THE
BIOLOGICAL BULLETIN
PUBLISHED BY
THE MARINE BIOLOGICAL LABORATORY
Editorial Board
JOHN M. ANDERSON, Cornell University
JOHN B. BUCK, National Institutes of Health
PHILIP B. DUNHAM, Syracuse University
W. D. RUSSELL HUNTER, Syracuse University
SHINYA INOUE, University of Pennsylvania
JOHN H. LOCHHEAD, University of Vermont
LEONARD NELSON, Toledo State College of
Medicine
MELVIN SPIEGEL, Dartmouth College
WM. RANDOLPH TAYLOR, University of
Michigan
ANNA R. WHITING, Oak Ridge National
Laboratory
CARROLL M. WILLIAMS, Harvard University
J. LOGAN IRVIN, University of North Carolina
DONALD P. COSTELLO, University of North Carolina
Managing Editor
VOLUME 132
FEBRUARY TO JUNE 1967
Printed and Issued by
LANCASTER PRESS, Inc.
PRINCE & LEMON STS.
LANCASTER, PA.
11
THE BIOLOGICAL BULLETIN is issued six times a year at the
Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Penn-
sylvania.
Subscriptions and similar matter should be addressed to The
Biological Bulletin, Marine Biological Laboratory, Woods Hole,
Massachusetts. Agent for Great Britain: Wheldon and Wesley,
Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London,
W. C. 2. Single numbers, $3.75. Subscription per volume (three
issues), $9.00.
Communications relative to manuscripts should be sent to Dr.
Donald P. Costello, Marine Biological Laboratory, Woods Hole,
Massachusetts, between June 1 and September 1, and to Dr.
Donald P. Costello, P.O. Box 429, Chapel Hill, North Carolina
27514, during the remainder of the year.
Second-class postage paid at Lancaster, Pa.
LANCASTER PRESS, INC., LANCASTER, PA.
CONTENTS
No. 1. FEBRUARY, 1967
PAGE
JACKSON, CHARLOTTE, AND ROBERT E. BLACK
The subcellular distributions of some hydrolytic enzymes in unfertilized
eggs of the sea urchin, Arbacia punctulata ........................... 1
JOHANSEN, KjELL, AND ROBERT L. VADAS
Oxygen uptake and responses to respiratory stress in sea urchins ...... 16
JONES, JACK COLVARD
Spermatocysts in Aedes aegypti (Linnaeus) ........................ 23
LEWIS, JOHN B.
Nitrogenous excretion in the tropical sea urchin Diadema antillarum
Philippi ........................................................ 34
MUN, A. M., L. B. CRITTENDEN AND BARBARA JEAN CLARKE
Induction of immunological tolerance by intracoelomic grafts in the 4-day
chick embryo ................................................... 38
PALMER, JOHN D., AND FRANK E. ROUND
Persistent, vertical-migration rhythms in benthic microflora. VI. The
tidal and diurnal nature of the rhythm in the diatom Hantzschia virgata . 44
PATTON, WENDELL K.
Studies on the commensal crab Domecia acanthophora with particular ref-
erence to modification of the coral host and feeding habits ............. 56
PIKO, LA jos, ALBERT TYLER AND JEROME VINOGRAD
Amount, location, priming capacity, circularity and other properties of
cytoplasmic DNA in sea urchin eggs ............................... 68
POTSWALD, HERBERT E.
Observations on the genital segments of Spirorbis (Polychaeta) ........ 91
SHAPPIRIO, DAVID G., DANIEL M. EICHENBAUM AND BRUCE R. LOCKE
Cholinesterase in the brain of the Cecropia silkmoth during metamorphosis
and pupal diapause .............................................. 108
SPIGHT, TOM M.
The water economy of salamanders : Exchange of water with the soil .... 126
STUNKARD, HORACE W.
Studies on the trematode genus Paramonostomum Lube, 1909 (Digenea:
Notocotylidae) ......................................... , ....... 133
in
iv CONTENTS
No. 2. APRIL, 1967
DANFORTII, CHARLES G.
Northern Pacific Gigantione (Isopoda) 147
EMERSON, DAVID N.
Surface area respiration during the hatching of encysted embryos of the
hrine shrimp, Artemia salina 156
FERGUSON, JOHN CARRUTHERS
Utilization of dissolved exogenous nutrients by the starfishes, Asterias
forbesi and Henricia sanguinolenta 161
HARRINGTON, ROBERT W., JR.
Environmentally controlled induction of primary male gonochorists from
eggs of the self-fertilizing hermaphroditic fish, Rivulus marmoratus Poey 174
JOHN, KENNETH R., MARC SEGALL AND LAWRENCE ZAWATZKY
Retinomotor rhythms in the goldfish, Carassius auratus 200
JONES, JACK COLVARD
Changes in the hemocyte picture of Galleria mellonella (Linnaeus) 211
LAMBERT, CHARLES C, AND CHARLES L. BRANDT
The effect of light on the spawning of Ciona intestinalis 222
MILKMAN, ROGER
Genetic and developmental studies on Botryllus schlosseri 229
PERSON, PHILIP. AND MARTIN B. MATHEWS
Endoskeletal cartilage in a marine polychaete, Eudistylia polymorpha . . . 244
SCHELTEMA, RUDOLF S.
The relationship of temperature to the larval development of Nassarius
obsoletus (Gastropoda) 253
STUNKARD, HORACE W.
The morphology, life-history, and systematic relations of the digenetic
trematode, Uniserialis breviserialis sp. nov., (Notocotylidae), a parasite
of the bursa Fabricius of birds 266
WATSON, J. A. L.
The growth and activity of the corpora allata in the larval firebrat,
Thermobia domestica (Packard) (Thysanura, Lepismatidae) 277
WHITTINGHAM, D. G.
Light-induction of shedding of gametes in Ciona intestinalis and Molgula
manhattensis 292
No. 3. JUNE, 1967
BARBIERI, FRANCISCO D., JORGE S. RAISMAN AND CESAR ALBARRACIN
Amylase and glycogenolysis in amphibian development 299
BROWN, F. A., JR., AND Y. H. PARK
Association-formation between photic and subtle geophysical stimulus pat-
terns— a new biological concept 311
DAVIS, WAYNE H., AND OLA B. REITE
Responses of bats from temperate regions to changes in ambient tempera-
ture 320
CONTENTS v
DOYLE, WILLIAM L.
Vesiculated axons in haemal vessels of an holothurian, Cucumaria frondosa 329
FISH, JOHN D.
The digestive system of the holothurian, Cucumaria elongata. I. Struc-
ture of the gut and hemal system 337
FISH, JOHN D.
The digestive system of the holothurian, Cucumaria elongata. II. Dis-
tribution of the digestive enzymes 354
JONES, MEREDITH L.
On the morphology of the nephridia of Nereis limnicola Johnson 362
PAX, RALPH A., AND RICHARD C. SANBORN
Cardioregulation in Limulus. II. Gamma aminobutyric acid, antagonists
and inhibitor nerves 381
PAX, RALPH A., AND RICHARD C. SANBORN
Cardioregulation in Limulus. III. Inhibition by 5-hydroxytryptamine
and antagonism by bromlysergic acid diethylamide and picrotoxin 392
SPIELMAN, ANDREW, SR. M. G. LEAHY AND VALERIE SKAFF
Seminal loss in repeatedly mated female Aedes aegypti 404
STAPLES, SUZANNE O., AND JAMES H. GREGG
Carotenoid pigments in the cellular slime mold, Dictyostelium discoideum 413
ERRATUM
In the legend to Figure 6, in the paper by Piko, Tyler and
Vinograd in the February, 1967, issue of THE BIOLOGICAL
BULLETIN (volume 132, no. 1, page 81), the first sentence
should read as follows: "Melting profiles of DXAs (in 0.15
M NaCl-0.015 M sodium citrate, pH 7) from L. Rictus."
Vol. 132, No. 1 February, 1967
THE
BIOLOGICAL BULLETIN
PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY
THE SUBCELLULAR DISTRIBUTIONS OF SOME HYUROLYT1C
ENZYMES IN UNFERTILIZED EGGS OF THE SEA
URCHIN, ARBACIA PUNCTULATA
CHARLOTTE JACKSON 1 AND ROBERT E. BLACK
Department of Biology, College of William and Mary, Williamsburg, Virginia 23185
The identification of specific cellular granules containing hydrolytic enzymes was
first made by de Duve and co-workers (1955) by centrifugal fractionation of rat-
liver homogenates. The name lysosome was proposed at this time for these granules
because several distinct acid hydrolases appeared to be located within them and to
be released in a paralleled manner in preparations subjected to disruptive treatments
such as freezing and thawing. At least 12 hydrolytic enzymes showing an acid pH
optimum are presently believed to be associated with the lysosome. These include
ribonuclease, acid deoxyribonuclease, acid phosphatase, phosphoprotein phosphatase,
cathepsin, collagenase, a//i/w-glucosidase, frrfa-N-acetylglucosaminidase, bcta-ghi-
curonidase, a//i/;a-mannosidase, and aryl-sulfatase (rf. review by Novikoff, 1961).
Although most of the work on the distribution of hydrolytic enzymes has been
with adult mammalian tissue, investigations of acid hydrolases have also been ex-
tended to some invertebrates. In most of these, however, only brief surveys have
been made of characteristic enzymes, and no attempts to isolate any specific granules
have been reported (rf. review by de Reuck and Cameron, 1963).
Lysosomal granules have been implicated in a few developmental processes in-
volving regression and resorption of embryonic cells, especially in Mullerian duct
rudiments of male chick embryos (Scheib-Pfleger and Wattiaux, 1962) and in the
tails of amphibian tadpoles undergoing metamorphosis (Weber, 1963). In the rat
egg Dalcq (1963) has observed granules in which high acid phosphatase was dem-
onstrated cytochemically and which stained metachromatically. The metachromatic
granules observed in invertebrate eggs appear to be of two types, designated as
alpha and beta granules by Pasteels and Mulnard (1957), who concluded that the
larger beta granules were not stained directly but received dye from the smaller
alpha granules. These investigators found by centrifugation that the alplia granules
are concentrated near the centrifugal extremity of the egg or blastomeres, while the
1 Submitted in partial fulfillment of the requirements for the Master of Arts degree. This
work was supported by NSF grant GB-711).
1
Copyright © 1967, by the Marine Biological Laboratory
Library of Congress Card No. A38-518
2 CHARLOTTE JACKSON AND ROBERT E. BLACK
beta granules are sedimented in the hyaline part of the cytoplasm. Dalcq (1963)
has recently concluded from comparative electron microscopic and cytochemical in-
vestigations of early developmental stages that the yolk platelets are a source not
only of nutrient material but also of definite organelles with phosphatase activity.
Pasteels and de Harven (1963) confirmed this by a series of electron microscope
findings which demonstrated the transformation of yolk platelets into microvesicular
bodies similar to the metachromatic granules observed in the living eggs, and the
rupture of the microvesicular bodies to release minute phosphohydrolase granules.
The only resemblances these granules have to lysosomes, however, are their ap-
parent ability to rupture, and their possession of acid phosphatase.
Attempts to identify specific acid hydrolase granules in embryonic tissue have
usually been made only by means of electron microscopy and cytochemistry. Dalcq
(1963) proposed at a recent symposium that the most direct approach to deter-
mining the presence or absence of particles containing specific hydrolases would be
the application of density gradient centrifugation to homogenates of eggs at various
stages, to determine whether a layer of particles containing an array of lytic enzymes
could be isolated. This present investigation is concerned with the problem of
attempting to isolate a fraction containing an appreciable concentration of acid
hydrolases from homogenates of unfertilized eggs of Arbacia punctnlata. To deter-
mine the distribution within the subcellular fractions differential centrifugation has
been applied to sucrose homogenates of both eggs and adult gut tissue. Isolation of
specific sets of granules containing acid hydrolase activity has been attempted only
on the large visible-granule fraction of the egg homogenate, by centrifugation on
sucrose density gradients. The preliminary results obtained by the differential and
density gradient centrifugation indicate the probable existence in the egg of more
than one type of large granule possessing hydrolytic enzymes.
MATERIALS AND METHODS
Biolological procedures
The animals used in this study, Arbacia punctulata, were collected along the
Eastern Shore region of the Chesapeake Bay. They were stored in the laboratory
at 23° C. in aerated, polyethylene aquaria filled with artificial sea water at 33/£0
made from Utility Seven-Seas Marine Mix (Utility Chemical Company, Patterson,
New Jersey). In order to obtain adult tissue, the entire gut was removed by dis-
section, placed in cold, artificial sea water, blotted carefully, and weighed. Eggs
were obtained by electrically inducing shedding in the females, using alternating
current at 30 volts (Harvey, 1954). The eggs were washed three times by settling
in artificial sea water prepared from reagent grade salts and distilled water. To
remove the jelly coat, the eggs were treated with acid sea water at pH 4.6-4.8 and
allowed to settle (Allen, 1957). The eggs were washed an additional time in sea
water buffered at pH 8.0 with 0.02 M tris-(hydroxymethyl)aminomethane (here-
after referred to as "tris") and packed in a hand centrifuge to prepare them for
homogenization.
Adult gut and eggs were prepared for fractionation by the same methods. A
10% homogenate of each tissue was made, based on a weight-volume ratio for the
gut tissues and on a volume/volume ratio in the eggs. The tissues were homog-
ENZYMES IN ARBACIA EGGS
enized at 0° C. in a hand-operated Tenbroeck glass homogenizer in 0.98 M sucrose
containing 10~3 M ethylenediamine tetraacetic acid (EDTA) and buffered at pH 7.5
with 0.05 .il/ tris. This buffered sucrose was used for suspending all of the cell
fractions obtained by later centrifugation. A starting material which was essentially
nuclei-free was prepared, based on the procedure of Berthet and de Duve (1951)
with modifications to allow for the differences in sucrose density. The 10%
homogenate was centrifuged for ten minutes at 2000 rpm for the gut homogenate
and at 500 rpm for the egg homogenate, in rotor No. 253 in an International PR-2
centrifuge at 0° C. The egg homogenate was centrifuged at the much lower speed
to remove the debris and whole cells without removing the majority of heavy
granules. The supernatant fluid was then decanted and saved. The sedimented
fraction from each tissue was rehomogenized in an additional 3 to 5 ml. of buffered
0.98 M sucrose-EDTA and recentrifuged at the same speed as before. The com-
bined supernatant fluids were used for the subsequent isolation of granules. The
final precipitate containing clumped nuclei, cell fragments, and whole cells was
discarded.
The various subcellular fractions were obtained bv a modification of the tech-
j
nique used by Applemans, Wattiaux and de Duve (1955). The procedure is out-
lined in the flow sheet given below. All of the operations described were carried
out at or near 0° C. The preparations were either used immediately or frozen at
-18° C. for up to 48 hours. Tests of the effects of freezing on enzyme activities
were made on whole homogenates. Xone of the enzymes for which data are re-
ported were decreased in activity as a result of freezing.
Fraction I from eggs was centrifuged on a layered sucrose gradient to separate
particles of different densities (de Duve, Berthet and Beaufay, 1959). The
separated granule layers were removed by pipetting from the top or by puncturing
the bottom of the tube and allowing the sucrose to drip out slowly.
The staining properties of the granules obtained on the density gradient were
examined by dividing Fraction I into three equal portions. One tube served as a
control; the other two contained 10 to 15 drops of 0.1% dye in a total volume of
5 ml. After an initial centrifugation at 17,500 times gravity to remove the excess
stain, the stained preparations were placed over the same density gradients as above,
and centrifuged for one hour at 90,000 times gravity.
Chemical procedures
Determinations of protein, nucleic acid, acid phosphatase, esterase, lipase, arvl-
sulfatase, />rta-galactosidase, ribonuclease (RNAase), and proteolytic activity were
attempted according to the methods outlined below. All enzyme reactions were run
at 25° C. Tests for linearity were made on whole homogenates in preliminary ex-
periments.
Before determinations of protein and nucleic acid were made, the samples were
extracted according to the method of Schmidt and Thannhauser (1945) three times
with cold 10% W/V trichloroacetic acid (TCA), twice with boiling ethanol-ether
(3:1 V/V) and twice with hot 5%, TCA at 90° for 15 minutes. The hot TCA
extracts were combined and used for nucleic acid determination by the ultraviolet
absorption procedure of Schneider (1957). The protein was suspended in 1 N
sodium hydroxide and determined with Folin-Ciocalteau reagent (Fisher Chemical
4 (. HARI.OTTE JACKSON AND ROBERT E. BLACK
Company) by tbe method of Lowry ct al. (1(>51 ). Standard absor1)ance curves
were prepared, using solutions of reagent grade RNA and of crystalline bovine
serum albumin, both obtained from Nutritional Biochemicals Corporation.
Assays of acid phosphatase, esterase, lipase, aryl-sulfatase and beta-ga.lactosida.se.
were attempted by using as substrates />-nitrophenyl phosphate, acetate, stearate,
sulfate and galactoside, respectively. All substrates were obtained from Nutritional
Biochemicals Corporation, except for ^-nitrophenyl phosphate, and />-nitrophenyl-
/3-D galactoside, which were obtained from the Sigma Chemical Company. The
/>-nitrophenol liberated from each substrate was determined in alkaline solution
(except as noted below ) at 400 mp. on the Beckman DU spectrophotometer, or with
filter No. 42 on the Klett-Summerson colorimeter. The details of each procedure
are presented below.
Acid phosphatase was determined by the method of Burch ct al. (1952).
Esterase was assayed at pH 7.1 by the method of Huggins and Laprides (1947),
using a standard curve of /'-nitrophenol at the same pH. The method used for
aryl-sulfatase depends on the rather small change in absorbance at 400 m/* which
occurs when />-nitrophenyl sulfate and enzyme are incubated at pH 5.1. The re-
action mixture contained 2.0 mg. of substrate, 0.03 M sodium acetate. pH 5.0 and
0.05 to 0.15 ml. of enzyme in a total volume of 3.0 ml. Readings were taken for 15
minutes on the DU or DB spectrophotometer against a blank containing buffer and
enzyme. Spontaneous hydrolysis of the substrate was negligible at this pH. A
standard curve was prepared from /'-nitrophenol at pH 5.1. Ribonuclease was
measured by the method of Dubos and Thompson (1938).
Attempts to measure protease activity at pH 5 were made by the method of
Anson (1938), using the liberation of tyrosine from denatured hemoglobin. This
method produced extremely variable results with both whole homogenates and sub-
cellular fractions. Whole homogenates gave high blank values, presumably because
of a high content of free tyrosine. The assay of frf fa-galactosidase was attempted
by the method of Wallenfels (1962), using />-nitrophenyl-&eto-galactoside. No
activity could be detected in the whole homogenates, and assays were not performed
on subcellular fractions. The assays for lipase, using />-nitrophenyl stearate (Hug-
gins and Laprides, 1947), were also negative in whole homogenates of eggs and gut.
Attempts to measure succinic dehydrogenase by the ferricyanide reduction method
of Bonner (1955) were made only on fresh preparations of Fraction I and its sub-
fractions. This reaction was complicated by the large amount of pigment in this
fraction, and it was decided that the data are probably unreliable.
RESULTS
Distribution of ensvtncs and nucleic acid in subcellular fractions
The distributions of acid phosphatase, esterase, and nucleic acid in the various
fractions obtained by differential centrifugation of the gut are shown in Table I.
The highest total acid phosphatase and esterase activities are found in the soluble
fraction, while the large granules contain only 10 to 15% of the total activity. The
microsomal fraction contains the highest specific activity of esterase while in most
experiments the specific activity of acid phosphatase was found to be rather uni-
formly higher in all the granule fractions than in the whole homogenate. The
ENZYMES IN ARBACIA EGGS
Outline of Fractionation Procedure
Whole homogenate in 0.()8 .!/
sucrose, 10 -"".V KDTA, 0.05 .17
tris, PH 7.5.
I
Centrifuged at 500-2000 rpm
for 10 minutes to remove
nuclei, whole cells, and
debris.
Precipitate discarded.
Precipitate suspended in
buffered 0.98 .17 sucrose.
Fraction I.
In eggs only, Fraction I re-
centrifuged on gradient of
densities 1.1513, 1.663,
1.1868, 1.1972, 1.2092, and
1.3163 for one hour at 90,000
times gravity. Granule
Subtractions A, B, C, D, and E.
Pre ipitate suspended in buffered
0.98 .17 sucrose. Fraction II.
Precipitate suspended in
buffered 0.98 M sucrose.
Fraction III.
Supernatant fluids
combined and centrifuged
at 23,000 times gravity
for 15 minutes. Washed
once.
Combined supernatant
fluids centrifuged at
90,000 times gravity
for 28 minutes. Washed
once.
Combined supernatant
fluids centrifuged at
90,000 times gravity
for 150 minutes. Washed
once.
Combined supernatant
fluids. Ftactir.n IV.
highest per cent of total nucleic acid, as determined hy the ultraviolet method, is
present in the supernatant fraction, representing very likely mostly soluble ribo-
nucleic acid. The percentage in the microsomal fraction, although slightly higher
than in the larger granules, is still quite low, possibly indicating that the cells are
poor in ribosomal ribonucleic acid. These findings may be a consequence of the
starvation of the animals prior to the fractionation of the tissue.
The distributions of enzyme activities within the subcellular fractions of the
eggs are presented in Table II. In addition to the enzymes examined in the gut
CHARLOTTE JACKSON AND ROBERT E. BLACK
TABLE I
Distribution of enzyme activities, nucleic acid and protein
in subcellular fractions of adult gut
Specific activity = /*M substrate converted/niin./mg. protein = factor indicated.
Standard errors are given. Number of experiments is indicated after each title.
Whole
I
II
III
IV
Total %
Acid phosphatase (4)
Per cent
100
15.5±1.1
8.9±2.0
8.2 ± 2.6
40. 8 ± 4.3
73.4 ± 3.2
Sp. act. X 10s
Esterase (2)
Per cent
6.0±0.8
100
11.9±1.5
9.3±2.0
11.9±2.0
8.7±0.0
12. 3± 1.2
22.6±11.0
3.9± 1.2
97.6± 0.6
138.2±12.0
Sp. act. X 105
Nucleic acid (2)
Per cent
1.4±0.0
100
1.5±0.0
8.7±2.4
2.0±0.9
5.2±1.3
4.4± 0.4
10.2± 2.0
1.6± 0.0
70.0± 5.7
94.1 ±0.2
Sp. amt. X 102
Protein (4)
Per cent
2.2±0.0
100
2.1±0.0
8.2 ±0.4
1.4±0.1
4.9±1.7
2.6± 0.6
4.7± 2.2
1.8± 0.4
70.0±11.2
87.8±15.0
subcellular fractions, determinations were made of aryl-sulfatase and RNAase. The
highest total acid phosphatase and esterase activities are present in the soluble
fractions, but the large granules contain from 25 to 50% of the total acid phospha-
tase activity and from 15 to 40% of the esterase activity. This difference in per-
centage of activity present in the large granules of the eggs is probably due to the
presence of granules which are not present in the gut. The highest specific activity
of esterase is found in the microsomal fraction (III) ; in this respect the distribution
TABLE II
Distribution of enzyme activities, nucleic acid and protein in
subcellular fractions of the egg
Specific activity for RNAase is expressed as ^g RNA solubilized/min.//jg protein.
All other activities are as in Table I.
Whole
I
II
ill
IV
Total %
Acid phosphatase (4)
Per cent
100
36.1 ±5. 2
11.3±2.9
5.8±0.9
62. 3± 7.9
11 5. 5 ±7. 6
Sp. act. X 105
Esterase (4)
Per cent
3. 5 ±0.4
100
7.1±0.6
23.2±6.4
6.8±2.0
12.3±2.8
13. 2 ±3.9
5.9±0.7
2.5± 0.5
69.2 ± 3.7
110.6± 7.2
Sp. act. X 106
Aryl-sulfatase (3)
Per cent
8.3±1.1
100
11.2±3.3
26.4±3.5
18.2±3.6
2.1±0.5
42.3±7.6
0.9±0.4
9.4± 2.1
33.5 ± 7.2
62. 9± 9.0
Sp. act. X 104
RNAase (3)
Per cent
2.4±0.6
100
1.9±0.6
32.1 ±2. 3
0.5±0.1
17.7±0.5
0.5 ±0.3
15.5±1.4
1.4± 0
83.6± 7.7
148.9± 9.0
Sp. act. X 102
Nucleic acid (3)
Per cent
4.7±0.8
100
4.5±0.8
20.3 ±1.4
8.0±2.9
14.2±3.2
9.8±2.4
32.8±7.0
8.8± 1.3
65.6±12.0
132.9±34.3
Sp. amt. X 102
Protein (7)
Per cent
7.7±0.2
100
4.7±0.7
24.3 ±3.5
10.4±0.8
7.7±1.2
34.4±5.0
4.1±1.2
11. 2± 2.3
60.3± 6.7
96.4±15.5
ENZYMES IN ARBACIA EGGS
B
E
FIGURE 1. Subfractions obtained from Fraction I by density gradient centrifugation. The
subfractions were obtained by centrifuging granules on the following densities of sucrose for one
hour at 90,000 times gravity: 1.1513, 1.1663, 1.1868, 1.1972, 1.2092, and 1.3163.
is similar to that in the adult intestine. The acid phosphatase in the egg also
resembles that of the gut in being rather consistently more concentrated in all
granule fractions than in the whole homogenate. The distribution of aryl-sulfatase
ranges from 20 to 30% in Fraction I and from 20 to 40% in Fraction IV. The
highest specific activity of RNAase, as well as the greatest nucleic acid : protein
ratios, were found in the microsomal fraction, Fraction III. The highest per-
centages of both RNAase activity and total nucleic acid are found in the soluble
fraction ; however, the recoveries of both these substances are quite high, when the
combined amounts in the separated fractions are compared to those in the whole
homogenates. It is interesting to note that considerable percentages of the recovered
nucleic acid and RNAase were found to be present in the visible granule fraction
(I), and in the intermediate granules (II), indicating the probable association of
8
CHARLOTTE JACKSON AND ROBERT E. BLACK
both substances with non-microsomnl particles. This association is examined in
more detail below.
/Density (jradient ccntrijmjatlon of Fraction I
In order to investigate the possible heterogeneity of the visible granules with
respect to their contents of hydrolytic enzymes, Fraction I was further centrifuged
in tubes containing several layers of sucrose solutions having different densities.
The separation of granules into layers of different densities as a result of this
centrifugation is depicted in Figure 1. In most experiments four separable layers
were obtained ; in one case a fifth, denser layer was also found. Because the centri-
fugation was performed for only one hour, it seems unlikely that complete separ-
ation of granules of different densities was achieved. This incompleteness of
separation, as well as a certain amount of mixing which occurred upon removal of
the different fractions, undoubtedly contributed to the variations in distribution and
activities reported below.
Assays of the hydrolytic enzymes in question, and of nucleic acid were per-
formed on the subtractions obtained by the gradient centrifugation. Tests for DNA
TABLE III
Distribution of enzyme activities, nucleic acid and protein in
subfr actions derived from fraction I
Specific activities are as in Tables I and II. The letters A through F represent subtractions of
different densities obtained by centrifuging Fraction I on a sucrose layer gradient. Sub-
fraction A has the lowest density.
A
B
C
D
E or F
Total %
Acid phosphatase (3)
Per cent of Fr. I.
28.0± 0.0
52.3±1.3
12.7±0.4
6.1 ±0.6
99. 1± 1.3
Sp. act. X 104
3.4 ± 0.1
1.1 ±0.0
0.5±0.0
0.5 ±0.0
Acid phosphatase* (2)
Per cent of Fr. I
51.0±10.0
9.2±2.1
3.2±2.2
3. 6 ±0.3
3.2±0.8 (Fj*
70.2± 8.0
Sp. act. X 104
0.5± 0.0
0.7±0.1
0.1 ±0.0
0.3 ±0.1
2.2±0.2 (F)*
Esterase (3)
Per cent of Fr. I.
8.6± 0.1
3 1.4 ±0.3
17.0±0.0
1.2±0.6
58.2± 1.0
Sp. act. X 105
3.3± 0.1
l.QdbO.O
2.1±0.1
0.2±0.1
Aryl-sulfatase (4)
Per cent of Fr. I
24.6± 5.2
39.8±5.9
17.8±3.3
15.9±4.6
25.1 (E)**
104.4±15.2
Sp. act. X 104
3.0 ± 0.7
1.7±0.2
2.2±0.4
3.6±0.7
7.2 (E)**
RNAase (3)
Per cent of Fr. I
38.8± 3.5
39.8±6.5
35.0±4.0
33. 7 ±2. 9
41.8 (E)**
161.3±10.0
Sp. act. X 102
8.7± 0.5
3. 2 ±0.8
16.2±4.6
12.0±1.4
26.0 (E)
Nucleic Acid (4)
Per cent of Fr. I
13.2+ 1.5
23.7±4.5
18.0±4.2
14.9±3.6
15.9 (E)**
73.3±9.1
Sp. amt. X 10-
9.1 ± 1.5
3.8 ±0.5
8.8±1.8
16.5±4.8
30.0 (E)**
Protein (8)
Percent of W.H.
5.3 ± 1.4
13.6±1.8
4.7±0.6
2.9±0.5
1.6 (E)**
27.9±2.2
Protein (2)*
Per cent of W.H.
13. 8± 2.4
2.0±0.1
5.9±0.5
1.5±0.3
0.2 (F)*
23.4± 2.2
* Eggs were washed in calcium-free water and Fraction I was centrifuged on the gradient for
90 minutes. Echinochrome granules were concentrated in Layer F.
** Layer E \vas obtained in only one experiment.
ENZYMES IN ARBACIA EGGS
A B C D
ACID PHOSPHATASE
A B C D
E ST E R A S E
— 4
LU
> 3
ABODE
SULFATASE
ABODE
RNASE
ABODE
NUCLEIC ACID
FIGURE 2. Average relative specific activities of enzymes and amounts of nucleic acid in
the subfraction granules isolated from Fraction I. The specific activity of the whole homogenate
is set at 1.0 for the calculations. The lines represent the range of values for three or four
experiments.
10 CHARLOTTE JACKSON AND ROBERT E. BLACK
were made by the diphenylamine procedure of Dische (1930) as modified by
Seibert (1940). These indicated that only trace amounts were present in each
fraction. The absorbance of the hot TCA extracts at 260 m/z is therefore tentatively
assumed to be a measure of ribonucleic acid content. The results of these deter-
minations are presented in Table III, and the average specific activities of enzymes
and amounts of nucleic acid in each granule fraction are shown in a series of histo-
grams in Figure 2. The results indicate considerable heterogeneity in granule
types. Acid phosphatase and esterase have the highest specific activities in the
granules of lowest density, Subtraction A. Acid phophatase in these granules has
a relative specific activity as much as nine times that of the whole homogenate,
while esterase is about four times as concentrated as in the whole homogenate.
Subfraction B contains the highest percentages of most of the enzymes assayed as
well as the highest percentage of total protein. This is assumed to be a result of
incomplete separation of granules as noted above. The echinochrome pigment
granules are also concentrated in this subfraction in most of the experiments ; how-
ever, see below. Pigment presumably derived from these granules was always
found in the fluid at the top of the density gradient. Aryl-sulfatase, a characteristic
lysosomal enzyme, has somewhat higher specific activities in Subtractions A and
in the denser granules D and E ; however, its distribution is fairly uniform through-
out all the subtractions isolated. Ribonuclease activity and nucleic acid are only
moderately high in Subfraction A ; however, the denser granules, especially D and
E, contain unusually high concentrations of both RNAase and nucleic acid.
In two experiments by one of us (R. B.) the eggs were washed several times in
calcium-free sea water before homogenization, and Fraction I was centrifuged on
the gradient for 90 minutes instead of 60, which resulted in a difference in the
distribution of both echinochrome granules and of acid phosphatase. In these ex-
periments the echinochrome granules did not release any pigment, but were driven
intact through all the density layers to the bottom of the tube. The granules were
evidently in osmotic equilibrium with the 2.5 M sucrose, since they immediately-
ruptured on resuspension in the 0.98 M buffered sucrose. The distribution of acid
phosphatase in these experiments is also indicated in Table III, with the echino-
chrome granule layer designated as "F." In these experiments the highest specific
activity of acid phosphatase was found to be associated with the echinochrome
granules, although the assays were complicated by the presence of the pigment.
The echinochrome granule fraction contains a very small percentage of the total
protein of Fraction I. The highest percentage of total protein and acid phosphatase
was found in Subfraction A in these experiments, while very little protein was
present in the B layer. These results appear to indicate that there are at least two
types of large, acid phosphatase-containing particles, one of which may be the
echinochrome granule.
One of the properties used to define the lysosome is the latency of enzymes in
the intact particle (De Duve, 1963). It was of interest to determine whether the
activities of acid phosphatase and esterase would be affected if the granules of Sub-
fraction A were subjected to different osmotic conditions. It must be noted that
under normal conditions of assay the granules would be ruptured in the dilute
reaction mixture. In order to test the effect of osmotic shock, granules of this
Subfraction were divided into two aliquots. The first was suspended in 0.05 M
ENZYMES IN ARBACIA EGGS 1 1
tris-lO3 M EDTA without sucrose, while the other was suspended in buffered
sucrose. Assays of acid phosphatase and esterase were performed in the usual
manner with the first aliquot, while for the second, all reagents were made up in
buffered 0.98 M sucrose in an attempt to maintain the granules in an intact state
during the reaction. The activity of acid phosphatase was increased 75% by the
rupture of the granules, while esterase activity remained unchanged. Treatment
with the dilute buffer caused an immediate clarification of the granule suspension.
Centrifugation of the clarified suspension at 90,000 g for one hour sedimented ap-
proximately one-third of the phosphatase and two-thirds of the esterase activity.
These results are somewhat inconclusive, since it is obvious that the enzymes of the
intact granules are not completely latent. One possibility is that some of the
granules are damaged in preparation ; however, it is apparent that they still possess
a semipermeable membrane. A second possibility is that the enzymes are really not
latent within the granules.
Microscopical observation and vital staining of granules
The granules in Subtraction A were observed under oil immersion in order to
estimate their size range. The spherical granules varied in size from about one to
three microns, with 60-70% being in the 1-1 1 micron range, and about 20% in the
2-2\ micron range. Only a very few granules were evident in the 3-3^ micron
range, probably less than one or two per cent. No size estimates are available for
granules of the heavier subtractions.
The staining of granules from Fraction I, prior to isolation on the sucrose
layers, gave conclusive results only with toluidine blue. With this stain Sub-
fractions A and E stained slightly, exhibiting a pale green color. Subtractions B,
C and D appeared to exhibit metachromasia to a considerable degree, with most of
the red color concentrated in Subtraction B. Neutral red and methyl red were also
predominantly taken up by Subtraction B. The presence of red echinochrome pig-
ment granules in this fraction interfered with the detection of its staining properties
with all dyes. It seems clear that the granules of Subtraction A, which possess the
highest specific activity of acid phosphatase, do not stain metachromatically in vitro.
A direct analysis of the nature and content of polysaccharides in the different
granules would appear to be desirable from the standpoint of correlating this finding
with that of Dalcq (1963), who reported that granules possessing acid phosphatase
activity also exhibited metachromatic properties.
DISCUSSION
From the results illustrated in Figure 2, it may be tentatively concluded that at
least two types of visible granules, differing in their content of hydrolytic enzymes,
exist in the egg. The first type, exemplified by Subtraction A, exhibits acid phos-
phatase and esterase activities. The second type, found in Subtractions D and E,
contains RNAase and nucleic acid, as well as a relatively high content of aryl-
sulfatase.
The granules of Subtraction A contain at least two of the hydrolytic enzymes
believed by de Duve (1963), Novikoff (1961) and others to be located within the
lysosomes. The absence of metachromasia in these granules after in vitro staining
12 CHARLOTTE JACKSON AND ROBERT E. BLACK
suggests that they may differ from the alpha and beta granules of Dalcq (1963) and
Pasteels and Mulnard (1957) in their polysaccharide content; however, these in-
vestigators worked only with fertilized eggs stained in vivo. Rehhun (1959) dem-
onstrated that in Spisula solidissiuia staining of the alpha and beta granules ap-
peared only after fertilization. In stratifying eggs of various species of sea urchins,
Immers (1960) expressed doubt that regions of mucopolysaccharide concentration
evident after in vivo staining corresponded to the metachromatic alpha and beta
granules of Dalcq and Pasteels because his staining was performed only on un-
fertilized eggs.
In spite of their high content of hydrolytic enzymes, it would be premature to
identify the granules of Subfraction A as lysosomes. De Duve (1963) has warned
that the present definition of the lysosome, although based primarily on his rat
liver tissue work, must not include any incidental details such as size and other
physical characters, osmotic properties, centrifugal behavior, mechanism of struc-
ture-linked latency, or sensitivity to individual disrupting treatments. If these
factors are therefore omitted in defining the lysosome, the essential characteristic
remaining is the association within a special group of cytoplasmic particles of a
number of soluble acid hydrolases of widely differing specificity. The accessibility
of these enzymes to the surrounding substrate must be restricted, making the
latency of the enzymes dependent on the structural complexes of the particles.
Such a definition would be broad enough to include the hydrolytic granules in Sub-
fraction A isolated from Arbacia eggs, if it could be shown more conclusively that
the accesssibility or activity of the enzymes in question is restricted by the granular
structure.
The finding that acid phosphatase activity is associated with the echinochrome
pigment granules is of considerable interest. Since these granules were ruptured
by dilution after recovery from the 2.5 M sucrose layer, the effects of different
osmotic treatments were not tested. Further experiments on these granules are in
progress.
We believe that the mitochondria are concentrated in Subfraction B, since it
contains the highest percentage and specific activity of succinic dehydrogenase ;
however, the data for this enzyme appear to be rather unreliable. This subfraction
is probably heterogeneous, since it contains the highest percentages of all enzymes
and of total protein.
The most dense granules in the visible granule fraction, recovered in Sub-
fractions D and E after gradient centrifugation, contain much higher specific con-
centrations of nucleic acid and RNAase than do the other visible granules. In
preliminary experiments we have made determinations of ammo acid incorporation
into protein of these heavy granules after giving unfertilized eggs a 10-minute pulse
with C14-phenylalanine. After such a pulse, the specific activity of Subfractions D
and E, calculated on the basis of nucleic acid content, is only about one-tenth that
of the microsome fraction ; these granules are therefore quite inactive in protein
synthesis, even in the fertilized egg. We have concluded that there is little con-
tamination from microsomes in this fraction. The existence of dense RNA bodies
in eggs has been reported by other workers. Raven (1945) demonstrated the
presence of heavy RNA particles in the centrifugal pole of stratified Linmaea eggs.
Pasteels (1958), by centrifuging Paracentrotus eggs, discovered "heavy bodies" of
ENZYMES IN ARBACIA EGGS 13
UNA, ranging from 1-3 microns, in the centrifugal cap region. This region, which
also contained the mitochondria, was intensely stained with pyronine. Pasteels
postulated that, in addition to being found in the ribosomes and in annulate mem-
branes within the egg, RNA could also be found in undefined structures that could
be linked to the mitochondria but which contained the most dense material in the
egg. Balinsky and Devis (1963) observed electron-dense granules in the young
oocytes of Xenopits lacris which presumably accumulated between adjacent mito-
chondria. Afzelius (1956) has also described "heavy bodies" which stain vitally
with toluidine blue in the sea urchin egg. Immers (1960) described dense RNA
granules which were separate from the mitochondria in the most centrifugal zone
of stratified eggs of Paraccntrotus lii'idns. A few workers have claimed that the
heavy yolk granules, especially in the Amphibia, contain an appreciable amount of
RNA '(Grant, 1953; Rounds and Flickinger, 1958), but others have shown by
histochemical and cytological studies that there is little or no RNA within the yolk
granules of most species examined. Collier (1960) found no evidence of either
RNA or proteolytic enzymes in the yolk granules of Ilyanassa obsolete!. The recent
work by Karasaki (1963) and Ohno ct al. (1963) revealed no evidence for the
presence of RNA in the yolk granules of Triturns pyrrhoyaster and Rana [>ipiais
embryos.
It may be tentatively concluded that the heavy granules in Subtractions D and
E correspond to those described by Immers and Pasteels. It is possible that the
annulate lamellae described by Pasteels (1958) are sufficiently dense to be included
in this fraction ; these structures consist of membranes to which bodies similar in
size and density to ribosomes are attached. No previous report has been made
concerning the association of RNAase with any large granule fraction in the egg :
however, a comparison of the present finding with that of Reid and Node (1959)
for granules of rat liver is of particular interest. These authors provided evidence
that acid RNAase was present in particles which were more rapidly sedimented
from homogenates than the lysosomes, indicating the possible existence of a separate
set of granules which contain this enzyme.
The distribution of activities of the hydrolytic enzymes in Fractions II and III
of the egg homogenates indicates the presence of these enzymes in submicroscopic
structures (see Table II). The possibility therefore exists that granules resembling
rat liver lysosomes in size may also be present in the egg. If such granules are
easily ruptured during preparation, as are liver lysosomes, this may account for the
high enzyme activities found in the soluble fraction. It is of interest that Fractions
II and III of the gut tissue of adult Arbacia (Table I ) contain approximately the
same specific activities of acid phosphatase and esterase as do the corresponding
fractions of the egg.
The heterogeneity of the populations of granules which contain hydrolytic
enzymes in the egg may be generally related to the timing with which different
enzymes become active during development. It is postulated that such a separation
of enzymes in different granules could result in the specific release or activation of
some hydrolases, but not others, at particular developmental stages. Furthermore,
partial segregation of the granules into different cells during cleavage may confer
different developmental potentialities on the daughter cells. Segregation of granules
and certain enzymes have been observed in numerous eggs exhibiting "mosaic"
14 CHARLOTTE JACKSON AND ROBERT E. BLACK
cleavage (cf. Brachet, 1950, for review) ; however, similar differentiation has not
heen observed in the sea urchin. Experimental testing of the latter hypothesis must
await the development of techniques for visual identification of the granules in
question.
SUMMARY
Differential centrifugation and density gradient centrifugation have heen applied
to nuclei-free homogenates of unfertilized eggs and adult gut of Arbacia, to deter-
mine the distributions of several hydrolytic enzymes and of nucleic acid and protein.
Two types of large visible granules have been partially separated from egg homog-
enates by gradient centrifugation. The first type is rich in acid phosphatase and
esterase ; the second contains sulfatase, RNAase and nucleic acid. The activities
of the above enzymes have also been determined in microsomal and soluble fractions
of the egg, and the distribution of acid phosphatase and esterase have also been
determined in the major subcellular fractions of the adult gut of Arbacia. In both
types of homogenates hydrolytic enzymes were found to be present in submicro-
scopic granules and in the supernatant fluid. The major difference in the two types
of material is that large granules containing the enzymes are present in the eggs
but not in the gut tissue.
LITERATURE CITED
AFZELIUS, B. A., 1956. The ultrastructure of the cortical granules and their products in the
sea urchin egg as studied with the electron microscope. Exp. Cell Res., 10: 257-285.
ALLEN, R. D., 1957. Isolation of cortical material from sea urchin eggs by centrifugation. /.
Cell. Comp. Physiol, 49: 379-394.
ANSON, M. L., 1938. The estimation of pepsin, trypsin, papain and cathepsin with hemoglobin.
/. Gen. Physiol, 22: 79-91.
APPLEMANS, F., R. WATTIAUX AND C. DE DUVE, 1955. Tissue fractionation studies. 5. The
association of acid phosphatase with a special class of cytoplasmic granules in rat liver.
Biochem. J., 59: 438-445.
BALINSKY, B. I., AND R. J. DEVIS, 1963. Origin and differentiation of cytoplasmic structures in
the oocytes of Xenopus laevis. Acta Embry. Morph. Exp., 6: 55-108.
BERTHET, J., AND C. DE DUVE, 1951. Tissue fractionation studies. 1. The existence of a
mitochondria linked, enzymically inactive form of acid phosphatase in rat-liver tissue.
Biochem. J., 50: 174-181.
BONNER, W. D., 1955. Succinic dehydrogenase. In: Methods in Enzymology, Vol. Ill, S. P.
Colowick and N. O. Kaplan, eds., Academic Press, Inc., New York.
BRACKET, J., 1950. Chemical Embryology. Interscience Publishers, Inc., New York.
BURCH, H. B., O. A. BESSEY, R. H. LOVE AND O. H. LOWRY, 1952. The determination of
thiamine and thiamine phosphates in small quantities of blood and blood cells. /. Biol.
Chcm., 198: 477-490.
COLLIER, J. R., 1960. The localization of ribonucleic acid in the eggs of Il\anassa obsoleta.
Exp. Cell. Res., 21: 126-136.
DALCQ, A. M., 1963. The relation to lysosomes of the in vivo metachromatic granules. In :
Ciba Foundation Symposium on Lysosomes, A. V. S. de Reuck and M. P. Cameron,
eds., Little, Brown and Co., Boston.
DE DUVE, C., 1963. The lysosome concept. In : Ciba Foundation Symposium on Lysosomes,
A. V. S. de Reuck and M. P. Cameron, eds., Little, Brown and Co., Boston.
DE DUVE, C., J. BERTHET, AND H. BEAUFAY, 1959. Gradient centrifugation of cell particles:
theory and applications. Prog. Biophysics and Biophysical Chem., 9: 325-369.
DE DUVE, C., B. C. PRESSMAN, R. GIANETTO, R. WATTIAUX AND F. APPELMANS, 1955. Tissue
fractionation studies. 6. Intracellular distribution patterns of enzymes in rat-liver
tissue. Biochem, /., 60: 604-617.
ENZYMES IN ARBACIA EGGS 15
DE REUCK, A. V. S., AND M. P. CAMERON, eds., 1963. Ciba Foundation Symposium on
Lysosomes. Little, Brown and Co., Boston.
DUBOS, R. J., AND R. H. S. THOMPSON, 1938. The decomposition of yeast nucleic acid by a
heat-resistant enzyme. /. Biol. Chcm., 124: 501-510.
DISCHE, Z., 1930. Some new characteristic color tests for thymonucleic acid and a micro-
chemical method for determining the same in animal organs by means of this test.
Mikrochcmic, 8: 4-32.
GRANT, P., 1953. Phosphate metabolism during oogenesis in Rana temporaria. J. Exp. Zoo/..
124: 513-543.
HARVEY, E. B., 1954. Electrical method of determining the sex of sea urchins. Nature, 173: 86.
HUGGINS, C, AND J. LAPRiDES, 1947. Chromogenic substrates. IV. Acyl esters of />-nitro-
phenol as substrates for the colorimetric determination of esterase. /. Biol. Chcm.,
170; 467-482.
IMMERS, J., 1960. Studies on cytoplasmic components of sea urchin eggs stratified by centri-
fugation. Exp. Cell Res., 19: 499-514.
KARASAKI, S., 1963. Studies on amphibian yolk. 5. Electron microscopic observation on the
utilization of yolk platelets during embryogenesis. /. infrastructure. Res., 9: 225-247.
LOWRY, O. H., N. J. ROSEBROUGH, A. L. FARR AND R. J. RANDALL, 1951. Protein measurement
with the Folin phenol reagent. /. Biol. Chcm., 193: 265-275.
NOVIKOFF, A. B., 1961. Lysosomes and related particles. In : The Cell, Vol. II, J. Brachet
and A. E. Mirsky, eds., Academic Press, Inc., New York.
OH NO, S., S. KARASAKI AND K. TAKATA, 1963. Histo- and cytochemical studies on the super-
ficial layer of yolk platelets in the Triturus embryo. Exp. Cell Res., 33: 310-318.
PASTEELS, J. J., 1958. Comparative cytochemistry of the fertilized egg. In : A Symposium on
the Chemical Basis of Development. W. D. McElroy and B. Glass, eds., Johns
Hopkins Press, Baltimore.
PASTEELS, J. J., AND E. DE HARVEN, 1963. Study by electron microscopy of the cytoplasm of
the fertilized and unfertilized egg of Barnca Candida. Arch Biol. (Liege), 74: 417-437.
PASTEELS, J. J., AND J. MULNARD, 1957. Toluidine blue in metachromasia in vivo and its
cytochemical analysis in eggs of B. Candida, G. angulata, and P. miliarius. Arch. Biol.
(Liege), 68: 115-163.
PASTEELS, J. J., AND J. MULNARD, 1963. The relations between alpha and beta meta-granules
in the egg of the bivalve mollusk Barnca Candida. Arch. Biol. (Liege), 74: 319-340.
RAVEN, CHR. P., 1945. The development of the egg of Limnaca stagnalis from oviposition till
first cleavage. Arch. Necrl. Zoo/., 7: 91-119.
REBHUN, L. I., 1959. Studies of early cleavage in the surf clam, Spisula solidissima, using
methylene blue and toluidine blue as vital stains. Biol. Bull., 117: 518-545.
REID, E., AND J. T. NODE, 1959. Liver ribonucleases. Ann. N. Y. Acad. Sci., 81: 619.
ROUNDS, D. E., AND R. A. FLICKINGER, 1958. Distribution of ribonucleoprotein during neural
induction in the frog embryo. /. Exp. Zool., 137 : 479-499.
SCHEIB-PFLEGER, D., AND R. WArriAux, 1962. fitude des hydrolases acides des canaux de
Muller de 1'embryon de Poulet. 1. Activites totales et solubles des canaux d'embryons
de 8 a 10 jours d'incubation. Dcv. Biol., 5: 205-217.
SCHMIDT, G., AND S. J. THANNHAUSER, 1945. A method for the determination of desoxyribo-
nucleic acid, ribonucleic acid and phosphoproteins in animal tissues. /. Biol. Chcm.,
161: 83-89.
SCHNEIDER, W. C., 1957. Determination of nucleic acids in tissues by pentose analysis. /;; :
Methods in Enzymology, Vol. Ill, S. P. Colowick and N. O. Kaplan, eds., Academic
Press, Inc., New York.
SEIBERT, F. B., 1940. Removal of the impurities, nucleic acid, and polysaccharide from tuber-
culin protein. /. Biol. Chem., 133: 593.
WALLENFELS, K., 1962. 5c/a-galactosidase (crystalline). In: Methods in Enzymology, Vol.
V., S. P. Colowick and N. O. Kaplan, eds., Academic Press, New York.
WEBER, R., 1963. Behaviour and properties of acid hydrolases in regressing tails of tadpoles
during spontaneous and induced metamorphosis in vitro. In : Ciba Foundation Sym-
posium on Lysosomes, A. V. S. de Reuck and M. P. Cameron, eds., Little, Brown and
Co., Boston.
OXYGEN UPTAKE AND RESPONSES TO RESPIRATORY
STRESS IN SEA URCHINS1
KJELL JOHANSEN2 AND ROBERT L. VADAS 3
Friday Harbor Laboratories and Department of Zoology,
University of U'asJiiin/ton, Seattle. U'ashinf/ton 9S1Q5
In discussions of oxygen consumption in invertebrates it is common practice to
classify animals as oxygen conformers or regulators. The conformers are often
referred to as being respiratory-dependent whereas the regulators are independent.
Such classifications imply that the animals either vary their oxygen uptake relative
to the oxygen availability in the external medium, or they show some degree of
regulation by maintaining oxygen uptake in spite of a decrease in the surrounding
oxygen availability. An animal showing respiratory independence usually becomes
respiratory dependent as the external oxygen tension (PiO2) decreases. The PiCX
at which this occurs is referred to as the critical oxygen tension. The concept of
respiratory dependence implies that the animal is at the mercy of the existing en-
vironmental conditions. Although such classifications may serve to clarify similari-
ties between various animals, they may obscure basic characteristics of animal
energy economy and of important factors affecting interaction of the organism and
its environment. We contend in opposition to current teaching that an animal's
oxygen uptake must have a closer correlation with internal oxygen tensions than
with the external oxygen availability. Recent advances in gas analyzing techniques
have simplified the measurement of internal oxygen tensions. The present report
is based on measurements of oxygen uptake (VCX) of three species of sea urchins.
Internal as well as external oxygen tensions were monitored as the animals were
subjected to a variety of external conditions.
MATERIALS
Two of the three species of the genus Strongylocentrotus used in these experi-
ments are extremely abundant members of the marine fauna in the vicinity of the
marine laboratory at Friday Harbor, Washington. .9. droebachiensis and S. jran-
ciscanus exhibit a marked subtidal zonation, with the former usually occupying the
lower intertidal and the upper subtidal zones, while the latter is generally found
several meters below zero tide. However, both species have been collected from
depths as great as 35 meters and as intertidal individuals in slightly exposed areas.
6". purpuratns, on the other hand, is mainly found intertidally in tide pools on very
exposed, wave-beaten rocks. Natural and transplanted populations of S. f>itrf>uratns
1 This work was supported by grant GB-4038 from the National Science Foundation, and
by NSF grant G-20901 to the Friday Harbor Laboratories for support of the junior author.
2 This work was done during the tenure of an Established Investigatorship of the American
Heart Association.
3 Botany Department, University of Washington, Seattle, Washington 98105,
16
OXYGEN UPTAKE IN SEA URCHINS 17
were located in limited areas around Friday I larbor. In addition, several collections
were made from the exposed coast, at Mukkaw Bay.
All animals were stored in running, oxygenated, sea water for at least one week
prior to use. \Yater temperatures varied from 9.5° to 10.5° C, and the animals
were maintained under relatively low light conditions (about 40 foot candles).
Upon completion of experiments, animals were returned to similar tanks and
observed over a period of several weeks for any signs of ill effects as a result of the
experimental procedures.
METHODS
Partial pressures of oxygen in the external water (PiO2) and in the coelomic
fluid (PcfO., ) were measured using a Beckman Spinco gas analyzer (model 160).
A special micro-cuvette permitted duplicate analyses of all samples (sample size:
0.05 nil.). The oxygen electrodes were calibrated with known gas mixtures or
solutions equilibrated to known gas composition. Repeated sampling of coelomic
fluid was accomplished by means of polyethylene catheters (P.E. 90) chronically
implanted through a small incision in the peristomial membrane with the catheter
tip protruding into the large coelomic cavity. Leakage around the catheter was
prevented by a purse string suture in the peristomial membrane. Measurements of
PCX were made at a temperature corresponding to the prevailing temperature in the
animal's immediate surroundings.
Oxygen uptake (VO2) was measured using a closed system at constant temper-
ature (10.0° C.) (Lenfant, 1961). The partial pressure of oxygen in the respirom-
eter was slowly reduced by the oxygen consumption of the animals. The size of
the respiration chamber was adjusted until a reduction in PiO2 from 150 mm. Hg to
10-20 mm. Hg occurred within a 6-8 hour period. VO2 was calculated as micro-
liters O,/g. wet weight/hour. Both PiO2 and PcfO2 were measured in successive
samples during the oxygen uptake experiments. An integral part of the investiga-
tion involved sampling of coelomic fluid from animals in their natural habitats.
Such samples were obtained while SCUBA diving and collected into greased glass
syringes by a needle inserted into the coelomic cavity through the peristomial mem-
brane. The samples were quickly brought to the laboratory for analyses of PcfO2.
RESULTS
Figure 1 shows a plot based on the average of three experiments utilizing six
animals. Oxygen uptake (VO.,) (left ordinate, open circles) arid PO2 in coelomic
fluid (PcfO2) (right ordinate. filled circles) are plotted against oxygen tension in
the ambient water PiO., in a closed system. In addition, values of PcfO2 obtained
from animals in their normal habitats have been plotted as a function of the PiO2
at which the samples were taken (open squares, Fig. 1).
In well aerated water there was always a gradient in PO2 between the coelomic
fluid and the ambient water (Figs. 1 and 2). Upon lowering of the PiO2, VO2
stayed relatively unchanged down to values of 60-70 mm. Hg. During this decrease
in the O2 availability of the ambient water, the oxygen tension in the coelomic fluid
characteristically showed two types of response patterns. In most cases PcfO.,
started to increase as soon as PiO2 was lowered (Figs. 1 and 2). In other cases
18
KJELL JOHANSEN AND ROBERT L. VADAS
//l/g/hr
mmHg
10-
Strongylocentrotus purpuratus .
-100
9-
°Vo2
• Poa coelomic fluid (Lab) ••
- 90 "I
^8-
° Poa coelomic fluid (Nature)
• • 0 o
-80-^
^
0 ° o
^
j. 7-
0 0 0 0
-70 -§
•5
°o»° °0 On
§
1- 6"
00 8 •
-60 ^
£
0° »°
••».
| 5-
O
-50.§
^
• • a
*x.
4-
c;
0
°
-40 |
^
* ° • • D
c;
^ 3'
,
-30 |
Q
o
iv
Q
v
-
-20 5
*.°* °
\ •
I* • *
- 10
° §
o B
20 40 60 80 100
Oxygen tension in water
120
140
FIGURE I. Oxygen consumption VO2 (left ordinate, open circles), and partial pressure of
oxygen in coelomic fluid, PcfO2 (right ordinate, filled circles), of Strongylocentrotus purpuratus
plotted against the partial pressure of oxygen in ambient water. PcfOa in samples obtained
from animals in nature are also plotted (right ordinate, open squares).
it was maintained but in no case did PcfO2 drop initially when PiCX decreased.
The increased or maintained PcfCX rapidly reduced the gradient in O2 tension be-
tween the coelomic fluid and the external environment. At the breaking point in
oxygen uptake, at a PiO, of 60-70 mm. Hg, the internal and external media were
essentially in equilibrium with respect to PO2. Some of the values for PcfO, were
considerably higher than the ambient water PCX (Figs. 1 and 2). Inadequate
stirring in the coelomic fluid space may have been responsible for the seemingly
paradoxical situation. However, Newell and Courtney (1965), working on re-
spiratory movements in a holothurian, observed a similar response and maintained
that the increased oxygen concentration of the coelomic fluid resulted from the
animal's ability to absorb water and from a delayed transfer of oxygen from the
respiratory trees to the coelomic fluid. Such an explanation, however, seems un-
likely in the case of sea urchins.
When the PiCX dropped below values of 60-70 mm. Hg, both VO2 and PcfCX
dropped sharply and at about the same rate (Figs. 1 and 2). At very low oxygen
tensions in the ambient water (19-20 mm. Hg) the VCX and PcfCX started to level
off and were observed to remain largely unchanged at that low PiO2 for periods of
at least 5-6 hours. VO2 was then reduced to approximately 15% of its value in
air-saturated water. All animals subjected to experiments of this nature completely
recovered after being transferred back to aerated sea water.
Figure 3 shows the composite results (average of 4 runs on each species) from
experiments designed to study the gas exchange of sea urchins when placed in air.
OXYGEN UPTAKE IN SEA URCHINS
19
The oxygen tension in the coelomic fluid followed a similar course for all three
species. During the first hour after transfer to air, the PcfO2 fell rapidly but sub-
sequently levelled off and remained relatively unchanged from the second to at least
the twelfth hour after initial exposure to air. All experimental animals recovered
after being returned to normal sea water. The water and air temperatures during
these experiments were 10° C. and 17° C., respectively.
mmHg
150
Strongylocentrotus purpuratus
100-
° Po2 water
• Po2 coelomic fluid
50-
§
i
100-
50-
0-
Strongylocentrotus droebachiensis
° Po2 water
• Po2 coelomic fluid
Time -Hours
FIGURE 2. Partial pressures of oxygen in ambient water and coelomic fluid plotted against
time for S. purpuratus and S. droebachiensis, when the urchins were in a closed respiration
chamber.
20
KIKI.I. inllAXSKX AND ROBKRT L. VADAS
I )lSCUSSION
The water vascular svslem of echinoids constitutes their primary means for
external gas exchange (Roller and Meyer, 1933, Steen, 1965). The system termi-
nates externally in the podia or tube feet which make up the major surface for gas
exchange. The external surface of echinoids is irrigated by ciliary currents. In
addition, the movement of the tube feet increases the circulation of the external
water. The water vascular system is lined internally by a flattened ciliated epithe-
lium responsible for maintaining the water current carrying the respiratory gases.
However, echinoids do not possess an effective system for internal oxygen transport.
mmHg-
140-
120-1
.SiooH
5
I
80-
60-
40-
20-
0-
iWaterS Air
0 Strongylocentrotus droebachiensis
• Strongylocentrotus purpuratus
D Strongylocentrotus franciscanus
T~
0
~r
2
— r
4
Time - Hours
FIGURE 3. Partial pressure of coelomic fluid during air exposure of S. drocbachiciisis.
S. piirpunitus and S. franciscanus.
A large coelomic space lined with a ciliated epithelium constitutes the link between
the water vascular system and the main mass of metabolizing tissue suspended
within the spacious coelomic compartment. Although this arrangement does not
insure an effective rate of transfer of oxygen from the ambient water to the respiring
tissues, one must keep in mind the obvious role of the large coelomic fluid space as
a storage compartment for the respiratory gases. Internal circulation based on fluid
movement in discrete vessels and capillary beds does not offer a comparable storage
factor. The importance of this in intertidal forms is amplified when considering
the long interruptions of effective external gas exchange during tidal exposure. The
role of the storage factor in respiratory exchange in holothurians has recently been
alluded to by Newell and Courtney (1965).
The present data show that the oxygen tension of the coeloir.ic fluid offers a
good indication of the oxygen uptake of the animal, i.e., that the rate of aerobic
OXYGEN UPTAKE IN SEA URCHINS 21
metabolism closely follows the oxygen availability in the internal environment
bathing the metabolizing tissues. This in essence is carrying the concept of respira-
tory dependence a step further using the internal rather than the external environ-
ment as the reference for the oxygen uptake. Additional evidence for the close
dependence of VO2 on the PO2 of the coelomic fluid was obtained in experiments
involving exposure of the urchins to hyperoxygenated water (PiO2: 350-500 mm.
Hg). This procedure led to a rapid increase in PcfO2 to values approaching that
in the ambient water. Simultaneous monitoring of the oxygen uptake showed a
conspicuous increase. It was imperative to extend these experiments over a long
period of time (10-15 hours) in order to separate the true level of oxygen uptake
from the mere storage of oxygen in the spacious coelomic fluid compartment.
Giese et al. (1966) suggest a similar dependence of VCX on the internal oxygen
concentration, but their experiments were not designed to offer direct information
on the problem.
The physiological significance of the close correlation between VO2 and PcKX
becomes paramount in view of our data concerning the PCX levels in coelomic
fluid sampled from animals in their normal environment. These data (represented
by the squares in Figure 1 ) demonstrate a very large variation in PcfCX which in
turn would indicate a similar variability in the actual oxygen uptake. One must
now remember that all of these samples of coelomic fluid were obtained from
animals located in water essentially air-saturated with oxygen (PiO2 > 130 mm.
Hg). Thus, in spite of a high CX availability in the ambient water the animal may
decrease external respiratory efforts with a consequent reduction of coelomic fluid
PCX and a reduction of overall CX uptake. In other words, rather than having its
metabolic rate changed at the mercy of the external environment, the animal can
make use of the phenomenon of respiratory dependence by lowering its internal CX
level in well oxygenated water and thus reduce the oxygen uptake, thereby con-
serving energy when it is not needed.
The applicability of this concept, in a larger biological sense, seems rather im-
portant in many lower forms where only intermittent needs for a high operational
ability are present, for instance, during feeding, escape and reproductive activities
or in cycles depending on internal biochemical events.
The actual values for oxygen uptake of ^. pitrpuratus obtained in the present
study are much lower than those reported by Giese et al. (1966), but they compare
well with the value listed for S. lividus (Spector, 1956).
The tolerance to air exposure as demonstrated in the present study is relevant
to the problem of intertidal distribution of sea urchins. Our results show that
during exposure to moist air at moderately increased temperatures (5° C. maximum
increase) the external gas exchange can support a steady oxygen uptake at a level
approximately 1/5 to 1/7 of the maximum in aerated water. Periods up to 15 hours
of exposure to the described conditions were compatible with survival for the species
investigated. The data warrant the suggestion that the endurance limit to air ex-
posure at low tide may be more dependent on temperature increases and desiccation
than on the ability of the urchins to exchange gases with air.
The capacity for compensatory alteration of the external gas exchange by means
of the water vascular system became evident from the experiments involving gradual
exposure to hypoxic water (Fig. 2}. All animals responded to this by a compensa-
22 KJELL JOHANSEN AND ROBERT L. VADAS
tion that maintained or even increased the PcfO2 as the PiO, dropped. It is inter-
esting to consider this compensation on the background of recent findings by Steen
(1965) who showed that the oxygen uptake measured during the course of his
experiments on Strongylocentrohis drocbachicnsis was only about one-tenth of the
theoretical capacity of their podia. Steen relates this apparent inefficiency to limita-
tions in the actual transport mechanisms of external and internal media. A com-
pensation as demonstrated seems all the more likely when the limitation in external
gas exchange is set by the movement of the respiratory media rather than by the
thickness and area of the exchange surface.
The assistance of Stewart Grant and Dennis Willows in the diving operations of
this study and the thoughtful comments by Dr. David Hanson are greatly appre-
ciated.
SUMMARY
1. Oxygen uptake (VO2) by sea urchins has been measured and correlated with
partial pressures of oxygen in the ambient water (PiCX) and in coelomic fluid
(PcfO,). PcfO, was also analyzed in samples obtained from animals in natural
environments using SCUBA-diving technique. In addition, changes in PcfO2
were recorded during prolonged air exposure of the animals.
2. The three species investigated, Strongylocentrotus purpuratus, S. franciscamts
and S. drocbachiensis, showed steady levels of VO2 until PiO, had dropped to 60-70
mm. Hg. At lower PiO, the oxygen uptake decreased corresponding to the rate
of decline of PiO2.
3. Changes in PcfO2 at decreasing PiO2 closely paralleled the changes in VO2,
except for a common initial compensatory increase in PcfO,. The data indicate that
PcfO2 represents a useful index of the level of VO2.
4. PcfO, in samples obtained from urchins in their natural habitat showed large
variations. Means by which the relationship between PcfO2 and VO2 may actively
be used by the animals in their energy economy are discussed.
5. Air exposure of the urchins while monitoring changes in PcfO, suggests that
external gas exchange in air is not a critical survival factor during tidal exposure.
LITERATURE CITED
GIESE, A. C, A. FARMANFARMAIAN, S. HILDEN AND P. DOEZEMA, 1966. Respiration during
the reproductive cycle in the sea urchin, Strongvloccntrotus purpuratus. Biol. Bull.,
130: 192-201.
HANDBOOK OF BIOLOGICAL DATA, 1956. Ed. W. S. Spector, W. B. Saunders Co., Philadelphia
and London : 584 pp.
ROLLER, G., AND H. MEYER, 1933. Versuche iiber die Atmung der Echinodermen (Asterias
rubens und Echinus miliaris). Biol. Zbl., 53: 655-661.
LENFANT, C., 1961. A method for measuring VC>2 and VCO2 of very small sea animals. /.
Appl. Physiol., 16: 768-770.
NEWELL, R. C., AND W. A. M. COURTNEY, 1965. Respiratory movements in Holothuria
forskali (Delle Chiaje). J. Exp. Biol., 42: 45-57.
STEEN, J. B., 1965. Comparative aspects of the respiratory gas exchange of sea urchins. Ada
Physiol Scand., 63: 164-170.
SPERM ATOCYSTS IN AEDES AEGYPTI (LINNAEUS) *
JACK COLVARD JONES
Department of Entomology, University of Maryland, College Park, Maryland 20740
It has been known for some time that groups of germinal cells in the testes of
various mosquitoes are separated hy delicate, membranous, transverse lamellae into
a series of chambers or compartments (Hurst, 1890; Kulagin, 1907 ; Cholodkowsky,
1905; and Lomen, 1914) which can be referred to as spermatocysts. The aims of
the present study were to determine the number and character of these compart-
ments during the life span of the Bangkok strain of the Yellow Fever mosquito
Aedes (Stegomyia) aegypti (Linnaeus) and to ascertain whether the spermatocysts
would reflect the sexual activity of adults.
Larvae were reared in batches of 100 in 250 ml. of water in stender dishes and
were well-fed. Pupae were sexed by examining the difference in their external
genitalia (see Christophers, 1960) and allowed to emerge as adults in a cubic-foot
screened cage, where they had continuous access to sugar water. The temperature
varied from 25° to 30° C. The testes were dissected into a small drop of Drosophila
saline (Ephrussi and Beadle, 1936) with fine needles and micro-forceps, and were
examined with and without phase contrast microscopy, either without coverslipping
or after slight flattening under a coverglass. Such flattening was often very useful
when the testes were heavily encased in fat body, but in many cases even after
flattening the fat body still completely obscured one or more portions of the testes
so that accurate counts could not be made. The fat body could sometimes be
removed by gently pushing the coverslip with a wet piece of filter paper which
caused the testes to roll over in the wet whole mount. In a number of cases the fat
body jacket was seen to be stripped away as the testes were being pulled out of the
body. The number of compartments was generally counted and categorized at a
magnification of 430 X. Often two to four counts were made on the same region
of a single testis and these were averaged. Frequently only one portion of a testis
was sufficiently visible for accurate counts ; hence, total numbers of cysts in Tables
II through VI include only the complete counts. Wherever possible, the character
of the cells in the different compartments was categorized as either undifferentiated
( = spermatogonia, spermatocytes, very early spermatids) or as partially to fully
differentiated. In some cases, however, only the number of cysts could be counted.
In some other cases, the presence of maturation could be detected but accurate
counts were not possible. Although it was often possible to discriminate partially-
from fully-differentiated spermatocysts, this distinction could not always be made
with certainty. Cells in the anteriormost portion of the testes were often very in-
1 This research was sponsored by N. I. H. Grant GM 06021 and by N. I. H. Development
Award GM 21,529. Scientific Article Number A 1289, contribution number 3829 of the Mary-
land Agricultural Experiment Station.
I am indebted to Dr. Arden O. Lea, Elliot S. Krafsur and Elizabeth Jones for many helpful
comments on the manuscript.
23
24
JACK COLVARD JONES
distinctly separated and the counting error is generally greater for this region than
for the partially- to fully-differentiated cysts. The precise ages of pupae and adults
were obtained by watching for the moment of ecdysis. These studies were made
in the summers of 1900, 1901 and 1966.
OBSERVATIONS
Larvae were dissected daily from the first through the seventh days after hatch-
ing, independently of their particular stadium. Forty-two gonads dissected from 34
larvae were identified as testes. The number of spermatocysts ranged from three
in one-day-old larvae to a maximum of 24 in a 6-clay-old fourth stage larva. The
mean numbers of cysts for the larval period varied from 3.8 to 22.5 (Table I).
Only one out of 8 testes from five-day-old larvae showed evidence of early differ-
entiation of the germ cells in the posteriormost compartment. The cells in this
case were pyriform spermatids. In none of the larvae examined were fully differ-
TABLE I
Daily speniiatocyst counts of Aedes aegypti (L.) during the larval stage
Number of spermatocysts
Days after
hatching
No. larvae
dissected
No. testes
No. counts
Range
Mean
1
4
4
6
3-5
3.8
2
5
6
10
4-10
7.3
3
5
7
7
10-18
14.3
4
6
7
15
16-22
19.9
5
7
8
11
15-21
18.0
6
4
4
->
21-24
22.5
7*
5
8
13
9-23
16.4
* Data collected 1966; other data 1960-1961.
entiated spermatozoa observed. Last stage larvae had from 15 to 24 testicular
compartments, with an average of about 20. During larval life, the testes increased
8 to 9 times in length and changed from small ovoid organs into long pyriform
structures. Most of the testes dissected from larvae were free of fat body up to the
fourth stage. In fourth stage larvae, one of the testes tended to be noticeably
larger than the other in the same individual. It is estimated that the number of
spermatocysts increases about five times during larval life. The maximum number
of cysts is reached several days before pupation.
During the pupal stage the number of spermatocysts ranged from 9 to 29, with
means fluctuating around 19 (Table II), thus indicating that the number of
testicular compartments remains essentially the same as that of fourth stage larvae.
Although 8 out of 9 testes dissected from 6 newly emerged "white" pupae (0-8
minutes old) had from one to four cysts (Table II) with spermatids in the spindle
form or with flagellum formation beginning, some other batches of newly emerged
pupae (0-10 minutes old) which were examined during these studies had no differ-
entiating cysts at all (11 testes from 6 pupae; data not shown). Indeed, a few
testes had not begun to differentiate sperm for as long as 21 hours after pupation.
SPERMATOCYSTS IX AKDES
25
In some batches of pupae, a few spermatocysts were matured as early as five hours ;
in others, fully matured spermatozoa were not seen until near the end of the first
day. The number of differentiated cysts definitely increases in old "black" pupae
about to emerge as adults (that is, within pharate adults), when as many as 13
matured cysts were found in one case. The matured sperm were not active within
these cysts. Sperm were never observed in the sperm ducts during the pupal stage.
Spermateleosis (process of differentiation of spermatozoa from spermatids)
always begins within the posteriormost compartment of the testes and proceeds
anteriorly until as much as 75% to 80% of the gonads may have differentiated cysts.
TABLE II
Spennatocyst counts of Aedes aegypti (L.) during the pupal stage
X umber of cysts/testis
Extent of differentiation
Age
No. dis-
sected
No.
testes
Totalf
None
Begun
Complete
Range
Mean
Range
Mean
Range
Mean
Range
Mean
0-8 min.*
6
9
11-17
14.8
0-4
2.3
0
13-21
17.1
25-45 min.*
5
9
14-19
16.7
1-2
1.2
0
—
15-20
17.5
3J-4 hr.
5
6
15-29
20.0
0-4
0.6
0
—
15-29
20.6
5 hr.
6
12
9-22
15.3
0-
0-5
1.8
9-23
17.0
5-6 hr.
5
8
15-24
21.3
0-
0-
15-24
21.3
16J-17£hr.
9
14
11-23
17.7
0.4
1.5
0-
13-23
19.2
18-21 hr.
10
13
14-23
18.9
1-4
2.0
0-
15-25
19.9
21-22 hr.*
4
7
10-19
14.8
0-1
—
3-6
4.7
15-23
20.1
24 hr.
3
5
13-20
16.0
1-4-
2.2
0-8^
3.4
18-24
21.6
27-29 hr.
3
4
11-15
13.0
—
—
4-6^
5.0
16-20
18.0
Old pupae*
7
9
9-12
10.8
0-1
0.2
6-12*
9.3
18-24
20.6
Old pupae**
5
8
6-10
7.9
—
—
6-13
8.7
12-20
16.5
t Totals do not include partial data.
* 1966 data; all other data are for 1960-1961.
* Pharate adults.
r/ Based on 10 measurements.
* Based on 3 measurements.
t Combined partially to fully differentiated cysts.
The anteriormost region of the testis always maintains an undifferentiated zone of
cysts. Although there is a tendency for a number of cysts in both testes within an
individual to begin to mature at the same time, individuals have been seen where
maturation had begun in one testis and not in the other. There was no correlation
between the presence or absence of a lumen in the vasa efferentia and the matura-
tion of the testes.
Within the matured testis the non-differentiated anterior portion is made up of
spermatocysts in bands or layers one to three or more cells thick in optical section.
The cells are relatively large and spherical and each has a large nucleus. (Before
the testis matures, all of the germ cells have this appearance and the cysts are small
anteriorly and progressively tend to enlarge posteriorly.) The spermatids pass
26
JACK COLVARD JONES
through 8 stages to become mature spermatozoa (Krafsur, 1964). After passing
through the pyriform and spindle stages, the cells progressively elongate and a
flagellum is formed. Cysts with early differentiating spermatids do not have a
distinct color. As differentiation proceeds, the cysts take on a distinct yellowish
brown cast with transmitted light. Differentiating cysts tend to be larger than non-
differentiated ones. The terminal cyst is the largest testicular compartment ; and,
after the first day of adult life, the spermatozoa actively move about within it. often
TABLE III
Sperniatocyst counts on testes of unmated Aedes aegypti (L.) during the adult stage
Numbers of cysts/testis
Extent of differentiation
Age and status
No. dis-
sected
No.
testes
None
Partially to fully
differentiated
Totalf
Range
Mean
Range
Mean
%
Range
Mean
Newly emerged ;
annotated
15
18
10-17
12.7
5-18
9.8
43.1
18-34
22.5
0 hour
7
12
5-11
9.2
3-14
9.5
48.2
13-25
19.7
5 hour*
6
11
5-9
7.5
7-12
9.8
56.6
15-21
17.3
£ rotated*
5
7
7-16
10.5
8-11
9.5
48.8
16-25
20.0
10 hour (i r)*
4
8
6-11
8.1
7-13
10.0
53.8
15-22
18.6
3 rotated
9
13
6-17
11.2
3-14
10.5
47.0
14-31
21.7
15 hour*
6
12
6-12
9.1
5-13
8.4
46.9
12-24
17.9
0-1 day
4
7
5-11
7.2
7-10
8.2
53.7
12-18
15.4
16|-24 hour;
rotated
9
14
9 15
11.2
6-14
9.3
45.5
18-26
20.5
1 day
10
13
6-10
9.1
4-9
6.7
43.3
11-19
15.8
24 hour*
5
8
6-16
9.5
6-11
7.7
44.0
12-24
17.5
2 days*
11
17
4-12
6.6
3-9
5.6
45.7
7-18
11.9
3 days*
7
12
3-10
7.0
3-6
4.8
41.9
7-16
11.9
4 days*
6
11
4-9
5.5
3-6
4.4
43.6
7-15
9.9
5 days*
5
10
4-13
7.4
2-8
4.5
37.7
7-18
12.1
6 days*
5
9
5-10
6.4
2-5
3.7
34.5
7-14
10.0
7 days*
6
10
5-14
7.5
2-6
3.9
35.8
8-18
11.2
7 days
10
18
3-11
8.2
2-5
3.0
27.1
7-14
11.2
10 days*
7
13
5-10
8.0
2-6
3.4
29.8
9-16
11.4
2 weeks
6
10
4-9
6.5
2-6
3.9
35.1
8-12
10.4
2 weeks
10
20
5-16
7.4
2-6
3.8
35.4
8-13
11.2
3 weeks
10
18
3-18
5.4
2-5
2.6
32.4
5-12
8.0
4 weeks
10
18
3-9
7.0
0-5
2.7
30.3
5-12
9.7
5 weeks
5
8
3-8
4.9
2-5
2.9
36.6
5-11
7.8
6 weeks
5
8
2-8
5.3
1-5
2.2
31.5
5-13
7.5
f Totals do not include partial data.
* Data collected 1966; all other data collected 1960-1961.
SPERMATOCVSTS IN AEDES 27
in dense, violently spinning whorls. The long threadlike cells in fully differentiated
compartments are tightly wound into ovoid balls, when the sperm are inactive in
situ. The testes of pupae and especially of the adults exhibit a wide number of
variations of the pyriform shape. The anterior end generally tends to be recurved
and the posterior end is either cuplike or in the shape of a funnel. The testis may
be bent into a C-shape. The middle portion may be compressed like a waist.
Testes were removed from unmated adults from the time of their emergence
through the sixth week. The number of spermatocysts did not increase significantly
at the time of adult emergence or thereafter. In the adult, the spermatocyst lamellae
frequently did not prevent active sperm from being able to move from chamber to
chamber in the differentiated region, but the sperm never moved into the non-
differentiated zone. The number of cysts was found to range from five to 34, with
means of 7.5 to 22.5 for the period of study (Table III). Undifferentiated com-
partments ranged from two to 18 (means of 5.3 to 12.7) and partially to fully
matured cysts varied from none 2 to 18 (means of 2.2 to 10.5) (Table III).
Out of 50 individual comparisons, 8% of the adults examined had both testes
of essentially the same size, while the remainder had one testis distinctly smaller
(by a factor of 1.3-fold) than the other in the same individual. Both large and
small testes tended to have approximately the same number and character of
spermatocysts. Over the 6-week period of study, no significant change could be
detected in the length of the testes in the 44 cases available for comparison. The
length of the large testes ranged from 340 to 737 microns, and the small testes
varied from 150 to 660 microns. There was no correlation between the size of a
testis and the direction of rotation of the terminalia among 10 individuals studied
in this regard.
During the first 24 hours of adult life, the male's terminalium rotates 180° and
the posteriormost compartment of each testis opens and a certain number of sperm
descend the spermatic duct (vas efferens plus deferens) and begin to fill the seminal
vesicles. Among many recently emerged adults, sperm were not present in the vas
efferens of one testis but were present in the duct of the other and subsequently
generally one duct contained more sperm than the other in the same individual. In
12 individual comparisons, there was no correlation between the presence or number
of sperm in the ducts and the size of the testes. While one testis clearly may be
the first to provide a portion of the initial supply of sperm to the seminal vesicles,
sperm from both testes are required to fill the vesicles.
During the first 24 hours of adult life, the mean number of non-differentiated
cysts per testis ranged from 7.2 to 12.7 (with an overall mean of 9.6) and the
differentiated cysts from 6.7 to 10.5 (with an overall mean of 9) (Table III). Al-
though sperm begin to fill the postgonadal system during this period, no significant
differences in the number or character of the cysts could be detected. With the
present material a deletion of two mature cysts per testis could not have been
detected.
Between the second and tenth days of adult life, when the sperm have already
filled the postgonadal system, the mean number of undifferentiated cysts per testis
ranged from 5.5 to 8.2 (with an overall mean of 7.1) and the matured cysts from
3 to 5.6 (with an overall mean of 4.2). These overall means differ from those of
2 The duct from this testis possessed numerous spermatozoa throughout its length.
28
JACK COLVARD JONES
0- to one-day-old adults by 2.5 fewer undifferentiated cysts and by 4.8 fewer differ-
entiated cysts per testis.
Jones and Wheeler (1965) reported 700 sperm in mature cysts, 740 sperm in
the spermatic ducts, and from 3700 to 6309 sperm in the seminal vesicles of unmated
Aedcs acfjvfiti. The filled postgonadal system would thus have from 4440 to 7049
spermatozoa. If these values are correct, spermatozoa within 6.3 to 10 matured
cysts would be needed to fill the postgonadal system. The mean deletion of 4.8
matured cysts per testis thus fits in with this requirement, and could account for a
supply of 6720 spermatozoa within the postgonadal system of A. acgypti. At least
2000 sperm could reach the postgonadal system during the first 24 hours after
emergence, and the remainder be delivered shortly thereafter.
TABLE IV
Spernwtocyst counts on testes of Acdes aegypti (L.) kept with approxi-
mately equal numbers of females for four to 7 weeks
Numbers of cysts/testis
Extent of differentiation
Age
No. dis-
sected
No. testes
Totalt
None
Partially to fully differentiated
Range
Mean
Range
Mean
%
Range
Mean
4 wks.
6
8
5-12
7.6
1-3
2.3
24.9
7-15
9.9
6 wks.
4
8
2-9
6.0
1-4
2.3
29.7
5-12
8.6
7 wks.
5
9
2-8
5.5
1-5
2.4
33. 0
4-12
7.4
f Totals do not include partial data.
During the six-week period of study, the total number of spermatocysts in
unmated males gradually declined from 22.5 to 7.5 (by a factor of 3) ; the un-
differentiated cysts decreased from 12.7 to 4.9 (by a factor of 2.6) ; and the differ-
entiated cysts were reduced from 10.5 to 2.2 (by a factor of 4.8) (Table III).
The mean percentage of compartments with differentiated sperm decreased from
56.69r< to a minimum of 27.1 9^, with an overall mean of 41.1 /f differentiated cysts
for the entire 6-week period of study. Viewed as a whole, there is a distinct and
significant trend for both undifferentiated and differentiated cysts to decline with
age in unmated adults, that is, in the absence of any loss of sperm from the repro-
ductive system.
If each mature cyst produces 700 spermatozoa, then the following calculations
can be made from the data in Table III: (1) About 13,500 sperm are present in
both testes before any or very few of them descend to the postgonadal system in
the newly emerged adult. (2) After 24 hours, 2720 to 4120 sperm have left the
testes. (3) During the next 9 days, there are from 4200 to 7840 sperm in the
testes and from 5660 to 9300 in the postgonadal system. (4) If the number of
sperm in each mature cyst does not change with time, then the amount of sperm in
the testes should gradually decrease as the supply in the postgonadal system in-
creases in unmated males. Between the second and tenth days, 3640 sperm should
SPERMATOCYSTS IN AEDES
leave the testes, and, between the second and sixth weeks, 2380 sperm should leave
the testes.
It is possible, however, that once the postgonadal system is filled, relatively
few or no additional sperm would be added thereafter. If this were the case,
spermatocyst walls could break down, leaving the same numbers of sperm within
the differentiated region of the testes but with fewer spermatocysts detectable
therein. With the present data, it is not possible to decide which of these is the case.
As shown in Table IV, data from males which had been caged continuously with
approximately equal numbers of females for four, 6 and 7 weeks did not differ in
number or character of spermatocysts from the data obtained from unmated males
of the same ages (Table III).
TABLE V
The number and character of spermatocysts in Aedes aegypti ( L.) after multiple copulation
Sex ratio and
cohabitation
time
X inn her males
dissected and
age when
dissected
No.
testes
Numbers of cysts/testis
Extent of differentiation
Totalf
None
Partial-complete
Range
Mean
Range
Mean
%
Range
Mean
1:10/3 days
controls
10/8 days
6/8 days
16
11
6 14
5-14
8.9
8.8
2-4
1-7
3.0
3.7
26.0
30.2
9-17
6-20
11.9
12.2
1:20/1 day
controls
1:20/2 days
controls
1:20/4 days
controls
11/3-4 days
9/3-5 days
17
16
3-13
4-17
8.7
9.3
2-10
2-7
4.8
4.2
36.9
31.6
10-17
6-18
13.6
13.3
3/7 days
6/7 days
6
10
7-10
6-13
9.0
8.8
2-7
2-5
5.5
3.1
38.0
26.1
12-17
9-16
14.8
11.9
3/14 davs
6/14 days
6
12
6-11
4-17
8.2
8.1
2-7
2-6
3.8
3.3
31.4
30.5
10-14
8-21
12.3
11.4
t Totals do not include partial data.
Since it had been found that the reproductive systems of males kept in the
presence of equal numbers of females for as long as 7 weeks could not be dis-
tinguished from those of unmated controls (Table IV), a series of cages were set
up containing varying combinations of previously unmated adults : ( 1 ) one cage of
10 males with 100 previously unmated females, (2) four cages of one male with 20
females and two cages of five males with 100 females. The adults were allowed to
co-habit for one to four days and the males were dissected. Frequent matings were
observed but the sexual history of individual males was not determined.
As shown in Table V, when the sex ratio was 1:10 or 1:20, the number and
character of the spermatocysts were basically the same as those of the unmated
controls. However, in 60% of the males the spermatic ducts and seminal vesicles
contained very few spermatozoa and these individuals generally had noticeably re-
duced accessory gland secretion, particularly when the sex ratio was 1:20. Pre-
sumably, those males with a reduced supply of sperm and accessory gland material
30
JACK COLVARD JONES
mated with more females than those whose supplies were not strikingly reduced.
Thus, depletion of sperm from the postgonadal system and of accessory gland
material did not appear to affect the general character of the testes.
This finding fits with the data of Jones and Wheeler (1965) which showed that,
after males had heen force-mated repeatedly, the posterior chamher of their testes
still had many spermatozoa (mean of 741). Together these data show that after
repeated matings only the sperm in the postgonadal system are used up and that
testicular sperm are not drawn down to replenish the supply as it is being removed.
To explore this problem further, three cages were set up, each with five males
to 30 females ; and, after co-habiting for 24 hours, the males were isolated for one,
two or three days before being dissected. After one day, three out of five males
had shrunken seminal vesicles with very few spermatozoa, and the ducts leading to
TABLE VI
The number and character of spermatocysts in Aedes aegypti (L.)
after being isolated following multiple copulation
Numbers of cysts/testis
Extent of differentiation
Days isolated
and sex ratio
No. testes
Totait
None
Partial-complete
Range
Mean
Range
Mean
%
Range
Mean
1 day 1:6
3
7-11
8.7
1-4
2.3
21.6
10-12
11.0
2 1:6
9
6-22
12.2
1-4
2.4
22.5
8-23
14.4
3 1:6
6
5-9
6.8
3-5
3.7
36.3
9-13
10.7
2 1:20
8
7-11
8.7
2-6
3.5
28.5
9-15
12.3
controls
1
4-17
9.1
1-6
3.5
26.8
6-18
12.2
5 1:20
13
5-11
7.4
2-6
3.9
35.7
8-16
11.2
controls
13
5-10
8.0
2-6
3.4
29.8
8-16
11.4
t Totals do not include partial data.
them had very few if any sperm ; the other two males had an obviously reduced
supply of vesicle sperm. Two days after isolation, three out of five males had
shrunken vesicles with very few spermatozoa, but two males had replenished the
sperm within the seminal vesicles. After three days, four out of five males had
replenished the sperm in their vesicles, and their accessory glands were filled with
secretion. Only one male still had very few sperm within his vesicles. Essentially
the same results were obtained with 5 males to 100 females (5 cages; co-habitation
time one, two, and four days).
As shown in Table VI, when the sex ratio was 1 :6, there does not appear to be
much difference between the number of differentiated cysts in testes of depleted
males and those of males which have largely replenished their sperm supply. If
4440 to 7049 sperm were removed from the testes to replenish their postgonadal
supply, then 6 to 10 matured cysts should have been needed and this would require
more sperm than would have been present within the 4.8 matured cysts which were
SPERMATOCYSTS IN AEDES 31
available in both testes. Three days after being isolated from females, however, the
males, after mostly replenishing their sperm supply, had more mature cysts than
the depleted males. Since there was an increase of 3.5 undifferentiated cysts per
testis two days after males were isolated from females, and since there was a
decrease of 5.4 undifferentiated cysts after the sperm supply had been replenished
on the third day, it seems likely that the 4.3 matured cysts were indeed all used up
and then replaced by maturation of new cysts derived from the undifferentiated zone
of the testes. Thus, it can be calculated that the 4.8 already matured cysts from
both testes would contribute only 3360 sperm to the postgonadal system, and that
the 5.4 extra undifferentiated cysts in each testis would produce a total of 10.8
matured cysts for both testes : of these 7.4 would replace and thus account for the
3.7 matured cysts seen in each testis of the replenished male, and the other 3.4 cysts
would contribute 2380 sperm to the postgonadal system to bring the total supply
there to 5740.
DISCUSSION
During the larval life of Aedes aegypti, the testes grow in size and the germ cells
greatly increase in numbers within them apparently near or around the time of each
larval ecdysis. In some old larvae (pharate pupae) spermatids may just begin to
differentiate in the posteriormost compartment of the testes. The general shape of
the gonads does not depend upon the presence of germ cells, as evidenced by those
males with agametic testes (Jones, 1961). Although agametic testes may still
possess a number of compartments, mostly or entirely at the anterior end, they
generally possess far fewer than do normal testes. Agametic testes are always
smaller than normal, thus showing that the growing number of germ cells leads to
a general increase in size of the testes (Jones, 1961 ).
The present observations indicate that differentiation of spermatids may begin
shortly before pupation occurs (that is, in pharate pupae) or they may not begin
for 6 to 21 hours. It is of considerable interest that differentiation can begin in
one testis without necessarily simultaneously beginning in the other. The entire
process of differentiation of spermatids into spermatozoa within a single cyst may
be completed within a five-hour period. Maturation always begins in the posterior-
most cyst. As many as 6 cysts may be maturing at the same time within a testis.
Matured cysts were found in all pupae after the first 24 hours. Maturation of
spermatocysts is preeminently a pupal event.
In old pupae just about to emerge as adults (that is, in pharate adults) 8.7 to
9.3 matured cysts were present in each testis. It can be calculated from this that
there are 6090 to 6510 spermatozoa within each testis at this time.
The present calculations indicate that while sperm begin to fill the postgonadal
system shortly after the adults emerge, this process is not completed until the second
day of adult life. It is estimated that about 10 matured cysts are required to fill
the postgonadal system with about 7000 spermatozoa and that both testes must con-
tribute to this supply.
In unmated males the number of spermatocysts declines in the absence of any
loss of sperm from the reproductive system. It is not clear whether this involves
an increase in the numbers of sperm within the postgonadal system as their num-
32 JACK COLVARD JONES
hers decrease in the testes or whether there are no changes in the numbers of
spermatozoa within different portions of the reproductive system but only a break-
down of spermatocyst lamellae.
After inseminating 6 females, the male quickly uses up all or nearly all of the
sperm in his postgonadal system but the numbers of sperm in the testes are not
immediately affected or drawn upon. When such males are isolated from females,
they replenish the sperm in their reproductive system in two to three days and the
numbers of spermatocysts do not clearly reflect this change. It is suggested that
replenishment cannot be achieved solely by the use of all the sperm within the
already matured cysts of the testes but requires the formation and maturation of
about 11 extra cysts.
SUMMARY
1. During the larval life of Aedes aegypti (L.), the testes greatly increase in
size and numbers of germinal cells, and the number of compartments (or spermato-
cysts) increases about five times, to a maximum of 24, usually several days before
pupation. Although the germ cells may begin the process of differentiation of
spermatids into spermatozoa within the terminal cysts of the testes just before
pupation occurs, fully differentiated spermatozoa were never observed in larvae.
Generally one testis is smaller than the other in fourth stage larvae, and this differ-
ence tends to persist throughout life.
2. Although a significant increase in the number of spermatocysts could not be
detected during pupal life, as many as 29 cysts were found among the testes ex-
amined during this period. Spermatids may transform into fully differentiated
spermatozoa within five hours. While the beginning of differentiation of the
spermatids may be delayed for as long as 21 hours after pupation, differentiated
sperm were always found after the first 24 hours. The number of differentiated
cysts increases during pupal life and it is calculated that 12,000 to 13,000 sperma-
tozoa are formed by both testes. Spermatozoa were never observed in the spermatic
ducts during the pupal stage.
3. Although a significant increase in the number of spermatocysts could not be
found during adult life, a maximum of 34 cysts were found among newly emerged
adults. The number of spermatocysts definitely declines with the age of unmated
adults. Spermatozoa begin to fill the postgonadal system during the day of adult
emergence. It is estimated that complete filling requires two days, and that the
sperm in about 10 cysts are required.
4. When the male uses up most or all of the sperm within his postgonadal system
after multiple matings, two to three days are required to replenish the sperm supply.
It is suggested that this must involve the formation and maturation of about 11
extra cysts, most of which are needed for replacement within the testes, the others
contributing to the supply of postgonadal sperm.
LITERATURE CITED
CHOLODKOWSKY, N., 1905. Uber den Bau des Dipterenhodens. Zcitschr. ivissenschaft. Zool.,
82: 389-410.
CHRISTOPHERS, S. R., 1960. Aedes aegypti. The Yellow Fever Mosquito. Its Life History,
Bionomics and Structure. Cambridge Univ. Press., 737 pp.
SPERMATOCYSTS IN AEDES
EPHRUSSI, B., AND G. W. BEADLE, 1936. A technique for transplantation for Drosophila. Amer.
Nat., 70: 218-225.
HURST, C. H., 1890. The post-embryonic development of a gnat (Culcx). Trans. Liverpool
Biol.Sci.,4: 170-191.
JONES, J. C., 1961. The internal reproductive anatomy of sterile male Aedes aegypti (Lin-
naeus). Mosq. Nezvs, 31: 118-119.
JONES, J. C., AND R. E. WHEELER, 1965. Studies on spermathecal filling in Aedes aegypti
(Linnaeus). I. Description. Biol. Bull., 129: 134-150.
KRAFSUR, E. S., 1964. The spermatogenesis of Aedes aegypti (L. ). M.S. thesis, Univ. of
Md., College Park, Md., 73 pp.
KULAGIN, N., 1907. Zur Naturgeschichte der Mucken. Zoo/. Anz.. 31: 865-881.
LOMEN, F., 1914. Der Hoden von Culex pipiens L. (Spermatogenese, Hodenwandungen und
Degenerationen). Jcnaische Zeitschr. Naturwissen., 52: 562-628.
NITROGENOUS EXCRETION IN THE TROPICAL SEA
URCHIN DIADEMA ANTILLARUM PHILIPPI
JOHN B. LEWIS
Bcllairs Research Institute of McGill University, St. James, Barbados, W . I.
Echinoids are considered to be predominantly ammonotelic in the excretion of
their nitrogenous metabolic end products (Prosser and Brown, 1961 ; Nicol, 1960).
Other excretory products have, however, been found in the perivisceral fluid and
tissue of echinoids or in the surrounding sea water of vessels in which specimens
have been enclosed. Delaunay (1931) found that considerable amounts of amino
acids were excreted by the urchins Paracentrotus and Strongylocentrotus, as well
as small amounts of urea, uric acid and other purines. Conheim (1901) found urea,
amino nitrogen, ammonia and purine bases in the coelomic fluid of urchins. Sanzo
(1907) found urea in several species and Myers (1920) found creatine, creatinine,
uric acid, urea and ammonia in Strongylocentrotus jranciscanus. Van der Heyde
(1923) found only uric acid in the coelomic fluid and intestine of Arbacia, while
Przylecki (1926) found only traces of uric acid in echinoids. Boolootian (1961)
has listed the amounts of nitrogenous excretory elements in the perivisceral fluid
found in a number of echinoids.
Excretion studies on echinoids have been mainly concerned with temperate and
cold-water species. The determination of the excretory products and rate of excre-
tion of Diadeina antillarum is thus of interest in terms of the physiology of a sea
urchin of widespread occurrence in the tropics.
The study was supported by a grant in aid of research from the National Re-
search Council of Canada.
METHODS
Experiments were performed on freshly collected specimens of 5 to 7 cm. test
diameter. Specimens were placed in covered glass jars filled with a measured
amount of filtered sea water, of between 2 and 4 liters. A control jar contained no
urchin. Tests were run for 4 hours at approximately sea temperature (26-28° C.).
At the completion of the trial the urchins were removed and the water filtered to
remove faecal matter and other solids. Analyses for excretory products in the
water were begun immediately.
Descending paper chromatograph techniques were used as qualitative tests for
excretory products in the "excretory water" as well as in the pericisceral fluid.
After electrolytic desalting (Baird & Tatlock desalting apparatus) 0.5 ml. of fluid
or water was evaporated on a paper pad. This pad was then affixed with a plastic
clip to the top of the paper chromatogram strip. Appropriate amounts (1-10 /xg.,
depending on sensitivity of method) of reagent quality samples of the substances
34
NITROGENOUS EXCRETION IN AN ECHINOID
35
being tested for were applied to paper pads and affixed to adjacent paper strips as
standards, according to the methods of Smith (1960).
Urea was tested for by the methods of Block et al. (1958), using phenol with
sodium hypochlorite and with acetone and dimethylaminobenzaldehyde (Smith,
1960). Uric acid and other purines were tested for by the method of Block et al.
(1958), using diphenylcarbazone on acidified mercuric acetate in ethanol and by
ultra-violet light (Smith, 1960). Creatine and creatinine were tested for by the
picric acid method (Block et al., 1958) and amino acids were identified with
ninhydrin.
Ammonia and total nonprotein nitrogen were determined quantitively by micro-
diffusion methods (Conway, 1962) . Kjeldahl treatments of sea water samples prior
to diffusion determinations were modified to use 200 ml. of water according to the
method of Barnes (1959). Amino acid quantitative determinations were a micro-
diffusion modification of the ninhydrin method of Sobel et al. (1945). Tests for
urea by diffusion \vere also run according to Conway (1962), using urease tablets
(British Drug Houses). Filtered sea water with and without urease were run as
controls and replicate samples were tested.
TABLE I
Mean hourly production in p.g. of excreted nitrogen by Diadema
Total
N.P.N.
NHsN
%NHsN
Amino N.
% Amino N.
Number of
animals
Duration exp.
in hours
162
99
61
42
26
12
4
245
157
64
71
29
12
4
RESULTS
In spite of repeated attempts to distinguish them in both perivisceral fluid and
in the "excretory water" of Diadema, urea, uric acid or other purines, creatine and
creatinine were not found by the methods used. Since these methods were sensitive
to a few micrograms of the detectable substances, it would appear that urea, purines
and creatine are only present in very minute amounts if indeed they are nitrogenous
excretory products. The amounts of urea, purines, and creatine and creatinine
noted by Boolootian (1961) for other echinoids were of the same order as the
amounts detectable by the chromatographic methods used here.
Amino acids, however, were found in sensible amounts, together with ammonia.
The results of analyses of excreted ammonia nitrogen, nonprotein nitrogen and
amino acid nitrogen in excretory water are shown in Table I.
The highest proportion of excreted nitrogen occurred as ammonia nitrogen in
two series of experiments. In the first series, urchins were freshly collected, while
in the second series the specimens had been previously fed for 12 hours on a diet
of fish meal.
The results of analyses of samples of perivisceral fluid showed no urea, purines,
creatine or creatinine. Substantial amounts of amino acids were detected on paper
chromatograms, however, and ammonia nitrogen was also present. The amounts
of ammonia nitrogen in the fluid were found to vary between 42 and 148 fig. per
36
JOHN B. LEWIS
100 ml. of fluid in freshly collected specimens. The mean content of 20 specimens
was 100 jug. of ammonia nitrogen per 100 ml. of fluid. The ammonia nitrogen con-
tent of the fluid of animals which had been fed on a protein diet for 24 hours was
markedly higher. Amounts of ammonia nitrogen in 16 specimens varied between
255 and 645 with a mean of 374 /xg. of nitrogen per 100 ml. of fluid.
Since no specific excretory organ is known for echinoids it is of interest to
compare the amounts of ammonia nitrogen found in the various tissues. The
amounts of ammonia nitrogen were obtained by grinding a known weight of tissue
in distilled water which was free of ammonia and subsequently determining the
ammonia by diffusion (Conway, 1962).
TABLE II
Alean ammonia nitrogen concentrations in ng./100 gm. in various tissues of Diadema
Organ
Ammonia N. /jg./100 gm.
No. specimens analyzed
Oesophagus
Caecum
159
195
12
12
Foregut
Hindgut
Rectum
317
400
155
12
12
12
Gonad
68
12
Muscle
61
12
Gills
75
12
The results of determinations of tissue ammonia contents are shown in Table TI.
The results show increasing concentrations of ammonia nitrogen in the gut
towards the rectum. Concentrations rose from 159 jug. in the oesophagus to 400 /^g.
per gm. of tissue in the hind gut. Concentrations in the gills, gonads and muscle
were comparatively low.
DISCUSSION
Like most other sea urchins, Diadema excretes the largest proportion of its
nitrogenous waste as ammonia but substantial amounts of amino acids were also
excreted. The amounts of ammonia nitrogen found in the perivisceral fluid are
comparable to those found in other urchins. Delaunay (1931) recorded a value of
240 /jg./100 ml. in Paracentrotus lividus and Myers (1920) found 80 /.ig./lOO ml. in
Strongylocentrotus franciscanus. Unlike other urchins whose excretory physiology
has been investigated, no urea, purines, creatine or creatinine have been found.
The ammonia content of the various tissues is of interest for it suggests in-
creased excretory activity towards the distal end of the hind gut. Sections of the
hind gut just preceding the rectum had more than twice the amount of tissue am-
monia than in the oesophagus and caecum. Progressive increase in tissue ammonia
along the gut in insects has been interpreted as denoting areas of excretory function
(Lennox, 1940; Staddon, 1955). However, the degree of differences in ammonia
content found in insects was far greater than occurred here in Diadema. The hind
gut was considered to have an excretory function in echinoids by Van der Heyde
(1923) and Delaunay (1931).
NITROGENOUS EXCRETION IN AN ECHINOID 37
SUMMARY
Diadema antillaruin is ammonotelic in its excretion of nitrogenous waste prod-
ucts. It excretes approximately 60% of its total nonprotein nitrogen as ammonia
and approximately 30% as amino acids. No urea, uric acid or other purine bases
were found to be excreted. A progressive increase in tissue ammonia content in
the intestine towards the rectum suggests that the hind gut has an excretory
function.
LITERATURE CITED
BARNES, H., 1959. Apparatus and Methods of Oceanography. Part One. Chemical. George
Allan and Unwin Lt, London.
BLOCK, R. J., E. L. DURRUM AND G. ZWEIG, 1958. A Manual of Paper Chromatography and
Paper Electrophoresis. Second Edition. Academic Press Inc. New York.
BOOLOOTIAN, R. A., 1961. Physical properties and chemical composition of perivisceral fluid :
Echinodermata. Reprinted from Biol. Handbook : Blood and other Body Fluids.
Federation of American Societies for Experimental Biology, pp. 339-344.
COHNHEIM, O., 1901. Versuche iiber Resorption, Verdauung und Stoffwechseln von Echinoder-
men. Zeitschr. Physiol. Chem., 33: 9-55.
CONWAY, E. J., 1962. Microdiffusion Analysis and Volumetric Error. Fifth Edition. Crosby
Lookwood & Son Ltd., London.
DELAUNAY, H., 1931. L'excretion azotee des invertebres. Biol. Rev. Cambridge Philos. Soc.,
6: 265-301.
LENNOX, F. G., 1940. Distribution of ammonia in larvae of Lucilia aiprina. Nature, 146: 269.
MYERS, R. G., 1920. A chemical study of the blood of several invertebrate animals. /. Biol.
Chem., 41: 119-143.
NICHOL, J. A. C, 1960. The Biology of Marine Animals. Sir Isaac and Sons, Ltd.
PROSSER, C. L., AND F. A. BROWN, JR., 1961. Comparative Animal Physiology. Second Edition,
Illustrated. W. B. Saunders Co., Philadelphia.
PRZYLECKI, ST. J., 1926. La repartition et la degredation de 1'acide urique chez les invertebres.
Arch. Internal. Physiol., 27: 157-202.
SANZO, L., 1907. Stickstaff-Stoffwechel bei marinen wirbellosen Tieren. Biol. Centralbl., 27:
479-491.
SMITH, I., 1960. Chromatographic and Electrophoretic Techniques. Volume I. Chroma-
tography. Second Edition. William Heinemann Medical Books Ltd. London.
SOBEL, A. E., A. HIRSCHMAN AND L. BESMAN, 1945. A convenient microtitration method for
the estimation of amino acids. /. Biol. Chem., 161 : 99-103.
STADDON, B. W., 1955. The excretion and storage of ammonia by the aquatic larva of Siales
lutaria (Neuropteri). /. Exp. Biol., 32: 84-94.
VAN DER HEYDE, H. C., 1923. Petites contributions a la physiologic comparee IV Sur
1'excretion chez les echinodermes. Arch. Neerland. de Physiol., 8: 151-160.
INDUCTION OF IMMUNOLOGICAL TOLERANCE BY INTRA-
COELOM1C GRAFTS IN THE 4-DAY CHICK EMBRYO
A. M. MUN, L. B. CRITTENDEN AND BARBARA JEAN CLARKE 1
Department of Zoology, University of Maine, Orono, Maine 04473, and U. S. Department of
Agriculture. Regional Poultry Research Laboratory, A.R.S., East Lansing, Michigan
When cells or antigens are injected into an embryo or a newborn animal a con-
dition of tolerance to the foreign stimulus may be induced. Although the mecha-
nism involved in this induction is not well known, we may entertain two possibili-
ties : ( 1 ) that the embryo may react with the cells or antigen, thereby revealing the
development of a certain level of immunological competence, and manifest either a
tolerant or an immune reaction depending on the dosage (Howard and Michie,
1962; Michie and Howard, 1962). However (2) the host embryo may also
nourish the proliferation of the foreign cells and permit the establishment of a
chimeric condition which is frequently obtained in tolerant animals (Billingham
ct al., 1952; Hasek and Hort, 1960; Stone ct al., 1965). Although the mechanism
is not clear, we may presume that the requirement for immunological competence
is not involved in such cases.
In the chick embryo, it is possible to examine these two alternatives as well as
elucidate the role of competence in the induction of tolerance by implanting cells or
antigens into the coelom of 4-day embryos, well before the onset of competence.
Immunologically competent cells, as measured by their ability to elicit a spleno-
megaly, are not detected until immediately after hatching (Solomon, 1961 ; Mun
et al., 1962). Solomon (1963; reported the sensitization of host lymphocytes as
measured by a depressed splenomegaly in the host during the eleventh to seventeenth
days of incubation. Ackerman and Knouff (1964) were able to identify certain
cell types which may be associated with the production of antibody in the thymus
of older 10- and 14-day chick embryos.
We ask first : Can tolerance be induced in the 4-day chick embryo ? If so, we
may next inquire : Would a greater degree of tolerance be obtained by the exchange
of cells or tissue from the same stage of development or by tissue from embryos
older than 10 days which may contain immunologically competent cells (Mun.
1965)?
MATERIALS AND METHODS
Two series of experiments were conducted, one at Orono, Maine, and the other
at East Lansing, Michigan. In the first series of experiments, the donor tissues
were obtained from a White Leghorn (WL) strain obtained from SPAFAS, Inc.,
Norwich, Conn., which has maintained for 10 or more years a closed flock with
continuous inbreeding, but not necessarily with brother-sister mating. The hosts
were derived from a cross between a Rhode Island Red male and a Barred Rock
1 Present address : Hope College, Holland, Michigan.
38
TOLERANCE INDUCTION IN THE CHICK
female (BR X RIR). .In the second series of experiments, donor tissues were
obtained from Line 7 embryos and hosts were derived from a cross between Line
151 and Line 6. These lines have been maintained as independent inbred lines
at the Regional Poultry Research Laboratory since 1939 (Crittenden et at., 1964).
The intracoelomic grafting technique has been previously described by Ham-
burger (1960) and Dossel (1954). The eggs were incubated for 86 to 96 hours
at 99° F. and 85% relative humidity. Embryos which had attained normal devel-
opmental stage 21 (Hamburger and Hamilton, 1951) with the allantois almost
in contact with the head were selected. A cut was made with a steel needle through
both the vitelline membrane and the somatopleure, in the small space between the
allantois and the head. The donor tissue, approximately 0.1 mm.3, was then
pushed through this opening and into the coelom toward the base of the allantois
with a curved and blunted glass needle. The operated eggs with their pointed ends
down were placed in the incubator and permitted to hatch.
TABLE I
Mean survival time of homografts in untreated hosts
Host
Donor
No. cases
MST* days
Standard
Range of survival
deviation
time (days)
BR X RIR
WL
53
8.1 ± 0.5
1.89
5-14
(SPAFAS)
151 X 6
Line 7
40
16.8 ± 2.0
6.29
6-31
* Plus and minus 95% confidence limits.
Two to 4 days after hatching, the chicks were divided into groups of four to six,
each group being made up of both operated, and unoperated or sham-operated
chicks. Each chick in the group then received a skin graft from a donor of the
same age and of the same strain as the previous embryonic donor. The donor
chick was discarded. The skin grafting technique developed by Cannon and Long-
mire (1952) was employed. The chicks were randomly numbered and not iden-
tified according to treatment. The grafts were read at one- or two-day intervals
for the first two to three weeks post-operation and later at weekly intervals. The
condition of the grafts in operated animals was compared with autografts as well as
grafts in unoperated controls, and the time at which the first signs of rejection
appeared was noted. Rejection was usually marked by a sudden darkening of
the graft, e.g., from pink or yellow to dark purple or brown, as well as a change
in the surface texture, e.g., from a soft, pliable condition to a smooth, hard surface
(Policy et al., 1960). After this stage, the hard scab which is formed eventually
drops off and the resultant bare area may persist for a variable period of time, until
the reappearance of host feathers oriented in the normal direction.
RESULTS
When skin grafts were exchanged between 2- to 4-day hatched chicks from the
same WL (SPAFAS) strain, eight out of sixteen grafts (50%) took successfully
and remained for more than 20 days. On the other hand, skin grafts from WL
(SPAFAS) donors on BR X RIR hosts were all rejected within 15 days (Table I).
40
MUM, CRITTENDEN AND CLARKE
TABLE II
Survival of skin grafts in BR X RIR hosts following intracoelomic grafts of various
tissues from WL (SPAFAS) embryos at different stages of development
No. of grafts
No. of grafts
Treatment
Total no. of cases
surviving 1 to
surviving more
20 days
than 20 days
4-day pharyngeal pouch, limb bud or lens
16
16
0
7- or 8-day spleen
25
25
0
12-day spleen
7
7
0
15-day lens
5
5
0
14- to 21-day spleen or thymus
80
71
9(11%)
None, or sham-operated
53
53
0
Fifteen experiments were conducted in which various tissues from WL (SPAFAS)
were implanted into the coelom of BR X RIR embryos. Pooled data from these
experiments show that limb buds, lens, or tissues from the region of the pharyngeal
pouches 3 and 4 of 4-day embryos, spleen from 7-, 8-, or 12-day embryos and lens
from 15-day embryos were not able to induce tolerance in BR X RIR hosts. How-
ever, tolerance was induced by spleen and thymus tissues from older 14- to 21-day
embryos in 11% of the cases (Table II). Because no striking differences in ability
of these two tissues to induce tolerance were observed in these preliminary studies,
the data were pooled.
Because of the small percentage of treated animals manifesting tolerance, these
experiments were repeated at East Lansing, Michigan, where embryos from highly
inbred lines were available. Twenty-two out of 29 (75%) skin grafts between
2- to 4-day Line 7 chicks took successfully and remained more than 50 days. Al-
though the homogeneity with respect to the histocompatibility loci in this particular
line is not yet complete (Crittenden et al., 1964), it is greater than that in the WL
(SPAFAS) strain (50%). Skin grafts from hatched chicks of Line 7 placed on
sham- or saline-operated or unoperated chicks of Lines 151x6 were all rejected
within 32 days (Table I).
Ten experiments were conducted in which donor tissues from Line 7 embryos
were implanted in the coelom of 4-day 151x6 embryos. Donor tissues were
TABLE III
Survival of skin grafts in 15 I X 6 hosts following intracoelomic grafts of various tissues
from Line 7 embryos at different stages of development
Graft survival time in days
Treatment
Total no.
of cases
1 to 32
33 to 50
More than
days
days
50 days
Intracoelomic grafts of
(1) 4-day embryonic limb, liver, pouches 3 and 4
64
52
4
8(12.5%)
(2) 9- to 18-day embryonic spleen, thymus, and
liver
44
16
10
18(41%)
(3) 1-day hatched chick spleen, thymus
16
10
4
2
Control or sham
40
40
0
0
Autograft 15 I X 6
39
1
0
38(97%)
TOLERANCE INDUCTION IN THE CHICK 41
obtained from limb buds, tbird and fourtb pharyngeal pouches, and liver of 4-day
embryos. Spleen, liver and thymus tissues were obtained from 9-day embryos to
1-day hatched chicks. Table III shows that a significantly greater proportion of
skin grafts lasting more than 50 days was obtained in chicks receiving intra-
coelomic grafts from older (9- to 18-day) embryos than from younger (4-day)
embryos (P < 0.005). Runts disease was observed in a few cases receiving
intracoelomic grafts from 1-day hatched chick tissues.
DISCUSSION
Tolerance can be induced in the chick embryo by joining their chorioallantoic
membranes on or about the 10th day of incubation (Hasek, Hraba and Hort, 1958)
or by cross-transfusion of blood on the 10th to 16th days of incubation (Terasaki,
Cannon and Longmire, 1958).
The present data show clearly that tolerance can also be induced in the 4-day
chick embryo by implanting various tissues into the coelom. If the initial steps
in the mechanism of tolerance induction involve the interaction of the foreign
antigen with immunologically competent cells, we may conclude that competent
cells are present in the chick embryo at this very early stage of development.
However, because immunologically competent cells as measured by other means
are not detected until at least after the 10th day of incubation, we may suggest that
the foreign donor tissues persist in the host environment and later react with com-
petent host cells as they appear (Mun et al., 1962). On the other hand, the
observation that tolerance induction was enhanced by older, more differentiated
tissue argues against the notion that tolerance is solely the result of mutual ex-
change, or persistence of donor tissue or cells in the host environment, and compels
us to consider the immediate impact of the foreign cells on the host environment.
There may be several explanations to account for the difference in the ability
to induce tolerance :
(1) The older tissues "took" better than grafts from younger donors. Yolpe
and Gebhardt (1965) observed in the frog that larger homografts, comprising two
complete lateral neural folds, survived and persisted indefinitely, while smaller single
lateral neural fold homografts were almost invariably eventually rejected. Thus,
the older donor grafts with a greater amount of antigen may demonstrate a larger
percentage of tolerant cases mainly because of their greater ability to survive in
the embryonic environment.
(2) On the other hand if we may assume that both older and younger grafts
take equally well, the greater ability of the older tissue to induce tolerance may
likewise be due to the amount of antigen. Howard and Michie (1962), Michie
and Howard (1962) and others have shown that a larger dose would result in
tolerance but a smaller dose of the same antigen would elicit sensitivity. However,
we found that intracoelomic grafting of larger pieces of tissues, almost two to three
times the usual size (0.1 to 0.3 mm.3), from either older or younger donors, did
not result in the induction of a greater degree of tolerance. The use of large pieces
of lens tissue from 15-day chick embryos also did not induce tolerance.
(3) The enhancement of tolerance by the grafts from older donors may also
be due to qualitative differences, as well as quantitative differences in antigen sup-
ply (Billingham and Silvers, 1962). Ebert (1951) discovered the appearance of
42 MUN, CRITTENDEN AND CLARKE
a spleen specific antigen on or about the 18th day of incubation. However, studies
on the development of the B blood antigens which are strongly associated with
histocompatibility in the chick reveal that they can be detected as early as the 7th
day of incubation (L. W. Johnson and W. E. Briles, personal communication).
(4) This leads us to consider another possibility: the impact of immunologi-
cally competent cells, which we may find in the 14- to 21 -day donor, on the host
environment. Jensen and Simonsen (1962) have observed in parabiosis experi-
ments in highly inbred mice, a facilitation of tolerance by the same antigenic stimu-
lus when the parabiont to become tolerant was exposed to a graft-vs.-host reaction
from its partner at the same time. The immunologically competent cells may
respond to the host antigen by proliferation and the release of greater amounts of
donor antigen, thus increasing their effective dosage very rapidly (Billingham and
Silvers, 1961, p. 127; see discussion by Burch and Burwell, 1965, p. 271). In the
chick, the immunologically competent donor cells may also act to stimulate pro-
liferation of the embryonic host spleen cells contributing to the observed organ
enlargement (Danchakoff, 1916; Biggs and Payne, 1961; DeLanney et al., 1962;
Mun and Burns, 1965). The role of these host-donor cell interactions in the
mechanism of tolerance induction remains to be explored.
We thank Mrs. Nancy McPhee Simpson for expert technical assistance. We
are grateful to Dr. James D. Ebert and Mr. Charles Kimmel for helpful sugges-
tions in the preparation of the manuscript. Wre also thank Dr. B. R. Burmester,
Director, U. S. Department of Agriculture, Regional Poultry Research Laboratory,
for making both invaluable inbred materials and research facilities available to us
at East Lansing, Michigan.
This investigation was supported by grant No. G-22431 from the National
Science Foundation to the University of Maine.
SUMMARY
1. In a series of experiments in which non-inbred material was used, 9 out of
80 skin grafts from a White Leghorn strain survived more than 20 days on Barred
Rock X Rhode Island Red hosts which had received intracoelomic grafts of spleen
and thymus from older (14- to 21-day) embryos of the same donor strain. Hosts
which had received intracoelomic grafts of pharyngeal pouches 3 and 4, limb buds
and lens from 4-day embryos, or spleens from 7-, 8-, or 12-day embryos or lens
tissue from 15-day embryos, rejected skin grafts from the same donor strain within
20 days.
2. When highly inbred material was used, tolerance was induced in Line 15 I
X 6 hosts by intracoelomic grafts of limb buds, liver, or pouches 3 and 4 from Line
7 embryos of 4 days. However, a significantly greater degree of tolerance was
induced by spleen and thymus tissues from older 9- to 18-day embryos of the same
donor strain. The possible impact of near-immunologically competent cells on host
cells in the induction of tolerance was considered.
LITERATURE CITED
ACKERMAN, G. A., AND R. A. KNOUFF, 1964. Lymphocyte formation in the thymus of the
embryonic chick. Anat. Record, 149: 191-216.
BIGGS, P. M., AND L. N. PAYNE, 1961. Pathological changes following the inoculation of chick
embryos with adult cells. I. Spleen cells. Immunology, 4: 24-37.
TOLERANCE INDUCTION IN THE CHICK 43
BILLINGHAM, R. E., AND W. K. SILVERS, 1961. Quantitative studies on the ability of cells of
different origins to induce tolerance of skin homografts and cause runt disease in neo-
natal mice. /. Exp. Zool, 146: 113-129.
BILLINGHAM, R. E., AND W. K. SILVERS, 1962. Some factors that determine the ability of
cellular inocula to induce tolerance of tissue homografts. /. Cell. Comp. Physiol.,
Suppl. 1.60: 183-200.
BILLINGHAM, R. E., G. H. LAMPKIN, P. B. MEDAWAR AND H. L. WILLIAMS, 1952. Tolerance
to homografts, twin diagnosis, and the freemartin condition in cattle. Heredity, 6:
201-212.
BURCH, P. R. J., AND R. G. BURWELL, 1965. Self and not-self: a clonal induction approach to
immunology. Quart. Rev. Biol., 40: 252-279.
CANNON, J. A., AND W. P. LONGMIRE, 1952. Studies of successful skin homografts in the
chicken. Ann. Siirg., 135: 60.
CRITTENDEN, L. B., L. W. JOHNSON AND W. OKAZAKI, 1964. Histocompatibility and erythro-
cyte antigen variability within highly inbred lines of White Leghorns. Transplantation,
2: 362-374.
DANCHAKOFF, V., 1916. Equivalence of different hematopoietic anlages (by method of stimu-
lation of their stem cells). I. Spleen. Amer. J. Anat., 20: 255-327.
DELANNEY, L. E., J. D. EBERT, C. M. COFFMAN AND A. M. MUN, 1962. On the chick spleen:
origin ; patterns of normal development and their experimental modification. Carnegie
hist. Washington, Contributions to Embryology, 37: 57-85.
DOSSEL, W., 1954. New method of intracoelomic grafting. Science, 120: 262-263.
EBERT, J. D., 1951. Ontogenetic change in the antigenicity of the chick spleen. Physiol.
Zool, 24: 20-41.
HAMBURGER, V., 1960. A Manual of Experimental Embryology. 221 pages. Univ. of Chicago
Press, Chicago, 111.
HAMBURGER, V., AND H. L. HAMILTON, 1951. A series of normal stages in the development of
the chick embryo. /. Morphol., 88: 49-92.
HASEK, M., AND J. HORT, 1960. Nonspecific tolerance of grafts and the dissociation of two
types of immunity. Nature, 186, 985.
HASEK, M., T. HRABA AND J. HORT, 1958. Embryonic parabiosis and related problems. Ann.
N. Y. Acad. Sci., 73: 570-574.
HOWARD, J. G., AND D. MICH IE, 1962. Induction of transplantation immunity in the newborn
mice. Transpl. Bull, 29: 91-96.
JENSEN, E., AND M. SIMONSEN, 1962. Induced tolerance after parabiosis : apparent facilitation
of tolerance by a simultaneous graft-trr^zw-host reaction. Ann. N. Y. Acad. Sci.,
99: 657-662.
