^ 2 - — ' ' 2 w z z

liAN~INSTITUTION NOIiflilXSNI NVINOSHJIIIMS SBiavaan LIBRARIES SMITHSONIAN INSTITI
Z s:, ^ ^ 5 ^2 Z

" I 3 (^^1#'''#'= '

I Q ^5^m. 2 m"/im

IWs“’s3iaVM0n^LIBRARIEs'^SMITHSONIAN INSTITUTION NOliDlUSNI NVINOSHilWS'"S3 I y V
S ,-r:r^ i " '" ‘

in

5/ CQ

o X^/V.DC>X _ o

nAN~'iNSTiTUTiON Noiiniii.SNi’l'NviNOSHiims'^sa lyvyan librari es^smithsonianiinstiti

iiNS saiavHan libraries Smithsonian institution NOiiniiiSNi nvinoshliws saiyv

</> Z ■, w Z » w Z in ,

- V - ^ .% ^ - /s:S^ ^

>* S § X^iVASkA^ ^ >■'" ^ ^^iClOSVAN^^ >*

lAN^INSTITUTlON'^NOlinillSNI NVINOSHiiV^s'^ S3 I « Vd 8 11 LI B RAR I ES^^SMITHSONIAN INSTITI
ay ^ \ ^ <n

in

Q ^ O ■■' X^vAsjii/ Q !^ ' O

IWS^S3 I d vy a n~’L ! B R AR I ES SMITHSONIAN INSTITUTION NOIinillSNl’^NVINOSHJ.mS S3 I
2: r- z ^ z: r- . 2:

JIAN""inST1TUTI0N NOliniliSNI NVINOSHIIIMS SBIHVHQII LIBRARIES SMITHSONIAN INSTIT

X

V ?

I^s'^'S3iyvy9n'”L*BRARlE$'^SMITHS0NIAN INSTITUTION NOliniliSN! NVINOSHXIl^S S3ia^
t/> ^ J2 - X .„. c/5

5 S g ffl

'IIAN"'lNSTITUTI0N~N0liniliSNl'"NVIN0SHilliMS^S3l avaa Ill'Ll BRAR I ES^SMITHSONIAN^INSTIT

“ z r- Z

UJ

VA^'

n

C/) ~ c/> — ^ V.

Liws saiavaan libraries Smithsonian institution NouniiiSNi nvinoshliws S3ia'i

(/) , z N, in -z. in z

a 2 S fMfWA g /#P

to

AlAN institution NOliniliSNI NVINOSHillNS S3IHVaan libraries SMITHSONIAN INSTIT

— (O zz '-, in zz ay n ^

^ CO ? in \ ^ •

'joiJ.niiiSNi NviNOSHiiws S3iav8an libraries Smithsonian institution NoiiniiisNi

z: in "Z. 2 to

.IBRARIES SMITHSONIAN INSTITUTION NOliniUSNI NVINOSHJ.IWS S3 I 8 V 8 8 n_
in ^ z

O v:4j-tixv>^ O

JOIiniliSNl"'NVINOSHimS^S3 l8V8an”*LIBRARI ES^SMITHSONIAN-INSTITUTION ^ NOliniliSNI
r~ , r- 2: ^ ^ '

CO

/:v/' ‘2

m

in

.IBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3 I 8V8 3 IT ,L1 B R AR I ES
g 2: , w z "■ 2- . I

lOliniUSNI NVINOSHLIWS S3l8V8an LIBRARIES SMITHSONIAN INSTITUTION „

= </> . z; _ «

CO

NOliniliSNI

O -'.W 1: o N>os»l^ i: o

^ I ^ I

.IBRARIES SMITHSONIAN INSTITUTION NOliflilXSNI MVINOSHLIINS S3l8Vaan LIBRARIES
z (- z r ^ -

— . ^ ^ £/> X ^ ^

NOliniliSNI NVINOSHilWS S3l8V8an LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI
z CO 2 in 2 ,N>'. CO

» iCliW iOii i

CO 2 (/) *,2 (O ^-^iinriArkor-,^

LIBRARIES SMITHSONIAN^ INSTITUTION NOIini!ISN!_NVINOSHlliNS S3ldVHan_ LIBRARIES

tK

o

r- 1°

^ \r

ffi vt-

NOliniliSNI ~'NVIN0SHill«S^S3 lavaan'^LIBRARI ES^SMITHS0NIAN"'|NSTITUT!0N NOliniliSNI
r- , z r- z S

O ■- O o

cu r™ DO

73
>

m ''W rn Xi^vos^^ m

CO z: CO E

LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3 I 8V8 3 H LI B R AR I ES

CO 2 -r <n 2 ^ ^ S

■\'^^ i %

g 5;.- 2 XirosH^X >• 2

NOliniliSNI NVIN0SHillNs'^S3iavaan^LIBRARIEs‘^SMlTHS0NIAN INSTITUTION NOliniliSNI

5 \ CO := in zz CO ^

/

■i

I

I

(

/

\

The Journal of

ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 14

SPRING 1986

NUMBER 1

THE JOURNAL OF ARACHNOLOGY

EDITOR: Oscar F. Francke, Texas Tech University

ASSOCIATE EDITOR: William B. Peck, Central Missouri State University
ASSISTANT EDITOR: James C. Cokendolpher, Texas Tech University
EDITORIAL ASSISTANT: Scott A. Stockwell, Texas Tech University
EDITORIAL BOARD: C. D. Dondale, Agriculture Canada; W. G. Eberhard,
Universidad de Costa Rica; M. E. Galiano, Museo Argentino de Ciencias
Naturales; W. J. Gertsch, American Museum of Natural History; N. F.

Hadley, Arizona State University; V. F. Lee, California Academy of Sciences;
H. W. Levi, Harvard University; E. A. Maury, Museo Argentino de Ciencias
Naturales; W. B. Muchmore, University of Rochester; M. H. Muma, Western
New Mexico University; N. 1. Platnick, American Museum of Natural
History; G. A. Polis, Vanderbilt University; S. E. Riechert, University of
Tennessee; M. E. Robinson, U.S. National Zoological Park; V. D. Roth,
Southwestern Research Station; J. S. Rovner, Ohio University; W. A. Shear,
Hampden-Sydney College; W. J. Tietjen, Lindenwood College; G. B. Uetz,
University of Cincinnati; C. E. Valerio, Universidad de Costa Rica; S. C.
Williams, San Francisco State University.

THE JOURNAL OF ARACHNOLOGY (ISSN 0161-8202) is published in
Spring, Summer, and Fall by The American Arachnological Society at Texas
Tech Press.

Individual subscriptions, which include membership in the Society, are $25.00
for regular members, $15.00 for student members. Institutional subscriptions to
The Journal are $35.00. Correspondence concerning subscriptions and
memberships should be addressed to the Membership Secretary (see back inside
cover). Back issues of The Journal are available from Dr. Susan E. Riechert,
Department of Zoology, University of Tennessee, Knoxville, TN 37916 U.S. A.,
at $10.00 for each number; the Index to Volumes 1-10 is available for $10.00.
Remittances should be made payable to The American Arachnological Society.
Correspondence concerning undelivered issues should be addressed to the Texas
Tech Press, Texas Tech University, Lubbock, Texas 79409, U.S. A.

Change of address notices must be sent to the Membership Secretary.

Manuscripts for The Journal of Arachnology are acceptable only from current
members of the Society. Detailed instructions for the preparation of manuscripts
appear in the Fall issue of each year, and can also be obtained from the Editor
and the Associate Editor. Manuscripts that do not follow those instructions will
be returned to the author(s) without benefit of review. Manuscripts and all
related correspondence must be sent to Dr. William B. Peck, Associate Editor,
Department of Biology, Central Missouri State University, Warrensburg,
Missouri 64093, U.S. A.

PROOFS, REPRINTS, and CHARGES: Authors will receive a reprint order
form along with their proofs. Reprints are billed at the printer’s current schedule
of costs. All authors will be billed by the American Arachnological Society for
page charges. The charge per journal page will be assessed as follows: $ 15.00-
papers acknowledging grant, institutional, or other support; $ 10.00-full members
who do not have (and do not acknowledge) support; $5. 00-student members who
do not coauthor with full members and who do not acknowledge support.

Haskins, M. F. and J. H. Shaddy. 1986. The ecological effects of burning, mowing, and plowing on
ground-inhabiting spiders (Araneae) in an old-field ecosystem. J. Arachnol., 14:1-13.

THE ECOLOGICAL EFFECTS OF BURNING, MOWING,

AND PLOWING ON GROUND-INHABITING SPIDERS
(ARANEAE) IN AN OLD-FIELD ECOSYSTEM

Mary F. Haskins' and James H. Shaddy

Science Division

Northeast Missouri State University
Kirksville, Missouri 63501

ABSTRACT

Cursorial spiders were studied in northeast Missouri from April-November 1980 in annually
manipulated old-fields, in fields undergoing succession from manipulations, and a control field.
Manipulations included burning, mowing, and plowing. Eleven cursorial families were collected in the
study. Pitfall traps were used as the collecting device. Spider communities were compared using Bray-
Curtis similarity indices. Seasonal and monthly spider and plant diversities were calculated using the
Shannon Index. Spider diversity was correlated with plant diversity during May. The relative abundance
of five spider species was correlated with the importance value of several plant species.

INTRODUCTION

Some of the current land management practices include burning, mowing, and
plowing. These practices along with natural disturbances often result in changes
in abundance, species composition, and diversity within both the plant and animal
communities (Lowrie 1942). Ecosystem disturbances appear to be important factors
in the evolution of plant and animal species (Denslow 1980, Connell and Slayter
1976, and Pickett 1976). Since plants are important to arthropods from the
standpoint of food and other factors any natural or man-made disturbance (e.g.
burning, mowing, and plowing) in the plant community may significantly alter
the arthropod community (Lowrie 1942). Relatively small differences in habitat
structure have been shown to affect significantly the abundance and type of spiders
present (Duffey 1978). Many workers have reported associations between arthropod
and plant diversity and succession (Almquist 1973, Muma 1973, Drew 1967, Duffey
1962, Chew 1961, Kajak 1960, and Elliot 1930 cited by Bultman et al., 1982;
Murdoch et al., 1972; Riechert and Reeder 1972; Lowrie 1948, 1942). Spider
diversity and abundance have been correlated with the physical structure of the
litter (Hagstrum 1970, Berry 1967, and Lowrie 1948 cited by Bultman et al., 1982;
Uetz 1979; Uetz 1975, Jocque 1973 cited by Uetz 1976).

Extensive reviews on prairie fire ecology may be found in Wright and Bailey
(1980) and Hulbert (1969). Studies on the effects of cultivation and mowing have
been reported by Duffey (1978), and Kajak et al. (1971). Diversity theories and
indices have been reviewed and modified by many workers (Berry 1967 cited by
Bultman et al., 1982, Pianka 1978, Uetz 1974, Huhta 1971, Goodall 1968, Lloyd

'Present address: Division of Biological Sciences, University of Missouri, Columbia, Missouri 6521 1.

2

THE JOURNAL OF ARACHNOLOGY

et al., 1968, Pielou 1966a, 1966b, and Lowrie 1948). Reviews concerning the use
of pitfall traps may be found in Merrett (1976), Uetz and Unzicker (1976), Gist
and Crossley (1973), Duffey (1972), Huhta (1971), Southwood (1966), and
Greenslade (1964).

Specifically, the objectives of this study were to: 1) identify the cursorial spider
fauna in fields which were manipulated (burned, plowed, and mowed) annually
for the past four years, fields manipulated four years ago and undergoing succession,
and a control field, 2) determine the active density, diversity, and similarity of
cursorial spider populations in these fields, 3) determine the effects of vegetation
on spider distribution, and 4) compare the capture efficiency of white and clear
pittraps (reported elsewhere).

METHODS

Spiders were sampled in an 8.1 hectare area located approximately 2 kilometers
south of Kirksville in northeast Missouri. The area lay fallow for approximately
20 years before the land management study originated. Seven fields employing three
common agricultural land management techniques: burning, mowing, and plowing
were used in the study. Fields were subdivided into four quadrats measuring
approximately 30 m by 40 m each. To reduce edge effect a buffer zone at least
9 m in width surrounded each field. Experimental design involved the administration
of one treatment per field: annually for one quadrat, every fifth year for a second
quadrat, and every 10th and 20th year for the third and fourth quadrats respectively
(Figure 1).

Field manipulations were started in 1976. At that time two fields (all eight
quadrats) were burned, two fields were mowed, and two fields plowed. Manipulated
fields were mowed and plowed in the fall, and burning treatments were conducted
in the spring. Area II fields served as duplicates of Area I to increase the sample
size. The seventh field was left untreated as a control. Samples for this study were
taken from quadrats that were manipulated annually for the past four years,
quadrats manipulated four years ago and undergoing succession, and a control
field which had been fallow for 20 years.

Pitfall trapping for cursorial spiders was conducted from April to November
1980. Six traps, 3 clear and 3 white, were placed approximately 10 m apart in
each quadrat. A total of 42 white and 42 clear traps were used in the study. The
initial trap color arrangement within each quadrat was random. Thereafter the
color arrangement and trap location were constant throughout the study. Traps
were placed in the ground for 7-day intervals approximately every 3 weeks. A
small amount of ethylene glycol was placed in each trap as a preservative. Clear
and white trap data were combined for the spider analyses in this paper.

Vegetation data, including frequency, density, cover, grass mat depth, and species,
were collected for area I quadrats. Three random one-square meter samples from
each quadrat were taken during the months of May, July, and September. Grass
mat depth was also recorded for each sample. Vegetation in Areas I and II appeared
to be similar. Vegetation samples were not taken, however, from Area II because
of time limitations.

Spider identifications were made to the species level for adults, and family level
for Juveniles. Several non-cursorial species were identified only to family.

HASKINS AND SHADDY— HABITAT MANIPULATION OF CURSORIAL SPIDERS

3

Fig. L — Map of study site using field and quadrat abbreviations. Abbreviations: I-area one fields,
Il-area two fields. Four year succession quadrats: burned-IB,IIB; mowed-IM,IIM; plowed-IP,IIP.
Annually manipulated quadrats: burned-IAB,IIAB; mowed-IAM,IIAM; plowed-IAP,IIAP. Control
quadrats-IAC,IC (AC used to distinguish between control fields). (1 cm = 30 m)

Comparisons of active densities were made for selected species between annually
treated quadrats and their succession counterparts. The Bray-Curtis similarity index
(Huhta 1979) was used in community comparisons between habitats which had
received different physical treatments, and between communities undergoing
succession from those physical treatments. The number of guilds and their percent
composition was compared with the study habitats.

Spider diversity using the Shannon (1948) Index was calculated. Previous studies
dealing with pitfall trapping of cursorial spiders have also used the Shannon Index
(Jocque 1973, Uetz 1975, 1976, 1979a, 1979b, Pielou 1966 cited by Bultman et
al., 1982). Diversity analyses were made for each quadrat using data for the trapping
season. Diversity indices between quadrats were tested for significant differences
with a t-test proposed by Hutcheson (1970 cited by Zar 1974). Trap dates were
then separated into three groups which corresponded closely with the vegetation
sampling dates. Diversity analyses were then made for each quadrat using the
grouped data.

Importance values for each plant species were calculated by adding relative
density, relative frequency, and relative dominance values for each species (Cox
1980). Linear correlation analyses were run using the importance values for plant
species and the number of adult spiders captured in selected species. Monthly plant
diversity was also calculated for the samples taken in area I using the Shannon
Index. Hutcheson’s test was used to determine significant differences in plant
diversity.

RESULTS AND DISCUSSION

Representatives of thirteen spider families, eleven of which are considered
cursorial or vagrant web builders, were captured. Forty-one genera with sixty-three

4

THE JOURNAL OF ARACHNOLOGY

species represented the eleven families, with a total of 1,989 adults and 2,945
juveniles (Table 1).

Chi-square analyses indicated that significant differences existed in the numbers
of adult cursorial spiders captured in many of the quadrats. The annually plowed
quadrats (area I and II) exhibited significantly fewer spiders than all quadrats except
the annually mowed (area I) quadrat. The differences were not significant for these
two quadrats (Table 2). The number of adult spiders captured in other annually
manipulated quadrats was usually significantly less than the numbers captured in
their succession counterparts. However, two exceptions were observed: higher
numbers captured in the annually mowed quadrat as compared to mowed quadrat
in area II, and no difference in the numbers captured between annually burned
and burned quadrats in area I. Three quadrats exhibited significantly higher
numbers of spiders than did the control quadrats. These quadrats included: plowed
(area I), annually mowed (area II), and burned (area II). Other succession quadrats
also exhibited higher numbers, but they were not statistically different, as compared
to the control quadrats.

Nine cursorial species representing three families accounted for approximately
65% of the adults captured. The species were: Neoantistea agilis, Hahnia ononidum,
Schizocosa saltatrix, Schizocosa retrorsa, Schizocosa bilineata, Schizocosa avida,
Pardosa saxatilis, Callilepis pluto, Zelotes laccus. The two species with the highest
active densities for the trapping season were S. avida and Z. laccus respectively.

Guild composition analysis revealed wolf spiders as the major guild in all quadrats
except the annually plowed area I quadrat where running spiders predominated,
and in a control quadrat where vagrant spiders were most abundant. Crab and
jumping spiders composed the highest percent guild composition in the annually
mowed and annually burned, area I, quadrats respectively (Figure 2).

Bray-Curtis analyses resulted in similarity indices between annually plowed
quadrats and plowed quadrats of 0.36 and 0.48. Similarity was higher
(approximately 0.65) between annually mowed and annually burned quadrats and
their succession counterparts (Table 3). As expected, annually manipulated quadrats
showed less similarity to the controls than did the succession quadrats. Bray-Curtis
similarity indices indicated the mowed and burned area II quadrats were the most
similar in species composition with a value of 0.79. Both of the annually plowed
quadrats exhibited low degrees of similarity with other quadrats. Similarity indices
were calculated for quadrats between treatment types within area I and area II
(i.e. areas I and II were not compared). Those quadrats having similarity indices
greater than 0.60 are included in Table 3.

Similarity indices were also calculated for the major guilds. Quadrats exhibiting
the highest similarity for the running spiders were mowed and annually burned
quadrats in area II with an index of 0.80. Wolf spiders had the highest similarity
in the mowed and burned quadrats (area II) with a value of 0.85. Their lowest
similarity was between the annually plowed quadrats of areas I and II, with a
value of 0.22. Wolf spiders in both annually plowed quadrats exhibited low degrees
of similarity with other quadrats. Vagrant spider populations were found to be
most similar in the annually mowed and burned area II quadrats with an index
of 0.94. Jumping spiders were found to be most similar in the annually mowed
and annually burned area II quadrats with a value of 0.72.

The number of species captured in each quadrat ranged from 13 to 31 with
an average of 23 species. The highest spider diversity for the season was found

HASKINS AND SHADDY HABITAT MANIPULATION OF CURSORIAL SPIDERS

5

Table 1. — Spider species represented in the study. Asterisk denotes web-building spiders known to
leave their webs when foraging for prey. (Ann == Annual, Sue = Succession, I and II = areas I and

ID

Adults Captured in Quadrats
Plowed Mowed Burned Control

Ann Sue Ann Sue Ann Sue

1,11 1,11 I, II 1,11 1, 11 1,11 1,11

Wolf spiders
Lycosidae

Lycosa avara (Keyser.)

L. punctulata (Hentz)

L. rahida Walck.

Lycosa sp.

Pardosa saxatillis (Hentz)

Pirata hiteorum Wallace and Exline
P. minutus Emerton
Schizocosa avida (Walck.)

5’. bilineata (Emerton)

S. retrorsa (Banks)

S. saltatrix (Hentz)

Schizocosa sp.

Lycosidae sp.

Running spiders
Clubionidae

Castianeira descript a (Hentz)

C. gertschi Kaston
C. longipalpus (Hentz)

C. variata Gertsch
Clubiona johnsoni Gertsch

C. mixta Emerton
Scotinella similis (Banks)

Scotinella fratella (Gertsch)

Clubionidae sp.

Gnaphosidae

Callilepis pluto Banks
Drassylus depressus (Emerton)

D. fallens Cham.

D. nannellus (Chamber, and Gertsch)

D. rufulus (Banks)

Drassylus sp.

Haplodrassus signifer Koch
Micaria elizabethae (Gertsch)

Sergiolus variegatus (Hentz)

Zelotes hentzi Barrows
Z. inheritus Kaston
Z. laccus (Barrows)

Gnaphosidae sp.

Crab spiders
Philodromidae
Thanatus sp.

Thomisidae
Xysticus sp.

Thomisidae sp.

Jumping Spiders
Salticidae

Agassa cyanea (Hentz)

Corythalia delicatula (Gertsch & Mulaik)

0,

0

0, 0

0,

I

0,

0

0, 0

0,

0

6,

2

2,

0

I, 0

0,

0

1,

0

0, 0

2,

0

3,

3

0,

0

3, 0

0,

0

1,

0

0, 0

I,

0

0,

1

0,

0

2, 0

2,

0

2,

2

2, 0

0,

3

1,

0

0,

10

27,21

3,

I

3,

5

13,18

2,

5

0,

I

0,

0

0, 0

0,

0

0,

1

2, 0

2,

0

4,

0

0,

0

3, 6

0,

3

1,

4

2, 0

3,

3

3,

2

3,

14

20,30

10,44

24,

16

47,46

32,60

11,1

1 1

I,

0

2,17

8,;

21

21,

15

1, 1

17,22

21,1

14

0,

5

6, 2

13,

16

1,

6

14,14

3,

6

7,

7

2,

0

24, 9

0,

3

5,

4

1, 4

2,

6

7,

3

0,

0

0, 0

0,

0

0,

I

0, 0

0,

0

0,

0

0,

3

0, 0

0,

0

3,

0

0, 3

0,

1

4,

I

1,

0

3,

0

0,

0

0,

0

0,

0

1,

I

0,

0

0,

1

2,

0

0,

0

1,

0

0,

0

0,

0

0,

0

0,

0

1,

I

0,

0

0,

0

0,

0

0,

0

0,

0

I,

0

I,

0

0,

0

0,

0

0,

0

0,

I

1,

0

0,

0

1,

0

0,

0

1,

0

0,

0

0,

0

1,

0

0,

0

0,

0

0,

0

0,

1

0,

0

0,

0

0,

0

0,

0

0,

0

0,

1

0,

0

0,

0

0,

0

0,

0

1,

0

1,

1

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

1,

0

0,

0

0,

0

0,

I

0,

0

0,

0

9,

11

1,

2

2,

2

0,

9

4,

11

1,

5

2,

0

2,

I

4,

3

4,

2

2,

0

6,

2

0,

1

2,

0

0,

0

2,

1

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

I,

0

0,

1

0,

0

0,

0

I,

0

4,

I

1,

0

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

0,

1

0,

0

0,

0

1,

0

0,

I

0,

0

0,

0

0,

0

1,

0

0,

0

0,

0

0,

I

0,

0

3,

5

0,

6

4,

3

0,

3

3,

9

2,

0

0,

0

0,

0

0,

0

0,

2

1,

0

I,

I

1,

2

0,

0

0,

0

0,

2

0,

1

0,

1

2,

0

0,

0

0,

0

0,

0

1,

1

0,

0

0,

0

I,

0

0,

0

0,

3

21,

18

7,33

28,23

11,

10

35,40

23,21

I,

0

3,

0

0,

1

I,

2

3,

1

0,

2

2,

0

1, 0

4, 3

4, 1

6,

2

3, 5

9, 4

9, 6

2, 5

22,27

16,18

24,

6

17,15

1 1,10

12,16

0, 2

0, 0

0, 0

0,

0

0, 0

0, 0

0, 0

0, 0

0, 0

1, 0

0,

0

0, 0

1, 0

0, 0

0, 0

0, 0

2, 0

1,

0

0, 0

0, 0

1, 0

6

THE JOURNAL OF ARACHNOLOGY

Table I. — Continued.

Eris aurantia (Lucas)

Evarchi hoyi (Peckham)

Marpissa mucosa (Clerck)
Metaphidippus canadensis (Banks)

M. galathea (Walck.)

Pellenes coecatus (Hentz)

Pellenes sp.

Phidippus clams (Keyser.)

P. me cooki (Peckham)

P. princeps (Peckham)

Sarinda hentzi Banks
Sitticus cursor Barrows
S. striatus Emerton
Talavera minuta (Banks)

Salticidae sp.

Vagrant spiders
Agelenidae

Cicurina cavealis* (Bishop & Crosby)
C. ludoviciana* (Simon)

Hahniidae

H. ononidum* (Simon)

Neoantistea agilis* (Keyser.)

Other guilds
Antrodiatidae

Antrodiaetus stygius (Coyle)
Oxyopidae

Oxyopes salticus (Hentz)

Mimetidae

Mimetus epeiroides Emerton
Ero sp.

Web-builders

0,

0

1,

1

0,

2

0,

1

4,

2

2,

1

0,

1

0,

0

0,

2

0,

3

2,

1

0,

0

0,

1

3,

1

0,

0

1,

0

0,

0

0,

0

0,

0

0,

0

1,

0

0,

1

0,

0

0,

0

2,

2

0,

0

1,

0

3,

0

0,

0

0,

0

0,

1

1,

0

1,

1

1,

0

0,

2

0,

2

0,

1

I,

0

0,

0

2,

1

0,

0

0,

0

I,

0

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

0,

0

1,

0

1,

0

0,

0

2,

0

1,

0

0,

0

0,

0

1,

1

0,

0

0,

0

0,

0

0,

1

I,

0

0,

0

0,

0

0,

1

1,

0

0,

1

0,

0

1,

2

0,

0

0,

1

0,

0

1,

2

0,

0

0,

0

0,

0

0,

0

0,

2

1,

4

3,

3

6,

3

3,

6

2,

I

0,

0

0,

0

0,

0

0,

0

0,

0

1,

0

0,

1

0,

0

0,

1

3,

0

2,

2

6,

1

5,

1

0,

2

0,

0

0,

I

0,

0

1,

0

0,

0

0,

0

0,

0

5, 4

2, 3

2, 0

0, 0

1,

4

1,

0

0, 0

0, 0

0, 1

0, 0

0, 0

0,

0

0,

0

0, 0

0, 0

0, 2

2, 0

2, 0

0,

0

1,

0

25,35

0, 4

13,11

5, 8

6,13

3,

2

7,

6

0, 3

6, 0

2, 0

0, 2

0, 0

0,

1

0,

1

0, 0

0, 1

3, 1

2, 0

4, 6

1,

4

0,

3

0, 0

0, 0

0, 0

0, 0

1, 0

0,

0

0,

0

0, 0

0, 0

0, 0

0, 0

0, 0

0,

0

0,

0

1, 0

Araneidae

30,45

23,36

34,35

28,37

44,34

30,28

46,37

Agelenidae

Agelenopsis kastoni Chamber. & Ivie

0, 0

0, 1

0, 0

0, 0

0, 0

0, 0

0, 0

Agelenopsis sp.

0, 0

0, 1

0, 0

0, 0

0, 0

0, 0

0, 0

Agelenidae sp.

0, 0

0, 0

0, 0

0, 0

1, 0

0, 0

0, 0

Dictynidae

15,37

8, 1

9, 4

4, 1

15, 0

7, 2

0, 1

in the plowed quadrat of area I with an index of 2.720. The lowest diversity
calculated was 2.203 and was in the annually plowed quadrat of area II (Table
4). Differences in diversity values were tested between quadrats. The diversity in
the plowed quadrat of area I was significantly higher than the diversities in the
annually plowed, annually burned, burned, and control quadrats of area I. The
area II plowed quadrat was significantly higher in diversity than the annually
mowed, and burned quadrats of area II.

Community complexity and diversity usually increase during early and mid-
succession serai stages. Arthropods generally increase in number as succession
progresses toward a climax state (Bultman et al. 1982). The succession quadrats
in this study should, therefore, be higher in diversity than the annually manipulated
quadrats. This was observed for the annually plowed (area I and II) and the annually
mowed (area II) quadrats when they were compared to their succession
counterparts. The annual burn quadrats (areas I and II and the area I mowed
quadrat exhibited low diversity values in the spring, but the diversities rapidly

HASKINS AND SHADDY^ HABITAT MANIPULATION OF CURSORIAL SPIDERS

7

Table 2. — Quadrats exhibiting significant differences between seasonal pitfall catches of spiders.
Asterisks denote: * p<0.05, ** p<0.01, *** p<0.00L

Adults Chi-square

Quadrats captured value

lAP

I P

39

183

92.1 12***

IIAP H P

67

172

42.255***

lAP lAM

39

90

19.379***

IIAP HAM

67

178

49.388***

lAP I M

39

156

69.005***

HAP II M

67

130

19.513***

lAP lAB

39

152

65.675***

HAP IIAB

67

155

34.094***

lAP I B

39

153

66.505***

HAP II B

67

197

63.034***

lAP lAC

39

162

74.049***

HAP lAC

67

162

38.585***

lAP I C

39

139

55.061***

IIAP I C

67

139

24.471***

I P lAM

183

90

31.004***

H P H M

172

130

5.566*

I P I C

183

139

5.742*

lAM I M

90

156

17.174***

HAM H M

178

130

7.172**

lAM lAB

90

152

15.376***

lAM

I B

90

153

15.818***

lAM lAC

90

162

20.003***

lAM

I C

90

139

10.061**

HAM

I C

178

139

4.555*

H M H B

130

197

13.321***

IIAB H B

155

197

4.775*

H B I C

197

139

9.670**

Percent Guild Composition of Spider Communities
for the Trapping Season

100

o

g75

o

13

Jso

<

o

1 25

o

0^

Spiders

Wolf

Running

Vagrant

Crab

.Jumping

I I I I I
A P A M A
P M B

I I I II li II II II II

BACAPAMAB
C P M B

Quadrats

Fig. 2. — Percent guild composition of spider communities for the trapping season.

THE JOURNAL OF ARACHNOLOGY

Table 3. — Bray-Curtis similarity values for the cursorial spider community for the trapping season
(0.00 no similarity, 1.00 maximum similarity).

Quadrat

Bray-Curtis

Quadrat

Bray-Curtis

comparisons

values

comparisons

values

<0.60

>0.60

>0.60

lAP I P

0.36

H M 11 B

0.79

IIAP II P

0.48

HAM H B

0.74

lAP I C

0.23

11 P H B

0.72

IIAP I C

0.36

lAM lAB

0.70

I P I C

0.50

H P H M

0.68

II P I C

0.56

HAM HAB

0.67

H P HAM

0.66

lAM I M

0.61

11 P HAB

0.66

HAM II M

0.68

I M I B

0.66

lAM I C

0.57

H M HAB

0.66

HAM I C

0.66

HAP HAB

0.63

I M I C

0.67

1 P lAB

0.63

11 M I C

0.57

I P I M

0.62

lAM I B

0.61

lAB I B

0.64

HAB H B

0.69

lAB I C

0.53

HAB I C

0.54

1 B I C

0.70

II B I C

0.60

increased. By July these diversities

were similar to the diversities observed for their

succession counterparts, and also of the controls. Hence no significant differences

in diversity were observed for these quadrats in

the overall trapping

season.

Shannon diversity indices were

also calculated for vegetation data collected in

area I (Table 5).

Previous studies have shown increased diversity in disturbed

grasslands and old-field habitats for the first one to four years (Hessing et al. 1981

and Bazzaz 1975). Diversity then decreases

somewhat as grasses become the

dominant species

of the habitat.

Increases in

diversity are again

observed after

Table 4. — Shannon diversity and J values for spiders in each quadrat for the trapping season.

Number of

Number of

Quadrat

species

adults

Diversity J component

captured

captured

value (

of evenness

lAP

14

36

2.305

0.873

I P

31

157

2.720

0.792

lAM

19

70

2.604

0.884

I M

29

126

2.601

0.773

lAB

21

132

2.349

0.771

I B

28

133

2.434

0.730

lAC

26

141

2.696

0.827

I C

23

117

2.351

0.750

HAP

13

60

2.203

0.859

11 P

26

144

2.577

0.791

HAM

24

160

2.326

0.732

H M

23

123

2.663

0.849

HAB

24

136

2.375

0.747

H B

23

185

2.240

0.714

HASKINS AND SHADDY— HABITAT MANIPULATION OF CURSORIAL SPIDERS

9

Table 5. — Shannon diversity and J evenness values for vegetation sampled in area 1.

Quadrat

Number of plant
species

Diversity

value

J component
of evenness

May

lAP

14

0.380

0.144

I P

29

2.649

0.787

lAM

24

1.390

0.437

I M

22

2.295

0.743

lAB

19

0.880

0.299

I B

20

2.207

0.737

lAC

10

1.419

0.616

I C

13

1.653

0.645

July

lAP

25

0.992

0.308

I P

32

2.878

0.831

lAM

27

2.475

0.735

I M

31

2.612

0.761

lAB

19

0.715

0.243

I B

24

2.296

0.722

lAC

14

1.700

0.644

I C

19

2.105

0.715

September

lAP

22

1.194

0.386

I P

34

2.965

0.841

lAM

31

2.812

0.819

I M

32

2.859

0.825

lAB

27

1.618

0.491

I B

30

2.753

0.809

lAC

16

2.034

0.734

I C

21

2.289

0.752

woody plants and shrubs become established in the community. This second increase
in diversity is initially rapid and then continues at a slower rate as succession
continues toward a climax state. In this study the most diverse vegetation was
observed during May in the plowed quadrat with a value of 2.649. This quadrat
also had the highest diversities in July and September. July and September diversity
values increased for most of the other quadrats. The diversity of the vegetation
in the succession quadrats was higher throughout the season than the diversity
of the control. Two quadrats, annually plowed and annually burned exhibited lower
vegetation diversities than the control did throughout the season. However the
diversity in the annually mowed quadrat was lower than the control’s diversity
in May, then increased, and was higher than the control’s diversities in July and
September.

Uetz (1975) found no relationship between cursorial spider and vegetation
diversity in a forest. However he observed significant positive correlations between
spider diversity and litter depth. Linear correlation analysis showed a significant
positive correlation (r = 0.747, p < 0.05, d.f. = 6) between the vegetation diversity
and spider diversity observed in May. A negative correlation was observed between
vegetation and spider diversities in all quadrats for the overall season. This
correlation was significant in only three quadrats: mowed (r = —0.864), annually
burned (r = —0.713), and one control (r = —0.726) (p < 0.05, d.f. = 6 in each
case). This was a result of the summer peak in spider abundance and diversity

10

THE JOURNAL OF ARACHNOLOGY

followed by a decline, whereas the plant abundance and diversity continued to
increase throughout the season. No correlations were observed between spider
diversity and grass mat depth.

Linear correlation analyses were made using importance values of plant species,
and the number of spiders captured. These calculations were made using the eight
quadrats from area I. Thirty-four significant correlations were observed between
plant importance and the number of hahniid and lycosid spiders captured. Twenty-
six correlations were observed in May, and eight in July (Table 6). Most of the
correlations occurred in the annually plowed quadrat of area 1. No correlations
were observed in September. This was probably a result of the decline in spider
abundance in the fall. Many of the correlations observed were significant because
the plant species occurred in only one quadrat or its importance value was highest
in that quadrat, and the number of spiders captured was highest in that particular
quadrat. The vegetation in the quadrats exhibiting correlations may be characteristic
of a particular habitat in which the spiders live. Whether or not the spiders were
actually interacting with the plants cannot be determined from the data.

Hahnia ononidum was positively correlated with Phleum pratense and Agrimonia
gryposepala and negatively correlated with Prunella vulgaris. Neoantistea agilis was
positively correlated with nine plants. Five of these plants occurred only in the
plowed quadrat (area I) where the active density of N. agilis was highest. The
plants were: Antennaria neglecta, Astor novae-angeliae, Potentilla simplex,
Solidago graminofolia, Ulmus americana. The four remaining plants: Agrostis alba,
Cerastium vulgatum, Chrysanthemum leucanthemum, and Erigeron strigosus,
exhibited high importance values in IP, and lower importance values in other
quadrats.

Two wolf spiders were positively correlated with plants in May. Pardosa saxatilis
was positively correlated in the annually plowed quadrat (area I) with seven plant
species: Agrostis alba, Antennaria neglecta, Astor novae-angeliae, Panicum
lanuginosum, Potentilla simplex. Prunella vulgaris, and Ulmus americana.
Schizocosa avida was positively correlated with only one plant, Astor sp., during
May. This correlation was observed in seven quadrats. Schizocosa saltatrix was
correlated with six plant species during May. The species were: Agrostis alba,
Antennaria neglecta, Astor novae-angeliae, Potentilla simplex, Ulmus americana,
and Solidago graminofolia. The first four species were also correlated with P
saxatilis and N, agilis.

Significant correlations in July for spiders and plants were observed only in wolf
spiders. Pardosa saxatilis continued to be correlated in the annually plowed area
I quadrat with the importance values of Panicum lanuginosum, Astor novae-
angeliae, and Potentilla simplex. It was also found to be correlated with Solidago
sp. and Composite sp. The latter species was not observed in the May samples.
However, Solidago sp. was present in both May and July. Whether or not the
spider is truly correlated with this species is questionable. Schizocosa avida in July
was positively correlated with Phleum pratense and Veronia baldwinii. Astor sp.
was no longer significantly correlated with S. avida although it was still present
in most of the quadrats. Schizocosa bilineata was also found to be correlated in
several quadrats with Phleum pratense in July.

In this study the spider and flora communities found in fields which were plowed
annually were quite different from those communities found in fields which were
mowed and burned annually. Although some cursorial spiders appeared to be

HASKINS AND SHADDY— HABITAT MANIPULATION OF CURSORIAL SPIDERS

1 1

Table 6. — Significant correlation coefficients from spider-plant linear correlation analyses. Asterisk
denotes p<0.05, d.f. = 6.

Month

Plant

Species

Spider

Species

Correlation

Coefficient

May

Phleum pratense L.

Hahnia ononidum

0.802*

May

Agrimonia gryposepala Wallr.

H. ononidum

0.919*

May

Prunella vulgaris L.

H. ononidum

-0.822*

May

Agrostis alba L.

Neoantistea agilis

0.807*

May

Antennaria neglecta Greene

N. agilis

0.764*

May

Aster novae-angeliae L.

N. agilis

0.764*

May

Cerastium vulgare L.

N. agilis

0.846*

May

Chrysanthemum leucanthemum L.

N. agilis

0.773*

May

Erigeron strigosus Muhl.

N. agilis

0.769*

May

Potentilla simplex Michx.

N. agilis

0.764*

May

SoUdago graminofolia (L.) Salisb.

N. agilis

0.764*

May

Ulmus americana L.

N. agilis

0.764*

May

Agrostis alba L.

Paradosa saxatilis

0.944*

May

Panicum lanuginosum Ell.

P. saxatilis

0.794*

May

Antennaria neglecta Greene

P saxatilis

0.956*

May

Aster novae-angeliae L.

P. saxatilis

0.956*

May

Potentilla simplex Michx.

P saxatilis

0.956*

May

Prunella vulgaris L.

P saxatilis

0.709*

May

Ulmus americana L.

P saxatilis

0.956*

May

Aster sp.

Schizocosa avida

0.794*

May

Agrostis alba L.

S. saltatrix

0.934*

May

Antennaria neglecta Greene

S. saltatrix

0.973*

May

Aster novae-angeliaeh

S. saltatrix

0.973*

May

Potentilla simplex Michx.

S. saltatrix

0.973*

May

SoUdago graminifolia (L) Salisb.

S. saltatrix

0.973*

May

Ulmus americana L.

S. saltatrix

0.973*

July

Panicum lanuginosum Ell.

Pardosa saxatilis

0.750*

July

Aster novae-angeliae L.

S. saltatrix

0.798*

July

Potentilla simplex Michx.

P saxatilis

0.798*

July

SoUdago sp.

P saxatilis

0.798*

July

Composite sp.

P. saxatilis

0.798*

July

Veronia baldwinii Torr.

Schizocosa avida

0.811*

July

Phleum pratense L.

S. avida

0.708*

July

Phleum pratense L.

S. bilineata

0.708*

correlated with the presence of certain plants it should not be concluded that there
is a causal relationship. More study is needed in the area of spider-plant
relationships. Annual plowing results in reduced spider and vegetation diversity.
Mowing and burning treatments that are annually or periodically administered and
periodic plowing may increase spider and vegetation diversity above that of a fallow
control field.

ACKNOWLEDGMENTS

Sincere appreciation is extended to Drs. William Peck and Charles Dondale
for specimen verification. Sincere thanks are extended to Randa Meiser for her
help in field and lab work.

12

THE JOURNAL OF ARACHNOLOGY

LITERATURE CITED

Bazzaz, F. A. 1975. Plant species diversity in old-field ecosystems in southern Illinois. Ecology, 56:485-
488.

Bultman, T, G. W. Uetz, and A. R. Brady. 1982. A comparison of cursorial spider communities along
a successional gradient. J. ArachnoL, 10:23-33.

Connell, J. H. and R. O. Slayter. 1976. Mechanisms of succession in natural communities and their
role in community stability and organization. Amer. Nat., 1 1 1(982): 119-1144.

Cox, G. W., 1980. Laboratory manual of general ecology. 4th edition. William C. Brown Company.
Dubuque, Iowa, 237 pp.

Denslow, J. S. 1980. Patterns of plant species diversity during succession under different disturbance
regimes. Oecologia, 46:18-21.

Duffey, E., 1972. Ecological survey and the arachnologist. Bull. British Arachnol. Soc., 2(5):69-82.

Duffey, E. 1978. Ecological strategies in spiders including some characteristics of species of pioneer
and mature habitats. Symp. Zool. Soc. London, 42:109-123.

Gist, C. S. and D. A. Crossley, Jr. 1973. A method for quantifying pitfall trapping. Environ.
Entomol., 2(5):95 1-952.

Goodall, D. W. 1968. Sample similarity and species correlation. Ordination of plant communities. Dr.
W. Junk B. V. Publishers, The Hague, pp. 99-149.

Greenslade, P. J. M. 1964. Pitfall trapping as a method for studying populations of Carabidae
(Coleoptera). J. Anim. Ecol., 33:301-310.

Hessing, M. B., C. D. Johnson, and R. P. Balder. 1981. Secondary succession of desert grassland
in north-central Arizona, USA, powerline corridors. J. Environ. Management, 13(l):59-70.

Huhta, V. 1971. Succession in the spider communities of the forest floor after clear-cutting and prescribed
burning. Ann. Zool. Fennici, 8:483-542.

Huhta, V. 1979. Evaluation of different similarity indices as measures of succession in arthropod
communities of the forest floor after clear-cutting. Oecologia, 41:1 1-23.

Hulbert, L. C. 1969. Fire and litter effects in undisturbed bluestem prairie in Kansas. Ecology, 50:874-
877.

Kajak, A., A. Breymeyer and J. Petal. 1971. Productivity investigation of two types of meadows in
the Vistula Valley. Ekol. Polska, 19(17):223-233.

Lloyd, M., J. H. Zar and J. R. Karr. 1968. On the calculation of information-theoretical measures
of diversity. Amer. Midi. Nat., 79(2):257-267.

Lowrie, D. C. 1942. The ecology of spiders of the xeric dunelands in the Chicago area. Bull. Chicago
Acad. Sci., 6(9): 161-189.

Lowrie, D. C. 1948. The ecological succession of spiders of the Chicago area dunes. Ecology, 29(3):334-
351.

Merrett, P. 1976. Changes in the ground-living spider fauna after heathland fires in Dorset. Bull. British
Arachnol. Soc., 3(8):214-22L

Murdoch, W., F. C. Evans and C. H. Peterson. 1972. Diversity and pattern in plant and insects. Ecology,
53:819-829.

Pianka, E. R. 1978. Evolutionary Ecology. Second edition. Harper and Row, Publishers. New York,
397 pp.

Pickett, S. T. A. 1976. Succession: an evolutionary interpretation. Amer. Nat., 1 10(971):107-1 19.

Pielou, E. C. 1966a. Shannon’s formula as a measure of specific diversity — its use and misuse. Amer.
Nat., 100:463-465.

Pielou, E. C. 1966b. The measurement of diversity in different types of biological collections. J. Theor.
Biol., 13:131-144.

Riechert, S. E. and W. G. Reeder. 1972. Effects of fire on spider distribution in southwestern Wisconsin
prairies. Proc. Second Midwest Prairie Conf., 1970, 75-90.

Shannon, C. E. 1948. A mathematical theory of communication. Bell Inst. Technol. J., 27:379-423,
623-656.

Southwood, T. R. E. 1966. Ecological methods with particular reference to the study of insect
populations. London: Methuen, 524 pp.

Uetz, G. W. 1974. Species diversity: a review. Biologist, 56:111-129.

Uetz, G. W. 1976. Gradient analysis of spider communities in a stream-side forest. Oecologia, 22:373-
385.

Uetz, G. W. 1979. Influence of variation of litter habitats on spider communities. Oecologia, 40:29-
42.

HASKINS AND SHADDY— HABITAT MANIPULATION OF CURSORIAL SPIDERS

13

Uetz, G. W. and J. D. Unzicker. 1976. Pitfall trapping in ecological studies of wandering spiders. J.
Arachnol., 3:101-111.

Wright, H. and A. Bailey. 1980. Fire ecology and prescribed burning in the great plains. USDA Forest
Service. General Technical Report INT-77.

Zar, J. H. 1974. Biostatistical Analysis. Prentice-Hall, Inc., Englewood Cliffs, N. J., 620 pp.

Manuscript received August 1983, revised December 1984.

Francke, O. F. and J. Boos. 1986. Chactidae (Scorpiones) from Trinidad and Tobago. J. ArachnoL,
14:15-28.

CHACTIDAE (SCORPIONES) FROM
TRINIDAD AND TOBAGO

Oscar F. Francke

Department of Biological Sciences
Texas Tech University
Lubbock, Texas 79409

and

Julius Boos

314 Northeast 26th Drive
Fort Lauderdale, Florida 33334

ABSTRACT

The scorpions belonging to the family Chactidae from Trinidad and Tobago are studied. A new
species, Chactas raymondhansi, from bromeliads on the highest mountains of Trinidad is described.
The type species of the genus Broteochactas Pocock, Broteochactas nitidus Pocock from Trinidad,
is resurrected from its synonymy under Broteochactas gollmeri (Karsch) from Venezuela. The genus
Auyantepuia Gonzalez-Sponga, from Venezuela, is synonymized under Broteochactas.

INTRODUCTION

Trinidad and Tobago are islands located on the continental shelf off the
northeast coast of Venezuela. Their proximity to the mainland, and their location
opposite the mouth of the Orinoco River have been used to hypothesize an
Orinoco-Guianan origin for their biota via rafting. Their scorpion fauna,
however, does not support this hypothesis at all. Kjellesvig-Waering (1966) noted
that the affinities of the scorpion fauna of Trinidad and Tobago were with the
Coastal Ranges of Venezuela to the east, rather than with the Orinoco River delta
to the south. Only one of the seven scorpion species reported from Trinidad by
Kjellesvig-Waering (1966) was regarded as an endemic; four were shared with
Venezuela and two with the Guianas. Likewise, one of the three species reported
from Tobago was considered an endemic; the other two were shared with both
Trinidad and Venezuela.

Considerable work done on the Venezuelan fauna during the past 20 years by
GonzMez-Sponga (summarized in his 1984 contribution) and additional collecting
in Trinidad and Tobago have changed the picture considerably. The three
chactids found in Trinidad and Tobago, including one new species described
below, are endemic. Two of the buthids (Microtityus rickyi Kjellesvig-Waering,

16

THE JOURNAL OF ARACHNOLOGY

and Tityus trinitatis Pocock) are now also considered endemic to the islands, and
four others are shared with the mainland. The high percentage of scorpion
endemism (55%) in Trinidad and Tobago reinforces the hypothesis put forth by
Kjellesvig-Waering (1966) that the Venezuelan Coastal Ranges and the Trinidad
Northern Range were continuous at one time; subsequently, faulting and erosion
during Pliocene or Pleistocene times led to the isolation of the island fauna. The
lack of shared species between the Orinoco River delta and Trinidad and Tobago
indicates that rafting is an unlikely dispersal mechanism in scorpions in general.

The terminology used herein follows essentially that of Stahnke (1970), except
for trichobothriotaxy (Vachon 1974), and metasomal carinae (Francke 1977).
Acronyms for specimen depositories are given in the acknowledgments, except for
those in the authors’ collections which are followed by their initials.

FAMILY CHACTIDAE

The systematics of this worldwide scorpion family are chaotic and unreliable.
The diagnostic characters given below will readily separate it from the Buthidae,
the only other family found in Trinidad and Tobago. These characters are also
useful for northeastern South America and the Caribbean region, as the families
Bothriuridae, luridae and Vaejovidae do not occur there naturally.

Diagnosis. — Sternum pentagonal rather than subtriangular. Tibial spurs absent,
prolateral and retrolateral pedal spurs present. Cheliceral fingers without ventral
teeth, movable finger with distal ventral tine considerably longer than the dorsal
tine. Pedipalps with trichobothrial pattern C of Vachon (1974): femur with three
trichobothria, tibia with 5-7 ventral trichobothria. Telson without a ventral tooth
under the sting.

Genus Chactas Gervais

This genus is known from Central America (Costa Rica and Panama) and
South America (as far south as Peru and Brazil). It differs from Broteochactas
Pocock, the only other genus in this family known from Trinidad and Tobago,
by having 5-6 instead of 7 ventral trichobothria on the tibia.

Five subgenera are presently recognized within the genus Chactas (Francke
1985). The new species described below belongs to the subgenus Andinochactas
Gonzalez-Sponga, characterized by having the tarsi armed with a ventral median
row of spines, and five ventral trichobothria on both the tibia and the chela of
the pedipalps.

Chactas (Andinochactas) raymondhansi, new species
Figs. 1-10

Type data. — Holotype male found on a bromeliad along the eastern ridge near
the summit of Mt. El Tucuche (1000 m), Northern Range, Trinidad, 26-VI-1982
(Betty Faber, Eddie Rooks and party). Deposited in the American Museum of
Natural History, New York. Paratypes listed under specimens examined.

Distribution. — Known only from bromeliads on the highest mountains on the
Northern Range of Trinidad.

FRANCKE AND BOOS— CH ACTIDS FROM TRINIDAD AND TOBAGO

17

5 6 7

Figs. 1-7. — Right pedipalp of holotype male of Chactas (Andinochactas) raymondhansi, new
species, showing trichobothrial patterns: 1, dorsal aspect of femur; 2, dorsal aspect of tibia; 3, external
aspect of tibia; 4, ventral aspect of tibia; 5, dorsal aspect of chela; 6, external aspect of chela; 7,
ventral aspect of chela.

18

THE JOURNAL OF ARACHNOLOGY

8 9 10

Figs. 8-10. — Tarsomeres on first right leg of holotype of Chactas (Andinochactas) raymondhansi,
new species, showing diagnostic curved spiniform seta prolaterally: 8, dorsal aspect; 9, prolateral
aspect; 10, ventral aspect.

Etymology. — This species is dedicated to Raymond A. Mendez and Hans E.
A. Boos, joint collectors of the first specimen seen by the junior author. That
specimen started the search that resulted in this publication.

Diagnosis. — Adult males 55-60 mm long, females 60-65 mm. Pedipalp tibia
(Fig. 4) and chela (Fig. 7) with five ventral trichobothria each. Pectinal tooth
counts 8-11. Tarsomere I of leg I with a strongly developed, curved spiniform seta
prolaterally (Figs. 8-10). Tarsi on all legs with 5-8 ventral median spines.

Male. — Measurements (lengths in mm) of holotype: carapace 9.5, metasomal
segment V 8.8; pedipalp femur 11.5, tibia 11.5, chela 20.5. Opisthosoma,
metasoma, chelicera and pedipalps dark red brown; sternites, legs and telson
yellow brown. Carapace and tergites acarinate. Pectinal tooth counts 9-11 (mode
= 9). Stigmata large, suboval. Sternite VII acarinate. Metasomal segments I-IV:
dorsolateral carinae weak, granulose; lateral supramedian carinae moderate,
granulose; lateral inframedian carinae on I weak, granulose, on II-IV obsolete;
ventrolateral keels weak, smooth; ventral submedian keels vestigial, smooth.
Segment V; dorsolateral and lateral median carinae vestigial to obsolete;
ventrolateral and ventromedian carinae weak, sparsely granulose. Pedipalps
elongate, femur and tibia longer than carapace, chela slightly over twice as long
as carapace. Trichobothrial pattern C (Vachon 1974): femur orthobothriotaxic

FRANCKE AND BOOS— CH ACTIDS FROM TRINIDAD AND TOBAGO

19

with 3 trichobothria (Fig. 1); tibia neobothriotaxic with 1 internal, 2 dorsal, 17
external, and 5 ventral trichobothria (Figs. 2-4); chela neobothriotaxic with 27
trichobothria, 5 ventrals (Figs. 5-7).

Female. — Measurements (lengths in mm) of largest female: carapace 10.3,
metasomal segment V 8.8; pedipalp femur 9.0, tibia 9.0, chela 18.6. Sternites, legs
and telson medium brown. Pectinal tooth counts 8-10 (mode = 9). Pedipalps not
as elongate as on male: femur and tibia shorter than carapace, chela less than
twice as long as carapace. Curved spiniform seta on tarsomere I of leg I not as
strongly developed as on male, although still very conspicuous. Other characters
as on male.

Variability. — Male sexual dimorphism in pedipalp length appears with the
penultimate molt; immature males resemble females morphometrically. Pectinal
tooth counts vary as follows: in males 14 combs with nine teeth, 1 comb with
ten teeth, and 1 comb with eleven teeth; in females 2 combs with eight teeth, 1 1
combs with nine teeth, and 6 combs with ten teeth. Finally, the diagnostic, curved
spiniform seta on tarsomere I is present in the smallest specimens examined (a
male and a female with carapace lengths of 2.9 and 3.0 mm, respectively;
probably represent second or third instars).

Comparisons. — The new species differs from Chactas (Andinochactas) gestroi
Kraepelin, the only other member of the subgenus, (1) by having a strong, curved
spiniform seta on tarsomere I of leg I, (2) by lacking dense punctations on the
metasoma, and (3) by having higher pectinal tooth counts on both sexes (range
8-1 1 and mode is 9 in C. raymondhansi; range 6-8 and mode is 7 in C. gestroi).

Remarks. — This species has been collected only from water-filled spaces
between leaf sheaths of the bromeliad Glomeropitcairnia erectiflora Mez. On live
specimens a large chamber extending from the stigmata and surrounding the
book lungs can be seen. This chamber may act as a reservoir for air if
summersion becomes necessary or unavoidable.

The same species of bromeliad occurs on the Paria Peninsula and Margarita
Island, Venezuela (Smith and Pittendrigh 1967). A search of this specific habitat
in those locations might reveal the presence of C. raymondhansi, or a close
relative, on the mainland.

Specimens examined. — TRINIDAD: Northern Range, on east ridge near summit of Mt. El Tucuche
(1000 m), in bromeliad, 26-VI-1982 (Betty Faber, Eddie Rooks and party), holotype male (AMNH);
same locality, in bromeliad 10 feet up small tree, 25-1-1981 (Raymond A. Mendez and Hans Boos),
one adult male paratype (JB); same locality, in bromeliad, 16-V-1982 (R. W. Bruce, Jr. and L.
Downie), one adult female paratype (JB); same locality, in bromeliad, 24-VI-82 (John Murphy and
Ron Humbert), one adult male paratype (FSCA); same locality, in bromeliad, VIII-1982 (M. Read),
one subadult female paratype (FSCA); same locality, in bromeliad, IX-1982 (M. Read), one subadult
male paratype (MAGS); same locality, in bromeliad, 1980 (M. Koo), one juvenile male paratype
(UWI); El Tucuche (3,000 ft.), 30 October 1937 (E. McC. Callan), one adult male paratype (USNM);
Northern Range, top of Mt. Aripo, in bromeliads, 9-X-1982 (M. Read and J. Seyjagat), one adult
male, one adult female, one subadult female, and one juvenile male paratypes (OFF); on ridge 1 mi
W of Mt. Aripo, XI-1982 (M. Read), one juvenile female paratype (JB); El Aripo (2,800 ft.), 25
March 1942 (C. G. Pittendrigh), one adult female paratype (USNM); El Aripo (3,085 ft.), 24 February
1944 (A. M. Adamson), one adult female paratype (USNM); Northern Range, summit of Morne Bleu,
in Glomeropitcairnia bromeliad, 4 March 1984 (M. Read), one juvenile female paratype (JB); St.
Augustine, 1941 (E. McC. Callan), one subadult female paratype (USNM).

20

THE JOURNAL OF ARACHNOLOGY

Genus Broteochactas Pocock

Broteochactas Pocock, 1893. Ann. Mag. Nat. Hist., ser. 6, 12:77 (Type species Broteochactas nitidus

Pocock 1893 by original designation).

Auyantepuia Gonzalez-Sponga, 1978. Escorpiofauna de la Region Oriental del Estado Bolivar, en

Venezuela, p. 75. (Type species Broteochactas scorzai Dagert 1957, by monotypy). NEW

SYNONYMY.

This genus is known from Panama, Colombia, Ecuador, Venezuela, the
Guianas, Brasil, and Trinidad and Tobago. The type species, originally described
from Trinidad, was synonymized under Broteochactas gollmeri (Karsch) from
Caracas, Venezuela, by Kraepelin in 1894, and has remained in synonymy ever
since. Comparison of material from Caracas, including the lectotype of B.
gollmeri (hereby designated), with specimens from both Trinidad and Tobago
revealed that these two taxa are not conspecific. Thus, B. nitidus is redescribed
below.

The genus Auyantepuia was created by Gonzalez-Sponga to accommodate
Broteochactas scorzai Dagert, from Estado Bolivar, Venezuela. This small species
differs from other Broteochactas spp. known from Estado Bolivar in three
characters, which is why it was placed in a separate genus. Eirst the dentition
in the pedipalp fingers terminates in two distal teeth, rather than three to five.
Second, the trichobothrium designated eb on the external series of the pedipalp
chela finger (Vachon 1974, Gonzalez-Sponga 1978) is found on the finger proper,
rather than on the base of the finger basal to the commissure. Einally, among
the supernumerary trichobothria on the external distal series on the tibia, et4 is
“petite” as opposed to ets and et6 being “petite.” When all species of
Broteochactas are considered, rather than only those from Estado Bolivar, B.
scorzai no longer stands out as a generically distinct entity.

The variation in the first character mentioned above appears to be size-related,
with smaller species having fewer terminal teeth than larger species. Broteochactas
scorzai and Broteochactas laui Kjellesvig-Waering, the two smallest species in the
genus, have two terminal teeth on both the fixed and movable fingers of the
pedipalp chela. Broteochactas nitidus and B. gollmeri, two medium sized species
have two terminal teeth on the fixed finger and three on the movable finger.
Finally, some of the larger Broteochactas spp. have three terminal teeth on the
fixed finger, and three to five on the movable finger. Thus, this character is not
taxonomically significant at the supraspecific level.

Regarding the second character used to diagnose Auyantepuia, B. nitidus and
B. gollmeri have trichobothrium eb in the same position as B. scorzai, on the
fixed finger proper, whereas B. laui and some of the larger Venezuelan
Broteochactas spp. have eb on the basal portion near the fingers’ commissure.
Thus, this character does not have the taxonomic significance ascribed to it by
Gonzalez-Sponga.

The third “character” mentioned above actually involves two different,
subjective problems, each of which is dealt with separately. First, a decision must
be made as to what constitutes a normal versus a “petite” trichobothrium. Among
the six trichobothria assigned to the external terminal (et) tibial series in
Broteochactas, some species have one with a small pit and short hair, and
another one with a normal pit but with a short hair (e.g., B. laui. Fig. 11); some

FRANCKE AND BOOS— CH ACTIDS FROM TRINIDAD AND TOBAGO

21

13

Figs. II-I3. — External aspect of pedipalp tibia (drawn on same scale) showing trichobothrial
variability in Broteochactas spp. An open circle denotes a normal trichobothrium, a half-closed circle
denotes a trichobothrium with a normal pit and a small seta, and a small dot denotes a
trichobothrium with a small pit and a small seta, ll, .fi. laui, from Tobago; 12, B. gollmeri, from
Venezuela; 13, Broteochactas sp. from Guyana.

Fig. 14. — Semi-diagrammatic illustration showing that positional variation in one small
trichobothrium can lead to erroneous homologies in designations: a = et4 and x = ets, however c =
ets and x = et4 (see text for further discussion).

species have two trichobothria with small pits and short hairs (e.g., B. gollmeri,
Fig. 12), and some have two with normal pits but short hairs (e.g., Broteochactas
sp.. Fig. 13). The second problem is that created by the numbering convention
of Vachon (1974), and used by GonzMez-Sponga (1978), and its implied
homologies (see Francke 1982, and Francke and Soleglad 1981, for further
discussion on this problem). In Fig. 14, a small trichobothrium appearing in
position “a” becomes et4, and “x” becomes ets; the same small trichobothrium in
position “c” becomes ets, and “x” is considered as et4; and in position “b” it is
a toss-up! The reputed difference between Auyantepuia, with trichobothria et4
small and ets normal (position “a” in Fig. 14), and Broteochactas, with et4 normal
and ets small (positions “b” and “c” in Fig. 14) is based on a false homology and
certainly does not justify the recognition of two separate genera.

Broteochactas nitidus Pocock
Figs. 15-20

Broteochactas nitidus Pocock 1893b:339-401, pi. 29, figs. 1-1 di.

Broteochactas gollmeri (in part): Kraepelin 1894:176-177, 1899:173, 1912:53; Pocock 1897:365-366,
1900:68; Mello-Leitao 1932:32, 1945:100.

Broteochactas gollmeri (misidentification): Roewer 1943:237; Waterman 1950:169; Kjellesvig-Waering
1966:126.

22

THE JOURNAL OF ARACHNOLOGY

Fig. 15-20. — Right pedipalp of adult male Broteochactas nitidus Pocock, from Trinidad, showing
trichobothrial patterns: 15, dorsal aspect of femur; 16, dorsal aspect of tibia; 17, external aspect of
tibia; 18, dorsal aspect of chela; 19, external aspect of chela; 20, ventrointernal aspect of chela.

FRANCKE AND BOOS— CHACTIDS FROM TRINIDAD AND TOBAGO

23

Type data. — Lectotype male and seven paralectotypes (hereby designated) from
Trinidad (L. E. Broadway). Permanently deposited in the British Museum
(Natural History). Examined.

Distribution. — Known only from Trinidad and Tobago.

Diagnosis. — Adults 30-35 mm long; dark brown to black, legs medium to dark
brown. Pectinal tooth count 8 in males, 7 in females. Metasomal segments III-
IV with ventrolateral keels moderate, granulose. Pedipalp chela with three
external trichobothria on finger (Fig. 18); dorsal margin of manus with dense,
coarse granulation; external aspect of manus with fine granules in reticulated
pattern. Chela length less than twice its width.

Male. — Measurements (lengths in mm) of lectotype male from Trinidad:
carapace 4.1, metasomal segment V 4.7; pedipalp femur 2.8, tibia 3.3, chela 6.3,
fixed finger 2.6. Opisthosoma, metasoma and pedipalps dark red brown; sternites,
chelicera and legs light brown; all areas heavily infuscate. Carapace and tergites
shagreened, acarinate except for tergite VII with few distal granules on
submedian and lateral carinal regions. Pectinal tooth count 8. Stigmata small,
round. Sternites smooth, acarinate. Metasomal segments I-IV; dorsolateral
carinae strong, granulose; lateral supramedian carinae moderate, granulose;
lateral inframedian carinae on I vestigial, subgranose, on II-IV obsolete;
ventrolateral carinae on I obsolete, on II vestigial and subgranose, on III and IV
weak to moderate and granulose; ventral submedian carinae on I and II obsolete,
on III vestigial and subgranose, on IV weak to moderate and granulose. Segment
V: dorsolateral and lateromedian carinae weak, finely granulose; ventral keels
moderate, granulose. Telson with moderately dense, small granulation. Pedipalps
stout, chela considerably wider (1.3-1.5X) than fixed finger length. Bothriotaxia
C (Vachon 1974): femur orthobothriotaxic with 3 trichobothria (Fig. 15); tibia
neobothriotaxic with 1 internal, 2 dorsal, 24 external and 7 ventral trichobothria
(Figs. 16 and 17); chela orthobothriotaxic with 26 trichobothria (Figs. 18-20).
Tarsomere II on all legs armed ventrally with two submedian rows of long setae.

Female. — Measurements (length in mm) of adult from Chancellor Hill,
Trinidad: carapace 5.0, metasomal segment V 5.0; pedipalp femur 3.5, tibia 4.0,
chela 7.6, fixed finger 2.9. Pectinal tooth count 7.

Variability. — The populations from Trinidad and Tobago are remarkably
uniform morphologically. However, specimens from mountainous areas tend to
be darker (almost black) than lowland forms. Pectinal tooth counts display very
low variability: in males 7 combs had seven teeth, 68 combs had eight teeth, and
1 comb had nine teeth; in females 2 combs had six teeth, 185 combs had seven
teeth, and 9 combs had eight teeth.

Comparisons. — Broteochactas nitidus differs from B. gollmeri, with which it
had been confused, as indicated below. First, in size: carapace length in adult
males is greater than 4.0 mm in B. nitidus, and less than 4.0 mm in B. gollmeri;
in females greater than 4.5 mm in B. nitidus, and less than 4.1 mm in B. gollmeri.
Second, in pectinal tooth counts: in B. nitidus 8 in males, and 7 in females; in
B. gollmeri 1 in males, and 6 in females. Third, in ornamentation on the dorsal
and external faces of the pedipalp chela: densely granulose in B. nitidus, smooth
and shiny in B. gollmeri. Fourth, in pedipalp chela proportions: 1.9 times longer
than wide in males of B. nitidus, 2.1 times longer than wide in males of B.
gollmeri; 1.8 times longer than wide in females of B. nitidus, 2.15 times longer
than wide in females of B. gollmeri. Finally, in development of the ventral

24

THE JOURNAL OF ARACHNOLOGY

submedian carinae on metasomal segments III and IV: weak to moderate and
granulose on B. nitidus, vestigial and subgranose on B. gollmeri.

Broteochactas nitidus differs from B. laui, with which it coexists on Tobago,
as follows. First, in size: carapace length in adults of B. laui is less than 4.0 mm.
Second, in pectinal tooth counts: 7 on males and 6 on females of B. laui. Third,
in the disposition of the external trichobothria along the fixed finger of the
pedipalp chela: eb basal to the finger commissure in B. laui (Fig. 25), and distal
to the commissure in B. nitidus (Fig. 19).

Specimens examined. — TRINIDAD: no date [29. IV. 1890 ?] (L. E. Broadway), lectotype male, two
paralectotype males, and five paralectotype females (BMNH); 3.4.1900 (W. Ince), one male, five
females (BMNH); Maraval Valley, 1913 (Thaxter), two females (MCZ); Port-of-Spain, Jan. 1913 (R.
Thaxter), two juveniles (MCZ); near Salibia, 28 April 1916 (H. L. Clark), one female (MCZ); La
Seira Valley, April 1916 (H. L. Clark), two males, two females (MCZ); Mt. Tucuche (in bromeliads),
28 May 1917 (J. B. Rorer), one female (MCZ); Port-of-Spain, 28 Jan. 1926 (W. S. Brooks), one male
(MCZ); Port-of-Spain, February 1942 (H. F, Loomis — A. V. Armour Exp.), two adult females
(USNM); Port-of-Spain, Chancellor Hill (1000 ft.), 20 March 1964 (E. N. Kjellesvig-Waering), three
adult males, one subadult female (FSCA); Simla, 18-25 April 1964 (no collector), four juveniles
(MCZ); Port-of-Spain, 15-V-1965 (E. N. Kj. Waering), two adult, one juvenile females (FSCA); Port-
of-Spain, Chancellor Hill (1000 ft.), 15-VH-1965 (E. N. Kj. Waering), four males, five females
(FSCA); Port-of-Spain, 1 May 1966 (E. N. Kj. Waering), one juvenile (FSCA); Port-of-Spain,
Chancellor Hill (1500 ft.), Aug. 1966 (E. N. Kjellesvig-Waering), one adult male, three adult females,
two subadult females (OFF), three males, one female (FSCA); Port-of-Spain, Ft. George Hill (1200
ft.), 15-VHI-1967 (Ricky Kj. Waering), one juvenile male (FSCA); Port-of-Spain, Chancellor Hill
(1500 ft.), March 1968 (E. N. Kj. Waering), four females (FSCA); Maraval, N of Port-of-Spain, 16
Dec. 1962 (E. N. Kj. Waering), one female (FSCA); Maraval, 4 Mar. 1963 (E. N. Kj. Waering), one
female (FSCA); 5 Mar. 1964 (W. R. Jones), one male, three females (one with 46 first instars)
(FSCA); Maraval, 6 March 1964 (W. R. Jones), one female (gave birth in captivity) FSCA); Maraval,
13 Mar. 1964 (W. R. Jones), two females (FSCA); Maraval, 12 April 1964 (E. N. Kj. Waering), one
male, one female (FSCA); Maraval, 15 April 1964 (E. N. Kj. Waering), one female (FSCA); Maraval,
23 April 1964 (E. N. Kj. Waering), one male (FSCA); Maraval, 12-1-1968 (E. N. Kj. Waering), one
female (with 27 first instars born in captivity) (FSCA); Maraval, Fondes Amandes Road, 2 Dec. 1967
(H. and J. Boos), four females (one with 44 first instars born in the wild) (FSCA); Maraval, Fondes
Amandes Rd., 24-V-1968 (J. Boos), one female (FSCA); Mayaro Beach, 2 mi N Plaisance (2 m), 15
Mar. 1964 (E. N. Kj. Waering), one female (FSCA); Mayaro Beach area, Ortoire, 12 June 1964 (E.
N. Kj. Waering), one female (FSCA); Mayaro Beach area, Ortoire Village, 20 Nov. 1964 (E. N. Kj.
Waering), one female (FSCA); Mayaro Beach, 15 Nov. 1965 (E. N. Kj. Waering), one male, one
female (FSCA); Mayaro Beach, 1 May 1966 (E. N. Kj. Waering), one female (FSCA); Mayaro,
Ortoire Village, 111-1968 (E. N. Kj. Waering), one female (FSCA); Mayaro Beach, Ortoire Village,
15-IV-1968 (E. N. Kj. Waering), three females (FSCA); Mayaro Beach, 15-V-68 (E. N. Kj. Waering),
one male (FSCA); Maracas, 1941 (E. McC. Callan), one adult female (USNM); Maracas Valley, 9-
IV- 1968 (Hans Boos), one male, one female (FSCA); milepost 6.4 Maracas Bay to Port-of-Spain Rd.
(800 ft.), 15-VI-1968 (Ricky Waering), one female (FSCA); milepost 6-1/4 on Maracas Bay Road,
20-VIII-1976 (J. Boos), one adult male (JB); Mt. Cathrine, 13-11-1977 (J. Boos), one adult male, one
juvenile female (JB); St. Joseph, Maracas Falls trail, 8-HI-1977 (J. Boos), two adult females (OFF);
Asa Wright Nature Center, Arima Valley, 16-VII-1978 (R. A. Mendez), one adult female (OFF);
Blanchisseuse Valley, l-H-1979 (R. Mendez and J. Boos), one adult male, one subadult female, one
immature male (OFF); El Tucuche (2,000 ft.), June 1944 (A. H. Strickland), one adult female with
27 first-instar young (USNM); Tucuche (2,500 ft.), 12 November 1944 (A. M. Adamson), one adult
and one juvenile males (USNM); Mt. Tucuche (3000 ft.), l-VI-1969 (Julius Boos), one male (FSCA);
trail to Mt. El Tucuche (under rock), 18-11-1979 (J. Boos), one adult male, one subadult female (JB);
in leaf litter nr. summit of Mt. El Tucuche, N. Range, 16-V-1982 (R. W. Bruce, Jr. and L. Downie),
two adult females (MAGS); Verdant Vale Road, off Blanchisseuese Rd., Arima, July 1982 (M. Read),
one adult female (JB); Simla, Blanchisseuse Rd., Arima, Sept. 1982 (M. Read), one adult female
(UWI); in bromeliad, summit of Mt. Aripo, Northern Range, IX-1982 (M. Read), one female (UWI);
Haven Hill farm, Talparo, Nov. 1983 (V. Quesnel), one juvenile female (JB); in rotting log, summit
of Mt. Aripo, Northern Range, 14 Dec. 1983 (M. Read), one adult female (JB); Northern Range
(2,000 ft.), 8 February 1942 (A. M. Adamson), one adult female (USNM); same locality (2,000 ft.).

FRANCKE AND BOOS CHACTIDS FROM TRINIDAD AND TOBAGO

25

24 March 1943 (A. M. Adamson), one adult male (USNM); Diego Martin, Petit Valley, 5 May 1964
(E. N. Kj. Waering), one male (FSCA); St. Anns, 12-XI-1968 (J. Boos), two females (FSCA); 3 mi
W Morne Bleu, 30-VI-1968 (J. Boos), one female (FSCA); 1 mi W Morne Bleu, 9-III-1969 (J. Boos),
two females (FSCA). TOBAGO: Speyside, I Sept. 1964 (Edgar Lau), one male (FSCA); Speyside,
28 Nov. 1969 (Egbert Lau), one female (FSCA); Speyside, .7.1970 (E. N. Kj. Waering), one female
(FSCA); road to Hillsborough Dam, May 1977 (Ray Mendez), one adult female (OFF); Mertichon
Estate nr. Speyside, 3-1V-1979 (Dave Hardy), one subadult female (JB); Pigeon Hill, nr.
Charlotteville, 4-V-I979 (Dave Hardy), one adult female (JB); Pigeon Peak, nr. Charlotteville, 26-27-
X-1982 (M. Read), two adult females (JB); near summit of Pigeon Peak, under stone in gulley, 26
Feb. 1984 (M. Read), two adult females (OFF).

Broteochactas laui Kjellesvig-Waering
Figs. 21-26

Broteochactas laui KjQWesvig-Waering 1966: 126-128, figs. 1-2; Gonzalez-Sponga 1974a:5, 1974b:300.

Type data. — Holotype female from Speyside (Paradise Inn Hotel), Tobago.
Permanently deposited in the Ameriean Museum of Natural History, New York.
Not examined.

Distribution. — Known only from Tobago.

Diagnosis. — Adults 22-25 mm long; medium brown and heavily infuscate, with
legs yellow brown and lightly infuscate. Pectinal tooth counts 6-7 (mode = 7) in
males, 5-7 (mode = 6) in females. Pedipalp chela with two external trichobothria
on finger (Fig. 25); dorsal and external aspects of chela densely and coarsely
granulose. Carinae on metasomal segments I-IV absent, except for lateral
supramedians on I and II and dorsolaterals on III and IV which are weak and
subgranose.

Female. — Measurements (length in mm) of adult paratype from Speyside,
Tobago: carapace 3.7; metasomal segment V 3.1; pedipalp femur 2.1, tibia 2.6,
chela 5.0, fixed finger 1.9. Opisthosoma and metasoma medium brown, heavily
infuscate; pedipalps dark red brown, heavily infuscate; sternites, chelicera and legs
yellow brown, lightly infuscate. Carapace and tergites acarinate, smooth and
shiny. Sternites smooth and shiny; stigmata small, round. Metasomal carinae on
I-IV: lateral supramedians on I and II, and dorsolaterals on HI and IV weak,
subgranose; all others obsolete. Segment V with dorsolateral keels weak, finely
granulose; ventral keels moderate, granulose. Telson shiny, with sparse small
granules. Pedipalps stout; chela with broad, stubby fixed finger. Bothriotaxia C
(Vachon 1974): femur orthobothriotaxic with 3 trichobothria (Fig. 21); tibia
neobothriotaxic with 1 internal, 2 dorsal, 24 external, and 7 ventral trichobothria
(Figs. 22 and 23); chela orthobothriotaxic with 26 trichobothria (Figs. 24-26).
Tarsomere II on all legs armed ventrally with two submedian rows of setae.

Male. — Measurements (lengths in millimeters) of adult paratype from Speyside,
Tobago: carapace 3.5; metasomal segment V 3.5; pedipalp femur 2.2, tibia 2.7,
chela 4.7, fixed finger 1.7. Differ from females primarily in pedipalp chela
proportions, having a wider manus and a shorter fixed finger; in pectinal
morphology, with one more tooth, and each tooth is 2.5 to 3 times longer than
on females.

Variability. — Pectinal tooth counts varied as follows: on males 1 comb with six
teeth, 3 combs with seven teeth (one male reported upon by Kjellesvig-Waering
1966); on females 1 comb with five teeth, 22 combs with six teeth, and 7 combs
with seven teeth.

26

THE JOURNAL OF ARACHNOLOGY

Figs. 21-26. — Right pedipalp of adult female Broteochactas laui Kjellesvig-Waering, from Tobago,
showing trichobothrial patterns: 21, dorsal aspect of femur; 22, dorsal aspect of tibia; 23, external
aspect of tibia; 24, dorsal aspect of chela; 25, external aspect of chela; 26, ventrointernal aspect of
chela.

FRANCKE AND BOOS— CH ACTIDS FROM TRINIDAD AND TOBAGO

27

Comparisons. — See the section under B. nitidus for differential characters with
that species.

Specimens examined. — TOBAGO: Speyside (under rocks in burrows; forest), 15-11- 1 965 (Edgar
Lau), one adult male, one adult female paratypes (FSCA); Speyside, 20 April 1965 (Edgar Lau), two
adult female paratypes (FSCA); between Charlotteville and Speyside, 29 Oct. 1982 (M. Read), one
adult female (JB); milepost 26 Windward road, among rocks and soil from walls of gulley in cocoa
plantation, 19-20 Feb. 1984 (M. Read), five adult, one subadult and three juvenile females (JB, OFF).

ACKNOWLEDGMENTS

We thank Dr. Mary Alkins, Department of Zoology, University of the West
Indies, St. Augustine, Trinidad (UWI), Dr. J. Coddington, U. S. National
Museum, Washington, D. C. (USNM), Dr. G. B. Edwards, Florida State
Collection of Arthropods, Gainesville, Florida (FSCA), Mr. P. D. Hillyard,
British Museum (Natural History), London (BMNH), Dr. H. W. Levi, Museum
of Comparative Zoology, Harvard University, Cambridge (MCZ), and Dr. M.
Moritz, Zoologisches Museum der Humboldt Universitat, Berlin (ZMB) for the
loan of specimens. Dr. M. A. Gonzalez-Sponga (MAGS) kindly donated some
Venezuelan scorpions to the senior author, and gave advice and loaned specimens
to the junior author. Special thanks to the many collectors of specimens who are
mentioned in the text, in particular Dr. Morley Read and Mr, John Seyjagat;
and to Mr. J. Santiago-Blay for locating specimens in the collection of the U.
S. National Museum. Finally, Dr. W. D. Sissom and Mr. S. A. Stockwell kindly
commented on a draft of the manuscript.

LITERATURE CITED

Francke, O. F. 1977. Scorpions of the genus Diplocenirus from Oaxaca, Mexico. J. ArachnoL, 4:145-
200.

Francke, O. F. 1982. Studies of the scorpion subfamilies Superstitioninae and Typhlochactinae, with
description of a new genus (Scorpiones, Chactoidea). Bull. Assn. Mexican Cave Studies, 8:51-62.
Francke, O. F. 1985. Conspectus Genericus Scorpionorum (Arachnida, Scorpiones). Occas. Papers
Mus., Texas Tech Univ., No. 98, 32 pp.

Francke, O. F. and M. E. Soleglad. 1981. The family luridae Thorell (Arachnida, Scorpiones). J.
ArachnoL, 9:233-258.

Gonzalez-Sponga, M. A. 1974a. Broteochactas sanmartini: nueva especie del Rio Caura, en la
Guayana de Venezuela. Monografias Cientificas “August© Pi Suner”, Caracas, No. 7, 16 pp.
Gonzalez-Sponga, M. A. 1974b. Una nueva especie de escorpiones de la Amazonia Venezolana
(Chactidae). Acta Biol. Venezuela, 8:299-313.

Gonzalez-Sponga, M. A. 1978. Escorpiofauna de la region oriental del Estado Bolivar, en Venezuela.

Consejo Nac. Inv. Cient. Tecnol. Caracas, Roto-Impresos C. A., 217 pp.

GonzMez-Sponga, M. A. 1984. Escorpiones de Venezuela. Cuadernos Lagoven, Cromotip, Caracas,

126 pp.

Kjellesvig-Waering, E. N. 1966 The scorpions of Trinidad and Tobago. Caribbean J. Sci., 6:123-135.
Kraepelin, K. 1894. Revision der Skorpione. II: Scorpionidae und Bothriuridae. Jahrb. Hamburg,
wiss. Anst., 1 1:1-248.

Kraepelin, K. 1899. Skorpiones und Pedipaipi. Das Tierreichs, 8:1-265.

Kraepelin, K. 1912, Neue Beitrage zur Systematics der Gliederspinnen. II. Chactinae. Jahrb.
Hamburg, wiss. Anst., 29:45-88.

Mello-Leitao, C. de 1932. Notas sobre escorpioes Sul-Americanos. Arq. Mus. Nac. Rio de Janeiro,
34:9-46.

Mello-Leitao, C. de 1945. Escorpioes Sul-Americanos. Arq. Mus. Nac. Rio de Janeiro, 40:7-468.

Pocock, R. I. 1893a. A contribution to the study of Neotropical scorpions. Ann. Mag. Nat. Hist.,
ser. VI, 12:77-103.

28

THE JOURNAL OF ARACHNOLOGY

Pocock, R. I. 1893b. Contributions to our knowledge of the arthropod fauna of the West Indies. —
Part I. Scorpiones and Pedipalpi; with a supplementary note upon the freshwater Decapoda of
St. Vincent. J. Zool. Linn. Soc., London, 24:374-409.

Pocock, R. I. 1897. Report upon the Scorpiones and Pedipalpi obtained on the Lower Amazon by
Messrs. E. E. Austen and F. Pickard-Cambridge during the trip of Mr. Siemen’s Steamship
“Faraday.” Ann. Mag. Nat. Hist., ser. VI, 19:357-368.

Pocock, R. 1. 1900. Myriapoda and Arachnida, in Report on a collection made by Messrs. F. V.
McConnell and J. J. Quelch at Mr. Roraima in British Guiana. Trans. Linn. Soc. London, ser.
II, 8:64-71.

Roewer, C. F. 1943. Ueber eine neuerworbene Sammlung von Skorpionen des Natur-Museums
Senkenberg. Senckenbergiana, 26:205-244.

Smith, L. B. and C. S. Pittendrigh. 1967. Bromeliaceae. Pp. 37-50, In Flora of Trinidad and Tobago.
Ministry of Agriculture, Industry and Commerce, Trinidad and Tobago. Government Printery,
Trinidad.

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

Waterman, J. A. 1950. Scorpions in the West Indies with special reference to Tityus trinitatis.
Caribbean Med. J., 12:167-177.

Vachon, M. 1974. Etude des caracteres utilises pour classer les families et les genres de Scorpions
(Arachnides). Bull. Mus. Nat. Hist. Nat. Paris (ser. Zool.), 104 ( 140):859-958.

Manuscript received August 1985, revised October 1985.

Forster, L. M. and F. M. Murphy. 1986. Ecology and behaviour in Portia schultzii, with notes on
related species (Araneae, Salticidae). J. Arachnol., 14: 29-42.

ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZII, WITH
NOTES ON RELATED SPECIES (ARANEAE, SALTICIDAE)^

Lyn M. Forster

Otago Museum
Gt. King St.
Dunedin, New Zealand

and

Frances M. Murphy

323 Hanworth Road
Hampton, Middlesex
TW12 3EJ, England

ABSTRACT

Field observations of Portia schultzii, a web-building jumping spider from Kenya, Africa, reveal
that it commonly inhabits the dense webs of Ischnothele karschi, a diplurid spider. As a detritus
mimic, this spider is inconspicuous whether in a web or on the move. Laboratory studies show that
it readily catches a variety of prey-spiders and insects (mainly Drosophila) in its own webs, in I.
karschi and other alien webs, as well as in non-web situations. Although the basic, visually-directed,
predatory sequence conforms to that of the typical salticids, several derived features, such as the
extremely slow and specialised locomotory movements, long periods of immobility in a cryptic
posture, the web strategies of ‘quivering’ and ‘dropping’, and shorter visual discriminatory distances
(not more than 10 cm) than most other salticids, enable this spider to exploit a web environment.

INTRODUCTION

Jumping spiders (Salticidae) have long been known for their proficiency and
versatility as cursorial predators (see Forster 1982a) so it has come as rather a
surprise to find that some species of the genus Portia Karsch specialise in the use
of webs to catch their prey. Web-building was first observed by one of us (EMM)
in the course of rearing P. schultzii Karsch collected from Kilifi, Kenya, in 1974
(see Wanless 1978b, Murphy and Murphy 1983) and it was soon evident that
these salticids seized insects detained by the silk. Although ‘web piracy’ by
members of this group had been recorded several times (e.g.. Gravely 1921,

'The preparation of this paper was supported by grants from the Scientific Distribution Committee,
New Zealand Lottery Board.

30

THE JOURNAL OF ARACHNOLOGY

Bristowe 1941) the fact that some of them also build their own three-dimensional
structures in which to lurk was unknown until that time. These same studies
confirmed that other spiders too were acceptable prey items and explained why
P. schultzii were frequently found in the dense, extensive webs of Ischnothele
karschi (Bosenberg and Lenz), a diplurid spider.

The discovery that P. schultzii construct and utilise their own webs as well as
invading those of other spiders in the pursuit of prey raises a number of
questions. How is prey detected, for instance, and do these salticids rely on vision
to the same extent as cursorial species? Do they behave like other members of
the family, or genus for that matter, and what are the advantages of this life style
to them? To investigate these questions, live P. schultzii were sent to LMF in New
Zealand and the present paper is the result of the collaborative studies thus
undertaken.

At much the same time, another series of investigations (Jackson and Blest
1982a, b) was begun in Australia, the subjects being Portia fimbriata (Doleschall)
which had been located in northern Queensland. Some of these studies paralleled
ours thus enabling us to make comparisons between the two species. Other Portia
spp. were observed and tested by FMM and these findings also provide useful
comparisons.

MATERIAL AND METHODS

Portia schultzii is a medium-sized salticid with a body length of about 7-8 mm.
Its long legs are quite slender but this is seen only in the two terminal segments
where the long, variously colored hairs covering the rest of the body are lacking.
Fringes and tufts on certain parts of the legs supplement this hairy covering, and
the many hues create a mottled effect.

Portia schultzii range from the Forest of Gedi, near Malindi, Kenya, in the
north, to Durban, South Africa, in the south, and westwards to the Rift Valley.
This species has also been recorded from West Africa and Madagascar. For more
detailed descriptions and distribution of P. schultzii and other Portia spp. see
Wanless (1978b) and Murphy and Murphy (1983).

Portia schultzii used in these studies were observed in their natural habitats in
Kilifi on the coast of Kenya from the Forest of Gedi to the Shimba Hills during
the following periods: 11-16 Aug. 1974: 29 Aug. - 24 Sept. 1977: 8 Aug. - 11
Sept. 1980. Spiders were located in a number of sites, viz., in the webs of
Ischnothele karschi (most commonly), on tree trunks and the walls of buildings,
in leaf litter and a stick pile, under a water container, in an old Cyrtophora web,
in a tangle of pholcid and uloborid webbery, and in their own webs.

Captive P. schultzii were housed in variously sized plastic containers (small, 4-
5 cm in diam. x 2-5 cm high; medium, 5-10 cm in diam. x 5-10 cm high; large,
10-20 cm in diam. x 15.25 cm high), with one or more corked openings, usually
supplied with moist cotton wool, and kept in room conditions (15-25°C).

Some of the tests with P. schultzii reported here were carried out in Kilifi but
most of the longer-term observations and experiments were undertaken in
England and New Zealand.

Investigations included: utilisation of silk and construction of webs, behaviour
in various alien webs with or without other intruders, prey preferences, as well

FORSTER AND MURPHY— ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZII

31

as the sequence of predatory reactions in web and non-web situations. In these
latter tests, attention was paid to the primary events and secondary subdivisions
established for Trite auricoma (Urquhart) and other salticid species (Forster
1977a). Briefly, these are:

Orientation: Alert (spider adopts a ‘watchful’ posture)

Swivel (spider turns to face source of movement)

Alignment (abdomen lines up with cephalothorax)

Pursuit: Walk (slow movement towards target)

Run (rapid movement towards target)

Stalk (very slow movement with lowered profile towards target)

Capture: Pre-crouch (low profile, 2 front pairs of legs forwards)

Crouch (low profile, 3 pairs of legs forwards, hind legs tensed)

Jump (leap towards target)

A variety of prey (see ‘Prey preferences’) was offered to R schultzii in small,
medium or large containers, in web and non-web situations.

Small (3-4 mm) lures (plasticene ball, tufts of cotton wool, bits of twig and
leaf) attached to clear nylon thread were moved about in the vicinity of P.
schultzii and their reactions observed. Other events were recorded and described.
More details of methods are given in text where necessary.

RESULTS

The variegated coloration and patterning of P. schultzii together with its
characteristic ‘folded legs’ resting posture (Fig. 1) combine to disguise the identity
of this spider in the web where, as an effective detritus mimic, it is almost
indistinguishable amongst the tiny bits of bark, seeds, dust and leaf fragments
which readily accumulate there. On the move on tree trunks, P. schultzii is all
but invisible and since, in the wild, it is most likely to commute between webs
via trees and shrubs, its crypticity is obviously designed to cope with the two
most likely situations in which it may find itself.

Distribution and Ecology. — The marked propensity of P schultzii to inhabit
the webs of /. karschi probably means that the distribution of these two species
is closely linked. Ischnotheie karschi have apparently benefited from the
modification of tropical rain forests by man since their webs are very abundant
in the partly cleared secondary bush which forms most of the gardens in Kilifi,
whereas they are far less abundant in the tropical rain Forest of Gedi. Perhaps
this has been advantageous for P. schultzii, too.

Ischnotheie karschi is a medium-sized diplurid (about 15 mm in body length)
which builds an extensive sheet web, occasionally on the ground, but more
commonly in shrubs at a height of about one metre above the ground. The dense,
central region of this web becomes greatly cluttered with plant debris thus
providing an apparent refuge not only for immature /. karschi, but also for a
great variety of spiders and insects, e.g. mysmenids (most common), scytodids,
palpimanids, prodidomids, pisaurids, mimetids, Ctenus sp., Orchestina sp.,
Argyrodes sp., Cyllobellus sp., Cosmophasis sp., Myrmarachne sp., in two
instances Thyeni inflata females with eggsacs (constructed between leaves
supported in the upper threads), and once an Olios exuviae, as well as insects
such as thysanurids, crickets and ants. Unfortunately, it was rarely possible to see

32

THE JOURNAL OF ARACHNOLOGY

*

#

I#*

Fig. 1. Often observed hanging motionless in its web for long periods of time, Portia schultzii
bears little resemblance to a spider. Its long legs are tucked against the body and are thereby
concealed, a mimetic attitude of the kind described by Robinson (1969). But if it sights likely prey,
the legs unfold and Portia begins a slow stealthy trek towards it.

what parts of the web P. schultzii usually occupied, or how they behaved therein,
although they could sometimes be seen wandering about on the periphery of the
web.

Fortuitously, an /. karschi web built in a fold of the trunk of a baobab tree
{Adansonia digitata) had not accumulated any fallen leaves or other debris and
it was possible to see further into the interior than usual. On three successive
days, a small P. schultzii was seen in the web but on the fourth day, three were
seen at much the same time. Two days later, a half-grown Portia as well as a
smaller one (alongside what was presumed to be its exuviae) were present
although they quickly vanished into the depths of the web. None was seen during
occasional observations over the next two weeks but when the web was
demolished only one half-grown Portia was revealed although there were several
small mysmenids, some thysanurids, as well as the resident owner.

These observation suggest that P. schultzii is nomadic, wandering from one web
to another, whereas P. fimbriata is relatively sedentary (Jackson and Blest 1982a)
and in any case, can move readily from one host web to another because of their
contiguity. Evidently P, schultzii prefer /. karschi webs, since they were never
found in the quite prevalent Stegodyphus webs and only occasionally in pholcid
or uloborid webs. Moreover, they were found in widely varying non-web
situations which suggests they were merely seeking for webs to invade.

A diplurid web usually harboured a single Portia at a time although very small
spiders were sometimes found together. Moreover, a survey of about 25 webs in
the Kilifi area showed that the overall distribution of P. schultzii was about one

FORSTER AND MURPHY— ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZII

33

individual to every third diplurid web. In one instance, a rolled-up leaf, hung in
the upper strands of an L karschi web, yielded a cluster of newly hatched Portia
spiderlings. Since the hair tufts did not appear until the third instar, spiderlings
were not immediately recognisable as P. schultzii offspring. Interestingly, Jackson
and Blest (1982a) report that P. fimbriata females construct special webs (Type
2) prior to the suspension of suitable leaves where the eggsacs are deposited.

Behaviour in ‘captive’ alien webs. — Portia schultzii. In captivity, a P. schultzii,
introduced into a small cage where thick webbery had been constructed by an
/. karschi, disappeared within the hour and was not seen again, apparently a
victim of the owner. It was a similar story with Stegodyphus sp. although the
Portia survived for several days before vanishing. However, in a large cage where
/. karschi had built a thick webbery extending from a retreat of dead leaves
bound with silk, the introduction of a half-grown P. schultzii, three mysmenids,
and later a number of L karschi spiderlings, led to rather different results.

Although one of the mysmenids lasted for about 4 months, two disappeared
within a week, and not long after, P. schultzii was seen to catch a young /.
karschi on the edge of the web. Two days later, fruitflies were put into the cage
whereupon the adult /. karschi emerged and took two, ignoring both Portia and
the spiderlings. Indeed, /. karschi was never seen to attack the spiderlings
although they were sometimes seen eating each other. Portia, however, often
seized wandering spiderlings.

Initially, P. schultzii remained near the floor of the cage at the opposite end
to /. karschi and generally ignored fruitflies in favour of spiderlings. After about
a month it moved up to the corked opening where the fruitflies were introduced
and subsequently took fruitflies as prey although there were still several
spiderlings available. Then, after 3 weeks, it moulted, moved back to the bottom
of the cage, and shortly afterwards disappeared, evidently a victim itself.

Portia labiata (Thorell): In 1979, three P. labiata were collected in Malaya. One
female, taken back alive to England, made a small sheet web on the underside
of its cage but did not respond to fruitflies dropped into the web as P. schultzii
had done. After being introduced into a medium-sized cage where an immature
Tegenaria had built a retreat and a web, P labiata soon seized the Tegenaria.
Returned to its home cage, P labiata was later observed hanging from the web
by its 3rd and 4th pairs of legs devouring its victim, a posture usually adopted
by P. schultzii when eating prey. In a pholcid web, P. labiata showed its ability
to move through the threads without alerting the pholcid but although it jumped
into the web after fruitflies did not succeed in catching any.

When P. labiata was placed on a wooden frame supporting a Meta web and
spider, it suddenly stirred after a long period of inaction and swivelled very
slowly through 30° towards the motionless Meta about 3 cm away. Aligning its
abdomen, Portia slowly stalked the Meta along the frame, finally crouching and
leaping at it from about 1 cm. After jumping back to the frame Portia began
feeding. Phiale is another salticid which attacks the spider Argiope argentata
(Fabricius) from outside the web although both victim and quarry then drop to
the ground (Robinson and Valerio 1977).

When a plasticene lure was jiggled briefly to one side of P labiata, it swivelled
very slowly through about 40°, aligned its abdomen, then began to creep slowly,
almost imperceptibly towards the now stationary lure until the side of the clear

34

THE JOURNAL OF ARACHNOLOGY

plastic arena was reached. It did not jump but other salticids often refuse to jump
at lures or live prey when barriers of glass or plastic intervene (Forster 1979b).

Portia durbanii Peckham and Peckham: Not all Portia species make webs. A
juvenile P. durbanii from South Africa (see Wanless 1978b) was reared by FMM
through several moults and showed no sign of web-spinning activity. Fed mostly
on fruitflies, it caught prey in typical salticid fashion, albeit much more slowly.
This species moults in the open.

Utilisation of silk by Portia schultzii. — When moving about in the open or
leaping across gaps, P schultzii laid down a dragline and attached it at various
points, as all typical salticids do. But once a number of aerial threads had been
established, this salticid showed a pronounced tendency to use these threads for
passage and this, in turn, increased the aggregation of aerial lines which readily
became attached to each other where they crossed. However, typical salticids
seldom use draglines in this manner nor do their silk threads adhere to each other
in such a way. Clearly, these modifications are fundamental to the life style
exhibited by P. schultzii.

When caged, P. schultzii immediately constructed a few aerial threads from
which it hung, upside-down. More silk was added to the preliminary structure
from time to time, resulting in a three-dimensional network supporting a
recognisable platform beneath which the spider usually rested. Mature males,
however, seldom built such platforms; indeed, their webs were much flimsier than
either female or juvenile structures.

In a web, whether its own or that of a host, P. schultzii almost always moved
about in a languid manner interspersing stepping movements with flexing and
stretching motions of the limbs, thus creating the impression of a ‘lolloping’ gait
(Figs. 2-3). This gait is strongly reminiscent of the locomotory activity of
gradungulid spiders (LMF. pers. obs.) which flex and stretch their long legs in
like manner as they move about the substrate. Perhaps the flexing, stretching
components of stepping actions in P. schultzii are neeessary to maneuvre the
extra length of the limbs while keeping the body close to the ground given that
‘long leggedness’ is a useful adaptation in web environments.

Lolloping movements differed little regardless of whether the spider was
approaching prey, exploring a new web, or walking along a substrate, although
in the first situation, a directional element was present. However, lolloping
seemed more pronounced during stalking than during walking or running but
perhaps this was merely a function of the distance of the forward motion
involved.

During the day, Portia remained quiescent for long periods, usually suspended
upside-down from its platform, or alternatively it sought a ‘vantage’ point from
which it apparently monitored activity in the web. For example, several captive
spiders regularly sat on cork plugs all day, facing the web, never moving from
there unless prey was introduced. If disturbed, their palps undulated rapidly.
Another jumping spider, Euophrys parvula Bryant, which sits for long periods on
the walls of houses keeping a ‘watchful eye’ on its visual territory, also undulates
its palps rapidly in the presence of intruders (Forster and Forster 1973).

Without fail, P. schultzii returned to their platforms at night. In one instance,
a spider devouring prey in a cage without a web, promptly constructed a web
at dusk (holding the prey meanwhile) and then continued its meal. This nightly

FORSTER AND MURPHY— ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZII

35

Figs. 2-3. — The ‘lolloping’ gait of Portia schultzii is apparently achieved by the long legs which
alternately (Fig. 2) flex and (Fig. 3) stretch as the spider makes its unhurried way towards the target.

behaviour is clearly analogous to that of many salticids which usually either make
new retreats at dusk or return to an old one.

Moulting. — Portia schultzii exuviae were observed on some occasions either
hanging in their own webs or in those of /. karschi thus leading to the
assumption that these salticids moult in the open (also see Jackson and Blest
1982a). On one occasion, the final stage of the moult of a caged spider was
observed. The moult skin was suspended in the web, attached to a thread by the
claws of a fourth leg; the spider, free except for its spinnerets, was hanging from
the moult skin, its pale, translucent legs still extended against the underside of
the body. Some slight flexing of the limbs was noticed and then suddenly the
spider began to ‘twirl.’ For at least 90 seconds both spider and skin ‘twirled’, a
relatively slow motion in which both were clearly visible. When the action ceased,
the spider was seen to be free of its exuviae and the legs in more normal
positions. The spider darkened up quickly and some time later it had assumed
its typical hanging posture in the web (see Fig. 1). Was this a freak happening
or might twirling be an adaptive feature of moulting to aid the spider in casting
off its skin in the open?

PREDATORY EVENTS

Predatory events were examined in four different situations (see Methods): in
two cages P. schultzii were in their own webs and in two they were without webs.

Typical and modified salticid responses. — Portia schultzWs primary responses
(Orientation, Pursuit and Capture) to prey were identical in form and sequence
in both the web and non-web situations to those of typical salticids but the
secondary components were frequently modified. Orientation (Alert, Swivel and
Alignment) always preceded approach although there was a marked tendency for
small-angle (<30°) swivels to be quite fast and for large-angle (>30°) swivels to
be abnormally slow. This suggests a distinction in the rate at which turning is
mediated either between the two pairs of lateral eyes or between the principal and
lateral eyes, a difference not recorded in other members of the family. However,
within the visual range of the principal eyes, a series of ‘tantalus-like’ turns (see
Land 1971) often occurred, such reactions being common to typical salticids. The
rate of alignment was typical, a movement probably also mediated by the
principal eyes.

During approach to the target, P schultzii walked, very occasionally ran, and
crept towards the target. Unlike typical salticids, however, walking along a

36

THE JOURNAL OF ARACHNOLOGY

substrate or negotiating the web network was punctuated by leg stretching and
flexing movements (lolloping) (Figs. 2-3) which tended to further reduce the
already slow rate of stepping. Nor did stalking result in the low profile adopted
by most other salticids (Forster 1977a) but no doubt Portia's long legs precluded
this posture on the ground and it would have had little value in the web.
Moreover, the pursuit of a target, however prolonged, was always linear, Portia
apparently being unable to diverge from a straight path unlike, e.g. Phidippus
(Hill 1979). If, for instance, a fruitfly or prey spider walked consistently around
the perimeter of a circular cage, Portia never walked around after it, although
it would swivel from time to time. This behavioural limitation is also
demonstrated by Trite planiceps Simon when its anterior-lateral eyes are occluded
(Forster 1979a); under such conditions this salticid also creeps very slowly and
only in a straight line towards the target, whereas normally it could give chase
at speed in any direction regardless of the inconsistency of the quarry’s
movements (Forster 1979b). The consequences of this is demonstrated by the
following test:

When P. schultzii and Trite auricoma were each presented simultaneously with
a fruitfly in small identical containers, T auricoma seized its quarry in 15 seconds
whereas Portia did not succeed for 1 Y2 hours. There were, in fact, only two
circumstances whereby Portia could catch the fruitfly. Either the fruitfly had to
remain motionless long enough for Portia to stalk it directly or it had to fly
between Portia's front legs. In brief, Portia cannot maintain visual contact with
swift, erratically-moving prey nor can it run after it, whereas T auricoma
possesses both capabilities, functions most likely mediated by the anterior-lateral
eyes (see Forster 1979a).

To complete the hunting sequence, P. schultzii halted at a critical distance from
the target, usually I.O to 1.5 cm, placed the first two pairs of legs forward (pre-
crouch) and then brought the third pair up (crouch), as typical salticids do (see
Forster 1977a), before lunging forward and seizing its quarry. Generally, the 4th
pair of legs did not leave the ground, or the web, so that, in effect, the spider
simply reached forward to grasp its prey. Nonetheless, in escape situations, these
spiders were able to move very fast and often jumped distances of several
centimetres during which the 4th pair of legs clearly left the ground.

Atypical tactics. — One tactic by which prey was seized has not been recorded
previously in salticids, as far as we know, although there are similarities to the
behaviour of Phiale (q.v.). A prey spider that ran beneath Portia, alert in its web,
could be captured when Portia dropped on it from distances of up to 8 cm above
it. The ’’drop”, always with Portia in a web, was extremely rapid so that details
of the movement could not be recorded by eye. If Portia missed the target, it
quickly hauled itself back to the original look-out site in the web and waited for
another attempt. Moreover, it usually positioned itself as before which is exactly
what Hill (1979) found in Phidippus pulcherrimus Keyserling when testing this
spider’s capacity for re-orientation on a substrate after jumping at prey in mid-
air. Clearly, too, Portia was alert for a second chance, for it could be seen
swivelling in the web as the object of its intentions moved about below it.
Normally, a subsequent drop was successful. However, dropping was a more
common occurrence from 2-4 cm above, regardless of whether the target was in
the web or on the ground, because longer drops were generally inhibited by web
structure. We never saw Portia attempt a drop if there was even the flimsiest of

FORSTER AND MURPHY— ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZII

37

silk structures between it and the target so we presume that the spider can
determine the presence of silk by sight. If, in the wild, P. schultzii lurked in the
lower portions of a web, it would be able to drop on insects and spiders moving
about on the ground below but such behaviour has not been observed.

A second web stratagem, not seen in typical salticids, was recorded as a
‘quiver.’ Portia schultzii always crept very, very slowly towards motionless prey,
especially if it was roughly equal in size to or larger than itself. When within
about 2-3 cm of its quarry, Portia might pause for up to 10 minutes, but in one
case for 25 minutes, during which time the body would quiver with very rapid
motions at fairly regular intervals. This movement was transmitted to the web
and hence to the target. It is difficult to say, unless one has seen both
performances, whether quivering is comparable to web vibration and tweaking in
P. fimbriata (Jackson and Blest 1982a) although it seems as if, in P. fimbriata,
these agitations of the web were caused by movements of the limbs whereas in
P. schultzii the entire body appeared to be involved. We never saw P. schultzii
tweak the web as these authors described for P. fimbriata. In almost all instances,
the quarry had remained completely motionless during the approach by Portia
so it seems most likely that quivering was designed either to enhance some aspect
of the spider’s visual perception of the target or to induce its movement.

Distances at which prey is detected. — In large, domed terraria, P. schultzii
constructed extensive but not dense networks of silk in the upper half of the
dome and mostly placed their platforms near the centre. From here, and the side
of the dome where they often sat, Portia had a range of vision extending some
10-15 cm across and above, with about 25 cm below.

Under illumination levels ranging from 1500 lux to bright sunlight, the distance
at which prey was detected was 9 or 10 cm. No responses to prey were ever
observed at distances greater than this, despite the fact that flies and spiders
introduced into the cage provided ample opportunity for long distance perception
and recognition to occur. Compare this behaviour to that of Trite auricoma, for
instance, which swivels towards a movement up to 75 cm away and approaches
suitable targets from about 20 cm (Forster 1977a, 1979a). Presumably, there is
little advantage to Portia in being alerted to prey at any great distance since its
slow rate of locomotion would seem to preclude it from ever reaching the target
in time to capture it. Nor was there the slightest indication that detection and
prey capture were mediated by other than vision; prey trapped out of sight in
the web but within 10 cm elicited not the slightest response from Portia.

At luminance levels of less than 100 lux, P. schultzii did not respond to prey
but between 100 and 500 lux they detected, approached and seized prey from
distances up to 6 cm. Above 500 lux response distances increased but reached
a maximum of about 10 cm at 1500 lux. More precise measurements obtained
from courtship and agonistic interactions (Forster, in prep.) confirmed that
reliable discriminations can only be made up to 10 cm away.

Reactions to lures. — Whether in or out of the web, P. schultzii treated all
moving lures as potential prey. Reactions included all elements of the salticid
prey-catching sequence: Alert (if target in front); Swivel (if target to side or rear);
Walk, Run or Stalk (depending on mobility of target); Follow (if target receding);
Pre-crouch, Crouch and Lunge (if target stationary but jiggled slightly).

These tests (n = 36) showed that target movement was a most important
stimulus parameter, apparently over-riding details of target shape and pattern.

38

THE JOURNAL OF ARACHNOLOGY

since P. schultzii did not discriminate between the items used as lures and actual
prey.

Other observations were as follows:

1. No reactions occurred at distances greater than 10 cm.

2. If movement of the target was erratic, Portia did not pursue it.

3. If the target was stationary, schultzWs approach was extremely slow (5 mm/
min - 100 mm/ min) and very long pauses (up to 25 mins) preceded pre-
crouch, crouch and lunge. In 43% of cases, the sequence was not completed.

4. If the target was stationary, but jiggled slight (cf. 3) Portia's approach was
much faster and the sequence was always completed.

5. If lures were trailed directly away from Portia, Portia followed — the faster
the lure receded, the faster Portia moved (up to a limit). Trite planiceps
behaves in a similar fashion under such predatory conditions (Forster, 1985).
Prey preferences. — Drosophila were the most readily available insects hence

were regularly offered to captive P. schultzii. The P. schultzii spiderlings,
mentioned previously, fed readily on them and in the tests designed to identify
predatory events they were seized almost as frequently as prey spiders. House flies
were less likely to be caught, Portia invariably retreating before their sudden
flights and buzzing onslaughts. If, however, a fly was left for several days in a
small cage, gradually becoming less active or entangled in the web, Portia might
eventually catch it. Although various insects were found in diplurid webs it was
not possible, owing to the opacity of the webs, to say whether Portia ever ate
any of them. But once, on a wall in Kilifi, Portia was seen eating a fly.

Offered one live and one dead Drosophila, Portia stalked and lunged at the
dead one, although it took 5 minutes to cover the two-centimetre distance to the
lunging point. Perhaps the live fly provided the initial stimulus but clearly the
stimulus properties of the dead fly were able to elicit the remainder of the
sequence. Moreover, there were other occasions on which Portia accepted dead
flies although details of the seizures were not witnessed. Hence movement is not
the only inducement needed for stalking and lunging.

Thomisids offered as prey when P. schultzii was off the web were rarely
captured, largely because these prey-spiders had a tendency to wave their front
pairs of legs at any ‘menacing object’, a ploy which certainly deterred Portia.
However, on two occasions, when Portia was in a web and a thomisid was
moving about on the ground several centimetres below, Portia dropped and
scooped it up. Theridiids, too, were often caught by this tactic, generally when
both prey and predator were in the web.

Small jumping spiders (unidentified) were only caught when both were on the
ground but capture depended on these prey-spiders either (i) remaining stationary
long enough for Portia to stalk them directly or (ii) walking away from Portia
in a straight line. This type of departure apparently induced Portia to follow in
much the same way as some male salticids follow a mate (Forster 1977c).

The ready availability of Cambridgea sp. meant that they were offered to P
schultzii more frequently than other prey spiders. All the tactics at Portia's
disposal were used against them but since Cambridgea tended to remain
motionless for long periods of time, very slow approaches were customary and
body quivers were commonly employed.

Araneids were more likely to be captured if they moved about. Portia regularly
stalked the motionless spiders but if unable to induce movements by quivering,

FORSTER AND MURPHY^ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZIl

39

or to obtain adequate stimulation by sidling around and scrutinising the araneid
from different angles, Portia would abandon the hunt.

Given a choice off the web between active prey (fruitflies, small salticids,
thomisids, Neoramia spp.) and relatively quiescent prey (Cambridgea spp.,
Araneus spp.), Portia invariably selected the laggards, stalking very slowly and
lunging only when its quarry stirred slightly.

DISCUSSION

Our observations show that Portia schultzii is primarily a web-dweller and that
the web it occupies may be of its own construction or that of an alien species.
Within the geographical limits of the present study, the most preferred alien web
appears to be that of a diplurid spider, Ischnothele karschi. We find, moreover,
that P schultzii readily catches prey when in a web and just as readily when away
from a web. As a web-dweller, therefore, the question is, does this spider catch
prey in the manner of a typical salticid or does it use methods more commonly
associated with other web-builders?

The evidence from this study is that P. schultzii is a salticid-like hunter and
that it has only secondarily adapted to a web-dwelling lifestyle. This conclusion
is based, first, on the fact that P. schultzii hunts solely by vision (also see Forster
1982a) and since the structure of the eyes is very similar to that of the typical
salticid (Forster, pers. obs.) the conclusion is that the peripheral detection of
prey, as evidenced by swivelling, is mediated by the lateral eyes and that the
ensuing pursuit and capture of prey are mediated by the principal eyes. Second,
P. schultzii employs the same set of predatory events as typical jumping spiders
and although two web tactics (quivering and dropping) have been added to the
basic salticid repertoire, both are guided by vision and neither appear to be
inherited from any other web-builders. Third, P. durbanii is shown to be a
cursorial hunter despite its known morphological relationship to the web-building
Portia species (Wanless 1978b). Moreover, all known Portia species possess two
claws and a claw tufts (Wanless 1978b), an acknowledged salticid feature which
links them to a cursorial predecessor (Gertsch 1979). Taken together, these
findings suggest that a number of Portia species, including P. schultzii have
secondarily adapted to a web-building and web-invading lifestyle.

The commonality of the visually specialised hunting strategy in salticids
suggests that the division of labour, which probably evolved to meet the needs
of a diurnal predator (Forster 1982b), was so successful that these spiders soon
outstripped their closest relatives. This basic cursorial strategy gave rise to a
variety of adaptations by which salticids were able to exploit a vast array of
terrestrial habitats and situations. One of the most successful of these — mimicry
— has taken many forms (see Reiskind 1977, Wanless 1978a, b, Platnick 1984,
Edwards 1984). Sarinda hentzi Banks, for example, is a Central American ant-
mimicking salticid which not only has the appearance of an ant, but also spends
much of its time in ant-like behaviour, being diverted only spasmodically for
bouts of salticid-like predation and reproductive activities (Forster 1982a and
pers. obs.). Portia schultzii, too, is a mimic, spending much of its time posing
as a piece of bark which has fallen into a web, reacting like a salticid only when
it sights potential prey or mate.

40

THE JOURNAL OF ARACHNOLOGY

In adapting to a web environment, P. schultziVs precursor retained the primary
visual elements of prey capture, i.e. orientation, pursuit and capture (Forster
1977a) because these visually mediated events characterise the predatory sequence
in the present-day P. schultzii. It was in the secondary components that
adjustments were required. In a web, it was clearly advantageous for large-angle
swivels to be performed slowly because fast ones might produce vibratory stimuli
or visual signals that would alert a neighbouring web-dweller or the host spider.
Slow, spasmodic approaches to prey would obviate the same kind of disturbances
and also be more appropriate when negotiating an alien web structure. ‘Dropping’
onto the prey from the relative safety of a vantage point does not carry the same
risk, hence rapid movement would be acceptable under these conditions.

Portia's failure to chase fast-running and erratically-moving prey either on or
off the web may mean that the anterior-lateral eyes are not involved in post-
detection behaviour as shown for Trite planiceps (Forster 1979a, 1985).

Moreover, because insects are usually immobilised by the silk, or prey spiders
quiescent, ‘quivering’ by P. schultzii may be an adaptive ploy by which movement
is induced in the quarry and visual cues enhanced, selection processes having
apparently ensured that its nature and frequency do not imitate those of a
trapped insect and so attract another predator to the scene.

Portia schultzii pierces prey after lunging at it, or scoops it up after ‘dropping’
whereas Jackson and Blest (1982a) observed that P. fimbriata swoops at, stabs
or picks up prey. Nevertheless, behavioural variability in P. schultzii does not
appear to be related to diversity in the type of prey encountered as shown by
P. fimbriata (Jackson and Blest 1982b) but rather to differences in prey mobility
as well as the conditions under which predation occurs (Forster 1985). Moreover,
P. schultzii does not ‘tweak’ the web as Jackson and Blest (1982a) describe for
P fimbriata for ‘quivering’, which has some similar characteristics, does not
appear to have the same functions.

In captivity, P. schultzii catches a wide range of spiders as prey and, unlike
P. fimbriata (Jackson and Blest 1982a), catches insect prey Just as readily. Portia
labiata exhibits a similar ability to catch prey both in its own webs as well as
alien webs and is also able to jump into and out of webs, a skill not
demonstrated by P. schultzii nor, it seems, by P. fimbriata, because Jackson and
Blest (1982a, b) make no mention of it. Moreover, the absence of web-building
in P. durbanii highlights the need for caution in making assumptions about the
evolutionary status of this genus until more species have been studied.

Many of the behaviours exhibited by known Portia species have been observed
independently in cursorial salticids. For example, a New Zealand Marpissa sp.
has been seen hunting on an araneid web (Forster, pers. obs.). An unidentified
salticid has been observed devouring ants while suspended from a thread, and
Phiale apparently makes a practice of leaping onto Argiope in its orb-web
(Robinson and Valerio 1977) as does Phidippus (Tolbert 1975). In behaviour
comparable to that of P. fimbriata taking insects from the chelicerae of the web
spider host (Jackson and Blest 1982a), an Indian Marpissa sp. robs ants of their
prey (Marson 1947). Similar prey robbery, described as kleptoparasitism, has
been reported for four Japanese salticid species, one of which has been seen
devouring the eggs of another spider (It 1977). Because the behaviour of relatively

FORSTER AND MURPHY-ECOLOGY AND BEHAVIOUR IN PORTIA SCHULTZII

41

few salticid species is known, it is likely that many more diverse strategies are
yet to be discovered.

ACKNOWLEDGMENTS

We would like to thank Mr. Neville Chalke for acting as a courier and bringing
Portia schultzii to Dunedin for study.

REFERENCES

Bristowe, W. S. 1941. The Comity of Spiders. Vol. 2. Ray Society, London.

Edwards, G. B. 1984. Mimicry of velvet ants (Hymenoptera: Mutillidae) by jumping spiders (Araneae:
Salticidae). Peckhamia, 2(4):46-49.

Forster, L. M. 1977a. A qualitative analysis of hunting behaviour in jumping spiders (Araneae:
Salticidae). New Zealand J. Zool., 4:51-62.

Forster, L. M. 1977b. Mating behaviour in Trite auricoma, a New Zealand jumping spider.
Peckhamia, 1:35-36.

Forster, L. M. 1979a. Visual mechanisms of hunting behaviour in Trite planiceps — a jumping spider.
New Zealand J. Zool., 6:79-93.

Forster, L. M. 1979b. Comparative aspects of the behavioural biology of New Zealand jumping
spiders (Araneae: Salticidae). Ph.D. dissertation. University of Otago, New Zealand. 402 pp.

Forster, L. M. 1982a. Vision and prey-catching strategies in jumping spiders. Amer. Sci., 70(2): 165-
175.

Forster, L. M. 1982b. Visual communication in jumping spiders (Salticidae). Chapter 5, In Spider
Communication: Mechanisms and Ecological Significance (P. M. Witt and J. S. Rovner, eds.).
Princeton Univ. Press, Princeton.

Forster, L. M. 1985. Target discrimination in jumping spiders. Chapter 13, In Neurobiology of
Arachnids. (F. Barth, ed.). Springer-Verlag, Frankfurt, Berlin.

Forster, R. R. and L. M. Forster. 1973. New Zealand Spiders, an introduction. Collins, Auckland,
London. 254 pp.

Gertsch, W. J. 1979. American Spiders (2nd Edition). Van Nostrand Reinhold Coy, New York.

Hill, D. E. 1979. Orientation by jumping spiders of the genus Phidippus (Araneae: Salticidae) during
the pursuit of prey. Behav. Ecol. Sociobiol., 5:301-322.

It, N. 1977. Observations on a strange plundering behaviour in salticid spiders. Acta Arachnol.,
27:209-212.

Jackson, R. R. and A. D. Blest. 1982a. The biology of Portia fimhriata, a web-building jumping
spider (Araneae: Salticidae) from Queensland: utilisation of webs and predatory versatility. J. Zool.
London, 196:255-293.

Jackson, R. R. and A. D. Blest. 1982b. The distances at which a primitive jumping spider, Portia
fimbriata, makes visual discriminations. J. Exp. Biol., 97:441-445.

Land, M. F. 1971. Orientation by jumping spiders in the absence of visual feedback. J. Exp. Biol.,
54:119-139.

Marson, J. E. 1947. The ant mimic, Myrmarachne plataleoides. J. E. Africa Nat. Hist. Soc., 19:62-
63.

Murphy, J. and F. Murphy. 1983. More about Portia (Araneae: Salticidae). Bull. British Arachnol.
Soc., 6(l):37-45.

Platnick, N. 1. 1984. On the pseudoscorpion-mimicking spider Cheliferoides (Araneae: Salticidae). J.
New York Entomol. Soc., 92(2): 169-173.

Reiskind, J. 1977. Ant-Mimicry in Panamanian clubionid and salticid spiders (Araneae: Clubionidae,
Salticidae). Biotropica, 9(1): 1-8.

Robinson, M. H. 1969. Defenses against visually hunting predators. Pp. 225-259, In Evolutionary
Biology (T. Dobzhansky, et al., eds.), vol. 3.

Robinson, M. H. and C. E. Valerio. 1977. Attacks on large or heavily defended prey by tropical
salticid spiders. Psyche, 84(1): 1-10.

42

THE JOURNAL OF ARACHNOLOGY

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

Wanless, F. R. 1978a. A revision of the spider genera Belippo and Myrmarachne (Araneae: Salticidae)
in the Ethiopian region. Bull. British Mus. Nat. Hist. (ZooL), 33(1): 1-139.

Wanless, F. R. 1978b. A revision of the spider genus Portia (Araneae: Salticidae). Bull. British Mus.
Nat. Hist. (ZooL), 34(3):83-124.

Manuscript received March, 1985, revised July 1985.

Ugolini, A., I. Carmignani and M. Vannini. 1986. Mother-young relationship in Euscorpius: Adaptive
value of the larval permanence on the mother’s back (Scorpiones, Chactidae). J. Arachnol., 14: 43-
46.

MOTHER-YOUNG RELATIONSHIP IN EUSCORPIUS: ADAPTIVE
VALUE OF THE LARVAL PERMANENCE
ON THE MOTHER’S BACK (SCORPIONES, CHACTIDAE)

A. Ugolini, 1. Carmignani and M. Vannini

Dipartimento di Biologia Animale e Genetica
Universita di Firenze
V. Romana 17, 50125 Firenze, Italy

ABSTRACT

Mortality rates were measured in Euscorpius flavicaudis larvae which had been made to stay on
their mother’s back (as actually occurs in nature) or had been separated from their mother, under
three different relative humidity levels. Mortality was greatest among larvae exposed to low RH levels
and in any cases, among larvae separated from their mother. The mother-offspring relationship likely
serves several functions: it is an obvious defense against predators and provides an optimum micro-
habitat selection. These experiments suggest that the mother may also provide a greater resistance
to dehydration, either by water-proofing the larvae or by refurnishing their water loss.

INTRODUCTION

The permanence of the larvae on the mother’s back after birth is an universal
feature among scorpions. In spite of investigations by Angermann (1957), Torres
and Heatwole (1967), Le Pape (1974), Vannini et al. (1978), Vannini and Ugolini
(1980) and Ugolini and Vannini (1983), evidence is still lacking as to the exact
adaptive value of this behavior.

Defense against predators and maintenance of the larvae at more optimal
microclimatic conditions seem plausible (Williams 1969, Maury 1969, Vannini et
al. 1978, Vannini and Ugolini 1981) but some kind of trophic exchange has also
been hypothesized (Alexander 1977).

The purpose of our work was to investigate whether larvae raised on their
mother’s back and larvae isolated from their mother exhibit the same survival
rate under stressful environmental conditions of low relative humidity.

MATERIALS AND METHODS

Pregnant females Euscorpius flavicaudis (Geer) were captured near Florence
(Italy) in August-September 1980-1982. In the laboratory the animals were reared
individually in small containers under natural conditions of temperature, relative
humidity and photoperiod. They were given Tenebrio molitor larvae and water
once a week. No food or water was administered during experiments.

Within 24 hours of birth, each litter was divided into two or three groups: 1)
larvae put back onto their mother (OM); 2) larvae separated from their mother

44

THE JOURNAL OF ARACHNOLOGY

Table 1. — Mortality rates (total number of deaths/total number of larvae tested) with respect to
relative humidity (RH) and treatment; OM, larvae on mother; WM, larvae without mother: n =
number of broods.

RH:

10%

60%

90%

treatments:

OM

47/132 (35.6%)

10/ 130 ( 7.7%)

5/112(4.5%)

WM

91/16 (71.2%)

32/139 (23.0%)

6/123 (4.9%)

n=

9

13

11

but grouped together in a single container (WM); and, in certain cases, 3) larvae
separated from their mother but also isolated one from another (IS). Each group
was exposed to one of three levels of RH (10%, 60% and 90%), obtained by using
CaCb, environmental RH and distilled water, respectively.

The larval mortality rate was then measured for each type of treatment and
defined as the ratio between the nymphs still alive after the first moult and the
initial members of each group.

RESULTS

The comparison of the experiments on larvae kept at 10%, 60%, and 90% RH,
on the mother and without the mother (Table 1), shows that the larval surviving
probability increases at higher RH levels, and, in general, in presence of the
mother.

The effect of the presence/ absence of the mother can be tested by applying the
Wilcoxon matched-pairs test (Table 2). The absence of the mother is shown to
increase the larval mortality rate at 10% and 60% RH but not at 90%.

It is possible then to compare the mortality rates at different RH levels by
applying the Mann-Whitney test (Table 3). The differences between 10% vs 60%
and 10% vs 90% RH are always significant, whereas the difference between 60%
and 90% RH is only evident among the larvae separated from the mother.

The presence of the mother is therefore reducing the larval water loss. The
larvae, when on their mother’s back, are usually highly aggregated. The larvae
without the mother are usually found aggregated at low RH, whereas they appear
largely scattered at high RH levels. The presence of the mother could then reduce
the larval water loss, independently from any substances exchange, by simply
inducing their aggregation.

The effect of the aggregation on the surviving rate was then measured
comparing (only at the lowest RH level) the behaviour of larvae with the mother,
without the mother but free of aggregating and without the mother and isolated
from each other (Table 4). The analysis of the results show the aggregation effect
is anyway quite negligible.

Table 2. — Comparison between mortality rates on and without mother. T = statistics of Wilcoxon
matched-pairs test (two-tailed); N = number of matched pairs whose difference is not zero; in
brackets, total number of pairs.

RH

10%

60%

90%

12(13)

10(13)

6(11)

T=

4

0

7

P

<0.05

<0.01

n.s.

UGOLINI MOTHER-YOUNG RELATIONSHIPS

45

Table 3. — Comparison between mortality rates on the mother (OM) and without the mother (WM),
at different RH levels. statistics of Mann-Whitney test (two-tailed).

RH;

10% vs 60%

10% vs 90%

60% vs 90%

(OM)

U =

5.5

5.0

67.0

P

<0.002

<0.002

n.s.

(WM)

15.5

7.5

22.5

P

<0.002

<0.002

<0.02

When is dehydration mostly affecting the larval survival rate? The 91.5% (279/
305) of the dead larvae were dead the moulting day or the day before.

DISCUSSION AND CONCLUSIONS

Larval aggregations are a universal and necessary feature of scorpion life
history. Such aggregation is accomplished both by larval behaviour (fallen larvae
climb back onto their mother’s back; Vannini et. al. 1978) and chemical cues
favoring their permanence on the mother (Torres and Heatwole 1967, Vannini
and Ugolini 1980, Ugolini and Vannini 1983).

It is possible that the mother provides protection for their young in a number
of ways:

i) when the larvae are on their mother’s back they are probably safe from a wide
range of potential predators (other scorpions, spiders, ants, centipedes; Polis et.
al., 1981) which are unlikely to harm larger scorpions;

ii) movement by the mother allows selection of more suitable microhabitats while
larvae are quite limited in their movements.

Our investigations show two additional advantages:

iii) by aggregating together on the mothers there is a passive decrease because
of boundary layer effects, of water loss and subsequent dehydration;

iv) the larvae on their mother’s back are more likely to survive the moult without
mishap.

Could advantage iv) result from trophic exchange between mother and young?
The larvae never participate in their mother’s meal, nor do they tend to take up
a position near her mouth-parts; this excludes any oral exchange of food and
water between the mother and her larvae. It is possible that the mother secretes
wax or water through her cuticle and this in turn is absorbed by the larvae.
Preliminary research using radioactive isotopes (tritium) shows that the
radioactive marker does pass from the mother to the larvae (Vannini et al. 1985).

Table 4. — Mortality rates at 10% RH of larvae on the mother (OM), without the mother and
aggregated (WM), without mother and isolated from each other (IS). Number of broods = 13.
T = statistics of Wilcoxon matched-pairs test (two-tailed). N — number of matched pairs whose
difference is not zero.

treatments:

OM:

OM: 36/ 102 (35.3%)

WM; 55/104 (52.9%)

IS: 63/106 (59.4%)

OM vs WM

OM vs IS

WM vs IS

N=

12

11

10

T=

14

7

23

P

<0.005

<0.002

n.s.

46

THE JOURNAL OF ARACHNOLOGY

ACKNOWLEDGMENTS

We wish to thank Profs. L. Pardi and A. Ercolini for reading the manuscript.
This study was supported by a grant of the Ministero della Pubblica Istruzione
(fondi local! 60%).

LITERATURE CITED

Alexander, A. J. 1977. Parental behaviour in scorpions. Abstr. 15th Int. Ethol. Conf., Bielefeld 1977.

Angermann, H. von. 1977. Uber Verhalten, Spermatophorenbildung und Sinnesphysiologie von
Euscorpius italicus Hbst. und verwandten Arten (Scorpiones, Chactidae). Z. Tierpsychol., 14:276-
302.

Le Pape, G. 1974. Sur quelque aspects des relations mere-jeunes chez trois especes de scorpions
Buthidae. Rev. comportement Anim., 8:261-264.

Maury, E. A. 1969. Observaciones sobre el ciclo reproductivo de Urophonius brachycentrus (Thorell
1877) (Scorpiones, Bothriuridae). Physis, 29:131-139.

Polls, G. A., W. D. Sissom and S. J. McCormick. 1981. Predators of scorpions: field data and a
review. J. Arid Environments, 4:309-326.

Torres, F. and H. Heatwole. 1967. Factors influencing behavioral interaction of female parent and
offspring in scorpions. Caribbean J. Sci., 7:19-22.

Ugolini, A. and M. Vannini. 1983. Larval behavior and phylogenetic relationships among scorpions.
J. Arachnol., 1 1:433-436.

Vannini, M. and A. Ugolini. 1980. Permanence of Euscorpius carpathicus (L.) larvae on the mother’s
back. Behav. Ecol. Sociobiol., 7:45-47.

Vannini, M. and A. Ugolini. 1981. Parental care in Euscorpius (Arachnida, Scorpiones). 17th Int.
Ethol. Conf., Oxford, 1981.

Vannini, M., A. Ugolini and C. Marucelli. 1978. Notes on the mother-young relationship in some
(Scorpiones, Chactidae). Monitore zool. Ital., (N. S.) 12:143-154.

Vannini, M., M. Balzi, A. Becciolini, 1. Carmignani and A. Ugolini. 1985. Water exchange between
mother and larvae in scorpions. Experientia, 41:1620-1621.

