of
the dorsal trough is usually produced distally into
some type of sclerotized structure. The different
types of structures present here may prove to
have considerable taxonomic value above the
species level. In all species of the Vaejovis niti-
dulus group thus far examined, this structure takes
the form of a pair of hooks (e.g.. Fig. 71) that
is always located basally on the dorsoental sur-
face of the distal lamina. I have observed com-
parable double hooks on hemispermatophores
of at least some representatives of other groups,
such as the Vaejovis mexicanus group (Sissom
1989a), a few Uroctonus {sensu Soleglad 1973),
Vejovoidus, and Paruroctonus. Some mexicanus
group species, such as V. mexicanus Koch, V.
granulatus Pocock, and V. maculosus Sissom lack
hooks altogether (Sissom 1989a). The hemisper-
matophores of other vaejovid groups examined
(i.e., Syntropis, Serradigitus, and the Vaejovis
eusthenura, punctipalpi, and intrepidus groups)
bear a broad flange along the ental margin of the
distal lamina which may be bluntly bifurcate dis-
tally, possibly representing a condition derived
from that seen in the aforementioned groups.
This flange typically terminates some distance
from the base of the lamina.

The capsular region is highly variable among
vaejovids, ranging from a very simple structure
to a highly developed system of lobes and pro-
cesses. The basal portion of the capsule is oc-
cupied by a folded canal that may function in
sperm transport. The inner lobe, when present,
is usually a fingerlike lobe projecting distally to-
wards the distal laminar base (very similar to the
condition found in scorpionids). It bears an ental
process (e.g., see Fig. 72) that may possess a series
of booklets and a curved, flange-like ectobasal
process. The ental process bears booklets in Syn-
tropis and the eusthenura, punctipalpi, and in-
trepidus groups of Vaejovis, but does not in spe-
cies of other groups, including the nitidulus group.
In the former case, the number of booklets ap-
pears to vary according to species and has re-
cently been used to separate species (Sissom,
1989b); these booklets were referred to as “cap-
sular spines” in that paper, but the former term
seems more appropriate. In the nitidulus group
species, the number and shape of lobes com-

SISSOM- SYSTEM ATICS OF THE VAEJOVIS NITIDULUS GROUP

27

prising the dorsoental portion of the capsule is
variable. For example, the median and basal lobes
may be produced into a distinct rim, appearing
pointed in the dorsal and ventral views (Figs. 73-
76), or they may be gently rounded (Figs. 79-
84). Two closely related species (perhaps sister
species), V. nigrescens and V. intermedins, pos-
sess median and basal lobes that are sharply
rimmed; however, the lobes are not so rimmed
in closely related species, V. curvidigitus and V.
solegladi. It is also interesting to note that round-
ed lobes occur in V. mitchelli, V. nitidulus, and
V. pococki] these three species are apparently close
relatives as well, based on their external anatomy
(see comparisons sections for these species). Fi-
nally, in addition to the median, basal, and outer
lobes, accessory lobes may be present on the dor-
soental aspect (e.g., V. curvidigitus, V. nigrescens,

V. nitidulus, and V. pococki).

Without doubt, as our understanding of hem-
ispermatophoric structure in vaejovids and the
related chactoid groups increases, many new
characteristics beyond the few mentioned here
will prove to be of taxonomic and phylogenetic
value.

ACKNOWLEDGMENTS

Numerous curators and institutions provided
assistance during the course of this study, either
through loans of specimens or by searching
through their collections for pertinent material.
For the loan of material, including type speci-
mens, I wish to thank E.-G. Burmeister, Zool-
ogische Staatssammlung, Munich (ZSM); G. B.
Edwards, the Florida State Collection of Arthro-
pods (FSCA), Gainesville; O. Elter, Museo ed
Istituto di Zoologia Sistematica della Universita
di Torino (TOR), Torino, Italy; H. W. Levi, the
Museum of Comparative Zoology (MCZ), Har-
vard University, Cambridge, Mass.; W. R. Lour-
engo and M. Vachon, the Museum National
d’Histoire Naturelle (MNHN), Paris; M. Moritz,
Zoologisches Museum der Humboldt-Univer-
sitat (ZMB), Berlin; N. I. Platnick, the American
Museum of Natural History (AMNH), New York;

W. J. Pulawski and V. F. Lee, the California
Academy of Sciences (CAS), San Francisco; J.
Reddell, the Texas Memorial Museum (TMM),
the University of Texas, Austin; S. A. Stockwell,
University of California at Berkeley (UCB); F.
R. Wanless and P. D. Hillyard, the British Mu-
seum (Natural History) (BMNH), London; and
L. Watrous, J. B. Kethley, and D. A. Summers,
the Field Museum of Natural History (FMNH),

Chicago. The following individuals kindly loaned
me specimens from their personal collections: M.
E. Braunwalder (MEB), O. F. Francke (OFF-
AMNH), R. N. Henson (RNH), S. A. Minton
(SAM), J. A. Nilsson (JAN), and M. E. Soleglad
(MES). I am extremely grateful for their assis-
tance. O. F. Francke also allowed me access to
specimens he had on loan from the Zoologisk
Museum, Copenhagen (ZMK).

For permission to collect in Big Bend National
Park, I am grateful to M. Fleming and the Na-
tional Park Service, Department of the Interior.
J. Reddell provided some information on the
cave record of V. rubrimanus and other localities
in Mexico. I also wish to extend a heartfelt thanks
to C. S. Colwell and C. and L. (nee Bom) Myers,
who accompanied me in the field.

L. P. Alberstadt, V. Fet, O. Francke, R. Krai,
D. E. McCauley, M. F. Miller, T. L. Page, G. A.
Polls, and S. A. Stockwell read various drafts of
the manuscript; their comments and suggestions
are sincerely appreciated. C. Myers, P. Malone,
and S. Hughes provided clerical assistance.

This research represents partial fulfillment of
the requirements for the Ph.D. degree at Van-
derbilt University and was partially supported
by a grant from the graduate school of that in-
stitution. Page charges were paid with the assis-
tance of Faculty Research and Development
Grants from Elon College.

LITERATURE CITED

Biicherl, W. 1959. Escorpioes e escorpionism no Bra-
sil. X. Catalogo da colegao escorpionica do Institute
Butantan. Mem. Inst. Butantan, 29:255-275.

Diaz Najera, A. 1964. Alacranes de la Republica
Mexicana: Identificacion de ejemplares capturados
en 235 localidades. Rev. Inst. Salubr. Enferm. Trop.,
Mexico, 24:12-30.

Diaz Najera, A. 1975. Listas y dates de distribucidn
geografica de los alacranes de Mexico (Scorpionida).
Rev. Inst. Salubr. Enferm. Trop., Mexico, 35:1-36.
Francke, O. F. 1977. Scorpions of the genus Diplo-
centrus from Oaxaca, Mexico (Scorpionida, Diplo-
centridae). J. ArachnoL, 4:145-200.

Francke, O. F. 1979. Spermatophores of some North
American scorpions (Arachnida, Scorpiones). J. Ar-
achnol., 7:19-32.

Hoffmann, C. C. 1931. Los Scorpiones de Mexico.
Primera parte: Diplocentridae, Chactidae, Vejovi-
dae. An. Inst. Biol., Mexico, 2:291-408.

Hoffmann, C. C. 1937. Nota acerca de los alacranes
del Valle de Mezquital, Hgo. An. Inst. Biol., Mexico,
8:201-206.

Hoffmann, C. C. 1939. Nuevas consideraciones ac-

28

THE JOURNAL OF ARACHNOLOGY

erca de los alacranes de Mexico. An. Inst. Biol.,
Mexico, 9:317-337.

Lamoral, B. H. 1979. The scorpions of Namibia
(Arachnida: Scorpionida). Ann. Natal Mus., 23(3):
497-784.

Sissom, W. D. 1989a. Systematic studies on

granulatus Pocock and Vaejovis pmillus Pocock, with
descriptions of six new related species (Scorpiones,
Vaejovidae). Revue Arachnol., 8(9): 13 1-1 57.

Sissom, W. D. 1989b. Redescription of Faeyovw oc-
cidentalis Hoffmann with a revised diagnosis for
Vaejovis subcristatus Pocock (Scorpiones, Vaejovi-
dae). Revue Arachnol., 8(1 1): 179-1 87.

Sissom, W. D. & O. F. Francke. 1985. Redescriptions
of some poorly known species of the nitidulus group
of the genus Vaejovis (Scorpiones, Vaejovidae). J.
Arachnol., 13:243-266.

Sissom, W. D., G. A. Polis, & D. D. Watt. 1 990. Field
and laboratory methods. In The Biology of Scor-

pions. (G. A. Polis, ed.). Stanford University Press,
Stanford, CA.

Soleglad, M. E. 1973. Scorpions of the mexicanus
group of the genus Vejovis (Scorpionida, Vejovidae).
Wasmann J. Biology, 31:351-372.

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

Vachon, M. 1974. Etude des caractdres utilises pour
classer les families et les genres de Scorpions (Ar-
achnides). Bull. Mus. natn. Hist, nat., Paris, 3e Sen,
No. 140, Zool. 104:857-958.

Williams, S. C. 1980. Scorpions of Baja California,
Mexico, and adjacent islands. Occ. Pap. California
Acad. Sci., 135:1-127.

Williams, S. C. & B. T. Berke. 1986. A new species
of Serradigitus from central California (Scorpiones:
Vaejovidae). Pan-Pacific Entomol., 62:350-354.

Manuscript received January 1989, revised July 1990.

1991. The Journal of Arachnology 19:29-39

A NEW SPECIES OF WOLF SPIDER,
SCHIZOCOSA STRIDULANS (ARANEAE, LYCOSIDAE)

Gail E. Stratton: Department of Biology, Albion College, Albion, MI 49224 USA

Abstract. Schizocosa stridulans new species is a sibling species to S. ocreata and S.
rovneri. Both males and females of S. stridulans are very similar to males and females of
S. ocreata and S. rovneri in coloration and genitalia, but are significantly smaller in carapace
length and width. Mature males of S. stridulans lack a distinctive tuft of bristles on the tibia
of the first pair of legs (present in mature males of S. ocreata, absent in mature males of 5'.
rovneri)-, however, the tibia, patella and the distal 1/3 to 1/2 of the femur of legs I of males
of S. stridulans are darkly pigmented. S. stridulans is found in mesic uplands leaf litter from
Tennessee, Kentucky, Illinois, Ohio, Missouri, Mississippi, and Alabama, and sometimes
co-occurs with S. ocreata. Male palps and female epigyna are figured for S. stridulans and
for S. rovneri for the first time.

The genus Schizocosa Chamberlin consists of
medium sized to large wolf spiders that are rel-
atively strong legged and keen sighted. Members
of this genus are characterized by conspicuous
and contrasting light and dark bands on the car-
apace and abdomen, a distinct embolus and ter-
minal apophysis in the palp of the male (Fig. 1),
and an excavated transverse piece in the median
septum in the epigynum of the female (Fig. 5)
(Dondale & Redner 1978). Schizocosa ocreata
(Hentz) is a member of this genus common
throughout woodlands of the eastern United
States. It is frequently called the brush-legged
spider because of conspicuous tufts of black bris-
tles and black pigmentation present on the tibia,
patella and basitarsus of the first pair of legs in
mature males (Fig. 11). In their revision of the
genus, Dondale and Redner (1978) noted that
there were occasional populations in which the
tufts of bristles were reduced or absent. A form
lacking tufts and black pigment on the first legs
of mature males has a distinct courtship and is
recognized as a distinct species {Schizocosa rov-
neri Uetz and Dondale 1979). S. ocreata and S.
rovneri do not interbreed unless the female is
anesthetized; thus, their dilfering courtship be-
havior serves as an isolating mechanism between
the two species (Uetz & Denterlein 1979). Fur-
ther studies by Stratton and Uetz (1981, 1983)
demonstrated that the two species are interfertile
when forced to mate.

A new species of Schizocosa is described and
figured herein that is sibling to both S. ocreata
and S. rovneri. Mature males of this new species

lack the bristles on the first pair of legs, but do
have conspicuous pigment on the distal 1/3 of
the femur and on the tibia. While at first it was
thought they may be hybrids between S. ocreata
and S. rovneri (Dondale, pers. comm.), a com-
parison of the morphology and behavior of
ocreata- rovneri hybrids (Stratton & Uetz 1983,
1986), with these clearly demonstrate that these
forms are not hybrids but are a distinct species.

METHODS

The anatomical description of 5. stridulans is
based on mature males and mature females. Spi-
ders collected as immatures were reared to ma-
turity in the laboratory. Anatomical terminology
follows that of Dondale and Redner (1978).

Scanning electron micrographs were done on
a JEOL JSM T200 scanning electron microscope
at 10 kv. Samples were prepared by cleaning ul-
trasonically for 3 min and then running samples
through a dehydrating series of alcohol dilutions.
They were air dried and mounted with silver
paint on SEM stubs. Internal aspects of females
were first cleared for 30 minutes in enzymatic
solution (contact lens cleaner: 1 0 mg/ 1 0 mis dis-
tilled water), dehydrated in alcohol dilutions, then
air dried and mounted.

In order to investigate patterns of co-occur-
rence and potential for overlap with related spe-
cies, collections of Schizocosa were made in mid-
west and southern USA forests from March to
July 1983-1986. Special emphasis was placed on
collecting from floodplain forests along major
rivers, and their corresponding uplands. Collec-

30

THE JOURNAL OF ARACHNOLOGY

lions from the Mississippi State Museum and
the Museum of Comparative Zoology were also
examined. In all collections, mixed assemblages
of species were noted.

Schizocosa stridulans, new species
Figs. 1, 5, 6, 13

Type Material.— Male holotype from Illinois,
Mason Co., Sand Ridge State Forest, June 1985
(G. Stratton and L. Hartz), deposited at the Mu-
seum of Comparative Zoology (MCZ), Harvard
University.

Etymology.— The species name refers to the
primary method of sound production by males
during courtship behavior.

Diagnosis. — 5. stridulans is significantly
smaller than either S. ocreata (as reported by
Dondale & Redner 1978) or S. rovneri (as re-
ported by Uetz & Dondale 1979; Table 1), al-
though the overlap in sizes of these three species
makes size an unreliable character (Table 1). Both
males and females are indistinguishable anatom-
ically from those of S. ocreata and S. rovneri
except for the pattern of pigmentation on the first
pair of legs of mature males. Both sexes key to
S. ocreata in the key provided in Dondale and
Redner’s (1978) revision of the genus. Females
can be confidently identified only when collected
in association with males. In males of S. stri-
dulans, the tibia, patella and distal 1/3 to 1/2 of
the femur are black (Fig. 1 3). There are fine black
hairs on the tibia of male S. stridulans distinct
from the tibial tufts of bristles found in the ma-
ture male S. ocreata (Fig. 1 1). Mature males of
S. rovneri lack both the tufts of bristles and the
solid pigmentation on the tibia of legs I (Fig. 1 2),
although these legs may be annulated. Males of
5”. stridulans, S. rovneri and 5. ocreata are iden-
tical with respect to length and angle of paleal
process of palp, median apophysis and with re-
spect to rugose prominence along the retrolateral
side of the paleal process (Figs. 1-3); this com-
pares with the palp of S. crassipes (Fig. 4), a more
southern species that has a smooth prominence
along the retrolateral side of the paleal process.
This last character corresponds with couplet #3,
p. 147 in Dondale and Redner’s (1978) key.

Mature females of 5. stridulans have a slight

darkening of the tibia, patella and basitarsus of
legs I, as compared with their other legs. Females
of S. stridulans, S. rovneri and S. ocreata all have
paired excavations in the transverse piece of the
median septum (Figs. 5, 7, 9). In each of these,
the distance between the surface excavations is
less than the width of one excavation. In females
of S. crassipes, the distance between the exca-
vations is greater than on the width of one ex-
cavation (refer to Dondale «fe Redner’s key to
females, couplet 5, p. 149 (1978)). Spermathecae
of S. stridulans, S. rovneri and S. ocreata are
illustrated in Figs. 6, 8, 10.

Males. —Total length, carapace length and car-
apace width as reported in Table 1. Carapace
brown; pale submarginal band slender, usually
distinct and undulating, rarely extending to car-
apace margins; pale median band as wide as pos-
terior lateral eyes (mean 0.85 mm), with smooth
margins and narrowing slightly in posterior third
of carapace. Sternum yellow brown. Chelicerae
brown, setaceous, with three uneven teeth on
promargin of fang furrow and three even teeth
on retromargin. Legs II to IV yellow with dark
annulations particularly on femur and tibia. Fe-
mur of leg I with black pigmentation on distal
half to third; tibia and patella of leg I usually
uniformly black (rest of leg yellow) (Fig. 1 3). Pig-
mentation on femur sometimes streaked. Tibial
brush in form of short black hairs that increase
the apparent width of the tibia by about 0.2 mm
(width of tibia: 0.54; width of tibia -I- hairs: 0.75;
1 5 specimens measured). Dark areas of leg I with
the appearance of a “five-o-clock shadow”. Dor-
sum of abdomen usually with heart mark (14 of
15 specimens), without chevrons. Cymbium of
palp without terminal macrosetae but with con-
centration of bristles. Palea of palp with long
distal process, and with a furrow marking off
rugose prominence on retrolateral side. Median
apophysis with distal margin convex and un-
dulating. Intromittent part of embolus slender
and pointed. Terminal apophysis with thickened
margin concealing base of intromittent part of
embolus (Fig. 1).

Females.— Total length, carapace length and
carapace width as reported in Table I. Coloration
similar to that of male but with the following

Figures 1-4.— Ventral aspect of left palp of Schizocosa species: 1, S. stridulans-, 1, S. rovneri-, 3, S. ocreata-, 4,
S. crassipes. ipe = intromittent portion of embolus; ma = median apophysis; ppr = paleal process; rp = rugose
prominence; sp = smooth prominence; ta = terminal apophysis. Scale bars = 200 microns.

32

THE JOURNAL OF ARACHNOLOGY

Table 1.— Comparison of total length, carapace length and carapace width of S. stridulans n.sp., S. ocreata
(data from Dondale & Redner 1978), and 5. rovneri (data from Uetz & Dondale 1979). Measurements are in
mm. Where sample size is greater than 10 individuals, carapace measurements are given as means ± their
standard deviations. Values followed by the same letter are not significantly different from each other (1 tailed
t test, P < 0.05).

S. ocreata

S. rovneri

S. stridulans

MALES

Total length (ranges)

(mean)

5.65-8.30

6.48-8.00

5.04-6.80

6.40 ± 0.43

Carapace length (means)

3.65 ± 0.43 A

3.73 A

3.25 ± 0.33 B

(ranges)

3.48-4.02

2.47-3.80

Carapace width (means)

2.78 ± 0.34 C

2.77 C

2.56 ± 0.24 D

(ranges)

2.57-2.95

2.04-3.10

Sample size

20

7

51

FEMALES

Total length (ranges)

(means)

7.30-10.40

6.01-7.95

6.56-11.36

8.09 ± 1.21

Carapace length (means)

4.00 ± 0.43 E

3.91 E

3.50 ± 0.40 F

(ranges)

3.45-4.28

2.63-^.27

Carapace width (means)

3.02 ± 0.31 G

2.93 G

2.68 ± 0.35 H

(ranges)

2.64-3.24

1.88-3.21

Sample size

20

7

61

exceptions. Pale median band on prosoma 0.96
mm wide behind eyes and usually narrowed in
posterior half of carapace. Chelicerae as in male.
Legs I to IV yellow with dark annulations. Tibia,
basi tarsus and occasionally patella of leg I darker
than on other legs, with annulations less distinct.
Epigynum with moderately deep atrium; median
septum with longitudinal piece broad posteriorly
and usually narrowing anteriorly with lateral
edges concave. Transverse piece with large paired
excavations, these excavations nearly meeting at
midline. In 7 of 1 5 individuals, these excavations
asymmetrical in size and sometimes in shape.
Distance between excavations varying from al-
most no space to a separation slightly less than
the width of a single excavation. Spermathecae
ovoid, smooth, separated by approximately their
width.

Courtship behavior.— Males of S. stridulans
clearly differ from S. ocreata and S. rovneri in
sexual behavior. The courtship of S. stridulans
consists of pulses of stridulation of the palp, in-
terspersed with tapping of the first pair of legs.
A full description of the courtship behavior, the
sounds produced during courtship, the variabil-
ity of the various components of the behavior,
and the results of attempted cross matings is in
preparation (Stratton, in prep.). Males of S. stri-

dulans will rarely court females of either S.
ocreata or S. rovneri. Females of the other spe-
cies are not receptive to courting males of S.
stridulans. Females of S. stridulans are not re-
ceptive to courting males of other species.

Geographic distribution and habitat.— Collec-
tions of S. stridulans have been made from
southern Ohio, Illinois, Kentucky, Tennessee,
Missouri, Alabama and Mississippi (Fig. 14), thus
giving it broad geographic overlap with both S.
rovneri and S. ocreata. The habitat of S. stri-
dulans is mesic uplands leaf litter, typically in
oak forests or oak hickory forests (Fig. 1 6). The
present study also extends the known range of
S. rovneri.

In two of eight localities visited in 1984 and
1985, S. stridulans was the only Schizocosa col-
lected in the uplands forests (Figs. 15, 16). In
three collections, 5. stridulans occurred in the
same habitat as S. ocreata (Figs. 15, 16), and in
one collection from Alabama it also occurred
with a population that is possibly an undescribed
species within this species complex (Stratton un-
publ. data). Table 2 summarizes 46 collections
of Schizocosa and indicates whether species were
collected alone, or co-occurred with other species
in the species group which includes 5. ocreata,
S. rovneri, S. stridulans and the southern species

STRATTON-5C///Z<9C05'.4 STRIDULANS, NEW SPECIES

33

Figures 5-10.— External and internal aspects of epigyna of Schizocosa: 5, external aspect of S. stridulans; 7,
external aspect S. rovneri\ 9, external aspect S. ocreata; 6, spermatheca of S. stridulans', 8, spermatheca of S.
rovneri', 10, spermatheca of S. ocreata. at = atrium; ex = excavation; ms = median septum; tp = transverse
piece. Scale bars = 100 microns.

34

THE JOURNAL OF ARACHNOLOGY

Figures 1 1-13.— Legs I of mature males of Schizocosa: 11,5'. ocreata\ 12, S. rovneri\ 13, S. stridulans.

S. crassipes and S . floridana. In most collections
each of these species was found alone, although
S. ocreata and S. rovneri sometimes co-occurred
as did 5. ocreata and S. stridulans. This suggests
that for the species that do co-occur, courtship

Table 2.— Co-occurrence of species within the Schi-
zocosa ocreata species complex. Each entry represents
a separate collection. Collections were by the author,
by Wayne Maddison and from the Mississippi State
Museum.

S.

S.

.S.

5.

5.

oc-

rov-

stri-

eras-

flori-

reata

neri

dulans

sipes

dana

S. ocreata

13

S. rovneri

3

12

S. stridulans

3

0

6

S. crassipes

1

1

1

4

S. floridana

0

0

0

1

1

behavior is potentially important as an isolating
mechanism. This has been studied extensively
for S. ocreata and S. rovneri (Stratton & Uetz
1981, 1983, 1986) but to a more limited extent
in S. stridulans and S. ocreata (Stratton, in prep).

More is known of the habitat preferences for
S. ocreata than for the other species in the genus,
and this preference appears to vary geographi-
cally. Dondale and Redner (1978) report that S.
ocreata tend to be found in moist areas relative
to S. crassipes and S. floridana. In North and
South Carolina, Missouri (Big Oak Tree State
Park), as well as in Sand Ridge State Forest in
central Illinois, S. ocreata was collected on the
floodplains of rivers or near wet areas. For ex-
ample, along the flood plain of the Tyger River
in S. Carolina, S. ocreata appeared to be the most
abundant wolf spider; in Big Oak Tree State Park,
a virgin floodplain forest along the Mississippi
River in Missouri, S. ocreata again appeared to

STRATTON -5C///Z(9CO&4 STRIDULANS, NEW SPECIES

35

be the most abundant wolf spider (Fig. 1 7). Col-
lections in Illinois, Kentucky, and Ohio yielded
S. ocreata from the drier uplands and often on
slopes above major rivers (Fig. 16), while other
species (particularly S. rovneri) were found on
floodplains and bottomlands (Fig. 17). Perhaps
the habitat “preference” of S. ocreata may par-
tially depend on geographic locality (and its many
associated factors) and/or possibly on the pres-

ence or absence of other competing species. Cady
(1983) in a study in south central Ohio, reports
that S. ocreata is closely restricted in its micro-
habitat and that its distribution and locomotor
activity are related to moisture and physical fea-
tures of the microhabitat. Cady found that S.
ocreata was more likely to be found in full leaf
litter rather than in sparse litter, and that the
species preferred areas of high soil moisture. He

36

THE JOURNAL OF ARACHNOLOGY

IL Ml IL MA

Figure 15.— Comparison of relative numbers of 5. stridulans n. sp. and related species in floodplain forest
and uplands forest along major river systems in the U.S. Midwest. The X-axis of each graph shows numbers
of spiders collected/person-hour (most collections were 2-3 person-hours). Collections were done by the author
in 1984 and 1985. From top and clockwise: IL MI = Illinois: Marshall Co., State Fish and Wildlife Area; IL
MA = Illinois: Bureau Co., Miller Anderson Nature Preserve; IL AL = Illinois: Alexander Co.; KY CB =
Kentucky: Hickman Co., Columbus-Belmont Battlefield State Park; AL Ld = Alabama: Lauderdale Co. (along
Tennessee River); MO B.O.T. = Missouri: Mississippi Co., Big Oak Tree State Park; MO TT = Missouri: Cape

STRATTON -^C/Z/ZOCOS/l STRIDULANS, NEW SPECIES

37

a>

■s

a

m

k.

«

n

E

Rivers:

IL IL Mi/IL IL

Key: ^ oereata

0 rovneri

1 I stridulans

^ species “B”

Mo TT

Mi Mo Mi Tn

Figure 16.— Comparison of relative numbers of Schizocosa species collected from uplands habitat. Key as in
Fig. 15.

suggests that microhabitat selection by S. oereata
is important in courtship.

The habitat of S. rovneri is generally floodplain
forests (Uetz & Dondale 1979; Stratton & Uetz
1981, and Fig. 17), although there are reported
collections of 5. rovneri from several upland hab-
itats in the Cincinnati area (G. W. Uetz, pers.
comm.) and in Illinois (Fig. 16). The spiders are
most frequently found in or on flattened mud
packed leaf litter, or in and on piles of drift that
are frequently found in these flood prone eco-
systems.

In Central Illinois (Mason County), S. oereata,
S. rovneri and S. stridulans were all found in
close proximity to each other. S. rovneri was found
in the Chautauqua National Wildlife Refuge,
along the Illinois River. It was also collected in
other floodplain forests along the Illinois River.
A population of S. oereata was found in a swampy
area within the Sand Ridge State Forest. An ad-
jacent area that was slightly higher in elevation
and slightly more mesic yielded S. stridulans.

These populations and their habitats were in-
vestigated in some detail and will be reported
separately (Stratton, in prep).

The distribution patterns within this complex
of three sibling species are intriguing. While S.
oereata and S. rovneri are sympatric and occa-
sionally syntopic, and while S. oereata and S.
stridulans are also occasionally syntopic, S. stri-
dulans and S', rovneri are apparently never syn-
topic. It appears that both S. rovneri and S. stri-
dulans are stenotypic, whereas S. oereata is
comparatively eurytypic. More investigations
with the detail of Cady’s (1983) study are needed
to understand the interactioh of these spiders
with their habitat.

It appears that courtship behavior of these spe-
cies may be very habitat specific. It is hypothe-
sized that the stridulatory component of the
courtship behavior in S. stridulans may be in-
audible (and ineffective) on anything but dry
leaves. The courtship behavior of S. stridulans
is most similar to that of S. crassipes and S.

Girardeau Co., Trail of Tears State Park; MO St C/Bab = Missouri: St. Louis Co., Babler State Park; IL Cl =
Illinois: Calhoun Co., Reds Landing Waterfowl Management Area; IL PM = Illinois: Jersey Co., Pere Marquette
State Park.

38

THE JOURNAL OF ARACHNOLOGY

Rivers:

24-

22-

20-

18-

16

14

12

10-

8-

6

4-1

2

Schizocosa species, Floodplains

IL Al

IL Ml

IL Cl

IL MA

MoStC

Key: rovneri

ocreata

IL

IL

IL

Mi/IL

Mo

Mi

Mi

Ty

Figure 17.— Comparison of relative numbers of Schizocosa species collected from floodplain habitats. Key as
in Fig. 15.

floridana (Stratton in prep), both of which are
also restricted to mesic habitats (Dondale & Red-
ner 1978; Stratton, unpubl. data). Through this
study and others, a more complete understand-
ing of Schizocosa stridulans will contribute to our
understanding of the evolution of this genus.

Material examined.— USA: lllinios: Mason Co., Sand
Ridge State Forest, May-June 1985 (G. Stratton and
L. Hartz), 15 males, 15 females (MCZ); Jersey Co.,
Pere Marquette State Park, 29 May 1984 (G. Stratton,
L. Williams), 4 males (GES). Ohio: Athens Co., Strouds
Run State Park, June 1986 (J. Rovner), 1 male (GES).
Missouri: St Louis Co., Babler State Park, 1 June 1984
(G. Stratton, L. Williams), 10 males (GES); Cape Gi-
rardeau Co., Trail of Tears State Park, Oak Forest Up-
lands, 25 June 1984 (G. Stratton and L. Williams), 18
m.ales, 1 1 females (MCZ). Tennessee: Lawrence Co.,
Davy Crockett State Park, ravine slope, 16 May 1983
(W. P. Maddison), 9 males, 7 females (MCZ); Knox
Co., near Powell, oak forest, 23, 30 June 1981 (G.
Stratton), 5 males, 1 2 females (MCZ). Kentucky: Row-
an Co., Daniel Boone National Forest, Twin Knob
Recreation Area, 14 May 1983 (W. P. Maddison), 4
males, 3 females (MCZ); Hickman Co., Columbus Bel-
mont Battle Field State Park, 3 June 1984 (G. Stratton,
L. Williams) 1 male (GES). Alabama: Lauderdale Co.,
above Tennessee River, 18 June 1984 (G. Stratton, L.
Williams), 1 male, 1 female (GES). Mississippi: Pon-
totoc Co., Natchez Trace Parkway, 17 May 1983 (W.
P. Maddison), 4 females (MCZ), 1 mi. SE of Ecru,
pitfall in deciduous forests May-June 1980 (W. H.

Cross), 7 males (MSM); Claiborne Co., Rocky Springs
Park, 17 May 1983 (W. P. Maddison), 5 males, 6 fe-
males (MCZ). (Note: MCZ refers to Museum of Com-
parative Zoology, Harvard University, MSM to Mis-
sissippi State Museum, and GES to the personal
collection of the author.)

ACKNOWLEDGMENTS

Thanks are extended to G. Uetz for his ongoing
interest and support and to C. Dondale for ex-
amining specimens and providing encourage-
ment. The comments of A. Brady and C. Don-
dale greatly improved the quality of the
manuscript. I wish to thank W. Maddison for
collecting wolf spiders, and to L. Williams and
L. Hartz for assistance with field work. Thanks
are extended to R. Schmitter for teaching me to
use the scanning electron microscope. B. Gib-
bons provided the drawings of the legs and M.
Knapp helped with some of the measurements.
Financial support was provided by the Cottrell
Research Corporation and Bradley Board for Re-
search and Creativity.

LITERATURE CITED

Cady, A. B. 1984. Microhabitat selection and loco-
motor activity of Schizocosa ocreata (Walckenaer)
(Araneae: Lycosidae). J. Arachnol., 11:297-307.
Dondale, C. D. & J. H. Redner. 1978. Revision of

STRATTON -5C///Z<9C(9S^ STRIDULANS, NEW SPECIES

39

the nearctic wolf spider genus Schizocosa (Araneida;
Lycosidae). Canad. EntomoL, 110:143-181.

Stratton, G. E. & G. W. Uetz. 1981. Acoustic com-
munication and reproductive isolation in two spe-
cies of wolf spiders. Science, 214:575-577.

Stratton, G. E. & G. W. Uetz. 1983. Communication
via substratum-coupled stridulation and reproduc-
tive isolation in wolf spiders (Araneae: Lycosidae).
Anim. Behav., 31:164-172.

Stratton, G. E. & G. W. Uetz. 1986. The inheritance
of courtship behavior and its role as a reproductive

isolating mechanism in two species of Schizocosa
wolf spiders. Evolution, 40:129-141.

Uetz, G. W. & G. Denterlein. 1979. Courtship be-
havior, habitat and reproductive isolation in Schi-
zocosa rovneri Uetz and Dondale (Araneae: Lycos-
idae). J. ArachnoL, 7:121-128.

Uetz, G. W. & C. D. Dondale. 1979. A new wolf
spider in the genus Schizocosa (Araneae: Lycosidae)
from Illinois. J. ArachnoL, 7:86-87.

Manuscript received August 1989, revised August 1990.

1991. The Journal of Arachnology 19:40-48

ONTOGENETIC AND SEASONAL CHANGES IN WEBS AND
WEBSITES OF A DESERT WIDOW SPIDER

Yael Lubin and Mandy Kotzman: Mitrani Center for Desert Ecology, Blaustein Institute
for Desert Research, Ben Gurion University of the Negev, Sede Boqer Campus 84990
Israel

Stephen Ellner: Biomathematics Graduate Program, Department of Statistics, North
Carolina State University, Raleigh, North Carolina USA

Abstract. Morphometric and nest position variables were used to examine the effects of
spider growth and seasonality on webs and websites of the desert widow spider, Latrodectus
revivensis Shulov (Theridiidae) in the Negev desert of Israel. The form of the web was similar
over the full range of spider body sizes. All morphometric variables had strong positive
correlations with spider size: larger spiders occupied larger nests in larger shrubs. However,
nest characteristics were more highly correlated with spider size than were website char-
acteristics. When the effect of spider size was removed by regression, more than 75% of the
remaining variance consisted of correlated variation in three groups of variables relating to
( 1 ) website characteristics (48%), (2) nest characteristics ( 1 8%) and (3) capture web placement
(12%). Most nest and website variables showed effects of seasonality that were independent
of spider size, and may be related to the thermal regime in the nest. The results indicate
that the relative quality of potential websites changes seasonally and with spider growth.
We suggest that the costs of relocating a web outweigh the advantages of reaching a new
website, with the result that spiders remain for some time in websites which have become
less suitable.

The habitat requirements of many organisms
change as they age, resulting in a shifts of their
“ontogenetic niche” (Werner & Gilliam 1984).
Ontogenetic changes in habitat may involve
changes in living sites, in food requirements or
in other factors which scale with body size. Such
size-related changes in habitat requirements may
have particularly important fitness consequences
for sedentary animals for which the possibilities
of moving to new sites may be limited (e.g., Sha-
chak & Brand 1983).

Web-building spiders are relatively sedentary
predators (Janetos 1 986). In most species the web
is primarily a prey-capture device whose location
and structure reflect the local distribution of prey
(Riechert & Luczak 1982; Janetos 1986; Riechert
& Gillespie 1986). Thus, studies of website re-
quirements have focused mainly on the effects
of prey abundance (e.g., T umbull 1 964; Gillespie
1981; Olive 1980, 1982; Vollrath 1985). The
changing requirements of developing spiders are
also likely to affect web structure and website
selection (Enders 1975; Vollrath 1987). How-
ever, these have not been examined systemati-

cally, and it is not known to what extent changes
in web and website characteristics are due to
growth, seasonal factors or other effects.

In this study, we use morphometric and nest
position variables to characterize ontogenetic and
seasonal changes in the webs and websites of the
desert widow spider Latrodectus revivensis Shu-
lov. Statistical analysis of these data allows us to
separate the variation in web and website char-
acters due to spider sex, size, season, and other
factors. In addition, we examine the patterns of
covariation among the morphometric variables
and their relationships with spider size.

NATURAL HISTORY AND METHODS

Natural history. — Latrodectus revivensis
(Theridiidae) is known only from the Negev des-
ert of Israel (Levi & Amitai 1983). Females ma-
ture in spring or summer (March to August) and
produce eggsacs throughout the summer and au-
tumn (May to September; Levy and Amitai 1 983).
Incubation time is about one month. Some young
emerge in mid- to late summer and overwinter
as juveniles. In other instances, eggs remain in

LUBIN ET AL.- VARIATION IN WEBS AND WEBSITES OF WIDOW SPIDER

41

Figure 1.— Schematic drawing of the web of L. revivensis, showing (a) website and (b) nest variables measured
in this study. NH = nest height, SH = shrub height, CD = distance from the nest to the capture web, CH =
height of the capture web, NT = total nest length, DN = length of dense silk layer, DB = length of debris layer,
DM = nest diameter at edge of debris layer, MD = maximum nest diameter.

the eggsac over the winter and the young emerge
the following spring.

Webs of L. revivensis are durable and long-
lasting structures which may persist for up to
several months (Zilberberg 1988). The web con-
sists of separate nest and prey-capture compo-
nents (Shulov 1948; Szlep 1965; Fig. 1). The nest,
built in a shrub, is connected by strong bridging
threads of variable length (a few centimeters to
over a meter) to a horizontal silk platform. The
platform is usually placed over an area of bare
ground beyond the edge of the shrub, and an
array of sticky capture threads is suspended from
the platform to the ground.

The nest of L. revivensis consists of a curved,
silk cone (Fig. 1). The top of the cone is covered
with a dense silk layer, while the lower section
is a more open mesh. In addition, the nest top
is covered with scattered debris which may in-
clude sand, pebbles, snail shells and feces, plant
material, exuviae and remains of prey. The dense
silk and debris layers are usually sparse or absent

on new nests, but may completely obscure the
upper half of an old nest.

The spiders are active at night and remain con-
cealed in the nests during the day. Nocturnal
activities include web repairs, renewal of the
sticky, capture threads and prey capture. Web
relocation also takes place at night.

Study area.— The study site was located on the
rocky slopes of the Halukim Ridge, near Sede
Boqer (30°50'N:34°46'E) in the central Negev
region of Israel. The ridge runs north-south and
is dissected laterally by dry watercourses pro-
ducing a relief of about 50 m. The area is arid
with highly variable winter rains (about 100 mm
annually) and is sparsely vegetated with a per-
manent shrubland (Evenari et al. 1982). Nests of
L. revivensis occurred in several shrub species,
including Zygophyllum dumosum, Artemesia
herba-alba, Reaumuria negevensis, Noaea mu-
cronata, and Hammada scoparia, and some-
times in clumps of annuals (e.g., Reboudia pin-
nata) and grasses.

42

THE JOURNAL OF ARACHNOLOGY

The study area was approximately 20 hectares
of a north-facing slope of a small wadi. To reduce
the effects of habitat heterogeneity, we limited
our search for spiders to the lower portion of the
slope, from the edge of the wadi bed to a rocky
outcrop about 50 m up the slope (Shivta and
colluvial formations; Olsvig-Whittaker et al.
1983).

Web and website measurements. — We located
and individually flagged and mapped webs. Webs
and websites were characterized with the follow-
ing measurements (Fig. 1): height of the nest,
height of the shrub, height of the capture plat-
form and its distance from the nest, total nest
length, length of the dense silk layer, the maxi-
mum length of the debris covering, nest diameter
at the lower edge of the debris layer and maxi-
mum nest diameter. We also determined the
compass orientation of the nest opening (nest
aspect) and the quadrant of the shrub in which
the nest was located (NE, NW, SE, SW). A total
of 350 nests and 226 spiders were sampled in
this manner between January and August 1987
and March and August 1988.

Spiders in occupied webs were sexed, classified
as juvenile, subadult or adult, and measured for
total body length and length of the tibia + patella
of leg IV. For all statistical analyses, we used
body length as a measure of spider size because
of convenience of measurement in the field. Body
length includes the expandable abdomen and may
be influenced by spider condition (Anderson
1974), unlike the more rigid cuticle of the leg
segments which does not change size during an
instar (Miyashita 1968). However, body length
was closely correlated with the length of the tibia
+ patella in L. revivensis {r^ = 0.92, n = 211),
suggesting that for L. revivensis, spider condition
did not significantly affect the length of the ab-
domen.

Statistical analyses.— Comparisons between
each pair of morphometric variables were made
using standard regression (linear or polynomial)
and correlation analyses. To remove heterosce-
dasticity, the dependent variable of each analysis
was transformed using the Box-Cox family of
power transformations with maximum-likeli-
hood choice of parameters (Ruppert 1 989; Krebs
1989).

Patterns of covariation among morphometric
variables were examined by principal compo-
nents analysis (PCA; Joliffe 1 986). We calculated
the variances and pairwise covariances of all
variables. We used pairwise rather than listwise

deletion of missing values; although the results
of the two methods were nearly identical, listwise
deletion greatly reduced the sample sizes for each
covariance. PCA identifies a sequence of uncor-
related “components” (axes) which are linear
combinations of the original variables. The first
axis is chosen to “explain” as much as possible
of the variance in the data, the second axis ex-
plains as much as possible of the remaining vari-
ance, and so on.

We applied PCA to a correlation matrix of the
raw data and of the Box-Cox transformed data
(Joliffe 1986). To determine the extent to which
spider size alone was the basis of these correla-
tions, we removed the effect of spider length by
regression: each value was replaced by its resid-
ual deviation from a regression on spider length.
Quadratic, rather than linear, regression was used
in order to stabilize variances. We then applied
PCA to the covariance matrix of the residuals.
PCA of the covariance matrix of residuals is use-
ful in this case, because the variables themselves
had already been standardized to equal variance
by the data transformations. Consequently, high
variance of the residuals indicates a weak cor-
relation with spider length.

RESULTS

Males and females.— Our measurements of
male’s webs were restricted to those of juveniles.
Adult males often remained in their own juvenile
webs {n = 1 7), built small nests lacking capture
webs attached to nests of females {n = 5), shared
nests of adult or juvenile females (n = 26 and 4,
respectively) or occupied abandoned nests of fe-
males (n = 4).

Webs of juvenile males and females differed
significantly in all morphometric variables ex-
cept distance from the nest to the capture web.
However, when we eliminated the difference in
size between males and females by comparing
only females of sizes equivalent to juvenile males
(<6.5 mm body length. Fig. 2), these differences
disappeared. Therefore, web and website char-
acteristics of juvenile males and females were
treated as a single data set.

Effects of spider length.— The total body length
of spiders in our sample ranged from 1.8-16.8
mm (x = 7.9, SD = 3.8, n = 246 spider mea-
surements; Fig. 2). In the following regression
analyses, spider length is treated as an error-free
independent variable (Snedecor & Cochran 1 967),
because the errors in measuring spider length
were small and independent of spider length over

LUBIN ET AL.-VARIATION IN WEBS AND WEBSITES OF WIDOW SPIDER

43

JM AM JF AF

STAGE, SEX

Figure 2.— Total body lengths (in mm) of L. revi-
vensis used in the study: boxes show the means (center
lines) and one standard deviation and vertical lines
show ranges for juvenile and adult males (JM and AM,
respectively) and for juvenile and adult females (JF
and AF, respectively). Juvenile males were all sub-
adults; j uvenile females included all immature and sub-
adult stages.

the range of sizes encountered (r’ = 0.04, P >
0.1; based on 5 replicate measurements of each
of 29 spiders). Pooling all measurements, the
standard deviation of the measurement error was
0.23 mm (95% Cl: 0. 1 9-0.29 mm), which is much
smaller than the standard deviation of spider
length in our full data set (SD = 3.84 mm, n =
252 spiders).

Spider length explained a significant amount
of the variation in all web and website variables
(Table \, P < 0.001 in all cases). The amount of
variation explained by spider length was higher
for variables that describe the nest itself (total

Table 2.— Allometric regression equations for nest
morphometric variables: In y - fio + a,\n X, where
X = nest total length. All regressions are significant at

P < 0.001 {Ho. a, = 0).

Variable

n

f-

ao

a,

Dense silk

290

0.67

0.21

0.70

Debris

291

0.79

-0.31

0.91

Nest diameter

287

0.84

-0.65

0.93

Maximum diameter

135

0.83

-0.78

1.01

nest length, lengths of dense silk and debris lay-
ers, and nest diameters) than for variables as-
sociated with nest placement in the shrub (shrub
height and nest height) or with the capture web
(height and distance from the nest).

Nest placement and allometry.— The height of
the nest in a shrub was closely correlated with
shrub height. Over the entire range of shrub
heights of 15 to 122 cm, nest height was ap-
proximately % of shrub height (r^ = 72.7%, n =
318). The allometric equation, nest height =
0.69(shrub height)"’®, only slightly improved the
amount of variance explained by shrub height
{P- = 73.0%).

Variation in the length of the debris layer, nest
diameter and maximum nest diameter can be
described by allometric regressions on total nest
length (Table 2). The length of dense silk in-
creased linearly up to nest lengths of approxi-
mately 75 mm, but did not increase with further
increase in nest length. Variation in this char-
acter was best described overall by the allometric
equation, dense silk length = 0.2(total nest
length)"'' {P = 0.67).

Table 1 . — Regression equations for the effect of spider length on nest and website morphometric variables.
The slopes of all of the regressions are significantly different from zero {P < 0.001). The equations are of the
form: = ao± a,x + a2X\ where x = spider length.

Variables

n

r~

\

a„

a,

a2

WEBSITE

Nest height

111

0.27

0.18

1.62

0.02

0.001

Shrub height

111

0.34

0.12

1.44

0.01

0.0001

Capture web distance

142

0.21

0.11

1.32

-0.004

0.0009

Capture web height

46

0.52

-0.06

0.89

-0.006

0.0001

NEST

Nest length

175

0.85

0.47

1.74

0.88

-0.025

Dense silk length

172

0.66

0.19

1.2

0.13

-0.005

Debris length

173

0.76

0.27

1.27

0.23

-0.008

Nest diameter

171

0.90

0.14

1.14

0.07

-0.002

Maximum diameter

109

0.92

0.18

1.12

0.13

-0.004

44

THE JOURNAL OF ARACHNOLOGY

Table 3.— Correlations among website and nest variables. Correlations based on the raw data are shown above
the diagonal; below the diagonal are the correlations among residuals, after removing the effect of spider length
by regression (see Methods: Statistical analyses). NH = nest height, SH = shrub height, CD = distance from nest
to capture web, CH = capture web height, NT = total nest length, DN = length of dense silk layer, DB = length
of debris layer, DM = nest diameter. * = nonsignificant correlations (P > 0.05).

NH

SH

CD

CH

NT

DN

DB

DM

NH

0.85

0.64

0.43

0.55

0.32

0.47

0.48

SH

0.81

0.54

0.46

0.63

0.37

0.54

0.56

CD

0.63

0.49

0.31

0.41

0.19

0.42

0.41

CH

♦

*

*

0.77

0.39

0.66

0.65

NT

0.25

0.23

*

0.42

0.71

0.87

0.91

DN

-0.14

*

-0.22

*

0.37

0.77

0.69

DB

*

♦

*

0.27

0.46

0.57

0.87

DM

*

♦

*

*

0.45

0.42

0.48

Covariation of web and website variables. — All

variables were significantly and positively cor-
related with each other (Table 3; < 0.0 1 , except

for the height of the capture web against its dis-
tance from the nest, 0.02 < P < 0.05). The re-
siduals had fewer significant correlations and
some negative correlations, indicating that many
correlations in the original variables were due to
the effect of spider length on all variables. As a
result, PCA of web and website variables using
the raw data was not very informative (Table 4).
The first axis, which accounted for 63% of the
variance, loaded evenly on all variables and sim-
ply reflects the positive correlation among the
variables; the remaining axes could not be in-
terpreted intuitively.

Clearer patterns were found in PCA of the re-
siduals after removing the effects of spider length.

Website variables (nest height, shrub height, and
distance to the capture web) accounted for 48%
of the residual variance (PCA axis 1, Table 4).
The second axis, which loaded mainly on nest-
concealment variables (dense silk and debris),
accounted for 1 8% of the residual variance, and
the third axis loaded mainly on capture web height
(12% of the residual variance). Thus over 78%
of the variance in web and website morphomet-
ric data may be summarized by four independent
axes of variation: spider size and the three clus-
ters of morphometric variables identified by the
first three PCA axes.

Seasonal differences in nest and website char-
acteristics.—Climatic conditions in the central
Negev differ considerably in winter and summer.
There is a “cold season” (November to April;
mean monthly temperature 13.4° C, range of

Table 4. — Principal components analysis (PCA) of website and nest variables. The first two PCA axes are
shown for the PCA based on the raw data (left side of table) and the first three PCA axes based on analysis of
the residuals, after removing the effect of spider length (right side of table). The percent of the variance explained
by each axis is shown. Variables are abbreviated as in Table 3.

PCA axes based on

Data

Residuals

Variable

1

2

1

2

3

% Variance
explained

63

16

48

18

12

NH

0.12

0.20

0.32

0.03

-0.07

SH

0.12

0.15

0.27

0.03

-0.10

CD

0.09

0.19

0.28

-0.06

0.10

CH

0.12

-0.05

0.03

0.17

0.41

NT

0.15

-0.07

0.04

0.14

0.03

DN

0.11

-0.15

-0.05

0.25

-0.20

DB

0.14

-0.10

0.01

0.22

-0.05

DM

0.14

-0.08

0

0.10

-0.05

LUBIN ET AL.- VARIATION IN WEBS AND WEBSITES OF WIDOW SPIDER

45

Table 5.— A comparison of spider website and nest variables in “hot” and “cool” seasons. Shown are means
and standard deviations of all measurements and probabilities for /-tests performed on the raw data (Pdata) and
on the residuals, after removing the effeet of spider length (P,es,d)- SL = spider length, MD = maximum nest
diameter; all other abbreviations as in Table 3. * = P < 0.05, ** = />< 0.01, *** = P < 0.001.

Variable

Cool

Hot

P dala

P rcsid

Mean

SD

N

Mean

SD

N

SL (mm)

7.2

2.8

126

8.6

4.5

126

—

NH (cm)

20.2

9.7

139

29.1

13.0

180

***

SH (cm)

31.7

11.6

140

43.4

18.2

180

***

CD (cm)

13.4

7.9

95

28.3

19.9

71

CH (cm)

12.3

2.7

32

11.1

6.3

16

ns

ns

NT (mm)

55.2

22.4

118

65.8

37.3

178

ns

DN (mm)

25.4

10.0

114

19.8

9.2

177

***

DB (mm)

31.6

11.8

114

31.9

19.8

178

ns

***

DM (mm)

22.4

9.3

112

26.1

16.3

176

*

MD (mm)

29.7

10.9

34

41.1

22.2

101

***

means 9.8-18.0° C) and a “hot season” (May to
October, mean monthly temperature 23.4° C, range
21.1-25.5° C).

There were statistically significant seasonal dif-
ferences in most of the morphometric variables,
both in the raw data and after removing the effect
of spider length (Table 5). In the hot season,
spiders were found in taller shrubs, built nests
higher above the ground and built capture webs
further from their nests. Nest diameter and max-
imum diameter were both greater in the hot sea-
son, but the dense silk layer was shorter in the
hot season. Nests were longer in the hot season,
but this appears to be the result of the seasonal
difference in spider length (Table 5). The height
of the capture web did not vary between seasons.

In the hot season, nests occurred more fre-
quently on the east side of shrubs than on the
west side (x- = 8.85, P < 0.005, n = 76 nests),
but their distribution with respect to the N-S axis
was random (x- = 0.47, P > 0.1). In the cold
season the distribution of nests with respect to
shrub quadrant was random (x^ = 0.728, P >
0A,n= 125). The orientation of the nest opening
was not significantly different from random in
both seasons (Rayleigh test; cold season: 106
nests; hot season: 68 nests), nor were there sig-
nificant differences among webs of juveniles and
adults in either season.

DISCUSSION

Scaling of nest and website components.— The
form of the web is remarkably constant over the
full range of spider sizes, from newly emerged
young to adults. Webs of juvenile males did not

differ from those of similar-sized juvenile fe-
males. Eggsac nests made by some females ap-
peared to be wider and more barrel-shaped than
nests made by subadult and juvenile females,
possibly to accommodate the large, spherical sacs.

Nest diameter and the lengths of the layers of
dense silk and debris all scaled allometrically with
total nest length. For linear dimensions of the
nest, isometry is the appropriate null model.
However, nest dimensions also scale to body size,
and the allometric equation (v = ax'’) is generally
a good descriptor of body size relationships (Pe-
ters 1983). Allometric scaling may indicate that
a functional relationship exists among the vari-
ables which depends on their geometry (La-
Barbara 1989).

Both the dense silk layer and the debris pro-
vide protection for the spider from predators,
whether mechanically or by crypsis (Konigswald
et al. 1990). In the initial stages of nest construc-
tion, L. revivensis builds a thin silk cap, consist-
ing of a few threads only (which will become the
top of the nest), and then descends to the ground
and carries up bits of debris which it attaches to
the top of the cap. Thus, protection from visu-
ally-orienting predators is obtained quickly and
with a minimal outlay of silk and activity, both
of which are energetically expensive (Lubin 1973;
Prestwich 1977). If the spider remains at that
website, both debris and dense silk are added to
the nest on successive nights.

Nests in the hot season had proportionately
shorter lengths of dense silk relative to the debris
layer. This may be related to the thermal regime
within the nest. Temperatures during the day

46

THE JOURNAL OF ARACHNOLOGY

inside the dense-silk portion of the nest were
consistently higher than in the lower, open mesh
portion of the nest, due to the greater flow of cool
air through the open mesh portion than through
the dense silk (Lubin et al. unpubl.). By increas-
ing the size of the debris layer without a con-
comitant increase in the dense silk, shade and
protection are provided without reducing airflow
through the nest.

Nests were generally placed % of the way up
the shrub, but were located significantly higher
in the hot season than in the cold season. The
portion of the shrub above the nest can provide
substantial shade and concealment from visual
predators (Konigswald et al. 1990). In summer,
spiders place their nests higher in shrubs (and in
taller shrubs), perhaps in order to take better
advantage of convective cooling. The placement
of large nests in large shrubs is intuitively ob-
vious, as small shrubs may provide insufficient
support and cover for large nests. It is less clear
why small nests are not found in large shrubs.
Given the tendency for nests to be built % of the
way up in shrubs, the upper branches of large
shrubs may be unsuitable (e.g., too widely spaced)
for suspending small nests.

Somewhat surprisingly, nest openings were
random with respect to compass orientation both
in summer and winter. In open habitats, a di-
urnal orb-weaver, Micrathena gracilis (Walcke-
naer), was shown to orient its web to reduce ex-
posure to direct insolation (Biere & Uetz 1981).
Similarly, the funnel openings of a desert age-
lenid, Agelenopsis aperta (Gertsch) tend to face
north in summer (Riechert & Tracy 1975). Both
species, however, were exposed regularly to di-
rect solar radiation, the former while sitting on
its web and the latter while basking and hunting.
Latrodectus revivensis is mainly nocturnal and
does not bask.

Sources of variation in web and website char-
acteristics.—The sources of variation in web and
website characteristics separate into three main
components: (1) spider size, (2) seasonal effects
and (3) residual variation. All morphometric
variables had strong positive correlations with
spider size: larger spiders occupied larger nests
in larger shrubs. Nest-size variables were more
tightly correlated with spider length than were
website or capture web variables. Seasonal dif-
ferences accounted for some variation in most
web and website characters after removing the
effect of spider length.

Nearly half of the variance not accounted for

by spider length or by seasonal effects consists
of correlated variation in website and capture
web characteristics (residual PCA axis 1). An ad-
ditional 18% was attributed to correlated vari-
ation in nest characteristics (axis 2). Distance to
the capture web was a component of axis 1; thus,
a larger shrub is associated with a greater distance
to the capture web. However, the variation in
the height of the capture web was also identified
as a separate component of the PCA (axis 3). We
conclude that the capture web and nest are rel-
atively independent structures, and factors af-
fecting capture web and nest placement may dif-
fer.

The residual variance in websites not account-
ed for by either spider size or seasonal effects
may be due to imprecise site selection (see Ja-
netos & Cole 1981), or to other factors that we
did not measure. Such factors may include: vari-
ation in body condition, hunger and reproduc-
tive status of the spider, spatial and temporal
variation in food supply. Riechert (1974) docu-
mented the importance of relatively short-lived
phenomena (e.g., the presence of ffowers and oth-
er insect attractants) in explaining the distribu-
tion of a desert web-building spider, Agelenopsis
aperta (Agelenidae), and showed that such cues
may influence the choice of a website (Riechert
1985). In orb-weaving spiders, which may re-
locate their webs frequently, site selection and
movement have been correlated with the avail-
ability of web supports (Enders 1975; Hodge
1 987a), the degree of disturbance to webs (Hodge
1987b) and food availability (Olive 1982; Voll-
rath 1985).

The residual variance in nest characteristics
may reflect differences in site quality, for ex-
ample, in the thermal regimes prevailing in dif-
ferent shrubs or the presence of suitable debris
for nest concealment. The amounts of dense silk
and debris might vary with immediate needs for
crypsis against a heterogeneous background, or
in response to the perceived risk of predation. In
some orb-web spiders, variable development of
stabilimenta (lines or zigzags of dense silk in the
orb; Robinson & Robinson 1970) and of other
web “decorations” (e.g., bits of debris) has been
correlated with their degree of exposure to visual
predators (Eberhard 1973; Lubin 1975, 1986).

These results suggest that the relative quality
of different websites changes seasonally and with
the spider’s ontogeny. Changes in nest require-
ments can be accommodated by modification of
the existing nest or the construction of a new nest

LUBIN ET AL.- VARIATION IN WEBS AND WEBSITES OF WIDOW SPIDER

47

at the same site, but changes in website require-
ments may necessitate moving to a new site.
While nests are modified regularly and are there-
fore tightly correlated with spider size, the large
variation observed in shrub and capture web
characteristics suggests that the spiders remain
for some time in websites that have become less
suitable. Nonetheless, website relocation occurs
several times in the spider’s lifetime (Zilberberg
1988) and the choice of new websites is influ-
enced by the spider’s size and by the time of year.
The decision to move to a new website may re-
flect a trade-olf between the advantages of reach-
ing a more suitable website and the costs of re-
locating, such as an increased risk of predation
and the energetic costs of movement and web
construction.

ACKNOWLEDGMENTS

B. Roth and A. Kotzman helped with the field
work and D. Ward and Y. Ayal commented on
a draft of the manuscript. We are grateful for
their help. The study was supported by the U.S.-
Israel Binational Science Foundation (Grant
#8600092 to YL and SE, and a Bergmann Me-
morial Research Award to SE). This is Publi-
cation No. 106 of the Mitrani Center for Desert
Ecology, Blaustein Institute for Desert Research.

REFERENCES CITED

Anderson, J. F. 1974. Responses to starvation in the
spiders Lycosa lenta Hentz and Filistata hibernalis
(Hentz). Ecology, 55:576-585.

Atkinson, A. C. 1985. Plots, Transformations and
Regression. Oxford University Press, Oxford.

Biere, J. M. & G. W. Uetz. 1981. Web orientation
in the spider Micrathena gracilis (Araneae: Aranei-
dae). Ecology, 62:336-344.

Eberhard, W. G. 1973. Stabilimenta on the webs of
Uloborus diversus (Araneae: Uloboridae) and other
spiders. Journal of Zoology, London, 171:367-384.
Enders, F. 1974. Vertical stratification in orb-web
spiders (Araneidae, Araneae) and a consideration of
other methods of coexistence. Ecology, 55:31 7-328.
Enders, F. 1975. Change in web site in Argiope spi-
ders (Araneidae). Amer. Midi. Natur., 94:484-490.
Evenari, M., L. Shanan & N. Tadmor. 1982. The
Negev: The Challenge of a Desert. Harvard Uni-
versity Press, Cambridge.

Gillespie, R. G. 1981. The quest for prey by the web-
building spider, Amaurobius similis (Blackwell).
Anim. Behav., 29:953-954.

Hodge, M. A. 1987a. Macrohabitat selection by the
orb weaving spider, Micrathena gracilis. Psyche, 94:
347-361.

Hodge, M. A. 1987b. Factors influencing web site

residence time of the orb weaving spider, Micrathe-
na gracilis. Psyche, 94:363-371.

Janetos, A. C. 1986. Web-site selection: Are we ask-
ing the right questions? Pp. 9-22, In Spiders: Webs,
Behavior, and Evolution (W. A. Shear, ed.). Stan-
ford University Press, Stanford, California.

Janetos, A. C. & B. J. Cole. 1981. Imperfectly optimal
animals. Behav. Ecol. Sociobiol., 9:203-209.

Joliffe, I. T. 1986. Principal Component Analysis.
Springer-Verlag, New York.

Konigswald, A., Y. D. Lubin & D. Ward. 1990. The
effectiveness of the nest of a desert widow spider in
predator deterrence. Psyche, 97:75-80.

Krebs, C. J. 1989. Ecological Methodology. Harper
and Row, New York.

LaBarbara, M. 1989. The evolution and ecology of
body size. Pp. 69-98, In Patterns and Processes in
the History of Life (D. M. Raup & D. Jablonski,
eds.). Dahlem Konferenzen 1986. Springer-Verlag,
Berlin.

Levy, G. & P. Amitai. 1983. Revision of the widow-
spider genus Latrodectus (Araneae: Theridiidae) in
Israel. Zool. J. Linnean Society, 77:39-63.

Lubin, Y. D. 1973. Web structure and function: The
nonadhesive orb-web of Cyrtophora nwluccensis
(Doleschall) (Araneae: Araneidae). Forma et Func-
tio, 6:337-358.

Lubin, Y. D. 1975. Stabilimenta and barrier webs in
the orb webs of Argiope argentata (Araneae: Ara-
neidae) on Daphne and Santa Cruz Islands, Gala-
pagos. J. Arachnol., 2:1 19-126.

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.). Stan-
ford University Press, Stanford, California.

Miyashita, K. 1968. Quantitative feeding biology of
Lycosa T-insignata Boes. et Str. (Araneae: Lycosi-
dae). Bull. Nat. Inst. Agric. Sci., 22:329-344.

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

Olive, C. W. 1982. Behavioral response of a sit-and-
wait predator to spatial variation in foraging gain.
Ecology, 63:912-920.

Olsvig- Whittaker, L., M. Shachak & A. Yair. 1983.
Vegetation patterns related to environmental factors
in a Negev desert watershed. Vegetatio, 54; 1 53-1 65.

Peters, R. H. 1983. The Ecological Implications of
Body Size. Cambridge University Press, Melbourne,
Australia.

Prestwich, K. N. 1977. The energetics of web-build-
ing in spiders. Comp. Biochem. Physiol., 57A:321-
326.

Riechert, S. E. 1974. The pattern of local web dis-
tribution in a desert spider: Mechanisms and sea-
sonal variation. J. Anim. Ecol., 43:733-746.

Riechert, S. E. 1985. Decisions in multiple goal con-
texts: Habitat selection of the spider, Agelenopsis
aperta (Gertsch). Z. Tierpsychol., 70:53-69.

Riechert, S. E. & R. G. Gillespie. 1986. Habitat choice

48

THE JOURNAL OF ARACHNOLOGY

and utilization in web-building spiders. Pp. 23-48,
In Spiders: Webs, Behavior, and Evolution (W. A.
Shear, ed.). Stanford University Press, Stanford,
California.

Riechert, S. E. & J. Luczak. 1982. Spider foraging:
Behavioral responses to prey. Pp. 353-386, In Spi-
der Communication: Mechanisms and Ecological
Significance (P. N. Witt & J. S. Rovner, eds.).
Princeton University Press, Princeton.

Riechert, S. E. & C. D. Tracy. 1975. Thermal balance
and prey availablity: Bases for a model relating web-
site characteristics to spider reproductive success.
Ecology, 56:265-284.

Robinson, M. H. & B. Robinson. 1970. The stabi-
limentum of the orb web spider, Argiope argentata:
An improbable defence against predators. Can. En-
tomol., 102:641-655.

Ruppert, D. 1989. Fitting mathematical models to
biological data: A review of recent developments.
Pp. 250-26 1 , In Mathematical Approaches to Prob-
lems in Resource Management and Epidemiology
(C. Castillo-Chavez, S. A. Levin & C. A. Shoemaker,
eds.). Lecture Notes in Biomathematics, Vol. 81.
Springer-Verlag, Berlin.

Shachak, M. & S. Brand. 1983. The relationship be-
tween sit-and-wait foraging strategy and dispersal
in the desert scorpion, Scorpio maurus palmatus.
Oecologia, 60:371-377.

Shulov, A. 1 948. Latrodectus revivensis sp. nov. from
Palestine. Ecology, 29:209-215.

Snedecor, G. W. & W. G. Cochran. 1967. Statistical
Methods. Sixth Edition, Iowa State University Press,
Ames, Iowa.

Szlep, R. 1965. The web-spinning process and web
structure of Latrodectus tredecimguttatus, L. palli-
das and L. revivensis. Proc. Zool. Soc., London, 145:
75-89.

Turnbull, A. L. 1964. The search for prey by a web-
building spider, Achaearanea tepidariorum (C. L.
Koch) (Araneae, Theridiidae). Can. Entomol., 96:
568-579.

Vollrath, F. 1985. Web spider’s dilemma: A risky
move or site-dependent growth. Oecologia, 68:69-
72.

Vollrath, F. 1987. Growth, foraging and reproductive
success. Pp. 357-370, In Ecophysiology of Spiders
(W. Nentwig, ed.). Springer-Verlag, Berlin.

Werner, E. E. & J. F. Gilliam. 1984. The ontogenetic
niche and species interactions in size-structured
populations. Ann. Rev. Ecol. System., 15:393^25.

Zilberberg, G. 1988. Behavioral ecology ofthe widow
spider, Latrodectus revivensis Shulov (Araneae: Ar-
aneidae). Shappirit (J. Israel Entomol. Center), 6:
52-76 (in Hebrew, with English summary).

Manuscript received February 1990, revised July 1990.

1991. The Journal of Arachnology 19:49-54

DISPERSAL AND SURVIVORSHIP IN A POPULATION OF

GEOLYCOSA TURRICOLA (ARANEAE, LYCOSIDAE)

Patricia R. Miller:' Department of Entomology, Mississippi Entomological Museum,
Mississippi State University, Mississippi State, Mississippi 39762 USA

Gary L. Miller: Department of Biology, The University of Mississippi, University, Mis-
sissippi 38677 USA

Abstract. A population of the burrowing wolf spider Geolycosa turricola in Mississippi
was monitored over a period of 4 years. Weekly censuses of the number of burrows that
were active, open but not active, or inactive were taken. The timing of the dispersal of
spiderlings was examined by use of caging experiments. A habitat manipulation experiment
was used to assess burrow site preferences. This population reproduced on a 2-year cycle;
no young were produced in even years. The results suggest that some dispersing spiderlings
construct burrows immediately after leaving their mother’s burrow while others overwinter
and build their first burrow during the spring. Two dispersing groups are identified and are
shown to have different survivorship properties. The importance of this dispersal strategy
in terms of subsocial behavior is discussed.

A number of field studies of the population
dynamics of the obligate burrowing wolf spiders
{Geolycosa) have been undertaken in recent years
(e.g., McQueen 1978, 1983; Conley 1985). For
the most part these studies have confirmed the
incidental observations of Wallace (1942): mul-
tiyear life cycles predominate (McQueen 1978),
dispersal of young from the maternal burrow oc-
curs in the early summer (McQueen 1978, 1983;

Conley 1985) and may be by ballooning (Miller
1984a; McQueen 1978), and mortality of spi-
derlings is high (Humphreys 1976; McQueen
1978). However, several questions regarding the
population dynamics of these spiders remain un-
answered. Chief among these are questions re-
lating to the timing of initial burrow construction
in relation to the onset of dispersal, burrow site
preference and tenacity, and the extent to which
the timing of dispersal and the size of the dis-
persing spider affects survivorship. Here we ad-
dress these issues in a multiyear study of the
dynamics of a population of Geolycosa turricola
(Treat).

METHODS

We studied a population of Geolycosa turricola
near Starkville, Mississippi, continuously be-
tween 1982 and 1985. The population inhabited

'Present address: Department of Science and Mathe-
matics, Northwest Mississippi Community College
Senatobia, Mississippi 38668 USA

a 1 ha Selma Chalk deposit (Harper 1857; Miller
1984b) surrounded on three sides by thick growths
of southern red cedar {Juniperus silicicola) and
on the other side by a dirt road. The predominant
vegetation, beard grass (Andropogon sp.), oc-
curred in large clumps interspersed with bare and
litter-covered ground.

A number of small isolated populations of G.
turricola occurred in similar habitats within a 6
km radius of the study population. These pop-
ulations were monitored periodically to deter-
mine the extent of interpopulation variation in
the timing of reproduction and dispersal.

In the fall of 1982 and early spring of 1983,
prior to the onset of the dispersal of young, all
burrows in the field were marked with numbered
surveyor’s flags. Beginning in the spring of 1983
the population was censused at approximately
weekly intervals between March and October of
each year and once a month at other times of
the year. During each census, a search of the field
was conducted and previously undiscovered bur-
rows of spiderlings were marked. We assumed
in this study that changes in burrow diameter
represented growth by the spider occupying that
burrow. Therefore, the largest diameter (mm) of
each newly discovered and previously marked
burrow was recorded. Burrow diameter has been
shown to be a good indicator of both spider age
(McQueen 1978) and size (McQueen 1978; Mil-
ler & Miller 1984) for Geolycosa spiders.

The state of each burrow was recorded as: ( 1 )

50

THE JOURNAL OF ARACHNOLOGY

active (burrow and burrow turret in good repair
and/or spider seen in burrow), (2) open/inactive
(burrow and turret in disrepair but burrow open-
ing present), (3) disappeared (not found; previ-
ously marked burrows only). Burrowing wolf spi-
ders may block the entrance of their burrows
with silk and debris during certain times of the
year, particularly during late summer through
spring (Miller & Miller unpublished data). Be-
cause of this, every burrow that was scored as
disappeared was reexamined in subsequent cen-
suses. Renewed activity at a burrow previously
marked as disappeared resulted in the reassign-
ment of that burrow as active.

We assumed that burrows that appeared active
were, in fact, occupied by a spider even though
this was not confirmed in every case. Errors in
this regard will lead to an overestimation of pop-
ulation size. However, our intent was to study
dispersal timing and survivorship not to make
estimates of population size or density per se.
We also assumed that burrows were occupied by
the original inhabitant. Studies of the activity
patterns of Geolycosa (e.g., McQueen & Culik
1981) suggest that this is a reasonable assump-
tion.

A preliminary examination of the census data
suggested that dispersal involved two groups of
spiderlings of different size (an early summer
group of small size and a late summer group of
a larger size.) To substantiate this, the size of
these two groups was compared. The survivor-
ship of these two groups was compared over two
winters (1983 and 1984).

Caging experiments were used to determine
whether the beginning of dispersal corresponded
to the beginning of burrow building activity in
G. turricola. In these studies, a wire cage with a
10 cm radius (fine mesh window screen) was
placed over six burrows that contained predis-
persal young. The inside walls of the cages were
examined several times each week and spider-
lings that were found on the cage wall or on the
ground outside the burrow were scored as dis-
persers and removed. Small crickets were intro-
duced into the burrow periodically (about once
each week) for food.

To determine whether newly dispersed spi-
derlings showed any initial burrow site prefer-
ence we conducted a habitat manipulation ex-
periment in 1985. Prior to dispersal we altered
the habitat within 3 m of eight burrows that con-
tained young. Each circular area was divided into
four pie-shaped areas of equal size, and one of

the following treatments was assigned to each
pie-shaped area: (1) control (no change), (2) litter
enhanced (addition of sufficient litter to make a
uniform, 2-cm cover in the area, (3) denuded (all
grass and litter removed, (4) litter removed (litter
raked out, grass left). Because of the small num-
ber of burrows used for this experiment, we were
not able to completely randomize the experi-
ment. We examined these experimental plots
during the normal census and recorded the num-
ber and size of the burrows constructed in each
plot subsection.

Statistical comparisons of average burrow di-
ameters between months were made with de-
pendent (paired) i-tests (tj using burrows that
were active in both of the months being com-
pared. Independent i-tests {t) were used to com-
pare distinct sets of burrows (e.g., newly discov-
ered burrows vs previously known burrows).

RESULTS

Comparisons of the reproductive timing of
nearby populations with that of the study pop-
ulation revealed that the study population had
an unusual breeding schedule characterized by
alternating years of production of young. Thus,
although some females matured each year in oth-
er populations, mature females were found only
in alternating years (odd years beginning in 1 983)
in the study population.

The tabulation of the census data for the gen-
eration of spiderlings hatched in the spring of
1983 is given in Table 1. The number of newly
discovered and previously marked active bur-
rows is given for each month. The total monthly
change in the number of burrows represents the
number of active burrows from the previous
month plus the number of newly discovered ac-
tive burrows, minus the number of burrows that
disappeared since the previous month (Table 1).

Burrow construction by spiderlings began in
July and new burrows were discovered through-
out the summer (Table 1). In 1 983 a total of 343
burrows were marked between July and October.
Most (8 1 .0%) were discovered in July and August
(Table 1). In that year, an average of 92.4% of
the burrows marked in one month were active
in the next month (92.7% July to August, 84.6%
August to September, 100% September to Oc-
tober).

There was an increase in the average burrow
diameter between July and August 1983. The
average diameter of the 141 burrows that were
established in July and recorded as active in Au-

MILLER & mLLEK-GEOLYCOSA POPULATION DYNAMICS

51

Table 1.— Survival history of a cohort of Geolycosa
turricola. Entries are the number of active burrows (see
text). Table includes only spiderlings that hatched in
spring 1983.

Active burrows

Date

New

Pr.

marked

Total

Diameter

(SD)

7-83

152

0

152

4.2(1.03)

8-83

126

141

267

7.0(1.78)

9-83

60

226

286

6.9(1.62)

10-83

5

286

291

—

Winter

3-84

0

139

139

8.2 (2.21)

4-84

136

139

275

8.3(1.97)

5-84

2

207

209

9.8 (1.61)

6-84

0

149

149

12.3(2.09)

7-84

0

139

139

14.7(2.61)

8-84

_

—

—

—

9-84

0

5

5

20.0 (3.74)

10-84

0

5

5

19.3 (3.04)

Winter

3-85

0

50

50

—

4-85

0

5

5

—

gust increased from X = 4.2 mm (SD = 1.03) to
X = 6.9 mm (SD = 1.79) {t^ = 1 1.5, #= 140,
P < 0.00 1). The average diameter of the burrows
that were established in August (X = 7.1 mm,
SD = 2.1 1) was significantly larger than the av-
erage diameter of the July burrows {t = 14.53,
df = 265, P < 0.001). There was not a difference
in the diameter of the burrows established in
September when compared to the new burrows
ofAugust(/= I33,df= 178, P < 0.001). Because
of the difference in initial burrow size between
the July and August spiders, the survivorship of
those two groups was examined separately.

Approximately 48% of the burrows that were
established in the summer and fall of 1983 (291;
Table 1) were recorded as active in early 1984
(139; Table 1). Of these, 31.6% were of the group
that initially established their burrows in July of
1983 and 68.4% were of the group of burrows
discovered in August of 1983 (Table 2). Fifteen
percent of those burrows that were active in 1 984
were still active in 1985. Approximately two-
thirds (76.2%) of these were from the original
August 1983 group (Table 2).

Forty-seven percent of the burrows that were
active in October 1983 reopened in March 1984.
The average diameter of these spring 1984 bur-
rows was significantly greater than the average

Table 2.— Percentage of overwinter survivors of and
burrow diameters of two groups of G. turricola spi-
derlings. July 1983 group includes spiderlings that es-
tablished burrows during July 1983, August 1983 group
includes spiderlings that established burrows between
August and October 1983 (see text).

Group

Initial
burrow size

% over-
winter

First winter, 1983

July 1983 group
(TV = 44)

4.2(1.03)

31.6

August 1983 group
(TV = 95)

7.1 (1.78)

68.4

Second winter, 1984
July 1983 group

(TV= 5)

4.1 (0.98)

23.8

August 1983 group
(TV= 16)

7.4(1.64)

76.2

for September of the previous year (l^ = 5.5 1,

= 139, F < 0.01; we obtained no burrow di-
ameter estimates for October of 1983). All of
these overwinter survivors remained active into
April at which time 136 new burrows were dis-
covered. The average burrow diameter of these
new burrows (X = 7.9 mm, SD = 2.93) was not
significantly different than that of the 139 over-
winter survivors {X = 8.7 mm, t = 0.80, df =
273, P > 0.05). The percentage of spiderlings
that survived through the spring and summer of

1984 was high {X = 85.1%; 100.0% March to
April, 75.3% April to May, 71.9% May to June,
93.3% June to July). In the late summer of that
year a large percentage of the burrows disap-
peared. A substantial number of these burrows
reappeared as active burrows in the spring of

1985 (35.9%; Table 1). During 1984 the average
burrow diameter increased from 8.2 mm to
around 20 mm (Table 1). Of the 50 burrows that
survived through the winter of 1984-1985, only
five of those were recorded active during August
of that year. One of those burrows is known to
have contained an adult spider with young dur-
ing the spring of 1985.

The average number of spiderlings taken from
cage walls (TV = 6 cages) during each month of
the 1983 caging experiment were: X = 7.5, SD
= 16.3, July; X = 7.6, SD = 23.4, August; X =
3.0, SD =12.0, September; X = 0.8, SD = 0.9,
October. Thus, 80% of the spiderlings found on
cage walls were found there during July and Au-
gust. Unaltered (control) plots contained a higher

52

THE JOURNAL OF ARACHNOLOGY

percentage of new burrows than any of the 3
treatment plots (control 40.5%, open 16.2%, lit-
ter removed 10.8%, litter added 32.4%.) Burrow
disappearance was higher in the treatment areas
(control 12.0%, open 26.3%, litter removed
14.0%, litter added 18.0%).

DISCUSSION

Although G. turricola was originally thought
to have a one-year life cycle (Wallace 1942), our
observations (Miller & Miller 1987) indicated
that the species has a two-year cycle with a single
reproductive period during the second year. This
is similar to that reported for other Geolycosa
(e.g., G. fatifera (Hentz), G. missouriensis ChdLm-
berlin, and G. pikei (Marx); Wallace 1942). Ma-
ture males appear only in late August and Sep-
tember prior to the final molt of the female. These
males may cohabitat with the immature female
for a short time prior to her last molt (Miller &
Miller 1986) at which time courtship and cop-
ulation occur. Once mated, females cover the
entrance to their burrow, overwinter there and
produce egg cases in the spring (Miller & Miller
1986). Males die after mating in the fall. The
young reach reproductive age during the fall of
their second year. In our study population, the
spiders that survived the winter of 1983-1984
were, thus, of the same cohort rather than prog-
eny of early and late breeding adults of the same
year. As we discuss below, the differences in the
diameter of initial burrows and the timing of the
establishment of these burrows are probably the
result of variation in behavior related to the de-
parture from the maternal burrow or the process
of burrow establishment itself.

Although the importance of the burrow during
all of the life stages of Geolycosa is widely ac-
cepted, there is still considerable uncertainty
about the timing of initial burrow establishment,
the factors that affect the positioning of the bur-
row and the extent to which spiders change bur-
row locations during their lifetime. With respect
to the establishment of the first burrow, the ques-
tion remains as to whether spiderlings build bur-
rows immediately following dispersal or whether
there is a delay between dispersal and burrow
establishment during which time spiderlings use
natural retreats. McQueen (1978) intimates that
construction of first burrows in G. domifex Han-
cock coincided with dispersal although little data
for this conclusion is given. Conley (1985) has
suggested that spiderlings of the western species
G. rafaelana (Chamberlin) overwinter in natural

retreats and build their first burrows in the spring.
However, our observations of several hundred
marked burrows of that species in Utah showed
that new burrows are constructed in the early fall
(Miller & Miller unpublished data). The results
presented here for G. turricola suggest that both
situations may exist in a single population of this
species. The onset of burrow construction co-
incides with the time at which the most spider-
lings were found on the walls of cages providing
evidence that some spiderlings construct bur-
rows immediately following dispersal from their
mother’s burrow. However, the census data show
that this pattern may not hold for all dispersing
spiderlings. A large number ( 1 36) of new burrows
appeared in the early spring of 1984. It is likely
that these are spiderlings that hatched in 1983,
overwintered either in their mother’s burrow or
in a retreat, and then constructed their first bur-
row in the spring (the strategy suggested by Con-
ley 1985). This is supported by the observation
that no young were hatched in the study popu-
lation in spring 1984, and the average diameter
of the spring 1984 burrows was nearly the same
as that of previously marked burrows that sur-
vived the winter of 1983 (Table 1). Moreover,
the Selma Chalk soil inhabited by this population
is usually dry and replete with small cracks dur-
ing the dispersal period. These soil cracks could
provide temporary retreats for dispersing spi-
derlings (Miller 1984a).

A number of researchers have shown that bur-
row density is not an important factor influenc-
ing the survival of Geolycosa spiders (e.g.,
McQueen 1983; Conley 1985). However, the im-
portance of the position of the burrow with re-
spect to physical features of the habitat and, thus,
possibly to critical resources, may be important.
McQueen (1983) observed that the burrows of
G. domifex were usually placed in unshaded open
areas. One of us (Miller 1984a) addressed the
issue of habitat preference in a series of labora-
tory studies with G. micanopy Wallace and G.
turricola. In those studies it was shown that the
tendency to establish a burrow was related to the
presence of vegetation and the feeding experience
of the spiderling. The study also showed that
these factors differed between species. The results
of the present study, though limited, corroborate
Miller’s (1984a) study by indicating that in G.
turricola burrow sites in grassy areas and grassy
areas with considerable litter are favored over
open, uncovered positions. The mortality of the
spiderlings that established burrows in vegetated

MILLER & MIUJEK-GEOLYCOSA POPULATION DYNAMICS

53

areas was somewhat reduced although the study
is too limited for a strong conclusion in that re-
gard.

There is considerable uncertainty about
whether Geolycosa spiders change burrow loca-
tions during their lifetime. It is generally thought
that such changes are uncommon (e.g., Wallace
1942; Conley 1985). Our observations of over
500 marked burrows in Mississippi (G. turricola;
this study) and southern Utah {G. mfaelana) tend
to support this (Miller & Miller unpublished data).
The month-to-month decrease in the number of
previously active burrows is much smaller than
the number of newly discovered burrows. Thus,
it is unlikely that those burrows marked as newly
discovered are actually spiders that have moved
burrow location. Nevertheless, it is possible (per-
haps likely) that some threshold of site tenacity
exists for most habitats. Indeed, McQueen (1978)
suggested that many individuals of the species
G. domifex change burrow locations during the
early spring. The results of the present study do
not conclusively rule out burrow position changes
in this species, but they do suggest that such ac-
tivity is uncommon in the population that we
studied.

The early survival of burrowing wolf spiders
is thought to be extremely low (Humphreys 1976;
McQueen 1978, 1983). Humphreys (1976) ob-
served that over three-quarters of the spiderlings
of G. godeffroyi (L. Koch) in the two smallest
size classes died. McQueen found that nearly all
(90%) of the young of the year of a population
of G. domifex in Canada died within several
months of hatching. Estimates of adult survi-
vorship suggest that fewer than 10% survive to
reproductive age (usually two to three years)
(McQueen 1 983). The results presented here cor-
roborate the observation of low survival to re-
productive age but suggest that high spiderling
mortality may not be the rule among the species
in this genus.

First winter survivorship in our population was
considerably higher than that of populations of
other species of Geolycosa (e.g., McQueen 1 983).
Nearly all of the first-year burrows marked as
active in 1983 remained active the following
spring. Further, the survivorship through the
summer of 1 984 appears to be high. The decrease
in the number of active burrows in the fall of
that year is primarily the result of ( 1 ) the mor-
tality of adult males that have left their burrows
to mate (and subsequently to die; see below), and
(2) burrow covering by mated females.

The reason for the high spiderling survival in
this population over the first winter is uncertain.
The factors influencing mortality of dispersing
Geolycosa spiderlings are unknown but are likely
to include predation, failure to find a suitable
burrow site, or parasitism. Mortality factors re-
lated to the density of burrows are probably not
important (McQueen 1983; Conley 1985). A
possible explanation is related to their dispersal
strategy. A portion of the broods of G. turricola
remain in the maternal burrow well beyond the
time when successful dispersal is possible where-
as other brood members disperse shortly after
hatching (Miller 1989). Humphreys (1983) re-
ported the existence of a phasic dispersal pattern
in the European tarantula Lycosa tarantula (L.).
He suggested that such a mixed dispersal strategy
might be an advantage in temporally varying en-
vironments. Miller (1989) hypothesized that spi-
derlings in these subsocial groups have a higher
chance of surviving the first winter because they
build deeper burrows than spiderlings that dis-
perse shortly after emergence from the egg case.
The relationship between burrow depth and sur-
vival was first recognized by Humphreys (1973,
1978). Although the results presented here do
not establish a direct link between subsociality
and the timing of burrow construction, they lend
support to that hypothesis. Spiderlings con-
structing burrows in August made larger burrows
and enjoyed a higher overwinter survival rate
than those constructing burrows in July.

It should be noted that if the spiderlings dis-
covered in August are participants in a subsocial
group, the timing of their dispersal is earlier than
that predicted from Miller’s (1989) laboratory
studies. This suggests that the extent to which
extended tolerance among Geolycosa brood mates
exists may be mediated by environmental con-
ditions. Clearly, studies of the spatial, temporal
and taxonomic variation in subsocial organiza-
tion in this genus are needed to delimit the nature
and strength of these environmental constraints.

In terms of the total number of active burrows
observed during the study, the survivorship over
the second winter appears low. Of the 1983 group
only 50 survived the winter of 1984-1985 and
were recorded active in the spring of 1985, and
only five of those (1.7% of 1983 active burrows)
were observed to be active beyond March of that
year. However, when viewed in terms of the spe-
cies’ life cycle, second winter survivorship is high.
If it is assumed that there is no sex-related mor-
tality then approximately one-half of the 1 39 spi-

54

THE JOURNAL OF ARACHNOLOGY

ders that were active in July of 1 984 were males.
As discussed above, these males would mature,
mate and die prior to the winter of 1984-1985.
The remaining spiders that survive would be-
come mature females and remain in their bur-
rows during the winter of 1984-1985 to produce
young the following spring. Thus, the fifty sur-
vivors of the winter of 1984-1985 represent a
majority of the spiders that would have had a
chance to survive that winter.

ACKNOWLEDGMENTS

We appreciate the comments of B. Main, J.
Weaver, T. Forrest and G. Stratton on early drafts
of this paper. We thank G. Baker, M. LaSalle,
C. Ware, J. Jenkins, and T. Furst for their help
in the field during portions of this study.

LITERATURE CITED

Conley, M. R. 1985. Predation versus resource lim-
itation in survival of adult burrowing wolf spiders
(Araneae: Lycosidae). Oecologia (Berlin), 67:71-75.
Harper, L. 1857. Preliminary report on the geology
and agriculture of the state of Mississippi. E. Barks-
dale Publishers, Jackson, MS.

Humphreys, W. F. 1973. The environment, biology
and energetics of the wolf spider Lycosa godeffroyi
(L. Koch 1865). Ph.D. thesis, Australian National
University, Canberra.

Humphreys, W. F. 1976. The population dynamics
of an Australian wolf spider, Geolycosa godeffroyi
(L. Koch 1865) (Araneae: Lycosidae). J. Anim. Ecol.,
45:59-80.

Humphreys, W. F. 1978. Thermal biology of Geo-
lycosa godeffroyi and other burrow inhabiting Ly-
cosidae (Araneae) in Australia. Oecologia (Berk), 3 1 :
319-347.

Humphreys, W. F. 1983. Temporally diphasic dis-
persal in siblings of a wolf spider: a game of Russian
roulette? Bull. Br. Arachnol. Soc., 5:124-126.

McQueen, D. J. 1978. Field studies of growth, re-
production, and mortality in the burrowing wolf
spider Geolycosa domifex (Hancock). Can. J. Zook,
56:2037-2049.

McQueen, D. J. 1983. Mortality patterns for a pop-
ulation of burrowing wolf spiders, Geolycosa dom-
ifex (Hancock), living in southern Ontario. Can. J.
Zook, 61:2758-2767.

McQueen, D. J. & B. M. Culik. 1981. Field and lab-
oratory activity patterns in the burrowing wolf spi-
der Geolycosa domifex (Hancock). Can. J. Zook, 59:
1263-1271.

Miller, G. L. 1984a. The influence of microhabitat
and prey availability on burrow establishment of
young Geolycosa turricola (Treat) and G. micanopy
Wallace (Araneae: Lycosidae): A laboratory study.
Psyche, 91:123-132.

Miller, G. L. 1984b. Ballooning in Geolycosa turri-
cola (Treat) and Geolycosa patellonigra Wallace: high
dispersal frequencies in stable habitats. Can. J. Zook,
62:2110-2111.

Miller, G. L. 1989. Subsocial organization and be-
havior in broods of the obligate burrowing wolf spi-
der Geolycosa turricola (Treat). Can. J. Zook, 67:
819-824.

Miller, G. L. & P. R. Miller. 1984. Correlations of
burrow characteristics and body size in burrowing
wolf spiders (Araneae: Lycosidae). Fla. Entomok,
67:314-317.

Miller, G. L. & P. R. Miller. 1986. Pre-courtship
cohabitation of mature male and penultimate fe-
male Geolycosa turricola (Araneae, Lycosidae). J.
Arachnol., 14:133-134.

Miller, G. L. & P. R. Miller. 1987. Life cycle and
courtship behavior of the burrowing wolf spider
Geolycosa turricola (Treat)( Araneae, Lycosidae). J.
Arachnol., 15:385-394.

Wallace, H. K. 1942. A revision of the burrowing
spiders of the genus Geolycosa (Araneae, Lycosidae).
Amer. Midland Natur., 27:1-61.

Manuscript received May 1 990, revised September 1990.

1991. The Journal of Arachnology 19:55-61

A REVISION OF THE GENUS ZORA
(ARANEAE, ZORIDAE) IN NORTH AMERICA

David T. Corey and Daniel J. Mott*: Department of Zoology, Southern Illinois Uni-
versity, Carbondale, Illinois 62901 USA

Abstract. The genus Zora C.L. Koch, 1 848 in North America includes two species: Zora
pumila (Hentz) and Zora hespera new species. Diagnoses, descriptions, distributions, and
natural history notes are presented.

The genus Zora consists of small to medium
entelegyne, ecribellate spiders. They may be rec-
ognized by having two claws with claw tufts, dis-
tinct longitudinal bands on the cephalothorax,
4-2-2 arrangement of the eyes and a series of
long overlapping spines on the first two tibiae
and metatarsi. The color pattern on the abdomen
is distinct and, in unfaded specimens, may be
useful in distinguishing the species.

The 17 species in the genus have a Holarctic
distribution with most of the species reported
from Europe and the Middle East. One species
is cited from the United States. Neither the genus
nor the family has been revised.

Zora has been placed in the Lycosidae (Dahl
&Dahl 1927), the Ctenidae(Petrunkevitch 1928;
Homann 1947), the Clubionidae (Kaston 1948)
and the Zoridae (Dahl 1912; Tullgren 1945; Leh-
tinen 1967; Kaston 198 1). Currently several oth-
er genera, whose relationships are debatable, are
also placed in the Zoridae.

METHODS

Specimens were examined and measured un-
der a stereo dissecting microscope with an ocular
micrometer mounted in one eyepiece. Microm-
eter units were converted to metric units and
these rounded to the nearest 0.01 mm. Drawings
were made with the aid of a squared-grid reticle
in one eyepiece of the dissecting microscope. A
total of 172 specimens was examined. Epigyna
were removed and cleared in clove oil.

Abbreviations for eyes: are as follows: AME
(anterior median eye), ALE (anterior lateral eye),
PME (posterior median eye), PLE (posterior lat-

' Present address: Department of Mathematics and Sci-
ences, Dickinson State University, Dickinson, North
Dakota 58601 USA.

eral eye), MOQ (median ocular quadrangle). Di-
mensions are given in the form: range, mean {X)
and standard error (SE).

Abbreviations for collections cited in text are:
AMNH = American Museum of Natural His-
tory, CAS = California Academy of Sciences,
MCZ = Museum of Comparative Zoology,
USNM = United States National Museum, SIUC
= Southern Illinois University at Carbondale,
JAB = Joseph A. Beatty.

Genus Zora C. L. Koch, 1 848

Lycaena Sundevall, 1832:265. Type species by mono-
typy, L. spinimana Sundevall. Preoccupied.
Hecaerge Blackwall, 1833:193. Type species Lycaena
spinimana (Sundevall). Preoccupied.

Lycodia (lapsus ?) Sundevall, 1833:22.

Dolomedes Walckenaer, 1837:348. (part) (3e race: Ru-
piariae).

Zora C. L. Koch, 1848:91. Type species Lycaena spi-
nimana Sundevall.

Psilothra Gistel, 1848: IX. Proposed replacement for
Hecaerge, preoccupied.

Katadysas Hentz, 1850:287. Type species by mono-
typy K. pumilus Hentz.

Thorell, 1869:43. Emendation of Ka?ac(y5a5

Hentz.

Diagnosis.— Zora can be distinguished from
most other eight-eyed spiders by the nearly
straight anterior eye row, and the strongly re-
curved posterior eye row, the eyes forming three
rows. They have 2 claws with claw tufts. The
presence of 6-8 pairs of long, overlapping spines
on tibiae I and II separate them from the Cten-
idae and Lycosidae.

Description.— Small spiders with a general or-
ange-brown color. A wide band extending back
from each posterior lateral eye (Figs. 5, 1 1). A
marginal band on each side of the cephalothorax.
Carapace highest in the region of thoracic groove.

56

THE JOURNAL OF ARACHNOLOGY

Figures. \-l. — Z. pumila: 1, 2, male palpus; 1, retrolateral aspect; 2, ventral aspect; 3-7, female; 3, tibia and
metatarsus leg I; 4, chelicerae; 5, dorsal view of female; 6, 7, epigynum; 6, dorsal aspect, cleared; 7, ventral
aspect.

COREY & MOTT-REVISION OF AMERICAN ZORA

57

Sternum with a dark spot near each coxa (Fig.
10). Labium wider than long. The legs heavily
spotted, leg IV longest, followed by I, II, III. Tarsi
with 2 claws and tufts. Tibiae I and II with 6-8
pairs of long, overlapping ventral spines. Meta-
tarsi I and II with 2-3 pairs of long, overlapping
ventral spines. Palpal tarsus of females and ju-
veniles with a pair of ventral spines. Retromargin
of chelicerae usually with two teeth, promargin
usually with 3 teeth. Apical segment of the pos-
terior spinnerets short and indistinct.

Distribution. — Holarctic. Specimens have been
reported from Mexico and Brazil by Marx ( 1890)
under the nomina nuda Zora californica and Z.
latithorax. The localities are likely to be in error.

Zora pumila (Hentz)

Figs. 1-7. Map 1

Katadysas pumilus Hentz, 1850:287, plate X, fig. 16,
1 male imm. Type destroyed.

Catadysas pumilus Thorell, 1869:43.

Zora pumila Wolmherg, 1882:156.

Z. spinimana Emerton, 191 1:403, plate V, figs. 5a-5b,
1 female. Not Z. spinimana (Sund.)

Z. pumilus Comstock, 1912:403, fig. 651; 1940:587.
fig. 651.

Diagnosis.— Males of Z. pumila are distin-
guished from males of Z. hespera by the larger
palp, and the angular conductor (Fig. 2). Females
have sperm ducts coiled anterior to the sper-
mathecae (Fig. 6). Zora pumila has been con-
fused with the European Zora spinimana (Sund.).
Examination of specimens of Z. spinimana
showed that they are distinct. The tibial apoph-
ysis of the male palp of Z. spinimana and Z.
nemoralis (Blackwall) has a broad base and is
bent distally. The distal end of tibial apophysis
in Z. spinimana is bifurcate and in Z. nemoralis
it is truncate. In female Z. spinimana the recep-
tacles are closer together than in Z. pumila and
Z. hespera and the sperm ducts turn forward and
parallel with the long axis of the body at their
anterior end. The course of ducts in Z. nemoralis
is different from Z. pumila and Z. hespera.

Description.— A wide marginal band on each
side of cephalothorax, wide paramedial band ex-
tending back from each posterior median eye and
a thin Y-shaped mark on abdomen. Legs heavily
spotted. Chelicerae as in Fig. 4.

Males: Measurements of 10 specimens (mm).
Total length 3.54-4.10 (3.74, 0.06). Carapace
length 1.69-1.95 (1.81, 0.03), width 1.25-1.54
(1.40, 0.03). Eye sizes and interdistances AME

0.08, ALE 0.08, PME 0.10, PLE 0.10; AME-
AME 0.08, AME-ALE 0.08, PME-PME 0.10,
PME-PLE 0.13, ALE-PLE 0.16; MOQ length
0.23-0.31 (0.28, 0.01), front width 0.23-0.28
(0.24, 0.01), back width 0.15-0.31 (0.28, 0.02).
Clypeus height 0.05-0.10 (0.09, 0.01). Chelicera
length 0.31-0.54 (0.46, 0.03), width 0.23-0.33
(0.27, 0.01). Total lengths of legs; I: 5.28-6.51
(5.74, 0.12), II: 4.92-6.05 (5.45, 0.12), III: 4.21-
5.92(5.14, 0.15), IV: 5.38-8.31 (7.39,0.27). To-
tal length of palp 1.36-1.95 (1.79, 0.06). Leg spi-
nation; tibiae I and II with 6-7 pairs of long,
overlapping ventral spines, metatarsi I and II
with 2 pairs of long, overlapping ventral spines.

Females: Measurements of 1 0 specimens (mm).
Total length 4.00-6.00 (4.78, 0.22). Carapace
length 1.29-2.31 (1.94, 0.10), width 1.10-1.80
(1.49,0.11). Eye sizes and interdistances AME
0.08, ALE 0.08, PME 0.10, PLE 0.10; AME-
AME 0.08, AME-ALE 0.10, PME-PME 0.13,
PME-PLE 0.15, ALE-PLE 0.16; MOQ length
0.26-0.35 (0.30, 0.01), front width 0.24-0.31
(0.27, 0.01), back width 0.28-0.37 (0.33, 0.01).
Clypeus height 0.05-0.13 (0.10, 0.01). Chelicera
length 0.30-0.77 (0.57, 0.04), width 0.26-0.33
(0.30, 0.01). Total length of legs; I: 5.59-6.90
(6.20, 0.15), II: 5.08-6.62 (5.89, 0.19), III: 4.49-
6.08 (5.41, 0.15), IV: 5.46-9.10 (8.02, 0.36). Total
length of palp 1.62-3.31 (2.25, 0.17). Leg spi-
nation; tibiae I and II with 8 pairs of long, over-
lapping ventral spines, metatarsi I and II with 3
pairs of long, overlapping ventral spines.

Distribution.— Eastern United States from
Massachusetts south to Florida, and west to Mis-
souri and Texas.

Natural history.— This spider hunts in tall
grass and bushes during daylight. The egg sac is
guarded by the female, but no protective retreat
is built (Kaston 1981). Females have been taken
from January through October, and males from
April through July. Immatures can be found year
round.

Specimens have been collected in pitfall traps
from a pond pine community in central Florida
(Corey & Taylor 1 988) and from a brushy prairie
in Johnson Co., Missouri. In North Carolina they
have been collected in broomsedge and young
pine litter with a Berlese funnel. A juvenile was
found in foam skimmed from Indian Creek,
Jackson Co., Illinois.

Material examined. — USA: Alabama; Lee Co.,
Auburn (N. Banks), immature (MCZ); Madison
Co., Monte Sano, December 1 940 (A. F. Archer),

58

THE JOURNAL OF ARACHNOLOGY

Map 1.— Distribution of Zom in North America.

immature (AMNH). Connecticut; Middlesex Co.,
Killingworth, 23 June 1935 (B. J. Kaston), 3 fe-
males, immature (USNM); New Haven Co., vi-
cinity of Norwalk, 2 July 1933 (W. Ivie), female,
immature (AMNH), Norwalk, August 1933 (W.
Gertsch), female (AMNH), Seymour, 30 May
1 965 (J. & W. Ivie), female, 2 immature (AMNH).
Florida; Orange Co., Univ. of Central Florida
Campus, May 1983 (D. T. Corey), female
(USNM). Georgia; Emanuel Co., N of Swains-
boro, 23 December 1952 (W. Ivie), immature
(AMNH). Illinois; Jackson Co., R1 W, TIOS, S25,
NW 1/4, 21 March 1975 (R. Parkin), immature
(SIUC). Massachusetts; Middlesex Co., Hollis-
ton, 5 May and 1 July 1923 (J. H. Emerton), 2
females, 2 immatures (MCZ), August 1928, fe-
male (MCZ), 9 September 1928 (N. Banks), im-
mature (MCZ), 15 September 1928 (N. Banks),
female (MCZ), 16 June 1929 (N. Banks), female,
3 immature (MCZ), Tyngsboro, 1 6 October 1 908
(J. H. Emerton), female (MCZ), 23 January 1910
(J. H. Emerton), female, immature (MCZ). Mis-
souri; Johnson Co., Knob Noster State Park, 1 5-
22 May 1978 (Peck), female (CAS), 16-26 May
1978 (Peck), immature (CAS), 11-16 April 1979

(PECK), 3 immatures (CAS); Phelps Co., Rolla,
Dry Fork Cr., 8-11 May 1951 (HEF, DLF), fe-
male (CAS); St. Louis, St. Louis Co., 13 May
1950 (W. Dowdy), immature (AMNH). New Jer-
sey; Atlantic Co., Great Egg Harbour River be-
tween Penny Pot and Weymouth, 3 May 1947
(H. Van Deusen), female (AMNH); Bergen Co.,
Ramsey, 23 June 1934 (W. J. Gertsch), female,
immature (AMNH), 5 June 1938 (B. J. Kaston),
4 females, 3 males, 2 immature (USNM), 5 July
1938 (W. Gertsch), male and female pieces
(AMNH); Hunterdon Co., Lambertville, May
1952 (W. Ivie), male (AMNH), June 1953 (W.
Ivie), immature (AMNH). North Carolina; Dur-
ham Co., Duke forest, Durham, 1 1-28 April 1935
(A. M. Chickering), male (MCZ), SE comer of
Co. road 1116 and Chapel Hill Blvd. junction,

I October 1963 (J. W. Berry), immature (JAB),

II November 1963 (J. W. Berry), immature
(JAB), 1 May 1964 (J. W. Berry), male (JAB), 3
June 1964 (J. W. Berry), female (JAB). Ohio;
Harrison Co., Hopedale, 17 May 1979 (R. Ur-
banek), male (JAB). Pennsylvania; Bucks Co., E
of Jamison, Horseshoe Bend of Neshaminy Cr.,
November (W. Ivie), 2 immature (AMNH), Au-

Figures. 8-14.— Z. hespera: 8, 9, male palpus; 8, retrolateral aspect; 9, ventral aspect; 10-14, female; 10,
sternum; 11, dorsal view of female; 12, chelicerae; 13, 14, epigynum; 13, dorsal aspect, cleared; 14, ventral
aspect.

60

THE JOURNAL OF ARACHNOLOGY

gust 1953 (W. Ivie), 5 females, 19 immatures
(AMNH), September 1953 (W. Ivie), female, 9
immatures (AMNH), October 1953 (W. Ivie),
female (AMNH), April 1954 (W. Ivie), 2 im-
matures (AMNH), May 1954 (W. Ivie), 4 fe-
males, 7 males (AMNH), June 1955 (W. Ivie),
1 1 females (CAS). Tennessee; Knox Co., 29 Jan-
uary 1982 (G. Tolbert), immature (CAS). Texas;
Anderson Co., 7 mi. E of Palestine, 17 July 1938
(Davis), 3 immature (AMNH); Fayette Co., 1 1
mi. N of La Grange, 18 July 1966 (J. & W. Ivie),
female (AMNH).

Zora hespera, new species
Figs. 8-14, Map 1

Types. — Female holotype from east of Pollock
Pines, El Dorado Co., California, in a pine forest,
25 June 1953 collected by V. Roth. In American
Museum of Natural History.

Etymology.— Name is derived from the Latin
word for west.

Diagnosis.— Males and females can be distin-
guished from Z. purnila by the genitalia (Figs. 1 ,
2, 6, 9, 13, 14). Conductor of male rounded an-
teriorly. Sperm ducts in female S-shaped. See Z.
purnila for differences between Z. hespera and
Z. spinimana and Z. nemoralis.

Description. — Narrow marginal band on each
side of the cephalothorax. Wide paramedial band
extends back from each posterior median eye.
Wide, lateral, longitudinal light colored bands on
the abdomen. Chelicerae as in Fig. 12.

Males: Measurements of 3 specimens (mm).
Total length 2.80-3.14 (2.95, 0.12). Carapace
length 1.40-1.62 (1.27, 0.08), width 1.24-1.32
(1.27, 0.03). Eye sizes and interdistances: AME
0.06, ALE 0.06, PME 0.06, PLE 0.08; AME-
AME 0.05, AME-ALE 0.04, PME-PME 0.08,
PME-PLE 0.11, ALE-PLE 0.16. MOQ length
0.14-0.26 (0.22, 0.05), front width 0.16-0.18
(0. 1 7, 0.0 1), back width 0.24. Clypeus height 0.06.
Chelicera length 0.36-0.42 (0.38, 0.02), width
0.20-0.22 (0.21, 0.00). Total lengths of legs; I:
4.00-4.98 (4.49, 0.69), II: 3.80-4.72 (4.26, 0.65),
III: 4.58 {N= 1), IV: 5.32-5.98 (5.65, 0.46). Total
length of palp 1.64-1.74 (1.68, 0.04). Leg spi-
nation tibiae I and II with 6-7 pairs of long,
overlapping ventral spines, metatarsi I and II
with 2 pairs of long, overlapping ventral spines.

Females; Measurements of 1 0 specimens (mm).
Total length 3.54-6.79 (4.77, 0.37). Carapace
length 1.53-2.13 (1.88, 0.07), width 1.23-1.72
(1.51, 0.05). Eye sizes and interdistances AME
0.08, ALE 0.08, PME 0.10, PLE 0.10; AME-

AME 0.05, AME-ALE 0.08, PME-PME 0.10,
PME-PLE 0.13, ALE-PLE 0.18. MOQ length
0.28-0.36 (0.31, 0.01), front width 0.18-0.26
(0.23, 0.01), back width 0.28-0.36 (0.31, 0.01).
Clypeus height 0.08-0.13 (0.10, 0.01). Chelicera
length 0.44-0.67 (0.56, 0.02), width 0.20-0.42
(0.30, 0.02). Total lengths of legs; I: 4.03-6.28
(5.21, 0.22), II: 4.56-5.82 (4.97, 0.15), III: 3.97-
5.69 (4.58, 0.17), IV: 5.36-7.36 (6.48, 0.17). To-
tal length of palp 1.56-2.17 (1.89, 0.08). Leg spi-
nation, tibiae I and II with 6-7 pairs of long,
overlapping ventral spines, metatarsi I and II
with 2 pairs of long, overlapping ventral spines.

Distribution. — Washington, Oregon, Califor-
nia, Idaho, and Arizona.

Natural history.— Zora hespera has been col-
lected throughout the year: adult males taken in
April, May, and June, females from February
through December. They have been found by
sifting through oak leaves or by using a Berlese
funnel.

Material examined. — All adults are paratypes.
USA: Arizona; Yavapai Co., 5 mi. S of Prescott,
23 April 1936 (Bishop), 2 immature (AMNH).
California; El Dorado Co., Riverton, 1 5 July 1934
(W. Ivie), female (AMNH), E Pollock Pines, 25
June 1953 (V. Roth), female (AMNH); Los An-
geles Co., Tanbark Flats, San Gabriel Mts., 20
June 1952 (M. Cazier, W. Gertsch, R. Schram-
mel), female (AMNH); Mariposa Co., Wawona
Camp, Yosemite Park, 17 September 1941 (W.
Ivie), 3 females, 2 immatures (AMNH); Napa
Co., Monticello Dam, 23 October 1957 (R. O.
Schuster), immature (AMNH), Toll road on
Mount St. Helena, 31 December 1953 (G. A.
Marsh, R. O. Schuster, V. Roth), female, 2 im-
matures (AMNH); Orange Co., Santa Ana Mts.,
13 September 1941 (W. Ivie), 3 immatures
(AMNH); Riverside Co., Idyllwild, San Jacinto
Mts., 18 June 1952 (W. J. Gertsch, M. Cazier,
R. Schrammel), immature (AMNH), 17 March

1957 (1. Newell), female (AMNH); San Bema-
dino Co., 1.6 mi. E of Seven Oaks, 23 March

1958 (1. Newell), female (AMNH), Arrowhead
Lake, 6 May 1936 (Bishop), immature (AMNH);
San Diego Co., 4.8 mi. S of Julian, 26 April 1959
(1. Newell), female (AMNH), Sweetwater River,
1 1/2 mi. N of Descanso, 26 March 1967 (E. &
R. Musillo & J. Ivie), female (AMNH); San Ma-
teo Co., S of Woodside, 17 September 1964 (J.
& W. Ivie), female (AMNH); Sonoma Co., 3 mi.
W of Glen Ellen, 31 December 1953 (Marsh,
Schuster, Roth), immature (AMNH), 15 Feb-
ruary 1954 (Roth, Schuster), female, immature

COREY & MOTT- REVISION OF AMERICAN ZORA

61

(AMNH), Sugarloaf Ridge State Park, Bald Mt.
Trail (450m), 20 November 1982 (V. F. Lee),
immature (CAS); Yolo Co., 5.4 mi. SW of Win-
ters, 29 May 1939 (FCR, LMS, ROS), immature
(AMNH), 23 April 1959 (F. C. Ramey), 2 fe-
males (AMNH). Idaho; Adams Co., 7 mi. NE of
Council, 17 October 1944 (W. Ivie), 3 immature
(AMNH); Idaho Co., Clearwater Creek near
Kooskia, 23 August 1940 (W. Ivie), female, im-
mature (AMNH). Oregon; Benton Co., W of Cor-
vallis, 20 March 1937 (JCC), female (AMNH);
Columbia Co., Goble, 22 March 1938, female
(AMNH), 23 April 1938, female, male, imma-
ture (AMNH); Coos Co., Charleston, 17 June
1952 (B. Malkin), male (AMNH); Josephine Co.,
Grave Cr., 30 May 1952 (V. Roth), female
(AMNH); Lane Co., 8 mi. S of Divide Guard
Sta., 27 July 1955 (V. Roth), female (AMNH);
Marion Co., St. Benedict, 2 May 1954 (J. Roth),
male (AMNH); Yamhill Co., Pea vine Ridge, near
McMinnville, November- December 1946 (K. M.
Fender), female, 7 immatures (AMNH). Wash-
ington; Yakima Co., Tieton River, about 10 mi.
E of Rimrock, 1 3 September 1965 (J. & W. Ivie),
immature (AMNH).

ACKNOWLEDGMENTS

We thank J. Reiskind for suggesting this genus
as a subject for revision. We are indebted to J.
A. Beatty for his advice on many problems as
they arose. We thank J. Wunderlich for sending
specimens of Zora spinimana. H. Levi and J. A.
Beatty made critical comments on an earlier draft
of the manuscript. The following people kindly
lent specimens for this revision: J. Coddington,
Smithsonian, US National Museum; J. A. Beat-
ty, Southern Illinois University; H. Levi, Mu-
seum of Comparative Zoology; N. 1. Platnick and
W. J. Gertsch, American Museum of Natural
History; and W. J. Pulawski, California Acade-
my of Sciences.

LITERATURE CITED

Blackwall, J. 1833. Characters of some undescribed
genera and species of Araneidae. Lond. Phil. Mag.
J. Sci., 3:187-197.

Comstock, J. H. 1912. The Spider Book. Garden
City, New York.

Comstock, J. H. 1940. The Spider Book. Revised
and edited by W. J. Gertsch. New York.

Corey, D. T. & W. K. Taylor. 1988. Ground surface

spiders in three central Florida plant communities.
J. Arachnol., 16:213-221.

Dahl, F. 1912. Arachnoidea. /« Handwort. Naturw.,
1:485-514.

Dahl, F. & M. Dahl. 1927. Spinnentiere oder Arach-
noidea. II. Lycosidae s. lat. (Wolfspinnen im wei-
teren sinne). Tierw. Deuts., 5:1-80.

Emerton, J. H. 1911. New spiders from New En-
gland. Trans. Connecticut Acad. Arts Sci., 16:385-
407.

Gistel, J. 1848. Naturgeschichte des Tierreichs fur
hohere Schulen. Stuttgart.

Hentz, N. M. 1850. Descriptions and figures of the
Araneides of the United States. Boston J. Nat. Hist.,
6:271-295.

Holmberg, E. L. 1882. Observations a propos du
sous-ordre des Araignees territelaires, specialement
du genre nordamericain Catadysas Hentz et de la
sous-famille Mecicobothrioidae. Bol. Acad. Arg., 4:
153-174.

Homann, H. 1947. DieNebenaugenderSpinnen(Ar-
anea). Ztschr. Naturforschg., 2b: 16 1-1 67.

Kaston, B. J. 1948. Spiders of Connecticut. State
Geological and Natural History Survey of Con-
necticut. Bull. 70.

Kaston, B. J. 1981. Spiders of Connecticut. State
Geological and Natural History Survey of Con-
necticut. Second edition. Bull. 70.

Koch, C. L. 1848. Die Arachniden. Vierzehnter Band.
Numberg, 210 pp.

Lehtinen, P. 1 967. Classification of the cribellate spi-
ders and some allied families, with notes on the
evolution of the suborder Araneomorpha. Ann. Zool.
Fenn., 4: 199-468.

Marx, G. 1 890. Arachnida in Scientific results of ex-
plorations by the U. S. Fish Commission steamer
“Albatross”. Proc. U. S. Nat. Mus., 12:207-21 1.

Petrunkevitch, A. 1928. Systema Aranearum. Trans.
Connecticut Acad. Arts Sci., 29:1-270.

Sundevall, J . C. 1832. Svenska Spindlarmes beskrifn-
ing. Fortsattning och slut. N. Act. Reg. Soc. Sci.
Upsal., 171-272.

Sundevall, J. C. 1833. Conspectus arachnidum. Lon-
dini Eothorum.

Thorell, T. 1869. On European spiders. N. Act. Reg.
Soc. Sci. Upsal., 7:1-108.

Tullgren, A. 1945. Svensk spindel fauna 3. Egentliga
spindlar. Araneae Fam. 5-7, Clubionidae, Zoridae
och Gnaphosidae. Stockholm.

Walckenaer, C. A. 1837. Histoire naturelle des in-
sects. Apteres. Paris. Tome 1.

Walker, F. 1851. List of spiders captured by F. Walker.
Ann. Mag. Nat. Hist., 7:157-158.

Manuscript received June 1990, revised September 1990.

1991. The Journal of Arachnology 19:62-66

OBSERVATIONS ON THE BEHAVIOR OF THE
KLEPTOPARASITIC SPIDER, MYSMENOPSIS FURTIVA
(ARANEAE, MYSMENIDAE)

Frederick A. Coyle, Theresa C. O’Shields and Daniel G. Perlmutter: Department of
Biology, Western Carolina University, Cullowhee, North Carolina 28723 USA

Abstract. Mysmenopsis furtiva, a tiny spider which lives in the funnelwebs of the Ja-
maican diplurid spider, Ischnothele xera, behaves both as a kleptoparasite and as a com-
mensal; it pilfers portions of its host’s prey and also captures and consumes minute insects
which are trapped in the host web and unnoticed or ignored by the host. Mysmenopsis
furtiva is able to ingest hemolymph from its host’s prey at a much faster rate than it can
ingest material from the insects it captures. Two of its stealth strategies are to move not at
all or slowly when the host is motionless and to synchronize its rapid movements with host
movements. The host’s anti-kleptoparasite behaviors suggest that the kleptoparasite has a
significant negative impact on the host.

Kleptoparasitic spiders (those which regularly
steal food from other species of spiders) are known
to occur in five families (Vollrath 1987; Griswold
& Meikle-Griswold 1987): 1) Theridiidae (Ar-
gyrodes species), 2) Dictynidae (Archaeodictyna
ulova Griswold and Griswold), 3) Salticidae
{Portia and Simaetha species), 4) Symphytog-
nathidae {Curimagua bayano Forster and Plat-
nick), and 5) Mysmenidae. Three genera of mys-
menids contain species that are definitely or very
probably kleptoparasitic. Two of these, Isela and
Kilifa, are recently described monotypic genera
living in the funnelwebs of African diplurid spi-
ders; Isela okuncana Griswold is a common
kleptoparasite of Allot hele terretis Tucker (Gris-
wold 1985), and Kilifia inquilina Baert and Mur-
phy is a common inhabitant of Thelechoris kar-
schi Bosenberg and Lenz webs (Baert & Murphy
1987). In the tropical American genus Mysme-
nopsis (the sister group of Isela) three species {M.
ischnamigo Platnick and Shadab, M. gamboa
Platnick and Shadab, and M. dipluramigo Plat-
nick and Shadab) regularly feed on the prey of
their diplurid spider hosts (Vollrath 1987), one
(M. archeri Platnick and Shadab) feeds on the
prey of its pholcid hosts (Baptista 1988), two {M.
capae Baert and M. cienga Muller) have been
observed living in Cyrtophora webs (Baert 1 990),
and seven others (M. palpalis (Kraus), M. cid-
relicola (Simon), M. monticola Coyle and Meigs,
M. furtiva Coyle and Meigs, M. hauscar Baert,
M. pachacutec Baert, and M. tibialis (Bryant))
have been observed living in diplurid webs (Plat-

nick & Shadab 1978; Coyle & Meigs 1989; Baert
1990; and Alayon pers. comm.). These obser-
vations and the evidence for host-kleptoparasite
cospeciation in the Jamaican species, M. mon-
ticola and M. furtiva (Coyle & Meigs 1989),
suggest that many of the 26 known Mysmenopsis
species may be found to be obligate kleptopar-
asites or at least highly dependent on a klepto-
parasitic life style.

Although Coyle and Meigs (1989) assumed that
the sister species M. monticola and M. furtiva
are kleptoparasites, no direct evidence of klep-
toparasitism was available. In this paper we de-
scribe observations on the interaction of M. fur-
tiva with its host, Ischnothele xera Coyle and
Meigs, observations which demonstrate conclu-
sively that M. furtiva is a kleptoparasite.

METHODS

One adult female M. furtiva (body length ap-
proximately 1.5 mm) was collected from an I.
xera web on 20 May 1990 about 15 miles east
of Kingston, Jamaica, very near the type locality
for both species (Coyle & Meigs 1989, 1990).
The kleptoparasite was then transported to our
lab in Cullowhee, North Carolina, and released
on 25 May into the web of a 14 mm long adult
female I. xera collected at the same place and
time. The host’s web was constructed between
two vertical panes of glass (15 x 24 cm) sepa-
rated by 1.5 cm thick strips of wood along the
sides and bottoms of the panes. The kleptopar-
asite and host were observed periodically during

COYLE ET AL.- BEHAVIOR OF MYSMENOPSIS FURTIVA

63

daylight hours over the next five weeks, and 9.5
h of behavior were recorded with a Panasonic
WV-D5000 video recorder with a Micro-Nikkor
55 mm close-up lens. Live prey, dropped into
the host web to trigger prey capture and feeding
bouts (iV = 8), included four mealworm beetle
larvae (Tenebrio, 1 1-1 5 mm long), an adult house
cricket {Acheta domestica, 16 mm long), three
fruit flies {Drosophila, 2 mm long), and several
booklice and collembolans (1-3 mm long). Es-
timates of the increase in kleptoparasite abdom-
inal volume during feeding were obtained by
measuring the width and length of the abdomen
on the video screen before and after feeding, con-
verting these to real dimensions using carapace
width as the reference scale, and then computing
the abdominal volumes using the equation for
an ellipsoid (prolate spheroid), which is similar
to the shape of the abdomen. Increase in abdom-
inal volume was then divided by feeding dura-
tion to obtain an estimate of the rate of food
intake.

RESULTS

The kleptoparasite spent much of the time mo-
tionless in a small region of the host’s web in the
upper half of the arena well away from (10-15
cm) the host’s normal resting position in the bot-
tom of the arena. The kleptoparasite did not con-
struct any obvious web, but it should be noted
that we did not use methods appropriate for de-
tecting very delicate silk constructs. When mov-
ing about the host’s web (chiefly during periods
of host foraging/feeding activity) the kleptopar-
asite periodically attached its dragline.

Feeding. —During all three phases of the host
prey captures (approach, capture, and carry; Coyle
& Ketner 1990) the kleptoparasite either re-
mained motionless, moved slowly, or retreated
a short distance away from the activity. Only
after the host had returned with the prey to the
vicinity of its retreat and had begun feeding did
the kleptoparasite move toward the feeding site.
Two of these approaches began 1 min after the
host commenced feeding (one mealworm cap-
ture and the cricket capture); the rest began 9-
25 min after.

During two host feeding bouts (mealworm lar-
va and cricket), the kleptoparasite climbed onto
the prey while the host was feeding and moved
slowly over the prey surface, touching it occa-
sionally with pedipalps and mouthparts as if
searching for digestible substrate (Fig. 1). On both
occasions the kleptoparasite boarded the prey far

from the host’s mouthparts; only once did the
kleptoparasite approach close to the host’s
mouthparts, but it made no attempt to feed there.
We observed the kleptoparasite feeding on host
prey three times. In the first instance, while the
host fed on the neck of the cricket, the klepto-
parasite fed on hemolymph seeping from the base
of the cricket’s third leg, which we had removed
before dropping the cricket into the web. During
this feeding, which lasted 16.5 min, the volume
of the kleptoparasite’ s abdomen more than dou-
bled, increasing by approximately 0.72 mm\ for
an intake rate of 0.044 ml/min. Three and a half
hours later, while the host was feeding on the
cricket’s body, the kleptoparasite fed very briefly
on the neck region of the cricket’s head, which
had been removed by the host. Several days later
during another host feeding bout, the kleptopar-
asite fed on fluid (presumably hemolymph) at
the host-severed end of half a mealworm while
the host fed on the other half nearby. This feeding
lasted 10.1 min, not including two brief pauses
near the end of the feeding bout, and doubled
the kleptoparasite’s abdominal volume, increas-
ing it by approximately 0.86 mm\ for an intake
rate of 0.085 ml/min. Whenever the kleptopar-
asite fed, its legs and pedipalps were motionless
but its abdomen usually swayed slowly and
slightly.

We attempted to determine whether the klep-
toparasite would capture small prey on its own.
When we dropped three fruit flies onto the host
web, the host captured all three within a few
minutes and the kleptoparasite remained virtu-
ally motionless. At another time we dropped some
collembolans and booklice onto the web, the host
captured and fed on the largest collembolan, and
the kleptoparasite did not approach any of the
insects. On another day we dropped a very small
booklouse (about half the volume of the klep-
toparasite) and two collembolans (a sminthurid
and an entomobryid, each about one-third to half
the volume of the kleptoparasite) onto the web.
The host did not respond, but, after about 7 min,
the kleptoparasite began to approach and even-
tually captured (grabbed with its first legs and
bit) and consumed the booklouse. During this
83 min feeding bout, the kleptoparasite’s abdo-
men increased slightly in volume, by approxi-
mately 0. 1 9 mm^ for an intake rate of 0.002 ml/
min. It then proceeded in a similar manner to
capture and consume the two collembolans,
feeding on the first for 80 min. Since these two
feeding episodes were not video-recorded, it was

64

THE JOURNAL OF ARACHNOLOGY

Figure 1.— Adult female Mysmenopsis furtiva searching for a feeding site on a Tenebrio larva, which is being
consumed by an adult female Ischnothele xera.

not possible to obtain abdominal volume in-
crease estimates, but the increases appeared small.

Occasionally the kleptoparasite repeatedly
reached with its legs and gathered a few fine
strands of silk to its mouthparts. Whether the
spider was simply breaking strands to reduce a
small region of the host web or actually digesting
and ingesting web protein could not be deter-
mined.

Kleptoparasite movements.— The form and
speed of kleptoparasite movement through the
web depended markedly on the host’s behavior.
When the host was not moving, kleptoparasite
movements were typically slow, especially if the
host was not feeding. During such slow move-
ment, rarely did more than two or three legs
move at a time, and forward progress, when it
occurred, was only 0.2-0. 6 body lengths per sec-
ond (mean = 0.41 ± 0.12, A^= 11). Slow move-
ments typically involved rather slow waving and
probing motions of the legs, particularly the an-
terior pairs. Rapid kleptoparasite movements
occurred only when the host was moving (am-

bulatory and prey manipulation movements) and
consisted of very short advances and/or rapid leg
waving/probing as well as (less commonly) lon-
ger advances. Forward progress was much faster
during these advances than during slow advanc-
es, ranging from 3.0-1 1.1 body lengths per sec-
ond (mean = 5.44 ± 2.57, N = 7).

During one bout of host feeding when both
kleptoparasite and host behaviors were simul-
taneously recorded, we tallied whether or not
kleptoparasite movements occurred during host
movements; every one of the 1 1 6 rapid (and many
of the slow) kleptoparasite movement events oc-
curred during host movement. The onset and
cessation of most of these kleptoparasite move-
ments coincided with the beginning and end of
the host movements. On occasion, when the host
was feeding, the kleptoparasite would make
moderately fast movements ( 1 .0- 1 . 7 body lengths
per second, 77 = 3) while the host was not mov-
ing, but this usually (over 75% of the time) hap-
pened when there was a large prey item or dense
silk between the kleptoparasite and the host.

COYLE ET AL.- BEHAVIOR OF MYSMENOPSIS FURTIVA

65

Host responses to the kleptoparasite.— While
the host was feeding, we observed certain host
behaviors that do not normally occur during
feeding in a kleptoparasite-free web, and were
therefore almost certainly anti-kleptoparasite be-
haviors. Silk application: On 20 occasions the
host interrupted feeding and applied silk to its
web, usually in the region between the klepto-
parasite and its feeding site (and/or unattended
pieces of prey). Frequently, much silk was ap-
plied (spinning duration = 7-113 s, mean = 25
± 28.1, = 12). Scouting/challenging: On 13

occasions the host released its prey, dashed or
walked quickly partway toward the kleptopar-
asite (which was usually approaching the host’s
feeding site or an unattended portion of prey),
paused, and then returned to the feeding site. The
non-persistent nature of this approach seems to
distinguish it from a prey capture approach.
Sometimes the kleptoparasite retreated in re-
sponse to the host’s advance. Occasionally the
host paused to apply silk during her return. Ag-
itatedfeeding: When the kleptoparasite was mov-
ing very close to or on the prey while the host
was feeding, prey manipulation by the host usu-
ally increased in frequency and intensity and was
sometimes accompanied by brief rapid tapping
of pedipalps on the prey or at the kleptoparasite.
Sometimes the kleptoparasite retreated during
these host behaviors. On a few occasions the host
suddenly carried the prey to a new position, leav-
ing the kleptoparasite behind. Web-biting: On
five occasions the host turned toward the klep-
toparasite, pulled part of the web toward its che-
licerae with its pedipalps, and quickly extended
and flexed its fangs into the silk in a biting move-
ment. Chasing: Three times we observed the host
chasing after the kleptoparasite, but every chase
was short and quickly aborted. The riskiest such
challenge for the kleptoparasite involved the host
lunging and striking at the spot where the klep-
toparasite had started its narrow escape, and then
feeding on the piece of prey on which the klep-
toparasite had been feeding.

DISCUSSION

Our observations show clearly that M. furtiva
is a kleptoparasite which readily approaches and
feeds upon prey captured by its host. Its feeding
does not require assistance from host digestive
enzymes, as evidenced by the kleptoparasite’s
successful ingestion of hemolymph at the crick-
et’s leg base, which had not been fed upon by
the host, and by its capture and consumption of

minute insects. Its stealthy behavior is also in-
dicative of a spider specialized for kleptopara-
sitism. Particularly noteworthy is its practice of
moving quickly only when the host is moving, a
stealth strategy also employed by Argyrodes ele-
vatasT aczanowski (V ollrath 1979) and one which
presumably takes advantage of the host’s prob-
able inability to separate informative web vibra-
tions from those it is generating. The possibility
that we observed M. furtiva ingesting its host’s
silk needs to be further explored; at least three
Argyrodes species feed on host silk (V ollrath 1981,
1987; Whitehouse 1986).

Our observations reveal that M. furtiva is also
an opportunistic predator which can detect, cap-
ture, and ingest tiny insects which are caught in
the host’s web and are unnoticed or ignored by
the host. The same capability has been observed
in kleptoparasitic Mysmenopsis species (M. isch-
namigo, M. gamboa, and M. dipluramigo) living
in diplurid webs in Panama (Vollrath 1978, 1987).
It would be interesting to know whether this kind
of commensalistic activity is a common and/or
crucial source of nutrition.

Clearly, this kleptoparasite ingested much more
material per unit of feeding time from host prey
than from the tiny insects it captured. The ap-
parent costs to M. furtiva of feeding as a klep-
toparasite (especially the risk of being captured
by the host) may be at least partly compensated
by the advantages of feeding on hemolymph, i.e.,
rapid ingestion and low digestive costs. Not only
can hemolymph be ingested without the time,
energy, and material costs of external digestion,
but many of its constituents are small molecules
which can be digested inexpensively. A cost/ben-
efit analysis of these alternate feeding strategies
should also consider the nutritional value of the
ingested food; since insect hemolymph is very
similar to intracellular fluids and contains rela-
tively high concentrations of amino acids, or-
ganic phosphates, proteins, and carbohydrates
(Rorkin & Jeuniaux 1974; Mullins 1985), it may
be nearly as valuable nutritionally per volume
as muscle and other tissues that are digested dur-
ing the consumption of minute prey.

Our observations suggest that Mysmenopsis
kleptoparasites, in spite of their small size, may
have an important negative impact upon their
hosts. Food stealing, the interruption of host
feeding, and the host’s (partly effective) anti-
kleptoparasite efforts, some of which are also per-
formed by the diplurid hosts of other Mysmenop-
sis species (Vollrath 1978, 1979, 1984), must in-

66

THE JOURNAL OF ARACHNOLOGY

crease the cost/benefit ratio of feeding for the
host. Just the existence of such anti-klepto par-
asite behaviors is indicative of a negative effect
by the kleptoparasite. When several Mysmenop-
sis adults and juveniles live together in one web
(a common situation in webs of the diplurids
Vollrath (1984) observed, adult female Ischno-
thele reggae Coyle and Meigs (Coyle & Meigs
1989), and probably also I. xera), the collective
cost to the host could be particularly important.

We suspect that a key reason why diplurid
webs are especially favorable for mysmenid klep-
toparasites is their persistence in time and space.
The first author’s field observations indicate that
adult female diplurids commonly occupy the
same web for one or more years. In addition, the
fine dense mesh, large size, asymmetry, and three-
dimensional nature of these webs, as well as the
high ratio of host size/kleptoparasite size (Fig.
1 ), should make it easier for a kleptoparasite to
avoid detection and capture.

These observations trigger many questions.
How important is the capture of tiny prey to the
economy of the kleptoparasite? Exactly how does
a group of these kleptoparasites affect the host?
Do the kleptoparasites interact aggressively, tol-
erate one another, or cooperate? How effective
are the host’s anti-kleptoparasite tactics? What
regulates the number of kleptoparasites in a web?
Further study of the interactions among these
mysmenid kleptoparasites and their hosts should
provide useful insights into the behavioral ecol-
ogy of kleptoparasitism and host-symbiont co-
evolution.

ACKNOWLEDGMENTS

We thank B. Freeman for assistance in the field
and R. Lumb for discussions about the biochem-
istry of insect hemolymph and digestion. This
study was supported by National Science Foun-
dation Grant BSR-8700298.

LITERATURE CITED

Baert, L. 1990. Mysmenidae (Araneae) from Peru.

Bull. Inst. Royal Sci. Nat. Belgique, 60:5-18.

Baert, L. L. & J. A. Murphy. 1987. Kilifia inquilina,
a new mysmenid spider from Kenya (Araneae, Mys-
menidae). Bull. British Arachnol. Soc., 7:194-196.
Baptista, R. 1988. On (Mysmenidae),

a kleptoparasite of Pholcidae. American Arachnol.,
(38):4 (abstract).

Coyle, F. A. & N. D. Ketner. 1 990. Observations on
the prey and prey capture behaviour of the funnel-
web mygalomorph spider genus Ischnothele (Ara-
neae, Dipluridae). Bull. Brit. Arachnol. Soc., 8:97-
104.

Coyle, F. A. & T. E. Meigs. 1989. Two new species
of kleptoparasitic Mysmenopsis (Araneae, Dipluri-
dae) from Jamaica. J. Arachnol., 17:59-70.

Coyle, F. A. & T. E. Meigs. 1990. Two new species
of Ischnothele funnelweb spiders (Araneae, Myga-
lomorphae, Dipluridae) from Jamaica. J. Arachnol.,
18:95-111.

Florkin, M. & C. Jeuniaux. 1974. Hemolymph: Com-
position. Pp. 256-308, In The Physiology of Insecta.
Vol. 5 (M. Rockstein, ed.). Academic Press, New
York.

Griswold, C. E. 1985. Isela okuncana, a new genus
and species of kleptoparasitic spider from southern
Africa (Araneae: Mysmenidae). Ann. Natal Mus.,
27:207-217.

Griswold, C. E. & T. Meikle-Griswold. 1987. Ar-
chaeodictyna ulova, new species (Araneae: Dictyn-
idae), a remarkable kleptoparasite of group-living
eresid spiders (Stegodyphus spp., Araneae: Eresi-
dae). Amer. Mus. Novitates, (2897): 1-1 1.

Mullins, D. E. 1985. Chemistry and physiology of
the hemolymph. Pp. 355-400, In Comprehensive
Insect Physiology, Biochemistry and Pharmacology.
Vol. 3 (G. A. Kerkut and L. I. Gilbert, eds.). Per-
gamon Press, Oxford.

Platnick, N. 1. & M. U. Shadab. 1978. A review of
the spider genus Mysmenopsis (Araneae, Mysmen-
idae). Amer. Mus. Novitates, (2661): 1-22.

Vollrath, F. 1978. A close relationship between two
spiders (Arachnida, Araneidae): Curimagua bayano
synecious on a Diplura species. Psyche, 85:347-353.

Vollrath, F. 1979. Behaviour of the kleptoparasitic
spider Argyrodes elevatus (Araneae, Theridiidae).
Anim. Behav., 27:515-521.

Vollrath, F. 1981. Energetic considerations of a spi-
der parasite-spider host system. Rev. Arachnol., 3:
37-44.

Vollrath, F. 1984. Kleptobiotic interactions in in-
vertebrates. Pp. 6 1-93, In Producers and Scroungers
(C. J. Barnard, ed.). Chapman and Hall, New York.

Vollrath, F. 1987. Kleptobiosis in spiders. Pp. 274-
286, In Ecophysiology of Spiders (W. Nentwig, ed.).
Springer- Verlag, Berlin.

Whitehouse, M. 1986. The foraging behaviours of
Argyrodes antipodiana (Theridiidae), a kleptopar-
asitic spider from New Zealand. New Zealand J.
Zool., 13:151-168.

Manuscript received August 1990, revised October 1990.

1991. The Journal of Arachnology 19:67-69

DEVELOPMENT AND REPRODUCTIVE POTENTIAL OF
FLORINDA COCCINEA (ARANEAE, LINYPHIIDAE)

Marianne B, Willey and Peter H. Adler: Department of Entomology, Clemson Uni-
versity, Clemson, South Carolina 29634 USA

Abstract. Development and reproduction of the red grass spider, Florinda coccinea (Hentz),
from South Carolina were studied under laboratory conditions (26 ± 2 °C). Both males and
females required five molts to reach maturity, although 10% of the males had one super-
numerary molt. Once mature, females lived approximately one month, or nearly twice as
long as males. Laboratory-reared females produced as many as six fertile egg sacs, whereas
field-collected females produced up to ten sacs. The first sac of laboratory-reared females
had the largest average clutch size - about 70 eggs. The reproductive capacity of females
mated with unmated males versus those mated with previously mated males was not sig-
nificantly different.

The red grass spider, Florinda coccinea (Hentz),
is common throughout southeastern North
America. Despite its abundance, biological in-
formation on this species, like that of the vast
majority of linyphiids, is sparse. The only bio-
logical study of F. coccinea is an unpublished
thesis on territoriality (Ross 1977). Because so
little is known about this species, we undertook
this study to determine the developmental and
reproductive biology under standardized labo-
ratory conditions, and to relate these aspects to
our field observations.

MATERIALS AND METHODS

We collected adult females of F. coccinea (N

= 34) from lily turf (Liriope muscaria) on the
Clemson University campus on 8 May 1989.
These spiders and all subsequent offspring were
held individually in plastic containers (3.7 cm
deep X 5.2 cm diameter) in an environmental
chamber (26 ± 2 °C, 65 ± 4% R. H., 14L:10D
photoperiod). Field-collected females had live
Drosophila melanogaster and moist cotton in their
containers at all times. All offspring had constant
access to moist cotton and were provided ap-
proximately ten small leafhoppers or flies per
day.

Developmental biology was determined as fol-
lows. On 6 June 1989, we removed ten spider-
lings from each of the first six clutches produced
by the field-collected females, and successfully
reared 22 females and 1 4 males to maturity. These
spiders (F,) were mated, and on 4 July 1989, we
removed a total of 60 of their offspring (Fj, 1 0

spiderlings from each of 6 clutches) and suc-
cessfully reared 21 females and 25 males to ma-
turity. For Fi and Fj spiderlings, we recorded the
duration of each post-emergence instar by ex-
amining containers daily for exuviae. Occasion-
ally (33.6%, N = 332) exuviae were not located,
and this caused variation in sample sizes. To test
for protandry, we determined the average num-
ber of days from the second molt to the final
molt for males and females. We used once-mat-
ed spiders to compare male (N = 13) and female
{N= 29) longevity from the date of maturity until
death.

To determine reproductive capacity of F. coc-
cinea , we monitored field-collected females and
their F, and Fj offspring. Fi females {N = 15)
were mated (8 days after the final molt) with
virgin males (A = 9, 7 days after final molt) or
with once-mated males (A = 6, 11-15 days after
final molt). F2 females (A = 18, 2-8 days after
final molt) were mated with virgin males (A =
1 5, 3-4 days after final molt) or with once-mated
males (A = 3, 7 days after final molt). We re-
moved females from each of their successive egg
sacs within 1 2 h of construction, and monitored
them for production of additional sacs. We re-
corded the date of construction of each egg sac
(A = 95), the clutch size (spiderlings plus un-
hatched eggs) (A = 93), and the date of first spi-
derling emergence from each egg sac (A = 50).
Oviposition times, clutch sizes, and spiderling-
emergence times did not differ significantly be-
tween females mated with virgin males and those
mated with experienced males (ANOVA, P >

68

THE JOURNAL OF ARACHNOLOGY

Table 1. — Days (x ± SE, N) required for development of Florinda coccinea in the laboratory (26 ± 2 °C). a =
supernumerary molts.

Intermolt interval

Generation

2-3

3-4

4-5

5-6

F,

6.4 ± 0.30, 35

4.6 ±0.13, 34

5.2 ± 0.12, 35

_

F2

5.5 ± 0.26, 40

4.0 ±0.17, 43

5.3 ± 0.11, 44

4.7 ± 0.33, 3“

0.05), so these groups were combined for sub-
sequent analyses. If there were no significant dif-
ferences between the Fi and F2 generations, data
for the two groups were combined.

We deposited voucher specimens of both sexes
in the Clemson University Arthropod Collec-
tion.

RESULTS

Developmental biology.— Both males and fe-
males required five molts to reach maturity, al-
though four males (10.2%) had one supernu-
merary molt (Table 1). The first of these five
molts occurred within the egg sac. Approxi-
mately 70% of second-instar spiderlings {N =
1 20) constructed webs after emergence from the
egg sac; the remainder, which did not construct
webs, died before the next molt. The number of
days required to reach maturity was not signif-
icantly different {F = 0.89, df = P = 0.3475)
between males {X = 15.8, SE = 0.41, A = 36)
and females {X = 15.2, SE = 0.30, N = 40),
although longevity of adult females {X = 27.8,
SE = 2.66 days, N = 29) exceeded that of adult
males {X = 16.1, SE = 3.15 days, N = 13) (/-
test, df= 40, P = 0.013).

Reproductive biology.— Of the laboratory-
reared females that produced egg sacs (N = 32),

more than half produced at least three sacs, with
the maximum number of sacs being six (Table
2). Intervals between production of the second
through fifth sacs did not differ significantly. The
first sac had a significantly larger average clutch
size than successive sacs (Table 3). Field-col-
lected females produced up to ten egg sacs {X =
4.4, SE = 0.46, N = 25); however, these values
are minima because females might have ovipos-
ited prior to collection. Egg sacs were not retained
in the web, but were constructed in the bottom
of the container, suggesting that in the field they
are deposited at the base of vegetation.

Spiderling-emergence times from successive
egg sacs {X = 12.8, SE = 0.23, A = 51 ) did not
differ significantly between generations (A = 0.3 1 ,
df = P = 0.5836) or among successive sacs {F
= 0.97, df=2,P^ 0.3858). Of the sacs in which
spiderlings were produced (74.2%, A = 93), young
were unable to emerge from 30.4%. The re-
maining sacs, including all fifth and sixth sacs,
contained only eggs.

DISCUSSION

Our study of F. coccinea is the first study of a
North American linyphiid to address develop-
mental rates within each stadium in males and

Table 2. — Days (x ± SE) between sequential ovi-
positions of Florinda coccinea in the laboratory (26 ±
2 °C). Times were not significantly different between
theF, and Fj generations (/"= 0.23, 1, P = 0.6354).

Adjusted means followed by different letters are sig-
nificantly different {P < 0.05; LSMeans procedure, SAS
Institute 1985).

Oviposition events

No. days between
oviposition events

A

Sac #1-Sac #2

4.8 ± 0.21a

25

Sac #2-Sac #3

5.7 ± 0.66a

19

Sac #3-Sac #4

5.3 ± 0.21a

10

Sac #4-Sac #5

6. 1 ± 0.76a

6

Sac # 5-Sac #6

9.7 ± 2.19b

3

Table 3.— Clutch sizes {X ± SE) of consecutive egg
sacs constructed by Florinda coccinea in the laboratory
(26 ± 2 °C). Clutch sizes were not significantly different
between the F, and Fj generations (F = 2.96, df= I,
P = 0.0893). Adjusted means followed by different
letters are significantly different (P < 0.05; LSMeans
procedure, SAS Institute 1985).

Egg sac
number

Clutch size

A

Sac#l

70.5 ± 2.71a

32

Sac #2

61.6 ± 3.06b

23

Sac #3

62.0 ± 2.13b

19

Sac #4

59.6 ± 3.03a

10

Sac #5

50.2 ± 7.75b

6

Sac #6

48.5 ± 9.96b

3

WILLEY & ADLER -REPRODUCTION OF FLORINDA COCCINEA

69

females. Shorter developmental times for F. coc-
cinea observed in our study relative to the few
developmental studies of Old World linyphiids
(Turnbull 1962; De Keer & Maelfait 1987a,
1987b), might be due to higher rearing temper-
atures and unlimited food in our study. Although
male spiders are generally smaller than females
and, therefore, require fewer molts (Foelix 1982),
male and female linyphiids are approximately
the same size and have the same number of molts
or, as in F. coccinea, males sometimes have su-
pernumerary molts.

Life cycles of linyphiids have been described
as either univoltine (Schaefer 1976; Christophe
1977), bivoltine (Baert 1978), polymorphic (Wise
1974, 1976), or diplochronic (Toft 1976; Wise
1984). F. coccinea in upstate South Carolina is
apparently multivoltine. However, the short gen-
eration time and overlapping cohorts from mul-
tiple egg sacs make it difficult to determine the
number of generations per year. We have found
all instars and adults of both sexes throughout
the year in South Carolina.

The developmental and reproductive biology
of F. coccinea is similar to that of other spiders.
For instance, a general trend among spiders is
for females to live longer than males (Foelix
1982), and for the first egg sac to have the largest
clutch size (Preston-Mafham & Preston-Mafham
1984). The production of six to ten egg sacs by
F. coccinea falls within the range found in other
linyphiids, although the clutch size of F. coccinea
was larger than that of Oedothorax fuscus (De
Keer & Maelfait 1987a) and Fwntinella pyr-
amitela (Austad 1982), perhaps because F. coc-
cinea was provided more food.

ACKNOWLEDGMENTS

We thank J. D. Culin for reviewing the manu-
script and H. S. Hill, Jr. for statistical advice.
This is technical contribution No. 3047 of the
South Carolina Agricultural Experiment Station,
Clemson University.

LITERATURE CITED

Austad, S. N. 1982. First male sperm priority in the
bowl and doily spider, Fwntinella pyramitela (Wal-
ckenaer). Evolution, 36:777-785.

Baert, L. 1978. Influence de la photo periodicite sur
la maturation ovarienne chez Gongylidiurn rufipes.
Rev. Arachnol., 2:23-27.

Christophe, T. 1977. Etude demographique d’une
population de I'araignee Macragus rufus (Wider)
(Linyphiidae). Bull. Soc. Zool. France, 102:187-196.

De Keer, R. & J. -P. Maelfait. 1987a. Laboratory
observations on the development and reproduction
of Oedothorax fuscus (Blackwall, 1834) (Araneida,
Linyphiidae) under different conditions of temper-
ature and food supply. Rev. Ecol. Biol. Sol., 24:63-
73.

De Keer, R. & J. -P. Maelfait. 1987b. Life history of
Oedothorax fuscus (Blackwall, 1834) (Araneae, Lin-
yphiidae) in a heavily grazed pasture. Rev. Ecol.
Biol. Sol., 24:171-185.

Foelix, R. F. 1982. Biology of Spiders. Harvard Uni-
versity Press, Cambridge, Massachusetts. 306 pp.

Preston-Mafham, R. & K. Preston-Mafham. 1984.
Spiders of the World. Facts on File, Inc., New York.
191 pp.

Ross, J. W. 1977. Evidence for territoriality in the
line-weaving spider, Florinda coccinea (Hentz). M.
S. Thesis, University of Tennessee.

SAS Institute. 1985. SAS User’s Guide: Statistics, 5th
ed. SAS Institute, Cary, North Carolina.

Schaefer, M. 1976. Experimentelle Untersuchungen
zum Jahreszyklus und zur Uberwinterung von Spin-
nen (Araneida). Zool. Jb. Syst., 103:127-289.

Toft, S. 1976. Life-histories of spiders in a Danish
beech wood. Natura Jutlandica, 19:5-40.

Turnbull, A. L. 1962. Quantitative studies of the food
of Linyphia triangularis Clerck (Araneae: Linyphi-
idae). Can. Entomol., 94:1233-1249.

Wise, D. H. 1974. Role of food supply in the pop-
ulation dynamics of the spider Linyphia marginaia.
Ph.D. Thesis, University of Michigan, Ann Arbor.

Wise, D. H. 1976. Variable rates of maturation of
the spider Neriene radiata (Linyphia marginata).
Am. Midi. Nat., 96:66-75.

Wise, D. H. 1984. Phenology and life history of the
filmy dome spider (Araneae:Linyphiidae) in two lo-
cal Maryland populations. Psyche, 91:267-288.

Manuscript received April 1990, revised October 1990.

1991. The Journal of Arachnology 19:70-71

RESEARCH NOTE

CENTRUROIDES HASETHI POCOCK, A JUNIOR SYNONYM
OF CENTRUROIDES TESTACEUS (DEGEER)
(SCORPIONES, BUTHIDAE)

DeGeer (1778) described Scorpio testaceus
from specimens collected in “Amerique” and,
since that time, no one has been able to assign
a correct locality to that taxon. This species, now
considered a valid member of the genus Centru-
roides, was redescribed by Sissom and Francke
(1983). Those authors discounted previous re-
cords of C. testaceus as being based upon misi-
dentifications, including long-accepted records
from Montserrat and Hispaniola (Haiti). Because
the two syntypes of C. testaceus represented dif-
ferent species, the lectotype designated by Sissom
and Francke remained the only known specimen
of C. testaceus.

While sorting through undetermined scorpion
material from the Field Museum of Natural His-
tory, Chicago, I had the opportunity to examine
a specimen of Centruroides hasethi hasethi Po-
cock from the island of Curagao. I was imme-
diately struck by the resemblance of the speci-
men from Curagao to the lectotype of C. testaceus
and borrowed the types of C. testaceus and C.
hasethi from their respective depositories. Com-
parison of the type specimens confirmed my sus-
picions that C. testaceus and C. hasethi were con-
specific.

Sissom and Francke (1983) mistakenly iden-
tified the lectotype of C. testaceus as a female
because its metasomal segments are not as long
and slender as those of males of most species of
Centruroides (including the male syntype of
Scorpio testaceus accompanying the lectotype).
Unfortunately, the lectotype was pinned and dried
and could not be sexed by the presence or ab-
sence of genital papillae. As a result, our mor-
phometric comparisons with C. hasethi were
based on females of that species. It is now clear
that the lectotype of C. testaceus is indeed a male,
and its morphometries and meristics are virtu-
ally identical with those of male C. hasethi from
Curagao (Bakker 1963).

Since Bakker’s (1963) study of the Centru-
roides populations of Curagao and neighboring

islands, C. hasethi has been considered polytypic,
with two distinct subspecies: C. hasethi hasethi
Pocock from Curagao and Bonaire, and C. has-
ethi arubensis (Bakker) from Aruba. Bakker
(1963) distinguished the two subspecies by the
following characters (based on comparisons be-
tween the populations of C. hasethi hasethi from
Curagao and C. hasethi arubensis from Aruba).
(1) in C. /z. hasethi, males have 27-29 pectinal
teeth and females 25-27 teeth; in C. h. arubensis,
males have 23-25 teeth and females 21-23; (2)
C. h. hasethi range up to 75 mm in body length,
whereas C. h. arubensis reach only 55 mm in
length; (3) C. h. hasethi have proportionately lon-
ger metasomal segments; and (4) C. h. hasethi
have proportionately longer pedipalpal femora
and patellae. Interestingly, according to mea-
surements and ratios published by Bakker (1963),
the morphometries of the Bonaire population are
in some cases intermediate between those of the
populations of Curagao and Aruba, suggesting
that further study of the taxonomic status of each
population may be warranted. Based on the ob-
servations above and on direct comparisons of
type specimens, there is no doubt that C. testa-
ceus and C. hasethi hasethi belong to the same
taxon. Further, because of the morphometric
similarities with specimens from Curagao, it is
probable that the lectotype of C. testaceus orig-
inated from that island, and I hereby restrict the
type locality of C. testaceus to Curagao, Neth-
erlands Antilles. Pending further investigation of
the various island populations, arubensis is con-
sidered a subspecies of C. testaceus.

As a consequence of the above observations,
the following synonymies are proposed: Centru-
roides hasethi Pocock, 1893 = Centruroides tes-
taceus (DeGeer, 1778); C. hasethi hasethi Po-
cock, 1893 = C. testaceus testaceus (DeGeer
1778); and C. hasethi arubensis (Bakker, 1963)
= C. testaceus arubensis (Bakker, 1963).

The author is grateful to T. Kronestedt (Na-
turhistoriska Riksmuseet, Stockholm) for allow-

RESEARCH NOTE

71

ing me to reexamine the lectotype of C. testaceus\
P. D. Hillyard (British Museum of Natural His-
tory, London) for allowing me to examine the
holotype of C. hasethi] and D. A. Summers, J.
Ashe, and R. F. Inger (Field Museum of Natural
History, Chicago) for allowing me access to their
undetermined scorpion material.

LITERATURE CITED

Bakker, M. A. 1963. A new subspecies of the scorpion
Rhopalurus hasethi. Studies on the Fauna of Cu-
rasao and Other Caribbean Islands, No. 70:102-

117.

1991. The Journal of Arachnology 19:71-72

DeGeer, C. 1778. Memoires pour servir a I’histoire des
Insectes. Stockholm, 7:337-349.

Sissom, W. D. & O. F. Francke. 1983. Redescription
of Centruroides testaceus (DeGeer) and description
of a new species from the Lesser Antilles (Scor-
piones: Buthidae). Occas. Papers Mus. Texas Tech
Univ., 88:1-13.

W. David Sissom: Department of Biology, Elon
College, Elon College, North Carolina 27244
USA.

Manuscript received July 1990, revised October 1990.

BOOK REVIEW

Polis, Gary A. 1990. The Biology of Scorpions.
Stanford University Press, Stanford, California.
587 pp. (Price $85.00.)

The editor of this volume, Gary Polis, and a
determined crew of nine contributors have com-
piled an impressive assemblage of fact and eso-
terica about world scorpions; 485 pages of text
in all and covering morphology, systematics, pa-
leontology, biogeography, population biology,
ecology, behavior, environmental physiology,
neurobiology, venom toxicology, and even my-
thology. The comprehensiveness of coverage and
the terse, highly readable style of the book suggest
what might be its hidden purpose— the attraction
of young scientists into an open field of over-
looked research. The first two chapters, a full
third of the book, give excellent preparation for
literature reading and research on scorpions.
There are diagrammatic summaries of basic
anatomy (Hjelle) and group systematics, with lu-
cid descriptions of biogeography and paleontol-
ogy (Sissom). The epic restudy of the scorpion
fossil record by Kjellesvig-Waering (published
1986) is carefully summarized so that even a
physiologist can follow the emergence of the group
from gill-breathing descendants of a eurypterid
line (Silurian aquatic forms), through the ap-
pearance of terrestrial fauna (probably upper De-
vonian) and their peak in species diversity during

the Carboniferous (at least 1 3 superfamilies com-
pared to the modem three). We are reminded
that the 1400 species surviving today are re-
markably similar to their Paleozoic ancestors al-
though mere remnants of what once was. There
are useful keys to modem families, subfamilies
and genera, with clear drawings of diagnostic fea-
tures used to distinguish them.

The middle third of this book is dedicated to
life history (Polis and Sissom), behavior (War-
burg and Polis), ecology (Polis) and predator-
prey relations (McCormick and Polis) of scor-
pions. Here we learn just how little is known
about such basic characteristics as embryology
(perhaps a dozen works in all, the most complete
published before 1900), post-natal growth and
sexual development (some live 25 years). Most
scorpions live in deserts or temperate regions of
the world as solitary, cannibalistic burrowers.
Given the impracticalities of working in deserts
at night, one is sympathetic to Polis’s assertion
that “ecology is the least known aspect of scor-
pion biology,” and is deserving of much greater
attention. Here Polis has been a vanguard, uti-
lizing portable UV lights (scorpion cuticle fluo-
resces under UV) to make broad-ranging obser-
vations of natural habits, population biology and
community stmcture. Many unanswered ques-
tions arise in these pages, especially concerning
tropical scorpions, which are virtually unstudied.

72

THE JOURNAL OF ARACHNOLOGY

This section is the book’s center of mass, and
while it may suffer from eclecticism, it should
provide strong stimulation for further field-ori-
ented research. Judging from the number of cof-
fee stains, these were for me the most difficult
and important chapters to read.

The last third of this volume contains excellent
summaries of environmental physiology (Had-
ley), neurobiology (Root) and venom biochem-
istry/pathophysiology (Simard and Watt). Here
we learn about the peculiar, often unique, aspects
of scorpion physiology and biochemistry that
have enabled these animals to invade harsh en-
vironments and to thrive as nocturnal insecti-
vores. For desert species, a water-impermeable
cuticle, low basal metabolism and conservative
periods of activity outside the burrow were pre-
dictable responses to heat and water stresses. But
who could have imagined the array of non-visual
cues these animals use in darkness to capture
prey and to locate suitable mates. Scorpion tox-
ins have attracted the greatest outside attention,
mostly from vertebrate physiologists and bio-
chemists who use them as tools for studies of
nerve and muscle excitability. As the primary
structures of these peptides become known, new

questions emerge concerning structure/function
relationships of protein macromolecules and
phylogenetic relationships within an ancient
group of organisms. Indeed, these invertebrate
incarnations of evil are very good preparations
for all sorts of biology, from molecular to evo-
lutionary levels of study. The book ends with a
practical chapter on field and laboratory methods
(collection, rearing, dissection, preservation) and
a wonderful synopsis of scorpion lore, both
mythological and historical, by Cloudsley-
Thompson.

Thumbing back through these pages, straight-
ening folded comers and reading notes penciled
in the margins, I am struck again by the overall
good quality of research on scorpions and the
care with which it has been presented here. One
senses in the doing and the writing a special af-
fection for these animals so long feared and ne-
glected by humans. It is as though justice has
finally been done and we are witness to the cer-
tain injection of a new subject into our science.

Philip H. Brownell: Department of Zoology,

Oregon State University, Corvallis, Oregon

97331 USA

INSTRUCTIONS TO AUTHORS

Manuscripts are preferred in English but may be accepted
in Spanish, French or Portuguese subject to availability of
appropriate reviewers. Authors whose primary language is not
English, may consult the Associate Editor for assistance in
obtaining help with English manuscript preparation. All man-
uscripts should be prepared in general accordance vrith the
current edition of the Council of Biological Editors ( CBE) Style
Manual unless instructed otherwise below. Authors are ad-
vised to consult a recent issue of the Journal of Arachnology
for additional points of style. Send three identical copies of
the typed material together with copies of illustrations to:
Gary L. Miller, Associate Editor
The Journal of Arachnology
Department of Biology
The University of Mississippi
University, Mississippi 38677 USA
Telephone: (601) 232-7495
FAX: (601) 232-5144
Bitnet: BYMILLER@UMSVM
Correspondence relating to manuscripts should also be di-
rected to the Associate Editor.

Manuscripts longer than 1 500 words should be prepared as
Feature Articles, shorter papers as Research Notes.

FEATURE ARTICLES

Title page.— Place in the upper left comer the complete name,
address, and telephone number of the author with whom proofs
and correspondence should be exchanged. Include FAX num-
bers and Bitnet addresses if available. Four lines below, type
the title, centered, all in capital letters. Another four lines be-
low, beginning with the left margin, type each author’s name
on a new line followed by the address.

Abstract.— The heading should be placed at the beginning
of the first paragraph set off by a period. In articles written in
English, a second abstract in an acceptable language may be
included pertinent to the nationality of the author(s) or geo-
graphic region(s) emphasized in the article. A second abstract
in English must be included in articles not written in the latter
language.

Text.— Double-space text, tables, legends, etc. throughout.
Three levels of heads are used. The first level (METHODS,
RESULTS, etc.) is typed in capitals and on a separate line.
The second level head begins a paragraph with an indent and
is separated from the text by a period and a dash. The third
level may or may not begin a paragraph but is underscored
and separated from the text by a colon. Use only the underline
method to indicate italics; DO NOT USE WORD PROCES-
SOR ITALICS CODES. Use the metric system unless quoting
text or referencing collection data. All decimal fractions are
indicated by the period regardless of language of the text.

Citation of references in the text.— Include within parenthe-
ses the surnames of the author(s) followed by the date of pub-
lication. A comma separates multiple citations by the same
author(s) and a semicolon separates citations by different au-
thors, e.g., (Smith 1970), (Jones 1988; Smith 1989), (Smith
1986, 1987; Smith & Jones 1989; Jones et al. 1989).

Literature cited section.— Use the following style:
Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid
composition of major ampullate gland silk (dragline) of Ne-
phila clavipes (Araneae, Tetragnathidae). J. Arachnol., 1 8:
297-306.

KrafR, B. 1982. The significance and complexity of com-
munication in spiders. Pp. 1 5-66, In Spider Communica-
tions: Mechanisms and Ecological Significance. (P. N. Witt
& J. S. Rovner, eds.). PrincetonUniversity Press, Princeton.

Footnotes.— Footnotes are only permitted on the first printed
page to indicate current address or other information con-
cerning the author(s). All footnotes are placed together on a
separate manuscript page.

Running head.— Author sumame(s) and an abbreviated title
must be typed all in caps and must not exceed 60 characters
and spaces. The running head is placed by itself on a separate
manuscript page.

Taxonomic articles.— Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Associate
Editor.

Tables.— Each table, with the legend above, should be placed
on a separate manuscript page. Only the appropriate horizontal
lines should be included. Use no footnotes; instead, include
all information in the legend. Make notations in the text mar-
gins to indicate the preferred location of tables in the printed
text.

Illustrations.— Original art work should be kept by the au-
thor(s) until final acceptance for publication. All art work must
be mounted on stiff board ready for reproduction. In line draw-
ings, pay particular attention to width of lines and size of
lettering when reductions are to be made by the printer. Mul-
tiple photos assembled on a single plate should be mounted
with only a minimum of space separating them. In the case of
multiple illustrations mounted together, each illustration must
be numbered sequentially rather than given an alphabetic se-
quence. Written on the back should be the name(s) of author(s),
an indication of top edge, and whether the art work should be
returned. The overall dimensions should be no more than 1 1
inches (28 cm) x 14 inches (36 cm). Larger drawings present
greater difficulty in shipping and greater risks of damage for
which the JOA assumes no responsibility. The author is en-
couraged to consider PMT (PhotoMechanical Transfer) repro-
ductions for large drawings and to consult an Editor for details
on appropriate reductions. Make notations in the text margins
to indicate the preferred position of illustrations in the printed
text.

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Adhere to the
following styles:

Figures 1-4.— A-us x-us, male from Timbuktu: 1, left leg;
2, right chelicera; 3, dorsal aspect of genitalia; 4, ventral aspect
of abdomen.

Figures 27-34. — Right chelicerae of species of A-us from
Timbuktu: 27, 29, 31, 33, dorsal views; 28, 30, 32, 34, prolater-
al views of moveable finger; 27, 28, A-us x-us, holotype male;
29, 30, A-us, w-us, holotype male; 31, 32, A-us z-us, holotype
male; 33, 34, A-us y-us, male. Scales = 1.0 mm.

Assemble manuscript for mailing.— Assemble the separate
sections or pages in the following sequence; title page, abstract,
text, figure legends, footnotes, running head, tables with leg-
ends, figures.

Page charges and reprints.— The current charge per journal
page will be assessed as follows: $50— nonmembers, $45 —
members with grant, institutional, or other support for page
costs; $25— regular members who do not have support for page
costs; $20— student members who do not author with full
members and who do not have support. Limited grants may
be available to help defray publication costs. Consult the As-
sociate Editor for details. Reprints are available only from the
printer at current costs.

RESEARCH NOTES

Instructions above pertaining to feature articles apply also
to research notes except that abstracts and most headings are
not used and the byline follows the Literature Cited section.

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 19 Feature Articles NUMBER 1

On South American Teminius (Araneae, Miturgidae), Norman I. Platnick
and Martin J. Ramirez 1

Systematic studies on the Nitidulus group of the genus Vaejovis, with de-
scriptions of seven new species (Scorpiones, Vaejovidae), W. David
Sissom 4

A new species of wolf spider, Schizocosa stridulans (Araneae, Lycosidae),

Gail E. Stratton 29

Ontogenic and seasonal changes in webs and websites of a desert widow

spider, Yael Lubin, Mandy Kotzman and Stephen Ellner 40

Dispersal and survivorship in a population of Geolycosa turricola (Araneae,

Lycosidae), Patricia R. Miller and Gary L. Miller 49

A revision of the genus Zora (Araneae, Zoridae) in North America, David
T. Corey and Daniel J. Mott 55

Observations on the behavior of the kleptoparasitic spider, Mysmenopsis
furtiva (Araneae, Mysmenidae), Frederick A. Coyle, Theresa C. O’Shields
and Daniel G. Perlmutter 62

Development and reproductive potential of Florinda coccinea (Araneae,

Linyphiidae), Marianne B. Willey and Peter H. Adler 67

Research Note

Centruroides hasethi Pocock, a junior synonym of Centruroides testaceus

(DeGeer) (Scorpiones, Buthidae), W. David Sissom 70

Book Review

The Biology of Scorpions (edited by Gary A. Polis), Philip H. Brownell ... 71

L

The Journal of
AP^CHNOLOGY

VOLUME 19

1991

NUMBER 2

THE JOURNAL OF ARACHNOLOGY

EDITOR: James E. Carico, Lynchburg College

ASSOCIATE EDITOR: Gary L. Miller, The University of Mississippi

EDITORIAL BOARD: J. E. Carrel, Univ. Missouri; J. A. Coddington, Na-
tional Mus. Natural Hist.; J. C. Cokendolpher, Lubbock, Texas; F. A. Coyle,
Western Carolina Univ.; C. D. Dondale, Agriculture Canada; W. G. Eberhard,
Univ. Costa Rica; M. E. Galiano, Mus. Argentine de Ciencias Naturales; M. H.
Greenstone, BCIRL, Columbia, Missouri; N. V. Horner, Midwestern State Univ.;
D. T. Jennings, Garland, Maine; V. F. Lee, California Acad. Sci.; H. W. Levi,
Harvard Univ.; E. A. Maury, Mus. Argentine de Ciencias Naturales; N. I. Plat-
nick, American Mus. Natural Hist.; G. A. Polls, Vanderbilt Univ.; S. E. Riechert,
Univ. Tennesse; A. L. Rypstra, Miami Univ., Ohio; M. H. Robinson, U.S. Na-
tional Zool. Park; W. A. Shear, Hampden-Sydney Coll.; G. W. Uetz, Univ. Cin-
cinnati; C. E. Valerio, Univ. Costa Rica.

The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the
study of Arachnida, is published three times each year by The American Arach-
nological Society. Memberships (yearly): Membership is open to all those inter-
ested in Arachnida. Subscriptions to The Journal of Arachnology and American
Arachnology (the newsletter), and annual meeting notices, are included with mem-
bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or
$90 (all other countries). Inquiries should be directed to the Membership Secretary
(see below). Back Issues: Susan E. Riechert, Department of Zoology, Univ. of
Tennessee, Knoxville, TN 37916 USA. Undelivered Issues: Allen Press, Inc.,
1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: George W. Uetz (1989-1991), Department ofBiological Sciences,
University of Cincinnati, Ohio 45221, USA.

PRESIDENT-ELECT: Allen R. Brady (1989-1991), Biology Department, Hope
College, Holland, Michigan 49423 USA.

MEMBERSHIP SECRETAR Y: Norman I. Platnick (appointed), American Mu-
seum of Natural History, Central Park West at 79th St., New York, New York
10024 USA.

TREASURER: Gail E. Stratton (1989-1991), Department of Biology, Albion
College, Albion, Michigan 49224 USA.

SECRETARY: James W. Berry (1989-1991), Department ofBiological Sciences,
Butler University, Indianapolis, Indiana 46208 USA.

ARCHIVIST: Vincent D. Roth, Box 136, Portal, Arizona 85632 USA.

DIRECTORS: Petra Sierwald (1989-1991), William A. Shear (1989-1991),
Matthew H. Greenstone (1990-1992).

HONORARY MEMBERS: C. D. Dondale, W. J. Gertsch, H. Homann, H. W.
Levi, A. F. Millidge, M. Vachon, T. Yaginuma.

Cover illustration: Female Phidippus mystaceus, (Araneae, Salticidae), Eastern USA, by G. B. Edwards.

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.

1991. The Journal of Arachnology 19:73-79

SPATIAL DISTRIBUTION OF LYCOSA TARENTULA
FASCIIVENTRIS (ARANEAE, LYCOSIDAE) IN A
POPULATION FROM CENTRAL SPAIN

Carmen Fernandez-Montraveta,* Rafael Lahoz-Beltra** and Joaquin Ortega*: *Depar-
tamento Psicologia Biologica y de la Salud, Universidad Autonoma, Cantoblanco,
28049-Madrid, Spain; **Departamento Matematica Aplicada, Universidad Complu-
tense, Ciudad Universitaria, 28040-Madrid, Spain

Abstract. The burrow spatial distribution pattern in a population of Lycosa tarentula
fasciiventris from central Spain was studied. The developmental stage of the individual
occupying the burrow, as well as the burrow spatial coordinates were measured during the
spring and summer. A three-dimensional distribution pattern was obtained and Morisita,
mean-crowding and variance-mean coefficient indices of burrow density were calculated.

The burrow distribution pattern changed throughout the study period. Subadult burrow
location shows a tendency toward instability whereas location stability of adult individuals
is greater. In both cases there is a tendency towards clumping, which is lesser in the case of
adult animals: if mean distances among burrows are compared between clumps, a tendency
towards regularity results in the latter case. The observed distribution pattern might be a
result of interspecific competition leading to a territorial system, with adult females con-
stituting the structural support of the population.

Resumen. Memos estudiado el patron de distribucion espacial de una poblacion de
Lycosa tarentula fasciiventris del centro de Espana. Durante la primavera y el verano, se
midieron las coordenadas espaciales de los nidos, asi como la fase de desarrollo del individuo
que lo ocupaba. A partir de los datos, se ha reconstruido el patron tridimensional de
distribucion, y se han calculado los indices de Morisita y mean-crowding, asi como el cociente
varianza-media de la densidad de los nidos. A lo largo del periodo de estudio se observa
una modificacion en el patron de distribucion de los nidos. Los nidos de los individuos
subadultos muestran una tendencia a la inestabilidad, mientras que los ocupados por in-
dividuos adultos tienen una localizacion mas estable. En ambos casos, se observa una
tendencia a la agregacion, que es menos marcada para los individuos adultos: si se comparan
las distancias medias entre los nidos dentro de cada agregado, aparece una tendencia a la
regularidad en el ultimo caso. El patron de distribucion espacial podria ser el resultado de
la competicion intraespecifica, que determinaria un sistema de tipo territorial, siendo las
hembras adultas el soporte estructural de la poblacion.

In the non-social species, competition usually
takes place in the form of struggling for a mate
or for food (Burgess & Uetz 1982) and will be
more severe between individuals of the same sex
and age because they have similar requirements
(McBride 1970; Dunbar 1986). This can lead to
interindividual spacing that reflects resource dis-
tribution (McBride 1970). When resources are
limited, competition is believed to be the main
determinant for the population spatial structure
(Riechert et al. 1973). Intraspecific aggression is
a way of competition (Wilson 1 975; Huntingford
& Turner 1987), and species and/or individuals
showing an active defense behavior may also be

the ones showing the greatest regularity in their
spacing patterns (Burgess & Uetz 1 982).

Aggressive and defensive behavior has been
reported in several spider families (Rovner 1 968;
Dijkstra 1969; Buskirk 1975; Riechert 1978;
Jackson 1980; Jacques & Dill 1980; Goist 1982;
Christenson 1 984; Nossek & Rovner 1984; Hodge
1986, 1987; Wells 1988). Usually, studies of ag-
gression as a means of competition have focused
on male-male interactions in reproductive con-
texts (Rovner 1968; Dijkstra 1969; Jackson 1980;
Goist 1982; Austad 1983). The study of com-
petition for food or spatial resources not related
to gaining access to a female received lesser at-

73

74

THE JOURNAL OF ARACHNOLOGY

tention (Buskirk 1975; Riechert 1978, 1980;
Hodge 1987). When competition for food re-
sources was studied, spacing was shown to be
actively maintained by means of agonistic inter-
actions (Buskirk 1975; Hodge 1986) and to fit
models ascribed to territorial systems (Buskirk
1975; Riechert 1978, 1982).

Most previous work on spatial distribution of
spiders has been carried out on web-building spe-
cies. The study of individual spatial distribution
of species of Lycosidae, non-web-building ones,
has not been too extensively put forward. Kuen-
zler (1958) reported that the spatial distribution
of three species of the genus Lycosa showed a
random pattern in a uniform environment. How-
ever, the active maintenance of spacing is a prev-
alent pattern (McBride 1970), and spiders seem
more often to be territorial animals (Riechert
1980; Maynard Smith & Riechert 1984).

Our laboratory studies on the agonistic be-
havior of adult female Lycosa tarentula fascii-
ventris, suggest that agonistic interactions are a
way of competing for burrows. Agonistic inter-
actions usually occur inside the burrows, and the
result of these interactions is the expulsion of
one of the contenders (Femandez-Montraveta &
Ortega 1990). Spiders of this species build tu-
bular burrows in the ground, with an opening to
the exterior which is sometimes surrounded by
a cylindrical structure (Ortega 1 986). Individuals
spend most of their time in their burrows, and
prey capture patterns seem to be related to them
(Ortega 1985).

If the agonistic behavior of adult females of
Lycosa t.fasciiventris is really a way of competing
for burrow sites, it might be expected that in-
dividual spacing patterns fit a non-random dis-
tribution, probably a regular one (Burgess & Uetz
1982). That distribution would primarily in-
volve individuals of the same age and sex. In this
paper, we measure the burrow spatial distribu-
tion pattern in a population of Lycosa t. fascii-
ventris in central Spain in order to determine
whether it fits the non-random distribution pat-
tern predicted (Riechert 1 980, 1 982).

METHODS

We studied a population of Lycosa t.fasciiven-
tris located in “El Goloso” (“Canal de Isabel 11”)
near the Universidad Autonoma de Madrid. The
study area was a rectangle 200 m long and 40 m
wide, with its four boundaries artificially limited
by a road and a three-sided metallic fence. This
area is characterized by having a sandy substrate

with a poor water table and herbaceous vegeta-
tion. The site was visited daily between 0900 and
1400 h during the period from April to August
1984.

The study area was marked in a grid, covering
the 200 X 40 m^ rectangle. Along the 200 m axis,
1 -m wide, parallel corridors were marked. These
corridors were exhaustively covered in the suc-
cessive visits, and the cartesian coordinates of
the burrows occupied by Lycosa t. fasciiventris
were recorded. Body measurements (prosoma
length and width, as well as length of the first
and the fourth leg pairs) and the developmental
stage of each burrow occupant were recorded.
Because it is difficult to accurately determine the
developmental stage of the spiders, they were
classified into three age categories: (1) subadult
individuals bom in 1983 (S-83), (2) subadult in-
dividuals bom in 1982 (S-82) and reaching their
adult instar in summer 1984, and (3) adult in-
dividuals. Immature individuals can be differ-
entiated with regard to their year of birth, since
they have markedly different sizes. With regard
to sex, only adults and immature individuals at
their penultimate instar can be differentiated.
Animals were marked by means of a label at-
tached to their prosoma.

A total of 1 3 1 burrows, only considering those
occupied in at least two successive visits, were
included in the total analysis. Since most of the
molts were not recovered, we were unable to
determine whether an unmarked occupant in a
burrow was the same individual that was found
previously. The chance that it was, according to
its measurements, was used as a criterion.

Because the development rate of animals dur-
ing the period studied is high, data were analyzed
at two different times. First, data from May were
taken into account. In this analysis, all the data
were included. Secondly, data corresponding to
July were analyzed. By that time, most of the
initially marked S-83 individuals had disap-
peared, S-82 individuals had become adults and
adult males usually did not occupy their burrows.
Thus, only data from adult females were includ-
ed in this second analysis.

The two sets of data were analyzed by intro-
ducing the coordinates of the burrows into a data
matrix. With this matrix, a three-dimensional
surface plot was obtained with the program
Golden Graphic System (Golden Software, Inc.)
in order to get a graphical representation of bur-
row density.

Data corresponding to July were subjected to

FERNANDEZ-MONTRAVETA ET AL.= SPATIAL DISTRIBUTION OF LYCOS A

75

Figure 1.— Study area plane with the location of the 38 adult female spider burrows which were included in
the second analysis (July data). The six marked areas represent the rectangles of 250 m^ superimposed on the
aggregation patches.

additional analyses. These burrows were plotted
as points on graph paper and counted using a
sample lattice of 336 cells of 25 mm^ The fre-
quencies of burrows/cell were then calculated.
The measures of aggregation we used were all
based on the above frequencies. These measures
were mean-crowding (Lloyd 1967), Morisita’s
index (Morisita 1959) and the variance-mean ra-
tio (Pielou 1977). The first index provided in-
formation about the mean number of burrows
per burrow co-occupying a cell, while the other
two visualized the spatial pattern of the burrows.
Significance of variance-mean coefficient was
measured by means of

t = [(sVx) - l]/(2/n - l)‘^

P < 0.05 (Kershaw 1973).

After the coordinates of the burrows were
graphically represented, several aggregation
patches clearly different from one another be-
came evident. Then a rectangle of 250 m^ was
superimposed on each patch, and only the bur-
rows which fell into these rectangles were chosen
for study (Fig. 1). The mean and the variance
values of the distance from each burrow to its
nearest neighbor were measured for each rect-
angle (Clark & Evans 1954). The mean distances
were compared between rectangles applying a test
of equality of means.

RESULTS

Table 1 shows the number of initially located
burrows that continue to be occupied in the fol-
lowing visits with regard to individual age class.
Data are from our first study phase.

The spatial distribution pattern of individuals
in May is shown in Fig. 2. Figure 3 shows the
spatial distribution pattern of individuals in July,
including only adult females. When the two dis-
tributions (Figs. 2, 3) are compared, a change is

noticed in which the area of stronger aggregation
in May disperses into several patches in which
burrow density is high in July.

For the whole data corresponding to July, the
variance-mean ratio (1.13) does not differ from
random {t = 0.590). The Morisita’s index (2.15)
reveals that burrows are distributed with a cer-
tain tendency toward patching. Mean-crowding
value (0.25), on the other hand, indicates the
mean number of co-occupants to be relatively
low. The mean distance to the nearest neighbor
as calculated in each area (Table 2) shows some
degree of spacing.

By comparing the mean distances to the near-
est neighbor between the different patches, we
found no significant differences among five of
them. Only two of the patches differ from one
another with regard to this parameter (Table 3).

DISCUSSION

A trend to burrow aggregation around certain
areas is shown in both spring and summer but

Table 1.— Burrow location stability with regard to the
developmental stage (age class) of the spider occupying
the burrow. Individual location was more unstable dur-
ing the early developmental stages than during the lat-
ter, and the entire number of animals was decreasing.
There was a reduction of 60% in the population size.
Disappearance of S-83 individuals accounts for 84%
of this reduction. (“ 34 of these individuals were found
to reach their adult stage during the study period.)

Age class

Numbers of burrows

Stable

Unstable

Total

Subadult

S-83

9 (12%)

66 (88%)

75

S-82

37» (74%)

13(26%)

50

Adult

6 (100%)

0

6

Totals

52 (40%)

79 (60%)

131

76

THE JOURNAL OF ARACHNOLOGY

Figure 2.— Three-dimensional representation of burrow density distribution during the early study period
(May). The axes labeled “x” and “y” represent the spatial coordinates of the area plane. The axis labeled “z”
represents burrow density.

is stronger in spring. Spiders remain compara-
tively grouped during their early developmental
stages (Fig. 2). A great instability of spatial lo-
cation is also shown during these stages. Then,
aggregation levels decrease, resulting in interin-
dividual spacing together with a greater stability.
These factors might result from increasing ag-
gressive trends (Riechert 1978, 1980) and also
from searching for suitable burrow locations,
leading to the less clumped distribution pattern
of adult individuals. At this point, interindivi-

dua! distance, as well as location stability, are
the greatest.

The spatial distribution of adult females might
be considered as non-random. The existence of
areas with differing density might reflect the het-
erogeneity of the study area conditions. In the
small areas, the trend is for the females to be
more regularly spaced out, suggesting spacing is
actively maintained. Since adult female inter-
actions occur in the laboratory resulting in one
female eventually running out of the burrow

FERNANDEZ-MONTRAVETA ET AL.- spatial distribution of LYCOS a

77

Figure 3. --Three dimensional representation of burrow density distribution during the latter study period
(July). Axes labeled “x” and “y” represent the spatial coordinates of the area plane. Axis labeled “z” represents
burrow density.

(Femandez-Montraveta & Ortega 1990), we
might consider the distribution pattern to be a
territorial system (McBride 1970; Riechert 1980;
Huntingford & Turner 1987; Hammerstein &
Riechert 1988). It might be an “overlapping ter-
ritorial system” in which individuals defend an
area around their burrow against intrusion, de-
creasing the attack intensity gradually as distance
to the burrow increases (McBride 1970).

Data supporting the idea that spatial distri-
bution of three Nearctic species of Lycosa was
random (Kuenzler 1958) were based on individ-
ual location during their activity periods, not on
the burrow location (two of the species not being
burrowers). The burrow must be a significant
resource as a shelter (Ortega 1986) as well as
protection against extreme temperatures (Hum-
phreys 1987; Riechert 1980). Moreover, it is

78

THE JOURNAL OF ARACHNOLOGY

Table 2.— Mean values (meters) and SE of nearest
neighbor distance in the studied subareas.

Areas

Mean

SE

1

1.20

0.34

2

0.56

0.13

3

0.50

0.04

4

0.66

0.18

5

0.57

0.07

6

0.45

0.19

probably associated with the predatory strategy
of this non-web-building species (Burgess & Uetz
1 982; Christenson 1 984) and is not only a shelter
to which animals return when disturbed (Kuen-
zler 1958).

The relatively great mean distance between the
burrows may be a consequence of the spacing
which results from agonistic interactions (Bus-
kirk 1975; Hodge 1986; McBride 1970) and might
reflect the minimal distance needed for intra-
specific cannibalism to be reduced (Burger 1981).
Territory size has been related to energy require-
ments and prey availability (Riechert 1978) and
might be stabilized at its greatest value, which
would correspond to extremely severe situations
(Riechert 1980). Interindividual distances ex-
ceeding the web size have also been found in
other species in which the web is considered to
be the hunting territory (Buskirk 1975). Conse-
quences of territorial behavior in Lycosa t. fas-
ciiventris then would be that the individual’s en-
ergy resources are probably assured. If territory
size were fixed, it might set limits to the popu-
lation size (Riechert 1980).

Because of their territorial distribution, adult
females might through time be the structural sup-
port of the population. The size of the population
studied shows sudden alterations that are prob-
ably due to dispersion and mortality during the
individuals’ early instars and, in the case of males,
mortality during adult instar. In a population
where females are competing for areas, whose
size is based on available shelter and food supply,
and where males are competing for females, the
female spacing pattern may influence the distri-
bution and size of the population to a greater
extent than does that of other individuals in the
population (Huntingford & Turner 1987).

ACKNOWLEDGMENTS

This study was carried out thanks to the per-
mission granted by “Canal de Isabel IF’ (Madrid)

Table 3.— P- value for comparisons between the
nearest neighbor distance in the studied subareas.

Areas

Areas 1

2

3

4

5

6

1 -

0.09

0.003

0.250

0.097

0.1

2 -

—

0.650

0.660

0.932

0.65

3 -

—

—

0.207

0.436

0.65

4 -

—

—

—

0.630

0.47

5

-

-

-

-

0.57

6

— ■

—

—

—

—

for visiting “El Goloso” deposit. We wish to ac-
knowledge the contribution in matters of style
and rendering into English made by Mark Youn-
german and two anonymous reviewers for their
valuable comments on the previous version of
this manuscript.

REFERENCES CITED

Austad, S. N. 1983. A game theoretical interpretation
of male combat in the bowl and doily spider {Fron-
tinella pyramitela). Anim. Behav., 31:59-73.
Burger, J. 1981. Super-territories, a comment. Amer.
Nat., 1 18:578-580.

Burgess, J. W. & C. W. Uetz. 1982. Social spacing
strategies in spiders. Pp. 317-322, In Spider Com-
munication. Mechanisms and Ecological Signifi-
cance (P. N. Witt & J. S. Rovner, eds.). Princeton
Univ. Press, Princeton.

Buskirk, R. E. 1975. Aggressive display and orb de-
fence in a colonial spider, Metabus gravidus. Anim.
Behav., 23:560-567.

Christenson, T. E. 1984. Behaviour of colonial and
solitary spiders of the Theriidae species Anelosimus
eximius. Anim. Behav., 32:725-734.

Clark, P. J. & F. C. Evans. 1954. Distance to nearest
neighbor as a measure of spatial relationships in
populations. Ecology, 35:445-453.

Dijkstra, H. 1969. Comparative research of the court-
ship behavior in the genus Pardosa. Ill: Agonistic
behavior in Pardosa amentata. Bull. Mus. Nat. Hist.
Nat., 2 ser., 41:91-97.

Dunbar, R. I. M. 1986. Aggression. Pp. 72-79, In
The Collins Encyclopedia of Animal Behaviour (P.
J. B. Slater, ed.). Collins, Oxford.
Femandez-Montraveta, C. & J. Ortega. 1990. El
comportamiento agonistico de hembras adultas de
Lycosa tarentula fasciiventris (Araneae, Lycosidae).
J. Arachnol., 18:49-58.

Goist, K. C. 1 982. Male-male competition in the orb
weaving spider Nephila clavipes. Ph.D. Diss., Tu-
lane Univ. Press, New Orleans.

Hammerstein, P. & S. E. Riechert. 1 988. Payoffs and
strategies in spider territorial contests: ESS-analyses
of two ecotypes. Evol. Ecol., 2:1 15-138.

fernAndez-montraveta et al.-spatial distribution of lycosa

79

Hodge, M. 1986. The relationships between agonistic
behavior and spatial organization in desert and trop-
ical Metepeira spp. from Mexico. Amer. Arachnol.,
34:5.

Hodge, M. 1987. Agonistic interactions between fe-
male bowl and doily spiders (Araneae, Linyphiidae):
Owner biased outcomes. J. Arachnol., 15:241-247.

Humphreys, W. F. 1987. The thermal biology of the
wolf spider Lycosa tarentula (Araneae, Lycosidae)
in Northern Greece. Bull. British Arachnol. Soc., 7,
4:117-122.

Huntingford, F. & A. Turner. 1987. Animal Conflict.
Chapman & Hall, London.

Jackson, R. R. 1980. The mating strategy of Phidip-
pus johnsoni. III. Intermale aggression and a cost-
benefit analysis. J. Arachnol., 8:241-250.

Jacques, A. M. & L. M. Dill. 1980. Zebra spiders use
uncorrelated asymmetries to settle contests. Amer.
Nat., 116:899-901.

Kershaw, K. A. 1973. Quantitative and Dynamic
Plant Ecology, 2nd ed. Arnold, London.

Kuenzler, E. J. 1 958. Niche relations of three species
of lycosid spiders. Ecology, 39, 3:494-500.

Lloyd, M. 1967. Mean-crowding. J. Anim. Ecol., 36:
1-30.

Maynard Smith, J. & S. E. Riechert. 1984. A con-
flicting-tendency model of spider agonistic behav-
iour: Hybrid-pure population line comparisons.
Anim. Behav., 32:564-578.

McBride, G. 1970. Theories of animal spacing: The
role of flight, fight and social distance. Pp. 53-68,
In The Social Use of Space in Animals and Man
(A. H. Esser, ed.). Plenum Press, New York.

Morisita, M. 1959. Measuring the dispersion of in-
dividuals and analysis of the distributional patterns.
Mem. Fac. Sci. Kyushu Univ., ser 2:215-235.

Nossek, M. & J. S. Rovner. 1984. Agonistic behavior
in female wolf-spiders (Araneae, Lycosidae). J.
Arachnol., 11:407-422.

Ortega, J. 1985. Qualitative and quantitative analysis
of predatory behaviour of Lycosa fasciiventris Du-
four (Araneae, Lycosidae). Biol. Behav., 10:55-65.

Ortega, J. 1986. Posibles relaciones entre el habitat
de Lycosa fasciiventris Dufour (Araneae, Lycosidae)
y su comportamiento. Bol. R. Soc. Espan. Hist. Nat.,
82, 1-4:121-129.

Pielou, E. C. 1977. Spatial patterns and their repre-
sentation by discrete distributions. Pp. 1 13-165, In
Mathematical Ecology. Wiley.

Riechert, S. E. 1978. Energy-based territoriality in
populations of the desert spider Agelenopsis aperta
(Gertsch). Symp. Zool. Soc. London, 42:21 1-222.

