5 N**“*>' 5 m >W 1 rn ~ „ N®.

HIINS^SB I ava 3 II^U B RAR I ES^SMITHSONIAN INSTITUTION^NOliDlIlSNI "”nvinoshumsws3 I avi

CO Z CO

I if%l § ,%k § life § j

1 i $3P i s w,

z I "Sr > Viav7 i ^ 5-* / § VJsssx >■

DNIAN_INSTlTUT10N^N0linxliSNI_NVIN0SHllWs"s3iaVaan\lBRARIEs“sMITHS0NIANjNSTlTU

O O ~ o n-vosh^ 0

z _J z «J 2 3 H

■lliins saiavaen libraries Smithsonian institution NoiiniusNi nvinoshiuns saiav:

z r- . — . ? <“ z r* 2 '

b.

• ■ IT- w- .-. _

_ CO — CO 5 IS [/) V ^

3NIAN INSTITUTION NOIlfllliSNl NVIN0SH1IIAJS S3 I H VH 8 11 LIBRARIES SMITHSONIAN” INST IT L

,y/ ^c£ “O 2^ v ,» . I/?

< AS . <£ < y<^SOv>v < . \X v

sKxS - A" =: ~ , , v = ,^'1 -

X
c O

i

U i > -a £ ’V- a '\ > X^Uixs^X 5

c O *- Z CO Z CO *Z

lliiajs sb lava an libraries Smithsonian institution NoimiusNi nvinoshiiws saiav

CO 2 CO 2 00 ”... CO

avI^Tx co w . >4 co co ^

*\

_ _ * O Q

)N!AN^1NSTITUT!0N NOlinillSNI^NVINOSHlIIAIS^SS I H VH a ll^Ll B RAR I ES^SMITHSONIAN^INSTITL
1 z ., z r- z «~

O tCTatit/T^ -*■ O T^STirp^ ~ X^,a O IZ /^o i

%}

m m - W £ XiSfoP' m Xt£5fTMX m * " ' m

CO _ CO ‘ £ CO 2 <y>

miAis S3 1 ava an libraries Smithsonian institution NoiiniiiSNi nvinoshiiiais S3iav

^ ^ ^ ^ ^ . Z * to z

$ 1 I 1 %f4^0 1 ^ ^ °

2 co X' z W “ Z M ^

NIAN INSTITUTION NOIlfUllSNI NVIN0SH1IW.S S3iavaaiT LIBRARIES SMITHSONIAN INSTITL

~ CO 2 \ 00 “ ~

SJlS ^CCxATITf/>N. ^ XITN. u . n Z

CO

O -S^VOSHV^ ” o " X^VAS'A'y q ^ 5

liiais S3 1 ava an libraries Smithsonian institution NoiiniiiSNi ^nvinoshiiiais ^saiav

z 5 r* - — ^ z r* 7-

o XTIcvTSw /^VASOaTTX O /<VASO/Ov ^ . — \V •■ ~

V/

CO

__ <Quirjjx m

nian ” institution NouniiisNrNviNosHiiws^s 3 1 ava a n~LI B R ar I eswsmithsonian±institi

Z ,w co Z CO z .... CO z w 1

o

co

X

BRARIES SMI
to

m x^vos^ix m

CO — CO £ CO —

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3IHVH9n _>

CO Z t to Z CO

X
to

> s I''' | I*' 5=

iimuLSNi NviN0SHims°°s3 1 ava a n librar i es^smithsonian institution NoiiniiisNi_Nvj

CO „ ^ X CO = ^ ^rcrr^ Z "

CO CO ^ w y/Zr¥L&-

^ O _ O

BRARIES SMITHSONIAN INSTITUTION NOlifllllSNI^ W¥IN0SH1IIA!S ^S3 I HVB 8 II LIBRARIES SMI
z «“ z r- ^ r=_z

m jo ^ x ^ co m f5

_ GO ± c/> X 5 00

mniiiSNi NViNOSHims saiavaan libraries Smithsonian institution NOlinillSNI

Z CO Z «... CO 2 V to Z

< £ . . ,S <4\-; 5 < .4K« £ ■•*

> ''W' 2 '\ > Xouu%2X 25

CO Z CO *. 2 w * z CO

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3iava9Sl LIBRARIES SMI

CO Z CO ™ . CO 5

uj ✓^v'sovTX. UJ ztMiTtTx \ yj /-:Xy C — in, /t^v'so^x CsJ

O

ilinillSNl"JNVIN0SHHWSZS3 I HVa a n“Yj B RAR I ESZSMITHS0NIAN~IN3TITUTI0N NOlinillSNI NVI
r* * z r* z r” z r*

x>

t

m z: \i*«S/ m Xivo^X ^ m X^osv^ ^ m

to — co £ co £ co

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIMQSH1IWS SBiavaail LIBRARIES SM,
to ^ ^ z co z eg z

5 x^TirtOx < A S S <

1 Jgfcs

1 * W X

>'

iinmsNi nvinoshiiwsws3 i avaa iui brar i es^smithsonian^institution coNoiiniiisMi_N¥i

2 \ ^ - z CO CO 2

o " “ x^wst^Z q ~ ■ 0

2 -J Z _j Z __

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IIAIS S3 lavas 11 LIBRARIES SM

2^ ocpP' <2 m n^d£>' w x^TjBt^z m X co x iinj^y m

co £ — co \ Z — co

MiniliSNI NVIN0SH1IWS SBIHVaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI N¥

■* — "• z v... co z

•-'''no

:

-

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 52 • ISSUE 1 • SPRING, 1967

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

1. Neural Adaptations in the Visual Pathway of Certain Heliconiine Butter-

flies, and Related Forms, to Variations in Wing Coloration. By S. L
Swihart. Figures 1-8 1

2. Preliminary Studies on The Isolation of Pterins from the Wings of Heli-

coniid Butterflies. By John Baust. Text-figures I-IV 15

3. Underwater Sound Production by Captive California Sea Lions, Zalophus

calijornianus. By Ronald J. Schusterman, Roger Gentry & James
Schmook. Plates I-V. 21

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park, Bronx Park, Bronx, N. Y. 10460, and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers, $1.50, unless otherwise stated in the Society’s catalog of publications. Second-class
postage paid at Bronx, N. Y.

Published June 23, 1967

1

Neural Adaptations in the Visual Pathway of Certain
Heliconiine Butterflies, and Related Forms, to Variations
in Wing Coloration1

S. L SWIHART
Department of Biology
N. Y. State University College
Fredonia, New York 14063

(Figures 1-8)

[This paper is a contribution from the William
Beebe Tropical Research Station of the New York
Zoological Society at Simla, Arima Valley, Trinidad,
West Indies. The station was founded in 1950 by the
Zoological Society’s Department of Tropical Re-
search, under Dr. Beebe’s direction. It comprises
250 acres in the middle of the Northern Range,
which includes large stretches of government forest
reserves. The altitude of the research area is 500 to
1800 feet, and the annual rainfall is more than 100
inches.

[For further ecological details of meteorology and
biotic zones, see “Introduction to the Ecology of
the Arima Valley, Trinidad, B.W.I.” by William
Beebe, Zoologica, 1952, 37 (13) 157-184.

[The success of the present study is in large meas-
ure due to the cooperation of the staff at Simla,
especially of Jocelyn Crane, Director. The author
particularly wishes to acknowledge the invaluable
assistance of Dr. Donald R. Griffin, Dr. Michael
Emsley, and Mr. John G. Baust],

Introduction

THE EXPERIMENTS which form the
basis of this paper were stimulated by
the well-known responsiveness of butter-
flies to visual stimuli of various colors. Behav-
ioral observations of this phenomenon include
those of Eltringham ( 1933), Ilse ( 1 937 ) . Mag-
nus (1956), and most particularly those of
Crane (1955, 1957).

Supported by grants (GB-2331 and GB-4218) from
the National Science Foundation.

Crane’s demonstration of the responsiveness
of Heliconius erato to orange-red stimuli in feed-
ing and courtship behavior prompted a series of
experiments designed to determine the physio-
logical basis of this special sensitivity to long
wavelengths. These experiments (Swihart, 1963,
1964, 1965) have demonstrated that: (1) the
visual system of H. erato includes at least two
types of receptors, one maximally sensitive to
blue-green and the other peaking in the red (ca.
620 m^); (2) the receptors interact with each
other producing an electroretinogram (ERG)
waveform with a distinct color component; (3)
specific neural pathways are associated with
these receptors. Recordings from the internal
chiasma, medulla interna, and the vicinity of
the optic nerve demonstrate a spectral sensitivity
which indicates a particularly close association
between the red receptors and the main pathway
of information to the brain.

The behavioral sensitivity of this species
seems, therefore, to be related not to any special
modification of the receptors, but rather to the
development of pathways which “selected” the
output from those receptors which transduced
information with special biological significance.

The basic hypothesis emanating from these
observations was that there was a selective ad-
vantage in developing neural mechanisms which
demonstrate disproportionate sensitivity to the
basic wing coloration, presumably because of
the role played by such colors in releasing court-
ship behavior. This hypothesis invited testing by

1

2

Zoologica: New York Zoological Society

[52: 1

conducting experiments on forms other than
erato , which possessed different wing coloration,
to determine how general this mechanism might
be, and the degree of adaptive flexibility afforded
by such a system. This paper reports the pre-
liminary results of such a comparative study.

Methods and Materials

Standard electrophysiological techniques were
employed. Intact organisms were rigidly mount-
ed with plasticene as previously described (Swi-
hart, 1963, 1964). Evoked potentials were re-
corded from the optic lobe with a semi-micro
(ca. 500 kohm), 3M KC1 filled, glass electrode.
On the basis of previous experiments (Swihart,
1965), a technique was developed for placing
the electrode almost directly into the medulla
interna, via a minute hole in the posterior (occi-
put) region of the head. Usually only very small
additional movements of the electrode were re-
quired to obtain the long-latency, negative
polarity, response characteristic of the medulla
interna (see fig. 15, Swihart, 1965).

In order to ascertain that the preparation was
continuing to yield “normal” responses to photo-
stimulation, ERGs were recorded simulta-
neously with a sub-corneal steel electrode. Ex-
periments were terminated if there was any
change in the ERG waveform.

Potentials were amplified with Grass P6 D.C.
amplifiers, displayed on a Tektronix four beam
564 oscilloscope, and photographed with a
Grass C-4 camera for subsequent measurement.

Photostimulation was accomplished with a
laboratory constructed stimulator which, auto-
matically, sequentially introduced a series of
15 narrow-band interference filters into an
optical system, and provided a 100 msec, stimu-
lus at each wavelength, at a preset and con-
stant interval, usually about one minute. The
stimulus duration was chosen as the shortest
which would ensure production of all the com-
ponents of the ERG (Swihart, 1964).

Preparations were aligned so that the stimu-
lating beam, focused by a microscope objective,
illuminated nearly an entire eye. The beam axis
was perpendicular to the center of the cornea.
Histological studies which have been conducted
on a number of the forms selected for this
study (Morpho, Agraulis, H. sarae, and H.
erato), and physiological experiments involving
the stimulation of small portions of the eye (of
H. erato), have not demonstrated any anatom-
ical or physiological differentiation between
various regions of the eye.

Peak transmission points of the interference
filters were fairly uniformly spaced throughout
the visible spectrum from 404 m,u. to 709 m/x.

Through the use of a specially-ground, color-
compensating filter, and sandwiched gelatin fil-
ters, the stimulus energy at each wavelength was
held constant (± 20%) as determined with an
Eppley thermopile.

Equipment was not available for making
small, calibrated adjustments in stimulus inten-
sity, hence “spectral sensitivity” (threshold en-
ergy) curves were not attempted, rather the
electrical magnitude of the response to a stand-
ard stimulus of about 5 x 103 microwatts/cm2
was determined.

Spectral efficiency curves were constructed by
running through the series of 15 filters three
times for each preparation. The magnitude of
the response to a particular stimulus was eval-
uated as a percentage of the largest response
to any filter in that particular series. The per-
centage responses to the three successive series
were then averaged to produce the curve for a
single preparation. The technique of recording
the amplitude of the response, seemed justified
since the magnitude of the medulla interna
response demonstrates a nearly linear relation-
ship to the log of the intensity of stimulation,
within the range of intensities studied. In this
range, a tenfold increase in white light intensity,
produces an increase in response magnitude of
about 30%.

Experimental material was normally captured
in the wild and maintained in large outdoor
insectaries. This technique permitted the selec-
tion and testing of only healthy animals.

Wing spectral reflectance characteristics were
measured with a Bausch & Lornb “Spectronic
20” spectrophotometer with reflectance attach-
ment. This equipment measures the reflectance
of an area approximately 2 mm x 8 mm. Only
specimens which appeared to be newly emerged
were used for such determinations.

It will doubtlessly be noted that the number
of specimens of each species investigated is fre-
quently very small. It is hoped that this will be
understood as related to certain problems which
were encountered, including: (1) the physical
difficulty of capturing healthy specimens of cer-
tain elusive and rare species, notably Philae-
thria and Morpho; (2) the extremely delicate
nature of these organisms which results in their
entering a state of “shock” if handled roughly.
This condition is evidenced by abnormal be-
havior, including strong positive phototaxis, and
highly aberrant evoked potentials, usually ac-
companied by very large (ca. 5 mV) low-fre-
quency spontaneous discharges; (3) the neces-
sity of limiting the observations to only those
individuals which produced strong “day” type
electroretinograms (Swihart, 1963); (4) the
problems in placing the micro-electrode into the

1967]

Swihart: Neural Adaptations of Certain Heliconiine Butterflies

3

intact organism in such a manner that it reaches
the very limited region near the medulla interna,
which yields the long-latency negative poten-
tials, without inducing operative trauma, and
its resultant massive spontaneous activity.

Six species of butterflies were selected for this
study. All belong to the family Nymphalidae.
Four are members of the subfamily Heliconii-
nae, while the Morphinae is represented by one
form ( Morpho), and the Nymphalinae by Vic-
torina.

Studies were begun on Heliconius same be-
cause of its fairly close relationship with the
subject of previous experiments, H. erato. In
Trinidad, erato is a small black butterfly (2Vi"
wingspread) with bright red spots on the fore-
wings. H. sarae is, in many respects, a very
similar butterfly, with the most obvious differ-
ence being the substitution of a pair of bright
yellow bands on each forewing. In addition
there is a moderately strong blue irridescence
on the hind wings.

Heliconius ricini was chosen because it offered
an “intermediate” color pattern to sarae and
erato. In ricini, the black “base” color of the
wings is broken by forewing yellow bands (as in
sarae ) , but a large red hindwing spot is also
present. H. sarae and ricini are the most closely
related of the forms tested. Both belong to the
same species-group as discussed by Emsley
(1965).

The third form chosen, Agraulis vanillae, was
selected because its color pattern is similar to
that of a number of other primitive Heliconiinae.
That is to say, the wings are predominantly
bright orange in color.

Even a cursory glance at the variety of wing
color patterns characteristic of the Heliconiinae
demonstrates a single basic plan. This is a long-
wavelength color with a contrasting dark brown
or black. It can be seen that the aforementioned
forms are representative of this basic pattern.
However, for purposes of investigating the phy-
siological adaptations of the visual pathway, it
seemed desirable to investigate forms with an
unusual wing coloration. Initially, it was hoped
that studies could be made on the rare, pre-
dominantly green heliconiine, Philaethria dido;
however, only a single healthy specimen of this
species was obtained, and observations were ex-
tended to the very similarly marked nymphalid.
V ictorina steneles.

Finally, to further extend the variety of wing
colors, observations were made on the blue-
winged Morpho peleides.

The results of the observations on each species
will be considered separately.

Results

Heliconius sarae thamar Hubner

A total of eight individuals that were tested
yielded acceptable responses, i.e., long-latency
negative potentials. A ninth individual produc-
ing such responses became erratic before the
experiment was completed; however, its per-
formance in the early stages of the experiment
gave a clear indication of its sensitivity.

It was found that the spectral efficiency curves
produced by these individuals were far from be-
ing identical to each other. Examination of the
curves revealed that they could be separated
into two distinct types, with each category being
fairly homogeneous. Figures 1 and 2 show these
two types of curves. Four individuals demon-
strated a sensitivity peak in the orange-red por-
tion of the spectrum (ca. 620 m^); the remain-
der were maximally sensitive to green.

In spite of the small sample size, the differ-
ences between these two curves is so great that
there is no statistical basis for doubting that
they represent two distinct populations. Thus,
for example, one obtains by comparing the per-
centage responses at a single wavelength (523
nr/t) , at value of 4.78, with a n = 7*. The proba-
bility that the responses represent different popu-
lations is thus greater than 0.995. If this analysis
is carried to additional points on the spectra,
no reasonable doubt can remain.

The red-sensitive curve peaks at 620 mp, and
in this respect is very much like the responses
recorded from erato. However, there is a dif-
ference between the two forms in their sensi-
tivity to shorter wavelengths. Thus, the response
of sarae to stimulation with short wavelengths
is, on the average, considerably greater than
erato (Fig. 2).

Figure 3 shows the spectral reflectance curves
of the erato red pigment, and the sarae yellow.
It can be seen that the yellow color could ac-
tually be described as “blue negative,” since it
actually reflects about the same amount of red
as does the erato pigment. If one, therefore, con-
siders the difference between these colors in
terms of a greater reflectance of mid and short-
er wavelengths, there is a parallel in the differ-
ences between the reflectance and spectral effi-
ciency curves in the two forms.

which is an approximation of a longer formula giv-
en by Welch (1947).

4

Zoologica: New York Zoological Society

[52: 1

100"

90"

80"

70"

iu
c n
z
o

CL

CO

LU

tr

60-

50"

z

(UJ

o

QC

UJ

OL

40"

30"

20"
10' •

0 I I I V— -i— ■ I I i- I I- r" I H

400 450 500 550 600 650 700

WAVE LENGTH IN MJJ

Figure 1.

One of the two types of spectral efficiency curves found to be characteristic of Heliconius sarae. Mea-
surements were based upon the magnitude of long-latency, summated potentials of negative polarity in
the vicinity of the medulla interna. See text for details of technique. Vertical bars indicate the limits of
observed variability between individuals, i.e., the maximum and minimum responses to any given wave-
length. Illustration is a semi-diagramatic representation of the dorsal surface: unshaded portions corre-
spond to the areas of yellow pigmentation. Shaded area represents the distribution of near black wing
pigmentation. Unless otherwise noted, all subsequent diagrams are to the same scale.

The other type of spectral efficiency curve
was recorded with the same frequency as the
red-sensitive one (four of each, with the pre-
viously mentioned ninth being red sensitive). It
would appear that the sensitivity of these or-
ganisms was largely determined by the green-
sensitive receptor system with a slight and vari-
able skew produced by the presence of the red
system.

At this time it is impossible to identify the
factors which contribute to the differences be-
tween these two types of individuals. All the
organisms were tested at approximately the
same time of day (late morning), and demon-
strated similar “day” type electroretinograms.
Three males and one female yielded the red-
sensitive curve, while three females and one

male produced the green curve. The age of the
individuals was not known since they were all
wild caught specimens.

It should be remembered, however, that be-
havioral observations on butterflies have demon-
strated rather dramatic shifts in the colors to
which they will respond, depending upon their
physiological state. Thus, Ilse (1937) showed
that while Pieris brassiccie selected red, yellow
or blue colors in its feeding behavior, it became
specially responsive to the color green when in-
volved in egg-laying behavior. It may well be
that such shifts in spectral sensitivity represent
a measurable index of the rather abstract con-
cept of "physiological state.”

Heliconius ricini (L.)

A total of eight specimens of this species pro-

1967]

Swihart: Neural Adaptations of Certain Heliconiine Butterflies

5

100-r
90- «

Id

CD

2

O

CL

CD

LU

CC

80- "

70- -

60- •

50

uj 40
o

QC

a 30'

20- •

10- *

0 L-h

400

H. erato

-I i i I I t i I - I I I -

450 500 550 600 650 700

WAVE LENGTH IN MJJ

Figure 2.

Alternate type of spectral efficiency curve that may be recorded from Heliconius sarae which demon-
strates far greater sensitivity to long wavelengths than does the curve in Figure 1. Also included in the
figure, for purposes of comparison, is the comparable average spectral efficiency curve recorded from
four individuals of the species Heliconius erato (dot-dash-line). Also figured is the distribution of red
pigmention in the forewings of H. erato.

duced acceptable responses. In spite of the fact
that the wing coloration of this form can be
considered to be intermediate between sarae
and erato (i.e., has both red and yellow mark-
ings), the responses were entirely similar to
sarae. Thus three of the individuals yielded
spectra peaking in the red (Fig. 4) (620 m/x),
with no significant differences from the sarae
red-sensitive curve. One of the specimens tested
was most interesting in that the nature of its
response changed during the course of the ex-
periments. The responses to the initial filter
series indicated a maximum sensitivity to red.
A gradual change took place, without any ap-
parent change in the waveform of the response,
so that by the time the organism was stimulated
by a fourth series of filters, the response was
plainly maximally sensitive to green light.

The remainder of the organisms tested pro-
duced the non-specific green-sensitive curve.

Agraulis vanillae vanillae (L.)

Five specimens of this primarily orange-col-
ored butterfly produced acceptable responses.
The averaged responses of these individuals pro-
duced a curve which is distinctively different
from any which had been previously recorded
(Fig. 5). In many respects, this spectrum can
be considered to be entirely intermediate be-
tween the two sarae spectra. Thus, Agra it I is
demonstrates a curve peaking in the orange, less
sensitivity to short wavelengths than the sarae
green curve, and less sensitivity to long wave-
lengths than the sarae red curve. These differ-
ences are so great that there is little or no over-
lap of even those responses demonstrating the

6

Zoologica: New York Zoological Society

[52: 1

Figure 3.

Spectral reflectance characteristics of various heliconiine wing pigments as measured against a magne-
sium carbonate standard. Intact wings were mounted on a nonreflective backing before being inserted
into the spectrophotometer. No differences in reflectance exist between similar appearing areas in the
various species, e.g., erato and ricini red are identical.

greatest variance from the mean in most of the
aforementioned portions of the spectrum.

It will be noted that, like erato, only a single
type of response was recorded from this species.

Philaethria dido dido Clerck and Victorina
steneles (L.)

It was hoped that observations could be made
on the green-winged heliconiine Philaethria as
it would provide a form with a markedly differ-
ent wing coloration. Unfortunately, not a single
wild specimen could be obtained. Several eggs
were located, however, and a single healthy
specimen was raised in the laboratory. Testing
of this individual did indeed produce responses
showing a distinctive peak in the green. Regret-
tably, no additional specimens were available
to confirm these observations. Because of the
shortage of Philaethria, it was decided to make
observations on the remarkably similar nympha-

lid Victorina. These forms are so similar that
they can be easily confused. The resemblance
is much more profound than mere superficial
appearance. Both forms have developed their
green coloration by a most interesting technique.
In the green areas of the wings, the scales are
either much reduced, or missing, and the pig-
mentation is within the wing membranes. The
coloration itself appears to be due to the pres-
ence of two pigments. One of these absorbs in
the blue (i.e., appears yellow) and can be ex-
tracted with ether, leaving an insoluble blue
compound behind (Fig. 6). It seems likely that
this unusual wing coloration is due to haemo-
lymph pigments (Hackman, 1952) since Vic-
torina’s blood and eggs are both bright grass-
green.

Four specimens of Victorina were tested and
found to resemble Philaethria in visual mecha-
nisms as well as in pigmentation. Figure 7 shows

1967]

Swihart: Neural Adaptations of Certain Heliconiine Butterflies

7

100"

90"

80"

70"

iLl

z 60"
o

CL
CD

C 50"

t—

2 40"
o

cr

S 30“

20"

10"

0 L=4 i 1 I — f -I ■" * I -i - I I i — i -

400 450 500 550 600 650 700

WAVE LENGTH IN MJU

Figure 4.

One of the two spectral efficiency curves characteristic of Heliconius ricini. Note the great similarity to
Figure 2. The other type of curve is similar to the same “green-sensitive” curve illustrated in Figure 1.
Unshaded areas of illustration correspond to the distribution of the yellow pigment, while the lightly
shaded area represents the hind-wing red spots.

the luminosity curve produced by these indi-
viduals which clearly peaks at about 565 mp..
There can be no doubt that this curve is dis-
tinctly different from the green-sensitive curve
of sarae. At virtually every wavelength longer
than the 565 m/x maximum, the Victorina re-
sponses are so greatly reduced that there is very
little, if any, overlap with even those sarae show-
ing the greatest deviation from the mean.

The curve recorded from the single Philae-
thria was entirely similar to that produced by
Victorina, and fell well within the limits of Vic-
torina variability.

Morpho peleides insularis Fruhstorfer

It was considered advisable to make tests on
a form with a primarily blue wing coloration.
Since there is no such heliconiine in Trinidad,
it was decided to utilize the large blue-violet

morphine Morpho peleides. This butterfly is well
known to show a special sensitivity to the color
blue. For many years, professional collectors
have used this fact to assist in the capture of this
elusive species, with its highly prized iridescent
physical coloration.

Two specimens of this species were fully
tested for spectral efficiency. The electrical re-
sponses produced by these organisms were
found to be so different from the responses of
the heliconiine that direct comparison is diffi-
cult. The electroretinograms show none of the
complexities, described in detail for erato (Swi-
hart, 1964) . The waveform consists of a B wave,
followed by a uniform, sustained negativity dur-
ing illumination. All the ERG components usu-
ally associated with the red-sensitive receptor,
such as the “dip” following the B wave, and
“off” effect, are totally absent.

8

Zoologica: New York Zoological Society

[52: 1

100- *
90- 4

80- *
70- *

iij

to

z

o

CL-

IO

Ll)

a:

60- •
50-'

ijj 40- *
o

cc

£ 30- •
20- •

10- •

0 -

WAVE LENGTH IN MJJ

Figure 5.

Single type of spectral efficiency curve recorded from five specimens of Agraulis vanillae. This species is a
nearly uniform orange color (light shading), with numerous small black (unshaded) spots on the wings.

These differences in the ERG waveform are
reflected in differences in evoked potentials, as
recorded with micro-electrodes. Unlike the
heliconiine, it was found that the evoked poten-
tials were generally directly proportional in
magnitude to the size of the ERG. The spectral
efficiency of these effects, as measured by the
ERG B wave magnitude, is indicated by Figure
8. The skew of this curve towards the blue, with
its peak at about 485 m/x is both obvious and
strikingly different from the curves produced by
the heliconiine.

Two additional specimens of this species were
tested in the course of work preliminary to the
experiments which form the basis of this report.
Full spectral efficiency curves, as determined
by the standard technique, were not calculated
for these individuals. They did, however, appear
to demonstrate the same ERG waveform and
great sensitivity to short wavelengths.

Discussion

It is unfortunately true that the fundamental
mechanisms of color vision remain largely un-
explained. This lack of knowledge is particu-
larly apparent in the case of invertebrates, where
even the method of coding color information
has not been identified as has been done for
certain vertebrates (e.g., Wagner et al., 1960;
Muntz, 1962).

Recent microelectrode studies (e.g., Horridge
et al., 1965) have revealed a vast complexity in
discharge patterns in the insect visual pathway.
There is, however, some question as to the ade-
quecy of such single fiber techniques to record
from those fibers most intimately involved in
the highest forms of behavior. There can be no
doubt that these techniques tend to select large-
diameter fibers. Such neurons have been repeat-
edly demonstrated to be associated with flight
or escape reactions, and hence can hardly be

1967]

Swihart: Neural Adaptations of Certain Heliconiine Butterflies

9

Figure 6.

Reflectance characteristics of the green wing areas of Philaethria dido and Victorina steneles. Since the
scales are much reduced, or absent in these regions, they are quite translucent and reflect poorly. Be-
cause of this fact, the wing reflectance was measured with a white backing. This technique increased
the height of the curve without altering its shape. Solid line illustrates the characteristics of the natural
wing coloration. Dot-dashed line indicates the reflectance of the wings after being extracted with ether.
Dashed line indicates the transmittance of the yellow-colored extract. The height of this curve was ad-
justed by plotting % transmittance.

2

considered as representative. For these reasons
semi-microelectrodes have been employed in the
current study, as it seems not unlikely that a
“summated” response reflects the nature of the
nervous activity with somewhat less bias.

In this connection, it should be remembered
that it was demonstrated (Swihart, 1965) that
there are fibers with discharge patterns related
to the magnitude of such summated potentials.
Thus, the curves presented in this report reflect
the discharge frequency (and hence the total
number of spikes per stimulus of standard dura-
tion) of at least some of the neurons in the visual
pathway.

Recently, Goldsmith (1965) has given evi-
dence that spectral sensitivity curves derived

from summated (ERG) potentials must be
carefully interpreted if screening pigments are
present. Such problems are not encountered in
the genus Helicortius, where the only pigmenta-
tion appears to be a nearly black substance, lo-
calized in granules within the iris pigment cells.

In the case of Agraulis and Victorina there
is a light orangish pigmentation within the cor-
neal cuticular layer. Tflis pigmentation does not
appear to penetrate to the deeper layers of the
eye, and hence probably could not produce the
effects described for Musca. However, the role
of such accessory pigments deserves further in-
vestigation.

A careful analysis of the techniques employed
in these experiments may suggest that they lack

10

Zoologica: New York Zoological Society

[52: 1

100'

90'

80

70'

Ld

CO

z 60'
o

CL

CO

£ 50'

UJ 40'
o

QC

£ 30*

20'

10'

0

WAVE LENGTH IN MJJ

Figure 7.

Spectral efficiency curve recorded from four specimens of Victorina steneles. The single specimen of Phi-
laethria dido demonstrated a similar sensitivity. Lightly shaded areas of illustration represent the distribu-
tion of the green coloration, as opposed to the dark brown-blackish pigmentation.

some of the controls commonly employed in the
investigation of primary visual events. It must
be remembered that this is primarily a com-
parative study and that the techniques employed
on one form were identical to those used on the
others. Thus, while it may not be possible to
interpret the results as representative of photo-
pigment absorption spectra, they are indicative
of real differences which exist between closely
related forms.

It is reasonable to enquire as to the origin
of these differences.

Analysis of the magnitude of the electroretino-
grams, in a method analagous to that described
in this paper, produces spectral efficiency curves
with no significant differences between the vari-
ous Heliconiinae. In every case the curves re-
sembled that described for H. erato (Swihart,
1963). This fact strongly suggests that the varia-

tions between forms cannot be related to dif-
ferences in the nature of the photopigments.

Alternatively one might suggest that the vari-
ations in sensitivity are due to differences in the
relative numbers of several different types of re-
ceptors, (e.g., Calliphora; Autrum & Burkhardt,
1961). However, an explanation based upon
such a rigid mechanism seems inconsistent with
the type of variability observed in sarae and
ricini.

For such reasons, it seems most reasonable
to interpret the observed variations in the sum-
mated responses to various colors, as a neural
phenomenon.

Turning to a consideration of the spectral
efficiency curves themselves, we find a most in-
teresting series in the responses of Agraulis,
sarae and erato. In considering these forms, it is
worth noting that casual behavioral observations

1967]

Swihart: Neural Adaptations of Certain Heliconiine Butterflies

11

IOOt

90- 4
80- *

70- •

UJ

CO

z

o

CL

CO

UJ

oc

60- •
50-'

LU

O

cr

UJ

CL

40- *
30-'

20-*

10-'

0 -

WAVE LENGTH IN MJJ

Figure 8.

Average spectral efficiency curve derived from two specimens of Morpho peleides, as determined by the
ERG B-wave magnitude. Very little variability between the individuals was observed. Illustration is at re-
duced scale. Central (hatched) areas of wings is iridescent blue, with contrasting brown-black margins.

by the author have indicated that consistently
stronger behavioral responses to specific colors
are produced by the species with the red and
orange markings. In the case of sarae, attempts
to demonstrate a behavioral sensitivity to yellow
or red have, as of this time, produced no posi-
tive responses in courtship (Crane, pers. com.).
This condition would seem to have a physio-
logical correlate in the existence of the two types
of spectral efficiency curves recorded from this
form, as opposed to the uniformity which char-
acterizes the two other species.

It would appear that considerable significance
can be attached to the variability between in-
dividuals producing each of the two different
sarae curves. Each of the curves demonstrates
a highly asymmetric distribution of the degree
of variability, which would hardly be expected
on the basis of a “normal” distribution of vari-
ance. The most likely interpretation of these

curves would seem to be that the green-sensitive
curve reflects the activity of a visual system re-
sponding primarily to the sensitivity of a blue-
green sensitive receptor, with a small and vari-
able contribution by a red-sensitive system. On
the basis of such an interpretation, the short
wavelength portion of this curve (400-525 m/x)
would probably reflect a portion of the inherent
sensitivity of the receptor system. Conversely,
the red-sensitive curve illustrates a high degree
of variability only at wavelengths below 525 m/x.
This would seem to imply that the system is re-
sponding primarily to a red-sensitive receptor
(maximum 620 m p.) , with a small and variable
contribution by the blue-green system. These
two maxima observed in sarae are the same as
those reported for erato (Swihart, 1964).

By comparison, the single type of spectral
efficiency curve characteristic of Agraulis dem-
onstrates a fairly uniform variability through-

12

Zoologica: New York Zoological Society

[52: 1

out the spectrum. This would seem to indicate
that the mechanism is not “dominated” by a
single receptor system, but rather reflects the
neural summation of the activity in several re-
ceptor types. Indeed, the peak of such a curve
may not correspond to any specific type of re-
ceptor, but, rather, indicate a region where the
overlapping sensitivity of two receptors “sum-
mates” to produce an “artificial” peak. It is,
therefore, interesting to note that the peak of
the Agraulis curve lies midway between the
maxima of the two receptors postulated for sarae
and erato.

Extending our analysis to the other forms
studied, we find that ricini is similar in all re-
spects to sarae. This is not particularly surpris-
ing when one considers both the extremely close
phylogenetic relationship between the two forms
(Emsley, 1965) and the basically similar wing
coloration (i.e., forewing yellow bands) .

Turning to the two similar green forms, Phi-
laethria dido and Victorina steneles, we find sev-
eral significant differences from the species pre-
viously considered. First is the rather obvious
shift in the peak from the 528 m/x region to
about 570 niju. The very small variance between
individuals at the peak of the curve suggests that
this is probably due to a difference in receptors
rather than a mechanism such as that postulated
for Agraulis.

The second remarkable feature of the Vic-
torina curve is the extremely attenuated response
to long wavelengths. It seems most unlikely that
the orange color of the cornea could be respon-
sible for the diminished responsiveness to long
wavelengths. While it is possible that this por-
tion of the curve represents the sensitivity of the
receptors, it seems much more probable that
some other mechanism is involved (e.g., inhi-
bition by a red receptor system).

It is difficult to extend the preceding type of
analysis to Morpho, since the responses of this
form (ERG and neural) are so different from the
preceding forms as to make a direct comparison
difficult, if not impossible. Regardless of the
final interpretation of the nature of the visual
mechanisms of this form, it is clear that virtually
all the recorded responses demonstrate a maxi-
mum sensitivity to the blue portion of the spec-
trum. It is obvious that this organism must pos-
sess a blue-sensitive receptor. The question is,
therefore, whether this represents a unique type
of receptor. The neural responses of the other
forms have given but little indication of any
special sensitivity to short wavelengths. Only
Victorina’s neural spectral efficiency curve dem-
onstrates the type of variability between indi-
viduals which could be interpreted as being

clearly indicative of the activity of a blue sensi-
tive system. On the other hand, spectral effici-
ency curves based upon the ERG B-wave dem-
onstrate a rather considerable sensitivity to short
wavelengths. This is true even in the case of the
primarily red-sensitive erato (Swihart, 1963).
It seems possible, therefore, that such a system
may be fairly commonly distributed among the
Nymphalids; however, evolutionary adaptation
has resulted in its contributing little or nothing
to the excitation of the information pathway
in the vast majority of species where the primary
wing coloration is in the long wavelength por-
tion of the spectrum.

To conclude this discussion, it is interesting
to speculate as to the evolutionary forces which
have contributed to the development of the weak
and variable responses characteristic of the
forms with the yellow forewing bands, i.e., sarae
and ricini. In considering this problem, one must
remember that there are two important factors
which have played a role in the development of
butterfly wing coloration. These are: protective
(warning or mimetic) coloration and the con-
servative force of sexual selection. These two
forces are frequently antagonistic. In fact, the
opposing pressures of these two factors are
believed (Brower, 1963) to have produced the
multiple cases of sexual dimorphism in wing
coloration found in butterflies.

As previously noted, many of the primitive
Heliconiinae are primarily orange in color. This
coloration appears to be due to a pterin pigment
(Baust, 1967). A small modification of this
molecule has produced the erythropterin pig-
mentation found in the red spots of erato, etc.
It seems not unlikely that this refinement of the
chromophore, which has produced a coloration
with greater purity, has allowed the refinement
of highly specific behavior patterns based upon
the releasing value of this striking color.

On the other hand, the yellow pigmentation
of sarae and ricini represents the development
of an entirely new type of pigment (an amino
acid, Brown, 1965) probably in response to
some other pressure. It is certain that these
forms have not lost the ability to synthesize a
pterin pigment since almost all the Heliconiinae
demonstrate minute red spots at the base of the
forewing. Employment of the yellow pigment
vice the red is hard to explain in terms of sexual
selection since the low color purity of this pig-
ment would make an inherently poor sign stimu-
lus. On the other hand, the yellow reflects about
207% more light in the visible spectrum than
does the red. It seems certain this more brilliant
pigmentation is considerably more effective as
warning coloration. The species has had to ac-

1967]

Swihart: Neural Adaptations of Certain Heliconiine Butterflies

13

commodate to this increased emphasis on warn-
ing coloration with a lessened dependence on
wing coloration as a courtship releaser. This
change appears to be reflected in the less specific
neural adaptation of the visual pathway to the
organism’s wing coloration.

It must be admitted that many of the ideas
which have been put forth must remain in the
category of speculation. To some extent, this
seems to be an inherent penalty for attempting
to penetrate the perceptual world of another
species.

Summary

Specimens of six species of butterflies were
examined using standard electrophysiological
techniques. Spectral efficiency curves were con-
structed for each species. For five of the species,
this was done on the basis of the magnitude of
long-latency, negative polarity, summated po-
tentials, associated with the activity of higher
order neurons in the vicinity of the medulla
interna.

Two species with yellow forewing spots (Heli-
conius sarae and H. ricini) produced similar re-
sults, i.e., individuals yielded one of two types
of curves, one a non-specific curve peaking in
the green, and another peaking in the red, with
a shape very similar to the spectral reflectance
of the yellow wing pigmentation. The orange
butterfly, Agraulis vanillae, produced a single
type of curve peaking in the orange. The two
green butterflies, Philaethria dido and Victorina
steneles, produced curves peaking in the green.
The blue butterfly, Morpho peleides, produced
very different electrical responses. The electro-
retinogram did not demonstrate the components,
which in the preceding forms, are associated
with a red-receptor system. The spectral effici-
ency curve based upon Morpho’ s ERG B-wave
demonstrated a maximum in the blue.

On the basis of these observations, and pre-
vious studies of a form with red markings (Heli-
conius erato), it is suggested that butterflies pos-
sess a neural mechanism which “selects” the
output from various receptors in such a manner
so as to make the visual system respond maxi-
mally to stimulation with colors approximating
the wing pigmentation.

References

Autrum, H. & D. Burkhardt

1961. Spectral sensitivity of single visual cells.
Nature, 190: 639.

Baust, J. G.

1967. The isolation of pterins from the wings of
heliconiine butterflies. Zoologica, 52:
00-00.

Brower, L. P.

1963. The evolution of sex-linked mimicry in
butterflies. Proc. XVI Int. Cong. Zook,
4:173-179.

Brown, K. S.

1965. A new L-cc amino acid from lepidoptera.
J. Am. Chem. Soc., 87:4202-4203.

Crane, J.

1955. Imaginal behavior of a Trinidad butterfly,
Heliconius erato hydra Hewitson, with
special reference to the social use of color.
Zoologica, 40:167-196.

1957. Imaginal behavior in butterflies of the fam-
ily Heliconiidae: changing social patterns
and irrelevant actions. Zoologica, 42:135-
145.

Eltringham, H.

1933. The Senses of Insects, Methuen. London.

Emsley, M. G.

1965. Speciation in Heliconius (lep. Nymphali-
dae): Morphology and geographic distri-
bution. Zoologica, 50:191-254.

Goldsmith, T. H.

1965. Do flies have a red receptor? J. Gen. Phys-
iol., 49:265-287.

Hackman, R. H.

1952. Green pigments of the hemolymph of in-
sects. Arch. Biochem. Biophysics, 41:166-
174.

Horridge, G. A., J. H. Scholes, S. Shaw &

J. Tunstall

1965. Extra cellular recording from single neu-
rons in the optic lobe and brain of the
locust. In The Physiology of the Insect
Central Nervous System, Ed. J. E. Tre-
herne and J. W. Beament, Academic, New
York.

Ilse, D.

1937. New observations on responses to colors
in egg-laying butterflies. Nature, 140:544.

Magnus, D. B.

1956. Experimental analysis of some “overopti-
mal” sign-stimuli in the mating behavior
of the fritillary butterfly Argynnis paphia
L. (Lepidoptera: Nymphalidae.) Proc.
Tenth Inter. Cong, of Ent., 2:405-418.

Muntz, W. R.

1962. Effectiveness of different colors of light in
releasing positive phototatic behavior of
frogs, and a possible function of the reti-
nal projection to the diencephalon. J.
Neurophysiol., 25:712-720.

Swihart, S. L

1963. The electroretinogram of Heliconius erato
(Lepidoptera) and its possible relation-
ship to established behavior patterns. Zoo-
logica, 48: 155-165.

14

Zoologica: New York Zoological Society

[52: 1

1964. The nature of the electroretinogram of a
tropical butterfly. J. Ins. Physiol., 10:547-
562.

1965. Evoked potentials in the visual pathway of
Heliconius erato (Lepidoptera) . Zoolog-
ica, 50:55-62.

Wagner, H. G., E. F. MacNichol &

M. L. WOLBARSHT

1960. The response properties of single ganglion
cells in the goldfish retina. J. Gen. Physiol.,
43, Suppl. 45-62.

Welch, B. L.

1947. The generalization of “Students” problems
when several different population vari-
ances are involved. Biometrika, 34:28-35.

2

Preliminary Studies
on

The Isolation of Pterins from the Wings of Heliconiid Butterflies1

John G. Baust2
Department of Biology
State University College
Fredonia, New York

(Text-figures I-IV)

[This paper is presented as a portion of a series
of studies on the Heliconiid butterflies which have
been supported by the National Science Founda-
tion and organized by Jocelyn Crane. The focal
point of these studies has been the William Beebe
Tropical Research Station of the New York Zoo-
logical Society at Simla, Arima Valley, Trinidad,
W.I. The station was founded in 1950 by the Zoo-
logical Society’s Department of Tropical Research
under the late Dr. Beebe’s direction.

[The success of the present study is in great part
due to the invaluable aid rendered by both Miss
Crane, director, and Dr. M. G. Emsley, who so
graciously contributed many of the specimens
needed. The author is particularly indebted to Dr.
Jerome H. Supple, Department of Chemistry, and
Dr. Stewart L Swihart, Department of Biology,
both of the State University College, for their ad-
vice and keen interest in the study. The author
wishes to gratefully acknowledge the gifts of sam-
ples of erythropterin, xanthopterin and rhizopterin
from Dr. E. L. Rickes, Merck and Co., Inc., Rah-
way, New Jersey],

Introduction

PTERINS HAVE BEEN isolated from a
variety of organisms, e.g., Tschesche &
Vester (1955) isolated erythropterin from
Mycobacterium lacitcola, Lecercq (1950) and

Supported by grant #GB-4218 from the National
Science Foundation.

2The author’s present address is the Department of
Physiology, Downstate College of Medicine, State Uni-
versity of New York, Brooklyn, New York.

Schopf & Becker (1933) from Hymenoptera,
Goto (1963) from Amphibia, and Forest &
Mitchell (1954) from Drosophilia, to mention
but a few. Pterins have also been isolated from
the wings and eyes of various Lepidoptera
(Pfleiderer, 1962; Schopf & Becker, 1933).
However, these studies have been limited to a
few moths and butterflies of the family Pieridae.
Essentially, this was due to the fact that it has
been generally believed that pterins existed only
within these groups of Lepidopterans (Ford,
1947; Ziegler-Giinder, 1955) . As a result of this
study, however, it has been demonstrated that at
least two Heliconiid butterflies contain pterins
as their principle wing pigment.

Analysis of the red wing patches of Heli-
conius erato adanis, a black, neo-tropical butter-
fly with two distinct red spots, has led to the
identification of erythropterin (Text-fig. I). Its
chemical structure has been described by Purr-
man & Eulitz (1948), Fieser & Fieser (1963),
and Tschesche & Korte (1951), and its proper-
ties by Fox (1953) and Albert (1954). A sec-
ond pterin has been detected in the wings of the
orange Heliconiid, Colaenis julia, but has not
yet been identified.

Methods and Materials

The red pigmented regions of the wild caught
Heliconius erato were removed. They were then
defatted with ethyl ether in a Soxhlet apparatus,
and the pigment extracted in a crude form with
methanolic HC1, evaporated and redissolved in

15

16

Zoologica: New York Zoological Society

[52: 2

COOH

ERYTHROPTERIN TRICYCLIC FORM

Text-fig. I. The proposed bicyclic and tricyclic structures of erythropterin.

methanol. Since only minute quantities of the
pigment were contained within the wing portions
used, identification was initially limited to paper
chromatographic techniques. Whatman filter
paper #1 was used, and chromatograms were
run in a butanol: acetic acid: water (4:1:5)
solvent system.

Ultra violet and visible spectra of the above
samples were recorded on a Beckman DK-2
Spectrophotometer.

Results

Initial experiments demonstrated that the
physical and chemical properties of the pigment
extract were consistent with those commonly
attributed to pterins (Cromartie, 1958). The
pigment was found to be insoluble in cold water
and most organic solvents, was degraded by
oxidation, and was melted with difficulty. It was
soluble in most acidic and basic media. These
observations suggested a more precise identifi-
cation on the basis of paper chromatography and
spectrophotometric comparisons with known
pterin samples.

Chromatograms of the crude pigment ex-
tracted with methanolic HC1 yielded two fluor-
escent spots with the previous mentioned solvent
mixture. The first had a pink fluorescence and
an Rf value of 0.33. These results were then
compared with chromatograms obtained from
what is believed to be pure erythropterin (Table

2):i. The chromatograms of the known and un-
known material were found to be identical in
R, values and in fluorescence for both sets of
spots. The erythropterin’s lower spot compared
favorably with the value obtained by Good &
Johnson ( 1949) .

The wing pigment and erythropterin spots
were analyzed spectrophotometrically, while
separately and individually eluted with methanol
from the paper. Strikingly similar spectra were
obtained in both the visible range (350-550m /x)
and in the ultra violet range (230-350 m,u.) for
all spots. The principle peaks were at approx-
imately 272, 458, and 490 m/r (Table 1 and
Text-figs. I & II). It should be noted that the up-
per and lower spots resulted in all but identical
spectra, with only minor differences within the
visible range. The lower erythropterin spot
(Good & Johnson, 1949) is thought to be the
bicyclic form, while the upper spot may be a
tricyclic isomer.

A second Heliconiid, Colaenis julia, was in-
vestigated briefly in an attempt to determine the
nature of its pigment. The spectral and chrom-
atographic data (Table 2) obtained from this
study showed that the orange color is due essen-
tially to a pterin. The structure of this particular

:lBoth the rhizopterin and xanthopterin samples were
compared in the same manner as the erythropterin and
wing pigment.

1967]

Baust: The Isolation of Pterins

17

c

o

U)

—

E

CD

c

fd

c

CD

o

CD

Q_

o i I ■ ■ ■ I — — — L— — — i^— a— — — a

250 300 350 400 450 500 5 50

wave length in mu

Text-fig. II. Ultra violet-visible spectra of the pigment extracted from H: erato with methanolic
HC1 and analyzed in a methanol solvent. Refer to Table I for numerical data for acid-base additions.

Table 1 . Ultra Violet and Visible Spectra*
(CH3OH solvent)

Methanolic HCi
extract

Erythropterin

273

270

Neutral**

458

459

480

481

490

490

235

235

Plus 5%

285

285

NaOH

462

463

485

485

499

498

272

270

Plus 5%

459

457

HC1

480

492

490

*wave length in m/x

**This is the respective order of the additions of NaOH
and HC1.

Table 2. Relative R(

Values of Pigments

Studied. (Solvent

—Butanol: Acetic

Acid: Water, 4:1:5)

Compound

Rr Value

H. erato pigment
(extracted with

0.10 (lower)
0.33 (upper)

methanolic HCI )
Erythropterin

0.10 (lower)
0.33 (upper)

Colaenis pigment
(extracted with
methanol)

0.53

Rhizopterin

0.49

Xanthopterin

0.39

pterin is as yet undetermined. Both its UV and
visible spectra are quite similar to those of the
erythropterin and the H. erato extract. How-
ever, relative intensities of the individual peaks

18

Zoologica: New York Zoological Society

[52: 2

Text-fig. III. Ultra violet-visible spectra of the erythropterin sample dissolved in methanol. Refer to
Table I for numerical data and for acid-base additions.

are quite different. Also, the R( of this particular
pterin is 0.55 as opposed to 0.10 and 0.33 for the
erythropterin forms. Chromatographic and spec-
trophotometric comparisons with samples of
rhizopterin and xanthopterin yielded dissimilar
results.

Heliconius sara, a black Heliconiid with two
yellow bands on each forewing, was also ex-
amined. Its yellow pigment is known to be a
new L-cc -amino acid (Brown, 1965). However,
it was checked in order to determine whether
or not a pterin was also contributing to the
yellow color. The resulting data were quite dis-
similar to those obtained from the H. erato or
C. julia specimens. The chromatographic and
spectrophotometric data were identical to those
of Brown and no indication of a pterin was
found.

Conclusion

From the above data it can be concluded that
erythropterin exists as a pigment within at least

one Heliconiid butterfly. Also, there is no reason
to doubt that it exists in other members of the
same family since many have the same distinct
red coloration on various portions of the body.
The pigment is no doubt located on the walls
of canals in the scales as it is in pierid butterflies
(Ziegler-Gunder, 1955).

The spectrum of the pigment extract is iden-
tical with that of the erythropterin. The fact that
two spots appear on chromatograms both with
the erythropterin and the methanolic HCI ex-
tract seem to indicate that this is probably not
the case. There is good indication that an equil-
ibrium exists between the two forms of this pig-
ment. Tschesche & Barkmeier (1955) and
Fieser & Fieser (1963) have suggested that
erythropterin may exist in equilibrium with a
tricyclic form (Text-fig. I) . Excellent support for
such an equilibrium is found in the fact that
when individual chromatogram spots were
eluted and re-run, two spots were again obtained.
Both had the same Rf values and the same fluor-

1967]

Baust: The Isolation of Pterins

19

Text-fig. IV. Ultra violet-visible spectra of the pigment extracted from Colaenis julia with methanol.

escence as did the original sample. A second
and less likely explanation is related to the rel-
atively unstable nature of pterins. It might be
possible to assume that certain changes in or
the loss of portions of a side chain or group
might cause a change in Rr without a corres-
ponding effect upon the UV spectra (Nawa,
Goto, et al., 1964) .

When comparing the visible spectra of Text-
figures II and III, it is obvious that a discrepancy
exists in regard to one of the principle peaks (480
m/x) . When the neutral H. erato pigment is made
alkaline, the 480 peak shifts out but returns upon
acidification as expected. The erythropterin,
however, does not do this. The 480 peak shifts
out with the addition of alkali but fails to return
upon acidification. This is attributed to the fact
that approximately equal quantities of acid were
added to the solutions of erythropterin and pig-
ment extract. It was later realized that excess
acid was needed to cause a complete re-shift.

An unidentified pterin has been detected in
the orange wings of Colaenis julia. It possesses
all the properties of pterins, has a spectra similar
in shape but not intensity to that of erythropterin
but has a different Rf.

References

Albert, A.

1954. The Pteridines. Fortschr Chem. Org.
Naturstoffe. 1 1: 350-403.

Brown, K. S.

1965. A new L-cc -amino acid from Leptdoptera.
Jour. Am. Chem. Soc., 87: 4202.

Cromartie, R. I. T.

1958. Insect Pigments. Ann. Rev. Entomol., 4:
59-56.

Fieser, L. F., & M. Fieser

1963. Topics in Organic Chemistry. Reinhold,
New York.

20

Zoologica: New York Zoological Society

[52: 2

Forest, H. S., & H. E. Mitchell

1954. Pteridines from Drosophila I. Isolation of
a yellow pigment. Jour. Am. Chem. Soc.,
76: 5656.

Ford, E. B.

1947. A murexide test for the recognization of
pterins in intact insects. Proc. Roy. Ent.
Soc. Lond., (A), 22: 72-76.

Fox, D. L.

1953. Animal Biochromes and Structural Colors.
Cambridge University Press, Cambridge.

Good, P. M., & A. W. Johnson

1949. Paper chromatography of pterins. Nature,
63: 31.

Goto, T.

1963. Uber die Verundderungen der Pterine in
den Entwichlungstadien von einem Frosch,
Rhacophorous schlegchii var. arborea. Ja-
pan Jour. Zool., 14: 68-81.

Lecercq, J.

1950. Occurrence of pterin pigments in Hymen-
optera. Nature, 165: 367.

Nawa, S., M. Goto, S. Matsuura, H. Kakizawa, &

Y. Hirata

1954. Studies on pteridines. Jour. Biochem., 41:
657-660.

Purrman, R., & F. Eulitz

1948. Uber die Fluglpigmente der Schmetter-
linge. XVI Zur Kenntnis des Erythropterin.
Ann. Phys. Lpz., 559: 169-174.

Schopf, C., & E. Becker

1933. Occurrence of pterins in wasps and butter-
flies. Ann., 507: 266-296.

Tschesche, R., & H. Barkmeier

1955. Pteridines. XI. Furanopteridine, a contrib-
ution to the constitution of erythropterin.
Ber. deut. chem. Ges., 88: 976-983.

Tschesche, R., & F. Korte

1951. The synthesis of erythropterin. Ber. deut.
chem. Ges., 84: 77-83.

Tschesche, R., & F. Vester

1953. Erythropterin from Mycobacterium lacti-
cola, Ber. deut. chem. Ges., 86: 454-459.

Ziegler-Gunder, J.

1956. Pterine: Pigmente und Wirkstoffe im Ter-
reich. Biol. Rev. Cambridge Phil. Soc., 31:
313-349.

3

Underwater Sound Production
by Captive California Sea Lions, Zalophus californianus

Ronald J. Schusterman
Roger Gentry
&

James Schmook
Stanford Research Institute
Menlo Park, California

(Plates I-V)

Introduction

Thus far, it has been shown that the Cali-
fornia sea lion, Zalophus californianus,
produces two types of underwater sounds
—clicks, or short-duration sound pulses, and
barks (Poulter, 1963; Schevill, Watkins & Ray,
1963). In contrast, the bottlenose porpoise pro-
duces a wide variety of sounds— clicks, whistles
or squeals, barks (Evans & Prescott, 1962), and
“cracks” (Caldwell, Haugen & Caldwell, 1962).
Although the clicks are used for echolocation by
the porpoise (Kellogg, 1961; Norris, 1964),
there is evidence indicating that a variety of
whistles have emotional and communicative sig-
nificance. Some whistle contours and the crack-
ing sounds have been said to be associated with
distress or fright reactions (Caldwell et al., 1962;
Lilly, 1962).

Underwater clicks by Zalophus have been re-
ported to occur usually when an animal was in
the final stages of searching for food (Poulter,
1963; Schevill et ah, 1963) or for an object
signalling food (Evans & Haugen, 1963; Schus-
terman, 1966). Most of these tests have limited
the range of behavior to those involved in feed-
ing activities. In order to determine whether
Zalophus is capable of emitting a greater variety
of underwater signals and calls than had been
previously reported, several California sea lions
were monitored while swimming freely under a
number of stimulus conditions.

Procedure and Apparatus

All observations and recordings were made
while animals were swimming untethered in an
oval tank constructed of redwood, measuring
15 feet by 30 feet and 6 feet deep, and filled
with 82 kiloliters of fresh water. Recordings of
the underwater sound productions by Zalophus
were made under the following conditions: (a)
conspecific social interaction; (b) orientation to
a mirror; (c) fleeing from a human observer.

Underwater sounds were continuously mon-
itored by a Channel Industries 275 hydrophone
(20 Hz to 150 kHz) and an Ampex 2044 ampli-
fier-speaker system (65 Hz to 13 kHz). Vocal
signals were periodically recorded on a Uher
4000-S tape recorder at 7.5 inches/second (40
Hz to 20 kHz).

Spectographs of the evoked signals were
made, using the Kay 661 sonograph. Either of
two analyzing bandpass filters (narrow or wide)
may be used with the Kay sonograph. The wide-
band filter has an effective bandwidth of 300
cycles, and the narrow-band filter has an effec-
tive bandwidth of 45 cycles. The analysis used
is indicated on each of the spectrographs pre-
sented. The use of this method for the analysis
of biological sounds has been described by
Borror (1960).

Recorded Sounds

Clicks. — Our preliminary analysis indicates

21

22

Zoologica: New York Zoological Society

[52: 3

that Zalophus produces a great variety of click
patterns. Although most of the click trains have
a duration of 2 seconds or less, many trains last
as long as 23 seconds with pauses of less than
0.5 second. The click repetition rate may vary
from less than 5 per second to 70 or 80 per
second, all within a given click train (separation
between clicks of 0.5 second or less).

Plate I is a spectrograph of clicks produced
by one sea lion while play-fighting with another
sea lion. When the tape which produced this
spectrograph was replayed, we noted that a pop-
ping sound seemed to be superimposed on the
clicks. This is indicated on the graph, we believe,
by the great variation in the frequency pattern.
Such a sound pattern is highly distinctive and
has been produced by only one of the animals
(Cathy) .

It is important to note that clicking sounds
were never emitted at fairly regular intervals by
any of the animals under any of the free-swim-
ming conditions. This is in marked contrast to
the behavior of the bottlenose dolphin, which
is reported to emit “exploratory” pings every
15 to 20 seconds. Such periodic signal emission
has been suggested as the sonar equivalent to
“glancing” in the field of vision (Kellogg, 1961 ) .

Barks.— This form of underwater vocalization
has most of its energy below 3500 Hz, although
some energy may be found at frequencies as high
as 8000 Hz. There is little variation in the dur-
ation of any given bark; they generally last from
200 to 300 milliseconds. Barks are sometimes
preceded by a series of clicks, as shown in Plate
II. The sounds shown in this plate were pro-
duced by a two-year-old male Zalophus while it
was fleeing from the experimenter, who was
attempting to drive the animal out of the testing
tank. During the experimenter’s initial attempts,
the animal swam rather rapidly while producing
long trains of clicks. As the action became more
intense, the clicks shifted into a series of barks.

Whinny .—A spectrograph of this vocalization
is shown in Plate III. It was frequently produced
by a 3.5-year-old female Zalophus (Bibi) during
an aggressive encounter. For lack of a better
term, we have called it the “whinny” sound,
since it sounds a little like a horse neighing. This
vocalization is often preceded by clicks or a
growl sound. The whinny sound typically lasts
for about 1.5 seconds and may be repeated three
or four times in succession. This whinny sound
may be the female counterpart of a male bark.
However, contrary to another report (Bonnot,
1951), we have heard females bark both in air
and submerged.

Buzzing.— A characteristic “buzz" sound from
a sea lion in a social situation is depicted in

Plate IV. This vocalization may actually be a
series of discrete sound pulses which occur so
rapidly that they take on a buzzing quality.

Bang or Crack.— This sound has thus far been
produced by two of our California sea lions
(a male and female). The sound was first heard
when Bibi was confronted with its mirror image
and was repeated several times over a period of
days, usually under the same circumstances.
Plate V shows a pair of these high-energy
“bang” sounds. The sound, which has always
been associated with extremely rapid swimming,
appears quite loud and mechanical to the human
ear, and, as the spectrograph shows, it is a broad-
band pulse with a rapid onset. Apparently, from
the description of Caldwell et cil. (1962), Zal-
ophus’ “bang” sound is very similar to high-
energy “crack” sounds produced by Tursiops
truncatus under conditions of fright. We have
recently heard similar sounds produced by
Steller’s sea lion ( Eumetopias jubatus) while per-
forming on an underwater visual discrimination
task.

Sound Production Mechanisms

Careful observations of Zalophus while it was
in the act of emitting underwater clicks have
indicated some movement in the area of the
throat or larynx; such movement appeared less
pronounced when the animal was silent. These
preliminary observations implicating the laryn-
geal area as the underwater sound-producing
site of Zalophus have been supported by experi-
mental evidence (T. C. Poulter, 1965). Using
a triangulation technique, Poulter found that the
site of underwater barking was the vocal cords
on the anterior portion of the larynx and that
the apparent point of origin of underwater clicks
was posterior to the vocal cords.

All of the underwater vocalizations that have
been described can apparently be produced with
the mouth and nostrils closed and therefore with-
out the emission of bubbles, or with the mouth
and nostrils partially opened and with the emis-
sion of bubbles. Moreover, clicks may be pro-
duced in air with the mouth closed or with the
mouth wide open. Barking sounds seem to show
the same basic frequency-intensity structure in
air and under water. However, clicking in air is
usually less intense and much less frequent than
under water. Although no systematic attempt has
yet been made to measure the intensity of Zalo-
phus' underwater clicks, there has been no diffi-
culty in monitoring these sounds even when the
background noise was considerable and the ani-
mal was as far as 5 to 6 meters from the hydro-
phone.

It is not clear how the “bang” sound of Zalo-

1967]

Schusterman, Gentry & Schmook: Zalophus californianus

23

pints is produced, i.e., whether it is made by the
sea lion’s vocal apparatus, by jaw-clapping, or
by some other mechanism such as the front flip-
pers causing an underwater cavitation as they
are thrust together and then parted during initia-
tion of a very rapid swim.

Discussion

Thus far, all of the underwater sounds pro-
duced by captive California sea lions have had
a pulsed structure and appear to be wholly or
partly a function of social or investigatory re-
sponsiveness. The shifting from clicking to bark-
ing or to a whinny sound under conditions of
either extra-specific or conspecific intimidation
suggests that these calls form a single system of
vocalization which changes as a function of the
level of physiological arousal (Duffy, 1957),
with barking indicative of a higher level of
arousal than clicking. This notion is similar to
that held by Andrew (1962, 1964), who has
developed the concept of “stimulus contrast” to
account for the vocalization of chicks and non-
human primates.

Although there are certain similarities be-
tween the sonar signals of the porpoise ( Tur -
siops truncatus) and the clicks of Zalophus cali-
fornianus, there are also great differences.
Whereas the clicks of the porpoise are very nar-
row columns of “noise” having their greatest
energy up to 30 kHz, with components of lesser
intensity reaching 170 kHz (Kellogg, 1961;
Norris, 1964), those sampled from Zalophus
thus far often contain at least traces of har-
monics and have their greatest energy at 500 Hz
to 4000 Hz, with possibly weak components ex-
tending to higher frequencies. Furthermore, re-
garding the porpoise, Norris reports that “dur-
ing fine discriminations where sight is impos-
sible, the environment is literally saturated with
tiny plosive clicks, up to 500-600 per second,”
(Norris, 1964, p. 320). Such rapid pulsing has
not been consistently produced by Zalophus.

Summary

Spectrographs are presented of underwater
sounds made by captive sea lions (Zalophus
californianus) under the following conditions:
(a) social interaction; (b) orientation to a mir-
ror, and (c) fleeing from the experimenter. These
animals produce a variety of vocal utterances
and sounds, including varying patterns of clicks,
barks, “whinny” sounds, “bangs,” and buzzes.
All sounds thus far recorded and analyzed have
a pulsed structure with dominant frequencies
ranging from 500 Hz to 4 kHz. The sounds ap-
pear to be wholly or partly a function of the

social and investigatory responsiveness of the

sea lion.

Bibliography

Andrew, R. J.

1962. The situations that evoke vocalization in
primates. Ann. N. Y. Acad. Sci. 102: 296-
315.

1964. Vocalization in chicks, and the concept of
“stimulus contrast.” Anim. Behav. 12:
64-76.

Bonnot, P.

1951. The sea lions, seals and sea otter of the
California coast. Calif. Fish and Game
37: 371-389.

Borror, D. J.

1960. The analysis of animal sounds. In: Animal
Sounds and Communication, edited by
W. E. Lanyon and W. N. Tavolga, Wash-
ington, D.C., Amer. Inst, of Biol. Sci.:
26-37.

Caldwell, M. C., R. Haugen & D. K. Caldwell

1962. High-energy sound associated with fright
in the dolphin. Science, 137: 907-908.

Duffy, E.

1957. The psychological significance of the con-
cept of “arousal” or “activation.” Psychol.
Rev. 64: 265-275.

Evans, W. E. & R. Haugen

1 963. An experimental study of the echolocation
ability of a California sea lion. Zalophus
californianus (Lesson). Bull. So. Calif.
Acad. Sci.. 62: 165-175.

Evans. W. E. & J. H. Prescott

1962. Observations of the sound production ca-
pabilities of a bottlenose porpoise: a study
of whistles and clicks. Zoologica 47: 121-
128.

Kellogg, W. N.

1961. Porpoises and Sonar. Chicago: Univ.

Chicago Press.

Lilly, J. C.

1962. Distress call of the bottlenose dolphin:
stimuli and evoked behavioral responses.
Science 139: 1 16-1 18.

Norris, K. S.

1964. Some problems of echolocation in ceta-
ceans. In: Marine Bio-Acoustics, edited by
W. N. Tavolga. New York: Pergamon
Press: 317-336.

Poulter, T. C.

1963. Sonar signals of the sea lion. Science 139:
753-755.

24

Zoologica: New York Zoological Society

[52: 3

1965. Location of the point of origin of the vo-
calization of the California sea lion, Zalo-
phus californianus. Proc. Second Ann.
Conf. Bio-Sonar, Menlo Pk., Calif.: Stan-
ford Research Institute: 41-48.

Schevill, W. E., W. A. Watkins & C. Ray

1963. Underwater sounds of pinnipeds. Science
141: 50-53.

SCHUSTERMAN, R. J.

1966. Underwater click vocalizations by a Cali-
fornia sea lion: effects of visibility. Psy-
chol. Rec. 16: 129-136.

EXPLANATIONS OF PLATES

Plate I.

Spectrograph of clicks emitted by a 3-year-old fe-
male California sea lion (Cathy) while play-fighting
with another sea lion (narrow band).

Plate II.

Spectrograph of clicks and barks produced by a 2-
year-old male (Tommy) while fleeing from the ex-
perimenter (wide band).

Plate III.

Spectrograph of Bibi’s “whinny” vocalization pro-

duced during an aggressive encounter with another
California sea lion (narrow band).

Plate IV.

Spectrograph of a “buzzing” sound emitted by Cathy
while swimming with another California sea lion
(narrow band).

Plate V.

Spectrograph of two "bang” sounds produced by a
3-year-old California sea lion (Bibi) while orienting
to a submerged mirror.

(ZALOPHUS CALIFORNIANUS

SCHUSTERMAN, GENTRY & SCHMOOK

PLATE I

TIME — seconds

SCHUSTERMAN, GENTRY & SCHMOOK

PLATE II

FREQUENCY kHz

— rooJ-^oiOT-vioo — pooJ-P>cn<D-sioo

UNDERWATER SOUND PRODUCTION BY CAPTIVE CALIFORNIA SEA LIONS
(ZALOPHUS CALIFORNIANUS)

UNDERWATER SOUND PRODUCTION BY CAPTIVE CALIFORNIA SEA LIONS
(ZALOPHUS CALIFORNIANUS )

SCHUSTERMAN, GENTRY & SCHMOOK

PLATE III

SCHUSTERMAN, GENTRY & SCHMOOK

PLATE IV

UNDERWATER SOUND PRODUCTION BY CAPTIVE CALIFORNIA SEA LIONS
(ZALOPHUS CALIFORNIANUS)

SCHUSTERMAN, GENTRY & SCHMOOK

PLATE V

FREQUENCY — kHz
— pooj-kcno^joo

NEW YORK ZOOLOGICAL SOCIETY

GENERAL OFFICE

630 Fifth Avenue, New York, N.Y, 10020

PUBLICATION OFFICE

The Zoological Park, Bronx, N.Y. 10460

Fairfield Osborn
President

Laurance S. Rockefeller
First Vice-President

OFFICERS

Henry Clay Frick, II
Robert G. Goelet
Vice-President

John Elliott
Secretary

David Hunter McAlpin
Treasurer

Eben W. Pyne
Assistant Treasurer

SCIENTIFIC STAFF

William G. Conway
General Director

John Tee- Van
General Director Emeritus

Allen Vinegar

ZOOLOGICAL PARK

William G. Conway . Director & Curator, Birds Joseph Bell

Hugh B. House Chairman &

Assistant Curator, Mammals
Grace Davall . . Assistant Curator, Mammals

& Birds

Victor H. Hutchison . . . Research Associate in

Herpetology

Associate Curator, Birds

, , Visiting Research Fellow,
Herpetology

Charles P. Gandal Veterinarian

Lee S. Crandall . . . General Curator Emeritus

& Zoological Park Consultant
John M. Budinger . . . Consultant, Pathology

Ross F. Nigrelli . .

Christopher W. Coates

AQUARIUM

. . . Director Robert A. Morris .... Associate Curator

Director Emeritus Louis Mowbray Research Associate in Field Biology

GENERAL

John L. Miller . . Editor <6 Associate Curator, Publications

Dorothy Reville ...... Assistant Editor Sam Dunton Photographer

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . . Director and Pathologist

Martin F. Stempien, Jr Assistant to the

Director & Research Associate in
Bio-Organic Chemistry
William Antopol . . . Research Associate in

Comparative Pathology

C. M. Breder, Jr Research Associate in

Ichthyology

Jack T. Cecil .... Research Associate in

Marine Virology

Harry A. Charipper . . Research Associate in

Histology

Kenneth Gold

Thomas Goreau

Myron Jacobs

. . Research Associate in
Radiation Ecology
. . Research Associate in
Marine Ecology
. , Research Associate in
Comparative Pathology

Klaus Kallman Geneticist

John J. A. McLaughlin . . Research Associate in

Planktonology

George D. Ruggieri, SJ. . Research Associate in

Experimental Embryology & Teratology

Martin P. Schreibman

Research Associate in Fish Endocrinology

INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin . . .

Peter R. Marler .
Jocelyn Crane .
Roger S. Payne .
Richard L. Penney

. , Director & Senior

Research Zoologist
Senior Research Zoologist
Senior Research Ethologist
. . Research Zoologist

, . Research Zoologist

Fernando Nottebohn . . . Research Zoologist

George Schaller Research Zoologist

Thomas T. Struhsaker . . . Research Zoologist

C. Alan Lill Research Fellow

O. Marcus Buchanon . . . Resident Director,

William Beebe Tropical Research Station

William G. Conway
Lee S. Crandall

EDITORIAL COMMITTEE
Fairfield Osborn
Chairman

Donald R. Griffin
Hugh B. House
Peter R. Marler

John L, Miller
Ross F. Nigrelli

ZOOLOGIC A

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 52 • ISSUE 2 • SUMMER, 1967

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

4. On the Survival Value of Fish Schools. By C. M. Breder, Jr 25

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park, Bronx Park, Bronx, N. Y. 10460, and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers, $1.50, unless otherwise stated in the Society’s catalog of publications. Second-class
postage paid at Bronx, N. Y.

Published September 22, 1967

4

On the Survival Value of Fish Schools

C. M. Breder, Jr.

The American Museum of Natural History
and

Cape Haze Marine Laboratory

Introduction

THE QUESTION of whether the typical
schools or other groupings of fishes have
survival value has often been raised, but
few investigators have gone into the matter in any
depth. One approach, which might be called the
anecdotal or the naturalist’s approach, is usually
given as a general verbal interpretation, based
on simple observation. Another, which might be
called the mathematician’s approach, is typically
given as a rigorous analysis of a schematic ab-
straction of a fish school, usually as an over-
simplification, in which prey and predators are
considered as making more or fewer encounters,
based primarily on random movements. The
first may be exemplified by Breder and Halpern
(1946), Hiatt and Brock (1948),Sette (1950),
Springer (1957), Milanovskii and Rekubratskii
(1960) and von Wahlert (1963). The second
may be illustrated by Brock and Riffenburgh
(1960) and Olson (1964). There is, of course,
merit in both these approaches, but neither, by
itself, would seem to be adequate to develop a
full understanding of the phenomenon. A third
approach would, of course, be the experimental
one, but there have been only two reports
directed toward the possible significance of
schools to survival (Williams, 1964; John,
1964). The recent great activity in the study
of schooling, on aspects other than possible sur-
vival value, has nonetheless useful data to con-
tribute to this subject.

The primary purpose of the present paper is
to indicate clearly that all fish schools are not
necessarily similar structures, nor that they
could be encompassed in a single formulation.
A considerable amount of material has been
examined and various theoretical considerations

have been drawn into the present study. This
treatment makes it possible to show, at least at
minimum, some of the complications necessar-
ily involved in any attempt to assign a specific
survival value to a given fish school under defi-
nite conditions of existence.

Valuable assistance has been freely given by
Dr. Eugenie Clark on matters concerning visi-
bility and certain aspects of assemblage and
by Dr. William N. Tavolga on features of un-
derwater sound and its consequences. The com-
plete manuscript has been critically examined
by Dr. James W. Atz. Drs. Donn E. Rosen and
W. N. Tavolga examined those sections in detail
pertinent to their interests. To these people the
author is grateful for the help rendered.

Definitions

As in most fields that are undergoing rapid
growth there is considerable variety in the usage
of words and terminology. This is a normal
symptom of an active and changing field. It is
brought about primarily by differences in the
interests and purposes of the earlier writers on
the subject. Evidently there are still too many
new facts and ideas developing to expect an
early stabilization or general agreement on usage.
Thus it behooves all workers in the area to indi-
cate scrupulously just how they are using any
terms that could possibly lead to confusion and
misunderstanding. Also readers should use great
care to be sure that they understand an author’s
precise meaning. In addition to definitions in
this section, differences in point of view and us-
age are indicated wherever clarification would
seem needed.

The word “school” has a long history of com-
mon usage in connection with fishes and many

25

26

Zoologica: New York Zoological Society

[52: 4

dictionaries give as a definition, “a large number
of fish swimming together,” or some equivalent.
The connotation would ordinarily be that if they
were swimming together they would be going
in the same direction, as opposed to churning
about or simply resting. Parr (1927), Atz
(1953) and Breder (1929 through 1965) have
used the word essentially in the sense of the
ordinary dictionary definition. Spooner (1931)
attempted to restrict the use of “school” to cover
only social groups, as opposed to groups drawn
to one place by non-social influences. Certainly
there is no objection to redefining a word for
technical purposes when such a modification of
usage is justified. However, it is seldom possible
to determine what motivations are effective in
the formation of a school. In most cases it is
difficult or impossible to define what determines
the formation of any type of fish group. There
is always at least a residue of both a social and
non-social influence present for the fishes must
at least tolerate each other and must be located
where they are because of non-social influences,
such as temperature, nearness to surface or bot-
tom, light, et cetera.

It is partly for the above reasons that the more
nearly objective and always recognizable mea-
sure, not concerned with what drives the fish
may or may not have, has been used here. Wil-
liams (1964) objects to this usage stating, “It
may be of some value to distinguish these two
phases of activity,1 but the difference between
social and non-social groupings is in greater
need of terminological distinction.” This, of
course, is a measure of the difference between
two approaches, needs and purposes. It very
nicely illustrates the point made in the preceding
comments. The field is still in such an uncon-
gealed state that it is possible for two very
thoughtful papers (Brock and Riffenburgh,
1960; Williams, 1964) to express essentially op-
posite points of view.

Milanovskii and Rekubratskii (1960) who
consider “. . . schooling behavior as one of the
adaptive features of a population of a single
species . . .” use the word “school” in an even
broader way than does Williams and indicate
that the usages of Parr, Keenleyside (1955)
and others are “one-sided” and “contain mech-
anistic elements.” At least there is agreement in
the present paper with their comment, “At any
stage of elaboration of the problem we find it
necessary to have a working hypothesis— tenta-
tive definitions of school and schooling behavior,
which should be based on the present level of

1 That is, “schooling” and “aggregating” in the sense
used here. This footnote mine.

our knowledge.”

Williams (1964) develops the idea that
schooling and aggregating are basically rooted
in a tendency to hide behind something, as a
response to “fright.” He carried out experi-
ments bearing on this idea, with a number of
species of essentially aggregating types of fishes,
as did John (1964) on Astyanax. Both found
that in a “blank” environment their fishes tend-
ed to stay together and formed aggregations or
“fright schools.” With these experiments there
is no disagreement, but an examination of them
may help to further clarify the different usages
and attitudes toward the word “school.” If the
correspondingly opposite experiments be made
of placing permanently-schooling fishes in an
environment of abundant and varied cover these
fishes will not hide behind anything, even if
completely isolated from others of their kind.
They merely go into a period of fast and erratic
swimming, evidently in search of companions
—behavior that looks surprisingly like “panic.”
It is not uncommon for them to exhaust them-
selves, collapse and promptly expire. For this
reason the term “obligate schoolers” would seem
to be appropriate in contrast to fish that may
be called “facultative schoolers.” Under such
conditions of isolation, obligate schoolers will
attempt to school with practically any fish, soli-
tary or not, that may be presented. These may
be very unlike, for example, Mugil “schooling”
with Canthigaster (Breder, 1949). Evidently it
is the motion of another swimming fish that in-
duces the otherwise isolated obligate schooler
to react, while they do not respond at all to
inert objects. Formal experiments hardly seem
necessary in this connection, as the action seems
to be entirely evident. The attempt to experi-
ment with these extreme types is, in any case,
difficult. They are notoriously difficult to even
establish and keep in aquaria. This is the prin-
cipal reason why scombriform, carangiform or
clupeiform species are seldom seen on display
in public aquaria.

Williams performed his experiments on An-
guilla rostrata (LeSueur), Hyphesobrycon flani-
meus Myers, Notropis antherinoides Rafinesque,
N. stramineus (Cope), Pimephales notatus
(Rafinesque), Xiphophorus hybrids, Poecilia
reticulata Peters, Lepomis cyanellus Rafinesque
and Colisa lalia (Hamilton-Buchanan) Not

2 Breder and Halpern (1946) and Breder and Roem-
hild (1947) performed somewhat related experiments in
which they analyzed the statistical deployment of a num-
ber of similar species of fishes, none of which were obli-
gate schoolers. All such work on aggregating forms,
while useful, is not adequate to determine the behavior
of obligate schoolers.

1967]

Breder: On the Survival Value of Fish Schools

27

one of these is an obligate schooler. They may
form aggregations in the non-polarized sense,
fright schools, schools in rapidly-flowing water
or other facultative assemblages. If this was all
there was to the matter no one would have
thought to differentiate chronic schoolers from
the others. It is here that the confusion about
this term and its usage arises. In an attempt to
clarify the present point of view, the following
details are brought together to enable a perhaps
clearer separation of the obligate from the
facultative.

To be considered obligate, schoolers must be
coherently polarized; can only be forced to stop
schooling momentarily, and then only by means
of considerable violence; and will not maintain
a state of random orientation. The group is per-
manent, excepting only when physical condi-
tions in the environment suppress the function-
ing of some essential system, usually the optical,
as on an extraordinarily dark night. Isolated
members display erratic locomotion and com-
monly cannot exist for long in the solitary
state. The drive to associate with others in a
body of great unanimity of orientation is clearly
a positive matter of great strength, quite unlike
the fragile schools of fright, or other temporary
mutual orientations seen in fishes which other-
wise are found more commonly in non-polar-
ized aggregations or as solitary individuals. For
fully evident mechanical reasons only schooling
fishes are able to form fish mills, a type of cir-
cular swimming which occurs regularly in obli-
gate schools. The non-polarized aggregations are
fully unable to form the mill structure. See
Breder (1965) for an extended discussion of
this phenomenon.

In the terminology proposed by Williams,
“school'’ refers to any group of fishes “. . . that
owes its persistence to social (but not sexual)
forces” and “aggregation” refers to . . groups
that arise by individuals independently seeking
the same localized conditions”, which is in
agreement with Spooner (1931). Probably all
groups contain an element of both “school”
and “aggregation” in the above usage, except
the obligate schoolers, as here used. That is,
the obligate schooler is so locked to its fellows
that it ignores other things in its environment
to a remarkable extent while the facultative
schooler clearly is more actively involved with
other environmental details, often ignoring his
fellows to the point of losing its group alto-
gether. Of this Williams was aware when he
wrote that . . . when one observes a dense con-

centration of the same species moving about in
a pelagic or other uniform habitat ... he is
probably safe in calling it a school . . . ,” but
that “schools and aggregations cannot be as con-
fidently distinguished in heterogeneous environ-
ments, and it must often happen that groups
are formed that owe their cohesiveness to both
schooling and aggregation in mutual reinforce-
ment (heterogeneous summation of Tinbergen,
1951).”

Thus it appears that what would seem to be
two very different positions are not as far apart
as might be thought for, in many cases, if not
all, by designating species by either system a
very similar listing would develop. That is to
say, what are designated as schools in the
present view are assembled on a great prepon-
derance of social tendencies, while aggregations
are assembled with a far greater content of gen-
eral environmental influence. This is precisely
the view propounded by Williams, and in an
area where so little is yet known, may be very
useful as a first approximation on what holds
the group together, that is, primarily social in-
fluences or non-social influences.

Analysis of Pertinent Details

The treatment of the available data in this
section has been broken down into several sub-
sections, bringing together the controlling in-
fluences of the environment and their effects
under various conditions of predation.

The influence of environment

A suitable point of departure is a considera-
tion of the sensory modalities that are dominant
in schooling fishes and the effects of various en-
vironmental influences on their functioning.

Visibility and transparency of water

It has been abundantly shown that vision is
necessary for the formation and maintenance
of fish schools, see for instance, Parr (1927),
Atz (1953), Breder (1959), and Blaxter and
Parrish ( 1965). Also, blind fish and fish in total
darkness are unable to maintain this highly
polarized arrangement. Obviously the trans-
parency of the water is of great importance
to any behavior so largely dependent on vision.
This was clearly recognized by Brock and Rif-
fenburgh (1960), in connection with vision’s
role in school maintenance, when they wrote,
“A consideration of the optical peculiarities of
water is pertinent in this connection. The dis-
tance an object of given size can be seen de-
pends upon two factors: the intercept angle at
the eye and the contrast difference between the

28

Zoologica: New York Zoological Society

[52: 4

object and the background. Due to backscatter
and light absorption an object of high contrast
will fade from sight regardless of size at a rela-
tively small distance, say 200 feet or less, even
in the clearest water. This means that for ob-
jects above a fairly moderate size, large enough
to give an intercept angle adequate for effective
vision at the distance where light absorption and
backscatter reduce contrast difference to a point
of invisibility, taken at 2 per cent, for man
(Duntley, 1952), any increase in size of the
object will not effectively increase the distance
at which it may be seen. The critical intercept
angle for the human eye is taken to be one min-
ute which would occur for an object 0.72 inches
in diameter at 200 feet.” This obviously gives
a measure of the extreme visibility range under
water, which in most places is not even closely
approached. It may be that this estimate, al-
though based on Duntley’s paper, is too high,
for he, in another place, wrote, “It is expected
that water having hydrological range3 as great
as 130 feet will be found in the Sargasso Sea
and in the Mediterranean.” In a personal com-
munication, Dr. Eugenie Clark estimated that
horizontal visibility as great as 180 feet occurred
off the Caribbean coast of Yucatan.