Williams, S. C. 1969. Birth activities of some North American scorpions. Proc. California Acad. Sci.,
37:1-24.

Manuscript received October 1984, revised July 1985.

Raven, R. 1986. A revision of the spider genus Sason Simon (Sasoninae, Barychelidae,
Mygalomorphae) and its historical biogeography. J. Arachnol., 14:47-70.

A REVISION OF THE SPIDER GENUS SASON SIMON
(SASONINAE, BARYCHELIDAE, MYGALOMORPHAE)
AND ITS HISTORICAL BIOGEOGRAPHY

Robert J. Raven

Queensland Museum
Gregory Terrace, Fortitude Valley
Queensland, Australia

ABSTRACT

The barychelid spider genus Sason is revised and includes six valid species: the type species, S.
robustum (O. P.-Cambridge 1883), 5'. andamanicum (Simon 1888), S. colemani sp. nov., S.
maculatum (Roewer 1963), S. pectinatum Kulczyhski 1908, and S. seychellanum Simon 1898. Sason
cinctipes (Pocock 1892) and S. armatoris Pocock 1900 are newly synonymized with 5'. robustum (O.
P.-Cambridge 1883), and Chrysopelma Roewer 1963 with Sason. Rhianus (- Rhianodes) and
Monodontium are transferred to the Barychelinae. Sason occurs in the Seychelles, India, Ceylon, the
Andaman Islands, New Guinea, to the islands of the Northwestern Pacific, and in northern Australia.
Its distribution is similar to that of other Indo-Pacific taxa; a vicariance hypothesis is proposed for
its historical biogeography.

INTRODUCTION

Sason Simon 1887 is a very distinctive barychelid genus that has included six
species distributed from the Seychelles in the western Indian Ocean, northwards
in the Andaman Islands, in Ceylon, southern India, to New Guinea (Roewer
1942). In general appearance, these spiders resemble migids (Simon 1892). They
are small, compact, and stout-legged, lack a strong rastellum and, unlike many
barychelids, their eyes are usually not on a tubercle and the eye group is
rectangular. The synonymy of Sason includes two generic names that allude to
the regal appearance of the spider: Sason, an abbreviation of the biblical name
Samson; and Sarpedon, (the first given name and a homonym) the legendary king
at the seige of Troy (Bonnet 1954). Roewer (1963) described a new genus,
Chrysopelma, from the Mariana Islands but judging by the number of genera
mentioned in his discussion he was unable to determine the closest relative. Sason
was not mentioned but is here considered its senior synonym.

The subfamily name Sasoninae derives from a tribe made by Simon (1892)
solely for Sason but Rhianodes (- Rhianus Thorell but invalidly emended to
Rianus by Simon), a putatively related genus of which Simon had seen no
material, was listed as “genus invisum et incertae sedis'* under the Sasoneae.
Simon (1903) made no further comment about “Rianus.” Roewer (1942) reflected
Simon’s initial decision by including Rhianodes in the Sasoninae and extended
the group to include the New Guinea genus Monodontium. The description of

48

THE JOURNAL OF ARACHNOLOGY

Monodontium Kulczyhski followed that of Sason pectinatum but Kulczyhski
(1908) associated the former with Barychelus and therefore the Barychelinae.
Since then, Benoit (1966, 1978) described further material of Sason seychellanum.

Most subsequent authors have followed Pocock (1903) who elevated Simon’s
Barychelinae to family rank and his tribes (including the Sasoneae) to
subfamilies. According to Bristowe (1938), Kishida (1930) elevated the Sasoninae
to a family without explanation. Benoit (1964) placed the Sasoninae in the
synonymy of the Barychelinae but continued to recognize the other two
barychelid subfamilies — Diplothelinae and Leptopelmatinae. Raven (1985)
continued to maintain the Sasoninae as a subfamily of the Barychelidae.

MATERIALS AND METHODS

Except for eye data, given in ocular eyepiece units, all measurements are in
millimeters. For brevity, several abbreviation techniques are used. Characters
present in all species, e.g., scopulae on metatarsi and tarsi I and II, are stated
only in the generic description. Only the presence of spines on legs is noted.
Rather than repeat the measured interval or object, a numeral is used to denote
the character. The numeral is followed by a comma and the value(s) of the
interval(s) in that specimen. For eyes, numerals denote the following: 1, number
of rows; 2, width of group at its midlength/ head-width through the same point;
3, ratio of front width: back width: length; 4, ratio of AME: ALE: PME: PEE;
5, ratio of MOQ (median ocular quadrangle) front width: back width: length; 6,
minimum eye interspaces of AME-AME, AME-ALE, ALE-PLE, ALE-ALE,
PME-PLE, PME-PME, respectively.

Leg segments in spine descriptions are abbreviated to their first two letters: fe,
femur; pa, patella; ti, tibia; me, metatarsus; ta, tarsus. Spine statements are
standard for the Araneae; other techniques not mentioned above are given in
Raven (1984).

Trichobothria: 1, approximate number per row and extent on tibiae; 2,
approximate number and extent on metatarsi; 3, number of clavate (c) and
filiform (f) on tarsi.

Spinnerets: 1-3, refer to posterior median spinnerets; 1, length; 2, mid-width;
3, separation of bases; 4-7 refer to lengths of segments of posterior lateral
spinnerets; 4, basal; 5, middle; 6, apical; 7, total.

SYSTEMATICS
SASONINAE SIMON

Diagnosis. — The Sasoninae differ from most other barychelids by the
combination of the conical apical segment of the posterior lateral spinnerets, very
low or absent eye tubercle, and edentate paired claws of males, and from
Ammonius by the presence of a row of cuspules on the labium in females and
the short cymbium in males.

Apical segment of posterior lateral spinnerets short, conical. Four spinnerets.
Eye group about twice as wide as long, not on distinct tubercle. Paired claws of
males with few teeth in one row or bare.

RAVEN — REVISION OF SASON

49

Genera included. — Sason Simon 1887, Cosmopelma Simon 1889, and
Paracenobiopelma Feio 1952.

Sason Simon

Sarpedon O. R-Cambridge 1883:353. Type species by monotypy Sarpedon rohustum O. P.-Cambridge
1883.

Sason Simon 1887:195. Replacement name for Sarpedon preoccupied in the Coleoptera by Sarpedon
Bonvouloir 1870.

Satzicus Simon 1888:286. Type species by monotypy Satzicus andamanicum Simon 1888. First
synonymized by Simon 1892.

Oecophloeus Pocock 1892:49. Type species by monotypy Oecophloeus cinctipes Pocock 1892. First
synonymized by Simon, 1892.

Chrysopelma Roewer 1963:113. Type species by original designation Chrysopelma maculata Roewer
1963. NEW SYNONYMY.

Diagnosis. — Sason differs from Paracenobiopelma by the absence of a distinct
clypeus, and from Cosmopelma by the presence of a line of cuspules on the
anterior edge of the labium in females.

Description. — Small, strongly patterned spiders. Carapace glabrous but with
numerous short bristles, especially in males. Caput low but arched medially.
Thoracic region slopes down from broad, shallow, slightly procurved or recurved
fovea. Eyes in three rows or two rows with strongly procurved front row. Back
row more or less straight. Eye group about twice as wide as long, rectangular.
Eye tubercle absent or low, and if present, usually excludes ALE. Clypeus absent.
Chelicerae short, sloping, with one row of teeth on furrow. Rastellum absent or
with two to four short spines. Maxillae rectangular; anterior lobe not
differentiated; heel acute, rounded; few cuspules in females. Lyra absent. Labium
rectangular, anterior edge straight, lateral edges almost parallel; males of some
species and all females armed with stout cuspules in line. Cuspules on maxillae
present or absent in males, always present in females. Sternum cordate with two
or three pairs of small, oval to round sigilla touching margin, and on sloping
edge. Labiosternal suture narrow, distinct. Legs stout, sometimes with distinct
annulations. Leg formula 4123. Scopulae entire but thin for full length of
metatarsi and tarsi I, II; divided, distal if present on metatarsi III, IV; divided,
thin if present on tarsi III, IV. Spines generally weak, few in number; often
present on femora, ventral patellae, rarely on metatarsi, never on tarsi. Preening
combs absent. Tarsi of females short, stout. Palpal claw and paired claws without
teeth or with one row of teeth in males and females. Claw tufts small, moderately
dense but never conceal claws entirely. Trichobothria in two short rows extending
to one-half to two-thirds of tibiae; distal group on metatarsi; broad band on tarsi.
Tarsi with both filiform and clavate trichobothria; bothria corrugiform. Tarsal
organ low, domed. Relative lengths of posterior lateral spinneret segments:
basal>middle>apical; apical segment with distal cluster of spigots. Spermathecae
with two receptacula, sometimes apically divided. Tibia I of males with prolateral
distal spur bearing megaspines. Males palp: tibia short; cymbium short, truncate,
undivided; bulb pyriform with tapering embolus.

Distribution and Natural History. — Sason is known from the Seychelles, the
Andaman and Mariana Islands, southern India, Ceylon, northern Australia, and

50

THE JOURNAL OF ARACHNOLOGY

Fig. 1. — Sason robustum (O. R-Cambridge), female, dorsal view. Scale line, 5 mm.

New Guinea. The retreat (fig. 19) consists of a very short tube with a door at
each end (Pocock 1900, Coleman 1981). The outer surface of the retreat is usually
impregnated with particles of soil and leaves.

Species included. — Sason andamanicum (Simon), Sason colemani, sp. nov.,
Sason maculatum (Roewer), Sason pectinatum Kulczyhski, Sason robustum (O.
P.-Cambridge), Sason seychellanum Simon.

Synonymy. — Roewer (1963) placed Chrysopelma in the Leptopelmatinae,
considering it unique in the absence of a clypeus and teeth on the tarsal claws.
However, Roewer’s subfamilial placement appears to be based upon
Petrunkevitch (1928) in which the key is erroneous through the “inversion” of one
of the key couplets of Simon (1903). Simon (1903) separated the Sasoninae plus
Leptopelmatinae by their "'area oculorum compactilis’,'" whereas Petrunkevitch
(1928) separated the Sasoninae from the Leptopelmatinae plus Barychelinae by

RAVEN — REVISION OF SASON

51

Fig. 2. — Sason robustum (O. R-Cambridge), male, dorsal view. Scale line, 5 mm.

52

THE JOURNAL OF ARACHNOLOGY

the non-compact eye group. Presumably, had Roewer (1963) used Simon (1903)
Chrysopelma maculata would have been placed in or related to Sason.

Relationships. — Sason uniquely shares with Paracenobiopelma the line of
cuspules on the anterior edge of the labium of females. (The maxillae have fewer
cuspules than the labium.) Because that condition is unique in the Barychelidae,
it is considered a synapomorphy and not a retention of the densely cuspulate
labium that is plesiomorphic in the Theraphosoidina (see Raven 1985).
Cosmopelma shares with Sason and Paracenobiopelma the similarly patterned
carapace, annulated legs, and broad chevrons on the abdomen — also unusual, if
not unique, conditions in the barychelids. Associated with that character
congruence, in both Cosmopelma and Paracenobiopelma the rectangular eye
group is on a tubercle that is very low or absent and, unlike Sason, is separated
from the edge of the carapace by a distinct clypeus. Males of Paracenobiopelma
and Sason also share the absence of biserially dentate claws — a condition found
widely in other barychelids and considered the family autapomorphy. Thus, one
synapomorphy of Sason and Paracenobiopelma is presumed to be the reduced
or edentate paired claws. Males of Cosmopelma are not known. The absence of
a clypeus and abdominal pattern in Cosmopelma dentata Fischel 1927 (the type
now lost) are sufficient to exclude it from the genus; the species may be correctly
placed in Trichopelma. Cosmopelma lacks the linear cuspules of both Sason and
Paracenobiopelma. Raven (1985) concluded that two hypotheses were equally
parsimonious. First, either the linear cuspules were gained in the Sasoninae and
lost only in Cosmopelma, and the clypeus was lost in the Sasoninae plus
Barychelinae but regained in Cosmopelma and Paracenobiopelma. The second
hypothesis is that the linear cuspules are synapomorphic for Sason plus
Paracenobiopelma (one step) and that the clypeus was regained in each of
Paracenobiopelma and Cosmopelma (two steps). Outgroup comparison supports
the latter hypothesis that, internal to the Sasoninae, is unaffected by outgroup
changes and uses two steps. The first, although also using two steps, is more
parsimonious because one of the steps applies also to the Barychelinae. In either
hypothesis the synapomorphies of the Sasoninae are the cephalic, leg, and
abdominal patterns, plus the edentate claws of males, but the first hypothesis is
correlated with the synapomorphy of the linear cuspules. If the presence of a
clypeus is considered plesiomorphic, intrafamilial homoplasies of the shape and
dentition of the labium and reduction of the posterior lateral spinnerets increase.

HISTORICAL BIOGEOGRAPHY OF SASON

The biogeography of Sason was little discussed when the genus was known
only from areas within and around the Indian Ocean. Pocock (1903) believed that
Sason arose near the Seychelles, India, and Ceylon prior to their separation “in
very early times (p. 353)” but attributed their occurrence in the Andaman Islands
to artificial introduction by man and suggested that the same may be true for
their occurrence in the Seychelles. Legendre (1979) suggested that the arboreal
nest of Sason allowed for its transport as flotsam in ocean currents. Raven (1980)
admitted that possibility in mygalomorphs generally and suggested the most
parsimonious mechanism would require that a gravid female made the voyage.
Main (1981a) attributed the origin of much of Australia’s mygalomorph fauna to

RAVEN — REVISION OF SASON

53

northern or southern invasions. In contrast, Kikkawa et al. (1981) found little
support among distributions of birds, beetles, and butterflies for large scale
invasions (through the Cape York corridor, at least). Williams (1981) also found
the invasive component of Crustacea in Australia was very small. Coyle (1983,
1985) observed that two other mygalomorph genera, Sphodros and Ummidia, can
disperse aerially and presented evidence that water gaps may be traversed during
such ballooning.

Main (1981b), although stating that mygalomorphs generally have “poor
powers of dispersal”, later made the contradictory claim that the occurrence of
mygalomorphs in New Guinea indicates that the groups “show exceptional
capacity for dispersal or are ecologically aggressive with an unusually high
invasive potential” (p. 587). Later, the theraphosids (accounting for about two-
thirds of the known species of mygalomorphs and to which Sason was
erroneously attributed) are considered “an extraordinarily “mobile” group (p.
589).” Neither the mechanism for the dispersal nor evidence for it has ever been
presented.

Alone, the dispersal ability of a group is insufficient to support a dispersal
explanation of their biogeography; areas of endemism (sometimes quite small) of
birds, butterflies, and fish falsify that notion repeatedly. As Platnick (1981) points
out, arachnologists have long been forced to concede the potential, however
small, of spiders to disperse. Thus, some mygalomorphs may disperse beyond
normal established ranges but that potential is not often realized. Clearly, once
an organism transcends one barrier it could be expected to transcend all similar
ones and so attain a very wide distribution.

Platnick (1981) rejected the need to consult dispersal abilities in discussing
biogeographies; the degree of endemism is the most informative component.
Sason species are endemic to relatively small areas. The species are
morphologically distinct. Intrarelationships of Sason species should therefore
reflect vicariance events in the historical biogeography of the genus.

Intrarelationships. — At present, two species groups are evident in Sason. S.
pectinatum and S. maculatum share the back eye row being wider than the front
(figs. 21, 27); and S. andamanicum and S. colemani share the “retreated” edge
of the distal first tibia (figs. 9, 11) and the absence of labial cuspules of the male
(figs. 7, 15). The plesiomorphic eye condition of barychelids is rectangular, thus
the rhomboidal condition is apomorphic. In S. robustum, the cuticle at the base
of the male tibial spur is not invaginated but the spur arises on a separate process
(fig. 35); the spur is the same shape as in S. andamanicum and S. colemani. No
spur is present in Paracenobiopelma, thus it cannot be used as an outgroup for
this character. It is not possible to establish whether the spur was
plesiomorphically distal and moved proximally with the resultant invagination
closing over in S. robustum. The correlation of the invaginated tibial cuticle with
the absence of cuspules (present in Paracenobiopelma and S. robustum) on the
labium of males of S. andamanicum and S. colemani indicates that it is
apomorphic; that is accepted here. Thus, three groups are evident: S. robustum
and S. seychellanum, which lack a synapomorphy, S. andamanicum plus S.
colemani, and S. pectinatum and S. maculatum. Without males of all species no
further grouping is possible. I predict that a synapomorphy will be found linking
S. robustum and S. seychellanum, and another linking S. andamanicum plus S.
colemani and S. pectinatum plus S. maculatum.

54

THE JOURNAL OF ARACHNOLOGY

At present, two areas of endemism (plus one default area) are recognizable:
eastern Indian Ocean plus Australia, northwestern Pacific Ocean, and western
Indian Ocean, respectively. Given an original widely spread southern occurrence,
the rafting of India would have produced two of those areas (eastern and western
Indian oceanic areas) and the uplifting of New Guinea would have produced the
third. Hence, the areas are consistent with geological events. Full resolution of
the area cladogram will come with recognition of synapomorphies of N. robustum
plus S. seychellanum, and of the other two groups.

KEY TO THE SPECIES
Females

1 . No scopuliform hairs on tarsi HI ........................................ 2

Scopula present but may be only thin and divided by setae on tarsi HI .......... .

S. robustum

2. Ocular area as wide in front as behind (Fig. 47) ........................... 3

Ocular area clearly wider behind than in front (Fig. 21) .4

3. Rastellum completely absent, rastellar area without even thick short setae .......

5'. seychellanum

At least two differentiated thick setae form weak rastellum (e.g., Fig. 29) ........

N. colemani

4. Eye group about twice as wide in front as long or wider ........... Y. pectinatum

Eye group clearly less than twice as wide in front as long (Fig. 21) ...............

— ........ — N. maculatum

Males

1. Cuspules present on labium and maxillae. , ,S. robustum

Cuspules absent on labium and maxillae 2

2. Teeth present on paired claws . N. colemani

Teeth absent on paired claws S. andamanicum

Sason andamanicum (Simon)
Figs. 3-9 Table 1

Satzicus andamanicum Simon 1888:287.
Sason andamanicum: Simon 1892:130.

Type. — Holotype male, Andaman Is., Port Blair (R. D. Oldham, MNHNP No.
9763, examined).

Diagnosis. — Differs from N. colemani by the absence of teeth on the paired
claws of the male and a rastellum.

Fovea recurved. Rastellum absent. Cuspules absent on maxillae and labium of
male. Scopulae thin, ventral, divided by scattered setae on tarsi HI, rudimentary
on tarsus IV; few distal scopuliform hairs on metatarsi HI. Tibia I of male with

RAVEN — REVISION OF SASON

55

Fig. 3. — Map showing distribution of Sason species. Symbols (from left): S. seychellanum (star),
N. robustum (black star); S. andamanicum (arrow); N. maculatum (hollow circle); S. pectinatum (solid
square); S. maculatum (dot).

prolateral distal spinose process directed entally. Palpal bulb squat. Spines
present on femur and tibia I. Paired claws without teeth. Females unknown.

Description. — Holotype male MNHNP No. 9763. Carapace 5.00 long, 4.75
wide. Abdomen 4.67 long, 3.42 wide.

Color in alcohol: carapace yellow brown with slight brown mottling on caput;
legs with faded pattern, without brown annulations, probably faded; abdominal
pattern faded. The leg annulations characteristic of Sason may be absent because
the type was in poor condition when first described.

Carapace: fovea broad, recurved. Bristles: about 20, long, erect on clypeal edge;
5 anteromedian; numerous long, erect, and thick, forming uniform covering on
caput and interstrial ridges. Eyes: all but ALE on low tubercle; 1, 3; 2, 0.49; 3,
73:77:39; 4, 24:17:9:12; 5, 52:56:31; 6, 7, 11, 23, 45, 2, 43. Chelicerae: long bristles
in two narrow dorsal bands; rastellum absent; 7 spaced promarginal teeth, 1 small
basomesal tooth on furrow.

Labium: 0.44 long, 0.88 wide; without cuspules. Maxillae: 1.20 long in front,
1.60 long behind, 0.72 wide; without cuspules. Sternum: 2.88 long, 2.28 wide;
only posterior and middle sigilla evident.

Legs: (Table 1). 413. Scopulae: thin, ventral, divided by scattered setae on
tarsus III; rudimentary on tarsus IV; metatarsus III with few distal scopuliform
hairs; tibia I with predistal prolateral process bearing entally directed megaspine.
Spines. Leg 1: fe, d2; ti, v4 + megaspine. Leg 2, missing. Leg 3, 0. Leg 4: fe,
d4. Palp, 0. Claws: bare. Trichobothria: 1, 10 for half; 2, 10 in distal one-third;
3, 6 c. and 6 f. Palp: bulb squat, pyriform, with narrow tapering embolus.

Spinnerets: 1, 0.34; 2, 0.14; 3, 0.14; 4, 0.72; 5, 0.22; 6, 0.12; 7, 1.06.

Distribution. — S. andamanicum is known only from the Andaman Islands in
the Bay of Bengal.

Material examined. — Only the type.

56

THE JOURNAL OF ARACHNOLOGY

Fig. 4-9. — Sason andamanicum Simon, holotype male: 4, eyes, dorsal view; 5, palpal bulb, ventral
view; 6, cephalothorax and abdomen, dorsal views; 7, sternum, maxillae, and labium; 8, palpal tibia,
cymbium, and bulb, retrolateral view; 9, tibia I, ventral view. All scale lines, I mm.

Sason colemani, new species
Figs. 3, 10-20 Table 2

Types. — Holotype male, paratype female. Australia: Queensland: Cairns, in
swamp on trees in short tubes covered by bark particles (17.xi.l980, N. C.
Coleman, QM S 1311, 1312); female paratype, same data (QM S 1313).

Table I. — Leg measurements of Sason andamanicum. Legs 2 are missing. Values are for holotype
male.

Leg 1

Leg 3

Leg 4

Palp

Femur

4.83

4.25

5.33

2.58

Patella

2.75

2.33

2.83

2.08

Tibia

3.50

3.50

5.00

2.08

Metatarsus

2.75

3.00

3.92

Tarsus

1.50

1.33

1.50

0.67

Total

15.33

14.41

18.58

6.91

Figs. 10-20. — Sason colemani, new species, holotype male, paratype female: 10-15, male; 10,
cephalothorax, dorsal view; 11, tibia and metatarsus I, prolateral view; 12, palpal tibia, cymbium, and
bulb, retrolateral view; 13, 14, abdomen, dorsal view (13), ventral view (14); 15, sternum, maxillae,
and labium; 16-20, female; 16, abdomen, ventral view; 17, cephalothorax and abdomen, dorsal view;
18, sternum, maxillae, and labium; 19, nest showing two doors; 20, spermathecae. Scale line: figs.
10-12, 15-18 = 1 mm; figs. 13, 14, 16 = 2 mm; fig. 19 = 5 mm; fig. 20 = 0.5 mm.

Diagnosis. — Differs from 5’. andamanicum by the presence of teeth on the
claws of the male, two coniform rastellar spines, and more spines on the legs.

Fovea almost straight. Rastellum consists of 2-3 coniform spines. Scopulae in
females divided on metatarsi and tarsi I, II; absent on metatarsi and tarsi III,
IV of males and females. Tibia I of males with prolateral and distal megaspine
on low ectally directed spur. Palpal bulb pyriform. Spines present on femora I-
IV and tibiae I and II of males, and ventral patellae and metatarsi I, II of
females.

Description. — Holotype male QM S 1311. Carapace 2.43 long, 2.28 wide.
Abdomen 2.50 long, 1.95 wide.

58

THE JOURNAL OF ARACHNOLOGY

Color in alcohol: carapace light brown with darker brown areas near eyes, on
caput, and interstrial ridges; chelicerae cream with brown areas; abdomen
dorsally brown with large white paired areas medially forming two large and
small lateral pairs in median triangular area, with irregular white spots
posteriorly; ventrally cream with small brown areas on booklung covers and
laterally; sternum edge brown; legs cream coloured with brown annotations on
distal and lateral femora, prolateral patellae, tibial joints, and distal metatarsi.

Carapace: fovea shallow, centrally placed, broad, transverse almost straight
with slightly recurved ends. Bristles: short curved in fringe on lateral margins,
posterior interstrial ridges, and few on anterior ridges; 5 thick anteromedian; 6
on clypeal edge; 1 long between AME; several between PME. Eyes: tubercle low;
1, 2; 2, 0.62; 3, 33:35:18; front row strongly procurved, back row recurved; 4,
7:9:4:5; 5, 19:25:12; 6, 3, 3, 6, 20, 1, 15. Chelicerae: short, almost glabrous;
rastellum consists of two distinct coniform bristles distally; 6-7 teeth on
promargin, no teeth or granules visible elsewhere.

Labium: 0.18 long, 0.50 wide; separated from sternum by shallow groove;
cuspules absent. Maxillae: 0.53 long, 0.30 wide; cuspules absent. Sternum: 1.30
long, 1.18 wide; posterior sigilla round, marginal, 0.06 across; other sigilla not
discernible.

Legs: (Table 2). 4123. Scopulae: present but very thin on short shiny tarsi and
metatarsi I, II; entirely absent on metatarsi and tarsi III, IV. Tibia I with small
megaspine on low mound prolaterodistally. Spines. Leg 1; fe, d2; pa, v2; ti, p2
v6 + megaspine; me, v4. Leg 2: ti, v6. Leg 3: fe, d3. Leg 4: fe, d5. Palp: pa, vl.
Claws: three teeth on ental edge of both claws on legs I; long curved with two
teeth on leg IV. Trichobothria: 1, 2-5 for half; 2,6; 3, 7-9 c. and f. Palp: bulb
spheroidal; embolus tapers to point, grooves absent.

Spinnerets: 1, 0.13; 2, 0.08; 3, 0.03; 4, 0.28; 5, 0.13; 6, 0.08; 7, 0.49.

Paratype female QM S 1312. Carapace 1.90 long, 1.64 wide. Abdomen 4.88
long, 3.75 wide.

Color in alcohol: carapace, dorsal abdomen, and legs similar to male but with
more distinct pattern; carapace pattern darker with paler areas behind eye group;
abdomen ventrally pallid with brown areas laterally and medially between
posterior booklungs; sternum entirely pallid.

Carapace: fovea broad, straight. Bristles: in three lines anterior to fovea; 3 very
long thick anteromedian; 1 long between AME; 2 long between PME; 6 on
clypeal edge. Eyes: not on tubercle; 1, 2; 2, 0.47; 3, 24:26:13; front row strongly
procurved, back row slightly recurved; 4, 8:10:5:7; 5, 12:17:9; 6, 3, 3, 5, 14, 1,
12. Chelicerae: short, geniculate, pallid with fine brown hairs; rastellum consists
of 2-3 coniform spines distally; 5 thick teeth, 5-7 granules basomesally on
promargin.

Labium: 0.75 long, 0.40 wide; 9 thick pointed cuspules. Maxillae: 0.70 long,

0. 55 wide; with 5-6 cuspules on inner edge, and thin spinules posteriorly.
Sternum: 1.90 long, 1.63 wide; posterior and middle sigilla evident as small paired
depressions.

Legs: (Table 2). 4123. Scopulae: distinct but thin, divided on metatarsi and tarsi

1, II; entirely absent on metatarsi and tarsi III, IV. Spines. Lemoral ‘spines’ are
long, thick, curved bristles. Leg T. fe, pi d3; pa, vl; ti, v3; me, vl. Leg 2: fe,
d3; pa, vl; ti, v3; me, v2. Leg 3: fe, d3. Leg 4: fe, d3. Palp: fe, d2 pa, v2; ti.

RAVEN — REVISION OF SASON

59

Table 2. — Leg measurements of Sason colemani. Values are for holotype male with allotype female
in parentheses.

Leg 1

Leg 2

Leg 3

Leg 4

Palp

Femur

2.28(1.72)

2.24(2.16)

2.08(2.04)

2.32(2.72)

1.05(1.48)

Patella

1.32(1.52)

1.16(1.20)

1.08(1.28)

1.02(1.72)

0.75(1.20)

Tibia

1.76(1.40)

1.72(1.40)

1.64(1.40)

2.32(2.36)

0.83(1.00)

Metatarsus

1.56(1.08)

1.48(1.00)

1.40(1.20)

1.84(1.60)

Tarsus

0.76(0.64)

0.68(0.76)

0.68(0.60)

0.68(0.72)

0.48(1.04)

Total

7.68(6.36)

7.28(6.52)

6.88(6.52)

8.18(9.12)

3.1 1(4.72)

v6. Claws:

long curved,

without teeth.

Trichobothria:

1, 5-8 for

half; 2, 5-6; 3,

4-6 c. and 9 f.

Spinnerets: 1, 0.20; 2, 0.08; 3, 0.20; 4, 0.33; 5, 0.18; 6, 0.08; 7, 0.61.
Spermathecae: each a short lobe with or without small, distal, ectal lobe.

Distribution, Habitat, and Remarks. — S. colemani is known only from a
natural swamp in the Botanical Gardens in Cairns, north Queensland. The retreat
consists of a very short tube with a door at each end (Fig. 19); as one door opens
the other is pressed closed. Retreats were found on the bark of trees. Two other
barychelid genera, Cyphonisia and Paracenobiopelma, make similar but slightly
longer retreats with more distance between the door hinges (Blandin and Celerier
1977, Feio 1952). The similar retreat is probably another synapomorphy of Sason
and Paracenobiopelma.

Material examined. — Only the types.

Sason maculatum (Roewer)

Figs. 3,21-26 Table 3

Chrysopelma maculata Roewer 1963:1 13, figs. 3d-f. NEW COMBINATION.

Types. — Holotype female, Korori, Palau Island (26. xi. 1947, H. Y. Dybas,
USNM, examined). Paratypes: female, Palau Island (August 1945, H. Y. Dybas,
SMF No. 12758); female, Ponape (March 1948, H. F. Dybas, deposition
unknown).

Diagnosis. — Differs from 5'. pectinatum in the relatively longer eye group.

Fovea broad with recurved ends. Rastellum consists of two coniform spines.
Scopulae in females absent on metatarsi and tarsi III, IV. Spines present on all
femora, and tibiae and metatarsi III, IV. Eye group about 1.5 times wider in front
than long, and clearly wider behind than in front.

Description. — Paratype female SMF No. 12758. Supplementary description to
Roewer 1963. Carapace 3.78 long, 2.72 wide. Abdomen 5.42 long, 3.58 wide.

Color in alcohol; carapace yellow brown with small brown areas on caput and
lateral margins; narrow brown areas radiating from fovea along interstrial ridges;
chelicerae yellow brown; legs yellow brown with large brown spots on posterior
femora I, II, prolateral and retrolateral on femora III, IV, prolaterally on patellae
and tibiae I, II, and dorsally on metatarsi I-IV; lateral brown markings on
patellae and tibiae III, IV; abdomen dorsally brown with large irregular brown
spots, ventrally fawn with brown markings posteriorly.

Carapace: fovea broad, straight with recurved ends. Bristles: one foveal pair;
3 posteriorly directed pairs in front of fovea; several on lateral caput; 1 long

60

THE JOURNAL OF ARACHNOLOGY

Figs. 21-26. — Sason maculatum (Roewer), paratype female: 21, eyes, dorsal view; 22, spermathecae;
23, 24, abdomen, ventral view (23), dorsal view (24); 25, cephalothorax, dorsal view; 26, sternum,
maxillae, and labium. Scale lines, 0.5 mm, except Fig. 22, 0.1 mm.

between AME; 5 long, curved between ALE. Eyes; not on tubercle; 1, 2; 2, 0.51;
3, 42:50:28; front row strongly procurved, back row slightly recurved; 4,
12:12:7:10; 5, 25:34:18; 6, 5, 6, 11, 21, 1, 22. Chelicerae: sparse bristles on
prodorsum; rastellum consists of 2 coniform spines medially on distal edge; 6
large and one small tooth, several small granules basomesally on promargin.

Labium: 0.66 long, 0.38 wide; with 10 blunt cuspules. Maxillae: 0.90 long in
front, 1.20 long behind, 0.52 wide; with 4-6 cuspules. Sternum: 1.84 long, 1.48
wide; all sigilla oval, touching margin.

RAVEN — REVISION OF SASON

61

Table 3. — Leg measurements of Sason maculatum. Values are for paratype female.

Leg 1

Leg 2

Leg 3

Leg 4

Palp

Femur

2.08

2.16

1.84

2.60

1.60

Patella

1.56

1.52

1.28

1.64

1.24

Tibia

1.36

1.32

1.32

2.00

1.00

Metatarsus

1.00

1.08

1.20

1.68

Tarsus

0.72

0.64

0.64

0.84

1.08

Total

6.72

6.62

6.28

8.76

4.92

Legs: (Table 3). 4123. Scopulae: entirely absent on metatarsi and tarsi III, IV.
Spines. Leg 1: fe, d5; ti, v5. Leg 2: fe, d5; pa, vl; ti, v5. Leg 3: fe, d5; ti, v7.
Leg 4: fe, d5; ti, v4; me, v6. Palp: fe, d3; pa, v3; ti, v7. Claws: palpal claw short
with two small teeth; others without teeth. Trichobothria: 1, 5-6 for half; 2, 6;
3, two divided bands of 4-6 c., 8 f.

Spinnerets: 1, 0.20; 2, 0.10; 3, 0.16; 4, 0.42; 5, 0.38; 6, 0.14; 7, 0.94.
Spermathecae: each receptaculum with short lateral lobe.

Distribution. — S. maculatum is known only from Saipan in the Marianas, and
Kusaie, Ponape, Truk, and Woleai in the Caroline atolls, north of New Guinea.
Material examined. — Only the types.

Sason pectinatum Kulczyhski
Figs. 3, 21-29

Sason pectinatum Kulczyhski 1908:450.

Type. — Holotype juvenile, northeastern New Guinea (1896, L. Biro. Museu
Nationalis Hungarici, examined).

Diagnosis. — Differs from S. pectinatum in the comparatively wider eye group
and stronger spines on legs I, IT

Fovea recurved. Rastellum consists of one or two spines. Scopulae of females
absent on metatarsi and tarsi III, IV. Spines only on femora to metatarsi I, II,
and femora and metatarsi III, IV. Eye group about twice as wide in front as long,
slightly wider behind than in front. Palpal claw with one or two small teeth.
Adults unknown.

Description. — Carapace 2.04 long, 1.80 wide. Abdomen 2.68 long, 1.84 wide.

Color in alcohol: carapace yellow brown with some faded areas on caput and
lateral margins; chelicerae yellow brown; legs yellow brown with faded
annulations on all patellae, tibiae, and metatarsi; abdominal pattern faded, only
medial brown area evident; ventrally pallid, also with no pattern.

Carapace: bristles: 3 anteromedian; 2 long, curved between ALE. Fovea broad,
straight. Eyes: not on tubercle; 1, 2; 2, 0.48; 3, 34:37:16; front row procurved,
back row slightly recurved; 4, 8:8:5:4; 5, 21:26:12; 6, 3, 3, 8, 21, 2, 18. Chelicerae:
bristles sparse on prodorsum; rastellum consists of 1 or 2 coniform spines
medially on distal edge; promargin with 6 large and one small tooth, 6-8 small
teeth basomesally.

Labium: 0.24 long, 0.48 wide; with one small, blunt and nine pointed cuspules.
Maxillae: 0.56 long in front, 0.70 long behind, 0.38 wide; with 7-12 cuspules.
Sternum: 1.24 long, 1.08 wide; sigilla not evident.

62

THE JOURNAL OF ARACHNOLOGY

Figs. 27-29. — Sason pectinatum Kulczynski, holotype juvenile; 27, eyes, 28, spinnerets, ventral view;
29, chelicerae, sternum, maxillae, and labium, ventral view. Common scale, 1 mm.

Legs. 4123. Scopulae: entirely absent on metatarsi and tarsi III, IV. Spines:
strong ventrally on legs I, II. Leg 1: pa, vl; ti, v5; me, vl. Leg 2: fe, d3; pa,
vl; ti, v5; me, vl. Leg 3: fe, d5; me, v3 (weak). Leg 4: fe, d4; me, v2 (weak).
Palp: fe, proventral 2; pa, p2, vl; ti, v6. Claws: palpal claw short with one or
two small teeth; others without teeth. Trichobothria: similar to S. maculatum.

Spinnerets: 1, 0.16; 2, 0.06; 3, 0.08; 4, 0.24; 5, 0.20; 6, 0.10; 7, 0.50.
Spermathecae: not evident.

Distribution and Remarks. — V. pectinatum is known only from north-eastern
New Guinea. Because the type is a juvenile, the above description and figures are
less informative than those of mature specimens, and no leg measurements are
given.

Material examined. — Only the type.

Sason robustum (O. P.-Cambridge)

Figs. 1-3, 30-45 Table 4

Sarpedon robustum O. R-Cambridge 1883:354, plate 36, fig. la-f.

Sason robustum Karsch 1891:273; Simon 1892:129, 130; Pocock 1900:173.

Oecophloeus cinctipes Pocock 1892c;49, pi. Ill, figs. 1, 2 (syntype female, Ceylon, Punduloya River,
E. E. Green, BMNH No. 1890.10.22.70, examined; female syntype, Kanthalai, BMNH No.

Figs. 30-36. — Sason robustum (O. P.-Cambridge), male: 30, eyes, dorsal view; 31, spinnerets, ventral
view; 32, abdomen, ventral view; 33, palpal bulb, ventral view; 34, cephalothorax and abdomen,
dorsal view; 35, tibia I, ventral view; 36, sternum, maxillae, and labium. All scale lines, 1 mm.

1898.3.21.3, examined; female and juvenile syntypes, Madras Jambunathan, BMNH No.
1923. xii.21. 33-34, examined). NEW SYNONYMY.

Sason cinctipes: Simon 1892:130; Pocock 1904:799.

Sason armatoris Pocock 1900:173; fig. 56 (syntype male, south western India, Travancore, Ponmudi,
coll. Fergerson, BMNH No. 1899.7.11.6, examined; male, Trevandrum, Feb. 1896, BMNH No.
1899.1.17.1). NEW SYNONYMY.

Type. — Lectotype female and paralectotype juvenile, Ceylon (G. H. K.
Thwaites, Hope Museum, Oxford, examined and here designated).

Diagnosis. — S. robustum females differ from those of other Sason species by
the absence of scopula on tarsus IV; males differ from those known of other
species by the presence of cuspules on the maxillae and labium.

Fovea broad, recurved. Rastellum absent. Scopulae: in females, distal, thin on
metatarsus III, sometimes absent on IV; thin but divided by broad setal band on

Figs. Sason rohustum (O. P.-Cambridge), female: 37, cephalothorax, dorsal view; 38,

spermathecae; 39, 40, abdomen, dorsal view (39), ventral view (40); 41, eyes, dorsal view; 42, sternum,
maxillae, and labium. All scale lines, 1 mm.

RAVEN— REVISION OF SASON

65

Figs. 43-45. — Sason robustum (O. P.-Cambridge), holotype female: 43, sternum, maxillae, and
labium; 44, cephalothorax and anterior abdomen, dorsal view; 45, spinnerets, ventral view. All scale
lines, 1 mm.

tarsus III, sometimes absent on IV; in males, divided by narrow band of setae
on tarsi III, absent on tarsi IV and metatarsi III, IV. Tibia I of males with
proventral process and spur. Palpal bulb spheroidal with distinctly demarcated
embolus. Spines present or absent on all femora and tibiae of males and females.

Description. — Male BMNH No. 1899.7.11.6. Carapace 4.64 long, 4.56 wide.
Abdomen 4.00 long, 3.75 wide.

Color in alcohol: carapace orange brown with brown radial marks; chelicerae
similar with darker areas; abdomen, dorsally brown with 3 paired white areas and
three brown inverted V’s, ventrally mottled with yellow brown booklung covers;
laterally brown. Legs with pattern faded, ?without brown annulations.

Carapace: fovea broad, recurved. Bristles: thinly distributed, short, brown; in
one row on lateral margins; 4, thick anteromedian. Eyes: on low tubercle; 1, 3;
2, 0.48; 3, 64:70:39; 4, 19:14:8:11; 5, 45:49:29; 6, 7, 10, 21, 40, 2, 37. Chelicerae:
short, geniculate; stiff, long bristles dorsally; rastellum absent; 1 small and 5 thick
teeth on promargin.

Labium: 0.40 long, 0.96 wide; 6 pointed cuspules anteriorly. Maxillae: 1.20 long
in front, 1.64 long behind, 0.76 wide; with 2 pointed cuspules. Sternum: 2.84
long, 2.08 wide; only posterior and middle sigilla evident — both pairs oval,
marginal.

Legs: (Table 4). 4123. Scopulae: distal two-thirds of metatarsi I, II; divided by
narrow setal band on tarsus III; absent on metatarsi III, IV and tarsus IV. Tibia
I with low prolateral spur set back from cuticle edge. Spines: Leg 1: fe, pi d3;
ti, v4 + megaspine. Leg 2: fe, pi d3; ti, v5. Leg 3: fe, pi d6 rl. Leg 4 missing.
Palp: fe, pi d3; ti, v2. Claws: two small teeth on legs I, II; bare on leg III.

66

THE JOURNAL OF ARACHNOLOGY

Table 4. — Leg measurements of Sason robustum. Values are for holotype female with MNHNP
male in parentheses.

Leg 1

Leg 2

Leg 3

Leg 4

Palp

Femur

3.04(4.83)

3.20(5.08)

3.04(4.67)

4.40(5.83)

2.48(2.83)

Patella

2.24(3.33)

2.48(3.08)

1.92(2.67)

2.80(3.33)

1.84(1.92)

Tibia

1.92(3.92)

2.08(4.17)

2.16(4.00)

3.28(5.17)

1.68(2.42)

Metatarsus

1.52(3.83)

1.28(3.83)

1.84(3.75)

2.56(4.67)

Tarsus

1.20(2.00)

1.04(1.83)

0.88(1.67)

1.12(2.00)

1.76(1.00)

Total

8.92(17.91)

9.09(17.99)

9.84(16.76)

14.16(21.00)

7.76(8.17)

Trichobothria: 1, 10 for half; 2, 12; 15-20 c. and f. Palp: bulb squat, pyriform,
with narrow tapering embolus.

Spinnerets: 1, 0.20; 2, 0.08; 3, 0.08; 4, 0.20; 5, 0.12; 6, 0.08; 7, 0,40.

Lectotype female Hope Entomological Collections, Oxford. Carapace 4.83
long, 4.17 wide. Abdomen 6.00 long, 4.58 wide.

Color in alcohol: carapace, chelicerae, and legs yellow brown; brown
annulations on proximal tibiae I~IV, and distal metatarsi and tarsi. Abdominal
pattern faded.

Carapace: Bristles: long, numerous on caput, fewer on thoracic region; 4
between AME; 4 long anteromedian; 2 long, several short on clypeal edge. Eyes:
not on tubercle; 1, 3; 2, 0.43; 3, 76:79:39; 4, 16:15:9:11; 5, 46:61:25; 6, 13, 13,
19, 45, 3, 54. Chelicerae: small, geniculate; rastellum absent; 7 thick teeth on
promargin, 4 granules basally.

Labium: 0.64 long, 1.16 wide; 8 thick pointed cuspules. Maxillae: 1.64 long in
front, 2.12 long behind, 0.92 wide; 5 blunt cuspules on inner edge, thin spinules
posteriorly. Sternum: 3.20 long, 2.76 wide; sigilloid depressions not evident.

Legs: (Table 4). 4123. Scopulae: lateral and strongly divided by setae on tarsi
111; absent on metatarsi 111, IV and tarsi IV. Spines. Leg 1: pa, vl; ti, v5. Leg
2: pa, vl; ti, pi v5. Leg 3: ti, v4. Leg 4: 0. Palp: pa, v3; ti, v6. Claws: 2-3 small
teeth on single keel; palpal claw (of female MNHNP No. 15063) with one distinct
tooth. Trichobothria: 1, 9; 2, 14; 13 f. and c.

Spinnerets: 1, 0.36; 2, 0.16; 3, 0.16; 4, 0.52; 5, 0.44; 6, 0.20; 7, 1.16.
Spermathecae: not examined.

Distribution and Remarks. — S. robustum is known only from Ceylon and
southern India. Pocock (1900), citing Simon (1892), diagnosed male Sason by the
cuspule-free labia and maxillae. However, Simon appears to have overlooked the
cuspules on the males in his collection and instead based his diagnosis of the
genus on the male of S. andamanicum.

Synonymy. — 5". cinctipes and N. armatoris differ from S. robustum only in
having less dense or less extensive scopulae on metatarsi and tarsi HI and IV,
and also in having spines on the femora — differences that would indicate that S.
cinctipes is a valid species. However, the differences are not congruent. The
extent and density of scopula hairs are not correlated on metatarsi and tarsi HI
and IV. Thus, in the absence of males that have similar characteristics to females
of 5’. cinctipes (spines on the femora and scopulae on metatarsi and tarsi HI and
IV) I conclude that only one species is present.

Material examined. — The types and the following: INDIA, female, 2 juveniles, Haragam (Aug.
1903, BMNH); female, south India, Yercaud, 1200 m (6.iii.l962, E. S. Ross and D. Cavagnaro, CAS);
juvenile, 2.5 km south of Toppur, 340 m (3.iv. 1962, E. S. Ross and D. Cavagnaro, CAS); female
?, Madras (8.iv.l962, E. S. Ross and D. Cavagnaro, CAS). CEYLON, female, Pundul Oya R, (E.

RAVEN — REVISION OF SASON

67

Figs. 46-49. — Sason seychellanum Simon, syntypes female; 46, 48, 49, ZMH specimen; 47, MNHNP
specimen; 46, sternum, maxillae, and labium; 47, eyes, dorsal view; 48, abdomen, dorsal view; 49,
cephalothorax, dorsal view. All scale lines, 1 mm.

E. Green, BMNH 1895.11.14.12.13); juvenile, (Dr. A. Willy, BMNH 1906. 11.14.4.6); female
(Holdeworth collection, BMNH No. 1875.12); male, female (MNHNP No. 15963).

Sason seychellanum Simon
Figs, 3, 46-49 Table 5

Sason seychellanum Simon 1898:370; 1903:915; Hirst 1911:381; Benoit 1966:213; 1978:407, figs, la-

d.

Types. — Lectotype female, Seychelle Islands (A. Brauer, 1895, ZMH,
examined). Paralectotypes: juvenile male, juvenile female same data (in ZMH,
examined); female, same data (MNHNP No. 15.220, examined). NEW
DESIGNATIONS.

Diagnosis. — Differs from S. andamanicum by the complete absence of scopulae
on legs III and IV, and more spinose legs.

Fovea more or less straight. Rastellum consisting of 3 coniform setae. Adult
males unknown. Scopulae in females absent on metatarsi and tarsi III, IV. Paired
claws of legs I, III with two teeth; palpal claw with 4 teeth.

Description. — Lectotype female ZMH. Carapace 3.92 long, 3.24 wide.
Abdomen 3.42 long, 2.58 wide.

Color in alcohol [of MT 122.898]: carapace yellow with black margin and two
roughly V-shaped areas on caput; femora with two brown areas distolaterally and
I-IV also with pair at half length of femora; patellae with distolateral brown
triangles; tibiae with three brown areas dorsally separated by two glabrous ovoid
bands; metatarsi brown in distal one-third. Abdomen dorsally with three large
white paired areas and posteriorly three irregular medial areas becoming smaller
towards spinnerets; laterally mottled; ventrally pallid with incomplete transverse
brown bands posteriorly.

Carapace: fovea broad, straight or very slightly procurved. Bristles: 5 on
clypeal edge; 1 thick between AME; 1 foveal pair; 5 anteromedian; 3-4 groups

68

THE JOURNAL OF ARACHNOLOGY

Table 5. — Leg measurements of Sason seychellanum. Values are for female MT

Leg 1

Leg 2

Leg 3

Leg 4

Palp

Femur

3.33

3.42

3.00

4.00

2.1 1

Patella

2.50

2.25

1.83

2.67

1.83

Tibia

2.33

2.25

2.33

3.17

1.83

Metatarsus

1.75

1.92

2.08

2.75

Tarsus

1.25

1.42

1.25

1.42

1.75

Total

1 1.16

11.26

10.46

14.01

7.91

each of 4-5 on margins; few on cephalothorax. Eyes: not on tubercle; 1, 2; 2,
0.47; 3, 53:54:29; front row slightly procurved, back row straight; 4, 10:12:9:9; 5,
27:28:20; 6, 7, 10, 11, 33, 2, 26. Chelicerae: bristles sparse on prodorsum;
rastellum not evident in ZMH but three distinct thorns in MT 122.898; 7 spaced
teeth on promargin, 4-5 small teeth basomesally. Presumably the rastellar spines
were damaged in the lectotype.

Labium: projects downward between maxillae; 0.36 long, 0.72 wide; 7 thick
pointed cuspules. Maxillae: 1.00 long in front, 1.32 long behind, 0.64 wide; 7-8
cuspules in straight diagonal line. Sternum: 2.08 long, 1.80 wide; separated from
labium by narrow shallow groove; all sigilla oval, about 0.08 long, and touching
margin.

Legs: (Table 5). 4123. Scopulae: thin but entire on metatarsi and tarsi I, II;
entirely absent on metatarsi and tarsi III, IV. Spines [from MT 122.898]: Leg 1:
fe, d5; pa, vl; ti, v5. Leg 2: fe, pi d5; pa, vl; ti, v6; me, vl. Leg 3: fe, d5; ti,
v4; v3. Leg 4: fe, d5; ti, v3; me, v5. Palp: fe, p2 d3; pa, v4; ti, v7. Claws: 2 teeth
on legs I, II; bare on leg IV; 4 long teeth on palpal claw. Trichobothria: 1, 7;
2, 5; 3, 13 c. and f.

Spinnerets: too inverted to measure. Spermathecae: two broad lobes narrowing
and apically divided into two broad mounds (Benoit 1978, fig. Id).

Distribution. — S. seychellanum is known only from the Seychelles.

Material examined. — The types and the following: Seychelles: female, Mahe, La Misere, 438m,
mixed wet forest (R L. G. Benoit and J. J. Van Mol, Mt 122.898); female, juvenile. Silhouette (Percy
Sladen Trust expedition, BMNH 1910.5.1.2-3).

ACKNOWLEDGMENTS

I am grateful to Dr. V. E. Davies (QM, Queensland Museum, Brisbane) for
making comments upon this manuscript. Lor locating and/or loaning material of
Sason, Cosmopelma, and Paracenobiopelma, I am grateful to Drs. M. Grasshoff
(SML, Senckenberg Lorschungsinstitut, Lrankfurt-am-Main), S. Mahunka
(Natural History Museum of Hungary, Budapest), W. Mathis, and J. Coddington
(USNM, U. S. National Museum, Washington, D. C.), G. Rack (ZMH,
Zoologisches Museum, Hamburg), W. Pulawski (CAS, California Academy of
Sciences, San Lrancisco), R. Jocque (MT, Musee Royal de L’Afrique Centrale,
Tervuren), J. Tenorio (Bernice P. Bishop Museum, Honolulu), I. Lansbury
(HMO, Hope Entomological Collections, Oxford), Ms. A. M. da Costa (MNRJ,
Museu Nacional, Rio de Janeiro), and also for gracious aid provided during
periods at their respective institutions, I am grateful to Mr. P. D. Hillyard
(BMNH, British Museum of Natural History, London) and Dr. M. Hubert
(MNHNP, Museum National d’Histoire Naturelle de Paris). Research involving

RAVEN— REVISION OF SASON

69

this manuscript was carried out while I was in receipt of a C.S.I.R.O. Post-
doctoral Fellowship with the superb facilities of the Department of Entomology,
American Museum of Natural History, New York, and was completed at the
Division of Entomology, Canberra, C.S.I.R.O. Australia. For all of that I am
very grateful.

LITERATURE CITED

Benoit, R L. G. 1964. Etudes sur les Barychelidae du centre africain (Araneae — Orthognatha). I. La
separation des sous-families. Rev. Zool. Bot. Afrique, 70:412-416.

Benoit, P. L. G. 1966. Les Barychelidae-Barychelinae africains et malgaches (Aran., Orthogn.). Rev.
Zool. Bot. Afrique, 74:209-241.

Benoit, R L. G. 1978. Contributions a I’etude de la faune terrestre des iles granitiques de I’archipel
des Sechelles (Mission P. L. G. Benoit — J. J. Van Mol 1972) (Araneae, Orthognatha). Rev. Zool.
Africain, 92:405-420.

Bristowe, W. S. 1938. The classification of spiders. Proc. Zool. Soc. London, ser. B, 108:285-321.
Blandin, R and M. L. Celerier. 1977. Observations sur les mygales terricoles recoltees a la Station
d’Ecologie Tropicale de Lamto (Cote d’Ivoire). Rev. Arachnol., l(2):75-83.

Bonnet, P. 1954. Difficulties de nomenclature chez les Araheides. VIII. Le nom de genre Sason. Bull.
Soc. Hist. Nat. Toulouse, 88:295-6.

Cambridge, O. R- 1883. On some new genera and species of spiders. Proc. Zool. Soc. London,
1883:352-365.

Coleman, N. C. 1981. A genus of spiders new to Australia. J. North Queensland Nat., 45(178): 14.

Coyle, F. A. 1983. Aerial dispersal by mygalomorph spiderlings (Araneae, Mygalomorphae). J.
Arachnol., 11:283-286.

Coyle, F. A. 1985. Ballooning behavior of Ummidia spiderlings (Araneae, Ctenizidae). Ibid., 13:137-
138.

Feio, J. L. de A. 1952. Uma curiosa migalomorfa aboricola — " Paracenohiopelma gerecormophila" g.

n., sp. n. (Araneae, Barychelidae). Bol. Mus. Nac. (n.s.) Rio de Janeiro, 1 13:1-19.

Fischel. W. 1927. fiber Siidamerikanische Aviculariiden. Zool. Anz., 74:59-74.

Hirst, A. S. 1911. No. XVIII. The Araneae, Opiliones and Pseudoscorpiones. In, “Percy Sladen Trust
Expedition to the Indian Ocean in 1905 under the leadership of Mr. J. Stanley Gardiner.” Trans.
Linnean Soc. London, Zool., ser. 2, 14:379-395.

Karsch, F. 1891. Arachniden von Ceylon und von Minikoy gesammelt von den Herren Doctoren P.
und F. Sarasin. Berlin Entomol. Zeits., 36:267-310.

Kikkawa, J., Monteith, G. B., and G. Ingram. 1981. Cape York Peninsula: major region of faunal
interchange. Pp. 1695-1743, In Keast A. (ed.). Ecological Biogeography of Australia. Dr. W. Junk,
The Hague.

Kishida, K. 1930. [Cited in Bristowe, 1938]. Lanzania, Tokyo, ii, 1930.

Kulczyhski, W. 1908. Araneae musei nationalis Hungarici in regionibus Indica et Australica a
Ludovico Biro collectae. Ann. Hist. Nat. Mus. Natl. Hungary, 6:428-494.

Legendre, R. 1979. Les Araignees et la derive des continents. Ann. Biol., 18:37-68.

Main, B. Y. 1981a. Eco-evolutionary radiation of mygalomorph spiders in Australia. Pp. 853-872, In
Keast A. (ed.). Ecological Biogeography of Australia. Dr. W. Junk, The Hague.

Main, B. Y. 1981b. Some zoogeographic considerations of families of spiders occurring in New
Guinea. Pp. 583-602, In Gressit, J. L. (ed.). Biogeography and Ecology in New Guinea. Dr. W.
Junk, The Hague.

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

Platnick, N. 1. 1981. Spider biogeography: past, present and future. Rev. Arachnol., 3:85-96.

Pocock, R. 1. 1892. Description of a new Trap-door spider from Ceylon. Ann. Mag. Nat. Hist., ser.
6, 9:49-51.

Pocock, R. 1. 1900. The fauna of British India, including Ceylon and Burma. Arachnida. Taylor and
Francis, London. 1900:i-xii, 1-279.

Pocock, R. 1. 1903. The geographical distribution of spiders of the order Mygalomorphae. Proc. Zool.
Soc. London, 1:340-368.

Pocock, R. 1. 1904. Arachnida, Pp. 797-805, In J. Gardiner (ed.). The fauna and geography of the
Maidive and Laccadive Archipelagoes, vol. 2, part 3. [Cambridge] University Press, Cambridge.

70

THE JOURNAL OF ARACHNOLOGY

Raven, R. J. 1980. The evolution and biogeography of the mygalomorph spider family Hexathelidae
(Araneae, Chelicerata). J. Arachnol., 8:251-266.

Raven, R. J. 1984. A new diplurid genus from eastern Australia and a related Aname species
(Diplurinae: Dipluridae: Araneae). Australian J. Zool., Suppl. ser. no. 96:1-51.

Raven, R. J. 1985. The spider Infraorder Mygalomorphae (Araneae): cladistics and systematics. Bull.
Amer. Mus. Nat. Hist., 182:1-180.

Roewer, C. F. 1942. Katalog der Araneae. Band 1. Friedlander and Sohn, Bremen, pp. 1-1040.

Roewer, C. F. 1963. Araneida: Orthognatha, Labiodognatha. Insects Micronesia, 3(4): 1-132.

Simon, E. 1887. [Observations sur divers arachnides: synonymies et descriptions.] Ann. Soc. Ent.
France, 6: 195.

Simon, E. 1888. Etudes sur les Arachnides de I’Asie meridionale faisant partie des collections de
rindian Museum (Calcutta). II. Arachnides recueillis aux iles Andaman par M. R. D. Oldham.
J. Asiatic Soc. Bengal, 57:282-287.

Simon, E. 1892. Histoire Naturelle des Araignees. Roret, Paris, vol. 1, pt. 1:1-256.

Simon, E. 1898. Etudes arachnologiques. 29^ memoire. XLVI. Arachnides recueillis en 1895 par M.
le Dr A. Brauer (de I’Universite de Marburg) aux iles Sechelles. Ann. Soc. Ent. France, 66:370-
388.

Simon, E. 1903. Histoire Naturelle des Araignees. Roret, Paris, vol. 2, pt. 4:669-1080.

Williams, W. D. 1981. The Crustacea of Australian inland waters. Pp. 1101-1138, In Keast A. (ed.).
Ecological Biogeography of Australia. Dr. W. Junk, The Hague.

Manuscript received July 1985, revised October 1985.

Tiedemann, K. B., D. F. Ventura and C. Ades. 1986. Spectral sensitivities of the eyes of the orb web
spider Argiope argentata (Fabricius). J. Arachnol., 14:71-78.

SPECTRAL SENSITIVITIES OF THE EYES OF THE ORB WEB
ARGIOPE ARGENTATA (FABRICIUS)

Klaus B. Tiedemann, Dora Fix Ventura and Cesar Ades

Department of Experimental Psychology
Institute of Psychology
University of Sao Paulo
Sao Paulo, Brazil

ABSTRACT

Spectral sensitivity curves to light between 425 and 650 nm for the four eyes of the spider Argiope
argentata were determined. Maximum sensitivity was observed at 530 nm for the secondary eyes and
at 525 nm for the principal ones. Chromatic adaptation did not affect this maximum, which suggests
that there is only one receptor and photopigment in this spectral region.

INTRODUCTION

Recordings from the anterior median eyes of jumping spiders demonstrated the
existence of different photoreceptor types, thus confirming earlier behavioral data
suggestive of color vision in these animals (Peckham and Peckham 1887, Crane
1949, Kastner 1950, Young and Wanless 1967). According to DeVoe’s (1975)
intracellular recordings, color vision in the principal eyes of the Jumping spider
Phidippus regius (C. L. Koch) is mediated by three photoreceptor types: UV cells
with peak sensitivity at 370 nm; green cells with peak at 532 nm, and UV-green
cells, peaking at 370 and 532 nm. In the secondary eyes of this spider DeVoe
found only green cells. In the jumping spider Menemerus confusus (Bdsenberg
and Strand), Yamashita and Tateda (1976) found, in addition to UV and green
receptors, a blue receptor at 490 nm and a long-wave-length receptor at 580 nm.
Their data were based both on intracellular recordings and on chromatic
adaptation effects on the electroretinogram (ERG).

Other families of spiders did not exhibit the same color vision abilities as found
in jumping spiders. Young and Wanless (1967) conducted a T-maze preference
study in seven spider families, and found that only Salticidae exhibited behaviors
suggestive of color vision. In ERGs from Lycosidae, DeVoe, Small and Zvargulis
(1969) found only one photopigment (green) in the secondary eyes. The principal
eyes contained variable amounts of this pigment in addition to an UV absorbing
one.

Color vision was only recently studied in spiders of the Araneidae family. The
anterior median eyes of Argiope bruenchini (Scopoli) and Argiope amoena (L.
Koch) were found to have three photoreceptors, with maxima at about 360 nm,
480-500 nm and 540 nm (Yamashita and Tateda 1978, 1981). No recordings were
reported for the secondary eyes.

72

THE JOURNAL OF ARACHNOLOGY

The present study, which was started before the publication of Yamashita and
Tateda’s reports (Tiedemann 1975), brings information about the color vision
system of Argiope argentata (Fabricius), an orb web spider whose behavior has
been extensively studied (Robinson 1969, Robinson and Olazarri 1971, Ades
1973). Argiope argentata lives in sunny areas and is diurnal, but it is capable of
hunting and of building its web in the dark. Its characteristic thermoregulatory
posture (Robinson and Robinson 1978) was shown to be controlled by light
rather than by heat (Ades and Kanner 1978). Its choice of the side of the web
on which to stand is also light dependent (Ades and Kanner 1979). It can rely
on visual stimuli when placing the first threads of the web, as do other
orbweavers (Tilquin 1942). There is no doubt that visual cues are of great
relevance to several aspects of behavior of this species. There is not, however, any
information regarding its color vision abilities. In this report we describe the
ERG measured spectral sensitivities of the dark-adapted eyes of Argiope
argentata, and the results obtained under chromatic adaptation, showing that it
has only one receptor type in the spectral range from 425 to 625 nm.

METHOD

Preparation. — The animal was lightly anesthetized with CO2 and fixed to a
black cardboard with adhesive tape over the legs. The cephalothorax was
immobilized with acrylic (Simplex dental acrylic). The recording electrode was a
cotton wick held in a glass tube pulled out at one end to a capillary tip and filled
with insect physiological solution. Contact with the recording system was done
through a nonpolarizable Ag-AgCl wire. The indifferent electrode was a
Microtode 415 (Transidyne General Corporation) inserted in a leg of the first or
second pair just below the surface. Such a preparation could last for at least 12
hours and longer than a week, if the spider received water by means of a capillary
tube. The temperature was maintained at 22 to 24° C during the recordings. ERGs
were measured with a Tektronix 122 AC amplifier with a 1 s time constant and
50 Hz filter, and a Tektronix 5103/ D 13 dual beam storage oscilloscope and
photographed with a Grass C-4 camera.

The preparation was placed in a light tight box inside a recording chamber.
The optical equipment consisted of a double Maxwellian view system with a
tungsten halide quartz lamp (Osram 58.8105 lOOW) as the light source, which
permitted focussing through a light pipe (Fiber Lite) of a 3 mm diameter light
spot on the eyes. Intensity was controlled by means of glass-mounted Kodak
Wratten 96 neutral density filters and duration of the stimulus flash through a
photographic shutter (Wollensack). Monochromatic light was filtered out through
an interference color wedge (Veril S-200-Leitz), corrected with neutral density
filters Kodak-Wratten 96 for equal light energy for wavelengths between 425 nm
and 675 nm in steps of 25 nm. The light passed through a 3 mm slit yielding
a halfband width of ca. 8.75 nm. A second light source, for the adapting light
was a Fiber Lite Illuminator Mod. 172 with a simple Maxwellian system,
perpendicular to the first. In the experiments with ultraviolet light, an
incandescent UV-lamp, which emitted light between 320 nm and 440 nm with a
maximum output at 380 nm (GE Purple X - 250W) was used in connection with
an UV filter with a maximum transmittance at 380 nm and an edge at 400 nm
(Kodak 18 A), directly on the preparation.

TIEDEMANN £7 AL.— SPECTRAL SENSITIVITIES

73

The stimulating light produced at the end of the light pipe a light intensity of
(4.24) . 10~^W at 500 nm, so that a 20 ms flash corresponded to about (2.76) .
10>® photons. The adapting light yielded, with a red filter with transmittance from
570 nm to infrared (Kodak 23 A), an intensity of (3.60) . lO'^W and with a
narrow band blue filter with maximum output at 430 nm (Kodak 48 B), (1.60)
. 10~^W. Calibrations of ND filters at each of the wavelengths used were made
with a Tektronix radiometer J 16 with J 6502 probe and a Zeiss DMR-21
spectrophotometer.

Procedure. — All recordings were done during the day, between 8 a.m. and 5
p.m. The preparation was dark adapted for at least 30 minutes prior to each
experimental session.

For each eye (anterior median, posterior median, anterior lateral and posterior
lateral eye) the energy-response functions were determined under conditions of
dark adaptation and red-, blue- and white-light adaptations for each of the
calibrated monochromatic lights between 425 and 675 nm. Data collected under
light-adapted conditions were preceded by 10 min of exposure to the adapting
light before measurements were begun. Energy-response functions were obtained
at each of these adapting conditions by measurements of ERG amplitude for
stimuli of the following mean attenuations: 0; 0.21; 0.53; 0.95; 1.26; 1.60; and 2.13
log. At each attenuation value the spectrum was scanned twice, at 25 nm steps,
in opposite directions, with 20 ms long flashes presented at 30 s intervals.

A standard of 550 nm was used throughout the experiment for frequent checks
of the stability of the preparation. All data reported are from preparations which
remained quite stable, within 2% variation.

RESULTS AND DISCUSSION

Waveform of the ERGs. — The waveform of the ERGs recorded in A. argentata
is similar for all four eyes and does not differ as a function of wavelength of the
test flash, as can be observed in Figure 1. Changes in waveform were only noticed
in connection with variation in light intensity. Recordings shown in Figure 2
exemplify such waveform changes. At high intensities the waveform was typically
a single rapid negative wave followed by a slow uniform decay. The small positive
wave is an artifact of capacity coupling of the AC recording amplifier. At low
intensities the ERG consisted of a double negative wave in all eyes except the
principal. In the latter, peak latency was intermediate between the two negative
waves found for the other eyes.

The shape of the ERG was also examined under chromatic light adaptation,
because it has been reported to change in species that have different
photoreceptor types, such as the jumping spider, M. confusus (Yamashita and
Tateda 1976). In the eyes of Argiope argentata no change in waveform was
observed as a function of wavelength of chromatic adaptation. This was a first
indication of the absence of additional photopigments in the tested part of the
spectrum (400-600 nm) in the species.

Spectral Sensitivity. — To determine the spectral sensitivity curves of the four
eyes, the ERG amplitudes in response to monochromatic lights presented to the
dark-adapted eye at 25 nm steps were measured as a function of light intensity
over a 2.0 log unit range. The energy-response functions thereby obtained are
plotted in Figure 3. The functions are closely parallel for each of the eyes, but

74

THE JOURNAL OF ARACHNOLOGY

anterior median eye posterior median eye anterior lateral eye posterior lateral eye

Wavelength

600 nm

425 nm

Fig. 1. — ERGs in response to stimuli of 425 nm and 600 nm obtained in the four eyes. Test flash
duration was about Is.

differ in slope from one eye to the other. The posterior median eyes exhibit the
shallowest curves, whereas the posterior lateral eyes the steepest ones. The
differences could be attributed either to different densities of photoreceptors, or,
as suggested by Dahl (pers. comm.), who found similar differences in the eyes
of Aphonopelma, to the presence of the tapetum in the eyes that exhibited the
lesser slope.

To draw the spectral sensitivity curves, the amplitude data collected above were
transformed according to the procedure described by Autrum and von Zwehl
(1964). This consisted of plotting the mean energy function for each eye and
finding the intensity that would have been necessary to elicit a given ERG
amplitude. The relative sensitivity is the reciprocal of that intensity.

The entire procedure was repeated under chromatic adaptation to blue and red
lights with the purpose of determining whether any of the eyes contained more
than one type of visual cell or pigment in the visible part of the spectrum. If this
were so, the spectral sensitivity curve obtained under selective chromatic
adaptation would be displaced relative to that obtained in the dark adapted state
(Wald 1968, Yamashita and Tateda 1976).

Figure 4 shows the spectral sensitivity curves for the four eyes, under dark
adaptation and the three light adapting conditions. As can be seen, chromatic
adaptation did not change the shape of the spectral sensitivity curves for any of
the four eyes. The smooth lines are Dartnall nomogram curves (Dartnall 1953).
For the lateral and posterior eyes, the best fitting was a 530 nm nomogram curve,
whereas for the anterior median eye a 525 nm nomogram curve yielded the best
fit. This discrepancy may be due not to a different photopigment but to
differences in the optic media between principal and secondary eyes (Young and
Wanless 1967). In the short-wave range of the spectrum the fit of the Dartnall

^ ^ _

2.10 1.60 1.08 .77 0

Relative intensity (log I)

Fig. 2. — ERGs from light adapted posterior lateral eye for 550 nm, light stimuli of five different
intensities. Test flash duration was 20 ms.

TIEDEMANN £7 /IL.— SPECTRAL SENSITIVITIES

75

Fig. 3. — Energy-response functions for dark adapted eyes obtained for five different wavelengths
(data from one preparation).

nomogram curves was not so good. The sensitivity was less than predicted by the
nomogram for all four eyes. The same type of result was found by DeVoe, Small
and Zvargulis (1969). This is not, however, the common finding in other
arthropods, the rule being that sensitivity in the near UV is higher than predicted
by the nomogram curves, as found in the bee (Autrum and von Zwehl 1964),
dragonfly (Autrum and Kolb 1968), Cailiphora and Periplaneta (Walther and
Dodt 1959), Musca domestica (Eckert 1971), Atta sexdens (Martinoya, Bloch,
Ventura, and Puglia 1975) and several species of Crustacea (Scott and Mote 1974).
UV sensitivity is nevertheless present in A. argentata, as revealed by tests made
with UV stimulation (GE Purple X lamp plus Kodak 18 A filter). These tests
showed that responses in the UV are present in all four eyes, but are higher in
the anterior median eyes than in the other ones. This is in agreement with
DeVoe's observations (DeVoe, Small and Zvargulis 1969, DeVoe 1972, 1975), who
also found higher sensitivity in the anterior median eyes.

The reason why we did not find evidence for more than one visual pigment
in the anterior median eyes of Argiope argentata, as found for the Japanese

76

THE JOURNAL OF ARACHNOLOGY

looh

1.75

c

0)

</>

>.50
o
cr

,25

anterior V \
median \

eye

I

1

i

»•
/ •

I

\

i.ooh

J L

posterior ^
median ^
eye

i

o

i

/

/ S

0

1

V/* — L.

* s

\8

J L

s;.5o

0

1

.25

anterior
lateral ^
eye

/ '

ft

. /'8

-i

» t

I

J lii.

posterior ■ /
lateral f'
eye

/

/

8^

A

1

'I

-IiJL,

450 500 550 600 650 450 500 550 600 650

Wavelength nm Wavelength nm

Fig. 4. — Spectral sensitivities of dark (open circle), white (solid circle), blue (square) and red
(triangle) adapted eyes (sensitivity expressed as I/IA, where I is the intensity needed for a standard
response at peak wavelength, and lA, the intensity needed for the same response at wavelength A).
The line drawn in each graph is a Dartnall nomogram for a photopigment with Amax at 530 nm
in secondary eyes, and 525 nm for the principal eyes.

species, is not clear. This failure to isolate a blue receptor (their “green” cell with
peak at 540 nm closely overlaps our data, which peaks at 530 nm) could be
attributed either to species differences, or to a lower sensitivity of our technique.
It could have also been due to the time of day in which the data were collected.
The blue cells found by Yamashita and Tateda were recorded at night. Our data
were always collected during the day. Alternatively, Yamashita and Tateda’s claim
for the existence of a “blue” receptor cannot be regarded as unquestionable.
There are two reasons for this. First, the fact that their claim is based on
recordings from only two cells. Second, the fact that the peak of the “blue” cell
(480-500 nm) falls very close to that of the “green” one. In fact, out of the 24
cells studied by them, only nine had a single peak (three were “UV” cells, two
were “blue” cells, and four, “green” ones). It is possible that the spectral
sensitivities obtained for the cells classified as “blue” represent the result of
electrical coupling between “UV” and “green” ones. A recent article by Horridge,
Marcelja, Jahnke and Matic (1983) shows interactions caused by coupling which
produced nine different types of spectral sensitivity curves in the butterfly retina.
Because coupling appears “to be the rule rather than the exception in visual
systems” (Shaw 1981), the identification of spectral sensitivity curves with single
pigments or single photoreceptors cannot be made without the meeting of a

TIEDEMANN £r/4L.— SPECTRAL SENSITIVITIES

77

number of criteria, as was done by Horridge et al. (1983) for their primary cell
types.

Alternatively, more recent work from our group (Souza, Menzel, and Ventura
1985) shows that the flash method of spectral sensitivity measurement yields
unreliable data, due to both the lengthy recording sessions, with unavoidable
baseline fluctuations, and to the changes in the state of adaptation of the cell with
the use of different spectral filters. A much more reliable method called “voltage
clamp” by Franceschini (1984) or “constant response” (Hertel and Ventura, in
press) avoids both problems and produces reproducible pigment — like spectral
sensitivity curves.

REFERENCES

Ades, C. 1973. A teia e a caga da aranha Argiope argentata. Doctoral Thesis, University of Sao
Paulo.

Ades, C. and E. Kanner. 1978. Orientagao postural na aranha Argiope argentata. Paper presented
at the 30'*’ Annual meeting of the Brazilian Society for the Advancement of Science, Sao Paulo,
Brazil, 1978. Ciencia e Cultura, 30:654 (Abstract).

Ades. C. and E. Kanner. 1979. Luz e travessia da teia pela aranha Argiope argentata. Paper presented
at the 3T‘ Annual meeting of the Brazilian Society for the Advancement of Science, Fortaleza,
Brazil, 1979. Ciencia e Cultura, 31:754 (Abstract).

Autrum, H. and V. von Zwehl. 1964. Die spektrale Empfindlichkeit einzelner Sehzellen der
Bienenauges. Z. vergleichende Physiol., 48:357-384.

Autrum, H. and G. Kolb. 1968. Spektrale Empfindlichkeit einzelner Sehzellen der Aeschniden. Z.
vergleichende Physiol., 60:450-477.

Crane, J. 1949. Comparative biology of salticid spiders of Rancho Grande, Venezuela. Part IV. An
analysis of display. Zoologica, 34:159.

Dartnall, H. J. A. 1953. The interpretation of spectral sensitivity curves. British Med. Bull., 9:24-30.

DeVoe, R. D., R. J. W. Small and J. E. Zvargulis. 1969. Spectral sensitivities of wolf spider eyes.
J. Gen. Physiol., 54:1-32.

DeVoe, R. D. 1972. Dual sensitivities of cells in wolf spider eyes at ultraviolet and visible wavelengths
of light. J. Gen. Physiol., 59:247-260.

DeVoe, R. D. 1975. Ultraviolet and green receptors in principal eyes of jumping spiders. J. Gen.
Physiol., 66:193-207.

Eckert, H. 1971. Die spektrale Empfindlichkeit der Komplexaugen von Musca (Bestimmungen aus
Messungen der optomotorischen Reaktion). Kybernetik, 9:145-156.

Franceschini, N. 1984. Chromatic organization and sexual dimorphism of the fly retinal mosaic. In
Photoreceptors (A. Borsellino and L. Cervetto, eds.), Plenum, New York.

Hertel, H. and D. F. Ventura, (in press). Spectral sensitivity of photoreceptors in the compound eye
of stingless tropical bees. J. Insect Physiol.

Horridge, G. A., L. Marcelja, R. Jahnke and T. Matic. 1983. Single electrode studies on the retina
of the butterfly Papilio. J. Comp. Physiol., 150:271-294.

Kastner, A., 1950. Reaktion der Hupfspinnen (Salticidae) auf unbewegte farblose und farbige
Gesichtsreize. Zool. Beitr., 1:13-50.

Martinoya, C., S. Bloch, D. F. Ventura and N. M. Puglia. 1975. Spectral efficiency as measured by
ERG in the ant (Atta sexdens rubropilosa). J. Comp. Physiol., 104:205-210.

Peckham, G. W. and E. G. Peckham. 1887. Some observations on the mental powers of spiders. J.
Morphol., 1:383.

Robinson, M. H. 1969. Predatory behavior of Argiope argentata (Fabricius). Amer. Zool., 9:161-173.

Robinson, M. H. and J. Olazarri. 1971. Units of behavior and complex sequences in the predatory
behavior of Argiope argentata (Fabricius). Smithsonian Contrib. Zoology, 65:1-36.

Robinson, M. H. and B. C. Robinson. 1978. Thermoregulation in orb-web spiders: new descriptions
of thermoregulatory postures and experiments on the effects of posture and coloration. Zool. J.
Linnean Soc., 64:87-102.

Scott, S. and M. I. Mote. 1974. Spectral sensitivity in some marine Crustacea. Vision Research,
14:659-663.

78

THE JOURNAL OF ARACHNOLOGY

Shaw, S. R. 1981. Anatomy and physiology of identified non-spiking cells in the photoreceptor-lamina
complex of the compound eye of insects, especially Diptera. In Neurones without impulses (A.
Roberts and B. M. H. Bush, eds.), Cambridge University Press, Cambridge.

Souza, J. M. de, R. Menzel and D. F. Ventura. 1985. Improved spectral sensitivity measurements in
bee photoreceptors. Brazilian J. Med. Biol. Res., 18:5-6.

Tiedemann, K. B. 1975. Sensibilidade espectral em Argiope argentata. Master dissertation, Univ. Sao
Paulo.

Tilquin, A. 1942. La toile geometrique des araignees. Presses Universitaires de France, Paris.

Wald, G. 1968. Single and multiple visual systems in arthropods. J. Gen. Physiol., 51:125-155.

Walther, J. B. and E. Dodt. 1959. Die Spektralsensitivitat von Insektenkomplexaugen im Ultraviolet
bis 290nm. Z. Naturfosch., 14b:273-278.

Yamashita, S. and H. Tateda. 1976. Spectral sensitivities of jumping spider eyes. J. Comp.
Physiol., 105:29-41.

Yamashita, S. and H. Tateda. 1978. Spectral sensitivities of the anterior median eyes of the web
spiders Argiope bruenchini and A. amoena. J. Exper. Biol., 74:47-57.

Yamashita, S. and H. Tateda. 1981. Efferent neural control in the eyes of orb weaving spiders. J.
Comp. Physiol., 143:477-483.

Young, M. R. and F. Wanless. 1967. Observations on the fluorescence and function of spider’s eyes.
J. Zool., 151:1-16.

Manuscript received January 1984, revised October 1985.

Herrero, M. V. and C. E. Valerio. 1986. Analisis de la actividad diaria de Aphonopelma seemanni
(Araneae, Theraphosidae) en Costa Rica. J. ArachnoL, 14:79-82.

ANALISIS DE LA ACTIVIDAD DIARIA DE
APHONOPELMA SEEMANNI
(ARANEAE, THERAPHOSIDAE) EN COSTA RICAi

Marco V. Herrero

Centro Universitario del Atlantico, Universidad de Costa Rica
Facultad de Ciencias Exactas y Naturales, Universidad NacionaF

and

Carlos E, Valerio

Escuela de Biologia, Universidad de Costa Rica

ABSTRACT

Variation in the daily activity patterns of Aphonopelma seemanni females was studied during the
rainy season in Guanacaste, Costa Rica. Hourly variations were significant (p < 0.05), with a constant
pattern throughout the season. The spiders are nocturnal predators that reach the activity peak
between 20.30 and 0.30 hours.

RESUMEN

La variacion en el patron de actividad diaria de las hembras de Aphonopelma seemanni fue
estudiada durante la epoca Iluviosa en Guanacaste, Costa Rica, encontrandose que los cambios
durante el dia fueron significativos (p < 0.05) con un patron constante durante toda la estacion. Las
aranas son eminentemente nocturnas y alcanzan su actividad maxima durante las 20:30 y las 0.30
horas.

INTRODUCTION

La ecologia de las aranas terafosidas en los tropicos ha sido poco estudiada,
siendo las especies costarricenses las mejor conocidas (Valerio 1979, 1980a, 1980b,
1982). En el caso de Aphonopelma seemanni (F. P. Cambridge) existe
informacion sobre estructura de tuneles (Herrero y Bolanos 1982) y sobre habitos
de vida, informandose que las hembras son basicamente nocturnas (Herrero et
al. 1983). Se destaca tambien la informacion de Minch (1978) sobre la actividad
diaria de Aphonopelma chalcodes del sur de Estados Unidos. En el presente

'Este trabajo fue financiado por la Vicerrectoria de Investigacion de la Universidad de Costa Rica.
2Direcci6n actual; Department of Entomology, Oklahoma State University, Stillwater, Oklahoma
74078.

80

THE JOURNAL OF ARACHNOLOGY

Cuadro 1. — Prueba de la menor diferencia significativa, usada para comparar promedios de
actividad horaria (diurna y nocturna) en Aphonopelma seemanni. Promedio subrayado con la misma
linea no son significativamente diferentes (P > 0.05).

Actividad
A. Diurna:

Horas (rango) 16.30 6.30

Promedios 0.38 0.36

8.30 14.30 10.30 12.30

0.32 0.32 0.19 0.16

B. Nocturna:

Horas (rango) 22.30 20.30 0.30 4.30 2.30 18.30

Promedios 0.97 0.96 0.93 0.88 0.86 0.76

trabajo se hace un analisis de la actividad diaria (diurna y nocturna) de esta
especie durante la epoca Iluviosa.

MATERIALES Y METHODOS

El estudio se realize en Pozo Azul de Abangares, Provincia de Guanacaste,
Costa Rica, en una colina rocosa de 7400 de area, poblada por gramineas
y arboles aislados. La zona corresponde a una formacion tropical humeda (Tosi
1969), con una precipitacion mensual tipica de 235 mm (maxima de 417 en
setiembre y minima de 13 en diciembre) y una temperatura ambiente (promedio
mensual) entre 24° y 27° C durante la epoca Iluviosa. Se efectuaron cuatro visitas
(junio, agosto, octubre y diciembre de 1981) y en cada visita fueron observados
42 tuneles (cuya localizacion habia sido previamente marcada con estacas de
madera) cada dos horas hurante un periodo de 24 horas.

Una arana se considero activa cuando se encontro en el horde del orificio de
entrada al tunel.

RESULTADOS

El anMisis de varianza indica que las diferencias entre los datos de las cuatro
visitas no son significativas (F = 0.65; p > 0.5906); siendo la diferencia horaria
altamente significativa (F = 26.94; p > 0.0001). Para comparar los promedios una
prueba de la “menor diferencia significativa” fue llevada a cabo (Cuadro 1).

En el cuadro 2 se presentan los resultados de la actividad diurna en terminos
de proporcidn de arahas activas por hora de observacion notandose muy poca
actividad ya que esta especie es predominantemente nocturna (Herrero et al.
1983). La actividad decrecid a partir de las 6.30 cesando casi totalmente al medio
dia, para aumentar durante las horas de la tarde (Fig. 1).

Como aspecto de interes bioldgico se observo que algunos tuneles estaban
cubiertos de seda entre las 6.30 y las 8.30. Entre las 16.30 y las 18.30 se
observaron en el area de estudio avispas pompilidas (posiblemente Pepsis) que
son depredadoras de estas aranas.

La mayor actividad se observo durante la noche, iniciandose a las 18.30 y
alcanzando su maximo a las 22.30 (Cuadro 2) para decrecer qradualmente en la
madrugada (Fig. 1).

HERRERO AND VALERIO— ACTIVIDAD DE APHONOPELMA SEEMANNI

Fig. 1. — Actividad diaria de hembras de Aphonopelma seemanni (se indica la desviacion estandar
para cada bora).

Una curva de regresion negativa significativamente diferente de cero (1.10 -
0.01 (X), r2 = 0.78, p > 0.048), elaborada con las observaciones entre las 20.30
y las 4.30 enfatizo el hecho de que la actividad decrecio en boras de la
madrugada.

CONCLUSIONES Y DISCUSION

Esta especie presenta un ciclo dial de actividad con un minimo a medio dia
y un maximo cerca de la media noche, que parece mantenerse sin variacion a
traves de toda la epoca Iluviosa.

Algunas especies de aranas “pica-caballo” (como se conoce en Costa Rica a las
terafosidas) son depredadores encontrados con frecuencia en terrenos de
vegetacion herbacea, como los pastisales cultivados para la cria de ganado. Su
impacto sobre las poblaciones de insectos herbivoros no ha sido evaluada y un
analisis sobre la actividad de las aranas puede aportar valiosa informacion para
iniciar tales estudios. En particular Aphonopelma seemanni cuya distribucion en
Costa Rica corresponde al Pacifico Seco (Valerio 1980b), es comunmente
encontrada en la sabana guanacasteca, sitio importante en la produccion de
ganado de came. For anos ha sido creencia popular que esta arana puede causar
lesiones locales en el ganado conocidas como “orinadas de arana”, en cuyo caso
la asociacion de estas aranas con el ganado seria inconveniente. Pero Herrero
(1980) demostro que el unico efecto asociado con sus secreciones externas es una
leve destruccion del musculo esqueletico en ratones blancos.

En lo referente a la actividad de los machos de esta especie, que es mas dificil
de estudiar por cuanto estos no viven en tuneles facilmente localizables, Valerio

82

THE JOURNAL OF ARACHNOLOGY

Cuadro 2. — Actividad diaria de Aphonopelma seemanni durante la epoca illuviosa en Costa Rica.
Resultados expresados en proporcion de individuos activos.

HORAS DE

OBSERVACION.

JUNIO

AGOSTO

OCTUBRE

DICIEMBRE

PROMEDIO

6:30

0.52

0.62

0.16

0.32

0.36

8:30

0.00

0.62

0.25

0.39

0.32

10:30

0.00

0.29

0.16

0.32

0.19

12:30

0.58

0.29

0.09

0.25

0.16

14:30

0.45

0.29

0.22

0.32

0.32

16:30

0.58

0.29

0.22

0.32

0.38

18:30

0.71

0.76

0.69

0.89

0.76

20:30

1.00

0.86

1.00

1.00

0.96

22:30

1.00

1.00

1.00

0.89

0.97

0:30

0.90

0.95

0.94

0.82

0.93

2:30

0.90

0.76

0.94

0.82

0.86

4:40

0.84

0.81

0.94

0.93

0.88

(1980) informo sobre actividad reproductiva en la epoca Iluviosa durante julio y
agosto. Aunque ban sido tambien colectados deambulando durante la epoca seca
(Herrero y Bolanos 1982).

AGRADECIMIENTOS

Agradecemos a los estudiantes Alvaro Morales y Ramon Vangas su asistencia
en el trabajo de campo y al Dr. Roger Bolanos sus valiosas sugerencias. Tambien
agradecemos al Dr. David Weeks, Departamento de Estadistica, Oklahoma State
University, su asesoramiento durante el analisis de este trabajo.

Ejemplares de referencias depositadas en el Museo de Zoologia, Universidad
de Costa Rica (MZUCR).

LITERATURA CITADA

Herrero, M. V. 1980. Caracteristicas generales y efectos locales de las secreciones externas de dos
especies de arana pica-caballo, Aphonopelma seemanni y Sphaerobothria hoffmanni (Araneae:
Theraphosidae) Tesis M. Sc., Universidad de Costa Rica.

Herrero, M. V. y R. Bolanos. 1982. Habitos de vida y tuneles de dos aranas pica-caballo de Costa
Rica (Araneae: Theraphosidae). Observaciones preliminares. Brenesia, 19/20:19-22.

Herrero, M. V. A. Morales Y R. Vargas. 1983. Patrones de actividad diaria de las hembras de
Aphonopelma seemanni (Araneae: Theraphosidae) durante la epoca Iluviosa en Guanacaste, Costa
Rica. Rev. Biol. Trop., 31(1):I61-162.

Minch, E. W. 1978. Daily activity patterns in the tarantula Aphonopelma chalcodes Chamberlin.
Novit. Arthrop., 1:1-34.

Tosi, J. 1969. Mapa Ecologico (Costa Rica). Centro Cientifico Tropical, San Jose, Costa Rica.

Valerio, C. E. 1979. Aranas terafosidas de Costa Rica (Araneae: Theraphosidae). IE Psalmopoeus
reduncus, redescripcion, distribucion y el problema de dispersion en terafosidas. Rev. Biol. Trop.,
27:301-308.

Valerio, C. E. 1980b. Aranas terafosidas de Costa Rica (Araneae: Theraphosidae). III.
Sphaerobothria, Aphonopelma, Pterinopelma, Citharacanthus y Stychoplastus. Rev. Biol. Trop.,
28:271-296.

Valerio, C. E. 1982. Aranas teraf6sidas de Costa Rica (Araneae: Theraphosidae). IV. Generos
Metriopelma y Cyclosternum, incluyendo especies de Panama. Brenesia, 19/20:407-423.

Manuscript received May 1985, revised December 1985.

Muchmore, W. B. 1986. Redefinition of the genus Olpiolum and description of a new genus
Banksolpium (Pseudoscorpionida, Olpiidae). J. Arachnol., 14:83-92.

REDEFINITION OF THE GENUS OLPIOLUM AND
DESCRIPTION OF A NEW GENUS BANKSOLPIUM
(PSEUDOSCORPIONIDA, OLPIIDAE)

William B. Muchmore

Department of Biology
University of Rochester
Rochester, New York 14627

ABSTRACT

The genus Olpiolum Beier is redefined on the basis of a redescription of the type species, Olpiolum
medium Beier. O. elegam (Balzan) is retained in the genus, but all other species previously assigned
there are removed. Olpiolum modestum Banks, redescribed, serves as the type species of the new
genus Banksolpium, and a new species, B. magnum, is added to the genus.

INTRODUCTION

In 1931 the genus Olpiolum was established by Beier, with the new species
Olpiolum medium as the type species. Since then, a dozen other species have been
placed in the genus. However, many of these have differed in significant ways
from the original diagnosis and the genus has been (both implicitly and explicitly)
emended several times (Beier 1954, 1959, Hoff 1964). As a result of these and
other considerations, the distinction between Olpiolum and Pachyolpium Beier
(1931) has been difficult to maintain (Hoff 1964, Mahnert and Schuster 1981,
Heurtault and Rebiere 1983).

In an attempt to fix the definitions of and distinguish clearly between the
genera Olpiolum and Pachyolpium, I have examined and redescribed the types
of the type species of the two. The present paper deals with Olpiolum;
Pachyolpium will be treated in another place.

Species which have been assigned to Olpiolum are as follows:

O. medium Beier 1931; Paraguay

O. elegans (Balzan 1890), Beier 1932; Paraguay

O. modestum (Banks 1909)?, Beier 1932; Brazil

O. peruanum Beier 1959; Peru

O. crassum Beier 1959; Peru

O. aureum (Hoff 1945), Hoff 1964; Puerto Rico

O. puertoricensis (Hoff 1945), Hoff 1965; Puerto Rico

O. confundens (Hoff 1945), Hoff 1964; Puerto Rico

O. monae Hoff 1964; Mona Is., Puerto Rico

O. amplum Hoff 1964; Jamaica

O. paucisetosum Muchmore 1977; Yucatan, Mexico

84

THE JOURNAL OF ARACHNOLOGY

O. fuscipalpum Muchmore 1977; Belize
O. machadoi Heurtault 1980; Venezuela

Of these, O. medium and O. elegans actually belong in the genus as redefined
below, O. modestum forms the basis for the establishment of a new genus, and
the others belong elsewhere.

Genus Olpiolum Beier

Olpiolum Beier, 1931:312, 1932:196; Hoff 1964:19.

Type species. — Olpiolum medium Beier, 1931.

Description. — Beier’s original description of the genus is as follows (1931:312):

Cephalothorax langer als breit, glatt, mit 4 Augen. Tergite ungeteilt, mit je 4-6 kleinen Borsten.
Palpen massig kraftig. Femur dorsal nahe der Basis mit einem Tasthaar. Giftkanal der
Palpenfinger kurz. Apodem des beweglichen Fingers klein. Das Tasthaar est etwas proximal der
Mitte gelegen, it etwas vor est, ist proximal von est, jedoch nicht an der Fingerbasis. Patella
des 1. Beinpaares kiirzer als das Femur, gegen dieses frei beweglich. Arolien einfach.

In Das Tierreich (1932:196), Beier changed this slightly, to read:

Cephalothorax deutlich langer als breit, ohne Quereindriicke, mit vier grossen, nahe beisammen
stehenden Augen. Tergite ungeteilt, die vorderen mit je 4, die mittleren mit je 6 Marginalborsten.
Palpen massig kraftig. Femur mit einem Tasthaar dorsal nahe der Basis. Giftkanal der
Palpenfinger kurz, Apodem des beweglichen Fingers klein. Patella des 1. Beinpaares kiirzer als
das Femur, gegen dieses beweglish. Arolien einfach. Das Tasthaar it etwas vor est, letzteres
etwas proximal der Mitte gelegen; ist zwischen it und der Fingerbasis.

These are essentially correct, but several features require better definition or
amplification.

Namely, a genus with the basic characters of the family Olpiidae, subfamily
Olpiinae, and tribe Olpiini (cf. Hoff 1964). The 4 eyes large, with bulging corneas,
as is common in Olpiinae. Setae on tergites very small and inconspicuous, as are
most other vestitural setae on animal. Palps moderately robust for an olpiin (cf.
Beier 1932:Fig. 228; present paper: Figs. 2 and 3). Tactile seta on dorsum of
palpal femur about '4 length of femur from proximal end. Venom ducts in chelal
fingers quite short; nodus ramosus in fixed finger distad of trichobothrium et,
while that in movable finger less than half distance from finger tip to
trichobothrium t. Trichobothrium est slightly proximad of middle of fixed finger;
it only slight distance distad of est; ist about midway between est and isb, or a
little closer to isb. On movable finger sb a little closer to b than to st. Basifemur
(femur) of leg I about 1.5 times as long as telofemur (patella). Arolia of pedal
tarsi simple, longer than claws. And in addition: Cheliceral flagellum of 3 setae.
All surfaces of palpal segments entirely smooth. Chaetotaxy of metatarsus of leg
IV, in terminology of Heurtault (1983:594), 2 dorsal, 2 ventral, 3 lateral external
(including a long tactile seta proximally), 7 lateral internal (or TT2-2-7-2 in my
abbreviated notation).

In brief, Olpiolum can be distinguished from all other genera of the Olpiini by
the location of both trichobothria est and it near the middle of the fixed chelal
finger, the very short venom ducts in the chelal fingers, the occurrence of only
6 setae on the middle tergites, and a single tactile seta on the palpal femur.

Remarks. — According to our present knowledge, Olpiolum must be considered
a small genus with only two species restricted to Paraguay, though, certainly,
future collections will extend its distribution. So, too, its systematic relationships

MUCHMORE— PSEUDOSCORPION GENERA OLPIOLUM KW BANKSOLPIVM

85

will become better understood when more is learned about the South American
Olpiidae.

While I do not believe that the chaetotaxy of the metatarsus (basitarsus) of leg
IV is important at the generic level (cf. Mahnert 1982:298), I have included the
numbers here for comparison with Heurtault and Rebiere’s figures for Olpiolum
and Pachyolpium (1983:594). A more fruitful discussion of this subject can occur
later, in reference to Pachyolpium, for which very many specimens are available
to demonstrate intra- and interspecific variations.

Olpiolum medium Beier
Figs. 1-5

Olpiolum medium Beier, 1931:312, Fig. 8 [erroneously labelled Apolpium medium]-, 1932:197, Fig.

228.

Material. — Female lectotype (here designated) and 3 paralectotypes (1 male, 2
tritonymphs) from “Paraguay, unter Holz und Laub, 18.7, K. Fiebrig, Nr. 672”;
all labelled syntypes. Nr. 27113, in the Museum fiir Naturkunde an der

Universitat zu Berlin. The lectotype has been cleared and mounted on a

microscope slide and examined; the paralectotype male was examined in

glycerine.

Description. — The description by Beier is sketchy, but generally accurate.
However, I give here a complete description based mainly on the mounted

lectotype female. The paralectotype male is a little larger than the lectotype.
Carapace and palps light brown, other parts lighter. Carapace a little longer than
broad, without transverse furrows; anterior margin slightly indented at center;
with 4 large, corneate eyes; surface mostly smooth, but with few reticulations
laterally and posteriorly; setae short and thin, difficult to count accurately, but
apparently with 6 across anterior margin and 4 at posterior margin. Abdominal
tergites and sternites undivided; tergites with 4-6 small, thin setae; setae of
sternites impossible to count accurately. Cribriform plates not visible.

Chelicera 2/5 as long as carapace; hand with 5 long, acuminate setae; flagellum
of 3 setae (1 long, 2 short), all finely denticulate; subapical lobe of movable finger
prominent and bifurcate (Fig. 1); galea long, slender, with 3 small terminal rami
(Fig. 1); lamina exterior present.

Palp rather heavy, with femur and tibia about equal in length (Fig. 2); all
surfaces smooth. Femur 3.0, tibia 2.5, and chela (without pedicel) 2.9 times as
long as broad; hand (without pedicel) 1.5 times as long as deep; movable finger
1.03 times as long as hand. Femur with a conspicuous tactile seta on dorsum
about length from proximal end. Trichobothria on chela as shown in Fig. 3:
est and it at about same level near middle of fixed finger, eb, esb, ib and isb
in a group at base of finger, ist midway between est and isb; on movable finger
sb closer to b than to st. Venom apparatus well developed in both fingers, venom
ducts short; nodus ramosus in fixed finger distad of trichobothrium et, in
movable finger less than half distance from tip to t. Fixed finger with 30 marginal
teeth, the distal ones cusped, the proximal ones low and flat; movable finger with
26 similar teeth.

Legs moderately robust. Leg I with basifemur about 1.5 times as long as
telofemur (Fig. 4). Leg IV with entire femur 2.7 and tibia 3.5 times as long as

86

THE JOURNAL OF ARACHNOLOGY

Figs. 1-5. — Olpiolum medium Beier, lectotype female: 1, end of movable finger of chelicera; 2,
dorsal view of right palp, 3, lateral view of left chela; 4, lateral view of leg I; 5, lateral view of leg

IV.

deep (Fig. 5). Chaetotaxy of metatarsus IV: T+2-2-7-2, tactile seta (T) very close
to proximal end. Arolia longer than claws, not divided.

Measurements (mm). — Figures given first for lectotype female, followed in
parentheses by those for paralectotype male. Body length 1.96 (1.52). Carapace
length 0.48(0.49). Chelicera 0.20(0.21) by 0.11(0.105). Palpal trochanter
0.245(0.26) by 0.13(0.14); femur 0.40(0.43) by 0.13(0.15); tibia 0.40(0.445) by
0.16(0.185); chela (without pedicel) 0.665(0.76) by 0.23(0.26); hand (without
pedicel) 0.34(0.36) by 0.23(0,26); pedicel 0.06 (0.065) long; movable finger
0.35(0.37) long. Leg I: basifemur 0.19(0.215) by 0.065(0.07); telofemur 0.125(0.14
by 0.065(0.075). Leg IV: entire femur 0.38(0.39) by 0.145(0.16); tibia 0.27 by 0.08;
metatarsus 0.155 by 0.05; telotarsus 0.13 by 0.04.

Remarks. — The measurements given above are slightly different from those
given by Beier (1931:313). This may be due to differences in our methods of
measuring or may be the result of Beier’s mistaken assumption that all 4
individuals were adults.

Olpiolum elegans (Balzan)

Olpiolum elegans Balzan, 1890:437, P1.16, Figs. 19, 19a, I9b,; 1891:549; Eilingsen 1910:390; Beier

1930:207.

Olpiolum e/egam (Balzan), Beier 1931:313; 1932:196, Fig. 227; Feio 1945:4; Hoff 1964:20.

MUCHMORE— PSEUDOSCORPION GENERA OLPIOLUM AND BANKSOLPIVM

87

O. elegans and O. medium seem very similar, with only slight differences in
size and proportions (cf. Beier 1931:313). Indeed, the types of O. medium are
apparently specimens which were identified as Olpiolum elegans by Ellingsen
(1910) and Beier (1930). Study of additional material may well prove the two to
be synonymous.

The type locality of this species is Rio Apa, Paraguay. Feio (1945) has also
recorded collections in Brazil and Argentina, but his identifications may not be
correct.

Olpiolum modestum Banks (1909) was referred tentatively to Olpiolum by
Beier (1932:197). Reexamination of the types of this species reveals that it belongs
in another genus, heretofore unrecognized.

Banksolpium, new genus

Type species. — Olpiolum modestum Banks, 1909.

Etymology. — The genus is named in honor of Nathan Banks, an early, great
pioneer in the study of American pseudoscropions and other arachnids.

Diagnosis. — A member of the family Olpiidae, subfamily Olpiinae, and tribe
Olpiini (cf. Hoff 1964), with relatively heavy palps, venom duct in fixed finger
of palpal chela reaching just to level of trichobothrium et, trichobothrium ist
lying distad of est, one tactile seta on palpal femur proximally, and cheliceral
flagellum of 3 setae.

Description. — Generally robust. Carapace and palps well sclerotized and darker
than other parts. Carapace longer than broad; anterior margin slightly indented
at middle; surface smooth, without transverse furrows; 4 large corneate eyes; 30-
40 vestitural setae, with 4 at anterior and 4 at posterior margins. Abdominal
tergites and sternites entire, smooth; pleural membranes longitudinally striate;
setae short and slender; middle tergites with 6-12 setae; anterior genital
operculum of male with 11-12 setae in 2 transverse rows, that of female with
about 4 setae in a single row; internal genitalia not obviously distinctive.
Chelicera about one-third as long as carapace; hand with 5 long setae; flagellum
of 3 setae (1 large, 2 small); subapical lobe of movable finger distinct; galea with
3 terminal rami, smaller in male than in female. Palps moderately robust for an
olpiin, femur and tibia about equal in length and hand (without pedicel) and
movable finger about equal in length (Figs. 7 and 8). All surfaces smooth. Femur
with conspicuous tactile seta in proximal third of dorsum. Chela with venom
ducts short, the nodus ramosus lying at or just proximad of trichobothrium et
in fixed finger and about midway between t and finger tip in movable finger.
Trichobothria on fixed chelal finger generally as in olpiins, but with ist lying
distad of est; on movable finger, st much closer to b than to st. Both fingers with
contiguous marginal teeth, well developed and cusped distally, but becoming
lower and rounded proximally. Legs rather robust, with entire femur 2. 2-2. 6 times
as long as deep. Leg I with basifemur 1.6- 1.8 times as long as telofemur. Arolia
of pedal tarsi simple and longer than claws.

Remarks. — Banksolpium can easily be distinguished from other genera of the
Olpiini by the placement of the trichobothria on the fixed finger of the chela,
notably the location of ist distad of est. The genus is presently known to include
only 2 species from eastern and southeastern Brazil.

88

THE JOURNAL OF ARACHNOLOGY

Figs. 6-10. — Banksolpium modestum (Banks), lectotype male: 6, genital opercula; 7, dorsal view of
right palp; 8, lateral view of left chela; 9, lateral view of leg I; 10, lateral view of leg IV.

Banksolpium modestum (Banks), new combination
Figs. 6-10

Olpium modestum Banks, 1909:148, no fig.

Olpiolum(?) modestum (Banks), Beier 1932:197.

Olpiolum modestum (Banks), Hoff 1964:20; Heurtault 1980:71; not Beier 1954:134, Fig. 3.

Material. — Lectotype male (WM 4534.01003) (here designated) and 2
paralectotype females from Pernambuco, Brazil, in the Museum of Comparative
Zoology, Harvard University; all specimens have been cleared, dissected, mounted
on slides, and examined.

Description. — The original description by Banks is very short and general and
not accompanied by illustrations. Therefore, a complete description is presented
here. Male and female similar. Moderately robust for an olpiin. Carapace and
palps well sclerotized and darker than other parts. Vestitural setae small and fine.
Carapace longer than broad; smooth and without transverse furrows; anterior
margin nearly straight or slightly indented at middle; 4 large, corneate eyes; about
30 setae, 4 at anterior and 4 at posterior margins. Abdominal tergites and
sternites entire, surfaces smooth; pleural membranes longitudinally striate. Tergal

MUCHMORE— PSEUDOSCORPION GENERA OLPIOLVM AND BANKSOLPIVM

89

chaetotaxy of lectotype 2:4:4:4:6:6:6:6: 10: 10:?:2; paralectotypes similar but with 6
setae on tergite 4. Sternal chaetotaxy of male 1 1:[2-2]:(0)4(0):(0)6(0):6:-, that of
females 4:(0)4(0);(0)6(0):6:-; anterior genital operculum of male with setae in 2
transverse rows (Fig. 6), that of female with setae in I row. Internal genitalia not
distinguishable.

Chelicera one-third as long as carapace; hand with 5 long setae; flagellum of
3 setae (1 long, 2 short), all finely denticulate; subapical lobe of movable finger
distinct and finely incised at tip; galea long, with 3 terminal rami, smaller in male
than in female; lamina exterior present.

Palp moderately heavy, with tibia slightly shorter than femur (Fig. 7); femur
3.3-3.35, tibia 2. 3-2. 4, and chela (without pedicel) 2.75-3.0 times as long as broad;
hand (without pedicel) 1.45-1.64 times as long as deep; movable finger 1.0-1.05
times as long as hand. All surfaces smooth. Femur with conspicuous tactile seta
on dorsum about 'A length from proximal end. Trichobothria on chela as shown
in Fig. 8: on fixed finger it nearer to et than to est, ist distad of est, and eh,
esb, ib and isb in a group at base; on movable finger sh much closer to b than
to St. In one paralectotype there is an extra trichobothrium on the external
surface midway between et and it. Venom apparatus well developed in both
fingers, ducts short; nodus ramosus in fixed finger at level of trichobothrium et,
in movable finger about half distance from tip to t. Fixed finger with 29-33
marginal teeth, the distal ones cusped, the proximal ones very low and flat;
movable finger with 20-22 similar teeth.

Legs robust; leg IV with entire femur 2.15-2.25 and tibia 3. 5-3. 7 times as long
as deep (Fig. 10). Leg I with basifemur about 1.65 times as long as telofemur
(Fig. 9). Chaetotaxy of metatarsus IV of lectotype T+1-2-8-2, tactile seta (T)
close to proximal end; others T+1-2-7-2 and T-M-3-7-2. Arolia longer than claws,
not divided.

Measurements (mm). — Figures given first for the lectotype male, followed in
parentheses by those for the two paralectotype females. Body length 2.0(2.1,2.25).
Carapace length 0.59(0.59,0.62). Chelicera 0.19(0.20,0.19) long. Palpal trochanter
0.26(0.27,0.27) by 0.14(0.14,0.16); femur 0.50 (0.51,0.51) by 0.15(0.155,0.155); tibia
0.42(0.45,0.45) by 0.175(0.19,0.195); chela (without pedicel) 0.755(0.81,0.82) by
0.25(0.28,0.30); hand (without pedicel) 0.39(0.42,0.44) by 0.265(0.25,0.30); pedicel
0.070(0.075) long; movable finger 0.41(0.44,0.43) long. Leg I: basifemur 0.235
(0.245,0.245) by 0.07(0.09,0.08); telofemur 0.14(0.155,0.15) by 0.08(0.08,0.09). Leg
IV: entire femur 0.48(0.50,0.49) by 0.22(0.23,0.22); tibia 0.32(0.33,0.34) by
0.09(0.09,0.095); metatarsus 0.16(0.18,0.18) by 0.06 (0.06,0.06); telotarsus
0. 13(0.1 3,0. 1 3) by 0.04(0.04,0.04).

Remarks. — As pointed out by Heurtault (1980:71-72), the specimen from
Venezuela described by Beier (1954) as Olpiolum modestum (Banks) is actually
referable to Olpiolum machadoi Heurtault. Importantly, the trichobothriotaxy of
the palpal chela and the numbers of setae on the abdominal tergites preclude its

being B. modestum.

Banksolpium magnum^ new species
Figs. 1 1 and 12

Material. — Holotype male (WM 612.03001) from Vigosa, Minas Gerais, Brazil,
6 July 1933, in Cornell University Insect Collection.

90

THE JOURNAL OF ARACHNOLOGY

Figs. II and 12. — Banksolpium magnum, new species, holotype male: 11, dorsal view of right palp;
12, lateral view of left chela.

Diagnosis. — Much like B. modestum but larger (carapace length 0.835 V5. 0.59)
and with more setae on carapace (40 V5. 30) and on abdominal tergites (5:6:9:11:-
v5. 2:4:4:4:-).

Description of male (female unknown). — A large species of the genus
Banksolpium as presently known. Carapace, palps, and posterior tergites well
sclerotized and reddish brown in color, other parts lighter. Vestitural setae
generally thin and short. Carapace a little longer than broad; smooth and without
transverse furrows; anterior margin straight; 4 large, corneate eyes; about 40
setae, 4 at anterior and 4 at posterior margins. Abdominal tergites and sternites
entire, smooth; pleural membranes finely, longitudinally striate. Tergal chaetotaxy
5:6:9:11:10:12:12:12:12:12:6:2. Sternal chaetotaxy 12:[2-2]:(0)6(0):(0)8(0): 12: 13: 14:
13:13:12:4:2; disposition of setae on genital opercula much as in B. modestum.
Internal genitalia not obviously distinctive.

Chelicera 0.4 as long as carapace; hand with 5 long setae; flagellum of 3 setae,
all denticulate; subapical lobe of movable finger distinct, entire; galea slender,
with 3 small rami; lamina exterior present.

Palp moderately heavy, with tibia slightly longer than femur (Fig. 11); femur
3.1, tibia 2.75, and chela (without pedicel) 3.1 times as long as broad; hand
(without pedicel) 1.7 times as long as deep; movable finger 0.92 as long as hand.
All surfaces smooth. Femur with tactile seta about one-fifth length from proximal
end. Trichobothria on chela as shown in Fig. 12: on fixed finger est eloser to

MUCHMORE— PSEUDOSCORPION GENERA OLPIOLVM BANKSOLPIVM

91

isb than to it, and ist distad of est; on movable finger sb much closer to b than
to St. Venom apparatus well developed in both chelal fingers, ducts short; nodus
ramosus in fixed finger just proximad of trichobothrium et, in movable finger
about midway between t and finger tip. Fixed finger with 46 contiguous teeth,
nearly all cusped; movable finger with 39 teeth, cusped in distal half but
becoming rounded and flattened proximally.

Legs more slender than in B. modestum; leg IV with entire femur 2.65 and tibia
4.15 times as long as deep. Leg I with basifemur about 1.85 times as long as
telofemur. Chaetotaxy of metatarsus IV is T-1-2-3-8-3. Arolia longer than claws,
not divided.

Measurements (mm). — Body length 2.57. Carapace length 0.835. Chelicera 0.33
by 0.17. Palpal trochanter 0.415 by 0.235; femur 0.77 by 0.25; tibia 0.82 by 0.30;
chela (without pedicel) 1.33 by 0.43; hand (without pedicel) 0.725 by 0.43; pedicel
0.105 long; movable finger 0.665 long. Leg I: basifemur 0.40 by 0.125; telofemur
0.215 by 0.125. Leg IV: entire femur 0.72 by 0.27; tibia 0.54 by 0.13; metatarsus
0.325 by 0.09; telotarsus 0.265 by 0.065.

Etymology. — The species is named magnum in reference to its being larger than

B. modestum.

Remarks. — The type locality in southeastern Brazil is some 1500 km from the
type locality of B. modestum on the eastern bulge of the country, indicating a
broad distribution of the genus. However, thus far no other congeneric species
are known.

ACKNOWLEDGMENTS

I am much indebted to the following for the loan of material: H. W. Levi of
the Museum of Comparative Zoology, M. Moritz of the Museum fiir Naturkunde
in Berlin, and L. L. Pechuman of the Cornell University Insect Collection.

LITERATURE CITED

Balzan, L. 1890. Revisione dei pseudoscorpioni del bacino dei fiumi Parana e Paraguay nell’America
meridionale. Ann. Mus. Civ. Stor. Nat. Genova, (2)9:401-454.

Balzan, L. 1891. Voyage de M. E. Simon au Venezuela (Decembre 1887-Avril 1888). 16e memoire.

Arachnides. Chernetes (Pseudoscorpiones). Ann. Soc. Entomol. France, 60:497-552.

Banks, N. 1909. New tropical pseudoscorpions. J. New York Entomol. Soc., 17:145-148.

Beier, M. 1930. Die Pseudoskorpione des Wiener Naturhistorischen Museums. 111. Ann. naturhistor.
Mus. Wien, 44:199-222.

Beier, M. 1931. Neue Pseudoscorpione der U. O. Neobisiinea. Mitt. Zool. Mus. Berlin, 17:299-318.

Beier, M. 1932. Pseudoscorpionidea I. Subord. Chthoniinea et Neobisiinea. Tierreich, 57:1-258.

Beier, M. 1954. Pseudoscorpiondea. Beitr. Fauna Perus, 4:1-12.

Beier, M. 1959. Zur Kenntnis der Pseudoscorpioniden-Fauna des Andengebietes. Beitr. neotrop.
Fauna, 1:185-228.

Ellingsen, E. 1910 Die Pseudoskorpione des Berliner Museums, Mitt. Zool. Mus. Berlin, 4:355-423.

Feio, J. L. A. 1945. Novos pseudoscorpioes da regiao neotropical (com a descri(;ao de uma
subfamilia, dois generos e sete especies). Bol. Mus. Nac., Rio de Janeiro, (N.S., Zool.), 44:1-47.
Heurtault, J. 1980. Le developpement postembryonnaire chez deux especes nouvelles de
pseudoscorpions Olpiinae du Venezuela. Rev. Nordest. Biol., 3:57-85.

Heurtault, J. and J. Rebiere. 1983. Pseudoscorpions des Petites Antilles 1. Chernetidae, Olpiidae,
Neobisiidae, Syarinidae. Bull. Mus. Natn. Hist. Nat., 5(A):59 1-609.

Hoff, C. C. 1945. The pseudoscorpion subfamily Olpiinae. Amer. Mus. Novitates, 1291:1-30.

92

THE JOURNAL OF ARACHNOLOGY

Hoff, C. C. 1964. The pseudoscorpions of Jamaica, part 3. The suborder Diplosphyronida. Bull. Inst.
Jamaica, Sci. Ser. 10(3): 1-47.

Mahnert, V. 1982, Die Pseudoskorpione (Arachnida) Kenyas. 3. Olpiidae. Monitore Zool. Ital. (N.
S.) Suppl. 16:263-304.

Mahnert, V. and R. Schuster. 1981. Pachyolpium atlanticum n. sp., ein Pseudoskorpion aus der
Gezeitenzone der Bermudas — Morphologie und Okologie (Pseudoscorpiones: Olpiidae). Rev.
suisse Zool., 88:265-273.

Muchmore, W. B. 1977. Preliminary list of the pseudoscorpions of the Yucatan Peninsula and
adjacent regions, with descriptions of some new species (Arachnida: Pseudoscorpionida). Assoc.
Mexican Cave Stud. Bull., 6:63-78.

Manuscript received September 1985, revised November 1985.

Valerio, C. E. 1986. Mygalomorph spiders in the Barychelidae (Araneae) from Costa Rica. J.
ArachnoL, 14:93-99.

MYGALOMORPH SPIDERS IN THE BARYCHELIDAE
(ARANEAE) FROM COSTA RICA^

Carlos E. Valerio

Escuela de Biologia
Universidad de Costa Rica
Ciudad Universitaria Rodrigo Facio
Costa Rica

ABSTRACT

Two species of Barychelidae, Psalistops venadensis and Trichopelma laselva, are described from
Costa Rica. These are the first records of this mygalomorph family from this country. Both species
were associated with tropical wet forests.

INTRODUCTION

Among the mygalomorph families, the Theraphosidae are particularly diverse
in Costa Rica with more than 30 species, including arboreal, diggers, forest-
dwellers, species associated with pastures, many lowland, and one-high-altitude
species (Valerio 1979, 1980). The Dipluridae are represented by three species and
the Ctenizidae by one (Zuniga 1980). The remaining families have no published
records, except for one paratropidid collected in 1930 and erroneously identified
as Anisaspis tuberculatus (a Caribbean species) by Reimoser (1940). I examined
the specimen and it is an immature Paratropis sp.

In this paper I record the presence of two species in the Barychelidae, a group
previously known, in the New World, only from South America and some
Caribbean islands. Additional specimens have been collected in Panama,
Honduras, Guatemala, and Mexico (W. Gertsch and V. Roth personal
communications).

Although some theraphosids are locally very abundant, long years of collecting
have yielded only three specimens of these barychelids and I think they are
relevant enough to merit taxonomic treatment.

FAMILY BARYCHFLIDAF SIMON

The species from Costa Rica possess a conspicuous rastellum on the anterior
face of the basal cheliceral segment, minute PMF, eye tubercule low and
scopulation of legs as follows: entire and thin pads on palpal tarsus and metatarsi
I and II, thin pad divided by a band of short setae on tarsus III. Scopulae absent
on tarsus IV and metatarsi III and IV. The labium bears some 25 cuspules on
the anterior half of its length (Figs. 7-8).

'Supported by a grant from Vicerrectoria de Investigacion, Universidad de Costa Rica.

94

THE JOURNAL OF ARACHNOLOGY

Map 1. — Map of Costa Rica indicating type localities for Psalistops venadensis (triangle) and
Trichopelma laselva (circle). Shaded area indicates tropical wet formations.

Genus Psalistops Simon

Psalistops Simon 1889, Simon 1892, Waterhouse 1902, Mello-Leitao 1923, Petrunkevitch 1928.

Easily distinghished from Trichopelma by the absence of a clear annular band
on tarsus IV. Lateral spinnerets with apical segment short and conical. Three
species are known from Venezuela, three from the Caribbean islands and two
from Brazil. (Bonnet 1949, Brignoli 1983).

Psalistops venadensis^ new species
Figs. 1-3, 6, 7, 10, 11; Map 1

Type. — Female from El venado, San Carlos, Alajuela Province (10° 33' N, 84°
24' W) collected by C. E. Valerio, deposited in the Museo de Zoologia
(Universidad de Costa Rica). An additional specimen (immature female) from
Tortuguero, Limon Province 10° 29' N, 83° 30' W) was examined and deposited
in the MZUCR.

Diagnosis. — Closely related to the Venezuelan P zonatus Simon, characterized
by the presence of two strong spines ventrally on the base of tibia III in addition

VALERIO— BARYCHELIDS FROM COSTA RICA

95

to two apical ones (Figs. 5-6), larger size (21 mm total length), dark coloration
and size of spermathecae (1.5 mm wide). Separated from the genotype P.
melanopygius Simon by the shape of spermathecae (head divided into large
protuberances vs. smooth, see Fig. 10). Note: the third Venezuelan species is R
tigrinus Simon, but the specimen labeled as type in the MNHN in Paris, number
10716 - 3.244, is a small prodidomid rather than a barychelid (the type must be
lost or misplaced). It can be separated from Trichopelma laselva n. sp. (besides
the generic characters) by its large AME (Fig. 2), the presence of three denticles
on tarsal paired claws (Fig. 11), considerably less cuspules on maxillary segments
and light body coloration.

Female. — Basal cheliceral segment with conspicuous rastellum. Sternum,
labium, maxillae and ventral side of coxae covered with short spines. Maxillae
with some 30 cuspules near medial border. Labium with some 25 cuspules
covering anterior half. Carapace covered with fine pubescence. All legs covered
dorsally with setae, all tarsi with few capitated hairs dorsally. Metatarsi I and II
with one strong distal spine on ventral side, metatarsi III and IV with two on
ventral side and metatarsi III alone with two such spines dorsally. Tibia III with

96

THE JOURNAL OF ARACHNOLOGY

Figs. 2-4. — Ocular area (dorsal view) and spinnerets (ventral view): 2-3, Psalistops venadensis; 4,
Trichopelma laselva. Scale line 1 mm.

only two strong spines on ventral side, located distally (Fig, 6). Tarsal claws with
series of three denticles (Fig. 11). Ocular area dark, prominent, with conspicuous
pair of setae anteriorly near clypeus, AME twice as large as ALE (Fig. 2).
Thoracic furrow transverse. General coloration light tan with ocular area black,
longitudinal series of five clear patches dorsally on abdomen and matching
transversal clear bands on sides (Fig. 1). Posterior lateral spinnerets long with
small triangular apical segment (Fig. 5). Measurements in mm: carapace 6.3,
sternum 2.8, abdomen 8.0, palp 11.5, leg I 15.0, leg II 13.5, leg III 13.5, leg IV
20.0. Spermatheca 0.9 in width, divided in two capitated regions, covered by
rounded spermathecal glands (Fig. 1).

Male. — Unknown.

Distribution. — Known from the Northern lowlands (Map 1).

Natural History. — Both specimens were collected under forest litter in disturbed
areas, in tropical wet formations.

Genus Trichopelma Simon

Trichopelma Simon 1888, Simon 1889, Simon 1891, Mello-Leitao 1923.

This genus is characterized by the presence of an annular pale band (suture?) on tarsus IV (better
seen from the sides and ventrally). One species is known from the Dominican Republic and two from
Brazil.

VALERIO— BARYCHELIDS FROM COSTA RICA

97

Figs. 5-9. — Right tibia III (ventral view) and prosoma (ventral view): 5, Psalistops zonatus; 6-7,
Psalistops venadensis; 8-9, Trichopelma laselva. Scale line 1 mm.

Trichopelma laselva new species
Figs. 4, 8, 9, 12, 13; Map 1

Type . — Female from Finca La Selva Station, near Puerto Viejo de Sarapiqui,
Heredia Province (10° 25' N, 84° 01 'W), collected by V. Roth, deposited in the
American Museum of Natural History.

Etymology. — The name refers to the type locality.

Diagnosis, — This species shows the suture ring on tarsus IV characteristic of
the genus. Valid comparisons cannot be made with its closest goegraphic
neighbor, T nitida Simon from the Caribbean, since it is known from a single
male specimen and sizes and proportions vary considerably with sex. Taken into
consideration that geographic ranges in mygalomorphs tend to be small (valerio
1979) and that T. nitida is isolated in a Caribbean island, I think it is safe to
consider this Costa Rican female a separate species.

Female. — Basal cheliceral segment with conspicuous rastellum. Sternum,
labium and maxillae covered with long black setae. Maxillae with some 40
cuspules anteriorly near medial border. Labium with some 25 cuspules on
anterior half (Fig. 8). Tarsi and metatarsi on all legs covered with spines. Tibia
HI with two slender spines ventrally near distal end (Fig, 9). Paired tarsal claw
with one denticle (Fig. 13). Ocular area prominent, anterior eyes in strongly
procurved line, AME smaller than ALE (Fig. 4). Thoracic furrow transverse.
General coloration dark brown, with pattern of light transverse bands on
abdomen, similar to that of P. venadensis (Fig. 1). Measurements in mm;
carapace 9.0, sternum 3.5, abdomen 12.0, palp 12.0, leg I 21.0, leg H 22.0, leg
III 19.0, leg IV 28.0. Spermathecae 0.9 in width, similar in shape to that of P.
venadensis n. sp. (Fig. 12).

98

THE JOURNAL OF ARACHNOLOGY

12-13, Trichopelma laselva. Scale line 0.5 mm.

Male. — Unknown,

Distribution. — Known only from the type locality (Map 1).

Natural History. — Collected in a densely forested area, in a tropical wet
formation.

ACKNOWLEDGMENTS

I am indebted to R. Raven for a critical review of the manuscript and for
sending me his unpublished keys for the mygalomorphs. I was able to study the
collections of the Museum Natural d’Histoire Naturelle, the Zoologischen
Museum Wien and the American Museum of Natural History, through the
generosity of M. Vachon, J. Gruber and N. Platnick.

LITERATURE CITED

Bonnet, R 1949. Bibliographia Araneorum. Vol. 11. Imprimerie Douladoure, Toulouse, pp. 3797,
4688-4680.

Brignoli, P. M. 1983. A catalogue of the Araneae. Manchester University Press, Manchester, 755 pp.

VALERIO— BARYCHELIDS FROM COSTA RICA

99

Mello-Leitao, C. F. 1923. Theraphosoideas do Brazil. Rev. Mus, Paulista, 13:1-438.

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

Reimoser, E. V. 1940 Wissenschaftliche der osterreischischen biologischen expedition nach Costa Rica.
Ann. Naturhist. Mus. Wien, 50:328-386.

Simon, E. 1888. Etudes arachnologiques 21. Descriptions d’especes et de genres nouveaux de
FAmerique Centrale et des Antilles. Ann. Soc. Entomol. France, 8:203-216.

Simon, E. 1889. Arachnides. In: Voyage de M. E. Simon au Venezuela. Ann. Soc. Entomol. France,
9:169-220.

Simon, E. 1892. Histoire naturelle des Araignees. Librarsic Roret, Paris, 256 pp.

Valerio, C. E. 1979. Aranas teraf6sidas de Costa Rica (Araneae, Theraphosidae): Psalmopoeus
reduncus, redescripcion, distribucion y el problema de dispersion en terafdsidos. Rev. Biol. Trop.,
27:301-308.

Valerio, C. E. 1980. Aranas terafosidas de Costa Rica (Araneae, Theraphosidae). Sphaerobothria,
Aphonopelma, Pterinopelma, Citharacanthus, Crypsidromus y Stichoplastus. Rev. Biol. Trop.,
28:271-296.

Waterhouse, C. O. 1902. Index Zoologicus. London, 421 pp.

Zuniga-Vega, C. M. 1980. Lista anotada de especies de aranas de Costa Rica. Brenesia, 18:310-352.

Manuscript received August 1985, revised January 1986.

Locket, N. A. 1986. Albinism and eye structure in an Australian scorpion, Urodacus yaschenkoi
(Scorpiones, Scorpionidae). J. ArachnoL, 14:101-1 15.

ALBINISM AND EYE STRUCTURE IN AN AUSTRALIAN
SCORPION, URODACUS YASCHENKOI
(SCORPIONES, SCORPIONIDAE)

N. A. Locket

Department of Anatomy and Histology
University of Adelaide
Box 498, G.RO.

Adelaide

South Australia 5001

ABSTRACT

Two albino individuals of Urodacus yaschenkoi (Birula 1903) were caught among a normally
pigmented population in South Australia. These are illustrated and discussed.

Microscopy of the eyes shows that the melanin granules normally present in the epidermis and in
the retina are represented by non-pigmented premelanosomes. Though the components of the eye are
all present, the rhabdoms are abnormal, consisting of reduced and disorganised microvilli. The
phaospheres, present in retinula cells, are also abnormal. The nature of the defects and their possible
consequence are discussed.

INTRODUCTION

Albinism, a specific lack of melanin pigment, is well known in a wide variety
of animals, including crabs, isopods and insects, but does not appear to have
been recorded in scorpions. The occurrence of depigmented arthropods, including
scorpions, is well known, most of them recorded from caves which provide a
totally dark environment in their depths. Commonly the pallor is associated with
the loss of eyes: Mitchell (1968, 1972) described the troglobite genus
Typhlochactas, eyeless pale scorpions from caves in Mexico, and Mitchell and
Peck (1977) later described another species of this genus, also pale and without
eyes, but from forest litter, not a cave. Francke (1977, 1978) has described
troglobite, i.e. obligate cave-dwelling, scorpions belonging to the family
Diplocentridae. One of these, Diplocentrus anophthalmus, is pale and has no
eyes. The other two species appear less extreme; D. mitchelli has minimal
pigmentation and markedly reduced eyes, but D. cueva has only slightly reduced
pigment and small median eyes; the lateral eyes are described as equal to related
epigean examples of Diplocentrus.

The eye structure of scorpions has been the subject of a number of studies,
from the early work of Lankester and Bourne (1883), Parker (1891), and
Scheuring (1913) to the electron microscope studies of Bedini (1967), Belmonte
and Stensaas (1975), Fleissner and Schliwa (1977), and Schliwa and Fleissner

102

THE JOURNAL OF ARACHNOLOGY

Fig. 1. — Diagram of scorpion median eye. Lens, 1, is a thickening of cuticle, 2. Epidermis, 3,
contains pigment which extends to limbus 4. Modified epidermal cells form transparent vitreous, 5.
Deep portion of retina contains pigment cells, 6, nerve fibre bundles, 7, and bodies of arhabdomeric
cells, 8. Retinula cell nuclei, 9, and phaospheres, 10, are deep to rhabdoms, 11. In light-adapted
condition rhabdoms are shielded by pigment, 12.

(1979, 1980). It was early recognised that the structure is different in the median
and lateral eyes; only the former will be considered in the present paper.

The median eyes (diagram, Fig. 1) occur close to each other on either side of
an ocular tubercle approximately in the middle of the carapace. They have a
single lens, derived from and continuous with the cuticle. Between lens and retina,
in the median but not the lateral eyes, is a layer of cells called the vitreous or
lentigen by various authors. This layer, continuous with the epidermis at the edge
of the eye, does not contain melanin granules, but the surrounding epidermis, and
particularly the cells at what may be called, by analogy with vertebrate eyes, the
limbus, are packed with these granules.

Separating the vitreous from the retina is a preretinal membrane, apparently
of cuticular origin, and continuous with a postretinal membrane of similar
structure.

The retina thus enclosed contains three classes of cells. The deep aspect of the
retina is lined by pigment cells, with flattened nuclei disposed parallel to the
postretinal membrane and packed with melanin granules. Interstitial pigment cells
occur within the retina at the level of the retinula cell nuclei. Between the outer
pigment cells and the bases of the retinula cells are the arhabdomeric cells
described by Schliwa and Fleissner (1979). The retinula cells, which show regional
specialisation, occupy most of the thickness of the retina. From their basal
aspects arise the nerve fibres, which form bundles amongst the pigment cells
before they penetrate the postretinal membrane and form the optic nerve. Further
details of retinal structure are given under Results.

The chance finding of two scorpions lacking the normal pigmentation in and
around the eye, and elsewhere in the body, and thus presumably albinos.

LOCKET— ALBINO SCORPION EYE

103

suggested that a comparison of their eyes with the normal might be of interest.
This account is now presented.