MICHIE, D., AND J. C. HOWARD, 1962. Transplantation tolerance and immunological immaturity.
Ann. N. Y. Acad. Sci., 99: 670-679.
MUN, A. M., 1965. Ontogeny of tolerance induction in the chick embryo. Amer. Zool., 5: 252.
MUN, A. M., AND E. R. BURNS, 1965. Donor-host cell interaction in homologous splenomegaly
in the chick embryo. Biol. Bull, 127: 467-477.
MUN, A. M., P. TARDENT, J. ERRICO, J. D. EBERT, L. E. DELANNEY AND T. S. ARGYRIS, 1962.
An analysis of the initial reaction in the sequence resulting in homologous splenomegaly
in the chick embryo. Biol. Bull, 123: 366-387.
POLLEY, C. R., A. E. GROSSE AND J. V. CRAIG, 1960. A skin grafting technique for use in
genetic studies with chickens. Transplantation Bull., 7: 425-428.
SOLOMON, J. B., 1961. The onset and maturation of the graft versus host reaction in chickens.
/. Embryol. Exp. Morphol., 9: 355-369.
SOLOMON, J. B., 1963. Actively acquired transplantation immunity in the chick embryo.
Nature, 198: 1171-1173.
STONE, W. H., R. G. CRAGLE, E. W. SWANSON AND D. G. BROWN, 1965. Skin grafts: delayed
rejection between pairs of cattle twins showing erythrocyte chimerism. Science, 148:
1335-1336.
TERASAKI, P. I., J. A. CANNON AND W. P. LONGMIRE, JR., 1958. The specificity of tolerance
to homografts in the chickens. /. ImmunoL, 81 : 246-252.
VOLPE, E. P., AND B. M. GEBHARDT, 1965. Effect of dosage on the survival of embryonic homo-
transplants in the Leopard Frog, Rana pipiens. J. Exp. Zool., 160: 11-28.
PERSISTENT, VERTICAL-MIGRATION RHYTHMS IN BENTHIC
MICROFLORA. VI. THE TIDAL AND DIURNAL NATURE OF
THE RHYTHM IN THE DIATOM HANTZSCHIA VIRGATA 1
JOHN D. PALMER AND FRANK E. ROUND
Dept. of Biology, Nezv York University, New York, N. Y. 10453, Marine Biological Laboratory,
Woods Hole, Massachusetts 02543, and Dept. of Botany, University of Bristol, Bristol, England
During ebb tide in certain intertidal mud and sand-flats, irregular areas of the
exposed substratum become green or golden brown in color. With the return of
the flooding tide — or often just prior to its return — the color fades and disappears.
Microscopic examination of these sediments reveals that the color may be due to
a superficial accumulation of enormous numbers of protozoans, small metazoans,
or more commonly, single-celled algae. These organisms dwell in the sediments
during tidal inundation and move up onto the surface sands during tidal exposure
—a behavior pattern called a vertical-migration rhythm. Dinoflagellates (Herd-
man, 1924), euglenoids (Bracher, 1919; Palmer and Round, 1965), a chrysomonad
(Faure-Fremiet, 1950), several species of diatoms (Fauvel and Bohn, 1907; Aleem,
1950; Callame and Debyser, 1954; Round and Palmer, 1966), and a zooxanthellae-
containing planarian (Gamble and Keeble, 1903) are all known to undergo these
tide-associated rhythms in vertical migration. At times these organisms are present
in such great numbers that one investigator (Herdman, 1924, p. 59) observed that
"... the diatoms were so abundant on the surface that their photosynthetic activ-
ity was distinctly audible as a gentle sizzling . . . and the sand was frothy with
bubbles of gas, presumably oxygen given off by them."
The rhythmic behavior of a few of these organisms has been studied in the
laboratory and found to persist in natural day-night conditions but in the absence
of the tide (Fauvel and Bohn, 1907; Bracher, 1919; Herdman, 1924; Faure-
Fremiet, 1950, 1951), and in constant conditions, i.e., constant temperature, con-
tinuous illumination of a constant intensity, and no tides (Palmer and Round,
1965; Round and Palmer, 1966). Our studies (loc. cit.) have revealed an inter-
esting and unexpected aspect of vertical-migration rhythms, namely, that in con-
stant conditions the rhythms of two species of Euglcna and eight species of dia-
toms are diurnal, rather than tidal, i.e, the 24.8-hour period of the rhythm — as
displayed in nature — is not expressed in the laboratory ; instead, the persistent
rhythm has a 24-hour period. These data, combined with certain field studies by
other investigators (e.g., Perkins, 1960), suggest that possibly all overt tidal
vertical-migration rhythms might actually represent underlying 24-hour rhythms
which are entrained and thus transformed by the tides in nature. Alone among
the modern studies in contradiction of this supposition, is the work of Faure-
Fremiet (1951) on the vertical-migration rhythm of the diatom Hantzschia am-
1 This work was supported by National Science Foundation grants GB-5045 to JDP, and
GB-4509 to the Marine Biological Laboratory.
44
A MIGRATION RHYTHM IN HANTZSCHIA 45
phioxys. He reports (p. 173) that when Hantzschia-bear'mg sand samples were
returned to the laboratory and "... exposed to diffuse light from the window, on
succeeding days (the longest period of observation being six days), the [Hantz-
schia] re-appeared on the surface of the sand at the same time as low tide in their
natural habitat." He used only the color change of the sand as an indication of
whether or not the cells were on the surface and stresses the lack of precision of
this type of observation. Because his work stands as an exception to our original
hypothesis, it has stimulated us to re-examine the vertical migratory behavior of
this organism in greater detail and using quantitative methods. Both field and
laboratory studies were carried out.
This diatom inhabits the intertidal sand-flats of Barnstable Harbor, Cape Cod,
Mass., and was previously identified in the paper of Faure-Fremiet (1951) as
H. amphio.rys. We have compared the diatom with collections held at the Phila-
delphia Academy of Sciences and the British Museum and find that it is H. viryata
var. intermedia (Grun.).2 During the summer months it tends to be the dominant
species of an algal community containing the diatoms Amphora, Navicula, Amphi-
prora, Plcurosigma, and Nitsschia; the dinoflagellate Amphidinium ; the euglenoids,
Euglena and Trachelomonas ; and the cyanophyceans Chroococcus, Merismopedia
and Oscillatoria. Preliminary studies indicate that all these subdominants also
undergo vertical-migration rhythms in the field and the laboratory.
METHOD
In order to obtain quantitative estimates of cell concentrations on the surface
sediments at any one time, a method previously described in the literature (Palmer
and Round, 1965) wras employed. In brief, just as the ebbing tide uncovered the
sampling station, numerous small pieces (9 mm.2) of ordinary microscope-lens-
cleaning tissue were placed on the sediment surface. The Hantzschia, in their
migratory ascent to daylight, moved up through the sediments and into the inter-
stices of the paper. Tissues were then periodically removed from the sediment
during tidal exposure, the diatoms washed out in a drop of water on a microscope
slide, and their numbers counted. Replicate samples were taken and averaged.
In order to collect cells for study in the laboratory, 10-mm. lengths of glass
tubing (35 mm. in diameter) were inserted into the sediment and removed with
a core of Hantsschia-bear'mg sand within them. These cores, still retained within
the glass rings, were placed in small Petri dishes and returned to the Marine Bio-
logical Laboratory at Woods Hole, Mass., where water was added to the moat-like
space between the outside of the glass ring and the inner wall of the Petri dish.
The samples were kept in Precision Scientific Incubators at a constant temperature
of 18° C. and overhead illumination of 110 foot-candles from Westinghouse, 15-watt,
cool-white fluorescent tubes. The cells were maintained in alternating light-dark
photoperiods (symbolized at L:D) with the light on between 0530 and 2000 hours
(the approximate time between sunrise and sunset), or in continuous illumination
(L:L). It should be pointed out, however, that inherent in vertical-migration
rhythms is a periodic sojourn beneath the sand surface, placing the organisms into
- We wish to thank Dr. R. Patrick and Mr. R. Ross for their help and the loan of type
material.
46
JOHN D. PALMER AND FRANK E. ROUND
semi-darkness. This obviously tends to negate the desired effect of the overhead
experimental light regime. The laboratory populations were also sampled with
the lens-paper technique and the average of 3-6 samples used for each cell count.
Because this technique unavoidably reduces the size of the sample populations,
during long-term observations, sampling on some days was intentionally omitted.
RESULTS
Field observations
The sampling station on the Barnstable Harbor sand-flats is uncovered by the
tide for an average duration of 4.5 hours once every 12.4 hours. Field observa-
tions were made at intervals during the summers of 1965 and 1966 and were timed
so that the presence of the cells on the surface during morning, midday, and eve-
ning low tides could be described and compared.
It was found that when low tide straddles the time of sunrise the cells do not
appear on the surface until shortly after the time of sunrise. Once they begin to
appear, their numbers rapidly increase to a maximum value which then remains
relatively constant until about 30-60 minutes before the return of high tide, at
which time the cells begin to re-burrow back into the substratum. A representa-
tive curve is described in Figure 1.
400 •
300 :
CM
£
G
200 •
O
100 •:•
o b. .
'• •; .-: '. *K: ,•_.. •..: •
5 6 ~7 8
Time of Day
FIGURE 1. Field observations of the vertical-migration rhythm in Hantzschia. Wavy lines
(HT) represent the times of high tide. Straight line (LT) subtending and connecting consecu-
tive wavy lines indicates time of low tide. The time of sunrise is represented by the boundary
between stippled and unstippled areas. The shaded horizontal bars supported by dashed lines
signify periods when the sands were covered by opaque canisters. See text for further description.
A MIGRATION RHYTHM IN HANTZSCHIA 47
During midday exposures, the cells begin to appear on the surface 15-30 min-
utes after the tidal water recedes, increase in number to a fairly constant density
which is maintained for about three hours, and then commence to re-burrow about
30 to 60 minutes in advance of the incoming tide. (Re-burrowing in anticipation
of actual return of flood tide is a common feature of vertical-migration rhythms
[Callame and Debyser, 1954; Palmer and Round, 1965 ; Round and Palmer, 1966].)
By the time of inundation only the remnants of the densest patches remain on the
surface and these cells re-burrow within 15 minutes after being submerged. A
small fraction of the cells is often washed away. Measurements of the sediment
water content during a tidal exposure showed that on a bright, windy day, the
water content drops as much as 14% below the value obtained just after exposure.
Up to one hour before the actual reflooding of the area — and simultaneous with
the time the cells begin to re-burrow — the interstitial water content of the sand
begins to increase steadily. This may well be the stimulus which initiates re-
burrowing in apparent anticipation of the return of high water. As will be shown
later, however, re-burrowing is under the control of a biological clock and the cells
need no external stimulus if this kind to re-burrow.
As late afternoon low tides approach the time of sunset, the diatoms do not
remain on the surface for the duration of low water, but instead re-burrow slightly
before sunset. When the sand flats are first exposed at 1630 hours, or later, the
cells never appear on the surface.
The diatoms never appear on the surface at night and they can be kept from
emerging on the surface during daylight by artificially darkening the sediments with
opaque canisters. Similarly, cells already on the surface can be made to re-burrow
by artificial darkening. This is seen in Figure 1, where the cells were artificially
darkened between 0700 and 0730 and again between 0845 and 0900. After the
first darkening the cells returned to the surface; after the second, they did not,
but at this time the rest of the population was also in the process of re-burrowing.
The upper curve in this figure represents untreated cells in an adjacent patch and
thus acts as a control for the darkening experiments.
Laboratory studies
The rhythmic behavior of Hantsschia was first studied in constant light and
temperature. Under these conditions the rhythm was found to persist for as long-
as eleven days. A representative experiment is seen in Figure 2. While night-
time values were not determined for this particular set of data, numerous other all-
night observations have adequately demonstrated that the cells never appear on the
surface at night, even when samples are maintained in L:L. The approximate
times of low tide in nature are indicated for each day, and show that the cells in
the laboratory appear on the surface in approximate synchrony with those in
Barnstable Harbor — strongly suggesting that the rhythm is actually tidal. How-
ever, the possibility exists that the rhythm may actually be one with a fundamental
period of 24 hours, which — as is common with most persistent rhythms — has become
circadian in constant conditions, and by chance, has a period of 24.8 hours. To
test this possibility, samples were placed in L:D, a condition which restricts solar-
day rhythms to their fundamental 24-hour period. Figure 2 shows that the cells
continue to appear on the surface later each day and again in approximate synchrony
48
JOHN D. PALMER AND FRANK E. ROUND
L-.L
100
•V^Jvil
x r^
~~T.
A
i
0
100
....*!
j-
Tf*.'*'.*
...
A
2
0
too
.
.*
T*.rr!
-.-.
••i
3
/*
4
1 0
S 100
"5
; °
s
o
•/
'.r.r—
^-^*
-1
L
..j
5
6
o
1
« too
1
i.
r-
^\
7
g o
too
#-- -*
1 \
r— •
8
A
c
o
to
•- •
]
^
10
-
II
J
\
j
..r
6 12 «
Time of Doy
•oof: ••?;•;;:>
A
/V
y
\
A
6 12 6
Time of Doy
FIGURE 2.
A MIGRATION RHYTHM IN HANTZSCHIA 49
with the tidal exposure in nature. The persistent rhythm is indeed a lunar-day
rhythm.
The 12.4-hour interval between successive low tides is such that in the summer,
when an afternoon low tide approaches the time of sunset, the following morning
low tide begins to overlap with sunrise. Therefore, twice each month, there are a
few days when the sand flats are exposed to sunlight twice a day : once in the
morning and again in the afternoon. As the afternoon exposure approaches dusk,
the diatoms abandon this phase of the rhythm and now appear on the surface
during the early morning low tide. This rather drastic change in phase was
studied in the laboratory.
Cells were collected during a mid-afternoon low tide and placed in L : D in the
incubators. Sampling began the next day and, as seen in Figure 3, the cells came
up strongly in the late afternoon. On the second day the afternoon peak virtually
collapsed, and by the third day the cells now appeared in the morning. The same
observations were carried out in L:L and, quite remarkably, the same change in
phase occurred. Both the observations in L:D and L:L have been repeated sev-
eral times with the same results.
DISCUSSION
Clock control of the rhythm
Persistent, tidal rhythms have been previously described for a variety of physio-
logical functions, e.g., oxygen consumption in crabs (Brown et al., 1954) and snails
(Sandeen et al., 1954) ; spontaneous locomotor activity in crabs (Bennett et al.,
1957; Naylor, 1958; Palmer, 1966), amphipods (Enright, 1962; Morgan, 1965),
and fish (Gibson, 1965), and in filtration rate in mussels (Rao, 1954). These
rhythms may be described as tidal rhythms (i.e., rhythms with periods of 12.4
hours), or better, as binwdal (or biphasic} lunar-day rhythms with periods of 24.8
hours. When studied in relation to a 24-hour-day scale, the dual peaks of the
lunar-day rhythm advance at a rate of 50 minutes/cycle across the solar day.
The overt lunar-day rhythm in Hantsschia differs considerably from the above
rhythms in two major ways. First, the rhythm is unimodal, i.e., the cells appear
on the surface only once every 24.8 hours. Secondly, the single maximum scans
across the hours corresponding to daylight at a tidal rate of 50 minutes/day and
then, in a matter of just a few days, rephases back to the morning hours again.
Any model derived to explain the rhythm in Hantsschia must take into account
these two unique properties of the rhythm.
The curves obtained in L:D, shown in Figure 3, indicate that when the supra-
surface phase of the rhythm reaches the dark portion of the imposed photoperiod
the rhythm rephases to the morning hours, suggesting that the times of "light off"
FIGURE 2. Persistence of the vertical-migration rhythm in constant light (L:L) and in
alternating light-dark period (L:D). In both conditions the rhythm displays a period of about
24.8 hours. Consecutive days run from top to bottom. Stippling indicates dark periods. X =
time of collection of samples. State of tide on day of collection symbolized as in Figure 1. De-
pressions in dotted lines represent times of low tide in nature on days when rhythm was studied
in the laboratory. For ease in comparison the data are expressed in percentages (the highest
cell count in each cycle was designated as 100 and all other values as percentages of this). In
no case was 100% less than 2.9 X 10s cells/cm.2.
50
JOHN D. PALMER AND FRANK E. ROUND
and "light on" act as guideposts for the extreme phase relationships of the rhythm.
By way of analogy, the scanning movements of the single peak across the day can
be likened to the movement of a typewriter carriage, which slowly and systemati-
cally— one letter at a time — moves across the instrument to the far carriage stop,
and is then rapidly swept back to the starting margin to begin another journey. The
carriage stops, which dictate the extent of movement of the carriage, could be
L'.D
Efe
asap
0) i*£
o ^
£• " ./.C.
K
r
Hf&$$
\
12
';Cj 200
i....
V. ,.
A
12
Time of Day
FIGURE 3. The phase change of the persistent vertical-migration rhythm in alternating
light-dark periods (L:D) and in constant light (L:L). Symbols the same as Figure 2. Ordi-
nate scale for all days given in upper right hand column.
A MIGRATION RHYTHM IN HANTZSCHIA
51
-18
FIGURE 4. Diagrammatic representation of the interaction of a 24.8-hour bimodal vertical-
migration rhythm ( here represented as a disk with opposing bulges ; each bulge signifying the
surface phase of the rhythm) and a 24-hour suppression-expression rhythm (represented as an
incomplete disk superimposed over the lunar-day rhythm). The shaded area of the disk is that
part of the 24-hour rhythm that suppresses the night-time phase of the migration rhythm, and
the open segment the part that allows the expression of the daytime supra-surface phase. Be-
cause the supra-surface phase of the lunar rhythm occurs 50 minutes later each day it even-
tually falls under the influence of the suppressive portion of the solar-day rhythm. As this
phase is inhibited the unexpressed early morning phase is expressed. The net result is an
apparent rephase of the migration rhythm.
likened to the times of "light on" and "light off" in the rephasing of the rhythm.
However, the analogy breaks down when the rhythm is re-examined in L:L as no
such obvious "stops" were then present, yet the same apparent rephase took place.
As a consequence, we prefer to adopt a working hypothesis based on the presence
of two interacting clock systems. One is a lunar-day clock which, of course,
measures periods of 24.8 hours. This clock controls a vertical-migration rhythm
characterized by tivo supra-surface phases, 12.4 hours apart. This rhythm is repre-
sented diagrammatically in Figure 4, by a rotating disk with diametrically opposed,
conspicuous bulges. Each of these lateral protrusions represents a supra-surface
phase of the rhythm and therefore each has a width equal to 4.5 hours. Coupled to
52 JOHN D. PALMER AND FRANK E. ROUND
tin- lunar-day clock is a solar-day clock which measures periods of 24 hours. This
horologue controls a rhythm which is characterized by two alternating phases : one
which suppresses tin- night-time supra-surface phase of the himodal vertical-migra-
tion rhythm, and a second phase which permits the expression of the migration
rhythm in the daytime. The action of this rhythm is represented in Figure 4, by
a partially shaded stationary disk superimposed over the disk representing the lunar-
day rhythm. The shaded section represents the suppressive role of this rhythm
and the open segment or "window" represents the portion in which the tidal rhythm
is not inhibited. The size of the "window" was determined by field observations
of the hours of daylight during which the cells appeared on the surface.
By means of such a dual mechanism, as the expressed phase of the lunar-day
rhythm occurs progressively later each day (50 minutes/day), it eventually coincides
with the suppressive phase of the solar-day clock and is inhibited. Concurrent with
this event, the opposite peak of the migration rhythm moves into the "window" and
is now expressed. The net result is an apparent rephase of the rhythm from after-
noon to morning hours.
The feasibility of such a hypothesis is strengthened by the studies of Xaylor
(1958), Barnwel'l (1963), Chandrashekaran (1965), and" Palmer (1966), which
all demonstrate that it is a very common feature for organisms which display per-
sistent lunar-day rhythms to have a solar-day clock system associated with, and
modifying, the lunar rhythms. Enright's conclusion (1963), that a single organ-
ismic process does not have simultaneous tidal and solar-day components, has now
been shown to be premature ; it was based on his interpretation of earlier work
(Bennett ci al., 1957; Naylor, 1958) and by his own work on an organism which
possessed only a tidal rhythm (Enright, 1962). Actually, in intertidal organisms,
the co-existence of solar-day and lunar-day components in a particular rhythmic
function is a commonly encountered pattern.
Role of lit/lit
The importance of light in the expression of the vertical-migration rhythm of
Hantsschia manifests itself in a variety of ways: (T) the cells never appear on the
surface during night-time low tides ; and when low tides straddle the time of light
and darkness, the cells appear on the surface only during the illuminated portion
of the tidal exposure, (2) cells may be prevented from appearing on the surface
by artificial darkening, and ( 3 ) cells already on the surface can lie made to re-
burrow by artificial darkening. It is therefore quite apparent that light is neces-
sary to bring the cells to the surface and to hold them there for some critical time.
However, light does not always have an attractive effect, for in L:L in the labora-
tory, the cells do not come to the surface during the times corresponding to night,
or daytime high tides. Reasoning deductively, this certainly indicates that the
Hantsschia must undergo a rhythmic change in responsiveness to light and this
rhythm must be of fundamental importance in their migratory behavior. While no
systematic studies have yet been made on the existence of a persistent tidal rhythm
in phototaxis in Hantzschia, field observations by Faure-Fremiet (1951) and
Palmer (1960) have demonstrated a sign reversal in the phototactic response of
this diatom during the supra-surface phase of its rhythm. The organisms were
A MIGRATION RHYTHM IX HAXTZSCHIA 53
found to respond positively to light during the initial and mid-portions of their
stay on the sediment surface and then become indifferent or negative to light just
before the return of the tide. A persistent rhythm in phototaxis is known for an-
other unicell. Emjlcna ( Pohl, 1948).
Other environmental factors also contribute to the migratory movements of
the diatom. Inundation by high tide water is of paramount importance in the
expression of the rhythm (Palmer, I960), and the fact that the diatoms re-
burrow when artificially darkened (burrowing being a specific directional move-
ment) indicates that geotactic orientation must also be important.
Adaptive nature of the rh\thm
It has been tacitly assumed by past investigators that a vertical-migration rhythm
represents a highly adaptive relationship with the environment. It is supposed that
these sand-dwelling organisms move out onto the surface in order to undergo
maximum photosynthetic activity during the daytime, and then re-burrow to avoid
being washed away by the returning tide (Ganapati ct a!., 1959). Certainly some
of the non-conformers who do not re-burrow before the flooding tide sweeps over
them are often seen to be washed away, thus supporting the latter half of the above
contention. However, Taylor and Palmer (1963) have described the photosyn-
thetic light-saturation curve for the benthic microflora community on Barnstable--
Hantzschia, of course, being a prominent member — and these results demonstrate
that sufficient light penetrates through the upper 1.5 mm. of sediment to enable the
cells to photosynthesize at above 90% of their maximum capacity. Full sunlight
is well above the optimum and actually inhibits photosynthesis somewhat. Quite
clearly, then, it is unnecessary for the cells to "risk" a journey onto the surface—
and the possibility of being washed away — in order to undergo efficient photosyn-
thesis. It may be that the response is just primarily a phototactic one, the adap-
tive significance of which is less obvious.
\Ye wish to thank Gary Tabor for technical assistance with the project.
SUMMARY
1. The diatom, Hantzscliia vinjata, appears on the surface sands of Barnstable
Harbor, Mass., during daytime low tides. Surface accumulations of this organism
reach such concentrations that the sand takes on a golden-brown color. As the
tide returns the cells re-burrow into the sand.
2. The cells can be prevented from emerging onto the surface sands at low tide
by artificially darkening the area with an opaque covering just as the tide recedes.
Cells already on the surface can be made to re-burrow by similarly placing them
in darkness.
3. The vertical-migration rhythm will persist in the laboratory in constant
illumination, constant temperature, and away from the influence of the tide for as
long as eleven days. During this time the cells remain in approximate synchrony
with the feral cells in nature.
54 JOHN D. PALMER AND FRANK E. ROUND
4. Jn nature, when the times of low tide approach sunset, the cells rephase their
rhythm to the early morning hours of daylight. Cells collected during late after-
noon low tides and returned to L:D or L:L in the lahoratory, undergo a similar
rephasing in an interval of just three days.
5. To explain the various unique properties of this rhythm, it is postulated
that the rhythm is a manifestation of an interacting dual-clock system: a lunar-day
clock which measures periods of 24.8 hours and is responsible for a himodal migra-
tion rhythm ; and a solar-day clock responsible for the suppression of the night-time
supra-surface phase of the migration rhythm.
LITERATURE CITED
ALEEM, A. A., 1950. The diatom community inhabiting the mud flats at Whitstable. New
Phytol, 9: 174-188.
BARNWELL, F. H., 1963. Observations on daily and tidal rhythms in some fiddler crabs from
equatorial Brazil. Bio!. Bull.. 125: 399-415.
BENNETT, M. F., J. SHRINER AND R. A. BROWN, 1957. Persistent tidal cycles of spontaneous
motor activity in the fiddler crab, Uca pugnax. Biol. Bull., 112: 267-275.
BRACHER, R., 1919. Observations on Euglcna dcscs. Ann. Bot.. 33: 93-108.
BROWN, F. A., JR., M. F. BENNETT AND H. M. WEBB, 1954. Persistent daily and tidal rhythms
of Os-consumption in fiddler crabs. /. Cell. Coinp. Physio!., 44: 477-506.
CALLAME, B., AND J. DEBVSER, 1954. Observations sur les mouvements des diatomees a la
surface des sediments marins de la zone intercotidale. J'ic Milieu. 5: 242-249.
CHANDRASHEKARAN, M. K., 1965. Persistent tidal and diurnal rhythms of locomotory activity
and oxygen consumption in Emerita asiatica. Zcitschr. vcrgl. Physio!., 50: 137-150.
ENRIGHT, J. T., 1962. The tidal rhythm of activity of a sand beach amphipod. Zcitschr. rcr</l.
Physiol.,46: 276-313.
ENRIGHT, J. T., 1963. Endogenous tidal and lunar rhythms. Proc. Int. Congr. Zool., 4: 355-
359.
FAURE-FREMIET, E., 1950. Rhythme de maree d'une Chroinnlia psainmophile. Bull. Biol.
France Bclgiquc, 84: 207-214.
FAURE-FREMIET, E., 1951. The tidal rhythm of the diatom Haiitcschia amphio.v\s. Bio!. Bull.,
100: 173-177.
FAUVEL, P., AND G. BOHN, 1907. Le rhythme des marees chez les diatomees littorales. C. R.
Seanc. Soc. Biol., 62: 121-123.
GAMBLE, F. W., AND F. KEEBLE, 1903. The bionomics of Conroluta roscoffensis, with special
reference to its green cells. Proc. Roy. Soc. London, Scr. B. 72: 93-98.
GANAPATI, P. N., M. V. LAKSHMANA RAO AND D. V. SUBBA RAO, 1959. Tidal rhythms of
some diatoms and dinoflagellates inhabiting the intertidal sands of Visakhapatnam
Beach. Curr. Sci., 11: 450-451.
GIBSON, R. N., 1965. Rhythmic activity in littoral fish. Nature. 207: 544-545.
HERDMAN, E. C., 1924. Notes on dinoflagellates and other organisms causing discoloration of
the sand at Port Erin. Proc. Trans. Liverpool Biol. Soc.. 38: 58-63.
MORGAN, E., 1965. The activity rhythm of the amphipod Corophiuin volutator (Pallas) and
its possible relationship to changes in hydrostatic pressure associated with the tides.
/. Anitn. Ecol, 34: 731-746.
NAYLOR, E., 1958. Tidal and diurnal rhythms of locomotory activity in Carciuus mamas (L).
/. Exp. Biol., 35: 602-610.
PALMER, J. D., 1960. The role of moisture and illumination on the expression of the rhythmic
behavior of the diatom, Hantsschia ainpJiioxys. Biol. Bull.. 119: 330.
PAI.MKR, J. D., 1966. Daily and tidal components in the persistent rhythmic activity of the crab,
Sesunini. Nature (in press).
PALMER, J. D., AND F. E. ROUND, 1965. Persistent, vertical migration rhythms in benthic
microflora. I. The effect of light and temperature on the rhythmic behavior of Euglcna
obtusa. J. Mar. Biol. Assoc., 45: 567-582.
A MIGRATION RHYTHM IN HANTZSCHIA 55
PERKINS, E. J., 1960. The diurnal rhythm of the littoral diatoms of the River Eden estuary,
Fife.' /. Ecol, 48: 725-728.
POHL, R., 1948. Tagesrhythmus im phototaktischen Verhalten der Euylcna gracilis. Zcitschr.
Nahirf., 3b: 367-374.
RAO, K. P., 1954. Tidal rhythmicity in the rate of water propulsion in Mytiltts, and its modi-
fiability by transportation. Biol. Bull., 106: 353-359.
ROUND, F. E., AND J. D. PALMER, 1966. Persistent, vertical-migration rhythms in benthic
microflora. II. Field and laboratory studies of diatoms from the banks of the River
Avon. /. Mar. Biol. Assoc., 46: 191-214.
SANDEEN, M. I., G. C. STEPHENS AND F. A. BROWN, JR., 1954. Persistent daily and tidal
rhythms of oxygen consumption in two species of marine snails. Ph\siol. Zoo/., 27:
350-356.
TAYLOR, W. R., AND J. D. PALMER, 1963. The relationship between light and photosynthesis
in intertidal benthic diatoms. Biol. Bull., 125: 395.
STUDIES ON DOMECIA ACANTHOPHORA, A COMMENSAL CRAB
FROM PUERTO RICO, WITH PARTICULAR REFERENCE TO
MODIFICATIONS OF THE CORAL HOST AND
FEEDING HABITS
WENDELL K. PATTON
Ohio U'cslcyan University, Dclait'arc, Ohio 43015
In the summer of 1965, a survey was made of the fauna of living portions of
Acropora colonies at La Parguera, on the south coast of Puerto Rico. Collections
and underwater observations were made in shallow depths (0-10 feet) at the west
end of Enrique Reef. This region has good coral growth but less wave action
than the southern, seaward face of the reef (Almy and Carrion-Torres, 1963).
Contrary to my expectations, the xanthid crab Domecia acanthophora (Desbonne
and Schramm ) was the only commensal found. Studies were made on the biology
of this animal and on its relationship with the host coral.
I am grateful to Dr. P. W. Glynn, Acting Director of the Institute of Marine
Biology of the University of Puerto Rico at Mayagiiez for reading portions of the
manuscript and for the assistance which he and the staff of the Institute offered
during my stay in Puerto Rico. Dr. John Garth, of the Allan Hancock Founda-
tion, kindly donated specimens of Doinccia hispida from the Galapagos Islands.
METHODS
The coral to be collected was covered as completely as possible with one or more
cloth bags while still in place, then broken off and returned to the laboratory for
study. Placing the coral inside bags was awkward but necessary since many of
the crabs would have escaped if the coral had simply been lifted from the water.
At the laboratory dock, the pieces of coral were weighed and carefully examined
while the cloth bags were rinsed in a bucket of water which was poured through a
fine mesh screen. The crabs found on each colony were collected and measured.
In addition, living crabs were studied both in nature and in the laboratory.
OBSERVATIONS AND CONCLUSIONS
Doinccia acanthophora is a small crab with a mottled brown and cream carapace.
Of 180 specimens examined, 162 were above 5.0 mm. in carapace breadth and
easily identifiable as to sex, 58 being males and 104 females. The largest male
had a carapace breadth of 15.0 mm. while the mean for males was 8.2 mm. For
females the comparable figures were 14.0 mm. and 8.4 mm. The smallest ovigerotis
female had a carapace breadth of 5.6 mm.
56
A CRAB COMMENSAL ON CORAL
57
host coral
Three species of Acropora are known from the Caribbean (Wells, 1956) and
all were found at Enrique Reef (Fig. 1). Acropora cerricornis (Lamarck) has
branches up to 25 nun. in diameter and grows in large loosely branching colonies
which would appear to offer the crabs very little shelter. About ten colonies were
examined carefully in the water and crabs were indeed found to be quite scarce,
with occasional individuals occurring at a fork or some other site which offered
a little protection.
FIGURE 1. A, Acropora prolijcra. B, Acropora cerricornis. C-E, Acropora palmata. C,
A colony with considerable peripheral branching. D, A colony with considerable algal-induced
vertical branching. E, A colony with little peripheral branching.
Acropora prolifera (Lamarck) is similar to the above species but has thinner,
more closely spaced branches and forms thickets offering considerable shelter. The
crabs seemed to have a patchy distribution, as certain regions of coral contained
four or five crabs in a space of 25 cm.3, while much larger and apparently equally
suitable regions had none at all.
Colonies of the third species, Acropora palmata (Lamarck), are much more
massive than those of the preceding two. and typically consist of a central trunk
with a number of flattened, horizontal sheets of coral spreading out laterally.
These colonies range up to 6 feet in height and are very abundant on the La Par-
guera Reefs. The peripheral portions of the colony are usually branched to vary-
ing degrees but as new outward growth occurs, the spaces between the older
58 WENDELL K. PATTON
branches are filled in, forming a solid, central plate of coral. The peripheral
branching was most delicate and extensive in small colonies in about eight feet of
water in the channel off the western edge of the reef and seemed least developed
in colonies on the seaward face of the reef. This observation is similar to that of
Almy and Carrion-Torres (1963) who found small finger-like peripheral branches
in colonies growing in the back-reefs. With regard to the central coral plates
which compose the bulk of the colony, many are quite smooth while others show
varying degrees of diagonal or vertical branching on the upper surface. These
branches are generally quite short. An interesting type of vertical branch has a
tuft of algae in the center (Fig. ID). This extends down to the base of the branch
and thus it seems likely that the plate was damaged in some way, allowing the
algae to settle, and that the coral has grown up around it. On one occasion a coral
plate was seen which contained round white spots where the coral tissue had been
removed. These spots were grouped in a manner similar to that often found for
the algae-tipped branches and were identified by Dr. Glynn as the work of the
polychaete Hcrmodicc carunculata (Pallas) (Marsden, 1962; Glynn, 1963). It
would seem that algae could settle easily on the exposed spots and that polychaete
predation may thus contribute to branch formation.
Acropora palmata contains relatively more specimens of Dornecia acanthophora
than do the preceding two coral species and was studied the most intensively. On
examination it is seen that the great majority of crabs are not merely sheltering
among natural features of the colony but instead are inhabiting structural deforma-
tions of the living coral, which, for lack of a better name, will be called "resting
places." The term gall should perhaps be reserved for more regular deformations
than those shown here. These resting places (Fig. 2) can be divided into three
general types : crevices, pits and spaces between vertical branches. I believe that
all of these types result from the response of the coral to the continuing presence
of a crab.
A crevice is the most inclusive category and is simply a space between a
branch and the adjacent coral. When a crab is removed from its crevice, it is seen
that the coral has grown away from and around the crab, forming a shelter for it.
In addition, the corallites in the region touched by the crab are either thickened and
rounded off or absent altogether. Crevices are most common in colonies where
much natural peripheral branching occurs and where they can be seen in all stages
of development. They can also be found under diagonal branches. I feel that a
crevice originates when a young crab settles in an available space and stays there
long enough for the coral to be modified by its presence. As the coral continues
to grow and the spaces between branches begin to be filled in, the two sides of the
crevice may grow so that the crab comes to occupy a slit extending down into the
colony. In some cases, the sides of the crevice will unite beyond the crab leaving
a space extending from one side of the colony to the other.
Another and common alternative is that the crab will be surrounded on all sides
but one by growing coral and the crevice then will be converted into a blind pit.
The coral inside a pit is, of course, dead, while the opening is surrounded by a
raised lip of corallite-free, living coral. These pits may be quite shallow or may
extend well down inside the coral skeleton, depending on the degree to which the
opening has remained in the region of active coral growth. Since Vaughan (1915)
A CRAB COMMENSAL ON CORAL
has reported that A. palmata may increase in diameter as much as 95 mm. a year,
it would seem that a well formed pit could develop in six months or less and that
a rudimentary crevice could develop in a matter of weeks.
Crabs are also found sheltering between certain vertical branches and these
spaces too are found to be modified. They are characterized by thickened corallites
and a slightly deformed growth of the branches involved.
It can thus be seen that the suitability of a colony for resting place formation
varies with the amount of diagonal, vertical and peripheral branching that it shows.
A further indication of the evident ease with which coral skeleton can be modi-
fied was provided by a colony of Acropora palmata in which the corallites were
thickened in an area brushed by the seemingly light touch of the expanded tentacles
FIGURE 2. Modifications induced in A. palmata by D. acanthophora. A, Early stage of
crevice formation in colony showing much peripheral branching. B, Older and better developed
crevice in the same colony. C, Crevice under a diagonal branch. D, Well developed crevice
at side of a colony. E, Crevice enclosed anteriorly. F, Pit. G, Pit. H, Modified space
between vertical branches.
60 WENDELL K. I'ATTON
of an adjacent tube worm. The well known fact that many branching corals,
including A. palmata, have different growth forms under different wave conditions
(Wood-Jones, 1907; Vaughan, 1919; Storr, 1964) also shows that the deposition
of coral skeleton can he much influenced by factors in the environment.
The above formation of resting places is not unique, as similar though more
regular modifications are known to be induced in their coral hosts by the various
gall-crabs of the family Hapalocarcinidae. The best known species, Hapalocar-
cinus inarsiipialis Stimpson has been well described by Potts (1915). Here the
male is free-living while the female forms galls in branching corals of the family
Pocilloporidae. The coral grows around the crab, eventually enclosing her except
for several small openings in the top of the gall. Potts noticed a malformation of
both corallites and polyps on the inside of the gall and regarded this and the
growth of the gall itself as the reaction of the coral to the crab's respiratory cur-
rents. Although respiratory currents may play a part in resting place formation
in Acropora palmata it seems more likely that the coral is responding to contact
with the crab itself. This is particularly true in the case of crevices and the modi-
fied spaces between branches where the coral is altered in the region actually
touched by the crab but not in the area in front of the crab against which the
exhalant respiratory current must press.
The remaining members of the Hapalocarcinidae form pits and crevices in
various massive corals (Fize and Serene, 1957). The only previously reported
case of decapod modifying Acropora is that of the uncommon Indo-Pacific gall
shrimp, Paratypton sicbcnrocki Balss.
The commensal
Doinccia acanthophora looks much like any free-living xanthid crab and as
evidenced by the individuals on A. ccri'lcornls and A. prolifcra is not dependent
on the existence of a resting place. This is in marked contrast to the gall-crabs
and gall-shrimp mentioned above which are very much modified structurally and
are seemingly unable to survive outside of their galls for any length of time. Fur-
thermore, D. acanthophora is not an obligate commensal of a particular group of
corals. Rathbun (1930) reported it "among sponges and branches of corals and
in holes of dead corals and stones" and mentioned the corals Mcandrina and Poritcs.
Rathbun (1921) found it on Acropora at Barbados. At La Parguera; Dr. P. W.
Glynn (personal communication) found this crab commonly on Acropora palinata
and in beds of Poritcs fitrcata in shallow water on reef flats.
The specimens of D. acantlwphora observed on Acropora in nature moved very
little. When disturbed, however, they could move very rapidly across the coral
colony. Those inhabiting A. pal mat a were very reluctant to leave their resting
places and generally would not do so until touched with forceps. In the laboratory
crabs showed a strong negative phototaxis and a low thigmokinesis. These traits
are of course shared by many benthic organisms. If there is competition among
crabs for resting places, some type of territorial behavior might be expected. This
was never observed.
Many colonies of Acropora showed no evidence of commensals and so were
not collected. Table I shows the results of several collections that were made.
A CRAB COMMENSAL ON CORAL
61
Note the fairly close correspondence which exists between the number of resting
places found on A. palmata and the number of crabs on the colony. This may be
due to predation of unprotected crabs by the many small fish which hover around
the coral or more likely to the tendency of the crab to keep moving until it is ade-
quately sheltered. The small excess of crabs over resting places may be due to
(a) crabs wandering over the colony, or (b) crabs inhabiting structurally unmodi-
fied shelters, or (c) two crabs inhabiting the same resting place. The first two
alternatives were noticed occasionally, the third only once. The vast majority of
resting places seen in nature were found to be inhabited.
Some differences were noticed in the crabs found on the two species of coral.
Of those shown in Table I, the largest taken from A. pro! if era was a female of
10.1 mm. carapace breadth while A. pal mat a contained 5 males and 13 females of
this size or larger. As can be seen from Table I the sex ratio is about equal in
crabs inhabiting A. pro! if era while on A. palmata there were twice as many females
as males.
TABLE I
Colonies and portions of colonies of Acropora collected from Enrique Reef
between July 20 and August 5, 1965
A. palmala
A
. prolifera
Colony number
1 2
3
4
5 6780
111 1 1
12
13 1 2
3
4567
8 Q
males
3
2
5
5 1
5
4
1 1 1
4
1 2
1 2
crabs ovigerous females
1 6
3
5
1 6 2
/
4
1 1 1
1 1
2 1
found non-ovigerous females
1 4
5
2 1 1
4
3
1 2
1
1 2
juveniles (under 5 mm.)
2
1 1 1
2
1
2
1
1
Total
2 13
5
17
1 14 2 3 2
0 IS
12
3
6 2
5
1251
4 3
Number of resting places found on
colony
1 1(1
4
14
1 11 1 3 2
(1 16
/
3
0 0
0
0000
(I (I
To investigate the mobility of D. acanthophora, the resting places on several
adjacent colonies of A. palmata were mapped and examined on 8 of the next 13
days. After each day's observation the crabs inhabiting the resting places were
destroyed. The results, shown in Table II, permit several observations to be made.
( 1 ) Crabs do not have to make their own resting places, but will readily inhabit
those they find vacant. Thus once a resting place is formed, it could be used by
many crabs during the life of the colony. (2) There is evidently considerable
movement of Dowecia on to the colony and if the originally vacated resting places
had been left undisturbed, they would probably all have been re-occupied in about
a week. (3) Much of the movement involved smaller individuals. It would have
been highly desirable to measure each crab, but unfortunately they were so agile or
else so well entrenched they could not be caught without destroying them. (4)
Some resting places are strongly preferred.
Feeding habits
Specimens of D. acantJwpJwra kept in the laboratory would hardly move unless
disturbed and refused to eat fragments of shrimp or fish. The main activities that
were observed were nicking movements of the first antennae and periodic extensions
and withdrawals of the second maxillipeds.
62
WENDELL K. PATTON
Most brachyuran crabs are carnivores or scavenging omnivores and their
stomachs contain fragments of more or less identifiable animal or plant material.
The stomachs of 14 D. acanthophora contained no animal organisms, no fragments
of muscle, chitin or other tissue and only a very occasional strand of algae. The
hulk of the material in the stomachs was a white, structureless mass containing
variable amounts of large bacteria. Small mineral grains and sponge spicules
were also commonly found. One crab contained many small yellowish cells about
10 microns in diameter which mav have been zooxanthellae.
TABLE II
Re -inhabitation of various resting places on adjacent areas of A. palamata.
crabs were destroyed following each day's observations
All
Day
1
2
3
4
7
8
9
11
14
Crevices a
P
—
—
—
—
—
—
b
P
—
L
M
M
—
M
S
M, S
c
P
—
—
—
—
—
S
—
—
d
P
P
—
M
—
—
—
—
—
e
P
—
—
—
S
—
—
—
—
Pits A
P
P
S
S
S
S
B
P
P
—
—
S
—
—
—
—
C
P
P
S
—
—
—
—
—
—
D
P
P
S
S
—
— •
S
—
—
Spaces between 1
—
—
—
—
—
—
—
—
—
vertical branches 2
P
P
—
—
—
—
—
—
—
3
P
—
P
—
M
—
S
—
—
4
—
—
—
—
S, M, L = small, medium or large sized crabs
P = crab present, size not noted
= no crab present
The mouthparts of two fairly typical non-commensal crabs, the shore crab,
Carcinns inacnas and a spider crab, Hyas coarctatns, are very similar even though
the crabs belong to different superfamilies (Borradaile, 1922; Hartnoll, 1963).
When the mouthparts of D. acanthophora (Fig. 3) are compared with those of the
above crabs, however, a number of marked differences can be seen. ( 1 ) The
dactyl of the third maxillipeds of Doinecia has a much longer tuft of terminal hairs.
(2) The armature of the dactyl of the second maxillipeds is very different. In
D. acanthophora the dactyl is considerably broadened and bears many stout setae
on its inner surface. These are arranged in rows with each row being composed
of setae of a different length. The longest ones are the most ventral, above which
are rows of progressively shorter and more dorsal setae. On either side of the
terminal portion of each seta there is a row of up to 12 lateral bristles. Each seta
ends in a peculiar tripartite paddle which turns upwards at a right angle. This
paddle is composed of the tip of the seta and a thickened bristle on either side.
The longer setae curve dorsally and have a greater number of lateral bristles than
A CRAB COMMKXSAL ON CORA I.
do the shorter ones. ( o ) The first maxillipeds and second maxillae have better
developed setae on their medial surfaces. (4) The mandibles are more weakly
calcified while the mandihnlar ]>al]> has a different shape and fewer setae.
The teeth of the gastric mill also differ from those found in typical brachynrans.
In ten species of spider crab, Hartnoll ( 1963 ) found that the lateral teeth of the
A-G
H
1 mm
.1 mm
FIGURE 3. A-G, Ventral view of right mouthparts of female Domccia acanthophora, 12.4
mm. carapace breadth. A, Third maxilliped. B, Second maxilliped. C, First maxilliped. D,
Second maxilla. E, First maxilla. F, Mandible. G, Paragnath. H, Dorsal view of spines on
distal end of the dactyl of the second maxilliped.
64 WENDELL K. PATTON
zygocarcliac ossicles had from 3-7 cusps and from 0-7 ridges. In Cardans they
have 5 fairly blunt cusps and <S ridges ( Potts, 1915). In />. acanthophora, how-
ex cr, the lateral teeth consist of 13-H> fairly slender cusps whose ])oints vary from
rounded to sharp and about 20 ridges.
In the light of the above observations it is probable that />. acanthophora re-
moves its food from the water with the second maxillipeds. It is not a typical
filter-feeder, however. The only possible filtering structure is the dactyl of the
second maxillipeds and this is of relatively small size and lacks the abundant pin-
nate setae usually found in the filter of filter-feeding crustaceans (Marshall and
Orr, 1960). The dactyls of ten crabs were carefully examined and although the
crabs had not been preserved until an hour after being caught, in nine cases the
setae contained amorphous material with inclusions of sand grains. In two cases,
the setae were largely covered with a sheet of this material while in the others
small pieces were entangled on some of the paddles. There were a very few small
particles on the setae which did not seem to be in a matrix of other material. This
material is very similar to that found in the stomach and is doubtless scraped off the
dactyl by the setae on the medial surfaces of the first maxillipeds and second and
first maxillae, and placed between the protruding mandibles.
A likely source of this amorphous material is the organic detritus drifting in
the water since descriptions of it agree closely with the stomach contents. Hunt
(1925) describes detritus from Plymouth as a pale brown flocculus containing liv-
ing micro-organisms, fragmented skeletal parts and sand grains. Riley (1963)
states that (p. 273) "organic aggregates in Long Island Sound commonly consist
of pale yellowish or brownish amorphous matrices with inclusions of bacteria, silt
particles and sometimes phytoplankton." Coral mucus is also a possible source
of the material eaten by I), acanthophora. It is shed abundantly by Acropora and
could contain sand grains and spicule fragments which settle on it. I doubt if it
is a major food, however. Crabs were never observed to eat mucus and do not
seem to move around the colony as they might if they were collecting it. One crab
was seen to remove a strand of mucus that had drifted into its mouthparts.
ZOOGEOGRAPHY
The genus Doiuccio occurs in tropical coral habitats around the world and
contains three species. The best known is D. hispida Eydoux and Souleyet. It
extends from the Red Sea across the Pacific to the west coast of America and is
apparently restricted to colonies of Pocillopora. The stomach contents of a speci-
men of D. hispida from the Galapagos were identical to that reported above for
D. acanthophora. The mouthparts of this specimen were very similar to those
of D. acanthophora and the two species doubtless have the same feeding habits.
Although Hapalocarcinus forms galls in Pocillopora, deformation of this coral by
D. hispida has not been reported.
The second species, D. ylabra Alcock, extends across the Indo-Pacific from
Madagascar to Tahiti and has been taken from Acropora (Garth, 1964; Patton,
1966). Resting place formation has not been reported but may occur on the
plate-forming species.
As to the third species, Guinot (1964) has shown that specimens of Domecia
from the tropical Atlantic do not represent I), hispida as previously supposed but
A CRAB COMMKXSAL ON CORAL 65
constitute a distinct, though very similar species, D. acantJwphora (Deshonne and
Schramm). Furthermore, the American specimens D. acanthophora forma acan-
thophora. which occur from South Carolina to Brazil, can he distinguished from
African specimens D. acanthophora forma africana Guinot. Little is known of the
habitat of the African form except that it is found among corals.
The genus Acropora is very well developed in the Indo-Pacific and has quite
a varied fauna of commensal decapods (Garth, 1964; Patton, 1966). The types
with fairly close branches have the most commensals but even species similar to
those of the Caribbean have more than one commensal species. The question then
arises as to why Doinccia acanthopJwra is the only decapod commensal with Puerto
Rican Acropora. The answer may well be an historical one. The present-day
hermatypic Atlantic corals are believed to result from a time when the Tethys Sea
connected the Atlantic with the Indian Ocean. This connection was permanently
broken in the Miocene. Later in the Tertiary the families Acroporidae and Pocillo-
poridae had an enormous development in the Indo-Pacific (Wells. 1956), producing
numerous species, many of which offer a great deal of shelter to commensals. It
seems likely that the development of the present extensive commensal faunas of
these two families would have followed or paralleled this expansion of their hosts.
If this was so, the Acropora fauna which evolved in the Indo-Pacific would have
had no opportunity to reach the Caribbean. There is no record of Acropora occur-
ring in the East Pacific (Durham and Allison, 1960) and the only possible tropical
connection between the Indo-Pacific and the Atlantic would have had to involve
the East Pacific and a Central American waterway.
Pocillopora, on the other hand, though absent in the Caribbean since the Miocene
(Durham and Allison, 1960), is common in the East Pacific and contains several
of the typical Indo-Pacific Pocillopora commensals including abundant Doinecia
hispida (Crane, 1947; Garth, 1948). The last Central American seaway was
closed either in the late Miocene (Durham and Allison, 1960) or the Pliocene
(Lloyd, 1963). Although the first record of Pocillopora in the East Pacific is
in the Pleistocene (Durham and Allison, 1960), it may have been there earlier
along with its commensals and prior to the closing of connections with the Atlantic.
If this was the case, then D. hispida could have crossed over and established itself
on Atlantic Acropora. The ability of Doinccia to feed on detritus could have
been sufficient reason for it being the only one of the Pocillopora commensals to
successfully cross into the Atlantic and transfer to a new host. Following the
separation of the two oceans, the Atlantic Doinecia would have evolved into a sepa-
rate species.
An alternative explanation for the presence of Doinecia in the Atlantic is that
the genus is older than the other commensal decapods and evolved before the
severing of the Tethys connection. This seems less likely since the great morpho-
logical similarity of D. liispida and D. acanthophora suggests that they have a
relatively recent common ancestry.
Sl'M MARY
1. The xanthid crab Doinccia acanthopJwra was collected from three species
of the coral Acropora at Enrique Reef, La Parguera, Puerto Rico.
66 WENDELL K. PATTON
2. The commensal was most commonly found on the flattened coral sheets of
Acropora paliiiata. Here the majority of crabs inhabit various types of structural
deformation of the coral which are called resting places. These are believed to
be formed by the growth of the coral around and to some extent away from a
resting crab.
3. Although undisturbed crabs remain quite motionless, ones which are dis-
turbed are capable of rapid movement over the colony. There is at least some
movement of crabs around the reef as vacated resting places will be re-occupied
by new crabs.
4. The mouthparts of D. acanthophora differ from those found in typical
crabs. In particular, the mandible is weakly calcined and the second maxilliped
possesses rows of peculiar paddle-tipped spines on the distal margin of the dactyl.
The most likely food for the animal seems to be organic detritus which it separates
from the surrounding water.
5. The genus Acropora harbors numerous commensal decapods in the Indo-
Pacific but apparently only one in the Caribbean.
LITERATURE CITED
ALMY, C. C, JR., AND C. CARRION-TORRES, 1963. Shallow-water stony corals of Puerto Rico.
Carihb. J. Sci.. 3: 133-162.
BORRADAILE, L. A., 1922. On the mouth-parts of the shore crab. /. Linn. Soc. London, 35:
115-142.
CRANE, J., 1947. Inter tidal brachygnathous crabs from the west coast of tropical America with
special reference to ecology. Zoologica, 32: 69-95.
DURHAM, J. W., AND E. C. ALLISON, 1960. The geologic history of Baja California and its
marine faunas. Systematic Zool., 9: 47-91.
FIZE, A., AND R. SERENE, 1957. Les Hapalocarcinides clu Viet-Nam. Arch. Mus. Hist. Nat.
Paris, (7) 5: 3-202.
GARTH, J. S., 1948. The Brachyura of the "Askoy" Expedition with remarks on carcino-
logical collecting in the Panama Bight. Bull. Aincr. Mus. Nat. Hist. 92 art. 1, pp.
1-66.
GARTH, J. S., 1964. The Crustacea Decapoda (Brachyura and Anomura) of Eniwetok Atoll,
Marshall Islands, with special reference to the obligate commensals of branching
corals. Micronesica, 1: 137-144.
GLYNN, P. W., 1963. Hcnnodicc carunculatu and Mithraculits sculptus, two hermatypic coral
predators. Assoc. Island Marine Lab. Caribb. 4th Meeting. Curacao; pp. 16-17.
GUINOT, D., 1964. Les trois especes du genre Dotnccia (Decapoda, Brachyura) : D. liispida
Eydoux & Souleyet, D. glabra Alcock, et D. acanthophora ( Desbonne & Schramm).
Crustaceana, 7 : 267-283.
HARTNOLL, R. G., 1963. The biology of Manx spider crabs. Proc. Zool. Soc. London, 141:
423-496.
HUNT, O. D., 1925. The food of the bottom fauna of the Plymouth fishing grounds. /. Mar.
Biol. Assoc., 13: 560-599.
LLOYD, J. J., 1963. Tectonic history of the South Central-American orogen. Aincr. Assoc.
Petrol Gcol. Men,., 2: 88-100.
MARSDEN, J. R., 1962. A coral-eating polychaete. Nature, 193: 598.
MARSHALL, S. M., AND A. P. ORR, 1960. Feeding and nutrition. In: T. H. Waterman (ed.),
The Physiology of Crustacea, vol. 1 : 227-258. Academic Press, New York.
PATTON, W. K., 1966. Decapod Crustacea commensal with Queensland branching corals.
Crustaceana, 10: 271-295.
POTTS, F. A., 1915. Hapalocarcinus, the gall-forming crab with some notes on the related
genus Cryptochirus. Papers Dept. Marine Biology Carnegie Institution, 8: 33-69.
A CRAB COMMENSAL ON CORAL 67
RATHBUN, M. J., 1921. Report on the Brachyura collected by the Barbados-Antigua Expedi-
tion from the University of Iowa in 1918. Univcrsitv of I two Stud. Nat. Hist., 9:
65-90.
RATHBUN, M. J., 1930. The cancroid crabs of America of the families Euryalidae, Portunidae,
Atelecyclidae, Cancridae and Xanthidae. Bull. U. S. Nat. Mus., 152: 1-609.
RILEY, G. A., 1963. Organic aggregates in seawater and the dynamics of their formation and
utilization. Liinnol. Oceanog., 8: 372-381.
STORK, J. F., 1964. Ecology and oceanography of the coral-reef tract, Abaco Island, Bahamas.
Spec. Pap. Geol. Soc. America, No. 79, 98 pp.
VAUGHAN, T. W., 1915. The geologic significance of the growth-rate of Floridian and Baha-
man shoal-water corals. /. ll'ash. Acad. Sci., 5: 591-600.
VAUGHAN, T. W., 1919. Corals and the formation of coral reefs. Smithsonian Institute Ann.
Rep.. 1917: 189-238.
WELLS, J. W., 1956. Scleractinia. In: R. C. Moore (ed.), Treatise on invertebrate paleon-
tology. Part (F) Coelenterata : F328-F444. Geol. Soc. Amer. & Univ. of Kansas
Press.
WUODS-IONES, F., 1907. On the growth forms and supposed species in corals. Proc. Zool.
Soc. London, 1907: 518-556.
AMOUNT, LOCATION, PRIMING CAPACITY, CIRCULARITY AND
OTHER PROPERTIES OE CYTOPLASMIC DNA IN
SEA URCHIN EGGS x
LAJOS PIKO,2 ALBERT TYLER AND JEROME VINOGRAD
Divisions of Biology and Chemistry,3 California Institute of Technology,
Pasadena, California 91109
The presence of large amounts of DNA in the cytoplasm of the mature egg of
many species of animals has been reported by a number of early workers (for sum-
maries see Brachet, 1962; Haggis, 1964; Grant, 1965; Monroy, 1965; Tyler and
Tyler, 1966b). Values several hundred times that of the nucleus have been re-
ported in eggs of sea urchins and frogs which have been the most extensively inves-
tigated material. However, the methods employed in the early determinations did
not clearly distinguish between DNA and materials, such as polysaccharides and
RNAs, that might interfere with the determinations, and, in fact, as the methods
have become more refined the reported values have dropped. Thus, in Paraccn-
trotus Ih'idns Hoff-JoYgensen (1954) obtained about 20 times the haploid (H)
value by microbiological assay and Whiteley and Baltzer (1958) obtained values
at the 32-cell stage by a fluorometric method that extrapolate to a similar value for
the unfertilized egg. In Hemicentrotus liridns Sugino ct al. (1960) reported
about 37 X H on the basis of thymidine determinations. Piko and Tyler (1965)
obtained approximately 13 X H and 8 X H, respectively, in Lytcchinns pictns and
Stronyyloccntrotits pnrpnratus by differential and buoyant density centrifugation
methods. Eberhard and Mazia (1965), from fluorometric measurements, esti-
mated about 180 X H in S. pnrf>iiratiis but indicated that the material that reacted
with the 3,5-diaminobenzoic acid dihydrochloride in their experiments might not
all be DNA. Baltus ct al. ( 1965), using a microfluorometric method, and Bibring
ct al. ( 1965), using centrifugation methods, found about 25 X H in Arbacia Uvula.
The location of the egg cytoplasmic DNA has also been uncertain. A few
years ago substantial evidence first appeared for the presence of DNA in the mito-
chondria of cells of a number of organisms, including chick embryo (Chevremont.
1962; Nass and Nass, 1963), mammalian tissues (Swift ct al., 1964; Schatz et al..
1964b), protozoa (Steinert ct al., 1958; Rudzinska ct al., 1964), molds (Luck and
Reich, 1964), yeast (Schatz ct al., 1964a), ferns (Bell and Miihlethaler, 1964).
maize (Ris, 1962). The evidence has accumulated since these first investigations
and DNA is now generally considered to be an integral part of the mitochondrion
(for reviews and further evidence see Gibor and Granick, 1964; Swift, 1965; Nass
1 Supported by grants from the National Science Foundation (GB-28) and from the Na-
tional Institutes of Health (GM 12777) and (CA 08014). The authors wish to acknowledge
the effective technical assistance of Peter N. Redington, Edward E. Vivanco and Robert Watson.
- This author during the latter part of this investigation has been on appointment as Chief,
Developmental Biology Laboratory, Veterans Administration Hospital, Sepulveda, California.
3 Contribution No. 3475 from the Division of Chemistry and Chemical Engineering.
68
CYTOPLASMIC DXA IX SKA URCHIN ECCS 69
ct nl.. 1965; Rabinowitz ct a!., 1965; Suyama and Freer, 1965; Dawid, 1966;
Corneo ct a!.. 1966; Sinclair and Stevens, 1966). It seemed possible, then, that
the egg cytoplasmic DNA might be entirely contained in these bodies. In fact a
calculation ( Piko and Tyler, 1965) based upon an estimate of the volume occupied
by mitochondria (ca. 9%) in sea urchin eggs (Shaver, 1956, 1957) and reported
values (c.y., Schatz et al., 1964a) of DNA in mitochondria, gives approximately
the amount of cytoplasmic DNA that has been found.
Baltus and Brachet (1962) (cf. Roller, 1963; Brachet, 1965) found about two-
thirds, at least, of the cytoplasmic DNA of frog's eggs to be associated with large
particles (pigment granules and yolk platelets) that sediment at low speed. Yolk
spherules of certain types, however, have been shown to be derived from mito-
chondria at least in some species (Lanzavecchia, 1960, 1965; Ward, 1962; cf.
Srivastava, 1965). In eggs of the clawed toad Xenopns lacvis, and the frog
Kana pipiens, Dawid (1965, 1966) has found DNA in the mitochondria. The total
obtained in the mitochondria! preparations accounts for some 65 to 8Q% of the
cytoplasmic DNA.
Comparisons have been made of cytoplasmic with nuclear DNA with respect
particularly to buoyant density and estimated molecular weight. Bibring ct al.
( 1965 ) indicate that in P. livid us it has a high molecular weight and a base com-
position similar to that of nuclear DNA. They also report that the buoyant
density of the DNA (presumably mostly cytoplasmic) extracted from eggs is similar
to that found in sperm. Garden et al. (1965) also found in Arbacia punctulata
that nuclear and cytoplasmic DNAs exhibit the same buoyant density in CsCl gra-
dients. In the experiments reported here the cytoplasmic DNA is found to djffer
in buoyant density from that of the nucleus.
The sedimentation behavior of this material indicated a similarity to the DNAs
of various viruses that are known to have a closed circular structure of uniform
circumference, as shown electron microscopically by \Yeil and Vinograd (1963)
for the DNA of the polyoma virus and by Kleinschmidt ct al. ( 1963) and Chandler
ct al. ( 1964) for the replicating form of bacteriophage 0X174. Also, as Vinograd
ct al. (1965) have shown, the circles are composed of double helices that are in
a superhelical form unless scissions are introduced into one or the other of the two
strands. Circular DNAs with circumferences ranging from 0.5 to 9.7 microns have
been found by electron microscopy in DNA preparations from boar sperm by
Hotta and Bassel (1965). Recent studies by Borst and Ruttenberg (1966) and
by Van Bruggen ct al. (1966) have shown the presence of circular DNA in mito-
chondria] preparations from chick and mouse liver and from beef heart. The
circles were of uniform circumference (ca. 5.45 microns). Sedimentation velocity
analyses revealed two components with standard sedimentation coefficients of 39-42 S
and 27-29S that correspond to the twisted and relaxed circular forms described by
Vinograd et al. (1965). Similar findings have been reported by Sinclair and
Stevens (1966) for mouse liver mitochondria. Our own studies indicate that the
DNA of the mitochondria of sea urchin eggs is also of the circular type as will be
reported here and in more detail in a subsequent paper.
Evidence that cytoplasmic DNA may be potentially active in oocytes or mature
eggs has been provided by experiments of Shmerling ( 1965 ) on sturgeon oocytes,
showing that DNA extracts that must contain predominantly cytoplasmic DNA
70 L PIKn, A. TYLER AND J. VINOGRAD
possessed priming activity for DXA and for RNA synthesis equal to that of DNA
rxlrartrd from the sperm. There is also evidence for in rivo activity of the cyto-
})lasmic DNA of eggs. Thus Mezger-Freed (1()63) has reported that artificially
activated enucleated frog eggs (with the nucleus in an attached exovate ) synthesize
DNA about as rapidly as do the fertilized eggs during early cleavage. Similarly
in sea urchins Baltus ct al. (1965) report synthesis of DNA by artificially activated
non-nucleate fragments and, in addition, the synthesis of RNA. In the present
experiments the DNA extracted from the mitochondria of sea urchin eggs was
found to be capable of serving as primer for RNA synthesis.
MATERIALS AND METHODS 4
(A) £(/</ and sperm samples
The sea urchins Lytecliinus pictits and Strongylocentrotus piirpnratus were
used in these experiments. The eggs were obtained by KC1 injection and handled
in artificial sea water as described elsewhere (Tyler and Tyler, 1966a). After
removal of the gelatinous coat in pH 5 sea water and thorough washing the eggs
were suspended in 0.55717 KC1 following three approximately 30:1 (v/v) wash-
ings in this solution. Samples were removed from the penultimate suspension for
counting (Tyler and Tyler, 1966a).
"Dry" sperm ( semen ) was collected as it exuded freely from dissected gonads
and diluted with artificial sea water to a stock solution of 0.5 to 1.0^ • Sperma-
tozoal counts were made with the Coulter electronic counter having a 30 /j. orifice.
As a check, counts were also made by hemocytometer.
(B) Preparation of hoinogcnates
The preparative procedure was similar in principle to that of Kay (1964);
namely, the use of detergent followed by salt extraction. In addition EDTA was
included further to reduce the possibility of nuclease activity and to serve as a
buffer supplemental to the egg material itself. In these experiments one volume
of packed KCl-washed eggs was mixed with three volumes of the homogenization
medium [4% sodium dodecyl sulphate (SDS), 0.08 717 ethylenediamine tetraace-
tate and 9% ethanol, pH 7.SJ and stirred gently for 20 to 30 minutes at 20° C,
with a Teflon rod. CsCl (optical grade, Harshaw Chemical Company) was then
added, with continued slow stirring for about one-half hour, to give the desired
final densities. These were 1.5 gm./cm.3 in the initial experiments in which at-
tempts were made by differential centrifugation to remove Dische-interfering mate-
rials and 1.70 gin. /cm." in the later experiments in which the DNA was isolated
by buoyant density centrifugation. Marker C14-DNA (see below) dissolved in
3.75 molal CsCl, when used, was added at this time in the ratio of 1 volume to 150
or 300 volumes of homogenate. In some cases the homogenates were stored at
4 Abbreviations used in the text: DNA, deoxyribonucleic acid; RNA, ribonucleic acid;
DNase, pancreatic deoxyribonuclease ; SDS, sodium dodecyl sulphate; EDTA, ethylenediamine
tetraacetate ; TCA, trichloroacetic acid; SSC, standard saline-citrate (0.15 M Nad, 0.015 M
Na citrate, pH 7) ; ATP, GTP, UTP, CTP, adenosine-, guanosine-, uridine-, and cytidine-
triphosphate, respectively.
CYTOPLASMIC DNA IN SEA URCHIN EGGS 71
-20° C. before the addition of CsCl, and this had no apparent effect on the
amount of DXA later obtained or on its properties.
(C) Marker DNA
Fertilized Lytcchiints Rictus eggs, at a density of 40,000 eggs/ml., were raised
until the gastruia stage in artificial sea water containing 50 /iC./ml. C14-thymidine
(30 /nC./AtM specific activity. New England Nuclear Corp.). After several wash-
ings with ice-cold sea water and 0.55 M KG, the embryos were homogenized with
SDS-EDTA and the DNA was isolated by buoyant density centrifugation in
CsCl solution. Fractions of 0.15 ml. were collected by puncturing near the bot-
tom of the centrifuge tubes and examined for absorption at 260 m^ and for radio-
activity. A single sharp peak of absorption, and of radioactivity, was obtained.
The fractions containing this material were pooled and stored in a freezer at
— 20° C. In two separate preparations made in this manner one had a specific
activity of 12 X 10'fi mC./mg. DNA and the other 16 X 10'". In the experiments
involving the use of marker DNA the amount of radioactivity was measured in
a Packard Tri-Carb scintillation counter at 50% counting efficiency by a method
described elsewhere (Tyler. 1966).
(D) DNA determinations
Measurements of the DNA content of the various preparations in the initial
experiments were made by the diphenylamine reaction of Dische (1930). This was
used on hot acid extracts (15 minutes extraction with an equal volume of 10%
TCA in a boiling water bath) of spermatozoal suspensions, and egg homogenates
and the various fractions thereof. Aliquots, mixed with Dische's reagent and
blank, respectively, were heated in a boiling water bath for 10 minutes, cooled
rapidly and allowed to stand at room temperature for one hour. The absorptions
were read with a Beckman DU spectrophotometer and DNA values calculated
by comparison with a deoxyadenosine standard. In some cases, spectral absorp-
tion curves between 500 and 700 m/* were taken with a Cary Model 15 spectro-
photometer.
For the purified materials isolated by the centrifugation methods the DNA
values were determined directly from measurements of the absorption at 260 mp,
assuming a value (Ogur and Rosen, 1950) of 0.050 mg./O.D. unit.
(E) Centrifugation
All high speed centrifugations were done in a Spinco model L or L2-65 ultra-
centrifuge with an SW-39 or SW-65 rotor. The conditions of centrifugation are
indicated in the individual experiments. The preformed density gradients em-
ployed in these experiments were prepared by use of a gradient-making device
of the type described by Leif and Vinograd (1964) but capable of filling three
tubes at one time.
The results of the buoyant density experiments were evaluated by the proce-
dures described by Vinograd and Hearst (1962). All buoyant densities were
calculated by the marker method. A value of 1.710 gm./cm.3 for E. coll DNA
was assumed.
72 L. PIK('), A. TYLER AND J. VINOGRAD
Following llit' experience of Sh:iver (ll'5()) \vith the sea urchins used in these
experiments, mitochondria! preparations were made hy homogenization of KC1-
washed eggs in 3 volumes of a solution containing 0.3 M sucrose, 0.36 M KG,
0.03 M Tris-HG and 0.003 M EDTA at pH 7.6. Homogenization was done hy
hand with a loose fitting Teflon pestle in a Potter-Elvehjem tube of about 2 cm.
diameter containing 15 ml. of suspension, for a total of 10 strokes in the cold at
relatively slow speed. The homogenates were first centrifuged for 10 minutes at
1200 to 1500 rpm ( SW-25 rotor of the Spinco model L centrifuge, 4° C. ) to remove
nuclei and large egg fragments. The supernatant was then centrifuged at 12,000
rpm in the same rotor for 20 minutes. The resulting pellet was resuspended in
homogenization medium and recentrifuged under the same conditions. Further
processing is described below.
(G) Priming actii'ity
The ability of the extracted DNA to serve as primer for RNA synthesis as
first described by Weiss and Gladstone (1959), was measured from the incorpora-
tion of C14-labeled CTP in a system containing also the other three trinucleotides
(GTP, ATP, UTP), and an RNA polymerase prepared from E. coll by the
method of Chamberlain and Berg (19(>2). We are indebted to Dr. K. Marushige
for these materials.
RESULTS AND COMMENTS
(A) DNA content
(1) Determination by Dischc reaction
When the "DNA" of homogenates of whole eggs of 5\ pitrpitratns and L. pictits
was directly assayed by the Dische reaction, the amount per cell ranged from 90
to 150 times that of the sperm. These are similar to the values reported by a
number of earlier investigators as noted above. However, as Elson ct al. (1954)
noted with sea urchin eggs, the color that is obtained is not that characteristic of
deoxyribose alone. In our tests the reaction mixture showed an absorption peak
at 530 m/A. Figure 1 illustrates an absorption curve for the egg material (curve A)
along with that of the sperm ( D ) and one for deoxyadenosine ( F ) .
One initial attempt to remove the interfering material was done by differential
centrifugation of the homogenates brought to a density of 1.5 gin. /cm.3 with CsCl.
After 14 hours of centrifugation at 35,000 rpm (Spinco SW-39 rotor) the material
separated into a small gelatinous bottom pellet, a rubbery top layer (occupying
about 5% of the tube when the final homogenate contains some 12 to \4% of eggs )
and a clear intermediate fluid. Dische reactions were run on the combined pellet
and intermediate fluid and on the top layer in the two species. Determinations
were also made on sperm. The top layer contained about four-fifths of the 595 in//,
absorption values of whole egg homogenates but the absorption curve is similarly
abnormal (curve B of Figure 1). The possibility that "trapped" DNA may be
contained in it is considered in a later section. For the combined pellet and clear
CYTOPLASMIC DNA IN SEA URCHIN EGGS
73
I
O.D.
0.10
0.05
500
650
550 600 650 500 550 600
WAVELENGTH, m/i WAVELENGTH, m/i
FIGURE 1. Spectral absorption curves of colors obtained with the Dische reaction. (A)
Whole egg homogenate of S. purpitrutns; (B) top layer and (C) combined pellet and clear
layer obtained after differential centrifugation of homogenates of S. pitrpitratiis eggs (see text) ;
(D) sperm of L. pictus; (E) pre-purified and alpha amylase-treated extract from eggs of
L. pic tits; (F) deoxyadenosine standard.
layer the following values (in micrograms "DXA" per million eggs) were obtained:
17.7, 19.3, 20.5, 22.0; av. 19.9
19.6, 16.2, 16.4; av. 17.4
.9. purpuratus
L. pictus
For the sperm the values were as follows :
.V. purpuratus
L.
0.72, 0.82; av. 0.77
0.84, 0.90; av. 0.87
The spectrum of the Dische-reacted combined pellet and clear layer (curve C of
Figure 1 J is somewhat less abnormal than that of the other preparations.
After exploration of a number of methods of disposing of the interfering mate-
rial the use of alpha amylase proved effective. Preparations that are obtained by
a single buoyant density separation show a turbid polysaccharide layer (see section
3) at the same level of the tube where the DXA is located. \Yhen this fraction,
after it is precipitated with alcohol and redissolved, is treated with alpha amylase,
as described in section 3, it gives a typical deoxyribose spectrum in the Dische
reaction, as illustrated in curve E of Figure 1. By this procedure the DXA values
obtained for the two species were :
L. pictus 8.0 pg./egg
S. purpuratus 3.6 pg./egg
These are similar to the values obtained by UV absorption measurements on puri-
fied DXA as described below.
(2) Evidence against "trapping" of DNA in the top la\cr
In order to examine the possibility that DXA might be trapped in the top layer
that forms upon centrifugation of homogenates in the CsCl solutions, use was
74 L. PIKn, A. TYLER AND J. VINOGRAD
made of the CMabeled gastrula DNA. In five experiments with L. plctns a
sample of the labeled DNA was added to the homogenate and determinations made
of the distribution between the top layer and the clear fluid after prolonged cen-
trifugation in the CsCl solutions employed for buoyant density separations (see
below). The values obtained in these experiments for the ratio of labeled DNA
in the top layer to that in the clear fluid ranged from 1:99 to 2:98.
(3) Direct determinations on purified material
Preparations of DNA were made from whole eggs of S. pnrpuratus (3 experi-
ments) and L. pictns (5 experiments) by buoyant density centrifugation of homo-
genates adjusted to a density of 1.70 gm./cm.3 with CsCl and containing radioactive
(marker) gastrula DNA. Centrifugation was for at least 50 hours at 35,000 rpm
at 10-12° C. in the SW-39 rotor of the Spinco model L centrifuge. A small
amount of solid CsCl is present at the bottom of the tube at the end of the run.
For collecting the fractions, then, the hypodermic needle is introduced above this
layer (approximately 7 mm. from the bottom). Usually 15 to 20 fractions were
collected and the radioactivity of small aliquots determined. The fractions com-
prising and surrounding those with the radioactive DNA were pooled and CsCl
solution (1.70 gm./cm.3 ; in 0.02 M Tris-HCl pH 7.6) added so as to give a volume
sufficient to fill the centrifuge tubes, which were then re-run as before.
It was noted, early in these experiments, that a band of visible turbidity, later
identified as polysaccharide (Piko and Tyler, 1965 ; Segovia et a!., 1965), appeared
in the region of the centrifuge tube where the marker DNA was located. This
material continued to appear at the level of the DNA upon repeated centrifuga-
tions and it obscured the O.D.200 readings. For removal of this material two
methods were explored. One was centrifugation of alcohol-precipitated and re-
dissolved fractions on preformed CsCl density gradients (1.22 to 1.65 gm./cm.3
for 4 hours at 35,000 rpm, 20° C.) in which the polysaccharide sediments (fa.
100S) well ahead of most of the DNA. The other method was simply to incubate
a solution (0.5717 KC1, 0.01 M Tris, 0.005 M EDTA, pH 7) of "the alcohol-
precipitated DNA and polysaccharide-containing fractions with a-amylase (Worth-
ington, 2 X crystallized, at 0.75 nig. /ml. for 1 hour at 37° C.). This method
proved to be the more effective. The digestion with alpha amylase was generally
done with the fractions collected after the first or second centrifugation. Following
this the buoyant density centrifugations and collection of the fractions were re-
peated two times.
After the final buoyant density centrifugation, absorbances of the fractions at
260 and 280 HI/A were read on a Beckman DU spectrophotometer. Radioactivity
(of the marker DNA) determinations on aliquots again served to locate the DNA-
containing region, and to provide an additional basis for quantitation. The frac-
tions collected after the final centrifugation by this procedure showed a single
O.D.2GO peak at a level corresponding to a density near 1.70 gm./cm.3 The marker
DNA showed a single peak of radioactivity in the region of 1.69. This is illus-
trated in Figures 2a and 21>. In one of these experiments an aliquot was treated
with DNase (Worthington, electrophoretically purified, 0.1 mg./ ml. at 37° C. for
30 minutes) before the final buoyant density centrifugation. This resulted in com-
plete elimination of both the O.D.,00, and the radioactivity, peaks.
CYTOPLASMIC DNA IN SEA URCHIN EGGS
75
DNA prepared in this way shows a typical absorption spectrum, as illustrated
in Figure 3. In 5 experiments with unfertilized eggs of L. pictus and 3 experi-
ments with .V. purpiiratns, in which the determinations were made by this method,
the following values were obtained for the content of DXA in micrograms pel-
million eggs.
7.9, 9.3, 7.5, 8.2, 8.4; av. 8.26 ± 0.30
L. pictus
S. purpitratus
3.5, 2.8, 3.6; av. 3.30 ± 0.25
(4) Extraction of DNA from eggs labeled during oogenesis
Further evidence for the effectiveness of the extraction procedure has been
obtained in an experiment in which the DNA was labeled radioactively during
oogenesis by the general procedure described by Tyler and Tyler (1966a).
In
OPTICAL DENSITY
AT 260 mM
-0.6
-0.4
-0.2
O.D.260
COUNT/MIN
(a)
COUNT/MIN
IN IOX
I25H
100-
75-
50-
25 H
OPTICAL DENSITY
AT 260 m/i
-0.6
-0.4
- 0.2
O.D.260
COUNT/MIN
(b)
COUNT/MIN
IN I5X
250 -1
200-
150-
100-
50-
o-o-o-o
o-o-o-o
1.0 2.0 3.0
L. PICTUS EGG DNA
4.0, Til 1.0 2.0 3.0
S. PURPURATUS EGG DNA
4.0ml
FIGURE 2. Cesium chloride gradients of DNA from unfertilized eggs of L. pictus and
S. purpiiratns. The Cu-labeled nuclear marker DNA (from L. pictus gastrulae) bands at a
somewhat lower density than the bulk of the cytoplasmic DNA (for procedures see text).
this experiment a female L. pictus received two successive injections, intracoelomi-
cally, of 200 microcuries of H3-thymidine (6 curies/mM) at a one-month interval
and the eggs were collected one month after the second injection. About 100,000
eggs were obtained, and tests on an aliquot showed about two-thirds of the radio-
activity to be in acid-precipitable form. Upon extraction by the procedures em-
ployed here (see section 3) all the labeled material, that was identified as DNA,
was found, upon buoyant density centrifugation, to be in a layer at a density near
1.70 gm./cm.3 This material contained about 0.1% of the originally injected
radioactivity. In addition there was an approximately equal amount of radio-
activity at the top of the tube. This material, upon treatment with preincubated
(1 hour) pronase (Calbiochem, final concentration 2 mg./nil. in 0.25 M CsCl,
0.005 AI Tris, 0.001 M EDTA, 5% ethanol, pH 7.6; incubated at 50° C. for 12
hours) lost more than 90% of its originally acid-precipitable radioactivity. It
76
L. PIK6, A. TYLER AND J. VINOGRAD
niav, then, he concluded that the extraction procedure yields practically all the
DNA obtainable from the egg.
(B) Presence of DNA in mitochondria and yolk
Two sets of experiments were run in which homogenates of L. pictus eggs (of
determined number) were subjected to differential centrifugation, as described
under Methods, so as to separate a 250 X g nuclear (N) pellet, a 18,000 X g
mitochondria + yolk (M + Y) pellet and a supernatant (S) fraction. Two or
three consecutive buoyant density centrifugations in CsCl solution were performed
. RICTUS GASTRULA DNA
L. RICTUS EGG DNA
240
260 280
WAVELENGTH, m/t
300
FIGURE 3. Ultraviolet absorption spectra of purified DNA in 0.015 M NaCl-0.0015 M sodium
citrate, pH 7, from gastrulae and unfertilized eggs of L. pictus.
on each fraction, as described in section A3, the treatment with alpha amylase being
applied on the fractions collected after the first centrifugation. From O.D.2GO
readings the following amounts of DNA in micrograms per million eggs were
obtained. The distribution in per cent of the total is given in parentheses.
Experiment 1: N == 1.51 (20.5%) ; M + Y == 5.03 (68%) ; S = 0.85 (11.5%)
Experiment 2: N = = 0.51 (8.5%) ; M + Y •= 3.84 (65%) ; S == 1.57 (26.5%)
In experiment 1 no separate determination of DNA content of whole eggs was
made, but if the average value of 8.26 pg. per egg from other experiments is taken
then the recovery here is about 90%. In experiment 2 parallel determination of
DNA content was made on an aliquot of the initial whole egg homogenate. This
CYTOPLASMIC UNA IX SEA URCHIN EGGS 77
gave a value of 8.40 micrograms per 10(i eggs, as corrected for 76 (/( recovery of
marker DXA. If we assume a similar recovery for the above fractions then the
corrected total for them is 7.8 micrograms. which would indicate very little, if any.
loss in the fractionation procedure.
As the results show, the hulk of the DXA is found in the M + Y fraction. The
DNA content of the N fraction is lower in experiment 2 than in experiment 1. In
experiment 1 this fraction had not been washed. Microscopic examination has
shown that some mitochondria and yolk spherules do sediment with this fraction.
This probably accounts, then, for the value of its DXA content being almost twice
that expected for the nuclei alone.
The S fraction is free of microscopically visible mitochondria and yolk particles,
and shows in the two experiments 11.5 and 26.5%, respectively, of the total DXA
content extracted from the eggs. \Yhether or not this DXA may be derived from
damaged mitochondria or yolk cannot be stated at present.
An aliquot of the M + Y fraction in the second experiment, with L. pictns. was
subjected to further processing by centrifugation on preformed linear gradients of
sucrose solutions (from 0.93 M to' 1.88 M sucrose in 0.003 M Tris, 0.0025 A/EDTA.
pH 7.6) in the S\Y-25 rotor of the Spinco model L at 25,000 rpm for 2 hours at 4°
C. 2.5 ml. of the suspension of M + Y fraction being layered on 25 ml. of gradient in
each tube. Under these conditions the yolk (Y ) remains on the top of the gradient
while the mitochondria (M) sediment as a band that is visible as a cloudy layer
some 4 mm. wide at a region of the tube where, as determined by subsequent
weighing, the density is 1.18 gin. /cm.3 This fraction and the top one were col-
lected, diluted with three volumes of 0.5 M KC1 (containing 0.05 M Tris, 0.005 M
EDTA, pH 7.6) and centrifuged at 12.000 rpm ( S\Y-25 rotor) for 20 minutes.
The pellets were suspended in SDS-EDTA solution, CsCl added to a density of
1.70 gin. /cm.3 and the solutions subjected to two buoyant density centrifugations
and fractionations, with intervening alpha amylase digestion, as described pre-
viously. The following values were obtained for DXA in micrograms per mil-
lion eggs.
Experiment 2: M == 2.07; Y - 0.72
The sum represents 73% of the amount of DXTA present in the A I + Y fraction,
as listed above.
In a separate experiment (3 ) an M + Y fraction of L. pic t us eggs was prepared
and all of this used for preparation of M and Y fractions as described in this section.
The following values for DXA content (micrograms per million eggs originally
extracted) were obtained.
Experiment 3: M " 2.47 ; Y -0.35
From these experiments it is clear that the bulk of the DXTA appears in the
mitochondrial fraction. In the two experiments (2 and 3) in which M + Y WPS
separated into M and Y the ratios (M:Y) of DXA content were 3:1 and 7:1.
respectively. From the sedimentation behavior, including the wide separation of
the two fractions, and from microscopic examination it is unlikely that the yolk
fraction contains any significant amount of mitochondria as such. Considering also
the lack of any appreciable trapping of marker DXA in that layer it is most reason-
78
L. PIK6, A. TYLER AND J. VINOGRAD
able to conclude that the DNA found therein is a component of the yolk spherules.
The differences in the relative amounts of DNA obtained from the mitochondria!
and yolk fractions in the two experiments may be explained by the sensitivity of
these particles to damage during the extraction procedures. The results of cesium
chloride buoyant density centrifugation of DNA from mitochondria and from yolk
are illustrated in Figure 4a, 41 >. Both DNAs behave similarly, forming bands at
somewhat higher density than the added radioactively labeled nuclear DNA.
In a preparation made by Dr. E. R. Berger, now of the Veterans Administra-
tion Hospital, Sepulveda, approximately 2140 mitochondria and 2280 yolk spherules
were counted on a montage of electron micrographs of a thin section (maximum
diameter) of an egg of L. pictns. From these figures, and values of 2.0 microns
for the diameter of a yolk spherule and equivalent spherical diameter of 0.8 micron
OPTICAL DENSITY
AT 260 m^.
-0.15
-0.10
-0.05
(a)
COUNT/MIN
IN I5X
150-
°-D-260
COUNT/MIN
100^
50-
OPTICAL DENSITY
AT 260 m/x
-0.15
COUNT/MIN
IN I5X
150-
O.D.,
'260
COUNT/MIN
-0.10
-0.05
(b)
100-
50-
2.0
3.0
ml
1.0
2.0
3.0
ml
FIGURE 4. Cesium chloride gradients of DNA from (a) isolated mitochondria and (b)
yolk of L. pichis. Each preparation contained added CMabeled marker DNA from L. pictus
gastrulae. The band of radioactivity is at a lower density than that of O.D.a* absorption for
both the mitochondrial and yolk-DNA.
for a mitochondrion, it may be estimated that there are some 80,000 yolk spherules
and some 200,000 mitochondria per egg. This corresponds to the approximately
3:1 ratio for the DNA found in mitochondria and yolk in experiment 2.
(C) Priming activity for RNA svntlicsis
Purified DNA from eggs of L. Rictus and S. pitrpnmtiis was tested for ability
to serve as primer for RNA synthesis. DNA that is preponderantly (more than
99%) nuclear was prepared from late blastulae and from plutei of S. pnrpuratus,
to serve as a basis for comparison. The measurements were made of the incor-
poration of radioactive label into RNA (material precipitable by 10% trichloroacetic
acid) in a system containing, in 0.25 ml., the following: 0.1 /xmole C14-cytosine
triphosphate (1.4 /iC.//*A/), 10 /uncles Tris buffer, pH 8, 1 /uncle MgCl2J 0.25
//.mole MnClo, 3 /mioles beta mercaptoethanol, 0.1 /^mole each of ATP, GTP and
CYTOPLASMIC DNA IN SEA URCHIN EGGS 7(J
UTP, and the RNA polymerase from E. coll. The following values were obtained
in terms of counts per minute (cpm at 30% counting efficiency) per 5 /xg. DNA
above a background of about 90 cpm for the complete mixture minus the DNA.
5. pnrpiiratus: Unfertilized egg DNA (80% cytoplasmic) : : 4,056 cpm
Blastula DNA (<\% cytoplasmic) =3,618 cpm
IMuteus DNA (< \% cytoplasmic) : 4,305 cpm
L. pictus: Unfertilized egg DNA (90% cytoplasmic) = 3,027 cpm
These initial values are all in the same general range. If only the nuclear DNA
of the egg preparations were active, the values for these preparations would have
been very much lower (one-fifth to one-tenth of those obtained). It may be con-
cluded, then, that the cytoplasmic DNA can function as primer for RNA synthesis.
(D) Sonic physical properties of the cytoplasmic ( niitochondrial) DNA
Detailed studies of various physical properties of the cytoplasmic DNA of
L. pictus are in progress and will be reported elsewhere. Here some preliminary
information is given concerning its density, melting behavior, sedimentation prop-
erties and microscopic appearance.
(1) Buoyant density. In the preparative buoyant density centrifugations in
CsCl solution of whole egg homogenates, with radioactive marker DNA included,
the O.D.2GO readings consistently show a peak at a higher density (ca. 1.70 gm./
cm.3) than the peak of radioactivity of the marker (ca. 1.69 gm./cm.3). This is
illustrated in Figures 2a, 2b. This is also true for the DNA obtained from iso-
lated mitochondria and yolk as shown in Figures 4a, 4b. In further buoyant
density centrifugations of purified whole egg DNA of L. pictus in the analytical
(Beckman Spinco Model E) centrifuge three bands were observed in scans at 265
in/*. The buoyant densities were 1.693, 1.703 and 1.719 gm./cm.3 (see Fig. 5).
The relative amounts of DNA in these three bands were of the order of 1:7:1.
Scans at 280 m^ again revealed three bands in which the ratios of the areas were
approximately the same as at 265 ni/i. In the same rotor sperm DNA and gastrula
DNA form single bands at 1.693 gm./cm.3 The 1.703 band evidently represents
the bulk of the cytoplasmic DNA. The nature of the 1.719 band is not, as yet,
known.
(2) Melting temperature. Determinations of melting profiles were made on
purified (as described in section A3) DNA preparations from spermatozoa, late
gastrulae and unfertilized eggs of L. pictus. Sedimentation analysis showed that
the latter preparation contained a negligible fraction of intact, i.e., covalently
closed, circular DNA. After dialysis and storage in one-tenth strength standard
saline citrate (SSC '•= 0.15 M NaCl ; 0.015 Na citrate; pH 7.0), the measurements
were made in SSC. From the profiles (see Figure 6) the average melting tem-
peratures (Tm) are 84.0° C. for sperm and gastrula DNA and 86.8° C. for the
unfertilized egg DNA. In the latter case the value is not entirely attributable to
the cytoplasmic DNA since there is some 10% each of nuclear DNA and an
unidentified component of a buoyant density in CsCl of 1.719. This may explain
the atypical shape of the melting curve of the unfertilized egg DNA (cf. also
Vinograd and Lebowitz, 1966). From the Tm for the whole egg DNA a guanine-
80
L IMK('), A. TYLER AND J. YINOGRAD
cvtosine (G-C) content of 42('/< is calculated (Marniur and Doty, 1962; Schild-
kraut and Lifson, 1965), which approximates the G-C content of 44%> calculated
( Schildkraut ct al., 1962) from the CsCl buoyant density of 1.703 gm./cm.3 of
the major peak (presumably mitochondrial DNA) obtained from unfertilized eggs.
For sperm DNA and gastrula DNA the Tm and the buoyant density indicate a
G-C content of 35 and 34</f , respectively.
(3) Sedimentation behavior. In four separate experiments in which prepara-
tions (see section A3) of DNA from whole eggs of L. pictits were centrifuged on
preformed linear CsCl gradients (1.30 to 1.40 gm./cnv'f in the SW-65 rotor for
LJ
o
<
03
cc
o
CO
CO
DENSITY »-
FIGURE 5. Tracings of direct scans at 265 m/u of buoyant density bands (in CsCl) of three
preparations of DNA of L. pic t us after 25 hours of centrifugation in the same rotor at 44,770
rpm in the Beckman model E centrifuge. L: density marker DNA (1.731 gm./cm.3) of Micro-
coccus lysodcikticits. Ai, A- and A.t : DNAs identified as nuclear in the three preparations and
with similar buoyant densities of 1.693. B : DNA identified as derived from mitochondria and
yolk and with a buoyant density of 1.703. C : Unidentified nucleic acid band of buoyant density
1.719.
2 hours at 50,000 rpm, 20° C.) the nuclear DNA sedimented to the bottom of the
tube and the cytoplasmic DNA sedimented in two distinct bands, corresponding
to sedimentation coefficients of about 23S and 28S, respectively. Approximately
equal amounts of the two fractions were obtained, the amounts varying in different
preparations. As noted in the introduction, the presence of two such components is
indicative of the two forms of circular DNA (twisted and open circles) described by
Vinograd ct al. (1965) and found in mitochondria by van Bruggen ct al. (1966).
Further studies on this material will be reported in a separate paper.
CYTOPLASMIC DNA IN SEA URCHIN EGGS
(4) Electron microscopic ohscn'ations. Purified DNAs from whole eggs,
from mitochondrial fractions and from gastrulae of L. pictus were prepared for
electron microscopy according to the method of Kleinschmidt ct al. (1965). For
this purpose a small amount (0.2 ml.) of a solution of ammonium acetate (1.5 M ;
pH 7) containing DNA at about 4 micrograms per ml. and cytochrome c at 0.1
mg. per ml. was allowed to flow down an inclined glass slide onto a solution of
0.1 M ammonium acetate in a large dish. Electron microscope grids coated with
formvar were touched to the surface of the solution, passed through 95c/r ethanol,
0.0001 M uranyl acetate solution in 0.001 M HC1 and isopentane. The prepara-
tions were examined in a Philips EM200 and micrographs taken at a film magni-
40
in
CO
>• 30
o
CD
cr
CD
<
LJ
CO
<
LJ
cr.
o
20
10
Tm = 84.0°C
A
75
80
85 90 95
TEMPERATURE (°C)
100
105
FIGURE 6. Meeting profiles of DNAs (in 0.15 M NaCl-0.015 M sodium citrate, pH 7)
from L. pictus. Curve A: DNA from sperm and from gastrulae. Curve B: DNA from whole
unfertilized eggs.
fication of 5000 X. (We are indebted to Mr. James Wetmur of the Division of
Chemistry for the use of uranyl acetate in this procedure.)
In the preparations from both whole eggs and mitochondrial fractions the
DNA was seen to be present mostly in the form of closed circular filaments, whereas
none of these were seen in the preparations from gastrulae. The latter is estimated
to contain less than \c/c of cytoplasmic DNA. Examples of the circular DNA
are shown in Figure 7. Both twisted and open circles are seen. A considerable
uniformity of size of circles was observed. Measurements of 144 perimeters gave
values ranging from 3.75 to 4. S3 microns with a mean of 4.45 and a standard devia-
tion of 0.25. We are indebted to Mr. Donald Blair of the Division of Chemistry
for providing us with the foregoing quantitative results.
L. l'IK('), A. TYLER AND J. VINOGRAD
DISCUSSION
(A) UNA conical
The present determinations of DNA content of the unfertilized eggs have given
values in the general range of those obtained by the more recent workers on sea
urchin eggs. However, the values are significantly lower than any previously
reported except for that of Marshak and Marshak (1953). By an isotope dilution
method they obtained a value of 10 X H (haploid) for Arbacia pitnctiilata but
attributed most of this to contamination with somatic cells and polar bodies and
concluded there was some 3 X to 4 X H of cytoplasmic DNA. They also con-
cluded, from the failure to obtain a Feulgen reaction, that the nucleus of the unfer-
tilized egg lacked DNA. However, measurements by Hinegardner (1961) on
isolated nuclei of E. niathaci and of S. pitrpuratus showed that these contain 1 X H
of DNA and others (e.g., Burgos, 1955 ) have obtained a positive Feulgen reaction.
In the present experiments, the areas under the buoyant density bands in the
analytical ultracentrifuge indicate the nuclear DNA to be present in approximately
the haploid amount in L. pictus eggs.
In the present work evidence has been presented that the material on which
the final determinations were made was, in fact, DNA. Also, the monitoring with
radioactively labeled DNA permitted an assessment to be made of the effectiveness
of recovery during the preparative procedures. This was reinforced by the results
of the experiment in which DNA was extracted from eggs in which it had been
labeled during oogenesis. It seems reasonable to conclude, then, that the present
values of 8.26 pg. and 3.30 pg. per egg for L. pictus and 6*. purpuratus, respec-
tively, are close to the actual content of macromolecular DNA in these cells. With
regard to other species of echinoids that have been examined, since none of these
have an egg size larger than that of L. pictus and since the nuclear DNA is closely
the same for various species (cf. Tyler and Tyler, 1966a), it seems reasonable to
expect that the total DNA should be in the same range as the values reported here.
The much higher values that have been reported, in the absence of substantial
evidence of specificity of the analyses, would seem then to be attributable at least
in part to the presence of interfering materials.
In the two species that have been used in the present work the difference in
DNA content correlates with differences in egg size. Correlation with egg size
may account for the much greater values that have been reported for amphibian
eggs in investigations in which attempts have been made to eliminate interfering
materials. Thus, Baltus and Brachet (1962) report 0.069 /xg. for the axolotl.
Haggis (1964) reports 27,000 X H for Runa pipicns, and Dawid (1965) gives
values 600 to 1000 X II for R. pipicns and Xenopus laeris.
(B) Presence in mitochondria and yolk
The present results show, as suggested earlier (Piko and Tyler, 1965), that
the bulk of the cytoplasmic DNA is present in the mitochondria. This accords
with the current findings on amphibian (R. pipicns and A', lacvis) eggs by Dawid
(1966) who reports that at least two-thirds of the DNA is associated with the
mitochondrial particles. As noted in the introduction the general occurrence of
DNA in mitochondria is now well established from investigations with various
organisms throughout the animal and plant kingdoms.
CYTOPLASMIC DKA IN SEA URCHIN EGGS
83
^fe??:;..v^.- V:'-^^Vv/-"v.;:^/:
- •*,*V "»'•««'. \, '* ' "': '•" ' '{ •'*',•.'•• '. '*' , • •'.
^<dp^''£ :
jN •• •'• cv^----'-' •- '-4-- ••*.••*•.'••• J7 •• •
-;vv^;,:.r--^^--';-;v->,v- -V-" vi- v
^\'-::X^S^:^v^';^-'i^
• I •«/;*; •;-.; VS-: .*'» '' <
\/.'--;^::,:;,-:
•L • •-'
. t
•c-
r" .
FIGURE 7. Electron micrographs of DXA from mitochondria (A) and of DXA from gastrulae
(B) of L. pic tits. The line represents one micron.
84 L. PIK6, A. TYLER AND J. VINOGRAD
Our results also show the presence of DNA in the preparations of yolk
spherules. Although there is a possibility that this DNA is simply adherent to
the yolk spherules this seems unlikely in view of the preparative procedures. The
evidence accords then with that of Baltus and Brachet (1962), who found about
two-thirds of the DNA (as detected by a fluorometric method) of eggs of Plcnru-
dclcs to be associated with the particles (mostly yolk) sedimentable at low speeds
(280 X <-/). It also accords with the evidence of Brachet and Ficq (1964, 1965),
obtained by use of radioactively labeled actinomycin, that DNA is an integral com-
ponent of the yolk spherules.
There is, in addition, convincing evidence, from studies on frogs (Lanzavecchia,
1960, 1965 ; Ward, 1962) that at least some of the yolk spherules, or possibly all,
are derived from mitochondria. This seems likely to be the case for eggs of ani-
mals in general (cf. Srivastava, 1965). If one assumes no increase in amount of
DNA upon transformation of a mitochondrion into a yolk spherule, then our finding
of a much lower content of DNA in the total mass of yolk than in the mass of mito-
chondria is readily understandable on the basis of the relative volumes of the two
particles. Thus in LytccJumts, the unfertilized egg contains, by our rough estimate,
about 200,000 mitochondria, and 80,000 yolk spherules. From the areas under the
bands in the analytical buoyant density centrifugations it appears that seven-ninths
of the total DNA, namely 6.4 /xg. per 106 eggs, is in these particles. This gives
2.3 X 10 17 grams per particle, or 1.4 X 107 daltons.
(C) Metabolic properties
The present results showing priming activity for RNA synthesis on the part
of the sea urchin cytoplasmic DNA add to the evidence (see introduction) for
such activity on the part of the cytoplasmic DNA of eggs of other species (cf.
Shmerling, 1965 for fish; Baltus ct a!., 1965 for sea urchins; and Dawid, 1965 for
Amphibia). That sea urchin egg mitochondria are capable also of protein syn-
thesis was shown earlier by Nakano and Monroy (1958) and Giudice (1960).
It appears, then, that in mitochondria of sea urchin eggs the complete systems of
RNA-dependent protein synthesis and DNA-dependent RNA synthesis are present,
as has been demonstrated for mitochondria of other origin (cf. Kroon, 1963; Kalf,
1964).
Detailed measurements are not as yet available concerning the extent to which
the mitochondrial activity accounts for the protein synthesis that occurs upon
fertilization in sea urchins, but according to Nakano and Monroy (1958) and
Giudice and Monroy (1958) this remains insignificant during the first three or
four hours of development. The evidence for potential activity of mitochondrial
DNA does not then alter the conclusions that have been drawn earlier from experi-
ments with non-nucleate fragments (Tyler, 1963, 1965; Denny and Tyler, 1964;
Brachet, Ficq and Tencer, 1963) and with actinomycin D (Gross and Cousineau,
1963, 1964) of the existence in the unfertilized egg of an inactive ("masked")
messenger RNA that becomes active upon fertilization. In fact the experiments
with actinomycin D provide particularly strong arguments against the possibility
that an activation of mitochondria might be responsible for the great increase in
protein synthesis that occurs upon fertilization, inasmuch as it is known (Kalf,
1964) that incorporation of amino acid into protein by intact mitochondria is sensi-
CYTOPLASMIC DNA IN SEA URCHIN EGGS 85
tive to actinomycin. Further arguments are provided by the fact, demonstrated
originally by Hultin (1961), that the difference between unfertilized and fertilized
eggs is exhibited also by cell-free systems which, from the method of preparation,
are evidently free of mitochondria.
(D) Physical properties
The buoyant density centrifugation in CsCl solutions in the preparative ultra-
centrifuge consistently showed the cytoplasmic DNA to have a higher density
than the nuclear, in the range of 1.70 to 1.71 gm./cin.3 A determination by the
analytical ultracentrifuge gives a value of 1.693 for the nuclear and 1.703 for the
bulk of the cytoplasmic DNA in L. pictus eggs, and a small (10% of the total
DNA) band at 1.719. While the latter is a nucleic acid band, as indicated by the
O.D.2(;o-O.D.2SO ratios and other properties, it could be a DNA-RNA hybrid rather
than DNA alone. Detailed studies of physical properties of the various components
are in progress and will be reported elsewhere.
For the nuclear and cytoplasmic DNA of other species there have been reports
both of similarities and of differences in buoyant density. Thus in Arbacia piinc-
tulata. Garden et al. (1965) reported similar buoyant densities for the two DNAs.
In Rana pipiens, Dawid (1965, 1966) reports that the two DNAs have the same
buoyant density (1.702 gm./cm.3) while in Xenopus laez'is the cytoplasmic DNA
is slightly denser (by 0.002 gm./cm.3) than the nuclear. In chickens 1.707
gm./cm.3 for the mitochondrial DNA, and 1.698 gm./cm.3 for the nuclear, are
reported by Rabinowitz ct al. (1965) and Borst and Ruttenberg (1966). The
latter, and also Sinclair and Stevens (1966) report similarity of buoyant density
for mitochondrial and nuclear DNAs of various mouse tissues, while Schneider
and Kuff (1965) report a somewhat lower buoyant density (1.699 gm./cm.3) for
mitochondrial DNA from rat liver as compared with nuclear DNA (1.703 gm./
cm.3). In different species of animals, then, the two DNAs may be alike or dif-
ferent in density. The presently available data indicate that even related species
may differ in this regard. Even within the same species differences may occur in
the buoyant density of mitochondrial DNA, as Mounolou ct al. (1966) have shown
for "petite" mutants of yeast.
Our determinations of melting temperature showed the cytoplasmic DNA to
have a Tin of 86.8° C. as compared with 84.0° C. for the nuclear DNA. As noted
in the results, this corresponds to a guanine-cytosine content of 42% as compared
with 35% for the nuclear DNA. This is in good agreement with the values (44
and 34%, respectively) calculated from the buoyant densities.
The sedimentation experiments revealed two main components in the prepara-
tions of cytoplasmic DNA. As noted in the introduction, according to the analysis
of Vinograd et al. (1965), this, along with other properties, indicated that the sea
urchin cytoplasmic DNA might be in the form of circles which could, also, be of
twisted and relaxed forms sedimenting at different rates. Examination by electron
microscopy has corroborated the circular form. Further studies of this material
are in progress. The data reported here show the circles to be of rather uniform
size with perimeter close to 4.45 microns. This is near the values reported for
mitochondrial DNA of chick and mouse liver and beef heart by van Bruggen et al.
86 L. PIK6, A. TYLER AND J. VINOGRAD
(1966) and of mouse and rat liver (and several other tissues) by Sinclair and
Stevens (1966).
On the basis of the present evidence from our material, and that of others, it
would appear that all of the cytoplasmic DNA may be in circular form, and that
filaments may represent breakage due to preparative procedures. In fact the re-
laxed circles are considered (cf. Vinograd ct a/., 1965) to result from the occur-
rence of one or more single-strand scissions in the native material and, as prepara-
tive procedures improve, the filaments and extended circles become less frequent, as
the twisted circles increase proportionately.
On the basis of the present figures it can be estimated that there are only one
or two circles (of 4.45 microns perimeter) per mitochondrion of eggs of L. pictus.
Whether the circular units are genetically alike in all mitochondria is one of the
many interesting questions now open for investigation.
Addendum: While this paper was in press, an article appeared by M. M. K.
Nass (1966) who reports that there are 2 to 6 circular DNA molecules per mito-
chondrion in mouse fibroblasts ( L cells ) .
SUMMARY
1. Values of 8.26 ± 0.30 pg. (9.5 X haploid amount) for Lytechinus pictus
and 3.30 ± 0.25 pg. (4.3 x haploid amount) for Strongyloeentrotus purpnratus
have been obtained for the DNA content per egg of these sea urchins. The methods
involved repeated CsCl-buoyant density centrifugations, digestion of interfering
polysaccharide, and monitoring of the procedures with added radioactively labeled
DNA. The final determinations were made on material characterized by several
criteria as highly purified.
2. Mitochondria! (M) and yolk (Y) fractions of differentially centrifuged
homogenates of L. pictus eggs contain the bulk of the cytoplasmic DNA. It is
uncertain to what extent the smaller variable amount (11.5 to 26.5%) found in
the supernatant may be derived from breakdown of M- and Y-particles.
3. For distribution between M- and Y-fractions the best present value is con-
sidered to be about 3:1. Since yolk spherules are approximately one-third as
numerous as mitochondria, the amount of DNA is estimated to be the same per
particle, namely, 2.3 X 10~17 grams.
4. Evidence is presented that the cytoplasmic DNA of eggs of L. pictus and
S. purpuratus can serve as primer in a DNA-dependent RNA-synthesizing system
with approximately the same activity as nuclear DNA.
5. The cytoplasmic DNA of L. pictus eggs shows a buoyant density of 1.703
gm./cm.3 ac compared with 1.693 for the nuclear. A third nucleic acid band, equal
in amount to the nuclear, has been found at a density of 1.719. The amount of
nuclear DNA corresponds to the sperm (haploid) value.
6. Melting temperatures in standard saline-citrate are 84.0° C. for sperm and
gastrula DNA and 86.8° C. for whole-egg DNA, indicating a guanine-cytosine
content calculated from these values as 35% and 42^, respectively. These are
similar to the values (34^ and 44f/ ) calculated from the buoyant densities.
7. Electron microscopic observations of DNA prepared from mitochondria of
L. pictus show almost exclusively circles that measure about 4.45 microns in cir-
CYTOPLASMIC DNA IN SEA URCHIN EGGS 87
cumference. It is estimated that there are one or two such circular filaments of
double-stranded DNA per mitochondrion or yolk particle.
8. Centrifugation of egg DNA of L. pictiis in preformed CsCl gradients has
revealed two main components with sedimentation coefficients of ca. 23 S and 28S,
indicative of the two forms of circular DNA.
LITERATURE CITED
BALTUS, E., AND J. BRACKET, 1962. Le dosage de 1'acide desoxyribonucleique dans les oeufs de
Batraciens. Biochim. Biophys. Acta, 61: 157-163.
BALTUS, E., J. QUERTIER, A. FICQ AND J. BRACKET, 1965. Biochemical studies of nucleate
and anucleate fragments isolated from sea urchin eggs. A comparison between fer-
tilization and parthenogenetic activation. Biochim. Biophys. Acta, 95: 408-417.
BELL, P. R., AND K. MUHLETHALER, 1964. Evidence for the presence of deoxyribonucleic acid
in the organelles of the egg cells of Ptcridium aquilimtin. J. Mol. Biol., 8: 853-862.
BIBRING, T., J. BRACKET, F. S. GAETA AND F. GRAZIOSI, 1965. Some physical properties of
cytoplasmic deoxyribonucleic acid in unfertilized egges of Arbacia li.rula. Biochim.
Biophys. Acta. 108: 644-651.
BORST, P., AND G. J. C. M. RUTTENBERG, 1966. Renaturation of mitochondrial DNA. Biochim.
Biophys. Acta, 114: 645-647.
BRACKET, J., 1962. Nucleic acids in development. /. Cell. Comp. Pliysiol.. 60 ( Suppl. 1) : 1-18.
BRACKET, J., 1965. Le role des acides nucleiques dans la morphogenese. Ann. Biol., 4: 21-48.
BRACKET, J., AND A. FICQ, 1964. Detection cytochimique, au moyen d'actinomycine radioactive,
de 1'acide desoxyribonucleique (DNA) cytoplasmique des oeufs de Batraciens. C. R.
Acad.. Sci. Paris, 258: 6258-6260.
BRACKET, J., AND A. FICQ, 1965. Binding sites of 14C-actinomycin in amphibian oocytes and an
autoradiography technique for the detection of cytoplasmic DNA. E.vp. Cell Res.,
38: 153-159.
BRACKET, J., A. FICQ AND R. TENCER, 1963. Amino acid incorporation into proteins of nucleate
and anucleate fragment of sea urchin eggs : effect of parthenogenetic activation. Exp.
Cell Res., 32: 168-170.
BURGOS, M. H., 1955. The Feulgen reaction in mature unfertilized sea urchin eggs. Exp.
Cell Res., 9: 360-363.
GARDEN, G. A., S. ROSENKRANZ AND H. S. ROSENKRANZ, 1965. Deoxyribonucleic acids of
sperm, eggs and somatic cells of the sea urchin Arbacia pitnctnlata. Nature, 205:
1338-1339.
CHAMBERLIN, M., AND P. BERG, 1962. Deoxyribonucleic acid-directed synthesis of ribonucleic
acid by an enzyme from Eschcrichia coli. Proc. Nat. Acad. Sci., 48: 81-94.
CHANDLER, B., M. HAYASHI, M. N. HAYASHI AND S. SPIEGELMAN, 1964. Circularity of the
replicating form of a single-stranded DNA virus. Science, 143: 47-49.
CHEVREMONT, M., 1962. Localization and synthesis of deoxyribonucleic acids in the cytoplasm
of somatic cells of vertebrates: the role of mitochondria. Biochew. J., 85: 25P-26P.
CORNED, G., C. MOORE, D. R. SANADI, L. I. GROSSMAN AND J. MARMUR, 1966. Mitochondrial
DNA in yeast and some mammalian species. Science, 151: 687-689.
DAWID, I. B., 1965. Deoxyribonucleic acid in amphibian eggs. /. Mol. Biol., 12: 581-599.
DAWID, I. B., 1966. Evidence for the mitochondrial origin of frog egg cytoplasmic DNA.
Proc. Nat. Acad. Sci.. 56: 269-276.
DENNY, P. C., AND A. TYLER, 1964. Activation of protein biosynthesis in non-nucleate frag-
ments of sea urchin eggs. Biochcm. Biophys. Res. Comnnin., 14: 245-249.
DISCHE, VON Z., 1930. Uber einige neue charakteristische Farbreaktionen der Thymonuklein-
saure und eine Mikromethode zur Bestimmung derselben in tierischen Organen mit
Hilfe dieser Reaktionen. Mikrocliemie, 2 ( Neue Folge) : 4—32.
EBERIIARD, A., AND D. MAZIA, 1965. Conditions for extraction of cytoplasmic DNA from
embryos of a sea urchin. Biochcm. Biophys. Res. Coinnnin.. 21: 460-465.
ELSON, D., T. GUSTAFSON AND E. CHARGAFF, 1954. The nucleic acids of the sea urchin during
embryonic development. /. Biol. C/iem., 209: 285-294.
88 L. PIK6, A. TYLER AND J. VINOGRAD
GIBOU, A., AND S. GRANICK, 1964. Plastids and mitochondria: Inheritable systems. Science,
145 : 890-897.
GIUDICE, G., 1960. Incorporation of labeled amino acids into the proteins of the mitochondria
isolated from the unfertilized eggs and developmental stages of Paracentrotus liridits.
Exp. Cell Res., 21: 222-225.
GIUDICE, G., AND A. MONROY, 1958. Incorporation of S^-methionine in the proteins of the
mitochondria of developing and parthenogenetically activated sea urchin eggs. Ada
Embryologiac et Morphologiae Expcrimentalis, 2: 58-65.
GRANT, P., 1965. Informational molecules and embryonic development. In: The Biochemistry
of Animal Development. R. Weber, Editor. Academic Press, New York, pp. 488-593.
GROSS, P. R., AND G. H. COUSINEAU, 1963. Effects of actinomycin D on macromolecule syn-
thesis and early development in sea urchin eggs. Blochcin. Biophys. Res. Coniiinin.,
4: 321-326.
GROSS, P. R., AND G. H. COUSINEAU, 1964. Macromolecule synthesis and the influence of
actinomycin on early development. Exp. Cell Res., 33: 368-395.
HAGGIS, A. J., 1964. Quantitative determination of deoxyribonucleic acid in embryos and
unfertilized eggs of Rana pipicns. Dcvcl. BioL, 10: 358-377.
HINEGARDNER, R. T., 1961. The DNA content of isolated sea urchin egg nuclei. Exp. Cell
Res. ,25: 341-347.
HOFF-J0RGENSEN, E., 1954. Deoxynucleic acid in some gametes and embryos. /;;: Recent
Developments in Cell Physiology. J. A. Kitching, Editor. Colston Papers, 7: 79-90.
HOTTA, Y., AND A. BASSEL, 1965. Molecular size and circularity of DNA in cells of mammals
and higher plants. Proc. Nat. Acad. Sci., 53: 356-362.
HULTIN, T., 1961. Activation of ribosomes in sea urchin eggs in response to fertilization.
Exp. Cell Res.. 25: 405-417.
KALF, G. F., 1964. Deoxyribonucleic acid in mitochondria and its role in protein synthesis.
Biochemistry, 3: 1702-1706.
KAY, E. R. M., 1964. Direct isolation of deoxyribonucleic acid from cells. Nature, 202: 390-
391.
KLEINSCHMIDT, A. K., A. BURTON AND R. L. SINSHEIMER, 1963. Electron microscopy of the
replicative form of the DNA of the bacteriophage 0X174. Science, 142: 961.
KLEINSCHMIDT, A. K., S. J. IVASS, R. C. WILLIAMS AND C. A. KNIGHT, 1965. Cyclic DNA
of Shope papilloma virus. /. Mol. BioL, 13: 749-756.
KROON, A. M., 1963. Inhibitors of mitochondrial protein synthesis. Biochiin. Biophys. Acta,
76: 165-167.
LANZAVECCHIA, G., 1960. The yolk formation in the frog oocytes. Electron microscope studies.
Symposium on the Germ Cells and Earliest Stages of Development, Pallanza. S.
Ranzi, Editor. Inst. Internatl. d'Embryologie and Fondazione A. Baselli, Milan, pp.
61-74.
LANZAVECCHIA, G., 1965. Structure and demolition of yolk in Rana csciilenta L. /. Ultrastr.
Res., 12: 147-159.
LEIF, R. C., AND J. VINOGRAD, 1964. The distribution of buoyant density of human erythrocytes
in bovine albumin solutions. Proc. Nat. Acad. Sci., 51: 520-528.
LUCK, D. J. L., AND E. REICH, 1964. DNA in mitochondria of Ncurospora crassa. Proc.
Nat. Acad. Sci., 52: 931-938.
MARMUR, J., AND P. DOTY, 1962. Determination of the base composition of deoxyribonucleic
acid from its thermal denaturation temperature. /. Mol. BioL, 5: 109-118.
MARSHAK, A., AND C. MARSHAK, 1953. Desoxyribonucleic acid in Arbacia eggs. Exp. Cell
Res., 5 : 288-300.
MEZGER-FREED, L., 1963. Evidence for synthesis of deoxyribonucleic acid in enucleated Rana
pipicns eggs. /. Cell BioL, 18: 471-475.
MONROY, A., 1965. Chemistry and Physiology of Fertilization. Holt, Rinehart and Winston,
New York.
MOUNOLOU, J. C., H. JAKOB AND P. P. SLONIMSKI, 1966. Mitochondrial DNA from yeast
"petite" mutants: Specific changes in buoyant density corresponding to different cyto-
plasmic mutations. Bioclicin. Biophys. Res. Coinnnin.. 24: 218-224.
NAKANO, E., AND A. MONROY, 1958. Incorporation of S35-methionine in the cell fractions of
sea urchin eggs and embryos. Exp. Cell Res., 14: 236-243.
CYTOPLASMIC DNA IN SEA URCHIN EGGS
MASS, M. M. K., 1966. The circularity of mitochondrial DNA. Proc. Nat. Acad. Sci., 56:
1215-1222.
NASS, M. M. K.. S. XASS .\\n R. A. AKZKLIUS, 1065. The general occurrence of mitochondrial
DXA. E.rp. Cell Res.. 37: 516-539.
NASS, S.. AXD M. M. K. NASS, 1963. Intramitochondrial fibers with DNA characteristics.
II. Enzymatic and other hydrolytic treatments. /. Cell Bid., 19: 613-629.
OGUR, E., AXD G. ROSEN, 1950. The nucleic acids of plant tissues. I. The extraction and esti-
mation of desoxypentose nucleic acid and pentose nucleic acid. Arch. Biochcm.. 25:
262-276.
PIKO, L., AXD A. TYLER, 1965. Deoxyribonucleic acid content of unfertilized sea urchin eggs.
Amcr. Zool., S: 636.
RABINOWITZ, M., J. SINCLAIR, L. DE SALLE, R. HASELKORX AXD H. SWIFT, 1965. Isolation of
deoxyribonucleic acid from mitochondria of chick embryo heart and liver. Proc. Nut.
Acad. Sci.. 53: 1126-1133.
Ris, H., 1962. Ultrastructure of certain self-dependent cytoplasmic organelles. Proceedings of
the 5th International Congress of Electron Microscopy, Vol. 2. Academic Press, New
York, pp. xx-1.
ROLLER, A., 1963. fitude de 1'acide desoxyribonucleique des oeufs de Batraciens : teneur et
localisation. Arch. Intern. Pliysiol. Bioch., 71: 139-140.
RUDZINSKA, M. M., P. A. D'ALESANDRO AND W. TRACER, 1964. The fine structure of Lcish-
inania donorani and the role of the kinetoplast in the Leishmania-leptomonad transfor-
mation. /. ProtozooL, 11: 166-191.
SCHATZ, G., E. HALSBRUXXER AXD H. TUPPY, 1964a. Deoxyribonucleic acid associated with
yeast mitochondria. Biochcm. Biophys. Res. Coiiunun., 15: 127-132.
SCHATZ, G., E. HALSBRUXNER AXD H. TUPPY, 1964b. Intramitochondriale Deoxyribonuklein-
saure in Saugetierzellen. Monatshefte filr Chemic, 95: 1135-1140.
SCHILDKRAUT, C. L., AND S. LiFSEN, 1965. Dependence of the melting temperature of DNA on
salt concentration. Biopolymers, 3: 195-208.
SCHILDKRAUT, C. L., J. MARMUR AND P. DOTY, 1962. Determination of base composition of
deoxyribonucleic acid from its buoyant density in CsCl. /. Mol. Biol., 4: 430-443.
SCHNEIDER, W. C., AND E. L. KUFF, 1965. The isolation and some properties of rat liver mito-
chondrial deoxyribonucleic acid. Proc. Nat. Acad. Sci., 54: 1650-1658.
SEGOVIA, Z. M. M., F. SOKOL, I. L. GRAVES AND W. W. ACKERMANN, 1965. Some properties of
nucleic acids extracted with phenol. Biochiin. Biophys. Acta, 95: 329-340.
SHAVER, J. R., 1956. Mitochondrial populations during development of the sea urchin. E.vp.
Cell Res.. 11: 548-559.
SHAVER, J., 1957. Some observations on cytoplasmic particles in early echinoderm develop-
ment. In: The Beginnings of Embryonic Development. A. Tyler, C. B. Metz and
R. C. von Borstel, Editors. American Association for the Advancement of Science,
Washington, D. C., pp. 263-290.
SHMERLIXG, ZH. G., 1965. Isolation and properties of DNA of sturgeon oocytes. Biokhiiniya,
30: 113-123 (in Russian) ; translation 30: 96-104.
SINCLAIR, J. H., AND B. J. STEVENS, 1966. Circular DNA filaments from mouse mitochondria.
Proc. Nat. Acad. Sci., 56: 508-514.
SRIVASTAVA, M. D. L., 1965. Cytoplasmic inclusions in oogenesis. Intermit. Rer. C\tol., 18:
73-98.
STEINERT, G., H. FIRKET AND M. STEINERT, 1958. Synthese d'acide desoxyribonucleique dans
le corps parabasal de Trypanosoma mega. E.rp. Cell Res., 15: 632-635.
SUGINO, Y., N. SUGIXO, R. OKAZAKI AXD T. OKAZAKI, 1960. Studies on deoxynucleoside
compounds. I. A modified microbioassay method and its application to sea urchin
eggs and several other materials. Biochiin. Biophys. Acta, 40: 417-424.
SUYAMA, Y., AND J. R. PREER, JR., 1965. Mitochondrial DNA from protozoa. Genetics, 52:
1051-1058.
SWIFT, H., 1965. Nucleic acids of mitochondria and chloroplasts. Amcr. Naturalist, 99: 201-
227.
SWIFT, H., N. KISLEV AXD L. BOGORAD, 1964. Evidence for DNA and RNA in mitochondria
and chloroplasts. /. Cell Biol., 23 : 91 A.
•'() L. PIK6, A. TYLER AND J. VINOGRAD
TYI.KK, A., 1963. The manipulations of macromolecular substances during fertili/ation and
early development of animal eggs, .liner. ZooL, 3: 109-126.
TYLER, A., 1965. The biology and chemistry of fertilization, .liner. Naturalist, 99: 309-334.
TYLER, A., 1966. Incorporation of aniino acids into protein by artificially activated non-nucleate
fragments of sea urchin eggs. liiol. Hull., 130 : 450-461.
TYLER, A., AND B. S. TYLER, 1966a. The gametes: Some procedures and properties. In:
Physiology of Echinodermata, Ch. 27. R. A. Boolootian, Editor. John Wiley and
Sons, Inc., New York, pp. 639-682.