Before distances as great as those mentioned
above are brought into the discussion, there are
considerations involving the geometry of fish
schools operating completely within the area
of full visibility, which can properly be dis-
cussed at this point. Since fish often tend to
accumulate into “balls,” see for instance Breder
(1959), they thereby also tend to occupy the
minimum space and show the least surface area.
This is also done by a droplet of fluid for purely
mechanical and geometrical reasons. The result
is to incidentally produce a figure of least con-
spicuousness and therefore to possess some pre-
sumed selective value. This we might call “pri-
mary selective value,” in which the direct re-
sponse to a stimulus, which may be a simple
physical condition, produces a result of definite
selective value. Viewed this way, it follows that
departure from the spherical form may be taken
as a measure of the extent to which other in-
fluences make the fishes independent of this or

3 Duntley (1952) defined “hydrological range” as fol-
lows. “The clarity of water can usefully be specified in
terms of hydrological range (v). This is the distance
measured along the path of sight, at which the apparent
contrast of any object seen against a deep water back-
ground is reduced to two per cent of its inherent value.
Along a horizontal path of sight hydrological range (v„)
is related to the transmittance (T) of the water (as mea-
sured by a hydrophotometer) by the equation
T = e-3.912x/v0 (4.1)

where x is the distance from the object to the observer.”
For the derivation of this expression see the original.

other similar constraints. In its place come oth-
er constraints, which may be thought of as “sec-
ondary survival values.” These are, of course,
the types of selective activity ordinarily referred
to as simply “selective values,” by evolutionists.
When the primary and secondary selective pro-
cesses both press in the same direction, it is often
difficult, if not impossible, to clearly separate
them, but it is here that one would expect the
development of great stability of behavior or
structure or whatever the selective processes
have been directing. Anyone familiar with
schooling fishes can attest to the strength and
rigidity of the habit. Departures from it are
clearly associated with special circumstances.
Some of these may be considered merely various
deformations of a primary tendency toward a
globular school, or even a non-polarized aggre-
gation, for in this feature both schools and ag-
gregations show similar tendencies.

Deformations may be related to groups form-
ing close to the water surface and spreading out
like a rising globule of very viscous oil. A simi-
lar case, on the bottom, would be like a globule
of heavy oil spreading out. In very shallow
water both surface and bottom would exert de-
formative influences. Also elongate schools are
normally associated with fish migrating or under
other kinds of highly directional travel.

As a more generalized concept of the geo-
metry of schools, their size and the restrictions
of lateral visibility in water, the following situa-
tion may be postulated. Given a case where a
single individual, prey or predator has a useful
visual range of, say, 30 feet, each solitary indi-
vidual fish may be considered at the center of
a sphere with a 30-foot radius. This is too much
of an oversimplification, however, for the re-
strictions on vision from above and below are
somewhat less than in any horizontal direction.
In the case of looking down, an object below
is more fully illuminated than any side view of
one at the same depth. In the case of looking up,
the fish is silhouetted against the illumination
from above. The resultant increase of visibility,
both up and down, increases the visual range
vertically to an extent determined by turbidity,
light angle, et cetera, except for the following
facts. These differences in visibility, owing to
direction, are the precise ones that are mini-
mized by countershading. In most clear open
waters countershading is notably efficient. Con-
sequently it is more nearly correct to think of
an individual fish as at the center of a geometric
figure approximating a very slightly prolate el-
lipsoid with its long axis vertical and the hori-
zontal axis, coinciding with that of the fish,
longer than the transverse axis.

1967]

Breder: On the Survival Value of Fish Schools

29

As a matter of simple geometry, several prop-
ositions follow. One fish or a “school” of two
has practically the same lateral range of vision
and there is little increase in the ability of two
fish, forming a “school,” to detect a predator,
over the ability of a single fish. This is because
the “inner side” of each is either blocked by its
companion or, if not, their fields of view are
almost completely duplicative. As a school in-
creases in the number of fishes the range of
vision increases proportionately to the area of
the side presented. As both prey and predators
wander about, there is thus twice the chance of
an encounter with two single fish, not encroach-
ing on each other’s field of vision, as with two
fish together in a school.

The above is precisely calculable and is in-
dependent of concerns of Duntley (1952) so
long as the fishes do not wander beyond their
mutual visibility ranges. Examples cover only
certain fishes indigenous to very shallow water,
or living near the surface where illumination
is not notably attenuated. Here small fishes such
as Jenkinsia or Sardinella are often preyed upon
by immature carangids and Sphyraena or ma-
ture Strongylura. The schools may be large, up
to over a thousand or more individuals, with
the predators cruising about with the prey in
full view. The predators may be typically soli-
tary ( Sphyraena ) or in small bands themselves
( Strongylura and Caranx). Any of them may
strike into a school and pick off their prey at
will, either alone or as two, or rarely a few,
actively-feeding predators. In the above named
fishes, multiple attacks are most common in the
carangids. Presumably the predators under such
situations are generally filled to satiation. Field
observations have shown that individuals com-
ing in from some distant point beyond the range
of visibility, and new to the school of prey,
usually pick off a few fish and then rest idly
nearby. From then on it is only occasionally
that one will dash in to take a single fish, with
extended idle intervals between. The length of
these intervals is presumably a measure of the
degree of digestive satiation which an individual
predator has reached. The situation above de-
scribed is one that can be generally found in
regions where such fishes abound and is ap-
parently the normal circumstances under which
they usually exist. This could be conceived of
as the degenerate limit of the situations involv-
ing no limitation on visibility, as earlier dis-
cussed. Here the predators are never under pro-
longed hunger and escape of a school unscathed
never occurs. Also here the maintenance of a
population of prey species must depend more
on reproductive potential or continued recruit-

ment from “safer” environments, with little or
no dependence on locomotor activity for escape
in flight. However, even within the limits de-
fined, the least healthy, alert or most awkward,
would on the average, be systematically elimi-
nated. From the standpoint of selection theory,
this in itself could be valuable to the long-time
survival of the population.

Sound production and its prevention

The problem of sound production by the
swimming efforts of schooling fishes or their
predators is presently unclear for several rea-
sons, see for instance Winn (1964), Wodinsky
and Tavolga (1964) and van Bergeijk (1964).
Ordinarily most fishes make no appreciable
sound incident to their locomotor activity but
may do so on sharp turns, see Moulton ( 1960) .
His observations check well with our own in
this respect, considering that different fishes
were under sonic observation. Very little on
swimming sounds has been reported by acous-
tical students on either individual fishes or fish
schools. This is most notable in observations
made in light. It is possible that there has been
selection tending to reduce activities and struc-
tures responsible for the production of sounds.
If this is the case, then schooling fishes that are
reported to produce sounds in the dark, see for
instance Takarev (1958), Shishkova (1958),
Moulton (1960) and Marshall (1962), could
represent an overriding nocturnal specialization
toward the prevention of too-wide dispersal
under lightless conditions. To predators with
sonar echo mechanisms, such as porpoises, fish
sounds or their absence would apparently make
little difference, if any. These forms are able
to feed by locating fishes by means of their
echo-ranging mechanisms alone, the data on
which is summarized by Norris (1964).

The above should not be interpreted to mean
that a complete silence is present in a school
of fishes, but only that its magnitude is too
small to be effective at distances under which
predators have to operate. The sounds noted by
Moulton ( 1960) when sharp turns are made by
fish schools are evidently only produced under
some fright-inducing stimulus. This means only
that fish already sensing the near presence of
a predator in their locomotor escape efforts, ex-
ceed some physical limit above which higher
sound levels are reached. This occurs at a time
when quietude is evidently no longer as im-
portant as flight.

Tavolga, in a personal communication, wrote
as follows about the quality of sounds produced
by a “smoothly” moving school, “The quality
of this noise is interesting in that it would tend

30

Zoologica: New York Zoological Society

[52: 4

to be random since all the fish-tail movements
are not perfectly in phase. Such a noise might
tend to be masked by ambient noise. Therefore,
even if a predator might be in the range of this
school noise, he might perceive it as only a slight
increase in ambient noise level, as might be
produced by wave action or some other physical
phenomenon.” These sounds would, of course,
be quite different from the various nighttime
sounds described by authors, often as clicks or
taps, and which are clearly not sounds made in-
cidental to locomotion.

A point to be considered about the above is
related to the information provided by the sonar
instruments such as those used by anglers to
locate fishes. These devices, because of the Dop-
pler effect, provide not only an indication of the
presence and species but also an estimate of the
size of the fish or fishes and the numbers
present. This information is based on the pulsa-
tions provided by the motion of swimming fish-
es, which are characteristic for most species and
sizes. Of course, the reflected high frequencies
used by these instruments, brought down to the
audible range by electronic means, are not iden-
tical with the low frequency, faint sounds pro-
duced by the fishes themselves. However, if
these are audible at all, they must have a beat
basically similar to that of the ultra-sonic re-
flected frequencies. It is certainly true that many
schools are so lacking in swimming synchroniza-
tion that only a broad band of low frequency
noise could be expected. However, schools vary
from those in which the individuals are com-
pletely out of phase to those that have well over
50 per cent of the members in good swimming
synchronization. Occasionally small schools,
usually of not more than a dozen individuals
as seen in various species of Mugil, Caranx and
a variety of scombrids, are clearly in near per-
fect phase. Schools, other than the ones lacking
any substantial synchronization, would intro-
duce a type of “noise” containing a beat, more
or less masked, but which should be able to
convey information to a predator, including es-
timates of species, size, number and direction
of travel. These thoughts introduce an unex-
plored area, including the extent of synchroni-
zation in fish schools, the reasons for its pres-
ence or absence and a study of its sonic product,
including volume and characteristic beat. All
this should be amenable to an instrumental ap-
proach. Indeed the schools without individuals
in phase may be an adaptation to the need for
the suppression of telltale sounds rather than
the other way around.

Bearing on this is the question of the ability
of fishes to detect the direction from which a

sound emanates. It has been argued by Harris
and van Bergeijk (1962), van Bergeijk (1964)
and Harris (1964) that far-field effects are
virtually non-directional for fishes, while near-
field effects are highly directional. Thus, a school
out of visual range and beyond the near-field
might not give a predator sonic cues as to its
location, but nonetheless, the sounds might
stimulate intensified ranging activities on the
part of the predator that could lead the latter
to its target on a basis of increasing intensity
of sound as it approached the school during ran-
dom searching. This is a matter distinctly dif-
ferent from following up a sound gradient, the
phenomenon whose existence has been ques-
tioned by van Bergeijk.

In order to present some idea of the areas
and limits of the near-field and far-field effects
and their somewhat complicated relationships,
the following comments and calculations are
given.

How far near-field directional cues extend
from a sound producing source will, to a con-
siderable extent, determine their utility to the
listener. This distance varies with the frequency,
being greatest at low frequencies and least at
high, and with the amount of the energy output
of the source. For instance, holding the energy
output constant, through the temperature range
at which Galeichthys emits its characteristic
“percolator”4 sound, approximately between 20
and 30°C, a frequency of 1000 Hz has the cal-
culated limit of its near field between 9 V2 and
9 3A inches from the origin, respectively. Other
values in feet follow:

Temp. Frequencies in Hz

°c

25

100

200

300

800

20

31'+

7'+

3'+

3'-

l'+

30

32'+

8'+

3'+

3'-

l'+

These relationships were calculated from the
given temperatures and frequency by means of
the empirical equation of Albers (1960).
c = 141,000 + 42 It - 3.7t2 + 110s + 0.018d,
where c = velocity in cm/sec, t = temp, in °C,
s = salinity in ppt and d = depth below surface
in cm. Using s = 34.8 ptt and d — 150 cm,
values of c were calculated for various values
of t. Changes in s and d were negligible for
present purposes and were held at the values
given, reducing the equation to

c= 145,829.7 + 42 1 1 — 3.7t2.

The values of the wavelengths were obtained
from the relationship

X = c/f

where X = wavelength and f = frequency in Hz.
From van Bergeijk (1964) the point of equal

4So designated by Kellogg (1953).

1967]

Breder: On the Survival Value of Fish Schools

31

amplitude of the pressure waves and the dis-
placement waves, from a pulsating bubble,
which he indicates as a convenient measure of
the range of the usefulness of near-field effects,
were calculated from the expression
n = X/2

where n = the distance of the point of equal
amplitude from the point of origin.5 The values
obtained are, of course, rather rough approxi-
mations, but are fully adequate for the present
discussion. The data on the temperature range
at which Galeichthys is sonic are original, hav-
ing been established for a certain locality in con-
nection with another project, only vaguely re-
lated to studies on schooling. Differences in the
attenuation of the various wavelengths con-
cerned are not significant within the spread of
frequencies here discussed (Albers, 1960). At
much higher frequencies, that is, within the k Hz.
range, there is some differential absorption, but
this is far removed from the sounds fish usually
produce. It should be emphasized, however,
that these calculations do not include the in-
fluence of the absolute energy of the original
signal which, of course, can be of great impor-
tance.

Since the range of hearing in fishes has been
calculated in general terms to run from about
100 to 3,000 Hz and the range important to the
lateral-line organs from about 20 to 500 Hz
(Harris, 1964), it follows that the statements
made here all fall within the accepted range of
fish auditory powers. Also, that when the pro-
ducer is separated from the receiver, “. . . both
near-field and far-field effects must be consid-
ered for the organs of hearing as well as the
organs of the lateral line.” At a frequency of
25 Hz, the wavelength is about 200 feet and at
1000 Hz it is about five feet.

From the preceding it should be possible to
estimate at about what distance a fish would
lose the directionality of, say, the percolator
sound by knowing the frequencies and tempera-
tures involved. Tavolga ( 1960) stated that there
was a predominance of frequencies around 300
Hz in these sounds, and his sonogram indicated
that they ranged to below 100 and above 800
Hz. If a fish loses its sense of directionality at
about the distance calculated, then if a fish was
receiving cues from a Galeichthys producing the
“percolator” sound at a frequency of 300 Hz or
higher, it would not be useful beyond something
less than three feet. However, in the spectrum
of this sound there are abundant frequencies of
200 and some of less than 100 Hz. Presumably

5 n is expressed in the same units used to measure
wavelength.

these would be considerably more attenuated
at their respective ranges which are about three
and one-half and eight feet. At a frequency as
low as 25 Hz the range reaches some 30 feet,
and one may assume that there are some effec-
tive frequencies between these two extremes, at
perhaps ten to 20 feet from the sound source. At
this distance the ability to receive directional
cues, especially at night, could be of great value,
as will be developed, especially since there is
some observed behavior of fishes that may be
accounted for by a range similar to the one
given above. In a personal communication. Dr.
Tavolga indicated that he has also observed dif-
ferences in the behavior of both “lost” schooling
fishes and predators that could perhaps repre-
sent a passing out of or into the limits of the
near-field.

Other influences

Other sensory modalities, such as olfaction
or taste, would not seem to be importantly in-
volved in the interactions of schooling fishes and
their predators, if at all, or at least there is no
clear evidence or theory which would indicate
such involvement. Brock and Riffenburgh
(1960) considered olfaction a possibility, writ-
ing, “. . . predators may attempt to remain with
a school of prey even though satiated, and it is
not unlikely that a large school of prey may
leave an easily detectable trail of odor for a
predator to follow,” but present no data to sup-
port this opinion. Skinner, Mathews and Park-
hurst (1962) concluded that the Schreckstoff
effect served to warn other members of a school.,
because “. . . alarmed fish communicate fright
by releasing a chemical substance into the
water.” This statement was questioned by Wil-
liams (1964) as follows, “Why then for com-
muncating a message for which speed of reaction
would be especially important, would fishes rely
on the slow process of chemical diffusion?” With
apparently a single exception, the Schreckstofj
reaction is confined to the Cypriniformes, an al-
most entirely freshwater order. This group does,
exhibit some schooling, usually in a facultative
form. Fishes of this group are: not to be con-
sidered as obligate schoolers. Strangely, in this
connection, Thines and Vandenbussche (1966)
indicate that in Rasbora the alarm substance
is more effective in the daytime, even in a
dark room. Pfeiffer (1962)/ has reviewed the
entire subject of the “fright reaction” and his
analysis indicates it to be rather remote from
the present problem.

Breaks in ontogeny., or more properly, points
at which step functions occur, such as, in the
case of certain fishes, pelagic from hatching,

32

Zoologica: New York Zoological Society

[52: 4

when they reach a sufficiently advanced but still
transparent post-larval stage, and encounter
shallow water, will permanently change their
attitudes, develop pigment and settle close to
the bottom. These, at this time, usually break
up their schools into single individuals or small
parties, as the life history unfolds. This type of
ontogenetic change seems to be present in a life
history where one stage is required to vanish
abruptly, so that the species concerned either
becomes a permanent schooler or abandons the
habit entirely.

The structure and size of schools

Brief reference has already been made to
the range of visibility under water and the rela-
tion of the conspicuousness of fishes to its
degree of transparency. Here a return is made
to that subject and its more immediate impli-
cations. Because of the considerable mathe-
matical difficulty of dealing with three-dimen-
sional structures of complex outline, see Cullen,
Shaw and Baldwin (1965), the case of a simple
surface type school, which is often not more
than one or two fish deep, will be discussed for
illustrative purposes.

It is not merely accidental that most fusiform
fishes, not in a school, usually face toward any
disturbance less than one that instigates imme-
diate flight. Aside from visual demands in an
animal that cannot turn its head alone, there is
an immediate reduction of conspicuousness, as
the frontal view is much less conspicuous than
the corresponding lateral aspect. Anyone who
has operated under water is well aware of the
phenomenon of having a fish effectively disap-
pear before one’s eyes merely because it had
turned so as to point at the observer. Such turn-
ing to face a disturbance is much less likely in
the case of a chunky fish such as an ostracid or
diodontid in which such a maneuver would do
little to alter its aspect. These, moreover, are
distinctly non-schooling types.6

The shape of schools

Since circles and spheres enclose the maxi-
mum amount of area or volume respectively
for a given perimeter or surface, it follows that
these or other shapes have a distinct bearing
on the conspicuousness of fish schools and ag-
gregations. For these reasons it could be argued
that the commonness of such approximations
as are found in real schools is a result of selec-
tion. As has, however, been indicated in other
connections, it happens that many non-living

6 All these comments are related to the less specifically
expressed view of Allee et al. (1949) and Allee (1951)
on the reduction of total area exposed by fishes in a
school.

systems show the same kind of behavior which
depends only on their innate cohesiveness. That
is, a drop of suitable oil in water of the same
specific gravity will be found to be spherical or
a drop of mercury on a flat surface will be found
to be a badly deformed sphere, flattened on one
side and of other curvature on the top side.
In other words, departures from the form show-
ing minimum surface may be considered as a
measure of some special influence. In this sense
the spherical schools discussed by Breder
(1959) and the flowing schools of Breder
(1951) all could be following simple physical
influences, with the first presenting the least
conspicuous form possible and the second ex-
posing a much greater area. The latter are us-
ually seen in very shallow water, commonly
shallow enough to eliminate the species’ pred-
ators. Also with the bottom and water surface
so close together only globular groups of small
size could occur, as for instance the globular
pods of Plotosus reported by Knipper (1953 and
1955) and observed and discussed by Clark,
in a personal communication. However, large
sheet-like schools can naturally “fit” most easily
into such vertically limited environments.
Where this dimension is greater, schools tend to
deepen, culminating in approximate spheres of
some bulk. Here also larger predators may swim
and view such gatherings from greater distances,
up to the point where visibility ceases and the
schools have protection not so much based on
their own geometry as on the peculiarities of
underwater vision. Springer (1957) considered
huge schools of small fishes, whose bulk at a
little distance could resemble some single large
creature, to have a discouraging influence on
possible predators. This would represent a case
where visibility instead of invisibility became
of positive advantage to the schoolers.

The problem of enormous schools

Data on details relevant to the present studies
are not yet available on the truly huge schools,
often involving many thousands of fishes, as
exemplified by the great assemblages which are
frequently formed by Clupea and Scomber.
Suggestive information, however, would seem
to indicate that they are not as uniform in their
size composition as smaller schools are usually
seen to be. It is conceivable that such lack of
uniformity may be based on the manner in
which they develop. If so, it may be that they
represent an agglomeration of all the smaller
schools in a given area. If, say, several hundred
schools, each normally uniform in size range
within itself, merged with others acceptably
similar, it could cause the assembled mass to
show a larger variation, from place to place

1967]

Breder: On the Survival Value of Fish Schools

33

within the whole group. If the combined schools
mixed sufficiently, large fish encountering much
smaller ones, a disruptive influence could de-
velop, or at least induce an internal realignment
so that the large fish were somewhat restricted
to one part of the group and the small to an-
other part, with intermediate fishes bridging
between them. Then more or less temporary
gradients in respect to size, or other character-
istics, could develop and stream about within
the group, establishing a continual movement
driven by the realignment activities of all indi-
viduals. This sort of continual adjustment, with
respect to locomotor facility is actually to be
seen, on a much smaller scale, in smaller schools,
and Breder (1965) thought that it formed the
basis of the continual small adjustments found
in most ordinary schools. This could easily lead
to a shearing action breaking up the different
size-groups into smaller, but still large schools.
Such effects may in fact be responsible for the
eventual disintegration of gigantic schools.7
Also, it has been shown by Hunter ( 1966) that
angular divergencies between school members
are greater between individuals of greater varia-
tion in size.

Milanovskii and Rekubratskii (1960) per-
formed some experiments with Phoxinus that
have an indirect bearing on the preceding com-
ments and on the amalgamation and disruption
of groups composed of merely facultative
schoolers, as follows.

“We noted that under natural conditions,
several schools of minnows which fed in the
same place, and which appeared from the out-
side to be one unit, reacted differently to changes
of the surrounding environment. In the begin-
ning of our observations, a school of small min-
nows was feeding; then a school of larger min-
nows approached cautiously, followed by the
school of largest minnows, even more cautious
and rapid than the fish of the first two schools.
All the fishes, small, medium and large, mingled
together and had we not seen them approaching
gradually we might have considered them to be
a single school. However, after some time, the
large minnows hid behind the nearest stone,
which they found somewhat downstream. From
their hiding place, they swam to the food,
grabbed it, and swam back. Such a phenomenon
of utmost cautiousness in the search for food
we designated by the term “withdrawal.” At the
slightest movement of the observer, the large

7 The finding of Allee and Dickinson ( 1954) that when

a Mustelus was as little as 6.7 per cent smaller than

another, the lesser dogfish would avoid the greater. This

does not imply aggression on the part of the large fishes.
This kind of avoidance is basic to the matters discussed
above.

minnows swam away, while the small and me-
dium-sized ones continued to feed undisturbed.
When the experimenter stretched his hand over
the feeding spot, the school of medium-sized
minnows fled while the smallest remained, flee-
ing only after the hand was immersed in the
water. Thus, fishes of three different schools
reacted in different ways to changes in the en-
vironment, while fishes belonging to each of the
three schools reacted as one whole. The natural
movements of fishes, obtaining food, fleeing in
the face of danger, etc., have definite signal val-
ues (of different orders of importance) for the
remaining fishes of the school. Among these
movements one can distinguish between search-
ing movements, alimentary movements and
movements of fear.” Also they wrote, again of
fishes in a stream, “The strongest biological
signal is the natural movement of fear. If, being
frightened by something, one or several fishes
move aside, the whole school follows them. We
tried to give the fishes food in such small quan-
tities that only one or two fishes could obtain
it. Once satiated, these specimens became more
fearful and went to shelter; they were followed
by all the other, still hungry, fishes.”8

The bearing that the various preceding notes
have on ideas concerning the survival value of
schooling is, among others, as before intimated,
that such massive groups may have adeterrent in-
fluence over approaching predators.9 However,
it is also reasonable that such an influence would
wane in a short time, to be replaced by an oppo-
site one based primarily on habituation of near-
by predators to such tremendous schools. The
slow drawing in of predators from perhaps a
considerable distance would be expected to fol-
low, because individuals of the prey species
concentrated in one place in an enormous mass
would proportionally restrict their numbers else-
where. Thus, a situation of positive survival
value could transform to a negative one, and
possibly also could become a force for the disin-
tegration of the huge group.10

8 These observations are also related to those of Breder
(1965) on the feeding of schools of very small Mugil.
The avoidance reactions these workers described is, no
doubt, caused at least partly by the general refusal of
fishes of slightly different sizes to mix.

9 Such a situation is probably related to or identical
with the “confusion effects” of Allee et al. (1949) and
Allee (1951). Also related to this is evidence that fishes
eat more when in groups than when alone (Allee, 1938).

10 According to the English translation of Nikolsky
(1963), the Russian usage is to apply “shoal” to such
large groups as those here under discussion and to limit
“school” to groups so small that presumably all mem-
bers could have visual or other contact with every other
member. In English and American usage “shoal” has
apparently always been used as a synonym of “school.”

34

Zoologica: New York Zoological Society

[52: 4

The maximum advantage, then, is enjoyed
by relatively small groups; that is, with addi-
tions of a few fish to a small group, the con-
spicuousness of the assemblage increases at a
much smaller rate than does the number of its
members. This advantage is lost, however, when
the number becomes so vast that the volume
occupied by the group, although remaining pro-
portional to the number of individuals, becomes
a conspicuous mass in terms of absolute size.

These various factors are necessarily influen-
tial in limiting the sizes of fish schools. Field
observation demonstrates that in a wide variety
of species this vague but very real “limit” is
not very large, at least under normal circum-
stances. Although Breder (1965) could find no
theoretical upper limit to the size that a fish
school might attain on a hydrodynamic basis,
such limitation may well be rooted in the aspect
here under consideration.

Williams suggests that the tendency for
schools to increase in size without limit until

. . the advantages of increased gregariousness
would be balanced by some disadvantage, such
as depletion of food in the center of a school.”
This is something that under ordinary condi-
tions would call for an extremely large school
because of the internal churning of schools, ex-
posing first one and then another of its mem-
bers to the periphery as well as the general con-
ditions of having the school move about or hold-
ing a position in a flow of water through it.

MacFarland and Moss (1967) were able to
measure dissolved oxygen within and outside
of large schools of Mugil cephalus Linnaeus.
They report that there was a reduction of the
oxygen concentration within the schools. Also
that there were areas of disruptive activity in
the locations showing the lowest oxygen read-
ings. These areas sometimes broke up into sev-
eral smaller schools. They refer such intra-
school activity to oxygen depletion, carbon diox-
ide increase and pFI reduction. As they indicate,
this could account, at least in part, for such be-
havior and may be a factor in limiting school
size on a basis of respiratory need.

Flere the problem of mill formation originally
analyzed by Parr (1927) and extended by Bred-
er (1965) is pertinent. Does mill formation
actually have deleterious11 effects on the fish
in a school or is an occasional occurrence of it
without significant effect on them, making an
interest in mills merely a matter of the mechan-
ics of its origin and eventual destruction? This

11 These could be extrinsic, possibly leading to greater
predation for instance, or intrinsic, holding the fish use-
lessly or dangerously in a place of poor feeding or other
disadvantage.

will have to remain an unanswered question, as
so far there appear to be no facts or ideas that
could begin a structure of theory building.

The relative size of prey and predator

The manner of feeding of predators on
schools would seem to have a distinct bearing
on the success of the school as a survival de-
vice. Commonly predator fishes may be seen
to dash into a school and pick off an individual
member and immediately retreat, usually swal-
lowing the fish whole. The predator seldom takes
more than one fish at a time, but returns again
and again, apparently until satiated. Typical
examples of this type of predator are Caranx,
Tylosurus and Sphyraena. This type of feeding
is probably the least disruptive and the most
conservative of the predators’ food supply.

Other manners of feeding on schools, as that
shown by Pomatomus, is destructive of much
more of the food supply than that described
above. Commonly an individual Pomatomus or
small group of them will race through a school
of smaller fishes, snapping right and left while
they go, leaving a trail of half-fish behind. Us-
ually it is the anterior end that is left, and this
probably means that less than half of each fish
destroyed becomes food for the predator.12 Simi-
lar modes of “wasteful” feeding on schooling
fishes have been described by Rich (1947) for
Xiphias , and Breder (1952) for Pristis. Wisner
(1958), however, exonerates Makaira from
such destructive activity, as flailing about with
its elongated rostral process in a school of much
smaller fishes.

In fishes the ratio of the size of prey to preda-
tor may vary widely, ranging from extreme
cases where the predator may be more than 20
million times the weight of its normal prey’s
weight, as for instance Manta preying on near
microscopic plankton.13 From this extreme the
ratio ranges to unity or even to cases in which
the prey may be larger than the predator, as in
Histrio and the extreme example of Chiasmo-
don. This range of differences in size has a
bearing on the nature of the utility of schooling.

The phenomenon of herding, for instance,

12 These mutilated fish-remains usually become food
of other types of fishes or invertebrates which otherwise
would be scavenging for other organic matter. Occas-
ionally some of them survive but are no longer members
of the schooling population. See Breder (1934) and
Gunter and Ward (1961) for records of this sort.

13 Based on a Manta of 3,000 lbs. compared to a
plankter of 0.1 oz., which is probably much too heavy
for the average plankton organism. The value given for
the difference in size is certainly minimal, possibly even
3 to 5 times too small.

1967]

Breder: On the Survival Value of Fish Schools

35

can only take place within certain relative size
ranges between prey and predator, for if the
two be of approximately the same size, the
predator’s approach becomes one of stalking,
and if the prey is vastly smaller, as above noted
for Manta, it becomes a matter of ranging
about in search of streaks of plankton where
neither stealth nor herding is involved.14

Schools, their models and discussion

The only serious mathematical treatment of
the possible protective value of schooling has
been presented by Brock and Riffenburgh
(1960). See also Brock (1962) for supple-
mentary data. This was followed by a note from
Olson (1964) who called attention to the work
of Koopman (1956a and b and 1957). The lat-
ter, which is concerned with the development
of “the theory of search” from the mathematical
approach, discusses cases involving situations
where both target and searcher are moving, as
in naval battles. Olson recognized the identity
of this with the situation of prey and predator,
especially among oceanic fishes. The contribu-
tions of both Brock and Riffenburgh, and Koop-
man are given in convincing mathematical
terms.

The usage of the word “school” by Brock
and Riffenburgh and by Olson is different from
the usage here employed, both in implication
and in context. In their usage, a school of fish
covers both schools and aggregations as here
used, irrespective of the individual orientations
or the distances between individuals, up to the
limit of the range of visibility and without ref-
erence to the drives and circumstances that cre-
ated the group.

As the equations of Brock and Riffenburgh do
not take the orientation of individuals into ac-
count, they apply equally well to either polar-
ized or non-polarized assemblages. One of the
marked characteristics of schools, in the present
sense, is that they consist of individuals spaced
a “standard” distance apart. Thus equation (28)
of Brock and Riffenburgh is applicable to
schools only when c, the distance between in-
dividuals in the group, is very small, for if it
becomes large, the polarization loosens and the
group can no longer be recognized as a closely
ordered array of fishes, all swimming side by
side in a common direction. This distance, (axis
to axis between adjacent fishes) is usually from

14 See Bigelow and Schroeder (1948) for a discussion
of herding in Alopias and Hiatt and Brock (1948) for
a discussion of it in Euthynnus. More complex prey-
predator relationships are described by Springer (1957)
for Rhincodon and others, by Fink (1959) for Porpoises
and Sardinops and by Bullis (1961) for Carcharinus
longimanus.

one-half to three-quarters the length of the in-
dividuals (Breder, 1954, 1965).

The whole possible confusion is further com-
plicated by the fact that Brock and Riffenburgh,
although dealing with “. . . assumptions . . . and
conclusions . . . not related to the observed be-
havior pattern of any particular species of fish
. . . ,” obviously are concerned primarily with
scombriform fishes, a group with which the sen-
ior author of that paper has had wide experi-
ence. These fishes form excellent material for
such studies, being one of the notable schooling
groups. It so happens, however, that as many
of these species age they tend to lose their strong
propensity to school. Consequently, at least in
the larger species such as Thunnus, the giant-
sized individuals occur as solitary fishes or at
least do not form the tightly organized schools
of their youth. Large fishes in general tend less
toward schooling than do small ones. This may
be associated with the fact that the larger the
fish, the less likely it is to fall prey to some
predator of still larger size. Certainly if school-
ing serves a protective function, the above
should naturally follow.

Lest any of the above comments be thought
a criticism of a very thoughtful piece of work,
this is to emphasize that these remarks are given
here only as a warning to the reader to beware
of possible misunderstanding because of differ-
ences in the usage of terms.

Koopman (1956a and b, 1957) divides his
work into three parts, which he describes as
follows. “I. The kinematic bases, involving the
positions, geometrical configurations, and mo-
tions in the searchers and targets, with particu-
lar reference to the statistics of their contacts
and the probabilities of their reaching various
specified positions. II. The probabilistic be-
havior of the instrument (eye, radar, sonar, etc.)
when making a given passage relative to the
target. III. The over-all result— the probability
of contact under general stated conditions, along
with the possibility of optimizing the results by
improving the methods of directing the search.”
Koopman considers much of his theory con-
cerned with the probability of situation to be a
special case of the theory of stocastic processes.
Obviously much of this has direct bearing on
predator and prey relationships, especially as
displayed by open water fishes.

Both Brock and Riffenburgh, and Olson ex-
press regret for the small amount of field data
available to compare with mathematical models.
The former wrote, “The general lack of field
data concerning the behavior pattern for a prey
species and its predator renders either the con-
firmation or refutation of conclusions reached

36

Zoologica: New York Zoological Society

[52: 4

in this paper by the elaboration of some scheme
of predator strategy rather futile.” The latter
wrote, referring to the Koopman equations,
“These are two basic equations, but to put rea-
sonable numbers in them is another matter.” For
similar reasons, no attempt will be made here
to apply any of these equations. Our intent is
to bring together the mathematical and obser-
vational aspects of work on fish schools, to pre-
sent some field observations hitherto unpub-
lished, and to give some general considerations
on the whole matter.

Although there is a large literature on prey
and predator relationships, almost none of it
is concerned with features that would seem to
have bearing on the problems of fish schools.
The work on bird flocks, such as those formed
by starlings, indicates that these are evidently
operating in a similar manner about as closely
as could be expected, considering the large basic
differences between birds and fishes, see for in-
stance Horstmann (1950).

In discussing the possible evolutionary course
of the schooling habit Williams wrote that “. . .
the lack of any apparent functional organiza-
tion is an eloquent argument for the conclusion
that the properties of schools have not been es-
tablished by natural selection on a basis of sur-
vival values.” By “functional organization” Wil-
liams means any or all specializations such as
“alarm notes,” markings displayed in flight, et
cetera. His detailed comments on the above are
followed by, “Evidence for such mechanisms in
fish schools would invalidate my position on
their lack of functional organization.” This ex-
treme position is here considered, at least, pre-
mature, as there are a number of valid instances
when just such mechanisms seem to be indi-
cated. Considering the difficulties in obtaining
adequate data and in interpreting their signifi-
cance, the slow progress in this direction is not
surprising. Relevant evidence suggestive of just
such “functional organization” is to be found
in practically all the current work on sound
production among aggregating and schooling
fishes, such as seen in Fish (1954), Kellogg
(1953), Moulton (1956, 1958, and 1960),
Tavolga (1958a, b, c and 1960), Marshall
( 1962), Stout (1963a and b), and Winn ( 1964) .
The consenus of these workers is in general that
there are two primary functions provided by the
sounds produced by fishes, evidently being
either of sexual or social significance. The evi-
dence that sound production is relevant to or-
ganization is indicated by various schooling
fish that become sonic only at night, when the
visual system is inoperable or only feebly so
(Takarev, 1958; Shishkova, 1958; Moulton,
1960; and Marshall, 1962).

Bearing on the question of functional organ-
ization of fish schools are recent, more refined
measurements of the spacing of individuals in
a school that have shown that both extrinsic
and intrinsic influences can vary these distances.
John (1966), working on Tilapia nilotica (Lin-
naeus) and Notemigonus crysoleucas (Mitchill) ,
for instance, showed that at very low light levels,
below 10-3 f. c., schools tended to break up and
that individuals served as, “. . . mutual distrac-
tions for one another and also as sources of
fright.” Previously, it had been thought that
schools in little light broke up merely because
of visual difficulties. This indicates that there is
a positive repelling factor involved, that appears
as light fades.

Hunter (1966), by means of computer tech-
niques, showed that schools of Trachurus sym-
metricus (Ayres) deprived of food swam at
greater distances from each other than did the
same fishes after feeding. Although schools and
aggregations appear to be leaderless, there are
some special cases, such as a white Carassius
being the focal point for aggregating by yellow
companions (Breder, 1959).

There is no disagreement with the William’s
view of how schools may have arisen, namely

. . that schooling could be expected to arise
in any species subject to aggregation.” In ac-
cordance with Williams’ definition of schooling,
this means, in effect, that fishes drawn to a given
area by some non-social influence may then in
some cases become social. He also wrote, “. . .
that a school is not an adaptive mechanism it-
self, but rather an incidental consequence of
adaptive individual behavior. The adaptation is
the reaction of each individual to the school.”

It seems most likely that schooling in fishes
arose from a wide variety of causes, including
some that are purely mechanical (Breder, 1965) .
Further speculations on this matter would seem
hardly to be worthwhile, until some time when
data and theory have reached a higher level of
development.

Evidently the schooling habit becomes estab-
lished because of purely mechanical or biological
reasons, but it would certainly be expected that
gene flow could re-enforce the habit, if it proved
to be advantageous to the group.

Levins (1962, 1963, 1964) expresses the idea
that the adaptive significance of gene flow is that
it permits appropriate response to long-term gen-
eral fluctuations of environment, while “. . .
damping the responses to local ephemeral oscilla-
tions.” This undoubtedly has bearing on the dis-
tribution of fish assemblages of all kinds. Levins
indicates that migration tends to increase the
above condition. It is noteworthy in this connec-

1967]

Breder: On the Survival Value of Fish Schools

37

tion that obligate schooling forms generally have
a large geographic range, produce large numbers
of young and commonly show migratory move-
ments. The population density is, of course, ex-
tremely high within the limits of the close con-
fines their schools delimit. It is, however,
extremely thin if their numbers are considered
in reference to the huge areas the schools pass
over, even more so if extensive migrations are
involved. The fact of schooling precludes nest
building or other protective reproductive modes
that presumably permit the production of fewer
young. The formation of great schools, and their
subsequent dissolution, as discussed herein, may
well exercise a regulatory role in the gene flow
of the species involved.

All that precedes in this paper could be used
to support the view that the functioning of prey-
fish schools, as well as of unpolarized aggrega-
tions, represents just another method of attain-
ing a manner of behavioral homeostasis. This
implies that schooling is effective against ex-
cessive predation through a wide range of activi-
ties, but fails when various limits are exceeded.
Backed up by adjustment of reproductive poten-
tial, all under the control of selective processes,
including those of both predator and prey, as
parts of a dynamic system, it is evidently suffi-
cient to produce a situation of considerable sta-
bility in the observed populations. While these
systems are probably not as closely controlled
as, for instance, the hydra populations of Slo-
bodan (1964), it would be extremely difficult
to attempt such analysis and experimental pro-
cedures on schooling fishes as he gives his ma-
terial. Nevertheless, it would seem that the basic
activity is similar. This view accepts schooling
as a biologically useful activity seen against the
appropriate ecological background.

Summary

1. The range of sight, limited as it is by trans-
parency of the water and the amount of light
present, governs the effectiveness of schooling
as a form of predation control, which varies
widely with environmental features.

2. The geometry of the school shape and its
motion affects the conspicuousness of schools
where water transparency permits good visibility.

3. In situations where visibility is not a limit-
ing factor, the system presents the degenerate
limit where schooling fails to protect effectively.

4. The general quietness of fish schools, ex-
cept under special conditions where some sound
may be inevitable or others in which it may be
desirable, suggests that there may have been
suppression of sound in schooling fishes, prob-
ably by way of selection.

5. The physical form and attitudes of the con-
stituent fishes bear on the effectiveness of schools
as a protective device, as do the shape and mo-
tion of them, schooling being associated chiefly
with streamlined fishes, less often with chunky
or odd-shaped fishes.

6. Sufficiently large schools may act as a re-
pellent to predators because of their size and
shape.

7. School size is related to the availability of
fishes of sufficient similarity of size to compose
a coherent group, as well as the mechanics of
flow within the group, and to this extent becomes
amenable to treatment by hydrodynamic means.

8. The size of the predators relative to the size
of the prey leads to “stalking” if the sizes are
about equal and to planktonic sifting if the prey
is extremely small compared with the predator.

9. The whole matter of schooling and aggre-
gating is looked upon as a mechanism of be-
havioral homeostasis and as such is subject to
the influences of selective processes.

Bibliography

Albers, V. M.

1960. Underwater acoustics handbook. Penn.
State Univ. Press: i-xiii, 1-290.

Allee, W. C.

1938. The social life of animals. New York,
W. W. Norton Co.: 1-293.

1951. Cooperation among animals. New York,
Schuman: 1-233.

Allee, W. C. and J. C. Dickinson, Jr.

1954. Dominance and subordination in the
smooth dogfish, Mustelus canis (Mitchill).
Physiol. Zool., 27(4): 356-364.

Allee, W. C., A. E. Emerson, O. Park, T. Park,
K. P. Schmidt

1949. Principles of animal ecology. Phila., W. B.
Saunders: i-xii, 1-837.

Atz, J. W.

1953. Orientation in schooling fishes. In Pro-
ceedings of a Conference on Orientation
in Animals, Feb. 6 and 7, 1953, Washing-
ton, D.C., Office of the Naval Research,
Department of the Navy, pp. 1-242.

Bergeijk, W. A. van

1964. Directional and non-directional hearing in
fish. In Marine Bio-acoustics, W. N. Tav-
olga (ed.), Proc. Symp. Lerner Marine
Lab., Bimini, Bahamas, pp. 281-298.

Bigelow, H. B. and W. C. Schroeder

1948. Sharks. Mem. Sears Foundation Marine
Res., no. 1, pt. 1, pp. 59-546.

38

Zoologica: New York Zoological Society

[52: 4

Blaxter, J. H. S. and B. B. Parrish

1965. The importance of light in shoaling, avoid-
ance of nets and vertical migration by her-
ring. Journ. Cons. perm. int. Explor. Mer.,
30(1): 40-57.

Breder, C. M., Jr.

1929. Certain effects in the habits of schooling
fishes, as based on the habits of Jenkinsia.
Amer. Mus. Novitates, 382: 1-5.

1934. The ultimate in tailless fish. Bull. N. Y.
Zool. Soc„ 37(5): 141-145.

1949. On the relationship of social behavior to
pigmentation in tropical shore fishes. Bull.
Amer. Mus. Nat. Hist., 94, art. 2: 83-106.

1951. Studies on the structure of the fish school.
Bull. Amer. Mus. Nat. Hist., 98: 1-28.

1952. On the utility of the saw of the sawfish.
Copeia, no. 2: 90-91.

1954. Equations descriptive of fish schools and
other animal aggregations. Ecology, 35
(3): 361-370.

1959. Studies on social groupings in fishes. Bull.
Amer. Mus. Nat. Hist., 117: 393-482.

1965. Vortices and fish schools. Zoologica, 50:
97-114.

Breder, C. M., Jr. and F. Halpern

1946. Innate and acquired behavior affecting
aggregations of fishes. Physiol. Zool., 19:
154-190.

Breder, C. M„ Jr. and J. Roemhild

1947. Comparative behavior of various fishes
under differing conditions of aggregation.
Copeia, no. 1: 29-40.

Brock, V. E.

1962. On the nature of the selective fishing ac-
tion of longline gear. Pacific Science,
16(1): 3-14.

Brock, V. E. and R. H. Riffenburgh

1960. Fish schooling: a possible factor in reduc-
ing predation. Journ. du Conseil, 25 (3):
307-317.

Bullis, H. R.

1961. Observations on the feeding behavior of
white-tip sharks on schooling fishes. Ecol-
ogy, 42: 194-195.

Duntley, S. Q.

1952. The visibility of submerged objects. Final
Report, Visibility Lab., Mass. Inst. Tech.,
pp. 1-74. Reissued Visibility Lab., Scripps
Inst. Oceanography (Chap. 1 through 4),
pp. 1-36.

Fink, B. D.

1959. Observations of porpoise predation on a
school of Pacific sardines. Calif. Fish and
Game, 45: 216-217.

Fish, M. P.

1954. The character and significance of sound
production among fishes of the Western
North Atlantic. Bull. Bingham Ocean.
Coll. 14: 1-109.

Gunter, G. and J. W. Ward

1961. Some fishes that survive extreme injuries
and some aspects of tenacity of life.
Copeia, no. 4: 456-462.

Hiatt, R. W. and V. E. Brock

1948. On the herding of prey and the schooling
of Euthynnus yaito Kishinouye. Pacific
Science, 2(4): 297-298.

Harris, G. G.

1964. Considerations on the physics of sound
production by fishes. In Marine Bio-acous-
tics, W. N. Tavolga (ed.), Proc. Sympos.
Lerner Marine Lab., Bimini, Bahamas, pp.
233-247.

Harris, G. G. and W. A. van Bergeijk

1962. Evidence that the lateral-line organ re-
sponds to near-field displacements of sound
sources in water. Journ. Acoustical Soc.
Amer., 34: 1831-1841.

Horstmann, E.

1950. Schwarm und Phalanx als iiberindividuelle
Lebensformen. Forsch. Spiekeroog Univ.
Hamburg, Stuttgart, no. 1 : 7-25.

Hunter, J. R.

1966. Procedure for analysis of schooling be-
havior. Journ. Fish. Res. Bd. Canada,
23(4): 457-562.

John, K. R.

1964. Illumination, vision and schooling of
Astyanax mexicanus (Filippi). Jour. Fish
Res. Bd. Canada, 21(6): 1453-1473.

1 966. Schooling of fresh-water fishes under vary-
ing illumination. Abstract. Bull. Eco. Soc.
Amer., Autumn Issue, Sept., p. 145.

Keenleyside, M. H. A.

1955. Some aspects of the schooling behavior of
fish. Behavior, 8(2, 3): 183-248.

Kellogg, W. N.

1953. Bibliography of the noises made by marine
organisms. Amer. Mus. Novitates, 1611:
1-15.

Knipper, H.

1953. Beobachtungen an jungen Plotosus anguil-
laris (Bloch). Veroffentl. uberseemus.
Bremen, 2(3): 141-148.

1955. Ein fall von “Killectiver Mimikry.” Um-
shau, 13: 398-400.

Koopman, B. O.

1956a. The theory of search. I. Kinematic bases.
Operations Research, 4: 324-346.

1967]

Breder: On the Survival Value of Fish Schools

39

1956b. The theory of search. II. Target detection.
Ibid.: 503-531.

1957. The theory of search. III. The optimum
distribution of searching effort. Ibid.: 613-
626.

Levins, R.

1962. The theory of fitness in a heterogenous
environment. I. The fitness set and adap-
tive function. Am. Nat. 96: 361-373.

1963. The theory of fitness in a heterogenous
environment. II. Developmental flexibility
and niche selection. Ibid.: 97: 75-90.

1964. The theory of fitness in a heterogenous
environment. III. The adaptive significance
of gene flow. Evolution, 18: 635-638.

MacFarland, W. N. and S. A. Moss

1967. Internal behavior in fish schools. Science,
156(3772): 260-262.

Marshall, N. B.

1962. The biology of sound producing fishes. In
Biological Acoustics, Haskell, P. T. and
F. C. Fraser (eds.). Symposia Zool. Soc.
London, no. 7: 45-60.

Milanovskii, Yu. E. and V. A. Rekubratskii

1960. Methods of studying the schooling be-
havior of fishes. Nauchnye Dok. Vysshe.
Shkoly, Biol. Nauki. No. 4: 77-81. (Eng-
lish translation from U. S. Off. Tech. Serv.,
Dept. Comm. OTS 63-11116.)

Moulton, J. M.

1956. Influencing the calling of sea robins (Pri-
onotus spp.) with sound. Bio. Bull., Ill:
393-398.

1958. The acoustical behavior of some fishes in
the Bimini area. Biol. Bull., 114: 357-374.

1960. Swimming sounds and the schooling of
fishes. Biol. Bull. Woods Hole, 114: 357-
374.

Nikolsky, G. V.

1963. The ecology of fishes. London and New
York, Academic Press, pp. i-xv, 325.

Norris, K. S.

1964. Some problems of echo location in ceta-
ceans. In Marine Bio-acoustics, W. N.
Tavolga (ed.), Proc. Sympos. Lerner
Marine Lab., Bimini, Bahamas, pp. 317-
336.

Olson, F. C. W.

1964. The survival value of fish schooling. Joum.
du Conseil., 29(1) : 115-116.

Parr, A. E.

1927. A contribution to the theoretical analysis
of the schooling behavior of fishes. Occas.
Papers Bingham Oceanogr. Coll., no I: 1-
32.

Pfeiffer, W.

1962. The fright reaction of-fish. Biol. Rev., 37:
495-511.

Rich, H. W.

1947. The swordfish and swordfishery of New
England. Proc. Portland Soc. Nat. Hist.,
4(2): 5-102.

Sette, O.

1950. Biology of the Atlantic mackerel (Scomber
scombrus ) of North America. Part II. Mi-
grations and habits. Fish. Bull., Fish and
Wildlife Serv., 51(49): 251-358.

Shishkova, E. V.

1958. Recording and study of sounds made by
fish. Trud. VNIRO, 36: 280.

Skinner, W. A., R. D. Mathews and

R. M. Parkhurst

1962. Alarm reaction of the top smelt, Ather-
inops affinis (Ayer). Science, 138: 681-
682.

Slobodkin, L. B.

1964. The strategy of evolution. Amer. Sci.,
52(3): 342-357.

Spooner. G. M.

1931. Some observations on schooling in fish.
Jour. Marine Biol. Assoc. United King-
dom, 17: 421-448.

Springer, S.

1957. Some observations on the behavior of
schools of fishes in the Gulf of Mexico
and adjacent waters. Ecology, 38(1): 166-
171.

Stout, J. F.

1963a. Sound communication during the repro-
ductive behavior of Notropis analostanus
(Pisces: Cyprinidae). Ph.D. thesis, Univ.
Maryland.

1963b. The significance of sound production dur-
ing the reproductive behavior of Notropis
analostanus (family Cyprinidae). Animal
Behavior, 11(1): 83-92.

Takarev, A. K.

1958. On biological and hydrodynamic sounds
produced by fish. Trud. VNIRO, 36: 272.

Tavolga, W. N.

1958a. The significance of underwater sounds pro-
duced by males of the gobiid fish, Bathy-
gobius soporator. Physio. Zool., 31: 259-
251.

1958b. Underwater sounds produced by males of
the blenniid fish, Chasmodes bosquianus.
Ecology, 39: 759-760.

1958c. Underwater sounds produced by two spe-
cies of toadfish, Opsanus tau and Opsanus
beta. Bull. Mar. Sci. Gulf and Carribean,
8: 278-284.

40

Zoologica: New York Zoological Society

[52: 4: 1967]

1960. Sound production and underwater com-
munication in fishes. In Animal Sounds
and Communications, Pub. No. 7, AIBS,
Lanyon, W. E. and W. N. Tavolga (eds.) :
93-136.

Thines, G. and E. Vandenbussche

1966. The effects of alarm substance on the
schooling of Rasbora heteromorpha
Duncker in day and night conditions.
Animal Behaviour, vol. 14, no. 2-3, pp.
296-302.

Tinbergen, N.

1951. The study of instinct. Oxford Univ. Press:
1-228.

Wahlert, G. and H. von

1963. Beobachtungen an Fischschwarmen. Ver-
offentlichungen Instit. fur Meeresfor-
schungin Bremerhaven, 8(2): 151-162.

Williams, G. C.

1964. Measurements of consociation among
fishes. Publ. Mus. Mich. State Univ., Biol.
Ser., 2(7): 349-384.

Winn, H. E.

1964. The biological significance of fish sounds.
In Marine Bio-acoustics, W. N. Tavolga
(ed.) Proc. Sympos. Lerner Marine Lab.,
Bimini, Bahamas, pp. 213-231.

Wisner, R. L.

1958. Is the spear of istiophorid fishes used in
feeding? Pac. Sci. 12: 60-70.

Wodinsky, J. and W. N. Tavolga

1964. Sound detection in teleost fishes. In Marine
Bio-acoustics, W. N. Tavolga (ed.) Proc.
Sympos. Lerner Marine Laboratory,
Bimini, Bahamas, pp. 269-280.

NEW YORK ZOOLOGICAL SOCIETY

GENERAL OFFICE PUBLICATION OFFICE

630 Fifth Avenue, New York, N.Y. 10020 The Zoological Park, Bronx, N.Y. 10460

Fairfield Osborn
President

Laurance S. Rockefeller
First Vice-President

OFFICERS

Henry Clay Frick, II
Robert G. Goelet
Vice-President

John Elliott
Secretary

David Hunter McAlpin
Treasurer

Eben W. Pyne
Assistant Treasurer

SCIENTIFIC STAFF

William G. Conway John Tee-Van

General Director General Director Emeritus

ZOOLOGICAL PARK

William G. Conway . Director & Curator, Birds F. Wayne King . , Associate Curator, Reptiles

Hugh B. House Chairman & Charles P. Gandal Veterinarian

Assistant Curator, Mammals Lee S. Crandall . . . General Curator Emeritus

Grace Davall . . Assistant Curator, Mammals & Zoological Park Consultant

& Birds John M. Budinger . . . Consultant, Pathology

Victor H. Hutchison . , . Research Associate in David W. Nellis Mammalogist

Herpetology James G. Doherty Mammalogist

Joseph Bell .... Associate Curator, Birds Donald F. Bruning ...... Ornithologist

AQUARIUM

Ross F. Nigrelli Director Robert A. Morris .... Associate Curator

Christopher W. Coates . . . Director Emeritus Louis Mowbray Research Associate in Field Biology

GENERAL

Edward R. Ricciuti . . Editor & Associate Curator, Publications

Dorothy Reville Assistant Editor Sam Dunton Photographer

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . , Director and Research Harry A. Charipper , . Research Associate in

Pathologist Histology

Martin F. Stempien, Jr. . . . Assistant to the Kenneth Gold Research Ecologist

Director & Bio-Organic Chemist Myron Jacobs . . . Research Neuroanatomist

George D. Ruggieri, S.J. . . . Coordinator of Klaus Kallman .... Research Associate in

Research & Research Experimental Embryologist Fish Genetics

William Antopol . . . Research Associate in Vincent R. Liguori . . . Research Associate in

Comparative Pathology Microbiology

C. M. Breder, Jr. . . . Research Associate in John J. A. McLaughlin . Research Associate in

Ichthyology Planktonology

Jack T. Cecil . . Research Associate in Virology Martin P. Schreibman . . Research Associate in

and Tissue Culture Fish Endocrinology

Martin P. Schreibman . . Research Associate in Fish Endocrinology

INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin .... Director & Senior Fernando Nottebohn . . . Research Zoologist

Research Zoologist George Schaller Research Zoologist

Peter R. Marler . . . Senior Research Zoologist Thomas T. Struhsaker . . . Research Zoologist

Jocelyn Crane . . . Senior Research Ethologist C. Alan Lill Research Fellow

Roger S. Payne Research Zoologist O. Marcus Buchanon . . . Resident Director,

Richard L. Penney .... Research Zoologist William Beebe Tropical Research Station

EDITORIAL COMMITTEE
Fairfield Osborn
Chairman

William G. Conway Donald R. Griffin Edward R. Ricciuti

Lee S. Crandall Hugh B. House Ross F. Nigrelli

Peter R. Marler

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 52 • ISSUE 3 • FALL-WINTER, 1967

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

1 . Longevity of the Naja naja philippinesis Under Stress of Venom Extraction.

By Enrique S. Salafranca. Text-figures 1-3 41

2. Combat and Its Ritualization in Fiddler Crabs (Ocypodidae) With Special
Reference to Uca rapax (Smith). By Jocelyn Crane. Text-figures 1-3;

Plate 1 49

Index To Volume 52 79

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park, Bronx Park, Bronx, N. Y. 10460, and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers, $1.50, unless otherwise stated in the Society’s catalog of publications. Second-class
postage paid at Bronx, N. Y.

Published December 30, 1967

5

Longevity of the Naja naja philip pinensis
Under Stress of Venom Extraction

Enrique S. Salafranca

Serum and Vaccine Laboratories, Bureau of Research and Laboratories, Department of Health,
Alabang, Muntinlupa, Rizal, Republic of the Philippines

(Text-figures 1-3)

I. Introduction

In order to ensure an adequate and uninter-
rupted supply of good quality venom of the
Philippine cobra, Naja naja philippinensis,
for antivenin production, the Serum and Vaccine
Laboratories (SVL) in Alabang, Muntinlupa,
Rizal, under the Bureau of Research and Lab-
oratories, Republic of the Philippines, maintains
a serpentarium. On a few occasions, requests
for venom for medical and research purposes
have been received and have been accordingly
filled. The presence of cobras in the laboratory
has every now and then attracted visitors. They
include foreigners and Filipinos, mostly students,
laboratory workers, and laymen. Some of them
seek information regarding cobra venom and
cobra antivenin production while others merely
wish to see the cobras out of curiosity.

The principal interest of the laboratory, be-
sides maintaining an adequate supply of the
venom, is to find out how long the cobras usually
live in captivity and how the stress of venom
extraction affects their longevity. It is expected
that the information gained in our serpentarium
will make it possible to prolong the life expect-
ancy of these animals and increase their life-
time venom yields under the conditions at the
SVL serpentarium. This study was conceived
and carried out for that purpose.

A review of the literature for the longevity of
cobras under the specific conditions to which
cobras kept in serpentaria for their venom are
subjected has yielded negative results. Refer-
ences encountered so far regarding longevity of

snakes are those of Baker (1951), Loveridge
( 1 946 ) , and Schmidt & Inger (1957).

Baker recorded a captive Pacific rattlesnake
(family Crotalidae) that lived for 20 years, but
doubted that wild ones could reach such age.
He further stated that few diamondbacks (genus
Crotalus) live past the first winter. Loveridge
states that apparently the record of longevity for
cobras kept in captivity is held by the two speci-
mens which lived for 13 years at the New York
Zoological Park. Schmidt and Inger state that
the greatest age on record for the reticulated
python (Python reticulatus) is 21 years; for the
Indian python (Python tnolurus), 17 years; for
the African python (Python sabae), about 15
years. In captivity, the diamond python (More-
lia argus) is known to live a little more than 4
years.

J. H. Mason,1 superintendent of the serum
department, South African Institute for Medical
Research, speaking about their serpentarium,
states that they milk their snakes once every-
fortnight and that “the death rate is rather high
as is usual when snakes are crowded together
and milked frequently.”

Cobras, like most wild animals, are very
nervous especially when confined in limited en-
closures in relatively great numbers. As in the
case with our serpentarium, where they have to
be maintained to ensure a continuous supply of
venom for immunization, many have even re-
fused to eat and literally starved themselves to

iPcrsonal communication, 1957.

41

42

Zoologica: New York Zoological Society

[52:5

death. The presence of men, cobra tenders and
visitors, and frequent handling for venom ex-
traction all contribute to increased irritation.
All these factors, especially the latter, seem to
exhaust them and shorten their lives.

II. Materials and Methods

A total of 2,075 cobras, Naja naja philip-
pinensis, were observed to supply the data for
this paper. These were collected during the
period December 3, 1959, through April 16,
1963. Except for occasional specimens caught
within and around the laboratory premises, all
were from the province of Camarines Sur,
Island of Luzon.

No attempt was made to determine the ages
of the specimens used inasmuch as no informa-
tion which may be used as a practical age in-
dicator was available to the author. Only speci-
mens 36 inches or more in apparent good health
and with intact fangs were used in this study.
These were caught in a manner to avoid any
injury to them. The maximum size noted was
59 inches with an average length of 42.73 inches
among 393 specimens measured at random.

A. Care and Management

The care and management of the cobras at
the SVL serpentarium conform with the facilities
available and the specific purpose for which they
are maintained.

1. Housing— The laboratory maintains two
roofless enclosures for housing these snakes.
The larger one measures 44.3 feet by 27.3 feet,
surrounded by a four-foot-high concrete wall
and five-foot-high reinforced half-inch mesh
wire screen atop the concrete wall. The inner
surface of the wall inclines inward so that a
plumb line dropped from the inner surface of
the upper end of the wall falls about four inches
from the inner surface of the bottom of the wall.
The walls have a two-and-a-half-inch overhang
all around the inside.

In the center is a rectangular tiled trough, 77
inches by 58 inches and 5 inches deep, which is
filled with water from a fountain. Here the
snakes usually water themselves after venom
extraction and at other times. Along the base
of the walls inside the enclosure is a concrete
strip 15V2 inches wide. On this are distributed
blocks 16 inches by 8 inches by 4 inches in
dimension. Each block has twin hollows, 2
inches by 5V6 inches by 8 inches. In these hol-
lows, where it is generally cool and dark, the
snakes remain most of the time. The rough sur-
face and corners of the blocks provide conven-
ient anchors for their skin when they start to
molt.

The smaller enclosure measures 22.6 by 25.3
feet. The surrounding wall is 3.7 feet high with
a 9-inch overhang on the inside. The reinforced
wire screen atop the wall is 3.8 feet high.

The tiled trough in the center is 35 inches
square and 3.6 inches deep. It is filled with water
from a fountain in its center. Hollow blocks are
also provided as in the larger enclosure.

2. Feeding.— White mice are made available
on the days following venom extraction for the
snakes to feed as they please.

3. Collection of the venom.— The preparatory
steps employed for controlling the snake have
been described elsewhere (De Leon & Sala-
franca, 1956).

The cobra is grasped gently but firmly by the
neck behind the angles of the jaw, with the
thumb and index finger assisted by the remain-
ing fingers of the left hand. The tail end is con-
trolled by the third and small fingers of the right
hand, with the thumb, index, and first fingers
holding the beaker for venom collection. The
latter is a “Pyrex” 100 ml. (50 X 65 mm.) beak-
er. A rubber diaphragm, a circular piece of all-
rubber surgical sheeting or the intact portions of
discarded inner tube (26" X 2.125") of a bicycle
tire of appropriate diameter, is tautly secured
across the opening of the beaker with a piece of
strong cotton twine. The diaphragm prevents
the fangs from jamming against the wall of the
beaker during the bite and thus prevents them
from breaking off, and because the snake is not
hurt in the process it bites more fully and holds
the bite longer than with the method and col-
lection apparatus previously employed (De
Leon & Salafranca, ibid.). The diaphragm also
prevents contamination of the venom with sa-
liva, soil, and other debris present in the mouth
of the snake.

As the beaker is brought opposite the snake’s
head, the cobra generally opens its mouth and
voluntarily bites. At times it appears to be so
eager to do so that venom spurts before it actual-
ly bites (the Naja naja philip pinensis is a “spit-
ting” cobra). The fangs pierce the rubber dia-
phragm and the venom may spurt out in a fine
stream with enough force to cause previously
collected venom in the beaker to froth, followed
by a few drops, or a trickle may be noted along
the wall of the beaker, or it may come out in
isolated drops.

Venom was collected once fortnightly, except
as indicated in the following controlled experi-
ments:

I. Effect of different schedules of venom ex-
traction on longevity .—In order to determine the
effect, if any, of the frequency of venom ex-
traction on longevity, replicate experiments were

1967]

Salafranca: Longevity of the Naja naja philippinensis

43

conducted on comparable groups of cobras,
which were subjected to different schedules of
venom extraction.

In the first experiment a batch of 1 14 cobras
was divided at random into three groups of 38
each. Venom was extracted once a week from
the first group, once in two weeks from the sec-
ond group, and once in three weeks from the
third group. All the snakes were allowed to move
freely in the larger enclosure and venom col-

lected according to the schedule indicated until
all of the specimens died. Color bands were
painted on the snakes to identify the different
groups.

A second experiment and a third, following
exactly the same plan, were conducted with a
batch of 192 cobras divided at random into
three groups of 64 snakes, and another batch of
90 divided into three groups of 30 (Table I,
Fig. 1).

Table I.

Relation Between Longevity and Three Schedules of Venom Extraction

Schedule of Venom Extraction

Experiment Number

A

Once a week

B

Once in 2 weeks

C

Once in 3 weeks

Experiment I
38 cobras per schedule

39.79*

51.53

59.03

Experiment II
64 cobras per schedule

56.06

94.35

123.81

Experiment HI
30 cobras per schedule

48.70

63.63

84.40

Over-all averages

48.18

69.84

89.08

*The figures under each schedule in each experiment are expressed in number of days and represent the arithmetic
average of all the respective data collected.

INTERVALS BETWEEN VENOM EXTRACTIONS IN WEEKS

FIGURE 1 COMPARATIVE MEAN LONGEVITY OF COMPARABLE
GROUPS OF COBRAS UNDER THREE SCHEDULES OF
VENOM EXTRACTION IN THREE EXPERIMENTS

44

Zoologica: New York Zoological Society

[52:5

II. Comparative longevity of cobra under
stress of venom extraction and of those from
which no venom was extracted.— A batch of 40
cobras was divided at random into groups of 33
and 7, respectively. From the first group, venom
was extracted fortnightly; from the second no
venom was extracted. The number of cobras in
the second group was kept as small as possible
mainly for economy reasons. All were kept in
the same enclosure under the same conditions
until all had died. The comparative data are pre-
sented in Table II.

III. Death rate and time (month) of the year.—
To determine any possible relation between the
time of the year and the death rate, beginning-
of-the-month inventories were determined start-
ing with December, 1959, through July, 1963.
Any acquisition during the month was added to
the corresponding inventory to arrive at the
monthly population figure. The total number of

deaths during the month was determined and the
monthly mortality rates were computed over a
period of 43 months, January, 1960, through
July, 1963 (Fig. 3).

In going over these observations and evaluat-
ing the data collected, one must be reminded of
the possible error which a greater or lesser per-
centage of carry-over of the specimens from one
month to the other, and consequently the rela-
tive ages of the specimens carried over from
month to month, may have on the resulting per-
centages.

IV. Longevity and time of collection.— In an
attempt to determine whether the time of the
year the snakes were collected had any in-
fluence on their subsequent longevity, the data
on all the batches included in this study and
which were subjected to a uniform, fortnightly
schedule of venom extraction were compiled
and are presented in Table III for this purpose.

Table II.

Longevity of Cobras Under Stress of Extraction Compared with Those from
Which No Venom Was Extracted

Number of

Schedule of Venom

Longevity in Days

G roup

Cobras

Extraction

Minimum

Maximum

Mean

I

33

Once in 2 weeks

39

118

74.90

II

7

No venom extraction

31

226

149.14

Table III.

Longevity in Batches of Naja naja philippinensis, All Subjected to
Fortnightly Venom Extraction

Number

Date of Arrival

Total No.
Specimens

Minimum

Maximum

Mean

I

December 3, 1959

229

5

147

94.67

II

March 8, 1960

491

2

202

87.86

III

June 7, 1960

555

2

175

84.31

IV

October 17, 1960

38

13

94

51.56

V

December 22, 1960

65

41

131

91.58

VI

February 9, 1961

30

7

90

50.70

VII

June 21, 1961

55

7

91

50.70

VIII

September 6, 1961

109

12

120

64.04

IX

December 22, 1961

24

31

105

82.79

X

January 29, 1962

29

74

142

102.31

XI

April 23, 1962

32

63

114

86.96

XII

June 25, 1962

32

57

131

83.78

XIII

September 4, 1962

29

23

118

54.28

XIV

November 4, 1962

15

63

97

73.33

XV

January 14, 1963

33

39

118

74.90

XVI

April 16, 1963

38

27

146

79.34

Total

1,804

Over-all averages

28

127

76.49

1967]

Salafranca: Longevity of the Naja naja philippinensis

45

B. Manner of Reckoning Longevity

The date of arrival and the number of the
specimens in each batch were recorded. Every
day the cobra tenders inspected their charges and
noted any deaths. The first death in a given batch
is designated as cobra number one, the second,
number two, and so on. The total number of days
each survived (from the date of their arrival in
the serpentarium) was recorded. When the last
specimen in the batch died, the arithmetic mean
longevity was determined. A total of 16 batches
were involved in this study.

III. Results and Discussions

Table I summarizes the results obtained in
replicate experiments comparing the effects of
the three schedules of venom extraction on the
longevity of the cobras. The data collected in
each case are expressed as the arithmetic average
of all the respective observations. An inspection

of the averages shows a definite increase in lon-
gevity as the frequency of the extraction de-
creases.

The differences observed in longevity in ex-
periments I, II, and III between groups 1 and 2
and between 1 and 3 were analyzed statistically
and were found to be significant.

Figure 1 is a graphic interpretation of the data
presented in Table I.

The results of the experiments to demonstrate
the difference in longevity in two comparable
groups of cobras— group I, from which venom
was extracted fortnightly, and group II, from
which no venom was extracted at all— are pre-
sented in Table II. The mean longevity obtained
for group II is for practical purposes twice that
obtained for group I.

Figure 2 gives a graphic picture of the com-
bined data in Tables I and II. It clearly shows that
venom extraction adversely affects longevity and

INTERVALS BETWEEN VENOM EXTRACTIONS
IN WEEKS

FIGURE 2 LONGEVITY OF COBRAS IN RELATION TO
THE FREQUENCY OF VENOM EXTRACTION

* NO VENOM EXTRACTION

46

Zoologica: New York Zoological Society

[52:5

that the degree to which longevity is adversely
affected is directly proportional to the frequency
of venom extraction.

To better appreciate the monthly death rates
and their trends and facilitate an interpretation
on the basis of observations made, both in the
serpentarium and in the fields during hunting
trips, Figure 3 has been prepared. It gives the
monthly mortality rate percentage computed on
the basis of cumulative monthly populations and
the cumulative monthly deaths covering a period
of 43 consecutive months, January, 1960, to
July, 1963.

In presenting the following interpretations of
Figure 3 the possible source of error men-
tioned above has been kept in mind and only
those supported by observations made both in
the serpentarium and in field trips are made:

1. Considering that, generally, June to Novem-
ber is our rainy season and December to May
our dry season, we have higher death rates dur-
ing the “wet” than during the “dry” season.
From the data obtained we get an over-all aver-
age death rate of 35.3% for the former and
19.3% for the latter. Death rates are lower dur-
ing the cooler part of the “dry” season, Decem-
ber to February, than during the warmer part,

March to May, reaching a peak in the latter
month.

2. The slump in death rate in June may be ex-
plained by the cooling effect of the first mild
rains of late May and early June.

In the serpentarium the only protected spaces
to which the cobras have access are the hollows
of the cement blocks, which offer limited in-
sulation to extremes of temperature and very
little protection to drenching rains. The earliest
rains experienced during late May offer relief
from the usually scorching heat, which reaches
a peak by May. This offers a logical explanation
for the sudden drop in death rate noted. With
more and heavier rains the earth becomes soggy,
water becomes stagnant, and the hollow blocks
are drenched. The skin of the snakes gets soaked
and soft, making them, possibly, more vulnerable
to systemic and skin diseases (as indicated by
their noticeably rundown condition and dull
roughened scales during this period). As the
rainy season progresses there is noted a rise in
mortality rate which reaches a peak in September.

3. As the rains abate toward the end of the
"wet” season, a gradual decrease in the death
rate is again noted.

Table III is a summary of the data on the

FIGURE 3

SHOWING THE TRENDS IN MONTHLY DEATH

RATES

1967]

Salafranca: Longevity of the Naja naja philippinensis

47

batches involved in this study, which were sub-
jected to the uniform schedule of fortnightly
venom extraction. From the data presented,
there appears to be no relation between the
month of the year the batches were acquired
and the mean survival times of the respective
batches.

IV. Summary and Conclusions

A total of 2,075 cobras were observed in the
serpentarium of the SVL over a period of 43
months.

The data collected indicate that:

1. Longevity is adversely affected by the
frequency of venom extraction; the more fre-
quent the extraction, the shorter the span of life.

Under the conditions at the SVL serpenta-
rium, cobras not subjected to venom extraction
that served as controls lived an average of 1 49. 1 4
days. Individual longevities ranged from 31 to
226 days.

Snakes from which venom was extracted once
in 21 days lived an average of 89.08 days, or
61% as long as the controls. Those from which
venom was extracted once every 14 days lived
an average of 69.84 days, or 46% as long as
the controls. Snakes from which venom was ex-
tracted every 7 days lived an average of 48.18
days, or 32% as long as the controls.

2. Monthly death rates computed on the basis
of cumulative monthly population and the cor-
responding deaths over a period of 43 months
indicate two peaks, the lesser peak occurring in
May, or about the height of the dry season, and
the greater peak in September, corresponding to
about the height of the rainy season.

3. An over-all average longevity of 76.49 days

was obtained from the data on 1,804 specimens
uniformly subjected to fortnightly venom ex-
traction. Individual longevities ranged from 2
to 202 days.

4. The time (month) of the year a batch of
specimens was collected does not appear to bear
any direct relation to the mean survival time of
that particular batch.

5. The severe rains increased monthly mor-
tality rates more than the extreme heat experi-
enced during the period of this study.

V. Acknowledgments

The author acknowledges with gratitude the
encouragement given by Dr. T. P. Pesigan, Di-
rector, Bureau of Research and Laboratories
(BRL), in setting up an atmosphere conducive
to scientific investigations and for his editorial
assistance in going over the manuscript. The as-
sistance extended by Dr. T. C. Banzon, Medical
Specialist, BRL, on statistical analyses and the
technical assistance rendered by Mr. Gelacio
Erica, cobra tender, are likewise greatly appre-
ciated.

References

Baker, E.

1951. Familiar animals of America, Harper &
Row, New York. 300 pp.

De Leon, W. & E. S. Salafranca

1956. Phil. J. Sc., 85:4.

Loveridge, A.

1946. Reptiles of the Pacific world. The Macmil-
lan Company, New York. 529 pp.

Schmidt, K. P. & R. F. Inger

1957. Living reptiles of world. Doubleday &
Company, Inc., New York. 287 pp.

6

Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)
With Special Reference to Uca rapax (Smith)1

Jocelyn Crane
New York Zoological Society,

New York, New York 10460

(Text-figures 1-3; Plate I)

[This paper is a contribution from the William
Beebe Tropical Research Station of the New York
Zoological Society at Simla, Arima Valley, Trinidad,
West Indies. The station was founded in 1950 by the
Zoological Society’s Department of Tropical Re-
search under the late Dr. Beebe’s direction. It com-
prises 250 acres in the middle of the Northern
Range, which includes large stretches of government
forest reserves. The altitude of the research area is
500 to 1,800 feet, with an annual rainfall of more
than 100 inches.

[For further ecological details of meteorology and
biotic zones see William Beebe, “Introduction to the
Ecology of the Arima Valley, Trinidad, B.W.I.,”
Zoologica, 1952, Vol. 37, No. 13, pp. 157-184.]

CONTENTS PAGE

I. Introduction 50

II. Historical Review 50

III. Materials. Methods, and Definitions .... 51

IV. Combat in Uca rapax 55

A. Relevant Characteristics of Opponents 55

1 . Phase 55

2. Relative size 55

3. Claw of same or opposite side

enlarged 56

B. Behavior Components and Their

Morphological Specializations 57

iThis study has been supported by the National Sci-
ence Foundation (G-1316, G-21071, and GB-3600).

1. Manus pushes 57

2. Manus rubs 57

3. Dactyl slides 57

4. Heel-and-ridges 57

5. Interlaces 58

6. Taps 59

7. Grips, flings, and upsets 59

C. Associated Activities 60

1. Withdrawals 60

2. Mero-suborbital frication 61

3. Downpushes 61

4. Removal of burrow intruders 61

5. Digging or prying out of opponent. . 61

6. Afterlunges 61

D. Categories 62

1. Combats between an aggressive

wanderer and a burrow holder 62

2. Combats between two burrow

holders 62

3. Mutual combats 63

E. Duration 63

F. Postcombat Behavior 66

V. Combat in Species of Uca Other Than
U. rapax 69

VI. Discussion 70

A. Functions of Combat in Uca 70

B. The Question of Adaptive Values of

Combat Ritualization 72

49

50

Zoologica: New York Zoological Society

[52:6

C. Derivation of Components 73

D. Comparisons with Ritualized Combat

in Other Animals 73

E. Areas for Further Research 73

VII. Summary 74

VIII. References 75

I. Introduction

Male fiddler crabs sometimes seize each
other’s large claws at the climax of
a fight. Physical damage practically
never occurs, although the stronger sometimes
flings the weaker inches away or flips him al-
together upside down. Almost all combats, how-
ever, stop short of a violent finish.

The work described in this contribution shows
that most combats are so fully ritualized that
the observer can detect no element of force. In
these encounters the end activity is the rubbing
or tapping of a correlated but different part of
the opponent’s claw. Morphological specializa-
tions include ridges, tubercles, and other struc-
tures with functions previously unknown.

At times the effective meeting of parts seems
to be achieved through cooperative movements
of the less-active crab. In the most elaborate
combats the components are performed by each
individual in turn, while his partner holds still.

Preliminary observations on ritualized com-
bat in Uca were made in the Indo-Pacific
(Crane, 1966). In some species the claws,
partly engaged, vibrated back and forth with a
clicking sound audible to the observer. Detailed
descriptions were not secured in the field, al-
though motion pictures recorded the pattern. In
other forms, pits and tubercles apparently served
as deterrents to forceful linkage of the chelipeds.
These observations, made incidentally during a
study of waving display, all showed the need for
a concentrated study of combat.

The fieldwork described below was accord-
ingly undertaken in Trinidad during 1966. A
socially advanced neotropical species, rapax,
was selected as the principal subject; compara-
tive observations on other species have begun.
The resulting contribution gives some basic in-
formation on the occurrence, organization, and
results of combat and discusses its possible func-
tions.

My thanks go to Pauline Thomas for Text-
figures 1 and 3, to lulie C. Emsley for Text-
figure 2, and to Kathleen Campbell for her aid
in the laboratory and in reviewing film.

Headquarters for the study was the New York
Zoological Society’s William Beebe Tropical
Research Station, Arima Valley, Trinidad, West
Indies.

II. Historical Review

The fighting proclivities of male fiddler crabs
have long been familiar to naturalists and are
even celebrated taxonomically in the species
names pugnax, pugilator, and bellator. Com-
pared with other conspicuous fiddler activities,
however, combats are so uncommon, short, fast,
and superficially similar that it is not entirely
surprising that their patterned complexities have
been overlooked. The infrequent reports have
been only roughly descriptive and wholly un-
analytical. Examples include Pearse (1912);
Dembowski (1925); Verwey (1930); Crane
(1958, 1966).

Intermale threat postures and movements in
Uca, which certainly should be regarded as rit-
ualized fighting, have been more fully reported,
especially by Altevogt (1957), von Hagen
(1962), and Schone & Schone (1963), in addi-
tion to the references just cited. This behavior is,
however, arbitrarily excluded from the bounds
of this contribution, where combat will be de-
fined (p. 53) as any behavior between male
Uca in which the claws of the chelipeds come
into contact. For the same reason, references to
waving display are excluded, although a great
deal of waving is directed at other males and in
many displays appears certainly to have devel-
oped in part from threat postures.

Intermale combat in other decapods occasion-
ally has been reported but, again, only in general
terms. It has been observed in other ocypodids,
as follows: in Dotilla, by Tweedie (1954) and
Altevogt (1957b); in Heloecius by Tweedie
(1954); in both these genera as well as in
Scopimera, Macrophthalmus, and Ilyoplax by
Crane (unpubl.). Reese (1964), in a review of
aggressive behavior in marine invertebrates, lists
intermale ritualized combat as recorded in only
three other genera of decapod crustaceans:
the grapsid crab Helice crassa, observed by Beer
( 1959); another grapsid, Pachygrapsus crass-
ipes, reported by Hiatt ( 1 948 ) , Bovbjerb ( 1 960),
and Schone & Schone (1963); and the gone-
placid crab Hemiplax hirtipes, recorded by Beer
(1959). Reese has not yet seen ritualized com-
bat during his own investigations of pagurids
although, as in a number of other decapods,
intermale threat behavior and dominance rela-
tions occur.

Study of the acoustical behavior of Uca is
now progressing (Salmon & Stout, 1962; Alte-
vogt, 1962, 1964; von Hagen, 1962; Salmon,
1965; Crane, 1966), but the sounds of combat,
which are occasionally audible to man, have yet
to be examined. No tape recordings were made
during the present study, although the above
references form part of its background. The re-

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

51

view by Guinot-Dumortier & Dumortier ( 1960)
on the morphology of stridulation in crabs gives
no examples where one crab is presumed to
stridulate against part of another individual.

A survey of ritualization in animals— in the
sense used first by Huxley (1914) and devel-
oped especially by Lorenz (1941), Baerends
(1950), and Tinbergen (1952)— is included in
Huxley et al. (1966). Ritualized combat is spe-
cifically discussed by Lorenz (1964, 1966a,
1966b).

III. Materials, Methods, and Definitions

Unlike the casual observation of fiddler com-
bat, its ethological study is slow and inconven-
ient. Not only does combat occur uncommonly,
but its characteristics are often difficult both to
determine and to record for a number of
reasons.

First, the physical conditions under which
combat is prevalent are stringent. At least in the
tropics, combat seems frequent only near new
and full moon, around the time of a diurnal low
tide, and during the optimum hours of waving
display for the species (Crane, 1958, p. 117).
The correlation of prevalence of combat with
the moon is, it seems at present, closer than that
of waving; the closer connection may, however,
be more apparent than real because of the rarity
of combats compared with the frequency of
waving and of threat postures. As with waving,
combats are more frequent in sunshine than in
cloudy weather, other conditions being equal.

A second difficulty lies in the fact that num-
erous combats are instigated by aggressive wan-
derers (loc. cit. p. 119; present contribution,
p. 53). It is now clear that these examples are
very often irregular. In working out the normal
patterns, therefore, it is important both to see
their beginnings and to observe the subsequent
behavior of each crab long enough to determine
whether an aggressive wanderer is one of the
protagonists. This ideal procedure is often im-
possible in practice because combats usually
start unpredictably and because afterward it is
frequently difficult to keep watch on two crabs,
one of which is moving rapidly away through
a crowded population.

Third, details of activities involving the cru-
cial inner surface of the manus are invisible in
many encounters, even when the observer sits
within inches of the combat.

Fourth, since most encounters start without
warning and end after a few seconds, filming
and acoustical recording are difficult.

There seem to be no short cuts to winning
knowledge of the patterns. Until their occur-

rence and outlines were established in the field,
it appeared unwise to depend on observations
made in outdoor terraria and in the laboratory.

Preliminary attempts here and elsewhere to
record the sounds of high-intensity fighting have
not yet been successful. Both aerial and contact
microphones have been used. Little time has so
far been spent on this aspect of the work, how-
ever, since the basic activity patterns first had to
be determined.

The data in this paper, therefore, are based
wholly on fieldwork undertaken in Trinidad dur-
ing the summer and fall of 1966 and on a review
of motion pictures photographed on earlier trips
to the Indo-Pacific and tropical America. The
method of study was observation, note taking,
scoring with and without binoculars, and pho-
tography with a 16-mm. Eastman Cine-special
camera equipped with a 150-mm. telephoto
lens. Combats were timed with a stopwatch and
by mentally counting off seconds; for details see
p. 53, under “Duration.”

Most of the observations were made on Uca
rapax (Smith), with comparative work on U.
cumulanta Crane. These species were studied
chiefly near the mouth of the Diego Martin
River, west of Port-of-Spain. Active popula-
tions were selected in undisturbed areas.

All of the counts were made on a small pla-
teau on the left bank of the river. Roughly
circular, this area was about 15 feet in diameter.
The terrain was typical for displaying popula-
tions of rapax— sandy mud that, although bare
of vegetation, was adjacent to grasses and man-
groves. The principal inhabitants were mature
individuals of rapax and cumulanta; at optimum
hours the great majority of males of both species
displayed. U. rapax was more abundant on the
top of the plateau, cumulanta on the damper,
sloping sides, closer to the river and a tributary
stream. The adjacent low ground supported
flourishing populations of both species that were
wholly without displaying individuals.

The portion of the plateau under observation
was an arc of 120° with an average radius of
7 feet. About 75 to 100 rapax males displayed
within this arc at any one moment during opti-
mum display conditions. Three to five aggres-
sive wanderers were usually active daily on the
entire plateau, with one or two usually within
the observation arc. To locate and count com-
bats the arc was scanned without binoculars, by
a single swing of the head averaging about 20
seconds; the swing was repeated at once if no
combat was seen. When a combat or the strong
threat of one was located, scanning ceased until
all observational data possible had been noted
from it, with or without binoculars.