MATERIALS AND METHODS

The two individuals now reported were caught in an area of scrub near Berri
in the Riverland of South Australia. The burrow mouths, which have a
characteristic appearance (Shorthouse 1971), were identified, and a plastic
vending machine cup buried so that its lip was level with the floor of the mouth
of the burrow. The trapped burrows were marked, and revisited about 1.5h after
sunset, when many scorpions had emerged from their burrows and fallen into the
cups. When the catch was inspected later indoors, it was noticed that two
specimens lacked the normal pigmentation. No other albinos have been caught
in previous or subsequent trapping in the same area. One albino was kept in the
laboratory for several months, during which time it was fed mealworms and other
insects. It showed no behavioural differences from normal examples.
Measurements of hand length and carapace plus first five metasomal segments
were made on this specimen following Shorthouse (1971), and it falls into
Shorthouse’s group C, i.e. fourth instar.

The second albino, of similar size, was killed and pieces of tissue, including the
eyes, were fixed for electron microscopy in 2% paraformaldehyde and 2.5%
glutaraldehyde in O.IM phosphate buffer, pH 7.3. The tissue was fixed for 2h,
postfixed in 1% osmium tetroxide in O.IM phosphate, dehydrated through an
ethanol series and taken through propylene oxide to Araldite, in which it was
embedded.

Sections along and perpendicular to the optic axis were cut for light and
electron microscopy on a Cambridge Huxley ultramicrotome, using glass knives.
Sections for light microscopy were cut at 1 /um, stained with methylene blue/
azure H followed by basic fuchsin/ borax, and mounted in immersion oil. Thin
sections were stained with alcoholic uranyl acetate followed by Reynolds’ (1963)
lead citrate, and examined in a JEOL 100c electron microscope.

Sections of the eye of a normal scorpion, of the same size and from the same
locality, were prepared as a control. Experiments showed that a primary fixative
containing 0.5g sucrose/ 100 ml gave improved fixation, and this was used for the
control tissue.

RESULTS

General appearance. — Normal specimens of Urodacus yaschenkoi (Birula, 1903)
(Fig. 2), are predominantly buff to pale brown. The mesosoma appears darker
and somewhat grey owing to the dark coloured viscera beneath the cuticle. The
aculeus, leg joints and tarsal claws, and the fingers of the chelae and chelicerae
are reddish to dark brown due to dense sclerotization. The fifth metasomal
segment and vesicle, and the eyes and their surrounds, are dark brown to black
due to pigmentation.

The albinos (Fig. 3) differed from the normal in the pallor of the metasoma
and vesicle, and complete absence of pigment from the eyes and their surrounds.
The mesosoma showed some of the normal darkness, but was noticeably paler

104

THE JOURNAL OF ARACHNOLOGY

Fig. 2. — Urodacus yaschenkoi, normal. The darkness of chelae and fingers, aculeus, tarsal claws
and leg joints is due to dark sclerotization whereas that of eyes and surrounds, vesicle and metasomal
segments is caused by pigmentation. Scale = 10 mm.

Fig. 3. — Urodacus yaschenkoi, albino. Sclerotized parts resemble normal, but eyes, vesicle and
metasomal segments lack pigment. Scale = 10 mm.

LOCKET— ALBINO SCORPION EYE

105

than the normal. Those parts that are normally dark due to sclerotization were
no less so than in the normal scorpion.

Eye structure. — The structure of the median eyes accords well with the
descriptions given by Bedini (1967), Fleissner and Schliwa (1977), Fleissner and
Siegler (1978) and Fleissner and Heinrichs (1982) and summarised in the
Introduction. Sections show that the overall architecture is the same in the
normal and albino. Figs. 4, 5. The cuticular lens has the same configuration, and
stains similarly. The columnar vitreous cells, with basal nuclei, are of similar
shape and size, and were transparent in both. The epidermis surrounding and up
to the limbus in the normal contain numerous pigment granules. These granules
are represented in the albino, but are not pigmented. They appear the same as
those in the retina, described below. The preretinal membrane, situated at the
junction of retina and vitreous, and the postretinal continuation of this
membrane, called the sclera by Bedini (1967), are well developed in both.

The retina contains the populations of cells outlined in the Introduction. The
normal retina (Fig. 4) has a layer of densely pigmented cells lining the eye cup
beneath the postretinal membrane, and extending to the limbus, where they lie
close to the pigmented epidermal cells. These cells have somewhat flattened
densely staining nuclei and their cytoplasm is packed with pigment granules.
Electron microscopy shows these as dense granules, in at least some cases clearly
membrane-bound (Fig. 12).

The ‘pigment’ cells are also present in the albino (Fig. 5) in which the
disposition of their nuclei is better seen than the normal, due to the absence of
pigment. Where the normal has melanin granules, the albino has an equal or
greater density of granules, but these are unpigmented. These granules are
membrane bound and have a finely granular content; some contain bars or rings
with a paracrystalline appearance. Similar granules occur among the melanin
granules in the normal, and they are considered to be premelanosomes, (Fig. 13).

A second population of pigment cells is present in the normal, interspersed
with the retinula cells in the substance of the retina. These cells have nuclei at
the same level as those of the retinula cells, and their perinuclear cytoplasm is
densely packed with melanin granules. They have processes that extend vitreally
between the retinula cells, and which also contain pigment granules.
Corresponding cells have not been identified with certainty in the albino.

Arhabdomeric cells with cell bodies located deeper in the retina than those of
retinula cells, and processes which penetrated between the retinula cells to the
level of the rhabdom bases were described by Schliwa and Fleissner (1979). Cells
conforming to their description are present in both normal and albino eyes, and
profiles corresponding to Schliwa and Fleissner’s description are present in
tangential sections of retina at the level of the rhabdom bases.

Axonal profiles containing dense vesicles, and identified by Fleissner and
Schliwa (1977) as neurosecretory fibres, are also present in normal and albino
retinas.

Most of the volume of the retina is made up by the retinula cells. These extend
from close against the lining pigment cells to the preretinal membrane. They are
interspersed near their bases by the arhabdomeric cells and intraretinal pigment
cells, but are the sole cell type present in the vitread half of the retina. The size
and extent of the retinula cells are similar in the normal and the albino. The
location and staining properties of their nuclei are also similar, and in both cases

106

THE JOURNAL OF ARACHNOLOGY

Fig. 4. — Normal median eye, coronal section. Retina, 1, is separated from lens, 2, by vitreous 3.
Epidermis is heavily pigmented, particularly at limbus, 4. Retina contains abundant pigment. Scale
= 100 /um.

Fig. 5. — Albino median eye, coronal section. Retina, 1, epidermis, 2, and limbus, 3, are devoid of
pigment. Scale = 100 /um.

the cells give rise to nerve fibres that form bundles in the deep part of the eye
before piercing the postretinal membrane to form the optic nerve.

The normal retinula cells each contain a rounded body called by Lankester and
Bourne (1883) a phaosphere, presumably from their dark appearance in stained

Fig. 7. — Retinal periphery, albino. Granules replace pigment in peripheral cells, 1, and retinula cells,
2, Rhabdoms, 3, and phaospheres, 4, are degenerate. Nuclei, and nerve fibres, 5, appear normal. Scale
= 100 /Ltm.

sections. The phaospheres, of which at least one appears to be present in each
retinula cell, are usually located deep to the nucleus and are of comparable size.
Ultrastructurally they appear as an aggregation of dense granules approximately
40 nm in diameter (Fig. 10). The phaospheres contain vacuoles that stain with

Fig. 6. — Retinal periphery, normal. Abundant pigment in peripheral pigment cells, 1, vitread end
of retinula cells, 2, and in substance of retina, 3. Rhabdoms, 4, and phaospheres, 5, well formed.
Scale = 100 jum.

LOCKET— ALBINO SCORPION EYE

108

THE JOURNAL OF ARACHNOLOGY

basic fuchsin but not with the blue dyes, and that are described as refractile by
other authors (Bedini, 1967). These vacuoles contain scattered granules, some
similar to those aggregated and others smaller and frequently arranged in a
paracrystalline array. Phaospheres have not been reported from other cell types,
and their function remains unknown.

The phaospheres in the albino are abnormal (Fig. 11). In most cases they are
represented by smaller and less regular clumps of dense amorphous material
without the normal granular substructure, though some show the normal rounded
profile and contain vacuoles.

The retinula cell processes vitread to the nuclei are in close mutual apposition,
and from the abutting processes the rhabdomeres are formed over a distance of
some 70 jum. Typically the cells are grouped in fives; the rhabdoms are
consequently star shaped in cross sections in most cases, though other patterns
are common, particularly in the periphery of the retina. Electron microscopy of
the rhabdoms confirms the observations of Bedini (1967) and Fleissner and
Schliwa (1977) that they consist of close arrays of uniform microvilli,
approximately 80 nm diameter and up to 2 fim in length (Fig. 8).

An unexpected finding in the albino is that the rhabdoms are markedly
degenerate. The length of the retinular cell along which rhabdomeres are formed
is much reduced, and the normal star pattern is disorganised or absent. Electron
microscopy shows that the normal grouping in fives is present, but that the
rhabdomeres are represented only by small and disorganised arrays of microvillar
material, lacking the density and regularity of the normal, (Fig. 9).

The normal retinular cells contain abundant melanin granules, some present in
the base of the cell close to the nucleus, and even extending into the nerve fibres.
Most granules are concentrated in the vitread ends of the cells where they
effectively screen the rhabdoms from incoming light. These melanin granules
appear identical to those in the pigment cells. Premelanosomes, membrane bound
and with laminated bars or rings in them, are also found among the granules
(Fig. 12).

The melanin granules are absent in the albino, their place being occupied by
numerous premelanosomes that appear similar to those of the normal, except that
they show a greater size range and are present in greater density (Fig. 13), These
premelanosomes show the same distribution along the cell as the melanin
granules of the normal, suggesting that they may have undergone circadian
migration as do the normal melanin granules (Scheuring, 1913; Fleissner, 1974),
or that they were permanently in the light regime position.

DISCUSSION

Before discussing the differences between the normal and albino examples we
may review the factors affecting the colouration of scorpions. As Cutler and
Richards (1972) point out, the dark colours commonly seen in arachnids are
partly due to brown exocuticle and partly to pigment granules present in the
epidermis, the features being present singly or alone.

Some scorpions are uniformly pale; Centruroides sculpturatus is a pale
yellowish or clay colour, and the newborn young of Urodacus manicatus, and of
other species illustrated in the literature, are white all over. This pallor is due to
poorly sclerotized cuticle in the case of the young, but in C. sculpturatus the

LOCKET— ALBINO SCORPION EYE

109

Fig. 8. — T. S. rhabdom, normal, E. M. Each retinula cell, 1, contributes microvilli to two arms
of the star-shaped rhabdom. Cytoplasm contains mitochondria, ribosomes and abundant melanin
granules. Scale = 5 jum.

Fig. 9. — T. S. rhabdom, albino, E. M. Groups of five retinula cells, 1, bear reduced and
disorganised microvilli. Mitochondria and ribosomes deficient. Pigment granules replaced by non-
pigmented premelanosomes. Scale = 5 pm.

cuticle is well sclerotized but pale, and there is little epidermal pigment. Other
scorpions are very dark; Heterometrus species from Asia and Urodacus manicatus
from Australia are cases in point. Of these U. manicatus certainly owes its
colouration to both dark cuticle and epidermal pigmentation. The fingers and

THE JOURNAL OF ARACHNOLOGY

1 10

Fig. 10. — Phaosphere, normal, E. M. Phaosphere consists of aggregated granules. Similar granules
present in cytoplasm and within vacuole. Paracrystalline array of smaller granules in some vacuoles.
Scale = 1 ;um.

Fig. 11. — Phaosphere, albino, E. M. Dense amorphous mass replaces well-defined granules. Scale
= 1 )um.

Fig. 12. — Pigment granules, normal, E. M. Melanin granules are dense and membrane-bound.
Adjacent premelanosome contains granular matrix and lamellae. Scale = 1 /im.

Fig. 13. — ‘Pigment’ granules, albino, E. M. Membrane-bound granules resemble normal
premelanosomes, but occur in greater size range. No melanin present. Scale = 1 ^tm.

LOCKET— ALBINO SCORPION EYE

111

aculeus are darkly sclerotized, but most of the darkness of the tergites is due to
epidermal pigment. Other scorpions are patterned; Lychas marmoreus owes its
pattern to epidermal pigment beneath a pale cuticle. Even the fingers are pale,
though well sclerotized, recalling the pallor of C. sculpturatus.

Whatever the amount of pigment elsewhere in the epidermis, the eyes are
almost always black. This is the case in C. sculpturatus, and the white newborn
scorpions have densely pigmented eyes. The only cases in which eye pigment is
usually lacking is in certain cavernicolous examples, e.g, Typhlochactas described
by Mitchell and Peck (1977) which has lost its eyes altogether. Another
troglobitic species, Diplocentrus mitchelU is described by Francke (1978) as
retaining some eye pigment, even though the rest of the body is depigmented and
the eyes are vestigial.

Comparing the albino and normal U. yaschenkoi we see that the darkness of
the fifth metasomal segment and vesicle, and that in and around the eyes is due
to pigment, but that the darkness of the chelae and chelicerae, the aculeus and
tarsal claws, and the leg joints is due to darkly sclerotized cuticle.

Hackman (1974) notes that some insects may have melanin within the cuticle
itself, besides the intracellular granules, but this does not appear to be the case
in the scorpions examined here.

The process of melanin formation is summarized by Hogan, Alvarado and
Weddell (1971) for the human retina: Rough endoplasmic reticulum within the
melanin-forming cell elaborates the enzyme tyrosinase, which is thought to be
transferred to the Golgi apparatus. Here it k packaged into the membrane-bound
premelanosomes, which contain a finely striated protein framework with which
the tyrosinase is associated. On this framework tyrosine is polymerised to form
melanin; when polymerisation is complete and the granule mature, no further
tyrosinase activity is demonstrable. In albinos the melanocytes contain vesicles
and premelanosomes, containing a protein framework but no melanin; such
individuals lack tyrosinase.

The present findings accord well with this scheme, despite the phylogenetic
disparity between scorpions and humans and the smaller size of the invertebrate
pigment granules. Melanin granules and premelanosomes are present in the
normal, but the albino contains only premelanosomes, many containing
laminated bodies but not melanin, and in a wider range of sizes than normal.

Accounts on vertebrate (Nyhan 1981) and insect (Hackman 1974) melanin
formation agree on the biochemical pathway. An enzyme, tyrosinase or o-
diphenyl oxidase, catalyses the conversion of tyrosine to dopa and then to dopa-
quinone. The remaining steps, which can proceed non-enzymatically, at least in
vertebrates, lead to moieties which undergo condensation to form a repeating
polymeric structure of high molecular weight. This dark pigment, eumelanin, is
bound to protein.

Hackman (1974) makes the point that the production of melanin, cuticular or
intracellular, and the hardening or sclerotization of cuticular proteins both
involve the conversion of tyrosine to dopa and dopa-quinone. The hardening of
cuticle, in many cases accompanied by darkening (Andersen 1980), involves the
tanning of a protein by quinones, which are derived from an o-diphenolic
compound by an oxidase. In at least some cases the diphenol is tyrosine from
the haemolymph (Pryor 1940, Fraenkel and Rudall 1940, 1947). The colour to
which cuticle eventually tans will depend on the proportion of quinones to

THE JOURNAL OF ARACHNOLOGY

1 12

residual diphenols; the quinoid, oxidised, state of the protein-bound material is
dark (Hackman 1959).

Though the tyrosine-dopa-quinone pathway may be involved both in the
hardening of cuticle and in melanin formation, the enzymes responsible for the
two trains of reactions are probably different. Dennell (1958) found that injection
of phenylthiourea into blowfly larvae completely suppressed the haemolymphal
tyrosinase activity, preventing the formation of the normal black banding of the
puparia, but not affecting their formation and sclerotization.

We may postulate a selective disability of tyrosinase in the present case, in
which the formation of melanin is suppressed in an individual with normally
sclerotized cuticle.

Fleissner (1974) confirmed Scheuring’s (1913) observation that the retinal
pigment granules normally undertake circadian migrations, being concentrated in
the vitread part of the retina by day, and retreating to the basal part by night.
Fleissner also showed that the sensitivity of the retina altered by 3 to 4 log units
with the movement of the pigment. It is clear from examining sections of the
retina in the light-adapted state that pigment is present vitread to the rhabdoms
in such quantities as greatly to attenuate the amount of light that would penetrate
to them.

Fleissner and Schliwa (1977) demonstrated neurosecretory fibres in the retina
and suggested that these control the pigment migration, later shown by Fleissner
and Fleissner (1978) to be mediated by the optic nerves. Section of the nerve did
not merely abolish the pigment movement, which would be expected if the
movement depended on visual impulses reaching the central nervous system, but
the pigment assumed the daylight position, normally associated with an
illuminated eye. They concluded that the pigment migration was controlled by
efferents in the optic nerve. The cell bodies of these neurosecretory fibres were
located in the supraoesophageal ganglion by Fleissner and Heinrichs (1982).

In view of the close association of the pigment with visual function it would
not be surprising to find a functional deficit due to the lack of pigment in the
albinos, but the structural disorganisation of the rhabdoms and of the
phaospheres is unexpected. It may be that the integrity of the rhabdoms depends
on protection from excess illumination, normally provided by the burrowing
habit and by the screening pigment. That the observed degeneration was not
caused by the laboratory regime is shown by the control, which was kept in the
same conditions, and in which the rhabdoms appear large and well ordered. The
nature and function of the phaospheres remain unknown, but it is interesting that
they too show degeneration in the absence of pigment and of intact rhabdoms.

The presence of the much reduced and degenerate rhabdoms in the albino
suggests that eye function was poor. Despite this, it had survived to the fourth
instar, managing to catch prey while being surrounded by normal individuals,
important potential predators. This survival of albino, and probably blind,
individuals confirms the relative unimportance of vision in scorpions.

In diurnal animals albinism may be disadvantageous, since pale individuals
may be more easily seen by predators. In dark environments this pallor will not
matter, and many cave forms are depigmented. Some cave dwellers have
degenerate eyes; others, including the Typhlochactas species described by Mitchell
and Peck (1977) and certain opiliones (Briggs 1974), have lost their eyes
altogether.

LOCKET— ALBINO SCORPION EYE

113

Though not a cave dweller, U. yaschenkoi does live in a burrow which spirals
underground, and so is dark at the end (Shorthouse 1971, Koch 1978, Shorthouse
and Marples 1980). The scorpion comes to the burrow mouth, and commonly
beyond it, at nightfall. This habit was certainly present in the albinos, because
the trapping method depends on it. This evening emergence suggests that the
circadian behavioural rhythm of the species was present, in spite of the absence
of pigment and disordered eye structure.

Consequences of albinism on scorpion metabolism other than those to the eyes
are not obvious. Cloudsley-Thompson (1979) discusses the possibility that
cuticular melanin may reduce water loss, and quotes Kalmus’ (1941) finding that
dark varieties of Drosophila withstand dehydration better than pale ones do.
Kalmus’ paper however does not distinguish unequivocally between darkness due
to cuticular sclerotization or melanin production, cuticular or epidermal. Dark
coloration is not associated with resistance to desiccation in scorpions. Edney
(1977) points out that black scorpions from rain forests have high cuticular
permeabilities while some pale desert scorpions have the lowest permeabilities
recorded from any arthropod. Supposed waterproofing properties of melanin are
probably unimportant in the present context, since U. yaschenkoi is a pale
species, the only pigmented regions, one metasomal segment, the vesicle and the
ocular area, representing a small fraction of the surface area.

Little can be said about the genetics of albinism in scorpions, since the
phenomenon is only known from the two specimens now described. Albinism in
man, which is due to a recessive gene, occurs with a frequency of about 1 in
20,000. If the rate is comparable in scorpions it would take a very large collecting
effort to determine the rate in the wild. Shorthouse and Marples (1982) have
shown that U. yaschenkoi takes at least 6y to reach maturity and that the
gestation period is 18 months. In his extensive study of the species Shorthouse
(1971) never observed mating, nor did he find females with first instar scorpions.
These features of the life history, with the burrowing habit, mean that breeding
experiments would be highly unlikely to succeed.

ACKNOWLEDGMENTS

I wish to thank my father, G. H. Locket, for introducing me to the study of
arachnids, and for his invaluable example and discussions over the years. I thank
my wife Deirdre for assistance in the field, Ms. Cathy Shanahan for advice on
localities and collecting methods and discussion of scorpion genetics. Professor
N. E. Hadley for valuable comments on the problems. Professor G. E. Rogers
for discussions on biochemistry and Mrs. Barbara Sheldon for typing the
manuscript.

LITERATURE CITED

Andersen, S. O. 1980. Cuticular sclerotization. Pp. 185-215, in Cuticle techniques in arthropods (T.

A. Miller, ed.). Berlin, Heidelberg, New York: Springer- Verlag.

Bedini, C. 1967. The fine structure of the eyes of Euscorpius carpathicus L. (Arachnida, Scorpiones).
Arch. Italiennes Biol., 105:361-378.

Belmonte, C, and L. J. Stensaas. 1975. Repetitive spikes in photoreceptor axons of the scorpion eye.
Invertebrate eye structure and tetrodotoxin. J. Gen. Physiol., 66:649-655.

THE JOURNAL OF ARACHNOLOGY

I 14

Briggs, T. S. 1974. Troglobitic harvestmen recently discovered in North American lava tubes.
(Travuniidae, Erebomastridae, Triaenonychidae: Opiliones). J. ArachnoL, 1:205-214.

Cloudsley-Thompson, J. L. 1979. Adaptive functions of the colours of desert animals. J. Arid. Env.,
2:95-104.

Cutler, B. and A. G. Richards. 1972. Sclerotization and the localization of brown and black pigments
in Chelicerates (Arthropoda). Zool. Jb. Anat., 89:404-421.

Dennell, R. 1958. The amino acid metabolism of a developing insect cuticle: The larval cuticle and
puparium of Calliphora vomitoria. II. The non-specific hydroxylation of aromatic amino-acids and
the production of polyphenols by the cuticle. Proc. Roy. Soc. London. B., 148:280-284.

Edney, E. B. 1977. Water balance in land arthropods. Berlin, Heidelberg, New York: Springer-Verlag,

282 pp.

Fleissner, G. 1974. Circadiane Adaptation und Schirmpigment verlagerung in der Sehzellen der
Medianaugen von Androctonus australis L. (Buthidae, Scorpiones). J. Comp. Physiol., 91:399-416.

Fleissner, G. and G. Fleissner. 1978. The optic nerve mediates the circadian pigment migrations in
the median eyes of the scorpion. Comp. Biochem. Physiol. A., 61:69-71.

Fleissner, G. and S. Heinrichs. 1982. Neurosecretory cells in the circadian clock system of the
scorpion Androctonus australis. Cell Tiss. Res. 224:233-238.

Fleissner, G. and M. Schliwa. 1977. Neurosecretory fibres in the median eyes of Androctonus australis
L. Cell Tiss. Res., 178:189-198.

Fleissner, G. and W. Siegler. 1978. Arhabdomeric cells in the retina of the median eyes of the
scorpion. Naturwiss., 65:210-21 1.

Fraenkel, G. and K. M. Rudall. 1940. A study of the physical and chemical properties of the insect
cuticle. Proc. Roy. Soc. London. B., 129:1-35.

Fraenkel, G. and K. M. Rudall. 1947. The structure of insect cuticles. Proc. Roy. Soc. London. B.,
134:1 1 1-143.

Francke, O. F. 1977. The genus Diplocentrus in the Yucatan peninsula, with description of two new
troglobites (Scorpionida, Diplocentridae). Bull. Assn. Mexican Cave Stud., 6:49-61.

Francke, O. F. 1978. New troglobite scorpion of genus Diplocentrus. Scorpionida, Diplocentridae.
Entomol. News, 89:39-45.

Hackman, R. H. 1959. Biochemistry of the insect cuticle. Proc. Int. Congr. Biochem. 4th. Vienna
1958, 12:48-62.

Hackman, R. H. 1974. Chemistry of the insect cuticle. Pp. 215-270, in. The Physiology of Insects.
Vol. VI. (M. Rockstein, ed.). New York, London, Academic Press.

Hogan, M. J., J. A. Alvarado and J. E. Weddell. 1971. Histology of the human eye. Philadephia,
W. B. Saunders.

Kalmus, H. 1941. The resistance to desiccation of Drosophila mutants affecting body colour. Proc.
Roy. Soc. London B., 130:185-201.

Koch, L. E. 1978. A comparative study of the structure, function and adaptation to different habitats
of burrows in the scorpion genus Urodacus (Scorpionida, Scorpionidae). Rec. West. Australian
Mus. 6(2):1 19-146.

Lankester, E. R. and A. G. Bourne. 1883. The minute structure of the lateral and central eyes of
Scorpio and Limulus. Quart. J. Micr. Sci., 23:177-212.

Mitchell, R. W. 1968. Typhlochactas, a new genus of eyeless cave scorpion from Mexico (Scorpionida,
Chactidae). Ann. Speleol., 23:753-777.

Mitchell, R. W. 1972. Typhlochactas elliotti, a new eyeless cave scorpion from Mexico (Scorpionida,
Chactidae). Ann. Speleol., 26:135-148.

Mitchell, R. W. and S. B. Peck. 1977. Typhlochactas sylvestris, a new eyeless scorpion from montane
forest litter in Mexico (Scorpionida, Chactidae, Typhlochactinae). J. ArachnoL, 5:159-168.

Nyhan, W. L. 1981. Inheritable abnormalities of amino acid metabolism. Pp. 305-481, in
Comprehensive biochemistry. Vol. 19A (A. Neuberger and L. L. M. van Deenen, eds.).
Amsterdam, Elsevier.

Parker, G. H. 1891. The eyes in scorpions. Bull. Mus. Comp. Zool., 13:173-208.

Pryor, M. G. M. 1940. On the hardening of the cuticle of insects. Proc. Roy. Soc. London B.,
128:393-407.

Reynolds, E. S. 1963. The use of lead citrate at high pH as an electron-opaque stain in electron
microscopy. J. Cell Biol., 17:208-212.

Scheuring, L. 1913. Die Augen der Arachniden. I. Die Augen der Skorpioniden. Zool. Jb. Abt. Anat.
Ontog., 33:553-588.

LOCKET— ALBINO SCORPION EYE

115

Schliwa, M and G. Fleissner. 1979. Arhabdomeric cells of the median eye retina of scorpions Part
1. Fine structural analysis. J. Comp. Physiol. A., 130:265-270.

Schliwa, M. and G. Fleissner. 1980. The lateral eyes of the scorpion Androctonus australis. Cell Tiss.
Res., 206:95-114.

Shorthouse, D. J. 1971. Studies on the biology and energetics of the scorpion Urodacus yaschenkoi.
Ph.D. Thesis. Australian National University.

Shorthouse, D. J. and T. G. Marples. 1980. Observations on the burrow and associated behaviour
of the arid-zone scorpion Urodacus yaschenkoi (Birula). Australian J. ZooL, 28:581-590.

Shorthouse, D. J. and T. G. Marples. 1982. The life stages and population dynamics of an arid zone
scorpion Urodacus yaschenkoi 1903). Australian J, EcoL, 7:109-1 18.

Manuscript received August 1985, revised December 1985.

. ^ V uW

' E*,, .i,w;i4{*'^'‘

' .(,'•«* ‘‘rtf /t’

:TS’-.f#hk •••'llfji.r

-A

ri«‘' 5' O ,■:<•>

Q?

r^J

•Ji&

‘ ■

« ^.^'illLr-ii ‘ir- ; «■ )-..

fc-'-iiai'

-V V*-k‘AM(jKV.^.^V- _ <

' '‘v'

1986. The Journal of Arachnology 14:1 17

RESEARCH NOTES

LAS CITAS \y¥.PACHYLUS CHILENSIS PARA LA ARGENTINA
(OPILIONES, GONYLEPTIDAE, PACHYLINAE)

Si bien la presencia de Pachylus chilensis (Gray) en Argentina ha sido senalada
en diversas ocasiones, las referencias son en su mayoria repeticiones de citas
anteriores. Solo cuatro de ellas corresponden a especimenes estudiados por el
autor respectivo: Thorell (1877, 1878) consigna la especie para Cordoba; Sorensen
para San Luis (1895) y Buenos Aires (1884), en tanto Roewer (1913) vuelve a
indicar su presencia en esta ultima localidad. La mencion de Roewer (1929) para
“Argentinien (Aconcagua)” debe ser atribuida seguramente a la provincia chilena
de ese nombre, como lo sugiere Ringuelet (1959); por otra parte, la referenda que
este autor — siguiendo a Roewer (1929) — hace para ‘‘Santa Cruz: Estrecho de
Magallanes” (Ringuelet, op. cit.) no es exacta, pues la cita original se limita a
consignar: “Magelhaes-Strasse.”

Aun excluyendo estas localidades cuestionables, P. chilensis parece ocupar en
Argentina un area de distribucion extensa; sin embargo, la especie no se halla
representada en colecciones aracnologicas del pais, ni fueron publicados nuevos
hallazgos. Por este motive, he procurado localizar el material estudiado por
Thorell, Sorensen y Roewer, consultando para tal fin al Naturhistoriska
Riksmuseet de Estocolmo (NRE), el Museo ed Istituto di Zoologia Sistematica
della Universita di Torino (MIZT), el Zoologisk Museum de Copenhage (ZMC)
y el Museum National d’Histoire Naturelle de Paris (MNHN), con los siguientes
resultados:

Thorell (1877, 1878) . — Menciona especimenes provenientes de Cordoba

(hembra y juvenil) de Pachylus granulatus C. L. Koch; esta especie es colocada
en sinonimia de P. chilensis por Sorensen (1884), involucrando con ello la cita
de Thorell. Dichos ejemplares (NRE, Coll. Thorell, N® 76/119, “Cordoba,
Weijenbergh coL”) corresponden en realidad a la especie Sphaleropachylus butleri
(Thorell).

Sorensen (1895). — Tambien aqui se trata de un error de identificacion, pues los
ejemplares examinados por Sorensen (MIZT, Op. 75, Provincia de San Luis,
'‘Pachylus chilensis') son 2 machos adultos y uno juvenil de S. butleri.

Sorensen (1884). — Segun el aracnologo danes, en el Museum Hauniensis
(Copenhage) existirian un macho y una hembra de P. chilensis, colectados por
Kjellerup en Buenos Aires durante el viaje de la Galatea. Al parecer, Sorensen
omitio rotular convenientemente el material, por lo que la localizacion de dichos
especimenes no es segura. En el ZMC se encuentran varios gonileptidos
conservados en seco, provistos de sendas etiquetas: “Buenos Ayres, Galatea”; uno
de ellos es un macho de P. chilensis, en tantos los demas son Acanthopachylus
aculeatus (Kirby), especie muy comun en dicha localidad. Es posible que ese
ejemplar sea uno de los referidos por Sorensen pero no tengo certeza de ello.

Roewer (1913). — Tal como lo consigna este autor, en el MNHN existe
efectivamente un lote de P. chilensis, 2 machos y 2 hembras procedentes de

1986. The Journal of Arachnology 14:1 18

Buenos Aires (MNHN, n- 1601, 41-97); en el rotulo se especifica “det. Roewer,
1912”, sin precisar fecha ni colector.

Lo expuesto plantea un amplio interrogante acerca de la inclusion de P.
chilensis en el elenco opiliologico argentine, toda vez que solo el material del
MNHN — y con margen para la duda, el del ZMC — coincide con lo indicado
en su referenda original. Cabe agregar que existen dos citas de esta especie
tambien para Montevideo, Uruguay (Sorensen 1902, Roewer 1913) ambas
referidas al mismo lote conservado en el Zoologisches Museum de Hamburgo
(ZMH); he tenido ocasion de consultar ese material (ZMH, Montevideo,
Ebrhardt leg. 1870), comprobando que se trata de un par de hembras de A.
aculeatus (aunque a juzgar por las etiquetas contenidas en el mismo tubo faltaria
alii un macho de P. chilensis).

Agradezco a los Doctores T. Kronestedt (NRE), O. Elter (MIZT), H. Enghoff
(ZMC), A. Munoz-Cuevas (MNHN) y G. Rack (ZMH) el envio del material
citado en este trabajo, y al Dr. E. A. Maury y el Sr. J. C. Cokendolpher la
lectura del manuscrito.

LITERATURA CITADA

Ringuelet, R. 1959. Los aracnidos argentinos del orden Opiliones. Rev. Mus. Arg. C. Nat., C. ZooL,
5(2): 127-439, lam. I-XX.

Roewer, C. 1913. Die Familie der Gonyleptiden der Opiliones-Laniatores. Arch. Naturg., 79A(4):1-
256, lam. I.

Roewer, C. 1929. Weitere Weberknechte III. Ill Erganzung der: ’’Weberknechte der Erde” 1923. Abh.
Natur. ver. Bremen, 27(2): 179-284, lam. I.

Sorensen, W. 1884. Opiliones Laniatores (Gonyleptides W. S. olim) Musei Hauniensis Naturh. Tidskr.
(ser. 3), 14:555-646.

Sorensen, W. 1895. Viaggio del dottor Alfredo Borelli nella Repubblica Argentina e nel Paraguay.

XVII. Opiliones Laniatores. Boll. Mus. Zool. Anat. Comp. Univ. Torino, 10(210): 1-6.

Sorensen, W. 1902. Gonyleptiden (Opiliones Laniatores). Ergeb. Hamb. Magalh. Sammelr., 2:1-36.
Thorell, T. 1877. Sobre algunos aracnidos de la Republica Argentina. Period. Zoolog., 2(4):201-218.
Thorell, T. 1878. Sobre algunos aracnidos de la Republica Argentina. Bol. Acad. Nac. Cienc.
Cordoba, 2(3):255-272.

Luis Eduardo Acosta^ Museo Argentine de Ciencias Naturales, Av. Angel
Gallardo 470, 1405 Buenos Aires, Argentina.

Manuscript received June 1985, accepted September 1985.

1986. The Journal of Arachnology 14:1 19

A RANGE EXTENSION OF THE PURSEWEB SPIDER
SPHODROS RUFIPES IN EASTERN KANSAS
(ARANEAE, ATYPIDAE)

Little is known about the behavior or ecology of atypid spiders, aside from
descriptions of tube-web construction and prey capture (McCook 1888, Poteat
1890, Bristowe 1958) and partial descriptions of the natural history of a few
species (Enock 1885, Muma and Muma 1945, Coyle and Shear 1981). Gertsch
and Platnick’s (1980) excellent paper provides a much needed taxonomic revision
of the group as well as known distributions and notes on natural histories of the
Nearctic species of atypids.

Atypids are patchily distributed, though locally abundant. The areas of
abundance are often difficult to locate and associate with climatic and ecological
variables, making their occurrence difficult to predict. They are more common
in the southern U.S., where they are found in forested areas with sandy soil, or
soil with a moderate to low clay content. In these areas, sites with southern
exposures, often along a stream or in fairly mesic situations, seem to be the most
prevalent purseweb habitats (Bristowe 1933, Muma and Muma 1945, pers. comm.
Teeter).

This note extends the range of Sphodros rufipes (Latreille) to the northwest
of previous records and is the first record of the species in Kansas. Previously
published distribution is from eastern Texas to northern Florida and northward
to southern Illinois and Rhode Island (Gertsch and Platnick 1980). S. bicolor and
S. milberti have recently been synonymized with S. rufipes (Gertsch and Platnick
1980).

Wandering adult males of S. rufipes have been found in two separate areas in
Douglas County in northeastern Kansas, syntopically with S. niger (Fitch 1963,
Gertsch and Platnick 1980). Males of S. niger and S. rufipes were captured in
a 2:1 ratio during the five week study period in May-July 1983 in baffle-type
pitfall traps (in prep. Morrow). The traps were placed at approximately 20-30 m
intervals in mixed hardwood forests at the University of Kansas Natural History
Reservation (NHR) in Douglas County, near Lawrence, Kansas and the
Breidenthal Tract, also in Douglas County, 3 km north of Baldwin, Kansas.
Traps were checked every other day at each site during the study period.

S. rufipes is absent from the list of spider species found on the NHR (Fitch
1963), however, a preserved specimen from the NHR was recently uncovered
from a private collection, an adult male S. rufipes, which had been misidentified
as S. fitchi. Adult males of N. rufipes and S. fitchi are easily confused. Both have
a black cephalothorax and abdomen and red legs. The two species can be
discriminated by the extent of the red coloration present on the legs. In males
of S. rufipes the femora and all distal leg segments are completely colored
carmine red. In males of S. fitchi the red coloration is limited to the dorsal
surface of the distal ends of the femora and all distal leg segments. Proximally,
the femora of S. fitchi are a darkish brown to black, the same color as the
cephalothorax. The shape of the sternum and the pattern of the sigilla on the
sternum is also distinctive in the two species. For a more detailed description of
characters, see Gertsch and Platnick (1980). Voucher specimens of male S. rufipes

1986. The Journal of Arachnology 14:120

were sent to the American Museum of Natural History, New York, and the
Museum of Comparative Zoology, Cambridge.

Both the NHR site and the Breidenthal Tract have mixed populations of S.
niger and S. rufipes. The possible occurrence of S. fitchi is currently being
investigated. Both sites are mixed hardwood forest, predominantly oak-hickory,
with moderate understory vegetation. In 1982 over 100 upright tube-webs located
at the bases of trees and shrubs had been marked at NHR and over 150 at the
Breidenthal Tract. Of these, approximately 20% contained spiders which had been
tentatively identified as adult females of S. niger (c.f., Fitch 1963, Beatty 1983).
A May 1984 census indicated only 20% of the originally marked population at
both sites was still present. Only one individual believed to be an adult female
of S. rufipes has been found. Unfortunately, this individual escaped before a
positive identification could be made. One adult female, believed to be 5'. fitchi,
has been found on a forested slope adjoining the Breidenthal site.

Adult females and immatures of both sexes are not known to leave their tube-
webs, except under extreme conditions. Wandering adult males may be found
seasonally. Over 40 males of S. niger and 20 of S. rufipes were captured during
a five week emergence in May-July 1983 at both locations. During the following
year, 35 males of S. niger and 4 of S. rufipes were captured. This seems to be
a large percentage of the population, given the total number of tube-webs
marked, probably contributing to the decline in population density over the past
two years. If these samples represent a typical emergence size, perhaps the total
population is larger than indicated, or males are capable of dispersing over large
distances ( 200 m). A mark, release and recapture experiment using males of S.
niger proved inconclusive. Alternatively, males may have a more rapid
maturation rate than presently suspected. A large number of medium sized tube-
webs were found empty or abandoned following the emergence. Current data
suggest that the 1983 emergence represents a peak year, possibly a cyclic
occurrence, in a population of fluctuating size. Demographic evidence indicates
a high mortality rate during extremely cold winters and hot summer drought
conditions. These temperature extremes may be the primary factors limiting the
ranges of these spiders.

A third species, S. fitchi, which had previously, though rarely, been collected
at the NHR in or near grasslands was not collected in pitfall traps at this site.
Advanced succession at the NHR could be responsible for a decrease in
abundance of this species. Large, uninhabited tube-webs have been found in
prairie and grassland areas at the NHR (c.f., Muma and Muma 1945). S. fitchi
and S. rufipes may prefer more open woodlands or ecotone situations to dense,
forested areas where S. niger is more common. Males of both S. fitchi and S.
rufipes share distinct carmine red legs, an unusual coloration among atypids. This
coloration may be associated with some form of mimicry or aposematic display
advantageous to wandering males (Coyle and Shear 1981).

LITERATURE CITED

Beatty, J. A. 1983. Discovery of the webs of the atypid Sphodros niger. Amer. Arachnol., 28:1 1.
Bristowe, W. S. 1933. Notes on the biology of spiders. — IX. The British species of Atypus. Ann.
Mag. Nat. Hist., ser. 10, 11:289-302.

1986. The Journal of Arachnology 14:121

Bristowe, W. S. 1958. The World of Spiders. Collins, London, 304 pp.

Coyle, F. A. and W. A. Shear. 1981. Observations on the natural history of Sphodros ahhoti and
Sphodros rufipes (Araneae, Atypidae), with evidence for a contact sex pherome. J. Arachnol.,
9:317-26.

Enock, F. 1885. The life history of Atypus piceus Suly. Trans. Entomol. Soc. London, pp. 389-420.

Fitch, H. S. 1963. Spiders of the University of Kansas Natural History Reservation and Rockefeller
Experimental Tract. University of Kansas Pubis., Mus. Nat. Hist., Lawrence, 202 pp.

Gertsch, W. J. and N. 1. Platnick. 1980. A revision of the American spiders of the family Atypidae
(Araneae, Mygalomorphae). American Museum Novitates, 2704:1-39.

McCook, H. C. 1888. Nesting habits of the American purseweb spider. Proc. Acad. Nat. Sci.
Philadelphia, pp. 203-20.

Muma, M. H. and K. E. Muma. 1945. Biological notes on Atypus bicolor Lucas (Arachnida).
Entomol. News, 56(5): 122-26.

Poteat, W. L. 1890. A tube-building spider. J. Elisha Mitchell Sci. Soc., 6(2): 134-47.

Willard Morrow, Department of Entomology, University of Kansas, Lawrence,
Kansas 66045 (Present address: University of Pennsylvania School of Veterinary
Medicine, 3800 Spruce Street, Philadelphia, Pennsylvania 19104).

Manuscript received October 1984, revised December 1984.

VERTICAL COPULATION CAN OCCUR IN LYCOSA
MALITIOSA TULLGREN (ARANEAE, LYCOSIDAE)

During studies on the sexual behavior of Lycosa malitiosa Tullgren we observed
several copulations occurring on the cage wall (Fig. 1). It has been assumed that
lycosid copulation (position II, Gerhardt, U. 1924. Arch. Naturgesch., 90:85-192)
always occurs in the horizontal plane. This view may result from the general use
of glass or plastic cages, which prevent lycosid spiders from climbing.

Lycosa malitiosa is a large, terrestrial species. It is often collected among stones
in sandy areas with short grass (Capocasale, R. M. y F. G. Costa, 1975. Vie et
Milieu, 25(1): 1-15). They have also been found in copula under stones (Costa, F.
G. and F. Perez-Miles, occas. obs.). Copulation is lengthy: 288 palpal insertions
during 99 min, at 21° C in laboratory conditions (Costa, F G. 1979. Rev. Brasil.
Biol., 39(2):361-376).

Spiders were separately housed in cylindrical cages with metallic mesh walls (12
cm diameter x 12 cm height) that were closed above and below with plastic dishes.
These cages are useful because they extend the locomotory freedom of lycosid
spiders, and they also improve the control of environmental parameters by external
equipment. Each male was stimulated with sexual pheromone (female silk threads)
and then introduced to the female’s cage. The male’s courtship — foreleg jerking
and palpal drumming, described elsewhere (Costa, F G. 1975. Rev. Brasil. Biol.,
35(3):359-368) — was generally seen while the spiders were on the cage wall.

Twenty-four copulations of virgin spiders occurred: 13 on the floor (horizontal
copulations), 10 on the wall (vertical copulations), and one that began vertically
(first 15 min) and, after descending to the floor, was completed horizontally, all
without any difficulties. Within vertical copulations, 6 occurred with the female

1986. The Journal of Arachnology 14:122

Fig. I. — Side view of a vertical copulation of Lycosa malitiosa (male facing down and female facing
up). To take this picture, the cylindrical cage wall was sectioned (Photography; R. M. Capocasale).

face down (and male face up), 3 with the female face up, and one with the
longitudinal axis of the couple parallel to the floor. No qualitative or quantitative
differences were found within vertical copulations, and none between vertical and
horizontal copulations (mean durations: 76 min and 75 min, respectively, at 27° C).

The couple’s orientation showed no influence of gravity on mating embrace or
copulatory pattern in L. malitiosa. The couple’s maintenance of a vertical
orientation may be attributed mainly to the female’s ability to hang on the wall,
since only legs IV of the male grasped the cage wall (Fig. 1). These results do
not agree with a view of a passive role (“cataleptic state”) for lycosid females in
copula.

We thank R. M. Capocasale and F. Perez-Miles for the comments, and A.
Postiglioni and J. S. Rovner for the English version of the manuscript. This work
was supported by “Programa Regional de Desarrollo Cientifico y Tecnologico de
O.E.A.”

Fernando G. Costa and J. Roberto Sotelo, Jr., Divisiones Zoologia Experimental
y Biofisica, Instituto de Investigaciones Biologicas Clemente Estable, Avenida Italia
3318, Montevideo, Uruguay.

Manuscript received October 1984, revised May 1985.

1986. The Journal of Arachnology 14:123

CHROMOSOMES OF LEIOBUNUM JAPONICUM JAPONICUM
AND LEIOBUNUM PAESSLERI (ARACHNID A, OPILIONES)

The chromosomes of sixteen harvestmen of the genus Leiohunum, have been
examined (Sokolow 1930, Parthasarathy and Goodnight 1958, Suzuki 1941, 1957,
1976a,b, Tsurusaki 1985). The early investigators used paraffin section and/or
squash methods and usually were unable to make precise karyological analyses
or to discriminate sex chromosomes. Recently air-drying techniques have become
available and Tsurusaki (1982, 1985) has used them on several harvestmen with
good results. In this note, we describe karyotypes of two species of Leiobunum
that were successfully analyzed by the same method.

Leiobunum japonicum japonicum Muller: 2n (5, 9) = 20, NF = 40 (Fig. 1).
Two juvenile specimens, one of each sex from Maruyama in Sapporo, Hokkaido,
Japan were dissected on 16 July 1982. Six adult males from Mt. Kiyosumi, Chiba
prefecture, Honshu were also prepared on 29 August 1984. The chromosome
analysis was based on examination of 35 spermatogonial metaphase plates. There
was one pair of heteromorphic chromosomes that consisted of a submetacentric
similar in size to chromosomes 1-4 and a small metacentric similar in size to
chromosome 8. We presume that this heteromorphic pair constitutes the sex

LO

}( KII t( K

9I (I 9) It Cl

X Y

n {< »« M If

I

2 » • %» f I » I ?•

Figs. 1-2. — Male karyotypes of two species of Leiohunum: 1, L. japonicum japonicum (2n - 20);
2, L. paessleri (2n = 22).

1986. The Journal of Arachnology 14:124

chromosomes, and that the larger member of the pair is the X and the smaller
the Y chromosome. The autosomes consisted of 9 pairs of metacentric
chromosomes. Numerous first meiotic metaphases from adult males invariably
showed 10 bivalents. No differences were detected between the males from the
two locations. Unfortunately, no mitotic metaphase plates good enough for
karyological analysis were obtained from the female. However, 20 chromosomes
could be clearly counted from one ovarian follicle cell. The diploid chromosome
number of 2n = 20 was previously reported for the male of this species by Suzuki
(1941).

Leiobunum paessleri Roewer: 2N (5) = 22. NF = 44 (Fig. 2). Adult specimens
were collected near Kuskonook, British Columbia, Canada on 19 October 1984
and stored at about 5°C until needed. One male was dissected on 22 February
and three more on 14 May 1984. Based on an examination of 13 well spread
spermatogonial metaphase plates, the diploid chromosome number was determined
to be 22. The presumed X and Y chromosomes were, respectively, a large
submetacentric, about the size of the largest autosomal chromosome, and the
smallest metacentric. The autosomes consisted of 10 pairs of meta- or
submetacentrics. Numerous first meiotic metaphases showed 1 1 bivalents without
exception.

Sharma and Dutta (1959) first reported the existence of heteromorphic
chromosomes in opilionids. Since then, Tsurusaki (1982, 1985, unpubl.) has
distinguished X and Y chromosomes in Paraumbogrella huzitai as well as six species
of the Leiobunum curvipalpe-gvonp. Similarly, in males of the two species reported
here, we distinguished a pair of heteromorphic chromosomes that probably
correspond to the X and Y. These results further support the prediction of Tsurusaki
(1982) that an XX-XY mechanism is the usual mode of sex-determination in
opilionids.

All Leiobunum species thus far studied have diploid chromosome numbers
between 16 and 26 (Sokolow 1930, Parthasarathy and Goodnight 1958, Suzuki
1941, 1957, 1976a, b, Tsurusaki 1985, unpubl.). In the two species considered here,
where 2n = 20 and 22, all the chromosomes were invariably found to be meta-
or submetacentric, with no acrocentrics or telocentrics. This stability in chromosome
structure has also been observed in almost all species of the curvipalpe-group
(Tsurusaki 1985, unpubl.), in which the diploid number fluctuates widely from 18
to 26.

We thank Dr. Ramesh Bhambhani of the University of Alberta and Mr. James
C. Cokendolpher of Texas Tech University for their valuable comments on the
manuscript. N. T. also thanks Prof. Mayumi Yamada, Hokkaido University, for
his constant encouragement.

LITERATURE CITED

Parthasarathy, M. D. and C. J. Goodnight. 1958. The chromosomal patterns of some Opiliones
(Arachnida). Trans. Amer. Microsc. Soc., 77:353-364.

Sharma, G. P. and G. P. Dutta. 1959. On the male heterogamety in Melanopa unicolor Roewer
(Opiliones-Arachnida). Res. Bull. (N.S.) Punjab Univ., 10:209-213.

Sokolow, I. 1930. Untersuchungen fiber die Spermatogenese bei den Arachniden. IV. Uber die
Spermatogenese der Phalangiiden (Opiliones). Z. Zellforsch. mikrosk. Anat., 10:164-194, T. 6-8.
Suzuki, S. 1941. On the chromosomes of some opilionids. Zool. Mag. (Tokyo), 53:101. (In Japanese).
Suzuki, S. 1957. On the three closely related forms of the genus Liohunum (Phalangiidae, Opiliones).
J. Fac. Sci. Hokkaido Univ., Ser. VI, Zool., 13:109-1 17.

1986. The Journal of Arachnology 14:125

Suzuki, S. 1976a. Cytotaxonomy in some species of the genus Leiohunum (Opiliones, Arachnida). Proc.
Japan Acad., 52:134-136.

Suzuki, S. 1976b. The genus Leiohunum C. L. Koch of Japan and adjacent countries (Leiobunidae,
Opiliones, Arachnida). J. Sci. Hiroshima Univ., Ser. B, Div. 1, 26:187-260.

Tsurusaki, N. 1982. Chromosomes of the Japanese gagrellid, Paraumbogrella huzitai Suzuki
(Gagrellidae, Opiliones, Arachnida). Bull. British Arachnol. Soc., 5:397-398.

Tsurusaki, N. 1985. Taxonomic revision of the Leiohunum curvipalpe-gvoup (Arachnida, Opiliones,
Phalangiidae). I. hikocola-, hiasai-, kohyai-, and platypenis- subgroups. J. Fac. Sci., Hokkaido Univ.,
Ser. VI, Zook, 24:1-42.

Nobuo Tsurusaki, Zoological Institute, Faculty of Science, Hokkaido University,
Sapporo 060, Japan and Robert G. Holmberg, Sciences, Athabasca University,
Athabasca, Alberta, TOG 2R0, Canada.

Manuscript received November 1984, revised June 1985.

ONTOGENETIC CHANGES IN THE WEB OF
EPEIROTYPUS SP. (ARANEAE, THERIDIOSOMATIDAE)

A recent summary of information on changes in spider web design during
ontogeny showed that when ancestral as opposed to derived designs can be
distinguished, adult webs are usually more derived than those of juveniles
(Eberhard, W. G. 1985. Psyche 92:105-117). This note on an undescribed species
in the theridiosomatid spider genus Epeirotypus demonstrates another possible
example of this pattern of change, and documents further differences between the
webs of young and old spiders.

Adult spiders in the genus Epeirotypus construct orb webs with a spring line
that is out of the plane of the orb, and the spider reels the line in as it sits at
the hub, thus pulling the web into a cone (J. Coddington pers. comm., pers.

obs.). In some species the web is built near a more or less vertical object such

as a tree trunk or a rock, and the spider reels in the entire spring line so that
its body is very close to or perhaps in some cases actually in contact with the

substrate as it waits at the hub (Fig. 1). These webs may be derived with respect

to those in which the spider is not next to a substrate as it holds the web tight
(other Epeirotypus, Theridiosoma, and Ogulnius, Wendilgarda galapagensis — see
McCook, H. C. 1889. American Spiders and their Spinningwork. I. Webs and
Nests, published by the author, Philadelphia; Coddington, J. in press, in Spider
webs and Spider Behavior [W. Shear, ed.], Stanford Univ. Press, Palo Alto; pers.
obs.), though other relationships are also feasible (J. Coddington, pers. comm.).
Sudden sounds nearby usually cause Epeirotypus spiders to release the reeled-up
spring line, thus making the web (and themselves) “snap” backward as in
Theridiosoma (McCook, H. C. 1889). This response presumably aids in prey
capture and defense against predators.

Sites and methods. — Epeirotypus sp. was studied 10-14 Feb., 1985 on mossy
tree trunks just inside a long narrow pasture (estimated 300 x 50 m) in the midst
of forest about 0.5 km down the road from the Tropical Science Center research

1986. The Journal of Arachnology 14:126

Fig. 1. — Large Epeirotypus sp. (probably fifth
instar) at hub of web. The spider (arrow) has
reeled in the entire spring line, and is touching or
nearly touching the moss to which the line is
attached.

Fig. 2. — Small Epeirotypus sp. (probably first
instar) at hub of web. Spider has reeled in only
a small fraction of the spring line.

Station in Monteverde, Costa Rica (elevation about 1200 m). The planes of most
orbs were more or less vertical, but angles were not measured. Webs were gently
coated with cornstarch before being measured with a ruler held near the web. The
angle between the web plane and the spring line was estimated visually. The
number of loops of sticky spiral was the average of the numbers of loops directly
above and directly below the hub, and the average space between sticky spiral
lines was calculated by dividing the number of loops by the distance from the
innermost to the outermost loop of sticky spiral directly above the hub and
directly below it. Spiders often replaced part or all of the spring line after the
web was powdered, then reeled in the new line and tensed the web again;
otherwise they did not usually alter their webs while I measured them. This

1986. The Journal of Arachnology 14:127

■D

(/)

10 15 20

Number of Radii

Fig. 3. — Number of radii versus average number of loops of sticky spiral in webs of small (dots),
medium (circles), and large (x) Epeirotypus sp. spiders.

species apparently passes through four or five instars before reaching maturity
(unpub. analysis of the distribution of sizes), but in this study spiders were
classified as ’’small” (av. max. carapace width 0.45 ± s.d. 0.05 mm — probably
corresponding to instars 1 and 2), “medium” (carapace width 0.67 + 0.03 m —
probably instar 3), and “large” (carapace length 0.77 ± 0.24 mm — probably
instars 4 and 5). No other spiders with similar webs (orbs with spring lines and
hub loops) were found on these trees, so the species identity of even the younger
instars is relatively certain. Voucher specimens of adults are deposited in the
Museum of Comparative Zoology, Cambridge, Massachusetts, and the U. S
National Museum in Washington, D.C.

Results. — The distance (measured along the spring line) from the spider to the
substrate while the spider held the web tensed varied with spider size. Averages
were 0.2 cm for large (N = 21), 0.4 cm for medium (N = 12), and 1.6 cm for
small spiders (N = 21); small spiders rested significantly farther from the substrate
than did large spiders (p < 0.01 with Mann- Whitney U Test).

1986. The Journal of Arachnology 14:128

The angle between the web plane and the spring line apparently increased with
spider size. Average estimated angles for small, medium and large spiders were
44, 49, and 65 respectively; the difference between small and large were significant
(p < 0.01 with G Test when angles were grouped in categories of 20-39, 40-60,
and 70-90, and with Mann Whitney U Test). One small spider had an orb without
a spring line.

Webs of small spiders had fewer radii and sticky spiral loops, and their sticky
spiral lines were closer together than those of larger spiders (p < 0.01 for all three
with Mann-Whitney U Test; average distances between loops of sticky spirals
were 0.16, 0.21, and 0.23 cm for small, medium and large spiders). As shown in
Fig. 3, the relationship between numbers of radii and sticky spiral loops in small
and large spiders’ webs did not change (analysis of covariance showed that the
relationship was not significantly different in the two groups).

There was no difference in the degree of above vertical asymmetry (above hub
vs. below hub) in the webs of large and small spiders. In both groups the average
space between sticky spiral loops was usually slightly larger below the hub than
above (totals 26 larger vs 14 smaller, p < 0.05 with Chi Squared Test), and
usually there were fewer loops below the hub (total 9 webs) than above (24 webs)
(p < 0.01 with Chi Squared Test; 21 webs had equal numbers of loops above
and below the hub).

Discussion. — One of the differences between the webs of small (younger) and
large (older) spiders is that the distance between smaller spiders and the substrate
measured along the spring line as the spider rests at the hub is larger — a character
that is probably less derived (more nearly like that of the probable ancestor of
this group) than is that of large spiders. Another difference, the angle the spring
line makes with the web plane, may follow this same pattern if spring lines were
derived from radial lines (spring lines and other lines running out of the orb
plane from the hub are laid as part of radius construction by some
theridiosomatids and anapids — Eberhard, W. G. 1982. EvoL 36:1067-1095;
Coddington, J. in press; pers. obs.). The polarity (primitive vs derived) of the
other web characters mentioned is not known. If further studies continue to
follow the trend for juvenile characters to be less derived, ontogenetic changes
in web design may be useful in deducing the direction of evolution in orb web
design in this and other groups.

I thank Jonathan Coddington for identifying the spiders and reading a
preliminary version of the manuscript, and the Vicerrectoria de Investigacion of
the Universidad de Costa Rica for financial support.

William G. Eberhard, Smithsonian Tropical Research Institute, and Escuela de
Biologia, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica.

Manuscript received July 1985, revised September 1985.

1986. The Journal of Arachnology 14:129

POSSIBLE PARTHENOGENESIS IN THE HUNTSMAN SPIDER
ISOPOD A INSIGNIS (ARANEAE, SPARASSIDAE)

On the 7th of August at Lake Bonney on the River Murray an immature
female Isopoda insignis (Thorell) huntsman spider (South Australian Museum
voucher specimen N1985143) was found beneath a piece of bark on a dead
Eucalyptus tree. It was caught, returned to Adelaide, and kept in a large plastic
container, being fed occasionally with a large blowfly. In October the spider
underwent its final moult to become a sexually mature female. During its
captivity the spider was kept in complete isolation from all other spiders. During
the early morning on the 26th of December a single eggsac was laid on the
bottom of the container. It was white, lenticular, 35mm in diameter, 5 mm in
height and attached by a few lateral non-viscid threads to the container bottom.
During incubation the mother ceased feeding and guarded the eggsac with her
body. The spiderlings hatched before dawn on the 21st of January. One hundred
and twelve spiderlings emerged from a large rough hole in the dorsal surface of
the eggsac. It was not possible to observe if the mother perforated the eggsac for
the young, as occurs in Delena cancerides Walckenaer (Windsor L. 1972.
Victorian Nat. 89:355-366, Coleman E. 1941. Victorian Nat. 58:88-90).
Examination of the eggsac revealed eight light yellow eggs which had become
crumpled through dehydration amongst the exuviae of the newly hatched young.

Sperm retention has been recorded by a female Paraplectanoides crassipes
Keyserling for over five years (Hickman V. V. 1975. Bull. British Arachnol. Soc.
3:166-173) with a 91% fertility rate. All other recorded cases of sperm retention
have been for eighty days or less, with fertility rates of less than 20% (Valerio
C. E. 1970. Bull. British Arachnol. Soc. 1:28). It is unusual for spiders with short
life spans to moult after mating, but huntsmen spiders may live several years
(Hickman V. V. 1967. Tasmanian Museum and Art Gallery, Hobart). However,
sperm retention requires a fully developed reproductive system, in particular
epigynal openings, not present in this specimen prior to its penultimate moult.
Females of this species have been observed only to become sexually receptive
after their final moult (Clyne D. 1971. Victorian Nat. 88:244-248, McKeown K.
1952. pp. 85-86 Angus and Robertson). It is therefore unlikely that the young
resulted from a mating by the mother, prior to capture.

The only other possible explanation is parthenogenesis. Parthenogenesis has
been known to occur in spiders for many years (Savory T. H. 1928. Sidgwick and
Jackson, London pp. 246-247) but is rare and usually results in a low fertility
rate. In this instance the fertility rate was 93%.

David C. Lake, South Australia Museum, North Terrace Adelaide, South
Australia 5000.

Manuscript received June 1985, revised October 1985.

1986. The Journal of Arachnology 14:130

WEB STRUCTURE AND BURROW LOCATION OF
SPHODROS NIGER (HENTZ) (ARANEAE, ATYPID AE)

The purse-web spider Sphodros niger (Hentz) has long been regarded as one
of the rarer North American mygalomorphs. In their recent revision, Gertsch and
Platnick (1980 Amer. Museum Novitates, 2704:1-39) list only 47 specimens
examined by them, though others doubtless are to be found scattered in various
collections.

The spider’s apparent rarity may be due principally to the fact that its habitat
is unknown. Whereas other atypids, such as Sphodros rufipes (Latreille) and
Sphodros abbotii Walckenaer build tubular webs attched to the bases of trees,
S. niger is not known to do so, and its burrow has rarely been found. Of the
47 specimens listed by Gertsch and Platnick (op. cit.) only 6 were females. The
males were captured mostly in pitfalls, or when they were seen wandering about
on the ground. Of the six females, one was taken from the stomach of a frog,
no information on web or burrow is given for four others, and one was taken
from a “tube in leaf mould.” The latter appears to be the only existing clue to
the location of this species’ burrows.

Recently it has been suspected (Gertsch and Platnick, op. cit.) that S. niger lays
the above-ground portion of its tube flat on or near the ground as the European
species of Atypus and the American Atypus snetsingeri Sarno are known or
believed to do. This suspicion has been confirmed by my discovery in 1981 of
the tubes and burrows of S. niger on Gibraltar Island, Lake Erie, Ottawa Co.,
Ohio.

The first web found, that of a tiny juvenile, was standing erect and attached
to a grass blade in the lawn near the northeast end of the island, at the site
shown in Figure 1. This tube was about 20mm high and 1.5mm in diameter.
Further search of the area disclosed 15-20 more tubes, all but one or two of
which lay flat on the ground, concealed by the grass. All of these tubes occurred
in an area measuring about 5 by 20 feet.

As is usual with atypids, the tubes were partially covered externally with
particles of soil and local debris, bits of grass blades in this case. There were few
very small tubes, no more than 2-3 were found in any one summer. The size
distribution of the tubes was otherwise made up of about equal numbers of tubes
of each size class. Almost all were Judged as being made by individuals ranging
from about half-grown to adult.

During the next tv/o summers (I have not visited the island more recently)
additional specimens were found at the same spot, though few were collected.
Searching in other parts of the island has revealed a few tubes in widely scattered
areas, but no more concentrations such as occur at the original site. One of these
other areas is shown in Figure 2, and a large, but somewhat worn (perhaps
abandoned?) tube found there is illustrated in Figure 3. This web was left
undisturbed. The next year no tube could be found in this spot.

Considering the habitats of S. abbotii and S. rufipes, the placement of the N.
niger burrows is surprising. The soil on Gibraltar Island is, in most places, hard
and rocky, and often quite dry. The subterranean portions of the S. niger tubes
sometimes twist tortuously around and between buried rocks. In very dry weather
the “aerial” portions of the tubes are often completely packed with soil.

1986. The Journal of Arachnology 14:131

Figs. 1-3. — 1. Location of original find and only dense colony of S. niger on Gibraltar Island, Lake
Erie, Ottawa Co., Ohio. The limits of the colony lie between the edge of the rocks and arrow tips.
(Photo by courtesy of Craig Holman, and the Columbus Dispatch.) 2. Another location where a few
burrows were found. 3. A large tube found at the location shown in Fig. 2. Arrows indicate ends
of the somewhat damaged tube.

presumably by the spider. One adult female was taken from the burrow beneath
the only one of these filled tubes I have excavated.

The exposed sites of the burrows may not be the preferred habitat of these
spiders. In the wooded sections of Gibraltar Island the soil is thin, more rocks
are present, and the bedrock is sometimes quite close to the surface. The open

1986. The Journal of Arachnology 14:132

areas may be the only places where the spiders can burrow successfully to a
sufficient depth.

Up to now the number of collected specimens of S. niger from Gibraltar Island
is only five: a male taken on the ground surface in 1978 by the late Michael
Glorioso, and an immature and three females all taken by me at the site of the
initial discovery of the webs. The total number of webs observed from 1981-1983
has been about 40-50, however.

It is unclear from the discussion by Gertsch and Platnick just how many
Sphodros species do extend their webs up tree trunks. Only two, S. abbotii and
-S’, rufipes, are unequivocally indicated as doing so, Coyle (personal comm.) states
that Sphodros atianticus Gertsch and Platnick and Sphodros coyiei Gertsch and
Platnick build vertical tubes attached to tree stumps or grasses, and that S.
atianticus has also been found in semi-horizontal tubes. Sphodros fitchi Gertsch
and Platnick has apparently been found in a burrow only once (Fitch H.S. 1963.
Univ. Kansas Mus. Nat. Hist., Misc. Pub., 33:1-202). The description of the tube
was brief, but implied that the tube was vertical. In southern Illinois, where N.
atianticus, S. niger and S. rufipes have all been collected (rarely), the only webs
found on tree trunks have either been empty or have contained S. rufipes.

I wish to thank Craig Holman and the Columbus (Ohio) Dispatch for
permission to use the photograph designated Figure 1. Mr. David Thrush kindly
made the other photographs for me. Michael Glorioso found the male S. niger,
recognized it, and presented it to me when I arrived on the island that summer.
Dr. Fred Coyle provided information on N. atianticus and S. coyiei habits. I am
grateful to all for their assistance.

Joseph A. Beatty, Department of Zoology, Southern Illinois University,
Carbondale, Illinois 62901.

Manuscript received July 1985, revised September 1985.

1986. The Journal of Arachnology 14:133

PRE-COURTSHIP COHABITATION OF MATURE MALE
AND PENULTIMATE FEMALE GEOLYCOSA TURRICOLA
(ARANEAE, LYCOSIDAE)

Pre-courtship cohabitation of mature male(s) and immature female spiders is
well known in some groups. The webs of subadult female araneids of several
genera (e.g., Nephila, Argiope) may contain many “suitor” males (Robinson and
Robinson 1980, Christenson and Goist 1979) and the immature females of some
wandering spiders (e.g., Salticidae, Clubionidae) tolerate mature males (Pollard
and Jackson 1982, Jackson 1977). However, no such phenomenon is known for
the family Lycosidae. Here we report on pre-courtship cohabitation of subadult
female and mature male Geolycosa turricola (Treat) and comment on the
significance of this observation.

As part of a study of the male courtship behavior of G. turricola, we made
evening observations of a number of burrows in a population near Starkville,
Mississippi. Courtship activities of Geolycosa are known to occur near dusk
(Wallace 1942). On the evening of 13 August 1984 we observed six burrows that
contained mature male spiders, each positioned face down at the burrow entrance
with the abdomen clearly visible from the surface. Each male was captured, and
the burrow was excavated. A penultimate female was found in each burrow. The
male-female pairs were held in the laboratory (females in artificially constructed
burrows, males separately in wire cages) until the female molted to maturity
between 5 and 6 days later (successfully in each case). Each pair was mated
successfully.

Cohabitation in spiders may serve to habituate females to the presence of the
male, thereby dampening her predatory instincts (Robinson 1982), or provide
males with the advantage of proximity, thereby increasing the chances of a
successful courtship (Christenson and Goist 1979). However, these advantages are
available to the male only if the female is stationary. Thus the cohabitation
phenomenon is thought to be associated primarily with web-building spiders or
wandering spiders that construct “nests” for reproduction (Robinson 1982).
Although the Lycosidae are generally considered vagabond spiders, the Geolycosa
have adopted a sedentary life style as obligate burrowers. Immature males and
females and mature females leave their burrows only for prey capture [e.g., we
found that the average prey capture distance of Geolycosa hubbeli (Wallace) in
Florida was less than 10 cm.]. Mature males abandon their burrows in search of
females. The advantages of cohabitation for male Geolycosa are probably
analagous to those of other spiders that exhibit the phenomenon.

Pre-courtship cohabitation has not been reported in the North American web-
building lycosids {Sosippus spp.). Brady (pers. comm.) notes that, on the several
occasions when he has observed mature males in the webs of females, the latter
have also been mature.

The mechanism by which male Geolycosa locate immature females is unknown.
Draglines deposited by receptive female wolf spiders are thought to aid the
searching male by effectively increasing the area of detection around the female
(Tietjen 1977). Contact sex pheromones in the silk and tactile stimulation
provided to the male by the dragline are both apparently important in this

1986. The Journal of Arachnology 14:134

process (Dondale and Hegdekar 1973, Tietjen 1977). Our observations (laboratory
and field) indicate that draglines probably are not as important in mate-finding
in Geolycosa. Only males wander any great distance from the burrow, and
although females deposit silk outside of the burrow, it is confined to a small area
containing the material that forms the turret. This is not to say, however, that
chemical cues associated with silk are not important in close range mate-finding
or other courtship behavior among Geolycosa. We have observed mature male
G. turricola perform preliminary courtship displays to the turrets (no burrow or
female) of a conspecific mature female.

We have not investigated the role of aerial sex-attractants or substrate and
airborne sounds in mate finding in Geolycosa. Airborne pheromones are known
to be important in some lycosids (e.g. Schizocosa saltatrix (Hentz) and S. ocreata
(Hentz), Tietjen and Rovner 1982). One is tempted to hypothesize that the turret
could serve as a broadcasting location for airborne pheromones. With regard to
sound production, we have observed a variety of palpal movements during
courtship, some of which may result in vibrations of the substrate (Miller and
Miller in prep.). Whether the immature female actively advertises her impending
readiness by producing vibrations of some sort is, however, unknown. Whether
the male produces a different signal prior to cohabitation than he does prior to
copulation also must be determined.

We thank G. Stratton, A. Brady, J. Rovner, and M. Robinson for their
comments on this manuscript. Voucher specimens are deposited in the Mississippi
Entomological Museum.

LITERATURE CITED

Christenson, T. E. and K. C. Goist. 1979. Costs and benefits of male-male competition in the orb
weaving spider, Nephila clavipes. Behav. Ecol. SciobioL, 5:87-92.

Dondale, C. D. and B. M Hegdekar. 1973. The contact sex pheromone of Pardosa lapidicina
Emerton. Canadian J. Zool., 51:400-401.

Jackson, R. R. 1977. Courtship versatility in the jumping spider, Phidippus johnsoni (Araneae:
Salticidae). Anim. Behav., 25:953-957.

Pollard, S. D. and R. R. Jackson. 1982. The biology of Clubiona cambridgei (Araneae, Clubionidae):

Intraspecific interactions. New Zealand J. Ecol., 5:44-50.

Robinson, M. H. 1982. Courtship and mating behavior in spiders. Annu. Rev. Entomol., 27:1-20.
Robinson, M. H. and B. Robinson. 1980 Comparative studies of the courtship and mating behavior
of tropical araneid spiders. Pac. Insects Monogr. No. 36. Bishop Museum, Honolulu.

Tietjen, W. J. 1977. Dragline-following by male lycosid spiders. Psyche, 84:165-178.

Tietjen, W. J. and J. S. Rovner. 1982. Chemical communication in lycosids and other species. Pp.
249-279, in Spider Communication: Mechanisms and Ecological Significance (P. N. Witt and J.
S. Rovner, eds.). Princeton University Press, Princeton.

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

Gary L. Miller and Patricia Ramey Miller, Department of Entomology,
Mississippi Entomological Museum, Drawer EM, Mississippi State University,
Mississippi State, MS 39762 (Present address: Department of Zoology, School of
Natural Sciences, Weber State College, Ogden, Utah 84408).

Manuscript received July 1985, revised October 1985.

1986. The Journal of Arachnology 14:135

A NEW SPIDER HOST ASSOCIATION FOR
MANTISPA VIRIDIS (NEUROPTERA, MANTISPIDAE)

Mantisfly larvae of the subfamily Mantispinae are predators within the egg sacs
of spiders (Redborg and MacLeod 1985). Laboratory studies by these workers
demonstrate that larval mantispids enter spider egg sacs by two strategies: 1)
direct penetration of existing egg sacs, or 2) boarding female spiders prior to egg
sac formation and subsequently entering the sac during its construction. The
larvae of some mantispid species are obligate egg sac penetrators, whereas other
larval mantispids are facultative penetrators or boarders (Redborg and MacLeod
1985).

Laboratory experiments with first instar larvae indicate that Mantispa viridis
Walker is an obligate egg sac penetrator (Redborg and MacLeod 1985,
Richardson 1976). Only one of 420 M. viridis larvae boarded spiders during
experimental trials by Redborg and MacLeod (1985), whereas 85% of the larvae
placed in jars containing spider egg sacs were found on or in the sacs 24 h later.
These authors suggested that the one record of a mantispid larva on a spider may
have been fortuitous. Although several first instar larvae of M. viridis boarded
spiders during initial experiments by Richardson (1976), she was unable to repeat
these results during subsequent trials involving several hundred mantispid larvae
and could not explain her initial results. She also reported that M. viridis larvae
penetrated the egg sacs of eight species of spiders in the laboratory.

The purpose of this note is to report the rearing of an adult M. viridis from
the egg sac of a Lycosa pulchra (Keyserling). The spider was carrying the egg
sac in typical lycosid fashion (i.e., attached to its spinnerets) when it was captured
on the ground on 12 March 1982 in Tyler State Park, Smith County, Texas. The
spider and egg sac were transported to Lawrence, Kansas, where the spider died
on 2 April. Upon opening the egg sac two weeks later, I found 95 surviving
spiderlings and a mantispid cocoon. An adult M. viridis emerged on 20 April and
lived until 25 May. These observations corroborate a report of high spiderling
survival within a mantispid-infested egg sac of Lycosa rabida (Rice 1985).

The known range of natural spider egg sac hosts of M. viridis consists of six
species in five families (Table 1). This is the first record of a lycosid spider serving
as a host for M. viridis. Stein (1955) reported the emergence of two M. viridis
adults from egg sacs that he misidentified as being those of a lycosid. Rather,
these egg sacs apparently belonged to a clubionid or gnaphosid spider (Redborg
and MacLeod 1985).

It is unlikely that the first instar larva of M. viridis boarded the L. pulchra
female prior to egg sac construction. The egg sac probably was penetrated after
it was constructed and being carried by the spider. Cupiennius salei, another
known host of M. viridis, also carries its egg sac (Melchers 1963). Also, larvae
of Mantispa styriaca and Mantispa vittata, which are known to feed on the eggs
of lycosid spiders, can penetrate spider egg sacs (Brauer 1869, McKeown and
Mincham 1948). Both of these species, like M. viridis, probably are obligate egg
sac penetrators (Redborg and MacLeod 1985). In contrast, first instar larvae of
Mantispa uhleri naturally board various spiders, including L. pulchra (Redborg
and MacLeod 1985).

1986. The Journal of Arachnology 14:136

Table 1. — Documented spider hosts of Mantispa viridis.

Family

Species

Reference

Theridiidae

Achaearanea tepidariorum

Valerio 1971

Araneidae

Argiope aurantia

Tolbert 1976

Mecynogea lemniscata

Hieber 1984

Agelenidae

Agelenopsis sp., probably pennsylvanica

Parfin 1958

Ctenidae

Cupiennius salei

Milliron 1940

Lycosidae

Lycosa pulchra

This report

I thank Kurt E. Redborg for bibliographic assistance. He and Marlin E. Rice
provided useful comments on the manuscript. Allen R. Brady identified the
Lycosa pulchra, which was collected during a field trip under the direction of
Robert E. Beer of the Department of Entomology, University of Kansas.

LITERATURE CITED

Brauer, E. 1869. Beschreibung der Verwandlungsgeschichte der Mantispa styriaca Poda und
Betrachtungen tiber die sogenannte Hypermetamorphose Fabre’s. Verb. Zool.-Bot. Ges. Wien,
19:831-840.

Hieber, C. S. 1984. Egg predators of the cocoons of the spider Mecynogea lemniscata (Araneae:

Araneidae): Rearing and population data. Florida Entomol., 67:176-178.

McKeown, K. C. and V. H. Mincham. 1948. The biology of an Australian mantispid {Mantispa
vittata Guerin). Australian Zool., 1 1:207-224.

Melchers, M. 1963. Zur Biologic und zum Verhalten von Cupiennius salei (Keyserling), einer
amerikanischen Ctenide. Zool. Jb. Syst., 91:1-90.

Milliron, H. E. 1940. The emergence of a neotropical mantispid from a spider egg sac. Ann. Entomol.
Soc. Amer., 33:357-360.

Parfin, S. 1958. Notes on the bionomics of the Mantispidae (Neuroptera: Planipennia). Entomol.
News, 69:203-207.

Redborg, K. E. and E. G. MacLeod. 1985. The developmental ecology of Mantispa uhleri Banks
(Neuroptera: Mantispidae). Illinois Biol. Monogr., 53. 130 pp.

Rice, M. E. 1985. Spiderling survival in a Mantispa (Neuroptera, Mantispidae) infested egg sac. J.
Arachnol., 13:139-140.

Richardson, M. W. 1976. Micropredators of spiders. Unpubl. Ph. D. thesis. Southern Illinois
University, Carbondale.

Stein, R. J. 1955. An insect masquerader. Nat. Hist., 11:472-473.

Tolbert, W. W. 1976. Population dynamics of the orb weaving spiders Argiope trifasciata and Argiope
aurantia (Araneae: Araneidae): Density changes associated with mortality, natality and migrations.
Unpubl. Ph. D. thesis. The University of Tennessee, Knoxville.

Valerio, C. E. 1971. Parasitismo en huevos de arana Achaearanea tepidariorum (Koch)(Aranea:
Theridiidae) en Costa Rica. Rev. Biol. Trop., 18:99-106.

Steven M. Roble, Museum of Natural History and Department of Systematics
and Ecology, University of Kansas, Lawrence, Kansas 66045. Present address:
Section of Amphibians and Reptiles, Carnegie Museum of Natural History, 4400
Forbes Avenue, Pittsburgh, Pennsylvania 15213.

Manuscript received August 1985, revised November 1985.

1986. The Journal of Arachnology 14:137

USE OF PHEROMONES BY MALES OF
PHIDIPPUS JOHNSONI (ARANEAE, SALTICID AE)

TO DETECT SUBADULT FEMALES THAT ARE ABOUT TO MOLT

Courtship versatility and a complex display repertoire have been described for
Phidippus johnsoni Peckham and Peckham, a common salticid species from
western North America (Jackson 1977). The nests of P. johnsoni are built under
rocks, under the bark of trees, and in similar places with low-intensity ambient
light. Males that encounter adult females away from nests use vision-dependent
displays (Type 1 courtship) similar to those traditionally associated with salticids,
but adult females inside nests elicit vibratory courtship (Type 2) that is
independent of vision. With a subadult (penultimate instar) female inside her nest,
the male first uses Type 2 courtship and then constructs an adjacent silken
chamber, cohabits with the female until she molts and mates when she matures.

Cohabitation durations, which were measured in nature and the laboratory
(Jackson 1978a), tend to be only a few days in duration (mean: 7 days;
maximum: 14 days), although the penultimate instar generally lasts much longer
(mean: 83 days; maximum: 145 days; Jackson 1978b). Both male and female
behavior may influence whether or not cohabitation will occur.

Males of P. johnsoni are known to begin vibratory courtship upon contacting
vacant nests of adult or subadult females (Jackson 1976, 1981), and silk-
associated, chemotactically-detected pheromones are probably involved. Previous
studies did not ascertain the stage (days before molting) at which nests of
subadults become attractive to males. This information was obtained for the
present paper, because male responses to vacant nests can provide an indication
of male sexual interest in subadult females without the subadult’s behavior being
a confounding factor.

MATERIALS AND METHODS

Juvenile P. johnsoni were collected in Berkeley, California in the spring of 1985
and taken to the laboratory in Christchurch. Maintenance, testing procedures,
and terminology were essentially the same as in earlier studies (Jackson 1981),
and only a few remarks will be provided here.

Tests were carried out by introducing males to cages with vacant nests. Vacant
nests were obtained by removing the resident subadult females 2-8 min before
introducing the males. The male’s response was recorded for 30 min after he first
contacted the nest. Only dense nests (4 or 5 on the scale provided elsewhere:
Jackson 1979) were used.

Males and subadult females were taken at random from the laboratory stock
for testing. A strictly random testing procedure, however, would have resulted in
a disproportionate number of tests being carried out on nests of subadults
nearing maturity, although this group was of lesser interest in this study. To
minimize this problem, subadults suspected to be nearly ready to molt (because
of their large abdomens, sedentary behavior, etc.: Jackson 1978b) were removed
from the stock of spiders used in testing.

1986. The Journal of Arachnology 14:138

DAYS

Fig. 1. — Sexual responses by males of Phidippus johnsoni tested with vacant nests of conspecific
subadult females of differing proximity to molting. 2 days: midpoint for 0-4 days; etc. See text for
numbers of tests. Response expressed as percentage of tests at each interval (0-4 days, etc.) in which
nests elicited sexual responses.

Each nest was used in only one test. Each male was tested no more than once
per day. After removal from their nests, subadults were given clean cages and the
number of days elapsing before they molted to maturity was recorded. If they
had not molted within 1 week of the last test, and if they had built a dense nest
in the interval, subadults were used in additional tests.

The performance by the male of any of the following behaviors while on a nest
was recorded as a sexual response: abdomen twitch (abdomen flicks up and
down), probe (legs I move backward and forward on silk), vibrate (with legs on
the nest, body flutters up and down). These behaviors normally occur only during
intraspecific interactions, and they are described in detail elsewhere (Jackson
1977).

RESULTS AND DISCUSSION

The number of days elapsing between the end of tests and molting by females
ranged from 0 to 47 days: 0-4 days, 10 tests; 5-9 days, 10, 10-14 days, 8; 15-19
days, 12; 20-24 days, 6; 25-29 days, 8; 30-34 days, 9; 35-39 days, 5; 40-44 days,
8; 45-49, 4. Nests of subadults more than 29 days from molting never elicited
sexual responses (Fig. 1). Nests of subadults less than 10 days from molting
always elicited sexual responses. The tendency for nests to elicit sexual responses
seemed to gradually increase over the period between 29 and 10 days from
molting.

1986. The Journal of Arachnology 14:139

These data suggest that females begin emitting a sex pheromone only late in
their penultimate instar. Within 10 days of molting, all are emitting pheromones
which make their nests highly attractive to males. For c. 3 weeks before this
either there are fewer females emitting pheromones or pheromone emission by
individual females is at a less effective level. Females apparently fail to make their
nests attractive to males if they are a month or more away from molting.

Peak nest attractiveness and cohabitation initiation apparently coincide: most
cohabitation durations observed in the laboratory and in nature were less than
10 days in duration, and none were more than 14 days. However, one-third to
one-half of nests tested when females were 2-4 weeks away from molting elicited
sexual responses from males. Whether these females would have cohabited is not
known but seems unlikely given the failure to observe cohabitation durations of
greater than 2 weeks in previous studies. Females probably begin to emit
pheromones before they are ready to cohabit. Whether this has any adaptive
significance for the female can not be inferred from exiting data.

In conclusion, cohabitation durations are the results of the combined influence
of male and female behavior as well as female pheromones. There is a limited
period near the end of the female’s penultimate instar during which males are
interested in cohabitation, but this period may be longer than the female is
willing to cohabit.

I thank Joan Buckley for preparing the histogram and Simon Pollard for
comments on the manuscript. Financial assistance was provided by grants from
the University Grants Committee of New Zealand and the Academic Staffing
Committee of the University of Canterbury. The New Zealand Ministry of
Agriculture provided an import permit.

LITERATURE CITED

Jackson, R. R. 1976. Predation as a selection factor in the mating strategy of the jumping spider
Phidippus johnsoni (Salticidae, Araneae). Psyche, 83:243-255.

Jackson, R. R. 1977. An analysis of alternative mating tactics of the jumping spider Phidippus
johnsoni (Araneae, Salticidae). J. Arachnol., 5:185-230.

Jackson, R. R. 1978a. The mating strategy of Phidippus johnsoni (Araneae, Salticidae): I. Pursuit
time and persistence. Behav. Ecol. Sociobiol., 4:123-132.

Jackson, R. R. 1978b. Life history of Phidippus johnsoni (Araneae, Salticidae). J. Arachnol., 6:1-29.
Jackson, R. R. 1979. Nests of Phidippus johnsoni (Araneae, Salticidae): characteristics, pattern of
occupation and function. J. Arachnol., 7:47-58.

Jackson, R. R. 1981. Nest mediated sexual discrimiriation by a jumping spider {Phidippus johnsoni).
J. Arachnol., 9:87-92.

Robert R. Jackson, Department of Zoology, University of Canterbury,
Christchurch 1, New Zealand.

Manuscript received August 1985, revised November 1985.

1986. The Journal of Arachnology 14:140

ON THE MALE OF SPHODROS PAISANO
(ARANEAE, ATYPIDAE)

Eight species (one Atypus and seven Sphodros) of atypid spiders are known
from North America (Gertsch, W. J., and N. 1. Platnick. 1980. Amer. Mus.
Novit., 2704:1-39). Of these, all save one are known from both sexes; Sphodros
paisano Gertsch and Platnick was described from females only. Material recently
acquired from Drs. J. and S. Peck of Carleton University included the first
known males of that species, described here. The format of the description
follows that used in the revision; the illustrations are by Dr. M. U. Shadab.

Sphodros paisano Gertsch and Platnick
Figs. 1, 2

Sphodros paisano Gertsch and Platnick, 1980:20, figs. 20, 30, 31 (female holotype from Rancho El

Milagro, Cruillas, Tamaulipas, Mexico, in Museum of Comparative Zoology, examined).

Diagnosis. — This species differs from all other known Sphodros except S.
abboti Walckenaer in having four-segmented posterior lateral spinnerets; as the
character appears during postembryonic development by subdivision of one of the
three normal segments, it seems to be a synapomorphy linking the two species
(with parallelism in some Atypus). Males of S. paisano can be distinguished from
those of S. abboti by the wider ventral ledge on the palpal conductor (Fig. 2).

Male (Brownsville, Texas). — Total length, including chelicerae, 11.80. Carapace
dark chestnut brown with ocular area darkest. Chelicerae dark brown dorsally,
reddish brown ventrally; sternum, labium, and palpal coxae orangeish brown.
Legs light brown except for orange tarsi and anterior metatarsi. Abdomen
irridescent green to blue dorsally, with long, dark brown scutum covering two-
thirds of dorsum; venter brown except for orangeish brown lung covers and
medially lightened spinnerets.

Carapace 3.74 long, 3.24 wide, irregularly pitted and roughened, with only a
few small setae along anterolateral edges, longest at middle, widest just behind
eyes, smoothly narrowed posteriorly, emarginated above pedicel. Pars cephalica
strongly elevated, set off by two posterolateral depressions, smoothly sloping
posteriorly. Pars thoracica level, thoracic groove deep, subquadrate depression
placed back three-fifths of carapace length, occupying one-eighth of carapace
width at that point.

Eyes on elevated tubercle occupying almost one-third of front width. Ratio of
eyes, anterior lateral: anterior median: posterior lateral: posterior median, 7:7:6:5.
Anterior row slightly wider than posterior, procurved from front, recurved from
above; medians separated by 1.5 times their diameter, by slightly more than their
diameter from anterior laterals. Posterior row recurved; medians separated by 4.5
times their diameter, by one-fifth their diameter from posterior laterals. Median
ocular quadrangle wider than long (16/7), narrowed in front (16/ 1 1).

Sternum 2.27 long, 2.05 wide; labium 0.45 long, 0.75 wide; both clothed evenly
with short, erect, black hairs. Sternum with eight sigilla, smallest pair at base of
labium, other pairs progressively larger, situated at level of posterior edges of
coxae I, II, and III, posterior pair separated by their length, farther from margins

1986. The Journal of Arachnology 14:141

Figs. 1, 2. — Left male palp of Sphodros paisano: 1, prolateral view; 2, retrolateral view.

of Sternum. Median edge of palpal coxae with two rows of long spinules followed
laterally by few irregular rows of small spinules. Chelicerae twice as long as wide
at middle, strongly elevated dorsally for most of length, flat prolaterally, set with
thick, strong setae distally; promargin with nine strong teeth, most proximal
greatly reduced in size; retromargin with thin band of black hairs.

Leg formula 4123. Legs slender, evenly clothed with short, black hairs, with
spines restricted to venter except for few dorsal spines on metatarsi (most
numerous on legs II and III). Tarsi long, flexible, distal three-quarters with
numerous false articulations, ventral surfaces with short, curved spines at sides
(most numerous on legs III and IV). Unpaired tarsal claws small, armed with
single weak tooth (strongest on leg IV); paired claws small, usually with five teeth
in single row. Leg measurements in mm:

I

11

111

IV

Palp

Femur

3.26

2.65

2.27

3.01

2.21

Patella

1.26

1.26

1.17

1.33

0.83

Tibia

1.75

1.45

1.48

2.09

1.37

Metatarsus

2.59

2.23

2.61

3.31

—

Tarsus

2.09

2.12

2.09

2.59

1.52

Total

10.95

9.71

9.62

12.33

5.93

Palp (Figs. 1, 2) with long, curved conductor folded ventrally to produce wide
ledge, bearing long, curved embolus; tibia incrassate, almost as high as long, with
retrolateral clump of spines.

Abdomen 3.73 long, 2.77 wide, clothed with weak, black hairs. Six spinnerets:
anterior laterals slender, 0.38 long; posterior medians separated by two-thirds
their basal width, 0.47 long; four-segmented posterior laterals with lengths as
follows: basal 0.47, median 0.61, subapical 0.48, apical 0.54, total 2.10.

New Record. — UNITED STATES: Texas: Cameron Co.: Sabal Palm Grove, Brownsville, May 31-
Aug. 10, 1983 (S. and J. Peck), 2 males, deposited in the American Museum of Natural History.

Distribution. — Texas and Tamaulipas, Mexico.

Norman I. Platnick, Department of Entomology, American Museum of
Natural History, New York, New York 10024.

Manuscript received and accepted November 1985

1986. The Journal of Arachnology 14:142

UMMIDIA TRAPDOOR SPIDER CAUGHT IN A STEATODA WEB
(ARANEAE: CTENZIDAE, THERIDIIDAE)

The following is an account of a large immature trapdoor spider, Ummidia
(Ctenizidae), being preyed upon by a mature female comb — footed spider,
Steatoda triangulosa (Walckenaer) (Theridiidae) in an underground water meter
housing. This event was observed by the junior author while he was collecting
spiders in Sunset, Montague Co., Texas on 15 March 1985.

The water meter housing was in sandy loam soil, the lid flush with the ground,
in a residential area. Upon removing the lid of the meter housing a typical
theridiid tangle web was observed. The Ummidia was swathed with a large
amount of silk and suspended with the ventral side up in the lower one-third of
the web. The silk was tightly wrapped around the abdomen and legs three and
four. The spider was immobile and appeared dead; however, when grasped with
forceps legs one and two moved slightly, suggesting that it had already been
bitten, but the venom had insufficient time or was incapable of totally paralyzing
a spider of this size. The female Steatoda scurried to shelter near the upper rim
of the housing when the lid was opened and was collected after the Ummidia was
removed.

General statements of theridiid’s ability to capture prey much larger than
themselves are mentioned by Gertsch (1979, American Spiders, Van Nostrand
Reinhold Co., New York 274 pp.) and Bristowe (1971, The World of Spiders,
Collins, London 304 pp.). No reference was found regarding Steatoda feeding on
large prey. In this instance the trapdoor spider was approximately 19 times larger
by weight than the Steatoda. This late instar trapdoor spider weighed 567 mg
(alcohol weight) and was 19 mm long (total length) and 7 mm wide at the
cephalothorax. The Steatoda weighed only 30 mg (alcohol weight), was 6 mm
long (total length), and 2 mm wide at the cephalothorax.

Norman V. Horner, Department of Biology, Midwestern State University,
Wichita Falls, Texas 76308; and Danny Russell, USDA, Soil Conservation
Service, Bowie, Texas 76230.

Manuscript received August 1985, revised October 1985.

1986. The Journal of Arachnology 14:143

HETERONEBO VACHONI AND HETERONEBO MUCHMOREI
ARE SYNONYMS (SCORPIONES, DIPLOCENTRIDAE)

I described Heteronebo vachoni in 1978 (Spec. Publ. Mus., Texas Tech
University, No. 14, 92 pp.) on the basis of material deposited at the Museum
National d’Histoire Naturelle, Paris, and labelled in Prof. Max Vachon’s
handwriting as originating from “Martinique, Sainte Croix, Cotton Valley,”
collected by R. P. Robert Pinchon. Subsequently, Francke and Sissom (1980,
Occas. Pap. Mus., Texas Tech University, No. 65, 19 pp.) described Heteronebo
muchmorei on the basis of material collected in St. Croix, U. S. Virgin islands,
by Dr. W. B. Muchmore. Sometime after the second publication. Dr. Muchmore
informed me that there is a Cotton Valley in St. Croix, U. S. Virgin Islands, and
further research revealed that there is no locality with that name in Martinique.

Direct comparison of the holotype and numerous “paratopotypes” of H.
vachoni against numerous specimens of both sexes of H. muchmorei from St.
Croix revealed that these two nominal taxa are conspecific. The characters given
earlier (Francke and Sissom 1980, loc. cit.), are slightly more variable than
previously known, and reflect populational differences within what I now
consider to be a single taxon endemic to St. Croix, U. S. Virgin Islands. This
species retains the name Heteronebo vachoni Francke, 1978, and its type locality
is hereby modified to Cotton Valley, St. Croix, U. S. Virgin Islands. I suspect
that R. P. Robert Pinchon collected the specimens there, and sent them to Prof.
Vachon from Martinique, thus the confusion on the original label. In the New
World the genus Heteronebo is known with certainty only from the Greater
Antilles and the U. S. Virgin Islands; it is not known from any of the Lesser
Antilles east of the Anegada Passage.

Oscar F. Francke, Department of Biological Sciences, Texas Tech University,
Lubbock, Texas 79409.

Manuscript received September 1985, accepted November 1985.

NONMENCLATURE NOTES

Opinion 1340. — The International Commission on Zoological Nomenclature
ruled under its plenary powers to conserve the name Attus otiosus Hentz, 1846
(Araneae, Salticidae) (Bull. Zool. Nomencl., 42:258, 1985).

On 6 December 1985 the Commission gave six months notice of the possible
use of its plenary powers in the following case:

Z. N. (S.) 2307. — THAIDIDAE Jousseaume, 1888 (Mollusca, Gastropoda) and
THAIDIDAE Lehtinen, 1967 (Arachnida, Araneae): proposals to remove
homonymy.

The Commission welcomes comments and advice from interested zoologists
(Bull. Zool. Nomencl, vol. 42, pt. 4).

THE AMERICAN ARACHNOLOGICAL SOCIETY

President:

Jerome S. Rovner (1985-1987)

Department of Zoology
Ohio University
Athens, Ohio 45701

Membership Secretary:

Norman I. Platnick (appointed)

American Museum of Natural History
Central Park West at 79th Street
New York, New York 10024

Secretary:

Brent D. Opell

Department of Biology
V.P.I.S.U.

Blacksburg, Virginia 24061
Directors:

Frederick A. Coyle (1985-1987), G. B. Edwards (1984-1986), Susan E. Riechert
(1985-1987).

Honorary Members:

P. Bonnett, W. J. Gertsch, H. Homann, fB. J. Kaston, R. F. Lawrence, H. W.
Levi, G. H. Locket, A. F. Millidge, M. Vachon, T. Yaginuma.

The American Arachnological Society was founded in August, 1972, to
promote the study of Arachnida, to achieve closer cooperation between amateur
and professional arachnologists, and to publish The Journal of Arachnology.

Membership in the Society is open to all persons interested in the Arachnida.
Annual dues are $25.00 for regular members, $15.00 for student members.
Correspondence concerning membership in the Society must be addressed to the
Membership Secretary. Members of the Society receive a subscription to The
Journal of Arachnology. In addition, members receive the bi-annual newsletter
of the Society, American Arachnology.

American Arachnology, edited by the Secretary, contains arachnological news
and comments, requests for specimens and hard-to-find literature, information
about arachnology courses and professional meetings, abstracts of papers
presented at the Society’s meetings, address changes and new listings of
subscribers, and many other items intended to keep arachnologists informed
about recent events and developments in arachnology. Contributions for
American Arachnology must be sent directly to the Secretary of the Society.

President-Elect:

William A. Shear (1985-1987)
Department of Biology
Hampden-Sydney College
Hampden-Sydney, Virginia 23943

Treasurer:

Norman V. Horner (1985-1987)
Department of Biology
Midwestern State University
Wichita Falls, Texas 76308

Archivist:

Vincent D. Roth

Box 136

Portal, Arizona 85632

Research Notes

Las citas de Pachylus chilensis para la Argentina (Opiliones,

Gonyleptidae, Pachylinae), Luis Eduardo Acosta 117

A range extension of the purseweb spider Sphodros rufipes
(Latreille) in eastern Kansas (Araneae, Atypidae),

Willard Morrow 119

Vertical copulation can occur in Lycosa malitiosa Tullgren
(Araneae, Lycosidae), Fernando G. Costa and

J. Roberto Sotelo, Jr 121

Chromosomes of Leiobunum japonicum japonicum and
Leiobunum paessleri (Arachnida, Opiliones),

Nobuo Tsurusaki and Robert G. Holmberg 123

Ontogenetic changes in the web of Epeirotypus sp.

(Araneae, Theridiosomatidae), William G. Eberhard 125

Possible parthenogenesis in the huntsman spider

Isopoda insignis (Araneae, Sparassidae), David C. Lake 129

Web structure and burrow location of Spohodros niger

(Hentz) (Araneae, Atypidae), Joseph A. Beatty 130

Pre-courtship cohabitation of mature male and penultimate
Geolycosa turricola (Araneae, Lycosidae), Gary L. Miller

and Patricia Ramey Miller 133

A new spider host association for Mantispa viridis

(Neuroptera, Mantispidae), Steven M. Roble 135

Use of pheromones by males of Phiddipus johnsoni (Araneae,

Salticidae) to detect subadult females that are about to

molt, Robert R. Jackson 137

On the male of Sphodros paisano (Araneae, Atypidae),

Norman /. Platnick 140

Ummidia trapdoor spider caught in a Steatoda web (Araneae:

Ctenizidae Theridiidae), Norman V. Horner

and Danny Russell 142

Heteronebo vachoni and Heteronebo muchmorei are synonyms

(Scorpiones, Diplocentridae), Oscar F Francke 143

Other

Nomenclature Notes

143

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 14 Feature Articles NUMBER 1

The ecological effects of burning, mowing, and plowing on
ground-inhabiting spiders (Araneae) in an old-field

ecosystem, Mary F Haskins and James H. Shaddy 1

Chactidae (Scorpiones) from Trinidad and Tobago,

Oscar F. Francke and Julius Boos 15

Ecology and behaviour in Portia schultzii, with notes
on related species (Araneae, Salticidae),

Lynn M. Forster and Frances M. Murphy 29

Mother-young relationships in Euscorpius: Adaptive value of
the larval permanence on the mother’s back, A. Ugolini,

/. Carmignani, and M. Vannini 43

A revision of the spider genus Sason Simon (Sasoninae,

Barychelidae, Mygalomorphae) and its historical

biogeography, Robert J. Raven 47

Spectral sensitivities of the eyes of the orb web spider
Argiope argentata (Fabricius), Klaus B. Tiedemann,

Dora Fix Ventura, and Cesar Ades 71

Analisis de la actividad diaria de Aphonopelma seemanni

(Araneae, Theraphosidae) en Costa Rica, Marco V. Herrero

and Carlos E. Valerio 79

Redefinition of the genus Olpiolum and description of a

new genus Banksolpium (Pseudoscorpionida, Olpiidae),

William B. Muchmore 83

Mygalomorph spiders in the Barychelidae (Araneae)

from Costa Rica, Carlos E. Valerio .93

Albinism and eye structure in an Australian scorpion,

Urodacus yaschenkoi (Scorpiones, Scorpionidae),

N. A. Locket 101

(continued on back inside cover)

Cover photograph, fluorescence of Sphaleropachylus butleri (Thorell), by J. C. Cokendolpher
Printed by the Texas Tech Press, Lubbock, Texas, U.S.A.

Posted at Lubbock, Texas, in June 1986

The Journal of

ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 14

SUMMER 1986

NUMBER 2

THE JOURNAL OF ARACHNOLOGY

EDITOR: Oscar F. Francke, Texas Tech University

ASSOCIATE EDITOR: William B. Peck, Central Missouri State University
ASSISTANT EDITOR: James C. Cokendolpher, Texas Tech University
EDITORIAL BOARD: C. D. Dondale, Agriculture Canada; W. G. Eberhard,
Universidad de Costa Rica; M. E. Galiano, Museo Argentino de Ciencias
Naturales; W. J. Gertsch, American Museum of Natural History; N. F.

Hadley, Arizona State University; V. F. Lee, California Academy of Sciences;
H. W. Levi, Harvard University; E. A. Maury, Museo Argentino de Ciencias
Naturales; W. B. Muchmore, University of Rochester; M. H. Muma, Western
New Mexico University; N. 1. Platnick, American Museum of Natural
History; G. A. Polis, Vanderbilt University; S. E. Riechert, University of
Tennessee; M. E. Robinson, U.S. National Zoological Park; V. D. Roth,
Southwestern Research Station; J. S. Rovner, Ohio University; W. A. Shear,
Hampden-Sydney College; W. J. Tietjen, Lindenwood College; G. W. Uetz,
University of Cincinnati; C. E. Valerio, Universidad de Costa Rica; S. C.
Williams, San Francisco State University.

THE JOURNAL OE ARACHNOLOGY (ISSN 0161-8202) is published in
Spring, Summer, and Fall by The American Arachnological Society at Texas
Tech Press.

Individual subscriptions, which include membership in the Society, are $25.00
for regular members, $15.00 for student members. Institutional subscriptions to
The Journal are $35.00. Correspondence concerning subscriptions and
memberships should be addressed to the Membership Secretary (see back inside
cover). Back issues of The Journal are available from Dr. Susan E. Riechert,
Department of Zoology, University of Tennessee, Knoxville, TN 37916 U.S. A.,
at $10.00 for each number; the Index to Volumes 1-10 is available for $10.00.
Remittances should be made payable to The American Arachnological Society.
Correspondence concerning undelivered issues should be addressed to the Texas
Tech Press, Texas Tech University, Lubbock, Texas 79409, U.S. A.

Change of address notices must be sent to the Membership Secretary.

Detailed instructions for the preparation of manuscripts appear in the Fall
issue of each year, and can also be obtained from the Editor and the Associate
Editor. Manuscripts that do not follow those instructions will be returned to the
author(s) without benefit of review. Manuscripts and all related correspondence
must be sent to Dr. William B. Peck, Associate Editor, Department of Biology,
Central Missouri State University, Warrensburg, Missouri 64093, U.S. A.

PROOFS, REPRINTS, and CHARGES: Authors will receive a reprint order
form along with their proofs. Reprints are billed at the printer’s current schedule
of costs. All authors will be billed by the printer for page charges. The charge
per journal page will be assessed as follows: $20. 00-nonmembers; $ 15.00-members
acknowledging grant, institutional, or other support; $ 10.00-full members who do
not have (and do not acknowledge) support; $5. 00-student members who do not
coauthor with full members and who do not acknowledge support. Members may
have page charges waived or reduced for a limited number of pages; a statement
of justification for waiver should accompany the manuscript when it is first sent
to the Associate Editor for review.

Tietjen, W. J. 1986. Effects of colony size on web structure and behavior of the social spider M alios
gregalis (Araneae, Dictynidae). J. Arachnol., 14:145-157.

EFFECTS OF COLONY SIZE ON WEB STRUCTURE
AND BEHAVIOR OF THE SOCIAL SPIDER
MALLOS GREGALIS (ARANEAE, DICTYNIDAE)^

William James Tietjen

Department of Biology
Lindenwood College
Saint Charles, Missouri 63301 U.S.A.

ABSTRACT

Groups of size 1, 2, 5, 10 and 20 Mallos gregalis were monitored under laboratory conditions with
the aid of a computer-controlled digital camera. Data collected included a measure of the density and
complexity of the silk comprising the nest, as well as activity levels and occupation of space within
experimental arenas.

Average web density and complexity was related to colony size, with the larger colonies building
more complex nests. I suggest that the greater web complexity would allow larger colonies greater
opportunities for the exploitation of marginal habitats. The webs built by the two smaller groupings
were similar to those built by solitary dictynids and indicated that M. gregalis may be a facultatively-
social spider. An estimate of mean silk deposition per spider indicated that members of the larger
colonies exerted less effort in web construction than spiders in the two smaller groupings.

Colony activity was related to group size and exhibited evidence for a group effect in the patterning
of activity bouts. It is possible that this would aid in coordinating colony behavior. Measures of both
web structure and colony activity indicated that the changes in colony behavior were not due to a
simple arithmetic effect (e.g., size 20 colonies were neither twice as active nor were their webs twice
as complex as colonies of size 10).

INTRODUCTION

Although most spiders are solitary, aggregating only early in life or during
mating, a few species are social (Shear 1970, Kullman 1972, Burgess 1978, 1979a,
Buskirk 1981). Unlike the groups seen among the insect societies, no araneid
associations studied to date show evidence of ethological nor morphological caste
systems (Wilson 1971, Burgess 1979b). For this reason, studies of social spiders
must confront the problem of how such complex groupings are behaviorally
organized. The present laboratory study describes some organizing features of
colony behavior in a social spider, Mallos gregalis, especially the effects of colony
size on the patterning of activity, nest complexity and web construction.

Mallos gregalis is a social species living in colonies that may cover three-
quarters of an 18-m tree with webbing. Up to 20,000 individuals of both sexes
and various stadia may inhabit such nests (Diguet 1909a, 1909b, Burgess 1979b,
Uetz 1983). Predation, feeding and web construction are communal, with little or
no cannibalism occurring among conspecifics (Witt et al., 1978).

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

146

THE JOURNAL OF ARACHNOLOGY

The web is an intricate, sheet-like affair, superficially resembling that of tent
caterpillar larvae {Malacosoma spp). The outer surface, (the primary prey-capture
site), has silk-reinforced tunnel openings to the web interior, silk-reinforced
runways, water-resistant papery areas, loosely-woven cribellate silk sheets, as well
as the remains of their prey, primarily muscid flies. The web interior has
numerous silk-lined tunnels and chambers which gives it a spongy appearance.
Spiders spend most of the day within these chambers, moving to the surface only
when it is disturbed by struggling prey. Within the web M. gregalis courts, mates,
constructs egg sacs and deposits silk to reinforce tunnels and chambers. At the
surface of the nest spiders capture prey, feed communally, eliminate excreta and
deposit silk for web expansion (Tietjen 1986).

MATERIALS AND METHODS

General Methods. — Approximately 350 Mallos gregalis from a colony collected
near Guanajuato, Guanajuato, Mexico by G. W. Uetz were maintained in a large
plexiglass cage (91.0 L X 47.0 W X 35.5 H cm) located near a window to provide
a natural photoperiod. This larger colony provided the “seed” individuals used
for the experimental colonies. Temperature remained fairly constant at 28.8 ±
2.97 sd°C. Water was provided daily by a fine mist sprayed on the web surface;
and animals were fed houseflies {Musca domestica) at weekly intervals.

Ninety one experimental colonies were housed in Petri dishes (52.2 cc) during
the course of the study (8/22/81-1/28/82). Five group sizes were examined (N
= 20, 10, 5, 2 or 1 adult females per container) to determine the effect of group
size on colony activity and web structure. Direct observation and preliminary
analyses of computer image data (see below) indicated that the rate of silk
deposition levels off by day five of nest growth. For this reason, all data were
collected from five-day-old experimental colonies. Previous data (Burgess 1979b,
Jackson 1979, Tietjen 1982) indicate that M. gregalis is nocturnal. Data collection
therefore began between 1500 and 1700 hrs, and experiments were terminated
between 0630 and 0900 EST. Experimental colonies were fed houseflies (M.
domestica) one day prior to their introduction to the arenas but were not fed
during the five-day growth period. Had spiders been fed during this period, web
structures would have been disrupted and measurements of nest complexity
would not have been possible. Water was provided daily by depositing a drop
near the edge of the Petri dish.

Buskirk (1981) indicates that coloniality in spiders may be related to high prey
availability. Thus, the possibility exists that the spiders used in these series of
experiments were stressed. Jackson (1980) maintained M. gregalis under
conditions of starvation for as long as 53 days, and Witt (personal
communication) maintained the spiders for four weeks without apparent ill
effects. In a similar vein, Witt, Scarboro and Peakall (1978) used radioisotope
techniques to determine the number of M. gregalis feeding on a single fly. Their
data indicates that, even after 24 hr, none of their colonies had spiders that had
all fed upon the prey. This minimal feeding level may be related to their low
metabolic rate (88.0 + 8.0 jul O2 /g body weight /hr); a value approximately !4
that of Araneus diadematus (Witt, personal communication). Under field
conditions, M. gregalis is likely to be presented with similar periods of high and
low prey availability. Burgess (1979b) reports that the nests are sometimes

TIETJEN— EFFECTS OF GROUP SIZE ON COLONY BEHAVIOR

147

surrounded by swarms of muscid flies, whereas prey availability at other times
is so low that it can not be accurately measured. His results are consistent with
the reports of Diguet (1909a, b) and Uetz (personal communication) who report
that M. gregalis occupies a xeric habitat with fluctuations in prey availability.
Although periods of high prey availability may have been important contributing
factors in the evolution of sociality in M. gregalis, a predictable and consistently
high abundance of prey is apparently not required for the maintenance of social
behavior in this species under both laboratory and field conditions. For these
reasons the short periods of low prey availability encountered in these series of
experiments are not likely to affect the general conclusions of this paper.

Recording Methods. — Detailed descriptions of recording methodologies are
provided elsewhere (Tietjen 1981, 1982). The system allowed for computer-
assisted scanning of colony activity, occupation of space within the arena, and
measurement of web structure and growth. In brief, a solid-state television
camera (Periphicon Type 511), connected via a suitable interface to a CDP-1802
microprocessor-based computer, was located 0.65m above the experimental
colony; and light was transmitted through the arena. The image in the camera’s
field of view was digitized into a 32 X 32 array providing 1024 picture elements
(= pixels) which could potentially be occupied by the spiders and their web.
Experimental arena shape reduced the usable recording area to about 620 pixels.
Spiders appeared as black dots on a white background with each animal
occupying a single pixel in the field of view.

The computer was programmed to record the occupation of space and activity
of experimental colonies at approximately 30-sec intervals providing a total of
126,134 exposures. Occupation of space within the arena was recorded on a
’’map” of the arena in computer memory for each exposure. The number of
moving spiders was determined by comparing sequential frames and recording the
number of spiders moving from one frame to the next.

By transmitting light through the colony it was also possible to estimate the
relative silk density at each pixel. High silk density areas (such as those associated
with runways, reinforcement threads and silk-lined chambers) occluded more light
than areas of low silk density. Thus, relative silk density and light intensity were
inversely proportional. The relative silk density was recorded in 1024 levels of
density for positions occupied by the colonies within the 32 x 32 grid. Blocks
occupied by spiders or excreta were ignored in the silk density analyses by
marking those pixels with the aid of a light pen. I compensated for possible
nonlinearities in lighting intensity across the field of view by subtracting a optical
density image of a control field from each of the silk density images.

Preliminary tests indicated that data (raw or transformed) did not fit
parametric assumptions (Sokal and Rohlf 1969). All analyses were therefore
nonparametric and were performed according to the methods of Conover (1971).
Even so, all data are presented as the means and their standard deviations.
Fourier analyses were based on the algorithm described by Owens (1981).

RESULTS

Web Structure. — Mallos gregalis readily built webs within the Petri dishes
during the five-day growth period. These webs exhibited many of the structural
features found in natural nests including chambers, silk runways and support

148

THE JOURNAL OF ARACHNOLOGY

Table 1. — Effects of group size on web structure. Measures of web structure are presented in
arbitrary units related to the optical density of the nest (see text).

Colony

Size

Number of

Colonies

Mean Silk
Density

Maximum Web
Density

Silk Deposition

Per Spider

20

18

105.1 ± 13.7

375.0 ± 30.6

5.3 ±0.7

10

17

78.1 ± 15.0

385.5 ± 30.6

7.8 ± 1.5

5

15

92.3 ± 11.9

305.3 ± 28.8

18.5 ± 2.4

2

20

45.3 ± 5.2

291.2 ± 60.6

22.7 ± 4.2

1

21

47.8 ± 4.2

205.0 ± 34.8

47.8 ± 4.2

threads (Tietjen 1985). The webs had their greatest silk density and structural
complexity (i.e., chambers and tunnels) near the perimeter of the arena, with
relatively simple, less dense sheets or silk-free areas found in the central parts of
the arenas.

The mean silk density differed among groups, and was proportional to colony
size (Spearman’s rho p = + 0.35, P<0.01; Kruskal-Wallis Test Test, P<0.001;
Table 1). The relationship between silk density and colony size was non-linear
with the three larger colonies exhibiting equal mean silk densities. Silk density
in these colonies was greater than, and significantly different from the two smaller
groups (Mann-Whitney Test, P<0.005). Visual examination of the webs (Fig. 1)
as well as quantitative indices of variation in silk density within nests (Maximum
web density [Table 1], standard error and variance to mean ratios among blocks)
indicated that the webs of the larger three colonies were more heterogeneous than
the smaller two groups (Spearman’s rho p = + 0.66, P<0.0001; Kruskal-Wallis
Test Test, P<0.001). The high silk density areas were associated with chambers
and tunnels in the web-interior, therefore variation in silk density is an index of
nest structural complexity.

A qualitative estimate of the average silk deposition per spider was calculated
by dividing the mean silk density of each web by the number of spiders in the
colony (Table 1; Fig. 2). These data indicate that the mean silk deposition per
spider is inversely proportional to colony size (Spearman’s rho p = —0.86,
P<0.001; Kruskal-Wallis Test, P<0.001).

Colony Activity. — Colony activity, recorded as the number of animals moving
per minute, was somewhat proportional to group size (Table 2). The relationship
was non-linear, however, with larger colonies exhibiting more activity than
expected, and smaller groupings showing less than the expected activity
(Spearman’s rho p = + 0.69, P<0.0001; Chi-square, P<0.001).

An examination of individual colony activity indicated that activity in the
larger groupings was not constant during the recording period, but rather was
clumped into discrete periods of high and low colony movement (Fig. 3). To
explore the possibility of periodicities in the behavior, the data were analyzed
using a fast Fourier transform and spectrum analysis. A full statistical analysis
of the power spectra (a measure of the relative ’’importance” of each frequency
component) was not possible with the available equipment. For this reason, only
the shortest significant period was recorded from the power spectra and
compared among colonies (Table 2).

Neither of the two smaller groupings exhibited periodicities above the
background noise (unresolved or nonsignificant periods) of the power curve,
indicating the absence of significant short-term activity bouts during the recording

TIETJEN— EFFECTS OF GROUP SIZE ON COLONY BEHAVIOR

149

Pig 1. —Effects of group size on relative silk density and structure. Representative silk density maps
are provided for a) Colony size = 20; B) Colony size = 5 and C) Colony size = 1. Vertical bars
represent relative silk density as determined by light transmission and displayed on a 0 to 10 scale
which was optimized to display low density silk areas. The circles at the top of the bars provide
additional cues as to the relative silk density; the larger circles representing denser areas of silk
deposition (those blocks having chambers and tunnels). Small dots on the X/Y plane are areas located
either outside the arena or blocks within the arena that were occupied by spiders or excreta. These
blocks were ignored in the analyses. Note the silk-free areas in A and B and the extended low-density
silk sheet in C. Also note the greater variation in web structure for the larger two colonies, and the
concentration of silk near the edge of the arena. Experimental colonies of size ten were intermediate
to those of size twenty and five. Groupings of size two were similar to those of size one.

150

THE JOURNAL OF ARACHNOLOGY

Number of Spiders per Colony

X///A Mean Silk Density [XXf Silk Dep./ Spider

Fig. 2. — Effects of colony size on silk deposition. The relative index is an optical measure of silk
deposition by colonies of size one-twenty. The mean silk density is presented as the averaged optical
density of the colonies in arbitrary units. The silk deposition per spider is a qualitative index of the
’’work per spider” in building a nest.

period. Similarly, colony groupings of sizes five and twenty each had one sample
with no significant periodicity while all of the remaining groups showed evidence
of short-term periodicity. Examination of these remaining power spectra indicated
that size-twenty colonies exhibited a shorter period than colonies of size five and
ten (Spearman’s rho p = —0.36, P<0.05; Kruskal-Wallis Test, P<0.025; Mann-
Whitney Test, P<0.01). Colonies of size 20, for example, showed peaks in activity
every 25.8 minutes whereas colonies of size five and ten had activity peaks
separated by over 40 minutes.

The recordings of occupation of space within the arena indicated that, in
general, spiders in all groupings mainly occupied the periphery of the arena, those
areas characterized by the highest silk density and complexity (Fig. 4).
Examination of these data also allowed the determination of two movement
indices: an occupation index expressed as blocks occupied per hour per spider,
and the percent inactivity. The first is an index of the space occupied by each
spider, whereas the second is a measure of the number of runs (for each
experimental group) which exhibited evidence of one (or more) animals showing
no movement during the recording period (Table 2).

The percent of inactive spiders was inversely proportional to colony size
(Spearman’s rho p = —0.82, P<0.0001 Kruskal-Wallis Test, P<0.001). The
occupation index varied directly with colony size (Spearman’s rho p = + 0.50,
P<0.01; Kruskal-Wallis Test, P<0.005). Even among the larger colonies, however,
the occupation index was below three blocks per hour per spider, indicating that
most of the spiders’ activity occurred within small areas of the web rather than
on the entire web surface (Table 2).

TIETJEN— EFFECTS OF GROUP SIZE ON COLONY BEHAVIOR

151

Table 2. — Effects of group size on colony activity and occupation by Mallos gregalis. The
coordination index is a measure of the shortest periodicity as determined by the Fourier analysis of
colony activity. Indices of activity and occupation (blocks occupied per hour) are presented as means
for each spider rather than colony means. Both the coordination index and maximum percent
inactivity are colony means.

Colony

Size

N

Activity
Moves/ min

Coordination
Index (min)

Occupation

Index

Maximum %
Inactivity

20

11

0.35 ± 0.1

25.8 + 4.4

2.0 + 0.7

99.4 + 1.5

10

10

0.33 + 0.2

41.5 + 4.0

2.6 + 0.6

73.7 + 27.5

5

10

0.22 ± 0.1

45.1 + 8.3

1.8 + 0.3

66.7 + 22.2

2

12

0.16 + 0.1

—

1.1 + 0.2

80.9 + 18.6

1

14

0.09 + 0.1

—

0.5 + 0.2

62.3 + 35.8

DISCUSSION

Evaluation of Methodology. — The computer-camera apparatus was non-
intrusive in the recording of animal position and activity and should have no
effect on behavior. The peak sensitivity of the camera (900 nm wavelength)
allowed monitoring using a light source which is outside the visual range of many
animals, including spiders (DeVoe and Zvargulis 1967, DeVoe 1972).

The use of the computer-controlled camera necessitated confining the spiders
within the limited area of a Petri dish. However, Mallos gregalis adapt well to
such conditions and will construct functional webs in a variety of containers
ranging in size from the arenas used in these series of experiments to room-sized
environmental chambers. Nonetheless, animals so restricted exhibit apparently
normal courtship, mating, prey-capture, and feeding behaviors over several
generations (Tietjen 1980, 1986). In addition, by confining the animals to a
relatively small and homogeneous area, intercolony variation for web structure
within a treatment group is decreased, thus allowing for comparisons among
groups using reasonable sample sizes.

The silk density analyses had a bias owing to the necessity of ignoring those
blocks occupied by spiders or excreta. This bias, however, would tend to reduce
web complexity estimates for the larger colonies, since spiders spent most of their
time near the perimeter of the arena (which had the highest silk density). The
error would have a greater effect on larger colonies, thus reducing their indices
of silk density variation among pixels. In a similar manner, the removal of those
blocks occupied by excreta would further reduce overall variability since most
waste is deposited in the central part of the arena, that area utilized as the
exterior of the web and having the lowest silk density estimates (Tietjen 1980).
The two smaller groupings had most of their web surface composed of low-
density silk, so ignoring excreta would tend to increase web variability. Thus, the
bias in both cases tends to make the measures of web complexity and variability
more conservative.

The delicate nature of the web and the availability of equipment made it
impossible to correlate the silk density data with an actual quantity of silk at each
pixel within the field of view. However, a visual comparison of silk density maps
with actual colonies indicated that the silk density image provided an accurate
sampling of high silk density sites, especially those associated with runways and
chambers within the nest. For this reason, the variation in silk density within a
colony is interpreted as a measure of web complexity. Those webs that were

152

THE JOURNAL OF ARACHNOLOGY

O'=»<r>«>c:>«ooo<r>'::>oo<=»*oo>o<=>c?oc:>o<=>ooo-o*o<2»ooo<=>oc:>oc:>o><r>-ooo

<^<a<3<^*OO*C><5C»O<D^C‘-jr-o^UT'0I^C0Cr*C>'— ••>«f^«»’ir>^r-^C0Cr*O’--'*5’'«'^*^tO'-0r-^Q0&*O'^C'-4r0^yT-s0r^C0Cr“0>

TIME (MIN)

Fig. 3. — Representative activity data for a colony of size twenty. The scale on the abscissa is
minutes since the beginning of the experiment. Note the clumping of the spiders’ movements into a
series of high and low activity bouts during the 8.5-hr recording period.

recorded by the computer as having little variation in silk density were
qualitatively more simple in construction than those with relatively greater
variation. The colonies with the greatest index of web complexity built webs with
a very spongy appearance due to the presence of numerous silk-lined chambers
and tunnels.

Web Structure as Related to Group Size. — Group size had a profound effect
on the density and complexity of nest structure which extended beyond an
arithmetic effect. Colonies of size 20, for example, did not simply deposit twice
as much silk as colonies of size 10 since web density was not directly correlated
with group size (the three larger colonies exhibited equal mean silk density).
These data suggest that, for the larger groupings, a minimal amount of silk was
required to support the construction of a nest which was to be occupied by
several spiders. Similar results showing adaptation to group size (e.g. non-
arithmetic changes in the behavior) are seen for web complexity, maximum silk
density and the overall form of the nest.

Nests built by the larger groups had most of the silk concentrated at the edge
of the arena. This outer section was characterized by greater structural
complexity, including the presence of silk-lined chambers, runways and tunnels.
The central portion of the arena had little or no silk, giving the larger nests a

TIETJEN— EFFECTS OF GROUP SIZE ON COLONY BEHAVIOR

153

Fig. 4. — Effects of group size on occupation of space within the arenas. Three representative
groupings are indicated: A) Colony size = 20; B) Colony size = 5 and C) Colony size = 1. As in Fig.
1, the vertical bars represent relative silk density. The circles at the top of the bars now indicate
occupation of blocks within the arena. Those blocks that were most often occupied have the largest
circles. Note the tendency for animals in each grouping to remain near the edge of the arena and
the lack of long-distance movements by colonies of size one. Tests with groups of size ten were
intermediate to those of size twenty and five. Groupings of size two were similar to those of size one.

154

THE JOURNAL OF ARACHNOLOGY

toroidal distribution of webbing. The greater density of silk near the periphery
was largely responsible for the increase in mean silk density with respect to group
size. In contrast, webs built by groupings of one or two spiders were far less
complex in construction, having few silk-lined retreats, and no runways or
tunnels. In addition, a finely spun silk sheet extended through the central portion
of the arena. Such sheets of silk were characteristic only of those webs built by
the smaller groups.

The greater silk density and complexity produced by the three larger groupings
would allow these colonies greater potential for exploitation of marginal habitats.
In the laboratory and field, groups of M. gregalis roll or tie together leaves of
the supporting foliage and construct a prey capture sheet on the surface of the
plant, a process which does not kill the underlying foliage. Within the interior
of these nests, the spiders build chambers and tunnels between the leaves.
Individual spiders do not build such structures under laboratory conditions, but
rather confine their building of a prey capture sheet to the surface of adjacent
leaves. By taking advantage of the structure provided by the supporting foliage,
the larger colonies would be afforded protection from sun, wind and rain, while
allowing M. gregalis to take advantage of leaf transpiration to cool the nest
(Tietjen 1986). In addition, a more dense and complex web would be expected
to provide greater protection from potential vertebrate predators (Buskirk 1981,
Rypstra 1979, Tietjen 1986). The web structure of the smaller test groupings
lacked the complexity seen in the larger groupings, most of the nests built by the
smaller groupings had only one or two silken chambers which could be used as
retreats. These nests are qualitatively similar to webs built by solitary dictynids
(Chamberlin and Gertsch 1953, Kaston 1948). Even among those groupings
having two members, the structure of the web was more similar to that seen in
solitary species than to social ones. This suggests that M. gregalis is capable of
shifting web-building behavior from a communal-cooperative mode of
construction to a solitary mode of behavior.

Under unrestrained conditions in the laboratory, individual M. gregalis may
leave the parent colony and build single-spider webs. Spiders occupying such
webs usually remained isolated from the original colony and were competent in
prey-capture behavior. These data, in conjunction with the quantitative
differences observed in the structure of the nests between the large and small
groupings, suggest that M. gregalis is, to some extent, a facultatively-social
spider.

The silk deposition per spider was inversely proportional to colony size.
Although these data do not represent the actual metabolic expenditures for
individuals, it is reasonable to assume that the differences observed among the
groups do reflect varying physiological burdens. If this is so, the data indicate
that another advantage of cooperative nest construction is a decrease in energy
expenditure per individual spider. A more complete explanation of this effect will
require an analysis of the caloric content of webs built by groups of differing size.

Colony Activity as Related to Group Size. — Measures of mean activity per
spider, patterning of activity bouts and occupation of space within the arena were
also affected by group size. As was the case for web structure, the changes
observed for each of these behaviors extended beyond an arithmetic effect and
instead represent shifts in the behavior of M. gregalis with changes in colony
density.

TIETJEN— EFFECTS OF GROUP SIZE ON COLONY BEHAVIOR

155

The data indicate that the mean activity of individuals in larger colonies was
greater than that of the smaller, with members of the largest two colonies being-
three to four times more active than isolates. This evidence, taken alone, might
indicate that a major disadvantage of group-living in M. gregalis is an increase
in metabolic costs due to greater activity. However, most of the activity observed
in the larger colonies was short-distance movements and turning in place as
shown by the mean number of blocks traversed per hour per spider, this
translates to an actual distance of less than 8mm per hour as compared to a
distance of 2-4mm per hour for the smaller colonies. Thus, although the
movement indices of the larger colonies are higher than those seen in smaller
groupings, the actual metabolic cost of the activity is likely to be low. It is
possible that greater efficiency in nest construction (e.g., work per spider) more
than offsets the slight increase in activity seen among the larger colonies, nearly
100% of the larger colonies had animals which did not move during the recording
period whereas smaller groupings exhibited less total inactivity (Table 2). This
suggests that in larger groupings, individuals may have greater opportunity for
extended periods of inactivity since they do not have to expend as much time
in nest construction. The Fourier analyses indicated that the larger colonies
organized their nocturnal rhythms into a series of high-activity bouts. Colonies
of size 20 showed greater coordination in the patterning of activity bouts as
evidenced by their shorter periodicity as compared to the smaller groupings.

The proximate cause for the patterning of activity seen in the larger groupings
is most likely due to web-transmitted vibrations. Thus the vibrations caused by
the movements of a single spider could induce movement in other colony
members. Similar effects have been described in several social species including
Cyrtophora citricola (Rypstra 1979), Metepeira spinipes (Uetz, in press), and
Oecobius civitas (Burgess 1976).

Vibrations transmitted through the web by walking spiders are extended below
the response window of 30-700 Hz recorded by Burgess (1979a), who suggested
that the movements of spiders are damped whereas vibrations of struggling prey
are enhanced. Although it is clear from Burgess’ work that fly vibrations are
accented by the web structure, the potential for low frequency vibrational
communication among nest mates is still possible for M gregalis. Hollar (in
Tietjen 1986) used a more sensitive photo-optic transducer than was available to
Burgess to record web movements and demonstrated that the vibrations of
normally walking spiders can be transmitted over distances of 10 cm or more.
Similarly, Jackson (1978) reports that the courtship vibrations of male M.
gregalis occur at about 10 Hz. These data suggest that the vibrations caused by
the movements of spiders could provide a means of intra-colony communication.

The occupation index and an examination of the space-utilization figures
indicate that members of the larger colonies occupy a greater area within the
Petri dishes than do those of the smaller groupings. Examination of the space-
utilization figures indicates that nearly all of the recorded activity was short-
distance movements (probably turning in place, silk depositions and grooming
movements). If the cribellate silk lining of chambers dampens the vibrations
caused by the movements of other spiders on the web, then animals located
within these structures could effectively remove themselves from the stimulation
caused by the activities of their nestmates. Casual observations suggest that
animals located outside chambers are more responsive to the movements of other

156

THE JOURNAL OF ARACHNOLOGY

spiders on the web than are those located within chambers. Thus, the structure
of the chambers could allow a proportion of the colony to maintain an in active
state without being stimulated to active behavior by the movement of other
colony members, as suggested by Tietjen (1982).

The organization of colony behavior among the Araneae does not depend on
coordination provided through ethological or morphological caste systems but
rather on mass action behavior, chemical communication, vibratory signals, and
periodicities in activity and occupation of web sites (Tietjen 1986). Web-
transmitted vibrations and interindividual interactions during activity peaks could
provide an efficient means for individuals to ascertain current colony conditions
related to population density and colony reproductive state. These effects,
coupled with web-position dependent behaviors, and chemical communication
may aid in organizing colony behavior.

ACKNOWLEDGMENTS

This research was supported by National Science Foundation Grant BNS 79-
10186 to W. J. Tietjen. I performed the construction of the computer-camera
apparatus at the laboratories of the North Carolina Foundation for Mental
Health Research. Data were collected while at the Department of Biology at
Georgia College. Statistical analyses were written and performed at Lindenwood
College. I am most grateful to Dr. P. N. Witt for the opportunities he provided
in electronics and programming as well as many helpful discussions during the
course of this research. George Uetz provided the spiders used in these series of
experiments. Dr. Jerome Rovner reviewed a preliminary copy of the manuscript.
The comments of S. E. Riechert and G. W. Uetz in their review of the manuscript
were greatly appreciated.