Riechert, S. E. 1980. The consequences of being ter-
ritorial: Spiders, a case study. Amer. Nat., 117:871-
892.

Riechert, S. E. 1982. Spider interaction strategies:
Communication vs. coercion. Pp. 281-313, In Spi-
der Communication: Mechanisms and Ecological
Significance (P. N. Witt & J. S. Rovner, eds.).
Princeton Univ. Press, Princeton.

Riechert, S. E., W. G. Reeder & T. A. Allen. 1973.
Patterns of spider distribution {Agelenopsis aperta
(Gertsch)) in desert grassland and recent lava bed
habitats, South Central New Mexico. J. Anim. Ecol.,
42:19-35.

Rovner, J. S. 1968. Territoriality in the sheet-web
spider Linyphia triangularis. Z. Tierpsychol., 25:
232-242.

Wells, M. S. 1988. Effects of body size and resource
value on fighting behaviour in a jumping spider.
Anim. Behav., 36, 2:321-326.

Wilson, E. O. 1975. Sociobiology. Harvard Univ.
Press, Cambridge.

Manuscript received August 1989, revised June 1990.

1991. The Journal of Arachnology 19:80-84

OWNER-BIASED AGONISTIC BEHAVIOR IN
FEMALE LYCOSA TARENTULA FASCIIVENTRIS
(ARANEAE, LYCOSIDAE)

Carmen Fernandez-Montraveta and Joaquin Ortega: Departamento de Psicologia Biolo-
gica y de la Salud, Universidad Autonoma, Cantoblanco, 28049-Madrid, Spain

Abstract. Matnces of the frequency of patterns of agonistic behavior of adult female
Lycosa tarentula fasciiventris throughout intra-individual sequences were analyzed by means
of an analysis of variance. Behavior differences were analyzed with regard to two factors:
female size and previous occupation of the burrow. Results show that females use different
tactics of agonistic behavior, depending on their previous occupation of the burrow, re-
gardless of the relative size differences.

Resumen. Mediante un analisis de varianza aplicado a las matrices de frecuencia de las
pautas en las secuencias intraindividuales de comportamiento agonistico, hemos analizado
las diferencias en el comportamiento de las hembras adultas de Lycosa tarentula fasciiventris
con respecto a dos variables: el tamano y la ocupacion previa del nido. Nuestros resultados
indican que las hembras utilizan diferentes tacticas de comportamiento agonistico segun su
ocupacion previa del nido, e independientemente de su tamano relativo.

When studying agonistic behavior, the lack of
intense aggression patterns has been explained
by means of a theoretical game model of maxi-
mizing the consequences of behavior on fitness,
given both the costs associated and the frequency
of using that tactic among members of the pop-
ulation (Maynard Smith & Price 1973; Hunt-
ingford & Turner 1987). In contexts in which
escalating a fight may be dangerous and infor-
mation about the opponent is easily assessed,
there may be different roles used to settle the
contest (Maynard Smith & Parker 1976). These
roles are determined by asymmetrical features
between individuals, which may or may not be
correlated with the individual winning ability or
the relative resource-holding potential (Maynard
Smith & Parker 1976; Hammerstein 1981). The
expenditure of energy an individual makes is as-
sumed to be adjusted to the value of winning the
contest and the probability of doing it (Maynard
Smith & Price 1973; Maynard Smith & Parker
1976).

Among spiders, several examples of biased ag-
onistic behavior have been reported, especially
among web-weaving species. Differences have
been found to be usually related to individual
size and previous residence (Buskirk 1975;
Riechert 1978; Hodge 1987), although size dif-
ference may also affect the result in this latter
case if greater than a critical value (Riechert 1978).

The agonistic behavior of adult females of the
lycosid species Lycosa tarentula fasciiventris Du-
four has been described from data obtained in
our laboratory. During agonistic interactions, fe-
males show stereotyped patterns of behavior, and
risk of bodily harm is relatively high. We think
animals exchange information throughout these
contests (Fernandez-Montraveta & Ortega 1 990)
and, since escalation risk is high, this informa-
tion can be expected to be accurate (Parker 1 974).
Animals will then be expected to accurately eval-
uate the contest and then use settlement strate-
gies based on asymmetrical features of previous
occupation of the burrow and size (Fernandez-
Montraveta & Ortega 1990). Since animals seem
to occup their burrows for a long period and there
is a relatively high investment of time and energy
in burrow construction, individual behavior could
be expected to differ with regard to residence,
resident females displaying a higher interaction
cost and persistence (Riechert 1 988).

In this paper, interindividual differences in the
frequency with which adult females of L. t. fas-
ciiventris show patterns of agonistic behavior are
measured with regard to the variables of size and
previous occupation of the burrow. We attempt
to verify if females of this species use different
behavioral tactics, depending on these interin-
dividual asymmetrical features.

80

FERNANDEZ-MONTRAVETA & ORTEGA-OWNER-BIASED BEHAVIOR IN LYCOSA

81

METHODS

In this study 40 adult female L. t. fasciiventris
were used. All the animals reached their adult
instar in the laboratory and were maintained in
isolation from the date of capture until obser-
vation. All the animals were from the same area,
near the Universidad Autonoma, 1 5 km north
of Madrid (Spain). Animals were weighed and
measured after undergoing their last molt.

Observations were made in 30 x 1 5 x 15 cm
terraria having a burrow constructed beside their
front walls. The interior of the burrow was visible
only during the observation time. Spiders were
observed randomly paired with regard to their
size, the intruder being moved to the observation
terrarium just before the observation began. The
resident female occupied the burrow inside the
terrarium for at least 7 days before the obser-
vation date. Only females usually occupying the
burrow, where they were fed, were taken into
account. We made a total of 73 different pair
observations, obtaining 34 interaction sequenc-
es. No animal was observed more than once a
day.

Agonistic patterns of behavior were described
(Femandez-Montraveta & Ortega 1990) from
records of all the movements and activities of
the animals during a minimum period of 30 min.
Considering behavioral patterns as variables and
individuals as cases, matrices were constructed
in which the absolute frequencies (the number
of times an individual exhibited a given behav-
ioral pattern during the sequence) were repre-
sented. The absolute frequency matrices were
transformed into relative frequency ones, rep-
resenting the proportion between the display of
each behavioral pattern and the total number of
elements of the sequence. In order to carry out
the analysis, an arc-sine transformation was ap-
plied to these resulting matrices.

Matrices were analyzed by means of a bifac-
torial analysis of variance, with the variables
“previous occupation of the burrow” and “size”
considered as the grouping factors. Two discrete
levels were considered for each of these: “resi-
dent/intruder” for the former and “larger/small-
er” for the latter. The analysis was applied by
means of the program 7D, belonging to the
BMDP87 package.

RESULTS

We analyzed 68 behavioral sequences from
both the resident females and the intruders. Ta-

xable \.—F and R- values for differences in the mean
frequencies of observed behavioral patterns relative to

female size
P < 0.05).

and previous

residence (*

indicates

Pattern

Factor

F

P

Motionless

Residence

6.81

0.0114*

Size

0.10

0.7531

Interaction

0.08

0.7738

Approach

Residence

4.83

0.0317*

Size

0.12

0.7358

Interaction

1.35

0.2509

Go away

Residence

22.87

0.0000*

Size

0.29

0.5936

Interaction

O.OI

0.9083

Contact

Residence

23.73

0.0000*

Size

0.00

0.9882

Interaction

2.88

0.0965

Pounce

Residence

36.39

0.0000*

Size

1.02

0.3166

Interaction

0.66

0.4221

Palpal drum

Residence

3.64

0.0641

Size

0.58

0.4501

Interaction

0.42

0.5199

Foreleg

Residence

49.33

0.0000*

extension

Size

0.01

0.9392

Interaction

0.56

0.4567

Capture

Residence

18.55

0.0001*

Size

0.00

0.9440

Interaction

3.13

0.0832

Tangle

Residence

16.00

0.0002*

Size

0.07

0.7934

Interaction

1.15

0.2892

ble 1 shows the results from the application of
the analysis. There is no statistical interaction
between the factors of “previous occupation of
the burrow” and “size” for any of the analyzed
behavior patterns (Table 1). There are significant
differences in the mean frequency of all the be-
havioral patterns with regard to the factor of
“previous residence,” regardless of size. F- values
are less than 0.05 in all cases, except for the
“palpal drumming” pattern.

Figures 1 and 2 show the mean frequency of
behavioral patterns for the variables of “previ-
ous occupation of the burrow” and “size”. In-
truder females more frequently use the patterns
of “Motionless”, “Approach” and “Go away”
(Fig. 1 ). On the other hand, resident females more
frequently use the patterns of “Foreleg Contact”,
“Pounce”, “Foreleg Extension Chelicerae
Spreading”, (“Threat”), “Tangle” and “Cap-

82

THE JOURNAL OF ARACHNOLOGY

MOTIONLESS APPROACH GO AWAY

PATTERNS

Figure 1. — Mean frequency of the behavior patterns more frequently shown by intruder females— “Motion-
less”, “Approach” and “Go away”— with regard to relative size (S = smaller, L = larger) and previous occupation
of the burrow (R = resident, I = intruder).

PATTERNS

Figure 2. — Mean frequency of using the behavior patterns more frequently shown by resident females—
“Contact”, “Pounce”, “Palpal drumming”, “Foreleg extension and chelicerae spreading”, (“Threat”), “Capture”
and “Tangle”— with regard to relative size (S = smaller, L = larger) and previous occupation of the burrow (R
= resident, I = intruder).

FERNANDEZ-MONTRAVETA & ORTEGA -OWNER-BIASED BEHAVIOR IN LYCOS A

83

ture” (Fig. 2). Mean frequency of “Palpal Drum-
ming” is also greater among this group, but is
not significantly different (Table 1).

DISCUSSION

From our results, a high degree of intraspecific
variability in the agonistic behavior of adult fe-
males of that spider species is obvious. When
studying spider behavior, interspecific differ-
ences have been emphasized (Hollander et al.
1973; Stratton & Uetz 1983; Suwa 1984) and a
lesser degree of attention has been paid to in-
traspecific variability (however, see Jackson 1 986;
Kronestedt 1986; Riechert 1988). Our results in-
dicate that differences shown by females of L. t.
fasciiventris in their agonistic behavior may be
related to their previous occupation of the bur-
row but, in no case, to their relative size.

If we consider that behavior patterns are or-
dered according to their intensity-defined by
the occurrence of contact (Glass & Huntingford
1988)— then behavior intensity seems to be
greater for resident females than for the intruders
in L. t. fasciiventris (Fig. 2). These latter tend to
use to a greater extent those behavioral patterns
of non-stereotyped approaching and retreating.
Therefore, the behavior shown by resident fe-
males seems to be associated with a higher cost
(Riechert 1988). These results can be interpreted
as if adult females of this species were using the
conditional strategy of “attack if resident and
retreat if intruder” (Maynard Smith 1 974) to set-
tle the contests, regardless of their size difference.

Since size may be thought of as a factor influ-
encing individual winning ability, and it is be-
lieved that animals exchange information about
that factor (Parker 1974), it might be expected
that spiders also adjust their behavior to their
size differences. In our study, previous occupa-
tion of the burrow was not associated with greater
physical ability since maintenance conditions
were the same for all individuals, both those used
as residents and those used as intruders. We think
that the female agonistic tactic is based essen-
tially on their previous residence, and that the
effect of size difference is related to the interac-
tion duration. This interpretation fits the inverse
relationship we found between interaction du-
ration and size differences in these kinds of en-
counters (Femandez-Montraveta & Ortega 1 990).

We think, therefore, that our results fit the
expectation that animals use an evaluator strat-
egy to settle their agonistic interactions and ex-

change accurate information about relative abil-
ity to hold the resource. The factor of previously
occupying the burrow apparently affects the as-
sumption of behavior patterns with higher attack
levels by resident females. This behavioral strat-
egy should allow the contests to be readily settled
for the resident female if size difference is not
too small. Otherwise contest duration might be
longer.

ACKNOWLEDGMENTS

We thank S. E. Riechert and G. W. Uetz for
reviewing this manuscript and especially for cor-
recting style and usage. We also thank M. A. Ruiz
for preparing the figures.

REFERENCES CITED

Buskirk, R. E. 1975. Aggressive display and orb-de-
fence in a colonial spider, Metabus gravidus. Anim.
Behav., 23:560-567.

Femandez-Montraveta, C. & J. Ortega. 1990. El
comportamiento agonistico de hembras adultas de
Lycosa tarentula fasciiventris (Araneae, Lycosidae).
J. Arachnol., 18:49-58.

Glass, C. W. & F. A. Huntingford. 1988. Initiation
and resolution of fights between swimming crabs.
Ethology, 77(3):237-249.

Hammerstein, P. 1981. The role of asymmetries in
animal contests. Anim. Behav., 29:193-205.
Hodge, M. A. 1987. Agonistic interactions between
female bowl and doily spiders (Araneae, Linyphi-
idae): owner biased outcomes. J. Arachnol., 1 5:241-
247.

Hollander, J. den, H. Dijkstra, H. Alleman & L. Vljim.
1973. Courtship behaviour as species barrier in the
Pardosa pullata group (Araneae, Lycosidae).
Tijdschrift. Ent., 116(l):l-22.

Huntingford, F. & A. Turner. 1987. Animal Conflict.

Champan & Hall, London. 448 pp.

Jackson, R. R. 1986. Communal jumping spiders
from Kenya: Intraspecific interactions, interspecific
nest complexes and cohabitation with web-build-
ing spiders. New Zealand J. Zool., 13:13-26.
Kronestedt, T. 1986. Ethospecies. Behavioral pat-
terns as an interspecific barrier. Actas X Congr. In-
tern. AracnoL, Jaca/Espana, 11:41-45.

Maynard Smith, J. 1974. The theory of games and
the evolution of animal conflicts. J. Theor. Biol.,
47:209-221.

Maynard Smith, J. & G. A. Parker. 1976. The logic
of asymmetric contests. Anim. Behav., 24:159-175.
Maynard Smith, J. & G. R. Price. 1973. The logic of
animal conflicts. Nature, 246:15-18.

Parker, G. A. 1 974. Assessment strategy and the evo-
lution of fighting behaviour. J. Theor. Biol., 47:223-
243.

84

THE JOURNAL OF ARACHNOLOGY

Riechert, S. E. 1978. Games spiders play: I. Behav-
ioral variability in territorial disputes. Behav. Ecol.
Sociobiol., 3:135-162.

Riechert, S. E. 1988. The energetic costs of fighting.

Amer. ZooL, 28(3):877-884.

Stratton, G. E. & G. W. Uetz. 1983. Communication
via substratum-coupled stridulation and reproduc-

tive isolation in wolf spiders (Araneae, Lycosidae).
Anim. Behav., 31:164—172.

Suwa, M. 1984. Courtship behaviour of three new
forms in the wolf spider Pardosa laura complex. J.
Ethol., 2:99-107.

Manuscript received October 1989, revised June 1990.

1991. The Journal of Arachnology 19:85-87

A MITOCHONDRIAL DNA RESTRICTION ENZYME
CLEAVAGE MAP FOR THE SCORPION
HADRURUS ARIZONENSIS (lURIDAE)

Deborah Roan Smith*: Insect Division and Laboratory for Molecular Systematics, Mu-
seum of Zoology, University of Michigan, Ann Arbor, Michigan 48109 USA

Wesley M. Brown: Laboratory for Molecular Systematics, Museum of Zoology and
Department of Biological Sciences, University of Michigan, Ann Arbor, Michigan
48109 USA

Abstract. Mitochondrial DNA (mtDNA) was prepared from a single individual of the
scorpion Hadrurus arizonensis Ewing. The total size of the mitochondrial genome was
estimated to be 13 850 to 14 000 base pairs. The mtDNA was surveyed for cleavage sites
using 1 7 six-base restriction enzymes and three four-base enzymes. The technique of double
digests was used to construct a map of the cleavage sites generated in this mtDNA by nine
six-base restriction enzymes.

We present a restriction enzyme cleavage map
and information on the size and base composi-
tion of the mitochondrial genome of the scorpion
Hadrurus arizonensis Ewing (luridae). This is the
first published restriction site map for an arach-
nid mitochondrial genome. This map will be use-
ful in systematic studies of Hadrurus and related
scorpions, in studies which use mtDNA markers
in the study of H. arizonensis population biology,
and as a guide for cloning and sequencing the H.
arizonensis mitochondrial genome. Excellent
discussions of animal mitochondrial DNA
(mtDNA) and its use in systematics, biogeog-
raphy and population biology can be found in
Avise et al. (1987), Brown (1985) and Moritz et
al. (1987).

METHODS AND MATERIALS

The scorpions used in this study were collected
from Cochise Co., Arizona in June, 1989 by K.
and D. Aiken, shipped live to the University of
Michigan, and stored at -80° C. MtDNA was
prepared from muscle tissue (tail, pedipalps) of
one adult using the methods described in Smith
and Brown ( 1 990). A second individual has been
kept as a voucher specimen. The mtDNA was
analyzed by digestion with restriction enzymes.

' Current address: Department of Entomology, Snow
Hall, University of Kansas, Lawrence, Kansas 66045

USA.

Aliquots of the mtDNA were digested with the
six-base and four-base restriction enzymes listed
in Table 1, using bulfer and temperature con-
ditions recommended by the manufacturers (Be-
thesda Research Laboratories, Boehringer
Mannheim Biochemicals, International Biotech-
nologies, and New England Biolabs). The re-
sulting fragments were radioactively end-labeled
with ^^P-deoxynucleotides and separated by elec-
trophoresis on 1% agarose and 4% polyacryl-
amide gels (Brown 1 980; Wright et al. 1 983) and
visualized by autoradiography. The sizes of the
DNA fragments were estimated by comparison
with size standards (Hindlll and Aval/Bglll di-
gests of wild-type lambda phage DNA and Haelll
digests of phage phi-X 1 74 DNA) run on the same
gels. The relative positions of restriction enzyme
cleavage sites were mapped by means of double
digests (Brown & Vinograd 1974).

RESULTS AND DISCUSSION

The tissues from the tail and one pedipalp of
one H. arizonensis yielded approximately 1.5 ng
of mtDNA. The mitochondrial genome of H.
arizonensis is small, approximately 13 850 to
14 000 base pairs (13.85 to 14.0 kilobase pairs
or kb). Table 1 shows the number of cleavages
generated by each restriction enzyme in H. ari-
zonensis mtDNA. Figure 1 shows the relative
positions of the cleavage sites generated by nine
of the enzymes.

85

86

THE JOURNAL OF ARACHNOLOGY

Table 1.— Six-base and four-base restriction endo-
nucleases used in this survey, and the number of cleav-
ages generated by each enzyme in the mitochondrial
DNA of Hadrurus arizonensis. (A = adenine,
T = thymine, C = cytosine, G = guanine, Py = any
pyrimidine, Pu = any purine, N = any base.)

Enzyme

Recognition

sequence(s)

Number of
recogni-
tion sites

AflII

CTTAAG

1

Asel

ATTAAT

>10

Aval

CPyCGPuG

0

BamHI

GGATCC

1

Bell

TGATCA

6

Cfol

GCGC

0

Dral

TTTAAA

>10

EcoO109

PuGGNCCPy

1

EcoRI

GAATTC

3

EcoT22I

ATGCAT

1

Haelll

GGCC

1

Hindi!

GTPyPuAC

4

Hindlll

AAGCTT

7

MspI

CCGG

0

Ndel

CATATG

2

PvuII

CAGCTG

1

Sad

GAGCTC

0

Spel

ACTAGT

4

SspI

AATATT

>10

Xbal

TCTAGA

0.

These data suggest that the mtDNA of H. ar-
izonensis, like that of insects such as Drosophila
(e.g., Clary & Wolstenholme 1 985) and the honey
bee {Apis mellifera L.; Crozier et al. 1 989) is rich

in adenine (A) and thymine (T) and poor in gua-
nine (G) and cytosine (C). Digestion with four-
base enzymes whose recognition sites contain only
cytosines and guanines, reveals one (Haelll) or
no (Cfol, MspI) cleavage sites. At the other ex-
treme, digestion with the 6-base enzymes Asel,
Dral and SspI, whose recognition sites contain
only A’s and T’s, revealed large numbers of
cleavage sites.

While most animal mtDNA’s fall in a rather
narrow size range of 1 6 to 18 kb, the total known
size range for animal mtDNA’s is from 39 kb in
the scallop Placopecten magellanicus (Gmelin)
(Snyder etal. 1987) to 14.285 kb in the nematode
Ascaris suum (Wolstenholme et al. 1 987). Nearly
all animal mtDNA’s that have been investigated
contain the same genes, coding for 22 transfer
RNA’s, 13 proteins and two ribosomal RNA’s
(Brown 1985; Moritz et al. 1987), plus a non-
coding control region (the D-loop in vertebrates
or AT-rich region of insects and other inverte-
brates). Variation in the size of animal mtDNA’s
is commonly due to variation in the size of the
control region, which may contain a variable
number of tandemly repeated sequences (e.g.,
Fauron and Wolstenholme 1980; Harrison et al.
1985; Densmore etal. 1985; Solignac etal. 1986)
but may also be due to duplications in coding
regions (e.g., Moritz & Brown 1986, 1987), or
small tandem repeats, inserts and deletions scat-
tered throughout the mitochondrial genome (e.g..
Powers et al. 1986). The unusually small mito-
chondrial genome of the nematodes is lacking
the ATPase 8 gene, and the tRNA’s have lost a

Hadrurus arizonensis mitochondrial DNA

10

1 1

12

13

14

Kb

H O Di "a >—

cS a

« ® ^

■s I .s

Z P3 ®

r-* ‘-I

Figure 1.— Cleavage map for mitochondrial DNA of the scorpion Hadrurus arizonensis, showing relative
positions of the cleavages generated by the enzymes AflII, BamHI, EcoRI, EcoT22I, Hindll, Ndel, Pvull, and
Spel. The single AflII site was arbitrarily chosen as the starting point for this linear map of the circular mtDNA
molecule. Subsequent sequencing (Smith unpubl. data) shows that the Spel and EcoRI sites at approximately
13.35 kb are 7 base pairs apart, with the Spel site preceding the EcoRI site. The scale is in thousands of base
pairs, or kilobase pairs (kb).

SMITH & BROWN-SCORPION MITOCHONDRIAL DNA

87

stem and loop structure. (Wolstenholme et al.
1987). This raises the possibility that the unusu-
ally small mitochondrial genome of H. arizo-
w/25/5 (13.85-14.0 kb) is also lacking a gene typ-
ically found in animal mtDNA. In this regard it
is interesting that the mtDNA of other chelicer-
ates, including the horseshoe crab, Limulus poly-
phemus (L.) (Xiphosurida), vinegaroon, Masti-
goproctus giganteus (Lucas) (Thelyphonida), and
the spiders Anelosimus eximius (Keyserling) and
A. studiosus (Hentz) (Araneae: Theridiidae), also
have small mitochondrial genomes of 14-15 kb
(Smith unpubl. obs.).

ACKNOWLEDGMENTS

We thank R. Hagen, G. Polis and M. Galindo-
Ramirez for comments on the manuscript, and
G. Polis for help in identification of the scorpi-
ons. This research was supported by NSF grant
BSR-8709661 to DRS, a grant from the Smith-
sonian Institute Scholarly Studies Program to J.
Coddington and DRS, and NSF and NIH grants
to WMB.

LITERATURE CITED

Avise,J.C., etal. 1987. Intraspecific phylogeography;
The mitochondrial DNA bridge between population
genetics and systematics. Ann. Rev. Ecol. Syst., 18:
489-522.

Brown, W. M. 1980. Polymorphisms in mitochon-
drial DNA of humans as revealed by restriction en-
donuclease analysis. Proc. Natl. Acad. Sci. USA, 77:
3605-3609.

Brown, W. M. 1985. The mitochondrial genome of
animals. Pp. 95-130, In Molecular Evolutionary
Genetics (R. J. MacIntyre, ed.). Plenum, New York.
Brown, W. M. & J. Vinograd. 1974. Restriction en-
donuclease cleavage maps of animal mitochondrial
DNA. Proc. Natl. Acad. Sci. USA, 71:4617-4621.
Clary, D. O., & D. R. Wolstenholme. 1985. The mi-
tochondrial DNA molecule of Drosophila yakuba:
Nucleotide sequence, gene organization, and genetic
code. J. Mol. Evol., 22:252-271.

Crozier, R. H., Y. C. Crozier & A. G. MacKinlay.
1989. The CO-I and CO-II region of honey-bee
mitochondrial DNA: Evidence for variation in in-
sect mitochondrial evolutionary rates. Molec. Biol.
Evol., 6:399-411.

Densmore, L. D., J. W. Wright & W. M. Brown. 1985.
Length variation and heteroplasmy are frequent in
mitochondrial DNA from parthenogenetic and bi-
sexual lizards (genus Cnemidophorus). Genetics, 1 1 0:
689-707.

Fauron, C. & D. Wolstenholme. 1980. Intraspecific
diversity of nucleotide sequences within the adenine
+ thymine-rich region of mitochondrial DNA mol-
ecules of Drosophila mauritiana. Drosophila mela-
nogaster and Drosophila simulans. Nucleic Acids
Research, 8:5391-5410.

Harrison, R., D. Rand & W. Wheeler. 1985. Mito-
chondrial DNA size variation within individual
crickets. Science, 228:1446-1448.

Moritz, C. & W. M. Brown. 1986. Tandem dupli-
cation of D-loop and ribosomal RNA sequences in
lizard mitochondrial DNA. Science, 233:1425-1427.

Moritz, C. & W. M. Brown. 1987. Tandem dupli-
cations in animal mitochondrial DNAs: Variation
in incidence and gene content among lizards. Proc.
Natl. Acad. Sci. USA, 84:7183-7187.

Moritz, C., T. E. Dowling & W. M. Brown. 1987.
Evolution of animal mitochondrial DNA: Rele-
vance for population biology and systematics. Ann.
Rev. Ecol. Syst., 18:269-292.

Powers, T. O., E. G. Platzer & B. C. Hyman. 1986.
Large mitochondrial genome and mitochondrial
DNA size polymorphism in the mosquito parasite,
Romanomermis culicivorax. Curr. Genet., 11:71-
77.

Smith, D. R. & W. M. Brown. 1990. Restriction en-
donuclease cleavage site and length polymorphisms
in mitochondrial DNA of Apix mellifera mellifera
and A. m. carnica (Hymenoptera: Apidae). Ann.
Entomol. Soc. Am., 83:81-88.

Snyder, M., et al. 1989. Atypical mitochondrial DNA
from the deep-seascallop Placopecten magellanicus.
Proc. Natl. Acad. Sci. USA, 84:7595-7599.

Solignac, M., M. Monnerot & J.-C. Mounolou. 1986.
Mitochondrial DNA evolution in the melanogaster
species subgroup of Drosophila. J. Mol. Evol., 23:
31^0.

Wolstenholme, D. R., et al. 1987. Bizzare tRNAs
inferred from DNA sequences of mitochondrial ge-
nomes of nematode worms. Proc. Natl. Acad. Sci.
USA, 84:1324-1328.

Wright, J., C. Spolsky & W. M. Brown. 1983. The
origin of the parthenogenetic lizard Cnemidophorus
laredoensis inferred from mitochondrial DNA anal-
ysis. Herpetologica, 39:410-416.

Manuscript received May 1990, revised July 1990.

1991. The Journal of Arachnology 19:88-92

EREMOCHELIS LAGUNENSIS, ESPECIE NUEVA
(ARACHNIDA, SOLPUGIDA, EREMOBATIDAE)

DE BAJA CALIFORNIA SUR, MEXICO

Ignacio M. Vazquez: Laboratorio de Acarologia, Laboratorio de Morfofisiologia Animal,
Facultad de Ciencias, UNAM, 04510 DF, Mexico

Abstract. Eremochelis lagunensis new species from Palo Extrano in Baja California Sur,
Mexico, is described and illustrated. The structure of the fixed finger of the male cheiicerae
of E. lagunensis looks like the fixed fingers of E. rossi Muma, 1986 but the fondal notch
and mesoventral groove of the fixed finger as well as the dentition of the movable finger of
E. lagunensis are different from that species. The females are similar to E. truncus.

Resumen. Se describe e ilustra a Eremochelis lagunensis especie nueva de Palo Extrano,
Baja California Sur, Mexico. La estructura del dedo fijo de los queliceros del macho de E.
lagunensis es similar a la de E. rossi Muma 1986; sin embargo, en E. lagunensis el espacio
de la muesca basal de los queliceros y el canal lateroventral del dedo fijo, asi como la
denticion del dedo movil es diferente a lo que presenta aquella especie. Las hembras se

parecen a las de E. truncus.

Uno de los ordenes de aracnidos muy poco
atendido por los aracnologos es el de los SoH-
fugos, en parte por ser escasos en las colecciones
y tambien por carecer de importancia medica y
economica. Sin embargo, este grupo de animales,
como otros, juega un papel ecologico relevante
como depredadores de diferentes tipos de insec-
tos y otros atropodos, incluso de vertebrados pe-
quenos. Martin H. Muma, de Estados Unidos,
dedico gran parte de su vida al estudio de los
solifugos de America del Norte, incluyendo el
norte de Mexico y las Antillas, describio un gran
numero de taxa nuevos de estas regiones,y tra-
bajo con aspectos de taxonomia (Muma 1951,
1962, 1970, 1985, 1986), biologia (Muma 1966)
y ecologia (Muma 1980) de especies de las fa-
milias Eremobatidae y Ammotrechidae. Otros
especialistas ban publicado sus hallazgos acerca
de solifugos mexicanos de las mismas familias
(Rowland 1974; Vazquez 1981).

El genero Eremochelis comprende veintiocho
especies ubicadas en seis “grupos de especies”:
arcus, bilobatus, branchi, andreasana, imperialis
y striodorsalis (Rowland 1974). Son solifugos de
talla mediana a pequena, alcanzando los 30 mm
de largo. Hasta la fecha se ban colectado al sur
de Estados Unidos en California, Nevada, Ari-
zona y Texas. Al Norte de la Republica Mexicana
se encontraron siete especies de este genero; en
Sonora E. sonorae Muma 1987 y E. imperialis

(Muma 1951); en Durango E. bilobatus Muma
1 95 1 ; en Baja California Norte y Sur E. flexacus
(Muma 1963), E. truncus Muma 1987 y E. an-
dreasana (Muma 1 962); y una especie en el vol-
can Popocatepetl en el Estado de Mexico E. rossi
Muma 1987.

Habiendo estudiado un lote pequeno de ejem-
plares provenientes de la Sierra y el Valle de la
Laguna, Baja California Sur, el material ba re-
sultado pertenecer a especies nuevas, una de las
cuales se describe a continuacion. Las medicio-
nes y el calculo de los radios se bicieron de la
misma forma que Brookbart y Muma (1987) y
Muma y Brookbart (1988), estan dados en mm
y se anotan en las Tablas 1 y 2. La nomenclatura
de los dientes basales “fondal teetb” de los que-
liceros es como la usada por Muma (1951). Se
utilizan las caracteristicas de las sedas ECCS
(Muma 1985) y los radios o cocientes CL/CW
(longitud/amplitud queliceral), PW/PL (ampli-
tud/longitud del propeltidio) y A/CP (longitud
de pedipalpos, patas I, patas IV/longitud de que-
liceros y propeltidio) a manera de complemento
de diagnosis y para bacer comparaciones en el
caso de las bembras.

Eremochelis lagunensis, especie nueva
Figuras 1 a 10.

Descripcion.— Holotipo macbo. Longitud to-
tal = 16.5 mm.

88

YAZQUEL-EREMOCHELIS NUEVA DE MEXICO

89

Tabla 1.— Medidas de tres machos (un holotipo y dos paratipos) de Eremochelis lagunensis, especie nueva.

Estructura

Largo

Ancho

Radios

Queliceros

3. 8-4.8

1. 3-1.6

CL/CW = 2.96

Propeltidio

1. 8-2.4

2.5-3. 1

PW/PL= 1.33

Pedipalpos

15.6-19.3

A/CP = 7.81

Patas I

10.0-13.4

Patas IV

18.6-23.1

Prosoma: Queliceros mas largos que los de otras
especies del mismo genero (Tabla 1), con man-
chas de color marron oscuro formando bandas
longitudinales delgadas casi negras, el resto es de
color amarillo palido. Complejo flagelar con la
seda apical plumosa cilindrica, no alargada, con
su extremo en forma de S (Fig. 5); sedas dorsales
lisas y cilindricas. Denticion queliceral como se
muestra en las figuras 1 y 2. Dedo fijo con dos
rebordes en forma de dientes, bien distinguibles,
sobre la orilla de un proceso laminar, formado
por el ensanchamiento del canal lateroventral, el
cual ocupa toda la parte ventral del dedo. Dicho
proceso es concavo en su cara interna (Fig. 6).
Con dos dientecillos en el espacio correspon-
diente a la muesca basal “fondal notch” del dedo
fijo. Hay cinco dientes basales extemos en orden
decreciente de tamano I, II, IV, III, V (Fig. 1) y
cuatro en la parte interna I, III, II, IV (Fig. 2).
Dedo movil con cuatro dientes en orden decre-
ciente de tamano como sigue: principal, anterior
y dos intermedios. El diente anterior esta sepa-
rado de los tres restantes por un espacio igual al
que ocupan los dientes intermedios juntos (Fig.
1). Sedas basales extemas del dedo movil = ECCS
(Muma 1985) como en la figura 10. Propeltidio
tan ancho como largo (Tabla 1), con manchas
oscuras a los lados de una banda media longi-
tudinal color amarillo claro, desde el tuberculo
ocular hasta el borde posterior. Las manchas os-
curas en pedipalpos y patas van desde los fe-
mures hasta los tarsos, excepto en las articula-
ciones. Metatarsos de los pedipalpos con espinas
gruesas impares, sin escopula. Con un par de
unas en cada tarso del primer par de patas.

Opistosoma: Color amarillo palido con los ter-
guitos marron oscuro: primer estemito posestig-
mal con un ctenidio (ctenidio = conjunto de se-
das formando un peine) ventral con cinco o seis
sedas gruesas (0.06 mm de ancho y 0.4 mm de
largo), planas, de color amarillo oscuro casi ana-
ranjado (Fig. 7).

Paratipo hembra inmadura coloreada. Lon-
gitud total =13.1 mm.

Prosoma: Coloracion de las hembras estudia-
das semejante a la del macho holotipo, excepto
que en una de ellas los terguitos del opistosoma
son de color claro uniforme. Denticion queliceral
como se muestra en las figuras 3 y 4. Dedo fijo
con dos dientes entre el anterior y el intermedio,
asi como entre el intermedio y el principal (Fig.
3). Denticion basal interna en orden decreciente
de tamano I, III, II (Fig. 4) y la externa I, II, III,
IV (Fig. 3). Dedo movil con dos dientecillos, casi
indistinguibles frente al diente anterior y dos pe-
quenos rebordes (Figs. 3 y 4); con un diente pe-
queno contiguo al principal, siendo el diente an-
terior y el principal casi del mismo tamano. Sedas
basales extemas del dedo movil (ECCS) como
en las figuras 9 y 1 1 .

Opistosoma: La forma de las placas genitales
no se ve claramente, ya que se trata de hembras
inmaduras; el esclerosamiento es poco y la co-
loracion tambien, aunque la forma que presenta
la hembra coloreada es como en la figura 8. La
hembra de color claro no presenta los repliegues
entre las placas genitales y las coxas IV como se
ve en la figura 8. La delineacion de los hordes
de las placas se efectuo por la diferencia de te-
gumento, es decir, sobre las placas hay sedas y
en el resto del estemito solo se ven pliegues. No
hay ctenidio sobre el primer estemito posestig-
mal.

Localidad tipo: Valle de La Laguna, Baja Cal-
ifornia Sur, MEXICO.

Etimologfa: El nombre se refiere a la localidad,
tanto al Valle como a la Sierra de La Laguna,
que son los sitios donde se colectaron los ejem-
plares.

Material estudiado. — MEXICO: Tres machos
y dos hembras. Un macho holotipo: Valle de La
Laguna, Baja California Sur, 20 mayo 1988, M.
Vazquez col., depositado en la coleccion 1. M.
Vazquez, Mexico, D.F. (VM). Dos hembras in-
maduras paratipos: Palo Extrano, Sierra de la
Laguna, Baja California Sur, 14 mayo 1986, F.
Cota y A. Cota col.; una hembra depositada en
VM y la otra en el American Museum of Natural

90

THE JOURNAL OF ARACHNOLOGY

Figuras \-\\.—Eremochelis lagunensis especie nueva, holotipo macho y paratipo hembra; 1, vista lateral
externa del quellcero derecho del macho; 2, vista lateral interna del quelicero derecho del macho; 3, vista lateral
externa del quelicero derecho de la hembra; 4, vista lateral interna del quelicero derecho de la hembra; 5, seda
plumosa apical del complejo flagelar del macho holotipo; 6, vista ventral del dedo fijo del quelicero derecho del
macho; 7, sedas del ctenidio opistosomal, sobre el primer segmento posestigmal; 8, esquema de las placas
genitales de la hembra inmadura; 9, sedas ECCS de la hembra inmadura de color claro; 10, sedas ECCS del
macho holotipo; 1 1, sedas ECCS de la hembra inmadura coloreada.

NAZQ\5EZ-EREM0CHELIS NUEVA DE MEXICO

91

Tabla 2.— Medidas de dos hembras inmaduras (paratipos) de Eremochelis lagunensis, especie nueva.

Estructura

Largo

Ancho

Radios

Queliceros

3. 1-3.2

1. 1-1.1

CL/CW = 2.86

Propeltidio

1.5-1. 5

2.2-2.4

PW/PL= 1.5

Pedipalpos

8. 1-8.7

A/CP = 5.6

Patas I

6.7-6.8

Patas IV

10.0-11.8

History (AMNH). Un macho paratipo: Palo Ex-
trafio, Sierra de la Laguna, Baja California Sur,
5 mayo 1986, F. Cota y A. Cota col, depositado
en AMNH. Un macho paratipo; Palo Extrano,
Sierra de la Laguna, Baja California Sur, 4 mayo
1983, M. Vazquez col., depositado en VM.

DISCUSION

En Eremochelis lagunensis especie nueva, la
seda apical plumosa del complejo flagelar no esta
alargada y es cilindrica, las sedas dorsales son
cih'ndricas lisas, caracteristico del genero. El ca-
nal lateroventral es amplio y concavo, lo que
hace que se relacione con las especies del grupo
bilobatus. El dedo fijo de los queliceros del ma-
cho de E. lagunensis es semejante al de E. rossi
en la forma de la muesca lateroventral, difiere
de esta por carecer de dientecillos en la muesca
del fondo. El dedo movil de los machos de E.
lagunensis es diferente al del macho holotipo de
E. rossi por tener el diente anterior y un dien-
tecillo intermedio bien distinguibles. Entre las
especies del grupo bilobatus, solo E. rossi se pue-
de considerar estrechamente relacionada con E.
lagunensis. Las sedas ECCS de E. lagunensis son
diferentes en numero y forma a las de E. rossi
aunque, como lo dice Muma ( 1 985), este caracter
debe ser probado ampliamente y no es definitivo
en todos los casos si se considera por separado.

Las hembras revisadas no son maduras y es
dificil hacer comparaciones a nivel de placas ge-
nitales. Sin embargo, al revisar las placas geni-
tales de nuestros ejemplares, nos dimos cuenta
que solo en la que tiene bien definido el color de
los terguitos opistosomales, se distingue la forma
similar a la hembra de E. bilobatus. Por otro
lado, los queliceros de las hembras estudiadas
son similares a los de la hembra de E. truncus,
excepto que en esta hay un diente intermedio
contiguo al principal y en nuestros ejemplares
hay dos. En el dedo fijo de las hembras revisadas
hay dos dientes detras del diente anterior y en la

hembra de E. truncus solo hay uno. Es posible
que nuestros ejemplares correspondan a hembras
de diferente especie ya que las sedas ECCS son
diferentes para cada hembra (ver Figs. 9 y 1 1 ) y
son diferentes a las del macho holotipo. Sin em-
bargo, puede ser que al manipular los queliceros
se hayan roto las sedas pequenas. La forma de
las placas genitales es muy parecida. Descarta-
mos la posibilidad de que la hembra coloreada
sea el estado juvenil de E. truncus y que el macho
de la especie que aqui se describe sea sinonimo
de tal especie, principalmente por las diferencias
en talla y tambien por las diferencias en denticion
queliceral. Sobre todo, la relacion entre longitud
queliceral y longitud general del cuerpo puede
distinguir a los machos y hembras de E. lagu-
nensis, aqui descritos, de otras especies (Tablas
1 y 2). En todo caso, las hembras estudiadas
pueden no corresponder a los machos descritos,
pero se comprobo que ellas son del mismo genero
y que se relacionan con las especies del grupo
bilobatus, por la forma de su placa genital en vias
de ser madura.

AGRADECIMIENTOS

Quiero- dedicar este trabajo a la memoria del
Doctor Martin H. Muma, quien me enseno todo
lo que se de solifugos y me animo en muchas
ocasiones para publicar el resultado de mis in-
vestigaciones. Por otra parte, agradezco el envio
del material del presente estudio a las Doctoras
Magdalena Vazquez y Maria Luisa Jimenez, co-
misionadas por el Centro de Investigaciones Bio-
logicas de Baja California Sur, A.C. para estudiar
la fauna terrestre de la Sierra de la Laguna.

LITERATURA CITADA

Brookhart, J. O. & M. H. Muma. 1987. Arenotherus
a new genus of Eremobatidae (Solpugida), in the
United States. Domestic Publication. 1 8 pp.
Muma, M. H. 1951. The arachnid order Solpugida
in the United States. Bull. Amer. Mus. Nat. Hist.,
97:31-142.

92

THE JOURNAL OF ARACHNOLOGY

Muma, M. H. 1962. The arachnid order Solpugida
in the United States, supplement 1. Amer. Mus.
Novitates, 2092:1-44.

Muma, M. H. 1966. Egg deposition and incubation
in Eremobates durangonus with notes on the eggs
of other species of Eremobatidae. Florida EntomoL,
49:23-31.

Muma, M. H. 1970. A synoptic review of North
American, Central American and West Indian Sol-
pugida (Arthropoda: Arachnida). Arthropods of
Florida and Neighboring Land Areas, 5:1-62.

Muma, M. H. 1980. Solpugid (Arachnida) popula-
tions in a creosotebush vs. a mixed plant associa-
tion. Southwestern Naturalist (25), 2:129-136.

Muma, M.H. 1985. A new, possibly diagnostic, char-
acter for Solpugida (Arachnida). Novitates Arthro-
podae, 2:1-4.

Muma, M. H. 1986. New species and records of Sol-

pugida (Arachnida) from Mexico, Central America
and the West Indies. Novitates Arthropodae, 2(3):
1-23.

Muma, M. H. & J. O. Brookhart. 1988. The Ere-
mobates palpisetulosus species-group (Solpugida:
Eremobatidae) in the United States. Cherry Creek
School District, Colorado, USA. 65 pp.

Rowland, J. M. 1974. A new solpugid of the genus
Eremochelis (Arachnida: Solpugida: Eremobatidae)
from California, with a key to males of the genus.
Occas. Pap. Mus. Texas Tech Univ., 25:1-8.
V^quez, I. M. 1981. Contribucion al Conocimiento
de los Solifugos de Mexico (Arachnida: Solifugae).
Tesis Profesional, Facultad de Ciencias, UNAM. 80

pp.

Manuscript received May 1990, revised August 1990.

1991. The Journal of Arachnology 19:93-96

NUEVOS APORTES AL GENERO PORRIMOSA ROEWER
(ARANEAE, LYCOSIDAE)

Roberto M. Capocasale: Institute de Investigaciones Biologicas, Clemente Estable, Di
vision Zoologia Experimental, Ave. Italia 3318, Montevideo, Uruguay

Resumen. Se describen la hembra de Porrimosa harknessi (Chamberlin) y el macho de
Porrimosa castanea (Mello-Leitao) hasta ahora desconocidos. Se da una clave de identifi-
cacion nueva y se ampHa la distribucion de las tres especies seguras de Porrimosa.

Abstract. The female of Porrimosa harknessi (Chamberlin) and the male of Porrimosa
castanea (Mello-Leitao) still unknown are described. A key to identification and a wider
distribution of the three good species of Porrimosa are given.

Porrimosa es un genero que plan tea dudas y
numerosas incognitas. Algunas de las causas que
motivan esa situacion son: (a) el escaso numero
de ejemplares disponibles en colecciones; (b) el
alto porcentaje de especies descriptas como vMi-
das pero basadas en ejemplares inmaduros; (c)
las dificultades factuales para evaluar caracteres
definitorios a niveles generico y especifico, de-
bido al mal estado de conservacion de los “ti-
pos”; (d) la ausencia casi total de informacion
biologica y ecologica sobre las especies.

Indudablemente Porrimosa es un grupo her-
mano (“sister group”) de Sosippus (Dondale
1986). Lo poco que se conoce sobre la biologia
de Porrimosa (Bucher 1974 en P. lagotis) y de
Sosippus (Qr2id.y 1962, en^. texanus] Brach 1976
en S. floridanus) reafirma esa conclusion; ir mas
lejos y establecer otras relaciones es solo es-
peculacion.

El objetivo de este articulo es comunicar los
resultados obtenidos despues de consultar y es-
tudiar el material prestado de casi la totalidad
de las colecciones de America del Sur, que tienen
ejemplares de Porrimosa. Estos aportes solo tie-
nen la pretension de complementar mi estudio
anterior (Capocasale 1982).

Dado que la informacion sobre este genero es
pobre, trate de aprovechar todo dato util recogi-
do con la finalidad de neutralizar dicha carencia.

Aqui se describen el macho de Porrimosa cas-
tanea Mello-Leitao y la hembra de Porrimosa
harknessi Chamberlin, hasta ahora desconoci-
dos. Tomando como base ese material nuevo,
hice un estudio comparativo de los escleritos pal-
pales en cada una de las tres especies seguras; de
donde surgio una clave de identificacion nueva.

ahora mas completa que la dada en Capocasale
(1982). Tambien se ampHa la distribucion geo-
grafica de algunas especies, con 22 localidades
nuevas, lo que permite comprobar que ni Por-
rimosa castanea ni Porrimosa harknessi estan
restringidas a algunas regiones como se suponia.

Teniendo en mente el me todo de trabajo de
Dondale (1986), quien establecio nuevas divi-
siones a nivel de subfamilia en Lycosidae, en-
fatizo la importancia de los escleritos palpales,
fundamentalmente de las apofisis lateral del con-
ductor y mediana del “tegulum,” como carac-
teres para hacer una identificacion segura. Esto
apoyado en la funcion de las mismas.

Abreviaturas. — MZUSP, Museu de Zoologia,
Universidad de Sao Paulo, Brasil; MNRJ, Museu
Nacional de Rio de Janeiro, Brasil; MRCN, Mu-
seu Riograndense de Ciencias Naturals, Porto
Alegre, Brasil; MNHN, Museo Nacional de His-
toria Natural de Montevideo, Uruguay.

CLAVE DE ESPECIES

HEMBRAS

1. Pieza transversa del “septum” aproximada-

mente 3 veces mas ancha que larga. El centre
donde intersectan ambas piezas septales
convexo, abultado, con 2 proyecciones a
ambos lados (Figs. 4-5). Espermatecas con
forma de C, con tuberculos en el apice, or-
ientadas hacia las areas laterales. Bolsas co-

pulatorias bilobuladas castanea

Pieza transversa del “septum” aproximada-
mente 4 veces mas ancha que larga (Fig. 6)
2

2. Espermatecas semejantes a P. castanea. Bolsas

copulatorias simples, con una protuberancia

93

94

THE JOURNAL OF ARACHNOLOGY

1

Figura l.—Porrimosa lagotis (Holmberg) (Uruguay, Canelones, Marindia, MNHN).

pequena, conica, cerca de los tubos copu-

latorios (Fig. 7) harknessi

Espermatecas semejantes as P. harknessi. Bol-
sas copulatorias simples lagotis

MACHOS

1 . Apofisis mediana del “tegulum” corta, conica,

canaliculada. Apofisis lateral del conductor
curva, diametro de la base ancho, terminada

en punta roma (Fig. 8) castanea

Apofisis mediana del “tegulum” alargada, can-
aliculada 2

2. Apofisis lateral del conductor curva, alargada

diametro de la base angosto, terminada en

punta roma (Fig. 9) harknessi

Apofisis lateral del conductor corta, muy cur-
va, diametro de la base ancho, terminada en
punta aguda (Fig. 10) lagotis

Porrimosa castanea (Mello-Leitao, 1 942)
Figs. 4-5, 8

Porrimosa castanea: Capocasale 1982:149.

Macho.— Cuerpo: largo 14 mm. Cefalotorax:
largo 7.3 mm, ancho 6.9 mm (1 ejemplar medi-
do), bandas laterales castano naranja, bandas
submarginales continuas, amarillo, 6 bandas ra-
diales castano naranja, convergentes hacia el sur-
co toracico, area ocular castano naranja. Ester-
non: largo 3.5 mm, ancho 3.0 mm, castano palido.
Dientes en el borde posterior intemo de los que-
liceros: cantidad 3. Femur I: largo 1 1 mm. Femur
11: largo 10 mm. Femur III: largo 9.5 mm. Femur
IV: largo 11.9 mm. Femures, patelas, tibias y
basitarsos de patas I-IV castano naranja. Ab-
domen: dorsal castano oscuro, areas laterales
castano oscuro, ventral castano amarillo. El ab-
domen insinua un levisimo diseno presunta-
mente perdido por el tiempo durante el cual el
animal estuvo en Hquido y por la tecnica de con-
servacion usada. Apofisis mediana del “tegu-
lum”: corta, conica, canaliculada (Fig. 8). Apofisis
lateral del conductor: curva, diametro de la base
ancho, terminada en punta roma (Fig. 8).

CAPOCASALE-NUEVOS APORTES AL GENERO PORRIMOSA

95

2

3

Figuras 2-10.— 2, 3, esquemas mostrando donde se tomaron las medidas en el epigino (a = ancho; 1 = largo)
y en la apofisis lateral del conductor (d = diametro de la base), respectivamente; 4, 5, epigino de Ponimosa
castanea (Mello-Leitao) (Brasil, Sao Joao, Agua Preta, MNRJ); 4, vista ventral; 5, vista anterior; 6, 7, Porrimosa
harknessi (Chamberlin) (Brasil, Riogrande do Sul, Passo Fundo, MRCN); 6, epigino (pt = pieza transversa de!
“septum”); 7, espermatecas (e) y bolsas copulatorias (be); 8-10, apofisis lateral del conductor (al) y apofisis
mediana del “tegulum” (am); 8, Porrimosa castanea-, 9, Porrimosa harknessi-, 10, Porrimosa lagotis.

96

THE JOURNAL OF ARACHNOLOGY

Distribudon.— Centro de Argentina, centro de
Peru y sur de Brasil.

Ejemplares examinados.— BRASIL: SaoJodo;
Agua Preta, set. 1927, 1 macho, 1 hembra
(MNRJ). ARGENTINA: Cdrdoba; Talumba (M.
Biraben), 1 hembra (MNRJ). BRASIL: Goias;
Rio Oliveira, 26 jun. 1942, 2 hembras (MNHN).

Porrimosa harknessi (Chamberlin, 1916)
Figs. 6, 7, 9

Porrimosa harknessi: Capocasale 1982:151.

Hembra.— Cuerpo: largo 17 mm. Cefalotorax:
largo 7 mm, ancho 5.3 mm (1 ejemplar medido),
bandas laterales amarillo castaho, bandas sub-
marginales contlnuas, amarillo, 6 bandas ra-
diales castano oscuro convergentes hacia el cen-
tro del surco toracico, area ocular castano negro.
Estemon: largo: 3.5 mm, ancho 2.5 mm, castano.
Dientes en el borde posterior intemo de los queli-
ceros: cantidad 3. Femur I: largo 5.8 mm. Femur
II: largo 5.7 mm. Femur III: largo 6.3 mm. Fe-
mur IV: largo 7.3 mm. Femures, patelas, tibias
y basitarsos de patas I-IV amarillo castano. Ab-
domen: dorsal una banda mediana castano os-
curo (el diseno muy semejante al de P. lagotis
(Fig. 1), areas laterales castano gris, ventral el
mismo tinte que el de las areas laterales. “Sep-
tum”: pieza mediana corta, pieza transversa
aproximadamente 3 veces mas ancha que larga
(Fig. 6). Espermatecas: con forma de C, los apices
orientados hacia las areas laterales. Bolsas co-
pulatorias: simples, con una protuberancia pe-
quena, conica, cerca de los tubos copulatorios
(Fig. 7).

Distribucion.— Sur de Brasil y sur de Peru.

Ejemplares examinados. — BRASIL: Rio-
grande do Sul, Passo Fundo, 12 oct. 1985 (A.
Lise), 1 macho, 1 hembra, 2 hembras juveniles
(MRCN).

Porrimosa lagotis (Holmberg, 1876)

Figs. 1, 10

Porrimosa lagotis: Capocasale 1982:151 (nec. P. la-
gotis (Mello-Lehao, 1941)).

Porrimosa lagotis: Platnick 1989:387.

Distribucidn.— Centro de Argentina, centro y
sur de Brasil y Uruguay.

Localidades nuevas.— ARGENTINA: Rw Ne-
gro-, Parana (MNRJ). BRASIL: Minas Gerais;

Morro Garza (MZUSP), Ribeirao (MZUSP); Sao
Paulo; Barnevi (MZUSP); Foz Itaquere
(MZUSP); Franca (MZUSP); Rio Claro
(MZUSP); Alto da Serra (MZUSP); Cananeia
(MZUSP); Parana; Villa Velha (MZUSP); Goias.
Cabeceiras (MZUSP); URUGUAY: Canelones;
Las Piedras (MNHN); Maldonado. Aigua
(MNHN), Sierra de las Animas (MNHN), Cerro
San Antonio (MNHN), Cerro Largo; Rio Ta-
cuari — ruta 8 (MNHN); Rivera; -ciudad-
(MNHN); Treinta y Tres; Quebrada de los Cuer-
vos (MNHN).

AGRADECIMIENTOS

A los Dres. H. W. Levi (MCZ), A. Lise (MRCN)
y A. Timotheo Da Costa (MNRJ) por el pres-
tamo de ejemplares, A. Brady y C. Valerio por
sus comentarios y sugerencias en la revision del
primer manuscrito, y a Exline-Frizzell Fund for
Arachnological Research por financiar la publi-
cation en The Journal of Arachnology (18:131-
141) de la primera parte de este estudio.

BIBLIOGRAFIA CITADA

Brach, V. 1976. Subsocial behavior in the funnel-
web wolf spider Sosippus floridanus (Araneae, Ly-
cosidae). The Florida Entomol., 59:225-229.
Brady, A. 1962. The spider genus Sosippus in North
America, Mexico, and Central America (Araneae,
Lycosidae). Psyche, 69:129-164.

Bucher, E. 1974. Observaciones ecol6gicas sobre ar-
tropodos del bosque chaqueno de Tucuman. Rev.
Fac. Ci. Ex. Fis. Nat. C6rdoba (N.S.). Biologia, 1:
35-122.

Capocasale, R. M. 1982. Las especies del genero Po-
rrimosa Roewer, 1959 (Araneae, Hippasinae). J.
Arachnol., 10:145-156.

Chamberlin, R. J. 1916. Results of the Yale Peruvian
Expedition of 1911: The Arachnida. Bull. Mus.
Comp. Zool. Harvard, 60:177-299.

Dondale, C. D. 1986. The subfamilies of wolf spiders
(Araneae: Lycosidae). Actas X Congr. Int. Aracnol.
Jaca, Espana, 1:327-332.

Holmberg, E. L. 1876. Aracnidos argentinos. An.
Agr. Argentina, 4:1-30.

Mello-Leitao, C. 1 942. Cinco aranhas novas do Peru.

Rev. Brasil. Biol., 2:429-434.

Platnick, N. 1989. Advances in Spider Taxonomy,
1981-1987: A Supplement to Brignoli’s. A Cata-
logue of the Araneae Described between 1 940 and
1981. Manchester Univ. Press, Oxford.

Manuscript received March 1990, revised September
1990.

1991. The Journal of Arachnology 19:97-101

MOTHER OFFSPRING FOOD TRANSFER IN
COELOTES TERRESTRIS {ARANEAE, AGELENIDAE)

Jean-Luc Gundermann, Andre Horel and Chantal Roland: Laboratoire de Biologie du
Comportement, Unite associee au CNRS, URA 1293, Universite de Nancy I, BP 239,
54506 Vandoeuvre-Les-Nancy Cedex, France

Abstract. Three different modes of maternal food supply have been reported in the sub-
social agelenid species Coelotes terrestris: prey provisioning, consumption of the mother’s
body, and regurgitation of nutritive fluids. Although the first two modes are well documented,
the latter is not fully assessed.

By comparing— in the absence of any prey— the weight variations in spiderlings either left
with their mother or isolated, and by simultaneously comparing the weight variations of
mothers, either isolated or left within the group of spiderlings, it was possible to see evidence
of a significant and long-lasting food transfer from the mother to her progeny. This food
transfer probably explains the high level of survivorship and reduction of cannibalism
showed by broods left with their mothers.

Close observation provided no direct evidence of mouth to mouth transfer. Rather, the
food transfer appears to involve the production and emission of miniature eggs by the
mother when in presence of spiderlings, a phenomenon which to date seems not to have
been noted among spiders.

Providing offspring with food is one of the
most important features of parental care, in that
it spares the young the many risks related to food
supply (Wilson 1971). In spiders, food may be
provided by the mother in various forms. The
yolk contained in the egg ensures the beginning
of the spiderling’s development during its life
inside the egg sac and the few days following its
emergence (Foelix 1982). This supply can be sup-
plemented, in several species, by oophagy of non-
hatched eggs inside the egg sac (Schick 1972;
Valerio 1974). In a certain number of species,
the mother provides food to a greater extent. Her
body may be a usual, or occasional, resource for
her progeny (Bristowe 1958; Cloudsley-Thomp-
son 1955; Kullmann 1972; Seibt & Wickler 1987;
Tahiri et al. 1989). Provisioning has also been
reported, either in the form of prey items sub-
dued by the mother and carried to the brood
(Brach 1976; Hirschberg 1969; Gundermann et
al. 1988; Tretzel 1961) or of fluids regurgitated
to the spiderlings (Locket 1926; Kullmann &
Zimmermann 1974). These fluids may be the
result of the partial digestion of the prey, or
sometimes, a production of the mother’s diges-
tive tract (Nawabi, cited by Collatz 1987).

We are currently studying the mother-young
interactions related to food in a European sub-

social species Coelotes terrestris (Wider) (KTafft
et al. 1 986), in particular the three modes of food
supply previously reported by T retzel (1961): prey
provisioning (Gundermann et al. 1988), moth-
er’s consumption (Gundermann 1989) and re-
gurgitation. This last mode is poorly document-
ed. Tretzel (1961) referred, for this species, to
only one definite observation, and our own ob-
servations are rare and dubious. Such a situation
could be explained, either by the actual scarcity
of the phenomenon— which would then lead to
question its functional significance— or by its lo-
calization during the nocturnal phase making the
observations difficult. Thus, the very first step of
the study, as exposed in the present paper, was
to demonstrate the existence of a significant food
transfer from the mother to her offspring, and to
try and find out its nature.

MATERIALS AND METHODS

The funnel-web agelenid Coelotes terrestris is
a terricolous spider common in European wood-
lands. The female weaves a silken tube under
stones, bark of dead logs, etc. From a lenticular
egg sac, 40 to 60 spiderlings emerge, stay in a
group inside the tube with their mother for about
one month, and then disperse and lead a solitary
life (Tretzel 1961; Gundermann 1989).

97

98

THE JOURNAL OF ARACHNOLOGY

Brood's size

Figure 1.— Weekly variations of the broods’ sizes
(median, quartiles). Open squares = treatment A{N =

1 7): orphaned brood; solid squares = treatment B (N
= 1 7): brood left with mother. E.: Emergence day. Mann-
Whitney [/-test: ns, non significant; *** P < 0.001.

In the present study, field inseminated females
were collected, then reared in the laboratory in
plastic boxes (15 x 9 x 7.5 cm) with transparent
sides and a bottom covered with a mixture of
sand and peat which was regularly moistened.
The rearing boxes were kept in a closed room,
at a temperature of 2 1 °C ( ± 2 °), fluorescent lights
providing a light of about 100 lux (12 L/1 2 D).

Two experiments were designed:

1 — In the first experiment, it was hypothesized
that if a food transfer did exist from the mother
to her young, spiderlings left with their mother
should present higher weights than orphaned spi-
derlings, while mothers left with their young
should present lower weights than isolated moth-
ers.

The experiment started from the broods’
emergence (most of them occurred within 2
weeks) and lasted 4 weeks. Throughout that time,
all the spiders were deprived of prey. Egg sacs
were randomly assigned to one of two treat-
ments:

—Treatment A (N = 17): mothers and young
separated;

—Treatment B {N = 17): mothers left with
their young.

Mothers and broods were weighed, first just
after emergence, then at the end of each week.
In order to minimize disturbance of the brood,
each weighing was limited to a sample of 10
spiderlings collected, as randomly as possible, by

aspiration using a pipette. The brood’s weight
was estimated by multiplying the individual mean
weight (given by the sampling) by the number of
spiderlings still alive. At the end of the experi-
ment some broods’ sizes fell below 10 individ-
uals; in this case, of course, the totality of the
brood was weighed.

2— The second experiment was designed to get
further information by trying to enhance mother-
young interactions. Ten groups of three day old
spiderlings were separated from their mothers for
24 h without any food. The subsequent reunion
of the mother with their young were thoroughly
observed and videorecorded.

RESULTS

Survivorship.— Orphaned broods (treatment
A) showed a very poor survivorship: at the end
of the experiment, the median rate of survivor-
ship (number of spiderlings still alive/initial size
of the brood x 100) fell to 12.5% (quartiles: 6.3-

25.5) . In contrast, survivorship stayed high in
treatment B (median rate: 8 1.8%, quartiles: 76.6-

88.5) . Actually the difference between the broods’
sizes of the two treatments was significant from
the end of week 1 on (Fig. 1) (Mann- Whitney
t/-test, P < 0.001). Direct observations and the
fact that no dead bodies could be found in the
vials lead to attribute this differential mortality
to a high incidence of cannibalism within or-
phaned broods.

Weights’ variations.- Statistical analysis
showed no significant difference between treat-
ments A and B, on emergence day, for either the
broods’ as well as the mothers’ weight (Mann-
Whitney U-test, ns).

From the end of week 1 on, the estimated
weights of the broods left with their mothers were
significantly higher than those of the orphaned
broods (Fig. 2a). The analysis of weights’ weekly
variations within each treatment (Table 1) shows
that, while orphaned broods’ weights (treatment
A) decreased (Wilcoxon matched-pairs signed-
rank test: P < 0.002 for weeks 2 and 3; P <
0.00 1 for week 4), weights of broods left with
their mothers (treatment B) varied in the op-
posite direction (Wilcoxon test: P < 0.002 for
weeks 1 and 2, ns for week 3, P < 0.02 for week
4).

The decrease in weight observed in treatment
A can be accounted for by basal metabolism and,
probably, by losses of material and energy brought
about by high intrabrood cannibalism. On the

GUNDERMANN ET AL.- MATERNAL FOOD TRANSFER

99

Broods weight (mg)

Mother’s weight (mg)

Weeks

Figure 2. —Weekly weight variations in mg (median,
quartiles). Open squares = treatment A (N = 17): or-
phaned brood; solid squares = treatment B (A = 1 7):
brood left with mother. E.: Emergence day. Mann-
Whitney D-test: * P < 0.0 1 ; *** P < 0.00 1 . a = Brood’s
weight variations; b = Mother’s weight variations.

other hand the low level of cannibalism in treat-
ment B cannot explain the increase in weight
observed in this treatment. Rather, the increase
suggests that some sort of food is provided to the
young by their mother.

This is confirmed by the study of the mothers’
weight variations (Fig. 2b). From the end of week
2, treatment B mothers were significantly lighter
than treatment A mothers. Furthermore, the
comparison of the weekly losses in weight be-
tween treatment A and treatment B (Table 2)
shows that, when left with their broods, the
mothers suffered significantly more severe weight
decrease than when isolated.

Modes of mother- young food transfer.— The
maintenance of weight gain in broods left with
their mothers over four weeks strongly suggested
that the foods transfer should be a relatively long-
lasting phenomenon. It was, thus, very unlikely
to consist of mere regurgitation of fluids extract-
ed from the prey eaten by the mother before the
beginning of the experiment. The transferred food
should, then, be produced by the mother herself

Close examination of rearing boxes, using a
stereomicroscope, showed, at times, spiderlings
eating substances deposited on the web. These
substances consisted of either clear yellow drops
of liquid, or brownish compact clusters of more
or less discernible eggs. The egg sizes varied
greatly, but were always smaller (0.3-0. 5 mm)
than the normal ones (0.7-1 mm). Some of the
egg-clusters were covered with a thin layer of silk,
similar to the internal layer of the egg sacs.

The second experiment was aimed at directly
observing the possible mother-young food trans-
fer. Actually, the reunion of the mothers with
their young after a 24-h separation did not induce
any special reactions other than an intensive
weaving activity of the mothers, except in one
group in which the young, unlike the others, dis-

Table 1.— Variations of the broods’ median weights per week, in mg (range), within each treatment. Treatment
A {N = 17): orphaned brood; treatment B (V= 17): brood left with mother. Wilcoxon matched-pairs signed-
rank test.

Treatment

Week 1

Week 2

Week 3

Week 4

Treatment A

-2.2

-11.9

-8.6

-2.8

Mothers isolated

(-47.3; +21.9)

(-32.7; +12.3)

(-29.2; +3.5)

(-23.4; +1.0)

P value 2-tailed

ns

<0.002

<0.002

<0.001

Treatment B

+ 19.2

+ 18.9

+ 7.0

+ 5.3

Mothers with broods

(-6.7; +71.9)

(-4.3; +72.1)

(-25.4; +38.3)

(-9.6; +31.6)

P value 2-tailed

<0.002

<0.002

ns

(0.02)

100

THE JOURNAL OF ARACHNOLOGY

Table 2.— Mothers’ median weight losses per week, in mg (range). Treatment A {N = 17): mother separated
from brood; treatment B (A = 1 7): mother left with brood. Comparison treatment A vs. treatment B for each
week: Mann-Whitney [/-test.

Treatment

Week 1

Week 2

Week 3

Week 4

Treatment A

-6.8

-4.7

-1.8

-2.5

Mothers isolated

(-3.4; -10.6)

(0.9; -8.3)

(-0.3; -4.5)

T

1

oo

Treatment B

-21.3

-24.1

-11.5

-6.0

Mothers with broods

(-5.9; -36.8)

(-15.0; -26.6)

(-7.0; -18.2)

(-3.7; -10.1)

P value 2-tailed

<0.02

<0.001

<0.001

<0.05

played a high level of activity and interacted fre-
quently with their mother.

In this group, from the beginning, a few spi-
derlings followed their mother in her numerous
movements and made attempts to come into
contact with her, mainly in the direction of her
hind-legs and opisthosoma (one of them was even
observed hanging from a spinneret). About 2 h
30 min after the reunion, the first emission of
substances took place with two more emissions
being recorded within the following 10 min. As
soon as the drops of substance were deposited
on the substratum they immediately attracted
groups of two or three spiderlings. During the
last emission the orientation of the mother’s
opisthosoma made it possible to clearly observe
the progressive exuding of substance from the
vaginal opening. A spiderling, which had suc-
ceeded in climbing on to the opisthosoma, came
to the vaginal opening, seized the small ball of
substance and eventually took it away.

CONCLUSION

In Coelotes (errestris food is actually trans-
ferred from the mother to her young. This trans-
fer occurs when no prey are available, even for
several weeks, thus indicating that food is pro-
duced by the mother herself. Even though the
digestive tract can not be definitely ruled out as
a site of production, the ovaries appear to be
playing a major role by producing miniature tro-
phic eggs. Observations showed that these eggs
are particularly attractive to the young and rap-
idly eaten. This might explain why they had not
been noticed before.

Such a transfer of food is likely to enhance
survivorship of young during food shortage in
field conditions, at the expense of the mother’s
own survivorship. But the adaptive significance
of this phenomenon remains to be more precisely
assessed; some incidental observations suggest

that food transfer could also occur in normally
fed groups.

Further investigations are also required to find
out the mechanisms involved in the process. To
what extent is ovarian production continuous?
Is ovarian physiology influenced by the presence
of the brood? Does tactile (or other) stimulation
of the mother by the spiderlings release the emis-
sion of ovarian substance, as suggested by our
observations?

As far as we know, this mode of maternal feed-
ing has never been previously reported in spi-
ders, but can be related to a somewhat similar
phenomenon described in the cricket Anuw-
gryllus muticus (West & Alexander 1 963). In this
sub-social insect, after interactions between the
nymphs and their mother (recalling those ob-
served in Coelotes terrestris), the nymphs are
provided with miniature eggs, that the authors
liken to the trophic eggs of ants (Brian 1953;
Wilson 1971). Actually, another mode of ma-
ternal feeding, not unlike what is here reported
in Coelotes terrestris, has recently been observed
in our laboratory (Tahiri et al. 1989). In two
species of the genus Amaurobius, A. ferox and A.
fenestralis, about 3 days after the emergence of
the offsprings, the mother lays an egg mass de-
prived of any silken envelope which is imme-
diately eaten by the spiderlings. This egg-laying
appears to be systematic and to represent the
only source of food for the young during the first
ten post-emergence days. These findings, togeth-
er with those of spiderlings feeding on eggs inside
the egg sac (Schick 1972; Valerio 1974), suggest
that egg cannibalism could play a significant role
in the reproductive strategies of many spider spe-
cies (see Polis 1981).

ACKNOWLEDGMENTS

We express our thanks to H. Gouth, D. Assi
and D. Moncotel for their assistance in this re-

GUNDERMANN ET AL.- MATERNAL FOOD TRANSFER

101

search. We are grateful to A. L. Rypstra and G.

W. Uetz for their criticism and advice on the

manuscript.

REFERENCES CITED

Brach, V. 1976. Subsocial behavior in the funnel-
web wolf spider Sosippus floridanus (Araneae; Ly-
cosidae). Florida EntomoL, 59:225-229.

Brian, V. 1953. Oviposition by workers of the ant
Myrmica. Physiol. Comp. OecoL, 3:26-35.

Bristowe, W. S. 1958. The World of Spiders. Collins,
London.

Cloudsley-Thompson, J. L. 1955. The life histories
of the British cribellate spiders of the genus Ciniflo
Bl. (Dyctinidae). Ann. Mag. Natur. Hist., 12:787-
794.

Collatz, K. G. 1987. Structure and function of the
digestive tract. Pp 229-238, In Ecophysiology of
Spiders (N. Nentwig ed.). Springer-Verlag, Berlin.

Foelix, R. F. 1982. Biology of Spiders. Harvard Uni-
versity Press, Cambridge, Massachusetts.

Gundermann, J. L. 1989. Etudes sur le comporte-
mente matemel et son implication dans les pheno-
menes sub-sociaux chez Faraignee Coelotes terrestris
(Wider). PhD, Universite Nancy I.

Gundermann, J. L., A. Horel & B. Krafft. 1988. Ma-
ternal food-supply and its regulation in Coelotes ter-
restris {Arme&e, Agelenidae). Behaviour, 107:278-
296.

Hirschberg, D. 1969. Beitrage zur Biologic, insbe-
sondere zur Brutpflege einiger Theridiiden. Z. wiss.
Biol., 179:189-252.

Krafft, B., A. Horel & J. M. Julita. 1986. Influence
of food-supply on the duration of the gregarious
phase of a maternal-social spider Coelotes terrestris
(Araneae, Agelenidae). J. ArachnoL, 14:219-226.

Kullmann, E. J. 1972. Evolution of social behavior
in spiders (Araneae; Eresidae and Theridiidae). Am.
Zool., 12:419-426.

Kullmann, E. J. & W. Zimmermann. 1974. Regur-

gitationfiitterungen als Bestandteil fur Brutfiirsorge
bei Haubennetz und Rohrenspinnen (Araneae,
Theridiidae und Eresidae). Proc. 6’" Intern. Arachn.
Cong., pp. 120-124.

Locket, G. H. 1 926. Observations on the mating hab-
its of some web-spinning spiders. Proc. Zool. Soc.
London, pp. 1125-1146.

Polis, G. A. 1 984. Intraspecific predation and “infant
killing” among invertebrates. Pp 87-104, In Infan-
ticide: Comparative and Evolutionary Perspectives
(G. Hausfater & S. B. Hrdy, eds.). Aldine, New-
York.

Schick, R. X. 1972. The early instars, larval feeding
and the significance of larval feeding in the crab
spider genus Misumenops (Araneidae, Thomisidae).
Notes Arachnol.Southwest., 3:12-19.

Seibt, U. & W. Wickler. 1987. Gerontophagy versus
cannibalism in the social spider Stegodyphus mi-
mosarum Paresi and Stegodyphus dumicola Pocock.
Anim. Behav., 35:1903-1904.

Tahiri, A., A. Horel & B. Krafft. 1989. Etude preli-
minaire sur les interactions mere-jeunes chez deux
espdces d'Amaurobius (Araneae, Amaurobiidae).
Rev. ArachnoL, 8:1 15-128.

Tretzel, E. 1961. Biologic, Oekologie und Brutpflege
von Coelotes terrestris (Wider) (Araneae, Ageleni-
dae). II Brutpflege. Z. Morph. Oekol. Tiere, 50:375-
542.

Valerio, C. E. 1974. Feeding on eggs by spiderlings
of Achaearanea tepidariorum (Araneae, Theridi-
idae), and the significance of the quiescent instar in
spiders. J. ArachnoL, 2:57-63.

West, M. J. and R. D. Alexander. 1963. Sub-social
behavior in a burrowing cricket, Anurogryllus mu-
ticus (De Geer). Ohio J. Science, 63:19-24.

Wilson, E. O. 1971. The Insect Societies. Harvard
University Press, Cambridge, Massachusetts.

Manuscript received November 1989, revised Septem-
ber 1990.

1991. The Journal of Arachnology 19:102-104

ON THE SPIDER GENUS FEDOTOVIA
(ARANEAE, GNAPHOSIDAE)

Vladimir 1. Ovtsharenko; Zoological Institute, Academy of Sciences, Universitetskaja
emb. 1, Leningrad 199034 USSR

Norman I. Platnick: Department of Entomology, American Museum of Natural History,
Central Park West at 79th Street, New York, New York 10024 USA

Abstract. Fedotovia Charitonov is removed from the synonymy of the laroniine genus
Eilica and placed as a valid member of the subfamily Gnaphosinae. Males of the type
species, F. uzbekistanica Charitonov, are described for the first time, and the species is
newly recorded from Kazakhstan and Mongolia.

The spider genus Fedotovia was established by
Charitonov (1946) for a single female from Ish-
kent, Uzbekistan, described as F. uzbekistanica.
In his original description, Charitonov provided
no comments on the possible relationships or
subfamilial placement of Fedotovia, but later
(Charitonov 1969) he placed the genus in the
“Gnaphoseae,” comparing it with Asemesthes
Simon and Gnaphosoides Hogg. The latter name
was synonymized with the laroniine gnaphosid
genus Eilica Keyserling by Platnick (1975); Plat-
nick and Shadab (1981) subsequently placed Fe-
dotovia as a synonym of Eilica as well. That syn-
onymy, however, was not based on an
examination of the type specimen, but only on
figures of the eye pattern and epigynum of F.
uzbekistanica provided by Charitonov (1946).

We have recently had the opportunity to ex-
amine Charitonov’s type, as well as newly col-
lected specimens, which indicate that despite the
numerous striking epigynal similarities with Ei-
lica (including the shape of the lateral epigynal
margins, the twisted posterior epigynal ducts, and
the coiled median epigynal ducts), Fedotovia is
neither a synonym of Eilica nor a member of the
Laroniinae. It lacks the retromarginal cheliceral
laminae characteristic of that subfamily and has
instead the retromarginal serrated keel charac-
teristic of the subfamily Gnaphosinae. Males,
newly described here, have an elaborate palpal
embolus unlike those of previously described
gnaphosine genera, and we therefore consider
dotovia a valid genus of the Gnaphosinae.

The removal of Fedotovia from the list of ge-
neric synonyms of Eilica (which thus includes

only Baeriella Simon, Caridrassus Bryant, Gna-
phosoides Hogg, Gytha Keyserling, and Laronia
Simon) returns the distribution of Eilica to a
more normal Gondwanan pattern, including only
species from Africa, India, Australia, and south-
ern parts of the Americas.

The format of the descriptions and abbrevia-
tions used follow those of Platnick and Shadab
(1975); all measurements are in mm.

Fedotovia Charitonov

Fedotovia Charitonov, 1946:24 (type species by orig-
inal designation Fedotovia uzbekistanica Charito-
nov, 1946); Charitonov, 1969:98.

Eilica: Platnick and Shadab, 1981:184 (in part).

Diagnosis.— Specimens of Fedotovia resemble
those of the African gnaphosine genus Ase-
mesthes in having a strongly recurved posterior
eye row, and median eyes that are much smaller
than the lateral pairs, and will therefore key out
to Asemesthes in the key by Dalmas (1921). Males
can be distinguished from those of Asemesthes
by the shorter tibial apophysis and the greatly
elongated embolus (Figs. 1-3), females by the
median epigynal scape (Fig. 4).

Description.— Total length 5 — 7. Carapace al-
most triangular in dorsal view, widest between
coxae II and III, flattened, smoothly narrowed
opposite palpal insertions, light brown with
darkened posterior margin and rows of stiff setae
radiating from thoracic groove. Cephalic area not
elevated, thoracic groove short, straight, longi-
tudinal. From front, anterior eye row straight,
posterior row recurved; from above, anterior row

102

OVTSHARENKO & PLATNICK-F££><9rOF/^ (GNAPHOSIDAE)

103

Figures 1-5. —Fedotovia uzbekistanica Charitonov; 1, left male palp, prolateral view; 2 same, ventral view;
3, same, retrolateral view; 4, epigynum, ventral view; 5, same, dorsal view.

slightly recurved, posterior row strongly re-
curved. PME flattened, irregularly triangular,
AME circular, ALE and PLE oval; PME and
AME subequal, much smaller than lateral eyes;
PME separated by less than their diameter, by
more than their diameter from PLE; AME sep-
arated by about their diameter, about as far from
ALE; MOQ slightly wider in back than in front,
longer than wide. Clypeal height roughly equal
to ALE width. Chelicerae with serrated keel on
retromargin bearing four or five teeth; promargin
with two closely set teeth. Endites obliquely de-
pressed, distally rounded but not as convergent
as in Gnaphosa, with anterolateral serrula. La-
bium long, extending two-thirds of endite length.
Sternum light brown with darkened, rebordered
margins and tiny sclerotized extensions to and
between coxae. Leg formula 4123. Legs light
brown, bearing numerous spines. Tarsi with two
elongate claws dentate only at their base, without
claw tufts; tarsi and distal portions of metatarsi
I and II with thick, dark scopula; metatarsi with-

out preening combs; trochanters unnotched. Ab-
domen grayish brown, males with small, shiny,
triangular anterior scutum. Six spinnerets; an-
terior laterals with six or seven long piriform
gland spigots set posteriorly from major ampul-
late gland spigots; posterior medians short, tu-
bular in males but widened and reflexed ante-
riorly in females. Male palp with short tibial
apophysis shifted dorsally, strong, distally hook-
shaped median apophysis, and elaborately coil-
ing embolus. Epigynum with pair of longitudinal
lateral margins and median scape; epigynal ducts
twisted, transversely oriented along posterior
epigynal margin, in large circular coils medially.

Distribution. —Known only from arid habitats
in the USSR (Uzbekistan, Kazakhstan) and
Mongolia.

Fedotovia uzbekistanica Charitonov
Figs. 1-5

Fedotovia uzbekistanica Chanionov, 1946:24, figs. 31,
32 (female holotype from Ishkent, Uzbekistan, USSR,

104

THE JOURNAL OF ARACHNOLOGY

in University of Perm, examined); Charitonov, 1969:

98.

Diagnosis.— With the characters of the genus
and genitalia as in Figures 1-5.

Male.— Total length 6.1 1. Carapace 2.81 long,
2.18 wide. Femur II 2.10 long. Eye sizes and
interdistances: AME 0.05, ALE 0.14, PME 0.12,
PLE 0.17; AME-AME 0.12, AME-ALE 0.05,
PME-PME 0.06, PME-PLE 0. 1 1 , ALE-PLE 0.13;
MOQ length 0.33, front width 0.22, back width
0.30. Leg spination (only surfaces bearing spines
listed): femora: I, II dl-1-0, pO-M; III, IVdl-
1-0, pO-M, rO-1-1; patellae III, IV pO-1-0, rO-

1- 0; tibiae: I v2-2-2; II pO-0-1, vlr-2-2; III dl-
0-0, p2-l-l, V2-2-2, r2-l-l; IV dl-0-1, p2-l-l,
v2-2-2, r2-l-l; metatarsi: I, II v2-0-0; III dl-0-
0, pl-2-2, V2-2-2, rl-1-2; IV dl-0-0, pl-2-2, v2-
lp-2, r 1-2-2. Palpal embolus originating basally,
extending over retrolateral surface of cymbium,
curling back to ventral surface of bulb distally
(Figs. 1-3).

Female.— Total length 5.67. Carapace 2.44
long, 2.03 wide. Femur II 1.75 long. Eye sizes
and interdistances: AME 0.08, ALE 0.17, PME
0.12, PLE 0.13; AME-AME 0.10, AME-ALE
0.04, PME-PME 0.03, PME-PLE 0. 1 3, ALE-PLE
0.09; MOQ length 0.30, front width 0.26, back
width 0.27. Leg spination as in male, except as
noted: patellae III, IV pO-0-0; tibiae: II pO-0-0;
III pi- 1-1; IV dl-0-0; metatarsus III dO-0-0, pO-

2- 2. Median epigynal scape situated between lat-
eral epigynal margins (Fig. 4); epigynal ducts ir-
regularly twisted posteriorly, coiled medially (Fig.
5).

Specimens examined (in Zoological Institute,
Leningrad, unless otherwise indicated).— USSR:
Kazakhstan: Chimkentskaya: Aksumbe, Kara-
tau Mountains, Suzakskii region, 16 June 1989
(A. A. Zyuzin), 5 females. Gurievskaya: Baskar-
gan Steppe, Ustyurt Plateau, Ustyurt Reserva-

tion, 15-28 May 1989 (A. A. Raikhanov, S. I.
Ibraev), 2 males, 6 females. Kzil-Orda: Dzhu-
samly, Karmakchinskii region, clay desert, 19
June 1989 (A. A. Zyuzin), 1 female. Uzbekistan;
Yakkabag: Ishkent, elev. 1 100-1300 m, 26 June
1942 (D. M. Fedotov), 1 female (holotype. Uni-
versity of Perm). MONGOLIA: Bayan Khongor
Aimak; Ekhingol, 30 Aug. 1977 (K. Monkhbay-
ar), 1 female.

ACKNOWLEDGMENTS

The work reported on here was done while the
second author was a guest of the Zoological In-
stitute, Akademia Nauk, Leningrad, and that
support is greatly appreciated. We thank A. S.
Utochkin and N. M. Pakhorukov of the Uni-
versity of Perm for access to type material, and
M. U. Shadab of the American Museum of Nat-
ural History for assistance with illustrations.

LITERATURE CITED

Charitonov, D. E. 1946. New forms of spiders of the
USSR. Izv. Est.-Nauchn. Inst. Molotovsk. Univ.,
12:19-32.

Charitonov, D. E. 1969. Material’i k faune paukov
SSR. Uchel. Zap. Permsk. Ord. Trudy. ICrasn. Znam.
Gos. Univ. Ime A. M. Gorkogo, 179:59-133.
Dalmas, R. de. 1921. Monographic des araignees de
la section des Pterotricha (Aran. Gnaphosidae). Ann.
Soc. Entomol. France, 89:233-328.

Platnick, N. I. 1975. A revision of the spider genus
Eilica (Araneae, Gnaphosidae). Amer. Mus. Novi-
tates, 2578:1-19.

Platnick, N. 1. & M. U. Shadab. 1975. A revision of
the spider genus Gnaphosa (Araneae, Gnaphosidae)
in America. Bull. Amer. Mus. Nat. Hist., 155:1-66.
Platnick, N. I. & M. U. Shadab. 1981. On the spider
genus Eilica (Araneae, Gnaphosidae). Bull. Amer.
Mus. Nat. Hist., 170:183-188.

Manuscript received August 1990, revised October 1990.

1991. The Journal of Arachnology 19:105-1 10

THE LIFE HISTORY OF EUSCORPIUS FLAVICAUDIS
(SCORPIONES, CHACTIDAE)

T. G. Benton: Department of Zoology, University of Cambridge, Downing St., Cam-
bridge, CB2 3EJ, United Kingdom

Abstract. The life-history of an English population of the scorpion, Euscorpius flavi-
caudis, was studied using morphometric measures of over 300 specimens. The study hy-
pothesizes seven instars. Evidence strongly suggests that there are two instars of adults:
some males and females mature at the sixth instar and some at the seventh. Larger scorpions
may have a higher per-season reproductive success but may have fewer reproductive seasons
owing to the extra time needed to mature.

Studies of scorpion life-histories face several
major difficulties. Firstly, scorpions are relatively
long-lived compared to other terrestrial arthro-
pods. For example, the Australian scorpion Uro-
dacus yaschenkoi (Birula) grows to maturity in
four years and may live an additional six (Short-
house & Marples 1982). Secondly, scorpions can
be very difficult to rear in the laboratory, often
suffering high mortality at molting (Francke &
Sissom 1984; Williams 1987; pers. obs.). Addi-
tionally, laboratory conditions may differ suffi-
ciently from those in the field to produce inac-
curate data. For example, Smith (1966) found
that the duration of the first instar of U. mani-
catus (Thorell) was strongly influenced by the
temperature at which the animals were main-
tained. Field studies of scorpion life-history can
be equally problematical, as marked individuals
lose their marks with each molt.

Many studies have resorted to estimating the
number of instars from measurements of many
specimens taken in the field (Polls & Sissom
1990). This method is approximate, but allows
the tentative calculation of the life-history pa-
rameters of the scorpion under study.

The chactid scorpion, Euscorpius flavicaudis
(de Geer) is a widespread southern European spe-
cies that has successfully colonized a port in
southern England. This colony at Sheemess, Kent
(5 1°26'N: 0°45'E), has existed for about 1 20 years
(Benton in press a). Over a two year period the
ecology and behavioral ecology of this species
was studied (Benton 1 990) and enough data were
collected to allow investigation of the life-history
of E. flavicaudis. The prime purpose of this paper
is to present data which strongly suggest that

there is more than one instar of sexually mature
adult in this population. The concurrent studies
allow, for the first time, the evolutionary signif-
icance of a scorpion life-history polymorphism
to be discussed.

METHODS

Between September 1987 and September 1989,
a total of 3 1 7 specimens of Euscorpius flavicaudis
were measured. Two methods of measuring were
used: (1) 208 live animals were measured in the
laboratory using a binocular microscope with op-
tical micrometer, and (2) 109 animals were mea-
sured in the field using dial calipers accurate to
0.1 mm. These latter scorpions were subdued
mechanically by using a modified 50 ml syringe.
The nozzle end was replaced by a fine nylon mesh
and the plunger was sheathed in cotton-wool.
Scorpions were placed in the barrel of the syringe
and pressed against the mesh. A small hole was
cut through the mesh to allow a pedipalp to be
pulled out. Thus immobilized, the pedipalp could
be measured quickly and accurately without
damaging the animal.

Most of the scorpions measured were adults
or sub-adults {N = 287). Measurements made
were: length and width of pedipalpal chela, length
of prosoma (taken along midline), width of ster-
nite V, weight and approximate whole-body
length. In some analyses, the two measures of
chela size were combined as the Chela-size Index
(CSI). CSI is used in preference to chela length
alone, as the ecological importance of chela size
is most likely related to their strength (Benton
in press b) and therefore volume. This is better
approximated by the CSI.

105

106

THE JOURNAL OF ARACHNOLOGY

Figure 1.— A log-log bivariate morphometric plot,
showing the 5 juvenile instars in the scorpion, Eus-
corpiusjlavicaudis(N= 163). Instars determined either
by morphological differences (e.g., first and second in-
stars) or cluster statistics (adults and fifth instars). White
squares = first instar; black squares = second; cross =
fourth; black triangle = fifth; white circle = adult fe-
males; black circle = adult males. Regression equation
of this relationship (including all instars other than 1 st):
3;= 1.139X - 0.079, = 0.974, A = 155.

The apparent objectivity of using cluster sta-
tistics in taxonomy has been frequently criticized
(e.g., Ridley 1986), so in my analyses I only use
cluster statistics to clump data when the number
of clusters has been determined, or hypothesized,
by some other method. The method of clustering
was to use “average neighbor distance” (see No-
rusis 1985) using the package SPSSx (SPSS Inc.,
Chicago, IL 60611). As two different methods of
measuring the scorpions were used, it is likely
that their associatied errors were different. Due
to this, in each analysis, care was taken to use
only those scorpions that had been measured us-
ing the same method.

RESULTS

Number of juvenile instars.— The bivariate
morphometric plot of prosomal carapace length
vs. width of the stemite on mesosomal segment
V, using all available data, reveals four clumps
of scorpions other than the adults (Fig. 1). These
clumps correspond to the first, second, fourth
and fifth instars. No third instars were measured
as a result of sampling error. Progression factors
for the juvenile molts were calculated (Table 1),
assuming that the third instar would clump mid-
way between the second and fourth. Data are
only available on the masses of the first and sec-
ond instars: 0.0084 ± 0.0025 g {N = 10) and
0.0070 ± 0.0013 g (iV = 5) respectively. The

Figure 2.— Relationship between chela-size index
(CSI = length of chela x width) and prosoma length
in adult males. Data have been separated into two
instars (see text), sixth (circles) and seventh (squares).
Regression line fits the relationship in the sixth instar:
seventh instar males have allometrically larger chelae.
Regression line: y = I2.9x - 37.1, = 0.88. Males

known to have deposited a spermatophore (filled sym-
bols) occur in both instars.

reduction in mass between the first and second
instar is due to the mass of the exuvium.

Number of adult male instars.— Sexually ma-
ture males can be distinguished on the basis of
two secondary sexual characteristics: the posses-
sion of a notch in the pedipalpal fingers and their
relatively longer pectines. The size-range of these
adult males is large. However, the largest male
and the second smallest deposited spermato-
phores, confirming their sexual maturity (Fig. 2).

During the course of field observations it be-
came apparent that a dimorphism in the size of
the pedipalpal chelae of adult males existed. Some
males had chelae comparable in size to adult
females; whereas others had disproportionately
large chelae (mean adult male CSI = 42.5 ± 9.2
mm^, range 24.3-66.6, N = 113; mean adult fe-
male CSI = 33.6 ± 7.3 mm^, range 18.6-51.4,
N = 78). To investigate this apparent dimor-
phism, the frequency distribution of chela lengths
was compared to the expected normal distribu-
tion. The observed distribution of chela sizes dif-
fers significantly from normal (x^ = 21.1, df= 7,
P < 0.001; x^ test of normality, Zar 1984:89)
(Fig. 3). The distribution is, in fact, bimodal,
indicating the population of adult males occurs
in two size-classes for chela length. The distri-
bution of prosoma lengths of these 1 1 4 males
also differs from normality (x^ = 17.2, df= 7, N
= 80, P < 0.025).

Having determined the existence of two size-

BENTON -LIFE-HISTORY OF EUSCORPIUS FLA VICAUDIS

107

Claw length (mm)

Figure 3.— The observed (filled bars) and expected (stippled bars) distributions of chela-lengths of 1 14 male
scorpions from Sheemess. The expected distribution is a normal distribution given the mean and standard
deviation of the observed. The distributions differ significantly (x^ = 21.1, 7, F < 0.001). The two modes

of the observed distribution correspond to two instars.

classes, cluster analysis was conducted on 66
males for which measures of prosoma length,
chela width and chela length were available, in
order to divide the adults more exactly. Figure
2 shows that the analysis splits the males into
two groups based on different relationships be-
tween the prosoma length and the chela-size in-

dex. Large males have disproportionately large
chelae.

The determination of the boundaries of the
clusters by the above analysis allows calculation
of the average weights and body size (prosoma
-I- mesosoma, measured as one unit). Large males
are, on average, 145% heavier than small males

Table 1 . — Progression factors (PFs) in E. flavicaudis. A PF is the multiplicative increase in dimensions between
instars. The first instar is atypical due to its incompletely sclerotized exoskeleton. The mean PFs for the other
instars are 1.30 ± 0.09 (prosoma length) and 1.34 ± 0.1 1 (width of stemite V). As there are no data for the
third instar, the PFs for the second-third and third-fourth molts are estimated by taking the square-root of the
increase between the second and fourth instars. Final instars often show a lower PF than earlier instars (e.g.,
Polis & Farley 1979).

Prosoma length

Progression

Width

Progression

Instar

[mm ± SD («)]

factor

[mm ± SD (n)]

factor

1st

1.10 ± 0.14(8)

1.50

1.34^

1.30

1.32

1.20

1.24 ± 0.24 (8)

1.16

1.442

1.40

1.43

1.22

2nd

1.65 ± 0.04 (12)

1.42 ± 0.08 (12)

4th

2.96 ± 0.14 (4)

2.95 ± 0.24 (4)

5th

3.86 ± 0.26 (8)

4.14 ± 0.40 (8)

Female 6th

5.11 ± 0.37 (20)

5.94 ± 0.59 (20)

Female 7th

6.15 ± 0.38 (19)

7.24 ± 0.45 (19)

5th

Male 6th

5.47 ± 0.39 (38)

1.42

1.17

5.34 ± 0.56 (49)

1.29

1.20

Male 7th

6.39 ± 0.28 (28)

6.53 ± 0.35 (30)

108

THE JOURNAL OF ARACHNOLOGY

(0.658 ± 0.124 g cf. 0.428 ± 0.115 g; Mann-
Whitney {720,23 = 36.5, P = 0.0001), and their
body length averages 117% longer (18.9 ± 1.6
mm cf. 16.2 ± 1.7 mm; Mann-Whitney Ujoii =
45, P = 0.0001).

Perhaps the most parsimonious explanation
for this dimorphism is that two instars of adult
males exist within the population. The mean
lengths of the males in each cluster allow the
calculation of putative progression factors (Table
1).

Number of adult female instars. —As with
males, females known to be mature (because they
gave birth in the laboratory) show a wide range
in size (Fig. 4). Animals assumed to be adult
females have a larger range of sizes than that of
adult males (prosoma length 4.46-7.22 mm of
5.07-7.14 mm in males). However no obvious
dimorphism is apparent within females, as they
do not have an allometric relationship between
chela size and body size. For all available vari-
ables there are no significant departures from
normality in the frequency distributions (x^ =
5 .1 , df = 1 , N = P > 0.50 for chela length;

= 10.26, df= 1 , N = 19, P > 0.10 for chela
width; x' = 2.1 1, #= 6, iV = 40, P > 0.90 for
prosoma length). It is however possible that there
are two instars of adult females, but the fre-
quency distributions overlap more than males
making females difficult to separate.

To investigate this further, cluster statistics
were used to divide adult females into two size-
classes (using the variables prosoma length and
CSI). The means of the clumps were then used
to calculate hypothetical progression factors (Ta-
ble 1) which agree well with what would be pre-
dicted if there were two instars.

Resightings of marked individuals in the field
suggest that adults can live for at least two years
(Benton in press a) and growth to maturity may
take three or more years. As post-maturation
molts are unknown in scorpions (Polis & Sissom
1990), adult sixth instars do not become adult
seventh instars, so we might expect females ma-
turing at the seventh instar to have a reproduc-
tive benefit. A higher per-season reproductive
success is the most obvious reason that natural
selection may favor the delay of female maturity
for several months and an extra instar. Under
laboratory conditions built to mimic closely the
scorpions’ habitat in England, 16 females gave
birth to viable broods (Benton in press b). Within
these females (where data are available) a sig-

Figure 4.— Relationship between chela-size index
(CSI = length of chela x width) and prosoma length
in adult females. Data have been separated into two
putative instars (see text), sixth (circles) and seventh
(squares). Females known to have given birth (filled
symbols) occur in both instars.

nificant relationship exists between female pro-
soma size and the weight of her brood {R^ = 0.36,
P, ,0 = 5.6, P < 0.05), and a positive relationship
exists between female size and number of young
(R2 = 0.23, F,,„ = 3.29, P = 0.10). Using the
size criteria, determined by the earlier cluster
analysis, to separate these females into two
groups, it can be shown that a “small” female
has on average 30 ± 5 young (range 26-38, N =
5) weighing 0.237 ± 0.048 g (range 0. 1 96-0.3 1 5,
N = 5); whereas a large female has 36 ± 9 young
(range 25-51, A = 10) weighing 0.327 ± 0.063
g (range 0.210-0.420, N = 9). The difference in
brood weights is significant (Mann-Whitney 6/5 9
= 6, P < 0.05), the difference in brood sizes is
not ({75,10= 15, P= 0.2).

DISCUSSION

From data reviewed in Francke and Sissom
(1984) the mean number of instars in scorpions
is 7.0 ± 1.2 (A = 57). In the present study there
appear also to be seven instars of Euscorpius
flavicaudis.

However, perhaps the most interesting aspect
of this life-history study is the evidence of finding
more than one instar of adults. The evidence for
this is that sexually mature males are dimorphic
in size. The means of the two size-classes are
separated by progression factors of the expected
size. If there were not two instars, a progression
factor of 1.48 (for increase in prosoma length)
for the fifth instar-adult male molt would be nec-
essary. This would be higher than that previously

BENTON-LIFE-HISTORY OF EUSCORPJUS FLA VICAUDIS

109

reported for scorpions (scorpion average =1.28
± 0.04, data from Polis & Farley 1979). Sexually
mature females do not occur in two size-classes,
but the range of female size is greater than for
males. It was therefore hypothesized that two
instars of adult female existed. Cluster statistics
were then used to divide the range of adult fe-
males into two groups. To test the hypothesis,
two predictions were made: (1) progression fac-
tors for the two groups should be within the nor-
mal range for scorpions, and (2) larger females
should have a greater reproductive success per
breeding attempt. Qualitatively, both these pre-
dictions were supported by data.

Life-history polymorphism is not unusual
among scorpions. Francke and Sissom (1984) list
32 species for which the sexually mature instars
had been determined. Of these species 47% (15)
showed more than one adult instar (three species
where only females mature at more than one
instar, six only males and six both). The only
record of a chactid scorpion with a life-history
polymorphism is for E. italicus, where Anger-
mann (1957) reports two instars of adult females.

No cases are known where scorpions continue
to molt after they mature (Polis & Sissom 1990),
so sixth instar adult scorpions are unlikely to
molt a seventh time. Thus, the sixth instar of
this population of E. Jlavicaudis consists of adult
males and females, and sub-adult males and fe-
males. Adult males are immediately distinguish-
able owing to their notch in the pedipalpal fin-
gers. However, the other categories of sixth instar
are not easily distinguishable. Unfortunately the
significance of the range of adult sizes was not
realized until after the field study of this popu-
lation had finished. In addition, the population
is small and has a threatened status (Benton in
press a) so I was unwilling to collect a large num-
ber of specimens for dissection. In total, 25 fe-
males (sixth and seventh instars) were preserved,
and on closer inspection one of these was found
to be a male without the obvious secondary sex-
ual characteristics: he was a sub-adult. The pro-
soma length of this individual was 4.73 mm, with
a CSI of 24.3 mm^, putting it within the range
of “adult females” (Fig. 4). Further, this sub-
adult male has an average progression factor of
1.28 when compared to the average dimensions
of the fifth instar scorpions which is within the
normal range in this species. This makes it much
more likely that this specimen is a small sixth
instar male rather than a very large fifth instar.

Given the state of preservation of the specimens
it was not possible to determine whether there
were sixth instar immature females amongst
them.

Although the observed life-history polymor-
phism in scorpions has previously been reported,
for the first time an attempt can be made to
explain its evolutionary significance. Seventh in-
star males are, on average, 1 .20 times larger than
sixth instar males; and seventh instar females
1.21 times larger than sixth (average of the pro-
gression factors for prosoma length, chela length,
chela width and body width). Behavioral and
ecological correlates of size do exist. In this spe-
cies, the scorpion’s primary offensive and defen-
sive weapons are its chelae. Chela size predicts
the outcome of fights between males for female-
occupied burrows during the mating season, and
large-clawed males can “persuade” otherwise
unwilling females to mate (Benton in press b). It
is interesting, therefore, that chela-size increases
disproportionately to body size between the sixth
and seventh instar adult males. Seventh instar
males, therefore, have a higher per-season re-
productive success than sixth instar males. In
females, the benefit of being large seems also to
be a greater reproductive success, in that large
females have heavier (and more) offspring than
smaller females. In Paruroctonus mesaensis rate
of food intake was shown to have a significant
effect on brood weights (Polis & McCormick
1987). This may result from the ability to in-
crease food intake rate, by capturing larger prey.
Larger scorpions win cannibalistic contests (Polis
1980; Benton 1990), so may obtain more food
in this way.

If scorpions have a higher per-season repro-
ductive success as seventh instar adults, why do
any mature at the sixth instar? The answer to
this must lie in the fact that seventh instar scor-
pions delay their maturity. In the study popu-
lation in England, females give birth in late sum-
mer (Benton in press b) and the first instar lasts
7 days (Benton 1991). Second instars overwinter
and molt the following spring to the third instar.
The duration of subsequent instars is unknown,
but laboratory studies uniformly show that instar
duration increases as scorpions age (W. D. Sis-
som, pers. commun.), so it is plausible that sev-
enth instar adults, by delaying maturity, miss one
mating season. Adult longevity is difficult to as-
sess, but the longest period between resightings
of an adult (male) in the field is 23 months. Lon-

110

THE JOURNAL OF ARACHNOLOGY

gevity for two or more mating seasons is prob-
able.

Two maturation strategies exist in this popu-
lation of scorpions: to mature at the sixth instar,
and be small, or to delay maturity until the sev-
enth instar and be large. It is possible that the
two strategies’ payoffs are frequency dependent,
and occur at an equilibrium value— the Evolu-
tionarily Stable Strategy (ESS, Maynard Smith
1 982). This subject is explored further in Benton
(in press b).

ACKNOWLEDGMENTS

Thanks are due to the Medway Ports Authority
for permission to work at Sheemess. M. Evans,
G. Miller, G. Polls, and W. D. Sissom com-
mented on the manuscript, as did S. Benton who
was a continual support during the project. The
project was funded by an SERC studentship, un-
der the supervision of W. A. Foster.

LITERATURE CITED

Angermann, H. 1957. Uber Verhalten, Spermato-
phorenbildung und Sinniesphysiologie von Euscor-
pius italicus Herbst und verwandten Arten. Z. Tier-
psychol., 14:276-302.

Benton, T. G. 1990. The Behavior and Ecology of
Scorpions. Ph.D. thesis, University of Cambridge,
UK.

Benton, T. G. 1991. Reproduction and parental care
in the scorpion Euscorpius flavicaudis. Behaviour,
117:20-28.

Benton, T. G. in press a. The ecology of the scorpion,
Euscorpius flavicaudis, in England. J. Zool. (Lon-
don).

Benton, T. G. in press b. Determinants of male mat-
ing success in a scorpion. Anim. Behav.

Francke, O. F. & W. D. Sissom. 1984. Comparative
review of the methods used to determine the num-

ber of molts to maturity in scorpions (Arachnida),
with analysis of the post-birth development of Vae-
jovis coahuilae Williams (Vaejovidae). J. Arachnol.,
12:1-20.

Maynard Smith, J. 1982. Evolution and the Theory
of Games. Cambridge Univ. Press, Cambridge.

Norusis, M. J. 1985. SPSSx Advanced Statistics
Guide. McGraw-Hill Book Co., New York.

Polis, G. A. 1980. The significance of cannibalism
on the demography and activity in a natural pop-
ulation of desert scorpions. Behav. Ecol. Sociobiol.,
7:25-35.

Polis, G. A. & R. D. Farley. 1979. Characteristics
and environmental determinants of natality, growth
and maturity in a natural population of the desert
scorpion, Paruroctonus mesaensis (Scorpionida:
Vaejovide). J. Zool., 187:517-542.

Polis, G. A. & S. J. McCormick. 1987. Intraguild
predation and competition among desert scorpions.
Ecology, 68:332-343.

Polis, G. A. & W. D. Sissom. 1 990. Life History. Pp.
161-223, In The Biology of Scorpions (G. A. Polis,
ed.). Stanford Univ. Press, Stanford, CA.

Ridley, M. 1986. Evolution and Classification: The
Reformation of Cladism. Longman, London.

Shorthouse, D. J. & T. G. Marples. 1982. The life
stages and population dynamics of an arid zone
scorpion Urodacus yaschenkoi (Birula 1903). Aus-
tralian J. Ecol., 7:109-118.

Smith, G. T. 1966. Observations on the life-history
of the scorpion Urodacus abruptus Pocock (Scor-
pionidae) and an analysis of its home sites. Austra-
lian J. Zool., 14:383-398.

Williams, S. C. 1987. Scorpion bionomics. Ann. Rev.
Entomol., 32:275-295.

Zar, J. H. 1984. Biostatistical Analysis. 2nd Ed. Pren-
tice-Hall Inc., Englewood Cliffs, NJ.

Manuscript received October 1990, revised December
1990.

1991. The Journal of Arachnology 19:1 1 1-1 14

HOMING BY CRAB SPIDERS MISUMENA VATIA
(ARANEAE, THOMISIDAE) SEPARATED
FROM THEIR NESTS

Douglass H. Morse: Graduate Program in Ecology and Evolutionary Biology, Division
of Biology and Medicine, Brown University, Providence, Rhode Island 02912 USA

Abstract. Nest-guarding female crab spiders Misumena vatia sometimes become dis-
placed from their nests on milkweed leaves. Experimentally displaced individuals usually
found their way back to their nests if put at the bottom of the stem containing their nest,
even though they had no silken lines to guide them. In repeat runs they performed similarly,
although returning more rapidly than in the initial runs. Spiders displaced several cm from
their nests recruited to them much less successfully than spiders at the base of the stem.

Finding lost nests may be important because
in unguarded ones.

The advent of parental care presents the par-
ents with a number of problems. One is to pro-
vide for their own young, because an individual
typically negates its fitness by tending unrelated
offspring. Solutions to this problem might either
be direct, as in identifying one’s own young, or
indirect, as in locating a rearing site, such as a
nest (Beer 1970; White 1971). If the parents pro-
vide extended care, they may have to forage for
themselves or for their young. Hunting for food
may present another problem, returning to the
site at which offspring have been left. Many of
the classic homing experiments test returns to
nest sites, and both olfactory and visual cues
have been implicated (e.g., Tinbergen 1951; von
Frisch 1967). As an alternative to recognition of
young or site, parents may remain in contact with
offspring throughout the period of care. In the
latter instance, parents may not have a well-de-
veloped recognition of either their site or off-
spring (Morse 1989). If such animals become
accidentally displaced they may have only a lim-
ited ability to relocate their sites, which may have
serious consequences for their offsprings’ suc-
cess.

Crab spiders Misumena vatia (Clerck) provide
an opportunity to explore and test the responses
of displaced individuals that normally do not
stray from their nest site. They lay a single clutch
of eggs (Gertsch 1939; Morse 1988), which they
guard, often until the young emerge from the egg
masses nearly a month later (Morse 1987). How-
ever, some individuals disappear from the nest

more offspring survive in guarded nests than

sites before their young emerge. Probably about
half of these adults die of senescence, and the
other half leave, or, occasionally, are preyed upon
at the nests (Morse 1987). Some of the spiders
that leave appear to become accidentally sepa-
rated from their nest sites (Morse 1989). Dis-
placement might occur if they are attacked or
otherwise disturbed and drop from the nest with-
out laying down silken lines as they do at other
times. Occasionally the spiders resort to this be-
havior when handled by humans, and under nat-
ural circumstances ants may cause the spiders to
give this reaction (Morse 1989), although they
usually do not nest in the presence of aggressive
ants. As a result, the spiders contact the ground
stratum without a line to retrace to their previous
site. Without a line, their ability to find their nest
may be compromised. These matters are im-
portant to the spiders, because guarded nests are
more successful than unguarded ones (Morse
1987, 1988, in prep.). The present study tests the
ability of individuals separated from their nests
to return to them, both with and without pre-
viously deposited silken drag-lines.

METHODS

I carried out these studies in a field in Bremen,
Lincoln Co., Maine. I have described this area
in detail elsewhere (Morse 1979, 1981). During
the summers of 1 988 and 1 990 1 tested the ability
of post-reproductive brooding spiders to find nest
sites from which I had displaced them. Since nest
sites on nonflowering common milkweeds As-

111

112

THE JOURNAL OF ARACHNOLOGY

Table 1.— Performance of nest-guarding spiders placed at various sites, a = Results in 1988 and 1990 did not
differ significantly, so data pooled, b = This individual did not return to nest on either first or second run.
c = Does not include one individual that did not return, d = Data from 1990 only, e = Does not include 10
individuals that did not return.

Manipulation

N

Return

to nest

Did not

return

to nest

Time to return

N S ± SD

Placed at bottom of nest stem'*

30

26

4

14

40.0 ± 32.4

Placed at bottom of nest stem a second time'*

15

14

lb

14

27.6 ± 23.6^

Placed on substrate**

30

12

18

5

75.8 ± 42.3‘‘’=

depias syriaca L., the most commonly used lo-
cations in the study area, averaged over 50 cm
above the ground, and nests on flowering milk-
weeks over 80 cm (Morse 1985), cues to the pres-
ence of their nests may be obscure or nonexistent
from the substrate below.

I placed spiders that were about to lay (Morse
1988) on the upper leaves of nonflowering milk-
weed plants and then put cages (50 cm x 50 cm
X 150 cm) of 0.2 cm x 0.2 cm metal screening
over these plants. The spiders always lay at night,
usually one to three days after being placed on
those sites (Morse 1985). In this set of experi-
ments I only used spiders that laid on the night
following placement in the cages, thereby mini-
mizing the probability that they would produce
a silken thread between the substrate and the
nest. In observations made in this and other stud-
ies, spiders that laid on the night following re-
lease invariably remained on the leaves of the
upper parts of the plants, and all observed shifts
in site took place at that height. Several individ-
uals began to manipulate their nest leaf within a
few hours (see Morse 1985) and subsequently
confined their activity to that leaf Individuals
that moved to the substrate or the screening of
the cage laid on a subsequent night and were thus
not included.

I removed post-reproductive individuals (not
over four days after laying) from their nests and
released them in two different places: 1) at the
base of the stems on which their nests were placed,
and 2) one (1988) or two (1990) days later, in
the grassy substrate underneath these plants at
the outer edge of the area covered by the nest-
stem’s leaves. Distances of releases from the stems
of the spiders’ nest plants using the latter crite-
rion (lengths of longest leaves) averaged 12.1 ±
1.5 cm (measured only in 1990). Individuals
placed on the stems in 1990 were run on the

stems in the same way a second time one day
later. This manipulation provided a comparison
with the initial run, the difference being the pres-
ence of a silken thread on the stem. All of the
1990 spiders were placed in the substrate one
day after the second run on the stems to test their
ability at finding the stem of their nest plant.

In testing responses of spiders put at the bot-
tom of their nest plants, I assumed a 50:50 prob-
ability that they would move up the stems toward
their nests by chance. This was a reasonable as-
sumption, given that they were placed in contact
with the stem, and that approximately half of the
path directions available to them if they moved
forward would subtend the stem immediately in
front of them.

RESULTS

Brooding spiders displaced to the base of the
stem of their nest plant returned to their nests
more frequently than predicted by chance over
a two-hour period, assuming a 50:50 predicted
level of choice, as above (Table 1: Z = 3.83, P
< 0.001 in a one-tailed binomial test). Individ-
uals run a second time on the following day per-
formed equally accurately (Table 1). Further, they
returned to their nests significantly more rapidly
on the second day than the first (Table 1: T =
16'/2, Z = 2.027, P < 0.02 in a one-tailed Wil-
coxon signed ranks, matched pairs test). This
difference could have been a response to the
threads that the spiders laid down on the stems
during the previous day’s ascent. Since I initially
placed these individuals on the plants shortly
before they deposited their egg mass, they were
very unlikely to have had access to a line on their
first run.

These spiders were then placed on the sub-
strate under the outer extremity of the leaves of
the nest plant, but nearer to its stem than to any

MORSE-NEST HNDING BY CRAB SPIDERS

113

other milkweed stem. They returned to their nest
significantly less frequently from here than from
the base of the stem (Table 1: G = 14.92, df =
\, P < 0.001 in a G-test). Further, the four in-
dividuals that failed to find their nests from the
bottom of the stem in the first test (Table 1) also
failed to find their way to the stem in this test.

DISCUSSION

These spiders clearly have well-developed
abilities to respond to displacement on the nest
plant. Simply moving upward on the stems would
suffice if they drop to the base of the stem, be-
cause they will soon reach nest height, where
parent spiders almost invariably lay down lines
among the nest leaf, adjacent leaves, and stem
proper (Morse 1985). This architecture results
from the periodic movements that the spiders
make in the immediate vicinity of their nest site,
laying down lines in the process, as a result se-
curing the nest tightly to the surrounding vege-
tation. In many instances a line would naturally
extend from the bottom of the stem to near the
top, a result of the spider’s initial recruitment
onto the plant. However, if they moved from the
leaf of an adjacent plant, no such line would exist.
The present experiments attempted to eliminate
the question of initially using lines by placing the
spiders onto the sites just before they built their
nest. The shorter recruitment time on the second
runs suggested that a line, when present, hastened
movement to their nest, although it did not elim-
inate the possibility of experience playing a role.

The spiders responded significantly more
poorly to greater displacement, suggesting that
falling off a plant without using a line is a drastic
action. Spiders do not appear to take this option
frequently. Under natural conditions, aggressive
ants may prompt this response most often. The
only ants observed to attack the spiders {Formica
sp. L.) have a patchy distribution in the study
area, and the spiders do not appear to build their
nests at sites frequented by them. However, nest-
ing spiders may not always be able to avoid ants,
because if aphids recruit to plants after the spi-
ders choose their nest sites, ants may recruit in
turn in response to the aphids.

Although some of the spiders displaced in the
substrate might eventually have found their nests,
occasional observations of individuals naturally
displaced from their sites shortly after laying, and
showing no signs of hunting, suggest that these
spiders may frequently be unable to relocate their

nests. Four such individuals located on vegeta-
tion 30-100 cm from their nests for either two
or three days were placed back on their nests,
and three of them remained there for one day or
more, strongly suggesting that although the pro-
pensity to guard remained, they did not have the
ability to relocate their nests (Morse 1987). Giv-
en the demonstrated importance of guarding
(Morse 1988), even moderate periods of absence
from these nests may appreciably increase the
chance of failure.

The performance of these post-reproductive
spiders is profitably compared with results from
analogous experiments run on pre-reproductive
adult female spiders searching for hunting sites
on flowering milkweed stems (Morse in prog-
ress). In contrast to the 26 of 30 post-reproduc-
tive spiders finding their nest sites (Table 1), only
16 of 32 pre-reproductive individuals selected
flowering stems when they were placed on the
base of them {G = 10.05, df= \, P < 0.01 in a
G-test). These results suggest that the parents’
success in relocating their nest sites involved traits
missing or poorly developed in the pre-repro-
ductive individuals. In contrast, post-reproduc-
tive individuals placed on the substrate did not
differ in nest-finding success from pre-reproduc-
tive individuals finding stems with satisfactory
hunting sites on the flowering plants [12 of 30
post-reproductive individuals successful (Table
1), vs. 34 of 76 pre-reproductive spiders: G =
0.20, df=\,P> 0.5 in a G-test].

ACKNOWLEDGMENTS

I thank R. G. Gillespie, R. S. Souter, and J.
K. Waage for comments on the manuscript. E.
B. Noyce generously permitted use of her prop-
erty. K. Cha, L. Heller, B. Hyman, N. McKay,
and E. Morse assisted with the experiments. This
work was supported by the National Science
Foundation (BSR85- 16279 and BSR90-07722).

LITERATURE CITED

Beer, C. G. 1970. Individual recognition of voice in
the social behaviour of birds. Adv. Stud. Behav., 3:
27-74.

Gertsch, W. J. 1939. A revision of the typical crab-
spiders (Misumeninae) of America north of Mexico.
Bull. Amer. Mus. Nat. Hist., 76:277-442.

Morse, D. H. 1979. Prey capture by the crab spider
Misumena calycina (Araneae: Thomisidae). Oecol-
ogia, 39:309-319.

Morse, D. H. 1981. Prey capture by the crab spider
Misumena vatia (L.) (Thomisidae) on three com-

114

THE JOURNAL OF ARACHNOLOGY

mon native flowers. Amer. Midi. Natur., 105:358-
367.

Morse, D. H. 1985. Nests and nest-site selection of
the crab spider Misumena vatia (Araneae, Thom-
isidae) on milkweed. J. Arachnol., 13:383-390.

Morse, D. H. 1987. Attendance patterns, prey cap-
ture, changes in mass, and survival of crab spiders
Misumena vatia (Araneae, Thomisidae) guarding
their nests. J. Arachnol., 15:193-204.

Morse,D. H. 1988. Relationship between crab spider
Misumena vatia nesting success and earlier patch-
choice decisions. Ecology, 69:1970-1973.

Morse, D. H. 1989. Nest acceptance by the crab spi-

der Misumena vatia (Araneae, Thomisidae). J. Ar-
achnol., 17:49-57.

Tinbergen, N. 1951. The Study of Instinct. Clarendon
Press, Oxford, England.

von Frisch, K. 1967. The dance language and ori-
entation of bees. Belknap Press, Cambridge, Mas-
sachusetts.

White, S. J. 1971. Selective responsiveness by the
gannet {Sula bassana) to played-back calls. Anim.
Behav., 19:135-141.

Manuscript received November 1990, revised December
1990.

1991. The Journal of Arachnology 19:1 15-121

ON EURASIAN AND AMERICAN TALANITES
(ARANEAE, GNAPHOSIDAE)

Norman I. Platnick: Department of Entomology, American Museum of Natural History,
Central Park West at 79th Street, New York, New York 10024 USA

Vladimir I. Ovtsharenko; Zoological Institute, Academy of Sciences, Universitetskaja
emb. 1, Leningrad 199034 USSR

Abstract. The North American spider genus Rachodrassus Chamberlin is newly syn-
onymized with the Old World genus Talanites Simon. The type species, Talanites fervidus
Simon from Israel, is redescribed, and the species of Talanites occurring in the Soviet Union
are revised. Four new species are described: T. mikhailovi from Kazakhstan, T. dunini from
Azerbaijan and Turkmenia, and T. moodyae and T. ubicki from California.

The North American spiders of the genus
Rachodrassus were revised by Platnick and
Shadab (1976), who recognized three species as
valid: R. echinus Chamberlin and R. exlineae
Platnick and Shadab from the southeastern Unit-
ed States, and R. captiosus (Gertsch and Davis)
from southern Texas and northeastern Mexico.
We have recently had the opportunity to com-
pare representatives of these species with Eur-
asian taxa that have been placed in the genus
Talanites Simon, and have concluded that the
New and Old World taxa are congeneric. We
present here a redescription of the type species
of Talanites, T. fervidus Simon from Israel, along
with a revision of the Soviet fauna of the group
and a description of two additional American
species from California.

In an unpublished thesis, Penniman (1 985) ob-
served that specimens of Rachodrassus lack pre-
coxal sclerites (i.e., sclerotized extentions of the
sternal margin that reach toward, and sometimes
between, the coxae). Because he considered this
feature synapomorphic for a large group of gna-
phosoids and clubionoids, Penniman suggested
that Rachodrassus is misplaced as a gnaphosid.
Although we have observed tiny but distinct pre-
coxal sclerites in some species, we agree that Tal-
anites may prove to be misplaced. In particular,
the posterior median eyes are often circular, rath-
er than irregularly shaped as in typical gnaphos-
ids, and the palpal endites may show only vague
traces of an oblique depression. The anterior lat-
eral spinnerets, however, are enlarged, heavily
sclerotized, tubular, and widely separated at their

base, as in other gnaphosids, and their piriform
gland spigots are widened (Platnick 1990, figs.
80-82). Similar piriform gland spigots occur in
some male (but not female) Clubionidae; in Tal-
anites, however, both sexes have widened piri-
form gland spigots, and females also have cylin-
drical gland spigots on the posterior median and
posterior lateral spinnerets that are lacking in
those clubionids but present in other gnaphosids.
We therefore retain Talanites in the Gnaphosi-
dae, at least until a better corroborated hypoth-
esis of its relationships can be supported.

The format of the descriptions and abbrevia-
tions used follow those of Platnick and Shadab
(1976); measurements (taken from type material,
unless otherwise indicated) are in mm.

Talanites Simon

Talanites Simon, 1893:363 (type species by original
designation Talanites fervidus Simon).
Rachodrassus Chamberlin, 1922:160 (type species by
original designation Rachodrassus echinus Cham-
berlin). NEW SYNONYMY.

Drassyllochemmis Gertsch and Davis, 1 940: 1 7 (type
species by original designation Drassyllochemmis
captiosus Gertsch and Davis). First synonymized
with Rachodrassus by Platnick and Shadab, 1976:4.

Diagnosis. — See Platnick and Shadab (1976:
4); the presence of two dorsal spines on tibia IV
has been corroborated for all the Eurasian species
examined, but some of those species lack the
second point on the median apophysis of the
male palp. Those males are readily recognizable
as Talanites, however, by the palpal conforma-

115

116

THE JOURNAL OF ARACHNOLOGY

Figures 1-4.— 1, 2, Talanites fervidus Simon; 3, 4, T. ubicki, new species: 1, left male palp, ventral view; 2,
same, retrolateral view; 3, epigynum, ventral view; 4, same, dorsal view.

tion, including a long and arched palpal tibia, a
wide prolateral embolus, and a greatly elongated
median apophysis.

Description.— See Platnick and Shadab (1976:
4).

Included species.— From America, T. echinus
(Chamberlin), NEW COMBINATION; T. exli-
neae (Platnick and Shadab), NEW COMBINA-
TION; T. captiosus (Gertsch and Davis, NEW
COMBINATION; and T. moodyae and T. ubicki,
new species; from Eurasia, at least T. fervidus
Simon and the four Soviet species discussed be-
low. Examination of the types and other speci-
mens of two further Soviet species, T. aculeatus
Charitonov (1946, 1969) and T. atscharicus
Mcheidze (1946), indicates that they do not be-
long to Talanites. The figures provided for T.
tibialis Caporiacco (1934) from India and Pa-
kistan indicate that it is also misplaced. Of the
other described species (from Greece, north Af-
rica, and Burma), little can be said until their
types can be examined.

Talanites fervidus Simon
Figs. 1, 2

Talanites fervidus Simon, 1893:363 (male syntype from
the Dead Sea area, Israel, in MNHN, examined).

Diagnosis.— Males resemble those of T. mik-
hailovi in having an excavated embolar base, but
the embolus (Fig. 1 ) is longer than in that species.

Male.— Total length 4.09. Carapace 1.88 long,
1.50 wide. Femur II 1.54 long. Eye sizes and
interdistances: AME 0.05, ALE 0.05, PME 0.05,
PLE 0.05; AME-AME 0.04, AME-ALE 0.03,
PME-PME 0.09, PME-PLE 0.08, ALE-PLE 0.03;
MOQ length 0.14, front width 0.14, back width
0.19. Leg spination: femora: I pO-1-1; II rO-0-0;
III rO-1-1; tibiae: I pl-1-1, v2-2-2; II v2-2-2, rO-
0-0; III vlp-2-2; metatarsi: II pO-1-0; IV pl-2-
2. Tibial apophysis directed retrolaterally,
scarcely wider than basal tibial spine; embolar
base excavated, tip twisted; median apophysis
without second point (Figs. 1 , 2).

Female.— Although Simon recorded both sex-
es, no females are now housed with the male
syntype described above.

Material examined . — Only the syntype, without date
or collector.

Talanites mikhailmi, new species
Figs. 17, 18

Type. —Male holotype from Dzhanibek, Ural,
Kazakhstan (30 June-6 July 1982; K. G. Mik-
hailov), deposited in ZIL.

PLATNICK & OVTSHARENKO-r^i^Af/T’E'S (GNAPHOSIDAE)

117

■Sm§M

Figures 5-8.—Talanites dunini, new species: 5, left male palp, ventral view; 6, same, retrolateral view; 7,
epigynum, ventral view; 8, same, dorsal view.

Etymology.— The specific name is a patronym
in honor of the collector of the holotype, Dr. K.
Mikhailov of Moscow State University.

Diagnosis.— Males resemble those of T. fer-
vidus in having an excavated embolar base, but
the embolus (Fig. 1 7) is shorter than in that spe-
cies.

Male.— Total length 4.90. Carapace 2.27 long,
1.69 wide. Femur II 1.28 long. Eye sizes and
interdistances: AME 0.06, ALE 0.09, PME 0.06,
PLE 0.07; AME-AME 0.06, AME-ALE 0.03,
PME-PME 0. 1 1 , PME-PLE 0.11, ALE-PLE 0.05;
MOQ length 0.19, front width 0.18, back width
0.23. Leg spination: femur I pO-1-1; tibiae; I v2-
2-2, rO-1-1; II v2-2-2, rl-1-1; metatarsus II pl-
1-0. Tibial apophysis short, wide, directed dis-
tally; embolar base excavated, tip twisted; me-
dian apophysis with second point (Figs. 17, 18).

Female Unknown.

Other material examined. —USSR; Kazakhstan: Ural;
Dzhanibek, 27-30 June 1975 (Y. I. Chernov), 1 male
(ZIL).

Talanites dunini, new species
Figs. 5-8

Types.— Male holotype and female allotype
from Saatly, Dzhafarkhan, Saatlinskii, Azerbai-

jan (31 August 1982; P. M. Dunin), deposited in
ZIL.

Etymology.— The specific name is a patronym
in honor of the collector of the types.

Diagnosis.— Males can be recognized by the
blade-shaped embolus (Fig. 5), females by the
short, flattened, and triangular epigynal hood (Fig.
7).

Male.— Total length 4.43. Carapace 2.03 long,
1.61 wide. Femur II 1.58 long. Eye sizes and
interdistances: AME 0.07, ALE 0.09, PME 0.07,
PLE 0.09; AME-AME 0.07, AME-ALE 0.03,
PME-PME 0. 1 3, PME-PLE 0.09, ALE-PLE 0.04;
MOQ length 0.20, front width 0.21, back width
0.27. Leg spination: femora: I pO-1-1; II rO-0-0;
tibiae: I, II v2-2-2; III vlp-2-2; metatarsi I, II
pi -0-0. Tibial apophysis short, directed retro-
laterally; embolus blade-shaped, with narrow
base; median apophysis without second point
(Figs. 5, 6).

Female.— Total length 5.87. Carapace 2.36
long, 1.76 wide. Femur II 1.61 long. Eye sizes
and interdistances: AME 0.07, ALE 0.10, PME
0.08, PLE 0.10; AME-AME 0.06, AME-ALE
0.03, PME-PME 0. 1 4, PME-PLE 0.13, ALE-PLE
0.06; MOQ length 0.26, front width 0.21, back
width 0.30. Leg spination: femora: I dl-1-0, pO-

Figures 9-\2. — Talanites fagei Spassky: 9, left male palp, ventral view; 10, same, retrolateral view; 11, epi-
gynum, ventral view; 1 2, same, dorsal view.

1-1; II dl-1-0; tibiae: I pO-0-0, v2-2-0; II pO-0-
0, v2-2-lp, rO-0-0; III vlp-2-2. Epigynum with
pocket-like lateral ridges and short, flattened, tri-
angular anterior hood (Fig. 7); spermathecae
rectangular (Fig. 8).

Other material examined.— USSR; Azerbaijan: Saa-
tlinskii: Saatly, Dzhafarkhan, 16 June-25 August 1982
(P. M. Dunin), 10 males, 1 female (ZIL). Turkmenia:
East Kopetdag: Miana-Chaach, 22-28 Apr. 1978 (G.
T. Kusnetsov), 5 males (ZIL); Krasnovodskaya: Kara-
Kala, Kara-Kalinskii, 4 May 1987 (A. A. Zyuzin), 1
male (ZIL).

Talanites fagei Spassky
Figs. 9-12

Talanites fagei Spassky, 1938:577, figs. 3, 4 (two male
and two female syntypes from Turkmenia and Ka-
zakhstan, in ZIL, examined).

Diagnosis.— Males can be recognized by the
distally expanded and bifid embolus (Fig. 9), fe-
males by the long, pointed anterior epigynal hood

(Fig. 11).

Male (Ustyurt).— Total length 4.54. Carapace
1.99 long, 1.61 wide. Femur II 1.60 long. Eye
sizes and interdistances: AME 0.07, ALE 0.09,
PME 0.08, PLE 0.09; AME-AME 0.06, AME-

ALE 0.03, PME-PME 0.11, PME-PLE 0.11,
ALE-PLE 0.05; MOQ length 0.21, front width
0.20, back width 0.27. Leg spination: femora: I
pO-2-1, rO-1-1; IV rO-1-1; tibiae: I v2-2-2, rl-1-
0; II v2-2-2, r 1-1-0; metatarsi: II pO-1-0; III p2-
2-2. Tibial apophysis short, directed retrolater-
ally; embolus distally expanded, bifid; median
apophysis without second point (Figs. 9, 10).

Female (Ustyurt).— Total length 5.96. Cara-
pace 2.18 long, 1.67 wide. Femur II 1.50 long.
Eye sizes and interdistances: AME 0.08, ALE
0.10, PME 0.08, PLE 0.09; AME-AME 0.06,
AME-ALE 0.03, PME-PME 0.14, PME-PLE
0.13, ALE-PLE 0.06; MOQ length 0.23, front
width 0.22, back width 0.30. Leg spination: fem-
ora: I dl-1-0, pO-1-1; II dl-1-0; tibiae: I pO-0-0,
v2-2-0; II pO-0-0, vlr-2-lp, rO-0-0; III vlp-2-2,
rO-1-1; IV vlp-2-2. Anterior epigynal hood
acutely pointed (Fig. 1 1); spermathecae angular
(Fig. 12).

Material examined.— USSR: Kazakhstan: Alma-Ata:
Alma-Ata, Apr. 1919 (V. Shnitnikov), 1 male (syntype,
ZIL), 15 May 1921 (V. Shnitnikov), 1 female (syntype,
ZIL); Kurty, Kurtinskii, July 1989 (A. A. Zyuzin), 1
male (ZIL). Gurievskaya: Ustyurt Reservation, Onere
River, Ustyurt plateau, 16-21 May 1989 (A. A. Raik-

PLATNICK & OYTSHARENKO-TALANITES (GNAPHOSIDAE)

119

Figures 13-16.— 13, 14, Talanites strandi Spassky; 15, 16, T. moodyae, new species: 13, left male palp, ventral
view; 14, same, retrolateral view; 15, epigynum, ventral view; 16, same, dorsal view.

hanov, S. I. Ibraev), 4 males, 7 females (ZIL). Kirgizia:
Ferghana Mountain ridge, 16 June 1984, evergreen
forest, elev. 1400 m (S. Zonstein), 1 male, 1 female
(AMNH). Russia: Suvorovskaya, Stavropolskii, July-
Aug. 1925 (N. Karancheva), 1 female (ZIL). Caucasus:
Kabardino-Balcaria, Naltshik, July 1925 (M. Karaitsh-
eva), 1 female (syntype, ZIL). Turkmenia: Central Ko-
petdag: Firyuza, 17 Mar.-26 Apr. 1979 (G. T. Kus-
netsov), 2 males, 1 female (ZIL). Serachskii: Agar-
Tshishme, Serachs, 26 May 1936 (L. Freiberg), 1 male
(syntype, ZIL).

Talanites strandi Spassky
Figs. 13, 14

Talanites strandi Spassky, 1940:353, fig. 1 (male ho-
lotype from Amvrosievka, Donetskaya, Ukraine, in
ZIL, examined).

Diagnosis.— Males can be recognized by the
folded and retrolaterally invaginated tip of the
embolus (Fig. 1 3).

Male (Dzhanibek).— Total length 7.61. Cara-
pace 3.28 long, 2.55 wide. Femur II 2.29 long.
Eye sizes and interdistances: AME 0.07, ALE
0.11, PME 0.09, PLE 0.10; AME-AME 0.08,
AME- ALE 0.04, PME-PME 0.14, PME-PLE
0.15, ALE-PLE 0.07; MOQ length 0.24, front

width 0.22, back width 0.32. Leg spination: fem-
ora; I pO-1-1, rl-1-0; II rl-l-l; III pl-1-1; IV
pl-1-1, rl-l-l; tibiae: I pl-1-1, v2-2-2, rl-l-l;
II V2-2-2, r2-l-l; III rl-1-2; IV pl-1-2, rl-1-2;
metatarsi: I pO-1-0, rO-1-0; II pi- 1-0, rO-1-0; III
p2-2-2. Tibial apophysis broad, shifted dorsally;
embolar tip large, folded, retrolaterally invagi-
nated; median apophysis with large second point
(Figs. 13, 14).

Female. —Unknown.

Material examined. — USSR: Kazakhstan: Ural:
Dzhanibek, 8- 1 1 Sept. 1 982 (K. G. Mikhailov), 2 males
(ZIL). Ukraine: Donttskaya: Amvrosievka, June 1912
(N. Spasskaja, S. Spassky), 1 male (holotype, ZIL).

Talanites moodyae, new species
Figs. 15, 16

Type.— Female holotype taken under a large
rock on the north slope of Rocky Hill, near Ex-
eter, Tulare Co., California (6 January 1983; M.
J. Moody, W. L. Abel), deposited in AMNH
courtesy of Ms. Moody.

Etymology.— The specific name is a patronym
in honor of the collector of the type, who first
recognized the species as new.

120

THE JOURNAL OF ARACHNOLOGY

Figures M-li. — Talanites mikhailovi, new species: 17, left male palp, ventral view; 18, same, retrolateral
view.

Diagnosis.— Females resemble those of T.
ubicki in having unusually small eyes, but can
be distinguished by their relatively short, wide
spermathecae (Fig. 1 6).

Male. — Unknown.

Female.— Total length 10.36. Carapace 4.43
long, 3.36 wide. Femur II 2.88 long. Eye sizes
and interdistances: AME 0.06, ALE 0.12, PME
0.08, PLE 0.09; AME- AME 0.10, AME- ALE
0. 12, PME-PME 0.24, PME-PLE 0.26, ALE-PLE
0.12; MOQ length 0.23, front width 0.22, back
width 0.40. Leg spination: femora: I dl-1-0, pO-
1-2, rO-1-0; II dl-1-0, pl-1-1, rO-1-0; III, IV dl-
1-1, pl-1-1, rO-1-1; tibiae: I v2-2-lp; II v2-2-2;
metatarsi: III pi -2-2, v2-2-2, r2-2-2. Epigynum
with distinct anterior margin, short lateral mar-
gins, and wide median plate (Fig. 15); sperma-
thecae short, wide (Fig. 16).

Other material examined.— None.

Talanites ubicki, new species
Figs. 3, 4

Type. — Female holotype taken under serpen-
tine floats along San Marin Drive, Novata, Mar-
in Co, California (7 March 1982; D. Ubick), de-
posited in AMNH courtesy of Mr. Ubick.

Etymology.— The specific name is a patronym
in honor of the collector of the type, who first
recognized the species as new.

Diagnosis.— Females resemble those of T.
moodyae in having unusually small eyes, but can
be distinguished by their longer spermathecae
(Fig. 4).

Male Unknown .

Female.— Total length 6.08. Carapace 2.67
long, 2.08 wide. Femur II 1.80 long. Eye sizes
and interdistances: AME 0.06, ALE 0.10, PME
0.05, PLE 0.07; AME-AME 0.07, AME-ALE
0.06, PME-PME 0. 16, PME-PLE 0. 1 5, ALE-PLE
0.05; MOQ length 0.18, front width 0.19, back
width 0.27. Leg spination: femora: I dl-1-0, pO-
0-1; II dl-1-0, pO-1-1; III dl-1-1, pO-1-1, rl-1-
1; IV dl-1-1, pO-1-1, rO-1-1; tibiae: I v2-2-0; II
v2-2-lp; metatarsi: III p2-2-2, v2-2-2, rl-2-2.
Epigynum with small anterior hood and weak
lateral margins (Fig. 3); spermathecae long (Fig.
4).

Other material examined.— One female taken with
the type, three females taken at the same locality ( 1 8
Dec. 1982), and one female taken at the same locality
(2 Jan. 1986), all in CDU.

PLATNICK & OVTSHARENKO-r^L^iV/reS (GNAPHOSIDAE)

121

ACKNOWLEDGMENTS

The work reported on here was done while the
first author was a guest of the Zoological Insti-
tute, Akademia Nauk, Leningrad (ZIL), and that
support is greatly appreciated. We thank M. J.
Moody of the California Department of Food
and Agriculture, and D. Ubick of the California
Academy of Sciences (CDU), for providing the
California specimens described above, J. Heur-
tault and C. Rollard of the Museum National
d’Histoire Naturelle, Paris (MNHN) for provid-
ing type material, G. Levy of the Hebrew Uni-
versity, Jerusalem, for information on the type
locality of T. fervidus, and M. U. Shadab of the
American Museum of Natural History (AMNH)
for assistance with illustrations.

LITERATURE CITED

Caporiacco, L. di. 1934. Aracnidi deH’Himalaia e del
Karakoram raccolti dalla missione Italiana al Ka-
rakoram (1929-VII). Mem. Soc. Ent. Ital., 13:1 13-
160.

Chamberlin, R. V. 1922. The North American spi-
ders of the family Gnaphosidae. Proc. Biol. Soc.
Washington, 35:145-172.

Charitonov, D. E. 1946. New forms of spiders of the
USSR. Izv. Est.-Nauchn. Inst. Molotovsk. Univ.,
12:19-32.

Charitonov, D. E. 1969. Material’i k faune paukov

SSR. Uchel. Zap. Permsk. Ord. Trudy. Krasn. Znam.
Gos. Univ. Ime A. M. Gorkogo, 179:59-133.

Gertsch, W. J. & L. I. Davis. 1940. Report on a
collection of spiders from Mexico. III. Amer. Mus.
Novitates, 1069:1-22.

Mcheidze, T. 1 946. Nov’e vid’ paukov v Gruzii. Bull.
Mus. Gaaorgie, 13(A):28 5-302.

Penniman, A. J. 1985. Revision of the and

pugnata groups of Scotinella (Araneae, Corinnidae,
Phrurolithinae) with a reclassification of phruroli-
thine spiders. Unpub. Ph.D. diss., Ohio State Uni-
versity, available through University Microfilms In-
ternational (no. 8510623).

Platnick, N. I. 1990. Spinneret morphology and the
phylogeny of ground spiders (Araneae, Gnaphoso-
idea). Amer. Mus. Novitates, 2978:1-^2.

Platnick, N. I. & M. U. Shadab. 1976. A revision of
the spider genera Rachodrassus, Sosticus, and Sco-
podes (Araneae, Gnaphosidae) in North America.
Amer. Mus. Novitates, 2594:1-33.

Simon, E. 1 893. Histoire naturelle des araignees. Paris,
l(2):257-488.

Spassky, S. 1938. Araneae palaearcticae novae. II.
Riga, Festschrift zum 60. Geburtstage von Professor
Dr. Embrik Strand, 4:573-582.

Spassky, S. 1940. Araneae palaearcticae novae. V.
Folia Zool. Hydrobiol., 10:353-364.

Manuscript received November 1990, revised December
1990.

1991. The Journal of Arachnology 19:122-125

DIPLOCENTRUS PEREZI, A NEW SPECIES OF SCORPION
FROM SOUTHEASTERN MEXICO (DIPLOCENTRIDAE)

W. David Sissom: Department of Biology, Elon College, Elon College, North Carolina
27244 USA

Abstract. Diplocentrus perezi, a new species of diplocentrid scorpion, is described from
the Mexican state of Veracruz, representing the first report of the genus from that area. A
female specimen from Tabasco is possibly referable to this species. The new species is most
similar to D. mexicanus Peters, but differs from that species in carapacial granulation,
tarsomere II spine formula, and carination of the pedipalps and metasoma.

Studies in the last decade have shown that the
genus Diplocentrus in southern Mexico and the
Yucatan Peninsula is quite diverse (Francke
1977a, 1977b, 1978). However, there are cur-
rently no known records of this genus from the
poorly-sampled southeastern coastal states. It is
the purpose here to describe a distinct new spe-
cies based on an adult male specimen from
southern Veracruz. A subadult female specimen
from Tabasco may also be referable to this spe-
cies.

Nomenclature and mensuration follows that
of Stahnke (1970), with the following exceptions:
carinal terminology and cheliceral measure-
ments are after Francke (1975, 1977a) and tri-
chobothrial terminology is after Vachon (1974).

Diplocentrus perezi, new species
Figs. 1-7

Type data.— Holotype male from San Martin
Tuxtla Volcano, Veracruz, Mexico (1300 m),
September 1 985 by Gonzalo Perez-Higareda; de-
posited in the American Museum of Natural His-
tory, New York.

Etymology.— This species is dedicated to Dr.
Gonzalo Perez-Higareda, who collected the type
specimen and made it available for study.

Distribution.- Known from the type locality
in southern Veracruz, with a possible record in
Tabasco.

Diagnosis.— Adult male 62 mm in length. Base
color dark orange brown to brown, with distinct
dusky markings throughout. Carapace with coarse
granulation restricted to area surrounding ante-
rior margin. Tergite VII moderately bilobed,
granulose. Pectinal tooth count 14-13. Metaso-

ma I-IV with 10 keels; dorsolateral carinae strong,
crenulate; ventrolateral and ventral submedian
carinae moderate, smooth to irregularly granu-
lar. Metasoma V with dorsolateral carinae mod-
erate, granular; ventrolateral and ventromedian
carinae strong with large spinoid denticles. Chel-
iceral chela length/chela width 1.27; fixed finger
length/chela width ratio 0.69; movable finger
length/chela length 1.00. Pedipalps: dorsal sur-
face of femur relatively flat, width distinctly
greater than depth; pedipalp patella with two
dorsal carinae, the dorsomedian strong, smooth
and the dorsoextemal weak, smooth; chela fixed
finger length/carapace length 0.95; movable fin-
ger length/carapace length 1.26; dorsal and ex-
ternal surfaces of chela palm moderately to
strongly reticulate, with outer palm carinae well
developed. Chela length/width ratio 2.86. Tar-
somere II spine formula: 3/4 4/4: 4/4 4/4: 5/5
5/5: 5/5 5/5.

Description.— Based on holotype male.

Prosoma: Carapace (Fig. 1) base color dark
orange brown with distinct dusky pattern. An-
terior portion of carapace covered with medium-
sized granules; area around ocular tubercle
densely, finely granular; remainder of carapace
sparsely granular. Coxostemal region yellow
brown to light brown, lustrous. Sternum with
about 9 pairs of setae; coxae sparsely setose.

Mesosoma: Tergites brown, with distinct dusky
pattern throughout. Tergites I-IV acarinate, V-
VII weakly monocarinate, with median carina
weak, smooth. Tergites I-II with minute granu-
lation on lateral portions; III-VI with dense min-
ute granulation interspersed with sparse, coarse
granulation. Tergite VII moderately bilobed, with

122

SISSOM-NEW DIPLOCENTRUS FROM VERACRUZ, MEXICO

123

each lobe granulose. Pectines pale yellow, with
14-13 teeth. Stemites uniformly yellow brown;
III-VI smooth, lustrous, moderately setose along
lateral and posterior margins. Stemite VII tetra-
carinate; lateral carinae moderate, unevenly
smooth; submedian pair weaker, smooth; about

10 pairs of reddish setae present.

Metasoma: Segments I-III dark orange brown;

IV and V slightly darker than preceding seg-
ments. Segment I 1 .09 times longer than wide;

11 1.33 times longer than wide; V 2.83 times
longer than wide. Segments I-IV: Dorsolateral
carinae strong, irregularly crenulate. Lateral su-
pramedian carinae on I-III strong, granular; on
IV moderate, subgranose. Lateral inframedian
carinae weak on all four segments, irregularly
granular. Ventrolateral carinae on I-II moderate,
smooth; on III-IV moderate, granular. Ventral
submedian carinae on I-III moderate, irregularly
granular; on IV vestigial, with some granulation
anteriorly. Intercarinal spaces with sparse, fine
and coarse granulation. Segment V: (Fig. 2) Dis-
tinctly narrower than segments I-IV, with lateral
sides subparallel. Dorsolateral carinae moderate,
granular. Lateromedian carinae vestigial, with a
few sharp granules anteriorly. Ventrolateral, ven-
tromedian, and ventral transverse carinae strong,
with distinctly enlarged, subconical granules (Fig.
2). Dorsal intercarinal space with dense fine gran-
ulation anteriorly; lateral intercarinal space
sparsely, coarsely granular; ventral intercarinal
spaces smooth, moderately setose.

Telson: (Fig. 2) reddish to orange brown. Ven-
tral surface of vesicle densely setose, proximally
with numerous sharp granules. Subaculear tu-
bercle strong, subconical, covered with setae and
fine white microchaetes.

Chelicerae: (Fig. 3) light yellow brown, lus-
trous, with distinct dusky mottling on dorsal sur-
face of manus; teeth dark reddish brown. Mov-
able finger with subdistal tooth closely apposed
to distal tooth.

Pedipalps: Base color orange brown, femur
lighter than patella and chela; distal end of chela
manus and fingers infuscate. Trichobothrial pat-
tern Type C, orthobothriotaxic (Vachon 1974).
Femur: (Fig. 4) Dorsointemal and ventrointemal
carinae strong, granulose; dorsoextemal carina
strong, irregularly granulose proximally, smooth
distally; ventroextemal carina obsolete. Internal
surface moderately granulose; dorsal surface flat,
moderately granular; ventral and external sur-
faces smooth. Patella: (Fig. 5) Dorsal aspect with

two smooth carinae; dorsointemal carina strong,
dorsoextemal carina weak; ventrointemal carina
moderate, granular; ventroextemal carina weak,
smooth. Basal tubercle of inner surface moder-
ate, followed distally by three to four large gran-
ules; remainder of inner surface covered with
moderately dense, fine granules. External surface
relatively flat, with very weak reticulations. Ven-
tral face slightly convex, essentially smooth. Che-
la: (Figs. 6,7) Dorsal marginal carina strong,
granulose; dorsal secondary carina weak, smooth;
digital carina strong, smooth; external secondary
carina weak, smooth; ventroextemal carina ob-
solete basally, but strong on distal one-fourth of
manus, smooth; ventromedian carina strong, es-
sentially smooth; ventrointemal carina weak,
smooth; two additional carinae on inner face,
these vestigial, smooth to feebly granular. Dorsal
and external faces of manus moderately reticu-
late (Fig. 6), ridges smooth. Inner and ventral
faces with irregular granulation and punctations.
Dorsal and external surfaces of manus sparsely
setose; internal and ventral surfaces moderately
to densely setose; fixed and movable fingers
densely setose. Inner margins of chela fingers with
moderate scalloping.

Legs: Contrasting in coloration with the body;
proximal segments yellow brown; tarsi light yel-
low.

Measurements (of holotype, in mm): Total
length, 62.2; carapace length 7.6; mesosoma
length 1 7.8. Metasomal segments: I length/width,
4. 7/4. 3; II length/width, 5. 3/4.0; III length/width,
5. 5/3. 9; IV length/width, 6. 3/3. 5; V length/width,
%.l/2.9. Telson length 6.8; vesicle length/width/
depth, 5.5/3. 1/2.5; aculeus length, 1.3. Chelic-
erae: chela length/width, 2.24/1.77; fixed finger
length, 1.23; movable finger length, 2.24. Pedi-
palps: femur length/width, 1 patella length/
width, 1. 'i/2.1-, chela length/width/depth, 15.4/
5. 4/3. 5; fixed finger length, 7.2; movable finger
length, 9.5.

Comparisons.— Diplocentrus perezi is most
similar to D. mexicanus Peters. From this spe-
cies, D. perezi may be distinguished by the fol-
lowing characteristics; (1) carapacial granulation
limited to area immediately surrounding anter-
omedian notch; (2) lower tarsomere II spine for-
mula (in D. mexicanus mexicanus 5/6 5/6: 6/6
6/7: 7/7 7/7: 7/7 7/7 and in D. mexicanus oax-
acae Francke 5/6 5/6: 6/7 6/7: 7/8 7/8: 7/8 7/
8); (3) the dorsal margin of the pedipalp chela is
relatively straight in D. perezi, but noticeably

124

THE JOURNAL OF ARACHNOLOGY

Figures 1-7.— Morphology of Diplocentrus perezi, new species: 1, dorsal aspect of carapace; 2, lateral aspect
of metasomal segment V and telson; 3, dorsal aspect of right chelicera; 4, dorsal aspect of right pedipalp femur;
5, external aspect of right pedipalp patella; 6, external aspect of right pedipalp chela; 7, dorsal aspect of chela.
Trichobothrial patterns are shown for pedipalpal structures.

sinuous in D. mexicanus', (4) the dorsal carinae
of the pedipalp patella are smooth in D. perezi,
but granulose or subcrenate in D. mexicanus-, (5)
the patellar et trichobothria form a distinct ob-
tuse angle in D. perezi, but are almost in a linear
arrangement in D. mexicanus-, and (6) in D. per-
ezi, the ventrolateral and ventral submedian ca-
rinae of metasomal segments I-IV are weaker
than in D. mexicanus and smooth to irregularly
granular, rather than distinctly granulose.

Diplocentrus perezi may be easily distin-
guished from D. tehuacanus Hoffmann, which is
found in Puebla, Morelos, Guerrero, and Oaxaca

by (1) the presence of two dorsal patellar carinae,
rather than only one; (2) by having strong retic-
ulations on the dorsal and external faces of the
pedipalp chela (in D. tehuacanus, the reticula-
tions are much weaker); (3) by having strong,
costate dorsolateral metasomal carinae (in D. te-
huacanus these are weak and smooth; (4) and by
having well developed ventrolateral and ventral
submedian carinae on metasomal segment IV (in
D. tehuacanus the ventrolateral carinae are feeble
and smooth and the ventral submedians more
or less obsolete). There are also conspicuous
morphometric differences: in D. perezi, the ped-

SISSOM-NEW DIPLOCENTRUS FROM VERACRUZ, MEXICO

125

ipalp chela fingers and the individual metasomal
segments are proportionately longer.

Comments.— A subadult female specimen col-
lected near Villahermosa, Tabasco, June 1985
by Gonzalo Perez-Higareda is possibly referable
to this species. The tarsomere II spine formula
for this female (4/4 4/4: 4/4 4/5: 5/5 5/5: 5/5 5/
5) does not differ significantly from that of the
male; the pectinal tooth count, 1 2-1 1 , is also very
close to that of the male (female diplocentrids
tend to have slightly lower counts than males).
It differs significantly from the holotype male in
that its pedipalp chelae are proportionately
broader, deeper, and more convex, with the ca-
rinae of the dorsal and external faces obsolete;
its metasomal carinae are slightly weaker; and
its carapace, pedipalps, and tergites are smooth
and lustrous. The characters on which these dif-
ferences are based are all known to be sexually
dimorphic in the genus, and consequently should
be interpreted with great caution. Given the ho-
lotype male’s morphology in these characters,
the predicted female morphology would be ex-
actly that exhibited by the specimen from Ta-
basco. However, the two specimens were col-
lected in different habitats and at different
altitudes: the female from lowlands and the ho-
lotype male from wet forest at 1 300 m (G. Perez-
Higareda, pers. comm., 1987). This leads to the
suggestion, at least, that they may represent dif-
ferent species. Because females of closely related,
but different, species of Diplocentrus are difficult
to distinguish, accurate determination of the spe-
cies identity of the female cannot be provided
until males are known from the Tabasco area.
The female specimen is deposited in the collec-
tion of the Estacion de Biologia “Los Tuxtlas”,
Universidad Nacional Autonoma de Mexico,
Catemaco, Veracruz.

ACKNOWLEDGMENTS

I am grateful to G. Perez-Higareda of the In-
stitute de Biologia Tropical Los Tuxtlas, Cate-
maco, Veracruz for making the holotype of D.
perezi and the female specimen from Tabasco
available for study. N. 1. Platnick of the Amer-
ican Museum of Natural History New York
kindly allowed me to examine the holotype of
Diplocentrus tehuacanus\ both he and H. W. Levi
of the Museum of Comparative Zoology provid-
ed other Diplocentrus material for comparative
purposes. Page charges were paid with the assis-
tance of a Faculty Research and Development
grant from Elon College.

LITERATURE CITED

Francke, O. F. 1975. A new species of Diplocentrus
from New Mexico and Arizona (Scorpionida, Di-
plocentridae). J. Arachnol., 2:107-1 18.

Francke, O. F. 1977a. Scorpions of the genus D/p/o-
centrus from Oaxaca, Mexico (Scorpionida, Diplo-
centridae). J. Arachnol., 4:145-200.

Francke, O. F. 1977b. The genus Diplocentrus in the
Yucatan Peninsula with description of two new trog-
lobites (Scorpionida, Diplocentridae). Assoc. Mex.
Cave Stud. Bull., 6:49-61.

Francke, O. F. 1978. New troglobite scorpion of ge-
nus Diplocentrus (Scorpionida: Diplocentridae). En-
tomol. News, 89:39-45.

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

Vachon, M. 1974. Etude des caracteres utilises pour
classer les families et les genres de Scorpions. (Ar-
achnides). Bull. Mus. Nat. d’Hist. nat., Paris, 3rd
series. No. 140, Zool., 104:857-958.

Manuscript received November 1990, revised January
1991.

1991. The Journal of Arachnology 19:126-135

WHEN IS THE SEX RATIO BIASED IN SOCIAL SPIDERS?:
CHROMOSOME STUDIES OF EMBRYOS AND MALE MEIOSIS
IN ANELOSIMUS SPECIES (ARANEAE, THERIDIIDAE)

Leticia Aviles': Museum of Comparative Zoology, Harvard University, Cambridge,
Massachusetts 02138 USA and Department of Integrative Biology, University of Cal-
ifornia, Berkeley California 94720 USA

Wayne Maddison': Department of Integrative Biology, University of California, Berke-
ley, California 94720 USA

Abstract. Embryo chromosome preparations of four species of social spiders of the genus
Anelosimus show that the two species known or suspected to form permanent, multigener-
ational colonies, A. eximius and A. domingo, have a highly female-biased primary sex ratio.
Anelosimus jucundus and A. studiosus, on the other hand, are shown to produce an even
number of males and females. The magnitude of the bias of A. eximius embryos is similar
to that reported for young preadult spiders of this species, therefore ruling out differential
mortality of juveniles as the cause of this species’ sex ratio bias. Chromosome counts of
nuclei in second division of A. eximius male meiosis indicate that nuclei destined to yield
sons and daughters are produced in equal numbers. Therefore, the sex ratio biasing mech-
anism in this species must act after male meiosis and before egg laying. The question of
how early the sex ratio bias arises still needs to be resolved in other social spiders. We
discuss some methodological and theoretical complications associated with measuring sex
ratios at different stages of the life cycle and present a fast and reliable technique to obtain
embryo chromosome preparations.

The occurrence of highly female-biased sex ra-
tios among adults of several species of social spi-
ders has been known since the 1960’s (Buskirk
1981), but there has been little study of exactly
when in the spiders’ life cycle the sex ratio bias
arises. Knowledge of the timing of the sex ratio
bias is important on at least two accounts: first,
from an evolutionary point of view, it would help
us determine whether differential parental in-
vestment is involved in biasing the sex ratio; and,
second, from a physiological point of view, it
would bring us closer to identifying the mecha-
nism by which the sex ratio bias is accomplished.

Fisher’s principle (Fisher 1930) states that, at
equilibrium, the total parental investment in off-
spring of each sex should be equal. Departures
from this equilibrium should bring about selec-
tion to restore an even sex ratio because indi-
viduals of the rare sex would have a reproductive
advantage. Exceptions to Fisher’s sex ratio prin-

‘ Present address: Dept, of Ecology and Evolutionary
Biology, University of Arizona, Tucson, Arizona 85721
USA.

ciple have been pointed out by Hamilton (1967),
who first noted that biased sex ratios can evolve
when the assumption of panmixia, implicit in
Fisher’s argument, is violated. Parasitic and fig
wasps (e.g., Werren 1980; Waage 1982; Herre
1985) and hummingbird-flower mites (Wilson
and Colwell 198 1) are notable examples. As first
noted by Aviles (1983, 1986), social spiders ap-
pear to represent another case in which Fisher’s
principle has been violated. In most social spi-
ders, however, this violation has not been con-
firmed because their sex ratio has been measured
late enough in the spider’s life cycle (usually
among adults) that higher male mortality during
the preadult or adult instars cannot be ruled out
as the cause of their sex ratio bias. As noted by
several authors (Leigh 1970; Chamov 1982;
Trivers 1985), if biased sex ratios are due solely
to mortality occurring after the end of the period
of parental investment then Fisher’s principle is
not violated. In three species, Anelosimus exi-
mius Keyserling, Achaearanea wau Levi and Ste-
godyphus dumicola Pocock, there is indirect ev-
idence that more females are actually being

126

AVILES & MADDISON- SOCIAL SPIDER SEX RATIO

127

produced. This evidence comes either from
young-preadult sex ratios estimated in isolated
natural colonies (Aviles 1983, 1986) or from
preadult/adult sex ratios measured among in-
dividuals raised under controlled conditions, ei-
ther from egg sacs (Vollrath 1986b; Lubin and
Crozier 1985; Lubin in press) or from a colony
maintained in the laboratory (Seibt and Wickler
1988). These measurements provide evidence of
an early bias, but may still be affected by mor-
tality during the juvenile instars.

In this paper we present a cytogenetic method
that makes it possible to directly determine the
sex of a developing embryo and, therefore, to
measure the sex ratio before mortality becomes
a factor. We apply this method to four species
of the genus Aneiosimus Simon (Levi 1956, 1 963)
present in Ecuador. Two of these, A. eximius and
A. domingo Levi, are among the most social in
the genus, with spiders that cooperate in prey
capture and brood care and share a permanent
communal nest generation after generation (A.
eximius references: Brach 1975; Tapia & De Vries
1980; Christenson 1984; A. domingo references:
Levi & Smith 1982; Rypstra & Tirey 1989). The
other two, A. jucundus O. P.- Cambridge and A.
studiosus Hentz, on the other hand, are known
to exhibit a less advanced form of sociality where
the offspring of a single female remain together
for the early part of their life cycle but disperse
before reaching adulthood (Brach 1977; Nentwig
& Christenson 1986). Unlike A. eximius, A. ju-
cundus and A. studiosus have been reported to
have 1 : 1 preadult or adult sex ratios (Fowler &
Levi 1979; Nentwig & Christenson 1986; Voll-
rath 1986b reared offspring from two A. jucundus
sacs and obtained an even sex ratio). Aneiosimus
domingo sex ratios have not been previously re-
ported.

Should it be confirmed that the embryo sex
ratios are biased, the next question to be an-
swered is what is the mechanism by which the
sex ratio bias arises. This question is of special
interest in spiders because, unlike haplodiploid
organisms that constitute most other known cases
of extreme sex ratio biases (Hamilton 1967), spi-
ders are diploid organisms with chromosomal
sex determination and therefore lack the oppor-
tunity to bias the sex ratio by choosing whether
or not to fertilize the egg. In this paper we ex-
amine the possibility that the earliest acting
mechanism, a bias in male meiosis leading to the
excess production of sperm destined to yield
daughters, may occur in A. eximius.

MATERIALS AND METHODS

Chromosome preparations were obtained from
embryos and males collected in Ecuador from
naturally occurring colonies of the four Aneio-
simus species. The sex of an embryo, or whether
a nucleus in a spider testis is going to become a
male- or a female-producing spermatozoid, can
be determined cytologically thanks to the differ-
ence in chromosome number between male and
female spiders. The most common mechanism
of sex determination in spiders involves two pairs
of X chromosomes, with the two members of
each pair present in females (X,X,X2X2) and only
one in males (X,X20) (White 1973). The sex of
an individual egg or of a developing spermato-
zoid can therefore be simply determined by ob-
taining its chromosome count.

Aneiosimus eximius egg sacs were collected in
July 1988 from two colonies, approximately 1.5
km apart, near the Recinto A. Perez Intriago,
ICm 1 1 3, Quito - Pto. Quito road (0°6'N:79°5'W).
Aneiosimus egg sacs were also collected near Per-
ez Intriago in June 1989 from one colony found
in a forest pocket near the Silanche river. Ane-
iosimus jucundus sacs were collected from two
colonies in Crucita, Manabi (0°52'S:80°33'W), in
August 1988 and A. studiosus near Calderon, Pi-
chincha (0°6'S:78°27'W), in July 1989. Males of
the four species were collected from the same
sites as the egg sacs, except that in addition one
male of A. studiosus was collected from El Tingo,
Pichincha (0°17'S:78°27'W). Egg sacs and males
were brought alive to the laboratory where the
preparations were made.

The technique we developed to obtain chro-
mosome preparations from individual eggs is de-
scribed in the Appendix. Basically, it is much
like typical acetic squash methods (e.g., Darling-
ton & La Cour 1975) which stain then squash,
except that it stains after squashing so as to yield
much superior squashing. The stage of devel-
opment of the eggs at the moment the sacs were
collected was not known. However, we found
that good preparations can be obtained from a
wide range of stages, from very young embryos
whose limb buds are just beginning to appear to
older ones with well formed buds prior to the
development of a cuticular covering. Once the
cuticle has been formed, squashing is not as good
and there are fewer dividing cells. All eggs in the
A. domingo and A. eximius sacs that appeared
to be developing normally were squashed (one
to three eggs in the first three A. eximius sacs

128

THE JOURNAL OF ARACHNOLOGY

1

Figure 1.— Metaphase I nucleus in A. eximius male
meiosis. Note 10 acrocentric bivalents and two X chro-
mosomes. Scale bar = 5 /irm.

were lost due to mishandling). In A. studiosus, a
random sample of around 30 eggs per sac was
chosen. In A. jucundus, a similar sample size was
chosen, though with the two egg sacs unevenly
represented due to differences in their develop-
mental stage. Sampling in A. jucundus could not
be entirely random because the eggs in a sac were
found to be widely asynchronous in their devel-
opment and the ones that were obviously too old
to yield good preparations were not squashed.

Chromosomes were counted under 1 OOOX us-
ing oil immersion. In the embryo preparations,
nuclei with countable chromosomes were sought

on the microscope slide until at least three were
found with the same chromosome count of either
22 (the male diploid complement) or 24 (female);
if three nuclei with 22 or 24 chromosomes could
not be found, the egg was deemed unscorable.
With few exceptions (see Table 1), around 90%
of the preparations for a given sac could be scored.
Since it is reasonable to suppose that the eggs
scored were a random sample of the e^s squashed
and since, with the only exception of A. jucundus,
the eggs squashed were all or a random sample
of those in a sac, the sex ratios obtained provide
a direct estimation of the primary sex ratios of
the species studied. Fewer than 90% percent of
the preparations in the A. domingo sacs 1 and 4
could be scored because, when the sacs were first
opened, the eggs were still too young to yield
enough cells for scoring. After the first one or
two eggs, these sacs were closed and the prepa-
rations continued at a later date. In the case of
A. jucundus, because of the asynchrony in the
development of the eggs in a sac, some were too
young or too old to yield good preparations.

Chromosome preparations from testes were
obtained by using one of three techniques: (a)
Feulgen, as done by Maddison (1982), (b)
squashing after staining with aceto-orcein, or (c)
the same technique as that described for the eggs.
One A. domingo, nine A. eximius, two A. jucun-
dus and three A. studiosus males were examined

2

3

'4 /

Ah. -

M ^

r • •

^ \1

%

• •

6

* ♦

♦

/)>

1

Figures 2-4.— A. eximius embryo chromosomes: 2, male embryo (22 chromosomes); 3, 4, female embryos
(24 chromosomes). Scale bar = 10 moi.

AVILES & MADDISON-SOCIAL SPIDER SEX RATIO

129

Figures 5, 6. —Second division nuclei of /I. eximius male meiosis: 5, two Metaphase II nuclei; top has 12
chromosomes, bottom, 10; 6, two pairs of early Telophase II nuclei, top has 12 chromosomes each member of
the pair, bottom, 10. Scale bar = 10 fim.

to confirm the chromosome complements of the
four species. The nine A. eximius males were
further examined to investigate the possibility of
a bias in male meiosis leading to the overpro-
duction of XX-bearing spermatids. A. eximius
slides were scanned systematically and all scor-

able nuclei found, scored. Counts were made of
either each second metaphase nucleus (Fig. 5) or
of each pair of early second telophase nuclei,
which nearly always occurred together (Fig. 6).
Confidence limits for the proportions at the 95%
level were obtained from binomial confidence

Table 1.— Number of female and male embryos present in individual egg sacs of four species of the genus
Anelosimus, as determined by their chromosome count: females, 24 chromosomes, and males, 22. The total
number of eggs in a sac, the number squashed for chromosomes and, of those, the percent that yielded preparations
whose chromosome count could be scored are given in columns 3-5.

Species

Sac no.

Eggs

Total

Squashed

% scorable

Females

Males

A. domingo

1

16

16

81

12

1

2

13

13

92

11

1

3

14

14

93

12

1

4

15

15

73

10

1

A. eximius

1

51

43

91

35

4

2

45

44

89

37

2

3

53

50

88

39

5

4

51

51

92

42

5

A. jucundus

1

73

16

75

6

6

2

71

43

63

14

13

A. studiosus

1

39

31

94

15

14

2

47

30

83

13

12

130

THE JOURNAL OF ARACHNOLOGY

Table 2. — Primary sex ratio of four species of the genus Anelosimus reported as the proportion of male embryos
contained in 2-4 egg sacs per species.

Species

No. sacs

No. eggs

Males

Females

Proportion

males

95% c.i.

A. domingo

4

50

4

45

0.08

0.02-0.19

A. eximius

4

172

16

153

0.09

0.05-0.14

A. jucundus

2

39

19

20

0.49

0.33-0.66

A. studiosus

2

54

26

28

0.48

0.34-0.62

interval graphs, as presented by Remington and
Schork(1985).

RESULTS

Chromosome complement of Anelosimus spp. —
Chromosome counts in male meiosis showed that
the male diploid complement in all four species
of Anelosimus examined is 20 autosomes + XXO
(Fig. 1), the typical complement for the family
Theridiidae (Suzuki 1954). Males, therefore, have
22 chromosomes, and females should have 24
chromosomes. As expected, developing eggs were
found to have either 22 or 24 chromosomes in
all four species examined (Figs. 2-4). The 20
autosomes, as well as the two X chromosomes,
are acrocentrics. One male of A. studiosus showed
an extra chromosome, possibly a supernumer-
ary.

Sex ratio of Anelosimus jucundus and

A. studiosus were found to have an even primary
sex ratio (Table 2), while the other two species,
A. domingo and A. eximius, were found to have
highly female biased primary sex ratios. Anelo-
simus domingo egg sacs contained a single male
out of 1 1 to 1 3 eggs and A. eximius sacs contained
from 2 to 5 males out of 39 to 47 eggs (Table 1).
The proportion of males found among embryos
from the four egg sacs of A. domingo is 0.08 and
of A. eximius, 0.09 (Table 2).

All the eggs in the A. domingo sacs and in three
of the A. eximius sacs were found to be develop-
ing normally. One of the A. eximius sacs (#1)
contained 1 dried up egg and 6 egg shells, most
likely the remains of eggs eaten up by two hy-
menopteran parasitic larvae found in the sac.

Is the sex ratio biased by male meiosis?. — By
the second division of male meiosis, nuclei des-
tined to become male-producing sperm have 10
chromosomes, those destined to become female-
producing sperm have 12. At telophase II (Fig.
6), the ratio of male-producing to female-pro-
ducing nuclei was found to be very close to 1 ; 1

(ratio 0.49, A = 105 pairs, 95% confidence in-
terval = 0.34-0.59), showing that male meiosis
is not biased. In the earlier stage of metaphase
II (Fig. 5) we obtained a slightly biased ratio
(ratio 0.39, N = 57, 95% confidence interval =
0.270.53) which however was not significantly
different from 1:1 (p > 0.11, by an exact two-
tailed test based on the binomial probabilities).
The slight bias observed in metaphase II is prob-
ably due to sampling error given that the number
of nuclei scored at this phase is smaller (57 vs.
105) and that at the later telophase II stage the
two types of nuclei occur in even numbers.

DISCUSSION

This study shows that the sex ratio among de-
veloping embryos of two of the most social spe-
cies of the genus Anelosimus, A. eximius and A.
domingo, is highly female biased. The bias in A.
eximius is of the same magnitude as that pre-
viously reported from preadult individuals of this
species (Aviles 1986 and unpublished data) and
from individuals raised from eggs (Vollrath
1986b). This shows that differential mortality of
the sexes during the juvenile instars is not re-
sponsible for the sex ratio bias and that the bias
results from an overproduction of females by the
time the eggs are laid. The sex ratio is therefore
biased throughout the life cycle, and parental in-
vestment in A. eximius, as in A. domingo, is
heavily skewed towards females. This removes
any doubts that this bias represents a violation
of Fisher’s principle, on the one hand, and push-
es back the moment at which the sex ratio biasing
mechanism must act to the period previous to
egg laying. In the other two species, A. jucundus
and A. studiosus, even primary sex ratios were
found.

The difference in sex ratio between A. jucundus
and 4. studiosus, on the one hand, and 4. eximius
and A. domingo, on the other, is consistent with
what is known about the mating system and pop-

AVILES & MADDISON-SOCIAL SPIDER SEX RATIO

131

ulation structure of these species. Anelosimus ju-
cundus and A. studiosus form colonies that dis-
integrate before its members (usually the progeny
of a single female) reach adulthood (Brach 1977;
Nentwig & Christenson 1986; Aviles unpub-
lished). Therefore, mating in these species takes
place among individuals from the population at
large and, as observed, an even sex ratio is ex-
pected. The colonies of A. eximius, on the other
hand, as a consequence of permanent sociality,
constitute isolated lineages whose members re-
produce by inbreeding generation after genera-
tion (Overal & Ferreira 1982; Vollrath 1982;
Smith 1982; Aviles 1983, 1986). According to a
model proposed (Aviles 1983, 1986, in prep.),
the isolated descent of many small lineages and
their rapid turnover rate would bring about the
conditions under which selection at the colony
level can override fisherian selection within col-
onies, making female-biased sex ratios evolu-
tionarily stable (see Frank 1987 for a different
model). The population structure of the fourth
species, A. domingo, is not yet known. However,
cooperation in this species extends through
adulthood (Rypstra & Tirey 1989) and multiple
egg-laying females and spiders of all instars occur
in the colonies (Aviles unpublished), suggesting
that sociality is permanent and that mating takes
place within the parental colony, as occurs in A.
eximius. The strongly female biased sex ratios
here reported lead us to predict that the popu-
lation structure of this species will also be found
to be highly subdivided and that the conditions
that favor the evolution of female biased sex ra-
tios in A. eximius are also present in A. domingo.

One of the questions opened by the present
study has to do with the mechanism by which
such a large overproduction of females is accom-
plished. This study rules out early death of em-
bryos as a possible mechanism since, with the
only exception of one sac, all eggs in the A. ex-
imius and A. domingo sacs examined were de-
veloping normally and almost all were scored.
The biasing mechanism must therefore act dur-
ing the period previous to the deposition of the
eggs. Our results also show that a bias in male
meiosis, the earliest acting possible mechanism,
does not occur in A. eximius: by telophase II,
nuclei destined to give rise to female- and male-
producing sperm occur in equal numbers. This
leaves the following stages as possible targets
during wich the biasing mechanism can act: the
final stages of spermatogenesis, sperm induction.

transfer of sperm to the female spermatheca,
sperm activation in the spermatheca previous to
the fertilization of the eggs, and fertilization it-
self. Once the eggs have been fertilized, the sex
ratio must be already determined, since, as Voll-
rath (1986b) points out, reabsorption of male
zygotes is not likely given that the sperm is added
to the eggs as they are being laid. Some form of
sperm selection, involving either differential
death of sperm, differential activation or sperm
competition, appears the most likely mecha-
nism.

The question of when the sex ratio is biased
still needs to be resolved in other social spiders.
Outside Anelosimus, with the already mentioned
exceptions of Achaearanea wau and Stegodyphus
dumicola for which rearing experiments have
been conducted, sex ratios in other social spiders
have only been measured in adults (Jackson &
Smith 1978; Riechert et al. 1986), in some com-
bination of adults and subadults (Pain 1964;
Kullman et al. 1971) or in some unspecified in-
star, presumably mature individuals (Darchen
1967; Main 1988; Jacson & Joseph 1973; Seibt
& Wickler 1988 for 5. mimosarum Pavesi). As
already mentioned, adult sex ratios are not suf-
ficient evidence that parental investment is bi-
ased, and, therefore, that Fisher’s principle has
been violated. When compared with data taken
at earlier stages, adult sex ratio data can never-
theless be useful as evidence that sex specific
mortality or migration occurs. However, because
in spiders males often mature at least one molt
earlier than females, measuring adult sex ratio is
much more involved than has been generally re-
garded. In social spiders, as in any other species
in which generations are discrete, either due to
seasonality or to recent establishment of the pop-
ulation or colony by just a few founders (Bradoo
1972; Darchen 1978; Lubin & Robinson 1982;
Aviles 1986; Main 1988; Seibt & Wickler 1988),
the difference in the number of molts makes the
proportion of adult males to adult females de-
pendent on the point in the colony life cycle at
which the sample is collected (e.g., see fig. 3 of
Aviles 1986). This might explain why some au-
thors have obtained some instances of male bi-
ased sex ratios (e.g., Vollrath 1986a,b; Riechert
et al. 1986, pp. 185, 186; Main 1988, p. 66), the
large variability found by most authors (Pain
1964; Bradoo 1975; Jacson & Joseph 1973;
Riechert et al. 1986; Vollrath 1986a,b; Seibt &
Wickler 1988), and the differences among them.

132

THE JOURNAL OF ARACHNOLOGY

To solve this problem one might measure adult
sex ratio as the proportion of adult males vs. the
chronologically equivalent preadult female in-
star. However, because of the different times the
sexes remain in those instars, this estimate is
biased against females (female spiders are only
temporarily in the preadult instar until molting
to adulthood while males of several cohorts ac-
cumulate in the adult instar). In general, the per-
sistence time in a particular instar should always
be taken into consideration when any two stages
in a life cycle are compared by vertical sampling.
The measurement that would more fairly com-
pare all males and females of the same cohort
would have to count adult males vs. females of
the same and all older instars to which they molt
while males are still around. However, even this
estimate would vary depending upon when in
the colony life cycle the sample was collected, if
males migrate or die earlier than females. For
these reasons, at the moment we do not have
good evidence of what the magnitude of the sex
ratio bias is in other social spiders or whether it
represents a theoretically interesting bias.

Given our current knowledge about the social
behavior and population structure of other social
spiders, however, our prediction is that their adult
sex ratio bias will also be found to result from
uneven parental investment. This prediction can
now be easily tested using the cytogenetic tech-
nique that we present here. It should be noted,
however, that in species in which parental care
extends beyond conception, measuring sex ratio
in embryos will not necessarily tell us all we want
to know about parental investment. Whether or
not biased at conception, the proportion of male
to female offspring, or their relative sizes, may
change during the period of parental care as a
result of differential mortality or differential al-
location of resources. To determine whether this
is the case, the sex ratio at the end of the period
of parental investment should also be estimated.
In social spiders, parental investment can be pre-
sumed to end at the instar at which the spider-
lings start to participate in the activities of the
colony and are therefore less dependent on the
parental generation (in A. eximius, this occurs at
about the same time when males are first rec-
ognizable due to their enlarged palpi, Aviles
1986). If this young preadult sex ratio is found
to be different from the embryo sex ratio, then
parental investment would need to be estimated
by integrating the numbers of male and female

offspring and the per capita investment in them
over the period of parental care. If the sex ratio
values are the same, and no size difference be-
tween the sexes is obvious, as has been found to
be the case in A. eximius , then the investment
ratio can be estimated from the numerical ratio
either among embryos or among young pread-
ults. In studies in which the sex ratio of a large
number of colonies needs to be estimated, mea-
suring the preadult sex ratio may be the only
feasible alternative. The preadult sex ratio, how-
ever, is probably more subject to empirical error
because, through time, random mortality or
asynchrony in the development times of the sex-
es would tend to increase the variance of the sex
ratio estimate. Aside from being perhaps more
accurate, embryo sex ratios have the additional
advantage of allowing an assessment of whether
there is variation for the sex ratio among the
progeny of different females (given that eggs of
a clutch are laid together in a sac).

The importance of knowing the primary sex
ratio is certainly not limited to social spiders
since issues of sex ratio and population structure,
sex specific demographic phenomena and sex ra-
tio variation are of general interest in spider bi-
ology. The cytogenetic technique that we present
here greatly simplifies the estimation of the pri-
mary sex ratio in spiders. It not only has obvious
advantages over time consuming egg-rearing
techniques which risk producing a biased esti-
mate if there is mortality (Fiala 1980), but it also
has several advantages over a previously de-
scribed technique for obtaining spider embryo
chromosomes (Matsumoto 1977; Tugmon et al.
1 990): it allows reliable preparation of individual
eggs and it is fast enough that population studies
become feasible. This technique has already been
successfully used in mites (M. Kaliszewski pers.
comm.) and it can probably be used with equal
success in other arthropods (see Crozier 1968 for
a more laborious technique used for insect pu-
pae). Widespread use of this technique will make
available quantitative sex ratio estimates of a
phylogenetically diverse set of species, so that
comparative studies to test specific predictions
of sex ratio and population structure become
possible. Should primary sex ratio biases be con-
firmed in the social species for which biased adult
sex ratios have been reported, we would then be
faced with the interesting question of how species
in five different spider families (Agelenidae,
Dyctinidae, Eresidae, Theridiidae and Thomis-

AVILES & MADDISON-SOCIAL SPIDER SEX RATIO

133

idae) have solved the common problem of de-
vising a mechanism by which to beat the odds
imposed by the meiotic process (Williams 1 979).

ACKNOWLEDGMENTS

We are very grateful for the support and the
laboratory facilities provided by the Dept, of Bi-
ology, Pontificia Universidad Catolica del Ec-
uador and the Museo Ecuatoriano de Ciencias
Naturales, where most of the chromosome slides
were prepared. G. Estevez assisted in the field
and in the laboratory, and J. Campana, in part
of the chromosome scoring; we thank them both
for their diligence and enthusiasm. J. Patton and
the Museum of Vertebrate Zoology generously
allowed us to use their cytogenetics lab at U. C.
Berkeley to carry out most of the scoring and
photographic work. M. Slatkin provided us shel-
ter. A. Rypstra, F. Vollrath and O. Krauss re-
viewed a previous version of the manuscript and
gave valuable comments for its improvement.
This work was supported by grants from the Na-
tional Geographic Society (#38 1 2-88), Sigma Xi,
and Harvard University (OEB graduate student
grants) to L. Aviles, and by an NSERC Canada
postdoctoral fellowship to W. Maddison.

LITERATURE CITED

Aviles, L. 1983. Proporcion de sexos en la arana
social Anelosimus eximius. Licenciatura thesis,
Pontificia Universidad Catolica del Ecuador, Quito.
Aviles, L. 1986. Sex ratio bias and possible group
selection in the social spider Anelosimus eximius.
Am. Nat., 128:1-12.

Brach, V. 1975. The biology of the social spider A ne-
losimus eximius (Araneae: Theridiidae). Bull. South.
Calif. Acad. Sci., 74:37^1.

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

Bradoo, B. L. 1972. Some observations on the ecol-
ogy of social spider Stegodyphus sarasinorum Karsch
(Araneae: Eresidae) from India. Oriental Insects, 6:
193-204.

Bradoo, B. L. 1975. The sexual biology and mor-
phology of the reproductive organs of Stegodyphus
sarasinorum Karsch (Araneae: Eresidae). Entomol.
Monthly Mag., 111:239-247.

Buskirk, R. 1981. Sociality in the Arachnida. Pp.
281-359, In Social Insects (H. R. Hermann, ed.)
vol. II. Academic Press, New York.

Chamov, E. 1982. The Theory of Sex Allocation.

Princeton University Press, Princeton.

Christenson, T. 1 984. Behaviour of colonial and sol-
itary spiders of the theridiid species Anelosimus ex-
imius. Anim. Behav., 32:725-734.

Crozier, R. H. 1968. An acetic acid dissociation, air-
drying technique for insect chromosomes, with ace-
to-lactic orcein staining. Stain Technology, 43: 1 7 1-
173.

Darlington, C. D. & L. F. La Cour. 1975. The Han-
dling of Chromosomes, Sixth Ed., Wiley and Sons,
New York.

Darchen, R. 1967. Une nouvelle araignee sociale du
Gabon Agelena republicana Darchen. Biol. Gabon-
ica, 3:31-42.

Darchen, R. 1978. Les essaimages de I’araignee so-
ciale, Agelena consociata Denis, dans la foret ga-
bonaise. C. R. Hebd. Seances Acad. Sci. Ser. D Sci.
Nat. 287:1035-1037.

Fiala, K. 1980. On estimating the primary sex ratio
from incomplete data. Am. Nat., 1 15:442-^44.

Fisher, R. A. 1930. The Genetical Theory of Natural
Selection. Dover, New York.

Fowler, H. G. & H. W. Levi. 1979. A new quasisocial
Anelosimus spider (Araneae: Theridiidae) from Par-
aguay. Psyche, 86:1 1-17.

Frank, S. 1987. Demography and sex ratio in social
spiders. Evolution, 41:1267-1281.

Hamilton, W. D. 1 967. Extraordinary sex ratios. Sci-
ence, 156:477-488.

Herre, E. A. 1985. Sex ratio adjustment in fig wasps.
Science, 228:896-898.

Jackson, R. R. & S. E. Smith. 1978. Aggregations of
Mallos and Dictyna (Araneae: Dictynidae): popu-
lation characteristics. Psyche, 85:65-80.

Jacson, C. C. & K. J. Joseph. 1973. Life history,
bionomics and behaviour of the social spider Ste-
godyphus sarasinorum. Insectes Sociaux, 20:189-
203.

Kullman, E., St. Nawabi & W. Zimmermann. 1971.
Neue Ergebnisse zur Brutbiologie cribellater Spin-
nen aus Afghanistan und der Serengeti. Zeitschr.
Kolner Zoo, 14:87-108.

Leigh, E. 1970. Sex ratio and differential mortality
between the sexes. Am. Nat., 104:205-206.

Levi, H. W. 1956. The spider genera Neottiura and
Anelosimus in America (Araneae: Theridiidae).
Trans. Amer. Micros. Soc., 75:407^22.

Levi, H. W. 1963. The American spiders of the genus
Anelosimus (Araneae, Theridiidae). Trans. Amer.
Micros. Soc., 82:30-48.

Levi, H. W. & D. R. Smith. 1982. A new colonial
Anelosimus spider from Suriname (Araneae: Ther-
idiidae). Psyche, 89:3-4.

Lubin, Y. D. in press. Patterns of variation in female-
biased sex ratios in a social spider. Biol. J. Linnean
Soc.

Lubin, Y. D. & R. H. Crozier. 1985. Electrophoretic
evidence for population differentiation in a social
spider Achaearanea wau (Theridiidae). Insectes So-
ciaux, 32:297-304.

Lubin, Y. D. & M. H. Robinson. 1982. Dispersal by
swarming in a social spider. Science, 216:319-321.

Main, B. Y. 1988. The biology of a social thomisid

134

THE JOURNAL OF ARACHNOLOGY

spider. Australian Arachnology [Austral. Ent. Soc.
Misc. PubL, Brisbane], 5:55-73.

Maddison, W. 1982. XXXY sex chromosomes in
males of the jumping spider genus Pellenes (Ara-
neae: Salticidae). Chromosoma (Berl.), 85:23-37.

Matsumoto, S. 1977. An observation of somatic
chromosomes from spider embryo-cells. Acta Ar-
achnol., 27:167-172.

Nentwig, W. & T. Christenson. 1986. Natural history
of the non-solitary sheetweaving spidzr Anelosimus
jucundus (Araneae: Theridiidae). Zool. J. Linnean
Soc., 87:27-35.

Overal, W. L. & P. R. Ferreira da Silva. 1982. Pop-
ulation dynamics of the quasisocial spider Anelo-
simus eximius (Araneae: Theridiidae). Pp. 181-182,
In The biology of Social Insects (M. E. Breed, C. D.
Michener & H. E. Evans, eds). Proc. of the IX Congr.
Int. Union for the Study of Social Insects, Boulder,
Colorado.

Pain, J. 1964. Premieres observations sur une espece
nouvelle d’araignees sociales, Agelena consociata.
Biol. Gabonica, 1:47-58.

Remington, R. D. & M. A. Schork. 1985. Statistics
with Applications to the Biological and Health Sci-
ences. Prentice Hall, New Jersey.

Riechert, S., R. Roeloffs & A. C. Echtemacht. 1986.
The ecology of the cooperative spider Agelena con-
sociata in Equatorial Africa (Araneae, Agelenidae).
J. Arachnol., 14:175-191.

Rypstra, A. L. & R. S. Tirey. 1989. Observations on
the social spider, Anelosimus domingo (Araneae,
Theridiidae), in southwestern Peru. J. Arachnol., 17:
368.

Seibt, U. & W. Wickler. 1988. Bionomics and social
structure of “Family Spiders” of the genus Stego-
dyphus, with special reference to the African species
S. dumicola and S', mimosarum (Araneidae, Eresi-
dae). Verb, naturwiss. Ver. Hamburg, 30:255-303.

Smith, D. R. 1982. Population genetics oi Anelosi-
mus eximius (Araneae, Theridiidae). J. Arachnol.,
14:201-217.

Suzuki, S. 1954. Cytological studies in spiders. III.
Studies on the chromosomes of fifty-seven species
of spiders belonging to seventeen families, with gen-
eral considerations on chromosomal evolution. J.
Sci. Hiroshima Univ., ser. B, 15:23-136.

Tapia, Y. & T. De Vries. 1980. Tolerancia y coop-
eracion en la arana social Anelosimus jucundus del
bosque tropical Rio Palenque, Ecuador. Rev. de la
Univ. Catolica Quito, VII(27):5 1-74.

Trivers, R. 1985. Social Evolution. The Benjamin/
Cummings Publishing Company, Inc.

Tugmon, C., J. D. Brown & N. V. Homer. 1990.
Karyotypes of seventeen USA spider species (Ara-
neae, Araneidae, Gnaphosidae, Loxoscelidae, Ly-
cosidae, Oxyopidae, Philodromidae, Salticidae and
Theridiidae). J. Arachnol., 18:41-48.

Vollrath, F. 1982. Colony foundation in a social spi-
der. Z. Tierpsychol., 60:313-324.

Vollrath, F. 1986a. Environment, reproduction and
the sex ratio of the social spider Anelosimus eximius
(Araneae, Theridiidae). J. Arachnol., 14:266-281.