52

Zoologica: New York Zoological Society

[52:6

The descriptions and counts in the following
pages are based on 337 of these combats ob-
served in sufficient detail to give information on
one or more of the topics examined. Except
where noted in the headings, therefore, the same
combats were not necessarily included in all the
tables or in the counts and percentages given in
the text. The periods of observation totaled 30.5
hours under favorable lunar, tidal, and daylight
conditions, distributed as follows during 1966:
August 17-19 and 30-31; October 13-17; No-
vember 26-29. Excluded from these totals and
dates are preliminary observations made before
all the combat components were determined,
along with their morphological specializations,
their general sequence, and the existence of mu-
tuality. Also excluded are combats occurring
during an hour or more of observation under
good conditions after every absence from the
field, for example between the October and
November sessions; for me this period of getting
back into practice was an essential safeguard
against false and incomplete observations. For
similar reasons, no observations are counted that
were incidentally made while photographing.
Finally, combats are excluded that occurred
during the first 15 minutes of every observation
period; it appeared that this length of time, for
this population of rapax, was more than enough
for the crabs to become habituated to the ob-
server’s presence, so that their subsequent be-
havior was unaffected. (In disturbed populations
and always when observing very large species a
blind would be highly advisable for studies such
as this.)

Tables I-VIII (pp. 55, 56, 60, and 64-67)
show the number of observed examples of vari-
ous kinds of combats and their components on
which the accounts of U. rapax in the following
pages are based.

Table IX (p. 69) lists species and numbers of
combats filmed in previous years in which pat-
terns were clear enough to give data now useful.

Text-figure 1 shows the location of the mor-
phological parts of the male’s large claw. A
review of the terminology may be helpful to
ethologists unfamiliar with crustaceans.

The claw consists of the two terminal segments
of the heterogonically enlarged first leg of the male.
This entire appendage is termed the major cheliped;
in this paper, since the minor cheliped is not men-
tioned, the adjective “major” will be omitted except
in a few cases of possible ambiguity. The segments
of the claw are (1) the propodus, consisting of the
manus and the pollex (“fixed finger” of some au-
thors), an extension of its lower part; and (2) the
dactyl (“movable finger”), which is attached to the
upper distal portion of the manus. The dactyl and
pollex together form the chela, the grasping part

A

B

carpal cavity dactyl base ridges

Text-fig. 1. Uca rapax. Claw of major cheliped,
to show location of structures used in combat. A,
outer view; B, inner view.

of the entire claw. The word heel is used for the
outer lower proximal portion of the manus, while
gape denotes the space between dactyl and pollex.
On the upper inner part of the manus is the carpal
cavity. Tubercles of importance occur on the outer
manus, on ridges of the inner manus, and close to
the gape. Those of concern in the present study are:
(1) on outer manus: enlarged tubercles, on either the
upper or lower part of the manus, depending on the
species; (2) on inner manus (palm): an oblique
tuberculated ridge running from the proximal lower
margin to the carpal cavity and sometimes con-
tinuing around the anterior margin of the cavity to
the dorsal margin of the manus; (3) on inner manus:
one or two tuberculated ridges, approximately ver-
tical and roughly parallel to base of dactyl, the more
proximal continuing ventrally in a curve around the
base of the gape between dactyl and pollex and
continuing for a way along the upper margin of the
pollex; (4) on pollex and dactyl; submarginal tu-
bercles and marginal teeth set along the gape, on
the lower edge of the dactyl and upper of the pollex.
Small tubercles or rugosities are often present on
the upper proximal part of the dactyl.

Definitions of general terms are, for the pur-
pose of this paper, as given below. Three con-
siderations determined their selection. The first
aim was to avoid misunderstanding in the use
of words and phrases that are not always em-
ployed in the same way by all ethologists or that

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

53

for this contribution needed limitation; an ex-
ample is the word “combat.” The second aim
was to define terms that might be unfamiliar to
workers who are not ethologists. Finally, specific
combat activities and associated activities, a
number of which are recorded for the first time
in this paper, are briefly defined in the list; fuller
descriptions, in context, follow on the indicated
pages. To facilitate reference, the entire list of
terms is given alphabetically rather than in any
topical order.

Actor — In Uca combat, the individual at the
moment performing the movements of a component.

Afterlunge — Feint by a burrow holder directed
toward departing opponent. This activity is often
associated with combat. (P. 61.)

Aggressive wanderer — A male Uca in the phase
described below.

Aggressive wandering phase — The crab moves
apparently at random through a population that
includes displaying males, punctuates his passage
with threats toward them, engages them in combat,
makes superficial burrow explorations, and attempts
unsuccessfully to mate. This phase is preceded by
one of nonaggressive wandering. In rapax it is fol-
lowed directly by the display phase, a predisplay
territorial phase being apparently absent. Instead,
the aggressive wanderer begins his waving display
as soon as he occupies a particular burrow. For
further details see below, p. 62.

Burrow holder — Equivalent, in rapax, to a crab
in the display phase, always visually characterized
by waving (see below). Because of differences found
in other parts of the genus in the relationship of
territoriality to display, the term “territory” and its
derivatives are avoided in this paper, except in dis-
cussing the few combats that took place between
burrows. In rapax the burrow itself is almost always
the focus of combat, yet because of the frequent
development of ritualization into a goal activity, it
seems that strongly territorial terms such as “bur-
row defender” should not be used, even though some
combats would make them wholly appropriate.
Individuals, including females, that are not in dis-
play phase often occupy a burrow indefinitely; since
they never engage in combat and never display,
however, they are not concerned in this contribu-
tion; a serviceable term for them would be “burrow
occupants.”

Combat — A general term for any behavior be-
tween male Uca in which the claws of the chelipeds
come into contact. Because of the usually high de-
gree of ritualization, it might be preferable to sub-
stitute the word “encounter,” thus avoiding the
“loaded" words “combat” and “fight.” Since the
behavior discussed undoubtedly has an aggressive
base, frequently shows overtly forceful components,
and probably often includes pushing elements effec-
tively masked by the ritualizations, it seems permis-
sible to use all three terms. In this paper, therefore,
“combat” is selected for general use; “encounter”
appears occasionally in the discussion of fully ritu-
alized combats; and “fight” is restricted to combats
with overtly forceful components.

Component — An activity that is a characteristic
part of combat and appears to the observer to be
distinct from adjacent actions.

The most clear-cut examples are so distinct and
so stereotyped that they may confidently be termed
fixed action patterns, in the sense developed by
Lorenz and Tinbergen and now often used in etho-
logical studies. These examples are characterized
not only by distinctive motor patterns but by juxta-
position of specialized morphological structures. In
all of them overt force appears to be absent and they
are here considered completely ritualized.

Other activities, however, are far less stereotyped
and show considerable variability connected neither
with intensity nor with transition to other combat
actions. All of these, instead, often appear instantly
adaptable to the changing circumstances of a fight.
All show overt use of force since the chelipeds are
used variously as a pushing, grasping, and lifting
organ. These activities are regarded as unritualized.
Any or all may need subdivision or other modifica-
tion. Only further study with emphasis on compara-
tive work within the genus can resolve the uncer-
tainties.

Therefore it seems that the use of “fixed action
pattern” would be at present a semantic disservice.
The more general word “component” is adopted
instead, in the same spirit shown by morphological
taxonomists when they feel it premature to use a
definite term such as “subgenus” and compromise
on the noncommittal “group.”

Dactyl slide — In rapax a ritualized component in
which one crab rubs his dactyl teeth along the upper
edge of his opponent’s dactyl. (P. 57.)

Display — Here confined to a rhythmic motion
of the major cheliped, species specific within the
genus; movements of other appendages often ac-
company it. Used interchangeably with “waving”
and “waving display.”

Display phase — The phase characterized by wav-
ing, burrow holding, threat, combat, and courtship.
In rapax it is preceded by the aggressive wandering
phase.

Downpush — One crab is pushed down his own
burrow by his opponent. An activity associated with
combat. (P. 61.)

Duration — Timing of a combat from the moment
at which the two chelipeds come into contact to
their separation immediately preceding the de-
parture of one of the opponents. Associated activi-
ties, ranging from preliminary threat behavior to
afterlunges, were not timed, either as part of the
combat proper or separately. Because of the diffi-
culties of observing closely certain other constitu-
ents of each combat — especially its start, the phases
of the opponents, their relative sizes, and their sub-
sequent behavior — accurate timing with a stopwatch
finally was discarded as a relatively expendable
activity. Instead, counting seconds mentally was
substituted, as in photography. Periodic checks with
the stopwatch showed this system to be fully accu-
rate within the broad limits of the two periods that
turned out to be important to this study. During

54

Zoologica: New York Zoological Society

[52:6

future fieldwork it will be feasible, now that the
basic work has been completed, to concentrate
adequately on duration, both of combats them-
selves and of associated activities. As will be seen
from the discussion, such records now become im-
perative.

Encounter — Fully ritualized combat. (See also

combat.)

Fight — Combat including forceful components.
(See also combat.)

Fling — A variable, unritualized component at the
close of a forceful ending. One opponent is pushed
backward in a skid or is partly overturned. (P. 59.)

Forceful component — In low-intensity combat —
manus pushes; in high-intensity combat — grips,
flings, and upsets. All are highly variable and seem
to be entirely unritualized, since the cheliped is
used variously in pushing, grasping, and lifting.

Forceful end — Unritualized behavior at the end
of a high-intensity fight, consisting of a grip by one
opponent and sometimes followed by a fling or by
the total upset of the other crab.

Grip — A forceful component consisting of the
seizure by ope opponent of the other’s claw. (P. 59.)

Heel-and-ridge — A high-intensity ritualized com-
ponent in which the actor places his dactyl outside
the manus of his opponent while the pollex passes
to the inner side, the palm, and rubs its oblique
ridge. (P. 57.)

Heteroclawed combat — In one opponent the claw
on the right side is enlarged, the other on the left.

High-intensity combat — Part of the claw of each
opponent comes between the dactyl and pollex of
the other. In forceful endings the claw tips may grip
the opponent’s claw; in fully ritualized encounters
they do not do so. In rapax the available compo-
nents consist of dactyl slides, heel-and-ridges, and
interlaces; a low-intensity component, the manus
rub, often initiates high-intensity combat.

Homoclawed combat — Both opponents have the
claw of the same side, either right or left, enlarged.

Instigator — In combats between two burrow
holders, the crab that approaches his future oppo-
nent; except for this approach, he does not neces-
sarily ever become an actor. In combats between an
aggressive wanderer and a burrow holder, the
wanderer is always the instigator. Because of the
usual high degree of ritualization, the words “ag-
gressor” and “attacker” are not used.

Interlace — A ritualized component prevalent in
heteroclawed combat. With the bases of both chelae
almost in contact, proximal claw teeth on one crab
rub against a distal manus ridge of the opponent.
(P. 58.)

Left-clawed — Cheliped on left side enlarged.

Low-intensity combat — Contact is confined to the
outer surfaces of the mani and chelae. In rapax the
components consist of the forceful manus push and
the ritualized manus rub. (P. 57.)

Manus push — See Low-intensity combat.

Manus rub — See Low-intensity combat.

Mero-suborbital frication — The rugosities on the
upper merus of the cheliped are rubbed against the
outer crenelations of the lower margin of the adja-
cent orbit. Behavior associated with combat. (P. 61.)

Mutual combat; mutual components — Both crabs
perform one or more of the components, either in
sequence or alternately. They can perform simul-
taneously only during manus rubs. (P. 63.)

Phase — A temporary condition of Uca charac-
terized by one of a number of general behavior
patterns. These have already been described (Crane,
1958). Of concern in this paper are the aggressive
wandering and display phases, defined above. The
term “phase” is retained here in preference to its
possible alternate, “state,” because the latter is often
used in work with other animals for conditions
controlled by long-term endocrine activity. At least
in the tropics phases appear regularly to last only
hours or days. The causes of phases remain unin-
vestigated.

Right-clawed — Cheliped of right side enlarged.

Ritualization — Any short definition or description
of ritualization in animals must be inadequate.
Furthermore, not all workers in ethology agree on
the boundaries of ritualization or, rarely, even on
the need for the concept. To me, however, much of
the social behavior in Uca would be unintelligible
without it. It is hoped that the following paragraphs
may serve as a frame of reference for the present
study of combat. Key contributions, classical and
recent, on ritualization include Huxley (1914, p. 491;
1966, pp. 249-271), Lorenz (1941, 1964, 1966a,
1966b), Baerends (1950), Tinbergen (1952, 1953),
Blest (1961), Hinde (1966, p. 432) and Thorpe
(1966).

In animal behavior ritualization is an evolu-
tionary process whereby social patterns develop
partly through transformations of movements and
structures that originally functioned during other
activities. For example, a bird’s courtship display
movements may include recognizable traces of feed-
ing and preening behavior. In the course of evolu-
tion both the motions and the associated morpho-
logical characteristics become simplified, exagger-
ated, or both. These changes often make a visual
display, for instance, conspicuous, unambiguous,
and distinct from that of related species. In short,
they produce an effective signal in communication.

In other cases, ritualization serves to reduce intra-
specific damage in combats between males. Here
ceremonial encounters replace injury-producing use
of weapons, while the advantages of aggression are
maintained. The ritualized motions are often inten-
tion movements of actual fighting, more or less
transformed and often, as in display, emphasized
by the enhancement of associated structures. Also,
as in display, parts of sequences from other contexts
may be altered and included. Sometimes these com-
bats are token tests of strength, and damage is
avoided at the climax by stereotyped signals of pos-

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

55

ture and motions. Sometimes force seems wholly to
be absent. Although ritualization in Uca combat
obviously belongs in this general category, both its
origins and its functions remain obscure, awaiting
comparative study within the genus. They are briefly
discussed, beginning on p. 72.

Ritualized combat — Encounters without forceful
components.

Ritualized components — In rapax these include
manus rubs, dactyl slides, heel-and-ridges, and in-
terlaces. No pushing, grasping, or lifting motions are
included.

Tapping — At the end of a ritualized component
the teeth of the dactyl or the pollex or both tap
against part of the opponent's dactyl or manus.
(P. 59.)

Territorial phase — See Burrow holder.

Threat — Aggressive postures or motions confined,
in male Uca , to individuals in the aggressive wander-
ing, territorial, and display phases. Although threats
usually precede combat, they occur far more often
than combats. As in other animals they should be
considered as classic examples of ritualized combat.
Already described briefly and illustrated elsewhere
(ibid., 1958, p. 123; 1966, p. 464), they are arbi-
trarily excluded from this contribution through the
selected definition of "combat." They will be further
examined in a subsequent publication on the com-
parative ethology of combat behavior.

Upset — Unritualized, final component, where one
crab is turned completely upside down. The rarest
of all components. ( P. 59. )

Wanderer — In the present paper used only for
“aggressive wanderer,” and then only where the
latter term has already appeared in the same para-
graph.

Wandering phase — In Uca a nonaggressive phase,
characterized by lack of attachment to a particular
burrow. It precedes the aggressive wandering phase

(ibid., 1958, p. 119). Not concerned in the present
paper.

Waving; Waving display — See Display.

Withdrawal — Behavior associated with combat,
where a burrow holder goes partway or entirely
down his own hole. (P. 60.)

IV. Combat in Uca rapax

A. Relevant Characteristics of Opponents

The pattern of a combat between rapax males
depends largely on three factors: the phase of
each individual; the relative size of the two
crabs; and the location— whether on the right or
left side— of the large claw of each.

1 . Phase— Males engage in combat only when
they are in either the aggressive wandering phase
or the waving display phase. In the latter con-
dition they are always burrow holders. Combats
take place only between an aggressive wanderer
and a burrow holder or between two burrow
holders; encounters between two aggressive
wanderers are unknown. When a wanderer
elicits combat from a burrow holder irregular
components are frequent and a ritualized en-
counter sometimes ends as a forceful fight. In
contrast, combats between neighboring males
are usually composed of highly regular, ritual-
ized components and practically never end in
force. These categories and their results will be
described beginning on p. 62.

2. Relative size— The larger crab in all com-
bats usually has the advantage. It is interesting,
therefore, that in the majority of combats includ-
ing an aggressive wanderer, the latter instigates
combat with a burrow holder larger than he.
See Table I.

Table I.

Uca rapax. Relative Sizes of Opponents in Combats Where Instigator Was Determined.
(Based on a total of 151 combats. Locality and dates as in Table II.)

Aggressive Wanderer vs. Burrow Holder

Aggressive wanderer (always instigator) larger 22

Aggressive wanderer smaller 66

Aggressive wanderer & burrow holder about equal 6

Total 94 94

Burrow Holder vs. Burrow Holder

Instigator larger 36

Instigator smaller 17

Opponents about equal 4

Total 57 57

Grand total 151

56

Zoologica: New York Zoological Society

[52:6

3. Claw of same or opposite side enlarged—
In rapax right- and left-clawed individuals oc-
curred in equal numbers in two population
samples (unpubl.). In Maracaibo, Venezuela,
the sample totaled 186; in the Trinidad study
area, 252. Tables II and III show no tendency
in combat for instigators to approach opponents

having the same claw enlarged as their own
rather than the claw on the opposite side. There-
fore, homoclawed and heteroclawed combats
are about equally likely to take place. Table III
shows clearly that the frequency of several of
the components described below is affected by
this characteristic.

Table II.

Uca rapax. Composition of 154 Combats Observed at Cocorite, Trinidad, October
13-17 and November 26-29, 1966.

(Note: Combats are listed only if observation is believed to have included the first component.)

Sequences of Components Combats Between an Combats Between Total

Aggressive Wanderer Two Burrow Holders
and a Burrow Holder

Homo-

clawed

Hetero-

clawed

Homo-

clawed

Hetero-

clawed

Manus push only

1

1

1

3

Manus push + manus rub

2

2

4

Manus rub only

19

10

18

10

57

Manus rub + dactyl slide

1

13

3

6

23

Dactyl slide only

Manus rub + dactyl slide +

1

3

1

1

6

heel-and-ridge

Manus push + manus rub +

1

1*

1

3

heel-and-ridge

1

1

Manus rub + heel-and-ridge

9**t

4*

4

2

19

Dactyl slide + heel-and-ridge

1

1

1*

3

Heel-and-ridge only

5*t

5**

1

11

Manus rub + interlace

5t

5

Heel-and-ridge -f- interlace

4

4

Interlace only

2t

2

4

Manus rub + heel-and-ridge -f interlace

1

1

Dactyl slide + heel-and-ridge + interlace

It

1

Manus rub + dactyl slide -f- heel-and-

ridge + interlace

3t

1

4

Manus rub + dactyl slide + interlace

2

1

3

Dactyl slide + interlace

1

1

2

Total

40

49

37

28

154

* 1 heel-and-ridge component not followed by tapping.

* 2 heel-and-ridge components not followed by tapping,
t 1 combat with forceful ending included.

Table III.

Uca rapax. Relative Frequency of Components in 154 Combats.
(From data in Table II.)

Component Frequency (%)

In 77 homoclawed combats

In 77 heteroclawed combats

Manus push

6

4

Manus rub

78

77

Dactyl slide

14

44

Heel-and-ridge

38

23

Interlace

3

29

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

57

B. Behavior Components and Their
Morphological Specializations
(Tables II and III)

All combats are based on seven components.
These are described below, in the order of their
usual occurrence. Component 1 or 2 or both
may form the entire combat; usually 2 precedes
one or more additional elements. Components
1 and 2 are regarded as low-intensity compo-
nents, and 3 through 6 comprise the highly
ritualized, high-intensity components. The in-
frequent forceful grips and upsets are grouped
under Component 7.

As shown in Table II, all the components
never occur in a single combat. The sequence
given below is altogether characteristic, but
reversals of order among high-intensity compo-
nents occasionally take place, particularly in
long mutual combats. These reversals may or
may not be followed by repetitions of previously
performed components. The combat pattern,
therefore, is not rigidly constructed of a fixed
sequence of interdependent components. The
omissions and irregularities are such that it does
not seem wise to stylize the data so far assembled
into a diagram of a “typical” combat sequence,
or even into any kind of flow diagram. Instead,
it is felt that tables present the complexities of
the subject in a more realistic manner.

1. Manus pushes— The chelipeds are held
flexed, the chelae partly open through slight
lifting of the dactyls. Meanwhile the lower,
smoother halves of the mani are pushed against
each other. It often seems impossible to decide
whether or not a manus push includes a rub
component or vice versa, and the totals given in
the tables are approximate.

2. Manus rubs— The same surfaces as in the
manus push are rubbed back and forth, longi-
tudinally, against each other. In rapax and re-
lated species the surfaces in contact are smooth,
rather than coarsely tuberculated as in some
other sections of the genus (Text-fig. 2). It is
possible that the rubbing over the smooth sur-
face facilitates the attainment of high-intensity
components. At these times the manus rub con-
tinues past the bases of the gapes until the chelae
are in contact externally and are free to proceed
to one or more of the following positions.

3. Dactyl slides (Text-fig. 3 A)— With the
chelae of both crabs partly open, the dactyl of
one moves on top of that of the opponent, at
about middle of its length, more or less at right
angles, while the pollex passes within the gape.
The approach may be from either the inner or
the outer side of the claw. Both chelae are by
then widely opened and the pollex does not

Text-fig. 2. Low-intensity combat in Uca. Semi-
diagrammatic view of position assumed by many
species throughout the genus during manus pushing
and rubbing. During homoclawed combat, as above,
the manus of each opponent sometimes lies against
the opponent’s pollex. In heteroclawed examples the
claws are usually lined up parallel, from manus heel
to chelae tips. The dorsal margins of the mani are
sometimes, as in the figure, at unequal heights, some-
times at the same level. See text, p. 57.

touch the opponent's fingers. Gentle maneuver-
ing for this position may continue for several
seconds. Once it is achieved, the rounded teeth
of the actor’s dactyl slide longitudinally back
and forth along the smooth middle portion of
the upper edge of the opponent’s dactyl. No
force seems to be used by either crab and no
attempt is made to use the claws as pincers, all
four tips being at all times free in the air. Except
for the sliding motion both crabs remain almost
motionless. An infrequent climax is the vibra-
tory tapping of the uppermost dactyl against
the one held quietly beneath it. In this species
the vibration is performed only by the crab with
the dactyl on top. Afterward the two crabs break
suddenly apart. When tapping does not occur,
the encounter either breaks off after at most
several seconds of slide or, infrequently, passes
into Component 4 or 5 below. Dactyl slides oc-
cur more often in heteroclawed than in homo-
clawed combat.

4. Heel-and-ridges (Text-figs. 3B, 3C; PI. I,
Figs. 1, 2.)— In rapax and many other species
the dactyl, longer than the pollex, curves down-
ward beyond it. This characteristic proves to be
of definite use in heel-and-ridging, where the
dactyl arches around the curving heel outside
the manus of the opponent. On its way toward
the heel, the dactyl tip appears to feel its way,
using as a guide the outer crease along the base
of the dorsal marginal ridge. Afterward, upon
reaching the proximal part of the manus, the
dactyl tip does not touch the heel except during
the climax to be described. Meanwhile the pol-
lex, shorter than the dactyl and virtually straight,
comes into contact with the oblique tubercu-

58

Zoologica: New York Zoological Society

[52:6

A

Text-fig. 3. High-intensity combat in Uca rapax. A, dactyl slide; B and C, heel-and-ridge with tapping
in homoclawed encounter; D and E, interlace in two forms of encounter. In B the crab on the left is the
actor, his dactyl tip touching his opponent’s heel; in C the same actor’s pollex on a reverse stroke is hitting
the opponent’s oblique ridge, his dactyl tip being now separated from the heel; the same pollex position is
used during the rubbing of the oblique ridge during the earlier part of the component. The interlace shown
in D illustrates the more usual, heteroclawed form of this component; the right-clawed crab on the right is
the actor, getting into position to bring dactyl teeth against opponent’s ridge along inner base of dactyl;
see also PI. I, Figs. 3 and 4. E shows the interlace in its less common form, in homoclawed combat; the
crab on the right is the actor, rubbing basal teeth of dactyl against basal teeth and outer tubercles of op-
ponent’s pollex. See text, p. 57 ff.

lated ridge of the inner manus. The pollex teeth,
in the seven examples where an adequate view
or film was secured, rub up and down along the
oblique ridge. At the climax, however, the actor
taps the ridge rapidly three or four times with his
pollex; on opposite strokes, when the pollex is
away from the ridge, the teeth near the dactyl tip
come into contact with the manus heel. At high-
est intensity the tapping is faster and of smaller
amplitude; the effect is vibratory (p. 59).
When crabs are not closely matched in size, the
parts coincide poorly and it is the acting crab—

not necessarily by now the original instigator—
that performs the tapping. No attempt at seiz-
ing and gripping has ever been detected when
the claws are in the heel-and-ridge position. A
heel-and-ridge, with or without tapping, may be
preceded by a manus rub or infrequently by a
dactyl slide. Occasionally in mutual, hetero-
clawed combat a heel-and-ridge is followed by
or alternates with an interlace.

5. Interlaces (Text-figs. 3D, 3E; PI. I, Figs.
3, 4.)— In this component the fingers of each
claw overlap the opponent’s manus, so that the

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs ( Ocypodidae )

59

bases of the gapes are almost or wholly in con-
tact. The sequence is characteristic of hetero-
clawed combats although not confined to them.
Typically the position is assumed by the crab
that has its dactyl against the inner side of the
opponent’s manus rather than against the outer
side, which is the position for a heel-and-ridge.
In a fully developed interlace the chela of each
crab is wide open through high elevation of the
dactyl; the tips of both pollex and dactyl are
wholly free from contact with any part of the
opponent. In this position the most proximal
teeth of the pollex come into contact with one
or both of the tuberculated ridges paralleling the
base of the opponent’s dactyl and rub up and
down along their course. At highest intensity
the rub follows along the longer, less-variable
subdistal ridge that continues from the dactyl
base down around the base of the gape and out
along the proximal, upper, inner portion of the
pollex.

The climax usually consists of frication by the
pollex teeth as just described, whereupon the en-
counter ends. This component may culminate in
serial taps similar to those following a normal
heel-and-ridge component but made, instead, by
the basal gape teeth against the subdistal ridge of
the manus in the interlace position. While a tap-
ping finale is the normal end of a fully developed
heel-and-ridge component, it is uncommon after
an interlace and, as when it follows a dactyl
slide, cannot be regarded as typical (p. 57).

An interlace is usually preceded by a manus
rub and less often by a dactyl slide. It also oc-
curs in mutual heteroclawed encounters when
one opponent, with the dactyl against the inner
side of the other’s manus, performs the interlace.
Usually it is the temporarily more active crab,
in passing from a low-intensity manus rub to
high-intensity, that assumes the heel-and-ridge
position, with the dactyl outside the manus;
hence the second crab arrives automatically in
a position appropriate for the interlace.

6. Taps— As already described, tapping occurs
most frequently at the end of a heel-and-ridge
component. Sometimes it is preceded by an
interlace or dactyl slide; rarely by a manus rub.
Always it consists of the rapid tapping of the
dactyl or, in most heel-and-ridge sequences, of
the dactyl and pollex alternately, against a par-
ticular part of the opponent’s claw. Rather slow
tapping of wide amplitude is performed both
by aggressive wanderers and by burrow holders.
Rapid taps of narrow amplitude are performed
only by a burrow holder, usually at the end of an
encounter beside his own hole. The several ex-
amples recorded on motion picture film appear

as blurs on frames exposed at 1/48 second. After
tapping, a combat often breaks off abruptly.

7. Grips, flings, and upsets— In contrast to
Components 2 through 6, the occasional force-
ful end of a combat is composed of irregular
elements that appear to be largely or wholly
unritualized. They may be grouped under grips,
flings, and upsets— descriptive terms for actions
that merge into one another.

A grip occasionally follows an unsuccessful
attempt by an aggressive wanderer to get into
the heel-and-ridge position; the fingers then slip
beyond the normal position and firmly seize the
base of the manus with one finger hooked into
the carpal cavity; even the carpus itself may be
grasped. Sometimes grips occur when two crabs
are grossly mismatched in size; then the larger
may seize the entire manus of the smaller cross-
wise between his fingers. More often the forceful
component consists only of a longitudinal grip,
with perhaps an undetected push, the actor then
opening his fingers. Almost always the crabs
then separate.

The term "fling” here includes those actions,
always starting with a grip, that result in a skid
or partial upset of the opponent. The momen-
tum of the actor’s pushing grip carries the re-
leased crab sliding backward, or he is thrust off
balance with some of his ambulatories off the
ground. “Upset” is confined to actions resulting
in the complete overturn of the opponent onto
his carapace, with all of his ambulatories in
the air.

Both the occurrence and the progress of these
forceful endings have so far proved to be un-
predictable. The following figures are relevant.
In 180 combats in which the endings were ade-
quately observed, 15 (8.3%) ended forcefully
(6 of these are included in the group of 154
combats analyzed in Table II and are so indi-
cated). In 14 fights the opponents were an ag-
gressive wanderer and a burrow holder; the
wanderer was larger than the burrow holder in
7 combats, smaller in 6, while his identity was
uncertain in 1 ; homoclawed and heteroclawed
combats were equally divided. In 6 fights the
forceful component consisted of a grip only, 5
others ended in flings, and only 3 in total up-
sets. The burrow holder was the actor both dur-
ing the preceding ritualized component and in
the grip and subsequent action in 7 combats, the
aggressive wanderer in 2, 1 of them resulting in
the eviction of the burrow holder; in 2 fights
both ritualized and force components were
mutual, 1 of them resulting in the second ob-
served eviction of a burrow holder; in 1 fight
the action consisted wholly in the eviction of an
aggressive wanderer that had slipped in while

60

Zoologica: New York Zoological Society

[52:6

the burrow holder was engaged with a neighbor.
The ritualized components immediately preced-
ing forceful endings were interlaces in 7 fights
and well-developed heel-and-ridges with taps in
3; antecedents to the grip in the remaining ex-
amples were irregular or improperly observed.
Finally, 5 fights were followed by reduced ag-
gressiveness of the wanderer and 2 by the dis-
possession of a burrow holder by a wanderer.

The single forceful ending to a combat be-
tween burrow holders was very short; the slight-
ly smaller instigator lightly seized the opponent’s
entire palm then released it and went home;
both crabs promptly resumed waving.

In brief, about 1 in 12 combats had a forceful
ending, usually consisting only of a brief grasp
by one crab of the other’s manus; 14 out of 15
fights were between an aggressive wanderer and
a burrow holder; in most the burrow holder was
the active crab at the end of the fight, seizing
the wanderer and administering the final push,
fling, or upset. Forceful components preceded
reductions in the wanderer’s aggressiveness in 5
combats and preceded the taking over of his
opponent’s burrow in 2. These changes in be-
havior will be further discussed beginning on
p. 66 and are included in Table VIII.

C. Associated Activities

1 . Withdrawals— Of the 154 combats observed
most completely, one quarter included some
degree of withdrawal by the noninstigating crab
into his own burrow. These withdrawals are
subdivided in Table IV, but the degrees indi-

cated form unsatisfactory comparison material.

As shown in Table I, in the majority of com-
bats between aggressive wanderers and burrow
holders, the wanderer, always the instigator, is
usually smaller than the burrow holder. Con-
versely, in encounters between burrow holders,
the instigator, which usually starts the combat
at the mouth of his opponent’s burrow, is most
often the larger crab.

Withdrawal of an approached crab, it seems,
would be a more likely action by a crab smaller
than his opponent than by one larger than the
instigator. It would be expected, therefore, that
in combats involving wanderers there would be
little withdrawal underground, since most wan-
derers are smaller than the burrow holders they
approach. In contrast it would seem predictable
that since burrow holders are usually engaged by
larger burrow-holding neighbors, withdrawals
among the first would be more numerous.

In the material at hand as shown in Table IV
such differences are not apparent. Withdrawals
of burrow holders in combat with aggressive
wanderers are more numerous than those be-
tween burrow holders. A principal reason is that
the figures for partial withdrawals in Column 3
are misleading elements in the totals. They are
misleading because partial withdrawals are not
often functional withdrawals from combat. In
these cases the burrow holder instead clearly
uses the upper part of his burrow as a firm foot-
hold for the prosecution of the encounter. In the
samples given in the table, as well as in a

Table IV.

Uca rapax. Combats Including Withdrawal of the Noninstigating Crab Into His Own Burrow.
(Based on a total of 154 combats. Locality, dates, material as in Table II, p. 56.)

Key: AW Aggressive wanderer

BH Burrow holder

1

Type of
Combat

2

Relative
Size of
Crabs

3 4 5 6

Degree of Withdrawal of Noninstigator

Total

Major side
appendages
above ground

Claw alone left above ground

Manus and Chela tip

chela flat on vertical in air
surface

Crab wholly
underground

AW & BH

AW larger

3

3

1

5

12

AW smaller

6

2

0

2

10

BH & BH

Instigator

larger

5

0

3

7

15

Instigator

smaller

1

0

0

0

1

Total

15

5

4

14

38

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

61

number of other combats witnessed too incom-
pletely to be included, I have seen these half-
withdrawn burrow holders engage most effec-
tively in all classes of combat. These engage-
ments ranged from the most unritualized and
violent upset of a wanderer to the delicate
mutual frications of well-matched neighbors.

Column 4 in the table is almost equally un-
satisfactory, giving no indications of the pre-
valence of examples where the manus and chela
are laid flat on the surface. Almost always this
behavior is shown by a burrow holder that does
not engage at all in combat with an approaching
aggressive wanderer but simply withdraws be-
low ground before the wanderer can touch him,
leaving the manus and chela in sight. Hence the
behavior properly belongs under the general
heading of threat rather than of combat as de-
fined here, and so is beyond the scope of this
paper. The wanderer may or may not make
prying motions at the claw with his own chela
or ambulatories, or may even stamp on it before
passing on. Only twice have I seen this type of
withdrawal during an encounter where both
opponents may have been burrow holders; these
are not included in the table since they occurred
in combats where the phase of the instigator was
not certainly determined.

These flat-clawed partial withdrawals from
aggressive wanderers have been observed often
in other species. In all, they seem to occur prin-
cipally in crowded populations. Unfortunately
counts and distance measurements have not
even been begun.

Where the chela tip appears vertically above
ground (Column 5), it seems probable that the
projecting tip represents a different degree of
withdrawal mood in the retired crab.

Withdrawal wholly underground (Column 6)
almost always occurs when the crab’s opponent,
whether aggressive wanderer or instigating
neighbor, is larger than he.

2. Mero-suborbital frication—\n two of the
encounters between burrow holders in which the
noninstigator withdrew, his opponent reached
briefly down into the burrow with his small claw.
The merus of the major cheliped meanwhile
rubbed three to four times against the enlarged
outer crenelations on the lower margin of his
orbit. I have also twice seen the motion per-
formed by rapax when wanderers reached down
occupied burrows with the legs of the side op-
posite the large claw and, in the Indo-Pacific,
by U. vocans (Herbst) in threat behavior when
both opponents were on the surface.

3. Downpushes —In four examples observed
in detail a crab actively pushed his opponent

down the latter’s own burrow. Three of the
combats were between burrow holders, the actor
using the low-intensity manus push. In one of
these the instigator was the smaller crab; the
larger pushed him back to his own burrow,
alternating manus pushes with manus rubs. The
fourth downpush marked the end of a meeting
between an aggressive wanderer and a burrow
holder; the downpush was the only noteworthy
part of the brief combat, which included a
manus rub and an irregular interlace. In this
fight the downpush was delivered by a final grip.

4. Removal of burrow intruders— Three times
when a burrow holder returned to his hole after
a combat nearby a third crab was in possession.
On two of these occasions the returning crab dug
the intruder out. On the third, the burrow hold-
er, which was much the larger crab, thrust his
major cheliped down the hole and flipped the
smaller out; the action was too rapid for me to
note details. In no case was the incident fol-
lowed by any of the components of regular
combat.

5. Digging or prying out of opponent— Only
once was a digging or prying out motion ob-
served during a regular combat. Here an ag-
gressive wanderer had engaged a slightly larger
burrow holder. At the end of an encounter com-
posed of a manus rub, dactyl slide, and heel-
and-ridge without tap, the burrow holder slid
slowly down his hole. The wanderer disengaged,
tried briefly to dig the other out with his major
cheliped, and then departed.

6. Afterlunges— The term “afterlunge” is giv-
en here to an action frequently associated with
combat, although the same or very similar be-
havior sometimes occurs in purely threat situa-
tions. With the cheliped more or less flexed, the
crab lunges toward the opponent— either po-
tential or recent. When associated with combat,
an afterlunge always follows the encounter,
sometimes so closely that perhaps it should then
be considered a component. Such distinctions
between afterlunges, however, would be too
blurred to be practical.

The afterlunge was counted in 54, or almost
half, of the 112 combats of all kinds in which
its presence or absence was noted. It was almost
always performed by the larger crab, and, ex-
cept in boundary encounters, occurred at the
mouth of his own burrow. Usually it followed a
combat between an aggressive wanderer and a
larger burrow holder— that is, the most frequent
kind of encounter; 65% of all afterlunges fol-
lowed such an engagement. Afterlunges were
also common after boundary combats between
burrow holders; one occurred, always per-

62

Zoologica: New York Zoological Society

[52:6

formed by a single crab, in 9 out of the 12 fully
observed boundary encounters; in 7 of these
examples the larger crab was the actor.

Particularly in combats including an aggres-
sive wanderer, the burrow holder sometimes
lunged just as the two crabs were separating and
through this movement may even have brought
about the wanderer’s departure; these examples
added an element of virtual force to an other-
wise wholly ritualized and often mutual en-
counter. More often the burrow holder did not
lunge until the wanderer, or instigating neigh-
bor, was already at a distance. Once a small,
instigating, burrow holder atypically lunged to-
ward his former opponent after his return to
his own burrow. Finally, after one of two force-
ful burrow seizures, the former aggressive wan-
derer waved and then lunged toward the dispos-
sessed crab, which was circling about in the
low posture characteristic of nonaggressive and
nondisplaying Uca. This unusual example was
the only combat seen where an afterlunge fol-
lowed a forceful ending.

D. Categories

The preceding accounts of behavior compo-
nents have mentioned two general classes ot
combats, those between an aggressive wanderer
and a burrow holder and those between two
burrow holders. The characteristics of each will
now be considered as a basis for a review of
combat results. Mutual combat occurs in both
categories but will be examined more closely
under subheading 3 below.

1. Combats between an aggressive wanderer
and a burrow holder— These combats are notable
for their irregularity, for the prevalence of
forceful grips, and for the frequent withdrawal
of a vigorous burrow holder from an encounter.
In first determining the combat repertory of any
species of Uca, in fact, it is misleading to con-
centrate on combats involving aggressive wan-
derers. Such descriptions might easily be as in-
accurate as reports on the nest-building behav-
ior of birds drawn from observations of indi-
viduals that have not fully reached breeding
condition.

Unfortunately for the observer, the combat
activities of fiddler crabs in this phase are often
more numerous and certainly more easily fore-
seen than are encounters between burrow hold-
ers. It is temptingly convenient to select for at-
tention an active aggressive wanderer and watch
him on his progress through the population.
Such a course is usually conveniently marked by
the threats of burrow holders on his route. If
his aggressive phase is well established, a num-
ber of combats may be thus observed. Once the

repertory of a species is partially known, of
course, the combats instigated by these wan-
derers form a rich source of information.

The combats may be composed of one or
most of the components previously described.
Often, however, the high-intensity components
—dactyl slides, heel-and-ridges, and interlaces—
are imperfectly attained or the motions are
atypical; tapping, if any, is of wide amplitude.

For example, several aggressive wanderers,
after an uneventful manus rub, inserted both
pollex and dactyl through the gape of the bur-
row holder’s chela, instead of inserting only the
pollex. No regular combat development by the
wanderer seemed possible from this position.
One such contest ended with the burrow holder’s
opening his chela wide and freeing himself. In
another the crabs broke apart after awkward
shaking and shoving. After each of these en-
counters the wanderer departed.

Similarly, an apparently clumsy attempt is
sometimes made by the wanderer to attain a
dorsal slide position. This the opponent thwarts
by raising his own dactyl high. The maneuver
does not happen in fully ritualized fights be-
tween burrow holders, when the opponent usu-
ally appears wholly unresisting and — one is
tempted to say— cooperative. When a slide posi-
tion is obtained, the wanderer sometimes saws
back and forth transversely across a single spot
on the opponent’s upper dactyl rather than in
the normal longitudinal direction.

The behavior of a burrow holder approached
by an aggressive wanderer is often atypical of
his normally aggressive display phase; this is
true even when, as is usually the case, the wan-
derer is the smaller crab (Table I) . As described
under “Withdrawals” (p. 60), the burrow
holder sometimes goes completely underground
either when approached or later in the course
of the encounters; sometimes he leaves the
manus and chela, bent at the carpus, flat on the
ground as in maracoani (Crane, 1958, Text-fig.
2). When the wanderer has passed, the burrow
holder emerges promptly and resumes display.

All except one of the 15 forceful endings took
place at the close of a burrow holder’s fight with
an aggressive wanderer.

2. Combats between two burrow holders—
These encounters usually occur at the mouth of
the burrow of one of the participants. In con-
trast to the usual procedure in combats started
by an aggressive wanderer, the instigator in this
category is usually the larger crab. A minority
of encounters between burrow holders take
place on or close to the boundary between two
territories.

The same two neighboring crabs occasionally

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

63

proceed through highly ritualized combats a
number of times during a single low tide. Some-
times a burrow holder seems to be attracted to
his neighbor’s vicinity by a combat between the
neighbor and another crab, usually an aggressive
wanderer. Irregularities in the performance of
the components are rare, combats brief, and
forceful endings practically absent.

3. Mutual combats— About one-third of all
combats are strongly mutual in the sense that
each crab performs at least one of the ritualized
components. In the 154 combats analyzed in
Tables II and III, mutual elements were detected
in 38% of the combats between an aggressive
wanderer and a burrow holder, and in 31% of
those between two burrow holders. In 87% of
the mutual combats including an aggressive
wanderer, the latter was smaller than the burrow
holder he approached; this figure is higher than
the proportion (67%) of smaller wanderers in
the entire combat sample (Table I).

In a few additional combats, which are in-
cluded in Tables V and VI, an aggressive wan-
derer’s only activity appeared to be the initial
approach to the burrow holder. After that the
burrow holder seemed to be the sole actor and
the wanderer eventually departed. Similar cases
were observed twice in combats between burrow
holders. No doubt many more such examples
were seen than are cited in the tables; they are
not included since the instigator’s inactivity,
whether he was a wanderer or a neighbor, could
be only suspected because the angle of observa-
tion was unsatisfactory.

The heart of the problem of instigator inac-
tivity lies in the observation of low-intensity
components, the manus rubs. The actor or ac-
tors are especially difficult to detect, even in
motion picture close-ups, because manus rub-
bing is the only component in which mutual
action can be performed simultaneously by both
crabs; this of course is because the necessary
juxtaposition of parts— outer manus to outer
manus— is much less precise than in the subse-
quent components of high-intensity combat.
Even in the analyses of highly mutual combats
the counts for mutual manus rubs are doubtless
low; when uncertain about the mutuality of a
rub, I counted it as absent.

In ritualized components of high intensity-
dactyl slides, heel-and-ridges, and interlaces— the
actors either perform the slide in turn or ex-
change roles, the former actor holding still while
the second crab performs the next component.
Any necessary shift in position is made, includ-
ing those demanded by a different size of claw.
Sometimes the temporarily inactive crab is not
only quiescent during and after the shift but

even appears to take a cooperative part in it.
For example, after a slide a retiring actor
brought his dactyl below that of his opponent,
into the latter’s gape, and then stopped moving.
This left the other crab in the actor’s position
for the slide that promptly followed.

In mutual heteroclawed combats the heel-
and-ridge is performed by the individual having
his dactyl outside the manus, the interlace by the
crab with the dactyl against the inner surface,
as described on p. 63. When the second crab
becomes the actor, he needs only to engage the
claws farther, to the gape base, to attain fully
the interlace position.

Only two examples of this type of hetero-
clawed mutuality appear in Tables V and VI,
since these were the only such combats observed
from their low-intensity beginnings, and for
those tables it seemed best to include only fully
observed examples. A total of 1 1 heteroclawed
combats were in fact partially witnessed which
included this alternate performance of heel-and-
ridging with interlacing; one record was secured
on motion picture film; all except one of these
additional examples took place between an ag-
gressive wanderer and a burrow holder. Other
mutual components recorded in the tables were
also seen in greater numbers than indicated, but
again observations were partial.

Although some mutual encounters were
among the best examples of ritualized combat,
with no discernible trace of force, in others
aggressive elements were scarcely disguised. In
these combats the shift from one actor to an-
other was clearly accompanied by irregular
pushing or by abruptly jerking motions of a claw
in the midst of a component.

E. Duration

A total of 104 combats of all classes, each
observed from its beginning, were approxi-
mately timed, as described on p. 53. The great
majority turned out to be very short encounters,
most of them lasting between about 3 and 8
seconds. The duration of a few ranged from 1
to 3 minutes. No records were obtained of com-
plete combats lasting between 20 and 60 sec-
onds; even the addition of incompletely ob-
served examples, however, indicated that the
combats of intermediate length were rare— rarer
than the few of long duration. It seem likely
that when sufficient data are accumulated the
curve will have two peaks.

Relevant subdivisions are given in Table VII.
Of the 104 combats timed, 45 (43%) were of
low intensity; all of these were short, and less
than one-third included pushing with, therefore,
a recognizable element of force. The others

Uca rapax. Mutual Activity in Encounters Between Aggressive Wanderers and Burrow Holders.

(Locality and dates as in Table II. p. 56. Encounters were included in this table only if they were

observed from a manus rub beginning.)

Key: AW Aggressive wanderer

64

Zoologica: New York Zoological Society

[52:6

o

J3

&

O

Ui

\—

P

X

TD

C

cs

T3

C

a

£

OX)

OX)

a

>.

x

a

u

p

C

>s

= £
OQ <,

5tf

CQ

* ^

03 <

~ u.

§ 2
R o
o C

a. ^
a.

O

0Q

R

R

O

a.

5

o

U

c

o

O.

£

o

U

£ "3

~ Q £

2 «a c
o 2 Si

U o

»• V.

W O)

■a S-a
S £ £
o t-n a

c

►> &
■2 °
S|
2 i

o

a

T3

» +
X> "3 X!

S- 5

3 >> 3

S ° m

SQ2

£ ^

D —

I-H K"*"*
CO r>

3 ta

3 -O

ca

T3 S'

3 2

ca

X. JS

+

* i)

3-2

3 >>

S O

(D D
OX) OX)
T3 T3

TD *0
C C

c3 «s

<U <U

<D

OX)

■§ "■§

C/5 •— < C/5

3 y d d

S II + +1

Ti-

cs

+ x

(DO) *5

X TD T3 £

22 2 g g

S <-> <-> a) a>

<5 03 03

SQQ22

0)

jo

■§ 2

s +i + +i

Manus rub 1 Manus rub + dactyl slide

+ dactyl slide Heel-and-ridge with taps

+ heel-and-ridge Interlace

with taps
-(- interlace

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs ( Ocypodidae )

65

os

UJ

Q

a

*

o

D

03

O

£

H

Z

w

w

H

u

ca

c«

os

w

H

Z

a

o

u

z

w

c

o

ft

a

crt O

O «H
2 0
O

-C £
* 2
8 5

§ -o
X) c

c on

X

3

e

3

o

o

c

W

. -a

O- D
>

2-g

I H

P

u

<

1>

g?

-J c/3

3

D

H

D

s

*

<3

Ci.

<3

g 5*

* X

o3 *p*
73 ^
-a g
d o

cd

O

hJ

X £
2 g

E x

G >
* 'O

S B

CD

"O j5

2 >>

-C JO

< ex

>" >>

£■
03
T3

a

>> 3
> O

>- 03

s X

^ (J
§ <

03

s

c

o

CX

£

o

U

c

o

a

E

o

O

"5- a >
■2 ^
£ s ^
0-0
U O

a . oj

■o S«

s £ s

o ^ a

<o £

>-

- jo

ft 2

C/5 C/5

3 3
C C

Cd 03

2 2

jo-§

D U

a-

C/5 C

3 03

c g

03

s +

03 03

(N

70

73

C

o> ^
X 73 X
X d

*— < cr> *— i
C/5 — c/5

3 >>3

5 S S

SQS

>»
s. I-1

£ 03
O T3

fc c

3 3

^ 03

4-> <D

03 C
m (N

<d

72

3 2
^ >,

3 O
2 os

£ ^

s +

a> <u -
CD CD
73 73 '

73 73
C C

^ ^ "

Slis

1 *
£ 2
U u

2 3
X x

y> co

cS cd

(N <N

CD

72

*n

X 73
D C
>- 03

3 « D-

S ++I

X 73
3 X

U. C/5
C/5 —

33

S u

S Q

-o >

•2^0
3 — 03

Sou
c « C

S3 73 •-

S + +

X

ON

Manus rub 1 1 at y’s burrow Manus rub

+ dactyl slide Dactyl slide

+ heel-and-ridge Heel-and-ridge with taps

with taps Interlace

+ interlace

66

Zoologica: New York Zoological Society

[52:6

Table VII.

U. rapax. Divisions of 104 Combats of Known Duration.

Intensity

Duration Less Than 20 Secs.

Duration More T han 60 Secs.

Total

With force Fully ritualized

With force

Fully ritualized

Low

10 35

0

0

45

High

4 46

4

5

59

Total

14 81

4

5

104

were composed of manus rubbing alone and al-
ways appeared wholly ritualized.

The remaining 59 combats (57%) each in-
cluded one or more high intensity components,
usually preceded by manus rubs. Eight (less
than 14%) included forceful endings; about five-
sixths (83%), therefore, were wholly ritualized.

As will be evident in the following sections,
the figures of greatest interest are those in the
duration divisions within high-intensity combat.
Here 50 (85%) of the combats were of short
duration, and of these 4(8%) ended forcefully.
In contrast, of the 9 (15%) long combats, 4
also ended forcefully, but the percentage (44% )
was much higher.

Seven of the 9 long combats had mutual com-
ponents; 3 of these 7 had forceful endings, in-
cluding the eviction of a burrow holder by an
aggressive wanderer. Both of the evictions in
the sample came at the end of a long fight. Six
of the long combats involved an aggressive wan-
derer, a proportion roughly in agreement with
wanderers’ occurrence in the combat sample.
Finally, 8 (88%) of the 9 long fights were
heteroclawed— a very different proportion from
the 50% characteristic of the sample.

In summary, in a sample of 104 combats all
but 9 lasted less than 20 seconds. Each of these
9 lasted more than a minute and contained high-
intensity components. Proportionately more of
these long combats occurred between right- and
left-clawed crabs, more had mutual components,
and more ended forcefully than did short
combats.

F. Postcombat Behavior

After most encounters between rapax males,
the opponents promptly resumed their precom-
bat activity. Aggressive wanderers passed on
through the population, instigating new combats
and engaging in other activities, as already de-
scribed (p. 53). Burrow holders, returning with
equal completeness to all their former activities,
first resumed waving.

Almost one quarter of all encounters suffi-

ciently observed, however, were followed by
detectable changes in behavior. Of all observed
combats, both opponents were watched long
enough in 148 examples to form a suitable basis
for an examination of such changes and of com-
bat composition when subsequent changes did
not occur. These alterations in behavior were of
two kinds: either the aggressiveness of an ag-
gressive wanderer was reduced or there was an
appreciable delay in the resumption of waving
by a burrow holder. Reduction of aggressiveness
in a wanderer and delayed waving by the burrow
holder never followed the same combat, nor was
waving ever delayed by both opposing burrow
holders.

Table VIII breaks down the 148 combats
where subsequent behavior was observed into a
number of potentially relevant subdivisions. The
first column, headed “Result,” divides the group
into those with behavior unchanged, waving
delayed, or aggression reduced. In the second
column, “Combat Class,” the opponents’ phases
and relative size are indicated, as in previous
tables, as well as, where necessary, the instigator
and the site of the combat. Under “General
Combat Composition” selected characteristics
are isolated; these show the relative prevalence
of low- and high-intensity combats, forceful
components, tapping, and mutual components.
It seemed that one or more of these aspects of
combat might be correlated with behavior
changes or the lack of them. However, no clear-
cut correlation emerges. For example, neither
tapping nor mutual components preclude either
a delay in resumption of waving or reduced ag-
gression; similarly, forceful endings are not
necessarily followed by subsequent behavior
changes. Nevertheless, certain trends are indi-
cated. The information given by the table, in
addition to the more limited data on combat
duration (pp. 63, 66; Table VII), form the
bases for the remainder of this section.

Class 1. Combats not followed by a detectable
change in behavior— The 112 combats in this
class comprise three-fourths of all in the sample;

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

67

x o
w J
CQ S.

H

a

a.

a

o

IS

3

.s

E

H

<

M

S

o

U

a

s &

£ _r
o ^

i-j <u

° ■§
^ H
h .5
g c§

£ CO

o a>
-ti

Oh

00 cd
£ £

^ c

nj c
c O
o a
a a
a o
° c

& -S

u

(L) — <

oo cd

a ^
>>

T. "O
cd

« £
CJ <L>
CO
CO

co O

w a

& £

2 % § 'f

fe fe oj

2 2 p 2

n\ ’’7s

§ o7

< <

3 3

m oa

73 *

3 3

2 co

< 1 S s s

>>

a>

-o a

<g§

O

a

g

o

U

o

-c

£

<0

O

03

&

o

£

co C

3 <D

O' 3

o g

z &

® VO <-1 VO cr\

IA1 (N T-.

s

fN

m co

VO

m

(N

£2
»-h <N

c3U

£ §

nK-C I! -C |T XJ

£ > I ~ X

< 5 03 03 CO

= O C ™
O- <D £
<u > uh cd
co cd cd X2
•& jg a o
3 <l> a c

CO -O cd 3

co uii 'a

<D co 2

u- a> c

^ X JX II £ ||

5 c* ^ co *3 £ =« g

< cd PC CQ CQ

00

w c o
a-$ t

£ % 5

=3 > X)

CO ^

O

C 1/5
3 «
, 7£ g

t-H QJ H— • _H

<L) 7>. G

T1 m <*> ‘P

_2 ^ co c
a> c

=3 «

$ £
<, cd

^ >, o '<l> jy
O 43 J2T3 w

68

Zoologica: New York Zoological Society

[52:6

61% were of high intensity. Force was detected
in about 12% of all combats in the class; there-
fore, 88% were fully ritualized; forceful ends
occurred in about 5%. Tapping was present in
more than 21% of the total, never in combats
where forceful ends occurred; where only high-
intensity combats were counted (since tapping
normally occurs only then), the percentage with
tapping was 35%. Mutual components were
noted in about 30%. Short combats, lasting less
than 20 seconds, formed 95% of the total in a
duration sample of 86 in this class.

Class 2. Combats followed by a delay in re-
sumption of waving— Fourteen of the 26 ex-
amples in this class were between an aggressive
wanderer and a burrow holder; 9 were between
two burrow holders where the larger was the
instigator; and 3, with the instigator question-
able, occurred near a territorial boundary be-
tween burrow holders. These 26 combats form
almost 18% of all combats in the sample. Wav-
ing was usually delayed by a crab at his own
burrow following combat with a larger indi-
vidual, either wanderer or neighbor. When a
smaller crab was the instigator at the burrow of
a larger crab, no postcombat waving delay oc-
curred. Of the 26 delays, 24 lasted less than two
minutes; the longer exceptions were by dispos-
sessed crabs and are described below. The short
delays included some subsequent behavior as-
sociated with combat, such as a number of the
total withdrawals.

Forceful components took place in a total of
15% of the class as follows: pushing occurred
in two low-intensity fights between burrow
holders; two high-intensity combats, both be-
tween a large aggressive wanderer and a smaller
burrow holder, resulted in eviction of the bur-
row holder. Because of the theoretical interest of
these two fights certain combat and postcombat
details are pertinent. Both included taps; one
showed mutual components; neither ended in a
physical upset; the duration of each was more
than one minute. After the first dispossession,
the wanderer descended the burrow, then
emerged and began waving. Meanwhile, the
former occupant assumed the low posture and
moved off giving no response to the threats of
neighbors. Eventually he retired down an empty
burrow some six feet away, where he plugged
the mouth and remained at least until the next
low tide. In the second example the wanderer
descended the burrow after the occupant left,
emerged within a few seconds, and then aban-
doned it to continue his precombat behavior as
a continuing aggressive wanderer. Meanwhile,
the evicted burrow holder had been circling
around in the low posture; as soon as the wan-

derer left, this crab resumed possession, de-
scending briefly, emerging and waving prompt-
ly. The entire delay, in addition to the combat’s
1.5 minutes, lasted 2.4 minutes, in contrast to
the indefinite delay in waving following the
first eviction.

Throughout Class 2 tapping was absent in all
fights with forceful ends except in the two just
described, which ended in eviction. Tapping
was present in 25% of the remainder, or almost
50% of those of high intensity without force.
Mutual components were absent in low-intensity
combats but present in 27% of those of high
intensity, including one of the evictions, and in
15% of the entire class. Short combats formed
67% of the total in a duration sample of 15.

Class 3. Combats followed by a reduction in
the aggressiveness of a wanderer— Of the 90
combats between an aggressive wanderer and a
burrow holder, 10 (11%) were followed by
reduced aggressiveness. These 10 combats were
distinguished as follows: In 40% the wanderer
was the smaller crab; 60% were of high inten-
sity and included all combats in which force
was detected; these forceful endings occurred in
83% of high-intensity combats or 50% of the
total in the class. Taps were included in 40%
of the total, most of them in combats with force-
ful ends. Mutual elements appeared in 20% of
the combats, all high-intensity, with and without
forceful ends. Short combats formed 83% of
the total in a duration sample of 6.

Comparison of combat characteristics in rela-
tion to subsequent activities— When the above
classes of combat are compared, five points
emerge that seem noteworthy in spite of the
small samples. First, in combats followed by the
reduced aggressiveness of a wanderer, forceful
endings were more numerous than in combats
either not followed by behavior changes or with
a subsequent delay in resumption of waving.
Second, long combats were most numerous in
the class followed by delayed waving, less so
among those resulting in reduced aggression,
and rare among encounters with no detectable
results. Third, mutual components were rela-
tively fewer in combats followed by changes in
behavior. Fourth, tapping was usually absent
from encounters with forceful endings; this
absence is probably correlated with the fre-
quently prompt cessation of combat after tap-
ping. When tapping did occur in the course of
a fight ending forcefully, subsequent behavior
was changed. Finally, after combat any changes
in behavior were usually shown by the smaller
crab.

Summation— because this study was made en-
tirely in the field among unmarked crabs, few

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

69

hints of summation were observed. I have in
fact obtained reliable notes on the subject in
only seven pairs of combats. Two of these in-
volved examples where the actor’s behavior
changed following the second encounter; in each
the encounter was less intensive than the first
and the change consisted in reduced aggressive-
ness of a wanderer.

When marked crabs are adequately observed,
it seems certain they will show that summation
plays a part in the effects of combat.

V. Combat in Species of
Uca Other Than U. rapax

Combat has now been filmed in fifteen species
of Uca in addition to U . rapax. These species are
scattered widely throughout the genus. The re-
sults are predictably inadequate, because no spe-
cial effort has yet been made to record the vari-
ous patterns in species other than rapax. As

stated earlier it is only recently that the complex
nature of fighting and its morphological special-
izations have become apparent. Hence, the lack
of a filmed record of any particular component
by no means precludes its existence.

Data on combat in Indo-Pacific and eastern
Pacific species recorded only in field notes have
been omitted from this paper. The observations
were made before the importance of ritualized
fighting was realized and at a period when my
attention was concentrated on display. During
those years all combats appeared too similar to
deserve full attention, in an essentially com-
parative study of species, since time in the field
was always limited.

Several statements can, however, now be made
with relative assurance in comparing combat in
other species with that in rapax. First, the Indo-
Pacific species listed in Table IX, with the ex-
ception of three species in the old, informally

Table IX.

Basic Components of Combat in Uca So Far Recorded on Motion Picture Film.

(Note: Minor elements are omitted because of insufficient data. Species are arranged in systematic
order, in accordance with a revision in preparation; the closest relations of U. lactea are
American species, as indicated.)

Manus

Manus Dactyl

Heel-

and-

Heel-

and-

Inter-

Region Species (Uca)

Push

Rub Slide

Hollow

Ridge

Tapping

lace

Indo-Pacific

dussumieri ( Milne-Edwards)

X

X X

coarctata (Milne-Edwards)

X

X

urvillei (Milne-Edwards)

X X

X

Dactyl

only

bellator (Petiver)

X

tetragonon (Herbst)

X

No

vocans (Linnaeus)

X

X

X

No

chlorophthalmus (Milne-Edwards)

X

Dactyl

only

inversa (Hoffman)

X

Dactyl

only

W. Africa

tangeri (Eydoux)

X

X

X

America

maracoani (Latreille)

X

X

insignis (Milne-Edwards)

X

rapax (Smith)

X

X X

X

Dactyl
& pollex

X

cumulanta Crane

X

X X

X

Dactyl
& pollex

X

inaequalis Rathbun

X

Indo-Pacific

lactea (deHaan)

X

X X

X

Dactyl
& pollex

America

deichmanni Rathbun

X

X

70

Zoologica: New York Zoological Society

[52:6

named “broad-fronted group,” show more ten-
dency to forceful fighting with less ritualization.
The three exceptions are chlorophthalmus, in-
versa, and lactea.

Second, the highly ritualized heel-and-ridges
of the New World crabs are represented in some
Indo-Pacific species by a lifting and gripping
motion differing from that in rapax as follows:
The pollex with its large distal or subdistal teeth
either grip in or do not usually pass beyond ap-
propriate hollows near the base of the pollex.
This position is in contrast to the rubbing and
tapping against the oblique ridge characteristic
of rapax , cumulanta, chlorophthalmus, and in-
versa. The dactyl, however, holds in all a posi-
tion similar to that in rapax, against the outer
part of the manus. Only in broad-fronted lactea
has the pollex been seen to make a series of taps
in the hollow at the inner base of the pollex.

Several species typically shift from low to
high intensity in a manner different from that of
rapax. In tangeri, maracoani , insignis, and in-
aequalis the chela tips are turned downward, the
mano-carpal joint raised, and the chelae en-
gaged in this position, after which they swing
once more to the horizontal. In all the other
species in which high-intensity encounters were
filmed, the chelae engaged in the horizontal po-
sition, usually at the end of a manus rub.

The variation in dactyl slides is great, as in
rapax. Whereas the detailed study of rapax
makes it possible to state the normal position of
the dactyl during a slide, in other species the
paucity of material and lack of information on
individual social situations make any further
statements unproductive.

VI. Discussion

Two challenging aspects of fiddler combat are
the obscurity of its functions and the complexity
of its ritualizations. As long as the functions of
fighting remain uninvestigated its ritualizations
can be only superficially understood. Because
of the new data presented, however, it seems
timely to comment on possible functions for
combat and to suggest reasons for its ritualiza-
tion. The section continues with preliminary re-
marks on the possible origins of the components
and on comparisons with combat in other ani-
mals. It closes with suggestions for further re-
search.

A. Functions of Combat in Uca

A general discussion of territoriality in fiddler
crabs lies beyond the scope of this paper. Here
the subject will be mentioned only to the extent
necessary to give a frame of reference for the
consideration of combat.

In this genus the territoriality of breeding

males appears to be not at all concerned with
the optimum distribution of the population in
respect to the food supply. Only males in the
display (waving) phase, during which mating
takes place, behave aggressively toward individ-
uals in the vicinity of their occupied burrows,
and then only toward males in the aggressive
wandering phase or toward trespassing neighbor
males. Aggressive wanderers, in contrast, neither
wave, mate, nor remain near a particular bur-
row. The territories of the displaying crabs are
not scattered equally through the population,
nor are they usually to be found in food-rich
locations.

Burrows of displaying crabs are often, in-
stead, concentrated in slightly higher, drier parts
of the intertidal zone, where food is less abun-
dant. Here males in other phases are few or ab-
sent. Moreover, crabs in the display phase feed
little in comparison with others, and when they
do so, they often move temporarily to the richer,
lower levels. Females in breeding condition in
some species occupy burrows close to those of
waving males; in others, receptive females move
actively through the displaying portion of the
population; after mating, while the eggs are de-
veloping, females usually inhabit burrows closer
to the low-tide levels than those of any section
of the population except the very young.

It appears rather that territoriality in Uca is
concerned directly and indirectly with reproduc-
tion. An advantage of territories on higher
ground may be that the displaying crabs are
there more conspicuous. This could obviously be
of importance not only in attracting receptive
females but in the mutual visual stimulation of
waving males. Again, the drier substrate may
transmit more effectively the vibration signals of
courtship. Finally, and probably importantly,
territories on higher ground are uncovered by
the tide for longer periods daily, thus giving
more time for courtship; at least in the tropics
receptive females are never abundant in any
population of waving males and courting time
is short at best. It must be emphasized, however,
that displaying males in a given species by no
means always gather into a lek-like formation;
very often they are scattered, although in clear
sight of one another, throughout the greater
part of a population. The relative breeding suc-
cess of individual males with burrows in differ-
ent locations remains wholly to be investigated.

Possible explanations of fighting remain far
from clear. It is unsurprising that each crab com-
ing into territorial and display phases acquires
a burrow in a spot appropriate for display and
mating. But, for two reasons, there appears to
be no need to obtain such a burrow through

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs ( Ocypodidae )

71

combat. First, all fiddlers of both sexes and of
all ages except the very young can dig new bur-
rows quickly and efficiently. Yet they do so only
in emergencies brought on by predators or by an
inrushing tide. In contrast, when a large crab
evicts a small one, he enlarges the burrow suit-
ably, with neither hesitation nor hurry. Further-
more, few areas are so crowded that there does
not appear to be ample space, empty and unde-
fended, among displaying males where new bur-
rows could be dug.

Second, and probably more important, in all
displaying populations that I have adequately
observed, empty burrows are plentiful among
those of displaying individuals. They have been
abandoned by crabs not at the time in a burrow-
holding phase; as the tide recedes, these crabs
simply dig their way to the surface and go else-
where. Yet an aggressive wanderer, unless his
aggressiveness has been reduced by a fight with
a burrow holder, either pays no attention to these
empty burrows, or pokes into them with chel-
iped or ambulatories, briefly and superficially;
then he moves on. These empty burrows always
appear to be considerably more numerous than
are the wanderers, so that no difficulty in en-
countering such a hole seems to be involved.

Although occasionally an aggressive wanderer
stops slightly longer at a burrow occupied by a
nondisplaying crab, any attempt to dig out such
a crab is rare and little effort is expended. I have
never knowingly seen a combat between a crab
that is not in display phase and a wanderer, or
between two wanderers.

In fiddler crabs no harems are maintained
and single females seem never to be direct causes
of intermale combat. Occasionally a male even
abandons an advanced courtship, attracted by a
combat between two other males.

The immediate goal of an instigator does not
in fact seem to be the taking over of a suitable
burrow as a center for display or a direct com-
petition for females. Rather, the apparent aim
is a combat with a displaying male.

The combat itself, as has been described in
detail for rapax, is usually fully ritualized, with
no apparent component of force; in most cases
it results in no detectable change in the subse-
quent behavior of either crab. One or the other
withdraws his claw from contact with that of his
opponent; the aggressive wanderer resumes his
progress through the population, threatening and
entering into new combats; the burrow holder
promptly resumes display, its intensity undi-
minished. In one such rapax combat in nine,
however, a wanderer’s aggressiveness was re-
duced; in one in 45, the burrow holder was dis-
possessed and the wanderer took over. Less

intensive observation of other advanced species
have yielded corroborative observations: The
wanderer’s behavior is similar and only rarely is
there a detectable result. In those instances
where a wanderer actually takes over a burrow
he sometimes assumes the display phase at once;
more often, he does not wave, but shortly aban-
dons the burrow and moves on, his aggressive
ness maintained and his territorial drive still in
abeyance.

With these figures in mind it seems likely that
combat may sometimes either advance or retard
the assumption of territorial and waving phases
by the wanderer. Summation, as suggested on
p. 68, may well play a part here that the field
techniques in use could only suggest. Combat,
then, may serve as a mechanism for ensuring
that suitable burrows for display are not taken
over by males in subbreeding condition; never-
theless, the availability of empty burrows, noted
above, forms an obvious argument against this
view.

The function of combat will now be examined
from the point-of-view of the burrow holder. If
this displaying crab is not vigorous enough or
sufficiently motivated to fend off an aggressive
wanderer, he may be in an inadequate condition
for breeding and should not, from the point of
view of selection, be left in a position to attract
receptive females. Yet many vigorous burrow
holders, in other species as in rapax, withdraw
partly or wholly from an incipient combat, even
in the frequent instances where the approaching
wanderer is the smaller crab; then the burrow
holders resume waving and courting promptly
and strongly when the wanderer has departed.
The role of this withdrawal behavior in the pat-
tern of combat remains puzzling.

After combat, however, one rapax in six de-
lays waving, while one in 45 loses his burrow,
with a consequent postponement of resumed
display. These relative numbers agree well with
impressions received in numerous more casual
observations of combat in other species.

In examining the possible selective values of
fighting and its ritualization a distinction should
be kept in mind between these two visible results
—namely reduced aggressiveness and delayed
waving. Since all burrow holders are in the dis-
play phase and, as part of that phase, in a threat-
ening and fighting mood toward both aggressive
wanderers and trespassing neighbors, it seems
that a postcombat reduction in aggressiveness
by a wanderer normally would result only if he
were not ready for territorial-display-mating be-
havior. In that case the “loss” of a combat would
be a selective advantage. On the other hand, a
reduction of display time for a burrow holder

72

Zoologica: New York Zoological Society

[52:6

through prolonged combat would be a disad-
vantage.

A reasonable suggestion, therefore, appears to
be that the ultimate value of combat, regardless
of the role of ritualization, lies in preventing
suboptimal males from wasting the breeding
time of the population by attracting receptive
females. This explanation does not, however,
account for all the facts. It takes no account of
combats, almost all fully ritualized and resulting
very rarely in waving delays, between vigorous
burrow holders. Again, the function of down-
pushes remains unexplained. Here one burrow
holder, far from endeavoring to take over the
burrow of his neighbor or at least to dig the oc-
cupant out and engage him in combat, simply
thrusts him forcefully underground before re-
turning to his own burrow and resuming display.

When viewed as a whole it seems that the
function of combat may lie primarily in stimu-
lating and synchronizing mating behavior. As
in so many other groups of animals where such
an effect is suspected, proof awaits work in
endocrinology and neurophysiology.

Similarly in need of the attention of physio-
logists are two strong impressions that recur
during fieldwork on Uca. One is that combat
may serve to release tension in the actively court-
ing section of the population. The other impres-
sion, particularly compelling when one is watch-
ing ritualized mutual encounters, is that combat
appears often to be in progress for its own sake.
The attention of a third crab is sometimes drawn
to a nearby combat; he may then either interrupt
or engage one of the participants after the end
of the first encounter. Even more suggestive are
the sequences of high ritualization discussed be-
low.

B. The Question of Adaptive Values of
Combat Ritualization

As shown in previous sections, a large major-
ity of combats in Uca rapax show no detectable
element of force and hence may be termed fully
ritualized. More casual observations on other
species indicate that ritualization is similarly
prevalent throughout the genus. In searching for
the selective advantages of ritualization the im-
mediate effects of individual combats have
proved unilluminating. As is well known, even
the most violent fights in Uca practically never
result in physical damage; no injury at all was
ever seen in rapax. It seems, therefore, that a
protective function, which has been considered
obvious in the ritualized encounters of many
well-armed animals, is not now of importance in
fiddler crabs.

Again, the data in this paper on rapax give

no evidence that ritualized encounters are any
more likely than the uncommon forceful fights
either to promote or to prevent behavior changes
in an opponent. This is true in general both of
reductions in the aggressiveness of a wanderer
and of delays in resumption of waving by a
burrow holder.

The only apparent advantage of ritualization
in rapax seems, rather, to lie in the shortening
of combats. The counts so far made indicate
clearly that ritualized encounters are not only
far more numerous than those including com-
ponents of force but also that they are shorter,
most lasting less than 10 seconds. In contrast,
forceful fights continuing more than one minute
are usual. This is true whether or not a forceful
combat results in subsequent visible behavior
changes for either crab. While this difference in
duration appears to have no obvious importance
for an aggressive wanderer, the shortening of
combats through ritualization may well be a
selective advantage through its effects on bur-
row holders.

This suggestion is based on both the ecology
and the mating behavior of Uca. Since they
court only during low tide, and are usually
further restricted by other requirements, both
meteorological and physiological, their periods
for courtship and mating are limited. Combat
and courtship cannot proceed simultaneously
and, in Uca, the combats of males seem to hold
no attraction whatever for females. Therefore it
seems clear that, by shortening combats, ritual-
ization ensures that courtship opportunities are
minimally reduced.

It may be that an important factor in waving
display lies in its stimulating effect on other
males, or in the synchronizing of breeding ac-
tivities. Here, too, a shortening of each combat
would advantageously shorten the time during
which one or two wavers did not contribute to
the communal effect.

One characteristic of ritualized combat be-
comes increasingly apparent with continued ob-
servation. This consists in the leisurely, formal-
ized, and wholly unforceful cooperation some-
times apparent between the two opponents. A
highly ritualized encounter in rapax may run
about as follows. An instigator, whether wan-
derer or neighbor, approaches a burrow holder.
A rub by one or both crabs, outer manus against
outer manus, usually follows. Next, the instiga-
tor sometimes holds perfectly still while his op-
ponent slowly eases his chela into the actor’s
slide position; the two crabs may then reverse
the role, the shift being accomplished slowly,
without fumbling, and with the apparent coop-
eration of the crabs. In a few moments they may

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs (Ocypodidae)

73

progress to a similar alternation of heel-and-
ridging or, in heteroclawed encounters, to an
alternation of heel-and-ridging with interlaces.
In other examples a single opponent may be the
actor throughout, the second crab holding him-
self quietly. When the actor breaks off, both
crabs move apart and resume their pre-encoun-
ter activities.

Observation of these encounters gives a
strong impression that they provide one or both
crabs with satisfactions that are not concerned
in direct goals, such as taking over a burrow or
evicting a trespasser; the activity itself seems to
serve as the goal. We know nothing at all yet
about the means of conferring satisfaction—
whether through the performance of the mo-
tions, or through the reception of associated
sensory stimuli.

If ritualization does indeed operate selectively
through shortening combats and thus providing
more time for courtship, then an obvious pres-
sure would be toward even shorter ritualized
encounters. Ultimately the action might be re-
duced to a token touch of mani or single rubs
of ridges by briefly overlapping chelae.

This trend is not apparent. According to our
present knowledge, the socially advanced species
have the largest repertoire of combat actions
and the most extensive structural specializations
for high-intensity encounters. If ritualization
shortens combats, then further elaboration could
nullify the effect. Occasional prolonged encoun-
ters in rapax, fully ritualized and elaborately
mutual, suggest that this process may prove to
be a factor in current evolution.

C. Derivation of Components

Too little comparative work has yet been
done to discuss the comparative ethology of
combat in Uca. Nevertheless it seems worth-
while to suggest two basic derivations that may
be kept usefully in mind during future studies.

It seems likely that high- and low-intensity
components have been derived from different
sources. These are, respectively, from forceful
fighting itself and from the fronto-lateral threat
gesture common to most crabs. If the direction
of evolution has been, as seems apparent, the
reduction of forceful combat through ritualiza-
tion, one logical point for the application of a
deterrent would be immediately before the grip.
The high-intensity components all appear to
have originated directly from unritualized force-
ful grips. All take place with the two claws
partly or wholly in a position for grasping each
other; when, rarely, a ritualized encounter pro-
ceeds to a grip, little or no change in basic claw
position is made. Unsurprisingly, specializations

appear to have been added one in front of an-
other, the interlace of rapax being perhaps the
closest now known to the original fighting grip.
The ultimate ritualization, then, would be logic-
ally the low-intensity manus rub.

However, the manus rub seems most credibly
to have evolved from the warding-off lateral
threat gesture, which is almost universally
present in crabs. But threat gestures in many
animals are themselves certainly to be under-
stood as ritualizations of fighting, where a
weapon, impressive size, or other potential ad-
vantage is exhibited, usually with exaggeration
or embellishment. By these criteria, the threat
gestures of crabs, including Uca , qualify as
ritualized fighting. It seems that in Uca , how-
ever, manus rubbing is a further ritualization
derived from a basic threat posture.

D. Comparisons with Ritualized Combat in
Other Animals

The ritualizations of combat through threat
displays are of course endlessly varied among
animals, perhaps culminating in the displays of
sound, color, and movement in many territorial
birds. In fiddler crabs parallels are close.

In encounters where two individuals come ac-
tually into contact, however, comparisons are
far less satisfactory. Checked point by point with
Uca combat, the harmless poking and heaving
of fighting beetles or the ritual butting and kick-
ing of ungulates are simple, coarse, and slow
while the associated structures seem relatively
unspecialized. A search of the literature has not
yet yielded, even among deer, an apt compari-
son. It seems that no buck rubs a certain antler
prong gently against a different prong, suitably
shaped, on his opponent’s armature, then holds
still while the other gets into position and rubs
in turn and, finally, passes on to a further step
in the sequence. Schaller’s ( 1967) detailed com-
parative study of combat in eight Indian ungu-
lates indicates no such refinements. There seems
to be in fact not even a rough analogy to perhaps
the highest Uca specialization of all, where
adaptations exist even to the different require-
ments of opponents with the ritualized weapon
on the same or opposite sides of their bodies. In
other decapod crustaceans it seems certain that
many spines and ridges, that have been as in
Uca of purely taxonomic interest, will prove to
be as functional as in fiddlers, and perhaps in
similar fashion.

E. Areas for Further Research

The following aspects of Uca combat partic-
ularly need attention before the wider implica-
tions of the subject can be investigated.

74

Zoologica: New York Zoological Society

[52:6

1 . Sensory aspects— It is not yet known wheth-
er the rubs and taps of combat are perceived by
the crabs and, if so, which senses are involved.
Acoustic elements will probably prove to be part
of the pattern of fiddler combat as in their other
types of social behavior; I have heard rasps and
clicks during examples of each of the compo-
nents. It seems likely, however, that tactile sen-
sations are also important. For that reason the
term “stridulation” has not been used in this
contribution.

2. N europhysiological and hormonal aspects
of phase and of behavioral changes following
combat— These basic problems are more than
ready for investigation. They should be under-
taken in a laboratory adjacent to wild popula-
tions, or at least in one equipped with a large
crabbery. This specialized terrarium should in-
clude provisions for artificial tides; for controlled
salinity, temperature, and light; for a substrate
natural for the species, changed at suitable in-
tervals; and for frequent restocking. In short,
adequate results could not be obtained from
fiddlers in finger bowls.

3. Duration of individual combats— The data
in the present paper on the timing of combats
are particularly inadequate in view of the tenta-
tive conclusion that the primary function of rit-
ualization in rapax is the shortening of combat.
In future work on all species, the behavior asso-
ciated with combat should certainly also be
timed.

4. Combat at night and underground— The
occurrence and form of encounters at night and
in burrows need examination. The deep pits at
the pollex base that in some Indo-Pacific forms
apparently serve as “brakes,” and are often
marked by puncture wounds, are represented in
rapax and its relations respectively by slight de-
pressions and faint scratches. No gripping has
ever been seen in rapax in this position; perhaps
such a grip is used only when an intruder enters
an occupied burrow. Observation burrows can
be devised in the laboratory.

5. Mathematical analysis— The distinctness of
components in most combats at first glance
makes a mathematical approach to combat
analysis seem feasible and attractive. But suc-
cess in such undertakings is endangered by a
number of pitfalls. Quantitative analysis could
be misleading unless based on accurate, detailed
knowledge of the full combat repertory and its
related behavior in the selected species, and on
influences of seasonal and other rhythms. The
principal threats to the significance of conclu-
sions include: the difficulty of observing the ex-
act beginnings of combats; the partial distinctions
between homoclawed and heteroclawed encoun-

ters; the existence of mutuality and the difficul-
ties in distinguishing it; the hard-to-gauge degree
of basic aggressiveness in contrast to the coop-
eration sometimes apparent; the need for definite
establishment of the phases of the opponents;
the frequent irregularities in combats involving
aggressive wanderers; and the difficulty of assess-
ing the effects of summation.

6. Marked crabs in a fenced wild population—
Many of the difficulties listed could be reduced
by fencing a suitable intertidal area, large enough
to give wanderers space to move about normally.

7. Comparative ethology — Contrary to my
earlier conclusions, it is now clear that interspe-
cific distinctions in combat behavior are numer-
ous enough to make their comparative study
highly rewarding. While the differences are not
nearly as striking as in display there is no ques-
tion but that they both illumine evolutionary
trends in species groups and show most interest-
ing steps in the development of combat ritualiza-
tion.

8. Instigation and frequency of combat— The
conditions leading to combat between individ-
uals and controlling its frequency in a popula-
tion deserve the closest attention. Most of the
following factors, all virtually uninvestigated,
will probably prove pertinent: precombat threat
behavior, length of time since phase of each
opponent began, age, individual variation, semi-
lunar and other rhythms, population density, and
infra-specific differences.

VII. Summary

This contribution concerns a field study of
combat between male fiddler crabs. Except for
preliminary comparisons with a few other spe-
cies, the report is based on several hundred com-
bats observed in a population of Uca rapax
(Smith) in Trinidad. Descriptions of the various
aspects of the subject were derived from between
104 and 180 examples watched or photographed
in sufficient detail for each purpose.

The combat pattern in rapax shows seven dis-
tinct components, five of them highly ritualized
and each of the five associated with particular
morphological specializations. The frequency of
three of the components is related to the juxta-
position of the large claws: when one crab has
the claw of the right side enlarged and his op-
ponent that of the left, the components tend to
be different from those observed when both
crabs have the large claw on the same side. Most
combats last only a few seconds and infrequently
include the unritualized components of force;
when they do so, the fights are often longer and
may last several minutes. Two or more ritualized
elements, usually performed in a fixed sequence,

1967]

Crane: Combat and Its Ritualization in Fiddler Crabs ( Ocypodidae )

75

compose most encounters. In each component
one crab rubs and sometimes taps with teeth on
his own claw a tuberculated ridge or other par-
ticular structure on the claw of his opponent.
Many combats are mutual in the sense that each
crab performs at least one of the components;
sometimes the same component is performed by
each crab in turn. The sensory aspects of ritual-
ized combat have not yet been investigated.

Combats divide sharply into two categories.
The first takes place between an aggressive wan-
derer and a burrow holder in the display phase;
its components are often irregular; practically
all combats ending in forceful grips and upsets
belong in this category. In the second, the op-
ponents are two neighboring burrow holders;
irregularities and force are both uncommon.
The relative size of the opponents is a factor in
the development of combats and in various asso-
ciated activities.

The great majority of combats, whether
wholly ritualized or not, result in no detectable
change in the behavior of either crab. Each
opponent continues either to wander aggres-
sively or to wave beside his own burrow. Only
twice, both after long fights, was a burrow holder
dispossessed by an aggressive wanderer with an
associated indefinite delay in further waving
and, therefore, courting activity. One combat in
nine results in a reduction of aggressiveness by
the wanderer; one in six is followed by a very
short delay in waving resumption by a burrow
holder. Combat does not appear to be concerned,
even indirectly, with the food supply; it neither
involves nor attracts females; it appears to be
unnecessary for the securing of suitable burrows
as display centers and mating sites.

It is suggested, therefore, that the functions
of combat must be sought in indirect physiolog-
ical needs connected with reproduction, such as
stimulation or the release of tension. It is further
suggested that the principal selective advantage
of ritualization lies not in the prevention of
physical injury or of loss of display phase but
rather in the shortening of combats. In this way
less of the brief time suitable for courtship is
lost in fighting.

VIII. References

Altevogt, R.

1957a. Untersuchungen zur Biologie, Okologie
und Physiologie Indischer Winkerkrabben.
Zeit. Morph, u. Okol. Tiere 46: 1-110.

1957b. Beitrage zur Biologie und Ethologie von
Dotilla blanfordi Alcock und Dotilla myc-
tiroides (Milne-Edwards) (Crustacea De-
capoda). Zeit. Morph, u. Okol. Tiere 46:
369-388.

1962. Akustische Epiphanomene im Sozialver-
halten von Uca tangeri in Sudspanien. Ver-
hand. Deutsch. Zool. Ges. 22: 309-315.

1964. Ein antiphoner Klopfkode und eine neue
Winkfunktion bei Uca tangeri. Sonder.
Zeit. Naturwiss. 24: 644-645.

Baerends, G. P.