REFERENCES

Burgess, J. W. 1976. Social spiders. Scient. Amer., 234(3); 100-106.

Burgess, J. W. 1978. Social behavior in group-living spider species. Symp. Zool. Soc. London, 42:69-
78.

Burgess, J. W. 1979a. Web-signal processing for tolerance and group predation in the social spider
Mallos gregalis Simon. Anim. Behav., 27:157-164.

Burgess, J. W. 1979b. The spider species Mallos gregalis: Behavioral adaptations for living in Mexico.

PhD. Dissertation. North Carolina State University, Raleigh, N.C.

Buskirk, R. E. 1981. Sociality in Arachnida. Pp. 281-367, In Social Insects (H. R. Hermann, ed.),
vol. II. Academic Press, New York.

Chamberlin, R. V. and W. J. Gertsch. 1958. The spider family Dictynidae in America North of
Mexico. Bull. Amer. Mus. Nat. Hist., 116:1-152.

Conover, W. J. 1971. Practical Nonparametric Statistics. John Wiley and Sons, Inc., New York. 462

pp.

DeVoe, R. D. 1972. Dual sensitivities of the cells in wolf spider eyes at ultraviolet and visible
wavelengths of light. J. Gen. Physiol., 59:247-269.

DeVoe, R. D. and J. E. Zvargulis. 1967. Spectral sensitivities of vision in wolf spiders and jumping
spiders. Fed. Proc., 26:655.

Diguet, L. 1909a. Sur I’aranignee mosquero. C. R. Acad. Sci. Paris, 148:735-736.

Diguet, L. 1909b. Le mosquero: nid d’araignee employe’ a Mexique comme pie‘ge a’mouches.
Naturaliste, Paris, 31:283-285.

Jackson, R. R. 1978. Male mating strategies of dictynid spiders with differing types of social
organization. Symp. Zool. Soc. London, 42:79-88.

TIETJEN— EFFECTS OF GROUP SIZE ON COLONY BEHAVIOR

157

Jackson, R. R. 1979. Comparative studies of Dictyna and Mallos (Araneae, Dictynidae) II. The
relation ship between courtship, mating, aggression and cannibalism in species with differing levels
of social organization. Rev. ArachnoL, 2:103-132.

Jackson, R. R. 1979. Comparative studies of Dictyna and Mallos (Araneae, Dictynidae) V. Tolerance
and resistance to starvation. Psyche, 87:21 1-220.

Kaston, B. J. 1948. Spiders of Connecticut. State Geological and Natural History Survey. Hartford,
Conn., 874 pp.

Kullman, E. J. 1972. Evolution of social behavior in spiders (Araneae: Eresidae and Theridiidae).

Amer. ZooL, 12:395-405.

Owens, A. J. 1981. On the use of Fourier transforms to explore biological rhythms. Byte: the Small
Systems Journal, 6:314-326.

Rypstra, A. L. 1979. Foraging flocks of spiders. Behav. Ecol. Sociobiol., 5:291-300.

Shear, W. A. 1979. The evolution of social phenomena in spiders. Bull. British ArachnoL Soc., 1:65-
76.

Sokal, R. R. and F. J. Rohlf 1969. Biometry. W. H. Freeman, San Francisco, 776 pp.

Tietjen, W. J. 1980. Sanitary behavior by the social spider Mallos gregalis (Dictynidae): Distribution
of excreta as related to web density and animal movements. Psyche, 87:59-73.

Tietjen, W. J. 1981. An inexpensive microcomputer-based image digitizer. Behavior Research Methods
and instrumentation, 13:347-350.

Tietjen, W. J. 1982. Influence of activity patterns on social organization of Mallos gregalis (Araneae,
Dictynidae). J. ArachnoL, 10:75-84.

Tietjen, W. J. 1986. Social spider webs with special reference to the web of Mallos gregalis. Pp. 100-
121, In Webs and Behavior (W. Shear, ed.). Stanford Univ. Press, Stanford.

Uetz, G. W. 1983. Sociable spiders. Nat. Hist., 92:62-69.

Uetz, G. W. 1985. Environmental and genetic influences on social grouping tendency of a communal
spider. PP. 000-000, In Evolutionary Genetics of Invertebrate Behavior (M. A. Huttel, ed.). Plenum
Press, New York.

Wilson, E. O. 1971. The Insect Societies. Harvard University Press, Cambridge Mass. 548. pp.

Witt, P. N., M. B. Scarboro and D. B. Peakall. 1978. Comparative feeding data in three spider species
of different sociality: Araneus diadematus CL, Mallos trivittatus (Banks) and Mallos gregalis
(Simon). Symp. ZooL Soc. London, 42:89-97.

Manuscript received July 1984, revised September 1984.

m

.fi-.t-i. I ■■■ VI

'■ ■■■■

,' VJJIJ .. *r^ a Hv J

I! ‘4u

‘ '"••iyw%(i.(i^.,«i •<■« . ■' < V, i .j

'T, ‘i. M:i;i«ff!^|(RfSA :.if^ ■■ '•T ^i' •».! 4^^ arV!*l-

V •■ '.fi'u;. •-.

»♦ I H<A f.^ti' * - , Wji'

r*V:W

'4

» I :■ f ' f. vV 1

^. ■&.j>Fiimb^ 'Wbiir-.<'>« :j

■/i- . . i ' ‘.'t' ■ :• : i i

'^iv; ■;-.::.'._i:: -~r;:x^.: . ^

,. ^ .! .}) v,v .

{t,-;- f . ■.' . f ... ■'.-
■;.-r ''.r.i^.'P'V >, • - "t

IM'V T'.l

,1 ■■ '! • * ‘ < ^“C

JE :

. *t frf

«■ N J.‘

U> t.>

...O. -, I *. ‘ ■ i *■■ -I •w \

* ' . *' * /'i ■ . jJ

•svuyj’ );;/**

. .•■ A,, . ■ •

_ C».-

• I*, i ' ■' *.. i-*' vV ■ "»* U tV*

^••■^ ^;^l:■!**i^'*••* V' . •■''-* •Vs ^ ' ife^.. - ■ ■■

;-«:r‘

Uetz, G. W. and K. R. Cangiaiosi. 1986. Genetic differences in social behavior and spacing in
populations of Metepeira spinipes, a communal-territorial orb weaver (Araneae, Araneidae). J.
ArachnoL, 14:159-173.

GENETIC DIFFERENCES IN SOCIAL BEHAVIOR AND
SPACING IN POPULATIONS OF METEPEIRA SPINIPES,
A COMMUNAL-TERRITORIAL ORB WEAVER
(ARANEAE, ARANEIDAE)^

George W. Uetz and Karen R, Cangiaiosi

Department of Biological Sciences
University of Cincinnati
Cincinnati, Ohio 45221 U.S.A.

ABSTRACT

Metepeira spinipes F. O. Pickard-Cambridge, a communal/ territorial orb-weaver from Mexico,
shows considerable geographic variation and temporal flexibility in spacing. A series of laboratory
studies was conducted to test whether the variation in spacing observed in the field is solely the result
of behavioral plasticity in response to environmental conditions, or the result of mechanisms inherent
in different populations (i.e., genetic differences in behavior). Spiders from source populations in
desert and moist tropical habitats were collected as eggs and raised in the laboratory under identical
controlled conditions. Measurements of three-dimensional spacing parameters in laboratory colonies
(nearest neighbor distance, within-colony density) have shown significant differences in spatial
organization between populations, suggesting a genetic basis to these differences. Behavioral
observations confirm that there are behavioral ecotypes within this species, with levels of sociality
adapted to the regions in which they occur.

INTRODUCTION

There are constraints on the social behavior of orb weaving spiders (families
Araneidae and Uloboridae) that make the evolution of sociality less likely in this
group. Unlike the sheet or tangle webbing of spiders in other families which
exhibit communal web building (Theridiidae, Agelenidae, Dictynidae, Eresidae,
Amaurobiidae, Oecobiidae, Pholcidae), the orb web cannot be built by more than
one spider. It is the result of a complex sequence of behaviors which are tightly
controlled by the genetic “program” of individual spiders (Witt and Reed 1965,
Witt et al. 1968), unlikely to be modified to include participation by others
(Lubin 1974, Burgess and Cangiaiosi 1982). Thus, cooperative web construction,
at least of the prey catching portions of the web, is precluded.

Most orb weavers are limited in their social organization to aggregations of
potentially of potentially competing individuals building and occupying their own
webs within a shared web foundation (Lubin 1974, Buskirk 1975, a,b, Fowler and
Diehl 1978, Uetz et al. 1982, Smith 1983). This type of social organization is
termed “communal/ territorial” (Jackson 1978), and is conceptually (and

‘Presented in a symposium, “Social Behavior in Spiders”, at the National Meeting of the American
Arachnological Society, June 27, 1984.

160

THE JOURNAL OF ARACHNOLOGY

sometimes evolutionarily) intermediate between the incipient sociality of
fortuitous spider aggregations, and more advanced social spiders (which exhibit
cooperation in prey capture, feeding, and brood care) (Buskirk 1981, Burgess and
Uetz 1982, Rypstra 1983).

These spiders must reconcile the demands of conflicting behavioral strategies —
communal aggregation and defense of space — while at the same time contend
with a mosiac of contiguous territories in three dimensions instead of two. The
result is a three-dimensional spatial arrangement of group members which is
somewhat fixed, yet may change every time webs are renewed (usually on a daily
basis) depending on the outcome of aggressive interactions between individuals.
The “territory” defended has indistinct boundaries, in that the silk connections
between the retreat, the orb or catching spiral, and the space web are connected
to and overlap with these of other individuals. In addition, this is a ’’multipurpose
territory” (Davies 1978), in that it may include a foraging site (the orb), a
habitation (the retreat), a mating site (orb and/or retreat), and a breeding/egg
laying site (the retreat). Thus, the spatial organization of communal/ territorial
orb weavers reflects a compromise solution to the conflicting selection pressures
of competition for space and food resources and for benefits gained from
communal foraging. An analysis of spatial organization in these spiders should
clarify the relative importance of those selection pressures in the evolution of
spider sociality.

For several years, we have studied the behavior and ecology of Metepeira
spinipes F. O. Pickard-Cambridge (Araneidae), a communal/ territorial orb-
weaving spider species found in central Mexico (F. O. Pickard-Cambridge 1903).
Previous studies have shown considerable between-population variation in the
spatial organization of this species (Burgess and Uetz et al. 1982, Uetz 1983, Uetz
et al., in press). This variation makes M. spinipes an interesting species with
which to investigate spacing, and raises questions about the influences of
environmental and genetic factors on the social organization of communal-
territorial spiders.

BACKGROUND

Metepeira spinipes occurs solitarily, but more frequently occurs in groups of
5 to 150 or more individuals (Burgess and Witt 1976, Uetz and Burgess 1979).
Metepeira spinipes is the only communal species in this genus (Levi 1977),
although there is one other species, Metepeira daytona Chamberlin and Ivie,
found in the Carribbean, which occasionally exhibits group web-building
(Schoener and Toft 1983). The genitalia of all specimens examined to date
resemble the drawings of Pickard-Cambridge in the species description. In
addition, preliminary studies of genetic differentiation between populations in
different geographic areas using starch gel electrophoresis tentatively indicate that
M. spinipes is a single species (Uetz et al., in press).

The web of individual Metepeira spinipes (Fig. 1) contains a three-dimensional
space (or barrier) web with a retreat, and a sticky orb connected to the retreat
by signal threads (Burgess and Witt 1976). Although the sticky orb webs are
taken down and renewed on a daily basis, a communal space web persists and
acts as a framework for the web building activities of numerous individuals. The

UETZ AND CANGIALOSI— SOCIAL SPIDER SPACING

161

resultant 3-dimensional colony consists of spiders inhabiting interconnected,
fixed, foraging locations, potentially changing positions on a daily basis.
Individuals maintain and defend orbs and retreats within the colony, and capture
their own prey.

Group size in M. spinipes varies with habitat (Uetz et al. 1982). In severe
habitats where prey availability is low (e.g. desert grasslands), spiders are
predominantly solitary or live in small groups. In sites where climate is benign
all year and insect abundance is great (e.g. moist tropical forest), colony size is
very large. Nearest neighbor distance decreases over the habitat gradient between
these sites and in general appears to be inversely related to prey availability (Uetz
et al. 1982, Uetz et al. in press). This relationship is further supported by the
results of field experiments, in which nearest neighbor distance increased in
colonies after relocation to prey-poor sites, except when colonies had prey
supplemented (by addition of cow dung to their new sites) (Uetz et al. 1982).

Metepeira spinipes can apparently tolerate conspecifics at closer distances in
areas where prey are more abundant, as Riechert (1978a) has found in
Agelenopsis aperta (Gertsch), a solitary, desert funnel web spider. Agelenopsis has
a minimum territory size, genetically set at an area that provides the spider with
sufficient prey biomass (Riechert 1981, Riechert in press). In contrast, M. spinipes
shows a rapid change in spacing when food availability changes. This may be
explained by the fact that orb weavers, unlike funnel web builders, renew their
web on a daily basis (Uetz 1985, Riechert and Gillespie, in press). In this
communal/ territorial orb weaver, rising hunger and aggression levels associated
with food deprivation may result in greater nearest neighbor distances when the
webs are rebuilt each day.

Observation of behavior and spacing patterns of spiders in laboratory cages
confirm the findings of field studies. Spiders in cages at low levels of prey

162

THE JOURNAL OF ARACHNOLOGY

availability were spaced at significantly greater distances from each other than
those in cages at maintenance or satiation levels of prey (Uetz et al. in press).
Instances of cannibalism (n = 9) were highest in the low prey availability cages
(30%), and very low (3%) in the other treatments. There were no significant
differences in nearest neighbor distances of spiders at maintenance and satiation
levels of prey. This suggests that there is an intrinsic lower limit of distance
(analogous to territory size) at which conspecifics can be tolerated.

Genetic differences in the degree of tolerance of conspecifics are suggested by
laboratory observations of differences in the behavior of field collected M.
spinipes from various localities. In particular, spiders from the tropical site
appear far more tolerant of each other. During transport, many individuals can
be placed together in small containers with water but no food, and survive for
up to two weeks without cannibalism. In contrast, spiders from the other
populations cannot exist under these conditions for more than a day or two. In
laboratory cages under identical conditions, field collected individuals from the
tropical site and desert site show distinct differences in web building behavior,
web structure, and inter-individual spacing. These differences are apparent from
the first day of introduction to the cages (Uetz et al. in press).

The questions raised by the results of these earlier studies concern the
possibility of some genetic control of social organization in this species. Could
the very wide range of group size and spacing seen in M. spinipes be solely the
result of behavioral plasticity in response to climate and food availability? Or, is
it possible that different social behavior or tolerance strategies are selected for in
different environments and are reflected in the differences seen between
populations? To answer this question, an experimental study was conducted in
the laboratory using a classic behavior genetics technique — rearing spiders from
different populations under identical, controlled conditions.

In this study, most proximal environmental (ecological) factors that might
influence spacing in M. spinipes (climate, predation, availability of prey, web site
availability, colony density) were controlled in the laboratory. Thus, differences
in nearest neighbor distance and other spacing parameters between populations
that might be attributable to genetic factors inherent in the populations can be
revealed. Experiential factors, and the interaction of genes and environment are
investigated in another study (Cangialosi and Uetz, in prep.)

METHODS

Representative field sites were chosen in three geographic regions where
populations of M. spinipes exhibit different levels of social behavior as indicated
by group size and spacing: (1) In the northern desert region, near San Miguel
de Allende, where M. spinipes is found in small groups with maximal inter-
individual distances. (2) In the tropical mountainside of Fortin de las Flores,
where the spiders are found in very large groups with minimal inter-individual
distances. (3) In the central valley of Mexico, in Tepotzotlan, an agricultural area
north of Mexico City, where the spiders are found in group sizes and with spatial
organization intermediate between (1) and (2) above. These sites represent the
range of variation in social spacing seen in this species, and are easily accessible,
being located off major highways. The vegetation and environments of these areas

UETZ AND CANGIALOSI— SOCIAL SPIDER SPACING

163

are described in Shelford (1963) and previous data on climate variables, prey
insect availability, and aspects of M. spinipes natural history are also available
(Uetz et al. 1982, Uetz 1985, Benton and Uetz in press).

Egg sacs were collected from the field sites in October 1982 and February 1983
and brought back to Cincinnati, Ohio for laboratory rearing under controlled
environmental conditions. Spiderlings used in the experiment were selected from
several egg sacs or egg sac strings. Each experimental group was a mixture of
eggs from six females, taken from different colonies of equal size at each site,
in order to minimize (or at least equalize) possible maternal effects. It is certainly
impossible to control for all differences between eggs, especially when collecting
them from females in the field (whose precise individual history is not known).
However, they were collected from sites (and in some cases, colonies) for which
previous ecological data are available, and this selection process undoubtedly
reduced the potential for error between experimental groups due to egg
differences within each group.

Spiders were housed in a walk-in environment room with light regulated on
the same daily regime occurring in their natural habitat (12 hrs light; 12 hrs
dark). Spiders were raised to adulthood under conditions of controlled climate
(Temperature 27° C; Rel. Humidity 75%). These conditions are optimal for
survival of M. spinipes, as established from previous research.

Spiders were raised communally through adulthood in groups of 30 individuals
in cages (100 cm X 75cm X 80 cm) where cultures of Drosophila and Musca flies
provided prey ad libitum. Three cages of spiders from each population (desert
and moist tropical forest) were established. Spiders were supplied with prey ad
libitum in order to control for hunger as an influence on spacing. Under
conditions of apparent food satiation, a minimal interindividual spacing level is
reached (Uetz et al. in press). In a separate study (Cangialosi and Uetz, in prep.)
spiders were reared individually in isolation to control for experiential effects.

The fixed colony size (30 spiders per cage) for this experiment was arrived at
after many attempts at rearing these animals at a variety of densities, and was
chosen for several important reasons. First, by keeping the number of spiders
constant in all cages, differences in spacing seen in the field due to the compound
influence of population density and dispersion are eliminated. Second, it is
important not to set up a false test of spacing differences — where spiders from
the desert are overcrowded, experiencing a density far greater than they would
in nature, while the tropical spiders spread out their webs in what to them may
be wide open space. Having 30 spiders in a cage yields a group size in the upper
range for the desert spiders (yet still encountered in nature with a frequency
0.05), and a density of approximately 40 spiders/ cubic meter (which is by no
means overcrowded — densities far great are seen frequently in nature). Likewise,
this group size and density are not unlike those seen for smaller colonies in the
tropical population under natural conditions.

The chosen colony size also serves as a conservative test of the tendency for
individuals from the tropical population to be spaced closer together, even when
space is not limiting. If inter-individual spacing expands, then it may be assumed
that in nature, some aspect of colony size and density (e.g., intrusion by other
spiders on individual space), is causing spiders to be spaced more closely together
Hixon, 1980, Schoener 1983). If cages housing individuals representing tropical
populations show smaller nearest-neighbor distances (NND) than cages with

164

THE JOURNAL OF ARACHNOLOGY

spiders from desert populations (as has been observed in previous studies), then
an inherent tendency toward closer association with conspecifics will be
demonstrated.

Using a fixed number of spiders per cage raises another set of problems that
must be addressed — what happens if there is differential mortality in cages? Might
this differentially affect nearest neighbor distance? As luck or fate may have it,
there was some mortality, but statistical analysis has shown that there is no
apparent directionality in its effect; NND and subsequent no. of spiders/ cage are
not significantly correlated for all cages from both populations (r = 0.21; p >
0.45).

Each of the cages was gridded with cm markers on the outside, so that the
exact position of each animal within the group could be determined using x, y,
z coordinates. Positions of individuals within colony cages were recorded at
regular intervals corresponding to developmental stages (immatures: instars 5-7;
pre-reproductive adults, reproductive adults), using the aforementioned
coordinates. This was done to insure that when measurements were taken, all
spiders (from both populations) were at approximately the same stage and size,
reflecting equivalent territory sizes. Data collected on location of individuals
within colonies were loaded into a computer program used to calculate nearest
neighbor distances using the following formula (Buskirk and Uetz 1982, Major
and Dill 1978):

Distance (1,2) = [(X2 -X,)^ + (Y2 -Yi)^ + (Z2 -z,)^]

Distance between all possible pairs of animals may be obtained using this
formula, and the computer program calculates means for 1st, 2nd, 3rd... Nth
nearest neighbors. Nearest neighbor distance is a widely used measure of animal
spacing, and provides a means of quantitatively comparing populations.

Density of spiders within the colony was also calculated. Because the number
of spiders per cage remains fairly equal between cages, this measure estimates the
internal cohesiveness of individual spacing within the group. A direct measure of
density was used by determining the amount of the cage occupied by the
communal web, and dividing that into the number of spiders per cage. Thus, if
there was any mortality, the measure of density was scaled by the size of the
spider colony and not nearest neighbor distance (as in other measures).

Spiders were observed in cages between 6:00 AM and 9:00 AM EST, which
is the time period during which webs are built and most behavioral interactions
occur (3 hours before “dawn”). A total of 9.5 hours observation time for the
desert population, and 8.25 hours of observation time for the tropical population
were accumulated in short periods (approximately 30 minutes — 1 hr at a time)
over several weeks. Numbers of interactions were totalled for each population,
and sequences of behavior during each interaction were recorded.

In Fortin de las Flores in August 1983, positions of nearest neighbors relative
to each other in 3-dimensional space were examined within a small colony
(approx. 100 individuals) and a much larger one (approx. 1500-2000 individuals),
determined from location data in both horizontal and vertical aspects.
Orientation of individuals facing webs (and thus the direction the web faces) was
recorded in xy plane, and set as zero degrees. Location of individuals relative to
each other in the xy plane (bearing) and in the xz plane (elevation) was recorded
with the vertical axis set at zero degrees.

UETZ AND CANGIALOSI— SOCIAL SPIDER SPACING

165

Table 1. — ANOVA: Nearest neighbor distances of spiders in laboratory cages (* = < 0.05).

Time

Period

Development

Stage

Source of Variation

F Value

1

immatures (instars 5-7)

Source Population

17.43*

Replicates

1.38

2

penultimate

Source Population

8.85*

Replicates

2.53

3

pre-reproductive adult

Source Population

24.64*

Replicates

4.84

4

reproductive adult

Source Population

35.74*

Replicates

0.60

5

adult females with eggs

Source Population

7.28*

Replicates

1.20

RESULTS AND DISCUSSION

In the laboratory, under identical controlled conditions, significant differences
were seen in the spacing of spiders from separate source populations at all stages
of their juvenile development (Table 1). The spiders from the moist tropical site
in Fortin de las Flores had a significantly lower nearest neighbor distance (NND)
than spiders from Tepotzotlan or San Miguel de Allende (Duncan’s Multiple
Range Test, p < 0.05). Spiders from the desert grassland population (San Miguel)
and the agricultural central valley (Tepotzotlan) showed no significant differences
in NND.

NND actually varies very little over several months’ time in the laboratory
(Figure 2), which suggests that under these conditions, the spiders maintained a
minimum NND. This suggestion is supported by a comparison of field and lab
data for adult females. In populations from San Miguel and Tepotzotlan, lab
NND was significantly lower than field NND (Student’s “t” test with unequal
variance correction, p < 0.05). For the populations from Fortin de las Flores,
no difference was seen in lab and field NND, perhaps because prey are so
abundant in the field there. At that site, however, female — female distances may
be misleading, because all ages are present at one time. In the lab, cohorts of
similar ages comprise the colonies.

An early concern in this study was that random mortality might cause the
numbers of spiders in each cage to be different, and thus influence NND.
However, within each time period or set of cages, NND and number of spiders
per cage (after mortality) were not correlated (Person’s “r” = 0.21; p < 0.45).
Even so, to guard against such a possibility, a measure of spacing independent
of NND was made by estimating the density within colonies (no. of spiders/ m^)
by measuring the volume of the webbing and dividing it into the number of
spiders in the cage. A comparison of the observed density and the expected
density (based on the no. of spiders/ volume of cage) (Figure 3) shows a
significant difference between spiders from the moist tropical site and those from
the other two areas (G test; p < 0.01). These data suggested that spatial
arrangements within colonies differ between populations, and in fact, consistent
differences in the spatial organization within colonies were revealed. When actual
positions of spiders were plotted in three dimensional graphic arrays, spiders from
Fortin tended to group in a cluster in the upper center of the cage, whereas

166 THE JOURNAL OF ARACHNOLOGY

Fig. 2. — Mean nearest neighbor distance ± 95% conf. limits) of three replicate M. spinipes colonies,
each set from three different field populations and reared over their life cycle in the laboratory under
identical conditions. (SMA = San Miguel de Allende; TPN = Tepotzotlan; FDLF = Fortin de las
Flores). Time periods represent developmental stages as indicated in text and Table 1.

spiders form Tepotzotlan and San Miguel tended to be dispersed throughout the
cage (Fig. 4a, b). These differences were consistent in all cages, and throughout
all time periods.

From these findings, it is apparent that populations from the extreme ends of
the range of spacing seen in nature show consistent differences in NND when
raised under identical conditions (conditions which should result in a minimal
level of distance between neighbors). Although the remote possibility of some
maternal or egg effect cannot be ruled out entirely, these data suggest some
genetic differences, perhaps influencing behavior and thus spacing. It is probable
that there are ecotypes (genetically distinct subpopulations), subspecies, or
emerging species populations showing differences with respect to social behavior,
which have spacing arrangements adapted to environments where they occur.
Genetic influences, acting along with environmental influences like climate and
food availability, result in the variation in social spacing observed in earlier
studies.

An important question that must now be raised concerns the nature of genetic
control of spacing behavior. How can genes influence the nearest neighbor
distance in a colony of spiders? There would appear to be two main ways in
which this could occur, and there are several lines of evidence supporting each
mechanism.

It may be possible that there are differences in the interaction strategies of
spiders that have become genetically fixed in each population (as has been found
for Agelenopsis by Riechert 1983, 1984) resulting in clear differences in the levels
of aggression shown. More aggressive spiders would be widely spaced, whereas

UETZ AND CANGIALOSl— SOCIAL SPIDER SPACING

167

CO

QC

LU

9

Q.

w

>

CO

z

LU

Q

SOURCE POPULATION

Fig. 3. — Density of M. spinipes in laboratory colonies (see text for explanation of observed and
expected values).

less aggressive ones would be spaced closer together. Spiders from the desert and
moist tropical forest populations, maintained in laboratory cages under identical
controlled conditions of this study, show differences in their interactive behavior.
In 9.5 hours of observation, spiders from the desert population had 40 encounters
in which agonistic behavior was seen (0. 178/spider-hr). In contrast, in 8.25 hours
of observations, the spiders from the tropical forest population had 119 such
encounters (0.560/ spider-hr). This apparent difference would at first appear
counter to both theory and previous findings. An explanation may be found in
the actual patterns of behavior seen in these interactions (Figs. 5a, b). In the
tropical population, where agonistic encounters are more frequent, a majority of
bouts are concluded with a single web pluck exchange (Fig. 5a). In the desert
population, bouts are more prolonged, and escalate more quickly to chasing and
grappling (Fig. 5b).

These findings are consistent with the predictions of game theory, as
demonstrated by Riechert (1982, 1983). Selection would favor an aggressive
behavior strategy that is appropriate for the level of resource availability in the
local environment. If space and food were the limiting factors, an aggressive
behavior strategy which results in the securing of a web site within the colony
at the optimum level of cost/ benefit would prevail. For the desert population,
where web sites and food availability are limited, a strategy of rigorous defense
of territory is appropriate. In tropical colonies, where prey are abundant and the
quality of web sites within the immense communal web may be more or less
equal, the appropriate strategy is for intruders to give up without a fight, and
move elsewhere. This may also be true of residents, which have often been
observed giving up a web to a more persistent intruder. The outcome of these
different strategies is that once settled into web locations, desert spiders move

168

THE JOURNAL OF ARACHNOLOGY

Fig. 4. — Location of individual spiders in laboratory colony cages (a) a representative FDLF cage;
(b) a representative SMA cage.

UETZ AND CANGIALOSl— SOCIAL SPIDER SPACING

169

considerably less often than tropical spiders do, and thus fewer aggressive
encounters occur among them.

It may also be possible that there is a fixed distance at which an intruder on
the web ‘‘territory” will be met with aggressive behavior, and that this distance
varies between populations. The result would be that some populations of spiders
are inherently more tolerant of neighbors at closer distances, and would be
spaced more closely than others. In communal/ territorial web building spiders,
the location of nearest neighbors in space, and the orientation of their webs may
relate to competition among colony members, or may somehow relate to prey
capture success. Observations of encounters between spiders suggest that the
presence of an intruder at some point close to the retreat or web will result in
the initiation of agonistic behavior sequences. We have observed that the location
of the intruder relative to the resident at the time this conflict is initiated varies,
which may provide some support for the concept of a variable, 3-dimensional
’’territory” in these spiders. Differences in territory size between populations
might be determined by whatever minimum area is required to provide an
individual with sufficient prey, based on its availability in the habitat, as has been
shown for Agelenopsis by Riechert (1978).

Field studies of spiders from the tropical population have shown that the
location of nearest neighbors and the orientation of their webs relative to each
other changes with colony size. In these studies, nearest neighbor location data
were plotted in hemispheric projection relative to a central point by making the
reference individual the one whose nearest neighbor is measured. This type of
projection (Fig. 6a, b) was used in order to determine any patterns in nearest
neighbor spacing (as in Major and Dill 1978). This also allows a determination
of the basic shape of the 3-dimensional territory. As colony size increases, spiders
become more tightly packed, with webs facing opposite directions, and a pattern
emerges (Fig. 6b) wherein nearest neighbors are located behind and below, or to
the side of a spider’s web (but never above or below in front of the web). This
suggests that territorial space within the colony is asymmetrical and compressible,
and that spiders are more tolerant of neighbors in locations where they pose less
of a threat to obtaining food. Data are not available for the desert population,
but it is possible that the size, shape and degree of compressibility of this 3-
dimensional territory may differ there.

The differences in social behavior observed, as well as morphological
differences between populations, might indicate that Metepeira spinipes is not a
single species. There are consistent size differences, and spiders from Fortin are
smaller and have shorter legs than spiders from the other populations. There are
also consistent differences in the abdominal folium and ventral markings.
However, specimens show considerable individual variation (including those from
the same colony), and look slightly different, presumably because the flexible
scape of the epigynum preserves at varying orientations, and the degree of
abdomen distention affects color. Research on genetic similarity of M. spinipes
populations from these and several other areas in Mexico using polyacridamide
gel electrophoresis, has shown all populations to be highly similar. Nei’s Index
of Similarity (Nei 1972) for these populations was > 0.96 (Uetz et al., in press),
a level of genetic similarity in the range of variation seen within many arthropod
species populations (Selander and Johnson 1973, Ayala 1976). These studies,
however, were done comparing only a few loci within a limited number of

DESERT POPULATION TROPICAL POPULATION

170

THE JOURNAL OF ARACHNOLOGY

o

o <u

-O

2 i

U c

•S

3

9 r:

o t:

1- ii
o o

<*1 u

C/3 'O

e c

cd

'•5 £

3

o

i|

!T 5
<U O

<i: G

UETZ AND CANGIALOSI— SOCIAL SPIDER SPACING

171

COLONY OF 100 SPIDERS COLONY OF 1500 SPIDERS

Fig. 6. — Hemispheric projection of M. spinipes nearest neighbor locations in Fortin de los Flores
(see text for explanation): (a) a colony of 100 spiders; (b) a colony of 1500-2000 spiders.

samples, and conclusions drawn are tentative. Attempts to mate individuals from
different populations have yielded inconclusive results. When pairs of males and
females were isolated no courtship or mating behavior was ever observed. Egg
sacs were produced, but the spiderlings never hatched, and no cause of the
hatching failure was apparent. We couldn’t determine if this failure to crossbreed
is indicative of species isolation, or the result of age incompatibility, or some
other factor.

CONCLUSION

Evidence presented here strongly suggests that both environmental and genetic
factors influence the variation in spatial organization observed in Metepeira
spinipes colonies. These findings must then mean that populations from the desert
and tropical forest habitat represent either separate genetic subgroups within the
species, or newly evolved species with different types of social behaviors. The
differences in social spacing and agonistic behavior seen in these populations are
likely to be adaptive, and result in improved survival and reproduction in the
environments in which they occur. These populations (or species) might then
represent early stages in the evolution of increasing social tolerance in spiders.

ACKNOWLEDGMENTS

We express our appreciation to a number of people for assistance in this
research. Thomas Kane, Gail Stratton, Michael Benton, Curtis Meininger, Bob
Hollis, Kathy Duncan, James Chapman and Debbie Fritz have all assisted in the
field, while Mary Klus, Bob Hollis, Tracy Troyan and Debbie Fritz have assisted
in the lab. We are especially grateful to Roger Steubing for statistical guidance
and computer consultation. We thank Leah Hammons for typing the manuscript.

172

THE JOURNAL OF ARACHNOLOGY

Portions of this research have been submitted in partial fulfillment of the
requirements for the degree of Master of Science in Biological Sciences at the
University of Cincinnati (Karen Cangialosi). This research has been funded by
grants from the National Geographic Society, the American Philosophical
Society, the Society of Sigma Xi, and the University of Cincinnati Research
Council.

LITERATURE CITED

Burgess, J. W. and G. W. Witt. 1982. Social spacing strategies in spiders. In Spider Communication,
Mechanisms and Ecological Significance (R N. Witt and J. S. Rovner, eds.). Princeton Univ. Press,
440 pp.

Burgess, J. W. and P. N. Witt. 1976. Spider webs: design and engineering. Interdisc. Sci. Rev., 1:322-
355.

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

Buskirk, R. E. 1975b. Coloniality, activity patterns and feeding in tropical orb-weaving spider.
Ecology, 56:1314-1328.

Buskirk, R. E. 1981. Sociality in the Arachnida. In Social Insects (H. R. Hermann, ed.). Vol. H, Ch.
7.

Davies, N. B. 1978. Ecological questions about territorial behaviour. Pp. 317-350, In Behavioural
Ecology: an Evolutionary Approach (J. R. Krebs and N. B. Davies, eds.). Sinauer Assoc.,
Sunderland, Mass.

Fowler, H. G. and J. Diehl. 1978. Biology of a Paraguayan orb-weaver, Eriophora bistriata (Rengger)
(Araneae, Araneidae). Bull. British Arachnol. Soc., 4:241-250.

Jackson, R. R. 1978. Comparative Studies of Dictyna and Mallos (Araneae: Dictynidae). I: social
organization and web characteristics. Rev. Arachnol., 1:133-164.

Levi, H. W. 1977. The orb weaver genera Metepeira, Kaira and Aculepeira in America North of
Mexico (Araneae: Araneidae). Bull. Mus. Comp. Zool., 148:185-238.

Lubin, Y. D. 1974. Adaptive advantages and the evolution of colony formation in Cyrtophora
(Araneae: Araneidae). Zool. J. Linn. Soc., 54:321-339.

Major, P. F. and L. M. Dill. 1978. The three-dimensional structure of airborne bird flocks. Behav.
Ecol. Sociobiol., 4:111-122.

Maynard-Smith, J. and S. E. Riechert. 1984. A conflicting tendency model of spider agonistic
behavior: Hybrid-pure line comparisons. Anim. Behav., 32:564-578.

Nei, M. 1972. Genetic distance between populations. Amer. Nat., 106:283-292.

Pickard-Cambridge, F. O. 1903. Arachnida: Metepeira, gen. nov. Biol. Centrali-Americana, 2:457-460.

Riechert, S. E. 1978a. Energy-based territoriality in populations of the desert spider, Agelenopsis
aperta (Gertsch). Symp. Zool. Soc. London, 42:21 1-222.

Riechert, S. E. 1978b. Games spiders play: behavioral variability in territorial disputes. Behav. Ecol.
Sociobiol., 3:135-162.

Riechert, S. E. 1979. Games spiders play H: Resource assessment strategies Behav. Ecol. Sociobiol.,
6:121-128.

Riechert, S. E. 1982a. The consequences of being territorial: spiders, a case study. Amer. Nat.,
117:871-892.

Riechert, S. E. 1982b. Spider interaction strategies: communication vs. coercion. In Spider
Communication, Mechanisms and Ecological Significance (P. N. Witt and J. S. Rovner, eds.).
Princeton Univ. Press, 440 pp.

Riechert, S. E. 1984. Games spiders play HI: Cues underlying context associated changes in agonistic
behavior. Anim. Behav., 32:1-15.

Riechert, S. E. (in press). Between population variation in spider territorial behavior: Hybrid-pure line
comparisons. In Evolutionary Genetics of Invertebrate Behavior (M. A. Huettel, ed.). Plenum
Press.

Rypstra, A. L. 1979. Foraging flocks of spiders — a study of aggregate behavior in Cyrtophora
citricola Forskal (Araneae: Araneidae) in West Africa. Behav. 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, 59:312-316.

UETZ AND CANGIALOSI— SOCIAL SPIDER SPACING

173

Schoener, T. W. and C. A. Toft. 1983. Dispersion of a small-island population of the spider Metepeira
daytona (Araneae: Araneidae) in relation to web-site availability. Behav. Ecol. Sociobiol., 12:121-
128.

Shelford, V. E. 1963. The Ecology of North America. Univ. Illinois Press, Chicago. IL.

Smith, D. R. 1983. Reproductive success of solitary and communal Philoponella oweni (Araneae:
Uloboridae). Behav. Ecol. Sociobiol., 11:149-154.

Uetz, G. W. and J. W. Burgess. 1979. Habitat structure and colonial behavior in Metepeira spinipes
(Araneae: Araneidae), an orb weaving spider from Mexico. J. Arachnol., 7:121-128.

Uetz, G. W., T. C. Kane and G. E. Stratton. 1982. Variation in the social grouping tendency of a
communal web-building spider. Science, 217-547-549.

Uetz, G. W. 1983. Sociable Spiders. Nat. Hist., 92:62-79.

Uetz, G. W. (in press). Web-building and prey capture in communal orb weavers. In Spider Webs
and Spider Behavior (W. A. Shear, ed.). Stanford Univ. Press.

Uetz, G. W., T. C. Kane, G. E. Stratton and M. J. Benton (in press). Environmental and genetic
influences on the social grouping tendency of a communal spider. In Evolutionary Genetics of
Invertebrate Behavior (M. A. Huettel, ed.). Plenum Press.

Witt, P. N. and C. F. Reed. 1965. Spider-web building. Science, 149:1190-1197.

Witt, P. N., C. F. Reed and D. B. Peakall 1968. A Spider’s Web. Problems in Regulatory Biology.
Springer Verlag. Berlin.

Manuscript received January 1985, revised October 1985.

Riechert, S. E., R. Roeloffs, and A. C. Echternacht. 1986. The ecology of the cooperative spider
Agelena consociata in Equatorial Africa (Araneae, Agelenidae). J. Arachnol., 14:175-191.

THE ECOLOGY OF THE COOPERATIVE SPIDER
AGELENA CONSOCIATA IN EQUATORIAL AFRICA
(ARANEAE, AGELENIDAE)^

Susan E. Riechert, Rosemarie Roeloffs and Arthur C. Echternacht

Department of Zoology
University of Tennessee
Knoxville, Tennessee 37996, U. S. A.

ABSTRACT

The ecology of the spider Agelenopsis consociata in rain forest habitats in Gabon was investigated
with reference to factors that might underlie its cooperative foraging behavior. Colonies consisted of
local clusters of from one to 27 nests and associated web traps. The composition of individual nests
also was highly variable, ranging from one adult spider to as many as 1100 adults. Nest survivorship
was a positive function of group size as estimated by the number of adult females in a nest. Analysis
of the energy budget of single adult females indicated that the high rate of extinction of small nests
might result from insufficient prey availabilities during the two rainy seasons when webs are destroyed
on 40% of the days. Individuals associated with larger nests do not experience this energy deficit
because individual investment in the web trap decreases with increasing colony size. Dispersal
problems may also favor the maintenance of groups: we observed heavy predation on individuals that
were released experimentally. On the negative side, individual foraging success and production of eggs
decreased with increasing nest size in our experimental groups, perhaps due to the deleterious effects
of interference. In addition, we consider that the continuity of generations permitted by the equatorial
environment fosters the cooperative life style compared to other adaptations that might be exhibited
in response to environmentally imposed energy deficits.

INTRODUCTION

Most work dealing with the evolution of cooperative behavior in spiders
involves the mechanisms by which this has been achieved; e.g., changes that occur
in spider behavior and patterns of association to permit communal living and the
sharing of resources (Shear 1970, Wilson 1971, Brach 1977, Buskirk 1981).
Equally important, however, are the ecological influences; external factors that
favor the development of cooperative traits. The study of these influences requires
field observation and experimentation. That the majority of the communal and
actually cooperative spider species are restricted to the lower latitudes, in fact,
suggests that some characteristic of the tropical environment is requisite to
interindividual tolerance and cooperative behavior in the Araneae which are, for
the most part, highly competitive and even cannibalistic.

In an attempt at identifying important parameters we have undertaken an
investigation of the behavioral ecology of the highly cooperative spider Agelena

'Presented in a symposium, “Social Behavior in Spiders,” at the National meeting of the American
Arachnological Society, June 17, 1984.

176

THE JOURNAL OF ARACHNOLOGY

consociata Denis (Agelenidae) which inhabits primary rain forests in equatorial
west Africa. Specifically, we consider its local distribution in the rain forest and
assess environmental effects on its success in colonies consisting of different
numbers of individuals.

AG ELEN A CONSOCIATA

Agelena consociata, one of two species of Agelena that inhabit the Old World
tropics, is a funnel web spider that occurs in groups of a few to hundreds of
individuals of all age classes (Chauvin and Denis 1965). These individuals share
a group nest and a web trap consisting of a flat sheet and attached vertical
scaffolding (Fig. 1). Unlike the solitary agelenids, there are multiple retreats,
rather than a single retreat, extending into the nest. There is variability both in
the composition of the individuals making up the nest and in its physical
structure. The latter is formed by the binding of branches and leaves with silk
(Pain 1964). Because groups of nests are frequently connected to one another by
scaffolding or part of the web sheet, and because individuals move freely among
these nests, we define a colony in this paper as consisting of one or more nests
in a local area.

Cooperation takes the form of group prey capture, web maintenance and
feeding (Krafft 1969). All larger individuals participate in construction of the web
trap and scaffolding, though the effort apparently is not a coordinated one
(Darchen 1965). Small prey are attacked by single individuals, whereas as many
as 25 spiders may be involved in the capture of larger prey that struggle violently
in the web (Chauvin and Denis 1965). Only one individual will drag a subdued
prey to a retreat. It is during transport that potential competition is observed,
with some pushing or shoving evident (our observations and Krafft 1969).
Although the winner of the pushing contest has the first opportunity to feed on
the prey, it frequently merely deposits the prey in the retreat and moves off
without feeding. Twenty to forty spiders may be observed simultaneously feeding
on a large prey item (e.g., Krafft 1969 and our observations). We have also
observed behaviors that suggest that adults regurgitate digested food to
spiderlings.

STUDY AREA

Agelena consociata is probably widely distributed throughout the rain forests
of equatorial west Africa, but its range is poorly known. The species has been
studied extensively only at a single locality near Makokou, in the Ogooue-Ivindo
region of Gabon (Darchen 1975, 1978, 1979, 1980, 1984, Pain 1964, Riechert
1985). Most of these investigations have been conducted at M’Passa, a field
station of the Institut de Recherche en Ecologie Tropicale (I.R.E.T.), which is
administered by the Centre National de la Recherche Scientifique et
Technologique of Gabon. M’Passa is located at approximately 00°34’N latitude,
12°50’ E longitude, on the west bank of the Ivindo River about 10 km southwest
of Makokou. The station, which lies at an average elevation of about 500 m, is
part of a 13,000 hectare United Nations Biosphere Reserve and is, thereby,
afforded at least nominal protection from human disturbance.

RIECHERT ET A L.— ECOLOGY OF AG ELEN A CONSOCIATA

111

Fig. 1. — Agelena consociata colony showing two nests (N) and associated web trap (W), scaffolding
(S), and nest retreats (R). Many more retreats are present than indicated.

With the exception of a laboratory clearing, the entire reserve is covered by
closed primary or old secondary evergreen rain forest. Our investigations were
conducted in a 140 hectare area of the reserve which has been divided into a grid
of one hectare quadrats, each 100 m on a side and bounded by paths about 1
m wide. The terrain is generally flat, but slopes gently from the northwest to the
Ivindo and Nyame Pende rivers to the southwest. Darchen (1980) identified
four types of forest within the grid system at M’Passa. Two of these can be
designated as layered forest and two as unlayered. The layered forests are
distinguished by differing heights of the understory, and the unlayered forests by
differences in the density of shrubs and vines under the tree canopy. A profile
of the annual climatic cycle at M’Passa is presented in Figure 2, together with
seasonal trends in insect biomass derived from those reported by Charles-
Dominque (1977) and our own data. Precipitation at M’Passa is very unevenly
distributed over the year and is the basis for recognizing four seasons (Charles-
Dominque 1977, Cruiziat 1966, Hladik 1978). The area receives an average of
1691 mm of precipitation annually, nearly 40 per cent falling during the period
September-November. This is the major wet season. It is followed by a minor
dry season, December-February, which is characterized by less precipitation and
maximum insolation. The minor wet season of March-May is a time of
“important rainfall and tropical storms” (Hladik 1978). Finally, the months of
June-August constitute the major dry season with minimum precipitation (less

178

THE JOURNAL OF ARACHNOLOGY

PRECIPITATION

%

100

75

50

25^

f " » '

RELATIVE HUMIDITY

Ma>

J FMAM J JASOND

JFMAMJ JASOND

MONTH

Fig. 2. — Seasonal variation in insect biomass and mean monthly climatic variables at M’Passa.
Climatic data are from records of the Gabonese Meterological Service station at Makokou. Insect
biomass is expressed as (bars) monthly means (except December) of grams dry weight of daily light
trap collections (from Charles-Dominque, 1977, Fig. 10), and as (triangles) milligrams dry weight per
trap per day (this study; see text for details).

than 5 percent of the annual mean), but also minimum temperatures and
insolation. Because of the constantly overcast skies, relative humidity remains
high and evaporation low during this season. As might be expected at a locality
so near the equator, monthly variation in temperature is slight, mean monthly
temperatures ranging from 21.7°C in June to 25° C in April. Agelena consociata
is active throughout the year at this locality.

RIECHERT £r/4L.— ECOLOGY OF AGELENA CONSOCIATA

179

METHODS

Colony Structure and Habitat Association. — Three 9 hectare study areas were
established within the reserve (Fig. 3). Two of the locations were selected to
provide data on spider habitat associations in layered and unlayered forest types.
The third area is representative of the habitat used by Darchen in his 1980 study
of the distribution of Agelena consociata in the reserve. Each study area was
inspected for colony locations by walking all 100 m transect lines within the areas
and traversing each one hectare quadrat within each area at 30 m intervals. The
locations of all nests in the plots were first mapped in February 1982. The
following measurements (in cm) were taken on each nest; height off the ground,
nest volume (maximum height by length by width), web sheet area (maximum
length by width), vertical scaffolding height and distances to other nests within
the colony. Twenty-one nests were collected off of the reserve for use in
regressing nest volume against spider numbers. The age and sex of each
individual were tallied as they were removed from the nests. We used the resulting
regression coefficients in estimating individual numbers in nests censused on the
study areas where destructive sampling was not possible.

In addition to the web structure measurements, the following habitat features
were recorded at 10 cm intervals along a 2 m transect beneath each web: the
presence of leaf litter under the nest, the presence of tree cover, the presence of
vegetation within 1 m above the nest, the presence of non-herb vegetation below
the nest, the presence of narrow (< 10 cm), medium (10-20 cm) and broad leaves
(> 20 cm) below and above the nest, numbers of leaf layers below the nests,
numbers of branches [narrow (< 1 cm), medium (1-4 cm), wide (4.1-16 cm) and
giant (> 16 cm)], below the nest and the presence of herbs. The transect was
oriented such that its 50 cm mark was positioned at the center of the nest and
it paralleled the longest axis of the web trap. The distance from the nest to the
canopy was estimated using a rangefinder.

For each nest within the two main study areas, a random site within the reserve
was located by picking random coordinates. The line intercepts were repeated at
each of these sites using the compass orientation determined for the line intercept
of the actual site and the respective web heights in the positioning of the sample.
These sites and their corresponding nest sites were not treated as paired samples
in subsequent analyses. Multiple discriminant analyses were applied to the
transect data to determine to what extent the habitat characteristics of nest sites
were representative of the general habitat and how habitat utilization might vary
with forest type.

Environmental Correlates. — Three additional environmental correlates were
considered: solar radiation, precipitation, and prey availability. At all nest sites
and their corresponding random sites within the two main study grids, solar
radiation striking the web at hourly intervals was scored at 0, partial, or total,
through visual censusing. In addition, precipitation readings were taken at ail
nests and associated random locations that were < 3 m in height. Rain gauges
were placed directly above or adjacent to the nests and at comparable nest
heights in the random sites. These precipitation estimates were referenced to those
provided by a rain gauge placed in the laboratory clearing. This allowed us to
use the station’s precipitation records in estimating the number of days per rainy
season in which web-damaging rains occur.

180

THE JOURNAL OF ARACHNOLOGY

Fig. 3. — Distribution of colonies of Agelena consociata in study plots at M’Passa at the initiation
of censusing in February 1982. Numbers refer to numbers of nests in the respective colonies. Plot
1 is in the center, 2 in the upper right, and 3 in the lower left of the figure.

Five nest sites, each from a different colony, were selected outside the study
areas for prey availability determinations. The nests were removed and sticky
traps measuring 6,525 cm^ were substituted. The trap consisted of horizontal and
vertical sheets of hardware cloth approximating the linear dimensions of the web
sheet and scaffolding, respectively, of the average A. consociata web trap. The
entire surface was coated with a thin coat of a sticky tree banding compound
(Stikem Special: Michel Pelton Co.). Random trap locations corresponding to
each actual nest-site were chosen as in the vegetation analyses. Potential prey
were collected from the traps each morning during the course of the censusing,
which was carried out in each of the four seasons. All prey caught on the traps
were scored as to Order and size class (0-4, 4.1-8, 8.1-12, etc. in mm). The
contents of each trap were placed in a small cloth bag and oven dried at 40° C
for 24 hours (The sticky material evaporated during the course of the drying
process). The samples were then weighed. The number of days over which the
sampling was completed for each season was determined by setting the standard
error of mean number of prey at ten percent and solving for number of days (13-
14 days in different seasons).

Extinctions. — During the course of each of 4 two month field seasons the fate
of each nest in the grids was followed through weekly censuses. The sampling
periods were; February-March (Minor Dry Season), 1982; June-July (Major Dry
Season), 1982; October-November (Major Rainy Season), 1983; and April-May
(Minor Rainy Season), 1984. As new nests were found during the course of
censusing, web and habitat measurement were collected as described for the initial
sample. At the beginnings and ends of each 2 month census period, nest volume

RIECHERT ET A L.— ECOLOGY OF AG ELENA CONSOCIATA

181

and web area measurements were repeated. In addition to the weekly censuses,
nests below 3 m in height were checked for web sheet and scaffolding damage
following rains during the Minor Wet Season of 1982.

Energy Budget. — A combination of field and lab studies were used to complete
comparative energy budgets for individuals at small (1-4 adult females) versus
large (>25 adult females) nests. We used sticky trap collections of potential prey
in estimating energy availability. The data collected to show seasonal variability
in prey availability were used in the case of the large nests. Using the same
methodology, sticky traps providing a surface area of 484 cm^ (the average trap
area associated with small nests) were also censused for prey numbers and
composition.

Estimates of prey consumption rates under unlimited food, the frequency of
foraging activity, and energy expenditure in construction of the web trap were
obtained from laboratory studies. Individuals maintained in plastic boxes in
groups of one, two, four and six individuals, respectively, were offered an
abundance of prey (moths, flies and disabled crickets) on a daily basis. Record
was kept of individual rates of consumption, weight gain, and egg production
over a two month period. Approximately 50 individuals representing each class
were measured. In the second experiment, 25 individuals were weighed and placed
in 0.5m3 containers. After 24 hrs, the individuals were removed and reweighed.
The webs each had built during the period were also collected and weighed. The
third experiment consisted of eighteen days of observation of the foraging activity
of five captive groups of individuals. Each group consisted of 25 individuals, 15
of which were adult females. All spiders were individually paint-marked using a
fast-drying enamel paint. Thirty minute watches were made of each group
following the introduction of prey once a day, and individuals active in web
construction, prey capture and feeding were noted.

Microbomb calorimetry was used in making mass conversions of joules and in
estimating what proportions of available prey were available for consumption by
the spiders. Twenty-five individuals of each of the major prey orders contacting
Agelena webs were collected, weighted, killed by freezing, dried as for the sticky
trap samples, and reweighed. Prey wet weight averaged 2.32 times dry weight.
The joule equivalent of 1 mg dry weight was 20.83 averaged over all prey types.
Finally, an average of 5.56% of the prey was ash. This quantity was subtracted
from the sticky trap estimates in calculating prey availabilities.

Population Structure. — During each of the four field study sessions, at least 20
nests were inspected to determine to what extent the age structure of Agelena
varies with season. Individuals were removed as the nests were dissected and were
tallied as to approximate age (spiderlings, juveniles, penultimates and adults) and
sex.

There is no known record of active dispersal by Agelena consociata, nor did
we observe such a phenomenon during the course of our study. We did observe
the destruction of nests by rain, falling objects and animal movements (birds and
mammals). A release experiment was thus performed to assess the survivorship
of individuals that lose contact with their nests during such stochastic events. In
the experiment, all preexisting Agelena nests within a 50 m radius of a chosen
site were recorded and marked with plastic flagging. Spiders were collected from
five different nests from other areas of the forest, were paint marked and then
released using the following protocol: 1) Forty-eight spiders were released

182

THE JOURNAL OF ARACHNOLOGY

individually and their movements followed until each had moved out of the
cleared release area; 2) The remaining spiders were released in groups of up to
five individuals. In the latter case individual movement within the release area
was not followed. Rather, a daily search was made for new nests built by released
spiders within the area. The location, size, and distance from the release site were
recorded for each new nest along with the identity of the resident spider.

Genetic relatedness of spiders within nests and colonies was assessed using
electrophoretic techniques. Twenty-five spiders were collected from each of one
to three nests/ colony for 30 colonies to allow estimation of degree of relatedness
for individuals within the same nest, the same colony, and separate colonies. The
collected spiders were subjected to starch gel electrophoresis in the Population
Genetics Laboratory at the University of Tennessee. Whole individual spiders
were ground using the methodology described in Selander et al. (1971). The gel
electrophoresis techniques used were similar to those described in the same paper.
The following loci were assayed: Esterases 1 and 2 (ESTl and EST2), Fumarase
1 and 2 (FUMl and FUM2), Galactosaminidase (GAM), Glutamic Oxaloacetic
Transaminase 1 and 2 (GOTl and GOT2), a-Glycerophosphate Dehydrogenase 1
and 2 (GPDl and GPD2), ;8-N-Acetylglucose-aminidase (HEX), Isocitrate
Dehydrogenase (IDH), Lactate Dehydrogenase 1 and 2 (LDHl and LDH2),
Malate Dehydrogenase (MDH), Malic Enzyme (ME), Mannose Phosphate
Isomerase (MPI), Octanol Dehydrogenase (ODH), Peptidase (PEP),
Phosphoglucose Isomerase (PGI), and Superoxide Dismutase (SOD). Of these,
three were polymorphic: PEP (three alleles), ESTl (three alleles), and EST2 (two
alleles).

RESULTS

Colony Structure. — Statistics pertaining to colony structure are presented in
Table 1 for the 29 colonies in existence at the time of the initial census (Fig. 3).
Colonies frequently were represented by more than one nest, and in 40 per cent
of these multiple cases nests were interconnected by either a shared web sheet or
scaffolding. Because it was impossible to determine the number of individuals
occupying different nests without destruction of the nest, a relationship was
established between the numbers of individuals and nest volume, using nests
collected off the reserve. We found that the highest correlation between nest
volume and colony size was achieved when only adult female occupants were
included (r^ = 96.8%). All of the colony size estimates used in this paper thus
were calculated from the following regression relationship:

No. of Adult Females = 0.0012 Nest Volume (cm^) — 4.13

Habitat Association. — The results of the discriminant analyses comparing the
vegetation characteristics of actual nest sites with random sites are presented in
Figure 4. Colony habitat associations are apparently non-random, differing
markedly in character from the general habitat available. Inspection of the
discriminant function coefficients representing each habitat variable indicate that
the nests tend to be constructed above multilayered shrubs under a full tree
canopy, but without branches immediately overhead (i.e., within 1 m).

RIECHERT £r/4L.— ECOLOGY OF AGELENA CONSOCIATA

183

Table 1. — Basic colony statistics for 29 colonies of Agelena consociata

COLONY STRUCTURE

Mean

Standard Error Range

Number Nests/ Colony 5.3

1.6

1-27

Nest Volume (cm^)

53094.2

879.4

120-1,848,000

Web Area (cm2)

5726.9

129.1

192-63,000

Scaffolding Height (cm) 131.4

0.7

0-550

Adult Females/ Nest

* 16

14-22

0-2213

* Median and 95% confidence interval based on

regression relationship:

No. - 0.0012 (Nest Volume)-

4.13. (r = 0.958)

NEST HEIGHT

Study Area

Forest Type

Nest Height

Mean

Standard Error

Plot 1

Layered, Low Understory

163.9

4.3

Plot 2

Unlayered, Lianas

287.6

1.7

Plot 3

Layered, High Understory

410.8

6.4

Nest height varies with colony location in the forest (Table lb). Nests are
highest in Plot 3 (Fig. 3) characterized by understory with little shrubbery. Nests
are lower in forest in which the understory is in the recovery phase (Plot 1) and
intermediate in the closed forest characterized by lianas and fallen trees (Plot 2).
Despite the differences in nest height, spider use of habitat appears to be similar
in the three habitats. Of particular interest is the comparison between spider
utilization of habitat features in our first two study plots with the section of
forest lacking an understory (Plot 3) which was characterized by Darchen (1980)
as prime Agelena habitat. The nests measured in this third grid were not included
in the initial discriminant analyses. Rather, scores for each nests were calculated
post facto using the discriminant coefficients derived from the analysis for each
habitat feature measured. Discriminant scores for nests censused in Plots 1 and
2 did not differ from those for the higher nests (Mann Whitney test, P > 0.25).

The vegetation features utilized by Agelena consociata might be predicted on
the basis of web structure alone, because the nest is constructed of curled
branches and leaves and must be supported by some underlying structure. In
addition, the vertical scaffolding requires an empty space but with points of
attachment present. There are, however, possible additional influences on habitat
association. Incident precipitation at actual nest sites, for instance, is significantly
less than that encountered at the corresponding random sites in the reserve (N
= 46, Sign Test: P < 0.001; median ratio of precipitation at nest sites to random
sites = 0.61). The variance in rainfall at random sites does not differ significantly
from that characteristic of nest sites (Siegel Tukey Test: P < 0.35). In one season,
at least, prey numbers and biomass are significantly higher at nest sites than at
random sites (Table 2). Solar radiation reaching the nests was estimated by
visiting each nest and associated random site once during each of 9 daytime hour
intervals (clear days only). The differences between light counts at nest sites and
associated random sites was highly significant (Sign Test: P < 0.0001), with more
direct solar radiation hitting actual nests (Mean = 3.12 ± 0.02 intervals) than
random sites (Mean = 1.94 + 0.02).

Extinctions. — The marked variability in colony sizes and nests sizes in
particular suggests a corresponding variability in colony success in the rain forest

184

THE JOURNAL OF ARACHNOLOGY

Fig. 4. — Results of discriminant analyses of habitat variables at Agelena consociata nest sites and
random sites in the original study plots (two upper right plots, Fig. 3). Scores represent sample
placement on the basis of vegetation characteristics at a particular site.

habitat. Inspection of the nest census data collected over the two year period of
this study shows that smaller nests went extinct significantly more often than did
larger nests (Mann Whitney Test: P < 0.0001). The median number of adult
females occupying the 37 nests that went extinct was 1 with a 95 per cent
confidence interval of 0-4 adult females/ nest. The median number of adult
females occupying nests that were not lost during the censusing was 16 with a
95 per cent confidence interval of 14-22. Most of these extinctions occurred
during the two wet seasons, as is indicated by the nest numbers present at the
end of the respective seasons (Table 3). This is despite the fact that larger nests
tended to receive more precipitation than did smaller nests (r = 0.38; P < 0.05).
Our nest and web censuses following rains in which precipitation levels were
recorded both at the sites and in a nearby clearing shows that at least 50 per
cent of the web trap and scaffolding are destroyed at 84 per cent of the nest sites
when 6 mm of precipitation is recorded at the clearing within a 24 hr period.
This quantity of rain is equivalent to between 20 and 40 mm of rainfall at the

RIECHERT ET/IL.— ECOLOGY OF AG ELEN A CONSOCIATA

185

Table 2. — Comparison of prey encounter estimates for paired Agelena consociata nest sites and
random sites over 4 seasons. Mean and standard errors mg dry weight/ dry. Probabilities refer to
results of sign tests. NS = Non-significant.

PREY BIOMASS

Season

Nest Sites

Random Sites

Probability

Major Wet

321.1 4.8

318.9

4.1

NS

Minor Wet

240.9 6.7

234.4

3.4

NS

Minor Dry

176.4 1.6

129.2

1.2

P<0.0001

Major Dry

48.0 0.7

30.6

0.5

NS

PREY NUMBERS

Major Wet

65.6

0.6

62.7

0.4

NS

Minor Wet

58.4

0.4

71.1

0.4

NS

Minor Dry

49.5

0.2

36.8

0.2

P<0.0001

Major Dry

38.1

0.2

39.6

0.2

NS

actual nest sites due to channeling by branches and foliage. Utilizing six years
of precipitation records for the clearing site on the reserve, we found that during
the two wet seasons totaling six months of the year, such web trap destruction
occurs on the average every two out of five days.

Energy Budget. — Solitary Agelena consociata expend an average of 360 ± 1.2
J per day in the capture of prey and in metabolic expenditures (N = 50
individuals measured over 60 days). An average of 6.5 ± 0.4 J is further put into
biomass and egg production per day when food is presented ad libitum. The
construction of a single layer of web trap costs an average of 366.8 + 2.8 J. Web
construction, then, doubles the daily energy expenditure of an individual spider.

Of the 52 nest sites censused for prey availability, only 27 per cent provided
the prey levels necessary to support web construction by a solitary individual at
the required frequency of two times/ five days. This is, in fact, an overestimate
because it assumes that a spider captures all prey that encounter its web.

By living in groups, Agelena consociata can overcome this limiting factor,
because web trap area in this social group decreases with increasing numbers of
individuals (Table 4). Thus, significantly smaller trap areas per individual are
associated with increasingly larger nests. Our activity census data indicate that
solitary individuals in captive nests were active in silk laying or prey capture in
every foraging period (median and 95 per cent confidence interval = 1), whereas
individuals belonging to nests containing 25 spiders were active only every third
foraging period (confidence interval = 3-4).

Population Structure. — The survivorship of Agelena consociata nests and
colonies may also be affected by population structure. Nest censuses show that
within the same season the proportion of adult males represented relative to
females varies considerably from nest to nest (range = 0-58 per cent of the adult

Table 3. — Number of Agelena consociata nests present in study grids at end of each season.

MONTHS

SEASON

NUMBER OF NESTS

Sept-Nov

Major Wet

52

Dec-Feb

Minor Dry

150

Mar- May

Minor Wet

82

June-Aug

Major Dry

144

186

THE JOURNAL OF ARACHNOLOGY

Table 4. — Change in web investment per adult Agelena consociata (cm^) with number of individuals
in a colony. Significant changes (Mann Whitney test) denoted by broken lines. (l=p < 0.05; 2=p <
0.001).

NUMBER OF INDI-
VIDUALS

MEAN

STANDARD ERROR

NUMBER OF NESTS

1- 5

930

421

43

ION

Ik

1

1

1

1

1

1

1

1

1

1

1

9

1

1

1

9

1

1

1

738

158

/

26-125

400

139

27

126-625

333

151

V. 2

626-3125

14

20

3

spiders present). Further, significantly more small nests (0-4 adult females) had
no males than larger nests (> 4 adult females: Chi square test, = 4.64, DF
= 1, P < 0.05). With such low numbers of males, smaller nests and single nest
colonies may become extinct because there is no sperm supply and hence no
production of offspring.

It is improbable that males immigrate from neighboring nests to supplement
a declining population. This conclusion is based both on the results of our release
experiments and genetic studies. After 4 weeks of following the release of 94
individuals onto the rain forest floor, a total of only 8 spiders had relocated and
all of these relocations constituted newly formed nests. Ten per cent of the 48
individuals whose dispersal was tracked after release, in fact, suffered predation
within the first hour after release. Electrophoretic analyses completed on these
populations also indicate that there is no migration between colonies. Nei’s (1972)
Genetic Identity scores (I) were calculated for attached nests (I = 0.9982 + 0.001)
and unattached nests (0.9982 ± 0.001) within colonies. Therefore, individuals in
nests within a colony are almost identical genetically and colonies may consist
of single families. Similarly high genetic identities among family groups have been
reported for other taxa (e.g., marmots, Schwartz and Armitage 1981). However,
nests separated by as little as 38 m (and occurring in different colonies) have
lower Genetic Identity scores (I = 0.9340 ± 0.005) and may be fixed for different
alleles at the same polymorphic locus.

Predation Efficiency. — One of the major explanations given for cooperative
behavior is the increased efficiency of feeding, particularly on large prey [See
Buskirk (1981) for review]. From the sticky trap data, however, it is apparent that
the majority of prey available to A. consociata are of smaller size classes, ones
that can be readily handled by solitary foragers (Fig. 5). Furthermore, we found
that food intake per spider decreases with group size (Fig. 6). This could reflect
an inhibitory effect of conspecifics on the feeding level of associated individuals
or it may indicate that as group size increases, individuals expend less energy and
thus require less food. In the same experiment, egg production rates decreased
with increasing group size (Fig. 7). This supports the first alternative: that
conspecifics inhibit the feeding activity of nest mates, an effect that increases with
group size (Fig. 7). Capture efficiency, then, does not appear to underly the
differential extinction of smaller nests.

RIECHERT ETAL.

ECOLOGY 0¥ AG ELEN A CONSOCIATA

187

Fig. 5. — Proportion of prey contacting Agelena consociata webs that can be readily captured by
single penultimate-adult spiders (prey lengths equal to or less than the body lengths of these spiders:
Riechert and Luczak 1982).

DISCUSSION

The behavior of an individual is governed by the degree of relative advantage
that behavior affords. This advantage is generally expressed in terms of individual
fitness or the number of surviving offspring an individual produces. Cooperation
in foraging, nesting, and defense against predators, however, does occur and the
origin of such behavior is difficult to explain by natural selection operating on
the individual. From a game theoretic view, if a population initially consists of
competitors, for instance, it is difficult for an altruistic mutant gene to invade,
because the payoff to the competitor when interacting with a cooperative
individual would always be higher (e.g., Axelrod 1984).

Three conditions have been identified as contributing to the evolution of
sociality: mutual benefit to the interacting individuals, kin selection, and
reciprocity. Under the first condition, cost/ benefit factors are such that it pays
individuals to interact in a cooperative manner. This would be the case if the
majority of prey available to A. consociata were larger in size than could be
readily captured by solitary individuals. [See Michener (1974) and West-Eberhard
(1975) for examples.] Under kin selection, relatives share more genes than the
population at large and thus, by aiding kin, individuals increase the survival of
copies of their own genes (Hamiton 1964a,b, Dawkins 1976). Reciprocity denotes
the exchange of altruistic acts occurring with a time lag: individual A benefits
B on day one and B benefits A at some later date (Trivers 1971). If there are
repeated interactions between individuals, and individual recognition in larger
groups, reciprocity can invade a competitive system, but only if the invasion is
by groups of individuals (Axelrod 1984).

Spiders are a particularly interesting group within which to examine the
underlying causes of cooperative behavior because the vast majority of the species
are highly competitive and even cannibalistic towards conspecifics. Why is the

188

THE JOURNAL OF ARACHNOLOGY

Fig. 6. — Results of Prey consumption feeding
experiments for individual Agelena consociata
maintained in cages housing indicated number of
individuals (N = ca. 50 individuals per category).
Bars represent mean total weight of prey consumed
per day; standard errors indicated by lines at tops
of bars.

exhibition of cooperative behavior limited in this group to the tropics and which
if any of the three conditions underlies it? Some clues to these problems are
provided in this ecological study of Agelena consociata.

A major limiting factor to the success of individual funnel spiders in the rain
forest habitat of Gabon is heavy rainfall. The energetic deficit incurred through
frequent web trap replacement during these periods is not balanced by food
intake at the majority of nest sites. Three adaptations might have been exhibited
by Agelena consociata in response to this energetic constraint. I) Spiders might
remain dormant during the rainy seasons. 2) Habitat selection might be refined
such that greater protection from rain damage is afforded. 3) Spiders might live
in family groups where individual energy expenditure in maintenance of the web
trap and other foraging activities is markedly reduced. The first two adaptations
are typically exhibited by temperate spider species, but, at least the first is not
particularly well suited to A. consociata. Spiders would have to be dormant
during six months of the year alternating three months of activity with three
months of dormancy. Besides the logistical problem associated with cuing activity
on and off on such a schedule when day lengths and temperatures vary little,
there is the additional fact that this dormancy would coincide with times of
maximum insect densities (Table 2). Spiders would, by necessity, be dormant
during periods when they could best maximize their intake of prey. (Note that
Agelena consociata is the only common sheetline weaver in this rain forest study
site.)

Agelena consociata exhibits a non-random association with habitat features,
one which does lend some protection from precipitation. The major obstacle to
increased use of habitat selection criteria to this species, however, is the
apparently high cost of dispersal. The release experiments demonstrate that A.
consociata moving across the forest floor suffer a high mortality rate to ants and

RIECHERT ECOLOGY OF AG ELEN A CONSOCIATA

189

0.7

Fig. 7. — Mean and standard error of egg
production (mg wet weight) per individual Agelena
comociata with respect to number of individuals in
feeding group (same experiment as depicted in Fig.
6).

Other invertebrate predators. All indications are that if dispersal does occur, it
is a rare event. We consider that the break up of colonies into smaller nests is
mainly achieved through the action of rain and falling limbs. Colonization, on
the other hand, probably involves passive transport of pieces of nests by bats that
live in them and by birds and mammals that may occasionally fly or walk
through them.

The long term cohesiveness of family groups is favored by the rain forest
environment for several reasons. First, the cost of dispersal is high. Second, the
apparent energetic advantage of construction of a group web trap overcomes the
limiting constraint of rain damage to nest success. These benefits apparently
outweigh the negative effects of group living on individual foraging efficiency
observed in our laboratory experiments. Finally, and what we consider key to the
tropical dilemma, is the fact that in this equatorial environment, year around
moderate temperatures allow a continuity of generations (Table 5) that is difficult
to achieve by the spiders in temperate environments. By Occam’s razor,
continuous colony function permits the evolution of cooperative behavior with
the minimum of adaptive modifications {sensu Wilson, 1975). It is this continuity

190

THE JOURNAL OF ARACHNOLOGY

Table 5. — Age class representation of Agelena consociata in nests by seasons.

PROPORTION OF INDIVIDUALS PROPORT. PROPORT.

NESTS

PER NEST NESTS BOTH AGES

PRESENT

WITH EGGS

SEASON ADULTS JUVENILES

MEAN

SE

MEAN

SE

Major Rain

0.12

0.03

0.88

0.03

—

1.0

Minor Dry

0.53

0.01

0.49

0.01

0.17

0.98

Minor Rain

0.57

0.02

0.26

0.03

0.33

1.0

Major Dry

0.63

0.01

0.36

0.01

0.93

1.0

that explains the fact that cooperative behavior is limited in spiders to tropical
areas.

Both mutual benefit and kin selection may underly the cooperative behavior
exhibited by Agelena consociata. An influence by reciprocity is less certain since
work by Krafft (1971, 1974) shows that there is no individual recognition, and
group sizes are frequently too large for the action of this phenomenon in the
absence of individual recognition. Delineation of the relative contributions of
mutualism and kin selection awaits further work, particularly with colonies of
intermediate-large size. Kin selection may not have been requisite to the
development of cooperative behavior in this species because of the marked
benefits associated with group living in the rain forest environment, but inclusive
fitness effects might have accelerated the development of the system.

ACKNOWLEDGMENTS

We thank Alain Pasquet and Deborah Smith for help with the field work and
the Centre National de la Recherche Scientifique et Technologique of Gabon for
the use of their facilities. We are especially indebted to Paul Posso, Director of
the Institute de Recherche sur I’Ecologie Tropicale of Gabon, for facilitating our
studies and making our brief visits to Libreville both pleasant and memorable.
Yael Lubin and Ann Rypstra provided helpful comments concerning a previous
draft of the manuscript. The study was completed under the auspices of two
grants from the Harry Prank Guggenheim Foundation.

LITERATURE CITED

Axelrod, R. 1984. The Evolution of Cooperation. Basic Books, New York.

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

Buskirk, R. E. 1981. Sociality in the Arachnida. Pp. 281-367, In Social Insects (H. R. Herman, ed.),
vol. 2. Academic Press, New York.

Charles-Dominque, P. 1977. Ecology and Behaviour of Nocturnal Primates. Columbia Univ. Press,
New York.

Chauvin, R. and J. Denis. 1965. Une araignee sociale du Gabon. Biol. Gabonica, 1:93-99.

Cruiziat, P. 1966. Note sur le microclimat de la strate inferieure de la foret equatoriale compare a
celvi d’une clairere. Biol. Gabonica, 2:361-402.

Darchen, R. 1965. Ethologie d’une araignee sociale, Agelena consociata. Biol. Gabonica, 1:117-146.

RIECHERT E:r^L.— ECOLOGY 0¥ AG ELEN A CONSOCIATA

191

Darchen, R. 1975. La foundation de nouvelle colonies d'Agelena consociata et d'Agelena republicana,
araignees sociales due Gabon. Problemes eco-ethologiques. C. R. Illeme Coll. Arachnol. Express,
franc., Les Eyzies. pp. 20-39.

Darchen, R. 1978. Les essaimages de I’araignee sociale, Agelena consociata Denis (Araneide,
Labidognathe), dans la foret gabonaise (III). C. R. Acad. Sc. Paris, 287:1035-1037.

Darchen, R. 1979. Relations entre colonies d’Agelenides sociaux du Gabon. Precision sur les
essaimages. 11. Bull. Biol. Francaise-Belgique, 117:3-29.

Darchen, R. 1980. Les populations d' Agelena consociata Denis, araignee sociale, dans la foret
primaire gabonaise. Leur repartition et leur densite. Ann. Sc. Nat., Zool., 2:19-26.

Darchen, R. 1984. Trois et apres .... Variations de la densite de la repartition des colonies de I’araignee
sociale Agelena consociata Denis dans la foret gabonaise. Rev. Arachnol., 5:185-191.

Dawkins, R. 1976. The Selfish Gene. Oxford Univ. Press, Oxford.

Hamiton, W. D. 1964a. The gentical evolution of social behaviour. 1. J. theor. Biol., 7:1-16.

Hamiton, W. D. 1964b. The genetical evolution of social behaviour. 11. J. theor. Biol., 7:17-32.

Hladik, A. 1978. Phenology of leaf production in rain forest of Gabon: Distribution and composition
of food for folivores. Pp. 51-71, In The Ecology of Arboreal Folivores (C. C. Montgomery, ed.).
Smithsonian Inst. Press, Washington, D. C.

Krafft, B. 1969. Various aspects of the biology of Agelena consociata Denis when bred in the
laboratory. American Zook, 9:201-210.

Krafft, B. 1970. Les rythmes d’activite d' Agelena consociata Denis: Activite de tissage et activite
locomotrice. Biol. Gabonica, 6:99-130.

Krafft, B. 1971. Les interactions entre les individus chez Agelena consociata, araignee sociale du
Gabon. Proc. 5th Internat. Congr. Arachnol., pp. 159-164.

Krafft, B. 1974. La tolerance reciproque chez I’araignee sociale Agelena consociata Denis. Proc. 6th
Internat. Congr. Arachnol., pp. 107-111.

Michener, C. D. 1974. The Social Behavior of the Bees. Belknap/ Harvard Univ. Press, Cambridge,
Massachusetts.

Nei, M. 1972. Genetic distance between populations. American Nat., 106:283-292.

Pain, J. 1964. Premieres observations sur especes nouvelle d’araignees sociales. Biol. Gabonia, 1:47-
48.

Riechert, S. E. 1985. Why do some spiders cooperate?, Agelena consociata, a case study. Insect
Behav. Ecol. symp. 1984. Florida Entomologist, 17:105-116.

Riechert, S. E. and J. Luczak. 1982. Spider foraging: Behavioral responses to prey. Pp. 353-385, In
Spider Communication: Mechanisms and Ecological Significance (P. N. Witt and J. Rovner, eds.).
Princeton Univ. Press, Princeton, New Jersey.

Schwartz, O. A. and K. B. Armitage. 1981. Social substructure and disperson of genetic variation
in the yellow-bellied marmot (Marmota flaviventris). Pp. 139-159, In Mammalian Population
Genetics (M. H. Smith and J. Joule, eds.). University of Georgia Press, Athens, Georgia.

Selander, R. K., M. H. Smith, S. Y. Yang, W. E. Johnson and J. B. Gentry. 1971. Biochemical
polymorphisms and systematics in the genus Peromyscus 1. Variation in the old field mouse.
Studies in Genetics VI (Univ. Texas Publ. 7103):49-90.

Shear, W. A. 1970. The evolution of social phenomena in spiders. Bull. British Arachnol. Soc., 1:65-
76.

Trivers, R. L. 1971. The evolution of reciprocal altrusim. Quart. Rev. Biol., 46:35-57.

West-Eberhard, M. J. 1975. The evolution of social behavior by kin selection. Quart. Rev. Biol., 50:1-
33.

Wilson, E. O. 1971. The Insect Societies. Belknap/ Harvard Univ. Press, Cambridge, Massachusetts.

Wilson, E. O. 1975. Sociobiology. Belknap/ Harvard Univ. Press, Cambridge, Massachusetts.

Manuscript received February 1985, revised November 1985.

‘s ': k! .

^ V'^'1

-* *y»ryih j

.n' ^.i.'t'

'-A , .r-/ m^. "

•< ■ ' Hfu^r ■ f' a^iKJlfi»-«^^^i'' ^

“-v .U -^VJJ N-' '^v;fj#*)^^e- 5W\^ ■■\

#*■ ■ ■

iy tf>v '* ,.■

Wnf.^'tr.'*r.<!»^|^, " ■ ■■' I

■’r«;i-i5 f.-'f'*;T 1?^ . ^Ti

i . L / «.•: v» y-vc.

, ^^.t43'w^teA

'j o 7«^'; ' ■•■’-^ f'l/ •■.' ' ®

--«’i ftjrit 71 - ’■• 'l-f- ' . .U

■■• V'i' M’ '/‘■^ A

■ * ' • • j- .. . . ■■" .;>^ !.i: : -..'O' ' j

'fir.M. i?»*h?,^’5i5

.,H . ■(*.,/ '« 5 '.*■ .r^Mfaiijl

V-^ *-,Vl. » ^ ^ r.T - ^

' ^./,T :•. "v-i .1' '.

,i7ii-'.i.

' ' ' YVf,0W|4in*‘S B^'k’ ’ .'■ ' <■; <1.> 'ii '■’ ’ *-/

■ V....»n-' -H' . i

■' ':#!■' ■'« = 'N‘'-»''fl||^ . ' ' rW^C '_ ^

’* * ■ ■■ V -^. ,vM . ’ ’■yV)

♦,-■'*• wi'i* <• * ‘f^>' ' j iiiUj^ilte! I •**"*’A 1^' -i^jj

' *^g' 1^^.' »

. ^ ' ‘ -^- :y} ;^»>*'..'». ■-■ '*>Wfi|''(3j / ''i^.

■ %-iKi O -'“a -

■^-'7 ^ -y .A’AS*

Rypstra, A. L. 1986. High prey abundance and a reduction in cannibalism; The first step to sociality
in spiders (Arachnida). J. Arachnol., 14:193-200.

HIGH PREY ABUNDANCE AND A REDUCTION
IN CANNIBALISM: THE FIRST STEP TO
SOCIALITY IN SPIDERS (ARACHNIDA)i

Ann Lundie Rypstra

Department of Zoology
Miami University
1601 Peck Boulevard
Hamilton, Ohio 45011, U.S.A.

ABSTRACT

Spiders of the species, Achaearanea tepidariorum (Araneae; Theridiidae), were held in enclosures
under two prey regimes. One group of 100 spiders was provided with 100 fruit flies (Diptera;
Drosophilidae; Drosophila melanogaster) per day and the second group of 100 spiders was provided
with 1000 fruit flies per day. The number of spiders in enclosures dropped during the first six days
in both groups. However, a higher rate of cannibalism in the low prey group caused the spider
numbers in that enclosure to drop more rapidly than the high prey group. A greater proportion of
the spiders in the low prey group settled into recognizable territories in the first six days than in the
high prey group. Spiders in the high prey group tended to move more and therefore encounter
conspecifics more frequently during observations. It is projected that tolerance, observed in these
normally solitary spiders when large amounts of prey are present, could lead to more complex
sociality if maintained under such conditions for an evolutionary period of time.

INTRODUCTION

Spatial patterns are a critical determinant of sociality in animals. If individuals
rarely, or never, come in contact with conspecifics then it is difficult to imagine
how sociality could evolve. Spiders are stereotyped as quick response predators
that are frequently cannibalistic. Given those characteristics, one would expect
that sociality would be extremely unlikely.

The spatial patterns displayed by many spider species appear to be strongly
influenced by the distribution of prey (Burgess and Uetz 1982, Rypstra 1983). In
species that maintain well-defined territories, those territories are small in
populations living with high prey availability relative to populations living where
prey are scarce (Riechert 1978, 1981, Uetz et al. 1982). In some natural
populations of solitary spiders, aggregations have been observed in association
with locally elevated insect abundances (Valerio and Herrero 1977, Honjo 1977,
Burgess an Uetz 1982, Rypstra 1985). In enclosures the number of web-spiders
that coexist is directly correlated with the amount of prey provided (Rypstra
1983). The fact that high prey levels can exert such control over the spatial

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

194

THE JOURNAL OF ARACHNOLOGY

patterns of solitary spiders makes it tempting to hypothesize that an abundance
of prey is prerequisite to the evolution of sociality in some species (Buskirk 1981,
Rypstra 1983, 1985).

The study reported here is an investigation of the behavior of a solitary spider
maintained in high densities within enclosures with a large amount of insect prey
available. In earlier experiments Rypstra (1983) found that intra-individual
tolerance increased and cannibalism decreased in several spider species when
maintained at extremely high prey levels in similar enclosures. Here, one of the
species that appeared tolerant, Achaearanea tepidariorum (Araneae, Theridiidae),
is observed more closely under two prey regimes. The goal of this study is to
determine more specifically what happens during the first few days within
enclosures as the spiders sort themselves out and an equilibrium density is
attained. Various behavior patterns including aggressive interactions and
cannibalism are quantified to gain a better understanding of how the high
densities are maintained.

METHODS

Mature or pentultimate females of Achaearanea tepidariorum were collected
around homes and buildings in the city of Hamilton, Butler Co., Ohio U.S.A.
All spiders were between 4. 5-5.0 mm in length. Prior to experimentation,
individuals were held in six dram vials for 48 hours during which they were
provided with water but no food. At the beginning of an experiment 100 spiders
were released into an enclosure measuring 2.5 x 2 x 2 m. Laboratory-raised
Drosophila melanogaster (Diptera, Drosophilidae) were released into these cages
each day at 1 100 h as prey for the spiders. In one enclosure 100 Drosophila were
released and in the second 1000 Drosophila were released daily.

The animals were observed during a three-hour period (1300-1600 h) each
afternoon for six days. At the end of each two-minute interval I recorded the
number of spiders involved in various activities. Special attention was paid to the
proportion of spiders participating in prey capture, feeding, spinning, and
interactions with other spiders. After the observation period, I removed any prey
remains and debris in the webbing and on the floor of the cage. It was possible
to sort out the dead spiders from the dead Drosophila. In addition I was able
to discern whether a particular item had been fed upon by the presence of silk
wrapping and a characteristic shrunken appearance.

RESULTS

In both experimental groups the total number of spiders alive in the enclosures
dropped during the six-day period (Fig. 1, Table 1). Significantly more spiders
were lost in the low prey treatment than in the high prey treatment (Wilcoxon
Paired Comparisons, p < 0.05) (Fig. 1, Table 1).

In both treatment groups about 13% of the spiders were involved in spinning
or web maintenance during observation periods (Chi-squared Test, p > 0.05)
(Table 2). However, more prey capture and feeding activity took place in the
group provided with 1000 flies per day versus the group with 100 flies available
(Chi-squared Test, p < 0.05) (Table 2).

RYPSTRA— PREY ABUNDANCE AND TOLERANCE AMONG SPIDERS

195

1 r r 1 — I 1

1 2 3 4 5 6

DAY

Pig \ — Total number of spiders living in enclosures maintained at two prey densities for six days.

I recovered close to 90% of the prey items over the course of the experiment
(Table 1). In the group which had been provided 1000 flies/ day, the number of
flies that appeared consumed dropped during the six-days of the experiment
(Kendall’s Tau = —0.87, p < 0.05) (Table 1). In the other group the number
consumed remained approximately the same over the whole period (Kendall’s Tau
= 0.20, p > 0.05) (Table 1). Of the spiders lost about 86% of them appeared to
have been cannibalized in both groups which translates into a higher number

Table 1. — Numbers of A. tepidariorum and the prey they ate during a six-day period in which they
were maintained in enclosures at either high prey (1000 fruit flies per day) or low prey (100 fruit flies
per day) densities.

Day

Number of
Living Spiders

Number in

Territories

Number Spiders
Cannibalized

Number of
Prey Recovered

Number of
Prey Fed Upon

Low Prey Group

1

91

0

4

96

74

2

78

0

6

90

79

3

67

3

15

89

80

4

40

4

21

97

76

5

25

10

18

98

82

6

19

14

6

95

76

Cumulative

Total

19

14

70

565

467

High Prey Group

1

90

0

3

911

672

2

82

0

4

856

714

3

76

1

3

898

628

4

70

2

7

932

588

5

66

2

7

842

487

6

61

4

6

927

431

Cumulative

Total

61

4

30

5366

3520

196

THE JOURNAL OF ARACHNOLOGY

Table 2. — Percent of three hour observation times that spiders in enclosures were engaged in the
activities of prey capture, feeding, web spinning and maintenance. All other behaviors including
agonistic interactions, simple movement and holding still in web were lumped in the “other” category.

Day

Percent Time Spent
Spinning

Percent Time Spent
in Prey Capture^

Percent Time Spent
Feeding^

Percent Time Spent
in Other Activities^

low prey

high prey

low prey

high prey

low prey

high prey

low prey

high prey

group

group

group

group

group

group

group

group

1

12.8

13.7

25.6

32.5

15.7

28.2

45.9

25.6

2

19.2

15.8

20.1

33.1

18.9

44.7

41.8

6.4

3

14.2

14.5

19.8

34.8

17.3

39.2

48.7

11.5

4

10.2

12.1

20.2

38.1

31.8

38.8

37.8

11.0

5

13.2

12.5

28.4

37.8

34.2

43.1

24.2

6.6

6

11.1

11.7

26.2

36.1

30.1

35.2

32.6

17.0

Total

13.4

13.4

23.4

35.4

24.6

38.2

38.5

13.0

'“A significant difference was found between low prey and high prey groups for these categories (p
<0.05).

cannibalized in the low prey treatment than in the high prey treatment
(Wilcoxon, p < 0.05) (Table 1).

Spiders in both experimental groups interacted with one another regularly
(Table 3). In a typical encounter one spider would orient toward the other with
its first two pairs of legs extended and give the webbing one or two firm jerks.
The approached spider usually reacted with a similar action. Interactions could
intensify with a series of such exchanges for up to three minutes. Most
interactions, however, ended after two to three exchanges spanning only a 40 to
80 sec time range (Table 3). Such bouts usually terminated with the retreat of
one of the spiders. In the spider group maintained with 100 flies/day the
aggressor was usually the one that retreated (Table 3). In the group maintained
at 1000 flies/day, the recipient of the aggression was more likely to retreat (Table
3). In about 10% of all such encounters one spider actively chased the other
away. Four observed aggressive bouts in the low prey experiment ended in
cannibalism (Table 3). Only one bout ended in cannibalism in the high prey
treatment group (Table 3). Significantly more interactions took place between
conspecifics in the high prey experiment than in the low prey experiment
(Wilcoxon, p < 0.05) (Table 3). Presumably a portion of this difference results
from the greater number of individuals in one group. However, even if the
number of interactions is standardized for the number of spiders present, each
group has a distinct level of aggressive interactions (Wilcoxon, p < 0.05) (Figure
2). This difference is more pronounced after the third day of each experiment
(Fig. 2). During more than 80% of the interactions a prey item was involved
(Table 3). Several interactions in the high prey treatment ended with the fly
abandoned by both spiders, whereas on only one occasion was a prey abandoned
in the low prey treatment group (Table 3). Interchanges observed were slightly
longer in duration in the low prey group than they were in the high prey group
(Table 3).

As the days progressed more of the webs characteristic of this species appeared
in the enclosures (Table 1). Although the enclosure into which the most prey were
provided filled up with webbing more quickly than the other, the low prey
enclosure contained more distinct webs than the other (Table 1). In addition,
single individuals were observed occupying the webs in the low prey group for

RYPSTRA— PREY ABUNDANCE AND TOLERANCE AMONG SPIDERS

197

Table 3. — Agonistic encounters between A. tepidariorum individuals in enclosures provided with
low prey (100 fruit flies per day) and high prey (1000 fruit flies per day).

Parameter

Low Prey Group

High Prey Group

Number of

Encounters Observed®

72

176

Number Encounters

Involving Prey Item®

60 (83%)

141 (80%)

Number in Which Prey

Item was Abandoned'’

1 (1%)

3 (2%)

Number in Which

Aggressor Retreated®

63 (88%)

60 (34%)

Number in Which

Recipient Retreated®

12(17%)

116(67%)

Number ending
in Cannibalism'’

4 (6%)

1 (0.6%)

Mean Duration
in seconds (range)®

58.4 (29.2-210.3)

41.5 (13.1-72.4)

^A significant difference was found between low prey and high prey groups for these categories (p
< 0.05).

'’There was insufficient data to test for a difference in these categories.

a two to three day span of time. Whereas spiders maintained in the high prey
container moved in and out of spaces in the webbing approximately every 21
min, so that it was not possible to identify specific territorial boundaries.

DISCUSSION

These results further document an increase in tolerance for conspecifics
displayed by the spider species, Achaearanea tepidariorum, at high prey densities
that has been reported before (Rypstra 1983). This species reduces the amount
of cannibalism, becomes non-territorial, and changes the nature of its interactions
with'^conspecifics when plenty of food is provided. Under high prey conditions

0.2 0-1

1 \ 1 1 1 1

1 2 3 4 5 6

DAY

Fig. 2. — Number of aggressive interactions observed per spider per observation hour. Interactions
took place among individuals maintained in enclosures at two prey densities for six days.

198

THE JOURNAL OF ARACHNOLOGY

the behavior of this species becomes very similar to that of regularly social species
that inhabit tropical regions (Buskirk 1981, Rypstra pers. obs.).

Prey abundance has been implicated in evolution of sociality in spiders in many
cases. Aggregations of the orb-weaver, Nephila clavipes (Araneae, Araneidae),
form around areas of high prey abundance in Peru (Rypstra 1985). Some species
that build big web conglomerates do so to take advantage of high insect densities
(Lubin 1973, 1974, Buskirk 1975, Rypstra 1979, Uetz et al. 1982). In addition
there are numerous reports of solitary species in temperate regions clustered in
areas where insects are abundant (Honjo 1977, Valerio and Herrero 1977, Burgess
and Uetz 1982). These examples indicate that high prey abundance is important
in the evolution of sociality via the parasocial route (sensu Wilson 1971). The
alternative route to sociality in spiders involves an extension of the social
tendencies frequently displayed by juveniles (Shear 1970, Buskirk 1981). Krafft et
al. (1986) have been able to prolong the juvenile social period by providing the
young with an abundance of food. Their experiments imply that food abundance
could also be an important aspect in the evolution of sociality via this second
pathway. Further evidence is provided by the geographical distribution of social
spider species. All species are tropical or subtropical (Buskirk 1981), and live in
areas characterized by consistently high insect abundances (Janzen 1973, Janzen
and Pond 1975, Rypstra 1986).

One of the functions of territoriality is to insure access to resources (Brown
1964, Morse 1980). When food is super-abundant and resources are not limited,
territorial behavior may break down. Carpenter and MacMillan (1976) developed
a model permitting them to predict when nectar-feeding birds would shift from
territorial to non-territorial states. The birds became non-territorial when resource
availability was high in a manner similar to that displayed by the spiders in this
study. When resources are super-abundant, no advantage is gained by defending
a specific area that functions to guarantee access to those resources (Brown 1964).

In the high prey group the frequency of interactions between pairs of
individuals is higher than in the low prey group (Fig. 2). Some part of the
difference is due to the fact that fewer of these individuals settled into territories
and more of them moved throughout the enclosure (Table 1). A similar difference
has been observed in some studies of bird flocks. Pulliam et al. (1974) reported
that individuals in flocks showed less aggression as conditions deteriorated. The
birds accomplished this by reducing the number of encounters per individual and
maintaining a defined inter-individual distance. This pattern is similar to what I
observed the spiders doing in this experiment.

In part, the high frequency of interactions between conspecifics in the high prey
treatment represents a stereotyped orientation response that they have to any
vibration in their web. The number of fruit flies and the number of spiders
moving throughout the intertwined webbing would naturally increase the number
of times the spiders would have to orient and respond. The reduction in the
intensity of the interactions is presumably because the spiders are not hungry and
therefore are not driven to escalate the interactions. Some highly social
cooperative spider species have a high rate of encounter interactions as they move
through the webbing (Vollrath and Rohde-Arndt 1982, Rypstra pers. obs.), which
operates as a form of communication network (Krafft 1982). With time these
encounters which now appear aggressive could ameliorate into a more standard
form of communication between individuals within a colony as sociality evolves.

RYPSTRA— PREY ABUNDANCE AND TOLERANCE AMONG SPIDERS

199

ACKNOWLEDGMENTS

Funding for this research was provided by the American Museum of Natural
History, Miami University (Hamilton Campus), and the Faculty Research
Committee of Miami University (Oxford Branch), D. H. Morse, E. J. Mury
Meyer and G. Stratton provided helpful comments on an earlier draft of this
manuscript. Thanks are extended to G. W. Uetz for organizing the symposium
that generated this manuscript.

LITERATURE CITED

Brown, J. L. 1964. The evolution of diversity in avian territorial systems. Wilson Bull., 76:160-169.

Burgess, J. W. and G. W. Uetz. 1982. Social spacing strategies in spiders. Pp. 317-351, In Spider
Communication: Mechanisms and Ecological Significance (P. N. Witt and J. S. Rovner, eds.).
Princeton University Press, Princeton, New Jersey.

Buskirk, R. E. 1975. Coloniality, activity patterns and feeding in a tropical orb-weaving spider.
Ecology, 56:1314-1328.

Buskirk, R. E. 1981. Sociality in the Arachnida. Pp. 281-367, In Social Insects. Vol. II (H. R.
Hermann, ed.). Academic Press, New York.

Carpenter, F. L. and R. E. MacMillan. 1976. Threshold model of feeding territoriality and test with
a Hawaiian honeycreeper. Science, 194:639-642.

Honjo, S. 1977. Social behavior of Dictyna follicola Bos et Str. (Araneae: Dictynidae). Acta
ArachnoL, 27:213-219.

Janzen, D. H. 1973. Sweep samples of tropical foliage insects: effects of seasons, vegetation types,
elevation, time of day, and insularity. Ecology, 54:687-708.

Janzen, D. H. and C. M. Pond. 1975. A comparison, by sweep sampling, of the arthropod fauna
of secondary vegetation in Michigan, England, and Costa Rica. Trans. Roy. Entomol. Soc.
London, 127:33-50.

Krafft, B. 1982. The significance and complexity of communication in spiders. Pp. 15-66, In Spider
Communication: Mechanisms and Ecological Significance (P. N. Witt and J. S. Rovner, eds.).
Princeton University Press, Princeton, New Jersey.

Krafft, B., A. Horel and 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.

Lubin, Y. 1973. Web structure and function: the non adhesive orb- web of Cyrtophora moluccensis
(Doleschall) (Araneae: Araneidae). Forma et Functio, 6:337-358.

Lubin, Y. 1974. Adaptive advantages and the evolution of colony formation in Cyrtophora (Araneae:
Araneidae) Zool. J. Linnean Soc., 54:321-339.

Morse, D. H. 1980. Behavioral Mechanisms in Ecology. Harvard University Press, Cambridge Mass.

Pulliam, H. R., K. A. Anderson, A. Misztal and N. Moore. 1974. Temperature-dependent social
behaviour injuncos. Ibis, 116:360-364.

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. 1981. The consequences of being territorial: spiders, a case study. Amer. Nat., 1 17:871-
892.

Rypstra, A. L. 1979. Foraging flocks of spiders: a study of aggregate behavior in Cyrtophora citricola
Forskal (Araneae; Araneidae) in west Africa. Behav. 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, 59:312-316.

Rypstra, A. L. 1985. Aggregations of Nephila clavipes (L.) (Araneae, Araneidae) in relation to prey
availability. J, ArachnoL, 13:71-78.

Rypstra, A. L. 1986. Web-spiders in temperate and tropical forests: relative abundance and
environmental correlates. Amer. Midland Nat., 115:42-51.

Shear, W. A. 1970. The evolution of social phenomena in spiders. Bull. British ArachnoL Soc., 1:65-
76.

200

THE JOURNAL OF ARACHNOLOGY

Uetz, G. W., T. C. Kane and G. E. Stratton. 1982. Variation in the social grouping tendency of a
communal web-building spider. Science, 217:547-549.

Valerio, C. E. and M. V. Herrero. 1977. Tendencia social en adultos de la Arana Leucauge, sp.

(Araneae, Araneidae) en Costa Rica. Brenesia, 10/ 1 1:69-79.

Vollrath, F. and D. Rohde- Arndt. 1983. Prey capture and feeding in the communal spider Anelosimus
eximius Z. Tierpsychol., 61:313-324.

Wilson, E. O. 1971. The Insect Societies. Belknap Press, Cambridge, Massachusetts.

Manuscript received April 1985, revised August 1985.

Smith, D. R. R. 1986. Population genetics of Anelosimus eximius (Araneae, Theridiidae). J.
Arachnol., 14:201-217.

POPULATION GENETICS OF ANELOSIMUS EXIMIUS
(ARANEAE, THERIDIIDAE)i

Deborah R. R. Smith

Department of Entomology
Cornell University
Ithaca, New York 148532

ABSTRACT

Anelosimus eximius is a cooperative, group-living neotropical spider. Colonies consist of up to
several thousand individuals, and colonies may be aggregated into local colony clusters. The colony
clusters are patchily distributed, and are often separated from their neighbors by a km or more. In
this study individuals were collected from colonies located in Panama and Suriname. These
individuals were subjected to horizontal starch gel electrophoresis and screened for polymorphisms
in 46 enzyme systems. A total of 51 scorable loci were found, of which seven were polymorphic. The
results were analyzed with Wright’s F statistics which were used to investigate the amount of genetic
differentiation in the population attributable to subdivision of the population into colonies, colony
clusters, local populations and the geographic regions of Panama and Suriname.

Most of the genetic differentiation in the A. eximius sampled was due to subdivision of the
population into colony clusters and into geographic regions. There was no evidence of differentiation
among colonies in a colony cluster, and little differentiation among collection sites within Panama
or Suriname. In contrast, within a local population, samples from adjacent colony clusters were
sometimes fixed for different alleles at one or more loci, and the Panama and Suriname samples were
fixed for different alleles at three loci.

INTRODUCTION

Anelosimus eximius (Keyserling) (Theridiidae) is among the best known of the
cooperative or quasisocial spiders. Its natural history has been studied most
recently by Brach (1975), Christenson (1984), Overal and Ferreira da Silva (1982),
Vollrath (1982), Vollrath and Rohde-Arndt (1983) and others (L. Aviles, Y.
Lubin, A. Rypstra, pers. comm.). Anelosimus eximius is found in rainforest and
second growth habitat from Panama to southern Brazil, and from Peru to
Trinidad and eastern Brazil (Levi 1963). The webs of A. eximius consist of a
large, more or less oval horizontal sheet of nonsticky silk, one or more retreats
made of green and dry leaves curled and held in place with silk, and vertical
threads (“knock down threads”) extending from the sheet and retreat to leaves
and branches above the web. Webs can be found from ground level to at least
20 m up in the forest canopy (personal observation). Webs range in size from
tiny structures 10-25 cm long containing only one or a few spiders, to large

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

^Current address: Museum of Zoology, Insect Division, University of Michigan, Ann Arbor, Michigan
48109

202

THE JOURNAL OF ARACHNOLOGY

colonies 2-3 m or more in length containing hundreds or thousands of
individuals.

Established colonies contain adult females and males, and immatures. The
social interactions of A. eximius have been described by Brach (1975),
Christenson (1984), Vollrath (1982), and Vollrath and Rohde-Arndt (1983). These
behaviors include cooperative construction, maintenance and cleaning of the web;
cooperative prey capture and feeding on large prey by several individuals; and
regurgitation feeding of young. Females regurgitate food for offspring other than
their own (Christenson 1984) and there is some evidence that some adult females
in colonies do not reproduce (Vollrath and Rohde-Arndt 1983; Overal and
Ferreira da Silva 1982). The adult sex ratio is highly biased towards females;
males make up only a small portion of the adult population of colonies: 5% to
22% (six colonies of 100 to 200 individuals; Overal and Ferreira da Silva 1982);
7% (4 of 55 adults in one web; Christenson 1984). The bias in sex ratio is already
apparent in the penultimate instar, when males and females can be distinguished
(Aviles 1983). Reproduction takes place year round, at least in some populations
(Overal and Ferreira da Silva 1982).

Two types of colony foundation have been described for this species: budding
and dispersal (Vollrath 1982). In budding a large established colony splits into
two or more webs, either because the web is broken by falling branches, debris
etc., or because small groups of spiders leave the main web and begin a new web
nearby (Overal and Ferreira da Silva 1982). During dispersal, large numbers of
mated females leave their original colony and disperse singly to build new solitary
webs. The females in the solitary webs may later be joined by other females,
apparently dispersing females whose new webs have failed. All newly founded
webs have a very high failure rate, but they stand a better chance for survival
if they are later joined by other females (Christenson 1984). Vollrath notes (1982)
that dispersing females always traveled alone, and only after the webs were built
did other females join. These multi-foundress colonies were generally in the
proximity of a larger established colony. Webs more distant from a large
established web were usually single female webs.

These two methods of colony foundation, budding and dispersal, may be the
cause of two superimposed patterns of colony distribution. Colonies or webs are
often found in local aggregations, or colony clusters which may contain two to
40 or more distinct webs. This pattern of distribution may be the result of
budding; if so, colonies within colony clusters should be genetically very similar.
The colony clusters themselves are patchily distributed, separated from
neighboring clusters by as little as a few tens of meters or as much as several
km. This may be partly due to new colony foundation by dispersal and founder
events.

Based on observations of natural history, it is likely that gene flow among the
colonies within a colony cluster is high. The webs are in close proximity,
sometimes touching, and may share parts of the knockdown threads. The extent
of gene flow among colony clusters is unknown. There are four potential avenues
for gene flow among colony clusters: 1) new colonies may be founded by
unrelated dispersing females; 2) males may disperse out of their natal colony
clusters and into different clusters to mate; 3) immatures or females may disperse
and join established colony clusters; and 4) spiders may be accidentally
transported to new webs (e.g. by wind, rivers, animals or humans).

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

203

Fig. 1. — Map of Browns Berg collection site, Brokopondo, Suriname. Solid lines indicate trails,
solid circles indicate Anelosimus eximius colonies or colony clusters, and solid diamonds are km
markers.

If colonies are founded by a single female lineage (one female or a group of
sibling females who mated before leaving their natal nest), and if there is no
subsequent migration into a colony by any age-sex class, then the individuals
within colonies and colony clusters would by necessity inbreed. In addition each
family lineage would be isolated from others in the population as it went through
cycles of dispersal, colony foundation, growth, budding and more dispersal.
Subdivision of the population into isolated genetic lineages may contribute to
more rapid or more frequent speciation than would be found in a more freely
mixing population.

Genetic variability at loci coding for structural proteins can be detected using
protein electrophoresis. This technique has been widely used in population studies
of both vertebrates and invertebrates (Selander and Whittam 1983, and references
therein). Pennington (1979) used protein electrophoresis to distinguish immatures
and females of similar species of Meta (Araneidae). The data presented here are
the results of a pilot study investigating the feasibility of using protein
electrophoresis to study genetic variation within and among colonies, colony
clusters, and populations of A. eximius. The questions addressed are:

1) Is there a high level of genetic similarity within colonies?

2) Is there a high level of genetic similarity within colony clusters?

3) Is there genetic differentiation among populations, and if so on what
geographic scale?

METHODS AND MATERIALS

Collection. — Anelosimus eximius were collected from colonies in Panama
(August 1983) and Suriname (April and May 1984). The Panama collections were
made in two locations. The first was along the El Llano — Carti Road in Panama
province east of the Panama Canal (9°15'N, 79°5'W), at an elevation of

204

THE JOURNAL OF ARACHNOLOGY

approximately 100-200 m. The area along the road was recently forested, but is
now mainly agricultural land and second growth forest. The second collection site
was outside of the town of El Valle in Penonome Province west of the Panama
Canal (8°37'N, 80°6'W). This area was higher in elevation (approximately 400-
500 m) in the bowl of an extinct volcano, and is covered with cooler, wet cloudy
forest. Here and at El Llano the spiders were collected by bagging a retreat and
cutting it out of the web.

Suriname collections were also made at two localities, the Browns Berg Nature
Park and the Voltzberg-Raleighvallen Reserve. The Browns Berg Reserve
(Brokopondo province, 4°50'N, 55°15'W) is located along the western shores of
the Brokopondo reservoir. The study area was on the Mazaroni plateau at an
elevation of 400-450 m. The vegetation here is rainforest which had been
selectively logged at some time in the past. Colony clusters were found along a
12 km trail across the plateau.

The Voltzberg-Raleighvallen reserve (Saramacca province,4°45'N, 56°10'W) is
located along the Coppename river in central Suriname. Park headquarters are
on Foengoe Island. Colonies were found on Foengoe Island, on the west bank
of the river (outside the reserve lands) and at the Voltzberg camp in the forest
east of the river. Both here and at Browns Berg spiders were collected by shaking
the web and catching the spiders in a bag or box as they jumped off the edge
of the web.

All spiders collected in Panama and Suriname were transported live to the
United States. The animals were starved for at least 1 week and stored frozen
at -70° C.

Electrophoresis. — Spiders were analyzed for enzyme polymorphisms using
horizontal starch gel electrophoresis. The techniques employed are described in
detail in May et al. (1979), May (1980), Harris and Hopkinson (1976), and
Brewer (1970); recipes for stains follow Harris and Hopkinson (1976) and Shaw
and Prasad (1970). Four buffer systems were used: “R” (Ridgway et al. 1970),
“C” (Clayton and Tretiak 1972), “M” (Markert and Faulhaber 1965), and “4”
(modified from Selander et al. 1971) (recipes for the buffers are given in
Appendix 1). Gels were made of 42 g (thin gels) or 70 g (thick gels) of a 1:1
mixture of hydrolyzed potato starch and electrostarch, and 300 or 500 ml
respectively of one of the four buffers. Individual spiders were homogenized in
2-3 drops of 0.05 M Tris HCl (Appendix 1). Heavy weight filter paper wicks
(Whatman #3 filter paper) approximately 0.5 x 3 x 8 mm were used to carry
samples of the homogenates; one spider provided enough homogenate for two to
four wicks. Running conditions followed those described in May et al. (1979) and
May (1980). After the proteins migrated a sufficient distance each gel was sliced
horizontally into four (thin gel) or six (thick gel) layers approximately 1.5 mm
thick and each slice was stained to indicate the position of a single enzyme.

A screen of 46 enzyme systems using each of the four buffer systems was
carried out to determine 1) which enzymes could be detected in A. eximius, 2)
which buffers gave the best results for each enzyme system, and 3) which enzymes
showed detectable polymorphisms. In the screen two thick gels of each buffer
system were run, for a total of 8 gels. Ten spiders were run on each gel; one
spider produced enough homogenate for four wicks, so that four replicates of the
ten spiders were placed on a gel. (Sometimes it was not possible to get four wicks
from a single homogenized spider, so samples from another ten spiders from the

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

205

Fig. 2. — Map of Voltzberg collection site, Saramacca, Suriname. Thick sold lines indicate trails,
solid circles indicate Anelosimus eximius colonies or colony clusters, and stippling indicates granite
outcrops in the form of bergs and plates. Colonies 8 and 9 are on Foengoe Island in the Coppename
River.

same web were used for one of the replicates). After the gels had been subjected
to electrophoresis they were cut into four “mini gels.” Each mini gel was sliced
horizontally into six layers, giving a total of 48 gel layers for each type of buffer.
One slice of each buffer type was stained for one of the 46 enzyme systems
investigated. The ten spiders in the replicates were chosen so that every buffer
and stain combination was tried on spiders from each of the four collection sites
and from as many different colonies as possible. The list of enzyme systems
screened and the buffers which gave best results is given in Appendix 2.

Based on the results of the screen, a survey of each colony was carried out
examining each apparently polymorphic locus on the buffer system giving the
best results. In the survey of polymorphic enzymes 7 to 14 spiders were sampled
from each of 23 webs for a total of 187 spiders. An additional 22 spiders were
sampled from one web in which heterozygotes were discovered (see results below).

Analysis. — Mean heterozygosity in the Anelosimus eximius samples was
calculated as the arithmetic mean of the heterozygosities calculated for every
scorable locus. Heterozygosity at a single locus was calculated as though all
members of the population were potentially freely interbreeding, so that
heterozygosity of a single locus with two alleles occurring with frequencies “p”
and “q” is 2pq; for three alleles with frequencies “p,” “q” and “r” it is 2(pq +
pr + qr), etc. I computed mean heterozygosities and standard errors according
to the method of Nei (1978) and Nei and Roychoudhury (1974) for populations

206

THE JOURNAL OF ARACHNOLOGY

t NP MUP ACP PGM EST4 ADH SODA

iL

ab cd ef ghijkl — pqr

t

Fig. 3. — Banding patterns observed for seven polymorphic enzymes, “o” is the point of origin on
the gel, and arrows indicate the direction of migration by the protein (+ towards positive terminal,
—towards negative terminal). The presumed genotype for each band is given in the arbitrary
numbering system used in the text (e.g. II, 12, 22) and in the percentage system described in
Allendorf et al. (1977). NP: a 1 1 or 100/100, b = 22 or 89/89; MUP: c = 22 or 158/158, d =
11 or 100/100; ACP: e = 1 1 or 100/100, f = 22 or 164/164; PGM: g = 11 or 100/100, k = 12 or
100/106.5, 1 = 22 or 106.5/106.5; ADH: m = 1 1 or 100/100, n = 22 or 42/42; SOD-4; O = 1 1 or
100/ 100, p = 33 or 107/ 107, q = 22 or 145/ 145.

consisting of the total A. eximius sample, the Panama and Suriname samples
separately, and for colony clusters which contained heterozygotes, using a
program written by Dowling and Moore (1984) modified for the IBM PC.

When colony clusters contained individuals heterozygous at a particular locus,
I calculated the frequencies of genotypes expected given the allele frequencies
calculated from the samples, and compared them with observed frequencies using
a Chi2 goodness-of-fit test.

I calculated Wright’s F statistics (Fdt, Fst and Fds) for each polymorphic locus
(Wright 1951, 1978; Hartl 1980). F statistics provide measures of the extent of
population structuring based on the departure of the frequencies of genotypes
observed from those expected under conditions of panmixia. Non-random mating
within populations and division of the population into sub-populations and
demes are examples of departure from panmixia which may lead to a reduction
in the observed frequency of heterozygotes compared to homozygotes. F statistics
are hierarchical in nature. Reduction in heterozygosity in the population as a
whole (Fdt), can be partitioned into that due to sub-populations within the total
population (Fst), and in demes within sub-populations (Fds). I performed two sets
of calculations; in both I treated the colony clusters as demes, and all A. eximius
sampled as the total population. In one set of calculations I treated the four
collecting sites (El Llano, El Valle, Browns Berg and Voltzberg) as sub-
populations, and in the other I considered the Panama and Suriname samples to
be sub-populations.

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

207

Table 1. — Presumed genotypes found for seven polymorphic loci in El Llano (EL), El Valle (EV),
Browns Berg (BR) and Voltzberg (VB) samples of Anelosimus eximius. N is number of individuals
sampled.

Locality

ACP

ADH

MUP

NP

SOD-4

PGM

EST-4

N

EL

11

11

11

11

11

11

11

26

EV

11

11

11

11

22

11

11,12,22

37

BR

22

22

22

22

22,33

11

22

23

VB

22

22

22

11,22

22,33

11,12„22

22

101*

*plus an additional 22 spiders sampled for PGM.

RESULTS

Collections. — At the El Llano collection site two colony clusters were found
about 1 km apart, each consisting of a single web. I collected spiders from each
of these two webs. At the El Valle site a group of webs was found along a 135
m portion of a path. At this site there were 40 webs ranging in size from 10 cm
across to 4 m across, and containing one spider to several hundred spiders. No
web was more than 22 m from its nearest neighbor, so that it appeared likely
that this was one very large colony cluster. I took spiders from five webs. Seven
colony clusters (which I labelled a A through G) were found at the Browns Berg
park. Cluster A consisted of four webs and cluster B consisted of three; the rest
consisted of single webs. The relative positions of these colony clusters are shown
in Figure 1. I collected spiders from each colony cluster except for cluster G,
which was 10 m up a tree. However, after collection a problem with excess
moisture killed all but the samples from two webs in cluster A and the sample
from cluster C. Ten colony clusters were found at the Voltzberg reserve. The
locations of these clusters are shown in Figure 2. Clusters 2, 3 and 7 contained
two webs each, the rest only single webs. I collected samples from each web in
each colony cluster.

Electrophoresis.— Thirty-seven of the 46 proteins screened gave clear and
scorable results on at least one buffer system. Ten enzymes were apparently coded
for by more than one locus (indicated in Appendix 2 by a numerical suffix, e.g.
MDH-1, MDH-2) giving a total of 51 scorable loci. Of these, 44 were
monomorphic for all individuals sampled, and only seven, or 13.7%, were
polymorphic. These seven were methylumbelliferyl phosphatase (MUP), alcohol
dehydrogenase (ADH), acid phosphatase (AC), an esterase (EST-4), a superoxide
dismutase (SOD-4), nucleoside phosphorylase (NP), and phosphoglucomutase
(PGM). SOD-4 was represented by three presumed alleles, the rest by two. The
banding patterns found, along with the presumed genotypes producing the
patterns, are presented in Figure 3.

The genotypes at these seven loci in each collection site are presented in Table
1. Of the seven polymorphic loci, three (MUP, ADH and ACP) show fixed
differences between the Panama and Suriname samples. NP and PGM were
monomorphic in the Panama samples, and EST-4 was monomorphic in the
Suriname samples. Thus in the Suriname sample three (5.9%) of the 51 loci
sampled were polymorphic, and in the Panama sample two (3.9%) were
polymorphic.

The gene frequencies found for EST-4, NP, PGM and SOD-4 in each web are
given in Table 2. Most samples taken from any one colony are made up of

208

THE JOURNAL OF ARACHNOLOGY

Table 2. — Gene frequencies of NP, SOD-4, PGM and EST-4 in webs sampled from El Llano (EL),
El Valle (EV), Browns Berg (BR) and Voltzberg (VB). The convention 1-1 is used to indicate colony
cluster 1, web 1, etc. N is number of individuals sampled from each web.

NP SOD-4 PGM EST-4 N

allele:

1

2

1

2

3

1

2

1

2

EL 1-1

1

0

1

0

0

1

0

1

0

14

EL 2-1

1

0

1

0

0

1

0

1

0

12

EV 1-1

1

0

0

1

0

1

0

0.86

0.14

9

EV 1-2

1

0

0

1

0

1

0

0.36

0.64

7

EV 1-3

1

0

0

1

0

1

0

0.64

0.36

7

EV 1-4

1

0

0

1

0

1

0

0.86

0.14

7

EV 1-5

1

0

0

1

0

1

0

0.86

0.14

7

BR A-1

0

1

0

1

0

1

0

0

1

8

BR A-2

0

1

0

1

0

1

0

0

1

7

BR C-1

0

1

0

0

1

1

0

0

1

8

VB 1-1

0

1

0

10

01

0

0

1

7

VB 2-1

0

1

0

1

0

1

0

0

1

7

VB 2-2

0

1

0

1

0

1

0

0

1

7

VB 3-1

1

0

0

1

0

1

0

0

1

7

VB 3-2

1

0

0

1

0

1

0

0

1

7

VB 4-1

1

0

0

0

1

0

1

0

1

7

VB 5-1

1

0

0

0

1

0

1

0

1

9

VB 6-1

1

0

0

1

0

0

1

0

1

8

VB 7-1

0

1

0

1

0

0.58

0.42

0

1

8*

VB 7-2

0

1

0

1

0

0.47

0.53

0

1

8**

VB 8-1

0

1

0

1

0

0

1

0

1

8

VB 9-1

0

1

0

1

0

1

0

0

1

8

VB 10-1

0

1

0

1

0

0

1

0

1

10

* three individuals unscorable.

** plus an additional 22 individuals to calculate frequencies of PGM alleles.

individuals identically homozygous at all loci sampled. Samples from colonies
within a cluster (the El Valle samples, cluster A from Browns Berg and clusters
2, 3, and 7 from Voltzberg) did not differ in alleles present. There are two
instances of heterozygosity. Within colony cluster 7 at Voltzberg the enzyme
PGM was polymorphic, and within the El Valle colony cluster the enzyme EST-
4 was polymorphic; in each instance all three potential genotypes were present
(11, 12, 22). In Voltzberg cluster 7, 30 spiders from one web and 5 from the other
in the cluster were scored for their PGM phenotype. In El Valle a total of 37
spiders were examined for their EST-4 phenotype. In both, the proportion of
individuals of each genotype in the cluster (data for the two webs which
comprised cluster 7 combined, and data for the 5 El Valle webs combined) did
not differ from that predicted by the Hardy- Weinberg equation given the allele
frequencies calculated from the data (Table 3). There were no significant
differences in the frequencies of the three genotypes (11, 12, 22) found in the two
webs which made up colony cluster 7 (chP = 1.75, 2 df, p = 0.42). Sample sizes
were not large enough to make comparisons among the five El Valle webs
sampled.

Comparison of the genotypes found within colony clusters presented in Table
2 with the spatial locations of colony clusters in Browns Berg and Voltzberg
(Figs. 1 and 2), shows that adjacent colonies (e.g. Voltzberg clusters 3 and 4,
separated by 70 m) may be fixed for different alleles of one or more enzyme.

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

209

Table 3. — Observed frequencies of genotypes found in Voltzberg 7 and El Valle 1 colony clusters
compared with the frequencies expected under Hardy-Weinberg conditions, given gene frequencies
calculated from the samples.

EST-4: El Valle webs 1 through 5 combined, frequency of allele 1
Genotype: 1 1

= 0.72, frequency of allele 2 = 0.28

12 22

observed

19

15

3

expected

chP = 0,2 d.f., p = 1

19.2

14.8

3

PGM: Voltzberg webs 1 and 2 from colony cluster 7 combined, frequency of allele 1
of allele 2 = 0.47

= 0.53, frequency

Genotype:

11

12

22

observed

8

21

6

expected

chi' = 0.93, 2 d.f., p = 0.63

9.8

17.4

7.7

Mean heterozygosity for the total A. eximius sample is 0.060 with a standard
error of 0.021, and the values for the Panama and Suriname samples are 0.017
(s.e. 0.012) and 0.024 (s.e. 0.012) respectively. Heterozygosity values within colony
clusters are zero, except for the El Valle cluster (0.0080) and Voltzberg cluster
7 (0.0099).

For each enzyme studied the biochemical phenotypes of 7 to 14 spiders from
each web were determined. In any individual web sampled, alleles that occurred
with a frequency of 0.19 or more would have been detected with a probability
of 0.05 or better. In future studies larger sample sizes (> 30 spiders per web) will
increase the probability that all alleles present in a colony are detected, and that
samples which show only one allele are representative of a monomorphic colony.
The pooled samples total 187 spiders; thus the conclusion of extreme
monomorphism in A. eximius is based on strong evidence.

F statistics for the seven polymorphic loci are presented in Table 4. Fdt
describes the reduction in observed heterozygosity in the population as a whole;
a value of zero implies that there is no difference between the observed and
expected frequency of heterozygotes, whereas values close to 1 imply a great
reduction in heterozygosity. Fst tells how much of the reduction in heterozygosity
observed in the population is associated with division of the total population into
sub-populations. A value of Fst close to zero indicates that little of the observed
loss of heterozygosity is associated with division of the total population into sub-
populations, and values close to 1 indicate that all or most of the observed loss
is associated with division of the population into sub-populations. Fds describes
what portion of the reduction in observed heterozygosity is associated with the
division of sub-populations (e.g. El Valle, El Llano, Browns Berg and Voltzberg)
into demes (colony clusters). A value of Fds close to zero would indicate that little
of the observed reduction in heterozygosity is associated with division of the sub-
populations into colony clusters, whereas values close to 1 indicate that most or
all of the observed reduction is associated with division of the sub-population
into demes.

All of the values of Fdt presented in Table 4 are large, indicating a large
reduction in the occurrence of heterozygotes in the population. The polymorphic
enzymes fall into two groups. For the enzymes EST-4, MUP, ACP and ADH all
of the observed reduction in heterozygosity is associated with division of the total

210

THE JOURNAL OF ARACHNOLOGY

Table 4. — Partitioning of total genetic variance in Anelosimus eximius into components associated
with division of the total population into demes (Eds) and sub-populations (Fst) for polymorphic
enzyme systems.

A. demes = colony clusters, sub-populations =

El Valle, El Llano, Browns Berg and Voltzberg.

Eds Fst

Fdt

NP

1.0

0.37

I.O

PGM

0.90

0.18

0.92

SOD-4

1.0

0.43

0.93

EST-4

0.0

0.93

0.93

MUP

0.0

1.0

1.0

ACP

0.0

1.0

1.0

ADH

0.0

1.0

1.0

B. demes = colony clusters, sub-populations =

Panama and Suriname.

Fds

Fst

Fdt

NP

1.0

0.32

1.0

PGM

0.91

0.05

0.92

SOD-4

1.0

0.17

1.0

EST-4

0.53

0.86

0.93

MUP

0.0

1.0

1.0

ACP

0.0

1.0

1.0

ADH

0.0

1.0

1.0

population sample into sub-populations, in particular into the Panama and
Suriname sub-populations, and none is associated with division of the local
populations into colony clusters. For the enzymes NP, PGM and SOD-4 the
observed reduction in heterozygosity is largely associated with division of the
total population and the sub-populations into colony clusters; relatively little is
associated with division of the total population into sub-populations. The F
statistics for EST-4, MUP, ACP and ADH imply that the Panama and Suriname
populations are highly differentiated. The corresponding values for the enzymes
NP, SOD-4 and PGM indicate that within large geographic areas (i.e. Panama
and Suriname) there is little differentiation among local populations but a great
deal of differentiation among colony clusters.

DISCUSSION

Social Structure and the Genetic Composition of Colonies. — Anelosimus
eximius is highly monomorphic; within the Suriname samples only three loci out
of 51 were polymorphic, and within the Panama samples only two. To interpret
these results, that is, to decide whether these levels of homozygosity are unusually
high or simply typical of spiders in general, it is necessary to compare these data
with studies of genetic variability in other cooperative and non-cooperative
spiders; however, very little comparative information is available. Riechert,
Roeloffs and McCracken are studying genetic variation in the cooperative West
African spider, Agelena consociata Denis (Agelenidae), and have found very few
polymorphic systems and few heterozygotes (Riechert, pers. comm.). Lubin and
Crozier (1985) examined 22 loci in 615 individuals of the New Guinea cooperative
spider Achearanea wau Levi (Theridiidae) taken from four localities, and found
only one polymorphic locus with two alleles. Using these data and Dowling and
Moore’s program for mean heterozygosity (1984), I calculated a mean

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

211

heterozygosity for A. wau of 0.02. I also calculated the mean levels of
heterozygosity from data for three non-cooperative species: two species of
Philophonella (Uloboridae) from Arizona (Smith, ms in prep) and for the
California trap door spider Bothriocyrtum californicum (O. P.-Cambridge)
(Ctenizidae) (Galindo-Ramirez and Beckwith 1983). Calculated mean
heterozygosity for Philophonella oweni (Chamberlin) is 0.085, s.e. 0.048 (11 loci
sampled in 18 individuals). For Philophonella sp. (undescribed species) calculated
mean heterozygosity is 0.12, s.e. 0.052 (11 loci sampled in 12 individuals). For
B. californicum, mean heterozygosity was 0.09, s.e. 0.06 (11 loci sampled in at
least 64 individuals). These three non-cooperative species show higher levels of
heterozygosity than the two cooperative species, perhaps because the cooperative
societies examined are characterized by high levels of inbreeding within colonies.
However, to establish that this is due to inbreeding and not to generally low
levels of heterozygosity in the genus Anelosimus or the family Theridiidae, similar
studies of related non-cooperative species are needed.

Colonies within a colony cluster do not appear to be genetically differentiated,
given the limitations imposed by sample size. This suggests that the colony
clusters are the result of budding rather than (for example) aggregations of
unrelated webs at particularly favorable sites. However, samples taken from
neighboring colony clusters, even when separated by as little as 70 m, may be
fixed for different alleles at one or more loci. This implies that adult males
probably do not leave their natal colonies and enter new webs to mate, and also
suggests that effective migration by females and immatures into established webs
does not occur. These data do not exclude the possibility that unrelated females
may join (either occasionally or routinely) to found a new colony during the
dispersal phase. The two colony clusters which contained heterozygotes might in
fact be the result of colony foundation by unrelated females.

Social Behavior, Population Structure, and Speciation. — The genetic variation
that occurs in the populations of A. eximius sampled is attributable to
subdivision of the population into demes (colony clusters), and into geographic
regions (Panama versus Suriname). Little or no variation is due to division of
the population into sub-populations within a geographic area. Thus in comparing
Browns Berg and Voltzberg, genetic variation appears to be attributable to
division of the population into colony clusters, not to differences between Browns
Berg and Voltzberg sub-populations as a whole. Thus, based on this sample, one
can say that the structure of the A. eximius population is a genetic mosaic, with
each piece in the mosaic consisting of one colony cluster. However, the samples
were relatively small and sampled only two points out of the large range of A.
eximius; thus what appears as a fixed difference at three loci between Panama
and Suriname populations may actually be two extremes in a continuum of allelic
frequencies across northern South America. Larger samples from Panama and
Suriname encompassing more colony clusters and more localities, and samples
from more of the range of A. eximius are needed.

Animal social systems have been proposed as an important agent shaping the
genetic structure of populations and rates of speciation. Mammals show a rapid
speciation rate and a high level of chromosomal diversity in comparison to other
vertebrate groups (Bush et al. 1977). Several authors have attributed this to the
mammalian propensity for social structuring (Wilson et al. 1975, Bush 1975),
proposing that many types of mammalian social structure lead to the formation

212

THE JOURNAL OF ARACHNOLOGY

of breeding units with small effective size and little gene flow among demes. This
theoretically would create the same genetic environment as a founder event —
a small, isolated gene pool in which stochastic events can act rapidly to fix
unique genotypes or chromosomal types that are maladaptive in the heterozygous
state and that would rapidly be eliminated by selection in a large panmictic
population.

This hypothesis has been investigated many times by many authors (e.g. Daly
1981 with wild rabbits; McCracken 1984 with two species of bats; Patton and
Feder 1981 with pocket gophers; review by Patton and Sherwood 1983). All
reported that gene flow among social units, particularly by dispersing young, was
sufficient to overcome any genetic effects caused by social structuring of the
population. The hypothesis that social structuring alone has an effect on
speciation rates does not appear to hold for mammals, the group for which it
was originally proposed; however the cooperative or quasisocial spiders may be
a perfect model for testing this hypothesis.

Templeton (1980a, 1980b) has provided a more refined analysis of the role of
founder events in speciation in the context of a more general discussion of types
of speciation. The three major elements in his analysis are: 1) the mechanism of
speciation; 2) the structure of the ancestral population; and 3) the type of split
between two potentially speciating populations.

1) Templeton (1980b) describes three mechanisms for speciation: genetic
transilience, chromosomal transilience, and divergence. In the genetic transilience
mechanism a founder event leads to a period of inbreeding during which alleles
are subject to selection for their contribution to fitness in the homozygous state
and against a more stable genetic background than (presumably) existed in the
ancestral population. But rather than relying on accumulation of fixations and
novel combinations at many loci, this model relies on changes in a few loci with
major developmental and regulatory effects leading rapidly to a new “adaptive
peak.” In chromosomal transilience, a chromosomal translocation which is
detrimental in its heterozygous form can quickly become established in small
populations and provide a rapid barrier to gene flow with the ancestral
population. Genetic divergence occurs by the accumulation of neutral and
adaptive changes in allelic frequencies in populations separated by physical
barriers, distance along a dine, host differences, etc.