Vollrath, F. 1986b. Eusociality and extraordinary sex
ratios in the spider Anelosimus eximius (Araneae:
Theridiidae). Behav. Ecol. SociobioL, 18:283-287.

Waage, J. K. 1 982. Sib-mating and sex ratio strategies
in scelionid wasps. Ecol. Entomol., 7:103-1 12.

Werren, J. H. 1980. Sex ratio adaptations to local
mate competition in a parasitic wasp. Science, 208:
1157-1159.

White, M. J. 1973. Animal cytology and evolution,
3rd. ed. Cambridge University Press, Cambridge.

Williams, G. C. 1979. The question of adaptive sex
ratio in outcrossed vertebrates. Proc. R. Soc. Lond.
B, 205:567-580.

Wilson, D. S. & R. K. Colwell. 1981. Evolution of
sex ratio in structured demes. Evolution, 35:882-
897.

Manuscript received August 1990, revised February
1991.

APPENDIX

PROTOCOL TO OBTAIN SPIDER EMBRYO
CHROMOSOME PREPARATIONS

1. Fix the egg. With a very fine needle, poke a small
hole in the egg and with the egg so skewered, place it
in a drop of fixative (3 parts absolute ethanol: 1 part
glacial acetic acid). We used electrochemically-sharp-
ened tungsten wire needles. Limb buds, if present, ap-
pear as a series of small white lumps as the fixative
enters the egg (a black background enhances visibility).
Remove the tip of the needle from the egg, and press
on the side of the egg so as to force the contents through
the small hole. If the egg is young enough the contents
can be squirted into a long thin string, which aids in
rapid fixing and later in breaking the contents in small
pieces. Tissue with nuclei is white; yolk without nuclei
is yellowish; if there is much yolk then some can be
discarded. Discard the empty chorion. Fix for 30 sec-
onds.

2. Squash the tissue. Place the fixed contents of the
egg in a small drop of 60% acetic acid on a microscope
slide. With two very fine needles, break the tissue into
small pieces. Place a cover slip on top, and squash the
tissue flat. This squashing is perhaps the most critical
step in the procedure: squashing too softly, sliding the
cover slip sideways while squashing, and air bubbles
should all be avoided.

3. Remove the cover slip and let dry the tissue. Freeze
the slide on dry ice at least several minutes and flip
the cover slip off with a razor blade. Wash off the acetic
acid for 20 seconds in a bath of absolute ethanol. Let

AVILES & MADDISON-SOCIAL SPIDER SEX RATIO

135

the slide dry at least ten minutes. If needed, the slide
can be left in this condition overnight or longer.

4. Stain the tissne and make permanent the prepa-
ration. After the slide has dried well, it can be stained
and made permanent. No doubt many different stains

could be used; we have used primarily a 3-4 minute
bath of acetocarmine. After staining, the slide can be
rinsed with appropriate solvents to prepare it for per-
manent mounting.

1991. The Journal of Arachnology 19:136-149

ULTRASTRUCTURE OF THE PRIMARY MALE GENITAL
SYSTEM, SPERMATOZOA, AND SPERMIOGENESIS OF
HYPOCHILUS POCOCKI (ARANEAE, HYPOCHILIDAE)

Gerd Alberti: Institute of Zoology I, University of Heidelberg, Im Neuenheimer Feld
230, D-6900 Heidelberg, F.R.G.

Frederick A. Coyle: Department of Biology, Western Carolina University, Cullowhee,
North Carolina 28723 USA

Abstract. Spermiogenesis and the ultrastructure of the testes, vasa deferentia, and sper-
matozoa of Hypochilus pococki, a palaeocribellate spider, are described. The sperm exhibit
many character states apparently plesiomorphic for spiders (an acrosomal complex com-
posed of a cone-shaped acrosomal vacuole and an acrosomal filament running to the end
of a nuclear canal, a postcentriolar nuclear elongation, an axonema with a 9x2 -f- 3 pattern,
centrally located mitochondria, and a rolling-up of the nucleus and flagellum at the end of
spermiogenesis) and two states that may be synapomorphic for araneomorphs (a stout
nucleus and a very pronounced nuclear elongation). The spermatozoa are delivered as
cleistospermia (individual spherical sperm cells each encapsulated in a secretory sheath). It
is argued that, for spiders, coenospermia are plesiomorphic and cleistospermia apomorphic.

The considerable effort devoted to under-
standing relationships among spider families has
produced varying views on the higher classifi-
cation of spiders. The most commonly accepted
classification scheme is that developed by Plat-
nick and Gertsch (1976) and Platnick (1977), and
further supported and developed by Forster et
al. (1987) and Coddington (1990). These cladistic
analyses favor a classification of spiders into two
suborders: Mesothelae and Opisthothelae, with
the latter composed of two groups, Mygalomor-
phae and Araneomorphae. The araneomorphs
contain the Paleocribellatae (consisting only of
the Hypochilidae) and its sister group, the Neo-
cribellatae (containing all other araneomorphs).
Examples of alternative views are those of 1)
Lehtinen (1967, 1978, 1986), who has argued that
mygalomorphs are more closely related to the
Mesothelae than to araneomorphs and that the
Araneomorphae is probably not monophyletic
(in particular he suggests that a cribellum might
have evolved independently in the filistatids) and
2) Eskov and Zonshtein (1990), whose recent cla-
distic analysis causes them to classify spiders into
two suborders which are very different from those
of Platnick and Gertsch (1976), i.e.; the Orthog-
natha (including the Liphistiomorphae, Thera-
phosomorphae, and Filistatomorphae) and the
Labidognatha (including the Geralycosomor-

phae, Dysderomorphae, Hypochilomorphae, and
Araneomorphae).

Since the discovery by Franzen (1956) that
sperm morphology is correlated with mode of
insemination, it has become evident that sperm
ultrastructure provides a rich source of charac-
ters for testing hypotheses of relationship (see
e.g.; Baccetti and Afzelius 1976, Franzen 1977,
Afzelius 1979, Baccetti 1979, 1985). Our knowl-
edge of spider sperm ultrastructure has increased
considerably since the classic paper of Osaki
(1969) (Baccetti 1970; Reger 1970; Rosati et al.
1970; Osaki 1972; Boissin 1973; Lopez & Boissin
1976; Juberthie et al. 1981; Lopez et al. 1983; Al-
berti & Weinmann 1985; Alberti et al. 1986). It
is of particular interest that four types of sperm
packaging have been found in spiders: 1) coe-
nospermia (capsules containing many individual
unfused sperm cells), 2) cleistospermia (individ-
ual sperm cells, each surrounded by its own
sheath), 3) synspermia (several fused sperm cells
forming a syncytium which is surrounded by a
sheath), and 4) so-called “spermatophores” (tubes
of secretion containing a row of highly ordered
individual sperm cells) (see Alberti 1990).

Coenospermia have been found electron mi-
croscopically in Mesothelae, Mygalomorphae,
and Filistatidae (Alberti & Weinmann 1985; Al-
berti et al. 1986; Alberti 1990). (Although Tuz-

136

ALBERTI & COYLE-SPERM ULTRASTRUCTURE OF HYPOCHILUS

137

et and Manier [1959] described in the clubionid,
Cheiracanthium sp., a “spermatophore” which
clearly represents a coenospermium, Alberti
(1990) could find only cleistospermia in Cheir-
acanthium punctorium.) Synspermia have been
observed only in certain haplogyne families (Se-
gestriidae, Dysderidae, Scytodidae, and Loxos-
celidae) (Alberti & Weinmann 1985; Alberti
(1990). The curious “spermatophores” have only
been found in Telemidae (Juberthie et al. 1981),
but may also be present in some other families
(e.g., Oonopidae, Tetrablemmidae; Brignoli 1978).
Cleistospermia have been found in all other ar-
aneomorph spiders which have been examined.

The Hypochilidae, which is universally re-
garded as a primitive taxon and believed by many
to be the sister group of all other araneomorph
spiders, is of special importance in understand-
ing spider phylogeny. In the following we de-
scribe, for the first time, the ultrastructure of
spermiogenesis, sperm cells, testes, and vasa de-
ferentia of a hypochilid, Hypochilus pococki Plat-
nick, and discuss some of the phylogenetic im-
plications of these results, with special attention
to the evolutionary polarity of modes of sperm
packaging.

MATERIALS AND METHODS

Three adult males of H. pococki were collected
on 9 October 1989 near Wolf Creek, 8 km south
of Cullowhee, Jackson Co., North Carolina. On
10 October the specimens were dissected and
fixed in cold 3.5% glutaraldehyde buffered at pH
7.4 (Sorensen-phosphate buffer) for 2 hours. The
fixative was then diluted with buffer solution (1:
4) and in this state the tissues were mailed to
Heidelberg where the specimens were rinsed with
buffer and postfixed for 2 hours with 2% buffered
OSO4 solution. Following further rinsing with
buffer the material was dehydrated with graded
ethanols and embedded in Araldite using pro-
pylenoxide as an intermedium. Ultrathin sec-
tions were obtained using a Reichert OM-U2
ultramicrotome. The sections were stained with
uranylacetate and lead citrate and were observed
with a Zeiss EM lOCR electron miroscope.

RESULTS

Testes.— The massive, elongate paired testes
are located ventrally in the opisthosoma. Sper-
miogenesis begins in cysts containing numerous
germ cells in the same stage of development.
Later the cysts become less compact and may be
confluent (Figs. 1 , 2). These cysts are formed by

extensions of large somatic cells characterized by
their irregular shape, interdigitations with neigh-
boring somatic cells, and numerous desmosomes
(Figs. 3-5, 24). Each somatic cell also contains
a large, irregularly shaped, electron lucent nucle-
us, numerous conspicuous cistemae of rough ER,
dictyosomes, several small mitochondria, and
various inclusions which are probably lyso-
somes. Some somatic cells are quite dense and
show signs of degeneration (dilated ER cistemae,
disintegrating cell apex, etc.) (Fig. 4). Narrow
extensions of the somatic cells reach between
germ cells only in an early stage of development.
Each testis is underlain by a thick basal lamina
supported by a similarly thick layer of collage-
nous fibres, and a muscularis adjacent to the hae-
mocoel (Fig. 3).

Vasa deferentia. — In the proximal parts of the
vasa deferentia, which are continuous with the
distal parts of the testes, the epithelial cells are
similar to the somatic cells of the testes, and the
basal components (basal lamina, collagenous
layer, muscularis) are similarly present. The lu-
men of the proximal region contains a homo-
geneous secretion and coiled spermatozoa which
are not yet surrounded by a secretory sheath.
This encystment occurs in the distal vas deferens
where the secretory activity is much higher. The
lumen here in the distal portion is filled with a
densely staining material, aggregates of granules
and spheres, and streaks which probably consist
of the same material that forms the sheaths of
the sperm cells (Figs. 6, 10). Large nuclei with
folded surfaces, cistemae of rough ER, many dic-
tyosomes, and secretory vesicles are conspicuous
in the epithelial cells. The apical surface of these
cells is provided with microvilli with many small
vesicles at their bases. Laterally these cells are
connected by extensive junctional complexes (Fig.
6). The basal plasmalemma of the cells is folded
and attached via hemidesmosomes to a rather
thin basal lamina (Fig. 7). The collagenous layer
is also thinner than in the testis. Within the mus-
cularis are many nerve endings (Fig. 8).

Spermatozoa. — The mature spermatozoa
found in the distal vas deferens are spherical (Fig.
9) and each one is encapsulated in a multilayered
0.3 Mm thick secretory sheath. The following
structures are observable in each sperm cell:

Acrosomal complex: The acrosomal com-
plex is composed of an acrosomal vacuole (or
vesicle) in the shape of an elongate hollow cone
and an acrosomal filament (Figs. 11, 12). The
acrosomal vacuole is located at the anterior end

138

THE JOURNAL OF ARACHNOLOGY

Figures 1, 2.— Spermiogenesis in Hypochilus pocockL 1, part of a cyst within the testis containing spermatids
in a very early stage of spermiogenesis. Note that cells are densely packed (arrows point to flagella of spermatids
located within narrow intercellular clefts). Chromatin condensation has just started. Acrosomal vacuoles are
already dense. X6,300. 2, advanced stages of spermiogenesis (compare Fig. 20). Cytoplasm of the spermatids
is rather electron lucent. Note extensive intercellular spaces between spermatids and nearly mature spermatozoa
(upper left). X4,000. AV = acrosomal vacuole, N = nucleus.

of the nucleus, runs parallel to the cell surface,
and does not protrude from the cell body. The
contents of the acrosomal vacuole are rather ho-
mogeneous and electron dense; only the inner-
most region shows distinct layers (Fig. 1 2). The
acrosomal filament, composed of many (actin?)
subfibers (Figs. 12, 22, 23), is rather thick and

runs through the subacrosomal space into the
nuclear canal down to its end (Figs. 11, 13).

Nucleus: The nucleus is the most prominent
structure of the sperm cell and is bent upon itself.
The main part of the nucleus is rather stout, it
describes approximately two-thirds of a circle
within the capsule (Fig. 9), and its convex pe-

ALBERTI & COYLE-SPERM ULTRASTRUCTURE OF HYPOCHILUS

139

Figures 3-5.— Spermiogenesis in Hypochilus pococki: 3, somatic cells from the testis close to proximal part of
vas deferens. Arrows point to basal lamina, arrow heads to collagenous layer. X6,300. 4, margin of cyst which
contains nearly mature spermatozoon. Note dense (degenerating) somatic cell (arrow). X12,600. 5, dictyosome
and rough ER close to nucleus of somatic cell. X20,000. D = dictyosome, ER = rough endoplasmic reticulum,
MU = muscle cell, N = nucleus, NO = nucleolus, SP = spermatozoon.

140

THE JOURNAL OF ARACHNOLOGY

Figures 6-8.— Spermiogenesis in Hypochilus pococki: 6, distal vas deferens containing dense secretion. Note
irregularly shaped nucleus, numerous vesicles, and extensive cell junctions (arrow heads). X9,450. 7, basal region

ALBERTI & COYLE-SPERM ULTRASTRUCTURE OF HYPOCHILUS

141

ripheral surface is smooth whereas its concave,
centrally directed surface is folded. The nucleus
extends beyond the axonemal base as a flattened,
so-called postcentriolar, nuclear elongation which
completes the circle and continues parallel to the
acrosomal vacuole reaching even behind the ax-
onemal base (Figs. 9, 11, 13). The nuclear elon-
gation is thus rather long (about one complete
circle). The close apposition of the nuclear elon-
gation to the acrosomal vacuole is quite distinc-
tive (Figs. 11, 13). The nucleus contains a pe-
ripheral canal which curves around its whole
length and contains the acrosomal filament (Figs.
11, 13).

Axonema: The implantation fossa, which is
moderately deep, includes, in front of the cen-
trioles, a homogeneous material (centriolar ad-
junct) and numerous small granules, presumably
glycogen (Figs. 9, 11, 13). The axonemal base is
marked by dense material which is opposed to
the peripheral tubules (Figs. 14, 16). The center
of the distal centriole includes three dense fibers
which are continuous with the dense material
connecting the three central tubules of the axo-
nema, which thus exhibits the 9x2+3 pattern
typical for most spiders (Figs. 14-17). The ax-
onema, which lacks a flagellar membrane, de-
scribes four to five coils within the cell body
(Figs. 13, 17). The A-tubules are denser proxi-
mally than distally (Fig. 17). Only in the very
distal part of the axonema are the central tubules
lacking.

Additional components: Some mitochondria
are found in the center of the cell together with
irregularly arranged dense streaks (Figs. 9, 11,
1 3). These streaks are probably condensed mem-
branes as is indicated by younger stages of sperm
development (see below; Fig. 24). Regions of cy-
toplasm devoid of organelles are studded with
moderately dense granules, which most likely
represent ^S-gly cogen (Fig. 9).

Spermiogenesis.— Young spermatids at the
beginning of chromatin condensation are densely
packed cells with spherical nuclei. The acrosomal
vacuole was already present in the stages we ob-
served and is located opposite a region of the

nucleus where chromatin condensation starts
(Figs. 1, 18). The acrosomal vacuole is slightly
inclined against the long axis of the cell and the
acrosomal filament thus runs obliquely into the
nuclear canal (Fig. 1 8). The nucleus is surround-
ed by a manchette of microtubules. Opposite the
acrosomal vacuole the nucleus invaginates to
form the implantation fossa (Figs. 19, 21). At
this pole of the cell, the mitochondria have as-
sembled. The flagellum extends into the inter-
cellular clefts left between the developing sper-
matids which are interconnected by narrow cell
bridges (Fig. 1). Some extensions of somatic cells
are also visible between the spermatids.

In a later stage, the nucleus narrows somewhat
towards its anterior end (Fig. 1 9). The acrosomal
vacuole is basally surrounded by a girdle of dense
material to which the manchette microtubules
are attached (Figs. 20, 22). The implantation fos-
sa is quite deep and includes the centrioles in a
tandem orientation (Fig. 21). The chromatin is
filamentous now. The surface of the cell becomes
irregular and between the cells are found more
extensive intercellular spaces containing a het-
erogeneous material (Figs. 2, 20). It appears as
if large quantities of cytoplasmic material are
discarded from the spermatids, predominately
from those regions which are close to the cell
bridges. These discarded “blebs” often include
large complexes of cistemae in different stages of
destruction (Fig. 21).

In this stage the cells have already elongated
and the nucleus exhibits a prominent nuclear
elongation (Fig. 20). The flagellum is partly sunk-
en into the cell and is consequently surrounded
by an invagination of the flagellum to form a so-
called flagellar tunnel (Figs. 20, 23). The nuclear
envelope shows distinct nuclear pores in its pos-
terior part (Figs. 19, 20). The acrosomal vacuole
is somewhat more dense, and the nuclear canal
with the acrosomal filament is complete, ap-
pearing as a prominent ridge on the surface of
the nucleus (Figs. 20, 21, 24). The mitochondria
are still located at the posterior end of the cell.
The cytoplasm is rather electron lucent, includ-
ing only few dense granules. Remarkably, we did

with deep infoldings of plasmalemma composing a basal labyrinth. The cells are attached to basal lamina with
hemidesmosomes (arrow heads). X16,000. 8, nerve ending at muscle cell underlying epithelium of vas deferens.
X30,000. M = mitochondrion, MU = muscle cell, N = nucleus.

Figures 9, 10.— Spermatozoa of Hypochilus pococki: 9, mature coiled spermatozoon within lumen of distal vas
deferens. Note multilayered secretory sheath. Arrows point to acrosomal filament. X32,000. 10, different secre-
tions within vas deferens. X25,000. AV = acrosomal vacuole, AX = axonema, CA = centriolar adjunct within
implantation fossa, GLY = glycogen, M = mitochondrion, N = nucleus.

ALBERTI & COYLE-SPERM ULTRASTRUCTURE OF HYPOCHILUS

143

Figures 11, 12. —• Spermatozoa of Hypochilm pococki: 11, nearly mature spermatozoon from testis. Note acrosomal
vacuole sectioned longitudinally, postcentriolar nuclear elongation, centriolar adjunct within implantation fossa,
numerous mitochondria, membranes and cistemae, glycogen, and axonema. Arrows point to acrosomal filament.
X30,000. 12, transverse section through acrosomal complex. Note concentric layers within acrosomal vacuole
close to subacrosomal space containing acrosomal filament, which is composed of subfibers. Dense cistemae
are opposed to acrosomal vacuole (arrow heads). X60,000. AV = acrosomal vacuole, AX- = axonema, CA =
centriolar adjunct, GLY = glycogen, M = mitochondrion, ME = membranes, N = nucleus, NE = nuclear
elongation.

not observe any membranous material within
the cytoplasmic matrix in this or earlier stages
(Figs. 2, 20).

Finally the chromatin condenses completely
to an almost totally dark structure leaving ex-
tensive areas of electron lucent nucleoplasm sur-
rounded by the nuclear envelope (Fig. 23). Also
in this stage there are no (other) membranes pres-
ent vrithin the cytoplasmic matrix; these only
appear after the coiling process, which presum-
ably occurs rapidly since no intermediate stages
were found. In these nearly mature sperm cells
distinct dense cistemae were found, some of
which parallel the nucleus and acrosomal vac-
uole (Figs. 24, also 11, 13). The cytoplasm is

rather homogeneous but some (glycogen?) gran-
ules are already concentrated at the axonemal
base (Figs. 11, 13). Further, in the center of the
cell is established a “dense body”, which later
becomes unrecognizable because of the general
condensation of the cytoplasm (Figs. 24, also 1 3).
The flagellum is completely incorporated, i.e.,
the axonema is without a flagellar membrane
(Figs. 24, also 11, 13, 17). The cell further con-
denses and finally achieves the stage of the ma-
ture spermatozoon. The cysts open and the sperm
cells, together with the intercellular fluid, are ex-
pelled into the lumen of the testis, which is es-
tablished by such confluent cysts and is contin-
uous with the lumen of the vas deferens.

144

THE JOURNAL OF ARACHNOLOGY

Figures 13-17.— Spermatozoa of Hypochilus pococki: 13, two nearly mature spermatozoa from testis. Note dense
nuclei with nuclear elongation and nuclear canal containing acrosomal filament (arrows). Membranes and
cistemae in part are parallel with nucleus. X20,000. 14, longitudinal section through axonemal base showing
modified distal centriole with central axis and proximal part of axonema surrounded by glycogen. X20,000. 15-
17, transverse sections: 15, proximal centriole within implantation fossa closely attached to centriolar adjunct.
X37,500. 16, axonemal base with central axis. Note accessory dense elements opposed to peripheral tubules.

ALBERTI & COYLE -SPERM ULTRASTRUCTURE OF HYPOCHILUS

145

DISCUSSION

The sperm cells of H. pococki are quite similar
to those of many other araneomorph spiders (Al-
berti 1990). Based upon observations of sper-
matozoa in other spiders (including Filistata in-
sidiatrix Forskal) and pedipalpate arachnids
(Uropygi and Amblypygi), we believe that two
of these similarities (the stout nucleus and the
very pronounced nuclear elongation) may be re-
garded as araneomorph synapomorphies and are
therefore supportive of Platnick and Gertsch’s
(1976) phytogeny. Many other Hypochiius char-
acter states appear to be plesiomorphic, e.g., the
presence of several mitochondria, membranous
material and a “dense body”, the high number
of glycogen granules, the rather simple implan-
tation fossa, the cone-shaped acrosomal vacuole,
and the acrosomal filament extending through
the whole length of the nuclear canal. These states
are also found in F. insidiatrix and many (other)
araneomorph spiders as well as in mygalo-
morphs, liphistiomorphs, and pedipalpate
arachnids (Alberti & Weinmann 1985; Alberti et
al. 1986; Alberti 1990; Osaki 1969; Phillips
1976; Jespersen 1978; Tripepi & Saita 1985; Al-
berti & Palacios- Vargas 1987). Thus our findings
are consistent with the interpretation that Hy-
pochilidae are ancient araneomorph spiders.

The discovery of cleistospermia in the hypo-
chilids leads us back to the question, first dis-
cussed by Bertkau (1877, 1878), of whether cleis-
tospermia or coenospermia are plesiomorphic for
spiders. Since only individual (not aggregated)
sperm cells have been found in uropygids and
amblypygids (the presumed sister group of spi-
ders), outgroup comparison supports the hy-
pothesis that cleistospermia are plesiomorphic.
If this is true, then coenospermia could be a syn-
apomorphy uniting the liphistiids, mygalo-
morphs, and filistatids (the only araneomorph
taxon in which coenospermia have been found
electron microscopically), a pattern consistent
with some of Lehtinen’s (1978) ideas and Eskov
and Zonshtein’s (1990) phytogeny.

If, on the other hand, coenospermia are prim-
itive for spiders, as Alberti & Weinmann (1985),

Alberti et al. (1986), and Alberti (1990) have
argued, cleistospermia could be a synapomorphy
for araneomorphs (with reversals in the filistatids
and Cheiracanthium sp.) or, more likely, cleis-
tospermia could have arisen two or more times
independently in the Araneomorphae. We favor
the hypothesis that coenospermia are plesiom-
orphic for spiders for the following reasons: 1)
Coenospermia have been found predominately
in spiders which are “primitive.” 2) Although
not found in amblypygids and uropygids, aggre-
gations of spermatozoa comparable to coenos-
permia are found in other arachnids (scorpions,
solpugids, opilionids, and certain mites) (Alberti
1990). 3) Given our current knowledge of the
distribution of coenospermia and cleistospermia,
and if, as our above-mentioned synapomorphies
suggest, the phytogeny of Platnick and Gertsch
(1976) is correct, it is more parsimonious to hy-
pothesize that coenospermia is the primitive state
(character state changes are required only within
the Araneomorphae) than that it is derived
(changes are required in the Mesothelae, Myga-
lomorphae, and Araneomorphae).

Other issues relevant to the evolution of modes
of sperm packaging deserve comment. First, the
observation that many cleistospermia are rather
similar to coenospermia, particularly with re-
spect to the (often multilayered) secretory sheath
and cellular components, and especially when
compared with the coenospermia of F. insidiatrix,
suggests that the evolutionary shift from coe-
nospermia to cleistospermia may be achieved
easily. The presence of coenospermia in one spe-
cies of Cheiracanthium (Tuzet & Manier 1959)
and of cleistospermia in another (Alberti 1 990)
also supports this hypothesis. Secondly, although
it is tempting to hypothesize that the syncytial
synspermia of some haplogynes and the multi-
cellular “spermatophores” of the telemids have
evolved from the multicellular coenospermia,
both types could also represent aggregates which
have originated independently from cleistosper-
mia (see Alberti 1990). Finally, in spite of the
greater amount of secretory sheath material
needed to package a given number of sperm as
cleistospermia rather than as coenospermia.

X24,000. 17, coiled axonema incorporated within cell body showing 9x2+3 pattern of tubules. In proximal part
of axonema, A-tubules are densely staining and central tubules are interconnected by dense material. X40,000.
AV = acrosomal vacuole, AX = axonema, CA = centriolar adjunct, DB = dense body, M = mitochondrion,
ME = membranes, N = nucleus, NE = nuclear elongation.

Figures 18-21. —Details of spermiogenesis in Hypochilus pococki: 18, early stage with acrosomal vacuole inclined
to longitudinal axis of nucleus. Acrosomal filament (arrow) runs obliquely into nuclear canal. X12,600. 19,
advanced stage with implantation fossa. Note nuclear pores at posterior part of nucleus (arrow heads). Fibrillar
chromatin condensation starts in posterior part of nucleus. X12,600. 20, more advanced stage. Chromatin is
completely fibrillar. A prominent postcentriolar nuclear elongation has developed. Note flagellar base with
flagellar tunnel sectioned only in its proximal part (arrow). Small arrow heads indicate nuclear pores. Nucleus
is provided with manchette microtubules attached to a dense girdle around the acrosomal vacuole (large arrow

Figures 22-24.— Spermatids of Hypochilus pococki: 22, tangentially sectioned acrosomal complex showing
acrosomal vacuole and acrosomal filament, composed of subfibers. Dense girdle with attached manchette mi-
crotubules. X48,000. 23, transverse section through flagellar tunnel around proximal part of flagellum (arrow
heads) of a spermatid. Note manchette microtubules. Arrow indicates nuclear envelope. X48,000. 24, spermatids
shortly after coiling and not completely condensed (note only 2 and 3 transverse sections of the axonema within
the cell in center of figure). Within cytoplasm dense cistemae are apparent now (arrow heads). Arrows point to
acrosomal filament. X12,600. AV = acrosomal vacuole, DB = dense body, M = mitochondrion, N = nucleus,
NE = nuclear elongation, SC = somatic cell.

heads). X16,000. 21, same stage as in Fig. 20 showing implantation fossa with centrioles in tandem position. At
right a cytoplasmic “bleb” containing irregular membranous network is detached from cell bridge region. X16,000.
AF = acrosomal filament, AV = acrosomal vacuole, CB = cell bridge, IF = implantation fossa, N = nucleus.

148

THE JOURNAL OF ARACHNOLOGY

transferring sperm as cleistospermia may be gen-
erally more effective, particularly in araneo-
morph genitalia, which often have narrower and
more sharply bent ducts than do nonaraneo-
morph genitalia. Cleistospermia are smaller than
coenospermia and therefore should be able to
move through such ducts with less resistence and
less chance of becoming stuck. A study of the
degree of correlation between sperm packet di-
ameter and genital duct diameter in spiders might
help test for such a functional relationship, one
which could play an important role in the evo-
lution of sperm packaging.

Spider sperm ultrastructure is evidently a rich
source of characters. As more taxa are examined
and as more is learned about the functional mor-
phology of spider genitalia (and therefore, per-
haps, the selective advantage of different types
of sperm), it may be possible to learn much about
spider phylogeny from comparative spermatol-
ogy.

LITERATURE CITED

Afzelius, B. A. 1979. Sperm structure in relation to
phylogeny in lower metazoa. Pp. 243-251, In The
Spermatozoan. (D. W. Fawcett & J. M. Bedford,
eds.). Urban and Schwarzenberg, Baltimore-Mu-
nich.

Alberti, G. 1990. Comparative spermatology of Ara-
neae. Acta Zool. Fenn., 190:17-34.

Alberti, G., B. A. Afzelius & S. M. Lucas. 1986. Ul-
trastructure of spermatozoa and spermatogenesis in
bird spiders (Theraphosidae, Mygalomorphae, Ara-
neae). J. Submicrosc. Cytol., 18:739-753.

Alberti, G. & J. G. Palacios-Vargas. 1987. Fine struc-
ture of spermatozoa and spermatogenesis of Schi-
zomus palaciosi, Reddell and Cokendolpher, 1986.
(Arachnida: Uropygi: Schizomida). Protoplasma,
137:1-14.

Alberti, G. & C. Weinmann. 1985. Fine structure of
spermatozoa of some labidognath spiders (Filista-
tidae, Segestriidae, Dysderidae, Oonopidae, Scyto-
didae, Pholcidae; Araneae; Arachnida) with re-
marks on spermiogenesis. J. MorphoL, 185:1-35.
Baccetti, B. 1970. The spermatozoa of Arthropoda.
IX. The sperm cell as an index of arthropod phy-
logeny. Pp. 169-181, In Comparative Spermatol-
ogy. (B. Baccetti, ed.). Academic Press, Orlando.
Baccetti, B. 1979. Ultrastructure of sperm and its
bearing on arthropod phylogeny. Pp. 609-644, In
Arthropod Phylogeny. (A. P. Gupta, ed.) Van Nos-
trand Reinhold, New York.

Baccetti, B. 1985. Evolution of the sperm cell. Pp.
3-58, In Biology of Fertilization. Vol. 2. Biology of
the Sperm. (C. B. Metz & A. Monroy, eds.). Aca-
demic Press, Orlando.

Baccetti, B. & B. A. Afzelius. 1976. The Biology of
the Sperm Cell. Monographs in Developmental Bi-
ology. Vol. 10. Karger, Miinchen.

Bertkau, P. 1877. Uber die Ubertragungsorgane und
die Spermatozoan der Spinnen. Verb. Naturh. Ver.
preuss. Rheinl., 34:28-32.

Bertkau, P. 1878. Versuch einer natiirlichen Anord-
nung der Spinnen nebst Bemerkungen zu einzelnen
Gattungen. Arch. f. Naturgesch., 44:351-410.

Boissin, L. 1973. Etude ultrastructurale de la sper-
miogenese de Meta bourneti Simon (Arachnides,
Araneides, Metinae). C.R. 2e Reun. Arachnol. Exp.
franc., Montpellier, 1973:7-22.

Brignoli, P. M. 1978. A few notes on a remarkable
South African troglobiotic spider, Congoderes lewisi
Harington, 1951 (Araneae, Telemidae). Rev. Suisse
Zool, 85:111-114.

Coddington, J. A. 1990. Ontogeny and homology in
the male palpus of orb-weaving spiders and their
relatives, with comments on phylogeny (Araneocla-
da: Araneoida, Deinopoidea). Smithson. Contrib.
Zool., (496): 1-52.

Eskov, K. Y. & S. L. Zonshtein. 1990. A new clas-
sification for the order Araneida (Arachnida: Chelic-
erata). Acta Zool. Fenn., 190:129-137.

Forster, R. R., N. 1. Platnick & M. R. Gray. 1987. A
review of the spider superfamilies Hypochiloidea
and Austrochiloidea (Araneae, Araneoraorphae).
Bull. Amer. Mus. Nat. Hist., 185:1-116.

Franzen, A. 1956. On spermiogenesis, morphology
of the spermatozoon, and biology of fertilization
among invertebrates. Zool. Bidrag., Uppsala, 31;
255-^82.

Franzen, A. 1977. Sperm structure with regard to
fertilization biology and phylogenetics. Verb. Dtsch.
Zool. Ges., 1977:123-138.

Jespersen, A. 1978. The fine structure of spermio-
genesis in the Amblypygi and Uropygi (Arachnida).
Zoomorphol., 89:237-250.

Juberthie, C., A. Lopez Sc J. Kovoor. 1981. Sper-
miogenesis and spermatophore in Telema tenella
Simon (Araneae: Telemidae) - An ultrastructural
study. Int. J. Invert. Reprod., 3:181-191.

Lehtinen, P. T. 1967. Classification of the cribellate
spiders and some allied families, with notes on the
evolution of the suborder Araneomorpha. Ann. Zool.
Fenn., 4:199-468.

Lehtinen, P. T. 1978. Definition and limitation of
supraspecific taxa in spiders. Symp. Zool. Soc. Lon-
don, 42:255-271.

Lehtinen, P. T. 1 986. Evolution of the Scytodoidea.
Pp. 149-157, In Proc. 9th Intern. Congr. Arachnol.
(W. G. Eberhard, Y. D. Lubin & B. C. Robinson,
eds.). Smithsonian Inst. Press, Washington, D. C.

Lopez, A. & L. Boissin. 1976. La spermatide d'Hol-
ocnemus pluchei (Scop.) (Arachnida, Araneida,
Pholcidae): Etude ultrastructurale. Bull. Soc. Zool.
France, 101:423-431.

Lopez, A., J.-C. Bonaric, & L. Boissin. 1983. Etude

ALBERTI & COYLE-SPERM ULTRASTRUCTURE OF HYPOCHILUS

149

ultrastructurale de la spermiogenese chez I’araignee
Pisaura mirabilis (Clerck, 1858) (Pisauridae). Rev.
ArachnoL, 5:55-64.

Osaki, H. 1969. Electron microscope study on the
spermatozoon of the liphistiid spider, Heptathela
kimurai. Acta ArachnoL, 22:1-12.

Osaki, H. 1972. Electron microscope study on sper-
miogenesis in the spider, Oxyopes sertatus. Japan.
J. ZooL, 16:184-199.

Phillips, D. M. 1976. Nuclear shaping during sper-
miogenesis in the whip scorpion. J. Ultrastr. Res.,
54:397.^05.

Platnick, N. I. 1977. The hypochiloid spiders: a cla-
distic analysis, with notes on the Atypoidea (Arach-
nida, Araneae). Amer. Mus. Novitates, (2627): 1-
23.

Platnick, N. 1. & W. J. Gertsch. 1976. The suborders
of spiders: a cladistic analysis. Amer. Mus. Novi-
tates, (2607): 1-1 5.

Raven, R. J. 1985. The spider infraorder Mygalo-

morphae: cladistics and systematics. Bull. Amer.
Mus. Nat. Hist., 182:1-180.

Reger, J. F. 1970. Spermiogenesis in the spider Pi-
saurina sp.: a fine structure study. J. Morphol., 1 30:
421-434.

Rosati, F., B. Baccetti & R. Dallai. 1970. The sper-
matozoa of Arthropoda. X. Araneids and the lower
myriapods. Pp. 247-254, In Comparative Sper-
matology. (B. Baccetti, ed.) Academic Press, New
York.

Tripepi, S. & A. Saita. 1985. Ultrastructural analysis
of spermiogenesis in Admetus pomilio (Arachnida,
Amblypygi). J. Morphol., 184:111-120.

Tuzet, O. & J. F. Manier. 1959. Recherches sur la
spermiogenese des araignees: Araneus cornutus, Te-
tragnatha pinicola, Philodromus poecilus, et Chir-
acanthium sp. Ann. Sc. Nat. ZooL, 12e ser.:9 1-103.

Manuscript received December 1990, revised March
1991.

1991. The Journal of Arachnology 19:150-152

RESEARCH NOTES

A TRAP TO CAPTURE BURROWING ARACHNIDS

In studies of population biology, it is often
necessary to determine size and reproductive sta-
tus of individuals and to mark them for later
recognition. This process should assure minimal
disturbance of study subjects and of their natural

surroundings. Current techniques for capturing
burrowing arachnids, however, often involve
disturbance, such as excavation. This destroys
burrows and risks injuring animals.

Several alternative techniques for capturing

burrow; C, door of trap; D, door-lever; E, balancing shaft; F, trigger-lever; G, trigger; H, locking hook; I, spider;
J, open vial to provide spider with shelter; K, slit to lock shaft; L, vial with live bait.

150

RESEARCH NOTES

151

large, burrowing, wandering spiders were tested
in a population study of Leucorchestris arenicola
Lawrence, a heteropodid (Henschel 1990).
When disturbance caused by excavation proved
unacceptable, pitfall traps were employed. Trap-
ping success was, however, low because spiders
usually detected and circumvented the edges of
pits. Furthermore, individual spiders could not
be targeted.

Therefore, I designed a container trap, de-
scribed here, which is cheap to make and easy
to operate. It capitalizes on a spider’s tendency
to probe when surrounded by a container. This
probing mechanically triggers closure of an ar-
tificial trapdoor that prevents the spider from
retreating into its burrow, thus capturing it inside
the container.

The sensitive trigger mechanism enables one
to capture burrowing arachnids having a mass
of 0.5 g or more. I have used it to capture more
than 1 00 spiders of two species and one scorpion
on surface slopes of 0-30° and in winds of 0-5
m/s.

The body of the trap (Fig. 1) is made of a
rectangular, flat-bottomed container (base ± 1 2
X 20 cm, height ±5 cm) with a transparent, air-
tight lid. A commercially available 2-liter plastic
container for food, such as an empty ice-cream
tub, is suitable. All other components besides
sample vials are made of 1 .5-mm-gauge stiff wire
and tape.

The description of components refers to labels
on Figure 1 . A hole of 4 x 4 cm (A) is cut into
the bottom near one end of the container. This
hole is larger than the natural trapdoor of a spider
burrow entrance (B) and is covered with a stiff
wire-rimmed 5x5 cm door (C), hinging on a
straight piece of wire attached to the body of the
trap. The door is held open by leaning a door-
lever (D), fixed to one side of the door, against
a balancing shaft (E) suspended across to the
other end of the trap. The heavier proximal end
of this balancing shaft rests on a trigger-lever (F)
connected to a wide, low-hanging trigger (G).

The trigger-lever and door-lever are circular
so that trigger sensitivity is less dependent on the
extent of overlap of contact points. If the trigger
is pushed only lightly (<0.1 g force = 9.806 x
10 ■’ Newtons), the heavier end of the balancing
shaft drops off the trigger-lever, moving the distal
end clear of the door-lever and the door closes
by gravity. Simultaneously, a broad hook (H)
drops onto the door to lock it (Fig. 2).

Figure 2.— Closed door of trap locked into place by
a wire hook.

The trigger is positioned away from the door
so that the spider (I) does not obstruct the slam-
ming door. Although a spider is capable of lifting
the door to enter its burrow, it cannot at the same
time lift the locking hook and the door. Deprived
of other shelter, it readily enters a darkened vial
(J) extending through the side of the trap. This
tube is later removed to manipulate the spider.

The trap has to be opened to set it. As the
trigger is very sensitive to wind until the lid is
closed, the balancing shaft can be locked into
position by forcing it into the narrowest top part
of a slit (K) in the wall of the trap. When the trap
is set and the lid closed, the balancing shaft is
loosened by lowering it into a wider section of
this slit until the balancing shaft is held only by
the trigger-lever.

Several factors increase trapping success.
Movements of live bait placed in a vial (L) out-
side the trap attracts the spider towards the trig-
ger. To overcome the spider’s initial reluctance
to step onto the artificial surroundings, the floor
of the trap is covered with sand. On slopes, the
trigger should be downhill of the door. In windy
conditions, shifting of the trap is prevented by
pegging it through its base behind the trigger.
Weight of the trigger and shapes of door- and
trigger-levers determine the minimum size of
arachnids that can be captured.

I thank the Foundation for Research Devel-
opment for funds, the Ministry of Wildlife Con-

152

THE JOURNAL OF ARACHNOLOGY

servation and Tourism of Namibia for facilities,
T. Harms for help and M. Seely for comments.

REFERENCES CITED

Henschel, J. R. 1990. The biology of Leucorches-
tris arenicola (Araneae: Heteropodidae), a burrow-
ing spider of the Namib Desert. In Current Research
on Namib Ecology— 25 Years of the Desert Eco-
logical Research Unit. (M. K. Seely, ed.). Transvaal

Museum Monograph No. 7. Transvaal Museum,
Pretoria, pp. 115-127.

Johannes R. Henschel: Desert Ecological
Research Unit of Namibia, P.O. Box 1592,
Swakopmund, Namibia

Manuscript received May 1990, revised August 1990.

1991. The Journal of Arachnology 19:153-154

NOVEL USE OF SILK BY THE HARLEQUIN
BEETLE-RIDING PSEUDOSCORPION,
CORDYLOCHERNES SCORPIOIDES
(PSEUDOSCORPIONIDA, CHERNETIDAE)

Pseudoscorpion use of silk in the construction
of nests for molting, brood production and hi-
bernation is well documented (Weygoldt 1969).
Silk for nest building is produced by glands in
the cephalothorax and is extruded through the
cheliceral galea (Chamberlin 1931). Males of
Serianus cawlinensis Muchmore also manufac-
ture a second type of silk in their rectal pocket
for use in the spinning of spermatophore signal
threads (Weygoldt 1966). Here, we describe two
additional functions of silk in the harlequin beetle-
riding pseudoscorpion, Cordylochernes scor-
pioides (L.).

Our research on the relationship between C.
scorpioides and Acrocinus longimanus (L.) has
established that the pseudoscorpion climbs un-
der the elytra of the large cerambycid to disperse
from old to newly-decaying trees (Zeh and Zeh
in prep.). Large males exploit this dispersal
mechanism by monopolizing beetle “subelytral
space” as a strategic site for intercepting and in-
seminating dispersing females. Whereas females
tend to disembark rapidly when beetles land on
fresh habitats, males may remain on beetles for
periods of at least two weeks.

An obvious tactical problem confronts these

Figures 1-4.— Beetle-riding tactics of the pseudoscorpion, Cordylochernes scorpioides: 1, two males each use
a chela to grasp an intertergal ridge of the harlequin beetle’s abdomen; 2, silken saftey harness connects male’s
pedipalpal chela with the beetle’s abdomen; 3, male on a silken, nest-like structure; 4, female uses silken thead
to descend from the beetle (lower right) while two males fight for control of the subelytral space (left).

153

154

THE JOURNAL OF ARACHNOLOGY

beetle-riding pseudoscorpions: they must remain
attached to the harlequin as it flies between trees.
Risk of detachment is exacerbated by the fact
that harlequin beetles fly with their bodies ori-
ented vertically (pers. obs.). The pseudoscor-
pions can avoid falling off by using their pedi-
palpal chelae to grasp intertergal ridges on the
beetle’s abdomen (Fig. 1). However, this method
is clearly inadequate for those males which re-
main on beetles for protracted periods in order
to defend mobile mating territories. Not only do
they experience numerous take-offs and flights,
but they must also have their pedipalpal chelae
free for mating. Males of C. scorpioides and all
other chemetids maintain their grasp on females
throughout mating (see Weygoldt 1969; Zeh
1987). Sudden flight of the beetle in response to
predation risk, for example, could put a mating
pseudoscorpion in danger of falling off.

Silk provides the solution to this dilemma. To
strap themselves securely to the beetle’s abdo-
men, males use their cheliceral galea to construct
silken “safety harnesses.” Initially, single threads
secure the pseudoscorpion’s chelae to the beetle’s
abdomen (Fig. 2). These are eventually elabo-
rated into a complex nest-like structure (Fig. 3).
Females have never been observed to make such
structures, although they do attach with single
threads.

In its interaction with the harlequin beetle, C.
scorpioides not only uses silk to stay on the beetle
but also as an aid to disembarkation. Pseudo-
scorpions can descend from their beetle host on
a silken thread (Fig. 4). By maintaining contact
with the beetle, this technique may provide dis-
persing individuals with the means to reconnoi-

ter new habitats and potentially use the thread
as a guide to re-board. We have not observed
dangling individuals climbing back up silken
threads (see Weygoldt 1969, fig. 20 for an ex-
ample of a pseudoscorpion climbing up a hair).
However, the thread can support the weight of
a pseudoscorpion (pers. obs.). A reconnoitering
pseudoscorpion could therefore remain attached
via the thread to a beetle which suddenly took
flight.

We thank W. B. Muchmore and V. Mahnert
for identifying the pseudoscorpions, V. F. Lee
and P. Weygoldt for useful comments on the
manuscript, and the Panamanian Institute Na-
cional de Recursos Naturales Renovables (IN-
RENARE) for permission to carry out the work.
Both authors gratefully acknowledge fellowship
support from the Smithsonian Tropical Research
Institute.

LITERATURE CITED

Chamberlin, J. C. 1931. The arachnid order Chelo-
nethida. Stanford Univ. Publ. Biol. Sci., 7:1-284.
Weygoldt, P. 1966. Spermatophore web formation
in a pseudoscorpion. Science, 153:1647-1649.
Weygoldt, P. 1969. The Biology of Pseudoscorpions.

Harvard Univ. Press, Cambridge.

Zeh, D. W. 1987. Aggression, density, and sexual
dimorphism in chemetid pseudoscorpions (Arach-
nida: Pseudoscorpionida). Evolution, 41:1072-1087.

David W. Zeh and Jeanne A. Zeh: Smithson-
ian Tropical Research Institute, APO Miami
34002-00 1 1 USA, or, Apartado 2072, Balboa,
Republica de Panama.

Manuscript received February 1991, revised April 1991.

1991. The Journal of Arachnology 19:155-156

ON THE SYNONYMY OF THAUMASTOBELLA MOUREI
MELLO-LEITAO AND ILDIBAHA ALBOMACULATA
KEYSERLING (ARANEAE, ARANEIDAE)

The little known araneid genus Thaumasto-
bella was created by Mello-Leitao in 1945. It is
monotypic, based on a single adult female from
Parana province in Brazil. Because the abdomen
of the type specimen is sclerotized, provided with
dorsal spines and a sclerotized ring around the
spinnerets, Mello-Leitao assigned the new genus
to the subfamily Gasteracanthinae. Since then,
little has been published on the genus. Brignoli
(1983) listed it as a genus on which nothing had
been published since the original description. Levi
(1985) pointed out that the status of the genus is
uncertain and that no representatives other than
the type specimen are known. Levi later in-
formed me (in litt. 1990) that the type specimen
is regarded as lost. However, based on the de-
scription and the three drawings provided by
Mello-Leitao (1945), Levi believed the type to
be an immature Micrathena.

In connection with an ongoing revision of the
genus Gasteracantha and a phylogenetic analysis
of the subfamily Gasteracanthinae I have been
looking for the type material of all gasteracan-
thine genera. Since Thaumastobella was origi-
nally placed in the gasteracanthines I also searched
for that material. In a recently published catalog
of type material deposited in “Museu de Historia
Natural “Capao da Imbuia”, Curitiba, Parana,
Brazil (Pinto-da-Rocha & De Fatima Caron 1 989)
I suddenly found the type species Thaumasto-
bella mourei listed under the family Salticidae.
The material was made available to me and an
examination revealed that Thaumastobella
mourei is in fact conspecific with Micrathena sac-
cata (C. L. Koch, 1836). Thaumastobella Mello-
Leitao, 1945 is therefore a junior synonym of
Micrathena Sundevall, 1833 and Thaumasto-
bella mourei Mello-Leitao, 1945 a junior syn-
onym of Micrathena saccata (C. L. Koch, 1 836).
The holotype matches perfectly with the descrip-
tion and illustrations of Micrathena saccata giv-
en by Levi (1985:490). Micrathena saccata is
already known from Brazil, but not further south
than Matto Grosso and this is therefore the
southernmost record of the species.

The generic name Ildibaha was synonymized
with Micrathena by Levi (1985) who also stated

that the type material of Ildibaha {Ildibaha al-
bomaculata Keyserling, 1892) is lost. Based on
Keyserling’s description and figure (Keyserling
1892:31, Tab. II fig. 29 & 29a, b) Levi concluded
that Ildibaha albomaculata is a junior synonym
of Micrathena flaveola (C. L. Koch, 1839) or a
related species of the triangularispinosa species
group. I recently found the syntypes of Ildibaha
albomaculata in the collection of the Naturhis-
torisches Museum, Wien (NMW) and was able
to examine them.

The material consists of three juvenile syn-
types (originally four according to the acquisition
ledger in NMW) from Blumenau, Brazil (26°55'S:
49°07'W). Two of the specimens are slightly
smaller (2.40 & 2.48 mm) and darker than the
third (2.68 mm), and the habitus of the two
smaller specimens is almost identical with the
juvenile specimen illustrated by Levi (1985, fig.
506, 507, Micrathena acuta) with only four ab-
dominal spines. The third specimen is lighter and
provided with 6 dorsal abdominal spines (Figs.
1, 2). I agree with Levi (1985) that Ildibaha al-
bomaculata is conspecific with a species in the
triangularispinosa species group, but do not think
it is possible to state which particular species
until more juvenile material of all known species
is available.

I thank R. Pinto-da-Rocha (Museu de Historia
Natural “Capao da Imbuia”, Curitiba, Parana)
and J. Gruber (Naturhistorisches Museum, Wien)
for making the material of Thaumastobella and
Ildibaha, respectively, available to me. Thanks
are also extended to H. W. Levi, J. Coddington,
N. 1. Platnick, C. Griswold and one anonymous
reviewer for comments and corrections to an ear-
lier draft of this note. The study was supported
by the Danish Natural Science Research Council
(grant no. 1 1-7440). The Smithsonian Institution
kindly provided space and facilities during the
study.

LITERATURE CITED

Bonnet, P. 1959. Bibliographia Araneorum, 2 (part

5). Toulouse.

Brignoli, P. M. 1983. A Catalogue of the Araneae

Described Between 1940 and 1981. Manchester

Univ. Press. Manchester.

155

156

THE JOURNAL OF ARACHNOLOGY

Figures 1, 2.— Habitus of Ildibaha albomaculata Keyserling, juvenile syntype from Brazil: 1, dorsal view; 2,
lateral view. Scale line = 1 mm.

Keyserling, E. 1892. Die spinnen Amerikas. Epeiridae,
part I. Niimberg, 4:1-208.

Levi, H. W. 1985. The spiny orb-weaver genera Mi-
crathena and Chaetacis (Araneae, Araneidae). Bull.
Mus. Comp. ZooL, 150:429-618.

Mello-Leitao, C. de. 1945. Tres novas especies de
Gasteracanthinae e notas sobre a subfamilia. Anais
da Academia Brasileira de Ciencias, 17:261-267.

Pinto-da-Rocha, R. & S. De Fatima Caron. 1989. Ca-
talogo do material-tipo da colegao de Arachnida

Rudolf Bruno Lange do Museu de Historia Natural
’’Capao da Imbuia", Curitiba, Parana, Brasil. Rev.
Brasil. Biol., 49:1021-1029.

Nikolaj Scharff, Zoologisk Museum, Universi-
tetsparken 15, DK-2100 Copenhagen, Den-
mark.

Manuscript received January 1991, revised February,
1991.

INSTRUCTIONS TO AUTHORS

Manuscripts are preferred in English but may be accepted
in Spanish, French or Portuguese subject to availability of
appropriate reviewers. Authors whose primary language is not
English, may consult the Associate Editor for assistance in
obtaining help with English manuscript preparation. All man-
uscripts should be prepared in general accordance with the
current edition of the Council of Biological Editors (CBE) Style
Manual unless instructed otherwise below. Authors are ad-
vised to consult a recent issue of the Journal of Arachnology
for additional points of style. Send three identical copies of
the typed material together with copies of illustrations to:

Gary L. Miller, Associate Editor
The Journal of Arachnology
Department of Biology
The University of Mississippi
University, Mississippi 38677 USA
Telephone: (601) 232-7495
FAX: (601) 232-5144
Bitnet: BYMILLER@UMSVM
Correspondence relating to manuscripts should also be di-
rected to the Associate Editor.

Manuscripts longer than 1 500 words should be prepared as
Feature Articles, shorter papers as Research Notes.

FEATURE ARTICLES

Title page.— Place in the upper left comer the complete name,
address, and telephone number of the author vrith whom proofs
and correspondence should be exchanged. Include FAX num-
bers and Bitnet addresses if available. Four lines below, type
the title, centered, all in capital letters. Another four lines be-
low, beginning with the left margin, type each author’s name
on a new line followed by the address.

Abstract.— The heading should be placed at the beginning
of the first paragraph set off by a period. In articles written in
English, a second abstract in an acceptable language may be
included pertinent to the nationality of the author(s) or geo-
graphic region(s) emphasized in the article. A second abstract
in English must be included in articles not written in the latter
language.

Text.— Double-space text, tables, legends, etc. throughout.
Three levels of heads are used. The first level (METHODS,
RESULTS, etc.) is typed in capitals and on a separate line.
The second level head begins a paragraph with an indent and
is separated from the text by a period and a dash. The third
level may or may not begin a paragraph but is underscored
and separated from the text by a colon. Use only the underline
method to indicate italics; DO NOT USE WORD PROCES-
SOR ITALICS CODES. Use the metric system unless quoting
text or referencing collection data. All decimal fractions are
indicated by the period regardless of language of the text.

Citation of references in the text.— Include within parenthe-
ses the surnames of the author(s) followed by the date of pub-
lication. A comma separates multiple eitations by the same
author(s) and a semicolon separates citations by different au-
thors, e.g., (Smith 1970), (Jones 1988; Smith 1989), (Smith
1986, 1987; Smith & Jones 1989; Jones et al. 1989).

Literature cited section.— Use the following style:
Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid
composition of major ampullate gland silk (dragline) of Ne-
phila clavipes (Araneae, Tetragnathidae). J. Arachnol., 18:
297-306.

Krafft, B. 1982. The significance and complexity of com-
munication in spiders. Pp. 1 5-66, In Spider Communica-
tions: Mechanisms and Ecological Significance. (P. N. Witt
& J. S. Rovner, eds.). PrincetonUniversity Press, Princeton.

Footnotes.— Footnotes are only permitted on the first printed
page to indicate current address or other information con-
cerning the author(s). All footnotes are placed together on a
separate manuscript page.

Running head.— Author sumame(s) and an abbreviated title
must be typed all in caps and must not exceed 60 characters
and spaces. The running head is placed by itself on a separate
manuscript page.

Taxonomic articles.— Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Associate
Editor.

Tables.— Each table, with the legend above, should be placed
on a separate manuscript page. Only the appropriate horizontal
lines should be included. Use no footnotes; instead, include
all information in the legend. Make notations in the text mar-
gins to indicate the preferred location of tables in the printed
text.

Illustrations.— Original art work should be kept by the au-
thor(s) until final acceptance for publication. All art work must
be mounted on stiff board ready for reproduction. In line draw-
ings, pay particular attention to width of lines and size of
lettering when reductions are to be made by the printer. Mul-
tiple photos assembled on a single plate should be mounted
with only a minimum of space separating them. In the case of
multiple illustrations mounted together, each illustration must
be numbered sequentially rather than given an alphabetic se-
quence. Written on the back should be the name(s) of author(s),
an indication of top edge, and whether the art work should be
returned. The overall dimensions should be no more than 1 1
inches (28 cm) x 14 inches (36 cm). Larger drawings present
greater difficulty in shipping and greater risks of damage for
which the JOA assumes no responsibility. The author is en-
couraged to consider PMT (PhotoMechanical Transfer) repro-
ductions for large drawings and to consult an Editor for details
on appropriate reductions. Make notations in the text margins
to indicate the preferred position of illustrations in the printed
text.

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Adhere to the
following styles:

Figures \-^.—A-us x-us, male from Timbuktu: 1, left leg;
2, right chelicera; 3, dorsal aspect of genitalia; 4, ventral aspect
of abdomen.

Figures 27-34.— Right chelicerae of species of A-us from
Timbuktu: 27, 29, 3 1, 33, dorsal views; 28, 30, 32, 34, prolater-
al views of moveable finger; 27, 28, A-us x-us, holotype male;
29, 30, A-us. w-us, holotype male; 31, 32, A-us z-us, holotype
male; 33, 34, A-us y-us, male. Scales = 1.0 mm.

Assemble manuscript for mailing.— Assemble the separate
sections or pages in the following sequence; title page, abstract,
text, figure legends, footnotes, running head, tables with leg-
ends, figures.

Page charges and reprints.- The current charge per journal
page will be assessed as follows: $50— nonmembers, $45 —
members with grant, institutional, or other support for page
costs; $25 — regular members who do not have support for page
costs; $20— student members who do not author with full
members and who do not have support. Limited grants may
be available to help defray publication costs. Consult the As-
sociate Editor for details. Reprints are available only from the
printer at current costs.

RESEARCH NOTES

Instructions above pertaining to feature articles apply also
to research notes except that abstracts and most headings are
not used and the byline follows the Literature Cited section.

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 19 Feature Articles NUMBER 2

Spatial distribution of Lycosa tarentula fasciiventris (Araneae, Lycosidae) in
a population from central Spain, Carmen Ferndndez-Montraveta, Ra-
fael Lohoz-Beltra and Joaquin Ortega 73

Owner-biased agonistic behavior in female Lycosa tarentula fasciiventris
(Araneae, Lycosidae), Carmen Ferndndez-Montraveta and Joaquin Or-
tega 80

A mitochondrial DNA restriction enzyme cleavage map for the scorpion
Hadrurus arizonensis (luridae), Deborah Roan Smith and Wesley M.

Brown , 85

Eremochelis lagunensis, especie nueva (Arachnida, Solpugida, Eremobati-

dae) de Baja California Sur, Mexico, Ignacio M. Vdzquez 88

Nuevos aportes al genero Porrimosa Roewer (Araneae, Lycosidae), Roberto
M. Capocasale 93

Mother-offspring food transfer in Coelotes terrestris (Araneae, Agelenidae),

Jean-Luc Gundermann, Andre Horel and Chantal Roland 97

On the spider genus Fedotovia (Araneae, Gnaphosidae), Vladimir I. Ovtsha-

renko and Norman I. Platnick 102

The life history of Euscorpius flavicaudis {^tCorpiorvQS, Chactidae), T. G.

Benton 105

Homing by crab spiders Misumena vatia (Araneae, Thomisidae) separated
from their nests, Douglass H. Morse Ill

On Eurasian and American Talanites (Araneae, Gnaphosidae), Norman I.

Platnick and Vladimir I. Ovtsharenko 115

Diplocentrus perezi, a new species of scorpion from southeastern Mexico

(Diplocentridae), W. David Sissom 122

When is the sex ratio biased in social spiders?: Chromosome studies of
embryos and male meiosis in Anelosimus species (Araneae, Theridi-
idae), Leticia Aviles and Wayne Maddison 126

Ultrastructure of the primary male genital system, spermatozoa, and sper-
miogenesis of Hypochilus pococki (Araneae, Hypochilidae), Gerd Al-
berti and Frederick A. Coyle 136

Research Notes

A trap to capture burrowing arachnids, Johannes R. Henschel 150

Novel use of silk by the harlequin beetle-riding pseudoscorpion, Cordylo-
chernes scorpioides (Pseudoscorpionida, Chemetidae), David W. Zeh
and Jeanne A. Zeh 153

On the synonymy of Thaumastobella mourei Mello-Leitao and Ildibaha

albomaculata Keyserling (Araneae, Araneidae), Nikolaj Scharff 155

°+ The Journal of
ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 19

1991

NUMBER 3

THE JOURNAL OF ARACHNOLOGY

EDITOR: James W. Berry, Butler University

ASSOCIATE EDITOR: Gary L. Miller, The University of Mississippi

EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E.
Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C.
Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ.; C. D. Don-
dale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galiano,
Mus. Argentine de Ciencias Naturales; M. H. Greenstone, BCIRL, Columbia,
Missouri; N. V. Horner, Midwestern State Univ.; D. T. Jennings, Garland, Maine;
V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ.; E. A. Maury, Mus.
Argentine de Ciencias Naturales; N. 1. Platnick, American Mus. Natural Hist.;
G. A. Polis, Vanderbilt Univ.; S. E. Riechert, Univ. Tennessee; A. L. Rypstra,
Miami Univ., Ohio; M. H. Robinson, U.S. National Zool. Park; W. A. Shear,
Hampden-Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa
Rica.

The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the
study of Arachnida, is published three times each year by The American Arach-
nological Society. Memberships (yearly): Membership is open to all those inter-
ested in Arachnida. Subscriptions to The Journal of Arachnology and American
Arachnology (the newsletter), and annual meeting notices, are included with mem-
bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or
$90 (all other countries). Inquiries should be directed to the Membership Secretary
(see below). Back Issues: Susan E. Riechert, Department of Zoology, Univ. of
Tennessee, Knoxville, TN 37916 USA. Undelivered Issues: Allen Press, Inc.,
1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: Allen R. Brady (1991-1993), Biology Department, Hope College,
Holland, Michigan 49423 USA.

PRESIDENT-ELECT: James E. Carico (1991-1993), Department of Biology,
Lynchburg College, Lynchburg, Virginia 24501 USA.

MEMBERSHIP SECRETARY: Norman I. Platnick (appointed), American
Museum of Natural History, Central Park West at 79th St., New York, New
York 10024 USA.

TREASURER: Gail E. Stratton (1991-1993), Department of Biology, Albion
College, Albion, Michigan 49224 USA.

SECRETARY: Brent Opell (1991-1993), Department of Biology, Virginia Poly-
technic Institute and State University, Blacksburg, Virginia 24061 USA.

ARCHIVIST: Vincent D. Roth, Box 136, Portal, Arizona 85632 USA.

DIRECTORS: Matthew H. Greenstone (1990-1992), George W. Uetz (1991-
1993), Charles E. Griswold (1991-1993).

HONORARY MEMBERS: C. D. Dondale, W. J. Gertsch, H. Homann, H. W.
Levi, A. F. Millidge, M. Vachon, T. Yaginuma.

Cover illustration: Female Phidippus mystaceus, (Araneae, Salticidae), Eastern USA, by G. B. Edwards.

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.

1991. The Journal of Arachnology 19:157-160

SEGREGATION STUDIES OF ISOZYME VARIATION IN
METAPHIDIPPUS GALATHEA (ARANEAE, SALTICIDAE)'

William W. M. Steiner and Matthew H. Greenstone: Biological Control of Insects
Research Laboratory, USDA, Agricultural Research Service, P.O. Box 7629,

Research Park, Columbia, Missouri 65205 USA

ABSTRACT. Three field-collected isofemale lines of Metaphidippus galathea were established as laboratory
colonies. The female parents and their progeny were electrophoretically examined for 1 3 proteins coding for 2 1
isozymes. Eight proteins were segregating for allozymes and four were analyzed for Mendelian inheritance.
Although sample sizes were small, the GOT-1 locus showed a tendency toward deviation from the expected
inheritance pattern. Nonconformance to genetic expectations may be due to multiple-mating, to selection effects
from laboratory rearing conditions, or to genetic drift.

Several studies now claim to show varying lev-
els of genetic variation in spiders based on allo-
zyme surveys (Steiner et al. in prep.; Roeloffs &
Riechert 1988; Terranova & Roach 1987a; Smith
1 986; Lubin & Crozier 1985; Cesaroni et al. 1981;
Manchenko 1981). Genetic variation is impor-
tant for the study of phenomena such as popu-
lation differentiation and interpopulation migra-
tion (Roeloffs & Riechert 1988; Smith 1986;
Lubin & Crozier 1985) and in defining taxonom-
ic differences (Terranova & Roach 1987b; Pen-
nington 1979). For the studies to be valid rep-
resentatives of genetic variability, it is important
that genetic inheritance of the allozymes be ver-
ified, as it is possible for enzyme variation to be
environmentally induced (Gerasimova & Smir-
nova 1986; Amason & Chambers 1987) or to be
found only during ontogeny (Korochkin & Mat-
veeva 1974). In this paper, we report the first
evidence based on parental-offspring correla-
tions to support Mendelian inheritance of allo-
zymes in spiders.

METHODS

Three female Metaphidippus galathea were
collected in mid-June of 1984 at the University
of Missouri’s Tucker Natural Prairie in central
Missouri. The Prairie is a tallgrass remnant lo-
cated 25 km east of Columbia, Missouri (Drew
1 947). The females were returned to the Biolog-
ical Control of Insects Research Laboratory,

'Any mention of a proprietary product in this article
does not indicate endorsement by the USDA-ARS.

United States Department of Agriculture, Agri-
cultural Research Service, where they were housed
in self-watering cages (Jackson 1974), fed early-
mid 5th instar Trichoplusia ni (Hubner) larvae
for maintenance, and kept at approximate rela-
tive humidities and temperatures of 70% and 25
°C, respectively. Approximately 1 7 days later all
three were observed with egg cases. Sixteen days
after that the egg cases hatched and the spider-
lings were maintained on Drosophila melano-
gaster Meigen. At 60 days of age, the spiderlings
and their maternal parents were frozen for elec-
trophoresis. Aging spiders for 60 days reduces
the possibility of ontogenic effects on electro-
phoretic pattern.

Starch gel electrophoresis was performed and
the resulting gels stained for enzyme activity as
described in Steiner and Joslyn (1979). Individ-
uals were run side-by-side on 1 cm thick gels
which could be horizontally-sliced into 1 mm
slices for protein histochemical staining. The
protein systems analyzed are listed in Table 1.
Two electrode chamber-gel buffer systems were
used. The first was a pH 8.2 LiOH/Boric Acid
electrode buffer with a pH 8.5 Trizma Base/citric
acid gel used to analyze the proteins ACPH, AL-
DOX, EST, PGM, and PGI. The second con-
tained citric acid and Trizma Base in both the
electrode chamber (pH 8.2) and the gel (pH 8.4)
and was used to analyze ADK, GOT, G-3-PDH,
HK, IDH, a-GPDH, MDH and 6-PGDH. Ma-
terial from the female and her offspring were run
on the same gels to aid in recognizing isozyme
banding homologies. The resulting segregation

157

158

THE JOURNAL OF ARACHNOLOGY

Table 1.— Enzymatic loci and their electrophoretic characteristics for the spider, Metaphidippus galathea.
Abbreviations: DH = dehydrogenase; E.C.N. = enzyme classification number assigned by the International Union
of Pure and Applied Chemistry; * = polymorphic protein system; S = protein quaternary structure, M = monomer
and D = dimer; N = number of loci observed for the individual protein (the total is 2 1 ); relative distance is the
anodal migration distance measured from the origin to the band on a 12% Sigma starch gel run at 70 mA/gel
for 12 h on the designated system in METHODS. For GOT-2, the protein migrated toward the cathode 3 mm
from the origin.

Protein

Enzyme

Classification

Number

Abbrev.

S

N

Relative
distance in mm

Acid phosphatase

3. 1.3.2

ACPH

M

2

18 and 15

Adenylate kinase

2.7.4.3

ADK

M

1

24

Aldehyde oxidase

1.2.3. 1

ALDOX

M

1

22

Esterase

3.1. 1.1

EST*

M

4

68, 60, 50 and 33

Glutamate oxaloacetate transaminase

2.6.1. I

GOT*

D

2

16 and -3

Glyceraldehyde-3-phosphate DH

1.2.1.12

G-3-PDH

—

1

16

alpha-Glycerophosphate DH

1.1. 1.8

a-GPDH*

D

2

20 and 10

Hexokinase

2.7. 1.1

HK

—

1

2

Isocitrate DH

1.1.1.42

IDH*

D

2

1 2 and 8

Malate DH

1.1.1.37

MDH

—

2

20 and 8

6-phosphogluconate DH

1.1.1.44

6-PGDH*

D

1

17

Phosphoglucomutase

2.7.5. 1

PGM

—

1

31

Phosphoglucose isomerase

5.3. 1.9

PGI

—

1

5

data were analyzed using the test with Yates
Correction Factor for small sample sizes.

RESULTS

A total of 2 1 loci encoding 1 3 proteins were
identified. Eight loci were polymorphic, but four
of these were esterases which we have not in-
cluded in this analysis. This was because some
individuals expressed overlap of alleles between
the esterase loci making genotype assignments
difficult. These esterase genes expressed the high-
est variation found.

The other four loci included GOT- 1 (migrating
to 16 mm), IDH-1 (migrating to 12 mm), a-
GPDH-1 (migrating to 20 mm) and 6-PGDH
(migrating to 1 7 mm). Heterozygotes of all these
loci were triple-banded while homozygotes were
single-handed suggesting a dimeric enzyme
structure consisting of two polypeptide chains
(Table 1).

In this study the maternal genotypes could be
determined directly from the gels, leading us to
infer the paternal genotype. This enabled us to
generate an expected genotype ratio in the F,,
assuming a single-pair mating took place. The
assumptions of male genotype and F, genotype
ratio were met at all loci although GOT-1 ap-
proached a significant departure. At that locus,
a deficiency of heterozygotes and an excess of the

common allele homozygote occurred which was
strengthened when similar mating types were
pooled. Assuming an alternative ratio of 1:2:1
only led to a higher value (x^ = 15.73, P <
0.001) due to a lack of GOT-U'' and decreased
expected numbers of GOT- PL

DISCUSSION

The segregation patterns we observed fit Men-
delian expectations. This is expected, since Men-
delian inheritance of allozyme genes is widely
reported in the literature for Drosophila, hu-
mans, plants and other organisms.

Only the segregation at GOT-1 approaches a
significant departure (x^ = 2.77,/’ = 0.105) which
can be explained in several ways. First, the GOT-
1 results may be anomalous since we cannot
completely exclude sampling error due to small
samples (genetic drift). Only further study can
verify or falsify this argument, although the
strength of the x^ statistic which includes the
correction for small sample sizes, and the rela-
tively higher number of surviving Fj would not
seem to support it. Or, it may be that selection
is favoring GOT-1” under the laboratory con-
ditions. However, it remains difficult to bring the
observed ratio into a 1 : 1 balance by just invoking
selection against or for a single genotype, and the
question arises as to what the selective factor

STEINER & GREENSTONE-SEGREGATION STUDIES OF ISOZYMES

159

Table 2.— Segregation statistics for four allozyme loci in the spider, Metaphidippus galatkea. Parental genotypic
crosses are indicated by A, B, and C. Sex ratio refers to female to male with female genotype known and male
genotype inferred. Expected F, genotypes are calculated from the expected genotypic ratio. In the genotype
codings, 5 designates the most commonly occurring band, 1-4 indicates faster migrating bands, and 6-9 indicates
slower migrating bands. We assume bands are indicative of the allelic state. For the chi-square tests, matings
with similar parental genotypes were pooled when individual chi-square tests were insignificant. Lack of sig-
nificant differences between observed and expected numbers of progeny is designated NS.

Locus

Parental

genotypes

F| sex
ratio

F, genotypes, obs/exp

Expected F,
genotypic

45

55

56

ratio

GOT-1

A

55 X 45

5:4

2/4

7/5

1:1

NS

B

55 X 45

1:4

1/2

3/2

1:1

NS

C

45 X 55

3:4

2/2

2/2

1:1

NS

6-PGDH

A

56 X 55

5:4

4/4

5/4

1:1

NS

B

55 X 55

1:4

4/4

Fixed

C

55 X 55

3:4

7/7

Fixed

a-GPDH-1

A

55 X 56

5:4

5/5

4/5

1:1

NS

B

55 X 56

3:4

2/3

5/4

1:1

NS

C

55 X 55

1:4

4/4

Fixed

IDH-1

A

55 X 55

5:4

9/9

Fixed

B

56 X 55

1:4

1/2

3/2

1:1

NS

C

55 X 55

3:4

7/7

Fixed

might be, assuming laboratory conditions are
more optimal to survival than natural condi-
tions. Obviously the other allozyme loci do not
reflect a selection pattern. It is also possible that
sperm competition is involved. In fact, Jackson
(1980) has found evidence for sperm competi-
tion in Phidippus, suggesting evidence for both
multiple mating and pre-zygotic selection. In our
case, assuming that GOT - 1 somehow plays a role
in any pre-zygotic selection further complicates
the issue.

An alternative explanation is that GOT-1
female A (Table 2) may have mated with two
males rather than one. Fertilization by an ad-
ditional GOT-1” genotypic male would make
half the female’s progeny homozygous; removing
that half from the nine offspring would then re-
sult in a 1:1 ratio of offspring resulting from the
mating with the GOT-U^ heterozygous male.

Although the evidence for either multiple-
mating and/or sperm competition is admittedly
weak, the suggestion that electrophoresis can be
used as a tool to study reproductive strategies in
spiders is appropriate. This approach should be
considered especially in view of Jackson’s (1980)
research. Questions concerning the extent of
multiple mating and the survival or fitness qual-
ities of the paternal parent could then be as-
sessed. In this way, population biology param-
eters previously undefined could lead to a better

understanding of spider ecology, behavior and
evolution.

ACKNOWLEDGEMENT

We would like to thank C. Morgan for his aid
in raising and caring for the spiders, and the De-
partment of Biological Sciences, University of
Missouri, Columbia for permission to work at
Tucker Prairie. We would also like to thank F.
Breden; Y. Lubin; I. McDonald; S. Riechert and
G. Uetz for comments and criticisms.

LITERATURE CITED

Amason, E. & G. K. Chambers. 1987. Macromolec-
ular interaction and the electrophoretic mobility of
Esterase- 5 from Drosophila pseudoboscura. Bio-
chem. Genet, 25:287-307.

Cesaroni, D., G. Allegrucci, M. Caccone, M. Sbordoni,
E. De Matthaeis & 1. Sbordoni. 1981. Genetic
variability and divergence between populations of
species of Nesticus cave spiders. Genetica, 56:81-
92.

Drew, W. B. 1947. Floristic composition of grazed
and ungrazed prairie vegetation in North-Central
Missouri. Ecology, 28:26^1.

Gerasimova, T. 1. & S. G. Smirnova. 1979. Maternal
effects for genes encoding 6-phosphogluconate de-
hydrogenase and glucose-6-phosphate dehydroge-
nase in Drosophila melanogaster. Develop. Genet.,
1:97-107.

Jackson, R. R. 1974. Rearing methods for spiders. J.
Arachnol., 2:53-56.

160

THE JOURNAL OF ARACHNOLOGY

Jackson, R. R. 1980. The mating strategy of Phidip-
pus johnsoni (Araneae, Salticidae): II. Sperm com-
petition and the function of copulation. J. Arach-
nol., 8:217-240.

Korochkin, L. I. & N. M. Matveeva. 1974. Genetics
of esterases in Drosophila of the Virilis group. II.
Sequential expression of paternal and maternal es-
terases in ontogenesis. Biochem. Genet., 12:1-7.

Lubin, Y. D. & R. H. Crozier. 1985. Electrophoretic
evidence for population differentiation in a social
Achaearanea wau (Theridiidae). Insectes Soc.
Paris, 32:297-304.

Manchenko, G. P. 1981. Allozymic variation in ylr-
aneus ( Arachnida, Aranei). Isozyme Bull.,

14:78.

Pennington, B. J. 1979. Enzyme genetics in taxon-
omy: diagnostic enzyme loci in the spider genus
Meta. Bull. Br. Arachnol. Soc., 4:377-392.

Roeloffs, R. & S. E. Riechert. 1988. Dispersal and

population-genetic structure of the cooperative spi-
der, Agelena consociata, in West African rain forest.
Evolution, 42: 173-183.

Smith, D. R. R. 1986. Population genetics of Ane-
losimus eximius (Araneae, Theridiidae). J. Arach-
nol., 14:201-217.

Steiner, W. W. M. & D. J. Joslyn. 1979. Electropho-
retic techniques for the genetic study of mosquitoes.
Mosq. News, 39:35-54.

Terranova, A. C. & S. H. Roach. 1987a. Genetic
differentiation in the genus Phidippus (Araneae, Sal-
ticidae). J. Arachnol., 14:385-391.

Terranova, A. C. & S. H. Roach. 1987b. An electro-
phoretic key for distinguishing immature and adult
species of the genus Phidippus (Araneae: Salticidae)
from South Carolina. Ann. Entomol. Soc. Amer.,
80:346-352.

Manuscript received January 1991, revised March 1991.

1991. The Journal of Arachnology 19:161-168

TWO NEW SPECIES OF NESTICUS SPIDERS
FROM THE SOUTHERN APPALACHIANS
(ARANEAE, NESTICIDAE)

Frederick A. Coyle and Augustus C. McGarity: Department of Biology, Western
Carolina University, Cullowhee, North Carolina 28723 USA

ABSTRACT. Diagnoses, descriptions, illustrations, and natural history data are presented for two new species
of Nesticus spiders: N. nasicus from epigean habitats in southwestern North Carolina, and N. gertschi from a
cave in eastern Tennessee. Nesticus nasicus appears to be the sister species of Nesticus brimleyi Gertsch, a cave-
dwelling species.

In his pioneering revision of North and Central
American nesticids, Gertsch (1984) predicted
that, as greater attention was focused on these
secretive cave- and litter-dwelling spiders, many
new species would be added to the already large,
apparently monophyletic, clade of 24 southern
Appalachian species of Nesticus. We describe here
a new epigean species, Nesticus nasicus, and a
new cave-dwelling species, Nesticus gertschi, both
of which, like the majority of the known southern
Appalachian species, have restricted ranges, are
allopatric, and live in eastern Tennessee and/or
western North Carolina (Map 1).

RELATIONSHIPS

Since Gertsch (1984) did not present a cladistic
analysis of relationships of Nesticus species and
since we have examined specimens of only two
of the 41 described species (and have therefore
relied heavily on Gertsch’s drawings and descrip-
tions), our hypotheses of relationship are es-
pecially tentative.

The numerous similarities between Nesticus
nasicus and Nesticus brimleyi Gertsch strongly
suggest that they are sister species. These simi-
larities include the following putative synapo-
morphies: 1) the broad, thin and translucent, ser-
rate distal process of the paracymbium [Figs. 1-
6]; 2) the massive size of the paracymbium [Figs.
1, 6]; 3) a median palpal apophysis with two
converging processes [Figs. 1, 6]; 4) a sharply
tapering tegular process hidden under the me-
dian apophysis [Figs. 1, 6]; and 5) a thick-walled
bulb-shaped spermatheca near each lateral bor-
der of the epigynum [Figs. 11, 14]. We postulate
that the common ancestor of N. nasicus and N.
brimleyi was, like several extant Nesticus species

(Gertsch 1984), a troglophilic species consisting
of both epigean and cave-dwelling populations,
and that its range, restricted to a cave-poor region
of the southern Blue Ridge Province, included
the areas now occupied by N. nasicus and N.
brimleyi (Map 1). We further suggest that the
epigean populations disappeared from the slight-
ly dryer (Clay et al. 1975) eastern portion of the
range leaving the N. brimleyi lineage isolated in
the humid refugium provided by the isolated
cluster of fissure caves it now occupies.

The relationships of N. gertschi are much less
clear. It shares with N. nasicus and N. brimleyi
the broad, translucent, spatulate, distal paracym-
bial process which appears to be unique to these
three species among all American Nesticus for
which males are known (Figs. 1-6, 1 5-1 7). How-
ever, there is no similarly distinctive female gen-
ital character state shared by N. gertschi and these
two species.

METHODS

The quantitative character values in Table 1
are an integral part of each description. These
characters are abbreviated and defined as fol-
lows: BL— body length; CL— carapace length;
CW— carapace width; CH— clypeus height (length
along median longitudinal line from edge of car-
apace to line connecting lowest edges of the two
ale’s, with clypeus horizontal); AMD, ALD,
PMD, PLD— maximum diameters of eye pupils
with each eye on horizontal plane; AMS— dis-
tance between AME and ALE (each eye inter-
distance is measured after positioning it on hor-
izontal plane); AS— distance between AME and
ALE; PMS—distance between PME’s; PS— dis-
tance between PME and PLE; IFL, IPL, ITL,

161

162

THE JOURNAL OF ARACHNOLOGY

Map 1 . — Distribution of all known southern Appalachian Nesticus species, based upon locality and habitat
records in Gertsch (1984). Each of the 16 species collected at only one or a cluster of neighboring localities is
represented by a single symbol. Known range boundaries of the nine more widely distributed species are
approximated by dotted or solid lines.

IML, ITarL— lengths of leg 1 articles (distance in
retrolateral view from proximal condyle to most
distal point on dorsal surface); EW — distance be-
tween lateral pockets of the epigynum (Fig. 1 2);
MSE— length of the caudal extension of the me-
dian septum (Fig. 12) (epigynum measurements
made with abdomen tilted so that ventral surface
of epigynum is on horizontal plane). Eye diam-
eters and appendage measurements were record-
ed from the left eye or appendage unless it was
damaged, missing, or not fully regenerated (in
which case the right structure was measured).

Measurements are in mm and were recorded
with a Wild M-5 stereomicroscope with 20 x oc-
ular lenses and an eyepiece micrometer scale. BE,
CL, CW, and leg measurements were performed
at 50 X and are accurate to 0.018 mm; all other
measurements were performed at lOOx and are
accurate to 0.009 mm. Internal (dorsal) views of

cleared (85% lactic acid) epigyna were drawn with
a compound light microscope fitted with a draw-
ing tube. We follow Gertsch’s ( 1 984) terminology
for genital anatomy. All specimens are deposited
in the American Museum of Natural History
(AMNH).

Nesticus nasicus, new species
Figs. 1-4, 7-14. Map 1.

Types.— Male holotype and one male and three
female paratypes collected under loose rocks 30
m outside west entrance of Cowee Mountain train
tunnel (1900 ft elev.), 1 mi W Dillsboro, Jackson
County, North Carolina (28 October [holotype,
2 females] and 1 1 November [other adult male]
1 990 and 20 April 1 99 1 [penultimate male, which
escaped in captivity, and female]; A. McGarity),
in AMNH.

Etymology.— The specific name, a Latin ad-

COYLE & McGARITY-NEW SPECIES OF NESTICUS SPIDERS

163

Figures 1-6.— Palpi of Nesticus holotypes, with paracymbial processes and some other structures labeled: 1-
4, N. nasicus\ 1, ventral; 2, dorsal; 3, retrolateral view of paracymbium; 4, medial (concave) surface of para-
cymbium; 5, 6, N. brimleyi\ 5, medial (concave) surface of paracymbium; 6, ventral.

jective, refers to the nose-like appearance of the
middle septum of the epigynum.

Diagnosis. — Males of N. nasicus are readily
distinguished from those of all other American

Nesticus species by the broad translucent distal
paracymbial process accompanied by a sharp-
tipped paradistal process (Fig. 3), and from all
but N. brimleyi by the distinctively shaped me-

164

THE JOURNAL OF ARACHNOLOGY

dian apophysis and tegular process (Fig. 1). The
following features of the palp, particularly the
position and shape of the paracymbial processes,
distinguish N. nasicus males from those of its
sister species, N. brimleyi: 1) ventromedial par-
acymbial process absent [Fig. 4] vs. present [Figs.
5, 6]; 2) dorsomedial paracymbial process pres-
ent [Figs. 2-4] vs. absent [Fig. 5]; 3) paradisial
paracymbial process narrow and pointed [Figs.
2-4] vs. broad and blunt [Fig. 5]; 4) dorsal par-
acymbial process tapers to a single point [Figs.
1-3] vs. truncate with three or more irregular
points [Figs. 5, 6]; 5) distal paracymbial process
rounded [Fig. 3] vs. angular; 6) base of tegular
process evenly curved and gradually tapering [Fig.
1] vs. a lobe-like shoulder [Fig. 6]; 7) tegular
process with one vs. two dorsal keels; 8) lateral
projection of median apophysis relatively long
and strongly curved [Fig. 1] vs. short and weakly
curved [Fig. 6]; 9) middle loop of seminal tube
broad at base and blunt [Fig. 1] vs. relatively
narrow at base and long [Fig. 6]. Additionally,
the legs of N. nasicus males are proportionately
much shorter [Table 1, ITL(100)/CL] than those
of N. brimleyi. Females of N. nasicus are most
readily distinguished from those of other Amer-
ican Nesticus species by their unique, medially
directed, epigynal pockets [Figs. 9, 11, 12, 14].
Although it was not possible for us to examine
N. brimleyi females, Gertsch’s (1984) figs. 138-
140 and description reveal the following dis-
tinctive N. nasicus traits: 1) lateral epigynal
pockets [Figs. 9, 11, 12, 14] not present on N.
brimleyi', 2) much darker abdominal pigmenta-
tion [Figs. 7, 8] than N. brimleyi', and 3) legs
proportionately much shorter [ITL(100)/CL =
142-161] than those of N. brimleyi [ITL(IOO)/
CL =214].

Males.— Table 1. Palpus [Figs. 1-4] with large
paracymbium with broad, translucent, serrate
distal process; sharp-tipped paradistal apophy-
sis; thin dark distomedial process; prominent
ventral process with distal edge turned outward;
thin, sharp-tipped, leaf-like dorsal process, ser-
rate and very thin on its ectal edge; and thin dark
dorsomedial process on base of dorsal process.
Tegular process tapers to sharp tip behind me-
dian apophysis, with dorsal keel visible in pro-
lateral view. Median apophysis with large, roughly
serrate, lateral process and prominent distal pro-
cess with twisted tip. AME’s with well-defined
lenses. Color very similar to that of females ex-
cept carapace less heavily pigmented.

Females,— Table 1. Epigynum [Figs. 9-14] with

prominent median septum with rather broad
rounded caudal projection; medially facing lat-
eral pocket on each side; two avocado-shaped
spermathecae (one ectal to each pocket) visible
in cleared dorsal view; keel-like rim borders each
atriobursal orifice, the two rims together forming
a V-shaped pattern in posterior view. AME’s [Fig.
8] with well-defined lenses. Color [Figs. 7, 8] of
appendages chiefly pale tan but whiter or darker
in places; carapace very pale tan with grey pig-
ment as in Fig. 8; abdomen dorsally with seg-
mental series of large paired lateral areas of grey
on white background.

Variation.— Although there is no noteworthy
variation within either population sample, the
two populations differ in the following female
characteristics, most of which are epigynum fea-
tures. 1) The caudal lobe of the median septum
of the Wolf Creek sample [n = 4] is absolutely
[MSE = 0.111-0.129, mean = 0.123 ±0.009]
and proportionately longer [MSE(100)/EW =
26.1-32.6, mean = 29.3 ±2.8] [Fig. 12] than that
of the type sample [n = 3] [MSE = 0.056-0.093,
mean = 0.074 ±0.018; MSE(100)/EW = 10.7-
20.0, mean = 15.6 ±4.7] [Fig. 9]. In posterior
view the dorsal contour of this lobe, enlarged as
it is in the Wolf Creek specimens, presents a
distinct transverse upcurved line [Fig. 13] not
present in the paratypes [Fig. 10]. 2) Each of the
two diagonal keels forming the V-shaped rim
bordering the atriobursal orifices and visible in
posterior view is entire in the Wolf Creek spec-
imens [Fig. 1 3] and not interrupted as in two of
the three paratypes [Fig. 10]. 3) The external lat-
eral epigynal pockets tend to be smaller and di-
rected more anteriorly in the Wolf Creek speci-
mens [Figs. 12, 14] than in the paratypes [Figs.
9, 1 1 ]. 4) The first leg articles are proportionately
longer in the Wolf Creek specimens [ITL(IOO)/
CL = 152-161, mean = 158 ±4.2)] than in the
paratypes [ITL( 1 00)/CL = 142-147, mean = 146
±2.7)].

These differences suggest that there may be
little or no gene flow across the 1 1 miles sepa-
rating these local populations. Determining
whether this is the case and whether these pop-
ulations are reproductively isolated requires the
collection and study of males from Wolf Creek,
a search for geographically intermediate popu-
lations, and, most importantly, cross-mating tri-
als. The small allopatric geographic ranges char-
acteristic of most of the southern Appalachian
species (Map 1) suggest that low vagility is a
common Nesticus trait.

COYLE & McGARITY-NEW SPECIES OF NESTICUS SPIDERS

165

Figures 7-14.— A. nasicus females; 7-11, paratypes; 7, lateral view of body; 8, dorsal view of body; 9-11,
epigynum; 9, ventral; 10, posterior; 11, dorsal (cleared); 12-14, specimen from Wolf Creek; 12, ventral; 13,
posterior; 14, dorsal (cleared). Scale lines: 0.5 mm for Figs. 7, 8; 0.2 mm for Figs. 9-14.

Natural history.-— The Cowee Mountain spec-
imens were living just outside the train tunnel
on the undersides of rocks which had fallen from
a high rock cut and accumulated in a wet leaf
litter-filled depression between the base of the
cut and the railway roadbed. The holotype male

and two females were collected under the same
large flat rock; the females were in small webs
consisting of a sparse asymmetrical mesh of silk
threads extending from the underside of this rock
to smaller rocks beneath. The adult female col-
lected on 20 April had just molted. The absence

166

THE JOURNAL OF ARACHNOLOGY

Table 1.— Quantitative character values for Nesticus species. Characters are defined in methods section of
text. All measurements in mm. Range, mean, and standard deviation given for large sample. * The second value
for each character is from the holotype. ** Except for CH, the first value for each character is from the holotype;
values for the second male are from Gertsch (1984), who erroneously indicated that the specimen he measured
was the holotype.

nasicus

brim ley i

gertschi

Males*

(« = 2)

Females
in = 7)

Males**

{n = 2)

Male

Female

BL

2.42, 2.66

2.41-3.07 (2.72 ± 0.26)

3.14, 4.00

3.48

3.42

CL

1.28, 1.43

1.18-1.33 (1.28 ±0.05)

1.74, 2.00

1.61

1.31

CW

1.11, 1.24

1.04-1.15 (1.11 ±0.04)

1.52, 1.75

1.37

1.48

CH

0.231, 0.259

0.176-0.231 (0.201 ± 0.017)

0.350, 0.361

0.296

0.250

AMD

0.028, 0.037

0.028-0.037 (0.032 ± 0.005)

0.037

0.037

0.037

ALD

0.102, 0.102

0.074-0.093 (0.087 ± 0.007)

0.111

0.083

0.093

PMD

0.093, 0.093

0.093-0.102 (0.095 ± 0.005)

0.093

0.083

0.083

PLD

0.093, 0.093

0.083-0.102(0.094 ± 0.006)

0.093

0.093

0.102

AMS

0.037, 0.065

0.037-0.056 (0.049 ± 0.007)

0.083

0.037

0.046

AS

0.046, 0.056

0.046-0.056 (0.049 ± 0.005)

0.074

0.074

0.065

PMS

0.093, 0.102

0.083-0.111 (0.096 ± 0.009)

0.139

0.129

0.129

PS

0.056, 0.065

0.046-0.074(0.057 ± 0.010)

0.083

0.083

0.074

IFL

2.07, 2.44

1.89-2.20 (2.07 ±0.12)

3.92, 4.80

3.63

2.89

IPL

0.56, 0.65

0.52-0.59 (0.57 ± 0.03)

0.80, 1.00

0.72

0.65

ITL

2.04, 2.44

1.74-2.09 (1.95 ±0.12)

4.14, 5.15

3.74

2.79

IML

1.87, 2.18

1.55-1.85 (1.72 ±0.10)

3.85, 4.50

3.37

2.48

ITarL

0.89, 1.04

0.83-0.93 (0.88 ± 0.03)

1.48, 1.50

1.35

1.15

EW

0.40-0.52 (0.45 ± 0.04)

MSE

0.056-0.129(0.102 ± 0.029)

MSE(100)/EW

10.7-32.6 (23.4 ±8.1)

ITL(100)/CL

159, 171

142-161 (153 ±8)

238, 258

232

213

AMD(100)/CH

12.0, 14.3

13.0-19.0 (15.8 ±2.4)

10.3

12.5

14.8

AMD(100)/CW

2.5, 3.0

2.4-3. 3 (2.8 ± 0.4)

2.4

2.7

2.5

of Nesticus within the dark but dry train tunnel
and on the surface of the moist north-facing rock
cut above the inhabited rock pile indicates that
N. nasicus requires both very high humidity and
very low light intensity.

The other N. nasicus population sample was
collected in the leaf litter of a mesic deciduous
forest on steep rocky north- and south-facing
slopes on each side of Wolf Creek. Tullgren fun-
nel extraction of this leaf litter collected on 14
and 16 November yielded two adult females,
four antepenultimate or penultimate females, two
penultimate males, two antepenultimate males,
and three younger juveniles (CW = 2.68-4.16
mm). These data and the presence of adult fe-
males at the type locality during both fall and
spring indicate that N. nasicus adults may occur
during most or all of the year, as is true for at
least some of the other species of Nesticus (Gertsch
1 984), and that mating and egg-laying may there-
fore occur during several months of each year.
Extended reproductive activity, an extreme case

being the year-round egg-laying exhibited by cave
populations of the nesticid Eidmanella pallida
(Ives 1935) and many other cave animals (Ho-
warth 1 983), may constitute a primitive Nesticus
cave-related trait which is expressed even in ep-
igean species like N. nasicus.

Distribution.— Known from two localities 11
miles apart in the mountains of southwestern
North Carolina (Map 1).

Other material examined. — NORTH CAROLINA:
Jackson Co., Wolf Creek, 5 mi S Cullowhee, 2400 ft
elev., deciduous forest leaf litter, 13 September 1990
(A. McGarity), 1 female; 24 October 1990 (A. Mc-
Garity), 1 female, 1 juv.; 14 November 1990(F. Coyle),
1 female, 10 juvs.; 16 November 1990 (R. Dellinger),
1 female, 1 juv.; 24 February 1991 (F. Coyle), 2 juvs.

Nesticus gertschi, new species
Figs. 15-20. Map 1.

Types. — Male holotype and one female para-
type collected 100 m inside Cedar Creek Cave
(1400 ft elev.), Cedar Creek, Greene County,

COYLE & McGARITY-NEW SPECIES OF NESTICUS SPIDERS

167

Figures 15-20.— N. gertschi: 15-17, holotype palpus; 15, ventral; 16, dorsal; 17, retrolateral view of paracym-
bium; 18-20, paratype epigynum; 18, ventral; 19, posterior; 20, dorsal (cleared).

Tennessee (16 March 1991; A. McGarity), in
AMNH.

Etymology. —The specific name is a patronym
in honor of Dr. Willis J. Gertsch, the first revisor
of American nesticids.

Diagnosis.— Among known species of Appa-
lachian Nesticus, N. gertschi is one of only three
species (including N. brimleyi and N. nasicus)
with a broad, thin, translucent, serrate, distal
paracymbial process (Fig. 1 7) and is the only one
with a single, very long, flat, broad tegular pro-
cess (Fig. 1 5). The collective shapes and positions
of other paracymbial processes also distinguish

this species (Figs. 15-17). The combination of
the curved, ectally-facing, external epigynal
grooves on each side of the median septum (Fig.
1 8), the large, sclerotized, internal anterior lobes
(Figs. 18, 20), and the elongate, nearly banana-
shaped spermathecae (Fig. 20) distinguish N.
gertschi females from those of all other Appa-
lachian Nesticus species.

Male.— Table 1 . Palpus (Figs. 15-17) with large
paracymbium with broad, translucent, slightly
serrate distal process, two subdistal processes on
dorsal edge, a thin, sharp-tipped, leaf-like dorsal
process, and a prominent ventral process with

168

THE JOURNAL OF ARACHNOLOGY

distal edge turned outward; tegular process long,
broad, and thin with distomedial angle expanded
toward distal lobe of median apophysis; median
apophysis with broad, spatulate, roughly serrate
lateral process and broad, angular, spatulate dis-
tal process. AME’s with well-defined lenses. Col-
or as in female except appendages darker tan
than those of female.

Female.— Table 1 . Epigynum (Eigs. 1 8-20) with
rather prominent broad median septum with lit-
tle, if any, caudal extension; depression on each
side of median septum with prominent, curved
ectally-facing groove with sclerotized rim and
more anteriorly and laterally a less conspicuous,
curved medially-facing rim; two elongate sper-
mathecae almost banana-shaped; two large, in-
ternal well-sclerotized lobes extending forward
at anterior of epigynum; shallow keel-like rim
borders each atriobursal orifice, the two rims
converging dorsally in posterior view. AME’s with
well-defined lenses. Color of appendages very pale
tan; carapace white to very pale tan with scat-
tered areas of faint grey pigment on pars ce-
phalica and around lateral border of pars thor-
acica, except dark amber to black around each
eye; abdomen dorsally with segmental series of
very small, faint, paired, lateral areas of grey on
lighter pale beige-grey background.

Natural History. — The two specimens were
collected in separate small concavites (19, 15 cm
high; 15, 16 cm wide; 10, 6 cm deep) in the cave
wall approximately 100 m from the entrance in
the moist dark zone of the cave. Each spider was
suspended back-downward from (or close to) the

ceiling of its concavity and in the upper denser
part of its web, a loose irregular mesh of threads
confined to the upper portion of the concavity.
When collected ( 1 6 March), both specimens were
in the penultimate instar. They were kept in the
dark at 15 °C and 85% relative humidity and
molted to adults five (male) and ten (female) days
later.

Distribution.- Known only from the type lo-
cality in the mountains of eastern Tennessee (Map

1).

Other material examined.— None.

ACKNOWLEDGMENTS

We thank N. I. Platnick (AMNH) for loaning
us the holotype of N. brimleyi and R. G. Bennett
for helpful comments on the manuscript.

LITERATURE CITED

Clay, J. W., D. M. On- & A. W. Stuart. 1975. North
Carolina Atlas. Univ. of North Carolina Press,
Chapel Hill.

Gertsch, W. J. 1984. The spider family Nesticidae
(Araneae) in North America, Central America, and
the West Indies. Texas Mem. Museum Bull., 31:1-
91.

Howarth, F. G. 1983. Ecology of cave arthropods.

Ann. Rev. EntomoL, 28:365-389.

Ives, J. D. 1935. A study of the cave spider, Nesticus
pallidus Emerton, to determine whether it breeds
seasonally or otherwise. J. Elisha Mitchell Sci. Soc.,
51:297-299.

Manuscript received May 1991, revised October 1991.

1991. The Journal of Arachnology 19:169-173

EVIDENCE FOR IDIOTHETICALLY CONTROLLED TURNS AND
EXTRAOCULAR PHOTORECEPTION IN LYCOSID SPIDERS

Jerome S. Rovner: Dept, of Zoological Sciences, Ohio University, Athens, Ohio 45701
USA

ABSTRACT. During some of the intervals between bouts of pheromone-stimulated courtship display, iso-
lated male Rabidosa rabida (Araneae, Lycosidae) perform a single pivot. In an investigation of the control of
this turning behavior, males were tested under four conditions. Two of these were visual, a uniform environment
or one with images, and two were non-visual, all eyes occluded or dim red lighting. The turning angle and the
tendency to change the direction of turning were measured for the first three conditions, and no significant
differences were found. This suggests that the turns are controlled idiothetically. Another parameter, the tendency
to perform the turns, was reduced under dim red light but not in blinded spiders under white light, the latter
suggesting the occurrence of extraocular photoreception.

When animals organize their behavior with
respect to spatial features, they use information
obtained from external directing stimuli (alloth-
etic orientation) or from internal sources (idioth-
etic orientation). The latter may depend either
on proprioceptive information or on central ner-
vous system (nonsensory) programs that contain
the necessary information for spatial execution
of movements (Schone 1984). The present study
deals with the question of whether rotational lo-
comotion that occurs during courtship behavior
in isolated male lycosid spiders is influenced by
visual stimuli or is under idiothetic control.

The display of Rabidosa rabida (Walckenaer)
occurs in discrete bouts. In each bout the palps
are waved in alternation, and then the right or
left leg I is extended coincident with palpal-pro-
duced sounds. Both of the latter elements end
abruptly in synchrony. A pause follows, during
which time a receptive female, if present, signals
her response. Thus, the male’s distinct bouts of
display alternate with inter-bout intervals, pro-
viding a basis for reciprocal signaling between
the sexes.

When this species’ display was quantitatively
analyzed (Rovner 1968), the data were obtained
from males in the presence of females. However,
when males in isolation are stimulated to display
by contact with the female’s sex pheromone, an
additional behavior occurs. During some of the
inter-bout intervals, such males perform a single
pivoting turn. Rovner (1991) hypothesized that

this rotational locomotion represents a compo-
nent of a local search pattern. Apparently, it is
added to the behavior of a male in the courtship
mode if he has failed to detect a responding fe-
male during the early phase of courtship.

Since turning behavior in animals can be in-
fluenced by goal-related images or by the level
of illumination (Schone 1984), I examined
whether such visual input plays a role in the
inter-bout turning behavior of isolated male R.
rabida. I tested spiders in well-lit arenas with or
without fixed images and also tested them under
non-visual conditions: in darkness (dim red light)
or after occlusion of the eyes.

METHODS

Fifty male and 1 0 female Rabidosa rabida (for-
merly Lycosa rabida) were collected as penulti-
mate instars in early July 1989 in a field in Ath-
ens County, Ohio. Spiders were not used until 1
week or more after the final molt. Methods of
maintenance and laboratory conditions during
testing were described previously (Rovner 1 989).

The testing arena was a glass bowl with a slop-
ing wall, on the inside of which a coat of flat,
pale green, non-toxic paint had been applied to
provide a uniform, non-reflecting surface. The
bottom was covered with a cardboard disk, over
which was placed a pale green sheet of paper 11.3
cm in diameter (about 100 cm^). The latter was
replaced with a fresh sheet for each test, so that
silk or chemicals deposited on the substrate by

169

170

THE JOURNAL OF ARACHNOLOGY

one spider could not remain to influence sub-
sequent individuals. A vertical cardboard barrier
visually isolated the arena from my location.

For most test conditions, non-directional il-
lumination was provided by a 32 W, soft- white,
circular fluorescent bulb centered over the arena
(height = 50 cm above the arena floor). The level
of illumination at the arena floor was about 700
lux (Gossen Luna-Pro meter), comparable to the
light level in a deciduous forest understory on a
sunny day.

For a test condition without visible light being
available to the spiders, a dim red light was pro-
vided by Kodak Safelight (No. 1 filter; 15 W
incandescent bulb) placed 20 cm above the arena
floor. The filter passed wavelengths >610 nm.
Even under bright white light, the sensitivity of
the largest eyes (posterior median) of wolf spiders
falls off sharply above 550 nm, especially if the
spiders are light-adapted when tested (DeVoe et
al. 1969), as was the case in the present study.
The single window in the laboratory was covered
with opaque material, and tests under this con-
dition were conducted after dark in the late eve-
ning. With this arrangement, a very low level of
red light (about 2-4 lux) reached the arena floor,
just sufficient for my direct observations but not
adequate for monitoring or recording by video.
Consequently, only the number of display bouts
and turns (not turning angles) was included in
the data for this test condition.

To determine whether the dim red light con-
dition insured total darkness for the spiders, I
ran a preliminary check on 20 males in the fol-
lowing manner. Pairs of males were vibrationally
isolated in separate cages that allowed visual con-
tact (Rovner 1 989) and were observed under the
dim red light until one of the males had initiated
walking and performed three passes across his
cage. None of the males showed a response to a
walking male under this condition. They sub-
sequently oriented and showed courtship display
to such a stimulus after I switched on a white
light (an exposed 10 W bulb 25 cm above the
cages that provided about 300 lux illumination).
On this basis, the dim red tight was judged sat-
isfactory for insuring darkness.

Observations and data recording in the three
test conditions run under white fluorescent light
were done with the aid of a video camera (JVC
model GX-8NU), a remote-controlled videocas-
sette recorder (Sony model SL-HFR70), and a
video monitor. A character generator (JVC mod-
el CG-C7U) provided an on-screen stopwatch

(reading to 0. 1 s) and titles identifying each test.
The video recorder was located on a separate
table 1.2 m from the testing arena to prevent
possible vibratory stimulation. The camera faced
obliquely upward toward a front-silvered mirror
clamped above the testing arena at a 45° angle
to the floor. This gave a dorsal view of the spider,
essential for later measurement of turning angles
(by use of a protractor placed over the still frame
on-screen). Due to limitations imposed by the
resolution of the video image, I could accurately
measure angles only to the nearest 5°.

Each male to be tested was transferred in a
plastic vial from its home cage to the arena. I
allowed the spider to slide gently onto a centrally
located stimulus source, a square section (6.25
cm^) of a larger piece of paper that had served
as the floor covering in a female’s cage for a
number of days. I then sat out of sight of the
arena and viewed the monitor. As soon as the
male began courtship display, I remotely acti-
vated the video recorder and kept it running for
the next 10 min. Twenty males were tested only
once, and 30 males were used twice, i.e., in two
of the four test conditions, mixed through all of
the treatments. When a male was used twice, the
tests were separated by several days, so as to
achieve a reasonable level of independence.

The four conditions examined were: (1) fixed
images on arena wall, (2) uniform arena wall, (3)
dim red light, and (4) occluded eyes. While tests
under the red light condition were run only in
the late evening, those under the other three con-
ditions were run throughout the day and evening.
Twenty males were tested under each condition.

The stimuli used for testing the influence of
images consisted of two identical silhouettes of
the front view of female wolf spiders, comparable
to shapes presented to salticid spiders by various
workers, including Crane (1949), but remaining
fixed rather than being moved like those usually
used for salticids. I attached the images to the
arena wall opposite each other and at floor level
so as to simulate the appearance of female con-
specifics resting at the arena’s edge. I used more
than one image to insure that the male had an
easy opportunity to pick up the potential stim-
ulus within his visual field soon after his entry
into the arena, no matter which direction he faced
initially. The maximum distance from the spider
to an image was well within the range of detect-
ability, based on data from a previous study
(Rovner 1989).

To occlude the eyes, I covered them with two.

ROVNER-TURNS IN LYCOSID SPIDERS

171

Table 1.— Inter-bout turning behavior in courting male Rabidosa rabida. There were no differences among
turning angles (Kruskal-Wallis, = 1.149) or series lengths of turns in the same direction (H^ = 1.316). Data
were based on 10 males/condition.

Condition

Number of

turns

Turning angle
(degrees)

Series of turns in
same direction

Turning angle x
series length

Fixed images

70

62 ± 45.9

2.6 ± 2.21

155°

Uniform wall

85

67 ± 55.3

2.8 ± 1.81

201°

Occluded eyes

107

61 ± 40.2

2.9 ± 1.90

183°

Grand mean ± SD

63 ± 47.0

2.8 ± 1.96

176°

separately applied, coats of water-based enamel
(Top Color Hobbylack, Pelikan AG). That this
insured complete occlusion had been established
previously (ibid.).

Data on turning angles were based on the mag-
nitude of each turn, irrespective of direction.
Where appropriate, I present these and other data
as X ± SD. Analyses of data involved Kruskal-
Wallis tests (corrected for ties) and t-tests of arc-
sine-transformed percentages (Sokal & Rohlf
1969).

RESULTS

Occurrence of turns.— During courtship, an
“inter-bout intervaP’ that was 6.7 ± 1.51 s in
duration followed each bout of display, and there
was a mean of 3.0 ± 0.83 display bouts/min.
After an early phase of courtship in which no
locomotion occurred during the inter-bout in-
tervals, the male pivoted in place during 2 1 .4%
of the subsequent inter-bout intervals. (A small
amount of forward locomotion sometimes oc-
curred during an inter-bout interval; however,
the nature of such linear locomotion was not
addressed in the present study.) A pause always
preceded inter-bout locomotion, during which

Table 2.— Tendency of male Rabidosa rabida to per-
form inter-bout turns. Data were based on 20 males/
condition. (For / = 1.96, P = 0.05.)

Condi-

tion

Total

turns

Total

bouts

Per-

cent

t

P

Fixed

images

Uniform

121

599

20.2

1.93

>0.05

wall

157

632

24.8

3.15

<0.01

Red

light

Occluded

93

537

17.3

2.31

<0.05

eyes

145

638

22.7

time a receptive female, had one been present,
would have performed her receptive display.
Turning never occurred during the male’s bouts
of courtship display.

Turning angle.— Turns resulted from forward
steps by the legs of one side of the spider and
reverse steps by the contralateral legs. (Some spi-
ders occasionally made very small turns of 1 5°
or less, resulting from a single remotion of one
or two anterior-most ipsilateral legs. Such cases
lacking bilateral appendage involvement were not
included in the analyses.) Turning angle, which
had a mean of 63 ± 47.0°, was independent of
the presence or absence of fixed, spider-like im-
ages and of whether the eyes were occluded or
not (Table 1).

Turning direction.— Viewed dorsally, turns
were either clockwise or counterclockwise, with
both directions equally represented in the data,
i.e., no handedness. Although in about one-third
of the cases a directional change occurred after
only one turn in the other direction, in the ma-
jority of cases the spiders performed a series of
turns in one direction, then a series in the other
direction. Since turns only occurred during inter-
bout intervals and only during about a fifth of
these intervals, it must be kept in mind that a
“series” of turns involved behavioral events sep-
arated by time and by other activity. The number
of turns in a series of unidirectional turns ranged
from two to eight {X = 2.8 ± 1.96) and was
independent of the presence or absence of fixed,
spider-like images and of whether the eyes were
occluded or not (Table 1).

Turning tendency.— The percentage of court-
ship bouts followed by turns was regarded as an
indicator of turning tendency (Table 2). Com-
parisons of arcsine-transformed percentages be-
tween treatment groups revealed significant dif-
ferences in two cases: (1) Spiders tested under
red light turned less often than those under white

172

THE JOURNAL OF ARACHNOLOGY

light in uniform arenas. (2) Blinded spiders under
white light turned more often than untreated spi-
ders under red light. A difference just shy of sig-
nificance was also noted: Untreated spiders ex-
posed to fixed images turned less often than those
surrounded by a uniform wall. Some of the for-
mer did maintain an initial orientation toward
an image for a period of time after their intro-
duction to the arena.

DISCUSSION

Orientation behavior in animals can involve
a mechanism that relies on external input or can
be controlled entirely by an internal mechanism
(Schone 1984). Data obtained in the present study
suggested that pivoting turns occurring during
courtship in isolated male lycosid spiders can be
performed independently of external stimuli. The
methods eliminated vibrational cues since no fe-
male was present; and directional lighting was
avoided as well. Testing in the presence or ab-
sence of fixed images and testing under greatly
different illumination levels were the approaches
used to determine the possible influence of cer-
tain visual stimuli on the orientation behavior
being studied.

When provided with fixed, spider-like images,
male R. rabida did not show significant differ-
ences in either turning angles or turning series
lengths from those of spiders in a uniform en-
vironment. Such data support the view of Hom-
ann (1931), who stated that the eyes of wolf spi-
ders are adapted for the detection of movement.
However, turning tendency was almost signifi-
cantly less for spiders exposed to the fixed images
compared to those in a uniform environment.
This resulted from some spiders having tem-
porarily held an orientation toward an image de-
tected at the time of introduction to the arena.
Crane (1949) also observed this occasionally in
salticids, which usually do not respond to a fixed
image. She suggested that when a spider is
dropped into the arena “the visual effect to the
spider may be similar to that obtained when the
stimulus is moving”.

The size of inter-bout turns was the same in
R. rabida with occluded eyes as in untreated in-
dividuals, which suggests that these turns are
controlled endogenously. Such self-steered turns
are well known in various arthropods and were
thoroughly analyzed in a series of studies on
courtship turning in the cockroach Blattella ger-
manica (Bell et al. 1978; Bell & Schal 1980;

Franklin et al. 1981). However, one cannot be
completely certain that turning behavior in any
arthropod is under idiothetic control until tests
employing Helmholtz coils are used to eliminate
the remote possibility of geomagnetic orientation
(Havukkala & Kennedy 1984).

The tendency of male R. rabida to turn during
the inter-bout interval was affected by the level
of illumination. Spiders under dim red light had
a lower turning tendency than that of spiders
under white light in uniform arenas and had the
lowest numbers of courtship bouts and inter-bout
turns of all groups. Interestingly, Frings (1941)
had found that R. rabida became less active
(“akinetic”) under reduced illumination; and he
judged this to be the reason (rather than negative
phototaxis) that the spiders ended up in the shad-
ed chamber of a choice box.

If male R. rabida have a reduced level of inter-
bout turning under dim red light, why then did
the spiders which also seemingly experienced
complete darkness due to occlusion of the eyes
show the same turning tendency as untreated spi-
ders under white light? Interestingly, Kapoor
(1971) found that blinded pumpkinseed fish re-
sponding to different levels of illumination
showed changes in turning angle like those of
untreated fish, and he noted that the fish’s pineal
photoreceptor can mediate such responses. Re-
cent electrophysiological studies by Yamashita
(1986) revealed that efferent neurons in the brain
of two species of araneid spiders were sensitive
to light. The behavioral data described here for
R. rabida raise the possibility that extraocular
photoreception also occurs in lycosid spiders, an
hypothesis that requires electrophysiological
confirmation in a future study.

ACKNOWLEDGMENTS

The video equipment was obtained with sup-
port from the Baker Fund of Ohio University. I
thank G. E. Svendsen for statistical advice, and
W. J. Bell, G. E. Stratton, and R. B. Suter for
their comments on earlier drafts of this paper.

LITERATURE CITED

Bell, W. J., S. B. Vuturo & M. Bennett. 1978. En-
dokinetic turning and programmed courtship acts
of the male German cockroach. J. Insect Physiol.,
24:369-374.

Bell, W. J. & C. Schal. 1980. Patterns of turning in
courtship orientation of the male German cock-
roach. Anim. Behav., 28:86-94.

Crane, J. 1949. Comparative biology of salticid spi-

ROVNER-TURNS IN LYCOSID SPIDERS

173

ders at Rancho Grande, Venezuela, Part IV. An
analysis of display. Zoologica, 34:159-215.

DeVoe, R. D., R. J. W. Small & J. E. Zvargulis. 1969.
Spectral sensitivities of wolf spider eyes. J. Gen.
Physiol., 54:1-32.

Franklin, R.,W.J. Bell &R.Jander. 1981. Rotational
locomotion by the cockroach Blattella germanica.
J. Insect Physiol., 27:249-255.

Frings, H. 1941. Stereokinetic and photokinetic re-
sponses of Lycosa rabida, Calosoma lugubre, and
Harpalus caliginosus. J. Comp. Psychol., 32:367-
377.

Havukkala, I. J. & J. S. Kennedy. 1984. A pro-
gramme of self-steered turns as a humidity response
in Tenebrio, and the problem of categorizing spatial
manoeuvres. Physiol. Entomol., 9:157-164.

Homann, H. 1931. Beitrage zur physiologic der Spin-
nenaugen. III. Das Sehvermogen der Lycosiden. Z.
vergl. Physiol., 14:40-67.

Kapoor, N. N. 1971. Locomotory patterns of fish
(Lepomis gibbosus) under different levels of illu-
mination. Anim. Behav., 19:744-749.

Rovner, J. S. 1968. An analysis of display in the
lycosid spider rabida Walckenaer. Anim. Be-
hav., 16:358-369.

Rovner, J. S. 1989. Wolf spiders lack mirror-image
responsiveness seen in jumping spiders. Anim. Be-
hav., 38:526-533.

Rovner, J.S. 1991. Turning behaviour during phero-
mone-stimulated courtship in wolf spiders. Anim.
Behav., 42:1015-1016.

Schone, H. 1984. Spatial Orientation: The Spatial
Control of Behavior in Animals and Man. Princeton
University Press, Princeton.

Sokal, R. R. & F. J. Rohlf 1969. Biometry. W. H.
Freeman, San Francisco.

Yamashita, S. 1986. Cerebral photosensitive neurons
in the orb-weaving spiders Argiope bruennichii and
A. amoena. Proc. 9th Intern. Congr. ArachnoL, Pan-
ama 1983: 332.

Manuscript received March 1991.

1991. The Journal of Arachnology 19:174-209

HAWAIIAN SPIDERS OF THE GENUS TETRAGNATHA:
L SPINY LEG GLADE

Rosemary G. Gillespie: Department of Zoology and Hawaiian Evolutionary Biology
Program, University of Hawaii, Honolulu, Hawaii 96822 USA

ABSTRACT. The Hawaiian archipelago is well known for some of the most spectacular species radiations
from single ancestors, although the occurrence of this phenomenon in spiders remains largely undocumented.
The present study introduces the radiation of the highly diverse spider genus Tetmgnatha in Hawaii. Preliminary
studies indicate that the Hawaiian Tetragnatha can be divided into distinct clades, and this paper describes
representatives of the Spiny Leg clade. These species are characterized by the many, robust spines on their legs,
and the abandonment of web-building activity. There are 12 species in this clade, ten of which are new and
described in this paper: T. tantalus n. sp., T. polychromata n. sp., T. hrevignatha n. sp., T. macracantha n. sp.,
T. waikamoi n. sp. and T. kauaiensis Simon (in the Green Spiny Legs group), T. kamakou n. sp. and T. perreimi

n. sp. (m the Green and Red Spiny Legs group), and
and T. mohihi n. sp. (in no distinct group).

The Hawaiian archipelago possesses some of
the most extraordinary faunal assemblages in the
world. Explosive diversification of species from
a single ancestor has occurred repeatedly, often
accompanied by radical shifts in morphology,
ecology and behavior. Some of the best examples
of this phenomenon can be found within the hon-
eycreepers (subfamily Drepanidinae in the Frin-
gillidae) (Berger 1981; Freed et al. 1 987), the land
snails (Cooke et al. 1960) and in the spectacular
radiation within the family Drosophilidae, with
over 500 endemic species (Kaneshiro and Boake
1987). This paper is the first in a series that will
document such a radiation in a genus of Ha-
waiian spiders.

Systematic studies on native spiders in Hawaii
are few, and, with the noted exception of thom-
isids(Suman 1964, 1970), and ecological studies
on the theridiid Theridion grallator Simon (Gil-
lespie 1989, 1990; Gillespie and Tabashnik 1989,
1990; Gon 1985), have been largely ignored for
almost a century. Even the studies of the late
1 9th century were very incomplete (Karsch 1 880;
Simon 1900; Okuma 1988c). Based on the col-
lection of R. C. L. Perkins, Simon (1900) rec-
ognized the speciose nature of one or a few genera
in four spider families: Theridiidae, Salticidae,
Thomisidae and Tetragnathidae. The usefulness
of this reference, however, is limited primarily
because Perkins’ spider collection, by his own
admission, was incomplete and unrepresentative
(Perkins 1913): spiders were collected only in

’. pilosa n. sp., T. quasimodo n. sp., T. restricta Simon

passing during his daylight searching for birds
and insects, or while he collected insects attracted
to a light at night. The majority of endemic Ha-
waiian spiders are strictly nocturnal and ex-
tremely difficult to find during the day (pers. obs.),
and they cannot be attracted by lights; it is there-
fore not surprising that they are under-repre-
sented in his collections. Also, recent studies
(Gillespie, in prep.) reveal that there was a good
deal of confusion in Simon’s assignation of spe-
cies. For example, he discusses the unique mor-
phological features of the “Spiny Leg’’ Tetmg-
natha Latreille, yet the holotype of one of the
three he describes bears no spines, while the
paratypes are mixed with those that do.

This study introduces the radiation of the long-
jawed orb-weaving spider genus Tetragnatha in
Hawaii, one of the most morphologically and
ecologically diverse group of spiders in the is-
lands. Consider what is known of the genus out-
side Hawaii: Of all spiders, Tetragnatha are
among the most abundant worldwide (Levi 1981).
They are also a very homogeneous group of spi-
ders, in both morphology (elongate bodies and
legs, and large chelicerae and endites [Kaston
1 948]) and ecology (Dabrowska Prot and Luczak
1968 a and b; Dabrowska Prot et al. 1968; Gil-
lespie 1986). They are characterized by the con-
struction of an orb web with an open hub (Wiehle
1963), the structure being extremely light and
fragile with low adhesiveness (Yoshida 1987). It
is generally built over water or in other wet places

174

GILLESPIE- HAWAIIAN TETRAGNATHA

175

(Gillespie 1987a). Construction of a web neces-
sitates ambush predation in the genus as a whole,
although individuals of certain species are ca-
pable of capturing prey without the use of a web
(Luczak and Dabrowska Prot 1966; Levi 1981;
Gillespie 1 987b). Now consider the genus in Ha-
waii: Here, in stark contrast to what is known of
the genus worldwide, the lineage is highly spe-
ciose (Simon 1 900), diverse in both morphology
and ecology. It now seems likely that there are
at least as many species endemic to Hawaii as
there are in the entire continent of Asia.

Preliminary phylogenetic studies using mor-
phological and molecular data (Croom, et al,
1991; Gillespie, Croom and Palumbi, in prep.)
indicate that the Hawaiian Tetragnatha can be
divided into distinct clades, each with its own
unique set of characteristics. At present we define
three (or four) major clades. This paper describes
the species in the Spiny Leg clade, i.e., the major
Spiny Leg species group. Cladistic analyses using
a total of 46 morphological and ecological char-
acters indicate that the Spiny Leg clade is mono-
phyletic (Gillespie, Croom and Palumbi, in prep.).
The same result is found using an independent
data set from mitochondrial DNA (Croom, et al,
1991). This paper itself, however, does not ad-
dress phylogenetic issues.

There are two distinct groups within the Spiny
Leg clade: the Green Spiny Legs {T. tantalus, T.
polychromata, T. brevignatha, T. macracantha,
T. waikamoi and T. kauaiensis) and the Green
and Red Spiny Legs {T. kamakou and T. per-
reirai). The remaining species {T. pilosa, T. quas-
imodo, T. restricta and T. mohihi) belong to nei-
ther group.

My criteria to recognize species are: 1) distinct
differences (internally homogeneous) in one or
more gross morphological characters; and 2) con-
sistent differences in genitalic structure. This
method is obviously a conservative means of
determining true species identity. Some may
judge the differences between certain populations
(e.g., T. kamakou and T. quasimodo on different
islands) sufficient to assign these to separate spe-
cies. However, mating experiments between these
populations reveal that coupling is possible, with
palpal insertion into the seminal receptacles (Gil-
lespie, in prep.), although I do not know whether
sperm transfer occurred. Future research may
determine these to be separate species, but in the
absence of evidence for reproductive isolation I
consider them different populations of a single
species. Further species may also be added to the

clade as more specimens are accumulated from
different areas, revealing hitherto unknown taxa.

METHODS

Characters examined.— Gross morphological
features were investigated using a dissecting mi-
croscope and illustrated using a camera lucida
attachment. For each individual examined, mea-
surements were taken of the separation between
each of the eyes, tooth pattern on the chelicerae
(both pro- and retromarginal), fang structure,
form and spination of the first and third leg (I
and III representing the greatest divergence in
leg function), and form and pattern of the dorsum
and venter of the abdomen, the carapace and
sternum. In order to estimate variability within
a taxon, and determine which features best char-
acterized a species, I attempted to measure at
least 6 individuals of each sex of each species,
with cursory observations on other individuals
once diagnostic characters had been identified.
These measurements were possible for all species
except T. tantalus females and T. perreirai, both
of which are localized and not common. At pres-
ent no female has been found for T. mohihi.

The genitalia of both sexes were examined us-
ing a compound microscope and illustrated using
a camera lucida. The female seminal receptacles
were dissected out, the muscle tissue digested
using Evans-Browning solution, and the struc-
ture cleared and mounted temporarily on a slide
in Hoyers medium. The male palps were ex-
amined by removing the left palp and placing it
temporarily on a slide in glycerol beneath a
moveable coverslip, allowing rotation of the
structure in order to determine the shape of the
conductor under low power. Palps and seminal
receptacles were subsequently stored in micro-
vials with the specimen.

Scanning electron microscopy was conducted
on the palps of paratype males. Palps were re-
moved from the body and placed in plastic cap-
sules with the central portion removed and nylon
mesh placed inside the capsule (to allow ex-
change of alcohol and COj, while retaining the
specimen). Filled capsules were put through an
alcohol series (70%, 85%, 95% and pure ethanol),
then dried with an Autosamdri-8 10 Critical Point
Dryer. Palps were removed from the capsules,
mounted on stubs using silver paste, then sput-
ter-coated with gold. Specimens were viewed us-
ing a Hitachi S-800 scanning electron micro-
scope.

176

THE JOURNAL OF ARACHNOLOGY

Diagnostic characters. — There are no univer-
sal “key” diagnostic characters for species in the
Spiny Leg clade. For example, the extraordinary,
complex spination of the femora of the 3rd tibia
is a unique and reliable character for identifying
T. pilosa. Among all other species, the spination
is simple, and there is almost no variation in this
character. Similarly, the unique structure of the
female seminal receptacles is one of the most
useful characters for identifying T. polychro-
mata, while in many of the other species, there
is too much inter-individual variation to use these
structures reliably. On the whole, at least for pre-
served specimens, males have many more useful
characters than females. Although the number
of teeth on the cheliceral margins is not reliable,
the pattern and shape of certain teeth (in partic-
ular the first two distal teeth on the promargin)
can be very useful. Similarly, the shape of the
tip of the conductor is usually reliable. I have
also found that, although scanning electron mi-
croscopy gives much more detail of the conduc-
tor tip, examination with a compound micro-
scope is sometimes more useful for revealing
subtle diagnostic features.

For females, the cheliceral armature is of lim-
ited usefulness. Spination of the tibia of the first
leg is a very useful “cue” for both sexes, but
should always be used in conjunction with an-
other character. Spination pattern on the femur
of the first leg is not reliable, while that on the
patella and metatarsus is almost invariable. Eye
patterns are very similar among species in this
clade, and, where there is variability, it is not
very reliable. The size of the eyes, in relation to
the amount of ocular area covered, can be useful.
In certain species, abdominal pattern (even in
largely faded alcoholic specimens) can be diag-
nostic, as can coloration of the venter and ster-
num. Leg banding and coloration of the carapace
are highly unreliable, as many species in the Green
Spiny Leg group change the color of these, ac-
cording (most likely) to habitat.

Terminology. — I have used the terminology of
Okuma (1987, 1988c) for the teeth on the chel-
iceral margins of the males (Fig. 1): ‘Gu’ (guide
tooth of upper row) is the small tubercle (may
be absent or almost tooth-like) on the distal pro-
margin of the chelicerae. Moving from the distal
end of the chelicerae, ‘sf is the first major tooth
on the promargin; ‘T’ is the second tooth, and is
often much larger; ‘rsu’ refer to the remaining
proximal teeth on the promargin, ‘a’ is the dorsal

cheliceral spur (apophysis) for locking the fem-
ale’s fang during mating. ‘AXF (auxiliary guide
tooth of lower row) is the small tubercle (may be
absent or almost tooth-like) on the distal retro-
margin of the chelicerae. Moving from the distal
end of the retromargin of the chelicerae ‘Gl’ (guide
tooth of lower row) is the first major tooth, ‘L2’
the second ‘L3’ the third etc. For females, the
cheliceral teeth are numbered from the distal end
‘Ur - ‘Un’ on the promargin and ‘LF - ‘L«’ on
the retromargin.

CHARACTERISTICS OF THE SPINY
LEG CLADE

The major characteristics of the clade are re-
lated to leg spination and predatory activity, these
being the synapomorphies that unite the species
in a single clade: 1) At least 4 (usually 5, some-
times 6) spines on both prolateral and retrolateral
sides of the 1 st tibia, and always 2 dorsal spines
on tibia I (most other Hawaiian species have 3
or fewer spines on both prolateral and retrola-
teral sides of the 1 st tibia). 2) Spines robust, usu-
ally between 30 and 100% length of carapace (the
spines on most other Hawaiian species are con-
siderably less than 30% length of carapace). 3)
Individuals do not build webs, either as adults
or immatures (all other Hawaiian species known
to date build webs). Some are very active, cur-
sorial predators, while others behave as more
typical sit-and-wait foragers, spending long pe-
riods hanging in mid-air, legs outstretched.

Natural history.- Spiders in this clade, as with
almost all the endemic Hawaiian Tetragnatha,
are exclusively nocturnal. They commence ac-
tivity only after complete darkness (1830-2000
hours), and terminate it before dawn. The peak
of activity is in the early part of the night, slowing
down at around 2330. During the daytime, in-
dividuals lie flat against the substrate that match-
es their own color: Leaves in the case of the
Green Spiny Leg group, rotten logs in the case
of the Green and Red Spiny Leg group, and bark
of any form in the case of T. quasimodo and T.
pilosa. Because of the difficulty of beating much
of the substrate with which these species are as-
sociated, I have found that directly capturing in-
dividuals at night is by far the most satisfactory
collecting technique.

The prey of this group are largely non-flying
insects, such as hemipterans and lepidopteran
larvae, with each species specializing on specific
prey (Gillespie, in prep.). The method of capture

GILLESPIE- HAWAIIAN TETRAGNATHA

177

Figure 1. — Diagram of cheliceral margins (A, promargin; B, retromargin) of male Tetragnatha indicating
terminology for teeth; from Okuma (1988c).

is similar to that of other tetragnathids; Spiders
bite the prey and hold it; they never wrap the
prey prior to immobilization.

Mating behavior has been observed in several
members of this clade. The strategy is that char-
acteristic of other tetragnathids (Levi 1981). There
is no evidence of courtship prior to mating. On
encountering each other, male and female appear
to be involved in a combative interaction, both
with their chelicerae and fangs outstretched. If
the sexual encounter is successful, the male locks
the fangs of the female against the spur (apoph-
ysis) on the dorsal surface of his chelicerae. He
then closes his fangs over those of the female, so
as to lock the female securely in position. The
cheliceral teeth themselves are not involved in
this locking mechanism.

Egg sacs are constructed in a manner that is
basically similar to that of other tetragnathids:
The ball of eggs, tightly wrapped in silk, is cov-
ered over with an additional “tent” of silk, se-
curely fastened to the substrate on all sides. The
form of the tent, however, is characteristic of a
species, often being dotted and blotched with
green and/or black, laid over the white threads.
Some species can even lay colored eggs (e.g., T.
brevignatha lays green eggs).

Distribution.— The Hawaiian islands are ar-
ranged within a chronological time frame, with
the northern island of Kauai the oldest at ap-
proximately 5 millions years, the big island of
Hawaii in the south the youngest at approxi-
mately 0.4 million years (Heliker 1989). The
Spiny Leg Hawaiian Tetragnatha show an inter-
esting pattern of distribution among the islands,
with the oldest island harboring three species
endemic to that island, while the youngest has
no species endemic to that island (Fig. 2). The
greatest diversity of species within this clade are
found on east Maui.

KEY TO SPECIES IN THE SPINY LEG
CLADE OF HAWAIIAN TETRAGNATHA

1. Males 2

Females 13

2. First tooth (‘si’) in form of strong, down-
curved wave, almost contiguous with erect,
pointed 2nd tooth (‘T’) (Fig. 123). Abdomen
widest in middle, medial distinct black in-
verted triangle just below mid- ventral line

T. quasimodo

First tooth weaker, not down-curved. Ab-
domen with no medial inverted triangle 3

178

THE JOURNAL OF ARACHNOLOGY

Figure 2. — Map of Ihe Hawaiian Islands, showing distribution of species in the Spiny Leg clade of Hawaiian
Tetragnatha (omitting T. quasimodo, which occurs on all islands shown except Kauai). Broken lines indicate
latitude and longitude. The perimeters of the major volcanic masses are outlined with marks converging towards
the summits of the volcanoes.

3. Femur of 3rd leg with at least 5 (up to 1 1)
strong, long ventral spines, more than 2x
width of femur (Fig. 1 14). Chelicerae short
(approx. 60% length of carapace); dorsal spur
short (approx. 9% length of carapace) (Figs.

1 09 and 111) T. pilosa

Femur of 3rd leg with no more than 3 rather
short (rarely more than width of femur) ven-
tral spines 4

4. Second tooth ‘T’ pointing rather sharply and

directly (not curved) upwards, away from
‘rsuF and towards ‘sF (Fig. 137) . . . T. restricta
‘T’ not pointing directly upwards from mar-
gin of chelicerae 5

5. Chelicerae long, > 80% length of carapace

(Fig. 55) 6

Chelicerae < 70% length of carapace (Fig. 29) 1 1

6. Apical projection of palpal conductor cap
straight, pointed and rather long (Figs. 22 and

1 54) T. polychromata

Apical projection of conductor cap curled ... 7

7. Conductor cap much higher than wide, apical
projection curled mostly laterally, tip pointed

(Figs. 48 and 157) T. macracantha

Conductor cap wider than high, apical pro-
jection curled mostly forward 8

8. Apical projection from conductor cap ap-
proximately as long as cap itself, pointing out
laterally in broad curl (Fig. 6 1 ). Cap uniform-
ly domed (Fig. 158). Dorsal spur on chelic-

erae without any bifurcation (Fig. 57)

T. waikamoi

Apical projection from conductor cap absent

or much shorter than cap itself 9

9. Backward projection of conductor cap well
below floor of cap itself, giving it appearance
of legionnaire hat (Figs. 87, 159 and 160)

T. kamakou

Backward projection of conductor cap at ap-
proximately same level as floor of cap itself 10

10. Conductor cap clearly divided into two sec-
tions by high ridge leading up from dorsal
side of stem (Figs. 9 and 153). Apical tip

pointed T. tantalus

Conductor cap with indistinct, low ridge di-
viding two sections (Figs. 74 and 161). Apical

tip blunt T. kauaiensis

11. Dorsal cheliceral spur long (18% carapace)

(Fig. 147). Promargin of chelicerae: Distance
from distal margin to ‘sF>> distance from
‘sF to ‘T’ (Fig. 1 45). Tibia I with 4 retrolateral
and 4 (or 3) prolateral spines (Fig. 149) . . .

T. mohihi

Dorsal cheliceral spur short (8-10% cara-
pace). Promargin of chelicerae: Distance from
distal margin to ‘sF not much more (<1.5 x)
than distance from ‘sF to ‘T’. Tibia I with 4
(or 6) retrolateral and 4 (or 6) prolateral spines

1 2. Tibia I: 4 retrolateral, 4 prolateral spines (Fig.

GILLESPIE- HAWAIIAN TETRAGNATHA

99). Legs distinctly banded, carapace and ab-
domen dark T. perreimi

Tibia I: 6 retrolateral, 6 prolateral spines (Fig.

33). Legs without banding, carapace and ab-
domen virtually unpigmented (bright green
in life) T. brevignatha

13. Femur III with numerous (8-10) long ventral

spines (Fig. 1 20) T. pilosa

Femur with no more than 2 ventral spines 14

14. Abdomen distinctly pyriform. Wide part
along medial line raised up into a flat medial
ridge (no lateral or dorso-lateral humps)

T. restricta

Abdomen not distinctly pyriform (diamond-
shaped or oval) 15

15. Abdomen diamond-shaped with sub-medial
distinct, small black inverted triangle, usually
drawn up into short, finger-like tubercle (Fig.

135). Sternum dark/dusky. Venter with
V-shaped bar down center. Color pattern
consists of various combinations of black,
brown and grey. Legs dark and distinctly
banded. Spider quite large (5. 3-8.8 mm) . ,

T. quasimodo

Abdomen without sub-medial, black tuber-
culate triangle. Sternum pale/ translucent.
Venter without V-shaped bar down center.
Color pattern usually green to green/red .... 16

1 6. Abdomen diamond shaped, exaggerated dor-
so-laterally into 2 lateral, rounded humps.
Color pattern various combinations of red
(on lateral humps) and dark green. Sternum
pale, venter uniformly colored. Legs dark,

distinctly banded 17

Abdomen elongate oval. Color bright green
in life, fading to pale yellow in alcohol (some
species capable of becoming darker according

to habitat). Legs usually pale 18

17. Chelicerae short, 52-56% length of carapace

(Fig. 102). Spider quite small (carapace 2.2-
2.4 mm). Promargin of chelicerae: Distance
between 1st and 2nd tooth 8-10% cheliceral
length. Leg spines relatively short (28-36%
length of carapace) (Figs. 105 and 106) ....

T. perreirai

Chelicerae 67-69% length of carapace (Fig.

88), carapace 2. 4-2. 8 mm. Promargin of che-
licerae: Distance between 1st and 2nd tooth
20-30% cheliceral length. Leg spines rela-
tively long (45-60% length of carapace) . . .

T. kamakou

1 8. Median lobe of seminal receptacles very large,
enveloping both dorsal and ventral bulbs (Fig.

28) T. polychromata

Median lobe of seminal receptacles smal-
ler, never enveloping either dorsal or ventral
bulbs 19

19. First tooth on retromargin of chelicerae ‘LT
larger than ‘L2’ (Fig. 37). Number of teeth

179

on promargin > number on retromargin.
Venter uniformly colored (particularly no-
ticeable in life). Chelicerae short (50-55%

length of carapace) T. brevignatha

First tooth on retromargin of chelicerae ‘LT
smaller than ‘L2’. Number of teeth on pro-
margin < number on retromargin. Venter
with distinct, narrow, median bar 20

20. Tibia I with 6 retrolateral and 6 (or 5) pro-
lateral spines 21

Tibia I with 5 retrolateral and 5 (or 4) pro-
lateral spines 22

21. Leg spines (length approx. 2.5 mm) equal to

or longer than carapace. Chelicerae long (60-
75% length of carapace) (Fig. 52)

T. macracantha

Leg spines (length approx. 1.5 mm) consid-
erably shorter than carapace. Chelicerae
shorter (55-65% length of carapace) (Fig. 1 3)

T. tantalus

22. Teeth on retromargin of chelicerae contigu-
ous, those on promargin nearly so (Figs. 75
and 76). Lateral eyes slightly separated from

each other (Fig. 77) T. kauaiensis

Teeth on retromargin of chelicerae well sep-
arated, as are those on promargin (Figs. 62
and 63). Lateral eyes contiguous (Fig. 64)

T. waikamoi

GREEN SPINY LEG GROUP

Characteristics. There are six species in this
group. Each of these has an elongate/oval ab-
domen, generally iridescent green with variable
red patterns superimposed. The legs are usually
rather pale and unbanded. The eyes are generally
small. The leg spines are long (44-105% length
of carapace). There are six species in this group:
T. tantalus, T. polychromata, T. brevignatha, T.
macracantha, T. waikamoi and T. kauaiensis.

Tetragnatha tantalus, new species
(Figs. 3-15 and 153)

Types. — Holotype male, allotype female from
Mount Tantalus, 1400 ft (427 m), Oahu Island
(25 October 1989), (coll. R.G. Gillespie and W.D.
Perreira), deposited in the Bishop Museum, Ho-
nolulu.

Etymology.— The specific epithet, regarded as
a noun in apposition, refers to the type locality
of the species, Mount Tantalus on the south-
eastern end of the Koolaus of Oahu.

Diagnosis. — r. tantalus is most easily con-
fused with T. polychromata. Males are distin-
guished as follows: (1) The distinctive conductor

Figures 3-\5. — Tetragnatha tantalus\ Male holotype. 3) Promargin of right chelicera; 4) Retromargin of left
chelicera; 5) Dorsal spur of chelicera, lateral view; 6) carapace, dorsal; 7) Right leg I, dorsal; 8) Right leg III,
prolateral; 9) Left palpus, prolateral. Female allotype. 10) Promargin of right chelicera; 1 1) Retromargin of left
chelicera; 12) Carapace, dorsal; 13) Right leg I, dorsal; 14) Right leg III, prolateral; 15) Seminal receptacles,
ventral. Scale bar (mm) at Fig. 12 applies to Figs. 3-6 and 10-12; at Fig. 8 to Figs. 7, 8 and 13, 14.

[Figs. 9 and 153] with the short apical projection
curling forward readily distinguishes it from all
others in the Green Spiny Leg group. (2) Tibia I
with 6 retrolateral, 2 dorsal, 6 [or 5] prolateral

spines [in T. polychromata tibia I has 5 retro-
lateral, 2 dorsal, 5 prolateral spines]; compare
Figs. 1 3 and 26. (3) First tooth on the male che-
licerae [‘si’] thicker than second [‘T’] and bent

GILLESPIE-HAWAIIAN TETRAGNATHA

181

up towards the top of the chelicerae [in T. po-
lychromata ‘sF is thinner than ‘T’, and projects
straight out]; compare Figs. 3 and 16. (4) Apical
tooth ‘Gu’ pronounced [in T. polychromata it is
small/absent]; compare Figs. 3 and 16. (5) Tip
of dorsal spur variably bifurcated or pointed [in
T. polychromata it is very pointed dorsally, slop-
ing sharply back ventrally]; compare Figs. 5 and
18.

Description. =//otovP^ male: i¥\g%. 3-9). Pro-
margin of chelicerae (Fig. 3): Distance between
‘Gu’ ‘sF and ‘T’ approximately equal, ratio of
distal end to ‘sF: ‘sF to ‘T’: ‘T’ to ‘rsuF 5:3:3 (2).
‘Gu’ pronounced, small and wide, flat-topped
tubercle; ‘sF robust, wide-based cone, pointed up
towards distal margin of chelicerae; much wider
than ‘T’, by 1 50% (100-155%), but shorter, 64%
height (51-78%). ‘T’ tall, thin, straight, dagger-
shaped. ‘rsu’ 4 (up to 7) straight spikes. Retro-
margin of chelicerae (Fig. 4): Total of 8 (up to
10) teeth. ‘AXF tiny notch; ‘GF and ‘L2’ strong,
stronger than rest of teeth on retromargin of che-
licerae. Dorsal spur long, shaped like slim, bent
finger (11.9% length of carapace); tip variably
bifurcated or pointed (Fig. 5). Cheliceral fang
slightly shorter than base, bent sharply over at
both proximal and distal ends. Length of ceph-
alothorax 1.9 mm (1.8-2. 2), total length 5.7 mm
(Fig. 6). Chelicerae slightly shorter (93%) than
length of carapace. Depression of thoracic fovea
indistinctly marked with broken semicircle on
prolateral margin. Leg spination similar to fe-
male, but spines shorter (Figs. 7-8). Femur I: 7
(6-8) prolateral, 2 dorsal, 7 retrolateral spines.
Tibia I: 5 (6) prolateral, 2 dorsal, 6 retrolateral
spines. Metatarsus I: 1 prolateral, 1 dorsal, 2
retrolateral spines. Femur III, no ventral spines.
Tibia III, 2 pairs of ventral spines and 2 single
spines. Coloration and eye pattern as in female.

Conductor Tip (Figs. 9 and 153): Conductor
cap clearly divided by high ridge leading up from
dorsal side of stem. Apical projection rather short
and curled forward.

Allotype female: (Figs. 10-15). PME separated
by approximately width of PME. Median ocular
area considerably wider posteriorly (Fig. 12).
Lateral eyes contiguous. Cheliceral margins: Pro-
margin (Fig. 10): series of 8 teeth ‘UT very ro-
bust, considerably wider but shorter (79%, 75-
85%) and well separated from (20%, 15-25%,
cheliceral length) ‘U2’ and ‘U3’. ‘U2’ and ‘U3’
of similar height, ‘U4’-’U8’ decreasing in size
proximally. Retromargin (Fig. 1 1): series of 9
teeth, ‘LI’ similar in height to ‘Ul’ and ‘L2’,

slightly separated from ‘L2’ and decreasing in
size proximally. Cheliceral fang quite long (ap-
proximately 90% length of base), tapering to
smooth point distally. Length of cephalothorax
2.1 mm (2.0-2. 5), total length 5.4 mm (4. 8-5. 8).
Cheliceiae shorter, 60% (55-65%) length of car-
apace. Legs unbanded, spines very distinct, but
considerably shorter (73%) than length of cara-
pace (Figs. 13, 14). Femur I: 8 (6-8) prolateral,
3 dorsal, 5 retrolateral spines. Tibia I: 6 (5) pro-
lateral, 2 dorsal, 6 retrolateral spines. Metatarsus
I: 1 prolateral, 1 dorsal, 2 retrolateral spines.
Femur III: 2 ventral spines. Tibia III: 2 pairs of
ventral spines and 2 single spines. Carapace pale
yellow (bright green in life) with indistinct fovea
marked by broken semicircle on prolateral mar-
gin. Sternum very pale yellow. Dorsum of ab-
domen uniformly pale yellow (bright green in
life), mostly plain, but sometimes with patches
of red (see color polymorphism below). Venter
pale whitish with distinct darker narrow band
running down midline.

Seminal receptacles (Fig. 1 5): Two bulbs linked
together in opposing “comma” shapes, each with
rather heavily sclerotized medial border. Neither
bulb greatly dilated at tip, and central portion
similar in width to bulbs. Median lobe smooth
doughnut shape that fits well within area defined
by outer limits of bulbs.

Color polymorphism. — Similar coloration and
its associated polymorphism are found in all the
currently known species of the Green Spiny Leg
group, T. tantalus, T. brevignatha, T. polychro-
mata, T. macracantha, T. waikamoi and T.
kauaiensis. All of these are bright lime green in
life, although all can exhibit color polymor-
phism, the most common polymorphism being
the presence of red patches on the dorsum of the
abdomen. These usually take the form of one or
series of red heart shapes. All species (except,
perhaps, T. tantalus, T. brevignatha and T. ma-
cracantha) are also capable of becoming much
more darkly pigmented, possibly due to envi-
ronmental conditions. This is particularly evi-
dent in T. polychromata and T. kauaiensis, both
of which can incorporate dark pigment (“melan-
ic” form), so gaining heavily banded legs, and
the dorsum of the abdomen becoming dark, mot-
tled green. However, the distinctive patterns
characteristic of species in the Green Spiny Leg
group are never similar to species outside this
group.

Material Examined. — This species is found in wet-

182

THE JOURNAL OF ARACHNOLOGY

Table 1.— Numbers of specimens collected at different sites (islands, volcanoes and elevations) through the
Hawaiian Islands.

Island Hawaii

(Mountain

Volcano Mauna Loa Saddle) Mauna Kea Kohala

South West West East East East East

Elevation (m X 1000) 1-2 1-2 0-1 1-2 0-1 1-2 0-1 1-2 0-1 1-2

T. tantalus

Male

Fern

Imm

T. polychromata

Male

Fern

Imm

T. hrevignatha

Male

Fern

Imm

T. macracantha

Male

Fern

Imm

T. waikamoi

Male

Fern

Imm

T. kauaiensis

Male

Fern

Imm

T. kamakou

Male

Fern

Imm

4 3

1 3

3 2 2 7

8 3 4
7 5 2
4 11 1

T. perreirai

Male

Fern

Imm

T. pilosa

Male

Fern

Imm

T. quasimodo

Male

16

37

Fern

19

50

Imm

58

62

T. restricta

Male

1

Fern

Imm

13

3

15

2

23

12

17

10

5

53

6

24

10

17

3

9

34

16

44

5

22

5 1

3

3 1

T. mohihi Male

Fern
Imm

mesic forest, only on Oahu Island, Koolau Mountains
(Table 1): Mount Tantalus, 1400 ft (427 m), 25-X-89
(R.G. Gillespie & W.D. Ferreira); Schofield-Waikane,
1910 ft (582 m), 30-IX-89 (R.G. Gillespie).

Tetragnatha polychromata, new species
(Figs. 16-28 and 154)

Types. — Holotype male from Peacock Flats,

Waianae Mountains, 1800 ft (550 m), Oahu Is-
land (18 August 1988) (coll. R.G. Gillespie and
C. Parrish), allotype female from Mount Kaala.
4000 ft (1220 m), Oahu Island (29 April 1990)
(coll. (R.G. Gillespie), deposited in the Bishop
Museum, Honolulu.

Etymology. — Poly (Greek) many; chromata
(Greek) colors. The specific epithet is an adjec-

GILLESPIE- HAWAIIAN TETRAGNATHA

183

Table L— Continued.

Hawaii

Maui

Molo-

kai

Lanai

Oahu

Kauai

Hua-

lalai

W.

Maui

Haleakala

Kam-

akou

Lanai-

hale

Wainaes

Koo-

laus

Waia-

leale

1-2

1-2

North

1-2

North East East

0-1 0-1 1-2

West

1-2

1-2

1-2

1-2 0-1

0-1

1-2

4

6

6

2

5

7

15

12

8

1 9

6 35

8 47

2 22 5

4 21 4

13 76 40

1 2

3

4 2

1 13

2 8

10 20

1 11

1 26

13 62

1 15

3 23

16 38

2

1

6

2 2 15

4 1 18

15 36

3
6
3

4 3 51 6

16 14 3 52 1

32 1 1 6 37 8

3 1

3 1

3

11

6

10

96

102

95

14

13

26

3

7

live referring to the presence of variable amounts
of red (if any) found on this vivid green species,
in addition to its ability to change color from
plain to melanic forms.

Diagnosis. — The distinctive conductor, which
lacks any form of curled tip or apical projection,
readily distinguishes T. polychromata from all
others in the Green Spiny Leg group (Fig. 1 54).

T. polychromata is most easily confused with T.
tantalus. These species can be distinguished as
mentioned above.

Description.— //o/orypc male: (Fig. 16-22).
Promargin of chelicerae (Fig. 16): Distance be-
tween ‘Gu’ ‘sF and ‘T’ approximately equal, ratio
of distal end to ‘sF: ‘sF to ‘T’: ‘T’ to ‘rsuF 4:3:3
(occasionally ‘sF may be little closer to ‘T’). ‘Gu’

1 84 the journal OF ARACHNOLOGY

Figures l6-2i. — Tetragnatha polychromata; Male holotype. 16) Promargin of right chelicera; 1 7) Retromargin
of left chelicera; 18) Dorsal spur of chelicera, lateral view; 19) carapace, dorsal; 20) Right leg I, dorsal; 21) Right
leg in, prolateral; 22) Left palpus, prolateral. Female allotype. 23) Promargin of right chelicera; 24) Retromargin
of left chelicera; 25) Carapace, dorsal; 26) Right leg I, dorsal; 27) Right leg III, prolateral; 28) Seminal receptacles,
ventral. Scale bar (mm) at Fig. 19 applies to Figs. 16-19; at Fig. 25 to Figs. 23-25; at Fig. 21 to Figs. 20-21; at
Fig. 27 to Figs. 26, 27.

small, often discernible only by hairs; ‘si’ tall,
straight, narrow cone, pointed up perpendicular
to margin of chelicerae. Much narrower than ‘T’,
by 63% (50-65%), and shorter, 50% height (36-

53%). ‘T’ tall, thin, straight, dagger-shaped, ‘rsu’
7 (5-7) straight spikes. Retromargin of chelicerae
(Fig. 17): Total of 1 1 teeth. ‘AXl’ tiny ill-defined
bump; ‘Gl’ and ‘L2’ strong, much stronger than

GILLESPIE- HAWAIIAN TETRAGNATHA

185

rest of teeth on retromargin. Dorsal spur long
{ 1 6.1% length of carapace, 1 5.5-20.0%), like slim,
bent finger, but with very pointed tip on dorsal
margin, sloping sharply back to ventral margin
(Fig. 18). Cheliceral fang slightly shorter than
base, bent sharply at both proximal and distal
ends. Length of cephalothorax 2.3 mm (1.3-2. 4),
total length 6.1 mm (Fig. 19). Chelicerae slightly
shorter (90%, 90-93%) than length of carapace.
Depression of thoracic fovea indistinctly marked
with broken semicircle on prolateral margin. Leg
spination similar to female, but spines shorter
(Figs. 20-21). Femur I: 9 prolateral, 3 dorsal, 5
retrolateral spines. Tibia I: 5 prolateral, 2 dorsal,
5 retrolateral spines. Metatarsus I: 1 prolateral,
1 dorsal, 2 retrolateral spines. Femur III: 1 ven-
tral spine. Tibia III: 2 pairs of ventral spines.
Coloration and eye pattern as in female.

Conductor Tip (Figs. 22 and 154): Angular,
flat-topped cap, terminating in smooth, straight
point without any form of curled apical projec-
tion.

Allotype female: (Figs. 23-28). PME separated
by just over width of PME (Fig. 25). Median
ocular area slightly wider posteriorly. Lateral eyes
contiguous. Cheliceral margins: Promargin (Fig.
23): series of 8 (7) teeth, ‘Ul’ slightly wider and
shorter (64%, 60-93%) than ‘U2’ and ‘U3’, and
well separated from them by 20% ( 1 8-26%) chel-
iceral length. ‘U2’ and ‘U3’ of similar height,
‘U4’-end decreasing in size proximally. Retro-
margin (Fig. 24): series of 1 1 teeth, ‘LI’ similar
in height to ‘Ul’, but smaller than ‘L2’ (56%);
teeth decrease in size proximally. Cheliceral fang
moderate (85% length of base), tapering to smooth
point at distal end. Length of cephalothorax 2.3
mm (2. 2-2.4), total length 6.4 mm (5. 5-6. 5).
Chelicerae 65% (60-70%) length of carapace. Legs
slightly banded, spines medium length, 75%
length of carapace (Figs. 26-27). Femur I: 7 pro-
lateral, 3 dorsal, 5 retrolateral spines. Tibia I: 5
prolateral, 2 dorsal, 5 retrolateral spines. Meta-
tarsus I: 1 prolateral, 1 dorsal, 2 retrolateral spines.
Femur III: 1 ventral spine. Tibia III: 2 pairs of
ventral spines. Carapace pale yellow (bright green
in life) with indistinct fovea marked by broken
semicircle around lateral margin. Sternum very
pale yellow. Dorsum of abdomen uniformly pale
yellow, although in life bright green, mostly plain,
but sometimes with patches of red (see color
polymorphism under T. tantalus). Venter pale
whitish with distinct darker narrow band run-
ning down midline.

Seminal receptacles (Fig. 28): Two bulbs linked

together in opposing “comma” shapes, each with
relatively heavily sclerotized medial border. Nei-
ther bulb greatly dilated at tip, and central por-
tion similar in width to bulbs. Median lobe: large
balloon, covering area much greater than that
defined by outer limits of bulbs.

Color polymorphism.— -See T. tantalus above.

Material Examined.- This species is found in wet-
mesic forest only on Oahu Island, Waianae Mountains
(Table 1); Waianae Kai, Waianae Mountains, 1900 ft
(580 m) 25-VI-88 (R.G. Gillespie & C. Parrish); Pea-
cock Flats, 1 800 ft (550 m), 1 8-Vni-88 (R.G. Gillespie
& C. Parrish); Summit of Mount Kaala, 4000 ft (1220
m) 29-IV-90 (R.G. Gillespie).

Tetmgnatha brevignatha, new species
(Figs. 29-41 and 155, 156)

Types. — Holotype male from Kaloko Road,
Hualalai, 3600 ft (1097 m), Hawaii Island (18
June 1989) (coll. R.G. Gillespie and C. Parrish),
allotype female from Hualalai, 3600 ft (1097 m),
Hawaii Island (30 July 1 988) (coll. R.G. Gillespie
and C. Parrish), deposited in the Bishop Muse-
um, Honolulu.

Etymology. — Brevis (Latin) short; gnathos
(Greek) jaw. The specific epithet is an adjective
referring to the short chelicerae of this species as
compared to others in the Green Spiny Leg group.

Diagnosis. — r. brevignatha is rarely confused
with other species, despite the fact that it and T.
waikamoi are the only species in the Green Spiny
Leg group known to date to have overlapping
ranges. The distinctive features that separate T.
brevignatha from all other species include: (1)
Short chelicerae [particularly so in males]; (2)
Venter uniformly colored, without a darker nar-
row band running down the midline. These char-
acters readily distinguish the species from T. wai-
kamoi and T. macracantha. The presence of a
small apical projection to the conductor cap
readily distinguishes it from T. polychromata and
T. tantalus. The most similar species in many
respects is T. kauaiensis. This species, however,
exhibits fundamental differences in cheliceral ar-
mature and leg spination.

Description.— //o/o/ype male: (Figs. 29-35).
Promargin of chelicerae (Fig. 29): Distance be-
tween ‘Gu’ ‘si’ and ‘T’ approximately equal, ratio
of distal end to ‘sP: ‘sF to ‘T’: ‘T’ to ‘rsul’ 4:3:3
(Note: applies to populations on Mauna Loa;
those from Maui and Mauna Kea appear to have
larger distance between ‘sF and ‘T’). ‘Gu’ absent;
‘sF well-developed, straight cone, pointed up per-
pendicular to margin of chelicerae; similar in

186

THE JOURNAL OF ARACHNOLOGY

Figures 29-4\. — Tetragnatha brevignatha\ Male holotype. 29) Promargin of right chelicera; 30) Retromargin
of left chelicera; 31) Dorsal spur of chelicera, lateral view; 32) carapace, dorsal; 33) Right leg I, dorsal; 34) Right
leg III, prolateral; 35) Left palpus, prolateral. Female allotype. 36) Promargin of right chelicera; 37) Retromargin
of left chelicera; 38) Carapace, dorsal; 39) Right leg I, dorsal; 40) Right leg III, prolateral; 4 1) Seminal receptacles,
ventral. Scale bar (mm) at Fig. 32 applies to Figs. 29-32; at Fig. 38 to Figs. 36-38; at Fig. 34 to Figs. 33-34; at
Fig. 40 to Figs. 39, 40.

GILLESPIE- HAWAIIAN TETRAGNATHA

187

width (87%, 75-100%) to T’, and only slightly
shorter, 71% height (this varies, 44-78%). ‘T’ tall,
thin, straight, dagger-shaped (sometimes slightly
bent up towards distal end of chelicerae). ‘rsu’ 5
(4-5) straight spikes. Retromargin of chelicerae
(Fig. 30): Total of 7 (6-8) teeth. ‘AXF absent;
‘Gr only very strong tooth, much stronger than
rest of teeth on retromargin. Dorsal spur short
(9.0% length of carapace, 6. 0-9. 9%), shaped like
fat, almost straight finger; tip pointed but not
sharply so (Fig. 31). Cheliceral fang distinctly
shorter than base, rather gently curved at both
proximal and distal ends. Length of cephalotho-
rax 2.2 mm (2. 0-2. 2), total length 5.8 mm (5.6-
6.0) (Fig. 32). Chelicerae much shorter (61%, 58-
64%) than length of carapace. Depression of tho-
racic fovea faint horseshoe-shape, with similarly
faint medial line running up from its anterior
margin. Leg spination similar to female, but
spines shorter (Figs. 33-34). Femur I: 7 prola-
teral, 3 dorsal, 7 retrolateral spines. Tibia I: 6
prolateral, 2 dorsal, 6 retrolateral spines. Meta-
tarsus I: 1 prolateral, 1 dorsal, 2 retrolateral spines.
Femur III: 2 ventral spines. Tibia III: 2 pairs of
ventral spines. Coloration and eye pattern as in
female.

Conductor Tip (Figs. 35 and 155): Smoothly
rounded cap, terminating in small apical projec-
tion that curls forwards.

Allotype female: (Figs. 36-41). PME separated
by approximately width of PME (Fig. 38). Me-
dian ocular area slightly wider posteriorly. Lat-
eral eyes contiguous (in representatives from Ha-
waii; usually - not always - well separated in
species from Maui). Cheliceral margins: Pro-
margin (Fig. 36): series of 8 (9) teeth ‘UF slightly
wider and shorter by 83% (60-100%) than ‘U2’
and ‘U3’, and separated from them by only 14%
(10-15%) cheliceral length. ‘U2’ and ‘U3’ of sim-
ilar height, with '■U4’-end decreasing in size prox-
imally. Retromargin (Fig. 37): series of 8 (7-9)
teeth, LI considerably larger (109% height, 105-
125%) than ‘L2’ and 70% height of ‘UF (range
70-140%); teeth decreasing in size proximally.
Cheliceral fang moderate (85% length of base),
tapering to smooth point at distal end. Length
of cephalothorax 2.2 mm (2. 0-2. 6), total length
5.2 mm (4. 5-5. 5). Chelicerae short, 53% (50-
55%) length of carapace. Legs unbanded, spines
medium length, 76% (70-80 %) length of cara-
pace (Figs. 39-40). Femur I; 8 prolateral, 2 dor-
sal, 7 retrolateral spines. Tibia I: 6 prolateral, 2
dorsal, 6 retrolateral spines. Metatarsus I: 1 pro-

lateral, 1 dorsal, 2 retrolateral spines. Femur III:
2 ventral spines. Tibia III: 1 pair of ventral spines
and 3 single spines. Carapace pale yellow (bright
green in life), depression of fovea unmarked.
Sternum very pale yellow. Dorsum of abdomen
uniformly pale yellow (bright green in life), most-
ly plain, but sometimes with patches of red (see
color polymorphism under T. tantalus). Venter
uniformly colored (particularly noticeable in life),
abdomen translucent green.

Seminal receptacles (Fig. 41): Two bulbs linked
together in opposing “C” shapes, each with rel-
atively heavily sclerotized medial border. Both
bulbs, in particular dorsal bulb, expanded at tips,
with constriction joining each to central portion.
Central portion similar in width to lower bulb,
with dorsal bulb wider than both. Median lobe
fits well within confines of upper and lower bulbs.

Color polymorphism. — See T. tantalus above.

Interisland Variation. — It is questionable
whether representatives from Maui and Hawaii
should be placed in different species. In deciding
to treat them as a single species, I took into ac-
count two factors: (1) only major difference be-
tween the islands is in separation of lateral eyes,
but this is not entirely consistent, and so unre-
liable [representatives on Maui tend to have well-
separated lateral eyes, whereas lateral eyes of those
on Maui are contiguous; but I have found one
individual on Maui with contiguous lateral eyes].
(2) Individuals from Mauna Kea on Hawaii are
more similar to those on Maui than they are to
others on Hawaii. In particular, the cap of the
conductor tip is broader in both of these popu-
lations (Fig. 1 56), than in other populations (Fig.
155). My suggestion is that the Maui population
was recently colonized by a representative(s) from
Mauna Kea. If this were true, it might also ex-
plain why T. brevignatha is the only member of
the Green Spiny Leg group to overlap with an-
other in the same group.

Material Examined.— This species is found in mesic
forest on Maui Island, wet-mesic on Hawaii Island
(Table 1): Hawaii Island, Puu Makaaia, Stainback
Highway, 4000 ft (1220 m), 14 and 21-X-90 (R.G.
Gillespie, D.J. Preston & 1. Felger) and 1 7-III-90 (R.G.
Gillespie & J.I.M. Gillespie); Laupahoehoe, 4120 ft
( 1 257 m) and 3200 ft (976 m), 1 9-X-90 (R.G. Gillespie,
D.J. Preston & J. Burgett); Laupahoehoe, 42 10 ft ( 1 284
m) and 4020 ft (1225 m), 13-III-90 (R.G. Gillespie &
J.I.M. Gillespie). Maui Island: East Maui, Waikamoi,
4400 ft ( 1 340 m), 8-VI-88 (R.G. Gillespie & A.C. Med-
eiros) and 8-II-90 (R.G. Gillespie & J. Burgett).

188

THE JOURNAL OF ARACHNOLOGY

Tetragnatha macracantha, new species
(Figs. 42-54 and 157)

Types. — Holotype male from Kipahulu Val-
ley, 4000 ft ( 1 220 m), Maui Island ( 1 5 May 1 990)
(coll. R.G. Gillespie and A.C. Medeiros), allo-
type female from Hanawi, 1500 ft, Maui Island
(11 May 1990) (coll. R.G. Gillespie, R. Rydell
and J. Burgett), deposited in the Bishop Museum,
Honolulu.

Etymology. — Makros (Greek) long; akantha
(Greek) spine. The specific epithet is an adjective
referring to the extraordinarily long spines on the
legs of this species, in particular the mature fe-
males, where the tibial spines are often as long
or longer than the carapace.

Diagnosis. — r. macracantha is most easily
confused with T. waikamoi, as both these species
are found on East Maui. Males can be distin-
guished as follows: (1) Tibia I with 6 retrolateral,
2 dorsal, 6 prolateral spines [in T. waikamoi tibia
I has 5 retrolateral, 2 dorsal, 5 prolateral spines].
(2) ‘si’ placed far down chelicerae, ratio of distal
end to ‘si’: ‘si’ to ‘T’: ‘T’ to ‘rsul’ 5:2:3 [in T.
waikamoi this ratio is 4:3:3, 4:3:4 or 3:3:4]. (3)
Apical tooth ‘Gu’ is absent [in T. waikamoi it is
pronounced]. (4) Conductor has a small apical
projection that curls forwards (Fig. 157) [in T.
waikamoi apical projection is very long and drawn
laterally outwards, terminating in a small for-
ward curl. Fig. 158]. These features also distin-
guish the species from others in the Green Spiny
Leg group.

Description.— male: (Figs. 42-48).
Promargin of chelicerae (Fig. 42): Distance be-
tween distal end and ‘si’ approximately equal to
distance between ‘si’ and ‘rsul’, ratio of distal
end to ‘si’: ‘si’ to ‘T’: ‘T’ to ‘rsul’ 5:2:3. ‘Gu’
absent; ‘si’ small peg, smaller than ‘T’ in width
(90%, 48-90%), much smaller in height (28%,
20-30%). ‘T’ tall, thin, dagger-shaped, very
slightly bent up towards distal end. ‘rsu’ 7 (5-7)
straight spikes. Retromargin of chelicerae (Fig.

43) : Total of 9 (8-10) teeth. ‘AXl’ absent; ‘Gl’
and ‘L2’ only slightly stronger than rest of teeth
on retromargin. Dorsal spur long (19.4% length
of carapace, 16.6-18.2%), shaped like slender,
bent finger, ending in distinctly blunt tip (Fig.

44) . Cheliceral fang almost same length as base,
abruptly curved at both proximal and distal ends.
Length of cephalothorax 2.0 mm (2.0-2. 2), total
length 5.1 mm (5.0- 5.5) (Fig. 45). Chelicerae
very slightly shorter (95%, 94-98%) than length
of carapace. Depression of thoracic fovea indis-

tinctly marked with pair of semicircles on lateral
margins, and with faint medial line running up
from its anterior margin. Leg spination similar
to female, but spines shorter (Figs. 46-47). Fe-
mur I: 7 prolateral, 3 dorsal, 5 retrolateral spines.

Tibia I: 6 prolateral, 2 dorsal, 6 retrolateral spines.
Metatarsus I; 1 prolateral, 1 dorsal, 2 retrolateral
spines. Femur III: 3 ventral spines. Tibia III; 1
pair of ventral spines and 3 single spines. Col-
oration and eye pattern as in female.

Conductor Tip (Fig. 48 and 157): Smoothly
rounded, very high-peaked cap, terminating in
small apical projection that curls forwards.

Allotype female: (Figs. 49-54). Eyes small, PME
separated by considerably more than width of
PME (Fig. 51). Median ocular area wider pos-
teriorly. Lateral eyes contiguous. Cheliceral mar-
gins: Promargin (Fig. 49): series of 7 teeth ‘Ul’
smaller and shorter (45% height, 45-65%) than
‘U2’ and ‘U3’, and separated from them by 21%
(20-40%) cheliceral length. ‘U2’ and ‘U3’ of sim-
ilar height, with ‘U4’-‘U7’ decreasing in size
proximally. Retromargin (Fig. 50): series of 9
teeth, LI smaller (73% height, 70-95%) than ‘L2’
and same height as ‘Ul’ (range 70-140% height);
teeth decreasing in size proximally. Cheliceral
fang moderate (82% length of base), tapering to
smooth point at distal end. Length of cephalo-
thorax 2.5 mm (2.0-2. 8), total length 5.5 mm
(5. 0-6.0). Chelicerae long, 61% (60-75%) length
of carapace. Legs unbanded, spines very long,
equal to or longer than length of carapace (Figs.
52-53). Femur I: 9 prolateral, 2 dorsal, 6 retro-
lateral spines. Tibia I: 6 prolateral, 2 dorsal, 6
retrolateral spines. Metatarsus I: 1 prolateral, 1
dorsal, 2 retrolateral spines. Femur III: 3 ventral
spines. Tibia III: 1 pair of ventral spines and 5
single spines. Carapace pale yellow (bright green
in life) with indistinct fovea marked with broken
semicircle around lateral margin. Sternum very
pale yellow. Dorsum of abdomen pale yellow in
alcohol, green in life, often with patches of red
(see color polymorphism under T. tantalus).

Venter pale white with distinct darker narrow
band running down midline.

Seminal receptacles (Fig. 54): Two bulbs linked ;
together in curl, so that upper bulbs run parallel I
to each other at midline, then make 90° turn to
connect to lower bulb. Neither bulb shows scle-
rotization except along margin where they con-
nect to each other. Both bulbs very slightly di-
lated, central portion forming narrow neck,
median lobe ill-defined.

Color polymorphism. — See T. tantalus above.

GILLESPIE-HAWAIIAN TETILAGNATHA

189

Figures 42-54. — Tetragnat ha macracantha\ Male holotype. 42) Promargin of right chelicera; 43) Retromargin
of left chelicera; 44) Dorsal spur of chelicera, lateral view; 45) carapace, dorsal; 46) Right leg I, dorsal; 47) Right
leg III, prolateral; 48) Left palpus, prolateral. Female allotype. 49) Promargin of right chelicera; 50) Retromargin
of left chelicera; 5 1) Carapace, dorsal; 52) Right leg I, dorsal; 53) Right leg III, prolateral; 54) Seminal receptacles,
ventral. Scale bar (mm) at Fig. 45 applies to Figs. 42-45; at Fig. 51 to Figs. 49-51; at Fig. 47 to Figs. 46-47; at
Fig. 53 to Figs. 52-53.

Material Examined. — This species is found in wet
forest only on Maui Island (Table 1): Kipahulu Valley,
2000 ft (610 m), lO-VI-89 (A.C. Medeiros) and 17-V-
90 (R.G. Gillespie & A.C. Medeiros); 3000 ft (9 1 4 m).

16-V-90 (R.G. Gillespie & A.C. Medeiros); 4000 ft
(1220 m). 1-V1-89(A.C. Medeiros) and 15-V-90(R.G.
Gillespie & A.C. Medeiros); 5000 ft (1524 m), 14-V-
90 (R.G. Gillespie & A.C. Medeiros); 6500 ft (1980

1 90 the journal of arachnology

Figures 55-61 . — Tetragnat ha waikamoh Male holotype. 55) Promargin of right chelicera; 56) Retromargin
of left chelicera; 57) Dorsal spur of chelicera, lateral view; 58) carapace, dorsal; 59) Right leg I, dorsal; 60) Right
leg III, prolateral; 61) Left palpus, prolateral. Female allotype. 62) Promargin of right chelicera; 63) Retromargin
of left chelicera; 64) Carapace, dorsal; 65) Right leg I, dorsal; 66) Right leg III, prolateral; 67) Seminal receptacles,
ventral. Scale bar (mm) at Fig. 58 applies to Figs. 55-58; at Fig. 64 to Figs. 62-64; at Fig. 60 to Figs. 59, 60;
at Fig. 66 to Figs. 65, 66.

GILLESPIE- HAWAIIAN TETRAGNATHA

191

m), 27-IV-88 (R.G. Gillespie & A.C. Medeiros). Han-
awi Valley, 1520 ft (463 m), 9-II-90 (R.G. Gillespie &
R. Rydell) and 1 l-V-90 (R.G. Gillespie, R. Rydell &
J. Burgett).

Tetragnatha waikamoi, new species
(Figs. 55-67 and 158)

Types. --Holotype male from Carruthers
Camp, Waikamoi, 61 50 ft (1876 m), Maui Island
(29 May 1988) (coll. R.G. Gillespie and C. Par-
rish), allotype female from Olinda, Waikamoi,
4460 ft (1360 m), Maui Island (15 July 1988)
(coll. R.G. Gillespie), deposited in the Bishop
Museum, Honolulu.

Etymology.— The specific epithet, regarded as
a noun in apposition, refers to the type locality
of the species, the Nature Conservancy of Ha-
waii’s Waikamoi Preserve on East Maui.

Diagnosis. — r. waikamoi is most easily con-
fused with T. macracantha. These species can be
distinguished as mentioned above. The distinc-
tive conductor, with its very long apical projec-
tion drawn laterally outwards and terminating in
a small forward curl (Fig. 158) readily distin-
guishes T. waikamoi from all others in the Green
Spiny Leg group.

Description.— //o/otype male: (Figs. 55-61).
Promargin of chelicerae (Fig. 55): Distance be-
tween ‘Gu’ ‘si’ and ‘T’ approximately equal, ratio
of distal end to ‘si’: ‘si’ to ‘T’: ‘T’ to ‘rsuF 4:3:3
(sometimes 4:3:4 or 3:3:4). ‘Gu’ present, large,
well-developed cone; ‘si’ medium-sized cone,
much smaller than ‘T’ in width (67%, 46-72%)
and in height (40%, 33-40%). ‘T’ very robust,
tall and straight, ‘rsu’ 5 (5-6) straight spikes. Re-
tromargin of chelicerae (Fig. 56): Total of 10 (9-
11) teeth. ‘AXr present and distinct; ‘Gl’ and
‘L2’ considerably stronger than rest of teeth on
retromargin. Dorsal spur long (18.3% length of
carapace, 18.1-18.5%), shaped like slender, bent
finger, ending in slightly rounded tip (Fig. 57).
Cheliceral fang distinctly shorter than base.
Length of cephalothorax 2.5 mm (2. 4-2. 8), total
length 6.1 mm (6. 0-7.0) (Fig. 58). Chelicerae al-
most same length (96%, 95-101%) as length of
carapace. Depression of thoracic fovea indis-
tinctly marked with pair of semicircles on pro-
lateral margins. Leg spination similar to female,
but spines shorter (Figs. 59-60). Femur I: 8 pro-
lateral, 3 dorsal, 4 retrolateral spines. Tibia I: 5
prolateral, 2 dorsal, 5 retrolateral spines. Meta-
tarsus 1: 1 prolateral, 1 dorsal, 2 retrolateral spines.
Femur III: 1 ventral spine. Tibia III: 2 pairs of

ventral spines. Coloration and eye pattern as in
female.

Conductor Tip (Figs. 61 and 158): Low, an-
gular cap leading to very long apical projection
drawn laterally outwards and terminating in small
forward curl.

Allotype female', (Figs. 62-67). Eyes small, PME
separated by just more than half width of PME
(Fig. 64). Median ocular area narrower posteri-
orly. Lateral eyes contiguous. Cheliceral margins:
Promargin (Fig. 62): series of 8 teeth ‘Ul’ wider
but shorter (70%) than ‘U2’ and ‘U3’, separated
from them by 10% (8-25%) cheliceral length.
‘U2’ and ‘U3’ of similar height, with ‘U4’-’U7’
decreasing in size proximally. Retromargin (Fig.
63): series of 9 teeth, ‘LF similar in height to
both ‘L2’ (87%, 85-100%) and ‘UF (100%, 98-
105%); teeth decreasing in size proximally. Chel-
iceral fang moderate (81% length of base), ta-
pering to smooth point at distal end. Length of
cephalothorax 2.4 mm (2. 2-2. 6), total length 5.4
mm (4. 5-6.0). Chelicerae medium length, 47%
(45-75%) length of carapace. Legs usually un-
banded, spines variable, 60% length of carapace
(Figs. 65-66). Femur I: 7 prolateral, 4 dorsal, 4
retrolateral spines. Tibia I: 5 retrolateral, 2 dor-
sal, 5 prolateral spines. Metatarsus I: 1 prolateral,
1 dorsal, 2 retrolateral spines. Femur III: 1 ven-
tral spine. Tibia III: 2 pairs of ventral spines.
Carapace pale yellow (bright green in life) with
indistinct fovea marked with broken semicircle
around prolateral margin. Sternum very pale yel-
low. Dorsum of abdomen pale yellow in alcohol,
green in life, often with patches of red (see color
polymorphism under T. tantalus). Venter pale
white with distinct darker narrow band running
down midline.

Seminal receptacles (Fig. 67): Two bulbs linked
together in opposing “comma” shapes, only low-
er bulb having relatively heavily sclerotized me-
dial border. Both bulbs, in particular dorsal bulb,
dilated, with central portion enveloped by me-
dian lobe. Median lobe smooth doughnut shape
that projects behind central portion, and projects
out into area approximately that defined by outer
limits of bulbs.

Color polymorphism. — See T. tantalus above.

Material Examined.— This species is found in wet
forest only on Maui Island (Table 1): West Maui, Puu
Kukui, 4550 ft (1387 m), 31-V-88 and l-VI-88 (R.G.
Gillespie & C. Parrish); East Maui, Waikamoi, 4400
ft ( 1 340 m), 8-VI-88 (R.G. Gillespie & C. Parrish) and
8-II-90 (R.G. Gillespie & J. Burgett); Waikamoi Flume,

192

THE JOURNAL OF ARACHNOLOGY

4400 ft (1340 m), 13-VIII-88 (R.G. Gillespie & C.
Parrish); Waikamoi, Carruthers Camp, 6150 ft (1876
m), 29-V-88 (R.G. Gillespie & C. Parrish) and 5-II-90
(R.G. Gillespie); Waikamoi, Honomanu Valley, 5200
ft (1585 m), 6-II-90 (R.G. Gillespie).

Tetragnatha kauaiensis Simon
(Figs. 68-80 and 161)

T. kauaiensis Simon (Simon 1900: 472, pi. XIX, fig.

9). Male holotype from Kauai, Halemanu, in the

Museum National d’Histoire Naturelle de Paris, ex-
amined. Okuma 1988c: 79-80, fig. 3.

Diagnosis. — r. kauaiensis is the only member
of the Green Spiny Leg group represented on
Kauai. In its melanic form, however, it might be
confused with T. pilosa and, perhaps, T. mohihi.
The diagnostic features are described under these
species.

Male: [holotype described by Simon (1900)
and redescribed by Okuma (1988c)]. Specimen
collected from Mohihi-Wailae Trail (DOFAW
Transect 5), 4000 ft (1220 m), Kauai Island (28
March 1990) (coll. R.G. Gillespie & C. Parrish):
Tooth arrangement on promargin of chelicerae
as follows (Fig. 68): ‘sP rather close to T’, ratio
of distal end to ‘sP: ‘sP to ‘T’: ‘T’ to ‘rsuF 3:2:
4. ‘Gu’ distinct notch that projects out; ‘sP small,
pointed spike, much smaller than ‘T’ in both
width (53%) and height (30%). ‘T' robust peak
directed out perpendicular to margin of chelic-
erae. ‘rsu’ 5 straight spikes. Retromargin of che-
licerae (Fig. 69): Total of 9 teeth, rather large,
pointed spikes. ‘AXP present, tiny notch; teeth
1 , 2 and 5-7 strongest teeth on retromargin. Dor-
sal spur fairly long, almost straight finger 17%
length of carapace; tip distinctly and equally bi-
furcate (Fig. 70). Cheliceral fang only slightly
shorter than base. Length of cephalothorax 2.3
mm, total length 5.8 mm. Chelicerae shorter
(70%) than length of carapace. Depression of tho-
racic fovea distinctly marked with inverted “Y”
shape (Fig. 71). Coloration and eye pattern as in
female. Leg spination similar to female, but spines
shorter (Figs. 72, 73).

Conductor Tip (Figs. 74 and 161): Cap pulled
strongly down at distal edge; apical projection
blunt tip that curls forward.

Female: Specimen collected from the Pihea-
Alakai Swamp Trail, 3800 ft (1158 m), Kauai
Island (8 June 1988) (coll. R.G. Gillespie & C.
Parrish). Eyes small, PME separated by well over
width of PME (Fig. 77). Median ocular area wid-
er posteriorly. Lateral eyes slightly separated.

Cheliceral margins; Promargin (Fig. 75): series
of 7 teeth ‘UF much shorter (50%) than ‘U2’
and ‘U3’, separated from them by 21% cheliceral
length. ‘U3’ slightly shorter than ‘U2’, with ‘U3’-
‘U7’ decreasing in size proximally. Retromargin
(Fig. 76): series of 10 teeth, rather tall, straight
spikes set close together. ‘LF smaller than ‘L2’
(71%), larger than ‘UF (1 18%); teeth decreasing
in size proximally. Cheliceral fang moderate (ap-
proximately 85% length of base), tapering to
smooth point at distal end. Length of cephalo-
thorax 2.4 mm (2. 2-2. 6), total length 5.5 mm
(4. 6-5. 9). Chelicerae medium in length, 51%
length of carapace. Legs usually unbanded, spines
variable, 58% length of carapace (Figs. 78-79).
Femur I; 6 prolateral, 3 dorsal, 4 retrolateral
spines. Tibia I: 5 retrolateral, 2 dorsal, 5 prola-
teral spines. Metatarsus I: 1 prolateral, 1 dorsal,
2 retrolateral spines. Femur III: no ventral spines.
Tibia III: 2 pairs of ventral spines. Carapace pale
yellow (bright green in life), fovea marked with
inverted “Y” shape. Sternum very pale yellow.
Dorsum of abdomen pale yellow in alcohol, green
in life, often with patches of red (see color poly-
morphism under T. tantalus). Venter pale white
with distinct darker narrow band running down
midline.

Seminal receptacles (Fig. 80): Two bulbs linked
together in opposing “comma” shapes, both up-
per and lower bulbs, as well as central portion,
having relatively heavily sclerotized medial bor-
der. Both bulbs equally dilated, with central por-
tion forming constricted “neck”. Median lobe
ill-defined.

Color polymorphism. — See T. tantalus ahowt.

Material Examined. — This species is found in wet
forest only on Kauai Island: Pihea-Alakai Swamp Trail,
3800 ft (1 1 58 m), 5-II-88 (R.G. Gillespie & A.C. Med-
eiros), 8-VI-88, 26-III-90, 22-VII-90 (R.G. Gillespie
& C. Parrish); Alakai Swamp, 3800 ft (1 158 m), 9-VI-
88 (R.G. Gillespie & C. Parrish); Mohihi Ditch, 3500
ft (1067 m), 27-111-90 (R.G. Gillespie & C. Parrish);
Mohihi-Wailae Trail (DOFAW Transect 5), 4000 ft
(1220 m), 28-III-90 (R.G. Gillespie & C. Parrish); Nu-
alolo Trail, Kuia, 3320 ft (1012 m), 21-VII-90 (R.G.
Gillespie & C. Parrish); Koaie Stream, 3700 ft (1128
m), 23-VII-90 (R.G. Gillespie & C. Parrish); Plateau
above Koaie Stream, 4000 ft ( 1 220 m), 24-VII-90 (R.G.
Gillespie & C. Parrish); Kokee/Kalalau Overlook, 4000
ft (1220 m), 27-VII-90 (R.G. Gillespie & C. Parrish).

GREEN AND RED SPINY LEGS GROUP

Characteristics. — There are 2 species in this
group, T. perreirai and T. kamakou. The derived

GILLESPIE- HAWAIIAN TETRAGNATHA

193

Figures 6^-S0. — Tetragnatha kauaiensis. 68) Promargin of right chelicera; 69) Retromargin of left chelicera;
70) Dorsal spur of chelicera, lateral view; 7 1 ) carapace, dorsal; 7 2) Right leg I, dorsal; 7 3) Right leg III, prolateral;
74) Left palpus, prolateral. Female allotype. 75) Promargin of right chelicera; 76) Retromargin of left chelicera;
77) Carapace and abdomen, dorsal; 78) Right leg I, dorsal; 79) Right leg III, prolateral; 80) Seminal receptacles,
ventral. Scale bar (mm) at Fig. 71 applies to Figs. 68-71; at Fig. 73 to Figs. 72-73; at Fig. 79 to Figs. 78-79.

(synapomorphic) features of this group relate pri-
marily to the shape and coloration of the ab-
domen, cephalothorax and legs, and the eye pat-
tern: Both species have a dark green/black
coloration to the diamond-shaped abdomen, with
a pattern that is highly distinctive in the presence
of a medial pair of maroon crescents that accen-
tuate the diamond shape, which may also be ex-

aggerated dorso-laterally to form 2 rounded
humps on either side of the midline. There is no
inverted triangle on the midline of the abdomen.
The pattern on the carapace is also very distinc-
tive, the entire dorsal surface dark, except for a
series of paler, roughly triangular-shaped, “is-
lands” that radiate out from the fovea. The legs
are heavily banded, usually dark. The eyes are

194

THE JOURNAL OF ARACHNOLOGY

Figures 81-94. — Tetragnatha kainakoir, Male holotype. 81) Promargin of right chelicera; 82) Retromargin of
left chelicera; 83) Dorsal spur of chelicera, lateral view; 84) carapace, dorsal; 85) Right leg I, dorsal; 86) Right
leg III, prolateral; 87) Left palpus, prolateral. Female allotype. 88) Promargin of right chelicera; 89) Retromargin
of left chelicera; 90) Carapace, dorsal; 91) Right leg I, dorsal; 92) Right leg III, prolateral; 93) abdomen, dorsal;
94) Seminal receptacles, ventral. Scale bar (mm) at Fig. 84 applies to Figs. 81-84; at Fig. 90 to Figs. 88, 90; at
Fig. 86 to Figs. 85, 86; at Fig. 92 to Figs. 91, 92.

GILLESPIE- HAWAIIAN TETHiGNATHA

195

rather large and often close together. The leg
spines are relatively short (28-58% length of car-
apace). There are 2 species in this group: T. ka-
makou and T. perreirai.

Tetragnatha kamakou, new species
(Figs. 81-94 and 159-160)

Types. — Holotype male, allotype female from
Kamakou, Puu Kolekole, 3950 ft (1205 m), Mo-
lokai Island (21 June 1988) (coll. R.G. Gillespie
and C. Parrish), deposited in the Bishop Muse-
um, Honolulu.

Etymology.— The specific epithet, regarded as
a noun in apposition, refers to the type locality
of the species, the Nature Conservancy of Ha-
waii’s Kamakou Preserve on Molokai.

Diagnosis. — r. kamakou is most easily con-
fused as indicated:

A. On Molokai with T. quasimodo. Either sex
can be distinguished as follows: (1) abdominal
pattern; (2) Sternum pale translucent yellowish
[black in T. quasimodo]-, (3) Venter uniformly
medium-brown [in T. quasimodo it has medial
pale fawn bar, narrower posteriorly]; (4) leg spi-
nation: Tibia I: 4 (or 5) medial, 2 dorsal, 5 lateral
spines (in T. quasimodo tibia I has: 4 medial, 2
dorsal, 4 lateral spines). Males lack distinctive
wave shape of first tooth ‘sP found in T. quasi-
modo, ‘sP being almost contiguous with ‘T’. The
conductor tip is also characteristic.

B. On Maui with: (a) T. quasimodo. The same
characters can be used to distinguish species, ex-
cept that leg spination in the two is often the
same on this island, (b) T. waikamoi in its more
melanic form. The most useful characters for
distinguishing these species are (1) abdominal
pattern. (2) Venter uniformly medium-brown (in
T. waikamoi it has a medial narrow darker bar).
(3) Leg spination: Tibia I: 4 medial, 2 dorsal, 4
lateral spines [in T. waikamoi tibia I has: 5 me-
dial, 2 dorsal, 5 lateral spines].

T. kamakou can be distinguished from T. per-
reirai in both sexes because of the much smaller
chelicerae in T. perreirai: In males, these are 69-
70% length of the carapace in T. perreirai, 87-
96% in T. kamakou. In females, these are 54%
length of the carapace in T. perreirai, 67-69% in
T. kamakou. The armature on the male chelic-
erae is entirely different in the two species, in
particular the shape and length of the dorsal spur
(9. 3-9. 8% length of the carapace and pointed in
T. perreirai, 15.6-18.7% and unequally bifur-
cated in T. kamakou).