1950. Specializations in organs and movements
with a releasing function. Symp. Soc. Exp.
Biol. 4: 337-361.

Beer, C. G.

1959. Notes on the behaviour of two estuarine
crab species. Trans. Roy. Soc. N. Z. 36:
197-203.

Blest, A. D.

1961. The concept of ritualisation. In W. H.
Thorpe and O. L. Zangwill (Eds.), Cur-
rent problems in animal behaviour. Cam-
bridge University Press, London and New
York. 424 pp.

Bovbjerb, R. V.

1960. Behavioral ecology of the crab, Pachygrap-
sus crassipes. Ecology 41: 668-672.

Crane, J.

1957. Basic patterns of display in fiddler crabs
(Ocypodidae, Genus Uca). Zoologica 42:
69-82.

1958. Aspects of social behavior in fiddler crabs
with special reference to Uca maracoani
(Latreille). Zoologica 43: 113-130.

1966. Combat, display and ritualization in fiddler
crabs (Ocypodidae, Genus Uca). In A
discussion on ritualization of behavior
in animals and man, organized by Sir
Julian Huxley. Phil. Trans. Roy. Soc.
Lond. Ser. B. No. 772; Vol. 251: 213-226.

Dembowski, J.

1925. On the “speech” of the fiddler crab, Uca
pugilator. Travaux Inst. Nencki 3: 1-7.

Guinot-Dumortier, D. & B. Dumortier

1960. La stridulation chez les crabes. Crustace-
ana 1: 117-155.

Hagen, 'H. O. v.

1962. Freilandstudien zur sexual -und- Fortp-
flanzungsbiologie von Uca tangeri in An-
dalusien. Zeit. Morph, u. Okol. Tiere 51:
611-725.

Hiatt, R. W.

1948. The biology of the lined shore crab, Pachy-
grapsus crassipes Randall. Pacific Science
2: 135-213.

Hinde, R. A.

1966. Animal behavior. McGraw-Hill Book
Company, New York. 534 pp.

76

Zoologica: New York Zoological Society

[52:6

Huxley, J. S.

1914. The courtship-habits of the great crested
grebe (Podiceps cristatus ) with an addi-
tion to the theory of sexual selection. Proc.
Zool. Soc. Lond. p. 491.

1966. Introduction. In A discussion on ritualiza-
tion of behaviour in animals and man,
organized by Sir Julian Huxley. Phil.
Trans. Roy. Soc. Lond. Ser. B. No. 772;
Vol. 251: 249-271.

Huxley, J. S. et al.

1966. A discussion on ritualization of behaviour
in animals and man, organized by Sir
Julian Huxley. Phil. Trans. Roy. Soc.
Lond. Ser. B. No. 772; Vol. 251: 247-526.

Lorenz, K. Z.

1941. Vergleichende Bewegnugstudien an Anti-
nen. J. fur Ornithol. 83: 137-213, 289-413.

1964. Ritualized fighting. In J. D. Carthy and
F. J. Ebling (Eds.), The natural history of
aggression. Academic Press, London and
New York. 39-50.

1966a. On aggression. (Trans, by Marjorie Kerr
Wilson.) Harcourt, Brace & World, New
York. 306 pp. (Originally published, 1963,
under the title Das Sogenannte Bose. Bor-
otha-Schoeler, Vienna.)

1966b. Evolution of ritualization in the biological
and cultural spheres. In A discussion on
ritualization of behaviour in animals and
man, organized by Sir Julian Huxley. Phil.
Trans. Roy. Soc. Lond. Ser. B. No. 772;
Vol. 251: 273-284.

Pearse, A. S.

1912. The habits of fiddler-crabs. Philippine J.
Sc. D, 7: 113-133.

Reese, E. S.

1964. Ethology and marine zoology. Oceanogr.
Mar. Biol. Ann. Rev. 455-488.

Salmon, M.

1965. Waving display and sound production in
the courtship behavior of Uca pugilator,
with comparisons to U. minax and U.
pugnax. Zoologica 50: 123-149.

Salmon, M. & J. F. Stout

1962. Sexual discrimination and sound produc-
tion in Uca pugilator Bose. Zoologica 47:
15-19.

SCHALLER, G. B.

1967. The deer and the tiger: a study of wildlife
in India. University of Chicago Press,
Chicago. 370 pp.

SCHONE H. & H. SCHONE

1963. Balz und andere Verhaltensweisen der
Mangrovekrabbe Goniopsis cruentata Latr.
und das Winkverhalten der eulitoralen
Brachyuren. Zeit. fur Tierpsychologie 20:
641-656.

Thorpe, W. H.

1966. Ritualization in ontogeny. I. Animal play.
In A discussion on ritualization of behav-
iour in animals and man, organized by Sir
Julian Huxley. Phil. Trans. Roy. Soc.
Lond. Ser. B. No. 772; Vol. 251: 311-319.

Tinbergen, N.

1952. “Derived” activities; their causation, bio-
logical significance, origin and emancipa-
tion during evolution. Quart. Rev. Biol.
27: 1-32.

1953. Social behaviour in animals. Methuen and
Co., Ltd., London. 150 pp.

Tweedie, M. W. F.

1954. Notes on grapsoid crabs from the Raffles
Museum Nos. 3, 4 and 5. Bull. Raffles Mus.
Singapore (25) : 118-127.

Verwey, J.

1930. Einiges fiber die Biologie ost-indischer
Mangrovekrabben. Treubia 12: 169-261.

EXPLANATION OF THE PLATE

Plate I

Combat in Uca rapax (from 16-mm. motion picture

film). See text, pp. 57—59.

Fig. 1. Tap following a heel-and-ridge in homo-
clawed combat. The actor is the crab on
the right. His dactyl is striking the heel of
his opponent’s manus while his pollex is
free.

Fig. 2. Same combat as in Fig. 1. Alternate stroke

showing the actor’s pollex against his op-
ponent’s invisible oblique ridge, on inner
side of manus.

Fig. 3. Interlace in a heteroclawed combat. The
actor is on the right. The teeth near his
dactyl’s base are starting to rub down-
ward against the ridges of his opponent’s
inner manus, which parallel the dactyl’s
base.

Fig. 4. Same, near end of downward stroke of rub.

CRANE

PLATE i

FIG. 1

FIG. 3

FIG. 4

FIG. 2

COMBAT AND ITS RITU ALIZATION IN FIDDLER CRABS (OCYPODIDAE)
WITH SPECIAL REFERENCE TO UCA RAPAX (SMITH)

[1967]

Zoologica: Index to Volume 52

79

Numbers in parentheses are the
series numbers of papers contain-
ing the tables, figures, or plates
listed immediately following.
Numbers in bold face indicate
text-figures.

A

Agraulis vanillae, 2, 3, 5-6, CO
5, 11-12, 13

B

Butterflies. See Agraulis vanillae ;
Heliconius ricini; Heliconius
sarae; Morpho peleides;
Philaethria dido ; Pterins;
Victorina steneles; Wing
coloration

C

Colaenis julia, 15, (2) Table II,
16-18, 19, (2) 4

Combat and its ritualization. See
Fiddler crabs

E

Erythropterin, 15, (2) 1, 16, (2)
Tables I— II, 18, (2) 3, 19
See also Pterins

F

Fiddler crabs, combat and its rit-
ualization:

Uca rapax, 49-77, (6) 1-3, (6)
Tables I-IX, (6) PL I
adaptive values, 72-73
associated activities, 60-62,
(6) Table IV

behavior components, mor-
phological specializations,
57-60, (6) 2-3
categories, 62-63, CO Tables
V-VI

characteristics of opponents,
55-56, (6) Tables I— III
comparison with other ani-
mals, 73

definitions, 52-55
derivation of components, 73
duration of combats, 63-66,
(6) Tables V-VI
functions, 70—72
further research, areas for,
73-74

postcombat behavior, 66-69,
(6) Table VIII
terminology, 52, (6) 1
species other than U. rapax,
69-70, (6) Table IX
Fish schools, survival value,
25-40

definitions, 25-27
environment, influence of :
olfaction, 31

INDEX

ontogenetic change, 31—32
sound production and pre-
vention, 29—31
visibility and transparency
of water, 27-29
prey and predator, relative
size, 34—35

structure and size of schools:
enormous schools, problem
of, 32-34

shape of schools, 32

H

Heliconius erato, 1—2, 3, 4, 6, ( 1 )
3, 10-11, 12, 13, 15-16, (2)
Tables I— II, (2) 2, 18

Heliconius ricini, 3, 4—5, CO 3-4,
10, 12-13

Heliconius sarae, 2, 3—4, CO 1-3,
5, 7, 10-11, 12-13, 18

L

Longevity of Naja naja philip-
pinensis. See Venom extrac-
tion

M

Morpho peleides, 2, 3, 7—8, COB,
12, 13

N

Naja naja philippinensis, longev-
ity under stress of venom ex-
traction, 41-47, C5) Tables I—

III, C5) 1-3

See also Venom extraction

P

Philaethria dido, 2, 3, 6-7, CO
6—7 12 13

Pterins, 15-20, CO Tables I— II,
C2) 1-4

Colaenis julia, 15, CO Table
II, 16-18, C2) 4, 19
Heliconius erato, 15-16, CO
Tables I— II, C2) 2, 18
Heliconius sara, 18
See also Erythropterin

S

Schools, fish. See Fish schools,
survival value

Sound production, underwater,
Zalophus caliiornianus,

21- 24, C3) Pis. I-V
recorded sounds:

bang or crack, 22, C3) PL V
barks, 22, C3) PL II
buzzing, 22, C3) PL IV
clicks, 21-22, C3) Pl. I
whinny, 22, C3) PL III
sound production mechanisms,

22- 23

spectrographs, 23, C3) Pis. I-V

Spectral efficiency curves, 2,
10-12, 13

Agraulis vanillae, 5, CO 5,
10-12, 13

Heliconius erato, 10-12
Heliconius ricini, 5, CO 4, 13
Heliconius sarae, 3-4, CO
1-2, 10-12, 13

Morpho peleides, 7-8, COB,
12, 13

Philaethria dido, 6, 7, CO 7,

12, 13

Victorina steneles, 6-7, C O 7,
12, 13

Spectral reflectance
characteristics, 2
Heliconius erato, 3, CO 3
Heliconius ricini, CO 3
Heliconius sarae, 3, CO 3
Philaethria dido, 6, CO 6
Victorina steneles, 6, CO 6

Spectrographs. See Sound produc-
tion

U

Uca rapax, combat and its rituali-
zation, 49-77, C6) 1—3, CO
Tables I-IX, C6) Pl. I
See also Fiddler crabs

V

Venom extraction, longevity under
stress of:

Naja naja philippinensis, 41-47,
C5) Tables I— III, CO 1-3
collection of venom, 42
adverse affect, 45, CO
Tables I-II, C5) 1-3, 47
controls, 44, CO Tables II,
47

death rate and time of year,
44, C5) 3, 46, 47
schedules, effects of, 42-43,
C5) Table I, C5) 1
time acquired, 44, CO
Table III, 46-47
feeding, 42
housing, 42

Victorina steneles, 3, 6-7, C 1 )
6-7, 12, 13

W

Wing coloration, neural adapta-
tions to in butterflies, 1-14,

Cl) 1-8

See also Spectral efficiency
curves; Spectral reflectance
characteristics

Z

Zalophus calif ornianus, 21—23,

C3) Pis. I-V

NEW YORK ZOOLOGICAL SOCIETY

GENERAL OFFICE

630 Fifth Avenue, New York, N.Y. 10020

PUBLICATION OFFICE

The Zoological Park, Bronx, N.Y. 10460

Fairfield Osborn
President

Laurance S. Rockefeller
First Vice-President

OFFICERS

Henry Clay Frick, II
Robert G. Goelet
Vice-President

John Elliott
Secretary

David Hunter McAlpin
Treasurer

Eben W. Pyne
Assistant Treasurer

SCIENTIFIC STAFF

William G. Conway
General Director

ZOOLOGICAL PARK

William G. Conway . Director & Curator, Birds F. Wayne King . . Associate Curator, Reptiles

Hugh B. House Chairman & Charles P. Gandal Veterinarian

Associate Curator, Mammals Lee S. Crandall . . . General Curator Emeritus

Grace Davall . . Assistant Curator, Mammals & Zoological Park Consultant

<6 Birds William Bridges Curator of Publications Emeritus
Victor H. Hutchison . . . Research Associate in John M. Budinger . . . Consultant, Pathology

Herpetology David W. Nellis Mammalogist

Joseph Bell .... Associate Curator, Birds James G. Doherty Mammalogist

Donald F. Bruning Ornithologist

AQUARIUM

Ross F. Nigrelli Director Nixon Griffis .... Administrative Assistant

Christopher W. Coates . . . Director Emeritus Robert A. Morris .... Associate Curator

Louis Mowbray Research Associate in Field Biology

GENERAL

Edward R. Ricciuti . . Editor & Associate Curator, Publications

Elysbeth H. Wyckoff . . . Production Editor

Dorothy Reville Photo Librarian Sam Dunton . Photographer

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . . Director and Pathologist Harry A. Charipper . . Research Associate in

Martin F. Stempien, Jr. . . . Assistant to the Histology

Director & Bio-Organic Chemist Kenneth Gold Marine Ecologist

George D. Ruggieri, S.J. . . . Coordinator of Myron Jacobs Neuroanatomist

Research & Experimental Embryologist Klaus Kallman Fish Geneticist

William Antopol . . . Research Associate in Vincent R. Liguori Microbiologist

Comparative Pathology John J. A. McLaughlin . Research Associate in
C. M. Breder, Jr. . . . Research Associate in Planktonology

Ichthyology Martin P. Schreibman . . Research Associate in

Jack T. Cecil Virologist Fish Endocrinology

Eva K. Hawkins Algologist

INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin .... Director & Senior Fernando Nottebohm . . . Research Zoologist

Research Zoologist George Schaller Research Zoologist

Peter R. Marler . . . Senior Research Zoologist Thomas T. Struhsaker . . . Research Zoologist

Jocelyn Crane . . . Senior Research Ethologist C. Alan Lill Research Fellow

Roger S. Payne Research Zoologist O. Marcus Buchanon . . . Resident Director,

Richard L. Penney .... Research Zoologist William Beebe Tropical Research Station

EDITORIAL COMMITTEE
Fairfield Osborn
Chairman

William G. Conway Donald R. Griffin Edward R. Ricciuti

Lee S. Crandall Hugh B. House Ross F. Nigrelli

Peter R. Marler

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 53 • 1968 • NUMBERS 1-8

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

NEW YORK ZOOLOGICAL SOCIETY

The Zoological Park, Bronx, N. Y. 10460
OFFICERS

Fairfield Osborn Laurance S. Rockefeller Robert G. Goelet

Chairman of the Board of Trustees President Executive Vice-President

Henry Clay Frick, II Chairman of the Executive Committee
John Pierrepont Vice-President Howard Phipps, Jr.

Treasurer Eben w Pyne Secretary

Assistant Treasurer

Edward R. Ricciuti
Editor & Curator,
Publications & Public Relations

Joan Van Haasteren
Editorial Assistant

EDITORIAL COMMITTEE

William G. Conway
Lee S. Crandall
Donald R. Griffin

Robert G. Goelet
Chairman

F. Wayne King

Hugh B. House Peter R. Marler

Ross F. Nigrelli

William G. Conway
General Director

ZOOLOGICAL PARK

William G. Conway . . . Director & Curator,

Ornithology

Hugh B. House .... Curator, Mammalogy
Grace Davall . . Assistant Curator, Mammals

& Birds

Walter AufFenberg . . Research Associate in

Herpetology

Joseph Bell . . Associate Curator, Ornithology

F. Wayne King .... Curator, Herpetology
Joseph A. Davis, Jr

Charles P. Gandal Veterinarian

Lee S. Crandall . . . General Curator Emeritus

& Zoological Park Consultant
William Bridges . Curator of Publications Emeritus
John M. Budinger . . . Consultant, Pathology

Ben Sheffy Consultant, Nutrition

James G. Doherty Mammalogist

Donald F. Bruning Ornithologist

Scientific Assistant to the Director

Ross F. Nigrelli . .

Christopher W. Coates
Nixon Griffis . . .

AQUARIUM

Director Robert A. Morris Curator

Director Emeritus U. Erich Friese Assistant Curator

Administrative Assistant Louis Mowbray . Research Associate in Field Biology

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . Director and Pathologist

Martin F. Stempien, Jr. ... Assistant to the
Director & Bio-Organic Chemist
George D. Ruggieri, S.J. . . . Coordinator of

Research & Experimental Embryologist
William Antopol . . . Research Associate in

Comparative Pathology
C. M. Breder, Jr. ... Research Associate in

Ichthyology

Jack T. Cecil Virologist

Jay Hyman

Harry A. Charipper . . Research Associate in

Histology

Kenneth Gold Marine Ecologist

Eva K. Hawkins Algologist

Myron Jacobs Neuroanatomist

Klaus Kallman Fish Geneticist

Vincent R. Liguori Microbiologist

John J. A. McLaughlin . . Research Associate in

Planktonology
Research Associate in
Fish Endocrinology

Martin P. Schreibman
Research Associate in Comparative Pathology

INSTITUTE FOR RESEACH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society's William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin .... Director & Senior

Research Zoologist
Peter R. Marler . . . Senior Research Zoologist

Jocelyn Crane . . . Senior Research Ethologist

Roger S. Payne Research Zoologist

Richard L. Penney .... Research Zoologist

Fernando Nottebohm . . . Research Zoologist

George Schaller Research Zoologist

Thomas T. Struhsaker . . . Research Zoologist

C. Alan Lill Resident Director

O. Marcus Buchanan . . . Resident Director,

William Beebe Tropical Research Station

Contents

Issue 1. April 15, 1968

PAGE

1 . Influence of Climate on the Distribution of Walruses, Odobenus rosmarus
(Linnaeus). I. Evidence from Thermoregulatory Behavior. By Francis

H. Fay & Carleton Ray. Plates I-IV; Text-figures 1-2 1

2. Influence of Climate on the Distribution of Walruses, Odobenus rosmarus

(Linnaeus) . II. Evidence from Physiological Characteristics. By Carleton
Ray & Francis H. Fay. Text-figures 1-9 19

3. Thermoregulation of the Pup and Adult Weddell Seal, Leptonychotes

weddelli (Lession), in Antarctica. By Carleton Ray & M. S. R. Smith.
Plates I-II; Text-figures 1-8 33

Issue 2. July 19, 1968

4. The Pentastomes, Waddycephalus teretiusculus (Baird, 1862) Sambon,

1922 and Parasambonia bridgesi n. gen., n. sp., from the Lungs of the
Australian Snake, Pseudechis porphyriacus. By Horace W. Stunkard &
Charles P. Gandal. Text-figures 1-6 49

5. Eastern Pacific Expeditions of the New York Zoological Society. Porcel-

lanid Crabs (Crustacea: Anomura) from the West Coast of Tropical
America. By Janet Haig. Text-figures 1-2 57

Issue 3. November 29, 1968

PAGE

6. Observations on the African Bushpig Potamochoerus porcus Linn, in

Rhodesia. By Lyle K. Sowls & Robert I. Phelps. Plates I-II; Text-
figures 1-8 75

7. The Breeding Biology of the Male Brown Bear (Ursus arctos). By Albert
W. Erickson, Harland W. Moosman, Richard J. Hensel, & Willard

A. Troyer. Plates I-IX; Text-figures 1-2 85

Issue 4. March 19, 1969

8. Host and Ecological Relationships of the Parasitic Helminth Capillaria
hepatica in Florida Mammals. By James N. Layne. Plate I; Text-figures

1-4 107

Index to Volume 53 123

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 53 • ISSUE 1 • SPRING, 1968

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

1. Influence of Climate on the Distribution of Walruses, Odobenus rosmarus

(Linnaeus). I. Evidence from Thermoregulatory Behavior. By Francis
H. Fay & Carleton Ray. Plates I-IV; Text-figures 1-2 1

2. Influence of Climate on the Distribution of Walruses, Odobenus rosmarus

(Linnaeus) . II. Evidence from Physiological Characteristics. By Carleton
Ray & Francis H. Fay. Text-figures 1-9 19

3. Thermoregulation of the Pup and Adult Weddell Seal, Leptonychotes

weddelli (Lesson), in Antarctica. By Carleton Ray & M. S. R. Smith.
Plates I-II; Text-figures 1-8 33

Erratum: Fall- Winter, 1967, Zoologica was Issue 3-4
containing papers 5 and 6, not Issue 3.

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park, Bronx Park, Bronx, N. Y. 13460, and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers, $1.50, unless otherwise stated in the Society’s catalog of publications. Second-class
postage paid at Bronx, N. Y.

Published April 15, 1968

1

Influence of Climate on the Distribution of Walruses, Odobenus
rosmarus (Linnaeus). I. Evidence from Thermoregulatory Behavior.

Francis H. Fay1 and Carleton Ray2
(Plates I-IV; Text-figures 1-2)

Introduction

THE walrus is one of a group of pinnipeds
typically associated with the ice front in
northern seas, but like most of the others,
it is by no means restricted to the front. Some
walruses occur as much as 1 ,000 miles south of
it in summer and some as much as 500 miles
north of it in winter. Within this range, the
walrus resides chiefly in the shallow waters of
the continental shelf, where its food of mollusks
and other benthic invertebrates is obtained at
depths of 80 meters or less (Vibe, 1950). The
area occupied by these mammals on a year-
round basis thus comprises parts of two marine
zoogeographic zones, the Arctic and the Boreal
or subarctic (Ekman, 1953; Zenkevitch, 1963),
but does not include the full extent of shallows
in either one. The failure of walruses to occupy
all of the shoals and inshore waters of the Arctic
Zone seems to be clearly a matter of their in-
ability to penetrate regularly into the polar ice-
pack or to obtain food in some areas where
mollusks are scarce (Fay, 1957) . Their occupa-
tion of only the northern part of the Boreal Zone
may be due to restrictions imposed by the
climate, which varies from subarctic in the

iBiologist, Arctic Health Research Laboratory, Pub-
lic Health Service, U.S. Department of Health, Educa-
tion, and Welfare, College, Alaska.

2Department of Pathobiology, School of Hygiene
and Public Health, The Johns Hopkins University,
Baltimore, Maryland. Formerly, Curator, New York
Aquarium, New York Zoological Society, Coney Island,
New York.

northern part to low temperate in the south. A
correlation between the southern limit of the
walrus’ range and isothermal lines was noted
more than a century ago by von Baer (1838,
fide Allen, 1880:91), and we find this to be
generally true today. The majority of these ani-
mals occurs in areas where monthly mean air
temperatures are from —15 to +5°C, and only
a few vagrants range south of the 10°C isotherm
at any time (Text-fig. 1). In the study reported
here, we set out to test the theory of a southern
climatic boundary, not by comparing distribu-
tion with thermal conditions, but by examining
the behavioral and physiological responses of
walruses to subarctic and temperate climates. In
this paper we report on the behavioral aspects
of the study; the physiological findings are re-
ported separately (Ray & Fay, 1968).

The material presented here is of two kinds.
First are descriptions of the postures and other
physical adjustments of walruses that affect the
amount of exposed surface area and could in-
fluence the rate of heat loss from the body to
the ambient. Second is a quantitative compari-
son of the weather when the animals were out
versus in the water. Walruses spend about as
much time out of the water as in it, and the
frequency and duration of their lying out seem
to be affected by the weather conditions at the
time (Shuldham, 1775, fide Allen, 1880:67;
Hayes, 1867:404; Nikulin, 1947). Since their
presence in or out of the water may depend also
on the normal alternation of activity and rest,
we have included an investigation of the normal
activity rhythm.

1

2

Zoologica: New York Zoological Society

[53: 1

90°E

Text-fig. 1. The north polar region, showing the present distribution of walruses in summer in relation to
the minimum extent of the permanent ice pack and the isotherm of 10° C mean air tempera-
ture for July.

Materials and Methods

Information on the behavior of free-living
walruses was obtained mostly by Fay during the
period 1952 to 1965, in the vicinity of St.
Lawrence Island, Alaska, just south of Bering
Strait. In the course of approximately 400 hours
spent hunting walruses with the Eskimos of that
area, at least 1,190 adults, subadults, and juve-
niles were seen, plus many calves not included
in the counts. More than 8,000 others were seen
during aerial surveys of the Bering Sea. Most of

these animals were sighted during the daytime,
between 0800 and 1600 hours, when more than
four-fifths of them were lying out on ice floes;
the rest were swimming or feeding in the water.
Although the Eskimos’ objective during the
hunts was to kill the animals for food, observa-
tion of undisturbed animals was usually possible
for several minutes before the shooting occurred.
In that time, behavior of potential thermoregu-
latory value was observed.

Most of the field observations were made in
May, when the Pacific walrus population was

1968] Fay & Ray: Influence of Climate on Distribution of Walruses. 1. Thermoregulatory Behavior 3

concentrated in the northern end of the Bering
Sea. The behavior of the animals in that time
and place is assumed to have been representative
of their responses to weather slightly warmer
than the mean for their year-round environment.
Observations in the vicinity of St. Lawrence
Island during January to March provided some
indications of the reactions to the coldest
weather to which walruses are ordinarily ex-
posed in the Bering Sea region; the island is the
northern limit of their range at that time (see
Brooks, 1954). For observation during the
warmest weather in summer, Fay went to Round
Island (58° 30' N, 160° W), Bristol Bay, Alaska,
the southernmost area regularly occupied by
Pacific walruses in that season. About 1,500
males were observed there during a four-day
visit at the end of June, 1958.

The data obtained in the field on the activity
rhythm and reactions to weather were mostly
notations of the time, location, and number of
animals seen in (swimming) or out (resting)
of the water. Only those animals that were un-
disturbed were counted; alarmed walruses in-
variably took refuge in the water. These notes
were later correlated with meteorological data
recorded by us or by personnel of a nearby
weather station. The data were biased to the
extent that it was usually not possible to observe
the animals during periods of extremely stormy
weather.

Our information on the behavior of walruses
in captivity was obtained from 1957 to 1963,
principally by Ray at the New York Aquarium.
The animals from which most of the informa-
tion was obtained were an adolescent male,
probably of Greenlandic origin, and a juvenile
female from the Bering Sea. Both had been in
captivity since infancy. Data on their activity
rhythm and reactions to weather were obtained
in 1960, during a six-month period of close sur-
veillance. At that time, the male was about five
years old, and the female was about one year
old. At regular intervals each day, notations
were made whether each animal was in the
water or hauled out on its resting platform, and
these were correlated with hourly weather re-
corded by the U. S. Weather Bureau, 17 Battery
Place, New York City. The weather records
were less applicable than on-site micrometeoro-
logical data would have been, but the conditions
described by them were similar to the general
weather at the aquarium. One notable excep-
tion to this was wind velocity; winds were
stronger and more frequent at the Battery than
at the aquarium, and within the sheltering walls

of their enclosures the animals were further
removed from the effects of all but strong winds.

From January 18 to May 30, notations on
activity were made several times daily between
0800 and 1600 hours and, after May 30, also at
0200, 0600, and 2200 hours. At 0800 hours
daily, throughout the six-month period, a nota-
tion was made also of the presence or absence of
feces on each animal’s resting platform. Since
the animals seldom defecated on the platforms
except when they stayed out for an hour or more
beforehand, we used this as an index of their
having spent some time out of water during the
night. In June and July, when both this index
and their regular nighttime observations were
recorded, they showed close agreement.

The captives were kept in separate, walled
enclosures about 20 meters apart and were visu-
ally, but not acoustically or olfactorily, isolated
from each other. Since vocal communication
between them was rarely detected, and since
neither animal was sexually mature, we are con-
fident that any intercommunication that did
occur did not seriously influence their behavior.
The male’s enclosure also held three gray seals,
Halichoerus grypus, but the walrus was domi-
nant over these and virtually unaffected by their
presence. Neither did the presence of human
spectators seem to distract either animal to the
extent that its activity rhythm was affected. Al-
though the spectators were not ignored while
the walruses were swimming, the animals were
largely oblivious to all human activity when they
hauled out to rest.’ The aquarium was open to
the public from 1000 to 1700 hours daily to May
30 and from 1000 to 2200 hours thereafter. The
only significant human influences on the be-
havior of the animals during that time were the
daily feedings; each animal usually hauled out
on its resting platform when the keeper arrived
with the food at 1 030 to 1100 hours and 1 5 30 to
1600 hours. Spot observations were recorded at
those times each day before feeding.

The juvenile female was one of 21 individuals
obtained at St. Lawrence Island between 1958
and 1963. All of these were very young animals
when captured, ranging in age from newborn to
three or four weeks old. They were held in pens
on the island for up to two weeks before being
transferred by aircraft to the aquarium. In that
time, we obtained information on their reactions
to the local weather and, in a few cases, to a
wide range of experimentally imposed thermal
conditions. Comparative information on their
reactions to hot summer weather was obtained
after their arrival in New York.

4

Zoologica: New York Zoological Society

[53: 1

Results

Regulation of Surface Exposure in Air
Four methods were used by walruses to regu-
late exposure of their body surface and appen-
dages to the ambient while at rest out of water.
These were huddling, posture, fanning, and se-
lection of substrates. These will be considered
separately, though they often occurred simul-
taneously.

Huddling. Of nearly 10,000 walruses seen by
us from small boats and aircraft, less than three
per cent were alone; the rest were in groups of
from two to several hundred. The mean size of
the groups tended to be smallest when the ani-
mals were in the water, larger when they were on
ice, and largest when they were on land (Table
I). One characteristic of each group resting on

Table I.

Group Size of Wild Pacific Walruses
in Relation to Their Location.3

No. of Animals

Location No. of No. of per Group

Animals Groups Range Mean

In water

339

94

1- 50

3

On ice

6274

388

1-600

16

On land

3254

21

1-850

155

a From unpublished data obtained by J. W. Brooks,
K. W. Kenyon, A. Thayer, and F. H. Fay during aerial
surveys and observation from small boats, Bering Sea,
1952 to 1962.

ice or land was the intense mutual contact be-
tween its members, which lay . . huddling like
swine, one over the other” (Cook, 1822:680).
This is at once apparent to anyone seeing a rest-
ing herd for the first time (Plate I), and it has
been mentioned many times previously in ac-
counts by naturalists and other explorers of
arctic regions (e.g., see review by Allen, 1880:
107-121, 178-180). In captivity, also, walruses
show a high degree of gregariousness and thig-
motaxis; when two or more of those that we
studied were kept in the same pen, they almost
invariably slept huddled together. We estimated
that the usual extent of contact in groups, both
in the field and in the aquarium, was about
20 per cent of the total body surface per animal.

The degree of mutual contact within groups
did not appear to vary seasonally, with latitude
or with air temperature or other weather condi-
tions. Herds on the beach at Round Island in
June, under clear skies and in 14°C air, were
apparently as tightly packed as those on the ice

in January when the sky was cloudy and air as
cold as— 19°C. In the aquarium, also, the ani-
mals huddled together to sleep, regardless of
whether they were cold, warm, or hot; under the
warmest conditions, both the wild and the cap-
tive animals showed signs of heat stress.

In the field, we observed another type of hud-
dling, characteristic of mother-and-calf pairs,
which we called “brooding” in as much as it
seemed to be of benefit to the calf alone. In this
case, there were distinct variations in the degree
of contact that seemed to be adaptively related
to the weather. For example, we noticed at first
that very few calves were in evidence during the
chilliest days, but in sunny weather with little or
no wind they were frequently seen standing or
lying on the ice beside the mother. Subsequently,
we discovered that in cold weather the calf was
usually situated against the mother’s breast, be-
tween her forelimbs, and so completely con-
cealed and sheltered that its presence was not
detected until the mother became alarmed and
began to flee. Calves in this position were esti-
mated to have at least 50 per cent of their body
surface in contact with that of the adult. When
removed from this maternal shelter and exposed
alone to the chilly weather, the calves reacted by
seeking protection from the wind, huddling
against any warm body or low-conductance ma-
terial, assuming a “fetal” position, and shiver-
ing violently. They were obviously chilled and
we concluded that the warmth derived from
maternal brooding was an important and, pos-
sibly, essential component of their environment.
In our experience, weather severe enough to
bring about these responses occurred, on the
average, in at least three out of four days during
the calving season (April-May), and sometimes
lasted for ten days at a time.

Posture and fanning. By varying their pos-
ture and the position of their appendages, wal-
ruses at rest are capable of controlling the
amount of exposed surface area. Maximum ex-
posure is attained by sprawling on the back with
head and neck extended and flippers out-
stretched and spread. Minimum exposure is ef-
fected by assuming the “fetal” position, with
head drawn in, back arched, and flippers pressed
tightly against the body. From comparative pho-
tographs of one calf in both positions (Plate II) ,
we estimate that the amount of surface exposed
in the fetal posture is only about three-fourths
as much as in the sprawling posture. That pos-
tural regulation of exposed surface is influenced
by ambient temperature was determined experi-
mentally. Four newly captured calves, each in
a separate wooden crate, were exposed to an in-

1968] Fay & Ray: Influence of Climate on Distribution of Walruses. 1. Thermoregulatory Behavior 5

crease of 3 or 4°C every 15 minutes. At the
lowest temperature, about 1°C, each animal
assumed an extreme fetal position with occa-
sional violent shivering. As the temperature was
raised, each became more relaxed and paid less
attention to keeping its appendages against the
body. At 10°C, the animals became fully relaxed
and lay either on the back or side; at 15°C they
began to sprawl and extend their appendages; at
18°C they became restless and began fanning
intermittently with their flippers; at 20°C they
were so restless that the experiment was
terminated.

The postural adjustments of isolated wild
adults in relation to air temperatures were simi-
lar to those of the experimental calves. There
was clearly a trend to fetal postures at low
temperatures (Plate III, fig. 5) and to sprawling
at high temperatures, but it was usual to find a
wide variety of postures under the latter condi-
tions (Plate III, fig. 6). We found that the fetal
posture was almost always assumed when the
animals first emerged from the water and was
maintained for some time, even when the air
temperature was relatively high. In air warmer
than 10°C, the emergent animals usually relaxed
to a more or less sprawling position after about
30 minutes, or when their skin became dry. We
assume that the change in posture reflected a
change in the rate of heat loss per unit of sur-
face, i.e., in relation to evaporative cooling.

Fanning by walruses in the aquarium was
often seen when ambient temperatures were
higher than 20°C but rarely at lower tempera-
tures. In general, fanning animals were visibly
hyperemic and hot to the touch, indicating rapid
dissipation of body heat. Fanning and super-
ficial hyperemia were seen also in the herds at
Round Island, where they were lying in the sun
in 13 to 14°C air (Plate IV, fig. 7). We did not
see fanning or hyperemia in animals on the ice,
even when air temperatures were as high as
7°C, but J. J. Burns (personal communication)
saw fanning by some adult males when the tem-
perature was about 8.5°C (Plate IV, fig. 8).

Selection of substrates. A group of three
newly captured calves, held in an outdoor pen
with a floor of snow, had access to a 1-meter-
square piece of plywood and approximately
equal areas of canvas and of fresh skin from an
adult walrus. The group elected to sleep hud-
dled on the wood, rather than on the snow or
other materials, and their persistent use of it
indicated that they found it preferable, perhaps
because of its lower conductance and specific
heat. Pieces of plywood were then provided to
several younger calves in separate pens, and

these also were consistently used as beds, in
preference to the ice and snow of the pen floor.

A positive response in cold weather to sub-
strates of low conductivity and specific heat was
suggested also by the behavior of the captives in
New York. When they hauled out to rest in the
winter, they clearly avoided snow-covered sur-
faces, and the female selected a wooden pallet
rather than an adjacent concrete platform. The
male walrus had only a wooden platform on
which to lie.

Probable selection of substrates by free-living
walruses was noticed only under the warmest
conditions at Round Island. Several times, rest-
ing animals were observed to grope about with
their hind flippers and, on touching a damp,
shaded rock, to press the spread flippers firmly
against it, as if aware of its coolness. Although
the temperature of the damp rocks in shade was
only 1 or 2°C lower than that of the air, they
were distinctly cooler to the touch because of
their high specific heat and conductivity.

Influence of Weather on Emergence

The opinion that walruses prefer to haul out
in sunny weather was expressed by Shuldham
( 1775, fide Allen, 1880:67) and Hayes
(1867:404), based on their observations of the
animals under natural conditions. We formed
the same opinion, independently, from our in-
itial nonsystematic observations of both wild
and captive animals and noted, furthermore,
that they seemed to stay in the water during
windy or stormy weather. The latter was noticed
also by Nikulin (1947) and Mansfield
(1958:115). These opinions were tested quan-
titatively by means of a sampling system, in
which periodic spot-observations of the activi-
ties of captive walruses were recorded by
impartial observers and correlated with meteoro-
logical data, supplied by the Weather Bureau.
Inasmuch as the emergence of the animals and
the duration of their exposure to weather might
be governed also by regular daily and seasonal
cycles of activity, the data were analyzed first
for evidence of activity rhythms.

Activity rhythm. Other than an opinion ex-
pressed by some Eskimos to Loughrey (1959:39)
that walruses feed mostly early in the morning
and haul out to rest in the remainder of the day,
there is no published information available on
the normal alternation of rest and activity in
these animals. We obtained an estimate of the
mean daily activity rhythm of wild walruses by
compiling a series of data on several thousand
that were sighted in the Bering Sea in the months
of January to June, 1952 to 1960 (F. H. Fay,

6

Zoologica: New York Zoological Society

[53: 1

K. W. Kenyon & A. Thayer, unpublished). A
comparable estimate for walruses in captivity
was obtained from more than 1,000 spot-obser-
vations of the two juveniles in the New York
Aquarium. January to July, 1960. Animals
sighted in the water were considered to have
been “active;” those sighted on land or ice were
considered as “inactive.” The relative number
of active animals per unit of time was found to
show a general circadian rhythm in both the
natural and the artificial environments (Text-
fig. 2). The animals tended to be most active in
the forenoon and evening and to haul out most
often in early morning and early in the after-
noon. The respective proportions of walruses in
and out of the water suggested that the wild
animals were less active in the daytime and more
active at night than were the captives, perhaps
because of differences in their feeding times.

We did not detect any significant changes in

the mean circadian rhythm per month, from
January to July, except in the intensity of ac-
tivity. In both the wild walruses and those in
captivity, more time was spent in the water in
January than in any other month (Table II).
The wild animals were out of the water most
often in February and March, at the height of
the mating season (Fay, unpublished); the cap-
tives were out most in March and April. From
April to July, the captives hauled out with in-
creasing frequency at night and decreasing fre-
quency in the daytime.

Activity in relation to weather: captive wal-
ruses. Four kinds of meteorological data were
utilized for comparison of the weather when
the animals were out of the water (inactive)
with that when they were in the water (active).
These were: air temperature, sky cover (in-
versely proportional to insolation) , precipitation
rate and wind velocity (Table III). Since, in a

Text-fig. 2. Comparative patterns of activity rhythms of wild and captive walruses, based on the percent-
age of occurrences of animals sighted in the water per hour.

1968] Fay & Ray: Influence of Climate on Distribution of Walruses. 1. Thermoregulatory Behavior 7

Relative

Activity

Table II.
of Walruses per

Month.

Locality and Time

Jan.

Feb.

Mar.

Apr.

May

June

July

Wild walruses, Bering Sea

Daytime, total sighted

323

261

10124

4569

5891

648

Daytime, per cent active

77

4

14

22

29

39

Captive walruses, N.Y. Aquarium . .

Daytime, total observations

46

103

142

122

136

239

149

Daytime, % when active

93

89

74

63

76

79

81

Nighttime, total observations ....

25

54

57

45

60

59

38

Nighttime, % when active

76

91

75

62

52

59

47

Table III.

Comparison of Monthly Mean Weather when Captive Walruses Were out of the Water
with That when They Were in the Water.

Wind,

Mean Mean 30 mph

Month No. of Mean Temp. Daytime Precip., or More

Observations (°C) Sky Cover3 mm/hr (% Occur.)

Out In Out In Out In Out In Out In

January 77 ~ 3 "43 2.1 2.4 83 69b 0 .18 0 21

February 11 92 2.2 3.0 81 60b .13 .43b 27 38

March 37 105 2.7 0.7b 44 58b .13 .80b 11 19

April 45 77 11.3 10.9 52 59 .29 .42 11 13

May 33 129 15.2 16.5b 56 60 .13 .90 0 8

June 102 259 20.1 21. lb 51 56 .62 .51 2 4

July 74 169 21.8 22.3 41 54 .19 2.86 0 2

3 Expressed as per cent of total sky obscured by clouds; inversely proportional to isolation.
bP < .05.

preliminary analysis, we found the correlations
of activity and weather to be unaffected by the
circadian rhythm, all of the data were treated
equally. Monthly means were compared by ap-
propriate statistical tests; dilferences, when
P < .05, were considered as significant. Water
temperatures during the six-month period, from
mid-January to mid-July, ranged from 4 to
1 8°C, respectively.

During January and February, the animals
stayed in the water most of the time (Table II),
and on the few occasions when they did haul
out, the weather was slightly cloudier and cooler
but less windy than when they were in the water.
High winds, 13.4 meters per second (30 mph)
or stronger, occurred very frequently in both
months and were correlated to a significant de-
gree with the sunniest weather.

March was actually a cooler month than Feb-
ruary, though there were more sunny days. The
male was out during the day more often than
in any other month; the female still remained
in the water most of the time. The weather when

either of them was out was warmer, sunnier,
drier, and less windy than it was when they were
in the water. High winds were still correlated
with sunny weather, but they occurred less often
than in February.

In April, the female was out of the water
more often than the male during the day, and
this relationship persisted through July. The
daytime weather while either of them was out
was slightly warmer and sunnier than when they
were in the water, but about equally wet and
windy.

By May, both animals were out less frequently
during the day and more frequently at night.
When they were out of the water in the daytime,
the weather was slightly sunnier and less windy,
but cooler and drier than when they were in the
water. In this and the succeeding month, the
high winds occurred with cloudy skies.

The tendency toward hauling out less fre-
quently by day and more frequently by night
increased through June and July. In both
months, the weather when the animals were out

Zoologica: New York Zoological Society

[53: 1

of the water in the daytime was slightly sunnier;
in both the day and the night it was less windy
but cooler and about as rainy as when they were
in the water.

Consistently, throughout the six-month pe-
riod, the animals tended to avoid exposure to
high winds, irrespective of the other conditions
at the time. In addition, they showed a negative
response to precipitation, especially in the cool-
er months, and a consistent affinity for sunshine
during March to July, i.e., when it was strong
enough to have a distinct warming effect. They
evidently were not influenced by “trace”
amounts of precipitation, but greater amounts
were clearly avoided. Precipitation of 0.5 mm/hr
or more occurred about three times more often
when they were in the water than when they
were out. Neither animal was out of the water
in precipitation greater than 0.5 mm/hr except
in June, one of the warmest months.

Mean air temperatures when the animals were
“out” were at no time greatly different from
those when they were “in” the water. The
slightly lower temperatures when they were out
in January-February apparently were related to
the greater sky cover (= less insolation); the
higher temperatures when they were out in
March-April apparently were related to the less-
er sky cover (= greater insolation). From May
to July, air temperatures were not correlated
with the amount of insolation, and in those
months both animals continued to haul out in
the sunnier weather but in slightly cooler air
than when they were in the water. During sunny
days in summer, when air temperatures rose to

25 °C or more, they spent most of their time in
the water, rarely hauling out for more than an
hour at a time. They tended to haul out mostly
at night in the warmest months. Though they
apparently became acclimatized to a certain de-
gree to the temperate climate, they consistently
avoided exposure to the greatest solar and at-
mospheric heat by escaping to the water.

Activity in relation to weather: wild walruses.
Our data from walruses sighted in the field are
less extensive than those from the captives, but
they also suggest an avoidance of high wind
in cold weather, irrespective of the sky cover
(Table IV). In January, all the animals sighted
when winds were 4 mps (10 mph) or stronger
were in the water, whereas most of those sighted
during lower wind velocities were on the ice.

Winds up to 9 mps (20 mph) seemed not to
deter the animals from hauling out in air at
—7 to 6°C in May, even under overcast skies;
indeed, in that month we saw more walruses on
the ice in windy weather than when it was calm.
In June at Round Island, also, most of the ani-
mals resting on a windward beach during a
squall with 10 to 12 mps winds and rain (air
12°C) were little affected. Though they were
appreciably more restless than they had been
earlier in the day in more moderate weather,
they showed no clear signs of withdrawing into
the sea ( 10°C). The highest rate of emigration
from the beaches of Round Island occurred dur-
ing the warmest afternoon (air 14°C, wind 0
to 2 mps, sky clear), and we interpreted this as
an indication of intolerance of excessive heat,
mostly from intense solar radiation.

Table IV.

Comparative Weather when Free-living Walruses Were Sighted out vs. in the Water
Near St. Lawrence Island, Bering Sea.

Month

Animals Sighted
No. % Out

Air

Temp.

(°C)

Sky

Cover

Occurrence
of Precip.

Wind

Velocity

(mps)

January

10

0

-27

clear

none

11-13

January

25

0

- 8

clear

none

9

January

11

0

-26

clear

none

4

January

3

33

-23

clear

none

0-2

January

25

100

-19

overcast

none

0-2

January

6

100

-12

overcast

none

0-2

January

25

100

- 3

overcast

snow

0-2

March

5

0

- 1

clear

fog

0-2

May

. . 310

97

2,3

clear

none

4-7

May

. . 418

87

1,2

clear

none

2-4

May

. . 96

44

—2 to 2

clear

none

0-2

May

58

98

-7 to 2

overcast

none

4-9

May

58

95

—2 to 6

overcast

none

2-4

May

. . 261

90

-5 to 3

overcast

none

0-2

June

3

0

4

clear

none

0-2

1968] Fay & Ray: Influence of Climate on Distribution of Walruses. I. Thermoregulatory Behavior 9

Discussion

The physical environment of pinnipeds com-
prises parts of both the hydrosphere and the
atmosphere, with, in some cases, nearly equal
amounts of time spent in each. The aquatic por-
tion is usually the more uniformly cold and
stable; the atmospheric portion is relatively un-
stable and heterogeneous and, at times, can be
either colder or warmer than the sea. The home-
otherm that inhabits both must possess unusual
thermoregulatory versatility. Reports from many
sources make it clear that this requirement is
met in pinnipeds not only by physiological
means but by extensive behavioral adjustments
as well. For example, in cold or stormy weather,
northern fur seals, Callorhinus, huddle together
or withdraw into the sea (Bartholomew & Wilke,
1956; Fay, unpublished); Weddell seals, Lepto-
nychotes, seek sunshine and shelter from the
wind (Smith, 1965; Ray & Smith, 1968); gray
seals, Halichoerus, avoid snow-covered surfaces
(Waters, 1965), and Steller sea lions, Eume-
topias, remain in the water (Kenyon & Rice,
1961). In warm weather, fur seals, Callorhinus
and Arctocephalus, seek shade and moisture,
expose areas of bare skin, and fan with their
flippers (Bartholomew & Wilke, 1956; Paulian,
1964); monk seals, Monachus, make wallows
in the damp sand or lie in the shade of a bush
or cave (Kenyon & Rice, 1959; van Wijngaar-
den, 1962) ; elephant seals, Mirounga, and South
American sea lions, Otaria, may escape the heat
altogether by staying in the water (Laws, 1956;
Vaz-Ferreira & Palerm, 1961). Some of these
tactics may considerably extend the thermal
comfort zone well beyond the capacity of phy-
siological mechanisms alone; others, such as es-
cape into the water, indicate that the limits
of the comfort zone have been exceeded.

We assume that adult walruses, like other
polar pinnipeds (Irving & Hart, 1957; Davidov
& Makarova, 1964), are fully adapted for ther-
moneutral existence in icewater, even while at
rest, for they spend the greater part of their life
there and may remain immersed for several days
or weeks at a time. They are capable of sleeping
in the water and frequently do so, yet at certain
times they seem more inclined to rest in air than
in the ostensible comfort of the sea. While out
of the water or in anticipation of hauling out,
they are notably selective of weather conditions,
generally seeking exposure to sunshine and
avoiding exposure to high winds and precipita-
tion. In addition to their selection of the more
favorable thermal conditions, usually warmer
than the sea, they employ heat-conserving be-
havior in all but the warmest weather. The usual
result is a relatively high, stable temperature

in the skin and appendages (Ray & Fay, 1968),
and we think that this is the principal benefit
derived from hauling out. The tissue most affect-
ed by it is the epidermis, the outermost layer of
the skin. Whereas, it is about as cold as the water
during immersion, it can become 30°C warmer
following emergence. Since epidermal mitosis
in pinnipeds probably occurs only at relatively
high tissue temperatures (Feltz & Fay, 1967)
and, perhaps, only when the animals are inac-
tive or asleep (Bullough, 1962; Bullough & Ry-
tomaa, 1965), growth and regeneration of the
skin, as in the molt and healing of wounds, may
be feasible only when the animals are at rest out
of the water. This is not a new theory (Laws,
1956; McLaren, 1958), but it is presented here
in a new context, with new support. We feel that
it could help to explain the conservative, thermo-
philic behavior of walruses and other polar pin-
nipeds when in air, in contrast to their apparent
comfort in the usually colder sea.

The principal behavioral adjustments of wal-
ruses that favor the conservation of body heat
when at rest in air are huddling, fetal posture,
and basking in the sunshine. Huddling may have
special significance for the calves, which possess
less than half as much physical insulation (hair
and blubber) as other arctic pinnipeds of com-
parable size. For the first two or three months
after birth, thermal compensation for their de-
ficiencies seems to be derived principally from
contact with their mother ( “brooding” ) . We con-
sider brooding by walruses as the behavioral
analogue of the woolly coat of young phocid
seals, in that it provides warmth and insulation
for the young animals while their blubber layer
is developing (Davydov & Makarova, 1964; Ray
& Smith, 1968). The young walrus also remains
mostly in the atmosphere during this critical
period, and the mother remains there with it.
On many occasions, we observed that the cows
with very young calves were extremely hesitant
to take the calves with them into the cold water
when threatened by hunters, whereas those with
older calves showed virtually no hesitancy at all.

The huddling of adult walruses at rest has
been recognized in a general way for a long
time, but its potential contribution to thermal
economy evidently has not been considered be-
fore. Among the Pinnipedia other than walruses,
gregariousness is common during the pupping,
mating, and molting periods, but huddling is un-
common. It seems significant that the walrus,
the most polar of the otarioid seals and the most
sparsely haired of all the pinnipeds, is also the
most thigmothermal. By extensive mutual re-
duction of surfaces exposed to the cold air and
substrate, the huddling walrus herd becomes a

10

Zoologica: New York Zoological Society

[53: 1

heat-exchanging and heat-conserving unit with
an advantage for arctic living. However, the per-
sistence of thigmotactic behavior under all ther-
mal conditions may place a limit on the amount
of climatic heat that can be tolerated. Huddling
is disadvantageous in warm weather, for it se-
verely obstructs the dissipation of heat from
the body.

By changing posture, the walrus is capable
of regulating the amount of exposed surface and,
thereby, of controlling the rate of heat loss. In
contrast to huddling, posture is adjusted accord-
ing to the ambient thermal conditions. The fetal
posture ( minimal exposure of surface) is clearly
a reaction to cold that has potential value for
conservation of body heat, whereas sprawling,
with extension of the appendages, undoubtedly
helps to accelerate cooling by exposing the great-
est surface area for dissipation of heat. When
very warm, walruses increase the convectional
heat loss from their body by fanning, usually
with one or both of the foreflippers. These are
small relative to body size but have a large sur-
face-to-volume ratio and can accommodate a
large volume of blood probably at a high flow
rate. The capacity of the hind flippers for trans-
ferring heat to the environment is enhanced also
by evaporation when they become wetted by the
animal’s watery excrement. We frequently no-
ticed also that the fore and hind flippers were
damp in the absence of any extrinsic supply of
moisture, but we were unable to determine the
source of the dampness. Sweat glands were not
found in any tissues from the bare flippers,
though they were abundant in skin from the
hairy parts of the body (Fay, unpublished).

Basking is another effective means for con-
serving body heat, largely by acquiring heat
from solar radiation. We assume that the dark
surface of the walrus’ body absorbs radiant heat
about as well as a black body, and that the short
hair serves to retain it somewhat better than a
bare surface. The hair may function also as a
baffle, protecting against excessive convectional
heat loss in all but the windiest weather.

Walruses in captivity at mid-latitudes showed
an affinity for sunshine from March to July
but tended to avoid prolonged exposure during
the warmest months, when the insolation was
about twice as strong as that in their native habi-
tat. The calves were more inclined to expose
themselves to it than were the juveniles, per-
haps because of their smaller size and less effec-
tive thermoregulatory system. However, we
observed, as did Reventlow (1951), that their

greater exposure seemed to be the cause of a
granular condition of the skin, tentatively iden-
tified as solar keratosis (cf. Mackie &
Mackie, 1963). This occurred during the molt,
in June, after the hair was shed and the skin
was unprotected from the direct rays of the sun.
In our animals, an acne-like condition often oc-
curred with it, possibly due to blockage of the
sebaceous ducts by an excess of keratin (van
Scott, 1959).

The ultimate behavioral response to thermal
conditions of the atmosphere is withdrawal into
the water (“escape”). With increasingly cold
weather, escape is preceded by the extreme
fetal posture and intense shivering; with increas-
ing warmth, it is the normal successor to sprawl-
ing and fanning. We believe that the range of
conditions under which escape does not occur
includes but slightly exceeds the “comfort zone.”
That is, we think that escape is not induced until
the animals become uncomfortably hot or cold.
Walruses that were acclimated for a year or
more to the temperate climate of New York
showed the escape reaction mostly when air
temperatures were lower than 0°C or higher than
25 °C, given sunshine, light winds, and a wet
concrete substrate on which to rest. Newly cap-
tured calves, on a dry, wooden substrate in
shade, were comfortable in still air only at tem-
peratures between 5 and 18°C, and even after
acclimation to warm weather in New York for
2 weeks, they tended not to lie out in air warmer
than 20°C. Wild adults on ice may occasionally
tolerate air as cold as — 35°C with strong winds
(Freuchen, 1935), but in our opinion, this is
more the exception than the rule. Whereas the
majority of those seen by us were on ice when
air temperatures were higher than — 20°C with
little or no wind, nearly all were in the water
when the air was colder or the winds were
stronger. The upper threshold of air tempera-
ture that induces wild adults to escape or remain
in the water seems to be between 10 and 15°C,
given sunshine, light winds, and a damp, rocky
substrate or ice.

Summary and Conclusions

1. The influence of climate on the distribu-
tion of walruses was investigated by observing
their behavioral reactions to weather in the
natural arctic and subarctic environment and in
the temperate climate at the New York Aquar-
ium. Walruses spend a large proportion of their
time out of the water and are, therefore, exposed
to conditions of the atmosphere nearly as often

1968] Fay & Ray: Influence of Climate on Distribution of Walruses. I. Thermoregulatory Behavior 11

as to those of the hydrosphere. Whereas they
can sleep in the water in apparent comfort, they
usually haul out on ice or land to sleep, espe-
cially during the spring and summer.

2. When at rest out of the water, they are
highly gregarious and tend to huddle together
at all times. This mutual reduction of exposed
surface is advantageous for conservation of heat
in cold weather, but it is a deterrent to their haul-
ing out or remaining out when the weather is
warm. Exposure of surface area is regulated
also by the sleeping posture, which is adjusted
for minimal exposure in cold and maximal ex-
posure in warm weather.

3. Walruses are most active in the water at
night and generally haul out to rest in the day-
time. In doing so, they usually seek exposure to
sunshine and avoid high winds and heavy pre-
cipitation. Their tolerance of wind and pre-
cipitation increases with rising air temperatures
and increasing insolation, while their affinity for
sunshine seems to remain unchanged. However,
they evidently cannot tolerate for long the in-
tense solar radiation in summer at mid-latitudes,
and the young may be adversely affected by it.

4. The principal benefit derived from their
hauling out to sleep seems to be the warming
of their peripheral tissues, which may require
heat and physical inactivity to fulfill their growth
and reparative functions. Sustained warmth of
the skin and appendages may be especially im-
portant for the molt, healing of wounds, and the
development and survival of the newborn young.

5. When the weather is excessively cold or hot,
the animals withdraw into the relative com-
fort of the sea after brief exposure or refrain
from hauling out altogether. Thus, the upper
and lower limits of their thermal tolerance are
recognizable from this escape reaction. These
limits may be expected to vary seasonally and,
perhaps, with age, sex, reproductive status,
health, and individuality.

6. The average limits of thermal tolerance of
adult Pacific walruses while at rest in air seem
to be between —20 and +15°C, given light
winds, moderate insolation, and a cool, damp
substrate on which to lie. Colder and warmer
conditions may occasionally be tolerated, but
only for short periods.

7. The highest air temperatures and most in-
tense insolation received in coastal areas at the

southern edge of the Pacific walrus’ present
range tend to induce the escape reaction. Warm-
er conditions, such as are found farther to the
south, could be expected to discourage them
to a greater extent from hauling out during the
day in the spring and summer months, when
they would ordinarily spend the most time out
of the water. We feel that the conflict with their
normal feeding, molting, and calving schedules
could be sufficient to deter them from extending
their range southward under present climatic
conditions.

Acknowledgments

The bulk of the data on which this study is
based was gathered in the course of zoonotic
disease investigations for the Arctic Health Re-
search Center (Fay) and in collecting and cura-
torial activities for the New York Aquarium of
the New York Zoological Society (Ray). Field
work was supported in part also by grants from
the Arctic Institute of North America, under
contractual agreements with the Office of Naval
Research. Transportation during some portions
of Fay's field work was provided by the Bureau
of Commercial Fisheries and the Bureau of
Sport Fisheries and Wildlife, U. S. Fish and
Wildlife Service.

We were assisted in the field by James W.
Brooks of the Alaska Department of Fish and
Game, Karl W. Kenyon of the Bureau of Sport
Fisheries and Wildlife, Charles Young of the
New York Aquarium, K. Richard Zinsmann of
the Arctic Health Research Center, and Steven
A ningayou, Winfred James, LawrenceKulukhon,
and Vernon K. Slwooko of Gambell, Alaska.
Facilities for the temporary holding of captive
calves were provided in Seattle by Edward John-
son of the Woodland Park Zoo. K. W. Kenyon
and Averill Thayer, Bureau of Sports Fisheries
and Wildlife, supplied some unpublished data,
and the keepers of the New York Aquarium
assisted with the acquisition of data from some
of the captive animals. James W. Brooks and
John J. Burns of the Alaska Department of Fish
and Game, M. Woodbridge Williams of the
National Park Service, and K. W. Kenyon pro-
vided some of the photographs. A draft of the
manuscript was reviewed by Drs. Laurence
Irving and L. Keith Miller of the University of
Alaska, Dr. A. W. Mansfield of the Fisheries
Research Board of Canada, Dr. Victor B.
Scheffer of the Bureau of Commercial Fisheries,
and Dr. Robert L. Rausch of the Arctic Health
Research Center. To each of these individuals
and agencies, we express our very sincere
appreciation.

12

Zoologica: New York Zoological Society

[53: 1

Literature Cited

Allen, J. A.

1880. History of North American pinnipeds; a
monograph of the walruses, sea-lions, sea-
bears, and seals of North America. U. S.
Geol. Geog. Surv. Territories, Misc. Publ.
No. 12.

Baer, K. E. Von

1838. Anatomische und zoologische Untersuch-
ungen fiber das Wallross (Trichechus ros-
marus) und Vergleichung dieses Thieres
mit andern See-Saugethieren. Mem. Acad.
Imper. Sciences, St. Petersburg, 1837, ser.
6, 4:97-236.

Bartholomew, G. A., and F. Wilke

1956. Body temperature in the northern fur seal,
Callorhinus ursinus. J. Mamm., 37 (3):
327-337.

Brooks, J. W.

1954. A contribution to the life history and
ecology of the Pacific walrus. Alaska
Cooperative Wildl. Research Unit, Special
Rept. No. 1.

Bullough, W. S.

1962. Growth control in mammalian skin. Na-
ture, 193 (4815): 520-523.

Bullough, W. S., and T. Rytomaa

1965. Mitotic homeostasis. Nature, 205 (4971):
573-578.

Cook, J.

1822. Voyages round the world performed by
Captain James Cook, F.R.S. J. Robins &
Co., London.

Davydov, A. F., and A. R. Makarova

1964. Changes in heat regulation and circulation
in newborn seals on transition to aquatic
form of life. Fiziol. Zhur. SSSR, 50 (7):
894. (Transl. Suppl., Federation Proc., 24
(4): T563-T566.)

Ekman, S.

1953. Zoogeography of the sea. Sidgwick & Jack-
son, London.

Fay, F. H.

1957. History and present status of the Pacific
walrus population. Trans. No. Amer.
Wildl. Conf., 22: 431-443.

Feltz, E. T., and F. H. Fay

1967. Thermal requirements in vitro of epider-
mal cells from seals. Cryobiology, 3 (3):
261-264 (December, 1966).

Freuchen, P.

1935. Mammals. Pt. II. Field notes and biolog-
ical observations. Rept. Fifth Thule
Exped., 2(4-5): 68-278.

Hayes, I. I.

1867. The open polar sea: a narrative of a voy-
age of discovery towards the north pole,
in the schooner “United States.” Hurd &
Houghton, New York.

Irving, L.

1956. Physiological insulation of swine as bare-
skinned mammals. J. Appl. Physiol., 9
(3): 414-420.

Irving, L., and J. S. Hart

1957. The metabolism and insulation of seals
as bare-skinned mammals in cold water.
Can. J. Zool., 35: 497-511.

Irving, L., and J. Krog

1955. Temperature of skin in the Arctic as a
regulator of heat. J. Appl. Physiol., 7 (4) :
355-364.

Kenyon, K. W., and D. W. Rice

1959. Life history of Hawaiian monk seal. Pacific
Sci., 13: 215-252.

1961. Abundance and distribution of the Steller
sea lion. J. Mamm., 42: 223-234.

Laws, R. M.

1956. The elephant seal (Mirounga leonina
Linn.). II. General, social and reproductive
behaviour. Falkland Is. Dependencies Sur-
vey Sci. Repts. No. 13.

Loughrey, A. G.

1959. Preliminary investigation of the Atlantic
walrus, Odobenus rosmarus rosmarus
(Linneaus). Wildl. Mgt. Bull., Ser. 1, No.
14, Canadian Wildl. Service, Ottawa.

Mackie, B. S., and L. E. Mackie

1963. Cancer of the skin. In S. W. Tromp,
et al., Medical biometeorology, pp. 481-
490. Elsevier Publ. Co., Amsterdam, Lon-
don, and New York.

McLaren, I. A.

1958. The biology of the ringed seal (Phoca
hispida Schreber) in the eastern Canadian
Arctic. Fisheries Research Bd., Canada,
Bull. No. 118.

Mansfield, A. W.

1958. The biology of the Atlantic walrus, Odo-
benus rosmarus rosmarus (Linnaeus) in
the eastern Canadian Arctic. Fisheries Re-
search Bd., Canada, Ms. Rept. Ser. (biol.).
No. 653.

Nikulin, P. G.

1947. [Biological characteristics of the shore
aggregations of the walrus in the Chukotka
Peninsula.] lzvestiia, T.N.I.R.O., Vladi-
vostok, 25: 226-228. (Fisheries Res. Bd.,
Canada, Transl. Ser., No. 115.)

1968] Fay & Ray: Influence of Climate on Distribution of Walruses. I. Thermoregulatory Behavior 13

Paulian, P.

1964. Contribution a l’etude de l’otarie de l’ile
Amsterdam. Mammalia, 28 (suppl. 1): 1-
146.

Ray, Carleton, and F. H. Fay

1968. The influence of climate on the distribu-
tion of walruses, Odobenus rosmarus (Lin-
naeus). II. Evidence from physiological
characteristics. Zoologica, 53 (1): 19-32.

Ray, Carleton, and M. S. R. Smith

1968. Thermoregulation in the Weddell seal.
Zoologica, 53 (1): 33-48.

Reventlow, A.

1951. Observations on the walrus (Odobenus
rosmarus) in captivity. Der Zoologische
Garten (NF), 18 (5/6): 227-234.

Scott, E. J. Van

1959. Significance of changes in pilosebaceous
units in acne and other diseases. In, S.
Rothman, (Ed.), The human integument,
normal and abnormal, pp. 113-126. Amer.
Assoc. Adv. Sci., Washington.

Smith, M. S. R.

1965. Movements of the Weddell seal in

McMurdo Sound, Antarctica. J. Wildl.
Mgt., 29: 464-470.

Vaz-Ferreira, R., and E. Palerm

1961. Efectos de los cambios meteorologicos
sobre agrupaciones terrestres de Pinni-
pedios. Rev. Facul. Humanidades & Cien-
cias, Univ. de la Republica (Montevideo),
19: 281-293.

Vibe, C.

1950. The marine mammals and the marine
fauna in the Thule District (northwest
Greenland) with observations on ice con-
ditions in 1939-41. Medd. om Gr0nl., 150
(6): 1-115.

Waters, W. E.

1965. Grey seal haulouts at St. Kilda. Proc. Zool.
Soc. London, 145: 158-160.

WlJNGA ARDEN, A. VAN

1962. The Mediterranean monk seal. Oryx, 6
(5): 270-273.

Zenkevitch, L.

1963. Biology of the seas of the U.S.S.R. Inter-
science (John Wiley & Sons, Inc.), New
York.

14

Zoologica: New York Zoological Society

[53: 1: 1968]

Fig. 1.

Fig. 2.

Fig. 3.
Fig. 4,

Fig. 5.

EXPLANATION OF THE PLATES

Plate I

Herd of adult female walruses resting on
an ice floe off Cape Lisburne, Alaska, sum- pIG
mer, 1937. Photo by M. Woodbridge Wil-
liams.

Herd of male walruses resting on Round
Island, Bristol Bay, Alaska, June 27, 1958.

Photo by Karl W. Kenyon.

Plate II

Walrus calf resting in “fetal” position
with near minimum exposure of body sur-
face. New York Aquarium, June 18, 1961.

Same calf, a few minutes later, in sprawl-
ing position with near maximum exposure
of body surface and appendages. Note
huddling calves in background.

Plate III

Adult female walrus sleeping in the “semi-
fetal” position. St. Lawrence Island,
Alaska, May 16, 1959. Air temperature at

the time was 3.5°C, the wind about 7 mps,
and the sky was clear.

6. A group of male walruses that had recently
emerged from the water. Round Island,
June 27, 1958. Note fetal posture of the
animal at center. Air temperature 13°C,
wind 1 mps, sunny with a high, thin over-
cast. Photo by Karl W. Kenyon.

Plate IV

Fig. 7. Herd of male walruses resting in the after-
noon sun. Round Island, June 24, 1958.
Note the animal sprawled on the rock at
center and the abundance of outstretched
flippers, many of them fanning (arrows).
Air temperature 14°C, wind about 1 mps,
sky clear. Photo by James W. Brooks.

Fig. 8. Two male walruses resting in the sprawl-
ing posture with flippers extended and
spread, Bering Strait, May 18, 1963. The
animal at left was fanning. Air tempera-
ture 8.5 °C, wind calm, sky clear. Photo by
John J. Burns.

FAY ft RAY

PLATE 1

FIG 1

FIG. 2

THE INFLUENCE OF CLIMATE ON THE DISTRIBUTION OF WALRUSES, ODOBENUS
ROSMARUS (LINNAEUS). I. EVIDENCE FROM THERMOREGULATORY BEHAVIOR.

FAY & RAY

PLATE II

FIG. 3

FIG. 4

THE INFLUENCE OF CLIMATE ON THE DISTRIBUTION OF WALRUSES, ODOBENUS
ROSMARUS (LINNAEUS). I. EVIDENCE FROM THERMOREGULATORY BEHAVIOR.

FAY a RAY

PLATE III

FIG. 5

FIG. 6

THE INFLUENCE OF CLIMATE ON THE DISTRIBUTION OF WALRUSES, ODOBENUS
ROSMARUS (LINNAEUS). I. EVIDENCE FROM THERMOREGULATORY BEHAVIOR

FAY & RAY

PLATE IV

FIG. 7

THE INFLUENCE OF CLIMATE ON THE DISTRIBUTION OF WALRUSES, ODOBENUS
ROSMARUS (LINNAEUS). I. EVIDENCE FROM THERMOREGULATORY BEHAVIOR.

2

Influence of Climate on the Distribution of Walruses, Odobenus ros-
marus (Linnaeus). II. Evidence from Physiological Characteristics.

Carleton Ray1 and Francis H. Fay2
(Text-figures 1-9)

Introduction

The principal objective of this and the fore-
going study (Fay & Ray, 1968) was to
test the theory first expressed by von Baer
( 1838) that walruses are prevented by warmer
climates from extending their range southward.
The rationale was that walruses are closely
adapted to the environment they occupy; that
they are highly mobile and could readily move
farther southward; that no physiographic bar-
riers prevent them from doing so, therefore they
must be inhibited by nonadaptation to some
physical or biotic factors of the more southerly
environments. Upon comparing the principal
physical and biotic characters of their range with
those of areas immediately to the south, we con-
cluded that climate was the most probable re-
strictive factor. We and others have observed
that, under certain conditions, walruses and
some other pinnipeds seem to be sensitive to
extremes of atmospheric and solar heat.

The influence of climate on the distribution of
animals has usually been evaluated indirectly by
correlating climatological data with the altitudi-
nal or latitudinal range of species (see review by
Allee, et al., 1949) . We have taken a more direct

iDepartment of Pathobiology, School of Hygiene
and Public Health, The Johns Hopkins University, Bal-
timore, Maryland. Formerly, Curator, New York Aquar-
ium, New York Zoological Society, Coney Island, New
York.

2Biologist, Arctic Health Research Laboratory, Public
Health Service, U.S. Department of Health, Education,
and Welfare, College, Alaska.

course by examining the specific reactions of in-
dividual animals to climatic and microclimatic
conditions. Walruses are adapted to the water of
the polar sea, which in terms of cooling power is
one of the coldest environments on earth. That
they are highly efficient at conservation of body
heat is indicated by their ability to sleep in the
cold water. However, they also spend a large
part of their time out of the water, where they
are exposed to a greater variety of thermal con-
ditions, the warmest and coldest of which evoke
behavioral signs of thermal stress, considered in
our previous paper. In this paper we report some
physiological responses associated with the ob-
served behavior.

The physiological thermoregulation of pin-
nipeds has been studied in recent years principal-
ly by Irving and associates (Irving, et al., 1935;
Scholander, 1940; Scholander, et al., 1950a,
1950b; Irving & Hart, 1957; Hart & Irving, 1959;
Irving, et al., 1962), largely using restrained
animals to facilitate measurement of physiologi-
cal characteristics. These animals were exposed
mostly to controlled environments, adequately
described by ambient temperatures alone. The
results indicated the animals’ capability for main-
taining a constant internal temperature exclu-
sively by physiological means, but they did not
describe the total thermoregulatory responses of
the free-living animal nor take into considera-
tion the complex thermal conditions of the na-
tural environment.

In this study, we worked mostly with animals
that were free to react to weather in their natural
environment or in their quarters at the New York

19

20

Zoologica: New York Zoological Society

[53: 1

Aquarium. In doing so, we hoped to gain insight
into the interrelationships of behavior and phy-
siology in thermoregulation, particularly as they
interact near the upper limits of the “tolerance
zone.” This zone is defined as comprising the
range of weather conditions tolerated by pin-
nipeds while at rest in air. When its limits are
exceeded, the animals withdraw to the water
(Fay & Ray, 1968) .

Materials and Methods

The wild animals studied were 14 adult and
subadult Pacific walruses, O. r. divergens (500
to 1560 kg), in the vicinity of St. Lawrence
Island, Bering Sea, and Round Island, Bristol
Bay, Alaska. Data on captive animals were ob-
tained from a juvenile male Atlantic walrus, O.
r. rosmarus (age four to five years, weight 500
to 600 kg), and a young female Pacific walrus
(age one month to two years, weight 54 to 254
kg) at the New York Aquarium. Data from
calves were obtained from 1 1 newly captured
animals (ages one day to two months, weights 59
to 73 kg) at Gambell, Alaska, at the Woodland
Park Zoo in Seattle, and at the New York Aquar-
ium, Coney Island, New York.

All data from wild adults and subadults were
obtained from animals that were at rest when
killed by rifle, mostly during the course of an
Eskimo hunt. Data from juveniles and calves
were obtained on living animals while they were
at rest and free to respond behaviorally to the
ambient conditions. Rectal temperatures were
taken at depths of 15 to 20 cm with mercury
rectal thermometers or Weston dial thermom-
eters that had been standardized at 37°C. All
other temperatures were taken with the latter.
Subsequent checks against telethermometers
(YSI) showed close agreement with Weston
thermometers, the chief difference being time to
reach equilibrium. Body skin temperatures were

measured on the lateral, ventral, or dorsal as-
pect, whichever was driest and most distant from
the substrate. Flipper temperatures were taken
on the webbing of the rear flippers about midway
between the tarsals and the tip.

Breathing and heart rates were counted on
resting animals, mostly coincident with measure-
ments of skin temperature. Breathing was
observable by nostril action, sound, or chest infla-
tion. Heart action was detected visually or by
axillary palpation.

Data from wet and dry animals were treated
separately. In all cases, ambient weather was
recorded at the site and level of the animal and
at the same time as physiological or behavioral
observations.

Results

Internal Temperatures

The mean body core temperature (thoracic
cavity) of ten subadult and adult walruses in an
ambient temperature range of —1 to 14°C was
36.6°C, or about 1.2°C lower than the mean for
terrestrial mammals of a similar size (Morrison
& Ryser, 1952). Rectal temperatures of the same
walruses were nearly identical to the core temp-
eratures (Table I). The rectal temperatures of
calves were significantly higher than those of the
subadults and adults, averaging 37.5 °C. Some of
the highest temperatures were measured in teeth-
ing calves in temperate climates, and it is con-
ceivable that temperatures were elevated as a
result of that condition. We also suggest that the
warmer conditions of captivity induced higher
rectal temperatures, but our data from calves in
cold conditions are too few to show this. A
diurnal fluctuation of body temperature also was
suggested. The mean rectal temperature of five
calves in midmorning was 38.2 ± 0.29°C, where-
as in the same calves in late afternoon it was
37.8 ± 0.18°C.

Table I

Internal Temperatures of Resting
Walruses at Air Temperatures of —1 to 25°C

Age of
Animals

Thoracic

No, Temperature °C

Observations Range Mean ± S.E.m

Rectal

Temperature °C
Range Mean ± S.E.m

1-6 months

36

35.3-39.0

37.5 ± 0.13

1

37.0

1-3 years

la

36.2

> 5 years

10

34.0-38.0

36.2 ± 0.42

10

34.0-38.0

36.6 ± 0.32

aData from Rausch, unpublished.

BODY SKIN °C BODY SKIN °C

1968] Ray & Fay: Influences of Climate on Distribution of Walruses. II. Physiological Characteristics 21

Temperature of the Skin When Wet
The body surface and flippers of moderately
active animals during immersion were 1 to 3°C
warmer than the water (Text-fig. 1 A) . This find-
ing is in accord with those of other workers
(Irving & Hart, 1957; Hart & Irving, 1959; Ray
& Smith, 1968) and indicates that, whether re-
strained or not, pinnipeds are usually about as
cool on the body surface as the water in which
they are immersed.

For at least five to 10 minutes after emerging

WATER °C

from the water, the walruses remained cool on
the body surface, usually within 8°C of the water
temperature, provided that the water was cooler
than 10°C and the air cooler than 20°C. The
body surface warmed more rapidly when air and
water temperatures were high, and the tempera-
ture of the flippers rose faster and more errat-
ically than did that of the body skin (Text-fig.
IB). Within an hour after emergence, the skin
of the body and flippers was usually dry and had
attained a relatively stable temperature, most

Text-fig. 1. Temperatures on the skin of the body and rear flipper web of walruses: (A) in water

— • — and immediately after emergence into air O . (B) in air within 10 minutes O

and within 20 minutes — A — after emergence.

22

Zoologica: New York Zoological Society

[53: 1

40-

emerge

almost dry

completely dry

35-

.A

flipper web

rr

3 30-
I-
<
oc.

lii

a.

s

UJ

-AIR #C

r

eoo^

25-

/

-WATER °C

20-

10

20 30 40

TIME MINUTES

50

60

Text-fig. 2. Example of the rate of warming of the body skin and rear flipper web of a captive
juvenile female walrus from the time of emergence into air until the skin was dry.

Text-fig. 3. Temperatures of the dry skin of the body and rear flipper web of newly captured calves
and wild adult walruses at rest in air. • A = newly captured calves. O A = wild adults. Curves de-
limit the usual upper limits of temperature.

BODY SKIN

1968] Ray & Fay: Influences of Climate on Distribution of Walruses. II. Physiological Characteristics 23

often higher than that of the air. The role of the
flippers in the dissipation of heat is indicated by
the high temperatures attained by them, espe-
cially in air warmer than body skin tempera-
ture. (Text-fig.2) .

Temperature of the Skin When Dry
At air temperatures from 0°C, to about 15°C,
the temperature of the dry skin on the body of
wild calves and adults was usually higher than
20°C but rarely higher than 32°C (Text-fig. 3).
The latter seemed to be an upper “limit,” beyond
which the skin temperature ordinarily did not
rise. Thus, the skin/air temperature curve of

wild walruses flattens out to a “plateau” in the
range from near freezing to about 15°C; at high-
er air temperatures, the curve again rises steeply.
This plateau was evident also in the captive wal-
ruses at the New York Aquarium, within a
higher range of ambient temperatures (Text-figs.
4-5) and has not been noticed previously in pin-
nipeds, perhaps because it does not occur in
restrained animals or occurs in them over a
narrower range of ambient temperatures (cf.
Irving & Hart, 1957) . It was detected recently in
unrestrained Weddell seals (Ray & Smith, 1968),
and we have also observed it in other unre-
strained pinnipeds at rest (unpublished). Eleva-

Text-fig. 4. Temperatures of the dry skin of the body and rear flipper web of a juvenile female
walrus at the New York Aquarium. Curves delimit the usual upper limits of temperature.

Text-fig. 5. Temperatures of the dry skin of the body and rear flipper web of a juvenile male walrus
at the New York Aquarium. Curves delimit the usual upper limits of temperature.

24

Zoologica: New York Zoological Society

[53: 1

tion of skin temperature above the 32°C plateau
coincided in calves and adults with the first be-
havioral signs of heat stress (Fay & Ray, 1968).

Under most conditions, the flippers were
somewhat warmer than the surface of the body
and showed an even greater tendency for sus-
tained high temperature over a wide range of air
temperatures (Text-figs., 3-5). In air warmer
than 0°C, the flippers were usually between 30
and 37°C; they were cooler than 25°C only when
wet or in contact with ice or snow. Brief, spon-
taneous fluctuations of 5 to 6°C in flipper tem-
perature were detected at ambient temperatures
lower than 10°C, and fluctuations of 2 or 3°C
were occasionally detected at higher air tem-
peratures.

The temperature of the calves’ flippers rose
above the 37°C level only in air warmer than
about 15°C, i.e., at or about the same ambient
temperatures in which the skin of the body ex-
ceeded the 32°C plateau and signs of heat stress
first appeared. The occurrence of flipper tem-
peratures of 38 and 39°C might have been indic-
ative of rising internal temperature, such as
could occur with increased metabolism or in-
adequate dissipation of heat.

Calves resting in air warmer than 15°C and
adults in air warmer than 10°C showed a dis-
tinct reddish cast on the body and flippers, in
contrast to the normal pallor of animals under
cooler conditions. Pale adults killed when the air
was —20 to 5°C were usually cool to the touch
and their skin and flippers scarcely bled at all
when slashed. Reddish adults killed in air of 13
to 14°C were contrastingly hot to the touch and
bled profusely when slashed. When these hot
animals were thoroughly bled out, their skin be-
came as pale as the cool animals’, demonstrating
that the redness and heat were due to vasodila-
tion and the resulting hyperemia. This hyper-
emic condition has often been called sunburn in
popular literature.

The contrast between hyperemic and ischemic
animals was especially noticeable at Round
Island, where most of the animals were in a
nearly hairless stage of their annual molt. When
in the 10°C water, the lightly pigmented adults
appeared nearly white, whereas they became
reddish after lying out on the beach for an hour
or more. When these reddish animals were
chased back into the water, their skin at once
regained its pallor.

Breathing and Heart Rates

Both breathing and heart rates appear to be
highly variable, even in resting calves. Minimum
breathing rates in calves exposed to stepwise in-
creases of still air temperature in shade declined
from 16 per minute at — 1°C to 4 per minute at

15°C, and rose again to 7 per minute at 18°C
(Text-fig. 6). These rates include brief periods
of apnea, especially prevalent at 10 to 18°C,
and each point in the graph represents the mean
of several counts. We did not notice panting
under any conditions. Minimum heart rates in
the same animals declined from 119 to 52 per
minute in the same temperature range, though
rates as high as 133 per minute were recorded
at about 15°C. Lacking any special equipment
for measuring heart rates, we were usually un-
able to determine them at ambient temperatures
lower than 5°C or even 10°C due to the animals’
frequent violent shivering. Since these were ani-
mals destined for display at the aquarium, we
did not expose them to temperatures higher than
19°C, at which they already appeared distressed.

Temperature Gradients in the Skin and Blubber

Temperatures of the tissues were measured at
several depths up to 20 cm in a few adults im-
mediately after they were killed. Some of these
animals were dry and on land or ice; others were
wet and on ice. Relatively steep temperature
gradients were indicated in most cases, and these
were taken up mainly in the skin and blubber
(Text-fig. 7) . Their lengths, taking the inner end
point to be 0.5 °C lower than the deep thoracic
temperature (Irving & Hart, 1957), ranged from
near 0 cm to about 15 cm. Compared with the
gradients measured in smaller pinnipeds (Hart
& Irving, 1959), these were much longer and
were not correlated in the same way with the
internal temperature-skin temperature differ-
ence. We assume that these dissimilarities were
due to the greater thickness of the walrus’ skin
and blubber, which should be expected to ac-
commodate a longer gradient if the tissues are
effective as insulation.

The shortest gradient amounted to virtually
no gradient at all. This was measured in an adult
male lying on a rocky beach in sunshine when
the air was 14°C (Text-fig. 71). The surface and
cutaneous tissues of this specimen were about
as warm as the interior of its body. Under similar
conditions but without sunshine, two other adults
had surface temperatures 5 to 7°C lower than
that of the body core (Text-figs. 7G, H). We at-
tribute the greater superficial warmth of the
first animal to solar radiation. For example, a
mercury thermometer and a Weston dial ther-
mometer exposed to the sun registered 22 and
34°C, respectively, and the surface of a walrus
cadaver nearby was 40.5 °C. Thus, the length
of gradient is not purely a function of air
temperature.

We were not equipped to measure tempera-
ture gradients in the living calves and had no
opportunities otherwise to determine the form of

1968] Ray & Fay: Influences of Climate on Distribution of Walruses. II. Physiological Characteristics 25

Text-fig. 6. Heart and breathing rates of newly captured walrus calves with relation to ambient
still air temperature in shade.

such gradients. Although the calves were about
twice as large as young harbor seals, such as
those used by Irving and Hart ( 1957) , the thick-
ness of their insulation (skin and blubber) was
about the same. Therefore, we assume that the
temperature gradients in their tissues were com-
parable to those of the seals, since gradient
length seems to be as much a function of the
thickness of the insulation as of the internal tem-
perature-skin temperature difference.

Discussion

Correlation of Thermoregulatory Behavior
and Physiology

The walrus is a homeothermic mammal with
an internal temperature of about 36.6°C, prob-
ably intermediate between those of other small-

er and larger pinnipeds (Bartholomew, 1954;
Bartholomew & Wilke, 1956). We assume that
this temperature is maintained in a balance be-
tween heat production and heat loss and that the
production of metabolic heat is usually at the
basal level in resting walruses, whether in or out
of the water. The loss of heat is closely regulated
by vasomotor and behavioral means, within the
limits imposed by the animal’s surface-to-volume
ratio and amount of physical insulation (hair,
skin, and blubber). We have described in the
previous paper the behavioral adjustments that
walruses make in response to their thermal en-
vironment, and we report here some indices of
their physiological adjustments. Most of our in-
formation was obtained from very young ani-
mals, up to one or two months old. From their
behavior alone, it was clear that these infants

26

Zoologica: New York Zoological Society

[53: 1

Text-fig. 7. Temperature gradients in the superficial tissues of adult wild walruses within a few minutes
of death. Skin and blubber thicknesses at the site of measurement are drawn to scale (thinner-skinned
animals are females). Air temperatures within 1 meter of the body are indicated by open circles to
the left of each section. A-D were wet and E-I were dry animals.

were not yet fully adapted to even the moderate
cold of the arctic springtime, but that they had
about the same tolerance of heat as the adults.
By correlating their physiology and behavior
(Text-fig. 8), we obtained a useful model with
which the adults’ reactions could be compared.