TYLER, A., AND B. S. TYLER, 1966b. Physiology of fertilization and early development. In:
Physiology of Echinodermata, Ch. 28. R. A. Boolootian, Editor. John Wiley and
Sons, Inc., New York, pp. 683-741.
VAN BRUGGEN, E. F. J., P. BORST, G. J. C. M. RUTTENBERG, M. GRUBER AND A. M. KROON,
1966. Circular mitochondrial DNA. Biochlm. Biophys. Acta, 119: 437-439.
VINOGRAD, J., AND J. HEARST, 1962. Equilibrium sedimentation of macromolecules and viruses
in a density gradient. Fortschr. Che HI. Org. Naturstoffe, 22: 372-422. Springer-
Verlag, Vienna.
VINOGRAD, J., AND J. LEBOWITZ, 1966. Physical and topological properties of circular DNA.
J. Gen. Physiol.,49: 103-125.
VINOGRAD, J., J. LEBOWITZ, R. RADLOFF, R. WATSON AND P. LAIPIS, 1965. The twisted cir-
cular form of polyoma viral DNA. Proc. Nat. Acad. Sci.. 53: 1104-1111.
WARD, R. T., 1962. The origin of protein and fatty yolk in Rana pipicns. II. Electron micro-
scopical and cytochemical observations of young and mature oocytes. /. Cell BioL, 14:
309-341.
WEIL, R., AND J. VINOGRAD, 1963. The cyclic helix and cyclic coil forms of polyoma viral
DNA. Proc. Nat. Acad. Sci. 50: 730-738.
WEISS, S. B., AND L. GLADSTONE, 1959. A mammalian system for the incorporation of cytidine
triphosphate into ribonucleic acid. /. Amcr. Chcm. Soc.. 81: 4118-4119.
WHITELEY, A. H., AND F. BALTZER, 1958. Development, respiratory rate and content of desoxy-
ribonucleic acid in the hybrid Paracentrotus$X Arbaciad. Pubhl. Stas. Zool. Napoli,
30 : 402-457.
OBSERVATIONS ON THE GENITAL SEGMENTS OF
SPIRORBIS (POLYCHAETA) l
HERBERT E. POTSWALD -
Department of Zoology, University of }\'ashington. Seattle, \\'ashm<iton 98105
The serpulid genus Spirorbis has attracted the attention of numerous investi-
gators for more than a century. Pagenstecher (1863) noted that Sp. pagcn-
stccheri exhihited hrood protection and was hermaphroditic. In subsequent ac-
counts (Agassiz. 1866; Fewkes, 1885; Schively, 1897; Bush. 1904; Elsler, 1907;
zur Loye, 1908; Borg, 1917; and others) Pagenstecher's observations have been
confirmed for all species examined. To date, the most comprehensive study on the
morphology and reproductive biology of Spirorbis is that of Bergan (1953a).
Bergan's study, although contributing valuable information, leaves many questions
unanswered and fails to give an adequate description of the genital segments. The
observations presented here were made in conjunction with an embryological study
of Spirorbis and serve to add new information concerning the morphology and
nature of the genital segments in this complex and enigmatic genus.
MATERIALS AND METHODS
Collection and maintenance of adult animals
Spirorbis (Laeospira) inb'rchi Levinsen, Sp. (Parade.riospira) Z'itreus Fabri-
cius. and Sp. ( De.riospira ) spirillum Linne were collected intertidally on San Juan
Island, Washington, periodically throughout the year, from 1960 to 1963. Spirorbis
spirillum was also frequently collected on hydroid colonies (Abietinaria sp.)
dredged 10-23 fathoms off San Juan Island. Spirorbis (Protolaeospira) ambi-
lateralis Pixell was collected by dredging off San Juan Island and was found most
frequently on Modiolus nwdiolus, Bahtnus nubilis, and Chlamys sp., and often in
association with Sp. ritreus. The system of classification followed is that first
used by Caullery and Mesnil (1897) and later adopted by Fauvel (1927). The
identities of the species reported on here were determined from descriptions given
by Bush (1904), Pixell (1912), Fauvel (1927), Berkeley and Berkeley (1952).
Bergan (19531)), and Pettibone (1954). In an examination of the Washington
species, the procedure was first to prepare a species description, as complete as
possible, before going to the literature to make comparisons. Of the several hun-
dreds of specimens examined, Sp. uiorchi and Sp. ambilateralis have always been
found to be sinistral and Sp. spirillum dextral. On the west side of San Juan
Island, Washington, however, there are extensive populations of dextral and sinis-
tral Sp. vitrcus. The sinistral form occurs together with the dextral form in about
1 Supported, in part, by precloctoral fellowship 1-F1-GM-20, 593-01 from U.S.P.H.S.
- Present address : Department of Zoology, University of Massachusetts, Amherst, Massa-
chusetts 01002.
91
92 HERBERT E. POTSWALD
equal numbers. They are exactly alike in all diagnostic characters except the
direction of coil. Observations on isolated brooding adults indicate that this is a
case of true genetic polymorphism (Potswald, 1965).
The animals were kept in the circulating sea water system at the Friday Harbor
Laboratories, or were brought to the Seattle campus where they were kept in the
Zoology Department's 10° C. cold room. In Seattle, no attempt was made to
provide food ; however, the animals remained in good condition for several months
at a time, and brooding individuals were always available.
Adult animals were observed both within their tubes and removed from their
tubes. In the latter case, animals were removed from their substratum by means
of a razor blade and then removed from their tubes by chipping away the cal-
careous secretion with dissecting needles. This is a relatively easy task except
for Sp. vitrens, which has an extremely hard tube.
Microtechnique
Adult worms removed from their tubes were fixed in a variety of fixatives
including : Bouin's fluid, Helly's with and without post-chroming, buffered formalin,
Flemming's with and without acetic acid, and Carnoy without chloroform. Fixed
material was dehydrated through ethyl alcohol and tertiary butyl alcohol, and
embedded in a mixture of 300 gm. of Fisher's Tissuemat (M.P. 60-62° C.) and
45 gm of dry piccolyte. This is the mixture recommended by Cloney (1961) but
without beeswax. Blocks, chilled in ice-water, were sectioned 5-6 microns at
room temperature.
A number of stains were used including Heidenhain's iron haematoxylin, Hei-
denhain's Azan, and Harris' or Ehrlich's haematoxylin with eosin counterstain.
Feulgen and PAS methods were followed according to McManus and Mowry
(1960).
In addition to routine paraffin technique, material was fixed in cold buffered
osmium tetroxide (Bennett and Luft, 1959) and embedded in Epon 812 according
to the method of Luft (1961). Thin sections, ^-1 micron, were cut on a Porter-
Blum ultra-microtome, using glass knives. After affixing to glass slides, the sec-
tions were stained with Richardson's stain (Richardson et al., 1960) for study with
the light microscope.
DESCRIPTION OF GENITAL SEGMENTS
Bergan (1953a), from his study of Sp. borealis, concluded that Spirorbis is
a simultaneous hermaphrodite and not protandrous as suggested by Hempelmann
(1931). Although providing histological evidence for his conclusion, Bergan fails
to give an adequate description of the genital segments. Aside from Franzen's
study (1956, 1958) on late stages of spermiogenesis and Dasgupta and Austin's
(1960) examination of chromosome numbers in spermatocytes, little information
concerning the gametes and their development in Spirorbis is available.
The observations presented here are based, for the most part, on a study of
Sp. morchi. Only where differences in detail have been found, will the other three
species be mentioned.
GENITAL SEGMENTS OF SPIRORRIS 93
Female Segments
Normal arrangement and anomalies
Germ cell proliferation is restricted to the abdominal or secondary segments;
the thoracic or primary segments never serve in this capacity. In a mature adult
every abdominal segment contains germ cells in various stages of formation. Each
genital segment contains a large coelomic space and is separated from its neighbors
by complete septa. In Sp. inarch i the first two abdominal segments are female
and the remaining abdominal segments male (Fig. 2). Sp. ambilateralis presents
the same arrangement, but the left halves of segments one and two are generally
larger than the right halves of the same segments. Of the dextral species, Sp.
vitrcus is like Sp. morcJii, while Sp. spirillum presents quite a different picture. In
local populations of Sp. spirillum, either the first three segments or first 3| segments
are female ; the left half of the fourth segment in the latter condition is male. The
anomalous condition is the most prevalent. According to Bergan (1953a), speci-
mens of Sp. spirillum collected in Oslofjord exhibit lateral asymmetry in sex
differentiation similar to that described above ; therefore, it would seem that such
anomalous sex differentiation is common in Sp. spirillum.
Although one might expect to find anomalies in a simultaneous hermaphrodite
such as Spirorbis, sex differentiation within the genus, with the notable exception
of Sp. spirillum, seems to be under rigid control and a specific arrangement of
female and male segments prevails for a given species. After examination of
several hundred specimens of adult Sp. morchi, only a few anomalies have been
observed. Lateral asymmetry was found in one animal in which the left half of the
second segment was male and the right half female. One case was observed in
which the first three abdominal segments were female instead of the first two as is
the normal condition. Finally, in a few instances, individuals were found to have
sperm and oocytes developing together in the same segment. The latter condition
generally occurs in the posterior male segments where the infrequent oocytes have
never been observed to enter into vitellogenesis. Only in two individuals have
oocytes and sperm been observed to develop together in the second abdominal seg-
ment, between a purely female and a male segment (Fig. 3). Oocytes developing
in the posterior part of the achaetous zone, as described by Bergan (1953a) for
Sp. boreal is, have not been found.
Development of the primary oocyte
The gonad is a discrete and constant organ composed of clumps of cells ar-
ranged in two retroperitoneal rows, mesial to the ventral nerve cords, and running
the length of the abdominal segments (Figs. 4, 5, 6, 7). In sexually mature adults,
the gonad is always larger in the female segments than in the male segments where
it is greatly compressed against the ventral ectoderm.
An examination of the gonad in the female segments reveals the presence of
a number of primordial germ cells in interphase. While in interphase, these are
the most distinctive cells in the adult body. The cells have almost spherical nuclei
5 to 6 microns in diameter, and the cytoplasm is reduced to a mere envelope. Scat-
tered around the periphery of the nucleus, in a predictable pattern, are spherical
94 HERBERT E. POTSWALD
clumps of Feulgen-positive chromatin ; the center of the nucleus resists staining and
appears transparent ( Fig. la). Counts, made from thick sections, indicate that
there are ten chromatin clumps per nucleus, this numher apparently being constant.
A nucleolus is absent. The cytoplasm is so reduced that little detail can be re-
solved in it with light microscopy.
Since the gonad is the site of constant proliferation, the primordial germ cells
are generally observed in various stages of mitosis. As nuclei enter prophase, the
thread-like projections appear to radiate out from the individual chromatin spheres
and the "unraveling" continues until the nucleus is completely occupied by typical
prophase chromosomes (Fig. la). Metaphase, anaphase, and telophase follow
in an orderly fashion. The telophase nuclei of the resulting daughter cells are
small and vesicular, with the diameter of the nuclei measuring about 2.5 microns
and that of the entire cell about 3.2 microns (Fig. Ib). It is assumed that the
mitotic divisions give rise either to more primordial germ cells which resume the
characteristic interphase condition or to a morphologically different cell type, the
early primary oocyte. The term "primordial germ cell" has been used for the
proliferating cell since a distinct gonial stage has not been recognized.
la Ib
FIGURE 1. Sf>. morchi: (la) section through primordial germ cells in interphase (I), and
prophase (Pr) ; (Ib) section through four primordial germ cells having just completed telo-
phase (T), and one primordial germ cell undergoing telophase reconstruction (TR). Drawing
made from a photomicrograph. Paraffin ; Feulgen. (1880X)
Early primary oocytes have a nuclear diameter of about 6 microns and a
cytoplasmic component somewhat larger than that found in the primordial germ
cell. The nucleus contains a single nucleolus and a diffuse chromatin network
(Figs. 4, 5). Premeiotic stages of prophase are passed within the gonad proper.
The delicate chromatin net of the interphase nucleus becomes coarser and more
irregular and, with further condensation, chromatin clumps appear not unlike those
found in the primordial germ cell. The nucleus has now become elliptical in shape
and measures about 7.5 microns in length (Fig. 5). Following condensation, the
chromatin bodies become frayed and give rise to delicate chromosomal threads.
As concerns succeeding stages of premeiotic prophase, only leptotene and early
pachytene, still in "bouquet" arrangement, have been observed (Figs. 4, 6). Scat-
tered throughout the gonad are a number of cells about the same size as oocytes in
meiotic prophase ; however, unlike the normal oocytes, these cells have structureless
nuclei often irregular in outline ( Fig. 4 ) . The role of the latter cells in normal
development is not clear, but it is possible that they are degenerating germ cells.
With the conclusion of the events of premeiotic prophase, the nucleus again
enters the diffuse or "confused" state and regains its spherical shape. A distinct
GENITAL SECMKXTS OF SPIRORBIS
nncleolus reappears along with several irregularly shaped pieces of chromatin ; the
basophilic cytoplasmic portion has increased and the diameter of the cell is now
about 11 microns. Oocytes of the latter size-class are found beneath the bulging
peritoneal covering of the gonad and at the base of the septa. As the oocytes
increase in size, they invade the septum and a progressive series can be found
extending dorsally from the gonad. The oocytes are located between the two thin
epithelial layers making up the septum (Fig. 7). \Yhile within the septum, the
cytoplasm of the oocytes is strongly basophilic, and little or no yolk deposition
occurs. In nuclei attaining a diameter of 13 microns, the single nucleolus starts
to fragment into a number of small Feulgen-negative bodies. Cells having this
nuclear diameter are about 22 microns in diameter and are starting to erupt into
the coelom. Their cytoplasm is filled with coarse membrane, a few scattered pro-
teid yolk granules, and numerous, small, spherical mitochondria (Fig. 8). In thin
sections of osmium-fixed material, small rod-like bodies about one micron long,
staining intensely with Richardson's stain, are found in the cytoplasm around the
nucleus. The bodies are often crescent-shaped and may represent Golgi material
(Fig. 8).
The bulk of vitellogenesis takes place in oocytes which have erupted through
the septum, and are floating free within the spacious coelom. There are no nurse
cells, as such, nor follicle cells associated with the oocytes. As cytoplasmic mass
increases, nuclei, which also steadily increase in size, become more eccentric in
position. In cells measuring 39 X 33 microns and having a nuclear diameter of
about 22 microns, yolk granules begin to arise in a rather localized area within
the cytoplasm. At the start of vitellogenesis, and even at its close, yolk granules
give only a slight PAS reaction which is not qualitatively affected by incubation
in diastase. Associated with the yolk granules are several lamellar stacks of
membrane. The membrane stacks are very sensitive to the fixative employed
and have been observed only in material fixed in either buffered osmium or Kelly's
fluid followed by post-chroming. As oocytes approach a diameter of about 50
microns, the lamellar stacks become concentric, thereby enclosing an internum.
Clusters of mitochondria are associated with the concentric bodies, and both pro-
teid and lipid yolk granules occur within the internum ( Fig. 9 ) . The description
given here corresponds in many respects to those given by several authors for
structures termed "yolk nuclei," which have been observed in various groups of
animals (see Raven, 1961, for survey). As vitellogenesis proceeds and the
cytoplasm becomes filled with yolk granules, the concentric structures disappear;
their fate has not been followed. There is little visible differentiation of the cortex
below the PAS-positive vitelline membrane, cortical granules apparently being
absent.
Maximum growth has been attained, and the primary oocyte is ready to be
spawned, when it measures about 165 microns X 132 microns. From the start of
vitellogenesis, yolk granules have increased in diameter from a fraction of a
micron, in the case of proteid yolk, to about 8 microns. The germinal vesicle has
become very wrinkled in appearance and has an average measurement of 26 X 13
microns. Nucleolar fragments, so abundant during vitellogenesis, have become
reduced both in number and in size. The germinal vesicle apparently does not
break down until after spawning has occurred. Animals ready to spawn contain
HERBERT E. POTSWALD
Cl
- -
-
GENITAL SEGMENTS OF SPIRORBIS 9?
two size-classes of oocytes in their coeloms : full grown primary oocytes and early
primary oocytes still within or attached to the septa. Animals have been observed
to spawn within 12 hours after releasing larvae, and since it takes about 30 days
from fertilization to larval release at 12° C, it is assumed that vitellogenesis occu-
pies about the same period of time.
From the present study, it is difficult to come to any definite conclusions as
to the origin of yolk. Nucleolar fragmentation, concentric lamellae, and possible
Golgi material have been mentioned, and it is conceivable that they all participate
in yolk formation. At the onset of vitellogenesis, the peritoneal cells lining the
coelomic cavity start to accumulate large droplets, both lipoidal and proteid in
nature, within their cytoplasm. The droplets accumulate to such a degree as to
cause the ordinarily flattened peritoneal cells to bulge into the coelomic cavity (Fig.
10). Apical portions of cells actually bud off and become free in the coelom.
Towards the end of vitellogenesis, the peritoneal cells again become flattened and
relatively devoid of inclusions. There is no evidence that the oocytes are phago-
cytic ; however, correlation between the onset and decline of the storage phenomenon
in peritoneal cells with that of vitellogenesis would seem to suggest that transfer
of material in some form takes place. Conceivably, such transfer could be in the
form of high molecular weight compounds. Finally, it should be mentioned that
it is not uncommon for one or two oocytes per segment to disintegrate midway
through vitellogenesis. It is not known what role, if any, this phenomenon plays
during normal development of the primary oocyte.
The above outline of events leading to the development of the primary oocyte
in Sf>. inorchi also holds true, in the main, for the other three species studied. In
Sp. vitreus, the behavior of the nucleolus is quite different from that in Sp. morchi.
FIGURE 2. Sp. inorchi: parasagittal section through the abdomen of a typical adult, show-
ing that the first two abdominal segments are female and the remaining abdominal segments are
male. Epon ; Richardson's stain. (110X)
FIGURE 3. Sp. morchi: parasagittal section through the second abdominal segment of an
adult, showing the simultaneous development of oocytes and spermatids within the same seg-
ment. Epon ; Richardson's stain. (700X)
FIGURE 4. Sp. morchi: frontal section through the gonad in a female segment showing
presumably degenerating oocytes (DO?), early oocytes (EO), and leptotene (L). Epon;
Richardson's stain. ( 800 X )
FIGURE 5. Sp. morchi: frontal section through the gonad of a female segment showing
early oocytes (EO) and a premeiotic oocyte (Pm) containing condensed, peripheral chromatin.
Epon ; Richardson's stain. (800 X )
FIGURE 6. Sp. morchi: cross-section through the gonad of a female segment showing an
oocyte in pachytene (Pa). Note the discrete nature of the gonad and the thin ventral epidermis
(E). Epon ; Richardson's stain. (800 X)
FIGURE 7. Sp. morchi: oblique cross-section through the achaetous region and first ab-
dominal segment showing a progressive series of oocytes extending dor sally from the gonad (G)
within the septum (S). Also illustrated are the ventral nerve cord (VN) and ventral ciliated
peritoneum (CP). Paraffin ; haematoxylin-eosin. (320X)
FIGURE 8. Sp. morchi: oblique cross-section through two adjacent female segments show-
ing a series of oocytes which have erupted through the septum into the coelom. Note the large
amount of coarse membrane present in the cytoplasm, the nucleolar fragments (NF), and pos-
sible Golgi material (Go). Only a few proteid yolk granules (YG) are present. Epon;
Richardson's stain. (800 X)
FIGURE 9. Sp. morchi: cross-section through an oocyte showing the eccentric position of
the germinal vesicle (GV) and concentric lamellae (CL) of membrane. Note absence of fol-
licle cells. Epon; Richardson's stain. (800 X)
98
HERBERT E. POTSWALD
GV
FIGURES 10-17.
GENITAL SEGMENTS OF SPIRORBIS (->(>
The nucleolus in early oocytes is homogeneous, hut just prior to vitellogenesis it
takes on the staining characteristics of an amphinucleolus. The cortex is acido-
philic and the medulla basophilic. As yolk granules hegin to appear, the cortex
separates into two spherical hodies which retain their identity throughout most of
the vitellogenetic period. The medulla also divides and starts to vacuolate and
extrude nucleolar material into the nucleoplasm (Fig. 11). Nucleolar vacuolation
also occurs in Sp. spirillum. The nucleolar products in both species can he observed
in germinal vesicles about to be spawned. At the level of light microscopy, there
is no evidence for nucleolar extrusion into the cytoplasm.
Bergan (1953a), although he does not describe them, claims to have found
abdominal nephridia in the female segments of the four species he examined. He
is of the opinion that the dimensions of the nephridia are such that they could
serve as genital ducts. It has not been possible to confirm Bergan's observations
on local species used in the present study. In Sp. morchi, for example, the ventral
peritoneum of the female segments is strongly ciliated but there is no duct arrange-
ment (Fig. 7). This ciliated patch of peritoneum probably represents a remnant
of the coelomostome, the habit of shedding oocytes through a coelomoduct having
been abandoned. Such remnants as a ciliated flap or patch on the peritoneum are
common in polychaetes which release gametes by rupture of the body wall or gut
(Dales, 1963). In Sf>. morchi, gravid female segments become greatly distended,
the ventral body wall measuring about 8 microns in thickness.
Male Segments
Spermatocyte development:
The mitotic events associated with primordial germ cell proliferation in the
male gonad are the same as in the female gonad. As in the female gonad, pre-
FIGURE 10. Sp. morchi: parasagittal section through a female segment showing the large
accumulation of lipid droplets within the peritoneal cells (P) and the thin epidermis (E) of
the body wall. Epon ; Richardson's stain. (630 X)
FIGURE 11. Sp. ritrcns: section through a primary oocyte showing the germinal vesicle
(GV) and nucleolar fragments undergoing nucleolar vacuolation (NB). Paraffin; haema-
toxylin-eosin. (1600X)
FIGURE 12. Sp. morchi: sagittal section through two adjacent male segments showing a
cluster of primary spermatocytes (PS) and a secondary spermatocyte (SS). Note the spheri-
cal Golgi material (Go) and mitochondria (M). Epon; Richardson's stain. (1600 X)
FIGURE 13. Sp. vitreus: primary spermatocytes showing spherical Golgi material (Go)
and cluster of mitochondria (M) at the opposite pole. Living material; phase-contrast.
(1600X)
FIGURE 14. Sp. morchi: sagittal section through two adjacent male segments showing pri-
mary spermatocytes (PS) just prior to first meiotic metaphase and secondary spermatocytes
(SS) just prior to second meiotic metaphase. Note inclusions in septal peritoneum (SP).
Epon ; Richardson's stain. ( 1600 X)
FIGURE 15. Sp. morchi: section through a male segment showing primary spermatocytes
(PS) in first meiotic metaphase and spermatids in the "complete ring stage" (CR). Epon;
Richardson's stain. (1260 X)
FIGURE 16. Sp. morchi: section through a male segment showing secondary spermato-
cytes (SS) in second meiotic metaphase. Epon; Richardson's stain. (1260 X)
FIGURE 17. Sp. morchi: section through a male segment showing early spermatids (ESp)
and late spermatids (LSp). Note the granular mitochondria (GM) concentrated at one pole
in the early spermatids and fusion of mitochondria to form four mitochondrial spheres (MS) in
the late spermatids. The four vacuoles (V) found in the sloughing cytoplasm of the late
spermatids alternate with the four mitochondrial spheres. Epon; Richardson's stain. (1925 X)
100 HERBERT E. POTSWALD
meiotic stages of prophase are difficult to find, only leptotene having been observed
on the male side. Also, as in the female gonad, there are a number of cells which
have the appearance of cells undergoing degeneration ; however, as in the case of
1 1 if female side, there is no clear evidence that these cells actually degenerate. The
primary spermatocyte erupts through the peritoneum of the septum and enters the
coelom with the nucleus in the "diffuse state." The fact that growth of the pri-
mary oocyte is initiated in the gonad probably accounts for the disparity in size of
gonad between the two sexes.
In the coelom, the primary spermatocyte reaches a diameter of about 8 microns.
The nucleus contains a fine chromatin network and a single nucleolus. Thread-
like mitochondria are scattered throughout the cytoplasm but are especially concen-
trated at one pole of the cell. At the opposite pole there is a spherical structure
which, in osmium-fixed material, is partially or completely surrounded by a sheath
of dark-staining material (Fig. 12). The same structure has also been observed
in living primary spermatocytes by means of phase-contrast (Fig. 13). It is
assumed that the spherical body is made up of Golgi material. Just prior to meta-
phase I, the chromosomes appear as distinct bivalents and the spherical body breaks
up (Fig. 14). The diplotene chromosomes contract greatly as they enter diakinesis
but unequivocal chiasma formations have not been observed. With the conclusion
of diakinesis, the chromosomes become arranged in metaphase (Fig. 15) and the
first meiotic division proceeds. Counts made from sectioned material and aceto-
orcein squashes indicate the haploid number to be about 10. Dasgupta and Austin
(1960) in five species of Spirorbis (Sp. borealis, Sp. coralHnae, Sp. pagenstechcri.
Sp. spirillum, and Sp. tridentatus) have found a uniform count of 2n = 20.
Secondary spermatocytes, resulting from the first meiotic division, enter into
an interphase condition. The single nucleolus and delicate chromatin network
reappear in the nucleus and the entire diameter of the cell averages 5.7 microns.
Thread-like mitochondria remain more or less concentrated at one pole while at
the opposite pole dense-staining bodies reaggregate to form the spherical body
which is assumed to be of Golgi material (Fig. 12). With the onset of prophase,
the nucleolus disappears and prophase chromosomes appear (Fig. 14). After con-
tracting to some extent, the chromosomes become arranged along the metaphase
plane and the second meiotic division occurs (Fig. 16).
Sp ermio gene sis
The early spermatid resulting from telophase II has a total diameter of 3.3
microns. A delicate chromatin net fills the nucleus and granular mitochondria
are concentrated at one pole within the scanty cytoplasm (Fig. 17). With the
initiation of spermiogenesis, the chromatin goes through a series of complex
changes. At the beginning of the cycle, the chromatin pulls away from the center
of the nucleus and starts to condense around the periphery just under the nuclear
membrane. Further peripheral condensation gives rise to what may be referred
to as the "interrupted ring stage." The mitochondria in the latter stage are still
granular in appearance and concentrated at one pole (Fig. 18). The next stage in
the developmental sequence is characterized by completion of the chromatin ring
and is referred to here as the "complete ring stage." At this stage, the mitochon-
GENITAL SEGMENTS OF SPIRORBIS 101
dri;i have fused into four spheres of uniform size which surround the point from
which the tail filament emerges (Fig. 15).
In material fixed in buffered osmium and stained with Richardson's stain, the
chromatin ring stains very intensely while the center of the nucleus stains hardly
at all. \Yith further development, the chromatin ring disappears except for a
dark-staining apical rim, and the nucleus becomes homogeneous. It has not been
possible to resolve the events involved in acrosome formation ; however, in the last
mentioned stage, a dark-staining body is found in the cytoplasm at one side of the
developing sperm head and may represent Golgi material (Fig. 19). The dark-
staining apical rim might be interpreted to be the deposited acrosome. After be-
coming homogeneous, the nucleus takes on a conical shape and starts to elongate
(Fig. 20). As the head continues to elongate, the cytoplasm moves backward so
as to encompass more and more of the tail filament. Four clear vacuoles appear
in the apical portion of the sloughing cytoplasm just above the mitochondrial
spheres. A cross-section taken through this region of the developing sperm reveals
that the vacuoles alternate with the four mitochondrial spheres (Fig. 17). The
vacuoles occur in all four species examined and are undoubtedly the "neutral red
vacuoles" described by Franzen (1956). The significance of the vacuoles is not
known.
Late spermatids in various stages of cytoplasm elimination form large plates
when artificially released into the sea water. This seems to be due to a mutual
stickiness and not to the presence of a cytophore or nurse cell arrangement ; at least
if such a system is present, it is not resolved at the light level. In the last phase
of cytoplasm elimination, the four mitochondrial spheres move posteriorly along
the tail filament to form the middle piece. In mature sperm, the middle piece
becomes homogeneous. The vacuoles are lost with the residual cytoplasm.
The morphologically mature sperm of Sp. uwrchi has a cylindrical head capped
with a distinctly pointed acrosome and has a total length of about 45 microns.
Sharply delimited from the head is the middle piece which is about three times the
length of the head. The tail filament is of the ordinary type and is just a little
more than twice the length of the middle piece (Fig. 21). In Sp. ainbilatcralis
the head of the sperm is also cylindrical hut the acrosome is neither as distinct nor
as pointed as in Sp. morchl. The middle piece is H times the length of the head
and the tail filament has a length equal to the total length of a Sp. uiorchi sperm
(Fig. 22). The average total length of a Sp. atnbilatcralis sperm is 57 microns.
In the two dextral species, Sp. vitreus and Sp. spirillum, the sperm are morpho-
logically quite different from those of the two sinistral species just described. In
both the dextral form and polymorphic sinistral form of Sp. I'itrcus, the sperm
head has the shape of a long, slightly bent, pointed cone, a middle piece slightly
less in length but of about the same thickness as the head, and a tail filament with
a length a little more than twice the combined length of head and middle piece
(Fig. 23). The sperm head of Sp. spirillum is longer, more sharply bent, and
more pointed than that of Sp. ritrcus. The middle piece is similar hut the tail
filament is considerably longer, having a length equal to that of an entire Sp.
ritrcns sperm (Fig. 24). In neither species is the acrosome clearly delimited from
the head. The average total length of a Sp. vitrcns sperm is 42.5 microns, whereas
102
HKKRKRT K. I'OTSWAI.I)
21
••
22
23
GENITAL SEGMENTS OF SPIRORBIS
the average length of a Sp. spirillum sperm is 5(^ microns. The descriptions
the dextral species are essentially in agreement with those given by Franzen (195(> ).
Arrangement of stages within the mule segments
Each male segment in a sexually mature animal contains stages of spermio-
genesis. Synchronous stages are found in clusters within the coelom, but there
is no predictable arrangement and each segment is autonomous. Generally, a
single segment contains clusters of primary spermatocytes, secondary spermatocytes,
and spermatids ; however, a single segment may or may not also contain meiotic
figures and mature sperm. As one would expect, the synchronous clusters of
primary spermatocytes are smallest in size, the cluster size reaching a maximum
with the spermatids and sperm. To account for this arrangement, there must be
a simultaneous proliferation of primary spermatocytes occurring periodically in
each male segment throughout the sexual period.
There is no clear evidence, at the level of light microscopy, for the existence of
nurse cells in male segments. It is interesting to note, however, that the peritoneal
cells lining the coelomic cavity accumulate inclusions not unlike those found in the
peritoneal cells of the female segments (Fig. 14). This accumulation of inclusion
bodies never reaches the degree observed in female segments.
The septa separating the male segments are, like those of the female segments,
made up of two thin epithelial layers of peritoneal origin. In each half of the male
segments, the peritoneal layer of the posterior septum is folded in upon itself in
such a manner as to form a short duct near the ventral floor of the coelom. Each
duct, the ventral portion of which is ciliated, ends blindly at the same level and
just lateral to the ventral nerve cords (Fig. 25). Mature spermatozoa are often
found in the ducts, but stages of spermatogenesis from primary spermatocytes to
late spermatids may also be found in the ducts. Bergan (1953a) refers to these
as abdominal nephridia but this can hardly be correct since their peritoneal origin
is obvious. It seems most likely that the ciliated ducts represent remnant
coelomostomes.
DISCUSSION
The most striking feature concerning the secondary segments of Spirorbis is
the fact that each functions as a genital segment and contains a well defined and
FIGURE 18. Sp. morchi: section through a male segment showing spermatids in the "incom-
plete ring stage" (IR). Epon ; Richardson's stain. (1925 X)
FIGURE 19. Sp. mdrclii: section through a male segment showing spermatids in the "homo-
geneous stage." Note the apical rim (AR), mitochondrial spheres (MS), forming vacuoles
(V), and dark-staining body which may be Golgi rest (Go). Epon; Richardson's stain.
( 1925 X )
FIGURE 20. Sp. indrclii: section through a male segment showing spermatids in the "conical
stage" (CS). Note that the acrosome (A) is now visible. Epon; Richardson's stain. (1925 X)
FIGURE 21. Sf>. morchi: nearly mature sperm. Middle piece still somewhat irregular in
appearance. Living material ; phase-contrast. ( 1140 X )
FIGURE 22. Sp. ambilateralis : mature sperm. Living material; phase-contrast. (1140 X)
FIGURE 23. Sp. ritrcus: mature sperm. Living material; phase-contrast. (1140 X)
FIGURE 24. Sf>. spirillum: mature sperm. Living material; phase-contrast. (1140X)
FIGURE 25. Sp. morchi: cross-section through a male segment showing a coelomostome
rudiment (CoR). Epon; Richardson's stain. (700X)
104 HERBERT E. POTSWALD
persistent gonad. Distinct gonads have been described in Sulmacina dysteri (Mala-
(|tiin, 1925), I:il<></nniti iuiplc.ra (Faulkner, 1930), Pomatoccros triqueter (Thomas,
1(>40; lyssum. 1(>57). and in tlie Arenicolidae (Fauvel, 1(>59; Matthews, 19(>2).
'\Yith the exception of the forms just mentioned, a discrete gonad is not charac-
teristic of polvchaetes. In a majority of polychaetes the germ cells arise from
rather indefinite patches of peritoneum (Parker and Haswell, 1957; Fauvel, 1959;
Dales, 1963).
Cells having features similar to those of the primordial germ cells in Spirorbis
have been reported in the three serpulid species mentioned above (Malaquin, 1925 ;
Faulkner, 1930; Jyssum, 1957). Jyssum (1957) in a study on oogenesis in
Pomatoccros triqueter refers to these cells as neoblasts and describes them as giving
rise to the female gamete. The interphase nuclei of the neoblasts contain periph-
eral chromatin clumps ; however, they are not as regular in size and form as are
those of the primordial germ cells in Spirorbis. Another difference between the
two is that the neoblasts contain one or two nucleoli. Jyssum describes the neo-
blasts as dividing and giving rise to gonia which, in turn, divide to give rise to
primary oocytes. The distinction between the two divisions is apparently based
on the thickness of the chromosomes at metaphase, the chromosomes of the neo-
blasts being thicker and more lumpy than the chromosomes in gonial metaphase.
Such a gonial stage, between primordial germ cell and primary cyte, has not been
observed in Spirorbis. A distinct gonial stage in Spirorbis is quite probable, but,
as yet, has not been identified. It will be shown, in a subsequent report, that the
primordial germ cells arise relatively early in the development of Spirorbis and can
be followed to the sites of gonad formation in the metamorphosed pre-adult.
Many of the problems associated with oogenesis have already been discussed.
The fact that most of the growth and maturation of the oocyte occur freely within
the coelom without the aid of attached follicle or nurse cells offers a number of
possibilities for experimental investigation of this type of oocyte development.
For example, it may be possible to culture oocytes in vitro and study the effects
of environmental conditions on growth and vitellogenesis.
There are a number of problems associated with spermiogenesis in Spirorbis
which are beyond the resolution afforded by the light microscope. It would be
of interest, for example, to study acrosome formation. Only a few electron micro-
scope studies of acrosome formation in invertebrates have been attempted (see
Cameron and Fogal, 1963) and some of the homologies between various types of
acrosomes are at a start of being elucidated. Another problem which would be
of interest is the origin and possible significance of the "neutral red vacuoles"
which are so apparent during spermiogenesis.
The mature sperm of Spirorbis is not of the simple or primitive type but is
modified. Two morphological types have been described here. Franzen (1956)
recognizes three morphological types in the genus: Sp. spirillum and Sp. I'itreus
share one type ; Sp, borcalis and Sp. granulatns share another type quite different
from the first ; and a third type, even more highly modified than the other two,
is found in Sp. pagenstcchcri. The descriptions given in the present study for
Sp. spirillum and Sp. I'itrcus agree in the main with those given for the same
species by Franzen. Sperm morphology of Sp. m'orchi and Sp. ambilateralis is
not consistent with any of the three types recognized by Franzen. There can be
GENITAL SEGMENTS OF SPIRORBIS 105
no doubt that the sperm morphology in the genus Spirorbis has some usefulness
as an auxiliary systematic character. In this connection it may again he pointed
out that the sperm morphology of the sinistral form of Sf>. litrcus is identical with
that of the dextral form.
Franzen's major thesis is that there is a definite relation between the morphology
of the sperm and the biology of fertilization. According to this thesis, inverte-
brates which discharge their gametes freely into the water retain a primitive type
of sperm which is characterized as consisting of a short rounded to conical head,
a small middle piece containing four to five mitochondria! spheres, and a tail
formed by a long flagellum. Invertebrates which have an altered biology of propa-
gation exhibit a modified sperm morphology. If the end product of spermatogene-
sis is a modified sperm, primitive characters are retained during spermiogenesis ;
the four mitochondrial spheres which appear in the spermatid of Spirorbis would
be such a character. In a discussion of the family Serpulidae, Franzen points
out the hermaphroditic and brooding nature of Spirorbis but admits that literature
on the reproductive biology of the genus is extremely incomplete. Speculating on
the mode of fertilization, he is of the opinion that the most natural way for it to
occur would be that sperm from a nearby animal are sucked into the tube and
there fertilize the eggs. If this were the case, sperm would not have to swim great
distances in order to reach the eggs. There is also evidence that at least certain
species of Spirorbis are capable of self-fertilization (Potswald, 1964; Gee and
Williams, 1965).
The writer washes to express his gratitude to Dr. Robert L. Fernald, Director
of the Friday Harbor Laboratories, for his valuable advice and assistance. Drs.
W. Siang Hsu and Paul L. Illg are thanked for their helpful suggestions.
SUMMARY
1. In all species of Spirorbis examined, the first two or three abdominal seg-
ments of mature adults are female and the remaining abdominal segments are male.
Both female and male gametes differentiate simultaneously in the same individual
and arise from a discrete and persistent gonad composed of primordial germ cells
arranged in two retroperitoneal rows, mesial to the ventral nerve cords, and running
the length of the abdominal segments.
2. Cytological events associated with the development of female and male
gametes are described. Differentiation of oocytes and spermatocytes occurs freely
within the coelomic cavity without the aid of attached nurse cells.
3. Although coelomostome rudiments are present in both female and male seg-
ments, functional genital ducts do not develop. Spawning is assumed to take
place by rupture of the body wall.
LITERATURE CITED
AGASSIZ, A., 1866. On the young stages of a few annelids. Ann. I. \ccinn Nat. Hist. N. V.
7: 303-343.
BENNETT, H. S., AND J. LUFT, 1959. S-collidine as a basis for buffering fixatives. J. Bi»ph\s.
Biochem. Cytol., 6: 113-117.
106 HERBERT E. POTSWALD
BERGAX. I'., 1953a. On the anatomy and reproductive biology of Spirorbis Daudin. Nytt
Mayasin Zoo/. Oslo, 1 : 1-26.
BERGAN, P., 1953b. The Norwegian species of Spirorbis Daudin. N\tt Mayasin Zoo/. Osln,
1 : 27-48.
BERKELEY, E., AND C. BERKELEY, 1952. Polychaeta sedentaria. Canad. Pacific Fauna. 9b:
132-138.
BORG, F., 1917. Uber die Spirorbisarten Schwedens. Zoo/. Bidray Uppsala, 5: 15-38.
BUSH, K. J., 1904. Tubiculous annelids of the tribes Sabellides and Serpulides from the Pacific
Ocean. Harriman Alaska E.rpcd., 12: 169-298.
CAMERON, M. L., AND W. FOGAL, 1963. The development and structure of the acrosome in the
sperm of Luinbricus tcrrcstris L. Canad. J. Zoo/., 41: 753-761.
CAULLERY, M., AND F. MESNIL, 1897. fitudes sur la morphologic comparee et la phylogenie
des especes chez les Spirorbis. Bull. Sci. France Bclyiquc, 30: 185-233.
CLONEY, R. A., 1961. Observations on the mechanism of tail resorption in ascidians. Aincr.
Zoo/., 1 : 67-87.
DALES, R. P., 1963. Annelids. Hutchinson and Co., London.
DASGUPTA, S., AND A. P. AUSTIN, 1960. Chromosome numbers in serpulids. Quart. J. Micr.
Sci., 101 : 395-399.
ELSLER, E., 1907. Deckel und Brutpflege bei Spirorbis. Zcitschr. wiss. Zoo/.. 87: 603-642.
FAULKNER, G. H., 1930. On the anatomy and histology of bud formation in Filograna iiuplc.ra.
J. Linn. Soc. London (Zoo/.), 37: 109-190.
FAUVEL, P., 1927. Polychetes sedentaires. Fannc dc France. Paris. 16: 388-405.
FAUVEL, P., 1959. Classes des annelides polychetes. Trait c de Zoologic, 5: 132.
FEWKES, J. W., 1885. On the larval forms of Spirorbis borcalis. Atncr. Naturalist. 19: 247-
257.
FRANZEN, A., 1956. On spermiogenesis, morphology of the spermatozoon, and biology of fer-
tilization among invertebrates. Zoo/. Bidray Uppsala. 31 : 355-482.
FRANZEN, A., 1958. On sperm morphology and acrosome filament formation in some Annelida,
Echiuroidea, and Tunicata. Zoo/. Bidrag Uppsala, 33: 1-28.
GEE, J. M., AND G. B. WILLIAMS, 1965. Self and cross-fertilization in Spirorbis borcalis and
S. patjcnstccheri. J. Mar. Biol. Assoc., 45: 275-285.
HEMPELMANN, F., 1931. Archiannelida + Polychaeta. Handbuch der Zoologie, 2: 90.
JYSSUM, S., 1957. Investigations of the neoblasts and oogenesis in the serpulid Pomatoceros
triquctcr L. Nytt Mayasin Zoo/. Oslo, 5: 5-10.
LOYE, J. F. ZUK, 1908. Die Anatomic von Spirorbis borcalis mit besonderer Beriicksichtigung
der Unregalmassigkeiten des Korperbau und deren Ursachen. Zoo/. Jahrb. Abt. Anat.
Ontoy. Tierc.26: 305-354.
LUFT, J. H., 1961. Improvements in epoxy resin embedding methods. /. Bioph\s. Biochcin.
Cytol.,9: 409-414.
MALAQUIN, A., 1925. Les cellules germinales (gonocytes) sont, au cours de la reproduction
asexuee de Salmacina dvstcri Huxley, la source de la proliferation blastogenique.
C. R. Acad. Sci. Paris, 180: 873-875.
MATTHEWS, L. H., 1962. The structure of the ovary. A. Invertebrates. In: volume I, The
Ovary, edited by Sir Solly Zuckerman. Academic Press, New York.
McMANUs, J. F. A., AND R. W. MOWRY, 1960. Staining Methods Histologic and Histochemi-
cal. Paul B. Hoeber, Inc., New York.
PAGENSTECHER, H. A., 1863. Untersuchungen iiber niedere Seetiere aus Cett. VII. Ent-
wickelungsgeschichte und Brutpflege von Spirorbis spirillum. Zcitschr. wiss. Zoo!.,
12: 486-496.
PARKER, T. J., AND W. A. HASWELL, 1957. A Text-book of Zoology. I. 6th ed. Macmillan
Co., London.
PETTIBONE, M. H., 1954. Marine polychaete worms from Point Barrow, Alaska, with addi-
tional records from the North Atlantic and North Pacific. Proc. U. S. Nat. Mus.
103: 341-345.
1' ix ELL, H., 1912. Polychaeta from the Pacific Coast of North America. Part I. Serpulidae
with a revised table of classification of the genus Spirorbis. Proc. Zoo/. Soc. London,
2 : 784-805.
GENITAL SEGMENTS OF SPIRORBIS 107
POTSWAI.II, H. E., 1964. The nature of the primordial germ cells and evidence for self-fertiliza-
tion in Spirnrbis ( Polychaeta, Serpulidae). Amcr. '/.ool., 4: Abstract 93.
PoTswAi.n, H. E., 1965. Reproductive biology and development of Spirorbis (Polychaeta, Ser-
pulidae). Doctoral Thesis, University of Washington, Dissertation Abstracts No.
65-8528.
RAVEX, CHR. P., 1961. Oogenesis : The Storage of Developmental Information. Pergamon
Press, New York.
RICHARDSON, K. C., L. JARETT AND E. H. FIXKE, 1960. Embedding in epoxy resins for ultra-
thin sectioning in electron microscopy. Stain Tcchnol.. 35: 313-323.
SCHIVELY, M. A., 1897. The anatomy and development of Spirorbis borcalis. Proc. Acad.
\'at. Sci. Philadelphia, 49: 153-160.
THOMAS, J. G., 1940. Pomatoceros, Sabclla and Amphitritc. Liverpool Mar. Biol Committee
Memoir XXXIII.
CHOLINESTERASE IN THE BRAIN OF THE CECROPIA SILKMOTH
DURING METAMORPHOSIS AND PUPAL DIAPAUSE1
DAVID G. SHAPPIRIO, DANIEL M. EICHENBAUM 2 AND BRUCE R. LOCKE2
Department /if Zoology, The University of Michigan, Ann Arbor, Michigan 48104
In giant silkmoths, pupal diapause results from the failure of neurosecretory
elements in the brain and corpora cardiaca to provide the hormonal stimulus re-
quired for initiation of further development ; the latter resumes, months later, when
the neurosecretory system recovers its function after exposure of the pupa to
appropriate temperatures and photoperiod (Williams, 1946, 1952, 1956; Williams
and Adkisson, 1964). For further resolution of the control of diapause and devel-
opment, a key problem is to define the physiological processes, within the brain,
that establish and later efface its endocrine impotence.
As one aspect of this problem, we have examined the behavior of cholinesterase
(ChE) in the brain of the Cecropia silkmoth. Our inquiry is based on the find-
ings of Van der Kloot (1955), who reported a disappearance of ChE and spon-
taneous electrical activity in the brain at the time of pupation, and their subsequent
reappearance together with the brain's recovery of endocrine activity. The changes
were found to be limited to the brain, since the thoracic and abdominal ganglia
retained normal levels of enzymatic and electrical activity throughout diapause.
On the basis of the close temporal correlations observed, and the possible func-
tional roles of ChE, Van der Kloot recognized that its behavior could account for
the neuroendocrine inactivation and reactivation of the brain. As is apparent from
current reviews on insect endocrinology and development (e.g., Wigglesworth,
1964; Gilbert, 1964), these observations have remained the most promising leads
to date on the control of neurosecretion and diapause in lepidopterous insects.
However, the aforementioned conclusions have recently been questioned as a
result of two brief reports ( Schoonhoven, 1963; Tyshtchenko and Mandelstam.
1965) containing electrophysiological observations supplemented by certain limited
enzymatic data. The observations were made on diapausing pupae belonging to
several families of Lepidoptera, and including a number of Cecropia. Since the
electrophysiological findings described in these two papers conflict with one an-
other, as well as with the observations of Van der Kloot, the net result has been
to create a rather uncertain picture of the extent of neural activity during diapause.
From the biochemical standpoint the picture is somewhat clearer but still equivocal.
Through the use of a manometric technique, Schoonhoven detected hydrolytic
activity in brains of diapausing pupae of a geometrid moth, Bnpalus piniarins L. ;
Tyshtchenko and Mandelstam, by histochemical methods, detected hydrolytic activ-
ity in brains of diapausing pupae of the silkmoth Antheraea pernyi Guer. These
1 This study was supported, in part, by Public Health Service Research Grant GM-06101.
2 Messrs. Eichenbaum and Locke held National Science Foundation Undergraduate Re-
search Participation Awards during this investigation.
108
CHOLINESTERASE IN SILKMOTH BRAIN 109
activities were ascribed to ChE. Though rather suggestive of ChE, both reports
omit mention of the biochemical details and controls required to confirm that the
activity was in fact enzymatic, and that it was due to the action of ChE rather than
that of other esterases known to occur in insect brain.
In the present study, we demonstrate by histochemical and quantitative methods
that substantial ChE activity persists throughout diapause in brains of Cecropia
and other silkmoths. Our findings provide comparative biochemical information
on the properties and specificity of ChE in silkmoth brain, and show that diapause
cannot be attributed to generalized absence of ChE, as seemed possible heretofore.
But, for reasons to be discussed subsequently, the present findings do not preclude
a regulatory role for one or more forms of esterase in the control of neurosecretion
and diapause. These findings have been announced previously in abstract (Shap-
pirio, Eichenbaum and Locke, 1965).
MATERIALS AND METHODS
1. Experimental animals
Brains from the following species of silkmoths were used : Hyalophora cecropia
(L.), Samia cvnthia- (Dru. ), and Anthcraca polyphcmus (Cram.). For con-
venience, these will be identified henceforth as Cecropia. Cynthia, and Polyphemus,
respectively. Most of the Cecropia were reared by us under nylon nets on wild
cherry trees. Other Cecropia. as well as all the Cynthia and Polyphemus, were
purchased from dealers in the northeastern and midwestern United States. The
insects were managed as described previously (Williams, 1946; Shappirio and
Williams, 1957). In addition to larvae and prepupae. we utilized: (a) "unchilled
pupae" kept at 25° C, in which diapause was found to persist for more than five
months after pupation; (b) "chilled pupae" stored at 6° C. to favor the prompt
return of endocrine activity, followed by the resumption of development, upon their
return to 25° C. ; and (c) "developing adults" derived from previously chilled
pupae allowed to terminate diapause at 25° C. Developing adults were staged by
use of the morphological and physiological criteria summarized by Schneiclerman
and Williams (1954). In additional experiments, adult moths were used four
days after their emergence at 25° C.
2. Initial preparation of tissue
For histochemical and quantitative studies, brains were excised from insects
anesthetized in carbon dioxide (Williams, 1946) and then briefly rinsed in Ringer
(Ephrussi and Beadle, 1936). In the case of larvae and prepupae, as well as
diapausing pupae, the optic nerves were severed just distal to the melanin granules
lying at their base ; other nerves were severed as close as possible to the surface of
the brain. In experiments on developing adults and adults, the optic lobes and
tracts were included with the brain itself, but antennal and other nerves were tran-
sected at the surface of the brain and thus excluded from analysis. Also excluded
was the subesophageal ganglion, by means of transverse cuts midway across both
roots connecting this ganglion to the brain. Further details on tissue preparation
are provided below.
110 SHAPPIRIO, EICHENBAUM AND LOCKE
3. Histochemistry
The size and fragile texture of silkmoth brains posed problems in tissue prepara-
tion which were sufficiently overcome only after extensive trials with various tech-
niques. The methods giving most reliable results are described here. Brains were
in certain cases prefixed in lO/f formalin (pH 7.0) for 1-2 hours at 2° C., and
then rinsed in Ringer for an equal time. Alternatively, post fixation was employed
as a variant in technique and to control for possible formalin-induced fixation arti-
fact ; postfixation was accomplished by exposure of cryostat sections, mounted and
sectioned as described below, to acetone at 2° C. for one hour before staining.
To provide a suitable matrix for sectioning, unfixed or fixed brains were em-
bedded in small pieces of fresh mouse liver (approximately 5-mm. cubes) and
immediately frozen in isopentane cooled to a viscous state by liquid nitrogen.
The resulting frozen block was then sectioned in a Universal cryostat ( — 15° C. )
at thicknesses of 8-10 /x. When possible, serial sections were placed in sequence on
several microscope slides.
For detection of ChE activity in cryostat sections, we employed Gomori's
(1952) modification of the Koelle (1951) method. In this procedure, acetylthio-
choline (AThCh), a thioester analogous to acetylcholine (ACh), serves as sub-
strate. Hydrolysis of the thioester yields thiocholine, which is precipitated at sites
of reaction in the form of copper thiocholine sulfate ; the latter is then converted to
copper sulfide for easier visualization of reaction sites within the section. In our
studies, an incubation time of 55-75 minutes at 20° C. proved optimal. After stain-
ing and conversion to copper sulfide, sections were lightly counterstained with
aqueous Ehrlich hematoxylin, and mounted in glycerin jelly.
In histochemical studies involving inhibitors, cryostat sections were incubated
for 30 minutes in Ringer containing the desired inhibitor concentration, before
exposure to reaction medium which also contained inhibitor at this concentration.
Control sections, serial when possible, were incubated and stained in parallel but
without inhibitor.
4. Quantitative enzymatic methods
To obtain more detailed information on the properties of ChE in Cecropia brain,
and to survey its behavior during the life history, we exploited the sensitive spectro-
photometric method introduced by Ellman (Ellman ct aL, 1961). This procedure
also uses AThCh as substrate. Thiocholine generated by hydrolysis reacts with
5,5'-dithio-/>?V2-nitrobenzoate (DTNB), incorporated in the reaction medium, to
yield a bright yellow color attributable to the thionitrobenzoate anion. The reaction
was followed at 412 m/* by means of a Beckman Model DU spectrophotometer.
Rapid assay of ChE activity in individual brains was possible with this method.
In most experiments, each freshly excised brain was homogenized in a micro-
size tissue grinder kept at 0° C., to yield 0.5 ml. of homogenate in 100 mM potas-
sium phosphate, pH 8.0. This volume was adequate for duplicate or triplicate
assays. In routine measurements of ChE activity in brains from animals at suc-
cessive stages in the life history, we used a reaction volume of 1.02 ml. at 25° C.,
containing the following reagents at the final concentrations shown : 0.75 mM
AThCh; 1.0 mM DTNB ; 100 mM potassium phosphate, pH 8.0; and homogenate.
CHOLINESTERASE IN SILKMOTH BRAIN 111
For other types of experiments, designed to examine the effects of pH, substrate
concentration, and other factors, this protocol was modified as appropriate for the
individual experiments noted under Results.
In experiments using inhibitors, the reaction was initially followed for 5 min-
utes in the absence of inhibitors, after which inhibitor was added to yield the desired
final concentration. When the reaction had stabilized, which occurred within 5
minutes, the rate was recorded for a further 5 minutes. In all experiments, cor-
rection was made for changes in absorption due to endogenous thiols. which were
very slight. Correction was also routinely made for changes in absorption due
to non-enzymatic hydrolysis of substrate, which was 10^- or less of the total reac-
tion rate under routine conditions of assay described in the preceding paragraph.
Suitable experiments established that the reaction rate, thus corrected, was propor-
tional to the enzyme concentrations used, was linear during the period of assay,
and was limited by the hydrolytic step of the reaction rather than by the steps
involved in color development. Since pH 8.0 lies close to the limit of buffering
action of the HPOr/FUPCV system, we also verified that this pH was maintained
within 0.05 unit in the course of reactions at the most rapid rates encountered in
this study.
5. Chemicals
AThCh and its homologues, propionylthiocholine and butyrylthiocholine
(PrThCh and BuThCh, respectively), were purchased as the iodides from Sigma
Chemical Co., St. Louis, Missouri. DTNB, eserine sulfate, and /m-(hydroxy-
methyl)-aminomethane (Tris buffer) were also Sigma products. The two Bur-
roughs Wellcome anticholinesterase compounds, 62C47 and 284C51J, were gen-
erously furnished by Burroughs Wellcome, Inc., Tuckahoe, N. Y. These code
names denote, respectively, 1.5-/}/V(4-trimethylammoniumphenyl)-pentane-3-one di-
iodide, and l,5-/^-(4-allyldimethylammoniumphenyl )-pentane-3-one dibromide.
Iso-OMPA (tetraisopropyl pyrophosphortetramide) was obtained from Koch-Light
Labs., Colnbrook, Bucks., England. Other chemicals were of analytical reagent
grade.
RESULTS
1. Histocheinical observations
Many general microanatomical features of the pupal brain in Cecropia are typi-
cal of the arthropod brains and central ganglia described in the treatise of Buliock
and Horridge (1965). The pupal brain is bilobate in structure, with most cell
bodies being located peripheral to large regions of neuropile. The latter consists
of fine nerve fibers and cytoplasmic processes of glial cells. It constitutes the
principal region of synaptic contact, and occupies substantial regions of the central,
lateral, and ventral portions of each lobe, as well as the interior part of the tissue
connecting the lobes.
When formalin-prefixed or acetone-post fixed cryostat sections were treated as
described under Methods, an intense deposit of histochemical reaction product was
observed in neuropile, as illustrated in Figure 1. Except for differences in frag-
mentation of tissue, which was greater after formalin treatment, prefixation and
112 SHAPPIRIO, EICHENBAUM AND LOCKE
postfixation yielded an identical histochemical pattern. No reaction product was
observed when AThCh was omitted from the reaction medium. Similar findings
were made in each of 15 unchilled and chilled Cecropia pupae which were judged
to he diapausing in terms of physiological and morphological criteria (Schneider-
man and Williams, 1954).
With our material, the histochemical method did not afford sufficient resolution
to permit us to ascertain the localization of ChE more intimately in the neuronal
or glial elements of neuropile. However, in favorable preparations, reaction prod-
uct was visualized in bands which apparently correspond to major prevailing direc-
tions of fiber tracts.
Initial insight into the specificity of the histochemical reaction in the neuropile
was gained with two agents, eserine sulfate and wo-OMPA ; the former is a general
FIGURE 1. Section of brain from diapausing pupa, stained for ChE. The photograph de-
picts an 8 fj. cryostat section, postfixed in acetone, which was stained histochemically using
AThCh as substrate and counterstained with Ehrlich hematoxylin. The darkly stained region
contains histochemical reaction product (CuS) and shows the localization of ChE activity
(X72).
inhibitor of ChE's and the latter is relatively selective for mammalian butyryl-
cholinesterase (BuChE) when applied at appropriate concentrations. In brains of
unchilled and chilled Cecropia pupae, the neuropile-associated esterase was found to
be abolished by eserine sulfate at 1O5 M and to be unaffected by wo-OMPA at
10-4 M.
Our observations on brains of post-diapausing Cecropia are limited in number,
but permit several conclusions. When pupae chilled for 16 weeks were placed at
25° C., a period of 7-10 days elapsed before the first externally visible sign of
adult development, namely, retraction of the leg epithelium from its overlying
cuticle. This signals the second day of adult development (Schneiderman and
Williams, 1954). In the present study, the histochemical pattern was found to
remain identical with that encountered in brains of diapausing pupae, when obser-
CHOLINESTERASE IN SlLKMOTH BRAIN
rations were made during the first four or five days after transfer of the chilled
pupae to 25° C. Subsequently, an increased area of deposition of reaction product
was observed, accompanied by a slight though seemingly significant increased
intensity of staining in neuropile. At the outset of adult development, esterase
activity was no longer confined to neuropile, but had spread laterally and ven-
trally to encompass regions of neuronal or glial cell bodies. The esterase reaction
in post-diapausing insects was inhibited fully by eserine sulfate at 10~5 M.
2. Properties of CliE in pupal brain as revealed by quantitative methods
In order to provide a meaningful basis for evaluating the behavior of ChE in
relation to the onset and termination of pupal diapause, and to afford comparative
biochemical insight, experiments were carried out to define optimal conditions of
assay for ChE and to determine the specificity of the enzyme or enzymes detected.
a. Effect of pH on ChE activity
A series of measurements was performed on the same homogenate of brains
pooled from unchilled diapausing pupae of Cecropia, using 100 mM phosphate or
Tris-HCl buffers in the range pH 6.5 to 9.0. Although the activity with Tris was
invariably lower than with phosphate (see below), in both cases enzyme activity
increased from pH 6.5 to 8.0 but showed little increase in the range 8.0 to 9.0.
For spectrophotometric assays, pH 8.0 was selected since this afforded a maximal
rate with relatively low non-enzymatic hydrolysis of AThCh. The rate of the
latter reaction increases markedly with further increase in pH, and at pH 9.0
exceeds the enzyme-catalyzed rate when the latter is kept within a range appro-
priate for meaningful assay. In experiments involving different pH's, suitable
controls confirmed that the absorptivity of the thionitrobenzoate anion was essen-
tially constant between pH 6.5 and 9.0.
b. Effect of composition and concentration of certain buffers
At pH 8.0, the reaction rate was found to be essentially similar when assays on
the same homogenates were carried out in the presence of 100 mM phosphate or
70 mM bicarbonate (the latter charged at 25° C. with S% carbon dioxide in oxy-
gen). Only one-quarter of this rate was observed with 100 mM Tris at the same
pH. The lower rate with Tris appears to be an inhibition rather than a failure of
activation, since combination of phosphate and Tris did not elevate the rate above
that observed with Tris alone. At 10 mM, the rate with phosphate was nearly
50% lower than at 100 mM. In view of these findings, a concentration of 100 mM
was used in routine assays.
c. Relationship of ChE activity to substrate concentration
To the best of our knowledge, this important relationship has not been con-
sidered in previous studies on ChE in silkmoth brains. Figure 2 illustrates the
striking dependence of ChE activity upon the concentration of the substrate, AThCh.
A final substrate concentration of 0.75 mM was found to yield the most rapid
reaction rate (Fig. 2). Above this concentration, one can readily observe that
114
SHAPPIRIO, EICHENBAUM AND LOCKE
activity decreases progressively, yielding a graph of a shape typical for acetyl-
cholinesterases (AChE's) in vertebrate and invertebrate preparations (Augustins-
son, 1949, 1903).
In order to permit comparison between data obtained under conditions found
optimal for assay of ChE in the present study, and the results of Van der Kloot's
(1955) investigation, the substrate-activity curve for pupal brain ChE was also
40
30
o
o
o
I 20
01
10
X
-log (AThCh)
FIGURE 2. ChE activity as a function of AThCh concentration. The solid line describes
enzymatic activity, which was corrected for spontaneous hydrolysis of AThCh at substrate con-
centrations above 10"* M. The broken line describes the rate of spontaneous hydrolysis of
AThCh. To provide sufficient material for the experiment illustrated, brains of six unchilled
Cecropia pupae were pooled before homogenization. Enzyme activity is shown in relative units :
one unit corresponds to a change in absorbance of 0.001 per minute.
determined under our "routine" assay conditions, but supplemented by sodium
chloride to yield a final concentration of 0.5 M. The latter had been employed in
Van der Kloot's experiments owing to the findings of Chadwick ct al. (1953) show-
ing that it yielded optimal ChE activity in homogenates of brains from flies. Our
experiments with Cecropia showed that the substrate-activity curves with and
without extra salt possess similar shapes. However, the addition of salt alters
CHOLINESTERASE IN SILKMOTH BRAIN 115
both the optimal substrate concentration and the reaction rate at any given sub-
strate concentration. In the presence of added salt, the substrate optimum was
slightly higher, and when comparisons were made at the substrate optima with and
without salt, that in salt was 22% lower.
d. Effects of selective esterase inhibitors
Information on the specificity of the ChE reaction in pupal brains was gained
with a variety of agents known to be selective on the basis of studies with vertebrate
and invertebrate preparations. The histochemical observations, described earlier,
were confirmed by the finding that eserine sulfate at 10~5 M completely abolished
enzymatic activity. The Burroughs Wellcome compounds, 62C47 and 284C51J,
also inhibited the reaction fully when tested at final concentrations of 10~6 M. In
contrast, wo-OMPA at 10"* M failed to inhibit ChE in homogenates of brain from
unchilled and chilled pupae.
Additional studies with inhibitors were carried out on brains from larvae,
prepupae, developing adults, and adult moths. In all cases, eserine at 10~5 M was
fully inhibitory. The agents 62C47 and 284C51J also fully inhibited ChE in larval
brains and in brains from animals at the outset of adult development, when tested
at 10~6 M. Iso-OMPA at 1O4 M was non-inhibitory in larvae, prepupae, and
developing adults at early stages ; but the brains of adult moths were slightly
affected, a concentration of 10~4 M inhibiting their ChE by 8-10%.
e. Reactivity of brain ChE with homologues of AThCh
Figure 3 illustrates the relative reaction rates obtained at a graded series of
concentrations of PrThCh and BuThCh. The homogenate used for the experi-
ment illustrated, which gave typical results, was derived from the pooled brains of
unchilled pupae. The same homogenate exhibited a rate of 24 (relative units as in
Fig. 3) when examined with AThCh at 0.75 mM. Thus, the reaction rate de-
creases in the order : AThCh > PrThCh > BuThCh. Although only a limited
range of concentrations was studied with the substrate homologues, it is clear that
the substrate-activity curve for PrThCh resembles that for AThCh in showing a
marked optimum substrate concentration, above which the reaction rate progres-
sively decreases. Moreover, the position of the substrate optimum for PrThCh is
rather close to that described (Fig. 2) for AThCh. In the range studied, reaction
rate increased slightly with increase in BuThCh concentration, but showed no clear
optimum. It was not possible to carry the studies with BuThCh to higher con-
centrations than are shown in Figure 3, inasmuch as the non-enzymatic breakdown
of BuThCh became excessive when its concentration was increased further. The
reactions with PrThCh and BuThCh were found to be fully sensitive to eserine
at 10-5 M.
3. ChE activity in Cecropia brain at successive stages in metamorphosis and dur-
ing diapause
In order to permit observations on as homogeneous a sample of insects as pos-
sible, ChE activity was determined in individuals of the same batch of Michigan-
116
SHAPPIRIO, EICHENBAUM AND LOCKE
reared Cecropia. Our 1964 crop, used for this purpose, was especially suitable
since the proportion of animals emerging precociously from pupal diapause was
unusually low ; only two individuals out of more than 800 initiated development and
emerged as moths during the first 6 months after pupation, without prior chilling
at 6° C.
Table I summarizes measurements of ChE activity in 46 individual brains
from animals at successive stages in the life history, ranging from late in larval life
through and beyond pupal diapause to the initial phase of adult development. All
measurements were made under the routine conditions of assay described under
Methods. As shown in the Table, ChE was readily detectable at all stages exam-
ined. Of special interest is the finding that ChE activity undergoes no decline
12
.
io-
/-s
/ \
/ \
2
I 8
/ \
/ \
1
|5
/
/
6
•*—
73
•
4
,
LJ
jc
0
-
2
•
•^
s'
s
s
.<'
- log (substrate)
FIGURE 3. ChE activity as a function of PrThCh and BuThCh concentrations. The upper
line describes activity for PrThCh ; the lower line relates to BuThCh. The activities have been
corrected for spontaneous hydrolysis of substrates. Enzyme activity is shown in relative units :
one unit corresponds to a change in absorbance of 0.001 per minute.
CHOL1NESTERASE IN SILKMOTH BRAIN
117
TABLE I
ChE activity in brain homogenates of Cecropia
Source of brain
Number
studied
Enzyme activity*
(jimoles AThCh, ''
brain-hr.)
Range
Remarks**
Larvae
4
0.24 ± 0.05
0.20-0.35
Late 5th instar, feeding
Prepupae
4
0.24 ± 0.02
0.22-0.28
Several hours before pupation
Fresh pupae
4
0.29 ± 0.03
0.24-0.33
2-22 hours after pupation
Unchilled pupae
6
0.32 ± 0.05
0.22-0.38
1-9 days after pupation
Unchilled pupae
8
0.33 ± 0.08
0.20-0.42
4-12 weeks after pupation
Chilled pupae
5
0.25 ± 0.03
0.22-0.31
8 weeks after pupation ; then 10
weeks at
6°C.
Chilled pupae
8
0.58 ± 0.04
0.51-0.62
8 weeks after pupation; then 13-20 weeks
at 6° C.
Developing adults
7
0.60 ± 0.05
0.51-0.69
2nd day of development
* Mean activity ± average deviation.
** Except where shown, all animals maintained at 25° C.
around the time of pupation, when the neurosecretory system becomes inactive, and
when ChE was previously reported to undergo precipitous disappearance (Van der
Kloot, 1955). It seems clear from Table I that except for possible minor fluctua-
tions, an essentially unchanging level of ChE activity persists in the newly pupated
animal, in unchilled pupae for at least 12 weeks after pupation, and in pupae
chilled up to 10 weeks at 6° C. In several of the groups of unchilled pupae, one
or two of the animals studied exhibited activity substantially higher or lower than
the mean recorded in Table I, but no upward or downward trend in activity was
noted during the indicated time intervals after pupation.
Table I also provides evidence for a rise in ChE activity during more prolonged
chilling at 6° C. In each of eight pupae, examined at intervals after periods of
chilling ranging from 13 to 20 weeks, the level of activity was at least double the
mean activity recorded after ten weeks' chilling. Evidently the rise in activity
occurred between the 10th and 13th weeks in this batch of animals. No trend of
change in activity was noted between the 13th and 20th week. The elevated level
was found to persist when chilled pupae were returned to 25° C. and allowed to
initiate adult development (Table I). At all stages shown in Table I, the enzy-
matic activity was fully inhibited by eserine at 10~5 M. An AThCh concentration
of 0.75 mM was found to be optimal in assays on larval and developing adult
brains, as had earlier been established for pupae.
A total of 50 Cecropia pupae from the same batch as those just considered was
returned to 25° C. after 16 weeks of chilling, and was used to provide animals at
successive stages during the maturation of the adult moth after termination of dia-
pause. As shown in Figure 4, ChE activity undergoes a six-fold increase during
this period, when expressed on a "per brain" basis. In the first two weeks of
adult development, activity rises progressively, but the high level thus attained
persists without large change until the time of adult emergence. For unexplained
reasons, a large variation in activity was observed in the later phases of adult
development and in adults.
118
SHAPPIRIO, EICHENBAUM AND LOCKE
4.0
o
o
LJ
.C
O
3.0
2.0
1.0
. r
• t
I — I — I — t-
-1 1—4-
-I— I 1-
4 8 12 16 20
Days in adult development
Adult
emergence
FIGURE 4. ChE activity in individual brains from developing adult and adult Cecropia.
Each point represents the average of duplicate determinations on one brain. Enzyme activity
is shown as micromoles AThCh per brain-hour. These units of activity are equivalent to those
summarized in Table I.
4. ChE activity in brains from various species of silkmoths
The preceding assays were carried out on brains from the same batch of
Michigan-reared Cecropia. To minimize the possibility that these data were excep-
tional, additional measurements of ChE activity were carried out, using several
species of silkmoths. In more than three dozen additional unchilled and chilled
Cecropia from different sources, ChE activity was encountered at approximately
the same levels recorded in Table I. Activity was also detected in unchilled
Cynthia pupae at about the same levels as in Table I. In unchilled Polyphemus
pupae, the activity was somewhat lower, but still detectable. Measurements on
Cynthia and Polyphemus were made using a final AThCh concentration of 0.75
mM; all activities were fully inhibited by eserine sulfate at 10"5 M. These obser-
vations provide additional evidence for essentially unchanging ChE activity in
pupae stored at 25° C. or for at least two months at 6° C.
DISCUSSION
1. Characteristics of ChE in silkmoth brain
a. Identification as AChE
Throughout most of the life history, the enzymatic activity detected by histo-
chemical and quantitative methods exhibits properties tentatively attributable to
CHOLINESTERASE IN SILKMOTH BRAIN 119
AChE. This conclusion follows from the sensitivity displayed to certain inhibitors,
from the reactivity toward thioester homologues, and from the substrate-activity
curve obtained using AThCh. The Burroughs Wellcome agents, 62C47 and
284C51J, are strongly inhibitory and selective for vertebrate AChE's (Augustins-
son, 1963), and also preferentially inhibit several ChE components of arthropod
central nervous systems that have properties similar to vertebrate AChE's (Wig-
glesworth, 1958; Maynard, 1964). The lack of sensitivity to iso-OMPA rein-
forces this finding, since uo-OMPA is known to be relatively selective for verte-
brate BuChE's including the pseudocholinesterase of mammalian brain (Aldridge,
1953; Austin and Berry, 1953; Pepler and Pearse, 1957). Further evidence for
AChE in the case of Cecropia derives from the finding that activity is much lower
toward BuThCh than toward AThCh. Finally, the pronounced reduction in
activity at the higher concentrations of AThCh and PrThCh which we employed
(Figs. 2 and 3) is typical of AChE's from vertebrate sources (Augustinsson, 1949,
1963). The combination of properties on the part of the Cecropia brain esterase
is similar to that described for brain ChE in several insect groups (Gilmour, 1961).
Insect ChE's possess properties that do not always lend themselves to convenient
classification using vertebrate-based terminology (Chadwick, 1963). Thus, despite
the similarities of Cecropia brain ChE to AChE, further study may reveal differ-
ences, and we use the term AChE with reservation.
In section 2.d of Results was mentioned the finding that ChE activity in adult
brain includes a small but significant fraction which is sensitive to wo-OMPA at
10-* M. This contrasts with the insensitivity to wo-OMPA at earlier stages in
the life history, and suggests that an additional ChE component, perhaps with
different specificity characteristics from AChE, may appear in the brain during
the later phases of adult development.
b. Changes in AChE during metamorphosis
On the basis of the present study, the only major changes in AChE activity
from the time of pupation through most of adult development are : (a) the two-fold
rise in activity which occurs during chilling at 6° C. ; and (b) the larger increase
in activity that accompanies the growth and morphogenesis of the adult brain.
With regard to the former, evidence is insufficient to determine whether the rise
is attributable to fabrication of new elements within neuropile and concomitant
synthesis of new ChE, or to enhanced titer of ChE within existing neuropile, or
to other factors. In any event, the rise in activity during chilling appears to repre-
sent the earliest biochemical signal so far reported of the brain's change in neuro-
endocrine status, which occurs during storage at the low temperature (Williams,
1946, 1956). With regard to the increase of ChE during adult development, it
seems most probable that the enzymatic changes in large part mirror the extensive
morphogenetic events occurring at this time. A study of the behavior of brain
esterases during adult development may yield useful information on changes at the
cellular and subcellular levels underlying the physiological and morphological
maturation of the adult brain.
c. Localization of ChE in pupal brain
In showing the presence of ChE in neuropile of pupal brain, our histochemical
findings agree with those of Wigglesworth (1958) on the bug Rhodnius prolixus,
120 SHAPPIRIO, EICHENBAUM AND LOCKE
and those of Salkeld (1961) on the milkweed bug, Oncopeltus fasciatus. In the
former study, Wigglesworth concluded that staining in the neuropile, with AThCh
as substrate, was largely confined to the glial elements. As used by us with
Cecropia pupae, resolution of the histochemical method was insufficient to permit
critical assessment of this important point. Wigglesworth also observed staining
outside the neuropile when other esterase substrates were used, and his overall
results supported the view that several esterases were present in the brain as a
whole. Further information regarding multiple esterases and their potential sig-
nificance will be presented in a subsequent section of this Discussion.
Wigglesworth (1958) noted in the first detailed histochemical study on localiza-
tion of ChE in the brain of an insect that the restriction or near-restriction of
AChE activity to neuropile shows certain kinship with the histochemical picture
derived from studies on amphibians. Thus, in many regions of frog brain (Rana
pipiens}, heaviest staining due to AChE is found in areas rich in synaptic termina-
tions and poor in cell bodies. However, in other areas of brain, neuron cell bodies
also possess high ChE activity (Shen, Greenfield and Boell, 1955; Koelle, 1963).
This situation in the frog differs markedly from that encountered in mammalian
brains, as typified by rat or cat, in which AChE is abundant in neuron perikarya
as well as in axonal and dendritic processes (Koelle, 1954; Pepler and Pearse,
1957). The localization of ChE in regions of synaptic contact has repeatedly
prompted the suggestion that it is functionally involved in transmission ; one can
speculate similarly regarding the AChE which we detect in neuropile of Cecropia
brain, but meaningful judgment on this point awaits further ultrastructural and
physiological evidence.
2. Analysis of present findings in relation to previous studies on Cecropia
The results of the present study clearly show that substantial ChE activity per-
sists in silkmoth brain throughout metamorphosis and diapause. Our findings
therefore contrast with those of Van der Kloot (1955) in which ChE was not
detected during diapause, but provide detailed evidence in favor of the conclusions
suggested by Schoonhoven (1963) and Tyshtchenko and Mandelstam (1965) to
the effect that ChE persists during the pupal diapause of lepidopterous insects.
Some of the discrepancy between our findings and those of Van der Kloot can
be resolved if one considers certain properties of ChE in pupal brains as revealed
in the course of the present study. It is worth noting that our spectrophotometric
method has about the same sensitivity as the manometric technique used by Van
der Kloot, since the minimum activity detectable by our method is close to that
stated by Van der Kloot (0.05 micromoles substrate hydrolyzed per brain-hour ) .
Although we have not compared the velocities of the ChE reaction in Cecropia with
respect to AThCh and ACh, it would not be surprising if the rate were higher
with AThCh, as this has been reported from time to time in the literature. Thus
choice of substrate may have contributed to our success in detecting enzymatic
activity during diapause. However, we believe that other factors are more sig-
nificant, as stated below.
We find that ChE activity in homogenates of Cecropia brain declines by as
much as 15% per hour in 100 mJ\I phosphate buffer at pH 8.0 and 0° C., and more
CHOLINESTERASE IN SILKMOTH BRAIN 121
rapidly at 25° C. The activity also declines slowly when homogenales are stored
in the frozen state at —15° C. ; about 2Q% is lost after one week's time. The loss
in activity would he significant in the case of manometric assays, which ordinarily
require one-half to one hour, but much less significant in our spectrophotometric
assays, which were completed within ten minutes. Thus, in the earlier study, it
appears likely that much activity would have been lost during the 1-40-day periods
of frozen storage prior to analysis, and during the manometric assays themselves.
There is ample reason to believe that our studies with AThCh yield information
with regard to substrate preferences, substrate-activity relationships, and inhibitor
sensitivities, that would also apply for ACh, as used in Van der Kloot's (1955)
study. Our confidence derives from a number of studies in which the behavior
of AThCh and ACh was compared, using the same ChE preparations (Heilbronn,
1959; Bergmann, Rimon and Segal, 1958; Elltnan ct al., 1961). Thus we suspect
that the 15 mM concentration of ACh used by Van der Kloot was supraoptimal for
AChE and inhibitory in effect. This follows from our finding that the optimal
AThCh concentration for pupal brain homogenates is only 0.75 mM, and that
activity is markedly reduced at higher concentrations in the manner typical of
AChE's (Fig. 2). Literature comparing substrate optima for AThCh and ACh
(Heilbronn, 1959; Bergmann, Rimon and Segal, 1958; Ellman et al., 1961),
though based on vertebrate preparations, shows the optima to be much closer than
the 20-fold difference separating our routine value for AThCh from that used by
Van der Kloot ( 15 mM) for ACh. Moreover, in our experience the 0.5 M sodium
chloride concentration incorporated in Van der Kloot's assay media yields a reac-
tion rate lower than with buffer alone, when rates at substrate optima are compared.
This conclusion concerning the effect of added salt conforms in large measure to
that of Wolfe and Smallman (1956) on brain ChE from flies, with ACh as sub-
strate. In the light of these arguments, and those of the preceding paragraph, we
conclude that the potential activity of ChE in brains from diapausing pupae was
not attained in earlier studies on Cecropia.
3. Neurophysiological status of the brain during pupal diapause
The present findings clearly show that at least one form of ChE persists in
neuropile throughout diapause. It now becomes important to know its localiza-
tion more precisely. In recent studies on the terminal abdominal ganglion of the
cockroach, Perlplaneta americana. Smith and Treherne (1965), using cytochemical
techniques at the electron microscope level, have defined several sites of esterase
activity. Eserine-sensitive esterase, presumably ChE, was found in association
with axonal membranes at apparent synaptic sites in neuropile. Other sites of
esterase activity were found both in and outside neuropile. If the neuropile-
associated esterase which we detect in brains of diapausing Cecropia occurs at
synaptic sites, and functions in transmission, then at least this element of neuronal
interaction would appear to remain patent throughout pupal diapause.
In view of this possibility, and in the light of the still-uncertain electrophysio-
logical status of the brain during diapause, a detailed reinvestigation by contem-
porary methods is clearly required. In studies of this type commenced recently,
Walcott (personal communication, 1965) has confirmed certain electrical activity
122 SHAPPIRIO, EICHENBAUM AND LOCKE
in the brains of diapausing silkmoth pupae; this supports the observations of
Schoonhoven (1963) and argues that any electrical "silence" of the diapausing
brain, as reported by Van der Kloot (1955) and Tyshtchenko and Mandelstam
(1965), must be restricted to certain brain regions if it occurs at all. It also now
becomes important to reassess the status of ACh itself following pupation, since
this (assayed as cholinergic substance effective upon clam heart in bioassays) was
reported to undergo precipitous disappearance at the outset of diapause, followed
by more gradual reaccumulation in unchilled and chilled pupae (Van der Kloot,
1955).
4. Multiple forms of esterase in relation to diapause and development
The arguments in Section 2 of this Discussion do not explain our failure to
detect large changes in esterase activity at the onset and termination of diapause,
as reported by Van der Kloot (1955). It is possible that our findings are com-
patible with those of Van der Kloot, but reflect the behavior of different esterases.
It has long been known that insect brains or heads contain a variety of esterases
(see reviews by Gilmour, 1961; Chadwick, 1963), though few data are available
on Lepidoptera. Recently, Maynard (1964) has characterized multiple esterases
in the nervous systems of crayfish and lobster. In the light of these seemingly
general attributes of arthropod central nervous systems, it is most probable that
silkmoth brain likewise contains esterases beyond the single ChE detected in the
present study. It is also worth noting that our use of AThCh at the rather low
optimal concentration of 0.75 mM may preclude detection of esterases other than
ChE's, or of ChE's having low activity toward the acetyl ester. Moreover, at
higher concentrations, ACh may be hydrolyzed via enzymes other than ChE's ; in
Van der Kloot's study ACh was used at 15 mM. Thus the prospect merits atten-
tion that the enzymatic changes described in his study are meaningful, and apply
to an esterase not detected in the present study. We look upon this prospect with
favor, since exploratory spectrophotometric and electrophoretic studies on Cecropia
brain, with various substrates, reveal several esterases including a component that
undergoes changes at the onset and termination of diapause.
We therefore conjecture that an esterase other than the A ChE described in
this report, and perhaps already manifested in Van der Kloot's (1955) study, will
be found to undergo changes that correlate with the neuroendocrine inactivation
and reactivation of the brain. Such a correlation between enzymatic and physio-
logical events would, of course, not in itself assure a causal role for the enzyme
in the control of neurosecretion and diapause. Nonetheless, attention will surely
center on its localization and properties in efforts to gain further insight into this
control at the molecular and subcellular levels.
Meanwhile, the results of the present study clearly oblige us to abandon the
attractive view that generalized disappearance and reappearance of ChE can account
for neuroendocrine changes in the brain, that in turn bring about the onset and
termination of pupal diapause. The history of biology is punctuated with occasions
where investigators, confronted with the need to revise an earlier theory, have
proceeded beyond their own data and sought to devalue the theory as a whole.
Slater (1958) discusses instances of this type in the history of the study of cellular
respiration. In present circumstances, we believe it prudent to continue to direct
CHOLINESTERASE IN SILKMOTH BRAIN 123
attention toward enzymatic events as part of the effort to understand the control
of neurosecretion and diapause.
We gratefully acknowledge helpful discussions and technical advice from Drs.
J. M. Allen, J. Martan, E. A. Maynard, and W. G. Van der Kloot. Dr. Charles
Walcott generously permitted us to cite his unpublished electrophysiological ob-
servations. We also thank Burroughs Wellcome, Inc., for gifts of ChE inhibitors.
R. S. Greathouse provided able technical assistance in a number of experiments.
SUMMARY
1. The localization and properties of cholinesterase in the brain of the Cecropia
silkmoth were investigated by histochemical and quantitative spectrophotometric
methods utilizing acetylthiocholine as substrate.
2. During pupal diapause, substantial activity was visualized in neuropile. At
the outset of adult development, activity was also detected in adjacent regions
occupied by neuronal or glial cell bodies.
3. Only one form of cholinesterase was detected with certainty. On the basis
of substrate-activity relationships for acetylthiocholine, propionylthiocholine and
butyrylthiocholine, and on the basis of its sensitivity to certain selective esterase
inhibitors, the enzyme has properties of an acetylcholinesterase.
4. Substantial and essentially unchanging enzymatic activity was detected dur-
ing pupation and most of pupal diapause, when the brain becomes endocrinologically
inactive. However, an approximate doubling in activity was detected during stor-
age of diapausing pupae at 6° C, apparently signalling the recovery of neuroendo-
crine competency by the brain. Subsequent growth and morphogenesis of adult
brain were found to be accompanied by a six-fold further increase in activity.
5. Cholinesterase activity also persists during diapause in the Cynthia and
Polyphemus silkmoths.
6. Consideration of the properties and optimal assay conditions for this enzyme
in pupal brain assists in explaining previous reports that it was undetectable.
7. The presence of substantial cholinesterase activity throughout metamorphosis
shows that a generalized disappearance and reappearance of the enzyme cannot be
responsible for inactivation and reactivation of the neurosecretory mechanism that
controls the onset and termination of diapause.
8. In the light of evidence for multiple forms of esterase in silkmoth brain, the
present findings do not preclude a possible role for one or more esterases as part
of the physiological mechanism controlling neurosecretion and diapause.
LITERATURE CITED
ALDRIDGE, W. N., 1953. Differentiation of true and pseudo cholinesterase by organophosphorus
compounds. Biochem. J., 53 : 62-67.
AUGUSTINSSON, K.-B., 1949. Substrate concentration and specificity of choline ester-splitting
enzymes. Arch. Biochem., 23: 111-126.
AUGUSTINSSON, K.-B., 1963. Classification and comparative enzymology of the cholinesterases,
and methods for their determination. In: Handbuch der Experimentellen Pharma-
kologie, Band 15, Cholinesterases and Anticholinesterase Agents (G. B. Koelle, sub-ed.),
Springer-Verlag, Berlin.
124 SHAPPIRIO, EICHENBAUM AND LOCKE
AUSTIN, L., AND W. K. BERRY, 1953. Two selective inhibitors of cholinesterasc. Biochem. J.,
54: 695-701.
BERGMANN, F., S. RIMON AND R. SEGAL, 1958. Effect of pH on the activity of eel esterase to-
wards different substrates. Biochem. J., 68: 493-499.
BULLOCK, T. H., AND G. A. HORRIDGE, 1965. Structure and Function in the Nervous Systems
of Invertebrates. W. H. Freeman and Co., San Francisco.
CHADXVICK, L. E., 1963. Actions on insects and other invertebrates. In: Handbuch der Ex-
perimentellen Pharmakologie, Band 15, Cholinesterases and Anticholinesterase Agents
(G. B. Koelle, sub-ed.), Springer- Verlag, Berlin.
CHADWICK, L. E., J. B. LOVELL AND V. E. EGNER, 1953. The effect of various suspension
media on the activity of cholinesterase from flies. Biol. Bull., 104: 323-333.
ELLMAN, G. L., K. D. COURTNEY, V. ANDRES, JR. AND R. M. FEATHERSTONE, 1961. A new and
rapid colorimetric determination of acetylcholinesterase activity. Biochem. PharmacoL,
7: 88-95.
EPHRUSSI, B., AND G. W. BEADLE, 1936. A technique of transplantation for Drosophila. Amcr.
Nat., 52:218-225.
GILBERT, L. L, 1964. Physiology of growth and development : endocrine aspects. /;; : The
Physiology of Insecta (M. Rockstein, ed.). Academic Press, New York.
GILMOUR, D., 1961. The Biochemistry of Insects. Academic Press, New York.
GOMORI, G., 1952. Microscopic Histochemistry ; Principles and Practice. University of Chi-
cago Press, Chicago.
HEILBRONN, E., 1959. Hydrolysis of carboxylic acid esters of thiocholine and its analogues.
3. Hydrolysis catalyzed by acetylcholine esterase and butyrylcholine esterase. Ada
Chem. Scand., 13: 1547-1560.
KOELLE, G. B., 1951. Elimination of diffusion artifacts in the histochemical localization of
cholinesterases, and a survey of their cellular distributions. /. PharmacoL Exp. Therap.,
103: 153-171.
KOELLE, G. B., 1954. The histochemical localization of cholinesterases in the central nervous
system of the rat. /. Comp. N enrol., 100: 211-228.
KOELLE, G. B., 1963. Cytological distributions and physiological functions of cholinesterases.
In: Handbuch der Experimentellen Pharmakologie, Band 15, Cholinesterases and
Anticholinesterase Agents (G. B. Koelle, sub-ed.), Springer- Verlag, Berlin.
MAYNARD, E. A., 1964. Esterases in crustacean nervous system. I. Electrophoretic studies in
lobsters. /. E.rp. Zoo!., 157: 251-266.
PEPLER, W. J., AND A. G. E. PEARSE, 1957. The histochemistry of the esterases of rat brain,
with special reference to those of the hypothalamic nuclei. /. Ncurochem., 1 : 193-202.
SALKELD, E. H. 1961. The distribution and identification of esterases in the developing embryo
and young nymph of the large milkweed bug, Oncopeltus fasciatus (Ball.). CanaA. J.
Zool.,39: 589-595.
SCHNEIDERMAN, H. A., AND C. M. WILLIAMS, 1954. The physiology of insect diapause. IX.
The cytochrome oxidase system in relation to the diapause and development of the
Cecropia silkworm. Biol. Bull., 106: 238-252.
SCHOONHOVEN, L. M., 1963. Spontaneous electrical activity in the brain of diapausing insects.
Science, 141: 173-174.
SHAPPIRIO, D. G., AND C. M. WILLIAMS, 1957. The cytochrome system of the Cecropia silk-
worm. I. Spectroscopic studies of individual tissues. Proc. Ro\. Soc. London, Scr. B,
147:218-232.
SHAPPIRIO, D. G., D. M. EICHENBAUM AND B. R. LOCKE, 1965. Cholinesterase in the brain
of the Cecropia silkmoth in relation to the control of neurosecretion and diapause.
Amer. Zool, 5: 698-699.
SHEN, S. C., P. GREENFIELD AND E. J. BOELL, 1955. The distribution of cholinesterase in the
frog brain. /. Comp. N enrol., 102: 717-743.
SLATER, E. C., 1958. Catalytically active hemoproteins, with special reference to the cyto-
chromes. Biochemical Society Symposia, 15: 100-101.
SMITH, D. S., AND J. E. TREHERNE, 1965. The electron microscopic localization of cholin-
esterase activity in the central nervous system of an insect, Pcriplaneta ainericaiia L.
J. Cell Biol., 26: 445-459.
CHOLINESTERASE IN SILKMOTH BRAIN 125
TVSHTCHENKO, V. P., AND J. E. MANDELSTAM, 1965. A study of spontaneous electrical activity
and localization of cholinesterase in the nerve ganglia of Antheraca ficrnyi Guer. at
different stages of metamorphosis and in pupal diapause. /. />/.?. PhysioL, 11: 1233-1239.
VAN UER KLOOT, W. G., 1955. The control of neurosecretion and diapause by physiological
changes in the brain of the Cecropia silkworm. Biol. Bull., 109: 276-294.
WIGGLESWORTH, V. B., 1958. The distribution of esterase in the nervous system and other
tissues of the insect Rhodiiius proli.ms. Quart. J. Micr. Sci., 99: 441-450.
WIGGLESWORTH, V. B., 1964. The hormonal regulation of growth and reproduction in insects.
Adi'an. Ins. PhysioL, 2: 247-336.
WILLIAMS, C. M., 1946. Physiology of insect diapause : The role of the brain in the production
and termination of pupal dormancy in the giant silkworm, Platysamia cecropia. Biol.
Bull., 90: 234-243.
WILLIAMS, C. M., 1952. Physiology of insect diapause. IV. The brain and prothoracic glands
as an endocrine system in the Cecropia silkworm. Biol. Bull., 103: 120-138.
WILLIAMS, C. M., 1956. Physiology of insect diapause. X. An endocrine mechanism for the
influence of temperature on the diapausing pupa of the Cecropia silkworm. Biol. Bull.,
110: 201-218.
WILLIAMS, C. M., AND P. L. ADKISSON, 1964. Physiology of insect diapause. XIV. An endo-
crine mechanism for the photoperiodic control of pupal diapause in the oak silkworm,
Antheraca pcrnyi. Biol. Bull, 127: 511-525.
WOLFE, L. S., AND B. N. SMALLMAN, 1956. The properties of cholinesterase from insects. /.
Cell. Comp. PhysioL, 48: 215-225.
THE WATER ECONOMY OF SALAMANDERS: EXCHANGE OF
WATER WITH THE SOIL
TOM M. SPIGHT 1
Department of Zoology, Duke University, Durham, North Carolina
Many species of amphibians spend their active periods at some distance from
surface waters, and, consequently, cannot depend entirely upon ponds, pools, and
streams for water to replace evaporative losses. These animals probably depend
upon the water stored in the soil, and the availability of this water to them is an
important parameter of their ecology.
Heatwole and Lim (1961) have shown that, at some water contents, the water
in the soil was available to the salamander, Plcthodon cinereus. They also showed
that the lowest soil water level at which the salamander could withdraw moisture
(the "absorption threshold") had the same moisture tension value in several soils.
They thus introduced the moisture tension scale as a scale of the availability of
soil water to salamanders.
In this study, I investigated the exchange of water between soil and salamanders
in the soil, using six salamander species and a range of soil moistures. The fol-
lowing questions were investigated : ( 1 ) How does the rate of water exchange vary
with the soil moisture? (2) What differences are there between different sala-
mander species? (3) What variables other than the species and the soil moisture
affect the exchange rates? (4) Can a salamander rehydrate fully in soil? With
the data from these experiments, it is possible to predict the availability of water
to these salamanders at any particular site.
MATERIALS AND METHODS
The soil used was a homogeneous mixture of several local soils. Its character-
istic curve, as determined on a porous plate apparatus (Soil Moisture Equipment
Company; see Marshall, 1959, Ch. 2) is given in Figure 1. Lots of soil contain-
ing 10%, 6%, 4% and \% water were packed into culture dishes. These lots
provided a range of moistures from above field capacity (soil containg about 9%
water) to air-dry soil (containing about 1% water).
The animals were dehydrated in air until they had lost approximately 15% of
their initial body weight (this loss, expressed as a percentage of the initial weight,
is the "dehydration deficit"). Each animal was then buried in a culture dish
filled with soil at one of the four moisture contents. The animals were removed
after six hours, and a soil sample was taken. Each salamander's bladder was
emptied, and the salamander was then weighed to the nearest 5 mg. on a torsion
balance (in the experiments with air-dry soils, the salamanders were not initially
dehydrated, and were left in the soil only 2, 3 or 4 hours). The weight change
1 Present address : Department of Zoology, University of Washington, Seattle, Washington
98105.
126
SALAMANDERS AND SOIL WATER
127
over the six-hour interval was assumed to be clue to water exchange, and was
expressed as a rate [mg./(cm.2 X day)] where the surface area was calculated
with Benedict's (1932) general formula:
(Surface Area) = 10 (Body Weight)2/3.
The soil sample was dried for 48 hours at 105° C. The soil moisture was calcu-
lated from the change in weight. Figure 2 shows the relation between soil moisture
and exchange rate for this initial 6-hour period.
Some of the animals which were losing weight were returned to the soil and
exposed for an additional 12- or 24-hour period. The weight changes over this
period were used to determine the variation in water loss with the length of the
exposure period.
All the animals which gained weight during the first 6-hour interval were
returned to the soil. These salamanders were removed for weighing at 24-hour
intervals until they had stopped gaining weight. The peak weight attained,
CO
LLJ
CO
14
,_
12
Q 8
CO
LLJ 6
C£
§4
8
2 4 6 8 10
SOIL WATER CONTENT (%OF DRY WEIGHT)
FIGURE 1. Characteristic curve of the soil used in the experiments.
Each point is the mean of three values.
128
TOM M. SPIGHT
+ 40
o
T3
O
-40
Ld
o
X
O
X
w
Ul
V)
O
U.
O
Ul
-80
-120
-160
-200
I I I 1
I I
D
0
a PL.JORDANI
o D. FUSCUS
+ D. MONTICOLUS
• D. QUADRAMACULATUS
o PS RUBER
T A. OPACUM
I
0 2 4 6 8 10 12
SOIL WATER CONTENT(PERCENT OF DRY WEIGHT)
FIGURE 2. The relationship between soil water content and the rate of water exhange by
six species of salamanders. The weights and dehydration deficits of the animals used in these
experiments are given in Spight (1966a).
expressed as a percentage of the original weight, was the rehydration success of
the animal.
All animals spent a final 48 hours in tap water. A weight increase during
this period was assumed to indicate that the animals were unable to complete
rehydration in soil.
Six species of salamanders were used in these experiments : the terrestrial,
lunged Ambystoma opacum, and these lungless salamanders: Desmognathns quad-
ramaculatus, from mountain streams; D. monticolus, from mountain stream banks;
D. juscus, from Piedmont streams and stream banks; Plethodon jordani, a fonest
floor form with no aquatic phase, and Pseudotrlton ruber, a terrestrial form with
an aquatic phase.
SALAMANDERS AND SOIL WATER 129
RESULTS
Moisture tension and species as variables
The rate of water exchange with unsaturated soil is a function of the water
content, and the function is the same for all of the species tested (Fig. 2).
The soil moisture tension is presumably a measure of water availability to
salamanders, and, if Figure 1 is superimposed on Figure 2, it can he seen that the
exchange rate is a function of the moisture tension. The moisture tension relation-
ship will remain constant from soil to soil, and rates observed in these experi-
ments, expressed as a function of the soil moisture tension, wrill serve as a basis
for predicting performance in other soils. The water content allowing a particular
exchange rate will, on the other hand, vary with the composition of the soil, and
can be predicted for another soil only when the characteristic curve is available
for that soil.
The range of moistures of critical importance to a salamander is the range
about its absorption threshold, since it is only in soil above that threshold that a
salamander can potentially remain in water balance. Within this range, the only
variable which determines the rate and direction of water exchange over the six-
hour measurement interval is the water content of the soil. The animals clustered
about the 2 atm. (about 5%) point have a wide range of weights (1 g. to 10 g.)
and dehydration deficits (8% to 20%) but all exchange water with the soil at rates
solely dependent upon the water content of the soil (Fig. 2).
I
Additional variables in water exchange
Soil moisture was the only variable which affected the rate and direction of
water flow between salamanders and soils when the soils had moisture tensions
between 0.3 atm. and 15 atm. and when the measurement intervals were equal.
In the saturated soils, however, the salamanders' rate of water uptake is correlated
with both weight and dehydration deficit. The exchange rate by an individual
salamander over a six-hour period in soil was also markedly different from that
over a 12-hour period.
If the data from the two individuals of each species which rehydrated in sat-
urated soil are compared, the differences between the members of each pair are
correlated with differences in body weight and dehydration deficit. For examples,
the larger of the pair of D. fuscus had the lower rate of gain, both animals having
the same dehydration deficit. For the pairs of D. inonticolus and Plethodon jor-
dani of the same weight, the animal in each pair 'with the greatest dehydration
deficit had the greatest rate of uptake. For pairs of salamanders of the species
Psendotriton ruber and A. opacum, the smaller animal had the greater deficit and
absorbed water from the soil more quickly. The data for D. quadramaculatus are
ambiguous. In all cases, for animals with similar dehydration deficits, the rates
of rehydration in soil were well below the average rates of rehydration in water
[rates in water ranged from 24 mg./(cm.2 X day) to 190 mg./cm.2 X day) ; see
Spight, 1966b].
The rate of loss by salamanders to air-dry soil (by the 12 animals with the
highest loss rates in Figure 2) is inversely correlated with the time the animals
spent in the soil. The four animals which spent only two hours in soil had an
130
TOM M. SPIGHT
average rate of loss of 190 mg./(cm.2 X day) ; all their rates fall within the range
of rates of loss by animals which were dehydrated in room air [for 60% to 70%
relative humidity at 20° C., animals lost 100 to 310 mg./(cm.2 X day) by evapora-
tion to the air; see Spight, 1966a]. In contrast, animals which spent 3 hours or
4 hours in soils of the same water content averaged, respectively, 140 and 110
mg./(cm.2 X day).
The mediating influence of the soil is particularly prominent in the air-dry
soils, for in these soils a small amount of water lost by a salamander to the sur-
rounding soil will lower the tension of that soil markedly (note the slope of the
characteristic curve in Figure 1 for soils containing less than 4% water). The
mediation was also evident in the moist soils. When one individual spent suc-
cessive intervals of 6 and 12 hours in the same soil, the loss rate was appreciably
lower over the longer interval (Table I).
TABLE I
Rates of water loss by individuals to moist soils during successive
periods of different lengths in the same soil
Rate of water loss
Species
Hours 1 to 6 mg./
Hours 6 to 18 (mg./
Soil moisture %
(cm.s X day)
(cm.2 X day)
Desmognathus fuscus
-18.2
-13.8
3.71
D. monticolus
-16.4
-11.4
3.46
D. monticolus
-25.0
- 9.4
3.70
D. quadramaculatus
-18.6
-11.4
3.42
D. quadramaculatus
-18.5
- 8.9
4.00
Plethodon jordani
-19.9
-10.0
3.64
Plethodon jordani
-27.5
-10.5
3.78
Ambystoma opacum
-16.6
-13.0
3.62
Pseudotriton ruber
-12.9
- 8.3
3.58
Pseudotriton ruber
-13.1
- 8.6
3.91
Rehydration ability
The salamanders can complete rehydration in saturated soil, although the rate
of gain is slow in comparison with rates of rehydration in water. Twelve animals
attained 93.2% of their initial weights (SE = 0.6) from soils with 9% to 11%
water, taking one to four days to complete the rehydration. These animals made
no further weight gains during 48 hours in water.
Animals exposed to soils containing 5% to 6% water showed initial weight
gains, but these soils dried below the absorption threshold before the animals were
able to complete rehydration. These animals subsequently attained 92.4% of their
original weights by rehydration in water (SE = 1.1, N = 12) ; this percentage is
not significantly different from the percentage attained by the group of salamanders
which were able to complete rehydration in soil (P > 0.50 that the difference has
arisen by chance alone).
It is postulated that a salamander can complete rehydration in any soil from
SALAMANDERS AND SOIL WATER 131
which it can gain any water, although it is doubtful that this can be shown in the
laboratory, since it is extremely difficult to maintain soil at a precise moisture tension.
DISCUSSION
Salamanders are found in both "wet" and "dry" communities, and species can
lie characterized by their "water requirements." Some species are wholly aquatic,
others semi-terrestrial, and others wholly terrestrial. Investigators, including
G. K. Noble (1931), have suggested that the distributions of different amphibian
species might be correlated with specific differences in their ability to absorb water
from various substrates. This study has shown that if the soil water at a par-
ticular site is available to a terrestrial species, it will also be available to a char-
acteristically aquatic species. In other words, in spite of well documented differ-
ences in "water requirements," knowledge of the availability of water in the substrate
of a habitat cannot be used to make predictions about which salamander species
will be able to occupy that habitat. Salamanders thus present another example
of Beament's (1961) generalization that, among closely related animals, even
species from quite different habitats have only minor differences in their physiologies.
Salamanders can absorb water from soils with moisture tensions less than 2
atm. The meaning of this value in typical natural situations can be clarified by
pointing out three particular points on the moisture tension scale. Soil is nor-
mally considered (1) to be at a tension of 0.33 atm. when the gravitationally-
induced draining of a rain-wet soil is complete, and (2) to be at a tension of 15
atm. when a plant growing in the soil becomes permanently wilted. Van Bavel
(1953) gives the third point (3) : on non-irrigated crop land near Raleigh, N. C.,
during the average year there will be 20 days during which the soil moisture rises
above 0.77 atm. (he considers this tension to be a stress level for agricultural
plants). It may be seen, then, that salamanders can obtain water from agricultural
land in North Carolina throughout most of the year.
Soil properties and moisture exchange
As the time the salamander spends in soil below the absorption threshold in-
creases, the rate of water loss by the salamander to the soil drops. This rate drop
is more prominent in the air-dry soils, but it is also evident in the wetter soils.
These rate drops reflect the water conductance of the soil used, and a related phe-
nomenon, the wetting front.
In unsaturated soils (soils with tensions greater than 0.33 atm.), a wetting
front, a locally steep gradient of moisture content, is formed between a wet region
and a drier region (Klute, 1952). The gradient is such that in very dry soil there
may be no movement of liquid water from a wet region (at about field capacity)
to an adjacent, very dry region (air dry; Bodman and Colman, 1944).
As water is lost by the salamander to the soil, it is absorbed by the immediately
adjacent soil, and a wetting front is formed. The water lost by the salamander
accumulates between the salamander and the edge of the wetting front. In this
zone, the humidity increases with the accumulation of water ; thus the accumulated
water lowers the humidity gradient from the salamander to the adjacent soil, and
the evaporation rate of the salamander decreases.
132 TOM M. SPIGHT
Soil water at some moisture tensions is available to many salamander species,
and the availability of the soil water at a particular site to a salamander can be
predicted. With this information, it is possible to approach old problems, such
as "does water act as a limiting factor in the distribution of salamanders," and
"what are the water problems of hibernating salamanders?" These experiments
have not shown that anurans have the same absorption threshold as do these
urodeles. Such a demonstration would make an interesting future study, and
would lead directly to the solution of the problem, "where do desert toads get
their water?"
I would like to express my gratitude to Dr. Paul J. Kramer and Dr. Peter J.
Bentley for advice and encouragement during the course of this study. I am
indebted to Nancy Nickerson for many of the animals, to Dave Hillier for the
moisture tension curve, and to Lee Miller for the soil. Dr. Bentley and Dr. Peter
W. Hochachka were kind enough to read the manuscript. Dr. J. R. Bailey super-
vised the work and identified the animals.
SUMMARY
Specimens of six salamander species were exposed to different soil moistures,
and rates of water exchange were calculated from the changes in weight observed.
Soil water was available to these salamanders in soils with moisture tensions be-
tween 0 atm. and 2 atm. At these tensions, salamanders could rehydrate fully,
and can therefore be expected to remain in water balance in any soil with a tension
in this range. The measurements should be useful in determining the suitability
of habitats for salamanders. The rate of water exchange between the salamander
and the soil was a function of the soil moisture tension. The rate of uptake from
saturated soils was correlated with the body weight and the dehydration deficit of
the salamanders. There were no differences in rate of exchange or in absorption
threshold between the different species. These characteristics of water exchange
between salamanders and soil are related to the properties of the soil.
LITERATURE CITED
BEAMENT, J. W. L., 1961. The role of physiology in adaptation and competition between ani-
mals. Sytnt>. Soc. Exp. Biol., 15: 62-71.
BENEDICT, F. G., 1932. The physiology of large reptiles. Carnegie Inst. Washington Publ. 425.
BODMAN, G. B., AND E. A. COLMAN, 1944. Moisture and energy considerations during down-
ward entry of water into soils. Soil Sci. Soc. Amer. Proc., 8: 116-122.
HEATWOLE, H., AND K. LIM, 1961. Relation of substrate moisture to absorption and loss of
water by the salamander Plethodon cinereus. Ecology, 42: 814-819.
KLUTE, A., 1952. A numerical method for solving the flow equation for water in unsaturated
materials. Soil Science, 73: 105-116.
MARSHALL, T. J., 1959. Relations between water and soil. Commonwealth Bureau of Soils,
Harpenden, England. Technical Communication 50.
NOBLE, G. K., 1931. The Biology of the Amphibia. McGraw Hill, N. Y.
SPIGHT, T. M., 1966a. Studies on the water relations of salamanders. M.A. thesis, Duke
University.
SPIGHT, T. M., 1966b. The water economy of salamanders : water uptake after dehydration.
Comp. Biochem. Physiol. (in press).
VAN BAVEL, C. H. M., 1953. A drought criterion and its application in evaluating drought
incidence and hazard. Agronomy J., 45: 167-172.
STUDIES ON THE TREMATODE GENUS PARAMONOSTOMUM
LUHE, 1909 (DIGENEA: NOTOCOTYLIDAE) x
HORACE W. STUXKARD
The American Museum of Natural History, Central Park West at 79th Street,
Nciv York, Netv York 10024
The genus Paramonostomum was erected by Liihe (1909) with Monostoma
alveatum Mehlis in Creplin, 1846 (syn. Monostoma aheifonnc Cohn, 1904) as
type. The species had been included by Monticelli (1892) in the genus Notoco-
tylus Diesing, 1839, but Liihe predicated that it is not congeneric with Notocotyhts
triserialls Diesing, 1839, type of Notocotylus. The species, P. alveatum, has been
reported from a large number of birds including Anas spp., Anser anser, Nyroca
marilla, Oedemia spp., Somateria mollissima, Cygnus spp., Branta spp., and Clan-
yula hyemalis. Some 20 additional species of Paramonostomum have been de-
scribed but distinctions between certain of them are very tenuous.
One life-cycle, that of P. alveatum, was reported by Kulachkova (1954). The
work was done at the marine station on Kandalaska Bay, in the southwest portion
of the White Sea, longitude 33° East and latitude 65.5° North. Hydrobia ulvae
was the intermediate host and harbored the asexual generations of the parasite.
Mme. Kulachkova published two short papers (1961a, 1961b) on seasonal infec-
tion of the mollusks and on the biology of the larval stages of P. alveatum. I am
indebted to Dr. GaltsofT, who graciously translated the Russian texts for me. The
studies of Mme. Kulachkova were occasioned by the mass mortality of young eider
ducks; in the period 22 June to 7 July, 1949, 321 chicks died from the infection.
As many as 50,000 worms were found in a single bird. The parasites penetrated
between the intestinal villi, with inflammation and destruction of the epithelium and
membranes. Fourteen per cent of the H. ulvae in the tide-pools were infected and
the cercariae, on emergence, encysted promptly on the shells of the snails from
which they had emerged. The cysts were 0.155 mm. in diameter and the worms
matured in 6-8 days in the birds. The tide-pools had sandy-gravelly bottoms and
the eider chicks, less than two weeks old, fed in these tide-pools where the shells
of the hydrobias carried from 10 to 25 cysts per snail. Birds older than two
weeks, fed in Fucus and mussel beds where the hydrobias were rare or absent, and
birds older than two weeks survived. In the examination of 5427 snails over a
four-year period, the rate of infection varied from 3.3% to 12%, with the greatest
incidence in July and August. Shedding of the larvae began at water tempera-
ture of 23° and massive discharge in the range 23° to 26°. This was usually in
the last week of June and first 10 days of July, although the dates varied with
weather conditions, but this was the time when the eider chicks were feeding in
the tide-pools.
1 Investigation supported by NSF, GB-3606, continuation of G-23561.
133
134 HORACE W. STUNKARD
Miriam Rothschild (1938) had studied the ccrcariae of the notocotylid trema-
todes and described six species of these cercariae from the snails, Peringia nlvac
and Hydrobia ventrosa. They were assigned to three different groups, based on
the form of the excretory system: the MONOSTOMI group, in which the anterior
transverse portion of the vesicle is a closed tubular circuit situated posterior to the
median eye-spot and cerebral ganglion; the IMBRICATA group, in which the
anterior portion of the vesicle forms a loop, between the eye-spots, which passes
anterior to the median eye-spot and cerebral ganglion; and the YENCHINGEN-
SIS group, in which there is an unpaired finger-like diverticulum which extends
anteriad from the transverse portion of the vesicle.
Stunkard (1965) reported on the examination of more than 4000 specimens
of Hydrobia salsa (Pilsbry, 1905) taken from brackish- water ponds, near Woods
Hole, Massachusetts, and the discovery of at least four species of notocotylid cer-
cariae, including the larval stages of Paramonostomum alveatum. The study has
been continued and a fifth notocotylid species has been identified (Stunkard, 1966a).
Two species have Yenchingensis type excretory systems ; the metacercariae develop
in the intestinal caeca of chicks and of domestic and eider ducklings, and belong to
the genus Notocotylus. One of these species is identical with Notocotylus minutus
Stunkard, 1960 and the other is described as a new species (Stunkard, 1966b).
One of the five cercariae belongs to the Imbricata Group ; it develops in the bursa
Fabricius of chicks and domestic ducklings and may be identical with Uniserialis
gippyensis Burton, 1958. The two other species have Monostomi type excretory
systems ; the metacercariae develop in the intestine of chicks and both domestic and
eider ducklings, and belong to the genus Paramonostomum. One is identified as
P. alveatum; and the other is Paramonostomum parvum Stunkard and Dunihue,
1931.
The methods for study of the Paramonostomum species are identical with those
described for the species of Notocotylus (Stunkard, 1966b). The procedure is
relatively simple. The infected snails were discovered by isolation ; the larvae were
studied alive, with and without the use of vital dyes. All stages were fixed and
stained for subsequent study. Duboscq-Brasil and AFAG mixtures were used for
fixation and specimens prepared as whole mounts were stained with Mayer's
paracarmine, Semichon's acetic carmine, or Ehrlich's acid haematoxylin. Adult
worms, sectioned serially in transverse and frontal planes, were stained with haema-
toxylin and erythrosin. Snails were isolated singly and metacercariae, encysted on
the light side of the container, were fed at two-day intervals to rats, mice, hamsters,
young herring gulls, chicks and both domestic and eider ducklings. The eider
chicks were provided through the kindness of Mr. Walter Welch and his associates
on the staff of the U. S. Fish and Wildlife Service, Boothbay Harbor, Maine.
The snails were killed to identify the larval stages and the final hosts were killed
to recover the developing and sexually mature worms. No infection was obtained
in a mammalian species. Both species of Paramonostomum developed to maturity
in chicks and in both eider and domestic ducklings, but sexually mature worms were
not recovered from gulls. The intestine of a gull killed two days after the inges-
tion of about 200 metacercariae contained a few dead, excysted worms. Three live,
juvenile specimens, identified as P. parvum, were recovered from the intestine of
another gull fed metacercariae on 29 June, 1964, and autopsied 22 July, 1964.
THE TREMATODE GENUS PARAMONOSTOMUM 135
The worms were the same si/.e ; fixed and stained they are 0.33 mm. long and 0.18
mm. wide, with very large ovaries and testes, but without eggs in their uteri. The
two species of Paramonostomum are distinguished primarily by differences in size
and size of organs. There are constant and apparently significant differences in
sizes of cercariae, metacercariae and sexually mature adults. Twelve specimens of
P. alveatum were left August 30, 1966, for several hours in pond-water and a large
number of eggs were expelled by the worms. Three days later six young labora-
tory-reared H. salsa were placed in the dish with the eggs and were observed to
eat some of them. The snails were then removed to fresh pond-water and a snail
sacrified on October 3, 1966, contained rediae, some of which contained developing
cercariae. The experimental infection of laboratory-reared H. salsa completes the
life-cycle and confirms the earlier observations. The findings of the present study
confirm the account of Mme. Kulachkova on P. alveatum.
DESCRIPTIONS
Paramonostomum alveatum (Mehlis in Creplin, 1846)
(Figs. 4-6)
Previous accounts include the inadequate redescription of the original specimens
by Monticelli (1892), the brief statement by Liihe (1909) when he erected the
genus Paramonostomum, and the account by Kossack (1911).
Adult (Fig. 4)
The worms are ovate, rounded posteriorly, more pointed anteriorly. Typically,
the edges of the body are turned ventrad and mediad, so the venter forms a cupuli-
form cavity, which suggested the specific name, alveatum. The opening is smaller
than the outline of the body. Fixed and stained sexually mature specimens measure
0.50 to 0.85 mm. in length and 0.40 to 0.53 mm. in greatest width, which is in
the posterior half of the body. Under pressure of a coverglass, the dimensions of
living worms are much greater. The cuticula appears to be smooth, but examina-
tion of living specimens under high magnification discloses exceedingly minute,
closely set spines, arranged in parallel rows, on the ventral surface. The muscula-
ture of the body wall is weak and movement is slight or sluggish. The pigment
from the ocelli of the cercaria persists in the parenchyma of the anterior end of the
body. The oral sucker is 0.06 to 0.065 mm. in diameter; the esophagus is short,
about the length of the oral sucker ; the caeca are dorsal in location and follow the
lateral contours of the body ; they turn mediad at the anterior ends of the testes,
pass between the testes and ovary, and terminate blindly behind the level of the
gonads. As noted by Rothschild (1941, p. 363, fig. 1), "it is well known that
in mature notocotylid trematodes the excretory vesicle becomes greatly complicated."
Essentially, large, ramifying dendritic branches arise from the lateral and medial
sides of the ring formed by the fusion of the collecting ducts of the cercaria, and
constitute a complex reticulum that permeates the parenchyma of the body. The
bladder, situated posterior to the ring, opens to the surface by a dorsal pore near
the caudal end of the body.
The testes are oval, lobed organs, situated in the extracaecal areas at the poste-
136
HORACE W. STUNKARD
I <b
FIGURE 1. Paramonostomnm parvuin, adult specimen, ventral view, from eider duck; worm,
0.37 mm. long.
FIGURE 2. P. parvuin, redia, pressed specimen, 0.93 mm. long.
FIGURE 3. P. parvuin, cercaria, ventral aspect, fixed without pressure in hot whirling
solution (Duboscq-Brasil), body, 0.17 mm. long.
FIGURE 4. P. alveatum, adult specimen, ventral view, from eider duck ; worm, 0.75 mm. long.
FIGURE 5. P. alveatum, redia, fixed without pressure, 0.82 mm. long.
FIGURE 6. P. alveatum, cercaria, ventral aspect, fixed under pressure of coverglass, body
0.32 mm. long.
THE TREMATODE GENUS PARAMONOSTOMUM 137
rior end of the body. They measure 0.12 to 0.18 by 0.10 to 0.14 mm. in diameter.
Sperm-ducts arise at the anteromedian faces and unite in front of Mehlis' gland to
form the vas deferens which passes anteriad, dorsal to the uterine coils. About
one-third of the body-length from the anterior end, the duct enlarges to form a large,
coiled, external seminal vesicle, the last loop of which extends forward on the right
side of the cirrus sac. The cirrus sac is curved, short and broad, 0.16 to 0.24 mm.
long and 0.10 to 0.13 mm. wide; the posterior portion contains the large internal
seminal vesicle; the pars prostatica is short, and the cirrus is eversible (Fig. 4).
The ovary is median, between the testes. It is a variably lobed organ, 0.08 to
0.15 mm. in diameter. The oviduct arises at the anterodorsal face of the ovary
and receives a short, common vitelline duct as it enters Mehlis' gland, which is
smaller and immediately anterior to the ovary. Mehlis' gland contains the ootype,
in which the eggs are formed. The initial coils of the uterus are filled with sperma-
tozoa. There are 10 to 12 transverse uterine loops that occupy the intercecal area
posterior to the cirrus sac. They are ventral to the digestive caeca and in pressed
specimens the vitellaria are almost contiguous with the uterine loops. The metra-
term is ventral to the cirrus sac and about one-half its length. The vitellaria are
extracaecal and extend from the testes to the level of the cirrus sac. The eggs are
operculate, have long polar filaments, measure 0.019 to 0.021 by 0.011 to 0.012 mm.,
and are embryonated when passed.
Rcdia (Fig. 5)
The rediae are vermiform, cylindrical to sausage-shaped, and vary in size from
young, colorless individuals with small germ balls to large, gravid rediae that may
extend to a length of 1.20 mm., and when retracted may be 0.50 mm. in width. The
size of the redia is largely determined by the number and size of the progeny in it.
The large rediae are golden-yellow to orange in color. The body has annular and
longitudinal muscles and when the longitudinal muscles are contracted, the wall
has a crenate margin. The pharynx increases to 0.055 mm. in diameter and the
esophagus is about the same length. In young rediae the intestine may extend
more than one-half the length of the body, but it becomes relatively shorter as the
redia is filled with offspring. The intestine is filled with decomposing snail tissue,
yellow droplets and blackish amorphous material. The excretory pores are lateral,
in the posterior half of the body. From each pore a common duct passes forward
for a short distance and divides into anterior and posterior branches, each of which
terminates in a flame-cell. The collecting ducts are coiled and are enclosed in
straight-walled sinuses as described by Rothschild (1935).
Cercaria (Fig. 6)
The cercariae leave the rediae before they are mature and complete their devel-
opment in the haemal sinuses of the snail. They emerge from the snails princi-
pally between 10 AM and 2 PM, and swim ordinarily for one to three or four hours.
They encyst on the shell or operculum of the snail from which they emerged or
on any hard surface. When irritated, e.g., placed in a solution of vital dyes, they
may encyst at once. When swimming, the body is contracted, bent ventrally; the
tail is extended and lashes violently. They are photosensitive and accumulate on
138 HORACE W. STUNKARD
the light side of the container. The body is oval to ovate, more pointed anteriorly,
convex dorsally and concave ventrally ; contracted it is circular, 0.20 to 0.25 mm.
in diameter; elongated it may be 0.38 by 0.15 mm. The tail is simple, 0.20 to
0.50 mm. long and when extended it is 0.02 to 0.03 mm. wide at the base. When
the body is extended the tail is contracted and vice versa. The posterolateral ends
of the body bear eversible and retractile locomotor appendages, 0.020 to 0.025 mm.
in diameter; when the body is extended they are close together, separated only
by the base of the tail which is ventral to them ; when the body is retracted they
are at the posterior, dorsolateral corners. While in the redia, the cercariae have
only two ocelli, with scattered pigment around the eye-spots, but by the time of
emergence often there is a third, median, ring-like condensation of pigment between
the ocelli, and dendritic strands of pigment extend posteriad along the digestive
caeca. The ocelli are 0.018 to 0.020 mm. in diameter; they are provided with
lenses and are situated dorsal and anterior to the ganglia of the nervous system.
The parenchyma becomes rilled with unicellular cystogenous glands ; the secretion
is in the form of bacilliform rods, 0.003 to 0.004 mm. in length and about one-half
as wide. The oral sucker is 0.038 to 0.045 mm. in diameter ; the esophagus is
about the same length and crosses the excretory ring dorsally, whereupon it opens
into the caeca. The caeca are dorsal and medial to the excretory ring as they pass
posteriad, but near the posterior end of the body they turn laterad across the ring
and end blindly as shown in the figure. The development of the excretory system
of the cercaria agrees completely with the account of Rothschild (1935) on C.
ephemera Lebour, 1907. In young cercariae the excretory pores are on the sides,
near the middle of the tail, but in mature larvae, the portion of the system in the
tail atrophies and a new excretory pore develops from the dorsal wall of the excre-
tory bladder. The ring, formed by the coalescence posteriorly and anteriorly of
the collecting ducts, is filled with concretions. In the posterior part of the ring
there may be four to six concretions at any level, while in the anterior part of the
ring the concretions may be disposed in a single row. They measure 0.003 to 0.006
mm. in diameter and often two or three are fused.
Metacercaria
In encysting, the cercaria attaches by the oral sucker ; the body is contracted to
circular form, and the cystogenous material is extruded on all sides. As the secre-
tion hardens, the tail, which is left outside, lashes itself free and the cyst, 0.15 to
0.16 mm. in diameter, is firmly attached to the sul (stratum. The cyst wall is rela-
tively impermeable and resists desiccation; the larva moves in the cyst and if not
dried, remains infective for a long period, weeks, possibly months.
Paramonostomum parvum Stunkard and Dunihue, 1931
(Figs. 1-3)
This species, described originally from specimens found in the intestine of an
unidentified duck, was recovered after feeding metacercariae to laboratory-reared
eider and domestic ducklings and to day-old chicks. The asexual generations were
found in Hydrobia salsa taken from Nobska and Oyster Ponds, brackish-water
areas that communicate with Vineyard Sound, near Woods Hole, Massachusetts.