Description. — male: (Figs. 81-87).
Promargin of chelicerae (Fig. 81): Distance be-
tween ‘Gu’ ‘sP and ‘T’ approximately equal, ratio
of distal end to ‘sP: ‘sP to ‘T’: ‘T’ to ‘rsul’ 4:3:3
(occasionally ‘sP may be little closer to ‘T’). ‘Gu’
present, small, flat-topped tubercle; ‘sP small,
pointed cone, directed perpendicular to cheli-
ceral margin, smaller than ‘T’ in both width (80%,
30-80%) and height (24%, 24-34%). ‘sP well sep-
arated from ‘T’, largest tooth (15.5% cheliceral
length), robust peak directed almost perpendic-
ular to margin of chelicerae. ‘rsu’ 5 (5-7) straight
spikes. Retromargin of chelicerae (Fig. 82): Total
of 14 (11-14) teeth, long battery of small pegs
‘GP and ‘L2’ largest, robust; 3-5 and 10-end
smallest. Dorsal spur long, curved finger, 18.7%
(15.7-18.7%) length of carapace; distinct semi-
bifurcated tip (dorsal side projects further for-
ward) (Fig. 83). Cheliceral fang approximately
93% length of base, bent over at both proximal
and distal ends. Length of cephalothorax 2.6 mm
(2. 2-2. 6), total length 6.7 mm (5. 3-6. 8) (Fig. 84).
Chelicerae only slightly shorter (96%, 87-96%)
than length of carapace. Depression of thoracic
fovea distinctly marked with inverted “V” shape.
Leg spination similar to female, but spines slight-
ly shorter (Figs. 85, 86). Femur I: 6 prolateral, 1
dorsal, 4 retrolateral spines. Tibia I: 5 prolateral,
2 dorsal, 5 retrolateral spines. Metatarsus I: 1
prolateral, 1 dorsal, 2 retrolateral spines. Femur
III: 1 ventral spine. Tibia III: 2 pairs of ventral
spines. Coloration and eye pattern as in female.
Shape of abdomen as in female, but medial di-
lation less pronounced.

Conductor Tip (Figs. 87 and 159): smoothly
rounded, high-peaked cap, terminating in small
apical projection that curls laterally outwards.

Allotype female: (Figs. 88-94). Eyes large, tak-
ing up most of ocular area (Fig. 90). PME sep-
arated by 70% width of PME. Median ocular area
wider at back than at front. Lateral eyes closely
contiguous. Cheliceral margins: Promargin (Fig.
88): series of 7 (8) short, thick teeth, ‘U2’-‘U4’
largest. ‘Ul’ 76% (40-80%) height of ‘U2’ and
well separated from it by 26% (20-30%) cheli-
ceral length; teeth decreasing in size proximally.
Retromargin (Fig. 89): series of 1 2 (9-12) slightly
smaller, robust teeth, 2 and 3 slightly larger than
rest, slightly separated from ‘LI’. ‘LI’ 77% (75-
85%) height of ‘Ul’ and 71% (50-90%) height of
‘L2’; teeth decreasing in size proximally. Cheli-
ceral fang approximately 80% length of base, ta-
pering to smooth point at distal end. Length of

196

THE JOURNAL OF ARACHNOLOGY

cephalothorax 2.7 mm (2. 4-2. 8), total length 6.7
mm (4. 5-7. 7). Chelicerae long, 69% (65-75%)
length of carapace. Legs banded, spines very dis-
tinct but relatively short, 54% (45-60%) length
of carapace (Figs. 91-92). Femur I: 5 (4) prola-
teral, 3 dorsal, 5 (4) retrolateral spines. Tibia I:
4 (5) prolateral, 2 dorsal, 5 (4) retrolateral spines.
Metatarsus I; 1 prolateral, 1 dorsal, 3 (2) retro-
lateral spines. Femur III: 1 ventral spine. Tibia
III: 2 pairs of ventral spines. Depression of ce-
phalothoracic fovea an elongate diamond shape.
Cephalothoracic pattern very distinct, broad “Y”
shape, with variable pairs of lateral wings radi-
ating from in front of thoracic fovea. Sternum
uniformly pale yellowish. Abdomen with pair of
distinct, medial crescents on either side; no black
inverted triangle (Fig. 93). Rest of pattern con-
sists of series of paired parallel marks running
down on either side of midline. Venter uniformly
dark.

Seminal receptacles (Fig. 94): Two bulbs linked
together in opposing “comma” shapes; sclero-
tization very weak. Both bulbs slightly expanded
at tips, lower bulb has long, rather narrow, stalk
joining it to central portion. Central portion sim-
ilar in width to upper bulb. Median lobe in form
of expanded balloon that projects out into area
approximately defined by outer limits of bulbs.

Color polymorphism. — I have found very little
evidence of any color polymorphism in either
members of Green and Red Spiny Leg group,
although some specimens are much darker than
others.

Interisland Variation. — Molokai representa-
tives have a longer dorsal spur (18.5-18.7% length
of carapace, as compared to 15.7-16.7% on
Maui). They are also a little larger (average length
of carapace 2. 5-2. 6 mm, as compared to 2.4-2. 5
on Maui) and the first leg is longer (10.0-10.2%
length of carapace as compared to 8. 3-8. 8% on
Maui). Maui representatives usually have 4 lat-
eral spines on the tibia of the first leg, whereas
there appear always to be 5 in Molokai repre-
sentatives. The conductor cap is also much
broader in Maui representatives (Fig. 160) than
those on Molokai (Fig. 159). This latter differ-
ence is striking, but at present I am considering
these populations to belong to the same species.

Material Examined. — This species is found only in
wet forest on Molokai and Maui islands (Table 1):
Molokai Island, Kamakou, 3800 ft (1158 m), 21-23-
Vl-88, 1-II-90 (R.G. Gillespie & C. Parrish); Kaun-
uohua Summit, Kamakou, 4535 ft (1382 m), 2-II-90

(R.G. Gillespie & J. Halloran). Maui Island: West Maui,
Puu Kukui, 4550 ft (1387 m), 31-V-88 and 1 -VI-88
(R.G. Gillespie & C. Parrish); East Maui, Waikamoi,
4400 ft ( 1 340 m), 8-VI-88 (R.G. Gillespie & C. Parrish)
and 8-II-90 (R.G. Gillespie & J. Burgett); Waikamoi
Flume, 4400 ft (1340 m), 13-VIII-88 (R.G. Gillespie
& C. Parrish); Waikamoi, Carruthers Camp, 6150 ft
(1876 m), 29-V-88 (R.G. Gillespie & C. Parrish) and
5-11-90 (R.G. Gillespie); Waikamoi, Honomanu Val-
ley, 5200 ft (1585 m), 6-11-90 (R.G. Gillespie); Bogs,
NE Rift Haleakala, 5500 ft (1676 m), 15-1-88 (R.G.
Gillespie & A.C. Medeiros), Kipahulu Valley, 4000 ft
(1220 m), 15-V-90 (R.G. Gillespie & A.C. Medeiros);
5000 ft (1524 m), 14-V-90 (R.G. Gillespie & A.C.
Medeiros); 6500 ft (1980 m), 27-1 V-88 (R.G. Gillespie
& A.C. Medeiros).

Tetragnatha perreirai, new species
(Figs. 95-108 and 162)

Types. — Holotype male from Mount Kaala,
Waianae Mountains, 4000 ft (1220 m), Oahu
Island (8 January 1990) (coll. W.D. Ferreira),
allotype female from Mount Kaala, Waianae
Mountains, 4000 ft (1220 m), Oahu Island (29
April 1990) (coll. R.G. Gillespie), deposited in
the Bishop Museum, Honolulu.

Etymology. — The specific epithet (possessive
noun) is in recognition of the collector of the
holotype, W.D. Ferreira, an excellent naturalist
and entomologist in the Hawaiian Evolutionary
Biology Frogram.

Diagnosis. — r. perreirai is most easily con-
fused with T. polychromata in its more melanic
form. The most useful characters for distinguish-
ing these species are: (1) Abdominal pattern. (2)
Venter uniformly medium-brown [in T. poly-
chromata it has a medial narrow darker bar]. (3)
Cheliceral length [in males, these are 69-70%
length of carapace in T. perreirai, 90-93% in T.
polychromata. In females, they are 54% length
of carapace in T. perreirai, 66% in T. polychro-
mata]. The armature on male chelicerae is en-
tirely different in the 2 species, in particular the
shape and length of the dorsal spur (9. 3-9. 8%
length of carapace and almost straight in T. per-
reirai, 1 5.5-20.0% and bent in T. polychromata).
It can be distinguished from T. kamakou as de-
scribed above.

Description.- male: (Figs. 95-101).
Fromargin of chelicerae (Fig. 95): ‘rsuF placed
close to ‘T’, ratio of distal end to ‘sF: ‘sF to ‘T’:
‘T’ to ‘rsul’ 4:4:2. ‘Gu’ medium-sized, pointed
cone, approximately same width and height as
‘sF; ‘sF robust, wide-based cone, smaller than
‘T’ in both width (61%, 60-63 %) and height

GILLESPIE- HAWAIIAN TETRAGNATHA

197

Figures 95-108. — Tetragnatha peneimv, Male holotype. 95) Promargin of right chelicera; 96) Retromargin of
left chelicera; 97) Dorsal spur of chelicera, lateral view; 98) carapace, dorsal; 99) Right leg I, dorsal; 100) Right
leg III, prolateral; 101) Left palpus, prolateral. Female allotype. 102) Promargin of right chelicera; 103) Retro-
margin of left chelicera; 104) Carapace, dorsal; 105) Right leg I, dorsal; 106) Right leg III, prolateral; 107)
abdomen, dorsal; 108) Seminal receptacles, ventral. Scale bar (mm) at Fig. 95 applies to Figs. 95-97; at Fig.
104 to Figs. 102-104; at Fig. 100 to Figs. 99, 100; at Fig. 106 to Figs. 105, 106.

(53%, 45-54%). ‘T’ robust peak, bent slightly up
towards distal end. ‘rsu’ 6 straight spikes. Retro-
margin of chelicerae (Fig. 96): Total of 8 teeth,
in addition to ‘AXf. ‘AXl’ as large as main teeth;
‘Gl’ and ‘L2’ wider than rest of teeth on retro-
margin. Dorsal spur short, stubby, pointed fin-
ger, 9.8% (9. 2-9. 9 %) length of carapace; tip

pointed on dorsal side (Fig. 97). Cheliceral fang
considerably shorter than base, bent very slightly
and smoothly over at distal end. Length of ceph-
alothorax 2.3 mm (2. 1-2.4), total length 5.7 mm
(Fig. 98). Chelicerae much shorter (69%) than
length of carapace. Depression of thoracic fovea
distinctly marked with inverted “Y” shape. Leg

198

THE JOURNAL OF ARACHNOLOGY

spination similar to female, but spines slightly
shorter (Figs. 99-100). Femur I: 5 prolateral, 5
dorsal, 4 retrolateral spines. Tibia I: 4 prolateral,
2 dorsal, 4 retrolateral spines. Metatarsus I: 1
prolateral, 1 dorsal, 2 retrolateral spines. Femur
III; 1 ventral spine. Tibia III: 2 pairs of ventral
spines. Coloration and eye pattern as in female.
Shape of abdomen as in female, but medial di-
lation less pronounced.

Conductor Tip (Figs. 101 and 162): smoothly
rounded, high-peaked cap, terminating in small
apical projection that has thin curl that projects
laterally outward at tip.

Allotype female-. (Figs. 102-108). Eyes large,
occupying most of ocular area (Fig. 104). PME
separated by just over half width of PME. Me-
dian ocular area wider at back than in front.
Lateral eyes contiguous. Cheliceral margins: Pro-
margin (Fig. 1 02): series of 6 medium-sized teeth
‘U2’ and ‘U3’ largest, well separated from ‘UF.
‘Ur 77% height of ‘U2’, separated from it by
18% cheliceral length; teeth decreasing in size
proximally. Retromargin (Fig. 103): series of 10
small teeth, set very close together and of similar
size (last couple smaller); 'L 1 ’ approximately same
height as both ‘UF and ‘L2’. Cheliceral fang ap-
proximately 83% length of base, tapering to
smooth point at distal end. Length of cephalo-
thorax 2.3 mm, total length 4.4 mm. Chelicerae
much shorter (54%) than length of carapace. Legs
dark and distinctly banded, spines very distinct,
although leg spines relatively short (approxi-
mately 33% length of carapace) (Figs. 105-106).
Femur I: 4 prolateral, 3 dorsal, 4 retrolateral
spines. Tibia I: 4 prolateral, 2 dorsal, 4 retrola-
teral spines. Metatarsus I: 1 prolateral, 1 dorsal,
2 retrolateral spines. Femur III: no ventral spines.
Tibia III: 2 pairs of ventral spines. Depression
of cephalothoracic fovea distinctly marked with
inverted “Y” shape on anterior margin. Cara-
pace dark brown with 3 pairs of pale triangular
blocks radiating from fovea. Sternum pale yel-
lowish. Abdomen oval/diamond-shaped, exag-
gerated dorso-laterally into 2 lateral, rounded
humps (Fig. 107). Color pattern consists of var-
ious combinations of red (on lateral humps) and
dark green. Venter uniformly colored.

Seminal receptacles (Fig. 108): Two bulbs
linked together in opposing “comma” shapes,
each with narrow sclerotized medial border. Both
bulbs, in particular upper, dilated at tip, with
central portion narrower. Median lobe ill-de-
fined, semi-collapsed balloon that projects out

into area approximately that defined by outer
limits of bulbs.

Color polymorphism.— See T. kamakou above.

Material Examined. — This species is found only in
bog habitat on Oahu Island. Waianae Mountains (Ta-
ble 1); Summit of Mount Kaala, 4000 ft (1220 m), 29-
lV-90 (R.G. Gillespie).

Tetragnatha pilosa, new species
(Figs. 109-122 and 163)

Types.- Holotype male from Mohihi Ditch,
approximately 3.2 km beyond the end of Camp
10 Road, 3500 ft (1067 m), Kauai Island (26
March 1990) (coll. R.G. Gillespie and C. Par-
rish), allotype female from the Alakai Swamp,
4000 ft (1220 m), Kauai Island (6 March 1988)
(coll. R.G. Gillespie and C. Parrish), deposited
in the Bishop Museum, Honolulu.

Etymology. — From pilus (Latin), hair. The
specific epithet is an adjective referring to the
extraordinarily “hairy” looking femora of the
third legs. These are very much longer and more
numerous than on any other species of Hawaiian
Tetragnatha.

Diagnosis. — r. pilosa is unlikely to be con-
fused with any other species because of its very
distinctive leg spination on femora of 3rd leg,
and its color pattern. It might be possible to con-
fuse it with T. kauaiensis in the more melanic
form of this species. The most useful characters
for distinguishing T. pilosa are (1) Leg spination
[especially on femora of 3rd leg]. (2) Cephalo-
thoracic pattern: Only T. pilosa has arms of “Y”
shape running parallel then turning sharply to-
wards stem. In males, the dorsal spur in T. pilosa
is much shorter [8. 5-9. 5% length of carapace
as compared to 13.9% in T. kauaiensis].

'Descnption.— Holotype male: i¥'\gs,. 109-1 15).
Promargin of chelicerae (Fig. 109): Distance be-
tween distal end and ‘sP approximately equal to
distance between ‘sF and ‘rsuF, ratio of distal
end to ‘sF: ‘sF to ‘T’: ‘T’ to ‘rsuF 5:2:3 (occa-
sionally 5:3:2). ‘Gu’ distinct, medium-sized cone,
almost same size as ‘sF; ‘sF rather small spike,
pointed slightly up towards distal end. Much nar-
rower than ‘T’, by 33% (30-50%), and shorter,
63% height (53-63%). ‘T’ robust, pointed very
slightly up towards distal end. ‘rsu’ 6 straight
spikes. Retromargin of chelicerae (Fig. 110):
Total of 8 (up to 10) teeth. ‘AXF absent. Dorsal
spur short (8.7% length of carapace, 8.6-9. 5%),
shaped like stubby, almost straight, finger, with

GILLESPIE- HAWAIIAN TETRAGNATHA

199

Figures \09-\12. — Tetragnatha pilosa\ Male holotype. 109) Promargin of right chelicera; 1 10) Retromargin
of left chelicera; 111) Dorsal spur of chelicera, lateral view; 1 12) carapace, dorsal; 1 13) Right leg I, dorsal; 1 14)
Right leg III, prolateral; 1 15) Left palpus, prolateral. Female allotype. 1 16) Promargin of right chelicera; 1 17)
Retromargin of left chelicera; 1 18) Carapace, dorsal; 1 19) Right leg I, dorsal; 120) Right leg III, prolateral; 121)
abdomen, dorsal; 122) Seminal receptacles, ventral. Scale bar (mm) at Fig. 1 1 5 applies to Figs. 111-11 5; at Fig.
1 18 to Figs. 1 16-118; at Fig. 1 14 to Figs. 113, 1 14; at Fig. 120 to Figs. 1 19, 120.

rounded tip (Fig. 111). Cheliceral fang distinctly
shorter than base and smoothly curved at both
proximal and distal ends (not bent sharply over).
Length of cephalothorax 2.4 mm (2. 3-2.4), total
length 5.9 mm (Fig. 1 1 2). Chelicerae much short-

er (62%, 56-64%) than length of carapace. De-
pression of thoracic fovea distinctly marked with
smoothly rounded “m” shape. Leg spination
similar to female, but spines shorter (Figs. 1 1 3-
1 14). Femur I: 6 prolateral, 4 dorsal, 2 retrola-

200

THE JOURNAL OF ARACHNOLOGY

teral spines. Tibia I: 6 prolateral, 2 dorsal, 5 (6)
retrolateral spines. Metatarsus I: 1 prolateral, 1
dorsal, 3 retrolateral spines. Femur III: 5 ventral
spines. Tibia III: 3 pairs of ventral spines. Col-
oration and eye pattern as in female.

Conductor Tip (Figs. 1 15 and 163): Cap low
and angular, pulled out laterally into “hooked-
nose” shape. Terminates in small hook that pro-
jects laterally forwards.

Allotype female: (Figs. 1 16-122). All eyes large,
occupying most of ocular area (Fig. 1 18). PME
separated by just over half width of PME. Me-
dian ocular area slightly wider in front than back.
Lateral eyes contiguous. Cheliceral margins: Pro-
margin (Fig. 1 16): series of 8 large, robust teeth
‘U2’ and ‘U3’ largest, well separated from ‘UF.
‘U r 71% height of ‘U2’, separated from it by
1 9% ( 1 8-23%) cheliceral length; teeth decreasing
in size proximally. Retromargin (Fig. 1 1 7): series
of 9 slightly smaller, robust teeth ‘L2’ and ‘L3’
largest, slightly separated from ‘LF (‘LF 67%
height of ‘UF and 62% height of ‘L2’); teeth
decrease in size proximally. Cheliceral fang short
(78% length of base), tapering to smooth point
at distal end. Length of cephalothorax 2.7 mm
(2. 5-2. 7), total length 5.5 mm (5. 3-6. 7). Chelic-
erae much shorter (54%) than length of carapace.
Legs distinctly banded, spines very distinct, al-
though leg spines relatively short (44 % length
of carapace) (Figs. 1 19, 120). Femur I: 6 prola-
teral, 4 dorsal, 3 retrolateral spines. Tibia I: 5 (6)
prolateral, 2 dorsal, 5 (6) retrolateral spines.
Metatarsus I: 2 prolateral, 1 dorsal, 3 retrolateral
spines. Femur III: 10 ventral spines. Tibia III: 3
pairs of ventral spines. Depression of cephalo-
thoracic fovea distinctly marked with rounded
“m” shape on anterior margin. Cephalothoracic
pattern distinct angular “Y” shape, unusual in
having arms run parallel before turning sharply
to converge at stem. Three (1-3) pairs of lines
radiating out from fovea to edge of carapace.
Sternum yellow with fairly wide dark border ex-
cept on anterior margin. Abdomen oval, pattern
as shown in Fig. 121. Venter pale grey with 2
pairs of silvery dots on either side of midline.

Seminal receptacles (Fig. 122): Two bulbs
linked together in opposing “comma” shapes;
sclerotization weak. Upper bulb dilated, lower
bulb narrower, same width as central portion;
not greatly dilated. Median lobe angular dough-
nut shape that projects out into area approxi-
mately that defined by outer limits of bulbs.

Color polymorphism. — Little evidence of this.

Material Examined.— This species has been found
only in wet forest on Kauai Island (Table 1): Pihea-
Alakai Swamp Trail, 3800 ft (1 158 m), 5-II-88 (R.G.
Gillespie & A.C. Medeiros), 8-VI-88, 26-III-90, 22-
VII-90 (R.G. Gillespie & C. Parrish); Alakai Swamp,
3800 ft (1158 m), 9-VI-88 (R.G. Gillespie & C. Par-
rish); Mohihi Ditch, 3500 ft (1067 m), 27-III-90 (R.G.
Gillespie & C. Parrish); Mohihi-Wailae Trail (DOFAW
Transect 5), 4000 ft ( 1 220 m), 28-III-90 (R.G. Gillespie
& C. Parrish); Nualolo Trail, Kuia, 3320 ft (1012 m),
21-VII-90 (R.G. Gillespie & C. Parrish; Koaie Stream,
3700 ft (1 128 m), 23-VII-90 (R.G. Gillespie & C. Par-
rish); Plateau above Koaie Stream, 4000 ft (1220 m),
24-VI1-90 (R.G. Gillespie & C. Parrish); Kokee/Ka-
lalau Overlook, 4000 ft (1220 m), 27-VII-90 (R.G.
Gillespie & C. Parrish).

Tetmgnatha quasimodo, new species
(Figs. 123-136 and 166-169)

Types. — Holotype male from Waianae Kai,
Waianae Mountains, 1900 ft (580 m), Oahu Is-
land (25 June 1988) (coll. R.G. Gillespie, J.S.
Strazanac and C. Parrish), allotype female from
Volcano Village, 3500 ft (1067 m), Hawaii Island
(17 June 1989) (coll. R.G. Gillespie and C. Par-
rish), deposited in the Bishop Museum, Hono-
lulu.

Etymology.— The common name of this spe-
cies is “Humpback Spiny”, because of the prom-
inent mid-dorsal peak of the abdomen. The spe-
cific epithet, regarded as a noun in apposition,
refers to Victor Hugo’s “Hunchback of Notre
Dame”.

Diagnosis. — r. kamakou and T. perreirai are
the only species with which T. quasimodo might
be confused. The abdomen in T. quasimodo is
widest in the middle, with a medial distinct black
inverted triangle just below the mid- ventral line.
Sternum dark-dusky. Legs banded and clothed
with robust spines. In the male, the first tooth
‘sP takes the form of a strong, down-curved wave,
almost contiguous with the erect and pointed 2nd
tooth ‘T’. Other distinguishing features have been
discussed above.

Description. —//o/ofvp^ male: (Figs. 1 23-129).
Promargin of chelicerae (Fig. 123): Distance be-
tween distal end and ‘sP very long, ratio of distal
end to ‘sP: ‘sP to ‘T’: ‘T’ to ‘rsuF 5:1:4 (occa-
sionally 6:1:3). ‘Gu’ present, small tubercle; ‘sP
large, very distinctive wave shape pointing prox-
imally, almost contiguous with ‘T’; almost ex-
actly same width as ‘T’, but considerably shorter,
39% height (37-47%). ‘T’ robust peak directed
perpendicular to margin of chelicerae (separation

GILLESPIE- HAWAIIAN TETRAGNATHA 201

Figures \23-l36. — Tetragnatha qiiasimodo\ Male holotype. 123) Promargin of right chelicera; 124) Retro-
margin of left chelicera; 125) Dorsal spur of chelicera, lateral view; 126) carapace, dorsal; 127) Right leg I,
dorsal; 128) Right leg III, prolateral; 129) Left palpus, prolateral. Female allotype. 130) Promargin of right
chelicera; 131) Retromargin of left chelicera; 132) Carapace, dorsal; 133) Right leg I, dorsal; 134) Right leg III,
prolateral; 135) abdomen, dorsal; 136) Seminal receptacles, ventral. Scale bar (mm) at Fig. 126 applies to Figs.
123-126; at Fig. 132 to Figs. 130-132; at Fig. 128 to Figs. 127, 128; at Fig. 134 to Figs. 133, 134.

202

THE JOURNAL OF ARACHNOLOGY

between ‘si’ and ‘T’ only 4-8% of cheliceral
length), ‘rsu’ 4 (up to 5) straight spikes. Retro-
margin of chelicerae (Fig. 124): Total of 8 teeth.
‘AXr small apical tubercle; ‘Gl’ and ‘L2’ strong,
much stronger than rest of teeth on retromargin;
3-4 very small pegs; 5-8 slightly longer, straight
pegs. Dorsal spur long, curved finger 20.8% (20.8-
21.0%) length of carapace; tip distinctly bifur-
cated, either equally or unequally (Fig. 125).
Cheliceral fang long (same length as base, bent
sharply over at both proximal and distal ends).
Length of cephalothorax 2.2 mm (2. 2-3. 2), total
length 6.0 mm (6. 0-7.0) (Fig. 126). Chelicerae
slightly shorter (93%, 92-94%) than length of
carapace. Depression of thoracic fovea distinctly
marked with broken semicircle shape. Leg
spination similar to female, but spines shorter
(Figs. 127-128). Femur I: 6 prolateral, 3 dorsal,
3 retrolateral spines. Tibia 1: 4 prolateral, 2 dor-
sal, 5 retrolateral spines. Metatarsus I: 1 prola-
teral, 1 dorsal, 2 retrolateral spines. Femur III:
no ventral spines. Tibia III: 2 pairs of ventral
spines. Coloration and eye pattern as in female.
Abdomen shape as in female, but median tu-
bercle less pronounced.

Conductor Tip (Figs. 129 and 166): smooth,
evenly rounded, helmet-like cap with tip that
hooks inwards, making it look like shepherd’s
crook.

Allotype female: (Figs. 1 30-136). All eyes rath-
er small (Fig. 1 32). PME separated by approxi-
mately width of PME. Median ocular area wider
at back than at front. Lateral eyes closely con-
tiguous. Cheliceral margins: Promargin (Fig. 1 30):
series of 8 (7) short, thick teeth ‘U2’ and ‘U3’
largest, well separated from ‘Ul’, ‘Ul’ 40% (range
37-50%) height of‘U2’, separated from it by 38%
(20-40%) cheliceral length; teeth decreasing in
size proximally. Retromargin (Fig. 131): series
of 9 slightly smaller, robust teeth ‘L2’ largest,
well separated from ‘LI’; ‘LI’ only 27% (20-
40%) height of ‘L2’, but same size as ‘Ul’; teeth
decreasing in size proximally. Cheliceral fang ap-
proximately 80% length of base, tapering to
smooth point at distal end. Length of cephalo-
thorax 3.0 mm (2. 5-3. 2), total length 7.7 mm
(5. 3-8. 8). Chelicerae quite short, 79% (60-80%)
length of carapace. Legs banded, spines very dis-
tinct, although relatively short (48% length of
carapace, 28-58%) (Figs. 133-134). Femur I: 5
(4) prolateral, 2 (3) dorsal, 5 (4) retrolateral spines.
Tibia I: 5 (4) prolateral, 2 dorsal, 5 (4) retrolateral
spines. Metatarsus I: 1 prolateral, 1 dorsal, 3 (2)

retrolateral spines. Femur III: no ventral spines.
Tibia III: 2 pairs of ventral spines. Depression
of cephalothoracic fovea distinctly marked with
broken semicircle on anterior margin. Cephal-
othoracic pattern distinct “Y” shape. Sternum
dark or dusky. Abdomen distinctly diamond
shaped, often exaggerated laterally, with sub-me-
dial distinct, small black inverted triangle, which
may which may be drawn up into short, finger-
like tubercle (Fig. 135). Color pattern consists of
various combinations of black, brown and grey.
Venter stippled silver with medial pale fawn bar,
narrower posteriorly.

Seminal receptacles (Fig. 136): Two bulbs
linked together in opposing “comma” shapes;
sclerotization rather strong, particularly along
median perimeter of lower bulb and central por-
tion. Both bulbs, in particular dorsal bulb, di-
lated, with central portion enveloped by median
lobe. Median lobe angular, squarish doughnut
shape that projects slightly behind central por-
tion, and projects into area defined approxi-
mately by outer limits of bulbs.

Color polymorphism.— This species exhibits
extraordinary diversity in color patterns: The
amount and location of black and brown patches
and lines vary tremendously; green and/or white
may often be present, and may sometimes even
be dominant colors. However, the median, in-
verted black triangle (drawn up, to a greater or
lesser degree, into a tubercle) is always present.
Similarly, the venter always has a medial pale
tan bar, narrower posteriorly, and the sternum
is always black (sometimes fading to dusky in
alcohol).

Interisland Variation. — I have examined 10
individuals of this species from each of Oahu,
Molokai, Maui and Hawaii, and 5 from Lanai.
There is considerable variation between islands.
But it seems like this variation is continuous,
without any clear-cut demarcations. At present,
therefore, I consider representatives on these is-
lands as populations of the same species. Con-
ductor tips can be compared for representatives
from Oahu (Fig. 166), Lanai (Fig. 167), Maui
(Fig. 168) and Hawaii (Fig. 169). Differences are
summarized in Table 2.

Material Examined.— This species is found in dry,
mesic and wet forest on all islands except Kauai (Table

l ) : Hawaii Island, Hakalau, Mauna Kea, 6 1 50 ft ( 1 876

m) , 12-X-90 (R.G. Gillespie, D.J. Preston & I. Felger).
Kipukas, Mauna Kea, 5800 ft (1770 m), 13-X-90(R.G.
Gillespie, D.J. Preston, J. Lepson & I. Felger); Kipukas,

GILLESPIE- HAWAIIAN TETRAGNATHA

203

Table 2.— Interisland variation in Tetragnatha quasi modo, new species, comparing the leg spines, bifurcation
of the dorsal spur, and the conductor tip of the male palp.

Oahu

Molokai

Lanai

Maui

Hawaii

Leg spines:

Retrolateral

5

4

4

4

5

Dorsal

2

2

2

2

2

Prolateral

5(4)

4

4

4

5(4)

Bifurcation of dorsal spur

Equal

Equal

Unequal

Unequal

Equal

Conductor tip:

Cap slightly curled

X

X

X

—

—

Backward hook present

-

-

-

X

X

Mauna Kea, 5440 ft (1658 m), 12-IIL90 (R.G. Gilles-
pie & J.I.M. Gillespie); Kipuka 8, Mauna Kea, 5240
ft, 25-VII-88 (R.G. Gillespie & C. Parrish); Kipuka 6,
Mauna Kea, 5050 ft, 25-VII-88 (R.G. Gillespie & C.
Parrish); Kipuka, Saddle Road, 2700 ft (823 m), 25-
VII-88 (R.G. Gillespie & C. Parrish); Wailuku River,
3500 ft (1067 m), l-VIII-88 (R.G. Gillespie & C. Par-
rish); Hualalai, 3600 ft (1097 m) 30-VII-88 (R.G. Gil-
lespie & C. Parrish); Kealakekua Ranch, 3740 ft (1 140
m), 3060 ft (933 m), 9-III-90 (R.G. Gillespie & J.I.M.
Gillespie); Puu Makaala, Stainback Highway, 4000 ft
(1220 m), 14-X-90 (R.G. Gillespie, D.J. Preston & 1.
Felger); Puu Makaala, Stainback Highway, 2090 ft (637
m), 3070 ft (936 m), 4010 ft (1222 m), 17-111-90 (R.G.
Gillespie & J.I.M. Gillespie); Puu Makaala, End Wright
Rd., 4300 ft 21-X-90 (R.G. Gillespie & D.J. Preston);
Kipahoehoe 4000 ft ( 1 220 m) 1 6-X-90 (R.G. Gillespie,
D.J. Preston & J. Kiyabu); Halepiula Road, Manuka,
3700 ft (1 128 m), 17-X-90 (R.G. Gillespie, D.J. Pres-
ton &J. Burgett); Laupahoehoe, 4 1 20 ft (1257 m), 3200
ft (976 m), 19-X-90 (R.G. Gillespie, D.J. Preston & J.
Burgett); Laupahoehoe, 2300 ft (700 m), 14-III-90, 3240
ft (988 m), 13-111-90, 4020 ft(1225m), 14-111-90,4210
ft (1283 m), 13-III-90, 6240 ft (1902 m), 15-III-90,
5140 ft (1567 m), 13-III-90 (R.G. Gillespie & J.I.M.
Gillespie); Mauna Loa Strip Rd, 5 5 1 0 ft ( 1 680 m), 3805
ft (1 160 m), 10-III-90 (R.G. Gillespie & J.I.M. Gilles-
pie); Thurston, Volcano, 4000 ft (1220 m), 31-VII-88
(R.G. Gillespie & C. Parrish); Kohala, 3780 ft (1 152
m), 27-VII-88, 28-VII-88 (R.G. Gillespie, W.D. Per-
reira, K.Y. Kaneshiro & C. Parrish). Maui Island, West
Maui, Puu Kukui, 4550 ft (1387 m), 31-V-88, 1-VI-
88 (R.G. Gillespie & C. Parrish). East Maui, Waika-
moi, 4400 ft (1340 m), 8- VI-88 (R.G. Gillespie & C.
Parrish); 8-II-90 (R.G. Gillespie & J. Burgett); Wai-
kamoi Flume, 4400 ft ( 1 340 m), 1 3-VIII-88 (R.G. Gil-
lespie & C. Parrish); Waikamoi, Carruthers Camp, 6 1 50
ft (1876 m), 29-V-88 (R.G. Gillespie & C. Parrish); 5-
11-90 (R.G. Gillespie); Waikamoi, Honomanu Valley,
5200 ft (1585 m), 6-II-90 (R.G. Gillespie); Hanawi,
1520 ft (463 m), 9-II-90 (R.G. Gillespie & R. Rydell);
1 l-V-90 (R.G. Gillespie, R. Rydell & J. Burgett); Bogs,

NE Rift Haleakala, 5500 ft (1676 m), 15-1-88 (R.G.
Gillespie & A.C. Medeiros); Kipahulu Valley, 2000 ft
(610 m), lO-VI-89, 4000 ft (1220 m), 1 l-VI-89 (A.C.
Medeiros); Kipahulu Valley, 2000 ft (610 m), 17-V-
90, 3000 ft (914 m) 16-V-90 (R.G. Gillespie & A.C.
Medeiros). Kipahulu Valley, 3000 ft (914 m), 16-V-
90, 4000 ft (1220 m), 15-V-90, 5000 ft (1524 m), 14-
V-90 (R.G. Gillespie & A.C. Medeiros). Kipahulu Val-
ley, 6500 ft (1980 m), 27-1 V-88 (R.G. Gillespie & A.C.
Medeiros). Pohakuokala, Crater Road, 5000 ft (1524
m), 1 l-V-90 (R.G. Gillespie & J. Burgett). Molokai
Island, Kamakou, 3800 ft (1158 m), 21-23-VI-88, 1-
11-90 (R.G. Gillespie & C. Parrish); Kaunuohua Sum-
mit, Kamakou, 4535 ft (1382 m), 2-II-90 (R.G. Gil-
lespie & J. Halloran); Lanai Island, Lanaihale, 3370
ft (1027 m), 14-VIII-90 (R.G. Gillespie & A.C. Med-
eiros); Oahu Island, Waianae Kai, 1900 ft (580 m),
25-VI-88 (R.G. Gillespie & A.C. Medeiros); Peacock
Rats, 1800 ft (550 m), 18-VIII-88 (R.G. Gillespie &
C. Parrish).

Tetragnatha restricta Simon
(Figs. 137-144 and 164)

Tetragnatha restricta Simon (Simon 1900: 473-474,
pi. XIX, fig. 10). Male holotype from Hawaii, Kona,
in the Museum National d’Histoire Naturelle de Par-
is, examined. Okuma 1988c: 83-84, fig. 7.

Diagnosis. — r. restricta is most easily con-
fused with T. quasimodo. The most useful char-
acters for distinguishing these species are (1) ab-
dominal shape: the flat topped abdomen of T.
restricta is very distinctive, contrasting with the
peaked abdomen of T. quasimodo, and (2) ab-
dominal pattern: males are readily differentiated
on the basis of their cheliceral armature, in par-
ticular the absence of the wave-like first tooth so
characteristic of T. quasimodo.

Male: [holotype described by Simon (1900)
and redescribed by Okuma (1988c)]. Specimen

204

THE JOURNAL OF ARACHNOLOGY

Figures \37-\44. — Tetragnatha restricta Simon.
Male: 137) Promargin of right chelicera: 138) Retro-
margin of left chelicera; 139) Dorsal spur of chelicera,
lateral view; 140) carapace, dorsal; 141) Right leg I,
dorsal; 142) Right leg III, prolateral; 143) Left palpus,
prolateral. Female: 144) Seminal receptacles. Scale bar
(mm) at Fig. 140 applies to Figs. 137-140; at Fig. 142
to Figs. 141, 142.

collected from Laupahoehoe, 3240 ft (988 m),
Hawaii Island (13 March 1990) (coll. J.I.M. Gil-
lespie & R.G. Gillespie): Eyes rather large (Fig.
1 40). PME separated by approximately half width
of PME. Median ocular area slightly narrower at

back than at front. Lateral eyes contiguous. Pro-
margin of chelicerae: ‘sP close to ‘T’, ratio of
distal end to ‘sP: ‘sP to ‘T’: ‘T’ to ‘rsuF 5:2:4
(Fig. 137). ‘Gu’ absent; ‘sP tiny bump, much
smaller than ‘T’ in both width (48%) and height
(16%). ‘T’ robust peak pointing rather sharply
and directly (not curved) upwards (pointing away
from ‘rsuF and towards ‘sP. ‘rsu’ 5 straight spikes.
Retromargin of chelicerae (Fig. 138): Total of 6
teeth. ‘AXP absent; ‘GP strong, much stronger
than rest of teeth on retromargin; 3-4 very small
pegs; 5-8 slightly longer, straight pegs. Dorsal
spur long, curved finger 13% cheliceral length;
tip distinctly and unequally bifurcated (Fig. 1 39).
Cheliceral fang considerably shorter than base.
Length of cephalothorax 1 .7 mm, total length 3.0
mm (Fig. 140). Chelicerae shorter (74%) than
length of carapace. Depression of thoracic fovea
distinctly marked with smoothly rounded in-
verted “U” shape. Coloration as in female. Leg
spination similar to female, but spines shorter
(Figs. 141-142).

Conductor Tip (Figs. 143 and 164): smoothly
rounded, high cap, terminating in distinct apical
projection that curls forwards and upwards.

Female allotype: Eyes similar to male. Cheli-
ceral margins: Promargin: series of 6 medium-
sized teeth ‘U2’ and ‘U3’ largest, well separated
from ‘Ur. ‘Ul’ same height as ‘U2’, separated
from it by 24% cheliceral length; teeth decreasing
in size proximally. Retromargin: series of 7 fairly
large teeth: ‘L 1 ’ much smaller than ‘U 1 ’ and ‘L2’.
Cheliceral fang approximately 84% length of base,
tapering to smooth point at distal end. Length
of cephalothorax 1.8 mm, total length 3.4 mm.
Chelicerae much shorter (53%) than length of
carapace. Legs quite dark, spotted under dorsal
spines on femora and banded on tibia. Leg spi-
nation (measured from another female from Ha-
waii Island since leg spines were absent from the
holotype: female from Laupahoehoe, 3240 ft (988
m), 1 3 March 1 990, coll. J.I.M. Gillespie & R.G.
Gillespie): Spines very distinct, although rela-
tively short (approximately 48% length of cara-
pace). Femur I: 5 prolateral, 3 dorsal, 4 retro-
lateral spines. Tibia I: 4 prolateral, 2 dorsal, 4
retrolateral spines. Metatarsus I: 1 prolateral, 1
dorsal, 2 retrolateral spines. Femur III: no ven-
tral spines. Tibia III: 2 pairs of ventral spines.
Depression of cephalothoracic fovea distinctly
marked with inverted “U” shape on anterior
margin. Carapace dark brown with 4 pairs of pale
lines radiating from fovea. Sternum dark. Ab-
domen pyriform in shape from above, raised up

GILLESPIE- HAWAIIAN TETRAGNATHA

along medial line, so that, when observed from
front, medial portion appears like flat plateau
across abdomen. Color pattern consists of vari-
ous combinations of grey and black, often with
rather dark line running longitudinally down
midline. Lateral lines may diverge near anterior
margin, running out towards lateral margins at
midline.

Seminal receptacles [from Maui representative
of species, Waikamoi Flume, 4400 ft (1340 m),
8 July 1988, coll. R.G. Gillespie & C. Parrish)]
(Fig. 144); Two bulbs linked together in oppos-
ing, rounded “C” shapes; sclerotization rather
strong, particularly along median perimeter of
both lower and upper bulbs and central portion.
Both bulbs slightly dilated. Median lobe irregular
doughnut shape that projects out into area ap-
proximately that defined by outer limits of bulbs.

Color polymorphism. — Little evidence of this.

Interisland Variation. — This species is found
mostly in mesic forest on both Hawaii and East
Maui. The primary difference between these 2
populations is in the tip of the conductor of the
male palp. Both populations have a smoothly
rounded, medium height cap, terminating in a
distinct apical projection that curls forwards and
upwards. However, the length of the apical pro-
jection is almost the same length as the cap on
Maui, whereas it is much smaller on Hawaii. The
only other difference is that ‘T’ and ‘rsuT are
closer in Maui representatives, the ratio; distal
end of chelicerae to ‘si’; ‘si’ to ‘T’; ‘T’ to ‘rsul’
5;3;2 rather than 4;2;4 or 5;2;4. However, in all
other respects, the species appear to differ little
on the two islands. At present, therefore, I con-
sider representatives on these islands as popu-
lations of the same species.

Material Examined. — //a»va/; Island, Hakalau,
Mauna Kea, 6150 ft (1876 m), 12-X-90 (R.G. Gilles-
pie, D.J. Preston & 1. Felger). Kipahoehoe, 4000 ft
(1220 m) 16-X-90 (R.G. Gillespie, D.J. Preston & J.
Kiyabu); Halepiula Road, Manuka, 3700 ft (1 128 m),
17-X-90 (R.G. Gillespie, D.J. Preston & J. Burgett);
Laupahoehoe, 4 1 20 ft, 3200, 19-X-90 (R.G. Gillespie,
D.J. Preston & J. Burgett); Laupahoehoe, 3240 ft (988
m), 13-III-90, 4020 ft (1225 m), 14-III-90 (R.G. Gil-
lespie & J.I.M. Gillespie); Mauna Loa Strip Rd, 6540
ft (1993 m) (R.G. Gillespie & J.I.M. Gillespie). Maui
Island. East Maui, Waikamoi, 4400 ft (1340 m), 8-VI-
88 (R.G. Gillespie & C. Parrish); 8-II-90 (R.G. Gil-
lespie & J. Burgett); Kipahulu Valley, 2000 ft (610 m),
17-V-90, 3000 ft (914 m) 16-V-90 (R.G. Gillespie &
A.C. Medeiros). Pohakuokala, Crater Road, 5000 ft
(1524 m), 1 l-V-90 (R.G. Gillespie & J. Burgett).

205

Figures \A5-\52. — Tetragnatha mohilii, male ho-
lotype. 145) Promargin of right chelicera; 146) Retro-
margin of left chelicera; 147) Dorsal spur of chelicera,
lateral view; 148) carapace, dorsal; 149) Right leg I,
dorsal; 150) Right leg III, prolateral; 151) abdomen,
dorsal; 152) Left palpus, prolateral. Scale bar (mm) at
Fig. 148 applies to Figs. 145-148; at Fig. 150 to Figs.
149, 150.

Tetragnatha mohihi, new species
(Figs. 145-152 and 165)

Types. — Holotype male from the Mohihi
Ditch, 3500 ft (1067 m), Kauai Island (21 March
1990) (coll. R.G. Gillespie and C. Parrish). Fe-
male unknown. Holotype deposited in the Bish-
op Museum, Honolulu.

Etymology.— The specific epithet, regarded as
a noun in apposition, refers to the type locality

206

THE JOURNAL OF ARACHNOLOGY

Figures 1 53-1 64. -Scanning electron micrographs of conductor tips of male palps (scale on each x 400): 153)
T. tantalus- 154) T. polycliromata-, 155) T. brevignatha (Hawaii); 156) T. brevignatha (Maui); 157) T. macra-
cantha', 158) T. waikainob 159) T. kainakou (Molokai); 160) T. kamakou (Maui); 161) T. kauaiensis', 162) T.
peneirak 163) T. pilosa-, 164)7'. restricta.

GILLESPIE- HAWAIIAN TETRAGNATHA

207

Figures 165-169.— Scanning electron micrographs of conductor tips of male palps (scale on each x 400); 165)
T. mohihi\ 166) T. quasi modo {Oahu)-, 167) T. quasimodo {hana\)\ 168) T. quasimodo (Mam)', 169) T. qiiasimodo
(Hawaii).

of the species, Mohihi Ditch, just beyond the end
of Camp 10 Road, on the flanks of Mount Wai-
aleale.

Diagnosis. — r. mohihi has many unique fea-
tures, and is unlikely to be confused with any
other species. The only potential candidates for
confusion would be T. pilosa and more melanic
form of T. kauaiensis. T. mohihi can be distin-
guished from either of these because of: (1) its
short chelicerae with a long dorsal spur [T. pilosa
has short chelicerae and a short dorsal spur, T.
kauaiensis has long chelicerae and long dorsal
spur], (2) distinctive abdominal pattern, and (3)
leg spination [T. mohihi has 4 retrolateral, 2 dor-
sal, 3(4) prolateral spines; T. pilosa has 6 (5)
retrolateral, 2 dorsal, 6 (5) prolateral and T.
kauaiensis has 5 retrolateral, 2 dorsal, 5 prola-
teral spines]. In males, the tip of the conductor
is very characteristic: the cap is much higher than
either T. pilosa or T. kauaiensis, and also has a
long apical projection, which is lacking in the
other two species.

Description.— //o/o/ype’ ma/e’.'(Figs. 145-1 52).
Promargin of chelicerae (Fig. 145): ‘si’ and ‘T’
very close together, and distance between distal

end and ‘sP approximately equal to distance be-
tween ‘sP and ‘rsul’, ratio of distal end to ‘sP:
‘sP to ‘T’: ‘T’ to ‘rsuT 5:1:4 (sometimes 4:1:5).
‘Gu’ present, small and inconspicuous; ‘sP dis-
tinct peak, directed perpendicular to cheliceral
margin, smaller than ‘T’ in both width (63%, 63-
67%) and height (43%, 43-50%). ‘sP close to ‘T’
(separated by 8.5% cheliceral length). ‘T’ large
(13.5% cheliceral length), robust peak directed
perpendicular to margin of chelicerae. ‘rsu’ 4 (3-
4) straight spikes. Retromargin of chelicerae (Fig.
1 46): Short series of 7 rather large spikes, well
separated ‘L3’ and ‘L7’ smallest. Dorsal spur long,
curved finger (13.9% length of carapace); tip not
bifurcated, although dorsal side projects slightly
further forward than ventral (Fig. 147). Cheli-
ceral fang approximately same length as base,
bent sharply over at proximal, and slightly at
distal, ends. Length of cephalothorax 1.7 mm
(1.7-1. 8), total length 4.5 mm (Fig. 148). Che-
licerae much shorter (66%, 66-68%) than length
of carapace. Eyes fairly large, PME separated by
approximately width of PME (Fig. 148). Median
ocular area wider at back than at front. Lateral
eyes contiguous. Legs pigmented under promi-

208

THE JOURNAL OF ARACHNOLOGY

nent spines (Figs. 149-150). Femur I: 5 prola-
teral, 3 dorsal, no retrolateral spines. Tibia I: 3
(4) prolateral, 2 dorsal, 4 retrolateral spines.
Metatarsus I: 1 prolateral, 1 dorsal, 2 retrolateral
spines. Femur III: no ventral spines. Tibia III: 2
pairs of ventral spines. Carapace with pair of
prominent lines in anterior region, converging
towards midline. Thoracic fovea marked with
circular indentation, tapering to thin line running
posteriorly. Border of carapace pigmented. Ab-
domen with pigmented margins, running slightly
in at midline, and turning into broken pair of
markings running posteriorly beyond midline
(Fig. 151).

Conductor Tip (Figs. 152 and 165): high,
rounded cap, terminating in long apical projec-
tion drawn laterally outwards and downwards
and terminating in very small forward curl.

Material Examined. — This species is found in mesic
forest only on Kauai Island (Table 1): Mohihi Ditch.
3500 ft (1067 m), 27-III-90 (R.G. Gillespie & C. Par-
rish).

ACKNOWLEDGEMENTS

This study was supported by grants from the
Hawaii Bishop Research Institute, the Hawaii
Natural Area Reserves System and the Nature
Conservancy of Hawaii. Additional support was
provided by the Bishop Museum, the Nature
Conservancy of Hawaii, Haleakala and Hawaii
Volcanoes National Parks, the Hawaii Branch of
the U.S. Fish and Wildlife Service and the Zo-
ology Department, U.H. Manoa. Helicopter sup-
port was provided by Haleakala National Park,
Maui Land and Pineapple Company and the Pa-
cific Tropical Botanical Gardens. I am deeply
indebted to the following for their assistance in
collecting specimens: Randy Bartlett, Jeff Bur-
gett, Hampton Carson, Ingrid Felger, Janet Gil-
lespie, John Halloran, Jim Jacobi, Kenneth Ka-
neshiro. Bob Lee, Lloyd Loope, David Lorence,
Tod Lum, Art Medeiros, Steve Montgomery,
Chris Parrish, Steve Perlman, Bill Perreira, Da-
vid Preston, Vince and Barbara Roth, Rob Ry-
dell. Bill Stormont, John Strazanac and Mark
White. Lee Golf allowed me to use his compound
microscope with camera lucida and Kenneth Ka-
neshiro his environmentally controlled facilities
to maintain and rear live specimens. I am also
grateful to the following landowners and prop-
erty managers who facilitated access to forest on
their property: Monty Richardson (Puu O Umi
and Kohala Forest), Jim Kiyabu (Kipahoehoe),

Sally Rice (Manuka), Sam Kuboto (Kealakekua),
Harry Yamamoto (Castle and Cook, Lanai) and
Maui Land and Pineapple Company (West Maui).
Thanks also to Sue Monden for making my
sketches look attractive, and to Marilyn Dunlap
and Tina Carvalho for help with the SEM. Also
to Henrietta Croom, Frank Howarth and Ste-
phen Palumbi for advice and discussion, and to
Jonathan Coddington, Gustavo Hormiga, Herb
Levi, Gary Miller, Norman Platnick and George
Roderick for careful and meticulous reviews of
the first draft.

LITERATURE CITED

Berger, A. J. 1981. Hawaiian Birdlife. University of
Hawaii Press, Honolulu.

Chickering, A. M. 1957. The genus Tetragnatha
(Araneae: Argiopidae) in Panama. Bull. Mus. Comp.
Zool., 116:302-354.

Croom, H. B., R.G. Gillespie, & S.R. Palumbi. 1991.
Mitochondrial DNA sequences coding for a portion
of the RNA of the small ribosomal subunits of Te-
tmgnatha mandibulata and Tetragnatha hawaiensis
(Araneae, Tetragnathidae). J. Arachnol., 19:21 0-2 1 4.
Cooke, C., Montague, J. & Kondo, Y. 1960. Revision
of Tornatellinidae and Achatinellidae (Gastropoda,
Pulmonata). B.P. Bishop Mus. Bull., 221:1-303.
Dabrowska Prot, E. & J. Luczak. 1968a .Spiders and
mosquitos of the ecotone alder forest (Carici elon-
gatae-alnetum) and oak pine forest {Pino querce-
tum). Ekologia Polska Seria A., XVI:46 1-483.
Dabrowska Prot, E. & J. Luczak. 1968b. Studies on
the incidence of mosquitos in the food of Tetrag-
natha montana Simon and its food activity in the
natural habitat. Ekologia Polska Seria A., XVI:843-
853.

Dabrowska Prot, E., Luczak, J. & Tarwid, K. 1968.
Prey and predator density and their reactions in the
process of mosquito reduction by spiders in field
experiments. Ekologia Polska Seria A., XVI:773-
819.

Freed, L. A., Conant, S. and Fleischer, R. C. 1987.
Evolutionary ecology and radiation of Hawaiian
passerine birds. Trends in Ecol. and Evol., 2:196-
203.

Gillespie, R. G. 1986. Between population compar-
ison of resource acquisition in the long jawed orb
weaving spider Tetragnatha elongata. Ph.D. disser-
tation; University of Tennessee, Knoxville, Ten-
nessee.

Gillespie, R. G. 1987a. The mechanism of habitat
selection in the long jawed orb weaving spider Te-
tragnatha elongata (Araneae, Tetragnathidae). J.
Arachnol., 15:81-90.

Gillespie, R.G. 1987b. The role ofprey in aggregative
behaviour of the long jawed orb weaving spider Te-
tragnatha elongata. Anim. Behav., 35:675-681.

GILLESPIE- HAWAIIAN TETRAGNATHA

209

Gillespie, R.G. 1989. Diet induced color change in
the Hawaiian happy face Spider Theridion gmllator
(Araneae, Theridiidae). J. ArachnoL, 17:171-177.

Gillespie, R. G. 1990. Costs and benefits of brood
care in the Hawaiian happy face s^iderTheridion
gmllator (Araneae, Theridiidae). American. Mid.
Nat, 123:236-243.

Gillespie, R. G. & B. Tabashnik. 1989. What makes
a happy face? Determinants of colour pattern in the
Hawaiian happy face spider Theridion grallat or. He-
redity, 62:355-363.

Gillespie, R. G. & B. Tabashnik. 1990. Maintaining
a happy face: Stable colour polymorphism in the
spider Theridion grallator (Araneae, Theridiidae).
Heredity, 65:67-74.

Heliker, C. 1989. The volcanic origin of the Hawaiian
Islands. Pp. 1 1-16, In Conservation Biology in Ha-
waii. Univ. of Hawaii Cooperative National Park
Resources Studies Unit, Honolulu.

Gon, S. M. III. 1985. Comparative behavioral ecol-
ogy of the spider Theridion grallator Simon (Ara-
neae: Theridiidae) in the Hawaiian archipelago.
Ph.D. dissertation; U. California at Davis. 322 pp.

Kaneshiro, K. Y. & Boake, C. R. B. 1987. Sexual
selection and speciation: issues raised by Hawaiian
drosophilids. Trends in Ecol. and EvoL, 2:207-2 1 1 .

Karsch, F. 1880. Sitzungs-Berichte der Gesellschaft
Naturforschender freunde zu Berlin. Jahrgang. Sit-
zung vom, 18:76-84.

Kaston, B. J. 1 948. How to know the spiders. 3rd ed.
Wm. C. Brown Co., Dubuque, Iowa.

Levi, H. W. 1981. The American orb-weaver genus
Dolichognatha and Tetragnatha north of Mexico
(Araneae: Araneidae, Tetragnathinae). Bull. Mus.
Comp. Zool. Harvard., 149:271-318.

Luczak, J. & Dabrowska Prot, E. 1 966. Experimental
studies on the reduction of the abundance of mos-
quitos by spiders. Bull. Acad. Polonaise Sci., Cl. 2:
14:315-320.

Okuma, C. 1987. A revision of the Australasian spe-
cies of the genus Tetragnatha (Araneae, Tetrag-
nathidae). Esakia, 25:37-96.

Okuma, C. 1988a. A revision of the genus Tetrag-
natha Latrielle (Araneae, Tetragnathidae) of Asia,
Part I., J. Fac. Agr. Kyushu Univ., 32:165-181.

Okuma, C. 1988b. A revision of the genus Tetrag-
natha Latrielle (Araneae, Tetragnathidae) of Asia,
Part I!., J. Fac. Agr. Kyushu Univ., 32:183-213.

Okuma, C. 1988c. Redescriptions of the Hawaiian
spiders of Tetragnatha described by Simon (Ara-
neae, Tetragnathidae). J. Fac. Agr. Kyushu Univ.,
33:77-86.

Perkins, R. C. L. 1913. Introduction (to Fauna Ha-
waiiensis). /«: Fauna Hawaiiensis, Vol. 1: xv-
ccxxviii. Ed. D. Sharp. Cambridge Univ. Press,
Cambridge.

Simon, E. 1900. Arachnida: Fauna Hawaiiensis, 2:
443-519, pis. 15-19.

Suman, T. W. 1964. Spiders of the Hawaiian Islands:
Catalog and Bibliography. Pacific Insects, 6:665-
687.

Suman, T. W. 1970. Spiders of the family Thomis-
idae in Hawaii. Pacific Insects, 12:773-864.

Wiehle, H. 1963. Tetragnathidae. Tierwelt Deutsch-
lands., 49: 1-76.

Yoshida, M. 1987. Predatory behavior of

tha praedonia (Araneae: Tetragnathidae). Acta Ar-
achnologica, 35:57-75.

Manuscript received March 1991, revised July 1991

1991. The Journal of Arachnology 19:210-214

MITOCHONDRIAL DNA SEQUENCES CODING EOR A
PORTION OF THE RNA OF THE SMALL RIBOSOMAL
SUBUNITS OF TETRAGNATHA MANDIBULATA AND
TETRAGNATHA HAWAIENSIS {ARANEAE, TETRAGNATHIDAE)

Henrietta B. Croom: Department of Biology, The University of the South, Sewanee,
Tennessee 37375 USA

Rosemary G. Gillespie and Stephen R. Palumbi: Department of Zoology, The University
of Hawaii at Manoa, Honolulu, Hawaii 96822 USA

ABSTRACT. A region of mitochondrial DNA coding for most of the third domain of the 12S rRNA of the
ribosomal small subunit has been sequenced from two spiders in the genus Tetragnatha (Araneae, Tetrag-
nathidae): a circumtropical species T. mandibulata and an endemic Hawaiian species T. hawaiensis. The
secondary structure of the spider ribosomal RNA shows strong similarity to that of insects. Across this region,
the two Tetragnatha sequences are 22% different. The T. mandibulata sequence is 36% different from the
homologous segment in Drosophila yakiiba and 51% different from the same segment in Homo sapiens. The

spider sequences are sufficiently variable to be useful in
the species in this genus.

A powerful approach to studying genetic re-
latedness of species involves DNA sequence
comparisons which can be used to estimate
branching order of phylogenetic trees as well as
evolutionary distance between extant taxa (Fel-
senstein 1988). Recent application of the poly-
merase chain reaction and direct sequencing has
accelerated efforts to examine a wide range of
taxa with DNA comparisons (Kocher et al. 1 989;
Martin et al. 1 990). To date the use of this meth-
od in studying spiders has not been reported.
Here we describe a procedure we have used to
amplify and to determine the sequence of nucle-
otidesofa 279-base-pair region of mitochondrial
DNA (mtDNA) from two species of the genus
Tetragnatha (Araneae, Tetragnathidae).

Sequencing DNA has the following advantages
over other techniques of genetic comparison; (1)
it has greater resolving power over a hierarchical
range of intraspecific to intergeneric compari-
sons, (2) sequences are easily compared with
known sequences from other species, and (3)
functional information on products encoded by
DNA allows strong inferences on the selective
importance of mutations observed, allowing
character weighting for sites that are not selec-
tively neutral (Kocher et al. 1989). Mitochon-
drial DNA was chosen for this work because its
matriarchal inheritance and lack of recombina-

studying genetic relationships among at least some of

tion make it often more instructive than nuclear
DNA in comparing taxa (Wilson et al. 1985).
This DNA evolves rapidly at the sequence level
in arthropods (DeSalle and Templeton 1988;
Palumbi and Benzie 1991) and has proven use-
ful for comparing recently-evolved taxa in Ha-
waii (DeSalle et al. 1987).

In order to determine nucleotide sequence in
a particular mtDNA segment, the segment must
first be amplified either by genetic cloning or by
using the polymerase chain reaction (Mullis and
Faloona 1987;Saikietal. 1988). The latter meth-
od depends upon knowledge of oligonucleotide
sequences that flank the segment of interest and
that serve as primers for enzymatic amplifica-
tion. Kocher et al. (1989) have described uni-
versal (highly-conserved) oligonucleotide se-
quences flanking a 300-base portion of the third
domain of the 1 2S ribosomal RNA gene that can
be used to amplify mtDNA from animals as di-
verse as humans and invertebrates. The conser-
vation of these primers makes them useful to
investigators sequencing the DNA of species for
which there is no previous sequence information
(Simon et al. 1990). We first used insect-specific
primers for the 12S rRNA region, slightly mod-
ified by C. Simon (pers. comm.) from the original
primers of Kocher et al. (1989), to amplify and
sequence the DNA of Tetragnatha mandibulata

210

CROOM ET Ah.-TETRAGNATHA MTDNA SEQUENCES

211

Walckenaer. We then designed a spider-specific
primer based on this sequence to amplify and
sequence the same region from a Hawaiian en-
demic species T.hawaiensis Simon.

MATERIALS AND METHODS

All solutions used were either sterilized or pre-
pared using sterile deionized, distilled water, and
all glass- and plastic-ware were sterile with the
exception of the Centricon tubes (see below). The
chelicerae and a front leg of each spider were
placed in 70% ethanol as voucher specimens.
Total (genomic) DNA was prepared by homog-
enizing a single spider in a 1.5 mL Eppendorf
tube in 200 mL of 25 mM Tris HCl (pH 7.5), 100
mM EDTA, 2% SDS, and 200 ntg/mL Proteinase
K, followed by incubation for 1-2 hr in a water
bath at 65 °C. The homogenate was extracted
first with phenol, previously-equilibrated with 1
M Tris HCl buffer (pH 7.5), then with 25:24:1
phenol/chloroform/isoamyl alcohol 1 to 4 times
(until all the protein-containing white interface
was removed), and finally with 24: 1 chloroform/
isoamyl alcohol. One-half volume of a 7.5 M
solution of ammonium acetate was added to the
extract to achieve a final concentration of 2.5 M
and mixed well before adding 2‘/2 volumes of
95% ethyl alcohol and mixing again. This solu-
tion was incubated at room temperature for 1 5
min to allow for DNA precipitation. The DNA
was pelleted by centrifugation at 1 4,000 x g for
1 5 min at room temperature, washed with 500
mL of 70% ethyl alcohol, dried under a vacuum,
and resuspended in 25 mL water. Five ixL of this
preparation was electrophoresed on a 0.8% aga-
rose gel in Tris-borate buffer and stained with
ethidium bromide as described in Sambrook et
al. (1989) to ensure that high molecular-weight
DNA was present. The preparation is stable in-
definitely when stored at -20°C.

One /iL of a 1:10 dilution in water of this DNA
and 2.5 units of DNA polymerase from Themus
aquaticus (Perkin-Elmer Cetus) were incubated
in 100 fxL of buffer containing 67 mM Tris HCl
(pH 8.8), 3 mM magnesium chloride, 16.7 mM
ammonium sulfate, each of four deoxynucleo-
side triphosphates at 200 mM , and each of two
primers at 1 according to the protocol of
Saiki et al. (1988). Thermal profile for 45 cycles
was as follows: (1) DNA melting for 1 min at 94
°C, (2) annealing for 1 min at 50 °C, and (3)
polymerization for 2 min at 72 °C to amplify the
double-stranded sequence. The primers and their

location in the Drosophila yakuba mtDNA se-
quence of Clary and Wolstenholme ( 1 985), were
12St-L (14503), a Tetragnatha-specific primer
designed by H. Croom: 5’-GGTGGCATTT-
TATTTTATTAGAGG-3’ and 12Sbi-H (14214),
an insect-specific primer designed by C. Simon:
5’-AAGAGCGACGGGCGATGTGT-3’. One
mL of the double-stranded product of the first
amplification was cycled under the same con-
ditions as above except that only one primer was
added to the incubation mixture. This produced
a single-strand sequencing template, which was
processed in a Centricon 30 microconcentrator
(Amicon) to concentrate and purify the DNA
product. The template was then sequenced by
the dideoxy chain termination method of Sanger
et al. (1977) as described by Engelke et al. ( 1 988),
using the primer that had not been used in the
second amplification. DNA from three individ-
uals of each species was sequenced in both di-
rections, and no intraspecific variation was ob-
served.

Homologous sequences of DNA from different
taxa were aligned to minimize deletions or ad-
ditions using software written by S. R. Palumbi
and C. Parrish. Pairwise percent differences were
calculated by counting only sites where both spe-
cies have nucleotides in our aligned sequences.
One strand of spider DNA was folded to show
its secondary structure using the folded sequenc-
es of insects as a guide (Clary and Wolstenholme
1987; Simon et al. 1990).

RESULTS AND DISCUSSION

The DNA sequences from the two Tetragna-
tha species were compared with the 12S rRNA
genes from both Homo sapiens and Drosophila
yakuba (Fig. 1). The two Tetragnatha sequences
are 22% different from each other, 36% different
from the homologous segment in Drosophila
(Clary and Wolstenholme 1985), and 51% dif-
ferent from the same segment in Homo (Ander-
son et al. 1981). In comparison, the Drosophila
and Homo segments differ by 45%. These values
are based solely on pairwise nucleotide differ-
ences, ignoring insertions and deletions, using
the alignments in Fig. 1. It is difficult to align
nonconserved regions of DNA from distantly-
related groups, so other alignments may yield
slightly different percentages. Likewise, the dif-
ferenees here are uncorrected for multiple mu-
tations at the same site, which has the effect of

212

THE JOURNAL OF ARACHNOLOGY

Tetragnatha mandibulata aacatgtttattaatcgacattacacgattatttt

Tetragnatha hawaiensis .. t ....... a atc .......... c .. .

Drosophila yakuba ... c ..... tg ....... t . a . c ..... . ggacc .

Homo sapiens . g . c . . . . ctg t . aac . c . . . . c . acc .

TACTTTTTTATAA ATTT TAT AT ACCTCCGTCC - - AGAATAAATTTT T AATA TTT

C AATA . T G .... TT ..... AAAA ..........

AAA . T . GTAATC . G .......... GT .. . TATC . . . . . . TT . . A . A . GA . TAA . AA

C. .CACC.CT.GC. .TC.GCC. G. .A. . TTC . . C . A . CCC . GA . G . AGGCTACA

TATTCAAAATAATATTATAATAATTTA GGTAAAGGTGTAGACTTTAAATTAGT -TT

AT . . C . . GA . . . . . . A . . . A . . T . AG . . A . . A

T.-.T.T.A....A...... TATC A .A.C. CT.A. .TT.A. . . .AA

A . G . A . GCGC . . GTACCC . CGT . AAG . CGTTA . . . C C . CA . G . GG . G . CAAG

AAATGTGTTACATTAAAAATTATTT AAGAATTATTTTTTATAA CAATATATGA

...... A ... AA ........... G ... . AAC . AA . C . T . . . . AGT . T . . A . .

T .... G ...... A .... -T ... ACGGATAAAA . . . . G . AA . . A . . . T . T . . .

G.C. . . . . . TTCT . CCCCAGAAAACTACGA . . GCCC . . . .G.AACTT. . GGGTC . .

AAGAGGATTTATAAGCTACTTTTTAATTAAAATTTTAACTTGAATTAAAAA--TAAATGCG

.. A .............. . TAA . . . T A ... T ...... . TA

. G . T GGT . . TA . AA . . . TAA. G . T . AA ... T ....... TT . GCTC .... ATAT

. G . T GC . . TA . AC . A- AG .G. .G.G.GC.T.G. . . . .-C. GGGCCC . G . . GCGC

Figure 1.— Comparison of Mitochondrial 12S Ribosomal DNA from Tetragnatha mandibulata, Tetragnatha
hawaiensis. Homo sapiens, and Drosophila yakuba. Dots represent positions that are identical to the top sequence,
and dashes represent gaps in the sequences required to maximize alignment. The sequence from the Drosophila
is that of Clary and Wolstenholme (1985) between their base pairs 14236 and 14502. The Homo DNA sequence
is that of Anderson et al. (1981) between their bases 1201 and 1475.

making the more distantly-related taxa appear
deceptively similar.

When using the polymerase chain reaction with
genomic DNA, one must always consider the
possibility that nontarget nuclear or mitochon-
drial DNA has been amplified. Using the method
of Palumbi and Wilson (1990), we separated
mtDNA from nuclear DNA of 25 specimens of
T. mandibulata on a cesium chloride gradient
before amplifying and sequencing the mtDNA
fraction. The sequence obtained was identical to
that in Fig. 1. In order to verify that the spider
sequences code for the third domain of 1 2S rRNA,
a single strand was folded to generate the sec-
ondary structure stabilized by hydrogen-bonding
between complementary bases. In all of the taxa
studied to date (Dams et al. 1988; Simon et al.
1990), the folded structure of this domain forms
helical paired stems and unpaired loops. The
structure obtained from T. mandibulata (Fig. 2)
is essentially the same as those of the other known
taxa. Conservation of secondary structure, de-
spite the large overall sequence differences among
taxa (Fig. 1), suggests we have sequenced a func-

tional ribosomal gene. The third domain of the
small rRNA encoded by nuclear DNA is both
larger than, and has a structure distinct from,
that represented in Fig. 2 (Woese et al. 1983;
Dams et al. 1988).

We have sequenced most of the homologous
region from 1 9 other spiders: Aphonopelma chal-
codes Chamberlin (Araneae, Theraphosidae),
Doryonychus raptor Simon (Araneae, Tetrag-
nathidae), and 17 endemic Hawaiian Tetrag-
natha taxa. In the case of the A. chalcodes, cesium
chloride gradient purified mtDNA was used for
the amplification instead of genomic DNA. We
found each of these sequences to be more similar
to the spider sequences in Fig. 1 than to those of
any other known taxa (Croom and Palumbi, un-
published). Such similarity suggests that neither
of the sequences reported in this paper is from
contaminating DNA.

Interestingly, 83% of all bases in the two Te-
tragnatha sequences are either A or T. In the
strand shown in Fig. 2, the frequencies for bases
are: 39% A, 43% T, 8% C, 10% G. The percent
AT across this region is 79% for Drosophila and

CROOM ET AL.-TETRAGNATHA MTDNA SEQUENCES

213

C C

G - C
C - G

Figure 2. — Mitochondrial DNA of Tetragnatha
mandibulata folded to show the secondary structure of
the third domain of 12S ribosomal RNA for which it
codes. Dashes represent hydrogen bonds between A
and T or C and G, and dots represent the weaker hy-
drogen bonds between T and G. The portion of the
sequence between the asterisks * is that of the primer
12St-L. Folding was based on the structure of Simon
et al. (1990).

53% for Homo. This is consistent with the ob-
servation that all known arthropods have high
AT content in this domain (Simon 1991).

The third domain of rRNA is highly conserved
across many taxa (Kocher et al. 1989). Hence,
the large difference (22%) between these two Te-
tragnatha species is surprising but not without
precedent. Palumbi and Benzie (1991) have found
similar percent diversity in the homologous
mtDNA region among species of shrimp in the
genus Penaeus. In addition, we have found 3-
1 3% variation in the homologous DNA from 1 8
different endemic Hawaiian tetragnathids that
appear (on morphological grounds) to have been
derived from a single introduction to the islands
(Gillespie, unpublished). Such high diversities
imply that these species have either diverged for
a long period or that their sequences have di-
verged at a rapid rate. We are currently using
these sequences, as well as those coding for mi-
tochondrial proteins, to conduct systematic anal-
ysis of this group which has undergone explosive
radiation (Gillespie, in press) in the Hawaiian
archipelago.

ACKNOWLEDGEMENTS

We thank Bailey Kessing, Chris Parrish, Chris-
tine Simon, and Rob deSalle for their help and
encouragement during this study. Lei-Anna
Willman provided excellent technical assistance.
This research was supported by NSF grant BSR-
8604969 and by the Faculty Research and Fac-
ulty Development Funds of The University of
the South.

LITERATURE CITED

Anderson, S., A. Bankier, B. G. Barrell, M. H. L.
deBruijn, A. R. Coulson, J. Drouin, I. C. Eperon,
D. P. Nierlich, B. A. Roe, F. Sanger, P. H. Schreier,
A. J. H. Smith, K. R. Stader & I. G. Young. 1981.
Sequence and organization of the human mito-
chondrial genome. Nature, 290:457^65.

Clary, D. O. & D. R. Wolstenholme. 1985. The mi-
tochondrial DNA molecule of Drosophila yakuba:
nucleotide sequence, gene organization, and genetic
code. J. Mol. Evol., 22:252-271.

Clary, D. O. & D. R. Wolstenholme. 1987. Dro-
sophila mitochondrial DNA: conserved sequence in
the A-l-T rich region and supporting evidence for a
secondary structure model of the small ribosomal
RNA. J. Mol. Evol., 25:1 16-125.

Dams, E., L. Hendriks, Y. Van de Peer, J-M. Neefs,
G. Smits & G. Vandenkempt. 1988. Compilation
of small ribosomal subunit RNA sequences. Nucleic
Acids Res. 16 supl.:r87-rl73.

DeSalle, R., T. Freedman, E. M. Prager, & A. C. Wil-
son. 1987. Tempo and mode of sequence evolu-
tion in mitochondrial DNA of Hawaiian Drosoph-
ila. J. Mol. Evol., 26:157-164.

DeSalle, R.& A. R. Templeton. 1988. Founder effects
and the rate of mitochondrial DNA evolution in
Hawaiian Drosophila. Evolution, 42:1076-1084.
Engelke, D. R., P. A. Hoener & F. D. Collins. 1988.
Direct sequencing of enzymatically-amplified hu-
man genomic DNA. Proc. Natl. Acad. Sci. USA,
85:544-548.

Felsenstein, J. 1988. Phylogenies from molecular se-
quences: inference and reliability. Ann. Rev. Gen.,
22:521-565.

Gillespie, R. G. 1991. Hawaiian spiders of the genus
Tetragnatha: I. Spiny leg clade. J. Arachnol., 19:
174-209.

Kocher, T. D., W. K. Thomas, A. Meyer, S. V. Ed-
wards, S. Paabo, F. X. Villablanca & A. C. Wilson.
1989. Dynamics of mitochondrial DNA evolution
in animals: Amplification and sequencing with con-
served primers. Proc. Natl. Acad. Sci. USA, 86:6 196-
6200.

Martin, A. P., B. D. Kessing, & S. R. Palumbi. 1990.
The accuracy of estimating genetic distance between
species from short sequences of mitochondrial DNA.
Mol. Biol. Evol., 7:485-488.

214

THE JOURNAL OF ARACHNOLOGY

Mullis, K. B. & F. A. Faloona. 1987. Specific syn-
thesis of DNA in vitro via a polymerase catalyzed
chain reaction. Methods in Enzymology, 155:335-
350.

Palumbi, S. R. & J. Benzie, in press. Large mito-
chondrial DNA differences between morphologi-
cally similar penaeid shrimp. Mol. Marine Biol. Bio-
technol.

Palumbi, S. R. & A. C. Wilson. 1990. Mitochondrial
DNA diversity in the sea urchins Strongylocentrotus
purpuratus and S. dwebachiensis. Evolution, 44:403-
415.

Saiki, R., D. Gelfand, S. Stoffel, S. Scharf, R. Higuchi,
G. Horn, K. Mullis & H. A. Erlich. 1988. Primer-
directed enzymatic amplification of DNA with a
thermostable DNA polymerase. Science, 239:489-
491.

Sambrook, J., T. Fritsch, & T. Maniatis. 1989. Aga-
rose gel electrophoresis. Pp. 6.2-6.18, In Molecular
Cloning, Second edition, Vol. 1 . Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, New York.

Sanger, F., S. Nicklen, & A. R. Coulsen. 1977. DNA
sequencing with chain-terminating inhibitors. Proc.
Natl. Acad. Sci. USA, 74:5463-5467.

Simon, C. in press. Molecular systematics at the spe-

cies boundary: Exploiting conserved and variable
regions of the mitochondrial genome of animals via
direct sequencing from amplified DNA. In: Molec-
ular Taxonomy, NATO Advanced Studies Institute.
(Godfrey M. Hewitt., ed.). Springer Verlag, Berlin.

Simon, C., S. Paabo, T. D., Kocher, & A. C. Wilson.
1990. Evolution of mitochondrial ribosomal RNA
in insects as shown by the polymerase chain reac-
tion. Pp. 235-244, In: Molecular Evolution,
U.C.L.A. Symposia on Molecular and Cellular Bi-
ology, New Series, Vol. 122. (M. Clegg & S. Clark,
eds.). Alan R. Liss, Inc., New York.

Wilson, A.C., R. L. Cann, S. M. Carr, M. George, U.
B. Gyllensten, K. M. Helm-Bychowski, R. G. Hig-
uchi, S. R. Palumbi, E. M. Prager, R. D. Sage & M.
Stoneking. 1985. Mitochondrial DNA and two
perspectives on evolutionary genetics. Biol. J. Linn.
Soc., 26:375-400.

Woese, C. R., R. Gutell, R. Gupta & H. F. Noller.
1983. Detailed analysis of the higher-order struc-
ture of 16S-like ribosomal ribonucleic acids. Mi-
crobiol. Revs., 47:621-669.

Manuscript received November 1990. revised July 1991 .

1991. The Journal of Arachnology 19:215-224

SEGMENTAL ANOMALIES IN RONCUS AFF. LUBRICUS
(NEOBISIIDAE, PSEUDOSCORPIONES) FROM YUGOSLAVIA

B. P. M. Curcic, R. N. Dimitrijevic, O. S. Karamata, and L. R. Lucic: Institute of

Zoology, Faculty of Science, University of Belgrade, 16, Studentski Trg, YU-11000
Beograd, Yugoslavia.

ABSTRACT. Malformations in the abdominal segmentation patterns were studied in two pseudoscorpion
species of the genus Roncus L. Koch, inhabiting Yugoslavia. A total of 36 abnormal examples were found out
of 4,825 specimens examined. All anomalous pseudoscorpions were dissected and subjected to the pathomor-
phological analysis. The frequency of the aberrant specimens was variable, depending on the locality, growth
stage, sex, and species. The following malformations were noted: hemimery; partial atrophy (single and multiple);
symphysomery (single and multiple); and combinations of different anomalies (combined hemimery and sclerite
enlargement; combined hemimery and symphysomery; combined partial atrophy and symphysomery; combined
partial atrophy and sclerite enlargement; combined atrophy or hemimery, symphysomery and sclerite enlarge-
ment; combined atrophy, symphysomery, helicomery and sclerite enlargement; and combined atrophy, heli-
comery and sclerite enlargement). Teratological variation of the abdominal sclerites has been confined mostly
to adults and, to a lesser degree, to tritonymphs. In addition, some specific features of the relative distribution
of various segmental deficiencies are considered. Finally, the probable causes of the genesis and development
of segmental anomalies in the pseudoscorpions studied have been also discussed.

Developmental anomalies are known to occur

in pseudoscorpions, the most common being the
malformations of various abdominal sclerites
(Legg and Jones 1988; Curcic 1989a, b). Aber-
rations may be produced internally or externally,
the latter being induced mechanically, chemi-
cally or physically. Internally induced aberra-
tions occur during development (including the
molting period). These malformations were first
noted by With (1905), Kastner (1927) and Hadzi
(1937), and later classified by Gilbert (1952) and
Pedder ( 1 965). More recently, Curcic (1980,1 988,
1989a, b), Curcic and Dimitrijevic (1982, 1984,
1986), Curcic et al. (1981, 1983), and Dimitri-
jevic (1985, 1990) have provided further ex-
amples and attempted to quantify the phenom-
enon.

In the majority of the species studied, the de-
formities of the abdominal sclerites were con-
fined to the tritonymph/adult, or maturation
molt. Aberrations were thought to represent post-
embryonic molt phenomena (Pedder 1965) but
their origin and development are still not suffi-
ciently understood at this moment. In Neobisium
carpaticum Beier, Curcic (1989a, b) and Curcic
and Dimitrijevic ( 1 982) found a teratological in-
cidence of 1.4% out of 1,300 specimens exam-
ined; and in N. sylvaticum (C. L. Koch) the in-
cidence was 2%. The majority of deficiencies

occurred at the maturation molt and were con-
fined mostly to males.

Teratological deformities, although not entire-
ly restricted to individuals undergoing the mat-
uration molt, occur most commonly at that time.
Such variations include alterations of segmen-
tation and change in structure of appendages
(Weygoldt 1969), tergal/stemal abnormalities,
variation in surface sculpturing and setal and tri-
chobothrial distribution patterns. In some cases
the abnormalities found in adults represent the
retention of tritonymphal features, or localized
neoteny (Gupta 1979;Gabbuttand Vachon 1963;
Weygoldt 1969). It is supposed that a slight dis-
ruption of the neurosecretory and endocrine sys-
tem or local variations in the interpretation of
the neurosecretions could lead to localized ter-
atological abnormalities.

Among the Cheliferinea, abdominal malfor-
mations have been described in Ellingsenius
sculpturatus (Lewis), Anatemnus javanus (Tho-
rell), Dactylochelifer latreillei (Leach), Synsphy-
ronus mimetus Chamberlin, Homs granulatus
(Ellingsen), Allochernes widen (C. L. Koch),
Lamprochernes nodosus (Schrank) and Alloch-
ernes dubius (O. P.-Cambridge) (With 1905;
Hadzi 1930; Gilbert 1952; Chamberlin 1949;
Beier 1955; Weygoldt 1969; Pedder 1965).

In the Chthoniinea, a number of segmental

215

216

THE JOURNAL OF ARACHNOLOGY

anomalies have been recorded from Chthonius
tenuis (L. Koch), C. aff. tetrachelatus (Preyssler)
and C. ischnocheles (Hermann) (Pedder 1965;
Dimitrijevic 1990).

Among the Neobisiinea, false scorpions have
been found with segmentation deficiencies in-
volving the dorsal and ventral sclerites. Such ab-
errations have been noted in Neobisium erythro-
dactylum (L. Koch), N. maritimum (Leach) and
N. muscorum (Leach)(Pedder 1965). Only re-
cently, comparative aspects of teratological vari-
ation have been studied in other neobisiid spe-
cies: N. carpaticum, N. macrodactylum (Daday),
N. cephalonicum (Daday), N. sylvaticum, N. fus-
cimamim (C. L. Koch) (Curcic 1980; Curcic and
Dimitrijevic 1982, 1984, 1985, 1986; Curcic et
al. 1983), N. bernardi Vachon and N. simoni (L.
Koch)(Dimitrijevic 1990). These studies have
revealed the heterogeneity of segmental anom-
alies affecting abdominal sclerites. Such varia-
tions include fusion, splitting and loss of tergites
and stemites. In addition, different combinations
of these anomalies may be present in a single
pseudoscorpion specimen (Curcic 1989a, b).

Quantitative and qualitative analysis of dif-
ferent samples of the genus Roncus L. Koch re-
vealed aberrations (anomalies) in the chelicerae,
pedipalps and abdomen but not in the walking
legs and the cephalothorax. Thus in R. lubricus
L. Koch, 1873, from Mt. Avala, near Belgrade,
Yugoslavia two or three teeth often grow (fuse)
together on both fingers of the pedipalpal chelae.
Furthermore, Curcic (1980) noted an anomaly
in the sclerite segmentation, which was mani-
fested in the fusion of tergites III-VI on each side
and in partial fusion of tergites VIII and IX of
the same specimen. Similar cases of sternal
anomalies have been recorded in the same spe-
cies by Curcic and Dimitrijevic (1984), where
the phenomena of sclerite fusion (symphyso-
mery), atrophy and enlargement have been not-
ed. As a consequence of these anomalies, some
parts of the aberrant stemites lack its normal
setae. Out of four other aberrant specimens (males
and females)(Curcic and Dimitrijevic 1984;
Dimitrijevic 1985), multiple tergal atrophy was
noticed in three specimens, while symphysomery
was observed in only one specimen. The atrophy
occurred on both anterior and posterior tergites
and symphysomery on more posterior tergites.
The percentage of aberrant specimens from these
samples ranged from 0.46% to 0.95% (Curcic and
Dimitrijevic 1986; Dimitrijevic 1985). Curcic and
Dimitrijevic ( 1 988) provided further evidence of

sclerite anomalies in R. lubricus from Mt. Avala
II. In the sample studied, different abdominal
malformations were found in 0.67% of all spec-
imens, the ratio of aberrant males to females
being 4: 1 . The following anomalies were noted:
partial atrophy, hemimery, (single) symphyso-
mery, and a combination of partial atrophy and
tergite enlargement (Curcic 1989b). As a con-
sequence of these aberrations, setal counts and
disposition were altered.

Anomalies in different body structures have
also been observed in some cavemicolous spe-
cies of Roncus inhabiting underground habitats
of the Dinaric Karst (Curcic 1988). Thus, in a
male of R. pripegala Curcic, the trichobothri-
otaxy of one movable chelal finger was altered
(5 trichobothria instead of 4). In another pseu-
doscorpion, R. aff. stussineri (Simon), a sym-
physomery of stemites VII and VIII was found,
while in R. timacensis Curcic and R. remesi-
anensis Curcic, there were also aberrations in the
structure of the chelicerae, in the cheliceral se-
tation, and in the structure of the chelal teeth
(Curcic 1983).

The primary purpose of this study was to an-
alyze the quantitative and qualitative variations
of abdominal anomalies in two Yugoslavian spe-
cies of Roncus pseudoscorpions, their frequen-
cies, common occurrences, and the possible fac-
tors affecting their development and distribution.

MATERIAL AND METHODS

We have analyzed the accidental and terato-
logical variation of abdominal deficiencies in a
population sample of R. aff. lubricus (sample A)
from the village of Obrez, near Belgrade, Yu-
goslavia, as well as in samples of R. aff. lubricus
(sample B) from the village of Asanovac and
from the village of Dubova (Ravniste), both near
Zitoradja, Yugoslavia. The numbers of speci-
mens collected in these localities are presented
in Table 2.

Samples of the two pseudoscorpion species
studied were obtained by sifting oak leaf litter
and humus over a period from April 1989 to
September 1990 (Obrez sample) and from Jan-
uary 1989 to May 1989 (Asanovac and Dubova
samples). Samples were taken once a month at
each of the two localities.

After dissection, all specimens were mounted
in gum chloral medium (Swan’s fluid) and ex-
amined carefully. The terminology for segmental
anomalies in this study is the same as used for
other arthropods (Balazuc 1948, 1967). This ter-

CURCIC ET AL.-SEGMENTAL ANOMALIES IN RONCUS

217

Table 1. — Frequency of different anomalies in Roncus aff. lubricus (A) and (B) (expressed as a percentage of
the total number of anomalies noted in each particular species sample). OBR = Obrez, ASA = Asanovac,
DUB = Dubova.

Species/locality

Segmental anomaly

(A)

OBR

(B)

ASA

(B)

DUB

Hemimery

4

-

-

Partial atrophy

single

—

17

28

multiple

4

—

—

Symphysomery

single

21

—

44

multiple

4

—

—

Combined hemimery and sclerite enlargement

25

49

28

Combined hemimery and symphysomery

—

17

—

Combined partial atrophy and symphysomery

8

—

—

Combined partial atrophy and sclerite enlargement

13

—

—

Combined (single or multiple) symphysomery and sclerite enlargement

—

17

—

Combined atrophy (or hemimery), symphysomery and sclerite enlargement

13

—

—

Combined atrophy, symphysomery, helicomery and sclerite enlargement

4

—

—

Combined atrophy, helicomery and sclerite enlargement

4

-

-

Total

100

100

100

minology has been somewhat modified by (Zurcic
and Dimitrijevic (1986) and Curcic (1989a, b)
to include the whole range of sternal and tergal
deficiencies which were observed in pseudoscor-
pions.

The false scorpions analyzed for the present
study belong to the R. lubricus species complex.
The two samples (A and B) are close to R. lu-
bricus [see Gardini (1983) for description of the
male lectotype from the United Kingdom] but
differ both from lubricus (sensu stricto) and from
each other in many important respects. It there-
fore seems that they belong to two new, previ-
ously undescribed specific taxa. Since their pre-
cise taxonomic status is considered elsewhere

(Curcic, Dimitrijevic & Karamata, in prep.) these
two relevant species have been designated as R.
aff. lubricus (A) and (B). A chi-square test was
used to verify the assumption of the possible sex-
linked inheritance of the abdominal deficiencies
in the analyzed species.

RESULTS

There was a total of 36 abnormal specimens
[23 of R. aff. lubricus (A) and 13 of R. aff. lu-
bricus (B)] (Table 3). Analysis of teratological
variation of segmental anomalies in the pseu-
doscorpions studied gave the following results;

Roncus aff lubricus (A)

Table 2.— The number of specimens of Roncus aff. lubricus (A) and (B) (by sex and growth stages) collected
from various sites. M = males, F = females, T = tritonymphs, D = deutonymphs, P = protonymphs,
OBR = Obrez, ASA = Asanovac, DUB = Dubova.

Sex/instar

Species

Site

M

F

T

D

P

Total

R. aff. lubricus (A)

OBR

1335

832

883

34

3084

R. aff lubricus (B)

ASA

512

433

237

—

—

1182

R. aff. lubricus (B)

DUB

242

225

91

1

—

559

Total

2089

1490

1211

35

-

4825

218

THE JOURNAL OF ARACHNOLOGY

Figures \-l .—Roncusa^. !ubricus{A) from Obrez, near Belgrade, Yugoslavia. Scale line = 0.5 mm. 1. - tergites
I-IV, female; 2. - tergites I-III, female; 3.- tergites II-IV, female; 4. - tergites I-IV, female; 5. - tergites I-IV,
female; 6. - tergites I-IV, female; 7. - tergites VIII-XI, female.

Village of Obrez. - Female (Fig. 1). A section
of tergite III is missing on the left. As a conse-
quence, the left parts of tergites II and IV are
enlarged to fill the missing part of tergite III; the
setation of this tergite is altered. Hemimery and
tergite enlargement.

Female (Fig. 2). The left part of tergite II is
completely missing, and the setae are unequally
distributed. Hemimery.

Female (Fig. 3). The right part of tergite III is
fused with the anterior mid-region of tergite IV.
A small section of tergite III is present on the
left. The number and disposition of setae on ter-
gite III are altered as a consequence of this anom-
aly. Partial helicomery, atrophy and tergite en-
largement.

Female (Fig. 4). A small part of tergite III is
missing on the left. Tergite II is enlarged and fills
the space where the missing part of tergite III
would otherwise be found. The setae on tergite

II are almost all missing, and their position on
tergites III and IV is changed. Atrophy and ter-
gite enlargement.

Female (Fig. 5). Right part of tergite III is fused
with tergite IV medially. A small part of tergite

III is isolated on the left and the median region
of this tergite lacks setae. Partial atrophy and
symphysomery.

Female (Fig. 6). The left part of tergite III is
missing, and the remaining part of this sclerite
is fused with the mid-anterior region of tergite
IV. Tergites II and IV are enlarged to fill the space
left vacant in tergite III. The setae on tergite III

are missing in part. Atrophy (possibly hemi-
mery), symphysomery and tergite enlargement.

Female (Fig. 7). Tergite X is reduced on the
right, and its number of setae is smaller than in
normal specimens. The right posterior and lat-
eral part of tergite IX is enlarged to fill part of
the space where the missing part of tergite X
would otherwise be found. Hemimery and tergite
enlargement.

Female (Fig. 8). Tergite VI is fused with tergite

VII medially. As a result of this anomaly, the
disposition of setae on tergite VI is altered. Sym-
physomery.

Female (Fig. 9). The mid-anterior part of ter-
gite VIII is lacking. The posterior border of the
preceding tergite is slightly enlarged posteriorly.
Partial atrophy and tergite enlargement.

Female (Fig. 10). Tergites VII and VIII are fused
on the left; in addition, a small part of tergite

VIII is missing on the left. Partial atrophy and
symphysomery.

Male (Fig. 1 1). Tergites III and IV are fused
medially and on the right. A small isolated sec-
tion of tergite III with 3 setae is present on the
right. The mid-region of tergite III is devoid of
setae. In addition, tergite 11 is slightly enlarged
posteriorly. Symphysomery, atrophy and tergite
enlargement.

Male (Fig. 12). The left part of tergite IV is
missing, and the following tergite is enlarged on
the left to fill the space left vacant by tergite IV,
hence the irregular setation on tergites IV and V.
Hemimery and tergite enlargement.

CURCIC ET AL.- SEGMENTAL ANOMALIES IN RONCUS

219

Figures S-IO. —Roncus aff. lubricus (A), from Obrez, near Belgrade, Yugoslavia. Scale line = 0.5 mm. 8. -
tergites V-VII, female; 9. - tergites VII,VIIL female; 10. - tergites VI-VIII, female.

Male (Fig. 13). The abnormalities affect five
tergites. First, tergite II is atrophied on the left
and tergite III is enlarged to fill the space where
the missing part of tergite II would otherwise be
found. Second, tergites IV-VI are fused on the
left. As a result of this anomaly, the number of
setae on tergites II-V is reduced, and their dis-
tribution is altered when compared with normal
specimens. Atrophy, hemimery, tergite enlarge-
ment and multiple symphysomery.

Male (Fig. 14). The right part of tergite II is
completely lacking, and the setae are unequally
distributed. Tergite III is enlarged to fill the space
left vacant in tergite II. Hemimery and tergite
enlargement.

Male (Fig. 1 5). Tergites II and III are fused on
the left, hence the irregular setation of tergite II.
Symphysomery.

Male (Fig. 1 6). The malformations affect three
tergites.The mid-region and right part of tergite
VI are weakly sclerotized, and its distribution of
setae is irregular (two setal rows on the right part
of the sclerite). Furthermore, small sections of

tergite VII (on the left) and tergite VIII (on the
anterior border) are missing. Multiple (partial)
atrophy.

Male (Fig. 1 7). Tergites VII and VIII are fused
on the left, hence the irregular distribution of
setae on tergite VII. Symphysomery.

Male (Fig. 18). Tergites VIII and IX are fused
together. In addition, small parts of tergites VIII
and IX are missing on the right, hence the irreg-
ular setation on tergites VII-IX. Partial atrophy
and symphysomery.

Female (Fig. 19). Part of stemite VIII is com-
pletely absent. Stemite IX is enlarged on the left
and fills the space where the missing part of ster-
nite VIII would otherwise be developed. The se-
tae on the remaining part of stemite VIII are
missing, and their position is altered in relation
to normal specimens. Hemimery and stemite en-
largement.

Male (Fig. 20). Stemites VIII and IX are fused
from the midline to the right side; hence, the
setae are missing on the right part of stemite VIII.
Symphysomery.

tl

t3

Figures 1 1-17 .—Roncus aff. lubricus (A), from Obrez, near Belgrade, Yugoslavia. Scale line = 0.5 mm. 1 1.-
tergites I-IV, male; 12. - tergites I-V, male; 13. - tergites I-VI, male; 14. - tergites I-III, male; 15. - tergites I-
III, male; 16. - tergites V-VIII, male; 17. - tergites VI-VIII, male.

220

THE JOURNAL OF ARACHNOLOGY

Figures \8-23. — Roncus aff. lubricus (A), from Obrez, near Belgrade, Yugoslavia. Scale line = 0.5 mm. 18. -
tergites VI-IX, male; 19. - sternites VI-IX, female; 20. - stemites VII-IX, male; 21. - tergites II-IV, tritonymph;
22. - tergites V-VIII, tritonymph; 23. - tergites IV-X, tritonymph.

Tritonymph (Fig. 21). Tergites III and IV are
almost completely fused. As a result of this mal-
formation, the setation of tergite III is altered
(the mid-region of this sclerite is devoid of setae).
Symphysomery.

Tritonymph (Fig. 22). The posterior mid-re-
gion of tergite VI is atrophied; hence, the setae
are missing in this area. In addition, tergite VII
is enlarged anteriorly to fill the space left vacant
by the tergite VI. Partial atrophy and tergite en-
largement.

Tritonymph (Fig. 23). The tergal anomalies
affect six tergites. Tergites V and VI are fused in
the middle and on the left; tergites VII and VIII
are fused on the right, as are tergites IX and X.
As a consequence, the mid-regions of tergites VI
and VII lack setae. Multiple symphysomery.

Roncus aff. lubricus (B)

Village of Asanovac. - Female (Fig. 24). In this
specimen, right sclerite halves are missing on
tergites II and III. The right anterior part of ter-

24. - tergites I-IV, female; 25.
28.- tergites VI-X, male; 29. -

lubricus (B), from Asanovac, near Zitoradja, Yugoslavia. Scale line = 0.5 mm.
- tergites V-VII, female; 26. - tergites VII, VIII, male; 27. - tergites II-V, male;
tergites V-VIII, male; 30. - tergites V-XI, male.

CURCIC ET AL. -SEGMENTAL ANOMALIES IN RONCUS

221

Figures 31-31 .—Ronciis aff. lubricus (B), from Dubova, near Zitoradja, Yugoslavia. Scale line = 0.5 mm. 31.
- tergites IX-XI, female; 32. - stemites VIII, IX, female; 33. - tergites VII-IX, female; 34. - tergites V,VI, female;
35.- tergites I-IV, male; 36. - tergites IV-VII, tritonymph; 37. - tergites IV-VI, tritonymph.

gite IV is enlarged to fill part of the space left
vacant in tergites II and III. Multiple hemimery
and tergite enlargement.

Female (Fig. 25). Tergite VII is reduced in size
in its anterior part on the right; hence, tergite VI
is slightly enlarged posteriorly to fill the space
where the missing part of tergite VII would oth-
erwise be found. Partial atrophy and tergite en-
largement.

Male (Fig. 26). The small anterior part of ter-
gite VIII is missing. Partial atrophy.

Male (Fig. 27). Right part of tergite IV is miss-
ing. Tergites III and V are enlarged and fill the
space where the missing part of tergite IV would
otherwise be found. Hemimery and tergite en-
largement.

Male (Fig. 28). Left part of tergite IX is miss-
ing. Tergites VII and VIII are fused on the right.
As a consequence of these anomalies, the distri-
bution of setae on tergites VII-IX is altered in
relation to normal individuals. Hemimery and
symphysomery.

Male (Fig. 29). The left half of tergite VII is
missing. Tergite VI is enlarged posteriorly, and
tergite VIII is enlarged anteriorly. Hemimery and
tergite enlargement.

Male (Fig. 30). The tergal anomalies affect six
tergites. Tergites VII-XI are fused (tergites VII-
IX on the left and tergites IX-XI on the right).
In addition, tergite VI is enlarged posteriorly on
the left. As a consequence of these anomalies,
the setation of tergites VII-XI is drastically al-
tered. Tergite enlargement and multiple sym-
physomery.

Village of Dubova (Ravniste) - Female (Fig.
31). Tergites X and XI are fused on the right.
The distribution of setae on tergite X is altered

(symphysomery). The mid-anterior part of ster-
nite IX is also missing (Fig. 32; partial atrophy).

Female (Fig. 33). A deficiency is found in ter-
gite VIII as manifested by the absence of the left
part. The adjacent region of tergite IX is enlarged
and partially fills the gap left by the missing half
of tergite VIII. Hemimery and tergite enlarge-
ment.

Female (Fig. 34). Part of tergite VI on the right
is missing. Partial atrophy.

Male (Fig. 35). The right parts of tergites II
and III are missing. The adjacent part of tergite
IV is enlarged anteriorly to fill part of the space
left vacant in tergites II and III. The number and
disposition of setae on tergites II and III are al-
tered. Multiple hemimery and tergite enlarge-
ment.

Tritonymph (Fig. 36). Tergites VI and VII are
fused on the right, hence the unequal distribution
of tergal setae. Symphysomery.

Tritonymph (Fig. 37). Tergites V and VI are
fused on the left. As a consequence of this anom-
aly, the distribution and number of setae are
changed in relation to normal specimens (Zlat-
kovic 1989). An isolated seta is present on the
left side of tergite V. Symphysomery.

The numbers of anomalies of males and fe-
males of R. aff. lubricus (A) and (B) were not
significantly different (x^ test, P >.05) for any
of the three locations. Thus, these data offer no
confirmation that a sex-linked inheritance of ab-
dominal deficiencies exists in the analyzed spe-
cies.

Comparison of the teratological variation in
the analyzed species of Roncus clearly showed
that this phenomenon was particularly marked
in R. aff. lubricus (A) (Table 1), less so in R. aff

222

THE JOURNAL OF ARACHNOLOGY

Table 3.— Abdominal abnormalities in different sexes and growth stages of Roncus aff. lubricus (A) and (B)
from three sites. M = males, F = females, T = tritonymphs, D = deutonymphs, P = protonymphs, OBR = Obrez,
ASA = Asanovac, DUB = Dubova.

Species

Site

Sex/instar

%

abnormal

males

%

abnormal

specimens

M

F

T D

P

Total

R. aff. lubricus (A)

OBR

9

1 1

3

—

23

39.3

0.7

R. aff. lubricus (B)

ASA

5

2

— —

—

7

71.4

0.6

R. aff. lubricus (B)

DUB

1

3

2

-

6

16.7

1.1

Total

15

16

5

—

36

Mean

42.4

0.8

lubricus (B) from Asanovac, and the least in R.
aff. lubricus (B) from Dubova (Table 1) and R.
lubricus (Curcic 1989b, see Table 5). Such obvi-
ous differences, or the different degrees of tera-
tological variation, may be due to different fre-
quencies of some segmental anomallies or sample
sizes (i.e., too small to give a detailed picture of
the whole range of teratological variability of the
species). Another explanation for the unequal
variation of teratological phenomena might be
the different susceptibility of various pseudo-
scorpion populations to the factors which cause
malformations. That both assumptions may be
correct is further supported by the similar un-
equal frequency of different anomalies in other
genera and species of the Neobisiidae, and par-
ticularly of the species of the genus Neobisium
(Curcic 1989a, b).

DISCUSSION

Postembryonic instars of each species were

found in each of the localities (Table 2), with the
absence of protonymphs of R. aff. lubricus (A)
from Obrez and Dubova, and deutonymphs and
protonymphs of R. aff. lubricus (B) from Asa-
novac.

In the more frequently occurring species, R.
aff. lubricus (A), 23 abnormal specimens were
noted (Table 3), while in R. aff. lubricus (B), the
numbers of aberrant examples in different sam-
ples ranged from six to seven (Table 3). The ma-
jority of abnormal specimens were adults, with
the exception of three tritonymphs of R. aff. lu-
bricus (A) and two tritonymphs of R. aff. lubricus
(B)(Table 3).

In R. aff. lubricus (A) from Obrez we found
that the frequency of anomalies of the abdominal
sclerites is 0.7%, whereas in R. aff. lubricus (B)
it varies from 0.6% to 1.1%, depending on the

site. In general these values correspond to the

frequency of anomalies in R. lubricus from other
sites, as noticed elsewhere by Curcic (1989b).
The highest percentage of these aberrations has
been noted in the sample of R. aff. lubricus (B)
from Dubova, and the lowest in the sample of
the same species from Asanovac. It is pertinent
to note that the frequency of segmental anom-
alies in R. aff. lubricus (A) and (B) is essentially
similar in different populations.

The analysis of the samples of R. aff. lubricus
(A) and (B) showed that teratological variation
of the abdominal sclerites was confined mostly
to adults [86.9% in R. aff. lubricus (A), and from
66.6% to 100% in R. aff. lubricus (B)].

The deficiencies in abdominal sclerites of the
species studied were variable. Thus, in R. aff.
lubricus (A) as many as 1 1 different single or
combined anomalies were noted. However, in
R. aff. lubricus (B) from Asanovac only four types
of aberrations were found, while in the same spe-
cies from Dubova we noted the existence of three
kinds of malformations. Among the reasons for
the unequal distribution of anomalies in different
samples might be the inappropriate sample size
or the unequal susceptibility to the factors caus-
ing such aberrations.

The analysis of the samples of R. aff. lubricus

(A) and (B) revealed that malformations of ab-
dominal sclerites are confined mainly to adults.
It is also evident that tergal anomalies are more
frequent than those affecting the stemites.

The sex-linked distribution of segmental
anomalies varies considerably, depending on the
sample studied. Thus, in R. aff. lubricus (A) only
39.3% of males were aberrant; in R. aff. lubricus

(B) , however, the percentage of abnormal males
varied from only 16.7% (Dubova) to as high as
7 1 .4% (Asanovac). However, Curcic ( 1 989b) re-
ported an incidence of 80.0% of aberrant males

CURCIC ET AL.- SEGMENTAL ANOMALIES IN RONCUS

223

in R. lubricus, and 50-100% aberrant males in
different species of Neobisiurn (with the sole ex-
ception of N. sylvaticum, where the frequency of
abnormal males was only 16.7%). One possible
explanation is that the genesis of abdominal de-
ficiencies is connected either with male- or fe-
male-linked inheritance, as noted elsewhere by
Curcic and Dimitrijevic (1985, 1986).

The study of the relative distribution of seg-
mental anomalies in the adult and tritonymph
stages of R. aff lubricus (A) and (B), as compared
to the same feature in different species of Neo-
bisium and Roncus (Curcic 1989a, b) revealed
the following; (1) hemimery, if present, is re-
stricted to the anterior part of the abdomen, (2)
atrophy is found mostly in the anterior abdom-
inal region, although it may occur also in the
posterior region, as already noticed by Curcic
(1989b), (3) symphysomery develops in the an-
terior, median and posterior sclerites, whereas in
Neobisiurn it was noted that this malformation
is present mainly in the central region of the
abdomen (Curcic 1989b), (4) the relative posi-
tion of sclerite enlargement is often correlated
with the presence and relative position of partial
atrophy or hemimery. Hence, the combination
of these anomalies develops both in the anterior
and in the posterior abdominal parts. On the
other hand, combined symphysomery and scler-
ite enlargement have been noted in the posterior,
while other combinations of different aberrations
are confined to the anterior sclerites. Therefore,
the evidence [furnished elsewhere by Curcic
(1989a, b)] obtained for i?. aff. lubricus in general
confirms the data for Neobisiurn and Roncus spe-
cies, (5) helicomery (combined with other mal-
formations) is confined to the anterior sclerites.
It is noteworthy that Curcic and Dimitrijevic
(1986) have found that the single case of heli-
comery was restricted to the posterior region of
the abdomen.

In each analyzed sample, segmental deficien-
cies are unequally distributed between represen-
tatives of different sexes. In R. aff. lubricus (A)
and R. aff. lubricus (B) from Dubova, the per-
centage of anomalous females was higher than
that of aberrant males. Only in R. aff. lubricus
(B) from Asanovac was the frequency of aberrant
males higher than that of females. In the majority
of Neobisiurn species which were analyzed by
Curcic (1989b), the percentage of anomalous
males was found to be higher than that of ab-
errant females.

In most cases, the development of sclerite
anomalies causes simultaneous malformations
in the setation of tergites and stemites. In other
words, changes in the number, size and dispo-
sition of setae result from deficiencies affecting
either parts of or whole sclerites.

It is already known that the majority of the
abdominal anomalies occur during the transfor-
mation of the tritonymphs into adults. A con-
siderably smaller number of specimens become
anomalous when transforming from the deuto-
nymph into tritonymphs (Curcic, Krunic and
Brajkovic 1 983; Curcic 1 989a, b; Legg and Jones
1988). The causes of the lower frequency of ab-
errations in earlier instars are unknown.

A number of anomalies are likely to be the
result of earlier mechanical injuries, at the ju-
venile or the adult stage. The most frequent con-
sequence of such lesions, apart from alteration
of setation, is the depigmentation of certain areas
of the sclerites. Such specimens are viable and
fully capable of reproduction.

Although the malfunction of the hormonal
system, as well as some environmental factors,
might be causes of various aberrations in ab-
dominal sclerites, it seems likely that genetic fac-
tors can also give rise to such aberrations (Gehr-
ing 1985). Among them, the factors of
metamerization should be especially mentioned.
There are a number of findings which would sup-
port this view: the constancy of teratological
variation in wild populations, a comparatively
similar incidence in percentages of abnormal
specimens in different populations, the noted de-
gree of qualitative diversity and specific features
of the distribution of different aberrations at var-
ious growth stages and in both sexes in each par-
ticular species.

ACKNOWLEDGMENTS

We are grateful to Zoran Ivkovic (Belgrade)
for his useful comments on some statistics used
in this paper. We are also grateful to Sigurd Nel-
son Jr. and Vincent F. Lee for the constructive
criticism of the manuscript.

LITERATURE CITED

Balazuc, J. 1948. La teratologic des Coleopteres et
experiences de transplantation sur Tenebrio molitor
L. Mem. Mus. Hist. Nat., Paris, 25:1-293.

Balazuc, J. 1967. Production experimentale d’an-
omalies de segmentation chez Mantis religiosa L.
Entomologiste, 23:45-50.

224

THE JOURNAL OF ARACHNOLOGY

Beier, M. 1955. Pseudoscorpionidea. S. African Anim.
Life, 1:263-328.

Chamberlin, J. C. 1949. New and little-known false
scorpions from various parts of the world (Arach-
nida, Chelonethida), with notes on structural ab-
normalities in two species. American Mus. Novi-
tates, 1430:1-59.

Curcic, B. P. M. 1980. Accidental and teratological
changes in the family Neobisiidae (Pseudoscor-
piones, Arachnida). Bull. British Arachnol. Soc., 5:
8-15.

Curcic, B. P. M. 1983. New cavernicolous pseudos-
corpions from Serbia. Rec. Rapp. Com. Karst Spe-
leol., Ed. Spec. Acad. Serbe Sci. Arts, Prcsid., 1 : 1 35-
150.

Curcic, B. P. M. 1988. Cave-dwelling pseudoscor-
pions of the Dinaric Karst. Acad. Sci. Art. Slov., Cl.
IV, Hist. Nat., Opera, 26, Inst. Biol. loannis Hadzi,
8:1-192.

Curcic, B. P. M. 1989a. Segmental anomalies in some
European Neobisiidae (Pseudoscorpiones, Arach-
nida) - Part I. Acta Arachnol., 37:77-87.

Curcic, B. P. M. 1989b. Segmental anomalies in some
European Neobisiidae (Pseudoscorpiones, Arach-
nida) - Part II. Acta Arachnol., 38:1-10.

Curcic, B. P. M., & R. N. Dimitrijevic. 1982. On
abnormalities of abdominal segmentation in Neo-
bisium carpaticum Beier (Neobisiidae, Pseudoscor-
piones, Arachnida). Revue Arachnol., 4:143-150.

Curcic, B. P. M., & R. N. Dimitrijevic. 1984. An
abnormal carapaco-abdominal junction in Neobis-
ium carpaticum Beier, 1934 (Neobisiidae, Pseudos-
corpiones). Arch. Sci. Biol. Belgrade, 36:9P-10P.

Curcic, B. P. M., & R. N. Dimitrijevic. 1985. Ab-
dominal deficiencies in four species of the Neobi-
siidae (Pseudoscorpiones, Arachnida). Revue Ar-
achnol., 6:9 1-98.

Curcic, B. P. M., & R. N. Dimitrijevic. 1986. Ab-
normalities of carapacal and abdominal segmenta-
tion in Neobisium Chamberlin (Neobisiidae, Pseu-
doscorpiones). Actas X Congr. Int. Aracnol., Jaca/
Espana, 1:17-23.

Curcic, B. P. M., & R. N. Dimitrijevic. 1988. Seg-
mental deficiencies in some Neobisiidae (Pseudos-
corpiones, Arachnida). TUB-Dokument. Kongr.
Tagungen, XI Europ. Arachnol. Coll., Berlin, 1:89-
97.

Curcic, B. P. M., M. D. Krunic, & M. M. Brajkovic.
1981. Further records of teratological changes in
the Neobisiidae (Arachnida, Pseudoscorpiones). Bull.
British Arachnol. Soc., 5:280-284.

Curcic, B. P. M., M. D. Krunic, & M. M. Brajkovic.
1983. Tergal and sternal anomalies in Neobisium
Chamberlin (Neobisiidae, Pseudoscorpiones,
Arachnida). J. Arachnol., 11:243-250.

Dimitrijevic, R. N. 1985. Patomorfoloska analiza
promena u segmentaciji abdomena kod nekih pred-
stavnika porodice Neobisiidae (Pseudoscorpiones,
Arachnida). M. Sc. Thesis, University of Belgrade.
72 pp.

Dimitrijevic, R. N. 1990. Neka obelezja postem-
brionalnograzvica i teratoloskih promena kod pseu-
doskorpija iz porodica Chthoniidae i Neobisiidae
(Pseudoscorpiones, Arachnida). Ph. D. Thesis, Uni-
versity of Belgrade. 24 1 pp.

Gabbutt, P. D., & M. Vachon. 1963. The external
morphology and life history of the pseudoscorpion
Chthonius ischnocheles (Hermann). Proc. Zool. Soc.
London, 140:75-98.

Gardini, G. 1983. Redescription of Rona/x

L. Koch, type-species of the genus Roncus L. Koch
1873 (Pseudoscorpionida, Neobisiidae). Bull. Brit-
ish Arachnol. Soc., 6:78-82.

Gehring, W. J. 1985. Homeotic genes, the homeobox
and the genetic control of development. Cold Spring
Harbor Symp. Quant. Biol., 50:243-251.

Gilbert, O. 1952. Three examples of abnormal seg-
mentation of the abdomen in Dactylochelifer la-
treillei (Leach), (Chelonethi). Ann. Mag. Nat. Hist.,
5:47-49.

Gupta, A. P. 1979. Arthropod phylogeny. Van Nos-
trand-Reinhold Co., New York, 762 pp.

Hadzi, J. 1930. Prirodoslovna istrazivanja sjeverno-
dalmatinskog otocja. 1. Dugi i Kornati. Pseudos-
corpiones, Prirod. istr. Kralj. Jugosl., 16:65-79.

Hadzi, J. 1937. Pseudoscorpioniden aus Siidserbien.
Bull. Soc. Sci. Skoplje, 17/18:13-38 and 151-187.

Kastner, A. 1927. Pseudoscorpiones. Biologic Tiere
Deutschlands (Berlin), 18:1-68.

Legg, G., & R. E. Jones. 1988. Pseudoscorpions (Ar-
thropoda; Arachnida). Keys and notes for the iden-
tification of the species. The Linnean Soc. London,
Synopses British Fauna (N. S.), 40:1-159.

Pedder, 1. J. 1965. Abnormal segmentation of the
abdomen in six species of British pseudoscorpions.
Entomologist (London), 98:108-1 12.

Weygoldt, P. 1969. The biology of pseudoscorpions.
Harvard University Press, Cambridge, xiv+ 145 pp.

With, C. J. 1905. On Chelonethi, chiefly from Aus-
tralian Region, in the collection of the British Mu-
seum, with observations on the “coxal sac” and
some cases of abnormal segmentation. Ann. Mag.
Nat. Hist., (7)15:94-143.

Zlatkovic, M. R. 1989. Segmentalne anomalije kod
pseudoskorpija Roncus i Neobisium (Pseudoscor-
piones) iz okoline Zitoradje. B. Sci. Thesis, Uni-
versity of Belgrade. 42 pp.

Manuscript received March 1991, revised May 1991.

1991. The Journal of Arachnology 19:225-226

RESEARCH NOTES

AN EXAMPLE OF ABNORMAL CARAPACO-ABDOMINAL
FUSION IN NEOBISIUM AFF. FUSCIMANUM
(ARACHNIDA, PSEUDOSCORPIONES, NEOBISIIDAE)

Teratological phenomena in various represen- 1982). The setation of tergite I has been drasti-

tatives of the pseudoscorpion family Neobisiidae
are extremely diverse, as shown by Curcic ( 1 980,
1989a, b), Curcic and Dimitrijevic (1984, 1985,
1986, 1990) and Curcic et al. (1981, 1983). The
majority of the observed malformations affect
the abdominal segmentation, thus causing either
tergal or sternal deficiencies. However, a number
of specimens have been found which exhibit var-
ious aberrations of the chelicerae, pedipalps, and
walking legs (Curcic 1980).

Records of carapacal deficiencies in the Neo-
bisiidae are very sparse. Only recently, an ab-
normal carapaco-abdominal junction (fusion) was
studied in a male of Neobisium carpaticum Beier
(Curcic and Dimitrijevic 1984, 1986). To our
knowledge, this is the only case of malformation
affecting the carapace which has been discovered
to date.

In a collection of pseudoscorpions made at
Asanovac, near Zitoradja, Serbia (Yugoslavia)
during March 1989, one anomalous deuto-
nymph of Neobisium aff. fuscimanum (C. L.
Koch) was found. The specimen was obtained
by sifting humus in an oak forest. In the speci-
men, only the carapace and the anterior tergites
were anomalous, while the stemites and the ap-
pendages were normal in all respects.

The purpose of this note is to describe the
phenomenon of carapaco-abdominal fusion in
the aberrant deutonymph. The carapace and the
first two tergites of this specimen are anomalous
(Fig. 1). The carapace lacks a chitinous section
on the right posterior margin; instead a thin and
transparent membrane is present in the area where
the missing part of the carapace would otherwise
be found. In addition, tergite I is fused with the
carapace along its mid-anterior region. As a con-
sequence of this deficiency, the carapacal setation
in this specimen is significantly altered in relation
to the normal setal complement for a deuto-
nymph, 4 + 6 -I- 6 + (6-7) = 22-23 (Curcic

cally changed, since the number of setae is greatly
increased and their distribution irregular. The
deutonymph of N . fuscimanum normally carries
6-8 setae on tergite I (Curcic 1982). Tergite II
bears 3 small setae on the left, but is devoid of
setae on the right. The tergal section with 6 setae
otherwise found anterior to tergite II might rep-
resent a part of this tergite, which perhaps had
been split originally into two transverse areas.
However, it is also possible that this demi-tergite
may represent a supernumerary sclerite, located
between tergites I and II. This assumption is based
on the presence of an additional row of setae on
this isolated tergal section. A similar case has
been noted in N. sylvaticum C. L. Koch (Curcic
and Dimitrijevic 1985). Altogether, five types of
abnormalities have been found to affect the car-
apace and abdominal tergites in this deuto-
nymph: (1) partial atrophy of the carapace, (2)
carapaco-abdominal fusion [symphysomery], (3)
partial atrophy of tergite II, (4) the occurrence
of a supernumerary sclerite and, (5) the alteration
of setation in both carapace and tergites.

The segmental anomalies in neobisiid species
mainly occur during the “maturation molt”, or
the transformation of tritonymph into adult
(Curcic 1989b). Considerably fewer specimens
become aberrant when transforming from deu-
tonymph into tritonymph, or from the proto-
nymph into deutonymph, as was shown by Cur-
cic ( 1 989a, b) and Curcic and Dimitrijevic ( 1 986).

The origin of the malformations in this spec-
imen of N. aff. fuscimanum is still unclear. We
assume that the genesis of the drastically mod-
ified carapaco-abdominal junction in this ex-
ample is provoked by genetic factors, especially
those affecting the metamerization period, as was
shown elsewhere by Gehring (1985) for repre-
sentatives of various invertebrates.

We are grateful to M. R. Zlatkovic (Zitoradja)
for his help in collecting pseudoscorpions.

225

226

THE JOURNAL OF ARACHNOLOGY

Figure \ .—Neobisium aS. fuscimanum (C. L. Koch).
Scale line = 0.5 mm. C+Tl = carapace and tergite I,
ST = supernumerary tergite (?), T2 = tergite II.

LITERATURE CITED

Curcic, B. P. M. 1980. Accidental and teratological
changes in the family Neobisiidae (Pseudoscor-

piones, Arachnida). Bull. British Arachnol. Soc., 5:
9-15.

Curcic, B. P. M. 1982. Postembryonic development
in the Neobisiidae (Pseudoscorpiones, Arachnida).
Serb. Acad. Sci. Arts, Belgrade, Monogr., DXLV,
Dept. Sci., 56:1-90.

Curcic, B. P. M. 1989a. Segmental anomalies in some
European Neobisiidae (Pseudoscorpiones, Arach-
nida) - Part I. Acta Arachnol., 37:77-87.

Curcic, B. P. M. 1989b. Segmental anomalies in some
European Neobisiidae (Pseudoscorpiones, Arach-
nida) - Part II. Acta Arachnol., 38:1-10.

Curcic, B. P. M., & R. N. Dimitrijevic. 1984. An
abnormal carapaco-abdominal junction in Neobis-
lum carpaticum Beier, 1934 (Neobisiidae, Pseudos-
corpiones). Arch. Sci. Biol. Belgrade, 36:9P-10P.

Curcic, B. P. M., & R. N. Dimitrijevic. 1985. Ab-
dominal deficiencies in four species of the Neobi-
siidae (Pseudoscorpiones, Arachnida). Rev. Arach-
nol., 6:91-98.

Curcic, B. P. M., & R. N. Dimitrijevic. 1986. Ab-
normalities of carapacal and abdominal segmenta-
tion in Neobisium Chamberlin (Neobisiidae, Pseu-
doscorpiones). Actas X Congr. Int. Aracnol., Jaca/
Espana 1986, 1:17-23.

Curcic, B. P. M., & R. N. Dimitrijevic. 1990. An
example of partial duplication of the abdomen in
Neobisium simoni (Pseudoscorpiones, Neobisiidae).
J. Arachnol., 18:113-115.

Curcic, B. P. M., M. D. Krunic, & M. M. Brajkovic.
1981. Further records of teratological changes in
the Neobisiidae (Arachnida, Pseudoscorpiones). Bull.
British Arachnol. Soc., 5:280-284.

Curcic, B. P. M., M. D. Krunic, & M. M. Brajkovic.
1983. Tergal and sternal anomalies in Neobisium
Chamberlin (Neobisiidae, Pseudoscorpiones,
Arachnida). J. Arachnol., 1 1:243-250.

Gehring, W. J. 1985. Homeotic genes, the homeobox
and the genetic control of development. Cold Spring
Harbor Symp. Quant. Biol., 50:243-251.

Bozidar P. M. Curcic and Rajko N. Dimitrijevic:

Institute of Zoology, Faculty of Science, Uni-
versity of Belgrade, Studentski Trg 16, YU-
1 1000 Beograd, Yugoslavia.

Manuscript received May 1991.

1991. The Journal of Arachnology 19:227-228

IN VITRO POST DISPERSAL BURROW SHARING AMONG
SPIDERLING GEOLYCOSA TURRICOLA
(ARANEAE, LYCOSIDAE)

An essential component in the evolution of
sociality in spiders is the extension of broodmate
tolerance and the delay, or reduction (e.g., Age-
lena consociata Denis; RoeloflFs & Riechert 1988)
of dispersal of the young (Shear 1 970). Subsocial
strategies (maternal social or extended tolerance
among broodmates) may be viewed as inter-
mediate between the solitary life style of most
species and the complex behavior of the social
species in that social cohesion is temporary
(though in some species it is prolonged, e.g.,
Nemisia caementaria, Buchli 1969). The nature
of the dispersal strategy in these temporarily so-
cial groups is of interest. In particular, the ques-
tion whether the mechanisms that end the sub-
social phase also trigger the onset of dispersal is
important since post-dispersal tolerance among
broodmates or spiderlings and their mother
would raise the possibility that the advantages
of subsociality reach beyond the proximate ben-
efits of food procurement or protection at the
nest site to processes of later development such
as nest site location and mate finding. To date,
there have been few studies of the dispersal strat-
egy of subsocial spiders (a notable exception is
Krafft et al. 1 985) and, thus, little is known about
the functional ecology of this process. Here I
report observations from a study of the burrow-
ing spider Geolycosa turricola (Treat) that suggest
that dispersal and the termination of tolerant
behavior may be independent events in this spe-
cies.

The biology and ecology of G. turricola has
been previously reported in detail (Miller & Mil-
ler 1987, 1991). Briefly, females produce egg cases
in their burrows in early spring and remain there
during the early development of the young (Mil-
ler & Miller 1 985). Some spiderlings may remain
with their siblings in the maternal burrow well
past the time when dispersal and burrow con-
struction is possible. During this time tolerant
spiderlings share large prey items and exhibit no
cannibalism or agonistic behaviors toward
broodmates (Miller 1989).

As part of a study of the mechanism that trig-

gers dispersal from these subsocial groups (Mil-
ler, in prep.), I observed the dispersal activities
of five broods (each with their mother; brood
size T = 82, SD = 15.3), each placed in paper
burrows positioned in the center of 1.5 m di-
ameter plastic swimming pools. Adult Geolycosa
readily accept paper burrows and routinely build
turrets on them. A sand substrate devoid of veg-
etation surrounded each burrow; the sand was
moistened each morning and afternoon with a
plant mister. The arenas were observed twice
each day for a period of four months (March-
June). Spiderlings that dispersed from all the ma-
ternal burrows (n = 78) easily constructed bur-
rows in the sand. Small crickets were provided
for food.

On six occasions (twice in one arena; once each
in the four other arenas) a spiderling dispersed,
constructed a burrow and, within a day, was
joined in that burrow by another spiderling. On
two occasions, (two different arenas) three spi-
derlings shared a single small burrow. The shared
burrows were located a considerable distance
from the maternal burrow (point of dispersal) {X
= 22.3 cm, SD = 6.2, n = 8) but were significantly
closer to the maternal burrow than burrows con-
taining single spiderlings {X = 43.1 cm, SD =
10.2, n = 78, all arenas combined; t = 5.64, P
< 0.001). The average diameter of the shared
burrows was not significantly different from that
of burrows containing only a single spider.

Upon their discovery, each shared burrow was
enclosed with a small screen cage (6 cm in di-
ameter) so that burrow desertion by one or both
of the spiderlings could be observed (small crick-
ets were provided for food). The period of bur-
row sharing ranged from 3 to 1 2 days. In five of
the six burrows shared by two spiderlings, one
of the spiderling deserted and constructed an-
other burrow within the wire cage. In the other
case one of the spiderlings disappeared and was
presumed to have been eaten. In the case of the
two burrows shared by three spiderlings, two of
the spiderlings deserted the burrow and each con-
structed burrows of their own within the cage.

227

228

THE JOURNAL OF ARACHNOLOGY

It is unclear how burrows were located by dis-
persing spiderlings. Groups of spiderling G. tur-
ricola held in Petri dishes deposit a considerable
amount of silk (Miller 1989) and there is some
evidence that silk is deposited by dispersing spi-
derlings (Miller unpubk). Such silk trails could
be followed by dispersing broodmates. The short
time between the establishment of a burrow by
a spiderling, and the joining of the spiderling by
another individual (less than a day in each case)
suggests such a process. Such a process might be
facilitated if spiderlings disperse in groups or
swarms such as in some social spiders (Lubin &
Robinson 1985). However, the presence of such
behavior would presumably result in nonran-
dom vectors of dispersal from the maternal bur-
row. Field and laboratory studies of the pattern
of burrow establishment of this species (Miller
& Miller 1991) do not show such directionality.

The burrow sharing observed here probably
does not reflect a paucity of suitable burrow sites
within the experimental chambers. Although
Miller ( 1 984) showed that spiderling G. turricola
prefer burrow sites that contain some vegetation,
different from the barren conditions of this ex-
periment, the number of shared burrows repre-
sented only 10% of the total number of spiders
that were observed to build burrows in the ap-
paratus indicating that the majority of spider-
lings found suitable burrow sites.

Burrow sharing could represent an avenue for
delaying the cost of burrow construction. Spi-
derling G. turricola and G. rafaelana are known
to disperse in late summer and then delay burrow
construction until the spring (Miller & Miller
1991, Conley 1985, respectively). Miller & Mil-
ler ( 1 99 1 ) showed that the size of the first burrow,
which is highly correlated to the size of the spi-
derling, is an important indicator of overwinter
survivorship in G. turricola. However, burrow
sharing has never been observed in the field.
Moreover, it is unclear what advantage burrow
sharing after dispersal has over subsocial toler-
ance in the maternal burrow.

Dispersal from subsocial groups is generally
viewed as simply the endpoint of the tolerant
phase, albeit somewhat delayed in comparison
to non-social species. In this view, dispersal serves
as the dependent variable measuring the re-
sponse of a brood to the events that alter social

behavior. The observations reported here suggest
that the factors that work to initiate dispersal
from subsocial broods may be independent of
those that imply the termination of mutual tol-
erance.

I appreciate the comments of Patricia Miller,
Micky Eubanks, Kari Benson, Chester Figiel,
Craig Hieber, Gail Stratton and an anonymous
reviewer. This work was supported by a Faculty
Development Grant from the Graduate School
of the University of Mississippi.

LITERATURE CITED

Buchli, H. H. R. 1969. Hunting behavior in the Cten-
izidae. American Zool., 9:175-193.

Conley, M. R. 1985. Predation versus resource lim-
itation in survival of adult burrowing wolf spiders
(Araneae: Lycosidae). Oecologia (Berlin), 67:7 1-75.
Krafft, B., A. Horel & J-M. Julita. 1 986. Influence of
food supply on the duration of the gregarious phase
of a maternal-social spider, Coelotes terrestris (Ara-
neae, Agelenidae). J. Arachnol., 14: 219-226.
Lubin, Y. D. & M. H. Robinson. 1982. Dispersal by
swarming in a social spider. Science, 216:319-321.
Miller, G. L. 1984. The influence of microhabitat
and prey availability on burrow establishment of
young Geolycosa turricola (Treat) and G. micanopy
Wallace (Araneae: Lycosidae): A laboratory study.
Psyche, 91:123-132.

Miller, G. L. 1989. Subsocial organization and be-
havior in broods of the obligate burrowing wolf spi-
der Geolycosa turricola (Treat). Canadian J. Zool.,
67:819-824.

Miller, G. L. & P. R. Miller. 1987. Life cycle and
courtship behavior of the burrowing wolf spider
Geolycosa turricola (Treat) (Araneae, Lycosidae). J.
Arachnol., 15:385-394.

Miller, P. R. & G. L. Miller. 1991. Dispersal and
survivorship in a population of Geolycosa turricola
(Araneae, Lycosidae). J. Arachnol., 19:49-54.
Roelofls, R. & S. E. Riechert. 1988. Dispersal and
population-genetic structure of the cooperative spi-
der, Agelena consociata, in west African rainforest.
Evolution, 42:173-183.

Shear, W. A. 1970. The evolution of social phenom-
ena in spiders. Bull. British Arachnol. Soc., 1:65-77.

Gary L. Miller: Department of Biology, the
University of Mississippi, University, Missis-
sippi 38677 USA.

Manuscript received September 1991, revised October
1991.

1991. The Journal of Arachnology 19:229-230

HOMOSEXUAL MATING BEHAVIOR IN MALE
DORYONYCHUS RAPTOR (ARANEAE, TETRAGNATHIDAE)

Homosexual behavior has been documented
in a large number of vertebrates, particularly pri-
mates, in which it plays an important function
in establishing dominance (Crook 1972). In birds,
homosexual behavior appears to be quite com-
mon (Armstrong 1942), and, in certain monog-
amous species, may lead to the establishment of
lesbian relationships (Hunt et al. 1984). Among
invertebrates, homosexual mounting behavior is
found in a number of insects (Kaneshiro & Gid-
dings 1987; Juberthie-Jupeau & Cazals 1989), as
well as phalangids (Bristowe 1929). Male-male
courtship display has been observed in several
spiders in the families Salticidae and Lycosidae
(Bristowe 1929). Homosexual mounting or mat-
ing behaviors, however, have, to my knowledge,
never been documented in spiders.

An elaborate mutual courtship is usually a nec-
essary component of sexual interactions between
spiders, and serves one or a number of functions
(Robinson & Robinson 1980; Suter 1990): (1)
reduce the risk of predation; (2) mutual arousal;
(3) species recognition; and (4) assessment of vir-
ginity. Prior to mating, therefore, courtship will
generally have resulted in mutual communica-
tion between the sexes. This would preclude the
likelihood of attempted copulation between in-
dividuals of the same sex.

One exception in this regard are tetragnathid
spiders, which lack any form of apparent court-
ship (Levi 1981). On encountering each other,
both the male and female tetragnathid appear to
interact combatively, with their chelicerae and
fangs outstretched. If the sexual encounter is suc-
cessful, the fangs of the female become locked
against the spur (apophysis) on the dorsal surface
of the male’s chelicerae. The fangs of the male
are then closed over those of the female, and the
pair are thus securely locked together (the chel-
iceral teeth themselves are not involved in this
locking mechanism). At this point the female
generally curls her abdomen anterioventrally, so
that her seminal receptacles are rotated forward,
thereby facilitating palpal insertion by the male.
The male moves one palp under the abdomen
of the female, then over the surface until it comes
in contact with the seminal receptacles. The oth-

er palp is held almost vertically up above the
carapace (palps are alternated during mating).

On 29th July 1990 I captured several Doryon-
ychus raptor Simon in Waiahuakua Valley (1 100
ft.) on the Hawaiian island of Kauai. These spi-
ders, endemic to Kauai, are robust (female 12-
15 mm, male 10 mm) and readily recognized by
the extremely long claws on the tarsi of the first
two pairs of legs (Simon 1900). Two penultim.ate
males were maintained together in a glass con-
tainer (22.5 cm X 22.5 cm x 25.0 cm) and fed on
laboratory-reared Drosophila grimshawi, molted
to maturity after 5 and 1 1 days respectively.
Twenty five days after capture (23rd August), I
observed these male spiders with their chelicerae
locked together (I did not witness the initial cou-
pling), one male with its fangs locked against the
dorsal apophysis of the other. The abdomens
were paraxial, neither curled under. The only
movement observed was that of the palps of both
males. These were used alternately to “search”
the underside of the abdomen of the respective
partner for 30-210 s before switching to the
alternate palp. This behavior continued for 1 7
minutes, although, because I did not observe the
initial encounter of the pair, the exact duration
of the interaction is unknown. After this time the
spiders disengaged naturally in a manner similar
to disengagement in normal inter-sexual en-
counters in tetragnathids. Neither appeared
harmed by the interaction.

The behavior observed was considered sexual
rather than aggressive. Aggression between in-
dividual tetragnathids (within or between sexes)
involves extension of the fangs and chelicerae,
followed by violent lunging, usually culminating
in the death or retreat of one, and sometimes
both, of the combatants (pers. obs.). Sexual be-
havior between male and female D. raptor is sim-
ilar to that of other tetragnathids (pers. obs.).
Neither participant struggles during a sexual in-
teraction, the only movement being alternation
of the palps and minor adjustments in position.
At the end, separation occurs by release of the
jaws without any lunging.

The observation of homosexual mating be-
havior in D. raptor argues for the absence of

229

230

THE JOURNAL OF ARACHNOLOGY

powers of sexual discrimination by males of this
species. In insects the exhibition of this type of
behavior among males appears to be a conse-
quence of mistaken identity. Among the Ha-
waiian drosophilids, this behavior has been as-
sociated with the absence of powers of sexual
discrimination in males, with female choice
playing the critical role in mating success (Ka-
neshiro & Giddings 1987). Male acceptance is
based on his performance during a complex
courtship display, which involves visual, tactile,
chemical and/or acoustic stimulation.

Among spiders, mistaken identity may explain
the observation of males courting other males
(Bristowe 1929). The potentially lethal effects of
allowing this behavior to continue on to attempt-
ed copulation is likely to have selected for some
form of sexual discrimination among male spi-
ders. It may be, however, that the unique cou-
pling behavior of tetragnathids, where the che-
licerae and fangs are securely locked, has allowed
the loss of male sexual discrimination.

As yet there is no information on female sexual
discrimination in tetragnathids. It may well oc-
cur, however, through the action of either pher-
omones (Tietjen & Rovner 1982), or mechanical
stimulation (Eberhard 1985).

This study was supported by grants from the
Hawaii Bishop Research Institute, the Hawaii
Natural Area Reserves System and the Nature
Conservancy of Hawaii. I am deeply indebted to
C. Parrish for helping me to collect specimens
and to K. Kaneshiro for use of his environmen-
tally controlled facilities to maintain live speci-
mens.

LITERATURE CITED

Armstrong, E. A. 1942. Bird Display: An Introduc-
tion to the Study of Bird Psychology. Cambridge
University Press, London.

Bristowe, W. S. 1929. The mating habits of spiders
with special reference to the problems surrounding
sexual dimorphism. Proc. Zool. Soc. London, 1929:
309-358.

Crook, J.H. 1972. Sexual selection, dimorphism, and
social organization in the primates. Pp. 23 1-28 1 , In
Sexual selection and the Descent of Man, 1871-
1971. (B. G. Campbell, ed.). Aldine, Chicago.

Eberhard, W. G. 1985. Sexual Selection and Animal
Genitalia. Harvard University Press, Cambridge,
Mass.

Hunt, G. L., Jr., A. L. Newman, M. H. Warner, J. C.
Wingfield & J. Kaiwi. 1984. Comparative behav-
ior of male-female and female-female pairs among
Western gulls prior to egg-laying. Condor, 86:157-
162.

Juberthie-Jupeau, L. & M. Cazals. 1989. Agressivite
de deux males en presence d’une femelle conspe-
cifique chez un coleoptere souterrain, Speomomus
delarouzeei Fairm. (Coleoptera, Catopidae). Mem.
Biospeol., Tome XVI: 183-187.

Kaneshiro, K. Y. & L. V. Giddings. 1987. The sig-
nificance of asymmetrical sexual isolation and the
formation of new species. Evolutionary Biology, 21:
29^3.

Levi, H. W. 1981. The American orb-weaver genus
Dolichognatha and Tetragnatha north of Mexico
(Araneae: Araneidae, Tetragnathinae). Bull. Mus.
Comp. Zool. Harvard, 149:271-318.

Robinson, M. H. & B. Robinson. 1980. Comparative
studies of the courtship and mating behavior of
tropical araneid spiders. Pacific Insect Monograph,
36, pp. 203-208.

Simon, E. 1900. Arachnida: Fauna Hawaiiensis, 2:
443-519, pis. 15-19.

Suter, R. B. 1990. Courtship and the assessment of
virginity by male bowl and doily spiders. Anim.
Behav., 39:307-313.

Tietjen, W. J. & J. S. Rovner. 1982. Chemical com-
munication in lycosids and other spiders. Pp. 249-
279, In Spider Communication: Mechanisms and
Ecological Significance. (P. N. Witt & J. S. Rovner,
eds.). Princeton University Press, Princeton, New
Jersey.

Rosemary G, Gillespie, Department of Zoology,
University of Hawaii at Manoa, Honolulu,
Hawaii 96822 USA.

Manuscript received December 1990, revised February
1991.

1991. The Journal of Arachnology 19:231-232

RESPONSE OF NEPHILA CLAVIPES TO MOCK PREDATION
CHANGES WITH THE PROXIMITY OF THE MOLT

Disturbance and predation attempts have been
recognized as being important determinants of
web-site tenacity in orb-weaving spiders (Eber-
hard 1971;Enders 1976). Response to simulated
predation by vibrating the web and dropping from
the web have been documented for a variety of
orb-weavers, including Nephila clavipes (Linn.)
(Araneae: Tetragnathidae) (Tolbert 1975). To in-
vestigate the role of failed predation attempts on
web-site abandonment in N. clavipes, I subjected
juveniles collected in the field to a strong stim-
ulus: a leg was pinched, causing autotomization.

Intact spiders of 0.3 to 0.4 cm leg I tibia +
patella length were collected from Barro Colo-
rado Island and Gigante Peninsula in the Barro
Colorado Island National Monument, Panama.
These spiders were housed in an insectary in the
laboratory clearing on the island, where they spun
webs in 23-25 cm diameter frames consisting of
two 0.5 cm strips of fiberglass fixed at right angles
and hung from one of the points of intersection.
These frames were uncovered, and the spiders
were always at liberty to move within the insec-
tary. Spiders were fed two 6 mm live prey items
daily, usually small moths or Trigona Jurine
stingless bees that were placed in the orb. After
about 10 days in captivity, five spiders were sub-
jected to a mock predator attack involving pinch-
ing the tibia with dissection forceps. Seven or 1 4
days later, the remaining five were attacked.

The response to mock-predation involving leg
autotomization was immediate abandonment of
the web, but not necessarily of the web site. Nine
spiders autotomized either a first or a second leg;
one spider eluded the forceps and escaped the
web. Six spiders abandoned the web site by drop-
ping on a drag-line and spinning an air-bome
bridge thread, and were later found building a
new web within the insectary. The remaining
four hid on the fiberglass web support and later
returned to the original web. Three of these four
spiders molted within five days of the attack
(range 2-4; mean 2.7 days), one molted seven
days after the attack. The spiders that abandoned
their webs were all more than five days away
from the next molt (range 6-8, mean 7.4 days).
The response to the attack was significantly af-
fected by the proximity of the molt (within five

days vs. more than five days to molting: N=10,
likelihood ratio “G” test = 6.43, df= 1 , p=0.005).

The failure of premolt spiders to abandon their
webs after a predation attempt is probably re-
lated to the physiological processes involved in
preparing for molting. Orb-web size declines in
the last four or five days before the molt, and
spiders cease spinning viscid orbs one to three
days before molting (Higgins 1990). About when
spiders cease spinning, the principal major am-
pullate silk glands are reconstructed (M. T ownley
and E. Tillinghast, pers. comm.). These glands
are the source of the orb-web frame lines and
probably also of the barrier- web silk (E. Tilling-
hast, pers. comm.). N. clavipes undergoes ec-
dysis suspended from the dorsal side of the orb
or the dorsal barrier web-hub connection (Hig-
gins 1990). Premolt spiders may have a greatly
reduced or negligible capacity to spin a new web,
and surviving the molt without a web is highly
unlikely.

Special thanks are due to M. Townley and E.
Tillinghast, whose discussions of unpublished
work concerning silk synthesis provided new in-
sight into these results. H. Drummond and J. L.
Osorno read and commented upon the manu-
script.

LITERATURE CITED

Eberhard, W.G. 1971. The ecology of the web of
Uloborus diversus (Araneae, Uloboridae). Oecolo-
gia, 6:328-342.

Enders, F. 1 976. Effects of prey capture, web destruc-
tion and habitat physiogamy on web-site tenacity
of Argiope spiders. J. ArachnoL, 3:75-82.

Higgins, L. E. 1988. Variation in web structure in the
orb-weaving spider Nephila clavipes and correlated
changes in life history. Ph. D. Dissertation. Uni-
versity of Texas at Austin.

Higgins, L. 1990. Variation in foraging investment
during the intermolt and before egg-laying in the
spider Nephila clavipes. i. Insect Behav., 3:773-783.
Horton, C. C. 1982. Predators of two orb-web spiders
(Araneae, Araneidae). J. Arachnol. 1 1: 447-449.
Reichert, S.E. & C. R. Tracy. 1975. Thermal balance
and prey availability: Bases for a model relating
web-site characteristics to spider reproductive suc-
cess. Ecology, 56:265-284.

Tolbert, W. W. 1975. Predator avoidance behaviors
and web defensive structures in the orb weavers

231

232

THE JOURNAL OF ARACHNOLOGY

Argiope aurantia and Argiope trifasciata (Araneae:
Araneidae). Psyche, 82:29-52.

Linden Higgins: Centro de Ecologia Universidad
Nacional Autonoma de Mexico, Apartado

Postal 70-275, Ciudad Universitaria, C.P.
04510 MEXICO. Current address: Depart-
ment of Zoology, University of Texas at Aus-
tin; Austin, Texas 78712 USA

Manuscript received March 1991, revised June 1991.

1991. The Journal of Arachnology 19:233-234

BOOK REVIEW

Dondale, Charles D. & James H. Redner.
1990. The Insects and Arachnids of Canada.
Part 17. The Wolf Spiders, Nurseryweb Spiders,
and Lynx Spiders of Canada and Alaska (Ara-
neae: Lycosidae, Pisauridae, and Oxyopidae).
Agriculture Canada Publication No. 1856. 383
pages. ($20.00 in Canada, $24.00 elsewhere).
Available from Canadian Government Publish-
ing Centre, Supply and Services Canada, Ottawa,
Canada KIA 0S9.

This, the third in a series of identification man-
uals for the spiders of Canada, treats the mem-
bers of the superfamily Lycosoidea, which are
recognized by the unique, grate-shaped form of
the tapeta of the indirect eyes. Included are Ly-
cosidae, with 14 genera and 107 species recorded
or believed to occur in Canada, Pisauridae, with
two genera and seven species, and Oxyopidae,
with two species in the lone genus Oxyopes. The
organization and format follow that of previous
contributions (Dondale & Redner 1978; Don-
dale & Redner 1 982). The introductory and anat-
omy sections are detailed, allowing this volume
to “stand alone,” and there is an extensive glos-
sary. Methodology is admirably explicit. As with
previous volumes, geographic scope is limited to
Canada and Alaska, and toward this end even
previously published figures were remounted and
renumbered, and new maps made providing no
new information but serving only to exclude the
continental United States.

Descriptions are concise, and effective diag-
noses are presented under “Comments.” Biolog-
ical information is provided wherever possible
and, drawing on an extensive bibliography of 273
entries, is comprehensive. Illustrations are many
(596 in all), including dorsal views of the cara-
pace and abdomen for all genera. Male palpi are
illustrated whole in ventral view and details of
the terminal division are supplied; epigyna and
vulvae are illustrated for females of all species.
Representative illustrations are labelled so that
the application of morphological terms is clear.
The illustrations are excellent for species iden-
tification and more than adequate for those who
wish a source of data on the genital morphology
of the taxa involved. Many figures are provided

with unlabelled arrows, which presumably point
out important features discussed in the text. New
keys are provided, in both official languages of
Canada, to genera within families and species
within genera. Keys are detailed with numerous
references to figures, and work well. In some cases
(e.g., Pardosa, Pirata), the new keys are a great
improvement. Given the rather strict geographic
demarcation of the work, utility of the keys ex-
cept in the immediate vicinity of Canada and
Alaska will probably be limited.

There are some minor nomenclatural prob-
lems. Hogna and Varacosa, both previously con-
sidered junior synonyms (the former of Lycosa
and the latter of Trochosa: Platnick 1989), are
treated as valid, though no discussions of their
new status are provided. How is Hogna to be
diagnosed from the European Lycosa, and what
are their relationships? What happened to Ra-
bidosa, which was still a valid genus at last look
(Platnick 1989)? But these are technical points
reflecting validity (a scientific decision), which is
beyond the scope of an identification manual,
and as an identification manual this work suc-
ceeds admirably.

A review of a work of this nature would be
incomplete without consideration of the pros and
cons of such regional faunal studies. More to the
point, in view of the American Arachnological
Society’s endorsement of the proposal for a biotic
survey of the United States (Kosztarab 1988), a
proposal that is slowly but inexorably making its
way toward realization, all readers of the Journal
of Arachnology should take time to consider
whether the scarce resources available for sys-
tematic biology are best utilized to produce re-
gional “faunas” of this kind. Whereas stated ben-
efits of regional surveys (e.g., Kosztarab 1988)
run the gamut from providing baseline data nec-
essary for monitoring environmental quality to
enhancing national security (!), three arguments
state the case forcefully: 1 . they provide widely
available keys and means for identification that
are useful to land-use planners and biologists of
all persuasions, specialists and novices alike; 2.
insofar as they accurately reflect the taxonomy
and distribution of species treated, they offer a
baseline for monitoring environmental changes,
and may provide data on endemism and poten-

233

234

THE JOURNAL OF ARACHNOLOGY

tial endangered status; and 3. regional emphasis
leads to decentralization, which appeals to leg-
islators and makes such studies potentially fund-
able (pork barrel systematics). These are not ar-
guments to be dismissed lightly! On the other
hand, arguments against the regional approach
are many (see especially Liebherr 1989; Pakaluk
& Wahl 1989). Regional studies generally offer
an incomplete treatment of natural groups or ar-
eas; and distributional data, while accurate for
the region treated, may not reflect the whole pic-
ture. Students participating in such studies are
often ill-prepared to compete for jobs, grants,
and tenure. Resources are focussed on countries
relatively rich in money (and poor in biodiver-
sity) while monetarily poor (and diversity-rich)
countries are neglected. Finally, regional studies
perpetuate the stereotype that systematics con-
sists largely of naming species, rather than its
more important contribution of a phylogenetic
context within which comparative biology be-
comes meaningful, and they divert scarce re-
sources from the latter pursuit.

In many ways this work represents a “best
case” scenario for a regional study. Dondale and
Redner have published six up-to-date mono-
graphs of North American Lycosidae which, when
added to Brady’s work on lycosids and oxyopids
and Carico’s work on pisaurids, provides the
sound monographic taxonomy necessary to un-
derpin such a regional study. The first author has
also produced an exemplary study of lycosid
higher classification (Dondale 1986). In view of
the quality and scope of that monographic work,
one may lament that Agriculture Canada BRC
has mandated that their researchers contribute
to this national series, and reflect that the con-
siderable talents and resources herein displayed
might have been better utilized to finish mono-
graphing the Lycosidae of North America rather
than to prepare this handsome but largely re-
dundant volume.

Needless to say, as an identification manual
this work is superior, and it will be indispensable
to any student of the terrestrial arthropods of
Canada and Alaska who has no access to the
primary literature.

LITERATURE CITED

Dondale, C. D. 1 986. The subfamilies of wolf spiders
(Araneae: Lycosidae). Adas X Congreso de Arach-
nologia, Jaca, Espana. 1:327-332.

Dondale, C. D. & J. H. Redner. 1978. The Insects
and Arachnids of Canada. Part 5. The Crab Spiders
of Canada and Alaska (Araneae: Philodromidae and
Thomisidae). Agriculture Canada Publication No.
1663. 255 pages.

Dondale, C. D. & J. H. Redner. 1982. The Insects
and Arachnids of Canada. Part 9. The Sac Spiders
of Canada and Alaska (Araneae: Clubionidae and
Anyphaenidae). Agriculture Canada Publication No.
1724. 194 pages.

Kosztarab, M. 1988. Biological Diversity: National
Biological Survey. Pp. 1-25. In: Proceedings of the
first annual symposium on the natural history of
lower Tennessee and Cumberland river valleys.
Clarksville, Tennessee. (Snyder, D., ed.). The Center
for Field Biology of Land Between The Lakes, Aus-
tin Peay State University.

Liebherr, J. K. 1989. An Open Letter Regarding Re-
cent Proposals for Nationalistic Entomological Sys-
tematics. Insect Collection News, 2:10-1 1.

Pakaluk, J. & D. B. Wahl. 1989. Systematic Ento-
mology and Biological Diversity. Insect Collection
News, 2:7-8.

Platnick, N. 1. 1 989. Advances in Spider Taxonomy,
1 98 1-1987; a supplement to Brignoli’s ‘A Catalogue
of the Araneae described between 1940 and 1981’.
Manchester University Press, Manchester, England.
673 pages.

Charles E. Griswold: California Academy Sci.,
Golden Gate Park, San Francisco, California
94118 USA

Manuscript received November 1991.

1991. The Journal of Arachnology 19:235-236

BOOK REVIEW

Izmailova, M. V. 1989. Fauna of Spiders of
Southeastern Siberia (In Russian). Irkutsk State
University, Irkutsk, USSR. 184 pp., figs.

This is a first book written about Siberian spi-
ders. It includes data on 341 species belonging
to 2 1 families from the southeastern Siberia. Data
are based on the author’s collections from 1967
to 1981 in the southern part of the Irkutsk Re-
gion, the Krasnoyarsk Region (Boguchansk Dis-
trict), the Buryatia and the western Chita Region.
Eighty species are listed only from the literature
and were not collected by the author. The chap-
ters include: the history of araneological research
in Siberia; physiographic description of East Si-
beria (very extensive and never referred to in
later chapters), faunal list with localities (pictures
of genitalia and synonymies are given for selected
species), data on habitat distribution and eco-
logical characteristics (for coniferous and mixed
forests, shrubs, swamps and rocky habitats) and
zoogeography.

The last monographic book on spiders from
the Soviet Union was Spiders of Tadjikistan by
E. M. Andreeva (1976). Thus, a serious treat-
ment of the regional Siberian fauna would have
been welcomed. However, this book is a disap-
pointment. It contains many errors and under-
represents the current knowledge of Siberian spi-
ders. Moreover, it does not adhere to the com-
mon standards for faunistic publications on spi-
ders.

The chapters of this book dealing with system-
atics are full of mistakes. Besides numerous mis-
spellings of Latin names, the author ignores (or
is not aware of) recent changes in taxonomy. She
lists many linyphiid species under their old ge-
neric names (e.g., Simula flavescens, instead of
Maro f. , Mengea warburtoni instead of Allo-
mengea w.). Synonyms Cornicularia karpinskii
and Wideria k. are listed as two different species.
Many generic names in the Araneidae and some
in the Theridiidae and Salticidae are outdated.
Junior synonyms of many species are not listed.
Although the author likely had at her disposal
very limited reference sources, Russian arach-

nologists often exchange information. Thus, there
appears to be little reason not to check all syn-
onymies and update references.

Of the 226 spider species collected or identified
by the author, 121 are represented only by one
or two adult specimens. Ten species are repre-
sented only by juveniles. Surprisingly, only one
drawing, that of Sitticus finschi, shows the gen-
italia for both males and females. All other 166
pictures show either palp or epigyne. Moreover,
judging from these pictures, many species are
misidentified; e.g., Pisaura mirabilis should be
P. ancora, Araneus grossus should be Aculepeira
carbonarioides, Zelotes subterraneans should be
Z. fratris, and many others. Synonymy of some
species is not checked, and they are listed twice,
under both valid name and a junior synonym.
Acantholycosa norvegica is listed the second time
as A.fedotovi; Alopecosa sibirica - as A. pinnata\
Alopecosa solivaga - as A. poecila\ Steatoda bi-
punctata - as Lithyphanthes corollatus, etc. Many
species are listed under the names that became
junior synonyms long ago. In one case, Izmailova
discovers a new homonymy (Gnaphosa punctata
Kulczynski and G. punctata Tullgren) but does
not discuss it and does not give a new name to
the junior homonym.

Two new species are described by M. Izmai-
lova: Alopecosa litvinovi and Pardosa “sp.n.”; the
latter one is not given any name. In both de-
scriptions, no diagnosis is provided, holotype
specimens are not designated, and the place of
their deposit is not given.

The distribution information for more than 70
species is incorrect (e. g., circum-Holarctic spe-
cies Gnaphosa borea is referred to as “endemic
of East Siberia”). Listed as “endemics of the Asi-
atic USSR” are G. borea, Clubiona interjecta,
Xysticus lectus {=X. britcheri), X. transsibiricus
(=X ephippiatus), Linyphia tridens {=Estrandia
grandeva)-, however, these spiders are found also
in North America and/or China, Mongolia and
Japan. Among the “first records for the USSR”
are Xysticus britcheri, Gnaphosa chajfanjoni,
Haplodrassus moderatus, Evarcha albaria-a.\\ of
which were recorded for the USSR before.

235

236

THE JOURNAL OF ARACHNOLOGY

Unfortunately, this book cannot serve as a
guide to Siberian spiders. It is outdated in ref-
erences and synonymy, and it has numerous mis-
takes.

Victor Fet: Department of Biological Sciences,
Loyola University, New Orleans, Louisiana
70118, USA.

Yuri Marusik: Institute of the Biological Prob-
lems of the North, Russian Academy of Sci- Manuscript received June 1991. revised August 1991.
ences, Magadan 685010, Russia.

INSTRUCTIONS TO AUTHORS

Manuscripts are preferred in English but may be accepted
in Spanish, French or Portuguese subject to availability of
appropriate reviewers. Authors whose primary language is not
English, may consult the Associate Editor for assistance in
obtaining help with English manuscript preparation. All man-
uscripts should be prepared in general accordance with the
current edition of the Council of Biological Editors (CBE) Style
Manual unless instructed otherwise below. Authors are ad-
vised to consult a recent issue of the Journal of Arachnology
for additional points of style. Send three identical copies of
the typed material together with copies of illustrations to:

Gary L. Miller, Associate Editor
The Journal of Arachnology
Department of Biology
The University of Mississippi
University, Mississippi 38677 USA
Telephone: (601) 232-7495
FAX: (601) 232-5144
Bitnet: BYMILLER@UMSVM
Correspondence relating to manuscripts should also be di-
rected to the Associate Editor.

Manuscripts longer than 1 500 words should be prepared as
Feature Articles, shorter papers as Research Notes.

FEATURE ARTICLES

Title page.— Place in the upper left comer the complete name,
address, and telephone number of the author with whom proofs
and correspondence should be exchanged. Include FAX num-
bers and Bitnet addresses if available. Four lines below, type
the title, centered, all in capital letters. Another four lines be-
low, beginning with the left margin, type each author’s name
on a new line followed by the address.

Abstract.— The heading should be placed at the beginning
of the first paragraph set off by a period. In articles written in
English, a second abstract in an acceptable language may be
included pertinent to the nationality of the authorfs) or geo-
graphic region(s) emphasized in the article. A second abstract
in English must be included in articles not written in the latter
language.

Text.— Double-space text, tables, legends, etc. throughout.
Three levels of heads are used. The first level (METHODS,
RESULTS, etc.) is typed in capitals and on a separate line.
The second level head begins a paragraph with an indent and
is separated from the text by a period and a dash. The third
level may or may not begin a paragraph but is underscored
and separated from the text by a colon. Use only the underline
method to indicate italics; DO NOT USE WORD PROCES-
SOR ITALICS CODES. Use the metric system unless quoting
text or referencing collection data. All decimal fractions are
indicated by the period regardless of language of the text.

Citation of references in the text.— Include within parenthe-
ses the surnames of the author(s) followed by the date of pub-
lication. A comma separates multiple citations by the same
author(s) and a semicolon separates citations by different au-
thors, e.g., (Smith 1970), (Jones 1988; Smith 1989), (Smith
1986, 1987; Smith & Jones 1989; Jones et al. 1989).

Literature cited section.— Use the following style:
Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid
composition of major ampullate gland silk (dragline) of Ne-
phila clavipes (Araneae, Tetragnathidae). J. ArachnoL, 18:
297-306.

Krafft, B. 1982. The significance and complexity of com-
munication in spiders. Pp. 1 5-66, In Spider Communica-
tions: Mechanisms and Ecological Significance. (P. N. Witt
& J. S. Rovner, eds.). PrincetonUniversity Press, Princeton.
Footnotes.— Footnotes are only permitted on the first printed
page to indicate current address or other information con-
cerning the author(s). All footnotes are placed together on a
separate manuscript page.

Running head.— Author sumame(s) and an abbreviated title
must be typed all in caps and must not exceed 60 characters
and spaces. The running head is placed by itself on a separate
manuscript page.

Taxonomic articles.— Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Associate
Editor.

Tables.— Each table, with the legend above, should be placed
on a separate manuscript page. Only the appropriate horizontal
lines should be included. Use no footnotes; instead, include
all information in the legend. Make notations in the text mar-
gins to indicate the preferred location of tables in the printed
text.

Illustrations.— Original art work should be kept by the au-
thor(s) until final acceptance for publication. All art work must
be mounted on stiff board ready for reproduction. In line draw-
ings, pay particular attention to width of lines and size of
lettering when reductions are to be made by the printer. Mul-
tiple photos assembled on a single plate should be mounted
with only a minimum of space separating them. In the case of
multiple illustrations mounted together, each illustration must
be numbered sequentially rather than given an alphabetic se-
quence. Written on the back should be the name(s) ofauthor(s),
an indication of top edge, and whether the art work should be
returned. The overall dimensions should be no more than 1 1
inches (28 cm) x 14 inches (36 cm). Larger drawings present
greater difficulty in shipping and greater risks of damage for
which the JOA assumes no responsibility. The author is en-
couraged to consider PMT (PhotoMechanical Transfer) repro-
ductions for large drawings and to consult an Editor for details
on appropriate reductions. Make notations in the text margins
to indicate the preferred position of illustrations in the printed
text.

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Adhere to the
following styles:

Figures \-^.—A-us x-us, male from Timbuktu: 1, left leg;
2, right chelicera; 3, dorsal aspect of genitalia; 4, ventral aspect
of abdomen.

Figures 27-34.— Right chelicerae of species of A-us from
Timbuktu: 27, 29, 3 1 , 33, dorsal views; 28, 30, 32, 34, prolater-
al views of moveable finger, 27, 28, A-us x-us, holotype male;
29, 30, A-us, w-us, holotype male; 31, 32, A-us z-us, holotyjje
male; 33, 34, A-us y-us, male. Scales = 1.0 mm.

Assemble manuscript for mailing.— Assemble the separate
sections or pages in the following sequence; title page, abstract,
text, figure legends, footnotes, running head, tables with leg-
ends, figures.

Page charges and reprints.— The current charge per journal
page will be assessed as follows: $50— nonmembers, $45 —
members with grant, institutional, or other support for page
costs; $25 —regular members who do not have support for page
costs; $20— student members who do not author with full
members and who do not have support. Reprints are available
only from the printer at current costs.

Waiver of Page Charges.— The Editor will consider requests
for a waiver of page charges for manuscripts which have been
accepted for publication. Requests should be made in writing
and should be submitted to the Associate Editor along with
the final revision of the manuscript. The author must certify
that neither grant nor institutional funds are available to cover
all of the publication costs. The Society will normally agree to
waive up to a maximum of 80% of the page charges, contingent
on the availability of funds; the author is expected to find
alternative sources for the remainder.

RESEARCH NOTES

Instructions above pertaining to feature articles apply also
to research notes except that abstracts and most headings are
not used and the byline follows the Literature Cited section.

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 19 Feature Articles NUMBER 3

Segregation studies of isozyme variation in Metaphidippus galathea (Ara-

neae, Salticidae), William W. M. Steiner and Matthew H. Greenstone 157

Two new species of Nesticus spiders from the southern Appalachians (Ara-
neae, Nesticidae), Frederick A. Coyle and Augustus C. McGarity ... 161

Evidence for idiothetically controlled turns and extraocular photoreception

in lycosid spiders, Jerome S. Rovner 169

Hawaiian spiders of the genus Tetragnatha: I. Spiny leg clade, Rosemary

G. Gillespie 174

Mitochondrial DNA sequences coding for a portion of the RNA of the small
ribosomal subunits of Tetragnatha mandibulata and Tetragnatha
hawaiensis (Araneae, Tetragnathidae), Henrietta B. Croom, Rosemary
G. Gillespie and Stephen R. Palumbi 210

Segmental anomalies in Roncus aff. lubricus (Neobisiidae, Pseudoscor-
piones) from Yugoslavia, B. P. M. Curcic, R. N. Dimitrijevic, O. S.
Karamata and L. R. Lucic 215

Research Notes

An example of abnormal carapaco-abdominal fusion in Neobisium aff.
fuscimanum (Arachnida, Pseudoscorpiones, Neobisiidae), Bozidar P.

M. Curcic and Rajko N. Dimitrijevic 225

In vitro post-dispersal burrow sharing among spiderling Geolycosa turricola

(Araneae, Lycosidae), Gary L. Miller 227

Homosexual mating behavior in male Doryonychus raptor (Araneae,

Tetragnathidae), Rosemary G. Gillespie 229

Response of Nephila clavipes to mock predation changes with the proximity

of the molt. Linden Higgins 231

Book Reviews

The Insects and Arachnids of Canada. Part 17. The Wolf Spiders, Nursery-

web Spiders, and Lynx Spiders of Canada and Alaska (by Charles D.
Dondale and James H. Redner), Charles E. Griswold 233

Fauna of Spiders of Southeastern Siberia (by M. V. Izmailova), Yuri Marusik

and Victor Fet 235

a

The Journal of
ARACHNOLOGY

\

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 20

■^^\THS0A///q7f\

I dUG ? 6 IVj- ^ j

1992 number 1

THE JOURNAL OF ARACHNOLOGY

EDITOR: James W. Berry, Butler University

ASSOCIATE EDITOR: Gary L. Miller, The University of Mississippi

EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E.
Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C.
Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ.; C. D. Don-
dale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galiano,
Mus. Argentino de Ciencias Naturales; M. H. Greenstone, BCIRL, Columbia,
Missouri; N. V. Horner, Midwestern State Univ.; D. T. Jennings, Garland, Maine;
V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ.; E. A. Maury, Mus.
Argentino de Ciencias Naturales; N. L Platnick, American Mus. Natural Hist.;
G. A. Polls, Vanderbilt Univ.; S. E. Riechert, Univ. Tennessee; A. L. Rypstra,
Miami Univ., Ohio; M. H. Robinson, U.S. National Zool. Park; W. A. Shear,
Hampden-Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa
Rica.

The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the
study of Arachnida, is published three times each year by The American Arach-
nological Society. Memberships (yearly): Membership is open to all those inter-
ested in Arachnida. Subscriptions to The Journal of Arachnology and American
Arachnology (the newsletter), and annual meeting notices, are included with mem-
bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or
$90 (all other countries). Inquiries should be directed to the Membership Secretary
(see below). Back Issues: Susan E. Riechert, Department of Zoology, Univ. of
Tennessee, Knoxville, TN 37916 USA. Undelivered Issues: Allen Press, Inc.,
1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: Allen R. Brady (1991-1993), Biology Department, Hope College,
Holland, Michigan 49423 USA.

PRESIDENT-ELECT: James E. Carico (1991-1993), Department of Biology,
Lynchburg College, Lynchburg, Virginia 24501 USA.

MEMBERSHIP SECRETARY: Norman 1. Platnick (appointed), American
Museum of Natural History, Central Park West at 79th St., New York, New
York 10024 USA.

TREASURER: Gail E. Stratton (1991-1993), Department of Biology, Albion
College, Albion, Michigan 49224 USA.

SECRETARY: Brent Opell (1991-1993), Department of Biology, Virginia Poly-
technic Institute and State University, Blacksburg, Virginia 2406 1 USA.
ARCHIVIST: Vincent D. Roth, Box 136, Portal, Arizona 85632 USA.
DIRECTORS: Matthew H. Greenstone (1990-1992), George W. Uetz (1991-
1993), Charles E. Griswold (1991-1993).

HONORARY MEMBERS: C. D. Dondale, W. J. Gertsch, H. Homann, H. W.
Levi, A. F. Millidge, M. Vachon, T. Yaginuma.

Cover illustration: SEM photomicrograph of the ocularium of the opilionid Odiellus pictus (Wood).
The species has a prominent trident of spines at the anterior border of its cephalothorax. Found in
the eastern United States. Photo by Steven Murphree.

Publication date: 14 August 1992

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.

1992. The Journal of Arachnology 20:1-17

HAWAIIAN SPIDERS OF THE GENUS TETRAGNATHA II.
SPECIES FROM NATURAL AREAS OF WINDWARD EAST MAUI

Rosemary G. Gillespie: Department of Zoology and Hawaiian Evolutionary Biology
Program, University of Hawaii, Honolulu, Hawaii 96822 USA

ABSTRACT: The spider genus Tetragnatha is highly speciose in the Hawaiian Islands, diverse in morphology,
ecology and behavior. The present study describes the distribution of 1 3 species of the genus from natural areas
on the windward northern and eastern sections of Haleakala volcano, Maui, primarily in the Nature Conservancy
of Hawaii’s Waikamoi Preserve and Haleakala National Park. Six species do not build webs, and have recently
been described; the remainder all build webs. The description of six new web-building species, T. thtuberculata
n. sp.,r. eurychasma n. sp., T. acuta n. sp., T. filiciphilia n. sp., T. stelarobusta n. sp., T. paludicola n. sp., is
the primary focus of this paper.

Spiders are one of the primary predatory groups
in Hawaiian ecosystems. Systematic studies on
the group, however, are very limited (Karsch
1880; Simon 1900; Suman 1964, 1970; Okuma
1988). The most comprehensive work was that
of Simon (1900), who worked on a small collec-
tion of Hawaiian spiders made by R. C. L. Per-
kins (Perkins 1913). Of all spider groups repre-
sented in the Hawaiian Islands, those of the genus
Tetragnatha are the most conspicuous, and per-
haps also the most widespread.

Outside Hawaii, representatives of the genus
Tetragnatha are among the more homogeneous
of spider genera in both morphology (elongate
bodies and legs, and large chelicerae and endites
[Kaston 1948; Levi 1981]) and ecology (Da-
browska Prot & Luczak 1968a, b; Dabrowska
Prot et al. 1968). In Hawaii, however, the highly
speciose genus is diverse in morphology, ecology
and behavior. Preliminary morphological and
molecular phylogenetic analyses (Croom, Gilles-
pie & Palumbi in prep.) suggest that there are
distinct clades of Hawaiian tetragnathids, each
with its own unique set of characteristics.

This paper documents 13 representatives of
the genus from two natural areas in the windward
northern and eastern sections of the east Maui
volcano, Haleakala: the Nature Conservancy of
Hawaii’s Waikamoi Preserve and Haleakala Na-
tional Park, the two areas abutting each other to
form an almost continuous swathe of native for-
est. The current systematic treatment is intended
to allow future publications on research I have
conducted on the ecology and behavior of the
species in these areas. Two additional sites on
windward Haleakala (Pohakuokala Gulch to the

west and Hanawi Valley to the NE) were sur-
veyed to determine the distribution of the 13
species across the mountain.

Haleakala National Park comprises a broad
strip of land (1 1,400 ha) running from sea level
in the east to the summit of Haleakala (3057 m)
in the west, part of its western edge bordering
the Waikamoi Preserve. The Preserve (2117 ha)
continues west (NW) from this border (2653 m)
running down to 1 340 m. Average annual rainfall
is generally high, increasing along a steep west-
to-east gradient from 2000 mm to 5000 mm,
with some areas exceeding 10,000 mm. The site
surveyed in Hanawi Valley, which lies to the
north of the National Park, was at 463 m. Po-
hakuokala Gulch, to the west of the National
Park, was surveyed at 1524 m.

The vegetation in the National Park changes
from disturbed Koa/’Ohi’a {Acacia koa/Metros-
ideros polymorpha) lowland wet forest in the east,
up through more pristine Koa/’Ohi’a stands to
’Ohi’a montane wet forest interspersed with
montane bogs in the west (Wagner et al, 1990;
Medeiros, pers. comm.). In the Preserve, the veg-
etation changes from the ’Ohi’a montane wet
forest at the border with the National Park, to
Koa/’Ohi’a montane mesic forest in the west.
The site examined at Hanawi is disturbed ’Ohi’a
lowland wet forest, while Pohakuokala is dis-
turbed Koa montane mesic forest. The dominant
plants vary according to forest type, but the most
common tree species are A. koa and M. poly-
morpha, as well as Clermontia arborescens. Ilex
anomala, Cheirodendron trigymim, Myrsine spp.
and Pelea spp. A number of species of ferns, in
particular Cibotium spp. dominate the understo-

1

2

THE JOURNAL OF ARACHNOLOGY

ry, along with Vaccinium calycinum, Broussaisia
arguta, Rubus hawaiensis and Alyxia oliviformis.

Of the 13 species of Tetragnatha found in the
Waikamoi Preserve and Haleakala National Park,
six species do not build webs, and are considered
in the Spiny Leg Clade of Hawaiian Tetragnatha,
which has recently been described (Gillespie
1991). The other species form a diverse group
of web-builders from an unknown number of
clades, the description of which is the main focus
of this paper. Five of these species are described
from specimens collected in the Waikamoi Pre-
serve. T. paludicola is described from specimens
collected from the bogs on the north east rift of
Haleakala (1676 m) in Haleakala National Park.

METHODS

Specimens were examined for both gross mor-
phological features as well as for more detailed
structure in the same manner as that used for
other Hawaiian Tetragnatha (Gillespie 1991). I
followed the terminology for cheliceral armature
used by Okuma (1987, 1988). In males, the teeth
on the promargin generally include ‘Gu’, a small
distal tubercle; ‘sP, the first major tooth; ‘T’, the
second (usually larger) tooth; and ‘rsu’, the re-
maining proximal teeth on the promargin. The
teeth on the retromargin generally include ‘AXF,
a small distal tubercle; ‘GF, the first major tooth,
‘L2’ the second ‘L3’ the third etc. ‘a’ is the dorsal
cheliceral spur. For females, the cheliceral teeth
are numbered from the distal end ‘Ul’— ‘U«’ on
the promargin and ‘LI’ — ‘L«’ on the retromar-
gin.

NON-WEB-BUILDING SPECIES
(SPINY LEG CLADE)

Tetragnatha brevignatha Gillespie

Tetragnatha brevignatha, a member of the
Green Spiny Leg group in the Spiny Leg clade,
was found only in a small section of mid-ele-
vation (1340 m) mesic forest on northern Ha-
leakala, in the NW comer of the Waikamoi Pre-
serve (Table 1).

Tetragnatha waikamoi Gillespie

Tetragnatha waikamoi, a second member of
the Green Spiny Leg group, was found only in
montane wet forest of northern Haleakala from
1310 m to 1876 m (Table 1). It was therefore
abundant in the more northern Waikamoi Pre-

serve, the only other place it was found being the
bogs on the NE Rift of Haleakala in the National
Park. To the west, the range of this species over-
laps with T. brevignatha in a very narrow zone;
to the east, the range comes close to that of T.
macracantha, but no overlap zone has yet been
found.

Tetragnatha macracantha Gillespie

Tetragnatha macracantha, the final member
of the Green Spiny Leg group in this region, was
found throughout the Kipahulu Valley of Hale-
akala National Park, from the lowest (610 m) to
the highest (1980 m) elevations (Table 1). In ad-
dition, it was found in the lowland disturbed
forest of Hanawi at 463 m. As mentioned, its
range comes close to, but has not been found to
overlap, that of T. waikamoi.

Tetragnatha kamakou Gillespie

Tetragnatha kamakou, a member of the Green
and Red Spiny Leg group in the Spiny Leg clade,
was found throughout montane wet forest of Ha-
leakala National Park and the Waikamoi Pre-
serve from 610 m to 1980 m (Table 1).

Tetragnatha quasimodo Gillespie

Tetragnatha quasimodo was found in abun-
dance in both mesic and wet forests from the
lowest (610 m) to the highest (1980 m) elevations
in Haleakala National Park and the Waikamoi
Preserve, as well as in Hanawi Valley and Po-
hakuokala Gulch (Table 1).

Tetragnatha restricta Simon

Tetragnatha restricta was found in mesic forest
at all elevations (610 m to 1524 m) in Haleakala
National Park, the Waikamoi Preserve and Po-
hakuokala Gulch (Table 1).

WEB-BUILDING SPECIES

Tetragnatha olindana Karsch

T. olindana was found only in low elevation
(610 m) wet forest, in Hanawi and the Kipahulu
Valley of Haleakala National Park (Table 1).

Tetragnatha trituberculata, new species
(Figs. 1-14, 85)

Types. — Holotype male and allotype female
from Waikamoi Gulch, Waikamoi, 1310 m, Maui
Island (7 January 1991), collected by D. J. Pres-
ton, deposited in the Bishop Museum, Honolulu.

GILLESPIE- HAWAIIAN TETR.4GNATHA

3

Table 1. — Distribution of species in the natural areas of windward east Maui. Locations are ordered from
west (Pohakuokala, 1524 m) to east (Kipahulu Valley at 610 m).

Poha-

Waikamoi Preserve

Hale.

Nat.

Pk.

N.E.

Haleakala National Park

Elevation (m)

kala

Olinda

ruthers

manu

Bogs

Hanawi

Kipahulu Valley

1524

1340

1876

1585

1676

463

1980

1524

1220

914

610

T. waikamoi

Male

1

4

2

1

Fern

8

13

3

11

Imm

9

10

20

8

T. brevignatha

Male

15

Fern

12

Imm

8

T. macracantha

Male

2

2

1

4

6

16

Fern

4

3

2

2

7

14

Imm

13

2

11

29

45

31

T. kamakou

Male

1

4

4

2

1

Fern

5

9

3

9

1

2

Imm

5

30

7

20

13

3

1

T. quasimodo

Male

10

5

1

2

4

Fern

3

5

9

3

1

1

7

3

4

10

6

Imm

6

12

20

2

2

3

4

4

18

14

T. restricta

Male

1

3

3

Fern

1

6

2

1

Imm

2

3

4

T. olindana

Male

4

1

2

2

Fern

5

1

12

7

Imm

35

2

25

10

T. trituber-

Male

1

6

1

1

culata

Fern

1

1

3

6

2

3

Imm

5

2

4

2

9

T. eurychasma

Male

1

7

2

2

1

1

1

Fern

1

5

2

4

3

8

4

2

3

Imm

1

8

9

6

5

4

1

3

T. acuta

Male

2

1

Fern

1

1

8

1

1

Imm

6

2

T. filiciphilia

Male

7

1

Fern

11

1

2

Imm

8

6

2

T. stelarobusta

Male

6

6

1

1

Fern

11

6

3

4

Imm

10

6

4

2

T. paludicola

Male

5

2

Fern

16

5

5

Imm

2

10

17

2

4

THE JOURNAL OF ARACHNOLOGY

Figures \-\A.— Tetragnatha thtuherculata\ Male holotype. 1) Promargin of right chelicera; 2) Retromargin of
left chelicera; 3) Dorsal spur of right chelicera, lateral view; 4) carapace, dorsal view; 5) Right leg I, dorsal view;
6) Right leg III, prolateral view; 7) Left palpus, prolateral view. Female allotype. 8) Promargin of right chelicera;
9) Retromargin of left chelicera; 10) Carapace, dorsal view; 11) Right leg I, dorsal view; 12) Right leg III,
prolateral view; 13) abdomen, dorsal (a) and lateral (b) views; 14) Seminal receptacles, ventral view. Scale lines
in mm. Scale of Figs. 1-3 indicated below 1; scale of Figs. 4, 8, 9, 10 indicated below 8; scale of Figs. 5, 6, 11,
12 indicated below 12.

GILLESPIE- HAWAIIAN TETRAGNATHA

5

Etymology.— Tri (Greek) three; tuberculum
(Latin) tubercle. The specific epithet is used in
its adjectival form and refers to the transverse
procurved row of tubercles across the abdomen
of this species.

Diagnosis. — T. trituberculata is not easily con-
fused with any other species. The most diagnostic
feature is the series of transverse abdominal lobes,
accentuated by the distinctive black pattern. Even
where the abdominal tubercles are reduced (as
in mature males), the pattern is highly diagnostic,
and does not appear to fade in alcohol.

Description.— //o/oiyppwa/e’.’ (Figs. 1-7). Pro-
margin of chelicerae (Fig. 1): Distance between
‘Gu’, ‘sf and ‘T’ approximately equal, ratio of
distal end to ‘sF: ‘sF to ‘T’: ‘T’ to ‘rsul’ 4:3:3.
‘Gu’ pronounced, small, cone-shaped tubercle;
‘sF medium-sized cone directed out perpendic-
ular to margin of chelicerae; narrower than ‘T’
by 90% (70-90%), and shorter, 60% height (50-
60%). ‘T’ moderately tall, robust, rocket-shaped,
‘rsu’ 4 (3-4) spikes, ‘rsul’ and ‘rsu2’ diverging
slightly along vertical plane. Retromargin of che-
licerae (Fig. 2): Total of only 4 (up to 7) teeth.
‘AXF small, but conspicuous triangular notch;
‘GF strong, similar width and height to ‘L3’, ‘L4’
and ‘L5’, much stronger than ‘L2’. ‘L2’ set farther
back into fang groove than other teeth on retro-
margin. Dorsal spur long, shaped like slim, bent
finger ( 1 4.0% length of cephalothorax); tip slight-
ly longer on dorsal surface (Fig. 3). Cheliceral
fang slightly shorter than base, bent sharply at
both proximal and distal ends. Cephlothorax 2.5
mm, total length 5.6 mm. Chelicerae shorter
(84%) than cephalothorax. Depression of tho-
racic fovea distinctly marked with dark lines ra-
diating out from center (Fig. 4). Leg spination
similar to female (Figs. 5, 6). Femur I: 6 prola-
teral, 5 dorsal, 6 retrolateral spines. Tibia I: 3
prolateral, 2 dorsal, 3 retrolateral spines. Meta-
tarsus 1: 1 prolateral, 1 dorsal, 2 retrolateral spines.
Femur III: 4 dorsal, 1 prolateral, no ventral spines.
Tibia III: 1 dorsal, 1 prolateral spine. Coloration
and eye pattern similar to female.

Conductor Tip: (Figs. 7 and 85). Smoothly
rounded, almost symmetrical, high-peaked cap,
terminating in small, downward-pointing, beak-
like tip.

Allotype female: (Figs. 8-14). Eye area heavily
pigmented, distance between PME smaller than
eye area itself (Fig. 1 0). Median ocular area wider
posteriorly. Lateral eyes loosely contiguous. Pro-
margin of chelicerae (Fig. 8): series of 7 teeth,

‘Ur very robust, considerably wider but shorter
(66%) than ‘U2’; separated from ‘U2’ by 13%
cheliceral length. ‘U2’-‘U7’ gradually decreasing
in size proximally. Retromargin of chelicerae (Fig.
9): series of 5 teeth, ‘LI’ slightly shorter (90%)
than ‘Ur, much smaller (69%) than ‘L2’. ‘LI’
contiguous with ‘L2’, teeth decreasing in size
proximally. Cheliceral fang short, approximately
77% length of base, tapering to smooth point at
distal end. Cephlothorax 2.9 mm, total length
7.6 mm. Chelicerae rather short, 60% length of
cephalothorax. Legs heavily spotted, banded with
dark brown (bottle green in life) on pale cream
(Figs. 1 1-12). Spines short (19% length of ceph-
alothorax) but robust. Femur I: 6 prolateral, 5
dorsal, 5 retrolateral spines. Tibia I: 3 prolateral,
2 dorsal, 3 retrolateral spines. Metatarsus I: 1
dorsal, 2 retrolateral spines. Femur III: 3 dorsal,

1 prolateral, no ventral spines. Tibia III: 1 dorsal
spine, 1 prolateral spine. Cephlothorax pale
brown, with a distinct fovea marked by very dark
lines radiating out from center (Fig. 1 0). Sternum
dusky black. Abdomen broad, deep, width and
depth both approximately 35% of total length.
Dorsum of abdomen green/brown (bright bottle-
green in life), with distinct black markings (Fig.
13a). Transverse procurved row of three tuber-
cles on abdomen, each accentuated by black
marks bordering all except median border in lat-
eral tubercles, and all sides (except for narrow
proximal and distal “window”) on medial tu-
bercle (Fig. 13b). Venter dark brown/black with

2 pairs of gold vertical bars on either side of
midline.

Seminal receptacles: (Fig. 14). Two bulbs linked
in tight opposing “comma” shapes, each well
sclerotized on medial border. Both bulbs equally
dilated. Central portion a robust stalk between
bulbs. Median lobe an angular balloon projecting
from stalk, fitting snugly sandwiched between
bulbs.

Color polymorphism. — Little evidence of this.

Material Examined. — This species is found in wet
forest only, from 1220 m to 1890 m in Haleakala Na-
tional Park and the Waikamoi Preserve (Table 1 ). Maui
Island'. Haleakala. Honomanu Gulch, 1876 m, 29-V-
88, 22-VI-89 & 5-II-90 (R. G. Gillespie & C. Parrish);
1585 m, 6-II-90 (R. G. Gillespie). Waikamoi Gulch,
1310 m, 13-VIII-88 (R. G. Gillespie & C. Parrish); 7-
1-91 (D. J. Preston); Bogs, NE Rift Haleakala, 1,676
m, 15-1-88, 16-1 88, 17-I-88& 18-1-88 (R. G. Gillespie
& A. C. Medeiros); Kipahulu Valley, 1220 m, 15-V-
90 (R. G. Gillespie & A. C. Medeiros); 1524 m, 14-V-
90 (R. G. Gillespie & A. C. Medeiros).

6

THE JOURNAL OF ARACHNOLOGY

Figures 1 5-2S. — Tetragnatha eurychasma-, Male holotype. 15) Promargin of right chelicera; 16) Retromargin
of left chelicera; 1 7) Dorsal spur of right chelicera, lateral view; 18) carapace, dorsal view; 19) Right leg I, dorsal
view; 20) Right leg III, prolateral view; 21) Left palpus, prolateral view. Female allotype. 22) Promargin of right
chelicera; 23) Retromargin of left chelicera; 24) Carapace, dorsal view; 25) Right leg I, dorsal view; 26) Right
leg III, prolateral view; 27) abdomen, dorsal view; 28) Seminal receptacles, ventral view. Scale lines in mm.
Scale of Figs. 15-18, 22-24 indicated below 23; scale of 19, 20, 25, 26 indicated below 25.

Tetragnatha eurychasma, new species
(Figs. 15-28, 86)

Types. —Holotype male from Honomanu
Gulch, Waikamoi, 1585 m, Maui Island (6 Feb-
ruary 1990), collected by R. G. Gillespie; allo-
type female from Carruther’s Camp, Honomanu

Valley, Waikamoi, 1876 m, Maui Island (29 May
1988), collected by R. G. Gillespie and C. Par-
rish, deposited in the Bishop Museum, Hono-
lulu.

Etymology. — Fury s (Greek) broad; chasma
(Greek) cleft, opening. The specific epithet is used

GILLESPIE- HAWAIIAN TETRAGNATHA

1

in its adjectival form and refers to the web of the
species, which, although of the basic tetragnathid
type (fragile, open hub), has generally very large
spaces {X = 1.38 cm, SD = 0.27, « = 8) between
the radial lines.

Diagnosis. — r. eurychasma is unlikely to be
confused with other species from Waikamoi. Its
distinctive black and silver coloration and
smoothly oval abdomen are characteristic of live
specimens. In alcohol, the most distinctive fea-
tures are its abdominal pattern, short leg spines
and cheliceral armature.

Description.— male: (Figs. 15-21).
Promargin of chelicerae (Fig. 15): Distance be-
tween ‘Gu’ and ‘si’ much greater than ‘si’ and
‘T’, ratio of distal end to ‘si’: ‘si’ to ‘T’: ‘T’ to
‘rsul’ 5:3:2. ‘Gu’ very small, inconspicuous, flat-
topped tubercle; ‘si’ sharp, wedge directed slight-
ly downwards towards ‘T’; narrower than ‘T’, by
63% (53-65%), and shorter, 53% (40-55%) height.
‘T’ moderately tall, directed perpendicular from
margin of chelicerae, but curved slightly up to-
wards ‘si’, ‘rsu’ 3 (up to 5) spikes, ‘rsuF slightly
closer to ‘T’ than to ‘rsu2’. Retromargin of che-
licerae (Fig. 16): Total of 8 (7) teeth. ‘AXl’ small,
almost square notch; ‘Gl’ strong, much taller than
all other teeth on retromargin; ‘L3’ next in size,
‘L2’ smaller, remainder of teeth considerably
smaller than ‘L2’. Dorsal spur long, shaped like
slim, bent finger ( 1 5.7% length of cephalothorax,
15.5-15.8%); tip pointed, upper margin project-
ing slightly beyond lower (Fig. 17). Cheliceral
fang considerably shorter than base, bent sharply
at proximal end and curved slightly at distal end.
Cephlothorax 1.7 mm (1.7-1. 8), total length 3.0
mm (2.9-3. 1). Chelicerae shorter (70%, 70-71%)
than cephalothorax. Cephalothoracic pattern a
distinct, dark flask shape, constricted at thoracic
fovea (Fig. 1 8). Leg spination similar to female
(Figs. 19, 20). Femur I: 4 prolateral, 3 dorsal, 3
retrolateral spines. Tibia I: 3 prolateral, 2 dorsal,
3 retrolateral spines. Metatarsus I: 1 prolateral,
1 dorsal, 3 retrolateral spines. Femur III: 3 dor-
sal, no ventral spines. Tibia III: 1 dorsal spine.
Coloration and eye pattern similar to female.

Conductor Tip: (Figs. 21 and 86). High-peaked
cap, leading out to narrow, straight, horizontal
projection (width similar to cap) which termi-
nates in small, downward-pointing, beak-like tip.