The calves resting in air at —1 to 3 or 4°C
assumed a tense fetal posture (minimum expo-
sure of surface) and shivered violently (Text-
fig. 8 ) . At the same time, the temperature on the
surface of the almost-dry to dry body ranged
from 7 to 30°C and similarly on the flippers
from 22 to 37°C. Breathing and heart rates were
the highest recorded at any temperature, and an
elevated metabolic rate was also indicated by the
intense shivering. Adults under similar condi-
tions showed comparable skin temperatures but
were clearly more comfortable and relaxed.

We have seen that adult walruses readily ex-
pose themselves to air temperatures as low as
— 20°C with light winds (Fay & Ray, 1968), and
they are known to lie in the open occasionally
in much colder weather (Freuchen. 1935).

The calves continued to shiver intermittently
up to an ambient temperature of 7 or 8°C but

only occasionally at temperatures of 10°C or
more. Their posture became more relaxed, and
there was a significant decline in the breathing
rate. Between 10 and 15°C, they became fully
relaxed and lay on their back or sides with flip-
pers limp and away from the body. Adults under
comparable conditions behaved in the same
way, and their skin temperatures, like the calves’,
were sustained at 25 to 32°C; the flippers of the
calves were sustained at 30 to 37 °C, i.e., near or
at the temperature of the body core.

Breathing rates of the calves declined to the
lowest recorded levels at an air temperature of
about 15°C, and minimum heart rates occurred
at 18 to 20°C. The calves, in general, became
restless at the latter temperatures, sprawling for
maximum exposure of surface and fanning with
the fore flippers. At this point also, skin and
flipper temperatures rose above the previously
sustained levels, indicating a major increase in
heat output and, we think, of heat production.
Ultimately, when air temperatures reached more
than 20°C the calves “escaped" into the water,
where they sometimes resumed their sleep.

The behavioral reactions of adult walruses

1968] Ray & Fay: Influences of Climate on Distribution of Walruses. 11. Physiological Characteristics 27

under comparably warm conditions were virtu-
ally the same as those of the calves. Sprawling
and fanning were observed when they lay out in
air at 8.5°C with intense midday insolation and
no wind (latitude 66°, mid-May), and at 14°C.
with afternoon sun and 2-mps breeze (latitude
58°, mid-June). Conditions warmer than the
latter are uncommon in the walrus’ range but
common to the south of it, where insolation is
more intense and air temperatures generally
higher.

The captive juveniles, with a history of one to
five years’ exposure to the temperate climate of
Coney Island, New York, responded behavior-

ally in a pattern comparable to the calves’. They
did not haul out often in air cooler than 3 or
4°C or warmer than 25°C, and when they did
emerge under these conditions, they stayed for
only a short time, usually returning to the water
before their skin dried off. However, under most
conditions, their skin was cooler than that of the
calves, and in no case did their skin or flipper
temperatures rise above the usually sustained
levels. We assume that these differences reflect
physiological maturation and acclimatization to
temperate climate, with more efficient conserva-
tion of heat in cold weather and better dissipa-
tion of heat when the weather was warm.

Text-fig. 8. Correlation between characteristic physiological and behavioral responses of newly cap-
tured walrus calves and ambient still air temperatures in shade.

28

Zoologica: New York Zoological Society

[53: 1

Limits of Thermoneutrality and
Thermal Tolerance

Without actual measurements of metabolic
rates, we could not determine the exact limits of
thermoneutrality in either the calves or the
adults, but the indirect evidence was strongly
suggestive of some general parameters. From
the calves’ highly variable skin and flipper tem-
peratures, high breathing and heart rates, and,
especially, intensive shivering at still air tem-
peratures of 5°C and lower, we estimate that the
lower limit (= critical temperature) of their
thermoneutral zone was probably about 5°C.
Their shivering alone was indicative that metab-
olism was increased above the basal rate. For
adults, the critical temperature is assumed to be
much lower, but we do not know how low. Judg-
ing from their ostensible comfort at air tem-
peratures near — 20°C with light wind, we are
confident that they can withstand at least that
much cold without elevation of their metabolism.

We estimate that the upper limit of thermo-
neutrality for the nonacclimatized calves was at
or near 18°C in still air and shade. This was
based on analogy with other homeotherms, in
which minimum breathing and heart rates occur
usually at or near the upper limit of thermo-
neutrality (e.g., Bartholomew & Fludson, 1962;
Hudson & Brush, 1962; Hudson, 1965), and
elevated body temperatures are correlated with
increased metabolism (e.g. Graham et al., 1959;
McNab & Morrison, 1963). Elevation of body
temperature in the walrus calves was indicated
by the rise of flipper temperature above 37 °C at
air temperatures over 18°C, for the flippers
could not have become warmer than the normal
body core unless the core temperature itself had
risen. Also possibly indicative were the high
rectal temperatures recorded occasionally from
calves introduced to temperate climate.

The upper limit of thermoneutrality for adults
is unknown but is believed to be similar to that
of the calves. This is suggested principally by
their comparative behavior (Fay & Ray, 1968).
Adults at the southern edge of their geographic
range in summer, in 14°C air with sunshine and
light breeze, showed signs of heat stress (hyper-
emia, sprawling, restlessness, fanning) compar-
able to those in calves in still air and shade at
18°C or more. The added heat from the sun in
that situation was equivalent to at least an addi-
tional 10°C of air temperature and was only
partly counterbalanced by the breeze. From their
behavior, we judged that the majority of these
animals would not have tolerated much warmer
weather. A high proportion of them had already
withdrawn into the sea.

The general scheme, as we envision it, of wal-

ruses’ reactions to the thermal climate in air is
shown in Text-fig. 9. Here, the scale of thermal
conditions (abscissa) is necessarily vague be-
cause individual animals will be influenced not
only by air temperature but by wind, solar radia-
tion, moisture, and the conductivity of the sub-
strate. These combine to present an “effective”
temperature that could not be considered in de-
tail here (cf. Ray & Smith, 1968).

Their reactions may also vary with individual
and seasonal differences in acclimatization, the
thickness of their insulation, their maternal
status, the molt, and their general physical and
nutritional condition. Molting animals seem to
be more sensitive to cold and less sensitive to
heat than any others; cows with newborn calves
and sick or exhausted animals tend to be least
sensitive to either. Tolerance will vary also with
the number of animals in the group. Large herds
can be expected to tolerate more intense cold
than isolated animals because of their huddling
and mutual improvement of the microclimate,
Conversely, huddling groups will be less tolerant
of heat because of their reduced surface area.
Indeed, the gregariousness and persistent hud-
dling of walruses may be among the most influ-
ential factors determining the limits of their
thermal tolerance.

The limits of tolerance are shown as being
somewhat wider than the estimated limits of
thermoneutrality. This is based mainly on the
reactions of the calves, in which escape evidently
was not induced until the metabolism had risen
above the basal level. In the captive juveniles
and the adults, the two zones may have coin-
cided exactly. We believe that wild walruses
usually avoid exposure to conditions outside of
their thermoneutral zone and that the principal
function of the escape reaction is the withdrawal
from unfavorable conditions. Whereas the at-
tainment of thermoneutrality may often be feasi-
ble in air, it probably is always feasible in the
water in healthy animals due to the normal con-
dition of vasoconstriction.

The question of why walruses haul out at all,
if they can sleep in the water with minimum
production of heat, may be answered by con-
sidering the metabolic requirements of their
skin. For conservation of body heat during im-
mersion, the skin is permitted to cool to about
the same temperature as the water and is largely
deprived of blood due to vasoconstriction. The
epidermis, which comprises the outermost layer,
in direct contact with the medium, is most
affected. At the low temperatures normally sus-
tained during immersion, the epidermis is evi-
dently in a semidormant state and is incapable
of performing its growth and reparative func-

1968] Ray & Fay: Influences of Climate on Distribution of Walruses. II. Physiological Characteristics 29

TOLERANCE ZONE

Brooding

ESCAPE

Very Cold

Cold C©@! Warm Very Warm

WEATHER/ CLIMATE

Ho!

Text-fig. 9. Schematic representation of dry skin temperature and its relation to the tolerance zone and
metabolic rate of walruses at rest in air. The critical physiological limits are not known for the walrus,
nor are the exact points of escape with response to extreme low temperature. Thermoneutrality is pre-
dicted within the entire tolerance zone and is considerably extended in calves by maternal brooding
(Fay & Ray, 1968). Acclimatization has the effect of shifting these predicted responses.

tions (Feltz & Fay, 1967). The optimum tem-
perature for epidermal growth seems to be near
30°C, which is attainable in the skin of polar
pinnipeds only when exposed to the air. By
means of behavioral regulation of their surface
area and its exposure to the air, walruses are

capable of sustaining the required skin tempera-
ture under a wide variety of climatic conditions.
This is reflected in the plateauing of the skin/
ambient temperature curve in the upper half or
more of the zone of thermoneutrality.

The highest dry skin temperature on the body

30

Zoologica: New York Zoological Society

[53: 1

during thermoneutrality is about 32°C. When
this temperature is exceeded, walruses become
restless, begin fanning, and ultimately withdraw
into the water. We have seen that this occurs,
mostly in huddling walruses, even in the rela-
tively cool climate at the southern edge of the
walrus’ range, and we believe that warmer cli-
mates would be intolerable for this reason alone.
Provided that these animals would retain their
usual pattern of diurnal rest and nocturnal ac-
tivity, as well as their tbigmotactic and helio-
philic behavior, we are confident that they could
not reside in comfort at lower latitudes in sum-
mer. Walruses exposed to temperate climates
while in captivity successfully avoided hyper-
thermia during the warmest weather by hauling
out only so long as they were cooled by evapora-
tion, or by hauling out at night. (Fay & Ray,
1968). However, because of their diurnal feed-
ing schedule in captivity, they were not as closely
bound to the normal activity rhythm. Some an-
cestral walruses evidently lived in warmer cli-
mates than their modern descendants ( Ray,
1960; Mitchell, 1961, 1962), and the morphol-
ogy of at least one of these suggests that it led
a more pelagic existence.

Summary and Conclusions

1 . The influence of climate on the distribution
of walruses was investigated by measuring some
parameters of physiological thermoregulation in
unrestrained animals at rest, under natural and
controlled conditions at ambient temperatures
from —1 to 25 °C. Data were obtained on the
temperatures of the body core, skin, and hind
flippers and on the breathing and heart rates and
temperature gradients. Some comparative data
't/e re obtained from young walruses that were
reared in captivity in a temperate climate.

2. Internal temperatures of wild adults and
calves were relatively labile, ranging from 34 to
39°C. Rectal temperatures of calves may have
fluctuated in response to several factors, includ-
ing teething, high air temperatures, and time of
day. The mean rectal temperature of ten sub-
adults and adults under natural conditions in —1
to 14°C air was 36.6°C.

3. Skin temperatures on the body during im-
mersion were within 3°C of water temperature
but rose rapidly to higher levels after emergence
and drying. In general, the flippers warmed more
rapidly than the skin of the body and attained
somewhat higher temperatures.

4. The skin and flipper temperatures of calves
in still air and shade did not rise continuously
with increasing ambient temperature, but leveled
off between air temperatures of 0 and 15°C and
then rose again under warmer conditions. The
upper limit of skin and flipper temperatures in
the plateau were about 32 and 37°C, respec-
tively.

5. The breathing rates of calves were highest
at air temperatures near 0°C. They declined to
a minimum at 15°C and rose again at 18°C.
Brief periods of apnea were most common in air
warmer than 10°C. The minimum heart rate oc-
curred at 18°C.

6. Temperature gradients in the skin, blubber,
and outer muscles were about five times longer
than those in young harbor seals under com-
parable conditions. This difference was corre-
lated with the greater thickness of the skin and
blubber in the walruses. Gradient length is as
much a function of the thickness of insulation
as of the internal temperature-skin temperature
difference. It is not a function of air temperature
when the walrus is dry and is affected by other
factors such as insolation.

7. The estimated lower limit of thermoneu-
trality (critical temperature) of the calves in still
air and shade is about 5°C; in adults it is prob-
ably lower than — 20°C. Adults are assumed to
be thermoneutral when at rest in the water.

8. The estimated upper limit of the zone of
thermoneutrality for isolated calves and adults
is about 18°C in still air and shade or its equiva-
lent under natural conditions. Animals in these
or warmer conditions showed elevated skin,
flipper, and body temperatures, as well as cuta-
neous hyperemia, restlessness, and fanning. Ulti-
mately, they avoided further hyperthermia by
withdrawing into the water.

9. The weather in spring and summer at the
southern edge of the walrus’ geographic range
is often warm enough to induce hyperthermia
and withdrawal to the water at a time when
basking may be particularly important, especial-
ly during annual molt. Without physiological
acclimatization and some major alterations of
their more stable behavioral characteristics, such
as diurnal inactivity, heliophilism, and huddling
(which are adaptive for cold climates and not
for warmth), walruses probably would not or
could not occupy areas with warmer weather.

1968] Ray & Fay: Influences of Climate on Distribution of Walruses. II. Physiological Characteristics 3 1

Acknowledgments

The bulk of the data on which this study is
based was gathered while collecting and per-
forming curatorial duties for the New York
Aquarium of the New York Zoological Society
(Ray) and in the course of zoonotic disease in-
vestigations for the Arctic Health Research
Laboratory of the U.S. Department of Health,
Education, and Welfare (Fay). Field work was
also supported in part by grants from the Arctic
Institute of North America, under contractual
agreements with the Office of Naval Research.
Transportation during one part of Fay’s field-
work was provided by the Bureau of Commercial
Fisheries, U.S. Fish and Wildlife Service.

We were assisted in the field by James W.
Brooks of the Alaska Department of Fish and
Game, Karl W. Kenyon of the Bureau of Sport
Fisheries and Wildlife, Charles Young of the
New York Aquarium, K. Richard Zinsmann of
the Arctic Health Research Laboratory, and
Stephen Aningayou, Winfred James, Lawrence
Kulukhon, and Vernon Slwooko of Gambell,
Alaska. We were also assisted at the Woodland
Park Zoo, Seattle, by Edward Johnson. At the
New York Aquarium we were assisted by Head
Keeper Charles Young and his men. Dr. Robert
L. Rausch of the Arctic Health Research Labo-
ratory supplied some unpublished data. A draft
of the manuscript was reviewed by Drs. Lau-
rence Irving and L. Keith Miller of the Uni-
versity of Alaska, Dr. A. W. Mansfield of the
Fisheries Research Board of Canada, Dr. Victor
B. Scheffer of the Bureau of Commercial Fisher-
ies, and Dr. Robert L. Rausch of the Arctic
Health Research Laboratory. To each of these
individuals and agencies, we express our sincere
appreciation.

Literature Cited

Allee, W. C., A. E. Emerson, O. Park, T. Park

and K. P. Schmidt

1949. Principles of animal ecology. W. B. Saun-
ders Company, Philadelphia.

Baer, K. E. von

1838. Anatomische und zoologische Untersuch-
ungen liber das Wallross ( Trichechus ros-
marus) und Vergleichung dieses Thieres
mit andern See-Saiigethieren. Mem. Acad.
Imper. Sciences, St. Petersburg, 1837, Ser.
6, 4:97-236.

Bartholomew, G. A.

1954. Body temperature and respiratory and
heart rates in the northern elephant seal.
J. Mammal., 35(2) : 2 1 1-2 18.

Bartholomew, G. A., and J. W. Hudson

1962. Hibernation, estivation, temperature regu-
lation, evaporative water loss, and heart
rate of the pigmy possum, Cercaertus
nanus. Physiol. Zool., 35(1) : 94- 107.

Bartholomew, G. A., and F. Wilke

1956. Body temperature in the northern fur seal,
Callorhinus ursinus. J. Mammal., 37(3):
327-337.

Fay, F. H., and Carleton Ray

1968. Influence of climate on the distribution of
walruses, Odobenus rosmarus (Linnaeus).
I. Evidence from thermoregulatory be-
havior. Zoologica, 53(1): 1-18.

Feltz, E. T., and F. H. Fay

1967. Thermal requirements in vitro of epi-
dermal cells from seals. Cryobiology,
3(3): 261-264.

Freuchen, P.

1935. Mammals. Pt. II. Field notes and bio-
logical observations. Rept. Fifth Thule
Exped., 2(4-5) :68-278.

Graham, N. Me C., F. W. Wainman, K. L. Blax-
ter and D. G. Armstrong
1959. Environmental temperature, energy me-
tabolism, and heat regulation in sheep. I.
Energy metabolism in closely clipped
sheep. J. Agric. Sci., 52:13-24.

Hart, J. S., and L. Irving

1959. The energetics of harbor seals in air and
in water with special consideration of sea-
sonal changes. Can. J. Zook, 37:447-457.

Hudson, J. W.

1965. Temperature regulation and torpidity in
the pigmy mouse, Baiomys taylori. Phy-
siol. Zook, 38(3) :243-254.

Hudson, J. W., and A. H. Brush

1964. A comparative study of the cardiac and
metabolic performance of the dove, Zen-
aidura macroura, and the quail, Lophortyx
californicus. Comp. Biochem. Physiol.,
12(2): 157-170.

Irving, L„ and J. S. Hart

1957. The metabolism and insulation of seals
as bare-skinned mammals in cold water.
Can. J. Zook, 35:497-511.

Irving, L., L. J. Peyton, C. H. Bahn, and R. S.
Peterson

1962. Regulation of temperature in fur seals.
Physiol. Zook. 35(4) :275-284.

32

Zoologica: New York Zoological Society

[53: 1: 1968]

Irving, L., O. M. Solandt, D. Y. Solandt, and
K. C. Fisher

1935. The respiratory metabolism of the seal
and its adjustment to diving. J. Cell. Comp.
Physiol., 7(1): 137-15 F.

McNab, B. K„ and P. Morrison

1963. Body temperature and metabolism in sub-
species of Peromyscus from arid and
mesic environments. Ecol. Monog., 33(1):
63-82.

Mitchell, E. D., Jr.

1961. A new walrus from the Imperial Pliocene
of southern California: with notes on odo-
benid and otariid humeri. L. A. County
Mus. Contrib. in Science, No. 44.

Mitchell, E. D., Jr.

1962. A walrus and a sea lion from the Pliocene
Purisima formation at Santa Cruz, Cali-
fornia: with remarks on the type locality
and geologic age of the sea lion Dusig-
nathus santacruzensis Kellogg. L. A.
County Mus. Contrib. in Science, No. 56.

Morrison, P. R., and F. A. Ryser

1952. Weight and body temperature in mam-
mals. Science, 1 16(3009) :23 1-232.

Ray, Carleton, and M. S. R. Smith

1968. Thermoregulation of the adult and pup
Weddell seal, Leptonychotes weddelli
( Lesson ) , in Antarctica. Zoologica, 53(1):
33-48.

Ray, Clayton

1960. Trichecodon huxleyi (Mammalia: Odo-
benidae) in the Pleistocene of southeast-
ern United States. Bull. Mus. Compar.
Zool., Harvard, 122(3) : 129-142.

Scholander, P. F.

1940. Experimental investigations on the respi-
ratory function in diving mammals and
birds. Hvalradets Skrifter, No. 22.

Scholander, P. F., R. Hock, V. Walters, and

L. Irving

1950a. Adaptation to cold in Arctic and tropical
mammals and birds in relation to body
temperature, insulation, and basal meta-
bolic rate. Biol. Bull., 99(2) : 259-27 1 .

1950b. Body insulation of some Arctic and tropi-
cal mammals and birds. Biol. Bull., 99(2) :
225-236.

3

Thermoregulation of the Pup and Adult Weddell Seal, Leptonychotes
weddelli (Lesson), in Antarctica.

Carleton Ray1 and M. S. R. Smith2
(Plates I-II; Text-figures 1-8)

Introduction

Some phocid seals are the most polar of
marine mammals and live in the thermally
most difficult of environments for homeo-
therms. They are confronted with extreme prob-
lems of heat conservation in the coldest of seas
and with problems of heat dissipation when
hauled out on the ice to pup, molt, or rest. Their
thermoregulation, therefore, illustrates some im-
portant facets of mammalian adaptation.

The Weddell seal, Leptonychotes weddelli
(Lesson), is typically an inhabitant of the shore
and fast ice of Antarctica, frequenting the most
southerly open water, leads, or access holes. It
probably lives at the lowest mean environmental
temperature of any mammal on a year-around
basis. The water of its environment is almost
always close to the freezing point even in sum-
mer, when air temperatures rarely rise above
5°C. Its biology has been reviewed by Wilson
(1907), Lindsey (1937), Bertram (1940),
Sapin-Jaloustre ( 1952) , Scheffer ( 1958) , Mans-
field (1958), King (1964), and Smith (1965,
1966). Recently, attention has been turned to its
underwater biology, for instance, Littlepage
(1963), Ray and Lavallee (1964), Kooyman

iDepartment of Pathobioiogy, School of Hygiene
and Public Health, The Johns Hopkins University, Bal-
timore, Maryland. Formerly, Curator, New York
Aquarium, New York Zoological Society, Coney Island,
New York.

2 Department of Zoology, University of Southampton,
United Kingdom. Formerly, Zoology Department, Uni-
versity of Canterbury, Christchurch, New Zealand.

(1965), Schevill and Watkins (1965), and Ray
(1965, 1966, 1967 ) . The species is an ideal sub-
ject for field study.

Previous work on the thermoregulation of
pinnipeds, such as the important studies of
Irving and Hart (1957) and Hart and Irving
(1959), has for the most part emphasized phys-
iological aspects in restrained animals: the skin/
ambient temperature regression, the metabolic
rates, the definition of thermoneutrality, and the
establishment of critical temperature limits. Fay
and Ray (1968) and Ray and Fay ( 1 968 ) have
used a somewhat different approach in which
behavioral and psysiological mechanisms were
considered simultaneously in wild and unre-
strained captive walruses. Their remarks and
methods apply here: in sum, that wild animals
rarely expose themselves for prolonged periods
to conditions where critical limits apply; that a
“tolerance zone” in which thermoneutrality is
maintained is more applicable for animals in
nature; that behavior (including “escape”) and
physiology are mutually responsible for the de-
limiting of this zone; and that a study of the
unrestrained animal helps to reveal the relation-
ship between physiology and behavior.

In the present study we report on work done
almost solely on unrestrained Weddell seals in
the field in Antarctica in an effort to delimit the
tolerance zone of the Weddell seal. In addition,
some years ago it was suggested to one of us
(Ray) by L. Irving (pers. comm.) that one
aspect of pinniped thermoregulation had been
ignored: the physiological change that the

33

34

Zoologica: New York Zoological Society

[53: 1

lanugo-clad, almost blubberless, mostly terres-
trial pup undergoes to become a more thinly
haired, thick-blubbered, amphibious postlanugo
animal. Since that time, the work of Davydov
and Makarova (1964) has appeared, in which
pup harp seals, Phoca groenlandica, were sub-
jected to metabolic tests. It was shown that
metabolism decreases directly with the accumula-
tion of subcutaneous fat and increases the toler-
ance of lower water temperatures. In the present
study, we have dealt with the Weddell seal prin-
cipally in air and have utilized almost identical
methods for pups, juveniles, and adults on the
assumption that the changes in the insulative
layers from birth to weaning are responsible for
no less significant changes in behavioral and
physiological thermoregulation. In contrast with
Davydov and Makarova, we have not utilized an
experimental approach, but rather an ecological
one. As in Fay and Ray’s work, we have paid
particular attention to adaptations for heat dis-
sipation near the upper tolerance limits, partly
owing to difficulties of observation in cold or
inclement weather.

Material and Methods

Data were gathered almost entirely from wild
and penned animals in Antarctica. Wild animals
were easily accessible by tracked vehicle from
McMurdo Station. Penned animals were two
adult females and their pups that were netted
and airlifted by helicopter to wood-slatted, 5-
by-5 meter pens placed on the sea ice 0.5 kilo-
meter west of McMurdo Station.

During the spring season when most of our
work was done, Weddell seals were mostly
gathered in rookeries or smaller nonbreeding
groups from Scott Base, Hut Point Peninsula, to
Cape Royds: i.e., between 77° 33' and 11° 52'
south latitude, and 165° 5’ and 166° 53' east
longitude. Temperatures are not so extreme as
they are in the interior of the continent, but are
more so than in similar arctic latitudes, owing to
higher velocity winds of colder temperatures
from the south polar plateau.

Physiological measurements were made on
slightly disturbed or undisturbed animals except
for thermal gradients in the superficial tissues of
adults, which were taken immediately after
death from animals killed for other purposes
(dog food, specimens) . A few observations were
also made for comparative purposes on captives
that were returned to the New York Aquarium,
but no physiological data from them are indi-
cated on our graphs. Skin samples were taken in
the field and returned to New York for tanning
and study. Respiratory and heart rates were

taken visually, or, in the case of the latter, oc-
casionally by means of a Cambridge Trans-
Scribe electrocardiograph. Temperatures were
taken with a six-channel Yellow Springs tele-
thermometer; the most efficient probe proved to
be a 20-gauge hypodermic type, which stabilized
rapidly. A few temperatures were taken with
Weston thermocouples when the telethermom-
eter was not available. Rectal temperatures were
taken on 15- to 30-cm penetration unless other-
wise specified.

One pup was kept crated for a time in front of
the biological laboratory for metabolic tests.
When the gut was empty, the pup was sealed for
one to two hours in a sheet-steel chamber, 1-by-
l-by-2-meters, from which air samples were tak-
en from two heights with 10-cc syringes through
rubber stoppers in the chamber. Gas analyses
were made with a Scholander 0.5-cc analyzer
(Scholander, 1947) .

Orientation and behavior studies were made
visually from the ground and from the air. Aerial
photography was used most to determine orien-
tation to the sun.

In an effort to standardize, we made ex-
tensive tests to determine the best location
on the seal’s body for the taking of skin tem-
peratures. In all cases, unless otherwise speci-
fied, temperatures are those from the dry body
or flipper perpendicular to the sun, or, if in
cloudy weather, farthest from the ice surface.
This minimized the effects of dampness and
vasoconstriction. Data from wet animals were
treated separately. In a few cases, we were able
to take skin temperatures from submerged or
recently emerged animals to note the extent of
vasoconstriction upon exposure to ice water.

For every physiological measurement, micro-
environmental weather data were taken. Being
cognizant that ambient shade temperature is not
the only parameter of weather that affects ther-
moregulation, we have used “effective tem-
perature” (cf. Eagan, 1964, & Folk, 1966),
calculated here as follows: the black-bulb tem-
perature in still air was taken in the seal’s micro-
environment as a measurement of insolation,
and from this temperature one degree centigrade
for every mile per hour of wind was subtracted.
There was a slightly better alignment of data
when using effective rather than ambient tem-
peratures (Text-figs. 1-4), but the major advan-
tage was that effective temperature gives a more
realistic environmental parameter when insola-
tion and/or wind are strong. For instance, Fay
and Ray ( 1968) have shown that insolation and
wind act independently of ambient temperature
in influencing behavioral thermoregulation.

1968] Ray & Smith: Thermoregulation of Weddell Seal, Leptonychotes weddelli (Lesson), in Antarctica 35

The dotted lines drawn on the graphs of Text-
figs. 1-4 are our estimates of maximal skin
temperatures in air. It is neither practical nor
realistic to use statistical methods (i.e., regres-
sions) in this work since “plateau" limits of
temperature during vasodilation (Ray & Fay,
1968), not “average” skin temperatures, are
probably more meaningful in adaptation.

Results

Internal Temperatures

Only five measurements were made rectally
but indications are that the internal temperature
of the Weddell seal is close to 37°C. Three rectal
temperatures of penned pups were all 37°C. The
importance of adequate penetration into the
rectum is illustrated by other two temperatures
taken in adults: 5°C circumanally and 28°C
upon only 7.5-cm penetration. Both of these
animals were lying on ice in subfreezing air.
Thermal gradients (Text-fig. 7) also indicate a
core temperature at or very near 37°C, as in-
dicated by deep blubber and muscle measure-
ments near this temperature.

Temperature of the Skin When Wet

We had few opportunities to obtain surface
temperatures on swimming, wild Weddell seals
(Table I). One slightly active adult in the water

five temperatures as high as 11°C. The highest
skin temperatures were obtained upon emer-
gence into air of the warmest effective tem-
perature, in agreement with data obtained for
walruses by Ray and Fay (1968).

Temperature of the Dry Fur and Skin

For adults, temperatures of the outer fur sur-
face and the skin were taken from the same lo-
cation on the body. For pups, the same applies,
but the 2.0-2.5-cm thickness of the lanugo made
possible the additional measurement of fur tem-
peratures 1 cm beneath the surface. For both
adults and pups, the skin temperatures of the
flippers were taken at the flipper tip and on the
membrane halfway to the tip from the ankle.

Adult fur-surface temperatures were usually
not more than about 10°C higher than effective
temperatures until about freezing, when the
gradient was greater (Text-fig. 1 ). Fur tempera-
tures rose as a straight regression with effective
temperatures. Skin temperatures were consider-
ably higher than fur temperatures. None was
below freezing, but some were as low as 0 to 7°C
at effective temperatures of —13 to — 21 °C. Skin
temperatures did not rise as a straight regression
with effective temperatures, but reached a pla-
teau of about 34°C at effective temperatures
over — 13°C.

Pup fur-surface temperatures were similarly

Table I

Skin Temperatures of Weddell Seals When in Water or Just Emerged

Effective Air
Temperature

Water

Temperature

Skin

Temperature

Flipper

Temperature

Remarks

-6.0

2.5

8.5

Adult, just emerged

-2.5

-2.0

Adult, in water

11.0

20.0

18.0

Adult, recently emerged

-19.0

10.0

2.0

Pup, just emerged

-19.0

-2.5

-1.0

Pup, just emerged

11.0

23.0

16.0

Pup, just emerged

of an access hole had a skin temperature a frac-
tion higher than the supercooled — 2.5°C water,
confirming the data of Irving and Flart ( 1957 ).
Just under the ice surface, the water temperature
was —1.9°, in which case the water/skin gradient
would be slightly smaller.

We were able to obtain data from five just-
emerged animals (Table I). All animals tested
had been out of water less than five minutes
and most about a minute, yet skin temperatures
rose in that short time to a maximum of 23 °C
and flipper temperatures to 18°C in air of effec-

low at low effective temperatures but did not
rise as fast as those of adults (Text-fig. 2). We
attribute the presence of fur-surface tempera-
tures lower than the effective temperatures (in
the upper range) to the fact that their fur was
often slightly damp from melted snow. The fur
temperature a centimeter below the fur surface
was considerably higher than the fur-surface
temperature. Skin temperatures were almost al-
ways high, plateauing at about 34°C at effective
temperatures from —20 to 10°C.

The flipper tip temperature of adults and pups

36

Zoologica: New York Zoological Society

[53: 1

Text-fig. 1. Temperatures of the dry fur surface
(— T) and skin (= x) of adult Weddell seals at
rest in air.

EFFECTIVE TEMPERATURE »C

Text-fig. 3. Temperatures of the flipper tip of
adult (= T) and pup (— •) Weddell seals at rest
in air. The flipper tip was often wet or damp.

Text-fig. 2. Temperatures of the fur surface (= •),
fur 1 cm beneath the surface (= o) and dry skin
(= x) of pup Weddell seals at rest in air.

Text-fig. 4. Temperatures of the dry flipper web
of adult (= t) and pup (= •) Weddell seals at
rest in air.

1968] Ray & Smith: Thermoregulation of Weddell Seal, Leptonychotes weddelli (Lesson), in Antarctica 37

rose as straight regressions with effective tem-
perature (Text-fig. 3). The temperature at the
tip of the flipper descended to — 4°C at an effec-
tive temperature of — 13°C and rose to 30°C at
an effective temperature near freezing. The flip-
per tip of both adults and pups was often damp
with melted snow, urine, or excrement. The flip-
per membrane of adults and pups was similar to
the body skin temperature, reaching plateaus of
about 34°C at effective temperatures over
-15°C (Text-fig. 4).

Data were also obtained for skin temperatures
from the base of the flipper (ankle) and the
pelvic region (hip) in an effort to establish a
gradient from flipper tip to body. However, these
results were indistinguishable from each other
and from temperatures of the body skin and
flipper web and so are not figured.

Vasoconstriction for body skin, where tem-
peratures would not rise to plateau levels, might
occur at effective temperatures below about
— 13°C for adults and — 15°C or lower for pups.
Our estimated slopes, shown by dotted lines on
Text-figs. 1-4, indicate that skin temperatures

might not descend to 0°C until effective tempera-
tures of at least — 30°C are attained, but data at
these low effective temperature levels are not
available.

Breathing and Heart Rates

Our few data, all from resting animals, are
presented in Text-figures 5 and 6. It is immedi-
ately apparent that pups had higher breathing
and heart rates than adults, reflecting higher me-
tabolic rates as well. For adults, the average
heart rate was 56 per minute and the average
breathing rate was 7.8 per minute. For pups,
these rates were 123 and 16.3, respectively. It
was noted that marked bradycardia occurred
when the breath was held and that both rates
were highly irregular. Heart rates taken by aid
of an electrocardiograph were relatively high ow-
ing to the disturbance caused to the animal. We
were unable to demonstrate a correlation of
these rates with effective temperature, as for
walruses in Ray and Fay (1968), though it is
possible that pup heart rates were minimal at
about — 5°C effective temperature.

ADULT

30

uj 20-1

3

oc

UJ

CL

< n

* SO

<

Ixt

a:

m

"r" t r

-20 -SO

EFFECTIVE TEMPERATURE °C

Text-fig. 5. Breathing rates of adult and pup Weddell seals at rest in air.

38

Zoologica: New York Zoological Society

[53: 1

. ADULT

! 50-

iii

s-

|llO

<z

u

a,

m
H
<
yd
£Q

70'

K

<

UJ

I

o

»* f#

-20 - 10

«

x

0

T »

10 20

EFFECTIVE

TEMPERATURE °C

Text-fig. 6. Heart rates of adult and pup Weddell seals at rest in air. o = rates taken by electrocar-
diograph.

Metabolic Rates

Seven tests of from 15 minutes to one hour
duration on a pup Weddell seal showed rather
consistent results (Table II). Oxygen consump-
tion averaged 35.3 cc/min/k3/4 and carbon
dioxide production was 25.5 cc/min/k3/4. The
respiratory quotient (R.Q.) was 0.727. High
metabolic rates for seals have previously been
shown by Scholander ( 1940), Scholander, et al.
(1950), Irving and Hart (1957) and Davydov
and Makarova (1964). Very few mammals
show such exception to the Benedict “mouse to

elephant curve” (Brody, 1945, p. 370; Kleiber,
1961, p. 201). Though we do not compare our
tests with those of others, because a pup was
used and some elevation of rate is to be expected,
we do wish to note the extremely high metabolic
rate found here.

Thermal Gradients in the Superficial Tissues

Temperature gradients were recorded through
the fur, skin, and subcutaneous tissues of living
pups and of adults immediately after they were
killed (Text-fig. 7) . In young pups without much

Table II

Metabolic Rates of a Pup

Weddell Seal

Chamber

Temperature

o2

cc/min/k3/4

C02

cc/min/k3/4

O

o &

to

op

to

Duration of Test,
Minutes

3 to 6

40.9

28.7

0.702

15

5 to 8

16.9

12.5

0.740

15

-3 to 0

26.5

18.8

0.709

15

-2 to 4

43.8

30.7

0.706

60

0 to 1

42.25

30.75

0.727

30

-1 to 3

39.8

30.5

0.766

30

-1 to 3

36.5

27.0

0.740

15

Averages

35.3

25.5

0.727

1968] Ray & Smith: Thermoregulation of Weddell Seal, Leptonychotes weddelli (Lesson), in Antarctica 39

blubber (7A & B), skin surfaces were 32 and
27°C at effective temperatures of —8 and 0°C.
Gradients were taken up mostly in the thick
lanugo coat. In contrast, a fat pup (7C), well
toward weaning and damp at the time of record-
ing, had a low skin temperature, illustrating the

ui

effect of dampness in inducing vasoconstriction.
Three examples of adults illustrate three con-
trasting gradients. The first (7D) contrasts tem-
peratures from the shaded side with that of the
sunny side of the seal. The gradient was not so
steep in shade as under direct insolation. Con-

DEPTH CENTIMETERS

Text-fig. 7. Temperature gradients through the superficial tissues of pup and adult Weddell seals
immediately after death. Skin and blubber thicknesses at the site of measurement are indicated be-
neath each section. Air temperatures within 1 meter of the animal are indicated by the open circle to
the left of each section. Lanugo — — . Adult hair = \ . Skin — t=| . Blubber = | : • : | .

Muscle = | mil . On section D, diamonds connected by a dotted line indicate a gardient through the

tissues on the shaded side of the seal.

40

Zoologica: New York Zoological Society

[53: 1

trasting was an example of elevation of skin
temperature to plateau levels (7E) in which the
gradient is mostly in the fur. The last (7F)
shows the effect of intense insolation in which
the skin was over 40°C and deeper tissues 38°C,
indicating that heat gained at the surface was
distributed to other parts of the body through the
circulatory system. Ray and Fay (1968) show
a similar situation in the walrus in which high
surface temperatures are induced by insolation.

Effect of Insolation on Fur and Skin

The lanugo of the Weddell seal is silver-tan to
gray in color, darkest on the back (Plate I) . Flat
wavy hairs form a thick coat about 2.0-2. 5 cm
thick. Short hairs of the juvenile coat are present
under the lanugo from birth and grow to replace
the lanugo at about four to six weeks of age. The
juvenile and adult coats are medium to dark
gray, darkest on the back, flecked with black and
white irregular blotches and spots, especially be-
low. The primary hairs are flattened and about
1 cm in length, pointing backward, but tending
to curve foreward when dry. There are fine sec-
ondary hairs about 0.5 cm in length. Scheffer
(1964) gives the density of primary hairs, on
formalin-preserved skin, as 34/cm2 with a maxi-
mum length of 1.3 cm, and the number of sec-
ondary hairs per unit as three to five.

Five tests were made of the effect of insolation
on tanned skins of pup and adult seals. Skins
were spread half in shade and half under strong
sun. Ambient shade and black-bulb tempera-
tures under sun were taken adjacent to the skins.
The results of the tests were almost identical and
examples of adult and pup will be considered
(Text-fig. 8 A & B ) . Air temperature in shade did
not change throughout the tests. Skin and fur-
surface temperatures in shade rose somewhat
higher than air temperatures in shade for both
adult and pup, probably owing to radiation re-
flected from a nearby white building. Black-bulb
temperatures in sun rose rapidly, but fur-surface
and skin temperatures of adult and pup rose
faster and to a higher level. Cloudiness occurred
about two-thirds through the tests and produced
a fall in black-bulb temperatures with a corre-
sponding fall in skin and fur temperatures.

Several explanations are possible for the rapid
and extensive rise in skin and fur temperature
under sun. The simplest is analogous to that of
Krog (1955) for pussy willow catkins. The
shiny, translucent hairs allow sunlight through
to the skin, where heat is absorbed, but they do
not allow as much heat radiation to escape. The
result is the familiar “green house” effect.

A second phenomenon, a lenticular effect of
the flattened hairs in concentrating the sun’s rad-

iation, may also be in evidence as suggested by
Fay (pers. comm.). Adult hairs are flatter and
more regularly arranged than pup lanugo, and
we see that adult skin temperature rises faster
and higher than that of the pup. However, in
the increasing cloudiness indicated, the adult
skin temperature fell, whereas the thick lanugo
trapped warm air so that no similar fall in the
pup skin temperature was observed.

The Role of Behavior

The Weddell seal does not exhibit the array
of thermoregulatory behavior displayed by wal-
ruses (Fay & Ray, 1968). Nevertheless, some
behavioral patterns are marked. The most nota-
ble of these is the prediliction of the Weddell
seal for sun. On hot, dry days Weddell seals were
observed basking on the surface in numbers not
seen in the same localities during inclement
weather. During sunny weather, the seals, espe-
cially pups, moved about on the ice; it appeared
to us that flippers were spread more than during
cloudy weather and that the seals sought the
sun’s ray by orienting with their long axis per-
pendicularly to them. To test the latter point,
seals were counted, principally from the air, and
their orientation to the sun recorded photograph-
ically (Table III). During cloudy weather about
half the seals were oriented perpendicularly to
the sun, but during sunny weather about four-
fifths were oriented in this way. Plate II shows
two examples of aerial photographs from which
data were taken.

During inclement or very cold weather, Wed-
dell seals are notably either sluggish in air or
retreat to water (Smith, 1965). However, escape
to water is usually not possible for mothers and
their pups. At these times, pups remain in the
lee of their mothers, moving out of the lee and
resuming activity only when effective tempera-
tures reach about — 20°C. We have some evi-
dence that mothers induce activity on the part
of the pups and/or that nursing keeps the pups
somewhat active even in the harshest weather.
This probably serves to prevent drifting in and
allows crystals of frozen moisture to fall off,
thus keeping the lanugo fluffy and dry. Plate I
shows a pup and its mother just after a severe
blizzard. The pup has resumed activity and is
thus dry and warm and free of snow.

Discussion

Scholander (1955), in reviewing the climatic
adaptations of homeotherms, has reasoned that
clues to thermoregulation are principally to be
found in the control of heat dissipation. He has
further stated that basal metabolic rate and in-
ternal temperature are not adapted to climate
and that physical insulation and adaptation of

TEMPERATURE *C TEMPERATURE *C

1968] Ray & Smith: Thermoregulation of Weddell Seal, Leptonychotes weddelli (Lesson), in Antarctica 41

Text-fig. 8. The effect of insolation on the fur and skin surfaces of tanned, dry adult (A) and pup

(B) Weddell seal skins. Air temperature in shade ( O ). Black bulb temperature in sun

( • ). Fur surface temperature in shade ( A )• Fur surface in sun ( A ). Skin

surface in shade ( □ ). Skin surface in sun ( H ).

42

Zoologica: New York Zoological Society

[53: 1

Table III

Orientation of Weddell Seals to Sun

Sky

Percentage of Total

Total Seals

Perpendicular to Sun

Clear-sunny

63

65

28

78

40

70

68

76

33

79

16

71

32

66

60

92

4

100

13

77

24

100

16

88

19

79

14

86

12

92

28

64

28

68

24

79

64

83

40

63

41

73

26

85

62

84

31

81

44

89

Average percentage 79.5

Cloudy or

38

50

partly

100

60

cloudy

64

70

68

53

40

45

33

58

46

39

34

56

82

49

76

47

49

53

35

51

41

56

90

52

Average percentage 52.8

critical temperatures are the major mechanisms
for heat dissipation control. These observations
strongly imply that skin temperature should
vary directly with ambient temperature in order
exactly to balance metabolic heat production
with heat dissipation. Irving and Hart (1957)
and Hart and Irving (1959) show such regres-
sions for restrained seals in experimental
conditions.

Fay and Ray (1968) and Ray and Fay
(1968) have found that for wild and unre-
strained walruses, a plateau of skin temperatures
rather than a straight regression is the case and

that behavior is at least as important as physical
or physiological mechanisms in thermoregula-
tion. Further, tolerance limits, rather than criti-
cal physiological limits, were indicated to be
significant in an animal that, by behavioral
means, is probably able to maintain thermo-
neutrality beyond the limits indicated by physiol-
ogy alone. The present study on the Weddell
seal confirms the observations made for the
walrus. Our results indicate, following the sug-
gestion of Brody (1945, p. 295), that although
critical values are theoretically interesting, they
probably have little significance under natural
conditions.

The Weddell seal is a very large species, prob-
ably the heaviest except the elephant seal, Mir-
ounga sp., and the longest except the elephant
seal and the leopard seal, Hydrurga leptonyx.
The lanugo of pups is thick, as described above.
The blubber appears to be the thickest of any
seal except the elephant seal. At birth the nose-
tail length is about 125 cm and the belly blubber
measures about 1 cm. The blubber is about 5 to
6 cm at weaning, when the pup has grown to
about 160 cm. Yearlings are 160 to 180 cm long
and have about 4 cm of blubber in November.
Females are inclined to be a little longer than
males. Large ones are about 300 cm, and the
blubber is 6 to 9 cm thick. Blubber is not of
constant thickness the year around but varies
with nutrition, nursing, and molting. Hart and
Fisher (1964) have given thickness of blubber
for several species, and the thickest given is that
of the Arctic harp seal, at 6.3 cm. Since the
tropical Hawaiian monk seal’s blubber is 4.5
cm, the thickness of blubber is probably of little
significance under conditions of vasodilation
when blubber is bypassed by blood flowing to the
surface. But blubber thickness is of significance
under conditions of vasoconstriction. Therefore,
a thick blubber becomes vitally important to
polar species only when such a species is swim-
ming in ice water or exposed to harsh conditions
in air.

We have presented data that clearly show that
the wild, unrestrained \yeddell seal, like the wal-
rus (Ray & Fay, 1968), vasodilates in air and
that the dry skin temperature reaches a high
“plateau” level of about 34°C. This high plateau
is reached, however, only on that part of the
seal not in contact with the snow or ice sub-
strate, against which the skin is vasoconstricted.
Heat gained on one surface of the seal is prob-
ably distributed through the body by the circu-
latory system. We have also shown that the
Weddell seal actively seeks the warmth of inso-
lation by orienting with the long axis of the body
perpendicularly to the sun’s rays.

1968] Ray & Smith: Thermoregulation of Weddell Seal, Leptonychotes weddelli (Lesson), in Antarctica 43

As shown by the time-temperature curves of
Text-fig. 8, Weddell seals are aided in the gaining
of heat by what is familiarly termed the "green-
house effect,” wherein heat gained by the sun’s
radiation is conserved by an overlying layer of
hair. This is a new observation for pinnipeds. It
is primarily an adaptation in which Weddell
seals are able facultatively to take advantage of
insolation, when the latter is sufficiently strong,
in an environment that is the coldest known for
mammals. The saving of metabolic heat due to
this efficient utilization of insolation is presum-
ably considerable.

Pups in lanugo appear to be better adapted to
temperatures well below freezing than to tem-
peratures close to or above 0°C. Wet hair has
little insulating value. At low temperatures,
dampness from birth, melting ice, urine, or ex-
crement freezes and falls off so that the lanugo
remains fluffy and dry. Observations of late-born
Weddell seals exposed to high temperatures and
dampness tend to confirm this observation. For
example, a seal estimated at about a week’s age
and still quite thin was observed during a warm,
wet snow fall on November 7. After a few sim-
ilar warm, wet days, this seal was found near its
mother, still thin. After nine days it died. We
speculate that the cause of death was chill ac-
companying dampness and near 0°C weather of
the period. We have subsequently observed that
early born pups gain weight faster than pups in
warmer, wetter conditions. The function of the
lanugo is to retain body heat and to prevent un-
due heat loss from skin, which must remain at
relatively high plateau temperatures to grow
(Feltz & Fay, 1967), and this is most effec-
tively performed when the lanugo is dry and
an effective insulator, even in very cold weather.

Insights into the critical limits of the Weddell
seal were obtained while transporting them from
Antarctica to the New York Aquarium and also
during their gradual acclimatization to temperate
climate there. In November 1963, six Weddell
seals were flown from Antarctica to New York
in about two days. Temperatures in the aircraft
ranged from 12 to 28°C, the latter during a
stopover in American Samoa, where air-condi-
tioning facilities were limited. There, after an
hour at 24 to 28°C, an adult male and female
became hyperactive, even showing signs of de-
lirium, tossing and rolling in their crates. Flipper
temperatures were as high as 39°C and body
skin as high as 37°C. Their breathing rates were
16 per minute, higher than any rates taken in
the field. Both animals died within 12 hours of
this exposure. Two pups showed similar stress,
one biting its flippers and the other showing

greatly increased activity when air temperatures
reached only 18°C. One of the pups also died
shortly after leaving Samoa. Two remaining
seals, a juvenile male and an adult female, re-
mained calm and lived. The juvenile male aston-
ished us by copiously taking water from a paper
cup held for it. Thus, the upper critical tem-
perature limit, suddenly imposed on Antarctic-
acclimatized animals, is probably not reached
at temperatures under 20°C.

The above mentioned female, juvenile, and
pup and two additional Weddell seals brought
from Antarctica in February 1965 underwent
gradual acclimatization during the New York
winter and spring. None of these showed ill ef-
fects from effective temperatures as high as 32
to 34°C, often lying out and basking, their skins
fully dry, under the heat of the summer sun.
They did not even choose to escape to the 12°C
water constantly available to them. Similar ac-
climatization is indicated for walruses by Ray
and Fay ( 1 968 ) , but the Weddell seal apparently
acclimatizes to even higher temperatures than
the walrus, as indicated by the lack of escape
behavior. It is, therefore, probable that upper
critical limits of acclimation and acclimatiza-
tion (see Hart, 1957, for further definition) are
not realistic parameters to be used for the Wed-
dell seal in its natural environment.

We could not observe seals at their probable
lower critical temperatures in the field and had
no means to expose Weddell seals to extremely
low temperatures in captivity. Wild seals were
observed to escape from severe weather condi-
tions lower than effective temperatures of — 30°C
by entering the water, and the number of seals
on the surface appears to be directly propor-
tional to effective temperature (Smith, 1965,
1966). However, judging from experimental
work on the harp seal, Phoca groenlandica,
(Irving & Hart, 1957), we would not expect the
lower critical temperature to be above — 40°C.
A similarly low critical temperature is indicated
here by virtue of the high skin temperatures
found at effective temperatures of — 20°C and
below.

Thus thermoregulation of polar amphibious
mammals, as pinnipeds, presents two different
situations: that in air, where we believe vasodila-
tion and high skin temperatures are normal, and
that in water, where vasoconstriction and low
skin temperatures are normal. In both, we think
the animal seeks to maintain thermoneutrality.
By behavioral means such as orientation to sun
in good weather or escape to water or retreat to
the lee of the wind in inclement weather, the
Weddell seal, as the walrus, seeks to avoid criti-
cal limits and to delimit its tolerance zone. The

44

Zoologica: New York Zoological Society

[53: 1

upper and lower in-air critical limits are not
realistic parameters for assessment of natural
adaptation but only indicate one of the many
imposed limits that may be superseded by be-
havior, as in the case of walrus brooding ( Ray
& Fay, 1968). We conclude that upper critical
temperatures are rarely if ever reached in the
natural environment of polar pinnipeds. Lower
critical temperatures are limited indications of
adaptation because both the walrus and the Wed-
dell seal rarely expose themselves to such low
temperatures but choose to escape to water and
maintain thermoneutrality there well before
lower limiting in-air temperatures are reached.

The primary differences between the walrus
and the Weddell seal appear to be found in the
tolerance zone, its relationship to distribution of
the species, and the behavior exhibited within it.
The walrus, especially the calf, is much more sen-
sitive to cold than the pup or adult Weddell seal.
Both species show some escape to water in in-
clement weather of —20 to — 30°C effective tem-
perature, but the walrus probably does^o nearer
its lower critical limit than the Weddell seal.
The lower critical temperature for walrus calves
is about 5°C and the lower limit of tolerance is
extended by brooding to a lower temperature.
The Weddell pup tolerates at least — 30°C. How-
ever, we wish to point out that much more work
is needed on unrestrained pinnipeds at their
lower tolerance limit. Data are better for higher
temperatures, and the walrus probably is limited
in its distribution by its upper tolerance tempera-
ture of about 18°C and the habit of persistent
huddling. The Weddell seal cannot be so limited.
It is able to withstand, and shows no escape
from, climate not remotely possible in the Ant-
arctic environment. In both, acclimatization is
possible, but, again, in the Weddell seal to a
higher temperature. Finally, the walrus exhibits
an array of behavior that serves to extend the
tolerance zone near to or beyond probable criti-
cal physiological limits whereas in the Weddell
seal few traits of behavioral thermoregulation
are shown. This difference might to an extent be
due to the presence of a coat of insulating hair
in the pup and adult Weddell seal, contrasted
with the practically bare-skinned walrus.

Summary and Conclusions

1. Parameters of physiological function and
behavior were measured in unrestrained Weddell
seals in an effort to delimit the tolerance zone
and its possible relationship to distribution. Sur-
face temperatures were obtained on the fur,
body skin, and rear flippers. Heart, breathing,
and metabolic rates were obtained. Tanned skins

were placed under shade and in sun in order to
measure fur- and skin-temperature changes with
time. Our measure of ambient temperature is
“effective temperature,” which includes the ef-
fects of wind and insolation.

2. Few measurements were made of rectal
temperature, but about 37 °C appears to be nor-
mal.

3. Skin temperatures during immersion ap-
pear to be only slightly higher than water tem-
perature. After the seals’ emersion into air, skin
temperatures rise rapidly, faster in warmer air.

4. Dry skin and flipper temperatures did not
rise as straight regressions with air temperature,
but leveled off at about 34°C at effective tem-
peratures above —13 to — 15°C for adults and
pups. Pups had skin temperatures in excess of
30°C at effective temperature of over — 21°C.
Fur surface temptratures, however, did rise
directly with effective temperatures.

5. There was little or no correlation between
effective temperature and breathing and heart
rates of adults or pups, except possibly for the
heart rate of the pup, which may have been
minimal at about —5°. Pups had extremely high
metabolic rates, in excess of any seal yet tested.

6. Gradients through the skin and blubber
show that most of the gradient is taken up in the
lanugo of the dry pup and in the fur, skin, and
blubber of the adult or wet pup. The blubber of
the adult Weddell seal is thicker than that of any
other phocid except the elephant seal.

7. Tanned skins under sun show a marked
“greenhouse effect” in which both fur and skin
temperatures rise faster and higher than black-
bulb temperatures. A possible lenticular effect
of the flattened translucent hairs of the adult is
indicated in which the adult skin is warmed
faster than that of the pup, but also cools faster
in shade.

8. Behavioral thermoregulation is not as
marked in the Weddell seal as in the walrus.
Seals show orientation to sun and seek lee from
winds and storms. They also show escape to
water from harsh weather conditions.

9. The upper critical limit of the nonclima-
tized Weddell seal in air is in excess of 20°C and
the lower limit in air is probably at least as low as
— 40°C. The upper limit is not a realistic param-
eter of in-air adaptation or of distribution. The
lower limit is probably at least — 40°C and is not
often experienced by most seals because before
that limit is reached they probably escape to
water. The seal, like the walrus, probably main-
tains thermoneutrality when at rest in either air
or water.

1968] Ray & Smith: Thermoregulation of Weddell Seal, Lepton ychotes weddelli (Lesson), in Antarctica 45

10. The principal difference between the pup
and the adult Weddell seal lies in the lanugo hair
of the latter. Its function has been found, as
expected, to provide an insulative coat in which
the thermal gradient may be taken up before
sufficient blubber is present for that purpose.
The lanugo is most effective when thoroughly
dry and so the pup Weddell seal is best protected
at low effective temperatures when moisture will
freeze and not wet fur or skin.

Acknowledgments

Investigations in Antarctica were supported
by grants from the Office of Antarctic Programs,
National Science Foundation, to the New York
Zoological Society (Ray); and by the Antarctic-
Division, New Zealand Department of Science
and Industrial Research and the zoology depart-
ment, especially Professor G. A. Knox and Dr.
Bernard Stonehouse, of the University of Can-
terbury (Smith). Some equipment was supplied
through a grant from the Arctic Institute of
North America to the New York Zoological
Society, under contractual agreements with the
office of Naval Research.

In the field, we received invaluable aid from
personnel of the McMurdo Station Biological
Laboratory, directed at the time by Dr. Donald
Wohlschlag of Standford University; from U.S.
Antarctic Research Program logistics personnel
at McMurdo Station; and from the leaders and
personnel of Scott Base (New Zealand). In par-
ticular, we wish to thank Lt. David O. Lavallee
(U.S.N., ret.), formerly of the Third Naval Dis-
trict; and E.T. Feltz of the Arctic Health Re-
search Laboratory, U.S. Public Health Service,
College, Alaska, both of whom were members
of the New York Zoological Society team in
Antarctica. Dr. Gerald Kooyman, then of the
University of Arizona, gave considerable aid and
advice on methodology.

The following have read and criticized this
paper in part or in its entirety: Dr. Victor E.
Scheffer of the Bureau of Commercial Fisheries,
U.S. Fish and Wildlife Service; Drs. Laurence
Irving and L. Keith Miller of the Institute of
Arctic Biology, University of Alaska; Dr. Fran-
cis H. Fay of the Arctic Health Research Labo-
ratory, U.S. Public Health Service; Dr. Arthur
W. Mansfield of the Arctic Biological Station,
Fisheries Research Board of Canada; and Dr.
Ross F. Nigrelli of the New York Aquarium,
New York Zoological Society.

To each of these individuals and agencies we
express our sincere appreciation.

Literature Cited

Bertram, G. C. L.

1940. The biology of the Weddell and Crabeater
seals, with a study of the comparative be-
haviour of the Pinnipedia. Brit. Mus. (Nat.
Hist.) Sci. Repts. Brit. Graham Land
Exped. 1934-1937, 1:1-139.

Brody, S.

1964. Bioenergetics and growth. Hafner Publish-
ing Co., New York.

Davydov, A. F. and A. R. Makarova

1964. Changes in heat regulation and circulation
in newborn seals on transition to aquatic
form of life. Fiziol. Zhur. SSSR, 50(7) :
894. (Transl. Suppl., Fed. Proc., 24(4):
T563-T566.)

Eagan, C. J.

1964. Effect of air movement on atmospheric
cooling power. Exerpted from “A review
of research on military problems in cold
regions.” AAL-TDR-64-28, Arctic Aero-
medical Laboratory, Ft. Wainwright,
Alaska, November 1964, 1-160.

Fay, F. H., and Carleton Ray

1968. The influence of climate on the distribu-
tion of walruses, Odobenus rosmarus (Lin-
naeus). I. Evidence from thermoregula-
tory behavior. Zoologica, 53(1) 1-18.

Folk, G. E.

1966. Introduction to environmental physiolo-
gy: environmental extremes and mam-
malian survival. Lea & Febiger, Philadel-
phia. 308 pp.

Hart, J. S., and H. D. Fisher

1964. The question of adaptations to polar en-
vironments in marine mammals. Fed.
Proc., 23(6): 1207-1214.

Hart, J. S., and L. Irving

1959. The energetics of harbor seals in air and
in water with special consideration of sea-
sonal changes. Can. Jour. Zool., 37:447-
457.

Irving, L., and J. S. Hart

1957. The metabolism and insulation of seals as
bare-skinned animals in cold water. Can.
Jour. Zool., 35:497-511.

King, J. E.

1964. Seals of the world. Brit. Mus. (Nat. Hist.),
London.

Kleiber, M.

1961. The fire of life. John Wiley & Sons, Inc.,
New York. 454 pp. *

46

Zoologica: New York Zoological Society

[53: 1: 1968]

Kooyman, G.

1965. Maximum diving capacities of the Wed-
dell seal, Leptonychotes weddelli. Science,
151(3717): 1553-1554.

Krog, J.

1955. Notes on temperature measurements in-
dicative of special organization in arctic
and subarctic plants for utilization of radi-
ated heat from the sun. Physiologica
Plantarum, 8:836-839.

Lindsey, A. A.

1937. The Weddell seal in the Bay of Whales,
Antarctica. Jour. Mammalogy, 18(2):
127-144.

Littlepage, J. L.

1963. Diving behavior of a Weddell seal winter-
ing in McMurdo Sound, Antarctica. Ecol.,
44(4) -.115-111 .

Mansfield, A. W.

1958. The breeding behaviour and reproductive
cycle of the Weddell seal ( Leptonychotes
weddelli Lesson). Sci. Repts., Falkland
Isl. Dep. Sur. No. 18.

Ray, Carleton

1965. Physiological ecology of marine mam-
mals. Bioscience, 15(4) : 274-277.

1966. The ecology of Antarctic seals. Antarctic
Journal, July-August, 1(4) : 143-144.

1967. Social behavior and acoustics of the Wed-
dell seal. Acoustics Journal, 2(4) : 105-106.

Ray, Carleton, and F. H. Fay

1968. The influence of climate on the distribu-
tion of walruses, Odobenus rosmarus (Lin-
naeus). II. Evidence from physiological
characteristics. Zoologica, 53 ( 1 ): 19-32.

Ray, Carleton, and D. O. Lavallee

1964. Self-contained diving operations in Mc-
Murdo Sound, Antarctica: observations of
the sub-ice environment of the Weddell
Seal, Leptonychotes weddelli (Lesson).
Zoologica, 49(3): 121-136.

Explanation

Plate I

Fig. 1. Female and nursing pup Weddell seals,
illustrating the color pattern and the tex-
ture of the fur.

Fig. 2. Penned female and pup Weddell seals just
after a severe storm. The female is still
covered with snow and has not yet re-
sumed full activity. The pup has resumed

Sapin-Jaloustre, J.

1952. Les phoques de Terre Adelie. Mammalia,
16:179-212.

Scheffer, V. E.

1958. Seals, seal lions and walruses. Stanford
University Press, Stanford, Calif.

1964. Hair patterns in seals (Pinnipedia). Jour.
Morph., 115(2) :291-301.

SCHEVILL, W. E., AND W. A. WATKINS

1965. Underwater calls of Leptonychotes (Wed-
dell seal). Zoologica, 50(1) :45-46.

SCHOLANDER, P. F.

1940. Experimental investigations on the respi-
ratory function in diving mammals and
birds. Hval. Skrift., No. 22.

1947. Analyzer for accurate estimation of respi-
ratory gases in one-half cubic centimeter
samples. Jour. Biol. Chem., 169:235-250.

1955. Evolution of climatic adaptation in ho-
meotherms. Evol., 9(1): 15-26.

SCHOLANDER, P. F„ R. HOCK, V. WALTERS, AND

L. Irving

1950. Adaptation to cold in Arctic and tropical
mammals and birds in relation to tem-
perature, insulation, and basal metabolic
rate. Biol. Bull., 99(2) : 259-271.

Smith, M. S. R.

1965. Movements of the Weddell seal in Mc-
Murdo Sound, Antarctica. Jour. Wild.
Man., 29(3) : 446-447.

1966. Studies on the Weddell seal in McMurdo
Sound, Antarctica. Ph. D. Thesis, unpub-
lished, University of Canterbury, Christ-
church, N.Z.

Wilson, E. A.

1907. Mammalia (whales and seals). Nat. Ant.
Exped., 1901-1904. Natural History, 2:
1-66.

activity with the result that the snow has
fallen off, leaving the lanugo fluffy and dry.

Plate II

Fig. 3. Weddell seals at random orientation on a
cloudy day.

Fig. 4. Weddell seals oriented perpendicularly to
the sun on a clear, sunny day.

RAY & SMITH

PLATE I

FIG. 1

FIG. 2

THERMOREGULATION OF THE PUP AND ADULT WEDDELL SEAL, LEPTO N YCHOTES
WEDDELLI (LESSON), IN ANTARCTICA.

RAY & SMITH

PLATE II

FIG. 3

FIG. 4

THERMOREGULATION OF THE PUP AND ADULT WEDDELL SEAL, LEPTONYCHOTES
WEDDELLI (LESSON), IN ANTARCTICA.

NEW YORK ZOOLOGICAL SOCIETY

GENERAL OFFICE PUBLICATION OFFICE

630 Fifth Avenue, New York, N.Y. 10020 The Zoological Park, Bronx, N.Y. 10460

Fairfield Osborn
President

Howard Phipps, Jr.
Secretary

OFFICERS

Laurance S. Rockefeller
First Vice-President

Henry Clay Frick, II

Robert G. Goelet
Vice-President

David Hunter McAlpin
Treasurer

Eben W, Pyne
Assistant Treasurer

Edward R. Ricciuti
Editor

William G. Conway
Lee S. Crandall

Robert G. Goelet

Elysbeth H. Wyckoff
Production Editor

EDITORIAL COMMITTEE
Fairfield Osborn
Chairman

Donald R. Griffin F. Wayne King

Hugh B. House Peter R. Marler

Ross F. Nigrelli

William G. Conway
General Director

ZOOLOGICAL PARK

William G. Conway . Director & Curator, Birds

Hugh B. House Chairman &

Associate Curator, Mammals
Grace Davall . . Assistant Curator, Mammals

& Birds

Victor H. Hutchison . . . Research Associate in

Herpetology

Joseph Bell .... Associate Curator, Birds
Donald F. Bruning . .

F. Wayne King . . Associate Curator, Reptiles

Charles P. Gandal Veterinarian

Lee S. Crandall . . . General Curator Emeritus

& Zoological Park Consultant
William Bridges Curator of Publications Emeritus
John M. Budinger . . . Consultant, Pathology

David W. Nellis Mammalogist

James G. Doherty Mammalogist

. . . Ornithologist

AQUARIUM

Ross F. Nigrelli Director Nixon Griffis .... Administrative Assistant

Christopher W. Coates . . . Director Emeritus Robert A. Morris Curator

Louis Mowbray Research Associate in Field Biology

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . . Director and Pathologist

Martin F. Stempien, Jr. . . . Assistant to the

Director & Bio-Organic Chemist
George D. Ruggieri, S.J. . . . Coordinator of

Research & Experimental Embryologist
William Antopol . . . Research Associate in
Comparative Pathology
C. M. Breder, Jr. . . . Research Associate in

Ichthyology

Jack T. Cecil Virologist

Harry A. Charipper . . Research Associate in

Histology

Kenneth Gold Marine Ecologist

Eva K. Hawkins Algologist

Myron Jacobs Neuroanatomist

Klaus Kallman . . . . . .Fish Geneticist

Vincent R. Liguori Microbiologist

John J. A. McLaughlin . Research Associate in

Planktonology

Martin P. Schreibman . . Research Associate in

Fish Endocrinology

INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad. West Indies]

Donald R. Griffin .... Director & Senior Fernando Nottebohm . . . Research Zoologist

Research Zoologist George Schaller Research Zoologist

Peter R. Marler . . . Senior Research Zoologist Thomas T. Struhsaker . . . Research Zoologist

Jocelyn Crane . . . Senior Research Ethologist C. Alan Lill Research Fellow

Roger S. Payne Research Zoologist O. Marcus Buchanan . . . Resident Director,

Richard L. Penney .... Research Zoologist William Beebe Tropical Research Station

^ i ^

X

SCIENTIFIC CONTRIBUTIONS OF THE

NEW YORK ZOOLOGICAL SOCIETY

VOLUME 53 • ISSUE 2 • SUMMER, 1968

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

4. The Pentastomes, W addycephalus teretiusculus (Baird, 1862) Sambon,

1922 and Parasambonia bridgesi n. gen., n. sp., from the Lungs of the
Australian Snake, Pseudechis porphyriacus. By Horace W. Stunkard &
Charles P. Gandal. Text-figures 1-6 49

5. Eastern Pacific Expeditions of the New York Zoological Society. Porcel-

lanid Crabs (Crustacea: Anomura) from the West Coast of Tropical
America. By Janet Haig. Text-figures 1 & 2 57

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park, Bronx Park. Bronx. N. Y. 10460. and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers, $1.50, unless otherwise stated in the Society's catalog of publications. Second-class
postage paid at Bronx, N. Y.

Published July 19, 1968

© 1968 New York Zoological Society. All rights reserved.

4

The Pentastomes, Waddycephalus teretiusculus (Baird, 1862)
Sambon, 1922 and Parasambonia bridgesi n. gen., n. sp., from the
Lungs of the Australian Snake, Pseudechis porphyriacus 1

Horace W. Stunkard2 and Charles P. Gandal3
(Text-figures 1-6)

Introduction

A full-grown specimen of Pseudechis
porphyriacus, taken in Australia and
purchased from a dealer, was received
at the New York Zoological Park December 10,
1965, and died January 14, 1966. During this
period it ate eight mice.

At autopsy, four large and 26 small pentas-
tomes, all gravid females, were removed from
the lungs. The pentastomes were killed and pre-
served in formalin. The large specimens, 45 to
50 mm. long, agreed in all particulars with de-
scriptions of Waddycephalus teretiusculus, a
species described originally by Baird (1862) as
Pentastomum teretiusculum from the mouth of
an Australian “copper-head” snake, Hoploce-
phalus superbus [Denisonia superba ], which died
in the gardens of the Zoological Society of
London. A detailed anatomical study of the spe-
cies was published by Spencer ( 1892) from ma-
terial found in Denisonia superba taken on King
Island, situated midway between the mainland
of Victoria and Tasmania, and additional speci-
mens that had been collected from the lungs of
P. porphyriacus taken in Victoria. According to
Spencer, the parasites were always firmly at-
tached by their hooks with the heads buried
deeply in the wall of the lung. A considerable
pull was required to dislodge the creatures and

investigation supported by Grant NSF GB 3606,
continuation of G 23561.

2Research Associate, the American Museum of Nat-
ural History, New York, New York 10024.

Veterinarian, the New York Zoological Park, Bronx,
New York 10460.

definite cavities were left. Spencer suggested that
the specimen found by Baird in the mouth of
the snake had migrated from the lung after the
death of the host. Heymons (1935) stated that
W. teretiusculus is widely distributed in Aus-
tralia, and in addition to D. superba and P. por-
phyriacus, the parasite had been reported from
the lungs of other Australian snakes: Notechis
scutatus, Demansia textilis, and Demansia psam-
mophis. Heymons (1941b) added that W. tere-
tiusculus not only is common in Australian
snakes but is parasitic in Elaphe radiata, a widely
dispersed species that occurs in southern China,
Bengal, the Malayan peninsula, Sumatra, and
Java.

New Material

The present specimens are clearly members
of the family Sambonidae Heymons, 1941,
which comprises three genera: Sambonia Noc
and Giglioni, 1922; Waddycephalus Sambon,
1922; and Elenia Heymons, 1932. The four
larger individuals may be assigned positively to
Waddycephalus teretiusculus (Baird, 1862). In-
deed, Heymons’ (1935, Fig. 1) drawing of the
species might represent one of our specimens.
These specimens are deposited in the American
Museum of Natural History, No. 12927. The
smaller individuals cannot be included in any of
the described genera, and a new genus, Para-
sambonia, is erected to receive the new species
that they represent. They are named Parasam-
bonia bridgesi, in honor of William Bridges, for
many years curator of publications and editor of
Zoologica.

49

50

Zoologica: New York Zoological Society

[53: 2

Parasambonia n. gen. : Based on female only.
Sambonidae; cephalothorax as wide as the ab-
domen; includes one or two cephalic annuli;
sides of abdomen almost parallel; hooks simple,
equal, median hooks slightly anterior to laterals;
annulation often indistinct; lateral lines not ap-
parent; about sixty annuli; uterine pore in sixth
annulus from posterior end. Type species: Para-
sambonia bridgesi n. sp.

Differential diagnosis : Sambonia and Elenia
are very similar; they are parasites of varanid
lizards; the cephalothorax is small, much nar-
rower than the abdomen, which is wide anterior-
ly and narrows to a stalklike posterior region; the
hooks are very small, have a trapezoid arrange-
ment with the anterior hooks somewhat larger
than the posterior-lateral pair; lateral lines are
recognizable. Waddycephalus is a parasite of
Australian snakes, but it differs from Parasam-
bonia in body form, gradually tapering poster-
iorly; the annulation is definite throughout with
weak lateral lines; the hooks are unequal, the
lateral pair smaller and anterior to the median
ones. The most noteworthy difference between
Waddycephalus and Parasambonia is the pres-
ence in the former genus of large glands, desig-
nated hook-glands by Spencer (1892), that ex-
tend along both sides of the intestine throughout
most of the body.

Specific description, Parasambonia bridgesi n.
sp.: The individuals of the smaller species were
transferred to alcohol; twelve were stained in
paracarmine or haematoxylin and mounted
whole (Text-fig. 1 ) ; two were cut in transverse
serial sections and stained with haematoxylin
and erythrosin. The specimens are very uniform
in size; they are slightly flattened ventrally and
convex dorsally. They vary from 25 to 30 mm. in

Fig. 1. Holotype, stained and mounted, ventral view, Parasambonia bridgesi.

1968]

Stunkard & Gandal: The Pentastomes, Waddycephalus teretiusculus and
Parasambonia bridgesi

51

length and 1.57 to 2.00 mm. in width. The ab-
domen is annulate but the annuli are often in-
distinct, especially at the posterior end and in
regions where pressure from distended uterine
loops obliterates the annulation. There is a slight-
ly narrower region, comprising six to eight an-
nuli, posterior to the cephalothorax and in the
remainder of the abdomen, except for the caudal
tip, the sides are almost parallel. The annuli in
the anterior constricted region are about 1.57
mm. wide and 0.30 mm. long; throughout most
of the body they are 1.65 to 1.90 mm. wide and
0.43 to 0.48 mm. long. In the specimen shown
in Text-figure 1, the annulus that bears the
uterine pore is 1 .20 mm. wide and 0.25 mm. long.
The posterior end of the body may be rounded
or tapered to the tip. The terminal annulus has
a slight dorsoventral cleft and the anus is located
ventrally at the base of the cleft. As noted, in
parts of the body the annuli are not distinct and
in the region posterior to the uterine pore, 1.30
to 1.80 mm in length, the number could be de-
termined in only two specimens. Although the
total number could not be counted with cer-
tainty, there are about 60 annuli. Apparently the
number varies slightly in different specimens.

No respiratory, circulatory, or excretory
organs were recognized. The body cavity is a
haemocoel; it contains a homogeneous fluid lack-
ing cellular elements. Movement of the blood is
effected by regional contractions of the general
musculature. Throughout most of its length, the
body cavity is occupied by loops of the uterus.
Lateral lines, described for related species, were
not observed. The cuticula measures 0.018 to
0.026 mm in thickness; it rests on a single layer
of hypodermal cells. Scattered among the hypo-
dermal cells on the dorsal and lateral sides of
the head region, there are numerous, small, mul-
ticellular glands, the “Hautdriisen” or “Stigmen-
driisen” of German authors. In addition, an ir-
regular, staggered row of these glands (Text-fig.
2) extends around the annuli in most of the ab-
domen. They are often inconspicuous and ap-
parently are absent near the posterior end of the
body. In number and arrangement they are sim-
ilar to those of Sebekia oxycephala as repre-
sented in Heymons (1935, Fig. 11). They vary
from 0.030 to 0.045 mm. in diameter, with a
transparent rim and a deeply staining core, some
0.010 mm. in diameter, which may be a central
duct filled with secretion. These glands are
bathed by the haemolymph of the body cavity
and their function is still uncertain. The cuticula
is turned in and lines the stomodeum, the proc-
todeum, the terminal portion of the uterus, and
the pockets in which the hooks are located.

The cephalothorax is flattened ventrally and

Fig. 2. Dermal glands on annuli, Parasambonia
bridgesi.

from a lateral or ventral aspect, two annuli ap-
pear to be covered by or incorporated into it. It
measures 1.60 to 1.90 mm. in width and 0.75
mm. to 1.20 mm. in length. The head is domed
dorsally; the cuticula and hypodermis are sepa-
rated from the underlying structures and the in-
tervening region is large haemal sinus, which in
fixed specimens is filled with a mass of coagu-
lated body fluid. The anteroventral portion of the
cephalothorax contains the muscular complex
that operates the hooks. The hooks (Text-fig. 3)
are simple, equal in shape and size, with the
lateral hooks situated slightly posterior to the

Fig. 3. Hook, Parasambonia bridgesi.

52

Zoologica: New York Zoological Society

[53: 2

median ones. A scheme for measurements of the
hooks was devised by Fain (1961). The meas-
urements, as denoted in Text-figure 3, are: AB,
0.14 to 0.15 mm.; AC, 0.28 to 0.33 mm.; BC,
0.21 to 0.25 mm.; AD, 0.32 to 0.37 mm.; CD,
0.18 to 0.21 mm. The measurements AD and
CD vary as the hook turns on the fulcrum. The
fulcrum is 0.48 to 0.52 mm. long; it is concave
ventrally, with two ventrolateral prongs that
articulate on either side with the lateral base of
the hook. Protractor muscles inserted on the
hook at B and retractors inserted at C operate
the rotary movements of the hook. Other mus-
cles, inserted on the base of the hook and on the
fulcrum, determine the position and orientation
of the hook in the cephalothorax.

The anteroventral face of the cephalothorax
bears two papillae, the “Tastpapillen” of Leuck-
art (1860), the “Sinnespapille A” of Haffner
(1926) and the “Frontalpapillen” of Heymons
(1935). They appear as protrusions of the
body wall, anterior to and midway between the
mouth and median hooks. In sections, they are
pyriform, wider internally, 0.017 mm. long and
0.011 mm. wide, and resemble the “Papille A”
of Porocephalus clavatus as figured by Haffner

(1926). The tip of the papilla bears a row of
sensory cells, supplied by a nerve from the
ganglion of that side, and contains the opening
of a duct, 0.026 mm. in diameter, lined with
low cuboidal epithelium, that extends dorsad
and posteriad, lateral to the median sinus and
above the musculature of the hooks. The duct
appears to subdivide and terminate among the
secretory cells located dorsal and lateral to the
esophagus and the neural ganglia.

The paired ganglia (Text-fig. 4) are ventral,
situated immediately posterior to the mouth.
Each is 0.08 to 0.09 mm. in lateral and 0.06
to 0.07 mm. in dorsoventral measurements, and
a commissure from the anterior part of the
ganglia passes dorsally around the esophagus.
Paired nerves extend from the ganglia to ce-
phalic structures and a principal midventral
pair extends posteriad through the length of the
body. In the dorsolateral region, posterior to the
esophagus and anterior to the intestine there
are masses of secretory cells, each 0.06 to 0.08
mm. in diameter, with nuclei that measure
0.016 to 0.02 mm. in diameter and contain
nucleoli that stain deeply. These cells appear
comparable to the “Kopfdriisen,” “Frontal-

Fig. 4. Cross section through the subesophageal ganglia, Parasambonia bridgesi.

1968]

Stunkard & Gandal: The Pentastomes, Waddycephalus teretiusculus and
Parasambonia bridgesi

53

driisen,” or “Hakendriisen” of Heymons
(1935) and other authors. The function of
these glands is problematical and the same is
true of the so-called “Parietaldriisen,” occa-
sional small groups of glandular cells situated
along the abdomen between the somatic muscu-
lature and the body wall.

The mouth is situated between the median
hooks, 0.50 to 0.60 mm. from the anterior end
of the body. It is oval, longer in the anteropos-
terior axis; the opening varies from 0.08 by 0.16
to 0.12 by 0.23 mm. The mouth is situated in
a cuticular frame, (Text-fig. 5), 0.41 to 0.48
mm. long and 0.21 to 0.25 mm. in greatest
width. The pharynx, dorsal to the mouth and
the esophagus, lined with cuticula, extends pos-
tered above the ganglia to open into the in-
testine in the third or fourth annulus. The
intestine is straight, leading to the anus at the
posterior end of the body, but the wall of the
gut has many large, internal folds (Text-fig. 6).

The ovary is a flattened tubular structure
with lateral wings. It is supported by a mesen-
tery and applied to the dorsal surface of the
intestine. It begins a short distance anterior
to the uterine pore, near the level of the middle
of the metraterm, and extends to the fourth
annulus from the anterior end of the body. Here
paired oviducts pass ventrad on either side of
the body and join in a wide tubular structure
that receives ducts from the two seminal re-
ceptacles. The receptacles are oval, 0.60 to 0.75
mm. long and 0.45 to 0.54 mm. wide. They are
situated on opposite sides of the body and may

5

Fig. 5. Circumoral cuticular frame, Parasambonia
bridgesi.

be at the same level or one may be anterior
to the other. After fertilization, the eggs pass
into the uterus. The uterus has a thin mem-
branous wall and is disposed in many loops
(Text-figs. 1, 6). The uterine pore is situated
on the ventral surface of the sixth annulus from
the posterior end of the body. The terminal por-
tion of the uterus, comparable to the metraterm
of the digenetic trematodes, is constricted, has
a strong muscular wall consisting of outer cir-
cular and inner longitudinal fibers, and has a
cuticular lining. It extends anteriad from the
uterine pore for 0.70 to 0.80 mm., where it
becomes continuous with the last thin-walled

Fig. 6. Cross section through the uterine pore, showing eggs, uterine coils at this level, and folded wall
of the intestine, Parasambonia bridgesi.

54

Zoologica: New York Zoological Society

[53: 2

loop of the uterus, which turns posteriad to the
level of the pore and then crosses to the oppo-
site side of the body, where it loops posteriad.
Newly formed eggs are oval, have smooth shells,
are remarkably uniform in size, and average
0.055 by 0.038 mm. The shells of older eggs
(Text-fig. 6) are 0.074 by 0.052 mm. and are
covered by cylindrical projections, 0.005 to
0.007 mm. long, which appear to be embedded
in a transparent matrix with a delicate mem-
branous covering. Eggs in the terminal portion
of the uterus contain fully formed larvae.

Type depository: American Museum of Nat-
ural History, holotype No. 12924; paratypes:
whole mounts and sections No. 12925; speci-
mens in alcohol, 12926.

Discussion

The first comprehensive study of the mor-
phology and life cycle of the pentastomes was
made by Leuckart (1860). He demonstrated
that Pentastomum denticulatum (Rudolphi,
1809) Rudolphi, 1819, is the larval stage ot
Pentastomum taenioides (Rudolphi, 1809) Ru-
dolphi, 1819, and that both are identical with
Linguatulida serrata Frohlich, 1789. In a foot-
note, Leuckart (1860, p. 2), noted that several
helminthologists, including von Nordmann and
van Beneden, had employed the generic name
Linguatula Frdhlich, 1789, which he conceded
was correct and in accord with the law of pri-
ority. But, influenced by the prestige of Ru-
dolphi, German investigators had assigned spe-
cies to the genus Pentastomum Rudolphi, 1819,
and since the majority of named species had
been described as members of Pentastomum,
Leuckart adopted the name, although it is a
junior synonym of Linguatula.

Classification: The classification of the pen-
tastomes is based largely on the work of Sam-
bon and Heymons, with a revision by Fain
(1961). In early literature the tongue-worms,
as they were called, were included in the family
Linguatulidae and comprised three genera: Lin-
guatula Frohlich, 1789, Porocephalus Hum-
boldt, 1809, and Pentastoma Rudolphi, 1819.
In a contribution published in two parts, Sam-
bon (1922a, June 15; 1922b, December 15)
made extensive additions to knowledge of these
parasites. He (1922a) divided the family Lin-
guatulidae into two subfamilies: Raillietiellinae,
based on Raillietiella Sambon, 1910, and Poro-
cephalinae n. subf., based on Porocephalus
Humboldt, 1809. The genus Linguatula was
placed in the subfamily Porocephalinae, an ob-
vious error, since it was the nomenclatural type
of the family and must be type of the subfamily
to which it is assigned. The two subfamilies

recognized by Sambon are morphologically
distinct, and Sambon noted that in the Raillie-
tiellinae the hooks are borne on stumplike pro-
trusions of the body wall; the female genital
pore is near the anterior end of the abdomen;
the uterus is straight and sacciform; and the
larva has six short, stumpy legs. In the sub-
family Porocephalinae, on the other hand, the
hooks are sessile; the female genital pore is at
or near the posterior end of the body; the
uterus is tubular, elongate, with numerous
windings; and the larva has four legs. Sambon
transferred several species to Raillietiella and
included the genus Reighardia Ward, 1899, in
the subfamily Raillietiellinae. The Porocepha-
linae were divided into three sections: I. Sebe-
kini, with the genera Sebekia Sambon, 1922;
Alofia Giglioni, 1922; and Leiperia Sambon,
1922; II. Porocephalini with the genera Poro-
cephalus Humboldt, 1809; Kiricephalus Sam-
bon, 1922; Armillifer Sambon, 1922; and Wad-
dycephalus Sambon, 1922; III. Linguatulini
with the genera Linguatula Frohlich, 1789, and
Subtriquetra Sambon, 1922.