THE TREMATODE GENUS PARAMONOSTOMUM 139
The specimens of experimental infection agree completely with the description of
worms of natural infection as reported by Stimkard and Dunihue (1931).
Adult (Fig. 1)
The worms measure 0.25 to 0.50 mm. in length and 0.20 to 0.35 mm. in width.
Only much flattened specimens exceed 0.50 mm. in length. In the original report,
the presence of spines on the cuticula was regarded as doubtful. With an abun-
dance of material, it has been possible to observe the presence of exceedingly
minute, closely-set spines on the ventral surface of the body. The spines are
arranged in fine, parallel rows. They are not visible on fixed and stained speci-
mens, but on living worms under high magnification they can be resolved by
careful focussing.
Redia (Fig. 2)
The rediae closely resemble those of P. alveatum; they occupy the haemal
sinuses of the snail and grow to a size of 0.65 by 0.13 mm. or 0.75 by 0.10 mm.
Daughter rediae emerge from the parental rediae while very small, much smaller
than the cercariae when they emerge, and while small they are active and migra-
tory. As the body cavity becomes filled with germ-balls and developing cercariae,
the rediae become more sluggish. The pharynx increases to a diameter of 0.035-
0.042 mm., and when pressed may measure 0.05 mm. in diameter. The esophagus
is approximately as long as the pharynx and the caecum varies with the size of
the redia. In young specimens the caecum is long, often more than one-half the
body length, but is relatively shorter as the redia enlarges. The caecum is filled
with partially digested snail tissue, yellow droplets and amorphous, blackish mate-
rial. The young rediae are colorless but as they grow they become more and more
filled with orange-yellow material. The excretory system is identical with that
of the redia of P. alveatum.
Cercaria (Fig. 3)
The cercariae differ from those of P. alveatum principally in size. They emerge
at about the same time of day, have the same swimming movements, accumulate
on the light side of the container, and encyst on the shells of the snails or other
hard surface in the course of one to three or four hours. They emerge from the
rediae while still immature and at this stage have considerable dark pigment around
the ocelli and in the anterior third of the body. On emergence from the snail, the
body is oval to ovate, more pointed anteriorly, convex dorsally and concave ven-
trally; contracted it is circular, 0.14 to 0.16 mm. in diameter; extended it may be
0.30 by 0.10 mm. The tail is simple, slender; it varies from one-half to three times
the length of the body. The locomotor appendage pits at the posterolateral ends
of the body are smaller than those of P. alveatum and diverge at an angle in speci-
mens killed without pressure in whirling, hot, fixing fluids (Fig. 3). The ocelli
are dorsal and anterior to the cephalic ganglia and measure about 0.015 mm. in
diameter; they are provided with lenses. The oral sucker is 0.029 to 0.036 mm.
in diameter ; the esophagus is approximately the same length ; the caeca follow the
140 HORACE W. STUNKARD
lateral contours of the body until they turn laterad and cross the excretory ring
dorsally, near the posterior end of the hody. The body is filled with cystogenous
cells ; the secretion appears as bacilliform rods, 0.002 to 0.003 mm. long and about
one-half as wide. There are 12 to 15 cells between the caeca in a transverse sec-
tion through the middle of the body. The excretory system develops as in all
notocotylid cercariae ; the ring passes posterior to the ganglia and ocelli and is
filled with concretions ; they vary from 0.003 to 0.006 mm. in diameter. On either
side, a recurrent tubule passes posteriad from the anterolateral faces of the ring;
the recurrent tubule bears tufts of long cilia and near the middle of the body
divides into anterior and posterior branches. Each branch bears three clusters of
flame-cells, probably three in each cluster, but not all cells have been observed,
as the cystogenous cells begin to fill with secretions before all the cells and tubules
of the midbody are recognizable.
Metac ere aria
The cercariae encyst promptly if irritated by agitation of the water or the
presence of toxic substances, e.g., solutions of vital dyes; otherwise they may swim
for one to three or four hours. They encyst on the operculum or shell of the snail
or any hard surface, including the wall of the container, always on the side toward
the light. The cysts measure 0.13 to 0.14 mm. outside diameter and 0.11 to 0.12
mm. inside diameter. A specimen was fixed while encysting on a slide ; the secre-
tion had produced a thin, flexible membrane and outside the membrane there was
a sheet of seta-like projections of cystogenous material, 0.030 mm. long and 0.003
mm. in diameter.
DISCUSSION
Rothschild (1941) reported efforts, continued for five years, to solve the life-
histories of the notocotylid cercariae that parasitize Peringia nlvac. Six species
of larvae were isolated and cysts were fed to laboratory-reared ducklings. Three
of the species belonged to the Monostomi group and three to the Yenchingensis
group of cercariae. All attempts to obtain adult worms from the Monostomi
cercariae were negative but she reported (p. 363), "Two species of the Yenchin-
gensis sub-group, however, developed in the intestinal ceca of the ducks, into flukes
of the genus Paranionostoiintin." Rothschild noted with some surprise that one
cercaria of the Yenchingensis group (Szidat and Szidat, 1933) and one of the
Monostomi group (Yamaguti, 1938) had been reported to develop into the same
species of adult, Notocotylus attenuatus. Although Rothschild obtained adult
specimens referred to the genus Paramonostomum, the specific identity of the
specimens was not determined. Kulachkova (1954) did not assign the cercariae
of P. alveatum to one of the larval groups. The recent studies of Odening (1966)
are particularly interesting ; he reported that the cercariae of Notocotylus ephemera
(Nitzsch, 1807); Notocotylus noyeri Joyeux, 1922; Notocotylus pacifcr (Noble,
1933) ; Notocotylus ralli Baylis, 1936; and Notocotylus regis Harwood, 1939, all
of which develop in fresh-water, pulmonate snails, belong to the Monostomi group
of cercariae ; whereas the cercariae of Catatropis vcrrucosa, which also develop in
fresh-water pulmonates, Segmentina nit Ida and Gyraulus albus. belong to the
Imbricata group. From the studies of Stunkard (I960) on the life history of
THE TREMATODE GENUS PARAMONOSTOMUM 141
Notocotylus minutus and (1966b) on Notocotylus atlanticus, it appeared that
cercariae of species of Notocotylus belong to the Yenchingensis group and mature
in the digestive caeca of birds, whereas the cercariae of Paramonostomum belong
to the Monostomi group and develop in the lumen of the intestine. However, the
statements of Rothschild and Odening do not permit such a correlation between
larval type, developmental site, and generic allocation, and the significance of the
larval groups remains obscure and equivocal.
The genus Paramonostomum contains some 20 described species, but the dis-
tinctions between certain of them are very tenuous. Existing descriptions are
based almost entirely on morphology of adult specimens, especially on position of
the genital pore and extent of the vitellaria. Some species are based on the descrip-
tion of a single individual, without adequate consideration for the variation that
always and inevitably occurs. It is admitted that morphological divergence results
from differences in age and degree of maturity, from extension and retraction of
entire specimens or of particular regions of the body, from the accumulation of
reproductive products and from procedures of examination, fixation and preserva-
tion, especially the degree of flattening under pressure of a coverglass. Measure-
ments made on living specimens may differ significantly from those made on the
same individuals after fixation and staining. Moreover, although specificity in
the molluscan host may be relatively restricted, representatives may develop in
final hosts as diverse as birds and mammals, with substantial structural modifica-
tions. Paramonostomum echinatum Harrah, 1922 and Paramonostomum pseudal-
veatum Price, 1931, were described from muskrats, Ondatra zibethica, whereas all
other species are from avian hosts. Swales (1933) reported, without description,
the finding of P. psendalvcatum in Branta canadcnsis taken in Nova Scotia. Lai
(1936) included P. parvum in a new genus, Neoparamonostomum, based on Para-
monostomum ionorne Travassos, 1921 and characterized by the location of the
genital pore and extent of vitellaria. Harwood (1939) discussed the genus Para-
monostomum, suppressed Neoparamonostomum as a synonym of Paramonostomum,
and considered the problems of generic and specific identity. He noted that Para-
monostomum differs from other notocotylid genera only in the absence of ventral
glands, and since these glands are frequently very difficult to locate, it is possible
that some of the species now assigned to Paramonostomum may ultimately be
found to belong elsewhere. The species of Paramonostomum were arranged in
two groups: the Alveatum group, short, oval, with vitellaria extending to the level
of the cirrus sac, to contain the species, alveatum, pseudalveatum, parvum, and
possibly ionorne ; and the Elongatum group, with elongate, spatulate bodies, sug-
gestive of Notocotylus, with a space between the vitellaria and the cirrus sac, to
contain all the other species. Although the observation of Harwood has merit,
Dunagan (1957, p. 581) commented, "Neither of these groups, however, possesses
characters inherent to one but not both. The division is, therefore, of little value
for systematic purposes." Harwood supplemented the description of P. parvum
by the study of specimens from the Helminthological Collection of the U. S. Na-
tional Museum, including No. 39598 from the intestine of a blue goose, Chen caeru-
lescens, collected by A. M. Fallis in Ontario, Canada, and No. 43148 from the
intestine of the American golden-eye, Glaucionetta clangula americana, collected by
D. K. Coburn at the Migratory Bird Refuge, Brigham, Utah. The worms meas-
142 HORACE W. STUNKARD
ured 0.69 to 0.80 mm. in length and 0.46 to 0.50 mm. in width. Harwood stated
(p. 337), "The specimens on which the present redescription is based are, if judg-
ment is based on size alone, more similar to Paramonostomum alveatum than to
P. parvum. They are referred to the latter species, because in the writer's opinion
size is an extremely variable character in trematodes, and because both the distri-
bution of the vitellaria and position of the genital pore are as described and figured
by Stunkard and Dunihue ; they differ in these respects from P. alveatum as figured
by both Liihe (1909) and Kossack (1911). These structural characters are re-
garded as more important than size differences." Harwood assigned the species
reported by Swales (1933) to P. parvum.
The studies of Harwood raise questions of specificity in the genus Paramo-
nostomum. The position of the genital pore relative to the bifurcation of the
digestive tract, and the extent of the vitellaria are variable ; indeed, Harwood ob-
served (p. 336), that the location of the genital pore "may not be wholly reliable
in specimens preserved in a contracted state, especially if the cephalic end is curved
ventrad." Observation of living specimens discloses much variation in location of
the genital pore as the anterior end of the body is extended and retracted. The
description of P. alveatum by Kossack (1911) was based on material from a num-
ber of host species and the specimens varied from 0.78 to 0.90 mm. in length and
0.50 to 0.56 mm. in width. The genital pore was described as ventral to the
intestinal bifurcation and the vitellaria occupied the middle third of the body.
Concerning the vitellaria, Kossack stated (p. 564), "Doch ist ihre Erstreckung
nicht ganz konstant, da sie haufig nach hinten bis zum Vorderrand der Hoden
reichen." In the present study, the cercariae, metacercariae, and adults are refer-
able to two different size groups. The adults of one group are less than 0.50 mm.
in length and are identified as P. parvum; adults of the other group are 0.55 to
0.90 mm. in length and are identified as P. alveatum. If there were specimens of
intermediate size, it would be feasible to include all in a single species, but since
all were from the same intermediate host-species and developed in the same final host-
species, the differences appear to be genetic. The specimens studied by Harwood
agree with the larger of the present species and may belong to P. alveatum.
Examination of published descriptions in the light of the above considerations
raises doubt concerning the validity of certain species. Paramonostomum pseudal-
veatum Price, 1931, from the muskrat, is very similar to P. parvum Stunkard and
Dunihue, 1931, from an unidentified duck, whose life-cycle is reported in the pres-
ent paper. The two species are virtually equal in size and shape ; in both the
cirrus sac is short and wide, with loops of the seminal vesicle and uterus extending
beside the sac ; the metraterm is short, not more than one-half the length of the
cirrus sac; the vitellaria extend from the testes to the level of the cirrus sac, and
the uterus has 8 to 11 transverse loops. Paramonostomum pseudalveatum has a
larger oral sucker, larger gonads, larger cirrus sac and a somewhat more anterior
location of the genital pore. If these features are the result of development in
different hosts, the two species may be identical. Similarly, Paramonostomum
brantae Bullock, 1952, agrees so completely with descriptions of P. alveatum that
the two specific concepts merge and P. brantae falls in synonymy. Bullock (1952)
noted the similarity and distinguished between the two species on the shape of the
ovary in P. alveatum, which he recognized as an unreliable character, and the
THE TREMATODE GENUS PARAMONOSTOMUM 143
larger cirrus sac of P. brantae. The figure of P. brantae shows the cirrus sac
expanded and filled with spermatozoa. Paramonostomum macrostomum Ku, 1938,
was described on a single specimen from Fulica atra taken at Soochow, China. A
somewhat larger single specimen from the same host, F. atra, taken at Lucknow,
India, was described by Baugh (1958) as Paramonostomum fulicai. Paramono-
stomum nettioni Baugh, 1958, from the common teal, Nettion crccca, is similar
morphologically and is intermediate in size between P. macrostomum and P. fulicai,
but information is inadequate to determine the specific status of these species. Two
species, Paramonostomum casarcum from Casarca rutila and Paramonostomum
querquedulum from Querquedula circia, were described by Lai (1936) in India.
Each species was described from a single specimen. The worms are approximately
the same size and morphological agreement is so complete and precise that specific
distinction is highly questionable. An item of reported difference is the location
of the genital pore, which in P. casarcum is at the posterior border of the oral
sucker, whereas in P. querquedulum it is slightly anterior to the intestinal bifurca-
tion. But the location of the pore shifts with extension and retraction of the ante-
rior end of the body and with the orientation of the oral sucker. If the sucker is
turned so that the mouth is subterminal, the esophagus appears short and bent and
the pore apparently is farther forward. Moreover, the two species described by Lai
(1936) are very similar to and may be identical with worms from ducks taken at
Soochow, China, and described by Hsu (1935) as Paramonostomum ovatum. The
description of Paramonostomum microstomum by Moghe (1932) is incomplete,
the uterus is represented in diagrammatic manner and the locations of the ovary
and Mehlis' gland are reversed. The specimens were from Philomachus pugna.r,
taken at Nagpur, India, and are similar to those described by Lai (1936). The
single specimen from Querquedula discors taken in Mexico and described by
Caballero (1942) as Paramonostomum obtortum, closely resembles the worms
described by Lai (1936). A distinct group, characterized by long cirrus sac and
short vitelline zones, includes Paramonostomum actiditis Cable, 1960, from chara-
driiform birds of Puerto Rico, and Paramonostomum histrionici Ching, 1962, from
Histrionicus pacificus taken near Friday Harbor, Washington. Three other species
are characterized by linear, spatulate bodies, very long cirrus sacs that extend to
the middle of the body and short vitelline zones. They are Paramonostomum
elongatum Yamaguti, 1934, from Olor beuncki jankowskii and Olor cygnus taken
in Korea; Paramonostomum bucephalae Yamaguti, 1935, from Buccphala clangula,
Tadorna tadorna, Spatula clypeata and Nyroca marila mariloides, taken in Japan ;
and Paramonostomum malerischi Dunagan, 1957, from the emperor goose, Philacte
canaganica, taken in Alaska. The description of new species from single speci-
mens is not commended and final determination of specific identity in the genus
Paramonostomum may depend on the discovery of life-cycles and the description
of larval stages.
SUMMARY
The account of Mme. Kulachkova (1954) on the life-history of Paramono-
stomum alvealtum is confirmed. The asexual generations and larval stages of both
Paramonostomum alveatum and Paramonostomum parvum occur in the proso-
branchiate snail, Hydrobia salsa, found in brackish-water ponds near Woods Hole,
144 HORACE W. STUNKARD
Massachusetts. Sexually mature worms have been obtained by feeding metacer-
cariae to day-old chicks and laboratory-reared eider and domestic ducklings. Adult
and larval stages of both species are described and figured. Problems of specific
identity in the genus Paramonostomum are discussed.
LITERATURE CITED
BAUGH, S. C, 1958. Contributions to our knowledge of digenetic trcmatodes. III. Proc.
Nat. Acad. Sci. India, Sect. B, 28: 205-226.
BULLOCK, W. L., 1952. Two new species of monostomes from the Canada goose with a review
of Paramonostomum alrcatum (Mehlis in Creplin, 1846). /. Parasitol., 38: 371-378.
CABALLERO Y CABALLERO, E. 1942. Description de un Paramonostomum (Trematoda: Noto-
cotylidae) encontrado en los patos silvestres del lago de Texcoco, V (1). Anal. Inst.
Biol, 13: 91-95.
CABLE, R. M., 1960. Digenetic trematodes of Puerto Rican shore birds. Sci. Sitrz'cy Porto
Rico and Virgin Islands; New York Acad. Sci., 17 (2) : 191-255.
CHING, HILDA L., 1961. Three trematodes from the harlequin duck. Canad. J. Zool., 39:
373-376.
DUNAGAN, T. T., 1957. Paramonostomum malcrischi n. sp. (Trematoda: Digenea : Notocotyli-
dae) from the emperor goose, Philacte canagica L. in Alaska. /. Parasitol., 43: 586-
589.
HARRAH, E. C., 1922. North American monostomes primarily from freshwater hosts. Illinois
Biol.Mongr.. 7: 225-324.
HARWOOD, P. D., 1939. Notes on Tennessee helminths. IV. North American trematodes of
the subfamily Notocotylinae. /. Tenn. Acad. Sci., 14: 332-341, 421-437.
Hsu, Y. C., 1935. Trematodes of fowls in Soochow. Peking Nat. Hist. Bull., 10: 141-150.
KOSSACK, W., 1911. Uber Monostomiden. Zoo!. Jahrb.. Syst., 31: 491-590.
Ku, C. T., 1938. New trematodes from Chinese birds. Peking Nat. Hist. Bull.. 13: 129-136.
KULACHKOVA, V. G., 1954. The life cycle and pathogenic significance of the trematode, Para-
monostomum alrcatum (Mehlis, 1846). Trudi Problemnikh i Trematisheskikh Sovesh-
chani. Akad. Nauk SSSR. No. 4: 118-122. (In Russian)
KULACHKOVA, V. G., 1961a. Annual and seasonal variations in infection of mollusks by larvae
of Paramonostomum alrcatum (Trematoda). Trudi Karelskogo Filiala Acad. Nauk
SSSR. No. 30 : 79-89. ( In Russian )
KULACHKOVA, V. G., 1961b. The biology of the larval stages of Paramonostomum alrcatum
(Trematoda) a parasite of the eider. Trudi Karelskogo Filiala Akad. Nauk SSSR.
No. 30: 90-91. (In Russian)
LAL, M. B., 1936. A review of the genus Paramonostomum Liihe, with descriptions of two
new species and remarks on the genera of the subfamily Notocotylinae. Proc. Ind.
Acad. Sci., 3: 25-34.
LUHE, M., 1909. Parasitische Plattwiirmer. I. Trematoden. In: Siisswasserfauna Deutsch-
lands. Heft 17: 215 pp.
MOGHE, M. A., 1932. Two new species of trematodes from an Indian ruff, Pliiltn/wchus fugnax
Gray. Parasitol., 24: 54-59.
MONTICELLI, F. S., 1892. Studii sui trematodi endoparassiti ; sul genere Notocot\le Diesing.
Boll. Soc. d. Nat. Nafoli. 1 s., 6: 26-46.
ODENING, K., 1966. Physidae und Planorbidae als Wirte in den Lebenszyklen einheimischer
Notocotylidae (Trematoda: Paramphistomidea). Zcitschr. Parasitcnk., 27: 210-239.
PILSBRY, H. A., 1905. A new brackish-water snail from New England. The Nautilus, 19:
90-91.
PRICE, E. W., 1931. Four new species of trematode worms from the muskrat, Ondatra zibethica.
with a key to the trematode parasites of the muskrat. Proc. U. S. Nat. Museum, 79:
1-13.
ROTHSCHILD, MIRIAM, 1935. Note on the excretory system of Ccrcaria ephemera Lebour, 1907
(nee Nitsch). Parasitol., 27: 171-174.
THE TREMATODE GENUS PARAMONOSTOMUM 145
ROTHSCHILP, MIRIAM, 1938. Notes on the classification of ccrcariac of the superfamily Noto-
cotyloidea (Trematoda), with special reference to the excretory system. Novitat.
Zo'oL, 41: 75-83.
ROTHSCHILD, MIRIAM, 1941. Note on the life history of the genus Paramonostomum Liihe,
with special reference to the excretory vesicle. /. ParasitoL, 27: 363-365.
STUNKARD, H. W., 1960. Studies on the morphology and life-history of Notocotylus minutus
n. sp., a digenetic trematode from ducks. /. ParasitoL, 46: 803-809.
STUNKARD, H. W., 1965. Studies on trematodes of the family Notocotylidae. Biol. Bull., 129:
425.
STUNKARD, H. W., 1966a. Further studies on digenetic trematodes of the family Notocotylidae.
Biol. Bull, 131: 409.
STUNKARD, H. W., 1966b. The morphology and life-history of Notocotylus atlanticus n. sp.,
a digenetic trematode of eider ducks, Somateria mollissima, and the designation Noto-
cot\'lns duboisi iioin. nor. for Notocotylus imbricatits (Looss, 1893) Szidat, 1935.
Biol. Bull., 131: 501-515.
STUNKARD, H. W., AND F. W. DUNIHUE, 1931. Notes on trematodes from a Long Island
duck with description of a new species. Biol. Bull., 60: 179-186.
SWALES, W. E., 1933. A review of Canadian helminthology. Canad. J. Res., 8: 468-482.
SZIDAT, L., AND U. SZIDAT, 1933. Beitrage zur Kenntnis der Trematoden der Monostomiden-
gattung Notocotylus Dies. Zcntralbl. Bakt. Parasitenk., Abt. 1, 129: 411^422.
YAMAGUTI, S., 1934. Studies on the helminth fauna of Japan. Part 3. Avian Trematodes II.
Japan J. Zoo/.. 5: 543-583.
YAMAGUTI, S., 1935. Studies on the helminth fauna of Japan. Part 5. Trematodes of birds
III. Japan J. Zool, 6: 159-182.
YAMAGUTI, S., 1938. Zur Entwicklungsgeschichte von Notocotylus attenuatus (Rud., 1809)
und N. magniovatus Yamaguti, 1934. Zeitschr. Parasitenk., 10: 288-292.
Vol. 132, No. 2 April, 1067
THE
BIOLOGICAL BULLETIN
PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY
NORTHERN PACIFIC GIGANTIONE (ISOPODA)
CHARLES G. DANFORTH 1
Biology Department, Glendalc College, Glcndale, California 91208
While carrying out research on the crabs of Eniwetok Atoll, Dr. Jens Knudsen
of Pacific Lutheran University noticed a laterally distorted specimen in the collection
of the Eniwetok Marine Biological Laboratory. He very kindly sent this crab to
the writer. The animal had been identified by Dr. J. Garth of the University of
Southern California, and had been collected by Dr. A. H. Banner of the University
of Hawaii. The host (Fig. 1A) had an isopod located in the left gill area, and this
ectoparasite was found to be a member of the Bopyridae family in the Epicaridea
suborder.
Some years ago the first record of a shore bopyrid for the Hawaiian Islands was
reported by Danforth (1963). At that time an exact identification was not made;
however, the assumption was that the form was in the "lone" or "Cepon" group.
Verification of the former hypothesis may be aided by comparison with the first
bopyrid to be reported from Eniwetok Atoll. This new form is in the genus
Gigantionc, and it and the Hawaiian specimen related to two species of Gigantione
reported by Shiino (1941, 1958). Thus it now appears that there are four species
of the genus in the northern Pacific, two of which have been previously undescribed.
Genus Gigantione Kossmann 1881
Gigantionc pratti n. sp.
Material : one pair.
Host: Phymodius ungulatits (Milne Edwards). Parasitized in the left branchial
region.
Locality: Bruce (Aniyaanii) Island, Eniwetok Atoll. Approximately 162° 28' E.
Long., 11° 28' N. Lat., in dead Acropora coral, at a depth of about 6 feet.
Date : collected on 23 February 1957.
FEMALE
Dimensions : 5.0 mm. greatest length, excluding lamellae ; 4.5 mm. greatest
width, at the third thoracomere.
1 Mailing address : 3612 Angelus Avenue, Glendale, California 91208.
147
Copyright © 1967, by the Marine Biological Laboratory
Library of Congress Card No. A38-518
148
CM. ARI.KS G. DANFORTH
A
C
D
B
All sketches, except that of the crab, are of Gigantionc pratti n. sp. Unless otherwise
indicated, each sketch was drawn by means of a camera lucida. Where a bar (I) is shown, it
represents 1 millimeter.
FIGURE IA. Phymodius uugulatns, drawn from a photograph. The portion of the carapace
which was removed is indicated by a dotted line.
FIGURE IB. Dorsal aspect of the female bopyrid. The extent of the marsupium is indicated
by a dotted line.
FIGURE Ic. First right pleopod of the female, with adjacent epimere. Drawn from a
photomicrograph.
FIGURE ID. First left pleopod of the female, with adjacent epimere.
FIGURE IE. Pleopod #4 of the female.
NORTHERN PACIFIC GIGANTIONE 149
Cephalon. Comprising 1 large lobe, deeply sunken into the thorax, with 2 ear-
like processes. There is a barely discernible anterior border or velum on the head,
having a notch just antero-medial to each of the processes. No eyes. No pigmenta-
tion. The tip of the oral cone can just lie seen from the dorsal aspect of the
parasite.
Thorax. Seven segments, with no pigment, and with only moderate axial
flexion. Most of the distortion to the right is due to unequal growth of the two
sides (Fig. IB). All the pereopods are present. The oostegites do not completely
cover the marsupium, leaving a slight gap at the third and fourth thoracomeres.
There is no hook on the first incubatory lamella (Fig. 2F), although a slight ridge
is present. Obvious, finger-like coxal plates are on the right side, being larger from
segments 1 through 3, and then decreasing through segment 7. The coxal plates
of the left side range posteriad from blade-like to anvil-shaped.
Abdomen. Considerably foreshortened, being about one-fifth of the total body
length, and hidden to a large extent ventrally by the swollen marsupium. Six seg-
ments evident, the sixth having 2 "Y-shaped" uropods (Fig. 2G). There are 5
pairs of pleopods, each pair being biramous, and having the exopods of numbers 2
through 5 relatively smooth and elongate, while the corresponding endopods are
long, thin, and slightly tuberculated (Fig. IE). The first pair of pleopods on either
side (Figs. 1C, ID) are much larger than the others, and tend to cover them. The
exopodite is smooth and blade-like, whereas the endopodites are tuberculated or
somewhat pinnately divided at the border. Each biramous pleopod is adjacent to a
smooth-edged, elongated epimeral plate.
MALE
Dimensions: Length 2.5 mm. Width 1.0 mm. at the fourth thoracomere.
Location. On the smaller side of the ventral abdomen of the female, with the
male's head in the same direction as that of its mate. Although the male was not
within the marsupium (Fig. IB), it was covered by the bulbous oostegites.
Cephalon. Blunt, with neither a border nor any processes. Eyes distinct, with
a reddish cast. First antenna of 2 articles, second antenna of 4 articles, tipped with
bristles. No pigment.
Thorax. Typical fusiform shape, widest at the fourth segment. Lateral plates
not unusual in appearance. Seven distinct segments and 7 pairs of pereopods. No
pigment.
Abdomen. Five separate, tapering segments, plus the telson. No pigmentation.
The lateral plates gently rounded, except the fifth pair, which is hooked posteriad.
Five pairs of pleopods, each with a short and a longer ramus (Fig. 21). The rami
are tube- or rod-like, and lie almost transversely to the abdominal axis. The short
ramus of each is lateral, with the longer ramus nearly meeting its counterpart at
the abdominal midline. The uropods are 2 in number, heart-shaped, with the
pointed end anteriorly. There is a posterior indentation on each which is quite
evident (Fig. 2J), rather than being merely a slight notching. The uropods are
plainly visible from the dorsal aspect of the male (Fig. 2H), and are slightly rough-
ened, but not hirsute.
1 50
CHARLES G. DANFORTH
-m
F
A
G
H J
FIGURE 2p. Internal face of the first oostegite of the female.
FIGURE 2c. Right uropod of the female.
FIGURE 2n. Dorsal aspect of the male bopyrid. Drawn from a photomicrograph.
FIGURE 2i. Second right pleopod of the male.
FIGURE 2j. Ventral aspect of the abdomen of the male. Drawn from a photomicrograph.
a = oral cone, b = notch in marginal velum, c = antenna, d = cephalic process, e = pereopod,
f = coxal plate, g = abdominal lamella, h = pleopod, i = uropod, j = position of male, k =
exopodite, 1 = endopodite, m = medial ramus.
REMARKS
As can be seen from the accompanying tables, G. pratti differs from all males in
the genus by the form of the uropoda. The male pleopods are similar to those of
sagamiensis, but there is pigmentation and a pair of uniramous, foliaceous uropods
NORTHERN PACIFIC GIGANTIONE
151
TABLE I
Major characteristics of Gigantione species
Name, size,
locale, host
Head
Thorax and
appendages
Abdomen and
appendages
Uropoda
bouvieri
"Ear-like"
Coxal plates nar-
1st 2 pleomeres fused
Shaped like a
9 = 3-4 mm.
processes.
row and folded.
in midline. Pleo-
two- fingered
c? = 0.5? mm.
Plates on all 7
meres 1-5 with
glove.
Azores, in
c?
segments
straight prolonga-
Pilumnus
No eyes.
tions. Pleopods al-
d1
W. Indies, in
most triramous, 2-5
Pea-shaped.
Hypoconcha
tubercnlated. 1st
pair largest.
d1
Dleopods 1-5 rod-like.
giardi
Processes.
2 "saillies" on an-
Plates 1-5 like those
Fleshy base with
9 = 20 mm.
Large, with
terior portions
of the pereon. #6
2 small cylin-
cf = 7 mm.
fleshy bor-
of each seg-
with 2 lamellae plus
drical branches.
Tuamoto, in
der.
ment. Lateral
nropods. Pleopods
d1
Xantho
d1
plates on 1-7.
like those of moebii.
2 leaf-like plates.
Eyes.
5 pairs large
d1
oostegites, # 1
Pleopods = "lami-
with internal
naires."
ridge.
hawaiiensis
Bilobed at
Long, thick lamel-
Long, thin lamellae on
Narrow base with
9 = 9.7 mm.
posterior.
lae or coxal
1-5. Pleopods sub-
2 elongate,
cf = 3.4 mm.
Processes.
plates on one
triramous ; more pin-
finger-like
Hawaii, in
c?
side, triangular
nate than tubercu-
branches.
Xantho
Eyes. Slight
on other side.
lated.
c?
pigmenta-
d]
cT
Kidney-shaped
tion.
Lateral pigmenta-
Pleopods unirainous
and pubescent.
tion.
rods. Slight pigmen-
tation.
ishigakiensis
Processes.
Elongate, taper-
Lamellae on 1-5 like
Shaped like a 2-
9 = 13mm.
Frontal
ing coxal
coxal plates. Pleopod
fingered glove.
C? = 3.2 mm.
lamina.
plates. Closed
1 largest, lamellar;
d1
Japan, in
d1
marsupium.
digitiform processes
Uniramous, folia-
Carpilius
Eyes. No
d1
on both rami. 2-5
ceous.
pigment.
Widest at seg-
are heavily tuber-
ment 6.
culated.
d1
5 pairs of rod-shaped,
uniramous pleopods.
moebii
No margin.
Marsupium
Small pleon plates.
Swollen base
9 = 15 mm.
Processes.
covered. Coxal
Pleopod 1 triram-
with 2 thin
c? = 3? mm.
plates heavy,
ous, relatively
branches.
Isle Maurice, in
d1
but not long.
smooth. Pleopods 2-
Ruppelia
Eyes. No
5 triramous, with
d1
pigment.
heavy tubercula-
Uniramous and
tion.
leaf-like.
d1
Pleopods 1-5 egg-
shaped.
CHARLES G. DANFORTH
TABLE I (continued)
Name, size,
locale, host
Head
Thorax and
appendages
Abdomen and
appendages
Uropoda
pratti
Processes.
Marsupium
Pleomeres 1-5 with
"Y-shaped,"
9 = 5.0 mm.
Notched
slightly open.
blade-like plates.
with a narr-
c? = 2.5 mm.
anterior
Coxal plates
Pleopod 1 largest,
row base.
Eniwetok, in
margin.
thin, long on
biramous. Endo-
Phymodius
larger side ; tri-
podites of other
cf
d1
angular on
pleopods slightly
Eyes. No
smaller side.
tuberculated.
Cordate, but still
pigment.
c?1
uniramous.
5 pairs of biramous,
rod-like pleopods.
rathbunae
No margin.
Coxal plates thin
Plates large and
Swollen base
9=4 mm.
No proc-
and finger-like.
finger-like on large
with 2 tapered
cf = 1 mm.
esses. Eyes.
Segments 1-4
side; triangular on
rami.
"Salomon Is-
d1
double on the
small side. Pleopod
lands," in
Eyes. No
left.
1 largest, with mar-
d1
Actaea
pigment.
ginal serrations.
Uniramous,
d1
foliaceous and
5 pairs of bulbous
hirsute.
pleopods.
sagamiensis
No margin.
Coxal plates from
Plates 1-5 tubercu-
Swollen base
9 = 3.7 mm.
No proc-
slightly pointed
lated and folded
with 2 short,
o71 = 1.4 mm.
esses.
to crescentic
back. 5 pairs of bi-
blunt rami.
Japan, in
and blunt. Mar-
ramous pleopods:
d1
Carpiliodes
d1
supium almost
endopod is filiform,
Uniramous,
Eyes.
closed.
tuberculated ; exo-
leaf-like.
pod more blunt.
Pleopod 1 is largest,
nearly triramous.
Pigmented. 5 pairs of
rod-shaped, biram-
ous pleopods.
on the latter. The G. pratti female has ear-like cephalic processes, as do all other
species except rathbunae and sagamiensis ; however in these, neither has a cephalic
margin, and in the case of rathbunae, eyes are present. The female bouvieri, as
illustrated by Nierst;rasz and Brender a Brandis (1931), has structurally different
coxal plates, and a partial fusion of pleomeres 1 and 2. The female giardi has a
fleshy cephalic border, and lamellae on pleomere 6. The female ishigakiensis has a
different frontal lamella, a closed marsupium, and digitiform processes on the exopo-
dite of the first pleopod. The female moebii lacks a cephalic margin, has differently
shaped coxal plates, shorter abdominal lamellae, and a triramous condition for
pleopod 1. Aside from the foregoing differences, the female pratti seems to be
unique in the possession of "Y-shaped" uropods. The uropoda of other females of
the genus are biramous, but the branches range from short and separated to fairly
long and distinct ; in none do they diverge abruptly from a narrow base.
NORTHERN PACIFIC GIGANTIONE 153
The male allotype, and female holotype have been deposited in the United States
National Museum, catalog number 113940. The host crab is catalog number 113939.
Named for Dr. Ivan Pratt, parasitologist at Oregon State University, who first
suggested to the writer that the epicarid isopods might be an interesting field of
study.
Gigantione hawanensis n. sp.
For added data and sketches, refer to Danforth (1963).
Material : one pair.
Host : Xantlw crassiiuonits Milne Edwards. Parasitized in the left branchial region.
Locality : tide pool at Diamond Head, Oahu, Hawaii.
Date: collected on 13 January 1962.
FEMALE
As described. The female is stated to have 6 pairs of abdominal lamellae, while
the drawing shows but 5 pairs.
MALE
As described. The pleopods are referred to as tubercles.
REMARKS
A reconsideration of this previously described Hawaiian form seems to indicate
that the genus is correctly Gigantione. The structure of the male, and some of the
features of the female, make it obvious that it does not belong in one of the existing
species. The "claw-shaped" appendages which were found free in the preservative
undoubtedly are the uropoda of the female.
The accompanying tabulation indicates the differences between G. hawaiiensis
and other species of the genus. The partially bilobed head of the female is unique,
and the pinnate structure of the pleopod rami is in contrast to the more commonly
found tuberculations of other forms. The male has pigmentation, as opposed to
others except sagamiensis, and the disc-like uropoda are distinctly at variance with
the foliaceous, pea-shaped, or cordate shapes illustrated by other males.
The specimens have been deposited in the United States National Museum,
catalog numbers: 110192 (larvae), 110191 (female holotype), and 110190 (male
allotype).
DISCUSSION
As stated by Bonnier (1900, p. 276) for Gigantione: "Deux caracteres suffisent
a caracteriser ce genre : la femelle adulte possede des lames pleurales sur tons les
somites, tant ceux du thorax que ceux de I'abdomen, et ses uropodes sont birames."
The species described so far are :
G. boitvieri Bonnier G. nwebii Kossmann
G. giardi Nobili G. pratti n. sp.
G. hawaiiensis n. sp G. rathbunae Stebbing
G. ishigakiensis Shiino (1941) G. sagamiensis Shiino (1958)
134 CHARLES G. DANFORTH
The major characteristics of these species are shown on the accompanying
tabulation (all female forms have 7 thoracomeres, 6 pleomeres, a single-lobed head,
and biramous uropoda). Reference was made to Paragigantione paplllosa Barnard,
since it is only a matter of degree between : "Eine Anzahl Coxalplatten am Pereion
sehr entwickelt (odor sehr abweichend gestaltet)," and, "Alle Coxalplatten am
Pereion nur massig entwickelt bis fehlend," as used in the key by Nierstrasz and
Brender a Brandis (1932, pp. 90, 91). However, the two genera are quite dis-
similar in many respects, so Paragigantione is not included in the table.
In reviewing almost any of the information on epicarids, one finds many in-
stances of contradiction or confusion. Some points are purely typographical errors,
others misinterpretation, etc. A disconcerting item for G. giardi is Nobili's (1906,
p. 270) statement: "Lames pleurales des segments de 1'abdomen conformees comme
celles du thorax ; sixieme segment pourvu aussi de deux lamelles, et de petits
uropodes charnus." If correct, this is entirely different from all other species in the
genus, and might lead one to place the form into the genus Orbionc. Unfortunately,
there is no illustration against which the description could be checked (such as is
the case in Bonnier's description of G. inoebii in which he mentions the presence of
6 pairs of pleopods, while his drawing shows 5 pairs). Further ambiguity is en-
countered in a key by Dakin ( 1931 ) , wherein Crassione is separated from Gigantione
on the basis of the uropods of the former being biramous, whereas those of the latter
are given as uniramous! Since Gigantione has biramous uropoda, it is indeed
fortunate that sketches of Crassione indicate that the specimen is in fact not
Gigantione. It is points such as these, coupled with accidental mislabeling or
identification, that indicate the great need for an evaluation of the available literature
on epicarids.
SUMMARY
1. Eight species of Gigantione have now been described. One was from the
north Atlantic, one from the Indian Ocean, two from the south Pacific, and four
from the north Pacific. Of these last, G. pratti and G". hati'aiiensis are new species.
The hosts of the different species have all been in separate genera, with the exception
of those for G. giardi and G. haivaiiensis, both of which were in Xantho.
2. Dr. Shiino is carrying out an intensive investigation of epicarids in the
Japanese archipelago, and the writer is in the process of preparing a monograph
covering the Epicaridea of the northern Pacific (except for those areas and forms
near Japan). Therefore, it should be expected that many new species and possibly
genera will be found in the Pacific as collecting continues.
LITERATURE CITED
BARNARD, K. H., 1920. Contributions to the crustacean fauna of South Africa. Number 6.
Further additions to the list of marine Isopoda. Ann. So. African Mus., 17: 319-438.
BONNIER, J., 1900. Contribution a 1'fitude des fipicarides : Les Bopyridae. Trav. Stat. Zool.
de Wimereux, 8 : 1-475.
DAKIN, W. J., 1931. On a new bopyrid parasite from the coast of New South Wales. Proc.
Linn. Soc. New South Wales, 56: 267-272.
DANFORTH, C. G., 1963. First record of a Hawaiian shore bopyrid (Isopoda: Bopyridae). J.
Parasit., 49 : 847-850.
NORTHERN PACIFIC GIGANTIONE 155
KOSSMANN, R., 1881. Studien iiber Bopyriden. I. Gigantionc inochii und Allegemeincs iihcr
die Mundwerkzeuge der Bopyriden. Zcitschr. f. wiss. Zool., 35 : 652-665.
NIERSTRASZ, H. F., AND G. A. BRENDER A BRANDTS, 1931. Papers from Dr. Th. Mortensen's
Pacific Expedition 1914-16. Epicaridea II. Vidcn. Mcdd. fra Dansk naturhist. Form.
K0bcnhavn, 91 : 147-226.
NIERSTRASZ, H. F., AND G. A. BRENDER A BRANDIS, 1932. Alte und neue Epicaridea. Zool.
Ans., 101 : 90-100.
NOBILI, G., 1906. Diagnoses preliminaires de Crustaces, Decapodes et Isopodes nouveaux
recueilles par M. le Dr. G. Suerate aux lies Touamotou. Bull. Mus. Hist. Nat., Paris,
12: 256-270.
SHIINO, S. M., 1941. Further notes on bopyrids from Kyusyu and Ryukyu. Annot. Zool.
Jap., 20: 154-158.
SHIINO, S. M., 1958. Note on the bopyrid fauna of Japan. Kept. Fac. Fish. Pref. Univ. Mie,
3: 29-73.
STEBBING, T. R. R., 1910. Isopoda from the Indian Ocean and British East Africa. In: Re-
ports of the Percy Sladen Trust Expedition to the Indian Ocean in 1905. Trans. Linn.
Soc. London, scr. 2, Zool., 14: 83-118.
SURFACE AREA RESPIRATION DURING THE HATCHING OF
ENCYSTED EMBRYOS OF THE BRINE SHRIMP,
ARTEMIA SALINA
DAVID N. EMERSON *
Department of Biological Sciences, University of Alaska, College, Alaska 99735
When the encysted embryos of Artemia salina are placed in water (hydration)
the embryos resume development. After an interval of time depending npon condi-
tions of incubation, excystment takes place in two stages. The first stage (emer-
gence) occurs when the hard outer cyst wall splits, and the embryo emerges head-
first within a hatching membrane. The second stage (hatching) occurs a few hours
later when a nauplius larva swims from the membrane and shell. The transition
from the encysted stage to the emerged stage depends upon an uptake of water,
mainly due to increased internal concentration of glycerol (Clegg, 1964) and pos-
sibly to an increase of free ammo acids at the same time (Emerson, 1967). The
uptake of water increases the volume of the developing embryo to cause the cyst
shell to split. There is consequently an increase of the surface area of the embryo
which is shown by scaled micro-photographs of Nakanishi ct al. (1962).
There are several studies of respiration of Artciiiia during development (Urbani,
1946; Dutrieu, 1960; Muramatsu, 1960; Emerson, 1963; Clegg, 1964). These
studies are difficult to compare because of different sources of cysts, possible differ-
ences in percentage of viable cysts, and experimental differences in the salinity and
temperature of the hatching solution. In spite of differences reported for the rate
of oxygen consumption of the embryos, some of these studies reveal a similar pat-
tern. The oxygen consumption rate increases rapidly within the first few hours
after hydration, and then remains constant for a time. A second increase occurs at
about the time of emergence. Von Bertalanffy and Krywienczyk (1953) have
shown that oxygen consumption of the nauplius and later stages of Artemia is pro-
portional to surface area. An increase of surface area during emergence could
account for the increase of respiration which occurs at the same time. The present
study demonstrates that oxygen consumption patterns of developing Artemia em-
bryos can be interpreted on the basis of surface rule respiration.
MATERIALS AND METHODS ,
1. Source of encysted Artemia embryos
The encysted embryos used in this study were obtained in 1965 from Ward's of
California (Monterey). The cysts were from Great Salt Lake, Utah.
1 Supported in part by NIH predoctoral fellowship (1-F1; Gm-21.084) while the author
was at the University of South Dakota, Vermillion.
156
HATCHING ARTEMIA SURFACE RESPIRATION 157
2. Respiration measurements
Oxygen consumption of the Artcmia embryos was measured with a Warburg
constant volume respirometer (Umbreit ct ol., 1959). Dry cyst samples weighing
10.0 mg. were placed in flasks ( IS ml. volume) with center well and sideann. The
flasks contained 2.5 nil. 0.5 M NaCl, and 0.2 ml. 20% KOH in the center well.
Readings were made at 1-2-hour intervals at 25° C. Calculations are expressed as
,wl. O2/hr./mg. dry cyst weight. Most of the experiments were carried out with no
agitation of the flasks and no antibiotics in the water. Since somewhat different
readings were obtained than in a similar set of experiments (Emerson, 1963), other
series of measurements were made with agitation at a rate of 60 complete oscilla-
tions/min.. and with antibiotics in the water (penicillin, 1000 units/ml, and strepto-
mycin, 100 /Ag./ml., Clegg, 1964). Microbial activity was evaluated at the end of
runs in which there were no antibiotics in the water by filtering off the brine shrimp
(Whatman #1 paper) and measuring oxygen consumption of the water over a
period of several hours.
3. Measurement of surface area of cysts and of emcryed embryos
All measurements were made with a dissecting microscope fitted with a cali-
brated eyepiece micrometer. The encysted embryos are spherical in shape. Surface
area was calculated directly from measurements of diameter, using the formula for
the surface area of a sphere (area -- 12.57 r2, where r -- radius). The emerged
embroys have a symmetrical shape resembling a pear. Measurements of emerged
embryos were drawrn on graph paper. . Each drawing was divided into sections.
Surface areas of the middle sections were calculated using the formula for the
curved surface of a right cylinder (area == 27i-rh, where h == altitude). The surface
areas of the two end sections were calculated as the curved surface of a right cone
(area = 7rr\/r2 + h2 ). Areas of individual sections were totaled to give the sur-
face area of the emerged embryo.
4. Percentage of emergence
The period of time when 50% of the embryos were fully emerged (Tso%E) was
estimated by periodic counts of the percentage of emerged embryos during develop-
ment.
RESULTS
Oxygen consumption rates during development are summarized in Table I.
The presence of antibiotics in the incubation media, or agitation of the Warburg
vessels does not significantly affect oxygen measurements (Table II). The surface
areas of encysted and emerged embryos are compared in Table III.
DISCUSSION
The following terms are used to describe the oxygen uptake pattern of Artemia
embryos during development (Table I). The hydration period is the first rapid
uptake of oxygen ; the differentiation period is a plateau during which the rate of
oxygen consumption remains about the same ; the emergence period occurs during
the second rise of oxygen consumption rate when most of the embryos are emerging ;
158
DAVID N. EMERSON
TAHLK I
Oxygen consumption of developing Artemia embryos in 0.5 ]\f NuCl at 25° C. The values
given are pi. O^/hr/.mg. dry cyst weight. The numbers preceded by ± signs
give confidence limits at the P5% level
Hours of development
Oxygen consumption
(15 determinations)
* Period of development
0
1-2
0.46 ± 0.08
Hydration
2-4
0.93 ± 0.02
4-6
1.06 ± 0.05
6-8
1.05 ± 0.11
8-10
1.17 ± 0.17
10-12
1.10 ± 0.17
Differentiation
12-14
1.20 ± 0.03
14-16
1.01 ± 0.15
16-18
1.21 ± 0.23
18-20
1.44 ± 0.23
20-22
1.66 ± 0.36
** Emergence
22-24
1.73 ± 0.36
28-30
1.94 ± 0.42
30-32
2.00 ± 0.49
Hatching
36-38
2.00 ± 0.26
* See text for explanation.
** The first emerged embryos were seen at 16 hours; T5o<r0E was at 24 hours.
and the hatching period is when oxygen consumption levels off again after Tso%E
Average rates of oxygen consumption of the differentiation and of the hatching
period of this study are compared with other studies (Table IV).
Muramatsu's measurements went only to 12 hours of development so that the
hatching period probably was not reached. Urbani's measurements probably repre-
sent oxygen consumption well past TSO%E, since the figure listed under hatching
period (Table IV) is a value for 50 hours of developing. These two studies will
not be considered in the following discussion.
TABLE II
Comparison of conditions for oxygen consumption measurements of developing Artemia em-
bryos through 20 hr. development in 0.5 M NaCl at 25° C. The numbers in
parentheses indicate the number of determinations. The numbers
preceded by ± signs give confidence limits at the 95% level
Condition
Total n\. O2/mg. dry
cyst (20 hr.)
*No antibiotics; not agitated (15)
Penicillin and streptomycin; not agitated (6)
Penicillin and streptomycin; agitated at 60 complete oscillations
per minute (6)
20.8 ± 1.3
19.9 ± 3.0
19.7 ± 2.2
* Measurement of filtered water at the end of the runs showed very little oxygen consumption
due to microbial activity.
HATCHING ARTEMIA SURFACE RESPIRATION
159
TABLE III
Surface area of Artemia embryos during development
Stage
* Surface area
Encysted embryo (differentiation period)
Fully emerged embryo (hatching period)
% increase of surface area
119,415
205,178
172%
* Averages of 10 measurements. Statistical variation is not shown because individual measure-
ments were almost identical.
The increase in oxygen consumption (Table IV) is very similar to the increase
of surface area (Table III) during emergence. This observation suggests that the
increase of oxygen consumption rate is proportional to an increase of surface area.
The pattern of oxygen consumption (Table I) can be interpreted as follows:
Oxygen consumption rises during the hydration period ( 1-3 hours in duration ;
Iwasaki, 1964) due to reactivation of metabolism of the dormant embryo. The
initial rate rises to a constant value which is limited by the surface area of the cyst
throughout the differentiation period. During this period, there is no cell division
(Nakanishi et a/., 1962; Emerson, 1963), no increase of DNA (Bellini, 1960;
Emerson, 1963) ; and no incorporation of tritiated thymidine (Emerson, 1963).
Tritiated thymidine is incorporated only after hatching (Emerson, 1964) as cells
start to divide (Nakanishi et al., 1962, 1963). The respiratory quotient remains
close to 1 during this period, indicating metabolism of carbohydrate (Dutrieu, 1960;
Muramatsu, 1960 ; Emerson, 1963 ; Clegg, 1964) which is probably trehalose
(Dutrieu, 1960). The rate of oxygen consumption increases during emergence,
and rises rapidly to a new peak limited by the surface area of the emerged embryo
and early nauplius. The respiratory substrate during and after emergence is prob-
ably lipid as indicated by lowered respiratory quotients ( Dutrieu, 1960 ; Emerson,
1963), and increase in lipase activity (Bellini and Lavizzari, 1958) and a decrease
in total lipids (Dutrieu, 1960; Urbani, 1959).
The present study shows that surface rule respiration can explain the pattern
of oxygen consumption during development of encysted Art curia embryos. Similar
patterns of respiration exist during the embryonic development of a variety of
animals (Boell, 1955). It would be interesting to see if surface rule respiration
TABLE IV
Average rates of oxygen consumption of the differentiation and of the
hatching period of Artemia
/il. Os consumed per hour
T? f
Diff. period
Hatch, period
Table I
1.11 fj,\./mg.
1.98 /ul./mg.
178
Emerson, 1963
1.66 Ail. /nig.
2.95 M./mg.
178
Dutrieu, 1960
1.30 Ail./mg.
2.23 Atl./mg.
172
Muramatsu, 1960
1.03 A»l-/mg.
1.45 Atl./mg.
141
Urbani, 1946
0.00009 Ail. /cyst
0.00019 Avl./cyst
211
160 DAVID N. EMERSON
applies for these animals, especially for sea urchins which have strikingly similar
patterns (Lindahl, 1939; Wright, 1963).
SUMMARY
Oxygen consumption of Art curia salina was measured during development in
0.5 M NaCl at 25° C. A pattern is seen in which the rate of oxygen consumption
increases rapidly within the first few hours after hydration, remains constant for a
time, and then increases rapidly again while most of the embryos are emerging.
This pattern is dependent upon surface area of the developing embryo. During
emergence, the surface area of the embryo increases 172% over the surface area
of the encysted embryo. During the same development period, oxygen uptake in-
creases by almost the same factor.
LITERATURE CITED
BELLINI, L., 1960. Osservazioni sugli acidi nucleici nello sviluppo di Artcmia salina Leach.
Riccrca Sci.. 30 : 816-833.
BELLINI, L., AND G. S. LAVIZZARI, 1958. Studio delle lipase nello sviluppo di Artcmia salina
Leach. R. C. Accad. Lined, 24: 92-95.
BERTALANFFY, L. VON, AND J. KRYWIENCZYK, 1953. Surface rule in crustaceans. Amer. Nat.,
87: 107-110.
BOELL, E. J., 1955. Energy exchange and enzyme development during embryogenesis. In :
Analysis of Development. Edited by Willier, Weiss, and Hamburger. W. B.
Saunders, Philadelphia.
CLEGG, J. S., 1964. The control of emergence and metabolism by external osmotic pressure and
the role of free glycerol in developing cysts of Artcmia salina. J. E.vp. Biol., 41 :
879-892.
DUTRIEU, J., 1960. Observations biochimiques et physiologiques sur le developpement d'Artemia
salina Leach. Arch. Zool. E.vp. Gen., 99: 1-133.
EMERSON, D. N., 1963. The metabolism of hatching embryos of the brine shrimp, Artcmia
salina. Proc. S. D. Acad. Sci., 42: 131-135.
EMERSON, D. N., 1964. Incorporation of tritiated thymidine by Artcmia salina. Proc. S. D.
Sci., 43: 90-95.
EMERSON, D. N., 1967. Some aspects of free amino acid metabolism in developing encysted
embryos of Artcmia salina, the brine shrimp. Comp. Biochcm. Physiol., 20: 245-261.
IWASAKI, T., 1964. Sensitivity of Artcmia eggs to the r-irradiation III. The sensitivity and
duration of hydration. /. Rad. Res., 5 : 91-96.
LINDAHL, P. E., 1939. Zur Kenntnis der Entwicklungsphysiologie des Seeigeleies. Zcitschr.
vergl. Physiol., 27: 233-250.
MURAMATSU, S., 1960. Studies on the physiology of Artemia embryos. I. Respiration and
its main substrate during the early development of the encysted embryo. Embrvohgia,
5: 95-106.
NAKANISHI, Y. Y., T. IWASAKI, T. OKIGAKI AND H. KATO, 1962. Cytological studies of
Artcmia salina. I. Embryonic development without cell multiplication after the
blastula stage in encysted dry eggs. Annot. Zool. Japan., 35 : 223-228.
NAKANISHI, Y. Y., T. OKIGAKI, H. KATO AND T. IWASAKI, 1963. Cytological studies of
Artcmia salina. II. Deoxyribonucleic acid (DNA) content and the chromosomes in
encysted dry eggs and nauplii. Proc. Japan Acad., 39: 306-309.
UMBREIT, W. W., R. W. BURRIS AND J. F. STAUFFER, 1959. Manometric Techniques, 3rd
edition. Burgess Publishing Co., Minneapolis, Minn.
URBANI, E., 1946. L'assuzione di 02 durante la vita latente ed il passaggio alia vita attiva.
Boll. Soc. Ital. Biol. Spcr., 22: 453-456.
URBANI, E., 1959. Protidi glucidi e lipidi nello svilluppo di Artcmia salina Leach. Ada
Embryol. Morph. E.vp., 2: 171-194.
WRIGHT, B. E., 1963. The biochemistry of morphogenesis. In : Comparative Biochemistry,
volume VI. Edited by Florkin, M. and H. S. Mason. Academic Press, New York.
UTILIZATION OF DISSOLVED EXOGENOUS NUTRIENTS BY THE
STARFISHES, ASTERIAS FORBESI AND
HENRICIA SANGUINOLENTA
JOHN CARRUTHERS FERGUSON
Department of Biology, Florida Presbyterian College, St. Petersburg, Florida 33733
In the past few years it has become increasingly evident that nutrition in many
types of invertebrate animals involves not only the ingestion of solid foods or
particulate matter, but also the utilization of dissolved organic materials commonly
found in the environment. While speculation on the significance of this latter source
of nutrients dates back at least to the work of Putter (1909), it remained for
Stephens and Schinske (1961) to provide the first clear evidence that dissolved
materials can be taken up by a wide variety of invertebrates. These workers demon-
strated that representatives of 10 phyla (including Echinodermata) could remove
glycine from dilute solutions in sea water. Stephens has continued his investiga-
tions and further described the uptake of dissolved amino acids and sugars by
several forms, notably the coral Fungia (Stephens, 1962), various annelids
(Stephens, 1963, 1964), and brittle stars (Stephens and Virkar, 1965, 1966).
In the course of my own studies (Ferguson, 1963a, 1963b), I have observed,
by the use of autoradiographic methods, that dissolved C14-labeled nutrients (glucose
and amino acids) appear to be readily taken up into at least the epidermal tissues of
Aster ias forbesi. I have suggested that this may represent the most important
source of nutrients to some of the more isolated superficial tissues of starfishes, and
that in species such as A. forbesi the epidermal absorptive process may be facilitated
"by enrichment of the medium with stray products released from the externally
digested food and by scavenging activities of pedicellariae" (Ferguson, 1963a, p.
79).
Most recently, Pequignat (1966) has reported detailed investigations on a num-
ber of echinoderms, including Asterias rubens, demonstrating digestion of various
types of nutritional products on the skin by glandular secretions and migrating
coelomocytes. While his observations are basically subjective in nature, he con-
cludes that at least some of the materials which are digested externally are absorbed
directly into the epidermis.
At this time, then, it appears that dissolved organic materials are utilized by
starfish (and many other invertebrates), and that at least some of the nutrients are
taken up directly by the body surface, thus by-passing the digestive tract. Further-
more, it is probable that in various species of echinoderms mechanisms, such as the
pedicellariae, have evolved which serve to enhance the availability of dissolved
nutrients to the integuments. There are, however, at least several important ques-
tions which are as yet unanswered. First, are dissolved nutrients taken up by the
digestive tract as well as the epidermis ? Second, do epidermally absorbed nutrients
1 Supported by NSF grants GB-2209 and GB-4994.
161
162 JOHN CARRUTHERS FERGUSON
become distributed throughout the body, or can they benefit only the superficial
tissues into which they are initially taken up ? And third, are there marked differ-
ences in the handling of exogenous nutrients by various species of starfishes ? The
present investigation has been directed toward these three points.
MATERIALS AND METHODS
The starfish used in these experiments were freshly collected specimens of
Asterias forbesi and Hcnricia sangninolenta obtained from the Supply Department
of the Marine Biological Laboratory. A few specimens of Asterias vulgaris were
also studied, but as these did not appear to react differently from A. forbesi further
work on this species was not continued. All the animals used were about 2 inches
in diameter. They were placed individually in beakers containing a medium con-
sisting of 50 ml. of filtered sea water and dissolved, C14-labeled nutrients. The
specimens were left in this medium for a period of 8 hours (except those sacrificed
at 1 hour), and then rinsed twice and placed in a holding tank of running sea water.
While retained in the holding tank they were provided with a number of small
clams to serve as food. The distribution of radioactivity in the tissues of groups of
animals was analyzed following periods of 1,8, and 72 hours, and 20 days, meas-
ured from the time the animals were first placed in the medium.
Two types of medium were used. One consisted of 0.5 microcurie (0.0033 mg.)
of a mixture of 15 uniformly C14-labeled, purified amino acids per 50 ml. of filtered
sea water. The manufacturer of the amino acid mixture (New England Nuclear
Corp. of Boston, Mass.) claims that it contains the "same relative proportions as
found in a typical algal protein hydrolysate." The other medium consisted of 0.5
microcurie (1.85 mg.) of uniformly labeled C14-glucose in each 50-ml. portion.
In order to measure the distribution of the labeled nutrients in the animals, each
specimen was dissected as follows : the rays were cut off as near to the disk as
possible. Incisions were then made up the lateral edges of each ray so that the oral
and aboral portions could be separated. Next, the digestive glands were pulled free
from the aboral portion. The disk was then picked up and each of its supporting
columns severed so that it could be opened and the stomach (both cardiac and
pyloric divisions) cut free.
As a result of this procedure five groups of tissue were obtained. These will
be referred to as the "disk," "oral body wall," "aboral body wall," "stomach," and
"digestive glands." The gonads were always included with the disk group, as their
state of development was not consistent enough to warrant a separate set of analyses.
Furthermore, preliminary studies had demonstrated negligible uptake of the nutri-
ents by these structures.
The groups of tissues were then processed in two different ways for analysis of
their radioactivity. The first method was designed to measure the total amount of
material actually absorbed and retained. In it, the tissues were digested at 100° C.
in test tubes with 1 ml. of 1 M NaOH in sea water. Digestion was enhanced by
adding 1 or 2 drops of 30% H2O2. When all of the soft tissues were uniformly dis-
persed, the contents of the tubes were decanted into tared, 1-inch, stainless steel
sample pans and dried in an oven. The radioactivity of each sample was measured
in a Nuclear-Chicago, low-background, G-M counter fitted with a "Micromil"
window. Corrections were made on the basis of infinite thickness and the counts
NUTRIENT UTILIZATION BY STARFISHES
163
compared to those of similarly prepared tissues to which known quantities of laheled
nutrients had been added. The corrected measurements of 27 such standard samples
had a mean deviation of 11.6%. The alkaline digestion was used in preference to
solubilizing in acid as it prevented the loss of carbonaceous endoskeletal material.
The second method was intended to determine the amounts of absorbed nutrient
materials which were retained by the tissues in a relatively unbound state. In this
procedure, each group of tissue was extracted 48 hours in 10 ml. of ethanol solution.
Based on the results of test runs, an 80% concentration of alcohol was found most
satisfactory for the amino acid samples and a 40% solution best for the samples
containing glucose. In both cases, duplicate 0.25-ml. aliquots of the extracts were
plated onto 1^-inch stainless steel sample pans, dried, and counted. Again, the
counts were compared to those of samples to which known quantities of tracers had
been added. The counts of 30 standard samples exhibited a mean deviation of 6.7%.
Two to 5 specimens of each starfish species were treated by both methods for
each of the 8 different combinations of time interval and type of medium employed.
RESULTS
Quantity of nutrients taken up
Almost all the animals used in the study absorbed significant amounts of the
labeled nutrients made available to them. In the experiments involving the amino
TABLE I
Distribution of exogenous amino acids taken up by starfish tissues.
(Expressed as % of initial quantity to which animals were exposed)
Time
Species
Disk
Oral
body wall
Aboral
body wall
Stomach
Digestive
gland
Total
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
1 hour
H. sang.
1.85
8.1
1.54
19.0
0.72
10.8
0.23
0.1
0.47
0.2
4.81
38.2
H. sang.
0.69
4.3
1.40
27.2
0.63
8.5
0.11
0.1
0.32
0.1
3.15
40.2
A. forb.
0.70
11.2
1.26
25.0
0.67
26.8
0.06
0.1
0.40
0.1
3.09
63.2
A.'forb.
1.35
13.7
2.89
34.0
1.79
10.1
0.08
0.0
0.48
0.1
6.59
57.9
8 hours
H. sang.
0.51
10.3
1.37
27.1
0.76
11.7
0.07
0.1
0.17
0.1
2.88
49.3
H. sang.
1.16
10.3
2.91
21.0
1.17
17.9
0.20
0.2
1.41
0.4
6.85
49.8
A. forb.
0.79
6.7
1.12
21.4
0.56
9.6
0.13
0.1
0.36
0.2
2.96
38.0
A. forb.
1.01
9.1
1.57
18.7
0.74
10.7
0.22
0.1
0.49
0.1
4.03
38.7
72 hours
H. sang.
1.01
11.5
1.41
32.3
0.59
15.9
0.46
0.2
0.70
0.5
4.17
60.4
H. sang.
0.95
14.6
1.44
37.5
1.53
11.7
0.27
1.6
0.35
1.9
4.54
67.3
A. forb.
1.18
9.0
3.15
23.3
2.67
17.4
0.11
0.1
0.80
0.7
7.91
50.5
A. forb.
1.52
13.4
4.15
22.0
1.80
20.2
0.35
0.1
1.62
0.9
8.44
56.6
20 days
H. sang.
0.46
9.3
0.49
25.2
0.24
14.2
0.11
0.3
0.11
0.5
1.41
49.5
H. sang.
0.86
11.0
0.69
25.2
0.36
27.4
0.17
0.2
0.22
0.4
2.30
64.2
H. sang.
0.69
12.7
0.85
27.3
0.43
20.6
0.11
0.3
— •
0.5
2.08 +
61.4
A. forb.
0.72
16.0
1.48
27.7
0.64
21.1
0.09
0.1
0.44
0.4
3.37
65.3
A . forb.
0.77
12.9
0.48
35.1
0.77
14.5
0.08
0.1
0.40
0.4
2.50
63.0
A. forb.
0.69
17.2
1.41
25.8
0.81
18.0
0.05
0.1
0.24
0.4
3.20
61.5
164
JOHN CARRUTHERS FERGUSON
TABLE II
Distribution of exogenous glucose taken up by starfish tissues.
(Expressed as % of initial quantity to which animals were exposed)
Time
Species
Disk
Oral
body wall
Aboral
body wall
Stomach
Digestive
gland
Total
Wet
vvt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
1 hour
H. sang.
0.86
0.2
1.22
0.4
0.51
0.2
0.14
0.0
0.34
0.0
3.07
0.8
H. sang.
1.10
0.3
1.49
0.7
0.58
0.4
0.24
0.0
0.33
0.1
3.74
1.5
A.forb.
0.55
1.2
1.16
3.2
0.69
0.8
0.13
0.6
0.57
0.1
3.10
5.9
A.forb.
1.14
2.1
2.04
5.7
1.19
2.1
0.21
0.0
1.36
0.1
5.94
10.0
8 hours
H. sang.*
1.04
7.8
1.42
3.5
0.81
6.6
0.20
6.5
0.38
14.5
3.85
38.9
H. sang.*
0.96
7.6
1.47
2.5
0.71
3.1
0.23
13.9
0.46
38.6
3.83
65.7
A. forb.
1.11
8.5
2.60
27.3
2.09
13.7
0.09
0.7
0.58
0.2
6.47
50.4
A. forb.
1.34
7.1
3.16
27.9
2.21
8.8
0.18
0.1
1.07
0.2
7.96
44.1
72 hours
H. sang.*
0.94
2.6
1.51
5.8
0.81
3.8
0.27
1.3
0.59
4.7
4.12
18.2
H. sang.
1.15
1.2
1.92
7.3
0.97
2.7
0.16
0.1
0.55
0.1
4.75
11.4
H. sang.
0.89
0.9
0.99
1.5
0.47
0.8
0.20
0.1
0.24
0.2
2.79
3.5
H. sang.*
0.77
0.9
1.30
1.5
0.84
1.7
0.10
0.6
0.31
1.2
3.32
5.9
A . forb.
0.92
4.7
3.44
13.5
2.56
23.0
0.16
0.1
0.95
0.4
8.03
41.7
A. forb.
0.88
9.1
1.97
34.4
1.21
6.9
0.10
0.1
0.67
0.2
4.83
50.7
20 days
H. sang.*
0.73
1.6
0.80
1.3
0.41
0.6
0.08
1.4
0.12
1.1
2.14
6.0
H. sang.*
0.69
2.3
0.71
1.3
0.82
1.2
0.11
3.0
0.17
4.0
2.50
11.8
H. sang.
0.63
0.6
0.61
1.1
0.32
0.5
0.11
0.1
0.13
0.2
1.80
2.5
H. sang.*
0.22
0.8
0.34
0.8
0.22
0.6
0.03
1.8
0.05
2.2
0.86
6.2
H. sang.
0.32
0.3
0.38
0.6
0.21
0.5
0.05
0.1
0.10
0.1
1.06
1.6
A. forb.
0.46
3.4
1.23
10.5
0.89
2.9
0.08
0.1
0.30
0.1
2.96
17.0
A . forb.
0.95
6.6
1.84
25.8
0.98
6.8
0.16
0.3
0.74
0.3
4.67
39.8
A.forb.
0.68
4.0
1.85
12.5
1.38
5.2
0.04
0.1
0.46
0.1
4.41
21.9
acid mixture (Table I), usually about 40 to 65% of the radioactive elements initially
present was removed. Interestingly, most of this uptake appeared to take place
during the first hour of incubation. In fact, with Aslerias, the mean total values for
absorption were less after 8 hours than they were after 1 hour ( Fig. 1 ) . Consider-
ing the variation between the different specimens, however, this apparent decrease
probably would not have been observed if a larger number of animals had been
tested.
Nevertheless, a very large proportion of the total uptake of the amino acids did
take place with both species early in the incubation period. While the causes of this
effect are uncertain, the property could have been due to at least two factors.
Firstly, certain of the types of the amino acids included in the mixture presumably
are more easily absorbed than others, and thus, these types would become rapidly
depleted from the medium. The less-easily absorbed amino acids remaining after
the first hour would be taken up more slowly over a longer period of time. Secondly,
the organism could release substances which would accumulate in the sea water and,
after an interval, some of these might reach concentrations sufficient to inhibit the
NUTRIENT UTILIZATION BY STARFISHES
165
absorption of the amino acids which had not yet been taken up. Such an inhibition
would be relatively easy to achieve considering the small quantities of labeled amino
acid used. Both of these phenomena have been observed in previous experiments
30
20
M
O
Q.
X
tu
- 0
o
030
20
10
A. fbrbesi
D OWAWSt DG
D OWAW St DG
D OW AW St DG
D OWAW St DG
H. sanguinolenta
D OWAW St DG D OW AW St DG D OWAW St DG D OW AW St DG
I hr. 8 hr. 72 hr. 20 d.
FIGURE 1. Quantities of absorbed amino acids found in five different body regions of
specimens of two species of starfishes. Values (% of initial exposure) refer to the percentages
of the total initial C14-labeled amino acid present in the medium which were recovered from the
different groups of tissues (mean 2-3 specimens). The entire bars represent the total uptake
(digest method) while the cross-hatched areas represent material remaining unbound (alcohol
extract method). D, disk; OW, oral body wall; AW, aboral body wall; St, stomach; DG,
digestive glands. Over the 20-day period there is little redistribution of the absorbed amino
acids. For further explanation, see text.
166
JOHN CARRUTHERS FERGUSON
30
20
ZlO
o
ex
X
LU
_ 0
o
30
20
10
A. f o r b e s i
l
I
I
D OWAW St DG DOWAWStDG D OWAWStDG D OWAW St DG
H. son guinolen to
D OWAW St DG D OW AW S t DG
D OWAWStDG
D OWAWSt DG
I hr.
8 hr.
72 hr.
20 d.
FIGURE 2. Quantities of absorbed glucose found in five different body regions of specimens
of two species of starfishes. Values (% of initial exposure) refer to the percentages of the
total initial C14-labeled glucose present in the medium which were recovered from the different
groups of tissues (mean 2-5 specimens). Symbols are the same as in Figure 1. Some specimens
of Henricia apparently have taken up the glucose directly into their digestive organs. For
further explanation, see text.
dealing with the uptake of amino acids hy isolated starfish organs (Ferguson, 1964,
and unpublished data).
In contrast to the time course of amino acid uptake which was observed, glucose
apparently was absorbed continually over the 8-hour incubation period. This fea-
NUTRIENT UTILIZATION BY STARFISHES 167
ture can be seen in Table II where the values for total per cent uptake by the 8-hour
specimens are many times those of the 1-hour specimens. The glucose solution,
unlike the amino acid mixture, was homogeneous. Also, the molar concentration of
the glucose was considerably higher than that of the amino acids (because of its
lower specific activity). Thus, while the percentages of glucose taken up appear to
be somewhat lower than those of the amino acids, the actual quantities were probably
much greater. Likewise, at the end of the incubation period the concentration of
glucose still remaining in the medium was greater than even the initial concentration
of amino acid used.
Distribution of the absorbed nutrients
With a few specific exceptions, practically all of the labeled nutrients which were
taken up from the two types of medium were absorbed by the body wall components
of the starfishes (Figs. 1 and 2). Very little (less than 1%) normally found its
way into the internal organs. Even after 20 days there generally was no increase in
the radioactivity of these structures which could be considered significant. The
greatest quantities of the nutrients were most often found in the oral portions of the
body wall. These substances were probably absorbed by the extensive surface of
the tube feet and other areas of the epidermis of this region.
The mean values for the distribution of the glucose absorbed by Henricia (Fig.
2) present a pattern markedly different from that observed in the other cases. In-
deed, in looking at the 8-hour specimens, the distribution is seen to be almost com-
pletely reversed ; the least activity is found in the oral body wall and the greatest
in the digestive glands. A study of the actual data which were recorded (Table II)
helps to clarify what has happened. A number of the specimens of Henricia
(marked *) show large values for the percentages of material taken up into their
internal organs and low ones for the uptake into external parts. Other individuals
of the species exhibit the opposite distribution and in this sense more closely re-
semble the specimens of Asterias. It appears, then, that the marked specimens
responded to some stimulus, probably the relatively high glucose concentrations
employed, by initiating a kind of feeding reaction in which the dissolved nutrient
was removed from the medium by the internal digestive organs. The same phe-
nomenon can also be noted in the data for some of the specimens which were ex-
tracted with alcohol (Table IV), but since the values recorded from these analyses
are quite a bit lower, the differences do not stand out as pronouncedly.
Loss of nutrients taken up
After the completion of the 8-hour incubation period, there was little change in
the total amino acid radioactivity observed in the various specimens (Fig. 1, Table
I ) . Apparently, the tissues had a strong affinity for the amino acids once they had
taken them up, and over the 20-day period did not release them back into the sea
wrater or lose them through metabolism and respiration to any significant degree.
There was, however, a very marked loss of radioactive glucose from animals
over the same period (Fig. 2, Table II). This reduction was most obvious in
Henricia, but clearly also took place in the specimens of Asterias. While no evi-
dence was obtained relative to the fate of this lost material, it most probably dis-
168
JOHN CARRUTHERS FERGUSON
appeared through the breakdown of the sugar by the cells and its release as
respiratory CO2.
Utilisation of the absorbed nutrients
The analyses of the alcoholic extracts of the experimental animals provide data
(Tables III and IV) through which additional insight may be gained into the ways
in which the absorbed nutrients are utilized. This method measures only the
labeled material which remains in a relatively "unbound" state after it is taken up.
TABLE III
Distribution of exogenous amino acids taken up by starfish tissues and retained in an un-
bound (alcohol-soluble) state. (Expressed as % of initial quantity to which
animals were exposed)
Time
Species
Disk
Oral
body wall
Aboral
body wall
Stomach
Digestive
gland
Total
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
1 hour
H. sang.
1.08
2.5
1.48
6.5
0.54
6.3
0.17
9.2
0.40
0.4
3.67
15.9
H. sang.
0.82
1.9
1.18
5.6
0.54
1.8
0.09
0.1
0.16
0.1
2.79
9.5
A . forb.
0.51
6.5
1.06
10.9
0.52
6.0
0.14
0.1
0.39
0.1
2.62
23.6
A.forb.
1.03
4.6
1.94
8.8
1.74
7.0
0.04
0.1
0.68
0.0
5.43
20.5
8 hours
H. sang.
1.06
1.6
1.20
4.7
0.65
2.8
0.11
0.1
0.41
0.2
3.43
9.4
H. sang.
1.24
1.6
2.53
3.4
1.23
2.6
0.71
0.1
0.12
0.1
5.83
7.8
A . forb.
1.88
4.7
3.90
7.4
3.58
4.4
0.10
0.1
0.92
0.4
10.38
17.0
A.forb.
3.21
5.4
5.10
6.6
3.29
5.6
0.29
0.2
0.99
0.0
12.88
17.8
72 hours
H. sang.
0.81
1.2
1.20
2.6
0.47
1.0
0.47
1.0
0.34
0.1
3.29
5.9
H. sang.
1.36
1.3
1.49
1.7
0.70
1.9
0.16
0.2
0.27
0.1
3.98
5.2
A.forb.
0.73
2.7
1.42
4.0
0.69
3.0
0.12
0.1
0.36
0.2
3.32
10.0
A.forb.
2.51
2.4
3.30
2.9
1.64
2.3
0.33
0.0
1.04
0.1
8.82
7.7
20 days
H. sang.
0.43
0.7
0.78
1.5
0.30
0.9
—
0.2
0.01
0.2
1.52 +
3.5
H. sang.
0.36
0.6
0.42
1.2
0.18
0.9
—
0.1
0.01
0.1
0.97 +
2.9
A . forb.
1.10
1.5
1.67
2.1
1.23
2.5
0.16
0.2
1.07
0.3
5.23
6.6
A.forb.
1.29
1.4
2.53
2.9
1.59
1.9
0.15
0.2
0.80
0.2
6.36
6.6
By comparison of these data with the results of the digestive method, an estimate can
l)e obtained of the relative proportion of bound and unbound material retained by
the cells at each period. These differences can be appreciated most easily with the
aid of the two figures, by comparing the dark areas of each bar with the total length
of the bar.
By such means it can be seen that, except for the initial periods, only a fraction
of the material taken up normally was recoverable in the extracts. In the case of
the amino acids (Fig. 1), the size of the soluble fraction decreased progressively in
Asterias over the 20-day interval, from a peak at the end of the incubation period
of 45% of the total amount absorbed to a low of 10% after 20 days. In Henricia,
the range was from 32% at 1 hour to 5% at 20 days. It is interesting that in both
NUTRIENT UTILIZATION BY STARFISHES
169
species over one-half of the absorbed amino acid was unextractable with alcohol
after only a single hour of incubation. Apparently, some of the absorbed amino
acid was bound up quite rapidly while the rest remained in a soluble pool in the
cells and was incorporated into proteins or metabolized much more slowly. There
could possibly be some exchange between the soluble pool and the bound state. If
such an exchange does occur, it presumably would also prolong the apparent time
required for the extractable fraction to diminish.
TABLE IV
Distribution of exogenous glucose taken up by starfish tissues and retained in an un-
bound (alcohol-soluble) state. (Expressed as % of initial quantity to which
animals were exposed)
Time
Species
Disk
Oral
body wall
Aboral
body wall
Stomach
Digestive
gland
Total
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
Wet
wt.
gm.
%
1 hour
H. sang.
0.70
0.1
1.49
0.4
0.59
0.2
—
0.0
0.27
0.0
3.05 +
0.7
H. sang.
1.50
0.1
2.02
0.3
0.87
0.2
0.16
0.0
0.41
0.0
4.96
0.6
A.forb.
0.76
1.9
1.70
4.9
1.15
2.1
0.03
0.0
0.67
0.0
4.31
8.9
A. forb.
1.61
1.5
3.98
4.1
2.78
2.1
0.17
0.1
1.27
0.0
9.81
7.8
8 hours
H. sang.
—
0.8
.
1.5
0.6
0.5
0.8
4.2
H. sang.
—
0.7
—
1.8
—
1.3
—
0.1
— .
0.1
4.0
H. sang.
1.02
1.8
1.19
1.6
0.55
1.1
0.10
0.6
0.43
2.0
3.29
7.1
H. sang.
1.92
0.5
1.84
1.0
1.11
0.8
0.22
0.2
0.73
0.2
5.82
2.7
A . forb.
—
4.1
—
16.8
—
4.6
. — -
0.2
—
0.1
—
25.8
A.forb.
—
5.7
—
15.1
—
8.5
—
0.0
—
0.1
—
29.4
72 hours
H. sang.
0.99
0.3
1.29
0.7
0.60
0.5
0.12
0.1
0.36
0.1
3.36
1.7
H. sang.
1.66
1.3
2.02
1.3
0.79
0.9
0.29
2.6
0.50
3.2
5.26
9.3
H. sang.
1.25
0.7
1.26
0.5
0.81
0.7
0.14
0.9
0.25
0.7
3.71
3.5
H. sang.
1.20
0.5
1.59
1.1
0.91
0.9
0.21
0.1
0.36
0.3
4.27
2.9
A.forb.
0.99
3.1
0.96
7.9
2.48
7.0
0.07
0.1
0.42
0.1
4.92
18.2
A. forb.
1.86
8.4
3.34
15.8
1.79
7.0
0.16
0.1
0.93
0.2
8.08
31.5
20 days
H. sang.
0.91
0.6
1.24
0.4
0.54
0.3
0.23
0.7
0.36
0.6
3.28
2.6
H. sang.
0.22
0.9
0.41
0.6
0.21
0.7
0.02
2.7
0.07
4.0
0.93
8.9
A.forb.
0.95
3.4
2.17
6.1
1.48
3.4
0.06
0.2
0.65
0.3
5.31
13.4
A.forb.
0.91
2.7
1.66
3.0
2.20
0.3
0.11
0.1
1.08
0.2
5.96
6.3
Glucose was handled quite differently by the cells than were the amino acids.
By the end of the first hour essentially all the glucose taken up was still unbound
(Fig. 2). After 8 hours nearly 60% of the total quantity absorbed remained ex-
tractable in Asterias. (The calculated values for Henricia are insignificant because
of the great amount of individual variation resulting from the apparent feeding
behavior exhibited by some of these specimens.) Whether or not more of the
glucose became bound cannot be determined from the data, since in the 3- and 20-
day specimens there was a progressive loss of radioactivity, practically all of which
appeared to be from the unbound material. The progressive disappearance of the
170 JOHN CARRUTHERS FERGUSON
unbound glucose seems to suggest that this fraction was the first to be metabolized
and lost as CCX. Again, some exchanges possibly could have occurred between the
two fractions.
DISCUSSION
These experiments complement earlier work on the utilization of exogenous
nutrients by starfish and confirm that at least two very different species of these
animals possess biochemical mechanisms which enable them to remove various types
of amino acids and glucose from sea water. These mechanisms apparently are
efficient in picking up nutrients from even very dilute solutions. The limits of
effectiveness of the absorptive machinery, however, have not been determined.
Neither has much evidence yet been gathered concerning its chemical and physical
properties.
While absorption probably can occur over all areas of the body surface, the
greatest activity takes place in the oral region. This locality doubtless has the
largest area of free surface, and very likely is more exposed to circulation of water
than the other parts. It includes the tube feet, that protrude into the medium, and
the entire region of the ambulacral groove, which is probably efficiently ventilated
by means of ciliary tracts. Such tracts have been described repeatedly in various
species, including A. forbesi (Budington, 1942) and H. sanguinolenta (Anderson,
1960; Rasmussen, 1965).
The full significance of the epidermal absorptive process is still uncertain. If
Pequignat (1966) is correct in his conclusion that epidermal digestion by skin
glands is a common phenomenon in starfish, one would expect the organisms to
possess adequate mechanisms for the absorption and utilization of the different kinds
of products released by such action. While the present investigation has indicated
that some types of amino acids and glucose may be taken up through the epidermis,
it is still undetermined if all the myriad types of organic compounds which pre-
sumably would be released through such a digestive process could be handled. In
fact, as mentioned previously, the pattern of uptake observed for the mixture of
amino acids suggests that certain types, representing nearly a third of the mixture,
may not be readily absorbed. Likewise, there is as yet no confirmation that carbo-
hydrates other than glucose can be utilized. Further investigations are contemplated
which will more fully evaluate the diversity of compounds which may be taken up
by epidermal mechanisms.
During the 20-day period in which the animals were studied there was little, if
any, indication that nutrients were passed on to the internal regions of the body
from the absorptive sites on the body surface. Very small amounts of radioactivity
were detected in the internal organs of a few of the test specimens after several
days, but since little consistency was seen, this activity was probably due to un-
avoidable contamination of the separate samples. Also, a few specimens may have
ingested some of the slime, mucus, and algae which accumulated on the walls of the
holding tanks, and this material could have picked up a slight amount of radio-
activity. In any case, as the values observed for the internal organs are too low to
be credited with significance, it should probably be concluded that epidermal absorp-
tion functions almost solely for the benefit of the superficial tissues.
NUTRIENT UTILIZATION BY STARFISHES 171
The apparent feeding reaction exhibited by some of the specimens of H.
sanyuinolcnta in the glucose medium is most interesting. Anderson (1960) care-
fully studied the structure and function of the digestive organs of this species and
concluded that its Tiedemann's pouches were a "hydrodynamic organ or flagellary
pump of prodigious effectiveness" (p. 393). He showed that Henricia was pri-
marily a filter-feeder and could take up and entrap such material as suspended
Mytihts sperm. Feeding experiments were also performed on Henricia by Rasmus-
sen (1965). These were more quantitative than Anderson's and served to confirm
further the great efficiency of this animal as a particle-suspension feeder.
The present observations reveal that the flagellary feeding mechanism described
by the above workers can also be effective in the utilization of dissolved nutrient
materials of relatively low molecular weight. The pumping mechanism of Henricia
is apparently so efficient that it "pays" the animal to take up solutions of nutrients,
provided they occur in at least minimal concentrations. The most significant aspect
of these observations, however, is not so much the uptake of the dissolved materials,
but rather, the nature of the stimulus which caused them to be taken up. Although
further verification is needed, the stimulus appears to have been the relatively
higher concentration (when compared to that of the amino acids) of the glucose
solution used. This was the only variable observed in the experiments other than
the type of compounds themselves.
Doubtless, in nature these animals frequently encounter various kinds of dis-
solved nutrients in equivalent or even greater concentrations than those used in the
experiments. Some of these probably come from the external digestion of relatively
solid organic substrates. The stomach of Henricia is rather unique among starfishes
in possessing numerous zymogen cells (cf. Anderson, 1960), which likely are a
source of enzymes for such a process. In the present experiments specimens were
often seen in an apparent feeding position, with their stomachs everted as button-like
protuberances applied against the algae-covered aquarium wall or between the valves
of a gaping clam. Under normal circumstances, digestive products released during
this activity would probably set off the pumping process. As the glucose in the
experiments seems to have elicited the same response as the natural stimulus, one
can conjecture that encounter by the animal of a significant concentration of dis-
solved nutrients in its environment could serve also as an effective stimulus for
initiating the pumping process. Once pumping is started, the soluble nutrients are
efficiently taken up into the internal digestive organs.
Henricia, then, seemingly obtains its nutrition through several different processes.
It depends primarily on the suspended and dissolved materials normally present in
the environment, but probably also can digest some solid food outside of its body.
These nutritional substances apparently are taken up by means of flagellary cur-
rents, and absorbed internally, or, at least in part, are directly assimilated by the
superficial tissues of the body which are also exposed to the substances.
An uptake of labeled glucose into the digestive system of Astcrias, comparable
to that observed in Henricia, was not noted. This difference in behavior probably
was due to the fact that Asterias is primarily a predator. While it lacks the complex
pumping apparatus possessed by Henricia, it does possess ciliated surfaces on its
stomach. It relies on currents produced on these surfaces to bring in concentrated
solutions of nutrients from victims digested externally by enzymes supplied from the
172 JOHN CARRUTHERS FERGUSON
digestive glands via gutters in the stomach wall (cf., Anderson, 1954). This process
is probably not altogether different from the pumping of Henricia. In a previous
note (Ferguson, 1963b), for example, I reported the uptake into the digestive
organs of Asterias of C14-labeled glucose and amino acids which had been injected
into small clams just before they were fed to the starfish. In that case, the presence
of the solid pieces of food appears to have stimulated the animals to activate their
feeding mechanism. As in Henricia, once feeding was initiated, uptake of the dis-
solved materials into the digestive organs proceeded rapidly.
The probability that epidermal absorption of exogenous nutrients is a continuous
process while normal feeding is generally a discontinuous one is perhaps quite
significant. In a sense, the two activities may balance each other as sources of
nutrition over a period of time. If such is the case, the internal regions of the body
may be seen as receiving nutrition almost exclusively via the digestive tract, while
the more external tissues would be nourished to a considerable extent directly
through the epidermis. One might suppose, then, that if an animal were prevented
from utilizing either one of the sources, it probably could not survive. In this vein,
investigations have shown that various species of starfishes can live long periods
with little or no visible food, but they cannot subsist indefinitely under such condi-
tions (Galtsoff and Loosanoff, 1939; Vevers, 1949). It would be much more diffi-
cult to design an experiment in which specimens were allowed to eat but completely
denied epidermal absorption. But since it has been determined that epidermal
absorption of nutritional materials does occur, and can take place to a significant
degree, it seems reasonable to conclude that such absorption is an important factor
in the economy of these organisms.
SUMMARY
1. Small specimens of A. forbesi and H. sanguinolenta were exposed to dissolved
C14-amino acids and glucose. The subsequent distribution of these materials was
then determined in the following five regions of the body: disk (including the
gonads), oral body wall of the rays, aboral body wall of the rays, stomach, and
digestive glands.
2. In all cases, large proportions of the labeled nutrients were taken up into the
external tissues. The largest amount was usually absorbed into the oral body wall,
which probably possesses a proportionately greater ventilated surface area than the
other regions.
3. Over a period of 20 days there was little indication of movement of the
externally absorbed nutrients into the internal organs. In this period, very little
loss of amino acid radioactivity was noted. The amino acids became progressively
less soluable in alcohol, suggesting that they were incorporated into the structural
proteins of the organism.
4. Glucose radioactivity declined progressively over the 20-day period. As
observed in Asterias, this decline occurred almost exclusively in the portion of
absorbed glucose that remained alcohol-soluble. This fraction was possibly used as
an energy source while the insoluble fraction became incorporated into more inert
elements.
5. A number of the specimens of Henricia appeared to pump up and absorb the
glucose medium into their digestive organs. This was interpreted as a form of
NUTRIENT UTILIZATION BY STARFISHES 173
feeding behavior possibly initiated by the relatively high concentration of glucose
used. The much less concentrated amino acid medium failed to initiate such a
reaction.
6. It is concluded that nutrition in starfish is probably a dual process involving
both a continuous epidermal absorption of dissolved exogenous materials for the
benefit primarily of the superficial tissues, and intermittent oral feeding to satisfy
the more general needs of the entire organism and especially of the internal organs.
LITERATURE CITED
ANDERSON, J. M., 1954. Studies on the cardiac stomach of the starfish, Asterias forbesi. Biol.
Bull, 107: 157-173.
ANDERSON, J. M., 1960. Histological studies on the digestive system of a starfish, Henricia,
with notes on Tiedemann's pouches in starfishes. Biol. Bull., 119: 371-398.
BUDINGTON, R. A., 1942. The ciliary transport-system of Asterias forbesi. Biol. Bull., 83:
438-450.
FERGUSON, J. C, 1963a. The physiological mechanisms of nutrient transport in the starfish,
Asterias forbesi. Doctoral dissertation, Cornell University.
FERGUSON, J. C., 1963b. An autoradiographic study of the distribution of ingested nutrients in
the starfish, Asterias forbesi. Amer. Zool., 3: 524.
FERGUSON, J. C., 1964. Nutrient transport in starfish. II. Uptake of nutrients by isolated
organs. Biol. Bull, 126: 391-406.
GALTSOFF, P. S., AND V. L. LOOSANOFF, 1939. Natural history and method of controlling the
starfish (Asterias forbesi, Desor). Bull. U. S. Bureau Fisheries, 49: 75-132.
PEQUIGNAT, E., 1966. Skin digestion and epidermal absorption in irregular and regular urchins
and their probable relation to the outflow of spherulecoelomocytes. Nature, 210: 397-
399.
PUTTER, A., 1909. Die Ernahrung der Wassertiere und der Stoffhaushalt der Gewasser.
Fisher, Jena.
RASMUSSEN, B. N., 1965. On taxonomy and biology of the North Atlantic species of the
asteroid genus Henricia Gray. Meddelelser fra Damnarks Fiskeri- og Havundersfigel-
scr,4: 157-213.
STEPHENS, G. C., 1962. Uptake of organic material by aquatic invertebrates. I. Uptake of
glucose by the solitary coral, Fungia scutaria. Biol. Bull., 123: 648-659.
STEPHENS, G. C., 1963. Uptake of organic material by aquatic invertebrates. II. Accumula-
tion of amino acids by the bamboo worm, Clvmcnclla torquata. Comp. Biochem.
Physiol, 10: 191-209.
STEPHENS, G. C., 1964. Uptake of organic material by aquatic invertebrates. III. Uptake of
glycine by brackish-water annelids. Biol. Bull., 126: 150-162.
STEPHENS, G. C., AND R. A. SCHINSKE, 1961. Uptake of amino acids by marine invertebrates.
Limnol. and Oceanog., 6: 175-181.
STEPHENS, G. C., AND R. A. VIRKAR, 1965. Accumulation and assimilation of amino acids by
the brittle star, Ophiactis simplex. Amcr. Zool., 5: 661.
STEPHENS, G. C., AND R. A. VIRKAR, 1966. Uptake of organic material by aquatic inverte-
brates. IV. The influence of salinity on the uptake of amino acids by the brittle star,
Ophiactis arenosa. Biol. Bull, 131 : 172-185.
VEVERS, H. G., 1949. The biology of Asterias rubens L. : growth and reproduction. /. Mar.
Biol. Assoc., 28: 165-187.
ENVIRONMENTALLY CONTROLLED INDUCTION OF PRIMARY
MALE GONOCHORISTS FROM EGGS OF THE SELF-
FERTILIZING HERMAPHRODITIC FISH,
RIVULUS MARMORATUS POEY
ROBERT W. HARRINGTON, JR.
Entomological Research Center, Florida State Board of Health,
Vero Beach, Florida 32960
The fact of genetic sex determination among teleosts is well established by sex
linkage (Gordon, 1957), although cytological demonstrations of fish sex chromo-
somes have not withstood critical scrutiny (White, 1954) until recently (Nogusa,
1960). Both male and female fish with phenotypic sex contrary to genotypic sex
have been produced by early treatment with sex steroids (Yamamoto, 1953-1961).
Can genetic sex determination in fishes be overridden also by external environmental
factors, as in some amphibians, is the question to which answers were sought in the
experiments to be reported here.
This question was first raised by effects on anuran sex determination of delayed
fertilization (overripeness of eggs) and of temperature obtained, respectively, by
Pfliiger (1882) and Witschi (1929). Comparable experiments on fishes have been
few ; their long duration with no assurance of negotiable results discourages investi-
gation. Under harsh contrasting experimental conditions it is extremely difficult to
rear fish through the early crises of ontogeny without excessive losses, and if
mortalities exceed a certain limit, the dilemma of a differential mortality of one sex
versus experimental induction of the other cannot be resolved.
Conclusive evidence of environmental influence on sex determination in teleosts
is lacking despite possible indications of such influence from experiments on one
species each of the genera Salmo (Mrsic, 1923), Bctta (Eberhardt, 1943), and
Anguilla (D'Ancona, 1950, 1960). Only by making explicit certain crucial defects
in these experiments passed over by reviewers can the rationale of our own experi-
ments and the cogency and singularity of their results be given their full context
(see Discussion).
The cyprinodontid fish used in the present study, Rivulus marmoratus Poey, is
unique among fishes so far as known in being comprised of natural, consistently
self-fertilizing hermaphrodites (Harrington, 1961, 1963; comments of Atz, 1964).
Its hermaphroditism is normal and not a laboratory artifact like that of Lebistes
reticulatus (Spur way, 1957), for example. Long deemed merely a nominal species
(Carman, 1895), R. marmoratus was revived as a valid species by Rivas (1945),
who rediscovered its types in the U. S. National Museum, but was unknown as a
living fish until it was found in Florida (Harrington and Rivas, 1958). Tissue
grafts between Florida wild-caught progenitors and their laboratory-reared descend-
ants gave the autograjt reaction (Kallman and Harrington, 1964), indicating that
they have the same genotype and that probably their wild antecedents also had re-
produced by self-fertilization, i.e. for upwards of 10 generations (see below).
174
LOW-TEMPERATURE-PRODUCED MALE FISH
175
Although an allegedly gonochoristic subspecies, R. inarmoratus bonairensis, was
described from the Antilles (Hoecleman, 1958) the same year that we found R.
inarmoratus in Florida, we have no evidence from the laboratory or from the wild
that females exist in Florida and so far have found no males in the wild. It came
as a surprise, therefore, when males appeared among hermaphrodites propagated in
our laboratory, especially since these were bright orange with the caudal ocellus
obsolescent, in sharp contrast to the hermaphrodites. The incidence of males has
SEA WATER
MORTALITY
BRIGHT
LIGHT
DIM
LIGHT
FRESHWATER
BRIGHT
LIGHT
DIM
LIGHT
HIGH
TEMPER-
ATURE
BO'C.
LOW
TEMPER-
ATURE
FIGURE 1. Plan and results of Experimental Series One, data in Table I. Individuals of
Rivulus marmoraius, each in its own jar, were exposed to the eight combinations of bright or
dim light, sea water or fresh water, high or low temperature. Exposure was from not later
than the I blastoderm stage until sexual maturity at high temperature or five months post-
hatching at low. Circles show the percentages of hermaphrodites and males and the percentage
mortality resulting from exposure to each extrinsic factor combination. The temperature of
the lowermost row of treatments was raised to 20° C. when 18° C. proved lethal in combination
with sea water, and the middle row, at 20° C., was added. Compare with the results of
Experimental Series Two (Table IV).
stayed below 5% through more than 10 uniparental laboratory generations, number-
ing over 350 fish, each isolated throughout life to exclude physiological interactions
of any kind except visual ones.
The appearance of an occasional male in clones otherwise composed of her-
maphrodites suggests some lability in the sex-determining mechanism through which
the genotype normally produces the hermaphrodite phenotype. It seems proper to
speak here of a hermaphrodite genotype, because as Atz (1964) observes, the
assumption is false that normal hermaphroditism cannot be genetically controlled,
176
ROBERT W. HARRINGTON, JR.
as is the sex of gonochorists. The present experiments were contrived to identify a
possible external environmental factor capable of causing a deviation to the male
phenotype during sex differentiation. Positive results were obtained in two series
of experiments, the first begun in August, 1961, the last completed in January, 1965.
MATERIALS AND METHODS
Few if any experiments on vertebrates can have used material as genetically
uniform as the Rivuhts marmoratus eggs used here. In Experimental Series One
TABLE I
Effects of external factors on the sex ratio of uniparental offspring of Rivulus marmoratus
hermaphrodites. Self-fertilized eggs were reared from outset of extraparental incubation
under various combinations of light intensity, salinity and temperature.
B, bright light; D, dim light; S, sea water; F, fresh water;
18/20°C., started at 18° C. but continued at 20° C.
Survivors
Non-survivors
Treatment
Surviv-
ing/
treated
Per-
cent-
age
sur-
Hermaphro-
dites
Males
Died in ovo
or at hatching
Died very
small
Extremely
abnormal;
discarded
vival
Total
%
Total
%
Total
%
Total
%
Total
%
30° C.
B S
6/10
60.0
6
100.0
0.0
4
100.0
0.0
0.0
DS
5/7
71.4
5
100.0
0.0
1
50.0
1
50.0
0.0
B F
7/12
58.3
6
85.7
1
14.3
2
40.0
3
60.0
0.0
DF
8/17
47.1
8
100.0
0.0
6
66.7
1
11.1
2
22.2
20° C.
B S
10/15
66.7
4
40.0
6
60.0
4
80.0
0.0
1
20.0
DS
11/16
68.8
2
18.2
9
81.8
3
60.0
0.0
2
40.0
B F
7/12
58.3
3
42.9
4
57.1
1
20.0
2
40.0
2
40.0
DF
7/14
50.0
2
28.6
5
71.4
2
28.6
5
74.1
0.0
18/20° C.
B S
1/19
5.3
0.0
1
100.0
16
88.9
0.0
2
11.1
DS
1/12
8.3
0.0
1
100.0
5
45.4
3
27.3
3
27.3
B F
4/8
50.0
1
25.0
3
75.0
1
25.0
1
25.0
2
50.0
DF
6/8
75.0
0.0
6
100.0
0.0
2
100.0
0.0
30° C.
All
26/46
56.5
25
96.2
1
3.8
13
65.0
5
25.0
2
10.0
20° C.
All
35/57
61.4
11
31.4
24
68.6
10
45.5
7
31.8
5
22.7
18/20° C.
All
12/47
25.5
1
8.3
11
91.7
22
62.9
6
17.1
7
20.0
( Fig. 1 and Table I ) , the fish surviving to be sexed hatched from eggs of hermaph-
rodites of two clones, 32 Clone-NA eggs and 41 Clone-DS eggs (Table II). The
fish of Table II are coded as they were when used in the graft tests providing the
evidence for these clones (Kallman and Harrington, 1964). Contrary to two data
(he. cit., Table III, #11 and #IV), however, Fish DS, Fish FT, and Fish NSU
all belong to the same clone, later interline grafts (unpublished) having given the
autograft reaction, showing that the previous rejections (#11 and #IV) were
mechanical and not immunological. In Experimental Series Two (Table IV), all
were Clone-DS eggs of Uniparental Laboratory Generations 9, 10, and 11, so that
LOW-TEMPERATURE-PRODUCED MALE FISH
177
TABLE II
Sex ratios of progeny of self-fertilized Rivulus marmoratus reared ab ovo from outset of extra-
parental incubation at either high or low temperature, shou'ing the same temperature
correlation regardless of parentage or clone.* Same data as in Table I
High temperature (30° C.)
Low temperature (18-20° C.)
Progeny
Percent-
Parent
surviv-
age
Hermaphrodites
Males
Hermaphrodites
Males
ing/
treated
survival
Total
%
Total
%
Total
%
Total
%
FT
19/42
45.2
5
100.0
0.0
2
14.3
12
85.7
NSU
2/2
100.0
1
100.0
0.0
0.0
1
100.0
DSP,
6/11
54.5
3
100.0
0.0
1
33.3
2
66.7
Fi
6/28
21.4
1
100.0
0.0
3
60.0
2
40.0
F2
8/14
57.1
1
100.0
0.0
4
57.1
3
42.9
NA
32/52
61.5
14
93.3
1
6.7
?
11.8
15
88.2
NSB
0/1
0.0
Totals
73/150
48.7
25
92.6
1
3.8
12
25.5
35
74.5
* Wild-caught fish FT, NSU, and DS and their uniparental descendants belong to the same
clone; wild-caught NA belongs to a different clone (Kallman and Harrington, 1964, and un-
published).
Ill
TABLE
Sex ratios of progeny of self-fertilized Rivulus marmoratus reared at either high or low temperature,
showing the same temperature correlation regardless of developmental stage at outset of treatment
( = outset of extraparental incubation). Same data as in Tables I-II. Stage 1 is the
fertilized egg before polar cap formation; at Stage 13c the blastoderm
encloses f of the yolk. For intervening stages see Harrington, 1963
High temperature (30° C.)
Low temperature (19-20° C.)
Develop-
mental
Surviv-
ing/
Percent-
age
Hermaphrodites
Males
Hermaphrodites
Males
stage at
outset
treated
survival
Total
%
Total
%
Total
%
Total
%
i
0/1
0.0
2
1/4
25.0
0.0
1
100.0
3
1/5
20.0
0.0
1
100.0
4
5/13
38.5
4
80.0
1
20.0
5
8/16
50.0
3
100.0
0.0
2
40.0
3
60.0
6
16/21
76.2
4
100.0
0.0
1
8.2
11
91.8
7
5/15
33.3
2
100.0
0.0
1
33.3
2
67.0
8a
4/6
66.7
2
100.0
0.0
0.0
2
100.0
8b
5/10
50.0
2
100.0
0.0
0.0
3
100.0
8c
2/6
33.3
2
100.0
0.0
8d
9/16
56.3
2
100.0
0.0
2
28.6
5
71.4
9
5/17
29.4
1
100.0
0.0
3
75.0
1
25.0
10
6/9
66.7
2
100.0
0.0
1
25.0
3
75.0
11
0/1
0.0
12b
3/4
75.0
2
100.0
0.0
0.0
1
100.0
13a
1/2
50.0
1
100.0
0.0
13c
2/3
66.7
0.0
->
100.0
Totals
73/150
48.7
25
96.2
1
3.8
12
25.5
35
74.5
178
ROBERT W. HARRINGTON, JR.
besides the immunological evidence that the fish at the outset of these generations
were of one clone (Kallman and Harrington, 1964), selfing through eight gener-
ations alone would have brought them to over 99^ homozygosity (Sinnott and
Dunn, 1939; p. 284). Alternatively, in the remote contingency of a homozygote-
preventing mechanism, they would share the same heterozygous genotype.
The wild-caught progenitors of Table II were isolated from date of capture.
Every other fish referred to in this report was kept in lifelong isolation begun at its
retrieval as a self-fertilized egg being emitted by its parent. Eggs of R. niarnwrafns
are laid after intraparental incubation for from a few minutes to 2\ days, vis. from
in Stage 1 (just fertilized) to in Stage 24 (prominent pectoral fin buds), as before
described (Harrington, 1963). Eggs for our experiments were sucked into a
pipette as they fell from laying hermaphrodites, kept at a water temperature of about
TABLE IV
The sex determination and differentiation of uni pa rental Rivulus marmoratus modified by tempera-
ture. Self-fertilized eggs from hermaphrodites of a single clone were reared under contrasting
temperature regimes, but with light intensity and salinity controlled. Compare with
Table I and Figure 1
Total
Hermaphrodites
Males
Mortality
Tc.T-r»t-.cit-o f iii-t. t-di-riT-nti
1 Clll lJt_IclL III" iCgllllv;
"66^
reared
No.
%
No.
%
No.
%
A) 25 ± 1° C. to maturity (control)
50
50
100
0
0
0
0
B) 25 ± 1° C. through hatching;
19.5 ± 0.5 °C. for the first 5
months post-hatching; 25 ±
1° C. thereafter to maturity
50
46
92
0
0
4
8
C) 25 ± 1 ° C. to at least Stage 16
but not beyond Stage 22a ;
then 19.5 ± 0.5° C. through
eclosion* and for 5 months
post-eclosion ; thereafter 25
± 1° C. to maturity
50
9
18
36
72
5
10
* Eclosion refers to either hatching or being cut out of the chorion.
25° C. Developmental stages at outset of experimental treatment varied according
to the experiment. Each egg of suitable stage was pipetted into its own jar. The
egg in its own jar was put under the conditions of its allotted treatment, encom-
passing extra parental incubation, hatching and subsequent life in this jar, except for
one low-temperature treatment (Table IV, B) begun with hatchlings.
Wide-mouthed, straight-sided, cylindrical, screw-top jars were used. These
were about 15 cm. high and 8 cm. in diameter, holding 950 ml., and were filled with
600 ml. of water. Plastic Petri-dish covers used as lids prevented escape, without
injury to jumping fish or interference with gas exchange. Fish have lived in these
jars for 45 months, and as many as 500 were kept concurrently, each in its own jar,
during otir experiments. Unremitting care was taken to exclude any possibility of
transfer between jars of physiological substances. A utensil inserted in one jar was
rinsed repeatedly in a container overflowing with fast-running water before being
inserted in another,
LOW-TEMPERATURE-PRODUCED MALE FISH 179
Jars going from propagating room (25° C.) to the low experimental temperature
(constant-temperature room) were moved immersed in a water bath at 25° C.
When the jar water reached the low temperature, the jars were taken out, wiped
off, and left in the constant-temperature room. Jars going to the high experimental
temperature were put in a water hath to raise the jar water to this high temperature.
Then the jars were moved to the constant-temperature room and left immersed for
the duration of the treatment in water baths thermostatically controlled to maintain
the high experimental temperature. These procedures were reversed when fish
were returned to the propagating room for post-treatment observation. The condi-
tions of each experimental treatment were maintained continuously. As new eggs
became available in temporal succession, each wras allocated to one of the treatments
singly and in its own jar. Each jar was removed from the treatment conditions
and returned to the propagating room when its fish reached functional sexual
maturity, or after five months in the case of low-temperature treatments.
The fish in jars returned to the propagating room for post-treatment observation
were kept in 40% sea water for the remainder of their lifelong isolation. If not
originally in 40% sea water, they were changed to it gradually over three days from
their former (treatment) salinity. The propagating room received light within the
natural daily photoperiod only, and the room temperature was constrained by an air
conditioner and thermostatically controlled heater to hold the water temperature to
about 25° C.
With the release of the hatchling from its egg chorion (terminology of Lord
Rothschild, 1958), feeding was begun, first with microworms (nematodes), then
with these mixed with brine shrimp (Artcmia} nauplii. A premature diet of
Artemia can cause death through intestinal stoppage, the shift to food of larger size
evidently being a crisis of ontogeny. Afterwards, brine shrimp alone were used.
Feeding was ad libitum; the amount squirted into the jar with a syringe was
adjusted to fish size and to volume of unconsumed food in its jar each morning.
Unless otherwise stated, food was introduced in water of the same salinity as in the
jar. After feeding began, the jar water was filtered weekly, by pouring through
filter paper on a glass funnel into a clean jar. The fish was transferred by syringe
or net to the filtrate when it was deep enough. The old jar was washed, and filtrate
and fish poured back into it. The new jar, the funnel, and the syringe or net were
washed, and, with fresh filter paper, used for the next filtering. Cloudy water was
replaced. New water replacing old or added to compensate for loss was of the same
temperature and salinity. Jar water was kept a pale blue with methylene blue.
Without this bacteriostatic dye eggs and larvae seemed to have a lower survival, but
this was not tested experimentally. One or more times a day, solid wastes, uneaten
food, and later on, eggs, if laid by the then mature fish, wrere sucked out with a
syringe.
The jars were monitored daily, at first for hatchlings or eggs in Stage 31. which
precedes hatching (Stage 32) under natural conditions, and for abnormal, sick and
dead eggs or hatchlings, later for the first external signs of sex differentiation.
Throughout the life of each hermaphrodite a daily record was kept of the number,
conditions, and stages of eggs found in its jar. Under certain treatments, environ-
mental cues normally triggering the hatching mechanism were either absent or
nullified (cj. Kinne and Kinne, 1962), because hatching proved to be another crisis
USO ROBERT W. HARRINGTON, JR.
of ontogeny. Prolonged delays caused deaths through inanition (hut see Har-
rington, 1959). In the first series of experiments, the mechanism could sometimes
be activated by focusing bright light on overdue eggs, but it became the practice to
cut from their chorions other embryos of the same age as those hatching naturally
or with light stimulation (Table V ) . In the second series, the crisis was circum-
vented by cutting out all embryos incubated at low temperature (Table IV, C) long
before the normal hatching stage, allowing the rest (Table IV, A and B) to hatch
naturally, and mortalities did not exceed statistically permissible limits in the
second series as they did in the first.
Hermaphrodites and males at the same temperature become externally recog-
nizable as such at about the same age and size. Hermaphrodites retain the caudal
ocellus possessed by both juvenile hermaphrodites and juvenile males, and acquire
no orange pigmentation later. Sperm production in hermaphrodites is not copious
enough to be visible as milt. The characteristic behavior pattern leading up to
oviposition (Harrington, 1961) is confined to hermaphrodites. Each hermaphro-
dite was performance-proven in lifelong isolation by laying eggs from which normal
fish hatched. Maturing males first acquire scattered, small orange spots on the
body and minute orange flecks in increasing density on the fins. An orange wash
later covers the whole body, as the caudal ocellus becomes obsolescent or vanishes
altogether. Handled males often release milt, which in hanging-drop suspension
shows active spermatozoa.
It is of fundamental importance to make clear that these are males ab initio,
sometimes called "pure males," but more exactly, primary male gonochorists. This
was established by serial sections of over 200 gonads from ontogenetic series of both
hermaphrodites and males, ranging from germ cell entry into genital ridges to
senility, as well as by contrast with testes of secondary male gonochorists, that may
arise from older hermaphrodites by involution of the ovarian component of the
ovotestes with proliferation of the testicular component, under conditions to be
described in a separate report. Embryological and histological details are beyond
the scope of the present report, but will be given elsewhere. A testis and an ovo-
testis, each in transverse section, and a male and hermaphrodite, both adults, are
shown in Figure 2.
EXPERIMENTAL SERIES ONE
The treatments were the eight combinations of high versus lo\v temperature, sea
water versus fresh water, and bright versus dim light (Fig. 1 and Table I). Eggs
ranged from the one-cell stage to Stage 13c (f blastoderm) at outset of treatment
(Table III), which began \\ithin five minutes after the egg was emitted by its
parent. The constant-temperature room thermostat was set at first to give a low
water temperature of 18° C., but this proved lethal in combination with sea water
(Fig. 1 and Table I). Survivors begun at 18° C. were continued at 20° C., and
eggs of a new set were started at 20° C. These low water temperatures actually
fluctuated between 20° C. and 21° C., being mostly closer to 20° C., and will be
referred to as 20° C. for convenience. The high-temperature jars were immersed
to the level of the water within them. They rested on a wire-screen platform above
the bottom, their lids just clearing the underside of the glass aquarium covers. The
bath water was circulated by air stones to equalize the temperature, held by thermo-
LOW-TEMPERATURE-PRODUCED MALE FISH
181
FIGURE 2. Young adult Rivulus marmoratus. a, hermaphrodite ; b, primary male gonocho-
rist ; c, cross section of right lobe of ovotestis ; d, its testicular component at higher magnifica-
tion ; e, cross section of right lobe of the testis of a primary male gonochorist, same magnification
as in c.
182
ROBERT W. HARRINGTON, JR.
FIGURE 3. Structural-functional abnormalities of Rivuhts mannoratus exposed to certain
light-salinity-temperature combinations. Compare with Figure 1. a, prolapsed oviduct; con-
fined to the dim-light, sea-water, high-temperature treatment; in 100% of the survivors; b,
pharyngcal hypcrplasiu ; confined to the dim-light, fresh-water, low-temperature treatments,
whether begun at 18° C. and continued at 20° C, or at 20° C. throughout; in 100% of the
LOW-TEMPERATURE-PRODUCED MALE FISH 183
stat within 30 ± 1° C. Jars contained either filtered 100% sea water (salinity.
36f/(f ) or honied drinking water ("Blue Crystal"). Illumination was provided by
four 40-watt \Yestinghouse Daylight fluorescent lamps suspended from above and
controlled by a time switch giving a 14-hour photoperiod (0500h-1900h ). Bright-
lit jars received direct light of low daylight intensity (425-520 lux). Dim-lit jars
were within a cube-shaped enclosure covered with black cloth on all but one side,
either in a water bath or on shelves. They received indirect light of low intensity
(2.55-21.00 lux), i.e., mostly above the lower end of the intensity range (3.5-400.0
lux) of civil twilight (see Nielsen, 1961, 1963). Before hatching occurred, jars
were examined once a day by flashlight ; afterwards, although the hatchlings could
feed by sight, each jar was taken out into the direct light for less than a minute
each day to be checked and cleaned.
RESULTS OF EXPERIMENTAL SERIES ONE
Of the 150 eggs allotted to the eight treatments, 73 or 40.7% survived to sexual
differentiation and functional maturity. Thirty-seven or 50.7% of these were her-
maphrodites ; 36 or 49.3% were males, an absolute number of males over seven
times the total number encountered before the experiments were performed. All
but one experimental male were from the low-temperature treatments (Table I).
Figure 1 shows the incidence of males and hermaphrodites and the mortalities as
percentages of the total number of eggs per treatment. In Table I, the data are
given in actual numbers and also as percentages of survivors and non-survivors.
There is no indication that the incidence of males versus that of hermaphrodites
was affected by the alternative salinities and light intensities in any of their four
possible combinations, even though two light-salinity-temperature combinations
resulted in far higher mortalities than the rest, viz.. either bright or dim light with
sea water at 18° C. (see above). This is obvious from Figure 1, and warrants
placing all high-temperature fish in one group and all low-temperature fish in
another, as is clone variously in Tables I-III. The correlation of male incidence
with low-temperature rearing holds regardless of parentage or clone (Table II),
or of embryonic stage at outset of treatment between Stages 1 and 13c (Table III ).
There are indications, besides, that death was caused differently among the
various light-salinity-temperature treatments, making it unlikely that the alterna-
tive temperatures per se caused alternative differential mortalities of hermaphro-
dites versus males. These indications are structural-functional abnormalities pecu-
liar to certain light-salinity-temperature treatments. The names given them here,
prolapsed oviduct, pharyngcal hypcrplasia, and kyphosis, are intended to be no
more than descriptive (Fig. 3). These abnormalities suggest that some deaths
came from extreme expression of the abnormality or stress peculiar to the light-
salinity-temperature treatment concerned.
Prolapsed oviduct was confined to the dim-light, sea-water, high-temperature
treatment, and appeared in 100% of the survivors. It may be defined as oviposi-
tion into a non-patent, exserted oviduct. The oviduct protrudes from the genital
survivors; c, kyphosis; of less than 100% incidence in the fish of two complementary treatments
both with dim light, one with fresh water and high temperature, the other with sea water and
low (18° C. changed to 20° C., and 20° C. throughout) ; often accompanied by non-buoyancy
and thinness of body ; commoner in the low-temperature treatments.
KS4 ROBERT W. HARRINGTON, JR.
pore as a flaccid, blind sac, filled with expelled eggs that ultimately break down.
The tip of one sac was snipped off, and subsequent ovipositions were successful.
The abnormality finally corrected itself in the other fish, and each in the end pro-
duced viable hatchlings, as in all the other treatments. Prolapsed oviduct occurs
infrequently among K. inannoratiis routinely propagated, and occasionally is fatal.
Pharyngeal hyperplasia was confined to the dim-light, fresh-water, low-tem-
perature treatments, whether begun at 18° C., and continued at 20° C. or at 20° C.
throughout. It occurred in lOO'/f of the survivors of these treatments, male or
hermaphrodite. It shows externally as permanently raised opercula, gaping widely
and exposing basibranchial swellings. One fish head was sectioned and found to
have profuse thyroid tissue, some of it apparently usurping branchial cartilage, so
that the condition may tentatively be diagnosed as thyroid hyperplasia, with the
reservation that no control material was sectioned.
Kyphosis (Rasquin and Rosenbloom, 1954) shows best in roentgenogram or
after clearing and staining with alizarin, but was intense enough here to show up
externally, although mild cases may have eluded recognition. Unlike the other
two abnormalities, it fell short of lOCC/f occurrence and was often associated with
non-buoyancy and thinness of body. Kyphosis occurred in fish of two comple-
mentary treatments both with dim light, one with fresh water and high tempera-
ture, the other with sea water and lowr temperature (18° C. changed to 20° C. and
20° C. throughout). It was more prevalent in the low-temperature treatments
and accompanied several early deaths.
Apart from these plausible causes of death, a large portion of the fish died from
failure to hatch (Table I), and some deaths were probably the delayed results of
weakness caused by abnormally prolonged deferment of hatching.
POST-TREATMENT OBSERVATIONS
Eggs entered treatments August 9— November 15, 1961 ; treatments ended Janu-
ary 9-May 16, 1962. Post-treatment monitoring extended to June 1, 1965, when
the last survivors were fixed for sectioning. Males and hermaphrodites died or
were killed each year of the post-treatment observation period, 1962-1965. Males
remained unchanged except for senile degeneration, but each year some hermaphro-
dites transformed to secondary male gonochorists (see Material and Methods).
All fish were autopsied. Eight primary males, all of the secondary males, and all
but one hermaphrodite were serially sectioned. Testes of secondary males are
larger than those of primary males and apparently yield more spermatozoa.
Nearly 60% of the 37 original hermaphrodites had become secondary males
by the time the last fish were killed at the end of May, 1965. Possibly some that
died earlier and some killed at the end of the post-treatment observation period
might also have changed over had they lived longer. Of the 16 fish dying as her-
maphrodites, eight died natural deaths, three of them egg-bound ; of those killed,
five w^ere egg-bound and probably doomed although otherwise healthy when killed,
while the remaining three were healthy when killed at the end of May, 1965.
The original 37 hermaphrodites, including those later transforming into sec-
ondary males, were kept alive, isolated and under daily observation, for 262-1376
(average, 1041) days post-hatching; the original 36 males, for 327-1314 (average,
945) days. The days from first to last egg laid by hermaphrodites dying as such
LOW-TEMPERATURE-PRODUCED MALE FISH 185
were 46-1117 (average, 694), bv hermaphrodites later changing to secondary males.
107-1167 (average, 711). Secondary males acquire orange pigmentation like that
of primary males, sometimes lose the caudal ocellus, which otherwise becomes obso-
lescent, but retain the basic hermaphrodite color pattern not shared with primary-
males, and are easily distinguished from primary males. The male attributes appear
gradually, making their earliest recognition variable, so that the change from func-
tional hermaphrodite to functional secondary male is best dated retrospectively, as
the day on which the last egg was evacuated. Secondarv males were kept alive
beyond the day on which the last egg was laid (last day as hermaphrodite) for
48-918 (average, 346) days. No secondary male gonochorist reverted to its
former (hermaphrodite) state. The serial sections provided detailed evidence of
the transformation of ovotestes into testes. There was no histological evidence of
change in the opposite direction.
The modalities of this sex inversion, from hermaphrodite to secondary male
gonochorist, will be analyzed more closely in a separate report. Primary sex
determination and differentiation are the concerns of the present report. These
are distinct from the phenomenon of sex inversion, and with one exception this
sex inversion took place after a life as a functional hermaphrodite as long as might
be expected to he the life span in the wild of a fish species of this small size.
EXPERIMENTAL SERIES Two
Although the mortalities incurred in Experimental Series One ( Eig. 1 and
Table I ) posed the formal dilemma of a differential mortality of hermaphrodites at
low temperature and of males at high versus an experimental induction of males by
low temperature and of hermaphrodites by high, nevertheless, the experiments
gave strong presumptive evidence of an induction of males by low temperature.
The dilemma concerned here usually presupposes a sex-determining mechanism
producing about 50% females to 50% males under optimum conditions. This is
clearly lacking or inoperative, because under routine laboratory conditions her-
maphrodites preponderate over males, no females have been encountered at all,
and only hermaphrodites have been encountered in the wild so far. On empirical
grounds the expectation is therefore not males and hermaphrodites (or females)
in equal numbers but rather a preponderance of hermaphrodites, an expectation
not out of line with the mode of reproduction of this species or with the known
genetic uniformity of the experimental fish.
The second series of experiments were undertaken to reconfirm the correlation
of male incidence with low-temperature rearing, resolve the dilemma of a selective
mortality of hermaphrodites versus induction of males at low temperature, and
delimit somewhat, if existent, that segment of ontogeny within which low tempera-
ture can cause a deviation to the male phenotype. These objectives required fewer
treatments with more individuals and lower mortalities than in the first series.
The three treatments of 50 individuals each were essentially three different tem-
perature regimes (Table IV, A-C). All 150 individuals received light of low-
daylight intensity. All were reared in fresh water until large enough (see above)
to be fed brine shrimp, which were introduced in 40% sea water, after which the
fish were gradually changed to 40% sea water, in which they spent the rest of
their lives.
RORKRT w. HARRINGTON, JR.
The individuals of Group A (controls) were reared throughout at a water
temperature of 25 ± 1° C. Those of Group B were under the same conditions
through hatching, but the hatchlings were reared for their first five months at 19.5
± 0.5° C., after which they were maintained at 25 ± 1° C. Those of Group C were
reared at 25 ± 1° C. until the eggs were at least in Stage 16 hut not beyond Stage
22a (cf. Harrington, 1963) ; then they were reared through eclosion (hatching or
cutting out of the chorion ) and for five months post-eclosion at 19.5 ± 0.5° C..
after which they were maintained at 25 ± 1° C. At the outset of the 19.5 ± 0.5° C.
interval of their treatment, 28 of the Group-C individuals that survived to maturity
were in Stage 16 (optic vesicles first visible as expansions of the forebrain), 8 in
Stage 17 (optocoeles), 3 in Stage 19 (optic cup, lens, and neurocoele), 3 in Stage
20a (optic lobes, neuromeres, pectoral fin-hud anlagen), 3 in Stage 201 ) (heart
pulses without blood circulation), and 1 in Stage 22a (circulation starting through
dorsal aorta and vitelline vessels).