Allotype female: (Fig. 22-28). PME separated
by less than half width of PME (Fig. 24). Median
ocular area almost square. Lateral eyes loosely
contiguous. Promargin of chelicerae (Fig. 22):
series of 6 teeth, ‘U 1 ’ moderate size, similar in

width, but shorter (66%) than ‘U2’; separated
from ‘U2’ by 13% cheliceral length. ‘U2’-‘U6’
gradually decreasing in size proximally. Retro-
margin of chelicerae (Fig. 23): series of 6 teeth,
‘LI’ slightly shorter (90%) than ‘Ul’, much
smaller (69%) than ‘L2’. ‘LI’ contiguous with
‘L2’, teeth decreasing in size proximally. Cheli-
ceral fang short, approximately 64% length of
base, tapering to smooth point at distal end.
Cephlothorax 1 .8 mm, total length 5.4 mm. Che-
licerae short, 50% length of cephalothorax. Legs
almost uniformly brown. Spines small, rather in-
conspicuous ( 1 9% length of cephalothorax). Fe-
mur I: 4 prolateral, 3 dorsal, 3 retrolateral spines.
Tibia I: 3 prolateral, 2 dorsal, 3 retrolateral spines.
Metatarsus I: 1 prolateral, 1 dorsal, 3 retrolateral
spines (Fig. 25). Femur III: 2 dorsal, no ventral
spines. Tibia III: 1 dorsal spine (Fig. 26). Ceph-
lothorax pale brown; fovea distinctly marked by
dark area, lines running short distances from an-
terior and posterior margins, broader lines run-
ning forward from anterior lateral margins to
lateral eyes (Fig. 24). Sternum uniformly pale
tan. Abdomen smoothly elongate oval, width and
depth both approximately 22% of total length.
Dorsum of abdomen silvery, with wide, longi-
tudinal medial dark bar (slightly undulating mar-
gins) running down length (Fig. 27). Venter pale
tan with 2 pairs of gold spots on either side of
midline.

Seminal receptacles: (Fig. 28). Upper bulb oval,
set at 45° to main angle of body, lower bulb shaped
like “top”, upper border peaked. Two bulbs
linked in opposing “C” shapes, lower bulb slight-
ly smaller, projecting slightly farther out than
upper. Bulbs joined by rather long, curved stalk.
Median lobe an irregular balloon projecting from
stalk fitting well inside area defined by bulbs.

Color polymorphism. — Little evidence of this.

Material Examined. — This species is found through-
out wet forests of Haleakala National Park and the
Waikamoi Preserve, but is most abundant at higher
elevations (Table 1). Maui Island: Haleakala. Hono-
manu Gulch, 1876 m, 29-V-88, 22-VI-89 & 5-II-90
(R. G. Gillespie & C. Parrish); 1585 m, 6-II-90 (R. G.
Gillespie). Waikamoi Gulch, 1310 m, 13-VIII-88 (R.
G. Gillespie & C. Parrish); 7-1-9 1 (D. J. Preston). Opana
Gulch, 1 340 m, 8-VI-88 (R. G. Gillespie & C. Parrish);
8-II-90 (R. G. Gillespie & J. Burgett). Hanawi Valley,
1340 m, 9-II-90 (R. G. Gillespie & R. Rydell). Bogs,
NE Rift Haleakala, 1676 m, 15-1-88, 16-1 88, 17-1-88
& 18-1-88 (R. G. Gillespie & A. C. Medeiros); Kipa-
hulu Valley, 610 m, 17-V-90, 914 m, 16-V-90, 1220
m, 15-V-90, 1524 m, 14-V-90 (R. G. Gillespie & A.
C. Medeiros).

THE JOURNAL OF ARACHNOLOGY

Figures 29-42. — Tetragnat ha acuta-, Male holotype. 29) Promargin of right chelicera; 30) Retromargin of left
chelicera; 31) Dorsal spur of right chelicera, lateral view; 32) carapace, dorsal view; 33) Right leg I, dorsal view;
34) Right leg III, prolateral view; 35) Left palpus, prolateral view. Female allotype. 36) Promargin of right
chelicera; 37) Retromargin of left chelicera; 38) Carapace, dorsal view; 39) Right leg I, dorsal view; 40) Right
leg III, prolateral view; 41) abdomen, dorsal (a) and lateral (b) views; 42) Seminal receptacles, ventral view.
Scale lines in mm. Scale of Figs. 29-32, 36-38 indicated below 36; scale of 33, 34, 39, 40 indicated beside 40.

Tetragnatha acuta, new species
(Figs. 29-42)

Types. — Holotype male from Honomanu Val-
ley, Waikamoi, 1 585 m, Maui Island (6 February
1990), collected by R. G. Gillespie; allotype fe-

male from Opana Gulch, Waikamoi, 1340 m,
Maui Island (8 February 1990), collected by R.
G. Gillespie and J. Burgett, deposited in the Bish-
op Museum, Honolulu.

Etymology.— Acuta (Latin) acutely angled. The

GILLESPIE- HAWAIIAN TETRAGNATHA

9

specific epithet is used in its adjectival form and
refers to the high, pointed abdomen of this spe-
cies.

Diagnosis.— The dark brown/black coloration
with transverse lines, and the single medial tu-
bercle on the abdomen are highly distinctive for
T. acuta.

Description.— //o/oiype male: (Fig. 29-35).
Promargin of chelicerae (Fig. 29): Distance be-
tween ‘Gu’, ‘sF and ‘T’ approximately equal, ra-
tio of distal end to ‘sF: ‘sF to ‘T’: ‘T’ to ‘rsuF 3:
3:3. ‘Gu’ pronounced, small, rounded tubercle;
‘sF wedge-shaped, directed downwards towards
‘T’; narrower than ‘T’, by 69%, and shorter, 49%
height. ‘T’ moderately tall, directed almost per-
pendicular from cheliceral margin, ‘rsu’ series of
4 spikes. Retromargin of chelicerae (Fig. 30): To-
tal of only 6 teeth. ‘AXF small, pointed cone;
‘GF strong, much stronger than all other teeth
on retromargin; ‘L2’ and ‘L3’ short and robust;
‘L4’ and ‘L5’ taller and narrower. Dorsal spur
shaped like thick, bent finger (18.8% length of
cephalothorax); tip minutely bifurcate (Fig. 31).
Cheliceral fang considerably shorter than base,
bent sharply at proximal end and curved at distal
end. Cephlothorax 2.2 mm, total length 5. 1 mm.
Chelicerae shorter (80%) than cephalothorax.
Cephalothorax very dark, with darker margins,
and dark “V” shape leading into thoracic fovea
(Fig. 32). Legs banded, spination similar to fe-
male (Figs. 33-34). Femur I: 3 prolateral, 2 dor-
sal, 3 retrolateral spines. Tibia I: 3 prolateral, 2
dorsal, 3 retrolateral spines. Metatarsus I: 1 pro-
lateral, 1 dorsal, 1 retrolateral spines. Femur III:
4 dorsal, no ventral spines. Tibia III: 1 dorsal, 1
prolateral spine. Coloration and eye pattern sim-
ilar to female.

Conductor Tip: (Fig. 35). High-peaked cap,
curved sharply, leading out to narrow, straight,
horizontal projection of similar width to cap,
which terminates in rounded, blunt tip.

Allotype female: (Figs. 36-42). PME separated
by just more than half width of PME (Fig. 38).
Median ocular area almost square. Lateral eyes
well separated. Promargin of chelicerae (Fig. 36):
series of 7 teeth, ‘UT rather small, upwardly
directed cone, much smaller (45% length) than
‘U2’; widely separated from ‘U2’ by 26% cheli-
ceral length. ‘U2’-‘U7’ gradually decreasing in
size proximally. Retromargin of chelicerae (Fig.
37): series of 6 teeth, ‘LI’ slightly taller (147%)
than ‘Ur and similar in size (95%) to ‘L2’. ‘LI’
well separated from ‘L2’; rest of teeth on retro-
margin of similar height. Cheliceral fang short,

approximately 69% length of base, tapering to
smooth point at distal end. Cephlothorax 2.3
mm, total length 5.8 mm. Chelicerae moderately
short, 62% length of cephalothorax. Legs with
wide proximal, medial and distal dark bands
(Figs. 39-40). Spines small, rather inconspicuous
(18% length of cephalothorax). Femur I: 4 pro-
lateral, 2 dorsal, 3 retrolateral spines. Tibia I: 3
prolateral, 2 dorsal, 3 retrolateral spines. Meta-
tarsus I: 1 prolateral, 1 dorsal, 1 retrolateral spines.
Femur III: 3 dorsal, 1 prolateral, no ventral spines.
Tibia III: 1 dorsal, 1 prolateral spine. Cephalo-
thorax very dark, with darker margins, and dark
“V” shape leading into thoracic fovea (Fig. 38).
Sternum dark brown or black. Abdomen quite
broad (width approximately 42% of total length),
very deep (depth approximately 57% of total
length) (Fig. 41b). Dorsum of abdomen dark gray/
brown, with variable transverse markings, much
more heavily marked on posterior margin, where
tines converge upwards towards single, pointed,
medial tubercle (Fig. 41a). Venter with rather
broad black line running longitudinally down
midline, bordered along length by paler stripes.

Seminal receptacles: (Fig. 42). Two bulbs
spherical-to-oval, upper rather boxing-glove-
shaped, lower round. Linked by rather long stalk
running almost horizontally. Each bulb fairly well
sclerotized on medial border. Both bulbs equally
dilated. Median lobe small balloon shape (small-
er than bulbs) projecting from stalk, situated in
free space between widely separated bulbs.

Color polymorphism.— The color and extent
of patterning in this species is highly variable.
Some species have a medial bar (pink, brown or
black, but darker than the base color) running
longitudinally from the anterior edge of the ab-
domen to the midline; in others this is absent.
All specimens examined to date have some form
of transverse line running across the midline,
converging on the medial protuberance. Also
variable transverse bars posterior to the midline.

Material Examined. — This species is scattered
throughout Haleakala National Park and the Waika-
moi Preserve, rarely abundant (Table 1). Maui Island:
Haleakala. Honomanu Gulch, 1585 m, 6-II-90 (R.G.
Gillespie). Opana Gulch, 1340 m, 8-VI-88 (R. G. Gil-
lespie & C. Parrish); 8-II-90 (R.G. Gillespie & J. Bur-
gett). Hanawi Valley, 1340 m, 9-II-90 (R. G. Gillespie
& R. Rydell). Bogs, NE Rift Haleakala, 1676 m, 15-1-
88, 16-1 88, 17-1-88 & 18-1-88 (R. G. Gillespie & A.
C. Medeiros). Kipahulu Valley, 1980 m, 27-IV-88 (R.
G. Gillespie & A. C. Medeiros).

1 0 THE JOURNAL OF ARACHNOLOGY

Figures 43-56. — Tetmgnatha filiciphilia\ Male holotype. 43) Promargin of right chelicera; 44) Retromargin
of left chelicera; 45) Dorsal spur of right chelicera, lateral view; 46) carapace, dorsal view; 47) Right leg I, dorsal
view; 48) Right leg III, prolateral view; 49) Left palpus, prolateral view. Female allotype. 50) Promargin of right
chelicera; 51) Retromargin of left chelicera; 52) Carapace, dorsal view; 53) Right leg I, dorsal view; 54) Right
leg III, prolateral view; 55) abdomen, dorsal view; 56) Seminal receptacles, ventral view. Scale lines in mm.
Scale of Figs. 43-45, 50, 51 indicated below 50; scale of 46, 52 indicated below 52; scale of 47, 48, 53, 54
indicated beside 53.

Tetragnatha filiciphilia, new species
(Figs. 43-56, 87)

Types. — Holotype male and allotype female
from Waikamoi Gulch, Waikamoi, 1310m, Maui
Island (8 July 1988), collected by R. G. Gillespie

and C. Parrish, deposited in the Bishop Museum,
Honolulu.

Etymology.— Filix (Latin) fern; philia (Greek)
affinity for. The specific epithet is used in its
adjectival form and refers to the tendency of this

GILLESPIE-HAWAIIAN TETRAGNATHA

species to build its web under the fronds of ferns,
of which tree ferns are one of the dominant groups.

Diagnosis. — In life, T. feliciphilia is immedi-
ately recognizable on the basis of its distinctive
green coloration, with the medial red bar on the
posterior of the abdomen. It might be confused
with members of the Green Spiny Leg clade, al-
though inspection of the legs, with their few, small
and weak spines (appear almost smooth to the
naked eye) readily indicates its lack of allegiance
to this clade. Specimens preserved in alcohol
might be confused with T. eurychasma, but is
easily recognized by the lack of paired gold spots
on the venter. Also, the (apparent) lack of any
distinct abdominal pattern, the pale coloration
of the cephalothorax, and (in males) the cheli-
ceral armature readily identifies T. feliciphilia.

Description.— Holotype male: (Figs. 43-49).
Promargin of chelicerae (Fig. 43): Distance be-
tween distal margin, ‘sP and ‘T’ approximately
equal, ratio of distal end to ‘si’: ‘sF to ‘T’: ‘T’ to
‘rsul’ 4:3:3 (3:3:4). ‘Gu’ absent, represented only
by few strong hairs; ‘sF medium-sized cone di-
rected out perpendicular to margin of chelicerae;
narrower than ‘T’, by 61% (60-95%), much
shorter, 51% height (50-55%). ‘T’ moderately
tall, robust, rocket-shaped, leaning slightly up
towards ‘sF. ‘rsu’ 4 straight spikes perpendicular
to margin of chelicerae. Retromargin of chelic-
erae (Fig. 44): Total of 7 (6) teeth. ‘AXF absent;
‘GF stronger than any other tooth on retromar-
gin. Dorsal spur short, shaped like straight finger
(9.6% length of cephalothorax, 9.5-10.0%); tip a
single, slightly blunt, point (Fig. 45). Cheliceral
fang considerably shorter than base, bent at prox-
imal end, slightly curved at distal end. Cephlo-
thorax 1.6 mm (1.4-1. 7), total length 3.9 mm
(3.0-4.0). Chelicerae much shorter (67%, 60-68%)
than cephalothorax. Cephalothorax pale yellow,
darker at depression of thoracic fovea, where dark
lines radiate forwards and laterally from sides
towards margin of cephalothorax (Fig. 46). Leg
spination similar to female (Figs. 47-48). Femur
I: 3 prolateral, 1 dorsal, 3 retrolateral spines.
Tibia I: 1 prolateral, 1 dorsal, 3 retrolateral spines.
Metatarsus I: 1 prolateral, 1 dorsal, 2 retrolateral
spines. Femur III: 2 dorsal, 1 prolateral spines,
no ventral spines. Tibia III: 1 dorsal, 1 prolateral
spine. Coloration and eye pattern similar to fe-
male.

Conductor Tip: (Figs. 49, 87). Moderately high,
rather pointed cap, drawn out laterally into nar-
row, straight, almost horizontal projection of

11

similar width to cap, which terminates in slightly
beaked, blunt tip.

Allotype female: (Figs. 50-56). PME separated
by approximately half width of PME (Fig. 52).
Median ocular area wider posteriorly. Lateral eyes
contiguous. Promargin of chelicerae (Fig. 50):
series of 6 teeth, ‘U 1 ’ medium sized, much small-
er (50% height) than ‘U2’; widely separated from
‘U2’ by 24% cheliceral length. ‘U2’-‘U6’ grad-
ually decreasing in size proximally. Retromargin
of chelicerae (Fig. 51): series of 7 teeth, ‘LI’ taller
(128%) than ‘Ul’, slightly smaller (91%) than
‘L2’. ‘LL, ‘L2’ and ‘L3’ well separated, ‘L2’ and
‘L3’ largest teeth on retromargin. ‘L4’-‘L7’ rather
small and close together. Cheliceral fang short,
approximately 80% length of base, tapering to
smooth point at distal end. Cephlothorax 1.7
mm, total length 4.8 mm. Chelicerae rather short,
62% length of cephalothorax. Legs uniformly pale
yellow. Spines very small and inconspicuous ( 1 5%
length of cephalothorax). Femur I (Fig. 53): 3
prolateral, 1 dorsal, 3 retrolateral spines. Tibia
I: 1 prolateral, 1 dorsal, 3 retrolateral spines.
Metatarsus I: 1 prolateral, 1 dorsal, 2 retrolateral
spines. Femur III (Fig. 54): 2 dorsal, 1 prolateral,
no ventral spines. Tibia III: 1 dorsal spine. Ceph-
alothorax pale yellow, with darker lines from an-
tero-lateral margins of cephalothorax converging
broadly towards fovea. Sternum uniformly pale
yellow to brown. Abdomen elongate oval, some-
times slightly domed, width and depth both ap-
proximately 23% of total length (Fig. 55). Dor-
sum of abdomen almost uniformly speckled
silver, iridescent lime green in life, with broad,
conspicuous medial red bar running from just
behind midline to posterior margin of abdomen.
Venter silvery with broad medial longitudinal
brown bar, expanded anterior to epigastric fur-
row.

Seminal receptacles: (Fig. 56). Upper bulb
elongate-oval, at about 45° to body axis; lower
bulb round-oval. Each bulb has fairly well scler-
otized medial border. Both bulbs dilated, but
lower smaller than upper. Median lobe angular
and irregular doughnut-shape arising from stalk
between fairly widely separated bulbs.

Color polymorphism. — Little evidence of this.

Material Examined.— This species occurs at mid and
lower elevations in Haleakala National Park and Wai-
kamoi Preserve (Table 1). Maui Island: Haleakala.
Waikamoi Gulch, 1310 m, 8-VII-88 & 13-VIII-88 (R.
G. Gillespie & C. Parrish); 7-1-9 1 . Opana Gulch, 1 340
m, 9-IV-88 & 26-V-88 (R. G. Gillespie); 8-VI-88 (R.
G. Gillespie & C. Parrish) & 8-II-90 (R. G. Gillespie

12

THE JOURNAL OF ARACHNOLOGY

Figures 51-10. — Tetragnat ha stelarobusta-, Male holotype. 57) Promargin of right chelicera; 58) Retromargin
of left chelicera; 59) Dorsal spur of right chelicera, lateral view; 60) carapace, dorsal view; 61) Right leg I, dorsal
view; 62) Right leg III, prolateral view; 63) Left palpus, prolateral view. Female allotype. 64) Promargin of right
chelicera; 65) Retromargin of left chelicera; 66) Carapace, dorsal view; 67) Right leg I, dorsal view; 68) Right
leg III, prolateral view; 69) abdomen, dorsal view; 70) Seminal receptacles, ventral view. Scale lines in mm.
Scale of Figs. 57-59, 64, 65 indicated below 65; scale of 60, 66 indicated below 66; scale of 61, 62, 67, 68
indicated beside 68.

& J. Burgett). Kipahulu Valley, 914 m, 16-V-90, 1220
m, 15-V-90 (R. G. Gillespie & A. C. Medeiros).

Tetragnatha stelarobusta, new species
(Figs. 57-70, 88, 89)

Types. — Holotype male from Waikamoi
Gulch, Waikamoi, 1 340 m, Maui Island ( 1 2 July

1988), collected by R. G. Gillespie; allotype fe-
male from Waikamoi Gulch, Haleakala, 1310m,
Maui Island (13 August 1988), collected by R.
G. Gillespie, deposited in the Bishop Museum,
Honolulu.

Etymology.— Stele (Greek) cylinder; robustus

GILLESPIE-HAWAIIAN TETRAGNATHA

13

(Latin) robust. The specific epithet is used in its
adjectival form and refers to the robust cylin-
drical abdomen with longitudinal striping.

Diagnosis. — r. stelarobusta can be recognized
by its elongate, cigar shape, the distinctive ce-
phalothoracic pattern (especially the medial pale
line), its large size and brown coloration with a
pattern that is longitudinal, never transverse.

Description.— //otovpe’ male: (Figs. 57-63).
Promargin of chelicerae (Fig. 57): Distance be-
tween ‘Gu’ and ‘si’ slightly greater than that be-
tween ‘si’ and ‘T’, ratio of distal end to ‘si’: ‘si’
to ‘T’: ‘T’ to ‘rsuT 4:3:4 (4:3:3). ‘Gu’ broad, ro-
bust, rounded tubercle; ‘sF rather small, narrow
(medium-width) cone directed out perpendicular
to margin of chelicerae; narrower (by 30%, 30-
60%) and shorter (by 44% height, 40-45%) than
‘T’. ‘T’ tall, very robust, rocket-shaped, leaning
very slightly up towards ‘sF. ‘rsu’ 4 (5) narrow,
straight spike perpendicular to margin of chelic-
erae. Retromargin of chelicerae (Fig. 58): Total
of 9 (6-9) teeth. ‘AXF robust, rounded tubercle;
‘GF very wide and robust, much stronger than
any other tooth on retromargin. Dorsal spur long,
shaped like curved finger (15.5% length of ceph-
alothorax); tip broad, very blunt, with evidence
of minute bifurcation (Fig. 59). Cheliceral fang
slightly shorter than base, bent at both proximal
and distal ends. Cephalothorax 3.0 mm (2.7-
3.0), total length 7.4 mm (7. 0-7. 5). Chelicerae
shorter (80%, 70-80%) than cephalothorax. Ce-
phalothoracic markings similar to female (Fig.
60). Leg spination similar to female (Figs. 61-
62). Femur I: 5 prolateral, 4 dorsal, 3 retrolateral
spines. Tibia I: 3 prolateral, 1 dorsal, 3 retrola-
teral spines. Metatarsus I: 1 prolateral, 1 dorsal,
2 retrolateral spines. Femur III: 3 dorsal, 2 pro-
lateral, no ventral spines. Tibia III: 1 dorsal, 1
prolateral spine. Coloration and eye pattern sim-
ilar to female.

Conductor Tip: (Figs. 63, 88, 89). Broad, low,
rounded cap, almost symmetrical, but drawn out
laterally into moderately narrow, straight, pro-
jection of similar width to cap, which terminates
in spatulate, slightly beaked, blunt tip.

Allotype female: (Figs. 64-70). PME separated
by just over half width of PME (Fig. 66). Median
ocular area almost square. Lateral eyes very
loosely contiguous. Promargin of chelicerae (Fig.
64): series of 5 teeth, with minute nipple on very
apex of tooth row (absent in some individuals).
‘U 1 ’ very wide, wedge-shaped, slightly wider but
considerably shorter (40%, 40-50%) than ‘U2’;
widely separated from ‘U2’ by 37% (35-
47%)cheliceral length, ‘U3’ much smaller than

‘U2’, ‘U3’-‘U5’ gradually decreasing in size
proximally. Retromargin of chelicerae (Fig. 65):
series of 7 teeth, ‘LI’ similar in shape and slightly
largerthan‘Ur (124% height, 1 15-125%), much
smaller (50% height, 50-67%) than ‘L2’. ‘LI’ well
separated from ‘L2’, remainder of teeth closer
together. Teeth gradually decreasing in size prox-
imally. Cheliceral fang moderately long, approx-
imately 84% length of base, tapering to smooth
point at distal end. Cephalothorax 3.6 mm (3.5-
3.8), total length 9.5 mm (9.3-12.0). Chelicerae
rather short, 64% (55-65%) length of cephalo-
thorax. Legs lightly spotted, at least on femora,
many of spots associated with spines (Figs. 67-
68). Spines small (21% length of cephalothorax),
but conspicuous because of dark pigment at base.
Femur I: 5 prolateral, 4 dorsal, 4 (3) retrolateral
spines. Tibia I: 4 prolateral, 1 dorsal, 3 retrola-
teral spines. Metatarsus I: 2 prolateral, 1 dorsal,
2 retrolateral spines. Femur III: 3 dorsal, 2 pro-
lateral, no ventral spines. Tibia III: 1 dorsal, 1
prolateral spine. Cephalothoracic pattern very
distinct: narrow, pale line running straight down
midline, formed by separation of pair of wide,
dark bands running down either side of midline
as straight columns which constrict and converge
towards midline, with small tendrils radiating
laterally from fovea (Fig. 66). Lateral margins on
posterior part of cephalothorax also dark. Ster-
num dark coppery brown with dark margins. Ab-
domen cigar-shaped, width and depth both ap-
proximately 18% of total length (Fig. 69). Dorsum
of abdomen with variable, straight-to-undulat-
ing dark marks on tan-brown. Venter with rather
broad black line running longitudinally down
midline, bordered along length by paler stripes.

Seminal receptacles: (Fig. 70). Two bulbs linked
in opposing crescent shapes. Both bulbs well di-
lated, upper bulb slightly larger than lower, lower
projecting considerably farther out than upper.
Median lobe takes form of doughnut projecting
from stalk, situated in free space between widely
separated bulbs.

Color polymorphism.— This species is highly
variable in the nature of the longitudinal lines
running down the chestnut-brown abdomen. In
some species, these are pale yellow, in other they
are very dark black. Their width is also variable.
Similarly, the extent of leg marking is variable,
with more pigmented species having heavy black
bars around the joints of their legs. The femoral
spotting, however, is invariably present.

Material Examined. — This species is common
throughout Haleakala National Park and Waikamoi

14

THE JOURNAL OF ARACHNOLOGY

Preserve (Table 1). Maui Island: Haleakala. Hono-
manu Gulch, 1876 m, 29-V-88, 22-VI-89 & 5-II-90
(R. G. Gillespie & C. Parrish); 1585 m, 6-II-90 (R. G.
Gillespie). Waikamoi Gulch, 1310 m, 13-VII1-88 (R.
G. Gillespie & C. Parrish). Opana Gulch, 1340 m, 8-
VI-88 & 12-VII-88 (R. G. Gillespie & C. Parrish); 8-
11-90 (R. G. Gillespie & J. Burgett). Bogs, NE Rift
Haleakala, 1676 m, 15-1-88, 16-1 88, 17-1-88 & 18-1-
88 (R. G. Gillespie & A. C. Medeiros); Kipahulu Val-
ley, 1524 m, 14-V-90 (R. G. Gillespie & A. C. Med-
eiros).

Tetragnatha paludicola, new species
(Figs. 71-84, 90, 91)

Types. — Holotype male and allotype female
from the bogs on the NE Rift of Haleakala, 1676
m, Maui Island (18 January 1988), collected by
R. G. Gillespie and A. C. Medeiros, deposited
in the Bishop Museum, Honolulu.

Etymology. — Palus (Latin) bog; colo (Latin) to
dwell in a place. The specific epithet is used in
its adjectival form and refers to the very wet,
boggy habitats to which this species is virtually
confined.

Diagnosis.— The most diagnostic feature of T.
paludicola in the field is the smoothly oval, bottle
green abdomen with red chevrons, and paired
yellow marks on the venter. The color mostly
fades in alcohol, but the cheliceral armature and
shape of the palpal conductor are still distinctive.

Description. — //o/o(.vpe male: (Figs. 71-77).
Promargin of chelicerae (Fig. 71): Distance be-
tween ‘Gu’, ‘sP and ‘T’ approximately equal, ra-
tio of distal end to ‘si’: ‘sP to ‘T’: ‘T’ to ‘rsuF 3:
3:3. ‘Gu’ pronounced, small, cone-shaped tu-
bercle; ‘sP medium-sized cone directed out and
slightly up from margin of chelicerae; same width
as ‘T’, but much shorter, 37% height (35-48%).
‘T’ tall, narrow, rather straight spike, ‘rsu’ 5 (4-
5) spikes, ‘rsuF and ‘rsu2’ slightly divergent. Re-
tromargin of chelicerae (Fig. 12): Total of 6 teeth.
‘AXP conspicuous cone-shaped notch; ‘GP strong
and robust, wider but of similar height to ‘L5’
and ‘L6’, much stronger than ‘L2’-‘L4’. Dorsal
spur long, shaped like slim, bent finger (13.7%
length of cephalothorax); tip considerably longer
on dorsal side (Fig. 73). Cheliceral fang shorter
than base, bent sharply at both proximal and
distal ends. Cephalothorax 2.6 mm (2. 2-2. 7), to-
tal length 5.7 mm (5. 4-5. 8). Chelicerae shorter
(75%) than cephalothorax. Depression of tho-
racic fovea in form of paired semicircles (Fig.
74). Leg spination similar to female (Figs. 75-
76). Femur I: 5 prolateral, 4 dorsal, 2 retrolateral
spines. Tibia I: 3 prolateral, 2 dorsal, 3 retrola-

teral spines. Metatarsus I: 1 prolateral, 1 dorsal,

3 retrolateral spines. Femur III: 3 dorsal, 2 pro-
lateral, no ventral spines. Tibia III: 1 dorsal, 1
prolateral spine. Coloration and eye pattern sim-
ilar to female.

Conductor Tip: (Figs. 77, 90, 91). Symmetri-
cal, high-peaked cap, terminating in smoothly
tapered, downward-pointing projection.

Allotype female: (Figs. 78-84). PME separated
by just less than width of PME (Fig. 80). Median
ocular area wider posteriorly. Lateral eyes con-
tiguous. Promargin of chelicerae (Fig. 78): series
of 6 teeth, ‘U 1 ’ very robust, considerably wider
but shorter (64%) than ‘U2’; widely separated
from ‘LJ2’ by 28% cheliceral length. ‘U2’-‘U7’
gradually decreasing in size proximally. Retro-
margin of chelicerae (Fig. 79): series of 7 teeth,
‘LF slightly shorter than both ‘UF (92%) and
‘L2’ (86%). ‘LF distinctly separated from ‘L2’,
teeth barely (if at all) decreasing in size proxi-
mally. Cheliceral fang short, approximately 66%
length of base, tapering to smooth point at distal
end. Cephalothorax 3.3 mm, total length 8.6 mm.
Chelicerae rather short, 65% length of cephalo-
thorax. Legs well spotted, banded with reddish
brown on yellow (Figs. 8 1 , 82). Spines short (23%
length of cephalothorax). Femur I: 5 prolateral,

4 dorsal, 2 retrolateral spines. Tibia I: 3 prola-
teral, 2 dorsal, 3 retrolateral spines. Metatarsus
I: 1 prolateral, 1 dorsal, 3 retrolateral spines.
Femur III: 3 dorsal, 2 prolateral, no ventral spines.
Tibia III: 1 dorsal, 1 prolateral spine. Cephalo-
thorax pale brown, with distinct double fovea
marked by darker lines along medial and anterior
borders (Fig. 80). Sternum black with central
translucent yellow area. Abdomen broad, deep,
width and depth both approximately 46% of total
length (Fig. 83). Dorsum of abdomen green/
brown (bright bottle-green in life), with distinct
paired marks running down midline (in life,
paired red chevron marks). Venter brown, 2 pairs
of gold vertical bars on either side of midline.

Seminal receptacles: (Fig. 84). Two bulbs linked
in tight opposing, almost closed “comma” shapes,
each well sclerotized on medial border. Upper
bulb larger and more dilated than lower; central
portion serves as a wide stalk between bulbs.
Median lobe an ill-defined balloon projecting
from stalk, virtually enclosed by bulbs.

Color polymorphism.— This species exhibits
continuous variation rather than polymorphism,
and this is evident only in living specimens, in
the form and extent of the paired red marks down
the midline.

GILLESPIE- HAWAIIAN TETRAGNATHA

15

Figures 1\-M. — Tetragnatha paludicola\ Male holotype. 71) Promargin of right chelicera; 72) Retromargin
of left chelicera; 73) Dorsal spur of right chelicera, lateral view; 74) carapace, dorsal view; 75) Right leg I, dorsal
view; 76) Right leg III, prolateral view; 77) Left palpus, prolateral view. Female allotype. 78) Promargin of right
chelicera; 79) Retromargin of left chelicera; 80) Carapace, dorsal view; 81) Right leg I, dorsal view; 82) Right
leg III, prolateral view; 83) abdomen, dorsal view; 84) Seminal receptacles, ventral view. Scale lines in mm.
Scale of Figs. 71-73, 78, 79 indicated beside 78; scale of 74, 80 indicated beside 80; scale of 75, 76, 81, 82
indicated beside 82.

16

THE JOURNAL OF ARACHNOLOGY

Figure 85-91. — Scanning electron micrographs of conductor tips of male palps: 85) T. trituberculata\ 86) T.
eurychasma: 87) T. filiciphilia\ 88-89) T. stelawbusta\ 90-91) T. paludicola. Scale: Figs. 85-88, 91 is 1000 x;
Figs. 89, 90 is 200 x.

Material Examined.— This species is found in very
wet forest only (Table 1). Maui Island: Haleakala. Bogs
on north east rift of Haleakala, 1676 m, 15-1-88, 16-
1-88, 17-1-88 & 18-1-88 (R. G. Gillespie & A. C. Med-
eiros). Kipahulu Valley, 914 m, 16-V-90, 1524 m, 14-
V-90 (R. G. Gillespie & A. C. Medeiros).

ACKNOWLEDGMENTS

This study was supported by grants from the
Hawaii Bishop Research Institute, the Hawaii
Natural Area Reserves System and the Nature
Conservancy of Hawaii. Additional support was
provided by the Bishop Museum, the Nature
Conservancy of Hawaii, Haleakala National Park

and the Zoology Department, U. H. Manoa. I
am deeply indebted to the following for their
assistance in collecting specimens: Art Medeiros,
Lloyd Loope, Mark White, Rob Rydell, David
Preston, George Roderick, Jeff Burgett, Ron Na-
gata, Chris Parrish and Paul Higashino. Lee Goff
allowed me to use his compound microscope with
camera lucida and Kenneth Kaneshiro allowed
me to use his environmentally controlled facil-
ities to maintain and rear live specimens. Thanks
also to Marilyn Dunlap and Tina Carvalho for
help with the SEM. Also to Henrietta Croom,
Frank Howarth and Stephen Palumbi for advice

GILLESPIE-HAWAIIAN TETRAGNATHA

17

and discussion, and to Jonathan Coddington and

Norman Platnick for careful reviews of the first

draft.

LITERATURE CITED

Dabrowska Prot, E. & J. Luczak. 1968a. Spiders and
mosquitos of the ecotone alder forest {Carici elon-
gatae-alnetum) and oak pine forest {Pino querce-
tum). Ekologia Polska Seria A. XVI:46 1-483.

Dabrowska Prot, E. & J. Luczak. 1968b. Studies on
the incidence of mosquitos in the food of Tetrag-
natha montana Simon and its food activity in the
natural habitat. Ekologia Polska Seria A. XVI:843-
853.

Dabrowska Prot, E., J. Luczak & K. Tarwid. 1968.
Prey and predator density and their reactions in the
process of mosquito reduction by spiders in field
experiments. Ekologia Polska Seria A. XVI:773-
819.

Gillespie, R. G. 1991. Hawaiian spiders of the genus
Tetragnatha: 1. Spiny leg clade. J. ArachnoL, 19:
174-209.

Karsch, F. 1880. Sitzungs-Berichte der Gesellschaft
Naturforschender freunde zu Berlin. Jahrgang. Sit-
zung vom 18:76-84.

Kaston, B.J. 1948. How to know the spiders. 3rd. ed.
Wm. C. Brown Co., Dubuque, Iowa.

Levi, H.W. 1981. The American orb-weaver genus
Dolichognatha and Tetragnatha north of Mexico
(Araneae: Araneidae, Tetragnathinae). Bull. Mus.
Comp. ZooL, 149:271-318.

Okuma, C. 1 987. A revision of the Australasian spe-
cies of the genus Tetragnatha (Araneae, Tetrag-
nathidae). Esakia, 25:37-96.

Okuma, C. 1988. Redescriptions of the Hawaiian
spiders of Tetragnatha described by Simon (Ara-
neae, Tetragnathidae). J. Fac. Agr. Kyushu Univ.,
33:77-86.

Perkins, R. C. L. 1913. Introduction (to Fauna Ha-
waiiensis). In Fauna Hawaiiensis. Vol. 1 :xv-ccxxviii.
(D. Sharp, ed.). Cambridge Univ. Press, Cambridge.
Simon, E. 1900. Arachnida: Fauna Hawaiiensis. Vol.
2:443-519, pis. 15-19.

Suman, T.W. 1964. Spiders of the Hawaiian Islands:
Catalog and Bibliography. Pacific Insects, 6:665-
687.

Suman, T.W. 1970. Spiders of the family Thomisidae
in Hawaii. Pacific Insects, 12(4):773-864.

Wagner, W. L., D. R. Herbst and S. H. Sohmer. 1 990.
Manual of the Flowering Plants of Hawaii. Bishop
Museum and Univ. of Hawaii Presses, Honolulu.

Manuscript received April 1991, revised July 1991.

1992. The Journal of Arachnology 20:18-24

PHRYNIDAE (AMBLYPYGI) FROM ANDROS ISLAND,
BAHAMAS, WITH NOTES ON DISTRIBUTION PATTERNS,
RECENT ORIGIN AND ALLOMETRY

D. Jonathan Browne: Dept, of Entomology, University of Pretoria, Pretoria 0002,
Republic of South Africa

ABSTRACT: Fieldwork on Andros Island produced two species of phrynid amblypygi, Phrynus margine-
maculatus and Pamphrynus viridiceps. New localities and biological data are presented. The species were found
to be completely sympatric in two of the three localities where they were collected. The presence of amblypygi
in the Bahamas is attributed to Recent dispersal from Cuba via Paleoprovidence, a land mass which emerged
during the lower sea-levels accompanying Pleistocene glacials. Dispersal from Florida and Hispaniola is rejected.
Significant isometric or allometric relationships between median prosomal length and pedipalp tibia length were
not detected. The work generally points to the lack of much basic information on amblypygid distribution and
bionomics.

Two species of phrynid amblypygi are known
from the Bahamas. Phrynus marginemaculatus
C. L. Koch is widespread in the northern Bahama
Islands (Quintero 1981) and Paraphrynus viri-
diceps (Pocock) is known from southern Andros
and New Providence Islands (Banks 1906; Mul-
linex 1975). From published records, both ap-
pear to be uncommon to rare in the Bahamas.
Furthermore, these species have never been re-
corded as being completely sympatric in the same
habitat (see Quintero 1983).

The general lack of collecting of all arthropods
in the Bahama Islands, and especially Andros
Island, was noted by the author several years ago
while undertaking literature searches. This was
surprising because Andros Island is the largest
island in the Bahamas (Fig. 1), and has the great-
est diversity of vegetation types which would
presumably harbor the greatest number of ar-
thropod species in the Bahamas. As a result of a
familiarity based on many trips to Andros Island
the author decided to undertake an extensive sur-
vey of this island, employing modem mass col-
lecting techniques not used by earlier workers.
Recent collecting has revealed new localities and
biological data for two Bahamian amblypygi spe-
cies on Andros Island. This report is offered as
a contribution to knowledge of the arthropods
of the West Indies.

METHODS

The project was intended as a general survey
of the arthropods of Andros Island, thus a wide

variety of collecting techniques were employed.
Collecting efforts from May through August 1987
(110 days) used flight intercept traps, malaise
traps, baited pitfall traps, shrub and tree beating,
grass sweeping, blacklighting and hand collecting
[the most successful in terms of obtaining am-
blypygi specimens]. The author was assisted by
a technician throughout the collecting period. In-
tensive hand collecting included investigating all
caves and “banana holes” found, which often
required the use of rappelling equipment to gain
entrance. In addition stones, rocks, leaf litter and
logs were turned and the spaces beneath the bark
of dead trees examined to a height of 2 m.

A large number of localities were sampled
(> 100) which included all the recognised vege-
tation zones many times over. Most recent plant
ecology workers have characterized ten distinct
Bahamian vegetation types; these are beach/
strand, coastal rock, coastal coppice, interior
coppice, pineland, savanna, scrub, freshwater
marsh, saltwater marsh, and mangrove (Nick-
rent et al. 1987).

All specimens recorded here were collected and
identified by the author and vouchers were de-
posited by Dr. S. B. Peck (Department of Biol-
ogy, Carleton University, Ottawa) in the arach-
nology collection of the American Museum of
Natural History.

The significance of allometric changes in am-
blypygi has been previously studied and detailed
by Quintero (1 983) for seven Cuban species. The
significance of allometric changes in Bahamian

18

BROWNE- PHRYNIDAE FROM ANDROS ISLAND, BAHAMAS

19

Figure 1.— Map showing major islands of the Bahamas in relation to the Antilles and Florida. Line around
islands represents probable extent of land area during Pleistocene glacial low sea levels. Paleoprovidence is the
name proposed for the largest and most central of these Pleistocene glacial islands. Water barriers to dispersal
were less significant during these times. Species may have been lost as sea levels rose, and the fauna was forced
into today’s more restricted island areas.

species has not been previously examined. Due
to the small sample size, immatures, females and
males were combined into a single data set (Table
1) as did Quintero (1983). The relationship be-
tween median prosomal length (x) and pedipalp
tibia length (y) was calculated using the Stat-
graphics simple regression package (Version 2.6).
Both linear and multiplicative models were fit to
the data. The best-fit model was chosen based
on R-squared values with the final relationship
between x and y expressed as a power curve
[y = <2x'’]. Comparison of actual and predicted
slopes (Futuyma 1986), using the Student’s t test,
followed techniques suggested by Sokal & Rohlf
(1981) and Hoel (1984). Interpretation of results
adhered to methods proposed by Futuyma (1986)
and Packard & Boardman (1987). Statistical data
are presented in Tables 2 and 3.

RESULTS AND DISCUSSION

Both Phrynus marginemaculatus and Para-
phrynus viridiceps were found to be resident on
Andros Island. Despite intensive and prolonged
searching, populations were located in only three

small (each roughly 0.5 km^) and isolated local-
ities of high interior coppice. No specimens of
either species were found in any of the other nine
recognized vegetation zones. Only a few speci-
mens of each species were collected so as not to
disrupt what appeared to be uncommon and iso-
lated populations. However, they were found to
be very abundant throughout the period from
May to August 1987. The two species bear the
following information.

Phrynus marginemaculatus

Distribution.— This species has the most up-
permost latitudinal distributional range of any
amblypygid species in the eastern part of North
America. It has been previously recorded from
Bermuda, southern Florida, Cuba, Haiti, Do-
minican Republic, Puerto Rico, Jamaica, and
Antigua (Quintero 1983) and is widespread in
the northern Bahama Islands (Quintero 1981).

New Records.— BAHAMAS: Andros Island (random
search of high interior coppices); CDC Farm, Cricket
Coppice, 19 July 1987, 1 female; Shot-gun Coppice,

20

THE JOURNAL OF ARACHNOLOGY

Table 1.— Measurements of Median Prosomal Length
(x) and Pedipalp Tibia Length 0')i in mm, from spec-
imens collected from Andros Island during July 1987.
A = Phrynus marginemaculatus, B = Paraphrynus vir-
idiceps.

Specimen .
number

A

B

V

y

.V

y

1

3.984

3.992

5.140

4.998

2

5.280

5.450

7.085

8.100

3

4.290

4.423

6.140

6.140

4

5.040

4.783

4.235

4.000

5

2.950

2.790

3.589

3.400

6

4.558

4.940

3.790

3.280

7

—

—

3.060

2.890

8

-

-

5.685

5.465

19 July 1987, 2 females (1 with 12 eggs), 4 males;
London Ridge, 24 July 1987, 1 male.

Variation. — Body length 6.32-13.98 mm; median
prosomal length 2.95-5.28 mm; left pedipalp tibia
length 2.79-5.45 mm. Color varies from wheat yellow
to black in both males and females.

Paraphrynus viridiceps

Distribution.— This species is limited to the
Bahamas and Cuba (Quintero 1983). Paraphry-
nus viridiceps has been previously recorded from
Cuba (Quintero 1983). The holotype male was
collected in New Providence and described by
Pocock (1893). Additional Bahamas records in-
clude Andros Island (South Bight) and New
Providence (Mullinex 1975).

New Records.— BAHAMAS: Andros Island (ran-
dom search of high interior coppices): CDC Farm,
Cricket Coppice, 19 July 1987, 1 male; London Ridge,
24 July 1987, 4 females, 2 males, 1 immature.

Variation. — Body length 6.29-15.04 mm; median
prosomal length 3.06-7.08 mm; left pedipalp tibia
length 2.89-8. 10 mm. Color varies from wheat yellow
to black in both males and females.

Local Distribution and Habitat Preferences:
High interior coppices occur on elevated parts
of Andros Island. This community is the most
diverse of the vegetation zones on the island.
Dominant woody plant species in this vegetation
zone include Bursera simaruba, Metopium tox-
iferum. Ficus aurea, Exothea paniculata, Calyp-
tranthes patlens, Drypetes diversifolia, Clusea ro-
sea, Psychotria angustifolia and Nectandra
coriacea (Nickrent et al. 1987). The surface of
the high coppice is very much eroded and slightly
depressed which tends to accumulate moisture.

They are protected from annual burning of the
surrounding pine forests and savanna (the two
most common vegetation zones). The canopy of
a coppice is dense with a cool and wet, but sparse,
understory layer. The vegetation grows on hon-
ey-combed limestone which affords the ambly-
pygids many small holes in which to retreat. Am-
blypygi appear to favor cool, wet habitats
(Quintero 1983) which explains their preference
for the high interior coppice. The coppices are
separated by wide stretches of arid savanna and
pine forest. Two of the three coppices in which
these species were found (London Ridge and Shot-
gun Coppice) lie approximately 60 km from each
other. The third coppice (Cricket Coppice) is
within four km of Shot-gun Coppice. Pocock
(1893) recorded Paraphrynus viridiceps from
southern Andros Island, which is over 100 km
from the populations recorded here, and per-
manently separated by a wide salt-water gap from
the central and northern sections of the island.
Although amblypygids are known to run rapidly,
their dispersal capabilities are unknown. No
specimens were found in the dry savanna or pine
forest. If non-habitable areas lie between cop-
pices, then four km may be as much a barrier to
gene flow as 60 km. At present it is not known
if amblypygids disperse between coppices in the
Bahamas. However, widespread habitat destruc-
tion has occurred several times on Andros Island
as a result of logging; therefore, the present ap-
parent isolation of these populations may be a
recent condition.

Sympatry: Partial sympatry of species ranges
between Phrynus and Paraphrynus has been re-
ported previously (Quintero 1983). However,
Quintero (1983) “doubts” that “species ranges
will overlap to a major extent” due to “compet-
itive exclusion”. On Andros Island Paraphrynus
viridiceps and Phrynus marginemaculatus were
found to be completely sympatric in the same
habitat. They were also observed to intermingle
freely. However, these observations do not dis-
prove the occurrence of competitive exclusion.
The absence of specific information relating to
specific niche requirements for either species
makes it difficult to support or refute competitive
exclusion.

Recent origin of Bahamian amblypy-
gids: Three possible sources of Bahamian am-
blypygi must be considered. These are Florida,
Cuba and Hispaniola. Dispersal of flora and fau-
na from Florida into the Caribbean is considered
to be a very rare event. Many insect groups (Eick-

BROWNE-PHRYNIDAE FROM ANDROS ISLAND, BAHAMAS

21

Table 2.— Statistics for regressions of Median Prosomal Length (x) versus Pedipalp Tibia Length (v), comparing
linear (L) and multiplicative (M) models. Data from specimens collected from Andros Island during July 1987.
A = Phrynus marginemaculatus, B = Paraphrynus viridiceps, a = intercept, b = slope.

Sp.

Model

a ± SE

b ± SE

SE estimate

A

L

-0.295 ± 0.59

1.08 ± 0.13

94.18

0.250

A

M

-0.153 ± 0.16

1.11 ± 0.11

95.92

0.053

B

L

-1.208 ± 0.45

1.23 ± 0.01

96.89

0.335

B

M

-0.338 ± 0.11

1.20 ± 0.07

97.81

0.057

wort 1988“Halictidae; Liebherr l98S—Platyn-
us; Nichols 1988~Scaratinae; Ramos 1988 —
Homoptera; Slater 1988 — Lygaeidae; Wilson
1988 — Formicidae; Peck 1989 — south Florida
insects) and trees (Tomlinson 1980) are believed
to have dispersed from the Caribbean north, via
Cuba or the Bahamas, to south Florida (Patter-
son & Stevenson 1977). The strong northern
movement, from the southern Caribbean, of
storms, prevailing winds and the Gulf Stream
supports this argument. The author agrees with
this assessment and with the implication that
southern movement of flora and fauna from
Florida to the Bahamas can be considered a very
rare event.

Immigration and certain residency could have
only been possible since the re-emergence of the
Bahamas after Pliocene flooding. The northern
Bahamas (including Andros Island) are part of
the Bahama Bank, considered to be exposed con-
tinental shelf (Lee 1951; Burke et al. 1984; Don-
nelly 1988). During the lower (some 100 m) sea-
levels accompanying Pleistocene glacials and as
recently as 18,000 years BP, the northern Ba-
hamas were broadly and continuously adjacent
to Cuba via a land mass known as Paleoprov-
idence, with the latter separated from the former
by only a few kilometres of open water (Fig. 1).
Immigration from Cuba during this time is the
most likely route, rather than the “stepping-
stone” route from Hispaniola via the southern
Bahamas. This hypothesis is reflected in the cur-
rent distribution of Bahamian trees. Sixty-three
species are common to the Bahamas and Cuba,
while only twenty-nine are common to the Ba-
hamas, Cuba, Hispaniola and the Lesser Antilles
(Patterson & Stevenson 1977). Therefore the
widespread presence of Phrynus marginemacu-
latus throughout the northern Bahama Islands
and the more restricted distribution of Para-
phrynus viridiceps to only a few of these islands
is a reflection of a once widespread distribution

of these species throughout Paleoprovidence.
Rising sea-levels and inundation of most of Pa-
leoprovidence have formed the present Bahama
Islands in the last few thousand years. This re-
duction in area would also have reduced species
numbers to present levels and restricted their
movement between the newly isolated islands.

Allometry: Both the linear and multiplicative
models had significant fits to the data for both
species (Table 2). However, the R-squared values
were higher for the multiplicative model, indi-
cating that a power curve is the best-fit line.

Allometries between closely related species are
not congruent; but significantly different inter-
specific differences between wild and laboratory
reared specimens have been reported, although
Quintero (1983) ascribes these differences to
methodology. Quintero (1983) reported isomet-
ric growth for seven Cuban species of amblypygi.
It is more probable that he proved allometric
growth for reasons which are detailed below.

Demonstration of an allometric or an isomet-
ric relationship depends on the value of both the
intercept and slope of the best-fit line. Packard
& Boardman (1987), in their comprehensive re-
view of allometric analysis, state: “When a plot
of some variable of interest yields a straight line
passing through the origin of a graph with linear
co-ordinates, the variable varies isometrically
with body size. When the line is curvilinear or
when it does not pass through the origin, how-
ever, the variable varies allometrically with body
size”. Since neither line passes through the ori-
gin, the Bahamian amblypygi species in this study
do not exhibit isometric growth (Table 2; Fig. 2).
It is also evident that Quintero (1983) demon-
strated allometric growth, rather than isometric
growth, as none of the best-fit lines which he
presented passed through the origin. This is not
an unexpected finding since size-related varia-
tion in most physiological variables is allometric
(Packard & Boardman 1987).

22

THE JOURNAL OF ARACHNOLOGY

P

E

D

I

P

A

L

P

T

I

B

I

A

L

E

N

G

T

H

MEDIAN PROSOMAL LENGTH

P

E

D

I

P

A

L

P

T

I

B

I

A

L

E

N

G

T

H

MEDIAN PROSOMAL LENGTH

Figure 2.— Allometric growth curves with linear co-ordinates; Median Prosomal Length (abscissa) versus
Pedipalp Tibia Length (ordinate), measurements in mm. Slopes of both graphs are not significantly different
from one and therefore an allometric relationship is not demonstrated. A = Phrynus marginemaculatus, B =
Paraphrynus viridiceps.

BROWNE-PHRYNIDAE FROM ANDROS ISLAND, BAHAMAS

23

Table 3.— Statistics for Student’s t test to determine whether the slope of the multiplicative regression differs
significantly from one for each of the Andros Island amblypygi species, lib = 1, then an allometric relationship
is not demonstrated (Futuyma 1986). This is the case for both species.

Species

t

N-2

P

Phrynus marginemaculatus

0.246

4

0.3 > P > 0.4

Paraphrynus viridiceps

0.529

6

0.3 > P> 0.4

While the data in this study did not exhibit an
isometric relationship, it cannot be concluded
that an allometric one is demonstrated by de-
fault. An allometric relationship also depends on
the value of the slope {b) (Futuyma 1986). Al-
lometry is demonstrated only if 1 <b> \.\ib =
1, then y is a constant proportion of x and al-
lometry is not demonstrated. This is the case
with both species of Bahamian amblypygi from
Andros Island (Table 3); therefore, allometry is
not demonstrated. This is in contradiction to
Quintero (1983) who reported an allometric re-
lationship for the same variables in seven species
of Cuban amblypygi. Quintero (1983) did not
detail his methodology so no reason for this dis-
crepancy can be determined at this time, save
for disproportionate sample sizes.

ACKNOWLEDGMENTS

The author wishes to thank the people of An-
dros Island, without whose co-operation this
project would never have been realised. Thanks
are especially due to Mr. D. Myles who ably
assisted under difficult field conditions. Dr. G.
Dennill (Dept, of Entomology, University of Pre-
toria) and Dr. S. B. Peck (Dept, of Biology, Carle-
ton University) provided invaluable editorial as-
sistance. Dr. S. Chown and Mr. M. Vogt (Dept,
of Entomology, University of Pretoria) kindly
assisted with the statistical analysis. To Mr. J.
C. Cokendolpher, Dr. G. L. Miller, Dr. D. Quin-
tero and an anonymous reviewer I wish to extend
my sincere thanks for suggesting many improve-
ments to this manuscript. I also wish to thank
Panagiotis Karandis, the mad Greek, for stim-
ulating conversation and procuring a seemingly
endless supply of mental lubricants. To my Zhitt
& Jive II companions, Lou “Captain Ahab” De
Vries, Mark “Robo-scrop” Vogt and Jon “Hard-
copy” Pio, my sincere thanks for their useful,
albeit few, twisted and rather bizarre, ideas. Prof.
C. H. Scholtz (Dept, of Entomology, University
of Pretoria), my PhD supervisor, kindly allowed
me to delve into arachnids when he probably
would have preferred me to keep looking at scar-

abaeoid wings. My tolerant parents, Ron and
Mary Browne, have kindly given me much en-
couragement and financial support while com-
pleting this project and over many years. Finally
I would like to thank Prebendary Philip Hus-
bands and Albert Ernest Browne who provided
most of the funding for the field work.

LITERATURE CITED

Banks, N. 1906. Arachnida from the Bahamas. Bull.

American Mus. Nat. Hist., 22:185-189.

Burke, K., C. Cooper, J. F. Dewey, P. Mann, & J. L.
Pindell. 1984. Caribbean tectonic and relative plate
motions. Pp. 31-63. In The Caribbean— South
American Plate Boundary and Regional Tectonics
(W.E. Bonini, R. B. Hargraves, & R. Shagam, eds.).
Geol. Soc. America Mem., 162.

Donnelly, T. W. 1988. Geological constraints on Ca-
ribbean biogeography. Pp. 15-37. In Zoogeography
of Caribbean Insects. (J. K. Liebheer, ed.). Cornell
University Press, Ithaca, New York.

Eickwort, G. C. 1988. Distribution patterns and bi-
ology of West Indian sweat bees (Hymenoptera:
Halictidae). Pp. 231-254. In Zoogeography of Ca-
ribbean Insects. (J. K. Liebheer, ed.). Cornell Uni-
versity Press, Ithaca, New York.

Futuyma, D. J. 1986. Evolutionary Biology, 2nd ed.
Sinauer Associates, Inc., Sunderland, Massachu-
setts. 600 pp.

Hoel, P. G. 1984. Introduction to Mathematical Sta-
tistics. 5th ed. John Wiley and Sons, Toronto. 435
PP-

Lee, C. S. 1951. Geophysical surveys on the Bahama
Banks. J. Inst. Petrol., 37:633-657.

Liebherr, J. K. 1988. Biogeographic patterns of West
Indian Platynus carabid beetles (Coleoptera). Pp.
121-153. In Zoogeography of Caribbean Insects, (J.
K. Liebherr, ed.). Cornell University Press, Ithaca,
New York.

Mullinex, C. L. 1975. Revision of Paraphrynus Mo-
reno (Amblypygida: Phrynidae) for North America
and the Antilles. Occ. Papers Calififomia Acad. Sci.,
116:1-80.

Nichols, S. W. 1988. Kaleidoscopic biogeography of
West Indian Scaratinae (Coleoptera: Carabidae). Pp.
71-121. In Zoogeography of Caribbean Insects. (J.
K. Liebherr, ed.). Cornell University Press, Ithaca,
New York.

24

THE JOURNAL OF ARACHNOLOGY

Nickrent, D. L., W. H. Eshbaugh & T. K. Wilson.
1987. The vascular flora of Andros Island. Uni-
versity of Illinois, Urbana, Illinois. 182 pp.

Packard, G. C. & T. J. Boardman. 1987. The misuse
of ratios to scale physiological data allometrically
with body size. Pp. 2 1 6-240. In New Directions in
Ecological Physiology. (M. E. Feder, A. F. Bennett,
W. W. Burggren & R. B. Huey, eds.). Cambridge
University Press, Cambridge.

Patterson, J. & G. Stevenson. 1977. Native trees of
the Bahamas. Bahamas National Trust, Nassau. 128

pp.

Peck, S. B. 1989. A survey of insects in the Florida
Keys: Post-Pleistocene land-bridge islands: Intro-
duction. Florida Ent., 72:603-612.

Pocock, R. I. 1893. Contributions to our knowledge
of the arthropod fauna of the West Indies. III. Di-
plopoda and Malacopoda, with a supplement on the
Arachnida of the class Pedipalpi. J. Linn. Soc. Zook,
24:473-544.

Quintero, D., Jr. 1981. The amblypygid genus P/trv-
nus in the Americas (Amblypygi, Phrynidae). J. Ar-
achnol., 9:1 17-166.

Quintero, D., Jr. 1983. Revision of the amblypygid
spiders of Cuba and their relationships with the Ca-

ribbean and continental American amblypygid fau-
na. Studies on the Fauna of Curasao and other Ca-
ribbean Islands, 196:1-54.

Ramos, J. A. 1988. Zoogeography of the Auchen-
orrhynchous Homoptera of the Greater Antilles
(Hemiptera). Pp. 61-11. In Zoogeography of Carib-
bean Insects. (J. K. Liebherr, ed.). Cornell Univer-
sity Press, Ithaca, New York.

Slater, J. A. 1988. Zoogeography of West Indian Ly-
gaeidae (Hemiptera). Pp. 38-61. In Zoogeography
of Caribbean Insects. (J. K. Liebherr, ed.). Cornell
University Press, Ithaca, New York.

Sokal, R. R. & F. J. Rohlf 1981. Biometry: The
Principles and Practice of Statistics in Biological
Research. 2nd ed. W. H. Freeman and Co., San
Francisco. 859 pp.

Tomlinson, P. B. 1980. The Biology of Trees Native
to Tropical Florida. Harvard University Printing
Office, Allston, Massachusetts. 480 pp.

Wilson, E. O. 1988. The biogeography of the West
Indian ants (Hymenoptera: Formicidae). Pp. 214-
23 1 . In Zoogeography of Caribbean Insects. Cornell
University Press, Ithaca, New York.

Manuscript received April 1991, revised August 1991.

1992. The Journal of Arachnology 20:25-34

WEB CONSTRUCTION BY MODISIMUS SP.
(ARANEAE, PHOLCIDAE)

William G. Eberhard: Smithsonian Tropical Research Institute and Escuela de
Biologia, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica

ABSTRACT. The behavior used by Modisimus sp. to construct its domed sheet web is more stereotyped and
organized than is apparent from the finished structure. A simple program involving attaching the non-sticky
dragline to the substrate beyond the previous limits of the web, and then filling in the newly formed angle is
probably used to construct the skeleton sheet and the tangle above it. A set of sticky lines is then laid, filling in
this sheet. Construction behavior resembles that of orb weavers in commencing with a skeleton scaffold of non-
sticky lines which is then filled in with other non-sticky lines, in adding sticky lines after the support structure
of non-sticky lines is complete, and in being organized around a central area.

It is commonly stated in general texts that
pholcid spiders make non-adhesive tangle webs
with little or no organization (e.g., Levi, Levi &
Zim 1968; Forster & Forster 1973; Foelix 1982;
Shinkai 1984; Shear 1986). Most accounts are
apparently based primarily on the webs of the
temperate species Pholcus phalangiodes (Fues-
slin). With the slow accumulation of data on
tropical pholcids, it is becoming clear that there
is a rich diversity of web forms in this family
(Eberhard & Briceno 1985; Deeleman-Reinhold
1986; Eberhard in press a on Physocyclus glo-
bosus Taczanowski), and that some pholcid webs
include sticky lines (Briceno 1985), and entan-
gling “screw threads” (Kirchner 1986).

Other than the mention of two stages of web
construction in two Modisimus spp. (Eberhard
& Briceno 1985), there are, to my knowledge, no
descriptions of how pholcid webs are built (or,
for that matter, of the construction of almost any
other non-orb web; Eberhard 1 990a). Given the
relatively isolated taxonomic position of Phol-
cidae (e.g., Lehtinen 1 967; Shear 1 986), the means
by which they produce aerial sheet webs with
sticky lines are likely to prove of interest in com-
parison with construction of webs in other fam-
ilies. This paper describes the construction of
such webs by a third Modisimus species.

METHODS

Spiders were observed during daylight hours
on 22-25 February, 1991 on fallen trees and but-
tresses in an overgrown cocoa orchard at La Sel-
va Biological Station, near Puerto Viejo, Heredia
Province, Costa Rica (el. about 50 m). At least
part of the construction of 2 1 different webs was

observed. Web construction was elicited by par-
tially or nearly completely destroying the web on
which the spider was found. Some webs were
coated with cornstarch before being destroyed. I
included observations of spiders that started to
build replacement webs 5-45 min after their webs
were destroyed.

Several partially completed webs were coated
with cornstarch. By recoating these webs lightly
when they were finished, it was possible to dis-
tinguish the order in which lines had been laid
(more heavily coated lines first, others later). The
stages of construction behavior in partial and
complete web replacement were similar, and the
two are combined in the descriptions below.

Samples of adult webs, collected by wetting
the edges of a microscope slide and then lifting
it through the web, were viewed at 400 x with
direct illumination.

Spiders were identified by C. Deeleman-Rein-
hold. The species, which seems close to M. pul-
chellus Banks 1929, is apparently undescribed.
Voucher specimens are deposited in the Museum
of Comparative Zoology, Cambridge, MA 02138
(Nos. 3604, 3606, 3611), and in the collection
of C. Deeleman-Reinhold (Sparrenlaan 8, Os-
sendrecht. The Netherlands). This species is dif-
ferent from the Modisimus species whose be-
havior was described previously (Eberhard &
Briceno 1983, 1985).

RESULTS

Finished webs. —Webs of Modisimus sp. were
found attached to large supporting objects such
as the buttresses of trees or fallen logs (Fig. 1).
Web sites were usually sheltered from at least

25

26

THE JOURNAL OF ARACHNOLOGY

Figure 1.— Webs of Modisimus sp. on the heavily
populated base of a tree trunk with deep indentations.
Note the variability in web design. Scale bar is 1 5 cm.

moderate rains. Webs typically included a more
or less dome-shaped sheet of relatively open
mesh, with a sparse tangle of lines above which
was more dense in the area above the top of the
dome (Figs. 2, 3). The height of the tangle was
generally between 0. 5-1.0 times the maximum
diameter of the sheet. The spider rested on the
underside of the sheet at the peak of the dome.
The dome was usually asymmetrical, with the
peak near a large object (e.g., the trunk of a tree).
The sheet on the side away from this object (the
“exposed” side) was usually larger.

The dome was oriented more or less horizon-
tally, so the peak was the uppermost part of the
sheet (Figs. 1, 2). The exposed side was usually
below the top of the dome, and its edge was often
close to horizontal (Fig. 2). Orientations and
shapes varied, however, with websites. For in-
stance, some sheets were nearly planar (Fig. 4),
while in other webs, built in small indentations
in tree trunks, the exposed side of the sheet was
nearly vertical (Fig. 5).

The lines in the sheet were not arranged in
geometrically regular arrays, but they showed
consistent patterns. Near the border of the ex-
posed side, a few lines in the sheet were relatively
straight (Figs. 2, 3); judging by the amount they
sagged when coated with cornstarch, these lines

were under more tension and/or were less exten-
sible than the others. The lines forming the edges
of the sheet were of this type. Other lines in the
sheet which intersected the edges often had a “V”
shape (Fig. 3). By powdering webs twice (see
Methods), it was determined that the long, straight
lines in the sheet were built during skeleton web
construction, while the others were laid during
sheet fill-in behavior (Fig. 6; see below). A further
pattern, more marked in some webs than in oth-
ers, was that the lines in the sheet near the ex-
posed edge formed a more open mesh than those
in the sheet near the peak of the dome (Figs. 2,
3, 5).

The size and shape of the web, as well as the
density of lines in the sheet varied substantially
between webs of the same individual. Replace-
ment webs seemed to be smaller, with less dense-
ly meshed sheets (cf Figs. 2 and 6), but no precise
measurements were made.

Samples of three finished webs collected on
microscope slides had lines of at least three dif-
ferent diameters. Many of the finest lines bore
rows of small spheres. Near the exposed edge of
the sheet these lines tended to run approximately
perpendicular to the border of the web, which
was formed by a relatively thick line. When a
drop of water was placed on one sample, then
allowed to evaporate, the spheres were reduced
to small “ghosts”, indicating that a major frac-
tion of each sphere was water soluble. The junc-
tions of thicker lines generally had masses of
material that were probably attachment discs. In
contrast, points where fine lines vwth balls crossed
other lines generally lacked such masses.

Construction behavior. — I distinguished three
types of construction behavior: extending the
skeleton web; filling in the skeleton web; and
filling in the sheet. Construction always began
with extension of the skeleton web, which was
followed by alternating bouts of filling in the skel-
eton and further extension. A bout of filling in
the skeleton was usually followed vwthout a pause
by filling in the sheet. This stage was less fre-
quently interrupted by other activities, though
occasionally a few attachments (probably filling
in the skeleton) were made to the substrate and/
or to web lines near a sheltered edge of the web.
Sometimes the spider temporarily ceased filling
in the sheet and rested at the top of the dome,
only to resume this behavior 1-10 min later.

1. Extension of the skeleton web: The area en-
compassed by the web lines was gradually ex-
tended by additions along an edge. The spider

EBERHARD-WEB CONSTRUCTION BY MODISIMUS SP.

27

Figures 2-3.— A finished web of a mature female Modisimus sp. seen laterally (Fig. 2), and a closeup of the
sheet on the exposed side of the web seen from above (Fig. 3). Heavy arrows mark relatively straight lines in
the sheet, while others in Fig. 3 mark “V” junctions at the exposed edge of the web. Note also the remains of
a previous sheet just below the new one (arrow), and the approximately horizontal edge of the exposed side
(nearest the viewer) in Fig. 2. Scale bars are 3 cm (Fig. 2) and 2 cm (Fig. 3).

extended the web by first attaching its drag line
one or more times to one or more lines already
present, then walking to the end of the line on
the web’s edge and then along the substrate away

from this web line (Fig. 7A). It usually moved
more or less horizontally on the substrate. The
spider attached its dragline to the substrate by
bending its abdomen ventrally to touch the spin-

28

THE JOURNAL OF ARACHNOLOGY

.Wapii

EBERHARD-WEB CONSTRUCTION BY MODISIMUS SP.

29

Figure 6.— Lateral view of the web built to replace the web in Figs. 2 and 3 (note collapsed web against the
tree trunk at right). The more heavily powdered lines of the replacement web formed the skeleton web that was
built before the spider began to fill in the sheet. The lines laid as the sheet was filled in are more lightly powdered.
Note the relatively long, straight lines of the skeleton web that are incorporated in the nearer, “exposed” portion
of the sheet. Scale bar is 1 cm.

nerets to the substrate, holding the dragline with
one extended leg IV as it did so (Fig. 7 A). Then
it returned along this newly laid line, attaching
its dragline one or more times to it or to other
web lines as it went (Fig. 7B). Often, especially
soon after building began, the spider immedi-
ately went to the other side of the same edge of
the web and extended it also in a similar manner
(Fig. 7B). Occasionally the trip to the opposite
side was abbreviated when the spider turned back
before reaching the substrate and returned along
the first side to extend it further. Up to six suc-
cessive extensions on alternate sides of the same
edge of the web were seen. A series of extensions
usually ended when the spider attached its drag-

line part way across the border of the web, and
turned to walk inward toward the area where the
top of the dome would be located (Fig. 7C).

Early stages of extension of the skeleton web
included lines laid above the plane where the
sheet would eventually be in the finished web.
In contrast, later extensions were always close to
the plane of the sheet. Most extension occurred
on the exposed edge of the web (e.g., side op-
posite the sheltering tree trunk). During web ex-
tension the spider seemed to walk more slowly
than during later stages.

2. Filling in the skeleton web: After one or more
web extensions, the spider moved around in the
space encompassed by the web lines, attaching

Figures 4-5.— Webs of Modisimus sp., illustrating variations on the basic domed form. The sheet of the web
of a mature individual (Fig. 4) was nearly horizontal, rather than being domed. The sheet of the web of an
immature individual was built in a small indentation in a tree trunk, and included a large, nearly vertical
extension of the exposed edge (Fig. 5). Scale bars are 8 cm (Fig. 4) and 3 cm (Fig. 5).

30

THE JOURNAL OF ARACHNOLOGY

its dragline to many of the lines it crossed. Each
attachment was made by moving the abdomen
ventrally toward the web line to which the drag-
line would be attached. At least one leg III grasped
this line just anterior to the spinnerets. On some
occasions the contralateral leg III also grasped
the web line, also apparently just anterior to the
attachment site. More rarely the IV leg ipsilateral
to the III on the line also held the web line, in
this case just posterior to the attachment site. In
one case, a leg II also held the line just anterior
to the leg III. I was unable to determine if any
legs consistently held the dragline just before or
during attachment. In some webs, but not others,
the spider dropped down from the web at least
once for 1-5 cm on a dragline, then ascended on
the same line without having made an attach-
ment.

Most filling in of the skeleton web was per-
formed in the central area of the web, especially
where the top of the dome of the finished sheet
would be. Filling in of the skeleton web differed
from web extension in that spiders were never
seen to turn back after an attachment and walk
along the line that had just been laid, even when
this involved attachments to the substrate. In
some cases, the spider did not fill in on the ex-
posed side of the web in the area encompassed
by the last several web extensions.

As during web extension, the spider consis-
tently moved beneath lines already laid while
filling in the skeleton, very seldom climbing up
past a line to make an attachment (two excep-
tions were seen). As a result, the lines laid while
filling in the skeleton tended to form bridges un-
der the more upward projecting portions of the
web which had been laid earlier (Fig. 8). Usually
successive lines soon came to be concentrated in
the plane where the sheet would be, but in some
cases the lines did not form a plane at first, and
the spider moved gradually lower, making a taller
tangle of lines before finally forming a plane which
would be the sheet. When I had only partially

destroyed the previous web, the earliest skeleton
web attachments were made to the very edge of
the broken sheet, while later attachments were
approximately 1 mm from the edge on the intact
sheet. This resulted in the plane of the new sheet
being slightly below the broken edge of the old
sheet.

The combined processes of extension of the
skeleton web and filling in the skeleton lasted 3-
10 min. In the “completed” skeleton web (all the
lines present when the spider began to fill in the
sheet), the smallest mesh was where the top of
the dome would be, often forming a small, nearly
horizontal platform of relatively uniform mesh
that was approximately the size of the spider as
it rested on the web (Fig. 6). This area had a
tangle of lines above it, and was surrounded by
a more or less planar extension that was pro-
gressively less densely meshed farther away from
this central area.

3. Filling in the sheet: The process of filling in
the sheet usually took approximately 5-10 min.
The spider walked in approximately straight lines
beneath the skeleton sheet, repeatedly drawing
silk from its spinnerets with its legs IV. The hind
legs pulled the line (or lines?) and then pressed
it upward against the sheet, where it stuck. The
legs IV usually moved nearly synchronously up-
ward, with one lagging slightly behind the other
(Fig. 9). Occasionally they moved with alternate
upward strokes, as described for similar behavior
in other Modisimus spp. (Eberhard & Briceno
1985).

Lines laid as the sheet was filled in were oc-
casionally attached by touching the spinnerets to
web lines as described above. Such attachments
were made almost exclusively to lines near the
edge of the web, and were immediately followed
by the spider abruptly moving toward the central
area, thus producing a “V” configuration of the
sheet fill in lines (e.g.. Fig. 3). No attachments
were made to most other web lines encountered
as the sheet was filled in. In one case, with fa-

Figure 7 A-C.— Schematic representation of two successive typical web extensions, seen from above. Lines
present before the web extension began are dotted, and attachments made during each figure are dots. Numbers
indicate the sequence of dragline attachments. A. The spider attached its dragline to a line at the edge of the
web (1), walked to the substrate along a line, and then walked away from this line before attaching its dragline
to the substrate (2). B. The spider returned along this newly laid line to the edge of the web and attached there
(3), then walked farther along this edge to the substrate on the other side and attached there beyond the previous
edge of the web (4). C. The spider returned along this newly laid line, attached its dragline part way across the
new edge of the web (5), and moved toward the interior of the web.

EBERHARD-WEB CONSTRUCTION BY MODISIMUS SP.

31

32

THE JOURNAL OF ARACHNOLOGY

Figure 8. — Lateral schematic representation of how the lower portion of the skeleton web was lowered and
smoothed as the spider attached lines exclusively on the underside of the web as the skeleton web was filled in.
Lines laid early are dotted, those laid later are solid.

vorable lighting in which the skeleton web lines
had been powdered, I noted that lines laid as the
sheet was filled in were not tense, and moved
slightly in very weak air currents.

I collected samples of two webs on microscope
slides just after the spider began to fill in the
sheet, and found that the fine lines with balls on
them that were common in finished webs were
nearly completely absent. Thus these presum-
ably sticky lines were added to the skeleton web
when the sheet was filled in.

DISCUSSION

There seems to be a simple organizing “prin-
ciple” at work during web construction by Mod-
isimus, in both the horizontal (Fig. 7) and the
vertical (Fig. 8) dimensions. The spider first ex-
tends the sides of the angle formed by the limits
of the web (e.g.. Fig. 7A), then fills in the space
between the new sides with further lines (e.g.,
Fig. 7B, C). In the horizontal dimension this pro-
cess occurs repeatedly, and involves new attach-

Figure 9. — Diagrammatic representation of move-
ments made as the spider filled in the sheet. The spider
moved across the underside of the skeleton web (hor-
izontal arrow), pulling a line or lines from its spinnerets
with its hind legs and pushing upward (vertical arrows)
against the web. All legs other than one leg I and the
two legs IV are omitted for clarity.

ments to the substrate. This enables the spider
to extend the web in accord with the open space
available. The spiders’ long legs permit them to
span relatively large spaces, and thus move easily
across irregularities in the substrate.

In order to perform horizontal web extensions
effectively, the spider must take into account
which side of the web already has lines present
when it turns after reaching the previous attach-
ment to the substrate (e.g., turn to its left in Fig.
7A). This is necessary if the spider is to extend
the web by laying its new line on the side of the
previous attachment which lacks lines, rather than
add another line to the area already covered.
Spiders seemed to make such distinctions quite
consistently, as I never saw a spider walk to the
leading edge of a web, then walk in the wrong
direction along the substrate and attach a line
and return to the web along it. Possibly this dis-
crimination was accomplished by having some
legs holding web lines other than the line along
which the spider moved out toward the periph-
ery. Memory of distances and directions trav-
elled (an apparently ancient and widespread ca-
pability in arachnids; see Eberhard 1988) may
also be involved.

In order for the lines and attachments which
are laid after a web extension attachment to the
substrate to extend the web, the spider must re-
turn along the line it has just attached to the
substrate, rather than along previously laid lines.
This may explain why spiders consistently held
the dragline with one leg IV as the attachment
to the substrate was made (Fig. 7). In contrast
with many orb weavers (e.g., Eberhard 1982),
Modisimus sp. frequently failed to hold the drag-
line in other situations.

The web construction behavior of Modisimus.
sp. and that of two other Modisimus species

EBERHARD-WEB CONSTRUCTION BY MODISIMUS SP.

33

(Eberhard & Briceno 1985), which also build
domed sheets with small tangles above, are prob-
ably very similar. All three species began con-
struction by laying a scaffold of thicker lines
without sticky balls on them. The first two stages
of construction behavior described here (exten-
sion and skeleton fill in) apparently correspond
to the “Phase I” of the other two Modisimus spp.
(Eberhard & Briceno 1985). These species also
have long, straight lines near the edge of the sheet.
The later stages of construction by all three spe-
cies consists of filling in the plane of the sheet,
using legs IV to pull out lines and then push them
against the sheet. Lines laid at this stage carry
sticky balls (Briceno 1985) in at least two of the
species. In all three species, attachments of both
lines in the skeleton web and of sheet fill in lines
to the edges of the skeleton web are consistently
made while one leg III holds the line to which
the attachment is being made just anterior to the
attachment site. Similar use of one leg III during
attachment of the dragline also occurs in the
pholcid Physocyclus globosus (Eberhard, un-
publ.).

The establishment of a sparse network of lines
followed by the addition of interconnecting lines
in Modisimus sp. also resembles the construction
process described for some theridiid spiders (La-
moral 1968). The pholcid differs, however, in
establishing the first lines in non-radial rather
than radial directions (e.g. compare Fig. 7 with
fig. 7 of Lamoral 1968), and also in having a
much tighter mesh away from the edge of the
skeleton web.

Pholcids are thought to be only very distantly
related to orb weavers (Lehtinen 1967; Shear
1986). A comparison of the organization of the
webs and web building behavior of Modisimus.
sp. with that of orb weavers suggests three basic
similarities. First, in both web types a scaffolding
of non-sticky lines is built first, and used to sus-
tain sticky lines laid later. Second, in both web
types the outlines of the scaffold are built first,
and then gradually filled in, first with other non-
sticky lines (although in the pholcid these two
stages were more often mixed together), and then
with sticky lines. Finally, construction of both
web types is clearly organized around a central
area (the hub of an orb, the peak of the dome of
the pholcid web).

These presumably independently derived sim-
ilarities support the view that some orb-associ-
ated traits, such as construction behavior that is
organized in a plane around a central area, and

construction of a non-sticky scaffold which is
then filled in with sticky lines, are not limited to
orb weavers (Eberhard 1 990a). However, in the
pholcid web the lines do not radiate from a cen-
tral area, as do the radii of an orb, and neither
sticky nor non-sticky lines are organized in cir-
cular or spiral patterns. Thus the behavioral sim-
ilarities are not reflected in the geometric pat-
terns of lines in the finished webs.

Another difference is that the pholcids did not
break lines and reconnect them during web con-
struction. The early “exploration” stage typical
of orb web construction seemed to be absent,
except for occasional descents without attach-
ments, which resembled similar descents of some
orb weavers (Eberhard 1990b). All other lines
laid from the start of construction were included
in the finished pholcid web. Probably this lack
of line replacement behavior is a primitive trait.
The absence of line removal was not due to the
pholcid being unable to cut lines, as on several
occasions during skeleton web construction a spi-
der neatly cut out a piece of debris and dropped
it free. In no case, however, did a spider break
a line, then attach its dragline to one broken end
and reel up the other as it walked on, as occurs
in many orb weavers (Eberhard 1982, 1990b;
Coddington 1986a,b; Shinkai 1990) as well as in
some theridiids (Szlep 1966; Eberhard in press
b).

Perhaps Modisimus. sp. cannot effectively re-
move lines already laid, and must correct early
mistakes in skeleton web construction by adding
subsequent short lines which change the outline
of the lower margin of the mesh which is being
formed, in effect replacing the earlier lines by
lowering the site where the sheet will be made.
This implies that at least some of the tangle above
the sheet may represent exploration behavior.
However, I was unable to discard the alternative
possibility that differences in the height and
numbers of lines in tangles represented adjust-
ments to particular website characteristics.

Several other pholcids make more or less
domed sheets {Blechwscelis sp. and Modisimus
spp.-Eberhard & Briceno 1985; Physocyclus glo-
6o.9M5-Eberhard in press a). More or less domed
sheets also occur in several other families such
as Diguetidae (Nuessly & Goeden 1984), Ther-
idiidae (Main 1976; Shinkai 1984), Hypochili-
dae (Shear 1969), Linyphiidae (Nielsen 1931;
Kaston 1948), and Araneidae (Kullmann 1964;
Shinkai 1984). The functional significance of the
domed form is not clear. Domed sheets might

34

THE JOURNAL OF ARACHNOLOGY

be designed to capture prey which is flying up-
ward, as a dome could work in a manner anal-
ogous to a malaise trap, using the prey’s tendency
to fly upward to channel it toward the spider.
However, Modisimus. sp. often built new webs
just above the remains of previous sheets (Fig.
2). These old webs would make it difficult for
prey to reach the new web from below, and thus
argue against the malaise trap interpretation, at
least for this species.

ACKNOWLEDGMENTS

I thank C. Deeleman-Reinhold for kindly
identifying the spiders, and Y. Lubin and G. Mil-
ler for comments on a previous draft. Financial
support was provided by General Research Funds
of the Smithsonian Tropical Research Institute,
and the Vicerrectoria de Investigacion of the
Universidad de Costa Rica.

LITERATURE CITED

Briceno, R. D. 1985. Sticky balls in webs of the spider
Modisimus sp. (Araneae, Pholcidae). J. Arachnol.,
13:267-269.

Coddington, J. 1986a. Orb webs in “non-orb weav-
ing” ogre-faced spiders (Araneae: Dinopidae): a
question of genealogy. J. Cladistics, 2:53-67.
Coddington, J. 1986b. The monophyletic origin of
the orb web. Pp. 317-363. In Spiders— Webs, Be-
havior, and Ecology (W. Shear, ed.). Stanford Univ.
Press, Stanford.

Deeleman-Reinhold, C. 1986. Leaf-dwelling Phol-
cidae in Indo-Australian rain forests. Pp. 45-48. In
Proceedings of Ninth International Congress of Ar-
achnology, Panama (W. Eberhard, Y. Lubin & B.
Robinson, eds.). Smithsonian Institution Press,
Washington.

Eberhard, W. G. 1982. Behavioral characters for the
higher classification of orb-weaving spiders. Evo-
lution, 36:1067-1095.

Eberhard, W. G. 1988. Memory of distances and
directions moved as cues during temporary spiral
construction in the spider Leucauge mariana (Ara-
neae: Araneidae). J. Ins. Behav., 1:51-66.
Eberhard, W. G. 1990a. Function and phytogeny of
spider webs. Ann. Rev. Ecol. Syst., 21:341-372.
Eberhard, W. G. 1990b. Early stages of orb construc-
tion by Philoponella vicina, Leucauge mariana, and
Nephila clavipes spiders (Araneae: Uloboridae and
Tetragnathidae) and their phylogenetic implica-
tions. J. Arachnol., 18:205-234.

Eberhard, W. G. in press a. Notes on the behaviour
and ecology of Physocyclus globosus (Araneae, Phol-
cidae). Bull. British Arachnol. Soc.

Eberhard, W. G. in press b. Chrosiothes tonala (Ara-
neae, Theridiidae): a web-building spider special-
izing on termites. Psyche.

Eberhard, W. G. & R. D. Briceno 1983. Chivalry in

pholcid spiders. Behav. Ecol. Sociobiol., 13:189-
195.

Eberhard, W. G. & R. D. Briceno 1985. Behavior
and ecology of four species of Modisimus and Ble-
chroscelis (Pholcidae). Rev. Arachnol., 6:29-36.

Foelix, R. 1982. Biology of Spiders. Harvard Univ.
Press, Cambridge, Mass.

Forster, R. R. & L. M. Forster. 1973. New Zealand
Spiders. Collins, London.

Kaston, B. J. 1948. Spiders of Connecticut. Bull. State
Geol. Nat. Hist. Surv., 70:1-874.

Kirchner, W. 1 986. Das Netz der Zitterspinne (Phol-
cus phalangioides Fuesslin) (Araneae: Pholcidae).
Zool. Anz. 216:151-169.

Kullmann, E. 1964. Neue Ergibnisse uber den Netz-
bau und das Sexualverhalten einiger Spinnenarten.
Z. Zool. Syst. Evol., 2:41-122.

Lamoral, B. H. 1968. On the nest and web structure
of Latrodectus in South Africa, and some observa-
tions on body colouration of L. geometricus (Ara-
neae: Theridiidae). Ann. Natal Mus., 20:1-14.

Lehtinen, P. T. 1967. Classification of the cribellate
spiders and some allied families, with notes on the
evolution of the suborder Araneomorpha. Ann. Zool.
Fenn., 4:199-468.

Levi, H. W., L. R. Levi, & H. S. Zim. 1968, Spiders
and their Kin. Western Pub. Co., Racine, Wise.

Main, B. Y. 1976. Spiders. Collins, London.

Nielsen, E. 1931. The Biology of Spiders, Vol. 11.
Levin & Munksgaard, Copenhagen.

Nuessly, G. S. & R. D. Goeden 1984. Aspects of the
biology and ecology of Diguetia mojavea Gertsch
(Araneae, Diguetidae). J. Arachnol., 12:75-85.

Shear, W. 1969. Observations on the predatory be-
havior of the spider Hypochilus gertschi Hoffman
(Hypochilidae). Psyche, 76:407-417.

Shear, W. 1986. Taxonomic glossary. Pp. 403-432.
In Spiders— Webs, Behavior, and Ecology (W. Shear,
ed.), Stanford Univ. Press, Stanford, California.

Shinkai, E. 1990. Observations on the web structure
and predatory behavior in Ogulnius pullus Bos. et
Str. Atypus, 96:19-24.

Shinkai, E. 1984. A Fieldguide to the Spider of Japan.
Tokai Univ. Press, Tokai, Japan.

Szlep, R. 1 966. The web structure of Latrodectus var-
iolus Walckener and L. bishopi Kaston. Israel J.
Zool., 15:89-94.

Manuscript received May 1991, revised August 1991.

1992. The Journal of Arachnology 20:35-39

VARIATION IN SCHIZOCOSA (ARANEAE: LYCOSIDAE),
METAPHIDIPPUS AND PHIDIPPUS (ARANEAE: SALTICIDAE)'

William W. M. Steiner and Matthew H. Greenstone: Biological Control of Insects
Research Laboratory, USDA, Agricultural Research Service, P.O.B. 7629, Columbia,
Missouri 65205 USA

Gail E. Stratton: Department of Biology, Albion College, Albion, Michigan 49224 USA

ABSTRACT. Allozyme variation was examined in five species of solitary spiders collected in Illinois and
Missouri including Schizocosa ocreata (Hentz), S. stridulans Stratton, A. rovneri Uetz & Dondale, Phidippus
clams Keyserling, and Metaphidippus galathea (Walckenaer). The average number of alleles per locus was small,
and the average heterozygosity ranged from 2.3 to 12.5. The percent of polymorphic loci ranged from 16.6 to
66.7%. For one species, P. clams, a Missouri population is compared to a population from South Carolina
(Terranova & Roach 1987). Overall, the genetic variability estimates are lower than those for other North
American spiders. However, the observed genetic variation is approximately four times higher than that observed
in communal spiders.

Genetic variation is generally considered nec-
essary, if not crucial, to a species’ ability to adapt
to changing environmental conditions. Recently,
Terranova & Roach (1987) reported electropho-
retic variation at 1 2 isozyme loci in 7 species of
the solitary spider genus Phidippus collected from
South Carolina. They found high estimates of
variation with 41.6% of loci polymorphic and a
mean heterozygosity of 1 1.7%.

In contrast, Lubin & Crozier (1985) investi-
gated 21 enzyme systems and found only one
polymorphic isozyme locus in the social spider
Achaeamna wau Levi of New Guinea. The vari-
ation occurred in only 6 of 30 naturally-occurring
colonies. Thus 24 of the colonies displayed no
polymorphism in the form of variable enzyme
systems in the 2 1 enzymes which were examined.

Lubin and Crozier hypothesized that social
spiders will have lower variation as a result of
close inbreeding within the colony. Their work
seems to be supported by subsequent studies of
social spiders and what is known about inbreed-
ing and genetic variation in eusocial Hymenop-
tera (Graur 1985; Berkelhamer 1983; Reeve et
al. 1983; Owen 1983). Smith (1986) examined
51 electrophoretic loci in South American and
Central American populations of the social spi-
der Anelosimus eximius Simon and found seven
loci segregating. A mean heterozygosity of 0.060

'Any mention of a proprietary product in this article

does not indicate endorsement by the USDA-ARS.

was found, with only two of seven colonies sam-
pled showing within-colony polymorphism.
Roeloffs & Riechert ( 1 988) sampled 44 nests from
17 colonies of Agelena consociata Denis from
Gabon in Africa. In this social spider, the mean
genetic distance between nests belonging to dif-
ferent colonies was significantly higher than that
among nests of the same colony, suggesting that
nests within an area were part of a single pan-
mictic unit. The mean heterozygosity for 22 loci
was 0.0 1 8 but each nest was polymorphic at 5.5%
of loci examined. On the average, one of the
seven loci was found to be polymorphic in each
nest. Uetz et al. ( 1 986) studied Metepeira species
of communal-territorial spiders considered “in-
termediate” in social versus solitary behavior.
They found heterozygosity ranging from 0.09 to
0.21 in three species; variation was lowest in the
most communal species.

In this report, we determine the extent of ge-
netic variation in additional genera of solitary
spiders from the United States. If solitary spiders
can be shown to have low to nonexistent genetic
variation, then the Lubin-Crozier hypothesis is
weakened and explanations other than inbreed-
ing must be invoked.

METHODS

Spiders were caught individually and taken
alive to the USDA, ARS Biological Control of
Insects Research Laboratory (BCIRL) where they
were frozen at -80 °C until electrophoresis was

35

36

THE JOURNAL OF ARACHNOLOGY

Table 1 Allozyme loci, abbreviations and enzyme classification numbers assigned by the International Union
of Pure and Applied Chemistry (1973). A dehydrogenase is designated by DH.

Protein

Enzyme

Commission

Number

No. of loci
encoded

Abbreviation

Acid phosphatase

3. 1.3.2

2

ACPH

Adenylate kinase

2.7.4.3

1

ADK

Esterase

3. 1.1.1

4

EST

Glyceraldehyde-3-phosphate DH

1.2.1.12

1

G-3-PDH

a-glycerophosphate DH

1.1. 1.8

2

a-GPDH

Hexokinase

2. 7. 1.1

1

HK

Hydroxy-|8-butyric acid DH

no number

1

H-(3-BDH

Isocitrate DH

1.1.1.42

2

IDH

Malate DH

1.1.1.37

2

MDH

Phosphoglucomutase

2.7.5. 1

1

PGM

Phosphoglucose isomerase

5.3. 1.9

1

PGI

6-phosphogluconate DH

1.1.1.44

1

6-PGDH

Glutamate oxaloacetate transaminase

2.6. 1.1

2

GOT

performed. The Schizocosa species were taken at
Sand Ridge State Forest {S. ocreata and S. stri-
dulans) or at Chautaqua National Wildlife Ref-
uge {S. rovneri) in Mason County, Illinois. Phi-
dippus clams and M galathea were collected from
the University of Missouri’s Tucker Natural
Prairie Preserve, a tall grass prairie remnant lo-
cated along Interstate Highway 70 in Callaway
County, Missouri (Drew 1947).

Starch gel electrophoresis and histochemical
staining were performed as described by Steiner
& Joslyn (1979), to reveal the proteins listed in
Table 1 . Protein abbreviations used in this study
are also listed there. Sigma starch (Sigma Chem-
ical Co., St. Louis, Missouri) was used at a con-
centration of 1 2%, and the 1 cm thick gels were
horizontally-sliced into 1 mm thick gel slices for
differential enzyme staining.

Two electrode-gel buffer systems were used
(Steiner & Joslyn 1979). The first consisted of a
LiOH/Boric acid (pH 8.2) electrode buffer and a
Trizma base/citric acid (pH 8.4) gel buffer on
which PGI, PGM, EST and ACPH were stained.
The second consisted of a continuous Trizma
base/citric acid electrode buffer (pH 8.1) and gel
buffer (pH 8.5) system on which were stained
GOT, 6-PGDH, HK, ADK, a-GPDH, IDH, G-
3-PDH, and The allelic bases of the

observed isozyme variations were determined
using traditional crossing methods to study in-
heritance patterns and are presented elsewhere
(Steiner & Greenstone, 1991). Allele homologies
among the different species were not determined.

The derived genetic statistics are as follows.

Average number of alleles/locus (A/L) is deter-
mined by counting the total revealed by electro-
phoresis across all loci within a species and di-
viding by the number of loci examined. The
percent of loci polymorphic (%LP) is determined
by dividing those which are segregating for two
or more alleles by the total number of loci ex-
amined within a species. The mean observed per-
cent heterozygosity (%H) is determined by
counting the total number of heterozygotes across
all loci in a species, dividing by the product of
the number of specimans analyzed x number of
loci analyzed, and taking the dividend x 100. In
general, as the allele frequencies at a locus be-
come equal, the higher the %H will be. High
numbers of alleles (high A/L) and of loci poly-
morphic (high %LP) may be indicative of mu-
tation rate or heterotic selection but do not nec-
essarily correlate with %H. Because of the low
numbers of individuals analyzed for each species
in this study, we do not presume gene frequencies
will be in Castle-Hardy-Weinberg equilibrium,
which could also affect expected heterozygosity
estimates, since these numbers may be biased
due to sampling error.

RESULTS

A total of 2 1 isozyme loci coding for 1 3 en-
zymes was observed across all species. Table 2
indicates each species’ genetic profile. The num-
ber of loci observed within a species differs from
one species to the next, depending on what could
be resolved and interpreted. Thus in each of the
Schizocosa spp. 1 2 enzyme loci are examined for

STEINER ET AL.- GENETIC VARIATION IN FIVE SPIDER SPECIES

37

Table 2.— Summary of genetic variability for five American Midwest species of arachnids. In the variable
loci, a maximum of only 2 alleles were observed for any one locus except where indicated in parentheses. The
number of heterozygotes observed are listed over the sample size for that locus. Designations for loci which
show no or poor results are indicated for those wishing to pursue genetic investigations of these or other spider
species. Abbreviations: B = blurry, not scorable; NA = not analyzed; NP = not present after staining; V =
variable but not scorable; see Table 1 for locus abbreviations.

Protein locus

Schizocosa

ocreata

Schizocosa

rovneri

Schizocosa

stridulans

Phidippus

clarus

Metaphidippus

galathea

ACPH-1

NA

NA

NA

1/22

0/27

ACPH-2

NA

NA

NA

nil

0/27

ADK-1

NA

NA

NA

0/22

5/27

EST-1

0/10

2/11

0/5

NP

9/27

EST-2

0/10

2/10

0/5

V, B

V, B

EST-3

NP

NP

NP

V, B

V, B (3)

EST-4

NP

1/10

NP

NP

V, B

a-GPDH-l

0/10

3/12

0/5

0/22

11/27

a-GPDH-2

NP

NP

NP

NP

0/27

GOT-1

B

B

B

0/22

9/27

GOT-2

B

B

B

0/22

0/27

G-3-PDH

0/10

0/12

0/5

0/22

0/27

H-/3-BDH

NA

NA

NA

0/22

NA

HK-1

0/10

1/12

0/5

0/22

0/27

IDH-1

0/10

NA

1/5

nil

4/27

IDH-2

0/10

0/12

0/5

NP

0/27

MDH-1

4/10

4/12

1/5

0/22

0/27

MDH-2

0/10

0/12

1/5

0/22

0/27

PGI

0/10

1/12

0/5

B

0/27

PGM

1/10

0/9

0/4

nil

0/27

6-PGDH

1/10

4/12 (3)

0/5

0/22

6/27

Alleles/locus

1.25

1.66

1.25

1.20

1.25

% polymorphic

16.60

66.70

25.00

37.50

45.00

% heterozygosity

5.00

12.50

5.00

2.30

9.60

variation compared to 1 6 in P. dams and 20 in
M. galathea. The most variable systems were the
esterases, but band overlap between loci on the
same gel slice sometimes prevented accurate
scoring of a particular esterase system. Only loci
which could be reliably scored as heterozygous
or homozygous were used to develop genetic pro-
files. 'Yh\isADK-2 and HK-2 were not scored and
were ignored as they showed up inconsistently
and then with only a trace of activity on the gels.
Only those esterase loci which could be clearly
seen and scored for heterozygosity were included
in the final estimate for percent loci polymorphic
(Table 2).

At most, we did not observe more than two
alleles at any polymorphic locus in any one spe-
cies with the exception of EST-3 in M. galathea
and 6-PGDH in 5”. rovneri which had 3 alleles
each (parentheses. Table 2). The average number
of alleles/locus ranged from 1.20-1.66. The per-
cent of loci segregating for two or more alleles

ranged from 16.6-66.7 with an average around
38.67%. The mean observed heterozygosity
ranged from 5.0-12.5 depending on the species
examined, with an overall mean of 6.8.

DISCUSSION

The percent of loci polymorphic which we
found in the Missouri P. dams is listed in Table
2 and is 37.5%, very similar to the average of
4 1 .6% observed for all 7 Phidippus species in the
study by Terranova & Roach (1987). However,
our average heterozygosity is only 2.3% for P.
dams and does not come close to the average
variability of 1 1.8% seen in the South Carolina
populations of this species.

The low levels of variation we observe in the
Missouri P. dams is in sharp contrast to that
observed in South Carolina. A third as many loci
are polymorphic in the 12 loci Terranova and
Roach examined compared to the 16 loci we can
score for the presence of variation. The variation

38

THE JOURNAL OF ARACHNOLOGY

observed in the South Carolina population oc-
curs at PGI,AAT, IDH-1, MDH-1, MDH-2 and
amylase, while that occurring in Missouri is at
the PGM, IDH-1, EST-2, EST-3, ACPH-1 and
ACPH-2 loci. The lower heterozygosity seen in
Missouri P. darns is one fifth that of South Car-
olina and is probably due to the polymorphic
loci in the Missouri population having lower gene
frequencies for alternative alleles.

Such differences in population genetic struc-
ture are indicative of adaptive processes at work,
and leave room for further question and study.
It may be that Midwest populations suffer more
often from severe, climatically induced, bottle-
necks in population density which are a conse-
quence of harsher winters. Even cultural prac-
tices of farmers might play a role if insecticide
use is higher in, say, the Midwest, acting to re-
duce genetic variation through direct selection
pressure. These possiblities could explain the dif-
ferences we observe in Missouri P. darns and
can be contrasted with effects due to more in-
herent phenomena such as breeding structure.
For example, Lubin (pers. commun.) points out
that low levels of variation might be a conse-
quence of breeding system as seen in solitary
wasps.

Given the small sample sizes, the five Midwest
species of solitary spiders have almost four times
the genetic variation observed in A. wan by Lu-
bin & Crozier (1985). Other solitary spiders, in-
cluding the genera Meta (Pennington 1 979), Nes-
ticns (Cesaroni et al. 1981), and Aranens
(Manchenko 1981), have relatively high levels
of genetic variability reflected as isozyme poly-
morphism as well. These results tend to support
the Lubin-Crozier hypothesis. Testing further the
robustness of the hypothesis requires more elec-
trophoretic studies of social spiders from more
diverse geographic areas and disparate temperate
zones and a study of solitary spiders from the
area where A. wan is endemic. Certainly the range
of genetic variation can vary greatly within a
genus as we see here in Schizocosa and as Ter-
ranova & Roach (1987) observed in Phidippus.
A more meaningful approach might be to look
at closely related social and non-social spiders
(e. g., see Smith 1987).

Finally, we would point out the significance of
having similar levels of variation in S. stridulans
and S. ocreata. These species often occur sym-
patrically and may share certain life history strat-
egies. Similarities in variability between sym-
patrically-occurring species is consistent with the

idea that micro-evolutionary or adaptive pro-
cesses transcend species status.

ACKNOWLEDGMENTS

We thank Clyde Morgan for his expertise in
caring for the Phidippns and Metaphidippus sam-
ples. We also thank the Department of Biological
Sciences, University of Missouri, Columbia for
permission to work at Tucker Prairie, and Felix
Breden, Yael Lubin, Ian McDonald, Susan
Riechert and George Uetz for comments and
suggestions, although the final interpretations
must remain our own.

LITERATURE CITED

Cesaroni, C., G. Allegrucci, M. Caccone, M. Sbordoni,
E. De Matthaeis & 1. Sbordoni. 1981. Genetic
variability and divergence between populations of
species of Nesticus cave spiders. Genetica, 56:81-
92.

Berkelhamer, R. C. 1983. Intraspecific genetic vari-
ation and haplodiploidy, eusociality, and polygyny
in the Hymenoptera. Evolution, 37:540-545.

Drew, W. B. 1947. Floristic composition of grazed
and ungrazed prairie vegetation in North-Central
Missouri. Ecology, 28:26-41.

Graur, D. 1985. Gene diversity in Hymenoptera.
Evolution, 39: 190-199.

International Union of Pure and Applied Chemistry
and the International Union of Biochemistry. 1973.
Enzyme Nomenclature. Elsevier Scientific Publ. Co.,
Amsterdam, pp. 443.

Lubin, Y. D. & R. H. Crozier. 1985. Electrophoretic
evidence for population differentiation in a social
spider Achearanea wau (Theridiidae). Insectes Soc.
Paris, 32:297-304.

Manchenko, G. P. 1981. Allozymic variation in Ar-
aneus ventricosns (Arachnida, Aranei). Isozyme Bull.,
14:78.

Owen, R. E. 1983. Difficulties with the interpretation
of patterns of genetic variation in the eusocial Hy-
menoptera. Evolution, 39:201-205.

Pennington, B. J. 1979. Enzyme genetics in taxon-
omy: diagnostic enzyme loci in the spider genus
Meta. Bull. British Arachnol. Soc., 4:377-392.
Reeve, H. K., J. S. Reeve & D. W. Pfennig. 1983.
Eusociality and genetic variability: a re-evaluation.
Evolution, 39: 200-201.

Roeloffs, R. & S. E. Riechert. 1988. Dispersal and
population genetic structure of the cooperative spi-
der, Agelena consodata, in West African rain forest.
Evolution, 42:173-183.

Smith, D. R. 1986. Population genetics of Anelosi-
mus eximius (Araneae, Theridiidae). J. Arachnol.,
14:201-217.

Smith, D. R. 1987. Genetic variation in solitary and
cooperative spiders of the genus Anelosimus (Ara-
neae: Theridiidae). In Chemistry and Biology of So-

STEINER ET AL.- GENETIC VARIATION IN FIVE SPIDER SPECIES

39

cial Insects. Proc. 10th Int. Congr. lUSSI (J. Eder
& H. Rembold, eds.). Verlag J. Pepemgy, Munich.

Steiner, W. W. M. & D. J. Joslyn. 1979. Electropho-
retic techniques for the genetic study of mosquitoes.
Mosq. News, 39:35-54.

Steiner, W. W. M. & M. H. Greenstone. 1991. Seg-
regation studies of isozyme variation in Metaphid-
dippus galathea (Araneae: Salticidae). J. ArachnoL,
19:157-160.

Terranova, A. C. & S. H. Roach. 1987. Genetic dif-
ferentiation in the genus Phidippus (Araneae, Sal-
ticidae). J. ArachnoL, 14:385-391.

Uetz, G. W., T. C. Kane, G. E. Stratton & M. J. Benton.
1986. Environmental and genetic influences on the
social grouping tendency of a communal spider.
Pp.43-53. /« Evolutionary Genetics of Invertebrate
Behavior; Progress and Prospects (M. D. Huettel,
ed.). Plenum Press, New York.

Manuscript received January 1991, revised December
1991.

1992. The Journal of Arachnology 20:40^6

SYSTEM ATICS OF HYPOCHILUS SHEARI AND
HYPOCHILUS COY LEI, TWO SOUTHERN APPALACHIAN
LAMPSHADE SPIDERS (ARANEAE, HYPOCHILIDAE)

Ronald P. Huff and Frederick A. Coyle: Department of Biology, Western Carolina
University, Cullowhee, North Carolina 28723 USA

ABSTRACT: A quantitative analysis of variation of genital characters among 16 population samples supports
Platnick’s (Forster et al. 1987) hypothesis that Hypochilus sheari Platnick and Hypochilus coylei Platnick are
separate species. Both the width and tip length of the conductor separate these species unambiguously, and an
index of spermathecal dimensions distinguishes the otherwise very similar females of each species. A conductor
tip synapomorphy indicates that these are probably sister species. Extensive field work indicates that these species
are confined to a 35 mile north-south mountain corridor in six counties of western North Carolina and are
separated by five miles. It is suggested that subtle deficits in habitat due to a constriction of the mountain
corridor and the presence of a watershed divide may have bisected the parent stock and may be maintaining
the allopatry of the daughter species. Both species appear to have a two year life cycle and commonly to feed
on gryllacridid crickets and cursorial spiders.

In a review of the hypochiloid and austrochi-
loid spiders (Forster et al. 1987), Platnick de-
scribed two similar species of Appalachian Hy-
pochilus which he separated on the basis of minor
differences in genital morphology observed in
small samples. Hypochilus coylei was described
from ten males and nine females from three
neighboring sites, and H. sheari was described
from three males and three females from a single
site 27 mi from the H. coylei sites. The primary
goals of this study are (1) to test more rigorously
Platnick’s hypothesis that these two morphs are
different species and the alternative hypothesis
that his samples might simply represent geo-
graphic variants of a single species, and (2) to
define accurately their geographic ranges. More-
over, we hope this study will foster the kinds of
research and management decisions needed to
protect the habitat of these unusual (Forster et
al. 1987) and geographically restricted spiders.

METHODS

During 1987 (August to November) and 1988
(March to November) we searched for Hypochi-
lus in the region around the sites where these
species were first collected. When a deme was
encountered, we collected only as many adults
of each sex as we felt we could without threat-
ening the deme’s survival.

Material Examined.-Collection data for all speci-
mens used in this study are listed below by species and

county (all in North Carolina). Specimens will be de-
posited in the AMNH. Each site (deme) is identified
on the map (Fig. 12) by its letter/number code.
Throughout this report we use the term “deme” for
each local population, and the terms “morph” or “spe-
cies” for all demes (collectively) of each putative spe-
cies.

Hypochilus sheari: BUNCOMBE CO.; S/ — Long
Branch Creek, 0.8 mi N on Long Branch Road from
Bee Tree Road, rock outcrop along road on right, elev.
3100 ft, 28 Sept. 1987, 3 males, 5 females; 6 Oct. 1987,
1 male, 1 female; 15 Oct. 1988, 10 males, 13 females,
1 egg sac; MCDOWELL CO.; 5'2-Buck Creek, 9.9
mi N of US 70 on NC 80, dry exposed rock face on
right of highway, elev. 3400 ft, 1 Sept. 1987, 10 males,
5 females; 5i— Curtis Creek, 5 mi up Curtis Creek road
from US 70, rock outcrop on left of road, elev. 2500
ft, 13 Sept. 1987, 3 males, 1 female; 16 July 1988, 1
male, 2 females; 24 Sept. 1988, 1 male, 2 females, 1
egg sac; 54— Mill Creek, 0.2 mi SE of Andrews Geyser,
large rock outcrop on left, elev. 3100 ft, 29 Sept. 1987,
1 male, 3 females; 6 Oct. 1987, 12 males, 15 females;
24 Sept. 1988, 3 males, 5 females, 1 egg sac; 55 —
Newberry Creek, 1.5 mi W on Newberry Creek from
Curtis Creek confluence, rock outcrop on either side
of road, elev. 2400 ft, 13 Sept. 1987, 7 males, 10 fe-
males; 24 Sept. 1988, 3 males, 2 females; YANCEY
CO.; 56— Upper Crabtree Falls, Blue Ridge Parkway,
elev. 3200 ft, 1 Sept. 1987, 6 males, 5 females.

Hypochilus coylei: BUNCOMBE CO.; C7— Garren
Creek, approx. 0.5 mi up Owenby Gap Road from
intersection of Eads Gap Road, disturbed rock above
stream on left, elev. 3100 ft, 19 Sept. 1987, 1 male, 4
females; C2— Round Mountain, 4 mi NE of NC 9 on
Bat Cave Road, rock outcrops on left above road, elev.

40

HUFF & COYLE-SYSTEMATICS OF TWO HYPOCHILUS SPECIES

41

Figures \-l .—Hypochilus genitalia: 1-2, H. sheari male palpus (traced from Forster et al. 1987) showing
measurement characters; 1, retrolateral; 2, prolateral; 3-6, approximately retrolateral views of palpal conductor;
3-4, H. coylei showing measurement characters; 3, from C7; 4, from C6; 5-6, H. shearr, 5, from SI; 6, from
S3; 7, H. sheari spermathecae from S4 showing measurement characters.

3200 ft, 9 Oct. 1988, 2 males, 2 females, 1 egg sac;
HENDERSON CO.; Ci- Hickory Creek, 1 mi SE
Gerton on US 74, rock outcrop on either side of road,
elev. 2575 ft, 13 Aug. 1987, 9 females; 3 Oct. 1987, 1
female; 6 Oct. 1987,2 females; 1 4 Oct. 1988,2 females;
C4 — Minihaha Falls, 0.8 mi N on NC 9 from US 74,

rock outcrop below road on left, elev. 1600 ft, 13 Aug.
1987, 2 males, 5 females; C5— Reedy Patch Creek,
10.5 mi E of 1-26 on US 64, rock face above creek on
right, elev. 2000 ft, 21 Aug. 1987, 1 male, 2 females;
C6— Tumbreeches Creek, approx. 1 mi SW of US 64,
rock outcrop on left, elev. 2400 ft, 3 Oct. 1987,6 males.

42

THE JOURNAL OF ARACHNOLOGY

Table 1 . — Descriptive statistics for the quantitative characters found most useful in distinguishing Hypochilus
sheari from Hypochilus coylei. Character abbreviations defined in Methods section. Measurements in mm.

Hypochilus sheari Hypochilus coylei

Character

n

Range

Mean

SD

n

Range

Mean

SD

Male

CTL

56

0.075-0.120

0.093

0.010

43

0.045-0.065

0.057

0.005

PCW

50

0.045-0.068

0.058

0.005

41

0.075-0.113

0.093

0.008

PCW(100)/CTL

50

41.7-83.9

62.5

8.3

41

132-221

164

19.8

Female

CW

69

2.50-3.75

3.11

0.24

57

2.85^.40

3.65

0.33

LSL

69

0.110-0.263

0.179

0.032

50

0.160-0.348

0.251

0.040

MXMD

68

0.103-0.170

0.135

0.016

48

0.070-0.128

0.101

0.013

MXMD(100)/LSL

68

57.7-131.8

77.5

14.4

48

28.8-55.6

40.9

7.1

MXMD(100)/CW

66

3.3-6. 1

4.4

0.6

48

2.2-3. 3

2.8

0.3

3 females; POLK CO.; C7— Cliflield Mountain, 2.5 mi
W of Mountain View Church, rock faces, elev. 1300
ft, 3 Sept. 1988, 5 males, 8 females; RUTHERFORD
CO.; C(S— Bat Cave Nature Preserve, 1 mi SE of Bat
Cave village, cliffs below cave, elev. 1600-1800 ft, 3
Oct. 1987, 2 males, 3 females; C9— Broad River, 1.3
mi E of NC 9 on US 74 (0.3 mi NW of start of Rainbow
Falls Trail), disturbed rock on either side of river, elev.
1300 ft, 3 Sept. 1988, 5 males, 4 females; C70— Rain-
bow Falls Trail, 1 .6 mi E of NC 9 on US 74, disturbed
rock along trail, elev. 1500-2200 ft, 13 Aug. 1987, 6
males, 8 females; 6 Oct. 1987, 13 males, 3 females; 28
May 1988, 4 females; 1 Nov. 1988, 1 egg sac.

Characters Examined: — In order to identify
characters that might distinguish one or more
demes, we focused on genital characters for two
reasons: 1) they tend to evolve rapidly and di-
vergently (Eberhard 1985, 1990), and 2) Platnick
(Forster et al. 1987) used genital characters to
differentiate H. coylei and H. sheari. We selected
potentially useful genital shape characters, de-
fined measurements which alone or in ratio form
would represent these shapes, recorded the val-
ues of these characters for every specimen, and,
with the Statview II (Abacus Concepts, Inc.) sta-
tistics program, generated scatter plots to identify
any clusters of individuals and/or demes with
distinctive values. Carapace width (CW) was used
in some of these bivariate analyses to control for
body size; CW measurements are accurate to 0.05
mm. A few other potentially useful non-genital
characters (e.g. pigmentation patterns) were also
surveyed.

Male quantitative characters.— The following
eight palpal dimensions were measured after re-
moving the palp: PL = length of palpal organ in
retrolateral view (Fig. 1); PBD = diameter of

palpal bulb in retrolateral view (Fig. 1); PTW =
width of palpal tarsus perpendicular to PTL at
distoventral apophysis in retrolateral view (Fig.
1); PTL = length of palpal tarsus in retrolateral
view (Fig. 1); CDL = conductor length in pro-
lateral view (Fig. 2); CTL = conductor tip length
in approximate retrolateral view (positioned so
CTL is maximized), defined as the distance from
the conductor tip to the edge of the conductor
along the line perpendicular to the line connect-
ing the tip of the conductor to the dorsal edge of
conductor (Fig. 3); CTH = conductor tip height
in retrolateral view, defined as the maximum
distance to lower edge of conductor tip from line
defined by CTL and perpendicular to CTL (Fig.
3); PCW = palpal conductor width in retrolateral
view, measured along the line originating at the
midpoint of, and perpendicular to, a line defined
as the maximum distance from point where con-
ductor edges overlap to inside edge of conductor
tip (Fig. 4). CTL, CTH, and PCW were measured
with the palpal organ in glycerine under a cov-
erslip on a depression slide at 250 x using a Wild
M-20 compound microscope with an ocular ret-
icle with two perpendicular scales; these mea-
surements are accurate to 0.0025 mm. The other
measurements were made at 100 x using a Wild
M5 stereomicroscope and are accurate to 0.009
mm.

Female quantitative characters. —The follow-
ing seven spermathecal dimensions were mea-
sured (Fig. 7): LSL = lateral stalk length, defined
as the maximum distance from junction of lateral
and median stalk to end of lateral stalk; MXLD
= maximum lateral stalk diameter perpendicular
to longitudinal axis of stalk; MNLD = minimum
lateral stalk diameter perpendicular to longitu-

HUFF & COYLE- SYSTEM ATICS OF TWO HYPOCHILUS SPECIES

43

dinal axis of stalk; MSL = median stalk length;
MXMD = maximum median stalk diameter;
MNMD = minimum median stalk diameter;
MXD = maximum distance from medial edge
of median stalk to lateral edge of lateral stalk.
The portion of body wall containing spermathe-
cae was removed, cleared, and mounted in lactic
acid on a glass slide under a coverslip. Measure-
ments were recorded from the right spermathe-
cae at lOOx using a Bausch and Lomb com-
pound microscope and are accurate to 0.005 mm.

RESULTS

Morphological Variation.— Ma/e characters:
T-tests reveal that the two morphs are signifi-
cantly dilferent (P < 0.001) for each of the eight
palpal characters. Although six of the eight show
broad overlap among demes and between
morphs, PCW and CTL clearly distinguish the
two species (Table 1, Fig. 8). Hypochilus coylei
(Figs. 3-4) has a wider conductor (PCW) and a
shorter, more sharply bent conductor tip (CTL)
than H. sheari (Figs. 5-6). When plotted against
each other, these characters produce better sep-
aration of the morphs (Fig. 8) than do any other
pair of characters. Only one of the pigmentation
differences described by Platnick (Forster et al.
1987) helps separate the larger samples we have
studied; the pigment patches on the edges of the
sternum are confluent in H. coylei but in H. sheari
are usually disjunct and seldom appear as a con-
tinuous band. The small finger-like apophysis
which we discovered extending from the retro-
dorsal edge of the conductor near its distal end
is usually much longer and easier to see (at 1 00 x )
in H. coylei (Figs. 3-4) than in H. sheari.

Female characters: T -tests showed that the two
morphs are significantly different (P < 0.00 1) for
each spermathecal measurement except MXD.
However, for each of these characters, there is
broad overlap between the two morphs and
among the demes. Nevertheless, highly confident
identification of females is possible when the two
characters with the least overlap, MXMD and
LSL (T able 1 ), are plotted against each other (Fig.

9) and when MXMD is plotted against CW (Fig.

10) . Hypochilus sheari has, on average, wider
median bulbs, shorter lateral spermathecae, and
a smaller carapace than H. coylei (Fig. 1 1). Nei-
ther of the pigment differences described by Plat-
nick (Forster et al. 1987) (endite pigmentation,
palpal tarsal pigment ring) appear to distinguish
any of the demes or separate the morphs.

Clinal variation: With one exception (LSL), no

0.12

0.10

PCW 0.08

0.06

0.04

o o8°o coylei
^ R °°L8

8

. sheari

•• • ••• •

• ti ••

0.04 0.06 0.08 0.10

CTL

0.12

LSL

MXMD

4.6
4.2

3.8

CW 3.4
3.0

2.6

0.06 0.10 0.14 0.18

MXMD

Figures 8-10.— Scattergrams of measurement char-
acters which best distinguish Hypochilus sheari from
H. coylei. Characters defined in Methods section. Mea-
surements in mm. 8, males; 9-10, females.

clinal pattern was found when the mean of each
quantitative character for each deme was ranked
(low to high) and mapped. LSL means increased
from north to south in H. sheari, but showed no
clinal pattern in H. coylei.

Habitat and Distribution.-The preferred (most
heavily populated) web substrate and habitat for
both morphs appears to be the same as that of
Hypochilus pococki Platnick (Fergusson 1972).
Webs are nearly always on vertical or overhang-
ing surfaces of rock outcrops and boulders which
are typically beside or near a stream in deciduous
or mixed deciduous/conifer forest. With the ex-
ception of two demes (S2 and C7), these spiders

44

THE JOURNAL OF ARACHNOLOGY

Figure 1 1 . — Spermathecae of Hypochilus sheari and H. coylei selected to illustrate range of variation; arranged
so that MXMD(100)/LSL values decrease from left to right and from top row to bottom row. Codes identify
demes (see Methods section).

were seldom found on relatively exposed and dry
roek outcrops. Curiously, areas of seemingly fa-
vorable habitat but devoid of Hypochilus were
found within the known geographic range of each
species (Fig. 12).

Both H. sheari and H. coylei are confined to a
35 mile north-south mountain corridor that is
bounded on the west by the relatively flat and
unforested basin of the French Broad River, on
the east by low (below 1400 ft elev.), relatively
flat Piedmont terrain, and on the north and south
by allopatric populations of H. pococki (Fig. 1 2).
The H. sheari demes are confined to the northern
half and the H. coylei demes to the southern half
of this corridor. The two species are separated
by an uninhabited five mile zone that includes
the divide between their respective watersheds.
All known H. sheari demes are in the Catawba
River watershed (Atlantic drainage), except for
SI, which lies in the French Broad watershed
(Gulf drainage). All H. coylei demes are in the
Broad River and Green River watersheds (At-
lantic drainage).

A search in the zone that separates the ranges
of H. sheari and H. coylei revealed no sites with
preferred substrate and habitat; and although
several sites with suboptimal but apparently suit-
able substrate and habitat were found, Hypochi-
lusnot present. Even such seemingly suitable
habitat is rare in this zone because here the

mountain corridor is constricted to a width of
only a few miles by convergent lobes of the French
Broad basin and Piedmont (Fig. 1 2) and includes
a divide separating the watersheds occupied by
the two species.

Life History and Reproductive Biology.— Like
H. pococki (Fergusson 1972, Coyle 1985), these
species appear to have a two-year life cycle. Me-
dium to large juveniles (that had overwintered)
constructed webs as early as March 15; these
individuals appeared to be large enough to reach
maturity later in the year. Adult females began
to appear in late May and adult males in late
July. Adult males disappeared after early Octo-
ber and adult females by the end of October.
Juveniles were recorded until late November.
New egg sacs were observed from early Septem-
ber until late October. Spiderlings emerged from
egg sacs in early May; no occupied egg sacs were
recorded later than mid-May.

Adult males were often found just outside the
webs of females. On several occasions (both spe-
cies and only after mid-September) single males
were observed motionless in webs of adult fe-
males, with the male either directly over the fe-
male or with the legs of one side held across her,
reminiscent of the post-mating position ob-
served by Fergusson (1972). Three H. sheari egg
sacs contained 44, 74, and 87 eggs, and two H.
coylei sacs held 62 and 93 eggs.

HUFF & COYLE- SYSTEM ATICS OF TWO HYPOCHILUS SPECIES

45

Figure 12.— Map of a ten county region of western North Carolina showing the known distribution of Hy-
pochilus sheari and H. coylei and neighboring sites where H. pococki has been collected. Each sampled deme of
H. sheari and H. coylei is identified with code (see Methods section). X’s mark sites with apparently suitable
Hypochilus habitat but found to be devoid of Hypochilus. French Broad River basin (stippled) ranges in elevation
from about 2000 to 2400 ft. Boundary between mountains (not stippled) and Piedmont (stippled) roughly follows
elevation of 1400 ft.

Feeding Biology.— Feeding was observed sev-
eral times in both morphs (H. coylei: n = 6, H.
sheari: n = S) and resembles that of H. pococki
(Fergusson 1972). The primary prey type was
gryllacridid crickets (//. coylei: n = 3; H. sheari:
n = 5) and cursorial spiders, including lycosids
(H. sheari: n = 2), gnaphosids {H. coylei: « = 1),
and pisaurids {H. coylei: n = 2; H. sheari: n =
1). No spiders were observed feeding on tipulid
flies even though they were abundant at many
demes.

DISCUSSION

The results of this study provide strong sup-
port for Platnick’s contention that these two Hy-

pochilus morphs represent distinct species. The
distinct genital morphologies of the two popu-
lation clusters (Table 1, Figs. 3-6, 8-10), the vir-
tual absence of clinal variation in the characters
studied (the demes closest to the geographic gap
separating the species are as distinct morpholog-
ically as those farthest from the gap), and the
absence of hybrid populations indicate that there
is no gene flow across the small geographic gap
separating them. Whether these allopatric mor-
phospecies are also separated by intrinsic repro-
ductive isolating mechanisms, and are therefore
biological species sensu Mayr ( 1 969), Dobzhan-
sky et al. (1977), and Futuyma (1986), can be
determined only by courtship and mating trials

46

THE JOURNAL OF ARACHNOLOGY

in laboratory situations with appropriate intra-
morph controls. However, the fact that the di-
agnostically most useful genital features (con-
ductor width and conductor tip shape) are those
which contact the female most intimately during
copulation, hints that H. sheari and H. coylei
cannot successfully interbreed. Male genitalic dif-
ferences probably are not just manifestations of
genetic divergence but may be important in in-
fluencing the female’s acceptance of sperm (Eber-
hard 1985, 1990) and, consequently, vital parts
of the actual mechanism of reproductive isola-
tion.

Among the many shared character states of H.
sheari and H. coylei, there is at least one probable
synapomorphy which supports the hypothesis
that these are sister species: the conductor tip is
abruptly and strongly bent and tapered so that
it resembles a beak (Figs. 3-6). We postulate that
some event, perhaps divide migration and sub-
sequent drainage capture, a common event in
the Southern Blue Ridge during Tertiary times
(Hack 1969), divided the parent species into two
geographic isolates distributed much like they
are today.

It is not obvious why Hypochilus demes do
not exist in the five mile zone separating these
species, but we suggest that, because of the drastic
constriction of the mountain corridor and the
presence of a watershed divide, this region does
not contain enough favorable substrate and hab-
itat to support persistent populations and allow
dispersal. Hypochilus species have not been ob-
served to balloon, and observations by Shear
(1969) and Fergusson (1972) indicate that only
adult males are highly likely to walk from one
outcrop to another.

ACKNOWLEDGMENTS

We thank R. Bruce and D. Pittillo for com-
ments on an early draft and R. Bennett and N.

Platnick for comments on the manuscript. A

Highlands Biological Station grant-in-aid to Huff

helped support the field work.

LITERATURE CITED

Coyle, F. A. 1985. Two-year life cycle and low palpal
character variance in a Great Smoky Mountain pop-
ulation of the lampshade spider (Araneae, Hypo-
chilidae, Hypochilus). J. Arachnol., 13:21 1-218.

Dobzhansky, T., F. J. Ayala, G. L. Stebbins & J. W.
Valentine. 1977. Evolution. W.H. Freeman & Co.,
San Francisco, 572 pp.

Eberhard, W. G. 1985. Sexual Selection and Animal
Genitalia. Harvard Univ. Press, Cambridge, Mass.
244 pp.

Eberhard, W. G. 1990. Animal genitalia and female
choice. American Sci., 78:134-141.

Fergusson,!. 1972. Natural history of the spider //y-
pochilus thorelli Marx (Hypochilidae). Psyche, 79:
179-199.

Forster, R. R., N. 1. Platnick & M. R. Gray. 1987. A
review of the spider superfamilies Hypochiloidea
and Austrochiloidea (Araneae, Araneomorphae).
Bull. American Mus. Nat. Hist., 185:1-1 16.

Futuyma, D. J. 1986. Evolutionary Biology. Sinauer
Assoc., Sunderland, Mass., 600 pp.

Hack, J. T. 1969. The area, its geology: Cenozoic
development of the Southern Appalachians. Pp. 1-
17, In The Distributional History of the Biota of
the Southern Appalachians, Part 1 -Invertebrates (P.
C. Holt, ed.). Virginia Polytechnic Institute, Blacks-
burg, VA.

Mayr, E. 1963. Animal Species and Evolution. The
Belknap Press, Harvard Univ. Press, Cambridge,
Mass., 797 pp.

Shear, W. A. 1969. Observations on the predatory
behavior of the spider Hypochilus gertschi Hoffman
(Hypochilidae). Psyche, 76:407-^17.

Manuscript received September 1991, revised October
1991.

1992. The Journal of Arachnology 20:47-51

ON THE FUNCTION OF HARLEQUIN BEETLE-RIDING IN THE
PSEUDOSCORPION, CORDYLOCHERNES SCORPIOIDES
(PSEUDOSCORPIONIDA: CHERNETIDAE)

David W. Zeh and Jeanne A. Zeh: Smithsonian Tropical Research Institute, Unit
0948, APO 34002-0948 USA, or, Apartado 2072, Balboa, Republica de Panama

ABSTRACT. The pseudoscorpion, Cordylochernes scorpioides, frequently occurs under the elytra of the
giant harlequin beetle, Acrocinus longimanus. Here, we assess four hypotheses/scenarios which have been pro-
posed to account for this phenomenon: (1) accidental boarding; (2) obligate symbiosis; (3) phagophily, and (4)
phoretic dispersal. Field and laboratory observations of embarkation behavior clearly refute the accidental
boarding hypothesis. Contrary to the obligate symbiosis scenario, pseudoscorpion offspring production does not
occur on the beetle and the primary habitat of C. scorpioides is decaying trees. The phagophily hypothesis, i.e.,
that pseudoscorpions mount harlequins for the primary purpose of preying upon the beetles’ phoretic mites, is
also not supported. Pseudoscorpions collected from trees were found to be in better nutritional condition than
beetle-riding individuals. Finally, evidence from a companion study supports the dispersal hypothesis, and also
indicates that large male C. scorpioides defend beetles’ abdomens as strategic sites for intercepting and insem-
inating dispersing females.

The reason why pseudoscorpions attach to
other organisms (generally termed phoresy) is the
subject of much debate and little data (see Much-
more 1971). In this paper, we evaluate several
competing hypotheses put forward to explain the
significance of the association between the cher-
netid pseudoscorpion, Cordylochernes scor-
pioides (L.), and the giant harlequin beetle, Ac-
rocinus longimanus (L.) (Cerambycidae).
Cordylochernes scorpioides is distributed
throughout the tropical forests of Central and
South America (Beier 1 948) where it is frequent-
ly found under the elytra of harlequin beetles
(Beier 1948; Beck 1968; Muchmore 1971; Zeh
& Zeh 1991, 1992). This cerambycid also carries
mites, occasionally in large numbers, both in
small pits on the outer surface of its fore-elytra,
and on its thorax, wings and abdomen (Fig. 1).
It has been hypothesized that C. scorpioides
climbs onto beetles; (1) accidentally; (2) for dis-
persal to new habitats, or (3) for “phagophily”,
i.e., to feed on the mites (for review see Much-
more 1971). The harlequin beetle/pseudoscor-
pion relationship has even been presented as ob-
ligate symbiosis in which the pseudoscorpions
live exclusively on the beetles (Ricklefs 1979).

Our three-year study of C. scorpioides (Zeh &
Zeh 1991, 1992) and A. longimanus (Zeh et al.
1992) has demonstrated that the primary habitat
of this pseudoscorpion is decaying trees in the

families Moraceae and Apocynaceae (e.g., Ficus
spp. L. and Parahancornia fasciculata (Poiret)).
Larval development of A. longimanus also oc-
curs in these habitats (Duffy 1 960; G. Tavakilian,
pers. comm.; pers. obser.). While all pseudo-
scorpion life stages were collected from trees, no
nymphs and only one newly-gravid female {N =
1 34 females) were taken from beetles, indicating
that pseudoscorpion presence on beetles is strict-
ly an adult phenomenon.

Field and laboratory observations in Panama
and French Guiana were not consistent with the
“accident” hypothesis. Pseudoscorpion embar-
kation involves a sequence of deliberate, stere-
otypical behaviors (Beck 1968) which appears to
be triggered by olfactory cues and beetle strid-
ulation (pers. obser.). Both in the field and when
placed in laboratory containers with harlequin
beetles, pseudoscorpions engaged in the char-
acteristic lifting of pedipalps (so-called “beck-
oning” movements, see Weygoldt 1969) before
moving rapidly to the posterior end of the beetle.
There, pseudoscorpions gained access to the “sub-
elytral space” by repeatedly pinching the beetle’s
abdomen, causing abdominal flexure and partial
opening of the elytra (Fig. 2). In addition, in
decaying Ficus trees, C. scorpioides occurs with
several other pseudoscorpion species, e.g., Lus-
trochernes sp. Beier, Parachernes plumosus (With)
and Semeiochernes armiger (Balzan). We have

47

48

THE JOURNAL OF ARACHNOLOGY

Figure 1. — Harlequin beetle with elytra opened to display female Cordylochernes scorpioides and unusually
heavy infestation of mites.

never found individuals of these species under
the elytra of harlequin beetles {N = 149 beetles).

By contrast, results of our research clearly sup-
ported the dispersal hypothesis, demonstrating
a pattern in which large numbers of adult pseu-
doscorpions boarded beetles on old, depleted trees
and disembarked on newly-fallen trees. The study
also revealed a novel aspect of the beetle/pseu-
doscorpion relationship. Large male C. scor-
pioides monopolize beetle abdomens as strategic
sites for intercepting and inseminating dispersing
females (Zeh & Zeh 1992).

METHODS

We tested the phagophily hypothesis by com-
paring pseudoscorpions taken from beetles with
individuals collected from within decaying trees,
using abdomen length as a measure of recent
food consumption. In pseudoscorpions, no fur-
ther molting occurs after the adult stage is reached
so that the fully-sclerotized pedipalps and ceph-
alothorax are fixed in size (Weygoldt 1969; Zeh
1987). However, the abdomen is only partially
sclerotized and enlarges with food intake.

Measurement of the abdomen length of beetle-
riding pseudoscorpions was restricted to indi-
viduals taken from beetles on newly-fallen, un-
decayed trees. We excluded pseudoscorpions
from beetles collected on older, dead trees with
evidence of beetle emergence. This was necessary
to ensure that pseudoscorpion abdomen length
reliably reflected the nutritional consequences of
beetle-riding. Pseudoscorpions climb onto har-
lequins soon after the beetles eclose from pupal
chambers within old trees (Zeh & Zeh 1992).
The post-teneral beetles then rapidly fly off in
search of newly-dead trees to mate and oviposit
(Zeh et al. 1992). Pseudoscorpions collected
from emerging beetles on old trees have therefore
only just embarked and are in a nutritional state
which largely reflects feeding within the tree and
not on the beetle. By contrast, pseudoscorpions
taken from a beetle captured on a new tree are
likely to have spent a significant period on the
beetle’s abdomen during its search for a freshly-
dead or dying tree.

In Panama, the research was carried out in
lowland tropical forest in the Parque Nacional

ZEH & ZEH- BEETLE-RIDING IN PSEUDOSCORPIONS

49

Figure 2. — Beetle-boarding behavior of C. scorpioides (see text).

Soberania. Eight female and 61 male pseudos-
corpions were removed from 58 harlequin bee-
tles taken from 7 newly-fallen, undecayed trees.
Sixty-nine female and 100 male C. scorpioides
were also collected from 16 tree populations in
the area. In French Guiana, 30 females and 54
males were removed from 34 beetles collected
from 3 newly-fallen trees along the Piste du Kaw,
50 km southeast of Cayenne. An equivalent
number of males and females was collected from
1 2 sympatric tree populations.

Two interrelated factors confound a direct
comparison of the mean abdomen length of
beetle-riding versus within-tree pseudoscor-
pions. First, although abdomen length varies with
nutrition, there are also significant positive cor-
relations between abdomen length and sizes of
pedipalpal and cephalothorax traits. Second, for
our analysis, it was particularly critical to take
account of these correlations. Male pseudoscor-
pions compete to establish mating territories on
the abdomens of beetles and, as a consequence,
beetle-riding males are significantly larger overall
than males randomly-sampled from trees (Zeh
& Zeh 1992). Therefore, even in the absence of

nutritional differences, mean abdomen length is
expected to be larger in beetle-riding pseudo-
scorpions.

Direct comparison of mean abdomen lengths

Table 1. — Results of principal components (PC)
analysis of nine traits of the pedipalps and cephalo-
thorax in C. scorpioides. PC analyses were carried out
separately by gender and geographic location.
Prop. = Proportion of total morphological variation
explained by PCI.

Trait

Trait loading on PCI

FG

female

FG

male

PN

female

PN

male

MFL

0.140

0.080

0.135

0.080

HL

0.139

0.121

0.092

0.1 14

HD

0.131

0.230

0.147

0.204

TL

0.077

0.104

0.134

0.106

TD

0.180

0.160

0.179

0.197

FL

0.182

0.098

0.137

0.103

FD

0.094

0.104

0.203

0.129

CL

0.121

0.064

0.089

0.054

CW

0.186

0.093

0.156

0.072

Prop.

0.637

0.897

0.61 1

0.886

50

THE JOURNAL OF ARACHNOLOGY

tijO

OJ

0)

PCI Score of Sclerotized Traits

Figures 3-6.— Comparison of relative abdomen length in C. scorpioides from trees versus individuals from
harlequin beetles (HB) (see text for explanation). Analysis carried out separately by gender and geographic
location (FG = French Guiana; PN = Panama).

in the two environments was carried out, using
an analysis of covariance (ANCOVA) which fac-
tored out the correlation between sclerotized trait
size and length of the abdomen. The ANCOVA
analysis is most clearly interpreted by first de-
riving a single, composite measure of size for the
nine sclerotized traits we measured. This was
obtained by performing principal components
analysis (PCA) in which the first principal com-
ponent (PCI) provides the single best measure
of overall size (see Bookstein et al. 1985).

Sclerotized traits included in the analysis were:
chelal movable finger length (MFL); chelal hand

length (HL) and depth (HD); tibia length (TL)
and depth (TD); femur length (FL) and depth
(FD), and cephalothorax length (CL) and pos-
terior width (CW) (see Chamberlin 1931). Mea-
surements were taken from photographs of live
individuals restrained with pedipalps fully ex-
tended under a glass slide (Kodak Technical Pan
film, Nikon FE2 camera, 55 mm Micro-Nikkor
lens with 55 mm extension). The negative image
of each specimen was then projected onto a com-
puter-linked digitizing tablet (Summagraphics
MM 1201) and the coordinates of 38 anatomical
landmarks on the dorsal outline of the body and

ZEH & ZEH- BEETLE-RIDING IN PSEUDOSCORPIONS

51

right pedipalp (plus two scale bar points) were
recorded. The 9 traits were then computed from
the coordinates. Principal component (PC) scores
were calculated from the covariance matrix of
In-transformed measurements. All traits loaded
positively on PCI (Table 1) which therefore rep-
resents a composite measure of overall size (see
Bookstein et al. 1 985). To avoid negative values,
the traits (measured in mm) were first multiplied
by 10 before natural logarithmic transformation.
Statistical analyses were performed using SAS©
(SAS Institute, 1988).

RESULTS AND DISCUSSION

In the analysis of covariance, abdomen length
represented the dependent variable, PCI score
of the sclerotized traits the covariate, and “envi-
ronment”, i.e., beetle versus tree, the indepen-
dent categorical variable. Results demonstrated
that, in both the Panamanian (PN) and French
Guianan (FG) populations, adjusted mean ab-
domen length (least squares mean or LSM) of
pseudoscorpions on trees (LSMxree) exceeded that
of beetle-riding individuals (LSMhb) in both fe-
males (FG: LSM-r^ee = 3.45, LSMhb = 3.36, F.,57
= 4.60, P = 0.036; PN: LSM-r,,e = 3.50, LSMhb
= 3.23, F, 74 = 13.88, P < 0.001) and males (FG:
LSM^^ee = 3.44, LSMhb = 3.36, F, ,05 = 4.94, P
= 0.028; PN: LSMi-^^e = 3.36, LSM„b = 3.26,
F,,,56 = 8.26, P = 0.005) (see Figs. 3-6). This
suggests that individuals within trees are better
nourished than their counterparts on beetles.
While our observations confirmed that beetle-
riding pseudoscorpions do prey on the harle-
quin’s mites, the results presented here were not
consistent with Vachon’s (1940) hypothesis that
phagophily is the primary motivation for the as-
sociation.

Finally, we suggest that, based on purely phys-
iological considerations, the phagophily hypoth-
esis seems flawed. Like other arachnids, pseu-
doscorpions feed by injecting digestive enzymes
into their prey and then sucking out the dissolv-
ing tissue (Weygoldt 1969). External digestion
enables C. scorpioides to exploit relatively large
prey such as the dipteran and coleopteran larvae
available in decaying trees. By contrast, such a
feeding technique seems particularly ill-suited for
specialization on the small mites resident on har-
lequin beetles.

ACKNOWLEDGMENTS

We thank W. B. Muchmore and V. Mahnert
for identifying the pseudoscorpions, W. B.
Muchmore and P. Weygoldt for useful com-
ments on the manuscript, and R. E. Strauss for
providing the digitizing data acquisition and dis-
tance computing programs. We also thank the
Panamanian Instituto Nacional de Recursos Na-
turales Renovables (INRENARE) for permis-
sion to carry out the work. Both authors grate-
fully acknowledge fellowship support from the
Smithsonian Tropical Research Institute.

LITERATURE CITED

Beck, L. 1 968. Aus den Regenwaldem am Amazonas
I. Natur und Mus., 98:24-32.

Beier, M. 1948. Phoresie und Phagophilie bei Pseu-
doscorpionen. Osterreich. Zool. Z., 1:441-497.
Bookstein, F. L., B. Chemoff, R. L. Elder, J. M. Hum-
phries, G. R. Smith & R. E. Strauss. 1985. Mor-
phometries in Evolutionary Biology: The Geometry
of Size and Shape Change with Examples from Fish-
es. Acad. Nat. Sci. Philadelphia Spec. Publ. 15.
Chamberlin, J. C. 1931. The arachnid order Chelo-
nethida. Stanford Univ. Publ. Biol. Sci., 7:1-284.
Duffy, E. A. J. 1960. A Monograph of the Immature
Stages of Neotropical Timber Beetles. British Mus.
Nat. Hist., London, U.K.

Muchmore, W. B. 1971. Phoresy by North and Cen-
tral American pseudoscorpions. Proc. Rochester
Acad. Sci., 12:79-97.

Ricklefs, R. E. 1979. Ecology, 2nd ed. Chiron Press,
New York.

SAS Institute Inc. 1988. SAS/STAT User’s Guide,
Release 6.03 Edition. Cary, North Carolina.
Vachon, M. 1940. Remarques sur la phoresie des
Pseudoscorpions. Ann. Soc. Ent. France, 109:1-18.
Weygoldt, P. 1969. The biology of pseudoscorpions.

Harvard Univ. Press, Cambridge, Massachusetts.
Zeh, D. W. 1987. Life history consequences of sexual
dimorphism in a chemetid pseudoscorpion. Ecol-
ogy, 68:1495-1501.

Zeh, D. W. & J. A. Zeh. 1991. Novel use of silk by
the harlequin beetle-riding pseudoscorpion, Cor-
dylochernes scorpioides (Pseudoscorpionida, Cher-
netidae). J. Arachnol., 19:153-154.

Zeh, D. W. & J. A. Zeh. 1992. Dispersal-generated
sexual selection in a beetle-riding pseudoscorpion.
Behav. Ecol. Sociobiol., 30:135-142.

Zeh, D. W., J. A. Zeh & G. Tavakilian. 1992. Sexual
selection and sexual dimorphism in the harlequin
beetle Acrocinus longimanus. Biotropica, 24:86-96.

Manuscript received February 1991, revised April 1991.

1992. The Journal of Arachnology 20:52-57

NEW SPECIES OF CRAB SPIDERS
FROM BAJA CALIFORNIA SUR (ARANEAE: THOMISIDAE)

Maria-Luisa Jimenez: Centro de Investigaciones Biologicas de Baja California Sur, A.
C., Apdo. Postal 128, La Paz, Baja California Sur, Mexico

ABSTRACT: Three new species of the genera Isaloides, Misumenoides and Tmarus from the Cape Region,
Baja California Sur, are described and illustrated.

Only two species of spiders from the Americas
are included in the genus Isaloides F. Pickard-
Cambridge: Isaloides puta (F. O. Pickard-Cam-
bridge 1890) from Mexico and Panama and I.
toussainti Banks 1903 from Cuba and Haiti (F.
O. Pickard-Cambridge 1903; Brignoli 1983 and
Bonnet 1957).

Spiders of this genus are typically six mm in
total length and with a carapace longer than wide
and flattened. The eyes are arranged in two trans-
verse and recurved rows with the anterior row
shorter; the lateral eyes are larger than the me-
dian eyes and are seated on separate tubercles.
The legs are generally short, pale brown, rather
thick, and very strong without scopulae. Legs I
and II are longer than III and IV. The tarsi have
two claws and claw tufts. The opisthosoma is
rather angular, longer than wide, and with four
or five pairs of circular red spots. The palpal tibia
of the male is without both retrolateral or ventral
apophyses, and the tegulum is flattened. The em-
bolus is long, slender and curled. The epigynum
of the female is with a wide excavated atrium
and without median septum or hood. The sper-
mathecae are small, varying in the shape ac-
cording to the species.

Twenty seven species of the genus Misume-
noides F.O. Pickard-Cambridge 1 900 have been
described worldwide (Brignoli 1983, Bonnet
1957), of which four occur in North America:
Misumenoides fonnosipes (Walckenaer) from
Canada and United States, M. parva (Keyserling
1880) from Mexico, Panama and Colombia, M.
aleatoria (Hentz 1847) from United States and
Canada, and M. annulipes (O.P. Cambridge 1891)
from Mexico, Guatemala and United States.

Members of this genus have large, flattened
bodies of 2.50-1 1.30 mm total length. The ca-
parace is low, smooth and convex, the body is
pale green and white along the lateral margins,

with red markings in the opisthosoma; several
erect setae and a white transverse carina are found
on the clipeus. The eyes are arranged in two
transverse recurved white rows, with the poste-
rior row more curved than the anterior row. The
lateral eyes are larger than the median eyes and
are seated on large conjoined tubercles. Legs I
and II are much longer and thicker than legs III
and IV. They are creamy white and have no spots
or bands, scopulae or claw tufts. They have a
pair of ventral macrosetae, and the tarsi have
with two claws. The opisthosoma is broad and
flat, off-white to yellow in color and lacking erect
setae. The palpal tibia of the male has an elab-
orate retrolateral apophyses, and a shorter sim-
pler ventral apophysis. The embolous is short,
spur-like, and arising near the distal end of te-
gulum. The epigynum of the female is nearly
sclerotized with a shallow atrium and a broad
hood. The spermathecae are broader than long
(Dondale & Redner 1978).

These spiders usually sit between petals and
stamens of blossoms, where they ambush polli-
nating insects of considerable size.

The genus Tmarus Simon 1875 has a wide
distribution with approximately 150 described
species worldwide (Bonnet 1959; Brignoli 1983).
Eight species of this genus are known from Mex-
ico (Jimenez 1987). Members of this genus are
recognized by their strong, dark brown dorsally,
rather convex carapace that is larger than wide
and conspicuously developed anteriad. The eyes
are in two transverse recurved rows. The lateral
eyes are longer than the median eyes and are
seated on large separate tubercles. The legs are
long with black spots, and no scopulae or claw
tufts. The tarsi have two claws. Legs I and II are
longer than III and IV. The opisthosoma is an-
gular at the lateral margins, longer than wide,
with a conspicuous dorsal tubercle at the pos-

52

JIMENEZ-NEW SPECIES OF CRAB SPIDERS

tenor end. It is mottled and dull in color. The
palpal tibia of the male has both ventral and
retrolateral apophyses. The embolus is broad, the
epigynum is lightly sclerotized, with a small hood.
The spermathecae are longer than wide, with sur-
face grooves (Gertsch 1939; Dondale & Redner
1978).

Isaloides yollotl, new species
(Figs. 1-4)

Types. — Male holotype from low deciduous
forest in Santiago, Baja California Sur, (13 Au-
gust 1989, F. Cota). The following paratypes are
from the type locality; six females and eight males
(13 August 1989, F. Cota). Type and one female
paratype will be deposited at the Collection of
the Acarology Laboratory, Fac. de Ciencias
Universidad Nacional Autonoma de Mexico,
with the exception of 1 3 paratypes which will be
deposited in the Arachnoiogical Collection of the
Centro de Investigaciones Biologicas de Baja
California Sur, A.C.

Etymology.— The specific name is derivated
from the Nahuatl word “yollotl” which means
“heart”. This is suggested by the epigynum shape.

Diagnosis.— Members of Isaloides yollotl n. sp.
resemble I. puta (O. P. Cambridge) in coloration
and body shape, but can be separated from those
of the other known similar species by the em-
bolus where the tip rests on a groove at the distal
edge of the bulb. Epigynum of the female has a
broader atrium and the spermathecae shape is
diagnostic.

Males.— Total length 3.80-4.85 mm, prosoma
1 .70-2. 1 5 mm long and 1 .60-2.00 mm wide (nine
specimens). Femur II 2.75-3.50 mm. Carapace
pale redish yellow, flattened, higher at level of
coxa III, with few white clavate setae; ocular area
darker. Anterior region of carapace with darker
radiating lines and a pale median area, with edges
somewhat dark. Eyes on small gray tubercles and
arranged in two transverse rows, anterior more
procurved than posterior; anterior median eyes
separated one diameter between them; anterior
lateral eyes bigger and red, separated 2.5 diam-
eters of one anterior median eye. Chelicerae with
two small teeth on both promargin and retro-
margin. Legs with sparse scopula and claw tufts;
patella and distal part of tibia and tarsus darker.
Femur I 3.10-3.55 mm, dark yellow with three
dorsal macrosetae, five prolateral, four retrola-
teral and two or three small ventral macrosetae;
tibia I 2.50-3.00 mm, with two dorsal macro-
setae, three prolateral, three retrolateral and no

53

ventral. Basitarsus I 2.55-2.80 mm, with three
prolateral macrosetae, three retrolateral and three
pairs of ventral macrosetae. Tibia III 0.95-1.10
mm, with two dorsal macrosetae, two prolateral,
two retrolateral and three ventral pairs. Opis-
thosoma slender, cleft mid-line, pale yellow, dor-
sally with a median dark band and five pairs of
red circular spots on lateral light bands; sides
pale, with grooves and dark setae; venter pale
and with a light spot in half-moon shape on each
side; spinnerets reddish yellow, lighter at tip. Tibia
of the palpus as long as wide without ventral or
retrolateral apophyses. Embolus long, slender and
curled, with the tip resting on an anterior distal
groove of the bulb; pars pendula terminating at
approximately 270° of bulb, with prolateral edge
strongly sclerotized (Figs. 1 , 2).

Female.— Total length 5.65-6.15 mm, pro-
soma 2.30-2.85 mm long and 2.05-2.50 mm
wide (six specimens). Shape and color similar to
the male, but with the body much paler. Femur
II 3.05-3.40 mm long. Femur I 3.00-3.35 mm,
with two dorsal macrosetae, four prolateral, four
retrolateral, none ventral. Tibia I 2.50-2.85 mm,
with two dorsal macrosetae, three prolateral, three
retrolateral, four ventral pairs. Basitarsus I 2.00-
2.25 mm, with no dorsal macrosetae, three pro-
lateral, three retrolateral and three ventral pairs.
Tibia III 0.90-1.25 mm, with two dorsal macro-
setae, two prolateral, three retrolateral and three
ventral pairs. Epigynum heart shaped, as long as
wide, with broad atrium and conspicuous cop-
ulatory openings; spermathecae small, and light-
ly curled; copulatory tubes wide and short (Figs.
3, 4).

Range. — Known only from the type locality.

Misumenoides quetzaltocatl, new species
(Figs. 5-8)

Types.— Male holotype from xeric shrub El
Comitan, 28 September 1987 (M. Vazquez).
Along with the following paratypes; four males
and three females, 7 October 1987 (F. Cota and
M. Jimenez), 7 October 1986 (M. Jimenez), 17
September 1987 (A. Cota), all from the type lo-
cality; Sierra de la Laguna, La Zorra canion, low
deciduous forest, 1 October 1987 (M. Jimenez);
Santiago, (13 August 1989 (A. Cota). The type
and one female paratype are deposited at the
Collection of the Acarology Laboratory, Facul-
tad de Ciencias, Universidad Nacional Auton-
oma de Mexico, and six paratypes, which are
deposited at the Arachnoiogical Collection of the

54

THE JOURNAL OF ARACHNOLOGY

view; 4. Epigynum, dorsal view.

Centro de Investigaciones Biologicas de Baja
California Sur, A.C.

Etymology.— The specific name is derived from
the Nahuatl word “quetzalli” which means “green
and beautiful plumage” and “tocatl” which means
“spider”. The name denotes the green coloration
of this species.

Diagnosis.— quetzaltocatl n. sp.
is most similar to M. parvus (Keyserling), but
differs from the latter in having the male palpus
smaller and the retrolateral apophysis without a
distal notch; the spermathecae of the female are
not round and the copulatory tubes originate dis-
tally.

Figures 5-8. — Misumenoides quetzaltocatl nov. sp.; 5. Palpus, ventral view; 6. Palpus, lateral view; 7. Epi-
gynum, ventral view; 8. Epigynum, dorsal view.

JIMENEZ-NEW SPECIES OF CRAB SPIDERS

Males.— Total length 2.57-2.67 mm, prosoma
1.15-1.25 mm long, and 1.27-1.47 mm wide
(five specimens). Femur II 1.50-1.90 mm. Car-
apace low, greenish orange, highest at level of
coxae III, with sparse setae, and with a white
transverse carina on front. Anterior region of
caparace light with white edges; ocular area pale.
Eyes in two transverse curved rows, anterior row
more recurved than posterior row; lateral eyes
seated on conjoined white tubercles; anterior me-
dian eyes separated by three times the diameter;
anterolateral eyes larger and separated from me-
dian eyes by one and a half the eye diameter.
Chelicerae reddish brown and without teeth on
either margin. Sternum dark yellow. Legs I and
II much longer and thicker than legs III and IV,
dark brown and strongly sclerotized, with sparse
scopula and with dorsal granulations, except on
the basitarsi and tarsi. Coxa and trochanter I and
II dorsally pale; legs III and IV pale yellow; joints
of the patella, tibia and femur II and IV with
white rings; femur I 1.50-1.92 mm, with four
dorsal macrosetae; tibia I 1.10-1.37 mm, with
two or three pairs of ventral macrosetae. Basi-
tarsus I 1.00-1.27 mm, with one prolateral mac-
roseta; one retrolateral, four or five ventral ma-
crosetae. Tibia III 0.45-0.50 mm, with no
macrosetae; trochanter IV with a shallow notch.
Dorsum of opisthosoma dark yellow, and broad,
with small sparse setae, with five dark spots;
sides dark; venter pale with a greenish spot at
median region; epigastric region dark yellow.
Tibia of palpus somewhat wider than long, with
a long retrolateral apophyses without notch (Fig.

5) ; ventral apophysis short and slender; embo-
lus short curved, arising distally on tegulum (Fig.

6) .

Female.— Total length 5.00-6.75 mm, proso-
ma 2.25-3.00 mm long and 2.37-2.90 mm wide
(three specimens). Shape and color similar to the
male but with the following exceptions: carapace
green with a white spot on anterior region; ocular
area with a dorsal median white line; sternum
pale; legs with white segments. Femur I 2.65-
3.35 mm with a dorsal macrosetae and three
prolateral macrosetae. Femur II 2.75-3.25 mm.
Tibia I 1.95-2.25 mm, with three ventral ma-
crosetae. Basitarsus 1 1 .62-2.00 mm with 10 ven-
tral macrosetae. Tibia III 0.75-1.00 mm, with
no macrosetae. Dorsum of opisthosoma whitish,
with sides darker, venter pale, with median green
spot. Epigynum wider than long with a shallow
atrium and with small hood; copulatory openings
located at median part (Fig. 7); copulatory tubes

55

short, rather thick; spermathecae longer than wide
(Fig. 8).

Range. — Known only from the type and lo-
cality of Sierra de La Laguna.

Tmams ehecatltocatl, new species
(Figs. 9-12)

Types. — Male holotype from xerich shrub El
Comitan, Baja California Sur, 15 August 1986
(M. Jimenez and F. Cota) and three females and
three males [15 August 1986, 17 September
1987], Sierra de la Laguna, Baja California Sur;
low deciduous forest, 5 October 1 986, 3 Novem-
ber 1987 (M. Jimenez, A. Acevedo and A. Cota).
The type and one female paratype will be de-
posited at the Collection of the Acarology Lab-
oratory, Fac. de Ciencias Universidad Nacional
Autonoma de Mexico. Seven paratypes are de-
posited at the Arachnological Collection of the
Centro de Investigaciones Biologicas de Baja
California Sur, A. C.

Etymology.— The specific name is a com-
pound Nahuatl word “ehecatltocatl” which
means “spider of the wind”.

Diagnosis. — rrnarws ehecatltocatl n. sp. most
resembles T. minutus Banks in structure but dif-
fers from that species in having the palpus of the
male with the ventral apophysis bifurcated, the
retrolateral apophysis ending in a bent tip, and
the embolus longer and thinner and the tegulum
with a median spur.

Males.— Total length 2.97-3.55 mm, caparace
1.07-1.42 mm long and 1.07-1.14 mm wide (four
specimens). Femur II 2.07-2.80 mm. Carapace
reddish brown anteriorly region with whitish ra-
diating lines and scattered black spots and lighter
areas; sides with reticular pigment, with macro-
setae arranged in three longitudinal rows. Eyes
on gray whitish tubercles; posterior median eyes
surrounded with small setae; front white, almost
horizontal. Chelicerae white, with sparse macro-
setae and without teeth on either margin. Ster-
num pale yellow with black spots. Legs pale yel-
low, with dark rings at distal part of basitarsus I
and 11. Legs III and IV paler and with sparse
scopula; ventral region of femur, patella and tibia
without dark spots. Femur I 2. 1 2-2.87 mm, with
two dorsal macrosetae, three retrolateral and two
prolateral. Tibia I 1.80-2.45 mm, with a dorsal
macrosetae, three prolateral, three retrolateral,
five ventral. Basitarsus 1 1 .60-2. 1 7 mm, with two
retrolateral macrosetae, two or three prolateral
and five ventral. Tibia III 0.77-1.10 mm, with
two dorsal macrosetae. Dorsum of opisthosoma

56

THE JOURNAL OF ARACHNOLOGY

Figures 9-\2. — Tmarus ehecatltocatl nov. sp.: 9. Palpus, ventral view; 10. Palpus, lateral view; 1 1. Epigynum,
ventral view; 12. Epigynum, dorsal view.

gray brown, with sparse setae arising on small
tubercles, with a pale longitudinal stripe; sides
with a black discontinuous band; venter pale;
epigastrium with a pale stripe and a median dark
spot and a pale band in each side; distal segment
of spinnerets with numerous dark setae. Tibia of
the palpus appoximately as long as wide, with a
bicuspid ventral apophyses, left cusp wider and
bigger, the right cusp smaller and truncated; re-
trolateral apohyses short and ending in a lateral
tip; tegulum with one median spur; embolus
rather short and thin, with tip lightly curved in
lateral view (Figs. 9, 10).

Female.— Total length 3.45-6.05 mm, proso-
ma 1 .25-1 .50 mm long and 1 .37-1 .50 mm wide.
General structure and color essentially as in male
(five specimens). Femur I 1.87-2.22 mm, with
one or two dorsal macrosteae, two prolateral,
three or four retrolateral. Femur II 1.87-2.30
mm. Tibia I 1 .47-1.67 mm, with two dorsal ma-
crosetae, three prolateral, two or three retrola-
teral, three vental. Basitarsus I 1.22-1.47 mm,
with two retrolateral macrosetae, eight ventral.
Tibia III 0.80-1.10 mm, with one dorsal macro-
setae, one prolateral and one retrolateral. Epi-
gynum round, somewhat sclerotized and without
hood, with anterior oval depression bounded by
setae (Fig. 1 1). Spermathecae longer than wide
and half-moon shaped, with series of shallow
transverse surface grooves (Fig. 12).

Range. — Known from the type locality and Si-
erra de la Laguna.

ACKNOWLEDGMENTS

I wish to express my gratitude to Dr. C. D.
Dondale of the Biosystematics Research Centre,
Ottawa, Canada for confirmating the new species
and also for his valuable comments on the manu-
script, to Dr. Anita Hoffmann for her review and
suggestions regarding this work and to Mr. Fran-
co Cota for his help in the collection of the spec-
imens. This work was supported by grants from
the Centro de Investigaciones Biologicas de Baja
California Sur, A.C., the Consejo Nacional de
Ciencia y Tecnologia (CONACyT), the Secretar-
ia de Programacion y Presupuesto (SPP) and the
World Wild Life Fund (WWF).

LITERATURE CITED

Bonnet, P. 1959. Bibliographia Araneorum. Tou-
louse, 2:1-5058.

Brignoli, P. M. 1983. A Catalogue of Araneae de-
scribed between 1940 and 1981. Manchester Univ.
Press. 755 pp.

Dondale, C. D. & J. H. Redner. 1978. The Insects
and Arachnids of Canada, Part 5, The crab spiders
of Canada and Alaska (Araneae: Philodromidae and
Thomisidae). Canada Dept. Agr. Pub. 1663. Otta-
wa, Canada. 255 pp.

Gertsch, W. J. 1939. A revision of the typical crab

JIMENEZ-NEW SPECIES OF CRAB SPIDERS

57

spiders (Misumeninae) of America North of Mex-
ico. Bull. American Mus. Nat. Hist., 76:277-422.
Jimenez, M. L. 1987. Dos nuevas aranas cangrejo
(Araneae, Thomisidae) de Mexico. J. Arachnol., 15:
395-399.

Pickard-Cambridge, F. O. 1903. Arachnida Aranei'
da. Biologia Centrali Americana, Vol. 2:89-192.

Manuscript received April 1991, revised August 1991.

1992. The Journal of Arachnology 20:58-63

SUPERCOOLING AND ITS ECOLOGICAL IMPLICATIONS
IN COELOTES ATROPOS {ARANEAE, AGELENIDAE)

Kefyn M. Catley': School of Biological Sciences, University College of Wales, Aberys-
twyth, Dyfed, Wales, United Kingdom

ABSTRACT. Field observations have shown Coelotes atropos to be winter-active and tolerant of a wide
environmental gradient. This study suggests that low temperature tolerance is achieved by a combination of
behavioral thermoregulation and physiological adaptation. It was found that the two populations studied, one
living at 732 m elevation and the other at sea level, were not significantly different in their ability to supercool.
However, a highly significant relationship between body weight and ability to supercool was demonstrated such
that immature stages are far more tolerant of low temperatures than adults. Juvenile spiders were not only able
to tolerate sub-zero temperatures, but also demonstrated an ability to cold acclimate. They were active in the
supercooled state and capable of silk production at -5 °C. Mechanisms which may account for the loss of
supercooling ability are discussed as well as the implications of such a change for habitat utilization and life

cycle strategy.

Poikilothermic arthropods living in northern
temperate zones have evolved a variety of over-
wintering strategies to maximize their fitness.
Spiders can be considered a model group in the
study of winter ecology (Schaefer 1977, 1987),
showing a variety of strategies of tolerating tem-
poral low temperature stress. Five basic life cycle
patterns can be distinguished: 1) Eurychronous
species that mature after two or more years and
therefore overwinter in various developmental
stages; 2) Stenochronous species that reproduce
in spring or summer and overwinter as imma-
tures, (Theridiidae, Salticidae and Lycosidae); 3)
Stenochronous species that lay eggs in autumn
and overwinter as spiderlings inside the egg case
(Araneidae); 4) Stenochronous species that re-
produce during winter (Linyphiidae); and 5) Di-
plochronous species that reproduce both in spring
and autumn and overwinter as adults.

Coelotes atropos (Walckenaer), an agelenid
spider, appears to be annual (type two), the main
overwintering stage being the juvenile spider.
During this period mortality due to low temper-
ature must be minimized. Thus one would expect
cryoprotectant synthesis, which facilitates su-
percooling of tissues, to be strongly selected for
in the juvenile spiders. Many passively overwin-
tering insects accumulate polyhydric alcohols in
their hemolymph (Kirchner & Kestler 1969;

'Present address: Department of Entomology, Com-
stock Hall, Cornell University, Ithaca, NY 1 4853-0999
USA

Kirchner 1973, 1987). However, the increased
osmotic pressure resulting from high polyol con-
centrations would require large scale physiolog-
ical and biochemical changes, which may not be
possible in a winter-active animal (Duman 1 977)
such as Coelotes atropos. Another factor impli-
cated in freezing point depression is thermal-
hysteresis protein (THP); these have been shown
to occur in insects (Husby & Zachariassen 1980)
and spiders (Duman 1979; Aunaas et al.,1983).
Two more important parameters to consider
when determining freeze tolerance are the type
of gut contents (the presence of ice nucleators)
and the level of dehydration (Somme 1982; Za-
chariassen 1982; Cannon & Block 1988). This
study was undertaken primarily to establish the
supercooling point (SCP) of Coelotes atropos, to
test whether this would differ between popula-
tions from two extremes of an altitudinal gra-
dient, and to test the spiders’ ability to cold ac-
climate. Further, I was interested to know whether
adults were more or less cold tolerant than ju-
venile spiders (this question has a bearing on the
phenomenon of the dead mother being canni-
balized by her overwintering spiderlings (Bris-
towe 1954)).

METHODS

Study sites.— Two study sites were chosen for
their degree of exposure to altitude (and therefore
temperature) and wind effects. Both habitats had
a plentiful supply of stones that were suitable for
C. atropos retreats. The Plynlimon site, Dyfed,

58

CATLEY-SUPERCOOLING IN COELOTES ATROPOS

59

Wales UK: During January and February 1989,
collections of spiders were made from rock scree
on the summit of Plynlimon Fawr (National Grid
Reference SN 789868) at an altitude of 752 m.
Animals were taken from stones in unstable scree
on the NW facing slopes and from more stable
scree on the SE slope of the summit. Both hab-
itats support a spider community which com-
prises C. atropos, Robertus lividus (Blackwall),
PoecUoneta globosa (Wider) and Centromerus
prudens (O. P. -Cambridge). The Arth valley,
Aberarth, Dyfed: During the same period animals
were collected from the sheltered wooded valley
of the River Arth (National Grid Reference SN
489625) at an altitude of 9.5 m, where stone-
strewn slopes beneath Quercus petraea (Mat-
tushka) Liebl. were found to support large pop-
ulations of C. atropos.

Supercooling point determination. —A Peltier
unit (a thermopile) was set on a stage such that
water cooled to 3 °C could circulate around a
heat sink under the stage. In order to achieve
sub-zero temperatures, it was necessary to cover
the stage with an insulated cover. The freezing
chamber consisted of two halves of an aluminum
dish (45 mm diameter), the bottom half of which
was in direct contact with the thermopile. Two
thermocouples (Copper/Constantan) were used
to record temperature, one connected to the
chamber, the other to the spider’s prosoma, thus
allowing simultaneous monitoring of the ani-
mal’s body temperature and the temperature of
the chamber. After collection, juvenile and adult
female spiders were stored in 5 cm x 2 cm vials
at 5 °C prior to being anesthetized with COj,
weighed, and placed in the freezing chamber. The
thermocouple was attached to the prosoma with
correcting fluid; however, for four of the largest
spiders it had to be attached with cellophane
adhesive tape. Thermocouples were connected
to Comark electronic thermometers, which in
turn were linked to a Bryan 2700 chart recorder.
Convention in low temperature experiments is
to decrease body temperature by 1 "C min’' (Salt
1961). This was achieved by use of a variable
power control to the Peltier unit, while moni-
toring the falling temperature curve with a stop
watch. Continual monitoring of the animal’s fall-
ing body temperature enabled determination of
the SCP exotherm, the sudden release of the la-
tent heat of fusion when the animal freezes spon-
taneously.

Animals used in the cold acclimation experi-
ment were placed in individual vials with small

In (Weight)

Figure 1. — Plot of supercooling points (SCP) against
the natural logarithm of body weight (g): for adult («
= 18 [triangles]), and juvenile {n = 29 [squares]) C.
atropos. Spiders were from both study sites.

amounts of moss as substrate. These were sub-
merged in a water bath containing ethylene glycol
and run at — 5 °C to — 2 °C (mean = —3.5 °C)
for 14 days. Light regimes were L. 10: D. 14,
closely following the ambient L/D cycle. Since a
period of 14 days was found to be adequate for
the spider Clubiona to acclimate (Duman 1979),
a similar time scale was adopted for C. atropos.
Throughout the experiment no animal was fed,
and none underwent ecdysis. A total of 93 ju-
venile and 18 adult spiders were used for the
statistical analyses using one-way ANOVA. Data
from the samples shown in Fig. 1 were combined
to test for differences in SCP between adults and
juveniles. Tests to detect differences between the
two populations of juveniles spiders were con-
ducted on a separate sample {n = 31). Spiders
used for the cold acclimation experiment were
also from an independent sample {n = 33).

RESULTS

Supercooling experiments.— Fig. 1 shows a plot
of SCP against the natural logarithm of body
weight for animals from both Plynlimon and Arth
sites. Adult spiders {n = 18) appear to have a
much higher mean freezing point than juveniles
(n = 29). Due to the paucity of adults at the

60

THE JOURNAL OF ARACHNOLOGY

Figure 2. — Plot of supercooling points (SCP) against
the natural logarithm of body weight (g): for adult (n
= 13 [triangles]), and juvenile (« = 14 [squares]) C
atropos. Spiders were from sea level, Arth site.

Plynlimon site {n = 5), data from the Arth site
(Fig. 2) were used to test for a difference in SCP
between adults > 100 mg (rz = 13) and juveniles
< 100 mg (« = 14). It proved to be highly sig-
nificant (One-Way ANOVA; F,, 25, = 46.66, P <
0.001). Consequently, adults and juveniles were
considered separately in population compari-
sons.

A further test showed there to be no significant
difference between the SCP of the Plynlimon {n
= 17) and Arth {n = 14) juvenile samples (One-
Way ANOVA: F„,29, < 1, NS; Fig. 3). Therefore,
ability to cold acclimate was tested using juve-
niles from both populations. Cold acclimation
to -3.5 °C for 14 days resulted in a significantly
enhanced mean SCP. Cold acclimated animals
mean SCP = — 6.1 °C, n = 33; whereas the mean
SCP for field fresh animals = —4.4 °C, n = 39
(One-Way ANOVA: F,,,™) = 17.49, P < 0.001;
Fig. 4). A test for an effect of location (Arth or
Plynlimon) on ability to cold acclimate proved
to be non-significant {P > 0.05).

Two observations of interest made during the
cold acclimation experiment related to the adults’
inability to survive sub-zero temperatures and
to silk production by the juveniles. Three adult
female spiders subjected to sub-freezing temper-
atures (-3.5 °C) died after 24 hours exposure.
All juvenile spiders {n = 33) not only survived
for the duration of the experiment (14 days), but
exhibited normal movements in the supercooled
state; silk synthesis continued apparently unhin-
dered and webs were fashioned on the substrate.

Natural History Observations. — In Britain the

-1 -2 -3 -4 “5 -6 -7
SCP°C

Figure 3. — Frequency distribution histograms of su-
percooling points (SCP) for juvenile spiders. A) Arth
sample {n = 14). B) Plynlimon sample (n = 17). There
was no significant difference between the populations
in their ability to supercool body fluids.

genus Coelotes contains two species; C. atropos
and Coelotes terrestris (Wider). Whereas their
distribution overlaps in southern counties, C.
atropos is much more common in the North and
West (Locket et al. 1974) where it is particularly
associated with high ground. However, its range
does not extend much into Scotland, even though
it is able to tolerate extremes of climate associ-
ated with mountain tops. While the C. atropos

SCP°C

Figure 4.— Frequency distribution histograms of su-
percooling points (SCP) for juvenile C. atropos samples
from both sites, juvenile spiders only. A) field fresh
spiders {n = 39). B) spiders after 14 days acclimation
at -3.5 “C (« = 33).

CATLEY- SUPERCOOLING IN COELOTES ATROPOS

61

population from Plynlimon Fawr (752 m) was
sampled during this study, the species has also
been recorded from several other Welsh moun-
tain habitats including Snowdon (> 920 m),
Cader Idris (893 m), and the Brecon Beacons
(887 m) (Bristowe 1938).

C. atropos is cryptozoic, and the web, often in
the form of a tube, is built under stones and logs.
A collar surrounds the opening of the retreat,
while the proximal end often bifurcates. Eggs are
laid in June and spiderlings eclose after a month
or so (Bristowe 1954). Petto (1990) reports that
populations of C. terrestris in Germany are bi-
ennial; mating occurs in the autumn and both
juveniles and adult females overwinter. Bristowe
(1954) states that in Britain C. atropos mates
during spring or early summer suggesting an an-
nual life-cycle; the loss of supercooling ability
demonstrated to occur in adult female C. atropos
is certainly consistent with such a strategy. After
emergence the spiderlings remain together for a
considerable period, often several months, dur-
ing which time they are fed and guarded by the
mother (Bristowe 1954). In casual observations
of C. atropos web sites over several seasons, I
have often observed dead adult females being
consumed by their spiderlings. This has been ob-
served by other authors (Bristowe 1954; Tretzel
1961) and should facilitate offspring survival un-
til the spring.

Whereas immature stages of C. atropos often
feed on various stages of Collembola (pers. obs.),
adults and sub-adults feed largely on Coleoptera
(Bristowe 1954; Tretzel 1961). Prey remains (el-
ytra) found in webs at the Plynlimon site indicate
that adult C. atropos feed largely on the following
predatory ground beetles: Family Carabidae,
Ptetrosticus madius F., Carabus problematicus
Herbst, Carabus arvensis Herbst, and Calathus
melanocephalus'L.-, Family Elateridae, Ctenicera
cuprea F.

DISCUSSION

Kirchner (1973) recognized three main cate-
gories of spider SCP distributions which clearly
reflect the animals’ overwintering microhabitat.
These range from the low SCP of Theridion no-
tatum (Clerck) (=Enoplognatha ovata (Clerck))
(mean = —26.1 °C) which overwinters in open
vegetation, to the high SCP of Meta menardi
(Latreille) (mean = — 4 °C) which lives in caves
that are subject to little temperature fluctuation.
A German population of Coelotes terrestris (also
cryptozoic) was shown to have a mean SCP of

-6.2 °C: this Kirchner placed in a medium-to-
low category of cold tolerance. The equally high
SCP (mean = —4.4 °C, n = 39) exhibited by C.
atropos in this study appears to be consistent
with its cryptozoic behavior.

Ability to withstand the lowest winter tem-
peratures that occur annually in a habitat will be
strongly selected for. Consequently, geographical
variation in supercooling ability and behavioral
thermoregulation (or both) should be expected
in populations with wide geographical or alti-
tudinal ranges (Somme 1 982), and should be most
strongly expressed in species that overwinter in
exposed conditions. Whereas differences in su-
percooling abilities have been detected between
separate populations in other arthropod species
(Macphee 1961;Hansen 1978), no significant dif-
ference could be detected between the two pop-
ulations of C. atropos with regard to their ability
to supercool. Cryptozoic thermoregulatory be-
havior seems to be a vital component in allowing
C. atropos to utilize hostile environments (i.e.,
mountain summits). Perhaps the effectiveness of
such behavior might explain why the species as
a whole has such a high SCP, and why high el-
evation populations have not evolved a lower
SCP. Coelotes atropos responded to sub-zero
temperature acclimation by enhancing its ability
to supercool. Such cold acclimation has been
shown to occur in many species of insects and
mites (Schenker 1983; Cannon 1986; Cannon &
Block 1988). However, Kirchner & Kullman
(1975) showed that supercooling ability in the
spiders Theridion sisyphium (Clerck) and T. im-
pressum (L. Koch), both of which overwinter in
unprotected vegetation, did not appear to be af-
fected by warm or cold acclimation. Whereas the
end products of cold acclimation, such as in-
creased levels of glycerol and other cryoprotec-
tants, are easily demonstrable in insects, the pre-
cise ecophysiological mechanisms of acclimation
are little understood. Indeed, only recently has a
start been made to elucidate the neural basis of
thermal reception and perception in spiders (Pulz
1986).

Another factor implicated in freezing point de-
pression is dehydration. Although the required
degree of desiccation was probably not reached
during the cold acclimation experiment reported
here, it must be borne in mind as a possible
contributory factor. Water loss can increase the
solute concentration of the hemolymph, thus de-
pressing freezing point, without necessarily re-
quiring further cryoprotectant synthesis. Finally,

62

THE JOURNAL OF ARACHNOLOGY

certain spiders have been shown to possess THP
in their hemolymph (Duman 1 979; Husby & Za-
chariassen 1980), and under natural conditions,
THP production should be strongly selected for
in a winter-active animal.

The remarkable ability of all juvenile spiders
(« = 33) to synthesize silk and construct webs
on the frozen substrate during the cold accli-
mation experiment warrants further study.
Throughout the fourteen day period (at -3.5 "C)
the animals exhibited normal coordinated move-
ments, with no sign of chill coma. Such obser-
vations are supported by Aitchison (1987) who
observed winter-activity in juvenile spiders of
several families in temperatures as low as -8 °C,
and Hagvar (1973) who reported copulation In
Bolyphantes index (Thorell) at sub-freezing tem-
peratures. Whereas overwintering juvenile spi-
ders may display both inhibition of ecdysis and
low metabolic rates (Schaefer 1987), ability to
move normally and produce silk at sub-freezing
temperatures might confer selective advantage if
it allowed food capture and consumption during
periods when temperatures rose above 6 °C. C.
atropos is capable of high food consumption at
8 °C and 10 °C but exhibits an arrested devel-
opment at 6 °C (Aitchison 1981). If, as the evi-
dence seems to suggest, C. atropos does consume
food during the winter, it seems likely that THP
will be synthesized in the midgut, thus prevent-
ing inoculative freezing. However, there is some
debate concerning the effectiveness of the filter-
ing process as a method of removing ice nucle-
ators during the feeding process in spiders in
general and Kirchner (1987) has suggested that
most nucleators may be removed by the process.
Ramsey (1964) has shown THP to occur in the
midgut of insects, but its occurrence in the mid-
gut of spiders has yet to be shown. Loss of su-
percooling ability in adult females may result
from physiological changes associated with oo-
genesis, or if THP is involved, its synthesis may
be mediated by the presence of juvenile hormone
(JH). THP regulation by JH does occur in certain
insects (Horwath & Duman 1983; Hamilton et
al. 1986), and recent evidence (Carrel et al. in
press), has shown for the first time that spiders
do utilize JH to regulate development. The pre-
cise mechanism notwithstanding, loss of cold tol-
erance after maturity should result in strong se-
lection pressure towards an annual life cycle. In
Argyroneta aquatica (Clerck) (Bromhall, 1988)
and in some lycosids (Schaefer 1977), where
overwintering occurs in both adult and juvenile

stages, no loss of supercooling ability occurs dur-
ing either stage. Conversely, Kirchner & Kull-
man (1975) found that supercooling ability in
Theridion spp. where overwintering occurs
mainly in the juvenile stages, did, as in C. atro-
pos, vary with age.

Ontogenetic loss of supercooling ability in C.
atropos combined with temporal and spatial cli-
matic fluctuations may result in a change of life
cycle strategy. At the Plynlimon site only five
adult females were encountered during a total of
four collecting trips (cf juveniles n = 37), whereas
at the Arth site, presumably as a consequence of
temperature amelioration by the nearby sea, adult
females were plentiful throughout the sample pe-
riod (during January and February 1988, tem-
peratures at the Arth site fell below freezing on
only 9 occasions, reaching a low of — 2.2 °C, with
a mean of - 1 . 1 °C for those days when the tem-
perature fell below zero). If there are severe low
temperatures early in the winter, a female may
die and be digested by (and thereby contribute
to the survivorship of) her spiderlings. If, how-
ever, the winter is unusually mild (as during the
period of the study 1988-1989), then adult fe-
males are able to survive and possibly reproduce
for a second time the following spring. Such a
strategy together with cannibalism of the dead
mother by her overwintering spiderlings, provide
the animal with a “bet hedging” system well able
to contend with most climatic eventualities.

ACKNOWLEDGMENTS
I thank the Nature Conservancy Council for
permission to collect specimens at the study sites,
both of which are Sites of Special Scientific In-
terest. My grateful thanks to Dr. Mike Ireland,
Dr. John Gee and Phil Lloyd, University College
of Wales, Aberystwyth for help and support, to
Adrian Fowles (NCC) for identifying the beetle
remains, and to John Dingly and Cynon Jones
for their help in the field. Dr. G. Sumner, St.
David’s University College, Lampeter supplied
weather data. Dr. Frederick Coyle, Dr. J. E. Car-
rel and Dr. C. W. Aitchison provided construc-
tive criticism of the manuscript. Finally, I thank
my wife Diana for her fortitude and encourage-
ment. This research was carried out in partial
fulfillment of the degree of B.Sc. at UCW Aber-
ystwyth, Wales.

LITERATURE CITED

Aitchison, C. W. 1981. Feeding and growth of Coc-
lotes atropos (Araneae, Agelenidae) at low temper-
atures. J. Arachnol., 9:327-330.

CATLEY-SUPERCOOLING IN COELOTES ATROPOS

63

Aitchison, C. W. 1987. Feeding ecology of winter-
active spiders. Pp. 264-273. In Ecophysiology of
Spiders. (W. Nantwig, ed.) Springer-Verlag, Berlin.

Aunaas, T., J. G. Baust & K. E. Zachariassen. 1983.
Ecological studies on arthropods from Spitsbergen.
Polar Research 1 (N.S.):235-240.

Bristowe, W. S. 1938. The Comity of Spiders. Ray
Society, London.

Bristowe, W. S. 1954. The World of Spiders. Collins,
London.

Bromhall, C. 1988. Argyroneta aquatica (Clerck)
overwinter behaviour and supercooling point. Bull.
British Arachnol. Soc., 7:225-228.

Carrel, J. E., J. K. Atkinson & G. D. Prestwich. in
press. Juvenile hormone binding protein in spider
hemolymph. American Arachnol. (abstract)

Cannon, R. J. C. 1986. Diet and acclimation effects
on the cold tolerance and survival of an Antarctic
springtail. British Antarctic Surv. Bull., 71:19-30.

Cannon, R. J. C. & W. Block. 1988. Cold tolerance
of microarthropods. Biol. Rev., 63:23-77.

Duman, J. G. 1977. The role of macromolecular an-
tifreeze in the darkling beetle Meracantha contractu.
J. Comp. Physiol., 1 15:279-286.

Duman, J. G. 1979. Sub-zero temperature tolerance
in spiders: the role of thermal-hysteresis factors. J.
Comp. Physiol., 131:347-352.

Hagvar, S. 1973. Ecological studies on a winter-active
spider Bolyphantes index (Thorell) (Araneida, Lin-
yphiidae). Norsk Entomol. Tidsskr., 20:309-314.

Hamilton, M. D., R. R. Rojas, & J. G. Buast. 1986.
Juvenile hormone modulation of cryoprotectant
synthesis in Eurosta solidaginis by a component of
the endocrine system. J. Insect Physiol., 32:971-
979.

Hanson, T. 1978. On seasonal changes in the glycerol
content of two different populations of Apatele psi
L. Esti NSV Tead. Akad. Toim. Biol., 27:302-305.

Horwath, K. L. & J. G. Duman. 1983. Induction of
antifreeze protein production by juvenile hormone
in larvae of the beetle Dendroides canadensis. J.
Comp. Physiol. B Biochem. Syst. Environ. Physi-
ol.,l 5 1:233-240.

Husby, J. A. & K. E. Zachariassen. 1980. Antifreeze
agents in the body fluid of winter active insects and
spiders. Experentia, 34:963-964.

Kirchner, W. 1973. Ecological aspects of cold toler-
ance in spiders (a comparative studyj.Pp. 271-279.
In Effects of Temperature on Ectothermic Organ-
isms: Ecological Implications and Mechanisms of
Compensation. (W. Wieser, ed.) Springer-Verlag
Berlin, Heidelberg, New York.

Kirchner, W. 1987. Behavioral and physiological ad-
aptations to cold. Pp. 66-77. In Ecophysiology of
Spiders. (W. Nantwig, ed.) Springer-Verlag, Berlin.

Kirchner, W. & P. Kestler. 1969. Untersuchungen
zur Kalteresistenz der Schilfradspinne Araneus co-
mutus (Araneidae). J. Insect Physiol., 15:41-53.

Kirchner, W. & E. Kullman. 1975. Uberwinterung
und Kalteresistenz der Haubennetzspinnerarten
Theridion impressum (L. Koch) und Theridion sis-
yphium (Clerck) (Araneae, Theridiidae). Dechen-
iana, 127:241-250.

Locket, G. H., A. F. Millidge & P. Merrett. 1974.
British Spiders, vol. 3. Ray Society, London.

Macphee, A. W. 1961. Mortality of winter eggs of
the European red mite Panonychus ulmi (Koch), at
low temperatures and its ecological significance. Ca-
nadian J. ZooL, 39:229-243.

Petto, R. 1990. Abundance and prey of Coe/o/c5 ter-
restris (Wider) (Araneae, Agelenidae) in hedges. Bull.
British Arachnol. Soc., 8:185-193.

Pultz, R. 1986. Temperature related behavior and
temperature perception in spiders. Pp. 41-48. In
Thermal Relations in Animals and Man. Biona rep.
4, Akad. Wiss. Mainz Fisher, Stuttgart.

Ramsey, J. A. 1964. The rectal complex of the meal-
worm, Tenebrio molitor L. (Coleoptera, Tenebri-
onidae). Philos. Trans. Soc. London B Biol. Sci.,
248:279-314.

Salt, R. W. 1961. Principles of insect cold-hardiness.
Ann. Rev. Entomol., 6: 55-73.

Schaefer, M. 1977. Winter ecology of spiders (Ara-
neida). Z. Angew. Ent., 8: 1 13-134.

Schaefer, M. 1987. Life cycles and diapause. Pp. 331-
347. In Ecophysiology of Spiders. (W. Nantwig, ed.)
Springer-Verlag, Berlin.

Schenker, R. 1983. Effects of temperature acclima-
tion on cold-hardiness of alpine microarthropods.
Rev. Ecol. Biol. Sol., 20:37-47.

Somme, L. 1982. Supercooling and winter survival
in terrestrial arthropods. Comp. Biochem. Physiol.,
73A:5 19-543.

Tretzel, E. 1961. Biologic, Okologie und Brutpflege
von Coelotes terrestris (Wider) (Araneae, Ageleni-
dae), Teil II: Brutpfleg. Z. Morph. Okol., Tiere 50:
375-542.

Zachariassen, K. E. 1982. Nucleating agents in cold-
hardy insects. Comp. Biochem. Physiol., 73A: 557-
562.

Manuscript received January 1991, revised October
1991.

1992. The Journal of Arachnology 20:64-66

RESEARCH NOTES

PARASITISM OF PSEUDOSCORPIONS (ARACHNIDA)
BY MERMITHIDAE (NEMATODA)

Records of nematode parasites of pseudoscor-
pions are rare and consist of brief reports by
Vachon (1949) and Harvey (1982) who cite un-
identified mermithids in European and Austra-
lian pseudoscorpions, respectively. Although
Vachon mentioned that the nematodes he re-
covered appeared to resemble juveniles of the
genus Hexamermis, it is not possible to base a
reliable generic determination on immature mer-
mithids. During a study on the life history and
teratology of pseudoscorpions in the Balkan re-
gion (Curcic et al. 1991), one of us (B. P. M. C.)
came across specimens containing nematodes.
The present paper reports these finds and sum-
marizes our knowledge of nematode-pseudo-
scorpion associations.

Samples of infected pseudoscorpions were ob-
tained by sifting through leaf litter and humus
over a period from April 1 989 to September 1 990
in a mixed oak forest in the village of Obrez, near
Belgrade, Serbia, Yugoslavia. Six females and
three males of Roncus aff. lubricus L. Koch (Neo-
bisiidae) from a total of2167 adults (1335 males
and 832 females) were infected with represen-
tatives of the family Mermithidae of the order
Mermithida (Fig. 1). This results in an overall
infection rate of 0.4% for the adults, with a 0.2%
infection rate for males and a 0.7% infection rate
for females.

Two specimens of Neobisium carpaticum Beier
collected from Mt. Avala near Belgrade were also
associated with nematodes. The first specimen,
a deutonymph, had a small, coiled nematode
(length = 349 Mm; width = 22 Mm) in the pedi-
palpal femur (Fig. 2). This may represent an early
developmental stage of a mermithid nematode,
although further growth would be restricted in
this region of the host. The second specimen was
a “dauer” stage of a representative of the order
Rhabditida attached by its head to an abdominal
sclerite of a mature female host. The specimen
was 567 Mm long and 38 Mm wide and repre-
sented a third stage juvenile enclosed in its sec-
ond stage cuticle. Such phoretic associations be-
tween soil arthropods and rhabditoid nematodes
are not uncommon.

Three mermithid-parasitized individuals of
Roncus aff. lubricus were dissected and the nem-
atodes removed. In all three hosts, the internal
tissues, including the gonads, were atrophied and
the body cavity was completely occupied by the
parasites (Fig. 1). Two of the three hosts con-
tained two parasites each while the third con-
tained a single mermithid. Often, when two mer-
mithids are present in a host, one is a female and
the other is a male (Poinar, pers. obs.). Such a
ratio favors reproductive activity and continu-
ation of the life cycle.

Nematodes removed from parasitized Roncus
aff. lubricus were cream colored and ranged in
length from 4.8 to 6.6 mm {x = 5.7; n = 5). The
greatest body width ranged from 164 to 189 Mm;
{x = 178 Mm; n = 5). Four of the mermithids
were in their late parasitic stage and the cuticle
had thickened in preparation for the following
free-living post-parasitic stage. In these individ-
uals a prominent cuticular appendage was pres-
ent, ranging from 31-44 Mm in length {x 40 Mm;
n = 4). The fifth mermithid was still in the middle
of its parasitic stage and did not yet possess an
appendage. In all specimens, both anterior and
posterior ends were rounded and six faint head
papillae could be detected. Cuticular cross fibers
were not evident. The trophosome extended an-
teriorly into the head region and posteriorly into
the tail region. It is likely that these parasites
belong to a new species and possibly genus. How-
ever, descriptions of mermithids should be based
on adult characters which are still unavailable to
us.