Noc and Giglioni (1922) erected the genus
Sambonia to contain Pentastomum clavatum
Lohrmann, 1889, a species that had been in-
cluded in the genus Sebekia by Sambon
(1922a). In a paper published on October 27,
Heymons (1922) described Raillietiella ma-
buiae and erected a new genus, Cephalobaena,
to receive a new species, described as Cephalo-
baena tetrapoda, taken from South American
snakes. He included Cephalobaena and Raillie-
tiella in a new family, Cephalobaenidae. But
Raillietiella had been named as a type of the
subfamily Raillietiellinae by Sambon (1922a),
and since the name of the family and type sub-
family must be formed from the name of the
type genus, the family name must be Raillie-
tiellidae, and Cephalobaenidae is a subjective
synonym. Sambon (1922b) suppressed Cepha-
lobaena as a synonym of Raillietiella, discussed
previously described species, and compiled a
list of the Linguatulidae, arranged according
to classification, definitive hosts, and geographi-
cal distribution.

Heymons (1935) published a comprehensive
treatise on the Pentastomida and presented a
revised system of classification. The group was
accorded the taxonomic status of a class. He
noted that the two names, Pentastomida and
Linguatulida, were equally available; and al-
though Linguatula antedates Pentastoma, he
chose to follow Leuckart and adopted Pentas-
tomida as the name for the class, because Ru-
dolphi had formulated a concept of the group,
whereas Frdhlich merely described a larva

1968]

Stunkard & Gandal: The Pentastomes, Waddycephalus teretiusculus and
Parasambonia bridgesi

55

from the lungs of the hare. Rudolphi’s concept
of the group was vague, however, since he in-
cluded Pentastomum in the Trematoda, be-
tween Polystomum and Tristomum on the mis-
taken belief that the hooks of the pentastomes
were comparable to the suckers and hooks of
the pectobothriid trematodes, which at the time
were believed to be anterior in position. Com-
parison with other invertebrate groups led Hey-
mons to the conclusion that the pentastomes
were derived from annelidlike progenitors; more-
over, their morphology and distribution sug-
gested an isolated and phylogenetically ancient
group, probably parasitic in extinct paleozoic
reptiles. In a recent study, Osche (1963) re-
affirmed the arthropod affinities of the pentas-
tomes.

In the class Pentastomida, Heymons (1935)
recognized two orders: Cephalobaenida, with
two families, Cephalobaenidae and Reighardii-
dae; and Porocephalida, with two families, Po-
rocephalidae and Linguatulidae. Since supra-
familial names are not subject to the rules that
apply to family names, the order Cephalobaen-
ida is acceptable, but as noted earlier, if Rail-
lietiella and Cephalobaena are in the same sub-
family, the name of the family must be Rail-
lietiellidae. The family Reighardiidae contained
only the genus Reighardia. In the family Poro-
cephalidae, Heymons recognized four subfam-
ilies: I. Porocephalinae with five genera; Poro-
cephalus Humboldt, 1809; Armillifer Sambon,
1922; Kiricephalus Sambon, 1922; Waddyce-
phalus Sambon, 1922; and Ligamifer Heymons,
1932. II. Sebekinae with three genera; Sebekia
Sambon, 1922; Leiperia Sambon, 1922; and
Alofia Giglioni, 1922. III. Diesinginae with two
genera; Diesingia Sambon, 1922; and Elenia
Heymons, 1932. IV. Samboninae, new subfam-
ily, to receive Sambonia Noc and Giglioni,
1922. The family Linguatulidae contained two
genera: Linguatula Frohlich, 1789, and Sub-
triquetra Sambon, 1922. The two orders recog-
nized by Heymons are virtually identical with
the two subfamilies of Sambon (1922a). Hey-
mons (1940) noted the agreement between
Sambonia and Elenia and suggested that the
two may be identical. In a later report, Heymons
(1941a, p. 323) stated “Die drei Gattungen,
Sambonia, Elenia and Waddycephalus stehen
einander sehr nahe.” Further comparison led
to the conclusion that the two genera, Sam-
bonia and Elenia should be retained and four
genera, Sambonia, Elenia, Waddycephalus, and
Ligamifer, were characterized. The subfamily
Samboninae was elevated to family status to
contain the first three genera. Relationship of
Ligamifer to the genus Armillifer was noted

with the comment (Heymons, 1941a, p. 325),
“Wo der Trennungsstrich zwischen Sambodi-
den und Armilliferiden zu ziehen ist, wissen
wir noch nicht.”

In a revision, Fain (1961) retained the two
orders Cephalobaenida and Porocephalida. Ce-
phalobaenida contained two families: Cephalo-
baenidae and Reighardiidae. The Cephalobaeni-
dae contained three genera: Cephalobaena with
a single species; Raillietiella with 19 species
arranged in five groups; and Megadrepanoides
Self and Kuntz, 1957, with two species. The
order Porocephalida was divided into two sub-
orders: Porocephaloidea and Linguatuloidea.
The Linguatuloidea contained the single family
Linguatulidae and the single genus Linguatula.
The suborder Porocephaloidea contained five
families: Porocephalidae, Sebekidae, Armillife-
ridae, Sambonidae, and Subriquetridae, a new
family, to receive the genus Subtriquetra, which
was transferred from the Linguatulidae. The
family Sambonidae contained three genera:
Sambonia, Elenia, and Waddycephalus. Fain
noted that infection by two different species
of pentastomes is a rare occurrence. It is in-
teresting also to note that with the exception
of two genera —Reighardia, whose only species
occurs in birds; and Linguatula, with species
only in mammals— all other pentastomes are
parasites of reptiles. Larval stages occur in
fishes and mammals, but specimens become
adult only in amniotes.

Literature Cited

Baird, W.

1862. Description of some species of Entozoa.
Proc. Zool. Soc. Lond. 113-115.

Fain, A.

1961. Les Pentastomides de l’Africa Centrale.
Musee Roy. de l’Africa Centrale, Ann.
Sci., Zool. Ser. 8, (92): 1-114.

Haffner, K. von

1926. Die Sinnesorgane der Linguatuliden, nebst
einer Betrachtung iiber die systematische
Stellung dieser Tiergruppe. Zeit. w. Zool.
128:201-252.

Heymons, R.

1922. Beitrag zur Systematik und Morphologie
der Zungenwiirmer (Pentastomida). Zool.
Anz. 55:154-167.

1935. Pentastomida. In Bronn’s Klassen u. Ord-
nunger des Tierreichs. 5(4): 1-113, 161-
268.

1940. Zur Kenntnis der Hautung von Samboni-
den und Sebekiden nebst Bemerkungen
iiber die Gattung Raillietiella (Pentasto-
mida). Zool. Anz. 130:89-91.

56

Zoologica: New York Zoological Society

[53: 2: 1968]

1941a. Beitrage zur Systematik der Pentastomi-
den. IV. Zur Kenntnis der Sambonidae.
Zeit. Parasitenk. 12:317-329.

1941b. Beitrage zur Systematik der Pentastomi-
den. V. Die Typenexemplare von Dies-
ingia megastoma. Zeit. Parasitenk. 12:330-
339.

Leuckart, R.

1860. Bau und Entwicklungsgeschichte der Pen-
tastomen nach Untersuchungen besonders
von Pent, taenioid.es und Pent, denticula-
tum. Leipzig u. Heidelberg. 160 pp.

Noc, F. and Giglioni, G.

1922. Linguatulids parasitic in monitors. The
new genus Sambonia. J. Trop. Med. Hyg.
25:276-286.

Osche, G.

1963. Die systematische Stellung und Phylo-
genie der Pentastomida. Zeit. Morph.
Okol. Tiere 52:487-596.

Sambon, L. W.

1922a. A synopsis of the family Linguatulidae.

J. Trop. Med. Hyg. 25:188-208.

1922b. A synopsis of the family Linguatulidae.
New species described since the publica-
tion of my “Synopsis of the family Lingu-
atulidae.” Part I. J. Trop. Med. Hyg.
25:391-428.

Spencer, W. B.

1892. The anatomy of Pentastomum teretiuscu-
lum (Baird). Quart. J. Micros. Sci. n.
ser. 34: 1-73.

5

Eastern Pacific Expeditions of the New York Zoological Society.
Porcellanid Crabs ( Crustacea : Anomura ) from the
West Coast of Tropical America

Janet Haig

Allan Hancock Foundation
University of Southern California

(Text-figures 1 & 2)

[This is the forty-seventh of a series of papers
dealing with the collections of the Eastern Pacific
Expeditions of the New York Zoological Society
made under the direction of William Beebe. The
present paper is concerned with specimens taken
on the Templeton Crocker Expedition (1936) and
the Eastern Pacific Zaca Expedition (1937-1938).
For data on localities, dates, dredges, refer to
Zoologica, Vol. XXII, No. 2, pp. 33-46, and Vol.

XXIII, No. 14, pp. 287-298.]

Contents page

Introduction 57

Ecological Considerations 59

Geographical Considerations 59

Systematic Considerations 60

Restriction of Synonymies 60

Systematic Discussion 60

Family Porcellanidae 60

Petrolisthes agassizii Faxon 60

P. edwardsii (Saussure) 60

P. glasselli Haig 61

P. polymitus Glassell 61

P. haigae Chace 61

P. armatus (Gibbes) 62

P. nobilii Haig 62

P. zacae, new species 63

P. robsonae Glassell 65

P. gracilis Stimpson 65

P. tridentatus Stimpson 65

P. galapagensis Haig 65

P. tonsorius Haig 66

P. holotrichus Nobili 66

P. platymerus Haig 66

P. ortmanni Nobili 66

P. lewisi lewisi (Glassell) 66

P. lewisi austrinus Haig 67

P. hians Nobili 67

Neopisosoma mexicanum (Streets) 67

Pachycheles chacei Haig 68

P. calculosus Haig 68

P. crassus (A. Milne Edwards) 68

P. biocellatus (Lockington) 68

P. vicarius Nobili 69

P. spinidactylus Haig 69

P. panamensis Faxon 69

P. trichotus Haig 69

Minyocerus kirki Glassell 70

Porcellana cancrisocialis G lassell 70

P. paguriconviva Glassell 70

Pisidia magdalenensis (Glassell) 71

Megalobrachium poeyi (Guerin) 71

M. garthi Haig 72

M. festai (Nobili) 72

M. tuberculipes (Lockington) 72

Polyonyx confinis Haig 72

Acknowledgments 72

Literature Cited 73

Introduction

During a three-year period, 1936 to
1938, two expeditions sponsored by
the New York Zoological Society and
under the direction of Dr. William Beebe ex-
plored the west coast of tropical America
aboard the yacht Zaca. The first, the Temple-
ton Crocker Expedition (1936), covered the
west coast of Baja California, the southern half

57

58

Zoologica: New York Zoological Society

[53: 2

of the Gulf of California, and Clarion Island
of the Revillagigedo group; the second, the East-
ern Pacific Zaca Expedition (1937-1938), pro-
ceeded along the coast of Mexico and Central
and South America as far south as Gorgona
Island, Colombia. (See Text-fig. 1.)

The porcellanid crabs from the 1936 Tem-
pleton Crocker Expedition (with the exception
of those collected at Clarion Island) were the
subject of a report by Glassell (1937). While
preparing a monograph on the Porcellanidae of
the eastern Pacific, I had hoped to examine

material of that family obtained during the ex-
pedition of 1937-1938; at that time, however,
it was packed away and not accessible for study,
except for a few lots, which were examined and
subsequently incorporated in the monographic
treatment (Haig, 1960) along with specimens
collected at Clarion Island during the 1936 ex-
pedition.

The bulk of the porcellanids from the East-
ern Pacific Zaca Expedition (1937-1938) has
since become available for examination, and
all the Porcellanidae collected during that ex-

X

25'- MAGDALENA*:
BAY

SAN LUCAS B.

ARENA BANK
GORDA BANKS

MAZATLAN

CLARION ISL.

BANDERAS B.

CHAMELA b\l>^

TENACATITA B.Tx
MANZANILLO^

SIHUATANEJO°-\0
ACAPULCO'

DULCE R.J
PORT ANGELES
PORT GUATULCO
SANTA CRUZ B.
TANGOLA-TANGOLA B
GULF OF FONSECA

(CONCHAGUITA ISL .. LA UNION
MEANGUERA ISL .. MONYPENNY P*T.
POTOSI R.. FARALLONE ISL.

GULF OF

(ALCATRAZ ISL.,
CEDflO ISL .
NEGRITOS ISL.,

SAN JUAN DEL SUR_r.^-JL s
PORT PARKER y ?C”TA\
MU RC l EL AGO B. ‘XoSh-'A*,
POTRERO GRANDE B.^

PORT CULEBRA-
BRAXILITO B.

PIEDRA BLANCA

NICOYA

0ALLE NAS B .

JASPER ISL.

PUNTARENAS - )

GULF OF DULGE

(-Wi).,

EASTERN
E X P E D

PACI FIC
T I O N S

N E V/ YORK

ZOOLOGICAL^ SOCIETY
SHORE

COLLECTING STATIONS

UVITA 1

PEARL 1

i V o

B J /

PAR IDA7 f

'■BAHIA ( °

ISL. /

HONDA \ L-

COI BA

ISL. ) ~

(hann.bal

•*»«) V ^

GORGONA ISL.*/ ™

J >

GA L APAGOS

IS.

N

ECUADOR

PERU

Text-fig. 1. Shore collecting stations of the Eastern Pacific Expeditions of the New York Zoological
Society. For exact locations of associated dredge stations, refer to Zoologica, vol. XXII, no. 2, and vol.
XXIII, no. 14.

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

59

pedition form the subject matter of the present
report. In general, the style of presentation fol-
lows that of Garth (1959, 1961, 1966), who
reported on part of the brachyuran crabs col-
lected by the expedition, the intertidal brachy-
gnaths having been dealt with earlier by Crane
(1947).

The material has been deposited in the Amer-
ican Museum of Natural History; in this re-
port the AMNH catalog number referring to
each lot is indicated in parentheses. A few speci-
mens from the group studied before 1960 were
donated to the Allan Hancock Foundation, and
these are identified (AHF) in the text.

Ecological Considerations

Porcellanid crabs are usually conspicuous
members of the coral community in tropical
seas. Most species are not obligate commensals
with corals, but may be found in a variety of
habitats that offer concealment. Nevertheless
certain species are predominantly coral dwell-
ers and have only rarely been found in other
situations: in the eastern Pacific these include
Petrolisthes glasselli, Petrolisthes polymitus,
Pachycheles biocellatus, Pachycheles vicarius,
and several others. Some species, on the other
hand, are generally found under stones in the
littoral or have been dredged from various kinds
of substrates, but occasionally turn up among
crabs and other animals taken from coral heads.
Petrolisthes edwardsii and Pisidia magdalenensis
are examples.

Of the 36 species of porcellanids from the
Eastern Pacific Zaca Expedition, 18, or 50 per
cent, were collected from corals. Crane (1947,
pp. 88-89) reported that “More than 50 heads
of coral, ranging in diameter from six inches
to more than two feet, were carefully hammered
open and their inhabitants collected.” The best
locality for this type of collecting, as far as
the Porcellanidae were concerned, was Sihuate-
nejo Bay, where 14 species were recovered from
corals in the intertidal zone. Eleven species were
collected from intertidal corals at Jasper Island;
ten at Uvita Bay; at least nine at Port Parker; and
eight at Port Culebra. At Port Guatulco, eight
species were found in corals obtained by diving
in W2 fathoms.

Crane (1947, p. 87) enumerated nine habitat
zones for the intertidal brachygnathous Brachy-
ura of the 1937-1938 expedition. Petrolisthes
zacae, described herein, was found living on
mud among mangroves; otherwise the coral hab-
itat zone was the only one specified for the Por-
cellanidae of the expedition. Thirteen species
not collected in the latter zone were probably

from “stones near midtide levels,” “stones near
low-tide levels,” or “tidepools.”

Dredged species included Petrolisthes robso-
nae and Polyonyx confinis, from mud and man-
grove leaves in 3 fathoms; and Porcellana can-
crisocialis, Porcellana paguriconviva, and Pisidia
magdalenensis, chiefly on sand, mud, crushed
shell, and rock substrates in 2Vi to 30 fathoms.

Among the species collected was an obligate
commensal, Minyocerus kirki, which lives in as-
sociation with sea stars and serpent stars. Por-
cellana cancrisocialis and Porcellana paguricon-
viva, free-living over a rather wide bathymetric
range, are frequently found with various species
of large hermit crabs in their shells; three P.
paguriconviva from the Eastern Pacific Zaca
Expedition were associated with a hermit crab
of unknown identity.

Geographical Considerations

As a result of the present study, the known
range of Pachycheles crassus is extended from
Balboa, Panama, to Sihuatenejo Bay, Mexico.
Smaller northward range extensions include
Petrolisthes tridentatus from Salinas Bay, Costa
Rica, to San Juan del Sur, Nicaragua; and
Pachycheles calculosus and Megalobrachium
festai from Acapulco to Sihuatenejo Bay, Mex-
ico. Petrolisthes robsonae is reported from La
Union, El Salvador, the first precise locality for
the species north of Panama although it was
recorded earlier from an unspecified area in
Mexico. Petrolisthes galapagensis is reported
from the Gulf of Nicoya, Costa Rica, the first
mainland record for the species.

Petrolisthes lewisi is recorded from the Gulf
of California south to Ectfador, but with a wide
gap in its known distribution. Since the two pop-
ulations in this discontinuous range are recog-
nized as subspecies, it may be assumed that they
meet somewhere in the intermediate area, which
is now narrowed with the extension of Petro-
listhes lewisi lewisi southward from Tequepa
Bay to Tangola-Tangola Bay, Mexico, and of P.
1. austrinus northward from Salinas Bay, Costa
Rica, to the Gulf of Fonseca.

In the present report, several species are noted
for the first time from certain countries visited
by the Zaca. These records may be listed as
follows:

New to Mexico: Pachycheles crassus.

New to El Salvador: Petrolisthes robsonae.

New to Nicaragua: Petrolisthes agassizii, P.
edwardsii, P. nobilii, P. tridentatus, P. tonsorius,
P. lewisi austrinus, Neopisosoma mexicanum,
and Pachycheles trichotus.

60

Zoologica: New York Zoological Society

[53: 2

New to Costa Rica: Petrolisthes galapagensis,
Pachycheles chacei, Megalobrachium garthi,
and M. tuberculipes.

Systematic Considerations

Two new species, neither of which is yet
known from any other source, are represented
in the collection of Porcellanidae from the Zaca
Expedition (1937-1938). One of these, Poly-
onyx confinis, has already been treated (Haig,
1960) ; the other, Petrolisthes zacae, is described
in the present report.

Sixty-five eastern Pacific porcellanid species
are now recognized as members of the Panamic
faunal province, which extends from the head
of the Gulf of California to the Gulf of Guaya-
quil, Ecuador, and includes a number of outly-
ing islands. Of these tropical species, fifteen ap-
pear to be restricted to the Gulf of California or
to the Cocos and Galapagos Islands, areas not
visited by the Zaca. The expedition obtained 36
of the remaining 50 species, or 72 per cent of
the total.

Restriction of Synonymies

For each species a reference is given to the
recent revision of eastern Pacific Porcellanidae
(Haig, 1960), which may be consulted for all
earlier references. Pertinent works published
since 1960 are also cited. Otherwise, the synony-
mies are restricted to the following references:
the work containing the original description of
the species; that first citing the name in its pres-
ent combination; and those containing the origi-
nal descriptions of its junior synonyms. For
Petrolisthes armatus and Megalobrachium poeyi,
species that occur in the Atlantic Ocean as well
as the Pacific, junior synonyms with an Atlantic
coast type locality are not cited.

Systematic Discussion
Family Porcellanidae

Petrolisthes agassizii Faxon

Petrolisthes agassizii Faxon, 1893, p. 174. Haig,

1960, pp. 24, 32, pi. 20 fig. 4; 1962, p. 174.

Range.— From Mazatlan, Gulf of California,
to Utria Bay, Colombia. Shore to 5 fathoms.
Material Examined.— 17 specimens from 4 sta-
tions:

Mexico

Tangola-Tangola Bay, December 8-13, 1937,
intertidal in coral, 1 male (12596).

Nicaragua

Corinto, December 28, 1937-January 7, 1938,
intertidal, 1 male, 3 females (12597).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
1938, 6-2 V2 fathoms, rocks, 5 males, 6 females
(12598).

Port Parker, Station 203, D-10, January 22,
1938,6-214 fathoms, rocks, 5 males, 6 females
(12598).

Measurements.— Males 5. 3-6. 5 mm., oviger-
ous females 6. 1-8.5 mm.

Breeding— Ovigerous females from Corinto in
late December or early January, and from Port
Parker in late January.

Remarks— The male specimen collected inte-
tidally at Port Parker was reported by Haig
(1960, p. 257).

Petrolisthes edwardsii (Saussure)

Porcellana edwardsii Saussure, 1853, p. 366, pi.

12 fig. 3.

Petrolisthes edwardsii Stimpson, 1858, p. 227.

Haig, 1960, pp. 24, 33, pi. 21; 1962, p. 175.

Range.— From Santa Maria and Magdalena
Bays, outer Baja California, and Los Frailes,
Gulf of California, to La Plata Island, Ecuador.
Isabel, Tres Marias, Revillagigedo, and Gala-
pagos Islands. Shore to 20 fathoms.

Material Examined— 53 specimens from 12
stations:

Mexico

Chamela Bay, November 17-20, 1937, inter-
tidal, 3 males (12599).

Tenacatita Bay, November 20, 1937, inter-
tidal, 4 males, 5 females, 3 young (12600).

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 2 females, 1 young (12601).

Port Angeles, December 1, 1937, intertidal, 3
males, 2 females ( 1 2602 ) .

Port Guatulco, Station 195, D-15, December
6, 1937, diving in IV2 fathoms, coral, 1 young
male, 2 young females, 2 young (12603).

Tangola-Tangola Bay, December 8-13, 1937,
intertidal in coral, 2 males (1 young), 1 female
(12604).

Nicaragua

Corinto, December 28, 1937-January 7, 1938,
intertidal, 2 females (12605).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?) , 9 males, 4 females ( 1 1788) .

Port Parker, Station 203, D-10, January 22,
1938, 6-214 fathoms, rocks, 2 females (12606).

Port Culebra, January 24-31, 1938, intertidal
from coral, 1 female (12607).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 male (12608).

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

61

Panama

Bahia Honda, March 13-19, 1938, low tide
under stones, 3 males (12609).

Measurements— Males 3.8-12.0 mm., nonovi-
gerous females 4. 2-8. 6 mm., ovigerous females
5.3-10.0 mm.

Breeding.— Ovigerous females from Tenaca-
tita and Sihuatenejo Bays in November, from
Corinto in late December or early January, and
from Port Parker and Port Culebra in January.

Remarks— A specimen from Chamela Bay is
parasitized by a bopyrid. The material collected
intertidally at Port Parker was recorded by Haig
(1960, p. 258).

Petrolisthes glasselli Haig

Petrolisthes glasselli Haig, 1957a, p. 33, pi. 8

figs. 1-3; 1960, pp. 24, 39, pi. 20 fig. 2; 1962,

p. 176. Chace, 1962. p. 623.

Range— From Cape San Lucas, Gulf of Cali-
fornia, to Gorgona Island, Colombia. Isabel,
Tres Marias, Revillagigedo, Galapagos, and
Clipperton Islands. Shore to 4 fathoms.

Material Examined.— 284 specimens from 5
stations:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 1 male, 3 females (12610).

Port Guatulco, Station 195, D-15, December
6, 1937, diving in IV2 fathoms, coral, 1 male, 1
female (12611).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?), 97 males, 121 females, 39 young
(11786).

Port Culebra, January 24-31, 1938, intertidal
in coral, 4 young (12612).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 8 males, 9 females (12613).

Measurements.— Males 4.1-10.0 mm., nonovi-
gerous females 4. 0-7. 8 mm., ovigerous females
5. 2-9. 6 mm.

Breeding— Ovigerous females from Sihuate-
nejo Bay in November, from Port Guatulco in
December, from Port Parker in January, and
from Uvita Bay in March.

Remarks.— This species, which has rarely been
found except in association with corals, is the
only eastern Pacific Petrolisthes with two epi-
branchial spines on either side of the carapace.

The specimens from Port Parker were pre-
viously reported by Haig (1960, p. 262).

Petrolisthes polymitus Glassell
Petrolisthes polymitus Glassell, 1937, p. 81, pi. 1

fig. 1. Haig, 1960, pp. 25, 41, pi. 22 fig. 1;

1962, p. 176.

Range.— From Espiritu Santo Island, Gulf of
California, to La Libertad, Ecuador. Tres Marias
and Galapagos Islands. Shore to 4 fathoms.

Material Examined.— 13 specimens from 4 sta-
tions :

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 1 female (12614).

Port Guatulco, Station 195, D-15, December
6, 1937, diving in IV2 fathoms, coral, 2 males,
5 females (12615).

Costa Rica

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 2 males, 1 female
(12616).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 1 male, 1 female (12617).

Measurements.— Males 4.0-5.0 mm., nonovig-
erous female 4.8 mm., ovigerous females 3.3-
4.9 mm.

Breeding— Ovigerous females from Sihuate-
nejo Bay in November, from Port Guatulco in
December, from Jasper Island in February, and
from Uvita Bay in March.

Remarks— This species was originally de-
scribed from a single specimen collected in the
Gulf of California during the 1936 Templeton
Crocker Expedition. It was taken over a wide
geographical area during various cruises of
Velero III and Velero IV.

Petrolisthes haigae Chace

Petrolisthes marginatus, Haig, 1960, pp. 25, 47,

pi. 20 fig. 1. Not P. marginatus Stimpson.
Petrolisthes sp., Haig, 1962, p. 177.

Petrolisthes haigae Chace, 1962, p. 620, text-

fig. 1.

Range— From Guaymas Bay, Gulf of Cali-
fornia, to Santa Elena Bay, Ecuador. Isabel, Tres
Marias, Revillagigedo, Galapagos, and Clipper-
ton Islands. Shore to about 10 fathoms (excep-
tionally to 22 fathoms) .

Material Examined.— 525 specimens from 10
stations:

Mexico

Tenacatita Bay, November 20, 1937, inter-
tidal, 1 female (12618).

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 14 males, 26 females (12619).

Acapulco, November 25-29, 1937, intertidal,
2 males, 7 females (12620).

Port Guatulco, Station 195, D-14, December
6, 1937, 4 fathoms, coral, 1 male (12621); D-

62

Zoologica: New York Zoological Society

[53: 2

15, December 6, 1937, diving in IV2 fathoms,
coral, 21 males, 35 females, 1 young (12622).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?) , 1 10 males, 146 females ( 1 1780) .

Port Parker, Station 203, D-9, January 22,
1938, 1 V2-4 fathoms, coral, 2 males, 1 young
(12623).

Port Culebra, January 24-31, 1938, intertidal
in coral, 90 specimens (12624).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 18 males, 13 females
(12625).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 16 males, 17 females (12626).

Panama

Bahia Honda, March 13-19, 1938, low tide
under stones, 1 male, 3 females (12627).

Measurements.— Males 2. 6-9. 5 mm., nonovig-
erous females 2. 4-7. 4 mm., ovigerous females
3. 5-9.2 mm.

Breeding.— Ovigerous females from Tenaca-
tita Bay, Sihuatenejo Bay, and Acapulco in No-
vember, from Port Guatulco in December, from
Port Parker and Port Culebra in January, from
Jasper Island in February, and from Uvita Bay
and Bahia Honda in March.

Remarks.— This species occurs abundantly
throughout its range. Chace ( 1962) showed that
it is distinct from Petrolisthes marginatus Stimp-
son, a closely related west Atlantic form.

The specimens collected intertidally at Port
Parker were reported by Haig ( 1960, p. 267) as
Petrolisthes marginatus Stimpson.

Petrolisthes armatus (Gibbes)
Porcellana armata Gibbes, 1850, p. 190.
Petrolisthes armatus, Stimpson, 1858, p. 227.

Haig, 1960, pp. 25, 50, pi. 19 fig. 2; 1962,

p. 178.

Range.— From Puerto Pehasco and San Felipe,
Gulf of California, to Independencia Bay, Peru.
Galapagos Islands. Shore to 10 fathoms. Also
occurs in western and eastern Atlantic.

Material Examined.— 32 specimens from 5
stations:

Nicaragua

Near Potosi River, Gulf of Fonseca, Decem-
ber 23-25, 1937, intertidal, 5 males, 7 females
(12628).

Costa Rica

Port Parker, January 12-23, 1938, intertidal,
6 males, 7 females (12629).

Port Culebra, January 24-31, 1938, intertidal,
3 males (12630).

Cedro Island, Gulf of Nicoya, February 12-
13 or 21-22, 1938, intertidal, 1 male, 1 female
(12631).

Panama

Bahia Honda, March 13-19, 1938, low tide
under stones, 1 male, 1 young female (12632).

Measurements.— Males 2.6-11.7 mm., non-
ovigerous females 3.9-5. 5 mm., ovigerous fe-
males 4. 5-7. 3 mm.

Breeding— Ovigerous females from Gulf of
Fonseca in December, from Port Parker in Jan-
uary, and from Cedro Island in February.

Petrolisthes nobilii Haig

Petrolisthes nobilii Haig, 1960, pp. 25, 55, pi. 1,

pi. 18 fig. 3.

Range.— From Cabeza Ballena, Gulf of Cali-
fornia, to Santa Elena Bay, Ecuador. Isabel Is-
land. Intertidal zone.

Material Examined— 16 specimens from 7 sta-
tions:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal under stones, 1 male (12633).

Port Angeles, December 1, 1937, intertidal,
1 female ( 12634) .

Tangola-Tangola Bay, December 8-13, 1937,
intertidal, 2 males, 2 females (12635).

Nicaragua

Corinto, December 28, 1937-January 7, 1938,
intertidal, 1 male, 1 female (12636).

San Juan del Sur, January 9-12, 1938, inter-
tidal, 2 males, 2 females (12637).

Costa Rica

Ballenas Bay, Gulf of Nicoya, February 25-
26, 1938, intertidal, 1 female (12638).

Colombia

Gorgonilla Island, March 30, 1938, intertidal
under rocks, 1 male, 1 female (12639).

Measurements— Males 6.7-10.4 mm., non-
ovigerous female 4.3 mm., ovigerous females
5.6-10.4 mm.

Breeding— Ovigerous females from Port An-
geles and Tangola-Tangola Bay in December,
from Corinto in late December or early January,
from San Juan del Sur in January, from Ballenas
Bay in February, and from Gorgonilla Island in
March.

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

63

Petrolisthes zacae, new species
(Text-fig. 2)

Types.— Female holotype, AMNH Cat. No.
12640, from Ballenas Bay, Gulf of Nicoya,
Costa Rica, February 25 or 26, 1938, intertidal
in mangrove mud. One male and one female,
paratypes, AMNH Cat. No. 12641, same data
as holotype.

Measurements.— Female holotype, length 8.6
mm., width 7.8 mm. Male paratype, length 5.0
mm. Ovigerous female paratype, length 7.8 mm.

Diagnosis.— Carapace finely rugose; no su-
praocular spine; a single epibranchial spine;
front broad, with three shallow lobes. Carpus of
chelipeds about two and a half times as long as
wide, inner margin with three narrow, wide-set

Text-fig. 2. Petrolisthes zacae. Holotype: a, carapace; b and c, chelipeds; d, e, and f, left walking legs 1,
2, and 3, respectively; g, dactyl of left first walking leg. Paratype: h, basal segment of right antennule;
i, right third maxilliped. (Scale a-f = 4 mm.; g, h = 2 mm.; i = 3 mm.)

64

Zoologica: New York Zoological Society

[53: 2

teeth, outer margin with four similar teeth; chela
long and slender, its outer margin spinulate.
Merus of walking legs with a few spines on upper
margin; dactyl with a single movable spinule on
lower margin.

Description.— C arapace finely rugose, espe-
cially along lateral and posterolateral margins;
dorsal surface flat, regions not strongly indicated
except for hepatics, which lie at a level below
that of rest of carapace, and protogastric lobes.
Front broad, flat or very faintly concave, with
three shallow, rounded lobes, median one broad-
er and more produced than laterals. No suprao-
cular spine. Orbits shallow, strongly oblique;
outer orbital angle subrectangular and some-
times produced into a minute spinule. A well-
developed epibranchial spine. No hairs on dorsal
surface of carapace; lateral portion with very
short hairs.

a series of long, slender, rather evenly spaced
spinules, more than 20 in number in the holo-
type, not developed on the proximal fourth of
the palm nor on the outer margin of the pollex.
Fingers smooth on dorsal surface, curved and
crossing at tips; outer margin of dactyl with a
ridge, produced into a sharp spinule at point
where the finger curves sharply inward and
crosses under pollex. A short, thick pubescence
on lower inner side of fingers, on dactyl extend-
ing more than halfway to tip. Outer margin of
chela with a fringe of very fine hairs, not obscur-
ing row of spinules.

Walking legs transversely rugose, and with a
few very fine, scattered hairs on margins. Merus
with a posterodistal spine on legs 1 and 2, none
on leg 3; anterior margin with a few spines (well-
developed in female specimens, weakly devel-
oped in the small male), as follows in Table I:

Table I

Spinulation of Walking Legs

Legs

Holotype 2
8.6 mm.

Paratype 2
7.8 mm.

Paratype $
5.0 mm.

Leg 1 (left)

4

3

2

Leg 1 (right)

3

3

2

Leg 2 (left)

3

3

2

Leg 2 (right)

3

(leg missing)

(leg missing)

Leg 3 (left)

2

(leg missing)

(leg missing)

Leg 3 (right)

2

(leg missing)

2

First movable segment of antenna with a
strongly projecting, rounded, spine-tipped lamel-
lar lobe; second granular, slightly produced at
proximal end of anterior margin; third smooth;
flagellum without hairs. Outer maxilliped and
antennule as shown in Text-figure 2.

Chelipeds subequal. Merus lightly rugose;
armed on inner margin with a strong pointed
tooth; two spines on dorsal surface near outer
margin, one at distal and the other near proxi-
mal end of segment. Carpus nearly two and a
half times as long as wide; dorsal surface nearly
smooth, evenly convex; armed on inner margin
with three low, narrow, wide-set teeth (the most
distal one not developed in the small male para-
type), their edges finely crenulate; outer margin
with four similar teeth, the most distal one bifid
and placed at outer distal angle, the proximal
two placed slightly on dorsal surface. Chela long
and slender, smooth, evenly convex; outer mar-
gin crenulate, some of the crenulations in the
form of short, close-set spinules; on dorsal sur-
face just to the inside of the crenulated margin,

Carpus with anterodistal spine well developed
on leg 1, obsolescent or absent on legs 2 and 3.
Propodus long and slender; two movable spin-
ules on lower margin in addition to the usual
posterodistal pair. Dactyl long and slender; a
“thumblike” projection about halfway along
lower margin, tipped with a movable corneous
spinule; lower margin otherwise unarmed. (The
spinule and the corneous fixed claw are paler in
color than indicated in Text-fig. 2 d-f.)

Remarks— Petrolisthes zacae is allied to a small
group of species, including P. armatus (Gibbes) ,
P. nobilii Haig, and P. robsonae Glassell, in
which the carapace is not transversely striate;
there are low, wide-set teeth on the inner margin
of the carpus of the cheliped; and the anterior
margin of the merus of the walking legs is armed
with only a few spines. It differs from all of
them by a combination of several characters,
and particularly by the form of the dactyl of the
walking legs. As far as I am aware, the structure
of the dactyl in this species is unique among
porcellanids of genus Petrolisthes. Assuming

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

65

that the types were collected in a situation typical
for the species, it is probably an adaptation for
living in mud.

Petrolisthes robsonae Glassell
Petrolisthes robsonae Glassell, 1945, p. 227,
text-fig. 3. Haig, 1960, pp. 25, 57, pi. 18 fig. 2.
Range— Mexico (specific locality not known)
to Guayaquil, Ecuador.

Material Examined.— ha Union, Gulf of Fon-
seca, El Salvador, Station 199, D-21, December
27, 1937, 3 fathoms, mud, mangrove leaves, 2
males, 1 female (12642).

Measurements— Males 6.5 and 7.0 mm., non-
ovigerous female 5.3 mm.

Remarks— An unusual characteristic of this
species is its ability to withstand great changes
in salinity (Haig, 1960, pp. 58-59). It has been
collected at both ends of the Panama Canal, its
occurrence on the Atlantic side of the Isthmus of
Panama probably being due to an accidental in-
troduction. It is also reported from Bellavista,
Panama City; Guayaquil, Ecuador; and an un-
specified locality in Mexico. With the discovery
of specimens among the material collected by
the Zaca during the 1937-1938 expedition, La
Union becomes the northernmost precise local-
ity known for the species.

Petrolisthes gracilis Stimpson
Petrolisthes gracilis Stimpson, 1858, p. 227
(nomen nudum); 1859, p. 74. Haig, 1960,
pp. 28, 79, pi. 27 fig. 2.

Range.— From Santa Maria Bay, Baja Cali-
fornia, and Punta Penasco and San Felipe, Gulf
of California, to Tangola-Tangola Bay, Mexico.
Tres Marias Islands. Shore; rarely to 25 fathoms.

Material Examined— 6 specimens from 3 sta-
tions:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal under stones, 1 male, 2 females (12643).

Port Guatulco, December 2-7, 1937, inter-
tidal, 1 male ( 12644) .

Tangola-Tangola Bay, December 8-13, 1937,
intertidal, 2 males (12645).

Measurements.— Males 2. 6-3. 9 mm., oviger-
ous females 2.7 and 4.3 mm.

Breeding— Ovigerous females from Sihua-
tenejo Bay in November.

Remarks— Except for a single specimen col-
lected at Tangola-Tangola Bay by the Velero III,
Petrolisthes gracilis has not been known south
of the Gulf of California. I suggested (Haig,
1960, p. 81) that the Tangola-Tangola Bay

record might be erroneous. However, the speci-
mens taken by the Zaca confirm the presence of
the species in southern Mexico.

Petrolisthes tridenfatus Stimpson

Petrolisthes tridentatus Stimpson, 1858, p. 227
( nomen nudum)', 1859, p. 75, pi. 1 fig. 4.
Haig, 1960, pp. 29, 81, pi. 25 fig. 4.

Range.— From Salinas Bay, Costa Rica, to
Puna Island, Ecuador. Intertidal. Also occurs in
western Atlantic.

Material Examined .—18 specimens from 5
stations:

Nicaragua

San Juan del Sur, January 9-12, 1938, inter-
tidal, 1 male, 3 females (12646).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?), 3 males, 2 females (11834).

Cedro Island, Gulf of Nicoya, February 12-13
or 21-22, 1938, intertidal, 3 males ( 12647) .
Jasper Island, Gulf of Nicoya, February 22-

25, 1938, intertidal, 2 males, 3 females ( 12648) .
Ballenas Bay, Gulf of Nicoya, February 25-

26, 1938, intertidal, 1 female (12649).
Measurements.— Males 3. 9-6.1 mm., nonovig-

erous females 3.8 and 4.0 mm., ovigerous fe-
males 3. 6-4. 5 mm.

Breeding— Ovigerous females from San Juan
del Sur and Port Parker in January, and from
Gulf of Nicoya in February.

Remarks— One of the specimens from Port
Parker was reported by Haig (1960, p. 287).
The range of the species is now extended slightly
northward from Salinas Bay to San Juan del Sur.

Petrolisthes galapagensis Haig

Petrolisthes galapagensis Haig, 1960, pp. 28, 84,
pi. 2, pi. 25 fig. 2.

Range.— Galapagos Islands. Shore to 2xh fa-
thoms.

Material Examined.— Jasper Island, Gulf of
Nicoya, Costa Rica, February 22-25, 1938, in-
tertidal, 4 males, 5 females (12650).

Measurements— Males 4. 2-6. 3 mm., nonovig-
erous females 5. 0-5. 5 mm.

Remarks— The above record is the first for
the species outside the Galapagos Archipelago.
In the Galapagos it frequently occurs with its
close relative Petrolisthes tonsorius Haig, and
the two species were encountered together at
Jasper Island as well. They are best distinguished
by the form of the cheliped: in P. galapagensis
the margins of the carpus are subparallel, while

66

Zoologica: New York Zoological Society

[53: 2

in P. tonsorius the inner carpal margin is pro-
duced into a strong lobe.

Petrolisthes tonsorius Haig

Petrolisthes tonsorius Haig, 1960, pp.28, 85, pi.

3, pi. 26 fig. 1.

Range— From Cape San Lucas, Gulf of Cali-
fornia, to Santa Elena Point, Ecuador. Revil-
lagigedo, Cocos, and Galapagos Islands. Shore
to 10 fathoms.

Material Examined— 36 specimens from 6
stations:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal under stones, 1 female (12651).

Port Angeles, December 1, 1937, intertidal,
2 males (12652).

Tangola-Tangola Bay, December 8-13, 1937,
intertidal, 4 males, 8 females, 2 young (12653).

Nicaragua

Corinto, December 28, 1937-January 7, 1938,
intertidal, 1 male (12654).

San Juan del Sur, January 9-12, 1938, inter-
tidal, 1 female (12655).

Costa Rica

Jasper Island, Gulf of Nicoya, February 22-

25, 1938, intertidal, 12 males, 5 females
(12656).

Measurements.— Males 4. 1-9.3 mm., nonovig-
erous females 3. 8-6. 2 mm., ovigerous females
3. 6-8. 3 mm.

Breeding— Ovigerous females from Sihua-
tenejo Bay in November, from Tangola-Tangola
Bay in December, and from Gulf of Nicoya in
February.

Petrolisthes holotrichus Nobili
Petrolisthes holotrichus Nobili, 1901, p. 14.

Haig, 1960, pp. 29, 102, pi. 29 fig. 4.

Range— From Salinas Bay, Costa Rica, to La
Libertad, Ecuador. Intertidal.

Material Examined.— A specimens from 2 sta-
tions:

Costa Rica

Ballenas Bay, Gulf of Nicoya, February 25-

26, 1938, intertidal, 1 female (12657).

Panama

Bahia Honda, March 13-19, 1938, low tide
under stones, 3 females (12658).

Measurements. — Nonovigerous female 2.9
mm., ovigerous females 3. 8-5.0 mm.

Breeding.— Ovigerous females from Gulf of
Nicoya in February and from Bahia Honda in
March.

Petrolisthes platymerus Haig
Petrolisthes platymerus Haig, 1960, pp. 29, 108,

pi. 4, pi. 29 fig. 3.

Range— Known only from Port Parker, Costa
Rica, and Taboguilla Island, Panama. Intertidal.

Material Examined.— Ballenas Bay, Gulf of
Nicoya, Costa Rica, February 25-26, 1938, in-
tertidal, 1 male (12659).

Measurements.— Male 4.5 mm.

Remarks.— This species was previously known
from only 14 specimens collected at two locali-
ties. It may have been overlooked by most col-
lectors because of its small size. The largest in-
dividual on record is the 5.2 mm. holotype, while
egg-bearing females range from 3. 5-4.9 mm.

Petrolisthes ortmanni Nobili

Petrolisthes ortmanni Nobili, 1901, p. 16. Haig,

1960, pp. 27, 112, pi. 23 fig. 3.

Range— From Puerto San Carlos, Gulf of
California, to Lobos de Afuera Islands, Peru.
Tres Marias Islands and Cocos Island. Shore to
3Vi fathoms.

Material Examined.— 57 specimens from 7
stations:

Mexico

Tenacatita Bay, November 20, 1937, inter-
tidal, 1 male (12660).

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 2 males, 4 females (12661).

Port Guatulco, December 2-7, 1937, inter-
tidal, 1 female (12662).

Port Guatulco, Station 195, D-15, December
6, 1937, diving in IV2 fathoms, coral, 10 males,
8 females (12663) .

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?), 11 males, 13 females (11831).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 3 males, 3 females
(12664).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 1 male (12665).

Measurements.— Males 2. 6-5. 6 mm., nonovig-
erous females 3. 2-4. 7 mm., ovigerous females
3. 6-6.6 mm.

Breeding— Ovigerous females from Sihua-
tenejo Bay in November, from Port Guatulco in
December, from Port Parker in January, and
from Gulf of Nicoya in February.

Remarks.— The specimens from Port Parker
were reported earlier by Haig (1960, p. 303).

Petrolisthes lewisi lewisi (Glassell)
Pisosoma lewisi Glassell, 1936, p. 287.

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

67

Petrolisthes lewisi Haig, 1957b, p. 7 (not new

records nor all of synonymy).

Petrolisthes lewisi lewisi, Haig, 1960, pp. 27,

113, pi. 23 fig. 1.

Range.— From Carmen Island, Gulf of Cali-
fornia, to Tequepa Bay, Mexico. Isabel and Tres
Marias Islands. Shore to 3 fathoms.

Material Examined— 11 specimens from 3
stations:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal under stones, 3 females (12666).

Port Guatulco, December 2-7, 1937, inter-
tidal, 2 males, 1 female (12667).

Tangola-Tangola Bay, December 8-13, 1937,
intertidal, 4 males, 1 female (12668).

Measurements.— Males 5. 0-5. 8 mm., nonovig-
erous females 3.0 and 5.8 mm., ovigerous fe-
males 4. 3-5.4 mm.

Breeding.— Ovigerous females from Sihua-
tenejo Bay in November and from Tangola-
Tangola Bay in December.

Remarks— The known range of Petrolisthes
I. lewisi is now extended southeastward from
Tequepa Bay to Tangola-Tangola Bay.

Petrolisthes lewisi austrinus Haig

Petrolisthes lewisi austrinus Haig, 1960, pp. 27,

115, pi. 5, pi. 23 fig. 2.

Range.— From Salinas Bay, Costa Rica, to
Santa Elena Point, Ecuador. Intertidal.

Material Examined— 5 specimens from 2 sta-
tions:

Nicaragua

Near Potosi River, Gulf of Fonseca, Decem-
ber 23-25, 1937, intertidal, 1 male (12669).

Panama

Bahia Honda, March 13-19, 1938, low tide
under stones, 1 male, 3 females (12670).

Measurements— Males 4.5 and 5.5 mm., non-
ovigerous females 2.6 and 4.7 mm., ovigerous
female 3.3 mm.

Breeding. — Ovigerous female from Bahia
Honda in March.

Remarks— The known range of subspecies
austrinus is extended northwestward from Sali-
nas Bay to Gulf of Fonseca. The area of contact
of the two subspecies of Petrolisthes lewisi,
which has yet to be determined, lies somewhere
between the latter locality and Tangola-Tangola
Bay.

Petrolisthes hiams Nobili

Petrolisthes hians Nobili, 1901, p. 17. Haig,

1960, pp. 26, 121, pi. 22 fig. 3.

Pisosoma flagraciliata Glassell, 1937, p. 82, pi.

1 fig. 2.

Range— From Santa Maria Bay, outer Baja
California, and Guaymas, Gulf of California, to
Santa Elena Bay, Ecuador. Isabel, Tres Marias,
and Revillagigedo Islands. Shore to 4 fathoms
(exceptionally to 18 fathoms).

Material Examined— I'M specimens from 8
stations:

Mexico .

Tenacatita Bay, November 20, 1937, inter-
tidal, 1 male (12671).

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 40 males, 44 females (12672) .

Acapulco, November 25-29, 1937, intertidal,
5 males, 4 females (12673).

Port Guatulco, Station 195, D-15, December
6, 1937, diving in Wz fathoms, coral, 4 males,

2 females (12674).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?), 7 males, 3 females (11835) .

Port Culebra, January 24-31, 1938, intertidal
in coral, 4 males, 6 females (12675).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 male (12676).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 8 males, 8 females (12677).

Measurements.— Males 1.6-5. 3 mm., nonovig-
erous females 2. 0-3. 9 mm., ovigerous females
1. 9-5.1 mm.

Breeding.— Ovigerous females from Sihua-
tenejo Bay and Acapulco in November, from
Port Guatulco in December, from Port Parker
and Port Culebra in January, and from Uvita
Bay in March.

Remarks.— Two ovigerous females from Si-
huatenejo Bay were parasitized by a bopyrid.
The specimens from Port Parker were previous-
ly reported by Haig (1960, p. 309).

Pisosoma flagraciliata, a synonym of Petro-
listhes hians, was based on material collected in
the Gulf of California during the 1936 Temple-
ton Crocker Expedition.

Neopisosoma mexicanuim (Streets)

Pachycheles mexicanus Streets, 1871, p. 225,

pi. 2 fig. 1.

Neopisosoma mexicanum, Haig, 1960, pp. 124,

127, pi. 30 fig. 2.

Range.— From Mazatlan, Gulf of California,
to Santa Elena Point, Ecuador. Galapagos Is-
lands. Shore to 10 fathoms.

Material Examined— Corinto, Nicaragua, De-

68

Zoological New York Zoological Society

[53: 2

cember 28, 1937-January 7, 1938, intertidal, 1
female (12678).

Measurements.— Ovigerous female 4.4 mm.

Remarks— N eopisosoma dohenyi Haig, which
occupies much the same area as N. mexicanum
and has been collected with it at Mazatlan and
Acapulco, was not taken during the Zaca expedi-
tion.

Pachycheles chacei Haig

Pachycheles chacei Haig, 1956, pp. 7, 9, pi. 1;

1960, pp. 134, 135, pi. 31 fig. 3.

Range— From San Jose, Guatemala, to Santa
Elena Bay, Ecuador. 1-4 fathoms. Also Atlantic
coast of Panama and Colombia.

Material Examined.— 23 specimens from 2
stations:

Costa Rica

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 male, 2 females
(12679).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 9 males, 11 females (12680).

Measurements.— Males 2. 4-4. 7 mm., oviger-
ous females 2.7-5. 2 mm.

Breeding.— Ovigerous females from Gulf of
Nicoya in February and from Uvita Bay in
March.

Remarks.— The Costa Rican specimens col-
lected by the Zaca bridge a considerable gap in
the known distribution of the species, which has
not been reported previously from the area be-
tween Acajutla, El Salvador, and Isla Verde,
Panama.

Pachycheles calculosus Haig

Pachycheles calculosus Haig, 1960, pp. 135, 136,

pi. 10, pi. 31 fig. 4.

Range— From Acapulco, Mexico, to La Li-
bertad, Ecuador. Shore to 4 fathoms.

Material Examined.— Sihuatenejo Bay, Mex-
ico, November 24, 1937, intertidal in coral, 2
males ( 12681 ) .

Measurements.— Males 4.5 and 4.7 mm.

Remarks.— The range of this species is now ex-
tended northwestward from Acapulco to Sihua-
tenejo Bay.

Pachycheles crassus (A. Milne Edwards)

Porcellana ( Pachycheles ) crassa A. Milne Ed-
wards, 1869, p. 128, pi. 26 fig. 12.
Pachycheles crassus, Haig, 1957b, p. 5; 1960,

pp. 134, 141, pi. 31 fig. 1, text-fig. 4.

Range.— From Balboa, Panama, to Gorgona
Island, Colombia. Shore to 4 fathoms.

Material Examined.— Sihuatenejo Bay, Mex-
ico, November 24, 1937, intertidal in coral, 2
males, 1 female (12682).

Measurements.— Males 3.9 mm. long, 4.8 mm.
wide and 4.8 mm. long, 6.3 mm. wide, ovigerous
female 5.6 mm. long, 7.9 mm. wide.

Remarks.— Pachycheles crassus was previous-
ly known from only nine specimens collected at
five localities. The Zaca material shows the
marked broadening of the carapace, particularly
in females, that is characteristic of the species.

The known range is now considerably ex-
tended northwestward, from Balboa to Sihua-
tenejo Bay.

Pachycheles biocellatus (Lockington)

Petrolisthes (P isosoma) biocellatus Lockington,

1878, pp. 396, 403.

Petrolisthes (P isosoma) gibbosicarpus Locking-
ton, 1878, pp. 396, 402.

Pisosoma aphrodita Boone, 1932, p. 53, text-

figs. 17-18.

Pachycheles biocellatus, Glassell, 1937, p. 84.

Haig, 1960, pp. 134, 144, pi. 32 fig. 1. Chace,

1962, p. 619.

Range.— From Espiritu Santo Island, Gulf of
California, to La Plata Island, Ecuador. Isabel,
Tres Marias, Revillagigedo, Clipperton, and
Galapagos Islands. Shore to 13 fathoms.

Material Examined— 112 specimens from 7
stations:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 22 males, 18 females (12683).

Acapulco, November 25-29, 1937, intertidal,
1 male, 1 female (12684).

Port Guatulco, Station 195, D-15, December
6, 1937, diving in Wi fathoms, coral, 1 male,
1 female (12685).

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?), 16 males, 11 females, 5 young
(11830).

Port Culebra, January 24-31, 1938, intertidal
in coral, 14 males, 16 females (12686).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 male, 3 females
(12687).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 1 male, 1 female (12688).

Measurements— Males 2. 6-7.0 mm., nonovi-
gerous females 3. 0-5. 5 mm., ovigerous females
2. 8-7. 9 mm.

Breeding— Ovigerous females from Sihuaten-

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

69

ejo Bay and Acapulco in November, from Port
Guatulco in December, from Port Parker and
Port Culebra in January, from Gulf of Nicoya
in February, and from Uvita Bay in March.

Remarks.— The specimens from Port Parker
were recorded earlier by Haig (1960, p. 315).
The species was collected in the Gulf of Cali-
fornia during the 1936 Templeton Crocker Ex-
pedition.

Pachycheles vicarius Nobili

Pachycheles vicarius Nobili, 1901, p. 19. Haig,

1960, pp. 134, 147, pi. 32 fig. 2.

Range.— From Acajutla, El Salvador, to Santa
Elena Bay, Ecuador. Shore to 4 fathoms.

Material Examined.— A3 specimens from 5
stations:

Costa Rica

Port Parker, January 12-23, 1938, intertidal
(in coral?), 7 males, 10 females, 1 young
(11832); 1 male, 1 female (AHF).

Port Culebra, January 24-31, 1938, intertidal
in coral, 3 males, 3 females (12689).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 male, 1 female
(12690).

Uvita Bay, March 2-4, 1938, intertidal in
coral, 9 males, 5 females (12691).

Panama

Bahia Honda, March 13-19, 1938, intertidal,
from Pocillopora coral, 1 male (12692).

Measurements.— Males 2. 8-6. 4 mm., nonovi-
gerous females 3. 2-5. 2 mm., ovigerous females
3. 8-6. 2 mm.

Breeding. — Ovigerous females from Port
Parker and Port Culebra in January, from Gulf
of Nicoya in February, and from Uvita Bay in
March.

Remarks.— The Port Parker specimens were
recorded and one of them illustrated by Haig
(1960, p. 317, pi. 32 fig. 2).

Pachycheles spinidactylus Haig

Pachycheles spinidactylus Haig, 1957a, p. 31, pi.

7 figs. 1-4; 1960, pp. 134, 153, pi. 33 fig. 2.

Range.— From Santa Maria Bay, outer Baja
California, and Cape San Tucas, Gulf of Cali-
fornia, to Port Utria, Colombia. Isabel Island.
Shore to 4 fathoms.

Material Examined.— 16 specimens from 3
stations:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 6 males, 7 females (12693).

Costa Rica

Port Culebra, January 24-31, 1938, intertidal
in coral, 1 female (12694).

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 male, 1 female
(12695).

Measurements.— Males 4. 9-7. 9 mm., oviger-
ous females 3. 5-8. 4 mm.

Breeding.— Ovigerous females from Sihuaten-
ejo Bay in November, from Port Culebra in Jan-
uary, and from Gulf of Nicoya in February.

Pachycheles panamensis Faxon

Pachycheles panamensis Faxon, 1893, p. 175.

Haig, 1960, pp. 134, 155, pi. 33 fig. 1; 1962,

p. 182.

Pachycheles sonorensis Glassell, 1936, p. 291.

Range— From Tiburon Island, Gulf of Cali-
fornia, to Santa Elena Bay, Ecuador. Isabel
Island. Shore to 4 fathoms.

Material Examined.— 58 specimens from 4
stations :

Mexico

Tenacatita Bay, November 20, 1937, inter-
tidal, 1 male, 1 female (12696).

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 25 males, 27 females, 2 young
(12697).

Port Guatulco, Station 195, D-14, December
6, 1937, 4 fathoms, coral, 1 young female
(12698).

Costa Rica

Uvita Bay, March 2-4, 1938, intertidal in
coral, 1 female (12699).

Measurements.— Males 3. 2-8.0 mm., nonovi-
gerous female 9 mm., ovigerous females 3. 6-9.1
mm.

Breeding.— Ovigerous females from Tenaca-
tita and Sihuatenejo Bays in November and
from Uvita Bay in March.

Remarks.— Material of this species was col-
lected in the Gulf of California during the 1936
Templeton Crocker Expedition and reported by
Glassell (1937) as Pachycheles sonorensis.

Pachycheles trichotus Haig

Pachycheles trichotus Haig, 1960, pp. 134, 157,

pi.’ 12, pi. 32 fig. 3.

Range.— Known only from Acajutla, El Sal-
vador, and Isla Verde, Panama. Probably inter-
tidal.

Material Examined.— Corinto, Nicaragua, De-
cember 28, 1937-January 7, 1938, intertidal, 1
male ( 12700) .

70

Zoologica: New York Zoological Society

[53: 2

Measurements— Male 3.8 mm. long, 4.1 mm.
wide.

Remarks— The single specimen collected by
the Zaca agrees very closely with the only speci-
mens previously known, the male holotype from
Isla Verde, Panama, and two ovigerous females
from Acajutla, El Salvador. The Corinto male is
smaller than the three types, which are nearly
identical in size:

Male holotype: 4.7 mm. long, 5.4 mm. wide.

Ovigerous female paratype: 4.7 mm. long, 5.3
mm. wide.

Ovigerous female paratype: 4.8 mm. long, 5.3
mm. wide.

Minyocerus kirki Glassell

Minyocerus kirki Glassell, 1938, p. 430, pi. 31.

Haig, 1960, p. 193, pi. 37 fig. 1, text-fig. 8;

1962, p. 185.

Range. — From Punta Penasco and San Felipe,
Gulf of California, to Realejo, Nicaragua. Shore
to 13 fathoms.

Material Examined— 12 specimens from 3
stations:

El Salvador

Cutuco, Gulf of Fonseca, December 21, 1937,
1 male, 3 females (AHF).

La Union, Gulf of Fonseca, Station 199, D-7
to D-16, December 27, 1937, 5-6 fathoms, 12
males, 12 females, 24 young (AHF).

Nicaragua

Monypenny Point, Gulf of Fonseca, Station
199, D-5 and D-6, December 24, 1937, 4-7
fathoms, 9 males, 11 females (AHF).

Measurements— Males 2. 9-3. 8 mm., nonovi-
gerous females 3. 5-3.7 mm., ovigerous females
3. 1-5.5 mm.

Breeding— Ovigerous females from all three
localities in the Gulf of Fonseca.

Remarks— All the material listed above was
reported by Haig (1960, p. 334) and is in the
collections of the Allan Hancock Foundation.

Specimens of Minyocerus kirki have been re-
ported living as commensals with sea stars,
Luidia Columbia (Gray) and Luidia phragma
H. L. Clark. Of specimens collected by the Zaca,
those from Cutuco were “around mouth of sea
star” and those from La Union and Monypenny
Point “on serpent stars and sea stars.” As I have
already noted (Haig, 1960, pp. 195 and 196),
the sea star was probably Luidia foliolata
Grube and the serpent star either Amphipholis
platydisca Nielsen, Ophiothrix spiculata Le-
conte, or Ophiolepis grisea H. L. Clark.

Porcellana cancrisocialis Glassell

Porcellana cancrisocialis Glassell, 1936, p. 292.

Haig, 1960, pp. 198, 200, pi. 38 fig. 2, text-fig.

9(2); 1962, p. 187.

Range— From Santa Maria Bay and Point
Tosco, outer Baja California, and Punta Pen-
asco, Gulf of California, to Santa Elena Bay,
Ecuador. Isabel Island. Shore to 54 fathoms.

Material Examined.— 14 specimens from 5
stations :

Mexico

Tenacatita Bay, Station 183, D-2, November
21, 1937, 30 fathoms, muddy sand, 1 male, 1
female (12701).

17 miles southeast by east of Acapulco, Sta-
tion 189, D-l, November 29, 1937, 20 fathoms,
sandy mud, algae, 2 males, 1 female, 1 young
(12702).

Port Guatulco, Station 195, D-ll, December
6, 1937, 5 fathoms, gray sand, crushed shell, 1
male (12703).

Costa Rica

Port Parker, Station 203, D-2, January 20,
1938, 10 fathoms, shelly sand, algae, 2 females
(12704).

Port Culebra, Station 206, D-l, January 30,
1938, 14 fathoms, sandy mud, 2 males, 3 females
(12705).

Measurements.— Males 3. 7-6.9 mm., nonovi-
gerous females 3.7 and 4.1 mm., ovigerous fe-
males 4.5-8. 6 mm.

Breeding— Ovigerous females from Acapulco
in November and from Port Parker and Port
Culebra in January.

Remarks.— Specimens of this species have fre-
quently been found living in association with
hermit crabs, but there is no evidence that this
was the case as far as the Zaca material is con-
cerned. The species was collected in the Gulf of
California by the 1936 Templeton Crocker Ex-
pedition (Glassell, 1937).

Porcellana paguriconviva Glassell
Porcellana paguriconviva Glassell, 1936, p. 293.

Haig, 1960, pp. 198, 203, pi. 38 fig. 1, text-

fig. 9(3); 1962, p. 185.

Range.— From Magdalena Bay, outer Baja
California, and Punta Penasco, Gulf of Cali-
fornia, to Taboga and Taboguilla Islands, Pan-
ama. Shore to 50 fathoms.

Material Examined.— 9 specimens from 4 sta-
tions :

Mexico

Port Guatulco, Station 195, D-2, December
4, 1937, 3 fathoms, sand, 1 male, 1 female
(12706).

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

71

Costa Rica

Port Parker, Station 203, D-2, January 20,
1938, 10 fathoms, shelly sand, algae, 1 female
(12707); D-13, January 22, 1938, 7-9 fathoms,
shells, algae, 1 male (12708).

Port Culebra, January 24-31, 1938, intertidal,
2 males, 1 female (12709).

Panama

Bahia Honda, Station 222, D-2, March 18,
1938, 4-8 fathoms, rocks, dead coral, 1 male, 1
young female (12710).

Measurements— Males 3.2-8. 1 mm., nonovi-
gerous female 3.0 mm., ovigerous females 7.3-
8.2 mm.

Breeding. — Ovigerous females from Port
Guatulco in December and from Port Parker
and Port Culebra in January.

Remarks— According to an accompanying
note, the specimens from Port Culebra were
found “on body of giant hermit inside shell.”
The “giant hermit” may have been Petrochirus
californiensis Bouvier, a large species with which
the types of Porcellana paguriconviva were asso-
ciated. However, the identity of the hermit crab
cannot be confirmed at this time because the
pagurids from the 1937-1938 Zaca Expedition
have not been located, according to Jocelyn
Crane.

Porcellana paguriconviva was collected in the
Gulf of California by the 1936 Templeton
Crocker Expedition (Glassell, 1937).

Pisidia magdalenensis (Glassell)

Porcellana magdalenensis Glassell, 1936, p. 295.
Pisidia magdalenensis, Haig, 1960, p. 209, pi. 38

fig. 4, text-fig. 10; 1962, p. 187.

Range— From Santa Maria Bay, outer Baja
California, and Petatlan Bay, Mexico, to Santa
Elena Bay, Ecuador. Shore to 25 fathoms.

Material Examined .—87 specimens from 8
stations:

Mexico

Port Guatulco, Station 195, D-2, December 4,
1937, 3 fathoms, sand, 1 male, 2 females
(12711); D-ll, December 6, 1937, 5 fathoms,
gray sand, crushed shell, 1 male ( 12712) ; D- 16,
December 7, 1937, 10 fathoms, sand, 1 male
(AHF) .

Nicaragua

Corinto, Station 200, D-l, December 29,
1937, 6V2 fathoms, mangrove leaves, 1 female
(12713); D-6, December 29, 1937, 2Vi fath-
oms, mangrove leaves, 4 males, 1 female
(12714); D-14, January 5, 1938, 3 fathoms,
mangrove leaves, 1 male, 1 female (AHF); D-

27 to D-30, January 7, 1938, 3 fathoms, 18
males, 23 females (12715).

Costa Rica

Port Parker, Station 203, D-2, January 20,
1938, 10 fathoms, shelly sand, algae, 4 males, 1
female (12716); D-4, January 22, 1938, 7 fath-
oms, gravel, algae, 1 female (AHF); D-7, Jan-
uary 22, 1938, 9-5 fathoms, shells, algae, 2
males, 2 females (12717); D-13, January 22,
1938, 7-9 fathoms, shells, algae, 2 males, 3 fe-
males (12718).

Port Culebra, January 24-31, intertidal in
coral, 2 males (12719).

Port Culebra, Station 206, D-l, January 30,
1938, 14 fathoms, sandy mud, 1 female (12720) .

Cedro Island, Gulf of Nicoya, Station 213,
D-7 and D-9, February 13, 1938, 4-6 fathoms,
mud, sand, crushed shell, 4 males, 5 females
(12721).

Golfito, Gulf of Dulce, Station 218, D-5,
March 9, 1938, 6 fathoms, mangrove leaves,
mud, shells, 1 male, 3 females (12722).

Panama

Bahia Honda, Station 222, D-2, March 18,
1938, 4-8 fathoms, rocks, dead coral, 1 male, 1
female (12723).

Measurements— Males 2. 5-4. 4 mm., nonovi-
gerous females 2. 2-3.0 mm., ovigerous females
2. 2-4.0 mm.

Breeding— Ovigerous females from Port Gua-
tulco in December, from Corinto in late Decem-
ber and early January, from Port Parker and
Port Culebra in January, from Gulf of Nicoya in
February, and from Gulf of Dulce and Bahia
Honda in March.

Remarks.— A portion of the material from
Port Guatulco, Corinto, and Port Parker was
reported earlier by Haig (1960, pp. 338, 339).

Megalobrachium poeyi (Guerin)

Porcellana poeyi Guerin, 1855, pi. 2 fig. 4.
Megalobrachium poeyi, Benedict, 1901, p. 136,

pi. 3 fig. 8. Haig, 1960, pp. 213, 214, pi. 16

fig. 4, pi. 39 fig. 1; 1962, p. 188.

Range.— From Salinas Bay, Costa Rica, to San
Francisco near Panama City, Panama. Shore to
25 fathoms. Also occurs in western Atlantic.

Material Examined.— 3 specimens from 2 sta-
tions:

Costa Rica

Cedro Island, Gulf of Nicoya, February 12-
13 or 21-22, 1938, intertidal, 1 male (12724).

72

Zoologica: New York Zoological Society

[53: 2

Panama

Bahia Honda, March 13-19, 1938, low tide
under stones, 1 male, 1 female (12725).

Measurements.— Males 3.5 and 3.7 mm., non-
ovigerous female 5.9 mm.

Remarks.— Although widely distributed in the
Caribbean area, this species appears to be con-
fined to Costa Rica and Panama on the Pacific
coast, where only 13 specimens are reported, in-
cluding the three cited above.

Megalobrachium garth! Haig

Megalobrachium garthi Haig, 1957a, p. 39, pi.

10 figs. 1-5; 1960, pp. 213, 220, pi. 16 fig. 7,

pi. 39 fig. 4.

Range.— From Turner Island, Gulf of Cali-
fornia, to Port Utria, Colombia. Tres Marias
Islands. Shore to 4 fathoms.

Material Examined.— 8 specimens from 4 sta-
tions:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 3 males, 1 female (12726).

Port Guatulco, Station 195, D-15, December
6, 1937, diving in IV2 fathoms, coral, 2 males
(12727).

Costa Rica

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 1 female ( 12728) .

Uvita Bay, March 2-4, 1938,’ intertidal in
coral, 1 female (12729).

Measurements.— Males 3. 9-6.0 mm., nonovi-
gerous female 2.8 mm., ovigerous females 5.0
and 5.1 mm.

Breeding— Ovigerous females from Sihuat-
enejo Bay in November and from Gulf of Nic-
oya in February.

Remarks.— The specimens from Costa Rica
are the first to be reported from the wide geo-
graphical area between Tangola-Tangola Bay,
Mexico, and Secas Islands, Panama.

Megalobrachium festai (Nobili)
Porcellanides festae Nobili, 1901, p. 21.
Megalobrachium festai. Haig, 1960, pp. 213,

226, pi. 16 fig. 10, pi. 40 fig. 3.

Range— From Acapulco, Mexico, to Santa
Elena Bay, Ecuador. Shore to 4 fathoms.

Material Examined— Sihuatenejo Bay, Mex-
ico, November 24, 1937, intertidal in coral, 4
males, 4 females (12730).

Measurements.— Males 2. 0-3. 8 mm., nonovi-
gerous females 3. 5-4. 3 mm., ovigerous female
2.2 mm.

Remarks.— Until now only 49 specimens of

Megalobrachium festai have been recorded, 41
of which were taken from sponges dredged off
Acapulco by the Velero IV. Material was col-
lected at three localities in Mexico and one each
in El Salvador and Ecuador.

The known range of the species is now ex-
tended northwestward from Acapulco to Sihuat-
enejo Bay.

Megalobrachium tuberculipes (Lockington)

Pachycheles tuberculipes Lockington, 1878, pp.

396, 404.

Megalobrachium tuberculipes, Haig, 1960, pp.

213, 227, pi. 16 fig. 1 1, pi. 40 fig. 4.

Range— From Punta Penasco and San Felipe,
Gulf of California, to Santa Elena Bay, Ecuador.
Shore to 10 fathoms.

Material Examined.— A specimens from 2 sta-
tions:

Mexico

Sihuatenejo Bay, November 24, 1937, inter-
tidal in coral, 1 male (12731).

Costa Rica

Jasper Island, Gulf of Nicoya, February 22-
25, 1938, intertidal in coral, 3 males (12732).

Measurements.— Males 2. 6-3.0 mm.

Remarks.— This species seems to be best
adapted for concealment in sponges. The speci-
mens collected at Gulf of Nicoya by the Zaca
Expedition are the first to be recorded between
Acapulco, Mexico, and Pearl Islands, Panama.

Polyonyx confinis Haig

Poly onyx confinis Haig, 1960, pp. 233, 234, pi.

17, text-fig. 12(3) .

Range— Known only from Corinto, Nicar-
agua.

Material Examined. — Corinto, Nicaragua,
Station 200, D-14, January 5, 1938, 3 fathoms,
mangrove leaves, male holotype (AHF 3817),
1 female paratype (AHF).

Measurements— Male holotype 2.7 mm. long
and 3.6 mm. wide, ovigerous female paratype
2.4 mm. long and 3.6 mm. wide.

Remarks.— This species, which is known only
from the two types collected by the Zaca, was
described by Haig ( 1960) . Both types are housed
in the Allan Hancock Foundation.

Acknowledgments

I wish to express my thanks to Jocelyn Crane,
Department of Tropical Research, New York
Zoological Society, for making the porcellanids
from this expedition available for study; and
to Dr. Dorothy E. Bliss and Dr. Arnold Ross,
Department of Living Invertebrates, The Amer-

1968]

Haig: Porcellanid Crabs (Crustacea: Anomura)

73

ican Museum of Natural History, who arranged

for the shipping of the collection to the Allan

Hancock Foundation and provided catalog num-
bers and other information.

Literature Cited

Benedict, J. E.

1901. The anomuran collections made by the
Fish Hawk expedition to Porto Rico. Bull.
U. S. Fish Comm., Pt. 2, 20:131-148, pis.
3-6.

Boone, Lee

1932. The littoral crustacean fauna of the Gala-
pagos Islands. Part II. Anomura. Zoo-
logica 14:1-62, text-figs. 1-19.

Chace, F. A., Ir.

1962. The non-brachyuran decapod crustaceans
of Clipperton Island. Proc. U. S. Natl.
Mus. 113:605-635, text-figs. 1-7.

Crane, Iocelyn

1947. Eastern Pacific expeditions of the New
York Zoological Society. XXXVIII. In-
tertidal brachygnathous crabs from the
west coast of tropical America with spe-
cial reference to ecology. Zoologica 32:
69-95, text-figs. 1-3.

Faxon, W.

1893. Reports on the dredging operations off
the west coast of Central America to the
Galapagos, to the west coast of Mexico,
and in the Gulf of California ... by the
U.S. Fish Commission steamer Albatross,
during 1891 . . . VI. Preliminary descrip-
tions of new species of Crustacea. Bull.
Mus. Compar. Zool. Harvard 24: 149-220.

Garth, I. S.

1959. Eastern Pacific expeditions of the New
York Zoological Society. XLIV. Non-in-
tertidal brachygnathous crabs from the
west coast of tropical America. Part 1:
Brachygnatha Oxyrhyncha. Zoologica
44:105-126, pi. 1, text-figs. 1-2.

1961. Eastern Pacific expeditions of the New
York Zoological Society. XLV. Non-in-
tertidal brachygnathous crabs from the
west coast of tropical America. Part 2:
Brachygnatha Brachyrhyncha. Zoologica
46:133-159, pi. 1, text-figs. 1-2.

1966. Eastern Pacific expeditions of the New
York Zoological Society. XLVI. Oxysto-
matous and allied crabs from the west
coast of tropical America. Zoologica 51:
1-16, text-figs. 1-2.

Gibbes, L. R.

1850. On the carcinological collections of the
United States, and an enumeration of
species contained in them, with notes on
the most remarkable, and descriptions of

new species. Proc. Amer. Assoc. Adv. Sci.
3:167-201.

Glassell, S. A.

1936. New porcellanids and pinnotherids from
tropical North American waters. Trans.
San Diego Soc. Nat. Hist. 8:277-304, pi.
21.

1937. The Templeton Crocker Expedition. IV.
Porcellanid crabs from the Gulf of Cali-
fornia. Zoologica 22:79-88, pi. 1.

1938. New and obscure decapod Crustacea from
the west American coasts. Trans. San
Diego Soc. Nat. Hist. 8:411-454, pis.
27-36.

1945. Four new species of North American
crabs of the genus Petrolisthes. Jour.
Wash. Acad. Sci. 35:223-229, text-figs. 1-4.

Guerin- Meneville, F. E.

1855. In R. de la Sagra, Historia fisica, politica
y natural de la isla de Cuba. Vol. 8, Atlas
de Zoologia. Crustaceos, Aragnides e In-
sectos. Paris. Pis. 1-20.

Haig, Janet

1956. The Galatheidea (Crustacea Anomura)
of the Allan Hancock Atlantic Expedition
with a review of the Porcellanidae of the
western north Atlantic. Allan Hancock
Atl. Exped. Rpt. 8:1-44, pi. 1.

1957a. Four new porcellain crabs from the east-
ern Pacific. Bull. South. Calif. Acad. Sci.
56:31-41, pis. 7-10.

1957b. The porcellanid crabs of the Askoy ex-
pedition to the Panama Bight. Amer. Mus.
Novitates ( 1865): 1-17.

1960. The Porcellanidae (Crustacea Anomura)
of the eastern Pacific. Allan Hancock Pac.
Exped. 24:1-440, pis. 1-41, text-figs. 1-12.

1962. Papers from Dr. Th. Mortensen’s Pacific
expedition 1914-1916. LXXIX. Porcel-
lanid crabs from eastern and western
America. Vidensk. Medd. fra Dansk
Naturh. Foren., 124:171-192, text-figs.
1-5.

Lockington, W. N.

1878. Remarks upon the Porcellanidea of the
west coast of North America. Ann. and
Mag. Nat. Hist. Ser. 5, 2:394-406.

Milne Edwards, A.

1869. In A. G. L. de Folin and L. Perier, Les
fonds de la mer. Etude intemationale sur
les particularites nouvelles des regions
sous-marines 1:128-130, pi. 26. Paris.

Nobili, G.

1901. Viaggio del Dr. Enrico Festa nella re-
pubblica dell’Ecuador e regioni vicine.
Decapodi e stomatopodi. Bol. Mus. Zool.
Anat. Comp. Univ. Torino 16 (415): 1-58.

74

Zoologica: New York Zoological Society

[53: 2: 1968]

Saussure, H. de

1853. Description de quelques crustaces nou-
veaux de la cote occidentale du Mexique.
Rev. et Mag. de Zool. Ser. 2, 5:354-368,
pis. 12-13.

Stimpson, W.

1858. Prodromus descriptionis animalium ev-
ertebratorum . . . Pars VII. Crustacea

Anomura. Proc. Acad. Nat. Sci. Phila.
10:225-252.

1859. Notes on North American Crustacea, no.
1. Ann. Lyceum Nat. Hist. New York
7:49-93, pi. 1.

Streets, T. H.

1871. Descriptions of five new species of Crus-
tacea from Mexico. Proc. Acad. Nat. Sci.
Phila. 23:225-227, pi. 2.

NEW YORK ZOOLOGICAL SOCIETY

GENERAL OFFICE PUBLICATION OFFICE

630 Fifth Avenue, New York, N.Y. 10020 The Zoological Park, Bronx, N.Y. 10460

OFFICERS

Fairfield Osborn

Laurance S. Rockefeller

Eben W. Pyne

President

First Vice-President

Assistant Treasurer

Howard Phipps, Jr.

Henry Clay Frick, II

Secretary

Robert G. Goelet
Vice-President

Edward R. Ricciuti Elysbeth H. Wyckoff

Editor & Curator, Production Editor

Publications and Public Relations

EDITORIAL COMMITTEE
Fairfield Osborn
Chairman

William G. Conway Donald R. Griffin

Lee S. Crandall Hugh B. House

Robert G. Goelet

William G. Conway
General Director

ZOOLOGICAL PARK

F. Wayne King
Peter R. Marler
Ross F. Nigrelli

William G. Conway Director & Curator, Birds Charles P. Gandal Veterinarian

Hugh B. House Curator, Mammals Lee S. Crandall . . . General Curator Emeritus

Grace Davall . . Assistant Curator, Mammals & Zoological Park Consultant

& Birds William Bridges Curator of Publications Emeritus

Victor H. Hutchison . . . Research Associate in John M. Budinger . . . Consultant, Pathology

Herpetology David W. Nellis Mammalogist

Joseph Bell .... Associate Curator, Birds James G. Doherty Mammalogist

F. Wayne King Curator, Reptiles Donald F. Bruning Ornithologist

AQUARIUM

Ross F. Nigrelli Director Nixon Griffis .... Administrative Assistant

Christopher W. Coates . . . Director Emeritus Robert A. Morris Curator

Louis Mowbray Research Associate in Field Biology

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . . Director and Pathologist

Martin F. Stempien, Jr. . . . Assistant to the

Director & Bio-Organic Chemist
George D. Ruggieri, S.J. . . . Coordinator of

Research & Experimental Embryologist
William Antopol . . . Research Associate in

Comparative Pathology
C. M. Breder, Jr. . . . Research Associate in

Ichthyology

Jack T. Cecil Virologist

Harry A. Charipper . . Research Associate in

Histology

Kenneth Gold Marine Ecologist

Eva K. Hawkins Algologist

Myron Jacobs Neuroanatomist

Klaus Kallman Fish Geneticist

Vincent R. Liguori Microbiologist

John J. A. McLaughlin . Research Associate in

Planktonology

Martin P. Schreibman . . Research Associate in

Fish Endocrinology

INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin .... Director & Senior Fernando Nottebohm . . . Research Zoologist

Research Zoologist George Schaller Research Zoologist

Peter R. Marler . . . Senior Research Zoologist Thomas T. Struhsaker . . . Research Zoologist

Jocelyn Crane . . . Senior Research Ethologist C. Alan Lill Research Fellow

Roger S. Payne Research Zoologist O. Marcus Buchanan . . . Resident Director,

Richard L. Penney .... Research Zoologist William Beebe Tropical Research Station

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 53 • ISSUE 3 • FALL, 1968

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

6. Observations on the African Bushpig Potamochoerus porcus Linn, in
Rhodesia. By Lyle K. Sowls and Robert J. Phelps. Plates I-II; Text-

figures 1-8 75

7. The Breeding Biology of the Male Brown Bear ( Ursus arctos) . By Albert
W. Erickson, Harland W. Mossman, Richard J. Hensel, and Willard
A. Troyer. Plates I-IX; Text-figures 1-2 85

Manuscripts must conform with Style Manual for Biological Journals (American Institute
of Biological Sciences). All material must be typewritten, double-spaced. Erasable bond paper
or mimeograph bond paper should not be used. Please submit an original and one copy of the
manuscript.

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park, Bronx Park. Bronx, N. Y. 10460, and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers, $1.50, unless otherwise stated in the Society’s catalog of publications. Second-class
postage paid at Bronx, N. Y.

Published November 29, 1968

© 1968 New York Zoological Society. All rights reserved.

6

Observations on the African Bushpig
Potamochoerus porcus Linn, in Rhodesia

Lyle K. Sowls1 and Robert J. Phelps2
(Plates I-II; Text-figures 1-8)

Introduction

ALTHOUGH the African bushpig is wide-
ly distributed and locally abundant it is
one of the continent’s least known large
mammals. Its elusive habits and the fact that it
feeds mostly at night make it better known by
its trail of damaged crops than by its appearance.
Consequently, except for records on museum
specimens which are relatively scarce, very little
factual information has been published on this
mammal.

This paper is an attempt to bring together
most of the existing knowledge of the bushpig
and to add new information based largely on
experience with animals raised in captivity.

Methods of Study

In November, 1962, farmers in the Salisbury
area of Rhodesia were asked through radio and
newspapers to notify us of any young bushpigs
found in the area. Seventeen young bushpigs,
taken from six litters, were obtained in this way
and held in captivity for periods of 13 to 26
months. Of this group all but two survived and
flourished. At the beginning these animals were
weighed and measured and their teeth examined
at weekly intervals. As they became larger they
were weighed and measured only once monthly.

'Arizona Cooperative Wildlife Research Unit, Univer-
sity of Arizona, Tucson, Arizona.

^Agricultural Research Council of Central Africa, Pax
House, Salisbury, Rhodesia.

To obtain skulls from pigs of known age the
animals were sacrificed at various ages between
1 3 and 26 months.

Several farmers in the Salisbury, Mazoe, and
Concession areas, where there are relatively high
bushpig populations, cooperated by gathering
weights and measurements of bushpigs killed.
One hundred and thirty-seven skulls in the Bula-
wayo Museum were examined to obtain infor-
mation on tooth eruption, presence or absence
of premolars, and reliability of the extension of
the maxilla for sex determination.

Findings

Dentition and Replacement

Normal Dentition.— The total number of teeth
for a normal bushpig can vary from 40 to 44
according to the following formula:

Incisors = 3 ; Canines l ; Premolars 3 or 4 ;

T T 1 or 4

Molars 3 X 2 = 40, 42 or 44

T

Ninety-two skulls of adults were examined. Of
this number, 73, or 79.3 percent, had 42 teeth;
16, or 17.4 percent, had 40 teeth; and only 3, or
3.2 percent, had 44 teeth.

For purposes of our study we numbered the
teeth as shown in Text-figure 1. This drawing
represents the most common situation where the
upper first premolars, but not the lower, are
present.

The tusks or canine teeth of the bushpig are

75

76

Zoologica: New York Zoological Society

[53: 3

Text-fig. 1. Normal dentition of the African bushpig and system of labeling individual teeth.

not nearly as long and conspicuous as those of
the warthog (Phacochoerus aethiopicus). The
lower canine is generally larger than the upper
and is sharpened by wearing action against the
upper canine.

Measurements were obtained of both upper
and lower tusks of both males and females.
Upper tusks of 23 males averaged 20.4 mm. and
of 14 females 17.4 mm. These figures were
(t = 1 .41 Tab. t. 05 for 35 d. f. = 2.03) not
significantly different. However, the lower tusks
of 29 males which averaged 47.5 mm. in length
were found to be significantly longer than 16
lower tusks of females which averaged 41 .4 mm.
in length (t=5.79, Tab. t .01 for 43 d. f.=2.69) .

Order of Eruption and Replacement of Teeth, —
At birth the bushpig normally has all four tem-
porary canines and the upper and lower third
incisors. The third premolar is the next tempo-
rary tooth to erupt in both the upper and lower
jaw. The first incisor in both jaws follows at
nearly the same time as the fourth temporary
premolar. The second premolar follows this. The
last temporary tooth to erupt is the second in-
cisor. By 15 to 17 weeks the temporary denti-
tion is complete. At this time the first permanent
tooth to erupt is the first lower molar, followed

shortly by the first upper molar. At about 43 to
45 weeks the permanent canines begin to ap-
pear. The upper canines appear about two weeks
ahead of the lower. The ages at which the various
teeth first appeared in the 15 captives are sum-
marized in Table I.