2. The genetic structure of the ancestral population can be considered as one
of two extreme types: a large panmictic population with high genetic variability,
or a population divided into many tiny demes with little gene flow between them.

3) The type of split between populations can also be considered as one of two
extremes: division of a population into two large, roughly equal populations, or
a separation of small populations from a large parental population.

Genetic transilience is most likely to occur when a small founder population
splits off from a large panmictic population; it is least likely to occur when the
founding population splits off from an ancestral population which is already low
in genetic diversity. In particular, it is unlikely to occur when the ancestral
population consists of small highly inbred demes, because the founder population
is unlikely to produce any novel genetic combinations by selection of alleles that
are more fit against a uniform and homozygous background because the genome
of the parent population has already been subject to selection for fitness in a
uniform genetic background.

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

213

Chromosomal transilience is most likely to occur when the ancestral population
consists of small inbreeding demes. Because chromosomal translocations are often
deleterious in the heterozygous state, they are likely to persist only in small
populations where they can rapidly become fixed by stochastic processes. Gradual
divergence, with or without a physical barrier to gene flow, is also most likely
to occur against a genetic background of small, isolated demes. Carson and
Templeton (1984) conclude that although speciation via founder events probably
occurs only under a restricted set of conditions, this mode of speciation may
nonetheless be important for certain groups of organisms, or for organisms living
under certain ecological conditions (but see Barton and Charlesworth 1984). The
cooperative social spiders may be one such group of animals. Anelosimus eximius
as it exists today apparently consists of many small inbred demes with little
genetic variability; when a founder event takes place (such as foundation of a new
colony), the emigrants also possess little genetic variability and there is little or
no scope for novel genetic combinations to enter a new adaptive peak. Thus it
is unlikely that A. eximius has given rise to new species by genetic transilience.
Speciation by chromosomal transilience is more probable for modern A. eximius
populations; if a chromosomal translocation occurred in one of the dispersing
foundress females, it could rapidly become established in a homozygous form
among her descendants. Finally, speciation by gradual divergence may also occur;
in particular, the viscous nature of the population, and subdivision into small
non-interacting demes, would accentuate the effects of natural geographic barriers
to gene flow.

Speciation by chromosomal and genetic transilience are not expected to lead
to high levels of genetic differentiation as detected by electrophoresis, producing
full species which may be very similar electrophoretically. As a result, genetic
distance measures are not expected to be good indicators of the species status of
populations in which genetic or chromosomal transilience are frequent speciation
mechanisms. Thus before any conclusions can be drawn about genetically
differentiated populations, such as the Panama and Suriname populations of A.
eximius, it will be necessary to do comparative studies of the genetic distances
among morphologically recognized species of Anelosimus.

ACKNOWLEDGMENTS

This work was supported by grant no. 2747-83 from the National Geographic
Society. I would like to thank T. Christenson, Y. Lubin, and D. Windsor for their
help in making the Panama collections. Stinasu, the Suriname Nature
Conservancy, and Dienst ‘sLands Bosbeheer Suriname, the Suriname Forest
Service, gave me permission to work and collect in the Browns Berg and
Voltzberg reserves. The staff of Stinasu, in particular F. L. J. Baal and K.
Mohadin, provided logistic assistance and smoothed out the paperwork; L.
Roberts of the Suriname Forest Service provided expert assistance in the field.
M. Galindo-Ramirez, G. Uetz and F. Vollrath gave welcome criticism on the
manuscript. I am especially grateful to R. Hagen for his help in making
collections in Suriname and his critical comments on the numerous drafts of this
manuscript.

214

THE JOURNAL OF ARACHNOLOGY

LITERATURE CITED

Allendorf, F. W., N. Mitchell, N. Ryman, and G. Stahl. 1977. Isozyme loci in the brown trout {Salmo
trutta L.): detection and interpretation from population data. Hereditas 86:179-190.

Aviles, L. 1983. Sex ratio in the social spider Anelosimus sp. IXth International Arachnology
Congress, Panama City, Panama, August 1-8, 1983.

Barton, N. H. and B. Charlesworth. 1984. Genetic revolutions, founder effects, and speciation. Annu.
Rev. Ecol. Syst., 15:133-164.

Brach, V. 1975. The biology of the social spider Anelosimus eximius (Araneae: Theridiidae). Bull. So.
Cal. Acad. Sci., 74:37-41.

Brewer, G. W. 1970. Introduction to isozyme techniques. Academic Press, New York.

Bush, G. L. 1975. Modes of animal speciation. Annu. Rev. Ecol. Syst., 6:339-364.

Bush, G. L., S. M. Case, A. C. Wilson and J. L. Patton. 1977. Rapid speciation and chromosomal
evolution in mammals. Proc. Natl. Acad. Sci. U.S.A., 74(9):3942-3946.

Carson, H. L. and A. R. Templeton. 1984. Genetic revolutions in relation to speciation phenomena:

the founding of new populations. Annu. Rev. Ecol. Syst., 15:97-131.

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

Clayton, J. W. and D. N. Tretiak. 1972. Amine citrate buffers for pH control in starch gel
electrophoresis. J. Fish. Res. Bd. Can., 29:1 169-1 172.

Daly, J. C. 1981. Effects of social organization and environmental diversity on determining the genetic
structure of a population of the wild rabbit, Oryctolagus cuniculus. Evolution, 35:689-706.

Dowling, T. E. and W. S. Moore. 1984. A program for estimating genetic variability within and
between populations. J. Heredity, 75:416.

Galindo-Ramirez, M. and R. Beckwith. 1983. Genetic diversity between isolated populations of the
California trap-door spider, Bothriocyrtum californicum (Mygalomorphae, Ctenizidae). IXth
International Arachnology Congress, August 1-8, 1984, Panama City, Panama.

Harris, H. and D. A. Hopkinson. 1976. Handbook of enzyme electrophoresis in human genetics.
American Elsevier, New York.

Hartl, D. 1980. Principles of population genetics. Sinauer Assoc. Inc., Sunderland, Mass.

Levi, H. W. 1963. The American spiders of the genus Anelosimus (Araneae: Theridiidae). Trans.
Amer. Micros. Soc., 82:30-48.

Lubin, Y. D., and R. H. Crozier. 1985. Electrophoretic evidence for population differentiation in a
social spider, Achaearanea wau (Theridiidae). Insectes Sociaux 32:297-304.

Markert, C. L. and 1. Faulhaber. 1965. Lactate dehydrogenase isozyme patterns of fish. J. Exp. Zool.,
159:319-332.

May, B. 1980. The salmonid genome: evolutionary restructuring following a tetraploid event. Ph. D.
Thesis, The Pennsylvania State University.

May, B., J. E. Wright, M. Stoneking. 1979. Joint segregation of biochemical loci in Salmonidae:
results from experiments with Salvelinus and review of the literature on other species. J. Fish. Res.
Bd. Can., 36:1114-1128.

McCracken, G. F. 1984. Social dispersion and genetic variation in two species of Emballonurid bats.
Z. Tierpsychol., 66:55-69.

Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of
individuals. Genetics, 89:583-590.

Nei, M. and A. K. Roychoudhury. 1974. Sampling variances of heterozygosity and genetic distance.
Genetics, 76:379-390.

Overal, W. L. and R. Ferreira da Silva. 1982. Population dynamics of the quasisocial spider
Anelosimus eximius (Araneae: Theridiidae). In The Biology of Social Insects (M. D. Breed, C. D.
Michener and H. E. Evans, eds.). Proceedings of the IXth Congress of the International Union
for the Study of Social Insects, Boulder, Colorado, August 1982.

Patton, J. L. and J. H. Feder. 1981. Microspatial genetic heterogeneity in pocket gophers: non-
random breeding and drift. Evolution, 35:912-920.

Patton, J. L. and S. W. Sherwood. 1983. Chromosome evolution and speciation in rodents. Annu.
Rev. Ecol. Syst., 14:139-158.

Pennington, B. J. 1979. Enzyme genetics in taxonomy: diagnostic enzyme loci in the spider genus
Meta. Bull. British Arachnol. Soc., 4:377-392.

Ridgway, G. W., S. W. Sherburne and R. D. Lewis. 1970. Polymorphism in the esterases of Atlantic
herring. Trans. Amer. Fish. Soc., 99:147-151.

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

215

Selander, R. K., M. H. Smith, S. Y. Yang, W. E. Johnson and J. B. Gentry, 1971. IV. Biochemical
polymorphism and systematics in the genus Peromyscus I. Variation in the old-field mouse
{Peromyscus polionotus). In Studies in Genetics VI., Univ. Texas Publ. 7103.

Selander, R. K. and T. S. Whittam. 1983. Protein polymorphism and the genetic structure of
populations. Pp. 89-114, in Evolution of genes and proteins (M. Nei and R. K. Koehn, eds.).
Sinauer Assoc., Inc., Sunderland, Mass.

Shaw, C. R. and R. Prasad. 1970. Starch gel electrophoresis of enzymes: a compilation of recipes.
Biochem. Genet., 4:297-320.

Templeton, A. R. 1980a. Modes of speciation and inferences based on genetic distances. Evolution,

34:719-729.

Templeton, A. R. 1980b. The theory of speciation via the founder principle. Genetics, 94:101 1-1038.
Vollrath, F, 1982. Colony formation in a social spider. Z. TierpsychoL, 60:313-324.

Vollrath, F. and D. Rohde-Arndt. 1983. Prey capture and feeding in the social spider Anelosimus
eximius. Z, TierpsychoL, 61:334-340,

Wilson, A. C., G. L. Bush, S. M. Case and M.-C. King. 1975. Social structuring of mammalian
populations and rate of chromosomal evolution. Proc. Natl. Acad. Sci. U.S.A., 72(12):5061-5065.
Wright, S. 1951. The genetical structure of populations. Ann. Eugen., 15:323-354.

Wright, S. 1978. The Evolution and Genetics of Populations. Volume 4 Variability within and among
natural populations. The University of Chicago Press, Chicago.

Manuscript received January 1985, revised October 1985.

Appendix I. — Recipes for buffering systems and running conditions for gels.

R: described by Ridgway, Sherburne, and Lewis (1970). Electrode buffer = 0.06 M lithium hydroxide,
0.3 M boric acid, pH 8.1; gel buffer = 0.03 M Tris, 0.005 M citric acid, 0.0006 M lithium hydroxide,
0.003 M boric acid. Running conditions = pre-run 75 mA, main run < 225 V and < 75 mA.

C: described by Clayton and Tretiak (1972). Electrode buffer = 0.04 M citric acid, pH adjusted to
6.1 with N-(3-Aminopropyl)-morpholine; gel buffer = electrode buffer diluted 1:10. Running
conditions = pre-run 75 mA, main run 75 mA and < 200 V.

M: described by Markert and Faulhaber (1965). Electrode buffer = 0.18 M Tris, 0.1 M boric acid,
0.004 M Na EDTA; gel buffer = electrode buffer diluted 1:4. Running conditions = pre-run 250 V,
main run 275 V and < 75 mA,

4:adjusted from Selander et al. (1971). Electrode buffer = 0.223 M Tris (pH 7.0), 0.094 M citric acid,
pH adjusted to 6.3 with NaOH; gel buffer = Tris and Tris HCl to pH 6,1, 0.003 M citric acid, final
pH adjusted to 6.7 with 1 M NaOH, Running conditions = pre-run 75 mA, main run 75 mA.

grinding or extraction buffer: 5.36 g Tris, 5.16 g Tris HCl, 8 1 H2O, pH 7.10.

216

THE JOURNAL OF ARACHNOLOGY

Appendix 2. — List of enzymes screened, standard abbreviations and buffer systems which gave the
best results in each case. The buffers are R, C, M, and 4 (see appendix 1 for references); NA indicates
no detectable activity on any gel/ buffer system.

ENZYME

ABBREVIATION

BUFFER

acid phosphatase

ACP

C

aconitase

AC

R

adenosine deaminase

ADA

4

adenylate kinase

AK

C

alcohol dehydrogenase

ADH

4

aldolase

ALD

R

alphaglycerophosphate dehydrogenase

AGP

R

aspartate aminotransferase

AAT

R

diaphorase

DIA

4

esterase

EST-I

R,M,4

EST-2

R,M,4

EST-3

R,M,4

EST-4

R,M,4

fructose diphosphatase

FDP

M

fumarase

FUM

4

galactosaminidase

GAM-1

C

GAM-2

C

beta-glucosidase

beta-GLU

NA

glyceraldehyde-3-phosphate

dehydrogenase

GAPDH

4

guanine deaminase

GDA

NA

glucokinase

GK

R

glucose-6-phosphate dehydrogenase

G6P

R

glucosephosphate isomerase

GPI

R,C,M,4

glutamic dehydrogenase

GDH

NA

glutamic pyruvic transaminase

GPT-1

C

GPT-2

C

glutathione reductase

GR

4

hexoseaminidase

HA-1

R

HA-2

R

hydroxybutyric dehydrogenase

HBDH

R

isocitrate dehydrogenase

IDH-1

C

IDH-2

C

lactose dehydrogenase

LDH-1

C

LDH-2

C

leucine aminopeptidase

LAP

NA

malate dehydrogenase

MDH-1

C

MDH-2

C

malic enzymes

ME

R

mannosephosphate isomerase

MPI

R

alpha-mannosidase

alpha-MAN

NA

methylumbelliferyl phosphatase

MUP

C

nucleoside phosphorylase

NP

4

octanol dehydrogenase

ODH

NA

peptidase with leucyl-alanine

PEP-LA

M

peptidase with leucyl-leucyl-leucine

PEP-LLL-1

M

PEP-LLL-2

M

peptidase with phenyl-alanyl-proline

PEP-PAP

M

phosphoglucomutase

PGM

4

phosphogluconate dehydrogenase

PGD

C

phosphoglycerate kinase

PGK

4

pyruvic kinase

PK

NA

protein

PRO

C

sorbitol dehydrogenase

SDH

R

SMITH— POPULATION GENETICS OF ANELOSIMUS EXIMIUS

217

Appendix 2. — cont.

superoxide dismutase

SOD-1

C,4

SOD-2

C,4

SOD-3

C,4

SOD-4

C,4

triosephosphate isomerase

TPI

R,C,M,4

xanthine dehydrogenase

XDH

NA

Krafft, B., A. Horel and 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.

INFLUENCE OF FOOD SUPPLY ON THE DURATION
OF THE GREGARIOUS PHASE OF A
MATERNAL-SOCIAL SPIDER, COELOTES TERRESTRIS
(ARANEAE, AGELENIDAE)i

Bertrand Krafft, Andre Horel and Jean-Michel Julita

Laboratoire de Biologic du Comportement
Universite de NANCY I
B.R 239

54506 Vandoeuvre-Les-Nancy, Cedex, France

ABSTRACT

In spiders, modes of social organization range from solitary to communal living. The “maternal-
social” spiders represent an intermediate step, where the young disperse before adulthood after a
gregarious phase of variable length depending on the species. Why does social life disappear in these
species? Several arguments lead us to involve trophic factors and to formulate the hypothesis that
food shortages, resulting from the increase in the young’s food requirements and from the limitation
of the colony’s prey supply, induce the spiderlings to emigrate.

To test this hypothesis, laboratory-reared colonies (mother T offspring) of the funnel-web spider
Coelotes terrestris (Wider) were given either an ad-libitum diet, or a reduced diet. We observed that
prey consumption increased with developmental age, and that under-fed colonies dispersed
significantly earlier than did ad-libitum-fed colonies. The significance of such social plasticity in the
evolution of spider societies is discussed.

INTRODUCTION

Taking into account the intensity, duration and complexity of intraspecific
interactions, spiders appear to exhibit quite a variety of social organization,
ranging from solitary to social living. Sociality is generally supposed to have
originated via two different evolutionary routes (Shear 1970, Burgess 1976, 1978,
Buskirk 1981): grouping of solitary individuals (Valerio and Herrero 1977), and
extension of the mother-offspring association (Kullmann 1972).

Intermediate forms of social organization, called “semi-social”, ’’periodic-social”
or “subsocial” (Shear 1970, Kullmann 1972, Krafft 1979) are of special interest.
By making it possible to investigate the costs and benefits of various types of
organization in various ecological conditions (Buskirk 1975, Fowler and Diehl
1978, Jackson 1978, Lubin 1980, Smith 1983, Christenson 1984), these diverse
forms provide insights about the possible causes and mechanisms of social
evolution.

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

220

THE JOURNAL OF ARACHNOLOGY

Since Riechert’s work (1976, 1978) on the territoriality of Agelenopsis aperta
(Gertsch), the importance of trophic effects on intraspecific interactions and on
social spacing has also been shown by a number of authors with other species
(for a review on social spacing, see Burgess and Uetz 1982). Recently, Smith
(1983) observed that in the facultatively communal uloborid spider Philoponella
oweni (Chamberlin) “communal groups are located at sites where insects are
particularly abundant. Solitary females appear to be located at sites where the
insects abundance... is at the ‘background level.’ A similar phenomenon was
observed and experimentally studied in another facultatively communal orb-
weaver, the araneid Metepeira spinipes F.P. Cambridge (Uetz, Kane and Stratton
1982).

However, nothing is known of a possible influence of trophic factors on the
“maternal-social” spiders (Burgess and Uetz 1982), i.e., those species exhibiting
maternal care, with the offspring staying in a group and being fed by their
mother for a variable period of time, depending on the species (Krafft and Horel
1979). The mother-offspring association represents a high concentration of
individuals in a limited space. We hypothesize that the food needs of the offspring
increase along with their development, and that after a lapse of time the food
needs exceed the food supply, which is limited by prey abundance in the habitat
and the prey catching capacity of the colony. Eventually, food shortage induces
the spiderlings to disperse. Accordingly, it should be possible to modify the
duration of the gregarious phase, by experimentally modifying the abundance of
prey.

We chose as a model a maternal funnel-web spider Coelotes terrestris (Wider),
belonging to a family (Agelenidae) which includes typical solitary species as well
as two highly social species: Agelena consociata Denis (Darchen 1965, Krafft
1970) and Agelena republicana Darchen (Darchen 1967).

Considering the relationship between social structure and trophic factors, we
studied the effect of prey availability on the duration of the gregarious phase of
this maternal social spider, by subjecting laboratory-reared colonies (mother +
offspring) either to a restricted or to an ad-libitum diet.

MATERIAL AND METHODS

Coelotes terrestris is a terricolous species, living under stones, bark of dead
logs, etc., common in many European forests. It usually spins a tube opening at
the ground surface through a small sheet web. The female lays 30 to 70 eggs.
The young hatch after a 3-4 week incubation period. They stay for about one
month in the tube, feeding on prey caught by the female. Insects are captured
on the sheet-web (diameter 10-20 cm), carried into the tube, then dropped in the
vicinity of the young. Feeding of young through regurgitation, though suspected
by Tretzel (1961) has never been demonstrated. The offspring stay with their
mother for about one month (34 days according to Tretzel 1961), and exhibit a
clumping tendency even in the mother’s absence (Horel, Roland and Leborgne
1979, Horel Leborgne and Roland 1982). Sometimes the mother dies before the
offspring’s dispersal, but the presence of a live mother does not prevent the
offspring from dispersing.

Inseminated females, collected in beech forests around Nancy (Lorraine-
France), were individually reared in the laboratory in glass boxes (20 x 20 x 12

KRAFFT £r/4L.— GREGARIOUS PHASE IN COELOTES TERRESTRIS

221

cm) containing a mixture of sand and peat to facilitate humidification, and a
block of plaster with a gallery where the female could spin its tube (Fig. 1).

The spiders were fed crickets {Gryllus domesticus, Gryllus bimaculatus). During
the experiment, medium size (about 14 mg) cricket nymphs were provided
exclusively. Such prey could easily be captured by the female (150-200 mg), but
not by the young (max: 30 mg).

Predation was restricted to the mother to make easier the control of the
colony’s food intake. The effect of the young’s predation is now under study.

After the spiderlings’ emergence, colonies were randomly assigned to 2
treatments: 1. “Set 10 colonies (median number of young: 37) were fed ad-
libitum. The number of live crickets was checked every day, and when necessary,
adjusted up to 6 in each box. This set provided data to evaluate prey capture
rate during development. 2. “Set — 9 colonies (median number of young: 36) had
a restricted diet. They were provided only one nymph per week. Under such
conditions, “set — ” ate approximately 15 times less that “set +”. The rearing
boxes were carefully examined daily, and the positions of the spiderlings inside
or outside the tube were checked.

RESULTS

Changes in colony food requirements. — A steady increase in the mean prey
capture rate of the adlibitum fed “set +”was observed after the emergence of the

222

THE JOURNAL OF ARACHNOLOGY

NUMBER OF WEEKS

Fig. 2. — Developmental changes in the prey capture rate of ad-libitum fed colonies (Mean daily
number of crickets captured during a week per colony). Statistical comparisons (sign test): Pre-
emergence week versus post-emergence week 1: p = 0.002; Post-emergence week 1 versus post-
emergence week 2: N.S.; Post-emergence week 1 versus post-emergence weeks 3-9: from p < 0.05 to

p <0.001.

young (Fig. 2). There was a significant difference between the pre-emergence week
and the first post-emergence week (Sign test: p = 0.002). This increase in the
capture rate can be considered as indicative of an increase in the spiderlings’ food
requirements, as other experiments under progress show that artificial
suppression, or natural dispersal, of the offspring reduce the mother’s prey
capture rate to its pre-emergence level.

Extension of the colony in the gallery. — Owing to the difficulty of observation
through the silk layer, it was not possible to determine the exact position of every
spiderling in the gallery. Thus, the gradual expansion of the colony was estimated
by the distance between the most distant visible spiderlings.

After emergence, the young spiders stayed close together, initially constituting
a very tight aggregate, then extending more and more in the gallery as they got
older (Fig. 3). Treatments did not differ in this respect (Sign test: N.S.).

Dispersal. — The difficulties of observation led us to choose as an index of
dispersal the date when, for the first time, one spiderling was located outside the
gallery during day-time. The accuracy of this index was tested by observing the
dispersal of a colony of 20 spiderlings from a rearing-box introduced into a large
enclosure (180 x 60 cm). Such an enclosure made it possible for the young to
widely disperse, while still being easily spotted. When outside the gallery, the
spiderlings went off as far as possible, and, once initiated, the process of going
out never reversed (Fig. 4). So, the first appearance of a young spider outside
the gallery can be considered to mark the beginning of dispersal.

KRAFFT £r/4L.-GREGARIOUS PHASE IN COELOTES TERRESTRIS

223

Fig. 3.— Expansion of the colony in the gallery (Mean size [in cm] of the aggregate every 5 days
± range): Solid lines = “Set — ” (Restrictively fed colonies, n = 9); Dotted lines = “Set +”(Ad-libitum
fed colonies, n = 10).

20-

>

tr

UJ

UJ

I

o

CO

K

3 10
O

CO

0

z

0 5 10 1b

DAYS AFTER THE BEGINNING OF DISPERSAL
Fig. 4. — Dispersal of a colony of 20 spiderlings in a large enclosure (Day 0 is the day preceding
the first spiderling’s going out).

224 THE JOURNAL OF ARACHNOLOGY

DAYS AFTER HATCHING

Fig. 5. — Cumulative distribution of the duration of the gregarious phase according to prey
abundance in two sets of colonies: Solid lines = “Set — (Restrictively fed colonies, n = 9); Dotted
lines — “Set +” (Ad-libitum fed colonies, n = 10) (Mann and Whitney test: U = 2; p < 0.001].

Figure 5 represents the cumulative frequencies, in both sets, of the gregarious
phase durations, i.e. the lapse of time between the hatching of the young from
the egg-sac and the beginning of their dispersal. Young of the restricted diet set
dispersed fairly earlier than those of the ad-libitum fed set (Mann-Whitney test
U = 2, p = 0.001). The median gregarious period was 64 days in “set +“, whereas
it was 31 days in “set -.”

DISCUSSION

This experiment shows, first, that the colony’s food requirements increase along
with the development of young. When these requirements eventually exceed the
supply which can be obtained by the colony, the spiderlings abandon their
communal behavior and disperse. Food shortage was produced in “set by the
low number of prey items provided; however, the factors responsible for “set
+”dispersal are not yet clearly known. Genetic factors cannot be ruled out
entirely. Other factors concerning the predatory behavior of the mother are also
unclear because of the limitations placed on the observation of her behavior by
our experimental design.

We have shown that well-fed young dispersed significantly later than did under-
fed young. This indicates that the temporary social structure of Coelotes terrestris
contains a certain degree of plasticity, because its duration can be modified by
prey availability. Such a plasticity is an argument in favor of the familial or
“maternal-social” origin of at least some spider societies (Krafft 1982).

The actual behavioral mechanisms related to dispersal of young are not yet
clearly known. They might involve a fading of interattraction, or an increase in
agonistic behavior induced by food competition, as observed in several solitary

KRAFFT GREGARIOUS PHASE IN COELOTES TERRESTRIS

225

or social species (Riechert 1976, 1978; Burgess and Uetz 1982, Rypstra 1985).
These phenomena, as well as the development of mother-offspring interactions
(namely during predatory activities), are currently under study in our laboratory.

It is very instructive to compare the results obtained with a maternal-social
sheet-web spider and those obtained with Metepeira spinipes. This orb-weaver
from Mexico, (as Philoponella oweni, Smith, 1983) may be solitary or live in
groups according to the habitat’s prey abundance. In field experiments, Uetz et
al. (1982) demonstrated in M. spinipes a “flexible social spacing pattern and rapid
responses to changes in prey availability.” Thus, although differing in their
geographical location, taxonomic position, web structure and type of social living,
these species show a similar plasticity in their social organization, depending on
prey abundance. Such a convergence is another argument in favor of a multiple
origin of sociality in spiders, and of the significance of trophic factors (among
other ecological factors) in these evolutionary events.

ACKNOWLEDGMENTS

We express our thanks to Dr. G. W. Uetz for his assistance in the translation
of the manuscript, and Drs. J. S. Rovner and T. H. Christenson for their useful
comments. This study was partially supported by C.N.R.S. Grant R.C.R 08 -
0723, and a D.G.R.S.T. Grant was attributed to J.M. J.

LITERATURE CITED

Burgess, J. W. 1976. Social spiders. Sci. Amer., 234:100-106.

Burgess, J. W. 1978. Social behavior in group-living spider species. Symp. Zool. Soc. London, 42:69-
78.

Burgess, J. W. and G. W. Uetz. 1982. Social spacing strategies in spiders. Pp. 317-351, In Spider
communication: mechanisms and ecological significance. (P. N. Witt and J. S. Rovner, eds.)
Princeton University Press, Princeton, 440 pp.

Buskirk, R. E. 1975. Coloniality, activity patterns and feeding in a tropical orb-weaving spider.
Ecology, 56:1314-1328.

Buskirk, R. E. 1981. Sociality in Arachnida. Pp. 281-387, In Social Insects (H. R. Hermann, ed.)
Academic Press, New York.

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

Darchen, R. 1965. Ethologie d’une araignee sociale, Agelena consociata Denis. Biologia Gabonica,
1:117-146.

Darchen, R. 1967. Une nouvelle araignee sociale du Gabon Agelena republicana Darchen. Biologia
Gabonica, 3:31-42.

Fowler, H. G. and J. Diehl. 1978. Biology of a Paraguayan colonial orb-weaver Eriophora bistriata
(Rengger) (Araneae: Araneidae). Bull. British Arachnol. Soc., 4:241-250.

Horel, A., C. Roland and R. Leborgne. 1979. Mise en evidence d’une tendance au groupement chez
les jeunes de I’araignee solitaire Coelotes terrestris. Rev. Arachnol., 2:157-164.

Horel, A., R. Leborgne and C. Roland. 1982. Effet du groupement et de la presence de la mere sur
le developpement des jeunes de I’araignee Coelotes terrestris (Agelenidae). C. R. Colloq.
Internation. U.I.E.I.S. (Barcelona), 207-214.

Jackson, R. R. 1978. Comparative studies of Dictyna and Mallos (Araneae, Dictynidae). I Social
organization and web characteristics. Rev. Arachnol., 1:133-164.

Krafft, B. 1970. Contribution a la biologic et a I’ethologie d' Agelena consociata (Araignee sociale du
Gabon) 1. Biologia Gabonica, 3:197-301.

Krafft, B. 1979. Organisation et evolution des societes d’araignees. Journal de Psychologic, 1:32-51.

226

THE JOURNAL OF ARACHNOLOGY

Krafft, B. 1982. Eco-ethology and evolution of social spiders. Pp. 73-84. In Social Insects in the
Tropics (P. Jaisson, ed.) Universite Paris-Nord, Paris.

Krafft, B. and A. Horel. 1979. Le comportement maternel et les relations mere-jeunes chez les
araignees. Reprod. Nutr. Develop., 20:747-758.

Kullmann, E. 1972. Evolution of social behavior in spiders (Araneae: Eresidae and Theridiidae).
Amer. Zool., 12:419-426.

Lubin, Y. D. 1974. Adaptive advantages and the evolution of colony formation in Cyrtophora
(Araneae: Araneidae). Zool. J. Linn. Soc., 54:321-339.

Lubin, Y. D. 1980. Population studies of two colonial orb-weaving spiders. Zool. J. Linn. Soc.,
70:265-287.

Riechert, S. E. 1976. Web-site selection in the desert spider, Agelenopsis aperta (Gertsch). Oikos,
27:31 1-315.

Riechert, S. E. 1978. Energy-based territoriality in populations of the desert-spider, Agelenopsis aperta
(Gertsch). Symp. Zool. Soc. London, 42:21 1-222.

Rypstra, A. L. 1985. Aggregations of Nephila clavipes (L.) (Araneae, Araneidae) in relation to prey
availability. J. Arachnol., 13:71-78.

Shear, W. A. 1970. The evolution of social phenomena in spiders. Bull. British Arachnol. Soc., 1:65-
76.

Smith, D. R. R. 1983. Ecological costs and benefits of communal behavior in a presocial spider.
Behav. Ecol. Sociobiol., 13:107-114.

Tretzel, E. 1961. Biologie, Oekologie und Brutpflege von Coelotes terrestris (Wider) (Araneae:
Agelenidae). Z. Morph. Okol. Tiere, 50:375-542.

Uetz, G. W., T. C. Kane and G. E. Stratton. 1982. Variation in the social grouping tendency of a
communal web-building spider. Science, 217:547-549.

Valerio, C. E. and M. V. Herrero. 1977. Tendencia social en adultos de la arana Leucauge sp.
(Araneae, Araneidae) en Costa Rica. Brenesia, 10/ 1 1:69-76.

Manuscript received November 1984, revised November 1985.

Darchen, R. and B. Delage-Darchen. 1986. Societies of spiders compared to the societies of insects.
J. Arachnol., 14:227-238.

SOCIETIES OF SPIDERS COMPARED TO THE
SOCIETIES OF INSECTS!

Roger Darchen and Bernadette Delage-Darchen

Station Biologique
24620 Les Eyzies, France

ABSTRACT

Since the earliest studies of social behavior in spiders, their social structure has often been
compared with that of social insects. A preliminary conclusion was that the degree of evolution of
spider societies was significantly lower than that found in insects. However, we wonder if the problem
has been correctly posed. In light of Wilson’s and Michener’s works, enriched by those of the French
School (Grasse, LeMasne), we accept the definition of insect societies with all the terms which seem
necessary to characterize them, namely inter-attraction and its multiple consequences, social
polymorphism and dominance. In our analysis, we evoke overlap of generations and the foundation
of societies to demonstrate the inherent contradiction of comparing social behavior of insects and
spiders. The sociality of spiders, which actually seems to exclude the dominance and hierarchy of
individuals, is paradoxically catalogued among inferior societies. With insight gained from recent
studies, we suggest here that social evolution in spiders has developed along a clearly alternate track,
which has rarely been followed in the animal kingdom. This type of egalitarian society is difficult
to achieve in nature, and thus it is quite rare; and social spiders, whose societies are based on this
principle, represent in fact very few species.

INTRODUCTION

Knowledge of social phenomena in spiders has developed considerably more
recently than that in insects. It was therefore inevitable that the latter should be
used as a reference point, especially as the first in-depth studies on the biology
of social spiders had been carried out by entomologists who were specialists of
social insects. As with insects, spiders show various levels of social evolution,
from the solitary species to those which live in groups all their life. However,
unlike insects, and despite the huge number of species, there are very few spiders
which have passed the ultimate steps to social life. It would appear that they have
followed quite a different path in their evolution, because under close
examination, their sociality is noticeably different from that of insects. The social
organization of insects can, to some extent, be compared to that of a number
of societies of vertebrates, whereas the schemes adopted by social spiders are
rarely realized in nature.

We shall not examine the whole picture of evolution of social phenomena in
insects and spiders here; it would not be useful as the work has already been done
by others (Shear 1970, Kullman 1972, Burgess 1976 Brach 1977, Krafft 1979,

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

228

THE JOURNAL OF ARACHNOLOGY

1982, Buskirk 1981). We shall, instead, compare the characters which are
considered to define eusociality in insects, and examine how social spiders may
be compared to this scheme. We shall then be in a better position to understand
what is essentially original in spider societies.

SOCIETIES OF INSECTS

Wilson (1971) and Michener (1969, 1974) suggest that a “eusocial” species is
defined by 1) an overlap of generations, with the mother often surviving its
offspring, 2)adults taking care of the young, and 3)the presence of castes.

This general definition needs to be enriched by what has been written on this
subject for a long time by other authors (for example LeMasne 1952, and Grasse
1938, 1952).

Inter-attraction in Societies of Insects and its Consequences. — The preliminary
basis necessary for any social life is the inter-attraction between individuals of a
conspecific group. This is certainly not particular to eusocial forms, and this is
probably why Wilson (1971) does not mention it. However, in highly evolved
societies, this factor is of prime importance, as inter-attraction is the compulsory
characteristic with which all individuals have to comply. For a eusocial insect
under experimental conditions, survival in isolated conditions is possible. Eusocial
insects in nature, however, are never solitary, except where the individual remains
isolated, but only during a transitory period, and at a very specific moment of
its existence.

The fact that in evolved societies, each individual is necessarily associated with
the other members of the group to which it belongs, leads consequently to an
interdependence between individuals. As eusocial insects live socially all their
lives, one finds in such a society, animals of all ages, and at all stages of
development; this distinguishes them from the lower forms of sociality, where the
individuals live in groups only at a particular stage of their life, such as the larvae
of the same generation (e.g., processionary caterpillars, Balfour-Browne 1926,
O’Byrne 1927). Consequently, the colony will have defined composition, and to
be in equilibrium, certain standards will have to be respected. Directly linked with
the structural pattern of societies and the mutual interdependence of individuals,
is the observation that these societies are often close to any individuals from
another colony (even conspecific).

Another important consequence of inter-attraction is the appearance of
collective tasks, i.e. some members of an insect society momentarily undertaking
the same task are capable of accomplishments that individuals working separately
cannot achieve (Darchen 1958). About this subject, Grasse (1952) wrote: “Fortune
and misfortune of social life: the group takes possession of the individuals’
potentials, increases them, exalts them and makes even new ones appear; but its
members lose their freedom of action and cannot subsist outside of it.” In other
words, the study of the isolated bee or termite is of limited value. There are, in
fact, so many physiological interactions, behavior patterns and regulations among
social insect groups that there can be no doubt about the integration of the
characteristics of the individuals of the group. In short, social insects groups
present problems that are unique to them.

From the species belonging to lower societies, to the eusocial ones, there is an
increasing complexity of tasks. However, the level of complexity of tasks is not

DARCHEN AND DARCHEN— COMPARISON OF SPIDER AND INSECT SOCIETIES

229

always proportionate to levels of complexity of the other features of sociality in
a given species. Thus, for example, constructions of the pine processional
caterpillar are far more sophisticated than the nests of some ants, although there
can be no comparison between a group of these caterpillars and a society of such
ants. It is, however, in the highly evolved societies that the highest sum of varied
collective activities can be found. Cleaning of nests, construction, “agricultural”
activities, and care of the young are all examples of behavioral development
schemes requiring complex coordination between individuals. Care of the young
represents in fact only one of the manifestations of sociality. This behavior can
also appear in species which, from another point of view are frankly solitary
(such as the forficulid Labidura ripaha) (Vancassel 1977, Caussanel and Karlinsky
1984). Thus, the overall level of activity issuing from the inter-attraction of social
insects is a more exact index of the level of sociality of the species considered,
than the simple consideration of the care of the young.

The evolved society even behaves as a super-organism which has its own laws
and surpasses those of the individual. Thus in the Apis mellifera bee, the groups
of wax-making workers in charge of construction have a degree of efficiency
which cannot be compared with the skill of an isolated worker who is unable to
build anything (Darchen 1958, 1978). The heat produced by the metabolism of
one bee is insignificant, but the whole hive maintains the temperature at around
31°C, thanks to subtle coordination between the animals which are grouped
around strategic points or disperse themselves and produce ventilation according
to the need to heat or cool the hive, mainly in relation to larvae. Finally, the
“stigmergy” behavior (Grasse 1939, 1952) means an interaction between the result
of a given act (state of construction, for instance) and the individual which
perceives it; this allows an adjustment, in time and space, of behavior with regard
to the task to be done, thus assuring the coordination of the operations in the
group.

In evolved societies, the collective tasks and regulation systems which modulate
them reach peaks of perfection, thanks to the existence of many communication
systems which interact. In this respect, the ultimate level reached in the evolution
of inter-attraction in eusocial insects, is the development of forms of language
(tactile, chemical) essentially needed for the coordination of the group life. The
society is not a simple summation of individuals, but the product of inter-action
between animals, which are behaviorally linked to one another.

Social Polymorphism: Castes and Hierarchies. — Another characteristic of insect
societies is social polymorphism. In some insects where a group effect is present
(acridids, aphids, etc.) a certain polymorphism can be noted, but no where does
it attain such an intensity as in eusocial insects. Except in a few cases (for
instance some wasps), the morphology and physiology of the adult individual is
determined during the larval stage. The soldier termite will never reproduce; the
worker will have feeding glands, atrophied ovaries and working wax glands; the
queen bee will never produce royal jelly or wax, but will be an egg laying
machine, etc. Thus, the social polymorphism of eusocial insects corresponds to
the existence of specialized castes. It allows for greater efficiency of individuals
and constitutes one of the unique traits of this type of insect.

The existence of hierarchies is an essential characteristic of insect societies, so
that two notions of hierarchy formation and evolved insect societies are
fundamentally linked. These hierarchies can in fact be constitutional or acquired:

230

THE JOUR'nAL of ARACHNOLOGY

(a) Constitutional hierarchies can be found, for example, in Apis mellifera where
the queen bee is morphologically, anatomically, physiologically and ethologically
different from the worker; (b)On the contrary, the queen wasps Polistes gallicus
the hierarchy can be acquired in time and according to the circumstances
(Deleurance 1957), In effect in spring time it is not infrequent that several young
queens group together to start a nest. Together, things work out better than if
alone, but these colonies are normally monogynous; this is why, after some days
of collaboration, dominance behavior appears among these queens, and a
hierarchy is created among them. The most dominant wasp systematically
devours the eggs laid by its companions and lays in their place. The other wasps
eventually accept this pattern, and do not lay any more. Their physiological
castration becomes real and they then take the rank of worker. Thus the castes
are being established de facto before the hatching of the first eggs which will only
then produce “normal” workers. It should be mentioned that in this species the
implementation of such a hierarchy is possible because there is not, as with Apis,
an anatomic gap between workers and queens.

Apart from these dominance phenomena and existence of castes, which are the
rule in social insects, polyethism can also be noted, i.e., the assignment of
different tasks to the same insect during its life time. The successive changes of
assignments are linked not only to the age of the insects, but also to individual
behavior in relation to some momentary needs of the society. The appearance of
polyethyism may occur in relation to behavioral dominance, but not necessarily.

The individuals belonging to a particular caste have a certain anatomy,
morphology, physiology and a behavior which distinguish them from those
belonging to other castes, but among females of the same group, no fundamental
genetic differences have been demonstrated. Caste determination is an induced
phenomenon, although sometimes it occurs very early in life [e.g., the case of the
ant Pheidole pallidula (Passera 1980) or the termite Schedorhinotermes (Renoux
1976)].

The Foundation of Insect Societies. — The insects belonging to an evolved
society are socialized all their life long and the colony forms a closed system: thus
the problem is how such societies can reproduce themselves.

The evolved societies essentially reproduce themselves according to two
mechanisms:

(1) There may be a temporary isolation of reproductive individuals, which
leave the original colony to start another one, e.g.. Termites go by pairs (male
and female). In the case of hymenoptera, these are fecund females (queens) who
only undertaken the foundation of colonies. They do so either alone or by groups
of queens, but in all cases these reproducing individuals produce the first broods
of workers without the assistance of other castes.

(2) Sociotomy and multiplication may occur by splitting, which constitutes
reproduction of colonies. In the case of sociotomy (called swarming in bees), a
colony breeds a young queen which will be fecund. At a certain period the colony
will divide into two, one section of the workers will stay with the young queen
and the other with the old one. The larvae will remain either with one of the
two groups, as in bees or will be shared between the two groups, as in army and
driver ants. There is not in this case at any moment a solitary phase for any
individual of the colony (Raignier 1972, Leroux 1982). The species which multiply
by splitting, in order to increase the number of colonies, also do not have a

DARCHEN AND DARCHEN— COMPARISON OF SPIDER AND INSECT SOCIETIES

231

solitary phase. This mechanism is found in the case of primitive termites or in
evolved ants of the genus Formica.

In termites, immature individuals may become mature when they have
remained for too long without any contact with the royal, couple. Apart from
the soldiers, which cease molting, and the white soldiers and nymphs, which are
very close to their imago molt, most of the individuals of these colonies retain
the potential of molting and having neotenic generation, even if they temporarily
serve the function of workers. The castes in primitive termites are thus relatively
dynamic and reproduction by splitting is long. In ants, the colonies which
multiply by splitting are to be found in the species with polygynous colonies;
these recruit queens through swarming processes. These queens can either come
from the mother colony or from neighboring colonies swarming at the same time.

When a colony of ants becomes too large, a group of individuals emigrate
further away. They first establish a camp of workers. This camp can organize and
recruit larvae and queens. Then a satellite colony is installed, and retains contact,
with the mother colony. Then after a period, one obtains either a group of
independent colonies which are genetically related or a super colony composed
of a certain number of satellite colonies. There is therefore no general and single
system for the foundation of colonies in social insects. Ants may be those which
show the greatest diversity in the modes of foundation.

One of the factors of sociality in insects which has been observed for a long
time is of course the overlap of generations. The mother must survive its
descendants and keep close links with them so that the society can survive. It is
indispensable that the reproductive individuals (at least) have a sufficiently long
life. In monogynous colonies, the life span of a colony is related to that of the
queen, as occurs in some species of ants and bees. A possibility of control exists,
however; in case of urgent need a colony can prepare a new queen (in Apis this
possibility is well known) but this system remains fragile (for example the number
of colonies of Leptothorax ants found without a queen is not negligible); such
colonies are destined to extinction (Plateaux 1970, Poussardin 1984). Eusocial
insects have therefore developed ways to solve this situation, and polygyny is one
of them. For example, the large colonies of Formica, the nests of wasps Polybia
and Nectarina are in principle immortal, their longevity is considerably greater
than that of any individual (for example the wasp nests of Synoeca cyanea, 60
years (Evans and West-Eberhard 1970). The overlap of generations in such species
reaches its peak and the longevity of the reproducing individuals does not affect
the chances of survival of the society.

As we have seen, the three definitions proposed by Wilson to characterize a
society of invertebrates clearly are too simplified, and need to be enriched by new
ones.

SOCIETIES OF SPIDERS

Inter-attraction in Spider Societies. — The first study defining features of
sociality in spiders was that of Kullman (1972). It was thus nearly
contemporaneous with the book of Wilson (the Insect Societies, 1971). For the
German author there are three real characteristics of social life in spiders: (1)
tolerance, (2) inter-attraction, and (3) cooperation. There is nothing to say against

232

THE JOURNAL OF ARACHNOLOGY

this definition, but it was certainly not established using the same criteria as those
given with insect societies in mind.

First, we can note that inter-attraction, which for Kullman (1972) is a pertinent
characteristic, was not mentioned by Wilson (1971), as this is probably implicit
for him. Yet we have seen that in insects various very important corollaries arise
from the idea of inter-attraction. As in insects, inter-attraction arises sometimes
in spiders whether or not they are completely social. But in the most social
species it means a compulsory link which unites the members of the group.
Usually no individuals live alone except for some females who are about to lay
eggs and leave the nest to start a new colony.

The fact that the isolated individual of social spiders is unknown in nature
poses, of course, the problem of the type of links which unite the animals among
themselves. In contrast with insects, each spider is theoretically able to satisfy its
needs alone. It can hunt, spin its web, lay eggs and thus survive outside this
society. Naturally, one is inclined to think that inter-attraction is linked to the
continuation of aggregative behavior of the young, and is essentially based on the
olfactory sense. Nevertheless, it would be very interesting to have more detailed
information on this phenomenon.

In this respect, it would be very instructive to make a comparative study on
the physiology and behavior of two very close species, e.g., Achaearanea
disparata (Denis) and Achaearanea tessellata (Keyserling). The first species is
strictly found in the Gabonese forests and is at this time unknown anywhere else
in the world. This is a social spider from all viewpoints and spends its entire life
in a group; it spins its web and hunts in association with its companions, and
forms spectacular colonies very high in the trees. Achaearanea tessellata is also
a tropical species, but very widespread on all continents. Curiously, these two
species are virtually morphologically identical. However, from an ethological
viewpoint there is no possible doubt as to the fact that Achaearanea tessellata
is solitary. What has thus happened during the evolution of the common stock
of these two spiders so that they now form two different species which are, at
the same time, so close and so different? A comparative study on this subject is
certainly desirable.

In the spider societies of all species that have been studied, there are no closed
social groups. It is possible to add to a given society individuals of any age of
the same species but coming from very different locations without causing any
fights. The tolerance which can be noted here, is thus one of the essential
characteristics of this type of social structure. Spiders recognize their own species
but show no restriction at the level of the social group. In accordance with the
“non-closing” of societies, the importance of the social groups might theoretically
be unlimited. However, in spiders, as in insects, each species has its standards and
the population does not increase indefinitely as, beyond a certain threshold, there
is a splitting of societies.

Overlap of generations is also in this instance indispensable for the survival of
the society. However, in Mallos gregalis (Simon) there are seasonal variations in
the age structure of the population of the society diminishes. In other species, as
in Anelosimus studiosus (Hentz), the societies disperse after some months and all
adult individuals separate to go here and there and initiate new families. It is
clear that the cyclic diminution of the number of larvae is also known in social
insects. It seems that the authors who have studied the social spiders with a short

DARCHEN AND DARCHEN— COMPARISON OF SPIDER AND INSECT SOCIETIES

233

cycle have some reluctance in conceding them a social status which in fact is not
refused for similar insects.

As with insects, we also find in social spiders a number of collective tasks
which make these groups real entities which are very different from a summation
of individuals living on their own. These collective tasks, which are called
“cooperation” by Kullman are one of the corollaries of the fundamental
phenomenon of inter-attraction. They concern the care of the nest (cleaning,
construction, repairs), hunting, care of the eggs and cocoons (Witt, Scarboro and
Peakall 1978). The analysis of collective tasks is relatively disappointing because
each animal often seems to work on its own and independently of the others;
however, cooperation in hunting activities has been well identified in Achaearanea
disparata for example, and building activities in Mallos gregalis. From an overall
viewpoint, ’’community behavior” is clearly present among the social spiders, and
the detail of the individual movements in spiders does not differ very much from
what has been observed in various insects.

In social spiders inter-attraction is accompanied, as in insects, by forms of
communication, and this communication is the crucial issue. The communication
system is spiders is largely based on pheromones and vibrations (Darchen 1965,
1975, Krafft 1982). The web plays an important role in the transmission of
“language” as it constitutes a substratum for the smells, is a vibratory system
which carries sorts of coded information. The mode of perception in spiders is
probably too remote from ours for us to understand the real role a web holds
for the animal that builds it and spends its life on it. If this mode of perception
and communication is fundamental in spiders which make webs, it can also be
found in wasps, for example Rhopalidia cincta where cardboard construction can
also, in a way, be used as a vehicle for vibratory information for coordinating
the activity of the different members of the society at a given time. The needs
of the community, inherent in any social life, have thus found ways of expressing
themselves (both in insects and spiders) which are essentially based on tactile
auditory and odoriferous perceptions.

The existence of this “language” is perhaps what makes the essential difference
between A disparata and A. tessellata. It is impossible to ignore the importance
of communication in the establishment of social life, and it is a pity that Wilson
has completely neglected it.

Absence of Castes in Social Spiders. — Although the presence of castes in social
insects is a fundamental characteristic of eusociality, any attempt to find them
in social spiders have remained unfruitful: no worker castes, no reproductive
castes, and nothing which resembles classical structures of insect societies can be
found in spiders.

Curiously, insect societies with more structure and hierarchy are regarded as
highly derived. In contrast, those societies where the relationships between
individuals fluetuate more will be catalogued as less derived, and the same
happens with spider societies.

The concept of eusociality (which goes along with that of hierarchy, where the
individual loses its autonomy to live in the group) obviously has a valorizing
connotation whereas the social forms which have not reached the top level of
hierarchy development are, we must admit, given a pejorative judgement. This is
certainly why spiders have never been granted the status of advanced or derived
societies; they could deserve this qualification which might in fact be of another
order.

234

THE JOURNAL OF ARACHNOLOGY

Thus, we humans show (but without confessing it, because this is probably
done unconsciously) some admiration of the more hierarchical forms of society
which are of an absolute monarchy type — and a much lower attraction for the
forms without castes of a democratic type. However, in most modern human
societies nobody would dare to give such a judgement. This finding is surprising
and it shows that research even in “pure” biology sometimes leads to types of
conclusions which are unconsciously biased by some cultural prejudices, and
nobody in fact notices them because these prejudices are more or less the same
for all.

In social spiders the tolerance pointed out by Kullman is of great importance.
We have seen that it can exist vis-a-vis conspecifics coming from other colonies.
We still find this tolerance in another form in these societies: first, the males are
neither excluded nor attacked; and second good relations exist between females,
each one having the opportunity to lay eggs and achieve the various tasks of the
society. But besides this tolerance, which is the fundamental rule, it should be
interesting to study the polyethism with marked animals, as research in this field
is sorely lacking. The data we have on Mallos gregalis for example show how
this field of research is interesting. This would allow us to have a more precise
idea of the succession of tasks carried out by the different individuals. It is
difficult to discuss in the absence of experiments, and it is only when we are in
possession of precise ethograms that we can assess the degree of cohesion which
links the members of a group in the various species of social spiders.

Foundation of Spider Societies. — As for insects, it is important to know how
a spider society starts as this is a source of information on the nature of this
society. Here again we have relatively few reliable observations (Darchen 1965,
1976, 1978, 1979, Kullman 1968, Wickler 1973, Jacson and Joseph 1973, Fowler
and Levi 1980, Lubin 1981, Vollrath 1982). A spider society may start either
through individuals, or through small groups of individuals who leave to make
a colony nearby, or through groups of mature and immature animals which
emigrate some distance from their original society.

In Agelena consociata individual separations have never been discovered, but
the two other types of splitting are known. The individuals which leave separately
are either gravid females or immature animals (Stegodyphus sarasinorum Karsch,
Grasse and Joseph, 1973). Anelosimus eximius also shows solitary foundation,
but not exclusively. Vollrath (pers. comm.) who has also studied the solitary
foundation of societies of the latter species notes that isolated females have very
few chances of survival but their chances are improved if several females join
forces. We have then a type of foundation which is comparable to the polygynous
type practiced by Polistes gallicus; in Anelosimus eximius, however, Vollrath does
not indicate dominance of behavior among groups of females. Here again this
great difference between the social systems of insects and spiders can be found,
which makes the close comparisons a bit difficult.

In studying the foundation of societies, overlap of generations and the
longevity or reproductive individuals are important factors to be taken into
consideration. Spider species establishing large colonies have freed themselves
from the constraint of individual longevity, somewhat like the social insects such
as ants of the genera Formica or Polyergus. For example, Vollrath mentions a
complex of colonies of Anelosimus eximius which is at least 14 years old, and
we have seen one Agelena consociata colony which we have observed for 10
years. Spiders reach this result without the existence of castes.

DARCHEN AND DARCHEN— COMPARISON OF SPIDER AND INSECT SOCIETIES

235

Table 1. — Summary comparison of characteristics of insect and spider societies.

Characters

Societies of Insects

Societies of Spiders

Inter-Attraction

Obligatory. In nature, an iso-
lated insect (excepted during
a transitory period of time)
does not exist, nor can it sur-
vive a long time.

Obligatory. In nature an iso-
lated spider (excepted during a
transitory period of time), does
not exist. However, an experiment-
ally isolated individual is able
to survive a long time.

Stability of Societies

Equilibrium preserved by numer-
ous internal controls.

No data.

Closure of Societies

Antagonism towards newcomers
from other conspecific
societies.

Tolerance. Societies open to
conspecific newcomers.

Castes and Social links

Hierarchical origin of castes.
Psychological castration
(Dominance).

No psychological castration. All
the adults are fecund. Apparent
equality of all members.

Origin of Castes

Trophic origin of castes.

No castes.

Communications

Chemical (pheromones), mechan-
ical (tactile).

Chemical (pheromones), mechanical
(vibratory, tactile).

Collective Works

Complex coordination of tasks.

Complex coordination of tasks.

Polyethism

Present, linked to the
physiological evolution of the
individuals.

Possible, but not enough data.

Foundation of Societies

Various types of foundation
-Isolation of reproductive
individuals

-no isolation of reproductive
individuals
-sociotomy
-splitting

Possible splitting; neotonic
workers in the primitive
societies of Termites, polygyny
of some ants societies and wasps

ones.

Scarce data.

These 2 types of foundations exist,
but the most successful one is
realised through splittings. The
splitting of the colonies are easy
because all the spiders are re-
producers.

Overlap of Generations

Yes, with some corollaries like
brood care and nest defense.

Yes, with some corollaries like
care and nest defense.

Cycle of societies

Annual and perennial.

Annual and perennial.

A “Super-Organism”?

The society is a super-organism
with its own “laws” transcend-
ing the ones of isolated
individuals.

Unclear due to the absence of
hierarchies. However, spider
societies are certainly greater
than the sum of the individuals
of the colony. Data are lacking.

The longevity of the founding mother is often proposed as a factor of the social
life in spiders. There would be in fact a certain conflict between the overlap of
generations which is indispensable to the appearance of the social life, and the
death of the mother following the feeding of larvae. In this respect, social
behaviors that provide individuals to take care of the brood, lead to economy
in the life of the laying mothers. Then it is possible to say that social life induces
the longevity of females and not the opposite (that longevity is a prerequisite to
social life). This is, of course speculation, but the present behavior where the

236

THE JOURNAL OF ARACHNOLOGY

mother takes care of the young and dies soon after may be considered as a stage
of the evolution of spiders towards sociality.

CONCLUSION

Tolerance, lack of caste and hierarchy, which appear among social spiders,
cannot be attributed to a lower range of sociality than that of the eusocial insects,
but rather, as we have tried to demonstrate, to a different kind of sociality.

A question comes to mind when considering social spiders: why has this type
of society not been more successful in nature? Why are there so few social species
despite the fact that they can be found in various families?

The studies which have already been carried out on these animals come to the
conclusion that among the advantages of spider sociality there is: (1) an economy
in the number of eggs laid; (2) better success of the hatching; (3) no repression
of female reproductive capability.

We may ask ourselves again why they did not conquer more of the world, as
they do not show aggression between different social groups within the species,
and thus have greater chances to co-exist. One could probably reply that there
are other constraints, such as difficulty for the societies to disperse over long
distances (cf., Vollrath 1982), and many other arguments which could be
considered in this respect.

Finally, it is certain that these spider societies are unique, given the mechanisms
of tolerance they have developed and which can be contrasted to the dominance
which generates castes in social insects (Table 1). It must be, that even in
invertebrates, democracy is a difficult goal to achieve.

ACKNOWLEDGMENTS

We thank Dr. H. G. Fowler and Dr. R. E. Buskirk for the helpful comments
on an earlier version of the manuscript. We are grateful to Dr. G. W. Uetz for
his help in the publication of this work.

LITERATURE CITED

Balfour-Browne, F. 1925. The evolution of social life among caterpillars. 3e Cong. Int. Entomol.,
Zurich, 1925 (1926):334-340.

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

Burgess, J. W, 1976. Social Spiders. Scient. American, 243:100-106.

Buskirk, R. E. 1981. Sociality in the Arachnida. Pp. 281-359, In H. R. Hermann, ed. Social Insects.
Vol. 2.

Caussanel, C. and A. Karlinsky. 1981. Les soins maternels d’un perce-oreille. La Recherche, 125:1004-
1007.

Darchen, R. 1959. L’Abeille, Insecte Social. J. Psychol. Norm. Pathi, 1:75-98.

Darchen, R. 1959. Les techniques de construction chez Apis mellifera. Ann. Sc. Nat,. 1:1 13-209.
Darchen, R. 1965. Ethologie d’une Araignee Sociale, Agelena consociata Denis. Biol. Gabonica,
1:117-146.

Darchen, R. 1967. Une nouvelle Araignee Sociale du Gabon, Agelena repubilcana Darchen. Biol.
Gabonica, 3:31-42.

DARCHEN AND DARCHEN— COMPARISON OF SPIDER AND INSECT SOCIETIES

237

Darchen, R. 1967. Ethologie de quelques Araignees Sociales. L’interattraction, la construction et la
chasse. C. R. V. Congress U. 1. E. I. S., Toulouse, 1365-1367.

Darchen, R. 1968. Ethologie d'Achaearanea disparata Denis, Araignee Sociale du Gabon. Biol.
Gabonica, 4:5-25.

Darchen, R. 1975. Les communications sociales chez Agelena consociata Denis. C. R. Acad. Sc.
Paris, 281:575-578.

Darchen, R. 1976. La fondation de nouvelles colonies d' Agelena consociata et d' Agelena republicana,
Araignees Sociales du Gabon. Problemes eco-ethologiques. C. R. Ill col. Arach. Fr., Les Eyzies,
20-39.

Darchen, R. 1976. Rhopalidia cincta, guepe sociale de la savane de Lamto (Cote d’Ivoire) (Hym.
Vesp.). Ann. Soc. Entomol. France, 12:579-601.

Darchen, R. 1978. Les essaimages lointains de I’Araignee Sociale, Agelena consociata, dans la foret
gabonaise (III). C. R. Acad. Sc. Paris, 287:1035-1037.

Darchen, R. 1978. Le comportement constructeur des Abeilles Sociales. Courrier du CNRS, 30:39-
45.

Darchen, R. and J.-C. Ledoux. 1978. Achaearanea disparata Denis, Araignee Sociale du Gabon,
synonyme ou espece jumelle d'A. tessellata, solitaiare? Rev. Arachnol., 1:121-132.

Darchen, R. 1979. Relations entre les colonies d’Agelenides Sociaux du Gabon. Precisions sur les
essaimages (II). Bull. Biol. France et Belgique, 1 13:3-29.

Darchen, R. 1980. Les populations d' Agelena consociata, Araignee Sociale, dans la foret primaire
gabonaise. Leur repartition et leur densite. Ann. Sc. Nat. Zool., 2:19-26.

Darchen, R. 1982. Problemes ecologiques souleves par une etude biometrique des individus de
colonies de I’Araignee Sociale, Agelena repubilcana Darchen. C. R. V. Col. Arachnol., Modene,
Pise.

Deegener, P, 1920. Soziologische Studien an Raupen, und Bemerkingen uber Licht und Statischen
Sinn. Arch. f. Naturgesch., 86:91-154.

Deleurance, E. P. 1957. Contribution a I’etude du comportement des Polistes (Hymenopteres,
Vespides). These, Fac. Sc. Paris Masson Edit., 222 pp.

Evans, H. E. and M. J. West-Eberhard. 1970. The Wasps. Univ. Michigan Press, Ann Arbor. 265

pp.

Fowler, H. G., Levi, H. W. 1980. A new quasi social Anelosimus spider (Araneae, Theridiidae) from
Paraguay. Psyche, 86:11-18.

Grasse, P. P. 1939. La reconstruction du nid et le travail collectif chez les Termites Superieurs. J.
Psych. Norm. Path., 30:370-376.

Grasse, P. P. 1952. La regulation sociale chez Is Isopteres et les Hymenopteres. Col. Intern. C. M.
R. S., 34:322-331.

Jacson, C. C. and K. J. Joseph. 1973. Life-history, bionomics and behaviour of the social spider,
Stegodyphys sarasinorum K. Insectes Sociaux, 20:189-204.

Krafft, B. 1971. Les interactions entre les individus chez Agelena consociata, Araignee Sociale du
Gabon. Proc. Int. Arachnol. Congr., 5th, 1970, 159-164.

Krafft, B. 1979. Organisation et evolution des Societes d’Araignees. J. Psychol., 1:23-51.

Krafft, B. 1982. Eco-ethology and evolution of social Spiders. Pp. 73-84, In Social Insects in the
Tropics.

Kullman, E. J. 1971. Versuche zur toleranz bei der permanent sozialen Spinnenart Stegodyphus
sarasinorum K. (Eresidae). C. R. 5. Congr. Arach. Brno.

Kullman, E. J. 1972. Evolution of Social Behavior in Spiders (Araneae; Eresidae and Theridiidae).
Ann. Zool., 12:419-426.

LeMasne, G. 1952. Classification et caracteristiques des principaux types de groupements sociaux
realises chez les invertebres. In Structure et physiologie des Societes Animales, Coll. Intern. C. N.
R. S., 1950.

Leroux, J. M. 1982. Ecologie des populations de Dorylines Anomma nigricans dans la region de
Lamto (Cote d’Ivoire). These Univers. Paris, pp. 157.

Lubin, Y. 1981. Dispersal by swarming in a social Theridiid. Amer. Arachnol., 24:18.

Michener, C. D. 1969. Comparative social behavior of bees. Ann. Rev. Entomol., 14:299-342.

Michener, C. D. 1974. The social behavior of the Bees. Harvard Univ. Press. Cambridge, 404 pp.
Mosebach-Pukowsky, E. 1937. Uber die Raupengesells chaften von Vanessa io und Vanessa urticae.

Zeitschr. Morph. Ockol. Tiere, 33:358-380.

O’Byrne, H. 1. 1927. Gregarious caterpillars. Proc. Missouri Acad. Sci., 3:103-108.

238

THE JOURNAL OF ARACHNOLOGY

Passera, L. 1980. La ponte d’eufs preorientes chez la fourmi Anelosimus eximius Nyl. Insectes
Sociaux, 27:79-95.

Plateaux, L. 1970. Sur le polymorphisme social de la fourmi Leptothorax nylanderi F. Ann. Sc. Nat.,
Zool., Paris.

Poussardin, C. 1984. Recherches sur la biologic et I’ecologie d’une population de Leptothorax
nylanderi F. (Hym., Formicidae) en foret d’Orleans. These 3eme cycle, Orleans.

Raignier, A. 1972. Sur I’origine des nouvelles societes des Fourmis voyageuses africaines (Hymenop.

Formicidae, Dorylinae). Insectes Sociaux. 3:158-170.

Renoux, j. 1976. le polymorphisme de Schedorhinotermes (Sjostedt) (Isoptera-Rhinotermitidae).
Insectes Sociaux, 23(3bis):279-494.

Shear, W. A. 1970. The evolution of social phenomena in spiders. Bull. British Arachnol. Soc., 1:65-
76.

Vancassel, M. 1977. Le developpement du cycle parental de Labidura riparia. Biol. Comp., 2:51-75.
Vollrath, F. 1982. Colony foundation in a social spider. Z. Tierpsychol., 60:313-324.

Wickler, N. 1973. Uber koloniegrundung und soziale Bindung von Stegodyphus mimosarum Pavesi
und anderen sozialen Spinnen. Z. Tierpsychol., 32:522-531.

Wilson, E. O., The Insect Societies. Harvard Univ. Press, Cambridge, 548 pp.

Witt, P. N. M. B. Scarboro and D. B. Peakall. 1978. Comparative feeding data in three spider species
of different sociability: Araneus diadematus Cl., Mallos trivittatu B. and Mallos gregalis S. Symp.
Zool. Lond., 42:89-97.

Manuscript received August 1984, revised December 1985.

Lubin, Y. D. 1986. Courtship and alternative mating tactics in a social spider. J. Arachnol., 14:239-
257.

COURTSHIP AND ALTERNATIVE MATING TACTICS
IN A SOCIAL SPIDERi

Yael D. Lubin

Smithsonian Tropical Research Institute
Balboa, Republic of Panama^

ABSTRACT

Males of a social theridiid spider, Achaearanea wau Levi, have a prolonged and complex courtship
in the communal web. They also mate opportunistically, by copulating with females engaged in
molting or feeding. Ritualized aggression occurs between males during attempts to mate with molting
females and during courtship. I describe the courtship, mating, and agonistic behaviors observed in
A. wau and the conditions in which the different behaviors occur and discuss ways in which the social
organization may have modified mating behavior. This is the first detailed description of the mating
system of a highly social species of spider.

INTRODUCTION

Mating behavior and the mating system are influenced by social organization
(Thornhill and Alcock 1983). Sexual selection in social species, as in nonsocial
ones, may produce (1) exaggerated traits associated with mating and (2)
synchronization of mating effort. Because communication in social groups is
complex and there are more opportunities for repeated interactions among
individuals, we might expect sociality to favor greater ritualization of both
courtship and agonistic behaviors (than in nonsocial species). Therefore, the
mating tactics of a social species should be instructive in elucidating the factors
that influence the evolution mating systems.

Many species in which complex male courtship displays occur have one or
more alternative mating tactics, used either by different kinds of individuals or
in different circumstances (Thornhill and Alcock 1983). (I use tactics throughout
to mean complex behavior sequences). Often males attempt to obtain matings
without engaging in time-consuming (and often risky) courtship displays. This
may be done either by sneaking copulations or by mating with females while they
are engaged in some other activity or are incapable of preventing the male’s
advances. Such tactics are often referred to as “opportunistic mating” (Robinson
and Robinson 1980; for reviews of alternative mating behaviors, particularly in
invertebrates, see Blum and Blum 1979, Thornhill and Alcock 1983). Alternative
mating behaviors were described in several species of solitary spiders (Robinson

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

^Present Address: Mitrani Center for Desert Ecology, Jacob Blaustein Institute for Desert Research,
84990 Sede Boqer, Israel.

240

THE JOURNAL OF ARACHNOLOGY

1978a, Robinson and Robinson 1980, Christenson 1984), but have not been
observed previously in social species.

Only fragmentary accounts are available of mating behavior of social spiders
(Robinson 1978b, Jacson and Joseph 1973); therefore, a detailed description of
mating in the social theridiid spider, Achaearanea wau Levi is warranted. I
describe here the different mating behaviors and the conditions in which they
occur, and discuss the possible functions of the different tactics and the evolution
of the mating system of A. wau. I show that (1) mating behavior in A. wau is
complex, involving stereotyped courtship and at least two alternative tactics that
do not involve courtship, and (2) mating is synchronized and spatially localized
in the colony. A quantitative analysis of alternative mating tactics is provided
elsewhere (Lubin, ms).

METHODS

Natural History. — Achaearanea wau Levi (Theridiidae) is a social spider found
in montane forest and forest edge in Papua New Guinea (Levi et al. 1982). The
communal web of a colony consists of a nonsticky, horizontal sheet with a
tangled maze of nonsticky threads (the barrier web) above it and one or more
curled-leaf retreats suspended in the barrier web by strong threads (Fig. 1). The
spiders sit in or below the leaf retreats during the day and are active mainly at
night; prey capture and web maintenance activities are cooperative, feeding takes
place in leaf retreats and spiders share prey. Females and their offspring overlap
in time. Females tend eggsacs and capture prey for the young (of other females
as well as their own). Accounts of colony structure and activities of spiders are
provided elsewhere (Lubin 1982, Lubin and Crozier 1985).

Achaearanea wau colonies have discrete, overlapping generations. No more
than two generations are present in a colony at any time, and adults of
consecutive generations do not overlap. Females mature at the seventh instar and
most females in a colony mature at the same time. Males are smaller than females
(body length: males 2 mm, females 4.5 mm) and mature at the fifth instar, about
3-4 weeks before the females (Table 1). The sex ratio is skewed towards females,
with about six females per male. Courtship and mating activities take place in
the parent colony over a period of 3-7 weeks, but most of the activity is
concentrated during a 2-3 week period (Table 1). Females produce one or two
eggsacs each, one to two weeks after mating. Dispersal is by means of mass
swarming of adult females and does not occur in every generation (Lubin and
Robinson 1983).

Observations and Analyses. — Observations of courtship and mating behaviors
were of spiders in colonies that were maintained in semi-captivity in the gardens
and coffee plantations of the Wau Ecology Institute (Wau, Morobe Province,
Papua New Guinea, elevation ca. 1100 m) and in natural populations in the
vicinity of Wau.

Colonies were collected from natural populations and released onto open
wooden frames (60x60x60 cm) provided with wire-mesh screens across their lower
faces (used to collect prey remains and other spider rejects) and perched on legs
at about 1.5 m above ground. There was no need to enclose the spiders
completely; when the collapsed web with leaf retreats (brought in from the field)

LUBIN— SOCIAL SPIDER MATING TACTICS

241

Fig. 1. — Web of A. wau showing the sheet (S), barrier web (B), and curled-leaf retreats (L).

was suspended from the frame, the spiders generally rebuilt the communal web
during the following two nights, sometimes attaching lines to the surrounding
vegetation as well as to the wooden frame. The colonies in semi-captivity were
in a habitat similar to the forest edge and treefall clearing habitats that they
normally occupy. The colonies were checked at least twice a week (generally
daily) to determine the onset of the various developmental stages (Table 1).

Courtship activities were observed in 10 colonies during the period October
1979-January 1981. Observations of males in these colonies were made for
periods of 1-20 days per colony, for a total of 60 days. Most observations were
of unmarked males: males were too small to mark with dots of acrylic paint. I
watched an individual male and recorded its activities for periods of 30 sec to
10 min, until he moved out of sight or became inactive. Timed observations of
individual males (a total of 678 min for all colonies combined) were supplemented
by routine, nightly censuses of spider activity in all colonies. Observations were
made mainly at dusk and at night, when the males were most active. I recorded
when and where females molted and if they were approached by males, as well
as any other activities in the colony (kleptoparasite activity, feeding, etc.). Details
of male behavior, particularly the courtship sequences, were observed with the aid
of a pair of 10 x 40 binoculars mounted on a tripod and fitted with a +3 diopter
close-up lens on one eyepiece. Super-8 movie films and videorecordings (using a
SONY AV-3400 videorecorder) were made at night using artificial lighting.

COURTSHIP AND MATING BEHAVIORS

Sexual activities in A. wau include searching for a mate, pre-copulatory
displays, female responses to male advances and finally mating (or mating
attempts). I also include male-male interactions and sperm induction, as these
behaviors occur only in the context of courtship and mating activities.

242

THE JOURNAL OF ARACHNOLOGY

Table 1. — Chronology of events related to courtship and mating activities in colonies of A. wau
maintained in semi captivity, (a) A = Mt. Missim, 1500 m; B = Wau Ecology Institute, 1200 m; C
= Mt. Kaindi (Kunai Creek), 1700 m; D = Mt. Kaindi (Namie Creek), 1600 m; E = Mt. Kaindi,
elevation unknown, (b) Number of days from the first appearance of mature males and females to
first observations of courtship or mating activities. Negative numbers indicate that males began
courting before females matured, (c) Colony collected while courtship was already in progress.

Colony

Days to First Courtship

Mating Season

No.

Origin (a)

Male

Female

Month

No. Days

I

A

4

-15

Feb.

39

2

C(c)

—

Mar.

—

3

D

28

0

May

26

4

E

29

3

May

28

5

B

28

0

Jun.

22

6

A

17

-6

Oct.

32

7

B

11

1

Nov.

31

8

B

15

2

Dec.

51

9

C

21

0

Dec.

36

10

B

11

-5

Dec.

29

Mean duration of mating season + S.D.

= 32.7 + 8.5 days

Courtship and mating took place in the communal barrier web, primarily in
the region of the curled-leaf retreats. A male became active and began to search
for a female. If he encountered an adult female (sometimes subadult females were
apparently not distinguished from adults), he either attempted directly to mate
with her (Fig. 2A), or constructed a mating arena and began courting (Fig. 2B).
A nearby female sometimes responded to the courtship display by approaching,
touching legs with the male and then hanging from a thread in a receptive
posture; the male then attempted to copulate. A searching male that encountered
a female in the process of molting attempted to copulate without courting (Fig.
2C). Males sparred over molting females, and in certain other contexts as well.
The different behaviors are described in detail below.

MALE REPRODUCTIVE BEHAVIORS

Sperm induction. — Sperm induction began with the male cutting out barrier
web threads and putting in his own draglines. The sperm “web” of A. wau is not
a web at all, but a single thread that is reinforced several times. The male then
dabbed sperm onto the thread with his abdomen and a shiny, white spot became
visible on the thread. While filling the palps, the male was supported in an
inverted position by all 8 legs. The palps were filled one at a time, alternating
left and right, and repeating the process several times. When first applied to the
droplet of sperm, the palp was vibrated rapidly; this was followed by slow
pulsating movements. As each palp was removed from the sperm droplet, it was
shaken rapidly a few times. The male then remained sitting and passed each palp
through the chelicerae (cleaning?), sometimes pausing to wave the palp up and
down gently for 10-30 sec. After sperm induction, the male remained sitting and
occasionally cleaning the palps for more than 15 min.

Sperm induction was observed on five occasions: twice in the morning and
three times in the afternoon and early evening, one male began construction of
the sperm web about 10 minutes after a copulation. Charging of palps after a

LUBIN— SOCIAL SPIDER MATING TACTICS

243

A.

B.

C.

Search

Attempt
to mount

Search

MATURE

FEMALE

Construct

arena

"1

Steal

arena

Copulate:

short

Cou rt

Female
, approach

Touch,
tap legs

Female
rec eptive

Mount, rap

Search

Attempt
to mount

I

MOLTING

FEMALE

Spar

Copulate :
long

Copulate :
short

Fig. 2. — Schematic diagram of male mating behaviors, given encounters with mature (non-molting)
females (A and B) or with a molting female (C). In courtship sequences with mature females (B),
males mount only if females adopt a receptive posture.

copulation has been reported in other spiders (e.g., Sicarius sp., Levi 1967).
Sperm deposition lasted 15 sec in one instance and about 36 sec in another. Two
complete sequences of charging of palps lasted 9 and 17 min, respectively. Each
subsequent palp application was of longer duration than the previous one; two
such sequences were 23, 27, 40, 62 and 31, 37, 42, 52, sec, alternating palps each
time.

Searching. — At the beginning of an activity bout, males walked around near
and below the leaf nests, and made tapping and rotary movements with legs I,
frequently touching females that they encountered. Searching graded into
courtship or into attempts to copulate with a female.

Are these male perambulations in the web in fact ‘searching’ behavior, and are
they specifically aimed at finding females? The evidence in support is as follows.
First, the tapping and rotary movements of legs I were indistinguishable from
exploratory movements used in other contexts (e.g., web-building or approaching
foreign objects in the web), which suggests that the spider is receiving tactile and/
or chemical cues. Second, the males were active almost exclusively in the
immediate vicinity of the leaf nests, where females could be found either feeding
or inactive. This is quite unlike the behavior of females, or juveniles or either sex
and of males before the onset of sexual behavior: when these spiders became
active, they moved away from the leaf nests into the barrier web or onto the sheet
where most colony activities (web repair, prey capture) took place. Finally, during
“normal” colony activities there was little physical contact between individuals,
whereas searching males frequently touched females and indeed, moved from one
female to the next.

244

THE JOURNAL OF ARACHNOLOGY

Courtship: the display. — Males courted in small lacunae in the barrier web
which they prepared just before courting (Fig. 3). The male cut out a small
volume of barrier web threads and laid down one or more dragline threads. These
threads might be considered functionally equivalent to the mating thread of
araneids (Bristowe 1958). Most display courts were situated near or just below
the leaf nests. Courtship was not always preceded by construction of mating
threads; apparently males displaced one another from display courts, or simply
used courts abandoned by other males. In practice, it was difficult to recognize
a display court unless one observed the male building it. Small, thread-free spaces
occurred naturally in the barrier web (for example, where an insect had been
removed from the barrier web) and the male’s mating threads were not
distinguishable (to the observer) from other silk lines.

The behavior of cutting out segments of the females’ web before or during
courtship, thereby reducing the capture area of the web, is sometimes interpreted
as a mechanism for minimizing the risk to a male of being attacked by the female
and treated as prey (e.g., Fecenia sp., Robinson and Lubin 1979), or as a mans
of cutting off the female’s escape route from the male (e.g., in Latrodectus
hesperus Chamberlin and Ivie, Ross and Smith 1979). Neither of these
explanations is applicable here, since the females are neither aggressive nor do
they show a tendency to escape from the web. More apposite is Rovner’s (1968)
suggestion that web removal reduces the possibility of interference from other
males and isolates the female from extraneous signals (such as those of insect
prey) that might interfere with mating.

Courtship involved two distinct displays, abdomen vibrating and twanging,
which were sometimes accompanied by tapping and rotary movements of first
legs. Abdomen vibrating (AVing) involved high frequency, low amplitude, vertical
oscillations of the abdomen from the pedicel. A single oscillation lasted about 1/
9 sec (2 frames of super-8 film at 18 fps) and displaced the mating thread by 2
mm. While AVing, all legs rested on threads and only the abdomen moved (Fig.
3). The vibrations induced were probably transmitted via the legs and dragline
to the mating thread and thence to threads in contact with the female.

Similar abdomen movements occur as part of the courtship behaviors of other
species, e.g., in the linyphiid, Frontinella pyramitella (Walckenaer) (Suter and
Renkes 1984) and in the theridiids Achaearanea tepidariorum (C. L. Koch)
(Bonnet 1935), Teridion pallens Blackwell (Locket 1926) and Latrodectus
hesperus (Ross and Smith 1979). In the theridiids Steatoda (— Teutana) grossa
(C. L. Koch) and S. bipunctata (Linnaeus), stridulatory organs on the abdomen
are coupled with abdomen vibrating movements to produce vibrations or clicks
(sometimes audible) that may be transmitted through the web (Gwinner-Hanke
1970). A stridulatory organ was not found in A. wau males examined under
scanning electron microscopy (H. W. Levi, pers. comm.).

Twanging is described by Robinson and Robinson (1980, p. 13) as “a
movement on mating threads in which the thread is rapidly tensioned and then
the tension suddenly released.” Twanging or plucking of threads occurs as a
component of courtship in many web spiders (e.g., A. tepidariorum, Bonnet 1935;
Teridion pallens, Locket 1926). In A. wau, the male pulled the mating thread by
flexing legs I and II and released it suddenly with legs I (and perhaps II), causing
a violent backwards jerk of the whole body. Both the spider and the mating

LUBIN— SOCIAL SPIDER MATING TACTICS

245

Fig. 3. — Male courting in the barrier web. Note that all legs hold threads.

thread were displaced during a twang which may produce large amplitude
oscillations of the mating thread (Fig. 4).

Tapping and rotary movements of first legs occurred during AVing and
between twangs. One leg I lightly touched the mating thread or nearby threads
in a slow, gentle motion that did not visibly deflect the thread. Sometimes the
tapping leg I did not strike a thread and continued to wave about in the air in
a circular movement, similar to leg I movements made when walking in the
barrier web and searching for females.

Courtship displays followed a stereotyped sequence, always starting with a bout
of AVing and followed by alternating bouts of twanging and AVing of variable
duration. A few patterns emerge from analyses of Super-8 films of courtship
sequences: (1) Twangs were often given in short bursts of 2-4 at a time. Intervals
between twangs within a burst were distributed bimodally, with peaks at 2/9 to
1/6 sec (4-6 frames of super-8 movie film, at 18 fps)and at 13/18 to 5/6 sec (IB-
IS frames; Fig. 5 A). (2) Bouts of AVing were most frequently 2/ 9-2/ 3 sec (4-12
frames) in duration, with a mean of 0.77 sec (SD = 0.75, range 0.06-4.28, N =
102; Fig. 5B). No regular pattern could be detected in either the frequency or
duration of alternating bouts of AVing and twanging within a display sequence.
However, individual variation was not studied.

AVing occurred off the mating thread in two contexts besides during courtship:
males AVed while following or attempting to mount females and during male-
male sparring bouts (described below). In neither of these contexts was AVing
accompanied by twanging.

Courtship: interactions. — Are male displays directed at particular females or
are they addressed to any female within ‘hearing’ distance? Searching behavior,
placement of the display thread and the display itself all seemed to be directed
at one or a group of females. The male investigated several females from close
range while searching; the display thread was attached at its proximal end to

246

THE JOURNAL OF ARACHNOLOGY

Imm

Fig. 4. — Twanging. Drawings based on a sequence of frames from Super-8 movie film. Two frames
are superimposed in each drawing, the first in solid lines and the second in dotted lines; frame
numbers are shown above each drawing. Note that in some frames not all legs were visible. The
backward jerking portion of the twang was seen as a blur on a single frame (15). Two twangs are
shown (frames 13-16 and 27-32). A small, forward movement precedes the twang (1-7, 7-13). The
vertical displacement of the thread was about 4 mm.

threads on which one or more females rested and while building the display
thread, the male frequently approached and touched these females; and finally,
the male displayed facing one or more females at a distance of 1-5 cm.
Apparently, however, the displays were not perfectly directional; females
sometimes approached displaying males from behind or from the side. Courting
males responded to frontal approaches by attempting to copulate in 89% of 89
observations, and in only 20% of 56 observations of approaches from other
directions (x^ 66.85, p < 0.001, df = 2).

The behavior of a displaying male on the approach of a female depends on
the response of the female (see Female Behaviors). Usually, there was mutual
touching and tapping of legs, after which the female adopted an acceptance
posture and the male mounted. Sometimes, however, the female approached the
male to within 1 cm or less and sat. The male either went up to the female,
touched her and resumed displaying, attempted to climb directly onto her ventral
side and copulate, or left to search for a new display site. The cues that trigger
the different male responses are not known.

Interference from other males during the display resulted in the termination of
courtship in nearly 10% of 197 courtship bouts in one colony. After such
interruptions, the ‘resident’ male either resumed displaying or searched for a new
site. Likewise, when a courting male failed to attract a female or when a female
approached and then rejected the male, he often moved away from the display
court and resumed searching.

COPULATION AND COPULATION ATTEMPTS

If a female adopted a receptive posture (see below) following the male’s display,
the male moved to the female’s ventral side and drummed on her epigynum with

LUBIN— SOCIAL SPIDER MATING TACTICS

247

NUMBER OF FRAMES

Fig. 5. — Frequency distributions of components of courtship, (a) Intervals between successive
twangs, based on Super-8 movie films at 18 fps (upper) and on videorecordings at approximately 30
fps (lower), (b) Durations of abdomen vibrating (AVing), in frames of Super-8 movie film at 18 fps.

the palps, alternating them rapidly. Males also drummed on the epigyna of
females that were busy feeding, in attempts to copulate. Montgomery (1903)
commented that male Achaearanea tepidariorum appeared to have difficulty
inserting the palp once the female adopted a receptive posture. This appeared to
be the case in A. wau. Palpal drumming may be a way of stimulating the female
to allow copulation and, as such, is an important part of courtship. Palpal
drumming may have a function similar to that of the pre-mating insertions in

248

THE JOURNAL OF ARACHNOLOGY

Frontinella pyramitella (Helsdingen 1965) or the post-mount courtship display of
Phiddipus johnsonii (Peckham and Peckham) (Jackson 1982).

If successful in stimulating the female, the male inserted one palp and pumping
movements of the abdomen could be seen. A male copulated only once, with one
palp, and then was usually chased away by the female. Copulations with mature
(non-molting) females were of short durations: all but one of nine timed
copulations lasted less than one minute, and all were terminated by the female.
There was no evidence of a copulatory plug, nor did males break off one of their
emboli, although both of these phenomena are known to occur in other theridiids
(e.g., Braun 1956).

If a female was not receptive there often ensued a frantic scramble on and
around her abdomen, as the males(s) tapped and drummed on her abdomen with
his palps and legs and attempted to gain access to her ventral side. This behavior
was observed in A. wau in three contexts. First, males pursued and attempted
to mount females that were attracted initially to the male courtship display, but
then rejected the male and moved away. I never observed a copulation following
such a pursuit. Second, when copulation attempts were interrupted by another
male, both males scrambled around the female as she moved away. Third, A. wau
males scrambled and sparred with one another on molting females, and even on
exuvia of newly molted females and on females that had died during molting (see
below).

During copulation with mature as well as with molting females, the male’s long
axis was parallel to that of the female and the two spiders faced in the same
direction. Sometimes, however, the movements of a recently molted female
attempting to dislodge a copulating male, caused the male to turn around so that
his body was at a right angle to that of the female, or even facing in the opposite
direction.

MALE-MALE INTERACTIONS

There were four possible outcomes to an encounter between two (or more)
males: (1) the males ignored one another, (2) they touched one another and one
or both moved away, (3) one male chased another away, and (4) they sparred and
one eventually fled, often chased by the other. Sparring males tapped, pushed and
grappled with one another, using the legs and palps. I never saw injury of any
sort to a male as a result of sparring, nor did I see males with missing
appendages. Sparring appears to be a form of ritualized fighting and occurred
only in the context of male sexual activities. No other form of fighting was
observed. Although males matured before females and were thus present as adults
in the colony for several weeks before courtship began, neither male-male
aggression nor male-male interactions of any other sort, were observed before the
onset of the mating season.

Sparring bouts and chases occurred most frequently when one male interfered
with courtship activities of another either during a display or while attempting
to mate with a female. They also occurred when males encountered one another
during searching and when building a mating thread. A searching male was more
likely than a displaying male to ‘give up’ and move away after an encounter
(Table 2).

LUBIN— SOCIAL SPIDER MATING TACTICS

249

Table 2. — Outcomes of male-male encounters during courtship and attempted copulations (C/AC)
and during searching (S), based on observations of males in 10 colonies. Group sparring events are
excluded. Male #1 is the instigator of the encounter (in searching males) or the resident in encounters
during C/AC. Possible outcomes were: either male #1 or male #2 left first, or there was no clear
winner (both males left or both remained inactive). Expected values are in parentheses. The null
hypothesis that there is no difference in the outcomes of encounters of C/AC and S males is tested
(G test, df=2). [G = 15.713, p < 0.005 (two-tailed)].

Male#l

Male#l

First male to leave

Male #2

Both/ Neither

Totals

C/AC

22 (23.2)

45 (34.0)

15 (24.8)

82

S

21 (19.8)

18 (29.0)

31 (21.2)

70

Totals

43

63

46

152

As a mating tactic, male interference and attempts to displace courting males
appear to be unsuccessful. In no instance did I observe an interfering male gain
access to a female (N = 152 attempts) and these contests generally ended with
one or both males moving away. Nonetheless, there may be other advantages: an
interfering male may disrupt the courtship display of another male sufficiently to
cause him to move away, whereupon the invading male can usurp the display site.
Displacements of this sort occurred, although more frequently the displaying male
chased the intruder away (Table 2). The possibility of size advantage in these
contests (Rovner 1968, Suter and Keiley 1984) was not examined.

Intense sparring bouts occurred over molting females and sometimes clusters
of males could be seen sparring over a molting female (Fig. 6) or over a recent
exuvium. The latter behavior suggests that chemical cues present in the molt
trigger male copulation attempts. A phenomenon that remains largely
unexplained is that of ‘spontaneous’ clustering of sparring males, with no female
to be seen nearby, nor even an exuvium to indicate that a molting female had
been there. Males entered a cluster, engaged in pushing and shoving matches, left
to sit in nearby threads or walked about tapping nearby females, and then
returned to the cluster to resume sparring. Such clusters persisted, with changes
in the number and identities of the contestants, for up to an hour.

How do such sparring contests start and what attracts males to them? Many
sparring contests (55% of 29 contests observed in one colony) began in the
presence of a molting or recently molted female or her exuvium, but continued
long after the females had moved away. Clusters of sparring males also shifted
location in the web as males left and entered them. Perhaps males are attracted
to such clusters initially by chemicals produced by molting females and, later, by
cues (chemical or vibrational) from the sparring males themselves.

Two explanations for the occurrence of these group sparring bouts are
suggested: (1) that they are a product of a state of high sexual “excitement” on
the part of the males and the likelihood of finding a molting female in the center
of such a cluster, and (2) that they involve the establishment of a dominance
hierarchy among males that, in the long run, determines access to females.

The occurrence of clusters of sparring males mainly early in the mating season
when molting females are present in colonies supports both explanations. During
this period males may be highly motivated to search for and attempt to mate with
molting females (see following sections for the reasons why this might be the
case). Likewise, a dominance hierarchy would be established early in the period
of sexual activity, as females are maturing and molting to adults. The presence

250

THE JOURNAL OF ARACHNOLOGY

Fig. 6. — A group of males of A. wau sparring over a recently molting female. The female is able
to move her legs and is pushing the males away.

of a dominance hierarchy implies individual recognition, a possibility which was
not examined. Further observations, particularly of marked males, are clearly
necessary.

AVing occurred in the context of male aggression (1) when several males
attempted to mate with a molting female, (2) during sparring bouts and (3)
during interference with a male that was courting or attempting to copulate.
AVing during male-male interactions appeared to be of lower frequency and
higher amplitude (slower and more pronounced oscillations of the abdomen) than
while courting a female. Possibly different messages are conveyed by differences
in the frequency and amplitude of the vibrations produced in AVing. Abdomen
twitching was observed in both courtship and male aggression in a salticid spider
(Jackson 1978a) and, indeed, the use of similar signals of different intensities in
both courtship and male-male aggression is common in many animals (c.f.. Cade
1979).

LUBIN— SOCIAL SPIDER MATING TACTICS

251

FEMALE BEHAVIORS

Mature, non-molting females responded to male courtship by tapping him with
legs I and 11. If the female was receptive, she adopted an acceptance posture, in
which she hung from the threads (the male’s mating threads?) by legs IV, while
legs I and II were held together either flexed over the cephalothorax or
downward and away from the body. As the male attempted to mate, the female
often held him with one or both legs III.

If a male persisted in drumming on the female’s ventral side without succeeding
in inserting the palp, the female then began to push him away with the legs, at
first gently, then more sharply and sometimes ending in a lunge at the male and
an apparent attempt to seize and bite him. The male reacted to the pushing by
either attempting to follow the female and scramble back onto her or moving
away, and to the lunge, by dropping several centimeters into the lower barrier
web. I never saw a female actually seize a male.

Much of the time, nearby females showed no response to courting males. When
a searching male approached and touched a female, she either remained
motionless or moved away. If the male persisted in following, she moved into the
lower barrier threads, or even under the sheet, pushing the male away with her
hind legs. Later in the reproductive season, females aggressively lunged at
courting or pursuing males and chased them into the lower barrier web. As in
the case of male-male sparring, I never saw any damage to a male thus ‘attacked’
by a female. Bonnet (1935) described similar behavior in Achaearanea
tepidariorum where older, mated females chased courting males aggressively, but
never harmed them.

COPULATIONS WITH MOLTING FEMALES

Molting females hung by legs I-HI, while their flexed legs IV held onto the web
near the cephalothorax. After molting, the spider remained hanging by a thread
from the exuvium in an inverted posture, such that all legs hung down and were
slightly flexed. One molt from last instar to adult female lasted 23 min from the
start of pumping movements to the first coordinated leg movements of the adult
female. During this period the female was defenseless. Another 22 min elapsed
before the female broke the attachment to the exuvium and moved away.

During the molt from penultimate instar to adult, females were sometimes
mated by one or more males. As many as 15 males were observed clustered
around a single molting female. I saw four instances of multiple copulations, one
of which involved four males in succession. Sometimes one male succeeded in
mating 2-3 times with the same molting female. This suggests that the male did
not transfer all of the sperm in one palp in the first insertion. Copulations with
molting females were terminated by the male, or as a result of interference from
other males, and were of longer duration than copulations with mature females
following courtship (molting females: n = 19, median = 60 sec, range 15-540 sec;
mature females: n = 9, median = 18 sec, range 6-129 sec; 0.01 < p <0.025,
Wilcoxon two-sample test).

Of 140 females observed molting in 8 different colonies, 40.7% were discovered
by males, and of these 59.6% were mated during molting, corresponding to 24.3%

252

THE JOURNAL OF ARACHNOLOGY

of all molting females. Females that molted at night in the vicinity of the leaf
retreats were more likely to be discovered and mated than were individuals that
molted during the day and away from the retreats. In one colony in which
molting females were observed throughout the mating season, 4 of 6 females that
molted at night were discovered by males who attempted to copulate with them,
whereas only 14.8% of 81 females that molted during the day were discovered
by males. Furthermore, about 2/3 of the molting females (109 of 152) were
observed during the day (0600-1600), a distribution that reflects a real preference
for molting during the day. Of 64 females that molted near leaf retreats, 52%
were discovered by males who then attempted to copulate with them, whereas
only 1 (2%) of 45 females that molted in the upper barrier web was discovered
(x2 = 29.81; p <0.001, df= 1).

There is an immediate risk to a female to being discovered and copulated with
during molting, aside from possible long-term disadvantages involving loss of
reproductive fitness. As the female’s exoskeleton begins to harden, it becomes
difficult for a copulating male to withdraw the embolus and he may become
‘stuck.’ Upon separating one such pair, I noticed a sticky secretion of unknown
origin covering the embolus and epigynum. As soon as the female was capable
of moving, she attempted to free herself from the copulating male by pushing him
with her legs III and IV. The male at first remained passive, then attempted to
withdraw the palp. Pairs often remained stuck for more than 12 hours. Males
became stuck in 40% of 40 observed copulations with molting females and on
at least four occasions (10% of all matings with molting females), both members
of the pair died while still in copula.

The risk of discovery by males during molting may be a strong selective
pressure for molting in locations and at times when males are least likely to be
active. Females molting during the day in the upper barrier web may be exposed
to risks of predation (by kleptoparasitic spiders of the genus Argyrodes
(Theridiidae), mimetids and salticids as well as by wasps and birds.) Although I
did not observe predation directly, I did find Argyrodes feeding on dead A. wau
individuals in the barrier web. Furthermore, females were found in the upper
barrier threads during the day only in two circumstances: during rare excursions
to capture prey or during the last molt. All other (pre-adult) molts occurred in
the vicinity of the leaf nests, and predominantly at night.

Matings with molting females give the appearance of forced copulations (a
term less emotionally charged than ‘rape;’ Thornhill and Alcock 1983). First,
molting females have no control over the mating; second, as soon as females can
move, they attempt to free themselves of the male; and third, females molt in
circumstances that minimize the chance of being found by males. Forced
copulations may be undesirable for females because (1) they allow no choice of
mate and (2) they carry a high risk (higher for the females, who loses all, than
for the male, who may have already mated with other females).

DISCUSSION

Variable Mating Tactics. — Intraspecific variation in male mating tactics is to be
expected when the ‘interests’ of males and females do not coincide exactly, as is
the case when females are choosier than males in accepting mates (Maynard

LUBIN— SOCIAL SPIDER MATING TACTICS

253

Smith 1982). The most thoroughly studied example of variable male mating
tactics in nonsocial spiders is that of a salticid, Phiddipus johnsonii (Jackson
1977, 1978a). Males of this species have three distinct tactics: they may court
females outside their nests with a species-typical, visual display; search for and
mate with females in nests, using vibratory courtship only; or cohabit in nests
of subadult females and copulate with a minimum of ceremony when females
molt to maturity. Unlike A. wau, females of P. johnsonii can terminate
copulations both in the nest and outside it. Thus, although males have alternative
mating tactics which differ in their probabilities of success and in the durations
of the resulting copulations, there is apparently no tactic by which males
circumvent female control of mating.

In many araneids that have large sexual size dimorphism (e.g. the Nephilinae
and some species of Argiope), the small males either court the female from a
distance on her own web, wait to copulate when the female is busy with prey,
or cohabit in webs of subadult females and copulate as soon as the female molts
to adulthood (Farr 1977, Robinson and Robinson 1980, Alcock 1984b). The
tactic of waiting until the female is feeding before approaching occurs in many
species, e.g. Meta segmentata (Clerck) (Araneidae), Zosis geniculatus (Oliver)
(Uloboridae), and Pisaura mirabilis (Clerck) (Pisauridae), where the male
provides a ‘gift’ of food before copulating (Gerhardt 1927, Bristowe 1958, Blanke
1974). Although females control mating in these cases, the likelihood of the male
being rejected may be reduced. Matings with recently molted females have been
reported sporadically in the literature, e.g. in a gnaphosid. Drassodes lapidosus
(Walckenaer) (Bristowe 1958, p. 119) and a clubionid, Clubiona robusta L. Koch
(Austin 1984). It is not clear from these reports if females are indeed capable of
rejecting males under these circumstances.

In these examples and in A. wau, a species-typical male courtship tactic is
replaced by a more direct tactic under specific conditions, namely, when the
likelihood of being rejected by the female is reduced because she is either
occupied or incapable of doing so. In contrast to the examples cited above,
opportunistic mating in A. wau takes an extreme form of forced copulation of
females that are in the process of molting (and are completely defenseless), and
often by several males in succession. The frequent occurrence of these behaviors
in A. wau (Lubin, ms.) suggests that female choice acts to increase the variation
in male mating success and that this provides strong selection on males to adopt
alternatives to courtship.

Male Fighting. — Male-male fighting occurs as well in other web-building
species, particularly in those in which males cohabit in females’ webs before or
after copulation (or both). Males of Nephila clavipes (Linnaeus) for example,
fight for a position at the hub of the female’s web (where access to the female
is easiest) and interfere with one another during mating attempts (Farr 1977,
Christenson and Goist 1979, Vollrath 1980). In the linyphiid, Frontinella
pyramitella, a male on a female’s web defends it against intruding males before
and after he copulates (Austad 1982, Suter and Keiley 1984) as may be the case
in some pholcids (Eberhard and Briceno 1983). Ritualized aggression occurs in
all of these species. Often contests are settled before any direct contact is made
and, if fighting breaks out, damages to the contestants occur infrequently. Austad
(1982) and Suter and Keiley (1984) suggested that intruding males of F.
pyramitella are often unable to assess accurately the ability of the resident male

254

THE JOURNAL OF ARACHNOLOGY

to defend the web and may do so only by increasing the intensity of the
aggressive interaction up to the point of a physical contest.

Male-male interactions in A. wan differ from those described above: of the
three types of interactions seen in A. wau — interference and displacement during
courtship and mating attempts, fights over access to molting females and group
sparring contests — only the first is similar to behaviors seen in non-social species.
Nonetheless, sparring over molting females and group fights may also be derived
from male-male interactions typical of non-social species. The highest intensity
interaction seen between males of A. wau was sparring, a ritualized fighting that
apparently never caused damage to the participants and at most resulted in one
male chasing another away. Sparring bouts were often preceded by, and
interspersed with, AVing, which in this context appears to be an aggressive signal
of lower intensity. Escalation of aggression from AVing to sparring carried
virtually no immediate risk of damage to the contestants in this social species;
there already exist strong inhibitions to damaging one another. The outcome of
a contest may depend on the stamina or persistence of the contestants rather than
on their ability to do damage. The reward may be immediate access to a molting
female, or as suggested earlier (and if individual recognition occurs), a higher
rank in a male dominance hierarchy which could determine access to females
later in the season.

Evolution of the Mating System of Achaearanea wau. — Comparable studies of
mating behaviors of other social spiders are lacking. Krafft (1969) observed a
male Agelena consociata Denis (Agelenidae) drumming his palps (on the web?)
as he approached a female, but the female moved away; Mallos gregalis (Simon)
(Dyctinidae) males courted by plucking the web and twitching their abdomens,
but the mating system was not described (Jackson 1978b); there are no published
observations of courtship or mating behavior in another social theridiid,
Anelosimus eximius Simon, although this is perhaps the best studied of all the
social spiders (Brach 1975, Vollrath 1982, Vollrath and Rohde-Arndt 1983, Tapia
and de Vries 1980, Christenson 1984). However, Vollrath (pers. comm.) observed
a male A. eximius fight with another male that was attempting to mount a
female. Clearly, none of these social species exhibits the frenetic, synchronized
mating season seen in A. wau. Achaearanea wau is apparently unique among the
social species studied until now in that (1) it has a communal mating season (2)
males have prolonged and complex courtship, (3) forced copulations are
common, and (4) males engage in ritualized sparring contests. Can these
behaviors be explained in the context of the social biology of A. waul

The synchronous, communal mating season in A. wau is a consequence of
colony dynamics and life history. New colonies are established by means of
synchronous swarming of mated females (Lubin and Robinson 1982). The stage
of a colony in its life cycle, rather than seasonal events, determines the timing
of swarming. Communal brood defense also imposes a degree of synchrony:
several days before females are ready to lay eggs, they remove the capture web,
leaving only the leaf nests supported by strong guy threads. The females hang
their eggsacs inside the leaf nests, seal the openings and brood the eggsacs for
1-2 weeks. Those females that cannot produce eggsacs during this period (due
perhaps to lack of resources) are unlikely to reproduce successfully later
(unpublished data).

LUBIN— SOCIAL SPIDER MATING TACTICS

255

The courtship of A. wau males remind one of the displays of males in leks,
and indeed, shares some of the characteristics of lekking behaviors. Males defend
small arenas that are used exclusively for courting females and copulating with
them; these arenas are concentrated in a relatively small area of the spider’s home
range (i.e., the colony); and, a female may visit (and perhaps compare) many
displaying males during a short period. Most lekking vertebrate (c.f. Bradbury
1981) and some lekking social insects (e.g., harvester ants; Holldobler 1976), use
traditional lekking sites. Achaearanea wau males, however, change courtship
arenas frequently and, indeed, the display site itself may be of little importance
except in relation to the position of females in the web.

‘Lekking’ in A. wau males can be understood in the contexts of male mating
strategy and the same colony life history characters discussed above. Given that
a male’s sole investment in his offspring is sperm, his best strategy is to put all
his efforts into finding mates and copulating with them. Finding females in the
colony presents little difficulty, particularly when females within a colony mature
simultaneously. Consequently, during a relatively short mating season, males
display vigorously and almost incessantly in sectors of the colony in which
receptive females are most likely to be found.

Both female choosiness and male eagerness might select (by means of sexual
selection) for elaborate and persistent male displays. Furthermore, given the
female’s apparent reluctance to accept mates, males should adopt opportinistic
mating tactics whenever possible as a means of circumventing female choosiness
(see above). In this respect, copulations with molting females may be particularly
advantageous (in spite of the risks), since a male may be able to completely fill
a female’s spermathecae and ensure that all of her offspring will be his own. As
molting females are a scarce resource, males are expected to compete for them
and fight over them.

The evolution of this mating system from one of a less social or solitary
theridiid must remain speculative, given the paucity of comparable observations
of other species. Achaearanea wau shares certain mating behaviors with other
theridiids {Achaearanea tepidariorum is the best known example), including
elements of the courtship displays (AVing, twanging and post-mount rapping); the
occurrence of numerous, short copulations that are generally terminated by the
female; and a relative lack of aggression between females and cohabiting males
(Bonnet 1935; Montgomery 1903). In addition, males of both A. tepidariorum
and of a subsocial species, A. mundula (L. Koch), congregate on webs of
receptive adult females and often cohabit their webs for several days. I predict
that male-male interactions and competitive male displays occur in these species,
although perhaps less ritualized and less frequent than in A. wau.

The differences between A. wau and these solitary or subsocial species are
related specifically to the social adaptations discussed earlier, low fecundity,
skewed sex ratios and dispersal by swarming of adult females resulting in
developmental synchrony within colonies. I propose that the unique mating
system of A. wau is the result of the modification of a ‘traditional’ theridiid
mating system to accommodate these social adaptations.

256

THE JOURNAL OF ARACHNOLOGY

ACKNOWLEDGEMENTS

This work was funded by a Smithsonian Institution Scholarly Studies Research
Award to M. H. Robinson and the author, and by the Smithsonian Tropical
Research Institute. I am grateful to M. H. Robinson, B. C. Robinson and the
late J. L. Gressitt, to my field assistant, Lukas K., and the staff of the Wau
Ecology Institute for helping in many ways. I thank M. J. Eberhard, W, G.
Eberhard, R. R. Jackson, B. Kotler, R. B. Suter and F. Vollrath for reading and
commenting on drafts of the manuscript, A. Decae for thought-provoking
discussions, and G. W. Uetz for realizing the publication of the symposium.

LITERATURE CITED

Alcock, J. 1979. The evolution of intraspecific diversity in male reproductive strategies in some bees
and wasps. Pp. 381-402, In Sexual Selection and Reproductive Competition in Insects (M. S. Blum
and N. A. Blum, eds.). Academic Press, New York.

Austin, A. D. 1984. Life history of Clubiona robusta L. Koch and related species (Araneae,
Clubionidae) in South Australia. J. Arachnol., 12:87-104.

Austad, S. N. 1982. First male sperm priority in the bowl and doily spider, Frontinella pyramitella
( Walckenaer). Evolution, 36:777-785.

Blanke, R. 1974. Rolle der Beute beim Kopulationsverhalten von Meta segmentata (Cl.) (Araneae,
Araneidae). Forma et Functio, 7:83-94.

Blum, M. S. and N. A. Blum, eds. 1979. Sexual Selection and Reproductive Competition in Insects,
Academic Press, New York.

Bonnet, P. 1935. Teridion tepidariorum C. L. Koch, araignee cosmopolite. Bull. Soc. Hist. nat.
Toulouse, LXVHl:335-386.

Braun, R. 1956. Zur Biologie von Teutana triangulosa (Walck.) (Araneae: Theridiidae, Asagenae). Z.
Wiss. Zook, 159:255-318.

Bristowe, W. S. 1958. The World of Spiders. Collins, London. 304 pp.

Cade, W. 1979. The evolution of alternative male reproductive strategies in field crickets. Pp. 343-
379, In Sexual Selection and Reproductive Competition in Insects (M. S. Blum and N. A. Blum,
eds.). Academic Press, New York.

Christenson, T. E. 1984a. Behavior of colonial and social- of the theridiid species Anelosimus eximius.
Animal Behavior, 32:725-734.

Christenson, T. E. 1984b. Alternative reproductive tactics in spiders. Amer. Zook, 24:321-332.
Christenson, T. E. and K. C. Goist. 1979. Costs and benefits of male-male competition in the orb-
weaving spider, Nephila clavipes. Behav. Ecok Sociobiok, 5:87-92.

Eberhard, W. G. and D. Briceno. 1983. Chivalry in pholcid spiders. Behav. Ecok Sociobiok, 13:189-
195.

Farr, J. A. 1977. Social behavior of the golden silk spider, Nephila clavipes (Linnaeus) (Araneae,
Araneidae). J. Arachnol., 4:137-144.

Gerhardt, U. 1927. Neue biologische Untersuchungen an einheimischen und auslandischen Spinnen.
Z. Morph. Okok Tiere, 8:96-186.

Gwinner-Hanke, H. 1970. Zum Verhalten zweier stridulierenden Spinnen Steatoda bipunctata Linne
und Teutana grossa Koch (Theridiidae, Araneae), unter besonderer Beriicksichtigung des
Fortpflanzungsverhaltens. Z. Tierpsychok, 27:649-678.

Helsdingen, P. J., van. 1965. Sexual behavior of Lethyphantes leprosus (Ohlert) (Araneida,
Linyphiidae) with notes on the function of the genital organs. Zook Mededelser, 41:15-42.

Jackson, R. R. 1977. Courtship versatility in the jumping spider Phiddipus johnsonii (Araneae:
Salticidae). Anim. Behav., 25:953-957.

Jackson, R. R. 1978a. An analysis of alternative mating tactics of the jumping spider Phiddipus
johnsonii (Salticidae, Araneae). J. Arachnol., 5:185-230.

Jackson, R. R. 1978b. Male mating strategies of dictynid spiders with differing types of social
organization. Symp. Zook Soc., London, 42:79-88.

Jackson, R. R. 1980. The mating strategy of Phiddipus johnsonii (Araneae, Salticidae): H. Sperm
competition and the function of copulation. J. Arachnol., 8: 217-240.

LUBIN— SOCIAL SPIDER MATING TACTICS

257

Jackson, R. R. 1982. The behavior of communicating in jumping spiders (Salticidae). Pp. 211-247,
In Spider Communication, Mechanisms and Ecological Significance (P. N. Witt and J. S. Rovner,
eds.). Princeton Univ. Press, Princeton, New Jersey.

Jacson, C. C. and K. J. Joseph. 1973. Life-history, bionomics and behavior of the social spider,
Stegodyphus sarasinorum Karsh. Insects Sociaux, 20:189-204.

Levi, H. W. 1967. Predatory and sexual behavior of the spider Sicarius (Araneae: Sicariidae). Psyche,
74:320-330.

Levi, W. H., Y. D. Lubin and M. H. Robinson. 1982. Two new Achaearanea species from Papua
New Guinea, with notes on other theridiid spiders (Araneae: Theridiidae), Pacific Insects, 24:105-
113.

Locket, G. H. 1926. Observations of the mating habits of some web-spinning spiders. Proc. Zool. Soc.
London, 22:1125-1146.

Lubin, Y, D. 1982. Does the social spider, Achaearanea wau (Theridiidae), feed its young? Z.
Tierpsychol., 60:127-134.

Lubin, Y. D. and R. C. Crozier. 1985. Electrophoretic evidence for population differentiation in a
social spider, Achaearanea wau (Theridiidae). Insectes Sociaux, 32:297-304.

Lubin, Y. D. and M. H. Robinson. 1982. Dispersal by swarming in a social spider. Science, 216:319-
321.

Maynard Smith, J. 1982. Evolution and the Theory of Games. Cambridge Univ. Press.

Montgomery, T. H. 1903. Studies on the habits of spiders, particularly those of the mating period.
Proc. Acad. Nat. Sci., Philadelphia, 1903:59-149.

Robinson, M. H. and Y. D. Lubin. 1979. Specialists and generalists: The ecology and behavior of
some web-building spiders from Papua New Guinea. 11. Psechrus argentatus and Fecenia sp.
(Araneae: Psechridae). Pacific Insects. 21:133-164.

Robinson, M. H. and B. Robinson. 1980. Comparative studies of the courtship and mating behavior
of tropical araneid spiders. Pacific Insects Monogr., 36:1-218.

Ross, K. and R. L. Smith. 1979. Aspects of the courtship behavior of the black widow spider,
Latrodectus hesperus (Araneae, Theridiidae) with evidence for the existence of a contact sex
pheromone. J. ArachnoL, 1:69-11 .

Rovner, J. S. 1968. Territoriality in the sheetweb spider, Linyphia triangularis (Clerck) (Araneidae,
Linyphiidae). Z. Tierpsychol., 25:232-242.

Savory, T. H. 1928. The Biology of Spiders. Macmillan Co., New York, 376 pp.

Suter, R. B. and M. Keiley. 1984. Agonistic interactions between male Front inella pyramitella
(Araneae: Linyphiidae). Behav. Ecol. Sociobiol., 15:1-7.

Suter, R. B. and G. Renkes. 1984. The courtship of Frontinella pyramitella (Araneae, Linyphiidae):
patterns, vibrations and functions. J. ArachnoL, 12:37-54.

Thornhill, R. and J. Alcock. 1983. The Evolution of Insect Mating Systems. Harvard Univ. Press,
Cambridge.

Vollrath, F. 1980. Male body size and fitness in the web-building spider Nephila clavipes. Z.
Tierpsychol., 53:61-78.

Vollrath, F. 1982. Colony foundation in a social spider. Z. Tierpsychol., 60:313-324.

Vollrath, F. and D. Rohde-Arndt. 1983. Prey capture and feeding in the social spider Anelosimus
eximius. Z. Tierpsychol., 61:334-340.

Manuscript received October 1984, revised January 1986.

■ ii

JJL '^o-waNTL-:* •■;♦.>;■■' ■■ ■' ''*'^'.s

_-Vv.#£;.-- • '../, ;< i' '■.‘■•«« ,a

:f J '■.

■ -f

v ■”■■■/, '■

‘■I*

■ (*^. v^-«]l.., ■ >tV-‘ ='JP.

tvr.-it> 'iM M

%«,,■■*•• I . ;;

S

• ■* ; ’

ii •.

3 .'1?

• ." J. ’

1-. « -1 - '

' . •»

■V .. .J

•'fr‘fi'’ii^%.'

, , . 4".0,-‘ '

' - -n- ■ ■■ '41?#

' *

‘ ■' ■ 2’**

* ■-■* -*

fir. ,)■ '

.4l6i^W5 ■ H *-«

ioifi‘^tt|.in^ ^ liN * :*.^i

t" „.Jt'

• !- ‘> ixm • « *' ^

»« i‘ti’'rt^i*t'<ii7i/%Ui

■’ '■ --VaS

: '.j.J*l8

-^srwttMo

.( '. s ) r

’ ‘- ■■ ■ ■ V J I ft • * '^

Buskirk, R, E. 1986, Orb-weaving spiders in aggregations modify individual web structure. J.
Arachnol., 14:259-265.

ORB-WEAVING SPIDERS IN AGGREGATIONS
MODIFY INDIVIDUAL WEB STRUCTURE^

Ruth E. Buskirk

Institute for Geophysics and Brackenridge Field Laboratory
University of Texas at Austin
Austin, Texas 78713

ABSTRACT

Orb-weaving spiders that live in groups modify their web structure and activity periods, in contrast
to solitary araneids that build more individually distinctive orbs. Individually marked spiders (n =
62) of the colonial Metabus gravidus (Araneidae) in Costa Rica were monitored for 5-day periods
during which all webs spun were measured. Variation in web characters (particularly web angle and
viscid spiral measures) within an individual is related to time of day and degree of aggregation.
Comparisons of six orb-weaving species suggest that spiders joining aggregation vary their web
characteristics when in groups.

INTRODUCTION

The geometric orb webs of spiders in the families Araneidae, Uloboridae, and
Tetragnathidae usually have characteristics that are typical of the genus or species
(Risch 1977, Eberhard 1982). In addition, individual spiders that build their orbs
in a controlled environment, with standarized web support structures and
attachment sites, often build orbs of the same size and pattern day after day (Witt
et al. 1968, Peters 1969). Changes in web structure with age of spider, molting
condition, nutritional status, configuration of attachment sites and time of day
have all been documented especially in laboratory work (Szlep 1958, 1961, Witt
et al. 1968, Peters 1970, Eberhard 1972, Ramousse and LeGuelte 1984).
Variability of web structure under natural conditions has not been as well
quantified.

Previous observations of the colonial spider Metabus gravidus (Cambridge)
(Araneae, Araneidae) in Costa Rica (Buskirk 1975a) suggested that orb structure
was highly variable within an individual, unlike the comparatively predictable
orbs found in laboratory studies of other araneids (e.g., Witt et al. 1968). For
this study I gathered data on webs of these colonial spiders, both those in
aggregations and a few isolated individuals, to determine the effect of colony
participation on orb structure. For additional comparison I analyzed webs of
other orb-weaving species which occasionally formed aggregations. This study
determines the extent of individual variation in orb webs in different species and
its relation to the tendency of the species to form aggregations.

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

260

THE JOURNAL OF ARACHNOLOGY

METHODS

The colonial orb-weaving spider M. gravidus is abundant at the study site
chosen on the Rio Guacimal (Monteverde, Costa Rica, elevation 1350 m).
Individually marked adult or subadult females were monitored hourly for periods
of 5 days in June or July (rainy season), during which every orb spun was
measured. Each of 62 individuals built at least 5 orbs, and some as many as 14
orbs, during the observation periods. About 15% of these orbs were webs rebuilt
in the same site. In this process the old web is destroyed, and the spider
immediately reinforces support lines and radii then completely renews the viscid
spiral. In all webs I measured radius, number of radii, number of spiral turns,
and angle from the horizontal.

RESULTS FOR A COLONIAL SPECIES

Orb web measurements of individual M. gravidus in colonies can vary more
than 30% within a period of a day or two. For example, the web formula
consisting of (A) number of radii, (B) number of spiral turns and (C) orb radius
in cm varied as follows for one adult female (day and time. A, B and C,
respectively): day 1 (07:00 h) — 13, 18, 15.0; day 1 (17:30 h) — 11, 15, 12.8; day 2
(12:00 h)— 12, 11, 12.2; day 3 (06:30 h)— 9, 16, 10.8; day 3 (17:00 h)— 10, 17, 14.0.
In a sample of webs the variability within individuals was as great as between
individuals for the orb radius and number of radii, as shown below.

One source of high variance in orb measures in M. gravidus is the time of day
at which orbs are built. These spiders may emerge from the retreat and build a
new orb at any daylight hour (Buskirk 1975a) and often respin webs once or
twice during the day. Webs either built or respun in the last two hours of the
diurnal activity period (Evening Webs in Table 1) have significantly smaller radii
(t = 2.39, p < 0.025), fewer radii (t = 2.70, p < 0.01) and larger mesh size (t
= 2.81, p < 0.01) than morning webs. Analysis of variance for radius length and
number of radii, however, indicates that even when morning and evening webs
are distinguished, variability within a spider is high (Table 1).

In 169 web-spinning observations spiders averaged a speed of nearly one cm/
sec in spinning the viscid spiral, but there was much variation over the day. At
dawn (06:00-07:00 h) the sticky thread of the viscid spiral was spun at a quicker
rate than in mid-morning (t = 2.41, p < 0.05) with no significant changes in
radius or mesh. In the evening period (16:00-18:00 h) an increased speed (1.4 cm/
sec) is accompanied by a smaller average radius (9.7 cm) and coarser mesh (6.1
sq. cm). These times of rapid spinning (dawn and dusk) coincide with heightened
insect activity, and the evening period is one of increased number of adult spiders
occupying and building orbs (Buskirk 1975a).

To test the effect of the colony participation on orb variance, I examined web
radius and number of radii in all webs spun by 62 individually marked spiders
during the five day observation period. Ten spiders that were temporarily isolated
and not building in colonies built less variable webs than the other 52 individuals
that built in colonies. For example, variance of the radius length (mean = 13.2
cm) in all orbs spun by a given individual ranged from 0.6-6. 7 in aggregated
individuals but was only 0. 8-3.0 in solitary spiders. Orb radius varied at least 50%

BUSKIRK— ORB WEBS IN AGGREGATIONS

261

Table 1. — Individual variability of orb webs in 62 Metahus gravidus adult female spiders within
a colony (p values are for One-Way ANOVA; ** indicates p < 0.10; NS indicates p > 0.10.) For
time of day analysis time blocks used were 06:00-09:00 h (morning webs), 10:00-12:00 h, 13:00-15:00
h, and 16:00-17:00 h (evening webs).

Morning Webs
Combined:

Evening Webs

Combined:

Radius

X ± s.d.

13.2 ±4.0 cm

11.7 ± 3.7 cm

ANOVA, by spider

** (p<0.05, df 5, 14)

** (p <0.02, df 5, 14)

ANOVA, by date

NS (p >0.10, df 5, 14)

NS (p>0.10, df 5, 14)

ANOVA, by time

All Webs:

** (p <0.02, df 3, 40)

Number of Radii:

X ± s.d.

11.5 ± 4.2

8.2 ± 5.3

ANOVA, by spider

**(p<0.05, df5, 14)

** (p<0.05, df 5, 14)

ANOVA, by date

**(p<0.05, df5, 14)

**(p<0.05, df5, 14)

ANOVA, by time

All Webs: **

(p<0.10, df 3, 40)

Mesh:

X + s.d.

3.9 ± 2.1 cm2

5.8 ± 3.9 cm2

ANOVA, by spider

NS (p> 0.10, df5, 14)

NS (p> 0.10, df5, 14)

ANOVA, by date

NS (p> 0.10, df5, 14)

NS(p>0.10, df5, 14)

ANOVA, by time

All Webs:

NS (p >0.10, df 3, 40)

of the mean radius for over a third of the clustered individuals. Previous
measurements (Buskirk 1975a) indicate that the average orb radius in the
direction of a close neighbor (within 25 cm) was shorter than other radii of the
same orb, presumably as a result of aggressive encounters with the neighbor
during orb construction (Buskirk 1975b).

EFFECTS OF CLUSTERING IN OTHER ORB-WEAVING SPIDERS

In other species of facultatively gregarious spiders, joining an aggregation may
also have an effect on web-building. I observed six other species of orb-weavers
at field sites in Texas and Costa Rica (see Table 2). The species examined
represent spectrum of gregarious tendencies from those like M. gravidus in which
spiders routinely cluster their orbs to those species which rarely cluster. Web
characters chosen for comparison were orb radius, angle of the orb’s plane from
horizontal and presence of any unique structural elements such as barrier webs.
I determined the percent of individuals aggregated from censuses of 30 or 50
individuals of each species. Additional spiders were surveyed for web measures.
Web data (radius and angle) from aggregated individuals were considered
different from those of solitary individuals if they were significant at the p < 0.05
level (Mann-Whitney U test).

As indicated in Table 2 some web characters varied with degree of aggregation.
In Texas webs of Nephila clavipes (Linnaeus) and Mecynogea lemniscata
(Walckenaer) are occasionally found in clusters, especially in open woodland or
edge areas where insects are abundant and web support structures are not limited.
Aggregated females of these species use the support lines of their neighbors as
web attachment points, but they do not generally modify their web by reducing
the knockdown strands or barrier webs. Presence of barrier webs in N. clavipes
may vary, but no consistent pattern in aggregations has yet been determined (L.
Higgins, pers. comm.). Similarly, Metazygia wittfeldae (McCook), which are
usually solitary, do not produce smaller or modified webs when they aggregate.

262

THE JOURNAL OF ARACHNOLOGY

Table 2. — Web characteristics in seven orb weaving species that aggregate occasionally (top of list)
or regularly (bottom of list). Yes = web radius or angle from horizontal was significantly different
in clumped individuals (Mann-Whitney U test, p < 0.05). *= in one case barrier web was reduced.

Species
and Site

Number

Censused

Percent
Sharing
Support Lines

Smaller

Orbs in

Clusters?

Different
Web Angle
in Clusters?

Orb Structure

Modified
in Clusters?

Metazygia wittfeldae
(Monteverde, Costa Rica)

50

6%

No

No

No

Mecynogea lemniscata
(Austin, Texas)

30

7%

No

No

No

Nephila clavipes
(Galveston, Texas)

50

10%

No

No

No*

Tetragnatha elongata
(Austin, Texas)

50

20%

Yes

No

No

Leucauge venusta
(Monteverde, Costa Rica)

50

34%

No

Yes

No*

Cyclosa caroli
(Monteverde, Costa Rica)

50

52%

Yes

Yes

Yes

Metahus gravidus
(Monteverde, Costa Rica)

50

90%

Yes

Yes

No

and the angle of the web continues to vary from just above horizontal to vertical
(Table 2).

Like M. gravidus , Tetragnatha elongata (Walckenaer) build simple wide-
meshed orbs, often over water. In tetragnathid aggregations I found orb radii to
be shorter, but no other structural modifications were noted. Solitary T. elongata
adults prefer to build horizontal orbs, but in aggregations the angle of the web
differed from horizontal. When adult females of Leucauge venusta (Walckenaer)
build in aggregations, they put their orbs in a more nearly horizontal plane and
spin less dense support threads than do solitary individuals. Despite the presence
of such modifications, the orb radius in these aggregated L. venusta webs did not
differ significantly from that of solitary webs. On the other hand, aggregated
individuals of Cyclosa caroli (Hentz) displayed a typical vertical web but
produced smaller orbs.

VARIABILITY IN ORB-BUILDING

Several factors are known to contribute to the dimensions and regularity of
web structures. Reed et al. (1965) and Eberhard (1972) found spiders can regulate
orb structure and central angles via measuring behavior of the first pair of legs.
During ontogeny, web mesh size increased in Araneus diadematus (Clerck) (Witt
and Baum 1960). Mesh width is directly correlated with leg length, especially in
adults (Risch 1977). Benforado and Kistler (1973) found that web radius increases
with body size within an age class, primarily as a result of differential feeding.
In comparison to individuals whose webs are destroyed, spiders permitted to
ingest their web daily produce subsequent orbs up to 15% larger in radius with
the number of spiral turns greater by 17-38% (Witt et al. 1968).

Species-specific patterns of orb-building and placement of radii account for
some differences in variability. For example, Nentwig (1983) found that the mesh
size and distance between spirals were much more variable in A. diadematus than

BUSKIRK— ORB WEBS IN AGGREGATIONS

263

in species such as Zygiella x-notata (Clerck) which insert additional radius lines
midway out from the hub. Web~building traits such as hub construction, spiral
attachment, and placement of barrier webs are typical of the genus (Eberhard

1982) . In species that normally add barrier webs near the orb, individuals joining
aggregations can construct webs with reduced or no barrier webs. Since barriers
are usually constructed later than the frame and initial orb, sometimes hours or
days later, the spider could receive feedback from the behavior of neighbors or
prey capture rate that prevents barrier construction.

Orb size can depend upon the size of available area in the substrate for frame
attachment. Adult A. diadematus in laboratory frames produced daily webs that
varied in radius by less than 5% (Witt et al. 1968). The number of radii (means
+ s.d. = 26.1 + 4.2) and number of spiral turns (21.4 + 8.2) were more variable.
Web radius and shape are modified, however, by A. diadematus as frame size
decreases, while the angle of the web is last to change (Szlep 1958). In a field
study of adult Argiope trifasciata (Forskal) and A. aurantia (Lucas), Brown
(1981) found some variance in web radius, but variation was much greater for
other measures such as web height. Studies of adult individuals of solitary orb-
weavers show relatively little variance in web radius under standarized conditions.

SIGNIFICANCE FOR SPIDER SOCIAL BEHAVIOR

Two general evolutionary pathways to sociality in spiders have been proposed,
one emphasizing prolonged parental care and cooperation within a cohort of
young and one depending upon tolerance and aggregation of unrelated
individuals (Shear 1970). Because orb-weavers must build orbs individually the
degree of cooperation in producing a communal web is limited (Buskirk 1981).
One might predict that stereotyped, individual web-building would be more
efficient and would confer a selective advantage to orb-weaving spiders in general.

On the other hand, there are ecological advantages for building webs in
aggregations. Field observations and manipulations of social and facultatively
social spiders have confirmed that spiders are more likely to aggregate in rich
prey patches and obtain as much or more food per individual in these groups
(Buskirk 1975a, Uetz et al. 1982, Rypstra 1983, 1985, Riechert 1985). The
tendency to cluster is strongest when food and web attachment sites are patchy.
Long-term aggregation may also provide protection from predators or climatic
changes. In addition, web clustering may result in more silk lines per individual
for entangling and slowing down prey (Burgess 1978).

In aggregations the higher density of spiders will be costly if there is much
effort spent in defending individual areas and if many supplantings of individuals
from their webs occurs. If species recognition and tolerance has evolved in the
behavioral repertoire or if high prey capture rates induce tolerance (e.g., Rypstra

1983) , then individual spiders can successfully occupy the aggregated webs for
long periods of prey capture.

Besides the orb-weavers addressed in this study, other araneids show some
modification of individual web structure when aggregated in favorable resource
sites. Metepeira spinipes (Cambridge) build orbs in aggregations of up to
hundreds of individuals, and group size is larger in areas of greater prey
availability (Uetz et al. 1982). Larger colonies are more dense, with smaller

264

THE JOURNAL OF ARACHNOLOGY

individual web areas, and individuals share space web and support threads (Uetz
and Burgess 1979). The congeneric Metepeira datona (Chamberlin and Ivie) in
the Bahama Islands sometimes form aggregations and may share retreats
(Schoener and Toft 1983). Web-sharing usually involves spiderlings or males
associating with large females, not aggregations of adult females. Solitary webs
of large spiders tend to be vertical and oriented to minimize the wind, while those
in aggregations are more variable in position.

Spiders that join aggregations can modify their web-building behavior to
increase both silk efficiency and tolerance of neighbors. In the groups surveyed
in Table 2 species that are regularly found in clusters show more modification
of orb structure. Patchy resources account for the ecological advantages of group
living in these spiders. Behavioral adaptability, in the form of variability in web
construction, allows these species to take advantage of the ecological
opportunities. Species with greater ability to modify individual foraging strategies
are more likely to be facultatively gregarious.

ACKNOWLEDGMENTS

I gratefully acknowledge the research facilities of the University of Texas
Brackenridge Field Laboratory and the Organization for Tropical Studies that
made this study possible. Parts of the research were supported by the Whitehall
Foundation and the Bache Fund. I thank Linden Higgins for helpful discussion
and George Uetz for organizing the Social Spider symposium of June 1984.

LITERATURE CITED

Benforado, B. J. and K. H. Kistler. 1973. Growth of the orb weaver, Araneus diadematus, and
correlation with web measurements. Psyche, 80:90-100.

Brown, K. M. 1981. Foraging ecology and niche partitioning in orb-weaving spiders. Oecologia,
50:380-385.

Burgess, J. W. 1978. Social behavior in group-living spider species. Symp. zool. Soc. London, 42:69-
78.

Buskirk, R. E. 1975a. Coloniality, activity patterns and feeding in a tropical orb-weaving spider.
Ecology, 56:1314-1328.

Buskirk, R. E. 1975b. Aggressive display and orb defense in a colonial spider. Metabus gravidus
(Araneae: Araneidae). Anim. Behav., 23:560-567.

Buskirk, R. E. 1981. Sociality in the Arachnida. Pp. 281-367, In Social Insects (H. R. Hermann, ed.).
Vol. II. Academic Press, New York.

Eberhard, W. G. 1972. The web of Uloborus diversus (Araneae: Uloboridae). J, Zook, 166:417-465.
Eberhard, W. G. 1982. Behavioral characters for the higher classification of orb-weaving spiders.
Evolution, 36:1067-1095.

Nentwig, W. 1983. The non-filter function of orb webs in spiders. Oecologia, 58:418-420.

Peters, H. M. 1969. Maturing and coordination of web-building activity. American ZooL, 9:223-227.
Peters, P. J. 1970. Orb web construction: interaction of spider {Araneus diadematus Cl.) and thread
configuration. Anim. Behav., 18:478-484.