Mermithid nematodes parasitize a wide vari-
ety of terrestrial invertebrates. Among the
Arachnida, they have been reported from spiders
and harvestmen (Poinar 1985) (Poinar & Early
1990) and scorpions (Poinar & Stockwell 1988)
as well as pseudoscorpions. Of the two spider
mermithids whose biology has been investigated,
both were shown to have indirect cycles involv-
ing spider predation on paratenic hosts contain-
ing the infective stages of the mermithids (Poinar
& Benton 1 986) (Poinar & Early 1 990). This type
of cycle may be widespread in predaceous hosts

64

RESEARCH NOTES

65

Figures 1, 2. — 1, Two parasitic mermithid nematodes filling the body cavity of a pseudoscorpion, Roncus aff.
lubricus. Bar = 540 nm. 2, A coiled unidentified nematode in the pedipalpal femur of a deutonymph of Neobisium
carpaticum. Bar = 36 fim.

and may occur in the present case with the pseu-
doscorpion mermithids. With spiders, the par-
atenic hosts are often aquatic detritivores such
as the immature stages of Trichoptera and
Ephemeroptera. When mature and ready to
emerge, the nematodes apparently drive the par-
asitized hosts to a water source; the mermithids
then exit the hosts. After maturation to the adult

stage, mating and oviposition occur. Immature
insects ingest the nematode eggs, which hatch in
their alimentary tracts. The newly emerged in-
fective stage mermithids penetrate the paratenic
host’s gut wall and enter the hemocoel where they
remain until being ingested by a spider.

It was not possible to determine whether the
life cycle of the pseudoscorpion mermithid is

66

THE JOURNAL OF ARACHNOLOGY

indirect, involving a paratenic host, or direct with
the adult stages in the same environment as the
host. Neither Vachon (1949) nor Harvey (1982)
commented on this point and their short reports
provide little more than the establishment of
mermithids in the body cavities of pseudoscor-
pions. However, Vachon did mention that the
ovary of the parasitized female was atrophied,
similar to the conditions found in the present
study.

Literature Cited

Curcic, B. P. M., R. N. Dimitrijevic, O. S. Karamata
& L. R. Lucic. 1991. Segmental anomalies in Ron-
cus alf. lubricus L. Koch, 1873 (Neobisiidae, Pseu-
doscorpiones) from Yugoslavia. J. ArachnoL, 19:
000-000.

Harvey, M. S. 1982. A parasitic nematode (Mermi-
thidae) from the pseudoscorpion “Sternophorus'’’
hirsti Chamberlin (Stemophoridae). J. Arachnol.,
10:192.

Poinar, Jr., G. O. 1985. Nematode parasitism of spi-
ders and harvestmen. J. Arachnol., 13:121-128.

Poinar, Jr., G. O. & C. L. B. Benton, Jr. 1986. Ar-
animermis aptispicula n. g., n. sp. (Mermithidae:
Nematoda), a parasite of spiders (Arachnida: Ara-
neida). Systematic ParasitoL, 8:33-38.

Poinar, Jr., G. O. & S. A. Stockwell. 1988. A new
record of a nematode parasite (Mermithidae) of a
scorpion. Revue d’NematoL, 1 1:316-364.

Poinar, Jr., G. O. & J. W. Early. 1990. Aranimermis
giganteus n. sp. (Mermithidae: Nematoda), a par-
asite of New Zealand mygalomorph spiders (Ara-
neae: Arachnida). Revue d’NematoL, 13:403-410.

Vachon, M. 1949. Ordre des Pseudoscorpions.
Pp.431-481. In: Traite de Zoologie (Grasse, P.-P.,
ed.) Vol. VI. Masson, Paris.

George O. Poinar, Jr. and Bozidar P. M. Curcic:
Department of Entomology, University of
California, Berkeley, California 94720 USA;
and Institute of Zoology, Faculty of Science,
Studentski Trg 16, YU- 11000 Beograd, Yu-
goslavia.

Manuscript received July 1991, revised September 1991.

1992. The Journal of Arachnology 20:67-68

CONJECTURES ON THE ORIGINS AND FUNCTIONS
OF A BRIDAL VEIL SPUN BY THE MALES OF
CUPIENNIUS COCCINEUS (ARANEAE, CTENIDAE)

Bristowe (1958) called the tiny web spun by
courting males of Xysticus cristatus (Araneae,
Thomisidae) over and around the female while
circling upon her a bridal veil. To my knowledge,
similar behavior of males has been reported for
Nephila davipes (Araneidae, Nephilidae; Rob-
inson & Robinson 1973, 1980), for Latrodectus
trededmguttatus (Araneidae, Theridiidae; Stem
«fe Kullmann 1981), for Tibellus oblongus (Ara-
neae, Philodromidae; Stem & Kullmann 1981),
^ov Ancylometes bogotensis (Araneae, Pisauridae;
Merrett 1988) and for Dictyna volucripes (Ara-
neae, Dictynidae; Starr 1988). In these species,
the male places a few threads over the female.
In Pisaurina mira (Araneae, Pisauridae), the fe-
male draws her legs I and II against her carapace
(in a flexed position) and the male wraps them
with a veil of silk prior to copulation (Bmce &
Carico 1988).

Spiders of the neotropical genus Cupiennius
live in close association with particular plants,
mainly bromeliads, on which they receive vi-
brations (e.g., from prey and mating partner) and
emit vibratory signals during courtship (Barth et
al. 1988). In a recent behavioral study of species
recognition and species isolation, we compared
hetero- and intraspecific communication in three
closely related species of the genus Cupiennius
(Barth & Schmitt 1991, Schmitt et al. 1990). In
9 out of 14 trials, the females of C. salei (which
are larger by about 30% than those of C coed-
neus) responded to the vibratory courtship of C.
coccineus males with their own vibratory court-
ship signals and both spiders finally met. When
mounting the female, three of the 9 males spun
attachment points onto the female’s legs and cir-
cled upon her for minutes while depositing silk
on her. The males inteirupted this behavior to
emit vibratory courtship signals. Whereas two
males copulated after a few minutes, the third
made increasingly longer excursions on the bro-
meliad consistently returning to the female and
continuing both his spinning behavior and vi-
bratory courtship. Finally, about two hours after
the first contact between the heterospecifics, he
was attacked and killed by the female when re-

turning to her. Obviously, the male silk did not
seriously affect the female’s mobility.

In Cupiennius, females are known to use silk
as draglines, to wrap and tie large prey (e.g., grass-
hoppers) to the substratum, to construct irregular
sheet webs partly or totally closing their retreats,
to build egg sacs and loose and irregular “nursery
webs” for the newly hatched spiderlings. Males
use silk as draglines, to build sperm webs, and
to immobilize large prey. During many years of
experience in breeding spiders of the genus Cu-
piennius, we never observed the behavior de-
scribed above for C. coccineus males. I suggest
four interpretations, which are not mutually ex-
clusive. (i) The male’s spinning is the same be-
havior as that shown when tying large prey to
the substratum. The male switches from court-
ship to predatory behavior and vice versa be-
cause the heterospecific female has attributes of
both mate and prey, (ii) The observed behavior
is a displacement activity. A heterospecific fe-
male that has attributes of mate and non-mate
and prey raises conflicts in the male. The kind
of displacement activity that arises (here: tying
of prey) is a matter of chance or of prevailing
attributes of the female, (iii) Bridal veiling is part
of the male’s repertoire which he uses only when
confronted with a particularly large and poten-
tially dangerous female. About 15% of the fe-
males responding to males during vibratory
courtship attack the approaching male, and males
smaller than females are at risk of being killed
by the female. Knowing this, we use pairs matched
in size for breeding. Thus bridal veiling could
never have been observed in conspecific pairs of
C. coccineus in our lab. (iv) Bridal veiling is an
atavism. Males regress to a behavior inherited
from, e. g., a pisaurid ancestor when confronted
with a particularly large and potentially danger-
ous female. The behavior has never been ob-
served in conspecific pairs for the same reason
as given in (iii). With the knowledge I have on
Cupiennius, 1 cannot refute any of the above con-
jectures. The above hypotheses are nevertheless
testable. For example, one can compare films of
males tying prey to the substratum with films of

67

68

THE JOURNAL OF ARACHNOLOGY

males spinning a bridal veil (hypothesis i) or one
can perform experiments using both female-larg-
er and male-larger pairs of conspecifics, the pre-
diction being to observe bridal veils with female-
larger pairs and no veils with male-larger pairs
(hypotheses iii and iv).

Let us assume now that application of silk onto
the female is part of the male’s courtship rep-
ertoire in the species enumerated above. What
functions does this behavior have? If the behav-
ior of the male handicaps the female during cop-
ulation or causes at least a brief delay of her
activity after copulation, which seems to be the
case in Pisaurina mira (Bruce & Carico 1988),
then aggressive acts of the female and/or post-
copulatory chases after the male (common in
Cupiennius coccineus and Latrodectus tredecim-
guttatus and present in Nephila clavipes', Chris-
tenson et al. 1985) should be less efficient. In this
view, the bridal veil “solution” is only purpose-
ful in those spider species in which females are
aggressive toward males during or after copula-
tion. Hence the number of species in which the
male applies silk to the female must depend on
female behavior. The close phylogenetic rela-
tionship of Ctenidae and Pisauridae on the one
hand, and the large taxonomic distance of Thom-
isidae and Dictynidae and Nephilidae and Ther-
idiidae from each other and from the other two
families (Homann 1975, Coddington 1990) on
the other hand, suggest that silk binding of the
female by the male has evolved separately sev-
eral times. But in view of the potentially lethal
weapons of spiders and of the fact that spiders
need to overcome cannibalistic tendencies when
mating, I wonder (as do Bruce & Carico 1988)
why this behavior is not more widespread among
male spiders. My answer is that aggressiveness
of females towards males during or after (and
even before) copulation is rare (Foelix 1982) or
at least much less common than spider folklore
says, which may explain why bridal veiling is so
rare. Most spiders appease their mates before
approaching them. The conjectures presented in
this paragraph can be corroborated or refuted by
investigating systematically the correlation be-
tween female and male behavior (and size) in as
many spider species as possible.

Supported by grant P7896 from the Austrian
Science Foundation (FWF) to Friedrich G. Barth.

LITERATURE CITED

Barth, F. G., H. Bleckmann, J. Bohnenberger & E.-A.
Seyfarth. 1988. Spiders of the genus Cupiennius
Simon 1891 (Araneae, Ctenidae) II. On the vibra-
tory environment of a wandering spider. Oecologia,
77:194-201.

Barth, F. G. & A. Schmitt. 1991. Species recognition
and species isolation in wandering spiders (Cupien-
nius specc., Ctenidae). Behav. Ecol. Sociobiol., 29:
333-339.

Bristowe, W.S. 1958. The World of Spiders. Collins,
London.

Bruce, J. A. & J. E. Carico. 1988. Silk use during
mating in Pisaurina mira (Araneae, Pisauridae). J.
Arachnol., 16:1-4.

Christenson, T. E., S. G. Brown, W. A. Wenzl, E. M.
Hill & K. C. Goist. 1985. Mating behavior of the
Golden-Orb-Weaving spider, Nephila clavipes: 1.
Female receptivity and male courtship. J. Comp.
Psychol., 99:160-166.

Coddington, J. A. 1990. Cladistics and spider clas-
sification: araneomorph phylogeny and the mono-
phyly of orbweavers (Araneae, Araneomorphae; Or-
biculariae). Acta Zool. Fennica, 190:75-88.

Foelix, R. F. 1982. The Biology of Spiders. Harvard
Univ. Press, Cambridge, Mass.

Homann, H. 1975. Die Stellung der Thomisidae und
der Philodromidae im System der Araneae (Cheli-
cerata, Arachnida). Z. Morph. Tiere, 80:181-202.

Merrett, P. 1988. Notes on the biology of the neo-
tropical pisaurid, Ancylometes bogotensis (Araneae:
Pisauridae). Bull. British Arachnol. Soc., 7:197-201.

Robinson, M. H. & B. Robinson. 1973. The stabi-
limenta of Nephila clavipes and the origins of sta-
bilimentum-building in araneids. Psyche, 80:277-
287.

Robinson, M. H. & B. Robinson. 1980. Comparative
studies of the courtship and mating behavior of
tropical araneid spiders. Pacific Insects Monogr., 36:
5-271.

Schmitt, A., M. Schuster & F. G. Barth. 1990. Daily
locomotor activity pattern in three species of Cu-
piennius (Araneae, Ctenidae): The males are the
wandering spiders. J. Arachnol., 18:249-255.

Starr, C. K. 1988. Sexual behavior in Z)/c?y«a vo/w-
(Araneae, Dyctinidae). J. Arachnol., 16:321-

330.

Stem, H. & E. Kullmann. 1981. Leben am seidenen
Faden. Kindler, Miinchen.

Alain Schmitt: Institut fiir Zoologie, Univer-
sitat Wien, Althanstr. 14, A- 1090 Wien, Aus-
tria.

Manuscript received August 1991, revised October 1991.

1992. The Journal of Arachnology 20:69-71

MALE OF THE BLIND CAVE GNAPHOSOID LYGROMMA ANOPS
(ARANEAE, GNAPHOSOIDEA, PRODIDOMIDAE)

FROM GALAPAGOS ISLANDS, ECUADOR

Peck and Shear (1987) described the prodi-
domid spider Lygromma amps from female
specimens collected in lava caves on Isla Santa
Cruz, Galapagos, Ecuador. This species is one of
three known eyeless gnaphosoid spiders, and one
of only two known gnaphosoid troglobites. The
lack of males left us unable to assess the rela-
tionships of L. amps.

Further field work by S. B. Peck in 1989 re-
sulted in the collection of a total of three males
from two localities. Below we describe and il-
lustrate the male, and provide some new thoughts
on the species’ relationships and biogeography.

Lygromma amps Peck and Shear
Figs. 1, 2

Lygromma anops Peck and Shear, 1987:106.

Male. — From Cueva Enrique Sevilla. Total
length, about 3.4 mm. Carapace 1.28 mm long,
1.12 mm wide. Eyes absent, carapace smooth,
without any indication of lenses. Chelicerae with
four denticles in promarginal row, single small
retromarginal denticle. Other characters closely
resembling those of female. Leg spination (meth-
od of Platnick & Shadab 1975): femora I, II: d2-
2-0, pO-0-2; III, IV: d 1 - 1-1 , pO-0- 1 , rO-0- 1 ; tibiae

1, II; V2-2-2, pO-0-1; III, IV: v2-2-2, pl-1-1, rl-
!-l, dO-1-0. Palpus (Figs. 1, 2) with bases of
retrolateral tibial apophyses fused. Embolus long,
robust; base of embolus broad, free from bulb,
shallowly, sinuously curved. Median apophysis
robust, sigmoid.

Material examined.— One male-from Cueva Enrique
Sevilla, 250 m elevation, 5 February 1989 (S. Peck); 2
males from 200 m elevation, Cuevas de Beliavista no.

2, March-April 1 989 (S. Peck); all from Isla Santa Cruz,
Galpagos.

Habitat notes.— The first locality represents a
new record for Lygromma anops. Previously
(Peck & Shear 1987) females and juveniles had
been taken in Cueva Beliavista No. 2 (Beliavista)
and Cuevas de Vargas (5 km NE of Santa Rosa).
Cueva Enrique Sevilla, like the others, is in the
moist transition {Scalesia) zone and is moder-

ately moist at all times, probably never drying
out and never flooding. Roth and Craig (1970)
noted what is probably Lygromma anops from
three juveniles in the Institut Royal des Sciences
Naturelles de Belgique, Brussels. These speci-
mens were taken from moist litter at the bottom
of a crevasse 10 m deep and about 800 m from
the dock of the Darwin Station. This is also known
as “Grieta Iguana,” and is the water source for
the Darwin Station. It is possible that Lygromma
anops, like many inhabitants of lava caves, col-
onizes new caves through interconnecting cracks
and crevices, and that these small spaces, ina-
cessible to man, are in reality its main habitat.
Thus it is not surprising that examples may be
found in any suitably cool and damp habitat that
the spiders can reach through this maze of tiny
“cavelets” (Peck 1990).

Evolutionary considerations.— In the context
of a detailed study of the spinneret morphology
of gnaphosoid spiders, Platnick (1990) has as-
signed Lygromma, formerly in the Gnaphosidae,
to a revalidated Family Prodidomidae. Platnick
and Shadab (1976b) give a discussion of the re-
lationships of the genus Lygromma. While they
were unable to analyze the evolution of the spe-
cies of the genus in a comprehensive fashion due
to missing data (many species are known only
from one sex), they did note that three species
from Venezuela {senoculatum, valencianum and
huberti) seemed to be closely related to each oth-
er but not to other species of Lygromma. The
discovery of the male of L. anops now allows the
inclusion of the Galapagos species in this group.
While Platnick and Shadab (1976b) did not sug-
gest any candidate synapomorphies, the males
of the four species differ from other Lygromma
in having the retrolateral apophyses of the male
palpal tibia with their bases close together (prac-
tically fused in anops), in having the embolus
long and originating as a separate sclerite on the
proximal surface of the bulb, and in the sigmoid
median apophysis. The epigynum of L. anops is
characterized by elaborate convoluted ducts.
These are known to occur in one of the Vene-
zuelan species, senoculatum (females of huberti

70

THE JOURNAL OF ARACHNOLOGY

Figures 1, 2.—Lygromma anops Peck & Shear: 1. Right palpus of male, retrolateral view; 2. Same, ventral
view.

and valencianum are unknown; in our original
description of L. anops we mistook Platnick and
Shadab’s illustrations of the epigynum of L. pe-
ruviana for those of huberti). Outside this group,
both convoluted ducts and a sigmoid median
apophysis are found in L. gertschi, a blind, cave-
inhabiting species from Jamaica which we orig-
inally suggested as a relative of L. anops, but this
species has a very short, distally arising embolus.
We doubt that eyelessness is a reasonable basis
for supposing close relationship; Lygromma con-
tains species with eight eyes, six, and none. Ly-
gromma simoni (Ecuador) and L. peruviana have
convoluted epigynal ducts, but males are not
known.

Are the characters we have mentioned syna-
pomorphies? Platnick and Shadab (1976a) have
suggested the Mexican genus Tivodrassus as the

sister group of Lygromma. The two Tivodrassus
species known from males have rather short em-
boli but have a sigmoid median apophysis. The
tibial apophyses are widely separated and have
a small, dark tooth between their bases. The epi-
gynal ducts are long and convoluted in all three
known species. Outgroup comparison thus sug-
gests that the long emboli and the approximated
bases of the tibial apophyses may be synapo-
morphies but that the sigmoid median apophysis
and convoluted epigynal ducts are pleisomorphic
within Lygromma. A second possible outgroup
genus is Tricongius (Platnick & Hofer 1 990). Tri-
congius amazonicus has convoluted epigynal
ducts and a rather long, basally arising embolus,
but the median apophysis is membranous, not
sigmoid, and there is a single, strong tibial
apophysis. Tricongius has a number of apomor-

RESEARCH NOTES

71

phies of its own, including a bizarre modification
of the cheliceral promargin. These questions of
relationships can be resolved only by analyzing
the full spectrum of characters in all prodidomid
genera.

Unfortunately each of the Venezuelan species
possibly related to L. amps is known only from
its type locality. However, numerous northern
South American soil arthropods have extended
their distrubution into the Isthmus of Panama,
a known source area for Galapagos biota, and
this species group of Lygromma may eventually
be found there. Again, lack of data about the
species composition and the distribution of
mainland forms hampers study of the historical
biogeography of the Galapagos soil and litter fau-
na.

The field work of S. B. Peck was supported in
part by operating grants from the Natural Sci-
ences and Engineering Research Council of Can-
ada. The administration of the Galapagos Na-
tional Park and the Charles Darwin Research
Foundation are thanked for allowing and aiding
field study in the protected habitats under their
care. W. A. Shear’s contribution to the study and
identification of Peck’s Galapagos collections has
been supported by a grant from Hampden-Syd-
ney College.

This paper is Contribution No. 457 from the
Charles Darwin Research Foundation.

LITERATURE CITED

Peck, S. B. 1990. Eyeless arthropods of the Galpagos

Islands, Ecuador: composition and origin of the

cryptozoic fauna of a young, tropical, oceanic ar-
chipelago. Biotropica, 22:366-381.

Peck, S. B. & W. A. Shear. 1987. A new blind cav-
emicolous Lygromma (Araneae, Gnaphosidae) from
the Galapagos Islands. Canadian Ent., 1 19:105-108.

Platnick, N. I. 1990. Spinneret morphology and the
phylogeny of ground spiders (Araneae, Gnaphosi-
dae). American Mus. Nov., 2978:1-42.

Platnick, N. I. & H. Hofer. 1990. Systematics and
ecology of ground spiders (Araneae, Gnaphosidae)
from central Amazonian inundation forests. Amer-
ican Mus. Nov., 2971:1-17.

Platnick, N. I. & M. U. Shadab. 1975. A revision of
the spider genus Gnaphosa (Araneae, Gnaphosidae)
in America. Bull. American. Mus. Nat. Hist., 155:
1-66.

Platnick, N. I. & M. U. Shadab. 1976a. A revision
of the spider genera Brass odes and Tivodrassus
(Araneae, Gnaphosidae) in North America. Amer-
ican Mus. Nov., 2593:1-29.

Platnick, N. 1. & M. U. Shadab. 1976b. A revision
of the spider genera Lygromma and Neozimiris
(Araneae, Gnaphosidae). American Mus. Nov.,
2598:1-23.

Roth, V. D. & P. R. Craig. 1970. VII. Arachnida of
the Galapagos Islands (excluding Acarina). Miss,
zooi. beige lies Galap. et en Ecu. (N. et J. Leleup,
1964-1965), 2: 107-124.

William A. Shear: Hampden-Sydney College,
Hampden-Sydney, Virginia 23943 USA

Stewart B. Peck: Department of Biology,
Carleton University, Ottawa, Ontario, Canada
KIS 5B6

Manuscript received August 1991, revised September
1991.

1992. The Journal of Arachnology 20:72

BOOK REVIEW

Cloudsley-Thompson, J. L. 1991. Ecophys-
iology of Desert Arthropods and Reptiles. Spring-
er-Verlag, Berlin, Heidelberg, New York. 216 pp.
77 Figs. (Price $98.00)

Cloudsley-Thompson has been writing about
the ecology and biology of desert animals for
over four decades. Drawing on that vast expe-
rience he has produced a book that is filled to
the hilt with detailed examples, often accom-
panied with photographs, of the peculiar behav-
iors and physiological characteristics of desert
reptiles and arthropods. His descriptions of such
behaviors as “sand-swimming” and “fog bask-
ing” will appeal to those of us who have come
to fancy the desert.

The book is divided into nine chapters that
focus on specific problems and adaptations of
desert life. Most of these chapters address issues
related to how reptiles and arthropods respond
to the physical constraints of the arid environ-
ment. Among these issues are thermal regulation
(chapter 4), water balance and nitrogenous ex-
cretion (chapter 5), phenology (chapter 6). Chap-
ter eight is an excellent treatment of burrowing,
mimicry, and adaptive coloration. Chapter nine
is an admirable review of important biological
interactions of desert animals (e.g., competition,
predation, etc.).

Arthropods and reptiles are discussed in sep-
arate subsections within each chapter and the
emphasis of the book is divided equally between
the two major groups. This organization, along
with the strong chapter and subsection intro-
ductions, makes it possible for one to read the
book from start to finish focusing only on either
reptiles or arthropods. With respect to arthro-
pods, Cloudsley-Thompson has drawn heavily
on the works of R. A. Bradley and G. Polls (Scor-
pions), C. S. Crawford (myriopods), E. B. Edney
and A. C. Marsh (insects), N. F. Hadley (scor-
pions and insects), W. F. Humphreys and B. Y.
Main (spiders); and his review of these works is
concise. The book is a valuable reference for those
interested in deserts.

Nevertheless, the book has a number of short-
comings that detract from its usefulness as a pri-

mary source on the ecology of desert animals.
Cloudsley-Thompson motivates the work by
suggesting that there is a need to compare and
contrast the adaptations of the two most suc-
cessful groups of desert animals, reptiles and ar-
thropods, to “..the various parameters of the des-
ert environment. .”(p. 1). Yet these “parameters”
are never clearly delineated save for the chapter
headings and, with few exceptions (most notably
the section on burrowing), little effort is given to
actually comparing the features of these two dis-
parate groups. Even the discussion of conver-
gence, where such comparisons could logically
be made, is divided into separate sections on
reptiles and arthropods. One is left wondering
whether the two groups are comparable.

Moreover, after 168 pages of excellent exam-
ples of adaptations to desert conditions, Cloud-
sley-Thompson, apparently under the strong in-
fluence of Bradshaw (1986), dismisses the
significance of those adaptations by suggesting
that most reptiles and arthropods are preadapted
for the desert extremes. He writes: “During the
course of this book it must have become appar-
ent that neither arthropods nor reptiles show par-
ticularly marked desert adaptations” (p. 1 69). An
astonishing statement from someone whose life’s
work has been given over to explaining how an-
imals survive in the desert.

I found some of the sections to be unneces-
sarily long and wordy. In particular, the discus-
sion of the theoretical aspects of parallel evolu-
tion and convergence is poorly developed. Also,
for those unfamiliar with either the reptiles or
the arthropods the index will be somewhat dif-
ficult to use since, with few exceptions, common
names are not included.

LITERATURE CITED

Bradshaw, S. D. 1988. Desert reptiles: a case of ad-
aptation or pre-adaptation? J. Arid. Environ., 14:
155-174.

Gary L. Miller: Department of Biology, The
University of Mississippi, University, Missis-
sippi 38677 USA.

Manuscript received September 1991.

72

INSTRUCTIONS TO AUTHORS

Manuscripts are preferred in English but may be accepted
in Spanish, French or Portuguese subject to availability of
appropriate reviewers. Authors whose primary language is not
English, may consult the Associate Editor for assistance in
obtaining help with English manuscript preparation. All man-
uscripts should be prepared in general accordance with the
current edition of the Council of Biological Editors ( CBE) Style
Manual unless instructed otherwise below. Authors are ad-
vised to consult a recent issue of the Journal of Arachnology
for additional points of style. Send three identical copies of
the typed material together with copies of illustrations to:

Gary L. Miller, Associate Editor
The Journal of Arachnology
Department of Biology
The University of Mississippi
University, Mississippi 38677 USA
Telephone: (601) 232-7495
FAX: (601) 232-5144
Bitnet: BYMILLER@UMSVM
Correspondence relating to manuscripts should also be di-
rected to the Associate Editor.

Manuscripts longer than 1 500 words should be prepared as
Feature Articles, shorter papers as Research Notes.

FEATURE ARTICLES

Title page.— Place in the upper left comer the complete name,
address, and telephone number of the author with whom proofs
and correspondence should be exchanged. Include EAX num-
bers and Bitnet addresses if available. Four lines below, type
the title, centered, all in capital letters. Another four lines be-
low, beginning with the left margin, type each author’s name
on a new line followed by the address.

Abstract.— The heading should be placed at the beginning
of the first paragraph set off by a period. In articles written in
English, a second abstract in an acceptable language may be
included pertinent to the nationality of the author(s) or geo-
graphic region(s) emphasized in the article. A second abstract
in English must be included in articles not written in the latter
language.

Text.— Double-space text, tables, legends, etc. throughout.
Three levels of heads are used. The first level (METHODS,
RESULTS, etc.) is typed in capitals and on a separate line.
The second level head begins a paragraph with an indent and
is separated from the text by a period and a dash. The third
level may or may not begin a paragraph but is underscored
and separated from the text by a colon. Use only the underline
method to indicate italics; DO NOT USE WORD PROCES-
SOR ITALICS CODES. Use the metric system unless quoting
text or referencing collection data. All decimal fractions are
indicated by the period regardless of language of the text.

Citation of references in the text.— Include within parenthe-
ses the surnames of the author(s) followed by the date of pub-
lication. A comma separates multiple citations by the same
author(s) and a semicolon separates citations by different au-
thors, e.g., (Smith 1970), (Jones 1988; Smith 1989), (Smith
1986, 1987; Smith & Jones 1989; Jones et al. 1989).

Literature cited section.— Use the following style:
Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid
composition of major ampullate gland silk (dragline) of Ne-
phila clavipes (Araneae, Tetragnathidae). J. Arachnol., 18:
297-306.

Krafft, B. 1982. The significance and complexity of com-
munication in spiders. Pp. 1 5-66, In Spider Communica-
tions: Mechanisms and Ecological Significance. (P. N. Witt
& J. S. Rovner, eds.). PrincetonUniversity Press, Princeton.
Footnotes.— Footnotes are only permitted on the first printed
page to indicate current address or other information con-
cerning the author(s). All footnotes are placed together on a
separate manuscript page.

Running head.— Author sumame(s) and an abbreviated title
must be typed all in caps and must not exceed 60 characters
and spaces. The running head is placed by itself on a separate
manuscript page.

Taxonomic articles.— Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Associate
Editor.

Tables.— Each table, with the legend above, should be placed
on a separate manuscript page. Only the appropriate horizontal
lines should be included. Use no footnotes; instead, include
all information in the legend. Make notations in the text mar-
gins to indicate the preferred location of tables in the printed
text.

Illustrations.— Original art work should be kept by the au-
thor(s) until final acceptance for publication. All art work must
be mounted on stiff board ready for reproduction. In line draw-
ings, pay particular attention to width of lines and size of
lettering when reductions are to be made by the printer. Mul-
tiple photos assembled on a single plate should be mounted
with only a minimum of space separating them. In the case of
multiple illustrations mounted together, each illustration must
be numbered sequentially rather than given an alphabetic se-
quence. Written on the back should be the name(s) ofauthor(s)
and an indication of top edge. The author should indicate
whether the figure should be one column or two columns in
width. The overall dimensions should be no more than 1 1
inches (28 cm) x 14 inches (36 cm). Larger drawings present
greater difficulty in shipping and greater risks of damage for
which the JOA assumes no responsibility. The author is en-
couraged to consider PMT (PhotoMechanical Transfer) repro-
ductions for large drawings and to consult an Editor for details
on appropriate reductions. Make notations in the text margins
to indicate the preferred position of illustrations in the printed
text.

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Adhere to the
following styles:

Eigures 1-4.— A-us x-us, male from Timbuktu: 1, left leg;
2, right chelicera; 3, dorsal aspect of genitalia; 4, ventral aspect
of abdomen.

Figures 27-34.— Right chelicerae of species of A-us from
Timbuktu: 27,29,31,33, dorsal views; 28, 30, 32, 34, prolater-
al views of moveable finger; 27, 28, A-us x-us, holotype male;
29, 30, A-us, w-us, holotype male; 31, 32, A-us z-us, holotype
male; 33, 34, A-us y-us, male. Scales = 1.0 mm.

Assemble manuscript for mailing.— Assemble the separate
sections or pages in the following sequence; title page, abstract,
text, figure legends, footnotes, running head, tables with leg-
ends, figures.

Page charges and reprints.— The current charge per journal
page will be assessed as follows: $50— nonmembers, $45—
members with grant, institutional, or other support for page
costs; $25— regular members who do not have support for page
costs; $20— student members who do not author with full
members and who do not have support. Reprints are available
only from the printer at current costs.

Waiver of Page Charges.— The Editor will consider requests
for a waiver of page charges for manuscripts which have been
accepted for publication. Requests should be made in writing
and should be submitted to the Associate Editor along with
the final revision of the manuscript. The author must certify
that neither grant nor institutional funds are available to cover
all of the publication costs. The Society will normally agree to
waive up to a maximum of 80% of the page charges, contingent
on the availability of funds; the author is expected to find
alternative sources for the remainder.

RESEARCH NOTES

Instructions above pertaining to feature articles apply also
to research notes except that abstracts and most headings are
not used and the byline follows the Literature Cited section.

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 20 Feature Articles NUMBER 1

Hawaiian spiders of the genus Tetragnatha II. Species from natural areas of
windward east Maui, Rosemary G. Gillespie 1

Phrynidae (Amblypygi) from Andros Island, Bahamas, with notes on dis-
tribution patterns, recent origin and allometry, D. Jonathan Browne 18

Web construction by Modisimus sp. (Araneae, Pholcidae), William G. Eber-
hard 25

Variation in Schizocosa (Araneae: Lycosidae), Metaphidippus and Phidippus
(Araneae: Salticidae), William W. M. Steiner, Matthew H. Greenstone
and Gail E. Stratton 35

Systematics of Hypochilus sheari and Hypochilus coylei, two southern Ap-
palachian lampshade spiders (Araneae, Hypochilidae), Ronald P. Hujf
and Frederick A. Coyle 40

On the function of harlequin beetle-riding in the pseudoscorpion, Cordy-
lochernes scorpioides (Pseudoscorpionida: Chemetidae), David W. Zeh
and Jeanne A. Zeh 47

New species of crab spiders from Baja California Sur (Araneae: Thomisidae),

Maria- Luisa Jimenez 52

Supercooling and its ecological implications in Coelotes atropos (Araneae,

Agelenidae), Kefyn M. Catley 58

Research Notes

Parasitism of pseudoscorpions (Arachnida) by Mermithidae (Nematoda),

George O. Poinar, Jr. and Bozidar P. M. Curcic 64

Conjectures on the origins and functions of a bridal veil spun by the males

of Cupiennius coccineus (Araneae, Ctenidae), Alain Schmitt 67

Male of the blind cave gnaphosoid Lygromma anops (Araneae, Gnapho-
soidea, Prodidomidae) from Galapagos Islands, Ecuador, William A.
Shear and Stewart B. Peck 69

Book Review

Ecophysiology of Desert Arthropods and Reptiles (by J. L. Cloudsley-

Thompson), Gary L. Miller 72

The Journal of
ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 20

1992

NUMBER 2

THE JOURNAL OF ARACHNOLOGY

EDITOR: James W. Berry, Butler University

ASSOCIATE EDITOR: Gary L. Miller, The University of Mississippi

EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E.
Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C.
Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ.; C. D. Don-
dale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galiano,
Mus. Argentino de Ciencias Naturales; M. H. Greenstone, BCIRL, Columbia,
Missouri; N. V. Horner, Midwestern State Univ.; D. T. Jennings, Garland, Maine;
V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ.; E. A. Maury, Mus.
Argentino de Ciencias Naturales; N. I. Platnick, American Mus. Natural Hist.;
G. A. Polls, Vanderbilt Univ.; S. E. Riechert, Univ. Tennessee; A. L. Rypstra,
Miami Univ., Ohio; M. H. Robinson, U.S. National Zool. Park; W. A. Shear,
Hampden-Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa
Rica.

The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the
study of Arachnida, is published three times each year by The American Arach-
nological Society. Memberships (yearly): Membership is open to all those inter-
ested in Arachnida. Subscriptions to The Journal of Arachnology and American
Arachnology (the newsletter), and annual meeting notices, are included with mem-
bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or
$90 (all other countries). Inquiries should be directed to the Membership Secretary
(see below). Back Issues: Susan E. Riechert, Department of Zoology, Univ. of
Tennessee, Knoxville, TN 37916 USA. Undelivered Issues: Allen Press, Inc.,
1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: Allen R. Brady (1991-1 993), Biology Department, Hope College,
Holland, Michigan 49423 USA.

PRESIDENT-ELECT: James E. Carico (1991-1993), Department of Biology,
Lynchburg College, Lynchburg, Virginia 24501 USA.

MEMBERSHIP SECRETARY: Norman I. Platnick (appointed), American
Museum of Natural History, Central Park West at 79th St., New York, New
York 10024 USA.

TREASURER: Gail E. Stratton (1991-1993), Department of Biology, Albion
College, Albion, Michigan 49224 USA.

SECRETARY: Brent Opell (1991-1993), Department of Biology, Virginia Poly-
technic Institute and State University, Blacksburg, Virginia 24061 USA.
ARCHIVIST: Vincent D. Roth, Box 136, Portal, Arizona 85632 USA.
DIRECTORS: Matthew H. Greenstone (1990-1992), George W. Uetz (1991-
1993), Charles E. Griswold (1991-1993).

HONORARY MEMBERS: C. D. Dondale, W. J. Gertsch, H. Homann, H. W.
Levi, A. F. Millidge, M. Vachon, T. Yaginuma.

Cover illustration: SEM photomicrograph of the ocularium of the opilionid Odiellus pictus (Wood).
The species has a prominent trident of spines at the anterior border of its cephalothorax. Found in
the eastern United States. Photo by Steven Murphree.

Publication date: 2 December 1992

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.

1992. The Journal of Arachnology 20:73-87

TEMPORAL AND SPATIAL SEGREGATION OF WEB~BUILDING
IN A COMMUNITY OF ORB^WEAVING SPIDERS

David Ward' and Yael Lubin; Mitrani Centre for Desert Ecology, Jacob Blaustein
Institute for Desert Research, Ben Gurion University of the Negev, Sede Boqer
84993, Israel

ABSTRACT. The temporal pattern of activity and spatial distribution of six species of nocturnal orb-weaving
spiders (Araneae: Araneidae and Tetragnathidae) were examined in coastal hedge vegetation in Israel. In Autumn,
small spiders of all species built their webs early in the evening and progressively larger spiders put their webs
up through the night. This activity pattern corresponded to the change in sizes of flying insects throughout the
night. There was no interspecific segregation in time of activity. Spiders were highly clumped in space, but
showed interspecific segregation only in web height. In Autumn, Nuctenea suspicax was the most abundant
species, while in Spring Singa lucina predominated. During the latter season, spiders had two periods of activity:
evening (at dusk) and morning (pre-dawn). Morning-active spiders had larger webs and larger dutches than
evening-active spiders. As in Autumn, there was little interspecific segregation in time of activity or in spatial
distribution. Spider removal experiments suggest that the timing of activity does not change following density
reduction, but that individuals that were previously inactive may take advantage of the newly available spaces.
The number of active spiders increased when sites for web attachments were added, supporting the hypothesis
that space availability limits spider activity. The results are discussed in terms of the importance of niche
partitioning in time and space.

The study of the factors influencing the dis-
tribution and abundance of animals has long been
fundamental to ecology (Andrewartha & Birch
1954). The dispersion of animals in time and
space has often been used to ascertain the influ-
ence of conspecifics and heterospecifics on the
behavioral ecology of a variety of animals (Da-
vies 1 978). Web-building spiders are particularly
suitable for such studies because: (1) they spend
much of their time in a fixed position, facilitating
measurement of dispersion, (2) they are easily
manipulated, by removal or supplementation ex-
periments, and (3) many species often coexist in
large numbers and in relatively small areas (Rob-
inson etal. 1974; Lubin 1978; HofFmaster 1985).

Some studies of web-building spiders have
shown that patterns of dispersion are related to
aggressive interactions among individuals, which
may influence website selection and web-build-
ing behaviour (Riechert 1981; Pasquet 1984; Le-
borgne & Pasquet 1987). Other studies, however,
have indicated that individuals may aggregate in
order to take advantage of clumped prey distri-
butions or to reduce predation risk (e.g., Lubin

' Present address: Department of Zoology, University
of British Columbia, Vancouver, British Columbia, V6T
IZ4 Canada.

1974; Uetz et al. 1978, 1982; Schoener & Toft
!983a). In some communities of orb-weaving
spiders, there is considerable separation of spe-
cies according to the vegetation types selected
(Enders 1973; Harwood 1974), heights at which
webs are placed (Enders 1974; Taub 1977; Olive
1980; Brown 1981), and types of prey taken (Ol-
ive 1980). These differences have been used as
indications of interspecific competition (Enders
1974; Brown 1981; Spiller 1984), although Spill-
er (1984) noted that seasonal reversals of com-
petitive advantage may occur. These studies con-
trast with that of HofFmaster (1 985) who showed
that, in multi-species orb-weaving spider com-
munities in Panama, the spatial distributions of
species were not significantly different from ran-
dom. HofFmaster (1985) suggested, like Wise
(1984), that interspecific interactions are not im-
portant in orb-weaving spider communities.

In this study, we examine the temporal and
spatial distribution of web-building behavior in
a guild of six nocturnal orb-weaving spider spe-
cies (Araneae, Araneidae and Tetragnathidae) in
coastal hedge vegetation at Ma’agan Michael, near
Haifa, Israel. The structural simplicity of the veg-
etation, combined with high densities of web-
building spiders, suggested a potential for space
limitation. Natural history observations indicat-

73

74

THE JOURNAL OF ARACHNOLOGY

ed that there was temporal segregation of web-
building behavior in these nocturnal orb-weav-
ing spiders. These changes in activity appeared
size-correlated, with larger spiders building their
webs later in the night than small spiders. We
set out to determine if the different species seg-
regated temporally and to assess the possible
causes of this activity pattern. We then experi-
mentally manipulated both spider densities and
the space available for web building in a series
of short-term experiments in order to examine
the interaction between temporal and spatial
scales of community organization.

METHODS

Study area and spider dispersion.— Six orb-
weaving spider species (Table 1) occur in the
almost homogeneous hedges of the perrenial
composite Inula viscosa (L.) Ait. surrounding fish
ponds at Ma’agan Michael on the Mediterranean
coast of Israel. These hedges are about 1-2 m
high and 1 m wide and are bounded on one side
by sand roads and by ponds on the other. The
vegetation is typical of coastal Mediterranean
pond-edge communities. Other plant species oc-
cur patchily, notably a reed {Phragmites sp.),
tamarisk trees {Tamariskia sp.) and a grass {Bw-
mus sp.). The study was conducted in Autumn
of 1986, 1988 and 1989 and in Spring 1989. We
measured ambient temperature and relative hu-
midity with a sling psychrometer (Bacharach,
Inc.) at the study site before each census. Cli-
matic conditions during the study periods were
similar: nighttime ambient temperatures record-
ed in Autumn near the ponds were 22-13.5 °C
and in Spring from 22-14 °C. The relative hu-
midity at night increased from 70 to > 90% from
early evening to about 2300 h and remained high
until dawn.

Plots were established in I. viscosa hedges. We
used three 15 m-long plots in our experiments
in September-October 1986 and 1988, one 18
m plot in our experiments in April-May 1989
and three 3 m long plots (each separated by one
meter) in October 1989. The plots were delin-
eated by vertical poles connected by string at a
height of 1 m above the ground, with the excep-
tion of the last mentioned experiments which
involved string supplementation, and were sub-
divided into 1 m^ sections.

Spiders with intact orb webs were denoted as
active. Active spiders in the study plots were

counted and identified to species at approxi-
mately 2 h intervals throughout the night. We
estimated spider size to the nearest mm, and
measured maximum web diameter and web
height above the ground (measured from the hub).
To determine the relationship between spider
length and web diameter, we removed spiders
and measured them in the laboratory with ver-
nier calipers. All mean measurements are given
± SE. Eggsacs of Singa lucina (Audouin) were
collected in May 1 989 to determine whether there
were changes in reproductive output associated
with different temporal activity patterns in this
species.

Activity of spiders was studied in conjunction
with trapping of flying insects (potential spider
prey) using a blacklight placed above a tray of
preservative (70% ethyl alcohol) in nearby hedge
vegetation. While the traps may not capture the
different prey types in the same proportions as
webs (Eberhard 1991), we believe that the tem-
poral distribution of the numerically dominant
insect groups in this structurally simple habitat
is adequately represented in our lighttrap sam-
ples. Earlier observations showed that orb webs
present in the hedges in early evening were choked
with small midges that emerge from the ponds
at dusk. Webs present later in the night were
largely free of midges, but had large, scale-lined
holes attributed to moth interceptions.

We used the standard Index of Dispersion, the
variance divided by the mean (I = sVx) to mea-
sure the randomness and clumping of webs: I =
1 denotes random dispersion and I> 1, clumped
dispersion. This index has a x“ distribution and
was tested against this for significance (Pielou
1977). We used the number of spiders in the 1
m^ quadrats for the purpose of this analysis. The
index of dispersion is affected by sample size,
although this effect is considered to be minimal
with the sample sizes we used (18 quadrats in
April-May 1989 to 45 in September-October
1986 and 1988) (Pielou 1977). We calculated the
index of dispersion separately for each census to
minimize errors due to possible lack of indepen-
dence of censuses (some spiders remained active
over more than one census period).

Vegetation density was measured by line tran-
sects at 1 m intervals along the hedge and at 40
cm increments above the ground to determine
whether there were differences in vegetation
structure among the quadrats. Vegetation den-
sity is expressed as the mean number of 5 cm

WARD & LUBIN-TEMPORAL AND SPATIAL SEGREGATION OF SPIDER ACTIVITY

75

sections of each transect line covered by vege-
tation (maximum = 20). In October 1988 and
in May 1989, spiders were censused in three ad-
ditional 6 m plots near the ponds to determine
if the different orb-weaver species had separate
habitat preferences. These plots, also in the hedg-
es, were chosen to represent a variety of hedge
habitats differing in vegetation structure.

Discriminant analysis was used to help elu-
cidate the usage of different habitat and envi-
ronmental selection features (vegetation density,
distance from the front or exposed side of the
plots, web height, distance along the plots) by
the various species. Discriminant analysis dis-
tinguishes among groups (spider species) by
weighting and linearly combining independent
variables (habitat features) into a new variable,
or discriminant function, which gives maximal
statistical separation of the groups (species)
(Green 1971). By extracting a second, orthogo-
nal, discriminant function, overlap is viewed in
a plane. As many discriminant functions are ex-
tracted as contribute to significant discrimina-
tion among groups. Plotting species centroids
along all relevant discriminant function axes
(those axes that are statistically significant in a
Wilks X test) gives a visual representation of
overlap along a reduced set of axes. Standardized
coefficients of the discriminant functions indi-
cate the associations of the function with each
of the original variables.

Experimental manipulations. — Experimental
manipulations of spider density and of space
available for web construction were performed
over periods of 3-5 days each in Autumn 1988
and 1989 and in Spring 1989.

Removal experiments: We conducted re-
moval experiments in order to test whether the
temporal stratification of activity was due to space
limitation acting on spiders building their webs
at preferred times of the night. In September

1988, we removed all spiders as they became
active each hour through the night in a 15 m
plot. The temporal stratification of active spiders
of the different size classes was compared with
that of active spiders in an adjacent unmani-
pulated 1 5 m plot.

To determine if changes in spider numbers
following removal could be explained by move-
ment of spiders from adjacent areas, we con-
ducted a second removal experiment in April

1989. We removed all active spiders in two cen-
tral 3 m plots and compared spider activity in

these plots with that in two adjacent 3 m plots
on either side.

In May 1989, we removed all the predawn
active spiders (henceforth “morning spiders”)
from two 3 m plots to determine whether spiders
active in the post-dusk period (henceforth “eve-
ning spiders”) would become “morning spiders”.
In another 3 m plot, all “evening spiders” were
removed, to determine whether “morning spi-
ders” would change their activity pattern to be-
come “evening active”. Activity in these exper-
imental plots was compared with that in two
neighboring control plots on either side. The re-
treats (curled leaves in which the spiders sat when
not active on their webs) of “morning” and “eve-
ning” spiders were marked with different colors
of paint to facilitate recognition.

Space-supplementation experiments: We
conducted space-supplementation experiments
by adding strings for web attachment to two 3
m plots in spring 1989. In one, string was tied
around the perimeter of the plot at 50 cm inter-
vals both vertically and horizontally. String was
also tied in horizontal and vertical planes through
the plot, with the effect that the plot was divided
into 50 cm x 50 cm x 50 cm cubes with a total
of 44 m of string. In the second experimental
plot, string was tied in the same manner, but at
25 cm intervals. Thus, this plot was divided into
25 cm X 25 cm x 25 cm cubes and contained
1 1 6 m of string. Another 3 m plot (control) had
only a single line of string demarcating the pe-
rimeter.

In these experiments, each plot was first cen-
sused for one night prior to string supplemen-
tation in order to determine the number of spi-
ders active under pre-experimental conditions.
The number of spiders active subsequent to sup-
plementation was then compared with the initial
density in each plot, and with the control plot
on the same night.

RESULTS

Temporal dispersion patterns.— Activity pat-
terns of spiders in Autumn 1 986 and 1988 showed
a peak of web-building at dusk, with additional
webs appearing through the night. There was a
significant increase in the size of new webs
(ANOVA, P < 0.05 for each of 8 nights in 1986
and 1 988) from early evening until morning (Fig.
1). Web diameter was highly positively correlat-
ed with spider body size (Spearman rank cor-
relations, 1986: R = 0.86, n = 62; 1988: R =

Web Diam., cm

76

THE JOURNAL OF ARACHNOLOGY

Table 1.— The species of orb-weaving spiders in hedge vegetation surrounding the ponds at Ma’agan Michael.
The two Tetragnatha species are as yet unidentified. They were distinguished by the presence of an unhumped
(Sp. A) or humped (Sp. B) opisthosoma.

Species (family)

Orbweb

Retreat

Singa lucina (Araneidae)

vertical

curled-leaf

Nuctenea suspicax (Araneidae)

variable

curled-leaf

Larinia chloris (Araneidae)

vertical

underside of grass-blade

Neoscona subfusca (Araneidae)

generally vertical

underside of leaf

Tetragnatha sp. A (Tetragnathidae)

variable

twig

Tetragnatha sp. B (Tetragnathidae)

variable

twig

0.75, n = 87, P < 0.0001 in both years). The
increase in web size throughout the night was not
due to a shift in species activity. In a two-factor
ANOVA (factors = species and time), there was
no significant interaction effect {P > 0.05) over
any of the census nights, indicating absence of
interspecific segregation in activity patterns.
Rather, a large proportion (>50%) of spiders
building their webs in the early evening were
immatures of all species.

Blacklighting of insects in Autumn showed that
insects had the same size-based temporal activity
patterns as the spiders. Many small insects were
active early in the evening and progressively larg-
er insects became active through the night (Fig.
2). At dusk, there was an emergence peak of
midges (Nematocera, Chironomidae) from the

ponds; by early morning, most active insects were
large moths (Lepidoptera).

In Spring 1989, we attempted to examine the
temporal segregation of web-building behaviour
more rigorously. However, at this time of year
few immature spiders occurred on the plots and
web-building was less evenly spaced throughout
the night than in Autumn. There were two main
periods of web-building, at dusk and again before
dawn (Fig. 4). Insect activity patterns (deter-
mined by blacklighting) followed the pattern de-
scribed above for Autumn; i.e., many small in-
sects appeared early in the night, and larger insects
increased in numbers towards morning (Fig. 3).

We marked the retreats of Singa lucina, which
was the predominant species present in Spring.
As the retreat is connected to the orb web by a

Fig. 1.— Changes in the diameter of new webs (means and 95% C. 1. for all species combined) constructed
throughout the night in Autumn 1988. Numbers of spiders are above each time period. Data for a single
representative night are shown.

WARD & LUBIN- TEMPORAL AND SPATIAL SEGREGATION OF SPIDER ACTIVITY

77

Time at end of sampling period, hOO

1930 2230 230 530

Time at end of sampling interval , hOO

Figs. 2, 3.— Numbers of insects of different size classes appearing throughout the night. Size classes are: hatched
= < 10 mm total body length, clear = > 10 mm total body length; 2. Autumn 1988; 3. Spring 1989.

78

THE JOURNAL OF ARACHNOLOGY

(/>

<D

Fig. 4.— Numbers of spiders (all species) on new webs (clear) and on old webs (hatched) throughout a 36 h
period in April 1989.

signal line, the activity of individual spiders could
be tracked. Over two census days in May 1989,
19.5% of the individuals were active on newly-
constructed webs in the evening only, 51.2% in
the morning only, and 29.3% were active both
morning and evening (« = 41 spiders total).
“Morning only” spiders had significantly larger
webs than “evening only” spiders (ANOVA, P
< 0.05; Table 2). There were also differences in
web geometry between spiders active at different

T able 2. — Web diameters of “evening only”, “morn-
ing only” and “morning and evening” (both) individ-
uals of S. lucina in April 1989. Web diameters of
“morning only” and “evening only” spiders differed
significantly (ANOVA, P < 0.05).

Activity

Web diameter

n

mean

SE

Morning

16.6 cm

0.7

44

Evening

1 1.0 cm

1.04

96

Both

14.2 cm

0.68

12

times. New webs constructed in the morning had
significantly more spirals per cm length of radius
(3.1 ±0. 1 7, « = 33) than webs constructed in the
evening (2.3 ±0. 1 5, « = 3 1 , ANOVA, Z’ < 0.0 1 ).

We counted the number of eggs laid by “morn-
ing only”, “evening only” and “morning and
evening” 5. lucina to determine possible fitness
consequences of the different activity patterns.
“Morning only” spiders laid significantly more
eggs (57.1 ±3.8 eggs per spider, n = 52) than
“evening only” spiders (49.8 ±5.9 eggs per spi-
der, n = \ \) (Mann-Whitney (7-test, P = 0.04).
“Morning and evening” spiders also produced
significantly more eggs (70.3 ±9.0 eggs per spi-
der, n = 9) than “evening only” spiders {P <
0.05). The difference in number of eggs between
“morning and evening” and “morning only” spi-
ders was not statistically significant (Mann-
Whitney (/-tests, P > 0.05). However, 2 of the
9 “morning and evening” clutches were parasit-
ized by an unidentified dipteran, significantly in-
creasing the variance in clutch size in this group.
There was no significant correlation (P > 0.05)
between spider size and egg number.

WARD & LUBIN-TEMPORAL AND SPATIAL SEGREGATION OF SPIDER ACTIVITY

79

Table 3.— Seasonal distribution of orb-weaving spiders on /. viscosa hedges at Ma’agan Michael: Percentages
of different species found in the study plots on selected evening (E) or predawn (M) censuses, n = Total number
of spiders.

Species

Nov 1988

Oct 1989

April 1989

May 1989

E

M

E

M

E

M

E

M

n

82

39

45

57

65

91

107

160

S. lucina

12.2

20.5

4.4

5.3

76.9

100

67.3

88.8

L. chloris

25.6

15.4

4.4

0

0

0

1.9

0

N. suspicax

4.9

23.1

37.8

35.0

16.9

0

24.3

10.6

N. subfusca

1.2

0

0

0

6.2

0

0.1

0

Tetragnatha A

1.2

0

2.2

0

0

0

4.7

0.1

Tetragnatha B

1.2

0

0

0

0

0

0.1

0

Immatures

53.7

41.0

59.7

59.7

0

0

0

0

Seasonal abundance and dispersion pat-
terns.—All orb-weaving species were present in
both Autumn and Spring sampling periods.
However, the relative abundances of the different
species and their age distribution varied season-
ally (Table 3). Singa lucina, Nuctenea suspicax
(O.P.-Cambridge) and Larinia chloris (Audouin)
together comprised about 50% of the spiders in
Autumn. In Spring, however, S. lucina alone
comprised over 70% of the total number of spi-
ders. N. suspicax was the second-most abundant
species in Spring (10-1 7% of individuals) and L.
chloris was rare. In Autumn, many small im-
matures of all species (< 2 mm body length) were
present (>50% of spiders in the evening census-
es), whereas in Spring, most spiders were larger
juveniles or adults and only very few (<0.01 %)
were small immatures.

In an attempt to elucidate the patterns of in-
terspecific distribution, discriminant function
analyses of both Autumn and Spring data were
conducted, using 3-dimensional location in the
quadrats and vegetation density in (1) the ex-
perimental plots before manipulation in Autumn
(November 1 988) and, (2) the experimental plots
in Spring (May 1989) and the three additional
hedge vegetation plots in May 1989. We first ran
discriminant function analyses comparing 5 mm
size classes of spiders, but derived no significant
discriminant function (Wilks X, P > 0.05) for
any of the censuses.

The discriminant analyses among species de-
rived two significant discriminant functions
(Wilks X, P < 0.05) for the Autumn data (Fig.
5). The first discriminant function explained 75%
of the variation in distribution, and was most
closely related to web height (standardized dis-
criminant coefficient, SDC = 0.60) and vegeta-

tion density (SDC = -0.54). The second dis-
criminant function, explaining an additional 1 7%
of the variation among species, was most closely
related to distribution along the length of the
hedge (SDC = 0.96). The Spring spider distri-
butions produced only one significant discrimi-
nant function which explained 87% of the vari-
ation in the data. This function was most closely
related to web height (SDC = 0.92).

To examine these patterns in greater detail, we
tested for (1) overall clumping of individuals
within the hedges and (2) differences among spe-
cies in their distribution within the plots. In Au-
tumn, the dispersion pattern of all spiders com-
bined was significantly clumped at most times
of night. The index of dispersion (I) was > 1 (i.e.,
non-random) for 9 of 14 censuses over 3 nights
(x^ tests, P < 0.001, « = 45 1 m^ quadrats).
However, all species were found in all quadrats
(though not always at the same census), and the
individual species were randomly dispersed along
the hedges both in evening and in early morning
censuses (x^ tests for S. lucina, L. chloris, N. sus-
picax and unidentified immatures on four census
dates, P > 0.05, « = 45 quadrats).

We tested whether N. suspicax and L. chloris
(the second- and third-most abundant species)
were significantly clumped away from the most
abundant species S. lucina. The null hypothesis
was that the likelihood of the nearest neighbour
being a conspecific or S. lucina was no different
to that predicted by random association of in-
dividuals, i.e., that there was no interspecific sep-
aration. The null hypothesis could not be rejected
(x^ tests, P > 0.05, « = 45 quadrats). Thus, N.
suspicax and L. chloris were not clumped away
from 5. lucina.

We examined the three-dimensional distri-

80

THE JOURNAL OF ARACHNOLOGY

.6

Nuctenea

(16)

.2

Unidentified

(222)

Lorinia

(43)

O

f—

O

h-

z

<

“.2

q:

o

w

Q

-1.0

Singa

(78)

i— — 1 1 1.^ I I

-1.0 -.6 -.2 .2 .6
DISCRIMINANT FUNCTION 2

Fig. 5.— Spatial separation of species on experimental plots in Autumn 1988. Separation of the species’
centroids in a two-dimensional space determined by discriminant function analysis. Nuctenea subfusca and
Tetragnatha spp. were excluded because of small sample sizes (< 10). Sample sizes in parentheses.

bution patterns of webs within the 1 quadrats.
The only difference among the species was along
the height axis: webs of 5’. lucina were signifi-
cantly lower in the vegetation than were those of
other species (ANOVA, P < 0.001; Fig. 6). This
difference was not due to differences in body size
among the species (ANOVA, P > 0.05). There
was no correlation within any species between
spider size and web height {P > 0.05).

Although we did not find any clear spatial seg-
regation among the different species in Autumn,
there could still occur density-dependent influ-
ences on the overall abundance of one species
on another. In such a situation, the slopes of the
regression of the abundance of one species on
another is a direct estimate of the competition
coefficient for that species pair (Hallet & Pimm
1 979). We tested for such effects among the three
most common species (5'. lucina, N. suspicax and
L. chloris) using pairwise linear regressions of
spider densities per quadrat. There were no sig-

nificant correlations between any pairs of species
{P > 0.05).

In Spring, the dispersion of all spiders com-
bined was highly clumped at all times (I » \, P
< 0.001 for 8 census dates, « = 18 quadrats per
census). S. lucina and N. suspicax were tested for
departures from randomness among the quad-
rats. N. suspicax was randomly distributed among
the quadrats (« = 1 8 quadrats, 3 census dates),
as was S. lucina in the evening censuses (« = 1 8
quadrats, 2 census dates). However, in the morn-
ing samples, the distribution of S. lucina was
significantly clumped (x^ tests, P < 0.001, n =
1 8 quadrats, 3 census dates). Clumping was as-
sociated with the high densities of webs of S.
lucina in the morning samples (14-24 spiders/
m^): there was a significant correlation between
the value and spider density for the 8 dates
tested (i? = 0.77, P = 0.01).

We tested for differences in species distribu-
tion in plots differing markedly in vegetation

120

6

2*

sz

-g ^0

78

16

43

222

+

i

Singa

_ — I __i -

Nuctenea Larinia

Immatures

Singa Nuctenea Neoscona Larinia

Figs. 6, 7.-6. Web heights (cm) of dominant species in Autumn. Means and 95% C. I. are shown for S.
lucina, N. suspicax, L. chloris and unidentified immatures (< 2 mm body length). Numbers of spiders are above
each census; 7. Web heights (cm) of dominant species in Spring, as in Fig. 6. Shown are web heights on the
experimental plot (April and May #1) and on three additional 3 m vegetation plots in May (May #2).

Web Diam., cm

82

THE JOURNAL OF ARACHNOLOGY

25

e

E xperimental

E

o

b

JD

a>

15

5

I

18

21

21

“T T

2230 0145

TIME, hOO

_j — _

0430

20

Control

22

15

10

33

“T"

17

41

_r 1 — 1 — ^

19 22 I 4

TIME, hOO

Figs. 8, 9.-8. Changes in web diameter (means and 95% C. I. of all species combined) throughout the night
on the (top graph) experimental plot (spiders removed through the night) in November, 1988. Numbers of
spiders are above each time period; 9. Changes in web diameter (means and 95% C. I.) on the (bottom graph)
control plot (unmanipulated), as in Fig. 8.

density (mean ± SE vegetation densities in three
plots: 6.0 ±1.52, 9.4 ±1.65, 10.4 ±1.5). We
found no significant differences in species distri-
bution among the plots (Kruskal- Wallis, F >
0.05).

In Spring, species composition (Table 3)
changed from April to May 1989. Singa lucina
and N. suspicax were dominant in both censuses.
In April, however, L. chloris was absent and
Neoscona subfusca (C. L. Koch) was the third

WARD & LUBIN-TEMPORAL AND SPATIAL SEGREGATION OF SPIDER ACTIVITY

83

Table 4.— Second removal experiment: Total num-
bers of spiders on control and removal quadrats on
days 1 and 2. Numbers of Y. lucina removed are shown
in parentheses. El, E2 = evening censuses of days 1
and 2, respectively; Ml, M2 = morning censuses of
days 1 and 2, respectively.

Day 1

Day 2

Quadrats

El

MI

E2

M2

Control

38

61

55

53

Removal

34 (26)

52 (50)

34

21

most abundant species. This was reversed in May,
when N. subfusca was rare (one spider) and L.
chloris was third in abundance.

The spatial distribution of webs also changed
from April to May (Fig. 7). In April, webs of N.
suspicax were significantly higher in the hedge
than one month later (ANOVA, P < 0.001) due
to an influx of small individuals in the later cen-
suses. In May, webs of L. chloris were signifi-
cantly higher in the hedge than either S. lucina
or N. suspicax (ANOVA, P < 0.05), and N. sus-
picax was significantly closer to the exposed edge
of the vegetation (facing away from the ponds)
than the other species (ANOVA, P < 0.05).

Removal experiments. — removal: In
the Autumn of 1988, we removed spiders as they
initiated web-building, expecting that if space was
the factor preventing simultaneous activity of
spiders, the larger spiders that are usually active
later in the night should initiate web-building
earlier. However, there was no significant differ-
ence in the web diameter of spiders at each time
period censused in the control and experimental
plots (Figs. 8, 9), indicating that larger spiders
did not take advantage of the space available to
put up webs earlier in the night.

Second removal: If space for web building is
limiting, we expected that the removal of active
spiders would open up new spaces and that new
individuals would move in to occupy them. To
test this, in Spring 1989, we divided an 18 m
section of hedge into six contiguous 3 m^ quad-
rats and removed all the active S. lucina (the
dominant species) from the two central quadrats
during one evening census and the following
morning.

New spiders became active in the removal
quadrats on both the evening and morning fol-
lowing the removals (Table 4). The increase in
the number of evening-active spiders in the re-

Table 5.— Third removal experiment: Numbers of spi-
ders observed on quadrat 2 from which all S. lucina
were removed on the second evening (E2); quadrats 3
and 4 combined, from which spiders were removed on
the first morning (M 1 ); and quadrats 1 and 6 combined,
which were unmanipulated (C). Shown are the num-
bers of S. lucina only, as other species occurred only
in small numbers. The numbers of spiders removed
are underlined.

Quad- .
rat

Day 1

Day 2

Day 3

El

Ml

E2

M2

E3

M3

E

23

25

27

20

24

22

M

41

76

22

21

—

—

C

33

46

25

42

32

37

moval quadrats was not significantly different
from that in the control quadrats. In the morning,
however, the control quadrats exhibited a decline
in numbers (from 61 to 53), whereas 21 new
spiders became active in the removal quadrats
(x^ = 7.44, P < 0.01). Thus, the increase in the
number of morning-active spiders on the re-
moval quadrats may be only partly explained by
movement of spiders from the control quadrats.

Third removal: There were more spiders ac-
tive in the morning than in the evening in Spring,
and morning-active spiders had significantly
larger webs and more eggs than evening-active
ones (see above). Therefore, we hypothesized that
morning was the preferred period of activity, but
that space limitation for web-building forced
some spiders to be active in the evening. Using
spiders whose activity period had been deter-
mined during the previous two days of obser-
vation, we tested this possibility by removing all
“morning only” spiders from two 3 m quadrats,
expecting that “evening only” spiders would be-
come active in the morning. We also removed
“evening only” spiders from another 3 m quad-
rat, expecting no change in the time of activity
of “morning only” spiders.

As in the second removal experiment, remov-
ing evening spiders had little effect on the activity
of spiders either the following morning or the
following evening (Table 5). On the third eve-
ning, new spiders replaced those that had been
removed on the second evening.

Removing morning-active spiders caused a
significant decrease in activity on the following
morning in comparison with the control plots (x^
test, P < 0.05). Activity was reduced also on the
evening following removal of morning-active

84

THE JOURNAL OF ARACHNOLOGY

Fig. 10. — Changes in web height of spiders following
space supplementation in October 1989. Shown are
means and 95% C. I. of web height in control (C), low-
string availability (L) and high-string availability (H)
plots. Numbers of spiders (all census days combined)
are shown above each plot.

spiders. There were no significant changes in the
numbers of spiders active in the morning and
evening on the control plots over the three days
of the experiment (x' tests, P > 0.05). Thus, the
appearance of 2 1 new webs on the morning fol-
lowing the removal of morning-active spiders
cannot be explained by movement of spiders from
the control plots into the removal plots.

In all three of the removal experiments, there
was no change in the web characters measured
(web diameter and the number of spiral threads
per cm ') with reduced population density
(ANOVA, P > 0.05), suggesting that web ge-
ometry is not sensitive to short-term changes in
spider density.

Space supplementation experiments. — In Au-
tumn 1988, we found a significant positive cor-
relation = 0.94, P < 0.05) between the num-
ber of active spiders and the number of spiders
attaching their webs to the string delineating the
plots. This suggested that space for building webs
was limited. To test this idea, we provided ad-
ditional web supports, using string to subdivide
two plots into squares of 50 cm^ (=low string

availability) and 25 cm^ (=high string availabil-
ity) respectively.

The number of spiders active did not increase
significantly in the high string-availability plot
on the night immediately after supplementation,
but increased four-fold on the second night (from
28-119 spiders; x" = 30.90, P < 0.001). Web
height decreased significantly as more supports
for web-building became available closer to the
ground (ANOVA, P < 0.05; Fig. 10). There was
no significant change in the number of spiders
active in the low-string availability plot after sup-
plementation (x^ tests, P > 0.05), nor in the con-
trol plot. Had these changes in activity in the
high-string availability plot been due to in-
creased disturbance, we would expect a reduction
rather than an increase in activity. Thus, we as-
cribe this change to extra supports provided for
web-building.

DISCUSSION

Temporal segregation of activity.— We have
shown that the spiders use stratified activity pe-
riods in Autumn to partition a homogeneous
habitat. In a review of resource partitioning in
animal communities, Schoener (1986) observed
that resource partitioning most commonly oc-
curs by division of habitat use, then of food and
only rarely, time. He suggests that theoretically
there is no advantage to temporal specialization
because no energetic gain can be derived from
not feeding during most time periods (for em-
pirical evidence, see Jaksic 1982). Temporal spe-
cialization should occur only if the risk of pre-
dation is large relative to the need for energy,
and then all species may specialize on the same
time period (Schoener 1986).

For the Ma’agan Michael orb-weavers, small
spiders that are active early in the evening can
prey on the large numbers of small Nematocera
emerging from ponds at dusk, while spiders for-
aging later in the evening have small numbers of
large prey available. Why are large spiders not
active earlier? Three possible explanations are:
( 1 ) Large webs are inefficient for trapping small
prey; (2) Webs are damaged or clogged by many
small insects, producing low rewards per unit
effort of silk production; (3) There is a greater
predation risk early in the evening than later on.

We favor the last explanation because early
activity would at least yield some food and a
damaged web may be renewed. For example, at
least some individuals of S. lucina in Spring re-

WARD & LUBIN-TEMPORAL AND SPATIAL SEGREGATION OF SPIDER ACTIVITY

85

newed their webs for both morning and evening
activity. Thus, prey arguments ( 1 and 2) do not
apply. However, to take advantage of dusk insect
emergences, webs must be built in the light. Large
spiders in large webs are more vulnerable than
smaller ones because they tend to be higher in
the vegetation and, in the case of N. suspicax,
closer to the exposed edge of the vegetation.

In Spring, temporal stratification was not cor-
related with either spider size or web size, al-
though we observed the same pattern of insect
activity as in Autumn, i.e., small insects active
early and larger insects active later in the night.
However, for S. lucina, spiral density was greater
in “morning spiders”, whose activity coincided
with that of larger flying insects, than in “evening
spiders”. This supports Eberhard’s (1986) hy-
pothesis that webs designed to intercept large
prey should have greater spiral density (and
therefore greater resilience) than orbs designed
for small prey. Further investigation is required
to establish whether this difference in spiral den-
sity is a behavioral response to prey size avail-
ability.

In Spring, “morning only” and “morning and
evening” individuals of S. lucina produced more
eggs than did “evening only” individuals. As these
differences were uncorrelated with body size, it
is not clear what prevented some spiders from
becoming active in the early morning. Our sec-
ond and third removal experiments indicate that
there was no major shift of activity from evening
to morning following removal of spiders from
plots. Active spiders did not take advantage of
the extra space provided in either evening or
morning periods, although spiders that were
hitherto inactive became active. Given the ad-
vantage to morning activity, it is puzzling that
some spiders retain their low-reward (evening)
activity periods. We suggest the following expla-
nations that bear further investigation: First, eve-
ning-active spiders may have a more reliable,
albeit lower quality, food resource (Nematocera)
than morning-active spiders. Second, short-term
experiments may allow insufficient time for spi-
ders to adjust their activity pattern to a new sit-
uation.

Spatial distribution. — Clumped distribution,
especially of small or subordinate individuals in
the presence of dominant individuals, may in-
dicate aggression (Pielou 1977). In spite of con-
siderable evidence of clumped distribution of
spiders in this study, we were unable to detect

strong interspecific interactions. There was little
interspecific separation of the species on any axis
examined, although L. chloris was on occasion
significantly higher in the vegetation than the
other species, and S. lucina significantly lower in
the vegetation in Autumn than the other species.
Divergence in web height selection has been found
in other studies of orb-weavers (Enders 1974,
1975; Tolbert 1975; Olive 1980; Brown 1981)
although species populations often switch ver-
tical positions in different studies (Brown 1981).
In our study, switches in web height selection
were also found, notably in N. suspicax (Fig. 7),
perhaps as an effect of a seasonal change in the
mean size of spiders.

Our study indicates that the slight interspecific
differences that occurred were not due to inter-
specific interaction, because there was no change
in the pattern of activity or in species compo-
sition with removal of potential competitors.
Only a single case of overt interspecific aggres-
sion was observed in 24 nights (a N. suspicax
removed the web of a L. chloris). A lack of move-
ment in response to short-term changes in con-
ditions may be typical of species with rolled-leaf
retreats (e.g., S. lucina and N. suspicax, see Table
1) which have relatively fixed websites.

There is little other evidence that interspecific
competition occurs in orb-weavers (Brown 1981;
Wise 1984). Manipulative studies indicate that
yearly differences in weather patterns may affect
population densities and foraging patterns as
much as, if not more than, competition (Wise
1981; Horton & Wise 1983). Thus, only Spiller
(1984) has shown that interspecific competition
occurs between orb-weavers. There too seasonal
reversal of competitive advantage occurred, so
the long-term effects of competition on the two
species studied may be minimal. Other studies
(Spiller & Schoener 1 988, 1 989; Schoener & Toft
1983b) have shown that predators (lizards) may
have more important effects on spider densities
than competition.

Although interspecific competition for living
space does not appear to play an overt role in
the community of orb-weavers at Ma’agan Mi-
chael, intraspecific competition (e.g., “scramble
competition”, see MacArthur 1972) may have
an important effect on activity. Space limitation
restricts the web-building activity of some spi-
ders, although the results of the string supple-
mentation experiments indicate that the spiders
may only react, over short periods, to a large

86

THE JOURNAL OF ARACHNOLOGY

increase in space availability. Other field exper-
iments have revealed significant intraspecific
competition between spiders (Coleboum 1974;
Wise 1975; Schaefer 1978; Riechert 1981), al-
though certainly not in all species studied (Wise
1981, 1983; Horton & Wise 1983).

An interesting possibility is that a “lottery for
living space” (Sale 1977) exists in this spider
community. Spiller (1984) has also suggested that
a lottery model best explained the seasonal re-
versal of competitive advantage between two
species of orb-weaving spiders. Chesson & War-
ner (1981) modelled the lottery system based on
competition for space. When the environment
varies such that each species has times when it
can have strong recruitments, the net effect is to
favor positive growth rates at low density for all
species (Chesson 1986).

On the hedges at Ma’agan Michael, S. lucina
and N. suspicax fluctuate in abundance and ex-
hibit a seasonal reversal in dominance. These
fluctuations are apparently associated with dif-
ferent reproductive periods of the species (May
for S. lucina, September-October for N. suspi-
cax). Natural history observations at Ma’agan
Michael indicate that there is considerable pre-
dation on spiders by leaf-gleaning warblers
(mostly Phylloscopus spp., Sylviidae), particu-
larly during the Spring and Autumn bird migra-
tions. We suggest that predation may influence
orb-weaver community structure in these hedges
as follows: The first spiders, regardless of species,
to become active after recruitment may establish
control over empty sites. By being the first spi-
ders to establish, they may grow quicker and
reach large sizes earlier. As observed intraspe-
cifically in S. lucina, those spiders using the pre-
ferred activity periods have the largest clutches.
However, predators will encounter more spiders
of the common species (especially because spider
species are mostly distributed at random within
the hedges) and this should result in greater pre-
dation on the dominant species. Vacant areas in
the hedge may be taken over by new recruits or
by previously inactive spiders. Therefore, differ-
ences in the time of recruitment will cause dif-
ferent species of spider to be dominant at differ-
ent times of year, while predation and/or other
environmental effects (e.g., storms) will facilitate
coexistence.

ACKNOWLEDGMENTS

We thank all the Ben Gurion University stu-
dents of the Field Ecology course who partici-

pated in the Ma’agan Michael experiments; R.
Gil, Y. Shak and the staff of the Hof HaCarmel
Field Study School (Ma’agan Michael) for their
help and permission to work in the Nature Re-
serve; and G. Levy (Hebrew University of Je-
rusalem) for identification of the spiders. This is
publication no. 143 of the Mitrani Centre for
Desert Ecology.

LITERATURE CITED

Andrewartha, H. G. & L. C. Birch. 1954. The dis-
tribution and abundance of animals. Univ. Chicago
Press, Chicago.

Brown, K. E. 1981. Foraging ecology and niche par-
titioning in orb-weaving spiders. Oecologia (Berlin),
50:380-385.

Chesson, P. L. 1986. Environmental variation and
the coexistence of species. Pp. 240-256, In Com-
munity ecology (J. L. Diamond & T. J. Case, eds.).
Harper & Row, New York.

Chesson, P. L. & R. R. Warner. 1981. Environmental
variability promotes coexistence in lottery compet-
itive systems. American Nat., 1 17:923-943.

Coleboum, P. H. 1974. The influence of habitat
stmcture on the distribution of Araneus diadematus
Clerck. J. Anim. Ecol., 43:401-409.

Connell, J. H. 1978. Diversity in tropical rain forests
and coral reefs. Science, 199:1302-1309.

Davies, N. B. 1978. Ecological questions about ter-
ritorial behaviour. Pp. 315-350, In Behavioural
ecology: an evolutionary approach (J. R. Krebs &
N. B. Davies, eds.) Blackwell Scientific Publications,
Oxford.

Eberhard, W. G. 1986. Effects of orb-web geometry
on prey interception and retention. Pp. 70-100, In
Spiders: webs, behavior and evolution (W. A. Shear,
ed.). Stanford University Press, Stanford.

Eberhard, W. G. 1991. Function and phylogeny of
spider webs. Ann. Rev. Ecol. Syst., 21:341-372.

Enders, F. 1973. Selection of habitat by the spider
Argiope aurantia Lucas (Araneidae). American Midi.
Natur., 90:47-55.

Enders, F. 1974. Vertical stratification in orb-web
spiders (Araneidae, Araneae) and a consideration of
other methods of coexistence. Ecology, 55:317-328.

Enders, F. 1975. The influence ofhunting manner on
prey size, particularly in spiders with long attack
distances. American Nat., 109:737-763.

Green, R. H. 1971. A multivariate statistical ap-
proach to the Hutchinsonian niche: bivalve mol-
luscs of centra! Canada. Ecology, 52:543-556.

Hallet, J. G. & S. L. Pimm. 1979. Direct estimation
of competition. American Nat., 1 13:593-599.

Harwood, R.H. 1974. Predatory behavior of
aurantia Lucas. Amer. Midi. Natur., 91:130-139.

Hoffmaster, D. K. 1985. Community composition
and nearest neighbour relationships in orb-weaving

WARD & LUBIN-TEMPORAL AND SPATIAL SEGREGATION OF SPIDER ACTIVITY

87

spiders: the product of aggression? Behav. Ecol. So-
ciobioL, 16:349-353.

Horton, C. C. & D. H. Wise. 1 983. The experimental
analysis of competition between two syntopic spe-
cies of orb-web spiders (Araneae: Araneidae). Ecol-
ogy, 64:929-944.

Jaksic, F. M. 1982. Inadequacy of activity time as a
niche difference: the case of diurnal and nocturnal
raptors. Oecologia (Berlin), 52:171-175.

LeBorgne, R. & A. Pasquet. 1987. Influences of ag-
gregative behaviour on space occupation in the spi-
der Zygiella x-notata (Clerck). Behav. Ecol. Socio-
biol., 20:203-208.

Lubin, Y. D. 1974. Adaptive advantages and the evo-
lution of colony formation in Cyrtophora (Araneae:
Araneidae). Zool. J. Linn. Soc., 54:321-339.

Lubin, Y.D. 1978. Seasonal abundance and diversity
of web-building spiders in relation to habitat struc-
ture on Barro Colorado island, Panama Canal Zone.
J. Arachnol., 6:31-51.

MacArthur, R. H. 1972. Geographical ecology. Harp-
er & Row, New York.

Menge, B. A. & J. P. Sutherland. 1976. Species di-
versity gradients: synthesis of the roles of predation,
competition and temporal heterogeneity. American.
Nat., 110:351-369.

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

Pasquet, A. 1984. Repartition de deux especes d’Ar-
aignees orbiteles, Amneus marmoreus (Clerck) et
Araneus diadematus (Clerck) dans une prairie en
friches. Biol. Behav., 9:321-331.

Pielou, E. C. 1977. Mathematical ecology. John Wi-
ley & Sons, New York.

Riechert, S. E. 1981. The consequences of being ter-
ritorial in spiders: a case study. American Nat., 117:
871-892.

Riechert, S. E. & A. B. Cady. 1983. Patterns of re-
source use and tests for competitive release in a
spider community. Ecology, 64:899-913.

Robinson, M. H., Y. D. Lubin & B. Robinson. 1974.
Phenology and natural history of web-building spi-
ders on three transects at Wau, New Guinea. Pacific
Insects, 16:1 18-162.

Sale, P. F. 1977. Maintenance of high diversity in
coral reef fish communities. American Nat., Ill:

337-359.

Schaefer, M. 1978. Some experiments on the regu-
lation of population density in the spider Floronia
bucculenta (Araneidae: Linyphiidae). Symp. Zool.
Soc. London, 42:203-210.

Schoener, T. W. 1 986. Resource partitioning. Pp. 9 1-

1 26, In Community ecology: pattern and process (J.
Kikkawa & D. J. Anderson, eds.). Blackwell Sci-
entific Publications, Melbourne.

Schoener, T. W. & C. A. Toft. 1983a. Dispersion of
a small-island population of the spider Metepeira
datona (Araneae: Araneidae) in relation to web-site
availability. Behav. Ecol. Sociobiol., 12:121-128.

Schoener, T. W. & C. A. Toft. 1983b. Spider popu-
lations: extraordinarily high densities on islands
without top predators. Science, 210:1041-1043.

Spiller, D. A. 1984. Seasonal reversal of competitive
advantage between two spider species. Oecologia
(Berlin), 64:322-331.

Spiller, D. A. & T. W. Schoener. 1988. An experi-
mental study of the effect of lizards on web-spider
communities. Ecol. Monogr., 58:57-77.

Spiller, D. A. & T. W. Schoener. 1989. Effect of a
major predator on grouping of an orb-weaving spi-
der. J. Anim. Ecol., 58:509-523.

Tolbert, W. W. 1975. Predator avoidance and be-
haviors and web defensive structures in the orb-
weavers Argiope aurantia Lucas and Argiope trifas-
ciata Forskai (Araneae: Araneidae). Psyche, 82:
29-52.

Uetz, G. W., A. D. Johnson & D. W. Schemske. 1978.
Web placement, web structure, and prey capture in
orb-weaving spiders. Bull. British Arachnol. Soc.,
4:141-148.

Uetz, G. W., T. C. Kane & G. E. Stratton. 1982.
Variation in the social group tendency of a com-
munal web-building spider. Science, 217: 547-549.

Wise, D. H. 1975. Food limitation on the spider
Linyphia marginata: experimental field studies.
Ecology, 56:637-646.

Wise, D. H. 1981. Inter- and intraspecific effects of
density manipulations upon females of two orb-
weaving spiders (Araneae: Araneidae). Oecologia
(Berlin), 48:252-256.

Wise, D. H. 1983. Competitive mechanisms in a food-
limited species: relative importance of interference
and exploitative interactions among labyrinth spi-
ders (Araneae: Araneidae). Oecologia (Berlin), 58:

1-9.

Wise, D. H. 1984. The role of competition in spider
communities: insights from field experiments with
a model organism. Pp. 42-53, In Ecological com-
munities: conceptual issues and the evidence (D. R.
Strong, D. S. Simberloff, L. G. Abele & A. Thistle,
eds.). Princeton Univ. Press, Princeton.

Manuscript received August 1991, revised January 1992.

1992. The Journal of Arachnology 20:88-93

HABITAT SEGREGATION BY SPECIES OF METAPHIDIPPUS
(ARANEAE: SALTICIDAE) IN MINNESOTA

Bruce Cutler: Electron Microscope Laboratory and Department of Entomology,
University of Kansas, Lawrence, Kansas 66045-2106 USA

Daniel T. Jennings: North Central Forest Experiment Station, 1992 Folwell Avenue,
St. Paul, Minnesota 55108 USA

ABSTRACT. Four species of Metaphidippm (Araneae: Salticidae) occupied different habitat types in Min-
nesota; M. ahzonensis was found exclusively in sand prairie; M. Jlavipedes was almost completely restricted to
conifers; M. insignis primarily inhabited open, non-canopy vegetation (e.g., grasslands); whereas, M. protervus
occupied most habitats, but most evidently shaded forest understory and wetlands. Reasons for such habitat
partitioning are conjectural. Size differences among the four species probably were not ecologically significant
based on Dyar’s constant; however, competition for prey may have influenced habitat selection.

Metaphidippus is one of the largest genera of
jumping spiders in North America (Richman &
Cutler 1978). Revision of the genus by other
workers will probably redefine the taxonomy and
introduce new generic names; however, the spe-
cies discussed here will remain in one genus.

When we first started collecting jumping spi-
ders in Minnesota, it quickly became evident that
different species were found only in specific hab-
itats. This was particularly noticeable in species
of Metaphidippus because our favorite collecting
methods— sweep netting and beating vegeta-
tion-garnered large numbers of these vegeta-
tion-inhabiting spiders. Over a 25 year period,
most parts of the state were visited and habitat
data recorded whenever salticids were collected.
The data was analyzed and a hypothesis for hab-
itat segregation (Dyar’s constant) considered.

METHODS

Specimens of Metaphidippus arizonensis
(Peckham & Peckham), M. Jlavipedes (Peckham
& Peckham), M. insignis (Banks), and M. pro-
tervus (Walckenaer) were collected predomi-
nantly by sweep netting; however, beating foliage
also yielded a few specimens. Collection heights
were not controlled or recorded. If necessary,
laboratory rearing was done in the case of an-
tepenultimate and penultimate instars to confirm
identifications based on adult genitalia. Spiders
were kept at ambient temperatures in Petri dish-
es with moist pieces of sponge, and fed Dro-
sophila adults and Tribolium larvae until mature.

To avoid sampling bias, individual sites were
counted only once even if repeatedly collected.
A site was considered a stand of vegetation iso-
lated from another stand by an intervening stand
of different vegetation, or by a large physical ob-
stacle. In many cases, collected sites were sepa-
rated by many kilometers; others were adjacent
and differed only in vegetation. All sites were in
Minnesota. Collecting dates were from April-
October.

Negative catches were not recorded; tabulated
data consisted only of samples that yielded spec-
imens. Carapace widths between row III eyes
were measured with an ocular micrometer for 50
mature females of M. Jlavipedes, M. insignis, and
M. protervus, and for 35 mature females of M.
arizonensis. Tukey’s Studentized Range (HSD)
Test (SAS 1985) was used for comparisons of
carapace widths among species at P < 0.05.

RESULTS

Figure 1 shows the Minnesota counties col-
lected and the species of Metaphidippus found.
Table 1 compares species presence/absence with-
in the different habitats. With only one excep-
tion, because of small sampling size (deciduous-
tree foliage), species were unequally distributed
within each habitat investigated (Table 1). We
conclude that 1) specific habitats support few (1-
3) species of Metaphidippus, and 2) species pres-
ence within a habitat usually is dominated by a
single species, less frequently by two species.

Habitat breadth or specificity (i.e., the number

88

CUTLER & JENNINGS -HABITAT SEGREGATION OF METAPHIDIPPUS

89

Table 1.— A comparison of species occurrence within and over all habitats studied in Minnesota, 1964 to
1989.

Habitat

Sum of individual collections/habitat by Metaphidippus species

arizonensis

flavipedes

insignis

protervus

Conifer foliage

0

41

0

0

Deciduous-tree foliage

0

1

0

3

Coniferous-tree understory

0

2

0

7

Deciduous understory

0

0

0

31

Wetland

0

0

1

10

Old field

0

0

1

6

Mixed meadow

0

0

13

10

Mesic prairie

0

1

19

1

Sand prairie

7

0

1

2

Crops

0

0

1

4

All

7

45

36

74

of habitats occupied by each species) also varied
considerably among the four species of Meia-
phidippus (Table 2). For example, M. arizonensis
was found only in sand prairie, whereas M. pro-
tervus was found in 9 of the 10 habitats inves-
tigated. However, these results must be inter-
preted with caution because sampling intensity
varied among habitats.

Dice-Lerra diagrams of carapace width mea-
surements are given in Figure 2 for the four spe-
cies of Metaphidippus. Means for all species pairs
were significantly different {P < 0.05), except the
pair M. insignis - M. protervus.

DISCUSSION

Species partitioning by habitat is a well known
phenomenon among many groups of animals
(Schoener 1974). Good examples exist for the
predominantly ground-dwelling lycosid spiders
in the genus Pardosa (Den Hollander & Lof 1974;
Greenstone 1980; Hallander 1970; Lowrie 1973;
Vlijm & Kessler-Geschiere 1967; Vogel 1972).
However, habitat partitioning among non-snare
building, vegetation-inhabiting spiders has been
little investigated. A few papers discuss the hab-
itat preferences of individual species (Jennings
1976, and papers cited therein; Jennings & Col-
lins 1987b), but rarely in the context of coexisting
phylogenetically related species. Turner & Polis
(1979) considered the members of a raptorial,
non-snare building guild of spiders on inflores-
cences of a coastal sagebrush community in Cal-
ifornia. Included were three species of the crab
spider genus Misumenops. Each species over-
lapped in occurrence on the shrubs, but the two

common Misumenops species were most fre-
quently found on different shrub species. Turner
& Polis (1979) concluded that it was unlikely that
widespread competition for food and space re-
sources occurred among guild members. Inter-
ference competition, i.e., interspecific predation
by guild members, was evoked as the determi-
nator of guild structure (Turner & Polis 1979).

Within Metaphidippus, species determina-
tions can be difficult. Misidentifications are pos-
sible, indeed probable. For example, it is likely
that M. exiguus (Banks) found on jackpine (Pinus
banksiana Lamb.) in Manitoba (Bradley & Hinks
1968) are M. flavipedes (Peckham & Peckham).
Nevertheless, clearly there are indicated habitat
preferences for Metaphidippus species in the lit-
erature (Allen et al. 1970; Berry 1970; Dondale
etal. 1979; Ives 1967; Jennings & Collins 1987a,
b; Legner & Oatman 1964; Lowrie 1968; Mason
&Paul 1 988; Stiettenroth& Homer 1987; Young
& Lockley 1990).

Table 2. —Comparison of Metaphidippus species oc-
currence among 10 habitats in Minnesota, 1964 to 1989.

Species

No. of

habitats
species
found in

Sum of
collections
yielding
species

M. arizonensis

1

7

M. flavipedes

4

45

M. insignis

6

36

M. protervus

9

74

All species

10

162

90

THE JOURNAL OF ARACHNOLOGY

Figure 1.— Localities of Metaphidippus species collected in Minnesota. M. arizonensis = ★, M. flavipedes =
□, M. insignis = O, M. protervus = A.

During our study, special efforts were made to
collect spiders on tamarack, Larix laricina (Du
Roi) K. Koch, because it is the only deciduous
conifer in Minnesota. Despite these efforts no
species of Metaphidippus was found, although
another jumping spider, Eris militaris (Hentz),
did occur. Interestingly, Ives ( 1 968) reported both
E. militaris and M. protervus from tamarack in
Manitoba; however, M. flavipedes is the expected
conifer-inhabiting Metaphidippus in Manitoba,

as reported by Bradley & Hinks (1968). In Min-
nesota, M. flavipedes was collected on all species
of conifers sampled except for tamarack and
northern white-cedar {Thuja occidentalis L.);
however, the latter was scarcely sampled. Strat-
ton et al. (1979) also sampled northern white-
cedar in Minnesota and found several genera of
salticids, but species of Metaphidippus wtre iden-
tified only to genus. In his investigation of spiders
on a small island in northern Lake Michigan,

Width (mm)

CUTLER & JENNINGS -HABITAT SEGREGATION OF METAPHIDIPPUS

91

15

1.45

1.4

1.35

1.3

1.25

1.2

1.15

1.1

1.05

1.0

0.95

0.9

0.85

0.8

A F I P

Metaphidippus species

Figure 2.— Dice-Lerra Diagram for distance between row III eyes in four Metaphidippus species. (A = Me-
taphidippus arizonensis, F = M. flavipedes, I = M. insignis, P - M. pwtervus).

Drew (1967) carefully collected from different
vegetation types including trees. M. flavipedes
was among the commonest species collected on
Juniperus communis (reported as J. depressus)

and on northern white-cedar, whereas M. pro-
tervus was commonest in the herb-shrub stratum
of the upland hardwood forest. Both species of
Metaphidippus occurred at lower frequencies in

92

THE JOURNAL OF ARACHNOLOGY

the old field community and in other commu-
nities (marshes, beach).

That small salticid species should partition by
type of space occupied, rather than successive
temporal occurrence, was predicted by Enders
(1975) based on previous habitat-sampling stud-
ies. The Metaphidippus species we investigated
had similar temporal occurrences of adults, i.e.,
many mature males and rare mature females in
September and October. Both sexes of all four
species are mature in May and June, with mature
females persisting into August. However, we did
not closely measure temporal succession at any
one site where two or more species were found.
Nevertheless, our data lends support to Enders’
hypothesis that species segregate by habitat.

One possible reason for habitat segregation by
Metaphidippus species is competition for similar
sized prey. However, with general collections such
as ours, the morphological information of the
specimens themselves is often the only data that
can be analyzed. Prosomal size differences were
statistically significant among all but one of the
six species-pair combinations, but it may not be
ecologically significant. In the laboratory, Homer
& Starks (1972) found that the average percent-
age difference of prosomal length between molts
of Metaphidippus galathea (Walckenaer) was 1 8%
(Dyar’s constant). Dyar’s constant has been
evoked as a means of determining the minimum
difference in ecological isolation for prey size
among different instars of a spider species (En-
ders 1976). The same explanation should ac-
count for size differences among closely related
species. The greatest percentage difference among
prosomal measurements in the species pairs dis-
cussed here was less than 13% (M. arizonensis
vs. M. protervus). Assuming that the average per-
cent difference (18%) between instars of M. gal-
athea also applies to species of Metaphidippus
found in Minnesota, then prosomal size differ-
ences among species apparently were not signif-
icant in determining ecological isolation. We
conclude that these species show a potential of
competing for similar sized prey based on the
absence of appreciable size differences.

This paper demonstrates that species of Me-
taphidippus occupy different habitats in Minne-
sota. Size differences among the Metaphidippus
species apparently are not great enough to pre-
vent competition for similar sized prey. Other
approaches, including experimental studies,
should provide some answers as to how habitat
separation is maintained.

ACKNOWLEDGMENTS

We thank those who have helped this study in
various capacities. Robert Dana, Minnesota; and
Ronald L. Huber, Kansas assisted with field col-
lections in Minnesota. Permission to collect on
properties under their supervision was gener-
ously provided by The Nature Conservancy-
Minnesota, and by Cedar Creek Natural History
Area-University of Minnesota. Portions of this
research were completed during the junior au-
thor’s tenure with the USDA, Forest Service,
Northeastern Forest Experiment Station, 180
Canfield Street, Morgantown, West Virginia. Dr.
Andrew Liebhold, USDA Forest Service, Mor-
gantown, West Virginia, assisted with data anal-
yses.

LITERATURE CITED

Allen, D. C., F. B. Knight & J. L. Foltz. 1970. In-
vertebrate predators of the jack-pine budworm,
Choristoneura pinus, in Michigan. Ann. Entomol.
Soc. America, 63:59-64.

Berry, J. W. 1970. Spiders of the North Carolina
piedmont old-field communities. J. Elisha Mitchell
Sci. Soc., 86:97-105.

Bradley, G. A. & J. D. Hinks. 1968. Ants, aphids,
andjackpine in Manitoba. Canadian Entomol., 100:
40-50.

Den Hollander, J. & H. Lof. 1972. Differential use
of the habitat by Pardosa pullata (Clerck) and Par-
dosa prativaga (L. Koch) in a mixed population.
Tijdschrift Entomol., 1 15:205-215.

Dondale, C. D., B. Parent & D. Pitre. 1979. A 6 -year
study of spiders (Araneae) in a Quebec apple or-
chard. Canadian Entomol., 1 1 1:377-380.

Drew, L. C. 1967. Spiders of Beaver Island, Michi-
gan. Michigan State Univ., Publications of the Mu-
seum, Biological Series 3:152-208.

Enders, F. 1975. The influence ofhunting manner on
prey size, particularly in spiders with long attack
distances (Araneidae, Linyphiidae, and Salticidae).
American Nat., 109:737-763.

Enders, F. 1976. Size, food-finding, and Dyar’s Con-
stant. Environ. Entomol., 5:1-10.

Greenstone, M. H. 1980. Contiguous allotopy of Par-
dosa ramulosa and Pardosa tuoba (Araneae: Lycos-
idae) in the San Francisco Bay Region, and its im-
plications for patterns of resource partitioning in the
genus. American Mid. Nat., 104:305-31 1.
Hallander, M. 1970. Prey, cannibalism, and micro-
habitat selection in the wolf spiders Pardosa chelata
O. F. Muller and P. pullata (Clerck). Oikos, 21:337-
340.

Homer, N. & K. J. Starks. 1972. Bionomics of the
jumping spider Metaphidippus galathea. Annals En-
tomol. Soc. America, 65:602-607.

Ives, W. G. H. 1967. Relations between invertebrate

CUTLER & JENNINGS- HABITAT SEGREGATION OF METAPHIDIPPUS

93

predators and prey associated with larch sawfly eggs
and larvae on tamarack. Canadian EntomoL, 99:
607-622.

Jennings, D. T. 1976. Spiders on black walnut. Amer-
ican Mid. Nat., 95:1 1 1-119.

Jennings, D. T. & J. A. Collins. 1987a. Spiders on
red spruce conifer foliage in northern Maine. J. Ar-
achnoL, 14:303-314.

Jennings, D. T. & J. A. Collins. 1987b. Coniferous-
habitat associations of spiders (Araneae) on red
spruce foliage. J. ArachnoL, 14:315-326.

Legner, E. F. & E. R. Oatman. 1964. Spiders on apple
in Wisconsin and their abundance in a natural and
two artificial environments. Canadian EntomoL 96:
1202-1207.

Lowrie, D. C. 1968. The spiders of the herbaceous
stratum of the Jackson Hole region V/yoming.
Northwest Sci., 42:89-100.

Lowrie, D. C. 1973. The microhabitats of western
wolf spiders in the genus Pardosa. EntomoL News,
84:103-116.

Mason, R. R. & H. G. Paul. 1988. Predation on larvae
of douglas-fir tussock moth, Orgyia pseudotsugata
(Lepidoptera: Lymantriidae), by Metapkidippus
aeneolus (Araneae: Salticidae). Pan-Pacific Ento-
moL, 64:258-260.

Richman, D. & B. Cutler. 1978. A list of the jumping
spiders (Araneae: Salticidae) of the United States
and Canada. Peckhamia 1:82-109.

Robinson, J. V. 1981. The effect of architectural vari-
ation in habitat on a spider community: experi-
mental field study. Ecology, 62:73-80.

SAS Institute. 1985. SAS User’s Guide: Statistics,
Version 5 Edition. SAS Institute, Cary, North Car-
olina.

Schoener, T. W. 1974. Resource partitioning in eco-
logical communities. Science, 185:27-39.

Stiettenroth, C. L. & N. V. Homer. 1987. The jump-
ing spiders (Araneae: Salticidae) of the Virginia Pen-
insula. EntomoL News, 98:235-245.

Stratton, G. E., G. W. Uetz & D. G. Dillery. 1979.
A comparison of the spiders of three coniferous tree
species. J. Arachnol., 6:219-226.

Turner, M. & G. A. Polis. 1979. Patterns of coexis-
tence in a guild of raptorial spiders. J. Anim. EcoL,
48:509-520.

Vlijm, L. & A. Kessler-Geschiere. 1967. The phe-
nology and habitat of Pardosa monticola, P. nigri-
ceps, and P. pullata (Araneae, Lycosidae). J. Anim.
EcoL, 36:31-56.

Vogel, B. R. 1972. Apparent niche sharing of two
Pardosa species (Araneida: Lycosidae). Armadillo
Papers 7, 1 3 pp.

Young, O. P. & T. C. Lockley. 1990. Autumnal pop-
ulations of arthropods on aster and goldenrod in the
delta of Mississippi. J. Entomol. Sci., 25:185-195.

Manuscript received February 1992, revised March 1992.

1992. The Journal of Arachnology 20:94-106

DEVELOPMENTAL PLASTICITY AND FECUNDITY
IN THE ORB-WEAVING SPIDER NEPHILA CLAVIPES

Linden E. Higgins: Department of Zoology, University of Texas, Austin, Texas
78712 USA

ABSTRACT. To document variation in several developmental parameters and the elfect of this variation on
female adult size and fecundity, marked individuals were followed in three disjunct populations of the widely
distributed spider Nephila clavipes (Araneae: Tetragnathidae). The sites chosen had very different physical and
biological conditions which were expected to affect the development of the animals. Several developmental
parameters were very plastic, such as weight gain and number of juvenile instars, varying both among and within
populations. In contrast, two important developmental parameters, growth per molt and pre-molt weight, were
constrained within each population but differed between tropical and temperate conditions. Constraining growth
per molt established a developmental trajectory, and variation of its slope and of the number of juvenile instars
were the primary causes of variation in adult female size and the correlated variation in the fecundity per egg
sac.

RESUMEN. Para explorar la variacion en parametros de ontogenia y la consequencia de esta variacion
para el tamano en hembras maduras, un estudio del campo usando individuos marcados fue hecho en tres
poblaciones disconectadas de la arana Nephila clavipes (Araneae: Tetragnathidae). Los sitios escojidos presen-
taban condiciones fisicas y biologicas muy distintas, los cuales se anticipaban a influir fuertamente en la ontogenia
de los animates. Algunos parametros ontegeneticos se demostraban muy plasticos, mostrando variacion tanto
dentro, como entre poblaciones, mientras que otros parametros fueron menos plasticos. Dos parametros im-
portantes en el cresimiento de las aranas, el cresimiento por muda y el peso antes de mudar, no variaban dentro
de cada poblacion, pero mostraban variacion entre condiciones tropicales y templadas. Inflexibilidad en el
parametro de cresimiento por muda produce una trayectoria ontogenetica. Variacion en el pendiente de tal
trajectoria y en el numero de estadios juveniles son los causas principales de la variacion observado en el tamano
de las hembras maduras.

In a variety of arthropods, variation in adult
size has been correlated with differences in male
competitive ability, voltinism, female fecundity,
and other parameters of fitness (e. g., Lawlor
1976; Harrington 1978; Christenson & Goist
1979; Eberhard 1982; Morse & Fritz 1987; At-
kinson & Begon 1987). In insects, correlations
of variation in size with environmental and bi-
ological factors have been utilized to develop
models describing the evolution of arthropod life
histories (Tauber & Tauber 1978; Masaki 1978;
Mousseau & Rolf 1989). Changes in size and
voltinism are often correlated with latitude and
altitude (Masaki 1978; Mousseau & Roff 1989;
Dingle, Mousseau & Scott 1990). However, few
studies have investigated the proximal cause of
variation in adult size: variation in juvenile de-
velopment (c/. Hugueny & Louveaux 1986).

In most insects and spiders, growth is deter-
minate and the age and size at maturity are gov-
erned by the number of instars, intermolt du-

ration, and growth at each molt. Variation in
juvenile development, due either to genetic dif-
ferences (Newman 1 988a; Mousseau & Roff 1 989)
or to environmentally induced plasticity
(Schmalhausen 1949; Steams 1983; Steams &
Koella 1 986; Newman 1 988b), will result in vari-
ation in adult size and age at maturity. Size and
age at first reproduction are correlated with fe-
male fitness in many invertebrates and ecto-
thermic vertebrates, changing female fecundity
and the probability of death prior to reproduc-
tion. First, female fecundity increases with in-
creased body size (Turnbull 1962; Toft 1976;
Harrington 1978; Seigel & Fitch 1984; Palmer
1985; Fritz & Morse 1985; Miyashita 1986;
McLay & Hayward 1987; Ford & Seigel 1989)
and decreases with increased age at maturity. In
an iteroperous annual organism, an early ma-
turing female will have more opportunities to
reproduce (Toft 1976, Suter 1990). Second, ju-
veniles of many organisms are more at risk from

94

HIGGINS-LIFE-HISTORY PLASTICITY AND FECUNDITY

95

Table L— Characteristics of sites used, including generations per year (facultatively bivoltine in Veracruz),
seasonality and annual rainfall, mean and standard deviation of prey capture rates (per 1 2 diurnal hours per
spider, superscripts refer to significantly distinct groups), and relative predation rates on juveniles less than 0.5
cm t + p (Higgins in press). Climate data sources: Panama: Leigh et al. 1982; Mexico: Garcia 1973; Texas: US
Meteorological Service. Prey-capture rates from Higgins and Buskirk, in press.

Site

Study period

Voltin-

ism

Seasonality

Rain-

fall

Prey capture

Predation

Panama

1/1985-771986

2

wet/dry

2.5 m

wet: 2 (9.75)“

high

dry: 1 (i.l2)'>

high

Veracruz

7=11/1986, 5/1987

1

warm/cold

4.5 m

2 (2.71)“

high

Texas

7-8/1985, 5/1988

1

warm/cold

1.1 m

1 (0.83)*’

low

predation than adults; therefore, the probability
of death prior to reproduction increases with the
duration of the juvenile stages (Bervin & Gill
1983; Steams 1983; fitter 1989; Higgins in press).

The orb-weaving spider Nephila davipes (Lin-
naeus) (Araneae: Tetragnathidae) is a broadly
distributed organism (Levi 1980) with striking
variation in adult female size. I anticipated that
variation in female size among populations ex-
isting under different conditions resulted from
variation in one or several juvenile develop-
mental parameters. Laboratory studies involving
males of this spider have revealed food-depen-
dent variation in both grov/th per instar and the
number of instars before sexual maturity, con-
tributing to variation in size at sexual maturity
(Vollrath 1983). To investigate the developmen-
tal causes of variation in female size and the
environmental correlates of the developmental
variation, I undertook field studies of marked
individuals in three disjunct populations of N.
davipes: Banro Colorado Island, Panama, coastal
Veracruz, Mexico and southeastern Texas, USA.
These data were combined with data concerning
female fecundity in the tv/o tropical populations
to document the fitness consequences of varia-
tion in adult female size.

Study Otgmasm.— Nephila davipes is an orb-
weaving spider found from the southeastern
United States to northern Argentina (Levi 1980).
The spiders have one post-hatching non-feeding
larval stage {sensu Foelix 1982) and molt to the
first instar before emerging from the egg sac. In
univoltine populations, a quiescent stage occurs
in the egg sac. The spiderlings spend one or two
instars together on a tangle web then disperse
(Kimmel & Grant 1980; Hill & Christenson
1981). This species is highly sexually dimorphic
(Levi 1 980). Mature males vary from 0.4=0. 7 cm
leg I tibia + patella length. Females reach sexual

maturity at a range of sizes, 0. 8=2.0 cm leg I tibia
+ patella length (Higgins pers. obs.). The spiders
lay their egg sacs away from the orb and do not
usually return, but build in a new site one or two
days after laying. The phenology varies among
populations, in part related to different weather
patterns (pers. obs.), but no marked, free-living
mature females have been observed to survive
more than five months (over 200 individuals in
7 sites). Among the three populations studied,
that on Barro Colorado Island, Panama, is bi-
voltine with peak female abundances in the early
rainy season and late-rainy to early dry season
(Lubin 1978; Higgins 1988) and that in Houston,
Texas, is univoltine with peak female abundance
in August-Septero,ber (Higgins 1988). In Los
Tuxtlas, Veracruz, Mexico, the population is fac-
ultatively bivoltine: there is normally only one
generation per year v/ith peak female abundance
in August-September (pers. obs.). Occasionally,
juveniles do not enter winter quiescence and these
individuals mature and reproduce in May.

Study Sites.— At all sites, I studied spiders
found in second growth (Texas, Panama) or pri-
mary (Veracruz) forest and along edges of trails
and abandoned roads. During the study, al! three
sites had maximum daily temperatures of about
27 °C (Garcia 1973; Leigh, Rand& Winsor 1982;
Higgins 1987). Patterns of rainfall, prey capture,
and predation varied among the three popula-
tions and between seasons on Barro Colorado
Island (Table 1). The prey-capture rate varied
with weather, and was significantly higher in Ve-
racruz and the Panama rainy season than in Tex-
as and the Panama dry season (Higgins & Bus-
kirk in press). The frequency of predator attacks
was significantly higher for small juveniles (<0.5
cm leg I tibia + patella length) within each pop-
ulation and higher in the tropical populations
compared to Texas (Higgins in press).

96

THE JOURNAL OF ARACHNOLOGY

Lowland Panama has a seasonal tropical cli-
mate (Leigh, Rand & Winsor 1982), the dry sea-
son normally lasts from January to mid-May.
The southern coast of Veracruz, Mexico, has a
wet tropical climate (de la Cruz & Dirzo 1987).
Although there is no regular dry season, there is
an unpredictable period of winter storms that
combine winds, low temperatures, and rain, be-
ginning between September and December and
lasting two to four months. The spiders were
studied in the eastern section of the biological
station “Los Tuxtlas” (Universidad Nacional
Autonoma de Mexico). Galveston County, Tex-
as has humid summers with little rainfall and
relatively cold winters lasting from November
to March (average minimum temperature 1 1 °C).
The spiders were studied in scrub forest at the
University of Houston Coastal Center (Higgins
1987). The seasonality of the climate in Panama
resulted in replication of dry and rainy condi-
tions between the three populations: Panama dry
and Texas, Panama rainy and Veracruz.

METHODS

The data were collected through repeated ob-
servations of marked individuals. Each spider
was measured (leg I tibia + patella length (t +
p), cm ± 2%, measured with Helios needle-nosed
calipers) without removing the animal from its
web. I individually marked spiders larger than
0.4 cm t + p on their legs (with “Testor’s” flat
enamel (Testor Corporation, Rockford, IL61 108,
USA)) and flagged their web sites. Spiders were
re-marked after molting. Individuals of less than
0.4 cm t 4- p were not marked, but their web
sites were flagged. Spider sex and reproductive
status were categorized as: immature (indeter-
minate sex), penultimate instar male, juvenile
female, mature male or female. Animals larger
than 0.5 cm tibia -I- patella length were assumed
to be juvenile females as males rarely reach that
size without showing secondary sexual charac-
teristics. Mature females have heavily sclerotized
external genitalia, distinguishing them from im-
mature females. During the study period (Table
1 ), I visited each individual regularly (nearly dai-
ly in Veracruz, Texas, and Panama in 1985; ev-
ery other day in Panama in 1986) until it could
no longer be found.

Growth,— Growth was divided into two dis-
tinct but interrelated measures: (a) growth per
molt (change in t -I- p) and intermolt interval
(days between molts), and (b) weight gain per

unit time. Between molts, the spiders gain weight
by expanding the abdomen volume. At the time
of the molt, the leg length and carapace size
change.

Growth per molt was determined through
comparison of pre- and post-molt t -I- p length.
I also measured t + p of discarded exoskeletons
found in the webs of recently molted spiders. The
pre-molt t -I- p length was not different from the
exuvia t + p length for the same individual {n
= 19, paired t test = -0.46, P {2 tailed) = 0.65).
Therefore, I included the length of the exuvia t
+ p in the analysis of growth per molt when pre-
molt t + p was unknown. If an individual was
observed for more than one molt, only data from
the first molt were included in the analysis of
growth per molt. I recorded intermolt intervals
(days between molts) for individuals that were
observed for more than one molt.

Weight gain over 14 day intervals by females
greater than 0.4 cm t -t- p was estimated in the
Panama and Veracruz populations. I chose two-
week intervals to reduce the variance in the rate
of weight gain during the intermolt cycle (Higgins
1988). As these spiders have approximately cy-
lindrical abdomens, I estimated abdomen vol-
ume from abdomen length (cm) and width (cm)
as: abdomen volume (cc) = (length) w (width/2)L
(I found that taking these measures with calipers
was less likely to cause web-site abandonment
than removing the spiders from their webs to be
weighed.) In Panama, I determined that the
weight of a spider could be estimated as a func-
tion of abdomen volume and leg I tibia + patella
length: weight (g) = 0.012 + 0.081 ((t -t- p)^) +
0.784 (abdomen volume); = 0.998 (n = 86,
= 17,701.29, P<0.001).

Reproduction.— To evaluate the effect of fe-
male size on fecundity, I collected first egg sacs
from marked free-living females in Panama and
Veracruz. Females observed molting to sexual
maturity were marked, and only those that were
followed until oviposition were included, there-
by avoiding age affects in reproduction (Suter
1 990). Changes in orb-web renewal behavior sig-
nal when a female is preparing to lay (Higgins
1990). I removed gravid females from the field,
weighed them, and placed them in 30 x 1 5 x
70 cm cages in an insectary in Panama or a non-
airconditioned laboratory in Veracruz, providing
live prey if a viscid spiral was built. Most females
laid eggs within 5 days of being collected, and
were subsequently weighed and released. Three

HIGGINS-LIFE-HISTORY PLASTICITY AND FECUNDITY

97

to five days after being laid, the egg sacs were
opened and the eggs removed, weighed and
counted. Whereas living eggs are yellow in color,
some egg sacs contained a few grey or dried black
eggs that I presumed to be infertile or dead. If
more than 10 eggs were grey or black, they were
counted separately and the yellow eggs were re-
weighed. In these clutches, I estimated mean egg
weight using only the fertile eggs. I calculated
relative clutch mass (RCM) as total egg mass
divided by post-laying weight of the female (Sei-
gel & Fitch 1984). Post-laying rather than pre-
laying weight was used to ease comparison with
the data available for other spiders (McLay &
Hayward 1986).

Statistical Analysis.— Many of the variables
collected are functions of spider size (t -h p) or
weight. After checking the subsets of data from
each site to assure that all had significant re-
gressions, preliminary ANCOVA were run to test
for significant interaction effects between the co-
variate (size or weight) and the factor in question
(site, generation or season). If significant inter-
action effects were found, indicating significant
difference in the slopes of the lines being com-
pared, a regression of the entire data set was done
saving the residuals; and these were analyzed with
ANOVA to test for significant effects of the factor
in question. If the comparison of regression lines
revealed no significant interaction effect (the lines
were parallel), the data were further analyzed with
ANCOVA, dropping the interaction term, to de-
termine if the functions had significantly differ-
ent y - intercepts (Sokal & Rohlf 198 1). All anal-
yses were done with SYSTAT, which uses a least-
squares algorithm for ANOVA and regression
analyses (Wilkinson 1987). Lastly, in biologically
significant cases where the null hypothesis was
not disproven, indicating similarity between
groups, an a posteriori power test was calculated.
This descriptive statistic gives the minimum dif-
ference the test could have detected at R = 0.05,
expressed as a percent of the mean value (N.
Fowler, pers. comm.).

RESULTS

Growth,— Growth was measured as three re-
lated factors: weight gain in 1 4 d intervals, leg I
tibia + patella growth per molt and intermolt
interval. Weight gain was compared between wet-
season Panama and Veracruz; the spiders’ web-
site tenacities were too low during the Panama
dry season to allow observations over two week

intervals. The rate of weight gain varied with size
(t + p) and sexual status (immature or mature
female). In juveniles of both sexes, weight gain
was a log function of the spider size (t + p) at
the beginning of the observation period, with no
difference between sites (ANCOVA: no inter-
action effect; site: Rd.io) = 2.17, R = 0.17; re-
gression: In(Aweight) = - 2.34 + 2.26 (t + p),
R2 = 0.626, F„,n) = 18.37, R = 0.001). Weight
gain by mature females was independent of size
(« = 1 1 , R^ = 0.002, ns) and a Mann-Whitney
f/-test showed no difference between Panama and
Veracruz {df= 1, t/ = 0.1 1, R = 0.84). Mature
females gained weight at a mean rate of 1.622
g/14 d (SD = 0.806).

The development of juveniles was compared
by examining the pre-molt weights and the change
in size (t + p) with each molt. Pre-molting ab-
domen volume (an estimate of weight), pre-molt
size (t + p), and post-molt size were related. The
volume of the abdomen was measured on the
day of the last pre-molt orb, 2-4 days before the
molt (spiders cease building orbs before molting).
Pre-molt abdomen volume was strongly corre-
lated with the pre-molt size (t + p) of the spider
and did not vary between the two tropical sites
(Fig. la. ANCOVA: no interaction effect, mini-
mum detectable difference = 28%. site: R,, ,7, =
0.505, R = 0.49; minimum detectable difference
= 28%. Regression: (abdomen volume )'^^ = 0.07
+ 0.63 (t + p), R2 = 0.89, R„,,8, = 146.0, R <
0.001). The post-molt t + p was highly correlated
with the abdomen volume of the individual on
the day of the last pre-molt orb and the rela-
tionship did not vary between the sites (Fig. lb.
ANCOVA: no interaction effect, minimum de-
tectable difference = 1.9%. site: R,i,i7) = 0.41, R
= 0.53; minimum detectable difference = 90%.
Regression: post-molt t + p = 0.04 + 1.88 (ab-
domen volume)''^- = 0.94, R,, ,8, = 286.2, R
< 0.001).

Intermolt interval in days was highly variable
within sites and positively correlated with the
spider size (Fig. 2). Comparison of Panama dry,
Panama wet, Veracruz and Texas showed no sig-
nificant effect of site or season (ANCOVA: n =
31. site: R(2,27) = 1-20, R = 0.32; season (Panama
dry and Texas vs. Panama wet and Veracruz):
R„,28, = 0.03, R= 0.86).

Growth per molt was compared through the
regression of post-molt size (t + p) on pre-molt
size (Fig. 3). Male and female growth per molt
was compared only for the Panama data set.

98

THE JOURNAL OF ARACHNOLOGY

cube root, pre-molt abdomen volume

Figure 1.— a. The relationship between pre-molt tibia -t- patella length (t -i- p) and pre-molt abdomen volume,
b. The relationship between pre-molt abdomen volume and post-molt t -I- p. (O = Panama wet; □ = Veracruz).

which has the largest number of observed male
molts. There was no difference between male
growth and juvenile and female growth (AN-
COVA: no interaction effect, minimum detect-
able difference = 0.8%. n = 28 males, 101 un-
sexed juveniles and females; sex: F,, ,26) = 0.6,
ns, minimum detectable difference = 24%). Spi-
ders in the three populations exhibited two dif-
ferent rates of growth per molt. The two gener-
ations observed on Panama and the Veracruz

population had the same growth per molt (Table
2) and these data were pooled in the final analysis
as “tropical” (between sites minimum detectable
slope difference = 0.60%, minimum detectable
intercept difference = 1 6%). In contrast, the pop-
ulation in Texas grew less per molt: the slope of
the regression line was significantly lower (AN-
COVA interaction term. Table 2). Analysis of
the residuals revealed that the early instars were
larger in Texas than in the tropics (ANOVA,

post-molt tibia+patella length, cm

HIGGINS -LIFE-HISTORY PLASTICITY AND FECUNDITY

99

tibia+patella

Figure 2.— The intermolt interval in days plotted against tibia + patella length (t + p) for all three populations:
days = 7.18 + 8.56 (t + p), n = 31, R- = 0.525, P < 0.001. (A = Veracruz; □ = Panama wet; ■ = Panama
dry; A = Texas).

premolt tibia+patella length

Figure 3.— Pre-molt t + p vs. post-molt t + p for Texas, Panama and Veracruz. The arrows mark the mean
adult female size in each population (Y). The two generations in Panama grew at the same rate (dry: y = 0.047
+ 1.300 X (R2 = 0.970); wet: y = 0.050 + 1.310 x {R^ = 0.985)). The Veracruz population grew at a similar
rate (y = 0.057 + 1.295 x (R^ = 0.992). The Texas spiders grew significantly less per molt (y = 0.068 + 1.183
X (R^ = 0.984)).

100

THE JOURNAL OF ARACHNOLOGY

Table 2.— ANCOVA of regression lines of growth per molt: Comparison between the two tropical populations
shows no difference in (a) slope or (b) intercept. Comparison of the tropical populations and Texas (c) reveals
significant difference in slope.

Source

SS

df

F

P

a. pre-molt size

41.48

1

17 342.9

<0.001

Veracruz vs. Panama (intercept)

0.001

1

0.451

0.503

interaction (effect on slope)

0.000

1

0.193

0.661

error

0.409

171

-

-

b. pre-molt size

42.14

1

17771.0

<0.001

Veracruz vs. Panama (intercept)

0.001

1

0.28

0.597

error

0.409

172

-

-

c. pre-molt size

43.71

1

6671.25

<0.001

tropical vs. Texas (intercept)

0.007

1

1.06

0.305

interaction (effect on slope)

0.075

1

11.47

0.001

error

1.61

246

-

-

tropical vs. Texas, n = 250, T'(i,248) = 12.30, P =
0.001), suggesting that the Texas spiders hatch
at a larger size.

Females matured at different mean sizes in
different populations (Fig. 3, Table 3). Compar-
ison of mature female t + p for the two gen-
erations on Panama, Veracruz, and Texas indi-
cated that the four groups were significantly
different (ANOVA: .g,, = 70.7, P < 0.001).

As the spiders in Panama and Veracruz grew the
same amount per molt, the variation in size at
maturity within these tropical populations re-
flects differences in the number of juvenile in-
stars. The small size of females in Texas reflects
the lower growth per molt and probably also a
lower number of juvenile molts.

Reproduction.— Reproductive effort was mea-
sured by the number of eggs and weight of the
first egg sac, compared between the two tropical
populations as a function of female size (t + p),
pre-laying weight, and post-laying weight. In ad-
dition, some free-living females were observed
to lay several egg sacs (uncollected), providing
an estimate of the interval between egg sacs. First
egg sacs were collected from 1 5 females on Pan-
ama (wet: 9 (June-September); dry: 5 (Decem-
ber-January), 1 (April)). Only two first clutches
were collected in Veracruz, but these were in-
cluded in the analysis for comparison. The low
sample size reflects the difficulty of following
marked individuals from the last molt to the first
oviposition in the dense Veracruz vegetation.
There was no significant difference in tibia +
patella length or final weight among females sam-
pled between the generations on Panama. Fe-

male size and final pre-laying weight were cor-
related and there was an effect of site/season (Fig.
4. ANCOVA t + p: = 10.494, P = 0.008;

site/season: iyj n) = 5.296, P = 0.024 (excluding
the April female because of undue influence in
the regression)). This reflects a relatively higher
final weight for female spiders of similar tibia +
patella length in Panama (June-September) com-
pared to Veracruz and Panama (December-Jan-
uary).

Only two females laid large numbers of infer-
tile eggs (3% and 5% of 1 35 1 and 1170 total eggs,
respectively), one from each generation in Pan-
ama. The number of good eggs and clutch weight
were significantly correlated with the pre-laying
weight of the females (Fig. 5) and positively cor-
related with female size (regression analysis,
number of eggs: slope = 956, = 0.30, iy,, ,3,

= 5.50, P = 0.035; clutch weight (g): slope =
0.553, = 0.15, ,3, = 2.24, ns). There was

no effect of site or season on the relationship
between female size and fecundity. Nor were there
differences between generations on Panama in
clutch weight {n = 17, Kruskall-Wallis = 2.44,
df = 2, ns) or egg number (« = 17, Kruskall-
Wallis = 1.38, df = 2, ns). Relative clutch mass
(RCM) was high; females laid on average 104%
of their post-laying weight in eggs (range 80-
129%) (Table 3). RCM did not vary with female
size or with site/season (ANCOVA: no interac-
tion effect, t + p: F), ,4, = 0.74, ns; site/season:
^(2. 14) = 1-15, ns). Non-egg weight gain was es-
timated by comparing estimated post-molt weight
to post-laying weight for five females, four in
Panama and one in Veracruz. In these females.

HIGGINS- LIFE-HISTORY PLASTICITY AND FECUNDITY

101

Table 3.— Mean female size (leg I tibia + patella, t -f p) at maturity and fecundity data for first egg sacs
collected in Panama and Veracruz; mean female size in Texas. RCM: relative clutch mass (=clutch mass/post-
laying weight); mean egg weight considers only viable eggs (= weight of good eggs/n good eggs). * (ANOVA P <
0.001), § (Fisher PSLD P < 0.05).

Generation

Mean t + p (SD, n)

Number of
egg sacs

Mean post-laying
weight (SD)

RCM

Mean egg wt (SD)

Panama June

1.57(0.14, 83)*

10

0.933 (0.26)

1.08 (0.14)

0.804 (0.06)§

Panama Dec.

1.43 (0.14, 46)*

5

0.809 (0.2)

i.OI (0.15)

0.724 (0.08)§

Veracruz

1.71 (0.15, 38)*

2

1.05

0.93

0.90

Texas

1.31 (0.14, 42)*

0

-

-

-

the weight of the eggs was equal to on average
65% of the weight gained between molting and
laying; non-egg weight gain averaged 0.21 g.

The mean egg weight (total weight of good
eggs/number of good eggs) was variable (overall
range 0.65-0.90 mg) but was not correlated with
female size, female weight or total number of
good eggs (regression analysis, t + p: Ty,, ,4, =
0.06, ns; weight: F(, ,4, = 1.27, ns; number of
eggs: F,, ,4) = 0.01, ns). Within Panama, mean
egg weight was greater in June-September clutch-

es than in December-January and April clutches
(Table 3; Fisher PLSD = 0.077, P < 0.05).

The exact number of days between final molt
and first reproduction is known for seven females
in Panama and three females in Veracruz. In
both populations, females laid within 30 days of
the final molt (Panama, range 18-29; Veracruz,
range 24-28). Observations of free living females
indicate that egg sacs are laid about 20 days apart,
and one free-living female in Panama laid five
fertile egg sacs.

E

g)

'o

O)

£

ro

I

o

Cl

tibia+patella length, cm

Figure 4.— Estimated adult female pre-laying weight vs. adult size (t + p). The slope is significantly positive,
and the June-September females tend to lay at a higher weight for their size, weight = — 1.59 + 2.35 (t + p),
{R^ = 0.52, n = 16; April female deleted (see text)). (A = Panama April; ▲ = Panama June-September; • =
Panama December-January; □ = Veracruz).

102 THE JOURNAL OF ARACHNOLOGY

0 12 3

o

<a

Vi

m

m

U)

O)

<D

o

n

E

Figure 5.— Female fecundity and clutch weight as a function of estimated pre-laying weight for spiders in the
Panama and Veracruz populations, a. Weight of eggs laid: clutch weight = 0.008 + 0.410 (spider weight) =
0.885). b. The number of eggs laid: n = 193.4 + 474.7 (spider weight) (R^ = 0.801). (A = Panama April; A =
Panama June-September; • = Panama December-January; □ = Veracruz).

HIGGINS-LIFE-HISTORY PLASTICITY AND FECUNDITY

103

DISCUSSION

Nephila davipes development reflects the in-
teraction of environmentally-induced individual
variation and population-specific developmental
constraints. Whereas food levels and perhaps also
foraging and defensive expenditures can cause
changes in growth rates and intermolt intervals,
there is little variation within a population in the
morphometric relationships involved in molt-
ing. This is true for both the size change per molt
and in the correlated parameter of pre-molt
weight. However, the rate of weight gain, the
intermolt interval and the number of juvenile
instars vary greatly within and among popula-
tions. Variation in female age and size at sexual
maturity within populations is probably related
to differences in the intermolt interval and num-
ber of juvenile instars. The instar at which a
female matures appears to reflect individual de-
velopmental history and the interaction of the
female’s chronological age and developmental
stage (instar) with the seasonality of the site. Size
at sexual maturity and total fecundity are deter-
mined by the growth per molt, the number of
molts, and the number of egg sacs produced.

Growth per molt in N. davipes is not depen-
dent upon sex or individual foraging success, and
minimum growth per molt appears to be pop-
ulation specific. Spiders of the Panama dry-sea-
son generation and in Texas have equivalent for-
aging success, and those of the Panama wet-season
generation and in Veracruz have equivalent for-
aging success (Higgins & Buskirk in press). How-
ever, the spiders under dry-season conditions in
Panama grew the same amount per molt as those
during the Panama wet-season or the Veracruz
population. Likewise, in the laboratory experi-
mental differences in prey availability did not
affect growth per molt in Panamanian spiders
(Higgins pers. obs.). Growth per molt in spiders
of both tropical populations was significantly
greater than growth per molt by spiders in Texas.
The population-specific growth per molt, the lack
of variation in size-specific pre-molt weight and
the relationship of pre-molt weight to post-molt
size imply that the growth per molt is controlled
by a population specific minimum pre-molt
weight. Increased feeding can cause an individual
to “skip” a molt, with a longer intermolt time
and greater tibia + patella growth at the molt
but reduction of feeding lengthened the intermolt
interval without decreasing the growth per molt

(Higgins pers. obs.). The inflexibility of the min-
imum growth per molt under experimental con-
ditions and between seasons in Panama indicates
that the minimum pre-molt weight may have a
strong genetic component.

The intermolt interval might therefore be the
time required to achieve necessary pre-molt
weight. Weight gain is highly plastic and depen-
dent on prey-capture rates (Turnbull 1962; Hig-
gins pers. obs.). An inverse correlation between
the rate of weight gain and instar duration in the
field was expected from laboratory experiments
(Turnbull 1962; Vollrath 1988; Hi^nspers. obs.)
and failure to observe significant differences in
instar duration among populations may have
been due to low sample sizes and high within-
population variation. The developmental con-
straint of fixed growth per molt and plastic in-
termolt interval is distinct from some insects and
at least one spider. Larvae of some Lepidoptera
and Coleoptera are known to molt even when
not gaining weight, apparently constrained by an
internal clock mechanism to a maximum inter-
molt interval (Beck 1950, 1973; Nijhout 1971).
Likewise, the spider Linyphia triangularis Clerck
is reported to molt at a variety of diet-dependent
weights (Turnbull 1962). However, the effect of
this variation in pre-molt weight on intermolt
interval and growth per molt was not presented.

The relative roles of individual history and
environment in determining the instar of sexual
maturity in spiders is not clear. In other spiders,
differences in environmental conditions over
small geographic distances are correlated with
differences in growth rates (Miyashita 1 986), size
(instar?) at maturity (Toft 1983; Benton & Uetz
1986), and weight gain by mature females (Wise
1975, 1979; Fritz Morse 1985). Total devel-
opment time can be shortened in part through
lowered pre-molt weight requirement, corre-
sponding to shorter intermolt periods and re-
sulting in lower growth per molt, manifested as
a reduced slope of the developmental trajectory.
It is possible that in habitats such as Texas, with
low foraging success and strong seasonal changes,
the total development time has shortened to per-
mit reproduction before the onset of winter. Ma-
ture females in Texas were smallest, reflecting
lower growth per molt and probably also mat-
uration at an earlier instar.

Differences among populations in the number
of juvenile instars were also related to the length
of the growing season. Females in Veracruz, with

104

THE JOURNAL OF ARACHNOLOGY

the longest growing season (April-August) were
largest. Females in Panama alternated, with fe-
males of the the first generation larger than those
of the second generation. The instar of maturity
in Panama was inversely related to early juvenile
feeding success. The females maturing in June-
September were immatures during the season of
low foraging success (the dry season), and were
larger at sexual maturity than the December-
January females. The latter were immatures dur-
ing the season of high foraging success (the wet
season). The inverse relationship with juvenile
feeding success is in contrast to earlier laboratory
studies, where it was found that male N. clavipes
matured at an earlier instar under a low, constant
feeding regime (Vollrath 1983). The first gener-
ation Panama juveniles may be taking advantage
of improving environmental conditions (end of
the dry season in May) by delaying maturity to
a larger instar. Their daughters, juveniles during
the rainy season, matured just before or as the
dry season began, apparently influenced by the
environmental cues heralding the dry season. This
hypothesis is supported by their lower pre-laying
weight compared to females of the June gener-
ation: they oviposited at low weights very early
in the dry season before drought conditions were
severe. It is striking that neither the relative clutch
mass (RCM) nor the number of eggs per clutch
were lowered; the lower mean egg weight may
be a necessary compensation for maintaining the
high number of eggs. The consequences of dif-
ferences in egg size to offspring size and survi-
vorship are unknown.

Female size and weight gain after sexual ma-
turity are correlated with fecundity in clutch
weight and number of eggs (Benton & Uetz 1986;
Miyashita 1986; Harrington 1978; Wise 1975;
Eberhard 1979; McLay & Hayward 1987). As in
N. clavipes, most post-molting weight gain by
mature female L. triangularis was involved in
egg production (Turnbull 1962). N. clavipes has
a high RCM compared to other species of spiders
(McLay & Hayward 1986). The size indepen-
dence of RCM has been found for spiders of six
other families (McLay & Hayward 1987), pill-
bugs (Lawlor 1976) and in several species of
snakes (Seigel, Fitch & Ford 1986).

Commitment to sexual maturity in female N.
clavipes probably reflects a complex interaction
between female chronological age, instar, and en-
vironmental seasonality. Because an individual
in the penultimate instar has partially developed

sexual characters, commitment to sexual matu-
rity is triggered at least two instars before sexual
maturity, the ante-penultimate instar. Two in-
stars before maturity at the end of the Pana-
manian rainy season corresponds to commit-
ment in October and November, preceding the
onset of the dry season by at least two months.
Several factors occurring in these months might
function as cues: changing photoperiod, the steady
decline in insect abundance between June and
September (Olive 1981; Smythe 1982), and in-
creased rainfall in October and November (D.
Winsor, pers. comm.).

In contrast, the arrival of the first winter storm
in coastal Veracruz is apparently not heralded
by cues to which the spiders respond. This weath-
er shift is temporally highly variable (occurring
between September and December), and in 1 986
many individuals did not reproduce before they
disappeared with the early October onset of the
winter storms (Higgins pers. obs.). These indi-
viduals were perhaps “taking a chance” in grow-
ing larger, increasing their reproductive output
through greater potential size, but risking total
failure if the weather changed before they could
lay (Lawlor 1976) or perhaps were delayed by
late emergence or poor foraging conditions.

The phenology of N. clavipes in each location
reflects a complex interaction between environ-
mental constraints and developmental con-
straints and plasticity. These spiders have some
ability to modulate their development: weight
gain and intermolt times are altered by changes
in prey availability (Higgins pers. obs.), and the
differences in growth per molt between temper-
ate and tropical populations may reflect local
adaptation to the differing strength of seasonality
(Toft 1983; Baldwin & Dingle 1986). The num-
ber of generations per year in populations of N.
clavipes thus far studied appears to be externally
regulated (pers. obs.) and therefore there is prob-
ably a greater increase in r (the intrinsic rate of
increase) by increasing individual fecundity than
by shortening generation time (Steams & Koella
1986). The relative fitness of individual females,
as measured by fecundity, is dependent upon all
of these factors as they affect growth, size at ma-
turity, and successful anticipation of seasonal
changes in weather.

ACKNOWLEDGMENTS

Invaluable voluntary field assistance was pro-
vided by J. Garcia Guzman and F. Conejo in

HIGGINS- LIFE-HISTORY PLASTICITY AND FECUNDITY

105

Panama and J. Villarreal and P. Batra in Vera-
cruz. Conversations with members of the Barro
Colorado Island community and logistical sup-
port from the Smithsonian Tropical Research
Institute staff were essential in the development
of this project. I am grateful to G. Cameron for
arranging access to the University of Houston
Coastal Center and to the Instituto de Biologia
for permission to work at Los Tuxtlas. In all
aspects of the project and the analysis, several
members of the faculty of the University of Texas
have provided assistance: R. Buskirk, B. Mc-
Guire, C. Pease, W. Reeder and M. Singer. Dis-
cussion and comments on various versions of
the manuscript were provided by them and by
G. Ceballos, C. Kelly, A. Pescador, D. Reznick,
R. Seigel, A. Cady, K. Cangialosi, and three
anonymous reviewers. N. Fowler provided valu-
able statistical advice. This work was supported
in part by Sigma-Xi, the University of Texas
Graduate Fellowship program, a National Sci-
ence Foundation Dissertation Improvement
Grant (BSR-84 13831) and a STRI short-term
fellowship, and was presented in partial fulfill-
ment of a doctoral degree at the University of
Texas at Austin.

LITERATURE CITED

Atkinson, D. & M. Begon. 1987. Reproductive vari-
ation and adult size in two co-occuring grasshopper
species. Ecol. Ent., 12:119-127.

Baldwin, J. D. & H. Dingle. 1986. Geographic vari-
ation in the effects of temperature on life-history
traits in the large milkweed bug Oncopeltus faciatus.
Oecologia, 69:64-71.

Beck, S. 1950. Nutrition of the European com borer,
Pyraustra nubilalis (HBN.). 11. Some effects of diet
on larval growth characteristics. Physiol. ZooL, 23:
353-361.

Beck, S. 1973. Growth and retrogression in larvae of
Trogoderma glabrum (Coleoptera: Dermestidae). 4.
Developmental characteristics and adaptive func-
tion. Annals Ent. Soc., 66:895-900.

Benton, M. J. & G. W. Uetz. 1986. Variation in life-
history characteristics over a clinal gradient in three
populations of a communal orb-weaving spider.
Oecologia, 68:395-399.

Bervin, K. A. & D. E. Gill. 1983. Interpreting geo-
graphic variation in life-history traits. American
Zook, 23:85-97.

Christenson, T. E. & K. C. Goist. 1979. Costs and
benefits of male-male competition in the orb-weav-
ing spider Nephila clavipes. Behav. Ecol. Sociobiok,
5:87-92.

de la Cruz, M. & R. Dirzo. 1987. A survey of the
standing levels of herbivory in seedlings from a
Mexican rain forest. Biotropica, 19:98-106.

Dingle, H., T. A. Mousseau & S. M. Scott. 1990.
Altitudinal variation in life cycle syndromes of Cal-
ifornia populations of the grasshopper Melanoplus
sanguinipes (F.). Oecologia, 84:199-206.

Eberhard, W. G. 1979. Rates of egg production by
tropical spiders in the field. Biotropica, 1 1 :292-300.

Eberhard, W. G. 1982. Beetle horn dimorphism:
making the best of a bad lot. American Nat., 1 19:
420-426.

Etter, R. J. 1989. Life history variation in the inter-
tidal snail Nucella lapillus across a wave-exposure
gradient. Ecology, 70:1857-1876.

Foelix, R. F. 1982. The Biology of Spiders. Harvard
University Press, Cambridge, Mass.

Ford, N. B. & R. A. Seigel. 1989. Relationships among
body size, clutch size and egg size in three species
of oviparous snakes. Herpetologia, 54:75-83

Fritz, R. S. & D. H. Morse. 1985. Reproductive suc-
cess and foraging of the crab spider Misumena vatia.
Oecologia, 65:194-200.

Garcia, E. 1973. Modificaciones al Systema de Clas-
sificacion Climatica de Koppen. Universidad Na-
cional Autonoma de Mexico Press, Mexico City.

Harrington, C. L. 1978. Field studies on reproduction
in the funnel-web spider Agelenopsis potteri. Oikos,
31:368-375.

Higgins, L.E. 1987. Time budget and prey of

clavipes (Linnaeus) (Araneae: Araneidae) in south-
ern Texas. J. ArachnoL, 15:401-417.

Higgins, L. E. 1988. Variation in web structure in the
orb-weaving spider Nephila clavipes and correlated
changes in life history. Ph. D. Dissertation. Uni-
versity of Texas at Austin.

Higgins, L. E. 1 990. Variation in foraging investment
during the intermolt and before egg-laying in the
Nephila clavipes. J. Insect Behav., 3:773-783.

Higgins, L. E. & R. Buskirk. In press. A trap-building
predator exhibiting different tactics for different as-
pects of foraging behaviour. Anim. Behav.

Higgins, L. E. In press. Developmental changes in
barrier web stmcture under differing levels of pre-
dation risk. J. Insect Behav.

Hill, E. M. & T. E. Christenson. 1981. Effects of prey
characteristics and web stmcture on feeding and
predatory responses of Nephila clavipes spiderlings.
Behav. Ecol. Sociobiok, 8:1-5.

Hugueny, B. & A. Louveaux. 1986. Gradient d’ari-
dite et variation latitudinale de la taille, dans des
populations de Calliptamus barbarus (Costa, 1836)
(Insecte, Orthoptore, Acrididae). Acta Ecologica Ecol.
Gener., 7:317-333.

Kimmel, D. L. & C. T. Grant. 1980. Attitude change
of Nephila clavipes spiderlings during communal life.
J. Arachnok, 8:53-58.

Lawlor, L. R. 1976. Molting, growth and reproduc-

106

THE JOURNAL OF ARACHNOLOGY

live strategies in the terrestrial isopod Armadilli-
dium vulgare. Ecology, 57:1 179-1 194.

Leigh, Jr.,E. G., A. S. Rand & D. M. Windsor. 1982.
The Ecology of a Tropical Forest: Seasonal rhythms
and long-term changes. Smithsonian Institution
Press, Washington D. C.

Levi, H. W. 1980. The orb-weaver genus

the subfamily Metinae and the genera Pachynatha.
Glenognatha and Azilia of the subfamily Tetrag-
nathinae North of Mexico (Araneae: Araneidae).
Bull. Mus. Comp. ZooL, 149:1-75.

Lubin, Y. D. 1978. Seasonal abundance and diversity
of web-building spiders in relation to habitat struc-
ture on Barro Colorado Island. J. Arachnol., 6:31-
35.

Masaki,S. 1978. Seasonal and latitudinal adaptations
in life cycles of crickets. Pp. 72-100, In Evolution
of Insect Migration and Diapause. (H. Dingle, ed.).
Springer- Verlag, New York.

McLay, C. L. & T. L. Hayward. 1987. Reproductive
biology of the intertidal spider Desis marina (Ara-
neae: Desidae) on a New Zealand rocky shore. J.
Zool. (London), 21 1:357-372.

Miyashita, T. 1986. Growth, egg production and pop-
ulation density of the spider Nephila clavata in re-
lation to food conditions in the field. Research in
Population Ecol., 28: 135-149.

Morse, D. H. & R. S. Fritz. 1987. The consequences
of foraging for reproductive success. Pp. 443-455,
In Foraging Behavior (A.C. Kamil, J. Krebs & H.
R. Pulliam, eds.). Plenum Publishing Co.

Mousseau, T. A. & D. A. Rolf. 1989. Adaptation to
seasonality in a cricket: Patterns of phenotypic and
genotypic variation in body size and diapause ex-
pression along a dine in season length. Evolution,
43:1483-1496.

Newman, R. A. 1988a. Genetic variation for larval
anuran (Scaphiopus couchii) development time in
an uncertain environment. Evolution, 42: 763-773.

Newman, R. A. 1988b. Adaptive plasticity in de-
velopment of Scaphiopus couchii tadpoles in desert
ponds. Evolution, 42:774-783.

Nijhout, H. F. 1981. Physiological control of molting
in insects. American Zool., 21:631-640.

Olive, C. W. 1981. Optimal phenology and body size
of orb-weaving spiders: Foraging constraints. Oec-
ologia, 49:83-87.

Palmer, J.O. 1985. Life-history consequences of body-
size variation in the milkweed leaf beetle: Labidom-
era clivicollis (Coleoptera: Chrysomelidae). Annals
Ent. Soc. America, 78:603-608.

Schmalhausen, 1. 1. 1949. Factors of Evolution: The
Theory of Stabilizing Selection, (translated by 1.
Dordick, 1986). The Univ. of Chicago Press.

Siegel, R. A. & H. F. Fitch. 1984. Ecological patterns
of relative clutch mass in snakes. Oecologia, 6 1 :293-
301.

Siegel, R. A., H. F. Fitch. & N. B. Ford. 1986. Vari-
ation in relative clutch mass in snakes among and
within species. Herpetologica, 42:179-185.

Smythe, N. 1982. The seasonal abundance of night-
flying insects in a neotropical forest. In The Ecology
of a Tropical Forest. (E. G. Leigh, Jr., A. S. Rand
& D. M. Windsor, eds.), Smithsonian Institution
Press, Washington, D. C.

Sokal, R. R. & F. J. Rohlf. 1981. Biometry. The
Principles and Practice of Statistics in Biological
Research, 2nd edition. W. H. Freeman and Co. San

Francisco.

Stems, S.C. 1983. The evolution of life-history traits
in mosquito fish since their introduction to Hawaii
in 1905: Rates of evolution, heritabilities, and de-
velopmental plasticity. American Zool., 23:65-75.

Stems, S. C. & J. C. Koella. 1986. The evolution of
phenotypic plasticity in life-history traits: predic-
tions of reaction norms for age and size at maturity.
Evolution, 40:893-913.

Suter, R. B. 1 990. Determinants of fecundity in Fron-
tinella pyrmitela (Araneae, Linyphiidae). J. Arach-
nol., 18:263-270.

Tauber, M. J. & C. A. Tauber. 1978. Evolution of
phonological strategies in insects: A comparative ap-
proach with eco-physiological and genetic consid-
erations. Pp. 53-71, In Evolution of Insect Migra-
tion and Diapause. (H. Dingle, ed.). Springer- Verlag,
New York.

Toft, S. 1976. Life-histories of spiders in a Danish
beech wood. Natura Jutlandica, 19:5-40.

Toft, S. 1 983. Life cycles of Meta segmentata (Clerck
1757) and Meta mengei (Blackwall 1 869) in Western
Europe (Arachnida: Araneae: Tetragnathidae). Ver.
Naturwiss. Ver. Hamburg, 26:265-276.

Turnbull, A. L. 1962. Quantitative studies of the food
of Linyphia triangularis Clerck (Araneae: Linyphi-
idae). Canadian Ent., 94: 1233-1249.

Vollrath, F. 1983. Relative and absolute growth in
Nephila clavipes (Arachnida: Araneae: Argiopidae).
Ver. Naturwiss. Ver. Hamburg, 26:277-290.

Vollrath, F. 1988. Spider growth as an indicator of
habitat quality. Bull. British Arachnol. Soc. 7:217-
219.

Wilkinson, L. 1987. SYST AT: The System for Sta-
tistics. SYSTAT Inc., Evanston, Illinois.

Wise, D. H. 1975. Food limitation of the spider Lm-
yphia marginata: experimental field studies. Ecol-
ogy, 56:637-646.

Wise, D. H. 1979. Effects of an experimental increase
in prey abundance upon reproductive rates of two
orb-weaving spider species. Oecologia, 4 1 :289-300.

Manuscript received October 1991, revised January
1992.

1992. The Journal of Arachnology 20:107-1 13

BALLOONING: DATA FROM SPIDERS IN FREEFALL
INDICATE THE IMPORTANCE OF POSTURE

Robert B. Suter: Department of Biology, Vassar College, Poughkeepsie, New York
12601 USA

ABSTRACT. Ballooning, the aerial displacement of a spider caused by friction between air and the spider
with its silk, has considerable ecological importance but remains poorly understood as a mechanical process.
The studies reported here provide insight into the mechanics of ballooning by way of experiments involving ( 1 )
stroboscopic measurement of the rates of fall of spiders slowed by known lengths of silk and (2) direct mea-
surement of the drag generated by moving air at the surface of spiders as those spiders change their postures.
The terminal velocities of spiders trailing silk, derived from their rates of fall at known distances from release,
were remarkably variable. The variability could not be attributed entirely to silk length and spider mass, and
in some cases silk length was not even statistically correlated with terminal velocity. Drag measurements made
on living spiders revealed that the variability in terminal velocities could be explained by variability in posture.

Some spiders, particularly small ones, travel
great distances by ballooning. The ecological cir-
cumstances under which ballooning occurs are
now well-documented (e.g., Richter 1970, Vugts
& Van Wingerden 1976), and data derived from
newly developed techniques (Greenstone et al.
1985) should increase our understanding of the
temporal, spatial, and geographical parameters
of ballooning. The physics of ballooning, in a
theoretical sense, has also been well elucidated,
both very generally (Vogel 1981) and in consid-
erable detail (Humphrey 1987). In contrast, our
empirical information about the physics of bal-
looning has been limited to deductions from mi-
crometeorological, behavioral, and morphomet-
ric observations of ballooning (e.g., Dean &
Sterling 1985, Eberhard 1987, Greenstone et al.
1987), and a single experimental study (Suter
1991).

In that earlier paper, I reported on the drag
produced by silk and by whole spiders under the
laminar flow conditions of a wind tunnel. The
most useful results of the study were a set of
equations relating the weights of spiders and the
lengths of their ballooning silks to their terminal
velocities and to the velocities of rising air nec-
essary to make them airborne. Because the ex-
periments involved spiders of several families,
many sizes, and a variety of shapes, the results
were quite general and therefore directly appli-
cable, in a sense, only to spiders of average shape
and a particular posture. Nevertheless, for a liv-
ing spider of a particular weight, the equations

allowed one to predict roughly its terminal ve-
locity at any known length of ballooning silk (Su-
ter 1991).

The experiments reported below constitute an
extension of that work, by testing the true ter-
minal velocities of falling spiders against the pre-
dictions generated in the earlier work.

METHODS

Spiders.— The great majority of ballooning
spiders have masses below 2 mg (Greenstone et
al. 1985, 1987), although much larger spiders are
found among the aerial plankton (Coyle et al.
1985). Because of that size distribution, only small
spiders were used in this study. In freefall trials,
five spiders were subjects: three were unidenti-
fied hatchling theridiids (0. 1 8, 0. 1 8, and 0.2 mg),
one was an unidentified immature theridiid (1.8
mg) and one was an immature Uloborus glo-
mosus (Walckenaer) (Uloboridae) (1.0 mg). In
the drag measurement trials, both subjects were
unidentified theridiids, one a hatchling (0.18 mg)
and one an immature (7.2 mg). The spiders were
captured from their webs in August and Septem-
ber 1 99 1 and were maintained in plastic vials at
1 00% R. H. for the few hours to few days between
capture and testing. They were not fed during
their captivity and were weighed within 24 hours
of testing.

Freefall.— Figure 1 shows the apparatus used
to measure the acceleration and velocity of spi-
ders falling through still air in a dark room. In
each trial, a spider was induced to attach its drag-

107

108

THE JOURNAL OF ARACHNOLOGY

nichrome

Figure 1.— Apparatus used to measure the velocity
of a spider falling through still air while trailing a known
length of dragline silk. Electrical current applied to the
nichrome wire loop at the top released the silk and
spider. The falling spider, illuminated by stroboscope
flashes at 100/s, was photographed as it fell past the
scroboscope and camera (not shown). The nichrome
wire loop could be adjusted to any height, allowing
experimental manipulation of the distance the spider
fell before being photographed.

line to a small nichrome wire loop (diameter: 0.5
cm). At some point during its subsequent de-
scent, electrical current was fed to the wire loop
which incandesced rapidly and released the spi-
der trailing its dragline (a hot-wire anemometer
indicated no measurable air flow, at 1 cm below
the wire loop, during the first 0.5 s after the loop
was energized). Dim light focused on a ruler near
the path of the dangling and then falling spider
allowed an observer to note and record the length
of silk with which the spider had been suspended
just prior to its release. A camera with its shutter
open, mounted so that it focused on both the
ruler and the path of the falling spider, recorded
the position of the spider during each flash of a
strobe (set at 100 flashes/s). Because of the very
brief duration of each flash, high speed film (Ko-

dak TMAX 3200) was used to record the spider’s
motion. The height of the nichrome wire loop
was adjustable, which made it possible to alter
the distance a spider fell before it appeared in
front of the camera.

Each spider in these trials was tested repeat-
edly, at different silk lengths and at different dis-
tances from its release. The resulting velocity vs.
distance lines were plotted together with theo-
retical curves derived from drag measurements
made on spiders and their silk in wind tunnel
experiments reported elsewhere (Suter 1991).
Many of the velocity vs. distance lines were long
enough to confirm that their shapes conformed
to the theoretical curves (e.g.. Figs. 3-5). I there-
fore assumed that all of the lines represented
segments of similarly shaped curves, and esti-
mated the terminal velocities by extrapolation
in cases in which the final three data points did
not indicate relatively constant velocity (e.g., Figs.
6-7).

Drag measurement. — Data derived from the
freefall trials suggested that variation in spider
posture strongly influenced the rates of fall of
spiders trailing draglines. Because posture was
difficult to ascertain from the images produced
during freefall trials, I designed the apparatus
shown in Fig. 2 to allow direct observation of
posture and simultaneous measurement of drag.
The apparatus constituted a miniature wind tun-
nel in which air near the aperture was in laminar
flow and had velocity (measured by a hot wire
anemometer) that could be closely controlled (at
air speeds less than 0.2 m/s, velocities varied less
than 0.008 m/s). The tests reported here took
place at an air velocity of 0.07 m/s.

Each spider in these trials was attached at the
posterior surface of the abdomen (with cyano-
acrylate glue) to the end of a wire 1.4 cm long
(diameter: 17.8 iim). The other end of this wire
was connected with paraffin to the end of a gal-
vanometer needle (see Fig. 2). An opaque plastic
target (1.2 x 2.0 mm) mounted on the galva-
nometer needle partially interrupted the beam of
a HeNe laser, so that very small movements of
the needle were detected as changes in the amount
of light falling upon a photodetector. DC output
from the photodetector was converted to a fre-
quency modulated signal by passing it through a
voltage-controlled oscillator circuit, and the FM
signal was recorded on one audio channel of the
videotape that was recording the activities of the
spider.

SUTER- BALLOONING: POSTURE AND TERMINAL VELOCITY

109

Air Source

nSSWS

■.SSAW

sSSSW

vSASAS

vSSSSA

v\SSS\

vSSSW

vSSSAA

\S\S%\

\SSSSS

^SS^S^

sS\S%\

HcNe

Laser

Air Source

Galvanometer

Figure 2.— Side view (a) and top view (b) of the apparatus used to measure drag generated by a living spider
in a laminar air flow while recording changes in its posture. Changes in the posture of the spider (above the
middle arrow in a, and inset) caused changes in friction between the moving air and the spider’s surface, and
that change in drag caused movement of the needle. The change in position of the needle was detected as an
altered amount of light falling on the photodetector (in b). Both the FM -encoded signal bearing drag information
and the video signal bearing posture information were recorded for later analysis.

110

THE JOURNAL OF ARACHNOLOGY

Distance from Release (cm)

Silk Length (cm)

Figures 3-1 2.— Each graph on the left contains data pertaining to a different spider as it fell, repeatedly, through
still air. In each graph, the top curve represents the velocity of an object falling through a vacuum; the other
two curves represent velocities calculated from Suter (1991) of a spider of the same mass as the test spider and
with the shortest length of trailing silk used in the tests shown (middle curve) or the longest length of trailing
silk (bottom curve). The shorter curves in each graph represent the velocities of the test spider in separate
freefalls: because of the experimental setup (see text and Fig. 1), curve segments originating nearer to the origin
are from trials in which a relatively long length of silk acted as the balloon, and segments originating at longer
distances from release are from trials with shorter silk lengths. Each graph on the right shows the relationship
between silk length and terminal velocity of each test spider in Figs. 3-7. The test spider identities, masses, and

SUTER- BALLOONING: POSTURE AND TERMINAL VELOCITY

111

I calibrated the apparatus at the end of each
trial by first turning off the air flow and then
giving the spider a series of coils of very fine
nichrome wire (diameter: 2.5 iim), each of mea-
sured length (and therefore of known weight).
Fortunately the spiders were very cooperative:
each grabbed and held on to each wire coil, ma-
nipulating it for several seconds then dropping
it, allowing me to record (as above) the effect of
each weight on the output of the photodetector.
Because output from the photodetector was re-
corded on the same tape that carried postural
information in the form of a video signal, sub-
sequent analysis of the relationship between drag
and posture was facilitated.

RESULTS

Spiders with trailing silk draglines passed the
camera lens at velocities that were in part de-
pendent on how far they had fallen. Figures 3-7
show the velocities of these falling spiders as
functions of distance from the release point (i.e.,
the elevation of the spider when its fall began).
Each figure presents the data from multiple re-
leases of the same spider, all plotted against three
theoretical curves. Those theoretical curves rep-
resent (1) the velocity of a spider falling in a
vacuum (no drag), (2) the velocity of a spider of
the same mass as the test animal, trailing the
minimum length of silk used by the test spider
during the trials (drag calculated from Suter 1991),
and (3) the velocity of a spider of the same mass
as the test animal, trailing the maximum length
of silk used by the test spider during the trials
(drag calculated from Suter 1991). Of the five
spiders tested in this study, only the immature
uloborid produced data that consistently corre-
sponded closely with expectations derived from
the wind tunnel experiments.

Because the distance from release varied in-
versely with silk length, individual curves near
the origins of Figs. 3-7 represent spiders with
relatively long draglines and curves far from the
origins represent spiders trailing relatively short
lengths of silk. Sometimes a spider fell more
slowly when trailing a longer dragline than when
trailing shorter draglines, but this relationship

Figure 13.— Drag generated on the surfaces of a ther-
idiid hatchling (0. 1 8 mg). The horizontal lines indicate
drag at different postures: (a) all legs extended and wav-
ing; (b) all legs loosely held against the body; (c) all legs
tightly held against the body. Note that a very small
change in posture (from b to c) resulted in a substantial
reduction in drag.

was by no means perfect. Significant inverse re-
lationships between silk length and terminal ve-
locity were found for both of the 0. 1 8 mg ther-
idiid spiderlings (Figs. 8, 9) but not for the other
three test animals (Figs. 1 0-12), all of which were
larger. The substantial deviations from the ex-
pected relationship led to the tentative conclu-
sion that posture must play a prominent role in
influencing terminal velocity.

Measurements of drag on spiders mounted in
a laminar air stream are shown in Figs. 13, 14.
Relatively small changes in posture for both very
small and somewhat larger spiders caused as
much as a 10-fold change in drag, even at the
very low air velocity used in this test.

DISCUSSION

A spider falling through still air accelerates un-
til the drag produced by air flowing past its body
and trailing silk just equals the pull of gravity:
at that point, acceleration is zero and the spider
is at its terminal velocity. Thus terminal velocity
is a function of the mass of the spider and various

velocity vs. silk length relationships were as follows: Figs. 3 & 8, hatchling theridiid, 0.18 mg, R = 0.90, P <
0.01; Figs. 4 & 9, hatchling theridiid, 0.18 mg, R = 0.88, P < 0.05; Figs. 5 & 10, hatchling theridiid, 0.20 mg,
R = 0.39, P > 0.05; Figs. 6 & 1 1, C/. glomosus, 1.0 mg, R = 0.59, P > 0.05; Figs. 7 & 12, immature theridiid,
1.8 mg, R = 0.49, P > 0.05.

112 THE JOURNAL OF ARACHNOLOGY

0 10 20 30 40 50 60 70

Time (s)

Figure 14. — Drag generated on the surfaces of a 7.2 mg theridiid. The postures shown at three points in the
graph are computer-enhanced images from individual video frames.

characteristics of the surfaces of the spider itself
and its silk. The relevant characteristics of the
surfaces have been discussed in theory by Hum-
phrey (1987), and I have discussed elsewhere
(Suter 1991) their relative importances as de-
duced from wind tunnel experiments. Both I and
Humphrey acknowledged a role of posture in
determining drag, but both of us considered its
role to be secondary, Humphrey because sim-
plifying assumptions needed to be made to make
the calculations tractable, and I because of my
use of dead spiders in the wind tunnel experi-
ments. The results reported here make clear the
need to reassess the role of posture in ballooning.

Figures 3-7 show that individual spiders trail-
ing silk sometimes fall much faster and some-
times much slower than predicted. Because these
large differences can occur between trials of a
single spider (e.g., Figs. 3, 7), they cannot be
attributed to differences in morphology. All of
the silk used by the spiders was dragline, the
structure of which is known and relatively in-
variant (Suter 1991), so that the differences in
velocities cannot be attributed to variations in
silk structure either. Finally, although silk length

is sometimes a significant component in the de-
termination of terminal velocity (Figs. 8, 9), it
by no means acts alone: even at a single silk
length in an individual spider, terminal velocity
can vary by a factor of 2.6 (e.g., Fig. 10: at 5 cm,
78 vs. 29 cm/s). This variability must arise,
therefore, from the single uncontrolled variable
in this system, the posture of the falling spider
relative to the direction of motion.

The influence of posture on the drag developed
by a spider in a moving air stream is clearly
evident in Figs. 13, 14. Even what appears to be
a small change, pulling the legs tightly to the body
from a posture where the legs are slightly more
loosely held, can result in a reduction in drag of
0.17 at a very low air velocity (0.07 m/s),
and that is for a spider with a weight of 1 .76 nN.
At much higher air velocities like those reached
by a falling spider of the same size (e.g., from
Fig. 3, 50 cm/s), the effect of this postural change
would be much greater [using Equation 4 from
Suter ( 1 99 1 ), the reduction in drag due to a com-
parable postural change at 50 cm/s would be 1 . 1 9
mN] and would easily account for the terminal
velocity differences seen in Figs. 8-12.

SUTER- BALLOONING: POSTURE AND TERMINAL VELOCITY

113

The importance of posture as an influence on
terminal velocity in ballooning spiders must, of
course, vary with the amount of silk that is used:
when a spider uses only a short length of silk,
the influence of posture will be greater than when
silk is very long and little of the drag on the silk/
spider system is developed by the spider’s body
and appendages. Unfortunately, almost nothing
is known about the amount of silk actually used
by ballooning spiders. That absence of data means
that the importance of posture in the travels of
ballooning spiders (even in their ability to con-
trol their elevation) cannot yet be assessed.

ACKNOWLEDGMENTS

I am indebted to Katherine J. Suter who as-
sisted in the data collection, to William G. Eber-
hard and Matthew H. Greenstone for their crit-
ical readings of the manuscript, and to Fernando
Nottebohm who generously provided me with
both laboratory and office space at The Rocke-
feller University Field Research Center, Mill-
brook, N. Y. The work reported here was sup-
ported in part by an equipment grant from the
Beadle Fund of Vassar College.

LITERATURE CITED

Coyle, F. A., M. A. Greenstone, A. L. Hultsch & C. E.
Morgan. 1985. Ballooning mygalomorphs: esti-
mates of the masses of Sphodros and Ummidia bal-

looners (Araneae: Atypidae, Ctenizidae). J. Arach-
nol., 13:291-296.

Dean, D. A., & W. L. Sterling. 1985. Size and phe-
nology of ballooning spiders at two locations in east-
ern Texas, USA. J. Arachnol., 13:1 1 1-120.

Eberhard, W. G. 1987. How spiders initiate airborne
lines. J. Arachnol., 15:1-9.

Greenstone, M. H., C. E. Morgan & A. L. Hultsch.
1985. Spider ballooning: development and evalu-
ation of field trapping methods (Araneae). J. Arach-
nol., 13:337-345.

Greenstone, M. H., C. E. Morgan, A. L. Hultsch, R.
A. Farrow &J. E. Dowse. 1987. Ballooning spiders
in Missouri, USA and New South Wales, Australia:
family and mass distributions. J. Arachnol., 15:163-
170.

Humphrey, J. A. C. 1987. Fluid mechanic constraints
on spider ballooning. Oecologia, 73:469-477.

Richter, C. J. J. 1970. Aerial dispersal in relation to
habitat in eight wolf spider species {Pardosa, Ara-
neae, Lycosidae). Oecologia, 5:200-214.

Suter, R. B. 1991. Ballooning in spiders: results of
wind tunnel experiments. Ethol. Ecol. & Evol., 3:
13-25.

Vogel, S. 1981. Life in Moving Fluids. Willard Grant
Press, Boston.

Vugts, H. V., & W. K. R. E. Van Wingerden. 1976.
Meteorological aspects of aeronautic behavior of
spiders. Oikos, 27:433-444.

Manuscript received December 1991, revised February
1992.

1992. The Journal of Arachnology 20:1 14-128

A REVISION OF SOME SPECIES OF RONCUS L. KOCH
(NEOBISIIDAE, PSEUDOSCORPIONES) FROM
NORTH AMERICA AND SOUTH EUROPE

Bozidar P. M. Curcic, Rajko N. Dimitnjevic, and Ozren S. Karamata: Institute of
Zoology, Faculty of Science (Biology), University of Belgrade, Studentski Trg 16,

1 1000 Beograd, Yugoslavia.

ABSTRACT. The available material of the species Roncus lubricus L. Koch, 1873, from North America and
South Europe has been studied. It was concluded that specimens of R. lubricus from the United States belong
to the nominal subspecies. Furthermore, it is assumed that the USA populations of this subspecies were probably
introduced by human activity. A new subspecies, R. lubricus pannonius, from Yugoslavia is described. A key
to the subspecies of R. lubricus is presented.

An analysis of the available type material from the collection of J. Hadzi has supported the elevation of two
of his subspecies to full specific rank: Roncus tenuis Hadzi, 1933, new status, and R. dalmatinus Hadzi, 1933,
new status, both from northern Dalmatia, Y ugoslavia. These two species were formerly regarded as subspecies
of R. lubricus, but this study revealed that they are not members of the R. lubricus group (since they both lack
microsetae proximal to eb and esb). Both species are considered endemics to the Balkan Peninsula. Most
diagnostic characters of the analyzed taxa are thoroughly described or figured. Some taxonomic interrelationships

and features of geographic distribution have been also

The genus Roncus was established in 1873 by
L. Koch for three new species. The first included
species, R. lubricus, was subsequently designated
as the type species (Beier 1 932). The systematics
of the genus is very poorly known. Setation and
morphometric characters are usually employed,
but they may be useful to distinguish between
endemic or cave-inhabiting species. A relative
“homogeneity” of these characters in most epi-
gean species does not permit their use for taxo-
nomic purposes (Gardini 1981, 1983). The so-
called “i?. lubricus" in fact represents a hetero-
geneous group of taxa widely distributed in west-
ern Europe and the northern Mediterranean re-
gion. Therefore, a thorough analysis of different
members of this group seems necessary in order
to establish sound criteria for delimiting the tax-
onomic status of different populations of “i?. lu-
bricus". Of the two existing type specimens of
each sex, Gardini (1983) designated and rede-
scribed the male of R. lubricus from Bloxworth,
United Kingdom, as the lectotype. Both speci-
mens are deposited in the collection of Rev. O.
Pickard-Cambridge in the University Museum
at Oxford. A full account of the external mor-
phology of R. lubricus from near Henley-on-
Thames, UK, was presented by Gabbutt & Va-
chon ( 1 967). Muchmore ( 1 969) collected a num-
ber of R. lubricus specimens from Rochester,

briefly discussed.

N.Y., USA. In 1981, one of us (BPMC) also ob-
tained a sample of the same species from the
USA, but from Cambridge, Massachusetts. Ad-
ditionally, we collected a number of specimens
of R. lubricus or of populations that were con-
sidered as closely related to this species (Curcic
1982, 1991; Poinar & Curcic, 1992), from Mt.
Avala and the village of Obrez, both near Bel-
grade, Yugoslavia (southeastern Europe). In ad-
dition, we analyzed four of the type specimens
of Roncus lubricus tenuis Hadzi, 1933, and R.
lubricus dalmatinus Hadzi, 1933, (two of each
subspecies) in order to verify their taxonomic
status. Both “subspecies” inhabit the northern
Mediterranean region.

The aim of this study was to verify precisely
the taxonomic position (status) of some North
American and South European representatives
assigned to R. lubricus, including those of the
type series of Hadzi’s two subspecies; to con-
tribute to the knowledge of variation between
and among different populations of R. lubricus-,
and to analyze their geographical distribution in
North America and South Europe.

METHODS

The samples from Rochester, Cambridge
(USA), and Obrez (Yugoslavia) were collected
by hand, by sifting leaf-litter and humus, or by

114

CURCIC ET AL.- REVISION OF SOME RONCUS SPECIES

115

Tullgren extractions from leaf litter and soil. Af-
ter collecting, all specimens were preserved in
70% alcohol with 5% glycerine added. The ac-
cessible type specimens of R. lubricus tenuis and
R. lubricus dalmatinus from Hadzi’s collection
(now in our possession) have been subjected to
redescription and further analysis. Measure-
ments (Tables 1-3) have been made in accor-
dance with Chamberlin (1 93 1) and Curcic (1 982,
1988), and the drawings in accordance with Cur-
cic (1982, 1988).

NEOBISIIDAE CHAMBERLIN, 1930

Roncus lubricus lubricus L. Koch, 1873
(Figs. 1-15; Table 1)

Specimens examined.— USA: Massachusetts: Cam-
bridge, 4 males, 8 females, and 1 tritonymph (collected
from leaf litter and humus on the grounds of the Amer-
ican Academy of Arts and Sciences), 18 July 1981 (B.
P. M. Curcic and P. S. Petrovic); New York: Monroe
Co., Rochester, 8 males and 6 females, (collected from
leaf litter and soil along base of wall shaded by syca-
more tree), 24 September 1967 (W. B. Muchmore).

Description. — Carapace somewhat longer than
broad (Fig. 1; Table 1). Epistome triangular and
pointed or apically blunt (Figs. 2, 3). Eyes well
developed. Anterior row with 4, ocular with 5-
7, median and intermedian rows each with 6-8,
and posterior row with 6 (rarely 5) setae.

Tergite I with 6 or 7 setae, tergite II with 8-1 1
setae, and tergites III-X each with 9-13 setae.
Male genital area; stemite II with a cluster of 1 4-
19 setae; of these, 8-10 longer setae on the pos-
terior sternal margin and the remainder in the
median and posterior region, thinning out an-
teriorly. Stemite III with 5-7 anterior and me-
dian setae, 9-1 1 posterior setae, and 3 supras-
tigmal microsetae on each side. Female genital
area: stemite II with a cluster of 9-15 setae in
form of an irregular triangle or circle; stemite III
with a transverse row of 1 1-15 posterior setae,
and 3 small setae along each stigma. Trito-
nymph: stemite II with 2 setae only, stemite III
with 9 setae, and 2 or 3 microsetae along each
stigma. On stemite IV, three microsetae along
each stigmatic plate in both adults and trito-
nymph. Stemite IV with 7 (tritonymph) or 10-
15 setae (adult); stemites V-X each with 11-15
setae. From stemite VII onwards, 2 median setae
slightly anterior to the row of marginal setae.
Cheliceral spinneret (galea) as a low hyaline con-
vexity (Fig. 4), slightly less prominent in males
than in females or tritonymph. Cheliceral palm

with 6 setae in both sexes and tritonymph, mov-
able finger with one seta. Fixed cheliceral finger
with 15-17 small teeth diminishing in size both
distally and proximally. Movable cheliceral fin-
ger with 10-13 teeth (Fig. 4). Flagellum with 1
short proximal blade and 7 longer blades distally.
Apex of pedipalpal coxa with 4 long, acuminate
setae. Pedipalpal trochanter with two small lat-
eral tubercles and with inconspicuous granula-
tions dorsally; femur with a single lateral tuber-
cle, granulation as in Figs. 12, 13, 15. Tibia
smooth. Chelal palm with some inconspicuous
granulations (Figs. 12, 13, 15); a group of 1 to 4
microsetae proximal to trichobothria eb and esb
(Figs. 9-1 1); a single tubercle on laterodistal side.
Fixed chelal finger with 57-64 (female), 59-68
(male), and 40 teeth (tritonymph); distal teeth
asymmetrical, followed by small, close-set, and
retroconical teeth. Movable chelal linger with 57-
62 female), 57-68 (male), and 42 teeth (trito-
nymph); only distal teeth pointed and retrocon-
ical, and these becoming rounded or square-
cusped teeth which extend as far as the level of
b. Sensillum between 1 6th and 22nd (male), be-
tween 12th and 20th (female), or at level of 1 1th
tooth (tritonymph). Sensillum in male is either
distal to or at level of sb\ in female, slightly prox-
imal to, distal to, or at the level of sb; and in
tritonymph, slightly proximal to st.

Chelal fingers longer than chelal palm and
slightly shorter than or equal to pedipalpal femur
(Table 1). Pedipalpal femur length almost equal
to carapacal length (Table 1). Trichobothriotaxy:
ist somewhat closer to est than to isb-, sb equi-
distant from b and st (Figs. 5-7). Trichobothrial
pattern of tritonymph as in Fig. 8. (Trichoboth-
riotaxy almost identical to R. lubricus from the
UK (Gabbutt & Vachon 1967)). Trochanteral
foramen pointed and sclerotized. Tibia IV, ba-
sitarsus IV, and telotarsus IV each with a long
tactile seta (Fig. 14). Tactile seta ratios, mor-
phometric ratios and linear measurements are
given in Table 1.

Distribution. — It is still impossible to outline
the true distribution of Roncus lubricus (Gardini
1983). Gardini suggested that in Europe this spe-
cies may be widespread in the United Kingdom,
France, and Belgium. In North America, R. lu-
bricus is distributed in the eastern United States.
Its presence in New York and Massachusetts
supports the assumption that it is a permanent
inhabitant of these regions. The findings of sub-
adult stages in the localities analyzed (Muchmore
1969) strongly confirm its ecological adaptability

116

THE JOURNAL OF ARACHNOLOGY

Figures \-i.—Roncus lubricus lubricus L. Koch, 1873, from Cambridge, Mass., USA: 1, carapace, male; 2,
epistome, male; 3, epistome, tritonymph; 4, cheliceral fingers, female; 5, pedipalpal chela, male; 6, pedipalpal
chela, male; 7, pedipalpal chela, female; 8, pedipalpal chela, tritonymph. Scales in mm.

as well as the presence of fully established life
cycle.

Interrelationships of R. lubricus from North
America and Great Britain.— The comparison of
measurements of various structures of R. lubri-
cus from the United States with the data pre-
sented by Gabbutt & Vachon (1967) and Gardini
(1983) for British specimens yielded some inter-
esting observations. North American specimens
show some minor differences in a number of
linear measurements and linear measurements
and morphometric ratios. However, some more
important differences have been noted in a num-
ber of characters, including the ratio of chelal
finger length to the chelal palm length which is

1.33=1.43 (females), 1.31-1.57 (males), and 1.29-
1 .47 (tritonymph) for UK specimens, as opposed
to 1.08-1.29 (females), 1.16-1.345 (males), and
1.19 (tritonymph) for US specimens; and the of
pedipalpal tibial length to breadth is 1.83-2.14
for females from the UK, as opposed to 2.14-
2.33 for US females. However, the majority of
other morphometric characters are close or iden-
tical.

Some slight differences between specimens
from UK and US are also manifested in the form
of pedipalpal articles; in general, the body size
and proportions of different structures are some-
what greater in US specimens. The noted small
distinctions in setation and other characters

CURCIC ET AL.-REVISION OF SOME RONCUS SPECIES

117

Figures 9-\5.—Roncus lubricus lubricus L. Koch, 1873, from Cambridge, Mass., USA: 9, microsetae proximal
to eb and esb, male; 1 0, microsetae proximal to eb and esb, male; 1 1 , microsetae proximal to b and esb, tritonymph;
12, pedipalp, female; 13, pedipalp, male; 14, leg IV, male; 15, pedipalp, tritonymph. Scales in mm.

Table 1.— Linear measurements (in mm) and morphometric ratios in Roncus lubricus lubricus L. Koch from
the US.

Character

Females

Males

Trito.

Body

Length (1)

1.75-1.94

1.84-2.995

1.97

Cephalothorax

Length (2)

0.56-0.73

0.62-4).74

0.46

Breadth

0.51-0.63

0.52-0.63

0.445

Abdomen

Length

1.03-2.29

1.17-2.33

1.51

Breadth

0.69-1.23

0.62-1.03

0.82

Cheiicerae

Length (3)

0.39-0.46

0.38-0.41

0.28

Breadth (4)

0.20-0.24

0.18-0.23

0.16

Length of movable finger (5)

0.26-0.32

0.25-0.29

0.20

Length of galea

0.01

0.01

0.003

Pedipalps

Length with coxa (6)

3.15-3.66

3.10-3.605

2.17

Ratio 6/1

1.10-2.03

1.08-1.81

1.10

Length of coxa

0.49-0.58

0.47-0.555

0.35

Length of trochanter

0.38-0.46

0.38-0.44

0.27

Length of femur (7)

0.46-0.74

0.63-4).73

0.43

Breadth of femur (8)

0.17-0.22

0.165-0.195

0.14

Ratio 7/8

3.285-4.00

3.46-4.29

3.07

Ratio 7/2

0.945-1.14

0.93-1.11

0.93

Length of tibia (9)

0.50-0.57

0.50-0.59

0.33

Breadth of tibia (10)

0.22-0.255

0.21-0.24

0.16

Ratio 9/10

2.14-2.33

2.22-2.46

2.06

Length of chela (11)

1.11-1.34

1.10-1.30

0.79

Breadth of chela (12)

0.31^.37

0.30-0.36

0.22

Ratio 11/12

3.35-3.72

3.50-4.07

3.59

Length of chelal palm ( 1 3)

0.51-0.61

0.48-0.58

0.36

Ratio 13/12

1.53-1.68

1.60-1.77

1.64

Length of chelal finger (14)

0.59-0.74

0.62-0.74

0.43

Ratio 14/13

1.08-1.29

1.16-1.345

1.19

Leg IV

Total length

2.18-2.53

2.205-2.495

1.57

Length of coxa

0.32-0.42

0.36-0.425

0.25

Length of trochanter ( 1 5)

0.26-0.315

0.27-0.31

0.18

Breadth of trochanter ( 1 6)

0.12-0.16

0.12-0.14

0.10

Ratio 15/16

1.93-2.42

2.14-2.38

1.80

Length of femur (17)

0.58-0.67

0.55-0.62

0.41

Breadth of femur (18)

0.20-0.23

0.185-0.23

0.13

Ratio 17/18

2.55-3.12

2.695-3.08

3.15

Length of tibia (19)

0.47-0.60

0.50-0.58

0.34

Breadth of tibia (20)

0.09-4). 12

0.10-0.12

0.10

Ratio 19/20

4.58-5.555

4.83-5.40

3.40

Length of basi tarsus (21)

0.185-0.21

0.18-0.21

0.14

Breadth of basitarsus (22)

0.07-0.09

0.07-0.08

0.06

Ratio 21/22

2.28-2.86

2.375-2.67

2.33

Length of telotarsus (23)

0.31-0.38

0.32-0.37

0.25

Breadth of telotarsus (24)

0.07-0.085

0.06-0.075

0.065

Ratio 23/24

4.00-5.43

5.14-5.67

3.85

TS ratio— tibia IV

0.56-0.64

0.50-0.62

0.59

TS ratio— basitarsus IV

0.17-0.25

0.17-0.20

0.275

TS ratio— telotarsus IV

0.34-0.41

0.34-0.43

0.39

CURCIC ET AL.- REVISION OF SOME RONCUS SPECIES

119

should be treated as the result of intraspecific
variability. One more item should be mentioned
here — the trans- Atlantic distribution of R. lu-
bricus. The explanation for such a distributional
pattern was presented elsewhere by Muchmore
(1969). This author suggested that the presence
of this species in the United States may be due
to its recent introduction by human activity.

Rmcus lubricus pannonius, new subspecies
(Figs. 16-30; Table 2)

aff. lubricus L. Koch, 1873; Curcic 1991:165.

Etymology.— After Pannonii (sing. Pannon-
ius), a group of Illyrian people, inhabiting Pan-
nonia, the territory in the Save Valley (Divkovic
1987); the type locality of this subspecies is lo-
cated in the Pannonian Plain.

Specimens examined.— Holotype male, allotype fe-
male, 9 paratype males, and 5 paratype females, from
the village of Obrez, near Belgrade, Serbia, Yugoslavia,
collected from leaf litter and humus in an oak forest,
over a period from May 1989 to September 1990 (B.
P. M. Curcic, R. N. Dimitrijevic, O. S. Karamata and
L. R. Lucic); deposited in the collections of the Institute
of Zoology, Faculty of Science, University of Belgrade,
Belgrade.

Description. — Epistome triangular, apically
pointed (Figs. 18, 19, 21). Eyes with flattened
lenses. Setal formulae: 4-I-6 + 2 + 4-I-2 + 6
= 24, and 4 + 6 + 2 + 3 + 2-1-6 = 23 setae
(the latter disposition of setae present in only one
specimen of each sex).

Tergite I with 6 or 7 setae, tergite II with 7-
1 1 setae, the subsequent tergites (III-X) each with
8-12 setae. Male genital area: stemite II with a
cluster of 14-21 setae; of these, 9-12 longer setae
are found along the posterior sternal margin and
remainder are smaller and thinning out anteri-
orly. Stemite III with 4—8 anterior and median
setae, 8-12 posterior setae and 3 (rarely 2, 4, or
5 ) microsetae along each stigma. Stemite IV with
3 (rarely 2 or 4) suprastigmal microsetae on ei-
ther side and 8-1 1 posterior setae. Female genital
area: stemite II with 7-14 small setae arranged
in form of two barely distinguishable groups;
stemite III with a transverse row of 10-14 setae
and 3 suprastigmal microsetae on each side. Ster-
nite IV with a row of 8-10 posterior setae and 3
microsetae along each stigma. Stemites V-X each
with 12-16 setae. Setae on stemites VII-X ar-
ranged in transverse rows, with no anterior and
median setae present.

The cheliceral spinneret (galea) as a low hya-

line convexity, slightly more prominent in fe-
males than in males. Cheliceral palm with 6 (in
both sexes), and the movable finger with only 1
seta. Fixed cheliceral finger with 18-23 small
teeth; movable cheliceral finger with 12-17 teeth
(Fig. 21). Flagellum with 1 short proximal blade
and 7 longer blades distally, all blades denticu-
late. Apex of pedipalpal coxa with 4 long setae.
Pedipalpal trochanter with 2 small lateral tuber-
cles and some inconspicuous denticulations dor-
sally, pedipalpal femur with a small lateral tu-
bercle and distinct granulations as in Figs. 29,
30. Tibia with some rare and inconspicuous el-
evations interiorly and laterally, or smooth. Che-
lal palm with distinct granulations interolaterally
and small and almost flattened elevations on its
exterolateral side. A patch of 1-4 microsetae
present proximal to eb and esb] also, 1-6 addi-
tional microsetae developed lateral and distal to
eb and esb. A tiny tubercle present on laterodistal
side of chelal palm. Fixed chelal finger with 55-

63 (male), and 52-63 teeth (female); distal teeth
of this finger are pointed and asymmetrical, fol-
lowed by small, close-set, and retroconical teeth.
Movable chelal finger with 57-63 (male) or 51-

64 teeth (female); only distal teeth pointed and
retroconical, these gradually give way to rounded
or square-cusped teeth which stretch as far as the
level of b. Sensillum distal to sb, between 17th
and 24th tooth (male), or between 1 5th and 22nd
tooth (female). Chelal fingers longer than chelal
palm and slightly shorter than pedipalpal femur.
Pedipalpal femur slightly shorter than carapace
(Table 2).

Trichobothriotaxy: ist slightly closer to est than
to isb, sb slightly closer to b than to si (Figs. 22-
24). Tibia IV, basitarsus IV, and telotarsus IV
each with a single tactile seta (Fig. 28). Tactile
seta ratios, morphometric ratios and linear mea-
surements are presented in Table 2.

Distribution.— Yugoslavia (Pannonian Plain),
epigean (in humus and leaf litter).

Interrelationships of R. lubricus pannonius and

the nominal subspecies.— As far as the morpho-
metric ratios and linear measurements are con-
cerned, the new subspecies can be easily distin-
guished from the nominal subspecies (inhabiting
western Europe and North America). In spite of
the fact that the majority of these characters are
either very close in both subspecies, or their val-
ues overlap, there exist a number of features which
are distinct in both subspecies. For instance, the
carapacal length in R. lubricus lubricus (the val-

120

THE JOURNAL OF ARACHNOLOGY

16 17

Figures \(y-2A.—Roncus lubrkus pannonius n. ssp., from Obrez, near Belgrade, Yugoslavia: 16, carapace,
male; 17, carapace, female; 18, epistome, male; 19, epistome, male; 20, epistome, female; 21, cheliceral fingers,
female; 22, pedipalpal chela, male; 23, pedipalpal chela, male; 24, pedipalpal chela, female. Scales in mm.

CURCIC ET AL.- REVISION OF SOME RONCUS SPECIES

121

0,1

I I

Figures 25-30. — Romchs lubricus pannonius n. ssp., from Obrez, near Belgrade, Yugoslavia: 25, microsetae
proximal to b and esb, male; 26, microsetae proximal to eb and esb, male; 27, microsetae proximal to eb and
esb, female; 28, leg IV, male; 29, pedipalp, female; 30, pedipalp, male. Scales in mm.

Table 2.— Linear measurements (in mm) and morphometric ratios in Roncus lubricus pannonius new sub-
species from Yugoslavia.

Character

Females

Males

Body

Length (1)

2.73-3.66

2.48-3.06

Cephalothorax

Length (2)

0.72-0.88

0.72-0.81

Breadth

0.60-0.69

0.58-0.63

Abdomen

Length

1.99-2.78

1.85-2.40

Breadth

0.75-1.10

0.82-1.17

Chelicerae

Length (3)

0.445-0.50

0.42-0.46

Breadth (4)

0.23-0.25

0.205-0.23

Length of movable finger (5)

0.315-0.36

0.29-0.315

Length of galea

0.01

0.01

Pedipalps

Length with coxa (6)

3.65^.015

3.51-3.89

Ratio 6/1

1.10-1.40

1.15-1.455

Length of coxa

0.58-0.63

0.50-0.60

Length of trochanter

0.425-0.47

0.42-0.47

Length of femur (7)

0.72-0.80

0.69-0.795

Breadth of femur (8)

0.21-0.24

0.21-0.23

Ratio 7/8

3.13-3.52

3.26-3.785

Ratio 7/2

0.91-1.04

0.925-1.09

Length of tibia (9)

0.60-0.675

0.59-0.66

Breadth of tibia ( 1 0)

0.25-0.30

0.27-0.29

Ratio 9/10

2.10-2.44

2.185-2.37

Length of chela (11)

1.28-1.44

1.26-1.41

Breadth of chela (12)

0.38-0.45

0.37-0.41

Ratio 11/12

2.93-3.37

3.191-3.51

Length of chelal palm (13)

0.59-0.69

0.57-0.665

Ratio 13/12

1.38-1.55

1.46-1.62

Length of chelal finger ( 1 4)

0.69-0.75

0.68-0.745

Ratio 14/13

1.09-1.17

1.12-1.28

Leg IV

Total length

2.67-2.84

2.56-2.775

Length of coxa

0.43-0.48

0.40-0.49

Length of trochanter (15)

0.32-0.36

0.29-0.34

Breadth of trochanter (16)

0.15-0.18

0.15-0.17

Ratio 15/16

1.89-2.19

1.88-2.27

Length of femur ( 1 7)

0.69-0.74

0.66-0.72

Breadth of femur ( 1 8)

0.25-0.26

0.22-0.26

Ratio 17/18

2.73-2.92

2.64-3.045

Length of tibia (19)

0.62-0.64

0.59-0.68

Breadth of tibia (20)

0.12-0.14

0.12-0.14

Ratio 19/20

4.57-5.17

4.615-5.58

Length of basitarsus (2 1 )

0.22-0.23

0.21-0.25

Breadth of basitarsus (22)

0.09-0.10

0.08-0.09

Ratio 21/22

2.20-2.555

2.33-2.875

Length of telotarsus (23)

0.38-0.40

0.34-0.38

Breadth of telotarsus (24)

0.08

0.075-0.08

Ratio 23/24

4.75-5.00

4.25-4.93

TS ratio— tibia IV

0.505-0.60

0.51-0.63

TS ratio— basitarsus IV

0.15-0.21

0.18-0.205

TS ratio- telotarsus IV

0.37-0.40

0.35-0.48

CURCIC ET AL.- REVISION OF SOME RONCUS SPECIES

123

ues for R. lubricus pannonius n. ssp. are enclosed
in parentheses) is 0.56-0.73 mm for females, and
0.56-0.74 mm for males (0.72-0.88 mm and
0.72-0.8 1 mm); the cheliceral length is 0.37-0.46
mm for females and 0.34-0.41 mm for males
(0.445-0.50 mm and 0.42-0.47 mm); the pedi-
palpal tibia length is 0.45-0.57 mm in females
and 0.50-0.59 mm in males (0.60-0.675 mm and
0.59-0.66 mm); the chelal palm breadth is 0.27-
0.37 mm in females and 0.27-0.36 mm in males
(0.38-0.45 mm and 0.37-0.41 mm), etc. In gen-
eral, the appendages and other body structures
in R. lubricus pannonius n. ssp. are larger than
in the nominal subspecies.

Apart from these morphometric distinctions,
there exist some qualitative differences between
the two subspecies, such as those manifested in
the disposition of some trichobothria on both
fixed and movable chelal fingers. For example,
in R. lubricus lubricus, the trichobothrium ist is
equidistant from est and isb, while in R. lubricus
pannonius n. ssp. it is slightly closer to est than
to isb. Furthermore, sb is equidistant from st and
b in the nominal subspecies, while in R. lubricus
pannonius n. ssp. it is closer to b than to st.

Interestingly enough, it is obvious that R. lu-
bricus pannonius n. ssp. belongs to the Balkan
endemic and relict fauna. The presence of this
form, as well as of some other non-cavemicolous
endemics (Curcic, in press), supports the fact that
the endemic differentiation in the Balkans has
been taking place not only in subterranean (Cur-
cic 1988) but also in epigean habitats.