Weights and Measurements

Only meager information on the size of wild
bushpigs has been published. A review of the
weight and body measurement data found in the
literature is summarized in Table II. Table III
gives the weights and standard measurements for
ten wild bushpigs examined in Rhodesia during

1962 and 1963. The field-dressed weight is that
of the head, skin, and body after heart, lungs,
and other viscera have been removed.

Growth

No data have been published on the growth
rate of wild, free-ranging bushpigs. In 1962 and

1963 we were able to raise 15 bushpigs from a
few weeks of age, when they were taken from
the nest. Weights and standard measurements
were taken at periodic intervals. These animals
were given canned milk and water when very
young and slowly transferred to a commercial-

1968]

Sowls and Phelps: Observations on the African Bushpig

77

Table I

Tooth Development of the African Bushpig as Determined
by Periodic Examinations of Captive Animals

Temporary Teeth
( Age at Tooth Eruption in Weeks)

No. Animals Range Mean

Permanent Teeth
( Age at Tooth Eruption in Weeks)

No. Animals Range Mean

Upper Jaw

Incisors, 1st

9

7-9

7.7

2

71-75

73.0

2nd

10

11-17

13.8

1

95

3rd

In at

birth

Canines

In at

birth

12

36-49

43.6

Premolars, 1st

No temporary tooth

13

30-41

33.1

2nd

12

9-15

11.8

3

71-78

74.4

3rd

9

4-6

5.0

3

4th

9

8-12

9.8

3

Molars, 1st

No temporary tooth

13

21-26

24.5

2nd

No temporary tooth

5

56-68

61.0

3rd

No temporary tooth

Lower Jaw

Incisors, 1st

9

3-7

5.5

2

72-75

73.5

2nd

10

9-14

10.8

1

88

3rd

In at

birth

Canines

In at

birth

10

45-53

47.8

Premolars, 1st

No temporary tooth

none

9

?

2nd

7

11-16

12.8

3

71-78

74.4

3rd

9

4-6

5.0

3

4th

8

8-12

9.6

3

Molars, 1st

No temporary tooth

14

20-26

22.8

2nd

No temporary tooth

4

56-68

63.0

3rd

No temporary tooth

Table II

Weights and Body Measurements of Bushpigs as Described in the Literature

Area of
Collection

A pprox.
Age

Sex

Weight

(lbs.)

Head

&

Body

Shoulder Hind
Height Foot

Ear

Tail

Reference

S. Africa

Adult

M

50.5

10.5

6.0

14.7

Shortridge (1934)

Kenya

M

51.0

10.5

5.5

14.0

Zambia

F

171

47.4

25.5 10.2

6.8

15.5

Ansell (collector),

Bulawayo Museum

S. Africa

M

51.1

8.9

12.4

Roberts (1951)

M

250

Shortridge (1934)

F

200

235

Shortridge 1934 from Kirby

Zambia

M

175

Robinette ( 1963) from Ansell

F

135

F

122

Robinette (1963) from Benson

type hog feed given twice daily. Text-figure 2
shows the rate of gain of males and Text-figure 3
gives the rate of gain of the females. An approxi-
mate rate of growth is represented by the curved
line which has been drawn by inspection.

When they were first obtained animals were
kept in small pens about ten feet square, at a
rate of four or five animals per pen.

At about four months of age all 15 bushpigs
were placed in one large pen with a total area of
about 900 square feet, of which about 30 percent
was sheltered.

Growth curves were not plotted beyond the
50-week mark. Several factors made the data
after this age less reliable. As the animals were
sacrificed the number of values obtained for the

78

Zoologica: New York Zoological Society

[53: 3

o

O

w

Q

O

a

Pi

S

o

«

&.

a.

S

in

u

PQ

a

d

« £

;*

Q

O

03

D

Z

w

72 o
.5 o
53 ^

53

TJ >.
ca -o

OJ O

53 PQ

>S 9)

^ k|

■«

-i -c -

^ bo
~~ ^ C-

> » 4-

£ id '

H

O O

V. <50

Ci. ^

^ C

o' -2

? o
2; -2
7 ^

<o

u O a

-- w (L>

o 8 E

73 o «
"O O o
c ^ d
2^2
« cd c3

q££

St

O

00

r4 vo ^ ^ Tt

hiriroh

in in

O © — < o VO

vo ©
M (N ro

VO <N »/N •— <
^ ^ in ^ <n in

^ O O i- m oo Tt
N (S N M N ^ CM

t|- ^ >n o h oo
*-* oo <N on i— i

lo *0 h- \c (N oo
>o o ^ ^ n t on

[i. n. S n. 5. S Ph

3

T3

<

a .2

.2 8
90 T3
C/3 ^

<D O
<-> M
C r£
O w

u

S3

T3

<

in (N

in vo

S£

=3

T3

<

3

JD C3

.S2 ‘w

’— - a>

<3 70

CO o

u JO

WEIGHT IN POUNDS WEIGHT IN POUNDS

1968]

Sowls and Phelps: Observations on the African Bushpig

79

170 r-

150 -
130 -

'/.■

■ ./

/: " ' ■
• • / •

MALES

S

_l I I 1 I I 1 L_

-J 1 I 1 L.

0 10 20 30 40 50 60 70 80 90

AGE IN WEEKS

Text-fig. 2. Rate of gain in weight of cap-
tive male bushpigs.

i i i i i i i i i i i i i i i i i i

10 20 30 40 50 60 70 80 90

AGE IN WEEKS

Text-fig. 4. Rate of increase of head and
body length, tail length, and ear length in
captive male bushpigs.

tive female bushpigs.

various ages became fewer and fewer; of the 15
animals obtained only four were females; the
effects of the experimentation with trypanosomi-
asis were not known. Although the bushpig is a
highly social animal and individuals appeared to
be compatible, unequal gain in weight and gen-
eral growth were probably caused by crowding,
especially when the animals were over one year
of age.

Text-figures 4 and 5 show the rate of increase
in length of ear, tail, and length of head and body
and Text-figures 6 and 7 give the rate of increase
in hind foot length and shoulder height.

Both sexes showed a rather uniform rate of

Text-fig. 5. Rate of increase of head and
body length, tail length, and ear length in
captive female bushpigs.

growth during the first year of life and gained
weight at the rate of about two pounds per week.

Some indication of the meaning of the growth
rates can be obtained by comparing them with
the weights and measurements of wild animals.
Old adults (Tables II and III) have been found
to vary in weight by about 40 pounds. Two wild
animals determined to be yearlings by the tooth
eruption pattern weighed nearly the same as cap-
tive animals at the age of one year.

The increase in head and body measurements
formed a steep curve until about 40 weeks. Com-

LENGTH IN INCHES LENGTH IN INCHES

80

Zoologica: New York Zoological Society

[53: 3

.!/■ ■ ■ SHOULDER HEIGHT

:/

MALES

T

iiii i i i i i i i — i — i — i — i — i — i

10 20 30 40 50 60 70 80 90

AGE IN WEEKS

Text-fig. 6. Rate of increase of height of
shoulder and length of hind foot in cap-
tive male bushpigs.

the captive animals and showed the least varia-
tion of all measurements.

The length of the tail among wild specimens
was found to vary by about 30 percent. The
flattening of the curve at about 50 weeks for the
small number of captive individuals on which
data were obtained apparently means that these
particular individuals had already reached adult
tail length.

Lens Weight as an Index to Age.— Among the
Suidae and related forms, lens weight as an index
to age has been studied in the warthog Phacochoe-
rus aethiopicus by Child, Sowls, and Richardson
(1965) and the collared peccary Pecari tajacu
by Richardson (1966).

Data on the lens weights of the bushpig are
scarce. However, we are showing the weight-age
relationship for the 12 sets of lenses from the
captive bushpigs in Text-figure 8. The lens
weight plotted is the average for the two lenses
which were oven-dried and weighed to the near-
est milligram.

Text-fig. 7. Rate of increase of height of
shoulder and length of hind foot in cap-
tive female bushpigs.

parisons with the figures on wild animals indicate
that this measurement continues until much later
in life than the 90 weeks which is represented by
the oldest animals in Text-figures 4 and 5.

Shoulder height in both sexes began to decline
in rate of increase among captive animals at
about 30 weeks, and like the length of head and
body did not show a definite flat curve at the 90
weeks at which the oldest animals were sacrificed
in these investigations. Like the length of head
and body the height at shoulder continues to in-
crease slightly after 90 weeks.

The length of hind foot and ear were the
earliest to show a leveling off of the curve among

Text-fig. 8. Lens weight-age relationship
in 1 1 specimens of bushpig.

Determining Sex from the Skull.— The male
bushpigs can be distinguished from the female
by a dorsal extension of the maxilla which ex-
tends from the base of the canine (PI. I, Figs.
1 & 2). This process has been described by
Roberts (1951), Shortridge (1934), and other
writers. Of the 92 adult skulls examined in this
study the record described all skulls with the
process as males and all without it as females ex-
cept one, which we suspect was incorrectly la-
beled.

Reproduction

Breeding Season and Litter Size.— Only mea-
ger information on reproduction in bushpigs ap-

1968]

Sowls and Phelps: Observations on the African Bushpig

81

pears in the literature. Asdell (1964) gives the
gestation period as four months. Phillips ( 1926)
says that in South Africa there is no definite
breeding season and that litters generally num-
ber four. Fitzsimmons (1920), however, gives
midsummer (December and January) as the
period of parturition in South Africa. Shortridge
( 1934) quotes Kirby, who gave the same period
and says litters are made up of five to six young.
In Zambia, Ansell (1960a) says that records of
recently born young are from October to March
and that the litter sizes range from two to six
with three to four the most common.

Among the specimens of bushpigs in the col-
lection of the National Museum of Rhodesia at
Bulawayo are 14 very young animals as follows:
three from the same litter taken in Zambia on
December 21, 1959, which were apparently col-
lected when only a few weeks old; two from one
litter collected by W. F. H. Ansell on February
10, 1950; three from one litter collected by B. O.
Williams from Turk Mine in Rhodesia in July
1960, which according to the tooth pattern and
measurements had apparently been born in
March; three collected together at Nagupande,
Rhodesia, on November 25, 1962; one collected
near Salisbury on January 7, 1963, which was
estimated to be about two weeks old; one from
Zambia collected on January 25, 1961; and one
collected by W. F. H. Ansell on March 30, 1956.

The records on time of parturition and litter

size obtained in this study are given in Table IV.
Most of the farmers to whom we talked in the
Mazoe-Salisbury-Concession area believed that
most of the young bushpigs in that area were
born between late November and about Febru-
ary 1. We conclude that in Rhodesia the prin-
cipal parturition season is the same as Ansell has
described for Zambia with the largest number of
animals being born in November, December,
and January. This period is during the summer
rainy season when the food supplies and the
young animals’ chances for survival would be
the best. This same type of timing of the parturi-
tion season and rainfall has been described by
Sowls (1964) for another swinelike animal, the
collared peccary.

One female in the captive herd gave birth to
three young on November 30, 1964, at the age
of 103 weeks. The mating in this instance, based
on the four-month gestation period, would have
occurred about August 1. At this time the sow
was 86 weeks old.

The sow makes a large nest of grass for the
young. These nests remain for many months
in the heavy woodland and resemble a small
weathered haystack. Most of the young that were
captured for us were taken from the nests by
African farm hands. Ansell refers to these nests
as bowers.

Table V gives the weight and measurements
of five young from two litters that were obtained

Table IV

Birth

Dates and Litter Size for Individual

Litters

...

Litter

Reference or

Date of Birth

Size

Observer

Concession, Rhodesia

11-23-62

7

Keats

Mazoe, Rhodesia

11-23-62

?

Wheeler

12-2-62

4+

Edwards

12-10-62

4+

Bothma

1-9-63

?

Douglas

Table V

Weight and Measurements for Newly Born

Bushpigs

Date

Sex

Est.
Age in
Days

Weight

(Grams)

Head

&

Body

Flesh measurements mm.

Hind Shoulder

Foot Ht. Tail

Ear

11-27-62

F

4 days

710

260

78

178

94

45

M

880

260

74

174

96

51

12-6-62

M

877

242

72

162

102

46

F

849

232

72

165

99

50

M

679

230

67

148

90

40

82

Zoologica: New York Zoological Society

[53: 3

by us when about four days old. At birth the
young have longitudinal yellowish or white
stripes on a brown background (see PI. II, Fig.
3). The color of individual litters varies con-
siderably. Some pigs are much darker than
others and have less distinct stripes. The stripes
slowly disappear as the animals grow and the
coat becomes reddish. Little evidence of stripes
remains after 24 weeks of age.

Behavior

No detailed studies on the behavior of the
bushpig have been made. Roberts (1951), Short-
ridge (1934), Phillips (1926), Ansell (1960b),
and other writers all describe the bushpig as a
gregarious animal. The herd size has been de-
scribed by various writers as containing from
four to 20 pigs.

Unlike the warthog, which is strictly diurnal,
the bushpigs feed almost entirely at night. Dur-
ing the daytime they seek refuge in the tall grass
and brush where they remain until darkness.
Ansell (1960b) says, however, that in remote
country they commonly move about in the day-
light when undisturbed.

In captivity young bushpigs become extremely
tame but tend to become dangerous as they reach
maturity. Among the captives which we raised
we found that injured animals were persecuted
by their pen mates; the injured pigs had to be
separated from the others when cuts or open
sores appeared on their bodies.

Bushpigs and Trypanosomiasis

No detailed information is available on the
species of trypanosomes that may be carried by
bushpigs. Some of the animals kept in the cur-
rent study were infected with a trypanosome,
fatal to domestic pigs, by inoculation of blood
from a heavily infected domestic pig. Not all the
bushpigs became positive for trypanosomes and
those that did apparently were not affected seri-
ously. The trypanosomes could no longer be de-
tected in blood smears taken a few weeks after
initial infection. This information was given to
the authors in a personal communication from
P. McKenzie.

Bushpigs infected with trypanosomes, even if
only for a short time, act as carriers of the
disease, and the frequency with which they are
fed on by tsetse flies determines the importance
of the animals in the epidemiology of trypanoso-
miasis. In Rhodesia, Glossina morsitans orientalis
Vanderplank and Glossina pallidipes Aust are
the major tsetse species. G. morsitans orientalis
falls into the group feeding mainly on suids and
bovids (Weitz, 1962), based on information de-
rived from identification of blood meals obtained

from flies in the field. Available information in-
dicates that suids provide 36.1 percent of the
meals of G. morsitans orientalis, and of the total
suid blood meals 14 percent are from the bush-
pig (Weitz, 1962). G. pallidipes falls into the
group classed as feeders mainly on bovids
(Weitz, 1962). Suid meals comprise 29.9 per-
cent of the total meals taken by this fly, and of
the suid meals 41 percent are from the bushpig
(Weitz, 1962). The importance of bushpigs in
the diet of tsetse in Rhodesia is established by
this data, and in other parts of Africa there are
tsetse species which are more partial to bushpig
blood.

All the tsetse infested country in Rhodesia
falls in the range of distribution of the bushpig
(Smithers, 1966), and the ranges of the two
major tsetse species are almost identical. Control
of tsetse by elimination of preferred host species
is attempted in some areas of Rhodesia, and
bushpigs are one of the animals hunted.

In January, 1965, and April, 1964, four of the
captive animals were used for experimental work
with trypanosomiasis. Blood samples from bush-
pigs inoculated with trypanosomes revealed short
periods of infection from which the animals
recovered, and weight curves did not appear to
be affected by the treatment. Further details on
this phase of study will be reported by Phelps
and Roth.

Food Habits

No detailed study of the foods taken by the
bushpig has been made. Fitzsimmons (1920)
says that in South Africa they take roots, bulbs,
and fruits. Shortridge (1934) agrees but adds
that they devour reptiles, eggs, and birds and
tells of one instance where the carcass of a bush-
buck was eaten by bushpigs. Roberts (1951)
says that the normal diet of the bushpig is roots
and edible vegetable matter. Phillips (1926) gives
a rather extensive list of plants eaten by bushpigs
in South Africa. He lists seven species of ferns,
nine species of monocotyledons, eight species of
dicotyledons, and 26 species under forest tree
fruits.

In agricultural areas surrounded by brushy
timber the bushpig makes domestic crops a large
part of his diet. Maize, groundnuts, and field
peas are heavily taken by bushpigs in Rhodesia
as the fruits become mature.

Control of Bushpigs

Because of their fondness for agricultural
crops they are classed as a major pest in the maize
growing area of Rhodesia. In these areas farmers
find it necessary to control the bushpig by what-
ever means possible. The animals destroy far

1968]

Sowls and Phelps: Observations on the African Bushpig

83

more of the crop than they can eat (Roberts,
1951), and being largely nocturnal in habit, they
are not easy to hunt by conventional methods.
The most successful method of hunting the ani-
mals is the use of dogs to track them to their
lairs which are generally in dense thickets. Even
when located, the bushpig is not easy to kill, and
can be a formidable adversary. Shotguns are the
most useful weapon in hunting the animals, the
range at which shooting is done being only a
matter of yards. Trapping has not been found to
be very effective against bushpigs (Roberts,
1951), and once a trap has been located the
animals will make wide detours around it. Bush-
pigs will sometimes eat carrion (Smithers, 1966)
and large animal carcasses are sometimes poi-
soned in areas infested with bushpigs. The ani-
mals may occur very close to settled areas, pro-
vided some thick cover remains in gullies or on
hill slopes. In such areas enough manpower may
be available to organize drives, and this method
of hunting sometimes achieves a degree of suc-
cess.

Acknowledgments

We wish to thank the many people who helped
us with this study. We want to especially thank
Patrick Maratos of Concession who helped
gather data. Also William Keats and Richard
Peak of Concession, Mr. Wheeler of Mazoe, Mr.
Edwards of Salisbury, Mr. Newmarsh of Salis-
bury, and Mr. Bothma of Umtali who furnished
us with young bushpigs. We want to thank R. H.
N. Smithers, Director of the National Museum
of Rhodesia for his assistance. Also Graham
Child, Curator of Vertebrates of the National
Museum, who assisted us and allowed us to use
data on the museum collections of bushpigs. We
wish to also thank E. B. Edney, Head, Zoology
Department of the University College of Rho-
desia and Nyasaland, for furnishing space to
raise young bushpigs. We appreciate the help of
Moses Nyamuremba, Michael Katere, and Shad-
rack Mushambande who cared for the animals.
We wish to acknowledge the National Science
Foundation for financial assistance on part of
the project.

Literature Cited
Ansell, W. F. H.

1960a. The breeding of some larger mammals in
Northern Rhodesia. Proc. Zoo. Soc. Lon-
don, 134 (2 ) : 25 1-274.

1960b. Mammals of Northern Rhodesia. The
Government Printer, Lusaka.

Asdell, S. A.

1964. Patterns of mammalian reproduction.
Comstock Publishing Associates, Cornell
University, Ithaca, N. Y. 670 pp.

Child, G., L. K. Sowls, and G. L. Richardson

1965. Uses and limitations of eye-lens weight
for ageing wart hog. Arnoldia (Rhodesia)
Series of Misc. Pub. Nat. Mus. of South-
ern Rhodesia, 1 (39): 1-2.

Fitzsimmons, F. W.

1920. The natural history of South Africa. Vol.
III. Longmann, Green & Co., London.
278 pp.

Phillips, John F. V.

1926. “Wild pig” (Potamochoerus choeropota-
mas ) at the Knysna: notes by a natural-
ist. South African Jour, of Science,
23:655-660.

Richardson, Gary L.

1965. Eye lens weight as an indicator of age
in the collared peccary ( Pecari tajacu).
Unpub. M.S. thesis. University of Arizona,
Tucson. 38 pp.

Roberts, Austin

1951. The mammals of South Africa. The Mam-
mals of S. Africa Book Fund, Capetown,
Central News Agency. 700 pp.

Shortridge, G. C.

1934. The mammals of South West Africa. Wil-
liam Heinemann, Ltd., London. 779 pp.

Smithers, Reay H. N.

1966. The mammals of Rhodesia, Zambia and
Malawi. Collins, London. 159 pp.

Sowls, Lyle K.

1964. Reproduction in the collared peccary
( Pecari tajacu ). In Comp. Bio. Reprod. in
Mammals. The Academic Press, London
and New York. 155-172.

Weitz, B.

1962. The feeding habits of tsetse ( Glossina )
species. W.H.O. Report on Trypanosomi-
asis.

84

EXPLANATION OF THE PLATES

Plate I

Fig. 1. Frontal portion of female bushpig skull.
Fig. 2. Frontal portion of male bushpig skull
showing upward extension of maxilla.

Plate II

Fig. 3. Young bushpigs two days old.
Fig. 4. Adult male captive bushpig.

SOWLS & PHELPS

PLATE I

FIG. 1

FIG. 2

OBSERVATIONS ON THE AFRICAN BUSHPIG POTAMOCHOERUS PORCUS LINN. IN RHODESIA

SOWLS & PHELPS

PLATE II

FIG. 4

OBSERVATIONS ON THE AFRICAN BUSHPIG POTAMOCHOERUS PORCUS LINN. IN RHODESIA

SOWLS & PHELPS

PLATE II

FIG. 3

OBSERVATIONS ON THE AFRICAN BUSHPIG POTAMOCHOERUS PORCUS LINN. IN RHODESIA

7

The Breeding Biology of the Male Brown Bear (Ursus arctos)1’2

Albert W. Erickson,3 Harland W. Mossman,3 Richard J. Hensel,4 and Willard A. Troyer4

(Plates I-IX; Text-figures 1-2)

Introduction

The BREEDING BIOLOGY of the
brown bear is known in only a general
way. Breeding occurs in the spring, usu-
ally in late May or June, and its timing does not
appear to vary significantly between wild or cap-
tive animals or throughout the wide expanse of
the species distribution (Dathe, ’61; Dittrich and
Einsiedel, ’61; DeVoto, ’53; and Murie, ’44).
The female exhibits a period of heat extending
up to two weeks and is polygamous. During this
time coital activity is recurrent but is interrupted
by days of nonbreeding (Dittrich and Kron-
berger, ’63) .

The age of puberty is unknown in the male
bear, but among female captives is usually at-
tained at three and a half years. The gestation
period in captivity has been reported as varying
between 194 and 278 days (Dittrich and Kron-
berger, ’63) . Despite this disparity, a large body
of evidence shows whelping to occur regularly
in late January and early February regardless

Reference to the brown bear here refers collectively
to the various so-called species of North American
brown and grizzly bears, and to the European and Eur-
asian brown bears. Recent taxonomic reviews conclude
that all of these are simply subforms of Ursus arctos
L. (Pocock, ’32; Erdbrink, ’53; Couturier, ’54; and
Rausch, ’62).

Supported in part by U.S.P.H.S. training grant
5TI-GM-723-04.

3Anatomy Department, University of Wisconsin Med-
ical School, Madison 53706. Dr. Erickson’s present
address is James Ford Bell Museum of Natural History,
University of Minnesota, Minneapolis 55455.

4Kodiak National Wildlife Refuge, Kodiak, Alaska
99615. Mr. Troyer’s present address is Kenai National
Wildlife Refuge, U.S. Fish and Wildlife Service, Kenai,
Alaska.

of when breeding occurs. Explanation for this
is that bears of the genus Ursus have a delayed
implantation wherein the fertilized eggs develop
to the blastocyst stage and lie quiescent in the
uteri for a long period of time. Implantation oc-
curs about the same time in all specimens re-
gardless of when breeding occurs (Wimsatt, ’63;
and Dittrich and Kronberger, ’63). Normally
the delays last slightly over half of the gestation
period and macroscopic embryos are not visible
until about the time of winter denning. The cubs
are born in an immature state during the so-
called hibernation period. Litters vary from one
to four cubs but are usually two or three.

Beyond breeding observations, the only spe-
cific information known to us on the reproduc-
tive biology of the male bear is a report by Dit-
trich and Kronberger (’63) on the histology of
the testes and epididymides of two captive bears
killed in August and October, respectively. On
the basis of spermatogenic activity and epidi-
dymal sperm observed in both animals, they
concluded that male brown bears retain repro-
ductive capability at least through October.

Methods and Procedures

The testes, epididymides, and vasa deferentia
of 127 brown bears were collected in Alaska
between May 20, 1961, and November 1 1, 1964.
The majority of the specimens were from Ko-
diak Island, but specimens were obtained also
from other areas of the state, particularly from
the Alaska Peninsula (Table I). Most of the
bears were killed by sport hunters. Additional
specimens were obtained from bears killed as
nuisances or by unilateral castrations of live-
trapped bears.

85

86

Zoologica: New York Zoological Society

[53: 3

Table I Results of Examinations of Testes, Epididymides,

Testis Tubules

Location

Specimen

Date

Skull 4

Age-

Testis 3

Activity4

Number

Meas.

(Years)

Wt. (Gms.)

Diam. (/r)

State

Kodiak Is.

14N

4-7-64

19%

1.2e

9.7*

NF, G

36M

5-4-62

18.2

95

NF, G

12M

5-7-62

19%

2.3e

17.2

113

NF, G

8N

5-8-64

20%

2.3a

11.8

NF, G

47M

5-11-63

22%

2.3e

15.0*

NF, G

49M

5-13-62

21%

2.3e

19.0*

NF, G

5M

5-14-62

24.0

112

NF, G

57M

5-19-63

19%

1.3e

11.0

NF, G

26M

5-19-62

20%

2.3e

19.4

95

NF, G

43N

5-19-64

21%

2.3e

16.7*

NF, G

K44

5-20-62

2.3a

6.3*

86

NF, G

65M

5-21-63

19%

1.3e

10.0*

NF, G

Anchorage

E306

7-14-63

11.0*

83

NF, G

Kodiak Is.

40N

10-7-64

15.0*

81

NF, G

49N

10-7-64

10.2*

NF, G

71M

10-8-63

21%

2.7e

10.2*

NF, G

2M

10-10-62

20%

2.7e

16.0

75

NF, G

3M

10-10-62

20%

2.7e

13.9

94

NF, G

31A

10-11-61

233/io

2.7a

20.1

NF, G

4M

10-12-62

21.9

118

NF, G

67A

10-12-63

2013/1g

11.7

81

NF, G

32A

10-17-61

23i/io

2.8a

19.2

NF, G

K97

10-17-63

21%

2.8a

16.5*

NF, G

34MK

10-19-61

22.9

NF, G

75M

10-19-63

24.2*

NF, G

47N

10-22-64

20

2.8e

23.1*

109

NF, G

20M

10-25-62

15.8

123

NF, G

52N

10-29-64

2.8e

20.8*

97

NF, G

18M

10-31-62

21%

2.9e

16.3

1 14

NF, G

46N

11-2-64

19.0

1.9e

6.5*

54

NF, G

5 ON

1 1-2-64

23%

2.8k

15.8*

NF, G

5 IN

11-3-64

20

2.9e

13.2*

96

NF, G

8M

11-5-62

22%

2.9e

22.1

NF, G

10M

11-8-62

20%

2.9e

15.1

NF, G

a For explanation of number designations see Table II.

Table II

Results of

Examinations of Testes, Epididymides,

Testis Tubules

Location

Specimen

Date

Skull1 Age 2

T estis 3

Height of

Activity4

Number

Meas. Yrs.: Months \yt. (Gins.)

Diam. (ft)

Epithel (/i)

State

Kodiak Is.

36N

5-7-64

42.5*

SN

59M

5-11-63

36.0*

S, G

13N

5-18-64

21% 3.3e

33.4

124

NF, G

21M

5-22-62

22 3.3e

25.6

196

35

SN

56M

5-24-63

43.0*

173

39

FS, A

E042

5-25-61

24.5

30.8

187

S, G

Paxson

E258

9-16-62

20.0

18.8

123

32

NF, G

Kodiak Is.

62A

10-4-62

22'/ic 3.9e

31.9*

154

51

SN, G

32N

10-5-64

2215/iq 3.9e

46.0

A, FS

3 ON

10-7-64

24% 4.9e

34.6*

SN, G

72M

10-9-63

24 4.9e

27.3

NF, G

9M

10-10-62

33.9

180

38

SN, G

73M

10-14-63

29.1

125

57

S, G

74M

10-17-63

25% 4.9e

21.4

NF, G

Legend for Tables I, II, III

1 Skull measurements: Length (occipital protuberance to margin of incisor) + width (outer edges of zygomatic arches).

2 Age: k = known age marked animal; a = approximate known age marked animal; e = estimated age.

3 Testis weight: testis + epididymis + vas deferens. The weights marked with an asterisk are preserved weights plus 10% (the
mean weight loss between fresh and preserved specimens).

4 Spermatogenic activity: FS = free sperm; SN = sperm nuclei or heads in Sertoli cells; S = primary or secondary spermato-
cytes; A = abnormal forms shed into lumen; G = Edematous (giant) cells in germinal epithelium or in lumen. NF = No formed
elements other than giant cells.

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

87

and Vasa Deferentia of Infantile Brown Bears3

Intertubular Area Epididymis Vas Deferens

General9

Character

Cytoplasmic

Abundance0 Vacuolation7

Diam. (p)

Height of
Epithel (p)

Lumen8

Contents

Coagulum9

Cytoplasmic10

Droplets

Lumen8

Contents

Coagulum9

Cytoplasmic1

Droplets

FI

E

+ + +

E

+ +

4" 4" +

IF

+

A, H

178

48

E

+

+

E

N

+

IF

+

A, H

196

50

E

+

+

E

+ +

+ +

FI

+

A, H

150

47

E

+ + +

+

E

+ +

+ +

FI

+

A, H

E

+

+ +

E

+

+ +

IF

+

A, H

E

+

+

E

+

+

IF

+ 4“

A, H

271

62

E

+ +

+ +

E

+

+ 4"

FI

+

N

FI

+

A, H

123

36

E

+ + +

N

E

FI

+

A, H

E

+

+

E

+

+ +

FI

+

N

169

60

E

+ +

+

E

FI

+

N

E

E

FI

+

A, H

E

N

+

E

FI

+

A, H

176

51

E

N

N

E

N

N

FI

+

N

E

E

FI

+

N

E

E

I

+

A, H

180

46

E

+

N

FI

+

N

194

57

E

+

+

E

N

+

FI

+

A, H

E

N

N

E

+ +

+ +

IF

+

A, H

205

51

E

N

N

E

N

+

IF

+

A, H

206

E

+ +

N

E

+

N

IF

+ +

A, H

E

E

IF

+ +

A, H

E

E

IF

+

A, H

E

E

FI

+

A, H

159

55

E, A

+

+

IF

+

A, H

215

E

+ +

+

E, D

N

N

IF

+

A, H

122

33

A

+++

+

FI

+

A, H

173

37

E

N

+

E

N

+

FI

+ +

A, H

150

35

E

+ +

+

IF

+

A, H

E

+

+

E

N

N

IF

+

A, H

206

E

+ +

N

E

+ +

N

FI

+

A, H

E, D

N

+

FI

+'

A, H

E, D

N

N

and Vasa Deferentia of Prepuberal Brown Bears

Tntertubular Area Epididymis Vas Deferens

General5

Character

Cytoplasmic

Abundance0 Vacuolation7

Diam. (p)

Height of
Epithel (/i)

Lumen8

Contents

Coagulum9

Cytoplasmic11

Droplets

1 Lumen8

Contents

Coagulum9

Cytoplasmic10

Droplets

E

E

I

+

A, H

A, S

+ + +

+

E, A, S

+

4“ 4"

IF

+

A, H

E, A

+ +

N

E

+ +

+

FI

+

A, H

182

45

A

+

N

A

+

+ "b

I

+ +

A, H

224

68

S, A

+ + +

+ +

S, A

+ +

N

IF

+

A, H

194

A

FI

+

A, H

IF

+

A, H

224

57

E

+

+

E

N

+

IF

+

A, H

A, S

A, S

N

N

FI

+

L

159

47

A

+

+

A

N

+

IF

+

A, H

E

E

FI

+

L

248

66

E, A

+ +

+

E, A

+

N

IF

+ +

A, H

167

E

+

+ +

E, D

N

N

IF

A, H

5 General character

of intertubular tissue: FI

= more fibrous tissue than

interstitial tissue; IF = more interstitial tissue than

fibrous tissue; I = predominantly interstitial tissue.

0 Leydig cell cytoplasmic abundance: + = low; -f— (- = med.; | | | = high.

7 Leydig cell vacuolation: N = little or none; A = abundant small vacuoles; L = large and small vacuoles; H = vacuoles
highly vesicular (frothy).

8 Lumen contents: S = apparently viable sperm; A = immature and abnormal forms; D = degradation products of tract;
E = empty; several entries indicate differences between ducts in order of decreasing occurrence.

9 Prevalence of coagulum in epididymis and vas deferens: N = little or none; + = low; + + = medium; +++ = abundant.

10 Cytoplasmic extrusions of epididymis and vas deferens: N = little or none; + = low; +■)- = medium; + + + = abundant.

88

Zoological New York Zoological Society

[53: 3

Table III

Results of Examinations of Testes, Epididymides,

Location

Specimen

Number

Date

Skull 1
Me as.

Testis 3
W t. (Gms.)

Testis Tubules
Height of
Diam. (n) Epithel (/t)

Activity4

State

Kodiak Is.

41M

4-2-63

29

94*

285

95

FS

42M

4-18-63

26%

83*

271

solid

FS

22M

4-22-62

23%

49

187

solid

FS

42N

4-28-64

25%

58*

FS

34N

5-1-64

60*

FS

44M

5-2-63

90*

FS

16M

5-4-62

28%

98

276

76

FS

45M

5-4-63

28%

47*

FS

1M

5-4-62

28%

95

262

95

FS

33N

5-5-64

28%0

85*

FS

43M

5-5-63

28%

84*

FS

31N

5-6-64

27%

110*

FS

13M

5-7-62

26%

77

209

57

FS

31M

5-8-62

89

257

76

FS

46M

5-9-63

26%

46*

FS

28M

5-10-63

79

FS

33M

5-10-62

27%

97

259

83

FS

Alaska Pen.

E231

5-10-62

72 74

FS, A

Kodiak Is.

61M

5-11-63

28%

86*

FS

Alaska Pen.

E305

5-11-64

92 90

FS, A

1396

5-12-64

76* 80*

266

59

FS

Kodiak Is.

7M

5-13-62

76

268

76

FS, A

15M

5-14-62

27

70

218

83

FS

6M

5-15-62

29

100

228

66

FS

40M

5-15-62

27%

94

218

66

FS

48M

5-15-63

28%

77

FS

53M

5-15-63

28

121*

237

83

FS

9N

5-16-64

71*

FS

11M

5-16-62

67

FS

14M

5-16-62

29%

56

228

66

FS

Alaska Pen.

3093

5-16-64

52*

FS

Kodiak Is.

54M

5-16-63

27

65*

FS

Alaska Pen.

3094

5-16-64

45*

FS

Kodiak Is.

24M

5-17-62

22%

73

200

47

FS

34M

5-17-62

28%

91

180

59

FS

3N

5-18-64

28%

111

FS

55M

5-18-63

26%

48*

FS

Alaska Pen.

3099

5-18-64

70*

FS

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

89

and Vasa Deferentia of Sexually Mature Brown Bears3

Imerlubular Area Epididymis V as Deferens

General5

Cytoplasmic

Height of

Lumen8

Coagulum9

Cytoplasmic10

Lumen8

Cytoplasmic1

Character

Abundance6

Vacuolation1

Diam. (/r)

Epithel (t<.)

Contents

Prevalence

Droplets

Contents

Coagulum9

Droplets

I

+ + +

L

268

64

S

N

+

S, A

N

+

I

+ + 4-

L

262

67

S, E

+

+

S

+

+

IF

+

N

253

62

S, A, E

+

+ +

A, S

+

N

I

++

L

S, A

+ + +

+

I

+ 4- 4-

L

S

S

I

++

A

s

s

I

4- 4- 4-

L

279

70

s

N

+

s

N

+

IF

+

A

s

s

+

+ +

I

+ 4~

A

271

66

S, A, E

N

+ + +

S, A

N

+

I

++

A

S

N

+

S

N

N

I

+++

A

s

N

+

IF

+

A

s

S

I

H — 1 —

L, H

275

63

s

N

+

S

N

N

I

++

L

307

59

S, A

N

+ +

S

N

+

I

+ 4~ +

L

s

N

+ +

S

+

+

I

+ 4- 4-

L, H

s

N

+ +

s

N

+

I

4- 4" 4~

L, H

319

47

S, E

N

+ +

s

N

+

I

++

L, H

S

+

+ +

I

+ + +

L, H

S

+

+ +

s

N

+

I

4- + 4~

L, H

317

64

S, A

+ +

+ + +

S, A

N

+

I

+++

L, H

326

66

S, A

N

+

S, D

N

+ 4~

I

+++

L, H

323

80

S, A

N

+ + +

S, A

N

+++

I

+ + +

L, H

262

49

S, A

N

+ +

S, A

N

++

I

+ + +

A, H

279

59

S, A

+

+ 4“ +

S, A

+

N

I

+ +

L, H

360

64

S

+

+ +

S, D, A

N

N

I

+ + +

L

S

N

N

S

+

N

I

++

A, H

342

71

S, A

+

+ + +

S, A

+

4~ 4~

I

+++

L

S

N

+

S

+

4-

I

+ +

A, H

234

59

S

+

+ +

s

N

+

IF

+

A, H

224

53

S, A, E

-j-

+

S, A

N

+

I

++

L, H

S

+ +

+

s

I

+++

L

s

N

N

s

N

+

I

4- +

A, H

243

59

s

+

+

I

+

A, H

234

63

E, S

+ +

+ +

E, S

N

++

I

+++

L

317

51

S, E

+ + +

N

S

N

N

I

4 — 1 — b

L, H

S

+

N

s

N

N

I

+ + +

A, H

s

+

+

s

N

N

I

+++

A, H

s

N

N

s

N

N

Table 111 continued on next page.

90 Zoologica: New York Zoological Society [53: 3

Table III (continued)

Results of Examinations of Testes, Epididymides,

Location Specimen

Number

Date

Skull 1
Me as.

T estis3
fVt. (Gms.)

Testis Tubules
Height of
Diam. (/r) Epithel in)

Activity4

State

Kodiak Is. 4N

5-19-64

48*

FS

15N

5-19-64

59*

FS

30M

5-19-62

251/2

78

265

76

FS

41N

5-19-64

73

FS

5 1M

5-19-63

26%

92*

247

66

FS

64M

5-19-63

27%

FS

27M

5-20-62

24%

64

216

59

FS

23M

5-22-62

27%

97

218

58

FS

50M

5-22-63

27%

80*

266

82

FS

52M

5-22-63

29%

113*

264

83

FS

17M

5-23-62

26%

70

230

66

FS

35M

5-23-62

28%

89

269

67

FS

58M

5-23-63

69*

253

65

FS

32M

5-24-62

28%

78

246

76

FS

37M

5-24-62

28%

87

263

75

FS

38M

5-24-62

28%

86

243

FS

19M

5-25-62

27%

83

294

77

FS, A

39K

5-25-63

87*

FS

63M

5-25-63

27%

52*

FS

39M

5-28-62

28

85

FS

29M

5-29-62

27

96

285

83

FS

68M

5-30-63

26%

FS

62M

5-31-63

27%

FS

Alaska Pen. E252

5-?-63

FS

Kodiak Is. 60M

6-1-63

52*

FS

66M

6-12-63

84*

FS

Alaska Pen. 1812

7-14-63

52*

253

77

FS

1820

7-17-63

84*

276

76

FS

1825

7-18-63

68*

247

76

FS

1827

7-19-63

78*

237

59

FS, A

1831

7-21-63

52*

209

60

FS, A

Kodiak Is. 78M

8-3-63

53*

206

66

A, FS

69M

10-1-63

25%

FS, A

70M

10-2-63

263/8

FS

42A

10-17-62

2215/ig

54*

184

44

FS, A

48N

11-4-64

22%

55*

FS, A

55N

11-10-64

28%

81*

199

47

SN, A

Kodiak Is. 54N

11-10-64

28%

56*

169

53

SN, A

3 For explanation of number designations see Table II.

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

91

and Vasa Deferentia of Sexually Mature Brown Bears3

Intertubular Area

Epididymis

V as Deferens

General0

Character

Cytoplasmic

Abundance6 Vacuolation1

Diam. (p)

Fleight of
Epithel (p)

Lumen8 Coagulum8 Cytoplasmic10

Contents Prevalence Droplets

Lumen8
Contents (

Coagulum9

Cytoplasmic1

Droplets

IF

+ +

A, H

S

N

N

s

N

N

1

+ + +

L, H

S

N

N

s

N

N

I

+ + +

A, H

267

52

S, A

N

+

s

N

+

IF

+ +

A, H

S

N

+

s

N

N

I

+ + +

L, H

293

63

S, E

N

+

s

N

+

I

+ +

A, H

S

N

N

s

N

N

I

+ +

A, H

253

59

s

+

+ +

s

+

+

I

+ H" +

L, H

275

68

s

+ + +

+

I

+

A, H

360

71

s

+

+

s

N

+

I

+ + +

L

333

56

s

+ +

N

s

N

N

I

+ + +

L

325

67

s

N

+

s

N

N

I

+ 4~ +

A, H

317

63

s

+

+

s

N

+

I

+ +

A, H

298

64

s

N

+

I

+++

A, H

309

71

s

N

+

s

N

N

1

++

A, H

326

71

S, E

+

+

s

N

+

I

+++

L, H

288

64

S, A

+ +

+

S, A

N

+

I

+++

L, H

355

74

S, A

N

N

s

N

N

I

+ ~b

A, H

s

+

+

s

N

+

I

+ + +

A, H

s

N

+

s

N

N

I

+ +

A, H

S, A

N

C

s

N

N

I

+ + +

L, H

271

62

s

N

+

s

N

+

I

A, H

S, E

1

++ +

L, H

S

N

N

s

N

N

I

+ +

A, H

S

N

N

I

+

A, H

S

N

N

S, D

N

N

I

+++

L

262

62

S

N

N

s

N

N

I

+ +

A, H

271

63

s

N

+

s

N

N

I

++ +

A, H

261

66

s

N

+

I

+ +

A, H

s

N

+

E, S

N

+

I

+ +

A, H

S, E

N

N

E, S

N

N

I

++

A, H

253

66

S, A

+

+

S, A

N

N

IF

+

A, H

224

64

S, A, D

+

+

S, A

S

s

I

++

A, H

234

59

S, D, A

N

+

IF

H — b

A, H

A, S

+

+

A, D, S

N

N

I

+ +

L

215

59

A, D, S

+ +

N

A, D, S

N

N

IF

++

A, H

262

39

A, S, D

+ +

+

E, A, D, S

N

N

92

Zoologica: New York Zoological Society

[53: 3

The reproductive tracts of six bears were of
known approximate age as established on the
basis of returns from animals live-trapped and
marked as cubs or yearlings (Troyer, et ah, ’62) .
The remaining specimens were estimated as to
sexual status on the basis of skull sizes, and
testicular weight and histological comparisons
with the known-age animals.

Male reproductive tracts were removed, the
testes dissected free from the tunica vaginalis,
and the epididymides and proximal segments of
the vasa deferentia left in contact with the testes
(PI. I, figs 1-3). When possible the testes were
weighed fresh; otherwise, the fixed weight was
determined and adjusted to the fresh weight
(Table II). The specimens were fixed in 10 per-
cent formalin and stored in 70 percent ethyl
alcohol. Histologic sections of all specimens
were prepared from the body of the epididymis
together with the underlying portion of the testis
and the adjacent portion of the vas deferens,
where available, and from other areas of the
reproductive tract for representative specimens.

Sperm morphology and maturation were de-
termined by examining sperm taken from vari-
ous areas of preserved epididymides and vasa
deferentia. The sperm were stained in a 1 per-
cent solution of osmium tetroxide and the ana-
tomical features of whole preserved sperm were
then studied and measured with the light and
electron microscope.

The data available for the specimens studied
are listed in Tables I to III. The specimens were
classed as being infantile, prepuberal, or sexu-
ally mature as determined from histological find-
ings. Infantile specimens exhibited little or no
spermatogenic activity. Prepuberal specimens
showed some spermatogenesis but not to a sexu-
ally functional state. Sexually mature bears rep-
resented one of three reproductive states: (1)
preseasonal— namely animals recovering from a
nonbreeding period; (2) sexually active— bears
in or near full breeding condition; or (3) post-
seasonal— animals showing spermatogenic de-
cline following the breeding season.

Included for each specimen, if available, are
the collection date, an age classification, body
weight, testicular weight, the combined length
plus zygomatic width of the skull, and the his-
tologic classification of the testis, epididymis,
and vas deferens (Tables I to III).

Observations

Estimates of Age and Reproductive State

Reliable procedures for estimating the ages
of bears are not available. Nonetheless the speci-
mens included in this study were readily identi-
fied as infantile, prepuberal, or adult on the

basis of testicular histology. It early became ap-
parent, too, that skull sizes and testicular weights
might also provide useful criteria for determin-
ing the reproductive maturity of bears and for
estimating their approximate ages, at least in
younger animals.

As seen in Text-figure 1 and Tables I to III,
the size of the brown bear’s testis appears to be
directly correlated with the size of the skull and
thus, presumably, with age. Although there are
individual and seasonal variations, it will be
noted that single testes of mature specimens
range upward in weight from approximately 50
gms; those of infantile bears weigh approximate-
ly 25 gms or less. The limited known-age speci-
mens available (Table I) suggested that the in-
fantile group included bears to three years and
occasionally older. The sexually mature speci-
mens were presumably four or more years of
age. The testis weights of presumed prepuberal
bears based on histological examination were
found to fall roughly between those of the in-
fantile and sexually mature animals but over-
lapped each group slightly (Text-fig. 2 and
Tables I to III). The overlap is attributed to the
different ages at which individual male bears at-
tain sexual maturity. Dittrich and Kronberger
(’63) have reported similar variations in the
time of sexual maturity in the female brown
bear.

Most male bears apparently attain prepuberty
in their fourth year and became sexually mature
at approximately four and one half years of age.
Presumably, as judged by the widely divergent
skull sizes of prepubal bears a few attain this
state in their third year and others not until their
fifth year (Tables I to III and Text-fig. 1). The
four sexually mature specimens with skulls mea-
suring only 22 and 23 inches in length plus width
and separate from the remaining mature ani-
mals, and the five infantile bears with skulls in
the same size range and separate from the re-
maining infantile specimens, provide further evi-
dence that the age of puberty is quite variable
among male bears (Text-fig. 1).

The relative age and sexual status of bears
may also be judged with fair accuracy from the
skull (Tables I to III and Text-fig. 1). Animals
with combined skull length and width measure-
ments exceeding 241A inches are quite assuredly
sexually mature specimens exceeding four years
of age. Conversely, specimens with skulls mea-
suring less than 22 inches are, generally, imma-
ture. Between these limits is a group of bears of
mixed sexual status, the majority presumably be-
ing prepuberal animals in their fourth year of
life (Tables I to III and Text-fig. 1).

While paired testes were obtained from only
a few bears, there was no indication that signi-

TESTIS WEIGHTS IN GRAMS TESTIS WEIGHTS IN GRAMS

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

93

130

no

90
70
50
30
10

19 20 21 22 23 24 25 26 27 28 29

SKULL LENGTH PLUS WIDTH IN INCHES

Text-fig. 1. Testis weights of brown bears as a function of skull size.

130

no

90
70
50
30
10

OOOOOOOOOOOOOOOOOOOOOOOO
— rvif*) — cvjfO — oam-cxjrQ — aifO—airo — cMfO — Mco

APR MAY JUNE JULY AUG SEP OCT NOV

DATE CHRONOLOGY

TESTIS WEIGHTS OF BROWN BEARS
AS A FUNCTION OF SEASON

TESTIS WEIGHTS OF BROWN BEARS
AS A FUNCTION OF SKULL SIZE

• INFANTILE
® PREPUBERAL
o MATURE

o

o o
®

® ®

®

1

t * •

•••:«*• •

®

®

o

o o O
o
ft)c

o 00 °o °f>P

Jo o
o o

° °°^<9

Text-fig. 2. Testis weights of brown bears as a function of season.

94

Zoologica: New York Zoological Society

[53: 3

ficant differences existed between left and right
testicular weights in the brown bear (Table III).
The heaviest testis weighed 121 gms. The mean
testis weight was 80 gms for the sexually mature
bears, 31 gms for the prepuberal bears, and 12
gms for the infantile bears. There were, however,
seasonal variations in the testicular weights of
bears. Among sexually mature specimens heavi-
est testes occurred during the breeding season in
May and June (Table III and Text-fig. 2) . Testi-
cular weights were then approximately one-third
greater than during the fall postbreeding season,
and conceivably at least twice as heavy as at the
time of maximum testicular regression in mid-
winter. The decline in weight is in large measure
attributable to shrinkage of the seminiferous
tubules during the nonbreeding season (Table
IV). A similar condition has been reported in
the black bear (Erickson and Nellor, ’64). In
contrast to the spring to fall testicular weight
decline in the adult bear, an opposite condition
occurs in the infantile bear. Here, fall testicular
weights exceed spring weights presumably be-
cause the infantile bear realizes substantial body
growth during the spring to fall period. It fol-
lows, therefore, that the fall testes of this sexual
class would weigh more since, as was shown in
Text-figure 1, a positive correlation exists be-
tween the body size and the testicular weight of
bears. There is a suggestion, nonetheless, that the
seminiferous tubules of the infantile bear under-
go slight shrinkage during the postbreeding sea-
son as in the mature animal (Table IV).

In contrast to both the infantile and adult
bear, the testes of prepuberal bears remain es-

sentially constant in weight from spring to fall
(Text-fig. 2). This is presumably due to slight
shrinkage of the seminiferous tubules following
the breeding season in the spring, but this is ac-
companied by some compensatory growth of the
body as a whole (Table III and compare PI. Ill,
fig. 12 and PI. IV, fig. 19).

Gross Features of the Male Reproductive Tract

Grossly the reproductive tract of the male
brown bear differs only in relative size from that
described for the black bear (Erickson and
Nellor, ’64). The species has a well-developed
os penis which measures up to eight and one-half
inches in length in older bears and, as in the
black bear, the penis is capable of extrusion from
its sheath only with the attainment of sexual
maturity. The testes are scrotal from infancy
and are held closely to the body except in the
adult animal during the breeding season. They
are then further removed from the body due
apparently to relaxation of the scrotum and their
enlargement through vascular engorgement and
tubule hypertrophy (Table IV, and PI. I, figs. 1
and 2). Concomitant with the attainment of
sexual maturity the tip of the scrotum becomes
bare and very darkly pigmented. The hair-free
patch is reduced in size and less obvious in late
fall animals due principally to scrotal shrinkage.
However, once sexual maturity is attained com-
plete refurring of the patch apparently does not
occur since it was observed in bears killed from
April through November. This character is thus
a useful one for identifying sexually mature
bears.

Table IV

Testis Tubule and Epididymal Duct Measurements of
Brown Bears as Related to Age and Season <a)

Reproductive Status

Season

Infantile

Prepuberal

Mature

No. Spec.

Ave.

Diam. (u)

No. Spec.

Ave.

Diam. (u)

No. Spec.

Ave.

Diam. (u)

Spring

(Apr. 2-May 29)
Tubule Diam.

4

97

5

159

31

245

Epididymal Duct Diam.

5

163

4

218

34

293

Summer

(June 12-Aug. 3)
Tubule Diam.

1

83

6

238

Epididymal Duct Diam.

-

-

-

-

5

254

Fall

(Sept. 1-Nov. 11)
Tubule Diam.

11

94

4

145

3

184

Epididymal Duct Diam.

11

180

4

199

3

237

<a) Includes only specimens whose tissue condition warranted measurements; average of five epididymal
and 10 tubule measurements.

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

95

The testis of the brown bear is roughly ovoid
in shape and is slightly compressed laterally (PI.
I, figs. 1 and 2). The organ is not particularly
large for the animal’s size. In the adult bear at
the height of the breeding season it measures
approximately 9x6x5 cm and weighs slightly
over 80 gms. Three bears, 1821, 1827, and 1831
had body weights of 705, 630, and 560 pounds
and testis weights, respectively, of 84, 78, and
52 gms. The epididymis of the bear is large and
tightly attached to the dorso-lateral surface of
the testis, as is the vas deferens which courses
back along the epididymis (PI. I, figs. 1 and 2).

The gross internal architecture of the testis is
shown in Plate I, figure 3. The fibrous tunica
albuginea is very heavy. The blood vessels of the
tunica vascularis hypertrophy markedly during
the breeding season and supply the major blood
needs of the testis. The vascularis layer is most
pronounced at the cephalic end of the gland, be-
coming diminished caudally as vascular elements
are passed into the testis (PI. I, fig. 3).

At the cephalic end of the testis portions of
the tunica albuginea pass into the gland as the
mediastinum testis, which consists of a core of
fibrous connective tissue, minor blood vessels,
and the rete testis. The mediastinum testis is
central in position and extends approximately
three-fourths of the length of the testis (PI. I,
fig. 3). This is continuous with finer laminae of
connective tissue, the septulae testis which ex-
tend radially into the testis encompassing the
individual lobuli testis, the germinative, and en-
docrine portions of the gland. The coiled semi-
niferous tubules connect with short, straight
tubuli recti which convey the sperm from the
apex of the lobules to a series of irregular spaces
lined with low cuboidal epithelium, the rete
testis, which extends throughout the mediasti-
num (PI. IX, fig. 51).

The Histology of the Male Reproductive Tract

Infantile hear— The seminiferous tubules of
the infantile bear are simple, undifferentiated
cords widely dispersed in a cellular connective
tissue. There is little apparent change in their
appearance in most bears until the animals reach
two years of age (PI. II, figs. 4 and 5) . The semi-
niferous tubules then show marked enlargement
and come to occupy a major portion of the testis
mass (PI. II, fig. 6) . The tubules of the two-year-
old bear at the time of the breeding season in
May measure slightly less than 1 00/x in diameter,
are lumenless and filled with an amorphous ma-
terial of uncertain origin— possibly Sertoli cell
cytoplasm. The epithelial membrane of the tu-
bules is two to four cells thick and for the most
part the cells resemble Sertoli cells with promi-

nent nucleoli (PI. II, fig. 6). However, certain
cells are hypertrophied and show mitosis indi-
cating that some among them are germ cells.
These do not develop beyond spermatocyte
stages, however, and they either degenerate in
situ or are passed to the tubule center, become
pyknotic, and disintegrate. Relatively few of
these immature cell forms are passed to the
epididymis and the vas deferens (PI. II, figs. 7
to 9) .

The interstitial cells of the two-year-old bear
are abundant but the tissue occupies a minor
portion of the testis. The cells contain relatively
little cytoplasm, are crowded closely together,
and the nuclei stain darkly (PI. II, fig. 6). It is
noteworthy, nonetheless, that bears of this age
class apparently produce abundant androgen as
indicated by the well-developed epididymis and
vas deferens (PI. II, figs. 7 to 9). These organs
appear fully functional, the epithelium being
pseudostratified columnar with stereo-cilia as in
the typical sexually mature mammal. With rare
exceptions, however, the epididymal ducts and
the vasa deferentia are empty of spermatogenic
elements during the spring breeding season (PI.
II, fig. 7). On the other hand, abundant secre-
tion products and an amorphous coagulum are
constant in these organs at all seasons (PI. II,
figs. 8 and 9) .

Following the breeding season in spring, the
reproductive tract of the two-year-old infantile
bear shows slight retrogressive changes. By Oc-
tober the seminiferous tubules undergo some
shrinkage. This is indicated by thickening and
wrinkling of the basement membranes, and an
increased density of the central cytoplasmic
complex (PI. Ill, fig. 10) . There is also a decline
in the cell population of the seminiferous tubules
and an accompanying increase in pyknotic shed
elements, although this was highly variable be-
tween bears. As in the spring infantile bear, most
of the cells lining the basement membrane are
Sertoli-like (PI. Ill, fig. 10).

The intertublar tissue in the two-year-old fall
infantile bear was more apparent than during the
spring, was frequently arranged in strands, and
the cytoplasmic-nuclear ratio of the Leydig cells
was slightly greater (compare PI. II, fig. 6 and
PI. Ill, fig. 10). In general, however, the his-
tology of the testis of the two-year-old bear does
not differ appreciably from spring to fall.

The same statement cannot be made for the
epididymis and the vas deferens. By November
these organs undergo marked retrogression from
their highly active state during the spring (com-
pare PI. II, fig. 7 and PI. Ill, fig. 11). By late
fall the ducts become reduced in size, lose their
cilia, and develop a thick, dense, connective tis-

96

Zoologica: New York Zoological Society

[53: 3

sue coat. While most of the ducts are empty, a
few contain abnormal and immature sperm cell
elements, and a heavy coagulum is a common
feature (PI. Ill, fig. 11).

Prepuberal bear.— Unfortunately no prepu-
beral bears were obtained prior to the breeding
season in May. However from this time on there
was a marked increase in the size of the semi-
niferous tubules to approximately 160/t (Table
IV) and the development of large, distinct tu-
bule lumina (PI. Ill, fig. 12). The interstitial
tissue at this time appeared sparse and com-
pressed between the swollen tubules. The Leydig
cells seemed to have a slightly greater cytoplas-
mic-nuclear ratio than in the infantile spring
bear (compare PI. II, fig. 6 and PI. Ill, fig. 14).

While the germinal epithelium of the spring
prepuberal bear was only four or five cells thick,
there was a considerable amount of spermato-
genic activity and spermatogenic stages were
quite readily differentiated from Sertoil cells
(Table I, PI. Ill, fig. 13 and PI. V, fig. 24). In-
terestingly, however, spermatogenesis became
arrested in the secondary spermatocyte and
spermatid stages and, as in the infantile bear,
the majority of these cells appeared to disinte-
grate in situ. A fair proportion were shed to the
tubule lumine, however, and passed to the epi-
didymis and vas deferens (PI. Ill, fig. 15 to PI.
IV, fig. 17). In this process multinucleate giant
cells were often formed (PI. Ill, figs. 13 and 15),
apparently by a coalescence of developing germ
cells in situ or a failure of the cell clones to
separate (Fawcett, 1961). These are thought to
indicate an animal in less than full reproductive
vigor (Parks, ’60).

As with the two-and-one-half-year infantile
bear, the epididymis and vas deferens of the
spring prepuberal bear is well developed with
abundant secretory products (PI. IV, figs. 16 and
17). The diameters of the ducts are increased,
however, over those of the infantile animal
(Table IV). A notable difference from the in-
fantile bear is the regular occurrence of imma-
ture and abnormal cell forms within the ducts
(PI. Ill, fig. 15). Relatively few of these reach
the vas deferens which suggests that their deg-
radation is continuous as they proceed down
the tract (PI. IV, figs. 16 and 17).

Following the breeding season the seminif-
erous tubules of the prepuberal bear show a
loss of turgor and activity (PI. IV, fig. 18). The
tubules become reduced in diameter (Table IV) ,
the basement membranes thicken and wrinkle,
and the tubule lumina are markedly reduced
(PI. IV, figs. 19 and 20). This is accompanied
by a reduction in the germinal epithelium to the
point that Sertoli cells predominate over 'germ
cells. Nevertheless spermatogenesis continues at

a diminished rate, the formed elements continu-
ing to degenerate either in situ or in passage to
the tubule lumina and epididymal ducts.

Ultimately, during late fall, defoliation of the
spermatogenic epithelium becomes so pro-
nounced that the only cells remaining within
the tubules appear to be Sertoli cells (PI. IV,
figs. 19 and 20). During spermatogenic decline,
as during other stages of the cycle, not all bears
or even all tubules of a given bear were in the
same reproductive state ( PI. IV, figs. 1 8 and 20) .
These differences were not attributable to the
“spermatogenic wave” phenomena.

During the declining phase of spermato-
genesis in the prepuberal bear the interstitial
tissue is strandlike, shows an apparent increase
in cytoplasmic amount, and the Leydig cell
nuclei are smaller and darker stained (compare
PI. Ill, fig. 14 with PI. IV, figs. 19 and 20). This
is accompanied by atrophy of the vas deferens
and of the epididymis (PI. IV, fig. 21 to PI. V,
fig. 24). The vas deferens is the first of these
organs to undergo degenerative changes (PI. V,
figs. 23 and 24). Its decline precedes even that
of the seminiferous tubules (compare PI. IV, fig.
19 and PI. V, fig. 24).

The first indication of retrogression of the vas
deferens and epididymis is disorganization of
the duct epithelia (PI. V, fig. 23). This process
becomes marked and is accompanied by a de-
crease in the size of the ducts (Table IV). By
November the epithelium of the vas deferens is
reduced to a layer of very low columnar cells,
the primary duct contents being degradation
products of the earlier pseudostratified epi-
thelium, a few abnormal germinal elements, and
coagulum (PI. V, fig. 24). By contrast, the epi-
didymis retains a functional appearance for at
least an additional month (PI. IV, fig. 21 and
PI. V, fig. 22) and in none of the prepuberal
specimens available to us did the epididymis
attain a state of involution similar to that shown
by the vas deferens (PI. V, fig. 24). Conceiv-
ably, however, such a state would have been
seen if we had had specimens of this group ex-
tending beyond early November. A notable fea-
ture during the declining phase was a marked
increase in the coagulum found within the ducts
(PI. IV, fig. 21 and PI. V, fig. 22).

The Sexually Mature Bear

Redevelopment phase.— The earliest mature
spring bear, 41 M, was taken April 2, 1963
(Table III). The activity state of this animal
was, however, well advanced over a number of
other specimens taken at later dates. Of particu-
lar interest among these was specimen 22 M
taken April 22, 1962. The small size of this
animal’s skull, and its delayed spermatogenic

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

97

state as compared with a majority of the other
mature specimens we examined, suggest that it
was just attaining sexual maturity (Table III and
PI. V, figs. 25 and 26). Among the obviously
sexually mature bears included in our collection,
specimen 42 M taken April 18, 1963, exhibited
the least advanced stage of spermatogenic de-
velopment (PI. V, fig. 27). The seminiferous
tubules of this specimen appeared quite uni-
formly rounded and the surrounding intertubu-
lar tissue was stringy and loose. This suggests
that the tubules were not as markedly swollen as
later in the cycle. In general appearance the
tubules were quite dense, and without a distinct
lumen. They were densely packed with devel-
oping spermatocytes and spermatids (PI. VI, fig.
28). A few tubules were, however, in a more
advanced state and exhibited all stages of sper-
matogenesis including mature spermatozoa (PI.
VI, fig. 29).

In the redeveloping bear testis abnormal cell
forms were frequently observed, particularly
multinucleated giant cells (PI. VI, fig. 30). The
high frequency of abnormal and immature ger-
minal elements seen in the epididymides and
vasa deferentia of the late redeveloping bears
suggests that the first germinal elements passed
are largely nonviable forms (PI. V, fig. 26, PI.
VI, figs. 29 and 33). A heavy coagulum was
also regularly noted in the lumina of the rede-
veloping tubules (PI. V, fig. 27, PI. VI, figs. 28
and 30). This was probably in part Sertoli cell
cytoplasm together with degradation products of
abortive germinal elements phagocytosed by
Sertoli cells (Vilar, ’65).

The interstitial tissue of the redeveloping
bear testis is abundant. The cells have a rela-
tively large cytoplasmic-nuclear ratio and the
abundance of large vacuoles in their cytoplasm
is striking (PI. VI, figs. 28 and 29).

While specimens of bear epididymides and
vasa deferentia were not available from bears
prior to late testicular redevelopment, the well-
developed state of these organs from the earliest
specimens available suggested that they became
fully functional before the testes produced suf-
ficient mature spermatozoa for fertile breeding
(PI. VI, figs. 31 and 32). And, as indicated
earlier, immature and abnormal cell forms were
regularly to be noted in the duct contents of the
redeveloping bear epididymis and vas deferens.
A further interesting character was a heavy and
marked vacuolation of the epithelia of the epi-
didymis and vas deferens. This is possibly a
precursor of the abundant secretion products
noted in these organs (PI. VI, fig. 32).

Fully active animals.— As has been mentioned
earlier, bears vary markedly in the time of at-

tainment of full breeding condition. Apparently,
however, all mature specimens are reproduc-
tively capable in May (Table 111).

At the height of breeding capability the sem-
iniferous tubules are much swollen and com-
pressed together and thus appear flattened or
polygonal in cross section (PI. VII, fig. 34) . The
tubules have distinct lumina and all stages of
spermatogenesis are to be seen in the germinal
epithelium. Abnormal sperm are rarely noticed.
While tufts of maturing spermatozoa are abun-
dantly embedded in the cytoplasm of Sertoli
cells, free sperm are seldom noticed in the tubule
lumina which suggests their rapid transport to
the epididymis (PI VII, fig. 34). The tubule
lumina are also free of the amorphous coagulum
which was commonly noted in the redeveloping
bear.

At the height of the breeding season the inter-
tubular interstitial tissue of the mature bear
appears sparse and dense (PI. VII, fig. 34). The
large, abundant cytoplasmic vacuoles noted in
the interstitial tissue of the redeveloping bear
are generally not noticed at this time except in
the abundant interstitial tissue which invades the
tunica albuginea and septulae testis (PI. VII,
figs. 38 and 39) .

The epididymis and vas deferens of the brown
bear at the height of the breeding season are
greatly distended with sperm (PI. VII, figs. 35
and 36). That the vas deferens, as well as the
epididymis, is an important organ for sperm
storage in this species is indicated by the fact
that it becomes so distended with spermatozoa
that the usual pseudostratified columnar epi-
thelium becomes low columnar (PI. VII, fig.
36). At the height of the breeding season ab-
normal or immature cell forms occur only rarely
in the ducts, and the epithelium of the epidid-
ymis and vas deferens are largely devoid of
the heavy vacuolations noted in the prepuberal
and in the preseasonal mature bear (PI. VII,
figs. 35 and 36). Bleb-like secretions are fre-
quently to be seen however at the tips of the
stereocilia (PI. IV, fig. 6). The source of
these is presumed to be the abundant vacuoles
noted in the epithelia of the vasa deferentia and
epididymides of the seasonal adult and pre-
puberal bear, and they are further the apparent
source of the abundant secretions noted among
the stereocilia in these animals (PI. IV, fig. 17
and PI. VI, fig. 32). It would appear, therefore,
that there is a build-up of secretion products
during the nonbreeding period and that the re-
lease of these products exceeds their build-up
during the breeding season.

Postseasonal animal.— Following the limited
breeding season in the spring, the male brown

98

Zoologica: New York Zoological Society

[53: 3

bear shows spermatogenic decline. The time of
the decline varies widely between bears (PI.
VII, figs. 37 and 39). The first sign of seminif-
erous tubule atrophy is the appearance of con-
siderable numbers of abnormal and immature
cell forms (PI. VII, figs. 37 and 38). Sperma-
togenesis continues but is arrested in the sec-
ondary spermatocyte and spermatid stages. This
condition closely parallels spermatogenesis as it
occurs in the prepuberal and the preseasonal
mature bear, particularly as regards the forma-
tion of multinucleated giant cells (PI. Ill, fig.
13; PI. IV, fig. 18 and PI. VI, fig. 30). During
the early phases of decline the germinal epi-
thelium becomes reduced to four or five cells in
thickness and tubule lumina become more dis-
tinct (PI. VII, fig. 39 and PI. VIII, fig. 40). This
is followed by tubule shrinkage as manifested
by a reduction of the luminar and a thickening
and folding of the basement membranes (PI.

VII, fig. 38 and PI. VIII, fig. 42) . Defoliation of
the spermatogenic epithelium continues until
ultimately many, if not most, of the tubules are
reduced to simple cords, the sole cell population
of which consists of a single layer of cells lining
the basement membrane as in senile testes (PI.

VIII, figs. 41 to 43). As in the prepuberal bear
during the postbreeding season, these cells ap-
pear to be Sertoli cells.

Since we had no specimens extending beyond
November, it was not possible to determine how
complete seminiferous tubule decline becomes.
The indications are, however, that the condition
becomes quite general at the height of decline,
presumably in midwinter when the animals are
in their winter dens.

The interstitial tissue of the mature bear testis
during the late postbreeding period appears
slightly more abundant than during the breed-
ing season. There is an apparent increase in the
amount of cytoplasm within the cells but the
vacuoles do not appear to attain the size of those
seen during the prebreeding season (compare
PI. VI, figs. 28 and 29 and PI. VIII, figs. 40 to
42). The interstitial tissue at this time is strand-
like, the nuclei being dense and dark staining,
and intercellular spaces are common. While
these spaces were perhaps in part artifacts, we
believe they represent decreases in cell size pos-
sibly coupled with edema. Concomitant with
seminiferous tubule atrophy during the post-
breeding season in the adult bear there is an
accompanying but slightly later decline of the
vas deferens and the epididymis (PI. VIII, fig.
44 to PI. IX, fig. 49). This is suggested by the
appearance of immature and abnormal cell
forms in the ducts before any discernible change
in the histology of these organs. The passage of
aberrant forms continues and ultimately they

predominate over normal forms in the duct con-
tents (PI. VIII, fig. 45). This is accompanied
also by a marked reduction in the sperm load.
In some instances the epididymal ducts are
largely empty of germinal elements before there
is any recognizable gross change in the epi-
didymis itself (PI. IX, fig. 46). In other cases
some disorganization of the epithelium lining
the epididymal ducts becomes obvious while
germinal elements are still evident. Surprisingly,
however, the vas deferens declines well in ad-
vance of the epididymis (compare PI. VIII, fig.
45 and PI. IX, fig. 48), the epithelium becoming
low cuboidal and the duct contents consisting
solely of degradation products and degenerate
sperm (PI. IX, fig. 48). In certain specimens
such a condition occurred as early as September
while in others a fairly healthy condition ex-
isted as late as November (Table III).

The limited number of fall specimens avail-
able to us did not afford a complete picture of
epididymal decline in the adult bear. It appears
likely that at the height of decline the epithelium
of the epididymis roughly resembles that at-
tained by the vas deferens. It is certain that dur-
ing this time there is a marked reduction in the
diameter of the epididymal ducts together with
a relative increase in the connective tissue sur-
rounding them (Table IV and PI. IX, fig. 49) .

Sperm morphology.— Mature brown bear
spermatozoa appear typically mammalian in
form with an overall length of 18.2p. The head
of the sperm is round to oblong in greatest
length-width aspect, and flattened with a slight
inflection of the head tip. The length and width
of the head are 1A and 4 p respectively. The
cylindrical midpiece measures 11.5 p and in
sperm taken from the epididymis it typically
retains a midpiece cytoplasmic droplet (PI. IX,
fig. 50). The attachment of the midpiece to the
head is slightly abaxial. The length of the tail is
57.0 p and the length of the terminal fibrilis
2.3 p. These measurements fall within the range
of values reported for other mammalian species
(Parkes, ’60).

Discussion

Although criteria for estimating the ages of
brown bears have not been developed, several
characters discussed in this study appear to have
merit as indices for estimating the reproductive
status and age of the male bear. The histology
of the testis, regardless of the season, appears
definitive for establishing the sexual status of
bears as being infantile, prepuberal, or sexually
mature. While a more extensive collection of
known-age specimens is necessary to determine
more closely the ages represented by each sexual
group, these preliminary data suggest that the
infantile class is represented by bears less than

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

99

three years of age and the sexually mature class
by bears four years and older. The prepuberal
class is mostly bears in their third year of life,
although a few two-year and four-year animals
are believed to be included also.

The weight of the testis and the size of the
skull appear equally useful as indices of the age
and sexual status of male bears. As judged from
the histology of the testis, infantile bears gener-
ally have single testis weights of less than 25
gms and skulls measuring less than 22 inches in
length plus width. Conversely, the single testis
weights of sexually mature bears ranged upward
from approximately 50 gms and the skulls sel-
dom measured less than 24 Vi inches in length
plus width. The testes and skulls of prepuberal
animals lie roughly between those of the in-
fantile and sexually mature groups with some
slight overlap with both as was also true for the
histological classifications. Nonetheless, skull
sizes and testicular weights, particularly if taken
together, appear to be fairly reliable criteria for
classifying bears as to sexual status and relative
age (Text-fig. 1). Other criteria described as
useful for identifying the sexually mature bear
were: capacity of the penis to be manually ex-
truded from its sheath, and the presence of a
darkly pigmented hairless patch at the tip of the
scrotum.

These studies show that the male brown bear
is a seasonal breeder, the reproductive period
beginning before and extending beyond the
breeding season of the female. As contrasted
with the more limited and consistent reproduc-
tive period of the female, that of the male ap-
pears to last four or five months and varies
markedly between individual bears. Certain
specimens are in full reproductive vigor well in
advance of the time of peak breeding while
others appear reproductively capable until near
the time of denning in November.

It seems, therefore, that the male brown bear
has a potential breeding season slightly exceed-
ing half of the year and encompassing the
greater portion of the den-free period. The dif-
fering testicular histology throughout this period
suggests, however, that individual bears are sex-
ually capable for perhaps only about half of the
potential breeding period (Table I). It is signifi-
cant that during the period of normal female
reproductive activity from late May to early
July (Dittrich and Kronberger, ’63) practically
all sexually mature bears are producing abun-
dant spermatozoa (Table III).

From a comparative point of view, certain as-
pects of the morphology and histology of the
brown bear reproductive tract are extremely in-
teresting. The high development of the epi-
didymis and the vas deferens of the immature

one-and-two-year old bear well in advance of
any spermatogenic activity was remarkable and
was seemingly typical of the state attained by
the sexually mature animal. To our knowledge
such an advanced state of development at least
two years in advance of the attainment of sexual
maturity is unique among mammals studied to
date with the exception of the closely related
black bear ( Erickson and Nellor, ’64) . The high
degree of spermatogenic activity attained by
the prepuberal bear a full year in advance of
sexual maturity was similarly striking.

Perhaps the most interesting feature of the
histology of the male bear’s reproductive tract
is the marked atrophic state exhibited by the
testis in the nonbreeding season. While the speci-
mens available to us did not permit a determina-
tion as to how general the atrophic state be-
comes, the indications were that the senile
appearance of the seminiferous tubules becomes
quite general at the height of decline, presum-
ably in midwinter when the animals are in winter
dens. This was suggested both by the almost
complete atrophy of the tubules during the non-
breeding season in the seasonally quite active
prepuberal bear (PI. IV, figs. 18 and 19), and
by the large proportion of senile tubules in the
late declining adult bear (PI. VIII, figs. 42-44).

In light of this it appears warranted to specu-
late as to the manner in which the germinal epi-
thelium becomes repopulated; whether it be
from a far advanced or more limited state of
atrophy. Three explanations may be advanced.
First it may be that the Sertoli-like cells remain-
ing in the tubules are really a mixture of sperma-
togenic and Sertoli cells but are not recogniz-
ably different. A second postulate is that the
cells are truly a single cell type but bipotential.
The third possibility is that portions of the
tubules fail to undergo the marked atrophy in-
dicated in PI. VIII, figs. 41 to 43 and these resid-
ual centers repopulate the tubules. Of these
possibilities the latter appears the most plausible
for the following reasons: As indicated in PI.
VIII, figs. 41 and 43, wide variations were seen
between the condition of tubules not only be-
tween bears but even within a given testis. Even
in animals showing a high percentage of senile-
type tubules, other tubules could be found show-
ing active spermatogenesis. It was interesting
too that seminiferous tubule decline appeared
in some cases to be affecting only individual
tubules (PI. VIII, fig. 41), but in other cases
whole lobuli testes (PI. VIII, fig. 43). This is
taken as evidence that atrophy begins at given
points within the tubules and then spreads pro-
gressively along their length.

We believe that seminiferous epithelial repop-

100

Zoologica: New York Zoological Society

[53: 3

illation (postulate 1) cannot be explained as re-
sulting from a mixture of cells in which sper-
matogonia and Sertoli cells are indistinguishable.
The irregularly spherical shape of the nucleus
and the large, dark-staining, eccentrically lo-
cated nucleolus is typical of Sertoli cells. None
of these nuclei showed a condition suggestive of
meiosis which further reduces the possibility of
their being spermatogonia, and therefore also
lowers the probability that they are bipotential
as suggested by postulate two.

Another interesting feature of the histology
of the brown bear testis is the marked degree to
which the interstitial tissue invades the tunica
albuginea (PI. VII, fig. 39). Furthermore, the
Leydig cells in these areas have a relatively large
cytoplasmic-nuclear ratio, in which their large
vacuoles are an important factor. Then too,
while the intertubular interstitial tissue of the
fully active animal appears to possess a much
lower cytoplasmic content than at other times in
the cycle, the cells in the tunic consistently main-
tained their seemingly high functional state.

The secretion products of the epididymides
and vasa deferentia are apparently proliferated
from their highly active stereo-ciliated epithelial
cells, and vary in amount between bears. In the
prepuberal bear at all seasons, and in the pre-
seasonal and early postseasonal sexually mature
bear, the basal cells of the epithelia of these
organs commonly contain large numbers of
vacuoles (PI. VI, fig. 32). A similar state is
rarely seen in the fully active animal. Rather the
vacuoles appear at this time to have migrated
toward the apical ends of the cells and the fre-
quently observed bleb-like formations among
the stereocilia may be a product in some way
related to them. This sequence suggests a build-
up of these products during the nonbreeding
season and a depletion (over build-up) during
the active reproductive period.

The abundant coagulum throughout the tract
is apparently of varied origin and again appears
to be related to the reproductive state of the
animal. It seems to be absent in the fully active
adult, present in low amounts in the prepuberal
animal, and quite abundant in the seasonal adult,
particularly during the preseasonal period (PI.
IX, fig. 46.) While some of this material may
conceivably be of Sertoli cell origin, the larger
amount appears to be degradation products of
degenerate sperm cells phagocytozed by Sertoli
cells (Vilar, ’65) and possibly by lymphocytes.
Further degradation of degenerate sperm ap-
pears to occur in the epididymis and vas deferens
and adds to the coagulum load in these organs.
Additional coagulum is believed to be added to
the ducts of the epididymis and the vas deferens
by liquifaction of the secretions of these organs

and by the addition of cytoplasmic material
from maturing sperm.

An additional interesting observation in the
declining bear is the frequent invasion of lym-
phocytes into and surrounding the seminiferous
tubules, the epididymides, and deferent ducts.
A similar condition frequently occurs in the pre-
puberal and preseasonal adult bear (PI. IV, fig.
19 and PI. VIII, fig. 45). While this condition
would normally suggest a pathologic state, it is
too general throughout our specimens to be at-
tributed to this. A possible explanation is that
the lymphocytes are involved in the breakdown
of both the epithelial debris from the ducts and
of the great numbers of degenerate germinal
cells.

Literature Cited
Couturier, M. A. J.

1954. Lours Brun. Grenable. 904 pp.

Dathe, H.

1961. Beobachtungen zur Fortpflanzungsbiol-
ogie des Braunbaren, Ursus arctos L. Der
Zool. Garten (NF), 25:235-250.

DeVoto, B. A.

1953. The journals of Lewis and Clark. Bernard
DeVoto, Ed. Houghton Mifflin Co., Bos-
ton. 504 pp.

Dittrich, L., and I. V. Einsiedel

1961. Bemerkungen zur Fortflanzung und Jug-
endentwicklung des Braunbaren (Ursus
arctos L.) im Leipziger Zoo. Der. Zool.
Garten (NF), 25:250-259.

Dittrich, L„ and H. Kronberger

1963. Biologisch-anatomische Untersuchungen
liber die Fortpflanzungs-biologie des
Braunbaren (Ursus arctos L.) und an-
derer Ursiden in Gefangenschaft. Zeit-
schrift for Saugetierkunde, 28:129-192.

Erdbrink, D. P.

1953. A review of fossil and recent bears of the
old world. Doct. Diss., Jan de Lange, De-
venter. 597 pp.

Erickson, A. W., and J. E. Nellor

1964. Breeding biology of the black bear. In:
The black bear in Michigan, Pt. I, Mich.
Agr. Exp. Sta. Research Bull. No. 4, Mich.
State Univ., East Lansing. 5-45.

Fawcett, D. W.

1951. Intercellular bridges. Exper. Cell Re-
search, Supplement B: 174-187.

Murie, A.

1944. The wolves of Mount McKinley. Fauna
of the Natl. Parks of the U.S., Fauna
Series No. 5. 238 pp.

1968]

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

101

Parks, A. S.

1960. Vol. I, Pt. II. Marshall’s physiology of
reproduction. A. S. Parks, Ed. Longmans,
Green and Co. Ltd., Toronto. 877 pp.

Pocock, R. I.

1932. The black and brown bears of Europe and
Asia. Pt. I, J. Bombay Nat. Hist. Soc.,
35:771-823.

Rausch, R. L.

1962. Geographic variation in size in North
American brown bears, Ursus arctos L.,
as indicated by condylobasal length.
Canad. J. Zook, 41:33-45.

Troyer, W. A., R. J. Hensel, and K. E. Durley

1962. Live trapping and handling of brown
bears. J. Wild. Mgt., 26:330-331.

Vilar, O.

1965. An electron microscopial study of the
phagocytosis of germinal cells by Sertoli
cells in tissue cultures. Anat. Record, 15 1 :
428-429.

WlMSATT, W. A.

1963. Delayed implantation in the ursidae, with
particular reference to the black bear
( Ursus americanus Pallas). In: Delayed
Implantation, Univ. Chicago Press, Chi-
cago. 49-76.

102

Zoologica: New York Zoological Society

[53: 3

Figs.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

EXPLANATION OF THE PLATES

Plate 1

1 & 2. Whole testis of a fully active bear show-
ing well-developed epididymis and vas
deferens. E-305, 5/11, X2/5.

3. Internal architecture of the active testis.
E-305, 5/11, X9/16.

Plate II

4. Testis of a one and one-half-year infan-
tile bear during the breeding season. Sem-
iniferous tubules are undifferentiated
cords widely dispersed in loose cellular
connective tissue. 65M, 5/21, X27.

5. Testis of a one and one-half-year infan-
tile bear during the postbreeding season.
Seminiferous tubules occupy a greater
portion of the mass of the testis than in
Plate II, figure 4, and the intertubular tis-
sue shows slightly greater interstitial cell
abundance. 46N, \-Vi, X34.

6. Testis of a two and one-half-year infan-
tile bear during the breeding season. Sem-
iniferous tubules are enlarged, lumen-
less, and occupy a major portion of the
testis. Epithelial lining is two to four cells
thick. The nuclei of most of these cells
resemble Sertoli cell nuclei; however, a
few cells hypertrophy and undergo cell
division but degenerate. Interstitial cells
are relatively abundant but contain little
cytoplasm and, consequently, occupy a
minor portion of the testis. 12M, 5/7,
X67.

7. Well-developed epididymis of the two-
and one-half-year bear during the breed-
ing season. The ducts are characteristic-
ally free of germinal elements. 12M, 5/7,
X67.

8. Abundant secretion products are common
in the nondegenerate epididymis of the
two-year infantile and prepuberal bear,
and in the preseasonal and early post-
seasonal adult bear. 18M, 10/31, X107.

9. Amorphous coagulum is common in the
nondegenerate vas deferens of the two-
year infantile and prepuberal bear, and in
the preseasonal and early postseasonal
mature bear. The presumed origin of the
material is liquefaction of secretion prod-
ucts and the breakdown products of de-
generate germinal elements. 12M, 5/7,
X67.

Plate III

Fig. 10. Testis of a two and one-half-year infan-
tile bear during the postbreeding period.
Seminiferous tubules are shrunken, base-
ment membranes thickened and wrinkled,
and the density of the central cytoplas-
mic complex increased. Interstitial tissue
is relatively more abundant than during
the breeding season with an increased
cytoplasmic-nuclear ratio. Cf. Plate II,
figure 8. 47N, 10/22, X32.

Fig. 11. Postseasonal retrogression of the two and
one-half-year infantile epididymis. Epi-
didymis is nonfunctional, stereocilia are
lost, and ducts reduced in size and sur-
rounded by a thick, fibrous connective
tissue coat. Most ducts are empty except
for a heavy coagulum. Cf. Plate II, fig-
ure 9. 52N, 10/29, X30.