RESULTS OF EXPERIMENTAL SERIES Two
The three objectives of this experimental series were achieved with decisive
results (Table IV). The zero mortality and absence of males at 25 ± 1° C. (Group
A) confirms the correlation of hermaphrodite incidence with high-temperature rear-
ing while showing that the temperature need not be as high as 30° C., as in the first
experimental series. This accords with the production of over 95% hermaphrodites
to under 5% males when temperatures fluctuated more widely about 25° C., during
routine rearing, suggesting that the low percentages of males hatched prior to these
experiments resulted from temperatures temporarily below a threshold within 19-
24° C. while these fish were traversing a critical segment of organogenesis thermo-
labile with regard to sex determination and differentiation. The correlation of male
incidence with low-temperature rearing was established unequivocally (Group C).
The dilemma of a selective mortality of hermaphrodites z'crsns induction of males
at low temperature and rice versa at high, is resolved, because the mortalities in-
curred in Groups A-C were zero, and only 8% and 10%, respectively. There can
be no doubt that low-temperature rearing caused a deviation to the male phenotype
during a critical phase of ontogeny.
The possible extent of this thermolabile phenocritical period of sex determination
and differentiation has been contracted at either end. Group B, with 92% survival
and all hermaphrodites, not only reinforces the results of Group A, but shows that
the thermolabile interval concerned does not extend beyond the end of Stage 31, the
last stage before hatching. Group C showed that cold treatment need not begin
before Stage 22a (blood circulation just established) to be effective in producing
males. This phenocritical period may span a much shorter segment of ontogeny
than its above-determined possible maximum extent. Only further experiments of
similar design can define it more closely.
DISCUSSION
Uniqueness of the present experiments and results
Of previous experiments investigating possible influences of extrinsic factors on
sex determination in fishes, two concerned that rather forced example of environ-
LOW-TEMPERATURE-PRODUCED MALE FISH 187
mental influence, overripeness of eggs. In rainbow trout with a delayed fertilization
of 21 clays, Mrsic (1923) reported 55% males, 33% females, and \2% interpreted
by him as having ovaries transforming into testes, an interpretation put in doubt by
later studies (see below, and Atz, 1964). With moderately delayed fertilizations,
he reported small excesses of females. The controls deviated little from the 1 : 1 sex
ratio. Mortalities were more than enough, however, to create the dilemma stated in
the introduction to this report: 88% with the 21-day delay, 60-70% with moderate
delays, 54% of the controls. Mrsic dismissed the dilemma by discounting the
mortality as having occurred too early in ontogeny to be the deciding factor, in an
argument based on an histological interpretation (see above) of the course of gonad
differentiation negated by later studies on rainbow trout (Padoa, 1939) and other
salmonids (Ashby, 1952; Robertson, 1953). In brown trout from late-fertilized
eggs Huxley ( 1923 ) found no significant departure from the 1 : 1 ratio.
Among broods of Siamese Fighting Fish, Eberhardt (1943) found the sex
ratios extremely variable, but under optimum conditions approximating the 1:1
ratio. By crowding during rearing he obtained statistically significant excesses of
males, and concluded that poor ("schlechte") space, food, and water conditions
favored differentiation in the male direction, i.e., opposite the genetic constitution.
Nevertheless, a selective mortality of females cannot be ruled out, because he did
not record the mortalities incurred in the experiments. Eberhardt rejected a selective
elimination of females not on the basis of the experiments themselves but by in-
ference from the results of rearing 25 other broods on scant food, so as to exaggerate
the usual high mortality of the first two weeks of life. Only -1 — \7 fish survived from
these broods of 100—400 hatchlings, and were well fed after the first two weeks.
The survivors of 12 broods were 41-50% females; the other broods had somewhat
lower percentages of females, but Eberhardt omitted details. The percentages of
females in these underfed broods do not adequately support his contention that be-
cause in the experiments deaths did not exceed 1 % after the first two weeks a selec-
tive mortality of females is ruled out. Furthermore, underfeeding and crowding
cannot a priori be equated with regard to selective mortality, nor can either a priori
be assumed without influence on sex determination.
A literature too large for more than summary treatment, reviewed in part by
Dodd (1960) and Atz (1964), but much of it obsolete, concerns environmental
influence on sex determination in the European eel. Grassi (1919) considered
temperature, salinity, and nutrition to exert such influence. Forty years later, his
student D'Ancona (1960; p. 67) was able to assert merely that his "own experi-
ments suggest the possibility of a phenotypic sex deviation under the influence of
experimental factors." Counter to an earlier report (D'Ancona, 1950) that sex
ratios ascribed to environmental influences were attributable to a differential migra-
tion of the sexes, he named crowding and high temperature as "favoring differentia-
tion toward the male sex," but the evidence is inconclusive and has since been put
further in doubt ( Sinha and Jones, 1966). It is unfortunate that a species so ill-
suited for settling the question of environmental influence on sex determination in
fishes became so closely linked with the question historically.
Each Rirnlus mannuratiis individual in the present experiments was reared ab
ot'o in its own container to preclude results attributable to crowding, which would
be indecisive as to the proximate extrinsic causal factor. A freemartin-like effect
188 ROBERT W. HARRINGTON, JR.
cannot be dismissed as a possible result of crowding, for not only can sex steroids
administered per us within a brief period of early ontogeny produce pbenotypes of
either sex in opposition to the genotypic sex (Yamamoto, 1953-1961), but there is
evidence (Egami, lc>54) for the uptake by fish in close confinement of estrogenic
substances released by other fish (also cf. comments of Lindsey, 1962; p. 304).
Among the fish crowded by Eberhardt ( 1943 ), for instance, rates of growth and sex
differentiation varied so much that he resorted to removing the faster-growing ones
when these could be sexed externally, each time reconcentrating the sexually in-
distinguishable ones remaining.
The induction of male gonochorists by incubating the eggs of R. nianiioratns at
low temperature detracts from the proposition that in eels high temperature favors
xx differentiation in the male direction, by demonstrating an environmental effect
on at least one species of fish the opposite of that on amphibians, in which it is well
established (Witschi, 1929, 1957; Piquet, 1930; Uchida, 1937) that high temper-
ature is male-inducing. R. iiiarinoratns was chosen for its hermaphroditism and
rare male incidence as possibly having a less homeostatic sex-determining mecha-
nism than gonochoristic fishes and thus being a more likely species for testing for
environmental influence on sex determination, but the results obtained with R.
iintniioratits raise expectations of analogous results with gonochoristic species of
fish. Observations in harmony with our results, but directed to other ends and not
excluding a selective mortality of the opposite sex, have in fact been made on two
gonochoristic fishes of promise for such experimental testing. In exploring ways of
rearing the cyprinodontid fish, Epiplatys chaperi, Van Doom (1962) obtained a
higher percentage of males at low temperatures. In experiments on meristic vari-
ation, Lindsey (1962) found that rearing conditions of high temperature and
crowding produced higher percentages of female sticklebacks, Gasterostens aciileatns.
The activity period of the sex-chromosome genes governing sex determination
is equated by Atz (1964; p. 215) with the period of ontogeny in which it is possible
with heterotypic hormones to reverse the sex of a gonochoristic fish, e.g., Oryzias
latipes (Yamamoto, /or. cit.). Atz remarks that at present it is problematic
whether a similar limited period could be ascribed to hermaphroditic species. Our
results bear indirectly on this question. Sex reversal in the sense of transformation
from one sexual phenotype (primary gonochorist) to the alternate one (secondary
gonochorist) is not concerned here, so that in either hermaphrodite or gonochorist
the interval by definition would not extend later than through the sexually indiffer-
ent and primary sex-differentiation stages. Both O. latipes and R. inannoratiis
hatch sexually undifferentiated and start eating at once as do other cyprinodontids.
Making use of these traits by feeding sex steroids to 0. latipes from the day after
hatching, Yamamoto caused sex reversals in the sense of producing primary
gonochorists of either sex contrary to genotypic sex. In the ontogeny of R. inar-
inoratits, the interval within which low-temperature rearing produced males in
opposition to the presumed hermaphrodite genotype (see above) begins after onset
of blood circulation (possibly long after) but ends before hatching. It remains to
be determined whether the thermolabile phenocritical period of sex determination in
R. niartnoratns is paralleled, overlapped, or succeeded by a hormonal lability in this
respect. In the same context, although the post-hatching hormonal lability of 0.
latipes is not paralleled by a post-hatching thermal lability in R. mannoratus, the
LOW-TEMPERATURE-PRODUCED MALE FISH 189
peroral administration of sex steroids by Yamatnoto in excluding pre-hatehing ef-
fects, leaves unknown whether the hormonal lability of O. latipes begins soon enough
to parallel or overlap the pre-hatching thermal lability of R. inannoratiis. In any
case, the much shorter thermolabile interval of sex determination in R. marmoratus
can be identified with the activity period of the sex-determining genes with as much
reason as the interval of 8-10 weeks post-hatching during which Yamamoto fed
0. latipes the steroids that caused sex reversals. This opens the possibility that
teleostean sex determination entails a two-stage sex differentiation, the first stage
with thermal lability, the second with hormonal lability.
Exclusion of alcrnative explanations
The male-inducing effect on R. niaruioratus of incubation at low temperature
emerges as a thermal effect apart from and undisturbed by the structural-functional
abnormalities (prolapsed oviduct, pharyngeal hyperplasia, kyphosis) produced by
certain (vide supra] specific light-temperature-salinity combinations of Experi-
mental Series One. The same effect was achieved without these abnormalities,
moreover, in Experimental Series Two, which avoided extremes of light intensity,
temperature, and salinity, except for low temperature. These abnormalities were
confined to dim-light treatments, and are attributable in part at least to hormonal
derangements, which further indicates the independence and priority of the thermal
effect on sex determination in R. niaruioratus. Prolapsed oviduct results pre-
sumably from either precocious ovulation or abnormal persistence of non-patent
oviduct, and in the European Minnow, Phoxinus phoxinus, for example, the oviduct
becomes patent only within the spawning season, under endocrine control (Bui-
lough, 1939). The pharyngeal hyperplasia and kyphosis in R. niaruioratus kept at
light intensities mostly within the range of civil twilight are reminiscent of the
thyroid hyperplasia and kyphosis in the characin, Astyana.r nie.ricaniis. kept in total
darkness, and ascribed to hormonal imbalance normally inhibited by light and
involving but not confined to the pituitary-thyroid complex (Rasquin and Rosen-
bloom, 1954). The dim-light treatments of R. niaruioratus began right after ovi-
position, at embryonic stages (Table III) not later than Stage 13c (f blastoderm),
but eggs of A. mexicanus spawned in the light failed to develop in the dark.
Rasquin and Rosenbloom placed in darkness specimens kept in the light their first
two months of life. Other causes of spinal curvature (Comfort, 1960, 1961) may
also have been involved, because not all A. nie.vicanus kept in darkness showed
kyphosis, and kyphosis was confined to R. niarnioratus of only two dim-light treat-
ments, in each of which it fell short of lOO'/f occurrence.
Despite the evident primacy of thermal influence on sex determination and
differentiation in R. niaruioratus. it would constitute the fallacy of misplaced con-
creteness to conclude that males were produced by low temperature to the complete
exclusion of influences from other extrinsic factors. The principle of complemen-
tarity as extended to biological phenomena (Meyer-Abich. 1956) is especially
relevant to environmental influences on the ontogenetic differentiation of aquatic
poikilo therms. To identify such influences requires polyfactorial analysis, with com-
binations of factors controlled as in the present experiments and in such as those of
Kinne and Kinne (1962), who observe that not onlv can one environmental factor
190
ROBERT W. HARRINGTON, JR.
modify the physiological effect of another, but a single factor reaching sufficient
intensity to modify the process under study may alter other environmental factors.
The obvious uncontrolled, dependent, extrinsic factor in these experiments on R.
nianiiorattts is dissolved oxygen, each egg having been incubated in its own jar of
stagnant water. Kinne and Kinne found stagnant (non-aerated ) water to have
70 ± 10% the concentration of dissolved oxygen in aerated (100% air-saturated)
water. From their nomograph (Kinne and Kinne, 1962, Fig. 2) can lie obtained
TABLE V
Extraparetital incubation periods of Rivulus ma.rmora.tus with various combinations of light intensity,
salinity, and temperature. B, bright light; D, dim light; S, sea water; F, fresh water.
Same eggs as in Tables I-III ; hatched unaided (starred), light-triggered hatching
(unencumbered numerals), cut from chorion (parentheses), started at 18° C.
and changed to 20° C. (italicized numerals), the rest at 20° C. throughout
Numbers of hatchlings
30° C.
18-20° C.
Num-
i r
days
Hermaphrodites
Male
Hermaphrodites
Males
BS
DS
BF
DF
BF
BS
DS
BF
DF
BS
DS
BF
DF
12
1*
14
3*
1*
2
2
15-16
1*
4*
2
18-19
1*
1
1
23-24
2
26-27
3
1
28
1*
30-31
1
1
1
35-36
2 + 1*
(D
(D
1
(1)1
37-38
1
(1)
(2)1
39-40
1
(1)1
1
2
1
(Dl
41-42
(! + /)
(D
1
1 + 1*
43-44
(D
;
45-46
(D
(D
(3)
2
2
51-56
(1)
(4)
14.5
15.0
21.0
19.4
27.0
39.0
38.5
35.3
33.5
40.7
41.0
42.1
44.0
Mean extraparental incubation periods in days.
the approximate 100% air saturation (ml. (X/L.) for each temperature-salinity
combination of our experiments, except those with temperature changed from 18°
to 20° C. Although the actual concentrations were less because the water was
stagnant, the 100% air saturation values permit an arrangement of the experimental
data in order of increasing oxygen concentration (Table VI). In contrast to the
decisive thermal influence on sex determination in R. uiannoratus, not only do
these oxygen values fail to uncover evidence of an effect ascribable to oxygen con-
centration, but in a pilot experiment 10 eggs incubated at 25° C, in 100% air-
saturated fresh water yielded 10 hermaphrodites.
LOW-TEMPERATURE-PRODUCED MALE FISH
191
Significance of tlic temperature efject per se
The production of male R. mannoratits by low-temperature incubation allows
its examination in relation to corresponding rates of embryonic development. The
developmental rates of Cyprinodon macularins exposed to a diversity of temper-
ature-salinity-oxygen combinations were measured by Rhine and Rhine (1962) as
numbers of clays from fertilization to certain embryonic stages, especially hatching.
These rates increased with increasing oxygen content, and decreased with increasing
salinity, the latter effect mediated by changing coefficients of oxygen absorption and
saturation in water. Both the retardation and the acceleration were increasingly
accentuated by increase in temperature. For comparisons of developmental rates
among R. inaniioratiis eggs exposed to different extrinsic factor combinations we
must rely on the incomplete data (Table V) of Experimental Series One, because
in Experimental Series Two exposure to low temperature was begun at a later and
wider range of embryonic stages and the low-temperature embryos were cut from
TABLE VI
The results of Experimental Series One and Two arranged in order of increasing oxygen concentration
at 100% air saturation. The actual concentrations were less, because the
water was stagnant. See Discussion and Tables I and IV
Approximate temperature
30° C.
20° C.
30° C.
25° C.
20° C.
20° C.
nil. O2/L. (100% air sat.)
4.3
5.3
5.6
6.0
6.6
6.6
Experimental Series
One
One
One
Two
One
Two
Total eggs
17
31
29
50
26
50
Percentage survival
64.7
67.7
51.7
100.0
53.8
90.0
Percentage male
0.0
48.0
3.4
0.0
30.8
72.0
Percentage hermaphrodite
64.7
19.7
48.3
100.0
23.0
18.0
their chorions far in advance of the normal hatching stage (see above and Table IV,
C). In Experimental Series One some eggs hatched unaided, others, with artificial
stimulation ; the embryos of the rest were cut out but only after some of the same
age had hatched, unaided or aided (Table V). Intraparental (pre-treatment) in-
cubations ranged at most from one to 24 hours (Table III and Harrington, 1963,
Table I), treatment starting right after oviposition. The incubation periods of
Table V might have diverged somewhat more had not many of them been ended
arbitrarily, but most of the embryos cut out or from eggs stimulated to hatch would
otherwise have perished unidentifiable as to ultimate sex type, as attested by
mortalities (Table I) ascribable to extraparental incubations protracted by failure
of the hatching mechanism.
The data of Table V, however imperfect, suffice to show a more delayed hatching
at high temperature in stagnant fresh water than in stagnant sea water, which is
paradoxical with reference to the eggs of Cyprinodon macularius ( see above and
Rhine and Rhine, 1962, Table X). Even if this perhaps resulted from impaired re-
sponsiveness of the hatching mechanism, it is no less interesting to find that the
single, anomalous male produced at high temperature in our experiments had an
extraparental incubation of 27 days in contrast to an average of 17 for the hermaph-
192 ROBERT W. HARRINGTON, JR.
rodites otherwise produced at 30° C. The arbitrary curtailment of the incubations
of many of the eggs at low temperature permits only the general comment that there
are indications of a possible tendency toward longer incubations among the eggs later
found to have yielded males and that the incubations at 20° C. were abnormally long
for R. iiiannorotiis. At the latitude of the wild-caught founder stock, which is at or
near the northernmost extent of the geographic range of this chiefly tropical species,
air-temperature daily minima between mid-April and mid-October form a plateau at
20° C., the daily means and maxima being much higher, of course. Most of the
potential extent of the as yet undefined natural spawning season of R. mannoratns
is excluded thereby from temperatures of sufficient duration low enough to produce
males. Sooner or later, however, some males may be expected to be found in the
wild at this latitude hatched from eggs incubated at lower temperatures toward the
extremities of the spawning season or perhaps subjected to less obvious, alternative
male-inducing conditions like the anomalous, lone male obtained at 30° C. (Tables
I-III, V and VI).
The complex effects of temperature per se on morphological differentiation and
the consequent impossibility of exactly equating developmental stages between
embryos incubated at contrasting temperatures hardly needs stating. The impre-
cision of the classical embryonic "stage" was illustrated by Hayes (1949) with the
comment added that hatching itself is not to be regarded as a stage, because it can
occur so variably. Nevertheless, with cautionary reservations and for want of any-
thing better, use must still lie made of such "stages," sometimes even hatching, as
was done by Kinne and Kinne (1962). Hatching as a stage is of normative
importance here only in that experiment (Table IV, B) of Experimental Series
Two in which low temperature treatment began with hatchlings from eggs incubated
at our standard laboratory temperature (25 ±1° C.). The mean extraparental
incubation of the eggs of Table IV, B was 15.3 ± 3.7 days, the total incubation
(extra- plus approximate intraparental) was 17.2 ± 3.7 days, and the feeding of
each hatchling for one full day at 25 ± 1° C. gave a mean of 19.4 ± 3.8 days before
transfer to low temperature. Accumulated laboratory records for 190 other eggs
incubated at 25 ± 1° C. yielded a mean extrapareutal incubation of 17.3 ± 4.5 days
and an approximate total incubation of 18.6 ± 4.4 days.
The very phenomenon under consideration, i'h., the production of males by
extraparental incubation at low temperature, may itself be the result of an un-
coupling of embryonic processes (cf. Hayes, 1949) by differential effects of low
temperature on two or more constituent rates of development, so as to change the
order of morphological events critical for sex differentiation. A paradigm for such
an effect is the delay by low temperature of medullary development in amphibian
gonads that feminizes males, at least temporarily (Uchida, 1937; Witschi, 1957).
Although medulla as a topographic term has been declared inapplicable to teleostean
gonads (D'Ancona, 1952), the bipotential gonocytes are sexualized as ovogonia and
spermatogonia, respectively, in heterologous somatic territories within the ovotestes
of several hermaphroditic fishes. Nor has uncertainty over the embryogenesis of
the heterologous tissues deterred postulations of inductor substances in fishes
analogous to the corticin and medullarin of Witschi, viz., gynogenine and audro-
genlne by D'Ancona (1949) , gynotennoiic and androtermone by Yamamoto (1962).
LOW-TEMPERATURE-PRODUCED MALE FISH 193
Implications for the interpretation of interse.vnality in fislics
Past studies of hermaphroditic fishes have been based at best on histological
sections of gonads from economically feasible numbers of fish, sampling as wide a
size range as collections provided. In most cases size was the sole criterion of rela-
tive age, an unreliable one for fishes, because there may be differential growth rates
and mortalities between the sexes, including determinate versus indeterminate
growth. Interpretations of otherwise adequate samples have been rendered incon-
clusive or incomplete by uncertainty over the relative ages of the fish coupled with
the fact that the effects on growth rate of sex inversion and reversal are unknown.
These difficulties are avoided with R. inannoratus. which is the first hermaphroditic
fish species to have been kept in the laboratory throughout life. The results of the
present experiments in conjunction with the daily observation of the fish of Experi-
mental Series One throughout their lives throw7 light on aspects of fish inter-
sexuality hitherto obscure, because the age and history of each fish were known
exactly.
Before applying the results of the present study to these aspects of fish inter-
sexuality, it is pertinent to reassess the extent to which the life span of R. mannora-
tiis was encompassed by the experimental and post-experimental observations of
Experimental Series One. Several tokens of senility (Comfort, 1960, 1961 ; Wai-
ford and Liu, 1965) appeared among these fish, e.g.. clouded cornea, emaciation,
exophthalmos, humped back, raising of scales, renal concretions. They were kept
alive as long as 1,376 days post-hatching; most of those killed were already in
extremis. The life span of R. inannoratits seems to be of the order of that of an-
other cyprinodont, the poeciliicl Lebistes retieulatits. Under laboratory conditions,
Lebistes has a limiting age of 2.000 days, 50% of age-dependent deaths occurring
by the end of 800-900 days (Comfort, 1961). Survival both of Rirnliis and of
Lebistes to the more advanced ages reached in the laboratory (see Post-treatment
Observations) is probably negligible in the wild.
Riz'iiliis inannoratus is the only fish species known to exemplify the ultimate
mode of synchronous hermaphroditism, normal self-fertilization. The only other
synchronously hermaphroditic fishes known are the serranids, Serramts cabrilla. S.
hepatus, S. scriba and S. snbligariits, none of which are claimed to be naturally self-
fertilizing (Clark, 1959; Reinboth, 1962; Atz, 1964). The life cycle of S.'snbli-
garius is incompletely known. The other three species mature and function first as
males, the next year first functioning as synchronous hermaphrodites, but it is not
known whether all start out as males or some start as hermaphrodites their first
year (Reinboth, 1962). On the other hand, R. inannoratiis is self-fertilizing from
outset of functional maturity, even when this is precocious, at higher than usual
temperatures. In a sense, therefore, neither these serranids nor R. inannoratus are
obligate synchronous hermaphrodites throughout life, some or all of the serranid
individuals serving as males their first year and some individuals of R. umrmoratits
ending their lives as functional secondary male gonochorists, at least under life-
prolonging laboratory conditions. The distinction between synchronous and suc-
cessional hermaphroditism in fishes thus seems to be one of degree. Of the two
forms of successional hermaphroditism, protogyny and protandry, protogyny is the
one toward which the synchronous hermaphroditism of R. inannoratus leans in
tending toward transformation in the male direction, but protogyny is characterized
194 ROBERT W. HARRINGTON, JR.
by temporal succession to the functional male state from the functional female state
and not from a fully functional synchronous hermaphroditism, primary and of long
duration, as in R. iinirinoratits.
The induction of primary male gonochorists by low-temperature incubation,
added to the spontaneous inversion of older hermaphrodites into secondary male
gonochorists, are attributes of R. inannoratiis recalling that primary and secondary
males occur also among wrasses (Labridae), e.g., Coris julis and Thallasoma bi-
fasciatuin (Reinboth, 1962). Although secondary male wrasses originate by sex
reversal from a primary female condition instead of by sex inversion from synchro-
nous hermaphrodites, perhaps all or some primary male wrasses arise from incuba-
tion at low temperature like the primary males of R. niannoratus, the no more likely
alternative being a homogamety-heterogamety (e.g.. XX-XY ) switch mechanism
yielding protogynous hermaphrodites versus primary males, respectively, or vice
versa. Wrasses have pelagic eggs of short incubation and spawning seasons
(Breder and Rosen, 1966) such as to encourage a search for differential latitudinal
occurrences of primary males correlated with differential thermal exposures of the
drifting eggs. Control of male coloration, however, seems to differ between Rivulus
and these wrasses. The inference of Zahl (1934), that the caudal ocellus of Rivitlus
urophthalmus is a sex-limited trait suppressed by testicular secretion is strengthened
by its presence in all immature R. iiianiioratus also, and even more by its persistence
in the hermaphrodites ( Figs. 2 and 3 ) , progressive extinction if these transform
into secondary males, and usually complete disappearance in primary males (Figs.
2 and 3 ) at sexual maturity. The striking dichromatism between primary and
secondary male wrasses seems to have a more complex hormonal control (Reinboth,
1962; comments of Atz, 1964).
Although Rivulus niannoratus exhibits a prolonged synchronous hermaphro-
ditism, its sexuality is more like that of wrasses, e.g., Coris julis, than of any other
hermaphroditic fishes, with either synchronous or successional hermaphroditism.
As far as they have been described, hermaphroditic fishes other than the wrasses
and R. inannoratiis have structurally bisexual gonads alone, although one or the
other of the heterologous gonadal territories or dispersed centers may be in a state
of persistent abortiveness (rudimentary hermaphroditism), prefunctional latency,
or postfunctional involution. The wrasses and R. nmrmoratus also have gonads of
bisexual structure. In the wrasses, these function first as ovaries, then, after a short
transitional period, as testes, their possessors changing from functional females to
functional secondary males. In R. inannoratiis, they function from the first as
ovotestes (Fig. 1, c and d), then after involution of the ovarian component and
further evolution of the testicular, as testes alone, their possessors changing from
self-fertilizing hermaphrodites for the greater part of their lives to functional sec-
ondary male gonochorists for the remainder. Unlike all other hermaphroditic fishes,
however, these wrasses and R. inannoratiis include, besides individuals with bi-
sexual gonads, a minority of primary male gonochorists, with testes of unisexual
structure (Fig. 2, e), in contrast to the bisexual structure of the testes of secondary
male gonochorists.
The testes of primary males of R. niannoratus, and of Coris julis (Reinboth,
1962), are like those of the true gonochoristic species making up the majority of
fishes, and can easily be told apart from testes of secondary male gonochorists, which
LOW-TEMPERATURE-PRODUCED MALE FISH 195
retain the oviduct functional during the female or hermaphroditic phase of the
bisexual gonad, and often ovarian residua hesides. Testes of true gonochoristic
fishes and of the primary males of R. mannoratus and Corls jnlis in being of uni-
sexual structure, without oviducal or ovarian vestiges, differ also from the testes of
protandrous fishes and of nominal gonochoristic fishes functionally gonochoristic
throughout life but with incipient bisexual structure. Testes of the last two cate-
gories are so similar that Reinboth ( 1962, Fig. 27 ; reprinted as Fig. 2 of Atz, 1964)
refers both of them to the same cross-section diagram (Fig. 27, c), which stands in
marked contrast to that for the testes of primary males of Coris jitlis (Fig. 27, a).
\Yhat is labeled oviduct in Reinboth's diagrams is of dual origin, the anterior paired
lumina derived from the entovarial sulci and nonhomologous with the intratesticular
sperm ducts according to Eggert (1933), the unpaired posterior duct arising
differently.
In the case of the wrasses, the alternative of primary male versus protogynous
hermaphrodite might possibly be decided by a genetic shift mechanism instead of by
low-temperature incubation, but this is not the case with R. mannoratus. In the
critical and decisive Experimental Series Two, not only were all individuals of the
same clone but they were the end products of selfing through 9-11 uniparental
laboratory generations. Since then, further evidence (unpublished) for the homo-
zygosity of these fish has been accumulating from hybridization-nnn-grafting experi-
ments. All of these lines of evidence of the overwhelming prevalence of hermaph-
rodites point to the likelihood that the fish of this clone are not only homozygous,
but, mferentially, homogametic as well, both hermaphrodites and experimentally
produced primary male gonochorists. The fact that both natural and experimentally
induced departures from the modal phenotypic expression of the hermaphrodite
genotype of R. mannoratus are in the male direction, makes it intriguing to find
that among amphibians steroid hormones seem capable of causing sex reversal only
in the homogametic sex, i.e., to the phenotype of the heterogametic sex (Witschi,
1957). In the fish, Orvzias Infixes, however, fully functional sex reversals in either
direction are produced with sex steroids ( Yamamoto, 1953-1961 ). In R. mannora-
tus, the hermaphroditic constitution is manifestly the epistatic one and the male
constitution, the hypostatic. While males result from low-temperature incubation
or from sex inversion late in life, females are non-existent. In pilot experiments
in which eggs were incubated at a descending series of constant temperatures, 31.2°
C. was the highest at which eggs survived, the 10 eggs concerned yielding 10
hermaphrodites.
Brecler and Rosen (1966) construe the hermaphroditism of R. mannoratus as
a mechanism evolved to compensate for the vicissitudes of the unstable coastal main-
land and oceanic island environments, allowing reproduction in spite of severe popu-
lation depletions by violent coastal storms. By the same tokens, if R. mannoratus
proves to be hermaphroditic throughout its range, which arches island-to-island
across the Caribbean from off the Venezuelan coast onto the Florida peninsula, it
is easy to see how island-hopping hermaphrodite castaways might found new colo-
nies more readily than males and females of gonochoristic species, with little chance
of meeting. Haldane (1957) argues similarly for the colonization of new rivers by
flood-transported hermaphroditic or parthenogenetic fish, while noting that if the
selfed immigrant ancestor were homogametic the progeny would be more likely to
196 ROBERT W. HARRINGTON, JR.
die out, because of the absence of males. Whatever the long-range prospects of
survival for R. inaniiorafiis, the graft tests for histocompatibility between wild-
caught progenitors and their earliest laboratofy descendants make it inescapable
that selling had gone on in the wild for some time. In the long run, perhaps enough
males (primary or secondary gonochorists ) are produced in the wild to contribute
genetic information occasionally from their clone to eggs of another clone, by
mating with a hermaphrodite with ovotestes out of phase so as to emit unfertilized
eggs. Hermaphrodites do occasionally emit a spate of unfertilized eggs (Harring-
ton, 1963) and we have had apparent success in fertilizing a few of these with sperm
from a male of different clone. Attempts to mate males with hermaphrodites have
resulted in sexual coaction, but the hermaphrodites, whether previously isolated or
not, have emitted eggs in stages of development so advanced as to indicate self-
fertilization long before the pairing. In sum, amphimixis between eggs from
hermaphrodites of one clone and sperm from primary or secondary male gonocho-
rists of a different clone is in the realm of possibility, but the evidence so far is that
it has occurred rarely if at all in the local populations from whence our stock was
derived.
The writer thanks E. S. Harrington and L. A. Webber for their faithful tech-
nical assistance, Drs. C. M. Breder, Jr., K. D. Kallman, and M. W. Provost for
reading the manuscript, Drs. J. B. Leonard and J. Packer for microphotographs,
Mr. J. O'Neal for macrophotographs, and Mr. W. Janse for executing Figure 1.
This is Contribution No. 167 of the Entomological Research Center, Florida State
Board of Health, supported by PHS grant #ES 00161 from the Office of Resource
Development, BSS (EH), and by PHS grant #CC 00274 from the Communicable
Disease Center, Atlanta, Georgia.
SUMMARY
1. Rii'iihts mannoratns is the only known hermaphroditic fish species naturally
self-fertilizing. Tissue grafts between wild-caught fish and their uniparental
laboratory descendants give the antoyraft reaction, indicating propagation by selfing
in the wild also. Only hermaphrodites have been found in the wild locally, although
selfing through more than 10 uniparental laboratory generations yielded a few pri-
mary male gonochorists, under 5% in contrast to over 95% that were hermaphro-
dites. Females seem to be non-existent.
2. Two series of experiments were undertaken to identify a possible environ-
mental factor able to cause a deviation to the male phenotype during sex differenti-
ation, on the working hypothesis that low male incidence in clones composed other-
wise of hermaphrodites indicated a lability in the sex-determining mechanism
through which the genotype normally produces the hermaphrodite phenotype.
3. Individuals of two clones, each in its own jar throughout life, were exposed
to the eight combinations of bright or dim light, sea water or fresh water, high or
low temperature (Experimental Series One). Exposure was from not later than
the 3 blastoderm stage until sexual maturity at high temperature or five months
post-hatching at low.
4. Over seven times the number of males previously encountered were obtained,
all but one from low-temperature treatments. Male production was correlated with
LOW-TEMPERATURE-PRODUCED MALE FISH 197
low-temperature rearing despite alternative light intensities and salinities and
structural-functional abnormalities (prolapsed oviduct, pharymjcal hyperplasia, ky-
phosis) peculiar to different dim-light, salinity-temperature combinations, and partly
attributable to hormonal derangements. Mortalities were high enough to present
the formal dilemma of a differential male induction versus hermaphrodite mortality
at low temperature and vice versa at high, but this dilemma was resolved by Ex-
perimental Series Two.
5. The Experimental Series One fish were monitored daily up to 1,376 days
post-hatching, by which time almost 60% of the hermaphrodites had changed to
functional secondary male gonochorists, the rest dying or killed as hermaphrodites,
some each year. Primary males remain unchanged except for senile degeneration.
Secondary 'males arise mostly late in laboratory-prolonged life, by involution
of the ovarian component of the ovotestes with further evolution of the testicular
component, the caudal ocellus fading or vanishing as they become orange like the
primary males.
6. In Experimental Series Two, mortalities were low and the structural-
functional abnormalities were absent. All individuals were kept at the same inter-
mediate salinity and light intensity : Group A, at moderate temperature throughout
to maturity; Group B, at the same temperature through hatching, at low temper-
ature the first five months post-hatching, thereafter at the moderate temperature;
Group C, at the moderate temperature up to stages from optic vesicle formation to
outset of blood circulation, then at low temperature through eclosion and for five
months post-eclosion. Group-C embryos being cut from their chorions to minimize
deaths from hatching failure.
7. The Group-A eggs yielded 100% hermaphrodites, the Group-B eggs, 92%
hermaphrodites and 8% deaths, the Group-C eggs, 72% males, 18% hermaphro-
dites, and 10% deaths. Exposure to low temperature from as late as outset of
blood circulation produced males.
8. The uniqueness of the present experiments and results, exclusion of alterna-
tive explanations, significance of the temperature effect per se, and the implications
of these findings for the interpretation of intersexuality in fishes are discussed at
length.
LITERATURE CITED
ASHBY, K. R., 1952. Sviluppo del sistema riproduttivo di Salmo trutta L., in condizioni normali
e sotto 1'influenza di ormoni steroidi. Rir. Biol., n. s. 44: 3-18.
ATZ, J. W., 1964. Intersexuality in fishes. In : Intersexuality in Vertebrates Including Man
(ed. by C. N. Armstrong and A. J. Marshall), Academic Press, London and New York.
BREDER, C. M., JR., AND D. E. ROSEN, 1966. Modes of Reproduction in Fishes. The American
Museum of Natural History, Natural History Press, Garden City, New York.
BULLOUGH, W. S., 1939. A study of the reproductive cycle of the minnow in relation to the
environment. Proc. Zoo/. Soc. London, Ser. A, 109: 79-102.
CLARK, E., 1959. Functional hermaphroditism and self-fertilization in a serranid fish. Science,
129: (3343) : 215-216.
COMFORT, A., 1960. The effect of age on growth resumption in fish (Lebistcs) checked by food
restriction. Gcrontologia, 4: 177-186.
COMFORT, A., 1961. The longevity and mortality of a fish (Lebistcs rcticnlatus Peters) in
captivity. Gcrontologia, 5: 209-222.
D'ANCONA, U., 1949. Ermafroditismo ed intersessualita nei Teleostei. Expcricntia, 5: 381-389.
198 ROBERT W. HARRINGTON, JR.
D'ANCONA, U., 1950. Determination et differenciation du sexe chez les poissons. Arch. Anat.
Micros, et Morph. E.r/>., 39: 274-292.
D'ANCONA, U., 1952. Territorial sexualization in the gonads of teleosteans. Anat. Rcc., 114:
666-667.
D'ANCONA, U., 1960. The life-history of the Atlantic eel. In: Hormones in Fish ( eel. by I.
Chester Jones), Symposia Zool. Soc. London, No. 1.
DODD, J. M., 1960. Genetic and environmental aspects of sex determination in cold-blooded
vertebrates. Mem. Soc. Rndocrinol., No. 7: 17-44.
EBERHARDT, K., 1943. Geschlechtsbestimmung nnd -differenzierung bei Betta splendcns Regan.
Zeitschr. jilr indukt. Abstuinni. nnd Vererbungslchrc, 81: 363-373.
EGAMI, N., 1954. Influence of temperature on the appearance of male characters in females of
the fish, Orvzias latipcs, following treatment with methlydihydrosterone. /. Fac. Sci.,
Univ. Tokyo, Sec. IV, 7 (2): 281-298.
EGGERT, B., 1933. Die Intersexualitat bei Knochenfischen. Zeitschr. iciss. Zool.. 144: 402-420.
CARMAN, S., 1895. The cyprinodonts. Mem. Mus. Comp. Zool., 19 ( 1 ) : 1-179.
GORDON, M., 1957. Physiological genetics of fishes. In : The Physiology of Fishes ( ed. by
M. E. Brown), Vol. 2, Academic Press, New York.
GRASSI, B., 1919. Nuove ricerche sulla storia naturale dell'Anguilla. Mem. R. Com. Talass.
Ital.,67: 1-141.
HALDANE, J. B. S., 1957. Sex determination in Metazoa. Proc. Zool. Soc. Calcutta, Mookerjee
Memorial Volume 1957: 13-28.
HARRINGTON, R. W., JR., 1959. Delayed hatching in stranded eggs of marsh killifish, Fmidulns
conflucntus. Ecology, 40 (3) : 430-437.
HARRINGTON, R. W., JR., 1961. Oviparous hermaphroditic fish with internal self-fertilization.
Science, 134 (3492) : 1749-1750.
HARRINGTON, R. W., JR., 1963. Twenty-four-hour rhythms of internal self-fertilization and of
oviposition by hermaphrodites of Rivulus mannoratits. Plivsiol. Zool., 36 (4) : 325-
341.
HARRINGTON, R. W., JR., AND L. R. RIVAS, 1958. The discovery in Florida of the cyprinodont
fish, Rivulus marmoratus, with a redescription and ecological notes. Copcia. 1958 (2) :
125-130.
HAYES, F. R., 1949. The growth, general chemistry, and temperature relations of salmonid
eggs. Quart. Rev. Biol., 24: 281-308.
HOEDEMAN, J. J., 1958. Rivulid fishes of the Antilles. Studio Fauna Curasao, 8: 112-126.
HUXLEY, J. S., 1923. Late fertilization and sex-ratio in trout. Science, 58 (1502) : 291-292.
KALLMAN, K. D., AND R. W. HARRINGTON, JR., 1964. Evidence for the existence of homozygous
clones in the self-fertilizing hermaphroditic teleost Rivulus marmoratus (Poey). Biol.
Bull., 126: 101-114.
KINNE, O., AND E. M. KINNE, 1962. Rates of development of embryos of a cyprinodont fish
exposed to different temperature-salinity-oxygen combinations. Canad. J. Zool., 40 :
231-253.
LINDSEY, C. C., 1962. Experimental study of meristic variation in a population of threespine
sticklebacks, Caster ostcus aculeatus. Canad. J. Zool., 40: 271-312.
MEYER-ABICH, A., 1956. The principle of complementarity in biology. Acta Biothcorctica, Scr.
A, 11: 57-74.
MRSIC, W., 1923. Die Spatbefruchtung uncl deren Einfluss auf Entwicklung und Geschlechts-
bildung, experimentell nachgepriift an der Regenbogenforelle. Arch. f. Entzv., 98: 125-
209.
NIELSEN, E. T., 1961. Twilight and the "crep" unit. Nature, 190 (4779) : 878-879.
NIELSEN, E. T., 1963. Illumination at twilight. Oikos, 14 (T) : 9-21.
NOGUSA, S., 1960. A comparative study of the chromosomes in fishes with particular consider-
ation on taxonomy and evolution. Mem. Hyogo Univ. Agric. (Biol.), 3(1): 1-60.
PADOA, E., 1939. Observations ulterieures sur la differenciation du sexe, normale et modifiee
par I'administration d'hormone folliculaire, chez la Truite iridee (Salmo widens).
Bio-Morphosis, 1: 337-354.
PFLUGER, E., 1882. Versiiche der Befruchtung iiberreifer Eier. Pfliiger's Arcliiv jilr gc sarnie
Physiol. des Mcnsclicn nnd der Tiere, 29: 76-77.
LOW-TEMPERATURE-PRODUCED MALE FISH 199
PIQUET, J., 1930. Determination clu sexe chez les batraciens en fonction de la temperature.
Rn>. Snissc Zoo!.. 37: 173-281.
RASQUIN, P., AND L. ROSENBLOOM, 1954. Endocrine imbalance and tissue hyperplasia in teleosts
maintained in darkness. Bull. Auicr. Mns. Nat. Hist., 104 (4) : 365-425.
REINBOTH, R., 1962. Morphologische und functionelle Zweigeschlechtlichkeit bei marinen
Teleostieren ( Serranidae, Sparidae, Centracanthidae, Labridae). Zool. Jahrb. Abt.
fiir Zool. und Physiol. dcr Tiere, 69: 405-480.
RIVAS, L. R., 1945. The discovery and redescription of the types of Rivulus inaniwratus Poey,
a cyprinodont fish from Cuba. /. H'asli. Acad. Sci., 35 (3) : 95-97.
ROBERTSON, J. G., 1953. Sex differentiation in the Pacific salmon Oncorhynchus kcta (Wal-
baum). Canad. J. Zool.. 31 (2) : 73-79.
ROTHSCHILD, LORD, 1958. Fertilization in fish and lampreys. Biol Rev., 33: 372-392.
SINHA, V. R. P., AND J. W. JONES, 1966. On the sex and distribution of the freshwater eel
(AmniilUt anguilla). J. Zoo!.. Land.. 150 (3): 371-385.
SINNOTT, E. W., AND L. C. DUNN, 1939. Principles of Genetics. 3rd ed., McGraw-Hill Book
Co., New York and London.
SPURWAY, H., 1957. Hermaphroditism with self-fertilization and the monthly extrusion of
unfertilized eggs in the viviparous fish Lebistcs reticulatus. Nature, 180 (4597) :
1248-1251.
UCHIDA, T., 1937. Studies on the sexuality of Amphibia. III. Sex-transformation in Hynohius
rctardatus by the function of high temperature. /. Fac. Sci. Hokkaido hnpcr. Univ.,
Ser. VI, 6: 59-70.
VAN DOORN, W. A., 1962. Geschlachtsbeinvloeding door temperatur. Hct Aquarium, 32 (9) :
208-209.
WALFORD, R. L., AND R. K. Liu, 1965. Husbandry, life span, and growth rate of the annual
fish, Cynolebias adloffi^ E. Ahl. Exp. Geront., 1: 161-171.
WHITE, M. J. D., 1954. Animal Cytology and Evolution. 2nd ed., Cambridge University
Press, Cambridge, England.
WITSCHI, E., 1929. Studies on sex differentiation and sex determination in amphibians. II.
Sex reversal in female tadpoles of R. sylvatica following the application of high
temperature. /. E.\-p. Zool., 52: 267-291.
WITSCHI, E., 1957. The inductor theory of sex differentiation. /. Fac. Sci. Hokkaido Univ.,
Ser. n. 13 (Professor T. Uchida Jubilee Volume) : 428-439.
YAMAMOTO, T., 1953. Artificially induced sex-reversal in genotypic males of the Medaka
(Orysias latipcs). J. Exp. Zool.. 123: 571-594.
YAMAMOTO, T., 1958. Artificial induction of functional sex-reversal in genotypic females of the
Medaka (Orysias latipcs). J. Exp. Zool., 137: 227-260.
YAMAMOTO, T., 1959. The effects of estrone dosage level upon the percentage of sex-reversals
in genotypic male (XY) of the Medaka (Oryzias latipcs). J. Exp. Zool., 141: 133-154.
YAMAMOTO, T., 1961. Progenies of sex-reversal females mated with sex-reversal males in the
Medaka, Orysias latipcs. J. Exp. Zool., 146: 163-179.
YAMAMOTO, T., 1962. Hormonic factors affecting gonadal sex differentiation in fish. Gen.
Couip. Endoerin., Suppl., 1: 341-345.
ZAHL, P. A., 1934. The hormonic inhibition of a somatic character in the cyprinodont fish,
Rivulus urophthalinus. Ainer. Nat., 68: 475-477.
RETINOMOTOR RHYTHMS IN THE GOLDFISH,
CARASSIUS AURATUS1
KENNETH R. JOHN, MARC SEGALL 2 AND LAWRENCE ZAWATZKY 3
Franklin and Marshall Collei/e, Lancaster, Pennsylvania 17604
We had preliminary evidence from histological preparations that the retina of
the goldfish, Carassins auratits, does show a persistent retino-motor rhythm in pro-
longed darkness, and we felt that the conclusions of Wigger (1941) stood in need
of verification. Wigger (1941) placed the fish in darkness at 1800 hr. and sampled
two fish at 2-hour intervals during the first day and at 1200 hr. and 2000 hr. the
second day. According to his graph, the cones elongated smoothly to their positions
of maximal dark-adaptation by 2200 hr. the first night, contracted sharply to a new
position by 2400 hr. and remained there until 0400 hr. Then they contracted
sharply to a location intermediate between the light- and dark-adapted positions by
0600 hr. and remained approximately in that position to the end of the experiment.
Wigger's results demonstrated that the rhythmic migration of cones in the goldfish
retina did not persist in constant darkness after the first 12 hours. He did not
distinguish between types of cones.
It is well known that within a species the distances of migration of visual cells
may vary considerably as do the thicknesses of the visual cell layers (Ali, 1963 ;
Engstrom and Rosstrop, 1963 ) and that single and double cones migrate different
distances (Walls, 1942; Nicol, 1965). In repeating these experiments on the gold-
fish, we followed the separate courses of migration of single and double cones.
MATERIALS AND METHODS
The experiments involved 6 groups of goldfish (mean total length 6.6 ± 0.3
cm.) obtained from Nolt's Ponds, Silver Springs, Pennsylvania.
Group A : 82 fish, conditioned to the natural diel cycle, were placed in the dark-
room at sunset, 2036 hr., on June 23, 1965. We fixed samples of 4 fish at 1-hour
intervals between 2100 hr. and 0500 hr., and then at sunrise, 0534 hr. Thereafter,
we fixed samples of 4 fish at 0800 hr., 1100 hr., 1300 hr., 1500 hr., 1800 hr., and
2036 hr.
Group B : 21 fish, conditioned as in group A, were placed in the darkroom at
sunset, July 5, 1965. We fixed three fish at 1300 hr. for 7 consecutive days.
Group C: 12 fish were conditioned for 45 days on an artificial cycle, 12 hr.
20 min. light and 11 hr. 40 min. dark. The light was turned off permanently at
1910 hr. (sunset) on November 19, 1965, and for three consecutive days we fixed
two fish at 2400 hr. and 1200 hr.
1 Supported by funds from the National Science Foundation (G- 19256).
- Present address : George Washington University, School of Medicine.
3 Present address : Johns Hopkins University, School of Medicine.
200
RETINOMOTOR RHYTHMS IN THE GOLDFISH 201
Group D : 8 fish, conditioned as in Group A, were placed in the darkroom at
sunset, June 13, 1966. We fixed two fish at 0100 hr. and 1300 hr. of the two
consecutive days.
Group E : 18 fish, conditioned as in Group A, were placed in the darkroom at
sunset on June 27, 1966. We fixed two fish at 1-hour intervals between 2100 hr.
and 0500 hr.
Group F : 4 fish, conditioned as in group A, were placed in the darkroom at
sunset on July 13, 1966. We fixed two fish at 0300 hr. and 0400 hr.
Light-adapted fish were fixed at 1200 hr. in July and November, 1965.
The fish were maintained in 77.5-liter aquaria equipped with filters and aerators.
They were fed daily before noon though not at a regular hour. They were not fed
during the experiments. The temperature was 25° ± 1° C.
The fish were fixed in Bouin's fluid in darkness (exclude light-adapted fish)
and remained in the fixative at least two hours before the corneas and lenses were
removed from the right eyes. The eyes were then dehydrated in an ethyl alcohol
series, cleared in xylene, and embedded in paraffin with \% beeswax.
In group C, radial and serial tangential sections were cut from a sector of
retina 2 mm. square, located 1 mm. ventral to the optic nerve. In all other groups,
sections were taken from the entire eye through the plane of the optic nerve on a
dorso-ventral axis. All sections were cut at 5 or 10 ju. and stained with Harris's
haematoxylin and eosin. In order to expose the cones for accurate identification
and measurement, some sections were bleached with the potassium permanganate-
sodium bisulfite method, and stained with 3% ferric chloride and eosin.
The interpretation of the behavior of the retina was based upon the following
measurements from each eye :
Group A : 5 measurements of the thickness of the visual cell layer, the location
of the single cones, and the location of both the long and short segments of the
double cones, all in a region of the retina 1.2 mm. ventral to the optic nerve. The
visual cell layer was defined as the distance between the ELM (external limiting
membrane) and the lamina basalis. The location of the cones was represented by
the distance from the ELM to the distal end of the cone ellipsoid. All measure-
ments were made with an ocular micrometer.
Group B : All measurements made as in group A.
Group C : Radial sections measured as in group A. On the serial tangential
sections, using a Whipple-Hauser ocular micrometer, we counted the numbers of
single and double cones in an area 670 p. square in each section beginning at the
ELM and progressing to the lamina basalis. The distance between elements in
successive sections was represented by the thickness of the sections, 5 /JL.
Groups D, E, and F : The measurements were made as in group A except that
10 measurements were made rather than 5.
For final comparisons the measurements were converted to RP values (location
of visual cell/thickness of visual cell layer X 100) described by Engstrom and
Rosstrop (1963).
RESULTS
All graphs of double cones from radial sections represent the measurements on
the long segment which had a mean length of about 3 p. greater than the short seg-
202
K. R. JOHN, M. SEGALL AND L. ZAWATSKY
70 -
UJ
< 50 -
O.
tr 30 H
II 1 1 1 1 1 1 1 II
SS 2200 2400 0200 0400 SR
TIME
IN
HOURS
FIGURE 1. Pattern of migration of double cones during first night of darkness. • Group A,
O group E, and 3 group F. SS = sunset, SR = expected sunrise, — = mean values.
merits. There was no change in the difference hetween the lengths of short and
long segments for eyes fixed at different times. The patterns of photomechanical
changes during the course of the first night are shown for douhle cones in Figure 1 ,
and for single cones in Figure 2. Each graph is a composite of results from
experimental groups A, E, and F. The points on the graph within a group are
not hased upon equal numbers of fish because some of the eyes did not produce
useful histological sections. In Figure 1, the points for group A at 0100 hr.,
0200 hr., and 0400 hr. represent three fish ; all other points represent 4 fish. The
point for group E at 2100 hr. represents one fish and all other points represent two
fish. Each point for group F represents two fish. In Figure 2, the points for
group A at 0100 hr. and 0400 hr. represent three fish ; 0200 hr., two fish; all other
points 4 fish. In group E, the point at 2100 hr. represents one fish, and all other
points represent two fish. Each point for group F represents two fish.
70 -
50 -
Q_
tr 30 H
SS 2200
2400 0200
0400 SR
TIME
IN
HOURS
FIGURE 2. Pattern of migration of single cones during first night of darkness. • Group A,
O group E, and 3 group F. SS = sunset, SR = expected sunrise, -- = mean values.
RETINOMOTOR RHYTHMS IN THE GOLDFISH
203
FIGURE 3. Photomicrographs of radial sections showing retinas in different states of
adaptation, (a) Light-adapted, 10 ^, H&E. (b) Noon dark-adapted after 17 hours in dark-
ness, 10 n, H&E. (c) Midnight dark-adapted after 5 hours in darkness, pigment bleached;
5/x, ferric chloride and eosin. The bar in (b) represents 100 /x. e = external limiting mem-
brane, 1 = lamina basalis, s = single cones, and d — double cones. 450 X.
The cones migrated through a cycle of positions during the course of the night,
but in each group of fish they followed an irregular and unique course with maxi-
mal extensions occurring at different times in the night. Between 2100 hr. and
0200 hr., the corresponding points of groups A and E almost always deviated in
opposite directions. The plot of the mean within that interval was relatively
smooth. Between 0200 hr. and 0500 hr.. the fluctuations in groups A and E
followed parallel courses and both exhibited a prominent positive shift in the slope
TABLE I
Mean RP values of double cones during the first 24 houis in darkness, based upon radial sections.
Sunset was at 2036 hr. and the expected sunrise was 0534 hr. Phis and minus values
represent one standard deviation. The p values are based upon Student's t-test
between successive time periods
Groups
Time
Number of fish
RP value
P
A, E, F
2100
5
53.6 ± 5.1
2200
6
59.2 ± 2.5
> .05
2300
6
59.5 ± 4.8
n. s.
2400
6
63.5 ± 5.8
n. s.
0100
5
64.4 ± 4.1
n. s.
0200
5
65.6 ± 3.4
n. s.
0300
8
59.1 ± 6.9
n. s.
0400
7
61.1 ±4.4
n. s.
0500
6
54.2 ± 2.9
> .01
A only
0534
4
59.3 ± 7.3
n. s.
0800
3
45.1 ± 1.1
> .05
1100
3
49.0 ± 3.6
n. s.
1300
4
50.2 ± 3.9
n. s.
1500
4
53.8 ± 6.4
n. s.
1800
4
53.7 ± 8.4
n. s.
2037
4
50.0 ± 3.8
n. s.
204
K. R. JOHN, M. SEGALL AND L. ZAWATSKY
TABLE 1 1
Mean RP values of single cones during the first 24 hours in darkness, based upon radial sections.
Sunset was at 2036 hr. and the expected sunrise was 0534 hr. Plus and minus values
represent one standard deviation. The p values are based upon Student's t-test
between successive time periods
Groups
Time
Number of fish
RP value
P
A, E, F
2100
o
41.2 ± 8.0
2200
6
57.6 ± 6.8
> .01
2300
6
59.3 ± 7.3
n. s.
2400
6
67.8 ± 8.1
> .05
0100
5
69.9 ± 7.8
n. s.
0200
4
69.8 ± 3.0
n. s.
0300
8
57.0 ± 6.7
> .01
0400
7
53.2 ± 6.9
n. s.
0500
6
42.8 ± 7.3
> .05
A only
0534
4
46.4 ± 13.0
n. s.
0800
i
30.8 ± 3.7
n. s.
1100
3
38.6 ± 0.7
> .05
1300
4
31.3 ± 7.8
n. s.
1500
4
32.6 ± 3.9
n. s.
1800
4
30.7 ± 4.3
n. s.
2037
4
30.3 ± 4.6
n. s.
of the curve between 0300 hr. and 0400 hr. The slope of the curve in that interval
for group F was negative.
The single cones showed a greater photomechanical shift than the double cones.
The photomicrographs in Figure 3 show that the distal margins of the ellipsoids
of single cones were nearer the ELM in the 1300 hr. retina and farther from the
ELM in the 0100 hr. retina than the double cones. The relative fluctuations be-
tween points in the graphs of single and double cones were about equal (Figs. 1
and 2), but the single and double cones may have shown some independent behavior.
70 -
UJ
< 50 -
>
Q.
tr 30 H
I
5
I
17
29
I
41
53
I
65
HOURS IN DD
FIGURE 4. Rhythmic shift of positions of double cones based upon serial tangential sections.
Group C.
RETINOMOTOR RHYTHMS IN THE GOLDFISH
205
70 -
LJ
Z)
< 50 -
o.
a: 30 -
I 0
I
5
I
17
I
29
HOURS
I
41
IN
T
53
I
65
DD
FIGURE 5. Rhythmic shift of positions of single cones based upon serial tangential sections.
Group C.
Note in Figure 1 that only three of the points for group A lie above those of group
E, while in Figure 2, 7 of the points for group A lie above those of group E.
Tables I and II summarize the combined statistical data for groups A, E and
F. As shown by the probability values derived from Student's t-test, there were
significant differences between the adjacent points from 2100 hr. to 2200 hr., 2300
hr. to 2400 hr., 0200 hr. to 0300 hr., and 0400 hr. to 0500 hr. The significant
differences practically alternated with insignificant differences. The data for group
A show that both single and double cones continued to contract beyond the time
of expected sunrise. The RP values reached a minimum at 0800 hr. and fluctuated
around that value through the remainder of the day.
Both single and double cones showed a persistent photomechanical rhythm in
constant darkness. This was first demonstrated by group C (Figs. 4 and 5) in
which the results were based upon serial tangential sections. Note that the
TABLE 1 1 1
Mean RP values of double cones through 65 hours in darkness, based upon serial tangential sections.
The eye at 0 hr. was light-adapted. Sections progress from the EL]\1 toward the lamina
basalts. Plus and minus values represent one standard deviation
Serial section sequence
Hours
Fish
RP
in DD
number
value
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
0
1
0
34
50
21
20.9 ± 5
5
2
0
0
0
0
0
0
0
0
0
2
17
29
18
82.5 ± 6
5
3
0
0
0
0
0
0
13
25
35
26
61.3 ± 7
17
4
0
0
0
0
0
0
3
5
2
11
19
17
8
53.4 ± 8
17
5
0
0
0
4
7
10
12
17
20
14
38.0 ± 9
29
6
0
0
0
0
0
0
0
0
0
1
1
13
29
30
19
70.4 ± 6
29
7
0
0
0
0
0
0
0
0
0
0
1
5
17
38
29
21
3
66.2 ± 17
41
8
0
0
1
1
0
0
7
17
18
20
16
5
54.4 ± 10
41
9
0
0
0
0
0
2
0
2
1
0
8
19
17
17
17
3
4
64.0 ± 10
53
10
0
0
0
0
0
0
0
0
0
0
0
0
6
12
26
46
37
21
1
72.5 ± 6
53
11
0
0
0
0
0
0
0
0
0
0
2
20
34
27
9
1
69.2 ± 9
65
12
0
2
10
17
17
19
17
8
0
1
24.9 ± 8
65
13
0
0
0
6
11
19
17
10
5
2
42.2 ± 10
206
K. R. JOHN, M. SEGALL AND L. ZAWATSKY
TAUI.K IV
Main RP values of single cones through 65 hours in darkness, based upon serial tangential sections.
The eye at 0 hr. was light-adapted. Sections progress from the ELM toward the lamina
basalis. Plus and minus values represent one standard deviation
TT' ,1,
Serial section sequence
Hours
r isil
RP
in DD
n u m -
bcr
value
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0
i
4
45
24
3
19. 8± 4
5
2
0
0
0
0
0
0
0
1
4
7
15
22
9
1
78. 7 ± 9
5
3
0
0
0
3
3
5
6
20
29
27
1
62.7±13
17
4
0
0
3
25
25
9
19
20
23
23
25
26
29
9
4
5
45.5±19
17
5
2
6
7
12
13
4
11
16
18
10
2
32.0 ±13
29
6
0
0
0
0
1
2
1
6
1
2
5
11
16
15
15
1
63.2±14
29
7
0
0
0
0
1
3
4
1
1
0
0
7
33
25
13
11
1
60.4±10
41
8
0
0
6
14
14
6
5
5
9
17
18
11
1
46.8±14
41
9
0
1
1
5
9
11
10
4
5
10
12
13
20
29
23
22
15
58.6±19
53
10
0
0
0
0
0
1
2
4
4
8
16
7
15
13
16
12
22
26
8
1
65.2±15
53
11
0
0
0
0
0
0
5
2
4
8
12
15
12
13
3
66.4±12
65
12
0
13
15
9
15
14
15
19
13
4
26.6±11
65
13
0
3
10
13
17
12
17
11
8
1
36.9±11
maximal RP values occurred on the first night. After the first night, the cones
did not migrate the full distances. Note also that on the first night the maximal
extension of single cones exceeded that of the double cones, and that subsequently
the single cones did not extend as far as the double cones. In the statistical sum-
mary of group C (Tables III and IV), the data for fish numbered 4 through 10
strongly suggested the existence of two populations of single cones. (We did not
attempt to distinguish between types of single cones.)
The persistent rhythm in the cones was again demonstrated by group D (Figs.
6 and 7). Although the data for group A were incomplete for these purposes,
they were included in the figures with the broken line suggesting the location of
the point that is missing at 29 hr.
During 8 days in darkness, as shown by eyes fixed at 1300 hr. (Table V), the
cones remained in the intermediate positions corresponding to the positions of the
70 -
LU
< 50 -
>
Q_
o: 30 -
r
5
17
HOURS
29
41
IN
DD
FIGURE 6. Rhythmic shift of positions of double cones based upon radial sections. • Group
A ; O group D. Note group A no sample taken at 29 hours.
RETINOMOTOR RHYTHMS IN THE GOLDFISH
207
70 -
LJ
D
< 50 -
o.
a: 30 -i
I 0
I
5
17
HOURS
29
41
IN
DD
FIGURE 7. Rhythmic shift of positions of single cones based upon radial sections. • Group
A, O group D. Note group A no sample taken at 29 hours.
first two or three days. The RP values became consistently smaller for the single
cones after the fourth day, but varied up and down for the double cones.
DISCUSSION
Although the retinas of different goldfish under the same conditions showed
some marked variations, mean values based upon two fish for each sampling period
adequately demonstrated the persistent circadian rhythm in the cones. There may
be no statistically significant differences between consecutive points based upon
small samples, but the probability of generating a cyclical function by chance is so
small that there is little doubt of the validity of the rhythm in the goldfish. If we
assumed a normal distribution for the variations between fish, the chance that the
values for two fish would both fall either above or below the center of the distri-
bution would be 1/4. The probability of getting the cycle over three days with
TABLE V
Mean RP values of single and double cones at 1300 hr. during 8 consecutive days in
darkness. Plus and minus values represent one standard deviation
RP value
Day
Number of fish
Single cones
Double cones
1
4
31.3 ± 7.8
50.2 ± 3.9
2
2
28.3 ± 2.7
51.2 ± 0.5
3
3
33.9 ± 4.5
57.2 ±3.7
4
1
30.6
59.5
5
1
29.2
46.9
6
2
22.8 ± 1.1
53.1 ± 3.6
7
1
19.2
31.7
8
2
20.1 ± 8.9
53.6 ± 6.5
208 K. R. JOHN, M. SEGALL AND L. ZAWATSKY
12 fish would lie 1/4096, and over two days with 8 fish it would he 1/256. The
reproducibility of the rhythm with different lots of fish further supported its validity.
Using small samples, one could not predict the exact time of occurrence of
maximal amplitudes in the retinomotor cycle as shown by the graphs in Figures 1
and 2, but the ease of obtaining the cycle indicated that the variations through the
middle of the night did not generally overlap with those of the day. The rhythms
were demonstrated by small samples collected at the mid-points of the nocturnal
and diurnal segments of the cycle. In fact, had only one fish been represented in
each sample, a cyclical function would have been obtained most of the time. In
only one of 12 fish, no. 9 (Tables III and IV), did the mean positions of the cones
deviate sufficiently to obscure the existence of a rhythm. The failure of Wigger
(1941) to observe a persistent retinomotor rhythm in the goldfish is inexplicable.
We do not know why the values for group D were consistently lower than
those of group A (Figs. 6 and 7). The fish were the same size, obtained at the
same time of year, and treated in the same way, but did represent different years.
We do not know whether the fish had different experiences prior to our purchasing
them. It is well established from work on other organisms that the circadian
rhythms are sensitive to a wide range of factors (Aschoff, 1965).
None of the cycles in this study or in the literature represent an individual fish.
They represent small samples and the variations show that the results from small
samples have limited comparative value. They also represent estimates of the
mean values for the populations from which the samples were drawn, but it could
be misleading to generalize about population values from small samples. Wigger
(1941) paid particular attention to the fact that the cones in darkness became
maximally extended before 2400 hr. and contracted about half the distance to a
stable position between 2400 hr. and 0400 hr. before contracting farther. Our
results (Figs. 1 and 2) illustrate that the time of maximum extension of the cones
for any individual or small number of fish might occur at any time from 2300 hr.
to 0300 hr. It would be interesting to be able to follow the course of migration
of cones in a single fish.
It is important to note that Wigger (1941) sampled fish every two hours while
we sampled fish every hour. If the graphs in Figures 1 and 2 were redrawn con-
necting points at two-hour intervals beginning at 2200 hr., the conspicuous fluc-
tuations would disappear and the curves would become relatively smooth and
symmetrical and would contradict our results as well as those of Wigger (1941).
It is apparent that the shape of the curve is a function of the sampling interval.
The mean values of the graphs in Figures 1 and 2 suggest a change in the
slope of the curve at 0300 hr. This change would lie attributed to groups A and
E, but the contrary results of group F suggest that the deflections of groups A
and E were caused by chance. Further work would be required to determine the
mean values and the nature of the variations for the population.
Engstrom (1960) described two types of single cones in the light-adapted gold-
fish retina and stated that the shorter type had no myoid process. This means that
the shorter cones would not migrate during adaptive changes in the retina. We do
not doubt the occurrence of two types of single cones, but a comparison of measure-
ments from light-adapted (no. 1) and dark-adapted (nos. 2 and 3) fish in Table
IV shows that all single cones migrated an average of 50 RP units, which means
RETINOMOTOR RHYTHMS IX THE GOLDFISH 209
that all cones possessed myoid processes. We think that any interpretation of mi-
gratory capacities of visual cells should be based on a comparison of light- and
dark-adapted eyes.
Our studies have shown that to obtain an eye in a state of maximal dark-
adaptation, one must choose eyes during the first night of darkness. Thereafter,
though the circadian rhythm would give a more dark-adapted eye at night than
during noon darkness, the RP values would be distinctly lower than they would
be on the first night. Arey and Mundt (1941) stated that the rhythm in the black
bullhead, Ameiurus nebulosus, persisted through 4 days of constant darkness, the
limit of their experiment.
Engstrom and Rosstrop (1963) interpreted this as a general guideline for
experimental designs. To assure that the eyes of the roach, Leuciscus rutihis, were
totally dark-adapted and free from the influence of a rhythm, they held the fish in
darkness for 4 days before initiating studies on retinal adaptation at low levels of
illumination. They illustrated the retina of a roach exposed to 10"e ft. c. and stated
(p. 155) that, "A histological comparison between eyes from '. . . 10~6 ft. c. . . .'
and the totally dark-adapted ones does not reveal any noticeable differences." The
dark-adapted eye was not identified, but was presumably one that had been in
darkness for 4 days. They also did not state the time of fixing the eye. Their
illustration of the retina of the roach looks like a 1300 hr. dark-adapted goldfish
retina and not like a 2400 hr. dark-adapted retina. See Figure 3 and Tables III
and IV for the relative positions of single and double cones in the 1200 hr. and
2400 hr. goldfish retinas. We think that the retina of the roach, after 4 days in
darkness, was exhibiting a persistent rhythm. The question on the longevity of
the persistent retinomotor rhythm in fishes has not been answered. At the end
of three days, it was well defined in the goldfish, and the conditions of the retinas
at 1300 hr. over a period of 8 days of darkness suggest that the rhythm was per-
sisting. At least, the retina was remaining in an intermediate condition, not a
dark-adapted condition.
Generalizations on the presence or absence of persistent retinomotor rhythms
in fishes appear to be based on inadequate factual support. Von Studnitz (1952),
citing the appropriate literature, mentioned two species in the discussion of rhythms,
the black bullhead and the goldfish. Since, according to Wigger (1941) the gold-
fish did not show a persistent rhythm, the sole evidence for such a rhythm rested
on the bullhead. Yet, AH (1961) cited von Studnitz (1952) for the statement
that rhythms occur in certain fishes. Engstrom and Rosstrop (1963) cited von
Studnitz (1940) as the authority for the statement that persistent retinomotor
rhythms are not general among fishes. There has not been sufficient study to
support any wide generalization on persistent retinomotor rhythm in fishes, but
the following quotation (p. 357) from Welsh and Osborn (1937) indicates that
such rhythms may be widespread: "Several species (not named in the paper)
other than Ameiurus were treated. . . . Not enough individual fishes were em-
ployed, however, to yield quantitative results, but without question, the phenomenon
is fairly widespread."
SUMMARY
1. The goldfish retina shows a persistent circadian rhythm. In constant dark-
ness for three days, the cones continued to shift positions in synchrony with the
210 K. R. JOHN, M. SEGALL AND L. ZAWATSKY
diel cycle. The amplitude of the shift decreased after the first night. A maximally
dark-adapted retina was obtained only on the first night.
2. Individual fish showed considerable variation in the time of occurrence of
maximal dark-adaptation. This condition might be attained at any time between
2300 hr. and 0200 hr. The mean values for all fish suggested that the curve of
progress of dark-adaptation for the population would be symmetrical with the
maximal dark-adapted condition occurring at mid-night.
3. The single and double cones showed some characteristic differences in their
behaviors. All cones migrated, but the relative excursions of single and double
cones changed after the first night. The migratory patterns suggested the existence
of two kinds of single cones.
4. The longevity of persistent rhythms in fish retinas is not known, but the
assumption that it ceases after 4 days is based upon misinterpretation of a state-
ment by Arey and Mundt (1941) about the black bullhead.
5. After 8 days in darkness the retina at 1300 hr. was in an intermediate state,
not dark-adapted. If anything it had drifted toward the light-adapted state.
LITERATURE CITED
ALI, M. A., 1961. Histophysiological studies on the juvenile Atlantic salmon (Salmon salar)
retina. II. Responses to light intensities, wavelengths, temperatures, and continuous
light or dark. Canadian J. Zool., 39: 511-526.
ALI, M. A., 1963. Correlation of some retinal and morphological measurements from the
Atlantic salmon (Salmo salar). Grozvth, 27: 57-76.
AREY, L. B., AND G. H. MUNDT, 1941. A persistent diurnal rhythm in visual cones. Anat.
Record, 79 (Suppl.), 5 (Abstr.).
ASCHOFF, J., 1965. Circadian Clocks. North-Holland Publ. Co. Amsterdam.
ENGSTROM, K., 1960. Cone types and cone arrangement in the retina of some cyprinids. A eta
Zool., 41 : 277-295.
ENGSTROM, K., AND E. ROSSTROP, 1963. Photomechanical responses in different cone types of
Leuciscits rutilus. Acta Zool., 44: 145-160.
NICOL, J. A. C, 1965. Retinomotor changes in flatfishes. /. Fish. Res. Bd. Canada, 22: 513-
516.
STUDNITZ, G. VON, 1940. Physiologie des Sehens. Retinale Primarprocesse. Akademische
Verlagsgesellsschaft. Leipzig.
STUDNITZ, G. VON, 1952. Physiologie des Sehens. Retinale Primarprocesse. 2nd ed. Akade-
mische Verlagsgesllschaft. Leipzig.
WALLS, G. L., 1942. The Vertebrate Eye and its Adaptive Radiation. Hafner Publ. Co.
New York.
WELSH, J. H., AND C. M. OSBORN, 1937. Diurnal changes in the retina of the catfish, Amciurus
ncbnlosns. J. Comp. Neural, 66: 349-359.
WIGGER, H., 1941. Diskontinuitat und Tagesrhythmik in der Dunkelwanderung retinaler Ele-
mente. Zcitschr. vergl. Physiol, 28 : 421-427.
CHANGES IN THE HEMOCYTE PICTURE OF
GALLERIA MELLONELLA (LINNAEUS) 1
JACK COLVARD JONES 2
Department of Entomology, University of Maryland, College Park, Maryland 20742
In this paper differential and total hemocyte counts were obtained and combined
with hemolymph volume determinations in order to estimate the changes which
occur in the hemocyte picture of the wax moth Caller ia melloneUa (Linnaeus)
from the eleventh through the twenty-first days of larval life, during which period
the larvae pass through successive phases of feeding, crawling, spinning a cocoon,
and preparing to pupate.
When reared by the method of Beck (1960), the larvae reach a fairly large size
within about 10 to 11 days. During the next 10 days or so they are particu-
larly suitable for hematological studies. Hemolymph for differential counts was
collected from manually immobilized, unanesthetized, 10- to 12-day-old larvae by
piercing an intersegmental membrane with a sharp needle. Hemolymph for differ-
entials from larger larvae was conveniently obtained either by cutting a proleg or
one of the protuberances on the last abdominal segment. The fresh, unfixed, and
undiluted hemolymph was collected directly on a slide and a coverslip added. The
cells were examined with a phase contrast microscope at X 970 and were classified
using the nomenclature of Jones (1962). From 200 to 1000 cells were identified
per preparation. All studies were made on larvae freshly taken from an incubator
held at 34° to 35° C.
Total hemocyte counts (cells per microliter) were generally made on the first
drop of hemolymph emerging from a cut proleg of both unfixed (= untreated) and
heat-fixed larvae.3 Heat-fixation consisted of immersing larvae in a water bath
at 55° C. for one minute. Hemolymph was quickly drawn to the 0.25 mark of
a Thoma WBC diluting pipette and then rapidly diluted with 2% acetic acid to the
1 1 mark. After shaking vigorously and discarding the first three drops from the
pipette, a double-lined hemocytometer was filled and the cells within 5 of the one-
millimeter ruled squares were counted.
Hemolymph volumes were determined using the method of Yeager and Munson
(1950), that is, by injecting the larvae with 10 microliters of \% amaranth red
in saline per gram body weight. Five larvae were used for each day of study.
The dye was allowed to circulate for 3 to 5 minutes and a proleg severed and the
1 This research was sponsored by N.I.H. Grant HE 5193 and by Department Award K 3
GM 21,529. Scientific Article Number A1304, Contribution Number 3846 of the Maryland
Agricultural Experiment Station.
2 I am most indebted to Mrs. Daisy P. Liu, Mr. Richard A. Werner, and Dr. Ronald E.
Wheeler for their assistance with various portions of this work. The manuscript was much
improved by the comments of Drs. Malcolm Lea, Yasukito Nittono, and Gertraude Wittig and
I am very grateful to them for their help.
3 Differential and total counts were made on separately reared batches of larvae.
211
212 JACK COLVARD JONES
hemolymph collected in a capillary tube. The intensity of the color was compared
to a series of known dilutions of the dye. The hemolymph volume percentages were
converted into microliters.
RESULTS
1. Differential hemocyte counts
During the last 10 to 12 days of larval life the following types of hemocytes
could be easily recognized in unfixed hemolymph examined with phase microscopy :
(1) prohemocytes, (2) plasmatocytes, (3) spherule cells, (4) adipohemocytes, and
(5) oenocytoids, as Ashhurst and Richards (1964) have previously noted. Be-
cause so many transitional forms were seen between prohemocytes and plasmato-
cytes, it was very difficult or impossible to separate them accurately for quantitative
work and these two types were combined into a common category which will, for
convenience, be termed "plasmatocytoids." Cells seemingly transitional between
plasmatocytoids and mature adipohemocytes were encountered during a definite
period of larval life, and a series of counts were made in which this apparently
intermediate category of cells was enumerated in addition to the other categories.
These intermediate cells are termed immature adipohemocytes. Mitotically divid-
ing hemocytes (probably prohemocytes) were counted and treated separately from
the other categories. Adipohemocytes and spherule cells were never seen in divi-
sion. The following types of hemocytes were not seen in Galleria mellonella
larvae : granular hemocytes, cystocytes, podocytes, and vermiform cells. A few
degenerating and unidentifiable hemocytes were encountered and they were so
categorized in many differential counts.
Differential counts are given in Table I. During the actively feeding period
(that is, from the eleventh through the fifteenth days of larval life), it is evident
that (1) the plasmatocytoids ranged from 90% to 100% (with an overall mean
of 96.7%), (2) adipohemocytes were consistently absent from the circulating
hemolymph, (3) spherule cells varied from none to 7% and averaged 1.4%, (4)
oenocytoids ranged from none to 8% and averaged 1.7%, (5) degenerating and
unidentifiable hemocytes varied from none to 3%, and (6) mitotically dividing
hemocytes (prohemocytes?) averaged 0.65%.
During the crawling, non-feeding, pre-cocoon-spinning period (approximately
between the sixteenth and seventeenth days), a few adipohemocytes were noted in
differential counts, and the number of dividing cells in such counts was reduced
to about one-half that of the actively feeding period.
As soon as the larvae start to spin their cocoons, however, immature adipo-
hemocytes suddenly increased to 12.6% and reached a maximum of 15.6% in
lightly cocooned larvae and thereafter declined. Mature adipohemocytes steadily
increased from 16% in the spinning period to a maximum of 57.1% in the newly
formed pupae. Spherule cells rapidly declined, following the lightly cocooned
period and were not observed in the pupae examined. Very few oenocytoids were
seen in young pupae.
According to the differential counts, as the larvae transform into pupae, (1)
plasmatocytoids decrease from about 96% to 41%, (2) immature adipohemocytes
suddenly appear, (3) mature adipohemocytes steadily increase, (4) spherule cells
HEMOCYTES OF GALLERIA
213
w
ca
H
5
a
^o *o
^> o
II
«J
-«
s
!^
|| 2
-« -^l1
ill
-2 I e
«
v.
s
s -s
o a
s-
"^
s~
"s X-
*s
/-•
***
s*
•s /•-
^
ir> rj ir> r
a r- r<
•51^^
h tvi r
•1 NC
u
•)
r*
r*
1 ro
4 C
: c
**
c<i O O C
»• (^^ ^™
5 0 C
5 O C
> C
3
C
; $
K
c
5°,^
c
3 C
3 C
>
O -H r^j -
O O O C
-i rN r
> o c
i csi r
3 0 C
1 *-H -
I
3 O C
•< C
3 C
4
?
r
C
^
T-
C
•j
^
^
c
H *•
3 C
H T-
> c
H
3
D
1-1 C-l
1 1 O
O 0
1
0
O
|
0
C
H
>
c
c
""" f~
ff ^m
C
J
Q
f<^ «-H ^H
1 1 1
0 O O
0
CN
I
1
O
£
c
>
c
3 Ic
H
c
S
55
tn
IO IO O* "-
4 (^) r<
•; oo c
: t- <
3 C
\
p<
•3 v«
3 r-
p(
V <D t~
r
M P1
•3 t-
4
m
CS r- 1 (N »-
H i — 1 *-
H *-H ^
H -H C
3 t-
H
H v
H r
4
r
C
parenth
O
r— oo oo "•
1
O O O C
^ IO T
3 O C
h to "
; o c
• 10 «•
3 O C
•* ^
> C
<
>
r*
^
"3 **
^ ^
~ t-
3 C
3
•^
C
1 fi v:
3 O C
3
3
\{
c
3 P
3C
•3 i/
3C1
3
^
C
•4
j
c
tn
Tt< •* -* ^
H ro v£
> ic c
5 t- C
3 "
•j
C
^ T
1- C
4
t-
» \o ^
o
0 C
3 VC
3
C
a
i—t *-H i— i r^
q -H c
5 -H -
H r-i »•
-i C
3
C
3 i-
* r*
4
H O r»
a
c
3 <-
i C
3
•d
cl
CN rf) ^ t-
1 1
^H o O C
» fff n
> 0 C
^ O «
>ic
^ OO "
3 O C
3 e-
3 C
4
3
f<
c
•> \c
3 G
3 «
3 C
•>
3
c
c
q ro i/
3i.
,
H
r:
C
1 M
3 C
•3 o
> c
4
3
tn
V
be
c
C3
a
3
c
N C
3
•(
I/
1 — i C
\
r
h N£
5 C
3
^
K
4.1
5
^
f
^
c
V.
3 C
-^ f
I "
•5 ,-
•^
c
^a
N "0 -
-1 ^— 1 I/
H
>
O
r
0 NC
4 T
3 f
< <
3
3
t-
u
•3
! counts wi
<
0 O C C
:> o c
: — c
3 tA C
OJ
a
3
r<
C
3
3
t-
r
t-
" r;
3 U
^
t-
h T
c
H
•
a
\<
f
\m
D OJ T-
3 ** f
3 O r
r^
H
J
\<
u
^
3 r
•3 \£
-i C
p
•) C
3 r-
\ *.
4 I/
N
*i
•>
C
t-
p
r\
*
Differentia
"
O 0 C C
: o c
5 e c
: c: -
r
C
l~
c^
V.
t-
u
r
-X
|\
^
"
4
t^
O
-*
t-~
i/^
1C
^o
*•
^
u
e>
s,
t-
u
r
•s ^
-i r
•> ^
4 v.
^
- C
•} ^
•4 C
•N
4 .
_^
^
H
>
*"
•^
c
N,
?N
C
~*.
h
3
~*
->
>
^00 "* >/
5 CN *-
H IO C
3 IO -
H t-
: z
N C
3
U
•5 f*; c
k
J
o
- c
- t
r
h
10 \o 10 ^c
o^ ^> <?> a
> t~ a
v ON C
3 O 01
N O C
D 10 a
N ON C
D C
i\ C
q
r\
C
u
^ >:
1 vi
h C
3 0
3
D
x£
t-
' * "
H
h
^
f f
H r;
") T
f f
t<
3
r
•r
>]
f
Cu
c
OO Ov C> C
^^ ^^ o^ '"
J,iJ,^
o^ o^ o°* c
> C
> O\ C
^ ON *~
1
J IO «
N ON C
5 C
5 ON C
H ON T-
1
•> ^ v£
N O- C
^ r"
3 ON C
H ON -
3 t^ I-
N 00 C
3 C
3 C
-• .-
- «•
N v£
3
3
u
t-
^
—
- c>
•( U
p •*
^
3 C
0
•^ *.
h -
\
H
-f
h
u
C
r<
•5 i~-i r
s 10 NC
H 0 f
") CN r
3
4
^
\£
C
r
H P<
3 VC
\ -
•1 f
•5 *-
3 r
H a
•3 T-
^
3
H
\<
^
c
3
H
4
Number
™llo
counted
0 0 O C
0 0 0 C
00 Ov 10 C
•— r^i r
5 O C
5 0 C
> *-> C
) ^-H r
5 0 C
5 0 C
5 rf C
4 01 r
3 C C
3 0 C
3 O C
-i t— i r
5 C
: ~
-. -
-i i-
5
3
H
C
C
r
3 C
: -
-. -
4 »•
3 C
3 C
3 C
H *•
3
3
H
C
\<
3 O C
3 0 C
3 O 0
01
K
D
C
C
c
3 C
3 C
: c
4 C
3 C
3 C
: ^
•j »•
3
3
4
H
c
c
c
3
3
4
tn
>>
a
!2
-H CN <T> r»
•> ^ ^
h 10 i/
3 NO vi
: r-
t-
- O
D C
^
C
3 OO -
H
z
>v C
r
H
4
T
H
D
o
a c
V
S
H
Number
T3
V
tn
•3
^f -f "o c
^
: >o c
H ^-
5 10 C
H *-
3 10 C
H T-
3 '/
H
•>
C
3 u
•; u
•5
n
•5 O r
1— 1
h
c
•^
3 C
H T
3 ^
H
3
C
v
3
^
*
i
.
0
1
t
V
'%> p ~
-3 <u 5
<u j: •-
u -4-j ~
u.
i
1
*!
a
M
— _
3
ft
i^cii ly
spinners
f\\ j
i
i
r,
-. in
J i-
2 c
^
J
_
i.igiiuy
cocooned
*
'
c
1
(
>*
1
1
j
3
j
j
"O
<U 0)
ril
Ma
214
JACK COLVARD JONES
decrease after larvae are lightly cocooned, (5) oenocytoids decrease, and (6)
dividing cells steadily decline.
Out of 56 cases where records were kept, 75% of the dividing hemocytes were
in metaphase, 23.2% in telophase, and 1.8% were in anaphase. Prophases could
not be recognized with the methods used.
2. Total hemocyte counts
Total hemocyte counts (THC) were made daily from the thirteenth through the
twenty-second days of life, from 150 unfixed and from 139 heat-fixed larvae. As
reported in Table II, unfixed THC are more variable (4.9- versus 3.2-fold mean
variation), and were consistently and significantly lower than heat-fixed counts at
greater than the 95% level (that is, twice the standard errors of daily unfixed and
TABLE II
Daily total hemocyte counts with standard errors from unfixed and heat-fixed
Galleria mellonella larvae
Days old
No. used
Cells per microliter
Unfixed
No. used
Heat-fixed
13
5
20,336
± 3360
5
50,448
± 4640
14
11
21,933
± 1804
10
49,686
±3200
15
16
23,704
±2210
10
42,992
± 3480
16
16
31,045
±2635
15
53,504
± 1661
17
22
25,252
± 1531
22
43,316
± 2869
18
16
27,609
± 1880
16
54,042
±4425
19
17
27,053
±2334
16
56,432
±5460
20
15
31,925
± 2779
15
48,724
±3739
21
16
37,744
± 2320
15
61,752
±3515
22
16
35,613
± 3545
15
54,672
±3488
Mean
28,221.4
± 2440
Mean
51,556.8
± 3648
heat-fixed counts did not approach an overlap on any of the days studied). In
both types of counts the numbers of cells per microliter of hemolymph increase
as larvae proceed toward the pupal stadium, and in both the increase is about the
same (I.e., from 17,000 to 18,000 hemocytes/microliter). In both, the most vari-
able counts were obtained on the seventeenth day. The unfixed counts from 20-
to 22-day-old larvae are comparable to the value of 33,200 hemocytes per microliter
reported by Stephens (1963) for larvae which were reared by a different method
and developed much more slowly. However, the consistently and significantly
higher counts from heat-fixed Galleria are definitely not in agreement with Stephens'
statement that heat fixation does not alter the counts.
Unfortunately only a few records were made on the status of the various larvae
during the above period. The records which were obtained, however, were as
follows. In 6 newly emerged larvae the unfixed THC averaged 22,666. During
the spinning of the cocoon, 32,584 cells were found per microliter from 10 unfixed
larvae. Three larvae from light cocoons had 33,067 cells per microliter. Unfixed
HEMOCYTES OF GALLERIA
215
THC from four larvae taken from dense cocoons amounted to 44,620 cells, and
counts from three prepupae came to 17,867 hemocytes. These few data suggest
that the THC increases as larvae spin their cocoons, that there is a further and
greater increase after they complete the cocoon-spinning process, and that the counts
begin to decrease in the prepupae. The counts from densely cocooned larvae and
pharate pupae are strikingly higher than those found by Shrivastava and Richards
(1965), possibly because they used chilled material and excluded the first two
drops of hemolymph.
3. Hemolymph volumes
Hemolymph volume determinations were made from 15- to 22-day-old unfixed
Galleria, all from a single batch of individuals. The larvae were all still feeding
in the medium on the seventeenth day. On the eighteenth day, four out of 5 larvae
had already spun a light cocoon. By the twenty-second day, they had all pupated.
As presented in Table III, the hemolymph volumes when viewed as percentages
TABLE III
Hemolymph volumes of 15- to 22-day-old Galleria mellonella (unfixed)
% Body weight
Calculated microliters
Age in
days
Status
Range
Mean
Range
Mean
15
Feeding
33-36
34.2
36.5-54.0
43.1
16
Feeding
33-36
33.6
63.2-73.6
67.7
17
Feeding
33-36
34.2
48.1-72.2
59.4
18
Lightly cocooned*
28-35
31.6
40.1-53.7
46.7
19
Medium cocoon
29-33
31.4
41.5-7.16
58.1
20
Dense cocoon
19-31
26.0
29.9-57.7
42.4
21
Dense cocoon
16-32
24.8
27.8-46.3
35.8
22
Pupae
16-18
16.4
17.0-24.5
18.8
* Only one of the larvae was beginning to spin a cocoon.
of the body weight remain level at about 34% during the feeding period and then
gradually decrease to less than 16.4% in newly formed pupae. Considered as
microliters, hemolymph volumes of feeding larvae averaged 56.7 and tended to
decline thereafter. Hemolymph volumes of cocooning or cocooned larvae averaged
45.7 microliters. With pupation, the volumes obtained were from less than 17 to
24.5, with a mean of about 19 microliters.
Hemolymph volumes were also made from a subsequent batch of 10 unfixed and
10 heat-fixed larvae of the same age. The hemolymph volumes were identical
(45.5 microliters).
4. Calculated hemocyte populations
The preceding information can be combined to indicate changes in the hemocyte
population within the entire insect. Thus, when THC values are multiplied by the
hemolymph volumes, it can be calculated that, at the 95% level, there are from
831,140 to 2,458,525 hemocytes available in the circulating hemolymph of unfixed
216
JACK COLVARD JONES
larvae from the fifteenth through the twenty-first clays, with means fluctuating
around 1,456,000 (Tahle IV). The daily mean hemocyte population from the
fifteenth through the twenty-first days varied hy a factor of only 1.2-fold. The
data suggest that the circulating hemocyte population in unfixed larvae remains
ahout the same up to pupation itself, at which time there is a very striking and
significant decrease (at the 95% level) so that more than one-half of the hemocyte
population is no longer circulating in unfixed newly formed pupae.
5. Calculated changes in the components o] the hemocyte population
Calculations on the components of the hemocyte population from the fifteenth
through the twenty-second days are presented in Tables V and VI. Assuming that
these data give a reasonable approximation towards the real situation, the following
estimations can be made. (1) During the feeding period, plasmatocytoids aver-
aged 1,490,569 (95% range = 802,050-2,419,189) and during cocooning they aver-
TABLE IV
Calculated circulating hemocyte populations in unfixed 15- to 22-day-old
Galleria mellonella
Age in days
Status
Range at 95% level
Mean
15
Feeding
831,140-1,212,144
1,021,642
16
Feeding
1,744,967-2,458,525
2,101,746
17
Feeding
1,318,086-1,681,852
1,499,969
18
Light cocoon
1,113,748-1,464,932
1,289,340
19
Medium cocoon
1,300,568-1,842,990
1,571,779
20
Dense cocoon
1,117,961-1,589,279
1,353,620
21
Dense cocoon
1,185,123-1,517,347
1,351,235
22
Pupae
536,232- 802,816
669,524
aged 563,308 (95% range = 411,238-816,445), amounting to an average decrease
of 927,261 cells (95% range = 390,812-1,602,744). With pupation, there was a
further average loss of 185,000 cells (95% range == 183,876-186,125), amounting to
an average total loss of 1,112,261 plasmatocytoids (95% range = 574,688-1,788,869)
as Galleria larvae transform into pupae. (2) From 520,000 to 684,123 immature
and 145,900 to 192,000 matured adipohemocytes appeared in the hemolymph of
lightly cocooned larvae, a range of 666,000 to 876,000 (average 771,025) cells
containing lipid inclusions. Immature adipohemocytes decreased by 439,627-569,695
(average 504,661) cells within the densely cocooned larval insect. With pupation,
there was a further decrease of 78,348-111,217 cells (average 184,996), thus
amounting to 517,975-680,912 immature adipohemocytes deleted during the larval-
pupal molt and ecdysis. (3) After mature adipohemocytes first appeared in the
hemolymph, they increase to a maximum of 955,929. Between the eighteenth and
twenty-first days, 600,726-764,023 (average 682,375) mature adipohemocytes were
formed, on each of which days they increased by 151,000-332,000. After pupation,
440,000-498,000 adipohemocytes were no longer circulating. (4) During larval
life, spherule cells averaged 16,950 and they were not observed in the pupae
examined. They reached a maximum on the sixteenth day, that is, before the
HEMOCYTES OF GALLERIA
217
TABLE V
Calculated mean changes in the components of the hemocyte population of unfixed
15- through 22-day-old Galleria mellonella
Numbers of circulating hemocytes
Age in days
Adipohemocytes
tlllCl St&tUS
Plasmatocytoids
Spherule
cells
Oenocytoids
Dividing
hemocytes
Immature
Mature
15 Feeding
985,884
0
0
15,325
18,389
3,065
15 Feeding
1,001,209
0
0
10,012
10,012
4,597
16 Feeding
2,007,167
0
0
56,747
35,730
4,203
16 Feeding
2,068,118
0
0
21,017
12,610
5,254
17 Feeding
1,390,471
0
71,998
7,499
28,499
9,000
18 Light cocoon
493,817
602,122
168,903
7,736
16,761
3,868
19 Dense cocoon
696,298
394,516
446,385
12,574
22,005
1,572
20 Dense cocoon
594,239
97,461
630,787
13,536
17,597
667
2 1 Dense cocoon
468,878
— -
851,278
8,107
22,971
1,081
22 Pupae
283,878
2,678
382,298
0
669
0
larvae began to spin a cocoon. (5) Oenocytoids fluctuated from 10,012 to 35,730
during larval life, with an overall mean of 20,508 and, like the spherule cells,
attained a maximum on the sixteenth day. (6) From 559 to 10,091 hemocytes
apparently divide in the hemolymph from the fifteenth through the twenty-first days
of life, the greatest number being present on the seventeenth day. Since it is not
known whether mitotic divisions occur throughout the day and since the duration
of the mitotic cycle is unknown, it is not possible to make correlations between
mitoses and changes in the hemocyte population.
In their radioautographic study, Shrivastava and Richards (1965) showed that
plasmatocytes of Galleria transform into adipohemocytes within 24 hours, and the
present hemocyte population calculations were examined to see if the changes in
the population of plasmatocytoids could be correlated with adipohemocyte popula-
TABLE VI
Calculated ranges at 95% level of plasmatocytoids and adipohemocytes
Adipohemocytes
Age in days and status
Plasmatocytoids
Immature
Mature
15 Feeding
802,050-1,169,719
0
0
15 Feeding
814,517-1,187,901
0
0
16 Feeding
1,666,443-2,347,891
0
0
16 Feeding
1,717,047-2,419,189
0
0
17 Feeding
1,221,866-1,559,077
0
58,170- 74,836
18 Light cocoon
426,565- 561,069
520,120-684,123
145,901-191,906
19 Dense cocoon
576,152- 816,445
326,443-462,590
369,361-523,409
20 Dense cocoon
490,785- 697,693
80,493-114,428
520,970-740,604
21 Dense cocoon
411,238- 526,519
746,627-955,929
22 Pupae
227,362- 340,394
2,145- 3,211
306,188-458,408
218
JACK COLVARD JONES
tion changes. The data in Table VII show the changes in the populations of
plasmatocytoids and of immature and mature adipohemocytes in terms of the 95%
ranges with the means in parentheses. ( 1 ) There is no correlation of changes in
the two populations between the sixteenth and seventeenth days (that is, 7.5 to 10.5
times more plasmatocytoids disappear than adipohemocytes appear). (2) If it is
assumed that all of the mature adipohemocytes already in circulation on the seven-
teenth day (71,998 cells) remain in circulation on the eighteenth day, then 96,905
new adipohemocytes would need to be formed from plasmatocytoids. Between the
seventeenth and eighteenth days, 896,654 plasmatocytoids disappeared and 771,025
adipohemocytes appeared which leaves a deficit of 125,629 plasmatocytoids unac-
counted for in a population of 1,264,842 (an error of 10%). This is interpreted
to mean that many (about 64%) plasmatocytoids transform into immature and
mature adipohemocytes between the seventeenth and eighteenth days. (3) Between
the eighteenth and nineteenth days the plasmatocytoids increased by 202,481 cells,
the immature adipohemocytes decreased by 207,606 cells, while mature adipohemo-
cytes increased by 277,482 cells. If all of the mature adipohemocytes of the eight-
TABLE VII
Estimated increases and/or decreases in populations of plasmatocytoids, immature and mature
adipohemocytes in unfixed Galleria mellonella. Ranges at 95% level; means in parentheses
Between
days
Plasmatocytoids
Adipohemocytes
Immature
Mature
16-17
17-18
18-19
19-20
20-21
-495.181 to 860,112 (-677,647)
-795,301 to 998,008 (-896,654)
+ 149,587 to 255,376 (+202,481)
- 85,367 to 118,752 (-102,059)
- 79,547 to 171,174 (-125.361)
+520,120 to 684,123 (+602,122)
-193,677 to 221,533 (-207,606)
-245,950 to 348,162 (-297.055)
+ 58,170 to 74,836
+ 87,731 to 117.070
+223,460 to 33 1,503
+ 151,609 to 217,195
+215,325 to 225,657
(+ 71,998)
(+ 96.905)
(+277,482)
(+184,402)
(+220,491)
eenth day remained in circulation on the nineteenth day, then 277,482 new mature
adipohemocytes would need to be formed. If all 207,606 immature adipohemo-
cytes which disappeared from the circulation between the eighteenth and nineteenth
days were transformed into mature adipohemocytes, this would leave only 69,876
mature adipohemocytes unaccounted for on the nineteenth day. This appears to
be an excellent correlation and implies that about 34% of the 168,903 immature
adipohemocytes in circulation on the eighteenth day transform into mature cells by
the nineteenth day. (4) Between the nineteenth and twentieth days the populations
of both plasmatocytoids and immature adipohemocytes appear to decrease simul-
taneously and far more immature adipohemocytes disappear than new ones form.
No correlations could be detected then between the various hemocvtes between the
j
nineteenth and twentieth days. (5) Between the twentieth and twenty-first days
the plasmatocytoids decreased by 125,361 cells and mature adipohemocytes increased
by 220,491 cells. If all 530,787 mature adipohemocytes of the twentieth day re-
mained in circulation and all of the circulating immature adipohemocytes trans-
formed into mature cells by the twenty-first day, this would still leave 123,030 ma-
ture adipohemocytes unaccounted for. If most of the 125,361 plasmatocytoids
which disappeared between the twentieth and twenty-first days transformed into
HEMOCYTES OF GALLERIA 219
mature adipohemocytes, this would account for the deficit and make an almost
perfect correlation.
Considering the many sources of error in calculations such as these, it is remark-
able that it was possible to make any correlations, and impressive that three out of
five of them appear so close. These correlations suggest ( 1 ) that many plasmato-
cytes transform into both immature and mature adipohemocytes between the
seventeenth and eighteenth days when larvae are spinning a cocoon, (2) that
between the eighteenth and nineteenth days, when the larvae are cocooned, mature
adipohemocytes are largely being formed by maturation of immature adipohemo-
cytes, and (3) that in pharate pupae mature new adipohemocytes are being formed
from both immature adipohemocytes and from plasmatocytoids between the twen-
tieth and twenty-first days.
DISCUSSION
In Prodenia larvae Yeager (1945) recognized and counted separately adipo-
hemocytes (his "spheroidocytes") and granular hemocytes (his "cystocytes").
Jones (1959) pointed out that when the adipohemocytes of Prodenia matured they
closely resemble the granular hemocytes. Yeager (1945) suggested that the adi-
pohemocytes were derived from prohemocytes and, since he observed mitoses
among adipohemocytes, they might also be considered as a self-perpetuating line
of cells in this insect. He suggested that the granular hemocytes of Prodenia were
derived from plasmatocytes. He observed mitoses among granular hemocytes,
though less commonly than in the plasmatocytes and adipohemocytes. In Boinbyx,
Nittono (1960) apparently combined Yeager's cystocytes and spheroidocytes into
a common category which he designated granular hemocytes. Earlier, Jones
(1959) had suggested that Yeager's "cystocytes" were comparable to the granular
hemocytes of other insects. In some insects, cells termed granular hemocytes are
quite distinct from both plasmatocytes and adipohemocytes : for example, in the
blood-sucking bug, Rhodnius prolixus (Jones, 1965), the granular hemocytes pos-
sess many uniform discrete inclusions and are not derived from plasmatocytes and
are not related to them. In Sarcophaga, changes in the population of cells termed
granular hemocytes (Jones, 1956) cannot be correlated with changes in the popula-
tion of plasmatocytes (Jones, unpublished data). The "granular hemocytes" of
Boinbyx are present in large numbers in one- to three-day-old larvae of the fifth
stage and they were frequently observed in mitotic division (p. 262) by Nittono.
The data in Nittono's Tables 4 and 6 were combined so that estimates could be
made of the components of the hemocyte population with time in both males and
females of the fifth stage. No correlations at all could be found between the popu-
lations of plasmatocytoids and adipohemocytes at any time. Dr. Nittono (personal
communication) has confirmed this. Can the granular hemocytes of Boinbyx and
Sarcophaga which are present throughout larval life and which do not appear to
be derived from plasmatocytes be compared with the adipohemocytes of Galleria
when the latter appear only near the end of larval life, do not divide, and are derived
from plasmatocytes ? Granular hemocytes and adipohemocytes may both be phago-
cytic and yet very different in their origins. There is no doubt that the granular
hemocytes in Rhodnius are not comparable morphologically or physiologically to
the adipohemocytes of Galleria. Until considerably more information is available
220 JACK COLVARD JONES
concerning the granular hemocytes and adipohemocytes, separate terms should be
retained. From the evidence now available it would seem that granular hemo-
cytes of some insects are not derived from plasmatocytes whereas adipohemocytes
of a number of the Lepidoptera arise by direct transformation of circulating plasma-
tocytes.
In Prodenia, plasmatocytoids decline from 86. 3 (/o in the first instar larvae to
34.2% in prepupae (calculations from Yeager's data, 1945). Granular hemocytes
appeared first in fourth-stage larvae and increased to 28% just before pupation
(Yeager, 1945). Adipohemocytes increased from 2.5% in first stage larvae to
38.6% in prepupae. Spherule cells reached their maximum (43.4%) in third-stage
larvae and declined to 6% the day before pupation (Yeager, 1945). During the
last three days before pupation of Prodenia, plasmatocytoids decreased from 41.7%
to 17.8%, adipohemocytes increased from 33.9% to 43.9%, granular hemocytes
increased from 5.2% to 28%, and oenocytoids decreased from 5.1% to 1.8%
(Yeager, 1945).
In last-stage Boniby.v larvae, plasmatocytoids reached a peak of 67.3% on the
seventh day and declined to 27.3% just before pupation. In the J 122 X C strain,
the granular hemocytes averaged 53.4% from the third through the fifth larval
stages and definitely increased near the end of each stadium. During the first
eight days of the last larval stage they averaged 45.7%, and during the last four
days they averaged 64.1% (Nittono, 1960).
Galleria resembles Prodenia and Boinhy.v in that plasmatocytoids decrease and
that hemocytes with many polysaccharide and/or lipid or other types of inclusions
increase prior to pupation. Galleria differs significantly from Prodenia and Bom-
byx in that their hemocytes with many polysaccharide and/or lipid or other types
of inclusions do not appear in the hemolymph for the first five days of the last
larval stadium. The hemocytes with lipid inclusions in Bombyx apparently are
not derived from plasmatocytes (at least no correlations between changes in these
two components of the population were detectable), whereas in Galleria there is
radioautographic evidence that hemocytes with lipid inclusions are derived from
plasmatocytes, and in three out of five cases it \vas possible to detect reasonably-
close reciprocal correlations between the changes in these two cell types.
SUMMARY
1. The hemocytes of Galleria mellonella (Linnaeus) larvae were identified and
differentially counted in unfixed hemolymph with phase microscopy. The numbers
of hemocytes per microliter of hemolymph were obtained from both unfixed and
heat-fixed larvae. Hemolymph volumes were determined by the amaranth red
method. These studies were made to determine what changes in the hematology
occur as the last stage larvae pass through distinctive phases in transforming into
pupae.
2. In differential counts, plasmatocytoids decrease, immature adipohemocytes
suddenly appear, and mature adipohemocytes steadily increase. Spherule cells,
oenocytoids and dividing hemocytes decrease as Galleria larvae develop into pupae.
3. The numbers of hemocytes per microliter of hemolymph increase as Galleria
larvae proceed towards the pupal stage in both unfixed and heat-fixed animals.
Counts were always significantly higher in heat-fixed than in unfixed larvae.
HEMOCYTES OF GALLERIA 221
4. The hemolymph volume is the same in both unfixed and heat-fixed larvae.
The hemolymph volume declines from about 34% (56.7 microliters) in precocoon-
spinning larvae to less than 16.4% (19 microliters) in newly formed pupae.
5. It is estimated from the various data presented that an average of 1,456,000
hemocytes remain in circulation within the hemocoele of unfixed larvae from the
fifteenth through the twentieth days of life, and that with pupation more than one-
half of these cells fall out of circulation.
6. In three out of 5 cases it was possible to correlate decreases in the plasma-
tocytoid population with increases in adipohemocytes. It is suggested that during
the spinning of a cocoon plasmatocytoids transform into both immature and mature
adipohemocytes, that when the larvae are densely cocooned mature adipohemocytes
are largely formed by the maturation of immature adipohemocytes, and that in
pharate pupae new mature adipohemocytes are derived from both immature adipo-
hemocytes and plasmatocytoids.
7. The hemocyte picture of Galleria is compared to that of Prodenia and
Bombyx. In all three of these Lepidoptera the plasmatocytoids decrease and the
hemocytes with many polysaccharide and/or lipid or other types of inclusions in-
crease prior to pupation. Galleria differs from the other two species in that their
hemocytes with lipid or other inclusions do not appear until about the sixteenth or
seventeenth days of larval life, do not divide, and in many cases are derived from
circulating plasmatocytes.
LITERATURE CITED
ASHHURST, D. C, AND A. G. RICHARDS, 1964. Some histochemical observations on the blood
cells of the wax moth, Galleria mellonella L. /. Morph., 114: 225-254.
BECK, S. D., 1960. Growth and development of the greater wax moth, Galleria mellonella (L.)
(Lepidoptera: Galleriidae). Wisconsin Acad. Sci., Arts and Letters, 49: 137-148.
JONES, J. C, 1956. The hemocytes of Sarcophaga bullata Parker. /. Morph., 99: 233-257.
JONES, J. C., 1959. A phase contrast study of the blood-cells in Prodenia larvae (Order Lepi-
doptera). Quart. J. Micr. Sci., 100: 17-23.
JONES, J. C., 1962. Current concepts concerning insect hemocytes. Amer. Zool., 2: 209-246.
JONES, J. C, 1965. The hemocytes of Rhodnius prolixus Stal. Biol. Bull, 129: 282-294.
NITTONO, Y., 1960. Studies on the blood cells in the silkworm, Bombyx mori L. Bull. Sericult.
Exp. Sta., 16: 171-266.
SHRIVASTAVA, S. C., AND A. G. RICHARDS, 1965. Au autoradiographic study of the relation be-
tween hemocytes and connective tissue in the wax moth, Galleria mellonella L. Biol.
Bull., 128: 337-345.
STEPHENS, J. M., 1963. Effects of active immunization on total hemocyte counts of larvae of
Galleria mellonella (Linnaeus). /. Ins. Path., 5: 152-156.
YEAGER, J. F., 1945. The blood picture of the southern annyworm (Prodenia cridania). J.
Agric.Res.,71: 1-40.
YEAGER, J. F., AND S. C. MUNSON, 1950. Blood volume of the roach Pcriplaneta americana
determined by several methods. Arthropoda, 1 : 255-265.
THE EFFECT OF LIGHT ON THE SPAWNING
OF CIONA INTESTINALIS
CHARLES C. LAMBERT 1 AND CHARLES L. BRANDT
Biology Department, San Diego State College, San Diego, California
Invertebrate embryologists have long known that a number of ascidians spawn
in response to light following darkness. Molgula manhattensis (Castle, 1896;
Conklin, 1905) and dona intestinalis (Castle, 1896; Conklin, 1905; Berrill, 1947)
normally spawn at dawn but can be induced to spawn at any time by keeping them
in the dark until needed ; then a short exposure to light causes them to spawn
(Costello et al., 1957). Styela partita spawns during the late afternoon (Castle,
1896; Conklin, 1905; Rose, 1939). Rose (1939) found that 5". partita could be
induced to spawn at any time by placing them in the dark for 12 hours, then sub-
jecting them to light for 11-12 hours, at the end of which time they spawn.
The physical factors controlling spawning in Corella par all elo gramma have been
extensively investigated by Huus (1939, 1941a, 1941b). This ascidian, which
normally spawns during the early morning, can be caused to spawn at any hour
by exposing dark-adapted animals to the light of a 60-candle bulb 25 cm. from
the aquarium for 2 minutes (Huus, 1939). Spawning begins within 30 minutes.
Huus termed this period between illumination and spawning the "dormant pe-
riod." Limiting temperatures for spawning were found to be 10°-24° C. (Huus,
1941a). The duration of the dormant period was determined to be temperature-
dependent (1941b), 11 minutes being required at 24° C. and 17 minutes at 10.5° C.
Huus hypothesized that light causes spawning by eliciting the production of some
unknown hormone; the temperature dependency of the dormant period, he stated,
tended to support this view.
The present study, on light-induced spawning by C. intestinalis, consists of
two series of experiments. The first series, using unmeasured white light, deter-
mined the minimum reliable dark-adaption time and the time required for spawning
after illumination. The second series, using quantified monochromatic light, deter-
mined the threshold dose of light energy necessary to cause spawning at different
wave-lengths. From these data an action spectrum for spawning is constructed.
MATERIALS AND METHODS
Experimental animals
dona intestinalis between 4 and 7 cm. in overall length were collected in Mission
Bay, San Diego, California. Only gravid individuals, identified by their full ovi-
ducts, were used in the experiments. Continuous illumination from the time of
collection until the dark-adaption period prevented uncontrolled spawning. The
animals were used only once, two days after collection.
1 Present address : Department of Zoology, University of Washington, Seattle, Washington
98105.
222
SPAWNING OF CIONA INTESTINALIS
Experimental apparatus and procedure
The preliminary experiments used white light from two Sylvania 40-watt day-
light fluorescent tubes 147 cm. above the animals. For these experiments, the
animals, each in a 400-ml. beaker of sea water, were placed on a 15°-17° C. water
table. Dark-adaption for periods of 45 to 60 minutes was accomplished by cover-
ing the water table. The animals were either continuously illuminated until
spawning occurred, or they were returned to darkness after a one-minute exposure
to light.
A Bausch and Lomb high intensity, grating monochrometer (1350 grooves/
mm.), equipped with a tungsten (quartz-iodine) 45-watt lamp as light source, was
used for monochromatic illumination. The band pass was 10 m^u. The beam from
the exit slit (exit lens removed) passed through a leaf shutter, then through the
end of the aquarium (10 cm. from monochrometer) to the animal. An approxi-
mately circular spot of light, 3 cm. in diameter, was formed on the animal by the
exit beam at this distance.
The monochromatic light intensity was measured with phototube C of a Photo-
volt electronic photometer, model 501 -M, placed in the same position relative to
the monochrometer as had been the end of the aquarium. The photometer was
calibrated at all wave-lengths used against a calibrated Reeder compensated vacuum
thermopile (RBL-500) and a Leeds and Northrup (22S4b) high sensitivity gal-
vanometer. A National Bureau of Standards 50-watt 115v. secondary Radiometric
Standard Lamp was used to calibrate the thermopile-galvanometer system according
to the method contained in form NBS-443.
The animals were suspended with nylon string, basal ends upward, in indi-
vidual 600-ml. beakers of sea water at 16°-17° C. After one hour of dark-adaption
the animals were removed from their beakers, placed in the experimental aquarium
and illuminated with monochromatic light, one at a time. While being illuminated,
the animals were pressed flat against the aquarium end by a flat flask. Since the
circle of illumination was only 3 cm. in diameter, only the siphonal end of the
animals received the light. After illumination, the animals were returned to their
beakers. The beakers were examined for ova 30 minutes after the last animal had
been illuminated. Those beakers that contained ova were recorded as positive.
Those that did not contain ova were brought into the well-lit laboratory (fluorescent
lighting) for 30 minutes, after which time they were again examined for the pres-
ence of ova. Beakers that now contained ova but which had not on removal from
the darkroom were recorded as negative, that is, the animals were capable of
spawning but had not been provoked to spawn by the amount of light energy
received.
The duration of exposure at the maximum intensity of a given wave-length
necessary to provoke spawning in two out of three animals, when a 10% shorter
exposure would not elicit spawning by two out of three animals, was taken as the
threshold duration of illumination for spawning at that wave-length. This threshold
duration was determined for wave-lengths between 400 m//, and 610 m/* in 15-m/A
increments. These values were then converted to threshold doses in quanta/mm2.
An action spectrum was constructed by graphing the reciprocal of the threshold
dose against wave-length.
224 CHARLES C. LAMBERT AND CHARLES L. BRANDT
RESULTS AND DISCUSSION
While light studies
Dark-adaption periods of 45-55 minutes followed by return to light resulted
in spawning by 26 out of 65 animals (40%). One hour of dark-adaption preceding
illumination elicited spawning in 47 of the 60 animals tested (78.3%). Following
the one-hour dark-adaption period, an average of 27.3 minutes elapsed before
spawning occurred. The one-hour dark-adaption period was used for all of the
following experiments.
A comparison of these results with dona and Huus's with Corolla demon-
strates clearly that these two ascidians have quite similar spawning responses to
light, the main difference being duration of the latent (dormant) period: at
14.5° C., Corella spawned 14.5 minutes after exposure to light; dona spawned
27 minutes after exposure at 15°-16° C.
The observation that Ciona spawns at dawn in the laboratory is an old one
(Castle, 1896; Conklin, 1905; Berrill, 1947), yet the most recent paper on the
spawning of this ascidian (Carlisle, 1951) curiously omits any reference to the
light-induced spawning of any ascidian. Carlisle (1951) was investigating the
spawning of Ciona intestinalis and Phallusia mainmilata in relation to two other
factors: the effect of injecting human chorionic gonadotropin and the effect of
ingesting gametes. Carlisle, without discussing the illumination of his laboratory,
stated that Ciona was never observed to spawn "spontaneously." We are quite
confident that in spite of the small number of animals involved in his studies (less
than 60), had his laboratory ever been darkened, spawning would have occurred,
provided the animals were ripe. Carlisle reported that either injection of chorionic
gonadotropin or ingestion of gametes provoked spawning in these two ascidians.
This spawning took place 20 hours after treatment, in contrast to the 27-minute
latent period established here. Carlisle further states that prior to treatment, no
corpora lutea were observed in dona's ovary. Millar's (1953) report that the
oviduct is always packed with ova prior to spawning has been fully confirmed by
our observations. Although the histological structure of the ovary was not exam-
ined in this study, the presence of ova in the oviduct implies that corpora lutea
should be found in the ovary. A re-evaluation of Carlisle's findings may be made
in the light of the observations reported here. Perhaps Carlisle did not provoke
spawning by his treatments, but instead induced ovulation. These two phenomena,
as demonstrated by the full oviduct prior to spawning, are quite separate in dona.
It should be stated here that Huus (1941a) found that Corella, prior to spawning,
has an empty oviduct, which suggests that in Corella spawning and ovulation are
either simultaneous or occur closer in time than in Ciona.
Monochromatic light studies
The action spectrum for spawning of C. intestinalis was obtained by illuminating
the animals at different wave-lengths and determining the threshold duration of
exposure required to evoke spawning at the maximum intensity of each of these
wave-lengths. Since the intensity of the incident beam at each wave-length was
known, the quantum requirement (the threshold dose for spawning) was easily
calculated.
SPAWNING OF CIONA INTESTINALIS
225
Since the intensity at each wave-length was different, it is possible that the
quantum requirement, determined on the basis of duration of exposure, might have
been different if the Reciprocity Law does not hold for some intensities used.
However, since the maximum difference in intensity between any two wave-lengths
was less than four times (Table I), and since the animals were most sensitive to
the wave-lengths showing the lowest intensity, this problem probably does not
seriously influence the shape of the action spectrum. Another drawback to this
method which became evident as the experiments progressed was that for a reason-
able exposure time (5 minutes) the energy output of the monochrometer was too
low in the red end of the spectrum to cause spawning. This fact also made it im-
possible to test for reciprocity, i.e., Intensity X Time equals a Constant Response,
at each wave-length used.
TABLE I
Experimental and derived data necessary to establish spawning threshold in quanta
Wave-length
m/i
Intensity
Threshold
Ergs/sec, /mm.2
Quanta X 10'V
sec. /mm.2
Duration
sec.
Dose
Quanta X lOu/mm.2
610
17.64
5.41
660
35.7
595
18.59
5.56
498
27.7
580
14.14
4.12
360
14.9
565
14.62
4.15
88
3.67
550
15.54
4.30
44
1.89
535
14.30
3.86
72
2.77
520
14.00
3.66
66
2.42
505
10.92
2.77
216
6.05
490
13.26
3.27
577
18.9
475
11.52
2.76
378
10.4
460
10.12
2.34
570
13.4
445
8.28
1.86
478
8.89
430
9.90
2.07
56
1.16
415
4.80
1.00
60
0.60
400
4.86
0.978
144
1.41
Table I presents the raw and derived data necessary to obtain the threshold
quanta requirements for spawning at all wave-lengths tested.
Of the 884 animals used in this study, 589 (66.6%) spawned in response to
light, either after illumination by monochromatic light or after return to the illumi-
nated laboratory.
Action spectrum for spa^vnin(J
Figure 1 is an action spectrum for photically induced spawning by dona
intestinalis. The reciprocals of the quantum thresholds from Table I are plotted
against wave-length to show the relative effectiveness of light of each wave-length
in inducing spawning. As can be seen from this figure, there are three peaks of
maximum effectiveness. Wave-length 415 in/A is most effective, requiring a dose
of light about one-third that of the next most effective wave-length, 550 m/x, to
226
CHARLES C. LAMBERT AND CHARLES L. BRANDT
induce spawning. Wave-lengths 520 m/x, and 550 in/x, are of nearly equal effective-
ness. This action spectrum for spawning by Ciuna inlestinalis suggests that a
hemoprotein is the light-absorber because of the great efficiency in the region of
the Soret band absorption and the characteristic peaks in the yellow. An ex-
amination of the absorption spectra of the hemoproteins led to cytochrome c as
a possible chromophore.
In Figure 2 the action spectrum for spawning in Ciona is replotted as the
Relative Effectiveness in Inducing Spawning as a function of wave-length. These
data are obtained by setting the reciprocal of the threshold dose in quanta/mm.2
at wave-length 415 m/x, equal to 100% Relative Effectiveness. The doses at all
other wave-lengths are then reduced to a percentage of the dose at 415 m/x.. On the
580
610
Wavelength
t
FIGURE 1. Action spectrum for light-induced spawning of Ciona intcstinalis
same figure (Fig. 2) are plotted data on the Relative Optical Density of reduced
horse heart cytochrome c. These data are calculated from those obtained by Mar-
goliash and Frohwirt (1959) by setting the optical density at wave-length 415 m/z
equal to 100% Relative Optical Density. At all other wave-lengths, the Relative
Optical Density is calculated as a percentage of the optical density at 415 m/x..
Comparison of these two curves shows that they are similar in many respects.
Oxidation of cytochrome c results in the following changes in its absorption spec-
trum: the major peak at 415 m/x, shifts to 410 m/x, and is lowered considerably, and
the peaks at 520 m/x, and 550 m/x, are replaced by a single peak at 528 m/x, ( Margo-
liash and Frohwirt, 1959). It is evident, therefore, that if cytochrome c is the
chromophore, it is in the reduced state. The maxima and minima of the action
spectrum fit quite well with those of the absorption spectrum. It will be seen,
however, that although the heights of the action spectrum maxima are of the same
relative order (415 m/x, > 550 m/x, > 520 m/x,) as those of the absorption spectrum,
the relative heights at 550 m/x, and 520 m/x, are different for the two spectra. While
the action spectrum for spawning closely matches the reduced cytochrome c absorp-
SPAWNING OP CIONA INTESTINALIS
227
tion spectrum, the resolution attained by our system is not sufficient to do more
than suggest that cytochrome c, or some other hemoprotein, may be the receptor
material.
The role of hemoproteins in photobiological processes has been extensively
investigated by Arvanitaki and Chalazonitis (1949, 1960, 1961). These workers,
studying the effect of monochromatic light on the visceral ganglion of the gastropod
Aplysia, have demonstrated that two chromophores are involved in light reception
as measured by the electrical activity of isolated neurons. These pigments seem
to act in antagonistic ways upon absorption of light. One pigment, a carotene-
protein, generally produces a hyperpolarization of the membrane potential and
c
5
o
o» c
o
•o
c
in
O
«
400
490
Wavelength
580
610
in
FIGURE 2.
A comparison of the absorption spectrum of cytochrome c (solid line) with the
action spectrum for spawning of Ciona intestinalis (dashed line).
inhibition of spiking. The other pigment, a heme-protein, produces a membrane
depolarization and the initiation of spiking. The pigments are contained in granules
just below the plasma membrane of the nerve cells, imparting a reddish hue to
the cells. It is hypothesized (Chalazonitis, 1964) that the heme-protein, upon
absorbing light, may pass an electron to the carotene-protein, thereby acting as a
photoconductor. This transfer of electrons within the membrane is then visualized
as opening channels for ionic flow. Thus a generator current is initiated which,
if of sufficient intensity, may initiate action potentials. It is tempting to suggest
that light absorbed by heme-proteins in Ciona may trigger a similar chain of events
leading eventually to spawning. This, of course, implies absorption of light and
action at the neuronal level. While it is true that the neural ganglion and numerous
nerves of Ciona were illuminated in these experiments, other pigmented structures
such as the tip of the gonopore and the neural gland, also received light. Studies
are now under way to attempt a localization of the light absorbers and to investi-
gate the neurophysiology of this response. Since the visceral ganglion is found
deeply buried in the viscera of the intact Aplysia, it is extremely unlikely that light
CHARLES C. LAMBERT AND CHARLES L. BRANDT
can reach it to cause a behavioral response in such an animal. It is possible that
the work reported here on the action spectrum for spawning of dona is the first
demonstration of a hemoprotein involvement in a photo-induced behavioral response
by any animal.
SUMMARY
1. The spawning of dona intestinalis with respect to light was studied, using
both white light and monochromatic light.
2. A one-hour dark-adaption period followed by exposure to light resulted in
spawning by 66.6% of the 884 animals tested.
3. Spawning occurs an average of 27.3 minutes after the onset of illumination.
4. Illumination need not be continuous until spawning occurs; the animals
spawn when returned to the dark after a short illumination period, provided they
have received enough energy.
5. The action spectrum for spawning suggests cytochrome c as a chromophore.
LITERATURE CITED
ARVANITAKI, A., AND N. CHALAZONITIS, 1949. Reactions bioelectriques neuroniques a la photo-
activation specifique d'une heme-proteine et d'une carotene-proteine. Arch. Sci. Physiol.,
3: 27-44.
ARVANITAKI, A., AND N. CHALAZONITIS, 1960. Photopotentiels d'excitation et d'inhibition de
defferents somata identifiables (Aplysia) activations monochromatiques. Bull. Inst.
Occanog. (Monaco), 57: No. 1164, 83 pp.
ARVANITAKI, A., AND N. CHALAZONITIS, 1961. Excitatory and inhibitory processes initiated by
light and infra-red radiations in single identifiable nerve cells. In : Nervous inhibition :
Proceedings of the Second Friday Harbor Symposium, Ernst Florey, Ed., Pergamon
Press, New York, pp. 194-231.