Ramousse, R. and L. LeGuelte. 1984. Strategies de construction de la toile chez deux especies
d’araignees {Araneus diadematus et Zygiella x-notata). Rev. Arachnol., 5:255-265.

Reed, C. F., P. N. Witt and R. L. Jones. 1965. The measuring function of the first legs of Araneus
diadematus Cl. Behaviour, 25:98-1 19.

Riechert, S. E. 1985. Why do some spiders cooperate? Agelena consociata, a case study. Florida Ent.,
68:105-116.

Risch, P. 1977. Quantitative analysis of orb web patterns in four species of spiders. Behav. Genet.,

7:199-238.

BUSKIRK— ORB WEBS IN AGGREGATIONS

265

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. 1985. Aggregations of Nephila clavipes (L.) (Araneae, Araneidae) in relation to prey
availability. J. Arachnol., 13:71-78.

Schoener, T. W. and C. A. Toft. 1983. 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.

Shear, W. A. 1970. The evolution of social phenomena in spiders. Bull. British Arachnol. Soc., 1:65-
76.

Szlep, R. 1958. Influence of external factors on some structural properties of the garden spider
{Aranea diademata) web. Folia Biol. 6:287-306.

Uetz, G. W. and J. W. Burgess. 1979. Habitat structure and colonial behavior in Metepeira spinipes
(Araneae: Araneidae), an orb weaving spider from Mexico. Psyche, 86:79-89.

Uetz, G. W., T. C. Kane and G. E. Stratton. 1982. Variation in the social grouping tendency of a
communal web-building spider. Science, 217:547-549.

Witt, P. N. and R. Baum. 1960. Changes in orb webs of spiders during growth {Araneus diadematus
Clerck and Neoscona vertebrata McCook). Behaviour, 16:309-318.

Witt, P. N., C. F. Reed and D. B. Peakall. 1968. A Spider’s Web. Springer- Verlag, New York. 107

pp.

Manuscript received June 1985, revised January 1986.

'fU'

jfejaitr '•*/?■ _

rv.-sai '

"vn ■■r^i ..A.iA li' -’^J

j<’*.

rr^t-'Mi V S“t

Ita

I

“ti

|\|* ■#♦^1* -Hi ■ » . < >■ ..j

’'5 ;

„>\.‘/ f.— ■,'..> •!!>«*• : • ,f-*‘-''.^l' '»»**- ''^.J.f*^ '"®^ -- :'>

S-.

Voilrath, F. 1986. Environment, reproduction and the sex ratio of the social spider Anelosimus
eximius (Araneae, Theridiidae). J. Arachnol., 14:267-281.

ENVIRONMENT, REPRODUCTION AND THE SEX RATIO OF
THE SOCIAL ANELOSIMUS EXIMIUS

(ARANEAE, THERIDIIDAE)*

Fritz Voilrath

Smithsonian Tropical Research Institute
Balboa, Republic of Panama^

ABSTRACT

Three populations of the neotropical theridiid Anelosimus eximius were studied, one in the
lowlands were the year shows distinct seasonal cycles, a second in the intermediate uplands, and a
third in the mountains where seasonal changes in climate are less pronounced. Colonies in the
lowlands showed a large monthly variation in egg production, presence of immature spiderlings and
colony biomass. Colonies in the mountains showed less variation. In all sites the sex ratios were
variable but were always skewed in favor of females. The sex ratio bias was independent of seasonal
factors. It was presumably also independent of other environmental factors, with the exception of
predation, which apparently was suffered more by females.

INTRODUCTION

Even in the tropics, seasonal change in the environment has an observable
influence on the presence of many short lived organisms. In tropical Panama
many terrestrial invertebrates like insects (Wolda 1977, 1978) and spiders
(Robinson and Robinson 1970, Lubin 1978, 1980) show clear annual cycles or
seasonal depressions in activity and/or abundance. Towards the end of the dry
season, for example, very few spiders are present in the lowlands whereas they
are quite abundant at most other times of the year. Colonies of the social spider
Anelosimus eximius, however, persist for many years at the same location and
are active in the dry season when most other spiders have vanished.

The effects of the environment on A. eximius were studied by comparing
colonies in several sites with different degrees of seasonal change. Colonies
generally raise about two generations a year and special interest was taken in the
reproductive aspect of a colony’s cycle, including egg production and the ratio
of the sexes. Like in other social spiders (Buskirk 1981) but unlike the solitary
spiders (in Lit.), the sex ratio of A. eximius is heavily skewed in favor of females
(Christenson 1984, Voilrath 1985, Aviles in press).

Anelosimus eximius (Keyserling) is a neotropical social theridiid (Levi 1956,
1972). Many individuals share a communal web (nest, colony) and cooperate in
web-building and prey capture (Simon 1891, Brach 1975, Tapia and DeVries
1980, Voilrath 1982, Christenson 1984). Colonies forming part of a cluster are

'Presented in a symposium, “Social Behavior in Spiders,” at the National Meeting of the American
Arachnological Society, June 17, 1984.

^Mailing address: Department of Zoology, South Parks Road, University of Oxford, Oxford, 0X1
3PS, UK.

268

THE JOURNAL OF ARACHNOLOGY

Fig. I. — Histogram displaying the size distribution of Anelosimus eximius collected in two colonies
in Panama. The length of the patella-tibia of the first leg was measured. Mature males were found
in size 2.4 mm, mature females from size 3 mm upwards. The roman numerals above the columns
indicate the instars (beginning with I at eclosion from the egg sac).

founded by budding, or by one female which may be joined by others
(Christenson 1984), all having been inseminated in the parental nest (Vollrath
1982). Colonies far from other colonies are presumably founded by single females
since emigrant females seem to travel alone (Vollrath 1982). A well established
colony can grow to contain many thousand individuals in one nest, and a colony
cluster with its offshoot nests may persist for many years in one site (Robinson
pers. comm., Vollrath unpubl.). However, few single female foundations are
successful and colony clusters are few and far between (Vollrath 1982,
Christenson 1984).

Adult males are smaller and weigh less than adult females (Levi 1963). It
appears that after emergence from the egg case the male undergoes four molts
to reach maturity, whereas the females mature in the fifth or sixth instar (Fig.
1). Both sexes reach maturity about two months following eclosion, (Aviles in
press) although growth can be slightly faster or considerably slower (Overal and
Ferreira 1982). In the field adult females have been observed to live for
approximately three months (Aviles in press), and under laboratory conditions
they may live up to 4 months, males lived up to two months in the lab (Vollrath
unpubl.).

The egg cases are spheres 4 mm in diameter, covered with tough grey silk and
containing from 17 to 53 eggs, depending on the locality and the nutritional state
of the colony (Vollrath unpubl.). The spiderlings are ready to leave the egg case
after about 20 to 30 days (Overal and Ferreira 1982, Aviles in press). They
emerge readily when the average relative humidity is high, but when it is low they
stay inside the protective egg case and may even die there. Egg sacs are
constructed in the center of the colony and are tended (guarded) there by
subadult or adult females. Tending entails carrying the egg sac in the chelicerae
and manipulating it with the mouth parts (Christenson 1984). It is possible that
the spiderlings cannot open the egg sac by themselves, and need a female’s help.
Egg sacs may be cannibalized — opened by members of the colony and the
contents eaten (Christenson 1984).

All members of the colony from instar three onward (with the notable
exception of the adult males) contribute towards colony maintenance, web
construction and repair, and prey capture (Tapia and DeVries 1980, Christenson
1984). Like most male spiders, males of A. eximius leave a dragline behind when

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMUS 269

moving about but they do not appear to do so in the orderly fashion of an
immature or a female Anelosimus repairing the nest or the capture threads. Male
prey only on small insects which they do not share. Males do, however, join
female feeding aggregations. Spiderlings of the first instar may feed on prey
captured and predigested by older instars. More often they are seen soliciting and
receiving regurgitation fluid from females (compare Kullmann and Kloft 1968,
Lubin 1982).

In Panama, A. eximius is found at forest edges or inside the forest, in clearings
or tree falls, where the webs are built at heights between 1 to 6 meters. The
communal web consists of a horizontal, tightly woven bowl (the retainer web)
with a loose network of vertical threads (the snare) above. Flying or jumping
insects are arrested in their path by the vertical threads and tumble onto the
retainer web, where they are chased by the spiders. The bowl section of the web
consists of many layers of silken threads, laid down continuously during the
entire life of the nest. It is remarkably strong, comparable to a fine fabric. Holes
ripped by falling twigs or cut by foraging ants are quickly repaired. In the dry
season, the spiders are less active, and leaves accumulate in the web changing the
appearance of the colony (compare Fig. 2 in Simon 1891; Fig. 1 in Brach 1975;
Fig. 1 in Vollrath 1982; Fig. 1 in Christenson 1984).

The numbers of spiders in a colony (colony size) is correlated with the size of
the bowl webbing (Fig. 2), which enabled me to estimate the spider population
of colony clusters from the size of individual webs. The estimate for one colony
complex in El Valle was about 40,000 spiders in 37 colonies (nests), stretching
30 meters along the edge of a small forest. The estimated mean colony size was
1130 spiders (± sd 1150).

In the same habitat, and often adjacent to colonies of A. eximius, one may
find single nests of the subsocial congeneric Anelosimus juncundus (O. P.-
Cambridge). On rare occasions, as observed in the El Valle highlands, the small
individual webs of single A. juncundus females were joined by a dense network
of threads, thus forming large webs containing several mature females and their
offspring. These webs correspond in their dimensions to colonies of A. eximius
(Fig. 2) and closely resemble nests of this species. Adult females in aggregations
of A. juncundus are highly aggressive towards one another and spread themselves
out inside the shared web structure. The immatures, in contrast, mingle freely.
In A. eximius individuals of all stages seem to be attracted to one another, during
the day they form dense aggregations of spiders in the retreats.

MATERIALS AND METHODS

The data presented here were collected during a study of A. eximius in Panama
(April 1978-May 1979), where three sites were studied in detail: (a) the lowland
semi-decidous forest (altitude 30 meters above sea level) on the Bohio Peninsula
(79° 49', 9° IT) and the Pipeline Road Site (Parque Nacional Soberania 79°45',
9° 10'), (b) a similar site at a higher elevation (300 meters) at the El Llano-Carti
Road (79° 4', 9° 15'), and (c) two locations in the cloud forest near El Valle
(elevation 880 meters, 80° 9', 8° 37'). Colonies were also collected from Anchiote
Road (80°, 9° 29'), Mahe-Bayano (78° 49', 9°81') and El Cope (80° 27', 8° 38').

The climate in lowland Panama has two distinct seasons: a rainy season from
May to December and a dry season from January to April. Although neither air

270

THE JOURNAL OF ARACHNOLOGY

Fig. 2. — Scattergram and correlation coefficient (r) of the size (volume) of the retainer web and
the number of Anelosimus eximius inhabiting it. Volume was calculated using the equation 2:3
phi*a*b*c (a = height of bowl, b = width, c = length of bowl). The formula of the regression line
was y = x*a+b. Each point represents one colony of A. eximius, the triangles represent colonies of
the subsocial congeneric A. juncundus.

temperature nor humidity changes drastically inside the forest (ESPI report 1976),
the presence or absence of rain and wind, as well as the seasonal distribution of
prey insects (Robinson and Robinson 1970, Wolda 1977, 1978) allow the
assignation of different environmental qualities to each season (Vollrath 1980); in
the lowlands, for instance, March is a rather severe month for spiders (Lubin
1978). The effects of the seasons are less pronounced in the higher elevations.

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMUS 271

especially in the cloud forest, where light afternoon showers fall throughout the
year (unpubl. RENARE data).

Suitable colonies were collected in the field and taken into the laboratory
where they were analyzed: the spiders were extracted, classified, and counted.
Classification was by size, and the different size classes were presumed to
represent instars (see Fig. 1). Suitability was judged by the apparent health of the
colony and by the ease with which the web could be tipped into a large plastic
bag without the loss of animals. Sometimes the area underneath a web was
covered with sheets which were later inspected for escapees, showing that only
strong shaking initiated escape responses in all but the first and second instar
spiderlings. Older instars escaped in similar proportions, which permits the
comparison of all collection data (even in the few cases where spiders might have
escaped during collection), as long as the data are represented as percentages. In
addition to sampling whole colonies (data in Table 1) I censused other colonies
in the field, without disturbance, repeatedly over a one year period, counting
spiders in situ. Most of these data were not used in the analysis (but see Fig.
7), since rarely was the vision good enough to count and classify all spiders.

The calculations of biomass (wet weight) use weight measurements of 5 live
spiders of each size class, which gave the following mean weights: size class 1 =
0.5 mg, 2 = 1 mg, 3 = 2 mg, 4 = 3 mg, 5 = 7 mg, female (6) = 15 mg and male
= 7 mg (Vollrath unpubl.). I give no standard variation since the spiders had been
preselected to be of roughly medium weight for their size class, as judged by the
state of their abdomens.

RESULTS

Effects of seasonality. — The three sites: lowland, intermediate and high
elevation were compared in the effects they exert on the population dynamics of
A. eximius. The colony cluster on the El Llano-Carti Road initially (April 1978)
consisted of over 20 large (about 2000 spiders each) to medium sized colonies
(about 600 spiders each). Later (June) many new colonies were founded in the
vicinity by single females which were often joined by other females. During the
time from May until September gravid females emigrated from most of the old
(mature) colonies (Vollrath 1982), departing from the top of the vertical threads.
It is unusual for Anelosimus to be active during the day in these parts of the
nest, and since I only twice saw a male here, I assume that males do not take
part in the emigration. In September the colonies in this site began to dwindle,
with many sickly or dead spiders hanging in the webbing. It is conceivable that
a virus had spread and contaminated all or most colonies. At the end of
December only two nests were left, much reduced in size. Because of the
combined effects of emigration and possible epidemic, the interpretation of data
from the El Llano site will only be referred to qualitatively.

Seasonal changes in the composition of size classes were more apparent in the
lowland sites of Bohio and Pipeline than in the highland El Valle sites (Fig. 3
and Table 1). The lowland colonies contained very few egg sacs in April, the end
and severest month of the dry season. With two exceptions however, they had
plenty of egg sacs throughout the rainy season and at the beginning of the dry
season. In March/ April I found only a single egg sac in 4 colonies, the mean

272

THE JOURNAL OF ARACHNOLOGY

! /

2 Ml III

Fig. 3. — Size class composition of colonies collected in the lowland (Bohio) and the highland sites
(El Valle). Each circular histogram represents the annual cycle (reading clockwise), the length of each
sector gives the relative proportion of spiders (in percent) of the particular size class (indicated by
the roman numeral) at the sampling month (compare Table 1).

number of egg sacs in 4 colonies was 6 in August/ September (Table la). The El
Valle colonies, in contrast, had egg sacs throughout the year: in 5 high elevation
colonies I found a mean number of 6.4 egg sacs during March/ April, and a mean
of 15.2 during August/ September (Table Ic). Even the El Llano site had a fair
number of egg sacs in April (Table lb).

The proportion of young in each colony also gives some indication about the
seasonality of reproduction. In the lowland sites no first or second instar young
were found in April/ May and very few in September/ October. In the highland
sites young were present throughout the year. In both sites the composition of
other instars varied. This is not astonishing since colonies even at the same site
seemed to differ from one another, in exposure to the elements, to prey, parasites
and predation. However, in the lowland site fourth and fifth instars showed
marked preponderances during certain times of the year, fourth instars at the end
of the rainy season and fifth instars at the beginning of the rainy season. In both
sites adult females were present throughout the year, so were males. The
difference between the sites is most easily expressed by the coefficient of variation
(CV, Simpson et al. 1960) of the instar composition. The CV for the lowland
colonies was CV = 60 (n = 14), for the high elevation colonies it was CV = 36
(n = 14): the colonies in the elevation site showed greater uniformity throughout
the year than those in the lowlands.

The proportional representation of instar numbers masks the true extent of the
difference between the two sites. The seasonal investment in reproduction is better
represented by the proportion of a colony’s biomass tied up in the different age
classes, which depicts the differences better (Fig. 4). The seasonal aspect in the
lowland colonies is apparent in the biomass distribution of the early instars which
is concentrated in February and September. In the lowlands, most of a colony’s
biomass at the beginning of the dry season consists of females, while during the
dry season a large proportion consists of young. At the beginning of the next
rainy season nearly all biomass is tied up again in subadult or adult females,

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMUS 273

Table 1, — Colony composition at the three study areas. The sample month is given first, followed
by the total number of spiders (spids), the number of adults (ads), the number of males ((5(5)^ the
adult sex ratio (ratio = males as proportion of adults), the number of egg sacs (egs), and finally the
ratios of females to egg sacs ($: eggs) and males to egg sacs (^; egs). Lowland colonies were collected
in: Bohio (a), Pipeline Road (b), Anchiote Road (c), Bayano-Mahe (d); two colonies are shown, which
were not collected, but where censused accurately in the field (*).

BIOHIO (a), PIPELINE (b), ACHIOTE (c), BAYANO (d)

Month

spids

ads

(5(5

ratio

egs

$:egs

(5:egs

Jan. (b)

493

74

12

0.17

21

3.0

0.6

Feb. (a)

413

32

4

0.13

0

—

—

March (d)

379

51

10

0.2

0

—

—

April (c)

213

90

6

0.07

0

—

—

(a)

574

72

12

0.13

1

60.0

12.0

(a)

365

86

14

0.17

2

36.0

7.0

May (b)

787

147

30

0.2

8

14.6

3.8

June (a)

425

65

3

0.05

11

5.6

0.3

(d)

459

160

24

0.15

17

8.0

1.4

Aug. (a)

844

95

4

0.04

1

91.0

4.0

(a)

999

120

5

0.04

9

12.9

0.6

Sept (b)

476

8

7

0.5

0

—

—

Nov. (a)

228

46

3

0.07

14

3.1

0.2

Dec. (b)

398

55

13

0.25

5

8.4

2.6

X

503

79

10.5

0.16

6.4

24.3

3.3

± sd

227

42

8.1

0.12

7.1

29.5

3.8

EL VALLE

Month

spids

ads

(5(5

ratio

egs

9:egs

(5:egs

Jan.

423

117

10

0.08

19

5.6

0.5

March

463

147

11

0.07

6

22.7

1.8

598

172

19

0.11

11

13.9

1.7

April

671

146

14

0.1

10

13.2

1.4

268

56

12

0.2

9

4.9

1.3

631

123

24

0.14

13

8.4

1.9

June

365

59

12

0.2

2

13.5

6.0

485

111

11

0.1

3

33.3

3.7

1002

340

32

0.09

76

4.1

0.4

July

738

168

17

0.1

40

3.8

0.4

298

77

12

0.17

3

21.7

4.0

Sept.

145

22

4

0.17

8

2.3

0.5

386

50

12

0.25

21

1.8

0.6

Nov,

619

64

16

0.25

4

12.0

4.0

X

506

118

14.7

0.15

16.1

12.2

2.1

± sd

221

80

6.8

0.06

20.0

9.7

1.7

EL LLANO-CARTI

Month

spids

ads

ratio

egs

$:egs

(5:egs

March

1305

301

42

0.14

0

—

—

April

434

43

15

0.35

14

2.0

1.0

1226

128

10

0.06

7

12.8

0.7

May

190

61

9

0.08

5

6.8

0.6

July

490

90

14

0.18

16

6.4

1.1

475

232

5

0.18

41

46.4

8.2

Aug.

210

26

3

0.12

2

7.7

0.7

Sept.

379

189

21

0.11

14

13.5

1.5

Sept.*

(200

52

0

—

0

—

-)

Nov.*

(270

18

12

0.66

0

—

-)

X

588

134

14.9

0.15

12.4

13.7

2.0

± sd

433

98

12.4

0.09

12.9

15.0

2.8

274 THE JOURNAL OF ARACHNOLOGY

CO

I l-ll lll-IV v-$

Fig. 4. — Representation of the distribution of colony biomass in the different size classes. The
biomass was calculated using the mean wet weight of 5 spiders for each class (see methods). For
explanation of the circular histograms see Fig. 3.

ready to reproduce and rear a second generation. No such variation was found
in the biomass distribution of colonies in the highland sites.

The quantitative observations on 38 colonies are supported by qualitative
observations on about a hundred colonies which were not disturbed (Vollrath
unpubl.). They allow the generalization that in different parts of the country
colonies in similar habitats show similar trends in egg production and instar
representation, with pronounced seasonal influences in the lowlands.

Colonies in the El Llano site were intermediate in their reproductive cycle
(Table lb). The average colony size was similar in all three sites. The average
ratios of adults to juveniles was also quite similar between the two upland
locations: 0.23 in El Llano and 0.23 in El Valle. It was considerably lower in the
lowlands (0.16) where colonies generally had a smaller proportion of adults. The
average number of egg sacs produced in the upland site was intermediate (12)
between the altitude site (16) and the lowlands (6). The ratio of adult females to
egg sacs was also intermediate (14) when compared with the lowlands (24) and
the highlands (12). These differences are not statistically significant because of the
large variation in all sites.

INDIVIDUALS PER CLASS TOTAL

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMUS 275

1000 =,

100-I

A

147

.2 —

54

74

.2

B

6

13 .2 3
~2 -3

27

.04

63

.1

1 18

.07

93 8

— 83 86 —
.04 _ _ /

.04 .04

MONTHS

Fig. 5. — Dynamics of three Anelosimus eximius’ colonies censused repeatedly in the field at the
lowland sites of Bohio and Pipeline Road in 1978-1979. The X-axis denotes the month (roman
numerals). The upper horizontal bars show the total number of spiders in the colony. The number
above each bar give the number of adult spiders, numbers under the bars denote the sex ratio (male
proportion of adults). The dark columns give the number of egg sacs. The actual numbers of spiders
of the different size classes (1 to 5 and (9$) is shown for the sample months by histogram outlines.

In order to study the dynamics of individual nests, many webs were examined
regularly in the field, with minimal disturbances. Only in few cases was the
visibility inside the colony consistently good enough to view 80 to 90% of all
spiders, thus allowing me to use these data. Figure 5 shows the changes in the
population of three colonies in a lowland site during the course of one year.
Colony A contained no egg sacs in September, colonies B and C had few egg
sacs in November and April. The age profiles of the three colonies differed in
several aspects, which may be explained by the differences in size. Colony A,
initially being the largest with 1000 individuals, maintained a stable age
distribution throughout the year. Both colonies B and C were growing in size,
B was the smaller of the two. The instar composition of the two colonies B and
C was similar in most months, often showing a depression in the middle instars.
No similar depression was found in the age structure of the mature colony A.
Other colonies, monitored with less accuracy (Vollrath unpubl.), showed similar
variations in egg production and age structure, again possibly correlated with
colony size.

In the El Valle site Anelosimus juncundus also seems to reproduce throughout
most of the year. There I found many nests, each containing one adult female
with eggs in the months of April, June, July, September and November. In the
other sites A. juncundus may be seasonal, but I found too few individuals to
make a definite statement. In the El Llano site 6 females and egg sacs were found
in July, and in the lowland site of Cerro Galera I found only two females with
eggs, in May. Although A. juncundus was at times found abundantly in the high
elevation and the intermediate sites, it appeared to be rare in the lowlands.

Sex ratios. — In all three habitats the adult sex ratios were highly variable but
(with a few notable exceptions) heavily biased in favor of females (Table 1, 2).
There were no differences in the mean sex ratios between the three habitats (0.16,

276

THE JOURNAL OF ARACHNOLOGY

Table 2. — Composition of new foundations in the El Llano-Carti site. The July column shows the
number of first instar spiderlings. 64 days later, in September, all had progressed to the fourth instar,
when the males show their sex. Only 4 males were then still immatures (included in ^(5), the other
males had just matured. Immature females are denoted by IV (fourth instar).

July

IV

September

$$

(5(5

mortality

sex ratio

32

12

8

4

25%

0.17

21

0

9

2

48%

0.18

78

4

30

2

54%

0.06

39

12

8

4

38%

0.17

40

20

4

2

35%

0.08

31

6

6

1

58%

0.08

41

5

15

2

46%

0.09

X

40

8

12

2

43%

0.12

± sd

17

6

9

1

11%

0.04

0.15, 0.15). On average, colonies in all three sites contained about five females
for every male.

We have seen that pronounced local environmental factors influence the
reproduction and instar composition of colonies. It is conceivable that the large
variance in the sex ratios could be the result of environmental influences,
comparable to the effect of temperature on the sex determination of a turtle egg
(Bull 1981). Such effects might show in a correlation between the colony sex ratio
and the time of year. No significant correlation was found (Fig. 6a), suggesting
that a direct seasonal influence on the sex ratio is not likely. What about indirect
influence?

Seasonality expresses itself in changes of the abiotic as well as the biotic
factors. In the lowland the exposure to desiccation by wind and sun increases
during the dry season when the trees are shedding their leaves. At the same time
the number of prey decreases while predation presumably increases (Robinson
and Robinson 1970). Inside a nest of A. eximius these factors are probably
buffered, the degree of insulation depending on colony size. For example, on a
sunny afternoon in the dry season the relative humidity was 74% inside a (1000
individual) lowland colony, when outside it was 60% (dry/ wet bulb temperature
measurements). Large colonies have large collections of leaves which provide
more shelter and also chambers of relatively constant conditions. Large colonies
may also be affected less by predation (Fig. 5, and Vollrath and Windsor in
press).

A scattergram of colony size plotted against the sex ratio shows no significant
trend if all data are lumped (Fig. 6b). However, a weak correlation (r = 0.36)
was found for the high elevation site, where the sex ratio of larger colonies was
more female biased. This correlation improved and became significant (r = 0.663,
0.05<P<0.01) when the number of mature spiders only (instead of all
individuals of a colony) was plotted against the sex ratio (Fig. 7a). If analyzed
this way, colonies in the lowlands which were censused repeatedly also showed
some correlation (r = 0.479, 0.1<P<0.05, Fig. 7c), whereas still no correlation
was found for the collected lowland colonies (Fig. 7b).

Colony size can be taken, to a very limited degree, as an indicator for the age
of a colony. Isolated small nests are generally new foundations, isolated large
nests are several generations older (Vollrath 1982). Since most of the colonies

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMVS 211

sex ratio
.5

• ■

■ •

month

■ — I —

JAN APR JUL OCT JAN

colony size
* of spiders

1000

500

100

B

.1

• BOHIO
■ EL VALLE

lex ratio

.3

Fig. 6. — Scattergrams of: (A) colony sex ratio and the time of year, and (B) colony sex ratio and
colony size. Sex ratios were calculated as the male proportion of all adults.

sampled had been taken from colony clusters, they may have been recent
offshoots of much older foundations. None of the new foundations that I either
observed to occur naturally or which I experimentally started, survived for longer
than 6 months (Vollrath 1982). However, I do have, from some of these natural
foundations, some information on the sex ratios of new colonies (Table 2). The
mean ratio (+ s.d.) in these colonies was 0.12 (+ 0.04), this is lower than the
primary sex ratio of 0.08 (± 0.01, Vollrath 1986) but higher than the average
ratio of mature colonies (0.15 ± 0.09).

Reproduction is measured by the number of offspring produced, and we saw
that reproduction was clearly seasonal in the lowland sites and may have had a
seasonal component in the highland site. Reproduction is presumably also a joint
effort of the male as well as female since no spider is known to be
parthenogenetic. Although the sex ratios showed seasonal trends, the possibility
remains that a correlation may be found between reproductive effort and the
relative abundance of either sex. The ratios of males or females to the number
of egg sacs present in a colony are given in Table 1. There were no differences
in the sex ratios between the two sites. There were, however, pronounced
differences in the average number of egg sacs produced by colonies in the two
sites, production was more than double in the altitude sites when compared with
the lowland sites (16 versus 6). Accordingly the ratio of females to egg sacs was
also lower: in the altitude site one egg sac was produced for every 12 females,
in the lowlands it was one egg sac for every 24 females. A similar trend was
found in the ratio of males to egg sacs. This analysis of colony composition and
presence of egg sacs shows that the ratio of females to egg sacs was generally
more than 10 to one, whereas the ratio of males to egg sacs was roughly two
to one (Vollrath 1986).

278

THE JOURNAL OF ARACHNOLOGY

EL VALLE

300 •

• BOHIO
■ EL LLANO
A PIPELINE

O others

A

E

3

C

200 •

’00 •

A

.1

200 -

COLONY REPREATS

■ B
• C

— r-

.3

.4

.5 sex ratio

Fig. 7. — Scattergram of adult colony size and sex ratio for (A) highland colonies, (B) lowland and
altitude colonies, and (C) colonies in the lowlands which were censused repeatedly. A and B represent
data from Table 1, C represents the colonies from Fig. 5.

DISCUSSION

Population dynamics and seasonality. — It emerges from the data on instar
composition of colonies, and even more from the calculations of colony biomass,
that the populations of A. eximius had different dynamics in the two main sites,
highlands and lowlands. It is most likely that these differences were due to the
differences in environmental parameters, which had a clear seasonal component.
It cannot be inferred from the data that the differences in colony composition
were due entirely to abiotic factors. Heavy rainfall, for example, destroys the
capture threads and may force “the colony” to spend much energy in frequent
rebuilding during the rainy season. On the other hand, active spiders dessicate
quickly in very dry conditions (Davies and Edney 1952). In order to conserve
water in the dry season colony activities are possibly reduced to the extent of
neglecting the web, thus foregoing prey capture and nutrient uptake. However,
it is more likely that a combination of biotic and abiotic factors was responsible
for the observed differences in the two sites. Most prey insects, for example, have
activity cycles clearly delineated by the seasons (Wolda 1977, 1978, ESPI report
1976). Predators, especially predatory arachnids and insects, will presumably also
show seasonality to match the temporal distribution of their spider prey. Lubin
(1978) has shown that most solitary spiders in the lowland forest of Panama have
distinct seasonal distribution and reproduction. My study suggests that the social

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMUS 279

spider A. eximius is less seasonal than its solitary sympatrics, whieh practically
disappear from the habitat in the late dry season.

The seasonality of A. eximius expresses itself most strongly in the presence of
egg sacs, it appears that in the lowlands reproduction stops during the height of
the dry season, in March and April. The highland colonies, on the other hand,
had egg sacs all year round. Although egg production was markedly different
between the two sites, the sex ratios showed no comparable trends.

Sex ratios and environment. — The sex ratios in colonies of A. eximius were
heavily biased in favor of females (Overal and Ferreira 1982, Christenson 1984,
Aviles in press, Vollrath 1982, Vollrath 1986). In some reptiles the sex ratio in
a population can be affected by the environment during the time of egg
development (Bull 1981), because ambient temperature has a modifying effect on
the hormonal control of sex determination. In spiders the sex is not determined
hormonally but by the chromosomal set of each cell individually [(viz the
presence of gynandromorphs, comparable to insects (White 1973)]. Still, abiotic
factors in the environment could conceivably influence the sex determination of
a developing egg. This possibility seems unlikely since the observed variation in
sex ratios, although quite large, showed neither seasonal nor locality differences
(Table 1). Moreover, colonies collected at the same time in the same location
showed as large a variation as colonies collected at different times in different
locations.

If abiotic factors are ruled out, could possibly biotic factors be responsible for
the sex ratio bias? Three main agents of biotic influence can be distinguished:
parasitism, predation and prey (nutrition). Of these nutrition is the most difficult
to separate from abiotic effects, since the activity of prey insects is in itself often
triggered by abiotic factors. Direct, long-term observations of prey capture for the
discussed colonies is lacking (but compare Nentwig 1985), therefore prey intake
was deduced from colony size, for this spider biomass may be a better measure
than number of individuals. Since environmental conditions might have been
unique at the time when the measured sex ratio of a colony was fixed (i.e. when
the adult spiders developed), adult numbers or biomass would have to be
compared with the sex ratio of each colony. However, different rates of the
mortality of males and females could easily mask any effect nutrition might have
had.

The highland sites showed a clear and significant negative correlation between
the number of adults and the sex ratio. A similar effect was apparent in the three
colonies censused repeatedly in the lowland site, but it was not visible in the other
lowland colonies. How can we explain the observation that colonies with fewer
adults contained relatively more males? If we assume the same mortality for both
sexes, this observation would lead us to deduce that fewer male eggs were laid
at times when prey was abundant, and the colony was growing. Such an effect
would be independent of absolute colony size, as a comparison of the altitude
and the censused eolonies shows. However, could not all the data, including the
observations on all lowland colonies, be better interpreted by the alternative
assumption that males and females experience different mortality?

In A. eximius the average primary sex ratio was much lower than the average
adult sex ratio observed in the field (Vollrath 1986), this suggests that many
females have either died or left the colony between hatching, maturation, and
older age. A. eximius' rarely venture far from the retreats, and they do not

280

THE JOURNAL OF ARACHNOLOGY

assist in web construction nor in prey capture. The females, on the other hand,
expose themselves to predators during the daily routine of capture-thread
construction, and to injury during the attack and subduction of potentially
dangerous prey insects (Vollrath and Rohde- Arndt 1983). They may even defend
the colony with considerable risks to their own lives (Vollrath and Windsor in
press). It would appear that female mortality was on average twice male mortality
(difference between primary and actual sex ratios: 0.08 vs 0.15), if the hypothesis
of higher female mortality is correct. Such an interpretation of the sex ratio data
would be supported by the observation that large colonies survive better than
small colonies (Vollrath 1982), an indication that in nests with many “workers”
more of them survive. This is supported by the observation that small

foundations had on the whole a lower mean sex ratio than the larger mature

colonies (0.12 vs 0.15). The data on new foundations comes from nests in which
the sexes are about to mature, or have Just done so (Table 2). It appears that

at this moment already more female than male spiderlings have died. After

maturation this ratio shifts even more towards the males (to 0.15), since male and
female adults are presumably more dissimilar than the juveniles.

Instead of mortality, female emigration might account for the rate of female
“disappearance”, which was 50% higher than that of the males. In A. eximius
emigration seems to be a well defined phenomenon, associated with a particular
behavior pattern (Vollrath 1982). Females move into the upper strands of the
capture area, where they depart, using their floating dragline as a bridge. This
behavior, combined with the observation that it occurs during the morning, a
time when the spiders normally are at rest in the retreats, makes emigration quite
obvious. The many hundreds of females I observed moving in the capture threads
prior to departure were only twice joined by a male, which both times retreated
back into the colony. Only one of the 380 single or multiple female foundations
examined contained a male which presumably had followed the females to their
new nest a few meters away from the likely parental colony (Vollrath unpubl.).
Emigration seems to occur at specific times of the year and the day, and it is
unlikely that females leave in any significant numbers at other times (Vollrath
1982). Males presumably leave their parental colony only rarely, and possibly
accidentally. It is concluded that emigration is not a likely explanation for the
observed bias in the sex ratios.

Summarizing the discussion on the sex ratio bias I want to conclude that the
sex ratio in Anelosimus eximius is probably not influenced by abiotic
environmental factors. The strong bias in favor of females is due to the fact that
many more female eggs are laid. This effect of a biased primary sex ratio is
slightly softened by higher female mortality due to predation, which is a biotic
vector of the environment with considerable local and temporal variation.

ACKNOWLEDGMENTS

I thank M. and B. Robinson, D. Windsor and Y. Lubin for many critical
discussions during the course of th study, and I thank M. Ridley, D. Smith and
especially Y. Lubin for helpful critique of the manuscript. The study was financed
by a Smithsonian Postdoctorate Fellowship and I thank the director and staff of
STRI for their continuous help. The American Arachnological Society generously
supported a trip to present this work at the meetings in New Orleans.

VOLLRATH— ENVIRONMENT, REPRODUCTION AND SEX RATIO OF ANELOSIMUS 281

LITERATURE CITED

Aviles, L. in press. Sex ratio bias in the social spider Anelosimus eximius, with comments on the
possibility of group selection. Amer. Nat.

Brach, V. 1975. The biology of the social spider Anelosimus eximius. Bull. S. California Acad. Sci.,
74:37-41.

Bull, J. J. 1981. Sex determination in reptiles. Quart. Rev. Biol., 55:3-16.

Buskirk, R. 1981. Sociality in Arachnida. Pp. 281-387, In Social Insects (H. R. Hermann, ed.).
Academic Press, New York.

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

Davies, M. E. and E. B. Edney. 1952. The evaporation of water from spiders. J. Exp. Biol., 29:571-

582.

ESPI Report. 1976. Environmental Science Program. Vol. IV, D. Windsor, ed. Smithsonian Tropical
Research Institute, Panama.

Kullmann, E. and W. Kloft. 1968. Traceruntersuchungen zur Regurgitationsfuetterung bei Spinnen.
Zool. Anz., 32:487-497.

Levi, H. W. 1956. The spider genera Neottiura and Anelosimus. Trans. Amer. Microsc. Soc., 75:407-
421.

Levi, H. W. 1963. The American spiders of the genus Anelosimus (Araneae, Theridiidae). Trans.
Amer. Microscop. Soc., 82:30-48.

Levi, H. W. 1972. Taxonomic-nomenclatural notes on misplaced Theridiid spiders, with observations
on Anelosimus. Trans. Amer. Misroscop. Soc., 91:533-538.

Lubin, Y. D. 1978. Seasonal Abundance and diversity of web-building spiders in relation to habitat
structure on Barro Colorado Island, Panama. J. Arachnol., 6:31-51.

Lubin, Y. D. 1980. Population studies of two colonial orb-weaving spiders. Zool. J. Linn. Soc.,
70:265-287.

Lubin, Y. D. 1982. Does the social spider Achaearanea wau (Theridiidae), feed its young? Z.
Tierpsychol., 60:127-134.

Nentwig, W. 1985. Social spiders catch larger prey: a study of Anelosimus eximius (Araneae,
Theridiidae). Behav. Ecol. Sociobiol., 17:79-85.

Overal, W. L. and Pa. Ferreira da Silva. 1982. Population dynamics of the quasisocial spider
Anelosimus eximius (Theridiidae). In The Biology of Social Insects (M. D. Breed, C. D. Michener
and H. E. Evans, eds.). Proceedings of the IXth Congress International Union for the Study of
Social Insects, Boulder, Colorado.

Robinson, M. H. and B. Robinson. 1970. Prey caught by a sample population of the spider Argiope
argentata in Panama. Zoo. J. Linn. Soc., 49:345-357.

Simon, E. 1891. Observations biologiques sur les Arachnides. Ann. Soc. Ent. France, 60:5-14.

Simpson, G. G., A. Roe and R. C. Lewontin. 1960. Quantitative Zoology. Harcourt, Brace and
World, Inc., New York.

Tapia, Y. and T. DeVries. 1980. Tolerancia y cooperation en la Arana social Anelosimus juncundus
del bosque tropical Rio Palenque, Ecuador. Rev. Univ. Cat. Ecuador, 8:51-74.

Vollrath, F. 1980. Male body size and fitness in the web-building spider Nephila clavipes. Z.
Tierpsych., 53:61-78.

Vollrath, F. 1982. Colony foundation in a social spider. Z. Tierpsych., 60:313-324.

Vollrath, F 1986. Eusociality and extraordinary sex ratios in spider. Behav. Ecol. Sociobiol., 18:283-
287.

Vollrath, F. and D. Rohde-Arndt. 1983. Prey capture and feeding in the social spider Anelosimus
eximius. Z. Tierpsych., 61:334-340.

Vollrath, F. and D. Windsor in press. Subsocial and social Anelosimus: a comparison especially of
nest defense. In Proceedings IXth Internatl. Arachnol. Congress (M. H. Robinson, ed.).

White, M. J. D. 1973. Animal cytology and evolution. 3rd ed., Cambridge Univ. Press, Cambridge,
UK.

Wolda, H. 1977. Fluctuation in abundance of some Homoptera in a neotropical forest. Geo-Eco-Trop,
31:229-257.

Wolda, H. 1978. Seasonal fluctuations in rainfall, food and abundance of tropical insects. J. Anim.
Ecol., 47:369-381.

Manuscript received, revised January 1986.

’4»/t|S»Vt 'swv.r4f»H

.Ai . .,» <- _ .i*^} X

‘‘ ^ ■'. V*'--

^ <.-■ ..,i

' ■.. ^■m.li ■, .-.‘I ,-■ . ■f^' j.Ti. ‘V'/.I rtKViltfcJlr

av; - ' 9 .. . .-• -: JjW ■ ' < •■

■■:\^.\ V ■ ‘ '' .’ ,f‘«* "**■-*?!

\. ’....»,' ► »i];--_X " .,|».,i'^<A» • r-"-' v|ijyS3l'' •■»-* 4 ,»• » '

'iMi'-

;'■,» ' -“/iJi' ■' ■'• kU I 4 ..ii'V ■ A ’ '* '■ '

'•’ ^ . .

■' ■ -u* 1',;.', I

ir-

»•,' i

■J. . *>.. vf 'i ■•

' « .^■^. -

v/ \,>‘. ‘j , fV*;-** ;'^ »'>; •■-.f-.- '■.•••' - '7 ■

•Ui'- • ■■ • .»-J ijiM>t!i3W^-.pWf. ;<''l .f.MO atwP'

•( ■pi. '0Sg;' ' .i'^ .■'k.<:>r-vtt#r,%'V

i ..■A/k.v^ 'a s«l:>t^.:.^

■■:■■■"■'■••’>

■,4i! A' * A.. ^-a

THE AMERICAN ARACHNOLOGICAL SOCIETY

President:

Jerome S. Rovner (1985-1987)

Department of Zoology
Ohio University
Athens, Ohio 45701

Membership Secretary:

Norman 1. Platnick (appointed)

American Museum of Natural History
Central Park West at 79th Street
New York, New York 10024

Secretary:

Brent D. Opell

Department of Biology
V.P.I.S.U.

Blacksburg, Virginia 24061
Directors:

Frederick A. Coyle (1985-1987), W. G. Eberhard (1986-1988), Susan E. Riechert
(1985-1987).

Honorary Members:

P. Bonnett, W. J. Gertsch, H. Homann, fB. J. Kaston, R. F. Lawrence, H. W.
Levi, G. H. Locket, A. F. Millidge, M. Vachon, T. Yaginuma.

The American Arachnological Society was founded in August, 1972, to
promote the study of Arachnida, to achieve closer cooperation between amateur
and professional arachnologists, and to publish The Journal of Arachnology.

Membership in the Society is open to all persons interested in the Arachnida.
Annual dues are $25.00 for regular members, $15.00 for student members.
Correspondence concerning membership in the Society must be addressed to the
Membership Secretary. Members of the Society receive a subscription to The
Journal of Arachnology. In addition, members receive the bi-annual newsletter
of the Society, American Arachnology.

American Arachnology, edited by the Secretary, contains arachnological news
and comments, requests for specimens and hard-to-find literature, information
about arachnology courses and professional meetings, abstracts of papers
presented at the Society’s meetings, address changes and new listings of
subscribers, and many other items intended to keep arachnologists informed
about recent events and developments in arachnology. Contributions for
American Arachnology must be sent directly to the Secretary of the Society.

President-Elect:

William A. Shear (1985-1987)
Department of Biology
Hampden-Sydney College
Hampden-Sydney, Virginia 23943

Treasurer:

Norman V. Horner (1985-1987)
Department of Biology
Midwestern State University
Wichita Falls, Texas 76308

Archivist:

Vincent D. Roth

Box 136

Portal, Arizona 85632

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 14

NUMBER 2

Symposium: Social Behavior in Spiders
George W. Uetz, University of Cincinnati, organizer and editor of symposium

Effects of colony size on web structure and behavior of the social
spider Mallos gregalis (Araneae, Dictynidae), William James

Tietjen 145

Genetic differences in social behavior and spacing in populations of
Metepeira spinipes, a communal-territorial orb weaver (Araneae,

Araneidae), George W. Uetz and Karen R. Cangialosi 159

The ecology of the cooperative spider Agelena consociata in Equatorial
Africa (Araneae, Agelenidae), Susan E. Riechert, Rosemarie Roeloffs,

and Arthur C. Echternacht 175

High prey abundance and a reduction in cannibalism: The first step to

sociality in spiders (Arachnida), Ann Lundie Rypstra 193

Population genetics of Anelosimus eximius (Araneae, Theridiidae), Deborah

R. R. Smith 201

Influence of food supply on the duration of the gregarious phase of a
maternal-social spider, Coelotes terrestris (Araneae, Agelenidae),

Bertrand Krafft, Andre Horel, and Jean- Michel Julita 219

Societies of spiders compared to the societies of insects, Roger Darchen

and Bernadette Delage- Darchen 227

Courtship and alternative mating tactics in a social spider, Yael D.

Lubin 239

Orb-weaving spiders in aggegations modify individual web structure,

Ruth E. Buskirk 259

Environment, reproduction and the sex ratio of the social spider

Anelosimus eximius (Araneae, Theridiidae), Fritz Vollrath 267

Cover photograph, Mallos gregalis Simon, by G. W. Uetz
Printed by the Texas Tech Press, Lubbock, Texas, U.S.A,
Posted at Lubbock, Texas, in August 1986

? The Journal of

ARACHNOLOGY

OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY

VOLUME 14

FALL 1986

NUMBER 3

THE JOURNAL OF ARACHNOLOGY

EDITOR: James E. Carico, Lynchburg College

ASSOCIATE EDITOR: William B. Peck, Central Missouri State University (Until
30 April 1987); Jerome S. Rovner, Ohio University (After 1 May 1987)
ASSISTANT EDITOR: James C. Cokendolpher, Texas Tech University
EDITORIAL BOARD: C. D. Dondale, Agriculture Canada; W. G. Eberhard,
Universidad de Costa Rica; M. E. Galiano, Museo Argentino de Ciencias
Naturales; W. J. Gertsch, American Museum of Natural History; N. F.

Hadley, Arizona State University; V. F. Lee, California Academy of Sciences;
H. W. Levi, Harvard University; E. A. Maury, Museo Argentino de Ciencias
Naturales; W. B. Muchmore, University of Rochester; M. H. Muma, Western
New Mexico University; N. 1. Platnick, American Museum of Natural
History; G. A. Polis, Vanderbilt University; S. E. Riechert, University of
Tennessee; M. E. Robinson, U.S. National Zoological Park; V. D. Rotb,
Southwestern Research Station; J. S. Rovner, Ohio University; W. A. Shear,
Hampden-Sydney College; W. J. Tietjen, Lindenwood College; G. W. Uetz,
University of Cincinnati; C. E. Valerio, Universidad de Costa Rica; S. C.
Williams, San Francisco State University.

THE JOURNAL OF ARACHNOLOGY (ISSN 0161-8202) is published in
Spring, Summer, and Fall by The American Arachnological Society at Texas
Tech Press.

Individual subscriptions, which include membership in the Society, are $25.00
for regular members, $15.00 for student members. Institutional subscriptions to
The Journal are $40.00. Correspondence concerning subscriptions and member-
ships should be addressed to the Membership Secretary (see back inside cover).
Back issues of The Journal are available from Dr. Susan E. Riechert, Department
of Zoology, University of Tennessee, Knoxville, TN 37916 U.S. A., at $10.00 for
each number; the Index to Volumes 1-10 is available for $10.00. Remittances
should be made payable to The American Arachnological Society. Correspon-
dence concerning undelivered issues should be addressed to the Texas Tech Press,
Texas Tech University, Lubbock, Texas 79409, U.S. A.

Change of address notices must be sent to the Membership Secretary.

Detailed instructions for the preparation of manuscripts appear in the Fall
issue of each year, and can also be obtained from the Editor and the Associate
Editor. Manuscripts that do not follow those instructions will be returned to the
author(s) without benefit of review. Manuscripts and all related correspondence
must be sent to Dr. Jerome S. Rovner, Associate Editor, Department of
Zoological Sciences, Ohio University, Irvine Hall, Athens, Ohio 45701, U.S. A.

PROOFS, REPRINTS, and CHARGES: Authors will receive a reprint order
form along with their proofs. Reprints are billed at the printer’s current schedule
of costs. All authors will be billed by the printer for page charges. The charge
per journal page will be assessed as follows: $20. 00-nonmembers; $ 15.00-members
acknowledging grant, institutional, or other support; $ 10.00-full members who do
not have (and do not acknowledge) support; $5. 00-student members who do not
coauthor with full members and who do not acknowledge support. Members may
have page charges waived or reduced for a limited number of pages; a statement
of justification for waiver should accompany the manuscript when it is first sent
to the Associate Editor for review.

Vincent, L. S. 1987. B. J. Kaston, American Araneologist 1906-1985: a biography and bibliography. J.
Arachnol., 14:283-291.

BIOGRAPHY

B. J. KASTON, AMERICAN ARANEOLOGIST 1906-1985:
A BIOGRAPHY AND BIBLIOGRAPHY

Leonard S. Vincent

Division of Biological Sciences
Fullerton College
Fullerton, California 92634

When B. J. Kaston did not attend the June, 1985, American Arachnological
Society meeting in Los Angeles, many of us who knew him well suspected his
health had worsened. A small delegation was sent to his home in nearby Santa
Ana to present him with an award for his special contribution to arachnology
and a huge home-made get-well card signed by the meeting’s attendants. B. J.
died two months later on August 24, 1985, in, as he would put it, his 80th year
of life.

In preparing this biography, I expected to find evidence of B. J.’s dedication
to araneology. To be sure, testimonials to his enormous contribution were
abundant. What was surprising was to find that his contributions were done in
his spare time — nights and weekends. B. J.’s academic appointments were
teaching positions, a situation which might have caused a less disciplined and
determined person to become a frustrated researcher and a mediocre teacher. B.
J. Kaston, however, turned a series of difficult situations into positive ones,
excelling as both teacher and researcher.

B. J. (Benjamin Julian) was born July 2, 1906, in New York City. His early
home life was less than ideal. His mother died when he was six and he disliked
his father, a real estate salesman. In later years, B. J. seldom wrote or spoke of
his parents or his two sisters and brother.

Unlike some other luminaries in arachnology, B. J. was not given to the study
of spiders early in his life. However, his high regard for intellectual achievement
was established early in his academic career. B. J. graduated June 24, 1926, from
the esteemed De Witt Clinton High School in New York City, scoring in the 99th
percentile in all of the comprehensive Regent’s Exams. On the other hand, he
kept a low social profile in high school; he did not join science clubs or
participate in other high school activities of record.

B. J. entered Ohio State University as a zoology major in 1926 and remained
one year before transferring to North Carolina State University at Raleigh. At
NCSU, B. J., also known at the time as “Doc”, earned a B. S. in zoology by
taking an extensive range of zoology courses with a minor in botany on top of
23 hours of chemistry. In spite of rigorous course loads, he managed to earn
membership in the Pine Burr Society (a local scholarship group). Phi Kappa Phi

284

THE JOURNAL OF ARACHNOLOGY

Fig. 1. — B. J. Kaston 1941. Fig. 2. — B. J. Kaston 1978 (courtesy H. Levi).

(a national honor fraternity), and the Leazar Literary Society (a debate and
oratory organization). He was also a member of the wrestling squad and
chemistry club. His interest in photography, recognized even then, was noted in
the 1930 NCSU year book Agromeck: “He was a familiar figure around the
campus chasing bugs and snakes. Of course, wherever he went he took his trusty
photographic outfit ...” This would aptly describe B. J. in his later years as well.
To support his studies at NCSU during the depression years, he worked as a
lifeguard and as a teaching assistant in zoology and botany.

At NCSU, B. J. met Elizabeth Haban who later became his wife, collecting
partner, rearer of spiders collected, and illustrator for his Ph.D. dissertation, for
How To Know The Spiders, and Spiders of Connecticut. Elizabeth persuaded
him, many years later, to accept the invitation of H. E. Jaques, the individual
in charge of the Pictured-Key Nature Series, to write How to Know the Spiders.
B. J. graduated from NCSU in June of 1930 and entered the Ph.D. program in
zoology at Yale in September of that year. There he met Alexander
Petrunkevitch, who suggested that B. J. study spiders. Previous to this, B. J. had
not expressed more than a casual interest in spiders. Four years later B. J.
finished his dissertation: “A Study of The Senses and Sense Organs Involved in
the Courtship of Some Vagabond Spiders.”

Much to his disappointment, B. J.’s first full-time job after graduating from
Yale in June of 1934 did not involve teaching. He obtained a position at the
Connecticut Experiment Station where he worked on elm beetles and Dutch Elm
Disease for the next four years. His tendency to work nights and weekends on
spiders became evident and continued throughout his long professional career.

VINCENT— B. J. KASTON BIOGRAPHY

285

While at the Experiment Station, he published 15 papers on spiders. He also
collected many of the spiders that were used in his major monograph, Spiders
of Connecticut, a project he conceived of doing as early as 1935. These spiders,
as well as the bulk of the rest of his collection, he donated to the Smithsonian
in 1984.

In 1938, he obtained his first teaching appointment in Gainesville, Georgia, at
Brenau College, a small, four-year liberal arts college for women, where he had
a heavy teaching load. His spare time included many trips to the American
Museum of Natural History to study their spider collection (a practice he retained
for years) and to visit his good friend Willis Gertsch. He often embarked on short
combined collecting and sightseeing tours which always seemed to include a visit
with fellow araneologist H. K. Wallace in Florida. He also continued writing
long, detailed letters to colleagues requesting or giving advice and information on
spiders or the latest news on other “spiderologists”, a habit maintained until his
death. Much of his non-teaching time in his early years at Brenau must have been
spent working on Spiders of Connecticut, which he finished in 1941. The seven
year delay (to 1948) in publication was a constant source of frustration for him.

B. J.’s tenure at Brenau was interrupted by a one-year sojourn (1943-1944) in
the army, identifying mosquitoes that were sent in from various army bases in
the southern U.S. as part of a malaria control project. Partial deafness in one
ear precluded him from other responsibilites. The president of Brenau managed
to communicate to the authorities that B. J.’s capabilities were more valuable in
the classroom, and B. J. was released from further service.

Judging from the comments of several of his former Brenau students, B. J. was
considered a great teacher. One wrote that he had “tremendous energies — and
managed to impart an excitement in all the courses he taught” and that “he
stimulated questions and thought and encouraged us whenever he saw a spark
of interest.” Others commented that he “made learning a joy,” and was a “teacher
one would never forget.”

After a summer research fellowship at Harvard in 1945, B. J. joined the faculty
of the Zoology Department of Syracuse University in New York. While there he
set to the task of updating the Spiders of Connecticut whose publication was still
delayed due to the war. B. J. remained at Syracuse University only one year
before accepting an offer from the Teachers College of Connecticut (now the
Central Connecticut State University). This offered close proximity to the Yale
libraries and the Connecticut Agricultural Experiment Station. Unfortunately,
there was no graduate program in biology; consequently, he recommended to
those students who expressed an interest in doing graduate studies on spider
biology that they work with H. K. Wallace at the University of Florida. John
F. Anderson was the only undergraduate student of B. J.’s to eventually complete
a dissertation on spiders and to continue working on them professionally. Many
others gained an appreciation for spiders and continued studying them as an
avocation. Some students benefited from funds which B. J. donated anonymously
for academic scholarships.

His major spare time project during his beginning years at Teachers College
was the preparation of what he referred to as his “little book”. How to Know
the Spiders. Since the book was limited in size by the publisher, B. J. felt that
it would not be able to contain much for the specialist but would better serve
the beginner in the field.

286

THE JOURNAL OF ARACHNOLOGY

At this time, B. J. was also in demand for his popular talks on nature
photography. His skills as a photographer had earned him a Five Star rating in
nature slides from the Photographic Society of America and a medal at the
Rochester International Exhibit. B. J. eventually donated most (several thousand)
of his slides to the American Museum of Natural History.

B. J. remained at Teachers College until his retirement in July of 1963. His
breadth of teaching experience was unusual: he taught every one of the 14 courses
that Teachers College and Brenau College offered in biology! In spite of his
teaching schedule, he still found time to advise and encourage students on matters
relating to graduate studies. A laboratory at Central Connecticut State University
will be given B. J.’s name in remembrance of his contributions.

In 1964, B. J. accepted a lectureship at San Diego State College and also
remarried. His new wife, Barbara (Szymanski), a former student of B. J.’s at
Teachers College and now a practicing psychiatrist in California, did graduate
work in cytogenetics at SDSC. Barbara worked with black widow chromosomes.

In 1967, B. J. received a three-year NIH grant to study black widows. In 1969,
he began the preparation of his revised edition of How to Know the Spiders,
which became available in 1972. He was president of the Arachnologists of the
Southwest for 1970 and 1971.

Thus, B. J. remained in high gear at San Diego State College for nine years
until his second retirement in 1973. In that year, he wrote Wallace of his planned
retirement, noting that he was “the last of our crowd to do so,” the crowd being
Gertsch, Lowrie, Muma, and Wallace. In the same letter he mentioned that he
would be taking on “some” of the editorial work of the newly formed Journal
of Arachnology. B. J. served on the editorial board for the next 12 years. He
was appointed (1980) to Associate Editor, a position he held until his death. Also,
in 1973 he attended the seminal meeting of the American Arachnological Society
in Silver City, New Mexico. B. J. participated in the founding of the A.A.S. and
later, 1978, was elected for a three-year term to the three-member Board of
Directors.

To gain a fuller appreciation of B. J. and the magnitude of his contribution
to araneology, one needs to go beyond a mere chronology of his life. An analysis
of his publications and correspondence with colleagues reveals his independent
nature, his perfectionism, his descriptive rather than experimental approach to
research, his dedication and determination, his forthrightness and liberal use of
a barbed pen, and his breadth of knowledge and interests regarding araneology
and araneologists.

B. J. published 86 papers; in all but six he was the sole author. Of the six,
he was the senior author of five. His collaborative papers dealt mostly with his
elm beetle work. Surely part of his lone araneology was due to his being isolated
in Georgia for many years and to the circumstance of most of his employment:
100 % teaching positions in colleges without graduate programs.

B. J.’s publications were legend for their accuracy and detail, both in style and
content. Herb Levi (pers. comm.) wrote, “I still cannot see how any man can
produce a work {Spiders of Connecticut, 1948) with so few misprints and so few
mistakes.” Martin Muma, as well as B. J. himself, referred to Kaston as an “old
maid” on terminology and detail.

The publications on spiders of B. J. Kaston began in 1935 and continued for
fifty years. His early major papers were on pheromones and courtship behavior.

VINCENT— B. J. KASTON BIOGRAPHY

287

The many shorter works were on parasites of spiders, black widow distributions,
and nomenclatural notes. Taxonomic papers first appeared in 1938. Shortly after
his monumental Spiders of Connecticut (1948), the first of his many reviews
appeared. Post- 1960 papers covered various morphological topics such as
intersexes, ocular anomalies, a review of little known aspects of spider behavior,
an interpretation of the evolution of webs, and a descriptive monograph on the
comparative biology of black widows. There were more papers on nomenclature,
many more reviews, and, of course, two more editions of How to Know the
Spiders, a large supplement to Spiders of Connecticut, and a revised Spiders of
Connecticut. The black widow on the cover of this edition of Spiders of
Connecticut reflected B. J.’s interest in the Latrodectus group.

B. J. continued writing until shortly before his death, since he could do this
in relative comfort. Other activities associated with his study of spiders were
gradually eliminated as various ailments worsened in his seventies. His poor
eyesight, angina, and arthritic back restricted him to sedentary activities such as
microscope work and editorial tasks. Eventually, peering into his scope became
too painful. He still was able to critique the spider works of others and he
determinedly continued to do so.

As a reviewer, B. J. was no less a perfectionist. Levi wrote, “B. J. was always
the first one to point out a missing comma or some other minor mistake
overlooked by the editor. I have always felt that I became more watchful lest B.
J. discover any error or misprint or missing comma in one of my papers.”

In manuscript reviews B. J. wanted to maintain high standards of writing,
precision of illustration, and thorough reporting of literature. It was the non-
araneologists and publishers who often received his sharpest, and often vehement,
attacks. For example, in closing his 1950 book review of Duncan’s Webs In The
Wind, B. J. wrote, “Why the editor and the publisher saw fit to inflict such a
volume on an unsuspecting public is incomprehensible. They would aid the
humanizing of science most by suppressing the sale of such a mass of
misinformation . . . .”

Clearly, B. J. was forthright in his reviews, no matter who the writer. In his
1950 review of his friend Gertsch’s American Spiders, B. J. used many
superlatives and closed with “ . . . the outstanding volume on these much
maligned but extremely interesting animals.” However, B. J. pointed out, “It is
regrettable that the morphology of the reproductive organs has not been
described sufficiently. ...”

B. J.’s own publications are but a part of his overall contributions to
araneology — a part, however, that can be quantified. The first edition of his “little
book,” How to Know the Spiders, sold over 16,000 copies. The second edition,
20 years later, sold close to 7,000 copies, and the last edition, 1978, has so far
sold close to 7,000 copies. The 1948 edition of Spiders of Connecticut is the most
widely used bulletin of the series (there are over 110), especially out of state.
Indeed, it was listed 66 times in Science Citation Index from 1965 to 1984.

B. J.’s crusade for perfection and thoroughness was just as evident throughout
his teaching career. His teaching assistants and students from the late thirties and
early forties found him to be a “born teacher.” Students from the 1950’s and
1960’s at Teachers College, who went on to get higher degrees in a biological
science, wrote the following: B. J. “was the epitome of the professor’s professor.
Always known for his candor . . . Respected by those who knew and sensed his

288

THE JOURNAL OF ARACHNOLOGY

dedication to the rigors of study and feared by those who were inept . . . [he]
instilled a deep and abiding love for the scientific method” and “in class, he was
a relentless pursuer ... if you didn’t prepare or think, God help you; he could
and would be merciless.” Lastly and perhaps most insightful: “He was similar to
other men given to strong beliefs, he was sometimes wrong, but never in doubt.”

It is within B. J.’s prolific correspondence with colleagues that one can best
sense his utter dedication to araneology. Reading his long, detailed letters to
Exline-Frizzell, Gertsch, Muma, and Wallace, letters that covered close to half a
century, one can observe certain aspects of the history of our discipline unfold,
and its contributors come and go or come and stay. B. J.’s letters — he would
sometimes write two or three a month to each of the above — usually consisted
of a brief greeting and inquiry into family matters before they would dive into
some aspect of araneology. Typically, he was either requesting information or,
more often, providing information.

Simple questions put to B. J. often received detailed answers. In a 1951 letter,
Harriet Exline-Frizzell referring to a statement in Spiders of Connecticut where
B. J. indicated his preference for the use of Agelenidae over Agalenidae, asked
B. J. if he was still of the same opinion considering that “the new rules call for
correction of caligraphy or orthography in deriving Latin terms.” This triggered
over a 500 word reply by B. J., part of which read: “ My opinion . . . still stands
as on page 278. I try to follow the rules carefully, and as stated in the footnote,
we have only Thorell’s opinion that there was a mistake in the translation from
the Greek. Actually, Walckenaer did not give any derivation, and on pg 51 in
Tabl. Aran, (of which I have a personal copy) appears this line ’G. Agelene
(Agelena) and in a footnote is given the species naevia. That it is not a
typographic error can be shown by the fact that the spelling is the same in the
running head of the page, and in the index. See opinion 34 of the Rules, and
read what I say on page 52 of my monograph. See also p. 132 in Thorell’s
“European Spiders”. Gaining momentum, B. J. continued “I am not adverse to
changing the orthography when it can be shown that there is justification, e.g.,
Chiracanthium, p. 369, since this is covered by the transcription matter in the
Rules. But I certainly am not going to change just because Pete, by fiat, issues
an edict!” This led to a further expose and criticism of Petrunkevitch’s preference
of Aranei over Araneae, and B. J.’s strong endorsement of Araneae over
Araneida.

Over the years, B. J.’s correspondence, in a large sense, served as unofficial
newsletters or a clearing house of sorts, since he often had knowledge of who
was doing what and where they were last seen. In these letters, he often expressed
his concern about the lack of jobs for araneologists, lamented the loss of several
of our colleagues to other professions, and referred to those who remained
biologists but never worked on spiders beyond their dissertations as “still-born”
Ph.Ds. All this with genuine sincerity.

B. J.’s outspoken participation in meetings of the A.A.S., his careful research
and reviews, his consummate editorial expertise — B. J., always the scientist and
teacher, will be missed. There will be no fourth edition of B. J.’s “little book”.

I thank the many friends, students, and associates of B. J.’s who contributed
to this biography. I owe special gratitude to: J. F. Anderson, T Cohn, A. Fuller,
W. Gertsch, E. (Haban) Howard, R. Judd, B. Kaston, H. Levi, W. Peck, M.
Muma, J. Rovner, E. Schlinger, B. Vincent, and H. K. Wallace.

VINCENT— B. J. KASTON BIOGRAPHY

289

BIBLIOGRAPHY OF B. J. KASTON

1935 The slit sense organs of spiders. J. Morph., 58:189-209.

1936 and W. B. Becker. Spathius canadensis Ashm., a parasite of Hylurgopinus rufipes (EichhofO- J.
Econ. Ent.; 29:807.

— .The senses involved in the courtship of some vagabond spiders. Entom. Americana, n.s., 16:97-
167.

— . The morphology of the elm bark beetle, Hylurgopinus rufipes (EichhofO- Connecticut Agr.
Exper. Sta. Bull., 387:613-650.

1937 The distribution of black widow spiders. Science, 85:74.

— . and D. S. Riggs. Studies on the larvae of the native elm bark beetle. J. Econ. Ent., 30:98-108.

— . Notes on hymenopterous parasites of elm insects in Connecticut. Connecticut Agr. Exper. Sta.
Buil., 396:351-361.

— . Structural anomolies in spiders. Bull. Brooklyn Ent. Soc., 32:104.

— .The black widow spider in New England. Bull. New England Mus. Nat. His., No. 65, p. 1-11, 11
figs.

— . and G. E. Jenks. Dipterous parasites of spider egg sacs. Bull. Brooklyn Ent. Soc., 32:160-165.

— . Notes on dipterous parasites of spiders. J. New York Ent. Soc., 45:415-420.

1938 Notes on little known New England Spiders. Canadian Entom., 70:12-17.

— . Family names in the order Araneae. American Midi. Nat., 19:638-646.

— . Mantispidae parasitic on spider egg sacs. J. New York Ent. Soc., 46:147-153.

— . A note on synonymy in spiders. Entom. News, 49:258-259.

— . Check list of elm insects. Connecticut Agr. Exp. Sta. Bull., 408:235-242.

— . and D. S. Riggs. On Certain habits of elm bark beetles. J. Econ. Ent., 31:467-469.

— . North American spiders of the genus Agroeca. Amer. Midi. Nat., 20:562-570.

— .New spiders from New England, with notes on other species. Bull. Brooklyn Ent. Soc., 33:173-
191.

— . Check list of the spiders of Connecticut. Connecticut Geol. Nat. Hist. Survey Bull., No. 60, p.
175-201.

— . Notes on a new variety of black widow from southern Florida. Florida Ent., 21:60-61.

1939 The native elm bark beetle in Connecticut. Connecticut Agr. Exp. Sta. Bull., 420:1-39.

— . and D. S. Riggs. Report of a survey of elm bark beetle populations in western Connecticut.
Connecticut Agr. Exp. Sta. Bull., 428:80-83.

1940 Another Mantispa reared. Bull. Brooklyn Ent. Soc., 35:21.

— . Flies and other insects that triumph over spiders. J. Tennessee Acad. Sci., 15:413.

1941 Mastophora cornigera (Hentz), a spider new to New England, In: Miscell. Notes, 40th Report
State Entom. [for 1940]. Connecticut Agr. Exp. Sta. Bull., 445:379-380.

1943 Supplementary notes on family names in the Order Araneae. Amer. Midi. Nat., 30:765-768.

1945 New spiders in the group Dionycha, with notes on other species. Amer. Mus. Nov., No. 1290, p.
1-25.

— . New Micryphantidae and Dictynidae with notes on other spiders. Amer. Mus. Nov., No. 1292,
p. 1-14.

— . Notes on nematode parasites of spiders. Trans. Connecticut Acad. Sci., 36:241-244.

1946 North American spiders of the genus Ctenium. Amer. Mus. Nov., 1306:1-19.

1948 Spiders of Connecticut. Connecticut Geol. Nat. Hist. Surv. Bull. No. 70, p. 1-874, 2198 figs.

1949 The spiders of Hispaniola by E. B. Bryant. [Review] Quart. Rev. Biol., 24:362.

— . The Salticidae of Panama by A. M. Chickering. [Review] Quart. Rev. Biol., 24:363.

1950 American Spiders by W. J. Gertsch. [Review] J. Econ. Ent., 43:231.

— . American Spiders by W. J. Gertsch. [Review] Quart. Rev. Biol., 25:208-209.

— . Webs in the Wind by W. Duncan. [Review] Quart. Rev. Biol., 25:209-210.

1951 Pseudoscorpions of Illinois by C. C. Hoff. [Review] Quart. Rev. Biol., 26:210.

— . Phalangida from tropical America by C. J. and M. L. Goodnight. [Review] Quart. Rev. Biol.,
26:210.

— . The spider genus Tmarus in Panama by A. M. Chickering. [Review] Quart. Rev. Biol., 26:404.

1952 Guide to some common Colorado spiders by W. van Riper. [Review] Quart. Rev. Biol., 27:216.
— .The arachnid order Solpugida in the United States by M. H. Muma. [Review] Quart. Rev. Biol.,

27:216.

— . The Life of the Spider by John Crompton. [Review] Quart. Rev. Biol., 27:313-314.

290

THE JOURNAL OF ARACHNOLOGY

1953 How to Know the Spiders. Wm. C. Brown, Dubuque, Iowa, 220pp.

1954 Is the black widow spider invading New England? Science, 119:192-193.

— . Structure and function of the genitalia in some American agelenid spiders by R, L. Gering.
[Review] Quart. Rev. Biol., 29:74.

. The spider genera Xysticus, Coriarachne, and Oxyptila, etc., by W. J. Gersch. [Review] Quart.
Rev. Biol., 29:266.

1955 A check list of Illinois spiders. Trans. Illinois Acad. Sci., 47:165-172,

1956 The genus Eustala in Central America by A. M. Chickering. [Review] Quart. Rev. Biol, 31:55-
56.

1957 Correct exposure with flash, for close-ups. J. Photog. Soc. Amer., 23:40-43.

1959 Anatomy and Physiology Laboratory Manual and Workbook. Edwards Brothers, Ann Arbor,
Mich., 200 pp.

— . Laboratory manual and study guide for Biology. Edwards Brothers, Ann Arbor, Mich., 139 pp.
— . The Tarantula by W. J. Baerg. [Review] Quart. Rev. Biol., 34:146-147.

— . Notes on pompilid wasps that do not dig burrows to bury their prey. Bull. Brooklyn Ent. Soc.,
54:103-113.

— . The spider family Dictynidae by R. V. Chamberlin and W. J. Gertsch. [Review] Quart. Rev.
Biol., 34:313.

— . The spider genera Achaearanea, Theridion, and Sphyrotinus from Mexico by H. W. Levi.
[Review] Quart. Rev. Biol., 35:153.

1961 A check list for directors arranging slide programs. Turtox News, 39:284.

— . Spider gynandromorphs and intersexes. J. New York Ent. Soc., 69:177-190.

1962 Ocular anomalies in spiders. Bull. Brooklyn Ent. Soc., 57:17-21.

1963 Deformities of external genitalia in spiders. J. New York Ent. Soc., 71:30-39.

— . Abnormal duplication of the epigynum and other structural anomalies in spiders. Trans. Amer.
Microsc. Soc., 82:220-223.

1964 The evolution of spider webs. Amer. Zook, 4:191-207.

1965 Some little known aspects of spider behavior. Amer. Midi. Nat., 73:336-356.

1966 Evolution of the web. Natural History, 75(4):126-133.

1967 Contribution a I’etude des Araignees de la famille des Lycosidae, etc., by Y. Guy. [Review]
Quart. Rev. Biol., 42:550.

1968 Remarks on black widow spiders, with an account of some anomalies. Entom. News, 79:113-
124.

1970 Comparative biology of American black widow spiders. Trans. San Diego Soc. Nat. Hist.,
16:33-82.

1972 How to Know the Spiders, 2nd. Ed., Wm. C. Brown, Dubuque, Iowa, 290 pp.

1972 Some name changes in spiders. Entom. News, 83:105-106.

— . Book reviews: Yates, Spiders of Southern Africa; Child, Australian Spiders; and Mascord,
Australian Spiders in Colour. Amer. ArachnoL, No. 7:11-14.

. On Tessarops maritima, a nomen oblitum in spiders. Entom. News, 83:117-118.

— . Some notes on pronunciation in Arachnology. Notes Arachnol. Southwest, No. 3:5-6.

— . Web making in young Peucetia. Ibid., pp. 6-7.

— . Book reviews: Main, 1964, Australian Spiders. Hickman, V. V., 1967, Some Common
Tasmanian Spiders, and Forster, R. R., 1967, 1968, 1970, The Spiders of New Zealand.
American Arachnol., No. 8:19-20.

1973 Four new species of Metaphidippus, with notes on related jumping spiders (Araneae, Salticidae)
from the eastern and central United States. Trans. Amer. Microsc. Soc., 82:106-122.

1974 Remarks on the names of families and higher taxa in spiders. J. Arachnol., 2:47-51.

— . Book review. Spiders of the United States by R. Headstrom. J. Arachnol., 2:205-207.

1975 Book review. New Zealand Spiders by R. R. and L. M. Forster. J. Arachnol., 3:211.

1976 Supplement to the Spiders of Connecticut. J. Arachnol., 4:1-72.

1978 How to Know the Spiders, 3rd. ed., Wm. C. Brown Co., Dubuque, Iowa, 272 pp.

. Recent bites by spiders other than Latrodectus and Loxoceles. California Vector Views, 25:25-
27.

1981 Spiders of Connecticut. Bull. Dept. Environm. Prot., State Geok Nat. Hist. Surv., 70:1020 pp.

. W. J. Gertsch: a bibliography and biography. Bulk Amer. Mus, Nat. Hist., 170:7-14.

1982 Additional ocular anomalies in spiders. J. Arachnol., 10:279-281.

VINCENT— B. J. KASTON BIOGRAPHY

291

1983 Book review. Spider Communication, Mechanisms and Ecological Significance, P. N. Witt and
J. S. Rovner (eds.). J. ArachnoL, 11:103-104.

— . Aldred, D. M. and B. J. Kaston. A list of Utah spiders, with their localities. Great Basin Nat.,

43:494-522.

1985 Book review. The spiders of Great Britain and Ireland by M. J. Roberts. J. ArachnoL, 13:275-
276.

Invited paper.

Downes, M. F. 1987. Postembryonic development of Lactrodectus hasselti Thorell (Araneae,
Theridiidae). J. Arachnol., 14:293-301.

FEATURE ARTICLES

POSTEMBRYONIC DEVELOPMENT OF
LATRODECTUS HASSELTI THORELL (ARANEAE, THERIDIIDAE)

Michael F. Downes

Zoology Dept., James Cook University of North Queensland
Post Office, James Cook University
Queensland, 4811, Australia

ABSTRACT

In northern Queensland the eggs of Latrodectus hasselti have a lower developmental temperature
threshold than do postembryos which in turn have a lower threshold than first instar spiderlings prior
to emergence from the egg sac. Emerged spiderlings, however, develop normally at temperatures
below that which would prevent emergence from the egg sac. Arrested development of free-living first
instar spiderlings may require a temperature of 10°C or less (Forster 1984). The subsequent stadia
for both sexes are in general shorter than for the American species of Latrodectus which mature later
and live longer than L. hasselti. For all species of Latrodectus for which temperature-controlled data
are available males have consistently longer early stadia than do females, partly compensating for an
asynchrony in time to maturity. L. hasselti males usually matured in the fourth instar, females in the
sixth.

INTRODUCTION

The genus Latrodectus Walckenaer occurs widely throughout the tropical and
temperate regions of the world. Since 1959 the Australian species was known as
Latrodectus mactans hasselti and was considered conspecific with L. mactans
Fabricius populations as far distant as the Mediterranean (Levi 1959, 1966).
However, this is no longer tenable, and the taxonomy of the genus is still largely
unresolved (Levi 1983). Recent investigations (Forster and Kingsford 1983,
Forster 1984) uphold the view that it is a species in its own right, Latrodectus
hasselti Thorell, and it appears to have become established in New Zealand
(Forster 1984) where two related species, L. katipo Powell and L. atritus
Urquhart occur (Forster and Kingsford 1983).

Being one of the few genera of spiders whose venomous bite has in some cases
been fatal to humans, Latrodectus has been the subject of numerous clinical
investigations and toxicological studies. In Australia an early survey of araneism
(see Maretic 1984) reported 98 cases, including seven deaths, of envenomation by
the bite of L. hasselti, but concluded that the mortality rate for the population at
large could not be estimated (Ingram and Musgrave 1933). Cleland (1942)
reported further Australian case histories and Hornabrook (1951) gave a clinical
account of L. katipo araneism in New Zealand. Sutherland and Trinca (1978)

294

THE JOURNAL OF ARACHNOLOGY

reported 2144 cases occurring in Australia and New Zealand in the eight years
to 1976.

The general and developmental biology of Latrodectus is best known for those
species occurring in the Americas, a particularly detailed and extensive account
being that of Kaston (1970). Rempel (1957) gave a comprehensive description of
the embryology of L. mactans.

The most notable reports on the postembryonic development of Latrodectus
are those of Deevey and Deevey (1945), Deevey (1949), McCrone and Levi (1964,
1966), Cariaso (1967), Kaston (1970), Gonzales (1979), Forster and Kingsford
(1983) and Forster (1984). Developmental studies of this kind have proved to be
of fundamental importance in the elucidation of complex taxonomic problems in
a number of spider genera including Latrodectus (Seligy 1970).

Some useful data on the postembryonic development of L. hasselti in Australia
were given by Softly and Freeth (1970); unfortunately, that study gave fewer
details for females than for males, and temperature was not controlled. Forster
(1984) was able to contribute significantly to our knowledge of L. hasselti
development using specimens introduced to New Zealand from Australia. The
latter study included the rearing of spiderlings to maturity at a constant
temperature of 25° C, enabling direct comparison with the present data and those
of some other authors.

The data presented in this report were obtained from laboratory studies
conducted in 1973, 1977, 1979 and 1981. Over these years it has become apparent
that although L. hasselti is more numerous in the field in summer, all stages of
its life cycle can be found throughout the year in northern Queensland, as
expected in a tropical region; field populations seemed unusually abundant in the
cooler seasons (June-August) of 1973 and 1975. Regrettably, field studies, other
than that of Smith (1971) which detailed the habitat preferences of L. katipo in
New Zealand, have to date been neglected in this region.

MATERIALS AND METHODS

The laboratory conditions under which this study was conducted, unless
otherwise stated, included a constant temperature of 25° C and a 14/10 hour
light/ dark cycle.

Latrodectus hasselti females of field origin in Townsville produced in the
laboratory the egg sacs required for the study. For determination of hatch times
and duration of early stadia, eggs were put into glass cavity blocks, the glass lids
of which were separated from the rim of the blocks by a layer of non-absorbent
cotton wool, with a little vaseline as adherent. This allowed for adequate aeration
and easy observation while excluding mites and fungal spores. Intact egg sacs
were observed simultaneously to disclose the effect, if any, that exposure of the
eggs in the described manner might have on the normal course of development.

The main series of developmental studies used spiderlings that had emerged
from intact egg sacs in the normal way. In all, 440 spiderlings were used, these
being derived from 17 egg sacs each of which was produced by a different female
parent. Each spiderling was isolated in a glass tube 50 mm x 25 mm with a
perforated plastic stopper.

Insects utilized as food for the developing spiders included Drosophila
melanogaster Meigen, Tenebrio molitor L. larvae (mealworms), muscoid flies and

DOWNES— POSTEMBRYONIC DEVELOPMENT OF LATRODECTUS HASSELTI

295

cockroach early nymphs. Prey of a size appropriate to each spider was given two
or three times a week; food supply was therefore varied and adequate but not
quantified. Water was not provided. Forster (1984) has confirmed that the
provision of water is not only unnecessary for rearing Latrodectus but may be
detrimental.

Weighing of food before and after feeding would be the only method of
quantifying food intake and even that would not allow feeding to be rigorously
correlated with development, for three reasons. First, disturbance of the spider
and its web by meticulously removing food remains would be detrimental, to a
varied and unknown extent, to normal development. Second, the amount of
energy expended in capturing prey varies enormously between individual spiders,
depending to a large degree on the extent of their webs. Third, the food provided
is only assumed to be nutritionally suitable. One recent study (MacKay 1982)
shows that L. hesperus Chamberlin and Ivie is an important predator of the ant
Pogonomyrmex rugosus Emery, but by and large natural prey preferences and
nutritional requirements are poorly understood. An instance of this is the
preference shown by L. mactans and L. variolus Walckenaer for housefly maggots
in the study of McCrone and Levi (1964); housefly maggots were rejected,
without exception, on the numerous occasions they were offered to L. hasselti in
the present study. Tenebrio larvae, however, were almost invariably eaten but
they were not used extensively as food because they were not easily trapped in
the web and they sometimes ate dead or weak spiders, posing a threat to spiders
undergoing a molt.

Of the 440 first instar spiderlings comprising the main series, 288 subsequently
provided the developmental data reported here. There were two sources of
attrition. First, data for specimens that died before their sex could be determined
were excluded. Second, ten specimens of each instar to maturity for each sex
were taken for a morphometric study; where this involved destruction of the
specimen before its sex was determined, the data on its development were
excluded.

It cannot be overemphasized that the results of this and similar studies, while
not invalid, can only be interpreted and compared in the light of the
methodological shortcomings. Without field comparisons, laboratory results are
artificial and confusing, and artificial diets have been shown to influence carapace
width and number of molts in lycosid spiders (Whitcomb 1978).

RESULTS AND DISCUSSION

For the present purpose eclosion is considered to be the shedding of the
chorion and to mark the close of the embryonic period. The stage following
eclosion and ending with the first true integument molt is termed the postembryo
and the stage following the first molt is termed the first instar. It is normally the
first instar which emerges from the egg sac in L. hasselti. Among the most recent
struggles with the terminology, since Peck and Whitcomb’s (1970) exhaustive
review, that of Peaslee and Peck (1983) is commendable.

The duration of development prior to emergence from the egg sac is
summarized in Table 1 for four of the five temperatures used in this part of the
study. Five egg sacs were incubated for four months at 15 °C but no development
occurred; on transfer to 25° C for a further six months there was still no

296

THE JOURNAL OF ARACHNOLOGY

Table 1. — Effect of temperature on the development of Latrodectus hasselti between oviposition and
emergence from the egg sac. Figures are given as mean numbers of days, with numbers of egg sacs in
parentheses (figures without (n) are derived estimates — see text). EXP = Eggs exposed for direct
viewing. INT = Egg sac left intact. COMB = Combined mean. NED = No further development. NE
= No emergence from egg sac.

Stage of
Development

Temperature (°C)

18

20

25

30

Oviposition

EXP

41 (3)

11 (9)

9 (4)

to hatching

INT

44 (9)

34(5)

13(21)

11 (3)

COMB

43(12)

34 (5)

12 (30)

10 (7)

NFD-(ll)

Hatching to

EXP

25 (1)

10 (9)

5 (4)

first molt

INT

17(5)

7 (3)

5 (3)

COMB

25 (1)

17(5)

10 (12)

5 (7)

NE-(l)

NE— (5)

First molt

EXP

to emergence

INT

COMB

11

6

Total

EXP

(Oviposition

INT

33 (47)

21 (19)

to emergence)

COMB

33 (47)

21 (19)

development although some of the eggs had not dehydrated. Development rarely
proceeded beyond the postembryo stage at 18°C and first instar spiderlings did
not emerge from the egg sac at 20° C; but after four months (from oviposition)
at these temperatures the spiderlings were still viable. Despite these effects, first
instar spiderlings that had hatched and emerged at 25° C developed normally, at
a reduced rate, at 18°C (see Table 2). In Western Australia L. hasselti eggs failed
to hatch at 9°C but spiderlings that had hatched at higher temperatures
withstood periods of 56 days at 7-9° C, emerging from their egg sacs and thriving
on return to room temperature (Softly and Freeth 1970). Similarly, Forster (1984)
reported tolerance of temperatures less than 10°C in first instar L. hasselti
spiderlings.

Whether these temperature tolerances are related to or comparable with the
seasonal overwintering that occurs in other Latrodectus species is unknown.
Smithers (1944) recorded overwintering of up to 212 days in egg sacs of L.
indistinctus O. P. Cambridge (also synonymized with L. mactans by Levi (1959)).
In Tasmania, where a similar phenomenon might be expected to occur, the
oviposition — emergence period for L. hasselti is 50 days in January (summer) and
70 days in November (spring) (Hickman 1967). Data for the winter months would
be interesting.

The upper limit of temperature tolerance for development to emergence is close
to 30° C (Table 1). Only 10 to 20 spiderlings emerged from four of the egg sacs
and the first instar spiderlings of one sac failed to emerge at all. 30° C would not
be an unusually high temperature in the field in Townsville; it seems likely that
L. hasselti web sites are normally positioned to preclude any prolonged exposure
to temperatures of 30° C or higher. However, Softly and Freeth (1970) reported
that at 37° C L. hasselti eggs hatched in eight days and the first instar spiderlings
emerged from the egg sac seven days later.

DOWNES— POSTEMBRYONIC DEVELOPMENT OF LATRODECTUS HASSELTI

297

Table 2. — Postembryonic development of Latrodectus hasselti at temperatures above and below
optimum. Figures are given as mean numbers of days with numbers of individuals in parentheses.
First stadium does not include the portion spent in the egg sac.

Temp.

(°C)

Sex

Stadium

1

2

3

4

30

Male

14(7)

10(7)

15(7)

12(4)

Female

10(9)

7(9)

12(9)

11 (5)

18

Male

29 (5)

27 (5)

47 (4)

Female

24 (6)

23 (6)

21 (6)

39 (6)

It is difficult to determine the point of hatching and that of the first molt in
intact egg sacs of L. hasselti. The former was judged largely as the point at which
the mass of eggs ceased to roll freely in the sac, and the latter involved a
darkening of the sac contents due to pigment development in the first instar
spiderlings. Table 1 clearly shows that either the incubation period prior to
hatching is overestimated by the above method or exposure of the eggs shortens
the time of incubation. Cariaso (1967) found the reverse: L. hasselti eggs, judged
from observation of intact egg sacs to hatch in 7.5 days at 27-29° C, took 9-13
days to hatch when exposed.

Since the time of emergence from the egg sac is not applicable in the case of
exposed eggs, the time between the first molt and emergence is given as a single
value derived by subtraction of the oviposition — first molt time from the mean
oviposition — emergence time.

Cariaso (1967) found a wider range (16-44 days) of times from oviposition to
emergence at 27-29° C than the corresponding ranges of the present study at 25° C
and 30° C which were 26-43 days and 17-24 days respectively. His mean of 28.7
days is reasonably consistent with the results in Table 1.

The three American species of Latrodectus studied by Kaston (1970) have
exactly corresponding hatch times at 25° C to that of L. hasselti given in Table
1; however, subsequent development up to emergence must be faster than it is
in the Australian species, emergence taking place between 26 days (L. variolus)
and 30 days (L. hesperus). This is peculiar in view of development rates after
emergence.

Figure 1 shows the results of the main series of developmental studies for
which 288 spiders, whose sexes were determined, provided data on the duration
of the post-emergence stadia. Mortality, to the extent it is meaningful in the
artificial conditions of the study, can be taken to be 20-25% (see Methods).

The appropriate data for the earlier (unsexed) stages are included to give a
complete picture of the ranges in the duration of all life-history stages; only the
period from the first molt to emergence from the egg sac is a derived estimate,
without a range, for the reasons given above.

Despite uncontrolled temperature the results of Softly and Freeth (1970) can
be reconciled with the ranges of the present data, except that Fig. 1 shows that
some L. hasselti males molt six times, once more than found for L. hasselti in
Western Australia. Five was the maximum number of molts undergone by L.
hasselti males in the study of Forster (1984), while Cariaso (1967) found a
maximum of but four molts for L. hasselti males in the Philippines.

Two features of the data in Fig. 1 are of particular interest when compared
with the results of those studies on the same or related species of Latrodectus

298

DAYS

40

THE JOURNAL OF ARACHNOLOGY

0 - H

H-M| vMi-E E-Ma.

1 2

3 INSTAR

4 5 6 7 8 INSTAR

Fig. 1. — Mean and range of stadia of Latrodectus hasselti at 25°C. Dark columns, males; light
columns, females; stippled columns, both sexes combined. O is a derived estimate (see text). Numbers
above columns are numbers of individuals. O = oviposition; H = hatching; M = molt; E =
emergence from egg sac.

DOWNES— POSTEMBRYONIC DEVELOPMENT OF LATRODECTUS HASSELTI

299

which were carried out at constant and comparable temperature (Cariaso 1967,
Forster 1984, Forster and Kingsford 1983, Kaston 1970, McCrone and Levi 1964,
1966). The first is that the two studies on L. hasselti show stadia durations
comparable with those given in Fig. 1, especially with regard to the absence of
any tendency towards significantly longer duration of the later stadia. By
contrast, Kaston’s (1970) results, and particularly those of McCrone and Levi
(1964, 1966), show that much longer periods of time are spent in the later instars
in the American species of Latrodectus. The results of Gonzales (1979), although
for specimens reared at a temperature that varied between 20-25° C, show the
same feature. A similar, though less marked, pattern of longer late stadia is found
in the New Zealand species (Forster and Kingsford 1983). The second notable
feature is that all six of these comparable studies consistently show, as does Fig.
1, that males spend longer in the early instars than do females. Again, the data
of Gonzales, though not constant temperature data, show precisely the same
feature. It has been suggested (Downes 1981) that this tendency compensates for
their having fewer instars to maturity than females, the latter phenomenon being
a direct result of the sexual dimorphism in size that has evolved in this genus.

Not surprisingly, lower and higher temperatures retard and accelerate
development respectively. Table 2 shows these effects. It is apparent that stadia
remain longer for males than for females regardless of temperature.

The shortest time between oviposition and maturity for a female L. hasselti was
58 days. Females usually matured in the sixth instar, males in the fourth. The
greatest longevity observed for a female L. hasselti was for a specimen that was
collected mature in the field and died 229 days later in the laboratory (a lifespan
well in excess of 300 days can be inferred from this). For a male, 136 days from
emergence was the maximum recorded longevity at 25° C, and 211 days from
emergence at 18°C. The American species of Latrodectus, then, develop more
slowly, mature later and live longer than L. hasselti of the Philippines and
Australia.

The development of the palpal organs, characteristic of maturing males, was
sometimes apparent in the second instar but was occasionally not evident even
in the third instar. Those individuals that died before their fourth instar,
therefore, unless clearly males, were excluded from the study. Of the remaining
individuals 111 were male and 177 female. Males tended to construct inferior
webs and expend more energy in trapping prey than did females; more males
than females may therefore have died before maturity. A similar situation may
have influenced the observations of Softly and Freeth (1970) who reported a
maleTemale sex ratio of 1:5 for L. hasselti. There is no reliable report of a sex
ratio other than 1:1 in the literature on Latrodectus.

There remains, in this and other similar studies, a marked variability in the
number and duration of the stadia in Latrodectus (see Kaston 1970), despite
stringent efforts to apply uniform conditions. Apart from the difficulties
(discussed above) of controlling food supply and feeding there is another factor,
other than inherent genetic variability, contributing to this: the production and
use of trophic eggs (Downes 1985). Reports of the occurrence of this
phenomenon continue to accumulate and it occurred in the present study as it
did in Kaston’s (1970) study of Latrodectus in the United States. Its effects,
investigated by Schick (1972) and Valerio (1974), may influence subsequent
development more than has hitherto been recognized. An extensive investigation

300

THE JOURNAL OF ARACHNOLOGY

has been made of egg-feeding in Theridion rufipes Lucas, one of four Townsville
theridiid spiders, including L. hasselti, whose first instar spiderlings have been
seen to feed on inviable eggs prior to emergence from the egg sac. The results
of that study will be available in the near future.

ACKNOWLEDGEMENTS

I thank Lyn Forster, Rhondda Jones, Barbara Main and Alister Spain for their
valuable comments on the numerous forerunners of this paper.

LITERATURE CITED

Cariaso, B. L. 1967. A biology of the black widow spider, Latrodectus hasselti Thorell (Araneida:
Theridiidae). Philippine Agric., 51:171-180.

Cleland, J. B. 1942. Injuries and diseases in Australia attributable to animals (insects excepted). Series
V; mammals, fish, spiders, mites and ticks et cetera, shell fish, sponges, protozoa. Med. J.
Australia, 3 October 1942:313-320.

Deevey, G. B. 1949. The developmental history of Latrodectus mactans (Fabr.) at different rates of
feeding. American Midi. Nat., 42:189-219.

Deevey, G. B. and E. S. Deevey. 1945. A life table for the black widow. Trans. Connecticut Acad.
Sci., 36:115-134.

Downes, M. F. 1981. Sexual dimorphism in Latrodectus (Araneae, Theridiidae). Australian J. Ecol.,
6:289-290.

Downes, M. F. 1985. Fecundity and fertility in Latrodectus hasselti (Araneae: Theridiidae). Australian
J. Ecol., 10:261-264.

Forster, L. M. 1984. The Australian redback spider {Latrodectus hasselti)-. its introduction and
potential for establishment and distribution in New Zealand. Pp. 273-289, In-. Commerce and the
Spread of Pests and Disease Vectors. (M. Laird ed.) Praeger, New York.

Forster, L. M. and S. Kingsford. 1983. A preliminary study of development in two Latrodectus
species (Araneae: Theridiidae). New Zealand Entomol., 7:431-438.

Gonzales, A. 1979. Observaciones bioecol6gicas sobre una especie del genero Latrodectus
(Walckenaer, 1805) del grupo mactans, de Sierra de la Ventana (Provincia de Buenos Aires,
Argentina) (Araneae, Theridiidae) III — Desarollo post-embrionario. Acta Zool. Lilloana, 35:97-
100.

Hickman, V. V. 1967. Some common Tasmanian spiders. Tasmanian Mus. Publ., 1 12 pp.

Hornabrook, R. W. 1951. Studies in preventive hygiene from the Otago Medical School. The katipo
spider. New Zealand Med. J., April 1951:131-138.

Ingram, W. W. and A. Musgrave. 1933. Spider bite (Arachnidism): A survey of its occurrence in
Australia, with case histories. Med. J. Australia, 1 July 1933:10-15.

Kaston, B. J. 1970. Comparative biology of American black widow spiders. Trans. San Diego Soc.
Nat. Hist., 16:33-82.

Levi, H. W. 1959. The spider genus Latrodectus (Araneae, Theridiidae). Trans. American Microscop.
Soc., 78:7-43.

Levi, H. W. 1966. The three species of Latrodectus (Araneae), found in Israel. J. Zool. (London),
150:427-432.

Levi, H. W. 1983. On the value of genitalic structures and coloration in separating species of widow
spiders {Latrodectus sp.) (Arachnida: Araneae: Theridiidae). Verb. Naturwiss. Ver. Hamburg,
26:195-200.

McCrone, J. D. and H. W. Levi. 1964. North American widow spiders of the Latrodectus
curacaviensis group (Araneae: Theridiidae). Psyche, 71:12-27.

McCrone, J. D. and H. W. Levi. 1966. Postembryological development of spiderlings from two
Peruvian Latrodectus populations. Psyche, 73:180-186.

MacKay, W. P. 1982. The effect of predation of western widow spiders (Araneae: Theridiidae) on
harvester ants (Hymenoptera: Formicidae). Oecologia (Berlin), 53:406-411.

DOWNES— POSTEMBRYONIC DEVELOPMENT OF LATRODECTUS HASSELTI

301

Maretic, Z. 1984. Discrimination between the terms Araneism, Araneidism and Arachnidism. Newsl.
British Arachnol. Soc., 40:5.

Peaslee, J. E. and W. B. Peck. 1983. The biology of Octonoha octonarius (Muma) (Araneae:
Uloboridae). J. Arachnol., 11:51-67.

Peck, W. B. and W. H. Whitcomb. 1970. Studies on the biology of a spider, Chiracanthium inclusum
(Hentz). Univ. Arkansas Agric. Exp. St. Bull., 753:1-76.

Rempel, J. G. 1957. The embryology of the black widow spider, Latrodectus mactans (Fabr.).
Canadian J. Zook, 35:35-74.

Schick, R. X. 1972. The early instars, larval feeding and the significance of larval feeding in the crab
spider genus Misumenops (Araneida: Thomisidae). Notes Arachnol. Southwest, 3:12-19.

Seligy, V. L. 1970. A method for studying spiders at various stages of development. Canadian J.
Zook, 48:406-407.

Smith, D. J. 1971. The habitat and distribution of the katipo spider at South Brighton Beach,
Christchurch, New Zealand. New Zealand Entomol, 5:96-100.

Smithers, R. H. N. 1944. Contributions to our knowledge of the genus Latrodectus (Araneae) in
South Africa. Ann. South African Mus., 36:263-312.

Softly, A. and G. E. Freeth. 1970. The laboratory maintenance of Latrodectus mactans. J. Inst. Anim.
Tech., 21:117-126.

Sutherland, S. K. and J. C. Trinca. 1978. Survey of 2144 cases of red-back spider bites. Australia
and New Zealand, 1963-1976. Med. J. Australia, 30 December 1978:620-623.

Valerio, C. E. 1974. Feeding on eggs by spiderlings of Achaearanea tepidariorum (Araneae,
Theridiidae), and the significance of the quiescent instar in spiders. J. Arachnol., 2:57-63.

Whitcomb, W. H. 1978. Lycosid life-history research: a critique. Symp Zook Soc. London, 42:423-
427.

Manuscript received July 1985, revised December 1985.

i ' * :; •.-«■' """"''"""ajMIBI ■ -^T

, V® ' ': H. ffl '

.o.'n'^.'' f'sKv ‘».. "jfitstiBW' ' .'c^^t'.-'K.' '> V-'

I. ■' ."-f '-*" ■ ' 'ij4 « ■' t- ■'^- - *

i’V’»*"mV!K'8'

i' ' a, . ^W'h .If ' '

'-Mt^-?rt^%»‘-'-|,-a4>ri / '■ - ' j^4. .r ,r>5|i

" Vi I

■",; je-

:5,.,v«t:rw- ^.;’:+'&-vJjr “ * ffli

■|j

%%; • Eii -/j^is^

‘

y-i

ife

tw

■Mt'

Jennings, D. T. and J. A. Collins. 1987. Spiders on red spruce foliage in northern Maine. J.
Arachnol., 14:303-314.

SPIDERS ON RED SPRUCE FOLIAGE IN NORTHERN MAINE'

Daniel T. Jennings

Northeastern Forest Experiment Station
USD A Building, University of Maine
Orono, Maine 04469

and

Judith A. Collins

Department of Entomology
University of Maine
Orono, Maine 04469

ABSTRACT

Spiders of 10 families, 16 genera, and at least 21 species were collected from foliage of Picea rubens
Sarg. in northern Maine. Foliage samples were collected from tree crowns using an extendable pole
pruner. Of 157 individuals, erigonids were numerically dominant (36%), followed by philodromids
(18%) and salticids (15%). Mean spider densities/ m^ of foliage area were nonsignificant among
sampling sites, but significantly more {P = 0.05) spiders were collected during the second sampling
period (24-25 July) than during the first sampling period (7-8 June). Differences in spider densities
between sampling periods were attributed to 1) seasonal activities and reproductive cycles of
individual species, 2) weather during collection of foliage samples, and 3) two sampling methods (i.e.,
pole pruner with clamping device or pole pruner with catchment basket). Spider-budworm
relationships and spider-habitat associations are discussed.

INTRODUCTION

Spiders are among the dominant predatory arthropods in forest ecosystems. In
northeastern spruce-fir forests, spider densities are estimated to range from about
75,000/acre (187,500/ hectare) (Morris 1963) to 125,000/acre (3 12,500/ hectare)
(Haynes and Sisojevic 1966a). Despite common occurrence and relative
abundance of spiders, little is known about the arboreal spider fauna of
coniferous tree species. Jennings (1976) summarized spider-faunal studies of
North American trees; only nine species of conifers were included. Recent
faunistic studies include spiders on balsam fir, Abies balsamea (L.) Mill., in New
Brunswick (Renault and Miller 1972); on white fir, Abies concolor (Gord. and
Glend.) Lindl. ex Hildebr., in California (Dahlsten et al. 1977) and in Oregon
(Mason and Torgersen 1983); and on red pine, Pinus resinosa Ait., white spruce,
Picea glauca (Moench) Voss, and white cedar. Thuja occidentalis L., in
Minnesota (Stratton et al. 1979).

'Mention of a commercial or proprietary product does not constitute endorsement by the U.S.
Department of Agriculture, the Forest Service, or the University of Maine.

304

THE JOURNAL OF ARACHNOLOGY

During our studies of the spruce budworm, Choristoneura fumiferana (Clem.),
in northern Maine (Collins 1985), we sampled budworms and spiders on foliage
of 160 red spruce, Picea rubens Sarg., trees. This paper reports spider species
composition on red spruce foliage, compares relative abundances and densities
during two sampling periods, and evaluates two sampling methods for arboreal
spiders.

METHODS

Study Area. — The study area was located on Great Northern Paper Company
lands about 60 km northwest of Millinocket, Piscataquis County, Maine.
Individual study sites (8) were located along the eastern edge of Township 4,
Range 12, WELS (45°58' North; 69° 13' West), about 2.6 km east of Ripogenus
Pond, and 2.5 km northwest of Soubunge Mt. (USGS, Harrington Lake
Quadrangle, 1954). Elevations were about 335 m.

Five strip clearcuts were chosen for study in a spruce-fir forest infested with
the spruce budworm. The strip clearcuts consisted of alternating uncut residual
stands and clearcut strips from which trees had been harvested in 1977. The strips
were oriented east-west, .x = 94°, range 90-96°. The study area had been sprayed
by the Maine Forest Service with Seven-4-oil® for spruce budworm suppression
in 1981. No further chemical-insecticide treatments were made, including the
study year of 1984.

Forest-Stand Characteristics. — Forest-stand parameters were measured on 16
variable-size plots established along the north and south edges of uncut residual
strips (Fig. 1): two plots at each of eight study sites (A, B, C...H). Within each
plot, a species inventory was taken of all trees > 1.0 cm in diameter at breast
height (dbh). Tree heights of three dominant or codominant softwoods were
measured (m) on each plot with a Haga® altimeter. Stand ages were determined
from tree increment cores {n = 48).

Foliage Samples. — Spruce budworm and spider populations were estimated
during two sampling periods. The first sampling, 7 and 8 June 1984,
corresponded with the budworm’s L3-L4 larval stages; the second sampling, 24
and 25 July 1984, corresponded with the budworm’s pupal stage. Within each
plot, 10 dominant or codominant red spruce were selected for sampling. Trees
were flagged and numbered so that the same trees could be used for both
population estimates.

For the first sampling, we used an extendable pole pruner equipped with a
clamping device (Stein 1969) to prune two midcrown branches (> 45 cm) from
each tree. The clamping device holds the branch after it has been severed from
the tree. Pruned branches were lowered or dropped to a ground cloth (3.6 x 3.6
m). All loose larvae and spiders were collected and preserved in vials containing
75% ethanol. Branches, associated vials, and labels were placed in clear plastic
bags (30 X 61 cm) and transported to the laboratory for examination.

The same procedure and sampling intensity were followed for the second
sampling, except that the pole pruner was equipped with a metal hoop and fine
mesh net (16 X 18 mesh, 46 cm diam, 69 cm deep) which formed a basket beneath
the cutting head of the pruner. Severed branches and any dislodged insects or
spiders were caught in the basket and lowered to the ground by disassembling
the poles.

JENNINGS & COLLINS— RED SPRUCE SPIDERS

305

RESIDUAL 1 i ^

STAND I 'i

f ! ** r

i tit i At

1 A* i V* V

i

CLEARCUT

STRIP

- — PLOT

SITI

H— PLOT— ^

J

E A

RESIDUAL ^ ^ t
STAND f i ^

Vi

1- PLOT ^

CLEARCUT

STRIP ^ ^ ^

t- — PLOT — H

Fig. 1. — Schematic of strip clearcut spruce-fir
forest showing residual stands, clearcut strips,
variable plot sizes (16), and two of eight sampling
sites (A, B, C . . . H), Piscataquis County, Maine.

In the laboratory, branches were stored in a walk-in cooler (3-4° C) until
examined. All branches were examined by trained technicians within two weeks
of collection. Foliated branch lengths and widths were measured (cm) and branch
areas (A) calculated by the formula: A = length x width/ 2 (Sanders 1980). Each
branch was cut into small, easily handled pieces (10-15 cm) with hand pruners.
The pieces were then examined carefully for spruce budworms and spiders. All
spiders were preserved in vials containing 75% ethanol. Population densities were
expressed as budworm larvae-pupae and spiders per m^ of branch foliage area.
Sample sizes were: 2 branches/tree x 10 trees/plot x 16 plots (2/site) = 320
branches/ sampling period, and 640 branches total.

Weather. — Because weather affects spider activity, temperature and rainfall
data were obtained from the Great Northern Paper Company’s weather station
maintained at Ripogenus Dam, about 12.2 km SSE of the study area (NOAA,
Climatological Data, New England, vol. 96 (6,7), 1984).

Spider Identifications. — Most of the spiders were identified with the
identification keys and species descriptions of Kaston (1981). Additional
consulted sources were: Chamberlin and Gertsch (1958) for the dictynids; Levi
(1957) for the theridiids; Dondale (1959) for the erigonids; Berman and Levi
(1971) and Levi (1974, 1977) for the araneids; Dondale and Rcdner (1982) for
the clubionids; Dondale and Redner (1978) for the philodromids and thomisids;
and Kaston (1973) for species of Metaphidippus. Following Platnick and Shadab
(1981), we considered Poecilochroa a synonym of Sergiolus.

Because species descriptions are based chiefly on the genitalia, only sexually
mature spiders were identified to species. Juvenile and penultimate stages were
identified to generic level only, except for those of Philodromus, which were
assigned to one of three species groups {rufus, aureolus, or histrio) based on color
patterns of the legs, carapace, and abdomen (Dondale 1961a; Dondale and
Redner 1968, 1975). Representative specimens of all identified species are
deposited in the arachnid collection, U.S. National Museum, Washington, DC.

Data Analyses. — Because the observed spider densities were not distributed
normally (Figs. 2 and 3) and variances were nonhomogenous (Hartley’s test), even
after transformations, we used nonparametric statistics for most analyses. The

FREQUENCY 35. FREQUENCY

306

THE JOURNAL OF ARACHNOLOGY

SPIDERS /BRANCH

Fig. 2. — Frequency distribution of spiders on red spruce foliage, first sampling, 7-8 June 1984,
scataquis County, Maine.

SPIDERS/BRANCH

Fig. 3. — Frequency distribution of spiders on red spruce foliage, second sampling, 24-25 July 1984,
Piscataquis County, Maine.

JENNINGS & COLLINS— RED SPRUCE SPIDERS

307

STP procedure (Sokal and Rohlf 1981) for multiple comparisons, based on the
Mann-Whitney U-test, was used to compare spider densities among sites within
sampling periods {P = 0.05). The Wilcoxon 2-sample test was used to compare
densities between sampling periods for individual sites {P = 0.05) and over all
sites {P = 0.01). Likewise, the Wilcoxon 2-sample test was used to test the
independence of distributions between spider foraging groups (web spinners,
hunters) {P = 0.05), and to compare spider densities between aspects (N, S
exposure) over all sites {P — 0.05). Simple linear regressions were used to test
for possible relationships between spider densities (dependent variable) and spruce
budworm densities (independent variable), where = coefficient of determina-
tion.

To estimate absolute populations, we converted spider densities per m^ of
foliage area to densities per ha (hectare) by the method of Morris (1955).
Populations were computed as:

. N

spiders/ ha = x spiders/ m^ of foliage • ( X

sp =l

BSAsp ), where

X BSAsp = sum of branch surface areas of spruce. The following equation was
used to calculate branch surface area per tree: BSAsp = 2.64 + 3.34 dbhcm, after
Dimond (unpublished data).

RESULTS AND DISCUSSION

Forest Stands.— Species composition by percent basal area (m^/ha) indicated
a predominantly softwood component of red spruce (69.5%); northern white
cedar (11.6%); white spruce (10.2%); white pine, Pinus strobus L. (2.7%); and
balsam fir (4.6%). The hardwood component was composed of paper birch,
Betula papyrifera Marsh. (1.5%), and red maple, Acer rubrum L. (<0.1%).

Red spruce trees averaged 13.7 ± 0.3 cm dbh and 12.4 + 0.4 m height. Mean
stand age was 62.1 ± 1.5 years.

Spruce Budworm Densities. — Mean larval and pupal densities per m^ of foliage
area were: L3-L4 x’s = 27.2 ± 2.6 over all sites, range 7.9 ±2.1 to 39.4 ± 9.1
among sites; pupal x’s = 5.4 ± 0.6 over all sites, range 3.3 ± 1.3 to 8.0 + 2.2
among sites. As expected, larval densities were significantly greater {P = 0.05)
than pupal densities among sites and over all sites (Collins 1985). No doubt, some
of the age-interval mortality between budworm larval and pupal stages was due
to spider predation; however, predation by spiders was not measured in this
study.

Spider Species Composition. — Spiders of 10 families, 16 genera, and at least
21 species were collected from foliage of red spruce in northern Maine (Table 1).
Of the 157 individuals collected, more were web spinners (54%) than hunters
(46%). More species of hunters (12) were collected than web spinners (9);
however, the same number of genera (8) were represented in each spider group.
The erigonids were numerically dominant (36%), followed by the philodromids
(18%) and the salticids (15%). Each of the remaining families accounted for fewer
than 10% of the total individuals.

308

THE JOURNAL OF ARACHNOLOGY

Table 1. — Spiders on red spruce foliage, Piscataquis County, Maine.

Family, Species

1st Sampling

2nd Sampling

WEB SPINNERS

J M F

J M F

%

Dictynidae

4

2.5

Dictyna hrevitarsus Emerton

1

Dictyna sp.

3

Theridiidae

4

2.5

Theridion montanum Emerton

3

Theridion murarium Emerton

1

Linyphiidae

10

6.4

Pityohyphantes costatus (Hentz)

1

4

Pityohyphantes sp.

5

Erigonidae

56

35.7

Grammonota angusta Dondale

1 2

1 11

Grammonota sp.

41

Araneidae

11

7.0

Eustala anastera (Walckenaer)

1

Bust ala sp.

1

Neoscona arahesca (Walckenaer)

1

Araneus sp.

5

Araniella displicata (Hentz)

1

Araniella sp.

2

Subtotals

1 3

57 1 23

85

54.1

HUNTERS

J

M

F

J

M

F

S

%

Gnaphosidae

2

1.3

Haplodrassus sp.

1

Sergiolus sp.

1

Clubionidae

3

1.9

Cluhiona canadensis Emerton

1

Cluhiona sp.

2

Philodromidae

29

18.5

Philodromus exilis Banks

1

1

Philodromus placidus Banks

2

Philodromus sp. {rufus grp.)

2

12

Philodromus sp. {aureolus grp.)

1

6

Philodromus sp. {histrio grp.)

4

Thomisidae

15

9.6

Misumena vatia (Clerck)

1

Xysticus discursans Keyserling

1

Xysticus punctatus Keyserling

1

Xysticus sp.

4

8

Salticidae

23

14.6

Salticus scenicus (Linnaeus)

1

Metaphidippus flavipedes

1

1

1

2

(G. & E. Peckham)

Metaphidippus sp.

2

15

Subtotals

9

2

4

49

1

7

72

45.9

Totals

9

3

7

106

2

30

157

100.0

JENNINGS & COLLINS— RED SPRUCE SPIDERS

309

Web spinners were represented by only two families, two genera, and two
species in the first sampling; this compares with five families, eight genera, and
nine species in the second sampling. Hunting spiders were represented by three
families, five' genera, and six species in the first sampling; this compares with five
families, six genera, and nine species in the second sampling. Combining web
spinner and hunter groups, the number of taxa doubled between sampling
periods; i.e., five families, seven genera, and eight species in the first, and 10
families, 14 genera, and 18 species in the second.

Juveniles of the Philodromus rufus group (Table 1) likely represent both P.
exilis Banks and P. placidus Banks; juveniles of the aureolus group probably are
P. pernix Blackwall, a common inhabitant of coniferous foliage. Juveniles of the
histrio group are most likely P. mysticus Dondale and Redner, a species collected
from black spruce, Picea mariana (Mill.) B.S.P., from white spruce, and from
balsam fir (Dondale and Redner 1975).

Spider Age and Sex Ratios. — The ratio of juvenile to adult spiders was 0.9 for
the first sampling and 3.3 for the second sampling. The ratio of males to females
was 0.4 for the first sampling and 0.1 for the second sampling. Seasonal activities
and reproductive cycles may account for some of these differences in life stages
between sampling periods. For example, some arboreal species attain reproduc-
tive maturity and lay eggs that produce young spiderlings in midsummer to late
summer (Dondale 1961b).

Spider Densities. — Only 5.3% of the 320 pruned branches had one or more
spiders at the first sampling (Fig. 2). During the second sampling, 24.7% of the
320 pruned branches had one or more spiders (Fig. 3). The maximum number
of individuals per branch was two during the first sampling, and six during the
second sampling.

Frequency distributions between spider foraging groups (web spinners, hunters)
were not significantly different {P = 0.05) by the Wilcoxon 2-sample test.

Mean densities of spiders per m^ of foliage area ranged from 0.0 to 1.4 for the
first sampling and from 1.5 to 16.6 for the second sampling (Table 2). None of
these densities were significantly different {P = 0.05) among sites by the Mann-
Whitney U-test, STP procedure (Sokal and Rohlf 1981). All sites showed
increases in spider densities from the first to the second samplings; all but sites
A, B, and D showed significantly greater densities by the Wilcoxon 2-sample test
{P — 0.05) for the second sampling period. Between sampling periods, overall
mean densities were significantly greater (Wilcoxon 2-sample test; P = 0.05) for
the second sampling period (Table 2). Possible factors contributing to these
differences were: 1) seasonal activities and reproductive cycles of the various
species, 2) weather during field collection of branch samples, and 3) sampling
methods.

We observed 92% more juveniles and 69% more adults during the second
sampling than during the first sampling. No doubt, some of these differences can
be attributed to the reproductive cycles of individual species and the production
of young spiderlings in midsummer. For example, Pityohyphantes costatus
(Hentz) produces eggs in early July in Quebec (Manuel 1984); Araniella displicata
(Hentz) reaches sexual maturity in May and June and produces eggs in early July
in Connecticut (Kaston 1981). However, juveniles of biennial species should have
been prevalent during both sampling periods. At least six of the species found
on red spruce foliage, Theridion murarium Emerton, Araniella displicata,

310

THE JOURNAL OF ARACHNOLOGY

Table 2. — Spider densities on red spruce foliage, Piscataquis County, Maine {n = 320 branches/
sampling). Column means within sampling periods are not significantly different by Mann-Whitney
U-test, STP procedure (Sokal and Rohlf 1981), P = 0.05. Row means (a, b) followed by the same
letter are not significantly different by Wilcoxon 2-sample test, P = 0.05. Overall means (x, y) are
significantly different by Wilcoxon 2-sample test, P< 0.01.

Site

1st Sampling

2nd Sampling

X j

(± S.E.)

X ! VCl

(± S.E.)

A

1.17a

(0.58)

1.47a

(0.76)

B

0.40a

(0.40)

2.34a

(1.00)

C

0.89a

(0.64)

9.69b

(2.91)

D

1.39a

(0.81)

5.42a

(2.47)

E

0.00a

(0.00)

7.20b

(2.29)

F

1.27a

(0.64)

16.57b

(5.86)

G

0.74a

(0.55)

8.67b

(1.88)

H

0.26a

(0.26)

5.75b

(1.95)

Overall

0.76x

(0.18)

7.14y

(1.68)

Misumena vatia (Clerck), Xysticus punctatus Keyserling, X. discursans Keyser-
ling, and Philodromus placidus, have biennial life histories (Dondale 1961b,
1976). Dondale (1961b, p. 785) noted that “biennialism is fairly prevalent among
tree and shrub spiders of the North Temperate Zone.” Juveniles of Xysticus and
Philodromus were collected during both sampling periods but were more
abundant during the second sampling. This suggests that weather or sampling
method differences were important.

Mean daily temperatures were about the same for both sampling periods —
17.2°C (7 June) and 20.8° C (8 June) for the first sampling, and 20.9° C (24 July)
and 17.2°C (25 July) for the second sampling. About equal amounts of
precipitation fell during both branch-collection periods, 1.6 cm (1st) and 1.1 cm
(2nd). During the first sampling, most of the trees (sites A-F) were sampled when
the foliage was dry; whereas, during the second sampling, most of the trees (sites
A-E) were sampled when the foliage was wet. We suspect that spider activity may
have been retarded during the second sampling due to the wet foliage, thereby
increasing the probability of collection. Conversely, spider activity probably was
not adversely affected by the clear, warm, and sunny weather during the first
sampling, thus diminishing the probability of capture. Lowrie (1971) found that
sweep-net catches of Oxyopes spiders were significantly greater in herbaceous
vegetation when the dew was heavy.

The two sampling methods used in this study also may have contributed to the
differences in spider densities between sampling periods. Some spiders may have
been lost with either sampling method, i.e., a pole pruner equipped with a
clamping device, or a pole pruner equipped with a basket. However, we suspect
that spiders were dislodged more readily and lost without the catchment basket.
Although Churcher (1981) showed there were no significant differences in
budworm counts on red spruce foliage sampled with and without baskets in New
Brunswick, spiders may be more susceptible to disturbance, particularly species
that build their webs between branch apices. Because manipulation of a pole
pruner equipped with a basket is difficult, adjacent branches often are brushed;
this may add spiders to a branch sample. Both sampling methods need to be
tested at the same time and place for spiders.

JENNINGS & COLLINS—RED SPRUCE SPIDERS

311

We found no significant difference in spider densities by aspect over all sites.
Trees north (southerly exposed) and trees south (northerly exposed) of the
clearcut strips had comparable populations per of foliage area; N tree x’s =
0.48 ± 0.20 (1st sampling), 8.67 + 1.71 (2nd sampling); S tree x’s = 1.05 + 0.33
(1st sampling), 5.60 ± 1.08 (2nd sampling). Although some individual sites varied,
overall means within sampling periods were not significantly different between
aspects by the Wilcoxon 2-sample test, P = 0.05.

Absolute population densities were estimated as 68,995 spiders/ ha for the first
sampling and 645,853 spiders/ ha for the second sampling. The latter density is
double that reported by Haynes and Sisojevic (1966a) for spiders in a balsam fir
stand in northern New Brunswick. We suspect that spruce may support more
spiders, both individuals and species, than fir. Stratton et al. (1979) found that
white spruce had the highest number of individuals and species compared with
red pine and white cedar in Minnesota.

Spider-Budworm Relationships. — Regression analyses indicated very weak
correlations between spider and budworm densities on red spruce foliage; =
0.004 {P = 0.28) for the first sampling, and R^ = 0.018 (P = 0.02) for the second
sampling, where n = 320 branches each sampling period. More refined sampling
techniques are needed to fully assess the density independent-dependent
relationships between spiders and spruce budworms.

All life stages of the spruce budworm — eggs, larvae, pupae, and adults — are
susceptible to predation by spiders. Some of the species found on red spruce
foliage during this study have been observed in previous studies feeding on
various stages of the spruce budworm. The salticid, Metaphidippus flavipedes (G.
& E. Peckham), feeds on eggs and first instars (Jennings and Houseweart 1978).
The web spinners, Theridion murarium and Araniella displicata, capture
budworm moths in their webs (Jennings and Crawford 1985). In laboratory tests,
Grammonota angusta Dondale readily captured and fed on second instars of the
spruce budworm (Haynes and Sisojevic 1966b). Dictyna phylax Gertsch and Ivie,
a species related to D. brevitarsus Emerton, caused considerable mortality to
second instars of the spruce budworm on “stocked” balsam fir branches in New
Brunswick (Renault and Miller 1972); survival of budworm larvae was only 3%
on branches with D. phylax and 60% on branches without spiders.

Detailed serological studies of spider predation on the spruce budworm by
Loughton et al. (1963) indicated that: 1) the erigonids, particularly Grammonota
pictilis { = G. angusta according to Dondale (1959)), were the most important
group because of their large numbers, 2) the theridiids were the most effective
predators, based on percentages showing positive feedings on budworm, and 3)
the salticids were important predators at all stages of larval development,
including the late instars. Mortality during the late larval stage influences
generation survival of the spruce budworm (Morris 1963). At low larval densities.
Watt (1963) estimated that only 0.46 larvae/ 10 ft^ (0.49 larvae/m^) of foliage
would have to be eaten by predators, including spiders, to account for a decrease
in population survival rate. Additional studies are needed to fully evaluate the
predatory impact of red spruce spiders on spruce budworm populations,
especially at low pest densities.

Spider Habitat Associations. — None of the species of spiders collected from red
spruce foliage are restricted to that habitat; most of the adult species have been
collected from two or more species of conifers (Jennings and Collins 1987). Our

312

THE JOURNAL OF ARACHNOLOGY

collections of Salticus scenicus (Linnaeus), Haplodrassus sp., and Sergiolus sp.
appear unusual because S. scenicus generally is associated with domestic habitats
(Kaston 1981) and species of both gnaphosid genera generally are ground dwellers
(Platnick and Shadab 1975; Kaston 1981). However, S. scenicus has been
collected from loblolly pine, Pinus taeda L., in Oklahoma (Bosworth et al. 1971);
Haplodrassus sp. from shortleaf pine, Pinus echinata Milk, and loblolly pine in
Arkansas (Peck et al. 1971); Poecilochroa sp. from white fir in Oregon (Mason
and Torgersen 1983) and from shortleaf and loblolly pines in Arkansas (Peck et
al. 1971); and P. montana Emerton from Douglas-fir, Pseudotsuga menziesii
(Mirbel) Franco, in British Columbia (Turnbull 1956).

The crab spider Misumena vatia frequently is collected from flowering shrubs
and forbs; however, it has also been taken from several conifer species (Jennings
and Collins 1987). Collections of Xysticus discursans usually are made by pitfall
taps and sweep nets in both grassland and wooded areas (Dondale and Redner
1978); our collections of this and other spider species from red spruce foliage
represent new, previously unknown habitat associations.

The remaining species of spiders found on red spruce foliage are arboreal in
habit; however, none are restricted to trees and to conifers. Some, e.g.,
Philodromus placidus and Xysticus punctatus, are more frequently collected from
coniferous trees than from deciduous trees.

CONCLUSIONS

With conventional budworm sampling techniques, our results indicate a
relatively sparse spider fauna (21 species) associated with red spruce foliage. No
doubt other sampling methods (e.g., knockdown sprays, beating-cloth collections,
destructive whole-tree samples) and sampling over a more extended period (e.g.,
April-October) would add additional species to the spider faunal list for red
spruce. The objectives of this study were to sample spiders during only two
restricted time periods of budworm development, L3-L4 larval stages and the
pupal stage. With additional samplings, spider species composition, age, sex ratio,
and density are all apt to change due to individual life histories, reproductive
cycles, and seasonal activities. Conventional budworm sampling methods are
useful for determining frequencies and densities of common associated spiders;
however, these methods are inadequate for determining important predator-prey
relationships. More refined sampling methods and direct or indirect methods of
assessing predation are needed (e.g., serological techniques for detecting target-
prey antigens or tagging potential prey with radioisotopes).

Our hypotheses regarding possible factors affecting spider densities (i.e.,
sampling methods, weather conditions during sampling, seasonal activities and
reproductive cycles of individual spider species, and tree physiognomies) need
further testing in independent studies with controlled conditions. Such studies
may elucidate important spider-tree associations.

ACKNOWLEDGMENTS

We thank the Great Northern Paper Company for permission to conduct these
studies on their lands. Richard A. Hosmer, USDA Forest Service, Northeastern

JENNINGS & COLLINS— RED SPRUCE SPIDERS

313

Forest Experiment Station, provided computer programming assistance; John B.
Dimond, Department of Entomology, University of Maine, provided field-crew
assistance for collection of branches. We also thank Drs. Frederick A. Coyle,
Bruce Cutler, David Wise, and Gail Stratton for their review comments. Donald
W. Seegrist, Northeastern Forest Experiment Station, provided statistical review.
This research was supported by funds provided by the USD A Forest Service,
Northeastern Forest Experiment Station, NE-1151, to the Department of
Entomology, College of Life Sciences and Agriculture, University of Maine,
Orono (Cooperative Research Agreement 23-805).

LITERATURE CITED

Berman, J. D. and H. W. Levi. 1971. The orb weaver genus Neoscona in North America. Mus.
Comp. Zool. Bull., 5:465-500.

Bosworth, A. B., H. G. Raney, E. D. Sturgeon, R. D. Morrison, and R. D. Eikenbary. 1971.
Population trends and location of spiders in loblolly pines, with notes of predation on the
Rhyacionia complex (Lepidoptera: Olethreutidae). Ann. Entomol. Soc. Amer., 64; 864-870.
Chamberlin, R. V. and W. J. Gertsch. 1958. The spider family Dictynidae in America north of
Mexico. Amer. Mus. Nat. Hist. Bull., 116. 152 pp.

Churcher, J. 1981. The utility of baskets fitted to pole pruners for the collection of spruce bud worm
larval samples. Canadian For. Serv. Res. Notes, 1:31-32.

Collins, J. A. 1985. Population enhancement and nesting height preferences of trap-nesting wasps
(Hymenoptera: Eumenidae) preying on spruce budworm (Lepidoptera: Tortricidae) in spruce-fir
forests of Maine. MS Thesis, Univ. of Maine. 65 pp.

Dahlsten, D. L., R. F. Luck, E. 1. Schlinger, J. M. Wenz, and W. A. Cooper. 1977. Parasitoids and
predators of the Douglas-fir tussock moth, Orgyia pseudotsugata (Lepidoptera: Lymantriidae), in
low to moderate populations in central California. Canadian Entomol., 109:727-746.

Dondale, C. D. 1959. Definition of the genus Grammonota with descriptions of seven new species.
Canadian Entomol., 91:232-242.

Dondale, C. D. 1961a. Revision of the aureolus group of the genus Philodromus (Araneae;

Thomisidae) in North America. Canadian Entomol., 93:199-222.

Dondale, C. D. 1961b. Life histories of some common spiders from trees and shrubs in Nova Scotia.
Canadian J. Zool., 39:777-787.

Dondale, C. D. 1977. Life histories and distribution patterns of hunting spiders (Araneida) in an
Ontario meadow. J. Arachnol., 4:73-93.

Dondale, C. D. and J. H. Redner. 1968. The imbecillus and rufus groups of the spider genus
Philodromus. Mem. Entomol. Soc. Canada, 55. 78 pp.

Dondale, C. D. and J. H. Redner. 1975. The fuscomarginatus and histrio groups of the spider genus
Philodromus in North America. Canadian Entomol., 107:369-384.

Dondale, C. D. and J. H. Redner. 1978. The crab spiders of Canada and Alaska (Araneae;

Philodromidae and Thomisidae). Canadian Dep. Agric. Publ., 1663. 255 pp.

Dondale, C. D. and J. H. Redner. 1982. The sac spiders of Canada and Alaska (Araneae: Clubionidae
and Anyphaenidae). Canadian Dep. Agric. Publ., 1724. 194 pp.

Haynes, D. L. and P. Sisojevic. 1966a. Predatory behavior of Philodromus rufus Walckenaer
(Araneae: Thomisidae). Canadian Entomol., 98:1 13-133.

Haynes, D. L. and P. Sisojevic. 1966b. The predatory behavior of Grammonota angusta Dondale
(Araneae: Erigonidae). For. Res. Lab., Fredericton, New Brunswick. Internal Rep., M-15. 42 pp.
Jennings, D. T. 1976. Spiders on black walnut. Amer. Midland Nat., 95:111-119.

Jennings, D. T. and M. W. Houseweart. 1978. Spider preys on spruce budworm egg mass. Entomol.
News, 89:183-186.

Jennings, D. T. and H. S. Crawford, Jr. 1985. Predators of the spruce budworm. U.S. Dep. Agric.,
Agric. Handb. 644. 77 pp.

Jennings, D. T. and J. A. Collins. 1987. Coniferous-habitat associations of spiders (Araneae) on red
spruce foliage. J. Arachnol., 14:315-326.

Kaston, B. J. 1973. Four new species of Metaphidippus, with notes on related jumping spiders from
the eastern and central United States. Amer. Microsc. Soc. Trans., 92:106-122.

314

THE JOURNAL OF ARACHNOLOGY

Kaston, B. J. 1981. Spiders of Connecticut. Bull. Connecticut State Geol. Nat. Hist. Surv., 70. 1020

pp.

Levi, H. W. 1957. The spider genera Enoplognatha, Theridion, and Paidisca in America north of
Mexico. Amer. Mus. Nat. Hist. Bull., 112:5-123.

Levi, H. W. 1974. The orbweaver genera Araniella and Nuctenea. Mus. Comp. Zool. Bull., 146:291-
316.

Levi, H. W. 1977. The American orbweaver genera Cyclosa, Metazygia and Eustala north of Mexico.
Mus. Comp. Zool. Bull., 148:61-127.

Loughton, B. G., C. Derry, and A. S. West. 1963. Spiders and the spruce budworm. Pp. 249-268. In
The dynamics of epidemic spruce budworm populations. (R. F. Morris, ed.). Mem. Entomol. Soc.
Canada 31. 332 pp.

Lowrie, D. C. 1971. Effects of time of day and weather on spider catches with a sweep net. Ecology,
52:348-351.

Manuel, R. L. 1984. The egg sac of Pityohyphantes costatus (Hentz) (Araneae, Linyphiidae) and its
phorid parasite. J. Arachnol., 12:371-372.

Mason, R. R. and T. R. Torgersen. 1983. Mortality of larvae in stocked cohorts of the Douglas-fir
tussock moth, Orgyia pseudotsugata (Lepidoptera: Lymantriidae). Canadian Entomol., 115:1119-
1127.

Morris, R. F. 1955. The development of sampling techniques for forest insect defoliators, with
particular reference to the spruce budworm. Canadian J. Zool., 33:225-294.

Morris, R. F, (ed.) 1963. The dynamics of epidemic spruce budworm populations. Mem. Entomol.
Soc. Canada 31. 332 pp.