Fig. 12. Testis of the prepuberal bear during the
breeding season. Seminiferous tubules are
markedly swollen and have a large lumen.
Interstitial cells are sparse, confined to
interstices between the tubules, and have
a slightly higher cytoplasmic-nuclear ratio
than in the spring infantile bear. 21M,
5/22, X27.

Fig. 13. High power showing spermatogenic activ-
ity becoming arrested in the spermatocyte
and spermatid stages in the prepuberal
bear. The formation of multi-nucleated
giant cells is a common feature. 21M,
5/22, X67.

Fig. 14. High power showing greater cytoplasmic-
nuclear ratio of the interstitial tissue of
the spring prepuberal bear as compared
to the infantile bear. Cf. Plate II, figures
2 and 3. 21M, 5/22, X67.

Fig. 15. Immature and abnormal germ elements
shed to the epididymis of the spring pre-
puberal bear. Most of the cells disinte-
grate in situ relatively few being passed
to the epididymis and even fewer to the
vas deferens. 21M, 5/22, X67.

1968]

Fig. 16.

Fig. 17.

Fig. 18.

Fig. 19.

Fig. 20.

Fig. 21.

Erickson, Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear 103

Plate IV

Prepuberal vas deferens showing limited
numbers of immature and abnormal germ
cells and the bleb-like secretions at the
tips of the stereocilia. 21M, 5/22, X38.

Vas deferens of the prepuberal bear show-
ing degenerate germ elements, including
several spermheads, and highly active
stereo-ciliated cells with abundant basally
located presumed steroid products. 59M,
5/11, X67.

Declining spermatogenic activity in the
postseasonal prepuberal bear. The tubules
are reduced in diameter, the basement
membranes thickened and wrinkled, and
the germinal epithelium and lumina re-
duced. Note the occurrence of lymopho-
cytes in the upper right tubule. 62A, 10/4,
X67.

Continued testicular decline in the fall
prepuberal bear. The lumina of the sem-
iniferous tubules are reduced markedly
and Sertoli cells predominate over the
germinal forms. Interstitial tissue is more
prominent than in the spring bear (Plate

III, fig. 13). The cells are arranged in
strands and have an increased cytoplas-
mic-nuclear ratio. Note the accumulation
of lymphocytes at the lower margin of
the upper left tubule. 73M, 10/14, X67.

Seminiferous tubules showing complete
loss of spermatogenic elements in the
postseasonal prepuberal bear. Note also
the difference in activity state between
various tubules within a given testis (PI.

IV. fig. 18). 62A, 10/4, X67.

Declining epididymis in the fall pre-
puberal bear. Epididymal ducts are be-
coming reduced in diameter and are being
enveloped by a thick, fibrous, connective
tissue coat. 62A, 10/4, X27.

Plate V

Epididymis of the late fall prepuberal bear
retaining a functional state beyond that
of the vas deferens (PI. V, fig. 23). Note
the coagulum in the ducts. 9M, 10/10,
X67.

Fig. 23. Declining vas deferens in the late fall pre-
puberal bear. The decline precedes that
of the epididymis (PI. V, fig. 22) and is
marked by a reduced duct diameter, and
disorganization of the epithelium. 9M,
10/10, X30.

Fig. 24. The prepuberal vas deferens in late de-
cline. Duct epithelium is reduced to a
single layer of low cuboidal cells. The
duct contents are degradation products
and a few abnormal cells. The ducts are
surrounded by a thick, fibrous, connective
tissue coat. Note that the decline of this
organ precedes that of the seminiferous
tubules (PI. IV, fig. 19). 73M, 10/14, X38.

early in the breeding season. The tubules
are rounded without distinct tubule lining
and are not markedly swollen as they are
later in the season. 42M, 4/18, X34.

Fig. 22.

Fig. 25. A presumed first-year adult showing sper-
matogenic activity in excess of that noted
for the prepuberal bear (PI. Ill, fig. 13)
and with more abundant interstitial tissue.
22M, 4/22, X67.

Fig. 26. Epididymis of presumed first-year adult
showing a relatively large number of
germinal elements, including a number of
sperm heads. This condition parallels that
of the fully mature bear at the approach
of the breeding season when abnormal
forms predominate. 22M, 4/22, X74.

Fig. 27. Spermatogenic activity in the adult bear

104

Zoologica: New York Zoological Society

[53: 3

Plate VI

Figs. 28 & 29. Differing states of spermatogenesis
between tubules. Interstitial tissue is
abundant, loose, and stringy with a high
cytoplasmic-nuclear ratio. The large size
of the vacuoles within the cytoplasm is
remarkable. 42M, 4/18, X67.

Fig. 30. Multinucleated giant cells commonly ob-
served in the redeveloping and declining
adult testis, and in the prepuberal and two-
year infantile testis. 6M, 5/15, X67.

Fig. 31. Well-developed state of the epididymis of
the early redeveloping bear. The sperm
load is light and abnormal forms are com-
mon. 42M, 4/18, X67.

Fig. 32. Redeveloping bear epididymis. Abundant
accumulations of material among the ster-
eocilia is a common feature. 1M, 5/4,
X67.

Fig. 33. Vas deferens of the redeveloping bear
showing early sperm load containing large
numbers of immature and abnormal ger-
minal elements. 42M, 4/18, X42.

Plate VII

Fig. 34. Testis of the fully active animal. The
seminiferous tubules are markedly swoll-
en and compressed together, with distinct
lumina and all stages of spermatogenesis.
Abnormal elements are rare. Interstitial
cells are sparse and exhibit a relatively
low cytoplasmic-nuclear ratio. 27M, 5/20,
X54.

Fig. 35. Epididymis at the height of the breeding
season containing a heavy sperm load
practically free of immature and abnor-
mal forms. 1396, 5/12, X67.

Fig. 36. Vas deferens at the height of the breed-
ing season. The heavy sperm load sug-
gests that the vas deferens is an important
organ for sperm storage in this species.
17M, 5/23, X67.

Fig. 37. Early spermatogenic decline in the adult
bear. Spermatogenesis is becoming ar-
rested in the spermatocyte and spermatid
stages and abnormal forms are increasing.
Interstitial cells are more abundant than
during the breeding season and have larger
amounts of cytoplasm (PL VII, fig. 34).
However, the cytoplasmic-nuclear ratio is
less than in the redeveloping testis (PI.
VI, figs. 28 and 29). 54N, 11/10, X89.

Fig. 38. Degeneration of the germinal epithelium
in the declining adult bear is accompanied
by tubule shrinkage, thickening of the
basement membranes, and the appearance
of intertubular spaces. 54N, 11/10, X67.

Fig. 39. Early declining testis showing reduction
of the germinal epithelium and the pecu-
liar and marked extension of the inter-
stitial tissue into the tunica albuginea.
1820, 7/17, X27.

1968]

Erickson , Mossman, Hensel, and Troyer: Breeding Biology of the Brown Bear

105

Plate VIII

Plate IX

Fig. 40. Further defoliation of the spermatogenic
epithelium of the declining bear. Note par-
ticularly the formation of multinucleated
giant cells probably by the coalesence of
developing germ cells. 1820, 7/17, X67.

Fig .41. Complete loss of spermatogenic elements
in individual tubules of the declining bear.
1831, 7/21, X30.

Fig. 42. High magnification showing that the re-
maining cell population is Sertoli cells.
1831, 7/21, X67.

Fig. 43. Marked germ cell degeneration affecting
a whole tubule while spermatogenesis is
still evident in adjacent tubules. 1831,
7/21, X30.

Fig. 44. Apparently functional epididymis in a
bear with marked spermatogenic decline
(PI. VII, fig. 37). The epididymal ducts
are largely empty except for limited ab-
normal forms, secretory products, and
coagulum. 54N, 11/10, X27.

Fig. 45. Epididymis of the declining bear. The
germinal elements are largely abnormal,
the duct diameter reduced, and the epi-
thelium disorganized. Heavy lymphocyte
invasion (left) is general. 55N, 11/10,
X67.

Fig. 46. Declining bear showing epididymis largely
free of germinal elements but with abun-
dant coagulum the source of which is
liquefaction of secretion products and de-
generate germinal cell forms. 54N, 11/10,
X67.

Fig. 47. Vas deferens of the declining bear just
prior to the major degenerative change
(PI. IX, fig. 48). 54N, 11/10, X34.

Fig. 48. Atrophic vas deferens in the postseasonal
bear. The epithelium is reduced to a single
layer of low cuboidal cells, and the ducts
are markedly reduced and surrounded by
a thick, dense, connective tissue coat. Duct
contents are largely debris and a few ab-
normal germ cells. 55N, 11/10, X34.

Fig. 49. Epididymis of the late postseasonal bear.

The ducts are reduced in diameter and
surrounded by a heavy connective tissue
coat. 55N, 1 1/10, X27.

Fig. 50. Brown bear epididymal sperm typically
retain a midpiece cytoplasmic droplet.
Sperm washed from 10 percent formalin-
fixed testis tissue and postfixed in Os04.
Overall length of sperm is 78.2 p,. E-305,
5/11.

Fig. 51 A portion of the very profuse rete testis of
the brown bear (PI. I, fig. 3). The epithe-
lium is typically a single layer of low cuboi-
dal cells. 62A, 10/4, X84.

ERICKSON, MOSSMANN, HENSEL a TROYER

PLATE I

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL & TROYER

PLATE II

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL & TROYER

PLATE III

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON. MOSSMANN, HENSEL & TROYER

PLATE IV

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL 8c TROYER

PLATE V

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL & TROYER

PLATE VI

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL & TROYER

PLATE VII

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL & TROYER

PLATE VIII

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR ( URSUS ARCTOS)

ERICKSON, MOSSMANN, HENSEL & TROYER

PLATE IX

THE BREEDING BIOLOGY OF THE MALE BROWN BEAR (URSUS ARCTOS)

NEW YORK ZOOLOGICAL SOCIETY

GENERAL OFFICE PUBLICATION OFFICE

630 Fifth Avenue, New York, N.Y. 10020 The Zoological Park, Bronx, N.Y. 10460

Fairfield Osborn
President

OFFICERS

Laurance S. Rockefeller
First Vice-President

John Pierrepont
T reasurer

Howard Phipps, Jr.
Secretary

Henry Clay Frick, II

Robert G. Goelet
Vice-President

Eben W. Pyne
Assistant Treasurer

Edward R. Ricciuti
Editor & Curator,

Publications and Public Relations

EDITORIAL COMMITTEE
Fairfield Osborn
Chairman

William G. Conway Donald R. Griffin

Lee S. Crandall Hugh B. House

Robert G. Goelet

William G. Conway
General Director

Joan Van Haasteren
Editorial Assistant

F. Wayne King
Peter R. Marler
Ross F. Nigrelli

ZOOLOGICAL PARK

William G. Conway . . . Director & Curator,

Ornithology

Hugh B. House .... Curator, Mammalogy
Grace Davall . . Assistant Curator, Mammals

& Birds

Victor H. Hutchison . . . Research Associate in

Herpetology

Joseph Bell . . Associate Curator, Ornithology

F. Wayne King .... Curator, Herpetology

Charles P. Gandal Veterinarian

Lee S. Crandall . . . General Curator Emeritus

<& Zoological Park Consultant
William Bridges Curator of Publications Emeritus
John M. Budinger . . . Consultant, Pathology

David W. Nellis Mammalogist

James G. Doherty Mammalogist

Donald F. Bruning Ornithologist

AQUARIUM

Ross F. Nigrelli Director Nixon Griffis .... Administrative Assistant

Christopher W. Coates . . . Director Emeritus Robert A. Morris Curator

Louis Mowbray Research Associate in Field Biology

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . . Director and Pathologist Harry A. Charipper . . Research Associate in

Martin F. Stempien, Jr. . . . Assistant to the Histology

Director & Bio-Organic Chemist Kenneth Gold ...... Marine Ecologist

George D. Ruggieri, S.J. . . . Coordinator of ^va Hawkins Algologist

Research & Experimental Embryologist Myron Jacobs Neuroanatomist

_ , . Klaus Kallman Fish Geneticist

William Antopol . . . Research Associate in Vincent R Liguon Microbiologist

Comparative Pathology John J. A. McLaughlin . Research Associate in

C. M. Breder, Jr. . . . Research Associate in Planktonology

Ichthyology Martin P. Schreibman . . Research Associate in

Jack T. Cecil Virologist Fish Endocrinology

INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin .... Director & Senior Fernando Nottebohm . . . Research Zoologist

Research Zoologist George Schaller ..... Research Zoologist
Peter R. Marler . . . Senior Research Zoologist Thomas T. Struhsaker . . . Research Zoologist

Jocelyn Crane . . . Senior Research Ethologist C. Alan Lill Research Fellow

Roger S. Payne Research Zoologist O. Marcus Buchanan . . . Resident Director,

Richard L. Penney .... Research Zoologist William Beebe Tropical Research Station

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

VOLUME 53 • ISSUE 4 • WINTER, 1968

PUBLISHED BY THE SOCIETY
The ZOOLOGICAL PARK, New York

Contents

PAGE

8. Host and Ecological Relationships of the Parasitic Helminth Capillaria
hepatica in Florida Mammals. By James N. Layne. Plate I; Text-figures

1-4 107

Index to Volume 53 123

Manuscripts must conform with Style Manual for Biological Journals (American Institute
of Biological Sciences). All material must be typewritten, double-spaced. Erasable bond paper
or mimeograph bond paper should not be used. Please submit an original and one copy of the
manuscript.

Zoologica is published quarterly by the New York Zoological Society at the New York
Zoological Park. Bronx Park. Bronx. N. Y. 10460. and manuscripts, subscriptions, orders for back
issues and changes of address should be sent to that address. Subscription rates: $6.00 per year;
single numbers. $1.50. unless otherwise stated in the Society's catalog of publications. Second-class
postage paid at Bronx. N. Y.

Published March 19, 1969

© 1969 New York Zoological Society. All rights reserved.

8

Host and Ecological Relationships of the
Parasitic Helminth Capillaria hepatica in Florida Mammals

James N. Layne1

(Plate I; Text-figures 1-4)

Introduction

The nematode Capillaria hepatica (Ban-
croft, 1893) typically occurs as an adult
in the liver of mammals, although in rare
instances adult worms may be found in extra-
hepatic sites. Capillaria hepatica is nearly cos-
mopolitan in its geographic distribution and ex-
hibits broad host tolerance. Although rodents
are the principal hosts, the parasite has been
found in a number of species in other orders, in-
cluding Primates, Lagomorpha, Carnivora, and
Artiodactyla. Both genuine and spurious human
infections have been reported (Ewing & Tilden,
1956; Lubinsky, 1956; MacArthur, 1924; Mc-
Quown, 1950, 1954; Otto et al, 1954).

Although C. hepatica has been recorded from
numerous hosts and localities in North America
(Table 1 ) , little is known about its status in pop-
ulations of particular species or in different hab-
itats or geographic regions. The most compre-
hensive data presently available on its incidence
in natural populations of North American mam-
mals are those of Freeman & Wright ( 1960) for
several species of rodents in Algonquin Park,
Ontario, Canada, and Layne & Griffo (1961)
for the Florida mouse, Peromyscus floridanus.
This paper presents additional information on
the hosts, geographic and ecologic distribution,
and incidence of C. hepatica in Florida.

1 Archbold Biological Station, Lake Placid, Florida
33852 and Department of Mammalogy, The American
Museum of Natural History, New York, N.Y. 10024.

Methods and Materials

A total of 2,524 specimens of 27 species of
mammals was examined. For the sake of com-
pleteness, the sample of Peromyscus floridanus
reported on previously (Layne & Griffo, 1961)
is included in the present analysis.

The majority of the animals was collected by
live trapping and killed with ether for examina-
tion. Traps were generally set in lines with one
or two traps per station and intervals of 30 to
50 ft. between stations, although in some in-
stances traps were set in a 50 or 100 ft. grid. The
remaining specimens were obtained by snap or
steel trapping, shooting, or as road kills. Data
recorded for each specimen usually included
locality, date of collection, habitat, sex, age,
body weight, external measurements, and repro-
ductive condition.

Occurrence of an infection was based pri-
marily on gross examination of the liver for the
characteristic lesions of C. hepatica (PI. I, fig.
1 ) . According to Freeman & Wright (1960), in-
fections of two or more weeks duration should
be reliably detected by this method. A further
indication of its validity is the fact that use of
a formalin-ether concentration technique in the
previous study (Layne & Griffo, 1961) did not
reveal any infections that had been overlooked
during routine autopsy. In the relatively few in-
stances where there was some doubt as to the
nature of a lesion, portions of tissue from the
area in question were teased apart in a drop of
water on a microslide, compressed with a cover
slip, and scanned at 100X for the presence of
ova (PI. I, figs. 2 and 3) or fragments of adult
worms.

107

108

Zoologica: New York Zoological Society

[53: 4

Table I

Summary of North American Host and Locality Records for Capillaria hepatica

Host

Locality and Authority

Order Primates

Ateles geoffroyi Mexico: Chiapas (Caballero & Grocott, 1952); Panama

(Foster & Johnson, 1939)

Cebus capucinus Panama (Foster & Johnson, 1939)

Order Carnivora

Canis familiaris U.S.:? Washington, D. C. and vicinity (Wright, 1930)

Order Lagomorpha

Sylvilagus floridanus U.S.: Oklahoma (Ward, 1934)

Order Rodentia

Marmota monax U.S.: Pennsylvania— zoo specimen (Doran, 1955)

Cynomys ludovicianus U.S.: Pennsylvania— zoo specimen (Weidman, 1925)

Citellus richardsonii Canada: Alberta (Brown & Roy, 1943); U.S. : Montana (Luttermoser, 1938)

Sciurus niger U.S.: Louisiana (McQuown, 1954)

Castor canadensis U.S.: Washington, D. C.— zoo specimen (Chitwood, 1934)

Thomomys talpoides Canada: Alberta (Lubinsky, 1956, 1957); U.S.: Wyoming (Dikmans,

1932; Law & Kennedy, 1932— cited from Lubinsky, 1956; Rausch, 1961)

Geomys bursarius U.S.: Pennsylvania— zoo specimen (Weidman, 1917)

Neotoma cinerea U.S.: Oregon (Rausch, 1961)

Peromyscus floridanus U.S.: Florida (Layne & Griffo, 1961; Layne, 1963; present study)

Peromyscus gossypinus U.S.: Florida (present study)

Peromyscus maniculatus .... Canada: Alberta (Lubinsky, 1956, 1957); Ontario (Freeman, 1958; Free-
man & Wright, 1960); U.S.: Alaska (Rausch, 1961); New York (Layne,
unpublished);? Washington (Dalquest, 1948); artificial infection, species
uncertain (Luttermoser, 1938)

Sigmodon hispidus U.S.: Florida (present study); Texas (Read, 1949)

Clethrionomys gapperi Canada: Ontario (Freeman & Wright, 1960); U.S. : New York

(Fisher, 1963)

Microtus chrotorrhinus U.S.: New York (Fisher, 1963)

Microtus pennsylvanicus .... Canada: Ontario (Freeman & Wright, 1960)

Synaptomys cooperi Canada: Ontario (Freeman & Wright, 1960)

Lemmus sibiricus U.S.: Alaska (Rausch, 1961)

Ondatra zibethicus Canada: Ontario (Freeman & Wright, 1960); Ontario— fur farm (Law &

Kennedy, 1932— cited from Lubinsky, 1956; Price, 1931; Swales, 1933);
U.S.: Louisiana (Penn, 1942); Maine (Meyer & Reilly, 1950); Michigan
(Ameel, 1942)

Mus musculus U.S.: Maryland (Luttermoser, 1938); Pennsylvania (Doran, 1955)

Rattus rattus Panama (Calero et al., 1950); U.S. : (Hall, 1916)

Rattus norvegicus .; Canada: Quebec (Firlotte, 1948); Panama (Calero et al., 1950) U.S.:

Illinois (Wantland et al., 1956); Maryland (Calhoun, 1962; Davis, 1951;
Habermann et al., 1954; Luttermoser, 1936;Shorb, 1931; Yokogawa, 1920);
New York (Herman, 1939); North Carolina (Harkema, 1936); Pennsyl-
vania (Herman, 1939) ; Washington, D. C. (Cram, 1928; Price & Chitwood,

1931).

Rattus rattus or

R. norvegicus U.S.: California;? Pennsylvania; Rhode Island; Washington, D.C.

(Hall, 1916)

Napaeozapus insignis Canada: Ontario (Freeman & Wright, 1960)

Order Artiodactyla

Tayassu pecari Panama (Foster & Johnson, 1939)

Sus scrofa U.S.: domestic— artificial infection (Tromba, 1959)

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

109

Infections were rated as to severity on the
basis of the appearance and abundance of le-
sions. Luttermoser (1938) considered the extent
of lesions rather than actual egg counts to be
better as an index of degree of infection. Cases
in which the liver had small, discrete, and widely
scattered lesions, often confined to a single lobe,
were regarded as light. Infections in which the
lesions were more numerous and extensively dis-
tributed were classified as moderate, whereas
those in which the lesions were almost continu-
ous and involved most or all of the hepatic lobes
were rated as heavy.

Collections were made in 22 counties of the
state, with emphasis on the north-central region.
An attempt was made to sample as wide a range
of habitats as possible. Although specimens were
taken only casually (e.g. as road kills) in some
habitat types, over 20 distinct vegetative types
were systematically live trapped or snap trapped
for small rodents. For present purposes these
may be grouped under nine major categories
which are briefly described below. More detailed
descriptions of most of these may be found in
Carr ( 1940) , Laessle ( 1942, 1958) , and Rogers
(1933). Rogers (1933) and Cooper et «/. ( 1959)
also give data on air and soil temperatures and
evaporation rates for some of the habitats in-
cluded here.

1. Hammocks. Mixed hardwood forests, in-
cluding mesophytic, upland, live oak, and cab-
bage palm hammocks. The first of these has
laurel oak (Quercus lauri folia), magnolia (Mag-
nolia grand i flora), and American holly (Ilex
opaca) as characteristic tree species. The soils
are rich in organic matter and usually moist. Up-
land hammocks are drier than the preceding,
with laurel oak, southern red oak (Q. falcata),
persimmon (Diospyros virginiana), mockernut
hickory (Cary a tomentosa), and loblolly pine
(Pinus taeda) being representative tree species.
Live oak hammocks, with live oak (Q. virgini-
ana) as the dominant tree and generally sparse
ground cover, are also comparatively xeric. Cab-
bage palm hammocks sampled consisted of
dense stands of cabbage palm (Sabal palmetto)
and other trees and shrubs. Soils were rich and
moist.

2. Pine-oak woodlands. Generally open stands
of pines and oaks occurring on sandy well-
drained soils, often on low ridges or hills.
Ground cover is generally rather sparse result-
ing in frequent patches of exposed sand. Two
types of pine-oak woodlands were sampled in
the study, one with longleaf pine (Pinus palustris)
and turkey oak (Q. laevis) as the chief tree spe-
cies, and the other with slash pine (P. elliottii)
and turkey oak. The first is widespread through-

out much of Florida, while the second is re-
stricted to a relatively small area in the south-
central part of the peninsula. Slash pine-turkey
oak woodlands are somewhat intermediate be-
tween longleaf pine-turkey oak woodlands and
sand pine scrub, described below, in over-all spe-
cies composition and environmental conditions.
Only one of 18 different pine-oak woodland
habitats from which collections were made rep-
resented slash pine-turkey oak.

3. Scrub. Stands of scattered pines with a dense
understory of small trees and shrubs, a number
of which are sclerophyllous. Herbaceous ground
cover is typically sparse, and open sandy patches
are common. This comparatively xeric vegeta-
tive type occurs on excessively well-drained fine
sandy soils. Sand pine (P. clausa) is generally re-
garded as the most characteristic species of true
scrub, and myrtle oak (Q. myrtifolia), live oak,
and Chapman oak (Q. chapmanii) are the prin-
cipal elements of the shrub layer. For purposes
of the present study, one station (Levy-19) with
all of the characteristics of scrub except for
slash instead of sand pine is included under this
category.

4. Flatwoods. Level pinelands found on poorly
drained soils. The particular species of pine
present as the dominant tree depends upon
edaphic conditions. Longleaf pine is typical of
better drained sites and slash pine of the more
moist areas. Ground cover in flatwoods is usual-
ly dense, consisting of varying proportions of
grasses, palmettos, and forbs. Shrubs and decid-
uous trees range from rare to abundant enough
to form a distinct understory, depending on fre-
quency of fire.

5. Swamps. Cypress (Taxodium distichum)
and hardwood swamps and bayheads. The latter
are low, moist to wet, dense stands of various
trees, shrubs, and vines. Sweetbay (Magnolia
virginiana) and loblolly bay (Gordonia lasian-
thus) are common species of bayheads.

6. Wetlands. Moist to wet prairies, marshes,
and lush herbaceous and brushy borders of road-
side ditches and ponds.

7. Dunes. Coastal sand dunes generally sparse-
ly vegetated with sea oats (Uniola paniculata)
and various other grasses and forbs.

8. Ruderal. Old fields in early successional
stages, pine plantations, weedy or grassy road
shoulders, and fence rows.

9. Miscellaneous. Buildings and lawns or gar-
dens in agricultural or residential areas. Except
for snap trapping for Mus musculus and Rattus,
no systematic collecting was carried out in these
habitats.

Although in many cases a given locality was
sampled only once, periodic trapping of selected

110 Zoologica: New York Zoological Society [53: 4

Table II

Species and Numbers of Florida Mammals Examined for Capillaria hepatica

Asterisk Indicates Habitat or Locality from which Infections were Recorded

Species

No.

exam.

Habitat 1

Locality 2

Didelphis marsupialis

6

HI, PI, SI, U3

A3, Hl,Lvl,Ul

Scalopus aquaticus

5

PI, M2, U2

A4, Lai

Cryptotis parva

10

PI, F6, Wl, R2

A10

Blarina brevicauda

6

H3, W3

A5, HI

Sylvilagus floridanus

12

H7, SI, FI, R2, U1

A3, Bl, H7, Lei

S. palustris

4

W3, R1

Al, Lvl, Pi2

Sciurus carolinensis

11

H4, M2, U5

A8, HI, Lvl, Pil

S. niger

2

PI, FI

Al, Lvl

Glaucomys volans

2

Rl, Ml

A2

Geomys pinetis

9

P3, SI, R4, U1

Al, Gl, H2, Lv4, Stl

Oryzomys palustris

32

S7, F5, W17, R2, Ml

A7, H6, Le5, Lvl3

Peromyscus polionotus

43

P3, S2, FI, D34, R3

A3, B3, E30, H6, 01

P. gossypinus

254

H39, P19, S126*, F9,

A55, B 12, Fr7, H38,

P. floridanus

1312

Sw32*, Wl, Rl, U5
H246*, P400*, S600*,

Lvl41*, Stl

A634, C4, G18, H75*,

Reithrodontomys humulis

2

F61, R16, U1
P2

Lv570*, Pa6, Pul, St4*
A2

Ochrotomys nuttalli

12

H3, P5, S4

A4, H2, Le2, St4

Sigmodon liispidus

427

H3, P32, S200*, F66,

A123, B4, E6, H36*, Lell,

Neotoma floridana

2

Swl8, W40, D9, R51*, U8
HI, SI

Lv213, Pi27, St6, Wl
Lvl, T1

Neofiber alleni

9

W9

A3, H5, Pul

Rattus norvegicus

1

Ml

Dwl

Rattus rattus

18

R4, Mil, U3

A2, Du2, M2, Pi 12

Mus musculus

328

H7, PI, S3, W13, D3, R2,

A303, H3, Lel9, 03

Urocyon cinereoargenteus

6

M297, U2
HI, PI, U4

A3, FI, HI, Pul

Procyon lotor

5

SI, Rl, U3

Al, HI. Lvl, M2

Mustela frenata

1

U1

Dul

Mephitis mephitis

1

D1

Del

Lynx rufus

4

HI, Rl. U2

A2, Del, T1

1 Habitat Key: H— hammocks, P— pine-oak woodlands, S-scrub, F— flatwoods, Sw—

swamps, W— wetlands, D— dunes, R— ruderal, M— miscellaneous, U— un-
known. Number of specimens examined from given habitat type follows
abbreviation.

2Locality (county) Key: A— Alachua, B— Brevard, C— Clay, De— DeSoto, Du— Duval, E— Escambia,

F— Flagler, Fr— Franklin, G— Gilchrist, H— Highlands, La— Lake, Le— Lee,
Lv— Levy, M— Monroe, O— Okaloosa, Pa— Palm Beach, Pi— Pinellas, Pu—
Putnam, Sa— Santa Rosa. St— St. Johns, T— ' Taylor, W— Walton, U— Un-
known. Number of specimens examined from given county follows ab-
breviation.

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

111

pine-oak woodland, scrub, flatwoods, hammock,
and ruderal habitat types was conducted in
order to obtain data on seasonal and long-term
trends in prevalence of C. hepatica in small
mammal populations. The numbers of samples
obtained from a given station ranged from two
in a two-year period to 21 over an eight-year
interval.

Results

Host and geographic distribution. The species
and numbers of mammals examined for C. hepa-
tica and localities represented are listed in Table
2. Eight per cent of the 2,524 specimens and
1 1% (3). of the 27 species examined had Capil-
laria infections. The species and localities from
which infections were recorded included the
Florida mouse, Peromyscus floridanus, from
Highlands, Levy, and St. Johns Counties; the
cotton mouse, Peromyscus gossypinus, from
Levy County; and the cotton rat, Sigmodon his-
pidus, from Highlands and Levy Counties. C.
hepatica has not been previously reported from
the cotton mouse, and Sigmodon constitutes a
new host record for Florida and the southeastern
United States generally.

Based on total numbers of specimens exam-
ined from all habitats combined, incidence of
infection was 15.4% in P. floridanus, 6.3% in
P. gossypinus, and 10.5% in Sigmodon. Because
infection rates in populations in the same and
different habitats varied considerably and sam-
ple sizes are unequal, these values probably do
not reflect the true status of C. hepatica in
these species as well as a mean incidence calcu-
lated from individual population means. Ex-
pressed in the latter manner, incidence was
13.2% in Florida mice, 2.5% in cotton mice,
and 6.3% in cotton rats.

Considering only those localities from which
infected animals of any species were collected
(Table 3), prevalence of C. hepatica was 34.2%
in Florida mice, 14.8% in cotton mice, and
and 21.1% in cotton rats. As there are signifi-
cant differences in the probability of infection in
different age groups, the above values may be
influenced by different age compositions of the
samples of each species. Thus incidence based
only on adult animals is probably a more satis-
factory value for comparative purposes. Adult
infection rates in P. floridanus, P. gossypinus,
and Sigmodon were 37.7%, 14.8% and 30.8%,
respectively. Neither the total nor adult infec-
tion rates in P. floridanus and Sigmodon differ
statistically when tested by chi-square (P > .05),
but both species differ significantly from P.
gossypinus (P < .005).

Florida mice and cotton rats tended to be
more heavily infected than cotton mice (Text-

fig. 1). The difference in severity of infections
between P. gossypinus and the other species is
significant (P < .05), but P. floridanus and Sig-
modon do not differ significantly in the propor-
tions of light, moderate, and heavy infections.

With one exception, all infections of C. hepa-
tica observed in this study were localized in the
liver as is typical. However, one cotton rat with
a massive early infection of the upper hepatic
lobes also had scattered adult worms in the
mesentery of the spleen and a “knot” of worms
about 14 -inch in diameter in the mesentery
adjacent to the liver. The worms at the latter
site appeared to be dead and in the process of
becoming calcified. No ova were observed when
teased fragments of the worms were examined
microscopically.

Infected livers of P. floridanus differed mark-
edly in appearance from those of the other
species. Even in the case of the lightest infec-
tions, the livers of Florida mice became prom-
inently lobulated, contrasting greatly with the
normally smooth surface of the organ (PI. I, fig.
1). Although the liver of one heavily infected
cotton rat had a slightly lobulated appearance,
all other infected livers of Sigmodon and cotton
mice were normal in appearance except for the
presence of lesions.

Ecological distribution. Infected mice were
recorded from five (hammocks, pine-oak wood-
lands, scrub, swamps, and ruderal) of the nine

floridanus gossypinus

Text-fig. 1. Severity of infections of C. hepatica in
three Florida rodent species.

112

Zoologica: New York Zoological Society

[53: 4

•2

'S.

a

O

u-

o

w

u

z

w

Q

u

z

&

3

T3

<.

C

o

•ts

o

c

.§)

Z S

o

% ><

o

^Z

&

J2 Z

z d

c S

o

'C.

cd

O

*-z

22 c o
* « J

o

VO

O

OO

O

q i i

o

O

i °

o 6

m

o

d

o

rd

»/n

*/n

o

r"H

o

"d-

o

r-

| |

i °

rd

1 1

i

V O

"d-

rd

ON

o

VO | |

o

i ^

T“l

T_H

rd

l/n

1 1

i

o

p

O

©

cd

q 1 1

o

i °

m

O

o

»/n

m

o

1

o

o

(N

m | |

_

i °

CO

i i

VO O rd

O

vo r- | |

i— i

ON

1 N

*— i cn

00 1 1

o

o

1 ^

00

© 1 1

o |

p

r-

1-H

i— I

rd

o

o

i ^

rd

° 1 1

O I

1

1

1 1

1

1

VO

i 00

*/n

ra i |

1

1

1

on

cs 1 1

1

1

O

o

1 9

° 1 1

° 1

1 ^

o

'd*

T_ 1

o

o

1 ^

rd

O | |

° 1

1 ^

VO

l 2

O

<N 1 |

r-~ 1

1 <*

<N ' '

ON

on

■*t

Cd

rd

o

O

m

i

O

1

00

vd

cd

cd

on

OO

o

d

ON

d

c*i

vo

VO

y—*

VO

m

o

o

o

T— '

Cd

on

Cd

VO

ON

_

rr

i

1

Cd

1— 1

rd

i

1

1

00

cc

00

r-

OO

vo

cr

1

1

r-

r-

r-

m

1

,—l

OO

p

p

cn

cd

p

O

o

p

1

o

|

vd

00

m

ON

vd

d

o

o

m

Cd

vo

T—l

on

o

o

o

,—l

y~> 1

Cd

m

VO

o

^t

1

|

<N

Cd

m

,_H

Os

_

m

m

rd

m

1

|

r— <

O

r-H

00

»o

Cd

rd

S'

cd

/«-v

Cd

m

On

O

CL> Tt-

T3

cd

1

rd

rd

|

1

C/5

> rd

(D

1

C/5

*”•

|

^ Js

Cd c/5

^3 1

-C

-a

O

ON

00

tn

cn

cd *o

u T3
ojj 0-

^ — ' C/5

o .

c

i— <

m

1

C

C

i) O c

z~: c

^ £

-f2

|

|

|

|

x:

X

c >,J3

cd

O 42

2

2

Of)

On

>

>s

>

On

>

On

>

On

>

o

o

H-J

a® j=

M)

cd X,

u Of)

o o

E ^

Swamp

Levy

2

<D

CD

<d

<U

hJ

a>

hJ

GO

55

r- v- ■

^ BX

cd w ^

•a X

3

1 ^

55

K

" Typical scrub vegetation but with slash instead of sand pine as noted in text.

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

113

major habitat categories sampled (Table 2). In-
cluding all mammals obtained from all habitats
combined, over-all infection rates were 0.3% in
hammocks, 1.1% in pine-oak woodlands, 26.0%
in scrub, 2.0% in swamps, and 1.1% in ruderal
situations. The data thus reveal a strong pre-
ponderance of infections in scrub habitats. This
indication of habitat specificity in the occurrence
of C. hepatica is further strengthened when
the data are examined in greater detail. All non-
scrub habitats from which infected animals were
collected were located within dispersal distance
of scrubs with moderate to high incidence of
C. hepatica.

Data from a locality in Levy County (Levy
—19) that was studied over an eight-year period
provide a particularly clear example of the close
correlation between the occurrence of C. hepa-
tica and scrub environments. The scrub vegeta-
tion at this site was restricted to a slightly ele-
vated area and graded abruptly into surround-
ing low bayhead and marsh habitats (Te-xt-fig. 2) .
Of the 418 mammals examined from this area,
46% of those collected in the scrub were in-
fected as compared to only 2% from the adja-
cent habitat types. It is also likely that the
infected animals from the adjoining habitats had
acquired their infections in the scrub.

Table 3 summarizes data on prevalence of
Capillaria in the three rodent species at all sta-
tions in which infections occurred in at least

one of the species. The incidence of infections is
strongly correlated with the extent to which a
species occurs in scrub and its general abun-
dance in this habitat. The data also indicate con-
siderable intraspecific variation in infection rates
within a given habitat type. Such variability
tends to be greater in Sigmodon and P. gossy-
pinus that in P. floridanus, with the last species
also exhibiting consistently higher infection
rates. There also appears to be some tendency
in at least the Florida mouse and cotton rat for
infection rates at different stations to vary in the
same direction.

Geographic differences in infection rates in
different species or populations of the same
species are not apparent from the present data.
There was as much variation in inter- and intra-
specific infection rates between stations 1, 3, 10,
19, and 28 in Levy County, all of which are
within 5 miles of one another, as between those
in different parts of the state.

Multiple samples from the same station show
yearly fluctuations in incidence of Capillaria.
However, there is some evidence from two P.
floridanus scrub populations sampled over an
eight-year period that infection levels in a given
population may vary within relatively narrow
limits for a considerable span of time (Text-fig.
3). The two populations in question are located
only about 5 miles apart, yet are separated by
unsuitable habitat.

100— i

so—

% —
4 0—

20—

7*V/ 4k.

-*■

0 s

SCRUB

L

F G S

F G S

Text-fig. 2. Relationships of habitats at station Levy- 19 and host species composition in each. Corre-
sponding data on infection rates given in text. Density of stippling indicates degree of soil moisture;
vertical scale somewhat exaggerated. Symbols: F — P. floridanus, G = P. gossypinus, S = Sigmodon
hispidus.

Levy 10

Levy 19

Text-fig. 3. Prevalence of C. hepatica infections in adult P. floridanus in two nearby scrub habitats from
1957 to 1964. Samples examined each year are given in parentheses following year.

Sex and age differences in infections. Table 4
gives the sex and age distribution of C. hepatica
in infected host populations. The age classes
used for the two Peromyscus species are based
on pelage features. Mice assigned to the juvenile
age class were still in the full gray juvenile pel-
age and showed no sign of molt on the dorsum.
Individuals undergoing the dorsal phase of the
postjuvenile molt were regarded as subadults,
while mice in which the postjuvenile molt had
been completed were assigned to the adult class.
Approximate chronological ages corresponding

to these pelage phases are under 6 weeks for
juveniles, 6 to 13 weeks for subadults, and over
14 weeks for adults (Layne, 1966; Pournelle,
1952). Age classes of cotton rats were based on
body weight as follows: juvenile, less than 40 g;
subadult, 40-70 g; and adult, above 70 g.

A relationship between age and infection rate
is evidenced by all species. No parasitized juve-
niles were found in any species, although sam-
ples of this age class are admittedly small.
Adult infection rates are significantly higher
(P < .05) than those of subadults in both Florida

Table IV

Sex and Age Differences in Capillaria hepatica Infections

Species

Age class

Male
N N

ex. inf.

%

Female
N N

ex. inf.

%

Total
N N

ex. inf.

%

P. floridanus

Juvenile

10

0

0

16

0

0

26

0

0

Subadult

40

9

22.5

28

4

14.3

67

13

19.4

Adult

236

98

41.5

262

91

34.7

498

189

38.0

P. gossypinus

Juvenile

4

0

0

2

0

0

6

0

0

Subadult

3

0

0

8

0

0

11

0

0

Adult

37

12

32.4

26

4

15.4

63

15

23.8

Sigmodon

Juvenile

17

0

0

11

0

0

28

0

0

Subadult

19

3

15.8

33

2

6.1

52

5

9.6

Adult

57

16

28.0

52

23

44.2

109

39

35.8

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

115

mice and cotton rats. An age effect in infection
rate also appears to persist into the adult class
of P. floridanus. Thirty-one nonparasitized males
from infected populations had a mean weight
of 28.2 g, while mean weight of 28 parasitized
males was 35.0 g. Corresponding values for 37
noninfected females and 12 infected females
were 29.5 g and 34.7 g respectively. It thus ap-
pears that infections are more prevalent in older
adults, assuming at least a rough correlation be-
tween weight and age in these field populations.

With the exception of adult cotton rats, more
males than females in each species and age group
had C. hepatica. Although the differences are
not statistically significant (P > .05) in any case,
they are suggestive and may reflect larger activ-
ity ranges in males than females.

Seasonal variation in infections. Monthly in-
fection rates in adult P. floridanus, P. gossy-
pinus, and Sigmodon from one extensively sam-
pled scrub locality (Levy— 19) are shown in Text-
fig. 4. The graphs are based on composite sam-
ples representing eight years of collecting and
thus reveal only average trends.

In the Florida mouse, infections appear to
be relatively low during the winter months and
to increase during summer and early fall. Al-
though the data are limited, a similar seasonal
trend is suggested for P. gossypinus. Peak levels
of infection in Sigmodon, however, appear to
come in the spring, although incidence during
the winter months tends to be low as in the case
of the two Peromyscus species.

Relationship between incidence and host den-
sity. Population levels and infection rates in two
scrub populations of P. floridanus are compared
in Table 5. The number of mice captured per
100 trap-nights is employed as the index of

Text-fig. 4. Seasonal variation in adult infection
rates at a single scrub station (Levy- 19).

Table V

Relationship Between Population Level and Incidence of C. hepatica in Adults
of Two P. floridanus Populations

Station 10

Station 1 9

Pop. index

Adult

Pop. index

Adult

Month and

(mice/100

Inf. rate

Month and

(mice/100

Inf. rate

year

trap-nights)

N

(%)

year

trap-nights)

N

(%)

Jan. 1961

49*

12

16.6

May 1960

20*

10

70.0

Jan. 1962

53*

2

0

May 1961

28*

17

70.5

May 1963

6*

7

85.7

Feb. 1959

48**

10

0

Feb. 1961

91*

6

0

July 1957

5*

4

100.0

May 1957

26**

36

2.8

Aug. 1960

17*

14*

9

80.0

75.0

July 1961

7

May 1961

24*

7

14.3

Apr. 1962

51*

9

0

Oct. 1960

21*

11

66.6

May 1963

16*

8

62.5

Sept. 1961

2*

1

100.0

Aug. 1958

38**

10

10.0

Dec. 1960

17*

9

70.0

Aug. 1960

23*

5

0

Dec. 1962

10*

5

62.5

*Trapline with 2 traps/station
**100 ft. grid, 1 trap/station

116

Zoologica: New York Zoological Society

[53: 4

mouse abundance. The combined data for the
two localities suggest a negative correlation be-
tween host abundance and degree of parasitism
in this species. Station 10, with an over-all popu-
lation index of 42 mice per 100 trap-nights, had
a mean infection rate of 10%, as compared to
station 19 with a mean population index of 14
and an adult infection rate of 78%. A similar
relationship is evident within each population
when comparison is made between abundance
and infection rates at the same season in differ-
ent years.

Discussion

This study indicates that Capillaria hepatica is
a relatively localized and uncommon parasite of
Florida mammals, yet probably is rather widely
distributed in the state. While records were ob-
tained from only three of the 22 counties in
which collecting was done, these localities are
widely separated, Levy and St. Johns Counties

being located on opposite sides of the northern
part of the peninsula and Highlands County in
the south-central part of the state.

Levels of infection previously reported in
rodents in North America vary greatly (Table 6),
although variations in sampling techniques and
methods of reporting results make critical com-
parisons and interpretations difficult. In general,
the present data suggest that the parasite occurs
with greater frequency in Rattus, particularly
urban populations, than in native species. Com-
pared to other native rodents, the incidence of
C. hepatica in the three Florida species is rela-
tively high. This is particularly true of P. flori-
danus and Sigmodon. The greater infection rate
given for the Florida mouse in this paper com-
pared to that reported earlier (Layne & Griffo,
1961) is due to additional collecting of this
species being concentrated in areas with higher
rates of infection. This in itself illustrates the

Table VI

Reported Infection Rates of C. hepatica in North America Rodents

Species

Locality

Incidence,
per cent

Source

Sciurus niger

Louisiana

3.7

McQuown, 1954

Peromyscus maniculatus

Ontario

9.4

Freeman & Wright, 1960

P. maniculatus *

Washington

“virtually all”

Dalquest, 1948

P. floridanus

Florida

2.9

Layne & Griffo, 1961

P. floridanus

Florida

15.4

Present study

P. gossypinus

Florida

6.3

Present study

Sigmodon hispidus

Florida

10.5

Present study

Clethrionomys gapperi

Ontario

2.8

Freeman & Wright, 1960

Ondatra zibethicus

Louisiana

less than
10-ca. 50

Penn, 1942

0. zibethicus

Maine

17+

Meyer & Reilly, 1950

O. zibethicus

Michigan

3

Ameel, 1942

Mus musculus

Maryland

4

Luttermoser, 1938

Rattus norvegicus

Quebec

6

Firlotte, 1948

R. norvegicus

Maryland

85.6

Luttermoser, 1936

R. norvegicus

Maryland

47.9

Shorb, 1931

R. norvegicus

Maryland

53.3

Calhoun, 1962

(semi-wild)

R. norvegicus

Maryland

35

Habermann et al, 1954

(semi-wild)

R. norvegicus

Maryland

94.1

Davis, 1951

R. norvegicus

New York

73.5

Herman, 1939

R. norvegicus

Pennsylvania

<30

Herman, 1939

R. norvegicus

North Carolina

2.6

Harkema, 1936

R. norvegicus

Washington, D. C.

77

Price, 1931

R. norvegicus (396)

Panama

12

Calero et al, 1950

-f R. rattus (4)

"Parasite not identified in paper but from description almost assuredly C. hepatica.

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

117

problems involved in attempting to assess the
real significance of differences in incidence
values given by various authors.

Herman (1939) noted that few of the infected
Rattus norvegicus examined from the New York
Zoological Park had the entire liver affected by
lesions and that in only 14% was more than
half of the organ involved. Luttermoser (1936)
similarly observed that infections in Baltimore
rats were of low intensity. Assuming generally
comparable criteria of extent of infection in the
above and present studies, both Florida mice
and cotton rats appear to have a greater propor-
tion of heavy infections than Rattus. Dalquest
(1948), presumably referring to C. hepatica.
stated that virtually all P. maniculatus collected
on Jones Island in the San Juan Island group off
the coast of Washington had greatly swollen
livers with a yellow, crystalline appearance. This
description would appear to fit the category of
heavy infection as used in this study. These
limited data suggest that native rodents may tend
to acquire more intense infections of C. hepatica
than Rattus.

An earlier study (Layne & Griffo, 1961) re-
vealed C. hepatica to be almost entirely confined
to populations of the Florida mouse living in
scrub or similar habitat types. The present data,
representing numerous other potential host spe-
cies and more extensive locality and habitat sam-
pling, provide further confirmation of a highly
restricted ecological distribution for this parasite
in Florida. All of the infected specimens of the
three host species were collected in scrub or simi-
lar habitats or from other habitats located near
scrub from which infected animals could readily
disperse. The over-all incidence of infection in
each species also is clearly correlated with the ex-
tent to which it is found in scrub habitats. Of the
three species, the Florida mouse is most charac-
teristic of scrub and has the highest incidence of
C. hepatica. Cotton mice and cotton rats occur
more commonly in other habitats. In the present
study, cotton rats were more abundant in scrub
than cotton mice. In addition, live trapping data
from permanent study plots indicated that cot-
ton rats living in scrub tend to be more sedentary
than cotton mice, many of the latter trapped in
this habitat appearing to be transient individuals.
This may explain the relatively low incidence
and intensity of C. hepatica infections in P.
gossypinus even from scrub stations with un-
usually high incidence of the parasite in P. flori-
danus and Sigmodon populations.

The pronounced habitat specificity of C. hepa-
tica in Florida is not evident in other parts of
the species’ range. Rather, the great ecological
diversity represented by its known hosts in North

America (Table 1) and elsewhere together with
its extensive geographic range suggest broad en-
vironmental tolerance. Furthermore, specific in-
formation on habitat relationships in other parts
of the range indicates that, unlike the case in
Florida, the parasite tends to have higher pre-
valence in more moist habitat types (Freeman &
Wright, 1960; Pavlov, Skrjabin et al., 1957,
cited from Freeman & Wright, 1960).

The basis of the marked restriction in the
ecological distribution of C. hepatica in Florida
is far from clear. Among the factors that might
be involved are distribution of suitable hosts,
substrate characteristics, host population dy-
namics, methods of egg-release and dissemina-
tion, and feeding habits of potential hosts.

There does not appear to be any strong cor-
relation between the ecologic distribution of C.
hepatica and mammalian hosts in Florida. The
wide variety of known hosts of this parasite in-
dicate that any of the small rodents involved
in this study would serve as a suitable host.
Moreover, the species often found infected in
scrub or similar habitats also commonly oc-
curred singly or in combination in other habitat
types from which C. hepatica was only rarely or
never recorded. In a number of cases such
habitats were actually continuous with scrub
with a high incidence of the parasite.

It is possible that scrub soils provide better
conditions for survival or embryonation of eggs
released from livers of the host than those of
other habitats included in this survey. In view
of the wide geographic range of the parasite and
the variety of its recorded hosts, it does not
seem likely that substrate conditions would have
such an important influence on its ecologic dis-
tribution in Florida. Furthermore, if substrate
conditions are so critical, it is difficult to recon-
cile the evidence for preferences for moist con-
ditions in other parts of the range with high
incidence in very sandy, highly drained soils in
Florida. The picture is further complicated by
the fact thatC. hepatica appears to be completely
absent from longleaf pine-turkey oak woodlands
which are also characterized by sandy, well-
drained soils, although there are important struc-
tural and chemical differences between the soils
of these habitats and true scrub.

Freeman & Wright (1960) believed that popu-
lation density played an important role in deter-
mining the incidence of C. hepatica in small
rodents in a local area in Ontario. Although
host population level may be an important factor
influencing establishment or persistence of C.
hepatica in certain habitats included in this
study, there is no convincing evidence that it is
a major cause of the observed habitat distribu-

118

Zoologica: New York Zoological Society

[53: 4

tion of the parasite. Scrub and related habitat
types in which infections occurred did not con-
sistently support higher populations of small
rodents than other habitats from which infec-
tions were never reported; nor does there seem
to be any significant correlation between host
abundance and infection rate at different sta-
tions in scrub habitat. In fact the data suggest
an inverse relationship. On the other hand, small
rodents are often scarce in longleaf pine-turkey
oak woodlands, and although other aspects of
the habitat might be suitable for C. hepatica, the
low host density might make establishment and
maintenance of the parasite difficult. Opposed
to this argument, however, is the fact that some
of the scrub habitats and the single slash pine-
turkey oak woodland station studied over a
period of several years at times had populations
as low or lower than some longleaf pine-turkey
oak stations yet maintained relatively high levels
of C. hepatica infections. Possibly the interval
of host scarcity is the critical factor in this situ-
ation; evidence indicates that periods of contin-
uous low population density may be considerably
more prolonged in longleaf pine-turkey oak
habitats.

Laboratory studies on the life cycle of C.
hepatica, reviewed by Freeman & Wright (1960)
and Wright (1961), indicate that infections are
acquired through ingestion of infective ova re-
leased from the liver of another host through
cannibalism, predation, or natural death and
decomposition. Ova freed through decomposi-
tion of the liver require a longer period for
embryonation and have lower viability than
those passed through the alimentary tract of
another animal.

Although little is known of the actual details
of the life cycle under natural conditions, it is
logical to assume that cannibalism and predation
are the most important of the commonly ac-
cepted egg-disseminating mechanisms. Freeman
& Wright (1960) concluded that cannibalism in
communal winter nests, rather than predation,
was the chief source of infections in deer mice in
Ontario, although their evidence was entirely cir-
cumstantial and subject to other interpretations.

Cannibalism does not appear to be an impor-
tant egg-releasing mechanism in Florida mam-
mals. Field data provide no evidence of com-
munal nesting or cannibalism in any of the three
host species, nor is there any reason to suppose
that if these phenomena were common they
would be more prevalent in scrub than other
habitat types in which the species are found.
This leaves predation as the more likely method
of egg-dissemination.

All other things being equal, higher predation

levels would seem to contribute to maintaining
C. hepatica in a small mammal population by
insuring a continuous supply of infective ova.
Present data are far too limited to allow defini-
tive conclusions concerning relative predator
abundance in the various habitats sampled. How-
ever, casual observations gave the impression
that potential small mammal predators such as
bobcats, foxes , raccoons, opossums, skunks,
feral pigs, snakes, and birds of prey tended to
be more common in scrub than in many of the
habitats studied. In fairness, it should be noted
that predator sign was probably more easily ob-
served in scrub than in some of the other habitats
studied. As in the case of rodent population
levels, however, abundance of potential pre-
dators was by no means associated only with
scrub habitats and thus cannot fully explain the
high incidence of C. hepatica there. Further-
more, as some of the scrubs with high incidence
of C. hepatica were small, many of the verte-
brate predators occurring there probably ranged
over other habitat types in the vicinity, thus
providing opportunity for wider dispersal of ova
from infected rodents eaten.

Certain aspects of the feeding behavior of
Florida mice, cotton mice, and cotton rats in
scrub seemed to be more specific to this habitat
type than the factors mentioned above. The two
Peromyscus species are essentially granivorous
in their dietary whereas the cotton rat is typically
herbivorous. However, field observations indi-
cate that all three species have generally similar
feeding habits when living in scrub. Acorns ap-
pear to be an important food source in scrub.
They are generally abundant in late fall and
early winter and decline steadily through winter
and spring. The rodent populations follow the
same annual cycle of abundance and decline. As
mast supplies become scarce in late winter and
spring, there is much evidence of digging by the
rodents for food. This behavior would seem to
increase the probability of exposure to infective
ova in the soil.

A preliminary attempt to obtain actual evi-
dence for this hypothesis in a scrub habitat (Levy
—19) with an unusually high level of C. hepatica
was unsuccessful. Fifty soil samples from the
surface to a depth of about 2 inches were col-
lected, many from within and around pits dug
by foraging mice, and examined microscopically
for ova. In addition, feces and stomach contents
of 30 P. floridanus, the species with the highest
incidence of infections at this station, were also
surveyed for eggs in the hope that the presence
of ingested eggs together with food remains
would provide some clue to the source of infec-
tions. No ova were detected in either case.

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

119

None of the above factors seems adequate
alone to account for the narrow habitat speci-
ficity of C. hepatica. It is possible, therefore, that
the ecologic distribution of C. hepatica is due to
a particular combination of such factors which
has a much higher probability of occurrence in
scrub than other habitat types. This set of condi-
tions might include 1) substrate suitable for
survival of ova, 2) sufficiently stable and high
enough populations of potential host species to
insure continuance of the cycle of parasitism,
3) an abundant enough supply of small mammal
predators to insure an adequate supply of infec-
tive ova, and 4) host foraging behavior conduc-
ive to exposure to ova. Although this explanation
is more consistent with the present data than a
single factor model, it is still difficult to con-
ceive of an interaction of a number of factors
being responsible for the sharply delimited hab-
itat distribution of C. hepatica. Thus, although
such a combination of conditions as noted above
may be generally prerequisite for the occurrence
of the parasite, some additional feature unique
to scrub and related habitats may actually be
the critical factor permitting its establishment
and maintenance.

Such a factor may be the presence of some
invertebrate species, most likely an insect, in
scrub habitats which may play a key role in the
transmission of the parasite. It may supplement
or replace vertebrate predators as the chief egg
releasing and disseminating agent through feed-
ing on dead mice or on feces of predators which
have eaten infected mammals. It may increase
probabilities of accidental ingestion of C. hepa-
tica eggs by potential hosts simply by contribut-
ing to broader dispersal of infective eggs in the
soil through its feces, or may be an even more
effective agent in maintaining the host-parasite
cycle through actually inhabitating and contam-
inating the nests of rodents. It is also possible
that mice might acquire infections directly by
feeding on the insect, particularly in times of
food shortage. The suggestion of seasonal vari-
ation in incidence of infections shown by the
present data is of interest in this connection. In
all host species, peak levels tend to occur in the
spring or summer months and low rates during
the winter months. Higher incidence thus corre-
lates with both a period of food scarcity in which
rodents might take more insects and warm
weather favoring insect activity.

With the exception of a study by Momma
(1930), the possible role of insects or other in-
vertebrates in dissemination of ova of C. hepa-
tica appears to have been ignored. Momma ex-
perimentally demonstrated that flies exposed to
C. hepatica eggs both ingested and picked them

up on the body and that the ova embryonated
normally after passage through the intestine of
the insect. He concluded that high summer and
low winter infection rates in Rattus norvegicus
in urban areas of lapan, together with low inci-
dence of eggs (in only 5 of 503 specimens
examined) in intestines of cats used for rat ex-
termination, was evidence that flies were the
primary method of egg dissemination.

Summary

A total of 2,254 specimens of 27 species of
Florida mammals was examined for infections
of the liver-inhabiting nematode Capillaria hepa-
tica (Bancroft, 1893). Collections were made in
22 counties of the state and in nine major habitat
types.

Infections were recorded in three rodents,
Peromyscus floridanus, P. gossypinus, and Sig-
modon hispidus, from three widely separated
localities. P. gossypinus constitutes a new host
record for the parasite and Sigmodon a new
record for Florida and the southeastern U.S.
generally.

Over-all incidence of infection was 15.4%
in P. floridanus, 6.3% in P. gossypinus, and
10.5% in Sigmodon. Prevalence in P. floridanus,
P. gossypinus, and Sigmodon from stations posi-
tive for infections in any species was 34.2%,
14.8%, and 21.1%, respectively. Severity of
infections was greater in P. floridanus and Sig-
modon than in P. gossypinus.

Infections were largely restricted to scrub and
related habitats, the rare occurrences in other
major habitat types in most cases being explain-
able on the basis of dispersal of infected animals
from scrubs. The basis of the narrow habitat
specificity of the parasite in Florida, which is
not apparent in other parts of its range, is un-
known. Such factors as host distribution, sub-
strate conditions, host and predator population
levels, and feeding habits of hosts do not, either
singly or in combination, appear to adequately
account for the marked restriction of the para-
site to scrub environments; and the possibility
of some insect being the key factor is suggested.

In all three host species, infection rates were
greater in older age classes, and in all age groups
males tended to have a higher infection rate than
females, although these differences were not sta-
tistically significant. Although seasonal variation
in infection rates was not pronounced, incidence
tended to be low in all species during the winter
months, with peak levels occurring in spring in
Sigmodon and summer in the Peromyscus spe-
cies. Population levels and incidence of parasit-
ism in two scrub populations exhibited an inverse
relationship.

120

Zoologica: New York Zoological Society

[53: 4

A summary of North American host and
locality records of C. hepatica is provided.

Acknowledgments

This research was supported by grant E-3730
from the U.S. Public Health Service. Grateful
acknowledgment is made to the following indi-
viduals for their valuable help and advice during
the course of the study: Dale E. Birkenholz,
James V. Griffo, Jr., William Wenz, William O.
Wirtz, II, and Dolores Smoleny.

Literature Cited

Ameel, D.

1942. Two larval cestodes from the muskrat.
Trans. Anrer. Microsc. Soc., 69:267-271.

Brown, J. H, and G. D. Roy

1943. The Richardson ground squirrel, Citellus
richardsonii Sabine, in southern Alberta,
its importance and control. Scien. Agric.
(Rev. Agron. Can.), 24:176-197.

Caballero, Y. C. E.. and R. G. Grocott

1 952. Nota sobre la presencia de Capillaria hepa-
tica en un mono arana (Ateles geofjroyi
vellerosns) de Mexico. Anal. Inst. Biol.
Mexico, 23:211-215.

Calero, M. C., P. Ortiz O., and L. de Souza

1950. Helminths in rats from Panama City and
suburbs. J. Parasit., 36:426.

Calhoun, J. B.

1962. The ecology and sociology of the Norway
rat. Publ. Health Serv. Publ. No. 1008,
U.S. Dept. Health, Education, and Wel-
fare, 288 pp.

Carr, A. F., Jr.

1940. A contribution to the herpetology of
Florida. Univ. Fla. Publ. Biol. Sci. Ser.,
3:1-118.

Chitwood, B. G.

1934. Capillaria hepatica from the liver of Cas-
tor canadensis canadensis. Proc. Helm.
Soc. Wash., 1:10.

Cooper, R. W.. C. S. Schopmeyer, and W. H. Davis
1.95 9. Sand pine regeneration on the Ocala Na-
tional Forest. Production Res. Rep. No.
30, U.S. Dept. Agr. For. Serv., 37 pp.

Cram, E. B.

1928. A note on parasites of rats (Rattns nor-
regicus and Rattns norvegicus alhus). J.
Parasit., 15:72.

Dalquest, W. W.

1948. Mammals of Washington. Univ. Kans.
Publ. Mus. Nat. Hist., 2: 1-444.

Davis, D. E.

1951. The relation between the level of popu-

lation and prevalence of Leptospira, Sal-
monella, and Capillaria in Norway rats.
Ecology, 32:465-468.

Dikmans, G.

1932. The pocket gopher, Thomomys fossor, a
new host of the nematode, Hepaticola
hepatica. J. Parasit., 19:83.

Doran, D. J.

1955. A catalogue of the Protozoa and Hel-
minths of North American rodents. III.
Nematoda. Amer. Midi. Nat., 53: 162-175.

Ewing, G. M., and I. L. Tilden

1956. Capillaria hepatica, report of fourth case
of true human infestation. J. Pediatrics,
48:341-348.

Firlotte, W. R.

1948. A survey of the parasites of the brown
Norway rat. Can. J. Comp. Med., 1:187-
191.

Fisher, R. L.

1963. Capillaria hepatica from the rock vole in
New York. J. Parasit., 49:450.

Foster, A. O., and C. M. Johnson

1939. An explanation for the occurrence of
Capillaria hepatica ova in human feces
suggested by the finding of three new hosts
used in food. Trans. Royal Soc. Trop. Med.
& Hyg., 32:639-644.

Freeman, R. S.

1958. On the epizootiology of Capillaria hepatica
(Bancroft, 1893) in Algonquin Park, On-
tario. J. Parasit., 44 (suppl.): 33.

Freeman, R. S., and K. A. Wright

1960. Factors concerned with the epizootiology
of Capillaria hepatica (Bancroft, 1893)
(Nematoda) in a population of Peromys-
cus maniculatus in Algonquin Park, Can-
ada. J. Parasit., 46:373-382.

Habermann, R. T„ F. P. Williams, and W. T. S.

Thorp

1954. Common infections and disease conditions
observed in wild Norway rats kept under
simulated natural conditions. Amer. J. Vet.
Res. 15:152-156.

Hall, M. C.

1916. Nematode parasites of mammals of the
orders Rodentia, Lagomorpha and Hyra-
coidea. Proc. U.S. Nat. Mus., 50: 1-258.

Harkema, R.

1936. The parasites of some North Carolina
rodents. Ecol. Monog., 6:153-232.

Herman, C. M.

1939. A parasitological survey of wild rats in the
New York Zoological Park. Zoologica,
24:305-308.

1968]

Layne: Host and Ecological Relationships of the Parasitic Helminth
Capillaria hepatica in Florida Mammals

121

Laessle, A. M.

1942. The plant communities of the Welaka area.
Univ. Fla. Publ. Biol. Sci. Ser., 4:1-143.

1958. The origin and successional relationships
of sandhill vegetation and sand-pine scrub.
Ecol. Monog., 28:361-387.

Law, R. G., and A. H. Kennedy

1932. Parasites of fur-bearing animals. Bull. No.
4, Ontario Dept. Game and Fisheries,
30 pp.

Layne, J. N.

1963. A study of the parasites' of the Florida
mouse, Peromyscus floridanus, in relation
to host and environmental factors. Tulane
Stud, in Zool., 11:1-27.

1966. Postnatal development and growth of
Peromyscus floridanus. Growth, 30:23-45.

Layne, J. N., and J. V. Griffo, Jr.

1961. Incidence of Capillaria hepatica in popula-
tions of the Florida deer mouse, Peromys-
cus floridanus. J. Parasit., 47:31-37.

Lubinsky, G.

1956. On the probable presence of parasitic liver
cirrhosis in Canada. Can. J. Comp. Med.,
20:457-465.

1957. List of helminths from Alberta rodents.
Can. J. Zool., 35:623-627.

Luttermoser, G. W.

1936. A helminthological survey of Baltimore
house rats (Rattus norvegicus). Amer. J.
Hyg., 24:350-360.

1938. An experimental study of Capillaria hepa-
tica in the rat and the mouse. Amer. J.
Hyg., 27:321-340.

MacArthur, W. P.

1924. A case of infestation of human liver with
Hepaticola hepatica (Bancroft, 1893) Hall,
1916; with sections from the liver. Proc.
Roy. Soc. Med. (Sec. Trop. Dis. and Par-
asit.), 17:83-84.

McQuown, A. L.

1950. Capillaria hepatica. Amer. J. Clinical
Path., 24:448-452.

Meyer, M. C., and J. R. Reilly

1950. Parasites of muskrats in Maine. Amer.
Midi. Nat., 44:467-477.

Momma, K.

1930. Notes on modes of rat infection with He-
paticola hepatica. Ann. Trop. Med. Para-
sit., 24:109-113.

Otto, G. F., M. Berthrong, R. E. Appleby, J. C.

Rawlins, and O. Wilbur

1954. Eosinophilia and hepatomegaly due to
Capillaria hepatica infection. Bull. John
Hopkins Hosp., 94:319-336.

Penn, G. H.

1942. Parasitological survey of Louisiana musk-
rats. J. Parasit., 28:348-349.

PoURNELLE, G. H.

1952. Reproduction and early post-natal devel-
opment of the cotton mouse, Peromyscus
gossypinus gossypinus. J. Mamm., 33:1-20.

Price, E. W.

1931. Hepaticola hepatica in liver of Ondatra
zibethica. J. Parasit., 18:51.

Price, E. W., and B. G. Chitwood

1931. Incidence of internal parasites in wild rats
in Washington, D. C. J. Parasit., 18:55.

Rausch, R.

1961. Notes on the occurrence of Capillaria
hepatica (Bancroft, 1893). Proc. Hel-
minth. Soc. Wash., 28:17-18.

Read, C. P.

1949. Studies on North American helminths of
the genus Capillaria Zeder, 1800 (Nema-
toda): I. Capillarids from mammals. J.
Parasit., 35:223-227.

Rogers, J. S.

1933. The ecological distribution of the crane-
flies of northern Florida. Ecol. Monogr.,
3:1-74.

Shorb, D. A.

1931. Experimental infestation of white rats
with Hepaticola hepatica. J. Parasit.,
17:151-154.

Skrjabin, K. I., N. P. Shikhobalova, and I. V.

Orlov

1957. Trichocephalidae and Capillariidae of ani-
mals and man and the diseases causes by
them. In: Fundamental Nematology. Vol.
6. Ed. K. I. Skrjabin. Acad. Sci. SSSR.
Moscow, pp. 5-585 (In Russian).

Swales, W. E.

1933. A review of Canadian helminthology. I.
Can. Res., 8:468-477.

Tromba, F. G.

1959. Swine as potential reservoir hosts of Hepa-
ticola hepatica. J. Parasit., 45:134.

Wantland, W. W., H. M. Kemple, G. R. Beers,

and K. E. Dye

1956. Cysticercus fasciolaris and Capillaria he-
patica in Rattus norvegicus. Trans. 111.
Acad. Sci., 49:177-181.

Ward, J. W.

1934. A study of some parasites of central Okla-
homa. The Biologist, 15:83-84.

Weidman, F.

1917. Dr. Weidman’s report. In: Fox, H., 1916,
Ann. Rept. Zool. Soc. Phila., pp. 33-40.

122

Zoologica: New York Zoological Society

[53: 4

Weidman, F.

1925. Hepaticoliasis, a frequent and sometimes
fatal verminous infestation of the livers of
rats and other rodents. J. Parasit., 12:19-
25.

Wright, K. A.

1961. Observations on the life cycle of Capillaria
hepatica (Bancroft, 1893) with a descrip-
tion of the adult. Can. J. Zool., 38 : 167-182.

Wright, W. H.

1930. Hepaticola hepatica in dogs. J. Parasit.,
17:54-55.

Yokogawa, S.

1920. A new nematode from the rat. J. Parasit.,
7:29-33.

EXPLANATION OF THE PLATE
Plate I

Fig. 1. Normal liver of P. floridanus (left) and one
showing lesions of Capillaria hepatica
(right). Note lobulation of infected liver.

Fig. 2. Ovum of Capillaria hepatica from liver of
P. floridanus. X625.

Fig. 3. Ova of C. hepatica in liver of P. floridanus.
X300.

LAYNE

PLATE 1

FIG. 1

FIG. 2

FIG. 3

HOST AND ECOLOGICAL RELATIONSHIPS OF
THE HELMINTH CAPILLARIA HEPATICA IN FLORIDA MAMMALS

[1968]

Zoologica: Index to Volume 53

123

Numbers in parentheses are the
series numbers of papers contain-
ing the tables, figures, or plates
listed immediately following; num-
bers in bold face indicate text-
figures; names in bold face indi-
cate new genera, species, or sub-
species.

B

brown bear, male

breeding biology of, 85-105, (7)
1-2, (7) Pis. MX
bushpig, African (Rhodesia)
observations on, 75-84, (6) Tables
I-IV, (6) 1-8, (6) PI. I

C

Capillaria hepatica

host and ecological relation-
ships in Florida mammals, 107-
122, (8) Tables I-VI, (8) 1-4, (8)
PI. I

ecologicaldistribution, 111-113
host and geographic distribu-
tion, 111

methods and materials, 107-
111

relationship between inci-
dence and host density, 115-
116

seasonal variation in infec-
tions, 115

sex and age differences in
infections, 114-115
Crustacea: Anomura. 57-74, (5)
Table I, (5) 1-2
considerations:
ecological, 59
geographical, 59-60
systematic, 60

restriction of synonymies, 60
systematic discussion, 60-72
Megalobrachium festai
(Nobili), 72
M. garthi Haig, 72
M. poeyi (Guerin), 71-72
M. tuberculipes (Lockington),
72

Minyocerus kirki Glassell, 70
Neopisosoma mexicanum
(Streets). 67-68
Pachychelces biocellatus
(Lockington), 68-69
P. calculosus Haig. 68
P. chacei Haig, 68
P. crassus (A. Milne Edwards),
68

P. panamensis Faxon, 69
P. spinidactylus Haig, 69
P. trichotus Haig, 69-70
P. vicarius Nobili, 69
Petrolisthes agassizii Faxon,
60

P. armatus (Gibbes), 62
P. edwardsii (Saussure), 60-
61

P. galapagensis Haig, 65-66

INDEX

P. g lasselli Haig, 61
P. gracilis Stimpson, 65
P. haigae Chace, 61-62
P. hians Nobili, 67
P. holotrichus Nobili, 66
P. lewisi austrinus Haig, 67
P. lewisi lewisi (Glassell), 66-
67

P. nobilii Haig, 62
P. ortmanii Nobili, 66
P. platymerus Haig, 66
P. polymitus Glassell, 61
P. robsonae Glassell, 65
P. tonsorius Haig, 66
P. tridentatus Stimpson, 65
P. zacae, new species, 63-65,
(5) Table I, (5) 2
Pisidia magdalenensis
(Glassell), 71

Polyonyx confinis Haig, 72
Porcellana cancrisocialis
Glassell, 70

P. paguriconviva Glassell,
70-71

L

Leptonychotes weddelli (Lesson)
thermoregulation of pup and
adult, in Antarctica, 33-46, (3)
Table I, (3) 1-9, (3) Pis. I ll
breathing and heart rates, 37
effect of insolation on fur and
skin, 40

internal temperatures, 35
material and methods, 34-35
metabolic rates, 38
role of behavior, 40
temperature of dry fur and
skin, 35-37

temperature of the skin when
wet 35

thermal gradients in superfi-
cial tissues, 38-40

O

Odobenus rosmarus (Linnaeus)
influence of climate on distribu-
tion of; evidence from physio-
logical characteristics, 19-32, (2)
Table I, (2) 1-9

breathing and heart rates, 24
correlation of thermoregula-
tory behavior and physiolo-
gy, 25-27

internal temperatures, 20
limits of thermoneutrality and
thermal tolerance, 28-30
materials and methods, 20
temperature gradients in skin
and blubber, 24-25
temperature of skin
dry, 23-24
wet, 21-23

influence of climate on distribu-
tion of: evidence from thermo-
regulatory behavior, 1-14, (1)
Tables I-IV, (1) 1-2, (1) Pis. I-IV
activity in relation to weather

captive walruses, 6-8
wild walruses, 8
activity rhythm, 5-6
huddling, 4

materials and methods, 2-3
posture and fanning, 4-5
selection of substrates, 5

P

Parasambonia bridgesi n. gen., n.
sp„ 49-56, (4) 1-6

differential diagnosis, 50
discovery and naming of, 49
new genera, 50
specific description, 50-54, (4)
1-6

pentastomes, 49-56, (4) 1-6
porcellanid crabs, 57-74, (5) Table
I, (5) 1-2

Potamochoerus porcus Linn., 75-
84, (6) Tables I-V, (6) 1-8, (6) Pis.
1-11

behavior, 82
control of. 82-83
dentition. 75-76

determining sex from skull, 80
food habits, 82
growth, 76-80

lens weight as index to age, 80
methods of study, 75
reproduction, 80-82
trypanosomiasis, 82
weights and measurements, 76
Pseudechis porphyriacus, 49

S

seal, Weddell

thermoregulation of pup and
adult, 33-46, (3) Table I, (3) 1-9,
(3) Pis. I II
snake, Australian

pentastomes from lungs of, 49-
56

U

Ursus arctos, 85-105, (7) Tables
I-IV, (7) 1-2, (7) Pis. MX

estimates of age and reproduc-
tive state, 92-94

gross features of male repro-
ductive tract, 94-95
histology of male reproductive
tract

infantile bear, 95-96
prepuberal, 96

methods and procedures, 85-92
sexually mature bear
fully active, 97
postseasonal, 97-98
redevelopment phase, 96-98

W

Waddycephalus teretiusculus, 49
walrus, influence of climate on
distribution of:

physiological characteristics,
19-32, (2) Table I, (2) 1-9
thermoregulatory behavior,
1-14, (1) Tables I-IV, (1) 1-2

NEW YORK ZOOLOGICAL SOCIETY

The Zoological Park, Bronx, N. Y. 10460

Fairfield Osborn
Chairman of the Board of Trustees

John Pierrepont
Treasurer

OFFICERS

Laurance S. Rockefeller
President

Henry Clay Frick, II
Vice-President

Eben W. Pyne
Assistant Treasurer

Robert G. Goelet
Executive Vice-President
Chairman of the Executive Committee

Howard Phipps, Jr.
Secretary

Edward R. Ricciuti
Editor & Curator,
Publications & Public Relations

Joan Van Haasteren
Editorial Assistant

EDITORIAL COMMITTEE
Robert G. Goelet
Chairman

William G. Conway F. Wayne King

Lee S. Crandall Hugh B. House Peter R. Marler

Donald R. Griffin Ross F. Nigrelli

William G. Conway
General Director

ZOOLOGICAL PARK

William G. Conway . . . Director & Curator, Charles P. Gandal Veterinarian

Ornithology Lee § Crandall . . . General Curator Emeritus

Hugh B. House .... Curator, Mammalogy <£ Zoological Park Consultant

Grace Davall . . Assistant Curator, Mammah william Bridges . Curator of Publications Emeritus

Walter Auffenberg . . . Research Associate in John M. Budinger . . . Consultant, Pathology

Herpetology Ben Sheffy Consultant, Nutrition

Joseph Bell . . Associate Curator, Ornithology James G. Doherty Mammalogist

F. Wayne King .... Curator, Herpetology Donald F. Bruning Ornithologist

Joseph A. Davis, Jr Scientific Assistant to the Director

AQUARIUM

Ross F. Nigrelli Director Robert A. Morris Curator

Christopher W. Coates . . Director Emeritus U. Erich Friese Assistant Curator

Nixon Griffis .... Administrative Assistant Louis Mowbray . Research Associate in Field Biology

OSBORN LABORATORIES OF MARINE SCIENCES

Ross F. Nigrelli . . . Director and Pathologist Harry A. Charipper . Research Associate in

Martin F. Stempien, Jr. . . . Assistant to the Histology

Director & Bio-Organic Chemist Kenneth Gold Marine Ecologist

George D. Ruggieri, S.J. . . . Coordinator of Eva K. Hawkins Algologist

Research & Experimental Embryologist Myron Jacobs Neuroanatomist

....... . , . _ , . . . Klaus Kallman Fish Geneticist

William Antopol . . . Research Associate in Vincent R. Liguori Microbiologist

Comparative Pathology John J. A. McLaughlin . . Research Associate in
C. M. Breder, Jr. ... Research Associate in Planktonology

Ichthyology Martin P. Schreibman . . Research Associate in

Jack T. Cecil Virologist Fish Endocrinology

Jay Hyman Research Associate in Comparative Pathology

INSTITUTE FOR RESEACH IN ANIMAL BEHAVIOR

[Jointly operated by the Society and The Rockefeller University, and including the Society’s William Beebe Tropical

Research Station, Trinidad, West Indies]

Donald R. Griffin . . .

Peter R. Marler .
Jocelyn Crane .
Roger S. Payne .
Richard L. Penney

Director & Senior
Research Zoologist
Senior Research Zoologist
Senior Research Ethologist
. Research Zoologist
. . Research Zoologist

Fernando Nottebohm . . . Research Zoologist

George Schaller Research Zoologist

Thomas T. Struhsaker . . Research Zoologist

C. Alan Lill Resident Director

O. Marcus Buchanan . . . Resident Director ,‘

William Beebe Tropical Research Station

. . vyi/

Q X^OSVAV^ W /X'5*’ — - '^2^' o X^/OSV^X M O X^OSV^X

2 «J 2 -J 2 _t 2

s saiavaan libraries Smithsonian institution NoiiniusNi nvinoshiiins saiavaai

2 Z £ 2 f“ ... 2

00 rn XXTcJX rn n^dcX w ^

fE to — to _ to _

N INSTITUTION NOIlOillSNI NVINOSHilWS S3IHVHan LIBRARIES SMITHSONIAN INSTITUTE

2 . v to 2 _ to 2 « . CO 2 v-

< ^ s .< 2 ..< <&\ ... =2 < ^v.

X
to

! ^211

> 2 '<2^"" > 2 ’*' >

CO “ 2 tO 2 00*2 CO

s saiavaan libraries Smithsonian institution NoiiniusNi nvinoshiiiais saiavaai

CO =p» to ^ to 2 . to

UJ X^^^TX w X^SP^/X t*J . tn Xt'^S'/X W X®T\ ^ UJ

Q ~ ''4?" o X^Y UL^/ __ Q

NJ|NSTITUTI0N:2N0linillSNl“JNVIN0SHlUAIS”S3 IHVHan^LlBRAR lES^SMITHSONIAN^INSTITUTIO

“ “ v Z r* 2 r-

i>

s saiavaan libraries Smithsonian institution NoiiniusNi nvinoshiiws saiavaai

to 2 •* 2 ■* CO 2 to

2 A. < X S xS^SX < V 2

x

to
O
2

>■' s S ,)^v >' 2

^institution ° NoiiniiiSNi__NviN0SHimswS3 i a va a n\i b rar i esu>smithsonian“institutioi

co _ r? X w _ S co

O xQftosvPX- — O Xjyos^ — O

2 2 -» 2 _S 2

s sa i a va an libraries Smithsonian institution NoiiniusNi nvinoshiiws S3iavaai

2 f~ z r; 2 r“ .v„ . 2

m x^x. d.c2 ^ W rn rn

N “INSTITUTION °° NOlinilJLSNrNVINOSHimS S3 i ava a it“li B R AR I ES^SMITHSONIAN INSTITUTI0

Z v to 2 to 2 *, </> 2 ,v-

g | ''f ~ | | i

to *• 2 tO 2 to *2 to

s S3 lava an libraries Smithsonian institution Noiimtism nvinoshiiws saiavaai

CO X CO 2 to — to

lu tn Xt*s°vX 5x1 . co XS<Bl£yX 1x1 /i@N Cx ui

o Xffl. Qp2 ~ '^s?- O x-'*1 _ XiOms^x o

n^institution ^NoiiniiiSNi^NViNOSHiiiAis saiavaan^LiBRARi ES2 SMITHSONIAN INSTITUTE

— “ > 2 r* 2

O X5st.!t<oX m ^ o o

V

° /?•?' — O Xl'VQS'^ _ O

AR I ES^SMITHSONIAN^INSTITUTION NOlinillSNl NVINOSHlIWS2 S3 I H VM 8 ! l^L ! B RAR I ES SMITH

p B » <f|> b " /gSjx |

’ l | ^

Z -V/S" m m v> p ^■'‘^'■•y V*

co \ ? 00

UllSNI NVINOSHimS S3 I BVB 8 11 LIBRARIES SMITHSONIAN INSTITUTION NOimillSN! NVINO
< s • < xn( s ^ ^ 5 « S z ^

t* — ^ . 'J — vm\\i iz! v • <C x<v^

_ 2 s " >

AR 1 ES,nSivilTHSOtsJIAN_ INSTITUTION NOlinillSNl NVIN0SH1I WS^S 3 lUVaail^LIBRARI es^smith:

— y* 5 co 5 CO

UJ w UJ X^H'X ^ XmSO^Xv Ui

-H
C

_ - O xv Z N^oJ^X 5 ^

UllSNl^NVINOSHllWS^SB I y VH a 11^ L I B R AR 1 ES^ SMITHSONIAN^ INSTITUTION s~NOIinilISNI~JNVINO!

. ° 5 | ra xgpgx o ^2^

> pc £ x§f' Jx; > - i 2

;o/ CO V5v" ~ 1 *“

03
S3

5 >2

m >SSf z m xgassy S ■ ' '■ m x^'5gs?' <" XCgSSZ p X

w/ M ^ ■■ — xK ^

ARIES SMITHSONIAN INSTITUTION NOIifUIlSNI “nVINOSHIIWS S3 I y VH a H ~L! B ^ AR I ES SMITH:

p, 2 > co 2; co ~ ■ rn

— < a*. 2 jadfe < 2

A | § %% 1 4$^ §

Z g >W ! t .

2 2 xs > 2 Xi^o^X > 2 • s^p* >

1I1SNI_NVIN0SH1IINS S3 I H V8 8 11 LIB RAR I ES^SMITHSONIAN^INSTITUTION ^NOlinillSNl* NVINO

^ r OT ~ *« = co i •

Ei

V

H f.
c L

pt

o TW “ X^TTTxX p. car ' ' — ' \^m^,

^ /X' — • ^ O N^A/osva^X _ Q X^vqsva^

^2. a j|

ARIES SMITHSONIAN INSTITUTION NOIifUIlSNI NVIN0SH1IIAIS S3iyvaaS1 LIBRARIES SMITf

— p; _ Z r* 2 r- 2

O 8-^X q xTTv^vaTv x^vTsoaT^x o

m - x^mis^x rn ^ y> N^iisgX m

__ jy) -— -

1I1SNI NVIN0SH1IINS S3 I yvy an LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNl" NVINO'

< XisoX 2 < X'C ^ 5? 2 $ ,-

«*— X^UfaJr^ — ‘ 'z S = .< X®

I 1 |F^I I IlfcL I 1

w i % %um t l %m t %

— > 2 "'*' > XJOcaSJiX S

<0 2 CO * 2 ^ * 2 to

* ^S SM1THS0NIAN_ INSTITUTION NOlinillSNl NVIN0SH1IWS S3IBVyan LI B RAR 1 ES SMITH!

, , 2 ^ ^ ~ v CO ~ CO

- .z ^ w w

0= ~ DC '■ ’ '

<

DC

o Ni^CDC^ “ X^TI^X 5 ^ 5

LIISNUNVINOSHIIINS S3IBVy ail^LIBRAR I ES*^ SMITHSON IAN*** INSTITUTION ~ NOlinillSNl ~*N VI NO

1 v r~ 2! r* 2; r— x,

rr, Xv ° 5 - JtM'a. m xfuiox °

A _ I- }\ » S$

'i/ —

%&ILWrj?/ », jf ( "

I'vosv'^/ ^

r> * f <"»