BERRILL, N. J., 1947. The development and growth of Ciona intesfinalis. J. Mar. Biol. Assoc.,
26: 616-625.
CARLISLE, D. B., 1951. On the hormonal control of the release of gametes in ascidians. /. Exp.
Zoo/., 28:463-472.
CASTLE, W. E., 1896. The early embryology of Ciona intestinalis Flemming (L.). Bull. Mus.
Comp. Zoo/., Harvard, 27: 201-280.
CHALAZONITIS, N., 1964. Light energy conversion in neuronal membranes. Photochcm.
Photobiol.,3: 539-559.
CONKLIN, E. G., 1905. The organization and cell lineage of the ascidian egg. /. Acad. Nat.
Sci., Philadelphia, 13: 1-119.
COSTELLO, D. P., ET AL., 1957. Methods for Obtaining and Handling Marine Eggs and Em-
bryos. Mar. Biol. Lab., Woods Hole, Mass.
Huus, J., 1939. The effect of light on the spawning in ascidians. Avhandlinger utgitt ar Dei
Norskc yidenskaps-Akadcini I Oslo I. Mat.-Naturv. Klassc, No. 4: 5-49.
Huus, J., 1941a. Effects of physical factors on the spawning in ascidians. Temperature limits
for spawning. Avhandlinger utgitt av Det Norske Videnskaps-Akademi I Oslo I.
Mat.-Naturv. Klasse, No. 8: 1-13.
Huus, J., 1941b. Effects of physical factors on the spawning in ascidians. Temperature and
latent period. Avhandlinger utgitt av Dct Norskc I'idenskaps-Akadcmi I Oslo I.
Mat.-Naturv. Klasse, No. 9: 2-12.
MILLAR, R. H., 1953. Ciona. L. M. B. C. Memoir XXXV. University Press, Liverpool.
MARGOLIASH, E., AND N. FROHWIRT, 1959. Spectrum of horse heart cytochrome c. Biochcin.
J., 71: 571-572.
ROSE, S. MERYL, 1939. Embryonic induction in ascidia. Biol. Bull., 77: 216-232.
GENETIC AND DEVELOPMENTAL STUDIES ON
BOTRYLLUS SCHLOSSERI *
ROGER MILKMAN
Marine Biological Laboratory, Woods Hole, Massachusetts 02543 and Department of Zoology,
Syracuse University, Syracuse, New York 13210
The colorful compound ascidian, Botryllus schlosseri, has great promise for
investigation in several important theaters of genetics, notably development and
natural variation. In order for its potential to he realized as an experimental
animal in these areas, a variety of preliminary studies have been undertaken. The
results of these studies are reported here.
Botryllus has been studied extensively by Bancroft (1903), Berrill (1941a,
1941b, 1951, 1961), Oka and Watanabe (1957, 1959, 1960), and Sabbadin (1958.
1959, 1960, 1962, 1964), as well as by Scott (1934), Watterson (1945), and
others. After Bancroft's early work on its natural history, development, species
structure, and its property of colony fusion, Berrill used Botryllus as one major
object of study in his broad and highly important series of contributions on devel-
opment. More recently, Oka and Watanabe (on Botryllus primly enus and Botryl-
loides) and Sabbadin have addressed themselves to additional developmental prob-
lems, as well as to the genetic analysis of pigmentation and compatibility.
Although Botryllus is well described in the literature (Berrill. 1950 ; Van Name.
1945) and is exceedingly common just below the low water mark on pilings, eel
grass, and under rocks, particularly in harbors, it is not a familiar organism even
to many marine biologists, and a brief description is therefore in order. Botryllus
colonies are of irregular shape and may be well over a foot in diameter, though
usually much smaller. Each colony (Fig. 1) is composed of rosette-like systems
of generally 5-18 sooids, each of which is like a solitary ascidian in form. The
zooids, together with a vascular system which pervades them and the areas between
•and around systems, and which consists of blood vessels and ampullae (Fig. 2), are
embedded in a gelatinous matrix which is maintained in a dynamic state by the
activities of numerous amoeboid cells. The zooids' long axes are radially arranged
in the systems. Their oral (incurrent) siphons are peripheral and open directly
to the water ; their atrial siphons open into the system's common atrial chamber
which in turn communicates with the outside via a common atrial opening. The
concentration of hydraulic force thus permits the powerful ejection of fecal pellets
(and sperm) ; accordingly the system may be thought of primarily as a unit of
egestion.
The oozoid resulting from the metamorphosis of a tadpole-type larva initiates
the asexual formation of a colony by budding. Throughout the life of the colony,
budding is synchronous, and when the buds become functional zooids, the previous
1 This was was supported by Research Grant GM 07810 of the National Institute of General
Medical Sciences, United States Public Health Service.
229
230
ROGER MILKMAN
System
Gelatinous test
Ampulla.
Blood vessel
Zooid
FIGURE 1. Habit sketch of B. schlosscri on glass.
generation of zooids is resorbed. Beginning with the oozooid, zooid size and pig-
mentation increase in each of the first five or ten asexual generations. During this
period, first functional sperm and finally mature eggs make their appearance. The
buds are produced at specific sites on the atrial wall, one per zooid at first, and later
up to four. The dependence of functional gonads upon a certain zooid size sug-
gested, and specific surgical experiments (Berrill, 1961) confirmed, that the degree
of differentiation is dependent on mass in a manner reminiscent of the findings of
Lopaschov (1935) and Grobstein and Zwilling (1953) in frog, chick, and mouse.
Colonies under suboptimal conditions may mark time or even regress while the
sequence of budding and resorption continues.
BOTRYLLUS GENETICS AND DEVELOPMENT
231
— Gelatinous test
taking over
regressing
~~ f
^ VBlood vessel
Ampulla
Common atnal siphon
Oral siphon
Common atrial aiphon
Alrial siphon
figment band
Oral siphon
Blood vessel
Ampulla,
FIGURE 2. Details of colony in Figure 1.
232 ROGER MILKMAN
Sabbadin (1958) has also caused right-left inversion of asymmetry by disturbing
early buds. This situs inversus is perpetuated faithfully in further atrially budded
generations. We have repeated these observations.
Since all the zooids in a colony are ordinarily derived by budding from a single
progenitor, the colony is a clone, and all the individuals share a color pattern which
is distinguished easily from most of the myriad color patterns of surrounding col-
onies. Age and environmental influences on color patterns exist (Watterson,
1945) but are subject to independent analysis. Contiguous colonies are delineated
by a clear discontinuity, generally bordered by tiers of vascular ampullae. Since
the colonies occasionally fuse (the possibility apparently being based on genetically
controlled affinities), mosaics do arise. These mosaics may become quite complex,
since each generation of zooids is resorbed when its buds mature, and the systems
can be rearranged radically — indeed, shuffled — as the number of zooids changes.
The basis of the color pattern lies in the number and distribution of three kinds of
pigment cell: blue (granular), white (granular — purine derivative), and orange
(carotenoid in solution) (Sabbadin, 1959).
Experiments by Sabbadin (1959, 1962, 1964) show that certain components of
color patterns are inherited in a simple Mendelian way. The availability of a large
number of potential markers, together with other useful properties, suggests the
feasibility of extensive genetic studies on Botryllus. This accessibility to classic
genetic analysis is fortunate in view of the major modern problems for whose
investigations it appears remarkably well suited. Sabbadin (1959) has pointed out
that the tremendous variety of color patterns, once resolved into the activities of
individual genes, would offer a way to study natural genetic variation by direct
observation. In addition, of course, subsequent studies of the individuals observed
and their progeny would add further substance to such an investigation. The pri-
mary concern of the present paper, however, is development. The remarkable
powers of regeneration shown by Botryllus put it in the same league as many
plants, such as carrot and tobacco ; moreover, its structural complexity and the
similarities of its larval development to that in some vertebrates add to its desira-
bility for study. Finally, it is the hope of many animal geneticists to establish cell
culture operations by which they can alternately treat cells as micro-organisms for
mutational and recombinational studies, and grow them into adult organisms for
detailed study of form and function. Thus an important part of the genetics of
development may be accessible to analysis in Botryllus. The experiments to be
described may be viewed as steps in this direction.
LABORATORY CULTURE
Botryllus can easily be maintained in laboratory culture, provided that certain
conditions are met. Cultures must be flat, rather than in clumps. The flat growth
habit is automatic when larvae settle on glass slides or similar flat objects and
metamorphose. Flat colonies from large mussel shells, boards, or similar natural
surfaces can easily be removed and allowed to attach to glass ; small clumps, if
attached to a flat surface, will also spread out by the movement (on the order of
several mm./day) of the existing systems and with subsequent growth.
The physical circumstances of the colonies in culture are critical. Beakers of
still sea water serve well. Larvae will attach to glass slides, which can then be
BOTRYLLUS GENETICS AND DEVELOPMENT
placed vertically, or even better, aslant with the Botryllns facing down. Horizontal
mounting, with the colonies upside down, is best of all : the fecal pellets drop away,
and contaminating filamentous algae are less likely to take hold. Saran wrap, to
which the larvae readily attach, can he floated on culture medium also. For ex-
amination, it is inverted and submerged ; it can be refloated when desirable. Also,
for fusion compatibility tests, the Saran can be cut and colonies approximated, with
a firm supporting substratum if desired. Zooids on the bottom of a vessel simply
do not do very well.
Turbulence results in the presence of fecal pellets and other comparably sized
detritus throughout the sea water : contact of such particles with the oral siphon or
branchial basket causes reflex cessation of pumping and reversal of water flow by
contraction of the body wall and thus prevents feeding and a normal flow of water
through the individuals. If such a situation persists, the colony degenerates. Ac-
cordingly, aeration and stirring, if employed, require careful design.
Botryllns is tolerant of salinity changes. Concentration or dilution of sea water
by 20% produces no ill effects, and short exposures to more extreme conditions
(including distilled water) can be survived. The use of Instant Ocean, an arti-
ficial sea water, is advantageous from several points of view. It contains no organic
substances and no predators or competitors, and it is a great convenience inland.
Under conditions where evaporation is controlled or compensated and micro-
organisms do not multiply explosively, weekly changes of water suffice.
Cultures grow well between 18° and 28° C. Since water temperatures survived
over the winter are much lower, it is likely that lower culture temperatures could
be used, but growth would be very slow at best. It is also probable that even
higher temperatures could be used, particularly where other conditions are optimal.
Cultures can sometimes survive for months without added food. The budding
cycle proceeds slowly with a gradual reduction in size and number of individual
zooids. This suggests that rapid regression and death are not due to starvation,
and thus that a major problem in laboratory culture is the control of other
organisms. Colony growth, of course, requires the addition of food (algae), either
via running sea water or from algal cultures. Cyclotella nana, a centric diatom,
appears to be the best food organism used so far. In f/2 medium (Guillard and
Ryther, 1962) made with Instant Ocean the algal cultures reach concentrations of
1 X 106 cells/ml. ; in f/2 medium made from sea water, 2 X 106 cells/ml. Botryllns
colonies grow well in concentrations of 0.5-2.5 X 105 cells/ml. Indeed, young
zooids under these conditions have on occasion developed four buds each, one more
than the three considered maximal till now (Berrill, 1961), and the four buds
have all become functional zooids in some cases.
Satisfactory feeding of any filter-feeder requires that two conditions be met :
first, there must be enough food, and second, the concentration of this food must
be high enough for an adequate feeding rate but not high enough to be harmful.
In the present case, a concentration of 1-2 X 105 cells/ml, is used. This is a safe
distance from the level at which the feeding system becomes clogged, too many
algae accumulate on the dorsal lamina, periodic regurgitation takes place, and death
eventually ensues. Botryllus is apparently not successful at intermittent feeding in
constant high concentrations of food. A concentration of 5 X 105 cells/ml., for
example, is accompanied by slow growth and poor appearance of the colonies.
234 ROGER MILKMAN
Higher concentrations generally cause regression after a day or so. On the other
hand, 0.5 X 105 (a concentration which also supports efficient feeding) has the
disadvantage of providing only { as many diatoms as 2 X 105 in a given vessel,
thus necessitating a volume four times as great.
With just a few newly metamorphosed oozoids, culture vessel volume is no
problem ; but with colony growth, the removal of algae from the medium becomes
rapid. On the basis of the time taken for a given colony to clear its water, I
estimate that each zooid can easily filter 2 X 10G algae per day. (Not all of these
are absorbed, as examination of the fecal pellets shows, but they are no longer
available.) Thus, if a suspension of 2 X 105 algae/ml, is provided each day, the
minimum culture vessel volume is about 10 ml. for each zooid. Since food intake
becomes slower as the concentration falls, doubling this volume would be even
better. A suspension of 1 X 105 algae/ml., allowing 40 ml./zooid, is probably
optimal.
Closed vessels (or vessels covered with glass plates, Parafilm, etc.] are obvi-
ously convenient, particularly since stirring is not required. Polyethylene, through
which oxygen and carbon dioxide can pass and water vapor .cannot, suggests itself
as a good cover (Walters and Williams, 1966). Constant light is acceptable and
permits the diatoms to photosynthesize and thus produce oxygen. It also favors
the growth of all algae, however, and may do more harm than good if certain
filamentous forms are present. Budding and gonad development proceed similarly
on all light regimes; only larval release (Costello et al., 1957) seems to be
influenced by light. In any event, large colonies require impractical volumes in
standing culture, so continuous flow systems maintaining the concentration of algae
within the desired range are preferable for them. The development of a recircu-
lating aquarium for filter-feeders would be useful : the problem is merely one of
finding an appropriate water filter.
For genetic studies, rapid growth and sexual maturation are desirable. Under
the culture conditions described, performance is satisfactory. In the progeny of
one mating, the colonies had from 5-22 zooids 22 days after metamorphosis (26
days after fertilization). Mature eggs are produced by cultured colonies less
than 1^ months after metamorphosis; and once eggs are produced, of course, a
new batch appears every 5-7 days, as long as conditions remain good.
Sabbadin (1960) has reported using Chlamydomotias (marine members of this
genus are now called Dunaliella) and Nitsschia (perhaps what is now called
Phacodactylum) to feed isolated colonies at Chioggia, on the Lagoon of Venice.
I am not convinced that I have given this combination an adequate test, but I have
not been successful with it.
Predators, such as the snail Mitrella lunata and probably some flatworms and
nematodes ; competitors, such as filamentous algae, sponges, encrusting ectoprocts,
Bugula, and entoprocts ; and bacteria (whose activities may be competitive or
direct) can all destroy cultures. In running sea water, Mitrella, sponges, and
ectoprocts become an increasing problem as the summer progresses. In isolated
culture, bacteria, algae, and entoprocts have proven more bothersome. Amphi-
pods, harpacticoid copepods, and a variety of ciliates seem to coexist peacefully in
Botryllus cultures as they do in the miniature jungles of wild colonies. But clean
colonies derived from washed unhatched larvae do best.
BOTRYLLUS GENETICS AND DEVELOPMENT
235
Although in the long run Botryllus cultures require the maintenance of favora-
ble conditions, they respond well to occasional rough treatment. For example,
small colonies on glass have survived exposure to air for ten minutes or longer
and microscopic examination under a coverslip for similar periods. Colonies
accustomed to 25° C. have survived a day in the refrigerator, but not much longer.
THE REPRODUCTIVE CYCLE
It is of particular importance when dealing with an organism capable of selfing
to have control of fertilization. Such control is achieved in B. schlosseri by fer-
tilizing isolated eggs with isolated sperm (Milkman and Borgmann, 1963). It is
believed that this is the first time external fertilization has been accomplished with
a compound ascidian, and it depends upon removing the eggs at the right time.
This in turn depends upon a detailed understanding of the timing of egg matura-
tion and sperm maturation in relation to one another and to the asexual cycle.
The present investigation has clarified these time relationships.
TABLE I
Timetable for one asexual generation
Day
Adult
Embryo
Testes
Bud
Egg
Testes
0
Takes over
Fertilized
0-1
Small
Small
1
Darkens
Raspberry
1-2
Grows and
2
Grows very
Tailbud
2-3
projects out
3
little
Wraparound
2-4
between zooids
4
Larva
3-5
pre-0
5
Released
4-remnants
Swells and
Full-
pre-0-0
6
Resorbed
takes over
sized*
0-1
* Enters atrial cavity, germinal vesicle breaks downs, egg fertilized.
Eggs develop in special chambers beside growing buds. They reach maturity
when the buds replace the previous zooid generation. Since this is a fairly
synchronous process (distant systems in a large colony may be several hours
apart), one can obtain hundreds, even thousands, of eggs from a good-sized colony.
As the new generation takes over, the eggs are pushed out of their chambers into
the atrial cavity of the swelling bud. The germinal vesicle is in clear evidence
in the eggs. During the next two to three hours, the following things happen
in parallel: (1) the old zooids shrink down and no longer contain (or release)
sperm; (2) the new zooids swell further and open their siphons; (3) the germinal
vesicle breaks down and the eggs soon become fertilizable.
It should be added that the new zooid's testes generally do not release mature
sperm until two days later. From this array of events, then, it follows that eggs
will not be fertilized by sperm from the same colony-clone unless sperm are not
forthcoming from elsewhere for two days, or unless the colony is so large that the
first eggs become accessible and fertilizable before the last old zooids, often virile
to the end, lose their mature sperm.
At the time of takeover, the testes contain very few mature sperm. That these
are not released is indicated by the failure of newly mature eggs in the same colony
236
ROGER MILKMAN
or another to be fertilized by them. The proportion of mature sperm in a testis
rises with time. The rate of release must be low at first ; the testes ultimately reach
a state of great fragility in which they contain nothing cellular except sperm oriented
in parallel ; at this point the sperm output of the colony must be several orders of
magnitude greater than at first. Table I shows the timetable of sexual and asexual
events in an adult colony. The budding cycle takes 5-7 days in the laboratory.
Illustrated in Table I is a 6-day cycle. The embryonic stages' designations used
here for convenience, reflect their appearance. The raspberry stage is a gastrula ;
the wraparound is still rather opaque and spherical, with the tail wrapped around.
Later the embryo clears and elongates into the larva, whose subsequent release
appears to be influenced by its light regime (Costello et al, 1957). The buds are
not visible until just before takeover, except in very flat colonies. Otherwise they
are concealed in the interior of the colony mass.
TABLE II
Distribution of testis stages vs. embryo stage in individual colonies
Testes in each stage (N)
Colony
Em for vo s ts.sc
0
1
2
3
4
5
i
Raspberry
0
1
22
2
0
0
2
Raspberry
0
4
23
3
0
0
3
Tailbud
0
2
22
1
0
0
4
Tailbud
0
0
9
11
1
0
5
Early wraparound
0
1
13
2
0
0
6
Wraparound
0
0
1
22
0
0
7
Larva
2*
1*
1
18
0
0
8
Larva
0
0
0
3
17
1
9
Tadpole
1*
1*
0
0
0
15
10
Tadpole rare
0
2
0
0
0
20
* Possibly taken accidentally from buds (see text).
Table II shows the degree of uniformity among the testes in a colony. Colonies
were staged according to their embryos, whose stages are very uniform indeed
under normal circumstances. Since these colonies were taken from a dock crowded
with Botryllus, their eggs were surely fertilized at the earliest possible moment.
Now within a given colony, the testes appear to be fairly synchronized, though there
is some scatter. (Exceptionally immature testes in an otherwise mature group may
have been taken accidentally from a bud.) On the other hand, a comparison from
colony to colony suggests that the phase relationship is not constant for the species,
though of course its range of variation is not great. This variability among colonies
must be kept in mind, for it, too, affects the possibility and time of selfing.
It should be clear from this description that there is no sudden onset of paternal
competence in a B. schlosseri colony. Mature sperm are seen well before they are
normally released ; crushed testes achieve a small percentage of fertilization at early
stages also. Table III illustrates the quantitative nature of testis maturity, corn-
pairing testis stage with per cent fertilization. Eggs from the same batch were placed
with crushed testes of various stages and cleavage was observed.
BOTRYLLUS GENETICS AND DEVELOPMENT 237
Conjectures involving storage of sperm or other complex mechanisms of fertili-
zation can be discarded because eggs can be taken at the right moment and fertilized.
Similarly, there is no evidence of egg-sperm incompatibility. The only technical
difficulty is that eggs isolated with germinal vesicles intact will never be fertilized ;
and it is probable that they do not mature until about an hour after breakdown.
Subsequently the eggs can be removed and fertilized. The actual removal consists
of slitting the zooids and gently pressing out the eggs : this is a very easy procedure,
and the zooids repair the damage within 24 hours.
Large wild colonies containing many eggs per zooid (I have removed as many
as 1 1 from one) can be staged and isolated about a day before takeover. As
Sabbadin points out (1959), the property of fusion places the clonal nature of any
wild colony in doubt, however, and it is certainly better to raise breeding colonies
from tadpoles. At any rate, for eggs each colony should be sequestered before the
new siphons open. For large colonies, a Incite container, divided into radial sectors,
has been used to isolate up to ten potential egg sources. This device, built to operate
like a reverse Botryllus, distributes water from a common central tube ; outflow is
TABLE III
Fertilization efficiency vs. testis stage
Testis stage
Eggs (N)
% Fertilized
1
55
0
3
135
15
31
80
28
3|
62
29
4 +
83
46
peripheral. Where running sea water is not available, large chambers containing
no food can be used. Here stirring involves no risk. A closed vessel, together with
the great efficiency of the Botryllus pumping system, raises the problem of selfing
once more (which is minimized by the constant washing of continuous flow).
Accordingly, egg sources should be washed and isolated shortly before the new
siphons open.
Eggs to be fertilized in vitro are placed in a Syracuse dish of (natural or arti-
ficial) sea water. Testes are then added. After the eggs are swirled to the center
of the dish, the testes are crushed, and the eggs are nested in a thick layer of sperm
like berries in whipped cream. Polyspermy is fortunately not a problem, and
fantastic quantities of sperm are required in comparison, for example, to the amounts
needed to fertilize sea urchin eggs. This situation, seen somewhat less spectacularly
in other tunicates (Costello et al., 1957), suggests that imperfections still remain
in the in vitro method, even though 100% fertilization can be achieved.
DEVELOPMENT OF FERTILIZED EGGS
Eggs fertilized in vitro, as well as early embryos removed from zooids, can
develop into mature larvae and subsequently metamorphose. Until recently, tech-
niques for permitting such in vitro development were complicated and unreliable at
best; this was a major obstacle to the use of controlled mating. Now, however, the
238 ROGER MILKMAN
simple expedient of placing the early embryos on a piece of filter paper in a vessel
containing Instant Ocean provides a good method of raising them. The filter paper
apparently serves two purposes : it provides for some circulation even where the
egg rests on its surface (Saran doesn't work as well), and it is not as hard as glass.
Eggs resting directly on glass become deformed and bacteria accumulate at the
point of contact and almost invariably attack and destroy the embryo by the time
of the gastrula stage or thereabouts. Filtered sea water is also satisfactory. If
evaporation is prevented, the water need not be changed, but a change seems to
result in healthier larvae. Hundreds can be raised in a finger bowl. Before eclosion,
the larvae are transferred to an appropriate vessel containing slides or Saran for
attachment, since the time of eclosion and that of attachment and metamorphosis are
variable and hard to control.
THE VASCULAR SYSTEM AND VASCULAR BUDDING
The vessels which pervade the zooids and outlying areas also connect to the
many vascular ampullae which are found in tiers at the periphery of the colony, in
rings around each system, and irregularly scattered throughout the matrix between
systems. Differential interference (Zeiss-Nomarski system) microscopy, which
permits undistorted high-magnification observation of optical sections of relatively
thick preparations, shows that the vessels and ampullae are very delicate structures.
Their walls are essentially one cell thick. Ordinary light microscopy of young
colonies under coverslips permits similar observations. The blood circulating
through the systems is moved, not only by the hearts in the various zooids, but by
contractions of the vascular ampullae. This is confirmed by motion pictures, again
using the Nomarski microscope, which show localized contractions within each
ampulla and thereby eliminate passive elastic contraction as the basis of their
periodic reductions in size. Moreover, removal of all the zooids and buds does not
stop circulation in the remaining outlying vessels. Circulation continues for hours,
and longer.
Isolated regions of the vascular system are of great interest because they can
regenerate entire zooids and, ultimately, whole colonies, in spite of their simple
structure and composition. Oka and Watanabe (1957, 1959) demonstrated vascular
budding in B. primigenus and in Botrylloides; Watkins (1958) suspected it in B.
schlosseri; and Byrne and I demonstrated it (Milkman and Byrne, 1961). In B.
schlosseri vascular budding has been seen only when all the zooids are removed, but
the possibility remains that under conditions of very rapid growth it occurs in intact
colonies also. During the first few days following excision of the zooids (and
buds ! ) , the ampullae consolidate into one or a few rather highly pigmented masses,
and these structures may now gradually begin to resemble miniature zooids, whose
size is that of zooids newly metamorphosed from larvae. Histological study of this
process has not yet been made in B. schlosseri, but even at the gross level there
appear to be differences between vascular budding in this species and those reported
by Oka and Watanabe. In less than a week, the tiny functional zooids are feeding
and growing.
This remarkable regenerative ability leads one to wonder if B. schlosseri cells
can be cultured and then be induced to form zooids in a manner analogous to vascu-
lar budding. It was considered useful first to make further inquiry into the nature
BOTRYLLUS GENETICS AND DEVELOPMENT
239
of the cells determining the characteristics of the zooids so produced. To this end,
advantage was taken of the ability of morphologically different colonies to fuse
(Bancroft, 1903). Colonies differing in color pattern were made contiguous; a
small proportion of them fused gelatinous tests and vascular systems, thus permit-
ting (indeed, necessitating) a complete interchange of blood cells. A week after
fusion, all zooids and buds were removed from the fused colonies, leaving only their
common vascular systems (and tests). Separate portions of each colony were
taken before fusion and maintained for comparison with the zooids to be regenerated.
In over 30 cases, the zooids produced by vascular budding in turn produced systems
identical to the parent systems originally present at the site of the bud. There
were no exceptions. This proves that, whatever the contributions of the freely
circulating blood cells, the fixed cells of the delicate vascular walls (or conceivably
of the test) determine the phenotype of the regenerated zooids (Milkman and
Therrien, 1965). Figure 3 illustrates the experiment. No buds were seen at the
original fusion border; this is probably a statistical matter. Perhaps a somatic
recombinant could be obtained if a bud of dual origin occurred.
FIGURE 3. Experiment on vascular budding in combination of two fused colonies. Note
similarity of system derived from vascular buds to systems derived from vascular buds to sys-
tems originally present at the same site. See text for further details.
In these experiments, it was necessary for the systems produced to reach a
steady-state with respect to pigment cell concentration. During the first several
atrial budding generations after the vascular bud is formed, the proportion of pig-
ment cells is unusually high, as it is in the original coalesced vascular tissue.
Gradually this proportion decreases and is maintained at a steady level correspond-
ing to that in the parent. This situation is reminiscent of the changes in pigmenta-
tion in the first few bud generations coming from an oozoid ; but in the latter case,
pigment cell concentration starts low and then rises to a maintained level (Bancroft,
1903; Watterson, 1945; Sabbadin, 1959). Moreover, when the old zooids are
resorbed, their pigment cells assemble in the ampullae and are gradually released
and taken up by the new generation of zooids. Thus buds begin by being rather
pale ; their sudden expansion at takeover time spreads out their relatively few pig-
ment cells and makes them paler ; and they gradually darken the next day. From
these observations, old and new, it can be seen that the distinctive pigmentation of
a mature Botryllus colony depends basically on its own particular genotype, and
that it is affected more immediately during the ontogeny of the colony by the relative
rates of formation of pigment cells and their concentration in a given stage (oozoid
or regenerating tissue).
240
KOGER MILKMAN
The obvious next step in this area of investigation, which has not yet been
undertaken, is an attempt at culturing vascular tissue, with or without blood cells.
It is of interest that Freeman (1964) has found in Pcrophora that only lympho-
cytes, of the several varieties of blood cells, are necessary for budding, which occurs
normally in that form at intervals at the growing end of the stolon, which carries
bloods among the zooids. Since he arrested all cell division by irradiation and
subsequently injected untreated cells of a given type, it is clear that these cells
form the body of the newly budded zooids, unless (as seems unlikely) large num-
bers of non-dividing cells are mobilized from existing zooids once the blood cells
initiate budding. Our findings do not rule out the participation of blood cells, but
they suggest that the structure of the Botryllits zooid formed in vascular budding
is controlled by the derivatives of the vascular tissue, or conceivably (though the
appearance of the bud lends no support to this alternative) cells in the test. The
quantitative changes in pigmentation suggest that different regions in the zooids
have different affinities for each kind of pigment cell and that a mass action relation-
TABLE IV
Results of representative crosses
Cross
Phenotypes
Genotypes
Phenotypes in offspring
1 (self)
2
3
PB X PB
pB X PB
Pb X PB
PP Bb X PP Bb
pp Bb* X PP BB
P-t bb X P-t Bb*
20 P:0p
15 P:0P
19 P:0p
15 B:5 b
15 B:0b
7 B:12 b
Pigment band: P presence, p absence. Black ground color: B presence, b absence.
* Bb genotype indicated by another cross of same parent (not shown here),
t At least one should be PP.
ship determines the disposition of pigment cells at any time. The mobility of pig-
ment cells that has been observed supports the view that, though they may lodge
in a particular place for a considerable time, they are never permanently fixed.
GENETIC CROSSES
The crosses we have performed so far have been preliminary in nature and lead
to three conclusions. First, selfing is general enough to suggest the absence of any
important self-incompatibility system (except, of course, the highly effective differ-
ence in time of maturation of eggs and testes in a given colony). Second, Sab-
badin's conclusion that the presence of a pigment band may be inherited in a simple,
dominant Mendelian fashion is supported, although neither his data from individual
crosses (Sabbadin, 1959, 1962, 1964) nor ours definitively exclude additional possi-
bilities. Third, it is clear that the tremendous variety of offspring produced from a
cross of any two colonies taken from nature defy extensive analysis : several genera-
tions of selfing are required to produce colonies sufficiently homozygous to be
useful for the study of a large number of traits. Such a program of selfing necessi-
tates laboratory culture methods capable of supporting sexual reproduction con-
sistently ; even now that we have such methods, any major degree of heterosis might
delay or prevent the acquisition of homozygous colonies.
BOTRYLLUS GENETICS AND DEVELOPMENT
Table IV contains the results of some representative crosses. Putative parental
genotypes for pigment bands are assigned tentatively ; it is not really clear at this
point that the inheritance of black ground color is simple. The numbers are quite
.small, but since these crosses were set up the techniques for getting good yields have
been improved greatly.
There is one detrimental result of in I'ii'o selfing of the usual type : when self-
fertilization takes place two days after the normal time, the larvae are not ready
for release at takeover time (see Table I ). The colony seems unable to adjust its
asexual schedule; the old zooids regress slowly and incompletely while they con-
tain larvae. Concomitantly, the buds do not complete their last stages of develop-
ment : they appear undersized and do not become contiguous with other zooids to
form normal systems. Thus the colony dies, though many larvae escape. In ex-
ceptional cases of earlier selfing due to fertilization by remaining zooids of the pre-
vious generation or by unusually advanced testes of the current generation, this
collapse may not occur. Actually, the fortunate expedient of refrigerating testes
up to four days may be employed ; sperm from these testes fertilize the eggs of the
next generation perfectly. In addition, the possibility also exists of separating a
colony into parts and staggering them at different temperatures.
YVhere traits are inherited in a simple Mendelian fashion, the alleles responsible
can be followed in populations. As Sabbadin has suggested (1959), BotryUus is of
particular interest because in animals two alleles associated with a striking pheno-
typic difference rarely both have high frequencies. The pigment band's presence
appears to be dominant over its absence, although Sabbadin (1964) believes that
multiple alleles account for some of the variants in pigment band form. At any
rate, among 100 colonies on the M. B. L. Supply Dock in Woods Hole, 63 had
pigment bands and 37 did not. If the conditions for the Hardy-Weinberg law ob-
tained, the frequency of the "absence" allele was V0.37 := 0.6, while that of the
"presence" allele, or class of alleles. was 1 -- 0.6 ~ 0.4. Extension of these observa-
tions over space and time should be quite easy and may lead to useful conclusions
about population sizes and related matters.
CHROMOSOME NUMBER
Colombera (1963), using either gallocyanin or gentian violet, together with pre-
liminary aceto-orcein or aceto-carmine staining, on testes, buds, and cleaving eggs,
has concluded that the haploid number of chromosomes in B. schlosseri is 16.
Therrien and I (Milkman and Therrien, 1965) studied cleaving eggs using the
Feulgen technique, blockading the cytoplasmic aldehyde groups with phenylhydra-
zine before hydrolysis. (Pronase had removed the chorion. ) We estimated the
haploid number to be 7 or 8. It is possible that this disparity has a real basis, or,
of course, that our conclusions are incorrect. Colombera points out that 16 is
rather high for ascidians, but that Tcthynui plicatitin, of the Styelidae (a family
close to. or including, the botryllids) also has a haploid number of 16.
HANDLING OF BLOOD
In my laboratory, Dr. Arnold Kahn has found it easy to remove and reinject
Botrvllits blood. As much as ] ml. of blood has been taken from a colony at one
\
242 ROGER MILKMAN
lime, suggesting tliat the alternate ])assive elastic stretching and active contraction
of ampullae are ordinarily responsible for the periodic reversal of peripheral blood
flow. Blood cells and test cells survive in culture for up to three days but do not
multiply. The ability to remove and inject blood easily and without injury permits
one to attempt to confer fusion compatibility. It is also conceivable that intravenous
feeding alone can support the growth of a Botryllus colony or isolated vascular
system, and that nutritional studies at this level might lead to a wide variety of
interesting findings.
GENERAL DISCUSSION AND CONCLUSIONS
The basic technical obstacles having been overcome, we can now look forward
to extensive genetic studies on Botryllus. Inland culture techniques permit year-
round propagation of strains, together with long-term crossing and selection pro-
grams. The ever-present risk of selfing, which cannot be controlled /;/ vivo, is
eliminated by the use of in vivo matings, which also permit multiple crosses involv-
ing one set of eggs or one set of testes. It may be concluded, then, that Botryllus
schlosscri is ripe for teaching and experimental use. In anticipation of its increased
popularity, in view of its appearance and habitat, and to remedy a current defect,
the vernacular name "harbor stars" is now suggested.
This work has been done with the collaboration and assistance of Sylvia Byrne
and Edward Therrien (NSF Undergraduate Research Participants), Martha Borg-
mann and Judith Pederson. Dr. Luigi Provasoli, Dr. Robert Guillard, and Airs.
Helen Stanley have been most generous with algal cultures, materials, and counsel.
Dr. Martha Baylor suggested the fusion-vascular budding experiment. Dr. Robert
D. Allen provided the Nomarski optics and made the movies. The illustrations are
by Julia S. Child.
SUMMARY
1. Properties of Botrvll/is schlosscri which give it outstanding promise for
studies in developmental genetics are reviewed.
2. Laboratory culture procedures, in vitro fertilization, and a method for raising-
embryos in vitro are described. Controlled successions of complete life cycles can
now be achieved in any laboratory.
3. Experiments involving colony fusion, subsequent vascular budding, and the
analysis of color patterns in resultant systems suggest that cells of the simple vessel
walls govern the morphology of the regenerated zooids.
4. Results of some preliminary genetic crosses are reported.
LITERATURE CITED
BANCROFT, F. W., 1903. Variation and fusion in compound ascidians. Prnc. Calif. Acad. Sci.,
scries 3, 3: 137-186.
BERRILL, N. ]., 1941a. The development of the bud in Bofryllus. Biol. Bull., 80: 169-184.
BERRILL, N. J., 1941b. Size and morphogenesis in the bud of Botryllus. Biol. Bull., 80: 185-193.
BERRILL, N. J., 1950. The Tunicata. Ray Society, London.
BERRILL, N. J., 1951. Regeneration and budding in tunicates. Biol. Rcr., 27: 456-475.
BERRILL, N. J., 1961. Growth, Development, and Pattern. Freeman, San Francisco.
BOTRvu.rs GENETICS AXD DEVELOPMENT 243
COLOMBERA, D., 1963. I cromosomi di Botrvllus schlosscri (Ascidiacea). h'ie. Sci., 33 (Il-b).
443-448.
COSTELLO, D. P., M. K. I ).\\ nisiix. A. EGGERS, M. H. Fox AXD C. HENLEV, 1957. Methods for
Olitaining and Handling Marine Eggs and Embryos. Marine Biological Laboratory,
Woods Hole, Mass.
FREEMA.V, G., 1%4. The role of the blood cells in the process of asexual reproduction in the
tunicate Pcrophoru -I'iriilis. J. /:.r/>. Zool.. 156: 157-183.
GKOBSTEIX, C., AND E. ZWILLING, 1953. Modification of growth and differentiation of chorio-
allantoic grafts of chick blastoderm pieces after cultivation at a glass clot interface.
/. £.r/>. Zool, 122: 259-284.
GriLLAKD, R. R. L., AXD J. H. RVTHER, 1962. Studies of marine planktonic diatoms. I. Cyclo-
tcllu liana Hustedt, and Dctonula conferracea ( Cleve) Gran. Canad. J. Microhiol.. 8:
229-239.
LOPASC HOV, G., 1935. Die Entwicklungsleistungen des Gastrulamesoderms. Biol. Zentrcilhl..
55: 6C6-615.
MILKMAX. R., AXD M. BoRGMAXx, 1963. External fertilization of Botrvllus schlosscri eggs.
Biol. Bull., 125: 383.
MILKMAX-, R., AXD S. BYRXE, 1961. Recent observations on Botrvllus schlosscri. Biol, Bull..
121: 376.
MILKMAX, R., AXD E. THERRIEX, 1965. Developmental and genetic studies on the compound
ascidian, Botrvllus schlosscri. Biol. Bull.. 129: 417.
OKA, H., AXD H. WATAXABE, 1957. Vascular budding, a new type of budding in Botrvllus.
Biol. Bull., 112: 225-240.
OKA. H., AXD H. WATAXABE, 1959. Vascular budding in Botr\Iloidcs. Biol. Bull. 117' 340-
346.
OKA, H., AND H. WATANABE, 1960. Problems of colony-specificity in compound ascidian s.
Bull. Marine Bin!. Sta. Asunnishi. 10: 153-155.
SABADDIX", A., 1958. Analisi sperimentale dello sviluppo delle colonie di Botrvllus schlosscri.
Arch Ital. Anat. EmbrwL, 63: 178-221.
SABADDIX, A., 1959. Analisi genetica del policn.matismo di Botrylltts schlosseri (Pallas)
Savigny (Ascidiacea). Boll. Zool.. 26: 221-243. (Translation by Roger Milkman
available on request.)
SABADDIX, A., 1960. Ulteriori notizie sull'allevamento e sulla biologia dei Botrilli in condizioni
di laboratorio. Arch. Occanoyr. Liinnol.. 12: 97-107.
SABADDIX, A., 1962. Bande intersifonali di pigmento purinico in Botryllits schlosscri (Ascidi-
acea) e loro determinazione genetica. Boll. Bool.. 29L: 721-726.
SABBADDIX, A., 1964. The pigments of Bntr\llns schlosscri and their genetic control. Ric. Sci..
34: (Il-b) : 439-444.
SCOTT, F. M., 1934. Studies on the later embryonic development of Tunicata : Botrylliis schlos-
scri and Ai/uirocciuiii constellation. Ph.D. thesis, Columbia University, Xew York.
VAN- NAME, W. G., 1945. The North and South American ascidians. Bull Am. Mits. N,if.
Hist., 84: 1-476.
WALTERS, D. R., AXD C. M. WILLIAMS, 1966. Reaggregation of insect cells as studied by a
new method of tissue and organ culture. Science, 154: 516-517.
WATKIXS, M. J., 1958. Regeneration of buds in Botrylliis. Biol. Bull.. 115: 147-152.
WATTERSOX, R. L., 1945. Asexual reproduction in the colonial tunicate, Botrylliis schlosscri
(Pallas) Savigny, with special reference to the developmental history of intersiphonal
bands of pigment cells. Biol. Bull., 88: 71-103.
ENDOSKELETAL CARTILAGE IX A MARINE POLYCHAETE,
EUDISTYLIA POLYMORPHA
PHILIP PERSON AND MARTIN B. MATHEWS
Special Research Lal'oratory fur Oral Tissue Metabolism,
V. A. Hospital, Brooklyn, New York 11209,
Department of Biochemistry, La Rabida-University of Chieai/o Institute,
('liica</o. Illinois 60649,
and the Mannc f-iiolof/ical Laboratory, ll'aods Hole. Massachusetts 02543
The nature of the endoskeletal tissues of certain marine worms has been the
subject of controversy in the past. Claparede ( 1873) described such tissues as part
of the tentacular and tentacle-supporting complexes in Alv.vicola infundibulum and
Spirographis spulUtnzani, and considered them to be cartilage. Viallanes (1885),
XowikofT ( 1912) and Hempelmann ( 1928) described such tissues in the above and
other polychaetes, and also considered them to be cartilage. However, in a brief
report, Krukenberg ( 1882) said he could extract neither gelatin nor "mucin" from
the tissues referred to above. He therefore concluded they could not be true
cartilage, and that instead they were "cartilage-like." Subsequently, several other
workers studied these endoskeletal tissues, and, basing their judgments primarily
upon Krukenberg's negative findings, also decided that the tissues in question were
not cartilage (Nicol, 1930; Evenkamp, 1931 ; Thomas, 1940).
The above problem is included in the broader question of the existence or non-
existence of cartilage tissues in invertebrates. XowikofT (1912), on the basis of
histologic, cytologic and also some chemical criteria, concluded that invertebrates
did possess true endoskeletal cartilage tissues, while Schaffer (1913, 1930) argued
that the invertebrate tissues in question were not cartilage. Schaffer (1930) con-
sidered that the invertebrate tissues in question were "chondroid," "chordoid" or
"cartilage-like," and that "true" cartilage was found only in the vertebrates ( 1930,
p. 210). In 1940, L. Hyman in her classic treatise on the invertebrates wrote that
true cartilage was absent in invertebrates, which, she said, tend to secrete external
noncellular rather than internal cell-containing hard parts (p. 281 ). Subsequently,
Ronier (1942) suggested that cartilage arose in the vertebrates as an embryonic
adaptation to stresses and deformations produced by rapid growth. Since then, the
view seems to have prevailed amongst most biologists that cartilage is a uniquely
vertebrate tissue, as expressed more recently, for example, by Pritchard (1956)
and Ronier (1964). Indeed, most textbooks and monographs on vertebrate bone
and cartilage, and most textbooks of invertebrate zoology, rarely mention inverte-
brate cartilage tissues. It should be noted, however, that a small number of indi-
viduals have continued to write of certain invertebrate cellular endoskeletal tissues
as being cartilage (Person and Fine, 1957; Raven, 1958; Lash, 1959; Person and
Philpott, 1963a).
In this paper, in further support of the contention that true endoskeletal cartilages
244
POLYCHAETE ENDOSKELETAL CARTILAGE
245
are found in invertebrate animals, \ve present a beginning study of the histology
and chemistry of cartilage tissues which form part of a rather complicated cellular
endoskeletal complex in the marine polychaetous annelid, Eudistylia polymorpha.
sp*&,£ %
<". ~+ .^/-v-
*£•**/*?* • ••* -i*i^" >•
~% 7* £rA -&>**&
'- i?^>^ ^^J^f^^M
^^i^^^Cv '£$T*f *&
;>;
.V*.
ep
ca
FIGURE 1. Endistylia polymorpha, 8 inches long, removed from tube and photographed alive.
FIGURE 2. Endistylia polymorpha, cross section of animal at level of base of crown :
cp = epithelium, ca = cartilage, ma = matrix, os = osteoid-like tissue; magnification 18X.
FIGURE 3. Endistylia polymorphic,, higher-power view of circle 1 in Figure 2: ep — epithe-
lium, ca = cartilage, ma — matrix ; magnification 200X.
FIGURE 4. Endistylia polymorpha, higher-power view of epithelial cells which appear to
be secreting matrix (ma) ; magnification 500 X.
246
PHILIP PERSON AND MARTIN B. MATHEWS
A""
mu
FIGURE 5. Eittlistylia pi>lyin<irp!i(i. higher-power view of circle #2 in Figure 2: o^ =
osteoid-like tissue, ca — cartilage ; magnification 200X.
FIGURE 6. Eudistylia polynwrplw, cross section through the base of a cluster of tentacles :
ca r — cartilage rods, arrows (f) show origin of cartilage rods from basilar mass of cartilage
(ca), os = osteoid-like tissue, bl = blood ; magnification 18X.
POLYCHAETE ENDOSKELETAL CARTILAGE 247
MATKRIALS AND METHODS
Eudistvlia pol\morpha (Fig. 1) was collected and shipped ria air express to
the Marine Biological Laboratory, Woods Hole, Mass., by Dr. R. C. Fay. Pacific
Biomarine Supply Co., Venice, California. The animals were active on arrival and
were kept in their tubes in a sea tank for several days, following which they were
sacrificed. For histologic study, tissues were placed in \Qc/c neutral formalin and
processed routinely to make hematoxylin-eosin stained sections. Fresh-frozen
(non-fixed) sections were cut to approximately 10-1 5 /A and were stained with
0.01% toluidine blue at pH 3.5. For chemical analyses, the tissues were dissected
and trimmed under a binocular microscope, and placed in cold acetone. For
analysis of acid mucopolysaccharides, the tissues were digested with crystalline
papain, and the acid mucopolysaccharides isolated and characterized chemically and
physically by procedures described elsewhere (Mathews and Glagov, 1966).
Hydroxyproline content was kindly determined by Dr. Milton Levy according to
the method of Stegemann (1958).
OBSERVATIONS AND RESULTS
Gross observations
The endoskeletal tissues of Eudistylia are located anteriorly in the animal as
supporting structures for the "feather-duster"-like tentacular complex which forms
the crown of the animals. The main clusters of tentacles, of which there are two
(see Fig. 1), are supported by a basal mass of cellular endoskeletal material, com-
posed of cartilage and an "osteoid-like" tissue. (As used in this paper, the term
"osteoid" refers to the vertebrate tissue identified histologically as the premineralized
matrix in which apatite crystals are later formed during the process of bone
formation.) These tissues are never mineralized. From this basal endoskeletal
complex are also given off individual cartilage rods, each of which supplies a fairly
rigid, but flexible, structural core for a single tentacle. From each rod, in turn
strands or columns of cells are given off, to provide individual structural cores for
the numerous villus-like pinnae, which project from each tentacle along its entire
length. The pinnae exert a beating motion, creating water currents which sweep
food down along the tentacular grooves, to facilitate ingestion by the mouth.
Histology
Basilar endoskdctal complex. In Figure 2. a low power cross section made at
the level of the basilar endoskeletal complex is shown. In this section we will
locate structures of interest, which will then be shown at higher magnifications.
The outer surface of the animal is covered by a tall columnar epithelium (f/0. In
the region of the endoskeletal complex, the epithelium overlies and appears to
secrete a relatively homogeneous eosinophilic matrix (ma). In some regions, this
FIGURE 7. Endistylia pulymorpha, higher-power view of cartilage rods (ca r), which sup-
port individual tentacles, and surrounding matrix (mo) which is continuous with osteoid-like
tissue seen in Figures 2, 5 and 6; magnification 85X.
FIGURE 8. Eudistylia [><>lyiin>rplia, sagittal section of a tentacle and its pinnae (pi) : ca r -
cartilage rod, -inn — muscle bundles, sc = strands of cells which support pinnae; magnifica-
tion 85 X.
248 PHILIP PERSON AND MARTIN B. MATHEWS
matrix is penetrated 1>v an extensive vascular complex, giving to the tissue a bone-
like or "osteoid-like" appearance (os) (it is stressed again that these tissues never
mineralize). Adjacent to the matrix and "osteoid-like" tissue are masses of carti-
lage tissue (ca). In Figure 3, a higher power view of the circular area #1 (marked
in Fig. 2) is shown. In this region, a relatively homogeneous character of the
matrix (ma) is evident, as well as the abundantly cellular nature of the adjacent
cartilage (ca). The cartilage matrix is comprised of thin seams of intercellular
material. In Figure 4, the eosinophilic, granule-laden, tall, columnar epithelial cells
are pictured at higher magnification. These cells hear a marked resemblance to
amelohlasts or odontoblasts of vertebrate tissues, as well as to the odontoblasts of
invertebrate radula-forming tissues in gastropod molluscs (Raven, 1958). Figure
5 shows a higher magnification of the circular area # 2 (marked in Fig. 2). Note
the strong resemblance of the "osteoid-like" material (os) to vertebrate bone tissue
in section. The cartilage in this region (ca) is quite cellular, but its intercellular
matrix is thicker and more rigid than that seen in the cartilage in Figure 3.
TABLE I
Chemical analysis of acid niucopolysaccharide of Eudistylia cartilage compared to theoretical
for chondroitin sulfate of mammalian cartilage, expressed as mole ratio to galactosamine
Eudistylia
Theoretical
Galactosamine
1.00
1.00
Glucosamine
0.03
0.00
Uronic acid
1.07
1,00
Nitrogen
Sulfate
1.20
1.88
1.00
1.00
Tentacular complex. In Figure 6 is a cross-section of the worm made at a level
slightly anterior to that from which the previous figures were made. The cartilage
rods (ca.r) which provide internal support for each of the tentacles are strongly
eosinophilic and stand out like vascular bundles seen in cross sections of plant
tissues. The origin of the tentacular cartilage rods (ca.r) from the basilar cartilage
masses (ca) is seen in the areas marked by the arrows (t). One can also see at
this level that the channels which penetrate the osteoid-like matrix (os) are part
of the vascular system which contains a spirographis heme-containing blood pigment
(/'/) (Crescitelli, 1945). Figure 7 shows several cartilage rods at higher magnifi-
cation (ca.r). A sagittal section of the portion of a tentacle, together with its
associated pinnae, is seen in Figure 8. The central rod of cartilage (ca.r) gives off
strands of cells ( s.c.) which course to the very tips of the pinnae (pi). Note the
beautiful, almost plant-like regularity in the cellular bio-architecture of these tissues.
Note also that thin bundles of muscle (inu) run between each of the pinnae. This
muscular system undoubtedly aids in the beating movements made by the feathery
pinnae.
Toluidine-blue metachromasia (not illustrated) was seen irregularly dispersed
in matrix and cells, throughout both the cartilage and "osteoid-like" tissues. A
detailed description and analysis of the metachromatic behavior of the tissues will
not be given at this time.
POLYCHAETE ENDOSKELETAL CARTILAGE 249
Chemistry
Results of analysis are summarized in Table I. A more detailed presentation
will appear in a separate publication elsewrhere. The yield of acid mucopolysac-
charide, corrected for moisture, was 2.5% of the acetone-dried weight of the tissues.
The glucosamine present was probably due to a trace contaminant. The uronic acid
value was close to unity within the range of variability in colorimetric determina-
tions. The excess of nitrogen over unity was accounted for by residual peptide
bound to the polysaccharide, represented mainly by serine and glycine.. The sulfate
content was unusually high. The preparation was hydrolyzed much more slowly
by testicular hyaluronidase than was chondroitin sulfate of mammalian cartilage.
This difference in rates of hydrolysis was due to the preparation's high sulfate
content, since the product of chemical desulfation by the method of Kantor and
Schubert (1957) was rapidly hydrolyzed by the enzyme.
The contribution of the sulfate to the polyelectrolyte properties of the polysac-
charide is indicated by electrophoretic mobilities on cellulose acetate relative to
chondroitin sulfate of 1.05 at pH 7.0 and 1.25 at pH 3.0.
The infrared spectrum closely resembled that of chondroitin sulfate C, rather
than chondroitin sulfate A of vertebrate cartilage, indicating primarily equatorial
sulfate conformation (Mathews, 1958). A minor difference was revealed by a
peak at 700 cm.-1, possibly due to sulfate not present in chondroitin sulfate C.
The number average molecular weight was determined by osmometry as near
10,000; the intrinsic viscosity was 0.32.
A 40.0-mg. net weight sample of the basilar endoskeletal complex had an
acetone-dried weight of 10.1 mg., and contained 0.320 mg. of hydroxyproline.
DISCUSSION
It is somewhat surprising to find in an annelid an endoskeleton of such com-
plexity as is seen in Eudistylia. Nevertheless such endoskeletons are not uncom-
mon, as we have verified by examination of related polychaetes, including Myxicola,
Spirographis and Sabella; and of course, as is known from references given in the
opening paragraph of this paper.
The histology of these tissues will be discussed briefly from several standpoints.
To begin with, the cartilage is of the cellular variety, possessing relatively thin
seams of matrix. Such cartilages are seen amongst invertebrates in tissues such as
the odontophore cartilages of gastropod molluscs, and the gill cartilage of young
specimens of Limulus polyphemus (Person and Philpott, 1963a). Amongst verte-
brates such cellular cartilages are also widely encountered as in the vertebrae of
Mustelus (dogfish), ear and xiphisternal cartilage of the young white rat, etc. (see
also Schaffer, 1930). The cellularity or relative lack of matrix (as compared with
cartilages of the hyaline variety) permits a remarkably plant-like organization of
Eudistylia cartilage in certain regions of its endoskeletal complex (see especially
Figs. 6, 7 and 8).
The marked resemblances between plant tissues and cartilage have been of
interest to biologists since the time of Schwann and Schleiden (Schwann, 1839),
and seems to be forgotten and rediscovered in each generation [see Godman and
Porter (1960) and Person and Philpott (1963b)]. It is worth stressing that while
the correspondence of plant and animal cells and subcellular organelles has been
250 PHILIP PERSON AND MARTIN B. MATHEWS
recognized as a keystone for the understanding of cell form and function, still, at
the tissue level of biological organization and above, there are few instances in
which plant and animal structures can he compared, with the noteworthy exception
of cartilage and plant tissues! Such correspondences may be either fortuitous or
indicative of the existence, in cartilage and certain plant tissues, of important
similarities in fundamental biochemical and physiological processes occurring at
tissue levels. However, because there is so little communication between botanists
and zoologists concerning biological processes at and above the tissue level of
organization, and because these correspondences may have a deeper biological
meaning, a fuller and more detailed study of the correspondences between plant and
cartilage tissues is underway in our laboratory.
Also, Johnson (personal communication) has said that the "osteoid-like" tissue
which forms part of the endoskeletal complex of Eudistylia is very similar in appear-
ance to forms of osteoid seen in a variety of human skeletal tissue tumors.
Finally, consideration of the morphology of the skeletal tissues of Eudistylia
leads to an intriguing possibility which may be of especial interest to students of the
evolution of animal skeletal systems : If the cartilage and "osteoid-like" components
of the skeletal complex of Eudistylia were mineralized, then one would have a struc-
ture strongly resembling the outer armor of the vertebrate ancestors, the ostraco-
clerms. For example, in Figure 3, the tissue components ep and ma, if mineralized,
might strongly resemble enamel. Where the ma is penetrated by vascular channels
to give the osteoid like-tissues, os (Fig. 5), and in regions where cartilage is found,
the appearance of the hypothetically mineralized components could easily resemble
many of the tissue components seen in dermal armor of the ostracoderms (see
Gregory, 1951 ; Stensio, 1927 ; Bystrow, 1959; Denison, 1963). It is to be definitely
understood that we are not advocating that ostracoderms are derived from annelids.
But the potential conversion of portions of EudistyJia's skeletal system to an
ostracoderm-like armor is most interesting, and merits further investigation.
The chemical, enzymatic and physico-chemical data indicate that the acid muco-
polysaccharide of Eudistylia cartilage closely resembles chondroitin sulfate of mam-
malian cartilage in many respects, but differs in its higher sulfate content. Although
most vertebrate cartilage chondroitin sulfates have sulfate : hexosamine ratios of
1:1, exceptionally high sulfate contents similar to those reported for Eudistylia are
characteristic of preparations from cartilage of the coelacanth (Lathneria], mem-
bers of the class of chondrichthyes, and also the hagfish (My.rine) (Mathews, in
Press a). The presence of serine and glycine as the main residual amino acids
suggests that the acid mucopolysaccharide of Eudistylia tissue may be covalently
bound to peptide via serine hydroxyl (with an adjacent glycine residue) in a manner
similar to that found for chondroitin sulfate of vertebrate cartilage (Mathews, in
Press a). Also, the acid mucopolysaccharide level of Eudistylia cartilage is within
the range of chondroitin sulfate levels of most vertebrate cartilages, which vary
from 1.5% to 20%. Thus, Eudistylia endoskeletal tissue, while unusual with re-
spect to the excess sulfate content of its acid mucopolysaccharide component, never-
theless falls within the range of biochemical criteria for vertebrate cartilage. In
this connection, Eudistylia is also not unique among invertebrates, and resembles
both Loligo and Limulns, whose cartilages also contain very similar chondroitin
sulfates (Mathews, Duh and Person, 1962).
POLYCHAETE ENDOSKELETAL CARTILAGE 251
The high level of hydroxyproline, together with the histologic appearance and
staining properties of the tissues, are strongly indicative of the presence of collagen
in the endoskeletal tissues of Eudistylia. The value of 3.2% hydroxyproline is
comparable to the value of 3.5% hydroxyproline in embryonic (18-day) chick
articular cartilage (Mathews, in Press b).
The inability to obtain a gelatin which sets (when cooled) from invertebrate
cartilages, was thought by Krukenberg (1882), Schaffer (1930) and others to con-
stitute evidence that collagen was not present in the tissues. Invertebrate collagens
are quite widespread in occurrence (Gross, 1963), but at the present time no worker
has yet reported isolating a setting gelatin from any invertebrate connective tissue,
so that the gelation phenomenon cannot be considered a necessary criterion for, or
essential property of, invertebrate collagens. At present, the identification of
collagen (invertebrate or vertebrate) also depends upon a characteristic wide-
angle x-ray diffraction pattern (Gross, 1963). Such patterns are not yet available
for Eudistylia tissues. Nevertheless, the presence of hydroxyproline in Eudistylia
and other invertebrate cartilage tissues, as well as a characteristic light- and electron-
microscope appearance (Person and Philpott, 1963a), makes it very likely that
collagen is present. In the work reported by Person and Philpott (1963a) a col-
lagen with 650 A band space is shown in Busycon cartilage.
The inability of early workers to detect "mucins," i.e., chondroitin stilfate-con-
taining components, in polychaete cartilage (Krukenberg, 1882) and in other in-
vertebrate cartilages (Schaffer, 1930) was also used as a major argument against
the existence of cartilage tissues in invertebrates. With the advent of better methods
for extraction and identification of these acidic polysaccharides, this argument is
eliminated (present data; see also Mathews, Dull and Person, 1963, for data on
chondroitin sulfates of Limulus and Loligo cartilages).
In view of the above, it is believed that the controversy over the existence or
non-existence of "true" cartilage in invertebrates should be reopened, and should
be answered in the affirmative because the invertebrate tissues in question : ( 1 ) are
composed of cells suspended in a relatively rigid matrix of varying abundance, (2)
are rich in acidic polysaccharides including chondroitin sulfates, and (3) have a
high collagen content. The above criteria are also those by which vertebrate
cartilages are designated.
Supported by grants from the U. S. Public Health Service, The National
Foundation and The Chicago Heart Association.
LITERATURE CITED
BYSTROW, A. P., 1959. The microstructure of skeleton elements in some vertebrates from the
lower Devonian deposits of the U.S.S.R. Acta Zool., 40: 59-76.
CLAPAREDE, E., 1873. Recherches sur la structure des Annelides Sedentaires. H. Georg, Geneve.
CRESCITELLI, F., 1945. A note on the absorption spectra of the blood of Eudistylia gigantca and
of the blood pigment in the red corpuscles of Cucumaria miniata and Molpadia inter-
media. Biol. Bull, 88: 30-36.
DENISON, R. H., 1963. The early history of the vertebrate calcified skeleton. Clin. Orthopaed.,
31: 141-151.
EVENKAMP, H., 1931. Morphologic, Histologie und Biologic der Sabelliden-specified Laonome
kroveri Malmyn und Euchone papillose M. Sars. Zool. Jahrb. Jena, Abt. Anat., 53:
405-534.
252 PHILIP PERSON AND MARTIN B. MATHEWS
GODMAN, G. C., AND K. R. PORTER, 1960. Chondrogenesis, studied with the electron micro-
scope. /. Biofihys. Biochcm. Cytol, 8: 719-760.
GREGORY, W. K., 1951. Evolution Emerging. Vol. II, Figs. 6.2; 6.3; 6.4; 6.8. Macmillan Co.,
New York.
GROSS, J., 1963. Comparative biochemistry of collagen. In : Comparative Biochemistry, ed.
by M. Florkin and H. S. Mason, Vol. 5 Part C, pp. 307-346. Academic Press, New
York.
HEMPELMANN, F., 1928-34. In : Handbuch der Zoologie, ed. by W. Kiikenthal and T. Krum-
bach, Zweite band, Zweite Halft, pp. 1-212, DeGruyter and Co., Berlin.
HYMAN, L., 1940. The Invertebrates. Vol. 1. McGraw Hill Book Co., New York.
KANTOR, T. G., AND M. SCHUBERT, 1957. A method for the desulfation of chondroitin sulfate.
/. Amer. Client. Soc., 79: 152-153.
KRUKENBERG, C. F. W., 1882. Vergleichend-Physiologische Studien, pp. 49-53. C. Winter,
Heidelberg.
LASH, J. W., 1959. The presence of myoglobin in the cartilage of a marine snail, Busy con.
Science, 130: 334.
MATHEWS, M. B., 1958. Isomeric chondroitin sulfates. Nature, 181: 421-422.
MATHEWS, M. B., In Press a. The molecular evolution of cartilage. Clin. Orthopacd.
MATHEWS, M. B., In Press b. Chondroitin sulfate and collagen in inherited skeletal defects of
chickens. Nature.
MATHEWS, M. B., J. DUH AND P. PERSON, 1962. Acid mucopolysaccharides of invertebrate
cartilage. Nature, 193: 378-379.
MATHEWS, M. B., AND S. GLAGOV, 1966. Mucopolysaccharide patterns in ageing human carti-
lage. /. Clin. Invest.. 45: 1103-1111.
NICHOL, E. A. T., 1930. The feeding mechanism, formation of the tube and physiology of
digestion in Sabella pavonina. Proc. Roy. Soc. Edinburgh, 56: 537-598.
NOWIKOFF, W., 1912. Studien iiber das Knorpelgewebe von Wirbellosen. Zcitschr. f. iviss.
Zool, 103: 661-717.
PERSON, P., AND A. FINE, 1957. Observations on the histology and oxidative metabolism of
gill cartilage from Limulus polyphemus. Biol. Bull., 113: 331.
PERSON, P., AND D. E. PHILPOTT, 1963a. Invertebrate cartilage. Ann. N. Y. Acad. Sci., 109:
113-126.
PERSON, P., AND D. E. PHILPOTT, 1963b. Limulus gill cartilage: a "plant-like" tissue. Biol.
Bull., 125: 355-356.
PRITCHARD, J. J., 1956. General anatomy and histology of bone. In : The Biochemistry and
Physiology of Bone, ed. by G. H. Bourne, pp. 1-23, Academic Press Inc., New York.
RAVEN, CHR. P., 1958. Morphogenesis in Molluscs. Pergamon Press, New York.
ROMER, A. S., 1942. Cartilage, an embryonic adaptation. Amer. Naturalist, p. 394.
ROMER, A. S., 1964. Bone in early vertebrates. In : Bone Biodynamics, ed. by H. M. Frost,
pp. 13-37. Little, Brown & Co., Boston, Mass.
SCHAFFER, J., 1913. tiber den feinen Bau und die Entwicklung des Knorpelgewebes und uber
verwandte Formen der Stutzsubstanz. IV. Zcitschr. f. iviss. Zool., 105: 280-347.
SCHAFFER, J., 1930. Die Stutzgewebe. In : Handbuch der Mikroskopischen Anatomic des
Menschen, ed. by W. von Mollendorf, .Zweiter Band, Zweiter Teil, pp. 1-390. Julius
Springer, Berlin.
SCHWANN, T., 1839. Mikroskopische Untersuchungen ueber die Ubereinstimmung in der
Struktur und dem Wachstum der Tiere und Pflanzen. Berlin.
STEGEMANN, H., 1958. Mikrobestimmung von Hydroxyprolin mit Chloramin T und p-Di-
methylaminobenzaldehyd. Zeitschr. Physiol. Chem., 311: 41-45.
STENSIO, E., 1927. The Downtownian and Devonian vertebrates of Spitzbergen. Part I.
Family Cephalaspidae. Skrift Svalbard Nordishavet, no. 12: 1-391.
THOMAS, J. G., 1940. Pomatoceros, Sabella and Amphitrite (Liverpool Marine Biol. Comm.
Memoir No. 33). Proc. Trans. Liverpool Biol. Soc., 43: 1-88.
VIALLANES, M. H., 1885. Sur la structure de la squellette branchiale de la Sabelle. Ann. Sci.
Nat. Paris, Ser. 6, 20(2) : 1-20.
THE RELATIONSHIP OF TEMPERATURE TO THE
LARVAL DEVELOPMENT OF NASSARIUS
OBSOLETUS (GASTROPODA) *• -
RUDOLF S. SCHELTEMA
Woods Hole Occanograplric Institution, Woods Hole, Massachusetts 02543
Temperature has long been suggested as an important factor regulating the
developmental rate, length of pelagic life, and mortality of larvae from benthic
marine invertebrate organisms. It is known, for example, that the rate of early
cleavage, within certain limits, is related directly to water temperature (vide
Costello et al., 1957). There have been a number of attempts by marine biologists,
especially with species of economic value, to relate the success of settlement during
any specific year to the sea water temperature at the time of larval development.
Among bivalve mollusks, the oysters Crassostrea virginica Gmelin and Ostrea edulis
L. and the clam Venus mcrcenaria L. have particularly been studied (e.g., Needier,
1940; Medcof, 1939; Korringa, 1952; Carriker, 1961, pp. 212-213).
Seno, Hori and Kusakabe (1926) determined the effect of temperature on the
early development of Ostrea gigas from the time of fertilization to the early shelled
larva. Clark (1935) examined the effect of reduced temperature on the early
development of Crassostrea virginica. Not until the development of adequate tech-
niques for growing larvae in mass culture from fertilization to settlement (Allen
and Nelson, 1911; Bruce et al., 1940) has it been possible to examine experi-
mentally the relationship of temperature to the development of molluscan larvae.
Loosanoff et al. (1951) and Loosanoff (1959) were the first to demonstrate suc-
cessfully in the laboratory the role of temperature throughout the entire period of
pelagic larval development of the bivalve mollusk, Venus tnercenaria. Subsequently
studies of comparable detail have been made by Walne (1958) with Ostrea edulis;
by Davis and Calabrese (1964) with Crassostrea virginica; by Stickney (1964)
with Mya arenaria L. ; and by Bayne (1965) with Mytilus edulis L. The very
interesting research on gastropod larvae by Lebour (1937) did not include experi-
mental work using mass culture techniques, as her primary concern was the identifi-
cation and description of veligers from the plankton. Similarly, Thorson (1946,
1950) has followed the development of gastropod larvae by examining plankton
tows periodically taken from the 0>esund, but he has not attempted to undertake
laboratory culture work as a means of understanding the relative importance of
environmental factors on pelagic larval development. It is the purpose of this
paper to describe such a laboratory study using the common marine intertidal
prosobranch gastropod Nassarius obsoletus Say.
1 Contribution No. 1815 from the Woods Hole Oceanographic Institution, Woods Hole,
Massachusetts.
2 This research was supported in part by Grants 17883 and GB-2207 from the National
Science Foundation. I wish to thank my research assistant, Mr. Gordon Enk, for his help
during the conduct of some of the experiments here described.
253
254 RUDOLF S. SCHKLTEMA
Nassarius obsoletus inhabits marine and estuarine intertidal flats from Chaleur
Bay in the Gulf of St. Lawrence to Cape Kennedy (Canaveral) in northern Florida.
Although the early cleavage stages of development are well known to emhryologists
(vide: Clement, 1962; Thompson, 1955), the later planktotrophic veliger larvae
were not described until recently (Scheltema, 1962a). Experimental studies have
determined the role of salinity in larval survival and development (Scheltema,
1962b, 1965). Some mechanisms which control the length of pelagic life and the
delay of metamorphosis are also known from previous experiments (Scheltema,
1961). Aspects of the ecology of the adults have been discussed by Dimon (1905),
Jenner (1956a, 1957), and Scheltema (1964).
This study is divided into two parts : ( 1 ) the relationship of temperature to the
rate of early embryological development within the egg capsule, as indicated by the
time required for emergence of veliger larvae into the sea; (2) the relationship of
temperature to growth and length of planktonic larval life. Before giving an ac-
count of the experimental work, however, I shall describe briefly the reproductive
habits of N. obsoletus, as these have not previously been recorded in any detail.
REPRODUCTION AND SPAWNING
The onset of spawning in Nassarius obsoletus differs with latitude and is directly
related to sea water temperature. As the species inhabits an environment where
temperature can be highly variable over short periods, the exact timing of repro-
duction is never very precise. It can be shown, for example, that at Beaufort, North
Carolina, the water temperature on the intertidal flats in early February may differ
as much at 5° C. between high and low tide; a change from 13° to 17° C. has been
recorded in the area on the flats where females of N. obsoletus occur. At Barn-
stable Harbor, Cape Cod, Massachusetts, the low-water temperature on the flats
inhabited by N. obsoletus increases abruptly in a period of about two to three weeks
from 13° C. in mid-May to 23° C. in early June. That the females respond to
elevation of water temperature by spawning can easily be shown by bringing snails
into the laboratory during mid-winter months. Under such conditions, when the
animals are fed, spawning commences within a week. Copulation occurs during
the same period as spawning.
Under natural conditions the process of gametogenesis is completed long before
the normal time for spawning. This is known from frequent anatomical and
histological examination of snail gonads throughout the period extending between
the cessation of spawning and the completion of gametogenesis. Three geographi-
cally separated populations were followed : ( 1 ) Barnstable Harbor, Cape Cod,
Massachusetts; (2) Beaufort, North Carolina; and (3) Charleston, South Carolina,
the latter two in somewhat less detail than the former. In the northern end of the
range, gametogenesis usually proceeds within six weeks after the cessation of
spawning, that is, sometime during late September. However, in the southern end
of the range, spawning is completed by mid-June (Jenner 1956b), but the onset of
gametogenesis is apparently delayed for several months. This delay needs confirma-
tion by more frequent observations. There is no question, however, that both in
New England and at Beaufort, North Carolina, gametogenesis has been completed
by late fall, i.e., mid- to late November. The attainment of sexual competence can
TEMPERATURE AND LARVAL DEVELOPMENT
readily be determined externally in the living intact organisms by the structure of
the penis in the male and by the pigmentation at the end of the oviduct in the female.
Natural spawning normally begins in February at the southern end of the species
range, about mid- to end-April in Delaware Bay and the south shore of Cape Cod,
and early June in Cape Cod Bay and Maine. Consequently, gametogenesis is com-
pleted almost six months before the natural spawning of populations found north of
Cape Cod and at least two months before spawning in populations south of Cape
Hatteras, North Carolina. Ecologically it is doubtlessly advantageous for the
species to spawn as soon as the sea water becomes warm enough for larval develop-
ment and the early gametogenesis allows great flexibility in the time of spawning.
The egg capsules of Ar. obsoletits are deposited on any solid object on the inter-
tidal flats, e.g., shells, Diofatra tubes, thallus algae, etc. Ankel (1929) has described
in detail the deposition of egg capsules by the European species, Nassarius rcti-
culatus L., and this account agrees in every essential detail with the process as it
occurs in N. obsoletus. A description of the egg capsules of N. obsoletus, along
with the characteristics distinguishing them from other members of the genus found
along the east coast of the United States, has been given by Scheltema (1962a) and
by Scheltema and Scheltema (1965).
RELATIONSHIP OF TEMPERATURE TO EMBRYONIC DEVELOPMENT AND THE
ESCAPE OF LARVAE FROM THE EGG CAPSULE INTO THE SEA
The larvae of Nassarius obsoletus after the completion of their embryonic devel-
opment emerge through an opening at the free end of the egg capsule. The precise
method by which the opening is made by the larvae at the time of their escape is
not understood, but its position at the distal end is structurally pre-determined at
the time of capsule formation.
The relationship of temperature to the time required between spawning and
emergence of veligers from their egg capsules can be demonstrated by a simple
experiment. Adults of N. obsoletus readily lay egg capsules upon the sides of
aquaria. If, shortly after their deposition, several hundred egg capsules are col-
lected and placed at regular temperature intervals, falling within the extreme range
at which they are normally found in nature, the effect of temperature on develop-
ment can be determined. A number of such experiments were performed at
temperature intervals of 28°, 19.5°, 16.5° and 11.5° C. Egg capsules laid within a
48-hour period were collected from snails that had been actively laying for several
weeks. For the purpose of the experiment, the median age of the egg capsules was
considered to be 24 hours. The exact time of deposition of each capsule is not
particularly meaningful as the degree of development of the eggs within each
capsule is known to vary at the time of attachment. Early in the spawning period,
capsules are occasionally retained within the oviduct of the female until development
of the embryos is almost completed to the veliger stage, but as the season of spawn-
ing proceeds there is normally little delay between the initiation of embryological
development and egg capsule attachment. The rates of development at different
temperatures were observed simultaneously from random aliquots taken from the
same "harvest" of egg capsules. At the beginning of the experiment each capsule
was examined to make certain that it was intact and had not been damaged during
its removal from the walls of the aquarium. Between 250 and 300 egg capsules
256
RUDOLF S. SCHELTEMA
were used at each temperature. Starting with the time at which larval emergence
first began, the number of empty egg capsules in each container was determined at
frequent intervals.
The results obtained in one such experiment are shown by the series of curves
in Figure 1, in which the ordinate gives the cumulative percentage of capsules from
which larvae had emerged and the abscissa the number of days since the deposition
of the capsules. The curves represent development at each of the different temper-
100
16
FIGURE 1. Percentage egg cases of Nassariiis obsolctus from which larvae have emerged
relative to the time since spawning occurred. The curves represent development at four different
temperatures: 28.0° C. (A); 19.5° C. (O) ; 16.5° C. (D) ; and 11.5° C. «>). The abscissa
gives the time in days since the deposition of egg capsules ; the ordinate is the cumulative
percentage of egg capsules from which larvae have emerged. The values along the abscissa are
approximate (± one day) as the egg capsules were laid over a 48-hour period.
TEMPERATURE AND LARVAL DEVELOPMENT
257
30<
20<
10'
i
10
DAYS
12
14
18
FIGURE 2. Time in days required between spawning and emergence of Nassarius obsolctus
larvae from egg capsules as a function of temperature (° C). The points indicate the number of
days necessary for emergence of 50% of the larvae. Results are from two geographically
isolated regions, Beaufort, North Carolina (O) and Cape Cod, Massachusetts (D)- No
significant difference is discernible in the results between egg capsules obtained from the two
populations. The data are derived from experiments shown in Figure 1, and from similar
experiments using egg capsules from snails obtained from Beaufort, North Carolina. From 250
to 300 egg capsules were used at each temperature in experiments with the Cape Cod
populations. Approximately 100 egg capsules were used at each temperature in the Beaufort
experiments.
atures. No larvae emerged from the capsules held at 11.5° C. during the course
of the experiment. It has previously been shown that at a temperature between 11°
and 13° C., embryos do not complete their development but that a large proportion
remain viable for a period of up to at least nine weeks (Scheltema, 1962a). When
returned to warmer water such embryos developed normally.
The relationship between temperature and the time required for the emergence
of veliger larvae from the egg capsules is best understood by reference to Figure 2,
where the number of days required for the liberation of veliger larvae from the first
50% of the capsules is indicated along the abscissa, and the temperature (° C.) at
which the development took place is shown along the ordinate. Whereas the time
required for emergence increases slightly between 28° and 20° C., about 0.25 day
per degree centigrade, it increases more rapidly at temperatures below 20° C. ;
between 20° and 16.5° C., there was an increase of two days for each degree of
258 RUDOLF S. SCHKLTKMA
lowering of the temperature. The figure also shows that there is no significant
difference in the effect of temperature on egg development in populations of snails
from Beaufort, North Carolina, and from Cape Cod, Massachusetts. These results
differ from those of Dehnel (1955) ohtained from several intertidal species of
gastropods along the west coast of North America. He found that when embryos
collected from different geographical regions were allowed to cleave at an identical
temperature, there was a clear difference in the developmental rate ; the relationship
appeared to be clinal.
THE RELATIONSHIP OF TEMPERATURE TO GROWTH RATE
A method for obtaining large numbers of Nassarius obsoletus veliger larvae and
for growing mass cultures to be used in experimental work has already been
described (Scheltema, 1962a). The cultures used in the present experiments were
10 liters in volume, each containing from 5000 to 10,000 larvae. Food used
throughout the duration of these experiments was Phaeodactylum tricornutum
Bohlin which was obtained from unialgal cultures.
Larvae which had emerged from a large number of egg capsules over a 24-hour
period were divided equally among 10-liter containers. The number of larval
cultures started was determined by the number of temperatures at which growth
was to be measured. A sample of the larvae was also taken at the beginning of
each experiment so that the initial size after emergence from the egg capsule could
be determined. Each larval culture was fed an identical quantity of food (ca.
200,000 cells/ml.) from the same unialgal culture of P. tricornutum. This amount
of food was great enough so that a slight excess remained after three days.
A sample of from 50 to 100 larvae was removed from each culture every third
day. At this time the water was also changed and new food cells were added. The
aliquot of larvae was preserved in 70% alcohol for later measurement.
The growth of larvae was estimated by measuring the shell length of 35 speci-
mens picked randomly from the larger aliquot described above. An ocular microm-
eter at a magnification of 100 X was used. The longest dimension of the shell of
a larva was considered to be the length.
The temperature of the cultures was maintained by means of water baths impro-
vised from commercial soft drink coolers. The maximum deviation from the stated
mean was 1.5° C., but the mean deviation was only ±0.5° C. Because all the
experiments extended over more than two weeks, these variations were not con-
sidered serious.
In the first series of experiments, larvae were grown simultaneously at either
three or four different temperatures. The results from one representative experi-
ment are shown in Figure 3, where the mean temperatures were 16.5°, 21.0°, 24.8°
and 29.5° C. From this experiment it was concluded that the optimum temperature
for growth under laboratory conditions was approximately 25° C. This was further
verified in three other experiments. At either higher or lower temperatures the
growth was significantly less. That the rate of growth is not uniform throughout
larval development, particularly at optimum temperatures, can also be seen in
Figure 3. The lowest temperature at which larvae successfully grew to metamor-
phosis was between 16° and 17° C.
TEMPERATURE AND LARVAL DEVELOPMENT
259
To determine the maximum effect of temperature upon growth I made a second
series of experiments. The growth rate at approximately 25° C, an optimum tem-
perature, was compared with that at 17.5° C., a value near the lowest temperature
at which larval development is completed to settlement. The results of one such
series of experiments are shown in the growth curves in Figure 4. Here the upper
cumulative growth curve is from larvae grown at 25.2° C. ; the lower curve repre-
sents growth under similar conditions except that the temperature was 17.5° C.
The minimum length at which the veliger larvae have been shown to metamorphose
30
TEMP
°C
25
20
250
300
350
400 450
LENGTH
500
550
600
650
FIGURE 3. The effect of temperature upon the growth of the planktonic veliger larvae of
Nassarius obsoletus. The ordinate gives the temperature range over which the larvae were
grown. The abscissa gives the length attained by the larvae. Each curve shown by a solid
line represents the total cumulative growth completed by the larvae within the number of days
indicated by the numeral over the curve. The amount of growth for any length of time and
for any temperature which was tested can be easily derived from the figure. The points on
the solid lines represent the actual experimental values obtained. Curves given with dashed
lines represent the average growth at intervening days and were derived by linear interpolation.
lies between 550 and 600 /x, but the median size is near 700 p. On the graph in
Figure 4 the inflection points on both curves are at approximately 600 //.. In order
to compare growth rates between two temperatures it is clearly necessary to con-
sider only those portions of the curves which precede the points of inflection. After
the median size for metamorphosis is reached (i.e., 700 /A) relatively little further
growth occurs. The maximum recorded size at metamorphosis is 950 p., but this
size is rarely attained by larvae. The length of the period following the completion
260
RUDOLF S. SCHELTEMA
of development (i.e., the attainment of 700 /x) is primarily dependent on a settle-
ment response of the larvae. This is further discussed below.
All the experiments, including those of the first series above for which no data
have thus far been given, are summarized in Table I. The data from all these
4,000
900
800
700
LENGTH
600
500
400
300
I
I
t
20
30
DAYS
FIGURE 4. Cumulative growth curves of the planktonic veliger larvae of Nassarius obsoletus.
The upper curve (O) represents cumulative growth at an average temperature of 25.2° C. ;
the lower curve (A) represents growth of larvae at 17.5° C. The attainment of the "creeping-
swimming" stage is indicated on each curve by an arrow. This shows the end of the "develop-
mental period" to the left of arrow and the beginning of the "delay period" to the right of
arrow. Note that the "developmental period" is approximately twice as long at 17.5° C. (21
days) as it is at 25.2° C. (10 days).
TEMPERATURE AND LARVAL DEVELOPMENT
261
experiments cannot be directly compared because ( 1 ) the larvae were not randomly
obtained from the same parents and therefore are not known to be genetically
similar, and (2) the experiments were not conducted simultaneously, using the
same algal culture, so that the quality of the food was not necessarily the same.
However, the data can be pooled and the results compared by using the value /
which is the per cent inhibition of growth (Scheltema, 1965). This may be com-
puted from the equation
AA
where A A is the change in length of the shell between the beginning and end of
an experiment at a temperature optimum for growth (ca. 25° C.) and A B is the
TABLE I
The growth of Nassarius obsoletus larvae at a near optimal temperature and at a minimum
temperature required for completion of development, showing the maximum
inhibition attributable to temperature
Per cent
Expt.
no.
Age in
days at
end of
expt.
Length
ft at
begin,
expt. (a)
Mean
temp.
°C.
Length**
/i at end
• expt. (A)
Growth
M
(A4)
Mean
temp.
°C.
Length**
n at end
expt. (B)
Growth
It
(\B)
inhibition
of growth
AA ~ AB - 100
A
I
12
280*
25.1
657± 6
377
17.3
436±5
156
59
II
12
280*
25.0
655 ± 8
375
17.2
498 ±8
218
42
III
13
280*
24.5
698±10
418
17.7
488 ±7
209
50
IV
12
268
26.4
569 ± 5
301
17.5
496 ±5
228
24
V
12
268
26.4
672 ± 7
404
17.3
492 ±5
234
42
VI
17
271
23.9
615±11
344
15.8
447±7
176
49
VII
12
262
23.8
529± 7
267
15.9
422±8
160
40
VIII
9
268
24.6
463± 5
195
16.5
361 ±5
93
52
IX
12
268
25.3
589 ± 7
321
16.6
413±6
145
55
mean = 46
* Estimated values.
** One standard error of the mean is indicated.
change in shell length at a minimum temperature required for the completion of
development.
The value of A A is determined by subtracting the initial length of the veliger
larvae at the beginning of the experiment from the length attained when the experi-
ment was terminated. Hence
A A= (A - a)
where a is the initial length at the time the larvae emerged from the egg capsule
and A is the final length of the larvae when grown at 25° C. Similarly A B is
obtained by subtracting the initial length a, from B where B is the final length of
the larvae grown at around 16° or 17° C. The time at which each experiment
was terminated was determined by the inflection point on that curve which repre-
sented the culture having optimum growth (i.e., 25° C.).
262 RUDOLF S. SCHELTEMA
With a single exception the values of / fall between 40% and 60% and the mean
per cent inhibition of groii'th attributable to temperature was approximately 46%.
This represents the average maximum-difference which can be accounted for by
temperature alone.
DISCUSSION
As the onset of spawning by Nassarius obsoletus is dependent upon temperature,
its timing is never precise. Gametogenesis is always completed several months
before spawning occurs, and consequently a short period of warming can very easily
initiate spawning. Such conditions occur when the low tide falls near noon on
clear sunny days during early spring. Experimental evidence now show's, however,
that embryos can survive over long periods in cold water, but at a sharply reduced
developmental rate. Somewhat similar results have been obtained with Nassarius
rcticulatus from the Black Sea (Bekman, 1941). It is very unlikely, from experi-
mental evidence, that embryonic development of N. obsoletus into free-swimming
veliger larvae is ever completed, under conditions in nature, before the water
temperature rises high enough to insure the completion of pelagic development.
Although an optimum growth of planktonic larvae in the experiments occurred
at 25° C, it is not clear whether this was an intrinsic characteristic of the veligers
themselves or whether growth was indirectly influenced by the effect of temperature
on the algal food. Phaeodactylum tricornutum does not long survive at tempera-
tures above 25° C. However, as the larvae were fed fresh algal cells every third
day and since an excess always remained in suspension at the end of this period, it
was believed that this effect must have been minimal. It is not possible, however,
to rule out such an indirect factor in the experimental results. Davis and Calabrese
(1964) have suggested that enzymes required to digest naked flagellates are active
at much lower temperatures than those involved in the digestion of certain other
food forms having thick cell walls. Very few dinoflagellates or diatoms can grow
and survive equally well at temperatures of 15° and 30° C. ; both their numbers
and food value to larvae may differ markedly at either of these extremes. It is
necessary, when relating experiments from the laboratory to natural conditions, to
take into account the effect of temperature on the principal phytoplankton organisms
upon which the larvae are likely to be feeding.
The length of pelagic larval life in some bivalves seems to be directly related
to the temperature and growth rate. Thus Loosanoff (1959) has shown that the
increase in pelagic larval life of Venus tncrcenaria is directly related to the decrease
in temperature. However, the results are somewhat obscured because the criterion
used to determine the length of larval life was the number of days required for
settlement to first begin. Such a criterion largely neglects the effect of a delay in
settling due to the lack of a desirable substratum, if indeed Venus mercenaria has
such a delay. Davis and Calabrese (1964, p. 648) have shown that in Crassostrea
virginica the last larvae to settle in their cultures usually have a planktonic life
almost two times as long as the earliest veligers to metamorphose. Bayne (1965)
has demonstrated that Mytilus e dulls in the absence of an adequate substratum for
settlement delays metamorphosis, and that this delay is accompanied by a gradual
decrease in growth rate to zero.
The larval life of Nassarius obsoletus can be divided into two periods. The
TEMPERATURE AND LARVAL DEVELOPMENT 263
first of these is one of rapid growth and morphological development and will be
termed the "developmental period." This is followed by a second period, the
"delay period," during which there is a gradual decrease in growth. The "develop-
mental period" ends at the inflection point on the cumulative growth curve (Fig. 4).
External morphological development has been completed to the creeping-swimming
stage (Scheltema, 1962a). The growth rate during the "developmental period" is
essentially constant if the environment remains reasonably so. Two physical factors
that are important in determining the slope of the growth curve, and consequently
the length of the "developmental period," are temperature and, under certain cir-
cumstances, the salinity of sea water (Scheltema, 1965). At the end of the "devel-
opmental period," metamorphosis first becomes possible. The "delay period" which
follows may vary greatly in its length. Its duration is largely determined by the
availability of the bottom sediment favorable for further post-larval life. The evi-
dence for delay in settlement and a response to bottom sediment in N. obsoletus has
already been given in previous papers (Scheltema, 1956, 1961).
SUMMARY
1. Development of the embryos of Nassarins obsoletus within egg capsules is
regulated by sea-water temperature. An increase in the time required between
spawning and the emergence of veliger larvae is slight between 28° and 20° C.,
about 0.25 day for each degree decrease in temperature. Between 20° and 16.5° C.,
the corresponding increase was 2 days per degree decrease in temperature. At
11.5° C., development was not completed and larvae did not emerge from their
egg capsules after nine weeks. However, a large proportion of these embryos
survived and developed normally through metamorphosis when placed at room
temperature.
2. The growth rate of planktonic veliger larvae of N. obsoletus was greatest at
approximately 25° C. The lowest temperature at which the development to meta-
morphosis was completed was at 16° to 17° C. There was a 46% inhibition in
the growth rate of larvae between the optimum temperature and the minimum
temperature at which development is completed.
3. The larval life of N. obsoletus veligers may be divided into two stages. The
first of these, the "developmental period," is one during which rapid growth and
morphological development occur. This is followed by the "delay period" char-
acterized by a gradual decrease in growth rate. Reduced temperature may influence
the rate of growth and consequently the length of the "developmental period." The
termination of the "developmental period" comes with the "creeping-swimming
stage." The duration of the "delay period" may be quite variable and is determined
by the availability of a favorable sediment for settlement.
LITERATURE CITED
ALLEN, E. J., AND E. W. NELSON, 1911. On the artificial culture of marine organisms. /. Mar.
Biol. Assoc.,S: 421-474.
ANKEL, W. E., 1929. Uber der Bildung der Eikapsel bei Nassa-arten. Verhandl. Dcntxch.
Zoo/. Gesellschajt, Zoo/. Anz. (Suppl. Baud}. 4: 219-230.
BAYNE, B. L., 1965. Growth and the delay of metamorphosis of larvae of Mytilus editlis L.
Ophelia (Copenhagen), 2: 1-47.
264 RUDOLF S. SCHELTEMA
BEKMAN, M. J., 1941. On the biology of the marine gastropod, Nassa rcticulata (Monter.) and
Nassa (Cydonassa) ncritca. Bull. Acad. Sci. U.R.S.S. (Scr. Biol.), 3: 347-352 (in
Russian).
BRUCE, J. R., M. KNIGHT AND M. W. PARKER, 1940. The rearing of oyster larvae on an algal
diet. /. Mar. Biol. Assoc., 24: 337-374.
CARRIKER, M. R., 1961. Interrelation of functional morphology, behavior, and autecology in
early stages of the bivalve, Mcrccnaria mcrccnaria. J. Elisha Mitchell Sci. Soc., 77:
168-241.
CLARK, A. E., 1935. Effects of temperature and salinity on early development of the oyster.
Prog. Rept., Atl. Biol. St., St. Andrezvs, N. B., 16: 10.
CLEMENT, A. C, 1962. Development of Ilyanassa following the removal of the D macromere at
successive cleavage stages. /. Exp. Zool., 149: 163-216.
COSTELLO, D. P., M. E. DAVIDSON, A. EGGERS, M. H. Fox AND C. HENLEY, 1957. Alethods for
Obtaining and Handling Marine Eggs and Embryos. Marine Biological Laboratory,
Woods Hole, 247 pp.
DAVIS, H. C., AND A. CALABRESE, 1964. Combined effects of temperature and salinity on
development of eggs and growth of larvae of M. mcrcenaria and C. virginica. Fishery
Bull. (U. S. Fish and Wildl. Ser.), 63: 643-655.
DEHNEL, P. A., 1955. Rates of growth of gastropods as a function of latitude. Ph\siol. Zool.,
28: 115-144.
DIMON, A. C., 1905. The mud snail: Nassa obsoleta. Cold Spring Harbor Monogr., 5: 1-48.
JENNER, C. E., 1956a. A striking behavioral change leading to the formation of extensive
aggregations in a population of Nassarhis obsolctus. Biol. Bull., Ill: 291.
JENNER, C. E., 1956b. The timing of reproductive cessation in geographically separated popula-
tions of Nassarius obsoletus. Biol. Bull., Ill: 292.
JENNER, C. E., 1957. Schooling behavior in mud snails in Barnstable Harbor leading to the
formation of massive aggregations at the completion of sexual reproduction. Biol.
Bull, 113: 328-329.
KORRINGA, P., 1952. Recent advances in oyster biology. Quart. Rev. Biol., 27: 266-308; 339-
365.
LEBOUR, M. V., 1937. The eggs and larvae of the British Prosobranchs with special reference
to those living in the plankton. /. Mar. Biol. Assoc., 22: 105-166.
LOOSANOFF, V. L., 1959. The size and shape of metamorphosing larvae of Venus (Mercenaria)
mercenaria grown at different temperatures. Biol. Bull., 117: 308-318.
LOOSANOFF, V. L., W. S. MILLER AND P. B. SMITH, 1951. Growth and setting of larvae of
Venus mcrcenaria in relation to temperature. /. Mar. Res. (New Haven), 10: 59-81.
MEDCOF, J. C., 1939. Larval life of the oyster (Ostrca virginica) in Biddeford River. J. Fish.
Res. Bd. Canada., 4: 287-301.
NEEDLER, A. W. H., 1940. Helping oyster growers to collect spat by predicting sets. Progr.
Rcpt. Atl. Biol. Sta., 27: 8-10.
SCHELTEMA, R. S., 1956. The effect of substrate on the length of planktonic existence in
Nassarius obsolctus. Biol. Bull., Ill: 312.
SCHELTEMA, R. S., 1961. Metamorphosis of the veliger larvae of Nassarius obsolctus (Gastro-
poda) in response to bottom sediment. Biol. Bull., 120: 92-109.
SCHELTEMA, R. S., 1962a. Pelagic larvae of New England intertidal gastropods. I. Nassarius
obsoletus Say and Nassarius vibex Say. Trans. Amer. Microsc. Soc., 81: 1-11.
SCHELTEMA, R. S., 1962b. Environmental factors affecting length of pelagic development in
the gastropod, Nassarius obsolctus. Amer. Zool., 2: 445.
SCHELTEMA, R. S., 1964. Feeding and growth in the mud-snail Nassarius obsoletus. Chesa-
peake Sci., 5: 161-166.
SCHELTEMA, R. S., 1965. The relationship of salinity to larval survival and development in
Nassarius obsoletus (Gastropoda). Biol. Bull., 129: 340-354.
SCHELTEMA, R. S., AND A. H. SCHELTEMA, 1965. Pelagic larvae of New England intertidal
gastropods. III. Nassarius trivittatus. Hydrobiologia, 25 : 321-329.
SENO, H., J. HORI AND D. KUSAKABE, 1926. Effects of temperature and salinity on the develop-
ment of the eggs of the common Japanese oyster, Ostrea gigas Thunberg. /. Fish. Inst.
Tokyo, 22 : 41-47.
TEMPERATURE AND LARVAL DEVELOPMENT 265
STICKNEY, A. P., 1964. Salinity, temperature and food requirements of soft-shell clam larvae
in laboratory culture. Ecology, 45: 283-291.
THOMPSON, J. R., 1955. The embryology and organogeny of the common eastern mud snail,
Nassariits obsolctits Say. M. S. thesis, Duke Univ., Durham, North Carolina.
THORSON, G., 1946. Reproduction and larval development of Danish marine bottom inverte-
brates with special reference to the planktonic larvae in the sound (0resund). Medd.
K 0111 in. Damn. Fiskeriog Hai'itudcrs. Scr. Plankton, 4: 1-523.
THORSON, G., 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol.
Rev. ,2$: 1-45.
WALNE, P. R., 1958. Growth of oysters (Ostrca ednlis L.). /. Mar. Biol. Assoc., 37: 415-425.
THE MORPHOLOGY, LIFE-HISTORY, AND SYSTEMATIC
RELATIONS OF THE DIGENETIC TREMATODE,
UNISERIALIS BREVISERIALIS SP. NOV.,
(NOTOCOTYLIDAE), A PARASITE
OF THE BURSA FABRICIUS
OF BIRDS1
HORACE W. STUNKARD
American Museum of Natural History, Central Park Urest at 79th Street,
Neu> York, New York 10024
The genus Uniscrialis was erected by Mary Beverley-Burton (1958) to contain
Uniserialis gippyensis, a new species from the intestinal caeca and bursae Fabricii
of mallard ducks. Anas platyrJtyncha platyrhyncha Linn., taken near Ipswich, Suf-
folk, England. The generic diagnosis stated, "Notocotylidae Liihe, 1909; body
flattened, small, rather pointed at the anterior end but rounded posteriorly ; cuticula
aspinose. Ventral surface with a single median longitudinal row of sessile glands.
Oral sucker terminal, pharynx absent ; esophagus short leading to intestinal bifurca-
tion, ceca simple ending blindly near posterior extremity. Ventral sucker absent.
Common genital pore ventral, median, anterior to intestinal bifurcation, near oral
sucker. Two lobed testes, posterior and extracecal in position : external vesicula
seminalis well developed. Cirrus sac elongate with internal vesicula seminalis ; cir-
rus unarmed. Ovary median, between the testes, immediately posterior to Mehlis'
gland; receptaculum absent. Uterus with ascending limb only, forming intracecal
transverse slings. Metraterm long, with thickened walls and opening at the genital
pore. Vitellaria follicular, in two lateral extracecal bands, running forward from
anterior border of testes. Excretory pore dorsal and posterior, receiving the two
main excretory vessels. Eggs numerous, small and operculate with long polar fila-
ments. Adults in intestinal ceca and bursa Fabricius of birds. Genotype : U.
gippyensis n. sp."
The genus was included in the subfamily Notocotylinae, family Notocotylidae,
and distinguished from other genera: Notocotylus Diesing, 1839; Catatropis
Odhner, 1905; Parainoiwstoiiiiiin Liihe, 1909; Quinqueserialis Skvorzov, 1934;
Hofmonostomum Harwood, 1939; and Tristriata Belopolskaia, 1953
MATERIAL AND METHODS
During the summer months of 1963, 1964, 1965 and 1966, over 5000 specimens
of Hydrobia salsa were examined for infection by larval trematodes. The snails
were identified by Dr. W. K. Emerson of the American Museum, New York. This
is a somewhat rare, prosobranchiate species, described by Pilsbry (1905) as
1 Investigation supported by Grant NSF-GB-3606, continuation of G-23561.
266
LIFE-CYCLE OF UNISERIALIS
267
Paludcstrlna salsa. During the summers of 1963 and 1964, these snails were
common in Nobska Pond near \Yoods Hole, Massachusetts, and most of the speci-
mens were taken from an area near the connection of the pond with Vineyard
Sound. In the fall of 1964, the pond was "treated" and most of the invertebrates,
including snails, were killed. In the summer of 1965. H. salsa was found in nearby
Oyster Pond and the study was continued. The results have been rewarding; ten
different species of larval trematodes have been recognized ; five of them are
notocotylid cercariae. The methods and procedures employed have been described
in earlier reports (Stunkard, 1960, 1966a, 1966b, 1967) on the morphology and life-
cycles of notocotylid species. In the (1966a) paper, the writer reported that the
five notocotylid cercariae included representatives of all three larval types dis-
Plate I
FIGURE 1. Uniso-iolis brcriscrialis, adult specimen, ventral view, somewhat flattened, fixed
and stained, 2.3 mm. long, from the bursa of a domestic duckling, 12 days after metacercaria
was fed.
FIGURE 2. Uniscrialis <iipp\cnsis Beverley-Burton, 1958, paratype specimen, ventral vie\v,
1.62 mm. long, from bursa of Anus platyrhyncha. The ventral glands are not visible in this
specimen ; their size and location on other worms were determined and added in the drawing.
268
HORACE W. STUNKARD
Plate II
LIFE-CYCLE OF UNISERIALIS 269
tinguished by Rothschild (1038) on differences in the structure of the excretory
system and designated as the Yenchingensis, the Monostomi, and the Imbricata
Groups. Two of the cercariae belong to the Yenchingensis Group ; they develop to
maturity in the digestive caeca of ducklings and were described by Stunkard
(1966b) as Xototvlus uihuttus Stunkard and Dunihue. 1931 and Notocotylns at-
lanticits sf>. nor. Two cercariae belong to the Monostomi Group ; they develop in
the lumen of the intestine of chicks and ducklings and were described by Stunkard
(1967) as Parainoiwstoinitiii ak'catinn ( Mehlis in Creplin, 1846) and Paramono-
stomnin parvnm Stunkard and Dunihue, 1931. The fifth cercarial species belongs
to the Imbricata Group ; these larvae develop in the bursa Fabricius of chicks and
ducklings, and form the subject of the present report.
These worms are similar in many respects to those from the intestinal caeca
and bursae Fabricii of mallard ducks, described by Beverley-Burton (1958) as
Uniserialis gippyensis ; compare Figures 1 and 2. According to Beverley-Burton,
the worms described as U. </i[>[>yensis have only one row of ventral glands, median
in position and five in number. The present specimens have the five median glands
and, in addition, four small lateral glands on either side, situated in the intervals
between the median glands. These small lateral glands are recognizable in most
living specimens but are rarely visible in fixed and stained preparations. Indeed,
the large, median glands can not be seen in many whole-mount individuals. It
appeared possible that the present worms are congeneric with those described by
Beverley-Burton and in an attempt to resolve the problem, the specimens of U.
</ipp\cnsis deposited in the British Museum (Natural History) were borrowed
through the kindness of Mr. Stephen Prudhoe. The material in the British Museum
consists of one slide on which there are three specimens, all lightly stained. The
largest and apparently the most representative specimen, ringed on the coverglass,
is reproduced (Fig. 2 ) ; it is 1.62 mm. long; 0.81 mm. wide; oral sucker, 0.14 mm.
in diameter ; distance from anterior end to base of cirrus sac, 0.66 mm. ; testes
measure 0.25 by 0.19 mm. ; ovary, 0.13 mm. in diameter; vitellaria extend 0.32 mm.
and are situated in the posterior half of the body. In this specimen the median
glands are not recognizable although they can be seen on the other two worms.
The other worms are slightly smaller but very similar, and all agree completely
with the specific description as given by Beverley-Burton. No lateral glands are
visible on any of the specimens, but the staining is faint and it is probable that if
present, they would not be recognizable.
The Imbricata cercariae are the largest of the notocotylid larvae found in H.
salsa. They emerge principally after 11 AM and some were swimming at 4 PM.
but most of them encysted after swimming for an hour or two. Feeding of encysted
metacercariae to chicks and domestic Pekin ducklings yielded developing and
FIGURE 3. U. brcriscrialis. redia, natural infection, fixed under a coverglass, 0.87 mm. long.
FIGURE 4. U. breviserialis, juvenile specimen, flattened under a coverglass, stained and
mounted, 0.75 mm. long, four days in domestic duckling.
FIGURE 5. U. breviserialis, juvenile specimen, flattened under a coverglass, stained and
mounted, 1.12 mm. long, six days in a domestic duckling.
FIGURE 6. U. breviserialis, young cercaria, from fixed and stained specimen, flattened
under a coverglass, with details added from sketches of living specimens.
FIGURE 7. U. brcriscrialis. juvenile specimen, flattened under a coverglass, stained and
mounted, 1.56 mm. long, nine days in a domestic duckling.
270 HORACE W. STUNKARD
gravid worms, all of which were taken from the bursa Fabricius. The worms adhere
tenaciously to the wall of the hursa and when removed, the caeca are hright red
with hloocl from the host. The corpuscles are clearly visible in fixed and stained
worms. Both median and lateral ventral glands are visible in juvenile worms, six
days in a chick, as well as in sexually mature specimens. A series of drawings
(Figs. 4-, 5 and 7) shows development of worms from four to nine days in the
bursa of domestic ducklings. The smallest sexually mature worm, fixed without
pressure, is 1.19 mm. long and 0.65 mm. wide; the vitellaria extend through a
distance of 0.24 mm.; the testes measure 0.18 by 0.16 mm.; the cirrus sac is 0.19
mm. long and 0.055 mm. wide; the ovary is 0.16 mm. long and 0.072 mm. wide;
Mehlis' gland is 0.080 mm. wide and 0.060 mm. long. The metacercariae were fed
August 14, 1965, and the bird was autopsied August 30, 1965.
DESCRIPTIONS
Adult (Fig. 1)
The body is ovate, much flattened, more pointed anteriorly and rounded pos-
teriorly. It is convex dorsally, concave ventrally, with the edges of the body turned
ventrad and mediad. Fixed and stained specimens measure 1.19 to 2.54 mm. in
length and 0.56 to 1.66 mm. in width. The cuticula is thin ; fine spines were observed
on the ventral surface of living worms but they do not show on fixed and stained
specimens. The body wall is delicate, the musculature is weak ; the longitudinal
muscles are best developed. The ventral surface bears five median, protrusible
glands that are conspicuous in living specimens. They are circular to oval in
outline with transverse slit-like openings. The anterior and posterior glands
measure 0.12 to 0.15 mm. in diameter; the three middle ones are somewhat larger
and measure 0.16 to 0.20 mm. in diameter. The most anterior gland is situated
about its diameter posterior to the base of the cirrus sac ; the second gland is at the
level of the anterior ends of the vitellaria ; the third gland is near the middle of the
vitelline zone ; the fourth gland is at or slightly posterior to the level of the caudal
ends of the vitellaria ; and the most posterior gland is at the ovarian level. The
lateral glands are smaller than the median ones; they measure 0.09 to 0.12 mm. in
diameter, and are situated in the intervals between the median ones. The three
anterior ones are in the fields of the digestive caeca and the most posterior glands
are ventral to the antero-median lobes of the testes.
The excretory pore is dorsal, about midway between the ovary and the posterior
end of the body. The bladder is small and the collecting ducts extend forward,
forming a loop that crosses the body anterior to the cerebral ganglia. Dendritic
branches from the longitudinal ducts form a complex network of excretory channels.
The oral sucker is terminal, 0.13 to 0.20 mm. in diameter; the mouth is slightly
ventral ; the esophagus is short, about the length of the sucker ; the caeca extend
posteriad, lateral to the uterine loops, turn mediad to pass between the testes and
ovary, and end blindly posterior to the testes.
The testes are situated in the extracaecal areas near the posterior end of the
body. They are deeply lobed, and vary in size from 0.18 by 0.16 mm. to 0.50 by
0.375 mm. Sperm ducts arise at the antero-median faces and unite a short distance
anterior to Mehlis' gland to form the vas deferens which passes anteriad, dorsal to
LIFE-CYCLE OF UNISERIALIS
the uterus. At about the level of the anterior ends of the vitellaria, it expands to
form a coiled external seminal vesicle which continues the forward course to the
cirrus sac. A coiled internal seminal vesicle occupies the posterior third to one-half
of the cirrus sac and is continued by the ejaculatory duct ; both are enclosed in
prostatic cells. The cirrus sac leads to the genital pore, located anterior to the
cerebral ganglia and at or near the level of the posterior border of the oral sucker.
The cirrus sac is dorsal to the metraterm and measures from 0.19 to 0.40 mm. in
length and 0.055 to 0.15 mm. in greatest width. It is located in the anterior fourth
of the body.
The ovary is lobed, usually longer in the antero-posterior axis and increases in
size as the worm matures. In a young specimen it may be 0.16 by 0.072 mm. and in
a large, fully mature worm it may be 0.28 by 0.20 mm. The oviduct arises at the
antero-dorsal margin and receives a common vitelline duct as it enters Mehlis'
gland, which is somewhat smaller and immediately anterior to the ovary. There
is no seminal receptacle and the initial coils of the uterus are filled with spermatozoa.
The uterus passes forward in intercaecal, transverse loops, 15 to 25 in number, to
communicate with the metraterm. The metraterm is somewhat shorter than the
cirrus sac, is ventral in position, has a weak muscular wall, and opens at the genital
pore posterior to the opening of the cirrus sac. The vitellaria consist of 15-20
discrete, irregularly shaped follicles which occupy the extracaecal areas from the
testes to the level of the external seminal vesicle. They extend through a distance
of 0.22 to 0.62 mm. and in large part are situated in the middle third of the body.
Collecting ducts course posteriad along their median faces and at the posterior end
of the vitellaria pass mediad, ventral to the caeca, then turn dorsad, joining above
Mehlis' gland to form a vitelline receptacle from which the short common duct
leads to the oviduct. The eggs are operculate, 0.019 to 0.020 mm. long, 0.011 to
0.013 mm. wide, provided with long polar filaments, and embryonated when passed.
Rcdia (Fig. 3)
The rediae are oval to sausage-shaped to elongate ; extended, they are cylindrical
with conical posterior ends. Small rediae are colorless, actively motile but without
feet, and one which measured 0.10 mm. in length contained a small daughter as well
as germ balls of developing cercariae. Locomotion is accomplished by contraction
of the circular and logitudinal muscles of the body wall. As the rediae grow and
become filled with progeny, movement is less and less active. Older rediae have
orange-yellow droplets, 0.002 to 0.006 mm. in diameter, in the body wall ; the
largest extend to a length of 1.00 mm. ; the specimen shown in Figure 3, fixed and
stained, is 0.87 mm. long and 0.23 mm. wide. In the older rediae the pharynx
measures 0.042 to 0.052 mm. in diameter, the esophagus is about the same length,
and the intestine, which extends to the middle of the body in young rediae, is
restricted to the anterior third or fourth of the body length. The birth pore is
ventral at the level of the esophagus. There are two excretory pores, one on either
side near the middle of the posterior half of the body. From each pore a duct passes
forward, just past the middle of the body where it divides into anterior and posterior
branches. Each branch terminates in a flame-cell, one at the level of the esophagus,
the other posterior to the excretory pore.
272 HORACE W. STUNKARD
C'crcaria ( Fig. 6)
The cercariae are large ; they emerge from the rediae in very immature condi-
tion, about one-half the size they eventually attain. On emergence from the redia
into the haemal sinus of the snail, the tail is so small and weak that the cercaria
cannot swim if liberated by crushing the snail. Figure 3 is made from sketches of
a young specimen. Alive, it extended to a length of 0.30 mm., fixed and stained it
is 0.20 mm. long and 0.12 mm. wide. In it the excretory ring is complete but the
common stem extends into the tail and the excretory pores are situated on the sides
of the tail. The excretory system develops in the manner described for Imbricata
cercariae by Rothschild ( 1938: Figs. 30, 31, 32, 34, 38), with the primary collecting
ducts fusing anteriorly to form the loop that extends across the body anterior to
the cerebral ganglia and the median eye-spot. As the cercaria matures in the
haemal sinuses of the snail, the ring becomes filled with the concretions 0.003-0.005
mm. in diameter and the portion of the excretory system in the tail atrophies as a
new definitive excretory pore develops from the dorsal wall of the expanding
excretory bladder. The study of the flame-cell pattern has been disappointing. It
is probable that the formula is 2 [ (3 + 3 + 3 ) + ( 3 + 3 + 3 )), as reported by Martin
(1956) and Odening (1966) for other notocotylid cercariae. In young cercariae
the formula is 2 |(1 + 1 + 1) + (1 + 1 + 1)], and in older cercariae the anterior and
posterior groups each have three cells, but the cystogenous cells fill so early that not
all the flame-cells and capillaries have been observed in the mid-region of the body.
As noted, the cystogenous cells fill the parenchyma and obscure other structures;
the secretion appears in the form of short, bacilliform rods. Normally emerged
cercariae vary from 0.30 to 0.60 mm. in length and 0.14 to 0.25 mm. in width. The
postero-lateral ends of the body bear eversible, retractile locomotor appendages
which function in creeping movements of the body. \Yhen the body is extended,
they are close together, separated only by the base of the tail which is between and
ventral to them ; as the body contracts they separate and serve as fulcra for the next
extension of the body. The tail is simple, slender, 0.04 to 0.08 mm. in width at
the base, and about the same length as the body. When either is contracted, the
other is elongated. The wall of the tail is composed of external circular and
internal longitudinal muscles which enclose loose parenchymal tissue. In swim-
ming, the body is contracted, bent ventrally, almost circular, while the tail is
extended and lashes vigorously. The ocelli are formed while the cercariae are in
the rediae (Fig. 3), and increase to a diameter of 0.016 to 0.024 mm.; they are
provided with lenses and are connected by short nerves to the cerebral ganglia.
The median eye-spot, usually lacking in young cercariae, often becomes well
organized as a dark ring in emerged individuals. Diffuse, dendritic strands of
pigment surround the ocelli, permeate the anterior end of the body and extend
posteriorly, especially along the digestive caeca. The oral sucker measures 0.04 to
0.05 mm. in diameter, the esophagus is about the same length ; it passes backward
below the commissure of the nervous system and above the excretory ring. Im-
mediately behind the level of the cerebral ganglia it communicates with the intestinal
caeca. The caeca extend posteriad, dorsal and medial to the excretory ring, which
they cross near the posterior end of the body to terminate in the extracaecal areas.
Deeply staining germinal cells, situated immediately anterior to the caudal end of
LIFE-CYCLE OF UNISERIALIS 273
the excretory ring, are the primordia of the gonads, and a strand of these cells
extends anteriorly in the median plane.
Metacercaria
Infected snails were identified by isolation. The cercariae begin to emerge about
1 1 :00 AM and swim toward the light side of the container. By 3 :00 PM, almost
all are encysted, on the shell of the snail from which they emerged, the wall of the
container, or on strands of algae. The cysts are the largest of the five notocotylid
species and average measurements are 0.195 mm. in external and 0.175 mm. internal
diameter. The worms do not develop in their cysts ; they are infective immediately
and become sexually mature in about two weeks in the bursae Fabricii of ducklings
and chicks.
DISCUSSION
The present specimens are very similar, morphologically, to Uniserialis gippv-
ensis Beverley-Burton, 1958. They are from the same site, the bursa Fabricius of
birds, and from the same or related host species. The principal difference is the
presence on the ventral surface of lateral glands which were not described for U.
gippycnsis. But these glands are rarely visible in fixed and stained specimens. If
they do occur in U. gippyensis, the present specimens are obviously congeneric with
those of Beverley-Burton, and on that presumption, they are described as a new
species, Uniserialis breviserialis. Type and paratype specimens are deposited in
the Helminthological Collection of the U. S. National Museum under the numbers
61,186 and 61,187. Specific differences between U. gippvcnsis and U. brcriscrialis
are recognizable in the length of the cirrus sac and the location of the gonads and
vitellaria. In U. gippycnsis the cirrus sac is about twice as long ; it extends one-
third of the length of the body ; in U. brcriscrialis it is short, less than one-fourth of
the body length. In Lr. gippyensis the reproductive organs are more posteriorly
situated ; the vitellaria are in the posterior halt" of the body whereas in U. brcri-
scrialis the vitellaria are situated largely in the middle third of the body.
The validity of the genus Uniserialis is questionable. Baer and Joyeux (1961 )
suppressed Uniserialis as identical with Notocotylits Diesing. 1839 and the presence
of lateral glands on the ventral surface of the present specimens seemingly supports
that action. But there are other considerations which may validate the genus
Uniserialis. Miriam Rothschild (1938) recognized three types of notocotylid cer-
cariae, designated the Yenchingensis, the Monostomi, and the Imbricata Groups,
respectively, based on the structure of the excretory system. Stunkard (1966a)
found that Notocotylits ininiitits and Notocotylits athuiticns have Yenchingensis-
type cercariae and develop in the intestinal caeca; that Paraiiwnostoiiiinn alreatinn
and Paramonostomum parr inn have Monostomi-type cercariae and develop in the
lumen of the intestine; whereas the present species, Uniserialis brcriscrialis, has
Imbricata-type cercariae and localizes in the bursa Fabricius. The apparent cor-
relation of cercarial type, generic allocation and sites of infection is disturbed by
the report of Rothschild (1941 ) that two species of Yenchingensis-type cercariae
developed in the intestinal caeca of ducks into flukes of the genus Paraiiionostoiiiuiii.
Furthermore. Odening ( 1966) reported that five species of Notocot\lus: X. pacifcr
(Noble, 1933) ; Ar. ephemera (Nitzsch, 1807) : N. noyeri Joyeux, 1922; Ar. regis
274 HORACE W. STUNKARD
1 larwood, 1939; and A', nilli Baylis, l(>3o, liave Monostomi-type cercariae, whereas
Catatro pis verruca so. ( Fn">hlich, 1789) has Imhricata-type cercariae.
Oclening stated that the cercariae of C. vcrnicosa lack eye-spots, have short,
stumpy tails, and encyst in the snails in which they are produced, iriz., Sajnientina
nitida (O.F.M.) and Gryaiilns albns (O.F.M.). The adults were raised in duck-
lings. The life-cycle of C. rcrntcosa as given by Odening recalls the account of
Joyeux (1922) who reported that stumpy-tailed cercariae without eye-spots from
Planorbis rotundatus Poiret developed in ducklings to adults which were identified
as Notocotylns attcunotiis. The adults were not described but Dubois (1951)
examined specimens deposited in the Zoological Institute of the University of
Xeuchatel and declared that the wrorms were not N. ctttenitatits but C. vcrnicosa.
Szidat (1930) had reported that Ccrcaria ephemera Nitzsch, 1807 from Planorbis
corncns is the larva of C. I'erntcosa, but L. and U. Szidat (1933) assigned the
adults to a new species, Notocotylus thicncuianni. Erkina (in Skrjabin et al., 1963)
described a large cercaria from Bithynia teutaculata and Bithynia leachi with three
eye-spots and a long tail as the larva of Catatropis vcrnicosa. Martin (1956)
described Catatropis johnstoni n. sp., and its life-cycle. The larvae were found in
the prosobranch snail, Ccrithidea California; they had long tails, eye-spots, and
belonged to the Imbricata-group of cercariae.
Discussing the reports of Erkina and Martin, Odening (1966: 229) stated, "Es
scheint kaum moglich, dass ein und dieselbe Art zwei ganz verschiedene Larven-
typen hat ; folglich kann es sich wohl nur bei einem der beiden Zyklen um den von
C. vcrnicosa handeln. Ob es sich nun hierbei um zwei Arten handelt, die als Adulti
kaum oder nicht unterscheidbar sind, kann vorerst nicht entschieden werden. . . . Es
erhebt sich die Frage, ob nicht jener merkwiirdige, von Joyeux entdeckte Typ
monostomer Cercarien fur die Gattung Catatropis characteristisch ware. Diese
Frage lasst sich in Anbetracht der Differenz zwischen den Angaben von Erkina
und den bier geschilderten Ergebnissen (sowie denen von Joyeux) iiber Catatropis
verntcosa nicht l)eanworten. Leider liegen keine Angaben iiber Entwicklung
anderer Catatropis-Arten vor, bis auf die Resultate von Martin (1956) iiber
'Catatropis johnstoni Martin, 1956.' Die Zugehorigkeit dieser Art zur Gattung
Catatropis ist jedoch fraglich, denn es heisst in der Diagnose: 'Median ventral
glandular ridge from ovarian to mid-cirrus level. Lateral ventral glands lacking.'
"Diese Art wiirde in dem gleichen Verhaltnis zur Gattung Catatropis stehen
wie die Gattung Uniserialis Beverley-Burton, 1958, zur Gattung Notocotylns.
Andererseits ist der von Martin beschriebene mediane Driisenkiel aus einzelnen
querovalen Driisen zusammengesetzt. Die Cercarie von 'Catatropis johnstoni'
gehort zur Tmbricata'-Gruppe (Rothschild), hat einen langen Schwanz und drei
Augen. Die zugehorigen Redien schmarotzen bei einem Prosobranchier, der an
der californischen Kiiste lebt. Die Gruppenzugehcirigkeit der von Erkina fiir C.
I'crnicosa beschriebenen Cercarie ist nicht eindeutig erkennbar ; aus den Zeich-
nungen konnte man vielleicht entnehmen, dass es sich um eine Cercarie der
'Yenchingensis'-Gruppe handelt (vgl. auch Sevcov und Zaskind, 1960). Der von
Joyeux entdeckte stummelschwanzige und augenlos Type monostomer Cercarien
wurde auch bei Parapronoceplialuni syniuictricinn Belopol'skaja, 1952, nachgewis-
sen (s. Skrjabin et al., 1955). Die Redien schmarotzen in Meeresprosobranchiern.
Die Cercarie gehort zur 'Monostomi'-oder zur Tmbricata'-Gruppe (die Gruppen-
LIFE-CYCLE OF UNISERIALIS 275
zugehorigkeit \vurde nicht angegeben, es kann nur die 'Yenchingensis'-Gruppe
ausgeschlossen \verden)."
I'niscrialis brcviscrialis, like species of Catatropis, has Imbricata-type cercariae.
The significance of groups of notocotylid cercariae, their generic allocations and
infective sites are yet dubious, and it is apparent that discrimination and discretion
will he required for a solution of the taxonomic problems in the family Notocotyli-
dae.
SUMMARY
Imbricata-type cercariae from Hydrobia salsa, a brackish-water, prosobranch
snail taken near Woods Hole, Massachusetts, emerge shortly before noon, are
photopositive and encyst after swimming for a few minutes to three to four hours.
Aletacercariae were fed to chicks and domestic ducklings and developed to mature
worms after about two weeks in the bursae Fabricii of these birds. Adult and larval
stages are described and figured. The worms belong in the family Notocotylidae
and are assigned to the genus, Uniserialis Beverley-Burton, 1958. Systematic prob-
lems of genera in the family are discussed.
LITERATURE CITED
BAER, J. G., AND CH. JOYEUX, 1961. Classe des Trematodes. /;/ : Traite de Zoologie, P. P.
Grasse, 4 : 561-692.
BEVERLEY-BURTON, MARY, 1958. A new notocotylid trematode, Uniserialis gippyensis gen. et
sp. nov., from the mallard, Anas f>lat\r1i\ncha [>l at \rli\nchn L. /. Parasitol., 44: 412-
415.
DUBOIS, G., 1951. fitude des trematodes Nord-Americains de la collection E. L. Schiller et
revision du genre Notocotyltis Diesing, 1839. Bull. Soc. Ncuchat. Sci. Nat., 74: 41-76.
JOYEUX, CH., 1922. Recherches sur les Notocotyles. Bull. Soc. Path. E.vot.. 15: 331-343.
MARTIN, W. E., 1956. The life cycle of Catatropis johnstoni n. sp. ( Trematoda : Notocotylidae).
Trans. Aincr. Micros. Soc., 75: 117-128.
ODENIXG, K., 1966. Physidae und Planorbidae als Wirte in den Lebenszyklen einheimischer
Xotocotylidae (Trematoda: Paramphistomidea). Zcitschr. Parasitcnk., 27: 210-239.
PILSBRY, H. A., 1905. A new brackish-water snail from New England. Nautilus. 19: 90-91.
ROTHSCHILD, MIRIAM, 1938. Notes on the classification of cercariae of the superfamily
Notocotyloidea (Trematoda), with special reference to the excretory system. Noi'it.
Zool., 41: 75-83.
ROTHSCHILD, MIRIAM, 1941. Note on the life histories of the genus I'urainonostoinuin Liihe,
1909 (Trematoda: Notocotylidae) with special reference to the excretory vesicle. /.
Parasitol., 27: 363-365.
SKRJABIN, K. I., 1953. Trematoden der Tiere und des Menschen. Grundlagen der Tremato-
dologie (Russich), vol. 8. (Cited after Odening, 1966.)
SKRJABIN, K. L, N. P. SICHOBALOVA, A. M. PETROV AND M. M. LEVASOV, 1963. Der Aufbati
der helminthologischen Wissenschaft und Praxis in the UdSSR. Bd. 2. (Russich).
Verl. Akad. Wiss. UdSSR. (Cited after Odening, 1966.)
STUNKARD, H. W., 1960. Studies on the morphology and life-history of Notocotylus ininiitus
n. sp., a digenetic trematode from ducks. /. Parasitol., 46: 803-809.
STUNKARD, H. W., 1966a. Further studies on digenetic trematodes of the family Notocotylidae.
Biol Bull, 131: 409.