KEY TO THE SUBSPECIES OF
RONCUS LUBRICUS

1. Trichobothrium ist equidistant from est and
isb, and sb equidistant from st and b. Pedi-
palpal tibial length 0.45-0.57 mm (females)
and 0.50-0.59 mm (males). Western Europe

and North America R. lubricus lubricus

Trichobothrium ist closer to est than to isb,
and sb closer to b than to st. Pedipalpal tibial
length 0.60-0.675 mm (females) and 0.59-
0.66 mm (males). Southeastern Europe . . .

R. lubricus pannonius, n. ssp.

Roncus tenuis Hadzi, new status
(Figs. 31-41; Table 3)

Roncus (Roncus) lubricus tenuis H&dii, 1933:166-170.

Specimens examined.— The type series consists of
two specimens (one of each sex); neither of these was
designated as the holotype by Hadzi (1933). Therefore,
we hereby designate the syntype male as the lectotype

and the syntype female as a paralectotype. The lecto-
type is mounted on a slide with the label “Roncus, 6
Malinska, 1929”. The paralectotype is mounted on a
separate slide and labelled “Roncus (Roncus), 9 17. IV
1927, under stone, Malinska”. This locality is situated
on the Island of Krk, in the northern Adriatic region
(Dalmatia), Yugoslavia.

Description.— Epistome tubercular (Fig. 33) or
triangular (Fig. 34). Eyes with almost flattened
lenses (Figs. 31, 32). Setal formulae: 4 + 6 + 2
+ 4 + 2 + 6 = 24 and 4 + 6 + 4 + 2 + 6 =
22. Tergites I-X bearing 6-10-1 1-1 1-10-10-10-
10-10-9 and 6-8-10-10-1 1-12-10-1 1-1 1-9 se-
tae. Male genital area: stemite II with 1 8 median
and posterior setae; of these, 10 setae along pos-
terior sternal margin and the remainder (8 setae)
medial, thinning out anteriorly. Stemite III with
3 or 4 microsetae on each side, 3 anterior and
median setae, and a transverse row of 1 0 setae.
Female genital area: stemite II with 7 small me-
dian and posterior setae of irregular distribution.
Stemite III with 9 posterior setae and 3 micro-
setae along each stigma. Cheliceral spinneret low
and flattened, but somewhat more prominent in
female than in male (Figs. 35, 36). Fixed cheli-
ceral finger with 15-18 small teeth. Movable
cheliceral finger with 1 1 or 1 2 teeth. Fixed chel-
iceral finger with 6, movable finger with 1 seta.
Flagellum with 1 short proximal blade and 7 or
8 longer blades distally. Apex of pedipalpal coxa
of the lectotype with 4 long setae (the paralec-
totype has 3 setae on the right, and 4 on the left
apex). Pedipalpal trochanter with a small tuber-
cle and some inconspicuous denticulations dor-
sally (Figs. 39, 40). Pedipalpal femur with an
extero-lateral tubercle and an intero-basal tu-
bercle; surface with granulations on its intero-
lateral side. Tibia tulip-shaped and smooth. Che-
lal palm with interior and lateral granulations
(Figs. 39, 40). Fixed chelal finger with 52-54 teeth,
and movable finger with 55-56 teeth. Only distal
teeth on the latter finger retroconical and point-
ed; these gradually merging into rounded or
square-cusped teeth which reach the level of b.
The patch of microsetae proximal to eb and esb
absent; a single tubercle on laterodistal side of
chelal palm evident. Sensillum at the level of
1 3th (male) or 1 8th tooth (female), either prox-
imal (male) or distal to sb (female).

Chelal palm ovate (dorsal view). Chelal fingers
only slightly longer than (male) or equal to chelal
palm (female), but shorter than pedipalpal femur
(Table 3). Pedipalpal femur length almost equal
to carapacal length (Table 3). Disposition of tri-

Figures 31 -4 tenuis Hadzi, new status, from Malinska, Yugoslavia: 31, carapace, male; 32, carapace,

female; 33, epistome, male; 34, epistome, female; 35, cheliceral fingers, female; 36, cheliceral fingers, male; 37,
pedipalpal chela, female; 38, leg IV, female; 39, pedipalp, male; 40, pedipalp, female; 41, pedipalpal chela, male.
Scales in mm.

CURCIC ET AL.-- REVISION OF SOME RONCUS SPECIES

125

chobothria: ist slightly closer to est than to isb
(or equidistant from these); seta sb equidistant
from b and st. The trichobothrial pattern as in
Figs. 37, 41. Trochanteral foramen pointed and
heavily sclerotized. Tibia IV, basitarsus IV, and
telotarsus IV each with a single tactile seta (Fig.
38). Tactile seta ratios are presented in Table 3.

Morphometric ratios and linear measurements
are presented in Table 3.

Distribution.— This species seems to be re-
stricted to the northern Adriatic region in Yu-
goslavia. To date, it is known from its type lo-
cality only (Malinska, Island of Krk).

Interrelationships of E. tenuis and R.. lubri-
cas.— There is no doubt that R. tenuis should be
given full specific rank. The differences between
R. tenuis and R. lubricus are clearly manifested
in many morphometric ratios and linear mea-
surements (see Tables 1, 3). Furthermore, the
two species differ in the presence/absence of a
patch of microsetae proximal to eb and esb (pres-
ent in lubricus, absent in tenuis), in the form of
the pedipalpal articles (attenuated in lubricus,
stout in tenuis), in the presence/absence of an
intero-basal tubercle on pedipalpal femur (pres-
ent in tenuis, absent in lubricus), in the form of
the epistome (small and tubercular in tenuis, long
and triangular in lubricus), in the ratio of chelal
finger length to chelal palm length (lower in ten-
uis, higher in lubricus), and in body size (smaller
in tenuis, greater in lubricus).

A thorough analysis of the external morphol-
ogy and biogeography of both taxa suggests that
R. tenuis does not belong to the R. lubricus group
of species (m'hose members have microsetae
proximal to eb and esb). It seems that this species
belongs to another species group, whose origin
is derived from the eastern Mediterranean or
Balkanic pseudoscorpion fauna.

Roncus dalmatinus Hadzi, new status
(Figs. 42-50; Table 3)

Roncus (Roncus) lubricus dalmatinus Hadzi, 1933: 1 70-

175.

Specimens examined.— According to Hadzi (1933),
the type series of this taxon consisted of “specimens”
collected in Omisalj on the Island of Krk, and in Split
(Meje), both in Yugoslavia. We have studied two males
from the type series, both from Split, Mt. Maqan (Meje).
These specimens are the only available syntypes. They
are both mounted on a slide labelled '‘Obisium {Ron-
cus) lubricus, Split-Maijan, in the vicinity of the pen-
sion "Split", Meje, 28.10-14. IV 1927”. The other type

specimens, if any, seem to be lost or damaged. There-
fore, we hereby designate the male specimen labelled
“ 1 ” as the lectotype and the male labelled “2” as the
paralectotype of this taxon.

Description. — Epistome triangular and point-
ed (Figs. 47, 48). Eyes small and with almost
flattened lenses (Figs. 42, 43). Setal formulae: 4
-i-6-l-2-i-4 + 2 + 7 = 25 and 4 + 6 + 2 +
4 + 2 + 8 = 26.

Tergite I with 6-10 setae, the following tergites
(III-X) each with 10-13 setae (mostly 1 1). Male
genital area: stemite II with 15-17 median and
posterior setae; of these, 10 longer setae along
posterior sternal margin and the remainder me-
dial and posteriorad, thinning out anteriorly.
Stemite III with 3 small setae along each stigma,
3-5 anterior and median setae and 9 or 10 pos-
terior setae. Stemite IV with 10 posterior setae
and 3 small setae along each stigmatic plate. Fe-
male genital area: unknown. However, according
to Hadzi (1933), stemite II of the female with
1 3 small median and posterior setae, and stemite
III v/ith 3 suprastigmatic setae on each side and
a posterior row of 14 setae; stemite IV with a
row of 9 setae and 3 small setae along each stig-
ma. Unfortunately, this female specimen was not
available, due to its probable loss or destruction.
In males, stemites IV-X each v/ith 12-15 setae
(mostly 13).

Cheliceral spinneret (galea) low and flattened
(Fig. 44). Fixed cheliceral finger with 18, and
movable finger with 1 5 small teeth. Flagellum of
8 or 9 blades, with 1 or 2 short proximal blades
and 7 longer blades distally (Fig. 49).

Apex of pedipalpal coxa (manducatory pro-
cess) with 4 long setae. Pedipalpal trochanter with
a small tubercle and inconspicuous denticula-
tions dorsally (Figs. 46, 50). Pedipalpal femur
with an extero-lateral tubercle and a pair of in-
terior and basal prominent tubercles; surface of
this podomere granulated interiorly and dorsally
(Figs. 46, 50). Pedipalpal tibia elongated and tu-
lip-shaped, with few inconspicuous granulations
interiorly and distally. Chelal palm ovate (dorsal
view), with some interior and exterior granula-
tions laterally (Figs. 46, 50). Fixed chelal finger
with 66-68 small teeth, and movable chelal fin-
ger with 60-62 teeth. Only distal teeth of the
latter finger retroconical and pointed, gradually
merging into rounded or square-topped teeth
which reach the level of b. Sensillum at level of
20th to 26th tooth (at the level of sb or slightly
distal to this trichobothrium).

A patch of microsetae proximal to eb and esb

Table 3. —Linear measurements (in mm) and morphometric ratios in Roncus lubricus tenuis Hadzi, new status,
and Roncus dalmatinus Hadzi, new status, from Yugoslavia.

R. tenuis

R. dalmatinus

Character

Male

Females

Males

Body

Length ( 1 )

2.19

2.49

2.86-2.91

Cephalothorax

Length (2)

0.61

0.67

0.85-0.87

Breadth

0.54

0.63

0.65-0.665

Abdomen

Length

1.58

1.82

1.99-2.06

Breadth

0.62

0.75

0.99-1.03

Chelicerae

Length (3)

0.37

0.41

0.47-0.48

Breadth (4)

0.195

0.21

0.23-0.24

Length of movable finger (5)

0.25

0.29

0.31-0.32

Length of galea

0.005

0.01

0.01

Pedipalps

Length with coxa (6)

3.21

3.28

4.31-4.375

Ratio 6/1

1.465

1.32

1.50-1.51

Length of coxa

0.50

0.51

0.65-0.73

Length of trochanter

0.39

0.39

0.51-0.555

Length of femur (7)

0.63

0.67

0.85-0.88

Breadth of femur (8)

0.195

0.21

0.255-0.26

Ratio 7/8

3.23

3.19

3.27-3.45

Ratio 7/2

1.03

1.00

0.98-1.035

Length of tibia (9)

0.54

0.55

0.74-0.75

Breadth of tibia (10)

0.24

0.25

0.34

Ratio 9/10

2.25

2.20

2.18-2.205

Length of chela (11)

1.15

1.16

1.48-1.54

Breadth of chela ( 1 2)

0.35

0.38

0.47-0.50

Ratio 11/12

3.285

3.05

2.96-3.28

Length of chelal palm (13)

0.57

0.58

0.74-0.78

Ratio 13/12

1.63

1.53

1.48-1.66

Length of chelal finger (14)

0.58

0.58

0.74-0.76

Ratio 14/13

1.02

1.00

0.97-1.00

Leg IV

Total length

2.355

2.385

2.67

Length of coxa

0.38

0.39

0.45

Length of trochanter (15)

0.31

0.32

0.34

Breadth of trochanter ( 1 6)

0.15

0.15

0.15

Ratio 15/16

2.07

2.13

2.27

Length of femur (17)

0.59

0.61

0.69

Breadth of femur (18)

0.22

0.22

0.25

Ratio 17/18

2.68

2.77

2.76

Length of tibia ( 1 9)

0.55

0.545

0.63

Breadth of tibia (20)

0.11

0.11

0.14

Ratio 19/20

5.00

4.95

4.50

Length of basitarsus (21)

0.195

0.20

0.23

Breadth of basitarsus (22)

0.10

0.08

0.10

Ratio 21/22

1.95

2.50

2.30

Length of telotarsus (23)

0.33

0.32

0.33

Breadth of telotarsus (24)

0.07

0.08

0.08

Ratio 23/24

4.71

4.00

4.125

TS ratio— tibia IV

0.60

0.58

0.58

TS ratio— basitarsus IV

0.21

0.21

0.19

TS ratio— telotarsus IV

0.33

0.30

0.35

Figures 42-50.— Roncus dalmatinus Hadzi, new status, from Mt. Marjan (Meje), Split, Yugoslavia: 42, car-
apace, male; 43, carapace, male; 44, cheliceral fingers, male; 45, pedipalpal chela, male; 46, pedipalpal trochanter,
femur and tibia, male; 47, epistome, male; 48, epistome, male; 49, flagellum, male; 50, pedipalp, male. Scales

128

THE JOURNAL OF ARACHNOLOGY

absent; a single tubercle on laterodistal side of
chelal palm evident. Chelal fingers as long as the
chelal palm (Table 3), but shorter than pedipalpal
femur. Pedipalpal femur length almost equal to
carapacal length (Table 3).

Disposition of trichobothria: ist slightly closer
to est than to isb; seta sb only slightly closer to
b than to st (or equidistant from these) (Fig. 45).
Trochanteral foramen pointed and heavily scler-
otized. Tibia IV, basitarsus IV, and telotarsus IV
each with a single tactile seta; tactile seta ratios
as in Table 3. Morphometric ratios and linear
measurements are presented in Table 3.

Distribution.— According to Hadzi (1933), this
species is present on the Island of Krk and at Mt.
Marjan, near Split (Dalmatia), Yugoslavia. How-
ever, the available type material at our disposal
comes from the vicinity of Split only. Therefore,
this species is now known from Middle Dalmatia
only, although it may be widespread on some
Adriatic islands.

Interrelationships of R. dalmatinus, R. lubri-
cus, and R. tenuis. — From R. lubricus, R. dal-
matinus is easily distinguished by the form of
the carapace (see also Tables 1 , 3), by the setation
of the posterior carapacal margin (6 setae in lu-
bricus, 7-8 in dalmatinus), by the form of pe-
dipalpal articles (more stout in dalmatinus, more
slender in lubricus), by the presence/absence of
two sclerotic knobs at the interior and basal re-
gion of the pedipalpal femur (present in dal-
matinus, absent in lubricus), by the presence/
absence of microsetae proximal to eb and esb
(present in lubricus, absent in dalmatinus), as
well as by some morphometric ratios and linear
measurements (Tables 1, 3).

The distinctions between R. dalmatinus and
R.tenuis are clearly manifested in the form and
size of the epistome (small and tubercular in ten-
uis, large and triangular in dalmatinus), in the
proportions of the carapace (only slightly longer
than wide in tenuis, considerably longer than wide
in dalmatinus), in the setation of the posterior
carapacal row (6 setae in tenuis, 7-8 in dalma-
tinus), in the form of the pedipalpal articles (more
stout in tenuis, more slender in dalmatinus), in
the number of sclerotic knobs on the interobasal
part of the pedipalpal femur (one in tenuis, two
in dalmatinus), as well as in some morphometric
ratios and linear measurements. According to the
absence of microsetae proximal to eb and esb, as
well as to the presence of sclerotic knobs on the
interior and lateral surface of pedipalpal femora.

it can be assumed that both R tenuis and R.
dalmatinus pertain to the same species group,
which is clearly distinct from R. lubricus and its
allies.

ACKNOWLEDGMENTS

Our sincere thanks are due to W. B. Much-
more, V. F. Lee, and G. Gardini, who offered
many valuable comments and suggestions for the
improvement of the text.

LITERATURE CITED

Beier, M. 1932. Pseudoscorpionidea. 1. Subord.

Chthoniinea et Neobisiinea. Tierreich, 57:1-254.
Chamberlin, J. C. 1931. The arachnid order Chelo-
nethida. Stanford Univ. Publ. (Biol.), 7:1-284.
Curcic, B. P. M. 1982. Postembryonic development
in the Neobisiidae (Pseudoscorpiones, Arachnida).
Serbian Acad. Sci. Arts, Monogr. 545, Dept. Sci.,
56:1-90.

Curcic, B. P. M. 1988. Cave-dwelling pseudoscor-
pions of the Dinaric Karst. Acad. Slov. Sci. Arts,
Opera, 26, Inst. Biol., 8:1-191.

Curcic, B. P. M. 1991. A new species of the genus
Roncus L. Koch, 1873 (Neobisiidae, Pseudoscor-
piones) from East Serbia. Mem. BiospeleoL, 18:165-
169.

Curcic, B. P. M. in press. Cave pseudoscorpions of

East Serbia. Serbian Acad. Sci. Arts, Monogr., Dept.

Sci.

Divkovic, M. 1987. Latinsko-hrvatski rijecnik za
skole. 4th ed. (reprint). Zagreb, Naprijed, 1-1 161.
Gabbutt, P. D. & M. Vachon. 1967. The external
morphology and life history of the pseudoscorpion
Roncus lubricus. J. Zool. London, 153:475-498.
Gardini, G. 1981. Roncus caralitanus n. sp. della
Sardegna meridionale (Pseudoscorpionida, Neobi-
siidae). Boll. Soc. Ent. Italiana, 113:129-135.
Gardini, G. 1983. Redescription of /wMcwi

L. Koch, 1873, type-species of the genus Roncus L.
Koch, 1873 (Pseudoscorpionida, Neobisiidae). Bull.
British Arachnol. Soc., 6:78-82.

Hadzi, J. 1933. Prinos poznavanju pseudoskorpijske
faune Primorja. Prirod. Istr. Kralj. Jugoslavije, 18:
125-192.

Koch, L. 1873. Uebersichtliche Darstellung der eu-
ropaeischen Chemetiden (Pseudoscorpione). Bauer
und Raspe, Nuernberg, Pp. 1-68.

Muchmore, W. B. 1969. A population of a European
pseudoscorpion established in New York. Ent. News,
80:66.

Poinar, Jr., G. O. & B. P. M. Curcic. 1992. Parasitism
of Pseudoscorpiones (Arachnida) by Mermithidae
(Nematoda). J. Arachnol., 20:64-66.

Manuscript received July 1991; revised September 1991.

1992. The Journal of Arachnology 20:129-133

NOTES ON MATING AND REPRODUCTIVE
SUCCESS OF CEROPELMA LONGISTERNALIS
(ARANEAE, THERAPHOSIDAE) IN CAPTIVITY

Fernando G. Costa and Fernando Perez-Miles: Division Zoologia Experimental, Insti-
tute de Investigaciones Biologicas Clemente Estable; Av. Italia 3318, Montevideo,
Uruguay

ABSTRACT. The reproductive success of a mating pair of Cewpelma longisternalis is reported. Courtship
and mating behaviour are described. The female molted, mated, built a retreat, and made a free egg sac under
laboratory conditions. For 49 days she remained in the retreat, until the juveniles emerged. Courtship, mating
and egg production are discussed and compared with data from other Mygalomorphae.

The biology of Theraphosidae in particular,
and Mygalomorphae in general, has not been
studied as thoroughly as that of the Araneomor-
phae. The state of Mygalomorphae systematics
is confusing and has only recently been globaly
focused (Raven 1985). Baerg (1928, 1958) car-
ried out the first biological studies on thera-
phosid natural history; and more recently Minch
(1978a, b, c; 1979a, b) studied Aphonopelma
chaicodes Chamberlin intensively. The Thera-
phosidae is a predominantly South American
family, yet the biology of most South American
species is unknown. Brazil & Vellard (1926) and
Biicherl (1951; 1952) described the reproductive
biology of several species, but only as addenda
to medical or systematic objectives. Galiano
(1969; 1973a, b; 1984) and Celerier (1986a) de-
scribed the postembryonic development and the
life cycles of several species, including Ceropel-
ma longisternalis Schiapelli & Gerschman (Ga-
liano 1973a). In spite of the frequent breeding
and commercialization of “tarantulas” in many
parts of the world, information on reproduction
obtained through mating in the laboratory is only
know from a brief note of Celerier (1986b).

The present paper initiates a series of studies
on the biology of some Uruguayan Mygalomor-
phae, contributing information regarding the re-
productive biology of C. longisternalis, a rela-
tively small-sized theraphosid spider (9.7 mm
carapace length). We describe the reproductive
success of a female that molted and mated in the
laboratory.

METHODS

An adult female and an adult male C. longis-
ternalis were captured in the Sierra de las Ani-

mas, Maldonado, Uruguay, on 1 7 March and 1 8
April 1989, respectively. They were maintained
separately in the laboratory in plastic petri dishes
(14.0 cm in diameter and 1.5 cm high) with wet
cotton wool. Tenebrio sp. larvae (Coleoptera) and
cockroach Blaptica dubia juveniles were provid-
ed weekly for food. The female molted on 23
March 1989, reverting to a virgin. Maximum and
minimum room temperatures were recorded dai-
ly. Room temperature, during captivity until
mating, was maintained close to 25 °C; thereafter
it was more variable (Fig. 1). After copulating
the female was placed in a cylindrical glass jar
measuring 17.0 cm in diameter and 6.0 cm high,
with a lid, and containing a 3 cm deep layer of
soil. She was fed mealworms and cockroaches
ad libitum, and provided with a small container
with water.

OBSERVATIONS

On 20 April 1989 the male and female were
placed together: the plastic petri dish with the
female was placed inside a larger cylindrical con-
tainer (17.0 cm diameter and 6.0 cm high) and
the lid was removed. Ten minutes later the male
was gently placed in the container. He advanced
slowly and soon made contact with the female.
The male’s courting behavior was characterized
by palpal drumming (up and down alternating
movements of the palpi touching the substra-
tum) and body vibrations (probably caused by
inward contractions of the third legs). The female
raised her body threateningly, extended the palpi
and forelegs upwards and opened her chelicerae.
The male tapped the female with his forelegs and,
pushing, clasped with his tibial spurs the female’s
open fangs. The female’s fangs penetrated be-

129

130

THE JOURNAL OF ARACHNOLOGY

time (months)

Figure 1 . — Room temperature variations during maintainance of copulated female Ceropelma longisternalis\
the superior line represents maximum daily temperatures and the inferior line minimum daily temperatures.

tween the external and the internal spurs, by the
retrolateral side of male forelegs, and closed
around the external apophysis (Figs. 2-5). The
basal tubercle of each foreleg metatarsus appar-

ently helped hold the female’s fang while the
metatarsus and tarsus remained against the basal
joint of her chelicera. The male’s second legs
surrounded the female cephalothorax and pulled

Figures 2-5. — Copulation in C. longisternalis: 2, Initial copulation position (drawing obtained from a pho-
tograph). The male (at left) grasps the female cheliceral fangs with the foreleg tibial spurs while surrounding and
attracting her with second legs; 3, The male attempts to insert the left palpus while the female (at right) rises
over the male (drawing obtained from another photograph); 4, Ventral view (slightly prolateral) of tibial meta-
tarsus joint of male right foreleg showing external (es) and internal (is) tibial spurs, and metatarsal basal tubercle
(bt); 5, Male right foreleg (at left) clasping female left cheliceae (at right), lateral view. Drawn from photographs,
notes and a reconstruction using preserved specimens.

COSTA & PEREZ-MILES- REPRODUCTION IN CEROPELMA

131

it towards him to raise the female, so that the
male palpi could approach the female genital zone
(Fig. 2). Legs III and IV of both spiders were
responsible for equilibrium. The male palpi os-
cillated in alternate fashion barely touching the
female’s abdomen while the male placed himself
underneath; initially the female lifted the ab-
domen a little avoiding contact with the extended
male palpi (Fig. 3).

Palpal insertions alternated regularly begin-
ning with the left palpus. There were five inser-
tions and the average duration of each was 50 s.
The female initiated weak leg movements 105 s
after commencement of copulation. Male body
vibrations were also recorded after three min
(during the fourth palpal insertion). The mating
pair partially lost its balance twice, 20 s and 306
s after commencing copulation; following the lat-
ter the female increased leg movements and the
male moved backwards freeing the female from
his grasp. An immediate new encounter showed
an active but non-aggressive female but did not
involve any clasping attempts by the male. Cop-
ulation measured from the first insertion lasted
5.3 min. Room temperature during copulation
was 23 °C. The male copulated one and four days
later with two other females.

The mated female mounded soil against one
side of the container and built a retreat there
against the glass so that we could partially ob-
serve her activity. The retreat was lined with silk
and its ceiling was against the lid. The female
reconstructed this silk ceiling each time it was
destroyed when we opened the jar to provide
food. Towards the beginning of spring (27 Sep-
tember, 160 days after copulation), the female
built a white oval egg sac with a 2.0 cm maxi-
mum diameter. The egg sac was not attached to
the retreat and the female kept her palpi and legs
in contact with it. Juveniles emerged 49 days
following oviposition. We then opened the re-
treat, counted the juveniles, measured the re-
treat, returned the juveniles to the female, and
closed the container (duration of manipulation:
40 min). Only 16 juveniles were found (three of
them with exuviae remains still attached) and 1 0
exuviae adhered to the retreat’s silk. The egg sac
silk capsule was not analysed and remained
pressed against the end of the retreat. The re-
treat’s characteristics were as follows: large silk
capsule over ground surface measuring 15.0 cm
long, 3.5 cm wide and 2.0 cm high, limited on
top by the lid, on the sides by the glass wall and
towards the center of the container by the soil

accumulated with silk. This capsule connected
with the underground portion through a hole of
approximately 2.5 cm diameter. The under-
ground portion, extended horizontally along the
bottom of the container and measured 7.0 cm
long, 2. 3-3. 2 cm wide and 2. 3-4.0 cm high.

The female did not rebuild the silk capsule
following manipulation or expel the egg sac’s
covering. Juveniles were frequently seen out of
the retreat 22 days after emergence. The female
molted 85 days after emergence of juveniles (8
February 1990, summer). Two juveniles still re-
mained in the retreat. The observations ended
four days later.

DISCUSSION

Immediate contact between male and female
prevented us from noticing whether the male can
detect the presence of a female by contact with
her silk. Nevertheless, one of us (Costa) observed
a sexual response in this and other male C. lon-
gisternalis when placed on female silk. Male be-
havior was: brusque frontward and downward
pushing movements of his body, which quickly
drew back keeping the tarsi fixed on the ground.
Platnick (1971) considered male Theraphosidae
capable of sexual recognition only by direct con-
tact with females (level I). But other authors have
observed courting responses in males in the pres-
ence of conspecific female silk: Baerg (1958) in
Dugesiella hentzi (Girard) and Minch (1979b) in
Aphonopelma chalcodes Chamberlin, and also
one of us (Costa) in species of Grammostola
Simon, Eupalaestrus Pocock, Oligoxystre Vel-
lard, Acanthoscurria Ausserer and Homeomma
Ausserer. These observations suggest that tacto-
chemical recognition is widespread in the family.

Male C. iongisternalis movements (palpal
drumming, body vibrations and advancing with
forelegs raised) when placed before the female,
female movements (threatening with palpi and
forelegs, raised body, half-open chelicerae) and
combined movements (tapping and entwining
forelegs) are similar to the courtship movements
observed in other Theraphosidae and other My-
galomorphae. Probably these movements con-
stitute species-specific stimuli transmitted via
acoustic and/or vibratory channels (through the
substratum) and the tacto-chemical channel dur-
ing physical contact. The apparently threatening
female response is a necessary condition for cop-
ulation enabling the male to grasp the female’s
fangs with his foreleg tibial spurs. This peculiar
grasp is the rule in Theraphosidae (Baerg 1928,

132

THE JOURNAL OF ARACHNOLOGY

1958; Minch 1979b; Bergo & Abe 1985; Costa
[pers. obs. in several Uruguayan species]), and
was also reported for Nemesia caementaria (La-
treille), Nemesiidae, by Buchli (1962), and for
Australothele jamiesoni Raven, Dipluridae, by
Raven (1988). It is surprising that Brazil & Vel-
lard (1926) and Biicherl (1951) did not describe
this conspicuous cheliceral clasping in Gram-
mostola spp.

The simultaneous attraction of legs II and the
fastening of forelegs onto the female chelicerae
neutralizes these dangerous instruments and po-
sitions the genital zone within reach of the male
palpi. A similar male attraction, either of the
body or of the basal part of the female’s legs, has
not been indicated in Theraphosidae, although
already reported in Dipluridae: Euagrus sp.,
Coyle ( 1986) and Australothele jamiesoni. Raven
(1988); in Hexathelidae: Macrothele calpeiana
(Walckenaer), Snazell & Allison (1989); in Ne-
mesiidae: Nemesia caementaria, Buchli (1962)
and Acanthogonatus tacuariensis (Perez-Miles &
Capocasale), Costa, pers. comm, in Perez-Miles
& Capocasale (1982).

Coyle (1985) briefly reviewed the different types
of sexual embrace in Mygalomorphae. Grasping
with legs II in C. longisternalis seems to disturb
the couple’s equilibrium. Clasping of chelicerae
forces this species to assume a raised position,
incompatible with copulation in the safety of the
usual narrow burrows of the females. Conse-
quently we would expect copulation to occur at
the burrow entrance, as Minch ( 1 979b) observed
for Aphonopelma chalcodes. The increased risk
of mating in an exposed situation would be com-
pensated for by the brief duration of copulation.

Baerg (1928, 1958) observed in Eurypelma
californica Ausserer and Dugesiella hentzi be-
tween 1-4 alternating palpal insertions in one
minute. Minch (1979b) reported an average of
seven alternating insertions during 2.3 min in
chalcodes. The C. longisternalis copulation was
longer than in these spiders, but the number of
insertions was intermediate between them.

Postcopulatory attack, and subsequent can-
nibalism, by the female was reported by Brazil
& Vellard (1926) and Biicherl (1951, 1952) in
several South American Theraphosidae. Our ob-
servations on C. longisternalis and other Uru-
guayan Theraphosidae show that males escape
undamaged after copulation. Biicherl (1951,
1952) stressed as exceptional the absence of fe-
male cannibalism in Grammostola mollicoma
(Ausserer) and G. longimana Mello-Leitao. Ber-

go & Abe (1985) reported a similar fact in Pam-
phobeteus sorocabae Mello-Leitao, a species cit-
ed by Biicherl (1952), however, as a postnuptial
devourer. Multiple copulation by males, a phe-
nomenon observed in this study and common
in Uruguayan theraphosids and other mygalo-
morphs (Baerg 1928, 1958; Minch 1979b; Coyle
1986; Coyle & O’Shields 1990) may be in part
the result of pronounced female longevity and a
resultant high ratio of adult females to adult males
(Coyle 1986).

The closed retreat built by C. longisternalis

when overwintering and laying eggs is similar to
the closed burrows observed by others: Baerg
(1958) in D. hentzi. Gabel (1972) in a non-de-
termined “tarantula”, and Minch (1979a) in A.
chalcodes. In the field juveniles and non-repro-
ductive females C. longisternalis build simple
silk tubes under stones, with little or no under-
ground component. The construction reported
here probably improves clutch viability by
damping external thermal and humidity varia-
tions. Its reconstruction ceased with the emer-
gence of juveniles. Duration of egg sac incubation
is comparable with that indicated by Biicherl
(1951) in Grammostola spp., Ibarra-Grasso
(1961) in G. burzaquensis Ibarra-Grasso, Baerg
(1958) and Whitcomb & Weems (1976) in Du-
gesiella hentzi, but shorter than that indicated
by Brazil & Vellard (1926) in Grammostola spp.

The very low number of juveniles emerging
from the egg sac may be due to laboratory con-
ditions. Two C. longisternalis egg sacs of similar
size obtained in the field contained 103 and 1 1 1
eggs, with an average diameter of 1 .7 mm. Larger
species of Theraphosidae lay more eggs: the small
Grammostola burzaquensis lays 100-200 eggs
(Ibarra-Grasso 1961), the large Acanthoscurria
atrox Vellard up to 2000 eggs (Lourengo 1978)
and the large Pamphobeteus spp. between 1 200-
2000 eggs (Bucherl 1951; Valente et al. 1985).
The postreproductive molt seems to occur later
in C. longisternalis than in other species (Brazil
& Vellard 1926; Bucherl 1951).

ACKNOWLEDGMENTS

We are indebted to Liliana Prandi for the
maintainance of spiders, to Roberto Capocasale,
Gabriel Francescoli, Barbara Y. Main and Fred-
erick A. Coyle for the critical reading of the
manuscript, to Eduardo Gudynas for his help
with drawings and to Ines Trabal for the trans-
lation into English.

COSTA & PEREZ-MILES-REPRODUCTION IN CEROPELMA

133

LITERATURE CITED

Baerg, W. J. 1928. The life cycle and mating habits
of the male tarantula. Quart. Rev. Biol., 3:109-1 16.

Baerg, W.J. 1958. The Tarantula. Univ. Kansas Press;
Lawrence, Kansas. 88 pp.

Bergo, E. S. & A. S. Abe. 1985. Observagoes sobre
o comportamento de corte e copula em Pamphob-
eteus sorocabae Mello-Leitao, 1923 Araneae - Ther-
aphosidae. Res. XII Congr. Brasil. Zool. Pp.46-47.

Brazil, V. & J. Vellard. 1926. Contribugao ao estudo
do veneno das aranhas (continuagao) (Segunda me-
moria). Mem. Inst. Butantan (Brasil), 3:243-299.

Biicherl, W. 1951. Estudos sobre a biologia e a sis-
tematica do genero Grammostola Simon, 1892.
Monogr. Inst. Butantan, 1:1-126.

Biicherl, W. 1952. Instintos matemais nas aranhas
brasileiras. Dusenia, 3:57-74.

Buchli, H. 1962. Note preliminaire sur Faccouple-
ment des araignees mygalomorphes Nemesia cae-
mentaria, Nemesia dubia et Pachylomerus piceus
(Ctenizidae). Vie et Milieu, 13:167-178.

Celerier, M.-L. 1986a. Energy cost of exuviae pro-
duction of spiders. Bull. British Arachnol. Soc., 7:
77-82.

Celerier, M.-L. 1986b. Elevage d’un mygale: ^co^/ra
griseipes, Pocock 1897 (Araneae, Theraphosidae).
Actas X Congr. Int. Aracnol. Jaca/Espaha, 1:56.

Coyle, F. A. 1985. Observations on the mating be-
haviour of the tiny mygalomorph spider, Microh-
exura montivaga Crosby & Bishop (Araneae, Di-
pluridae). Bull. British Arachnol. Soc., 6:328-330.

Coyle, F. A. 1986. Courtship, mating, and the func-
tion of male-specific leg structures in the mygalo-
morph spider genus Euagrus (Araneae, Dipluridae).
Proc. IX Int. Congr. Arachnol., Panama. Pp. 33-
38.

Coyle, F. A. & T. C. O’Shields. 1990. Courtship and
mating behavior of Thelochoris karschi (Araneae,
Dipluridae), an African funnel web spider. J. Ar-
achnoL, 18:281-296.

Gabel, J. R. 1972. Further observations of the ther-
aphosid tarantula burrows. Pan-Pacific EntomoL,
48:72-73.

Galiano, M. E. 1969. El desarrollo postembrionario
larval en Grammostola pulchripes (Simon, 1891)
(Araneae, Theraphosidae). Physis (Buenos Aires),
C. 29:73-90.

Galiano, M. E. 1973a. El desarrollo postembrionario

larval en Theraphosidae (Araneae). Physis (Buenos
Aires), C. 32:37-A6.

Galiano, M. E. 1973b. El desarrollo postembrionario
larval de Avicularia avicularia (Linnaeus, 1758)
(Araneae, Theraphosidae). Physis (Buenos Aires),
C. 32:315-327.

Galiano, M. E. 1984. Datos adicionales sobre el ciclo
vital de Acanthoscurria sternalis Pocock, 1 903 (Ara-
neae, Theraphosidae). Rev. Soc. Ent. Argentina, 43:
45-55.

Ibarra-Grasso, A. 1961. Datos biologicos sobre
Grammostola burzaquensis Ibarra Grasso 1 946 y su
distribucion geografica (Araneae, Theraphosidae).

Neotropica, 7:7-12.

Lourengo, W. R. 1978. Notas sobre a biologia de
Acanthoscurria atrox Vellard, 1 924 (Araneae, Ther-
aphosidae). Rev. Brasileira Biol., 38:161-164.

Minch, E. W. 1978a. Daily activity patterns in the
\3.Tanln\dLAphonopelmachalcodesC\ia.mher\m. Bull.
British Arachnol. Soc., 4:231-237.

Minch, E. W. 1978b. The effectiveness of chemical
defenses against predation by the tarantula Aphon-
opelma chalcodes Chamberlin. Bull. British Arach-
nol. Soc., 4:275-278.

Minch, E. W. 1978c. Some notes on the egg sacs of
Aphonopelma chalcodes (Araneae: Theraphosidae).
Pan-Pacific Entomol., 54:153.

Minch, E. W. 1979a. Burrow entrance plugging be-
haviour in the tarantula Aphonopelma chalcodes
Chamberlin (Araneae: Theraphosidae). Bull. British
Arachnol. Soc., 4:414-415.

Minch, E. W. 1979b. Reproductive behaviour of the
X&ran\n\2i Aphonopelma chalcodes Chamberlin (Ara-
neae: Theraphosidae). Bull. British Arachnol. Soc.,
4:416-420.

Perez-Miles, F. & R. M. Capocasale. 1982. Aranas
del Uruguay, IV. Hallazgo de una tercera especie
del genero Pycnothelopsis: Pycnothelopsis tacuarien-
sis sp. nov. (Araneae, Pycnothelidae). Com. Zool.
Mus. Hist. Nat. Montevideo, 11:1-7.

Platnick, N. 1971. The evolution of courtship be-
haviour in spiders. Bull. British Arachnol. Soc., 2:
40-47.

Raven, R. J. 1985. The spider infraorder Mygalo-
morphae (Araneae): Cladistics and systematics. Bull.
American Mus. Nat. Hist., 182:1-180.

Raven, R. J. 1988. Preliminary observations on the
mating behaviour of the Australian mygalomorph
spider Australothele jamiesoni (Dipluridae, Arane-
ae, Arachnida). Mem. Queensland Mus., 25:471-
474.

Snazell, R. & R. Allison. 1 989. The genus Macrothele
Ausserer (Araneae, Hexathelidae) in Europe. Bull.
British Arachnol. Soc., 8:65-72.

Valente, M. M., I. Mombrum de Carvalho & M. J.
Bauab-Vianna, 1985. Microanatomia do aparelho
genital femenino de Pamphobeteus sorocabae (Po-
cock, 1901) (Aviculariidae - Araneae). Res. XII
Congr. Brasileiro Zool. Pp. 46.

Whitcomb, W. H. & H. V. Weems, Jr. 1976. The
tarantula (family Theraphosidae). Florida Dept. Agr.
7 Consumer. Serv., Division Plant Industry, Ento-
mol. Circular, 169:1-2.

Manuscript received April 1991, revised August 1991.

1992. The Journal of Arachnology 20:134-136

DESCRIPCION DEL MACHO DE ACHAEARANEA JEQUIRITUBA

(ARANEAE, THERIDIIDAE)’

Alda Gonzalez^: Facultad de Ciencias Naturales y Museo, Paseo del Bosque s/n, 1900
La Plata, Argentina

ABSTRACT. Description of the male of Achaeranea jequirituba (Araneae, Theridiidae). The male oiAchaear-
anea jequirituba from Misiones province (Argentina) is described and new distributional records and biological
considerations are given.

RESUMEN. Se describe el macho de Achaearanea jequirituba de la provincia de Misiones, se amplia su
distribucion geografica y se aportan consideraciones biologicas.

El genero Achaearanea Strand, esta amplia-
mente distribuido por todo el mundo y escasa-
mente representado en nuestro pais. Revisiones
de las especies americanas fueron publicadas por
Levi (1955, 1959, 1963, 1967). Levi (1963) sos-
tiene que pocas especies son citadas para Argen-
tina debido a la falta de especimenes en las co-
lecciones. Algunas especies de Achaearanea son
dificiles de separar de las de Theridion, parti-
cularmente cuando solo la hembra es conocida.
Muchas de las nuevas especies descriptas por
Levi para America del Norte, Central y del Sur
comprenden solamente la descripcion de los
ejemplares hembras. Achaearanea jequirituba fue
descripta por Levi en 1 963. En el presente trabajo
se describe el macho dt Achaearanea jequirituba,
hasta ahora desconocido para la ciencia, se am-
plia su distribucion geografica a la Argentina y
se aportan datos sobre la biologia.

METODOS

Los estudios efectuados se realizaron sobre
material colectado en el campo. En viajes real-
izados a la provincia de Misiones en el ano 1 989,
se hallaron en las barandas de las pasarelas de
los saltos de las Cataratas del Iguazu, ejemplares
hembras que fueron identificados como
Achaearanea jequirituba y ejemplares machos no
descriptos hasta ese momento. Del material co-
lectado se seleccionaron cinco parejas de machos

' Contribucidn N° 195 del Centro de Estudios Paras-
itologicos y de Vectores (CEPAVE).

^ Facultad de Ciencias Naturales y Museo. 1 900 La
Plata, Argentina. (Investigadora, CONICET).

y hembras que copularon en el laboratorio. Cua-
tro de las hembras que se colectaron estaban
gravidas y desovaron en el laboratorio produ-
ciendo diez ootecas. Se criaron los juveniles
provenientes de esos desoves y los machos ob-
tenidos resultaron Achaearanea jequirituba,
identificandose asi al macho de la especie que se
describe por primera vez.

Los dibujos se llevaron a cabo con camara
Clara bajo microscopio binocular estereoscopico.
Las medidas se expresan en milimetros.

Abreviaturas utilizadas: MLP, Museo de Cien-
cias Naturales de La Plata; MACN, Museo Ar-
gentino de Ciencias Naturales “Bernardino Ri-
vadavia”.

Achaearanea jequirituba Levi
Figs. 1-2

Achaearanea jequirituba Levi, 1963, Bull. Mus. Comp.
Zool. 129:226, figs. 98-99. (Hembra holotipo de Je-
quirituba, Sao Paulo, Brasil, en el American Muse-
um of Natural History).

Descripcion. Macho. Cefalotorax ancho en su
parte media-posterior. Con surco toracico tri-
angular. Ojos medianos anteriores mayores que
el resto, 0.09 mm de diametro, separados entre
si por 0.06 mm, subcontiguos a los laterales. Ojos
posteriores casi en linea recta. Diametro de los
medianos posteriores 0.064 mm, separados entre
si por mas de su diametro (0.07 mm). Ojos la-
terales contiguos. Ojos laterales posteriores se-
parados de los medianos posteriores por 0.046
mm. Cefalotorax, dorsalmente pardo claro. Ven-
tralmente pardo oscuro con hordes negros. Patas
pardo claro con los extremos distales de patela,
tibia y metatarso negros. Palpos pardo claro. Ab-

134

GONZALEZ- DESCRIPCION DEL MACHO DE ACHAEARANEA JEQUIRITUBA

135

domen alargado. Dorsalmente predomina el col-
or negro, con lineas grises blanquecinas distri-
buidas irregularmente, sin manchas definidas.
Ventralmente negro con pequenas manchas
blancas alrededor de la zona anal. Medidas: Lar-
go total 1.90 mm. Cefalotorax: largo 0.96 mm,
ancho 0.70 mm. Pata I: femur 1.54 mm, patela
+ tibia 1.38 mm, metatarso 1.22 mm, tarso 0.50
mm. Pata II: femur 1.04 mm, patela + tibia 0.94
mm. Pata III: femur 0.70 mm, patela + tibia
0.64 mm. Pata IV: femur 1.07 mm, patela +
tibia 0.96 mm, metatarso 0.56 mm, tarso 0.42
mm. Longitud relativa de las patas 1-4-2-3.

El palpo con un embolo corto. El epigino fue
descripto por Levi, 1963, figs. 98-99.

Distribucion: BRASIL. Guanabara, Rio de Ja-
neiro. PARAGUAY. ARGENTINA. Misiones,
Parque Nacional Iguazu.

Material examinado: R. ARGENTINA, Misiones,
Parque Nacional Iguazu, 1 2 N° 17.154 MLP, 1 S N°
17.159MLP, noviembre 1988 (Gonzalez); 2 2 N“ 17.155
MLP, 27 2 N° 17.157 MLP, 6 3 N° 17.160 MLP, se-
tiembre 1989 (Gonzalez); 1 2 N° 8.882. 1980 MACN
(Galiano).

Biologia; Las hembras fueron encontradas
dentro de sus nidos. Estos consisten en una hoja
arrollada, suspendida de sus extremes por hilos
de tela en el angulo formado por dos parantes
de la baranda de las pasarelas. La hoja presen ta
un extreme totalmente cerrado y el otro con una
pequena abertura por donde entra y sale la arana.
Los individuos se alojan en su interior, cercanos
al extreme cerrado. Generalmente las hembras
permanecen en el nido junto al desove. Los ma-
chos fueron hallados compartiendo el nido con
las hembras o suspendidos de los hilos de la tela
que sostienen al nido.

Las ootecas son de color bianco, de aspecto
algodonoso y de forma piriforme. Son de pe-
queno tamano, miden promedio 2.9 mm de largo
(maximo 3.32 mm, minimo 2.64 mm) y pro-
medio 2.6 mm de ancho (maximo 2.75 mm,
minimo 2.48 mm).

Los huevos son esfericos, blancos y miden entre
0.58 mm y 0.62 mm de diametro. El total de
huevos de las diez ootecas examinadas fue de
298. El promedio resultante fue de 18 huevos
por ooteca, con un maximo de 32 y un minimo
de 15.

El desarrollo postembrionario es similar en el
tipo de eclosion y numero de estadios (desde la
eclosion hasta la dispersion) al de otras especies
de la familia Theridiidae, descriptos en trabajos

Figuras 1, 2.— Palpo izquierdo. 1) ventral; 2) lateral.

136

THE JOURNAL OF ARACHNOLOGY

anteriores (Gonzalez 1979, 1981, 1984, 1986,
1988). El lapse entre el desove y la dispersion de
los juveniles es de promedio 22.5 dias (maximo
25, minimo 17) considerando los individuos de
1 0 ootecas estudiadas.

LITERATURA CITADA

Gonzalez, A. 1979. Observaciones bioecologicas sobre
una especie del genero LatrodectusWalckm'a.ei, 1805,
del grapo mactans, de Sierra de la Ventana (Prov-
incia de Buenos Aires, Argentina) (Araneae, Ther-
idiidae). Ill ~ Desarrollo postembrionario. Acta Zool.
Lilloana, 35:95-110.

Gonzalez, A. 1981. Desarrollo postembrioario de La-
trodectus mirabilis, Latrodectus corallinus y Latro-
dectus antheratus (Araneae, Theridiidae). Physis (Bs.
As.), Secc. C, 39:83-91.

Gonzalez, A. 1982. E! desarrollo postembrionario de
Tidarren sisyphoides (Walckenaer) (Araneae, Ther-
idiidae). Physis (Bs. As.), Secc. C, 41:87-91.
Gonzalez, A. 1 984. Desarrollo postembrionario y ev-
olucion de los organos mecanoreceptores de Latro-
dectus diaquita Carcavallo y estudio de la tricobo-
triotaxia de Lactrodectus quartus Abalos (Araneae,
Theridiidae). Physis (Bs. As.) Secc. C, 42:1-5.

Gonzalez, A. 1987. Ciclo biologico y desarrollo de
Steatoda retorta Gonzalez (Araneae, Theridiidae).
Revta. Soc. Ent. Argentina, 44:185-197.

Gonzalez, A. y A. L. Estevez. 1988. Estudio del de-
sarroilo postembrionario y estadisticos vitales de
Theridion rufipes Lucas, 1846 (Araneae, Theridi-
idae). Revta. Bras. Ent., 32:499-506.

Levi, H. 1955. The spider genera Coressa and
Achaearanea in America North of Mexico (Araneae,
Theridiidae) American Mus. Novitates, 1718:1-33.

Levi.H. 1959. The spider genera Achaearanea, Ther-
idion and Sphyrotinus from Mexico, Central Amer-
ica and West Indies. (Araneae, Theridiidae). Bull.
Mus. Comp. ZooL, 121:57-163.

Levi, H. 1963. American spiders of the genus
Achaearanea and the new genus Echinotheridion
(Araneae, Theridiidae). Bull. Mus. Comp. Zool., 129:
189-240.

Levi, H. 1967. The theridiid spider fauna of Chile.
Bull. Mus. Comp. Zooi., 136:1-20.

Manuscript received November 1990, revised December
1991.

1992. The Journal of Arachnology 20:137-143

A REVIEW OF META (ARANEAE, TETRAGNATHIDAE),
WITH DESCRIPTION OF TWO NEW SPECIES

Yuri M. Marusik: Institute of Biological Problems of the North, Russian Academy of
Sciences, Magadan 685010, Russia

Seppo Koponen’: Centre d’etudes nordiques, Universite Laval, Ste-Foy, Quebec GIK
7P4, Canada

ABSTRACT. The validity of the genera Meta C. L. Koch and Metellina Chamberlin & Ivie is supported,
mainly based on genital structure and living habitats. Two new species, closely related to and earlier partly
included in Meta menardi (Latreille) (limited to Europe), are described: M. americana n. sp. (eastern North
America) and M. manchurica n. sp. (Russian Far East). The additional species of Meta are M. bourneti Simon
(southern Europe, Caucasus and Canary Islands) and M. dolloff\x.Vi (California).

The status of the genera Meta C. L. Koch, 1836
and Metellina Chamberlin & Ivie, 1 94 1 has been
unstable. Levi (1980) treated them as valid gen-
era in his North American revision. Although
this concept has been followed by some recent
authors (Levy 1985; Marusik 1985, 1986; Ya-
ginuma 1986; Coddington 1990), some others
have considered Meta a senior synonym of Me-
tellina (Wunderlich 1987).

The genus Meta sensu stricto includes three
known species, Meta menardi (Latreille), M.
bourneti Simon and M. dollojfl^Yi. The first
mentioned has been regarded as a Holarctic spe-
cies. The material previously treated as M. men-
ardi appears to include three very closely related
species, with allopatric distributions. The two
new species will be described here.

THE GENERA META AND METELLINA

We cannot agree with the synonymization of
Meta and Metellina made by Wunderlich ( 1 987).
As Levi (1980) clearly pointed out there are
marked differences in body shape and in genitalia
of both sexes; see figs. 95-135 in Levi (1980) and
figs. 49, 54 in Coddington (1990). The main dif-
ferences in genitalia between genera Meta and
Metellina are: epigynum, embolus, conductor and
paracymbium sclerotized in Meta and not in Me-
tellina, embolus free in Meta and covered by
conductor in Metellina, base of embolus com-
plex (apophyses) in Meta compared to the simple

' Current address: Zoological Museum, Univer-
sity of Turku, SF-20500 Turku, Finland

base in Metellina, conductor directed prolater-
ally in Meta and more horizontally in Metellina,
and epigynal openings posterior in Meta and
ventral in Metellina. Coddington (1990) pointed
out the difference in the sperm duct routing
through tegulum in Meta (complex) and Metel-
lina (simple). The size ol Meta varies from 8-17
mm and that of Metellina from 4-8 mm, and the
abdomen of Meta is almost as high as long. The
species of Meta live in caves (however, see Pen-
nington 1979) and only a few species of Metellina
occur in cave entrances, the rest living outside
of caves.

The placement of Meta at the family level is
also open to discussion. Traditionally it has been
included in Araneidae (e.g.. Locket & Millidge
1953), and Simon (1929) placedM^ta in the sub-
family Tetragnathinae. Locket et al. (1974) trans-
ferred Meta into Tetragnathidae, and recently it
has been considered either in the family Metidae,
or in the subfamily Metinae vffthin Tetragnath-
idae or within Araneidae (see e.g., Heimer &
Nentwig 1982; Roberts 1985; Levi 1986; Wun-
derlich 1986). Differences in anatomy and gen-
italia between metids and tetragnathids have been
pointed out, e.g., by Palmgren (1978) and Cod-
dington ( 1 990). We prefer to include Meta in the
subfamily Metinae within the family Tetrag-
nathidae until new data is available.

THE GENUS META C. L. KOCH

For description and figures of Meta bourneti,
known from southern Europe, Caucasus and Ca-

137

138

THE JOURNAL OF ARACHNOLOGY

nary Islands, and of M. dolloff, from California,
see Locket & Millidge (1953), Levi (1980) and
Roberts (1985).

Meta americana, new species
Figures 1-4, 14

Types. — Male holotype, male paratype and fe-
male paratype from USA, Pennsylvania, NE of
Jamison, Horseshoe Bend, Neshaminy Cr.
(40U6'N, 75°03'W); June 1954, leg. W. Ivie; de-
posited in the American Museum of Natural His-
tory, New York. One male and one female para-
type from the same locality and date; deposited
in the Zoological Museum, Moscow State Uni-
versity, Russia. Three male paratypes and one
female paratype from Canada, Ontario, Egan-
ville, Bonneshere Caves (45°30'N, 77°00'W); 12
June 1972, leg. S. Peck; deposited in the Cana-
dian National Collections, Ottawa.

Etymology.— The specific name refers to the
distribution area.

Diagnosis.— Af. americana differs from close-
ly related M. manchurica and M. menardi in the
shape of copulatory organs. M. americana has
the widest tegulum and numerous small denticles
on conductor (Figs. 1, 14). Unlike M. menardi,
M. americana has clearly divided apophysis on
the embolus base (Fig. 2) as does M. manchurica.
The ridge on the embolus base of M. americana
is higher and wider than that of M. manchurica
(Figs. 1, 2 and 5, 6). The paracymbium is distally
not so rounded as that of M. manchurica and
M. menardi (Figs. 3, 7, 12). Females of M. amer-
icana can be recognized by the shape and col-
oration of the epigynal bulge which is wide and
darker than the basal part of the epigynum (Fig.
4).

Description.— 5/ze.‘ total length of males 7.0-
10.1 mm, of females 7.8-13.7 mm; carapace
length of males 3. 7-5.0 mm, of females 4. 0-5. 9
mm; carapace width of males 3. 0-4.0 mm, of
females 3. 3-4. 5 mm; leg I (femur- patella -I- tibia-
metatarsus + tarsus) of males 9.5-8.0-10.5 mm,
7.2-10.2-11.1 mm and 7.5-10.4-11.2 mm, of
females 6. 5-8. 2-8. 5 mm and 7. 5-9. 5-9. 8 mm.
Coloration, see Levi’s (1980) description and figs.
113-115.

Male palp: see Figs. 1-3, 14, and figs. 124-127
in Levi (1980). Conductor with numerous small,
thin denticles, tegulum wide.

Epigynum: see Fig. 4 and figs. 1 1 6-1 20 in Levi
(1980). Bulge darker than basal part of epigyn-
um.

Distribution.— Eastern North America (main-

ly east of the Mississippi River); the northern-
most records are from Newfoundland, Nova
Scotia, southern Quebec and northern shore of
Lake Superior; in the south it reaches the north-
ern parts of Georgia, Alabama, Arkansas and
Oklahoma, the southernmost record being from
Louisiana, Baton Rouge (Levi 1980: map 5).

Material examined.— Type material, and 13 males
and 1 7 females from Canada (Nova Scotia, New Bruns-
wick, Quebec and Ontario; in the Canadian National
Collections and in the Zoological Museum, University
of Turku). In addition, Prof. H. W. Levi kindly com-
pared present figures with a number of specimens from
the USA (in the Museum of Comparative Zoology,
Harvard University).

Meta manchurica, new species

Figures 5-9, 15

Types. — Holotype male from Russia, Primor-
ski (Maritime) Province, 20 km SW from Par-
tizansk, entrance of a cave; 2 May 1978, leg. A.
Lelei. Two female paratypes Primorski Prov.,
Khasan District, Nerpichya Bay, cave; 3 Septem-
ber 1978, leg. B.P. Zakharov. Types are depos-
ited in the Zoological Museum, Moscow State
University, Russia.

Etymology.— The specific name refers to the
distribution area.

Diagnosis.— Af. manchurica can be distin-
guished from closely related Af. americana and
Af. menardi by copulatory organs. Tegulum of
Af. manchurica is somewhat thicker than that of
Af. menardi but thinner than of M. americana
(Figs. 1, 5, 10). Denticles of conductor thick as
in Af. menardi but shorter and more numerous
(Figs. 15-16). Unlike in Af. menardi and Af.
americana, apophyses of the embolus base are
pointed in M. manchurica, and ridge at the em-
bolus base is low and short (Fig. 6). Females of
Af. manchurica can be recognized by the wide
epigynal bulge (Figs. 8, 9).

Description.— 5/zd; male 10.2 mm, females
1 1.6-12.2 mm; carapace length of male 4.8 mm,
of females 5.2 mm; carapace width of male 4.0
mm, of females 4. 0-4. 4 mm. Leg I (femur-patella
+ tibia-metatarsus + tarsus) of male 7.5-10.1-
1 1.4 mm, offemales 7.8-10.0-1 1.0 mm and 8.0-
10.5-10.6 mm. Coloration as in Af. menardi but
darker.

Male palp: see Figs. 5-7 and 15. Conductor
with numerous thick denticles, base of embolus
with pointed apical and dorsal apophyses, ridge
on base of embolus low and short.

Epigynum: see Figs. 8-9. Wide pale bulge, bas-

MARUSIK & KOPONEN-REVIEW OF META, WITH TWO NEW SPECIES

139

Figures l-A.~Meta americana n. sp., types from Pennsylvania; I, left male palp; 2, embolic apophysis; 3,
paracymbium; 4, epigynum (posterior view). Scale bars: 0. 1 mm.

140

THE JOURNAL OF ARACHNOLOGY

9

Figures 5-9.— Meta manchurica n. sp., types from Primorski Province, Russia: 5, left male palp; 6, embolic
apophysis; 7, paracymbium; 8, epigynum (ventral view); 9, epigynum (posterior view). Scale bars: 0.1 mm for
Figs. 5-7 & 9; 0.25 mm for Fig. 8.

al part of epigynum is darker than the apical one Meta menardi (Latreille, 1 804)

(bulge). Figures 10-13, 16

Distribution. - Maritime Province of Russia, Diagnosis. -See that of M. americana and M.

possibly also Japan and Korea. manchurica

Material examined. - Only types. Description. — See Wiehle (1931), Locket &

MARUSIK & KOPONEN- REVIEW OF META, WITH TWO NEW SPECIES

141

Figures 10-13.— Meta menardi (Latreille) from Germany: 10, left male palp; 1 1, embolic apophysis; 12, para-
cymbium; 13, epigynum (posterior view). Scale bars: 0.1 mm.

Millidge (1953) and Roberts (1985) and Figs. 1 0-
13, 16.

Size: total length of males up to 12.5 mm
(specimens from Sweden), 10-11 mm (Roberts
1 985), of females up to 15.1 mm (from Sweden),

12-15 mm (Roberts 1985); carapace length of
males up to 6.0 mm (Sweden), 5-5.5 mm (Wiehle
1931), of females up to 6.6 mm (Sweden), 6-6.5
mm (Wiehle 1931).

Male palp: see Figs. 10-12, 16.

142

THE JOURNAL OF ARACHNOLOGY

14 15 16

Figures 14-16.— Tip of conductor of Meta: 14, M. americana n. sp.; 15, M. manchurica n. sp.; 16, M. menardi
(Latreille). Scale bars: 0.1 mm.

Epigynum: see Fig. 13.

Distribution.— Western Europe. Probably also
North Africa (Algeria) and perhaps Middle East
(Syria) (Roewer 1942; however, see Levy 1987).
The northernmost known records are from Brit-
ish Islands, including Ireland and Scotland
(Locket et al. 1974), from the Arctic circle on
Norwegian coast, central parts of Sweden and
southwestern comer of Finland (Hippa et al.
1 984). The easternmost records are from Latvia,
and from Moldova and other Transcarpathian
areas (K. G. Mikhailov, unpubl. catalogue).

Material examined. —Several males and females from
Germany, Sweden and Norway (in the Institute of Bi-
ological Problems of the North, Magadan, in the Swed-
ish Museum of Natural History, Stockholm, and in the
Zoological Museum, University of Turku). Drawings
of the material from (eastern) Germany, Hinter Sachs.
Schweiz, Gr. Winterberg, S-Kuppe (420-490 m); June
1978, leg. S. Heimer.

ACKNOWLEDGMENTS

We wish to thank the following persons for
material, discussions and other help: C. D. Don-
dale & J. H. Redner (Ottawa), S. Heimer (Dres-
den), T. Kronestedt (Stockholm), H. W. Levi
(Harvard), D. M. Logunov (Novosibirsk), K. G.
Mikhailov (Moscow) and N. 1. Platnick (New
York).

LITERATURE CITED

Coddington, J.A. 1990. Ontogeny and homology in
the male palpus of orb-weaving spiders and their
relatives, with comments on phylogeny (Araneocla-
da: Araneoidea, Deinopoidea). Smithsonian Contr.
Zool., 496:1-52.

Heimer, S. & W. Nentwig. 1982. Thoughts on the
phylogeny of the Araneoidea Latreille, 1 806 (Arach-

nida, Araneae). Zeitschr. Zool. Syst. Evol. Forsch.,
20:284-295.

Hippa, H., S. Koponen & R. Mannila. 1984. Inver-
tebrates of Scandinavian caves. 1. Araneae, Opi-
liones, and Pseudoscorpionida (Arachnida). Ann.
Entomol. Fennici, 50:23-29.

Levi, H. W. 1 980. The orb-weaver genus Mecynogea,
the subfamily Metinae and the genera Pachygnatha,
Glenognatha and Azilia of the subfamily Tetrag-
nathinae north of Mexico (Araneae: Araneidae). Bull.
Mus. Comp. Zool., 149:1-74.

Levi, H. W. 1986. The Neotropical orb-weaver gen-
era Chrysometa and Homalometa (Araneae: Te-
tragnathidae). Bull. Mus. Comp. Zool., 151:91-215.

Levy, G. 1987. Spiders of the genera Araniella, Zyg-
iella, Zilla and Mangora (Araneae, Araneidae) from
Israel, with notes on Metellina species from Leba-
non. Zool. Scr., 16:243-257.

Locket, G. H. & A. F. Millidge. 1953. British spiders
11. Ray Society, 13:1-449.

Locket, G. H., A. F. Millidge & P. Merrett. 1974.
British spiders III. Ray Society, 149:1-315.

Marusik, Y. M. 1985. A systematic list of the orb-
weaving spiders (Aranei: Araneidae, Tetragnathi-
dae, Theridiosomatidae, Uloboridae) of the Euro-
pean part of the USSR and the Caucasus. Trudy
Zool. Inst. Leningrad, 139:135-140.

Marusik, Y. M. 1986. A redescription of types of
certain orb-weaving spiders (Araneidae, Tetrag-
nathidae) from S. A. Spassky collection. Vest. Zool.,
6:19-22.

Palmgren, P. 1978. Taxonomic position of the genus
Meta (Araneida). Ann. Zool. Fennici, 15:241-242.

Pennington, B.J. 1979. The colour patterns of diurnal
Meta menardi (Latreille). Bull. British Arachnol. Soc.,
4:392-393.

Roberts, M. J. 1985. The spiders of Great Britain
and Ireland. Vol I. Harley Books, Colchester. 229
PP-

Roewer, C. 1942. Katalog der Araneae 1. Bremen.
1040 pp.

MARUSIK & KOPONEN- REVIEW OF META, WITH TWO NEW SPECIES

143

Simon, E. 1929. Les Arachnides de France, 6:533-
772. Paris.

Wiehle, H. 1931. Spinnentiere oder Arachnoidea. VI.
Agelenidae - Araneidae. Tierwelt Deutschlands 23.
46 + 136 pp.

Wunderlich, J. 1 986. Spinnenfauna gestem und heute:
Fossile Spinnen in Bernstein und ihre heute leben-
den Verwandten. Straubenhardt. 283 pp.
Wunderlich, J. 1987. Die Spinnen der Kanarischen

Inseln und Madeiras. Taxonomy & Ecology 1 (Lan-
gen). 433 pp.

Yaginuma, T. 1986. Taxonomic notes on Japanese
spiders. (II): Araneus, Neoscona, Metellina, Cispius,
Heptathela. Fac. Let. Rev. Otemon Gakuin Univ.,
20:187-200.

Manuscript received April 1991, revised September 1991.

1992. The Journal of Arachnology 20:144-145

RESEARCH NOTES

COHABITATION OF SIX SPECIES OF SPIDERS IN WEBS
OF CYRTOPHORA MOLUCCENSIS (ARANEAE, ARANEIDAE)
IN MOOREA, FRENCH POLYNESIA

Webs of colonial spiders often harbor many
species of spiders in addition to the original ar-
chitects of the web (Buskirk 1982). On Moorea
(French Polynesia) I noticed several different
species in webs of the colonial spider Cyrtophora
moiuccensis (Doleschall) (Araneidae). On 26-28
April 1991,1 collected all inhabitants of five small
colonies and found five species in addition to C
moiuccensis (Table 1). Leucauge granulataQNal-
ckenaer)(Araneidae) built its own webs entirely
or partially supported by the frame threads of
the colonial spider’s web. Pholcus ancoralis L.
Ivoch (Pholcidae) also built its own web, but did
so in the upper barrier web of C. moiuccensis.
Tangaroa tahitiensis (Berland)(Uloboridae) con-
structed its own web, typically in the lower bar-
rier web and sometimes on the frame threads. I
found Argyrodes argentatus O.P. -Cambridge
(Theridiidae) in all parts of the colonial web but
failed to note whether it built its own webs. Also,
I did not notice whether the single specimen of
Theridion adamsoni Berland (Theridiidae) con-
structed its own web.

Argyrodes argentatus has been found in C.
moiuccensis webs on the Pacific island of Yap
(Berry 1987), and its congeners are common
kleptoparasites in the webs of other colonial and
noncolonial spiders (Kaston 1 965). The other four
species are not known to be frequent residents
of webs other than their own (J. Beatty, pers.

commun.). However, Rypstra (1979) and Lubin
(1974) noted Leucauge species building webs at-
tached to the guylines of Cyrtophora citricola
(ForskAal) and C. moiuccensis colonies, respec-
tively. Archaearanea tepidariorum (C. L. Koch)
(reported as Theridion tepidariorum) has been
found in C. citricola webs as well (Kaston 1965).
Within the Uloboridae, Uloborus species cohabit
with Cyrtophora and another colonial spider,
Stegodyphus sarasinorum Karsch (Rypstra 1 979).
Jackson & Rowe (1987) describe Pholcus ancor-
alis as a web-invading spider, but did not find it
in Cyrtophora webs.

Spiders that inhabit colonial spiders’ webs may
be obligatory kleptoparasites (e.g., many Argy-
rodes species) or facultative commensals that
build their own food-capturing webs within those
of the social spider. Cohabitation of the second
type may be selected for in habitats where good
web sites are rare (Buskirk 1986) or in which
disturbances like rain or falling leaves are com-
mon (Riechert et al. 1986); here webs of the co-
habiting species may gain protection from their
location within the sturdier web of the social
spider. Sharing webs also reduces the cost of silk
production for the cohabiting spider (e. g., Jakob
1991). Finally, the large size of colonial spider
webs allows them to occupy open spaces that are
high in insect traffic (Lubin 1974), and cohabi-
tants may benefit from this access to prey. With

Table 1.— Spiders found in five webs of Cyrtophora moiuccensis near Paopao, Moorea, French Polynesia.
Juvenile males were recognized by their swollen palps. Juvenile C. moiuccensis that were as large or larger than
males were termed juvenile females and their number is shown in parentheses next to that of juvenile males.

Adults

Juveniles

Total

Males

Females

Males

Unsexed

Cyrtophora moiuccensis

13

7

6(25)

35

86

Argyrodes argentatus

3

10

8

19

40

Leucauge granulata

0

3

0

8

11

Pholcus ancoralis

5

3

0

20

28

Theridion adamsoni

0

1

0

0

1

Tangaroa tahitiensis

4

11

4

52

71

144

RESEARCH NOTES

145

the exception of A. argentatus (if it is an oblig-
atory kleptoparasite), web cohabitation by so
many species on Moorea may have been en-
couraged by environmental conditions. I col-
lected webs near the end of the rainy season, and
storms that brought down both rain and bits of
vegetation were common. The web of C. mol-
uccensis is very strong (Lubin 1974); by building
within its confines the other species could use it
as a framework that could both withstand the
blows of falling objects and could deflect these
objects from their own frailer webs. It would be
interesting to determine whether cohabitation is
a less common occurrence during Moorea’s dry
season.

I thank managers Rick and Bonnie Steger and
the directorship of the University of California,
Berkeley, Field Station for access to their prop-
erty, and James Fullard for providing room and
board. Joe Beatty kindly identified the spiders,
and both he and Yael Lubin improved the manu-
script with their reviews. Research was funded
by a Natural Sciences and Engineering Research
Council of Canada (NSERC) 1967 Science and
Engineering Scholarship to H. C. P., and an
NSERC Operating Grant from the University of
Toronto to James Fullard.

LITERATURE CITED

Berry, J. W. 1987. Notes on the life history and be-
havior of the communal spider Cyrtophora mol-
uccensis (Doleschall) (Araneae, Araneidae) in Yap,
Caroline Islands. J. ArachnoL, 15:309-319.
Buskirk, R. E. 1982. Sociality in the Arachnida. Pp.
28 1-367, In Social Insects. Vol. II. (H. R. Hermann,
ed.). Academic Press, London.

Buskirk, R. E. 1986. Orb-weaving spiders in aggre-
gations modify individual web structure. J. Arach-
noL, 14:259-265.

Jackson, R. R. & R. J. Rowe. 1987. Web-invasion
and araneophagy by New Zealand and Australian
social spiders. New Zealand J. Zook, 14:139-140.

Jakob, E. M. 1991. Costs and benefits of group living
for pholcid spiderlings: losing food, saving silk. Anim.
Behav., 41:711-722.

Kaston, B. J. 1965. Some little known aspects of
spider behavior. American Midi. Natur., 73:336-
356.

Lubin, Y. D. 1974. Adaptive advantages and the evo-
lution of colony formation in Cyrtophora (Araneae:
Araneidae). Zook J. Linn. Soc., 54:321-339.

Riechert, S. E., R. Roeloffs, & A. C. Echtemacht. 1986.
The ecology of the cooperative spider Agelena con-
sociata in Equatorial Africa (Araneae, Agelenidae).
J. Arachnok, 14:201-217.

Rypstra, A. L. 1979. Foraging flocks of spiders: a
study of aggregate behavior in Cyrtophora citricola
ForskAal (Araneae; Araneidae) in West Africa. Be-
hav. Ecok Sociobiok, 5:291-300.

Manuscript received November 1991, revised January
1992.

Heather C. Proctor*: Department of Zoology,

Erindale College, University of Toronto, Mis-
sissauga, Ontario, L5L 1C6 Canada.

'Current address: Department of Biological Sciences,
University of Calgary, Calgary, Alberta, T2N 1N4
Canada.

1992. The Journal of Arachnology 20:146-147

WEB-MONITORING FORCE EXERTED BY THE SPIDER
WAITKERA WAITAKERENSIS (ULOBORIDAE)

The purpose of this study is to determine if
the resting force expressed by the primitive, orb-
weaver Waitkera waitakerensis (Chamberlain
1946) is similar to that of the orb-weaver Ulo-
borus glomosus (Walckenaer 1841) and less than
that of the triangle-weaver Hyptiotes cavatus
(Hentz 1847) (Opell 1987a). This is important
because Opell (1987a) used U. glomosus to rep-
resent orb-weavers in a study which concluded
that triangle-web spiders exert more force on a
horizontal resting line than do orb-weavers.

I chose the monotypic genus Waitkera for this
study because it is one of the two most primitive
members of the uloborid clade that is a sister
clade of the larger assemblage that includes Ulo-
borus, Hyptiotes, and Miagrammopes (Codding-
ton 1990). Unlike members of the other primi-
tive orb-weaving genus Tangaroa, which are very
small (Opell 1983), W. waitakerensis is similar
in size to the aforementioned genera (Table 1).
Like members of these genera, this species has a
well developed tracheal system, characterized by
tracheae that pass through the pedicel and enter
the legs (Opell 1979, 1987b), a pattern consid-
ered plesiomorphic for the family Uloboridae.
Like U. glomosus, W. waitakerensis hangs be-
neath the hub of its web with legs extended, while
it waits for prey to strike its web (Opell pers.
obs.). Therefore, although U. glomosus and W.
waitakerensis are phylogenetically distant, there
are no morphological or behavioral features that
suggest their resting forces should greatly differ.

Following the methods described by Opell
(1987a), I used a glass needle strain gauge to
measure the resting forces expressed by adult fe-
male W. waitakerensis. This species is found only
on New Zealand’s north island (Opell 1979),
where I studied two populations, one in a city

park in Hamilton (sample size 36) and another
from the Waitakere Mountains near Piha (sam-
ple size 19). I recorded the temperature at which
each force measurement was taken and the live
weight of each spider. These data were compared
with those of adult female U. glomosus and H.
cavatus, as measured by Opell (1987a).

Using a Shapiro-Wilk W-statistic, I first de-
termined if the resting forces of each population
or species were normally distributed. If they were,
I used a t test {i) to compare means; if they were
not, I used a Wilcoxon 2-sample test (W). Except
for the temperature at which force measurements
were taken, all values from the Hamilton and
Piha populations of W. waitakerensis were sim-
ilar (Weight: t, P = 0.178; Force: t, P ^ 0.267;
Force/weight: W, P = 0.2 1 2). Mean temperatures
were very similar (22.44 ±1.44 °C and 23.00
±0.00 °C, respectively) and their statistical dif-
ference (W, P = 0.03 1) is attributable to the uni-
form temperature at which the Piha population
was measured. Therefore, the following com-
parisons combine the values of the two W. wai-
takerensis populations.

Table 1 compares the absolute and weight-
specific resting forces of the three species. The
mean resting force of W. waitakerensis is greater
than that of U. glomosus {t, P = 0.016) but did
not differ from that of H. cavatus {t, P = 0.052).
The weight-specific resting force of W. waitak-
erensis Q. 25 X 10“5 N/mg greater than that
of U. glomosus (W, P = 0.004) and 0.77 x 10“^
N/mg less than that of H. cavatus (W, P = 0.000 1).
The mean temperatures at which the resting force
of U. glomosus and H. cavatus were measured
were nearly identical (Table 1) and that at which
W. waitakerensis was measured was only 2.5 °C
lower. Therefore, it is unlikely that temperature

Table 1.— Comparison of the weights and resting forces of three uloborid species. Mean ± standard deviation
(sample size is indicated by boldface).

Uloborus glomosus

Waitkera

waitakerensis

Hyptiotes cavatus

Live weight (mg)

Resting force (10~'* Newtons)

Resting force/live weight (10^^ N/mg)
Temperature (°C)

9.93 ± 4.65 (45)
1.07 ± 0.30(40)
1.21 ± 0.29 (40)
25.0 ± 0.7 (45)

8.92 ± 2.52 (55)
1.21 ± 0.26(57)
1.46 ± 0.41 (55)
22.6 ± 1.2 (57)

6.76 ± 3.05 (42)
1.34 ± 0.37 (42)
2.23 ± 0.69 (42)
25.2 ± 0.9 (42)

146

RESEARCH NOTES

147

differences had a major influence on the observed
differences in resting forces.

The weight-specific resting force expressed by
JV. waitakerensis is intermediate between and
statistically different from those of the orb-weav-
er U. glomosus and the triangle-weaver H. ca-
vatus. However, the weight-specific forces of the
two phylogenetically distant orb-weaving ulo-
borids are more similar to one another than ei-
ther is to that of H. cavatus. This upholds Opell’s
(1987a) conclusion that triangle-web uloborids
that actively monitor their webs express more
web-monitoring force than do orb-weaving ulo-
borids that hang from the hubs of their webs
while waiting for prey to strike. However, the
fact that W. waitakerensis expresses greater
weight-specific resting force than U. glomosus
may indicate a trend toward the reduction of
web-monitoring forces within orb-weaving ulo-
borids. Support for this is found in the reduced
tracheal systems of members of the higher orb-
weaving uloborid genera Daramuliana, Octo-
noba, Phiioponella, Ponella, Purumitra, and Zo-
sis {Opdl 1979, 1987b).

James E. Carrel and Edward K. Tillinghast
made helpful suggestions on this manuscript. I
am grateful to Denis and Jill Gibbs for their hos-
pitality during my stay in Hamilton and to Denis
Gibbs for showing me where W. waitakerensis
could be found. Permission to collect this species

in the Waitakere Ranges Centennial Memorial

Park was granted by the Auckland Regional

Council Parks Committee. This study was sup-
ported by National Science Foundation grant No.

BSR-8917935.

LITERATURE CITED

Coddington, J. A. 1990. Ontogeny and homology in
the male palpus of orb-weaving spiders and their
relatives, with comments on phylogeny (Araneocla-
da: Araneoidea, Deinopoidea). Smithsonian Contr.
Zool., 496:1-52.

Opell, B. D. 1979. Revision of the genera and tropical
American species of the spider family Uloboridae.
Bull. Mus. Comp. Zool., 148:433-549.

Opell, B. D. 1983. A review of the genus Tangaroa
(Araneae, Uloboridae). J. Arachnol., 11:287-295.

Opell, B. D. 1987a. Changes in web-monitoring forc-
es associated with web reduction in the spider family
Uloboridae. Canadian J. Zool., 65:1028-1034.

Opell, B. D. 1 987b. The influence of web monitoring
tactics upon the tracheal systems of spiders in the
family Uloboridae. Zoomorphology, 107:255-259.

Opell, B. D. 1990. The relationship of book lung and
tracheal systems in the spider family Uloboridae. J.
Morph., 206:211-216.

Brent D. Opell: Department of Biology, Vir-
ginia Polytechnic Institute and State Univer-
sity, Blacksburg, Virginia 24061 USA.

Manuscript received January 1992, revised April 1992.

1992. The Journal of Arachnology 20:148-150

STEP-COUPLED FLUCTUATIONS IN PROSOMAL PRESSURE
MAY CONSTRAIN STEPPING RATES IN WHIPSCORPIONS

(UROPYGI)

Giant whipscorpions, Mastigoproctus gigan-

teus (Lucas) (Uropygi), tend to walk slowly, using
step cycle periods of about one second or longer,
and do not maintain shorter step periods even
when stimulated to move faster (Shultz 1992).
Results from a recent electrophysiological study
(Shultz 1991) suggest a possible explanation for
this preference. Whipscorpions lack extensor
muscles at the femur-patella joint (Shultz 1989)
but extend this joint with hydraulic pressure gen-
erated through compression of the prosoma.
Baseline prosomal pressure at preferred step pe-
riods is maintained at about 60 torr above min-
imum resting pressure, but there is also a cyclical
pressure fluctuation superimposed on the ele-
vated or ‘active’ baseline (Fig. I). This fluctua-
tion, which may have a peak amplitude as high
as 20 torr over the active baseline, is apparently
caused by rapid flexion of the femur-patella joint
in members of the fourth leg pair during the re-
covery phase (protraction) of their step cycles
(Shultz 1991). I hypothesize that by maintaining
their preferred step periods, whipscorpions avoid
flexing the femur-patella joint of one member of
the fourth leg pair against the pressure surge
caused by flexion in the other member and there-
by minimize the mechanical and energetic inef-
ficiencies of hydraulic locomotion. This paper
refines quantitative predictions of the hypothesis
and presents preliminary supporting evidence
derived from a frequency distribution of step pe-
riods.

Referring to Fig. 1 , the hypothesis predicts that
whipscorpions should prefer a step period (T) in
which the decay time (hysteresis) of the pressure
surge (D) is less than the time between the end
of protraction in one member of the fourth leg
pair and the beginning of protraction in the other
(At). To ascertain empirically, one can de-
termine the difference between By (the lag time
between the onset of protraction in one member
of the fourth leg pair and the onset of protraction
in the other for a given step period) and Ct (the
duration of protraction for a given step period).
Thus for a given step period (T)

(1) At = Bt-Ct

Actual values for Bx and Cj were determined
from kinematic and electrophysiological analy-
ses of walking whipscorpions. Video analysis re-
vealed that members of the fourth leg pair step
1 80'’ out of phase at all step periods (Shultz 1991),
as is typical of segmental leg pairs in arthropods.
Thus By, the time between the onset of protrac-
tion in one leg and the onset of protraction of
the other, is always equal to about one half the
step period; that is, B^ = 0.5T.

The value Q, the duration of protraction for
a given step period, was determined by con-
ducting a least-squares regression analysis on se-
lected electromyographic parameters of the tro-
chanter-femur levator and depressor muscles
within one member of the fourth leg pair. The
levator and depressor are antagonistic muscles,
with the levator activating at the onset of pro-
traction (recovery phase) and the depressor ac-
tivating at the onset of retraction (propulsive
phase). Levator cycle duration was regarded as
the step period (T) and represented the indepen-
dent variable. The time between activation of
the levator and activation of the depressor was
regarded as the duration of protraction (Q) and
represented the dependent variable. All mea-
surements were made in seconds and were ac-
curate to ±0.005 s. Least-squares analysis of 340
steps from four individuals produced the regres-
sion equation Ct = 0.309T ± 0.024 (r^ = 0.86).

Empirically derived values for Bt and Ct were
then substituted into Equation 1

(2) At = 0.5T - (0.309T ± 0.024)

= 0.191T - 0.024

The hypothesis predicts that whipscorpions

should prefer step periods in which At is greater
than D, the decay time of the step-coupled pres-
sure surge. Thus the minimum preferred step
period can be predicted by substituting the em-
pirically derived value of D for At in Equation
2 and then solving the equation for T. The value
for D was determined by recording prosomal

148

RESEARCH NOTES

149

m

I \/VA/\/VA/^

R4

L4

Aj I — I Bj [— I Cj I I

J

►

TIME

Figure 1 . — Diagram illustrating the relationship between step-coupled fluctuations in prosomal pressure and
stepping pattern of the fourth leg pair. Each pressure surge is caused by flexion of the femur-patella joint in one
of the fourth walking legs during the recovery phase of the step cycle. The pressure surges labled ‘L’ are caused
by flexion of the left fourth leg (L4) and those labled ‘R’ are caused by flexion in the right fourth leg (R4). The
time needed for the pressure surge to decay is indicated by ‘D’. In the step diagram for R4 and L4, the black
bars represent periods of retraction (propulsive phase) and the open regions represent protraction (recovery
phase). Variables associated with the step diagram include Aj (time between end of protraction in one of the
fourth legs and beginning of protraction in the other at step period T), (time between beginning of protraction
in one of the fourth legs and beginning of protraction in the other at step period T) and Cj (duration of protraction
in the fourth leg pair at step period T). The hypothesis developed here predicts that whipscorpions should prefer
step cycles (T) in which D is less than A^, where Aj = Bj - C^. At step periods where D is greater than A^,
pressure surges will overlap causing an increase in the pressure baseline and requiring greater exertion by flexor
muscles.

pressure from a freely walking whipscorpion
through a pressure transducer affixed to the ar-
throdial membrane of the femur-patella joint in
leg 1 (see Shultz 1991 for details). Pressure decay
times were measured in 1 7 steps where step pe-
riod was sufficiently long that the pressure surge
decayed to the active pressure baseline. Mean
decay time of the pressure surge was found to be
0.130 s (SE = 0.006). This value was then sub-
stituted for Aj in Equation 2, and the equation
was solved for T. On the basis of these data, the
hypothesis predicts that whipscorpions should
prefer step periods greater than about 0.80 s.

This prediction was compared to a frequency
distribution of step periods {n — 483) from elec-
tromyographic records of seven whipscorpions.

The structure of the distribution is consistent
with the predictions of the hypothesis (Fig. 2).
The number of steps with Tong’ cycle periods
(i.e., greater than 0.8 s) is substantially greater
than those with ‘short’ cycle periods (i.e., less
than 0.8 s). More importantly, a preference for
‘long’ over ‘short’ step periods is suggested by a
pronounced increase in the frequency of steps in
the 0.80-0.89 s range compared to the 0.70-0.79
s range.

These results are consistent with the hypoth-
esis that Mastigoproctus uses long step periods
to avoid mechanical and energetic inefficiencies
associated with flexing the femur-patella joint of
one member of the fourth leg pair against high
prosomal pressures generated by flexion in the

150

THE JOURNAL OF ARACHNOLOGY

PERIOD (sec)

Figure 2. — Frequency distribution of step cycle pe-
riods representing 483 steps from seven whipscor-
pions. The hypothesis developed here predicts that
whipscorpions should prefer step periods longer than
0.80 s. The structure of the frequency distribution is
consistent with this prediction, but this comparison
cannot be regarded as a statistical test of the hypothesis.

other. If whipscorpions were to walk at shorter
step periods (i.e., higher speeds), step-coupled
pressure surges would overlap thereby increasing
the active baseline pressure and requiring greater

forces from the flexor muscles. It is not known
whether these results apply to spiders and other
arachnids that use hydraulic pressure for leg ex-
tension.

I thank Abbot S. Gaunt, Thomas E. Hether-
ington and R. Z. German for equipment loans.
This research was supported by grants-in-aid of
research from the Exline-Frizzell Fund for Ar-
achnological Research (California Academy of
Sciences) and Sigma Xi.

LITERATURE CITED

Shultz, J. W. 1989. Morphology of locomotor ap-
pendages in Arachnida: evolutionary trends and
phylogenetic implications. Zool. J. Linn. Soc., 97:
1-56.

Shultz, J. W. 1991. Evolution of locomotion in
Arachnida: the hydraulic pressure pump of the giant
whipscorpion, Mastigoproctus giganteus (Uropygi).
J. Morph. 210:13-31.

Shultz, J. W. 1992. Muscle firing patterns in two
arachnids using different methods of propulsive leg
extension. J. Exp. Biol., 162:313-329.

Jeffrey W. Shultz: Department of Biological Sci-
ences, University of Cincinnati, Cincinnati,
Ohio 45221-0006, USA

Manuscript received March 1991, revised November
1991.

1992. The Journal of Arachnology 20:151-152

FEMALE SPIDERS (ARANEAE: DIPLURIDAE, DESIDAE,
LINYPHIIDAE) EAT THEIR OWN EGGS

Some spiders may eat the eggs of other spiders
(Hallas 1988; Willey & Adler 1989; Jackson
1990), but female spiders have rarely been re-
ported to consume their own eggs. Shaw (1989)
observed an egg-guarding female of Clubiona re-
clusa O. P.-Cambridge (Clubionidae) consume
half of her eggs, and Downes (1987) observed
females of the genus Crossopriza, presumably C.
lyoni, (Pholcidae) consuming one or more of their
eggs.

We observed females of three additional ara-
neomorph species, Florinda coccinea (Hentz)
(Linyphiidae), Ixeuticus martins (Simon) (Desi-
dae) and Ixeuticus robustus (L. Koch) (Desidae),
and one mygalomorph species, Thelechoris stria-
tipes (Simon) (Dipluridae), consuming entire
clutches of their own eggs and all egg-sac silk
(Table 1). Some of these oophagous females also
produced other egg sacs, both fertile and infertile,
that were not consumed and, as noted in Table
1 , the sequence of oviposition of these sacs was
variable. All of the egg-eating females of F. coc-
cinea, /. martins and I. robustus had mated once,
but we do not know whether the T. striatipes
females had mated.

Florinda coccinea females (n = 66) were re-
moved from their egg sacs within 1 2 h of con-
struction, and I. martins (« = 21) and 1. robustus
{n = 13) females were removed from their egg
sacs within 72 h of construction; therefore, ad-
ditional oophagous events might occur when fe-
males are allowed constant access to their sacs.
Thelechoris striatipes {n= 13) females remained
with their egg sacs until spiderling emergence,
and emergence required 14 days and 17 days for
two clutches where developmental time was cal-
culated.

Gravid F. coccinea females were collected by
MBW on 8 May 1989 in Clemson, South Car-
olina, and maintained in the laboratory (26 ± 2
°C, 65 ± 4% relative humidity, 14L:10D pho-
toperiod) in Clemson. Spiders were held indi-
vidually in plastic containers (3.7 cm deep x 5.2
cm diameter). The first female of F. coccinea to
consume her eggs was field-collected; the second
was an F2 female reared and mated in the lab-
oratory. Gravid I. martins and I. robustus fe-
males were collected by MBW in April 1991 in

Christchurch, New Zealand, and maintained in-
dividually in 4.5 cm deep x 9.75 cm diameter
plastic containers in the laboratory in Clemson;
the five females that consumed eggs were from
the F2 generation. All females of the above spe-
cies were given constant access to moist cotton,
were fed daily, and were offered a variety of prey,
including German cockroaches, house flies, and
tachinid flies.

The T. striatipes females were collected by FAC
on 15 April 1989 in Kenya, Africa (Tsavo West
National Park at Kitani Lodge) and maintained
in 31 cm length x 16 cm width x 8 cm deep
plastic shoe boxes in the laboratory (24 ± 2 °C,
1 2L: 1 2D photoperiod) in Cullowhee, North Car-
olina. These spiders were fed one Tenebrio larva
every 10 days, occasionally supplemented by
cricket nymphs or house flies. Spiders had con-
stant access to moist cotton.

Because web construction and oviposition be-
havior of the four species were similar to these
behaviors in the field, we did not add substrata
to any containers. All females attached their webs
to the sides of the containers.

In nature, F. coccinea constructs sheet webs in
low vegetation and oviposits among the vege-
tation, rather than retaining egg sacs in the web.
Spiders drop from their webs when disturbed, so
females could possibly contact their sacs in the
vegetation. In the laboratory, egg sacs were con-
structed in the bottom of the container, and fe-
males did not remain in close proximity to their
egg sacs. Ixeuticus martins females live in funnel
retreats in crevices and construct their egg sacs
within the retreat (Forster 1970). Ixeuticus ro-
bustus females also live in funnel retreats in crev-
ices, and it is likely that they construct egg sacs
within the retreat because females in the labo-
ratory remain in close contact with their egg sacs.
Both /. martins and 7. robustus females have been
observed killing prey and then moving away from
the prey to allow their spiderlings to feed (MBW,
pers. obs.), so it is likely that females remain in
contact with their spiderlings in nature.

In nature, T. striatipes females live in perennial
funnel webs, construct their egg sacs in the wall
of their tubular silken retreats, and remain with
the brood through spiderling emergence and dis-

151

152

THE JOURNAL OF ARACHNOLOGY

Table 1.— Oophagy by Thelechoris striatipes (Dipluridae), Florinda coccinea (Linyphiidae), Ixeuticus martins
(Desidae) and Ixeuticus robustus (Desidae). a = unknown which day within given range the eggs were consumed,
b = Fertile (F), Infertile (I), Consumed (*), c = females were sacrificed within a month after the egg sacs were
consumed. Under I. robustus, female #2 produced and ate two sacs.

Species and
female

identification

Date of oviposition
of consumed sacs

Date of consumption

Sequence of sacs produced*

Thelechoris striatipes'^

1 10 July 1989

10-11 July 1989

♦

2

12 July 1989

15-20 July 1989

*

3

18 July 1989

18-20 July 1989

*

Florinda coccinea

1

16 May 1989

16-19 May 1989

*,F,F,F,I,I,I

2

6 Aug. 1989

7 Aug. 1989

F,*,F,F

Ixeuticus martins

1

26 Nov. 1989

26-29 Nov. 1991

2

2 Jan. 1992

2-4 Jan. 1992

F,F,I,I,F,F,*,I,I,I

3

23 April 1992

24-28 April 1992

I, I, I,*

Ixeuticus robustus

1

26 Nov. 1991

26-30 Nov. 1991

F,F,*,F,F

2

13 Dec. 1991

13-18 Dec. 1991

*,*,F,F,F,I

2

23 Dec. 1991

23-27 Dec. 1991

persal (FAC, pers. obs.). In the laboratory, fe-
males constructed capture webs and retreats in
the shoe box arenas. Spiders captured prey in
these webs, and often constructed egg sacs in
their retreats and spent much of their inactive
time upon or near the sacs.

The egg-eating behavior we observed was pos-
sibly triggered by abnormal conditions associ-
ated with captivity and is rarely, if ever, practiced
in nature. However, if oophagy were directed
toward infertile, damaged, or otherwise inviable
eggs, it could be an adaptive strategy to recycle
nutrients and thereby decrease losses. We do not
know whether the consumed eggs were fertile,
and we have no evidence that females are capable
of detecting whether their eggs are fertile. How-
ever, the fact that several of the infertile clutches
produced by four of the oophagous females were
not consumed (Table 1) suggests that oophagy
has not evolved as a consistent response to clutch
infertility.

We thank P. H. Adler, L. Higgins, and P. A.
Zungoli for reviewing the manuscript. This is
technical contribution No. 3291 of the South
Carolina Agricultural Experiment Station, Clem-
son University.

LITERATURE CITED

Downes, M. F. 1987. Crossopriza {lyonil) (Araneae,

Pholcidae) eats her own eggs. J. Arachnol., 15:276.

Forster, R. R. 1970. The spiders of New Zealand,
Part III. Otago Museum Bulletin No. 3, Dunedin,
New Zealand. 1 84 pp.

Hallas, S. E. A. 1988. The ontogeny of behaviour in
Portia fimbriata, P. labiata, and P. schultzi, web-
building jumping spiders (Araneae: Salticidae). J.
Zool., London, 215:231-238.

Jackson, R. R. 1990. Predatory versatility and intra-
specific interactions in Cyrba algerina and Cyrba
ocellata, web-invading spartaeine jumping spiders
(Araneae: Salticidae). New Zealand J. Zool., 17:157-
168.

Shaw, M. R. 1989. Why did a Clubiona reclusa eat
her own eggs? Newsl. British Arachnol. Soc., 56:6,7.

Willey , M. B. & P. H. Adler. 1989. Biology of Peu-
cetia viridans (Araneae, Oxyopidae) in South Car-
olina, with special reference to predation and ma-
ternal care. J. Arachnol., 17:275-284.

Marianne B. Willey: Department of Entomol-
ogy, 1 14 Long Hall, Clemson University, Box
340365, Clemson, South Carolina, 29634-0365
USA

Frederick A. Coyle: Department of Biology,
Western Carolina University, Cullowhee,
North Carolina 28723 USA

Manuscript received May 1992, revised July 1992.

INSTRUCTIONS TO AUTHORS

Manuscripts are preferred in English but may be accepted
in Spanish, French or Portuguese subject to availability of
appropriate reviewers. Authors whose primary language is not
English, may consult the Associate Editor for assistance in
obtaining help with English manuscript preparation. All man-
uscripts should be prepared in general accordance with the
current edition of the Council of Biological Editors (CBE) Style
Manual unless instructed otherwise below. Authors are ad-
vised to consult a recent issue of the Journal of Arachnology
for additional points of style. Send three identical copies of
the typed material together with copies of illustrations to:

Gary L. Miller, Associate Editor
The Journal of Arachnology
Department of Biology
The University of Mississippi
University, Mississippi 38677 USA
Telephone: (601) 232-7495
FAX: (601) 232-5144
Bitnet: BYMILLER@UMSVM
Correspondence relating to manuscripts should also be di-
rected to the Associate Editor.

Manuscripts longer than 1 500 words should be prepared as
Feature Articles, shorter papers as Research Notes.

FEATURE ARTICLES

Title page.— Place in the upper left comer the complete name,
address, and telephone number of the author with whom proofs
and correspondence should be exchanged. Include FAX num-
bers and Bitnet addresses if available. Four lines below, type
the title, centered, all in capital letters. Another four lines be-
low, beginning with the left margin, type each author’s name
on a new line followed by the address.

Abstract.— The heading should be placed at the beginning
of the first paragraph set off by a period. In articles written in
English, a second abstract in an acceptable language may be
included pertinent to the nationality of the author(s) or geo-
graphic region(s) emphasized in the article. A second abstract
in English must be included in articles not written in the latter
language.

Text.— Double-space text, tables, legends, etc. throughout.
Three levels of heads are used. The first level (METHODS,
RESULTS, etc.) is typed in capitals and on a separate line.
The second level head begins a paragraph with an indent and
is separated from the text by a period and a dash. The third
level may or may not begin a paragraph but is underscored
and separated from the text by a colon. Use only the underline
method to indicate italics; DO NOT USE WORD PROCES-
SOR ITALICS CODES. Use the metric system unless quoting
text or referencing collection data. All decimal fractions are
indicated by the period regardless of language of the text.

Citation of references in the text.— Include within parenthe-
ses the surnames of the author(s) followed by the date of pub-
lication. A comma separates multiple citations by the same
author(s) and a semicolon separates citations by different au-
thors, e.g., (Smith 1970), (Jones 1988; Smith 1989), (Smith
1986, 1987; Smith & Jones 1989; Jones et al. 1989).

Literature cited section.— Use the following style:
Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid
composition of major ampullate gland silk (dragline) of Ne-
phila clavipes (Araneae, Tetragnathidae). J. Arachnol., 18:
297-306.

Kraffit, B. 1982. The significance and complexity of com-
munication in spiders. Pp. 1 5-66, In Spider Communica-
tions: Mechanisms and Ecological Significance. (P. N. Witt
& J. S. Rovner, eds.). PrincetonUniversity Press, Princeton.
Footnotes. — Footnotes are only permitted on the first printed
page to indicate current address or other information con-
cerning the author(s). All footnotes are placed together on a
separate manuscript page.

Running head.— Author sumame(s) and an abbreviated title
must be typed all in caps and must not exceed 60 characters
and spaces. The running head is placed by itself on a separate
manuscript page.

Taxonomic articles.— Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Associate
Editor.

Tables.— Each table, with the legend above, should be placed
on a separate manuscript page. Only the appropriate horizontal
lines should be included. Use no footnotes; instead, include
all information in the legend. Make notations in the text mar-
gins to indicate the preferred location of tables in the printed
text.

Illustrations.— Original art work should be kept by the au-
thor(s) until final acceptance for publication. All art work must
be mounted on stiffboard ready for reproduction. In line draw-
ings, pay particular attention to width of lines and size of
lettering when reductions are to be made by the printer. Mul-
tiple photos assembled on a single plate should be mounted
with only a minimum of space separating them. In the case of
multiple illustrations mounted together, each illustration must
be numbered sequentially rather than given an alphabetic se-
quence. Written on the back should be the name(s) of author(s)
and an indication of top edge. The author should indicate
whether the figure should be one column or two columns in
width. The overall dimensions should be no more than 1 1
inches (28 cm) x 14 inches (36 cm). Larger drawings present
greater difficulty in shipping and greater risks of damage for
which the JOA assumes no responsibility. The author is en-
couraged to consider PMT (PhotoMechanical Transfer) repro-
ductions for large drawings and to consult an Editor for details
on appropriate reductions. Make notations in the text margins
to indicate the preferred position of illustrations in the printed
text.

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Adhere to the
following styles:

Figures \-4.—A-us x-us, male from Timbuktu: 1, left leg;
2, right chelicera; 3, dorsal aspect of genitalia; 4, ventral aspect
of abdomen.

Figures 27-34.— Right chelicerae of species of A-us from
Timbuktu: 27, 29, 3 1 , 33, dorsal views; 28, 30, 32, 34, prolater-
al views of moveable finger; 27, 28, A-us x-us, holotype male;
29, 30, A-us, w-us, holotype male; 31, 32, A-us z-us, holotype
male; 33, 34, A-us y-us, male. Scales = 1.0 mm.

Assemble manuscript for mailing.— Assemble the separate
sections or pages in the following sequence; title page, abstract,
text, figure legends, footnotes, running head, tables with leg-
ends, figures.

Page charges and reprints.- The current charge per journal
page will be assessed as follows: $50— nonmembers, $45 —
members with grant, institutional, or other support for page
costs; $25— regular members who do not have support for page
costs; $20— student members who do not author with full
members and who do not have support. Reprints are available
only from the printer at current costs.

Waiver of Page Charges.— The Editor will consider requests
for a waiver of page charges for manuscripts which have been
accepted for publication. Requests should be made in writing
and should be submitted to the Associate Editor along with
the final revision of the manuscript. The author must certify
that neither grant nor institutional funds are available to cover
all of the publication costs. The Society will normally agree to
waive up to a maximum of 80% of the page charges, contingent
on the availability of funds; the author is expected to find
alternative sources for the remainder.

RESEARCH NOTES

Instructions above pertaining to feature articles apply also
to research notes except that abstracts and most headings are
not used and the byline follows the Literature Cited section.

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 20 Feature Articles NUMBER 2

Temporal and spatial segregation of web-building in a community of orb-
weaving spiders, David Ward and Yael Lubin 73

Habitat segregation by species of Metaphidippus (Araneae: Salticidae) in

Minnesota, Bruce Cutler and Daniel T. Jennings 88

Developmental plasticity and fecundity in the orb-weaving spider Nephila

clavipes, Linden E. Higgins 94

Ballooning: Data from spiders in freefall indicate the importance of posture,

Robert B. Suter 107

A revision of some species of Roncus L. Koch (Neobisiidae, Pseudoscor-
piones) from North America and South Europe, Bozidar P. M. Curcic,
Rajko N. Dimitrijevic, and Ozren S. Karamata 114

Notes on mating and reproductive success of Ceropelma longisternalis (Ara-
neae, Theraphosidae) in captivity, Fernando G. Costa and Fernando
Perez-Miles 129

Descripcion del macho de Achaearanea jequirituba (Araneae, Theridiidae),

Alda Gonzalez 134

A review of Meta (Araneae, Tetragnathidae), with description of two new

species, Yuri M. Marusik and Seppo Koponen 137

Research Notes

Cohabitation of six species of spiders in webs of Cyrtophora moluccensis

(Araneae, Araneidae) in Moorea, French Polynesia, Heather C. Proctor 144

Web-monitoring force exerted by the spider Waitkera waitakerensis (Ulo-

boridae), Brent D. Opell 146

Step-coupled fluctuations in prosomal pressure may constrain stepping rates

in whipscorpions (Uropygi), Jejfrey W. Shultz 148

Female spiders (Araneae: Dipluridae, Desidae, Linyphiidae) eat their own

eggs, Marianne B. Willey and Frederick A. Coyle 151

L.

:H)T

The Journal of
ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 20

1992

NUMBER 3

THE JOURNAL OF ARACHNOLOGY

EDITOR: James W. Berry, Butler University

ASSOCIATE EDITOR: Gary L. Miller, The University of Mississippi

EDITORIAL BOARD: A. Cady, Miami (Ohio) Univ. at Middletown; J. E.
Carrel, Univ. Missouri; J. A. Coddington, National Mus. Natural Hist.; J. C.
Cokendolpher, Lubbock, Texas; F. A. Coyle, Western Carolina Univ.; C. D. Don-
dale, Agriculture Canada; W. G. Eberhard, Univ. Costa Rica; M. E. Galiano,
Mus. Argentine de Ciencias Naturales; M. H. Greenstone, BCIRL, Columbia,
Missouri; N. V. Horner, Midwestern State Univ.; D. T. Jennings, Garland, Maine;
V. F. Lee, California Acad. Sci.; H. W. Levi, Harvard Univ.; E. A. Maury, Mus.
Argentine de Ciencias Naturales; N. 1. Platnick, American Mus. Natural Hist.;
G. A. Polis, Vanderbilt Univ.; S. E. Riechert, Univ. Tennessee; A. L. Rypstra,
Miami Univ., Ohio; M. H. Robinson, U.S. National Zool. Park; W. A. Shear,
Hampden-Sydney Coll.; G. W. Uetz, Univ. Cincinnati; C. E. Valerio, Univ. Costa
Rica.

The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the
study of Arachnida, is published three times each year by The American Arach-
nological Society. Memberships (yearly): Membership is open to all those inter-
ested in Arachnida. Subscriptions to The Journal of Arachnology and American
Arachnology (the newsletter), and annual meeting notices, are included with mem-
bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or
$90 (all other countries). Inquiries should be directed to the Membership Secretary
(see below). Back Issues: Susan E. Riechert, Department of Zoology, Univ. of
Tennessee, Knoxville, TN 37916 USA. Undelivered Issues: Allen Press, Inc.,
1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA.

THE AMERICAN ARACHNOLOGICAL SOCIETY

PRESIDENT: Allen R. Brady (1991-1993), Biology Department, Hope College,
Holland, Michigan 49423 USA.

PRESIDENT-ELECT: James E. Carico (1991-1993), Department of Biology,
Lynchburg College, Lynchburg, Virginia 24501 USA.

MEMBERSHIP SECRETARY: Norman 1. Platnick (appointed), American
Museum of Natural History, Central Park West at 79th St., New York, New
York 10024 USA.

TREASURER: Gail E. Stratton (1991-1993), Department of Biology, Albion
College, Albion, Michigan 49224 USA.

SECRETARY: Brent Opell (1991-1993), Department of Biology, Virginia Poly-
technic Institute and State University, Blacksburg, Virginia 24061 USA.
ARCHIVIST: Vincent D. Roth, Box 136, Portal, Arizona 85632 USA.
DIRECTORS: Matthew H. Greenstone (1990-1992), George W. Uetz (1991-
1993), Charles E. Griswold (1991-1993).

HONORARY MEMBERS: C. D. Dondale, W. J. Gertsch, H. Homann, H. W.
Levi, A. F. Millidge, M. Vachon, T. Yaginuma.

Cover illustration: SEM photomicrograph of the ocularium of the opilionid Odiellus pictus (Wood).
The species has a prominent trident of spines at the anterior border of its cephalothorax. Found in
the eastern United States. Photo by Steven Murphree.

Publication date: 2 April 1993

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.

1992. The Journal of Arachnology 20:153-156

SPIDER (ARANEAE) TAXA ASSOCIATED WITH
MANTISPA VIRIDIS (NEUROPTERA: MANTISPIDAE)

Jeffrey R. Brushwein', Kevin M. Hoffman and Joseph D. Culin: Department of
Entomology, Clemson University, Clemson, South Carolina 29634-0365 USA

ABSTRACT. Egg sacs of 25 species of spiders in 14 families were found to contain immatures of Mantispa
viridis in northwestern South Carolina, bringing the total spider taxa associated with this species to at least 29
species in 15 families. Thirty-one of the 124 M. viridis infested egg sacs had two or more mantispids in them.
However, only three of these sacs produced two or more adult mantispids, with two sacs producing two adults
each and the third sac producing four adults.

Larvae of Mantispa viridis Walker, a member
of the mantispid subfamily Mantispinae, are
predators of spider eggs. First instars of M. viridis
locate spider egg sacs and penetrate through the
surrounding silk to gain access to the eggs, where-
upon they develop through two additional and
relatively immobile instars prior to pupation
within the sac (Richardson 1976; Redborg &
MacLeod 1985). Ten species of spiders in the
families Agelenidae, Araneidae, Clubionidae,
Ctenidae, Lycosidae, Theridiidae, and either the
Clubionidae or Gnaphosidae have been associ-
ated with M. viridis (Milliron 1940; Stein 1955;
Parfin 1958; Valerio 1971; Tolbert 1976; Hieber
1984; Redborg & MacLeod 1985; Roble 1986;
Hoffman & Brushwein 1 992).

In the first reported rearing of mantispine lar-
vae, Brauer (1869) noted that although more than
one first instar of the European Mantispa styriaca
(Poda) would enter single egg sacs in the labo-
ratory, only one would develop to the adult. Lat-
er, three to eight larvae of Eumantispa harmandi
(Navas) were reported developing inside single
egg sacs, although no information was given on
larval survival or adult emergence (Kishida 1929;
K. Kishida, pers. comm, in Bristowe 1932). Sub-
sequent studies have documented the emer-
gences of from two to seven adult mantispids
from single egg sacs (McKeown & Mincham 1 948;
Downes 1985; Monserrat & Diaz- Aranda 1989),
including M. viridis (Parfin 1958; Richardson

‘Current address: 517 Lake Avenue, Lehigh Acres,
Florida 33936 USA.

1 976; W. W. Tolbert, pers. comm, in Richardson
1976).

Previous studies on M. viridis indicate that this
species feeds on eggs of a broad taxonomic and
ecological range of spiders and that more than
one larva can successfully develop inside single
egg sacs. The present paper reports the results of
field studies conducted from 1982 through 1986
to document the spider taxa exploited by M.
viridis in northwestern South Carolina, to deter-
mine the frequency with which multiple larvae
attack single egg sacs, and to determine the num-
ber of adults which successfully develop in such
multiply-infested sacs.

METHODS

Spider egg sacs and associated female spiders
were collected from 1 982 through 1 986 by visual
searching in various habitats within a 20 km
radius of Clemson, South Carolina. The most
frequently searched locations were wooded areas
bordering Lake Hartwell and fields along South
Carolina State Highway 123. Microhabitats sam-
pled included foliage of hardwoods and conifers,
ornamental shrubs, herbaceous vegetation, be-
neath tree bark, on the surface of the ground,
under stones and fallen logs, in burrows of Geo-
lycosa sp. (Lycosidae), and the outside surfaces
of various buildings. All egg sacs located during
these searches were collected. Egg sacs were
opened in the laboratory and examined with a
stereomicroscope. Sacs with mantispids inside
were retained and the number of larvae and co-
coons present were recorded, and those without

153

154

THE JOURNAL OF ARACHNOLOGY

Table 1. — Spider taxa associated with the immature stages of Mantispa viridis Walker. Taxa are arranged
alphabetically, and incorporate taxonomic changes compiled by Platnick ( 1 989) and Dondale and Redner ( 1 990).

Family

Genus and species

Reference

Agelenidae

Agelenopsis sp.,

prob. pennsylvanica (C. L. Koch)

Parfin 1958

Agelenopsis sp.

this report

Anyphaenidae

Teudis mordax (O. P. -Cambridge)

this report

Araneidae

Araneus pegnia (Walckenaer)

this report

Araniella displicata (Hentz)

this report

Argiope aurantia Lucas

Tolbert 1976, this report

Argiope trifasciata (ForskSl)

Tolbert 1976

Cyclosa turhinata (Walckenaer)

this report

Cyclosa sp., prob. turbinata

this report

Metepeira labyrinthea (Hentz)

this report

Neoscona arabesca (Walckenaer)

this report

unidentified genus, prob. Neoscona

this report

Clubionidae

Cheiracanthium inclusum (Hentz)

this report

Clubiona sp.

Hoffman and Brushwein 1992

Clubionidae or Gnaphosidae, undetermined

Stein 1955

Corinnidae

Castianeira sp.

this report

Ctenidae

Cupiennius salei (Keyserling)

Milliron 1940

Lycosidae

Gladicosa pulchra (Keyserling)

Roble 1986

Varacosa avara (Keyserling)

Hoffman and Brushwein 1992

unidentified genus

this report

Oxyopidae

Peiicetia viridans (Hentz)

Fink 1968, 1987, this report

Philodromidae

Philodromus imbecillus Keyserling

this report

Pisauridae

Pisanrina mira (Walckenaer)

this report

Salticidae

Habronattus coecatus (Hentz)

this report

Phidippus clarus Keyserling

this report

Phidippus mystaceus (Hentz)

this report

Plexippus paykulli (Audouin)

this report

Tetragnathidae

Tetragnatha sp.

this report

Theridiidae

Achaearanea nipicola (Emerton)

this report

Achaearanea tepidariorum (C. L. Koch)

Valerio 1971, this report

Latrodectus mactans (Fabricius)

this report

Thomisidae

Misumenoides fonnosipes (Walckenaer)

this report

Tmarus angidatus (Walckenaer)

this report

Uloboridae

Uloborus glomosus (Walckenaer)

this report

mantispids were discarded. The numbers and
identities of uninfested sacs were not recorded,
but an estimated 350-700 egg sacs were exam-
ined during the course of the study. Egg sacs with
larvae were placed in larval rearing cells while
mantispid cocoons were placed in vials designed
for maintaining adult mantispids. Rearing con-
tainers and environmental conditions were as
described by Brushwein & Culin (1991). First
instars of M. viridis were identified by the dorsal
banding pattern on the thorax and abdomen, and
second and third instars were identified by the
characteristic shapes of the thoracic legs and tenth
abdominal segments of each instar (Hoffman &
Brushwein 1992).

In cases where field-collected egg sacs which
contained mantispids were not associated with
female spiders but still contained viable eggs or
spiderlings, surviving spiders were reared to ma-
turity on a variety of larval Lepidoptera and adult
Diptera. Rearing conditions and procedures were
the same as those used to maintain adult man-
tispids (Brushwein & Culin 1991). Spiders were
identified by using the keys of Kaston (1948,
1978) and Roth (1985) and by comparison with
previously identified specimens in the Clemson
University Arthropod Collection (CUAC), De-
partment of Entomology. Voucher specimens of
immature and adult mantispids and the associ-
ated spiders are deposited in the CUAC.

BRUSH WEIN ET AL. -SPIDER ASSOCIATIONS OF M ANTISPA VIRIDIS

155

Table 2. — Incidence and magnitude of multiple infestations of single spider egg sacs by M. viridis immatures
and the maximum number of adult mantispids reared per sac.

Family

Genus and species

Number
of sacs
infested

Number
of sacs
with > 1

Maximum Maximum
number number
inside sac of adults

Agelenidae

Agelenopsis sp.

58

21

16

1

Araneidae

Argiope aurantia Lucas

4

1

5

1

Metepeira labyrinthea (Hentz)

21

3

2

1

Neoscona arabesca (Walckenaer)

1

1

2

1

unidentified, prob. Neoscona

1

1

4

4

Pisauridae

Pisaurina mira (Walckenaer)

3

1

2

2

Salticidae

Phidippus darns Keyserling

9

1

3

1

Theridiidae

Latrodectus mactans (Fabricius)

6

2

2

2

RESULTS AND DISCUSSION

Egg sacs of 1 24 spiders contained immatures
of M. viridis in the Clemson area. These spiders
belonged to 25 species in 23 genera representing
14 families, bringing the total spider taxa asso-
ciated with M. viridis to at least 29 species in 26
genera from 15 families (Table 1). Three of the
species were previously associated with M. vir-
idis and 20 are newly associated, while the status
of the unidentified species of Agelenopsis and of
Lycosidae as previously or newly associated taxa
can not be clarified in the absence of species-
level identifications. Eight species had more than
one egg sac associated with M. viridis. Six of these
eight species had at least one egg sac infested
with two or more immatures and are listed in
Table 2. The other two species were the uniden-
tified Cyclosa species with two singly-infested egg
sacs in a single web and Peucetia viridans (Hentz)
with two singly-infested sacs.

Egg sacs containing more than one M. viridis
larva were relatively common and accounted for
25% (31 of 124) of the total number of infested
sacs (Table 2). However, multiple adults were
reared from only 9.7% (3 of 3 1) of the multiply-
infested sacs. Also, although as many as 1 6 im-
matures were found inside single sacs, no more
than four developed into adults from any one
sac. Failure of larvae to develop in multiply-
infested sacs was most likely due to either star-
vation or intraspecific aggression. First instars
become relatively immobile shortly after feeding
commences and second and third instars possess
very reduced legs. Therefore, developing larvae
are trapped inside egg sacs and are vulnerable to
starvation if the available eggs are depleted by
other larvae. Multiple adults of M. viridis were

able to develop in single egg sacs of Pisaurina
mira (Walckenaer), Latrodectus mactans (Fabri-
cius), and an unidentified large araneid, possibly
because the spiders are relatively large and pro-
duce large egg sacs. Mortality caused by conspe-
cifics also may play a role in multiply-infested
sacs. Richardson (1976) noted that it was not
uncommon for second and third instars of M.
viridis to kill other larvae when reared together
in the laboratory. Unfortunately, many of the
larvae in multiply-infested sacs in the present
study were already dead and somewhat dessi-
cated when the sacs were first examined, making
a conclusive determination of the cause of their
fate impossible.

ACKNOWLEDGMENTS

We thank Kurt E. Redborg, Coe College; Stev-
en Roble, Carnegie Museum of Natural History;
and Mitchell E. Roof and B. Merle Shepard, both
of Clemson University, for their helpful com-
ments on earlier drafts. This is Technical Con-
tribution No. 3117 of the South Carolina Agri-
cultural Experiment Station, Clemson University.

LITERATURE CITED

Brauer, F. 1869. Beschreibung der Verwandlungs-
geschichte der Mantispa styriaca Poda und Betrach-
tungen iiber die sogenannte Hypermetamorphose
Fabre’s. Verb. Zool.-Bot. Ges. Wien, 19:831-840.
Bristowe, C. S. 1932. Mantispa, a spider parasite.

Entomol. Mon. Mag., 68:222-224.

Brushwein, J. R. & J. D. Culin. 1991. Modified rear-
ing and maintenance techniques for Mantispa viridis
(Neuroptera: Mantispidae). Florida Entomol., 74:
446-452.

Dondale, C. D. & J. H. Redner. 1990. The insects
and arachnids of Canada. Part 1 7. The wolf spiders.

156

THE JOURNAL OF ARACHNOLOGY

nurseryweb spiders, and lynx spiders of Canada and
Alaska. Araneae: Lycosidae, Pisauridae, and Oxy-
opidae. Agric. Canada Publ. 1856. 383 pp.

Downes, M. F. 1985. Emergence of /I

imbecilla (Gerstaecker) (Neuroptera: Mantispidae)
from the retreat web of Mopsus pencillatus (Karsch)
(Araneae: Salticidae). Australian Entomol. Mag., 12:
54.

Fink, L. S. 1986. Costs and benefits of maternal be-
haviour in the green lynx spider (Oxyopidae, Peu-
cetia viridans). Anim. Behav., 34:1051-1060.

Fink, L. S. 1987. Green lynx spider egg sacs: sources
of mortality and the function of female guarding
(Araneae, Oxyopidae). J. ArachnoL, 15:231-239.

Hieber, C. S. 1984. Egg predators of the cocoons of
the spider Mecynogea lemniscata (Araneae: Ara-
neidae): rearing and population data. Florida En-
tomol., 67:176-178.

Hoffman, K. M. & J. R. Brushwein. 1992. Descrip-
tions of the larvae and pupae of some North Amer-
ican Mantispinae (Neuroptera: Mantispidae) and
development of a system of larval chaetotaxy for
Neuroptera. Trans. American Entomol. Soc., 118:
159-196.

Kaston, B. J. 1948. Spiders of Connecticut. Con-
necticut St. Geol. Nat. Hist. Surv. Bull. 70. 874 pp.

Kaston, B. J. 1978. How to Know the Spiders, 3rd
ed. Brown Publ. Co., Dubuque, Iowa. 272 pp.

Kishida, K. 1929. On the oviposition of a clubionid
spider, Chiracanthium rubicundulum. Lansania,
Tokyo, 1:73-74. (in Japanese).

McKeown, K. C. & V. H. Mincham. 1948. The bi-
ology of an Australian mantispid (Mantispa vittata
Guerin). Australian Zook, 1 1:207-224.

Milliron, H. E. 1 940. The emergence of a neotropical
mantispid from a spider egg sac. Ann. Entomol. Soc.
America, 33:357-360.

Monserrat, V. J. & L. M. Diaz-Aranda. 1989. Esta-
dios larvarios de los neuropteros ibericos. V: Man-
tispa styriaca (Poda, 1761) (Planipennia: Mantis-
pidae). Neuroptera Int., 5:189-204.

Parfin, S. 1958. Notes on the bionomics of the Man-
tispidae (Neuroptera: Planipennia). Entomol. News,
69:203-207.

Platnick, N. I. 1989. Advances in Spider Taxonomy
1981-1987. Manchester Univ. Press, Manchester,
England. 673 pp.

Redborg, K. E. & E. G. MacLeod. 1985. The devel-
opmental ecology of Mantispa uhleri Banks (Neu-
roptera: Mantispidae). Illinois Biol. Monogr. 53. 130

pp.

Richardson, M. W. 1976. Micropredators of spiders.
Ph.D. dissertation. Southern Illinois Univ., Car-
bondale, Illinois. 211 pp.

Roble, S. M. 1986. A new spider host association for
Mantispa viridis (Neuroptera, Mantispidae). J. Ar-
achnoL, 14:135-136.

Roth, V. D. 1985. Spider Genera of North America.
American ArachnoL Soc. 176 pp.

Stein, R. J. 1955. An insect masquerader. Nat. Hist.,
11:472^73.

Tolbert, W. W. 1976. Population dynamics of the
orb weaving spiders Argiope trifasciata and Argiope
aiirantia (Araneae, Araneidae): density changes as-
sociated with mortality, natality and migrations.
Ph.D. dissertation, Univ. of Tennessee, Knoxville,
Tennessee. 186 pp.

Valerio, C. E. 1971. Parasitismo en huevos de arana
Achaearanea tepidariorum (Koch) (Aranea: Theri-
diidae) en Costa Rica. Rev. Biol. Trop., 18:99-106.

Manuscript received 27 May 1992, revised 31 August
1992.

1992. The Journal of Arachnology 20:157-164

LIFE CYCLE AND HABITAT PREFERENCE
OF THE FACULTATIVELY ARBOREAL WOLF SPIDER,
GLADICOSA PULCHRA (ARANEAE, LYCOSIDAE)

Micky D. Eubanks* and Gary L. Miller: Department of Biology, University of
Mississippi, University, Mississippi 38677 USA

ABSTRACT. The life history and habitat preference of the wolf spider Gladicosa pulchm were investigated
in several populations in Mississippi. Gladicosa pulchra has a one year life cycle with spiders changing from
forest floor to tree trunk habitats in late summer or early fall. During the fall, spiders were found almost exclusively
on trees (93% of observed spiders in 1989 and 80% of observed spiders in 1990). Males were observed to inhabit
trees earlier in the year than females. Spiders did not climb trees smaller than 2 cm in diameter at breast height.
Most individuals were collected at heights less than 2.5 m, and spiders were primarily oriented face down while
on trees. The role that environmental factors play in this animal’s habitat preference is discussed.

Although the wolf spiders (Lycosidae) are pri-
marily ground dwellers, some species are occa-
sionally found on low vegetation (e.g., Lycosa
caro/mewsw Walckenaer, L. timuqua Wallace, and
L. rabida Walckenaer, Kuenzler 1958; Bames
1953, L. hentzi Banks, Miller et al. 1988), small
low tree branches (e.g., L. rabida, Kuenzler 1958),
and, less often, the trunks of trees (e.g., L. ha-
waiiensis Simon, Gon 1985). One species for
which the use of arboreal habitats appears to be
a more significant aspect of its life history is
Gladicosa pulchra (Keyserling). Brady (1986) re-
ported that mature males and females of this
species were commonly collected from the tranks
of trees but rarely from the forest floor. Miller
and Miller (pers. obs.) have found the extensive
use of tree trunk habitats during certain times of
the year predominately by penultimate and ma-
ture individuals in both pine and hardwood hab-
itats in Florida and Mississippi. These obser-
vations, particularly the dearth of sightings of
individuals in habitats other than trees, imply
that G. pulchra spiders may undergo an abrupt
habitat change, moving from forest floor to trees.
However, no detailed ecological study of this
species has been undertaken, and the phenology
and significance of the use of arboreal habitats
are unclear. In particular, it is not known whether
tree trank habitats serve primarily as sites for

‘Current address: Department of Entomology, 1300
Symons Hall, University of Maryland, College Park,
Maryland 20742-5575.

foraging, reproduction, or as a refuge from pred-
ators.

The purpose of this study was to delineate the
life cycle of G. pulchra for the purpose of un-
derstanding the importance of the use of the ar-
boreal habitat in this species. In particular, we
describe (1) the life history, (2) the timing of tree
climbing behavior, (3) tree size preference, (4)
height of spiders on trees, (5) the orientation of
spiders on trees, and (6) sex differences in ar-
boreal habitat use.

METHODS

Habitat Preferences and Life History. — We
conducted the study in a 3 ha mixed-hardwood
woodlot located on the University of Mississippi
campus, near Oxford (Lafayette County), Mis-
sissippi. The study site was composed predom-
inantly of a canopy and understory of Quercus
spp. with abundant leaf litter (> 15 cm depth in
some locations). We made weekly nighttime sur-
veys of the area between 30 March-19 Novem-
ber 1989, and between 1 February- 15 October
1990. Periodic observations were made during
February, June, and July of 1991, but no mea-
surements were made. For every individual G.
pulchra observed on a tree we determined: (1)
the diameter [cm] at breast height [DBH] of the
tree, (2) the height [cm] of the spider on the tree,
(3) the vertical orientation of the spider mea-
sured as degrees to the nearest 1 0° with face down
orientation designated as 0°, and (4) the number
of additional G. pulchra spiders on the tree. All
spiders that were collected in 1989 and 1990

157

158

THE JOURNAL OF ARACHNOLOGY

were taken to the laboratory where immatures
were reared to adulthood to assess sex and de-
velopmental stage at capture; however, not all
spiders that were observed were collected.

To determine the size distribution of trees in
the woodlot we measured the DBH of every tree
within 2 m of three north-south transects (each
200 m in length). The transect samples repre-
sented approximately 4% of the area of the wood-
lot. Weekly average temperatures and rainfall ac-
cumulations were obtained from the USDA
National Sedimentation Laboratory weather sta-
tion located approximately five km from the field
site in Oxford, Mississippi.

In addition to the detailed observations made
at the primary study site, we made periodic ob-
servations (approximately once each month) at
four other areas of Lafayette County, Mississip-
pi. We propose a life history of this species based
on all of our observations in all study areas.

Statistical Analyses.— To determine whether
individual G. piilchra select trees of a specific
size, we compared the size (DBH) distribution
of trees climbed by spiders to that of trees sam-
pled along the transects with a Kolmogorov-
Smimov (KS) goodness-of-fit test (test statistic
designated d„3,,; Sokal & Rohlf 1 980). Regression
analysis was employed to determine if there was
a seasonal trend in the DBH of trees selected by
spiders. A two-tailed Durbin-Watson test was
used to determine whether the error terms of the
regression were uncorrelated (Neter et al. 1991).
The two-tailed test is performed by testing sep-
arately for positive and negative autocorrelation
where the Type I error for the combined test is
2a (Neter et al. 1991).

We tested whether the orientation of males and
females was uniformly distributed around a cir-
cle using the Rayleigh test (test statistic denoted
Z; Zar 1984). The mean direction, the 95% con-
fidence interval of the mean, the Z statistic, and,
r, a unitless measure of magnitude or concentra-
tion that ranges from 0 (indicating so much dis-
persion that a mean direction cannot be de-
scribed) to 1 .0 (where all the data are concentrated
at the same direction) are presented (Zar 1984).

Contingency table analysis (G-statistic, Sokal
& Rohlf 1980) was used to compare the fre-
quency with which adult male and female, pen-
ultimate male and female, and immatures were
present on trees during each year. We employed
logistic regression analysis (Hosmer & Leme-
show 1989) to determine if the sex of a spider

ADULT MALES AND FEMALES
Courtship and Copulation

IMMATURE SPIDERS
Mature

(Forest Floor/Trees)

ADULT FEMALES
Overwinter

(Forest Floor/Burrows)

SPIDERLINGS
Emerge From Egg Case
(Forest Floor)

ADULT FEMALES
With Egg Case
(Forest Floor/Burrows)

Figure 1. — Proposed life history of Gladicosa pul-
chra. Adults are present on trees in the fall of the year
where courtship and copulation is thought to occur.
Females overwinter gravid and produce egg cases in
the spring while inhabiting burrows. Females move
about on the forest floor with spiderlings on their ab-
domen in late spring. Spiders then climb trees as they
approach maturity during the late summer and early
fall.

was related to the date collected, height of the
spider at the time of collection, or the DBH of
the tree on which the spider was observed. Lo-
gistic regression analysis is used to test for re-
lationships between a discrete (usually dichoto-
mous) dependent variable and discrete or
continuous independent (predictor) variables. In
our analysis the presence or absence of females
represented the dichotomous dependent variable
and the variables DBH, height of collection site,
and date of collection or observation represented
the predictor variables. A hierarchy of models
including all subsets of the independent variables
was examined. The likelihood ratio (denoted D),
which is analogous to the residual sum of squares
in linear regression, is reported and the test sta-
tistic G was employed to determine what vari-
ables were significant predictors of the response
variable (Hosmer & Lemeshow 1989). Logistic
analysis was calculated on the BMDP statistical
package.

Student’s t test was used to determine if the
average DBH of trees occupied by more than one
spider was larger than the average DBH of trees
occupied by a single spider. To determine if there
were seasonal differences in the number of trees
occupied by more than one spider, we used lo-
gistic regression as described above with the
number of spiders per tree as the dependent vari-
able. Because of the small number of sampling
dates during 1 99 1 , we did not attempt to analyze
seasonal trends for that year.

EUBANKS & mUJEK-GLADICOSA PULCHRA LIFE HISTORY

159

Table 1.— Summary of collections of G. pulchra for
1989 and 1990. The (*) indicates that males and fe-
males were not reared to adulthood to access sex, and
the (**) indicates that immatures were not included in
the totals.

On trees

On forest floor

Fe-

Males males

Males

Fe-

males

1989-(30 Mar.

-19 Nov.) {n

= 91)

Immature

0

4

0

0

Penultimate

18

14

0

1

Adult

6

43

1

4

Total

24

61

1

5

1990-(1 Feb.-

15 Oct.) (n =

153)

Immature

56*

—

18*

—

Penultimate

16

13

5

0

Adult

18

20

3

4

Total

34**

33

8»*

4

RESULTS

Life History. — A proposed life history of G.

pulchra is diagrammed in Fig. 1. During 1989
and 1 990, adult males and females were collected
predominantly on trees in the fall. Courtship and
copulation were not observed during either year,
however, during February of 1991 we collected
two gravid females. These females were found
in vertical burrows approximately 10 cm deep
each with a turret constructed of leaf litter. Dur-
ing this same month, three mature females and
one immature spider were collected from the for-
est floor and a mature female was taken from a
tree at another site located on the University of
Mississippi, Oxford campus. Patricia Miller (pers.
comm.) observed a mature female with an egg
sac occupying a burrow located in her yard 1.2
km north of Oxford, Mississippi. This burrow
had a turret constructed from grass. During both
census ye -is, no adult males were collected dur-
ing the spring.

Habitat Preferences.— Of the 91 spiders col-
lected during 1989 headlight censuses, 6 (7%)
were found on the forest floor (Table 1). Only
one, an adult female, was collected prior to 21
August, the first date that spiders were collected
from trees (Table 1). The first date that spiders
were collected from trees coincided with the first
week of the year during which the average weekly
temperature reached 29 °C or higher in both 1989
and 1990 (21 August 1991 and 2 June 1991,
respectively).

Proportion o< Trees

y-

Proportion ot Trees

0.6 -

jHlM litBmBi H -- - -

0 10 20 30 40 60 60 70 80

Proportion of Trees

DBH (cm)

Figure 2. — Top, Frequency of occurrence of trees in
sixteen, 5 cm size classes (Diameter in cm at Breast
Height) for trees selected along transects; Middle, trees
climbed by spiders in 1989; and Bottom, trees climbed
by spiders in 1990.

The average DBH of trees selected along the
transects was 8 cm (« = 196, SD = 19; Fig. 2),
and that distribution of tree sizes also differed
significantly from the distribution of trees se-

160

THE JOURNAL OF ARACHNOLOGY

Height of Spiders on Trees

1989 & 1990

Number

Height (cm)

Figure 3.— The number of Gladicosa pulchm collected at each height (cm) on trees during 1989 and 1990 at
the study site on The University of Mississippi campus, Oxford, Mississippi.

lected along the transects (KS test, = 0.64,
n, = 90, = 196, P < 0.01).

The two-tailed Durbin-Watson test for auto-
correlation revealed no positive or negative cor-
relation among error terms in the regressions of
DBH against collection date for either 1989 or
1 990 (test for positive correlation, 0,98, = 1 .663,
n = 180, D,99o = 1.814, n = 108, P < O.OI; test
for negative correlation, D,989 = 6.656, n = 180,
D,99o = 2.186, « = 108, P < 0.01). There was no
seasonal change in tree size preference during

1989 or 1990 (Regression analysis, = 0.02,

MSE = 518.89, df= 139, P > 0.05; = 0.02,

MSE = 360.32, df = 89, P > 0.10, 1989 and

1990 respectively).

In 1989, 93% of the 91 spiders collected were
found on trees (Table 1). Adult females were
found significantly more often on trees than any
other sex/stage combination in 1989 (contingen-
cy table analysis, Gadj = 17.79, n = 85, P <
0.001). During the 1990 census, 80% of the 153

spiders collected were found on trees. In 1990
no sex/stage combination was present signifi-
cantly more often than any other (contingency
table analysis, G^dj = 0.37, n = 61, P > 0.5).
Immature spiders were first collected from trees
on 2 June 1990, and 85% of the 13 spiders col-
lected on that date were on trees. Adult spiders
were not collected on trees until late August, as
in 1989.

The height of spiders on trees averaged 145
cm in 1989 and 80 cm in 1990. No seasonal
changes in height were observed during either
year (Regression analysis, R^ = 0.02, MSE =
3979.55, df= 89, P > 0.05; R^ = 0.04, MSE =
5549.43, #= 107, P > 0.05, 1989 and 1990
respectively). During both years most spiders were
at heights below 2.5 m (Fig. 3).

During both years the vertical orientation of
spiders differed significantly from random, with
most individuals adopting a face down position
on the tree (mean direction = 0°, n = 101, 95%

EUBANKS & MILLER- GLAD/COSA PULCHRA LIFE HISTORY

161

1989 1990

Figure 4.— Orientation of Gladicosa pulchra collected from trees in 1989 and 1990. Figures shov/ the mean
direction (P), magnitude (r), and number of obser'/ations {N), 0° = facing down. Each darkened circle represents
two observations.

Cl ±10, r = .82, Z = 67.91, P < 0.001; mean
direction = 0°, « = 57, 95% Cl ±10, r = 0.91,
Z = 47.20, P < 0.001; 1989, 1990 respectively;
Fig. 4). Males, females, and immatures all had
mean directions = 0°.

The proportion of spiders collected or ob-
served during 1989 that were female was 0.73,
significantly different from a 1 : 1 maie/female sex
ratio (x^ P < 0.001). Logistic regression analysis
results showed that date was the best predictor
of the sex of an individual spider (D = - 56.246,
G = 8.81, 0.005 < P > 0.001). Female spiders
were more likely to be found on trees later in the
season than males. In 1990, the overall propor-
tion of females in the population was 0.47 and
did not differ significantly from a 1:1 male/fe-

male sex ratio (x^ P > 0.05). In 1990, date was
the best predictor of the sex of an individual
spider collected from a tree (D = —20.554, G =
9.2, 0.005 < P > 0.001), and as in 1989, females
were more likely to be found later in the season
than males.

During both years few spiders were observed
to occupy trees that contained other G. pulchra
(16 out of 238 observations, 7%). During 1989
trees occupied by a single spider had a signifi-
cantly smaller mean DBH than those with two
or more spiders (t = -4.53, P < 0.001), but not
in 1 990 (t = 0. 1 , F > 0.05). The number of trees
occupied by more than one spider at one time
significantly increased during the season in 1 989
(D = -26.451, G = 12.881, df= 6, P < 0.05),

162

THE JOURNAL OF ARACHNOLOGY

but this was not the case during 1990 (D =
-22.040, G = 10.013, df= 1,P> 0.1).

DISCUSSION

Although we collected immature individuals
during February of 1991, we believe that the
observations reported here support the sugges-
tion of a one-year life cycle for G. pulchra, with
courtship and copulation occurring in the fall
(Fig. 1). Although the location of courtship and
copulation is uncertain, the abundance of pen-
ultimate and mature spiders found on trees in
the fall of 1989 and 1990 suggests the possible
importance of the arboreal habitat in the repro-
duction of this species. Females apparently over-
winter gravid and produce egg sacs in burrows
during the spring. It is unclear whether females
construct burrows or whether they usurp com-
pleted or partially completed burrows of another
species. It is evident, however, females expend
some energy in the construction and mainte-
nance of the turret.

Spiderlings emerge from egg sacs in the spring
and reach sexual maturity during the early fall.
Once the spiderlings emerge, the female aban-
dons the burrow and wanders about the forest
floor carrying the young on her abdomen in the
manner typical of non-burrowing wolf spiders.
Adult males apparently do not overwinter, and
are present predominantly on trees during the
fall.

Our observations showed that in 1989 and 1990
G. pulchra is present primarily on trees during
the late summer and early fall. Spiders were rare-
ly collected from the forest floor or from trees
prior to the fall, but females were found in bur-
rows during the spring. These observations sug-
gest that individuals move from the forest floor
(either from burrows or from the leaf litter) to
trees during the late summer or early fall. Pre-
vious studies of spiders have documented vari-
ations in microhabitat use by different sizes or
different life stages of conspecifics (e.g., Waldorf
1976; Hallander 1970). Hallander (1970) re-
ported that spiders of the genus Pardosa stratified
their use of leaf litter habitats according to body
size in order to limit the effects of cannibalism.
However, the change from the forest floor habitat
to a tree dwelling life style observed in G. pulchra
is substantially more dramatic than most habitat
shifts observed in spiders. Animals may change
habitats to balance the conflicting demands of

minimizing the risk of mortality and maximizing
food intake (e.g., Werner & Gilliam 1984, Wer-
ner et al. 1983, Gilliam & Fraser 1987, Werner
& Hall 1988, Pierce 1988, Gotceitas & Colgan
1990, Gotceitas 1990) or to avoid intraspecific
competition or size-specific predation (Werner
& Gilliam 1 984). The relative importance of these
two factors in the habitat change of G. pulchra
will be reported elsewhere.

In both 1 989 and 1 990 adult females and pen-
ultimate males were occasionally observed to oc-
cupy positions on the same tree. The asynchrony
in the timing of maturation of males and females
that is indicated by these observations is not un-
common in spiders (e.g., Austin 1984; Berry 1987;
Miller & Miller 1987). Such differences in the
timing of maturation may lead to pre-courtship
cohabitation (e.g., Robinson & Robinson 1980;
Christenson & Goist 1979; Pollard & Jackson
1982; Jackson 1977; Miller & Miller 1986). Typ-
ically, if cohabitation occurs, the mature male
seeks a stationary subadult female and waits with
her until she matures. However, our observa-
tions indicate that penultimate males ascend trees
earlier in the year than females, which typically
climb as adults. Thus, if the proximity of males
and females on trees represents pre-courtship co-
habitation in this species, then both the behav-
iors associated with the phenomenon (e.g., one
sex moving to locate another) and the relative
timing of the maturation of the sexes are different
from that observed in other spiders. It is im-
portant to note that the proximity of males and
females of different stages on the same tree may
be the result of other processes (e.g., foraging,
predator avoidance) not related directly to re-
productive behavior. Moreover, the majority of
our observations were of single individuals on a
tree (see below).

Our observations indicate that the onset of tree
climbing in G. pulchra is not exclusively triggered
developmentally. The sex ratio of collected spi-
ders was strongly female biased in 1989, but not
in 1990. Additionally, the beginning of the tree
dwelling phase came later in the year in 1989
than in 1990. In 1990, spiders climbed trees ear-
lier in the year (June vs. August) and earlier in
their life cycle (immature vs. penultimate or
adult). Thus, spiders climbed trees at different
stages in the two years. Although, considerably
more work will be required to determine the rel-
ative influence of physical and biological factors
in this behavior, our study suggests the impor-

EUBANKS & M.IUJEB.-GLADICOSA PULCH114 LIFE HISTORY

163

tance of rainfall as a critical physical parameter.
Total rainfall was greater in the summer of 1989
than in the summer of 1990. Humidity and soil
moisture should be directly influenced by rain-
fall, and these two factors have been shown to
play an important role in the microhabitat se-
lection of spiders (Cady 1984; Reichert & Tracy
1975).

Individual G. pulchra were never collected on
trees smaller than 2 cm DBH during this study.
Larger trees may provide larger foraging areas or
refuge from forest floor predators. We have no
information about the relative availability of prey
on trees and the forest floor, but spiders were
often observed feeding on ants and moths on
trees. However, a potential predator of G. pul-
chra, the large wolf spider Lycosa georgicola
Walckenaer, is common on the forest floor. We
have seen individuals of that species climb small
saplings but we made only a single observation
of L. georgicola on a tree larger than 2 cm DBH.

The average height of spiders on trees did not
change as the season progressed in 1 989 or 1 990.
Spiders were predominantly found at heights be-
tween 1.5-2. 5 m, and individuals of all stages
(immatures, penultimates, and adults) typically
adopted a face down orientation while on the
tree. In the absence of a compelling physiological
explanation for the face-down behavior, we sug-
gest that such an orientation is the most practical
for intercepting prey and/or mates that originate
from below. The relatively constant height
through the season supports this notion.

Although we observed trees that held more
than one spider, most of the spiders seen during
this study were the lone occupant of the tree from
which they were collected. Many field studies
have established that density-dependent effects,
such as competition for web sites, occur in web-
building spiders (Schaefer 1978; Reichert 1979;
Wise 1981). Web-building spiders tend to stratify
web construction in a manner that reduces com-
petition. Our observations may indicate an anal-
ogous form of habitat stratification or territori-
ality.

ACKNOWLEDGMENTS

We wish to thank Patricia R. Miller, Gail
Stratton, Timothy G. Forrest, Joel Trexler, Kari
Benson, and Chester Figiel for their constructive
comments on various versions of this paper. Noel
Hunt assisted with the field work. This research
was supported by a grant from the Exline-Frizzell

Fund for Arachnological Research, California
Academy of Sciences.

LITERATURE CITED

Austin, A. D. 1984. Life history of Clubiona robusta
L. Koch and related species (Araneae, Clubionidae)
in South Australia. J. ArachnoL, 12:87-104.

Barnes, R. D. 1953. The ecological distribution of
spiders in non-forest maritime communities at
Beaufort, North Carolina. Ecol. Mon., 23:316-337.

Berry, J. W. 1987. Notes on the life history and be-
havior of the communal spider Cyrtophora mol-
uccensis (Doleschall) (Araneae, Araneidae) in Yap,
Caroline Islands. J. Arachnol., 15:309-319.

Brady, A. R. 1986. Nearctic species of the new wolf
spider genus Gladicosa (Araneae: Lycosidae). Psy-
che, 93:285-319.

Cady, A. B. 1984. Microhabitat selection and loco-
motor activity of Schizocosa ocreata (Walckenaer)
(Araneae, Lycosidae). J. Arachnol., 1 1:297-307.

Christenson, T. E. & K. C. Goist. 1979. Costs and
benefits of male-male competition in the orb weav-
ing spider, Nephila clavipes. Behav. Ecol. Sociobiol.,
5:87-92.

Gilliam, J. F. & D. F. Fraser. 1987. Habitat selection
under predation hazard: test of a model with for-
aging minnows. Ecology, 68:1856-1862.

Gon, S. M. 1985. A Hawaiian wolf spider, Lycosa
hawaiiensis Simon foraging in the top of a Metros-
ideros polymorpha tree. J. Arachnol., 13:393.

Gotceitas, V. 1990. Foraging and predator avoid-
ance: a test of a patch choice model with juvenile
bluegill sunfish. Oecologia (Berlin), 83:346-351.

Gotceitas, V. & P. Colgan. 1990. The effects of prey
availability and predation risk on habitat selection
by juvenile bluegill sunfish. Copeia, 1990:409-417.

Hallander, H. 1970. Prey, cannibalism and micro-
habitat selection in the wolf spiders Pardosa chelata
O.F. Muller and P. pullata Clerck. Oikos, 21:337-
340.

Hosmer, D. W. & S. Lemeshow. 1989. Applied Lo-
gistic Regression. John Wiley and Sons, New York,
New York.

Jackson, R. R. 1977. Courtship versatility in the
jumping spider, Phidippus johnsoni (Araneae: Sal-
ticidae). Anim. Behav., 25:953-957.

Kuenzler, E. J. 1958. Niche relations of three species
of lycosid spiders. Ecology, 39:494-500.

Miller, G. L. & P. R. Miller. 1986. Pre-courtship
cohabitation of mature male and penultimate fe-
male Geolycosa turricola (Araneae, Lycosidae). J.
Arachnol., 14:133-134.

Miller, G. L., P. R. Miller & A. R. Brady. 1988. Ad-
hesive hairs in lycosid spiders of various life styles,
including the occurrence of claw tufts in Lycosa hen-
tzi Banks. Bull. British. Arachnol. Soc., 7:213-216.

Neter, J., W. Wasserman, M. H. Kutner. 1991. Ap-

164

THE JOURNAL OF ARACHNOLOGY

plied linear statistical models, 3rd edition. Irwin,
Homewood, Illinois.

Pierce, C. L. 1988. Predator avoidance, microhabitat
shift, and risk-sensitive foraging in larval dragon-
flies. Oecologia (Berlin), 77:81-90.

Pollard, S. D. & R. R. Jackson. 1982. The biology
of Clubiona cambridgei (Araneae, Clubionidae): In-
traspecific interactions. New Zealand J. Ecol., 5:44-
50.

Reichert. S. E. 1979. Games spiders play. III. Re-
source assessment strategies. Behav. Ecol. Socio-
biol., 6:121-128.

Reichert, S. E. & C. R. Tracy. 1975. Thermal balance
and prey availability: Basis for a model relating web-
site characteristics to spider reproductive success.
Ecology, 56:265-284.

Robinson, M. H. & B. Robinson. 1980. Comparative
studies of the courtship and mating behavior of
tropical araneid spiders. Pacific Insects Monogr. No.
36. Bishop Museum, Honolulu.

Schaefer, M. 1978. Some experiments on the regu-
lation of population density in the spider Floronia
biicculenta (Araneida: Linyphiidae). Symp. Zool. Soc.
London, 42:203-210.

Sokal, R. R. & F. J. Rohlf 1980. Biometry. W. H.
Freeman and Company, New York.

Waldorf, E. S. 1976. Spider size, microhabitat selec-
tion, and use of food. American Mid. Nat., 96:76-
87.

Werner, E. E. & D. J. Hall. 1988. Ontogenetic habitat
shifts in bluegill: the foraging rate-predation risk
trade-off. Ecology, 69:1352-1366.

Werner, E. E. & J. F. Gilliam. 1984. The ontogenetic
niche and species interactions in size-structured
populations. Ann. Rev. Ecol. Syst., 15:393-425.

Werner, E. E., J. F. Gilliam, D. J. Hall & G. G. Mit-
telbach. 1983. An experimental test of the effects
of predation risk on habitat use in fish. Ecology, 64:
1540-1548.

Wise, D. H. 1981. Inter- and intraspecific effects of
density manipulations upon females of two orb-
weaving spiders (Araneae: Araneidae). Oecologia
(Berlin), 48:252-256.

Zar, J. H. 1 984. Biostatistical analysis, 2nd edition.
Prentice-Hall, Englewood Cliffs, New Jersey.

Manuscript received 22 September 1 991, revised 20 June
1992.

1992. The Journal of Arachnology 20:165-172

ABUNDANCE AND ASSOCIATION OF CURSORIAL SPIDERS
FROM CALCAREOUS FENS IN SOUTHERN MISSOURI

Thomas L. Bultman: Division of Science, Northeast Missouri State University,

Kirksville, Missouri 63501 USA

ABSTRACT. I systematically sampled, by pitfall trap, spiders in three different but nearly contiguous fens
(Seep, Prairie and Forested) as well as those in the surrounding habitat (Oak-hickory forest) in southern Missouri.
Samples were taken biweekly from May through August. Fifty-five species of cursorial spiders were found, of
these, 17 occurred only within one or more of the fens and not in the surrounding habitat. All habitats, and
particularly fens, were dominated by wolf spiders. Jaccard’s similarity coefficient showed spider faunas of the
Seep and Prairie fens were more similar to one another than to that of the Forested fen. While both Seep and
Forested fen habitats harbored spider species quite different from the surrounding Oak-hickory forest, many
species from the Forested fen were also found in the Oak-hickory habitat (56.3%). Abundances of spiders were
greatest in May and declined through summer in all habitats, except the Seep fen in which abundances remained
fairly constant. Tests of interspecific association among species within Lycosidae, Clubionidae, and Gnaphosidae
gave no evidence of negative associations of species, which could theoretically result from interspecific com-
petition, among the three fen habitats. I conclude that the spider faunas associated with the fens are quite distinct
from that of the surrounding habitat with each fen harboring a somewhat unique assemblage of spiders and that

there is no evidence of competitive exclusion of species

The Ozark Mountains are a biologically rich
area of the central United States with many en-
demic plant (Thom & Wilson 1980; Thom &
Iffrig 1985) and animal species. Of the com-
munities found there, fens are perhaps the most
distinctive. Fens are boggy areas with saturated
soils caused by seepage. Within the Ozark Moun-
tains, fens are most common in the Salem Pla-
teau region (Thom & Wilson 1980). Several un-
usual geological factors contribute to the
abundance of fens in the Salem Plateau region.
First, the region is characterized by Cambrio-
Ordovician dolomite bedrock covered by a thick
residuum formed from in situ weathering of the
parent material. The residuum freely accepts and
transmits water (Orzell 1983). Second, there are
several regional springs that are recharged in the
area. As water moves through the residuum from
hillsides and uplands the soil becomes saturated.
In lowland areas within the Salem Plateau region
fens develop under the influence of a constant
supply of cool water. The source of the water is
thought to be springs recharged by storage
groundwater (Aley 1978). Fed by cool-water
springs, fens of Missouri support a flora more
typical of latitudes much farther north.

Recent work shows that the flora of Missouri
fens is markedly different from that of the sur-

from the fen habitats.

rounding vegetation (Orzell 1983). Yet, the ar-
thropod fauna associated with fens is poorly
known. Systematic sampling of fen spiders has
not been done. The goal of my study was to
describe the spiders associated with the various
fen types in the Grasshopper Hollow Fen com-
plex in southern Missouri. Additional objectives
were to assess the seasonal abundance patterns
of spiders, to assess if species exclude one another
from some fens and to evaluate how unique fen
spiders are, compared to those from the habitat
surrounding the fens. A goal of my work is to
provide knowledge about the biotic distinctness
of the fens so that appropriate management prac-
tices of these protected and unique habitats can
be initiated.

METHODS

The study was conducted in the largest (7 hec-
tares) and most diverse fen complex in Missouri:
Grasshopper Hollow Fen in Reynolds County.
The complex is heterogeneous, being composed
of several fens that differ somewhat vegetatively
(Orzell 1983). One of the fens in the complex,
the Seep fen, is heavily dominated by various
species of sedges [Carex interior Bailey, C. lep-
talea Wahl, and C. subserecta (Olney) Britt.] and
rush (Juncus dudleyi Wieg.). Another fen type,

165

166

THE JOURNAL OF ARACHNOLOGY

the Prairie fen, contains an interesting mix of fen
(cord grass, Spartina pectinata Link.) and prairie
plants (big bluestem, Andwpogon gerardii Vitm.;
Indian grass, Sorghastrum nutans (L.) Nash;
prairie dock; Silphium terebinthinaceum Jacq.;
and swamp orange coneflower, Rudbeckia ful-
gida Ait.). A third fen type, the Forested fen,
contains the following calcifile tree species: Quer-
cus mulenbergii Engelm., Q. rubra L., Fraxinus
americana L., Alnus serrulata (Ait.) Willd., and
Carpinus caroliniana Walt. The habitat sur-
rounding the fen complex is mesic oak-hickory
forest typical of southern Missouri (Nelson 1985).

I assessed vegetative structure at the study sites
monthly. Foliage height diversity (FHD) was de-
termined by the point-quadrat method using a
calibrated (four 0.4 m intervals) pole that was
diagonally placed at 1 m intervals along a ran-
domly selected 8 m transect within each habitat.
The number of stems contacting the pole at each
height interval was recorded. Data generated from
samples were input into the Shannon (1948) di-
versity index to give the foliage height diversity,
as has been done by others (e. g., Willson 1974).
The technique provides a measure of the struc-
tural diversity of the habitat, a feature than may
strongly influence arthopods (Murdock et al.
1972).

Ten pitfall traps (13 cm diameter; 9 cm deep)
were placed in 1989 within each fen and the
surrounding habitat to sample cursorial spiders.
Traps were placed at 5 m intervals along a ran-
domly-placed 50 m transect, with the exception
of the Forested fen which was less than 50 m in
diameter. Here, traps were placed randomly
within the habitat. Traps were placed in fens
during the first and third week of each sampling
month; sampling occurred on two weeks each
month to reduce environmental pertebation to
the protected study area. A mixture of ethylene
glycol and water (1:1) was used as a preservative
in traps. Spiders were sampled in May, June, July
and August. Wooden covers raised above traps
by legs allowed arthropod entry but discouraged
mammals from entering and falling into traps.
Pitfall traps do not give a true measure of density,
rather they sample the number of cursorial spi-
ders moving in an area for a given time or the
“active density” (Uetz 1977). Yet, they are useful
in that they allow continuous sampling in varied
habitats and provide an adequate estimate of the
number of species of cursorial spiders over a
wide range of habitats (Uetz & Unziker 1976).

Several pitfall traps were disturbed during the
study and could not be included in the analysis.
Therefore, for some months, some habitats had
fewer than 10 pitfall trap samples. Numbers of
individuals from these pitfall traps were stan-
dardized to make them comparable to habitats
that did not have disturbed traps. For example,
two pitfall traps in May from the Prairie fen were
disturbed. Numbers of spiders collected in the
remaining eight traps were multiplied by 10/8.
In June, all 10 traps in the Prairie fen were dis-
turbed, apparently by a small vertebrate.

Fen and non-fen habitats were compared at
the species level using Jaccard’s (1908) coeffi-
cient,

Sj = a/a + b + c

where a is the number of species collected in both
habitats A and B, b is the number of species
collected in habitat B but not in habitat A, and
C is the number of species collected in habitat
A but not in habitat B.

Species identification was made by the author,
except for several species of Lycosidae and Gna-
phosidae, which were identified or their identity
verified by systematists. Only adult specimens
(with the exception of a few individuals whose
color pattern and general morphology allowed
unquestionable matching with mature individ-
uals) are included in the data set since immatures
could rarely be reliably identified to species.

I calculated the variance ratio (VR) to test si-
multaneously for significant associations among
species present at the three fen habitats for mem-
bers of the Lycosidae, Clubionidae, and Gna-
phosidae. The variance ratio was calculated as

VR = Syar

where S? is the variance in total species number
and (T? is the total sample variance for occur-
rences of the S species in the samples from the
three fens (see Schluter 1984). If VR > 1, then
the species exhibit a positive association and if
VR < 1 , a negative association exists among the
species at the fens. The statistic, W, was com-
puted to test whether deviations of VR from 1.0
were significant. The statistic, W, equals N(VR),
where N is the number of habitats; W has a 95%
probability of falling between limits given by the
chi-square distribution (see Schluter 1984)

X^.025,N ^ ^ X .975,N-

BULTMAN-CURSORIAL SPIDERS FROM MISSOURI FENS

167

RESULTS AND DISCUSSION

A total of 55 species of cursorial spiders was
collected from the fen and non-fen habitats (Ta-
ble 1). These spiders represented eight families.
Of these species, 17 were found only in one or
more of the fens and not in the Oak-hickory
forest. Only one species was restricted to just the
Oak-hickory forest. Both Seep and Prairie fens
contained fairly diverse wolf (8 and 1 1 species,
respectively) and running (7 and 1 2 species, re-
spectively) spider groups. Overall, the Prairie fen
was the most speciose of the habitats sampled,
while the Forested fen had the fewest number of
species (Table 1).

Comparison of species similarity among the
various habitats reveals several interesting fea-
tures (Table 2). The coefficient of similarity be-
tween the spider faunas at the Seep and Prairie
fens is quite high (Sj = 0.457 1 ; Table 2), but both
of these faunas are quite dissimilar to that found
at the Forested fen. The Seep fen spider fauna is
also not similar to that found in the Oak-hickory
forest. The spider community associated with the
surrounding habitat is truly much different than
that found in the Seep fen. In contrast, the spider
species associated with the Forested fen and the
Oak-hickory forest are fairly similar (Sj = 0.2727;
Table 2). Species associated with the Prairie fen
show some similarities with those from the Oak-
hickory forest (Sj = 0.2500; Table 2), but less so
than species from the Forested fen. Overall, eight
of the 55 species collected in the study were found
only in pitfall traps placed in the Prairie fen.
Moreover, since these data represent only three-
trap months (due to disturbance in June), it is
likely that one or more species endemic to the
Prairie fen may have been missed. In sum, while
the Prairie fen and particularly the Seep fen tend
to harbor spider species unlike the surrounding
non-fen habitat, the Forested fen contains spe-
cies also found in the surrounding Oak-hickory
forest.

Looking at individual spider species reveals
similarities and differences among habitats at the
species level (Table 1). Pirata insularis Emerton
was found at all sites, but was particularly com-
mon in the three fen habitats. Seep and Prairie
fens both shared a numerically abundant species,
Pardosa saxatilis (Hentz) (Table 1). One species,
Pirata insularis, a small wolf spider that is often
found near water (Kaston 1981), comprised
64.4% of all individuals in the Seep fen.

The Forested fen and the Oak-hickory forest
shared nine species in common. Particularly
abundant species that they both contained were:
Schizocosa ocreata (Hentz), Drassylus novus
(Banks), Gnaphosa sericata (L. Koch) and Xys-
ticus fraternus Banks. The overlap in species is
not unexpected, given that both habitats are for-
ests and therefore share many abiotic features.
While the Forested fen was most similar to the
Oak-hickory forest, it did share some species with
the other fens. Most notable are Pirata insularis,
a species common in all three fens, but rare in
the non-fen habitat; P. arenicola Emerton, which
was found only in the Forested and Prairie fens;
and, Drassylus creolus Chamberlin & Gertsch,
also a member of Prairie and Forested fens. Of
interest is Phrurotimpus borealis (Emerton), a
species common to Oak-hickory and Forested
fens, but more common in the latter. The range
of this species is reported to be east of the Mis-
sissippi River and north of Georgia (Kaston
1978). Its occurrence in southern Missouri is
somewhat unexpected, since it is frequently found
in more northern deciduous forests. Its presence
in and near the Forested fen supports the hy-
pothesis that the fens are refugia for species nor-
mally found at more northern latitudes.

In terms of community structure, the most dis-
tinctive spider assemblage occurred at the Seep
fen. This community was characterized by a
rather sparse spider fauna (20 species; Table 1)
and an extreme numerical dominant in Pirata
insularis (64.4% of all individuals). The structure
of the spider community at the Seep fen parallels
the relatively simple structural diversity (Fig. 1 )
of the fen’s sedge-dominated vegetation.

At the family level, lycosids dominated all
habitats, particularly fens (Fig. 2). Of all habitats,
the Seep fen was most heavily dominated by
lycosids, with 95.7% of all individuals belonging
to the Lycosidae (Fig. 2). In contrast, the cur-
sorial spider faunas associated with the Prairie
fen. Forested fen and Oak-hickory forest were
composed of 83.3%, 64.8%, and 58.4% lycosids,
respectively (Fig. 2). A preponderance of wolf
spiders in the samples probably reflects the pitfall
trap sampling technique employed. As a measure
of “active density”, pitfall traps would be ex-
pected to differentially capture highly active spi-
ders like lycosids. However, visual inspection
and hand sifting through litter also revealed that
lycosids were by far the most common spiders
present in the fen study sites (pers. obs.). While

168

THE JOURNAL OF ARACHNOLOGY

Table 1.— Species list and seasonal abundances for cursorial spiders from pitfall trap samples taken in 1989.
Abundances corrected for missing pitfall traps (see text).

Habitat

Family

Seep

Prairie

Forested

Oak-hickory

Agelenidae

Agelenopsis pennsylvanica (C. L. Koch)

0

2

3

0

Cicurina robusta Simon

2

0

0

0

Hahniidae

Neoant istea magna (Keyserling)

1

9

0

2

Lycosidae

Allocosa funerea (Hentz)

3

1

0

1

Arctosa virgo (Chamberlin)

0

0

0

29

Lycosa helluo Walckenaer

21

3

0

0

Lycosa rabida Walckenaer

8

24

0

0

Lycosa sp. A

0

0

8

17

Pardosa saxatilis (Hentz)

135

193

0

0

Pardosa moesta Banks

0

1

0

0

Pirata alachuus Gertsch & Wallace

18

5

14

15

Pimta insularis Emerton

391

28

87

2

Pirata arenicola Emerton

0

2

17

0

Schizocosa bilineata (Emerton)

2

29

0

1

Schizocosa crassipes (Walckenaer)

0

0

0

98

Schizocosa ocreata (Hentz)

3

11

77

51

Schizocosa saltatrix (Hentz)

0

8

0

2

Trabea aurantiaca (Emerton)

0

0

1

0

Gnaphosidae

Callilepis imbecilla (Keyserling)

0

4

0

1

Drassylus aprilinus (Banks)

0

1

0

0

Drassylus covensis (Banks)

0

0

0

1

Drassylus creolus Chamberlin & Gertsch

1

5

3

0

Drassylus diximis Chamberlin

2

1

0

0

Drassylus eremitus Chamberlin

2

1

0

1

Drassylus niger (Banks)

0

5

0

0

Drassylus novus (Banks)

0

0

18

17

Gnaphosa fontinalis Keyserling

0

0

0

4

Gnaphosa sericata (L. Koch)

0

0

29

41

Haplodrassus bicornis (Emerton)

0

0

0

3

Haplodrassus signifer (C. L. Koch)

0

4

0

0

Herpyllus vasifer (Walckenaer)

0

0

0

1

Rachodrassus exlineae PI. & Sh.

0

0

0

2

Zelotes duplex Chamberlin

0

1

0

3

Z. hentzi Barrows

0

0

0

1

Clubionidae

Castianeira descripta (Hentz)

1

1

0

0

Castianeira longipalpis (Hentz)

0

9

0

22

Castianeira variata Gertsch

6

0

0

0

Castianeira sp. A

4

0

0

0

Clubionoides excepta (C. L. Koch)

2

4

2

0

Micaria elizabethae Gertsch

0

2

0

0

Phrurotimpus alarms (Hentz)

0

0

19

7

Phrurotimpus borealis (Emerton)

0

0

11

4

BULTMAN-CURSORIAL SPIDERS FROM MISSOURI FENS

169

Table 1.— Continoed.

Habitat

Family

Seep

Prairie

Forested

Oak-hickory

Thomisidae

Misumenops sp. A

2

3

0

0

Oxyptiia sp. A

0

I

0

0

Philodromus sp. A

0

1

0

0

Tmarus sp. A

0

1

0

0

Xyticus emertoni Keyserling

0

0

0

12

Xyiicus ferox (Hentz)

0

0

0

1

Xyticus fraternm Banks

0

0

19

27

Xyticus funestus Keyserling

0

2

0

!

Salticidae

Onondaga lineata (C. L. Koch)

I

0

0

0

Paraphidippus sp. A

0

0

0

3

Sitticus sp. A

2

4

0

0

Salticid sp. A

0

0

4

0

Dictynidae

Dictyna sp. A.

0

0

3

0

Total number of individuals

606

366

370

315

the Seep fen was heavily dominated by members
of the Lycosidae, the Prairie and Forested fens
contained significant numbers of non-lycosid
spiders. Members of the Gnaphosidae and Club-
ionidae comprised substantial proportions of the
total number of individuals at both the Prairie
(6.0% and 4.4%, respectively) and the Forested
fens (10.2% and 6.0%, respectively; Fig. 2). The
Oak-hickory forest differed from the fens, par-
ticularly Seep and Prairie, in having large pro-
portions of Gnaphosidae (20.3%) and Thomisi-
dae (1 1.1%). In sum, family level comparisions
somewhat mirror those at the species level; both
the Seep and Prairie fen spider faunas were more
similar to one another than they were to the For-
ested fen, while spiders from the latter was more
similar to those from the Oak-hickory forest.

Abundances of cursorial spiders were greatest
in May and declined through the summer in the
Prairie and Forested fen and the surrounding Oak-
hickory forest. In contrast, spiders from the Seep
fen occurred in fairly constant numbers over the
sampling season (Fig. 3). Differences among fens
in seasonal trends in abundance are primarily
due to the occurrence of P. insularis. Mature
males and females were found in large numbers
throughout the four month sampling period. In

contrast, other species showed early summer
peaks in abundance of mature individuals. Since
few immatures were included in the analysis,
species at the Prairie and Forested fens showed
a decline in abundance over the summer. If im-
matures could have been included in the anal-
ysis, then spider abundances for these fens would
probably not have declined in this way. Thus,
differences in seasonal abundance patterns among
fens (Fig. 3) reflect abundances of adults, not
immatures and adults. Nonetheless, it is of in-
terest that P. insularis exhibited a phenology fair-
ly atypical of the species in this study; mature
adults were abundant from May through August.

Table 2.— Jaccard’s coefficient of similarity for com-
parisons of the presence/absence of species at the fen
and non-fen habitats.

Seep Prairie

Forested

Oak-

hickory

Seep

- 0.4571

0.1613

0.1667

Prairie

—

0.1892

0.2500

Forested

Oak-hickory

—

0.2727

170

THE JOURNAL OF ARACHNOLOGY

• — '9 Seep
O — O Prairie
V — V Forested

May June July August

Figure 1. — Foliage height diversity (FHD) for fen and non-fen habitats during the four sample months.

100 -i

8

c

(Q

■o

Xi

<

o

>

’■J3

W

Q)

tr

Seep

Prairie Oak-Hickory Forested

^ Lycosidae
^ Gnaphosidae
^ Clubionidae
10 Thomisidae
□ Other

Site

Figure 2. — Family composition of spiders sampled by pitfall trap at fens and surrounding non-fen habitat.

BULTMAN- CURSORIAL SPIDERS FROM MISSOURI FENS

171

o

z

5

35

30

25

20

15

10

5

0

Seep 0' — '0

Prairie O O

Forested V— -V
Oak-Hickory □

\

MAY JUNE JULY AUGUST

Figure 3.— Abundance of cursorial spiders sampled by pitfall trap at fens and surrounding non-fen habitat.
Bars equal 1 SEM.

Females of F. insularis were collected with egg
cases in June, July, and August. It is not clear if
these represent multiple generations or one gen-
eration that breeds over a long period of time.
That penultimate females were only found dur-
ing the month of May argues in favor of the latter
possibility.

Tests for interspecific association (Schluter
1984) within species of the Lycosidae, Clubi-
onidae, and Gnaphosidae among the three fen
habitats showed no negative associations be-
tween species (Table 3). These tests evaluate
whether species of the relatively diverse wolf and
running spider families show nonrandom pat-
terns of presence/absence across the three fens.
One might expect that members would preempt
resource space within a fen and thereby prohibit
the presence of another species of that family.
That is, each fen might not have a full comple-
ment of species due to competitive exclusion from
some fens by some species. The test for inter-
specific association however, gives no evidence
for this conclusion. There is no interspecific com-
petition acting to effect the observed pattern of
presence/absence by wolf and running spiders in
the fens.

CONCLUSIONS

My results show that spiders from the Prairie
fen and particularly the Seep fen differ markedly
from those from the surrounding Oak-hickory
forest. In contrast, the spider fauna from the For-
ested fen is more similar to that at the Oak-
hickory forest. Abundances of adult spiders de-
clined over the summer in all habitats, except
the Seep fen. Tests for interspecific association
among wolf and running spider species at the fen
habitats gave no evidence for competitive dis-

Table 3. — Results of Variance Ratio analysis as a
test of interspecific species association among wolf and
running spiders (Clubionidae and Gnaphosidae) among
the three fen types. Schluter’s (1984) Variance Ratio
is VR. W is the test statistic associated with the vari-
ance ratio test. * = Number of spiders present. ** =
Spider species show a significant positive association
among the 3 fen habitats.

Guild

S*

VR

W

P

Wolf spiders

13

1.9

5.7

ns

Running spiders

20

1.4

4.2

<0.05**

172

THE JOURNAL OF ARACHNOLOGY

placment. Furthermore, spiders collected from
fens showed differences at the species and family
levels among fen types. That cursorial spiders
differ from fen to fen is noteworthy. Spiders, as
generalist, mobile predators, might not be ex-
pected to respond to the somewhat subtle flo-
ristic differences between fen types. The fact that
they do underscores the biotic differences be-
tween these habitats. Spiders, as common sec-
ondary consumers, are extremely important
predators in natural ecosystems (Riechert 1974)
and as such are excellent biological indicators of
community- and ecosystem-level organization.
Differences in spider faunas associated with hab-
itats should indicate fundamental biological dif-
ferences between those habitats. Implications of
my findings are that management of the Grass-
hopper Hollow complex should incorporate the
autonomy that exists between fen types. The fen
complex should not be treated as a single unit
with one management plan.

ACKNOWLEDGMENTS

P. Miller, N. Platnick and J. Redner kindly
made taxonomic identification of difficult spec-
imens. This work was supported through funding
from the Missouri Department of Conserva-
tion’s Small Grants Program.

LITERATURE CITED

Aley, T. J. 1978. Ozark hydrology: a perspective
model. Missouri Speleology, Vol. 18, 185 pp.
Jaccard, P. 1908. Nouvelles recherches sur la distri-
bution florale. Bull. Soc. Sci. Nat., 44:223-270.
Kaston, B.J. 1978. How to know the spiders. Wm C.
Brown Company, Publ., Dubuque, Iowa.

Kaston, B.J. 1981. Spiders of Connecticut. Connect-
icut State Geol. and Nat. Hist. Surv. Bull. 2nd Ed.,
70:1-1020.

Murdoch, W. W., F. C. Evans, & C.H. Peterson. 1 972.
Diversity and patterns in plants and insects. Ecol-
ogy, 53:819-829.

Nelson, P. W. 1985. The terrestrial natural com-
munities of Missouri. Missouri Nat. Areas Com-
mittee. Jefferson City, Missouri.

Orzell, S. L. 1983. Natural area inventory and flo-
ristic analysis of fens in selected southeastern Mis-
souri counties. MS thesis. Southern Illinois Univ.,
Carbondale, Illinois.

Riechert, S. E. 1974. Thoughts on the ecological sig-
nificance of spiders. BioScience, 24:352-356.

Shannon, C.E. 1948. A mathematical theory of com-
munication. Bell Inst. Tech. J., 27:379-423, 623-
656.

Schluter, D. 1984. A variance test for detecting spe-
cies associations, with some example applications.
Ecology, 65:998-1005.

Thom, R. H. & G. Iffrig. 1985. Directory of Missouri
natural areas. Missouri Dept, of Conserv., Jefferson
City, Missouri.

Thom, R. H. & J. H. Wilson. 1980. The natural
divisions of Missouri. Trans. Missouri. Acad. Sci.,
84:9-23.

Uetz, G. W. 1977. Coexistence in a guild of wan-
dering spiders. J. Anim. Ecol., 46:531-542.

Uetz, G. W. & J. D. Unzicker. 1976. Pitfall trapping
in ecological studies of wandering spiders. J. Ar-
achnol., 3:101-111.

Willson, M. 1974. Avian community structure and
habitat structure. Ecology, 55:1017-1029.

Manuscript received 22 March 1992, revised 15 August
1992.

1992. The Journal of Arachnology 20:173-178

WEB ORIENTATION, THERMOREGULATION, AND PREY
CAPTURE EFFICIENCY IN A TROPICAL FOREST SPIDER

Leslie Bishop and Sean R. Connolly: Department of Biology, Earlham College,
Richmond, Indiana 47374 USA

ABSTRACT. No correlation was found between the web angle and the directional web orientation (in relation
to the sun) for the orb webs of Leucauge regnyi in the Luquillo Forest of Puerto Rico. These data suggest that
the web angle of L. regnyi is not a thermoregulatory response. In addition, prey capture efficiency of sticky traps
placed at the mean angle of webs of L. regnyi is compared with traps placed at vertical and horizontal orientations
in three sites in this tropical forest. Prey captured in sticky traps indicate a trend of prey availability in this
ecosystem: vertically oriented traps catch fewer prey than horizontal traps or traps at the mean angle of web
orientations; at sites of little or intermediate ecological disturbance, mean angle orientations catch more prey
than horizontal orientations. Although confirmation that traps and spiders capture the same types of prey is
lacking, it may be that in this tropical forest system, more prey are made available to spiders with horizontal
and nearly horizontal web orientations than to spiders with vertical webs.

INTRODUCTION

A number of orb-weaving spiders, including
Metabus, Tetragnatha, Conoculus, Uloborus,
Gasteracantha and Leucauge, weave webs that
vary in their orientation relative to the degree to
which they are in vertical versus horizontal planes
(Eberhard 1971, 1989, 1990;Buskirk 1975; Cha-
con & Eberhard 1980). However, no conclusive
explanation has been offered for this horizontal
rather than vertical orientation (Eberhard 1 990).
It has been postulated that: 1) horizontal orbs
receive less damage by the wind (Eberhard 1971),
but others (Biere & Uetz 1981) found that wind
did not affect web orientation; 2) the oscillations
of nearly horizontal orb webs effectively inter-
cept slow flying dipterans (Craig et al. 1985;
Eberhard 1990); and 3) a more horizontal ori-
entation may allow a spider to build closer to a
microhabitat containing abundant prey (Eber-
hard 1990), or to utilize limited support struc-
tures for web building (Buskirk 1975). A corre-
lation between web size and angle found by
Eberhard (1988) may also be due to habitat struc-
ture. Some webs of varying angle, however, have
been observed where structure is available for
vertical webs and a non-vertical orientation does
not appear to increase proximity to an advan-
tageous microhabitat (Eberhard 1971, 1990).

Web angle may also affect rates of prey capture.
Chacon & Eberhard (1980) and Craig (1987) dis-
cuss three factors determining the prey capture

efficiency of a spider web: the probability of in-
sect encountering the web, the absorption by the
web of the prey’s kinetic energy, and the reten-
tion of prey on the web. Vertical webs appear to
be favored in at least the first and last of these
criteria. Chacon & Eberhard (1980) found that
vertical sticky traps in an open field caught nearly
three times as many prey as horizontal sticky
traps and twice as many as inclined (45°) traps.
Eberhard (1986) suggests that horizontal webs
decrease effective distance between silk strands
for prey moving horizontally, and thus make such
orientations energetically costly. Horizontal webs
do not retain prey as well as vertical webs, since
prey freeing themselves from strands of vertical
webs tend to fall into lower strands (Eberhard
1989).

Habitat structure may affect this apparent re-
lationship between web angle and prey capture.
Spiders that build nearly horizontal webs occur
in forest and edge microhabitats (Milne & Milne
1980), as well as across streams (Buskirk 1975)
and in deserts (Eberhard 1971). Perhaps insect
flight patterns differ according to habitat struc-
ture. Eberhard (1990) has suggested that hori-
zontal webs may capture prey falling from above.
Numbers of prey falling from above may be high-
er in a forest than in an open field, since forest
vegetation grows to a height from which prey
may fall and be intercepted by a web.

To date, most studies on the directional ori-
entation (North-South) of orb webs have not em-

173

174

THE JOURNAL OF ARACHNOLOGY

phasized prey capture efficiency, [but see Tol-
bert’s work on Argiope (1979) and various studies
of colonial Metepeira spp. (Uetz & Cangialosi
1986; Uetz 1989; Uetz & Hodge 1990)]. Most
studies attribute a thermoregulatory role to the
directional orientation of webs (Carrel 1978;
Tolbert 1979; Biere & Uetz 1981; Caine & Hei-
ber 1 987). Yet only a few studies have speculated
on possible thermoregulatory aspects of web an-
gle (vertical vs horizontal): Krakauer (1972) sug-
gests that the slight angle of the webs of Nephila
clavipes facilitates postural thermoregulation, and
Tolbert (1979) argues that a vertical orientation
allows a spider to heat up rapidly early and late
in the day by exposing a large surface area to the
sun, and keeps it from overheating at midday by
exposing a small surface area to the sun. Thus
there may be important thermoregulatory as-
pects of web angle. Specifically, the interaction
of vertical angle and directional orientation may
affect the surface area of the spider exposed to
direct sunlight, and may, therefore, play a ther-
moregulatory role.

The tropical orb weaver Leucauge regnyi (Te-
tragnathidae) (Simon), that inhabits the Luquillo
forest of Puerto Rico, has been observed to build
webs of varying angles (vertical-horizontal) in
habitats ranging from treefall gaps to forests with
dense canopies (Bishop, pers. obs.). Consequent-
ly, information on the web building behavior and
ecology of this species might contribute to our
understanding of the thermoregulatory impli-
cations of web angle and the prey availability
patterns of the forest ecosystems inhabited by
horizontal orb-weavers. This study attempts to
investigate what factors affect the angle of the
orb webs of L. regnyi. We will test two specific
hypotheses: 1) the angle of the web of L. regnyi
is a behavioral thermoregulatory response, and
2) the angle of the webs of L. regnyi maximizes
prey capture.

METHODS

Study Site. — During June 1991, we studied
Leucauge regnyi at the El Verde field station in
the Luquillo Experimental Forest of Puerto Rico,
a tropical tabonuco forest (Brown et al. 1983).

Hypothesis 1: Thermoregulation.— We tested
Hypothesis 1 in edge microhabitats where sun-
light exposure is greatest and thermoregulatory
mechanisms should be most pronounced. We
measured the angles, in degrees of departure from
vertical, of 200 webs using a clinometer (Suunto
PM-5/360 PC). Webs were measured where the

lowest value could be obtained, i.e., along the
line on which the web was most vertical. When
webs were not strictly planar, we used the mean
of the web slant above (« = 1) and below (n =
1) the hub. We recorded the compass direction
of the sun in the morning (1000 h) and the af-
ternoon (1500 h). We also recorded the direction
faced by the side of each web on which the spider
rested. We measured the smallest angle between
the direction faced by each web and each of the
compass directions of the sun, respectively,
yielding two values, a morning and afternoon
directional angle, each between 0° and 1 80°, for
each of the 200 webs. These values represented
the web’s orientation relative to the sun, and they
were examined for correlations with the angle of
the spider webs.

Hypothesis 2: Prey Capture Efficiency.— We
collected prey capture data at three sites of vary-
ing levels of disturbance due to the passage of
Hurricane Hugo ( 1 989) through the forest. These
areas of varying disturbance expose the spiders
to differences in prey availability due to changes
in forest structure among the sites (Bishop, pers.
obs.). The high degree of variability among the
sites should provide a rigorous test of the hy-
pothesis: despite habitat type, the angle of the
web maximizes prey capture in this forest eco-
system. The three sites used in this study were
the following — Least disturbance (Site 1): Most
of the canopy-level trees were left standing after
the hurricane, but with foliage and many branch-
es damaged. By June 1991, the canopy was grow-
ing back. Because this area was relatively undis-
turbed, the understory remained relatively open.
Intermediate disturbance (Site 2): This site was
characterized by many fallen trees and branches,
and dense successional growth (primarily Cecro-
pia) generally 4-6 m tall. Most disturbance (Site
3): This site was in a treefall gap with very dense
plant debris on the forest floor. In June 1991,
successional growth was short (< 3 m).

All prey capture data were collected using sticky
traps, consisting of embroidery hoops 25 cm in
diameter and covered on both sides with cheese
cloth and pest glue (Stickem Special, R. Seabright
Industries, Emeryville, California) and placed
approximately 1.5 m off the ground. Traps were
generally hung on the branches of saplings or on
dead, fallen trees. First, in order to determine
whether we would need to separate data from
traps of varying directional orientation in testing
Hypothesis 2 (Castillo & Eberhard 1983), we
placed nine traps facing North/South and nine
facing East/West and counted the prey inter-

BISHOP & CONNOLLY ■=■ WEB ANGLE OF LEUCAUGE REGNYI

175

DIRECTIONAL ANGLE (MORNING)

Figure L— Scattergram for the directional angle rel-
ative to the morning sun (1000 h) and vertical angle
of the orb webs of Leucauge regnyi (n = 200) in the
Luquillo Experimental Forest of Puerto Rico. No cor-
relation was found (N = 0.001, p > 0.05), so the equa-
tion and line are not shown. Measurements are in de-
grees.

cepted during a 24 h period. The results were
analyzed for significant differences using a G-test
(Sokal & Rohlf 1987).

We then used sticky traps to test prey avail-
ability at each of the three disturbance sites. We
positioned eight traps suspended by wire from
the vegetation at each of three orientations for a
72 h period to determine prey availability: ver-
tical, horizontal, and the mean angle of the actual
webs of Leucauge regnyi at that site. We chose
a three day sampling period to minimize error
due to variation in prey availability caused by
daily differences in weather or other periodical
parameters, and we interspersed traps of different
orientations within an area of approximately 20
m^.

The mean angle orientation was determined
from measurements of 50 web (all adult females)
angles at each site, according to the equation for
means from a normal distribution given by Krebs
(1989). Mean angles at the sites were 55.5°, 63.6°,
and 60.5° for sites 1, 2 and 3, respectively and
variation among the means was significant (one-
factor ANOVA; F = 3.5 1 8, T = 0.032). Fisher’s
test indicates that only the difference between
sites 1 and 2 was significant.

We counted the insects captured on each trap
and used G-tests to determine if there were sig-
nificant differences in the number of prey cap-
tured at each of the three sticky trap angles at
each site. Multiple G-tests were used to allow

DIRECTIONAL ANGLE (AFTERNOON)

Figure 2.— Scattergram for the directional angle rel-
ative to the afternoon sun (1500 h) and vertical angle
of the orb webs of Leucauge regnyi {n = 200) in the
Luquillo Experimental Forest of Puerto Rico. No cor-
relation was found (A = 0.001, p > 0.05), so the equa-
tion and line are not shown. Measurements are in de-
grees.

testing among specific orientations, so only high-
ly significant results {P < 0.01) were accepted.

RESULTS

Thermoregulation.— The direction of the
morning sun relative to web orientation was
measured at 95° at 1000 h on 22 June 1991, and
the direction of the afternoon sun was measured
at 290° at 1 500 h on the same day. Web orien-
tation relative to the sun ranged from 0° to 1 80°
in both the morning and afternoon with no clear
modal value. Web angles varied betv/een 0° and
90°, with a mean of 59° (SD = 17.86, n = 200
webs). We found no correlation between web po-
sition relative to the sun and the angle of the
webs in the morning or afternoon (morning: ri
= 0.001, p >> 0.05, n = 200; afternoon: U =
0.001, P>> 0.05, n = 200) (Figs. 1, 2).

Prey capture.— We found no significant dif-
ferences in the number of insects captured by
North/South versus East/West facing sticky traps
(G-test: Gadj = 0.007, P > 0.9), so we did not
separate sticky traps by direction for sampling at
the three sites.

At Site 1 (least disturbance), horizontal traps
caught significantly more prey than vertical traps
(G-test: Gadj = 43.118, P < 0.001), and mean
angle traps caught significantly more prey than
vertical traps (Gadj = 167.413, P < 0.001) and

176

THE JOURNAL OF ARACHNOLOGY

eoo

o

UJ

&

CL

SITE 1

■ Vertical
^ Horizontal
H Mean angle

SITE 2

PLOT TYPE

SITE 3

Figure 3.— Comparison of the total numbers of in-
sects caught by vertical, horizontal, and mean angle
sticky traps in three different plot types in the Fuquillo
Experimental Forest of Puerto Rico. The mean angle
was determined by measuring the angles of 50 webs of
Leucauge regnyi at each site and calculating the means
of those webs (55.5°, 63.6°, and 60.5° for sites 1, 2 and
3, respectively). Horizontal and mean angle traps caught
significantly more prey than vertical traps at all three
sites (G-test: p < 0.001 in each case). Mean angle traps
caught significantly more prey than horizontal traps at
sites 1 and 2 (G-test: p < 0.001), but there were no
significant differences between the numbers of prey
caught by horizontal and mean angle traps at site 3 (p
> 0.1).

horizontal traps = 42.200, P < 0.00 1 ). Like-
wise, at Site 2 (intermediate disturbance), hori-
zontal traps caught significantly more prey than
vertical traps (G^dj = 55.839, P < 0.001), and
mean angle traps caught significantly more prey
than vertical traps (G^dj = 161.970, P < 0.001)
and horizontal traps (G^dj = 29.073, P < 0.001).
At site 3 (most disturbance), horizontal traps
again caught significantly more prey than vertical
traps (Gadj = 216.558, P < 0.001), and mean
angle traps caught significantly more prey than
vertical traps (Gadj = 239.659, P < 0.001), but
there was no significant difference between hor-
izontal and mean angle traps, although mean an-
gle traps caught slightly more prey (Gadj = 0.619,
P > 0.1) (Fig. 3).

DISCUSSION

These results suggest that the web angle of Leu-
cauge regnyi maximizes the web’s prey capture
efficiency and is not influenced by the orientation
of the web relative to the sun. This does not
necessarily mean that thermoregulation exerts no

influence on web angle. For example, Krakauer
(1972) has argued that web angle may facilitate
postural thermoregulation in Nephila clavipes.
Because we did not quantify posturing behavior,
further research is required before concluding that
this facilitation does or does not occur in L. reg-
nyi.

The prey capture patterns from our data in a
tropical forest contrast with the pattern of the
open field studied by Chacon & Eberhard ( 1980),
in which vertical traps caught the most prey and
horizontal traps caught the fewest. In our study,
vertical sticky traps caught significantly fewer prey
than either horizontal or mean angle traps, and
neither vertical nor horizontal traps caught more
prey than the traps positioned at the mean angle
of the spider webs measured. The fact that mean
angle traps at the least and intermediate distur-
bance sites caught significantly more prey than
horizontal traps suggests that the advantage of
the inclined orientation used by Leucauge regnyi
may be even greater in these two sites.

The data suggest that insect flight patterns in
forest systems differ from those in open fields,
as meteorological data on wind movement would
imply (Pedgley 1982). Vegetation structure may
be partly responsible for this. Forest structure
probably allows for more vertical movement of
insects, and this would account for results in
which horizontal and inclined traps catch more
prey relative to vertical traps than in an open
field. Because mean angle traps did not catch
significantly less prey than horizontal traps at any
sites, and in fact caught more at two sites, it is
unlikely that this pattern can be attributed ex-
clusively to prey falling from above, as suggested
by Eberhard ( 1 990). The smaller horizontal sweep
of mean angle traps would result in fewer prey
captured if this was the sole explanation (Eber-
hard 1986). Rather, it appears more likely that
mean angle traps interfere with more flight pat-
terns than horizontal traps. It may also be more
difficult for insects to see and avoid horizontal
and mean angle traps (Craig 1 990).

Although the reasons for the differences in prey
capture found in this study cannot yet proceed
beyond speculation, the results themselves raise
significant problems for further research in the
web ecology of spiders in forest systems. In light
of the apparent greater efficiency of more hori-
zontal orientations in intercepting available prey,
the vertical orientation of most forest orb-weav-
ers is surprising. However, before vertical ori-
entations are ascribed wholly to other factors.

BISHOP & CONNOLLY- WEB ANGLE OF LEUCAUGE REGNYI

177

such as thermal stress (Tolbert 1979; Caine &
Heiber 1 987) or structure for web building (Eber-
hard 1988), data on the planar prey availability
of the actual prey of both horizontal and vertical
orb weavers are needed. To avoid the inaccuracies
that sticky traps would bring to such a study,
actual webs with spiders would have to be reori-
ented (Eberhard 1989) and observed during the
foraging periods of the spiders studied. Further,
data on horizontal and vertical orb v/eavers found
in the same microhabitats would provide infor-
mative comparisons.

Although sticky trap data can be used for com-
parisons and for the evaluation of web charac-
teristics such as vertical angle, drawing conclu-
sions about v/eb ecology from sticky trap data
has been treated with skepticism (Castillo &
Eberhard 1983; Eberhard 1990), due to the dif-
ferences in the qualities of sticky traps and orb
webs. Most notably, the sticky traps we used are
much more visible than orb webs; thus, the fre-
quencies and types of prey caught in our study
may be different than those of actual webs. Sticky
traps only mimic the encounter function of orb
webs and capture potential prey items, not nec-
essarily the actual prey of Leucauge regnyi. Fur-
thermore, we did not classify the prey caught in
our traps by taxon or size. However, because
small dipterans constituted the overwhelming
majority of sticky trap captures, and the same
group also constitutes most of the diet of L. reg-
nyi (Bishop, pers. obs.), and because the data
showed an extremely high level of significance
{P < 0.001 in all instances of significant differ-
ences), the general trend may well be biologically
significant.