Peck, W. B., L. O. Warren, and 1. L. Brown. 1971. Spider fauna of shortleaf and loblolly pines in
Arkansas. J. Georgia Entomol. Soc., 6:87-94.

Platnick, N. 1. and M. U. Shadab. 1975. A revision of the spider genera Haplodrassus and
Orodrassus. Amer. Mus. Novitates, No. 2583. 40 pp.

Platnick, N. 1. and M. U. Shadab. 1981. A revision of the spider genus Sergiolus (Araneae,
Gnaphosidae). Amer. Mus. Novitates, No. 21 17. 41 pp.

Renault, T. R. and C. A. Miller. 1972. Spiders in a fir-spruce biotype: abundance, diversity, and
influence on spruce budworm densities. Canadian J. Zool., 50:1039-1046.

Sanders, C. J. 1980. A summary of current techniques used for sampling spruce budworm populations
and estimating defoliation in eastern Canada. Canadian For. Serv., Great Lakes For. Res. Cent.
Rep., O-X-306. 33 pp.

Sokal, R. R. and F J. Rohlf. 1981. Biometry. 2nd ed. W. H. Freeman Co., New York. 859 pp.

Stein, J. D. 1969. Modified tree pruner for twig sampling. USDA For. Serv., Rocky Mount. For.
Range Exp. Sta. Res. Notes, RM-130. 2 pp.

Stratton, G. E., G. W. Uetz, and D. G. Dillery. 1979. A comparison of the spiders of three coniferous
tree species. J. Arachnol., 6:219-226.

Turnbull, A. L. 1956. Spider predators of the spruce budworm, Choristoneura fumiferana (Clem.) at
Lillooet, B. C., Canada. Proc. 8th Pac. Sci. Congr., HIA: 1579-1593.

Watt, K. E. F. 1963. The analysis of the survival of large larvae in the unsprayed area. Pp. 52-63. In
The dynamics of epidemic spruce budworm populations. (R. F. Morris, ed.). Mem. Entomol. Soc.
Canada 31. 332 pp.

Manuscript received July 1985, revised January 1986.

Jennings, D. T. and J. A. Collins. 1987. Coniferous-habitat associations of spiders (Araneae) on red
spruce foliage. J. Arachnol., 14:315-326.

CONIFEROUS-HABITAT ASSOCIATIONS
OF SPIDERS (ARANEAE) ON RED SPRUCE FOLIAGE

Daniel T. Jennings

Northeastern Forest Experiment Station
USD A Building, University of Maine
Orono, Maine 04469

and

Judith A. Collins

Department of Entomology
University of Maine
Orono, Maine 04469

ABSTRACT

Coniferous-habitat associations were determined for 16 species of adult spiders collected from red
spruce foliage in northern Maine. Most of the spiders have been found on three or more (range one
to 15) conifer species. Spider-conifer habitats were positively correlated (r = 0.96) with geographic
states and provinces. Mean conifer-habitat associations did not differ between web spinner and hunter
species. Significantly more species of spiders from red spruce foliage were associated with northern
conifers than with western conifers, but not with southern conifers.

Comparison of the red spruce fauna with 16 selected spider-tree faunal studies showed that 1)
Sorensen’s similarity quotients (QS) were significantly greater for genera than for species, and 2) mean
species QS values were significantly greater for northern than for southern but not for western
conifers. The spider fauna on red spruce is closely allied with that on other northern conifers;
however, none of the 16 spider species are restricted to conifers.

INTRODUCTION

The arboreal spider fauna of North American trees has received little attention;
fewer than 25 coniferous and deciduous tree species have been studied in detail
(Jennings 1976). Recent studies include: spiders on red pine, Pinus resinosa Ait.,
white spruce, Picea glauca (Moench) Voss, and northern white-cedar, Thuja
occidentalis L., in Minnesota (Stratton, Uetz, and Dillery 1979); spiders on white
fir, Abies concolor (Gord. and Glend.) Lindl. ex Hildebr., in California (Dahlsten
et al. 1977) and in Oregon (Mason and Torgersen 1983). Additional spider-tree
association records are scattered throughout the araneological and entomological
literature; most records are incidental to other study objectives.

During our investigations on natural enemies of the spruce budworm,
Choristoneura fumiferana (Clem.), we collected spiders of 10 families, 16 genera,
and 21 species from foliage of red spruce, Picea rubens Sarg., in northern Maine.
Mean spider densities during two sampling periods and spider-budworm

316

THE JOURNAL OF ARACHNOLOGY

relationships were discussed in an earlier paper (Jennings and Collins 1987). Here
we summarize the known coniferous-habitat associations for each of the 16
species found on red spruce foliage, and compare the spider fauna on red spruce
with that of other coniferous species in North America. Our purposes for these
comparisons were: 1) to identify (by spider and conifer species) the known spider-
coniferous habitat associations, 2) to determine possible relationships between
spider-coniferous habitats and geographic states and provinces, 3) to determine
the degree of similarity between the spider fauna on red spruce and that found
on other conifers, and 4) to determine possible patterns of association or
differences in association among geographic regions. This information may be
useful to forest pest managers who are concerned with identifying potentially
important predators of coniferous pests.

METHODS

Habitat Associations. — We searched the araneological and entomological
literature for spider-tree faunal studies and for habitat-association records of
spiders collected from North American conifers. Standard reference sources were
consulted including Biological Abstracts, Zoological Record, and Centre
International de Documentation Arachnologique Liste. In addition, a DIALOG
search was made from the AGRICOLA database at the National Agricultural
Library, Beltsville, Maryland. The database search was limited to North
American literature that included information and identities of both spiders and
trees.

Because tree habitats may vary by spider species, and spiders identified to
generic level only may include numerous species, we restricted our search to the
16 species of adult spiders found on red spruce foliage (Jennings and Collins
1987). In searching the older literature, recent spider synonymies were considered,
e.g., Pityohyphantes phrygianus (Koch) [= R costatus (Hentz)] (Kaston 1981),
Neoscona minima F.O.R-Cambridge [= N. arabesca (Walck.)] (Berman and Levi
1971), and Araneus displicata (Hentz) [= Araniella displicata (Hentz)] (Levi
1974). Consistent with Dondale (1959), records of Grammonota pictilis (O. P.
Cambridge) collected on balsam fir, Abies balsamea (L.) Milk, in New Brunswick
were considered to belong to G. angusta Dondale.

Habitat associations were recorded by spider and tree species. Only coniferous
tree species were considered, though many of the red spruce spiders also occur
on broad-leaved deciduous trees. Common and scientific names of trees follow
Little (1979). Both evergreen and cone-bearing species of the families
Cupressaceae, Pinaceae, and Taxodiaceae were included.

Data Analyses. — Although our literature search did not represent a random
sample (i.e., all known available studies were included), we considered the data
generated by these faunal studies to meet random-sample criteria (i.e., all
potential coniferous habitats were available for study; none were selectively biased
for our comparison purposes). We used nonparametric procedures (Sokal and
Rohlf 1981) for statistical comparisons. The Wilcoxon two-sample test was used
to test for differences between and among means (P = 0.05). For comparisons
involving more than two means, all 2-mean combinations were performed.

Correlation analysis (Sokal and Rohlf 1981) was used as a measure of
association between spider-coniferous habitats and geographic states and

JENNINGS & COLLINS— CONIFEROUS-HABITAT ASSOCIATIONS

317

provinces. We defined a spider-coniferous habitat as the collection of a spider
species from or on a conifer species, e.g., Dictyna brevitarsus Emerton collected
from foliage of Abies balsamea (L.) Mill, in New Brunswick (Renault 1968).
Regression analysis was not used because we were interested in the degree of
association (interdependence) between the two variables, not the dependence of
one on the other. Habitat records without geographic locality were excluded from
the analysis.

Sorensen’s similarity quotient (QS), as defined by Price (1975), was used to
determine the degree of similarity between the spider fauna on red spruce and
that found on other coniferous species. The formula used was: QS = 2c x 100/
(^7 + b), where a = the number of spider genera or species in study A; b = the
number of spider genera or species in study B; and c = the number of spider
genera or species common to both studies.

To determine possible patterns of association among geographic areas, each
conifer species was assigned to one of three broadly defined regions — northern,
southern, or western — based on distributional ranges of trees given by Little
(1979). We then calculated mean QS values for each region and performed
statistical tests to determine possible differences and associations among regions.
Likewise, mean numbers of spider species in common with red spruce were
determined and compared for each region.

RESULTS AND DISCUSSION

Coniferous-Habitat Associations. — The 16 species of adult spiders found on red
spruce foliage and their habitat associations with North American conifers are
summarized in Table 1. With only one exception, Xysticus discursans Keyserling,
most of the spiders have been collected from three or more conifer species (range
1 to 15). Included are habitat affinities with balsam fir, Abies balsamea (L.) Mill.,
and white fir, A. concolor (Gord. and Glend.) Lindl. ex Hildebr.; oldfield
common juniper, Juniperus communis var. depressa Pursh., and Juniperus spp.
(may include /. chinensis L., eastern red cedar, J. virginiana L., and Rocky
Mountain juniper, J. scopulorum Sarg.); tamarack, Larix laricina (Du Roi) K.
Koch, and western larch, L. occidentalis Nutt.; Norway spruce, Picea abies (L.)
Karst., white spruce, P. glauca (Moench) Voss, and red spruce, P. rubens Sarg.;
jack pine, Pinus banksiana Lamb., sand pine, P. clausa (Chapm. ex Engelm.)
Vasey ex Sarg., lodgepole pine, P. contorta var. latifolia Engelm., shortleaf pine,
P. echinata Mill., Jeffrey pine, P. jeffreyi Grev. and Balf., ponderosa pine, P.
ponder osa Dough ex Laws., red pine, P. resinosa Ait., eastern white pine, P.
strobus L., Scotch pine, P. sylvestris L., loblolly pine, P taeda L., and Virginia
pine, P virginiana Mill.; Douglas-fir, Pseudotsuga menziesii (Mirb.) Franco;
baldcypress, Taxodium sp.; and northern white-cedar. Thuja occidentalis L.

Predictably, coniferous-habitat associations were positively and highly corre-
lated (r = 0.96) with geographic states and provinces (Fig. 1). This indicates that,
within limits, the more localities sampled, the greater the likelihood of finding
more coniferous-habitat associations. However, numbers of available coniferous
species vary among states and provinces, and some spider species have limited
ranges.

Considering habitat association by spider foraging group indicates that web
spinners have been found and reported from more species of conifers (x = 7.9

318

THE JOURNAL OF ARACHNOLOGY

Table 1. — Coniferous-habitat associations of spiders from red spruce foliage.

Conifer species

Locality

Reference(s)

Dictyna brevitarsus Emerton

Abies balsamea

New Brunswick

Loughton et al. (1963), Renault (1968)

Renault and Miller (1972)

Picea rubens

Maine

Jennings and Collins (1987)

Picea sp.

New Brunswick

Renault and Miller (1972)

Pinus strobus

Wisconsin

Coppel and Smythe (1963)

Theridion montanum Emerton

Abies balsamea

New Brunswick

Loughton et al. (1963), Renault (1968)

Renault and Miller (1972)

Picea rubens

Maine

Jennings and Collins (1987)

Picea sp.

New Brunswick

Renault and Miller (1972)

New Mexico

Levi (1957)

Pinus ponderosa

New Mexico

Levi (1957)

Theridion murarium Emerton

Abies balsamea

New Brunswick

Loughton et al. (1963), Renault (1968)

Renault and Miller (1972)

Abies concolor

Oregon

Mason and Torgersen (1983)

Juniperus communis

var. depressa

Michigan

Drew (1967)

Juniperus spp.

Kansas

Heinrichs and Thompson (1968)

New Mexico

Levi (1957)

Picea abies

Massachusetts

Taylor (1928)'

Picea rubens

Maine

Jennings and Collins (1987)

Picea sp.

New Brunswick

Renault and Miller (1972)

Connecticut

Kaston (1981)

Pinus banksiana

Michigan

Allen et al. (1970)

Indiana

Lowrie (1948)

Pinus echinata

Arkansas

Peck et al. (1971)'

Pinus ponderosa

New Mexico

Levi (1957)

Pinus resinosa

Ontario

Martin (1966)

Pinus strobus

Wisconsin

Coppel and Smythe (1963)

Massachusetts

Taylor (1928)

Pinus sylvestris

Massachusetts

Taylor (1928)

Pinus taeda

Arkansas

Peck et al. (1971)

Oklahoma

Bosworth et al. (1971)

Pinus virginiana

Maryland

Howden and Vogt (1951)

Pinus sp.

Connecticut

Kaston (1981)

Pseudotsuga menziesii

British Columbia

Turnbull (1956)

Thuja occidentalis

Michigan

Drew (1967)

Pityohyphantes costatus (Hentz)

Abies balsamea

New Brunswick

Loughton et al. (1963), Renault (1968)

Renault and Miller (1972)

Abies concolor

Oregon

Mason and Torgersen (1983)

California

Ohmart and Dahlsten (1979)

Picea glauca

Quebec

Manuel (1984)

Picea rubens

Maine

Jennings and Collins (1987)

Picea sp.

New Brunswick

Renault and Miller (1972)

Pinus contorta

Alberta

Powell (1971)

var. latifolia

Pinus resinosa

Minnesota

Stratton et al. (1979)

Connecticut

Bean and Godwin (1955)

Thuja occidentalis

Michigan

Drew (1967)

JENNINGS & COLLINS— CONIFEROUS-HABITAT ASSOCIATIONS

319

Table 1. — Continued.

Grammonota angusta Dondale

Abies balsamea

New Brunswick

Loughton et al. (1963), Renault (1968)
Renault and Miller (1972)

Larix laricina

Manitoba

Ives (1967)

Picea rubens

Maine

Jennings and Collins (1987)

Pice a sp.

New Brunswick

Renault and Miller (1972)

Pinus banksiana

Manitoba

Bradley and Hinks (1968)

Eustala anastera (Walckenaer)

Abies balsamea

New Brunswick

Levi (1977)

Juniperus spp.

Kansas

Heinrichs and Thompson (1968)

Nebraska

Worley and Pickwell (1927)

Larix occidentalis

Levi (1977)

Picea glauca

New Brunswick

Levi (1977)

Picea rubens

Maine

Jennings and Collins (1987)

Pinus banksiana

Michigan

Allen et al. (1970)

Pinus clausa

Florida

Levi (1977)

Pinus echinata

Arkansas

Peck et al. (1971)

Pinus taeda

Arkansas

Peck et al. (1971), Levi (1977)

Taxodium sp.

Levi (1977)

Neoscona arabesca (Walckenaer)

Abies balsamea

New Brunswick

Renault (1968)

Abies concolor

Oregon

Mason and Torgersen (1983)

Juniperus communis

Michigan

Drew (1967)

var. depressa

Juniperus sp.

Berman and Levi (1971)

Larix laricina

Berman and Levi (1971)

Picea rubens

Maine

Jennings and Collins (1987)

Pinus echinata

Arkansas

Peck et al. (1971)

Pinus strobus

Wisconsin

Coppel and Smythe (1963)

Pinus taeda

Arkansas

Peck et al. (1971)

Oklahoma

Bosworth et al. (1971)

Pinus virginiana

Maryland

Howden and Vogt (1951)

Araniella displicata (Hentz)

Abies balsamea

New Brunswick

Loughton et al. (1963)

Renault and Miller (1972)

Juniperus communis

Michigan

Drew (1967)

var. depressa

Picea glauca

Minnesota

Stratton et al. (1979)

Picea rubens

Maine

Jennings and Collins (1987)

Picea sp.

New Brunswick

Renault and Miller (1972)

Pinus banksiana

Manitoba

Bradley and Hinks (1968)

Michigan

Allen et al. (1970)

Pinus echinata

Arkansas

Peck et al. (1971)

Pinus jeffreyi

California

Dahlsten (1961)'

Pinus ponderosa

California

Dahlsten (1961)

Pinus resinosa

Ontario

Martin (1966)

Pinus strobus

Wisconsin

Coppel and Smythe (1963)

Maine

Procter (1946)

Pinus taeda

Arkansas

Peck et al. (1971)

Oklahoma

Bosworth et al. (1971)

Pseudotsuga menziesii

British Columbia

Turnbull (1956)

Clubiona canadensis Emerton

Abies balsamea

New Brunswick

Loughton et al. (1963)

Renault (1968)

Renault and Miller (1972)

320

THE JOURNAL OF ARACHNOLOGY

Table 1. — Continued.

Picea rubens
Pice a sp.

Thuja occidentalis

Abies balsamea
Abies sp.

Juniper us sp.

Picea glauca

Picea rubens
Pinus strobus

Thuja occidentalis
Abies balsamea

Juniperus communis
var. depressa
Picea rubens
Picea sp.

Pinus banksiana

Pinus echinata
Pinus resinosa
Pinus sp.

Pinus taeda

Juniperus communis
var. depressa
Picea rubens
Pinus banksiana
Pinus echinata
Pinus taeda
Pseudotsuga menziesii

Picea rubens

Abies balsamea
Picea glauca
Picea rubens
Pinus banksiana

Pinus echinata
Pinus resinosa
Pinus strobus
Pinus taeda

Pseudotsuga menziesii

Picea rubens
Pinus echinata
Pinus taeda

Maine

Jennings and Collins (1987)

New Brunswick

Renault and Miller (1972)

Michigan

Drew (1967)

Philodromus exilis Banks

New Brunswick

Renault (1968)

Dondale and Redner (1968)
Dondale and Redner (1968)

Ontario

Nova Scotia

New Brunswick

Dondale (pers. comm.)

Maine

Jennings and Collins (1987)

Ontario

Nova Scotia

Dondale (pers. comm.)

Ontario

Dondale (pers. comm.)

Philodromus placidus Banks

New Brunswick

Loughton et al. (1963)

Renault (1968)

Renault and Miller (1972)

Michigan

Drew (1967)

Maine

Jennings and Collins (1987)

New Brunswick

Renault and Miller (1972)

Manitoba

Bradley and Hinks (1968)

Michigan

Allen et al. (1970)

Arkansas

Peck et al. (1971)

Minnesota

Heimer et al. (1984)

Maine

Procter (1946)

Arkansas

Peck et al. (1971)

Misumena vatia (Clerck)

Michigan

Drew (1967)

Maine

Jennings and Collins (1987)

Manitoba

Bradley and Hinks (1968)

Arkansas

Peck et al. (1971)

Arkansas

Peck et al. (1971)

British Columbia

Turnbull (1956)

Xysticus discursans Keyserling

Maine

Jennings and Collins (1987)

Xysticus punctatus Keyserling

New Brunswick

Renault (1968)

Minnesota

Houseweart and Kulman (1976)

Maine

Jennings and Collins (1987)

Manitoba

Bradley and Hinks (1968)

Michigan

Allen et al. (1970)

Arkansas

Peck et al. (1971)

Minnesota

Heimer et al. (1984)

Wisconsin

Coppel and Smythe (1963)

Arkansas

Peck et al. (1971)

Oklahoma

Bosworth et al. (1971)

British Columbia

Turnbull (1956)

Salticus scenicus (Linnaeus)

Maine

Jennings and Collins (1987)

North Carolina

Ramsey (1941)'

North Carolina

Ramsey (1941)

Oklahoma

Bosworth et al. (1971)

JENNINGS & COLLINS— CONIFEROUS-HABITAT ASSOCIATIONS

32!

Table 1. — Continued.

Metaphidippus flavipedes (G. & E. Peckham)

Abies halsamea

Maine

New Brunswick

Jennings and Houseweart (1978)
Loughton et al. (1963)

Renault (1968)

Renault and Miller (1972)

Drew (1967)

Juniperus communis

Michigan

var. depressa
Picea glauca
Pice a rube ns
Picea sp.

Pinus banksiana

New Brunswick

Manitoba

Michigan

Arkansas

Ontario

Minnesota

Maine

Houseweart and Kulman (1976)
Jennings and Collins (1987)
Renault and Miller (1972)
Bradley and Hinks (1968)

Allen et al. (1970)

Peck et al. (1971)

Martin (1966)

Heimer et al. (1984)

Coppel and Smythe (1963)

Peck et al. (1971)

Turnbull (1956)

Drew (1967)

Pinus echinata
Pinus resinosa

Pinus strobus
Pinus taeda
Pseudotsuga menziesii
Thuja occidentalis

Minnesota

Wisconsin

Arkansas

British Columbia
Michigan

'Collections not separated by tree species.

± 1.6) than have hunters {x = 5.8 ± 1.2); however, the means for these two
groups were not significantly different by the Wilcoxon two-sample test.
Coniferous foliage is a suitable habitat substrate for both spider groups. Stratton
et al. (1979) found greater percentages of web-builders than hunters on pine,
spruce, and cedar in Minnesota; but the relative composition of the two foraging
groups was not significantly different among tree species.

Exotic conifers excepted, i.e., Picea abies and Pinus sylvestris, significantly
more {P = 0.02) species of spiders from red spruce foliage were associated with
northern conifers (x =7.4 + 1.5) than with western conifers (x = 2.9 ± 0.8), but
not with southern conifers (x = 4.4 ± 1.9). This supports our hypothesis that the
spider fauna on red spruce is closely allied with other northern conifers; however,
elements may also occur on southern conifers, and to a lesser degree on western
conifers. Mean numbers of red spruce species associated with western and
southern conifers did not differ significantly.

Among individual species, Theridion murarium Emerton showed the greatest
range of coniferous habitats, being found on at least 15 species of conifers. Six
other spider species have been collected from seven or more conifer species:
Araniella displicata (Hentz) (12), Metaphidippus flavipedes (G. & E. Peckham)
(11), Xysticus punctatus Keyserling (9), Eustala anastera (Walckenaer) (8),
Neoscona arabesca (Walckenaer) (8), and Pityohyphantes costatus (Hentz) (7).
The remaining spiders from red spruce foliage are reported from six or fewer
conifers. Apparently Xysticus discursans has been collected and reported from
only one conifer, Picea rubens\ however, this association may be accidental
because the normal habitat for this spider is near the ground (Dondale and
Redner 1978).

The number of coniferous-habitat associations for individual spider species
does not necessarily correspond with or indicate their habitat specificity. None of
the 16 species found on red spruce are restricted to that tree species, or to

322

THE JOURNAL OF ARACHNOLOGY

CO 16

h-

<

t 14

m

<

^ 12
CO

o '0

Fig. 1. — Relationship between spider-coniferous tr
habitats and states and provinces; correlation 8
analysis, r = 0.96 (Sokal and Rohlf 1981). Each g
point represents the number of conifer species and ^ 6

states and provinces from which one of the red ^
spruce spiders {n= 16) has been reported. uj ^

0

0 2 4 6 8 10 12 14 16

STATES - PROVINCES

conifers. For example, many of the species, including Theridion murarium,
Metaphidippus flavipedes, Araniella displicata, Eustala anastera, and Neoscona
arabesca, also occur on a variety of broad-leaved trees and shrubs. Misumena
vatia (Clerck) and Clubiona canadensis Emerton are frequently collected from
shrubs and forbs; Salticus scenicus (Linnaeus) is found in synanthropic habitats
(Kaston 1981). Xysticus discursans is usually taken by pitfall traps and sweep nets
in both grassland and wooded areas (Dondale and Redner 1978). Grammonota
angusta Dondale, Philodromus placidus Banks, and Xysticus punctatus are
typically found on conifers.

Spider-Faunal Studies.— Sorensen’s similarity quotients (QS values) for 16
spider-tree faunal studies are shown in Table 2. Not included are studies lacking
complete species lists (Dahlsten et al. 1977; Renault and Miller 1972; Stratton et
al. 1979) and studies where tree-habitat association was uncertain (Fox and
Griffith 1976). As expected, QS values were significantly greater {P < 0.01) for
genera than for species, i.e., more genera than species were shared in common
among the studies. QS values for species were generally < 30, no doubt because
sampling methods and intensities varied considerably among the studies.
However, despite these differences, the red spruce fauna showed more similarities
with spider faunas on northern conifers than on southern or western conifers.
Mean species QS values were significantly greater {P = 0.03) for northern
conifers {x = 18.4 + 2.2) than for southern conifers (x = 9.0 ± 1.5), but not
for western conifers (x = 13.2 + 0.8). Southern and western QS means also did
not differ significantly; however, the level of statistical significance {P = 0.08)
shows possible distinction. With only two exceptions, Larix laricina and Pinus
resinosa, QS values for northern conifers were generally > 15; whereas, those for
southern and western conifers were generally < 15.

The apparent similarity among spider faunas on northern conifers is also
evidenced by comparable QS values, particularly for tree species often found
growing in the same forest stand, e.g., Abies balsamea, Pinus strobus, and Thuja
occidentalis (Table 2). Surprisingly, Larix laricina, a common resident of
northeastern spruce-fir forests, had a very low QS value; conversely, Juniperus
communis var. depressa, an inhabitant of old fields and cutover forests, had a
relatively high QS value. We are unable to explain these dissimilarities.

JENNINGS & COLLINS— CONIFEROUS-HABITAT ASSOCIATIONS

323

Table 2. — Comparison of red spruce spider fauna with spider faunas of other coniferous species
by geographic region. QS = ^'orensen’s similarity quotient.

State-

QS value

Conifer species

Province

Genera

Species

Reference

NORTHERN REGION

Abies balsamea

New Brunswick

30.4

19.5

Renault (1968)

Abies balsamea

New Brunswick

37.7

24.0

Loughton et al. (1963)

Juniperus communis

Michigan

57.1

27.8

Drew (1967)

var. depressa

Larix laricina

Manitoba

51.8

6.1

Ives (1967)

Pinus banksiana

Michigan

46.2

21.4

Bradley and Hinks (1968)

Pinus banksiana

Michigan

38.5

18.8

Allen et al. (1970)

Pinus resinosa

Ontario

25.8

10.5

Martin (1966)

Pinus strobus

Wisconsin

59.5

19.4

Coppel and Smythe (1963)

Thuja occidentalis

Michigan

38.5

18.2

Drew (1967)

SOUTHERN REGION

Juniperus spp.

Kansas

37.5

9.5

Heinrichs and Thompson (1968)

Pinus echinata and

North Carolina

25.9

3.2

Ramsey (1941)

P. taeda^

Pinus echinata and

Arkansas

27.7

10.3

Peck et al. (1971)

P. taeda'

Pinus taeda

Oklahoma

32.4

11.6

Bosworth et al. (1971)

Pinus virginiana

Maryland

32.3

10.5

Howden and Vogt (1951)

WESTERN REGION

Abies concolor

Oregon

54.6

14.0

Mason and Torgersen (1983)

Pseudotsuga menziesii

British Columbia

36.7

12.5

Turnbull (1956)

'Collections not separated by tree species.

In addition to geographic region, tree

height

and growth form may also

influence resident spider faunas. For example, within the same northern region,
spider faunas on young, plantation red pine and on tamarack generally had low
QS values compared with other species. Martin (1966) concluded that young red
pine seedlings planted in an old field do not form an influential part of the
community, and ecological conditions remain those of an old field. No doubt the
sparse, clumped needles of tamarack provide fewer microhabitats for spiders than
the dense, dispersed foliage of spruce. Stratton et al. (1979) found the highest
number of spider species and individuals on foliage of white spruce compared
with red pine and northern white-cedar.

Notably absent from these faunistic comparisons, including our own study, are
investigations involving two or more coniferous species sampled at the same time,
place, and intensity. For such studies, we predict that QS values will be much
higher, particularly for conifers growing in the same forest stand and of similar
growth form, height, and age. Under these conditions, faunal similarities may
approach the QS-50 limit, below which communities are arbitrarily considered
distinct (Price 1975). None of the species faunistic comparisons in Table 2
approached this limit, probably because the conifers studied were widely
separated in time, place, and sampling methodology.

Finally, the paucity of information about spider-coniferous habitat associations
is evident from both Tables 1 and 2. Apparently less than 30% of the 82 native
species of conifers (Little 1979; families Cupressaceae, Pinaceae, and Taxodiaceae)
have even been examined for spiders, much less studied in detail. The maximum

324

THE JOURNAL OF ARACHNOLOGY

number of conifer species sampled in any one state or province was four;
likewise, only one conifer, Picea glauca, has been studied in five states and
provinces. We conclude that additional faunistic studies are needed: 1) to define
the spider fauna of any one conifer species, and 2) to determine habitat
specificities and associational ranges of individual spider species. Such
information will help to elucidate potential predator-prey relationships involving
spiders and pests of conifers.

ACKNOWLEDGMENTS

We thank Daniel Starr, who at the time of this study was with USDA Forest
Service, National Agricultural Library, Beltsville, Maryland, for assistance with
literature searches. Richard A. Hosmer provided computer programming
assistance and is with the USDA Forest Service, Northeastern Forest Experiment
Station, Orono, Maine.

We are grateful to our reviewers, Bruce Cutler, Charles D. Dondale, James H.
Redner, and Dave Walter for their constructive comments, and to Donald W.
Seegrist, Biometrician, Northeastern Forest Experiment Station, for statistical
review.

Portions of this research were supported by funds provided by the USDA,
Forest Service, Northeastern Forest Experiment Station, NE-1151, to the
Department of Entomology, College of Life Sciences and Agriculture, University
of Maine, Orono (Cooperative Research Agreement 23-805).

LITERATURE CITED

Allen, D. C., F. B. Knight, and J. L. Foltz. 1970. Invertebrate predators of the jack-pine budworm,
Choristoneura pinus, in Michigan. Ann. Entomol. Soc. Amer., 63:59-64.

Bean, J. L. and P. A. Godwin. 1955. Description and bionomics of a new red pine scale, Matsucoccus
resinosae. For. Sci., 1:164-176.

Berman, J. D. and H. W. Levi. 1971. The orb weaver genus Neoscona in North America. Mus. Comp.
Zool. Bull., 5:465-500.

Bosworth, A. B., H. G. Raney, E. D. Sturgeon, R. D. Morrison, and R. D. Eikenbary. 1971.
Population trends and location of spiders in loblolly pines, with notes of predation on the
Rhyacionia complex (Lepidoptera: Olethreutidae). Ann. Entomol. Soc. Amer., 64:864-870.

Bradley, G. A. and J. D. Hinks. 1968. Ants, aphids, and jack pine in Manitoba. Canadian Entomol.,
100:40-50.

Coppel, H. C. and R. V. Smythe. 1963. Occurrence of spiders on eastern white pine trees infested
with the introduced pine sawfly. Diprion similis (Htg.). Univ. Wisconsin Forestry Res. Notes No.
102. 7 pp.

Dahlsten, D. L. 1961. Life history of a pine sawfly. Neodiprion sp., at Willits, California
(Hymenoptera: Diprionidae). Canadian Entomol., 93:182-195.

Dahlsten, D. L., R. F. Luck, E. 1. Schlinger, J. M. Wenz., and W. A. Copper. 1977. Parasitoids and
predators of the Douglas-fir tussock moth, Orgyia pseudotsugata (Lepidoptera: Lymantriidae), in
low to moderate populations in central California. Canadian Entomol., 109:727-746.

Dondale, C. D. 1959. Definition of the genus Grammonota with descriptions of seven new species.
Canadian Entomol., 91:232-242.

Dondale, C. D. and J. H. Redner. 1968. The imbecillus and rufus groups of the spider genus
Philodromus. Entomol. Soc. Canada Mem., 55. 78 pp.

Dondale, C. D. and J. H. Redner. 1978. The crab spiders of Canada and Alaska (Araneae:
Philodromidae and Thomisidae). Canadian Dept. Agr. Publ., 1663. 255 pp.

JENNINGS & COLLINS^CONIFEROUS-HABITAT ASSOCIATIONS

325

Drew, L. C. 1967. Spiders of Beaver Island Michigan. Publications of the Museum, Michigan State
Univ., 3(3): 153-208.

Fox, R. C. and K. H. Griffith. 1976. Predation of pine cinaran aphids by spiders. J. Georgia
Entomol. Soc., 11:241-243.

Heinrichs, E. A. and H. E. Thompson. 1968. Parasites, predators, and other arthropods associated
with Choristoneura houstonana (Lepidoptera: Tortricidae) on Juniperus species in Kansas.
Canadian Entomol., 100:33-39.

Heimer, S., W. Nentwig, and B. Cutler. 1984. The spider fauna of the Itasca State Park (Minnesota,
USA). Faun. Abh. Mus. Tierk. Dresden, II, Nr. 6, 1983:119-124.

Houseweart, M. W. and H. W. Kulman. 1976. Life tables of the yellowheaded spruce sawfly,
Pikonema alaskensis (Rohwer) (Hymenoptera: Tenthredinidae) in Minnesota. Environ. Entomol.,
5:859-867.

Howden, H. F. and G. B. Vogt. 1951. Insect communities of standing dead pine {Pinus virginiana
Mill.). Ann. Entomol. Soc. Amer., 44:581-595.

Ives, W. G. H. 1967. Relations between invertebrate 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. Midland Nat., 95:111-119.

Jennings, D. T. and J. A. Collins. 1987. Spiders on red spruce foliage in northern Maine. J.
Arachnol., 14:303-313.

Jennings, D. T and M. W. Houseweart. 1978. Spider preys on spruce budworm egg mass. Entomol.
News, 89:183-186.

Kaston, B. J. 1981. Spiders of Connecticut. Bull. Connecticut State Geol. Nat. Hist. Surv., 70: 1020

pp.

Levi, H. W. 1957. The spider genera Enoplognatha, Theridion, and Paidisca in America north of
Mexico. Amer. Mus. Nat. Hist. Bull., 112:5-123.

Levi, H. W. 1974. The orbweaver genera Araniella and Nuctenea. Mus. Comp. Zool. Bull., 146:291-
316.

Levi, H. W. 1977. The American orbweaver genera Cyclosa, Metazygia and Eustala north of Mexico.
Mus. Comp. Zool. Bull., 148:61-127.

Little, E. L., Jr. 1979. Checklist of United States trees (native and naturalized). For. Serv., U.S. Dep.

Agr., Washington, D.C. Agr. Handbk., No. 541. 375 pp.

Loughton, B. G., C. Derry, and A. S. West. 1963. Spiders and the spruce budworm. Pp. 249-268. In
The dynamics of epidemic spruce budworm populations. (R. F. Morris, ed.). Mem. Entomol. Soc.
Canada 31. 332 pp.

Lowrie, D. C. 1948. The ecological succession of the spiders of the Chicago area dunes. Ecology,
29:334-351.

Manuel, R. L. 1984. The egg sac of Pityohyphantes costatus (Hentz) (Araneae, Linyphiidae) and its
phorid parasite. J. Arachnol., 12:371-372.

Mason, R. R. and T. R. Torgersen. 1983. Mortality of larvae in stocked cohorts of the Douglas-fir
tussock moth, Orgyia pseudotsugata (Lepidoptera: Lymantriidae). Canadian Entomol., 115:1119-
1127.

Martin, J. L. 1966. The insect ecology of red pine plantations in central Ontario. IV. The crown
fauna. Canadian Entomol., 98:10-27.

Ohmart, C. P. and D. L. Dahlsten. 1979. Biological studies of bud mining sawflies, Pleroneura spp.
(Hymenoptera: Xyelidae), on white fir in the central Sierrra Nevada of California. HI. Mortality
factors of egg, larval, and adult stages and a partial life table. Canadian Entomol., 1 1 1:833-888.

Peck, W. B., L. O. Warren, and 1. L. Brown. 1971. Spider fauna of shortleaf and loblolly pines in
Arkansas. J. Georgia Entomol. Soc., 6:87-94.

Powell, J. M. 1971. The arthropod fauna collected from the comandra blister rust, Cronartium
comandrae, on lodgepole pine in Alberta. Canadian Entomol., 103:908-918.

Price, P. W. 1975. Insect Ecology. John Wiley & Sons, New York, 514 pp.

Procter, W. 1946. Biological survey of the Mount Desert Region Incorporated. Part VH. Wistar
Institute of Anatomy and Biology. Philadelphia., 566 pp.

Ramsey, H. 1941. Fauna of pine bark. Elisha Mitchell Sci. Soc., 57:91-97.

Renault, T R. 1968. An illustrated key to arboreal spiders (Araneida) in the fir-spruce forests of New
Brunswick. Canada Dep. Fisheries and Forestry, For. Res. Lab., Fredericton, New Brunswick.
Internal Rep. M-39, 41 pp.

Renault, T. R. and C. A. Miller. 1972. Spiders in a fir-spruce biotype: abundance, diversity, and
influence on spruce budworm densities. Canadian J. Zool., 50:1039-1046.

326

THE JOURNAL OF ARACHNOLOGY

Stratton, G. E., G. W. Uetz, and D. G. Dillery. 1979. A comparison of the spiders of three coniferous
tree species. J. Arachnol., 6:219-226.

Sokal, R. R. and F. J. Rohlf. 1981. Biometry. 2nd ed. W. H. Freeman and Company. New York,
859 pp.

Taylor, R. L. 1928. The arthropod fauna of coniferous leaders weeviled by Pissodes strobi (Peck).
Psyche, 35:217-225.

Turnbull, A. L. 1956. Spider predators of the spruce budworm, Choristoneura fumiferana (Clem.) at
Lillooet, B.C., Canada. Proc. 8th Pacific Sci. Congr. Vol. IIIA: 1579-1 593.

Worley, L. G. and G. B. Pickwell. 1927. The spiders of Nebraska. University Studies, University of
Nebraska, 27 (1-4): 1-129.

Manuscript received October 1985, revised January 1986.

Rovner, J. S, 1987. Nests of terrestrial spiders maintain a physical gill: flooding and evolution of silk
constructions. J. Aracnol., 14:327-337.

NESTS OF TERRESTRIAL SPIDERS MAINTAIN
A PHYSICAL GILL: FLOODING AND THE
EVOLUTION OF SILK CONSTRUCTIONS

Jerome S. Rovner

Department of Zoological Sciences
Ohio University, Irvine Hall
Athens, Ohio 45701

ABSTRACT

Individuals of Dysdera crocata and Ariadna bicolor usually drown during the first day of
submersion in aerated water (21 to 25° C). However, if submerged while in their nests, they remain
capable of activity for up to 10 days. Those residing underwater for the average of 3 days or for
a longer period cannot continue to rely on the original oxygen supply of the nest’s air store, which
is less than a 2-day supply. A decrease in dissolved oxygen in the water, as measured in closed
systems, indicates that the bubble held in the nest acts as a physical gill, with oxygen uptake averaging
3 to 4 /ifi, h"‘ for the nests of adult spiders. Thus, the spider’s survival underwater depends on the
nest’s preventing the Ege effect from diminishing the bubble. Flooding, which could occur in any of
the terrestrial habitats of various ancestral arthropods, should be included among the several factors
hypothesized to have favored the evolutionary origin of silk nests in spiders and similar constructions
in certain other arthropod taxa.

INTRODUCTION

The use of silk by submerged spiders to maintain a bubble that acts as a
physical gill has been shown to be significant only in the aquatic species
Argyroneta aquatica (Clerck) (Schollmeyer 1913, Braun 1931, Thorpe 1950).
While the ability of coastal species to survive tidal sumergence attracted the
attention of many workers (e.g., Abrahams 1924, 1926, Arndt 1915, Bethge 1973,
Bristowe 1923, 1930, 1931, Heydemann 1967, Lamoral 1968, McQueen and
McLay 1983, McQueen et al. 1983, Schaefer 1974, 1976), only McQueen and
McLay included tests of spiders that were inside silk nests. They found that the
physical gill effect did not substantially increase the survival time underwater in
the intertidal spider Desis marina (Hector), since this spider’s nest, built beneath
the holdfasts of bull kelp, was always within a cavity that held an air supply
adequate for many days of submergence. But what of inland, terrestrial, ground-
dwelling spiders trapped in their nests during rain-caused flooding, wherein the
nest surface itself is directly covered by water for days in many cases?

Many spiders carry a bubble when submerged, which was shown to act as a
physical gill in two intertidal species (Lamoral 1968), just as it does in many
insects (Ege 1915, Thorpe 1950). In all spiders examined so far — even Argyroneta
aquatica — the hydrofuge hairs are not specialized like the critically spaced

328

THE JOURNAL OF ARACHNOLOGY

hydrofuge hairs of certain aquatic insects; therefore, spiders cannot maintain a
bubble or air film as a permanent plastron (Thorpe 1950). The bubble gradually
disappears during a number of hours because of the outward diffusion of
nitrogen (the “Ege effect”). Without a bubble, the spider cannot extract oxygen
from the water, and it drowns.

Like the wintertime web of A. aquatica, the silk nests of various terrestrial
spiders might prevent the dissolution of the entrapped bubble, thereby
maintaining a physical gill during prolonged submergence. I decided to investigate
this possibility, being curious about how our local, ground-dwelling spiders
survive rain-caused flooding.

Since many or most terrestrial spiders climb vegetation to escape rising flood
waters (Cooke 1962), I chose to study two species which spend much or all of
their time in silken structures built among or beneath stones. Dysdera crocata C.
L. Koch (Dysderidae) builds closed sacs used for molting, brood care, and
retreats that may be occupied for over a month (Cooke 1965) and also can be
found beneath lumps of mud in marshes (Cooke 1967). Ariadna bicolor (Hentz)
(Segestriidae) dwells in a tubular retreat from which it darts for prey capture. (To
save space, I shall use the term “nest” henceforth for both types of structures.)
I assumed that these spiders tend to remain in their nests at the onset of flooding
in nature, since they always did so in laboratory simulations. Such behavior was
observed in the field in a terrestrial clubionid (Clubiona phragmitis C. L. Koch)
by Bristowe (1931), who found the spiders “enclosed in dense silken cells attached
to the underside of stones the tops of which were only just appearing above the
surface of the lake, which had overflowed its ordinary boundaries.”

In the present laboratory study the spiders survived as long as 10 days in their
nests under water that was relatively warm, while they typically drowned within
12-24 h if outside the nest in such water. Measurements made with oxygen
electrodes showed that oxygen diffuses from the water into the nests, thereby
supporting an hypothesis that the nest-maintained bubble serves as a physical gill,
a heretofore unsuspected adaptive advantage of the nests of terrestrial spiders. A
possible flood-survival role for the silk added to the burrows of ancestral spiders
was suggested by this study.

MATERIALS AND METHODS

I collected the two species of spiders beneath or between loose bricks and
stones in locations shaded by vegetation (summer and fall, 1983 and 1984; Athens
County, Ohio, USA). Vertically standing plastic vials (24 mm diameter, 50 mm
height for small individuals; 32 mm diameter, 70 mm height for large) housed
most spiders. Nests were built where the vial wall and bottom came together,
often beneath a piece of plastic screen that leaned against the wall. Drinking
water was injected through the air hole of the cap every few days. Some D.
crocata were housed in plastic cages (7.0 X 12.5 X 7.0 cm high). D. crocata
preyed on isopods, while A. bicolor fed on Drosophila melanogaster and
Tenebrio molitor. Room temperatures ranged from 21 to 25° C, depending on the
time of year. All the housing containers were placed behind and beneath
cardboard sheets to maintain low illumination, like that found beneath stones. I
used a Kahn electronic analytical balance (model TA-450) to weigh spiders.
Means are followed by standard errors of the mean.

ROVNER-SPIDER NESTS MAINTAIN A PHYSICAL GILL

329

Survival Times After Submersion. — To compare resistance to drowning in
spiders without vs. within their nests, I submerged them 10 cm below the surface
in aerated aquarium water. Spiders without nests were given a strip of paper for
a foothold. I used a syringe to withdraw the bubble trapped beneath the screen
cover of the vial after submergence. The vials were held horizontally beneath
glass bowls and had their screen-covered openings facing the center of the 5-
gallon aquarium, where bubbles issued from an air stone connected to an air
pump, a setup modified from Lamoral (1968). Cardboard sheets over and around
the aquarium insured dim illumination, except at brief intervals when I removed
them to view the spiders. Spiders in nests were made partly visible by
backlighting with microscope illuminators during the brief inspection. Water and
home-vial air temperatures were similar, which minimized acclimation effects.

Spiders without nests were removed from the water after a limited period,
usually 12, 18, 24, or 36 h; placed ventral side up (if unable to stand upright);
and given 1 day to recover. Spiders within nests were checked at intervals; I
tapped the vial and, if there was no response, prodded the nest with a dissecting
needle inserted through the screen cover of the vial. If still no response, the vial
was removed from the water and the nest was opened. Such spiders soon showed
activity and, therefore, could have remained submerged a longer time. These
individuals constituted the category of spiders that had “limited residency.”

The majority of nest submersions involved “unlimited residency,” in which case
the spider was removed from the aquarium only after it had “chosen” to leave
its nest. Since spiders without nests can survive some number of hours
underwater, the time recorded for the duration of each unlimited residency
represents the time spent in the nest plus some unknown amount of time in the
water outside the nest, the latter not exceeding the nighttime interval (maximum
= 8h).

I estimated nest volume in D. crocata by filling empty nests with water from
a 0.01 ml-calibrated syringe; values were rounded up to the nearest 0.05 ml. Since
the spider would occupy some space (roughly 15%), nest volumes exceeded
maximum possible bubble volumes.

Oxygen Diffusion into Submerged Nests. — To test an hypothesis that the nest-
trapped bubble serves as a physical gill, I used closed systems based on two
setups, A and B, each providing a check for the other. I compared the decrease
during 2 h in the percent oxygen saturation of previously well-aerated aquarium
water surrounding an inhabited nest to that of such water (fresh sample)
surrounding the same nest, now empty and flattened. A similar approach,
employing a micro-Winkler technique, established a physical gill function for the
bubble adhering to the body in two species of intertidal spiders (Lamoral 1968).

In setup A, a housing jar served as the test chamber and a Markson oxygen
meter (model 230) for the measurements. I added a piece of plastic screen shaped
to encourage nest construction by the spider dwelling near the bottom of each
jar (jar size: 25 mm diameter X 58 mm high).

To make each measurement I used the following procedure. After turning the
jar to a horizontal position, I placed a tiny (2.0 X 8.5 mm) stirring bar inside
near the opening and taped a magnet to the jar’s exterior at the bar’s location,
the latter insuring control of the bar’s location during subsequent steps.
Returning the jar to a vertical position, I slowly filled it with aerated aquarium
water. Next, the jar was screwed upward into a cap through which projected the

330

THE JOURNAL OF ARACHNOLOGY

vertically clamped oxygen probe. (Aquarium sealer filled the tiny gap between the
cut surface of the cap and the probe.) I then placed the probe and its now-
attached jar in a horizontal position and removed the magnet. (If a bubble were
trapped, appropriate steps were repeated to eliminate such a source of error.)
Finally, the probe/jar were positioned 5 mm above the center of a magnetic
stirrer, so that the bar would spin at the far end of the jar, away from the spider’s
nest (Fig. 1).

For the initial and each of the two subsequent hourly oxygen readings, I
switched on the stirrer and oxygen meter, allowing 1.5 min for equilibration
before taking the reading; the stirrer and meter were then switched off. Readings
were made at 22 to 24° C, depending primarily on room temperature but also,
as to the initial reading, on my handling of the jar during the closure procedure.
Because of the latter, the oxygen value for the initial reading was later calculated
to the higher value it would have at the 1 to 2° lower temperature recorded at
the two subsequent hourly readings. (A thermistor inside the probe of the
Markson meter provided all the temperature readings in setup A.) The volume
of water present in the closed jar with the probe tip inserted was 19 m£.

Setup B involved a YSI oxygen monitor (model 53), a macro-bath stirrer
assembly (YSI 5302), and a circulator (Forma 2066). The advantage of this setup
was in having a constant water temperature (±0.02°) throughout each run and
among all runs. I used 25°, the lowest temperature of the range recommended
for the YSI oxygen monitor, to stay close to the temperatures used in setup A,
as well as to avoid subjecting the spiders to a significant change in temperature
from that in their home cage; previous thermal history affects the metabolic rate
measured at a sufficiently different temperature (Anderson 1970). In setup B the
volume of water was approximately 25 mil and, unlike setup A, the stirring bar
spun and the oxygen electrode was on throughout the 2-h period. Individuals
ranged in mass from 70 to 158 mg (mean =116 + 36).

The disadvantage of setup B was that the bath-stirrer required that the stirring
bar be placed on the bottom of the chamber beneath a screen. This prevented
use of the measurement chamber as a site for nest building. Instead, laboratory
nests were built inside 10 mm-diameter X 25 mm plastic screen (no. 18 mesh)
cylinders offered to some of the spiders in their home cages (Fig. 2). A few nests
collected intact in the field were placed into stainless-steel screen (no. 30 mesh)
holders; two such nests were subsequently revealed to be shared with broods —
a 121-mg female with 80 young (88 mg) and a 158-mg female with 115 young
(155 mg) (Fig. 3).

The method of calculating oxygen consumption was as follows. Oxygen
concentration (mg/ ft) at each temperature was obtained from an appropriate table
of oxygen solubilities in water (Wetzel 1983). (Since such tables are for pure
rather than aquarium water, values determined in the present study are only
approximate.) Multiplication by 0.70046 converted the value for mass to a
volumetric value (mft/ft (ibid.), which provided the basis for values expressed as
juft/ mft. Multiplication of the latter by the total volume of water in the chamber
(19 or 25 mft) gave an approximate value for the total volume of oxygen present
in such a volume of water under ideal aerated conditions. This figure was then
multiplied by the difference in percent oxygen saturation between the initial and
final readings, thus yielding a rough estimate of oxygen decrease in the water

ROVNER-SPIDER NESTS MAINTAIN A PHYSICAL GILL

331

Fig. 1 — Setup A for determining oxygen diffusion into submerged nests of Dysdera crocata. The
spider is visible v^ithin the backlighted nest, which was built upon plastic mesh at the bottom of the
originally vertical jar (jar diameter = 25 mm). The stirring bar is located to the left of the probe,
which projects through the cap (at right). This photograph was not taken during an actual run.

during the 2-h period. Half of this value then expressed the approximate drop
in oxygen concentration per hour h~^).

Since a determination of actual oxygen consumption was not a goal of my
study, I did not control for possible minor effects of diel variation. Starving the
spiders prior to testing did minimize the effect of food assimilation (Anderson
1970).

Fig. 2. — Plastic screen cylinder (length = 25 mm) containing a silken nest built by an adult female

Dysdera crocata.

332

THE JOURNAL OF ARACHNOLOGY

Fig. 3. — Field-collected, debris-covered brood nest of Dysdera crocata just removed from the water
after a 2-h run in setup B. The interior and inhabitants remained dry during submergence in the
swirling water. (Body length of now-emerged female = 13 mm)

RESULTS

Survival Times After Submersion. — Spiders submerged without nests {N = 49)
crawled around the vial at intervals, Z). crocata moreso than A. bicolor. D.
crocata lost its adhering bubble and drowned sooner than did A. bicolor. When
removed from the water after 12 h submergence, 10 of 14 D. crocata had
drowned; all 10 of those submerged for 18 h drowned (mean for drownings =
15.0 h). (Tests of 60 spiderlings of D. crocata showed them to be no more
resistant to drowning than the much larger instars: 9 of 20 drowned in a 6-h test;
19 of 20 in a 12-h test; and all of 20 in a 24 h test, all tests at 24° C.) Since
A. bicolor often survived 12 h in preliminary trials, it was tested for longer
periods: 8 of 11 drowned within 18 h; 7 of 9 within 24 h; and all of 5 within
36 h (mean for drownings = 24.2 h) (Fig. 4).

Spiders submerged while in their nests {N = 35), although inactive for most
of the time, occasionally groomed, deposited silk, and re-positioned themselves;
some molted. (I did not determine whether spiders in nests that had not been
submerged showed similar levels and types of activity. I was interested only in
whether spiders in submerged nests entered a state of immobility suggestive of
very low metabolic rates, precluding ‘"normal” behavior; they did not.) Seven
spiders that had not responded to direct prodding of their nests emerged from
their subsequently opened nests; i.e., they had not drowned during the period of
“limited residency.” During one inspection an A. bicolor left its nest (after 5 days
of submergence), wandered in the water for several minutes, and then returned
to its nest for an additional day of residency.

ROVNER— SPIDER NESTS MAINTAIN A PHYSICAL GILL

333

Dysdera crocata

j j j j

6 7 8 9 10

r

0

T

3

Ariadna bicolor

■ ■ ■

■ ■ ■ ■ ■ ■

_j 1 j , j j j

4 5 6 7 8 9 10

DAYS

o no nest; drowned ■ in nest; unlimited residency

☆ no nest; survived ♦ in nest; limited residency

Fig. 4. — Duration of underwater survival in two species of terrestrial spiders tested within {N =
35) or without (N = 49) their nests.

Throughout nest residency, spiders were inside a typically ovoid bubble that
had stabilized at a length greater than that of the spider’s body, although initial
reduction in size had occurred following submergence. For example, after 6 days
a D. crocata still had a bubble large enough to allow the spider to make a 180°
turn and to stand with its legs spread as if on an open surface. However, in some
very thin-walled nests, the bubble decreased to a size little wider than the spider’s
tightly drawn-in legs; indeed, a few individuals rested with their anterior legs
projecting through a nest opening into the water. The nest volumes in D. crocata
ranged from 0.15 to 0.60 ml for solitary individuals and from 0.75 to 0.80 ml
for nests shared by females and their broods.

Most spiders eventually left the nest, even though the bubble was still present
inside. They took part of this nest bubble with them when leaving, carrying the
adhering bubble primarily on the opisthosoma and sometimes being buoyed up
in the water to a variable extent as a result.

For spiders residing at least 1 day, the duration of nest residency in D. crocata
was 3.1 ± 0.9 days {N — 10); maximum = 10 days (21°; 43 mg spider). For those
A. bicolor residing at least 1.5 days, nest residency = 3.2 ± 0.4 days {N — 25);
maximum = 7 days (23°; 20 mg spider) (Fig. 4). Although not included in the
above data, even females with spiderlings or eggs eventually left the nest;
however, they remained near the nest with the aid of a dragline.

Oxygen Diffusion into Submerged Nests. — Decreases in relative oxygen
saturation readings were greater for water surrounding inhabited nests than for
water surrounding empty nests (Table 1). Decreases in readings were due partly

334

THE JOURNAL OF ARACHNOLOGY

Table 1. — Mean (+ S.E.M.) decrease during 2 h in the relative reading for percent oxygen
saturation of water surrounding nests with bubble-enclosed spiders vs. water surrounding empty nests.
Setup A had 19 ml of water; B had 25 ml. Significance of differences (one-tailed paired r-tests, d.f.
= N — 1) and mean volume of oxygen entering an inhabited nest are given. Abbreviations: n, number
of spiders; N, number of paired runs; d, mean difference; D. c., Dysdera crocata; A. b., Ariadna
bicolor. Two females (*) shared their nests with broods.

Setup

Species

n

N

Inhabited

nest

Empty

nest

d

t

P

O2 entry

hr"')

A

D. c.

3

1

10.1± 1.5

4.6 + 0.6

5.5

3.51

<0.01

3.1

A. b.

1

2

6.0+ 1.0

1.0+ 1.0

5.0

2.50

NS

3.0

B

D. c.

5

5

30.8+ 4.6

25.2+ 4.1

5.6

4.63

<0.005

4.1

D. c.*

2

2

62.5+18.6

36.5+20.6

26.0

13.00

<0.025

18.8

to downward drift (which preliminary tests showed to occur at a higher rate in
the equipment of setup B than in that of setup A); also I did not assess the
measurement errors and the mechanical and electrical drift associated with the
use of these electronic oxygen probes. I therefore regarded all readings as relative
measures of p02. Furthermore, I assumed (but had no evidence) that the rate of
drift affected experimental and control readings within each setup similarly, and
that the difference between experimental and control readings was an absolute
measure of oxygen diffusion into the nests.

The approximate volume of oxygen that entered the nest of each solitary D.
crocata averaged 3 to 4 h~* (22 to 25°) (Table 1). The diffusion rates for nests

with solitary spiders were similar between setups A and B, suggesting that both
systems of measurement were reliable. On the basis of mass, group B of solitary
D. crocata (106 + 16 mg) showed oxygen diffusing into nests at a mean rate of
0.038 iii mg"^ h~' (25°). (Lacking the mass of one D. crocata of group A, I could
not make such an estimate from the data of that group.)

DISCUSSION

Spiders submerged while in their intact nests were able to survive for much
longer than those outside of nests: D. crocata inside nests endured submergence
up to at least 16 times (mean = 5.0 times) longer, and A. bicolor did so up to
at least 7 times (mean = 3.2 times) longer than did conspecifics submerged
without nests. Furthermore, the 10- and 7-day maxima for the two species may
not be the upper limits for their survival underwater, since they reflected the
spiders’ “decision” to leave the nest-trapped bubble. Nonetheless, the persistence
of the bubble for these many days, unlike its disappearance from the body of
“naked” spiders within a matter of hours, indicates clearly that the silken
construction can maintain a physical gill far more effectively than can the spider’s
setae. Also, the ability of the spiders to perform various behaviors, as well as
molt, while within the underwater nest shows that the physical gill was able to
provide for an adequate level of oxygen diffusion from the aerated water.

The volume of oxygen entering the nest of each solitary D. crocata, as
estimated from measured changes in percent oxygen saturation of the
surrounding water in closed systems, was 3 to 4 h~‘. This falls within the range

of about 1 to 5 juft h~* determined as the oxygen consumption rates for 10- to
100-mg individuals of the intertidal spider Desis marina in air and at a cooler

ROVNER— SPIDER NESTS MAINTAIN A PHYSICAL GILL

335

temperature (17.5°C) (McQueen et al. 1983). The uptake into the nest on the
basis of mass (0.038 juft mg~' h~^ at 25° C) in D. crocata was about 30 percent
lower than the 0.055 mg"^h~^ oxygen consumption rate for D. marina of

similar mass ‘(100 mg; 17.5°C) in air, which was a rate lower than that of any
previously tested terrestrial spider (ibid.).

The very low rate for oxygen diffusion into nests of D. crocata may reflect one
or more of several factors: (1) D. crocata is more “primitive” than D. marina,
based on the dysderid’s possession of paired, anteriorly located tracheal spiracles
and the desid’s having a single, posterior spiracle. Anderson (1970) found that
spiders with relatively primitive respiratory systems generally have lower
metabolic rates than do spiders with more advanced respiratory systems. (2)
Spiders residing in nests underwater may be less active or have lower metabolic
rates than spiders exposed in air-filled glassware, the latter being the arrangement
used in oxygen consumption studies. (3) The presence of some material in the
nest wall or at the air-water interface may slow gas diffusion. McQueen et al.
(1983) suggested that spiders may produce an excretory product or other material
that could have caused a reduced rate of diffusion across the bubble-water
interface in their study of Desis marina. Nonetheless, while the rate of diffusion
estimated here for the nests of D. crocata may be lower than the rate of oxygen
consumption by this species in air, it is clear that the nest-trapped bubble
provides a physical gill, just as does the temporary bubble carried on the body
of the marine intertidal spiders studied by Lamoral (1968).

Conservative use of the data on oxygen uptake into nests, together with liberal
estimates of the amount of oxygen available in the nest’s initial air store, indicate
that the air store is not adequate for prolonged submersion in D. crocata. For
example, a 100-mg individual using only 3 h~‘ of oxygen needs 72 /xft day~';

however, even the largest nest of a solitary D. crocata holds <0.60 ml of air, i.e.,
<120 jLtft of oxygen, which is less than a 2-day supply. Likewise, the D. crocata
that survived 10 days underwater was in a 0.20 ml nest, i.e., <40 of oxygen.
Even if it could have survived by consuming oxygen at a reduced rate of as little
as 1 lA h~* (and had the ability to take up oxygen at eventually very low partial
pressures), it had less than a 2-day supply in the air store. Thus, the nest’s
maintenance of a physical gill is probably essential for much or most of the
submergence period of spiders residing in nests as long as, or longer than, the
average of 3 days.

“The hazard of flooding is a very real one for mygalomorph spiders, whether
they live in rainforest or deserts” (Cloudsley-Thompson 1983), just as it is for
burrowing lycosid spiders living in salt lakes (McKay 1976), and for numerous
other ground-dwelling forms. My findings suggest that the silk nest enables such
terrestrial spiders to endure a temporary aquatic existence imposed occasionally
during their lives. This is particularly important for any spider trapped while
molting (a lengthy process of many hours in primitive spiders) as well as for
females and their eggs or offspring, which in primitive forms typically share a
nest prior to dispersal of the spiderlings.

Secretion of a silk precursor probably arose in ancestral spiders for the
protection of eggs or sperm (J. Shultz, in prep.). Since the earliest spiders were
most likely burrow dwellers (Decae 1984), the first non-reproductive use of silk
was probably to line the burrow. Due to the surface tension of water, a relatively
simple meshwork would have been sufficient to keep excess soil water from

336

THE JOURNAL OF ARACHNOLOGY

entering a burrow. If the meshwork were later extended to cover the burrow
opening, a bubble would have been trapped during flooding, which could occur
in any of the habitats of ancestral spiders.

Protection from drowning may have been important for the ancestors of other
silk-producing arthropods. Pseudoscorpions use anteriorly secreted silk to build
nests for molting, brood care, and overwintering, as well as nests for intertidal
sites (Gabbutt 1966, Weygoldt 1969). As to silk-using insects, Hinton (1953)
stated that the cocoons of Lepidoptera and Hymenoptera pupating on the ground
maintain a physical gill during rain-caused flooding. The same may be true for
the silken tunnels of the Embioptera (webspinners). Thus, survival during
flooding may have been one of the selection pressures underlying the evolutionary
origin or diversification of silk use within each of several arthropod taxa in which
shelter construction is the primary use for this secretion. In these groups, as well
as in spiders, early adaptations for the improved production and application of
silk probably were favored by the multiple advantages derived from building
some type of enclosure — protection from predation, prevention of desiccation
during drought, and maintenance of a physical gill during flooding.

ACKNOWLEDGMENTS

This study was supported by an Ohio University Baker Fund Award. I thank
W. D. Hummon for an equipment loan and for critical reading of the manuscript,
M. C. Modrzakowski for an equipment loan, M. L. Trivett for permitting me
to use her property for collecting spiders, and J. W. Shultz and V. D. Roth for
pointing out some relevant papers.

LITERATURE CITED

Abraham, H. C. 1924. Some mygalomorph spiders from the Malay Peninsula. Proc. Zool. Soc.
London, 1924:1091-1124.

Abraham, H. C. 1926. A marine spider of the family Attidae. Proc. Zool. Soc. London, 1926:1357-
1363.

Anderson, J. F. 1970. Metabolic rates of spiders. Comp. Biochem. Physiol., 33:51-72.

Arndt, C. H. 1915. Some insects of the between tides zone. Proc. Indiana Acad. Sci., 1914:323-333.
Bethge, W. 1973. Okologisch-physiologische Untersuchungen iiber die Bindung von Erigone
longipalpis (Araneae, Micryphantidae) an das Litoral. Faun.-dkol. Mitt., 4:223-240.

Braun, F. 1931. Beitrage zur Biologie und Atmungsphysiologie der Argyroneta aquatica Cl. Zool. Jb.
(Syst.), 62:175-262.

Bristowe, W. S. 1923. A British semi-marine spider. Ann. Mag. Nat. Hist., Ser. (9) 12:154-156.

Bristowe, W. S. 1930. Notes on the biology of spiders. H. Aquatic spiders. Ann. Mag. Nat. Hist.,
Ser. (10) 6:343-347.

Bristowe, W. S. 1931. Notes on the biology of spiders. IV. Further notes on aquatic spiders, with
a description of a new species of pseudoscorpion from Singapore. Ann. Mag. Nat. Hist., Ser. (10)
8:457-465.

Cloudsley-Thompson, J. L. 1983. Desert adaptations in spiders. J. Arid Environ., 6:307-317.

Cooke, J. A. L. 1962. Spiders from tidal marshes near Flatford Mill, Suffolk. Entomol. Monthly
Mag., 98:21-22.

Cooke, J. A. L. 1965. A contribution to the biology of the British spiders belonging to the genus
Dysdera. Oikos, 16:20-25.

Cooke, J. A. L. 1967. Factors affecting the distribution of some spiders of the genus Dysdera
(Araneae, Dysderidae). Entomol. Monthly Mag,, 103:221-225.

ROVNER— SPIDER NESTS MAINTAIN A PHYSICAL GILL

337

Decae, A. E. 1984. A theory on the origin of spiders and the primitive function of spider silk. J.
Arachnol., 12:21-28.

Ege, R. 1915. On the respiratory function of the air stores carried by some aquatic insects (Corixidae,
Dystiscidae, Notonecta). Z. Allg. Physiol., 17:18-124.

Gabbutt, P. D. 1966. An investigation of the silken chambers of the marine pseudoscorpion
Neobisium maritimum. J. Zook, London 149:337-343.

Heydemann, B. 1967. Die biologische Grenze Land-Meer im Bereich der Salwiesen. Franz Steiner
Verlag, Wiesbaden.

Hinton, H. E. 1953. Some adaptations of insects to environments that are alternately dry and flooded,
with some notes on the habits of the Stratiomyidae. Trans. Soc. British Entomol., 1 1:209-227.

Lamoral, B. H. 1968. On the ecology and habitat adaptations of two intertidal spiders, Desis
formidabilis (O. P. Cambridge) and Amaurobioides africanus Hewitt, at “The Island” (Kommetjie,
Cape Peninsula), with notes on the occurrence of two other spiders. Ann. Natal Mus., 20:151-
193.

McKay, R. J. 1976. The wolf spiders of Australia (Araneae: Lycosidae): 8. Two new species inhabiting
salt lakes of western Australia. Mem. Queensland Mus., 17:417-423.

McQueen, D. J. and C. L. McLay. 1983. How does the intertidal spider Desis marina (Hector) remain
under water for such a long time? New Zealand J. Zool., 10:383-392.

McQueen, D. J., L. K. Panned and C. L. McLay. 1983. Respiration rates for the intertidal spider
Desis marina (Hector). New Zealand J. Zool., 10:393-399.

Schaefer, M. 1974. Experimentelle Untersuchungen zur Bedeutung der interspezifischen Konkurrenz
bei 3 Wolfspinnen-Arten (Araneida: Lycosidae) einer Salzweise. Zool. Jb.(Syst.), 101:213-235.

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

Schollmeyer. A. 1913. Argyroneta aquatica: Biologic mit besonderer Beriicksichtigung der Atmung.
Ann. Biol, lacustre Brux., 6:314-338.

Thorpe, W. H. 1950. Plastron respiration in aquatic insects. Biol. Rev., 25:344-390.

Wetzel, R. G. 1983. Limnology, 2nd ed. Saunders Publ., New York.

Weygoldt, P. 1969. The Biology of Pseudoscorpions. Harvard Univ. Press, Cambridge, Massachusetts.

Manuscript received August 1985, revised January 1986.

V, i!)

"1/ „

' 'i wipi^' ' 'tslw^i^? VT?^') -*• Ir- ' ' '-‘i’-'fffl^;

:'-|j>^

& ' ' #». " ^ •

rV>'^

i|h V ii_

mm".

'% '''’^v '^

T^^:' '‘‘:.„,i „M ., ^ ■„ - ■" ,.:- V^'tt

«UP.»S-;'

' ' ^ *fe“' ''*''® ■^‘*

'''■'■'At*

. . /^i

3^

h.n

,t-.Vi.'‘'7^l.'^/Vv

Git

* ,. J; ^

■■■•■'■*' ■ ■"

i'tu" ■■ '■■'Vi' ., '„‘ ■ , 'jjJ* - •■' ;T^

*^-. ■""■4* #* •"“> '*■ ■ '" s - : ',

!i

.JSII’ ' % "' ■ " ' ' '^ ■'

'{>

Eberhard, W. G. 1987. Web-building behavior of anapid, symphytognathid and mysmenid spiders
(Araneae). J. Arachnol., 14:339-356.

WEB-BUILDING BEHAVIOR OF ANAPID, SYMPHYTOGNATHID
AND MYSMENID SPIDERS (ARANEAE)

William G. Eberhard

Smithsonian Tropical Research Institute
and

Escuela de Biologia, Universidad de Costa Rica
Ciudad Universitaria, Costa Rica

ABSTRACT

The web building behavior of species of Anapis and Anapisona (Anapidae), Patu (Symphytogna-
thidae), and Maymena and Mysmena (Mysmenidae) is homologous with orb construction of other
araneoids. Possible behavioral synapomorphies linking these three families, and linking Anapidae with
Mysmenidae are proposed.

INTRODUCTION

The families Anapidae and Mysmenidae were recently separated from the
family Symphytognathidae (Forster and Platnick 1977) on the basis of
morphological differences. These authors speculated that web building behavior
would provide “some of the best clues” to resolve the present difficulties in
understanding the relationships between these and other families formerly
grouped in Symphytognathidae. Recent work with other orb-weaving spiders has
shown that some details of web building behavior are indeed useful in indicating
relationships (Eberhard 1982, Coddington 1986a and b, in prep.), and this paper
is an attempt to use direct observations of building behavior by six species and
indirect evidence from finished webs of at least seven others as indicators of the
relationships between these three families. The observations, which are generally
in agreement with the fragmentary notes of previous authors, are also compared
with those of orb-weavers in the families Araneidae and Theridiosomatidae.

Anapid webs and behavior are better known than those of the other two
families. Horizontal orb webs are built by species in the genera Anapis,
Anapisona (Platnick and Shadab 1978, Eberhard 1981, Coddington 1986a),
Chasmocephalon (Forster 1959, Coddington 1986a), Conoculus (Shinkai and
Takano 1984), Risdonius (Hickman 1938), and probably Pseudanapis (Forster
and Forster 1973). In the first three genera there are one or a few radii above
the plane of the orb, and a few sticky lines are attached to them. Several details
of the construction behavior of Anapis and Anapisona are given in Eberhard
1981 and 1982, and Coddington (1986a) noted that both horizontal and non-
horizontal radii of Chasmocephalon shantzii Gertsch are laid before the sticky
line is produced. Some anapids’ webs have “supplementary” radii that differ from
the other radii both in being thinner (Hickman 1938) and in the angles they make

340

THE JOURNAL OF ARACHNOLOGY

with Sticky lines (e.g., Coddington 1986a), but the construction behavior
associated with these lines has not been carefully described.

In Symphytognathidae at least four species of Patu plus an unidentified Puerto
Rican genus are known to construct fine-meshed horizontal orbs (Marples 1955,
Forster 1959, Forster and Platnick 1977, Coddington 1986a), while Symphytog-
natha globosa Hickman makes a “web ... of a few irregular threads in a more
or less horizontal plane . . . the web seems to be made on the under-surface of
the stones . . . The threads do not appear to be adhesive.” (V. V. Hickman in
Forster and Platnick 1977). There are no direct observations of the spiders’
building behavior.

In Mysmenidae Marples (1955) noted that the webs of Mysmena {=Tamesesia)
rotunda Marples and M. acuminata Marples consist “of a set of threads radiating
in all directions from a centre . . . The space between the radials is filled with
threads of a sticky silk, so fine that the droplets can only be seen under the
microscope . . . the general impression is of an orb web in three dimensions.”
Photographs of webs of M. jobi (Kraus) (Shinkai 1977), M. guttata Bishop and
Crosby and M. sp. in Coddington 1986a conform to this description. The web
of Maymena ambita (Barrows) however resembles the orbs of anapids
(Coddington 1986a). Marples (1955) observed that when M. acuminata spins
sticky lines “the spider keeps going quickly out along different radials. Apparently
it attaches a thread to a radial and carries the other end to the centre and out
along another radial to attach it there. The web is built from the periphery
inwards.” Coddington (1986a) also found that radial and frame lines were laid
before sticky lines in M. sp., and that in the center of M. guttata webs some radii
ended on other radii rather than all ending at the central point or system of lines
(hub).

MATERIALS AND METHODS

The spiders of this study are all minute, the smallest, Mysmena sp., being less
than 0.5mm long at maturity. The lines they spin are thus normally invisible to
the naked eye except under unusually favorable light conditions. It was
nevertheless possible to acquire a relatively detailed picture of web construction
by combining several techniques. Paths of moving spiders were followed closely
and used as indicators of the positions of lines already in place. A strong
headlamp was used to observe lines at favorable angles to the light. Those angles
between lines that are indicative of tensions on the lines were watched especially
closely to deduce whether the spider broke certain lines as it moved (see Eberhard
1981 for a description of this technique). The entire set of lines which a spider
had laid was made visible periodically by breathing gently on the web. Due to
the very humid microhabitats in which the webs were built (tropical rain forest
leaf litter), this caused tiny drops of water to condense on all the lines, thus
making the entire web easily visible (this technique was discovered independently
by J. Carico in humid forests in New Zealand). The drops soon evaporated,
leaving the web undamaged. The spider was generally disturbed briefly by being
blown on, but in all but one case it resumed building, sometimes so soon
afterward that the droplets were still on the lines, allowing for still more detailed
observations.

EBERHARD— WEB-BUILDING BEHAVIOR

341

I assumed, based on studies of larger and more easily observed species (e.g.,
Jacobi-Kleeman 1953, Eberhard 1982) that many details of web construction are
highly stereotyped; thus when I succeeded in seeing a given detail of behavior
clearly, I assumed that similar behaviors were executed in homologous situations
which were less easily observed^

Finally, both partially built and finished webs were collected on microscope
slides and examined with a compound microscope. It was found that when the
edges of the slide were wetted just before the slide was pressed against the web,
the positions of web lines were usually only slightly distorted by the process of
collection.

Some of the species mentioned here are identified only to genus due to present
incomplete taxonomic knowledge of the groups. Voucher specimens are deposited
in the Museum of Comparative Zoology, Cambridge, Mass. 02138 U.S.A. The
numbers and letters after names refer to labels in the vials with the specimens.
Webs were photographed after being lightly coated with cornstarch or talcum
powder.

The behavior reported here was compared to that of other orb weavers with
respect to characteristics useful in the characterization of families and subfamilies
(Eberhard 1982). The italicized letters and numbers {Al, B2, etc.) used below
designate character states used in that study.

There is some inconsistency in the literature regarding terms that designate
radii made before and after sticky spiral construction. Szlep (1961) used the terms
“ordinary” and “additional”; Eberhard (1977) spoke of “original” and “supple-
mentary” radii, and Coddington (1986a) termed them “structural” and “accessory”
radii. Although the lines may have evolved independently in Uloboridae and the
araneoids, they are geometrically and probably functionally analogous, so it
seems desirable to standardize terminology. Some previous terms have imprecise
or misleading connotations: ordinary and original imply character transforma-
tions which have not been established; and radii which are not “structural” are
by implication not part of the structure. An additional problem is that some
obvious alternative names (primary, secondary) have already been used to
distinguish different radii of the first type that are laid at different stages of
radius construction (e.g., LeGuelte 1966). I thus propose that radii laid before
sticky spiral construction commences be called “elementary” radii, and those after
the sticky spiral is complete “supplementary” radii. These terms are used in the
descriptions that follow.

Throughout this paper “sticky” lines are those which have small balls of liquid
distributed along their lengths, and “non-sticky” lines are those which lack such
balls.

RESULTS

Anapidae.— Webs: The webs of the individuals whose building behavior was
observed, Anapisona simoni Gertsch (No. 2166), Anapis calima Platnick and
Shadab (SJ1-39-A1), and Anapis sp. (SJ1-69-K), shared several characteristics.
They were horizontal orbs with one or a few non-sticky lines running upward
from the hub and variable numbers of sticky lines attached to these lines (Figs.
1-3). In the web of Anapis calima only about 10-20 elementary radii were

342

THE JOURNAL OF ARACHNOLOGY

Fig. 1 — Web of Anapis anchicaya Platnick and Shadab with a single non-sticky line above the hub
that has no sticky lines attached to it. Photograph is 7.4 cm wide.

fastened together at the hub, and the supplementary radii were either attached
to the elementary radii or ended on sticky lines near the hub (Fig. 4).

Anapidae.— Building Behavior: Frame construction, which was observed only
in Anapisona simoni, was similar to that of araneids and theridiosomatids
(Savory 1952, pers. obs.) both with respect to the sequence of attachments made

Fig. 2 — Web of Anapis heredia Platnick and Shadab with a single non-sticky line running upward
from the hub. The slanting lines attached to this line are excursions of the sticky spiral (see text).
The large object just above the spider at the hub is an egg sac. Photograph is 4.7 cm wide.

EBERHARD— WEB-BUILDING BEHAVIOR

343

Fig. 3 — Web of Anapisona simoni Gertsch with multiple non-sticky lines running upward from the
hub. Numerous sticky lines are attached to these lines. Photograph is 15.6 cm wide.

to the radii bordering the sector that was spanned by the new frame line, and
the fact that a single new radius was laid in the process of each frame
construction. Construction of the rest of the elementary radii, also observed only

Fig. 4 — Web of Anapis calima Platnick and Shadab. Many supplementary radii are broken near
the hub. A single line runs upward from the hub to an egg sac, and pulls the hub slightly above
the plane of the rest of the orb. Photograph is 10.0 cm wide.

344

THE JOURNAL OF ARACHNOLOGY

in A. simoni, continued after the frames were complete, and was similar to typical
araneid and theridiosomatid behavior (character state FI of Eberhard 1982) in
that a single radius was laid for each trip away from the hub, and the line laid
on the way out from the hub was apparently broken and rolled up and replaced
by a new one laid on the trip back (breakage was deduced from observations of
angles between web lines and from observations of a small white mass of loose
silk at the center of the hub — in araneids similar masses accumulate during radius
construction as a result of broken radial lines being rolled up and left there —
Eberhard 1982). The construction behavior for radii out of the plane of the orb
was the same in A. simoni as that for horizontal radii, and they were laid
interspersed in the sequence of construction of horizontal radii. No hub loops
were laid by this species during radius construction.

Subsequent behavior of at least two species deviated from typical araneid
behavior. In Anapisona simoni radius construction was followed by hub
construction, but then, instead of laying a temporary spiral of non-sticky silk, the
spider immediately began laying sticky spiral. Eor each segment of sticky line it
moved all the way from the edge of the web in to the hub, then back out again
(character state H3 of Eberhard 1982). Anapis keyserlingi Gertsch also failed to
build a temporary spiral.

Sticky spiral construction began at the edge of the planar part of the web (not
verified but probably also true in Anapis calima, which was observed laying only
the last five, innermost loops of sticky spiral), and the spider gradually worked
inward as do all orb weavers. In all species the spider faced away from the hub
as it moved out a radius prior to attaching (character state A1 or A4 of Eberhard
1982), then turned briskly 180° to attach near the innermost sticky line already
in place. I could not be certain in Anapisona simoni and Anapis calima whether
the spider touched the inner sticky line before turning; it apprarently did touch
it in Anapis keyserlingi (SJ1-69-K).

Sticky lines were laid to the radii above the plane of the web during sticky
spiral construction in both Anapisona simoni and Anapis keyserlingi (SJ1-69-K)
and were actually continuous with the planar sticky spiral. The spider attached
the sticky line to a horizontal radius, went to the hub as usual, but then climbed
one of the radii above the hub rather than going back out the next horizontal
radius. It walked along this radius for approximately the same distance it had
walked inward toward the hub, then attached the sticky line. From here it
returned to the hub, and either walked out another horizontal radius (not the one
it had started from) and attached the sticky line there, thus producing an inverted
“V” of sticky line above the orb, or else it went back up another radius above
the web plane, attached the sticky line there to form a more or less horizontal
segment of sticky line above the web plane, and then returned to the hub and
went out a horizontal radius (e.g. Fig. 3). In Anapisona simoni sticky lines were
laid above the plane of the orb only during the early part of sticky spiral
construction, so all sticky lines above the orb were attached near the orb’s
periphery. Photographs of Anapis heredia Platnick and Shadab webs (Fig. 2,
Coddington, 1986a) suggest the same is true for that species.

When the sticky line was all in place, the spiders performed several behaviors
which have not been described in any other orb-weavers. Anapis calima laid
another series of radial lines (supplementary radii). Each line was laid by walking
out a pre-existing radius, moving along the frame, attaching the drag-line to the
frame, and then returning to the hub and apparently attaching the line there or

EBERHARD— WEB-BUILDING BEHAVIOR

345

near there. One Anapis keyserlingi Gertsch (No. 2166) made only a single
supplementary radius; one Anapisona simoni did not make any (Fig. 3) while
another, figured in Coddington (1986a), made only a few. There was a clear
tendency to place consecutive supplementary radii on nearly opposite sides of the
hub. No obvious hub lines were laid during the intervals between trips to lay new
supplementary radii. The total number of supplementary radii in one web of
Anapis calima was about 48.

The path traced by the spider as it returned to the hub during supplementary
radius construction indicated that the radial line laid on the trip away from the
hub was broken near the frame and was replaced with another as the spider
moved back to the hub. The spider descended far below the web plane, and
climbed nearly directly upward as it arrived at the hub. The placement of
supplementary radii resulted in a characteristic radius-sticky spiral pattern in
finished webs in which the sticky spiral changed direction whenever it crossed an
elementary radius, but crossed supplementary radii without deviating (Figs. 1, 2
1 1). No supplementary radii were laid out of the plane of the orb.

Finally, in all species (except perhaps Anapis calima — I was not able to observe
it well enough to be certain) the spider performed another unique behavior
somewhat similar to hub destruction by theridiosomatids (character state G4 of
Eberhard 1982). Starting at the hub, the spider moved a few steps out an
elementary radius, turned 180° to face toward the hub, and broke the radius. The
spider evidently attached its drag-line near the outermost broken end and then
payed out a length of silk, as the sticky spirals attached to that radius moved away
from the hub. The spider returned to the hub as it slackened the radius or just
afterward, reeling up the inner broken end of the radius and replacing it with
the drag-line which it then attached at the hub. Since at least some supplementary
radii were broken at or near the inner edge of the sticky spiral in finished webs
(see Figs. 9 and 11), these lines were probably also broken during the loosening
process, but I was unable to resolve this detail. Coddington (pers. comm.) found
this to be the case in heredia and Anapisona simoni. Consecutive loosening
operations tended to occur on more or less opposite sides of the web. Loosening
operations on radii above the orb, which caused the orb to become less conical
and more nearly planar, were interspersed with those on the others.

The loosening process continued until most or all of the elementary radii were
lengthened; one Anapisona simoni did not loosen all of the radii, and microscopic
inspection of two Anapis keyserlingi webs (SJ1-69-H, SJ1-69-K) showed that 15
of 18 and 11 of 11 elementary radii had been broken near the hub and
presumably loosened.

Finally the spider finished the web by laying two or three tight hub loops
connecting the new inner ends of the planar elementary radii (determined by
direct observation in Anapisona simonv, in the other two species I was not certain
that lines were laid as the spider turned slowly at the hub, but microscopic
examination of two finished Anapis keyserlingi webs (No. 2166, SJ1-69-K)
revealed two to three hub loops — Fig. 5). As it turned while making the hub, the
spider removed the mass of rolled-up lines at the center of the hub and
apparently ingested it (the white speck disappeared). The hub spiral caused sharp
deflections of the radii where it was attached to them (Fig. 5), indicating that the
hub line was relatively tight. Finally the spider assumed its waiting position at
the hub. In none of the species observed did it flex the web, and when disturbed

346

THE JOURNAL OF ARACHNOLOGY

it climbed one of the radii above the orb. In contrast to many theridiosomatids,
the spiders did not destroy any of their webs as they moved.

Symphytognathidae.— Webs: The web of the only symphytognathid observed,
Patu sp. (No. 2194), was identical to the description of anapid webs with
supplementary radii above except there were no lines above the plane of the orb.

Symphytognathidae.— Building Behavior: The presence of a white speck at the
center of the hub in the web of Patu sp. after radius construction ended suggested
that radius construction behavior was similar to that described above (character
state FI of Eberhard 1982). After spinning about five loops of temporary spiral
which spiralled outward from near the hub (character state HI of Eberhard 1982)
the spider began spinning sticky spiral, starting from the edge and working
inward. It faced away from the hub as it moved out each radius preparatory to
attaching (apparently ^7 of Eberhard 1982), then turned 180° to attach. It did
not maintain contact with the temporary spiral as it laid sticky spiral (Z)7 of
Eberhard 1982). When the sticky spiral was finished, the spider immediately
began laying supplementary radii as described above except that it did not “sag”
as dramatically below the orb plane as it returned to the hub. I was thus not
certain whether the new supplementary radial lines were broken and rolled up as
the spider moved back to the bub, but the white speck of loose silk at the hub
did seem to grow. Finally the spider loosened the elementary radii as described
above for anapids, and then turned slowly about three times at the hub and the
white speck at the center disappeared (probably G4 of Eberhard 1982).

Mysmenidae.— Webs: Two web forms were observed in this family. The webs
of Mysmena sp. or spp. (Nos. 1034, 1122, 1173, 1174, 1175, 1188, 1366, 1679,
and 2195) were similar to the Mysmena webs described and figured by Marples
(1955) and Coddington (1986a). They consisted of about 20-30 non-sticky lines
radiating in three dimensions from a central area or hub, with many sticky lines
attached to them (Fig. 6). Some of the radii were attached directly to a support
(leaf, twig, etc.), but most ended on short frame lines which were in turn attached
to supports. Only two to four radii were attached to a single frame line.
Microscopic examinations of webs were complicated by the webs’ three
dimensional structure, but in two cases (SJ1-69-M and SJl-C) a clear hub was
found. In both cases many radial lines did not extend all the way to the hub but
terminated on other radial lines. Not all of the radial lines near the hub of the
web of SJ1-69-M were of equal thickness.

The cloud of fine, slack sticky lines was attached to the radii in the space
between the hub and the frames. These lines were so numerous and dense that
it was not possible to discern any pattern in their arrangement. Microscopic
examination showed that these sticky lines were thinner than the radial lines.

The webs of Mysmena sp. or spp. (Nos. 1608, 1687, 1689, SJ1-69-N, and SJl-
69-G) were different, being anapid-like horizontal orbs with a superstructure
above. The building behavior of Maymena will thus be described separately
below.

Mysmena.’— Building Behavior: Frame construction, which followed “explora-
tory” behavior, marked the discernable beginning of web construction, and was
identical to that seen in most araneoids (above). Observations of changes in
radius-frame angles during frame construction indicated that the radial line laid
on the trip away from the hub was broken near the new frame and replaced as
the spider returned to the hub. Subsequent elementary radii were also laid with
typical araneid behavior (FI of Eberhard 1982). The formation of a white speck

EBERHARD-WEB-BUILDING BEHAVIOR 347

T?

Fig. 5 — Drawing of the hub of a finished web of
Anapis keyserlingi Gertsch as seen on a slide
under a compound microscope. Those lines
marked with black dots had apparently been
broken and repaired during web construction,
^ since near the innermost edge of the sticky spiral
each one had a mass of material (like that
indicated by the arrow) which was probably
pyriform silk used to attach lines together. The
three lines marked with triangles were probably
above the plane of the web; sticky lines were not
attached to them in the central area of the web.

at the hub during radius and frame construction supported this conclusion, and
microscopic examination of two webs collected at the end of this stage also
revealed a tangled mass of lines at the hub which apparently corresponded to this
white speck. No hub lines were laid during radius construction (direct
observations were confirmed by microscopic examinations of the two unfinished
webs). The spider “loosened” a few radii, moving a few body lengths away from
the hub and then turning back and returning to the hub as do some other
araneids (Eberhard 1981). The first radii were not concentrated in a single plane;
I could not discern any pattern in the order in which radii were added.
Coddington (1986a) also saw Mysmena build radii and frame lines before starting
the sticky spiral.

After all or nearly all of the radii and frames had been laid, the spider began
laying sticky lines without first making a temporary spiral {H3 of Eberhard 1982).
Coddington (1986a) also noted the absence of a temporary spiral in a Mysmena
web. The spider moved quickly to the end or near the end of a radius, attached
a sticky line there, then went quickly back to the hub and out another radius
to attach again (Fig. 7). As it moved out a radius it faced away from the hub,
but I was not sure if it made contact with the innermost sticky line already
attached there {Al or A4 of Eberhard 1982). Coddington (1986a) thought some
attachments were made without contacting the inner loop. One spider which
moved more slowly clearly paused just before each attachment to make several
quick pulling movements with alternate strokes of legs IV (Eberhard 1981 gives
evidence that such movements in other species result in more sticky line being
pulled from the spinnerets). This same spider “pushed” the new sticky line away
from its body with one leg IV just before the attachment was made {Cl of

348

THE JOURNAL OF ARACHNOLOGY

Fig. 6 — “Three dimensional orb” of Mysmena sp. (No. 1188). The spider is at the point of
convergence of lines in the center of the web. Photograph is 4.3 cm wide.

Eberhard 1982). Two spiders made most or all of the early attachments to radii
above the hub.

Each successive sticky line was attached to the hub side of the point where the
previous innermost sticky spiral segment was attached. The distances between
successive attachments varied, the largest being more than twice the smallest.

The spider probably slid one tarsal claw along the radius as it moved out from
the hub since the tiny water drops which were sometimes present on the radius
were removed when the spider passed, and a single apparently larger drop was
left just inside the point where the sticky line was attached to the radius.

The estimated angles between radii to which spiders made successive
attachments of sticky lines were recorded in two webs. Although the data are
somewhat suspect since angles were estimated rather than measured and I was
not confident that all of my estimates were accurate, it appears that spiders chose
radii which made angles of less than 90° more often than expected by chance.
Of 243 pairs of attachments, 1 1 1 were to radii which made angles of less than
90°, and 76 were to radii which made angles of greater than 90° (56 others made
angles close to 90°) (Chi Squared =6.55, /7<0.01 assuming equal numbers of radii
at less than and more than 90°. If such a tendency exists, it would explain why
sticky lines are at least sometimes concentrated near the periphery of the web
(Fig. 6, Coddington 1986a) since a sticky line laid from one radius to another
180° opposite would run near the hub.

One spider which was laying sticky spiral was disturbed when I blew on and
apparently damaged its web. The spider remained immobile for several minutes,
then laid several radial lines before returning to sticky spiral construction.

When the sticky spiral was complete, the spider loosened some of the radii near
the hub (perhaps all in some cases) in the manner described above for anapids.
That the radii were indeed broken and lengthened in this process was confirmed
in one case by noting that a small white spot on the radius moved away from

EBERHARD— WEB=BUILD1NG BEHAVIOR

349

Fig. 7— Drawing of Mysmena sp. early in the process of sticky spiral construction. Straight lines
(solid and dashed) are the non-sticky radii and frame lines. The dotted, curved lines are the sticky
spiral. The spider travels all the way in to the hub and then back out (large arrows) between one
attachment of the sticky spiral and the next (small arrows).

the hub as the spider moved back toward the hub. In several other cases I saw
that the spider made quick alternate pulling movements with its legs IV
apparently on the dragdine as it neared the hub, probably pulling additional silk
as it did so. In one case the loosening behavior was at first confined to radii in
the lower half of the web, and some upward-directed radii were not loosened. In
finished webs that were examined microscopically, 5 of 23 radii in the web of
one (SJ1-69-M) and 14 of the 19 radii which could be followed in the web of
another (Fig. 8) had “pyriform masses” on them similar to those on the
elementary radii of anapids and Maymena (below) that had been broken and
lengthened.

When the loosening was completed, the spider slowly turned about 360° at the
hub (3 of 3 cases), making movements which may have involved attaching the
radii there together. Two completed webs on slides had a single loop of hub line
(Fig. 8). As the spider turned, its anterior end was on the white speck at the
center of the hub, and when it finished turning the speck was gone, presumably

350

THE JOURNAL OF ARACHNOLOGY

ingested (apparently G4 of Eberhard 1982). One apparently finished web however
had a tangled mass of threads of the hub which probably corresponded to the
white speck. Most radial lines did not extend all the way to the hub, but instead
formed a branching pattern (Fig. 8) similar to that of some theridiosomatids
(McCook 1889). Coddington (1986a) saw a similar pattern of radii in the web
of M. guttata.

The spider rested at the hub of the finished web. It did not tighten the web
as it waited, and when disturbed it moved out along one of the radii without
noticeably damaging the radii or frame lines.

Building Behavior: The building behavior of Maymena sp. (SJl-
69-D) and No. 2168) resembled that of the anapids described above in most
respects. A white speck present at the center of the hub during sticky spiral
construction suggested the same type of radius construction behavior {FI of
Eberhard 1982). Several loops of temporary spiral were laid spiralling away from
the hub {HI of Eberhard 1982). Miscroscopic examination of the web of SJl-
AA showed that the temporary spiral lines (broken in the finished web) were
made of lines of at least approximately the same thickness as those of the
elementary radii.

The spiders moved extraordinarily rapidly during sticky spiral construction and
were thus difficult to observe, but it was clear that they faced away from the hub
as they moved out radii to attach the sticky spiral and came close to touching
the inner loop of sticky spiral already in place {A1 of Eberhard 1982). They did
not maintain contact with the temporary spiral as they worked near the edge of
the orb {D1 of Eberhard 1982). It appeared that one leg IV pushed the sticky
spiral just as it was attached to a radius {Cl of Eberhard 1982), but I could not
be certain. The spider attached sticky line to the radii above the orb during sticky
spiral construction as in anapids. In one case (No. 2168) several such lines were
laid early during sticky spiral construction, then none were laid until the sticky
spiral was between two thirds and three quarters complete, and then 10-15 more
were laid in quick succession. Since these webs had temporary spirals, the spiders
must have walked to the hub on top of rather than beneath the radii, and
somehow they kept the sticky spiral free of entanglement with the radius. I was
not able to see how these feats were accomplished.

After finishing the sticky spiral, the spider paused at the hub for 10-30 seconds,
then began to lay supplementary radii. The spider clearly sagged far below the
plane of the orb as it returned to the hub, and thus apparently broke the line
it had laid as it moved away from the hub. In one case (SJ1-69-D) it was clear
that successive supplementary radii tended to be laid on nearly opposite sides of
the web, and also that the spider made a series of quick, alternating pulling
movements with its legs IV as it reached the hub, suggesting that additional line
was being pulled from the spinnerets. Microscopic examinations of the webs of
SJ1-69-D, SJ 1-69- A A, and SJ1-69-G (Fig. 9) showed that the supplementary radii
were much thinner than elementary radii. They could only barely be distinguished
at 450X, while elementary radii were distinct even at 200X. These examinations
also confirmed that only elementary radii were attached to the hub, and that the
sticky spiral lines were attached only to elementary radii. No supplementary radii
were laid out of the plane of the orb.

Spider No. 2168 was “distracted” by capturing and eating a series of three prey
as it laid supplementary radii, and no further behavior was observed, but SJl-
69-D followed supplementary radius construction by cutting each elementary

EBERHARD- WEB-BUILDING BEHAVIOR

351

t

Fig. 8 — Drawing of lines in the center of a finished three dimensional orb of Mysmena sp. (SJl-
C) which was pressed onto a flat glass slide and observed under a compound microscope. The lines
with a black dot had pyriform masses on them farther from the hub, indicating that they had been
broken and repaired during construction. Only 19 of the 24 radial lines could be followed far from
the hub, and others may have also had pyriform masses. Lines on which others converged in the
central area were thicker.

radius near the hub, lengthening it, and then reattaching it. At least some
supplementary radii were also cut while elementary radii were being lengthened.
I also noted pulling movements with legs IV as the spider returned to the hub
during this process. The accumulation of broken supplementary radii and sticky
spiral lines which had stuck together formed a “halo” of silk around the spider
that was visible even in unpowdered webs (Fig. 10). Examinations of webs under
the microscope confirmed both that all elementary radii were lengthened and
reattached at the hub, and that all supplementary radii ended on the sticky spiral
or an elementary radius and none of them reached the hub (Figs. 9, 1 1).

Finally the spider laid two loops of hub line around the accumulation of loose
silk at the center of the hub, and in the process removed this silk (apparently
G4 of Eberhard 1982). The apparently complete web of SJ1-69-D, however, had
a small mass of loose silk at the center of the hub.

Finished Webs of Other Species. — Some of the distinctive behaviors described
above result in webs whose designs are also distinctive, allowing one to deduce
details of building behavior from finished webs. Supplementary radii, distin-
guished by lack of deflection of sticky spiral lines crossing them, greater sag
under the weight of cornstarch used for photography, and failure to pull frame
lines to which they are attached out of straight lines, occur in the webs of the

352

THE JOURNAL OF ARACHNOLOGY

Fig. 9 — Drawing of the central portion of the web of a Maymena sp. (No. SJ1-69-G) web that
was collected on a slide and observed under a compound microscope. Thin lines with dots represent
sticky lines. All elementary radii (thicker lines with larger dots) are connected to the hub, while
supplementary radii (thinner lines with dots) end on sticky lines or elementary radii. The 5:00 sector
of the web was evidently damaged during collection. The lines extending upward at 11:30 were
probably above the plane of the orb. Pyriform masses are indicated by dark spots on the elementary
radii (arrow).

anapids Anapis anchicaya Platnick and Shadab, A. felidia P. and S., A. herediae
P. and S., the mysmenid Maymena sp. (No. 1687), and the symphytognathid Patu
saladito Forster and Platnick. One web of Anapisona hamigera (Simon) clearly
did not have supplementary radii.

When a spider which has laid supplementary radii then loosens the elementary
radii, the pattern of threads near the hub is like that in Fig. 11, a pattern not
seen in the web of any orb weaver which does not perform this behavior. This
pattern is clear in photographs of the webs of all the species just mentioned.

EBERHARD WEB-BUILDING BEHAVIOR

353

Fig. 10 — Maymena (?) sp. (No. 1687) at the hub of its unpowdered web. The circular line (“halo”)
around the spider is the accumulation of sticky and non-sticky lines formed when the spider replaced
the hub after finishing supplementary radius construction (see Fig. 9). Photograph is about 3 cm wide.

DISCUSSION

Studies of web building behavior of other groups of orb-weavers have shown
that many details are quite conservative (Eberhard 1982), so although the
numbers of species of anapids, mysmenids and symphytognathids whose webs

Fig. 11 — Powdered web of Maymena (?) sp. (No. 1608). The white object above the hub is an egg
sac. Supplementary radii and the “halo” of lines around the hub are visible. Photograph is 5.6 cm
wide.

354

THE JOURNAL OF ARACHNOLOGY

and building behavior have been observed is still small, the fact that they share
apparently derived characters never seen in better studied groups is strong
evidence that they are closely related. The data are combined with those of
Eberhard (1982) and Coddington (1986a) and summarized in Fig. 12. Several
synapomorphies are suggested. The term “radius” is used in Fig. 12 only for lines
to which sticky lines are attached (thus excluding the “spring line” of some
theridiosomatids). The relationships are in agreement with those of Coddington
(1986a) based on other characters in addition to web form.

Several assumptions were made in constructing the cladogram. Convergent loss
of characters was considered more likely than convergent origin; thus both
ancestral possession of temporary spiral with covergent losses in anapids and
some mysmenids and ancestral possession of supplementary radii with subsequent
losses in some anapids and some mysmenids were preferred over alternative
possibilities. Loosening radii was taken to be derived with respect to absence of
this behavior since loosening involves an additional behavior. The decision to
attribute late construction of the definitive hub and the use of leg IV to hold the
radius to which the sticky spiral has just been attached away from the sticky
spiral line (Eberhard 1981) to the presumptive ancestor of the non-araneid
families was arbitrary, and early hub construction and not sliding leg IV could
equally well be considered derived characters of araneids, unless uloborids and/
or dinopids are taken as the sister group of araneoids. As argued in Eberhard
(1981), holding the radius away with leg IV is probably associated with relatively
small spider size in relation to the distances between lines in the orb. I was not
able to observe the mysmenids and symphytognathids in sufficient detail to
ascertain whether they perform this behavior. If the relationships in Fig. 12 are
correct, at least some of them probably do.

Loss of both temporary spiral and supplementary radii in Mysmena webs is
probably a consequence of the three dimensional orb design, since it is difficult
to imagine an effective three-dimensional temporary spiral, and it is probably
impossible to lay supplementary radii through a dense three-dimensional array of
sticky lines (the spider must move through them, however, when it captures prey).
The secondary loss of supplementary radii proposed for anapids is supported by
the observation of Anapis sp. (No. 2166) laying a single (vestigial?) supplementary
radius, and the low numbers of supplementary radii in the webs of Anapisona
simoni (Coddington 1986a) and Anapis atuncela Platnick and Shadab.

“Radial anastomosis” (convergence of some radii before they reach the hub)
was not included in the characters used to make the cladogram even though it
may be a useful taxonomic character (Coddington 1986a) because I was unable
to determine whether the occasional supplementary radius terminating on an
elementary radius near the hub (e.g. Fig. 7) was actually attached there or
whether it fell against the elementary radius as the spider broke lines while
loosening the elementary radii. It was clear that elementary radii did not
“anastomize” in the orbs of any of the species studied other than Mysmena .

Coddington (1986a) pointed out the similarity between supplementary radii of
symphytognathoids and the additional radii spun by some young uloborids (Szlep
1961, Eberhard 1977), and suggested that the possibility of homology should be
considered in more detail. Supplementary radii are apparently similar to the
uloborid lines in not being produced by the ampullate glands (at least they have
very different diameters from those of elementary radii and hub lines). They differ
however in being single lines placed on the web in radial directions rather than

EBERHARD— WEB-BUILDING BEHAVIOR

355

SOME WITHOUT WEBS

MYSMENIDAE

'''known in Theridiosomatidae and ^napidae; no certain observations
OF Symphytognathidae or I^Iysmenidae

Fig. 12 — Most likely system of relationships between different families based on behavioral
characters (see text for assumptions made in analysis). Proposed synapomorphies are underlined.

clouds of fine fibers that often do not run radially as in uloborids (Eberhard
1977). Homology is thus unlikely.

The building behavior of Mysmena sp., the radial design of its web, and the
orb web of the related Maymena clearly support the idea of Marples (1955) that
Mysmena webs are best considered as three-dimensional orbs. Radii and frame
lines are laid first, and involve behavior apparently identical to that of other
araneoid orb-weavers. The tendency to lay some radii out of the plane of the orb,
which also occurs in anapids and Maymena, is accentuated. As in all known orb-
weavers, the sticky line is laid from the edge of the web moving inward so that
each successive sticky spiral attachment to a radius is to the hub side of the last.
As with the sticky lines of anapids and Maymena which are attached to radii out
of the plane of the orb, all Mysmena sticky lines are very slack and sag under
their own weight. When the sticky spiral is finished the spider lowers the tension

356

THE JOURNAL OF ARACHNOLOGY

on most radii by breaking them and lengthening them just as do all known orb-
weaving anapids, Maymena, and the symphytognathid Patu. Finally, the
accumulation of rolled-up lines at the hub is eliminated, apparently by ingestion,
and the only circular hub lines which persist in the finished web are added after
the web is otherwise complete as in all anapids, those theridiosomatids with hubs
(Eberhard 1982, Coddington 1986a), and Maymena. Thus Mysmena behavior is
apparently homologous at many points with that of species making typical orbs.

ACKNOWLEDGEMENTS

I am grateful to Norman Platnick for kindly identifying many tiny specimens.
Jonathan Coddington and M. J. West-Eberhard read a preliminary version of the
manuscript. Financial support was provided by the Vicerrectoria de Investigacion
of the Universidad de Costa Rica.

LITERATURE CITED

Coddington, J. 1986a. The monophyletic origin of the orb web. In Spiders — Webs, Behavior, and
Evolution. (W. A. Shear ed.). Stanford Univ. Press.

Coddington, J. 1986b. Orb webs in “Non-orb weaving” ogre-faced spiders (Araneae: Dinopidae): a
question of genealogy. Cladistics, 2:53-67.

Eberhard, W. G. 1977. The webs of newly emerged Ulohorus diversus and of a male Uloborus sp.
(Araneae: Uloboridae). J. Arachnol., 4:201-206.

Eberhard, W. G. 1981. Construction behavior and the distributions of tensions in orb webs. Bull.
British Arachnol. Soc., 5:189-204.

Eberhard, W. G. 1982. Behavioral characters for the higher classification of orb-weaving spiders.
Evolution, 36:1067-1095.

Forster, R. R. 1959. The spiders of the family Symphytognathidae. Trans. Roy. Soc. New Zealand,
86:269-329.

Forster, R. R. and L. M. Forster, 1973. New Zealand Spiders. Collins, London.

Forster, R. R. and N. 1. Platnick. 1977. A review of the spider family Symphytognathidae (Arachnida:
Araneae). Amer. Mus. Novitates, 2619:1-29.

Hickman, V. V. 1938. On a dipneumone spider (Risdonius parvus, gen. et sp. n.), the female of which
has reduced palpi. Proc. Zool. Soc. London, 108:655-660.

Jacobi-Kleemann, R. 1953. Ueber die Locomotion der Kreuzspinne Aranea diademata beim Netzbau
(nach Filmanalysen). Z. vergl. Physiol., 34:606-654.

LeGuelte, L. 1966. Structure de la toile de Zygiella x-notata CL (Araignees, Argiopidae) et facteurs
que regissent le comportement de I’araignee pendant la construction de la toile. These Pub. Univ.
Nancy. 1-77.

Marples, B. J. 1955. Spiders from Western Samoa. J. Linnaean Soc. Zool., 42:453-504.

McCook, H. C. 1889. American Spiders and their Spinningwork. Vol. 1. published by the author,
Philadelphia.

Platnick, N. 1. and M. Shadab. 1978. A review of the spider genus Anapis (Araneae: Anapidae) with
a dual cladistic analysis. Amer. Mus. Novitates, 2663:1-23.

Shinkai, E. 1977. Spiders of Tokyo HI (in Japanese). Acta Arach., 27:321-336.

Shinkai, E. and S. Takano, 1984. A Field Guide to the Spiders of Japan. Tokai Univ. Press, Tokyo.
Szlep, R. 1961. Developmental changes in the web-spinning instinct of Uloboridae: construction of
the primary type web. Behaviour, 27:60-70.

Manuscript received October 1985, revised February 1986.

Raven, R. J. 1987. A new mygalomorph spider genus from Mexico (Nemesiinae, Nemesiidae,
Arachnida). J. Arachnol., 14:357-362.

A NEW MYGALOMORPH SPIDER GENUS
FROM MEXICO (NEMESIINAE, NEMESIIDAE, ARACHNIDA)

Robert J. Raven

Queensland Museum
P.O. Box 300, South Brisbane
Queensland, Australia

ABSTRACT

A new mygalomorph genus, Mexentypesa, and its type species Mexentypesa chiapas, new species,
are described from southern Mexico. The presence of claw tufts in association with biserially dentate
paired claws in both males and females allies the genus to both the theraphosids and nemesiids; its
sister group is suggested to be Calisoga.

INTRODUCTION

During a revision of mygalomorph genera previously placed in the Diplurinae,
I examined two new genera that presented a new combination of characters that
challenged my hypotheses of relationships of nemesiids and cyrtaucheniids (see
Raven 1985). The problem arose because both of those genera have claw tufts,
unlike most other co-familial taxa but like the theraphosoids. In mygalomorphs,
claw tufts are found in barychelids, theraphosids, one paratropidid genus, plus
the nemesiid Neodiplothele (see Raven 1985). That occurrence prompted me to
treat the presence of claw tufts as a synapomorphy of the Theraphosoidina
(Theraphosidae, Paratropididae, plus Barychelidae). Because the Nemesiidae is
the sister group of the Theraphosoidina, the synapomorphic nature of claw tufts
in the latter must be re-examined. The first of those genera is described here for
an apparently unknown species.

MATERIALS AND METHODS

All eye measurements are given in ocular eyepiece units with interspaces
expressed in diameters of an AME. All other measurements are in millimeters.
Abbreviations are standard for the Mygalomorphae, except possibly for MOQ,
median ocular quadrangle.

Claw tufts are each considered to be a dense cluster of hair arising from a
separate cuticular pad lying between the outer edges of the bases of each paired
claw and the tarsus. Hairs arising from the cuticle on the legs (as in some
Australian Aname species) are merely extensions of the scopulae. A pseudoseg-
mented tarsus is evident as a single ventral transverse weakness in the cuticle or
as a recticulated area of weak cuticle resembling cracking mud; the effect of either
condition is the downward flexion of the tarsus.

358

THE JOURNAL OF ARACHNOLOGY

Mexentypesa, new genus

Type. — The type species of this genus is Mexentypesa chiapas, new species. The
name is feminine and derives from a combination of Mexico and Entypesa, which
the genus resembles.

Diagnosis. — Females may be distinguished from barychelids, paratropidids, and
theraphosids by the biserially dentate paired claws, and from the barychelids
Monodontium and Troglothele and the nemesiid Spelocteniza by the presence of
well developed eyes and/or a digitiform apical segment to the posterior lateral
spinnerets. Males are unique in the combination of biserially dentate paired
claws, claw tufts, and all tarsi pseudosegmented. Both sexes are readily
distinguished from ail other nemesiids, save Neodiplothele (which has only two
spinnerets), by the presence of claw tufts and from the cyrtaucheniid by lacking
a third claw on any leg.

Description. — As for species.

Mexentypesa chiapas, new species
Fig. 1-9, Tables 1, 2

Types. — Male holotype, female paratype, from Ocosingo (altitude 900 m,
Chiapas, Mexico, June 25, 1950 (C. and N. Goodnight, L. J. Stannard);
deposited in AMNH.

Diagnosis. — As for genus.

Description. — Holotype male. Carapace, chelicerae, and legs reddish brown;
sternum, maxillae, and labium yellow brown; abdomen dorsally brown with white
mottling, ventrally with brown areas on pallid yellow, booklung covers orange.

Carapace 4.16 long, 2.72 wide; few black setae on caput; lateral margins with
strong setae for two thirds of carapace; caput flat, as high as thoracic region
behind fovea; fovea short, recurved, with posterior indentation.

Eight eyes in two rows on distinctly raised tubercle; ratio of
AME:ALE:PME:PLE, 10:9:6:8; from above, front row straight, back row
recurved. Group occupies 0.46 of headwidth, 1.88 times wider than long. MOQ
1.47 times wider behind than long, and 1.16 times wider behind than in front.
Eye interspaces: AME-AME, 0.3; AME-ALE, 0.1; ALE-PLE, 0.1; PME-PLE, 0;
PME-PME, 1.3.

Chelicerae short, slightly geniculate, with strong setae dorsally; rastellum of six
stout spines (each about four times longer than wide at the base) curving down
over fang. Fang smooth, slender. Intercheliceral face not modified; promargin
with four large and five small spaced teeth, basomesally with four small teeth.
Labium 0.25 long, 0.70 wide, without cuspules, two large sigilla in labiosternal
suture; with transverse ridge medially. Maxillae 1.13 long in front, 1.25 long
behind, 0.60 wide, anterior lobe smooth, distinct, without serrula; eight cuspules
on inner angle, not on mound; lyra absent. Sternum 2.04 long, 1.84 wide, domed
in cross-section; six sigilla oval, marginal, posterior pair 0.14 long.

Leg formula 41(2=3). Tarsi I-IV curved, pseudosegmented. Leg I incrassate;
tibia incrassate with distinct megaspine on retroventral edge, metatarsus with
distinct, proximal cuticular thorn retrolaterally. Preening combs absent. Scopulae:
light, distal on metatarsus I, light on tarsus I, very thin on metatarsus and tarsus

RAVEN— NEW MEXICAN NEMESIID MYGALOMORPH

359

Figs. 1-9 — Mexentypesa chiapas, new species: 1-7. Male holotype: I, cephalothorax, lateral view;
2, tibia, metatarsus, and tarsus I, retrolateral view; 3, cephalothorax, chelicerae, and abdomen, dorsal
view; 4, palpal tibia, cymbium, and bulb, prolateral view; 5, sternum, maxilla, and labium; 6,
abdomen and spinnerets, ventral view; 7, distal chelicera and fang, oblique anterior view. 8, 9.
Paratype female: 8, spermathecae, ventral view; 9, cephalothorax and chelicerae, dorsal view. Scale
lines = 4 mm for figs. I, 3; 1 mm for figs. 4, 7, 8; 2 mm for others.

II, only scattered scopuliform setae on tarsi III and IV, absent elsewhere. Spines
(no spines on tarsi): leg I, femur pi, d4, patella 0, tibia p2, v5 + megaspine,
metatarsus v2 + process; leg II, femur pi, d4, patella p2, tibia p2, v6, metatarsus
pi, v5; leg III, femur p3, d3, r4, patella p2, rl, tibia, p2, d3, r2, v6, metatarsus,
p6, dl, r4, v7; leg IV, femur p2, d3, rl, patella p2, rl, tibia p2, dl, r4, v6,
metatarsus p4, d2, r3, v8.

360

THE JOURNAL OF ARACHNOLOGY

Leg 1

Leg 2

Leg 3

Leg 4

Palp

Femur

2.75

2.50

2.31

2.94

1.69

Patella

1.75

1.63

1.31

1.75

1.00

Tibia

1.75

1.50

1.31

2.13

1.25

Metatarsus

1.69

1.50

2.13

3.19

—

Tarsus

1.13

1.06

1.13

1.38

0.63

Total

9.07

8.19

8.19

11.39

4.57

Palp (Fig. 4) with pyriform bulb; embolus lacks keels and tips of both appear
to have been broken; cymbium without scopula, divided into two similar lobes,
apically with four thick and four thinner spines; tibia incrassate. Spines: femur
pi, d3, patella pi, patella pi, tibia p2, v4, tarsus 8 apical.

Paired claws with two long, juxtaposed rows each of about nine teeth; unpaired
claws absent on all tarsi; on all legs, two small thin tufts on distinct cuticular
extension.

Ten to 15 trichobothria in row extending for full length of tibiae, curving row
of 8-10 on metatarsi, and 10-12 proximal of crack or pseudosegmentation. Tarsal
organ a low dome with 3-4 shallow central concentric ridges; cuticle surface
smooth; bothria corrugiform almost to base.

Abdomen 4.11 long, 2.89 wide; lungbook apertures broad, slit-like. Posterior
median spinnerets 0.43 long, 0.30 wide, 0.25 apart. Basal, middle, apical segments
of posterior lateral spinnerets 0.75, 0.70, 0.65 long, respectively; apical segment
digitiform.

Paratype female. Carapace orange brown with brown markings radiating along
strial edges, on lateral carapace, and caput; chelicerae orange brown; legs yellow
brown with brown annulations on distal and proximal metatarsi, annulations
lateral on tibiae and patellae, distolateral on femora; sternum, maxillae, and
labium yellow brown; abdomen similar to male with less white.

Carapace 3.16 long, 2.68 wide; setae weaker than in male; caput flat, as high
as thoracic region behind fovea; fovea short, straight.

Ratio of AME:ALE:PME:PLE, 7:10:8:10; from above, front row slightly
procurved, back row recurved. Group occupies 0.47 of headwidth, 1.94 times
wider than long, slightly wider behind than in front. MOQ 1.53 times wider
behind than long, 1.35 times wider behind than in front. Eye interspaces: AME-
AME, 0.4; AME-ALE, 0.3; ALE-PLE, 0.3; PME-PLE, 0.1; PME-PME, 2.0.

Chelicerae similar to male but rastellum entirely absent; promargin with seven
spaced teeth, basomesally with few granules. Labium 0.18 long, 0.64 wide,
without cuspules. Maxillae 0.96 long in front, 1.16 long behind, 0.60 wide,
anterior lobe less distinct than in male; six cuspules on inner angle. Sternum 1.54
long and wide; all sigilla oval, marginal.

Leg formula 41(2=3). Only tarsi IV distinctly cracked. Preening combs absent.
Scopulae: very thin to absent on metatarsi I, II, almost absent on tarsi I, II,
absent elsewhere. Spines (no spines on tarsi, femora I-IV with one long dorsal
basal spine): leg I, femur d4, patella 0, tibia 0, metatarsus v3; leg II, femur d4,
patella 0, tibia 0, metatarsus v4; leg III, femur dl, rl, patella p2, rl, tibia p2, dl,
rl, v3, metatarsus p3, d3, r3, v7; leg IV, femur dl, rl, patella pi, tibia p2, dl, r3,
v4, metatarsus p3, d3, r3, v8.

RAVEN-NEW MEXICAN NEMESIID MYGALOMORPH

361

Leg 1

Leg 2

Leg 3

Leg 4

Palp

Femur

2.20

1.92

1.88

2.00

1.44

Patella

1.48

1.36

1.12

1.44

0.96

Tibia

1.40

1.16

1.04

1.72

0.96

Metatarsus

1.08

1.12

1.52

2.56

—

Tarsus

0.76

0.80

0.80

1.08

0.96

Total

6.92

6.36

6.36

8.80

4.32

Palp with thin scopulae on tarsi and four spines on ventral tibia.

Claws, tufts, tarsal organ, and trichobothria similar to male.

Abdomen 4.08 long, 2.68 wide; lungbook apertures broad, oval. Posterior
median spinnerets 0.32 long, 0.14 wide, 0.26 apart. Basal, middle, apical segments
of posterior lateral spinnerets 0.50, 0.40, 0.38 long, respectively; apical segment
digitiform. Spermathecae two (Fig. 8), each a short lobe divided for apical one
half into two dissimilar ectally directed lobes.

Distribution. — Known only from the type locality in Chiapas, Mexico.

RELATIONSHIPS

Mexentypesa lacks the keels on the palpal bulb and intercheliceral tumescence
that would otherwise qualify it for membership in the Pycnothelinae (see Raven
1985). It shares with the Californian genus Calisoga (Nemesiinae) the
pseudosegmented tarsi of males and the digitiform apical segment of the posterior
lateral spinnerets. If those two genera are considered sister groups, an hypothesis
is required to explain the degree of difference in scopulae development: very
dense in Calisoga but almost nil in Mexentypesa. In the current sister group of
Calisoga, Nemesia plus Brachythele, the scopulae are also weak but generally not
so weak as in Mexentypesa. Most parsimoniously, the presence of scopulae is
considered an autapomorphy of Calisoga (congruent with the clavate setae on the
upper inner faces of the chelicerae), and the weak scopulation of Mexentypesa
is either homologous to the condition in Nemesia plus Brachythele or simply a
species autapomorphy. Hence, the sister group of Mexentypesa is considered to
be Calisoga.

The inclusion of Mexentypesa in the Theraphosoidina is rejected for three
reasons. First, the inclusion of Mexentypesa in the Nemesiidae requires only the
acquisition of claw tufts, thus incurring one additional step in the family
cladogram (Raven 1985, Fig. 1). Second, its inclusion in the Theraphosoidina
requires several homoplasies (not including those also required in the
Nemesiidae): biserially dentate paired claws in females; absence of theraphosoid
coupling spurs; long distinct maxillary lobe; numerous maxillary and labial
cuspules; and pseudosegmented tarsi in males. Third, its association with the
barychelids with biserial dentition of the claws in females {Troglothele,
Monodontium) would require the proposal of other homoplasies, e.g., digitiform
apical segment of the posterior lateral spinnerets, weak leg scopulae, plesiomor-
phic eye group configuration. However, males of those barychelid genera are not
known; hence, their relationships are in doubt and no further discussion is
warranted. In summary, Mexentypesa is currently accommodated in the
Nemesiidae with the minimum of homoplasy.

362

THE JOURNAL OF ARACHNOLOGY

ACKNOWLEDGEMENTS

I am grateful to Dr. Norman 1. Platnick, Department of Entomology,
American Museum of Natural History, New York, for allowing me access to the
collection in his care, and to Dr. Frederick A. Coyle, Western Carolina
University, Cullowhee, who also reviewed the manuscript.

LITERATURE CITED

Raven, R. J. 1985. The spider infraorder Mygalomorphae (Araneae): cladistics and systematics. Bull.
Amer. Mus. Nat. Hist., 181:1-180.

Manuscript received August 1985, revised March 1986.

Wharton, R. A. 1987. Biology of the diurnal Metasolpuga picta (Kraepelin) (Solifugae, Solpugidae)
compared with that of nocturnal species. J. Arachnol., 14:363-383.

BIOLOGY OF THE XI. METASOLPUGA PICTA

(KRAEPELIN) (SOLIFUGAE, SOLPUGIDAE)
COMPARED WITH THAT OF NOCTURNAL SPECIES^

Robert A. Wharton

Desert Ecological Research Unit
RO. Box 953
Walvis Bay 9190

• 2

South West Africa/ Namibia

ABSTRACT

Metasolpuga picta (Kraepelin) is a diurnal solifuge restricted to the Namib Desert of S'WAj
Namibia. Burrowing, mating, oviposition, and feeding behaviors are described and compared with
other solifuge species. Thermoregulatory behavior, previously undocumented for solifuges, is
associated with diurnal habits. Populations were bivoltine in 1979, whereas all solifuges previously
studied have been univoltine. The biology of M. picta is otherwise similar to that of nocturnal species
from North African, South-west Asian, and North American deserts.

INTRODUCTION

Despite extensive work on some North American species (Muma 1966a-e,
1967), solifuges are relatively poorly studied (Cloudsley-Thompson 1977, Savory
1977). Since the publication of Cloudsley-Thompson’s (1977) review, Muma
(1974a,b, 1975a,b, 1979, 1980a,b), Gore and Cushing (1980), Thaler (1982), and
Aliev and Gadzhiev (1983) have published additional biological information; and
Aruchami and Sundara Rajulu (1978), Alberti (1979, 1980), Haupt (1982), and
Bauchhenss (1983) have contributed morphological studies. Nevertheless, detailed
biological data have been published for only four of the approximately 1000
described species of solifuge arachnids (Heymons 1902, Cloudsley-Thompson
1961a,b, Junqua 1966, and Muma 1966c). Published biological data on Solifugae
are restricted primarily to members of the families Galeodidae, Eremobatidae,
and Ammotrechidae. Little or nothing is known about the other nine solifuge
families.

Cloudsley-Thompson (1977) stated that solifuges are nocturnal, with only “a
few small colourful species” exhibiting diurnal habits. Previous biological studies
have dealt almost exclusively with these nocturnal species. However, there are

'Approved as TA No. 18797 by the Texas Agricultural Experiment Station.

^Present address: Department of Entomology, Texas A&M University, College Station, Texas 77843,
U.S.A.

364

THE JOURNAL OF ARACHNOLOGY

several large, diurnal solifuge species in the arid regions of southern Africa
(Lawrence 1963, Wharton 1981).

Based on trap catches of 14 individuals. Holm and Scholtz (1980) determined
that Metasolpuga picta (Kraepelin) was diurnal, and restricted to the interdune
valley habitat of the Namib Desert dune ecosystem at Gobabeb (South West
Africa/ Namibia). Its biology is detailed here for comparison with its better
known nocturnal relatives. Prior to Holm’s studies of dune arthropods (Holm
and Edney 1973, Holm and Scholtz 1980), only five specimens of M. picta had
ever been collected (Kraepelin 1899, Purcell 1899, Lawrence 1963, 1965, 1967).
The following account represents the first detailed biological study of a member
of the family Solpugidae.

MATERIALS AND METHODS

Solifuge behavior was observed between November, 1978 and February, 1980
within a 5 km radius of Gobabeb (23°24'S, 15°03'E), in the Namib Desert of
South West Africa/ Namibia. This region consists of a southern dune habitat and
a northern gravel plain bisected by the sporadically flowing Kuiseb River. Holm
and Edney (1973), Holm and Scholtz (1980), Robinson and Seely (1980), Seely
and Louw (1980), and Wharton and Seely (1982) have described and figured the
dune and gravel plains habitats in this region. Nearly all observations were made
directly in the field, on both sides of the Kuiseb River. Eleven individuals were
also maintained for varying periods in laboratory cages to supplement field data
on oviposition, feeding, response to temperature, and intraspecific aggression.

Although most solifuges, including those from the Namib Desert, readily climb
out of pitfall traps, M. picta was unsuccessful in doing so unless the sides of the
traps were exceptionally rough. As a consequence, populations could be readily
sampled throughout the year using smooth-sided, metal, bowl-shaped traps (20
cm diameter x 12 cm deep). These traps could be quickly set in the sand and
easily removed at the end of a trapping period with little disturbance to the
habitat. Traps were placed in an interdune valley (1 km wide x 5 km long) on
10 different occasions throughout 1979 to record phenology and obtain estimates
of population age structure. A minimum of 57 traps was used during each
trapping period. Traps were monitored at least twice daily until a total of 24 M.
picta had been captured. During June and July, decreased solifuge activity
precluded capture of 24 solifuges within a 2 week period, and trapping was
therefore terminated before the desired quantity was obtained. Trapped solifuges
were taken to the laboratory, where they were measured and, if adult, sexed.
Adult males were readily recognized by the presence of a flagellum, females by
differences in development of the genital operculum. All identifications were done
by the author. Solifuges were either maintained in the laboratory until the end
of the current trapping period, or immediately marked and released (none of the
75 marked individuals was ever recovered). Cheliceral length was measured to ±
0.1 mm using vernier calipers. Measurements of other body parts proved
considerably less reliable, based on repeated measurements of the same
individual: the opisthosoma was capable of considerable stretching; endpoints of
legs and palps were difficult to pinpoint; and the prosoma was flexible and
subject to some distortion depending on position of chelicerae. While there was
some evidence of wear on the cheliceral tips of a few of the larger adults, this

WHARTON— BIOLOGY OF METASOLPUGA

365

was not observed in immatures. Results of trap catches were plotted as size-
frequency distributions, using 0.5 mm size classes. The 0.5 mm size class was
based on a 0.4 mm range in measurements of 2nd instars.

Meterological data for the study period (including ground surface tempera-
tures) were obtained from the First Order Weather Station, Gobabeb. Additional
temperature records were made using a YSI® rapid-response telethermometer,
calibrated to correspond with Gobabeb Weather Station measurements. A Met-
tler® balance was used for weights recorded in the text. Where not otherwise
stated, averages reported in the text are means (m) + standard deviations. The
term “pebble” is used repeatedly in the text, and refers to small pieces of quartz
and CaCOs weighing 5 gm or less.

RESULTS AND DISCUSSIONS

Phenology and Diel Periodicity. — Metasolpuga picta was observed in the field
between 0915 and 1845. Trap catches indicated that they were only active during
the daylight hours, confirming the data presented by Holm and Scholtz (1980).
Of 65 solifuges encountered in the field, only two were observed after 1745, and
seven before 1200. This activity was confirmed during pitfall trap censuses (Fig.
1), where 79% of the individuals were trapped in the afternoon.

Both adult and immature M. picta were observed in the field during every
month except June and July. No individuals were trapped during a 2-wk
interdune valley pitfall trap census in June; and only one immature. M. picta
during a 9-day census in July. A pitfall trap census was not conducted in August.
Results of remaining censuses are shown in Fig. 1.

Solifuges were active in the field at ground surface temperatures between 40
and 6 PC. Nevertheless, caged individuals maintained at lower ambient
temperatures of 25-30° C still actively burrowed, and quickly responded to
stimulation of leg IV setae. Two individuals placed at 20 ± 1°C, however, became
lethargic and responded only weakly to the same stimuli. At 5-10°C, M. picta
became completely torpid. Of approximately 190 individuals subjected to
refrigeration at 5-10° C for up to 30 min during census periods (to facilitate
cheliceral measurement), all recovered completely, usually within 10 min. There
was no noticeable difference in response to refrigeration among the various
instars.

Reduced M. picta activity during June and July was correlated with low
temperatures (a relationship first noted by Heymons (1902) for solifuges). In
1979, June, July, and August (austral winter) were the only months in which
mean monthly ground surface temperatures at 1400 were 40° C or less (Table 1).
Assuming an activity threshold between 20 and 25° C, however, temperatures in
overnight burrows (see burrowing section below) are more likely to be a limiting
factor for activity than are surface temperatures. During June and July, monthly
means at 1400, 10 cm below the surface, were 23.5 ± 1.8 and 22.0 ± 1.9°C
(more than 10° C lower than the means for January and February). It is thus
possible that M. picta remained lethargic in burrows during cool days or weeks,
but emerged on warm days throughout the mild Namib Desert winter.

Based on trap catches (Fig. 1) and field observations, the interdune valley
population was essentially bivoltine in 1979. Eggs were deposited primarily in the

366

THE JOURNAL OF ARACHNOLOGY

6-

4-

2-

5

5-9X1

Fig. 1. — Size-frequency distributions represent-
ing 8 pitfall trap samples of Metasolpuga picta
(Kraepelin) taken from an interdune valley popu-
lation in 1979; 24 solifuges per trapping period.
Adults > 10.5 mm combined with the 10.0-10.5
mm class.

2-- rn-TirTI 24-28 k

n-i R-r-u-unJl-rL a-i7V

3579

rfiVn-uruTI

nJl-Thr-JT-un

11-18IV

24-26111

8-10III

21-23 II

6-911

I I I I I I I I H— I- I I I I I I

3456789 10

Chelceral L. (mm)

spring (September-October) and again in late summer (February-early April).
Data presented by Heymons (1902), Cloudsley-Thompson (1961b), Muma (1963,
1974a), and Junqua (1966) on 11 other solifuge species suggest that they are all
univoltine.

Figure 1 shows the appearance of the first motile instar ( =2nd instar), with
cheliceral length 2. 5-3.0 mm, beginning in November and March of 1979.
Cloudsley-Thompson (1977) has stated that in the Sudan and western North
America, production of young solifuges occurs in the summer, during the time
of annual desert rains. Andrewartha and Birch (1954) and Nichols et al. (1976)
have discussed similar correlations between precipitation and natality in desert
invertebrates and vertebrates. Rainfall in the Namib is sporadic (Schulze 1969,
Robinson and Seely 1980), and production of young M. picta was not correlated
with it. In 1979, for example, total monthly rainfall was >10 mm only in June.
In 1978, such rains occurred in February, March, and April. In 1980, they did
not occur. Similarly, production of young Eremobates durangonus Roewer
(Muma 1966e) occurs even when it doesn’t rain (Muma, personal communication).

Cheliceral length of adult males from the interdune valley ranged from 7.2-11.0
mm (m = 8.7 ± 0.9; N = 42), and adult females from 7.2-12.3 mm (m = 10.4
± 1.1; N = 28). However, about half the individuals in Fig. 1 with cheliceral
length >7.5 mm were still immature. Adults from the gravel plain were larger,
with female cheliceral length 8.0-16.2 mm (m = 12.7 ± 2.8; N = 12). Only one
male, with cheliceral length of 13.0 mm, was available for measurement from the
gravel plain. These findings support those of Junqua (1966) that adult size is not
fixed.

Cheliceral lengths of all individuals measured from the interdune valley
population (trap catches plus hand-collected specimens) are shown in Fig. 2. The
peak at 2.8 mm represents the first active or motile instar ( = second instar). The

WHARTON^BIOLOGY OF METASOLPUGA

367

Table 1. — Mean (m) monthly ground surface temperatures at 1400, during 1979 pitfall trap
censuses. No trapping conducted in August and October. No Metasolpuga picta trapped during 2-
week census in June; only one M. picta trapped during 9-day census in July.