STUNKARD, H. W., 1966b. The morphology and life-history of Notocotylus atlanticus n. sp., a
digenetic trematode of eider ducks, Soinatcria inollissiina, and the designation, Noto-
cot\lus duboisi nom. nov., for Notocotvlus inibricatus (Looss, 1893) Szidat, 1935.
Biol. Bull., 131: 501-515.
276 HORACE W. STUNKARD
STUNKARD, H. W., 1967. Studies on tin- tmnatode genus, Paramonostomum Liihe, 1909
(Digenea: Notocotylidae ). liwl. Hull., 132: 133-145.
STUNKAKD, H. W., AND F. W. DUNIHUE, 1931. Notes on trematodes from a Long Island
duck with description of a new species. Biol Hi/11.. 60: 179-186.
SZIDAT, L., 1930. Die Parasiten des Hausgefliigels. 4. Notocittylus Diesing und Catatropis
Odhner, zwei, die Blinddarme des Gefliigels bewohnende monostome Trematodengat-
tungen, ihre Entvvicklung und Uebertragung. Arch. f. Gefluuclk., 4: 105-114.
SZIDAT, L., AND URSULA SZIDAT, 1933. Beitrage zur Kenntnis der Trematoden der Monostomi-
dengattung Notocotylus Dies. Zcntrb. Bakt., Abt. I, 129: 411-422.
THE GROWTH AND ACTIVITY OF THE CORPORA ALLATA IN
THE LARVAL FIREBRAT, THERMOBIA DOMESTICA
(PACKARD) (THYSANURA, LEPISMATIDAE)
J. A. L. WATSON
CSIRO, Division of Entomology, Canberra, Australia a
Althougli the post-embryonic development of the Apterygota ( including the
Archaeognatha and Thysanura ) is essentially continuous and progressive, it shows
some ahrupt changes similar to those associated with the metamorphoses of higher
insects. In particular, the integument of newly hatched larvae lacks scales, although
it bears setae ; if scales develop, they do so either at the second molt, as in Archae-
ognatha, or at the third, as in most lepismatid Thysanura (Delany, 1957).
The processes of scale and bristle formation in Thysanura resemble those found
in the higher insects (Schmidt, 1959; Richter, 1962). Furthermore the develop-
ment of scales in Lepisma resembles metamorphic changes in that it appears to
depend on the hormonal system of the insect (Piepho and Richter, 1959; Richter,
1962). Thus fragments of the integument from newly hatched Lcpisina, if im-
planted into adults, molt when the adult molts and develop scales, suggesting that
the control is humoral. However, the time that elapses between hatching and
ecydsis to the fourth larval stage in Lepisina averages 17.8 days, as opposed to
approximately 43 days for each stadium in the adult (Sweetman, 1952), so that in
Piepho and Richter's experiments, the epidermal cells could have had time to differ-
entiate independently between implantation and the subsequent adult ecdysis.
Other recent, experimental studies on the hormonal control of molting and re-
production in lepismatids have emphasized that the endocrine systems of these
insects are similar in structure to those of the winged insects and that, at least with
respect to the initiation of molting and the deposition of yolk, they appear to func-
tion in the same way (Yashika, 1960; Watson, 1963a, 1964a, 1965 and unpublished
results; Rohdendorf, 1966).
The question therefore arises : If the development of scales in Thysanura is not
a spontaneous event, is it regulated by the corpus allatum? In other words, do these
ametabolous insects show an early stage in utilization of the allatal secretion as a
morphogenetic agent ?
The present paper describes experiments to check for the existence of humoral
control over scale formation in lepismatids, and examines the physiological activity
of the corpora allata during the life of the firebrat. Brief notes on the work have
already appeared (Watson, 1936b, 1965). Attempts to influence scale formation
in early larvae and in regenerating integument have so far yielded inconclusive
results and will not be documented here.
1 Much of the work described in this paper \vas carried out in the Developmental Biology
Center, Western Reserve University, Cleveland, Ohio.
277
278 J. A. L. WATSON
MATERIALS AND METHODS
1. Experimental animals
All experiments were performed on the firebrat, Thcnnobla domestica (Pack-
ard). Larvae and adults were reared under optimal conditions, at 37° C. and 84%
relative humidity, as described in Watson (1964a).
2. Implantations
Implantations were carried out much as described by Piepho and Richter
(1959). First stage larvae less than three hours old were anesthetized with carbon
dioxide. The terminalia were removed from the abdomen and the insect was then
cut across at the junction of the thorax and abdomen. The abdominal fragment
was placed under 0.9% saline until implantation. An adult firebrat three to five
days after ecdysis, and presumed to be on the point of initiating a molt (cf. Watson,
1964a), was anesthetized for one or two minutes with carbon dioxide, the abdomen
was descaled and the larval fragment implanted through a slit in the third abdominal
tergum ; the wound was sealed with a paraffin-beeswax mixture. The adult was
returned to 37° C. and 84% relative humidity.
The recipients of implants were fixed in neutral formalin at various times after
implantation, and were sectioned serially for examination of the larval fragment.
3. Histological measurements
Studies of the corpora allata of other insects have indicated that changes in the
physiological activity of the gland may involve changes in the volumes of cytoplasm
and nucleus, commonly expressed as a ratio between cytoplasm and nucleus, with
or without changes in the number of cells (cf. Pflugfelder, 1958; Scharrer, 1964).
In the case of the firebrat, however, Watson (unpublished data and below) has
shown that neither the nuclear diameter nor the number of cells alters during short-
term fluctuations in the size of the corpora allata, fluctuations that are correlated
with changes in the physiological activity of the gland.
The volume of cytoplasm in each allatal cell is therefore an appropriate measure
of physiological activity ; and as nuclear volume is constant, total cell volume is an
equally valid index.
Firebrats in which such histological measurements were to be made were fixed
in neutral formalin and sectioned serially at 4-8 p.
(a) Numbers of allatal nuclei
Nuclei and fragments were counted in all sections of the corpora allata. The
resulting numbers were corrected for fragmentation by the formulae of Marrable
(1962).
(b) Nuclear diameter
The major and minor axes of the almost spherical nuclei were measured in 10
allatal cells in each of a series of 24 firebrats of various ages. An average diameter
was calculated for each animal, and the individual means were averaged over the
sample. The resulting average, 6.323 /A (S.E. =0.067/0, was used for all later
calculations involving nuclear volumes.
CORPORA ALLATA IN LARVAL THERMOBIA 279
(c) Volume of the corpora allata
Volumes were measured by summing the areas of all the sections of the corpora
allata, using a squared eyepiece micrometer, and multiplying by the thickness of
the section.
All other cytological statistics used in this paper were calculated from the above
three measurements.
4. TIic assay of all at a! activity
The juvenilizing activity of the corpora allata was assayed by a modification of
the Polyphemus test (Gilbert and Schneiderman, 1960). Chilled pupae of the
saturniid Antheraca polyphetnus (Cram.) were placed at 25° C., until the first
signs of epidermal retraction from the facial window indicated the beginning of
adult development. Maxillae (which contain the corpora allata) or the ventral
halves of heads sufficient to provide approximately 100 allatal cells (see below)
were implanted into the midrib of the pupal antenna. Dorsal halves of the head,
or an approximately equal volume of body tissue, served as control in the other
antenna, and in later comparative experiments corpora allata from two different
stages were implanted into the two antennae. The antennae of the resulting moth
were scored for pupal characters on the scale of Gilbert and Schneiderman (1960).
As the critical period for the action of the juvenile hormone ends shortly after re-
traction of the facial epidermis, the juvenilizing effect of the implant is a reflection
of its secretory state at or immediately after the time of implantation.
5. Analyses of data
Statistical techniques referred to below are described in Siegel (1956) or
Bailey (1959).
RESULTS
1. Timetable of development
At 37° C. and 84% R.H. growth from hatching to ecclysis to the fourth larval
stage, when the scales first appear, occupies an average of 8.5 days ; the mean lengths
of the first three stadia, with standard errors, are 24.9 ± 0.47 hours, 69.9 ± 0.49
hours and 109.1 ± 5.14 hours. The larvae commence feeding in the third stadium,
the residual yolk then being exhausted.
The sequence of epidermal events during the third stadium appears in Figure 1.
At 18 hours after ecdysis, the epidermis appears inactive ; mitosis follows, in most
cases some time between the 18th and 48th hour, so that by the 60th hour, pycnotic
epidermal nuclei are present in all larvae. Extensive RNA synthesis, as reflected
by the basophilia of the trichogen and presumptive scale-forming cells, generally
commences between the 18th and 48th hours, but strong basophilia may not develop
until the third day. The epidermis starts to retract between 60 and 72 hours, the
first scale-forming processes appear by 84 hours, and ecdysis follows approximately
24 hours later.
Thus any mechanisms influencing scale formation must act before the 48th hour
after ecdysis, by which time the presumptive scale-forming cells are becoming
recognizable,
280
J. A. L. WATSON
1 }
Enlarged scale-forming cells
Age in third stadium ( hours )
96
FIGURE 1. Timetable of epidermal events during the 3rd stadium of Thennobia.
2. Implantation experiments
The experiments on implantation unequivocally confirmed the interpretations
of Piepho and Richter (1959). The implants had molted and developed scales
when the adult recipients had molted, five days or more after implantation ; and in
a few cases, scales had formed within 48 hours of implantation (Fig. 2). Clearly,
scale formation it not simply a spontaneous, time-dependent differentiation ; the
CORPORA ALLATA IN LARVAL THERMOBIA
281
potential for it is present in first stage larvae, and can be realized immediately in an
appropriate environment. As some of the implants which had developed scales still
contained abundant yolk in close proximity to the integument, it seems unlikely that
a component of yolk directly inhibits scale formation. The "appropriate" environ-
ment may therefore be one which differs hormonally from that of the earl}- larvae,
as Piepho and Richter (1959) have suggested.
Could such a difference involve the juvenile hormone? In the present study,
the histophysiological aspects of this question have been considered, particularly the
pattern of growth in the corpus allatum and the secretory activity of the gland dur-
ing larval and adult life.
3. The post- embryonic <jrowth of the corpus a/latitin
(a) The number of cells in the corpus allatum
The number of cells in each corpus allatum increases as the firebrat grows.
"Watson (1963a), working with data not corrected for fragmentation, described
linear growth in the adult corpus allatum, and it now appears that the linear rela-
tionship extends through the larval stages (Fig. 3). The fitted regression in Figure
FiiH'KE 2. Section of integument from Tlicnnobia hatchling implanted into molting adult,
and left for 48 hours. The prematurely-formed scales are indicated by arrows ; the marker
represents \() p..
282
J. A. L. WATSON
Weigh! of
FIGURE 3. Relationship between the weight of a firebrat and the number of allatal cells.
The fitted regression has the form: y = 10.61 + 1.94x (modified after Watson, 1963a).
3 has the form :
y = 10.61 + 1.94x,
where y == the number of cells in each corpus allatum and x -- the weight of the
firebrat in mg. The 5c/c confidence limits for the regression coefficient are 1.80 and
2.08, so that the regression line does not intersect the origin, P < 0.001.
TABLE I
The numbers of allatal nuclei in larval Thermobia
Instar
Average
Range
N
1st
13.9
9.6-18.4
6
Early — mid-2nd
Late 2nd
17.2
16.9
11.2-24.0
15.1-18.6
8
5
Early 3rd
Mid-3rd
15.6
17.4
12.0-21.6
12.4-22.1
10
10
Late 3rd
18.5
14.0-23.5
5
Early 4th
Mid-4th
16.2
16.1
12.4-18.6
12.4-19.7
6
11
CORPORA ALLATA IX LARVAL THERMOBIA
283
15
0-0
3 4 5 6 7 S 9 I
3 45676910
Weight of insect ( mg )
10 30 40 50 60 708090100
FIGURE 4. Changes in volume of allatal cells during the life of Thcnnobia. (Volumes
for each of the first four instars are grouped sequentially, rather than by actual weight.)
O = 1st and 2nd instar ; • = 3rd instar ; A = 4th and later instars.
284
J. A. L. WATSON
Not onlv is the net growth in the corpus allaUun linearly related to the size of
the lirchrat, lint the growth is progressive; there are no cycles of increase and
decrease within a stadium. Watson (unpublished results) has documented this for
the adult, and an analysis for the first four larval stages appears in Tahle I. None
of the samples differs significantly from any other within that instar (Mann-
Whitney "I"' test).
(b) The volume of the corpus allatum
In addition to the changes in allatal volume due to the progressive increase in
the number of allatal cells, there are changes due to fluctuations in the volumes of
1-3
n MU r
^ — i — ^
•^- — i —
^ — — j — — r-
.* A
1-2
-
1-1
-
..
10
-
\
•
9
•8
-
• •
?. ;•
T... **
• 7
•6
8 5- i
\ !
•5
•
•3
• 1
•o
i i
(
) i
234
5678
9 10 11 12 13
Age ( days )
FIGURE 5. Changes in volume of allatal cells during the first four stadia of the firebrat.
existing cells. As mentioned above, these changes do not involve nuclear volume,
which remains constant throughout, but depend on increases and decreases in the
amount of cytoplasm. The relevant data appear in Figures 4 and 5 ; Figure 4
shows the period from hatching through adult life (including reproducing females)
and Figure 5 presents the early larval life in detail.
The data document two points. First of all, once the scales have formed, the
volume of cytoplasm in each allatal cell increases steadily until the attainment of
sexual maturity. Thus the cell volumes in each of the categories, 4th instar - ; 5
CORPORA ALLATA IN LARVAL THKRMOBIA
285
nig., >5 nig. - ; 10 ing., and > 10 ing. (excluding mated females) (Fig. 4) ex-
ceed those in the previous category. P < 0.002 ( Mann-Whitney "U" test). At
weights of 8-10 nig., when the females start to mate, the cell volume may increase
further, correlated with the development of the eggs (Watson, 1965 and unpublished
results).
Secondly, there are changes in the allatal cytoplasm (hiring the first four stadia.
During the first twro larval stages, the corpora allata maintain a constant cell volume
(Fig. 5; P > 0.10, Mann-Whitney "I"' test). After the second ecdysis, however,
the volume of cytoplasm decreases ; the cell volumes for early and mid-third stage
larvae differ from those for first and second stage larvae, P < 0.001 (Mann-
Whitney "U" test), as do the ratios between cytoplasm and nucleus. After the
mid-third stage, when the scales have appeared on the new cuticle, the volume in-
creases again ; the cell volumes and ratios of cytoplasm to nucleus for the late third
stage exceed those for the mid-third stage, P < 0.01 > 0.001, but do not differ
from those of the fourth stage, P > 0.10 (Mann-Whitney "U" test).
The question then arises : Do these changes in the volume of the corpus allatum
reflect changes in the physiological activity of the gland ?
4. The juvenile hormone activity of the corpora allata
An initial series of assays based on the implantation of either maxillae or the
ventral halves of the head into the left antenna of Polyphemus pupae with body or
TABLE II
The distribution of juvenile hormone activity in heads and body fragments of Thermobia
Implant
Stage of
donor
No. adults
emerged
L + ve
R + ve
L > R
L. antenna
R. antenna
1st
Ventral head
Body fragments
11
6
2
5
4th-5th
Ventral head
Body fragments
13
7
0
7
Adults
Maxillae
Dorsal head or labium
4
3
0
3
dorsal cephalic tissue as control in the right antenna, showed that the corpora allata
from first stage and fourth or fifth stage larvae and from young adults possess
juvenile hormone activity (Table II). Measured over the entire sample, the allatal
activity exceeded that shown by an equal or greater amount of body tissue (P
0.0154, Fisher exact probability test). The antennae of the resulting adult moths
generally showed slight to considerable thickening of the midrib in the vicinity of
the allatal implant, with fusion of the barbs up to half the width of the antenna,
equivalent to scores of 0 to 3 on the scales of Gilbert and Schneiderman (1960). In
the right antenna, only bodies from the first stage showed any juvenile hormone
activity. In no case was there any general, systematic effect such as Yashika (1960)
obtained with Ctenolepisma, but Yashika implanted far greater quantities of allatal
tissue, and into the abdominal hemocoel.
The experiment was then extended to assay the activity of early larval corpora
allata. The ventral halves of five heads were implanted into each antenna, a differ-
ent instar being used as donor for each side. The antennae of the resulting adult
286
J. A. L. WATSON
were scored us above, and the diflcrcnccs between (be two antennae were analyzed
bv a sign test : ibe magnitude ot tbe differences was disregarded, as the scoring
tuble is based on an ordinal scale.
Tbe results appear on Table 111. As the Table shows, the corpora allata reach
their minimal activity early in the third stage ; the activities of the first and second
stages do not differ from each other ; and there is a marked rise in activity between
the third and fourth stages. Thus the juvenilizing activity of the corpus allatum
is related to the volume of cytoplasm in the gland. In other words, the activity of
the corpus allatum is minimal at the time that scale formation is induced.
5. Tlic regulation of the corpus allatinn
The coincidence between the exhaustion of yolk reserves and the minimal activity
of the corpus allatum suggests that the two may be connected ; the activity of the
TABLE III
Juvenile hormone activity in the corpora allata of early larval Thermobia
Comparison
L < R
L > R
Interpretation and probability
(Sign test)
L. antenna
R. antenna
1st
2nd
6
4
1
2nd > 1st: P = 0.062
2nd
early 3rd
1
3
7
2nd > early 3rd :
P = 0.035
early 3rd
late 3rd
8
1
2
Late 3rd ;> early 3rd :
P = 0.055
early 3rd
4th
11
1
0
4th > early 3rd :
P = 0.006
corpus allatum may be controlled, directly or indirectly, by the quantity or quality
of food.
The allatal volume was therefore measured in third stage larval firebrats that
were isolated immediately after ecdysis, placed under optimal physical conditions
and provided only with cellulose ("starved" firebrats). It was found that such
larvae generally became moribund and died within four days, despite the fact that
cellulose was ingested ; none survived to molt into the fourth stage. The corpora
allata were much smaller in the "starved" firebrats than in normal animals of the
same age, attaining a constant minimal volume within 48-60 hours after ecdysis
(Fig. 6) ; P == 0.004 (Mann-Whitney "U" test).
Third instar firebrats that had been fed on cellulose for 3 or 4 days were then
fed with cereal containing protein, and the allatal volumes were measured 24, 48 and
72 hours after resumption of normal diet, and immediately after ecdysis to the fourth
stage which occurs about 10 days after the second ecdysis. As the data in Figure 6
show, an increase in allatal volume followed the resumption of feeding, except in
animals which had become too moribund to recover (e.g. low values at 48 and 72
hours). The overall reduction in allatal volume persisted at least until ecdysis to
the fourth stage, at which time the corpora allata were still smaller than in control
animals (Fig. 5), P = 0.012 (Mann-Whitney "U" test), although they exceeded
the volumes at the end of starvation, P = 0.018 (Mann-Whitney "U" test).
CORPORA ALLATA IN LARVAL THERMOBlA
287
» ' | moribund )
| moribund t
Protein obsent
•+M-
. Protein present
Age
third stadiun
FIGURE 6. The effect of starvation on the allatal volume of 3rd instar Thermobia.
Thus irrespective of whether the effect of starvation is greater on the corpus
allatum than it is on other tissues, it is evident that nutrition can influence the
volume of allatal cytoplasm and hence, the physiological activity of the gland.
DISCUSSION
The data presented above strengthen the suggestion of Piepho and Richter
(1959) and Richter (1962) that the post-embryonic development of lepismatids is
regulated by changing titers of juvenile hormone. More specifically, one may cor-
relate the appearance of scales during the third molting cycle with a precipitate fall
in the activity of the corpora allata at the time of the second ecdysis, a fall possibly
connected with the exhaustion of food reserves.
The question then arises : Why should scales persist when the allatal activity
increases in later instars?
It is impossible to make any precise estimates of what the actual titers of
juvenile hormone might be. If, however, it is assumed that the secretory potential
of the corpus allatum is directly related to the volume of its cytoplasm, a supposition
which is probably an oversimplification, but about which there is little critical infor-
mation, it follows that the maximal titer of juvenile hormone that the gland can
produce will depend on the volume of cytoplasm per unit volume of insect (cf.
2SS
J. A. L. WATSON
70
60
50
O)
'«
X
-
0>
E
i
O
40
a
o
o
30
20
10
0-5 1-0 50
Weight of insect ( mg )
10-0
50-0
FIGURE 7. Changes in the amount of allatal cytoplasm per mg. body weight during the life of
Thenuobla. For the interpretation of curves A and B see text. Symbols as in Figure 4.
CORPORA ALLATA I \T LARVAL THERMOBIA
Novak, 1954). Thus in Thermobla the maximal titer would depend on the numher
of allatal cells and the average volume of cytoplasm contained in each cell. The
pattern of growth in the corpus allatum ensures that the numher of cells per nig.
hody weight decreases rapidly as the larva increases in size. At hatching, there are
approximately 63 allatal cells per mg. body weight, the numher of allatal cells then
averaging 13.9 (Table I) and the weight, 0.22 mg. ; in the mid-late third stage, the
number has fallen to approximately 41 cells per mg., the weight then averaging 0.43
mg. and the number of allatal cells, 17.7; at sexual maturity, with a body weight of
approximately 10 mg.. there are some 6 allatal cells per mg. ; and the number
approaches the asymptotic value, 3.88, in old age, at weights of 50 mg. The decline
in titer that this pattern of growth could provide is shown in Figure 7B, based on
the median value for the volume of cytoplasm per allatal cell (1,358/x3) and the
regression in section 3 (a) above.
However, this decline is partially offset by the progressive increase in the volume
of allatal cytoplasm. The observed values for the volume of allatal cytoplasm per
mg. body weight appear in Figure 7 and curve A is fitted to these points. It is
evident that despite the cytoplasmic changes, the range of values in the third instar
is not exceeded in later larval or adult life, even in reproducing females.
Furthermore, even if the titer of juvenile hormone were to exceed the values of
the early third stage, it seems probable from studies of higher insects that the
dedifferentiation of the epidermal cells would not be immediate (cf. Wigglesworth,
1954; Lawrence, 1966). Thus inactivation of juvenile hormone from the onset of
the molting cycle would prevent the prolonged exposure to the hormone which
appears to be a prerequisite for dedifferentiation. It is therefore of interest that
such inactivation occurs, at least in the adult firebrat ; the activity of exogenous
hormone ( Cecropia extract) persists during the reproductive phase of the adult
stadium, but not during the molting phase (Watson, unpublished results; cf.
Watson, 1964a, 1964b).
I wish to acknowledge the hospitality of Professor H. A. Schneiderman. of the
Department of Biology, Western Reserve University ; and the financial assistance
of the Lalor Foundation and the Queen Elizabeth II Fellowships Committee,
together with grants from the National Institutes of Health and the National
Science Foundation to Professor Schneiderman. I also wish to thank Mrs. M.
Hudack for her help with the histology.
SUMMARY
1. The integument of the firebrat, Thcrmobia doiucstica, lacks scales until the
molt from the third to the fourth larval stage, but retains them in all subsequent
instars.
2. Implantation experiments confirm earlier findings that the scaleless integu-
ment of first stage larvae will develop scales prematurely when implanted into a
molting adult, implying that the formation of scales is Immorally determined.
3. The number of cells in the corpora allata increases progressively throughout
the life of the firebrat, and the relationship between the number of allatal cells and
the weight of the firebrat is expressed by the regression :
y :: 10.61 + 1.94x,
290 J. A. L. WATSON
where y == the number of cells in each corpus allutum and x -- the weight of the
insect in ing. The regression does not intersect the origin.
4. The size of nuclei in the corpus allatum remains constant throughout life,
but the volume of cytoplasm can alter. The minimal cytoplasmic volume coincides
with the deposition of the first scale-bearing cuticle. The amount of cytoplasm
then increases abruptly, the level continuing to rise slowly throughout the rest of
larval life. A further increase may occur in mated females.
5. The juvenile hormone activity of the corpora allata, when assayed on pupae
of the silkmoth Anthcraea polyphcnnis, correlates with the volume of cytoplasm in
the gland, and is minimal in the third stage.
6. The activity of the corpora allata is influenced by the intake of food,
specifically protein.
7. Dedifferentiation of the epidermis is prevented by the pattern of growth in the
corpus allatum, which indicates that the secretory potential of the gland, expressed
as the volume of allatal cytoplasm per milligram body weight, is greater during the
first through third instars than at any later stage, and by the inactivation of juvenile
hormone during the molting cycle.
LITERATURE CITED
BAILEY, N. T. J., 1959. Statistical Methods in Biology. English Universities Press, London.
DELANY, M. J., 1957. Life histories in the Thysaniira. Ada Zool. Cracov., 2: 61-90.
GILBERT, L. L, AND H. A. SCHNEIDERMAN, 1960. The development of a bioassay for the juvenile
hormone of insects. Trans. Amcr. Micr. Soc., 79: 38-67.
LAWRENCE, P. A., 1966. The hormonal control of the development of hairs and bristles in the
milkweed bug, Oncopeltus fasdatus, Ball. /. Exp. BioL, 44: 507-522.
MARRABLE, A. W., 1962. The counting of cells and nuclei in microtome sections. Quart. J.
Micr. Set., 103: 331-347.
NOVAK, V. J. A., 1954. Growth of the corpora allata during the postembryonal development
in insects. Mem. Soc. Zool. Tchccosl., 18: 98-133.
PIEPHO, H., AND A. RICHTER, 1959. Zur Entwicklungsphysiologie des Schuppenkleides bei
Ur-Insekten, Untersuchungen am Silberfischen Lcpisma saccharina L. Biol. Zbl.,
78: 855-861.
PFLUGFELDER, O., 1958. Entwicklungsphysiologie der Insekten. Geest u. Portig, Leipzig.
RICHTER, A., 1962. Uber die Entwicklung der Schuppenorganen und der Genitalanhange in
Abhangigkeit vom Hormonsystem bei Lepisma saccharina L. Arch. f. Entw., 154: 1-28.
ROHDENDORF, E., 1966. Der Einfluss der Allatektomie auf adulte Weibchen von Thermobia
doincstica PACKARD (Lepismatidae, Thysanura). Zool. Jahrb. Physiol., 8: 685-693.
SCHARRER, B., 1964. Histophysiological studies on the corpus allatum of Lciicophaea maderae
IV. Ultrastructure during normal activity cycle. Z. Zellforsch., 62: 125-148.
SCHMIDT, E. M., 1959. Untersuchungen zur Cytologie des Wachstums in der postembryonalen
Entwicklung von Lcpisma saccharina L. Zool. Jahrb. Anat., 77: 357-410.
SIEGEL, S., 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill Book
Co., Inc., New York, Toronto and London.
SWEETMAN, H. L., 1952. The number of instars among the Thysanura as influenced by the
environment. Trans. 9th Int. Congress of Entonwl., 1: 411-415.
WATSON, J. A. L., 1963a. The cephalic endocrine system in the Thysanura. /. Morph., 113:
359-373.
WATSON, J. A. L., 1936b. Control of postembryonic development in the firebrat, Thermobia
doincstica (Packard). Amer. Zool., 3: 533.
WATSON, J. A. L., 1964a. Moulting and reproduction in the adult firebrat, Thermobia domestica
(Packard) (Thysanura, Lepismatidae) — I. The moulting cycle and its control. /.
Ins. Physiol., 10: 305-317.
CORPORA ALLATA IN LARVAL THERMOBIA 291
WATSON, J. A. L., 1964b. Moulting and reproduction in the adult firebrat, Thermobia domcstica
(Packard) (Thysanura, Lepismatidae) — II. The reproductive cycles. /. Ins. Phvsiol.,
10: 399-408.
WATSON, J. A. L., 1965. The endocrine system of the lepismatid Thysanura and its phylogenetic
significance. Proc. 12th Int. Congress of Entomol., p. 144.
WIGGLESWORTH, V. B., 1954. The Physiology of Insect Metamorphosis. Cambridge University
Press, Cambridge.
YASHIKA, K., 1960. Studies on the neurosecretory system in Apterygota. II. Development
of the corpus allatum in Ctenolepisma and its juvenile action on the Philosamia-pupa.
Mem. Coll. Sci. Kyoto, So: B., 27 : 83-88.
LIGHT-INDUCTION OF SHEDDING OF GAMETES IN CIONA
INTESTINALIS AND MOLGULA MANHATTENSIS '
D. G. WH ITT INCH AM
Johns Hopkins School of Hygiene & Public Health, Baltimore, Maryland
Castle (1896) and Conklin (1905) observed that the release of gametes in three
species of solitary ascidians (Ciona intestinalis, Slycla partita and Molgula man-
Jiattensis) occurred at definite times during the daylight period. Ciona intestinalis
and Molgula uianliattcnsis spawn one to one and a half hours before sunrise and
Styela partita spawns during the late afternoon. Grave (1921, 1937) found, in the
colonial ascidian Amaroucium const ellatum, that the greatest release of larvae
occurred at and just before sunrise but they continued to be liberated in small
numbers throughout the daylight period. He reported a similar phenomenon in
BotryUits sclilosscri ; in this colonial form the larvae are released in increasingly
greater numbers as the day advances, finally reaching a maximum at noon. Rose
(1939) recorded that Stvela partita could be induced to shed their eggs by subject-
ing them to 11-12 hours of light prior to the desired time of spawning. He observed
that the natural time for shedding occurred in the laboratory between 4 and 7 PM.
Furthermore, spawning under experimental conditions could be induced on four
or five successive days by controlling the illumination. As yet there has been no
study made of the characteristics of the illumination necessary to cause shedding.
In the present study the effects of the intensity and the wave-length of light upon
shedding have been investigated in two solitary oviparous ascidians which shed at
dawn — Ciona intestinalis and Mohjula uianliattcnsis. Before this was done, how-
ever, it was necessary to determine the exact times of shedding in these two species.
MATERIALS AND METHOD
Ciona intestinalis and Mohjitla uianliattcnsis were obtained from the supply
department at the Marine Biological Laboratory, Woods Hole, Massachusetts,
during the months of June, July and August, 1965. These animals were kept for
several days in laboratory aquaria ; those ascidians which were sexually mature and
shedding gametes were selected for experimental purposes, and after being subjected
to a particular treatment they were discarded.
All experiments were conducted in a room where the ascidians were subjected
to a standard day consisting of 12 hours of light and 12 hours of darkness. In order
that two series of experiments could be carried out simultaneously, animals were
kept in black-painted light-tight boxes during the period of 12 hours of darkness.
1 This work was carried out during the summer of 1965 at the Marine Biological Laboratory,
Woods Hole, Massachusetts, while the author was a recipient of a Pennsylvania Plan Fellowship
in the Department of Animal Biology, School of Veterinary Medicine, University of Pennsyl-
vania, Philadelphia. Support was made available through training grant 5T1-HD 26-04 from
the National Institutes of Health.
292
INDUCTION OF SHEDDING IN ASC1D1ANS
In one experiment light was provided from 9 AM to 9 PM, while in the second
experiment light was provided from 9 PM to 9 AM. In the method outlined ahove
some of the experiments were carried ont under conditions of lighting which were
completely reversed from the conditions existing in the animals' natural hahitat.
No apparent differences were ohserved between animals shedding at approximately
their normal time and those shedding in the reversed system of lighting.
The ascidians were kept individually in fingerbowls containing 250 ml. of filtered
sea water which was changed every 12 hours. The temperature of the sea water was
recorded at the time it was placed in the fingerbowl and at the end of the 12-hour
period just before it was replenished. The fingerbowls were placed on a black
surface to keep the reflection of light from the source to a minimum. (Reflecting
power of a black painted matt surface is less than l%.) The shedding response
was recorded as either positive when gametes were released or negative when no
gametes were shed.
Intensity
In the first series of light-intensity experiments, a standard light source — 100-
watt bulb — was used and the intensity varied by placing the ascidians at different
distances from the light source. The intensity of illumination from a constant light
source varies inversely as the square of the distance from that source. Theoretical
values of light intensity at various distances from the light source may be calculated
from the following equation :
Intensity of illumination at A dr,-
2
Intensity of illumination at B dA
where f/A and dH are the distances of A and B from the light source, respectively,
since manufacturers supply data on the intensity of illumination at 1 foot.
The shedding response was measured at 1 foot, 2 feet, 4 feet, and 8 feet from
the light source and the intensity of illumination was measured with a light meter
to check the theoretical calculations. Groups of six animals were placed in each
treatment and their shedding responses recorded on four successive days. In these
experiments the time when shedding occurred after the animals were introduced
into the light was tabulated for both species of ascidians.
In a second series of light-intensity experiments four different light sources—
60-watt, 40- watt, 25-watt, and 15-watt bulbs — were used at a standard distance
of eight feet from the animals.
Wave-length
Four Kodak Wratten gelatin niters were used to determine the effects of the
wave-length of light upon shedding. The four filters had the following char-
acteristics : ( 1 ) Filter No. 2B absorbed light at wave-lengths of 390 m/x and below,
the approximate colors absorbed were ultraviolet and violet light. (2) Filter No.
16 absorbed light at wave-lengths of 500 m/A and below, the approximate color range
absorbed being blue, blue green, and those colors absorbed by filter No. 2B. (3)
Filter No. 25 absorbed light at wave-lengths of 600 m/*, and below, the approximate
color range absorbed being green and yellow plus those colors absorbed by filter
294
D. G. WHITTINGHAM
No. 16. (4) Filter No. 89B absorbed ligbt at wave-lengths of 700 in/A and below,
the approximate color range absorbed being orange and red plus those colors
absorbed by filter No. 25. Each filter measured 10 cm. by 12 cm. and it was
placed in a darkroom safelight. A fixed intensity of 32 foot-candles was used ; this
intensity had been found to produce optimum shedding responses in previous studies
on the effect of light-intensity on this process. Shedding responses were recorded
on two successive days. Three separate trials were made with dona intestinalis;
in each trial nine animals were allotted to a treatment. Molgula manhattensis was
not as plentiful and 6 animals were used per treatment and 2 trials were made.
RESULTS AND DISCUSSION
1. Time of shedding
In the first series of light-intensity experiments the time when shedding com-
menced after the ascidians were introduced into light was recorded. The data for
TABLE I
Mean shedding times in minutes for dona intestinalis recorded at 4 different
intensities of light
Six animals per light-intensity, shedding responses recorded on 4 consecutive days
Day
1
2
3
4
Light intensity
(foot-candles)
Number
shedding
Mean
shedding
time
Number
shedding
Mean
shedding
time
Number
shedding
Mean
shedding
time
Number
shedding
Mean
shedding
time
130
32
8
15
Total number shedding
Mean shedding time
6
6
4
2
18
3.67
7.17
3.0
2.0
4.5 ±3. 22
6
5
4
3
18
3.17
5.2
2.25
4.0
3. 67 ±1.87
4
5
4
5
18
2.5
3.6
3.5
2.6
3.06±1.11
4
5
5
3
17
3.75
6.8
6.0
2.67
5. 12 ±3. 33
Mean shedding time for 24 animals during 4-day period — 4.07 min. ±2.60.
Ciona intestinalis and Molgula manhattensis are presented in Tables I and II,
respectively. In each species 24 animals were allotted at random to the 4 light
intensities so that the responses of 6 animals were observed at each light-intensity.
The animals were stimulated to shed on 4 occasions, 24 hours apart. Mean shed-
ding times at the 4 light-intensities are given for each group of 6 animals on the 4
days when shedding responses were observed. In addition, the number of animals
in these groups giving positive shedding responses on each occasion is tabulated.
In Ciona intestinalis analyses of variance showed no significant differences in the
shedding time of the animals between 4 levels of light-intensity. A similar analysis
of the data from Molgula manhattensis indicated a difference in the shedding time
of the animals between intensities on the first day (0.01 > P > 0.001) but no differ-
ences were found on the three subsequent days. At the highest light-intensity (130
foot-candles) the mean time taken for the 6 animals to shed after exposure to light
(16.2 min.) was less than the mean times for the three lower intensities (26.5 min.,
23.2 min., and 22.2 min.). It seems unlikely that this is a true effect of light-
INDUCTION OF SHEDDING IN ASCIDIANS
295
TABLE II
Mean shedding times, in minutes for Molgula manhattensis recorded at 4 different
intensities of light
Six animals per light-intensity, shedding responses recorded on 4 consecutive days
Day
1
2
3
4
Light intensity
(foot-candles)
Number
shedding
Mean
shedding
time
Number
shedding
Mean
shedding
time
Number
shedding
Mean
shedding
time
Number
shedding
Mean
shedding
time
130
S
16.2
.S
23.8
2
JS.5
2
27.5
32
6
26.5
6
23.67
6
25.5
1
31.0
8
6
23.2
4
27.75
4
26.25
2
26.0
1.5
5
22.2
4
24.25
5
24.0
5
22.2
Total number
22
19
17
10
shedding
Mean shedding
22. 78 ±5.36
24. 68 ±3. 53
25. 59 ±3.98
24.9 ±3. 73
time
Mean shedding time for 24 animals during the 4-day period — 24.16 min. ±4.46.
intensity upon the time of shedding since it is not repeated on the subsequent days ;
however, these animals may have adapted themselves to this high intensity of light
on the latter three days of the experiment.
Mean shedding times for both species were calculated from the data obtained
during the whole experiment. In dona intestinalis shedding occurred in 22/24 of
the animals used and 74% (71/96) positive shedding responses were obtained
TABLE III
Shedding response patterns of Ciona intestinalis and Molgula manhattensis observed
on 4 consecutive days
Shedding response patterns
Day
Ciona intestinalis
Light-intensities
Molgula manhattensis
Light-intensities
(Foot-candles)
Type
i
2
3
4
130
32
8
1.5
Total
130
32
8
1.5
Total
1
1
1
1
1
4
5
3
12
1
1
1
3
6
2
1
1
1
0
1
5
2
1
9
3
1
1
0
1
1
1
4
1
1
0
0
2
2
1
1
2
5
1
0
1
1
1
1
2
1
1
2
6
1
0
1
0
1
1
7
1
0
0
1
8
1
0
0
0
1
1
1
1
2
9
0
1
1
1
2
2
10
0
1
1
0
1
1
11
0
1
0
1
1
1
12
0
1
0
0
1
1
13
0
0
1
1
14
0
0
1
0
15
0
0
0
1
16
0
0
0
0
1
1
2
1
1
1 = shedding
0 — no shedding
296
I). <i. WHITTIXUI \M
during the 4 days. This species commenced shedding gametes 4.07 min. ± 2.60
after exposure to light. In Molgnla inanhattcnsis 23/24 animals shed and 71%
(68/9(> ) positive shedding responses were obtained during the 4 days; the overall
mean time when shedding commenced was 24.16 min. ± 4.46 after exposure to
light. Therefore, from the data above it has been shown that these two species of
ascidians have their own characteristic shedding time. If the light stimulus mediates
its effect ria the neural ganglion and neural gland complex of these animals, the
response of these structures to the light-stimulus occurs much more rapidly in Ciona
intestinalis than in Molgitla inanhattcnsis.
Since the behavior of each animal was observed on 4 successive days its overall
response pattern can be represented by a vector of 4 elements, 1 and 0 representing
shedding and not shedding, respectively. There are thus 16 possible response pat-
terns. The distribution of the shedding response patterns of the two species ob-
served on the four consecutive days is listed in Table III. Fifty per cent (12/24)
TABLE IV
Percentage of dona intestinalis and Molgula manhattensis shedding in response to
d ifferen t light-in tensities
Light intensities
(foot-candles)
Number of animals
% Shedding
Ciona inteslinalis
Molg2ila manhattensis
130
6
100.00
83.33
32
6
83.33
83.33
8
6
66.67
66.67
1.5
6
50.00
66.67
1.17
12
41.67
58.33
0.67
12
41.67
66.67
0.40
12
16.67
58.33
0.21
12
0
25.00
of Ciona intestinalis shed on four days whereas only 2S% of Molgula manhattensis
shed consecutively over a similar period. The highest shedding responses occurred
on the first two days of the experiment ; the fall in response over the last two days
may have been due to the lack of food materials or to the handling of the animals.
2. Light-intensity
The shedding responses of the two ascidians to 8 different light-intensities are
presented in Table IV. These responses were recorded on the second day of
exposure to the various light-intensities. The high intensities of light did not
inhibit the shedding response. In Ciona intestinalis shedding was reduced to
16.67% at 0.40 foot candle and was completely inhibited at 0.21 foot-candle. The
response in Molgula manhattensis was reduced to 25% at the latter light-intensity.
The active contractions of the animals associated with the light-stimulus (Hecht.
192C> ) and the shedding of gametes (Castle, 1896) were diminished in both species
at the lowest light-intensity (0.21 foot-candle). It may be suggested that the
release of gametes is a reflex associated with the muscular contractions of the animal
stimulated by exposure to light.
INDUCTION OF SHEDDING IN ASCID1ANS
297
In the present study, it has been shown that these species shed at low as well
as high light-intensities which would allow them to adapt to a fairly wide range of
naturally occurring hahitats (Van Name, 1945).
3. Wave-length
The data obtained showing the shedding response of the two species when the
wave-lengths of light were restricted to certain regions of the spectrum are pre-
sented in Tahle V. In hoth species the shedding response was reduced to below
50% when light was absorbed up to 600 m^ and totally inhibited \vhen absorbed up
to 700 m/x. These results indicate that the excitatory wave-lengths occur between
500 m/x. and 700 m//,. A constant light-intensity was used in these experiments—
32 foot-candles — and therefore no interaction between light-intensity and wave-
length was shown ; however, further investigation of this aspect may help in the
understanding of the shedding phenomenon. In the experiments described, the
TABLE V
Percentage of Ciona intestinalis and Molgula manhattensis shedding in response to
restricted wave-lengths of light
Ciona inlestinalis
Molgula manhattensis
Wave-length absorption
Number of
animals
% Shedding
Number of
animals
% Shedding
390 nip and below
27
81.84
12
91.67
500 nip and below
27
62.96
12
83.33
600 nip and below
27
18.52
12
33.33
700 nip and below
27
0
12
0
wave-length range which stimulates shedding has good transmission through sea
water (Jerlov, 1964) and these wave-lengths would reach the animals in their
natural habitat.
The author gratefully acknowledges the advice and guidance given to him by
Dr. J. D. Biggers during the course of the work and in the preparation of the
manuscript.
SUMMARY
1 . The time of shedding of gametes and the effects of the intensity and the wave-
length of light upon the process have been investigated in Ciona intestinalis and
Molc/itla manhattensis.
2. The time of shedding was recorded from 24 animals in each species on four
consecutive days. It was found that Ciona intestinalis and Mohjula manhattensis
commenced to shed their gametes 4.07 min. ± 2.60 and 24.16 min. ± 4.46, respec-
tively, after exposure to light. Twelve out of 24 Ciona intestinalis shed on 4 con-
secutive days and 6 out of 24 MoUjula manhattensis shed consecutively over a simi-
lar period.
298 1). G. WHITT1NGHAM
3. Intensities ranging from 0.21 to 130 foot-candles (f.c.) were used to study
light-intensity effects. High intensities did not inhibit shedding but lower intensities
—0.67 f.c. and 0.40 f.c. — produced a marked reduction in the shedding response.
At 0.21 f.c. shedding was completely inhibited in Ciona intestinalis and reduced to
25% in Molgula manhattensis.
4. In both species, shedding was partially inhibited when light was absorbed up
to 600 m/t and totally inhibited when absorbed to 700 m//,. This indicated that the
excitatory wave-lengths lie between 500 in//, to 700 m/i.
LITERATURE CITED
CASTLE, W. E., 1896. The early embryology of Ciona intestinalis (L). Bull. Mns. Comp.
Zool.,27: 201-280.
CONKLIN, E. G., 1905. Organization and cell lineage of the ascidian egg. /. Acad. Nat. Sci.,
Philadelphia. 13: 1-21.
GRAVE, C., 1921. Amaroucium constellatum (Verrill) II. The structure and organization of
the tadpole larva. /. Morfh., 36: 71-91.
GRAVE, C., 1937. In : Culture Methods for Invertebrate Animals. Ed. Galtsoff et al., Corn-
stock, Ithaca, pp. 556-564.
HECHT, S., 1926. The effect of exposure period and temperature on the photosensory process
in Ciona. J. Gen. PhysioL, 8: 291-301.
JERLOV, N. G., 1964. Colour niters to simulate the extinction of daylight in the sea. Extrait
du Journal dit Conscil International Pour I'Exploration de la mcr, 20: 156-159.
ROSE, S. M., 1939. Embryonic induction in the ascidians. Biol. Bull., 77: 216-232.
VAN NAME, W. E., 1945. The North and South American ascidians. Bull. Amcr. Mns. Nat.
Hist., 84: 1-476.
Vol. 132, No. 3 June, 1967
THE
BIOLOGICAL BULLETIN
PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY
AMYLASE AND GLYCOGENOLYSIS IN
AMPHIBIAN DEVELOPMENT1
FRANCISCO D. BARBIERI, JORGE S. RAISMAN AND CESAR ALBARRACIN 2
Institnto dc Biologia and Institute de Quimica Biologica, Facultad de Bioquimica, Quimica y
F armada, Univcrsidad National de Tucuman, R. Argentina
Since the carbohydrate reserves in amphibian eggs are mainly stored as gly-
cogen (Gregg, 1948), attempts have been made to estimate its utilization during
development as a function of glycogen breakdown. The most reliable data show
that glycogen level begins to decrease when time of gastrulation starts (Brachet and
Needham, 1935 ; Gregg. 1948). This has been confirmed by microchemical (Heat-
ley and Lindahl, 1937; Jaeger, 1945) as well as histochemical methods (Woerde-
mann, 1933; Raven, 1935) which have also shown that glycogenolysis is stronger
in those cells involved in the morphogenetic movements of gastrulation. Besides,
the determination of the respiratory quotient (R.Q.) has given values concordant
with those results. It is true that Earth (1946) has found a constant value of
about 0.9; but Brachet (1934) and Boell (1955) have reported low values during
segmentation, with a tendency to increase up to a value close to 1 at the time
of gastrulation.
Less is known about the egg enzymes involved in glycogenolysis. Some evi-
dence was reported indicating that glycogen breakdown could be accomplished
through phosphoroclastic (Cohen, 1954; Gregg et al, 1964) as well as through
amylolytic pathways (see Urbani, 1962, for a review of the subject). The most
outstanding news reported by the Italian author was the description of a ^-amylasic
activity in eggs of Rana escnlcnta and Biijo vulgaris.
Nothing is known, however, about glycogenolytic enzymes in Bufo arenarum.
The limited information we possess on glycogen utilization during development
agrees with the above reported results. Thus, a glycogen loss in eggs could be
detected after the onset of gastrulation (Barbieri and Gil, 1962) ; and the R.Q.
values were found to increase from about 0.6 during segmentation, up to 1 at the
time of gastrulation (Legname and Barbieri, 1962).
1 This work was supported in part by grants from the Consejo Nacional de Investigaciones
Cientificas y Tecnicas (R. Argentina).
2 Fellow of the Consejo Nacional de Investigaciones Cientificas y Tecnicas (R. Argentina)
under the direction of Dr. J. M. Olavarria (Institute de Quimica Biologica).
299
Copyright © 1967, by the Marine Biological Laboratory
Library of Congress Card No. A38-518
300 F. D. BARBIERI, J. S. RAISMAN AND C. ALBARRACIN
The present paper has two purposes. The first is to report some additional
data about glycogen contained in the eggs of the toad. The second is to present
our first observations about the properties and biological role of an amylase found
in the same material.
MATERIAL AND METHODS
Biological material
Oocytes of Bufo arenarum obtained by injection of pituitary glands preserved
according to Pisano (1956) were artificially fertilized. Development was allowed
to proceed in W% amphibian Ringer's solution without bicarbonate at laboratory
temperature. Prior to homogenization, the jelly coat was dissolved by ultraviolet
irradiation or by a neutralized thioglycolic acid solution.
Cellular fractionation
The eggs were homogenized, unless otherwise stated, with two volumes of
chilled 0.25 M sucrose solution containing 0.001 M ethylenediatnine-tetraacetic
acid (disodium salt). The homogenization was carried out in an ice-cooled Potter
type glass homogenizer. In order to establish the localization of glycogen and
amylase, cellular fractions were isolated by differential centrifugation in Sorval
RC-2 or Christ Universal refrigerated centrifuges. The homogenate was centri-
fuged, 20 minutes each time, at 1500 g, yielding a pellet containing nuclei, yolk
platelets and pigment granules (fraction I), and at 10,000 and 25,000 g, sediment-
ing the mitochondria (fractions II and III). The remaining supernatant will be
referred to as fraction IV.
Chemical methods
As a mild procedure for the extraction of glycogen, the HgCl2 method of
Mordoh et al. (1966) was adopted. A comparison of HgCl2-extracted glycogen
with cold trichloracetic acid-extracted glycogen, prepared from the same batch of
eggs, has shown that they do not differ from the standpoint of the properties con-
sidered in this paper. As a standard method for the estimation of glycogen the
phenol-sulfuric acid method as described by Dubois et al. (1956) was employed.
The iodine reaction was performed in the presence of calcium chloride, according
to Krisman (1962). Liver glycogen from adult specimens of the same species
was used as a standard for both methods, and, in some experiments, it was ex-
tracted from the same females wherefrom the eggs used for analysis had been taken.
For the determination of amylase activity the reaction mixture contained,
except where otherwise stated, 1.2 mg. of glycogen, 0.2 M phosphate buffer at
pH 7.2, Q.I M NaCl and 0.02 ml. of crude enzyme preparation, in a total volume
of 0.08 ml. Incubation time was 20 minutes at 37° C., and the reaction was
stopped by heating 2 minutes at 100° C. Reducing power was determined accord-
ing to Somogyi and Nelson (Ashwell, 1957).
Absorption spectra of the color reactions with iodine were determined with a
Beckman DU spectrophotometer, and the photometric readings with the phenol-
sulfuric and Somogyi-Nelson methods were performed with a Spectronic 20
(Bausch & Lomb).
AMYLASE AND GLYCOGENOLYSIS
301
en
oz
o
in
CD
0.4-
0.3
380
420
460
500
540 rnp
FIGURE 1. Absorption spectra of four glycogen preparations (0.05%) in the presence of iodine
reagent. 1 : unfertilized eggs ; 2 : rat liver ; 3 : toad liver ; 4 : tail bud embryos.
RESULTS
Glycogen
Egg glycogen dissolved in water never exhibits the milky appearance of the
same concentration of liver participate glycogen. Differential centrifugation of
egg glycogen in aqueous solution, as well as its precipitation with varying concen-
trations of ethanol, allowed a rough estimate of its extensive polydispersity.
When studying the glycogen-iodine complex, egg glycogen was compared with
liver glycogen of the same species as well as with rat liver glycogen prepared by
TABLE I
Changes in glycogen as shown by the iodine color reaction
Clutch
Stage
Iodine
phenol-sulfuric
E400/E460
1
Second cleavage
0.98
1.09
Neural fold
0.73
1.21
2
Unfertilized egg
1.02
0.98
Tail bud
0.70
1.06
3
Unfertilized egg
1.09
0.97
Tail bud
0.85
1.17
302
F. D. BARBIERI, J. S. RAISMAN AND C. ALBARRACIN
the same procedure. We found that the absorption spectrum of rat liver gly-
cogen in the presence of iodine reagent presented, in addition to the absorption
maximum at about 460 m/j, as reported by Krisman (1962), a second Amnx at
410 in/A. Both maxima have about the same height and are separated by a slight
depression (Fig. 1).
With respect to the toad, a significant difference could be detected between
liver and oocyte glycogen. While liver glycogen showed a Amax at about 400-
410 m/A, oocyte glycogen presented a Amax between 450 and 460 m/x (Fig. 1).
The affinity of glycogen for iodine was expressed by the ratio of the amounts
of glycogen as determined by the iodine and the phenol-sulfuric acid methods, the
value of this ratio being taken as 1 for the standard (Krisman, 1962). In the case
of the iodine method, estimations were based on the average of extinctions at 400
and 460 m/x. The values of this ratio varied among different batches, but were
generally slightly higher for oocytes than for liver glycogen.
When glycogen is extracted from developing eggs after gastrulation the
absorption maximum shifts from 460 towards 400 m/x. (Fig. 1). This displace-
ment is expressed in Table I by the ratio of absorbancies at 400 and 460 mju.
(E400/E4GO). Besides, the values of the ratio iodine/phenol-sulfuric indicate a
fall of glycogen affinity for iodine over 2Qc/c. This change of affinity could also be
detected in four glycogen fractions arbitrarily isolated by fractionated precipitation
with ethanol (Fig. 2).
2.0-
1.5-
1.0-
0.5
OOCYTES
MUSCULAR RESPONSE
FIGURE 2. Histograms expressing the values of the iodine/phenol-sulfuric ratio in four glycogen
fractions precipitated with 20, 30, 40 and 67% ethanol (reading from left to right).
AMYLASE AND GLYCOGENOLYSIS
303
100
I so
u
40
20
PH
-3
-1
log[cr
FIGURE 3 (left). pH optimum. Open circles: egg amylase ; solid circles: mammalian sali-
vary amylase.
FIGURE 4 (right). Effect of chloride on enzyme activity.
In connection with a point to be discussed later, it is important to add that egg
glycogen does not seem to be associated to particles sedimeriting as fraction I. In
our working conditions, less than 10/£ of the total amount of glycogen could be
detected in that fraction.
Amylase activity
The homogenates of B. arenarum eggs degrade glycogen in the same conditions
as a-amylase. Thus, the pH of maximum activity is about 6.8 and it decreases
sharply out of the range 6.4-7.8 (Fig. 3). The enzymatic activity is enhanced
by chloride ions with an optimum concentration between 0.01 and 0.1 M (Fig. 4).
Zir+ is inhibitory and 10^5 M HgCU showed no effect on the enzymatic activity
(Table II). A Km value of 3.7 mg./ml. was found.
TABLE II
Effects of ZnCl-2 and HgCli on amylase activity
Additions
Final concentration
(M)
Enzyme activity*
Fraction I
Salivary amylase
ZnCU
ID"3
10~2
87
70
79
74
10-1
0
8
HgCl2
10~5
98
—
Activity without salts taken as 100.
304
F. D. BARBIERI, J. S. RAISMAN AND C. ALBARRACIN
TABLE 1 1 1
Rc< overy of amylase activity as a function of the procedure followed in the preparation of fraction I
Preparation
Enzyme activity (%)
Sediment
Supernatant
In 0.25 M sucrose
In 0.25 M sucrose washing once
In distilled water
79
2
29
21
98
71
After fractionation of oocyte homogenates, most of the activity appears localized
in fraction I (nuclei, yolk and pigment granules), though in a very labile fashion.
In fact, it suffices to wash the pellet only once with sucrose solution, or to
homogenize the eggs with water, in order to loose the enzymatic activity of fraction
I (Table III). We cannot decide as yet to which particles the enzyme is asso-
ciated, but it does not seem to be linked to yolk platelets of major or medium size,
TABLE IV
Recovery of amylase activity from 0.25 M sucrose breis as a function of centrifugal force
Enzyme activity (%)
Sediment
Supernatant
1500 g/20 min.
500 g/30 sec.
84
2
16
98
100 •
ao
- 40
20
FIGURE 5. Enzyme activity of fractions I (nuclei-yolk-pigment) and IV (supernantant) as
a function of developmental age. Abscissa, embryonic stages : 0, unfertilized eggs ; 6, early
cleavage ; 12, gastrulation ; 17, tail bud.
AMYLASE AND GLYCOGENOLYSIS
305
TABLE V
Activity of fraction I on glycogen as shown in vitro
Minutes of incubation
Reducing power (%)
lodine/phenol-sulfuric
E«00/E460
0
3
1.15
0.91
5
25
1.08
0.97
12
59
0.97
1.05
20
100
0.83
1.13
The incubation mixture contained: 2.4 mg. of glycogen and 0.05 ml. of enzyme in 0.004 M
maleic acid-KOH buffer and 0.1 M NaCl (total volume 0.6 ml.).
0.8
0.7-
0.6-
m
GO
04-
0.3 J
380
420
460
500
540
FIGURE 6. Glycogen breakdown by fraction I of unfertilized eggs, as shown by the iodine
method. Incubation times are indicated in minutes.
306 F. D. BARBIERI, J. S. RAISMAN AND C. ALBARRACIN
since after sedimentation of most of the yolk mass at a low centrifugal field (500 g
for 30 seconds) the activity remains in the supernatant (Table IV). Neither
does it seem to be associated with the pigment granules, since an important part
of these particles also sediments at 500 g.
Although the level of amylase activity remains constant up to the end of the
neurula stages, the intracellular localization of the enzyme seems to change as a
function of embryonic development (Fig. 5). While the activity of fraction I,
which is the most important in the oocytes, decreases during development, the
activity of fraction IV, i.e., of the supernatant, increases at the same time. It is
after gastrulation when the major part of the enzymatic activity appears in frac-
tion IV at the expense of fraction I. The properties of this "soluble amylase" were
found to be the same as those described for the enzyme linked to fraction I of the
unfertilized egg.
The in vitro attack of glycogen by the enzyme of fraction I was also studied
with the iodine color reaction. After incubating oocyte glycogen with a suspen-
sion of fraction I of the same origin, it was isolated by the HgCl2 method and
analyzed with the iodine reagent. The data collected in Table V and Figure 6
show that the optical properties of the glycogen iodine complex change in the same
way after glycogen breakdown in vivo or in vitro.
DISCUSSION
The iodine color reaction has proved useful to get some additional informa-
tion about glycogen utilization in amphibian eggs. Let us point out first that the
different absorption spectra reported for the toad and rat liver glycogens agree
with previous observations indicating that the properties of the glycogen-iodine
complex depend on the origin of the polysaccharide (Schlamowitz, 1951 ; Manners,
1957). Much more knowledge will be needed before the real basis of these differ-
ences may be understood ; some evidence, however, is available showing that the
iodine reaction reflects some structural aspects of the glycogen molecule. Thus,
it has been shown that the value of the specific absorptivity coefficient is a function
of the chain length and the value of the Amax is related to the degree of branching
of the polysaccharide (Swanson, 1948; Thoma and French, 1960; Archibald et al.,
1961; Bailey and Whelan, 1961). This holds also true for the reaction per-
formed following the technique used in this paper (Krisman, 1962).
We have pointed out that after gastrulation, when a consumption of glycogen
can be already detected, a fall of its affinity for iodine as well as a shift of the
Amax from 460 towards 400 m/*, takes place. On the basis of Krisman's (1962)
results, it can be assumed that a shortening of external branches has occurred.
It is pertinent to observe that at this stage, when the iodine spectra of egg and
liver glycogens become similar, we are dealing in both cases with actively metabo-
lized glycogens. Oocyte glycogen, on the other hand, with a Anjax at 460 mp.,
should be characterized by a relatively slow turnover rate. This last assumption
finds some additional support in the following facts : ( 1 ) The apparent stability
of glycogen level, even during egg segmentation (Barbieri and Gil, 1962) ; (2)
the low respiratory activity with a R.Q. about 0.6 (Legname and Barbieri, 1962) ;
(3) the negligible amounts of lactic acid contained in normal eggs (Barbieri and
AMYLASE AND GLYCOGENOLYSIS 307
Salomon, 1963). The apparent lack of activity exhibited by this cell and in this
conection the widespread idea of viewing the unfertilized egg as an "anesthetized
cell" (Brachet, 1960) should also be taken into account. Therefore, the iodine
method seems to reveal, in the unfertilized egg of B. arenarum, the presence of a
more "complete" or nearly "untouched" glycogen molecule.
We have found that glycogen isolated from eggs, as well as from several other
sources, exhibits a high degree of polydispersity (Staudinger, 1948; Stetten et al.,
1956; Manners, 1957; Barber et al., 1965; Mordoh et al., 1966). The fact that
most of its molecules, independently of their size, seem to be simultaneously metab-
olized, as shown by the iodine method, also agrees with previously reported results
(Stetten and Stetten, 1960; Barber et al., 1965).
Concerning our first observations in connection with the enzymes involved in
the breakdown of egg glycogen we have established the presence of an enzyme with
the properties of a mammalian a-amylase. Although only crude preparations have
been used, some chromatographic controls of the reaction products, as well as the
requirement of chloride ions, seem to exclude the presence of a relatively important
glucosidase or phosphorylase activity in our working conditions.
It has already been mentioned that a /?-amylasic activity has been described in
eggs of Rana esculenta and Bufo vulgaris (Urbani, 1962) , although no conclusive
evidence has been provided. It is not unlikely that what has been taken for a
/3-amylase was really an a-amylase. During short incubation periods, such as those
utilized by the Italian authors, a-amylase attacks only the outer branches of gly-
cogen, with the formation of linear oligosaccharides (Olavarria and Torres, 1962).
This means that during this first step of enzyme action, the analytical methods
employed would not allow a clear-cut distinction between the two amylolytic path-
ways. Besides, while a-amylases exhibit a pH optimum between 6 and 7 and
/?-amylases an optimum below pH 6 (Fischer and Stein, 1960; French, 1960),
"both" amylases of B. vulgaris eggs were found to have the same optimum at pH
7.2 (Scollo Lavizzari, 1956). We have also shown that the enzymatic activity of
B. arenarum eggs remains unaffected in the presence of 10~5 M HgCl2, which is
known to inhibit /?-amylase activity at that concentration. Finally, for the time
being and at the present stage of our knowledge, /3-amylases should be circum-
scribed to the plant kingdom (French, 1960).
A point of special interest concerns the intracellular localization of this enzyme.
LpVtrup (1955), in Amblystoma mexicanum eggs, as well as Urbani and collabora-
tors, in eggs of Anura, have found that the amylolytic activity should not be linked
to yolk. Contrarily, as it has been shown, in the eggs of B. arenarum it should be
associated to particles sedimenting at the same rate as the yolk platelets of minor
size. This disagreement could be explained by the extreme lability of the enzyme-
particle association. In fact, we have shown that if B. arenarum eggs are homoge-
nized in water, as by the above-mentioned authors, most of the enzymatic activity
remains in the supernatant after sedimentation of nuclei, yolk and pigment. In
this connection, it is interesting to point out that when Urbani and Scollo Lavizzari
(1955) measured amylase activity on parts of dissected embryos in the tail-bud
stage they found the greatest activity in the portion richest in yolk; but as they
found no activity in isolated yolk platelets, their conclusion was that the enzyme
was localized in the protoplasm of the vitelline cells.
308 F. D. BARBIERI, J. S. RAISMAN AND C. ALBARRACIN
The observation that amylase seems to be progressively released from its sup-
porting particle in the course of development has a promissory value concerning
the intracellular localization of its substrate. Thus, on the assumption that gly-
cogen is not linked to particles of fraction I, it is tempting to speculate that we
are dealing with a regulatory mechanism of enzyme activity based on the spatial
orientation of enzyme and substrate. It is true that glycogen could also be linked,
in a very labile fashion, to the same particle to which amylase appears associated,
but we do not count as yet with any experimental evidence supporting this possi-
bility. Besides, there are reasons to believe that this regulatory mechanism is
operative in vivo : ( 1 ) At the beginning of development, amylase activity appears
restricted to fraction I ; (2) as the activity of this fraction decreases throughout
development, the activity of the supernatant increases (total activity remaining the
same) ; (3) only one enzyme is involved in both fractions as far as we can judge,
considering the properties analyzed in the present work; and (4) the changes
suffered by glycogen, as shown by the iodine method, were the same after being
attacked in vivo and in vitro.
L0vtrup (1955) has found that amylase activity in the non-yolk fraction of
eggs of Amblystoma seems to increase through the stage of gastrulation, remaining
unchanged once neurulation sets in. If we suppose that in these eggs the enzyme
is linked to some participate elements and that homogenization leads to a partial
detachment of the former, we may assume that L0vtrup's graph representing
"amylase synthesis" actually is an expression of the passage of the enzyme from
the particles to the supernatant.
A regulatory enzymatic mechanism of this kind, depending upon the spatial
orientation of enzyme and substrate, has already been proposed to explain the con-
trol of respiration in amphibian eggs (Spiegelman and Steinbach, 1945). It is
not unlikely that such a mechanism was more generalized in these eggs than is
currently believed. In this sense, our results give a new support to the view that
yolk, more than as a simple reservoir of materials for the building up of the embryo,
might function as an active part in metabolic control (Barth and Barth, 1954;
Wallace, 1961).
Taking into account our limited knowledge about the function of amylase in
adult tissues, as well as the important role that it seems to play in the breakdown
of amphibian glycogen, further investigations along this line are being programmed.
Appreciation is expressed to Dr. Nicolas Bazan for his assistance at the begin-
ning of this work, and to Dr. Peter Seeligmann for critical reading of the
manuscript.
SUMMARY
1. An enzyme with the properties of a mammalian a-amylase in the eggs of
the toad Bufo arenarum is described. The enzyme appears associated in a very
labile fashion to particles sedimenting at 1500 g for 20 minutes in 0.25 M sucrose
solution.
2. Egg glycogen is polydisperse and does not seem to be linked to the same
particles to which amylase appears associated. At the beginning of development
it reacts with iodine in a different way than liver glycogen of the same species, as
AMYLASE AND GLYCOGENOLYSIS 309
was shown by their absorption spectra; after gastrulation, when glycogen is sup-
posed to be actively metabolized, the spectra of embryo and adult glycogens be-
come similar.
3. Some evidence is presented indicating that egg glycogen in the course of
development is degraded by the action of amylase, which would be progressively
released from its compartment.
LITERATURE CITED
ARCHIBALD, A. R., I. D. FLEMING, A. M. LIDDLE, D. J. MANNERS, G. A. MERCER AND A.
WRIGHT, 1961. a-l,4-Glucosans. Part XL The absorption spectra of glycogen- and
amylopectin-iodine complexes. /. Chem. Soc., 1183-1190.
ASH WELL, G., 1957. Colorimetric analysis of sugars. In: Methods in Enzymology, edited by
S. P. Colowick and N. O. Kaplan, Academic Press, New York.
BAILEY, J. M., AND W. J. WHELAN, 1961. Physical properties of starch. I. Relationship be-
tween iodine stain and chain length. /. Biol. Chem., 236: 969-973.
BARBER, A. A., W. W. HARRIS AND G. M. PADILLA, 1965. Studies of native glycogen isolated
from synchronized Tetrahymena pyriformis (HSM). /. Cell Biol., 27: 281-292.
BARBIERI, F. D., AND J. A. GIL, 1962. Utilizacion del glucogeno en huevos en desarrollo de
Bujo arenarum. Arch. Bioq. Quim. Farm. — Tucumdn, 10: 41-47.
BARBIERI, F. D., AND H. SALOMON, 1963. Some experimental studies on glycogen and lactic
acid in the developing eggs of Bufo arenarum. Ada Embryol. Morphol. Exper., 6:
304-310.
EARTH, L. G., 1946. Studies on the metabolism of development. J. Exp. Zool, 103: 463-486.
BARTH, L. G., AND L. J. BARTH, 1954. The Energetics of Development. Columbia University
Press, New York.
BOELL, E. J., 1955. Energy exchange and enzyme development during embryogenesis. In:
Analysis of Development, edited by B. H. Willier, P. A. Weiss and V. Hamburger,
W. B. Saunders Co., Philadelphia.
BRACKET, J., 1934. fitude du metabolisme de 1'oeuf de grenouille (Rana jusca} au cours du
developpement. I. La respiration et la glycolyse, de la segmentation a 1'eclosion.
Arch. Biol., AS: 611-727.
BRACKET, J., 1960. The Biochemistry of Development. Pergamon Press, London.
BRACKET, J., AND J. NEEDHAM, 1935. fitude du metabolisme de 1'oeuf de grenouille (Rana fusca)
au cours du developpement. IV. La teneur en glycogene de 1'oeuf de la segmentation
a 1'eclosion. Arch. Biol., 46: 821-835.
COHEN, A. I., 1954. Studies on glycolysis during the early development of the Rana pipicns
embryo. Physiol. Zool., 27: 128-141.
DUBOIS, M., K. A. GILLES, J. K. HAMILTON, P. A. REBERS AND F. SMITH, 1956. Colorimetric
method for determination of sugars and related substances. Anal. Chem., 28: 350-356.
FISCHER, E. H., AND E. A. STEIN, 1960. a-Amylases. In: The Enzymes, edited by P. D.
Boyer, H. Lardy and K. Myrback, Academic Press, New York.
FRENCH, D., 1960. /3-Amylases. /;;: The Enzymes, edited by P. D. Boyer, H. Lardy and K.
Myrback, Academic Press, New York.
GREGG, J. R., 1948. Carbohydrate metabolism of normal and of hybrid amphibian embryos.
J. Ex p. Zool, 109: 119-134.
GREGG, J. R., AND S. L^VTRUP, 1950. Biochemical gradients in the axolotl gastrula. C. R. Lab.
Carlsberg, Ser. Chim., 27: 307-324.
GREGG, J. R., J. J. MAC!SAAC AND M. A. PARKER, 1964. Anaerobic glycolysis in amphibian
development. Homogenates. Biol. Bull., 127: 256-270.
HEATLEY, N. G., AND P. E. LINDAHL, 1937. Studies on the nature of the amphibian organiza-
tion center. 5. The distribution and nature of glycogen in the amphibian embryo. Proc.
Roy. Soc. London, Ser. B., 122: 395-402.
JAEGER, L., 1945. Glycogen utilization by the amphibian gastrula in relation to invagination and
induction. /. Cell. Comp. Physiol., 25: 97-120.
310 F. D. BARBIERI, J. S. RAISMAN AND C. ALBARRACIN
KRISMAN, C. R., 1962. A method for the colorimetric estimation of glycogen with iodine.
Anal. Biochem., 4: 17-23.
LEGNAME, A. H., AND F. D. BARBIERI, 1962. El metabolismo respiratorio en el desarrollo
inicial de Bnfo arenarum. Arch. Bioq. Quim. Farm. — Tucumdn, 10: 31-39.
L^VTRUP, S., 1955. Chemical differentiation during amphibian embryogenesis. C. R. Lab.
Carlsberg, Ser. Chim., 29: 261-314.
MANNERS, D. J., 1957. The molecular structure of glycogens. Advances in Carbohydrate
Chemistry, 12: 261-298.
MORDOH, J., C. R. KRISMAN AND L. F. LELOIR, 1966. Further studies on high molecular weight
liver glycogen. Arch. Biochem. Biopliys., 113: 265-272.
OLAVARRIA, J. M., AND H. N. TORRES, 1962. Mechanism of action of liver a-amylase. /. Biol.
Chem., 227: 1746-1751.
PISANO, A., 1956. Metodo para mantener la hipofisis de anfibio fisiologicamente "in vitro."
Arch. Farm. Bioq. — Tucumdn, 7 : 387-392.
RAVEN, CH. P., 1935. Experimented Untersuchungen iiber den Glycogenstoffwechsel des
Organisationszentrums in der Amphibiengastrula. II. Proc. Kon. Ned. Akad. v.
Wetensch., Amsterdam, 38: 1107-1109.
SCOLLO LAVIZZARI, G., 1956. II comportamento delle alfa e beta amilasi durante lo sviluppo
embrionale e larvale di Bufo vulgaris. Ric. Scicnt., 26: 2146-2152.
SCHLAMOWITZ, M., 1951. On the nature of rabbit liver glycogen. II. Iodine absorption spec-
trum. /. Biol. Chem., 190: 519-527.
SPIEGELMAN, S., AND H. B. STEINBACH, 1945. Substrate-enzyme orientation during embryonic
development. Biol. Bull., 88: 254-268.
STAUDINGER, H., 1948. tlber natiirliche Glykogene. Makromol. Chcm., 2: 88-108.
STETTEN, D. W., AND M. R. STETTEN, 1960. Glycogen metabolism. Physiol. Revs., 40: 505-534.
STETTEN, M. R., H. M. KATZEN AND D. STETTEN, 1956. Metabolic inhomogeneity of glycogen
as a function of molecular weight. /. Biol. Chcm., 222: 587-599.
SWANSON, M. A., 1948. Studies on the structure of polysaccharides. IV. Relation of the iodine
color to the structure. /. Biol. Chcm., 172: 825-837.
THOMA, J. A., AND D. FRENCH, 1960. The starch-iodine-iodide interaction. Part 1. Spectro-
photometric investigations. /. Amcr. Chem. Soc., 82: 4144-4147.
URBANI, E., 1962. Comparative biochemical studies on amphibian and invertebrate development.
In: Advances in Morphogenesis, edited by A. Abercrombie and J. Brachet, Academic
Press, New York.
URBANI, E., AND G. SCOLLO LAVIZZARI, 1955. Studio delle amilasi nello sviluppo embrionale e
larvale di Bnfo vulgaris. Ric. Scicnt., 25: 1151-1159.
WALLACE, R. A., 1961. Enzymatic patterns in the developing frog embryo. Develop. Biol., 3:
486-515.
WOERDEMANN, M. W., 1933. Ueber den Glycogenstoffwechsel des Organisationszentrums in der
Amphibiengastrula. Proc. Kon. Ned. Akad. v. Wetensch., Amsterdam, 36: 189-193.
ASSOCIATION-FORMATION BETWEEN PHOTIC AND SUBTLE
GEOPHYSICAL STIMULUS PATTERNS— A NEW
BIOLOGICAL CONCEPT1
F. A. BROWN, JR. AND Y. H. PARK
Department of Biological Sciences, Northwestern University, Evanston, Illinois 60201
Living systems have recently been demonstrated to be able to distinguish among
strengths and vector directions of magnetic (Brown, 1962a; Palmer, 1963; Picton,
1966), electrostatic (Brown, 1962b), and gamma radiation fields (Brown, 1963;
Brown and Park, 1964) of the order of magnitude of those of their natural terres-
trial environment. Such responsiveness and a number of its properties have been
assayed by quantitative studies of orientational tendencies of organisms as diverse
as Paramecium, Volvox, planarians, snails, and fruit flies. Demonstrations of such
properties as (a) that the maximum capacity of mud-snails to resolve direction of
a horizontal magnetic vector occurs at the strength of the local natural one (Brown,
Barnwell and Webb, 1964), (b) that effects of brief exposures to magnetic fields
deviating slightly from the earth's may persist for many minutes following restor-
ation of the natural field (Brown, Barnwell and Webb, 1964; Brown and Park,
1965a), (c) that a sense of geographic direction in the absence of all obvious en-
vironmental cues can be duplicated by a response to experimental horizontal mag-
netic vectors (Brown, 1962a; Brown, Webb and Barnwell, 1964), and (d) that
monthly (lunar day) periodisms in behavior can be abruptly phase-shifted by alter-
ing direction of the horizontal magnetic vector (Brown and Park, 1965b), suggested
that response to geomagnetism might play some normal role in the lives of the
organisms.
In addition, numerous unpublished observations during extensive studies with
planarians and mud-snails collectively suggested that organisms possessed some
kind of "memory" for geographic directions which did not depend upon any obvious
cues. The hypothesis was suggested that the living systems might form associa-
tions between their ambient fields of obvious factors and the concurrent pervasive
three-dimensional complex of electromagnetic forces of their environment. To
test this hypothesis the following series of simple experiments were designed and
conducted.
MATERIALS AND METHODS
About 20,000 Dugesia dorotocephala were collected in the field on one day in
September, 1965, and were maintained in the laboratory for the duration of the
study. They were kept in enameled steel containers with aluminum covers that
excluded nearly completely the laboratory illumination. The containers were set
in running tap water whose temperature ranged systematically through the year
from about 19° C. in July to about 5° C. in February. The worms were exposed
to the diffuse laboratory illumination for about 2 hours twice a week while they
were fed beef liver.
irThis study was aided by a contract with the Office of Naval Research (1228-30) and by
a grant from the National Science Foundation (GB 3481).
311
312 F. A. BROWN, JR. AND Y. H. PARK
Employing apparatus and methods that have been described in some detail in
earlier reports (Brown, 1962a, 1963) on planarian orientational tendencies, sam-
ples of the worms taken from the stock supply were assayed usually between 9 and
11 AM in each of four types of experimental series. Unvarying were the worms'
field of illumination and all other obvious orienting influences in the worms'
environment.
The fixed light field for the worms consisted of ( 1 ) a point-source directly
above the origin of the one-inch test course and hence non-orienting, (2) a hori-
zontal point-source directly behind the initially directed worm, and (3) a horizontal
point-source at right-angles to the right of the initial path direction. In this 3-light
field the mean path of the test worms always veered to the left, reflecting their
negative phototaxis. The degree of the turning was quantified as the angular
deviation from straight forward (0°) of the worms' path at the end of a 1-inch
free run. The variables were geographic orientation of the whole apparatus and
hence initial orientation of the worms, and time. In each of the four types of
experimental series, five or six worms were placed in the apparatus at the begin-
ning of the series and were test-run repeatedly until the the series was finished.
Then these worms were permanently discarded.
Series IA comprised determining in immediate succession the mean of each of
three 15-path samples, requiring about 5 minutes for each sample, for initially
North-directed flatworms and then rotating the apparatus with its contained worms
clockwise by 180° to South and assaying again the mean paths for three imme-
diately following 15-path samples. Series IB involved the same procedure and
sequence except that the apparatus was rotated 180° counterclockwise. Series IA
and IB were carried out first with equal frequency, and fresh worms were always
used for the second.
Series II A involved a fully parallel procedure to I A, except that the worms
were assayed first while South-directed followed by 180° clockwise rotation of the
apparatus to North and reassay of the worms. Series IIB was like IIA except
that the direction of apparatus rotation was counterclockwise.
All experimenters worked with equal frequency with both Series I and Series II.
Two different, identically constructed, orientation apparatuses were employed and
usually on any given day both a Series I and a Series II were being conducted
concurrently. The study was extended rather uniformly over a period of 10
synodic months — from October 8, 1965, through July 29, 1966. By this means all
uncontrolled geophysical variables were essentially randomized. The number of
series observed for each calendar month was as follows :
Series I N -» S Series II S -» N
1965 1 October (8) 13 11
2 November 19 16
3 December 20 20
1966 4 January 26 24
5 February 28 28
6 March 23 24
7 April 21 20
8 May 20 21
9 June 16 17
10 July (29) 21 19
207 200
GEOMAGNETISM AND LIGHT RESPONSE
313
The mean turning response was calculated for the three 15-path samples for
the initial direction, North or South. Then the mean turning was determined for
the first, second and third 15-path samples immediately following the 180° direc-
tional change. The values following clockwise rotations were computed separately
from those following the counterclockwise ones.
The data for each of the four series for the 10 months were next reduced to
mean turning for each day of the synodic month from full moon minus 15 days to
full moon plus 15 days, and three-day moving means for these monthly variations
were calculated. Such a moving mean was employed to provide a more dependable
indication of any systematic variation related to moon phase since each value could
be the mean of a sample of 18 to 21 days of data instead of only 5 to 8, and at the
same time, appropriately less emphasis would be accorded single monthly-day means
which by chance had been based upon a smaller number of days.
RESULTS
In Figure 1 are depicted the mean monthly variations obtained for the worms
initially directed Northward for the two independent series conducted consecu-
tively on a given day, together with the mean monthly variations of the same two
21°
23'
<
a
z
<
FM
+ 14
FIGURE 1. A. Mean monthly variation in path of Dugesia initially directed North with
orienting light sources behind and to the right of the worms during a 15-minute assay period.
B. Same, immediately after 180° clockwise rotation of the whole apparatus to South. C and D.
Repeat of same except with counterclockwise rotation. Ordinate : Degrees of turn to left.
Abscissa : Days relative to full moon.
314
F. A. BROWN, JR. AND Y. H. PARK
zr
23
X
H
til
520*
f «. *. /
: '• / *
fL.
FIGURE 2. A. Mean monthly variation in path of Dugesia initially directed South with
orienting light sources behind and to the right of the worms during a IS-minute assay period.
B. Same, immediately after 180° clockwise rotation of the whole apparatus to North. C and D.
Repeat of same, except with counterclockwise rotation.
worm-samples after rotation of the apparatus to South, clockwise and counter-
clockwise. Generally similar to one another are the two independently obtained
monthly patterns of variation of the worms while Northbound (Fig. 1, A and C).
Quite different, however, are the monthly patterns of variation following the 180°
rotation. The two independently derived patterns for these South-directed worms
(Fig. 1, B and D) resemble one another rather well despite the opposite directions
of the preceding rotation from North to South.
In Figure 2 are shown the comparable monthly patterns for the worms initially
directed Southward and thereafter rotated to North. Notable in this figure are
several things. First, the two independently determined monthly patterns for the
South-directed worms (Fig. 2, A and C) are both quite different from the patterns
for the South-directed worms immediately following rotation from North (Fig. 1,
B and D). They also show some striking differences between them, though for
both the maximum for right turning occurs 5 to 7 days before full moon and the
minimum occurs after full moon. But equally evident is an apparent strong tend-
ency for the monthly pattern of variation of the worms after rotation to North to
repeat, in general, the same monthly pattern shown by them when previously
South-directed (Fig. 2, B and D). Again, as in Series I A and IB, the character
of the pattern after rotation appeared independent of the direction of the rotation.
In Figure 3 are plotted the mean monthly patterns, with about a quarter of a
cycle repeated, and now centered on new moon. The data for the two series in-
GEOMAGNETISM AND LIGHT RESPONSE
315
itially North-directed have been averaged together, neglecting direction of rotation,
as have also the two initially South-directed ones. In Figure 3A the monthly
variation of the South-directed worms following initial North-direction has been
temporally displaced by 180°. Evident by inspection is the fact that the worms
rotated to South, after North, have a monthly variation of closely the same form
23 t
FIGURE 3. A. Combined values for clockwise and counterclockwise rotated worms first
North-directed (open circles) and secondly South-directed (dots). The South-directed lunar
monthly relationship has been 180° phase-shifted relative to the North-directed. B. The com-
parable combined data for the worms first South-directed (dots) and then North-directed (open
circles).
316
F. A. BROWN, JR. AND Y. H. PARK
as that which was present when initially North-directed but has become temporally
displaced by 180° and has a slightly greater amplitude. Again is evident the quite
different situation for the initially South-directed worms where the pattern shows
a strong tendency to repeat the same form and lunar phase relationships after the
180° rotation to North and even to have essentially the same mean amplitude of
variation (Fig. 3B).
DISCUSSION
The form and the phase relations of the monthly variation in turning of the
initially North-directed worms resemble closely the pattern that has existed stead-
ily over a continuous five-year investigational period (Brown, 1962a, 1963; Brown
and Park, unpublished observations). The variation is overt during Autumn and
Winter but in Spring and Summer becomes somewhat obscured by greater variance
of the samples. It is, however, readily evident during the latter two seasons as
mean monthly cycles of the same gross form and phase relations but of significantly
decreased amplitude.
The striking phase-shift of the monthly pattern observed after the initially
North-directed worms were rotated to South resembles a comparably altered
monthly-pattern phase-shift previously reported to follow an abrupt experimental
2°c
2
£ Zf
VJ-
NM
FM
\
\
A \
NM
FM
\
NM
CONTROLS
REVERSED MAGNETIC VECTOR
FIGURE 4. The mean monthly variations of North-directed worms during mornings and
afternoons in the earth's natural field (open circles) and 30 to 40 minutes after experimental
reversal of the horizontal magnetic vector at 0.05 gauss, and, indeed, IS to 25 minutes after
removal of the reversed experimental fields (dots). (Redrawn from Brown and Park, 1965b,
to illustrate the 180° temporal phase-shift.)
GEOMAGNETISM AND LIGHT RESPONSE 317
reversal of the ambient horizontal magnetic vector. When the reversed vector
was only about 25% of the strength of the natural geophysical one, the alteration
was essentially immediate, but when it was about 25 times the natural field, the
alteration was completed only after about 40 minutes of transient states (Brown
and Park, 1965b). The latter study had been conducted both mornings and after-
noons and the mean monthly cycle-form differed slightly from that obtained for
mornings only during the same period. But notable in the results, replotted for
comparative purposes in Figure 4, was a very similar 180° temporal shift of the
monthly pattern of variation. In neither the present study involving rotation of
the apparatus nor the earlier one with reversal of the horizontal vector of magne-
tism could the results be explained as a simple cycle inversion. In both, the worms
appeared to exchange their turning relations with respect to full moon and new
moon but at the same time to retain their characteristic asymmetrical cycle-form
through what amounted essentially to a 180° cycle-displacement along the temporal
axis. In other words, the total monthly pattern was included in the full moon-new
moon "exchange." The similarity of the results from these two kinds of experi-
ments, apparatus rotation and magnetic-field rotation, suggests strongly that the
chief factor in effecting the results following the 180° rotation of the worms and
apparatus in this experiment from North to South is the direction of the ambient
horizontal magnetic vector.
The worms are negatively phototactic. While after the rotation of the appara-
tus from North to South the worms appeared to vary in the strength of their nega-
tive phototaxis in a nearly opposite manner, the total explanation cannot be a simple
sign-change in phototaxis in view of the detailed asymmetrical pattern participating
in the observed alteration.
The worms which were rotated to North into an asymmetrical field with light
sources to South and East, after a 15-minute period of residence while South-
directed in an asymmetrical field with light sources to North and West, clearly did
not respond like worms placed freshly from diffuse illumination into the North-
directed apparatus. There must have been an influence of the earlier specific light-
magnetic field relationship still persisting after the worms were rotated. And since,
comparably, the monthly pattern for South-directed worms with the North and
West light sources, after residence for 15 minutes in the asymmetrically lighted
field with sources to South and East while North-directed, differed substantially
from that of worms placed initially in the South-directed field, there must, again,
have been a persisting influence of the earlier specific light-magnetic field vector
relationship.
It is not known at this time why there should have been the very conspicuous
180° displacement of the characteristic North-directed monthly pattern after the
180° rotation of the worms to South while, on the contrary, there was an equally
conspicuous tendency of the worms rotated 180° from South to North to retain
not only essentially the same form but also the same phase relationship of their
monthly pattern. Hence, while this study indicates the existence of a capacity of
the worms to form associations between vector directional components of such
overt environmental factors as light on the one hand and such a subtle pervasive
geophysical factor as ambient magnetism on the other, and that these associations
may persist for at least many minutes, we are still far from a complete understand-
ing of their nature, properties, and biological significance.
318 F. A. BROWN, JR. AND Y. H. PARK
From Figure 3A it is seen that after rotation of the apparatus from North to
South there appears superimposed on the general 180° -shifted pattern an apparent
tendency toward exaggeration of the left-turning behavior over full moon, especially
for 1 to 3 days immediately after full moon. This is the time of month of maxi-
mum left-turning in the monthly pattern of the worms which are initially South-
directed. Correspondingly, in Figure 3B, after rotation of the apparatus from
South to North, the major observed difference in otherwise rather similar patterns
is an exaggeration of the amount of right-turning over the period of full moon.
The last is the normal behavior of the worms when initially North-directed. In
brief, these observations suggest that after apparatus-rotation, the mean response
pattern for the subsequent period of approximately 15 minutes contains a mixture
of two components: (a) a persisting influence of the preceding light-magnetic field
vector relationship to which the worms had been exposed, and (b) a characteristic
pattern for the worms for that particular geographic direction, other factors equal.
Of great interest will be the results of experiments directed toward determining the
rate of acquisition of these light-subtle field associations and duration of their
persistence.
This study suggests strongly that a living system is able essentially to code an
ambient geographic pattern of illumination on a 360° geographic grid of such a
subtle geophysical vector field as that of magnetism, and to retain this spatially
coded information for at least many minutes. Such a coded geographic "cycle"
may provide concurrent alternative, or mutually-supplementing, directional cues,
which serve during homing and navigation. The disclosure of such an organismic
capacity to code an illumination pattern upon a subtle geophysical variation related
to geographic direction provides a basis for postulating that the living system, for
any fixed geographic direction, is able to code temporally varying information of
such an overt factor as illumination on a comparable 360° temporally varying
subtle geophysical grid such as that related to the solar day or lunar day. Also
supporting such an hypothesis are the gradually accumulating observations that
the mechanisms of organismic orientation in time and space possess a common
denominator (Brown, 1965). Such an organismic capacity may well serve as a
fundamental basis of the phenomena of biological rhythms and clocks and of the
apparent clock-dependent astrotaxes.
The "circa" character (not exactly solar-day or lunar-tidal length) of many
observed periodisms of animals and plants in unvarying illumination offers no
obstacle to an hypothesis that coded, recycling temporal "tapes" are essential, under-
lying components of the biological clock system, since these odd periodisms may
be accounted for as simply a systematic slippage of the cyclic coded patterns along
the "tapes" (autophasing).
SUMMARY
I
1. Evidence is presented that an organism is able to form associations between
concurrent ambient vector patterns of light and such a pervasive ambient environ-
mental component as geomagnetism.
2. These associations appear to persist for at least many minutes.
3. Some of the implications of this newly disclosed, extraordinary biological
capacity for the still unresolved mechanisms of biological clocks and compasses
are discussed.
GEOMAGNETISM AND LIGHT RESPONSE 319
LITERATURE CITED
BROWN, F. A., JR., 1962a. Responses of the planarian, Dugesia, and the protozoan, Paramecium,
to very weak horizontal magnetic fields. Biol. Bull., 123 : 264-281.
BROWN, F. A., JR., 1962b. Response of the planarian, Dugesia, to very weak horizontal elec-
trostatic fields. Biol. Bull, 123 : 282-294.
BROWN, F. A., JR., 1963. An orientational response to very weak gamma radiation. Biol.
Bull, 125 : 206-225.
BROWN, F. A., JR., 1965. A unified theory for biological rhythms : Rhythmic duplicity and
the genesis of "circa" periodisms : In : Circadian Clocks. Ed. J. Aschoff. North
Holland Publ. Co., Amsterdam.
BROWN, F. A., JR., AND Y. H. PARK, 1964. Seasonal variations in sign and strength of gamma-
taxis in planarians. Nature, 202 : 469-471.
BROWN, F. A., JR., AND Y. H. PARK, 1965a. Duration of an after-effect in planarians following
a reversed horizontal magnetic vector. Biol. Bull., 128 : 347-355.
BROWN, F. A., JR., AND Y. H. PARK, 1965b. Phase-shifting a lunar rhythm in planarians by
altering the horizontal magnetic vector. Biol. Bull., 129 : 79-86.
BROWN, F. A., JR., F. H. BARNWELL AND H. M. WEBB, 1964. Adaptation of the magnetorecep-
tive mechanism of mud-snails to geomagnetic strength. Biol. Bull., 127 : 221-231.
BROWN, F. A., JR., H. M. WEBB AND F. H. BARNWELL, 1964. A compass directional phenomenon
in mud-snails and its relation to magnetism. Biol. Bull., 127 : 206-220.
PALMER, J. D., 1963. Organismic spatial orientation in very weak magnetic fields. Nature,
198: 1061-1062.
PICTON, H. D., 1966. Some responses of Drosophila to weak magnetic and electrostatic fields.
Nature, 211: 303-304.
RESPONSES OF BATS FROM TEMPERATE REGIONS
TO CHANGES IN AMBIENT TEMPERATURE
WAYNE H. DAVIS AND OLA B. REITE 1
Department of Zoology and Department of Physiology & Biophysics,
University of Kentucky, Lexington, Kentucky 40506
Although bats in temperate regions are usually thought to migrate to warmer
climates or to retreat underground beyond the frost line for the winter, many
species commonly encounter freezing temperatures. Big brown bats (Eptesicus
juscus) winter in buildings in areas where the outdoor temperature may drop to
below —30° C. (Nero, 1959). This bat appears in numbers in caves only during
very cold weather, and individuals move to and from their hibernating sites through-
out the winter (Mumford, 1958). Pipistrels (Pipistrellus subflavus), Indiana bats
(Myotis sodalis) and little brown bats (Myotis lucifugus) all hibernate in caves.
The pipistrel is a hardy bat whose numbers present at the place used for hibernation
depend upon the severity of the weather (Davis, 1959). Indiana bats enter the
caves in mid-autumn, and occasionally die of cold when hibernating too near the
entrance. The little brown bat enters hibernation in early autumn, and some indi-
viduals retreat into crevices, apparently in response to cold weather. A few bats
of this species also succumb to cold during hibernation (Davis and Hitchcock,
1965). The red bat (Lasiurus borealis) is a tree-dwelling species which winters in
regions where temperatures frequently stay well below freezing for days (Davis
and Lidicker, 1956).
As ambient temperature drops below 0° C. a hibernating bat may respond in
one of three ways. It may increase its body temperature and arouse from hiberna-
tion ; it may remain in hibernation and maintain its body temperature by increasing
the metabolism sufficiently to compensate for increased heat loss ; or it may remain
passive during cooling and eventually freeze. Maintenance of a relatively stable
body temperature by metabolic compensation has been reported for Lasiurus
borealis and Myotis hicifugus (Reite and Davis, 1966).
The present investigation was undertaken to relate the known differences in the
ecology of the five species of bats mentioned above to possible differences in their
response to changes in ambient temperature.
MATERIALS AND METHODS
Experimental animals (Table I) were captured in Kentucky. All were obtained
from buildings or caves except the red bats, which were netted near certain cave
entrances where they regularly appear in late summer and early fall. Upon arrival
at the laboratory, the bats were placed in a temperature-controlled room and kept
1 International Postdoctoral Research Fellow, U. S. Public Health Service (Award No.
1 F05-TW-687-01). Present address: Institute for Experimental Medical Research, Ullevaal
Hospital, Oslo, Norway.
320
TEMPERATURE RELATIONS OF BATS
321
overnight at 5° C. Some were restrained by taping the wings to a wooden block,
and a copper-constantan thermocouple connected to a Brown electronic potenti-
ometer was inserted into each bat's rectum. Other bats were left unrestrained with
rectal thermocouples taped in place, and still others without any recording device
were put into individual jars covered with a screen. The next morning temperature
was lowered or increased stepwise by units of 2-10° C. Ambient and rectal tem-
peratures were recorded continuously, and the bats were observed for signs of
activity. Electrocardiograms were taken on two or more individuals of each species.
The leads were fastened to the skin of the bats with alligator clips, and the signals
recorded with a Sanborn amplifier and recording system.
RESULTS
Ambient temperature of 5° C.
After staying overnight in the room with the temperature control set at 5° C.,
all specimens appeared to be dormant except for P. subflai'iis and E. fuscus, of
which species single bats were occasionally found to arouse spontaneously and
re-enter the dormant state even when undisturbed. Most unrestrained bats sus-
TABLE I
Experimental animals
Number
Species
Dates of collection
Body weights (g.)
Males
Females
L. borealis
8
2
Sept. -Oct.
9-13
M. lucifugus
15
10
Sept. -Dec.
7-10
M. sodalis
8
9
Oct. -Dec.
8-10
E. fuscus
8
8
Sept.
19 25
P. subflavus
19
6
Oct. -Dec.
5-6
pended themselves by their feet from the screen covering the jars. L. borealis
brought the large furred interfemoral membrane up over the body, covering the
wings, except at the wrists, and all the ventral surface up to the chin (Fig. 1).
Rectal temperatures of dormant bats at 5° C. were always less than one degree
above ambient. Heart rates varied considerably among species (Table II).
Spontaneous and rapid changes in heart rate, not accompanied by noticeable
changes in conditions of torpor or rectal temperature, were noted in E. fuscns and
P. subflavus.
Increasing temperature
Stepwise increase in ambient temperature could be performed up to about 10°
C. without any change in the appearance of the bats. All seemed to remain dormant.
When ambient temperature was changed from 10° C. to 15° C., arousal began in
all species except L. borealis. In one series of experiments which included 8-12
unrestrained individuals of each of E. fuscus, P. subflavns, M. hicijmjus and M.
sodalis, 3-4 bats of each species were active within 60-80 minutes of the increase
322
WAYNE H. DAVIS AND OLA B. REITE
I
npgp*"
FIGURE 1. Male red bat, L. borcalis, in dormancy at 5° C.
in temperature. In experiments with L. borcalis there were no signs of arousal at
15° C. In two individuals of this species taken to higher temperatures, arousal was
induced at about 20° C.
During passive warming to 10° C., rectal temperatures remained within 1° C.
above ambient. Bats which began to arouse after ambient temperature was raised
to 15° C. showed an increase in rectal temperature. In E. fitscus the temperature
reached 35-38° C. in 1-2 hours.
Heart rates after passive warming of M. Incifiu'/iis, E. fiiscus, and L. borcalis to
10° C. are given in Table II. Arousal from the dormant state was always accom-
panied by a rapid increase in heart rate.
TABLE II
Heart rates in dormant bats. Counts were made over several 30-60-sec. periods chosen at random
from 6-8 recordings each lasting 5-10 min. and taken at intervals of at least 1 hour
Ambient temperature (° C.)
Heart rate (beats/min.)
L. borealis
5
10-16
L. borealis
10
16-22
HI. lucifugus
5
24-32
M. lucifiigus
10
44-56
E. fuscus
5
42-62
E. fust-us
10
64-88
M . sodalis
5
36-62
P. subflavus
5
24-80
TEMPERATURE RELATIONS OF BATS
323
Decreasing temperature
Striking differences among species were evident in response to decreasing
ambient temperature to 0° C. and below. E. fuscus invariably aroused from dor-
mancy and became active within 40-120 minutes. Both restrained and unrestrained
individuals were able to remain active for several hours, even at an ambient tem-
perature as low as —5° C. Attempts to make them re-enter dormancy by keeping
them overnight at —3° C. to —5° C. were unsuccessful, but this could easily be
achieved by changing ambient temperature to 5° C. About half the P. snbflavus
studied aroused following a temperature change from 5° C. to 0° C. None aroused
when the temperature was lowered from 5° C. to —5° C. in one step. M. lucifugus
remained dormant when ambient temperature was lowered to 0° C. Further lower-
40|—
35
30
25
20 —
"
-5
15 —
Temperature setting changed
• Rectal temperature
x Ambient temperature
10 20 30 40 50
Time (min)
60
70
80
FIGURE 2. Changes in rectal temperature of E. fuscus during arousal from dormancy
in response to lowering of ambient temperature.
ing to —5° C. induced arousal in a few, whereas an abrupt change from 5° C. to
-5° C. induced arousal in all bats of this species. M. sodalis responded similarly.
L. borealis remained dormant both during stepwise and abrupt lowering of ambient
temperature from 5° C. to —5° C. Abrupt exposure of L. borealis to subfreezing
temperatures induced an almost instantaneous increase in the rate and depth of
respiration.
Following a change of ambient temperature from 5° C. to between 0° C. and
-5° C., dormant E. fuscus showed a slight decrease in rectal temperature followed
within about 30 minutes by a rapid increase (Fig. 2). The increase in rectal tem-
perature was accompanied by increasing heart rate. As reported previously (Reite
and Davis, 1966), L. borealis and M. Incifiit/its show an increase in the difference
324
WAYNE H. DAVIS AND OLA B. REITE
between rectal and ambient temperature and an increased heart rate when ambient
temperature is gradually decreased. A similar response was found in M. sodalis.
P. siibftai'its, tested for response to an abrupt lowering of ambient temperature,
showed low rectal temperature even after more than one hour at — 5° C. A com-
parison between unrestrained P. snbflavns (5 individuals) and M. Incifiigns (7 in-
dividuals) by exposing them to a temperature change from 5° C. to — 5° C. revealed
a marked difference in their responses. After one hour the rectal temperatures of
P. subflavits ranged from --1.5° C. to 0° C. Four of the M. lucifugus were fully
active with rectal temperature ranging from 34° C. to 39° C., and the other three
were in various stages of arousal (rectal temperatures 5-14° C.). P. subflavus
kept at —5° C. for 3—4- hours died.
50 —
40
10
x Supercooled bat
• Dormant bat
-5 -2.5 0 2.5 5
Rectal temperature (°C)
7.5
10
FIGURE 3. Changes in heart rate in ,17. Incifitc/its during passive warming of a supercooled
individual from —5° C. to 5° C. and a dormant one from 5° C. to 10° C.
Supercooling and freezing
Rectal temperatures of restrained dormant M. lucifugus, M. sodalis and L.
borcalis kept at -5° C. stayed I—I-0 C. above ambient for 2-3 hours, the central
body temperature being probably still higher than that of the rectum. During this
time, the bats continued breathing and maintained increased heart rate. When kept
at —5° C. for longer periods the breathing ceased and rectal temperature dropped
to ambient, indicating that the temperature gradients between central parts of the
body and the periphery were disappearing. Only bats in this latter condition will
be termed supercooled, although local tissue temperatures in bats of the former
category were also well below the freezing point of tissue fluids. Spontaneous
tissue freezing followed by death occurred in some of the supercooled bats.
Whether or not freezing occurred was determined by inspection. In those bats
TEMPERATURE RELATIONS OF BATS
which did not freeze, heart beats continued and rectal temperature remained con-
stant at —5° C. Several bats of each species were removed from the cold room
and exposed to room temperatures after having been kept in the supercooled state
with ceased breathing for 5-8 hours, while others in the same condition were left
in the cold room, where the temperature was then changed to 5° C. We tested all
bats for ability to respond to stimuli by forceably opening the eyes and mouth,
extending the wings, and probing the bodies with fingers. Unlike the situation
during dormancy, the bats lacked muscle tone and did not respond to stimuli. After
8-16 minutes at an ambient temperature of 24° C. the bats which had been removed
from the cold room began slight movements of the feet and soon resumed breathing.
\Yithin 30—15 minutes they had recovered completely and were capable of normal
flight. They were kept overnight in the laboratory, tested again for normal flight,
and released in apparently good condition. In the supercooled bats left in the cold
room at 5° C., heart rate increased slowly during rewarming (Fig. 3), breathing
was resumed, and these bats also recovered completely.
No attempt was made to test survival time of supercooled bats which had ceased
breathing. \Ye succeeded in keeping one individual of M. lucifiigiis supercooled at
—6.5° C. for half an hour and rewarming it without ice formation. This bat also
survived. Occasionally, during exposure to sub-zero temperatures, freezing oc-
curred in peripheral tissues of bats which had not become supercooled. These
animals recovered if removed from the cold room before ice formation had taken
place in more central parts of the body. However, when spontaneous freezing
occurred in already supercooled animals it seemed inevitably to be lethal. Freezing
in supercooled bats was accompanied by a rapid rise in body temperature to a level
between —0.5° C. and --1° C., and the process could be induced by pricking with
a needle. Supercooling to below — 5° C. for any length of time seemed difficult, as
further lowering of the temperature greatly enhanced the tendency to spontaneous
ice formation.
DISCUSSION
Since Eisentraut (1934) published one of the first accounts on the different
physiological states of bats with respect to ambient temperature, numerous reports
have followed. The literature is covered in a recent review (Stones and Wiebers,
1965).
Hibernating bats have different ways of surviving when ambient temperatures
drop below freezing. According to our experiments, E. fuscits will awaken, a re-
action which also seems to be normal in other hibernators (Hock, 1960). During
extremely cold weather it is not unusual for bats of this species to be seen in flight
in buildings. Common speculation among students of Chiroptera is that over-
heating of the hibernating sites by the heating systems of the buildings is the likely
cause of arousal. From the present results we believe it is more probable that the
hibernating sites become too cold, which causes the bats to move to warmer places.
This assumption is supported by the observation that E. fuscus undergoes con-
siderable intercave movement at air temperatures below freezing (Mumford, 1958).
L. borealis responds to freezing temperatures by increasing its metabolism
enough to maintain its body temperature above a dangerous lower limit (Reite and
Davis, 1966). These animals do not hibernate in caves, and it would be to their
326 WAYNE H. DAVIS AND OLA B. REITE
disadvantage to arouse, since arousal would use far more energy than regulation
during dormancy. Red bats also remain dormant when ambient temperature rises.
Our observations indicate that whereas other species will arouse at temperatures
between 10° C. and 15° C., red bats remain in dormancy up to a temperature of
about 20° C. unless handled or otherwise disturbed. This is an important adapta-
tion. These animals are exposed to wide fluctuations in temperature in their
natural environment, and cannot afford to become active until it is warm enough
to obtain enough food to compensate for loss of stored energy. Thus in winter
they do not fly unless it is warm enough for insect flight. Davis and Lidicker
(1956) found that red bats became active only on days when temperatures rose to
19° C. or above. Constantine (1958) observed the closely related L. serninolus
hibernating in their natural environment, and reported that they awoke and flew
only when environmental temperatures reached 21° C. Both reports give support
to the present findings. L. borcalis joins other species of bats in swarming at the
caves in early fall (Davis, 1964), but never hibernates there. Occasionally, red
bats enter rooms in certain caves, cannot find their way out again, so hang up and
become dormant. Such bats invariably perish (Myers, 1960), perhaps being unable
to arouse spontaneously at cave temperatures. Thus red bats seem to be so adapted
to survival outside that they are unable to survive in caves.
The anatomical structure and the behavior of L. borealis are better modified for
survival at low temperatures than those of any of the other species studied. Except
for the ears and parts of the wings, this bat is completely furred. The furred inter-
femoral membrane and the long tail, in relation to body length, is probably of sig-
nificance in heat conservation during hibernation when the bat uses the interfemoral
membrane to cover the ventral surface ( Fig. 1 ) . The short rounded ears may also
enable this bat to tolerate exposure to cold better than the other species.
The only bat in which the response was not what might be expected from
previous knowledge of its behavior in its natural environment was P. sub flams.
Since bats of this species move into the caves in numbers only after periods of
freezing weather, we would expect them to arouse as temperatures approach and
go below freezing. However, our experiments showed that they aroused only in
response to moderate lowering of ambient temperature. The reason may be that
their small size does not allow them to generate enough heat to exceed the heat
loss when ambient temperature is lowered abruptly to —5° C. Even in larger bats
the increase in body temperature is slow during the initial steps of arousal. The
low tolerance of P. sitbflai'its to supercooling makes it reasonable to believe that
their natural way of responding is to come out of hibernation. They may survive
outside the caves during early moderate cold periods and arouse and move into the
caves when the most severe part of the winter is approaching. Folk (1940) has
suggested that the first cold periods of winter may indicate to bats the suitability of
their resting place for hibernation.
Both Kayser (1940) and Hock (1951) noted increased respiratory exchange
in dormant bats exposed to temperatures near 0° C. The justification for con-
sidering this as a true thermoregulatory response is supported by the present
observations in bats of the species L. borcalis, M . lucijiigns and M. sodalis, which
established an increased difference between rectal and ambient temperature when
the latter was decreased.
TEMPERATURE RELATIONS OF BATS 327
If taken as an indicator of metabolism, tbe heart rate in dormant bats at 5° C.
should reflect the relative efficiency or depth of hibernation in the different species.
Of the species studied, L. borealis may be considered best adapted for hibernation.
Those with the highest heart rates (E. fnscns, M. sodalis and /'. snbflants) should
be more prone to arousal. This assumption is supported by the finding of spon-
taneous rapid changes in the heart rate of dormant E. fnscns and P. subflai'iis,
and also by the observation that individuals from these species occasionally became
active and re-entered dormancy even when kept at a stable ambient temperature of
5° C. M. IncifiKjns is intermediate. \Yhether this bat will respond to a lowering
of ambient temperature by increasing its metabolism enough to compensate for
increased heat loss (thermoregulation ) or by arousal from dormancy, seems to
depend on the abruptness and severity of the cold exposure. Seasonal differences
may also be present. The heart rate of M. Incifiu/iis in dormancy at 5° C. is in
the same range as that reported by Johansen and Krog (1959) for the birchmouse,
a hibernator of comparable size.
Cooling of bats to about -5° C. without formation of ice in the body is in
agreement with previously obtained results ( Kalabuchow, 1935 ). The slow increase
in heart rate in supercooled M. Incifngns during rewarming from —5° C. to 5° C.
( Fig. 3 ) corresponds fairly well to the rate change found in isolated hearts of this
species over that part of the same temperature range where such studies have been
performed (Michael and Menaker. 1963). The heart rate in dormant M. lucifugus
at temperatures of 5-10° C. is also in the same range as that of the isolated heart.
These observations suggest that in supercooled bats with ceased breathing and in
dormant bats at neutral ambient temperatures (5-10° C. ) the heart is not under
any neural influence. This is different from the situation in dormant bats of L.
borealis and .17. Incifiujns exposed to stepwise lowering of ambient temperature
from 5° C. In these bats the heart rate increases with decreasing temperature
(Reite and Davis. 1966).
Supercooling of bats could be of significance for survival during short term
exposure to sub-zero temperatures, a situation which may occur following a change
of wind direction at the entrance to a cave used for hibernation. However, super-
cooling is an unstable condition and must be transient.
\Ye wish to thank Roger \Y. Barbour for photographing the dormant red bat,
and Loren D. Carlson, Eugene C. Crawford and Marion D. Hassell for reading
the manuscript.
SUMMARY
1. Responses to upward and downward changes in ambient temperature from
5° C. were studied in dormant bats of the species E. fnscns. P. snbflavns, M. sodalis,
M . lncifii(/ns and L. borealis. Rectal temperatures and heart rates were recorded.
2. Except for L. borealis which did not arouse until ambient temperature
reached about 20° C., all species responded by arousal from dormancy when the
temperature was increased to 15° C.
3. The effects of decreasing ambient temperature varied considerably among
species. E. fnscns invariably aroused from dormancy. L. borealis never aroused
but showed a thermoregulatory response by increasing its metabolism to compensate
328 WAYNE H. DAVIS AND OLA B. REITE
for the increase in heat loss. The responses of the other species depended upon the
abruptness of the temperature change. Abrupt lowering of ambient temperature
tended to induce arousal in M. lucijii</iis and M. sodalis, whereas these species re-
sponded similarly to L. borealis when exposed to gradually decreasing temperature.
P. snbflavns usually aroused in response to a gradual decrease in ambient temper-
ature, but seemed unable to arouse in response to abrupt lowering of temperature.
4. Bats of the species L. borealis, AI. lucijugus and AI. sodalis supercooled to
-5° C. showed cessation of breathing, but slow heart beats continued for several
hours. Passive rewarming was necessary for survival.
5. Many of the known differences in the ecology of the studied species of bats
are reflected as differences in their response to changes in ambient temperature in
the laboratory.
LITERATURE CITED
CONSTANTINE, D. G., 1958. Ecological observations on lasiurine bats in Georgia. /. Mammal.,
39 : 64-70.
DAVIS, W. H., 1959. Disproportionate sex ratios in hibernating bats. J. Mammal., 40: 16-19.
DAVIS, W. H., 1964. Fall swarming of bats at Kentucky caves. Bull. Natl. Spclcolog. Soc.,
26 : 82-83.
DAVIS, W. H., AND H. B. HITCHCOCK, 1965. Biology and migration of the bat, Myotis lucijugus,
in New England. /. Mammal., 46: 296-313.
DAVIS, W. H., AND W. Z. LIDICKER, JR., 1956. Winter range of the red bat, Lasittrns borealis.
J. Mammal., 37 : 280-281.
EISENTRAUT, M., 1934. Der Winterschlaf der Fledermause mit besonderer Beriicksichtigung der
Warmeregulation. Z. Morphol. Okol. Ticrc, 29 : 231-267.
FOLK, G. E., JR., 1940. Shift of population among hibernating bats. /. Mammal., 21: 306-315.
HOCK, R. J., 1951. The metabolic rates and body temperatures of bats. B'wl. Bull, 101: 289-
299.
HOCK, R. J., 1960. Seasonal variations in physiologic functions of arctic ground squirrels and
black bears. In : Mammalian Hibernation. Ed. by C. P. Lyman & A. R. Dawe.
Bull. Mus. Comp. Zool, 124: 155-169.
JOHANSEN, K., AND J. KROG, 1959. Diurnal body temperature variations and hibernation in
the birchmouse, Sicista betulina. Amcr. J. Physio!., 196: 1200-1204.
KALABUCHOW, N. L, 1935. Anabiose bei Wirbeltieren und Insekten bei Temperaturen unter 0°.
Zool. Jahrb., Abt. allg. Zool. Physio!., 55: 47-64.
KAYSER, CH., 1940. Les echanges respiratoires des hibernants a 1'etat de sommeil hivernal.
Ann. Physiol. Physicochim. Biol., 16: 127-221.
MICHAEL, C. R., AND M. MENAKER, 1963. The effect of temperature on the isolated heart of
the bat, Myotis lucijugus. J. Cell. Comp. Physio!.. 62 : 355-358.
MUMFORD, R. E., 1958. Population turnover in wintering bats in Indiana. /. Mammal., 39 :
253-261.
MYERS, R. F., 1960. Lasiurus from Missouri caves. /. Mamma!., 41 : 114-117.
NERO, R. W., 1959. Winter records of bats in Saskatchewan. Blue Jay, 17 : 78.
REITE, O. B., AND W. H. DAVIS, 1966. Thermoregulation in bats exposed to low ambient
temperatures. Proc. Soc. Exp. Biol. Mcd., 121 : 1212-1215.
STONE-S, R. C., AND J. E. WIEBERS, 1965. A review of temperature regulation in bats
(Chiroptera). Am. Midi. Naturalist, 74: 155-167.
VESICULATED AXONS IN HAEMAL VESSELS OF AN
HOLOTHURIAN, CUCUMARIA FRONDOSA
WILLIAM L. DOYLE
Department of Anatomy. University of Chicago. Chicago, III. 60637 , and
Mount Desert Island Biological Laboratory, Salisbury Cove, Maine
There is only fragmentary evidence of neurosecretory activity in echinoderms
(Fontaine, 1962; Bullock and Horridge, 1965). A primitive haemal system occurs
as a rete in the sea cucumber, Cucumaria. In this rete we have found nerve strands
containing large dense-cored vesicles in axons which are distributed to the non-
striated muscle of the vessel wall.
Specimens were collected from Frenchman Bay, Maine, and kept in running
sea water at 15° C. for a few days. Segments of the haemal rete were ligated to
prevent contraction, excised and fixed for electron microscopy. Fixation in a 3.5%
glutaraldehyde in phosphate buffer was followed by treatment with \% osmium
tetroxide. Other specimens were fixed in \% osmium tetroxide with 0.3 M sucrose
in the fixative, followed by 10% formalin. Tissues were embedded in epoxy resin
and sections stained with uranyl acetate and lead citrate.
The wall of the haemal vessel has three layers. The outer layer consists of
coelomic epithelial cells, nerve strands and non-striated muscle cells. The inter-
mediate layer consists of a thick, distinctly filamentous basal lamina adjacent to
the muscle fibers and a deeper connective tissue consisting of a mucoid matrix con-
taining fibers with a periodicity of 640 A. The inner layer consists of more or less
contiguous cells with processes embedded in the fibrous layer, constituting an
endothelium. The "endothelial" cells have fine structural features similar to those
of fibroblasts in higher forms.
In the outer layer the coelomic epithelial cells have a columnar peripheral por-
tion containing large oil droplets and, at the level of the nucleus, a prominent Golgi
region. The basal portion of these cells extends as two or more tapering processes
which pass between the fibrous portions of the muscle cells to reach the basal
lamina (Figs. 1, 6, 7 F). In these processes masses of fine parallel filaments fill
the terminal portions. Adjacent to the basal lamina the fibrous portions of the
muscle cells are oriented in the plane of the basal lamina. Most of the fibers are
circumferential but a few are longitudinal. There are cytoplasmic processes from
the muscle cell extending into the basal lamina. The nuclei and most of the
cytoplasm of the muscle cells are found peripherally among the processes of the
coelomic epithelial cells but most of the mitochondria are adjacent to the myofila-
ments. Rather narrow strands of cytoplasm may connect the nuclear and fibrillar
portions of the muscle cells (Fig. 1).
Passing between the numerous processes of the coelomic and muscle cells are
groups of axons forming nerve strands (Fig. IN). At bifurcations of the rete the
strands of axons apparently cross each other (Fig. 4). Nerve strands three to five
microns in diameter are common in the wall of the haemal rete (Figs. 2, 3). From
329
330
WILLIAM L. DOYLE
these strands single axons are distributed to the fibrous portions of the muscle cells
(Fig. 6). Individual axons vary considerably in cross-sectional area (Fig. 4) and
there are frequent expansions which may be rather empty or containing accumula-
tions of mitochondria, lipid and dense-cored vesicles.
Within the axons microtubules are preserved after glutaraldehyde fixation (Fig.
3) and are quite uniformly 260 A in diameter. A small amount of granular lipid
is present. There is a variety of sizes of vesicles in the axoplasm and rows of large
(0.2 to 0.3 micron ) membrane-bounded bodies with a clear space surrounding a
' •
N
L
F
M
* **„'
-
FIGURE 1. Cross-section of the inner portion of the outer layer of the wall of the haemal
vessel. Basal lamina, B, at lower left corner and muscle fibers, M, adjacent. The slender dense
cell processes extending diagonally across the figure are extensions of coelomic epithelial cells
containing filamentous masses, F, and a few mitochondria and lipid granules. A cell with lipid
granules, L, is at upper left. In one muscle cell, M, sections of four mitochondria are located
among the muscle fibrils and have associated lipid granules. The larger groups of mitochondria
in adjacent cells are in the cytoplasmic portions of other muscle cells. Cross-sections of parts
of two nerve strands, N, are evident midway in the upper and right hand edges of the figure.
Osmic fixation, 7900 X.
VESICULATED AXONS IN CUCUMARIA
331
FIGURE 2. Longitudinal section of a nerve strand with several axons running from upper left
to lower right. Within the axuns are rows of dense vesicles. Glutaraldehyde fixation, 8500 X.
FIGURE 3. Parallel axons in a nerve strand containing microtubules 260 A in diameter and
membrane-bounded dense vesicles 0.1 to 0.3 micron in diameter. The individual axons approxi-
mate 0.5 micron in width. Glutaraldehyde fixation, 27,000 X.
332 WILLIAM L. DOYLE
FIGURE 4. Section through a plexus of axons illustrating variations in diameter. Diameters
at lower right range from 0.4 to 1.5 microns. Other axons may be identified by the presence of
microtubules and specific dense vesicles. The small dense granules in axons are lipid.
Glutaraldehyde fixation, 9200 X.
dense core. These specific vesicles are often elongate and the density of the core
varies. In less dense cores the contents can be seen to consist of aggregates of
smaller granules or dense vesicles.
No ganglia have been found in the wall of the rete and only occasional single
nerve cell bodies along the nerve strands. These cell bodies are distinguished by a
folded nucleus, a dilated endoplasmic reticulum which has more ribosomes associated
with it in the perinuclear region than in peripheral zones, relatively few mitochon-
dria, a Golgi region, microtubules in the cell processes and in some cases large
numbers of vesicles. In one such cell (Fig. 5) the cytoplasm is filled with a variety
of heavy-walled vesicles with granular and vesicular contents. In Figure 5 these
vesicles are in close association with the Golgi region and give the impression of
arising from it. The contents of these vesicles are much less dense than seen in
the vesicles found in the axons and some look like multivesiculate bodies. In
general, however, the contents are more heterogeneous than seen in multivesiculate
bodies and in the elongate forms the contents form denser aggregates. In other
cell bodies the dense-cored vesicles have been fewer in number but uniformly more
dense, with only a few showing a multivesiculate appearance. Wherever found, the
densest particles frequently give some evidence of aggregated composition. It is
uncertain whether these bodies are the precursors of those found in the axons but
both the heterogeneous and the dense-cored vesicles are confined to the nerve cell
VESICULATED AXONS IN CUCUMARIA
333
body and axons. No similar vesicles have been found in coelomic epithelial cells,
and in hundreds of sections containing muscle cells we have found only a few
similar granules in muscle cell cytoplasm.
Sections of cell processes distant from the cell body have been found containing
much more uniform populations of a hundred or more dense-cored vesicles.
Dilations of the axons in the nerve strands also show accumulations of several
vesicles and a few mitochondria and lipid granules.
Axons leaving the nerve strand are often about 0.6 micron in diameter and
taper gradually to 0.2 to 0.3 micron at their terminations at the muscle cells.
Single dense vesicles may fill the cross-section of the axon termination.
At the muscle cell surface the axons terminate as slightly flattened processes
with a distinct intercellular space between the membranes of muscle and nerve cells.
/
'"
k
FIGURE 5. Portion of a cell body with a tangential section of the Golgi region surrounded
by vesicles of a variety of sizes including some multivesiculate forms. The internal vesicles are
of differing densities and degrees of aggregation. Glutaraldehyde fixation, 16,000 X.
334
WILLIAM L. DOYLE
y . '
Vo
•u
V
B
FIGURE 6. Section of a muscle fiber, M, with axon terminals. One axon contains the
specific dense vesicles and adjoins the muscle cell membrane at the right. Another, without
specific granules, inserts into an invagination, I, of the muscle. Osmic fixation, 11,800 X.
There is no evident membrane specialization in the regions of approximation.
Some of the nerve terminations are in channels of the muscle cell surface with
overlapping muscle cell processes. Some axons insert into invaginations of the
muscle cell surface but always surrounded by distinct extracellular space (Figs.
6, 7, 8, 9).
DISCUSSION
Cytochemical and electron microscopic evidence for the presence of glycogen is
negative in these tissues. The large oil droplets of coelomic epithelial cells and of
muscle cells have been observed breaking up into small granules of irregular outline
as previously described in the respiratory tree ( Doyle and McNiell, 1964) and this
lipid may substitute for the glycogen of higher forms. In our preparations the small
granules of lipid have irregular outlines and appear as densely stained particles in
all of the cells of the outer wall of the haemal rete.
The present evidence on the occurrence of large (0.3 /A) vesicles containing
dense aggregates in axons distributed to non-striated muscle cells suggests a
possible neurosecretory function in this primitive vascular system. Very little is
known of the organization of this part of the nervous system in these organisms.
In common with other neurosecretory vesicles the ones present in these axons stain
intensely. They appear to arise in the cell body in proximity to the Golgi region.
They are distributed along the axons and do not accumulate at the terminals which
VESICULATED AXONS IN CUCUMARIA
335
-£v4 ',
N
« f •
, *> ' & '
* *-.
; ?
«» '
B
• f
V
I
N
•
?
' :•*• %
t \ ^
J \ »
\$ *'< »i ':• *//• ^
^ >-.;.- «v\%
* * * -» .. * « .
•*,
\ A
FIGURES 7-9. Cross-sections of fibrous portions of muscle cells with tubular invaginations,
I, containing axons. Osmic fixation, 35,000 X.
336 WILLIAM L. DOYLE
lack a terminal expansion. The specific vesicles are with rare exceptions confined
to the nerve cells and their processes. A specific search for evidence of discharge
or transfer of these vesicles has revealed a very few instances of the presence of
similar structures in the muscle cell cytoplasm. These very few instances may in
fact represent evidence for transfer of vesicles of a transmitter suhstance but we
have no evidence that this is so. In one instance four dense-cored vesicles were
seen in the non-fibrous portion of the muscle cell cytoplasm and in another three.
In other instances single bodies were observed in the fibrous portion. No evidence
has been found of fusion with the cell membrane or discharge from the axon.
Ultrastructural relationships of nerve processes and smooth muscle cells have
been described and reviewed recently by Thaemert (1966) while Lever et al.
(1965) have reported on axon terminals in the arteriolar wall. The occurrence of
small specific vesicles has been commonly reported in these studies in higher forms.
The close contiguity of nerve and muscle and occurrence of channels in the muscle
cell are similar to the relations found in Citcmiiaria. The specific dense vesicles
found in Ciiciiiiiaria are much larger structures and their neurosecretory nature
remains to be established. Similar structures have been reported in Hydra by Lentz
and Barrnett (1965).