The significant differences in actual spider web
angle among the sites may not be solely related
to prey capture efficiency, although such a pos-
sibility can not be ruled out. Vegetational differ-
ences among sites may affect web angle by chang-
ing the structure available for web building
(Buskirk 1975; Gillespie 1987; Eberhard 1988),
or web angle may vary as a response to some
unknown parameter. The variation in orienta-
tion among forest sites with different structure,
along with the implications of the sticky trap
results, add to current knowledge of the selective
forces acting on spider webs, but the contrast that
our findings present with past research further
complicates the problem of thoroughly assessing
the role of these forces on the evolution and ecol-
ogy of orb web orientation. Most importantly,

they indicate that relevant parameters may vary
considerably among ecosystems.

ACKNOWLEDGMENTS

This research was funded by a grant for un-
dergraduate research from the Ford Foundation,
awarded through Earlham College to Leslie Bish-
op. Additional support was provided to Bishop
by the Exline-Frizzell Fund of the California
Academy of Sciences, and an Oak Ridge Asso-
ciated Universities Faculty Travel Grant. We
would like to express our gratitude to the staff of
the El Verde Field Station (Center for Energy and
Environmental Research, University of Puerto
Rico) for the use of their facilities and for pro-
viding access to the tropical study sites. We also
thank Ann Rypstra and George Uetz for discus-
sion and suggestions during the planning of this
project, and Andrea Condit and Sandra Encalada
for assistance in the field. Susan E. Riechert, Wil-
liam Eberhard, and A1 Cady provided valuable
reviews of the manuscript.

LITERATURE CITED

Biere, J. M. & G. W. Uetz. 1981. Web orientation
in the spider Micrathena gracilis (Araneae: Aranei-
dae). Ecology, 62:336-345.

Brown, S., A. E. Lugo, S. Silander & L. Liegel. 1983.
Research history and opportunities in the Luquillo
Experimental Forest. Institute of Tropical Forestry,
SO-44.

Buskirk, R. E. 1975. Coloniality, activity patterns
and feeding in a tropical orb-weaving spider. Ecol-
ogy, 56:1314-1328.

Caine, L. A. & C. S. Heiber. 1987. Web orientation
in the spider Mangora gibberosa (Hentz) (Araneae,
Araneidae). J. Arachnol., 15:263-265.

Carrel, J. E. 1978. Behavioral thermoregulation dur-
ing winter in an orb-weaving spider. Symp. Zool.
Soc. London, 42:41-50.

Castillo, J. A. & W. G. Eberhard. 1983. Use of ar-
tificial webs to determine prey available to orb
weaving spiders. Ecology, 64:1655-1658.

Chacon, P. & W. G. Eberhard. 1 980. Factors affecting
numbers and kinds of prey caught in artificial spider
webs, with considerations of how orb webs trap prey.
Bull. British Arachnol. Soc., 5:29-38.

Craig, C. L. 1987. The ecological and evolutionary
interdependence between web architecture and web
silk spun by orb web weaving spiders. Biol. J. Linn.
Soc., 30:135-162.

Craig, C. L. 1990. Effects of background pattern on
insect perception of webs spun by orb-weaving spi-
ders. Anim. Behav., 39:135-144.

Craig, C. L., A. Okubu & V. Andreasen. 1985. Effect
of spider orb-web and insect oscillations on prey
interception. J. Theor. Biol., 1 15:201-21 1.

178

THE JOURNAL OF ARACHNOLOGY

Eberhard, W. G. 1971. The ecology of the web of
Uloboriis diversus (Araneae: Uloboridae). Oecolo-
gia, 6:328-342.

Eberhard, W. G. 1986. Effects of orb-web geometry
on prey interception and retention. Pp. 70-100, In
Spiders: webs, behavior, and evolution. (W. A. Shear,
ed.). Stanford University Press; Stanford, Califor-
nia.

Eberhard, W. G. 1988. Behavioral flexibility in orb
web construction: effects of supplies in different silk
glands and spider size and weight. J. Arachnol., 16:
295-302.

Eberhard, W. G. 1 989. Effects of orb web orientation
and spider size on prey retention. Bull. British Ar-
achnol. Soc., 8:45-48.

Eberhard, W. G. 1990. Function and phylogeny of
spider webs. Ann. Rev. Ecol. Syst., 21:341-372.

Gillespie, R. G. 1987. The mechanism of habitat
selection in the long-jawed orb-weaving spider Te-
tragnatha elongata (Araneae, Tetragnathidae. J. Ar-
achnol., 15:81-90.

Krakauer, T. 1972. Thermal responses of the orb-
weaving spider, N. clavipes. American Midi. Nat.,
88:245-250.

Krebs, C. J. 1989. Ecological Methodology. Harper
and Row, New York.

Milne, L. & M. Milne. 1980. The Audubon Soc. Field
Guide to North American Insects and Spiders. Al-
fred A. Knopf, New York.

Pedgley, D. E. 1982. Windbome Pests and Diseases:
the meteorology of airborne organisms. Halsted
Press, New York.

Sokal, R. R. & F. J. Rohlf. 1987. Introduction to
Biostatistics. W. H. Freeman and Company, New
York.

Tolbert. W.W. 1979. Thermal stress of the orb weav-
ing spider Argiope trifasciata (Araneae). Oikos, 32:
386-392.

Uetz, G. W. 1989. The “ricochet effect” and prey
capture in colonial spiders. Oecologia, 81:154-159.

Uetz, G. W. & K. R. Cangialosi. 1986. Genetic dif-
ferences in social behavior and spacing in popula-
tions of Metapeira spinipes, a communal-territorial
orb weaver (Araneae, Araneidae). J. Arachnol., 14:
159-173.

Uetz, G. W. & M. A. Hodge. 1990. Influence of hab-
itat and prey availability on spatial organization and
behavior of colonial web-building spiders. Nat. Geog.
Res., 6:22-40.

Manuscript received 24 October 1991, revised 29 August
1992.

1992. The Journal of Arachnology 20:179-188

NUMERICAL RESPONSE TO PREY ABUNDANCE BY
ZYGIELLA X-NOTATA (ARANEAE, ARANEIDAE)

David A. Spiller; Department of Zoology and Center for Population Biology,
University of California, Davis, California 95616 USA

ABSTRACT. To test the effect of prey abundance on the orb spider Zygiella x-notata I conducted two field
experiments at the same site. In the first experiment, from 20 June to 9 September 1978, I augmented prey
abundance in two plots interspersed within four control plots; at the end of the experiment the mean number
of spiders was 2.7 times higher in prey-augmented plots than in control plots, and the difference was highly
significant {P = 0.005). In the second experiment, from 25 June to 16 September 1979, I augmented prey
abundance in one plot located in between two control plots; at the end of the experiment the number of spiders
was > 3.0 times higlier in the prey-augmented plot than in each control plot. Although numbers of spiders in
the 1979 experiment could not be analyzed statistically because the treatment was not replicated, they support
the results of the 1978 experiment. I monitored the phenology of the population at the experimental site and
an unmanipulated population at a different site throughout 1978 and 1979. In both years reproduction began
earlier in the experimental population than in the unmanipulated population. In 1979 I collected eggs sacs in
both populations. The experimental population contained more eggs and heavier eggs than those than in the
unmanipulated population. Within the experimental population, egg sacs in the prey-augmented plot contained
heavier eggs than those in control plots.

Numerous comparative and experimental
studies indicate that spiders often experience food
shortages in nature (review in Wise 1992). How-
ever, the evidence that spider population sizes
respond numerically to temporal fluctuations in
prey abundance is equivocal. Some observation-
al studies found positive correlations between
abundances of spiders and their potential prey
(e.g., Wingerden 1978) but others found no cor-
relation (e.g., Greenstone 1978). Furthermore, a
positive correlation could be due to both spiders
and their prey responding to some other envi-
ronmental factor. An experimental study in rice
fields showed that spider densities increased in
areas where Drosophila flies were released rela-
tive to control areas (Kobayashi 1975); unfor-
tunately, problems in experimental design and
presentation of statistical tests make it difficult
to assess the observed response (Wise 1 992). Sev-
eral studies showed that spiders moved from ar-
eas with low prey abundance to areas with high
prey abundance (Riechert & Gillespie 1986); this
“aggregative response” (Hassell & May 1974)
changes the predator’s distribution, but not nec-
essarily the predator’s population size. Extensive
studies of Agelenopsis aperta (Gertsch) demon-
strated that reproductive success was related to
prey availability (Riechert & Tracy 1975); how-
ever, population size was limited by suitable web

sites (via territoriality), and was not influenced
by prey abundance (Riechert 1981).

Several possible factors could restrict a nu-
merical response by spiders. Riechert & Lockley
(1984) suggested that the extent to which spiders
respond numerically is limited by long genera-
tion tim,es of spiders relative to their prey and
by strong self limitation within spider popula-
tions. Field experiments indicate that predators
may often reduce spider populations (Askenmo
1977; Gunnarsson 1983; Wise 1982; Pacala &
Roughgarden 1984; Polls & McCormick 1986;
Spiller & Schoener 1988). Frequently, both spi-
ders and their predators eat some of the same
types of prey (Polls et al. 1989; Spiller & Schoe-
ner 1990). Therefore, a numerical response by
spiders could be impeded by spider predators
responding numerically to the same prey. Wise
(1992) concluded that although many field ex-
periments have shown that food shortages limit
spider growth rate and fecundity, more experi-
ments are needed to determine whether these
parameters are translated into increased popu-
lation size in the next generation.

Zygiella x-notata (Clerck) is a common orb
spider in coastal areas of California (Gertsch 1 964;
Levi 1974). In this study, to test for a numerical
response by Z. x-notata I conducted two prey-
augmentation experiments at the same site. In

179

180

THE JOURNAL OF ARACHNOLOGY

Table 1.— Mean (1 SE) web radii and estimated mean
web areas for different size classes of Z. x-notata. Mean
web areas were estimated by assuming that webs were
circular (which they were approximately).

Class

Body

length

(mm)

n

Web

radius

(cm)

Esti-

mated

mean

web

area

(cmq

Hatchlings

<2.0

40

3.3 (0.1)

34

Juveniles

2.0-3.9

36

6.3 (0.2)

125

Subadults

4.0-5. 9

42

8.4 (0.3)

222

Adult females

>6.0

103

11.4 (0.3)

415

addition, I compared some life-history charac-
teristics of the population at the experimental
site to those of an unmanipulated population at
a different site.

METHODS

Censusing procedures.— ZygtW/a x-notata
adult females are about 6. 0-8.0 mm in body
length (measured from the chelicerae to the pos-
terior end of the abdomen). Adult males are about
4. 5-6. 5 mm. In this study I divided the spiders
into 4 age/size classes: hatchlings (< 2.0 mm),
juveniles (2.0-3. 9 mm), subadults (4. 0-5. 9 mm)
and adults (> 6.0 mm and smaller adult males).
During field censuses I counted the numbers of
spiders in each class. To test my accuracy in size
determination, I assigned 90 individuals to these
size classes in the field, and then collected and
measured them with an ocular micrometer; 83%
were assigned to the correct size class.

I censused two Z. x-notata populations re-
peatedly for over two years. The first population
inhabited a group of 1 1 abandoned cabins lo-
cated about 100 m inland from the beach at Coal
Oil Point, which is approximately 2 km west of
the University of California, Santa Barbara Cam-
pus. Most spiders nested underneath ledges that
were 1.5 m above the ground and surrounded
each cabin. I censused all individuals within 0.5
m of the ledges (total area censused was 162 m^)
about once a month from November 1977 to
June 1980. The second population inhabited an
assemblage of large boulders, constructed for
erosion control, at the base of a cliff on the Uni-
versity of California, Santa Barbara Campus
Beach. The assemblage was about 100 m long
and 6 m wide, and was about 0.5- 1.0 m above

mean higher high water. The spiders built their
orbs in caves formed by the boulders and caught
mostly flies that bred in drift kelp on the beach.
I censused all individuals within a 48 m x 6 m
section of boulders about once a month from
October 1977 to February 1980. The first pop-
ulation was unmanipulated whereas the second
was prey augmented (see below).

Beginning in July 1 978, 1 recorded the number
of egg sacs and the numbers of arthropods in the
spiders’ webs and being consumed by the spiders
in each site. For each census I computed an index
of prey availability by dividing the total number
of arthropods counted by the estimated total area
of webs censused. I computed the total area of
webs censused from the numbers of spiders with
webs in each size class and the estimated mean
web area of each size class. Mean web area was
estimated by measuring the radii of a large sam-
ple of webs in the field (Table 1).

Prey-augmentation experiments,— On 19 June
1978, 1 divided the 48 m x 6 m section of boul-
ders into six contiguous 8 m x 6 m plots, and
then censused each plot. From 20 June to 9 Sep-
tember 1978 I augmented prey abundance in the
second and fifth plots; thus, there were two treat-
ment plots and four control plots (Fig. 1). I chose
this arrangement, rather than assigning treat-
ments randomly, to ensure that treatment plots
were interspersed within control plots, as sug-
gested by Hurlbert (1984). To augment prey
abundance, I put large quantities of drift kelp at
the bottom of caves within the treatment plots
during the first few days of the experiment, and
added smaller amounts of kelp and sea water
about twice a week. I censused each plot at about
1-2 wk intervals from 25 June to 10 September
1978, and at about 2-3 wk intervals from 24
September to 10 December 1978.

On 24 June 1979, 1 divided the same 48 m x
6 m section of boulders into three contiguous 16
m X 6 m plots, and then censused each plot.
From 25 June to 16 September 1979 I put kelp
in the center plot; thus, there was one treatment
in between two controls (Fig. 1). I chose this
arrangement so that prey was augmented in a
different location in this year than in the previous
year. I censused each plot about once a month
from July to December 1979.

Egg sac collections.— From 2 September to 1 5
October 1979 I collected all egg sacs at the un-
manipulated site and all that I could find in each
plot at the experimental site (I might have missed
some sacs located inside crevices at the experi-

SPILLER-NUMERICAL RESPONSE BY ZYGIELLA X-NOTATA

181

mental site). During this time period I visited
each site at least twice a week and collected each
new sac produced. All the eggs from each sac
were counted, dried at 60 °C for 24 hr, and
weighed together. I did not analyze sacs contain-
ing hatchlings.

Analyses.— To assess whether adding kelp to
the treatment plots increased prey availability
during the experiment in 1978, I computed the
mean index of prey availability (number of prey/
m- web) recorded in each plot from 25 June (first
census after I began adding kelp) to 1 0 September
(one day after my last addition), and then com-
pared the mean indices in treatment and control
plots with a one-tailed t test. To test the overall
treatment effect on Z. x-notata in 1978, I per-
formed a one-tailed t test on total number of
individuals in each plot on 10 September; in
addition, I analyzed the change in total number
of individuals in each plot from 1 9 June (the day
before the experiment) to 10 September. Prey-
availability indices, numbers of egg sacs and
numbers of each age/size class recorded during
each census are given for descriptive purposes,
but they are not statistically analyzed. I present
the data recorded during the 1979 experiment
for descriptive purposes, but the treatment effect
cannot be tested statistically because kelp was
added to only one plot.

From the data on egg sacs collected in 1979,
I analyzed two variables: number of eggs per sac
and mean biomass per egg per sac. I treated each
of the three plots in the experimental site and
the unmanipulated site as four separate groups.
For each variable I performed a one-way ANO-
VA with three contrasts. Two contrasts tested
the variation among plots within the experimen-
tal site (treatment vs. control 1 + control 2; con-
trol 1 vs. control 2) and one contrast tested the
variation between sites (experimental site [all
plots] vs. unmanipulated site). The purpose of
these tests is to describe the extent to which the
variables in the areas differed, but they cannot
be interpreted directly as tests of the hypothesis
that the spiders are food limited.

RESULTS

Life-history observations,— The population at
the unmanipulated site exhibited an annual life
cycle (Fig. 2). In 1978, a cohort emerged in early
spring, matured during late spring and early sum-
mer, reproduced during late summer and fall,
and declined during fall and winter. The offspring
produced in 1978 overwintered in egg sacs and

1978

Cl Tl

C2 C3

T2 C4

1979

C 1

T

C2

Figure 1.— Spatial design of the prey-augmentation
experiments. In 1978 prey abundance was augmented
in plots T1 and T2. In 1979 prey abundance was aug-
mented in plot T (see Methods for details).

emerged during late winter and early spring 1979.
This cohort matured during late spring and early
summer, reproduced during late summer and fall,
and declined during fail and winter. Note that I
collected 19 egg sacs from 2 September to 15
October 1979; 27 egg sacs were produced from
16 October 1979 to 15 January 1980. The off-
spring produced by the 1979 cohort emerged
during late winter and early spring 1980 and de-
veloped later in the spring. Prey-availability in-
dices were relatively high during winter and
spring, and were relatively low during summer
and fall.

The phenology of the experimental population
was more complex (Fig. 3). Hatchlings emerged
in fall 1977 but they either died or emigrated
during severe winter storms. (During a storm in
January 1978 I observed waves breaking on the
boulders; the next day no spider was present in
the site.) The site was apparently recolonized in
late winter and spring 1978 by juveniles and su-
badults. This cohort matured in late spring and
early summer 1978, and some adult females re-
produced in early summer. A second cohort
emerged later in the summer and developed in
fall 1978. Two cohorts emerged in 1979. The
first emerged during late winter and early spring
and matured later in the spring. Some adult fe-
males reproduced in early summer 1979. The
second cohort emerged later in the summer and
developed in fall 1 979. Note that this cohort was
relatively small because I collected all visible egg
sacs within the censused area from 2 September
to 1 5 October 1979; hatchlings that were counted
during this time period emerged either from hid-
den sacs within the censused area or from sacs
outside the censused area.

Prey-availability indices tended to be higher
at the experimental site than at the unmanipu-

182

THE JOURNAL OF ARACHNOLOGY

Figure 2. — Prey-availability indices (no. prey/m^ web) and numbers of Z. x-notala in the unmanipulated site.
Prey availability index and number of egg sacs were not recorded in the initial 10 censuses. Egg sacs were
collected from 2 September to 15 October 1979.

lated site, particularly during the summer when
kelp was added to the treatment plots. For all
censuses, mean ± 1 SD of the prey-availability
indices at the experimental site and the unman-
ipulated site were respectively, 45.8 ± 42.9 and
16.6 ± 23. 1. For censuses in summer (1978 and
1979), mean ± 1 SD of indices at the experi-
mental site and the unmanipulated site were re-
spectively, 60.4 ± 40.7 and 3.79 ± 1.82.

Prey-augmentation experiment in 1978. — On
1 9 June (one day before kelp was added to treat-
ment plots), prey-availability indices and num-
bers of spiders in treatment and control plots did
not dilfer significantly (Table 2). During the ex-
periment mean prey-availability indices were
significantly higher in treatments than in con-
trols. On 10 September mean number of spiders
was 2.7 times higher in prey-augmented plots
than in control plots, and the difference was high-
ly significant. Numbers of spiders increased in

all plots from 19 June to 10 September; the in-
crease was significantly greater in treatments than
in controls.

Details in Fig. 4 show that in July mean num-
bers of adults became higher in treatments than
in controls, whereas mean numbers of smaller
individuals were nearly identical in treatments
and controls. In August mean numbers of egg
sacs became higher in treatments than in con-
trols. In September hatchlings emerged from the
egg sacs and mean numbers of immature spiders
became higher in treatments than in controls.
Shortly after I stopped adding kelp to the treat-
ments, prey-availability indices were about equal
in treatments and controls, but numbers of spi-
ders remained higher in treatments for a few
months.

Prey-augmentation experiment in 1979.— On
24 June (one day before kelp was added to the
treatment plot) prey-availability indices and

SPILLER-= NUMERICAL RESPONSE BY ZYGIELLA X-NOTATA

183

Figure 3.— Prey-availability indices (no. prey/m^ web) and numbers of Z. x-notata in the experimental site.
Prey-availability index and number of egg sacs were not recorded in the initial eight censuses. Egg sacs were
collected from 2 September to 15 October 1979. Arrows depict time periods when kelp was added to treatment
plots.

Table 2.— Prey-availability indices (no. prey/m^ web) on 19 June (one day before kelp was added to treatment
plots), mean of the indices from 25 June (first census during kelp additions) to 10 September (one day after the
last kelp additions) and total numbers of spiders and change in numbers (number on 10 September minus
number on 19 June) in each plot during the 1978 prey-augmentation experiment. [‘ one-tailed test; each test
was performed on the numbers given in each column.]

Plot

No. prey/m^ web

Mean of

25 June to

19 June 10 Sept.

Total number of spiders

19 June

10 Sept.

Change

Treatment 1

10.2

84.2

55

491

436

Treatment 2

11.3

108.6

63

680

617

Control 1

9.3

8.3

133

327

194

Control 2

6.0

16.4

80

182

102

Control 3

16.3

19.3

45

173

128

Control 4

64.1

57.2

37

183

143

II

0.65

3.96

0.44

4.61

6.14

P

0.553

0.008'

0.677

0.005'

0.002'

184

THE JOURNAL OF ARACHNOLOGY

0^

bJ

m

PREY/m^ WEB

TREATMENTS
0---0 CONTROLS

25 9 30 13 27 10

JUNE JULY AUG SEPT

OCT

19

NOV DEC

Figure 4. — Mean ± 1 SE prey-availability indices {no. prey/m^ web) and numbers of Z. x-notata in treatment
and control plots during the 1978 prey-augmentation experiment. The arrow depicts the time period when kelp
was added to treatment plots.

numbers of spiders were similar in the treatment
and control plots (Fig. 5). On 29 July prey-avail-
ability index and number of egg sacs were higher
in the treatment than in controls but numbers of

spiders remained about the same. On 25 August
prey-availability index and numbers of egg sacs,
hatchlings, subadults and adults were higher in
the treatment than in controls. On 16 September

SPILLER-- NUMERICAL RESPONSE BY ZYGIELLA X-NOTATA

185

LlJ

QQ

TREATMENT
CONTROL 1
CONTROL 2

JUNE

JULY

AUG SEPT

OCT NOV DEC

Figure 5.— Prey-availabiiity indices (no. prey/m^ web) and numbers of Z. x-notata in the treatment and control
plots during the 1979 prey-augmentation experiment. The arrow depicts the time period when kelp was added
to treatment plots. Egg sacs were collected from 2 September to 1 5 October.

prey-availability index and numbers of all spider
classes were higher in the treatment than in con-
trols; total number of spiders was >3.0 times
higher in the treatment than in each control plot.
(Note that this year egg sacs were collected from

2 September to 15 October.) Prey-availability
indices and numbers of spiders remained higher
in the treatment than in controls for a few months
after I stopped adding kelp to the treatments.

Egg sacs collected in 1979. — Within the ex-

186

THE JOURNAL OF ARACHNOLOGY

Table 3. —Mean (1 SE) number of eggs and mean biomass per egg in Z. x-notata sacs collected from 2
September to i 5 October 1979. Prey was augmented in the treatment plot at the experimental site. [‘ one-tailed
test.]

Site

Plot

n

No. of
eggs/sac

Mean biomass
(mg)/egg/sac

Experimental

Treatment

27

74.1 (3.6)

0.163(0.004)

Control 1

13

66.7 (6.0)

0.146 (0.006)

Control 2

6

63.7(8.8)

0.145 (0.003)

Unmanipulated

19

51.5(2.9)

0.130(0.005)

ANOVA:

df

SS J

P

Number of eggs

Treatment vs. Controls 1 + 2

1

814.6

2.45

0.06 12‘

Control 1 vs. Control 2

1

37.6

0.11

0.7377

Experimental vs. Unmanipulated

Error

1

61

3299.9

20 261.5

9.93

0.0013'

Mean biomass/egg

Treatment vs. Controls 1 + 2

1

0.00303

7.30

0.0044'

Control 1 vs. Control 2

1

0.00000

0.00

0.9544

Experimental vs. Unmanipulated

Error

1

61

0.00561

0.02529

13.53

0.0003'

perimental site mean biomass per egg was sig-
nificantly greater in the treatment than in control
plots (Table 3). Numbers of eggs per sac tended
to be higher in the treatment than in control
plots, but the difference was not significant at the
0.05 level. Numbers of eggs and mean biomasses
per egg in the two control plots were very similar.
Number of eggs and mean biomass per egg were
significantly greater in the experimental site than
in the unmanipulated site.

DISCUSSION

Phenologies of unmanipulated and experi-
mental populations differed considerably. The
unmanipulated population reproduced during late
summer and fall, and the next generation emerged
during late winter and early spring in the follow-
ing year. The experimental population began to
reproduce in early summer, and some of the next
generation emerged later in the same season. In
addition, egg sacs produced in the experimental
population contained more eggs and heavier eggs
than those produced in the unmanipulated pop-
ulation.

Field experiments on other web spiders showed
that growth rates or fecundities were influenced
by food supply (Wise 1975, 1979; Spiller 1984;
Spiller & Schoener 1990). My indices of prey
availability during summer 1978 and summer

1979 were much higher at the experimental site
than at the unmanipulated site. This suggests that
the differences between populations in life-his-
tory characteristics were influenced by higher prey
abundance at the experimental site. However,
this interpretation should be taken with caution
for two reasons. First, differences between sites
in physical factors (e.g., temperature) might have
influenced life-history characteristics. Second, my
indices of prey availability did not take into ac-
count the size of individual prey; therefore, were
prey sizes larger at the unmanipulated site than
at the experimental site, comparisons between
sites could be misleading.

Within the experimental population, egg sacs
in the treatment plot contained heavier eggs than
those in controls. Although this difference was
statistically significant, the analysis does not
demonstrate that it was caused by prey abun-
dance because in 1979 the treatment was not
replicated (Hurlbert 1984). Hence, comparisons
among plots within the experimental population
are subject to the same caveats as the comparison
between populations. However, the fact that the
treatment plot was in between the two control
plots, and egg biomasses in control 1 and control
2 were nearly identical, provides compelling ev-
idence that prey abundance influenced egg bio-
masses. Number of eggs per sac tended to be

SPILLER- NUMERICAL RESPONSE BY ZYGIELLA X-NOTATA

187

higher in the treatment plot than in the controls
but the difference was not significant at the 0.05
level. Possibly, the number of eggs that a repro-
ductive female produces in a sac is determined
before the biomass of each egg. Therefore, if an
adult female moved from a control plot to the
treatment plot after number of eggs was deter-
mined, the increased food supply might have
increased egg biomass but not number of eggs.

The prey-augmentation experiment in 1978
demonstrated that Z. x-notata responded nu-
merically to prey abundance. During July and
August numbers of adults became higher in treat-
ments than in controls. Three different mecha-
nisms could have produced this result: 1 . adults
moved from control plots to treatment plots, 2.
adult survivorship was higher in treatments than
in controls, or 3. developmental rate of imma-
tures was higher in treatments than in controls.
Because marked individuals were not followed
during the experiment I cannot assess the im-
portance of these possible mechanisms. Follow-
ing the increase in adults, numbers of egg sacs
became higher in treatments than in controls;
subsequently, the numerical response became
more pronounced when the second generation
emerged in September. During the 1979 exper-
iment numbers of spiders became substantially
higher in the treatment plot than in control plots.
I could not statistically analyze the results of this
experiment because the treatment was not rep-
licated in 1979. However, the data support the
overall results of the 1978 experiment.

Rypstra’s (1983) enclosure experiments showed
that food abundance influenced densities of sev-
eral web-spider species; interestingly, some sol-
itary species exhibited some degree of coloniality
when prey abundance was high (Rypstra 1986).
Although Z. x-notata is typically solitary, other
studies found that individuals were attracted to
conspecific silk, and that some individuals re-
duced their web sizes in response to crowding
(Leborgne & Pasquet 1 987a, 1 987b). In this study,
Z. x-notata webs were occasionally attached to
one another in treatment plots when prey avail-
ability was high. Such behavior might have fa-
cilitated the numerical response by Z. x-notata.

An important factor that accounted for the
numerical response by Z. x-notata was the emer-
gence of a second generation in the same year.
Many spider species have obligatory annual life
cycles, and would probably not exhibit such a
marked numerical response within a season
(Riechert & Lockley 1984). Thus, the extent to

which spiders respond numerically may depend
on the behavior and phenology of the species.

ACKNOWLEDGMENTS

I wish to thank Joseph H. Connell for his en-
couragement and support. I thank M. Fawcett
and N. Spiller for field assistance, H. Levi for
identifying specimens, and L. Bishop, M. Green-
stone, T. Schoener, and D. Wise for comments.
I was supported by NSF grant BSR90-20052
while preparing the manuscript.

LITERATURE CITED

Askenmo, C., A. von Broemssen, J. Eckman & C. Jans-
son. 1977. Impact of some wintering birds on spi-
der abundance in spruce. Oikos, 28:90-94.

Gertsch, W. J. 1964. The spider genus Zygiella in
North America (Araneae: Argiopidae). American
Mus. Novit., 2188:1-21.

Greenstone, M. H. 1978. The numerical response to
prey availability of Pardosa mmulosa (McCook)
(Araneae: Lycosidae) and its relationship to the role
of spiders in the balance of nature. Symp. Zool. Soc.
London, 42:183-193.

Gunnarsson, B. 1983. Winter mortality of spruce-
living spiders: effect of spider interactions and bird
predation. Oikos, 40:226-233.

Hassell, M. P. & R. M. May. 1974. Aggregation of
predators and insect parasites and its effect on sta-
bility. J. Anim. EcoL, 43:567-594.

Hurlbert, S. H. 1984. Pseudoreplication and the de-
sign of ecological field experiments. Ecol. Monogr.,
54:187-21 1.

Kobayashi, S. 1975. The effect of release

on the spider population in a paddy field. Appl. Ent.
Zool., 10:268-274.

Leborgne R. & A. Pasquet. 1987a. Influence of con-
specific silk-structures on the choice of a web-site
by the spider Zygiella x-notata (Clerck). Revue Ar-
achnol., 7:85-90.

Leborgne R. & A. Pasquet. 1987b. Influences of ag-
gregative behavior on space occupation in the spider
Zygiella x-notata (Clerck). Behav. Ecol. Sociobiol.,
20:203-208.

Levi, H. W. 1974. The orb-weaver genus Zygiella
(Araneae: Araneidae). Bull. Mus. Comp. Zool., 146:
267-290.

Pacala, S. & J. Roughgarden. 1984. Control of ar-
thropod abundance by Anolis lizards on St. Eusta-
tius (Neth. Antilles). Oecologia, 64:160-162.

Polls, G. A. & S. J. McCormick. 1986. Scorpions,
spiders and solpugids: predation and competition
among distantly related taxa. Oecologia, 71:111-
116.

Polls, G. A., C. A. Myers & R. D. Holt. 1989. The
ecology and evolution of intraguild predation: po-
tential competitors that eat each other. Annu. Rev.
Ecol. Syst., 20:297-330.

188

THE JOURNAL OF ARACHNOLOGY

Riechert, S. E. 1981. The consequences of being ter-
ritorial: spiders, a case study. American Nat., 1 17:
871-892.

Riechert, S. E. & R. G. Gillespie. 1 986. Habitat choice
and utilization in web-building spiders. Pp. 23-48,
In Spiders: Webs, Behavior, and Evolution. (W. A.
Shear, ed.) Stanford Univ. Press, Stanford, Califor-
nia.

Riechert, S. E. & T. Lockley. 1984. Spiders as bio-
logical control agents. Annu. Rev. Entomol., 29:
299-320.

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, 56:265-284.

Rypstra, A. L. 1983. The importance of food and
space in limiting web-spider densities: a test using
field enclosures. Oecologia, 59:312-316.

Rypstra, A. L. 1986. High prey abundance and a re-
duction in cannibalism: the first step to sociality in
spiders (Arachnida). J. Arachnol., 14:193-200.

Spiller, D. A. 1984. Competition between two spider
species: experimental field study. Ecology, 65:909-
919.

Spiller, D. A. & T. W. Schoener. 1988. An experi-
mental study of the effect of lizards on web-spider
communities. Ecol. Monogr., 58:57-77.

Spiller, D. A. & T. W. Schoener. 1990. Lizards reduce
food consumption by spiders: mechanisms and con-
sequences. Oecologia, 83:150-161.

Wingerden, W. K. R. E. van. 1978. Population dy-
namics of Erigone arctica (White) (Araneae, Liny-
phiidae) 11. Symp. Zool. Soc. London, 42:195-202.

Wise, D. H. 1975. Food limitation of the spider L/n-
yphia marginata: experimental field studies. Ecol-
ogy, 56:637-646.

Wise, D. H. 1979. Effects of experimental increase
in prey abundance upon female reproductive rates
of two orb-weaving spider species. Oecologia, 41:
289-300.

Wise, D. H. 1982. Predation by a commensal spider,
Argyrodes trigonum, upon its host: an experimental
study. J. Arachnol., 10:1 1 1-1 16.

Wise, D. H. 1992. Spiders in Ecological Webs. Cam-
bridge Univ. Press. Cambridge, UK.

Manuscript received 8 June 1992, revised 23 October
1992.

1992. The Journal of Arachnology 20:189-199

A NEW SPECIES OF NORTH AMERICAN TARANTULA,

APHONOPELMA PALOMA

(ARANEAE, MYGALOMORPHAE, THERAPHOSIDAE)

Thomas R. Prentice: Department of Entomology, University of California, Riverside,
California 92521, USA

ABSTRACT. Aphonopelma paloma new species, is distinguished from all other North American tarantulas
by its unusually small size and presence of setae partially or completely dividing the scopula of tarsus IV in
both sexes. Both sexes also are characterized by a general reduction of the scopula on metatarsus IV. Males are
characterized by a swollen third femur.

In 1939 and 1940 R. V. Chamberlin and W.
Ivie described almost all of the currently recog-
nized North American theraphosid spiders. De-
spite the acknowledged significance of their work,
it is difficult to apply Chamberlin’s keys with
much success even in dealing with specimens
from type localities, primarily because their small
sample sizes did not allow variational assess-
ment. Eleven of these species descriptions were
based on single males, five on single females, and
three on two males each (Chamberlin & Ivie 1 939;
Chamberlin 1 940). By increasing the sample size
and working with both sexes, I describe here a
new species of tarantula, Aphonopelma paloma,
based on characters that have been selected after
evaluating variation. Because of deficiencies in
variational data the values of characters such as
spination, meristic and morphometric ranges, and
various ratios as taxonomic tools have not been
assessed adequately. Therefore, specific ratios and
morphometric comparisons as well as tables
showing spination patterns and leg and palp seg-
ment lengths and ranges are provided. These data
should be useful in future taxonomic studies. A
brief account of the natural history of A. paloma
is included.

METHODS

Measurements were made using an American
Optical 570 stereoscope equipped with an eye-
piece micrometer accurate to 0.1 mm. Leg and
pedipalp measurements were made on the side
appearing most normal (e.g., not in the process
of regeneration). Trochanters and coxae mea-
surements were performed from ventral aspect,
other leg and pedipalp segments from dorsal as-
pect (Coyle 1971). Carapace length was taken

with anterior and posterior edges in the same
plane. All ink drawings except femora were aided
by a camera lucida. Palpal bulb and seminal re-
ceptacles were cleared in 10% NaOH (for 12 hr.
at 50 °C.) prior to illustration. Scanning electron
micrographs were taken with a JEOL JSM C35.
Abbreviations for eyes are standard for Araneae.
For leg spination, abbreviations are as follows:
a = apical, b = basal, d = dorsal, e = preapical,
L = left, m = medial, p = prolateral direction, r
= retrolateral direction, R = right, usu. = usually,
V = ventral, var. = variable, 0.33, 0.50, etc. =
approximate fraction of the total segment length
a spine is from the proximal end. Color refer-
ences are from the color charts in the Munsell
Book of Color and refer to color of live specimens
under natural light conditions.

Aphonopelma paloma, new species
Figs. 1-12

Types.— -Holotype male, allotype female from
Pinal County, Arizona, 3 mi. NE exit 151 off I-
8, 17-18 November 1989. Paratype males: 15
November 1986 - 1, 18-19 November 1989 -
2, 1 7 November 1 990 - 2; Paratype females: 1 8
Nov. 1989 - 2. Holotype and allotype deposited
in the American Museum of Natural History.

Etymology.— The specific epithet is from the
Spanish word paloma (dove) which was used in
the plural to describe a vast plain in the south-
western desert of Arizona - Palomas Plain - where
there is an abundance of these tiny tarantulas.

Diagnosis.— differs from other North
American theraphosid spiders by the presence of
setae partially or completely dividing tarsus IV
scopula and, to a lesser extent, tarsus III scopula
(Figs. 1-3). Additional characters that further dif-

189

190

THE JOURNAL OF ARACHNOLOGY

ferentiate A. paloma from other small species are
its unusually small size, reduced metatarsal sco-
pulation, and swollen third femora of males. Of
the four small species that may overlap A. pal-
oma in size {Chaunopelma radinum Chamberlin
1940; A. marxi Simon 1891 (Bonnet 1939); A.
phasmus ChamberYm 1940; and A. ingrami Jung
1975 (Jung 1975; invalid name from unpub-
lished thesis, types deposited under that name
in the American Museum of Natural History)),
only A. ingrami (name used for reference only)
males have been found to do so but lack swollen
third femora. Males from the remaining three
species also lack swollen femora. All genera =
Rhechostica Simon 1892 (Raven \ 9S5) = Aphon-
opelma Pocock 1901 (Bull. Zool. Nomen. 48 (2)
June 1991, Opinion 1637).

Description. — Male: Holotype male. Length
14.5 mm. Carapace length 5.7 mm, carapace
width 5.1 mm carapace width/carapace length
0.89. Tibia I/metatarsus I, 1.09. Leg span 47
mm. Entire tarantula black except silvery reflec-
tion from black carapace hairs, grayish silver
chelicerae, and interspersed long straw-colored
abdominal setae. Ventral aspect also black ex-
cept orange color of labium and anterior most
portions of maxillae. Leg and palp segment
lengths are in Table 1. Third femur noticeably
swollen (Figs. 4-6), width, viewed from above,
1.7 mm, femora I and IV at widest point 1.1
mm. Lower process of tibial spur with one me-
gaspine, long upper process with two shorter me-
gaspines (Fig. 7).

Chelicerae clothed in pale grayish-silver pu-
bescence, longer silverish setae interspersed on
dorsal surfaces. Cheliceral width 2.7 mm. Chel-
iceral width/carapace width, 0.53. Promargins of
each fang furrow with eight macro teeth.

Carapace clothed with medium long dense
black hairs (Munsell, 5Y 2/1), not closely ap-
pressed. Thoracic groove a transverse pit. Ce-
phalic region not rising abruptly from thoracic
region, but in gradual arch. Extreme posterio-
dorsal surface with scattered medium length bris-
tles, shorter bristles just anteriad.

Ocular area compact and elevated, turret com-
paratively low. Anterior eye row slightly pro-
curved; AME round, diameter 0.3 mm, separat-
ed from each other by their diameter; ALE oval,
length 0.8 X AME diameter; AME-ALE O.lx
AME diameter. Posterior eye row nearly straight,
at most slightly procurved; PME rounded, 0.7 x
AME diameter, separated from each other by

Figures 1-3.— T. paloma, new species. Scanning
electron micrographs showing setae (arrows) dividing
scopula on right tarsi, ventral: 1, paratype male, tarsus
III (top), tarsus IV (bottom), x 24; 2, center section of
tarsus IV in Fig. 1, x200; 3, paratype female, tarsus
III (top), tarsus IV (bottom), x 36.

1.5 X AME diameter; PLE rounded, 0.7 x AME
diameter; PME-PLE nearly contiguous; ALE-PLE
0.4 X AME diameter; AME-PME 0. 1 x AME di-
ameter; median ocular quadrangle wider poste-
riorly.

Abdomen with short black (Munsell, N 1/)

PRENTICE-NEW SPECIES OF APHONOPELMA

191

Figures 4-6.—Aphonopelma paloma, new species. Right femora of holotype male showing relative widths,
dorsal: 4, femur IV; 5, femur I; 6, femur III.

pubescence, long dark brown setae with straw-
colored apices (Munsell, 5 YR 7/10, 10 YR 8/
6) and longer attenuated setae, bases dark brown,
distal halves straw-colored (Munsell, 10 YR 8/
6) copiously interspersed. Longest setae less dense
on venter than dorsum. Circular patch of black
type I (Cooke et al. 1972) urticating hairs (Fig.
8) covering posteriodorsai % of abdomen.

Legs black and hirsute; black pubescence in
addition to medium long dark brown to black
setae with straw-colored apices, medium density,
and very long straw-colored setae with dark
brown bases, low density. All tarsi fully scopulate
with setae dividing tarsus IV scopula on proxi-
mal 0.60. Extent of metatarsal scopula: meta-
tarsus 1, 0.67; metatarsus II, 0.50; metatarsus III,

192

THE JOURNAL OF ARACHNOLOGY

Figure 1 .—Aphonopelma paloma, new species. Left
tibia I of holotype male showing general shape of spur
processes and megaspine location, prolateral aspect;
mgs = megaspine.

distal 0.25 divided by setae; metatarsus IV, distal
0.17, divided by setae.

Spination: leg I, metatarsus ld(p0.33) 2v(lap
lam), tibia 2d(lp0.33 lpO.67) L4v(lrb lrO.50
2re) R5v(lrb 2r0.50 2re), femur Id(pe); leg II,
metatarsus ld(p0.40) 4v(lap lam lar lrO.50),
tibia 2d(lp0.33 lpO.67) L4v(lpe Ire lrO.50 Irb)
R5v(lpe Ire irO.50 2rb), femur Id(pe); leg III,
metatarsus 4d(lpe Ire lp0.60 lr0.60) R5v(lap
2ar lmO.50 lpO.50) L7v(lap lam. 2ar 2p0.50
lrO.50), tibia 4d(lp0.33 lrO.33 lpO.75 lrO.75)
R4v(2pe lpO.50 lmO.50) L5v(2pe Ire lpO.50
lmO.50); leg IV, metatarsus 4d(lpe Ire lpO.50
lrO.50) 9v(5a 2r0.40-0.75 lpO.50 lmO.50), tibia
R3d(lp0.25 lrO.25 lrO.75) L3d(lp0.25 lrO.25
lpO.75); palp, tibia 2v(lp0.50 lrO.50), femur
Id(pe).

Sternum widest between bases of second and
third coxae, sigilla at margins of sternum op-
posite coxae I, II and III, posterior sigillum larg-
est. 37 cuspules on proximoventral surface of
right maxilla, 39 on left. 40 labial cuspules.

Table l.~ Aphonopelma paloma, holotype male: leg
and pedipalp length (mm).

Leg

I

II

III

IV

Palp

Coxa

2.3

2.0

1.8

1.9

1.9

Trochanter

1.3

1.2

I.O

1.1

1.2

Femur

6.1

5.5

5.0

5.8

3.6

Patella

2.7

2.4

2.1

2.4

1.8

Tibia

4.8

4.0

3.4

4.7

3.4

Metatarsus

4.4

4.4

4.7

6.0

Tarsus

2.7

2.7

2.7

3.0

1.0

Total leg
length

24.3

22.2

20.7

24.9

12.9

Female: Allotype female. Length 16.2 mm.
Carapace length 5.2 mm, carapace width 4.2 mm,
carapace width/carapace length 0.81. Tibia 1/
metatarsus I, 1.17. Leg span 33 mm. Female
gray; legs and carapace bronze-gray (Munsell, 10
YR 4/2) and abdomen black-gray (Munsell, 10
YR 3/1). Ventral aspect also gray except orange-
brown color of labium and anterior halves of
maxillae. Third femur not swollen as in male.
Leg and palp segment lengths in Table 2.

Cheiicerae clothed in grayish pubescence with
long silverish setae as in holotype; cheliceral width
3.1 mm. Cheliceral width/carapace width, 0.74.
Promargin of right fang furrow with eight ma-
croteeth, left fang furrow with seven macroteeth.
Carapace clothed in moderately short, dense
bronze-gray hairs more closely appressed than
in male. Numerous black attenuated setae, api-
cally straw-colored, interspersed. Thoracic groove
transverse, slightly procurved. Cephalic region
rising abruptly from thoracic region, in profile
steeper and higher than in male. Eye arrangement
similar to holotype; AME diameter slightly great-

Figure Aphonopelma paloma, new species. Type

I urticating hair, Scanning electron micrograph, x780.

PRENTICE- NEW SPECIES OF APHONOPELMA

193

Table 2.—Aphonopelmapaloma, allotype female: leg
and pedipalp length (mm).

Leg

I

II

III

IV

Palp

Coxa

2.0

1.6

1.4

1.7

1.9

Trochanter

1.1

0.9

0.8

1.0

1.1

Femur

3.9

3.4

3.0

4.0

3.0

Patella

2.2

1.9

1.8

2.0

1.6

Tibia

2.8

2.3

1.9

3.1

2.2

Metatarsus

2.4

2.2

2.4

3.6

Tarsus

1.8

1.6

1.6

2.0

1.9

Total leg
length

16.2

13.9

12.9

17.5

11.7

er than 0.2 mm, separated from each other by
0.7 X AME diameter; ALE round (appear ovoid
when viewed from above), slightly smaller than
AME; AME-ALE 0.2 x AME diameter; PME
round, diameter somewhat greater than 0. 1 mm,
separated from each other by 1.8 x AME di-
ameter; PLE somewhat ovoid, length 0.8 x AME
diameter; PME-PLE, ALE-PLE, AME-PME as
in holotype.

Abdomen clothed in short dark gray pubes-
cence. Color of medium long and longest setae
as in holotype. Medium long setae copiously in-
terspersed, longest setae sparse, confined mostly
within circular patch of black urticating hairs on
posterior % of abdomen.

Legs covered with bronzish-gray pubescence
in addition to short and medium length black-
based straw-gray setae. Longest setae as in ho-
lotype but less dense. All tarsi fully scopulate,
setae dividing tarsus IV proximal 0.88, tarsus III
proximal 0.33. Extent of metatarsal scopula:
metatarsus I, 0.67; metatarsus II, 0.50; metatar-
sus III distal 0.33, divided by setae; metatarsus
IV, few scattered scopula hairs on distal end.

Spination: Leg I, metatarsus !v(am), tibia
lv(p0.40), femur Id(pe); leg II, metatarsus

ld(p0.50)4v(lap lam lar lr0.40), tibia ld(p0.50)
Rlv(r0.50) L(0); leg III, metatarsus 4d(lpe Ire
lpO.50 lrO.50) R6v(lap lam 2ar lpO.50 lrO.67)
L7v(2ap lam 2ar lpO.67 lmO.67), tibia
2d(ip0.50 IrO.50) R3v(lpe lp0.60 lm0.60)
L2v(lpe lp0.60); leg IV, metatarsus 4d(lpe Ire
lpO.50 lrO.50) R6v(lap lam lar lpO.50 lmO.75
lrO.50) L8v(lap lam 2ar lpO.67 lpO.88 lmO.67
lrO.88); palp, tibia 3v(2pe Ire), femur Id (pe).

Widest point of sternum and position of sigilla
as in holotype. 43 maxillary cuspules on proxi-
moventral surface of right maxilla, 48 on left. 36
labial cuspules.

Variation.— No evidence has been found to
correlate intraspecific variation with the geo-
graphical distribution of A. paloma. Allometric
and other morphometric variations most likely
reflect phenotypic plasticity. However, because
of their possible value in species designation,
specific variational ranges and ratios are included
in this section.

Males: 10 (nine with leg measurements in-
cluding holotype). Overall length 9.9-14.1 mm.
Carapace length 4. 2-6. 2 mm, width 3.6-5. 5 mm.
Carapace width/carapace length 0.85-0.95, mean
0.89. Cheliceral width/carapace width 0.49-0.59,
mean 0.54. Eight cheliceral macroteeth most
common (7, 7 in one individual, 9, 8 in another,
and 7, 8 in a third). Tibia I, in all specimens,
longer than metatarsus I; tibia I/metatarsus I,
1.04-1.11, mean, 1.08. Metatarsus IV/carapace
length >1 as in holotype. Leg and palp segment
length ranges are in Table 3. Femur III 1.36-
1 .58 X width of femora I and IV. Femur I equal
to or slightly longer than femur IV. Variation in
leg and palpal spination is recorded in Table 4
showing approximate position and percent oc-
currence of each spine. Relatively little variation
was found in the palpal bulb. Interspecific dif-
ferences in bulb morphology have proven to be
of some value in distinguishing theraphosid spe-

Table 3.—Aphonopelma paloma, 9 males, holotype: range of leg and pedipalp segment lengths (mm). For
example, smallest specimen = 1.7 mm, largest specimen = 2.4 mm.

Leg

I

II

III

IV

Palp

Coxa

1. 7-2.4

1.5-2. 1

1. 3-1.9

1. 4-2.2

1.4-2. 1

Trochanter

0.9-1. 3

0.8-1. 2

0.8-1. 1

0.8-1. 3

0.9-1. 4

Femur

4.4-6.3

4.0-5. 8

3.6-5.2

4.4-6. 1

2.7-3. 7

Patella

2.0-2.7

L7-2.4

1. 6-2.3

1. 7-2.4

1.4-1. 9

Tibia

3. 5-4.8

2.9-4.2

2.5-3. 5

3.6-5.0

2.6-3. 3

Metatarsus

3. 2^.6

3. 1-4.6

3.3-5.0

4.4-6.5

Tarsus

1. 8-2.7

1.9-2. 7

1. 9-2.7

2.2-3.2

1.0-1. 2

194

THE JOURNAL OF ARACHNOLOGY

Table 4.—Aphonopelma paloma, male (9 males, holotype) spination. Approximate position and percent
occurrence of each leg and palpal spine. Spination abbreviations are defined in the Methods section of text.
Dorsal (d) refers to upper half, ventral (v) to lower half of segment. An asterisk (*) indicates present on at least
one of segment pair in all specimens; two asterisks (**) indicate present on at least one of segment pair except
in the smallest specimen.

100-80%

79-60%

<60%

d-Me-I

v-Me-I

l-am(lOO)

l-pl/3-l/2(69)

l-ap(44)

l-ar(6)

l-pl/4-l/2(15)

d-Ti-I

l-p2/3(100)

l-pl/3(89)**

v-Ti-I

l-er(lOO)

2nd-er(67) at least on 1

2nd-er(56)

l-l/2(m or p)(100)

of leg pr.

l-b(p or m)(56)

2nd-v p 1/2(11)
2nd-bp(22)

3rd-b(p or m)(l 1)

d-Fe-I

l-ep(lOO)

d-Me-lI

l-pl/2(83)**

v-Me-lI

3-a(lp Im lr)(94)*

4th-a(22)

l-l/2(r or m)(100)

l-additional(17)

d-Ti-Il

l-p2/3(94)*

l-pl/3(83)*

l-ep(ll)

v-Ti-II

2-e(lp lr)(94)*

2nd-ep(17)

l-r2/5-2/3(100)

2nd-er(6)
l-em(l 1)
l-br(5)

1-m 1/2(28)
l-var.(38)

d-Fe-II

d-Me-lII

2-e(lp lr)(100)

2-l/2(lp lr)(83)*

l-ep(61)

v-Me-IlI

4-a(lp Im 2r)(89)*

5th-a(25)

2-l/2(lp Im or r)(94)*

3rd-v(17)

d-Ti-III

2-2/3(lp lrusu.)(78)**

l-pl/3(56)

l-rl/3(33)

v-Ti-III

l-ep(94)*

l-er(67)**
l-pl/2(78)
l-l/2(r or m)(72)

2nd-ep(28)

d-Me-IV

2-e(lp lr)(83)**
l-rl/2(83)*

1-p 1/2(72)

v-Me-lV

4-a(lp Im 2r)(100)

5th-a(r)(78)

6th-a(ll)

3-v(lp Ir usu.)(100)

4th-v(78)**

7th-a(ll)

5th-v(33)

d-Ti-IV

l-r2/3(89)*

l-r3/5(33)

l-pl/3(33)

l-epdl)

l-p3/5(6)

v-Ti-IV

l-ep{100)

1-p 1/2(27)

l-er(50)

2nd- 1 /2m usu.(61)

2nd-ep(44)

d-Ti-Palp

l-p3/5(33)

l-ep(17)

v-Ti-Palp

l-pl/2(61)

1-r 1/2(40)
l-var.(l 1)

d-Pt-Palp

I-P(ll)

d-Fe-Palp

l-ep(56)

PRENTICE-NEW SPECIES OF APHONOPELMA

195

0.5 mm

Figure 9 .—Aphonopelma paloma, new species. Left
palpal tarsus of paratype male showing cymbium, bulb,
and convolution of the receptaculum seminus, ventral
aspect, ad = apical division of buib; bd = basal divi-
sion; cy = cymbium; md = middle division; rs = re-
ceptaculum seminus.

cies. Apical and middle divisions of A. paloma
bulb subequal, basal % of apical division tapering
rapidly, distal third uniformly narrow (Fig. 9).
Extent of scopula; metatarsus 1, 0.50-0.88; meta-
tarsal II, 0.40-0.60; metatarsus III, 0.20-0.40;
metatarsus IV, 0.20 to few scattered hairs. Tarsus
IV scopula divided by stout setae often forming
a band, proximal 0.50 to entire; tarsus III scopula
divided by weaker setae, proximal 0.0-0.88;
metatarsus III and IV divided by setae, meta-

tarsus III sometimes by scattered setae. 30-60
maxillary cuspules per maxilla, 28-57 labial cus-
pules.

Females: 1 1 (including allotype). Overall
length ranging from 1 2.2-1 9.0 mm averaging 1 5.0
mm. Little color variation in newly molted fe-
males. Faded individuals lighter gray-brown or
gray-bronze. Carapace length ranging 4.3-6. 1
mm, width 3. 6-5.0 mm, carapace width/cara-
pace length 0.8 1-0.84, mean 0.82. Carapace width
greater than length of femur I and IV. Cheliceral
width/carapace width 0.67-0.76, mean 0.70.
Eight cheliceral macro teeth most common (9, 8
in one individual, 8, 7 in another, 10, 9 in a
third). Tibia I, as in males, longer than metatar-
sus I; tibia 1/metatarsus I 1.19-1.29, mean 1.22.
Metatarsus IV shorter than both carapace length
and width. Leg and pedipalp segment length
ranges are in Table 5. Femur III not sv/ollen as
in male. Femur I equal to or shorter than femur
IV. Variation in leg and palp spination is re-
corded in Table 6 showing approximate position
and percent occurrence of each spine. Only two
forms of seminal receptacle shape have been
found (Fig. 10, 1 1), both in paratypes. The scler-
otized nature of each form suggests that variation
may result from phenotypic plasticity. Extent of
metatarsal scopula: metatarsus I, 0.67-full;
metatarsus II, 0.50-0.60; metatarsus III, 0.25-
0.40; metatarsus IV, 0-0.17. Tarsus IV scopula
divided by stout setae often forming a setal band,
tarsus III proximal 0.40-0.67 divided by weaker
setae except undivided in one specimen. Meta-
tarsus III and IV divided by setae. 38-70 max-
illary cuspules per maxilla, 3 5-5 5 labia! cuspules.

Distribution. — Habitable terrain is in the SW
Arizona desert from 680-1950 ft. elevation, on
relatively flat and sandy bajadas and plains con-
ducive to burrowing. Major vegetation types in-
clude creosote bush, saguaro cactus, ocotillo.

Table 5.— Aphonopelma paloma, 10 females, allotype: range of leg and pedipalp segment lengths (mm). For
example, smallest specimen— 1.6 mm, largest specimen— 2.3 mm.

Leg

I

II

III

IV

Palp

Coxa

1. 6-2.3

1. 4-2.0

1.2-1. 8

1. 4-2.0

1. 6-2.2

Trochanter

0.9-L2

0.8-1. 0

0.7-0.9

0.8-1. i

0.9-1. 1

Femur

3.1-4.6

2.7-^.0

2.5-3.6

3.3-4.8

2.5-3.6

Patella

1.8-2. 5

1. 5-2.3

L4-2.0

1. 7-2.3

1.4-1.9

Tibia

2.3-3.4

1. 9-2.7

1.5-2. 3

2.4-3. 5

1. 7-2.5

Metatarsus

1. 9-2.8

1. 8-2.7

1. 9-3.0

2.8-M.2

Tarsus

1. 3-2.2

1.3-2. 1

1.3-2. 1

1. 7-2.4

1. 6-2.3

196

THE JOURNAL OF ARACHNOLOGY

Table 6.—Aphonopelma paioma, female (10 females, allotype) spination. Approximate position and percent
occurrence of each leg and palpal spine. Spination abbreviations are defined in the Methods section of text.
Dorsal (d) refers to upper half, ventral (v) to lower half of segment. An asterisk (*) indicates present on at least
one of segment pair in all specimens; two asterisks (**) indicate present on at least one of segment pair except
in one specimen.

100-80%

79-60%

<60%

v-Me-I

l-am(lOO)

l-ap(59}

1-61/3(5)

d-Ti-I

1 -bp 1/4(5)

v-Ti-I

l-l/2(r or m)(27)
l-er(18)

d-Fe-I

l-ep(18)

d-Me-II

i-p 1/2(36)

v-Me-II

3-a(lp Im lr)(86)*
l-rl/3-l/2(86)*

2nd-vl/3-l/2(14)

d-Ti-II

i-pl/2(45)

v-Ti-II

d-Me-III

2-e(lp lr)(100)

2-l/2(lp lr)(95)*

!-rl/2(59)

v-Me-III

4-a(lp Im 2r)(95)*

2-l/2(lp Ir or m)(86)*

5th-a(p)(5)

d-Ti-III

l-r2/5-3/5(86)*

l-p2/5-3/5(45)

l.r3/4-7/8(5)

v-Ti-III

l-ep(82)

2nd-ep(14)

1-1/2(45)

2nd- 1/2(1 4)

d-Me-IV

2-e(lp lr)(82)
l-rl/2(86)**

l-pl/2(55)

v-Me-IV

4-a(lp Im 2r)(95)*

3rd-v(var.)(68)

5th-a(36)

2-l/2(ip Iror m){95)*

6th-a(5)

4th-v(var.)(27)

d-Ti-IV

l-rl/2-7/8(82)**

l-r2/5(27)

v-Ti-IV

l-ep(73)

2nd-e(p usu.)(32)
1-1/2(45)

4th-v(var.)(5)

d-Ti-Palp

l-p(14)

v-Ti-Palp

3-e(lr 2p usu.)(95)*

l-rl/2(36)

l-pl/2(i8)

d-Fe-Palp

l-ep(73)

cholla (Pinal and Pima Co.), acacia, palo verde,
and bur sage (Ambrosia).

Natural history.— ^4. paioma appears to have
a very limited activity cycle. Open burrows have
been observed only between the last week of Oc-
tober and mid-December. After unplugging bur-
row entrances the small tarantulas deposit silk-
bound earth and accumulated debris from bur-
rows either adjacent to or removed from the en-
trances by up to 15 cm. The resulting crescent-
shaped mounds are typical of the species and
have not been observed in other U.S. thera-
phosid species (Fig. 12). The maximum open
burrow density corresponds to the height of the
breeding season in mid-November. After the first

heavy winter rains, often in December, the cres-
cent mounds are flattened and burrows cannot
be found.

Entrance hole diameters for mature females
range from 5-10 mm. Smaller diameter burrows
have been excavated and contained immature
individuals. Burrow mouths are often lined with
a thin layer of silk extending for up to a few mm
onto and below the ground surface in contrast to
most U.S. theraphosids which generally have a
thicker layer extending several mm in either di-
rection. Burrow depths range from 30-62 cm,
but most often are about 50-54 cm, with very
few less than 46 cm. Burrows are primarily ver-
tical with little horizontal deviation. Toward the

PRENTICE -NEW SPECIES OF APHONOPELMA

197

Figures 10, 1 \ .—Aphonopelma paloma, new species.
Spermathecae of two paratype females illustrating two
forms found; 10, bulbs separated by ~ 0.5 mm, egg
canal closed; 1 1, bulbs separated by ~ 0.75 mm, egg
canal open. Anterior direction indicated by arrow.

bottom they widen into horizontal or oblique
terminal chambers. One or two additional side
chambers are often found along the burrow, gen-
erally at junctions where it makes a brief hori-
zontal run or otherwise changes direction.

Most activity is nocturnal. Mature females and
immatures can be observed in the evening hours
either carrying excavated materials with their
chelicerae and pedipalps or waiting near or just
inside their burrows for passing prey. Females
have not been observed more than 20 cm from
the burrows. Prey consists of beetles (many in
the Tenebrionidae), harvester ants (Myrmicin-
ae), and arachnids, including other spiders and
small scorpions. Remains of these have been
found in excavated mounds and in the terminal
chambers.

In the late fall breeding season, males can be
found wandering by day in search of female bur-
rows, especially from late morning until early
afternoon. No active males have been found after
dark. Mating is diurnal and occurs outside the
burrow. Courtship and mating behavior in A.
paloma is typical of the genus (pers. obs.). Upon
finding a female burrow, the male generally taps
the ground near the burrow entrance with his
first pair of legs, often quite forcefully. Gentle
substratum drumming with the palps commonly
accompanies leg tapping, but the two behaviors
are not simultaneous. The spider sporadically
alternates leg tapping and palp drumming with
a stridulating vibration in which his body moves
in a rhythmic up and down fashion with femur
III positioned almost vertically (stridulation
mechanism has not been isolated). Each stridu-

Figure 12.— Aphonopelma paloma, new species. Photograph of burrow entrance showing typical crescent-
shaped excavation mound.

198

THE JOURNAL OF ARACHNOLOGY

lating pulse is of short duration, often less than
three s. The stridulating posture is also seen at
various intervals during the course of the male’s
wandering. In two large U.S. theraphosid species
{A. reversum Chamberlin and A. eutylenum
Chamberlin), vibrations from male stridulation
can be detected by the females from at least 1.2
m through a combination of 1 2 cm earth, 1 6 mm
Plexiglass, and approximately 2.4 m tubular
stainless steel (pers. obs.). Females of these spe-
cies responded by rapidly drumming their first
two pairs of legs. This drumming behavior is also
seen in A. paloma females in response to male
stridulation. These behaviors suggest long dis-
tance communication between male and female
tarantulas. The male continues courtship behav-
ior until the female emerges from the burrow and
contact is made. After mating both spiders usu-
ally run (female often walks) in opposite direc-
tions, the female back into her burrow, the male
several cm away.

It is not known when the female constructs her
cocoon and deposits eggs. Tarantulas of many
other U.S. species lay eggs in spring or early sum-
mer (Baerg 1958; Gertsch 1979). Mostd. paloma
spiderlings have dispersed by September, and
none can be found with females when burrows
are unplugged in October or November. Four
burrows excavated between 1-3 September 1991
contained young. Of the two burrows at the Pinal
Co. site, one contained a female with five young,
one of which had the shortest carapace length,
1.5 mm, the other contained one juvenile (no
female present), had the longest carapace length,
2.1 mm, and was the largest (length 6.1) of all
young found. Of the two remaining burrows at
Sentinel (Maricopa Co.) site 1 contained a female
with three young, the other contained one young
(no female present). All young appeared to be in
their second or third instar. I assume that more
than five young emerge from most egg sacs since
dissected females have been found with approx-
imately 40-100 developing eggs but evidence of
when and how young dispersed has not been
found. Populations of these tarantulas are often
subdivided into loose aggregations where burrow
densities have been observed as high as 14 per
10.7 m\

One species of pompilid wasp was reared from
a parasitized A. paloma female (Sentinel, Mari-
copa Co. site) and was identified as Hemipepsis
ustulata ustulata Dahlbom. No other predators
are known.

DISCUSSION

This study found size, general reduction of
metatarsal scopula, division of tarsal scopula by
setae, and swollen third femora of males to be
reliable characters in distinguishing A. paloma
from other recognized North American Aphon-
opelma. Overlap in size between valid Aphono-
pelma species and A. paloma has not been seen.
However, Jung (1975) described a small species
within which overlap with the largest A. paloma
males occurred. Metatarsal scopula in A. paloma
is more reduced than in the other species. Di-
vision of tarsal scopula by setae (either partially
or wholly), in itself, separates this new species
from all other Aphonopelma (Chamberlin 1940;
Raven 1985). The swollen third femur of the
male is also unique in this genus.

Based on the work presented here, it appears
that many of the characters used by Chamberlin
to distinguish theraphosid species may be highly
variable within a species. Nine of the 24 couplets
in his male Aphonopelma key are based on rel-
ative number and position of leg and palpal
spines. For example, Delopelma and Gosipelma
subgenera were differentiated on the basis of two
and four submedian spines, respectively, on the
prolateral face of the palpal tibia. I have found
that d. paloma type males (also with tibia I longer
than metatarsus I) have 0, 1, or 2 submedial
prolateral spines. Furthermore, Delopelma was
divided into two sets of species by the number
of ventral spine levels between the base and spur
of tibia I, two species with 1 or 2 spine levels on
tibia I and 1-2 ventral spines on the palpal tibia,
and two species with 4 or 5 spine levels on tibia
I and no ventral spines on the palpal tibia. A.
paloma males have 1, 2, or 3 levels of spines on
tibia I and 0, 1 , or 2 ventral spines on the palpal
tibia. Similarly, males of an undescribed Cali-
fornia species (all specimens from a single lo-
cation) have 2-4 submedial prolateral spines, 1-
4 ventral spines on the palpal tibia, and 2-4 levels
of ventral spines on tibia I. All of these examples
suggest that spination is more variable within
species than previously thought. Reliable spi-
nation patterns may emerge as additional vari-
ational data accumulate for each species.

Chamberlin used relative lengths of tibia and
metatarsus I to differentiate Aphonopelma from
Gosipelma and Delopelma species. Although this
character appears to be valuable in distinguishing
between some morphologically similar allopath-
ic species or between sympatric species within

PRENTICE -NEW SPECIES OF APHONOPELMA

199

geographical isolates, ratio reversals of tibia 1/
metatarrsus I may represent opposite ends of
morphoclines within extensive continuous pop-
ulations.

Raven (1985) maintained that Dugesiella
shares with Rhechostica = Aphonopelma (ICZN,
Opinion 1637) the thomlike setae on the pro-
lateral coxae and that no other known characters
merited its continued separation from Aphono-
pelma. Jung (1975) did not feel that prolateral
coxa I setation warranted generic or subgeneric
designation because he found a gradual reduction
of swollen setal bases from Dugesiella to Chaun-
opelma. He found that generic designations of
North American tarantulas were based on char-
acters that are shared in different degrees by all
the genera. In spite of the uniqueness of A. pal-
oma, there is little use in considering it or any
other North American theraphosid as other than
Aphonopelma until variational ranges have been
determined for all proposed species. Only by ex-
amining sufficient numbers of each of these spe-
cies will their variational ranges become com-
parable and the taxonomic significances of these
various characters be realized.

Material examined.— The type specimens and the
following: USA: Arizona; Maricopa Co., 2 mi. N. Sen-
tinel, 690 ft. elev., 1 8 November 1989, 1 female, 4 Jan.
1990, 1 female, 26 Oct. 1990, 2 males, 2 females, 10
Nov. 1990, 1 female; Rd. 355, 1 mi. S. Granite Reef
Aqueduct, 1400 ft. elev., 10 Nov. 1990, 1 female, 11
Dec. 1990, 1 female; Pima Co., 3 miles E. Ajo, 1780
ft. elev., 18 Nov. 1989, 1 male; 0.1 mile N. boundary
Organ Pipe Cactus National Monument, 1950 ft. elev.,
18 Nov. 1989, 1 male; 15 miles. S. Ajo, 1650 ft. elev.,
30 March 1990, 1 female.

ACKNOWLEDGMENTS

Special thanks are extended to Gordon Gordh
and Michael Adams for providing encourage-
ment and expertise in assisting me in the prep-
aration of this description and to Wendell Icen-
ogle for his time and effort in helping me to
acquire the knowledge I have of the North Amer-
ican tarantulas. My special thanks also to David

Headrick and Mario Moratorio for their labo-
ratory assistance, to Willis Gertsch and Vincent
Roth for their on-going encouragement of my
work on U.S. tarantulas, and to Frederick Coyle
for his valuable comments in his review of this
manuscript.

LITERATURE CITED

Baerg, W. 1958. The tarantula. University of Kansas
Press, Lawrence.

Bonnet, P. 1939. Bibliographia Araneorum. Vol. II,
parts 1-5, Vol. III. Toulouse.

Chamberlin, R. V. 1940. New American tarantulas
of the Family Aviculariidae. Bull. Univ. Utah, Vol.
30, No. 13. Univ. of Utah Press, Salt Lake City.
Chamberlin, R. V. & W. Ivie. 1939. New tarantulas
from the southwestern states. Bull. Univ. Utah, Vol.
29, No. 11. University Press, Univ. of Utah, Salt
Lake City.

Cooke, J. A. L., V. D. Roth & F. A. Miller. 1972.
The urticating hairs of theraphosid spiders. Amer-
ican Mus. Nov., No. 2498.

Coyle, F. A. 1971. Systematics and natural history
of the mygalomorph spider genus Antrodiaetus and
related genera (Araneae: Antrodiaetidae). Bull. Mus.
Comp. ZooL, Vol. 141, No. 6.

ICZN. 1991. Opinion 1637. Aphonopelma Pocock,
1901 (Arachnida, Araneae): gives precedence over
Rhechostica Simon, 1892. Bull Zool. Nomen. 48.
Gertsch, W.J. 1979. Mygalomorph spiders. Pp. 100-
1 26, In American Spiders, 2nd Edition. Van Nos-
trand Reinhold Co., New York.

Jung, A. K. S. 1975. Morphological relationships
among five species of sympatric tarantulas (Ara-
neae: Theraphosidae), with descriptions of four new
species. Master’s Thesis, University of Florida.
Munsell Color Company, Inc., 1958. Munsell Book
of Color. Cabinet Edition. Baltimore, Maryland,
U.S.A.

Raven, R. J. 1985. The spider infraorder Mygalo-
morphae (Araneae): cladistics and systematics. Bull.
American Mus. Nat. Hist., 182:1-175.

Simon, E. 1891. Liste des esptees de la famille des
Aviculariides qui habitent I’Amerique du nord. Act.
Soc. Linn. Bord., 44:307-326.

Manuscript received I January 1992, revised 7 July
1992.

1992. The Journal of Arachnology 20:200-206

SEX RATIO IN THE SOCIAL SPIDER DIAEA SOCIALIS
(ARANEAE: THOMISIDAE)

David M. Rowell: Division of Botany and Zoology, Australian National University,
GPO Box 4, Canberra City 2601 Australia

Barbara York Main: Zoology Department, University of Western Australia,

Nedlands, WA 6009 Australia

ABSTRACT. Sex ratio data from embryos and adults are compared in the social thomisid Diaea socialis. The
relative proportions of males and females do not differ significantly between the two data sets, indicating that
a sex ratio bias already exists at the time of fertilization. A statistical comparison with published data for the
theridiid Anelosimus eximius shows a different ratio but similar pattern for the two types of data. A molecular
procedure for determining whether the overproduction of females results from a bias in the sperm or differential
success of sperm in fertilization is described.

In those social spiders where the adult sex ratio
is known, it is female-biased (Buskirk 1981; Voll-
rath 1986). However, Fisher’s principle (Fisher
1930) predicts that selection will favor an equal
parental investment in offspring of both sexes. It
would follow, then, that any bias in sex ratio
should result from forces acting after dispersal.
On the basis of the correlation between sex ratio
and social behavior it would seem reasonable to
hypothesize that the two phenomena are in some
way causally linked. Skewed sex ratios have sim-
ilarly been reported in colonies of Diaea socialis
Main, with an observed male:female ratio of
0.2126 (Main 1988).

The life-history and behavior of D. socialis
were described in Main (1988). Colonies are
founded by individual gravid females which use
silk to bind together eucalypt leaves to form a
nest. The brood spiderlings remain with the
mother after hatching, add leaves cooperatively
to enlarge the nest, share prey, mature in the nest
and mate with siblings. Generally, only mated
females migrate from the nest, over the Septem-
ber to November period. The sex ratios reported
for this species (Main 1988) have been assessed
on adult and preadult morphology only, and con-
sequently the observed bias may reflect differ-
ential life span or mortality between the sexes.

Alternatively, it is possible that the sex ratio
is skewed from birth as a result of meiotic drive
or some other factor. To determine whether such
processes are operating requires the ascertain-
ment of gender as soon as possible after fertil-

ization, and preferably before hatching. Using
cytological techniques, Aviles and Maddison
(1991) have demonstrated a primary female bias
in embryos of the social species Anelosimus ex-
imius Keyserling and A. domingo Levi (propor-
tion of males = 0.08 and 0.09, respectively) which
contrasts with a more even sex ratio for non-
social species in the genus.

In spiders it is relatively easy to determine the
gender of embryos using cytological techniques
because of their unusual sex determination
mechanism. In 81% of the spider species that
have been analyzed there are two sex chromo-
somes involved in sex-determination (White
1973). Males possess one copy of each of these,
while females cany' two copies of each (four in
all). Consequently, females have two more chro-
mosomes than males, and so can be distin-
guished simply on the basis of chromosome
number. This method of sex-determination is
referred to as XjXjO, and it would appear to be
the ancestral condition in spiders (White 1973).
This system has been modified to X.XjXjO in
the Australian huntsman spiders (Rowell 1985)
and X1X2X3X4O in the sparassid Hetewpoda sik-
kimensis (Datta & Chatterjee 1983). In the ma-
jority of the Oxyopidae analyzed, there has been
a reduction in sex chromosome number resulting
in an XO system similar to that prevalent in the
insects (Datta & Chatterjee 1983). Thus, among
different species, females may possess one, two,
three or four more chromosomes than males.

Over 300 species of spider have been analyzed

200

ROWELL & MAIN-SEX RATIO IN A SOCIAL SPIDER

201

Figure 1.— Male meiosis in D. socialis (2N = 24).
Note the presence of eleven bivalents plus two heter-
omorphic X-chromosomes (arrow).

cytologically, and the kind of sex-determination
system described above is virtually universal. In
only two cases, the sparassid Delena cancehdes
Walckenaer and the salticid genus Pellenes Simon,
do aberrant systems exist (Rowell 1990; Mad-
dison 1982), both involving rearrangements be-
tween the X-chromosomes and autosomes. These
modified systems are very distinctive, however,
and can easily be identified from male meiotic
preparations.

In this study, sex ratios of adult and subadult
D. socialis are compared to embryonic sex ratios
determined by cytological means, to ascertain
whether the observed bias arises before hatching,
or as a result of differential mortality of the sexes.
In addition, the ratios observed are compared
with those reported by Aviles and Maddison
(1991) from cytological analysis and Aviles (1986)
from morphology for the social species Anelo-

simus eximius, to determine whether a particular
primary sex ratio may exist in social spiders.

METHODS

All specimens, both eggs and males, used in
the cytological analyses were collected by one of
us (BYM) from sites in the vicinity of Torbay
west of Albany, southwestern Western Australia.
The sex ratio was also determined from adult
and subadult spiders from colonies collected in
the same areas by BYM (subsequently consid-
ered by Main (1988)) and from Pemberton by T.
A. Evans. In younger colonies at two sites a small
proportion of juvenile individuals possessed no
characters of either sex. No reliable morpholog-
ical characters have been found for sexing youn-
ger spiders, and the small size of D. socialis makes
dissection of large numbers of individuals im-
practical. However, the authors’ previous expe-
rience with this species has shown that the rate
of development within and between the sexes in
a colony is concerted (but with males maturing
before females). Hence swelling of the palp would
have been apparent were these individuals male.
Consequently, in colonies possessing a mixture
of adults and juveniles, the juveniles were scored
as females. The validity of this assumption was
borne out in the statistical analysis, where it was
shown that the sex ratios within these colonies
were not statistically different from those of adult
colonies. Colonies in which the majority of in-
dividuals were juvenile were not included in the
analysis, nor were degenerate colonies with less
than five individuals, or those without represen-
tatives of both sexes (two colonies qualified for
the latter).

Chromosome preparations were made from
testes and embryos using the techniques de-
scribed in Rowell (1985) which differs from that
detailed by Aviles and Maddison (1991). Not
every embryo yielded spreads which could be

Table 1.— Heterogeneity and deviation from even sex ratio in embryonic data for D. socialis. * P < 0.01.

Site

Males

Females

Total

Male

proportion

Coombes Rd, Torbay

5

13

18

0.28

Coombes/Harding Rds, Torbay

3

8

11

0.27

Coombes/Harding Rds, Torbay

4

12

16

0.25

Coombes/Harding Rds, Torbay

4

9

13

0.31

Total

Deviation from 1:1 G = 12.08*

16

Heterogeneity G =

42

0.122 ns.

58

0.28

202

THE JOURNAL OF ARACHNOLOGY

2

)> cue !! fO

ta M u H<t ••

^ X, X

1 2

H K n I*

la li II II ••

• •

Figures 2, 3. — Karyotypes from D. socialis embryos: 2, male (2N = 24); 3, female (2N = 26). Note the presence
of supernumerary microchromosomes in the two karyotypes.

counted with certainty, but given the marked
developmental synchrony within each egg sac,
those counted are held to represent a random
sample from each colony.

Data were analyzed for deviation from a sex
ratio of 1:1 and homogeneity among egg sacs,
nests and sites using the G-test of Sokal and Rohlf
(1981). This statistic has a chi-square distribu-
tion but is more reliable than the traditional chi-
square test when sample sizes are small.

The data obtained for D. socialis were com-
pared, by G-test, against the data given for Ane-
losimus eximius by Aviles and Maddison (1991)
and by Aviles (1986).

RESULTS

Chromosomal analysis. — Both meiotic and
mitotic preparations were successfully obtained
from testes. Mitotic preparations were examined
for 58 embryos from four egg sacs in all.

ROWELL & MAIN-SEX RATIO IN A SOCIAL SPIDER

203

Table 2. — Summary of data obtained from morphological analysis of adult and subadult D. socialis specimens.
Colony data, heterogeneity within sites, male proportion and number of colonies with a female bias. Treatment
of outliers discussed in text. * P < 0.001.

Site

No. of
colonies

Female bias

Signifi-
cant Present

Hetero-

geneity

G value

n

Male

propor-

tion

Pemberton 26/7-1/8 Areal

10

7

9

8.03 ns

214

0.23

Pemberton 26/7-1/8 Area2

9

8

9

8.17 ns

240

0.15

Puls Rd, Torbay 26/3/89

1

1

1

29

0.28

Puls Rd, Torbay 18/4/90

4

4

4

7.82 ns

104

0.17

Puls Rd, Torbay 23/8/87

1

1

1

34

0.17

Puls Rd, Torbay 14/10 & 5/1 1/90

8

5

6

43.55*

213

0.26

(without outliers)

5

4

5

3.536 ns

102

0.29

Curinup Rd 3/6/90

6

5

6

11.07 ns

89

0.12

Curinup Rd 14/10/90

2

0

0

0.07 ns

16

1.67

Total (pooled by site)

44

31

39

26.80*

939

(outliers removed)

39

30

38

12.35 ns

812

0.19

Figure 1 shows male meiosis, which involves
1 1 bivalents and the X-chromosomes. Of im-
portance here is the fact that two X-chromo-
somes are visible, indicating that D. socialis em-
ploys the XjXjO system common to most spider
species. This was reflected in the mitotic prep-
arations from embryos (Figs. 2, 3), where indi-
viduals possessed either 24 or 26 chromosomes.
Clearly, the former represent male embryos and
the latter females. In many of the preparations
a number of chromosome fragments or micro-
chromosomes (ranging from one to three) was
visible (Figs. 2, 3).

Table 1 shows the numbers ofmale and female
embryos identified for each egg sac analyzed. In
every case, the number of females exceeded the
number of males, and the results were statisti-
cally homogeneous among the egg sacs. The
pooled data deviate significantly from a 1 : 1 sex
ratio, with an overall male proportion of 0.28.

Morphology.— Table 2 shows a summary of
the data obtained from morphological analysis
of adult and subadult D. socialis specimens. The
ratios among colonies and sites were quite uni-
form, with only five deviant colonies (one from
Puls Road, 14 October 1990, two from the same
locality on 5 November 1990 and both of the
Curinup Road, 1 4 October 1 990, colonies). Since
these colonies were all collected in mid-October
or later, which is towards the end of the dispersal
phase, it is probable that they represent the rem-
nants of larger colonies following emigration of
many of the spiders. Moreover, two of the col-

onies were unusual in that they possessed resi-
dent clubionid predators. Consequently, the sex
ratios in these collections are considered to be
non-representative of the adult sex ratio of the
population as a whole, and these colonies were
removed from further analyses.

In total, 38 of the remaining 39 colonies showed
a female bias, and in 33 of these the bias was
statistically significant. The overall sex ratio de-
termined from the morphology was not signifi-
cantly different from the overall embryonic ratio
(Table 3).

Comparison with Anelosimus eximius. — Habit
3 shows the results of comparisons of embryonic
and morphological data between D. socialis and
A. eximius from Aviles (1986) and Aviles and
Maddison (1991). For both species the embry-
onic ratios do not differ significantly from those
of adult and subadult spiders, but a greater bias
is apparent in the morphological data. For both
types of data, however, the female bias is sig-
nificantly greater in A. eximius.

DISCUSSION

Karyotype of Diaea socialis. — Tht chromo-
some number and sex determination mechanism
of D. socialis differ from those reported for D.
subadulta Rosenberg & Strand, the only other
species of the genus which has been examined
karyotypically (Suzuki 1952). D. subadulta pos-
sesses 26 autosomes and one X-chromosome in
the male and consequently a complement of 26
autosomes and two X-chromosomes in the fe-

204

THE JOURNAL OF ARACHNOLOGY

male can be inferred. Within D. socialis, how-
ever, the chromosome number and karyotype
form are conserved, with identical male com-
plements present in populations from Victoria
and the Australian Capital Territory, separated
from the Western Australian study site by dis-
tances of over 3000 km, with large areas of in-
tervening desert (unpublished data). This con-
trasts with cytological data from two other wide-
ranging spider species, Delena cancerides and Se-
lenops australiensis Koch which show marked
geographical variation in chromosomal number
and karyotypic form across their range (Rowell
1990, DMR unpublished).

Sex ratio in D. socialis and A. eximius.— The
embryonic sex ratio bias in D. socialis is consis-
tent with the data from adult and subadult spi-
ders. Thus it can be concluded that, as in A.
eximius, the bias is present before hatching.

The tests for heterogeneity indicate that there
is a marked conservatism in the sex ratio of D.
socialis both between sites and through time, as
the data are derived from observations over four
years. Moreover, there does not appear to be any
relationship between month of collection and sex
ratio, except during the period of dispersal, over
the October/November period. This is an im-
portant observation because such uniformity may
not be expected if the ratio is influenced by some
environmental factor. For example, if the sur-
vival of male embryos or the success of male
determining sperm were influenced by temper-
ature, considerable variation in ratio would be
expected between years, or even within one sea-
son.

The mechanism involved in producing the fe-
male bias observed is not apparent; however,
some conclusions can be drawn. As in A. eximius
(Aviles & Maddison 1991), few infertile D. so-
cialis eggs were observed, and certainly insuffi-
cient to make up the shortfall in males even if it
were assumed that every one represented a dead
male embryo. The little “infertility” that is ap-
parent in this species is usually associated with
egg parasitism (BYM unpublished). In A. exi-
mius and D. socialis, meiosis proceeds normally
through to anaphase II, and so it is possible to
narrow down the time at which the mechanism
acts in both species to the period between the
end of meiosis and fertilization. Since sex is de-
termined by the chromosomal composition of
sperm, it follows that male-determining sperm
are either present in lower numbers than female-

Table 3.— G-test comparisons involving morpho-
logical and chromosomal data sets from A. eximius
(Aviles 1986, Aviles & Maddison 1991)andZ). socialis.
*{P < 0.01), ** (P < 0.001).

Analysis

G Value

A. eximius, heterogeneity among

colonies, morphology

24.108 ns

A. eximius, embryos x morphology

0.078 ns

D. socialis, embryos x morphology

2.602 ns

A. eximius embryos x D. socialis

embryos

10.454*

A. eximius morphology x D. socialis

morphology

24.496**

determining sperm, or they compete less well for
fertilization of the egg.

To gain any finer resolution would require the
use of techniques capable of distinguishing be-
tween male and female sperm, such as sedimen-
tation rate or DNA content. Alternatively, if an
X-linked probe were available, it might be pos-
sible to determine the ratio of the marker to total
DNA in males, females and a sperm sample us-
ing a technique as simple as dot blotting. If male-
and female-determining sperm are present in
equal proportions, the ratio of the marker to total
DNA in male tissue would equal the ratio in the
sperm sample, and be half that found in female
tissue (because females possess twice as many X
chromosomes as males). If, however, the bias
arises from a deficiency of male-determining
sperm, X chromosomes would be present in more
than half of the sperm; and the ratio would be
intermediate between that of male and female
tissue.

Hamilton (1967) pointed out that Fisher’s
principle is based on the assumption of panmix-
ia, and that selection on groups where inbreeding
within a single clutch is the norm may favor the
production of fewer males, thus minimizing
competition between male offspring and the con-
sequent wastage of reproductive effort. This
would appear to be a probable explanation here
since the life histories of both D. socialis and A.
eximius suggest marked inbreeding. This is fur-
ther supported by preliminary data from the so-
cial spider Delena cancerides. This species has
been demonstrated to be outbred by the use of
enzyme electrophoresis (Rowell 1990), and on
the basis of dissection of hatchlings it would ap-
pear to have an even sex ratio (Rowell, unpub-
lished).

ROWELL & MAIN-SEX RATIO IN A SOCIAL SPIDER

205

An alternative explanation discussed in detail
by Aviles (in press) is that, in continuously in-
breeding lineages of social species where colonies
are initiated by individual mated females, selec-
tion at the colony level plays a major role. Future
success of a given lineage is a function of the
number of successful females rather than the ab-
solute number of offspring. Consistent with this
is that in the solitary leaf-rolling thomisids of the
genus Cymbacha L. Koch, which presumably
have an even sex ratio, dispersal occurs during
the juvenile stage; and, as in many species, there
is likely to be a high mortality in the free-living
stage of the life cycle in both sexes. Moreover,
the chances of free-living females finding a mate
rely on a relatively high density of males. In con-
trast, because only adult females of D. socialis
(which are assured of a mate within the colony)
disperse, the biased ratio increases the proba-
bility of a lineage surviving to future generations.
Main (1988) noted that only one in five incipient
nests survived to maturity, and this does not take
into account mortality during emigration prior
to leaf-rolling.

Aviles’ model is based on the hypothesis of
Colwell (1981) which has received some criti-
cism (see, for example, Charlesworth & Toro
1982). Nevertheless the life history of D. socialis
appears to follow the parameters of the model
very closely. Indeed, further study of D. socialis
may be of particular value in establishing the
validity of this model.

It has long been known that a female bias exists
in the eusocial Hymenoptera (Hamilton 1967),
and in this context it is interesting to note that
the presence of nonreproductive females has been
demonstrated convincingly only once in spiders
(Vollrath 1986). However, a behavioral parallel
with true eusociality was also recognized in D.
socialis, where it appeared that some females
forfeited reproduction while still working to the
benefit of the colony (Main 1988).

The phenomenon of skewed sex ratio in social
spiders is particularly interesting because social
behavior has arisen independently in a number
of spider families (Burgess 1978). This would
imply that either selection for a bias is very strong
in social spiders, or that such an alteration in sex
ratio is easily achieved. While the mechanism
by which this is achieved is obscure, it is probable
that it is the same in A. eximius, A. domingo and
D. socialis, because they act at a very precise time
in the three species. Moreover, given the ob-
served plasticity in sex ratio among the species.

the difference in sex ratio between D. socialis and
A. eximius may imply a difference in the optimal
ratio, determined by their life histories. Thus a
detailed comparison of the lifestyles of these spe-
cies may shed light on the particular factors that
influence sex ratio.

ACKNOWLEDGMENTS

We thank T. A. Evans for collection and counts
of specimens from the Manjimup region. We
would also like to thank Ian Scott and Angela
Higgins for their help with data analysis and
preparation of the manuscript. Leticia Aviles of
the University of Arizona provided particularly
helpful comments on the manuscript. The Zo-
ology Department of the University of Western
Australia provided some of the facilities neces-
sary for this work.

LITERATURE CITED

Aviles, L. 1986. Sex ratio bias and possible group
selection in the social spider Anelosimus eximius.
American Nat., 128:1-12.

Aviles, L. in press Interdemic selection and sex ratio:
a social spider perspective. American Nat., 000:000-
000.

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). J. Arachnol., 19:126-135.
Burgess, J.W. 1978. Social behaviour in group-living
spider species. Symp. Zool. Soc. London, 42:69-78.
Buskirk, R. E. 1981. 4. Sociality in the Arachnida.
Pp. 281-367, In Social Insects II (H. R. Hermann,
ed.). Academic Press, Inc., San Diego, California.
Charlesworth, B. & M. A. Toro. 1982. Female biased
sex ratios. Nature, 298:494.

Colwell, R. K. 1981. Group selection is implicated
in the evolution of female-biased sex ratios. Nature,
290:401-404.

Datta, S. N. & K. Chatterjee. 1983. Chromosome
number and sex-determining mechanism in fifty-
two species of spiders from north-east India. Chrom.
Inf Serv., 35:6-8.

Fisher, R. A. 1930. The genetical theory of natural
selection. Dover, New York.

Hamilton, W. D. 1967. Extraordinary sex ratios. Sci-
ence, 156:477-488.

Maddison, W. P. 1982. XXXY sex chromosomes in
males of the jumping spider genus Pellenes (Ara-
neae: Salticidae). Chromosoma, 85:23-37.

Main, B. Y. 1988. The biology of a social thomisid
spider. Pp. 55-74, In Australian Arachnology. (A.
D. Austin & N. W. Heather, eds.) The Australian
Entomological Society, Misc. Pub. No. 5.

Rowell, D. M. 1985. Complex sex linked fusion het-
erozygosity in the Australian huntsman spider De-

206

THE JOURNAL OF ARACHNOLOGY

lena cancerides (Araneae: Sparassidae). Chromo-
soma, 93:169-176.

Rowell, D. M. 1 990. Fusion heterozygosity in Delena
cancerides Walck. (Araneae: Sparassidae): an alter-
native to speciation by monobrachial fusion. Ge-
netica, 80:139-157.

Sokal, R. R. & F. J. Rohlf. 1981. Biometry. 2nd ed..
W. H. Freeman & Co. San Francisco.

Suzuki, S. 1952. Cytological studies in spiders. II.
Chromosomal investigation in the 22 species of spi-
ders belonging to four families, Clubionidae, Spar-
assidae, Thomisidae and Oxyopidae, which consti-

tute Clubionidea, with special reference to sex
chromosomes. J. Sci. (Hiroshima University) Ser.
B, Div. 1 : Zoology, 15:23-136.

Vollrath, F. 1986. Eusociality and extraordinary sex
ratios in the spider Anelosimus eximius (Araneae:
Theridiidae). Behav. Ecol. Sociobiol., 18:283-287.

White, M. J. D. 1973. Animal cytology and evolu-
tion. Cambridge University Press. Cambridge, En-
gland.

Manuscript received 1 April 1992, revised 27 July 1 992.

1992. The Journal of Arachnology 20:207-21 1

SURVIVORSHIP OF WOLF SPIDERS (LYCOSIDAE)
REARED ON DIFFERENT DIETS

George W. Uetz, Jennifer Bischoff, and Joseph Raver: Department of Biological
Sciences, University of Cincinnati, Cincinnati, Ohio 45221-0006 USA

ABSTRACT. Observations from previous studies have indicated that lycosid spiders often die before maturing
when raised on only one prey type. Two wolf spider species (Lycosa helluo collected from Florida, and Lycosa
sp. collected from Kentucky) were used to test the hypothesis that diet affects survivorship. Siblings from one
egg sac of each species were divided into two groups of 50 spiderlings each, and reared under identical conditions
with different diets. The polytypic diet consisted of crickets {Acheta domesticus), fly grubs (Sarcophaga bullata),
cockroaches (Periplaneta americana), mealworms (Tenebrio molitor), beetles (Dermestes sp.), and an occasional
supplemental orthopteran collected from the field. The monotypic diet consisted only of crickets (A. domesticus).
There was significantly lower survivorship of spiders raised on monotypic prey in both species, although the
pattern of mortality over time varied between species. There were also significant differences in certain body
size parameters (cephalothorax width, total leg I length, patella-tibia length) measured at maturity between
spiders raised on polytypic or monotypic diets in one species (L. helluo). In addition, Lycosa helluo raised on
polytypic diets reached sexual maturity earlier than those reared on monotypic prey. These results suggest that
there are fitness-related consequences of dietary breadth in spiders, and support the hypothesis of Greenstone
(1979) that lycosids require a mixed diet.

INTRODUCTION

The dietary breadth of generalist predators,
including spiders, is usually thought to indicate
a strategy of opportunistic prey capture (Riechert
& Luczak 1 982, Riechert & Lockley 1 984, Riech-
ert & Harp 1987; Uetz 1990, 1992). However,
there is growing evidence that many animals (es-
pecially herbivores) maintain a mixed diet for
nutritional reasons (Belovsky 1978; Slansky &
Rodriguez 1987). Greenstone (1978) found that
lycosid spiders {Pardosa ramulosa (McCook)) did
not switch to more abundant or profitable prey
items as prey density changed. Optimization of
critical nutritional requirements (amino acids,
fatty acids, etc.) may be why lycosids maintain
a mixed diet in the field despite high abundance
of single prey species (Greenstone 1979).

Observations from a number of previous stud-
ies have indicated that lycosid spiders often die
before maturing when raised on a diet composed
of only one prey type (Miyashita 1 968; Van Dyke
& Lowrie 1975; C. D. Dondale, J. S. Rovner,
pers. comm.). This may be true for other spider
families as well (e. g., Agelenidae - Riechert &
Harp 1987; Linyphiidae - D. H. Wise, pers.
comm.). In particular, juveniles raised on a diet
of Drosophila melanogaster do not survive past
the 4th or 5th instar (Van Dyke & Lowrie 1975;

K. Redborg, pers. comm.), suggesting the pos-
sible absence of critical nutrients in this prey
species. In contrast, reduced survivorship was
not seen in rearing studies with lycosids and oth-
er spider families where a variety of insect species
were available as prey (Eason 1969; Peck &
Whitcomb 1970). In this study, we tested the
influence of diet on survival and development
of lycosids through adulthood by rearing spiders
under identical controlled conditions, but feed-
ing them on polytypic and monotypic diets.

METHODS

Two species of lycosid spiders were used in
this study: Lycosa helluo Walckenaer, collected
from Highlands Hammock State Forest in High-
lands County, Florida, and Lycosa sp. (possibly
an undescribed member of the L. helluo group),
collected from a power line right-of-way along
the Licking River in Kenton County, Kentucky.
Females of each of these species, carrying egg
sacs, were collected and brought into the labo-
ratory.

After emergence of spiderlings, and dispersal
from the female’s abdomen, 100 siblings from
each species were separated from the female and
assigned at random to experimental groups of 50
spiderlings each. Spiderlings were reared at first

207

208

THE JOURNAL OF ARACHNOLOGY

(a) Lycosa helluo

(b) Lycosa sp.

I ' MONOTYPIC DIET POLYTYPIC DIET |

Figure 1. — Survivorship (in days) of laboratory pop-
ulations of lycosids raised on different diets. Solid line-
polytypic diet; dashed line— monotypic diet. Arrows
indicate approximate earliest onset of sexual maturity.
a: Lycosa helluo from Florida; b: Lycosa sp. from Ken-
tucky.

in 10 cm long, 1 cm diameter glass tubes with
cotton plugs at each end, then transferred to semi-
clear cylindrical plastic containers (11.8 cm high,
15.3 cm diameter). Rearing conditions for spi-
ders were identical except for diet: 1 2: 1 2 hr light/
dark cycle, constant 27 °C temperature, relative
humidity 65-75%. Water was available ad libi-
tum in rearing containers from soaked cotton
plugs and/or small shell vials with water and a
cotton plug at one end.

Insect prey (three to five individuals selected
for size approximately ten percent less than spi-
der size) were provided twice weekly. The pol-
ytypic diet consisted of crickets {Acheta domes-
ticus (L.)), flesh fly grubs {Sarcophaga bullata
Park), cockroaches {Periplaneta americana (L.)),
mealworms {Tenebrio molitor L.), beetles {Der-
mestes spp.) and occasional orthopterans (Tet-
tigoniidae, Acrididae) collected from the field.

Lycosa helluo

-e- POLYPi'PIC DIET ♦ MONOTYPIC DIET

Figure 2. — Cumulative (percent) maturation curves
for laboratory populations of Lycosa helluo from Flor-
ida raised on different diets. Open squares/solid line-
polytypic diet; filled squares/dashed line— monotypic
diet.

The monotypic diet consisted only of crickets
{Acheta domesticus).

Spider survival was monitored with each feed-
ing, or at least once weekly, throughout devel-
opment until sexual maturity. Molts were re-
corded when exuviae were observed in the
container. At approximately four weeks after
reaching sexual maturity, a subsample of surviv-
ing female spiders from each experimental group
was weighed and measured. Body size parame-
ters measured included body length (BL), ceph-
alothorax width (CW), abdomen width (AW),
total leg length (TLL) of one leg I (chosen at
random), patella-tibia length (Pa-Ti) of that same
leg, and live weight.

RESULTS

In both species, differences in survivorship were
apparent between experimental groups (Figure
la, b). These differences between monotypic and
polytypic prey treatments are significant (Kol-
mogorov-Smimov test: values = 0.41 (L.

helluo)-, 0.37 {Lycosa sp.); P < 0.001 for both).
Patterns of survivorship varied between species,
as mortality occurred at different rates and at
different times of the life cycle. In L. helluo, dif-
ferential mortality was apparent in the first 30
days (Fig. la). For L. helluo fed a polytypic diet,
approximately 96% of spiders survived past 250
days, and about 85-90% reached sexual matu-
rity. The mortality rate (no. dying/no. alive at
start X 100) for this treatment was 12.2%. For

UETZ ET AL- SURVIVORSHIP OF WOLF SPIDERS ON DIFFERENT DIETS

209

Table 1.— Body size measurements of adult female wolf spiders fed on different diets.

Body size parameter

BL

cw

AW

TLL

Pa-Ti

Weight

(mm)

(mm)

(mm)

(mm)

(mm)

(g)

Lycosa helluo

Monotypic diet:

Mean

27.09

9.64

10.27

30.69

10.85

1.73

SD

1.32

0.84

1.27

2.67

0.85

0.22

n

18

18

18

17

17

17

Polytypic diet:

Mean

27.80

10.26

10.21

32.87

11.67

1.79

SD

1.34

0.79

1.15

2.69

0.89

0.22

n

13

14

15

12

15

12

/-test

1.42

2.02

0.14

2.07

2.58

0.71

P value

ns

<0.05

ns

<0.05

<0.05

ns

Lycosa sp.

Monotypic diet:

Mean

20.32

8.00

7.89

23.05

9.41

0.69

SD

0.75

0.36

0.51

1.95

0.81

0.09

n

6

5

6

6

6

6

Polytypic diet:

Mean

21.07

7.68

8.33

22.04

8.81

0.81

SD

1.51

0.41

0.57

2.06

0.77

0.14

n

11

9

11

11

11

11

/-test

1.07

1.46

1.49

0.92

1.39

1.7

P value

ns

ns

ns

ns

ns

ns

L. helluo fed a monotypic diet, survivorship was
fairly constant after the initial decline, but only
75% readied sexual maturity (the mortality rate
for this treatment was 28%). These results con-
trast sharply with survivorship patterns of Ly-
cosa sp. (Fig. lb); both treatment groups for this
species experienced a 25% decline in survivor-
ship in the first 30 days. Survivorship curves for
Lycosa sp. diverged after approximately 80 days,
with only 25% of the spiders reared with a mono-
typic diet surviving to adulthood. The mortality
rate for the monotypic treatment was 75.5%. In
contrast, >70% of those reared on a polytypic
diet survived to adulthood (a mortality rate of
28.5%).

Owing to difficulty in observing female geni-
talia on live specimens, precise data on age at
sexual maturity were only available for L. helluo
(Fig. 2). Distributions of maturation times were
significantly different for the two diet treatments
(Kolmogorov-Smimov test: = 0.37; P <

0.05). Spiders raised on a polytypic diet reached

sexual maturity earlier than those raised on a
monotypic diet; differences in age at maturity are
significant (Mann- Whitney U-test: U, = 482; P
< 0.05). The median age at maturity for poly-
typic diet treatment individuals was estimated
to be 337 days; for monotypic diet treatment
individuals, 387 days (precise dates could not be
determined in all cases, as spiders were moni-
tored on a weekly basis).

Body size parameters measured at maturity
were significantly different between experimental
treatments for one of the two species studied
(Table I). For L. helluo, significant differences
between treatments were seen in cephalothorax
width (T = 2.02; P < 0.05), total leg length (T
= 2.07; P < 0.05); and patella-tibia length (T =
2.58; F < 0.05), with spiders raised on polytypic
diets larger in all measures. In contrast, for Ly-
cosa sp., no significant differences in body size
parameters v/ere seen (although lowered survival
and consequent smaller sample sizes may have
influenced this result).

210

THE JOURNAL OF ARACHNOLOGY

DISCUSSION

Results of this study provide strong support
for earlier hypotheses regarding the importance
of a mixed diet for lycosid spiders and other
spider species (Peck & Whitcomb 1970; Green-
stone 1979; Riechert & Harp 1987). It is im-
portant to note that the monotypic diet used in
this study was composed of a domesticated prey
animal, which was itself reared under artificial
conditions with an unknown diet. It is possible
that the diet fed to crickets in culture may have
lacked a nutrient requirement critical for spiders
(although not for crickets), and an experimental
study with a monotypic diet of field-collected
crickets might have yielded a different result.
Walcott (1963) reported that Achaearanea tepi-
dariorum (Koch) had poor survivorship when
fed mealworms whose diet was limited to stan-
dard mealworm bran. However, when meal-
worms were fed vitamin-enriched commercial
bran cereals, spider survivorship was improved
dramatically. This result may also explain why
lycosids suffer high mortality when fed on a diet
of Drosophila, an insect known to lack a require-
ment for linoleic and linolenic acid in its diet (K.
Redborg, pers. comm.). While these findings
clearly suggest that rearing of spiders in the lab-
oratory could be enhanced by providing a variety
of prey species, it also raises questions about the
role of dietary mixing in the field.

The differences in survivorship, age at matu-
ration, and size at maturity seen in this study
between spiders fed on a polytypic versus a
monotypic diet suggest that there are clear fitness
consequences of dietary breadth. Differential
mortality rates, with approximately 2. 3-2. 6 times
greater mortality for spiders fed a monotypic diet,
suggest that selection pressure for dietary mixing
could be strong. Although in this study ail other
factors were controlled, the lack of dietary mix-
ing might affect spiders in the field in other ways
as well. For example, if physical condition were
affected by dietary breadth, differences in vul-
nerability to predation and parasitism might re-
sult. Moreover, earlier maturation and larger size
at maturity might well confer other fitness ad-
vantages on spiders with mixed diets (Uetz 1 992).
It is well known that larger spiders are more likely
to win contests over territory and or mates (Aus-
tad 1983; Christenson 1984; Suter&Keiley 1984;
Riechert 1 986; Uetz & Hodge 1 990). Spiders ma-
turing early in the breeding season might have
access to more potential mates, and have a longer

time to feed before laying eggs. In addition, off-
spring of spiders breeding earlier might have a
competitive size advantage over other broods,
and might even cannibalize them (Edgar 1969).

This demonstration of differential mortality
and other fitness-related consequences of diet
provides strong support for the hypothesis that
dietary mixing is adaptive in spiders (Greenstone
1979). While the proximate mechanisms by
which spiders maintain a mixed diet remain un-
clear, there is new evidence that foraging behav-
iors affecting diet choice in spiders are geneti-
cally-based (Hedrick & Riechert 1989; Riechert
1991), and therefore potentially subject to selec-
tion. As spiders are considered important model
organisms for research in ecology and behavioral
genetics, as well as potential agents of agricultural
pest management (Wise 1984; Riechert & Lock-
ley 1984; Uetz 1992), further study of the adap-
tive significance of diet in these animals deserves
attention.

ACKNOWLEDGMENTS

This study was supported in part by funding
from the University of Cincinnati Research
Council, the Department of Biological Sciences,
and the Arachnological Research Fund. We ap-
preciate the assistance of Sam Marshall, who col-
lected the Florida specimens, and Allen Brady
and Pat Miller, who kindly verified the species.
We are also grateful to Cara Hardesty, Nilay Pa-
tel, Michelle Gingery, Alan Hensley, and James
McDonough for assistance in rearing spiders, and
Dave Clark and Veronica Casebolt for various
other assistance. Gail Stratton, A1 Cady, Jerry
Rovner and Matt Greenstone provided insight-
ful comments on the manuscript.

LITERATURE CITED

Austad, S. N. 1983. A game theoretical interpretation
of male combat in the bowl and doily spider (Fron-
tinella pyramitela). Anim. Behav., 31:59-73.
Belovsky, G. E. 1978. Diet optimization in a gen-
eralist herbivore: the moose. Theor. Popul. Biol.,
14:105-134.

Christenson, T. E. 1984. Alternative reproductive
tactics in spiders. American ZooL, 24:321-332.
Eason, R. R. 1 969. Life history and behavior of Par-
dosa lapidicina Emerton (Araneae: Lycosidae). J.
Kansas EntomoL Soc., 42:339-360.

Edgar, W. D. 1969. Prey and predators of the wolf
spider Lycosa lugubris. J. Zool. London, 1 59:405-
411.

Greenstone, M. H. 1978. The numerical response to
prey availability of Pardosa rarnulosa (McCook)

UETZ ET AL.- SURVIVORSHIP OF WOLF SPIDERS ON DIFFERENT DIETS

211

(Araneae: Lycosidae) and its relationship to the role
of spiders in the balance of nature. Symp. Zool. Soc.
London, 42:183-193.

Greenstone, M. H. 1979. Spider behavior optimises
dietary essential amino acid composition. Nature,
282:501-503.

Hedrick, A. & S. E. Riechert. 1989. Genetically-based
variation between two spider populations in for-
aging behavior. Oecologia, 80:533-539.

Miyashita, K. 1968. Growth and development of Ly-
cos^ t-insignata Boes. et Str. (Araneae: Lycosidae)
under different feeding conditions. Appl. Ent. Zool.,
3: 81-88.

Peck, W. B. & W. H. Whitcomb. 1970. Studies on
the biology of a spider Chiracanthium indusum
(Hentz). Agr. Exp. Sta. Div. Agr. Univ. Arkansas
Bull, 753:1-76.

Riechert, S. E. 1986. Spider fights: a test of evolu-
tionary game theory. American Sci., 74:604-610.

Riechert, S. E. 1991. Prey abundance vs. diet breadth
in a spider test system. Evol. EcoL, 5:327-338.

Riechert, S. E., & J. M. Harp. 1 987. Nutritional ecol-
ogy of spiders. Pp. 645-672, In Nutritional Ecology
of Insects, Mites, and Spiders (F. Slansky & J. G.
Rodriguez, eds.). John Wiley & Sons, New York

Riechert, S. E. & T. Lockley. 1984. Spiders as bio-
logical control agents. Ann. Rev. Entomoi., 29:299-
320.

Riechert, S. E. & J. Luczak. 1982. Spider foraging:
behavioral responses to prey. Pp. 353-384, In Spi-
der Communication: Mechanisms and Ecological
Significance (P. N. Witt & J. S. Rovner, eds.).
Princeton Univ. Press, Princeton, New Jersey.

Slansky, F. & J. G. Rodriguez. 1987. Nutritional

Ecology of Insects, Mites, and Spiders. John Wiley
& Sons, New York.

Suter, R. B. & M. Keiley. 1984. Agonistic interactions
between male Frontinella pyramitela (Araneae: Lin-
yphiidae). Behav. Ecol. SociobioL, 15:1-7.

Uetz, G. W. 1990. Prey selection in web-building
spiders and evolution of prey defenses. Pp. 93-128,
In Insect Defenses: Adaptive Mechanisms and
Strategies of Prey and Predators (D. L. Evans & J.
O. Schmidt, eds.). State Univ. of New York Press,
Albany.

Uetz, G. W. 1992. Foraging strategies of spiders.
Trends Ecol. Evol., 7:155-159.

Uetz, G. W. & M. A. Hodge. 1990. Influence of hab-
itat and prey availability on spatial organization and
behavior of colonial web-building spiders. Nat.
Geogr. Res., 6:22-40.

Van Dyke, D. & D. C. Lowrie. 1975. Comparative
life histories of the wolf spiders Pardosa ramulosa
and P. sierra (Araneae: Lycosidae). Southwest Na-
tur., 20:29-44.

Walcott, C. 1963. The effect of the web on vibration
sensitivity in the spider, Achaearanea tepidariorum
(Koch). J. Exp. BioL, 40:595-61 1.

Wise, D. H. 1984. The role of competition in spider
communities: Insights from field experiments with
a model organism. Pp. 42-54, In Ecological Com-
munities: Conceptual Issues and the Evidence (D.
R. Strong, S. S. Simberloff, L. G. Abele, and A.
Thistle, eds.). Princeton Univ. Press, Princeton.

Manuscript received 8 June 1992, revised 16 August
1992.

1992. The Journal of Arachnology 20:212-216

ANTIPREDATOR BENEFITS OF SINGLE- AND MIXED-SPECIES
GROUPING BY NEPHILA CLAVIPES (h.)

(ARANEAE, TETRAGNATHIDAE)

Margaret A. Hodge' and George W. Uetz: Dept, of Biological Sciences, University of
Cincinnati; Cincinnati, Ohio 45221-0006 USA

ABSTRACT. The golden silk spider, Nephila clavipes (L.), is known to live both solitarily and in single-species
aggregations. In Veracruz, Mexico, N. clavipes is also found in association with the colonial orbwsMev Metepeira
incrassata F.O. Pickard-Cambridge (Araneae: Araneidae). This study compared the frequency of predation
attempts on solitary, intraspecifically aggregated and colony associated N. clavipes. Solitary N. clavipes suffered
greater relative predation than those in single-species groups or those associated with M. incrassata colonies.
We also compared the distance at which the three categories of N. clavipes were able to detect and respond to
a simulated predation attempt. Both intraspecifically grouped and colony associated N. clavipes had significantly
greater response distances than did solitary individuals, indicating that they could respond to a predation threat
sooner. These data support predictions that grouped spiders may benefit from lower predation and/or an early
warning system.

Although spiders, as predators, have been well
studied, evidence for the types and especially the
frequencies of predation on spiders is lacking.
Predation rates on solitary spiders may be quite
high (Askenmo et al. 1977; Gunnarsson 1983),
especially in tropical areas (Rypstra 1984; Voll-
rath 1985). There is evidence that spider webs
are irritating to vertebrate predators such as birds
(Horton 1980; Eisner & Nowicki 1983), and it
has been suggested that the dense webbing of
colonial spiders might act to deter predation (Lu-
bin 1974; Robinson & Robinson 1976; Rypstra
1979). It has also been hypothesized that colonial
spiders may benefit from an “early-warning sys-
tem”, in which the intertwined webbing trans-
mits alarm signals from other spiders or vibra-
tions from predators, causing spiders to take
evasive action (Lubin 1974; Rypstra 1979; Uetz
1985). To date, however, actual evidence for an-
tipredator benefits in colonial spiders is limited
(Spiller & Schoener 1989).

There are several ways that animals in groups
can avoid predation: shared vigilance and early
warning signals may increase the chances of
predator detection (Pulliam & Caraco 1984); re-
duction of individual risk as a result of being one
of many possible prey present (Hamilton 1971);

'Dept, of Biology, The College of Wooster, Wooster,
Ohio 44691 USA

predator deterrence arising from collective mim-
icry or aposematism (Morse 1980); group de-
fense, such as mobbing. While the above ideas
were developed with regard to single-species
grouping, reduction of predation frequency has
also been reported in a variety of mixed-species
groups (reviewed in Morse 1977 and Barnard &
Thompson 1985). However, it is also possible
that individuals in mixed-species groups might
experience a unique disadvantage with respect
to predation if they are different in appearance
from the majority of the group. There is evidence
that odd group members are selectively preyed
upon in heterospeciflc groups (Mueller 1975;
Milinski 1977; Ohguchi 1978).

The goal of this study was to determine wheth-
er predator avoidance benefits exist for spiders
in heterospeciflc groups. Mixed-species associa-
tions have been noted to occur between colonial
web-building spiders and other spider species
(Lubin 1974; Sabath et al.l974; Bradoo 1972,
1979; Rypstra 1979; Berry 1987; Lopez 1988;
Hodge 1990). Spiders in the genus Nephila (Te-
tragnathidae) often form intraspeciflc aggrega-
tions (Shear 1970; Moore 1977; Farr 1977; Ryp-
stra 1985; Higgins 1988). Other species of web-
building spiders associate with Nephila groups
(McCook 1890; Strusaker 1969; Yoshida 1988)
and Nephila have been found in association with
colonial orbwea vers (Jackson 1986; Hodge 1990).

212

HODGE & UETZ- ANTIPREDATOR BENEFITS OF GROUPING BY NEPHILA

213

Because they are found solitarily, in single-spe-
cies and in mixed-species groups, Nephila are
ideal for comparing of the costs and benefits of
these three different living situations.

METHODS

We studied predation on a natural population
of female Nephila davipes (L.) (Araneae: Tetrag-
nathidae) on the grounds of the Hotel Posada
Loma, Fortin de las Flores, Veracruz, Mexico,
where they exist solitarily, in single-species groups
and in association with a colonial orb-weaving
spider, Metepeira incrassata F.O. Pickard-Cam-
bridge (Araneae: Araneidae) (Hodge 1990). The
habitat is classified as high semi-evergreen selva
(Gomez-Pompa 1977) and the study site con-
sisted of approximately 1.5 hectares of unen-
closed botanical gardens. Evidence of possible
predation was recorded during daily prey capture
observations (Hodge 1990) performed between
August 18-September 9, 1988. During these ob-
servations, groups and individuals on the grounds
were inspected during the morning, and again
before dark. A predation attempt was assumed
to have occurred if an individual was present on
her web between approximately 0600 h and 1 200
h, but was absent from the web and the area
immediately surrounding the web before night-
fall, and evidence of predation was found. Such
evidence consisted of the presence of a large hole
in the orb-web, which is a reliable sign of pre-
dation attempts by birds (Higgins, in press). It is
possible that such damaged webs could also be
caused by large insects impacting the web and
then escaping. However, in such cases the spider
usually rebuilds all or part of the web rather than
relocating. When Nephila voluntarily relocate,
they usually do so during the night, after ingesting
the orb web (Horton 1982; Higgins, in press;
Hodge, pers. obs.).

The average population size of each of the three
categories of N. davipes was estimated from three
different censuses performed during the study
period (Table 1). The number of N. davipes in
each category in the area of the predator attack
observations was recorded during each census.
Relative predation on each category of N. da-
vipes was estimated by calculating the percentage
which disappeared due to apparent predation.
The mean group size of intraspecific groups in
the population during the study period was 3.76
± 2.34 (n = 20 groups).

To test the “early-warning” hypothesis, pre-

Table 1.— Estimated relative predation attempts on
solitary, single-species and mixed-species associated N.
davipes between August 18 and September 9, 1988
(estimated population size = mean number of indi-
viduals in each category based on three censuses ±
standard deviation).

Estimated

Evidence of

population

predation

size

attempts

Solitary individuals

36 ± 2.08

12(33%)

Intraspecific groups

77 ± 2.52

8 (10%)

Mixed-species groups

43 ± 5.13

4 (9%)

dation attempts were simulated by disturbing
webs of arbitrarily chosen individuals by tapping
with a stick (the “pencil-poke” method: Tolbert
1975). Tapping was initiated as far from the cen-
ter of the orb as possible, usually where the web
was attached to vegetation. If no response was
elicited, the disturbance was moved toward the
orb at 1 cm intervals until a response was elicited,
or until the center of the orb (where the spider
sits) was reached. The distance (cm) from the
disturbance at which each spider performed an
antipredator response was recorded. Behaviors
scored as antipredator responses were: drop from
web, run off web or shake web. All of the web
disturbances were performed on a single day
(September 12, 1988) to control for possible ef-
fects of differences in temperature or other en-
vironmental variables on spider response. The
conditions were cloudy, so no spiders were under
heat stress, and the ambient temperature was 24
°C. The mean size of intraspecific groups in-
volved in these manipulations was 3.28 ± 1.98
{n = 1 groups). The size of the Metepeira in-
crassata colonies ranged from approximately
500-1000 spiders.

RESULTS

Solitary N. davipes experienced a greater per-
centage of predation attempts than did single-or
mixed-species groups, which had similar levels
(Table 1; G-test, P < 0.01). Response distances
of solitary N. davipes subject to the mock-pre-
dation disturbance were significantly less than
those of individuals in single- or mixed-species
groups (Table 2; Kruskal- Wallis, P < 0.05). The
primary difference was between grouped webs
and solitary webs {P < 0.05; non-parametric
multiple comparison procedure, Zar 1 984); there
was no difference between response distances of

214

THE JOURNAL OF ARACHNOLOGY

Table 2.— Comparison of distances at which re-
sponse was elicited to simulated predation by solitary,
single-species and mixed-species associated N. clavipes
(response distance = mean ± standard deviation; n =
number of individuals in each category).

Response

distance (cm)

n

Solitary individuals

21.5 ± 26.7

20

Intraspecific groups

41.9 ± 25.3

26

Mixed-species groups

38.4 ± 33.6

19

single- or mixed-species groups (Table 2). Thus,
there was no early-warning benefit to N. clavipes
associated with groups of M. incrassata exceed-
ing that of individuals in single-species groups.

DISCUSSION

These data support predictions that grouped
spiders benefit from lower predation and/or an
early-warning system. They do not, however,
support the hypothesis that odd group members
(i. e., N. clavipes in M. incrassata colonies) are
selectively preyed upon. Nephila in both single-
and mixed-species groups suffered a lower fre-
quency of predation attempts and had greater
response distances than solitary individuals.
However, there was no evidence that individuals
living in M. incrassata colonies had any greater
advantages than those living in single-species
groups. The similarly low predation frequency
on both types of groups may be related to hesi-
tation on the part of wasp or bird predators to
attack grouped webs. An alternative explanation,
however, may be related to a foraging related
benefit of group living. Nephila in single and
mixed-species groups capture significantly more
prey biomass, and hence are larger in body size
than the solitary individuals (Hodge 1 990). De-
pending on the size of the predator, there may
be limits to the size of spider that they will attack.

The greater response distances observed for
grouped webs was most likely a function of more
silk between the “potential predator” and the
spider than exists in solitary webs, and thus a
greater distance over which vibrations could be
detected. In natural situations, individual escape
behaviors might transmit vibrations and elicit
evasive behavior among other members of the
group. Two factors may account for the lack of
difference in response distance between single-

and mixed-species associated N. clavipes. First,
it was only possible to use somewhat peripheral
spiders as focal animals in mixed-species groups
(which typically contained 500-1000 spiders) to
avoid reaching into and partially destroying the
colony. The amount of silk between the experi-
menter and the spider was therefore probably not
much different in single-species or mixed-species
situations. In addition, even if the distances had
been different, there may be an upper limit to
the distance that disturbance vibrations can be
transmitted before they attenuate. Therefore,
there may be some upper limit of group size
beyond which no additional early-warning ben-
efit will accrue. However, the possibility does
exist that had a different experimental protocol
been used a slightly greater response distance may
have been detected for N. clavipes in M. incras-
sata colonies, since these colonies do have more
silk, and hence, a greater potential distance for
signal transmission.

Existing evidence suggests that foraging ben-
efits most likely select for tolerance and coloni-
ality in spiders (Lubin 1974; Rypstra 1989; Uetz
1988, 1989) and conditions of very high prey
density favor the formation of mixed-species as-
sociation as well (Rypstra 1979, 1983; Hodge
1 990). Large spider colonies may actually be a
liability with respect to predation since they po-
tentially attract more egg-sac parasites (Lubin
1974; Rypstra 1979; Buskirk 1981; Smith 1982;
Heiber & Uetz 1990). However, the results of
this study indicate that some antipredator ben-
efits may result from group-living. While not
necessarily the driving force behind the evolu-
tion of social tendencies in spiders, such anti-
predator benefits may, in some cases, be an ad-
vantageous fortuitous effect of both single- and
mixed-species associations between web-build-
ing spiders.

ACKNOWLEDGMENTS

We thank S. Marshall, L. Higgins and K. Can-
gialosi for helpful comments on an earlier ver-
sion of this manuscript. Funding for this project
was provided by grants from a National Science
Foundation Doctoral Dissertation Improvement
Grant BSR-8601078, the University of Cincin-
nati Research Council, Sigma Delta Epsilon
(Graduate Women in Science); Sigma Xi, and
the Exline-Frizzel fund (California Academy of
Sciences).

HODGE & UETZ-ANTIPREDATOR BENEFITS OF GROUPING BY NEPHILA

215

LITERATURE CITED

Askenmo, C., A. von Bromssen, J. Ekman & C. Jans-
son. 1977. Impact of some wintering birds on spi-
der abundance in spruce. Oikos, 28: 90-94.

Barnard, C. J. & D. B. A. Thompson. 1985. Gulls
and plovers: the ecology and behaviour of mixed-
species feeding groups. Columbia University Press,
New York.

Berry, J. W. 1987. Notes on the life history and be-
havior of the communal spider Cyrtophora mol-
uccensis (Doleschall) (Araneae, Araneidae) in Yap,
Caroline Islands. J. ArachnoL, 15:309-319.

Bradoo, B. L. 1972. Life history and bionomics of
Idris sp. nov. (Scelionidae: Hymenoptera) egg par-
asite of Uloborus ferokus, a commensal of social
spider Stegodyphus sarasinorum Karsch. Zool. Anz.,
188:43-52.

Bradoo, B. L. 1979. Uloborus ferokus sp. nov. (Ar-
aneidae: Uloboridae) a commensal of Stegodyphus
sarasinorum Karsch. Bull. British. ArachnoL Soc.,
4:353-355.

Buskirk, R. E. 1981. Sociality in the Arachnida, Pp.
281-367, In Social Insects. (H. R. Hermann, ed.).
Vol. II. Academic Press, New York.

Eisner, T. & S. Nowicki. 1 983. Spider web protection
through visual advertisement: role of the stabili-
mentum. Science, 219:185-187.

Farr, J. A. 1977. Social behavior of the golden silk
spider TVep/zi/fl clavipes. J. Arachnol., 4:137-144.

Gomez-Pompa, A. 1973. Ecology of the vegetation
of Veracruz, In Vegetation and vegetational history
of northern Latin America. (A. Graham, ed.). El-
sevier Publishing Co., Amsterdam.

Gunnarsson, B. 1983. Winter mortality of spruce-
living spiders: effect of spider interaction and bird
predation. Oikos, 40:226-233.

Hamilton, W. D. 1971. Geometry for the selfish herd.
J. Theor. Biol, 31:295-211.

Heiber, C. S. & G. W. Uetz. 1990. Colony size and
parasitoid load in two species of colonial Metepeira
spiders from Mexico (Araneae: Araneidae). Oecol-
ogia, 82:145-150.

Higgins, L. in press. Developmental changes in bar-
rier web structure under differing levels of predation
risk. J. Insect Behav., 00:00-00.

Hodge, M. A. 1 990. The behavioral ecology of mixed-
species groups of web-building spiders. Ph.D. Dis-
sertation. Univ. of Cincinnati; Cincinnati, Ohio,
USA.

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

Horton, C.C. 1982. Predators of two orb-web spiders
(Araneae, Araneidae). J. Arachnol., 11:447^49.

Jackson, R. R. 1986. Communal jumping spiders
(Araneae: Salticidae) from Kenya: interspecific nest
complexes, cohabitation with web-building spiders.

and intraspecific interactions. New Zealand J. ZooL,
13:13-26.

Lopez, A. 1988. Nephila inaurata inaurata (Walck-
enaer) and other Reunion Island orb- weavers. Newsl.
British Arachnol. Soc., 51:2-3.

Lubin, Y. D. 1 974. Adaptive advantages and the evo-
lution of colony formation in Cyrtophora (Araneae:
Araneidae). Zool. J. Linn. Soc., 54:321-339.

McCook, H. C. 1890. American spiders and their
spinning work. Vol. 11. Acad. Nat. Sci. Philadelphia.

Milinski, M. 1977. Experiments on the selection by
predators against spatial oddity of their prey. Z.
TierpsychoL, 43:31 1-325.

Moore, C. W. 1977. The life cycle, habitat and vari-
ation in selected web parameters in the spider Neph-
ila clavipes Koch (Araneidae). American Midi. Nat.,
98:95-108.

Morse, D. H. 1977. Feeding behavior and predator
avoidance in heterospecific groups. Bioscience, 27:
332-339.

Morse, D. H. 1980. Behavioral Mechanisms in Ecol-
ogy. Harvard University Press, Cambridge, Mas-
sachusetts.

Mueller, H. C. 1975. Hawks select odd prey. Science,
188:953-954.

Ohguchi, O. 1978. Experiments on the selection
against colour oddity of water fleas by three-spined
sticklebacks. Z. TierpsychoL, 47:254-267.

Pulliam, H. R. & T. Caraco. 1984. Living in groups:
is there an optimal group size?. Pp. 122-147, In
Behavioural Ecology: An Evolutionary Approach,
2nd ed.,(J. R. Krebs & N. B. Davies, eds.). Blackwell
Scientific Publications, Oxford.

Robinson, M. H. & B. Robinson. 1976. The ecology
and behavior of Nephila maculata. a supplement.
Smithson. Contrib. Zool., 218:122.

Rypstra, A. L. 1979. Foraging flocks of spiders: a
study in aggregate behavior in Cyrtophora citricola
Forskal (Araneae: Araneidae) in West Africa. Be-
hav. Ecol. SociobioL, 5:291-300.

Rypstra, A. L. 1983. The importance of food and
space in limiting web-spider densities: a test using
field enclosures. Oecologia, 59:312-319.

Rypstra, A. L. 1 984. A relative measure of predation
on web-building spiders in temperate and tropical
forests. Oikos, 43:129-132.

Rypstra, A. L. 1985. Aggregations of Nephila clavipes
(L.) in relation to prey availability. J. Arachnol., 1 3:
71-78.

Rypstra, A. L. 1989. Foraging success of solitary and
aggregated spiders: insights into flock formation.
Anim. Behav., 37:274-281.

Sabath, M. D., L. E. Sabath & A. M. Moore. 1974.
Web, reproduction and commensals of the semi-
social spider Cyrtophora moluccensis (Araneae: Ar-
aneidae) on Guam, Mariana Islands. Micronesica,
10:51-55.

Shear, W. A. 1970. The evolution of social phenom-

216

THE JOURNAL OF ARACHNOLOGY

ena in spiders. Bull. British Arachnol. Soc., 1:65-
76.

Smith, D. R. R. 1982. Ecological costs and benefits
of communal behavior in a presocial spider. Behav.
Ecol. Sociobiol., 13:107-114.

Spiller, D. A. & T. W. Schoener. 1989. Effect of a
major predator on grouping of an orb-weaving spi-
der. J. Anim. Ecol., 58:509.

Strusaker, T. T. 1969. Notes on the spiders Uloborus
WM«r//or (Chamberlain & Ivie) and Nephila ciavipes
(L.) in Panama. American Midi. Nat., 82:61 1-613.

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.

Uetz,G.W. 1985. Ecology and behavior of Metepe/ra
spinipes (Araneae: Araneidae) a colonial web-build-
ing spider from Mexico. Nat. Geogr. Res. Rep., 19:
597-609.

Uetz, G. W. 1988. Group foraging in colonial web-
building spiders: evidence for risk-sensitivity. Be-
hav. Ecol. Sociobiol., 22:265-270.

Uetz, G. W. 1989. The “ricochet effect” and prey
capture in colonial spiders. Oecologia, 81:154-159.

Vollrath, F. 1985. Web spider’s dilemma: a risky
move or site dependent growth. Oecologia, 68:69-
72.

Yoshida, M. 1988. The spiders intruding into the
webs of Nephila clavata - a consideration in relation
to the densities of spiders. Acta Arachnol., 36: 1-10.

Zar, J. H. 1984. Biostatistical Analysis. 2nd ed. Pren-
tice-Hall, Inc., Englewood Cliffs, New Jersey.

Manuscript received 7 January 1992, revised 13 April
1992.

1992. The Journal of Arachnology 20:217-221

DISPERSAL OF THE SPIDERLINGS OF
XYSTICUS EMERTONI (ARANEAE, THOMISIDAE),
A LITTER^DWELLING CRAB SPIDER

Douglass H. Morse: Graduate Program in Ecology and Evolutionary Biology,

Division of Biology and Medicine, Box G-W, Brown University, Providence, Rhode

Island 02912 USA

ABSTRACT. Dispersing spiderlings of Xysticus emertoni (Thomisidae), a litter-dwelling species, placed on
sites similar to their nests on leaves usually descended quickly into the vegetation below. However, those placed
on nearby goldenrod (Solidago spp.) flowers remained significantly longer, sometimes hunting, before dropping
lower in the vegetation. They seldom ballooned from either site. Xysticus dispersal behavior differs markedly
from that of the thomisid Misumena vatia, a flower-inhabiting species, which balloons regularly from leaves
and remains much longer on goldenrod flowers than Xysticus.

Resource spacing and availability play an im-
portant role in an individual’s predisposition to
disperse (Southwood 1962; Richter 1970; Dingle
1984). In turn, dispersal patterns influence other
aspects of the life styles of the individuals in
question. Combined, these factors should exert
a dominant impact on patterns of gene flow and
consequent population structure. It is therefore
instructive to compare the dispersal patterns of
related species with extremely different life styles.

Two thomisid crab spiders, Xysticus emertoni
Keyserling and Misumena vatia (Clerck), pro-
vide such a comparison. Xysticus is primarily an
inhabitant of herbaceous vegetation and litter in
fields and pastures of eastern North America
(Comstock 1940; Dondale & Redner 1978). Al-
though it sometimes hunts in flowers at the top
of the herbaceous layer, Xysticus occurs much
less frequently, remains for shorter periods, and
as an adult experiences considerably lower hunt-
ing success in flowers than does Misumena (Morse

1983) . The latter species concentrates its activ-
ities, especially as an adult, at flowers (Morse

1984) . Nevertheless, Xysticus sometimes places
its nests in emergent herbaceous plants, as does
Misumena. This placement presents Xysticus
spiderlings with an excellent opportunity to dis-
perse.

Although thomisid spiders regularly occur in
samples of aerial “plankton” (Glick 1939; Salm-
on & Homer 1977; Greenstone et al. 1987), re-
ports often do not distinguish between species or
genera. Most frequently, reference is made to

Misumenops F. Pickard-Cambridge and Misu-
menoides F. Pickard-Cambridge in these catches,
because these are the thomisid genera that dom-
inate the aerial fauna— Xysticus is extremely rare
in aerial catches (M. H. Greenstone, pers. comm.).
Other than for Richter’s (1970, 1971), Green-
stone’s (1982), and Miller’s (1984) studies on
lycosids, no experimental effort has been made
to compare the ballooning by different species of
a spider family, although Tolbert (1977) pro-
vides some comparative information on two ara-
neid species.

I therefore tested dispersal patterns of Xysticus
emertoni spiderlings for comparison with recent
results (Morse in press) on the dispersal of Mis-
umena vatia spiderlings. In the latter study I es-
tablished that Misumena readily balloon after
emerging from their egg sacs, although the pro-
pensity to do so is strongly influenced by the
substrate they occupy. If the nest sites offer rich
foraging opportunities for the spiderlings, pri-
marily flowering goldenrods (Solidago spp.) in
my study area, they are very likely to remain and
feed in them. Their consequent foraging success
and subsequent growth at these sites decrease
their probability of ballooning. However, Mis-
umena of all ages live above the litter in their
habitat, concentrating their activities in flowers,
a characteristic associated with a nearly contin-
uous quest for insect food. In light of the much
less frequent occupation of the somewhat ex-
posed sites by Xysticus, and the seemingly more
homogeneous areas that they occupy low in the

217

218

THE JOURNAL OF ARACHNOLOGY

vegetation, one may predict a considerably lower
probability of ballooning than for Misumena,
even though Xysticus nests are often placed in
positions that make this type of dispersal pos-
sible.

In the present study, I tested the propensity of
second instar Xysticus spiderlings, newly emerged
from their egg sacs, to balloon in experimental
situations similar to those under which I had
tested Misumena spiderlings. I then compared
the two species, paying particular attention to
the implications of these differences for dispersal,
potential gene flow and population structure in
general.

METHODS

I found several Xysticus emertoni nests on
broad-leaved vegetation in the study area while
pursuing other work. I measured key parameters
of their locations as I encountered them, but made
no special attempt to hunt for them in the litter
layer. Thus, I do not claim that those nests are
typical of all X. emertoni. Young were taken from
these nests for release experiments (discussed be-
low), and others were taken from broods laid in
the laboratory, for a total of 22 broods.

I released over 200 newly-emerged, second in-
star Xysticus spiderlings from locations that re-
sembled one of their frequent nest sites, leaves
of the common milkweed Asclepias syriaca\ and
inflorescences of nearby goldenrod clones {Soli-
dago juncea and 5. canadensis) that attracted
large numbers of tiny insects. Five young from
a single brood were placed on a substrate (milk-
weed leaf or goldenrod inflorescence) at a time,
the maximum number that an observer could
carefully watch and record under these condi-
tions. These densities frequently occur when they
emerge from a nest. All statistics were run on the
responses of groups of spiderlings, with only one
group used from a brood in any experiment. Pri-
or to release the spiderlings were lightly dusted
with powdered red micronite dye, which in-
creased their visibility to the observer. Earlier
experiments with Misumena spiderlings had
demonstrated that this manipulation did not af-
fect their subsequent behavior (Morse in press).
The studies were carried out in a field in Bremen,
Lincoln Co., Maine, an area that I have described
in detail elsewhere (Morse 1979, 1981).

I observed these spiderlings continuously dur-
ing the first two hours following release, or until
they had all dispersed. If any remained at the

end of two hours I censused them twice or more
daily to determine the approximate time at which
they dispersed from the substrate on which they
were released. I recorded movements of these
spiderlings, the time they remained on the sub-
strates to which they were introduced, and the
methods by which they left these sites (balloon-
ing, dropping on lines, etc.). I then compared
these results with those fox Misumena spiderlings
that had been exposed to similar experiments
(Morse in press).

RESULTS

Location of Xysticus nests in the field.— I found
nine Xysticus emertoni nests during 1989-1991,
of which a majority (five) were on milkweed, two
on aster {Aster sp.), one on chokecherry {Prunus
virginiana), and one on raspberry {Rubus sp.).
The nests were constructed at the ends of leaves,
and on all but the aster the distal tip was turned
under the rest of the leaf and secured, the eggs
laid between the two resulting thicknesses of leaf,
and the sides drawn tight by silk into a compact
nest. These nests resembled those of Misumena
(figure 1, Morse 1985), except that Xysticus
mothers ensconced themselves inside their nests,
rather than guarding the nests from the outside.
In contrast, Xysticus folded the narrow aster
leaves twice, permitting a nest to be fashioned
by essentially providing a third side from plant
material. In a sample now exceeding 1 500, 1 have
never seen a Misumena nest built by folding a
leaf in the latter way.

Nests were all located in low vegetation at a
mean height of 54.9 (± 12.1 SD) cm, in vege-
tation of? 1.3 ± 17.8 cm. The nests on milkweed
most frequently occupied the third pair of leaves
from the top, ranging from the second to the fifth
from the top.

Movement of Xysticus spiderlings on various
substrates.— Yysi/cwi' spiderlings have a strong
tendency to descend into the litter when placed
on substrates similar to the ones on which their
parents normally build their nests. This behavior
occurred both in young placed on the normal
sites (milkweed leaves) and on nearby goldenrod
flowers at the peak of bloom (Table 1), a source
at which large numbers of tiny insects, potential
prey of these spiderlings, often congregate. How-
ever, they responded quantitatively differently to
these two substrates, remaining significantly lon-
ger on goldenrod than on milkweed (Table 1) (P
< 0.05 in a two-tailed Mann-Whitney U-test).

MORSE- DISPERSAL OF XYSTICUS SPIDERLINGS

219

Table 1. — Time (min) that newly emerged Xysticus spiderlings remained on different substrates, with similar
results on Misumena for comparison (Morse in press), a = Misumena remained 1 76. 1 ± 25 1 .5 min on milkweed,
3820.0 ± 3435.5 min on goldenrod (Morse in press), b = 19 of 50 Misumena released on milkweed were
observed to balloon, but none of 50 released on goldenrod were observed to balloon (Morse in press).

■ , Method of dispersal

Time remained

Substrate

Year

N

(X ± SD)

Drop

Line

Balloon

Not Known

Milkweed""

1987

92

8.1 ± 11.9

64

17

gb

3

1990

65

18.0 ± 21.0

53

4

0

8

Goldenrod""

1990

38

114.8 ± 169.1

11

0

2»

25

1991

10

144.4 ± 203.8

5

0

0

5

Further, numbers of spiderlings observed quick-
ly dropping on lines directly to the litter from
milkweed far exceeded those dropping from
goldenrod {P < 0.002 in a two-tailed Mann-
Whitney LZ-test). This difference, combined with
observations of two Xysticus spiderlings captur-
ing tiny midges on goldenrod during these pe-
riods, suggested that they used these flowers as
hunting sites. This result is consistent with oc-
casional observations of early-instar Xysticus on
these sites under unmanipulated conditions (Ta-
ble 2).

The most frequent movement from the release
sites was directly into the litter, either on vertical
lines or by crawling down the vegetation (Table
1). These movements occurred significantly more
frequently than dispersal via horizontal lines to
adjacent grass stems {P < 0.01 for milkweed,
P = 0.05 for goldenrod in Wilcoxon matched
pairs signed ranks tests). The outcome of most
moves on horizontal lines probably did not differ
in function from the vertical movements, how-
ever, since all such individuals that I could follow
on horizontal lines used their new locations as
staging sites from which to move into the litter,
rather than for aerial dispersal. Some of the latter
movements could have been a consequence of

the wind moving the line-laying spiders from a
vertical to horizontal position while they were
leaving the original sites. Although the results
clearly indicated that Xysticus does balloon, it
was the least common of the activities recorded
in these releases (Table 1).

Movement of Xysticus and Misumena spider-
lings.—Although Xysticus spiderlings resembled
Misumena spiderlings in remaining significantly
longer on goldenrod (foraging substrate) than on
milkweed (nest substrate), they stayed on both
of these sites only a small fraction of the time
that Misumena did (Table 1), differences that are
highly significant {P < 0.002 for goldenrod, P <
0.00 1 for milkweed in two-tailed Mann-Whitney
(7-tests). These results are consistent with their
parents’ habits and with the relative scarcity of
Xysticus spiderlings on goldenrods and other
flowers. In censuses of spiderlings on goldenrods
only two Xysticus spiderlings were found on over
900 inflorescences, in comparison to over 600
Misumena spiderlings (Table 2).

Also striking is the difference in dispersal modes
of Xysticus and Misumena spiderlings from the
release sites. Although a majority of Xysticus spi-
derlings left these sites for the litter, either di-
rectly by lines or by crawling down the vegeta-

Table 2.— Numbers of Xysticus and Misumena spiderlings on flowering goldenrod. a = Each clump is distinct
from others and probably a clone, b = Involves daily counts over I'h to 3 ‘/2-week period following release of a
total of 404 Misumena spiderlings.

Sample of goldenrod

Number
of clumps""

Number of
flowering
stems

Number of
Xysticus
spiderlings

Number of
Misumena
spiderlings

Randomly chosen

25

111

1

248

> 5 m from Misumena nest

10

151

0

27

< 1 m from Misumena nest

12

439

0

359

Release of 4 Misumena broods'"

4

42

1

-

220

THE JOURNAL OF ARACHNOLOGY

tion, Misurnena spiderlings never descended to
the litter (Table 1). This difference is highly sig-
nificant {P < 0.002 for both milkweed and gold-
enrod in two-tailed Mann-Whitney [/-tests), as
is the difference in frequency with which the two
species balloon from milkweed {P < 0.002 in a
two-tailed Mann-Whitney [/-test) (Table 1). Al-
though a few Xysticus were observed to balloon
off goldenrod as well as milkweed in the trials,
Misurnena were observed to balloon only off
milkweed. Nevertheless, the frequency of bal-
looning from goldenrod by Xysticus was so low
that the two species did not differ significantly
{P > 0.05 in a two-tailed Mann-Whitney [/-test).
Since many Xysticus left goldenrod within the
two-hour period of continuous observation (Ta-
ble 1), the probability of observing them bal-
looning from goldenrod was far higher than for
Misurnena. Misurnena remained for long peri-
ods, often several days, on goldenrod (Table 1),
and they were only censused a few times a day
after the original observation period.

DISCUSSION

The behavior of Xysticus spiderlings resem-
bled that of Misurnena spiderlings in that dis-
persal time was related to substrate in both spe-
cies. This difference strongly suggests that they
discriminate between sites. However, actual times
required to disperse were considerably shorter
for Xysticus than for Misurnena. This brief ten-
ure is consistent with Xysticus'% distribution at
similar heights as adults (Morse 1983).

Richter (1970) and Greenstone (1982) have
provided the only previous experimental studies
on between-species differences in ballooning pat-
terns, although they have all been performed in
the laboratory using artificial wind, heat, and light
sources. Young wolf spiders of different Pardosa
species vary in their ballooning tendencies, which
Richter attributed to the abundance and stability
of their habitats, and Greenstone to the predict-
ability of the habitats. Propensity to balloon dif-
fered inversely with each of these traits. Abun-
dant habitats often have a low level of patchiness.

This study thus suggests that spatial patchiness
may be added to the variable of temporal patch-
iness as a factor affecting ballooning. Both thom-
isids live in similar habitats, but as a result of
their markedly different demands on these hab-
itats, they probably view patchiness at strikingly
different scales. For most or all of its stages, Mis-
umena depends on insects drawn to extremely

patchy flower resources, but Xysticus does not
depend primarily on this resource. Even when
Xysticus does hunt on flowers, it remains there
for much shorter periods than Misurnena, with
a periodicity suggesting that its poor hunting suc-
cess may account for the short tenure (Morse
1983). Because Xysticus obtains a major part of
its prey away from the flowers, within the veg-
etation and litter layers, its resources are not like-
ly to be as patchy as those of Misurnena. Xysticus
should therefore balloon less frequently than
Misurnena, as observed. Thus the two species
appear to respond to the same habitat in dis-
tinctly different ways, notwithstanding their sim-
ilar size and close phylogenetic relationship.

These results, plus the infrequent natural pres-
ence of Xysticus spiderlings on goldenrod, sug-
gest that the latter spider seldom moves above
the litter layer. Consequently, its dispersal dis-
tances as juveniles are likely to be low. Gene flow
should thus be much lower in Xysticus than in
Misurnena, which should in turn generate dif-
ferences of population structure in the two spe-
cies. Nevertheless, since Xysticus sometimes bal-
looned, it clearly retains the ability to initiate
long-distance movement.

ACKNOWLEDGMENTS

I thank M. H. Greenstone, J. D. Parrish, G.
Stratton, and R. B. Suter for reading a draft of
this manuscript. My research on Misurnena is
supported by the National Science Foundation
(BSR85-16279 and BSR90-07722). I thank H.
Heller, J. Kotanchik, N. McKay, and J. Rollen-
hagen for assistance with the field work, E. B.
Noyce for kindly permitting use of the study site,
and M. H. Greenstone for information on tho-
misid genera in aerial samples.

LITERATURE CITED

Comstock, J. H. 1940. The spider book, revised and
edited by W. J. Gertsch. Cornell University Press,
Ithaca, New York.

Dondale, C. D. & J. H. Redner. 1978. The insects
and arachnids of Canada, Part 5. The crab spiders
of Canada and Alaska. Canada Dept, of Agriculture
Publ. 1663:1-255.

Glick, P. A. 1939. The distribution of insects, spiders,
and mites in the air. United States Dept, of Agri-
culture Tech. Bull., 671:1-150.

Greenstone, M. H. 1982. Ballooning frequency and
habitat predictability in two wolf spider species (Ly-
cosidae: Pardosa). Florida Entomol., 65:83-89.
Greenstone, M. H., C. E. Morgan, A.-L. Hultsch, R.

MORSE -DISPERSAL OF XYSTICUS SPIDERLINGS

221

A. Farrow, & J. E. Dowse. 1987. Ballooning spi-
ders in Missouri, USA, and New South Wales, Aus-
tralia: family and mass distributions. J. Arachnol.,
15:163-170.

Miller, G. L. 1 984. Ballooning in Geolycosa turricola
(Treat) and Geolycosa patellonigm Wallace: high
dispersal frequencies in stable habitats. Canadian J.
ZooL, 62:2110-2111.

Morse, D. H. 1979. Prey capture by the crab spider
Misumena calycina (Araneae: Thomisidae). Oec-
ologia, 39:309-319.

Morse, D. H. 1981. Prey capture by the crab spider
Misumena vatia (L.) (Thomisidae) on three com-
mon native flowers. American Midi. Natur., 105:
358-367.

Morse, D. H. 1983. Foraging patterns and time bud-
gets of the crab spiders Xysticus emertoni Keyserling
and Misumena vatia (Clerck) (Araneae: Thomisi-
dae) on flowers. J. Arachnol., 1 1:87-94.

Morse, D. H. 1984. How crab spiders (Araneae:
Thomisidae) hunt at flowers. J. Arachnol., 12:307-
316.

Morse, D. H. 1985. Nests and nest-site selection of
the crab spider Misumena vatia (Araneae, Thomis-
idae) on milkweed. J. Arachnol., 13:383-390.

Morse, D. H. in press. The relationship between dis-
persal by spiderlings from their nests and earlier
foraging patch decisions made by their mothers.
Ecology.

Richter, C. J. J. 1970. Aerial dispersal in relation to
habitat in eight wolf spider species (Pardosa, Ara-
neae, Lycosidae). Oecologia, 5:200-214.

Richter, C.J.J. 1971. Some aspects of aerial dispersal
in different populations of wolf spiders, with par-
ticular reference to Pardosa amentata (Araneae, Ly-
cosidae). Misc. Pap., Landouwhogeschool, Wagen-
ingen. The Netherlands, 8:77-88.

Salmon, J. T. & N. V. Homer. 1977. Aerial disper-
sion of spiders in North Central Texas. J. Arachnol.,
5:153-157.

Southwood, T. R. E. 1962. Migration of terrestrial
arthropods in relation to habitat. Biol. Rev., 37:
171-214.

Tolbert, W. W. 1977. Aerial dispersal behavior of
two orb-weaving spiders. Psyche, 84:13-27.

Manuscript received 2 January 1992, revised 3 Septem-
ber 1992.

1992. The Journal of Arachnology 20:222-226

COURTSHIP BEHAVIOR AND SEXUAL CANNIBALISM
IN THE SEMI-AQUATIC FISHING SPIDER,
DOLOMEDES FIMBRIATES (CLERCK) (ARANEAE: PISAURIDAE)

Goran Arnqvist: Department of Animal Ecology, University of Umefi; S-901 87
UmeS, Sweden

ABSTRACT. The courtship behavior of the semi-aquatic Pisaurid fishing spider Dolomedes fimbriatus was
examined in the laboratory. Male courtship was triggered by the presence of female drag-lines, presumably by
a female sex pheromone since males did not respond with courtship to male drag-lines. Male courtship behavior
included vibratory signaling (water surface waves), leg-waving, and following female drag-lines. Vibratory sig-
naling was a major courtship component, and signals were produced at a regular rate (mean rate: 8.33 ± 1.53
s, n = 97). Irrespective of whether females were mated or unmated, females were very aggressive towards males,
and sexual cannibalism prior to copulation occurred in 6.6% of the female attacks on males (« = 76). The capture
success rate of females depended on whether the male was attacked from a distance or from immediate proximity.
The occurrence of sexual cannibalism of courting males by virgin Dolomedes females is discussed, and it is
suggested that this behavior of fishing spiders may represent an adaptive female strategy rather than mistaken
identity.

The semi-aquatic fishing spiders of the genus
Dolomedes (Pisauridae) are large spiders that in-
habit various freshwater habitats such as the
shoreline of streams and lakes (Carico 1973). The
main prey of fishing spiders are other aquatic
and semi-aquatic invertebrates and vertebrates,
and terrestrial invertebrates trapped at the water
surface (Bleckmann & Lotz 1987; Zimmermann
& Spence 1989). The sensory ecology of fishing
spiders with regards to predation has been rather
well examined (Carico 1973; Williams 1979; Ro-
land & Rovner 1983; Bleckmann & Barth 1984;
Bleckmann & Rovner 1984; Bleckmann 1985).
In contrast, the reproductive behavior of fishing
spiders has been thoroughly described only for
the Nearctic species D. triton, where vibratory,
tactile, chemical, and visual communication all
play important roles during courtship (Roland &
Rovner 1983; Bleckmann & Bender 1987). Such
information of the Holarctic species D. fimbria-
tus is limited, and only scattered observations in
older studies are available (Pappenheim 1903;
Gerhardt 1926; Schmidt 1953, 1957).

The purpose of the current study is to describe
male courtship behavior as well as female re-
sponse to male courtship in D. fimbriatus, and
to evaluate the relative roles of vibratory, tactile,
chemical, and visual stimuli during courtship.
Dolomedes spiders are known to exhibit sexual
cannibalism (females killing and consuming
males) (Schmidt 1957; Zimmermann & Spence

1989; Foelix 1982; Elgar 1992), and the occur-
rence of sexual cannibalism in D. fimbriatus is
quantified and discussed.

METHODS

Adult D. fimbriatus were collected from a dense
population at Sirapsbacken (64° 22' N, 19° 28'
E), an alluvial meadow by the river Vindelalven
in northern Sweden. A total of 25 females and
17 males was captured on June 17-18, 1991.
Individuals were placed individually in plastic
aquaria (0.45 x 0.25 m) filled with water to a
depth of 8 cm. Each aquarium was provided with
three pieces of floating styrox elements (15x5
cm) which served as resting sites. Ambient tem-
perature was 20 °C ±1 °C, and males were fed
daily one water strider (Gerris odontogaster) per
individual. Females were fed one water strider
and one field cricket {Gryllus bimaculatus, > 20
mm body length) per individual per day.

Behavioral observations were made in trials,
where one male was introduced to a female in
her aquarium for 50 minutes. When introduced,
males were carefully placed at the water surface
as far away from the female as possible, on the
opposite side of the aquaria. During each trial,
the behavior of the spiders was observed visually
and videotaped for subsequent behavioral anal-
yses from slow motion replays. A total of 36 trials
was performed. In a first round of trials (« = 25),
all females were exposed to a male. In a second

222

ARNQVIST-COURTSHIP BEHAVIOR IN DOLOMEDES FIMBRIATUS

223

round {n = 11), all females that had not yet laid
eggs were allowed a second exposure to a male.
Males were numbered individually and used for
behavioral trials in numerical order. Thus in-
dividuals were chosen systematically for the be-
havioral trials, and each female was used 1-2
times and each male 2-3 times (if not eaten by
a female in the first round). After the behavioral
trials, females were fed (see above) and allowed
to lay and tend their eggs. Ten days after the date
of egg laying, the egg sac was opened. By this
time the embryos in fertilized eggs were clearly
visible, and the fertilization rate of the egg batch
could thus be recorded.

In order to determine the role of chemical
stimuli relative to that of visual and tactile in
triggering male courtship behavior, a series of
experiments was performed. In addition to the
behavioral trials described above (treatment I, n
= 36), males were also introduced (in numerical/
systematical order by being carefully placed at
the water surface) into one of three types of
aquaria: male aquarium with a male (treatment
II, n = 5), uninhabited new and thus clean aquar-
ium with a styrox element from a female aquar-
ium (treatment III, n = 6), and uninhabited new
and clean aquarium with a styrox element from
a male aquarium (treatment IV, n = 5). The sty-
rox in treatments III and IV was covered with
drag-lines from their previous environment. In
each of these trials (treatments II-IV), the spiders
were observed for 50 minutes, and it was re-
corded only whether or not the introduced male
exhibited courtship.

RESULTS

Of the 25 females brought to the laboratory,
1 7 were penultimates (performed their ultimate
moult in the laboratory prior to the behavioral
trials) and eight were adults. Seventeen females
were thus virgin, whereas the mating status of
the females collected as adults was unknown at
the time of the behavioral trials. However, the
mating status of these females could be deter-
mined after the experiments since unmated D.
fimbriatus females eat their unfertilized eggs after
egg deposition (Schmidt 1957), which is also the
case in Pardosa wolf spiders (Lycosidae) (Vlijm
et al. 1963; Kessler 1970). After the behavioral
trials, 1 6 females laid unfertilized eggs that were
partially consumed by the female (15 collected
as penultimates and one collected as adult). Nine
females laid and tended their eggs (two collected
as penultimates and seven as adults). In five cases

the egg batches were incompletely fertilized (fer-
tilization rate 25-50%) and in four cases the egg
batches were completely fertilized (fertilization
rate > 95%). Females that tended their egg sac
laid 350.8 ± 79.4 eggs per egg sac {n = 9).

Courtship. — When a male D. fimbriatus was
introduced into an aquarium inhabited by an
adult female (treatment I), the initial male re-
sponse was an “announcement display”. This
behavior commenced on average 3.09 ± 3.21
min (« = 36) after introduction to the aquarium,
apparently always when the male first made
physical contact with a female drag-line (irre-
spective of whether the drag-line was placed on
the water surface or on a styrox element). The
announcement display consisted of two major
components; vibratory signals and leg-waving.
Males produced vibratory signals on both sub-
strate types, by a slight elevation of the body
followed by a sudden jerky lowering of the ab-
domen, causing concentric water surface waves
to spread out from the male (when vibratory
signals were performed at the water surface).
These jerks were made in the same manner as
in D. triton males (Bleckmann & Bender 1987).
Vibratory signals were produced at a regular rate,
once every 8.33 ± 1.53 s (« = 97), and males
often continued to produce vibratory signals
throughout the trials (50 min). During the inter-
vibrational pauses, males occasionally per-
formed leg-waving, where the spider lifted and
waved legs I. When leg- waving, the male alter-
nated between the left and right leg in an irregular
vertical waving-pattem with the legs typically
being held extended forward, straight and stiff.
Leg- waving was often combined with a rapid tap-
ping with legs I on the substratum, especially
when on styrox, presumably producing addi-
tional vibrations. When placed in a female
aquarium males moved very slowly, typically a
few mm/min, normally following a female drag-
line with the palps and legs I. Vibratory signals,
leg-waving, leg-tapping, and very cautious ad-
vance along a female drag-line during the inter-
vibrational pauses were alternated. Males per-
formed announcement display for 29.36 ± 14.71
min (n = 36) during each trial.

Males seemed unable to visually detect mo-
tionless females, for males frequently (1.42 ±
0.39 times per trial) passed in very close range
(nearest distance between tips of legs less than 2
cm) behind them, apparently without detecting
their exact location.

The role of chemical stimuli.— Males always

224

THE JOURNAL OF ARACHNOLOGY

responded with courtship when placed in a fe-
male environment (treatments I and III), but
never exhibited courtship in male environment
(treatments II and IV) (x^ = 53.0, df = 3, P <
0.001; Table 1). In treatment III, males typically
commenced with courtship signaling upon first
physical contact with a styrox element from a
female cage. Males did not respond to drag-lines
of other males (treatment IV).

Female response to male courtship. — Females
typically remained motionless during the behav-
ioral trials with legs III and IV anchored to styrox
and the anterior two pairs of legs resting upon
the water surface, and no female courtship was
observed. However, the females exhibited two
different responses to male courtship which will
be described below; attack from a distance and
passivity. A total of 89 male-female interactions
was observed. The first interaction in a trial oc-
curred after 26.6 ± 19.7 min {n = 36), and 76%
of the interactions consisted of female attacks on
males from a distance. The average attack dis-
tance was 6.76 ± 3.53 cm {n = 26). In 97% of
these long distance attacks males avoided cap-
ture by rapid evasive movements, often includ-
ing a change of direction (approximately 90°)
making it more difficult for females to pursue
males. In three of these cases, the males were
seized by the females but escaped by rapidly au-
totomizing legs. However, in the other 3% of the
attacks, males were caught and cannibalized by
females.

In 24% of the interactions, females remained
passive when the male approached. Males mak-
ing physical contact with a female immediately
started to tap vigorously with legs I and II on
her legs and abdomen for 2.03 ± 1.42 min, after
which the male immediately mounted the female
from behind and attempted copulation. When
mounting, males climbed up onto the abdomen
of the female, turned around (facing in a direc-
tion opposite that of the female) and slid some-
what sideways in order to reach the epigynum
with its palps. Only two successful copulations
occurred. In both cases, only one palp was in-
serted and the females involved were virgin fe-
males collected as penultimates which later laid
and tended incompletely fertilized egg batches.
Although no data on exact palpal insertion times
were collected, they were both very brief (< 15
sec) in accordance with the observations made
by Schmidt (1957).

During the phase of physical contact (male on
top of female), lasting 1.36 ± 0.83 min, the fe-

Table 1.— The occurrence of male Dolomedes fim-
briatus courtship behavior in different experimental
treatments in the laboratory.

No. of
trials
yield-
ing
court-
Total ship
no. of re-

Treatment

trials

sponse

I. Male introduced to a female in fe-
male aquarium

36

36

II. Male introduced to a male in male
aquarium

5

0

III. Male introduced to a new aquar-
ium with styrox element from fe-
male

6

6

IV. Male introduced to a new aquar-
ium with styrox element from
male

5

0

male either remained totally passive (62% of the
cases) in which case the male attempted copu-
lation and then made a sudden vertical jump
followed by rapid withdrawal, or females sud-
denly attacked the male (38% of the cases). When
attacking a male in physical contact, females
caught and cannibalized males in 37.5% of the
cases, a success rate significantly higher than long
distance attacks (Fisher exact contingency table
test [two-tailed], P = 0.016). In total (for both
types of attacks, n = 76), males were caught and
cannibalized in 6.6% of the attacks.

There were no apparent differences in response
to male courtship between mated and unmated
(virgin) females, and females did not differ sig-
nificantly in level of aggression; females attacked
the male in 73.1% of the interactions involving
mated females {n = 26) and in 76.2% of the
interactions involving unmated females {n =
63)(Fisher exact contingency table test [two-
tailed], P = 0.791).

DISCUSSION

In general, the courtship behavior of D. fim-
briatus includes many components typical for
Pisaurid and Lycosid courtship, such as vibra-
tory signaling, leg-waving, and female drag-line
following (Barth 1982; Foelix 1982; Robinson
1982; Tietjen & Rovner 1980, 1982; Roland &
Rovner 1 983; Bleckmann & Bender 1 987). Com-
munication between the sexes in D. firnbriatus

ARNQVIST- COURTSHIP BEHAVIOR IN DOLOMEDES FIMBRIATUS

225

during courtship probably relies on vibratory
(jerks and leg-tapping on substratum), visual (leg-
waving), and chemical stimuli during the initial
phase and tactile stimuli (leg-tapping on female)
during later phases. This is similar to the court-
ship behavior of D. triton, where vibratory com-
munication has been shown to play a major role
during courtship (Roland & Rovner 1983; Bleck-
mann & Bender 1987). However, while the water
surface wave signals of D. triton are produced at
irregular intervals (Roland & Rovner 1983;
Bleckmann & Bender 1987), D. fimbriatus pro-
duced these signals at a regular rate suggesting
that interspecific differences occur in the pattern
of signaling. There may be additional differences
between the species with regards to signal wave
parameters, e.g., frequency content, duration and
amplitude (Bleckmann & Bender 1987).

In D. scriptus and D. triton, male courtship
behavior is triggered by chemical stimuli, and
visual/vibrational cues are not required (Kaston
1936; Roland & Rovner 1983). The results of
the current study show that this is the case also
in D. fimbriatus, and suggest that the female sex
pheromone is bound to the female’s drag-line.
Distance chemoreception does not seem to occur
in D. fimbriatus (cf Tietjen & Rovner 1982).

Several previous studies have demonstrated
that Dolomedes females may be very aggressive
towards courting males, and that sexual canni-
balism may occur (Gerhardt 1926; Schmidt 1957;
Roland & Rovner 1 983). Further, in a field study
of D. triton, Zimmermann & Spence (1989)
showed that one of the most important prey items
of adult female fishing spiders was adult males,
confirming that sexual cannibalism in natural
populations is an important phenomenon in this
group of spiders. D. fimbriatus females in the
current study were extremely aggressive towards
males, and female mating status did not affect
the level of aggression. Despite vigorous court-
ship, males were usually attacked even when ap-
proaching virgin or incompletely mated females
in most cases, and several instances of sexual
cannibalism occurred. It is further worth noting
that females had a significantly higher attack suc-
cess rate when attacking males in immediate
proximity compared to attacks from a distance.

There is a current controversy over the evo-
lution of sexual cannibalism (Buskirk et al. 1 984;
Gould 1984; Elgar 1992). Sexual cannibalism af-
ter copulation may represent an adaptive male
paternal investment strategy (Thornhill 1976;
Buskirk et al. 1984), while sexual cannibalism

prior to copulation may be adaptive even for
virgin females (Elgar 1991; Newman & Elgar
1991). Sexual cannibalism may also simply rep-
resent cases of “mistaken identity” (Robinson
1982; Elgar 1992). Dolomedes females may be
characterized as sit-and-wait predators. The pos-
terior legs are normally anchored to a firm object
(e.g., a rock or the vegetation), while the anterior
legs are spread out on the water surface (Carico
1973; Williams 1979). Studies of the sensory
ecology of fishing spiders have demonstrated that
females in this typical position have a sophisti-
cated system for detecting and interpreting water
surface wave vibrations, and that females are
capable of discriminating prey and non-prey gen-
erated surface waves (Williams 1979; Roland &
Rovner 1983; Bleckmann & Barth 1984; Bleck-
mann & Rovner 1984; Bleckmann 1985). In a
study of the vibratory component of courtship,
Bleckmann & Bender (1987) concluded that the
courtship surface wave signals of male fishing
spiders are insufficient to release female prey cap-
ture behavior since they lack prey wave char-
acteristics, and that females should be able to
identify males solely on the basis on vibratory
cues. This conclusion, combined with the fact
that sexual cannibalism in fishing spiders occurs
primarily before copulation (Gerhardt 1926;
Schmidt 1957; Roland & Rovner 1983; this
study), suggests that cannibalism of courting
males by virgin fishing spider females might rep-
resent an adaptive female strategy rather than
cases of mistaken identity. Virgin D. fimbriatus
females may benefit from killing and consuming
a courting male provided that the risk of re-
maining unmated is low (see Newman & Elgar
1991). However, it is difficult to distinguish be-
tween different models of sexual cannibalism,
and future work should be directed at testing the
various assumptions and predictions of the mod-
el of Newman & Elgar (1991).

ACKNOWLEDGMENTS

Thanks are due to C. Otto for constructive
comments on the manuscript, and to S. Diehl
for providing assistance with the experiments.

LITERATURE CITED

Barth, F. G. 1 982. Spiders and vibratory signals; sen-
sory reception and behavioral significance. Pp. 67-
122, In Spider communication: mechanisms and
ecological significance. (P. N. Witt & J. S. Rovner,
eds.). Princeton Univ. Press, Princeton, New Jersey.
Bleckmann, H. 1985. Discrimination between prey

226

THE JOURNAL OF ARACHNOLOGY

and non-prey wave signals in the fishing spider Do-
lomedes triton (Pisauridae). Pp. 2 1 5-222, In Acous-
tic and vibrational communication in insects. (K.
Kalmring & N. Eisner, eds.). Parey Verlag, Berlin.

Bleckmann, H. & F. G. Barth. 1984. Sensory ecology
of a semi-aquatic spider (Dolomedes triton). II. The
release of predatory behavior by water surface waves.
Behav. Ecol. Sociobiol., 14:303-312.

Bleckmann, H. & M. Bender. 1987. Water surface
waves generated by the male pisaurid spider Do-
lomedes triton (Walckenaer) during courtship be-
havior. J. Arachnol., 15:363-369.

Bleckmann, H. & T. Lotz. 1987. The vertebrate-
catching behavior of the fishing spider Dolomedes
triton (Araneae, Pisauridae). Anim. Behav., 35:641-
651.

Bleckmann, H. & J . S. Rovner. 1984. Sensory ecology
of a semi-aquatic spider (Dolomedes triton). I. Roles
of vegetation and wind-generated waves in site se-
lection. Behav. Ecol. Sociobiol., 14:297-301.

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

Carico, J. E. 1973. The Nearctic species of the genus
Dolomedes (ArancaQ-. Pisauridae). Bull. Mus. Comp.
Zool., 144:435-488.

Elgar, M. A. 1991. Sexual cannibalism, size dimor-
phism, and courtship behavior in orb-weaving spi-
ders (Araneidae). Evolution, 45:444^48.

Elgar, M. A. 1 992. Sexual cannibalism in spiders and
other invertebrates. Pp. 128-155, In Cannibalism:
ecology and evolution among diverse taxa. (M. A.
Elgar & B. J. Crespi, eds.). Oxford Univ. Press, Ox-
ford.

Foelix, R. F. 1 982. Biology of spiders. Harvard Univ.
Press, Cambridge, Massachusetts.

Gould, S. J. 1984. Only his wings remained. Nat.
Hist., 93:10-18.

Gerhardt, U. 1 926. Weitere untersuchungen zur Bio-
logic der Spinnen. Z. Morphol. Okol. Tiere, 6: 1-77.

Kaston, B. J. 1936. The senses involved in the court-
ship of some vagabond spiders. Entomol. Ameri-
cana, 16:97-167.

Kessler, A. 1970. Egg production in Pardosa. I. In-

fluence of mating on the egg-ripening period in Par-
dosa lugubris (WaIckenaer)(Araneae, Lycosidae).
Bull. Mus. National D’Hist. Nat., suppl. I, 2:98-
101.

Newman, J. A. & M. A. Elgar. 1991. Sexual canni-
balism in orb-weaving spiders: an economic model.
American Nat., 138:1372-1395.

Schmidt, G. 1953. Fine deutsche Spinne, die Wir-
beltiere friBt, Orion 7/8.

Schmidt, G. 1957. Einige Notizen fiber Dolomedes
fimbriatus (Cl.). Zool. Anz., 158:83-97.

Pappenheim, P. 1 903. Beitrage zur Kenntnis der En-
twicklungsgeschichte von Dolomedes fimbriatus
Clerck. Z. Wiss. Zool., 74:109-154.

Robinson, M. H. 1982. Courtship and mating be-
havior in spiders. Ann. Rev. Entomol., 27:1-20.

Roland, C. & J. S. Rovner. 1983. Chemical and vi-
bratory communication in the aquatic pisaurid spi-
der Dolomedes triton (Araneae: Pisauridae). J. Ar-
achnol., 1 1:77-85.

Thornhill, R. 1976. Sexual selection and paternal in-
vestments in insects. American Nat., 1 10:153-163.

Tietjen, W. J. & J. S. Rovner. 1980. Trail-following
behaviour in two species of wolf spiders: sensory
and etho-ecological concomitants. Anim. Behav.,
28:735-741.

Tietjen, W. J. & J. S. Rovner. 1982. Chemical com-
munication in Lycosids and other spiders. Pp. 249-
279, In Spider communication: mechanisms and
ecological significance. (P. N. Witt & J. S. Rovner,
eds.). Princeton Univ. Press, Princeton, New Jersey.

Vlijm, L., A. Kessler & C. J. J. Richter. 1963. The
life-history of Pardosa amentata (Cl.) (Araneae, Ly-
cosidae). Ent. Ber. 23:75-80.

Williams, D. S. 1979. The feeding behavior of New
Zealand Dolomedes species (Araneae: Pisauridae).
New Zealand J. Zool., 6:95-105.

Zimmerman, M. & J. R. Spence. 1989. Prey use of
the fishing spider Dolomedes triton (Pisauridae, Ara-
neae): an important predator of the neuston com-
munity. Oecologia, 80:187-194.

Manuscript received 25 May 1992, revised 11 Septem-
ber 1992.

1992. The Journal of Arachnology 20:227-228

BOOK REVIEW

Martin Lister’s English Spiders, 1678. Trans-
lated by Malcolm Davies and Basil Harley, ed-
ited by John Parker and Basil Harley. Harley
Books, Colchester, England. 160 pp.; Hardbound
£49.95 [$87.65], paperback £24.95 [$43.80], +
shipping overseas £3.50 [$6.15].

Some years ago a student, well versed in arach-
nid literature, told me that it was Martin Lister
(1638-1712) who first separated species by pal-
pal morphology. As no English translation was
available, I never checked Lister’s Latin volume
to see whether the story was correct. This is the
first time Lister has been translated into English.
Only the spider part is included, not the second
part, which is on freshwater and marine mol-
lusks. There is, however, a 1778 translation of
Lister’s work into German.

Lister described molting, courtship and cop-
ulation of spiders, and also feeding of its young
by Theridion sisyphium. Lister correctly sepa-
rated males from females by the swollen palpi
of the males. He observed that spiders have no
penis but use their swollen palpi to touch the
females abdomen, while “two-eyed spiders’’ (op-
ilionids) have a penis. He counted eyes and legs
and leg articles, and described eight-eyed spiders
with anal appendages, a six-eyed spider {Dys-
dem), and two-eyed spiders (now called Opi-
liones and Acari) that lacked a pedicel. He dif-
ferentiated eggs and eggsacs and generalized that
small spiders produce few eggs, large spiders pro-
duce many, that some spiders carry egg sacs
around with them, and that spiders have no lar-
vae. He observed that spiders produce numerous
threads at the same time from the anal append-
ages, and can ascend by submitting themselves
to a gentle breeze or can attach a bridge of air-
borne silk to an object some distance away. This
observation is correct although some present day
spider books have misinformation, stating that
the spider walks down one stem and up another
to make the bridge (impossible according to W.
Eberhard, pers. commun.). Lister recognized that
the spider’s skills are innate, not learned. Lister
speculates that skins of all later molts may be
present in spiderlings from the start. Spiders feed
on insects or each other and wasps prey on spi-
ders.

Lister described how orb-webs are made: the
radii constructed first, from the middle to the
periphery; the first spiral laid from the center out,
the final spiral laid from outside in; sometimes
a hole in the hub made at the end; the use of
knots or glue to tie threads together.

He contradicts 23 traditional views held by
Aristotle, Pliny and his own contemporary,
Moulfet. According to the editors. Lister’s cita-
tions of old literature are not always correct.

As Lister practiced medicine and published
papers on medical matters, (although he must
have spent all of his spare time watching spiders
and snails), his comments on the use of spiders
and webs as medicines might be of interest. How-
ever, he provides only an unconvincing list with-
out comments. He lists macerations for warding
off fevers; spiders steeped in olive-oil or rose-
water for earache; a wax salve from spiders ap-
plied to the navel for hysteria, swelling of the
spleen and boils; spiders from rosewater for ces-
sation of lactation, gout, ringworm and other
spreading skin diseases; cobwebs or their ash for
stanching the flow of blood and for healing open
sores and inflammations; spiders’ eggs from oil
of spikenard applied to an aching tooth or for
tertian fever. Lister wrote before double blind
experiments were used to demonstrate efficacy.
Lister tried to investigate venoms, and noted that
spiders are not noxious when eaten.

Each of the thirty-eight spiders he recorded for
England and Wales (and considered to be all there
were) is illustrated and described in about two
pages, including habitat and life history infor-
mation. The spiders are also grouped in a table
according to the type of web made and the num-
ber of eyes. The editors have matched Lister’s
spiders with species known from the area, and
have assigned them their present day Linnaean
binomials. The editors comment on how they
matched the name to each of the Lister species,
but if Lister differentiated only one species of a
genus (e.g., Tetragnatha), I wonder whether he
might have mixed observations from several
species he did not differentiate. Was the spider
fauna in Lister’s time the same as today when
the population of England is nearly 1 0 times larg-
er?

Lister’s detailed descriptions and numbering

227

228

THE JOURNAL OF ARACHNOLOGY

of spiders are in marked contrast to the skimpy
work of some authors of the 1 8th and early 1 9th
century, who gave names accompanied by one
sentence descriptions. These later works have led
to subjective interpretation of species names,
which is the reason that strict priority of names
in spiders and other invertebrates does not nec-
essarily lead to stable nomenclature.

Lister was a friend of John Ray and other no-
tables of the era, and these friendships were
maintained partly through correspondence. The
editors looked through surviving correspon-
dence in British archives and reproduced ex-
cerpts dealing with spiders. The letters, written
in English with inconsistent 17th century spell-
ing, are here produced in readable modem En-
glish, with a plate showing a reproduction of an
original letter.

Even in the 17th century there were nasty dis-
putes. Did Martin Lister or Edward Hulse make
the first observation of spiders emitting gossamer
and sailing through the air? It seems that Lister
was the first to make the observation, but Hulse
got ahead to publish.

Unlike later authors. Lister used alcohol as a
preservative and he reports that green coloration
washed out. In England as late as the 1870’s
spiders were still speared and dried on insect
pins.

This translation into modem English uses
modem anatomical terms. The editing is superb.
Present day ethologists should check this volume
before making statements on who made a first
spider observation.

Nowhere did I find that he used palpi to dif-
ferentiate species. The repeated statement that

males are distinguished by their palpi refers to
differentiating males from females.

This was the first time that I examined the
1778 German translation in the Harvard Uni-
versity Library. It differs by having a list of pre-
1778 publications about spiders. Names have
been attached to the species for which Lister had
assigned numbers, and these binomial names are
preceded by “List.” or “Lister”; reference is made
to Linnaeus in footnotes. Also, additional infor-
mation and plates are provided; such additional
information is followed by the abbreviation G.,
(for the junior editor). In Roewer’s catalog ( 1 942)
and in Bonnet (1955) some of these names are
assigned to Lister rather than to the editors Mar-
tini & Goeze; some I could not find. These old
names are probably all synonyms and hom-
onyms; the less said about them, the better.

LITERATURE CITED

Bonnet, P. 1955. Bibliographia Araneorum, Tou-
louse, 2:1-918.

Martini, F. H. W. & Goeze, J. A. E. 1778. Martin
Listers Naturgeschichte der Spinnen iiberhaupt und
der Engellandischen Spinnen insonderheit aus dem
Lateinischen iibersetzt und mit Anmerkungen ver-
mehrt. Quedlinburg und Blankenburg. 302 pp., 5
pi.

Roewer, C. F. 1942. Katalog der Araneae von 1758
bis 1940. Bremen, 1:1-1040.

Herbert W. Levi: Museum of Comparative Zo-
ology, Harvard University, Cambridge, Mas-
sachusetts 02138, USA.

Manuscript received 31 October 1992.

INSTRUCTIONS TO AUTHORS

Manuscripts are preferred in English but may be accepted
in Spanish, French or Portuguese subject to availability of
appropriate reviewers. Authors whose primary language is not
English, may consult the Associate Editor for assistance in
obtaining help with English manuscript preparation. All man-
uscripts should be prepared in general accordance with the
current edition of the Council of Biological Editors (CBE) Style
Manual unless instructed otherwise below. Authors are ad-
vised to consult a recent issue of the Journal of Arachnology
for additional points of style. Send three identical copies of
the typed material together with copies of illustrations to:

Gary L. Miller, Associate Editor
The Journal of Arachnology
Department of Biology
The University of Mississippi
University, Mississippi 38677 USA
Telephone: (601) 232-7495
FAX: (601) 232-5144
Bitnet: BYMILLER@UMSVM
Correspondence relating to manuscripts should also be di-
rected to the Associate Editor.

Manuscripts longer than 1 500 words should be prepared as
Feature Articles, shorter papers as Research Notes.

FEATURE ARTICLES

Title page.— Place in the upper left comer the complete name,
address, and telephone number of the author with whom proofs
and correspondence should be exchanged. Include FAX num-
bers and Bitnet addresses if available. Four lines below, type
the title, centered, all in capital letters. Another four lines be-
low, beginning with the left margin, type each author’s name
on a new line followed by the address.

Abstract.— The heading should be placed at the beginning
of the first paragraph set off by a period. In articles written in
English, a second abstract in an acceptable language may be
included pertinent to the nationality of the author(s) or geo-
graphic region(s) emphasized in the article. A second abstract
in English must be included in articles not written in the latter
language.

Text.— Double-space text, tables, legends, etc. throughout.
Three levels of heads are used. The first level (METHODS,
RESULTS, etc.) is typed in capitals and on a separate line.
The second level head begins a paragraph with an indent and
is separated from the text by a period and a dash. The third
level may or may not begin a paragraph but is underscored
and separated from the text by a colon. Use only the underline
method to indicate italics; DO NOT USE WORD PROCES-
SOR ITALICS CODES. Use the metric system unless quoting
text or referencing collection data. All decimal fractions are
indicated by the period regardless of language of the text.

Citation of references in the text.— Include within parenthe-
ses the surnames of the author(s) followed by the date of pub-
lication. A comma separates multiple citations by the same
author(s) and a semicolon separates citations by different au-
thors, e.g., (Smith 1970), (Jones 1988; Smith 1989), (Smith
1986, 1987; Smith & Jones 1989; Jones et al. 1989).

Literature cited section.— Use the following style:
Lombardi, S. J. & D. L. Kaplan. 1990. The amino acid
composition of major ampullate gland silk (dragline) of Ne-
phila clavipes (Araneae, Tetragnathidae). J. Arachnol., 18:
297-306.

Krafit, B. 1982. The significance and complexity of com-
munication in spiders. Pp. 1 5-66, In Spider Communica-
tions: Mechanisms and Ecological Significance. (P. N. Witt
& J. S. Rovner, eds.). PrincetonUniversity Press, Princeton.
Footnotes.— Footnotes are only permitted on the first printed
page to indicate current address or other information con-
cerning the author(s). All footnotes are placed together on a
separate manuscript page.

Running head.— Author sumame(s) and an abbreviated title
must be typed all in caps and must not exceed 60 characters
and spaces. The running head is placed by itself on a separate
manuscript page.

Taxonomic articles.— Consult a recent taxonomic article in
the Journal of Arachnology for style or contact the Associate
Editor.

Tables.— Each table, with the legend above, should be placed
on a separate manuscript page. Only the appropriate horizontal
lines should be included. Use no footnotes; instead, include
all information in the legend. Make notations in the text mar-
gins to indicate the preferred location of tables in the printed
text.

Illustrations.— Original art work should be kept by the au-
thor(s) until final acceptance for publication. All art work must
be mounted on stiff board ready for reproduction. In line draw-
ings, pay particular attention to width of lines and size of
lettering when reductions are to be made by the printer. Mul-
tiple photos assembled on a single plate should be mounted
with only a minimum of space separating them. In the case of
multiple illustrations mounted together, each illustration must
be numbered sequentially rather than given an alphabetic se-
quence. Written on the back should be the name(s) of author(s)
and an indication of top edge. The author should indicate
whether the figure should be one column or two columns in
width. The overall dimensions should be no more than 1 1
inches (28 cm) x 14 inches (36 cm). Larger drawings present
greater difficulty in shipping and greater risks of damage for
which the JOA assumes no responsibility. The author is en-
couraged to consider PMT (PhotoMechanical Transfer) repro-
ductions for large drawings and to consult an Editor for details
on appropriate reductions. Make notations in the text margins
to indicate the preferred position of illustrations in the printed
text.

Legends for illustrations should be placed together on the
same page(s) and separate from the illustrations. Adhere to the
following styles;

Figures 1-4.— A-us x-us, male from Timbuktu: 1, left leg;
2, right chelicera; 3, dorsal aspect of genitalia; 4, ventral aspect
of abdomen.

Figures 27-34.— Right chelicerae of species of A-us from
Timbuktu: 27, 29, 3 1 , 33, dorsal views; 28, 30, 32, 34, prolater-
al views of moveable finger; 27, 28, A-us x-us, holotype male;
29, 30, A-us, w-us, holotype male; 31, 32, A-us z-us, holotype
male; 33, 34, A-us y-us, male. Scales = 1.0 mm.

Assemble manuscript for mailing.— Assemble the separate
sections or pages in the following sequence; title page, abstract,
text, figure legends, footnotes, running head, tables with leg-
ends, figures.

Page charges and reprints.— The current charge per journal
page will be assessed as follows: $50— nonmembers, $45 —
members with grant, institutional, or other support for page
costs; $25 — regular members who do not have support for page
costs; $20— student members who do not author with full
members and who do not have support. Reprints are available
only from the printer at current costs.

Waiver of Page Charges.- The Editor will consider requests
for a waiver of page charges for manuscripts which have been
accepted for publication. Requests should be made in writing
and should be submitted to the Associate Editor along with
the final revision of the manuscript. The author must certify
that neither grant nor institutional funds are available to cover
all of the publication costs. The Society will normally agree to
waive up to a maximum of 80% of the page charges, contingent
on the availability of funds; the author is expected to find
alternative sources for the remainder.

RESEARCH NOTES

Instructions above pertaining to feature articles apply also
to research notes except that abstracts and most headings are
not used and the byline follows the Literature Cited section.

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 20 Feature Articles NUMBER 3

Spider (Araneae) taxa associated with Mantispa viridis (Neuroptera: Man-
tispidae) Jeffrey R. Bmshwein, Kevin M. Hoffman, and Joseph D.

Culin 153

Life cycle and habitat preference of the facultatively arboreal wolf spider,
Gladicosa pulchra (Araneae, Lycosidae) Micky D. Eubanks and Gary

L. Miller 157

Abundance and association of cursorial spiders from calcareous fens in

southern Missouri Thomas L. Bultman 165

Web orientation, thermoregulation, and prey capture efficiency in a tropical

forest spider Leslie Bishop and Sean R. Connolly 173

Numerical response to prey abundance by Zygiella x-notata (Araneae, Ar-

aneidae) David A. Spiller 179

A new species of North American tarantula, Aphonopelma paloma (Araneae,

Mygalomorphae, Theraphosidae) Thomas R. Prentice 189

Sex ratio in the social spider Diaea socialis (Araneae: Thomisidae) David

M. Rowell and Barbara York Main 200

Survivorship of wolf spiders (Lycosidae) reared on different diets George

W. Uetz, Jennifer Bischoff, and Joseph Raver 207

Antipredator benefits of single- and mixed-species grouping by Nephila cla-
vipes (L.) (Araneae, Tetragnathidae) Margaret A. Hodge and George
W. Uetz 212

Dispersal of the spiderlings of Xysticus emertoni (Araneae, Thomisidae), a

litter-dwelling crab spider Douglass H. Morse 217

Courtship behavior and sexual cannibalism in the semi-aquatic fishing spi-
der, Dolomedes fimbriatus (Clerck) (Araneae: Pisauridae) Gdran
Arnqvist 222

Book Review

Martin Lister’s English Spiders, 1678 (translated by Malcolm Davies and

Basil Harley, edited by John Parker and Basil Harley) Herbert W. Levi 111

23 SI XI

ffi/93 98115

(X

m

II NviNOSHiiws saiHvaan libraries Smithsonian institution NoiiniiiSNi nvinoshj

S SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3IBVdan LIBRARIES SMITHSOI
(/> 2 V 2: ^ ^ w z ...(/>

< s < S ,<

I .
c/>

^

> 2 'CLS^ > ■'^ur’ 5 >

2 cn z £/) Z {/) *. Z

ji NviNOSHilws S3 1 a va an libraries Smithsonian institution NoiiniiiSNi nvinosh

— CO ^ (/) CO —

O

S^SMITHS0NIAN”‘|NSTITUTI0N NOliDiliSNl'^NVlNOSHiltMS S3iavaan LIBRARI ES^SMITHSO

2 [I ^ [I ' ^ 2

O r-r, 5 xCtTsfirlyX ® O

m ^ m I ;;

±. CO £ CO £ CO _

II NviNosHims S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinosh
2 ^2 ^ 5 ^ ^ ^ ^ 2 » ‘z’ 2:

- ^ ^ ^

2 -H Wy/zmj:, ? =< z

^ J^Y A' ^ z 'KW^AJJK

UJ

cc

<

ex

CQ

S 5 • W >

CO 2 CO '2 CO '" 2 CO

S SMITHSONIAN INSTITUTION NOIiniliSNI_ NVINOSHillAIS S3iavaan LIBRARIES SMITHSO

<0 — CO 2 \ ^ ^ CO

liJ Xrt^TiroX /ft V UJ

II NVINOSHillAIS S3 lava an libraries SMITHSONIAN institution NOlinillSNI nvinosh

2 r- z r-

ra

50
>

m ^ m ^ m ^ m

co^. £ CO £ CO £ CO

S SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHIIINS S3iavaan LIBRARIES SMITHSC

CO 2 V CO 2 CO 2 -.• CO

2 . < 2 , <

2 =5 .. o:/ ' g

CO
X

X o

> 5 x^;v_Dc/ j, 2 >

JI NviNOSHiiiMs'^ss I ava a n libraries^^smithsonian institution NoiiniiiSNi nvinosh

— CO ~ CO ^ CO —

to

Q O xic^pc^ O

S^SMITHSONIAN'^INSTITUTION NOliniliSNI "^NVINOSHIIINS S3iavaan~'LIBRARI ES^SMITHSC
2 [3 2 r- , 2 r- 2 ,

“ mr "' ' ^

_ z _j z

iiyvaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws ssiuvuan

r- 2 r- ^ z r- z

2 . vtxj 2 m 2

t 1

■ » s I

(fl zz Ui — (J)

STITUTION NOlinillSNI NVINOSHilWS S3iavaan LIBRARIES SMITHSONIAN INSTITUTION
fn Z ^ W 2!

=; ../ -I .,y../ Ay,. Z .v«V^v -1 2

i

n a va a LI b rar i Es‘^SMiTHS0NiAN^iNSTiTUTi0N'^N0iiniiiSNi^NviN0SHiiws‘^s3 1 ava a n

//\ f/) •“ tfi

o:

<

.IXilK-V :: VcSWPy m g ='■%'«

STITUTION ^NOIiniliSNl“'NVlNOSHilUS^S 3 lavaan'^LIBRARI ES^SMITHS0N1AN~'|NST1TUTI0N
2 r- , 2 r- 2

O ^^Tt-TiriTN. vv O xt^jSTiriT^ ia O

_ {/) “ tn E (/i

1 1 ava an libraries smithsonian institution NoiiniiiSNi NViNOSHims S3iavaan

to Z * to 2 to

J, ^ J, - 2 >

STITUTION ‘^N0liniliSNl_NVIN0SHillAIS^S3 I a Va a n_LI B RAR 1 es^^smithsonian^institution

_ V -X-5NJS. < (N
y 03

aiavaan libraries smithsonian institution NoiiniiisNi nvinoshiiws S3iavaan

2 r- Z r- z

(/} — <A — CO

iSTiTUTiON NoiiniiiSNi NViNOSHiiiAis S3iavaan libraries smithsonian institution

to 2 w Z ... w 2 ,v

- y y -I /y.-/ yi* ' Z -I '

o X .'Cw- AJf' o Xs ^

to :/^y£w to to

o -Sy/zM. X o

o

- Z to 2 to * Z (/5

3iavaan libraries smithsonian institution NoiiniiisNi NviNosnims S3iavaan

to X to — CO

O Dt>^ ~ O >s^"l D>-^ Q

ISTITUTION N0liniliSNI~'NVIN0SHillMS^S3 I avaan~^LIBRARI ES^SMITHSONIAN“'lNSTITUTION
X r- , z r- Z

O v^TtrlrTTN. O ><^Tir77S^ — ,,^7^ r>-^ O