Month

(m)

# days to
trap 24 M. picta

# traps

Feb.

57.9

3

57

3

63

Mar.

55.1

3

63

3

65

Apr.

49.1

8

61

May

42.7

14

60

June

36.0

-

60

July

40.2

-

52

Aug.

40.4

-

-

Sept.

47.1

5

59

Oct.

50.9

-

-

Nov.

55.6

5

60

first instar is inactive and passed within the oviposition chamber in the
Solpugidae; and the existence of a post-embryo was not determined during this
study. These data suggest that there are at least four active instars before
adulthood. The peak between 7.5 and 8.0 mm consists of nearly an equal number
of immatures and adult males (only one female). Since smaller adults are almost
entirely males, either females have one more instar than males, or male and
female chelicerae grow at different rates.

These data on cheliceral lengths serve only as a rough estimate for number of
instars; and Tranche and Sissom (1984) have recently discussed the limitations of
such indirect methods. Such estimates are of value as a basis for future work,
however, especially in light of the difficulty of rearing solifuges in captivity.
Similar estimates are available for only two other species. Muma (1966c)
estimated 8-10 instars for E. durangonus; and Junqua (1962, 1966) estimated six
instars for Othoes saharae Panouse. Junqua (1962, 1966) also stated that, because
of considerable observed variation in size, some individuals probably achieved
adulthood at the 4th instar, and others at the 9th or 10th instar.

Only five males were trapped in censuses from February through May (Fig. 1).
Although males dominated trap catches in late September (15 of 23 adults), they
had largely disappeared by early November (3 of 24 individuals, 3 of 15 adults).
This change in dominance from slightly more males than females in September
to many more females than males in November was corroborated by field
observations. Muma (1963, 1966e, 1974a) discussed an apparently similar seasonal
appearance of males before females in some North American species. The
appearance of males before females supports the hypothesis that males may have
one less instar than females. These data further suggest that adult male M. picta
are short-lived; and die after a relatively brief mating period. This can be
attributed to the high energy demands and low food intake observed during mate
location (discussed below). Females may be longer-lived because of the time
required for egg development and oviposition following mating. Supporting data
for M. picta, however, is meager. A single field-collected female oviposited after
one week in the laboratory, and died 2 weeks following oviposition.

Cloudsley-Thompson (1961b), Junqua (1962, 1966), and Muma (1963, 1966c)
have also presented evidence suggesting adults are short-lived; and that solifuges

368

THE JOURNAL OF ARACHNOLOGY

Fig. 2. — Size-frequency distribution for Metasolpuga picta (Kraepelin) chelicerae measured on
solifuges collected from interdune valley habitat, 1/ 1979-1/ 1980 {N = 257). Includes solifuges from
pitfall traps plus those collected by hand. Peaks at 2.8, 3.8, 5.3, and 6.3 mm suggest the presence of at
least four nymphal instars between the first (immobile and spent in oviposition chamber) and the
adults (starting at 7.2 mm). Peaks beyond 6.3 mm difficult to interpret because they represent
mixtures of immatures and adults.

do not live more than one year. Junqua (1966) showed that male O. saharae lived
for only about one week, but Muma (1966c) kept male E. durangonus alive for
up to one month.

Burrowing. — Solifuge burrowing behavior has been described by several
workers (Hutton 1843, Turner 1916, Kingston 1925, Junqua 1966, Muma 1966d,
Cloudsley-Thompson 1977). Burrows are constructed for digestion, oviposition,
ecdysis, hibernation, and retreats during inactive periods (Muma 1966d,
Cloudsley-Thompson 1977). Solifuges also burrow as a displacement activity
following disturbance (Cloudsley-Thompson 1961b).

Although solifuges studied elsewhere use rocks, logs, or other surface debris as
shelters, or to conceal burrows, these were generally not present on the sandier
substrates preferred by M. picta. The few large rocks available were not utilized
by this species. Thus the generalization by Newlands (1978) that in southern
Africa only hexisopodids occurred in sandy areas, and that members of other
solifuge families sheltered under stones, does not hold for at least some Namib
species. Otherwise, burrowing by M. picta was generally similar to that reported
for other solifuge species.

Metasolpuga picta either excavated a new burrow or occupied an abandoned
one at the end of each activity period. Although chelicerae are used extensively
in burrowing by other solifuge species (as detailed in reviews by Muma 1966d
and Cloudsley-Thompson 1977), they were utilized less readily by M. picta.
Instead, the sandy soil was rapidly excavated primarily with the second pair of
walking legs. As in other solifuges (Muma 1966d: 254), M. picta tossed soil out
beneath the opisthosoma (in canine fashion) while excavating (Fig. 3). Chelicerae
and the first pair of walking legs were used only rarely to loosen more compact
soil particles. Chelicerae were also used to carry out larger pebbles during
excavations. One adult male and two large immatures were each observed
transporting pebbles weighing 0.6-0. 7 gm. The pebbles weighed 2-5 times that of
the solifuges (wet weight).

Position reversal within the tunnel during burrowing was accomplished with a
simple turn rather than the more complicated operation described by Muma

WHARTON— BIOLOGY OF METASOLPUGA

369

Fig. 3. — Metasolpuga picta (Kraepelin) in the
act of burrowing. From original kindly furnished
by M. Seely.

(1966(1) for several North American species. Plowing was accomplished by
lowering the chelicerae onto the base of the accumulated refuse, and pushing the
soil out of the hole. The palps and first pair of legs were held on either side of
the pile. Burrow construction (total or partial) was observed in the field for 7
males, 9 females, and 4 immatures; and all excavated in this fashion. Burrow
length ranged from 50-300 mm (N = 13), and angle of penetration about 30-70°
from horizontal.

Solifuges plugged the entrance of overnight burrows with excavated sand 72.2%
of the time {N = 18). All newly excavated burrows were plugged, but 62.5% of
the solifuges occupying old, abandoned burrows (tV = 8) left the entrance
completely open. Both Turner (1916) and Cloudsley-Thompson (1961a) noted
irregular plugging of burrows. Junqua (1966) and Muma (1966d), however,
reported that burrows were regularly plugged.

Excavation of oviposition chambers by females differed from excavation of
burrows used as overnight shelters. Of four females observed preparing
oviposition chambers in the field, one completed the task in 25 min, but the
others took at least 60 min. Muma (1966d) reported a similar range of burrowing
times for both males and females of 4 North American species. Mean excavation
time of M. picta overnight burrows, however, was only 7 min {N = 5).

Oviposition burrows were plugged much more thoroughly than overnight
shelters, and it was difficult to trace them for measurements of burrow depth.
Only three of the presumed oviposition burrows were successfully excavated, and
these were 70, 120, and 170 mm deep. Mean vertical depth of overnight burrows
was 54 mm {N = 8).

The dimensions of M. picta burrows compare favorably with those reported for
most other solifuge species. Berland (1932) reported burrows of Galeodes up to
240 mm deep. Cloudsley-Thompson (1961a) described variation in burrow length
of 100-200 mm. Muma (1966d) recorded burrow lengths of 6-54 mm for
immatures and males, and 38-229 mm for females of 6 different species. Muma
(1966d) also noted variation in length and depth of burrows used for egg
deposition. Junqua (1966), however, reported a “remarkably constant” burrow
configuration for O. saharae and a burrow length of about 80-90 mm.

Excavation of overnight shelters, as well as digging during prey and mate
location (discussed below), was generally rapid. Thus the general conclusion by

370

THE JOURNAL OF ARACHNOLOGY

Muma (1966d: 251) that solifuge “burrowing, although conducted with extreme
vigor and activity, seems to be accomplished inefficiently and laboriously” is not
appropriate for this species. In M. picta, there are six spatulate setae posteriorly
on the second pair of walking legs. Most other members of the family Solpugidae
have five apically pointed setae in this position. The spatulate setae are clearly
adaptations to facilitate digging in sandy substrates.

Gore and Cushing (1980) concluded on the basis of four observations that
males of a North American species normally returned to the same sheltered
depressions for at least three consecutive nights to conserve energy. Unfortu-
nately, the authors did not indicate whether the solifuges were marked to allow
individual recognition or whether they were all given the same mark. Also, data
on two of the four observed males (Gore and Cushing 1980: Table 1) may have
been misinterpreted. In one case, only the depression was marked, rather than
the solifuge, and the solifuge which occupied the depression on a subsequent day
may not have been the same. In a second case, apparently two different
individuals occupied a single depression on successive nights. In the case of M.
picta, it is unnecessary to invoke homing as a means of conserving energy, since
burrows abandoned by other arthropods and small vertebrates are plentiful, and
readily occupied by this species.

Foraging. — Solifuges are extremely active, with most of their above-ground
activity spent foraging or (in adult males) locating mates. Although Muma (1967)
has interpreted much of their activity as simply “investigating their surroundings,”
I interpret such activity as typical foraging behavior in a cursorial predator.

Few field observations on solifuge foraging behavior have been made because
species previously studied were nocturnal, and were difficult to follow in the dark.
Diurnal solifuges in the Namib Desert were also difficult to observe because they
were easily disburbed, extremely fast, and surprisingly cryptic.

Sixty-five M. picta were observed foraging in the field. Fifteen of these were
followed continuously for at least 15 min each. Females and immatures searched
the habitat by thoroughly covering the area in a rapid, zig-zag pattern. They
probed all burrows, pebbles, and small pieces of detritus. Dead grass clumps and
the few larger stones present in the study area were also thoroughly searched
when encountered. However, nearly all of the foraging observed during this study
occurred in the open. Actual distances covered by females and immatures during
foraging generally could not be determined because of their rapid, zig-zag motion.
However, seven individuals followed for periods ranging from 5-87 min traveled
a mean straight-line distance of only 106.8 m/h. The single individual running
more or less in a straight line traveled 978.9 m/h. This latter figure is comparable
to velocities achieved by males.

Searching behavior of adult males differed from that of females and all
immatures. Adult males traveled at nearly the same speed as females and
immature, but rarely exhibited the intense, zig-zag searching behavior of the
latter. They thus covered much greater straight-line distances (1062.1 ± 593.1 m/
h; A = 9, range of observations, 5-100 min). When extensive searching of a
limited area by males was observed, males searched in a more deliberate, circling
pattern. Only two of 18 males found searching in the field searched in a zig-zag
pattern for more than 10% of the observation time.

One immature with a body length of 13 mm was observed for 1.5 h on the
gravel plain. It foraged mostly in the open, pawing at pebbles about a third its

WHARTON— BIOLOGY OF METASOLPUGA

371

body length or less. It used primarily the second pair of walking legs for digging,
and the palps and the first pair of walking legs for perception. Although there
were several dead grass clumps in the area being searched, the solifuge only
foraged through one of these. The solifuge dug one burrow, but it collapsed
shortly after the start of excavation. The solifuge also cleared the entrance to an
abandoned burrow, and stayed within the tunnel for 4 min before resuming
foraging. On four occasions, the solifuge stopped for 3 sec, then resumed
foraging. The only other breaks in the foraging pattern occurred when four small
invertebrates of undetermined identity were dug up from around four different
pebbles and immediately fed upon. The solifuge covered an area of roughly 40
m^ in 1.5 h. This foraging pattern was typical of all females and immatures
observed, with variation limited to number of holes dug or occupied, number of
feeding bouts or pauses unrelated to feeding, and amount of time spent around
grass clumps.

Females and immatures fed, or attempted to feed, on every invertebrate
encountered in the field. Males were observed feeding only twice, even though
total observation time for males (5.5 h) was slightly more than that for females
and immatures. Moreover, males on seven different occasions ran into potential
prey (species eaten by similarly sized females or immatures) without attempting
to feed on them.

Prey items were located primarily by direct contact (as detailed by Bolwig
(1952)) during running, or while pawing around pebbles. Because of the frequency
with which M. picta entered holes for short periods during foraging, it is possible
that prey were also trapped and eaten in these holes. Field-captured prey of
females and immatures were rarely identified because of their small size and the
speed with which they were captured and ground up in the chelicerae. The single
item positively identified was the tettigoniid Comicus sp.

Two of the nocturnal Comicus were excavated by a single solifuge. In each
case, the solifuge dug three small holes 30 mm apart, with each hole about 30
mm from the cricket. The resulting pattern formed a square, with the cricket at
one corner and excavations at the other three corners. The solifuge quickly
alternated from one hole to the next, spending about 5 sec at each hole. At the
first excavation site, the cricket emerged from its burrow 4 min after the solifuge
began excavating, and jumped away. The solifuge pursued, but was unable to
follow after two jumps of the tettigoniid. In the second excavation, the solifuge
dug for 3 min, then ran immeditely to the buried cricket and pulled it out of
the sand with chelicerae and palps. The solifuge macerated the cricket for 15 sec,
then carried it into a hole. The solifuge remained in the hole for 5 min before
resuming foraging.

In both cases, there was no visible evidence of the cricket or its burrow on the
surface of the soil. Thus, M. picta is capable of detecting prey other than visually
or by direct contact. Muma (1966b) has discussed the interactions of sight, touch,
and vibration in prey location by solifuges. Prey location is even more complex
than this, however. Detection by means of a kairomone is possible, since the
malleoli are chemoreceptors (Brownell and Farley 1974); and they touch the
ground while the solifuge forages (Cloudsley-Thompson 1961a). Moreover,
detection of sound waves through sandy soils has been shown to be an important
method of prey detection in scorpions (Brownell 1977, Brownell and Farley 1979),
and this method may be used by solifuges as well.

372

THE JOURNAL OF ARACHNOLOGY

Caged M. picta ate all soft-bodied prey offered. These included termites
(Hodotermes mosambicus (Hagen) and Psammotermes allocerus (Silvestri),
mealworms (Tenebrio molitor L.), moths, roaches, and silverfish. Larger
Coleoptera larvae were chewed apart at intersegmental membranes. Tips of the
chelicerae were then used to scoop out soft parts from the otherwise intact
exoskeleton. Numerous zophosine and two adesmiine Tenebrionidae were also fed
upon in this fashion after head and prothorax had been severed from the rest
of the body. In the field, only two encounters between adult beetles and M. picta
were observed. In the first encounter, the solifuge picked up the beetle, but
dropped it after one sec. In the second encounter, the solifuge touched the beetle
with its palps, then ignored it. The hard exoskeleton of beetles and the normally
rapid inspection of potential prey items by the solifuge would seem to preclude
adult Coleoptera from being an important component of M. picta's diet, even
though this species is clearly capable of eating them.

Although Newlands (1978) stated that solifuges relied solely on the chelicerae
to capture prey, M. picta occasionally brought prey to the chelicerae with the
palps. Muma (1966b) has provided a good discussion of variation in the use of
chelcierae and palps in prey capture.

The cheliceral teeth of male solifuges are often smaller, blunter, and/or less
numerous than those of females and immatures of the same species. This is also
true of M. picta. Reduced cheliceral teeth are correlated with reduced feeding by
the males, and are also of benefit to the male during courtship (described below),
since larger, sharper teeth might puncture the female opisthosoma during the
massaging phase. The reduced cheliceral teeth and marginal interest in feeding
support the hypothesis derived from phenological data that males are very short-
lived.

Junqua (1966) accurately noted that potential prey of desert solifuges are often
scarce, in part because solifuges feed largely on relatively soft-bodied
invertebrates. Solifuges are also voracious predators (e. g. Hutton 1843, Bernard
1897, Heymons 1902, Turner 1916, Lawrence 1949, Cloudsley-Thompson 1958);
and Muma (1966b) and Cloudsley-Thompson (1977) have reviewed their feeding
habits. Extreme activity, voracious appetities, and widely dispersed prey (at least
for desert species) are interrelated: to locate a sufficient quantity of food in a
desert, solifuges must forage over large areas (or be highly specific such as some
of the termitophilous species). By increasing foraging speed, they cover more
territory per unit time, and therefore increase the probability of encountering
prey. But increased foraging speed results in increased water loss, and is energy
expensive. Therefore, to replenish energy stores used in foraging, solifuges must
eat large quantities.

Opisthosomal Elevation. — Of the 15 solifuges observed foraging on the surface
for more than 15 min, five frequently elevated the opisthosoma in scorpion-like
fashion (Fig. 4), and one did so only rarely (3X in 1.7 h). The five which
regularly exhibited this behavior were observed when ground surface tempera-
tures were 55-60° C. Only three other M. picta were observed for more than 15
min at these temperatures. Although none raised the opisthosoma, two regularly
stopped in the shade of grass clumps for 0. 5-3.0 min periods.

Berland (1932), following Walter (1889), interpreted opisthosomal elevation as
a defense reaction in the Galeodidae. Cloudsley-Thompson (1949) hypothesized
that this behavior mimics scorpions, especially in some of the Rhagodidae.

WHARTON— BIOLOGY OF METASOLPUGA

373

Fig. 4. — Elevation of opisthosoma by Metasol-
puga picta (Kraepelin) as an apparent thermoreg-
ulatory response. From original kindly furnished
by C. K. Brain.

However, since the opisthosoma is the softest part of the body (easily punctured,
for example, by insect mandibles), elevation may be as important in protection
(Pocock 1898) as in mimicry.

In M. picta, opisthosomal elevation was observed only during activity at high
ground surface temperatures, and never during flight or defense behavior (see
description of predator avoidance below). I therefore interpret opisthosomal
elevation as a thermoregulatory response. The scorpion Opisthophthalmus
latimanus (Koch) also elevates the opisthosoma as an adaptation to heat stress
(Alexander and Ewer 1958). Krakauer (1972) and Robinson and Robinson (1974)
have also discussed this behavior in detail for the spider Nephila clavipes
(Linnaeus). The predictions of Robinson and Robinson (1974) regarding complex
thermoregulatory behavior for cryptic species at low latitudes fit well for M.
picta.

Predator Avoidance and Predators. — Metasolpuga picta responded to vibrations
and tactile stimuli, and was also visually oriented. Flight reponse consisted of
traveling several meters at greatly increased speed, with chelicerae and prosoma
usually angled toward the source of the disturbance. This was followed either by
resumption of normal foraging behavior and speed or a sudden stop, depending
on strength of stimulus. When motionless, M. picta was not readily detected in
the field by the human eye from a distance > 2 m. The orange propeltidium and
chelicerae resembled the many orange pebbles in the interdune valley and gravel
plain. The yellow and black pattern on the rest of the body broke up the outline,
and thus contributed to the crypsis. Solifuges in the field exhibited flight
responses to a hovering sarcophagid fly, a hunting wasp (bembicine Sphecidae),
and to other solifuges as well as the author. However, flight response and crypsis
are more logically adaptations to visually oriented predators, such as vertebrates,
than to these other disturbances.

374

THE JOURNAL OF ARACHNOLOGY

Gore and Cushing (1980) concluded that there was little evidence of natural
predators of solifuges. However, a number of mammals (Bothma 1966, 1971,
Smithers 1971, Viljoen and Davis 1973, Roer 1975, Bigalke 1978, Nel 1978),
reptiles (Louw and Holm 1972, Haacke 1976, Robinson and Cunningham 1978,
Holm and Scholtz 1980), and birds (Distant 1892, Willoughby 1971, Brain 1974,
Brain and Brain 1977, Dixon 1981, Clark et al. 1983) are known to eat solifuges
in southern Africa, and there is one report of a hymenopteran predator (Bristowe
1973). During the present study, one large sparassid (Araneae) was observed
capturing and eating a male M. picta\ and M. picta chelicerae were found in one
Ludwig’s bustard {Otis ludwigii Ruppell) dropping and several raptor pellets (=
regurgitations).

Of 80 pellets collected from beneath one roost on the gravel plain (occupied
primarily by a greater kestrel, Falco rupicoloides A. Smith), 43 contained M.
picta chelicerae, or moveable cheliceral fingers probably belonging to M. picta.
There was an average of 1.2 complete sets (one set per solifuge) per pellet. Four
other raptor pellets, two from an interdune valley and two from the gravel plain,
were collected during the study period. All four contained M. picta chelicerae,
and averaged 5.2 cheliceral sets per pellet. Chelicerae of M. picta were not found
in several cape fox {Vulpes chama (A. Smith)), black-backed jackal {Canis
mesomelas Schreber), or lizard {Meroles cuneirostris (Strauch), Aporosaura
anchietae (Bocage), and Mabuya occidentalis (Peters)) droppings collected during
this study. Thus, raptors may be the most important predators of larger solifuges
active on the surface. This would help explain their visual sensitivity to objects
moving overhead (such as the fly and wasp noted above). Predation of lethargic
solifuges in subterranean burrows may also be important, but would be extremely
difficult to document.

Ruggiero et al. (1979) studied the effects of certain prey characteristics on
kestral predatory behavior in a North American species. They found that the
highest rates of attack were elicited by moving, familiar prey. Aberrant movement
enhanced acceptability of familiar prey. They hypothesized that operation of these
two factors should lead to selection for a uniform, immobile response in species
which are heavily preyed upon. The quick flight and sudden stop of M. picta fits
this model. Pianka and Pianka (1970) and Pianka (1971) have described similar
escape responses for agamid lizards in sparsely vegetated habitats in Australia.

Mate Location. — Mate location may be summarized as follows; 1) male
searches in a straight-line pattern; 2) female in burrow emits chemical or auditory
cues; 3) male encounters cues, and switches from straight-line pattern to more
intensive circular or criss-cross pattern; or, 3a) male directly encounters female
on surface; 4) from step 3, male locates area directly above buried female and
initiates digging; 5) female responds by coming to soil surface; 6) from steps 3a
and 5, male touches female and jumps back; 7) male grabs female opisthosoma
in his chelicerae.

Males of M. picta located conspecific females either by direct contact or by
indirect detection through the soil surface. Five males were observed to detect
females beneath the soil surface. In two of the encounters, different males
successfully mated with the same female. In each case, the female was excavated
at a spot showing no visible evidence of a burrow. The burrow opening which
the female entered after the first mating was several centimeters away from the
point of excavation by the second male. In the third encounter observed, a male

WHARTON— BIOLOGY OF METASOLPUGA

375

Started excavating in an area where the soil was uneven (indicating recent
burrowing), but with no visible burrow entrance. In all three cases, females came
to the surface in response to the males’ digging. Thus male digging may serve
as a stimulus to the female rather than as an attempt to actually excavate her.
In the fourth and fifth observations, males discovered open burrows occupied by
females, and waited at the entrance until the females appeared.

Another five males detected females by direct encounter, while females were
either excavating at the entrances of burrows or foraging. Only one of the
encounters resulted in mating. This mating occurred after a male ran into a
foraging female, then ran into her again a few minutes later as she was starting
to dig a burrow. One additional mating was observed in the field, but the initial
encounter was not witnessed.

The intense, zig-zag searching behavior of females is replaced in males by a
pattern of long, straight-line runs. These long-distance runs (one male traveled 2
km in 1.7 h) were interrupted at intervals by intensive searching of restricted
areas. Such searches were characterized by slow circling or pacing back and forth,
rather than the high intensity zig-zagging of females and immatures. With one
exception, these more restricted searches led to eventual discovery of females in
burrows. The three females in closed-off burrows were located more quickly than
those actively excavating.

Unreceptive females (including four of the five encountred while burrowing)
exhibited three different responses. They often ignored males, continuing their
foraging or burrowing without interruption. Alternatively, some exhibited a weak
aggressive response towards the males. This consisted of raising the prosoma,
extending the palps, and opening the chelicerae slightly. More rarely, burrowing
females took one step towards the males before running back down the burrow
for another load of soil.

Males which encountered unresponsive females in the act of burrowing
remained at the burrow until the female closed off the entrance. Three such males
remained for 47, 101, and 123 min before excavating their own overnight shelters.
During these vigils, males usually stood near the entrance, with their palps

overhanging the hole.

As in detection of concealed prey (described above), observations on mate
location may be interpreted as evidence for either chemical or auditory cues (or
both). If the malleoli are truly chemoreceptors (Brownell and Farley 1974), then
their role in detection of a female pheromone is strongly suggested by their
greater size in males. A single observation favors auditory over chemical cues: a
searching male stopped in an area lacking a visible burrow or surface anomaly,
and detected a large sparassid concealed within a trap-door covered tunnel. The
solifuge exhibited none of the typical flight responses of disturbed M. picta when
the spider sprang from its tunnel. Instead, the solifuge approached the spider with
palps extended until direct contact was made. At this point, the sparassid
pounced on the solifuge, killed it, and pulled it into its burrow. The sparassid
was roughly the same size and shape as an adult M. picta, and the male solifuge
exhibited the same behavior towards the sparassid as did other males
encountering female M. picta. In this instance, the ability of the sparassid to
mimic auditory stimuli of female M. picta can be explained by the similar size
of these two burrowing arachnids. However, mimicry of chemical cues by the
sparassid is also possible, and cannot be completely ruled out without further
testing.

376

THE JOURNAL OF ARACHNOLOGY

Males were unable to successfully pursue unresponsive females encountered on
the soil surface, and lost them within seconds after the first encounter. Two males
ran within 5 and 10 cm respectively of females burrowing and foraging on the
surface without contacting them, and apparently without detecting them. One of
these males, as well as a third one, ran by other females at least once before
direct contact was made minutes later. These observations suggest that males of
M. picta are unable readily to detect or follow females on the surface when
passing within short distances of them, unless physical contact is actually made.
Moreover, the sense used to locate mates (or prey) buried beneath the surface
is apparently insufficient or at best inefficient for locating individuals on the
surface in this species. These observations do not necessarily contradict the
hypothesized role of vision in predator detection since the eyes are dorsally placed
in solifuges, and much better adapted to detect motion from above, than laterally.

Mating. — Males initiated courtship by grabbing the posterior end of the female
opisthosoma with their chelicerae, while touching the female prosoma and
chelicerae with their palps. Females folded their legs up against their bodies, and
their palps against their chelicerae, when grabbed in this fashion. Receptive
females remained quiescent as the male thoroughly massaged the opisthosoma
with his chelicerae. Males gradually worked forward onto the prosoma and
female chelicerae. Unreceptive females began struggling as males moved
anteriorly. In both observations on unreceptive females, females broke away, and
males were unable to capture them. In the four successful matings observed,
males massaged females in this fashion for 2.5 min. At the completion of the
massaging phase, males were standing directly over females, both facing the same
direction. Males held the females’ legs and palps against their bodies (as described
by Heymons (1902) and Junqua (1966)). Males then released a single
spermatophore from the genital pouch at the base of the opisthosoma. Males
next moved quickly backwards, used their chelicerae to pick up the
spermatophore lying on the female’s dorsum, lifted the female opisthosoma to a
vertical position, and inserted the cheliceral tips and spermatophore directly into
the female’s genital opening. The flagellum was also inserted into the genital
opening at this time. Although the flagellum in the family Solpugidae has an
immovable base, the shaft is flexible. The shaft was held at a 45-90° angle from
its resting position while inserted in the female’s genital opening, but quickly
returned to lie along the chelicerae after mating. Males held females in this
position for 2. 5-3.0 min. They used their palps (and to a lesser extent their first
pair of legs) to hold the females; and constantly probed into the genital opening
with the tips of the chelicerae. Females recovered from their torpor while in this
position, and terminated mating by struggling free from the males. Total mating
time was 5-6 min (N = 3). Two males ran off quickly after the female struggled
free, but the other two remained in place. After mating, males wiped the tips of
their chelicerae in the soil and flexed them for about 1 min before resuming a
typical searching behavior. Two of the females re-entered burrows from which
they emerged in response to males’ digging stimuli, a third found and entered
another burrow, and the fourth ran off.

Mating in solifuges has been described by Heymons (1902), Cloudsley-
Thompson (1961b), Amitai et al. (1962), Junqua (1962, 1966), and Muma
(1966c,e, 1967). All previous observations were limited to three genera in two
families (Eremobatidae and Galeodidae). Detailed field observations were made
only by Heymons (1902) and Junqua (1966).

WHARTON-BIOLOGY OF METASOLPUGA

377

Heymons (1902) and Junqua (1966) stressed the importance of the suddenness
of the ‘‘attack” by the male in subduing the female (inducing torpor) and
preventing her counter-attack. The reports by Amitai et al. (1962) and Muma
(1966e, 1967) also suggest that the attack phase was critical. However, the females
they studied were generally more passively subdued. In all species studied to date,
including M. picta, both pedipalps and cheliceral massaging played a role in the
initial phase of mating. Hemons (1902) and Junqua (1966) attempted to
determine the relative importance of palps and chelicerae in mating, but were
largely unsuccessful. Junqua (1966), however, was able to induce the catatonic
state in receptive females with his fingertips, and Heymons (1902) by using
forceps. Thus the submissive state in solifuges is produced mechanically, as a
response to sudden attack. Heymons (1902) equated it to a similar phenomenon
commonly seen in vertebrates. This state was also produced very briefly in two
M. picta during initial rough handling when captured in the field.

Heymons (1902) stressed the importance of female receptivity on induction of
torpor in Galeodes. Female receptivity was also important in M. picta, since
males were unable to hold onto unreceptive females. A few observations (Junqua
1966, Wharton unpublished) suggest that torpor, once induced, was maintained
even when the male was removed. Cloudsley-Thompson’s observations on mating
in Galeodes grand Pocock also suggest this. Thus, continued massaging during
mating may not be necessary to maintain the torpid condition of the female.

In M. picta and the Galeodidae, chelicerae were used to massage the genital
opening and to quickly transfer the spermatophore to the opening. Spermato-
phores were carried by only one chelicera in Galeodes (Amitai et al. 1962,
Cloudsley-Thompson 1961b), and between both fixed cheliceral fingers in the
galeodid Othoes (Junqua 1966). In the nocturnal or crepuscular galeodids,
spermatophores were deposited, probably on the soil, while the female’s abdomen
was bent back over the prosoma. In M. picta, however, the spermatophore was
deposited on the female’s dorsum before her opisthosoma was pushed back over
her head. This reflects either a basic difference between galeodids and solpugids,
or a need to protect the spermatophore from the soil by the diurnal solpugid in
response to high ground surface temperatures. In Eremobatidae, Muma (1966e)
observed direct transfer of the spermatophore from the genital opening of the
male to that of the female. Nevertheless, male eremobatids still massaged the area
around the genital opening before and after transfer. This massaging may
function in part in liberating the sperm from the spermatophore.

Mating was of approximately equal duration in all species studied thus far.
Massaging of the female by the male chelicerae, especially after spermatophore
transfer, was of longer duration in M. picta than in most other species. Recovery
from torpor after spermatophore transfer was also longer in M. picta than in
galeodids, but comparable to that in eremobatids. Heymons (1902), Amitai et al.
(1962), Cloudsley-Thompson (1961b) and Junqua (1966) all reported at least some
males picking up females with their chelicerae and carrying them about. This was
not observed in M. picta.

Use of the flagellum has not been previously reported for any solifuge. Junqua
(1966), in the most detailed study yet published on mating behavior, was unable
to discern the function of this structure. Lamoral (1975) presented morphological
evidence that the flagellum operates in storage and transmission of an exocrine
secretion. Based on Lamoral’s work, Cloudsley-Thompson (1977) suggested a role
in “brief displays of territoriality among males during the mating season.”

378

THE JOURNAL OF ARACHNOLOGY

Male solifuges often have longer appendages than females of the same species.
The longer legs and palps are correlated with the manipulation and containment
of the female during mating, and also with the long-distance mate-searching
behavior of the male.

Oviposition. — Actual oviposition by M. picta was observed only once. A
female confined in a cage dug a deep burrow one week after capture. Construction
of this burrow lasted several hours, and was similar to that described above for
presumed oviposition burrows in the field. The female spent considerably more
time packing soil to close off the burrow than she did during construction of
overnight shelters. The female emerged from this oviposition chamber after 5
days, leaving 63 eggs behind. She then covered over her emergence hole. The
female remained lethargic for nearly 2 weeks, did not feed, and did not return to
the eggs. She then ate actively for one day before she died.

Oviposition in Solpugidae was previously observed by Lawrence (1947, 1949).
He recorded the nocturnal Zeria caffra (Pocock) producing 192 eggs and Z.
hostilis (White) 64 eggs. Muma (1966a) presented data for the Eremobatidae; and
summarized findings for other Solifugae. He also noted that the number of eggs
per mass was intraspecifically variable.

Confinement. — The effects of caging and temperature must be considered when
interpreting solifuge behavior. This is particularly true for diurnal species such as
M. picta. For caged M. picta, movement of the terrarium from a shaded
laboratory into direct sunlight altered behavior. Solifuges which responded only
weakly to tactile stimuli at 20-25° C in shaded cages usually responded quickly
and aggressively to the same stimuli when cages were placed in the sun at 30-
35°C.

Burrowing and feeding by M. picta was influenced by caged conditions. Unless
the soil was of sufficient depth, quality, and compactness, burrows were not
constructed efficiently, frequently collapsed, and often were not completed. Caged
solifuges often huddled on the surface throughout the night instead of seeking
shelter. Caged solifuges generally ate more than was available to them in the field
per unit time. Moreover, heavily sclerotized invertebrates (such as many of the
tenebrionids) were more readily eaten in cages than in the field. Ability or
willingness to feed on adult beetles is directly related to continued contact in
caged environments.

Mating behavior was also influenced by confinement, and especially by size of
the arena. Field observations on M. picta suggest that males are generally able
to flee from unreceptive females, or from females with which they had recently
mated. When caged, however, fighting and resulting cannibalism are more
common (e.g. Muma 1966b, Cloudsley-Thompson 1977). In confinement, males
of M. picta were usually unable to approach females undetected, females
generally exhibited more aggression towards males than in the field, and males
were unable to flee from unresponsive females without constantly running into
them again. Polis and Farley (1979) studied cannibalism-minimizing behavior
during mating in the scorpion Paruroctonus mesaensis Stahnke. In the field,
mating behavior in M. picta produces comparable results.

Construction of the oviposition chamber, incubation, and the maternal care
observed by some workers (e.g. Hutton 1843, Lawrence 1949, Junqua 1966,
Cloudsley-Thompson 1967) may be similarly affected by laboratory confinement.
While maternal care in the Galeodidae is probable, it has not been observed in

WHARTON— BIOLOGY OF METASOLPUGA

379

Eremobatidae and Ammotrechidae (Turner 1916, Muma 1966a). Observations on
Z. caffra (Lawrence 1949) and M. picta need to be supplemented to determine
whether or not maternal care occurs in Solpugidae.

Muma (1966e, 1967) noted that caged individuals invariably reduced activity,
became lethargic, and eventually died. He termed this behavior “taming”; and
Cloudsley-Thompson (1977) has unfortunately followed this usage. This behavior
would seem to be a natural result of confining any highly active animal. By
greatly reducing the foraging area, and constantly offering food, the solifuge
cannot balance energy expenditure with food intake. Moreover, contamination
caused by increasing food and waste products in a confined area may also affect
activity.

Since the great majority of published data on solifuge biology was obtained
from caged individuals, behavioral comparisons must be made with caution.
More field data are needed to verify and expand the limited biological
information available from studies of caged individuals.

Distribution. — No attempts were made to determine the range of M. picta, but
several predictions can be made, based on the above findings. Leg morphology
(especially spatulate setae) and rapid burrowing behavior indicate that this species
is limited to sandy substrates. The construction of discrete burrows, and M.
picta's inability to burrow in loose sand, further restrict this species’ activities to
more compact soils (and thus to inderdune valleys rather than the loose sands
of the Namib dune slopes). Distribution north of the Kuiseb River and east of
the dune ecosystem should be fairly limited due to the harder substrates. Suitable
sandy conditions exist along the coast, but cooler temperatures brought about by
the Benguela Current may preclude activity there.

SUMMARY

Metasolpuga picta is a diurnal, bivoltine species active throughout most of the
year on compact, sandy soils in the Namib Desert. Thermoregulatory capabilities
are vital to species such as this which are active at high temperatures in habitats
largely devoid of vegetational and mineralogical shelters; and this topic merits
further investigation. Activity at high daytime temperatures is counterbalanced by
lethargy induced by the same cooler temperatures (20-25° C) at which many
nocturnal species are active. Feeding, burrowing, and mating behaviors are
similar to those reported for other solifuge species, but there are some differences.
The voracious nature of solifuges was generally confirmed, with the exception of
adult males. Both prey and mates concealed beneath the surface can be detected
by M. picta, but the mechanism for such detection is unknown. Differences were
observed in the types of burrows constructed or occupied, with overnight shelters
being constructed much more rapidly than previously reported for solifuge
burrows. Mating was similar to that reported for galeodids, but use of the
flagellum during mating has not been previously recorded. Its function is still
uncertain since possible uses (rupture of spermatophore, destruction of sperm or
spermatophores from previous matings, transport of spermatophore up oviduct)
do not explain species specificity in flagellum configuration. The absence of the
flagellum in certain solifuge families leaves its function even more difficult to
explain. The effect of vertebrate predators on solifuge behavior needs to be more

380

THE JOURNAL OF ARACHNOLOGY

thoroughly explored, especially for nocturnal species. Vertebrates, especially small
raptors, are important predators of M. picta.

ACKNOWLEDGMENTS

The support of the C.S.I.R., Transvaal Museum, and staff of D.E.R.U. during
my stay at Gobabeb is greatly appreciated. The Division of Nature Conservation
and Tourism, South West Africa/ Namibia, provided facilities and kindly granted
permission to work in part of the Namib-Naukluft Park. I wish to thank C. K.
Brain, W. Haacke, E. Holm, R. Lawrence, P. leRoux, W. Savary, M. Seely, R.
Tilson, and H. Williams for discussions on various aspects of solifuge biology;
and especially T. Friedlander, J. Roth, and an anonymous reviewer for supplying
criticisms of an earlier draft. C. K. Brain and M. Seely also kindly granted
permission to use their photographs. I am particularly grateful to F. von Blottnitz
for translating some difficult German passages, and S. Bethune and M.
Vandevalle for technical assistance. Much of this work could not have been
completed without the able assistance of S. Falconer and V. Gray, and to them
I am deeply indebted. Finally, I acknowledge my debt to previous workers,
especially M. Muma and J. Cloudsley-Thompson, for providing a basis on which
to work, and for helpful comments on the submitted draft.

LITERATURE CITED

Alberti, G. 1979. Licht-und electronenmikroskopische Untersuchungen an Coxaldruesen von
Walzenspinnen (Arachnida: Solifugae). Zool. Anz., 203:48-64.

Alberti, G. 1980. Zur Feinstruktur des Hodenepithels und der Spermien von Eusimonia mirabilis
Roewer 1934 (Solifugae, Arachnida). Zool. Anz., 204:345-352.

Alexander, A. J. and D. W. Ewer. 1958. Temperature adaptive behaviour in the scorpion,
Opisthophthalmus latimanus Koch. J. Expt. Biol., 35:349-359.

Aliev, Sh. 1. and A. T. Gadzhiev. 1983. Solpugids (Arachnida, Solifugae) in the Azerbaijan SSR. Izv.
Akad. Nauk Azerbaijan SSR, Ser. Biol., 4:43-46.

Amitai, R, G. Levy, and A. Shulov. 1962. Observations on mating in solifugid Galeodes sulfuripes
Roewer. Bull. Res. Council Israel (B) Zool., 11:156-159.

Andrewartha, H. G. and L. C. Birch. 1954. The Distribution and Abundance of Animals. University
of Chicago Press, Chicago. 782 pp.

Aruchami, M. and G. Sundara Rajulu. 1978. An investigation on the poison glands and the nature
of the venom of Rhagodes nigrocinctus (Solifugae: Arachnida). Nat. Acad. Sci. Letters (India),
1:191-192.

Bauchhenss, E. 1983. Morphology and ultrastructure of sensilla ampullacea in Solifugae (Chelicerata:

Arachnida). Int. J. Insect Morphol. Embryol., 12:129-138.

Berland, L. 1932. Les Arachnides. Encycl. Entomol., (A) 16:33-43.

Bernard, H. M. 1897. ‘Wind scorpions’, a brief account. Sci. Prog., 1:317-343.

Bigalke, R. 1978. Mammals. Pp. 981-1048. In Biogeography and Ecology of Southern Africa. (M.

J. A. Werger and A. C. Van Bruggen, eds.). Dr. W. Junk B. V., The Hague.

Bolwig, N. 1952. Observations on the behavior and mode of orientation of hunting Solifugae. J.

Entomol. Soc. Southern Africa, 15:239-240.

Bothma, J. du P. 1966. Food of the silver fox Vulpes chama. Zool. Africana, 2:205-210.

Bothma, J. du P. 1971. Reports from the mammal research unit: 2. Food habits of some Carnivora
(Mammalia) from Southern Africa. Ann. Transvaal Mus., 27:15-26.

Brain, C. K. 1974. The use of microfaunal remains as habitat indicators in the Namib. South African
Archeol. Soc. (Goodwin Ser.), 2:55-60.

Brain, C. K. and V. Brain. 1977. Microfaunal remains from Mirabib: some evidence of palaeo-
ecological changes in the Namib. Madoqua, 10:285-293.

WHARTON— BIOLOGY OF METASOLPUGA

381

Bristowe, W. S. 1973. A new lycosid spider in the Seychelles and an endemic eumenid wasp which
stores spiders. Entomologist, 106:220-222.

Brownell, P. 1977. Compressional and surface waves in sand: used by desert scorpions to locate prey.
Science, 197:479-481.

Brownell, P. H. and R. C. Farley. 1974. The organization of the malleolar sensory system in the
solpugid, Chanbria sp. Tissue and Cell, 6:471-485.

Brownell, P. H. and R. D. Farley. 1979. Prey-localizing behavior of the nocturnal desert scorpion,
Paruroctonus mesaensis: orientation to substrate vibrations. Anim. Behav., 27:185-193.

Clark, W. H., D. R. Frohlich, and P. L. Comanor. (1982) 1983. Shrike predation on the scorpion
Anuroctonus phaiodactylus (Wood) and on a solpugid (Scorpionida: Vaejovidae; Solpugida). Pan-
Pacific Entomol., 58:164.

Cloudsley-Thompson, J. L. 1949. Notes on Arachnida, 9. Do Solifugae mimic scorpions? Entomol.
Monthly Mag., 85:47.

Cloudsley-Thompson, J. L. 1958. Spiders, Scorpions, Centipedes, and Mites. Pp. 87-98 (revised ed.,
1968, pp. 105-123). Pegamon, Oxford. 228 pp.

Cloudsley-Thompson, J. L. 1961a. Some aspects of the physiology and behavior of Galeodes arabs.
Entomol. Exp. Appl., 4:257-263.

Cloudsley-Thompson, J. L. 1961b. Observations on the natural history of the “camel-spider” Galeodes
arabs C. L. Koch (Solifugae: Galeodidae) in the Sudan. Entomol. Monthly Mag., 97:145-152.
Cloudsley-Thompson, J. L. 1967. Reproduction in Solifugae. Entomol. Monthly Mag., 103:144.
Cloudsley-Thompson, J. L. 1977. Adaptational biology of the Solifugae (Solpugida). Bull. British
Arachol. Soc., 4:61-71.

Distant, W. L. 1892. Are the Solpugidae poisonous? Nature, 46:247.

Dixon, J. E. W. 1981. Diet of the owl Glaucidium perlatum in the Etosha National Park. Madoqua,
12:267-268.

Francke, O. F. and W. D. Sissom. 1984. Comparative review of the methods used to determine the
number of molts to maturity in scorpions (Arachnida), with analysis of the post-birth development
of Vaejovis coahuilae Williams (Vaejovidae). J. Arachnol., 12:1-20.

Gore, J. A. and B. S. Cushing. 1980. Observations on temporary foraging areas and burrows of the
sun spider, Ammotrechula penninsulana (Banks) (Arachnida: Solpugida). Southwest. Nat., 25:95-
102.

Haacke, W. 1976. The burrowing geckos of Southern Africa, 2 (Reptilia: Gekkonidae). Ann.
Transvaal Mus., 30:13-28.

Haupt, J. 1982. Hair regeneration in a solfugid chemotactile sensillum during molting (Arachnida:

Solifugae). Wilhelm Roux’s Arch. dev. Biol., 191:137-142.

Heymons, R. 1902. Biologische Beobachtungen an asiatischen Solifugen. Abh. Preussischen Akad.
Wiss., 190:1-65.

Hingston, R. W. 1925. Nature at the Desert’s Edge; Studies and Observations in the Bagdad Oasis.
Whitherby, London. 299 pp.

Holm, E. and E. B. Edney. 1973. Daily activity of Namib Desert arthropods in relation to climate.
Ecology, 54:45-56.

Holm, E. and C. H. Scholtz. 1980. Structure and pattern of the Namib Desert dune ecosystem at
Gobabeb. Madoqua, 12:5-39.

Hutton, T. 1843. Observations on the habits of a large species of Galeodes. Ann. Mag. Nat. Hist.,
12:81-85.

Junqua, C. 1962. Donnes sur la reproduction d’un solifuge: Othoes saharae Panouse. C. R. Acad.
Sci., Paris, 225:2673-2675.

Junqua, C. 1966. Recherches biologiques et histophysiologiques sur un solifuge Saharien Othoes
saharae Panouse. Mem. Mus. Natn. Hist. Nat., Paris (A), 43:1-124.

Kraepelin, K. 1899. Zur Systematik der Solifugen. Mitt. Naturh. Mus. Hamburg, 16:187-259.

Krakauer, T. 1972. Thermal responses of the orb-weaving spider Nephila clavipes (Araneae:
Argiopidae). Amer. Midland Nat., 88:245-250.

Lamoral, B, H. 1975. The structure and possible function of the flagellum in four species of male
solifuges of the family Solpugidae. Proc. 6th Int. Congr. Arachnol., pp. 136-141.

Lawrence, R. F. 1947. Some observations on the eggs and newly hatched embryos of Solpuga hostilis
White (Arachnida). Proc. Zool. Soc. London, 117:429-434.

Lawrence, R. F. 1949. Observations on the habits of a female solifuge, Solpuga caffra Pocock. Ann.
Transvaal Mus., 21:197-200.

382

THE JOURNAL OF ARACHNOLOGY

Lawrence, R. F. 1963. The Solifugae of South West Africa. Cimbebasia, 8:1-28.

Lawrence, R. F. 1965. New and little known Arachnida from the Namib Desert, S. W. Africa. Sci.
Papers Namib Desert Res. Stn., 27:1-12.

Lawrence, R. F. 1967. Additions to the fauna of South West Africa: solifuges, scorpions and
Pedipalpi. Sci. Papers Namib Desert Res. Stn., 34:1-19.

Louw, G. N. and E. Holm. 1972. Physiological, morphological and behavioural adaptations of the
ultrapsammophilous Namib Desert lizard, Aporosaura anchietae (Bocage). Madoqua, (11)1:67-85.

Muma, M. H. 1963. Solpugida of the Nevada Test Site. Sci. Bull. Brigham Young Univ. (Biol. Ser.),
3:1-13.

Muma, M. H. 1966a. Egg deposition and incubation for Eremobates durangonus with notes on the
eggs of other species of Eremobatidae (Arachnida: Solpugida). Florida Entomol., 49:23-31.

Muma, M. H. 1966b. Feeding behavior of North American Solpugida (Arachnida). Florida Entomol.,
49:199-216.

Muma, M H. 1966c. The life cycle of Eremobates durangonus (Arachnida: Solpugida). Florida
Entomol., 49:233-242.

Muma, M. H. 1966d. Burrowing habits of North American Solpugida (Arachnida). Psyche, 73:251-
260.

Muma, M. H. 1966e. Mating behavior in the solpugid genus Eremobates Banks. Anim. Behav.,
14:346-350.

Muma, M. H. 1967. Basic behavior of North American Solpugida. Florida Entomol., 50:1 15-123.

Muma, M. H. 1974a. Maturity and reproductive isolation of common solpugids in North American
deserts. J. Arachnol., 2:5-10.

Muma, M. H. 1974b. Solpugid populations in southwestern New Mexico. Florida Entomol., 57:385-
392.

Muma, M. H. 1975a. Two vernal ground-surface arachnid populations in Tularosa Basin, New
Mexico. Southwest. Nat., 20:55-67.

Muma, M. H. 1975b. Long term can trapping for population analyses of ground-surface, arid-land
arachnids. Florida Entomol., 58:257-270.

Muma, M. H. 1979. Arid-grassland solpugid population variations in southwestern New Mexico
USA. Florida Entomol., 62:320-328.

Muma, M. H. 1980a. Comparison of three methods for estimating solpugid (Arachnida) populations.
J. Arachnol., 8:267-270.

Muma, M. H. 1980b. Solpugid (Arachnida) populations in a creosotebush vs. a mixed plant
association. Southwest. Nat., 25:129-136.

Nel, J. A. J. 1978. Notes on the food and foraging behavior of the bat-eared fox, Otocyon megalotis.
Bull. Carnegie Mus., 6:132-137.

Newlands, J. 1978. Arachnida (except Acari). Pp. 685-702, In Biogeography and Ecology of Southern
Africa. (M. J. A. Werger and A. C. Van Bruggen, eds.). Dr. W. Junk B. V., The Hague, 1439

pp.

Nichols, J. D., W. Conley, B. Batt, and A. Tipton. 1976. Temporally dynamic reproductive strategies
and the concept of r- and K- selection. Amer. Nat., 1 10:995-1005.

Pianka, E. R. 1971. Ecology of the agamid lizard Amphibolurus isolepis in Western Australia. Copeia,
1971:527-536.

Pianka, E. R. and H. D. Pianka. 1970. The ecology of Moloch horridus (Lacertilia: Agamidae) in
Western Australia. Copeia, 1970:90-103.

Pocock, R. 1. 1898. On the nature and habits of Pliny’s Solpuga. Nature, 57:618-620.

Polis, G. A. and R. D. Farley. 1979. Behavior and ecology of mating in the cannibalistic scorpion,
Paruroctonus mesaensis Stahnke (Scorpionida: Vaejovidae). J. Arachnol., 7:33-46.

Purcell, W. F. 1899. New and little known South African Solifugae in the collection of the South
African Museum. Ann. South African Mus., 1:381-432.

Robinson, M. D. and A. B. Cunningham. 1978. Comparative diet of two Namib Desert sand lizards
(Lacertidae). Madoqua, 11:41-53.

Robinson, M. D. and M. K. Seely. 1980. Physical and biotic environments of the southern Namib
dune ecosystem. J. Arid Environ., 3:183-203.

Robinson, M. H. and B. C. Robinson. 1974. Adaptive complexity: the thermoregulatory posture of
the golden-web spider, Nephila clavipes, at low latitudes. Amer. Midland Nat., 92:386-396.

Roer, H. 1975. Zur Kenntnis der Chiropterenfauna Suedwestafrikas. J. South West Africa Sci. Soc.,
29:105-127.

WHARTON— BIOLOGY OF METASOLPUGA

383

Ruggiero, L. F., C. D. Cheney, and F. F. Knowlton. 1979. Interacting prey characteristic effects on
kestrel predatory behavior. Amer. Nat., 1 13:749-757.

Savory, T. 1977. Arachnida (2nd ed.). Academic Press, London, New York, San Francisco. 340 pp.
Schulze, B. R. 1969. The climate of Gobabeb. Sci. Papers Namib Desert Res. Stn., 37:5-12.

Seely, M. K. and G. N. Louw. 1980. First approximation of the effects of rainfall on the ecology
and energetics of a Namib Desert dune ecosystem. J. Arid Environ., 3:25-54.

Smithers, R. H. N. 1971. The mammals of Botswana. Mem. Mus. Nat. Hist. Rhodesia., 4:1-340.

Thaler, K. 1982. Die Primaerlarve der Walzenspinne Gylippus cf. cypriotica Lawrence (Arachnida,
Solifugae, Karschiidae). Mitt, schweizerischen entomol. Ges., 55:93-95.

Turner, C. H. 1916. Notes on the feeding behavior and oviposition of a captive American false spider.
J. Anim. Behav., 6:160-168.

Viljoen, S. and D. H. S. Davis. 1973. Notes on stomach contents analyses of various carnivores in
Southern Africa (Mammalia: Carnivora). Ann. Transvaal Mus., 28:353-363.

Walter, A. 1889. Transcaspische Galeodiden. Zool. Jb. (Syst.), 4:1094-1 109.

Wharton, R. A. 1981. Namibian Solifugae (Arachnida). Cimbebasia Mem., 5:1-87.

Wharton, R. A. and M. Seely. 1982. Species composition and biological notes on Tenebrionidae of
the lower Kuiseb River and adjacent gravel plain. Madoqua, 13:5-25.

Willoughby, E. J. 1971. Biology of larks (Aves: Alaudidae) in the Central Namib Desert. Zool.
Africana, 6:133-176.

Manuscript received January 1986, revised March 1986.

. ...

''%M^

S&:..' -fej

Terranova, A. C. and S. H. Roach. 1987. Genetic differentiation in the genus Phidippus (Araneae,
Salticidae). J. Arachnol., 14:385-391.

GENETIC DIFFERENTIATION IN THE GENUS PHIDIPPUS
(ARANEAE, SALTICIDAE)i

A. C. Terranova and S. H. Roach

Cotton Production Research Unit
Agricultural Research Service
U.S. Department of Agriculture
RO. Box 2131

Florence, South Carolina 29503

ABSTRACT

Electromorph profiles of seven Phidippus species found in South Carolina were compiled by
acrylamide gel electrophoresis. Twelve gene loci from 10 enzymes were analyzed. Mean heterozygosity
of the seven species populations was 0.117 with an average proportion of polymorphic loci of 0.416.
The mean genetic distance of all spider populations was 0.56 + 0.04. P. otiosus (Hentz) and P. regius
C. L. Koch were more closely related to each other (I = 0.781) than were any other pair of jumping
spiders.

INTRODUCTION

Roach and Edwards (1984) reported the occurrence of seven species of
Phidippus in South Carolina, and Terranova and Roach (1987) developed an
electrophoretic key to identify these seven species in both immature and adult
forms.

In a partial revision of the genus Phidippus, Edwards (1980) divided the genus
into two subgenera. One subgenus contains the South Carolina species Phidippus
putnami (Peckham and Peckham), and the other subgenus contains all other
species observed in South Carolina.

According to the classification of Edwards (1980), P. whitmani Peckham and
Peckham and P clams Keyserling belong to the cardinalis group. P. audax
(Hentz), P. otiosus (Hentz) and P. regius C. L. Koch belong to the audax group,
and P. princeps Peckham and Peckham belongs to the princeps group which
Edwards (1980) states may be related to the audax group.

Edwards (1980) presented a cladogram of phylogenetic relationships among
these seven species of Phidippus and others not reported from South Carolina
based primarily on morphological characters. He further presented evidence of
courtship relationships which supported his revised classification scheme
(Edwards 1980).

In this report, we present information to elucidate the genetic relationships
among the seven members of South Carolina Phidippus. This information is

'In cooperation with the South Carolina Agricultural Experiment Station. Mention of a commercial
or proprietary product does not constitute endorsement by the USDA.

386

THE JOURNAL OF ARACHNOLOGY

Table 1. — Enzyme systems studied and their polymorphic (P) or monomorphic (M) status in one
or all species of Phidippus. Presumed subunit structure of observed electromorphs is indicated.

Enzyme

Enzyme

commission

number

Abbreviation

and

locus

Polymorphic (P)

or

monomorphic

(M)

SubuniT

no.

Aspartate aminotransferase

2.6.1. 1

Aat-1

P

2

Amylase

3. 2.1.1

Amy-1

P

1

Fumarase

4.2.1.2

Fum-1

M?

7

a-Glycerophosphate dehydrogenase

1.1. 1.8

a-Gpd-l

P

9

Isocitrate dehydrogenase

1.1.1.42

Idh-1

P

2

Idh-2

P

2

Malate dehydrogenase

1.1.1.37

Mdh-1

P

2

Mdh-2

P

2

Phosphoglucose isomerase

5. 3.1.9

Pgi-1

P

2

Phosphoglucomutase

2.7.5.1

Pgm-1

P

1

Superoxide dismutase

1.15.1.1

Sod-I

M

9

Tyrosinase

1.10.3.1

Tyr-1

P

1

'‘1 = Monomeric protein; 2 = dimeric proteins.

derived from the patterns of genetic variability at 12 presumptive gene loci in
each Phidippus. The techniques of polyacrylamide gel electrophoresis and
isozyme detection were used to obtain this information.

METHODS

Spiders used in these studies were either field-collected Phidippus mainly from
eastern South Carolina or progeny obtained from laboratory-reared spiders. The
species P. audax, P clams, and P. princeps occurred widely and were fairly
abundant, whereas P. otiosus, P putnami, P regius, and P whitmani were rarely
collected due to their scarcity or their habitat specificity.

The electrophoretic techniques and the isozyme staining techniques used have
been previously described (Terranova 1978, 1980, 1981a,b). A total of 12 gene loci
represented by 10 enzyme systems were studied in every species in which they
were present (Table 1).

Analysis of electrophoretic data was accomplished using program GELDAT
(Terranova and Smith unpub.). Population statistics and dendograms were based
on Ferguson (1980). Genetic divergence between species was determined by the
genetic similarity (I) and genetic distance (D) statistics of Nei (1972). Similarity
values (I) are a measure of the proportion of loci at which the allelic constituents
of two populations are essentially identical. Similarity values range from zero (no
identity) to one (complete identity). Genetic distance (D) values estimate the
effective number of complete allelic substitutions per locus which have
accumulated since any two populations diverged from a common ancestor.

RESULTS

The data presented in Table 2 show that the percent of polymorphic loci ranges
from 16.7% in P putnami to 58.3% in P. audax, P clams, and P. whitmani with

TERRANOVA & ROACH— GENETIC DIFFERENTIATION IN PHIDIPPUS

387

Table 2. — Levels of genetic variation in seven Phidippus species observed in South Carolina.

Species

Individuals

sampled

Mean % ± S.E.
heterozygotes
expected per
locus

Percent

polymorphic

locP

Mean no.

alleles
per locus

P audax

58

17.2 ± 5.0

58.3

1.83

P clams

38

11.8 ± 4.3

58.3

1.83

P otiosus

11

6.1 ± 3.9

25.0

1.33

P princeps

33

17.3 + 5.7

50.0

1.75

P putnami

7

5.9 ±4.3

16.7

1.17

P regius

21

4.7 ± 3.3

25.0

1.25

P. whitmani

10

17.9 + 5.1

58.3

1.75

^Frequency of the most common allele <0.99.

an average for all seven species of 41.6% out of 12 loci. The expected mean
frequency of heterozygous individuals per locus ranges from 4.7% in P. regius to
17.9% in P. whitmani, and the average for the genus is 11.7 + 4.5%. The lower
percent of heterozygotes per locus observed in the species P. otiosus, P putnami,
and P regius as compared to the remaining species is probably due to the small
number of field-collected spiders. For example, the low heterozygosity observed
in P regius is more than likely the result of most specimens belonging to the
same family. In this case, only one wild adult was ever found. The remaining
specimens were collected as spiderlings, presumably from the same mother. The
electromorph [= allele, see Berlocher (1980) for discussion] frequencies at the 12
genetic loci as shown in Table 3.

The levels of genetic variation found in the genus Phidippus compared to other
invertebrates show that they are characteristic of invertebrates in general. The
data presented by Ferguson (1980) show a mean percent polymorphism of 51%
and a mean percent heterozygosity of 15.5% for a wide assortment of
invertebrates compared to 41% and 11.7%, respectively, for Phidippus.

Estimates of genetic similarity (I) and genetic distance (D) between all species
pairs are given in Table 4. Genetic distance ranges from 0.247 for P. regius vs.
P. otiosus to 0.846 for P. princeps vs. P. putnami with an average value of D
for all pair-wise comparisons of 0.56 ± 0.04. Thus, on average, about 0.56
electrophoretically detectable electromorphic substitutions per locus have
occurred in the separate evolutions of any two species.

Figure 1 shows the distribution of genetic similarities among loci. Generally,
pairs of species are essentially identical (I = <0.95) at nearly 52% of the loci and
completely different (I = <0.005) at about 32% of the loci, and few loci have
genetic similarities in the broad range between 0.05 and 0.95. Again, our data
correlate nicely with those reported for other invertebrate species (e.g. Drosophila
willistoni group; Ayala et al. 1974a,b). Ferguson (1980) comments that this U-
shaped distribution may be representative of most outcrossing sexual organisms.
Among others, Ferguson (1980) and Avise and Ayala (1976) state that the
significance of this bimodality of distribution is that for a given locus, any two
species are either identical in electromorphic composition or completely distinct
with unique electromorphs. This is important from a systematist’s point of view
in that it is not essential to sample many individuals at each locus since
electromorphic frequency differences are not an important feature of interspecific
studies (Ferguson 1980).

388

THE JOURNAL OF ARACHNOLOGY

Table 3. — Frequencies of electromorphs in seven species of Phidippus.

Gene &

electro-

morph

Species

audax

clams

otiosus

princeps

putnami

regius

whitmani

Aat-1

36

-

0.013

-

-

-

-

-

46

0.053

-

-

-

-

-

-

55

0.947

0.987

1.000

1.000

1.000

1.000

1.000

Amy-1

23

-

0.048

-

-

-

-

-

30

-

0.077

-

-

-

-

-

34

-

-

0.667

-

-

-

-

36

0.121

-

-

-

-

-

-

40

-

0.712

0.208

0.204

-

0.976

0.182

45

0.638

0.067

0.125

0.278

1.000

0.024

0.500

50

0.233

-

-

0.315

-

-

0.273

55

-

0.096

-

0.093

-

-

-

58

0.009

-

-

-

-

-

-

63

-

-

-

0.111

-

-

0.045

Fum-1

55

-

1.000

1.000

1.000

-

-

-

58

1.000

-

-

-

1.000

1.000

1.000

Idh-1

38

0.588

0.900

-

0.222

-

-

0.786

50

0.412

0.100

1.000

0.778

1.000

1.000

0.214

Idh-2

55

0.921

1.000

1.000

0.852

1.000

1.000

0.938

65

0.079

-

-

0.148

-

-

0.063

a-Gpd-l

116

-

-

-

0.256

-

-

0.167

120

1.000

-

-

-

-

-

-

124

-

1.000

1.000

0.750

-

1.000

0.917

129

-

-

-

-

1.000

-

0.083

Mdh-1

21

-

0.056

0.062

-

-

-

-

25

0.589

-

-

-

0.429

-

-

30

-

0.944

-

-

0.571

-

-

35

0.411

-

0.938

1.000

-

0.700

1.000

43

-

-

-

-

-

0.300

-

Mdh-2

123

0.905

0.941

0.938

0.950

0.786

0.900

0.836

143

0.095

0.059

0.062

0.050

0.214

0.100

0.164

Pgm-1

65

-

-

-

1.000

-

-

0.600

70

1.000

0.067

1.000

-

1.000

1.000

0.400

76

-

0.333

-

-

-

-

-

Pgi-1

45

-

0.041

-

0.147

-

-

-

55

1.000

0.959

1.000

0.853

1.000

1.000

-

60

-

-

-

-

-

-

0.143

Sod-1

134

-

-

1.000

-

-

-

-

138

-

-

-

-

-

1.000

-

143

-

-

-

-

1.000

-

-

156

-

1.000

-

1.000

-

-

-

159

1.000

-

-

-

-

-

1.000

Tyr-1

27

0.066

-

-

-

-

-

-

33

0.920

1.000

-

-

-

-

-

37

-

-

-

-

1.000

-

-

44

0.020

-

1.000

1.000

-

1.000

1.000

In order to more easily visualize the information in Table 4, we evaluated the
data by an agglomerative clustering procedure, the unweighted pair-group method
with arithmetic means (UPGMA, Sneath and Sokal 1973). The resultant
dendogram is presented in Figure 2. The relationship based on the biochemical
similarities between Phidippus species are discussed in relation to the
interpretations of Edwards (1980).

TERRANOVA & ROACH-GENETIC DIFFERENTIATION IN PHIDIPPUS

389

Table 4. — Matrix of genetic distance (above diagonal) and genetic similarities (below diagonal)
between species in Phidippus based on 12 loci. Distances and similarities calculated using the method
of Nei(1972).

Species

Species

audax

clarus

otiosus

princeps

putnami

regius

whitmani

P. audax

.

0.558

0.633

0.77I

0.423

0.672

0.443

P clams

0.573

-

0.480

0.416

0.763

0.726

0.752

P otiosus

0.531

0.619

-

0.261

0.647

0.247

0.502

P princeps

0.463

0.660

0.770

-

0.846

0.4I6

0.402

P putnami

0.655

0.466

0.524

0.429

-

0.672

0.829

P regius

0.51I

0.484

0.781

0.659

0.511

-

0.301

P. whitmani

0.642

0.47 1

0.606

0.669

0.437

0.740

-

DISCUSSION

The cladogram presented by Edwards (1980) depicting hypothetical phylogenies
of the Phidippus species was constructed utilizing morphological characters
considered primitive or derived. His observations on the overall courtship
behaviors of Phidippus proved to be a good indicator of these phylogenetic
relationships (Edwards 1980).

Both the cladogram of Edwards, based on morphological characters, and our
dendogram based on biochemical electromorph differences agree in one important
aspect; that of the relationship between P. otiosus and P. regius.

Edwards (1980) stated that the two species, P. otiosus and P. regius, appeared
to be closely related, and further, that P. otiosus is assumed to be the parent
species because it is hypothesized that P. regius evolved as an isolated population
on islands formed when the Florida peninsula was partially submerged during

LU

a.

60

0.025 0.1 75 0.325 0.4.75 0.625 0.775 0.925

SIMILARITY

Fig. I. — Percentage of loci within a given range of genetic similarity values in the comparisons
among seven species of South Carolina Phidippus.

390

THE JOURNAL OF ARACHNOLOGY

Fig. 2. — Dendogram of South Carolina Phidip-
pus based on the unweighted pair group method
with arithmetic mean (UPGMA). Scale is in units
of genetic identity, F

otiosus

regius

princeps
whi tmani
clarus
audax
putnami

0.4 0.6 0.8 1.0

GENETIC IDENTITY

prehistoric times. He based these assumptions on morphological similarities, and
on field and laboratory observations. Morphologically, the basic structure of
body markings and the genitalia are similar in both species, but the color patterns
are different. The ranges of both species overlap greatly, and both species have
been observed sharing the same ecological niche. The most convincing evidence
presented by Edwards (1980), however, is the fact that hybridization attempts
between P. otiosus males X P. regius females were successful in the laboratory
and produced viable young. Edwards (1980) also collected apparent hybrids in
xeric woods habitat, and he recorded that a colleague. Dr. John F. Anderson,
collected a male P. otiosus in copula with a female P. regius, 19 September 1980
(Edwards 1980).

Our results support Edwards’ hypothesis that these two species are closely
related. From the paired comparisons of Table 4, these two species show the least
overall genetic divergence (D = 0.247) and the most overall genetic identity (I
= 0.781). Thus, about 24.7 electrophoretically detectable electromorphic
substitutes per 100 loci have occurred between P otiosus and P. regius as
compared to 84.6 between P princeps and P. putnami or compared to an average
of 56 substitutions per 100 loci in the evolution of the genus as a whole. These
results compare favorably with other surveys in which hybridizing species are
completely distinct in allelic composition at almost one third to one half of their
loci (Avise 1974).

These results are interesting when looked at in light of Edwards’ (1980)
discussion on hybridization in which he states that hybridization was successful
between P. otiosus and P. regius from a similar part of their range (northern
peninsular Florida), but unsuccessful when attempts to hybridize the same two
species were made with individuals from widely separated sources. A thorough
study of these two species throughout their ranges by isozyme analysis and
hybridization studies could reveal a number of significant factors concerning their
evolution.

Other similarities between the cladogram of Edwards (1980) and our
dendrogram exist. Edwards states that the courtship behavior of P. clarus has
more in common with the audax group than to its apparent morphological
relatives in the cardinalis group. The relationship of P. clarus to P. otiosus and
P. regius in our data would tend to substantiate this in that the latter two species
are classified by Edwards as belonging to the audax group. Our data, like
Edwards’, also show that P. clarus and P. whitmani are related. However, our

TERRANOVA & ROACH— GENETIC DIFFERENTIATION IN PHIDIPPUS

391

data differ from Edwards’ in that P. audax is more distantly related to the
previously named species. P. princeps is apparently more related to the P
witmani, P otiosus, P regius group than to P audax. This would concur with
Edwards’ premise that P. princeps may be related to the audax group. Our data
also differ from Edwards’ in that P. putnami is also distantly related to P. audax.

Overall, our results fit the classification of Edwards with the exception of P.
audax. By including other congeneric species, a more certain phylogeny is
possible for the genus Phidippus, as has been shown with other invertebrate
groups. In Drosophila for example, Nair et al. (1971), Yang et al. (1972),
Lakovaara et al. (1972) found agreement between similarity based on classical
systematic criteria, and those based on electrophoretic data. Avise (1974)
concluded that attempts to utilize electrophoretic data for classifying closely
related species would appear justified.

LITERATURE CITED

Avise, J. C. 1974. Systematic value of electrophoretic data. Syst. Zool., 23:465-481.

Avise, J. C., and F. J. Ayala. 1976. Genetic differentiation in speciose versus depauperate phylids:

evidence from the California minnows. Evolution, 30:46-58.

Ayala, F. J., M. L. Tracey, L. G. Barr, J. F. McDonald, and S. Perez-Salas. 1974a. Genetic variation
in natural populations of five Drosophila species and the hypothesis of the selective neutrality of
protein polymorphism. Genetics, 77:343-384.

Ayala, F. J., M. L. Tracey, D. Hedgecock, and R. C. Richmond. 1974b. Genetic differentiation during
the speciation process in Drosophila. Evolution, 28:576-592.

Berlocher, S. H. 1980. An electrophoretic key for distinguishing species of the genus Rhagoletis
(Diptera: Tephritidae) as larvae, pupae, or adults. Ann. Entomol. Soc. Amer., 73:131-137.

Edwards, G. B. 1980. Taxonomy, ethology, and ecology of Phidippus (Araneae: Salticidae) in eastern
North America. Ph.D. Dissertation, Univ. of Florida. 354 pp.

Ferguson, A. 1980. Biochemical Systematics and Evolution. John Wiley and Sons, New York. 194
PP-

Lakovaara, S., A. Saura, and K. Falk. 1972. Genetic distance and evolutionary relationships in the
Drosophila obscura group. Evolution, 26:177-184.

Nair, P. S., D. Banck, and K. Kojima. 1971. Isozyme variations and evolutionary relationships in the
Mesophragmmatica species group of Drosophila. Studies in genetics VI. Univ. Texas Publ.,
7103:17-28.

Nei, M. 1972, Genetic distance between populations. Amer. Natur., 106:283-292.

Roach, S. H., and G. B. Edwards. 1984. An annotated list of South Carolina Salticidae (Araneae).
Peckhamia, 2:49-57.

Sneath, P. H. A., and R. R. Sokal. 1973. Numerical Taxonomy. W. H. Freeman, San Francisco. 573
PP-

Terranova, A. C. 1978. A multifunctional electrophoresis system: specifications, construction, and
operation. USDA Publ., ARS-S-I79/ June. 61 pp.

Terranova, A. C. 1980. Diazo transparencies of polyacrylamide slab gels. Anal. Biochem., 107:443-
445.

Terranova, A. C. 1981a. A rotating temperature-controlled water bath for isozyme development in
polyacrylamide slab gels. Anal. Biochem., 112:232-235.

Terranova, A. C. 1981b. Polyacrylamide gel electrophoresis of Anthonomus grandis Boheman
proteins: profile of a standard boll weevil strain. USDA Publ., ARR-S-9/ July. 48 pp.

Terranova, A. C., and S. H. Roach. 1987. An electrophoretic key for distinguishing immature and
adult species of the genus Phidippus (Araneae: Salticidae) from South Carolina. Ann. Entomol.
Soc. Amer. (in press).

Yang, S. Y., L. L. Wheeler, and 1. R. Bock. 1972. Isozyme variation and phylogenetic relationships
in the Drosophila bipectinata species complex. Studies in genetics VIE Univ. Texas Publ.,
7203:213-227.

Manuscript received October 1985, revised April 1986.

■ ■ ■•*' , ■ ^ ’.V'-V

' " '

4^.^; ^ -

■■ 'P< ''ik, ,1tfei^4jtMr’^\

‘z '■■

l.Vfe:

-S’* # .

^■/4 ' ■■ii5^'*'4H

5.i«i^:.:‘:i „. ,v.f% j^'S S^ o'! ppfU^^M '9i

;■ -^'■?’'^lll^^. V 'i'- '■> '^""^1-0 ^ .Ui!i(|^|-Urr^.,?Tf.>^4

M'ilSSiJ^ V f ' ^ ^

^ H-

■', . ® -f. ... ■;,{ >>i/.Or: ;W,'a4j' ,'.' .'.T' • .'■ ..>?:» ■;,;(, ''.i- '\ji> VMf#irfil>-'i\

?'• ,>C> ' ^■^>i''!‘'r’:^ ' vi^r-' '. y^’'-; •' . }' ,*'«'* J-i-* j,\?j<||J''=!“iffi?'

. _ .

m

vW(#’»^«<^'''fi«' w!'S...,'iS|| ■ ■■ '■

C,4kS

.0

■^^•« •' wii-j •,*\.4?'A " ' ■ ;! ' :<^> i: r™ '*’ -■ ■"’^.' .m -PI-

-jTh-

msi>l'^>.' »^t#(«irtmh V «Wwif4«tt.iAS- ««<’4SkfcES^

(£^;p:^iu*:i3rr»7_. i>w .

f'^V-.-i, ' '-iK,i.«,li«:fl!iW..

:MXvS'ejli>f<Hf»

/Ai<

1987. The Journal of Arachnology 14:393

RESEARCH NOTES

PHIDIPPUS A UDAX (ARANEAE, SALTICIDAE) PREDATION
UPON A CICADA {TIBICEN SP.) (HOMOPTERA, CICADIDAE)

In Washington Co., Mississippi, at 1200 on 30 July 1985, an annual cicada
(Tibicen sp.) (Homoptera) was observed perching in a plum tree in a residential
area. This cicada was situated one meter above the ground, 0.5 m from the main
trunk on a branch ascending at a 45° angle. Ten cm further up the branch was
perched a penultimate instar female of Phidippus audax (Hentz) (Araneae,
Salticidae). When the cicada began to sing, the jumping spider immediately
changed the orientation of its body by 180 degrees so that it was facing the
cicada. The spider then approached until it was within 5 cm, paused ca. 5
seconds, then jumped upon the cicada. The cicada immediately stopped singing
and fell to the ground with the spider’s chelicerae attached to the frons region.
On the ground the spider remained attached and the cicada did not move for 5
minutes, whereupon both predator and prey were placed in a 95 mm diam, 65
mm high screen-clear plastic container and moved indoors for further
observations. At an ambient temperature of ca. 25° C and local photoperiod, the
spider in the next 5’/2 hours appeared to feed from seven different attachment
sites on the cicada as follows: 1200, frons between clypeus and right eye; 1430,
prothorax between dorso-ventral margin and right procoxa; 1445, prothorax at
rear edge of right operculum; 1500, abdomen between left ventral segments 4 and
5; 1530, abdomen at ventral midline of 2nd segment; 1545, prothorax between
metacoxae; 1600, prothorax at left dorso-ventral margin. The prey was finally
abandoned at 1730, with the spider positioning itself in an upper corner of the
container at the maximum possible distance from the cicada. Observations at 15
minute intervals until 2400 indicated a lack of spider movement. When
observations were resumed at 0600 the next morning, the spider was wandering
about the container but did not touch the dead cicada. This behavior continued
until 0800, when the spider was removed to another container.

The body length of the P. audax was 12 mm and the body length (excluding
wings) of the Tibicen sp. was 38 mm, indicating a predatoriprey size ratio of
1:3.2. The weight of the cicada after the spider had completed feeding was 0.77
g and the weight of the spider after feeding was 0.35 g, indicating a predator:prey
weight ratio of ca. 1:2.2.

We believe these observations to be important because of the paucity of
information both on the predators of cicada species and on the prey of P. audax.
The principal predators of adult cicadas are probably birds (e.g. Maier 1982),
though large orb-web species are probably also important (e.g. Fitch 1963). We
have found no records in the literature of Phidippus spp. predation on adult
cicadas (e.g. Edwards 1980). It is generally believed that salticids are limited to
prey smaller than themselves or to soft-bodied/ defenseless forms (Enders 1975).
In the genus Phidippus, this has been documented by field observations on P.
johnsoni (Jackson 1977) and by laboratory experiments with P. coccineus
(Gardner 1966). Phidippus audax may be an exception to this generalization,
however. In the field, this species has been observed attempting to catch flying

1987. The Journal of Arachnology 14:394

dragonflies considerably larger than itself (Fritch 1963), and has been observed
eating prey (cockroach, grasshopper, wasp) of much larger size (Bilsing 1920,
Edwards 1980). In laboratory experiments, prey larger than itself has been
consumed, though prey the same size as the spider seems to be preferred (Givens
1978, Freed 1984).

Robinson and Valerio (1977) describe a technique employed by several
neotropical salticids for capturing prey larger and/or stronger than themselves.
After jumping upon a large prey in an arboreal situation, the spider drops below
the substrate on its dragline and holds the prey with its chelicerae, letting the prey
hang below while envenomation or exhaustion occurs. This technique may have
been attempted by the P. audax individual described herein, but the cicada prey
may have been too heavy and was not supported by the dragline.

The observed predation by P. audax on a singing and stationary cicada leads
to the intriguing suggestion that the spider was able to orient to the prey by
receiving auditory and/or vibratory signals and that these signals were sufficient
to release the final attack on the prey. Prior to the start of our observations, the
spider may have seen the cicada fly to the branch site and begun to move toward
the site. When our observations began, the spider may have just arrived and
made a reorientation to the prey based on non-visual cues. Thus the initial
orientation to the prey may have been by visual detection of movement, thought
to be a requirement for eventual prey capture (Land 1971). There is sufficient
general evidence to suggest that the spider could have detected auditory and/or
vibratory signals from a singing cicada (Barth 1982), though the demonstration
of this specific phenomenon has yet to be recorded.

LITERATURE CITED

Barth, F. G. 1982. Spiders and vibratory signals: sensory reception and behavioral significance. Pp.
67-122, In Spider Communication: Mechanisms and Ecological Significance. (P. N. Witt and J. S.
Rovner, eds.). Princeton University Press, Princeton.

Bilsing, S. W. 1920. Quantitative studies in the food of spiders. Ohio J. Sci., 20:215-260.

Edwards, G. B., Jr. 1980. Taxonomy, ethology, and ecology of Phidippus (Araneae: Salticidae) in
eastern North America. Ph.D. Thesis, University of Florida, Gainesville. 354 pp.

Enders, F. 1975. The influence of hunting manner on prey size, particularly in spiders with long attack
distances (Araneidae, Linyphiidae, and Salticidae). Amer. Natur., 109:737-763.

Fitch, H. S. 1963. Spiders of the University of Kansas Natural History Reservation and Rockefeller
Experimental Tract. Univ. Kansas Mus. Nat. Hist., Misc. Publ., 33:1-202.

Freed, A. N. 1984. Foraging behavior in the jumping spider Phidippus audax: bases for selectivity.
J. Zook, Lond., 203:49-61.

Gardner, B. T. 1966. Hunger and characteristics of the prey in the hunting behavior of salticid spiders.
J. Comp. Physiol. Psychol., 62:475-478.

Givens, R. P. 1978. Dimorphic foraging strategies of a salticid spider {Phidippus audax). Ecology,
59:309-321.

Jackson, R. R. 1977. Prey of the jumping spider Phidippus johmoni (Araneae: Salticidae). J.
Arachnok, 5:145-149.

Maier, C. T. 1982. Observations on the seventeen-year periodical cicada, Magicicada septendecim
(Hemiptera: Homoptera: Cicadidae). Ann. Entomok Soc. Amer., 75:14-23.

Robinson, M. H., and C. E. Valerio. 1977. Attacks on large or heavily defended prey by tropical
salticid spiders. Psyche, 84:1-10.

Timothy C. Lockley and Orrey P. Young. P.O. Box 3, Stoneville, Mississippi
38776.

Manuscript received November 1985, revised February 1986.

1987. The Journal of Arachnology 14:395

PREY OF THE STRIPED LYNX SPIDER
OXYOPES SALTICUS (ARANEAE, OXYOPIDAE),

ON COTTON IN THE DELTA AREA OF MISSISSIPPI

Spiders have received considerable attention as potentially important predators
of arthropod pests in agroecosystems (Riechert and Lockley 1984). Much of this
attention has been focused on the striped lynx spider, Oxyopes salticus Hentz
(Young and Lockley 1985). This species is particularly abundant in cotton, to the
extent that it is often the most abundant beneficial predator (e.g., Faster and
Brazzel 1968, Dean et al. 1982). The known prey of O. salticus in cotton include
such economically important pests as the tobacco budworm and bollworm
(Heliothis spp.) (Whitcomb 1967, McDaniel et al. 1981), and the mirid plant bugs
Lygus lineolaris (P. de B.) (tarnished plant bug) (Whitcomb and Bell 1964),
Adelphocoris rapidus (Say) (rapid plant bug) (Kagan 1943), and Pseudatomosce-
lis seriatus (Reuter) (cotton fleahopper) (Almand 1974). A survey of the literature
has made it quite clear, however, that very little is known about the prey of O.
salticus in cotton, particularly the proportion of harmful and beneficial insects in
its diet (Young and Lockley 1985).

Observations and collections were conducted on a farm in Sunflower County,
Mississippi, during June-July, 1983. A 16.4 ha field planted in “Stoneville 213”
cotton was subdivided into forty 0.41 ha plots as a part of a long-term
experiment involving aldicarb treatments (Scott et al. 1985). Eight untreated
check plots were randomly distributed in this field and were the sites for the
research reported herein.

Field observations were conducted one day each week for five consecutive
weeks during the hours of ca. 0730-1030, a period of maximal O. salticus activity
(pers. obs.). Cotton plants were examined until an individual O. salticus was
discovered. It was then observed (< 15 minutes) until a prey was captured,
whereupon the spider was disturbed and forced to release its prey. Field
identification of the prey was then attempted, but if unsuccessful the specimen
was brought back to the lab for examination under magnification. Approximate
body length measurements of the predator and prey were obtained in the field.

During 11 h 15 min of field observations over a five week period, 48 O. salticus
were detected capturing or already in possession of prey (Table 1). These prey
included adult and/or immature forms of 14 insect species in five orders. Eight
of these species represent new feeding records for O. salticus, and significantly
increase the previous total of 33 species of prey (Young and Lockley 1985).
Seventy-one percent of the 48 prey items were in the orders Hemiptera and
Homoptera, with nymphs of the tarnished plant bug representing the most
frequently captured prey (35%). Of the 14 species of prey, five can be considered
beneficial, eight harmful, and one {Aedes sp.) neutral. The five beneficial species,
however, represent only 10% of the prey items. Thus O. salticus may have
minimal direct impact on the beneficial arthropod population.

The mean body lengths of each instar of O. salticus ranged from 3.1 to 6.0
mm with a mean sample length of 4.85 mm. Eighty-one percent of the spiders
were > 4.3 mm and 7th instar or older. The body length of the prey ranged from
2 to 7 mm with a mean of 4.19 mm. Twenty-one percent of the prey were >
6.0 mm, the maximum size of the predator. Although the frequency distribution
of the various sizes of prey is unknown, these data indicate that most captured

1987. The Journal of Arachnology 14:396

Table 1. — Prey of Oxyopes salticus observed in cotton, Sunflower County, Mississippi. H =
harmful, B = beneficial, N = neutral.

Taxon

Prey Stage

No. of

Observations

Length of
Prey (mm)

Economic

Status

Orthoptera

Tettigoniidae

Neoconocephalus sp.

nymph

2

5,6

H

Hemiptera

Anthocoridae

Orius insidiosus (Say)

adult

1

3

B

Lygaeidae

Geocoris punctipes (Say)

nymph

1

3

B

Miridae

Lygus lineolaris (P. de B.)

nymph

17

3 (8 obs.).

H

adult

2

4 (7 obs.), 5, 6
6

H

Pseudatomoscelis seriatus (Reuter)

nymph

1

3

H

" "

adult

3

4

H

Pentatomidae

Podisus maculiventris (Say)

nymph

1

2

B

Homoptera

Cicadellidae

Chlorotettix viridus (V.D.)

nymph

2

4

H

" "

adult

1

6

H

Empoasca fabae (Harris)

nymph

3

3

H

" "

adult

1

5

H

Graphocephala versuta (say)

adult

1

6

H

Lepidoptera

Noctuidae

Trichoplusia ni (Hubner)

larva

2

6,7

H

Spodoptera ornithogalli (Guenee)

larva

1

6

H

Diptera

Culicidae

Aedes sp.

adult

7

3 (3 obs.).

N

Dolichopodidae

Condylostylus sp.

adult

1

5 (4 obs.)

4

B

Syrphidae

Syritta pipiens L.

adult

1

5

B

prey were about the same length as the predator. This may be a result, however,
of the preponderance of relatively large O. salticus in the sample. Previous
research has indicated that small and young O. salticus may have difficulty
capturing prey as large or larger than itself, but that large and mature O. salticus
are quite capable of obtaining as large/ larger prey (Young and Lockley 1986).

The absence of evidence for O. salticus predation on conspecifics or any other
spider is noteworthy. Most spiders will feed on almost any suitable-sized animal
that they may encounter, including spiders (Turnbull 1960). Studies of ground-
foraging spiders in agricultural situations have demonstrated substantial predation
on spiders, including cannibalism (Edgar 1969, Kiritani et al. 1972). Since almost
all spiders can be considered beneficial in that they frequently feed on pest
species, a low predation rate on spiders by O. salticus is a very significant
characteristic for the most abundant spider in a cotton field.

1987. The Journal of Arachnology 14:397

LITERATURE CITED

Almand, L. K. 1974. Seasonal abundance, dispersal, and control of the cotton fleahopper on certain
host plants. Ph.D. Thesis, Texas A&M Univ., College Station. 67 pp.

Dean, D. A., W. L. Sterling, and N. V. Horner. 1982. Spiders in eastern Texas cotton fields. J.
ArachnoL, 10:251-260,

Edgar, W. D. 1969. Prey and predators of the wolf spider Lycosa luguhris. J. Zool., 159:405-41 1.

Kagan, M. 1943. The Araneida found on cotton in central Texas, Ann, Entomol. Soc. Amer., 36:257-

258.

Kiritani, K., S. Kawahara, T. Sasaba, and F. Nakasuji. 1972. Quantitative evaluation of predation
by spiders on the green rice leafhopper, Nephotettix cincticeps, by a sight-count method. Res.
Popul. Ecol., 13:187-200.

Faster, M. L., and J. R. Brazzel. 1968. A comparison of predator populations in cotton under
different control programs in Mississippi. J. Econ. Entomol., 61:714-719.

McDaniel, S. G., W. L, Sterling, and D. A. Dean. 1981. Predators of tobacco budworm larvae in
Texas cotton Southw. Entomol, 6:102-108.

Riechert, S. E., and T. C. Lockley. 1984. Spiders as biological control agents. Ann. Rev. Entomol,

29:299-320.

Scott, W. R, J. W. Smith, and G. L. Snodgrass. 1985. Response of cotton arthropod populations
in cotton to various dosages of aldicarb applied in the furrow at planting time. J. Econ. Entomol,

78:249-257.

Turnbull, A. L. 1960. The prey of the spider Linyphia triangularis (Clerck) (Araneae, Linyphiidae).
Canadian J. Zool, 38:859-873.

Whitcomb, W. H. 1967. Field studies of predators of the second-instar bollworm, Heliothis zea
(Boddie) (Lepidoptera: Noctuidae). J. Georgia Entomol. Soc., 2:113-118.

Whitcomb, W, H,, and K. Bell 1964. Predaceous insects, spiders, and mites of Arkansas cotton fields.
Arkansas Agric. Exp. Sta., Bull, 690:1-84,

Young, O. R, and T, C. Lockley. 1985. The striped lynx spider, Oxyopes salticus (Araneae:

Oxyopidae), in agroecosystems. Entomophaga, 30:329-346.

Young, O. R and T, C, Lockley. 1986. Predation of striped lynx spider, Oxyopes salticus (Araneae:
Oxyopidae), on tarnished plant bug, Lygus lineolaris (Heteroptera: Miridae): a laboratory
evaluation. Ann. Entomol Soc. America,, 79:879-883.

Timothy C. Lockley and Orrey P. Young. P.O. Box 3, Stoneville, Mississippi

38776.

Manuscript received December 1985, revised February 1986.

1987. The Journal of Arachnology 14:398

MATERNAL BEHAVIOR IN APOLLOPHANES PUNCTIPES
(O. PICKARD-CAMBRIDGE) (ARANEAE, THOMISIDAE)

Maternal defense of eggs has been reported in a number of spider families (see
reviews by Burgess, J. W. 1978. Symp. ZooL Soc. Lond., 42:69-78, and Buskirk
R. E. 1981. in Hermann, H. (ed.) Social Insects. Vol. 2, Academic Press, New
York). Buskirk cites a study showing that the thomisid Philodromus caespiticolis
defends its egg sacs, but to my knowledge the present note is the first report of
maternal behavior in the closely related genus Apollophanes.

A female A. punctipes was found on 13 Feb. 1983 (first half of the dry season)
resting on an egg sac which was on the inner surface of one of three dried leaves
which had been fastened to a twig with silk lines attached near their petioles. The
twig was about 1.5 m above the ground on a Anona cherimolia tree in a grassy
yard near San Antonio de Escazu, Costa Rica (el. 1300m). The white silk of the
sac formed a relatively thin wall through which the outlines of individual eggs
were clearly distinguishable. The leaf with the egg sac was also fastened loosely
with silk near its midpoint to the adjacent leaf.

After partially separating the leaves, I dropped a small ant onto the sac, and
the spider immediately pounced on it. Seizing it in her cheliceae, she ran to the
lower leaf surface below the sac and dropped the ant to the ground. Ten more
essentially identical performances were elicited in the next 30 minutes by
dropping as many other ants onto the sac.

Several times after dropping the ant, the spider remained on the leaf at or just
below the lower edge of the sac with only her rear legs on the sac. When I
dropped the ant onto the upper part of the sac where it did not touch the spider
and was probably out of her sight because of the curve of the leaf, the spider
nevertheless responded immediately, suggesting that she sensed the ant via
vibrations of the silk.

There are several interesting parallels between the maternal behavior of A.
punctipes and that of the salticid Lyssomanes jemineus (Eberhard, W. G. 1974.
Bull. Br. Arachnol. Soc. 3:51.). Females of both species lay eggs at relatively
exposed, aerial sites, and both defend them against ants by causing the ants to
fall — the salticid employing quick flicks of its front legs, and the thomisid by use
of its chelicerae. Both also make reduced egg sacs — that of the thomisid having
relatively thin walls and that of the salticid being almost completely eliminated.

I thank J. H. Redner and C. D. Dondale for identifying the spider, and the
Vicerectoria de Investigacion of the Universidad de Costa Rica for financial
support.

William G. Eberhard. Smithsonian Tropical Research Institute, and Escuela de
Biologia, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica.

Manuscript received November 1985, revised February 1986.

1987. The Journal of Arachnology 14:399

FOSSIL EVIDENCE OF SPIDER PARASITISM
BY ICHNEUMONIDAE

A piece of amber containing a spider and associated insect larva was sent to
the author by Mr. Jake Brodzinsky of Santo Domingo. The amber originated
from the Dominican Republic, was golden brown, transparent and measured 10
by 5 mm. The insect larva, which was located just anterior to the spider’s
abdomen, appeared to represent a first instar ichneumonid. The body of the
spider was 2.2 mm long and the curved legs extended another 1.4 mm on either
side of the body. The adjacent insect larva was 0.29 mm long and 0.04 mm wide.
It was 0.16 mm from the spider’s abdomen. It possessed a distinct head capsule,
12 recognizable abdominal segments and a smooth integument. The spider was
identified as an immature male Clubionidae.

Members of the tribe Polysphinctini (Ichneumonidae: Hymenoptera) are
contemporary parasites of spiders or predators on spider eggs. The fossil form
would fall into the former category. According to Nielsen, (1923. Entomol.
Meddel., 14:137-205; 1935. Entomol. Meddel., 19:191-215) who accumulated a
considerable amount of information on these “ichneumonid flys”, the early instar
larvae can remain for months on a spider in an inactive state. Most spider-
parasitic Polysphinctini first sting the spider host into unconsciousness, then
deposit their egg dorsally or laterally at the base of the spider’s abdomen or on
the posterior declivity of the cephalothorax (in the fossil form, the parasite is
found just adjacent to the base of the spider’s abdomen in a dorsal-lateral
position). The eggs are attached by a mucilaginous material and the hatched larva
sits exposed on the abdomen of its spider host and periodically feeds by
puncturing the spider’s integument with its mandibles and sucking up the
extruded hemolymph. First instar larvae of polysphinctinids are described by
Nielsen (1923, 1935) as having a large, sclerotized head, 13 body segments of
diminishing width and a bare or minutely-spined integument. First instar larvae
of representatives of Polysphinctini that were in the collection at Berkeley

Figs. 1-2 — Fossil first stage larva of a member of the Polysphinctini in Dominican amber:
(Magnification 20X). 1, position of the larva (arrow) in relation to the spider; 2, general shape of
larva showing distinct head capsule (arrow) (Magnification SOX).

1987. The Journal of Arachnology 14:400

measured 0.45 mm in length. However, the tribe is composed of several genera
and a size variation of first stage larvae would be expected.

On the basis of size, morphology and location near the base of the spider’s
abdomen, it is concluded that the fossil larva is a member of the tribe
Polysphinctini and that spider parasitism by these ichneumonids was well
established some 20-40 million years ago. This is roughly the age of Dominican
amber (Lambert et al., 1985. Archaeometry, 27:43-51). Fossil associations
showing evidence of parasitism, such as this one, are rare in the geological record.
I thank Dr. E. Schlinger for identifying the fossil spider and providing me with
present-day larvae of Polysphinctini for comparative studies.

George O. Poinar, Jr. Department of Entomological Sciences, University of
California, Berkeley, California 94720.

Manuscript received January 1986, revised March 1986.

1987. The Journal of Arachnology 14:401

X-RAY STERILIZATION OF MALE GOLDEN-WEB SPIDERS
NEPHILA CLAVIPES (ARA^EAE)

In some spider species, females mate with more than one male (Austad 1984).
Male sterilization can be used to determine whether a particular male fertilizes a
disproportionate number of a given female’s eggs (Parker 1970). A popular
method requires some females to mate first with an intact male and second with a
sterile; other females are presented males in the reverse order. Differential rates of
egg hatching between these two groups reveal patterns of male priority in
fertilization. There are, however, only a few published papers on irradiation and
male sterility in spiders (Jackson 1980, Vollrath 1980, Austad 1982). Austad
(1982) presented a relatively detailed methodology which we used as a guide.
Because Austad’s bowl and doily spiders {Frontinella pyramitela (Walck.)) and
our Nephila clavipes (L.) differ in size and aspects of natural history, a
presentation of our methods and findings might be useful to others contemplating
sterilization procedures.

This study was done during the summer of 1984 at the F. Edward Hebert
Center of Tulane University, about 15 km south of New Orleans. Subjects were
selected in mid-July, early in the mating season, because early season eggs are
more likely to hatch than later eggs (Christenson, Wenzl and Legum 1979). Sixty
adult males were collected from small orbs (probably of their own construction)
that contained sperm webs. This insured that males were in the same phase of
spermatogenesis at the time of irradiation. Light brown abdomen coloration and
the presence of the sperm web indicated that collections were made within three
days after the final molt.

After collection, males were refrigerated until inactive and then placed singly in
a 1 cm cardboard cube. Twenty received 2000 rad of X-ray radiation, 20 received
5000 rad, and 20 served as controls, being treated in the same manner as
experimentals except for irradiation. Radiation was generated by a General
Electric Maximar® machine at 250 kv through a filter of 1/4 mm copper and 1
mm aluminum situated 39 cm above the center of the holding box. Output was
129 rad/ min. These specifications are based upon those that proved effective for
Austad (1982).

Males were placed individually with a female collected in her penultimate
instar. Subjects were housed in Fiberglas screen boxes (123 X 62 X 62 cm). We
placed males with females prior to the female’s final molt since unrestrained N.
clavipes females mate just after the final molt (Christenson, Brown, Wenzl, Hill,
and Goist 1985). However, in this species, irradiated males could have been
mated to virgin adults since adult females will mate while feeding and such
matings result in egg hatching (Wenzl 1980). Males were removed 25 days after
irradiation.

When assessing fertilization priority between two males, it is necessary to
demonstrate that both males actually mated with the female. Or, if one wished to
study eggs fertilized by irradiated sperm, it would be necessary to observe mating
to insure that fertilization had occurred since unfertilized eggs and eggs fertilized
by inviable sperm are similar in appearance. To document mating, we observed
subjects during a morning census and several 1-min time samples taken
throughout the day. Mating and egg data presented in this paper are based upon

1987. The Journal of Arachnology 14:402

the 38 pairs that were observed to mate and in which the female lived long
enough to produce an egg clutch.

Females laid eggs in the field boxes, with the first appearing about four weeks
after the final molt. Eggs were left in the boxes until hatching since laboratory
conditions tend to reduce hatching success (unpublished observations). Fourteen
of the controls, 13 of the 2 krad group, and 11 of the 5 krad group produced a
clutch of eggs. Eggs of thirteen of the control sacs hatched compared to none of
the 2-krad or 5-krad clutches = 33.874, df = 2, p < 0.0001). Eggs of the three
groups were of a normal yellow coloration and could not be distinguished by the
unaided eye. Eggs fertilized by control males {n — 2 clutches) and irradiated
males (/7 = 3 clutches) were examined microscopically. Group differentiation was
possible about one week prior to hatching, with developmental details obvious on
the surface of the eggs fertilized by sperm of control males.

Irradiation did not seem to affect the positioning of males on the female web
nor the vigor of overt male sexual behavior. A male who appears on a
unrestrained female’s web will position himself near the hub; if two or more are
present, usually the largest male will defend this position (Christenson and Goist
1979). All males but one (a 2-krad animal) had assumed position near the hub by
the morning after introduction. The remaining male did so by the following day.
Groups did not statistically differ in terms of males who mated the day the female
molted; all but one control, one 5-krad, and three 2-krad males were observed
mating (T^ = 1.856, df = 2, p < 0.395). All males were observed to mate for the
first time within two days of the female’s final molt. All initial mating occurred
within 15 days after irradiation.

As Austad (1982) notes, to assess the effectiveness of sterilization procedures
used to examine sperm competition and determine paternity, it is necessary to
note if the irradiated sperm are motile. To assess motility, we examined
microscopically sperm for six females (two controls, two 2-krad, and two 5-krad)
26 days after males were irradiated and at least two weeks after mating had
occurred. Collection was made at this time since Nephila sperm do not become
flagellate (and motile) in the spermathecae until about ten days after mating
(Brown 1985). Spermathecae were removed from the abdomen, placed on a clean
slide in an uncontrolled drop of saline, cut with a razor, the halves squashed, and
sperm immediately examined under a Leitz® phase-contrast microscope. All six
sets of spermathecae contained many flagellate, motile sperm. Inviability of eggs
of females in the irradiated groups was not due to sperm being aflagellate since
oviposition occurred an average of 27 days after copulation, when most sperm are
still motile (Brown 1985).

Irradiation influences the lifespan of sperm. Five pairs of spermathecae were
removed and examined sixty days after irradiation. One control female and two
2-krad females contained many flagellate sperm, while two 5-krad females
contained mostly dead sperm. This presented no problem in N. clavipes as
fertilization occurs at oviposition, about 27 days after final molt.

Irradiation did not influence male longevity. Sixteen control males, 17 2-krad
males, and 16 5-krad males (x^ = 0.223, df = 2, p < 0.895) were alive twenty five
days after irradiation. To more fully assess effects on longevity, fifteen males were
placed with a second female, after the death of the original female with which
they were housed. Males residing with an adult female live longer than those
which are not (unpublished observations). Eight controls lived an average of 52.6

1987. The Journal of Arachnology 14:403

(s.d. = 13.6) days and seven 5-krad animals 46.9 (s.d. = 9.4) days (F 1,13 =

0. 887, /7 > 0.10).

We do not know exactly how long the initial radiation treatment affects sperm
production. -As our males were housed with the female for 25 days, they had
ample opportunity to mate more than once, first after the final molt, and during
subsequent days while the adult female fed. If an irradiated male eventually
produced viable sperm, then it is possible that at least some of the eggs of a
second female sac would hatch. Thirty three of our females produced a second
clutch. Of the 12 second control clutches, 12 second 2-krad clutches, and 9 second
5-krad clutches, 11 (91.7%), 0%, and 0% hatched, respectively (T^ = 28.875, df —

1, p < 0.0001). Our initial radiation treatment was effective for at least 25 days
which is comparable to “at least 10 days” noted by Jackson (1980) for the
jumping spider Phidippus johnsoni. To assess longevity of irradiation effects,
further data are needed on the normal course of spermatogenesis in this species
and egg productivity of females mated with one irradiated male.

We thank L. E. LaChance of the USD A Metabolism and Radiation Research
Laboratory and J. W. Haynes of the USDA Boll Weevil Research Laboratory for
helpful comments. This research was supported by NSF grant BNS-83 17988 to
TC.

LITERATURE CITED

Austad, S. N. 1982. First male sperm priority in the bowl and doily spider Frontinella pyramitela.
Evolution, 36:777-785.

Austad, S. N. 1984. Evolution of sperm priority patterns in spiders. In Sperm competition and the
evolution of animal mating systems (R. L. Smith, ed.). Academic Press, New York.

Brown, S. G. 1985. Mating behavior of the golden-orb weaving spider, Nephila clavipes: II. Sperm
capacitation, sperm competition, and fecundity. J. Comp. Psychol., 99:167-175.

Christenson, T. E., S. G. Brown, P. A. Wenzl, E. M. Hill, and K. C. Goist. 1985. Mating behavior of
the golden-orb weaving spider, Nephila clavipes: I. Female receptivity and male courtship. J.
Comp. Psychol., 99:160-166.

Christenson, T. E. and K. C. Goist. 1979. Costs and benefits of male-male competition in the orb
weaving spider, Nephila clavipes. Behav. Ecol. Sociobiol., 5:87-92.

Christenson, T. E., P. A. Wenzl, and P. Legum. 1979. Seasonal variation in egg hatching and certain
egg parameters of the golden silk spider Nephila clavipes. Psyche, 86:137-147.

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

Parker, G. A. 1970. Sperm competition and its evolutionary consequences in the insects. Biol. Rev.,
45:525-567.

Vollrath, F. 1980. Male body size and fitness in the web-building spider Nephila clavipes. Zeit.
Tierpsychol., 53:61-78.

Wenzl, P. A. 1980. Mating behavior and egg production in the orb weaving spider, Nephila clavipes.
Unpublished master’s thesis, Tulane University.

Terry Christenson, Department of Psychology, Tulane University, New
Orleans, Louisiana 70118; Joseph Schlosser, M.D., 1521 St. Charles, New
Orleans, Louisiana 70130; and Jeffrey Cohn, and Leann Myers, Department of
Psychology, Tulane University, New Orleans, Louisiana 70118.

Manuscript received June 1985, revised April 1986.

1987. The Journal of Arachnology 14:404

BOOK REVIEWS

Barth, F. G. (Ed.) 1985. Neurobiology of Arachnids. Springer-Verlag, New
York. 385 pp. 174 textfigs. (price $69.50).

In this first monograph on the subject of arachnid neurobiology, most of the
coverage (approximately 70 percent) concerns what has been the primary interest
of the majority of workers in the field — sensory systems and behaviors mediated
by the senses. Nevertheless, the reviews of other aspects of neurobiology
sufficiently round out the book and broaden its usefulness.

The central nervous system of arachnids has attracted the attention of relatively
few investigators, compared to that of insects and crustaceans. It is treated here
in chapters on anatomy (Babu) and ontogeny (Weygoldt), followed by one on
neurosecretion (Legendre).

In the lengthiest sections of the book, early chapters on vision in spiders —
morphology and optics (Land), photoreceptor functioning (Blest), and spectral
sensitivity (Yamashita) — are complemented later by a study of visual discrimina-
tion in salticids (L. Forster) and two examinations of visual and idiothetic
orientation in agelenids (Corner and Claas; Mittelstaedt). Likewise, coverage of
mechanoreceptors — tactile hairs and spines, and scopula hairs (Foelix); trichobo-
thria (Reissland and Corner); and slit sensilla (Barth) — provides a basis for
understanding later chapters on the diverse functions of the spider’s vibration
sense (Barth) and the role of proprioception in mediating locomotion (Seyfarth).
An additional aspect of the sensory system, the occurrence of peripheral synapses
in arachnids, is presented by its discoverer (Foelix).

Much of the relatively limited information on arachnid motor systems is
summarized in a chapter that includes the author’s (Sherman) special area, neural
control of the heartbeat. How motor output is affected by sensory and central
nervous system mechanisms to yield locomotion in scorpions is the topic of the
following chapter (Root). The book’s final paper (Fleissner and Fleissner)
presents an analysis of the neurobiological clock underlying the circadian rhythm
of visual sensitivity in scorpions.

While all of the chapters provide useful reviews of their topics, many also
include especially large sections devoted to very recent research or contain
heretofore unpublished findings. Noteworthy here are the reviews by Blest on the
fine structure of spider photoreceptors, by Yamashita on the spectral sensitivity
of spider photoreceptors, by Seyfarth on spider proprioception, by Mittelstaedt
on spider navigation, and by the Fleissners on circadian clocks in scorpions.
Although some of the content of a few chapters (the two by Barth and the one
by L. Forster) can be found in previous contributions by their authors to recent
books on spider communication and on sense organs, nowhere else can one find
the wealth of information and completeness of coverage of the major topics of
arachnid neurobiology as in this excellent monograph that Barth has brought
together. The availability of this publication should increase the possibility that
arachnids will be brought into the laboratories of some of the other
neurobiologists, who so far have generally ignored this class of arthropods.

1987. The Journal of Arachnology 14:405

Unlike those many monographs which suffer from publication delays, this one
is as up to date as possible, with some of the contributions containing references
to papers published in 1985. The book is handsomely illustrated, with clear
diagrams and high-quality electron micrographs. Typographical errors are
minimal and not distracting, reflecting the carefulness and language skills of the
editor. Indeed, although he and about half of the contributors are German, all
of the chapters are written clearly in English, reflecting Barth’s desire to reach
as broad an audience as possible.

Jerome. S. Rovner, Zoological Sciences, Ohio University, Athens, Ohio, 45701.

Eberhard, W. G. 1985. Sexual Selection and Animal Genitalia. Harvard
University Press, Cambridge, Massachusetts and London, England. 244 pp. (Price
$25.00).

Why are genitalia, particularly male genitalia, so useful as taxonomic
characters for distinguishing closely related species? Most of us probably assume
that animals evolve genital differences to prevent interspecific mating. Or put
another way, animals other than arachnologists also use genitalia for species
identification. W. G. Eberhard’s provocative new book contends that this
traditional interpretation is erroneous, and elaborates a new hypothesis, namely
that male genitalia have evolved so rapidly and divergently in response to female
mate preference among males of the same species.

In form, the book is an extended argument. Chapter 1 documents the
phenomenon. Male intromittant organs are useful in species-level taxonomy in
animals ranging from arthropods, planarians and nematodes to sharks, snakes
and rodents. Rapid evolution of male genitalia may therefore be one of the most
ubiquitous patterns in nature. Chapters 2-4 consider several previous hypotheses,
including the species identification hypothesis, and explains why none of these
hypotheses adequately explains observed patterns of genital diversity.

Chapter 5 outlines Eberhard’s female choice hypothesis. Briefly, the hypothesis
assumes that females prefer as mates certain conspecific males on the basis of
some genital trait. Even if there is no other adaptive advantage associated with
this trait, the fact that females prefer males exhibiting the trait gives those males
a fitness advantage, and the trait will spread. If females prefer a more elaborate
form of the trait, then the trait will become more and more elaborate until its
advantage relative to mating is balanced by its disadvantage relative to survival.
A key feature of this argument is that females will prefer traits which are
arbitrary with respect to other aspects of male fitness. Thus male genitalia can
be expected to evolve the elaborate and bizarre morphology with which
arachnologists are so familiar.

Chapters 6-11 consider arguments and evidence for and against Eberhard’s
hypothesis. Among other things, he demonstrates convincingly that numerous
mechanisms exist by which females might influence which of several mates
actually succeed in fertilizing her eggs — a necessity for the theory to work. Both
in these, and previous, chapters Eberhard uses his extensive knowledge of the
arachnid literature to support his arguments. Of more than 700 references cited,
nearly 100 pertain to arachnids.

1987. The Journal of Arachnology 14:406

Perhaps the most impressive aspect of this book, aside from the exciting new
idea presented, is the creativity with which Eberhard milks the existing literature
for information bearing on his, and competing, hypotheses. For instance, he
reasons that if genitalia diverged because of their use in species identification,
then species which have little opportunity to make species identification errors
would be expected to show less divergence than species which frequently contact
close relatives. He tests this prediction by examing divergence and elaboration of
genitalia in geographically isolated species (island endemics and host-specific
parasites, for which hosts are habitat islands) compared with divergence in
nonisolated near relatives.

In compiling his arguments, Eberhard by necessity also compiles a fascinating
survey of the diversity of animal mating mechanisms. We learn of such
phenomena as hypodermic insemination (males insert their genitalia through the
female body wall and deposit sperm in the body cavity), exploding spermato-
phores, and genital scoops for removing females’ stored sperm. Therefore, even
if the book weren’t so well written, so excellently illustrated, and so imaginatively
conceived, it would still be worth owning.

Steven N. Austad, Department of Organismic and Evolutionary Biology,
Harvard University, The Biological Laboratories, Cambridge, Massachusetts
02138.

NOMENCLATURE NOTES

Opinion 1394. — The International Commission on Zoological Nomenclature
ruled under its plenary powers to conserve the names Centrums limpidus Karsch,
1879 and Centruroides ornatus Pocock, 1902 (Arachnida, Scorpiones) (Bull. Zool.
Nomencl., 43(2): 144).

The ICZN also published an application (Bull. Zool. Nomencl., vol. 43, part
4) to confirm Thomisus hirtus Latreille 1819 as the type species for Heriaeus
Simon, 1875 (Arachnida, Araneae). Comment or advice is welcomed by the
Commission and should be sent to The British Museum (Natural History),
London, England. Please refer to case no. 2447.

407

THE JOURNAL OF ARACHNOLOGY

Instructions to Authors

GENERAL COMMENTS

Manuscripts are acceptable in English, French, Portuguese, or Spanish, and must be TYPED
DOUBLE- OR TRIPLE-SPACED THROUGHOUT. Good quality paper should be used and erasable
bond is specifically excluded. Leave margins at least 1.5 in. (4 cm) on the left and 1 in. (2.5 cm)
on the top, bottom, and right. Do not hyphenate any words at the right margin. Avoid using right
justification in manuscripts prepared on word processors. Do not edit your own manuscript. Italics
are permitted only to indicate scientific names; only the TITLE, PRIMARY HEADINGS (e.g.,
INTRODUCTION, etc.) and RUNNING HEADS should be typed in capital letters. Follow the
recommendations of Council of Biological Editors (CBE) Style Manual, Third Edition, unless
indicated otherwise below. To facilitate prompt review by two or more referees, send to the Associate
Editor three identical copies of the typed material, one of which should be designated as the
“original,” together with copies of the illustrations.

Manuscripts longer than about 1500 words (five double-spaced typewritten pages in Elite, six
double-spaced typewritten pages in Pica) should be prepared as Feature Articles, shorter papers as
Research Notes. Manuscripts and all related correspondence must be sent to the Associate Editor:
Dr. Jerome S. Rovner, Department of Zoological Sciences, Ohio University, Irvine Hall, Athens,
Ohio 45701.

FEATURE ARTICLES

Arrange the various part of feature articles in the following sequence: (1) mailing address, (2) title,
(3) by-line, (4) abstract, (5) body of text, (6) acknowledgments, (7) literature cited, (8) figure legends,
(9) footnotes, (10) running head, (11) tables with legends, (12) figures. Put only parts 1, 2 and 3 on
page 1, and start page 2 with part 4. All other pages are numbered consecutively.

Mailing address. — Include the complete address and telephone number of that author with whom
all correspondence with the Editors should be handled.

Title. — When using common or scientific names in the title, be sure to include in parentheses the
appropriate higher taxa (order, family, . . .). Include footnote indication if appropriate (e.g., to
acknowledge grant support), but type footnote separately (see part 9 above). The title should be typed
in capital letters and may not exceed 48 characters and spaces per line.

By-line. — Include the name(s) of author(s) as customarily written (less titles) and complete
address(es), including zip code or other postal designation. Include footnote indication(s) if
appropriate (e.g., to indicate change of address), but type footnote separately (see part 9 above).
Leave three spaces between title and by-line, and four spaces between name(s) and address(es).

Abstract. — The abstract should be a summary of the basic findings, and should not exceed two to
three percent of the text in length. Papers in a language other than English must be accompanied by
an English abstract as well as an abstract in the language of the text. Papers written in English but
focusing on a subject pertaining to a country where another of the accepted languages is used, may
include an abstract in that language.

Body of text. — Use whatever form seems best to accommodate the necessary description of
research. Be concise. Conform to standard references for abbreviations, i.e. the CBE Style Manual
for scientific abbreviations, and a good language reference such as The Little, Brown Handbook for
abbreviations of English words. Use the metric system for all measurements (the English system is
acceptable only when transcribing locality data accompanying museum specimens), and note that
abbreviations of metric units of measurements are not punctuated (e.g., mm and km, but ft. and mi.).
Citations should be in the following form: Bellrose (1950); Bellrose (1950:33); or (Bellrose 1950). The
complete scientific name of a species or genus, including author(s), must be given the first time they
are mentioned in the text. Use single-line notation for fractions [e.g., 1/4 and not %; (4-12)/3 and
not^].

The following special directions apply to authors of taxonomic papers:

(a). Do not use abbreviations in a primary heading to indicate that a new name or a new combination
is being proposed (e.g., A-us x-us, new species, rather than A-us x-us, n. sp. or comparable
abbreviations).

408

(b) . Keys must be typed as follows:

1. Use Arabic numerals to designate the leading entry of the couplet 2

Do not designate the second entry of a couplet, either by means of numbers, letters or other
marks A-us x-us

2. Type numbers flush to left margin, and start entry on fifth space .A-us y-us

Subsequent lines of any entry must also be idented five spaces .A-us z-us

(c) . Synonymies must follow the abbreviated style shown below:

A-us x-us Jones 1930:3, 1935:9; Russell 1945:453; Smith 1954a: 16, 1954b:678; Cooper and Lim
1955:18 (in part).

A-us y-us Bates 1932:18, fig. 4. NEW SYNONYMY.

A-us z-us: Miranda 1948:98 (misidentification); Harris 1951:3 (in part?), {nec A-us z-us
Zimmer).

(d) . Lists of specimens examined of a given taxon must be the last items typed in the treatment of
that taxon as they will be set in smaller type. Adhere to the following style for listing specimens
examined: Country: State or comparable political subdivision; County or District, detailed locality
(elevation), 14 July 1945 (collector): 2 males, 5 females (acronym of institution where specimens are
deposited), next detailed locality within that County, etc.; next County in the same State; etc.: next
State in the same Country: etc. Next Country: etc. Punctuation rules are very simple. Use a period
to separate countries, a colon to separate states, a semi-colon to separate counties, and a comma to
separate specific localities.

Acknowledgments. — Avoid overlooking persons who have in some substantial way assisted with the
work. Authors of taxonomic papers should spell out the name and indicate parenthetically the
acronym of institutions where specimens studied are deposited, if not mentioned elsewhere in the text.

Literature Cited. — Include only those publications to which reference is made in the text. Adhere
to the CBE Style Manual and refer to a previous issue of The Journal of Arachnology for style. Do
not abbreviate place names in journal citations. Repeat name(s) of author(s) in case of multiple
entries.

Figure Legends. — Provide one legend for each illustration which will be reproduced separately, or
for each “plate” consisting of several illustrations. Adhere to the following styles:

Figs. 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.

Figs. 27-34. — Right chelicerae of species of A-us from Timbuktu: Figs. 27, 29, 31, 33. — Dorsal
views; Figs. 28, 30, 32, 34. — Prolateral views of movable finger; Figs. 27-28: A-us x-us, holotype male;
Figs. 29-30: A-us w-us, male; Figs. 31-32: A-us z-us, holotype male; Figs. 33-34; A-us t-us, male. Scale
= 1.0 mm.

Type all figure legends consecutively on the same page(s), double spacing lines within each legend
and leaving four spaces between legends. Keep in mind that 99 characters and spaces represent one
printed line, and that 4 mm for each printed line must be subtracted from the maximum height
permissible on full-page illustrations.

Footnotes. — Footnotes will be permitted on page one when it is appropriate to acknowledge grant
support and to indicate change of address, etc. No footnotes will be permitted on any other page.
Type footnotes together on a separate page.

Running head. — At the top of each right hand page will be printed the author(s)’ last name(s) and
an abbreviated title. This running head may not exceed 60 characters and spaces, must be typed in
capital letters, and a hyphen must separate the name(s) from the title.

Tables with legends. — Prepare all tables precisely as they are to be typeset. Construct tables as
simply as possible. Include the legend at the top of each. Make marginal notations in the text which
clearly indicate the appropriate point of insertion of each table. Note that in these instructions
“graphs” are regarded as “figures,” not “tables.”

The size of the printed page imposes a limit on the size of tables that can be accepted. In a normal,
upright table these limits are 99 characters and spaces per line and 56 lines in height, including legend.
Sideways tables are difficult to typeset and present problems during paste-up; therefore, they will be
accepted in exceptional cases only, as most tables that are “too wide” can usually be rearranged to
fit into the more desirable, upright position. Any table that exceeds the dimensions given above will
not be accepted for publication. Tables must be typed double-spaced throughout, and the legend must
be in the following style:

Table 1.— Incorporate into the legend all information necessary to make the data presented clearly
understandable. Do not use footnotes.

409

Figures. — Figures are the only parts of the paper for which the author is responsible for preparation
of camera-ready material, and therefore they must be executed with great care. Drawings should be
made with black drawing ink on high quality, smooth, white, heavy paper. Drawings made in some
manner not employing black ink (e.g., shaded pencil drawings) are difficult to print, more so than
photographs. When such a technique is used, the drawings should be fairly dark, have good contrast,
and be well executed. The largest dimensions on the printed page on which illustrative material (with
legends) will be printed is 127 by 200 mm. Decisions as to size of originals, placement of several
smaller drawings on a single page, etc., should be based upon this format. Where several smaller
drawings are to appear together on a single page, it is the responsibility of the author(s) to furnish
an assembled “plate” precisely as it should appear in print. An effort should be made to efficiently
utilize the available space within the limits of the original “plate.” The “plate” must be prepared to
allow for inclusion of the legend beneath the drawings on the same printed page. Number all drawings
consecutively using Arabic numerals; do not designate illustrations with letters, and never designate
plates. Labeling should be neat, secure, and done with a method which produces flat characters, e.g.,
pressure-sensitive transfer characters or a lettering apparatus employing black ink. Photographs
should have good contrast, be glossy or semi-glossy, and printed slightly larger (up to 2X) than
anticipated published size. Photographs and inked drawings should not be grouped together in a
single plate, because they will be difficult to print. Illustrative material of any kind larger than two
times the anticipated printed size will not be accepted. Larger illustrations are difficult to handle and
are easily damaged or lost in the mail. Authors preparing larger illustrations should consult local
photo shops and print firms about the availability of PMT’s (PhotoMechanical Transfer) for reducing
their illustrations to an acceptable size. PMT’s are high contrast, direct positive-to-positive
reproductions on semi-glossy paper, which will also give the author(s) a good idea of how the
illustrations will look when printed. PMT’s can be obtained at a reasonable cost in a number of sizes,
usually 8 by 10 in. or 9 by 12 in., with any amount of figure reduction desired within that space.
Sending PMT’s rather than originals avoids the loss of art work in the mail, and if the PMT’s get
lost or damaged they can be replaced. Illustrations, PMT’s, and photographs must be securely
mounted on stiff white paper or mounting board (Scotch Magic Transparent Tape can be used on
the white portions of illustration paper). On the back of each plate include author’s name(s),
abbreviated title of manuscript (or running head), figure numbers included, and orientation {-
topside). Illustrative material that is poorly prepared, too large, or not mounted properly will be
returned to the author.

RESEARCH NOTES

Arrange the various parts of your research notes in the following sequence: (1) mailing address,
(2) title, (3) body of text, (4) by-line, (5) figure legends, tables with legends, and illustrations. Follow
instructions given above for feature articles unless otherwise indicated below. Do not include
footnotes anywhere. References to grant support and all other acknowledgments should be made in
a statement in the body of the text. If an indication of change of address is desired, it should be
included parenthetically after the credited institution as “present address.” For citation of literature,
follow the same instructions as for Feature Articles. The by-line must be typed in paragraph form
after the body of the text or the literature cited when one is present. When possible. Research Notes
will take priority over Feature Articles in the publication sequence.

THE AMERICAN ARACHNOLOGICAL SOCIETY

President:

Jerome S. Rovner (1985-1987)

Department of Zoology
Ohio University
Athens, Ohio 45701

Membership Secretary:

Norman I. Platnick (appointed)

American Museum of Natural History
Central Park West at 79th Street
New York, New York 10024

Secretary:

Brent D. Opell

Department of Zoology
V.P.LS.U.

Blacksburg, Virginia 24061
Directors:

Frederick A. Coyle (1985-1987), W. G. Eberhard (1986-1988), Susan E. Riechert
(1985-1987).

Honorary Members:

P. Bonnet, W. J. Gertsch, H. Homann, R. F. Lawrence, H. W. Levi, G. H.
Locket, A. F. Millidge, M. Vachon, T. Yaginuma.

The American Arachnological Society was founded in August, 1972, to
promote the study of Arachnida, to achieve closer cooperation between amateur
and professional arachnologists, and to publish The Journal of Arachnology.

Membership in the Society is open to all persons interested in the Arachnida.
Annual dues are $25.00 for regular members, $15.00 for student members.
Correspondence concerning membership in the Society must be addressed to the
Membership Secretary. Members of the Society receive a subscription to The
Journal of Arachnology. In addition, members receive the bi-annual newsletter
of the Society, American Arachnology.

American Arachnology, edited by the Secretary, contains arachnological news
and comments, requests for specimens and hard-to-find literature, information
about arachnology courses and professional meetings, abstracts of papers
presented at the Society’s meetings, address changes and new listings of
subscribers, and many other items intended to keep arachnologists informed
about recent events and developments in arachnology. Contributions for
American Arachnology must be sent directly to the Secretary of the Society.

President-Elect:

William A. Shear (1985-1987)
Department of Biology
Hampden-Sydney College
Hampden-Sydney, Virginia 23943

Treasurer:

Norman V. Horner (1985-1987)
Department of Biology
Midwestern State University
Wichita Falls, Texas 76308

Archivist:

Vincent D. Roth

Box 136

Portal, Arizona 85632

Research Notes

Phidippus audax (Araneae, Salticidae) predation upon a cicada
(Tibicen sp.) (Homoptera, Cicadidae), Timothy C. Lockley and

Orrey P Young ; 393

Prey of the striped lynx spider, Oxyopes salticus (Araneae, Oxyopidae),
on cotton in the delta area of Mississippi, Timothy C. Lockley and

Orrey P Young 395

Maternal behavior in Apollophanes punctipes (O. Pickard-Cambridge)

(Araneae, Thomisidae), William G. Eberhard 398

Fossil evidence of spider parasitism by Ichneumonidae, George O. Poinar, Jr. . . 399
X-Ray sterilization of male golden-web spiders Nephila clavipes
(Araneae), Terry Christenson, Joseph Schlosser, Jeffrey Cohn, and
Leann Myers 40 1

Book Reviews

Neurobiology of Arachnids, by F. G. Barth, Jerome S. Rovner 404

Sexual Selection and Animal Genitalia, by W. G. Eberhard,

Steven N. Austad 405

Other

Nomenclature Notes 406

Instructions to Authors 407

CONTENTS

THE JOURNAL OF ARACHNOLOGY

VOLUME 14 Biography NUMBER 3

B. J. Kaston, American Araneologist 1906-1985: a biography and
bibliography, Leonard S. Vincent 283

Feature Articles

Postembryonic development of Lactrodectus hasselti Thorell (Araneae,

Theridiidae), Michael F Downes 293

Spiders on red spruce foliage in northern Maine, Daniel T Jennings

and Judith A. Collins 303

Coniferous-habitat associations of spiders (Araneae) on red spruce

foliage, Daniel T Jennings and Judith A. Collins 315

Nests of terrestrial spiders maintain a physical gill: flooding

and evolution of silk constructions, Jerome S. Rovner 327

Web-building behavior of anapid, symphytognathid and mysmenid

spiders (Araneae), William G. Eberhard 339

A new mygalomorph spider genus from Mexico (Nemesiinae, Nemesiidae,

Arachnida), Robert J. Raven 357

Biology of the diurnal Metasolpuga picta (Kraepelin) (Solifugae, Solpugidae)
compared with that of nocturnal species, Robert A. Wharton. ... . . 7. . ,'.363

Genetic differentiation in the genus Phidippus (Araneae^ Salticidae),

A. C. Terranova and S. H. Roach 385

(continued on back inside cover)

Cover photograph, fluorescence of Sphaleropachylus butleri (Thorell), by J. C. Cokendolpher
. - Printed by the Texas Tech Press, Lubbock, Texas, U.S. A.

, .. Posted at Lubbock, Texas, in May 1987

1>.

TUTioN MouniiiSNi imv. voSHiiiAis S3iavaan libraries Smithsonian institution

c/> 2! ^ 2:; V .• . </) 2!

< ^ „< 2: ^

> ^ >

^ in 2: </) * 2 — (/)

Viian LIBRARIES SMITHSONIAN INSTIIUTION NOlinillSNI NVINOSHXIt^S SBIBVaail

0^ IT _ <n

Q ^ O pc^ ^ O

rUTION^NOITniliSNI~‘NVlNOSHillNS S3iaVaan~'LIBRARI ES^ SMITHSONIAN”' institution

2 r“ :z r“ z: ■“

vaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws saiavaan

in

X
in

o

2 ■ 2 >■''- 2

(/) -<1- C/)’‘^*' ]Z CO —

UTioN '^oiiniusNi_NviNosHiiws S3iavaan libraries Smithsonian institution

CO X \ ^ ^ ^ —

q::

<
tr

O O Xi;o^ - O

2 _J 2

vaan..L‘BRARiEs_SMiTHsoNiANjNSTiTUTiON NoiiniiiSNi NviNOSHiiws S3iavaai

/ i i

> (f# H ^ s ^ /g

70

in ^ in ^ CO \ ^ ,

rUTION NOlinillSNI NVINOSHIIINS SaiBVaail libraries SMITHSONIAN INSTITUTIO^
CO Z <n z .y. in 2 .V.-

X\

^ z ■ CO 2: CO * — (/) 1

vaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiins saiavaai

CO X CO X CO

ruTioN NoiiniiiSNi“*NvmosHiiiMS saiavaan'^LiBRARi es^ Smithsonian institution

V 2 r- Z r- I

2 tn ■ • _ to £ CO

vaan libraries Smithsonian institution NoiiniiiSNi nvinoshihns saiavaan

2 '• Z v CO 2 CO

rUTION NOIJ.nLliSNI_NVINOSHillAIS SaiaVHaiT libraries SMITHSONIAN INSTITUTION

CO z \ ^ 71 in zi ^

c/5 zz (/> K zz N wi. v-' ! c/j —

isNi MviNOSHilws S3iavyan libraries Smithsonian institution NoiiriLUSNi n

2 ^ W Z W 2 W Z

r ^ [%P. .^51 g

5 2 Ni^JiosSX S > ■•<18^ 2

lES' SMITHSONIAN INSTITUTION NOlifliliSNI NVINOSHLIlMs‘”s3 I U VB 3 IT LIBRARIES S

(n 2 U5 — . , U5 5 w

5 5 5 “

LSNI~NVINOSHilWS^S3 I BVB 3 n“‘LI B R AR I ES^SMITHSONIAN“‘lNSTITUTION^NOIiniUSNI_N

c/5 E c/> 2 c/5 ^ c/>

\ES SMITHSONIAN INSTITUTION NOlifliliSNI NVINOSHillAIS S3idVHan LIBRARIES^^S

Z H CO z w z ^

<

^ JMZ'Wa o /•i^

O X

= Ni^S?' I |,^ g ^-

2 ■••^’ 2 W 2

LSNI_NVINOSHLliMS SBIBVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOIinLli.SNI_h

2: \ ^ — CO :r c/5

O ^<^os\^ « O

2 — ^ Z

lES SMITHSONIAN^INSTITUTION NOliHiliSNI NVINOSHilWS S3 I BVB 8 ll^LI B R AR I ES

m CO m to

H
LTi

~ CO £ — (/> \ £ CO —

iSNI NVINOSHillAIS S3l8Vaan LIBRARIES SMITHSONIAN INSTITUTION NOliHiliSNI I

Z CO Z CO z to z

< s s: ... s: <

z
o

CO

X V^J

^ t

[lES^SMITHSONlAN INSTITUTION NOliHiliSNI NVIN0SHilWS^S3 1 d V8 8 H LIBRARIES ^

(/> X iO zz . . (fi X c/5

O p 4>>" _ x^uiis,^ O II O

iSNl'^NVINOSHillNS S3 I dVd 8 Il’^LI B RAR 1 £S^SM!THSONIAN“*INSTITUTION^NOIiHiliSNI^I
r“vZ r- Z r- 2

DO

zo
>

m c >w ^ X^ASiii^x m m Xi^osv^ zz m

{/> ' zz (/) £ CO £ CO

lES SMITHSONIAN INSTITUTION NOlinLliSNl NVINOSHilWS S3IBVHan LIBRARIES !

z , crt z (/> z ■■« t2

>• -^212::^ 2 >' 2 > 2

iSNl_NVINOSHi.IWs‘^S3 I B VB 3 I T^LI B RAR I ES^SMITHSONIAN^INSTITUTION *”NOIiniliSNI_ I

-7 \ CO :r: co r!: ,