This work was supported by a grant, GB 3035, from the National Science
Foundation.
SUMMARY
Segments of the primitive haemal rete of the holothurian, Ciictunaria, were fixed
in glutaraldehyde and in osmic acid, embedded in epoxy resin, sectioned for electron
microscopy and stained with uranyl acetate and lead citrate. Multifibered nerve
strands were found among the epithelial cell processes of the wall of the haemal
vessels. Individual axons containing large (0.2 to 0.3 micron) membrane-bounded
dense-cored vesicles are distributed to the non-striated muscle cells. The vesicles
arise in association with the Golgi region of the neurone and large numbers are
found in proximal cell processes. The vesicles containing dense aggregates are
distributed along the axons, with a few present at the tapered terminal portions
at the muscle cell.
LITERATURE CITED
BULLOCK, T. H., AND G. A. HORRIDGE, 1965. Structure and Function in the Nervous Systems of
Invertebrates. San Francisco : Freeman.
DOYLE, W. L., AND G. F. McNiELL, 1964. The fine structure of the respiratory tree in
Cucumaria. Quart. J. Micr. Sci., 105: 7-11.
FONTAINE, A. R., 1962. Neurosecretion in the ophiuroid, Ophiopholes. Science, 138: 908-909.
LENTZ, T. H., AND R. J. BARRNETT, 1965. Fine structure of the nervous system of Hydra.
Amer. ZooL, 5 : 341-356.
LEVER, J. D., J. D. P. GRAHAM, G. IRVINE AND W. J. CHICK, 1965. Vesicular axons in relation
to arteriolar smooth muscle in the pancreas. (Brit.} J. Anal., 99: 299-313.
THAEMERT, J. C, 1966. Ultrastructural interrelationships of nerve processes and smooth
muscle cells in three dimensions. /. Cell Biol., 28 : 37-49.
THE DIGESTIVE SYSTEM OF THE HOLOTHURIAN, CUCUMARIA
ELONGATA. I. STRUCTURE OF THE GUT AND HEMAL SYSTEM
JOHN D. FISH1
Dove Marine Laboratory, University of Newcastle Upon Tyne, England
The structure and function of the echinoderm digestive system have been the
subjects of several recent papers. Anderson (1953, 1959) has made valuable
contributions to the study of digestion in asteroids, and both Stott (1955) and Fuji
(1961) have studied the structural and functional aspects of the echinoid gut by
means of histological and histochemical techniques. As early as 1883 Hamann gave
detailed accounts of the gut histology of the holothurians, Leptosynapta and
Holotlinria, and more recently Stott (1957) has studied the alimentary canal and
associated structures in Holothuria forskali. Choe (1962) has given an account
of gut structure and digestive enzymes found in Stichopus japonicus, and the
feeding and digestive processes of this holothurian have been studied by Tanaka
(1958). However, the process of digestion in holothurians is still not fully under-
stood. The function of the hemal system is open to controversy, and the role of
the amoebocytes in digestion has yet to be conclusively demonstrated. To provide
a fuller understanding of the process of digestion it is necessary for further detailed
histological and histochemical studies to be accompanied by the results of phys-
iological studies. This paper forms an introduction to the study of digestion in
Ciiciunaria clongata, and deals with the histology and histochemistry of the gut.
It is intended that a second paper will deal with the distribution of the digestive
enzymes.
It is a pleasure to acknowledge the help and encouragement given by Dr. J. B.
Buchanan, who supervised this study. Dr. D. B. Lewis gave valuable assistance
with the photography. The work was supported by a research studentship from the
Department of Scientific and Industrial Research.
MATERIAL AND METHODS
Specimens of Cucumaria were collected oft" the Northumberland coast from
depths of about 20 fathoms. Those animals required for histological and histo-
chemical studies were treated with a suitable fixative on the day of capture.
The different gut regions were dissected out in sea water and fixed in a suitable
fluid. The material was processed and embedded according to the nature of the
histological and histochemical techniques to be applied. ( 1 ) For general cell
structure, tissues were fixed in Heidenhain's "Susa" made with sea water, embedded
in paraffin wax and sectioned at 6/x. For finer histological structure and the
identification of secretory granules, tissues were fixed in Zenker-formol. (2) For
the demonstration of mucin and similar compounds (acid polysaccharides), tissues
1 Present address : Department of Zoology, University College of Wales, Aberystwyth, Wales.
337
338 JOHN D. FISH
fixed in Heidenhaiii's "Susa" were stained in dilute aqueous solutions of aluminum-
methylene blue (Heath, 1962), alcian blue at pH 3, and mucicarmine. To demon-
strate the metachromatic staining of acid polysaccharide elements, sections were
stained overnight in dilute aqueous solutions of toluidine blue (0.01%). The pH
at which the acid mucopolysaccharide lost the ability to bind with methylene blue
(methylene blue extinction, M. B. E.) was determined by staining sections over-
night in dilute solutions (0.01%) of aqueous methylene blue at different pH values.
In all cases staining was followed by rapid dehydration in 95% and absolute
alcohol. (3) For general recognition of lipid deposits tissues were fixed in Baker's
formol-calcium, soaked in 5% potassium dichromate for 24 hours at 60° C,
embedded in gelatine, sectioned at 10-15 /A on a freezing microtome, and stained
with Sudan black. For the detection of phospholipid, material was fixed in
Baker's formol-calcium and treated by Baker's acid-hematin method accompanied
by the pyridine extraction test applied to sections fixed in weak Bourn's fluid
(Baker, 1946). Material fixed in Baker's formol-calcium, post chromed, embedded
in gelatine and sectioned as above, was stained in 1% aqueous Nile blue at 60° C.
and differentiated in 1% acetic acid for the demonstration of acidic lipids (Cain,
1947). (4) For the demonstration of glycogen and related compounds, material
was fixed in a weak Bouin's fluid, paraffin-embedded, and sections exposed to the
periodic acid-Schiff reaction. Control slides exposed to the action of 1% malt
diastase in a phosphate buffer at neutrality differentiate between glycogen and
other Schiff-positive substances. (5) Identification of proteins. A full account
of methods for the identification of proteins is given by Pearse (1960) in Appendix
5, page 791.
(i) Identification of protein. Mercury-bromphenol blue method. (Formalin-
fixed, paraffin-embedded.)
(ii) Identification of tyrosine. Millon reaction. (Baker modification),
(iii) Protein-bound NH2. Nmhydrin-Schiff method. (Fixative: 85% ethanol.
Paraffin sections.)
(iv) Identification of tryptophan.
(a) Dimethylaminobenzaldehyde (D. M. A. B.) nitrate method. (Formalin-
fixed, paraffin sections.)
(b) Naphthyl ethylenediamine method: (Formalin-fixed, paraffin sections.)
A stronger color was produced by this method than by the D.M.A.B.
method.
(v) Identification of arginine. Sakaguchi reaction. (Susa-fixed, paraffin
sections. )
GUT NOMENCLATURE AND MORPHOLOGY
There is confusion between the present systems of nomenclature used for the
holothurian gut, primarily because of the morphological variation between species,
and because the names of the different gut regions appear to have been assigned
by analogy with the mammalian gut, rather than being based on functional differ-
entiation. Stott (1957) has listed the nomenclatures used by Oomen (1926),
Cuenot (1948) and Stott (1957). Choe (1962) has given a nomenclature for the
gut of Stichopiis japonicits, yet none of these is suitable for the gut of Cucumaria.
DIGESTIVE SYSTEM OF CUCUMARIA. I
339
The system of nomenclature used throughout this study is as follows: pharynx,
esophagus, stomach, constriction, intestine I, intestine II and cloaca. Each of these
regions is morphologically clearly differentiated from the others (Fig. 1). By
using terms which are familiar in the description of mammalian digestive systems,
it is not intended that any functional comparisons should he drawn. Such names
are retained only until a nomenclature based on functional differentiation can be
given.
The first region of the gut, the pharynx, lies in the center of the aquapharyngeal
bulb, and upon emergence into the body cavity it takes the form of the esophagus.
The esophagus is slender yet conspicuous, having patches of black pigmentation
at its anterior end. It is followed by a much broader and thicker-walled stomach,
which is usually of similar length, but its pink coloration contrasts with the grey
color of the esophagus. Following the stomach is a short, thin-walled region,
Pharynx
Oesophagus
\
Stomach
Constriction
Intestine II
Intestine
Cloaca
FIGURE 1. The gut of Cncumaria clongata. For details of gut nomenclature see text.
340
JOHN D. FISH
w:< ML
' •'•'V.VJ rfi
IB
11?"^ : :','$$
•,»;'•«.•. f . • ;, j.- .
. »
;4\
^SS^S^M. ?
i«
^'"«1
•• i?
FIGURE 2. Longitudinal section through the junction between stomach and constriction.
Compare the development of circular muscle in the stomach (S) with that in the constriction
(C). Masson's trichrome. 1 cm. = 100 /it.
FIGURE 3. Transverse section of intestine I, showing muscle cell bodies (a), muscle layer
(indicated by arrows), and connective tissue-fluid complex (b). C is an amoebocyte held in
the fluid complex. Masson's trichrome. 1 cm. = 10 p.
FIGURE 4. Circular muscle fibers of the stomach teased to show muscle cell bodies (indi-
cated by arrows). Masson's trichrome. 1 cm. = 20 /*.
DIGESTIVE SYSTEM OF CUCUMARIA. I 341
approximately 2-3 mm. long, which is bounded anteriorly and posteriorly by pro-
nounced constrictions. Throughout this study this region of the gut is known
as the constriction. The green color of the constriction contrasts vividly with the
brown of the intestine and the pink of the stomach. The intestine of Cuctunaria
is typical of holothurians in being very long. Intestine I is about half as long
again as intestine II. There is no color difference between these two parts, but
the change from I to II is made clear by the change from the convoluted gut wall
of intestine I to the relatively smooth-walled intestine II. The right and left
respiratory trees open into the most posterior part of intestine II, from which
point the gut is known as the cloaca. The cloaca is relatively short and runs to
the posterior extremity of the animal.
STRUCTURE OF THE GUT
The gut wall consists of a number of distinct layers which can be described
as follows :
(a) An outer covering of ciliated serosal epithelium which in places is so thin
that it can only be detected by the presence of its nuclei.
(b) To the inside of the ciliated epithelium is a distinct layer of cells which
are thought to be the cell bodies of the circular muscle fibers which lie to the
inside of them.
(c) A muscle layer which is variously represented in the different regions of
the gut. Outer circular and inner longitudinal muscle fibers are present in all
gut regions. In the stomach the muscle bands show their maximum development,
and are chiefly responsible for the thickness of the gut wall. At the junction be-
tween the stomach and the constriction there is a marked change in the muscula-
ture of the gut (Fig. 2). In the constriction the circular muscle is reduced to a
band of fibers about 4 ^ wide, while the longitudinal muscle is present as a few
scattered fibers. This condition persists throughout the intestine and cloaca.
(d) A connective tissue layer, associated with which is a fluid continuous with
that in the hemal system. The fluid component is variously represented in the
different regions of the gut and is described fully below.
(e) The mucosal epithelium which, except in the stomach, is chiefly responsible
for the thickness of the gut wall. It is composed of a single layer of tall slender
cells, among which are several cell types described fully below.
Muscle cell bodies
A distinct layer of cells, varying from 10 to 15 ju, thick and lying to the outside
of the circular muscle (Fig. 3), is thought to comprise the cell bodies of the cir-
cular muscle fibers. The layer is present in all gut regions, and is covered by the
serosal epithelium. In stained preparations cut both transversely and longitudi-
nally it is difficult to interpret the relationship between the cell bodies and the fibers
since both are densely packed. Even when pieces of the gut wall are teased and
then stained with Masson's trichrome. the relationship is still obscure. There is
no indication that these areas might be fiber bundles of a nerve layer, and the
original contention that these cells are muscle cell bodies is held in view of the
following observations. Preparations of esophagus, constriction and intestine, cut
342
JOHN D. FISH
transversely and stained in Masson's trichrome, failed to show muscle cell bodies
lying along the length of the circular fibers. In all the above regions the circular
muscle fibers form a narrow layer, approximately 12-15 /A wide in the esophagus,
compared with the thickness of a few fibers in the other regions. In these regions
the muscle cell bodies must lie to the outside of the fibers. The cell bodies are
highly distended and must be connected to individual fibers by way of short necks.
In the stomach, where the circular muscle attains a thickness of 75 /JL, muscle cell
bodies can be clearly seen lying along the length of many of the fibers (Fig. 4).
If the innermost fibers of the stomach were to have cell bodies arranged in the
manner described for other gut regions, the connection between cell body and fiber
would be via a neck in the region of 70 /A long. The presence of a thick overlying
layer of densely packed muscle fibers makes such a connection unlikely. It is
suggested that only the outermost circular fibers of the stomach have this highly
distended type of cell body, whereas all fibers have this arrangement in the regions
where the development of circular muscle is not as extensive. Although it has
i.e.
FIGURE 5. Semi-diagrammatic representation of the relationship between muscle cell bodies
and the circular muscle fibers as seen in longitudinal section: c.e., coelomic epithelium; m.c.b.,
muscle cell bodies ; c.m., circular muscle ; l.m., longitudinal muscle ; c.f., connective tissue-fluid
complex ; i.e., mucosal epithelium.
proved impossible to demonstrate conclusively that this arrangement exists, the
proposed relationship between cell bodies and the individual muscle fibers is shown
diagrammatically in Figure 5. Electron micrographs are necessary to give a clear
picture of the arrangement which exists in this part of the gut wall.
The connective tissue-fluid complex
The connective tissue layer attains its maximum development at the bases of
the villus-like projections of the esophagus where it doubtless acts as a supporting
framework. The most interesting aspect of this complex is the fluid component
which is continuous with the fluid of the hemal system (Fig. 6). In living prepara-
tions the hemal fluid has a viscous appearance, whilst on fixation it appears to
become "gelled." Amoebocytes are present in this fluid medium (see Fig. 3).
Histochemical tests indicate that the fluid is periodic acid-Schiff-positive (diastase-
fast) (Fig. 7), and contains tryptophan, arginine, tyrosine and reactive NH2
groups, together with an acid mucopolysaccharide.
DIGESTIVE SYSTEM OF CUCUMARIA. I
343
The mucosal epithelium
The mucosal epithelium is composed of a single layer of tall, slender cells which
have centrally placed nuclei. Cells specialized to produce currents in the lumen
of the gut are absent. In the esophagus the cells are formed into villus-like projec-
tions which have a connective tissue framework, and the individual cells are inter-
spersed with large, conspicuous mucous gland cells (Fig. 8). Histochemical tests
show that the glands contain an acid mucopolysaccharide with methylene blue
extinction below pH 2. Negative results were obtained with the periodic acid-
Schiff technique (Table I). Only a few of the glands extend to the basement
membrane of the epithelium; those which do have a swollen basal portion (6-7 /A).
The majority of the mucous glands are interspersed among the distal parts of the
epithelial cells, and open by way of short necks into the esophageal lumen. In
fixed preparations stained in aluminum-methylene blue, the contents of the glands
appear distinctly granular. In the pharynx the mucous glands are similarly
distributed.
TABLE I
Histochem-istry of the gland cells
Test
Esophageal glands
Constriction glands
Intestine I glands
Toluidine blue
Gamma
Gamma
Negative
metachromasia
metachromasia
Alcian blue, pH 3
Positive
Positive
Negative
Aluminum-methylene blue
Positive
Positive
Negative
Mucicarmine
Positive
Positive
Negative
Methylene blue extinction
Below pH 2
Below pH 2
Negative to
methylene blue
Periodic acid-Schifi reagent
Negative
Negative
Negative
In the stomach the cells of the mucosal epithelium are covered by a cuticle which
has a thickness of about 2 p (see Fig. 15). Mucous glands are absent and the
gland cells which have been demonstrated by Hamann (1883) in the stomachs of
Leptosynapta and Holothuria are lacking in Cucumaria. Chains of secretory
granules found in the lining epithelial cells of the stomach of Echinus esculentus
(Stott, 1955) and Strongylocentrotiis intertnedius (Fuji, 1961) have not been
demonstrated, and the most conspicuous feature of the structure of the stomach wall
of Cucumaria is its heavy musculature.
The epithelial cells of the constriction are formed into stout villus-like projec-
tions similar to those of the esophagus, and interspersed among individual cells
there are numerous mucous gland cells which invariably extend to the basement
membrane of the epithelium (Fig. 9). The base of the mucous glands has a
diameter (6/x) much greater than that of the neighboring epithelial cells. These
glands show histochemical reactions similar to those of the esophagus (Table I),
and when stained in aluminum-methylene blue the contents appear granular. Sec-
tions of the esophagus and constriction stained in aluminum-methylene blue, made
with polychrome methylene blue (Microme salt no. 1041 — E. Gurr), show differ-
ences in the staining reaction of the mucous glands after the preparations have been
344
JOHN D. FISH
Lumen
8
FIGURE 6. Transverse section of intestine I and the dorsal henial sinus, showing the
continuity between the hemal fluid (a), and the gut fluid complex (b). Masson's trichrome ;
1 cm. = 15 fi.
FIGURE 7. Longitudinal section of intestine II, showing the fluid complex. Periodic acid-
Schiff; 1 cm. = 50/M.
DIGESTIVE SYSTEM OF CUCUMARIA. I 345
stored for a few days. Mucous glands in the esophagus retain the brilliant blue-
coloration characteristic of sulfated mucopolysaccharides, while those of the
constriction change from blue to reddish purple. This would seem to indicate
differences in the chemical nature of the secretion from these glands.
Preparations of the constriction stained in Heidenhain's iron hematoxylin and
Masson's trichrome show two characteristic features. First, the swollen basal
portion of the mucous glands is represented by light-colored areas in arcades be-
tween the bases of the epithelial cells. Secondly, there is a faintly stained "fringe"
area, permeated by the ducts of the mucous glands, which represents the distal
portion of the epithelial cells (Fig. 10). Throughout this fringe region histo-
chemical tests show the presence of an acid mucopolysaccharide which has the same
histochemical reactions as the glands of the esophagus and constriction, yet distinct
gland cells are absent. Preparations stained in Heidenhain's iron hematoxylin also
show secretory cells which contain chains of secretory granules ( Fig. 11) similar
to those described by Anderson (1953) in the pyloric caeca of Asterias forbesi.
It has proved difficult to clearly establish the relationship between the secretory
granules and the secretory cell, but in most cases the granules extend in rows
towards the free end of the cell.
The organization of the epithelial cells throughout the intestine and cloaca is
similar to that in the constriction, yet mucous glands and secretory granules are
absent. Interspersed among the distal portions of the epithelial cells in intestine I,
distinct gland cells are present which open into the lumen of the intestine (Fig.
12). These cells have only been demonstrated using Heidenhain's iron hema-
toxylin, and the contents appear granular. The nucleus is situated in the proxi-
mal half of the cell. Similar gland cells have been shown in the intestine of Holo-
thuria by Hamann (1883). The distal portion of the epithelial cells of the intes-
tine, corresponding to the "fringe" zone of the constriction, show faintly positive
reactions for acid mucopolysaccharide. As in the "fringe" of the constriction,
distinct gland cells are absent.
Storage cells
The mucosal epithelial cells in all regions of the gut hold deposits of lipid (Fig.
13). which constitutes an important food reserve of the animal (Fish, 1967) . The
lipid is stored in the form of droplets which lie both above and below the nucleus.
In the distal region of the epithelial cells the droplets are generally small and
sparsely distributed while in the basal portion they appear to have coalesced into
larger globules. The Nile blue technique reveals that acidic lipids are prominent
in the composition of the lipid deposits. The acid hematin test (Baker, 1946),
accompanied by pyridine extraction, gave doubtful results for the presence of
phospholipid. Lipid deposits are also present in the much inflated cell bodies
of the circular muscular fibers. Sudan black staining shows that the muscle cell
bodies are crowded with lipid droplets, the histochemistry of which is the same
as that of lipid stored in the epithelial cells.
FIGURE 8. Transverse section of esophagus, showing mucous glands. Aluminum-methylene
blue ; 1 cm. = 35 /*.
FIGURE 9. Longitudinal section of the constriction, showing mucous glands (indicated by
arrows). Aluminum-methylene blue; 1 cm. = 35 /*.
346
JOHN D. FISH
12
;<h
.*>
13
FIGURE 10. Transverse section of the constriction. Basal portion of mucous glands indi-
cated by arrow. Note lightly stained "fringe" zone at f. Masson's trichrome ; 1 cm. = 40 /JL.
FIGURE 11. Transverse section of the constriction, showing chains of secretory granules.
Heidenhain's iron hematoxylin ; 1 cm. = 30 /j..
FIGURE 12. Longitudinal section of intestine I, showing gland cells (g). Heidenhain's
iron hematoxylin; 1 cm. = 30 /a.
FIGURE 13. Transverse section of the constriction, showing lipid deposits. Note concen-
tration of lipid in basal portion of cells. Frozen sections, Sudan black ; 1 cm. = 55 p.
DIGESTIVE SYSTEM OF CUCUMARIA. I
347
THE HEMAL SYSTEM
Typical of holothurians there is a close association between the gut and the
hemal system. The system in Cucumaria is shown diagrammatically in Figure 14,
and consists of two main sinuses — the dorsal and the ventral. There is no rete
mirabile or complicated network of lacunar tufts such as is found in Holothuria
forskali (Stott, 1957) and other large aspidochirotes, yet transverse connections
between different parts of the same sinus are evident (Fig. 14). A direct route
between the hemal system and the gut is provided by the continuity which exists
Ventral sinus
Dorsal sinus
FIGURE 14. Hemal system of Cucumaria, showing dorsal and ventral sinuses and
their connecting strands.
348
JOHN D. FISH
between the henial fluid and the connective tissue-fluid complex of the gut (see
Fig. 6).
The ventral sinus runs along the length of the intestine and constriction, yet at
the anterior end of the constriction the sinus ceases to exist as a separate channel, and
serial sections at this point show that it passes diffusely through the stomach wall
until it reaches the connective tissue-fluid complex (Fig. 15). The author is not
aware of a similar system in any other holothurian ; the more usual arrangement
is for the ventral vessel to continue along the length of the stomach and esophagus
until it reaches the hemal ring surrounding the pharynx.
Throughout its course the ventral sinus is in close association with the gut
wall. At the anterior region of intestine I the sinus gives off several transverse
; ^*' ,*Ju ••-../ ' i*
^^^1* ^%*« Jfc A .jJl : ^* t^T™* •
'** »
Wfe IfclL^ -
15
FIGURE 15. Serial longitudinal sections through the junction between stomach and con-
striction, showing the ventral hemal sinus (v) merging with the stomach wall at points indi-
cated by arrows. Note the cuticle (c) covering the mucosal epithelial cells of the stomach.
Masson's trichrome ; 1 cm. = 40 ^.
connections which join with that part of the ventral sinus which is' associated with
the posterior part of intestine I and the anterior part of intestine II.
The dorsal sinus runs along the complete length of the gut on the side which
is attached by the dorsal mesentery. It is connected to the intestinal wall by
numerous branches, and shortly after the commencement of its coitrse along intes-
tine I, it gives off a single transverse connection which joins the part of the dorsal
sinus which is associated with the anterior part of intestine II. Anteriorly the
dorsal sinus diminishes towards the pharynx. The presence of a hemal ring has
not been satisfactorily demonstrated, yet this may be due to its delicate nature
and the fact that it is believed to lie directly behind the water vascular ring.
DIGESTIVE SYSTEM OF CUCUMARIA. I
349
ce
ce
I
17
m
FIGURE 16. Dorsal hemal sinus seen in transverse section. Note coelomic epithelium, ce ;
muscle layer, m; and amoebocytes within the sinus (a). Heidenhain's iron hematoxylin,
1 cm. - 10 M.
FIGURE 17. Transverse connecting strand of the ventral hemal sinus seen in transverse
section. Note coelomic epithelium, ce; muscle layer, m; and amoebocytes within the sinus (a),
Masson's trichrome ; 1 cm. — 10 /*.
350 JOHN D. FISH
Although the small size and delicate nature of the hemal sinuses make it
difficult to ohtain good sections, available evidence suggests that all parts of the
system have the same histological structure (Figs. 16 and 17). There is an outer
layer of coelomic epithelium which contains strands of connective tissue. The thick-
ness of this layer varies from 4-5 /*, to 10-12 //.. A thin hut distinct layer of cir-
cular muscle fibers is found to the inside of the coelomic epithelium. Associated
with the circular muscle fibers there are a few scattered longitudinal fibers, to the
inside of which is an indistinct layer of connective tissue. In all sections of the
hemal sinuses the circular muscle fibers have shown as a distinct layer. This con-
trasts with Stichopns cliloronohts (Sivickis and Domantay, 1928), which has an
indistinct muscle layer in the hemal sinus. The lacunar tufts of the rete mirabile of
Actinopyga were found by Hyman (1955) to be without a muscle layer.
Sections of the hemal sinuses treated with aluminum methylene blue and the
periodic acid-Schiff technique gave negative results.
DISCUSSION
The results of histological and histochemical observations on the gut of C.
elongata present several interesting features. The relationship between the muscle
cell bodies and the individual circular muscle fibers poses problems as to the reasons
for, or advantages of, such a system. Nichols (1959) has described a similar
arrangement of muscle cell bodies in the ampullae of the tube feet of Echinocar-
dhim. Such an arrangement has not previously been recorded in the gut wall of
holothurians. The development of this system is perhaps associated with the lack
of connective tissue in the outer layers of the gut wall. The layer of inflated,
densely packed and interlocking muscle cell bodies may function as an anchorage
system for the circular muscle fibers. In the absence of connective tissue in this
region, such an arrangement would be important during phases of strong muscle
contraction.
Lipid stores are held in the muscle cell bodies as well as in the cells of the
mucosal epithelium. Nichols (1959) found that in Echlnocardluni, glycogen was
held in the muscle cell cytoplasm as a food store. Glycogen has not been detected
in either the body wall or gut of Cucumaria. The failure to obtain a positive Baker
test for the presence of phospholipid is interesting, in that Anderson (1953) and
Karnovsky et al. (1955) failed to demonstrate phospholipid in Asterias. How-
ever, as pointed out by Karnovsky ct al., the failure of the test may be due to the
low phosphorus content of the phosphatide fraction and not due to the low concen-
tration of phosphatide. In the acid hematin test, the hematin is presumed to react
with the phosphate radical. Although the specificity of the Baker test is estab-
lished (Casselman, 1952), its sensitivity has never been determined, and it may
be that in spite of negative results, phospholipids are present in the constitution
of the lipid deposits. It must also be emphasized that as marine invertebrates
have low melting point lipids (Giese, 1966), the technique used in the histochemical
localization of the lipid deposits is itself questionable, as it involved incubation in
5% potassium dichromate for 24 hours at 60° C.
Results similar to those obtained for the histochemistry of the hemal fluid in
Cucumaria have been recorded by Millot and Vevers (1964) for the axial organ
secretion in echinoids. These authors suggest that the axial organ is suitably
DIGESTIVE SYSTEM OF CUCUMARIA. I 351
positioned to act as an endocrine organ, and they have shown that considerahle
quantities of secretion leave the glandular recesses of the organ. Millot (1966)
has further suggested that the reactions of the axial organ may be part of a
"defensive injury response." In holothurians there is little agreement between
authors on the existence of an axial gland. Cuenot (1891) claims that the part
of the coelom giving rise to the axial gland disappears during embryonic develop-
ment, while other workers described a connective tissue network to the side of the
water ring which they considered to be an axial gland (vide Hyman, 1955). It
is significant to note that even though the axial organ in echinoids may secrete
fluid into the hemal system, Millot and Vevers (1964) believed it unlikely that
amoebocytes arose there. Furthermore, Holland ct al. (1965) were unable to find
evidence either for or against the participation of the axial organ in amoebocyte
production.
Although the structure and possible functions of the hemal system have been
investigated by a number of authors, the functions of the system have not been
conclusively demonstrated. A number of investigators have reported a contractile
nature for parts of the hemal system (Kaw'amoto, 1927; Prosser and Judson, 1952;
Boolootian and Campbell, 1964; see also Hyman, 1955). Prosser and Judson
(1952) further demonstrated that in holothurians the contractions were myogenic,
being accelerated by adrenalin and slowed by atropin. Burton (1964) has shown
that despite evidence of contractility of the sinus in regular echinoids, the full sig-
nificance of this is not yet clear, and it would appear unlikely that the hemal system
functions as a true circulatory system. The experiments and histological observa-
tions of Enriques (1902), Oomen (1926), and Schreiber (1930, 1932a, 1932b),
led to the hypothesis that the holothurian hemal system played an important role
in digestion, in that amoebocytes contained within the hemal fluid were believed
to carry digestive enzymes into the gut, and carry away the products of digestion.
Contrary to these earlier reports it has recently been suggested that sugars may
cross the gut wall by active transport (D'Agostino and Farmanfarmaian, 1960;
Rundles and Farmanfarmaian, 1964). It has further been shown that the hemal
sinuses are not significantly involved in nutrient transport in either echinoids or
holothurians (Farmanfarmaian and Phillips, 1962; Farmanfarmaian, 1963). Re-
sults of histochemical tests applied to sections of gut material of Ciicinnaria are
also contrary to the hypothesis of Enriques, Oomen and Schreiber. These results
show that parts of the mucosal epithelium of the gut appear to be capable of secret-
ing digestive enzymes. The constriction has abundant mucous glands and chains
of secretory granules, and intestine I has conspicuous gland cells. The distribu-
tion of mucous glands, secretory granules and gland cells would appear to indicate
that at least the constriction and intestine I are sites of enzyme production and
secretion. The possibility of ascribing a zymogenic function to parts of the lining
epithelium of the gut will be considered more closely in the second part of this
study when the results of the distribution of digestive enzymes are discussed.
SUMMARY
1. A system of nomenclature is given for the gut of Citcninaria elongata. The
different regions of the gut have been named as follows : pharynx, esophagus,
stomach, constriction, intestine I, intestine II, and cloaca.
352 JOHN D. FISH
2. The gut wall is composed of five distinct layers: (a) an outer serosal epi-
thelium; (h) muscle cell bodies of the circular muscle fibers; (c) a muscle layer
with outer circular and inner longitudinal fibers; (d) a connective tissue-fluid com-
plex, the fluid component of which is continuous with the fluid in the hemal system ;
(e) the mucosal epithelium, which is composed of a single layer of tall slender cells.
3. Interspersed among the cells of the mucosal epithelium are mucous glands,
secretory granules and gland cells. Mucous glands are present in the esophagus
and constriction ; secretory granules in the constriction, and gland cells in intestine
I. Cells specialized to produce currents in the lumen of the gut are absent.
4. Stores of lipid are held in the cells of the mucosal epithelium and in the
muscle cell bodies of the circular muscle fibers. Glycogen deposits have not been
demonstrated.
5. The histology of the hemal system has been studied and the role of the hemal
system in digestion is discussed.
6. From the distribution of gland cells and secretory granules it is suggested
that the mucosal epithelial cells of the constriction and intestine I are sites of
digestive enzyme production and secretion.
LITERATURE CITED
ANDERSON, J. M., 1953. Structure and function in the pyloric caeca of Asterias forbcsi. BioL
Bull., 105: 47-61.
ANDERSON, J. M., 1959. Studies on the cardiac stomach of a starfish, Patiria miniata (Brandt).
BioL Bull., 117: 185-201.
BAKER, J. R., 1946. The histochemical recognition of lipine. Quart. J. Micr. Sci., 87: 441-447.
BOOLOOTIAN, R. A., AND J. L. CAMPBELL, 1964. A primitive heart in the echinoid Strongylo-
centrotus purpuratus. Science, 145: 173-175.
BURTON, M. P. M., 1964. Hemal system of regular echinoids. Nature, 204: 1218.
CAIN, A. J., 1947. The use of Nile Blue in the examination of lipoids. Quart. J. Micr. Sci.,
88: 383-392.
CASSELMAN, W. G. B., 1952. Observations concerning the specificity of the acid haematin test
for phospholipids. Quart. J. Micr. Sci., 93: 381-383.
CHOE, S., 1962. Biology of the Japanese common sea cucumber, Stichopus japonicus Selenka.
226 pp. Pusan, Korea. (In Japanese with English summary.)
CUENOT, L., 1891. fitudes morphologiques sur les echinodermes. Arch. BioL, 11: 313-608.
CUENOT, L., 1948. Anatomic, ethologie et systematiques des echinodermes. In: Traite de
Zoologie, ed. P. Grasse, T. XI. Masson, Paris.
D'AGOSTINO, A. S., AND A. FARMANFARMAIAN, 1960. Transport of nutrients in the holo-
thurian Leptosynapta inhacrens. BioL Bull., 119: 301.
ENRIQUES, P., 1902. Digestione, circolazione e assorbimento nelle oloturie. Arch. Zool. ItaL,
1 : 1-58.
FARMANFARMAIAN, A., 1963. Transport of nutrients in echinoderms. Proc. XVI. Int. Cou</.
Zool., 1: 118.
FARMANFARMAIAN, A., AND J. H. PHILLIPS, 1962. Digestion, storage and translocation of
nutrients in the purple sea urchin Strongylocentrotus purpuratus. BioL Bull., 123:
105-120.
FISH, J. D., 1967. The biology of Cucumaria clongata. (Echinodermata : Holothuroidea. )
/. Mar. BioL Assoc., 47: 129-143.
FUJI, A., 1961. Studies on the biology of the sea urchin, IV. Histological observation of the
food canal of Strongylocentrotus intcnncdiits. Bull. Fac. Fisheries, Hokkaido Univ.,
11: 195-202.
GIESE, A. C., 1966. Lipids in the economy of marine invertebrates. Plivsiol. Rerietvs, 46:
244-298.
GURR, E., 1960. Encyclopaedia of Microscopic Stains. Leonard Hill, London.
DIGESTIVE SYSTEM OF CUCUMARIA. I
HAMANN, O., 1883. Beitrage zur Histologie der Echinodermen. Die Holothurien. Zcitschr.
Ti'i'jj. Zoo/., 39: 145-190.
HEATH, I. D., 1962. Observations on a highly specific method for the histochemical detection
of sulphated mucopolysacharides, and its possible mechanisms. Quart. J. Micr. Sci.,
103: 457-475.
HOLLAND, N. D., ]. H. PHILLIPS AND A. C. GIESE, 1965. An autoradiographic investigation of
coelomocyte production in the purple sea urchin (Strongylocentrotus pnrpuratns).
Biol. Bull, 128: 259-270.
HYMAN, L. H., 1955. The Invertebrates : Echinodermata. The Coelomate Bilateria. Mc-
Graw-Hill Book Co., New York.
KAWAMOTO, N., 1927. Anatomy of Caitdina cliilcnsis. Tolwkii Univ. Sci. Refits. Scr. 4 Biol.,
2: 239-265.
KARNOVSKY, M. L., S. S. JEFFRY, M. S. THOMPSON AND H. W. DEANE, 1955. A chemical
and histochemical study of the lipids of the pyloric caecum of the starfish Asterins
forbcsi. J. Biophysic. Biochem. Cytol., 1: 173-182.
MILLOT, N., 1966. A possible function for the axial organ of echinoids. Nature, 209: 594-596.
MILLOT, N., AND H. G. VEVERS, 1964. Axial organ and fluid circulation in echinoids. Nature,
204: 1216-1217.
NICHOLS, D., 1959. The histology of the tube feet and clavulae of Echinocardinin cordatnin.
Quart. J. Micr. Sci., 100: 73-87.
OOMEN, H. A. P. C., 1926. Verdauungsphysiologische Studien an Holothurien. Puhhl. Staz.
Zoo/. Nafoli, 7: 215-297.
PEARSE, A. G. E., 1960. Histochemistry. Theoretical and Applied. Churchill, London.
PROSSER, C. L., AND C. L. JUDSON, 1952. Pharmacology of the haemal vessels of Stichopus.
Biol. Bull., 102: 249-251.
RUNDLES, C., AND A. FARMANFARMAiAN, 1964. Absorption and transport of D-glucose in the
intestine of Tliyoue briareus. Biol. Bull.. 127: 387-388.
SCHREIBER, B., 1930. Studi sull' assorbimento intestinale nelle oloturie. Pubbl. Staz. Zoo!.
Napoli, 10: 235-277.
SCHREIBER, B., 1932a. Pigmenti e secrezioni nel sistema digerente nelle oloturie. Pnhhl. Staz.
Zoo!. Nafio/i, 12: 18-60.
SCHREIBER, B., 1932b. Experiment] per lo studio dell' assorbimento intestinale nelle oloturie.
Arch. Zoo/. Ital., 16: 865-870.
SIVICKIS, P., AND J. DOMANTAY, 1928. The morphology of a holothurian, Sticlwpits chloro-
notus Brandt. Phillifiiiie J. Sci.. 37: 229-332.
STOTT, F. C., 1955. The food canal of the sea urchin Echinus escnlentits L. and its functions.
Proc. Zoo/. Soc. London. 125: 63-85.
STOTT, F. C., 1957. Observations on the food canal and associated structures in the holothurian
Holotluiria forskuli Delle Chiaje. Proc. Zoo/. Soc. London, 129: 129-136.
TANAKA, Y., 1958. Feeding and digestive processes of Stichopus japonicus. Bull. Fac.
Fisheries, Hokkaido Univ., 9: 14-28.
THE DIGESTIVE SYSTEM OF THE HOLOTHURIAN, CUCUMARIA
ELONGATA. II. DISTRIBUTION OF THE DIGESTIVE ENZYMES
JOHN D. FISHi
Dove Marine Laboratory, University of Newcastle Upon Tync, Enyland
Several enzymes have been recorded from holothurian digestive tracts and in
extracts of the gut wall. Oomen (1926) demonstrated a protease, amylase, maltase
and a weak lipase in Holothnria, and Van der Heyde (1922) reported protease,
invertase and lipase, but no amylase in Thyonc. The digestive fluid and extracts
of the gut wall of Caiidina chilensis were found by Sawano (1928) to contain lipase,
maltase, invertase, glycogenase and a protease resembling trypsin. Choe (1962)
detected amylase, cellulase, pectinase and dipeptidase in gut extracts of both the
green and red varieties of Stichopus japonicus. He also demonstrated small
amounts of lipolytic enzyme capable of digesting simple ester, glyceride and
higher fatty acid. It is now generally accepted that holothurians have abundant
carbohydrate-splitting enzymes and a proteolytic enzyme similar to trypsin. Al-
though Choe (1962) has given details of the amount of enzyme activity recorded
from the different parts of the intestine of SticJiofms japonicus, there is little infor-
mation available concerning the distribution of digestive enzymes in holothurians.
In the present study, extracts of the different gut regions of Ciicnniaria were tested
for proteases, carbohydrases and lipases, so as to determine the distribution of
enzymes as well as the enzyme complement.
It is interesting to note that Pequignat ( 1966) has recently demonstrated "skin
digestion" in echinoids. He has shown that the mucous coating of the body surface
is capable of digesting a wide range of food materials. He observed spherule coelo-
mocytes "creeping" out to the external mucous coating where they eventually dis-
integrated "while probably releasing digestive enzymes." Although Pequignat has
not studied holothurians, he claims that it is unlikely that they should behave any
differently. However, in terms of overall nutritional requirements it is unlikely
that "skin digestion" can be of serious significance to animals with well developed
digestive tracts.
I am grateful to Dr. J. B. Buchanan for advice and encouragement. The work
was supported by a research studentship from the Department of Scientific and
Industrial Research.
MATERIALS AND METHODS
Specimens of Cuciiinaria were collected off the Northumberland coast during
July and August, 1966, and the enzyme extract was prepared on the day of capture.
1 Present address : Department of Zoology, University College of Wales, Aberystwyth,
Wales.
354
DIGESTIVE SYSTEM OF CUCUMARIA. II
Preparation of enzyme extract
Extracts were prepared of the esophagus, stomach, constriction, intestine I and
intestine II. Details of gut nomenclature are given by Fish (1967).
The different regions of the gut were carefully dissected from a number of
animals and placed in separate containers. After removal of the gut contents the
material was quickly rinsed and then blotted dry. The gut material was weighed
and added to twice its own weight of glycerol and homogenized for 10 minutes in
an M.S.E. homogenizer. The volume obtained was diluted with an equal volume
of filtered sea water and centrifuged for 15 minutes. The supernatant was col-
lected and sea water of an equal volume to this supernatant was added to tJic residue
which, after mixing, was centrifuged for a further 15 minutes. The second super-
natant was added to the first to give the final enzyme extract which was filtered
through a Whatman No. 4 paper. Toluene was added to prevent putrefaction.
Incubation with substrate solutions was started on the day following preparation,
the extracts being stored overnight at 4° C.
Estimation of proteases
Proteolytic enzymes were estimated using the formol titration method of
So'rensen, as described by Davis and Smith (1955). For speed and convenience
titration by indicator was preferred to the potentiometric titration recommended
by Dunn and Loshakoff (1936). One-tenth per cent phenolphthalein in absolute
alcohol was used as the indicator, and the enzyme-substrate mixture was titrated
against approximately 0.3 N NaOH. For all titrations an "Alga" micrometer
syringe was used instead of a burette. The volume of alkali delivered was accurate
to 0.001 ml. The results are expressed as the amount of hydrolysis per hour per
ml. of enzyme extract by using arbitrary units, i.e., 0.01 ml. of 0.3 N NaOH - : 10
units of hydrolysis.
The enzymes studied, together with the respective substrate solutions, were as
follows :
Enzyme system Substrate (1% aqueous solution)
Trypsin a Benzoyl-L-arginine, ethyl ester
Aminotripeptidase Triglycine
Glycylglycine dipeptidase Glycylglycine
Leucine aminopeptidase L-Leucyl-glycylglycine
Carboxypolypeptidase Chloracetyl-L-tyrosine
Estimation of carboli\drases
Amylase, invertase and maltase were estimated by using starch, sucrose and
maltose solutions as the respective substrates.
One nil. of gut extract plus 2 ml. of substrate solution were incubated for 12
hours at 20° C. Quantitative estimations were carried out by volumetric estimation
of the cuprous oxide which was formed on reduction of a cupric salt by the products
of enzyme hydrolysis. The method employed was that of Bertram!, described by
Plimmer (1920) with the following modifications.
Instead of boiling the reagent-sugar mixture over a bunsen flame for three
minutes the mixture was heated in a boiling water bath for 15 minutes. It was
356 JOHN D. FISH
cooled and centrifuged at 3500 r.p.m. for 15 minutes, and the supernatant carefully
decanted, leaving a deposit of cuprous oxide. Centrifugation is quicker and less
tedious than removing the precipitate by nitration through a special asbestos filter
as described in the original method. It was found necessary to avoid transference
of the solution from one tube to another because the cuprous oxide becomes
adsorbed onto the walls of the tube as it is being precipitated. Transferance of the
solution was found to result in a considerable loss of cuprous oxide. Consequently,
the whole procedure, from incubation to titration, was carried out in the same tube.
As a control experiment 2 ml. of substrate solution were incubated without gut
extract and treated as described above, so as to account for any hydrolysis of the
substrate not due to enzyme action.
The modified Bertram! method is satisfactory when dealing with the quantitative
estimation of amylase, and invertase, as the substrate solutions used are not reducing
sugars. However, as the method depends upon the reduction of alkaline cupric
sulfate by the products of enzyme hydrolysis — reducing sugars — it is useless when
testing for maltase, as the substrate used, maltose, is itself a reducing sugar. To
overcome this, 10 ml. of cupric acetate solution (Barfoed's reagent) were used in
place of alkaline cupric sulfate. Glucose, the product of enzyme hydrolysis,
reduces Barfoed's reagent, giving a precipitate of red cuprous oxide, whereas it is
unaffected by maltose.
Titration values were converted to mg. of glucose liberated by enzyme hydrolysis
by reference to calibration curves obtained with glucose under conditions stated
above. Although invertase on hydrolysis yields molecules of fructose as well as
molecules of glucose, the reducing power of fructose is so similar to that of glucose
(Plimmer, 1920), that for the purpose of constructing the calibration curves, the
products of enzyme hydrolysis were regarded as molecules of glucose. Results are
expressed as mg. of glucose liberated per hour per ml. of enzyme extract.
Qualitative estimation of carbohydrases
The extract used in the qualitative estimation of carbohydrases was prepared
by the technique described above, using the complete gut from several animals. In
each test 1 ml. of extract was incubated at 20° C. for 12 hours with 2 ml. of the
respective substrate solution. Tests were made for amylase, invertase, maltase,
lactase, glycogenase and cellulase.
Supplies of Laminarin and Fucoidin, substrates for cellulase activity, were
obtained in the soluble powder form from the Seaweed Research Institute, Inveresk,
Midlothian, Scotland.
Estimation of esterase and lipase
Esterase and lipase were estimated quantitatively by using the method of
Nachlas and Seligman (1949). Beta-naphthyl esters were used as substrates.
Broadly speaking, the esters of short-chain fatty acids (C2-C4) are split by
esterases, and the long-chain esters (Cs-upwards) by lipases, yet according to
Nachlas and Seligman (1949), there is a considerable degree of overlapping in
enzymatic hydrolysis by these two enzymes. Enzymes splitting the substrate beta-
naphthyl acetate (C2) are here regarded as esterases, those splitting beta-naphthyl
DIGESTIVE SYSTEM OF CUCUMARIA. II
357
Units Of
hydrolysis
10 -
5 -
Aminotripeptidase
Aminopept idase
I I I I I
I I I I I
Carbo xy -
polypept idase
I I I I I
Mg.
glucose
0-5
0-5 •
I I
Amy lase
I I I I I
Invert ase
FIGURE 1. Enzyme activity recorded from the different regions of the gut. A, proteases; B, car
bohydrases ; O, esophagus; S, stomach; C, constriction; I, intestine I; II intestine II.
358
JOHN D. FISH
TABLE I
Activity of f>n>teolytic enzymes from the different gut regions
Amino-
tripeptidase
Amino-
peptidase
1 Mpcptidase
Carboxy-
polypeptidase
Trypsin
Esophagus
2.2 ± 0.9
1.4 ± 0.4
7.4 ± 1.4
1.2 ± 0.5
1.3 ± 0.2
Stomach
2.6 ± 1.4
1.4 ± 1.0
6.6 ± 1.7
1.4 ± 0.4
1.2 ± 0.4
Constriction
3.7 ± 1.4
2.4 ± 1.1
8.0 ± 0.7
1.8 ± 0.8
1.6 ± 1.0
Intestine I
8.9 ± 2.1
3.8 ± 1.3
12.2 ± 2.1
2.0 ± 0.7
3.0 ± 1.3
Intestine II
9.8 ± 4.5
2.5 ± 0.3
14.0 ± 2.5
1.1 ± 0.6
2.2 ± 0.8
Activity expressed as units of hydrolysis/ml, of extract/hr.
standard deviation of 5 determinations.
Each value is the mean and
laurate (C]2) are described as "esterase-lipase" and enzymes splitting beta-naphthyl
stearate (C1S) as lipases.
Beta-naphthol is liberated by enzymatic hydrolysis and an azo dye is produced
by coupling the free naphthol with a tetrazoninm salt. The colored compound was
extracted with ethyl acetate and measured colorimetrically. All measurements were
made using the Unicam Spectrophotometer S.P.600 at a wave-length of 540 m/x.
The colorimeter readings were converted to mg. of beta-naphthol by reference to a
calibration curve obtained using known quantities of beta-naphthol. The results
are expressed as mg. beta-naphthol liberated per hour per ml. of enzyme extract,
RESULTS AND CONCLUSIONS
Results are featured in Tables I-IV, and Figure 1, and indicate the presence
of a variety of digestive enzymes in gut extracts of Cuciunaria. An endopeptidase
of a trypsin-like nature and several exopeptidases are present, and although these
were detected in extracts of all gut regions, maximum activity was without exception
recorded from the intestine. Maltase, amylase, and invertase were readily detected,
each having maximum activity in the constriction, and glycogenase has been de-
tected in extracts of the whole gut. Lactase and cellulase have not been detected
in Cuciunaria. The distribution of amylase is interesting, in that of the enzymes
TABLE II
Results of the qualitative estimation of carbohydrases
Enzyme
Substrate
Reagent employed
Result
Amylase
1% Starch soln.
Fehling's soln.
+ +
1 nvertase
5% Sucrose soln.
Fehling's soln.
+ +
Maltase
5% Maltose soln.
Barfoed's reagent
+ +
Lactase
2% Lactose soln.
Barfoed's reagent
—
Glycogenase
Saturated soln. of
Fehling's soln.
+
glycogen
Cellulase
1% Laminarin soln.
1% Fucoidin soln.
Fehling's soln.
—
1% Sodium alginate
+ + = Strongly positive; + = positive; -- = negative.
DIGESTIVE SYSTEM OF CUCUMARIA. II
359
TABLE 1 1 1
Activity of carbohydrases from the different gut regions
Maltase
Invertase
Amylase
Esophagus
Stomach
0.51 ± 0.2
0.58 ± 0.2
0.20 ± 0.3
0.17 ± 0.2
0
0
Constriction
1.38 ± 0.3
0.34 ± 0.2
0.49 ± 0.2
Intestine I
1.07 ± 0.2
0.25 ± 0.3
0.26 ± 0.2
Intestine II
0.66 ± 0.2
0.16 ± 0.1
0
Activity expressed as mg. reducing sugar/ml, of extract/hour,
standard deviation of 5 determinations.
Each value is the mean and
studied, it is the only one \vhich is not found throughout the gut. Choe (1962)
detected amylase in the first and second small intestines and the anterior and
posterior parts of the large intestine of Stichopus faponicus.
A strong esterase activity and a weaker "esterase-lipase" have been detected
with practically uniform distribution throughout the gut. The ability of the extracts
to hydrolyze beta-naphthyl stearate was so poor that the results are not given. It
is unlikely that the amount of enzyme activity recorded is within the limits of
accuracy of the method. Oomen (1926), using amyl-acetate and ethyl butyrate
esters as substrates, and Sawano (1928) using olive oil, both recorded a weak lipase
in Holothuria and Caudina, respectively. It is unfortunate that there is a consider-
able degree of overlapping in enzymatic hydrolysis by esterases and lipases even
when using purified beta-naphthyl esters as substrates. However, it can be con-
cluded that gut extracts of Citcitiimria hydrolyze short-chain fatty acids (C2), and
intermediate-chain fatty acids (C12), yet it is doubtful whether they can hydrolyze
long-chain fatty acids (C18 and upwards).
DISCUSSION
Enriques ( 1902 ) suggested that digestive enzymes were carried by amoebocytes
from the hemal system into the digestive tract. Oomen (1926) and Schreiber
(1930, 1932a, 1932b) found that extracts of the hemal wall contained a protease,
invertase, amylase and maltase, yet during his experiments Oomen found that
extracts of the stomach wall contained more of these enzvmes than did the hemal
TABLE IV
Activity of lipolytic enzymes from the different gut regions
Esterase
"Esterase-lipase"
Esophagus
4.02 ± 0.4
0.26 ± 0.2
Stomach
3.03 ± 0.5
0.17 ± 0.2
Constriction
3.23 ± 0.4
0.13 ± 0.2
1 ntestine I
3.78 ± 0.4
0.28 ± 0.3
Intestine II
3.29 ± 0.4
0.16 ± 0.2
Activity expressed as mg. beta-naphthol/ml. of extract/hour. Each value is the mean and
standard deviation of 5 determinations.
360 JOHN D. FISH
extract or the digestive fluid. However, the presence of digestive enzymes in an
extract of the hemal wall was accepted as more or less verification of the hypothesis
of Fnriques. According to Fren/.el (1892, vide Oomen, 1926), the walls of the
rete mirahile are glandular, and their secretion is taken up by the amoebocytes and
transported via the hemal system to the gut. The amoebocytes pass through the gut
wall and between the epithelial cells into the gut lumen where they burst to release
their contents. Although Hamann (18X3 ) demonstrated gland cells in the stomachs
of Holothuria and Leptosynapta, there are no records relating to the possible secre-
tion of digestive enzymes by cells surrounding the lumen of the gut. The work of
earlier authors appears to have been accepted without confirmation by detailed
histochemical and physiological studies.
Studies on the digestive enzyme systems in Citcninaria have shown that amylase
is present only in extracts of the constriction and intestine I, with a peak density
in the constriction. Maltase and invertase are present in all gut regions, but there
is always an obvious density peak in the constriction. Proteolytic enzymes show
their greatest activity in the intestine. In Cncitniaria there is no rete mirabile, and
all parts of the hemal system have the same histological structure (Fish, 1967).
If the supply of digestive enzymes is dependent upon the entry into the gut of
loaded amoebocytes, then this would seem to suggest two hypotheses as regards
the sites of enzyme secretion. Either the enzymes are secreted solely from the walls
of the transverse branches, which are the larger channels of the system, or they are
secreted in all parts of the hemal system. If the first hypothesis is valid and
amoebocytes carry digestive enzymes from the transverse branches of the hemal
system into the gut, then to account for the results given above (with particular
reference to amylase), there must be some mechanism which ensures that amoebo-
cytes carrying amylase pass only into the constriction and intestine I. If the second
hypothesis is true and amoebocytes carry digestive enzymes from any of the hemal
channels, then to account for the distribution of amylase, the dorsal and/or ventral
hemal sinuses must produce enzymes needed by the particular part of the gut to
which the sinus is attached. If this is the case, then the amoebocytes need only
pass through the adjacent gut wall. It is suggested that neither of these hypotheses
is tenable. The distribution of digestive enzymes can be correlated with the results
of histological and histochemical tests applied to the different gut regions of Citcn-
inaria (Fish, 1967). Secretory granules and gland cells have been demonstrated in
the constriction and the intestine, respectively, and it is from extracts of these
regions that the highest degrees of enzyme activity have been recorded. These
results and observations lead to the conclusion that cells of the lining epithelium of
the gut secrete digestive enzymes. This does not preclude the possibility that
amoebocytes carry enzymes. These enzymes may play some part in the nutrition
of the whole animal, or their presence may be attributed to the metabolic require-
ments of the amoebocytes themselves. Useful information would be gained from a
detailed study of the enzyme histochemistry of the holothurian digestive tract.
SUMMARY
1. Estimations of proteases, carbohydrases, and lipases have been made on
extracts of the different gut regions of Cucumaria elonyata.
DIGESTIVE SYSTEM OF CUCUMARIA. 11 361
2. An endopeptidase of a trypsin-like nature and several exopeptidases have
been detected, all with maximum activity in the intestine. Cellulase and lactase
have not been detected, yet maltase and invertase were found in all regions of the
gut. and showed their maximum activity in the constriction and intestine I.
Amylase was found only in the constriction and intestine I. A strong esterase and
a weaker "esterase-lipase" have been detected with practically uniform distribution
throughout the gut, yet it is doubtful whether a true lipase, hydrolyzing long-chain
fatty acids (C1S and upwards), is present.
3. The work of earlier authors dealing with digestive enzymes and the sites of
enzyme production in holothurians has been summarized. Results of quantitative
estimations of digestive enzymes in Cucumaria, coupled with the results of previous
histological and histochemical studies, lead to the conclusion that digestive enzymes
are secreted by cells bordering the gut lumen.
LITERATURE CITED
ANDERSON, I. M., 1953. Structure and function in the pyloric caeca of Astcrias forbesi. Biol.
Bull., 105: 47-61.
CHOE, S., 1962. Biology of the Japanese common sea-cucumber, Stichopus japonicus Selenka.
Fisheries College, Pusan National University.
DAVIS, N. C, AXD E. L. SMITH, 1955. Assay of proteolytic enzymes. In: Methods of Bio-
chemical Analysis. Ed. D. Click. Vol. 2 : 215-257. Interscience Publishers, New York.
DUNN, M. S., AND A. LOSHAKOFF, 1936. Quantitative investigations of ammo-acids and pep-
tides. I. Quantitative fonnol titration by means of the glass electrode. /. Biol. Cliciu..
113: 359-369.
ENRIQUES, P., 1902. Digestione, circolazione, e assorbimento nelle oloturie. Arch. Zool. Ital.,
1: 1-58.
FISH, J. D., 1967. The digestive system of the holothurian, Citcitinariu cloiif/ata. I. Structure
of the gut and hemal system. Biol. Bull.. 132 : 337-353.
HAMANN, O., 1883. Beitrage zur Histologie der Echinodermen. Die Holothurien. Zcitschr.
n'iss. Zool, 39: 145-190.
NACHLAS, M. M., AND A. M. SELIGMAX, 1949. Evidence for the specificity of esterase and
lipase by the use of three chromogenic substrates. /. Biol. Chan., 181 : 343-355.
OOMEN, H. A. P. C., 1926. Verdauungsphysiologische Studien an Holothurien. Pitbbl. Stas.
Zoo!. Nal>oli,7: 215-297.
PEQUIGNAT, E., 1966. 'Skin digestion' and epidermal absorption in irregular and regular urchins
and their probable relation to the outflow of spherule coelomocytes. Nature. 210:
397-399.
PLIMMER, R. H. A., 1920. Practical Organic and Biochemistry (Rev. ed). Longmans, Green
Co., London.
SAWANO, E., 1928. On the digestive enzymes of Caiidina chilcnsis (J. Muller). Sci. Rep.
Tohoku Univ. (4) Biol., 3: 205-218.
SCHREIBER, B., 1930. Studi sull' assorbimento intestinale nelle oloturie. Pubbl. Stas. Zool.
Napoli, 10: 235-277.
SCHREIBER, B., 1932a. Pigmenti e secrezioni nel sistema digerente nelle oloturie. Pubbl. Staz.
Zool. Napoli, 12: 18-60.
SCHREIBER, B., 1932b. Experiment! per lo studio dell' assorbimento intestinale nelle oloturie.
Arch. Zool. Ital., 16: 865-870.
VAN DER HEYDE. H. C., 1922. On the physiology of digestion, respiration and excretion in
echinoderms. Dissertation, Amsterdam.
ON THE MORPHOLOGY OF THE NEPHRIDIA OF
NEREIS LIMNICOLA JOHNSON
MEREDITH L. JONES
Division of Worms, Museum nf Natural History, Smithsonian Institution,
n\,s!nn</t,m,D. C. 20560
In the past century there have been a number of reports of observations on the
nephridia of the polychaetous annelids. These have ranged from passing notes to
detailed morphological treatments. Some of these have been concerned with all
families of the polychaetes or with general comments (Benham, 1891 ; Ehlers. 1864-
68; and Goodrich, 1895, 1945) ; some dealt only with the nephridia of the so-called
errant forms (Aiyar, 1933; Page, 1906; and Goodrich. 1897, 1898 and 1900) ; and
most have considered the nephridia of the so-called sedentary worms.
Relatively little information has been published on the detailed morphology of
the nephridia of the Nereidae. Goodrich (1893) described the nephridia of Nereis
dh'ersicolor and found that the nephridial canal could be divided into four regions
which differed in distribution of cilia, diameter of the lumen and extent of tubule
convolution. Page (1906) worked on Perincrcis cultrijera, observing nephridial
histology and reporting on the uptake of neutral red by nephridia in living animals.
Krishnan (1952) described the nephridial morphology of three nereid species with
contrasting salinity tolerances: Namalycastis indica (Southern) (euryhaline) ;
Nereis chilkacnsis Southern (relatively stenohaline, in slightly brackish environ-
ments) ; and Perinereis innitia (Savigny) (stenohaline, in fully marine situations).
Krishnan not only described the nephridia of the three genera, but also related the
size of the nephridia and the amount of nephridial vascularization to the ability of
the polychaetes to tolerate lowered salinities. He concluded that the large size of
the nephridia of N. indica, as well as their rich supply of blood vessels, represents
an adaptation for a euryhaline existence. He suggested that there is a direct
excretion of water from blood vessels to the lumen of the nephridial canal and
showed that there is a shrinking and collapse of nephridial blood vessels in specimens
of N. indica which had been acclimatized to full-strength sea water. Finally, the
nephridial morphology of Nereis vc.villosa Grube has been described (Jones, 1957 )
and it was noted that the nephridial canal of this species is ciliated throughout and
that there are three general regions along the length of the canal, based on the
lumen diameter 1 and the amount of convolution. In addition, a reconstruction of
the nephridial canal was presented.
Nereis linmicola Johnson has recently been used as an experimental animal by
a number of workers. The species was originally described (Johnson, 1903) from
Lake Merced, a fresh-water lake which has served as a water supply for the city of
1 It should be noted that in Figure 8, Jones, 1957, the scale line of the diagram of the
nephridial canal of Nereis I'c.rillosu should read 100 micra, not 50; further, Figure 6 is reversed,
left for right.
362
NEI'HKIDIA OF NEREIS LIMNICOLA 363
San Francisco, California (for a short account of the history of the lake, see Smith,
1958, p. 61). Subsequently, there were no published reports of the species until
Smith (1958) re-collected material from Lake Merced for physiological observa-
tions.
Hartman (1938) described and recorded (1944) X emit lies In/hti from Marin
and Sonoma Counties, north of San Francisco Bay. Smith ( 1950) described
embryonic development in specimens of this species from the Salinas River, south
of San Francisco, near Monterey, and showed that it is a viviparous self-fertilizing
hermaphrodite. Later, Smith (1953) studied the distribution of the species along
the Salinas River and reported observations on the salinity cycle of the river over a
three-year period and the effect of salinity changes on the distribution of the
polychaete.
After his re-collection of Nereis limnicola, Smith (1959b) compared the type
specimens of Nereis limnicola Johnson (1903) with specimens of Neanthes light!
Hartman (1938) and concluded that Neanthes I'njhti is a junior synonym of Nereis
linniicola. Neantlies I'njhti was referred to Nereis japonica Izuka (1908) by Edith
and Cyril Berkeley (1956, p. 269), who pointed out the close morphological simi-
larities between Nereis japonica and Nereis dh'ersicolor O. F. Miiller. Smith
(1958) compared, in considerable detail, specimens of Nereis limnicola from Cali-
fornia, Washington, and British Columbia, with specimens of Nereis japonica from
[apan, and specimens of Nereis dh'ersicolor from Scotland, England, Denmark,
Finland, France, and New Hampshire. Smith presented strong arguments for the
separation of these three species, which are reproductively and geographically
isolated. Their close morphological and ecological similarities are emphasized by
Pettibone (1963, pp. 160-161) who referred all three species to Nereis (Hedistc).
Hartman (1960) referred N. limnicola, N. light! and N. japonica to Neanthes
dh'ersicolor and Imajima and Hartman (1964) referred Nereis japonica Izuka to
Neanthes dh'ersicolor. However, I prefer to follow Smith (1958, 1959b),
Khlebovich (1963), and Pettibone (1963) in considering the three species, N.
dh'ersicolor O. F. Miiller, N. japonica Izuku, and N. limnicola Johnson, as distinct,
but closely related, species of Nereis (Hediste).
Nereis limnicola has been utilized by Smith for physiological studies (1957,
1959a), who found that the species can control the influx of pond water, distilled
water, and extreme dilutions of sea water at 13° C. but has no control at tem-
peratures of l°-2° C. Later, Smith (1963), in comparing N. dh'ersicolor, N.
limnicola, and Nereis (NeantJies) siiccinea, found that N. limnicola had the lowest
salt loss rate of the three species when placed in lowered salinities, but (Smith,
1964) that both N. limnicola and Ar. siiccinea have an equal D.O influx at a body
weight of about 100 mg., even though Ar. limnicola takes up less water when both
species are subjected to an equal external osmotic gradient.
Stephens (1964) made observations of the uptake of glycine by N. linniicola
and N. siiccinea. He found that the uptake by the latter is greater by an order of
magnitude than in the former and suggested that the uptake takes place across the
body wall. Stephens further suggested that glycine uptake and osmoregulation are
incompatible, since glycine uptake becomes less, and even ceases, when the salinity
of the medium is lowered into the range wherein the worms are hyper-regulating.
Oglesby (1965a), in comparing water and chloride regulation in N. limnicola,
364 MEREDITH L. JONES
N. sitccinea. N. I'e.villusa. and Laconercis cithcri (Webster), reported that N.
liiiDiico/a shows the best ability to ululate osmotic concentration and exhibits the
least change in water content of the entire body with varying salinities. Further,
Ogleshy (1965b) has shown that the chloride exchange rate is lowest in N. lintnicola
and suggested that this may be due to a low chloride permeability, the worms be-
coming essentially impermeable to chloride in fresh water.
In his paper treating of viviparity in Nereis liinnieola. Smith ( 1950 ) dealt with
worms inhabiting the lower Salinas River, Monterey County, in central Cali-
fornia. The Salinas River presents a difficult situation for aquatic forms. It is not
a large river and appears to serve mainly as a run-off channel for the fall, winter,
and spring rains from its watershed. After the spring run-off, a sandbar is formed
across the mouth of the river, forming a "blind" estuary of the river, as defined by
Day (1951). This serves to dam the river flow until the following winter when
the press of run-off may be sufficient to break through to the sea. The latter is not
necessarily an annual event and the river mouth mav remain blocked for several
years.
Smith's studies (1950) were concerned with the worms found in two areas.
One ("Area A" on Smith's map, p. 425) was a "muddy channel in a SaUcornia
marsh near the mouth of the Salinas River," and the other (Area B) was "a sandy
stretch of river about four miles upstream, where the general aspect is that of fresh
water." Smith further gave the range of salinities for these areas as 20% to 115%
sea water for Area A and 1% to 3% sea water for Area B, the salinity of a given
area being a function of rainfall and season. He pointed out that the effect of the
more extreme low salinities was possibly damped by the residual salt in the soil of
the surrounding substrate. Smith (1953) later provided more detailed ecological
observations on the lower Salinas River and referred to seven numbered locations
(location 3 -- Area A, above; location 4 — Area B).
During the rainy season and the period of run-off, the SaUcornia marshes of the
lower Salinas River may be flooded with fresh water. Within a few days the
salinity of the overlying water is decreased immensely. This dilution prevails for
a variable period until the last of the seasonal excess is sluiced into the sea. At
this time, as the fresh wrater recedes, the marshes may be inundated with sea water
at spring tides. As the flow of the river decreases, the channels in the marsh
become isolated from the main river. Through the summer, isolated ponds are
subjected to the evaporative effect of the sun, although they are relieved by sporadic
rains and heavy fogs. During this time, the salinity of the overlying water reaches
its annual high that is maintained until the fall rains, when the rising river inundates
the isolated areas and dilutes them to their annual low. In the descriptions and
discussion to follow, the worms of Area A will be referred to as the down-river
population.
Throughout the year there is near-fresh water over Area B, several miles
up-river from Area A. This condition is relieved only during times of extremely
high tides when the river flow is slackening, following the winter rains, at which
times incursions of salt water may reach Area B. The worms found in this area
are probably near the extreme fresh-water end of their range, for Smith (1953,
location 5), has not found them more than | mile up-stream from the collection site
in Area B. The worms of Area B will lie referred to as the up-river population.
NEPHRIDIA OF NEREIS LIMNICOLA 365
Considering the brief ecological description above, several questions become
apparent. Is it reasonable to expect that groups within the same species, differing
only in the salinity of their environments, will exhibit morphological differences?
Or, because of a physiological adaptability, will these groups show no significant
anatomical differences? If there are differences, will they lie manifest in the
nephridium ? It was in an attempt to answer these questions that this morphological
comparison of the nephridia of Nereis liumicola from environments of different
salinity was undertaken.
THE NEPHRIDIUM OF NEREIS LIMNICOLA (Ui'-RivER FORM)
The worms used in the following section were collected in Area B (Smith,
1950). The salinities of the water flowing over them were less than 2.5% sea
water. The worms were relaxed by the use of dilute alcohol, fixed in Benin's,
serially sectioned in paraffin at 6 micra, stained in Harris' hematoxylin, and counter-
stained in eosin.
\Yithin the same animal, indeed, within the same segment, there may be a wide
variation in the overall size of the nephridia. In one case, a pair of nephridia in the
same segment were observed in which the sizes differed in the order of 1:2. These
approximated the extreme differences between pairs of nephridia, as w*ell as between
unpaired nephridia. They were both approximately the same width and length
(250 micra), but differed', however, in that one was about 240 micra in height
while the other was about 400 micra (for lengths and numbers of segments of the
worms examined, see Table II). The approximate volume of the former was
0.0106 mm.3 and the latter, 0.0140 mm.3
If one views the nephridium of Nereis liumicola from Area B, with the purpose
of comparing it to that of Nereis rc.villosa (Jones, 1957), one is immediately im-
pressed by the vast number of blood vessels in contact with, and buried in, the
tissue of the nephridium (Figs. 1. 2, and 3, BV). The shape of the nephridium
also contrasts with that of N. t'c.rillosa. \Yhereas the nephridium of N. vexillosa
is globular and possesses a smooth, even surface, that of N. liumicola shows a sug-
gestion of a division into regions. The dorsal half of the nephridium is oval in
cross-section, and is somewhat compressed antero-posteriorly. The ventral half is
nearly circular in cross-section, and is more or less hemispherical. At the equator
of the hemisphere, the post-septal canal enters the nephridial mass in company with
the ventral segmental vessel, that ramifies over the surface of the nephridium (Fig.
1 , PS ) . At the point of entry of the post-septal canal there is a slight swelling.
The surface of the nephridium shows a slight indication of the internal canal in the
more vascularized portion, while the other half of the nephridium externally shows
a well-defined canal.
The sectioned nephridium of the up-river form of Nereis liumicola shows much
the same aspect as Nereis t'e.rillosa (Figs. 1, 2, and 3), and a number of nuclei are
scattered throughout the sectioned area. Further, there are occasional areas of
vacuolation, but not to the extent of those observed in N. ve.villosa. On the whole,
the perforations or sections of tubule lumen observed in the sections of the nephridia
of the up-river form of N. liumicola presented the same appearance as those of N.
vexillosa.
366
MEREDITH L. JONES
'
\
\
-50
w
-50n 1
NEPHRIDIA OF NEREIS LIMNICOLA 367
The ciliation of the nephridial canal (Figs. 2 and 3. CL). as noted in this form,
did not seem to differ significantly from the pattern seen in A'. I'e.rillosa. Cilia were
noted throughout the length of the canal, from the nephrostome to within 40 to 50
micra of the nephridiopore. As hefore, there seemed to he no distinct division in
the nephridial canal on the hasis of its ciliation.
As mentioned above, the walls of the tuhules are only occasionally distinct.
\\hen present, they consist of vacuolated areas around the periphery of the perfora-
tion. They give the appearance of a clear ring around the lumen, and may have
some intradivision in the form of faint, thin walls. In the "non-walled" perforation,
the fine network of the interstitial tissue comes up to the canal boundary, and no
basement membrane is visible. In all probability, there actually is or was a wall
present, but staining and/or fixation techniques may not have been adequate to
bring it out. A variation of this last type of wall occurs when the area immediately
surrounding the perforation appears to be more heavily stained than the adjacent
interstitial tissue (Figs. 2 and 3, C\Y). By careful examination, it is seen that
this darkening is due to the presence of a more concentrated net system and many
granular inclusions.
In the nephridium of Xereis rc.villosa it was noted that blood vessels were at a
minimum, approaching and possibly contacting the nephridial system at only two
points. In Xereis liinnicola from the up-river area, it is readily seen that the
nephridium is penetrated throughout by many vessels. In the main, they are con-
fined to the more peripheral areas, but many branches pass through the center of
the mass (Figs. 1, 2 and 3, BV). The ventral segmental vessel, after it approaches
the nephridial mass in company with the post-septal canal, ramifies over the lateral
face of the nephridium and at several points passes dorsally into the interior. The
ventral portion of the nephridium has no internal blood vessels, while the dorsal
half contains more vessels than it carries on its surface. Occasionally, there are
blood vessels on the surface of the nephridial mass which seem to have sunken into
the tissue. They are not surrounded by nephridial tissue, but are in close contact
with it over about 180° to 200° of their circumference in section.
In contrast to the long post-septal canal of Xereis vcxillosa, this structure in the
up-river form of Xereis liinnicola is extremely short (Fig. 1, PS. and Fig. 9A ) .
the length of the former being 250 micra. and that of the latter about 175 micra.
Key to lettering: BV, blood vessel; CL, cilia; CP, cytoplasmic processes of nephrostome;
CM, mass of cilia; CW, nephridial canal wall; XB, band of nuclei of nephrostome ( = septal
band); NC, nephridial canal; NS, nephrostome; PS, post-septal canal; VS, ventral segmental
vessel.
Figures 1-4, up-river form of Nereis lininicolu ; Figures 5-8, down-river form of N.
limnicola.
FIGURE 1. Dorsal view of right nephridium and associated nephrostome; specimen RB.
FIGURE 2. View of nephridial tissue ; specimen S-2.
FIGURE 3. Detailed view of nephridial tissue and associated blood vessels ; specimen S-2.
FIGURE 4. Nephrostome ; specimen RB.
FIGURE 5. Dorsal view of left nephridium ; specimen SB.
FIGURE 6. View of nephridial tissue at junction of medial (left) and lateral (right)
regions ; specimen S-3.
FIGURE 7. Detailed view of nephridial tissue and associated blood vessels ; specimen S-3.
FIGURE 8. Nephrostome ; specimen S-3.
368 MEREDITH L. JONKS
Further, the diameter of this portion of the nephridial canal is slightly larger in the
up-river form of N. I in mi cola. Proceeding through the post-septal canal from the
nephridial mass toward the nephrostome, the cross-sections of the isolated canal
show the same structure noted in N. rc.rillosa. The wall appears to he vacuolated,
with occasional larger nuclei. Through the proximal portion of the post-septal
canal, the nuclei are well-scattered along the proximal portion of the post-septal
canal, but become more concentrated toward the middle portion, where there are
about twelve visible in each section (6 micra thick). Throughout this part of the
canal, cilia are visible, distributed around the inner boundary of the wall. The
diameter of the lumen is about 15 micra, basally, near the nephridial mass, and
distally narrows to about 7 micra. These conditions prevail throughout the proxi-
mal 100 micra of the post-septal canal. \Yhere the lumen is at its narrowest,
immediately distal to the region just described, and at about the level of the passage
of the post-septal canal through the septum, the nuclei within the walls are quite
concentrated ; 28 to 30 are distributed fairly evenly around the lumen in a section.
This concentration of nuclei, in what might be called a "septal band" (Fig. 4, NB),
is not so great nor so extensive as in N. vexillosa ; a comparison shows a more
restricted area in N. linmicola.
B
^—
= n
100/v
FIGURE 9. Graphic representation of post-septal canals associated with nephridia of Nereis
liinnicola. A. Up-river form ; specimen RB. B. Down-river form ; specimen SB. N, nephro-
stome ; n, nephridial mass.
Immediately anterior to the narrowing of the canal and the nuclear concentra-
tion, the canal widens to form the nephrostome, its extreme width being 25 to 30
micra at its mouth. All along the walls of the funnel the nuclei are scattered evenly,
as they were at the beginning of the post-septal canal. As in N. vc.villosa. the
lumen of the nephrostome is almost choked by the tangle of cilia lining it (Fig. 4.
CM). Many fairly large club-like structures, the protoplasmic processes, occur at
the opening of the nephrostome (Fig. 4, CP ). In some of these are found the same
type of inclusion that appears in the interstitial tissue and tubule walls of the
nephridial mass. It is difficult to make out the exact structure of the processes, for
they stain weakly, and, at times, are intermeshed with the cilia that originate in
the walls of the nephrostome and the liases of the processes. From the tops and
sides of most of the processes, long cilia project into the open mouth of the funnel.
It is to be noted that the processes in this species are different from those found in
N. "i'c.villosa ; in N. vexillosa, they are long and thin, while in N. linniicola, they are
stout and nearly pyriform.
Figure 10 shows the diameter of the canal of the up-river form of Nereis
liinnicola from the entrance of the post-septal canal into the nepbridium, to the
nephridiopore (the canal chosen was the largest of tbe nephridia observed in detail).
NEPHRIDIA OF NEREIS LIMNICOLA
369
The canal has an overall length of 2232 micra and a mean diameter of 18.9 micra
(the length of the canal of the smallest nephridium was found to he 1800 micra
and its mean diameter was 19.6 micra). There appear to he four different regions
with respect to lumen diameter. The first, diameter about 24 to 30 micra, extends
from the entrance of the post-septal canal for about 800 micra ; the second, whose
diameter is in the range of 36 to 40 micra, extends for another 600 micra ; the third,
the narrowest part, about 12 to 27 micra, runs for approximately 200 micra before
grading into the final portion ; the last portion, about 750 micra in length, increases
from 27 to 48 micra, then decreases somewhat irregularly until it reaches the
nephridiopore, where its diameter is 6 micra. The sudden widening of the canal
just prior to the nephridiopore gives the appearance of an ampulla.
In the last portion of the nephridial canal, as the nephridiopore is approached,
the wall of the lumen seems to become thicker and more dense. Closer inspection
shows that the network of the interstitial tissue has become more concentrated in
the immediate area of the canal and that there seems to be an increase in the number
FIGURE 10. Graphic representation of the inner diameter and wall thickness of a nephridial
canal of Nereis liinuicola from up-river (reconstructed from sectioned material). PSC, post-
septal canal ; NPR, nephridiopore.
of granular inclusions contained in this net. There are large nuclei scattered
through the wall of the canal, until, at a point about 40 micra from the external
opening, a more regular distribution is assumed, with three or four nuclei apparently
in the same plane. This continues to the last 8 micra, where large nuclei are
clustered around the canal, and the wall loses its identity in the surrounding tissue.
As was the case in N. ve. \~illosa, there is no ciliation in the canal through the last
40 to 50 micra. Though an ampulla was not so obvious in N. vexillosa, N. limnicola
usually shows an ampulla (Fig. 10), or a suppression of one, just interior to the
nephridiopore. In cases of suppression, the lumen as seen in section is tripartite,
with the walls pressed together until the lumen cross-sectional area is at a minimum.
Whether this is an artifact of fixation, a morphological anomaly, a sphincter-like
device for closing the canal, or an adaptation providing a greater surface-volume
ratio for more efficient resorption or excretion, is not clear, but, as this type of
structure was fairly common, the condition seems most probably related to
resorption-excretion or to canal closure.
370 MEREDITH L. JONES
THE Nici'ii Kimi'M OF NEREIS LIMNICOLA (DowN-RivER FORM)
The worms used as a basis for the following description were collected at Area
A (Smith, 1950). The salinity of the water standing over them was at least 47.5%
sea water. (It is necessary to point out that one of the worms, S-3, was obtained
from Dr. Ralph I. Smith, who had adapted it from 81 (/c sea water to 106%. It is
assumed that consideration of this worm is not remiss, for this salinity is well within
the range reported for the species and at none of these salinities are the worms
osmoregulating. ) As before, the worms were relaxed, fixed in Bouin's, sectioned
serially at 6 micra, stained in Harris' hematoxylin, and counterstained with eosin.
One of the most obvious characteristics of the nephridia of the down-river form
of Nereis Ihiinicolo is their shape. Whereas in the up-river form there was a
NPR
lOOjj
FIGURE 11. Graphic representation of the inner diameter and wall thickness of a nephridial
canal of Nereis liiiiuicolu from down-river (reconstructed from sectioned material). PSC,
post-septal canal ; NPR, nephridiopore.
slight compression in the dorsal portion and the ventral half was hemispherical, in
the worms from the Salicornia marsh, there is a general compression of the entire
nephridium (Fig. 5). This usually is seen to occur parallel to the axis of the
post-septal canal which projects obliquely, anteriad and mediad. In extreme
dimensions, the nephridium measures about 400 micra long, 350 micra high, and
150 micra thick (through the medial half). The lateral half is approximately 50
micra thick, and the approximate volume is 0.0122 mm.3
As just intimated, there is an external division into medial and lateral halves,
the medial half being elliptical in cross-section, and the lateral half being extremely
compressed to about one-third the thickness of the other. In the extreme dorsal
sections of the nephridium the two halves are entirely separate.
In comparing the sectioned nephridia of the down-river and the up-river forms
of N. limnicola, the first glance at those from down-river would lead one to doubt
NEPHRIDIA OF NEREIS LIMNICOLA 371
that the two were at all related. A considerahle reduction occurs in the diameters
of the canal lumen, and, to a lesser degree, a reduction in the number of blood
vessels (Fig. 5). In addition, the interstitial tissue seems more dense than that in
either the up-river form or in N. I'c.i'illosa.
The diameter of the tubule lumen is, in most portions, as little as 1 to 2 micra.
In some cases, the lumen of the canal is almost completely closed and only a pin-
point of clear area is visible by careful focusing. Under these circumstances no
measurement is possible, and in graphing the tubule diameter (Fig. 11), these
perforations were considered to be one micron or less in diameter. At other times,
the perforations were obscured either by heavy ciliation or a turning of the canal
within the 6-micron thickness of the section. It is possible that this general narrow-
ness of the canal might have been an artifact caused by osmotic factors during
fixation and/or relaxation. However, there were portions of the tubule present in
the same section, with diameters comparable to those of N. vexillosa and the up-
river form of N. limnicola (Figs. 6 and 7). In some, a well-defined boundary was
visible, complete with a basement membrane separating the tubule wall from the
interstitial tissue. In others, there was an irregular area of vacuolation surrounding
the perforation. In still others, the network of the interstitial tissue extended up
to the canal and the poorly-defined wall may have been due to a thickening of the
network or to the presence of extremely fine participate material. It was impossible
to make an accurate judgment here, for the nature of the darkening was not
resolvable, microscopically. It is interesting to note in these worms from down-
river, that not all of the peripheral areas of the nephridial perforations were well-
defined, but that all were surrounded by an area that stained darker with hema-
toxylin than the interstitial tissue. Indeed, in the case of those lumen perforations
which were closed most tightly, the darkened areas helped to locate the fine canal
openings.
As stated above, the down-stream form stands in contrast to its up-stream
counterpart in the lesser amount of vascularization of the nephridium. Of the two
nephridial halves referred to, the more lateral is the more vascularized (Fig. 6,
BV). The nephridium of the down-river form is extremely well-supplied with
a network of small blood vessels that ramify over its surface (especially that of the
lateral half). These find their origin in the ventral segmental blood vessel, which
itself proceeds over the anterior face of the medial half of the nephridium, and
finally departs toward the parapodium, about 60 micra from the body wall. In the
central part of the nephridium, this vessel gives rise to a branch that remains in
contact with the medial half until immediately before ventral contact is made with
the body wall (Fig. 5, VS). This last branch and the large ventral segmental
vessel are the only blood vessels in contact with the medial half of the nephridium.
In the up-river form, the post-septal canal is fairly short and has a lumen diam-
eter much the same as that of the main canal in the nephridial mass ; in the down-
river form (worms of comparable size) the post-septal canal is twice as long (about
250 micra) and the lumen is almost entirely closed at many points (Fig. 9B). The
post-septal canal wall of the down-river form is also much thinner, about 1 to 2
micra for the most part. The small size of the tubule makes it difficult to trace
from the nephridial mass to the nephrostome, for it is closely applied to the ventral
segmental vessel throughout (Fig. 8), and at times, in transverse section, resembles
372 MEREDITH L. JONES
a small cell attached to the hlood vessel. The nuclei which are visible within the
tube wall possess little chromatin. The "septal band" separating the post-septal
canal from the nephrostome is composed of a concentration of nuclei, but is not
so extensive as the bands previously described for the up-river form and N. vexillosa
(Fig. 8, NB).
The nephrostome (Fig. 8), which extends anteriorly about 100 micra from the
dark band, has walls that appear to be solid, and there are no large vacuoles within
them. The walls seem to be about the same density as the interstitial tissue of the
nephridial mass. Around the margin of the nephrostome, the protoplasmic proc-
esses are stout, club-shaped structures that give rise to long cilia (Fig. 8, CP).
Their shape would seem to bear out Goodrich's (1945) statement concerning the
specificity of these structures, for they are similar to those observed in the up-river
form, but differ from those of Nereis vexillosa and N. diversicolor. As before, the
number of cilia originating inside the funnel is sufficient to clog the lumen (Fig.
8, CM).
Figure 11 shows the diameter of the nephridial lumen of the down-river form
of Nereis limnicola. It is seen that the lumen is quite narrow at its beginning
(of the order of 1 to 6 micra) and gradually increases in size, until at the three-
quarter mark, it is consistently larger. Beyond this point, it undergoes a series of
irregularities, grows extremely wide, closes once more, and finally becomes fairly
uniform close to the nephridiopore. It is fully walled throughout; at, and just
subsequent to, its widest part, the wall is at its thinnest ; also, the wall thickens
considerably as it approaches the nephridiopore. In the case of the nephridium
upon which the diagram is based, the length of the canal within the nephridial mass
is about 3864 micra, with a mean diameter of 9.3 micra.
The region of the nephridiopore of the down-river form of N. limnicola is essen-
tially the same as that of the up-river form. As the nephridial canal approaches
the body wall, the walls of the canal thicken, and contain large, relatively clear,
nuclei. At times, the area shows the same compression as described for the up-
river form.
DISCUSSION
Several points emerge from the descriptions above : the nephridia of both the
up-river and down-river forms of N. limnicola are more highly vascularized than
those of N. vexillosa; the nephridia of the up-river form are more highly vas-
cularized than those of the down-river form; the down-river form possesses a
longer and more narrow nephridial canal than the specimens from up-river ; and
the nephridial blood vessels of both forms do not come into contact with the
nephridial canal.
Krishnan (1952) found that the nephridia of Namalycastis indica, a euryhaline
species, were larger and more heavily vascularized than those of the other nereids
he studied. He also found that some of the nephridial blood vessels were in inti-
mate contact with the canal wall and that, in the case of worms acclimatized to
full-strength sea water, there was a lessening of the blood supply to the nephridia,
in terms of shrunken and collapsed vessels. He suggested (p. 248) that the
reduced blood supply might indicate that these nephridia "... are probably doing
less osmotic work than in the normal forms living in fresh water." Krishnan fur-
NEPHRIDIA OF NEREIS LIMNICOLA
373
ther postulated that there is a direct relationship between the size of nephridia and
the osmoregulatory ability of the species in question, and that the ability of a nereid
to osmoregulate also was reflected, not only by the amount of nephridial vascu-
larization, but by the proximity of blood vessels to the nephridial canal.
It would seem from the series of three species considered by Krishnan that
there is, indeed, a correlation between nephridial size and the ability to osmoregu-
late; but it should be noted that Nereis vexillosa, a stenohaline, relatively high-
salinity species, possesses nephridia nearly as large as those of Nauialycastis indica
(Jones, 1957). Further, the nephridia of the up-river form of Nereis limnicola,
which one would assume to lie osmoregulating, are larger than those of N. indica,
but quite a bit smaller than those of the down-river form which one would assume
to be doing less osmoregulatory work.
TABLE I
Derivation of Indices of Excretory Capacity of nephridia from specimens of J\ereis limnicola,
from up-river (S-2), adapted from low to high salanity (S-l),from down-river (S-3),
adapted from high to low salinity (S-4)
A
B
c
D
E
F
Worm
Number of
sections
counted
Assumed
total number
of canal sections
(A X3)
Assumed
length of canal
(B X 6 M)
Number of
segments
Length (/j)
Index of
Excretory
Capacity
XC X2DA
V, E )
S-2
128
384
2304
50
38,000
6.982
100
300
1800
50
38,000
5.455
98
294
1764
50
38,000
5.345
86
258
1548
50
38,000
4.691
S-l
237
711
4266
42
30,000
11.945
194
582
3492
42
30,000
9.778
S-3
290
870
5220
61
35,000
18.191
274
822
4932
61
35,000
17.192
224
672
4032
61
35,000
14.054
252
756
4536
61
35,000
15.811
S-4
243
729
4374
62
33,000
16.436
194
582
3492
62
33,000
13.121
Clearly, some character other than size, alone, allows these various nereids to
survive in a dilute medium. Krishnamoorthi (1963b, 1963c) invoked size as a
criterion of regulatory ability but, in addition, suggested that the length of the
nephridial canal, as embodied in his "Index of Excretory Capacity" (== length of
excretory surface, in microns/length of worm, in microns ; "excretory surface" is
defined as the average length of nephridial canal multiplied by the average number
of nephridia per worm), was also a reflection of osmoregulation. Krishnamoorthi
found that the indices of excretory capability were correlated with the distribution
of four polychaetes, as he found them in the River Adyar and the nearby Bay of
Bengal (Krishnamoorthi, 1963a) : Diopatra variabilis Southern, index -- 0.350,
salinity range = 20-26% 0 ; Euclymene insecta (Ehlers), 0.310, 20-26%,? ; OnitpJiis
eremita Audoin and Milne Edwards, 0.247, 30-34% c ; and Loimia medusa (Sa-
vigny), 0.225, 30-34%c. Although an extended series of pertinent observations
374 MEREDITH L. JONES
was not conducted on the length of nephridia] canals of the up-river and down-
river forms of N. limnicola, certain assumptions can be made. If one assumes
that the number of canal sections counted in every third nephridial section (Table
I, column A) is a reasonable estimate of one-third of the total number of canal
sections, then, by multiplying by three (Table I, column B) and by the thickness
of the sections, 6 microns (Table I, column C), one can arrive at an estimate of
the length of a given nephridial canal. If this number is multiplied by twice the
number of segments of the worm and this is, in turn, divided by the worm's length,
in microns, one obtains Krishnamoorthi's Index of Excretory Capacity. Depending
upon which nephridium and which population is chosen, the indices vary from
4.691-6.982 for the up-river forms to 14.054-18.195 for the down-river forms
(Table I, column F). Worms cross-adapted from high to low and from low to
high salinities give intermediate indices.
It would seem, intuitively, that the up-river population would have need of a
greater "excretory capability," yet it has the lowest indices of the specimens of
N. limnicola considered here. In addition, the lowest of the index values are more
than ten times those found by Krishnamoorthi. Clearly, then, the Index of Excre-
tory Capacity, itself, can not give an adequate idea of the osmoregulatory capabilities
of a polychaete living in a low-salinity or fresh-water habitat.
Krishnan (1952) and Krishnamoorthi (1963b, 1963c) also have suggested that
there is a correlation between the amount of vascularization and the ability to
osmoregulate. Although subjective observations of the amount of nephridial vas-
cularization of N. limnicola would seem to confirm this, I have not found a satis-
factory method of quantifying these differences.
Yet another nephridial parameter might lie considered, in addition to overall
nephridial size, relative length of nephridial canal, and nephridial vascularization.
Reduced to essentials, the survival of an animal with a permeable integument in
a hyposmotic medium depends on (a) its ability to control its volume and, in
effect, to slow or stop the osmotic inflow by hydrostatic pressure; (b) its ability
to tolerate a dilution of its body fluids; or (c) its ability to counteract the dilutive
effect of the osmotic inflow by the rapid excretion of water. Although the first
two possibilities are outside the purview of the present work, observations have
been made above, which bear on the third.
A number of papers have appeared which have been concerned with various
physiological responses of N. limnicola to dilute or fresh water media. All of these
postulate that there must be some means of volume control (Smith, 1963), a means
of modifying salt loss rate (Smith, 1963), and/or a means of increasing the ability
of the worm to eliminate excess water (Smith, 1959a, 1963; Oglesby, 1965b).
It has been noted that there is an apparent difference in the diameter of the
nephridial canal of the two forms of N. limnicola considered here. In an effort
to establish the statistical validity of these apparent differences, a number of
nephridia of both forms were examined (Table II). Using 14 nephridia from six
different up-river specimens and ten nephridia from four different specimens from
the down-river area, all of the perforations of sectioned nephridial canals were
measured in every third section of each nephridium. The results of all measure-
ments of all nephridial canals of both forms were cast as frequency distributions
(Fig. 12), and it was found that the mean canal diameter of the up-river forms
NEPHRIDIA OF NEREIS LIMNICOLA
375
TABLE 1 1
Collection data and various measurements of specimens of Aereis limnicola
considered in the present study
Collection data
Mean
Salinity
Body
Length
canal
Mean ± 2 S.E.
\\ orrn
at death
segments
mm.
diameter
(M)
Date
Salinity
Crt
Nereis limnicola, Up-river
S-2
10 Dec. 1950
0.55% SW
0.75% SW
50
33
22.84
21.28-24.40
S-2
10 Dec. 1950
0.55% S\Y
0.75% SW
50
33
16.82
15.64-18.00
S-2
10 Dec. 1950
0.55% SW
0.75% SW
50
33
11.03
10.03-12.03
S-2
10 Dec. 1950
0.55% S\V
0.75r; S\Y
50
33
10.66
9.66-11.66
RB
6 May 1951
2.45% SW
2.45% SW
52
—
30.36
28.18-32.54
RB
6 May 1951
2.45% SW
2.45% SW
52
—
21.59
19.63-23.55
S-10
6 May 1951
2.45%, SW
2.45%, SW
46
—
19.27
18.09-20.45
S-10
6 May 1951
2.45% SW
2.45%, SW
46
—
17.74
16.60-18.88
49C
13 Jan' 1949
1.49%SW
1.49%SW
—
—
31.87
30.57-33.17
49C
13 Jan. 1949
1.49%S\V
1.49% SW
—
—
27.26
25.08-29.44
49 D
13 Jan. 1949
1.49% SW
1.49%SW
—
• —
20.07
18.23-21.91
49D
13 Jan. 1949
1.49% SW
1.49% SW
—
—
18.51
16.27-20.75
49E
13 Jan. 1949
1.49%SW
1.49% SW
• —
—
8.59
7.55- 9.63
49E
13 Jan. 1949
1.49%SW
1.49% SW
—
—
9.21
8.07-10.35
All data
—
—
—
• —
—
20.79
20.48-21.10
pooled
Nereis limnicola, Down-river
S-3
10 Dec. 1950
81.00%SW
106.00% SW
61
35
9.79
8.97-10.61
S-3
10 Dec. 1950
81.00% SW
106.00% SW
61
35
9.54
8.72-10.36
S-3
10 Dec. 1950
81.00% SW
106.00% SW
61
35
9.09
8.37- 9.81
S-3
10 Dec. 1950
81.00% SW
106.00% SW
61
35
8.83
8.15- 9.51
SB
6 May 1951
48.00% SWT
48.00% SW
64
—
4.83
4.23- 5.43
SB
6 May 1951
48.00% SW
48.00% SW
64
—
4.18
3.74- 4.62
S-13
6 May 1951
48.00% SW
48.00% SW
68
— •
4.63
4.07- 5.19
S-13
6 May 1951
48.00% SW
48.00% SW
68
—
4.33
3.89- 4.77
51 A
21 Feb. 1951
47.50% SW
47.50% SW
57
—
11.21
10.11-12.31
51A
21 Feb. 1951
47.50% SW
47.50% SW
57
—
10.88
9.64-12.12
All data
—
—
—
—
—
8.49
8.19- 8.79
pooled
Nereis limnicola, Cross-adapted
S-l
29 Apr. 1951
0.55% SW
118.00% SW
42
30
22.19
20.85-23.53
S-l
29 Apr. 1951
0.55% SW
118.00% SW
42
30
11.89
11.17-12.61
S-4
1 June 1951
81.00% SW
0.80% SW
62
33
17.32
16.40-18.24
S-4
1 June 1951
81.00%, SW
0.80% SW
62
33
13.58
12.82-14.34
Nereis vexillosa, San Francisco Bay
VI
3 May 1951
73-90% SW
73-90%, SW
—
—
16.92
15.56-18.28
VI
3 May 1951
73-90% SW
73-90% SW
—
—
16.75
15.53-17.97
V2
3 May 1951
73-90% S\V
73-90% SW
—
—
14.17
13.13-15.21
V2
3 May 1951
73-90% SW
73-90% SW
—
—
15.11
13.67-16.55
V4
3 Mav 1951
73-90% SW
73-90% SW
—
—
15.67
14.01-17.33
V4
3 May 1951
73-90% SW
73-90% SW
— •
—
15.73
14.33-17.13
376
MEREDITH L. JONES
was 20.79 niicra (one standard error ~ 0.31 micra) and that of the down-river forms
was 8.49 micra (one standard error -- 0.15 micra). Utilizing the "Student" t test,
it was found that there was, indeed, a significant difference between the mean canal
diameters of the two forms (£ — 38.44). This also can be interpreted as the
difference between the two means being 38.44 times the standard error of this
difference.
The results above, however, may not be so straightforward as they might seem.
If the mean canal diameter (±2 standard errors) of each nephridium examined is
plotted against salinity (Fig. 13), it is seen that there is a rather large spread of
the data derived from the up-river forms. Indeed, the results from three of the
up-river worms (S-2, 49C and RB) indicate that there is a real difference be-
tween and among the diameters of the nephridial canal in the same animal, and
10
15
20
25 30 35 40 45
Micrometer units (I unit = l.
50
55
60
65
70
FIGURE 12. Frequency polygons showing the difference in nephridial canal diameter
between the up-river forms of Nereis limnicola (based on 14 nephridia from six specimens) and
the down-river forms (based on ten nephridia from four specimens).
two specimens (S-2 and 49E) have nephridial canals whose diameters are not
significantly different from at least some of those from the down-river locality.
In addition to observations of the nephridia of worms sacrificed directly from
the salinities in which they were collected, examinations were made of the nephridia
of two cross-adapted worms. In the case of specimen S-l (originally up-river),
the adaptation was from 0.55% sea water to 118% sea water and of S-4 (originally
down-river), from 81% sea water to 0.80% sea water. The general aspect of the
nephridia of both S-l (Figs. 14 and 15) and S-4 (Figs. 16 and 17) is strikingly
similar to the nephridia of the up-river population of Nereis limnicola. The aver-
age canal diameter of S-4 (Fig. 13) falls among the lower values of the up-river
forms, while that of S-l is comparable to the larger canal diameters of the up-river
population, even though S-l was acclimatized to 118% sea water just before it
was sacrificed.
NEPHRIDIA OF NEREIS LIMNICOLA
377
«tu-
49C
RB
ff
30
;
"":
i
i
—
S-2 I.,
i
Si
—
;
7!
al diameter
CV)
0
fl-\ 49D
ilfl I";
•
1
•3 Up-river (a II data pooled)
I V4 1 j
c
• '1
s-io
Wp- ;' j
o
: L S-4
, viy
51 A J.
-
Uli 49E
i) *
to
1
©Down-river (all data pooled) xy
S-3
5
n^XSB
-
S ~ 1 o
n-
0.5
1.0
45 10 20
Salinity (% seawater)
30 40 50
100
150
FIGURE 13. Graph showing the relationships among all nephridial canals considered.
Horizontal lines represent mean canal diameters for each nephridium examined and the vertical
lines, two standard errors above and below the mean. Specimen numbers are referable to Table
II. Symbols surrounded by solid lines represent down-river forms; those with dashed lines,
up-river forms; those with central solid circles, cross-adapted specimens; and those with central
open half-circles, Nereis vexillosa from San Francisco Bay.
For comparison, Figure 13 also includes data based on observations of the
nephridia of Nereis vexillosa from the Berkeley Yacht Harbor, San Francisco Bay
(Jones, 1957). Although the salinity of the environment at the time of collection
was not determined, the means ±2 S.E. for specimens V1( V2, and V4 are clustered
around an estimated salinity range, i.e., 73-90% sea water (Jones, 1957, p. 407).
The nephridial canals of N. vexillosa are significantly larger than those of the
down-river forms of N. limnicola and are of comparable size to half of the up-river
forms.
A comparison of the data of Table II indicates that, in the case of down-river
forms, there is no difference between canal diameters of nephridia from the same
segment (specimens S-3 and S-13) or from succeeding segments (specimen SB).
In up-river forms, there is a significant difference in canal diameters of nephridia
from the same segment in two of five cases (specimens S-2 and 49C), and in the
one case of nephridia from succeeding segments (specimen RB). In both of the
adapted specimens, S-l and S-4, there is also a significant difference in the case
of nephridia from the same segment.
378
MEREDITH L. JONES
NC
14
•200 v-
'CL-
15
50 jj-
NC
*VS
NC-
i
is ^, V.;
Lettering as in Figures 1-8.
Figures 14 and 15, up-river form of N. limmcola, adapted from 0.55% sea water to 118%
sea water; Figures 16 and 17, down-river form of AT. liinnicola, adapted from 81% sea water
to 0.80% sea water.
FIGURE 14. Dorsal view of right nephridium and associated nephrostome ; specimen S-l.
FIGURE 15. View of nephridial tissue ; specimen S-l.
FIGURE 16. Dorsal view of right nephridium and associated nephrostome ; specimen S-4.
FIGURE 17. View of nephridial tissue ; specimen S-4.
Because of these apparently conflicting observations, that is, the small diameter
of the nephridial canals of 49E and some of those of S-4 from up-river, and the
large diameter of S-l, it is apparent that some physiological and/or physical mecha-
nism, in addition to nephridial canal diameter, operates to allow N. limnicola to
survive in dilute media.
That a larger canal diameter is advantageous in coping with lowered salinity
is suggested by S-4 which apparently developed a larger nephridial canal as it was
acclimatized from 81 to 0.80% sea water. That an environment of higher salinity
does not necessarily evoke a comparable diminution of canal diameter is suggested
by S-l which apparently maintained a larger canal diameter in one of the measured
nephridia while it was acclimatized from 0.55 to 118% sea water.
It would seem, then, that even though the annual fluctuations of salinity in the
down-river area may be far greater than those up-river, nephridia with a relatively
small diameter are adequate to the osmotic stresses placed on the worms in this
area. On the other hand, the nephridia of the down-river forms appear to be more
plastic in their response to a fresh-water or near-fresh-water medium ; quite pos-
sibly, nephridial activity, insofar as water excretion is concerned, may be aug-
NEPHRIDIA OF NEREIS LIMNICOLA 379
merited or superseded by some other mechanism. Finally, there appears to he a
general trend toward nephridia of large lumen diameter in the up-river forms,
although this is not invariably the case.
I would here extend thanks to Drs. Ralph I. Smith, Kenneth B. DeOme, and
Howard A. Bern, all of the Department of Zoology, University of California
(where much of the basic work of this study was carried out), to Dr. Larry C.
Oglesby of Reed College, and to Drs. Marian H. Pettibone and Clyde Roper,
Smithsonian Institution ; the advice, criticisms, and suggestions of all of these have
been gratefully received, if not always followed. I am particularly obliged to Dr.
Smith for the use of a number of specimens of N. limnicola from his collections,
both sectioned and un-sectioned.
SUMMARY
1. The morphology of the nephridia of specimens of the polychaete worm,
Nereis limnicola Johnson from areas of different salinity in the estuary of the
Salinas River is described.
2. Generally, the canal diameters of the nephridia of the up-river (low salinity )
forms are larger than those from down-river ( high salinity ) ; the nephridia of the
up-river forms are more highly vascularized than those from animals found in
higher salinities. This suggests that the nephridial canal acts to rid the animal of
the excess water brought into its body by osmotic influx.
3. Nephridial canal diameters of worms adapted from low to high and from
high to low salinities approach those of the animals from low salinity ; this suggests
that a larger canal diameter is efficacious in coping with the osmoregulatory prob-
lems presented by a dilute medium, and that canal diameter is not very important
in higher salinities.
4. Inconsistencies in the correlation of large nephridial canal diameter with low-
salinity suggest that other mechanisms are utilized in meeting the stresses imposed
by an environment of low salinity.
j j
5. Krishnamoorthi's Index of Excretory Capacity is derived for a number of
nephridia ; the results indicate that the Index and/or the nephridia of N. limnicola
do not seem to be comparable with Krishnamoorthi's observations on polychaetes
of India.
LITERATURE CITED
AIYAR, R. G., 1933. On the anatomy of Marphysa gravel vi Southern. Rcc. Indian Mas.,
Calcutta, 35: 287-323.
BENHAM, W. B., 1891. Nephridium of Lin/ihriciis and its blood supply with remarks on the
nephridium of chaetopods. Quart. J. Micr. Sci., 32: 293-334.
BERKELEY, E., AND C. BERKELEY, 1956. A new species and two new records of Polychaeta from
eastern Canada. Canadian J. Zoo!., 34: 267-271.
DAY, J. H., 1951. The ecology of South African estuaries. Part I. A review of estuarine
conditions in general. Trans. Roy. Soc. South Africa, 33: 53-91.
EHLERS, E., 1864—68. Die Borstenwiirmer (Annelida Chaetopoda) nach systematischen und
anatomischen Untersuchungen dargestellt. Leipzig, W. Engelmann, 748 pp.
FACE, L., 1906. Recherches sur les organes segmentaires des Annelides Polychetes. Ann. Sci.
Nat. Zoo/. Paris. Scr. 9, 3: 261-410.
GOODRICH, E. S., 1893. On a new organ in the Lycoridea, and on the nephridium in Nereis
diversicolor O. F. Mull. Quart. J. Micr. Sci., 34: 387-402.
GOODRICH, E. S., 1895. On the coelom, genital ducts and nephridia. Quart. J. Micr. Sci., 37:
477-510.
380 MEREDITH L. JONES
GOODRICH, E. S., 1897. On the nephridia of Polychaeta. Part I. On Hesione, Tyrrhena and
Ncphthys. Quart. J. Micr. Sci., 40: 185-196.
GOODRICH, E. S., 1898. On the nephridia of Polychaeta. Part II. Glyccra and Goniada.
Quart. J. Micr. Sci.. 41 : 439-457.
GOODRICH, E. S., 1900. On the nephridia of the Polychaeta. Part 111. The Phyllodocidae,
Syllidae, Amphinomidae, etc., with summary and conclusions. Quart. J. Micr. Sci.,
43: 699-748.
GOODRICH, E. S., 1945. The study of nephridia and genital ducts since 1895. Quart. J. Micr.
Sci.. 86: 113-392.
HARTMAN, O., 1938. Brackish and fresh-water Nereidae from the northwest Pacific, with the
description of a new species from central California. Univ. California Publ. Zoo]., 43:
79-82.
HARTMAN, O., 1944. Polychaetous annelids from California including the description of two
new genera and nine new species. Allan Hancock Pacif. Experts., 10: 239-307.
HARTMAN, O., 1960. An account of the nereid worm, Ncanthes diversicolor, comb. n. in the
Caspian Sea and its more extensive distribution. Zoo/. Zliuni., 39: 35-39 (in Russian,
English summary).
IMAJIMA, M., AND O. HARTMAN, 1964. The polychaetous annelids of Japan. Allan Hancock
Found. Publ., Occas. Papers, No. 26, pp. 1-452.
IZUKA, A., 1908. On the breeding habit and development of Nereis japonica n. sp. Annot.
Zool. Japan, 6: 295-305.
JOHNSON, H. P., 1903. Fresh-water nereids from the Pacific Coast and Hawaii, with remarks
on fresh-water Polychaeta in general. Mark Anniv. Vol., Art. 10, pp. 205-223.
JONES, M. L., 1957. On the morphology of the nephridium of Nereis vexillosa Grube. Biol.
Bull, 113: 407-413.
KHLEBOVICH, V. V., 1963. On the systematic position of the nereid from the Caspian Sea.
Zool. Zlmrn., 42: 129-131 (in Russian, English summary).
KRISHNAMOORTHI, B., 1963a. On the distribution of six species of polychaetes in the Adyar
Estuary, Madras. /. Mar. Biol. Assoc. India, 5: 97-102.
KRISHNAMOORTHI, B., 1963b. Gross morphology and histology of nephridia in four species of
polychaetes. Proc. Indian Acad. Sci., B, 57 : 195-209.
KRISHNAMOORTHI, B., 1963c. Volume regulation in eggs, larvae and adults of a brackish-water
polychaete, Diopatra variabilis (Southern). Proc. Indian Acad. Sci., B, 57: 275-289.
KRISHNAN, G., 1952. On the nephridia of Nereidae in relation to habitat. Proc. Natl. Inst.
Sci. India, 18: 241-255.
OGLESBY, L. C., 1965a. Steady-state parameters of water and chloride regulation in estuarine
nereid polychaetes. Comp. Biochem. Physiol., 14: 621-640.
OGLESBY, L. C., 1965b. Water and chloride fluxes in estuarine nereid polychaetes. Comp.
Biochem. Physiol., 16: 437-455.
PETTIBONE, M. H., 1963. Marine polychaete worms of the New England region. 1. Families
Aphroditidae through Trochochaetidae. U. S. Natl. Mus. Bull, 227: 1-356.
SMITH, R. I., 1950. Embryonic development in the viviparous nereid polychaete Neanthes
lighti Hartman. /. Morph., 87: 417-455.
SMITH, R. I., 1953. The distribution of the polychaete Neanthes lighti in the Salinas River
Estuary, California, in relation to salinity, 1948-1952. Biol. Bull, 105: 335-347.
SMITH, R. L, 1957. A note on the tolerance of low salinities by nereid polychaetes and its
relation to temperature and reproductive habit. L'Annee Biol., 33: 93-107.
SMITH, R. L, 1958. On reproductive pattern as a specific characteristic among nereid poly-
chaetes. Syst. Zool., 7: 60-73.
SMITH, R. I., 1959a. Physiological and ecological problems of brackish waters. Proc. Twen-
tieth Annual Biol. Colloq., Oregon State College, pp. 59-69.
SMITH, R. I., 1959b. The synonymy of the viviparous polychaete Neanthes lighti Hartman
(1938) with Nereis limnicola Johnson (1903). Pacif. Sci.. 13: 349-350.
SMITH, R. L, 1963. A comparison of salt loss rate in three species of brackish-water nereid
polychaetes. Biol. Bull., 125 : 332-343.
SMITH, R. L, 1964. D2O uptake rate in two brackish-water nereid polychaetes. Biol. Bull.,
126: 142-149.
STEPHENS, G. C., 1964. Uptake of organic material by aquatic invertebrates. HI. Uptake of
glycine by brackish-water annelids. Biol. Bull., 126: 150-162.
CARDIOREGULATION IN LIMULUS. II. GAMMA AMINOBUTYRIC
ACID, ANTAGONISTS AND INHIBITOR NERVES
RALPH A. PAX i, - AND RICHARD C. SANBORN
Department of Biological Sciences, Purdue University, Lafayette, Indiana 47907
and the Marine Biological Laboratory, UToods Hole, Massachusetts 02543
The neurogenic beat of the Liunihts heart has long been regarded the classic
example of a neurogenic rhythm (Carlson, 1909). As in other neurogenic hearts,
such as those of Crustacea, the rate and strength of beating can be decreased by
stimulation of cardioinhibitory nerves arising from the central nervous system
(Carlson, 1905; Heinbecker, 1933; Pax and Sanborn, 1964). In Limidus the de-
crease in heart rate is not tightly coupled to stimulation of the inhibitor nerves, a
time lag in the response occurring both at the beginning and at the end of the
stimulation periods. It is probable, therefore, that inhibition in the Lhnulus heart
is chemically mediated (Pax and Sanborn, 1964).
The nature of the chemical mediator of inhibition is not known. 5-Hydroxy-
tryptamine (5-HT, serotonin) has been reported to slow the rate of rhythmic dis-
charge from the isolated cardiac ganglion (Burgen and Kuffler, 1957). However,
in other neurogenic hearts, 5-HT and related compounds have excitatory effects
(Kerkut and Price, 1964).
Gatnma-aminobutyric acid (GAB A) has also been reported to inhibit the
Li nt u I us heart (Burgen and Kuffler, 1957). This compound has inhibitory effects
on neuromuscular phenomena in a wide variety of other animals. It is present in
lobster inhibitory motor neurons but not in excitatory motor neurons (Kravitz
ct al., 1963). At the crustacean neuromuscular junction, it mimics the action of
the inhibitory transmitter both postsynaptically and presynaptically (Dudel, 1965;
Takeuchi and Takeuchi, 1966) and in the crustacean cardiac ganglion GABA
closely mimics the action of the inhibitor (Florey, 1957; Maynard, 1961). From
this evidence it appears possible that GABA or a GABA-like compound may be
responsible for cardioinhibition in the Linniliis heart. We report here results of
experiments exploring this possibility more fully.
A primary requirement of any supposed transmitter is that, when artificially
applied, it mimics in all respects stimulation of the prejunctional structure (Mc-
Lennan, 1963). Stimulation of the cardioinhibitory nerves in Linntliis results in
a decrease in rate and strength of beating of the intact heart, a decrease in the
number of units discharging in the cardiac ganglion during each burst of electrical
activity and a decrease in the total duration of each burst (Carlson, 1905 ; Hein-
becker, 1933; Pax and Sanborn, 1964). We have tested the ability of exogenously
applied GABA to mimic these actions of the inhibitor nerves.
1 Present Address : Department of Zoology, Michigan State University, East Lansing,
Michigan.
2 Predoctoral Fellow of the National Institute of General Medical Sciences, N. I. H.
381
RALPH A. PAX AND RICHARD C. SANBORN
Since data obtained by application of supposed transmitters to tbe cardiac
ganglion are at best equivocal, we have also followed a second line of investigation.
Compounds which block the action of the endogenous transmitter should similarly
antagonize the effects of exogenously applied GABA.
this purpose we have used picrotoxin, a compound capable of blocking the action
of GABA in other systems (Van der Kloot ct al., 1958). \Ye have tested picro-
toxin for its ability to block the action of the endogenous transmitter, i.e.. block the
action of the inhibitor ueivts. We have also tested picrotoxin for its ability to
block the action of the supposed transmitter artificially applied to the heart. For
MATERIALS AND METHODS
Source and maintenance of animals
Adult Limit Ins polyphemus, maintained as previously described (Pax and San-
born, 1964), 20 to 25 cm. maximal width, were used in all experiments. They
were shipped by air express at two-week intervals from the Supply Department,
Marine Biological Laboratory, Woods Hole, Massachusetts, and maintained in
moist excelsior at a temperature of 5° C. Responses of animals so maintained
did not vary for at least six weeks.
Animal preparations
Isolation of the heart from Limit! us requires removal of the tough dorsal exo-
skeleton. This is best done by sawing through the exoskeleton just lateral to the
underlying heart and joining the lateral cuts with transverse anterior and posterior
cuts so that a rectangular piece of isolated exoskeleton overlying the heart may be
removed by lifting and scraping it free of the underlying tissues. Once this piece
of exoskeleton has been removed the internal extensor muscles of the opisthosoma
dorsal to the heart in the cephalothorax and the epidermal tissue overlying it in
the opisthosoma can be dissected away. The intact heart can then be removed.
Stimulation of inhibitor nen'es
As we suggested earlier (Pax and Sanborn, 1964), stimulation of the inhibitor
nerves near the ventral nerve ring is undesirable since they also contain fibers which
innervate muscles. We have since been able to locate inhibitor fibers as they enter
the heart dorsally. At these sites the nerves apparently consist exclusively of
cardioinhibitory fibers.
Perfusion of the isolated heart
After removal from the animal the heart was placed in a V-shaped Incite
chamber 15 cm. in length. The heart was ligated anteriorly in the secondi segment
and posteriorly a cannula was inserted into the lumen of the heart through the
cardiac muscle.
Tension on the heart walls and the amount of intra-lumenal pressure both
influence the rate and the strength of beating of the heart (Carlson, 1907). Longi-
tudinal tension approximating that on the heart in situ was obtained by stretching
the heart to a length equal to that present before removal from the animal.
GABA, ANTAGONISTS, AND LIMULUS HEART
In order to maintain an intra-lumenal pressure, a gravity-feed reservoir of Qiao's
(1933) saline solution (0.44 .17 XaCl, 0.009 M KC1, 0.037 M CaCl, ) was con-
nected to the cannula at the posterior of the heart. The hearts were perfused at
the rate of 20 ml. per minute, the route of the perfusion fluid being from the lumen
of the heart out through the ostia and lateral arteries to the exterior. The total
volume of fluid in the chamber was maintained at 10 ml. by providing an overflow
in the chamber near the anterior end of the heart.
Recording of data
Electrical activity was recorded from the cardiac ganglion of the intact heart
by dissecting it free of the heart muscle in the second and the third segments and
placing it over hooked platinum electrodes. From the isolated cardiac ganglion,
electrical activity was recorded by stringing the ganglion through a series of 12
platinum loop electrodes spaced five mm. apart. During the course of a single
experiment any of these electrodes could be chosen to be used as recording electrode.
Measurement of mechanical activity of the heart muscle was obtained with a
Statham GlOb displacement transducer (maximum displacement 0.15 oz.).
Experimental metJwds and dnnjs
All drugs were dissolved in Qiao's (1933) saline as shortly before use as prac-
ticable. Parallel reservoirs of saline and drug solution were connected to the
chamber through a two-channel stopcock so that perfusion could be alternated by
a turn of the barrel.
Data reduction
Heart rates in Liinulits vary greatly from animal to animal (Pax and Sanborn,
1964). Moreover those hearts which have an initial high rate of beating tend to
have a greater change in rate during inhibition than do those which have an initial
low rate of beating. For these reasons we have, when measuring changes in rate,
used each animal as its own control and expressed all rates as relative heart rates.
Relative heart rate is defined as the ratio of the experimentally altered rate to the
control rate. Thus relative rates of less than one are indicative of inhibition and
values greater than one indicate excitation. In a similar manner, all data on
strength of contraction of the heart muscle are expressed as relative strengths.
In drug perfusion experiments relative rates and contraction strengths were
calculated from the mean rates and contraction strengths during the last two minutes
of perfusion. In stimulation experiments relative rates were calculated from the
mean rates during the entire time of stimulation.
RESULTS
Gamma-aminobutyric acid
Perfusion of GABA through the intact isolated Lhniilus heart results in a de-
creased heart rate. A typical result of such perfusion is shown in Figure 1, and
the results of 21 such perfusions in 15 different hearts are plotted in Figure 2.
The solid line in the figure is the regression line for these data as determined by
384
RALPH A. PAX AND RICHARD C. SANBORN
l | I I M | | I I ! !
i
LI
2 MIN
FIGURE 1. The response of the isolated heart to perfusion with GAB A. The record is
continuous from upper left to lower right. One hundred ml. of 5 X 10~5 M GABA were per-
fused during the time between the two arrows.
the method of least squares ; the standard error is indicated by the clashed lines.
The slope of the regression line is -- 0.36, the standard error 0.23. At all concen-
trations of GABA tested, the rate-slowing effect is readily reversible by perfnsion
with drug-free saline (Fig. 1).
The strength of contraction of the heart also decreases with GABA perfusion
(Fig. 1). In Figure 3 the relationship between relative strength and concentration
FIGURE 2. Relation of relative heart rate to concentration of perfused GABA. Each point
represents a single perfusion. The solid line is the regression line determined by least squares
and the dashed lines are the standard error of the regression line.
GABA, ANTAGONISTS, AND LIMULUS HEART
385
1.0
T
H
a
0.4
0.2
10
S 50 100
CABA (M X I0~7)
250
500
FIGURE 3. Relation of relative contraction strength to concentration of GABA perfused
through isolated hearts. The solid line is the regression line determined by least squares and
the dashed lines represent the standard error of the regression line.
of perfused GABA is plotted for 19 perfusions in 13 different hearts. The slope
of the regression line in this case is only -- 0.17 compared to the slope of — 0.36 for
rate changes. Thus a concentration of GABA sufficient to reduce heart rate by
50% reduces contraction strength by less than 20%.
Although GABA reduces the rate at which rhythmic bursts of electrical activity
occur in isolated cardiac ganglia, it causes no readily apparent changes in the pat-
tern of the individual bursts. In Figure 4 the pattern of a typical burst of electrical
0.5 SEC
FIGURE 4. The pattern of electrical activity in an isolated ganglion. The upper trace
shows a representative burst before drug treatment ; the lower trace a representative burst after
bathing the ganglion for one minute in 1 X IQ'5 M GABA.
386
RALPH A. PAX AND RICHARD C. SANBORN
activity recorded from the fourth segment of an isolated cardiac ganglion in drug-
free saline is compared to a typical burst of electrical activity recorded from the
same segment of the same isolated cardiac ganglion after bathing in 1 X 10~5 M
GABA for one minute. At the time of recording the rate of rhythmic bursting has
been reduced by 50% but, contrary to the changes seen in the pattern of the burst
during stimulation of the inhibitor nerves, there is neither a decrease in the duration
of the burst nor a lesser number of discharges in a particular burst. Bathing the
ganglion for one minute in drug-free saline is sufficient to return the rate of rhythmic
bursting to the pre-treatment level.
1.0
0.8
06
bl
I
O.4
0.2
0.0
t
fPICROTOXIN
to
IS
20 23 30
TIME (MIN)
40
45
FIGURE 5. Effect of picrotoxin on cardioinhibitory nerves. Each point represents the mean
relative heart rate for four hearts during stimulation of the inhibitor nerves. Vertical lines
extend one standard deviation on either side of the mean. During the time between the two
arrows 100 ml. of 10"3 M picrotoxin were perfused.
L-glutamic acid at a concentration of 10~3 M perfused through the isolated
heart reduces the strength of contraction of the heart muscle to a barely detectable
level but does not change the heart rate. At 10~5 M it has no measurable effects on
rate or strength. In like manner carnitine (gamma-aminobutyric-beta-hydroxy-
betaine) perfused through the heart at 10~4 M causes a marked decrease in strength
of contraction but causes no measurable change in rate.
Picrotoxin
The ability of picrotoxin to block the action of the inhibitor nerves was tested
in four isolated hearts. In each experiment the inhibitor nerve was stimulated near
its junction with the cardiac ganglion in the fourth heart segment. Stimulation
was given for 40 seconds out of every five minutes. During the first four such five-
GABA, ANTAGONISTS, AND LIMULUS HEART
387
minute intervals drug-free saline was perfused. In the fifth and sixth intervals 100
ml. of 10~3 M picrotoxin were perfused and then during the next four five-minute
intervals drug-free saline was again perfused.
In each case picrotoxin alone caused an increase in heart rate, the mean rate
being 24.9 beats per minute before picrotoxin perfusion and 34.6 beats per minute
after picrotoxin perfusion. To compensate for this drug-induced rate increase, the
relative rates in the portion of the experiment when picrotoxin was used were
computed by comparing the ratio of the rate during stimulation to that obtained
immediately before stimulation. Both rates were thus measured in the presence of
the drug.
In each of the four hearts the inhibitor nerves were less effective during picro-
toxin perfusion. In two of these this decreased effectiveness preceded the increase
1.0
o.e
PICROTOXIN PICROTOXIN + OABA
s 10 is eo
TIME (MINI
FIGURE 6. Rate changes in a heart perfused with GABA alone (2 X 10~5 M) and with GABA
plus picrotoxin (1 X 10"3 M). See text for details.
in heart rate. The block of the inhibitor nerves, therefore, is not merely a reflection
of the increased heart rate caused by the picrotoxin. The mean relative rate ob-
tained by stimulation before treatment with the drug was 0.19 (SD = 0.15) i.e.,
stimulation caused an 81% decrease in rate, while after picrotoxin treatment the
mean relative rate was 0.76 (SD = 0.10). Thus, in the presence of the drug,
stimulation decreased the rate by only 24%. A "t" test for the difference between
these two means showed it to be significant (P > 0.99). In Figure 5 the mean
relative rate produced by stimulation of the inhibitor nerves in the four hearts
before, during and after perfusion with picrotoxin during each of the ten stimulation
periods is shown. The mean decrease in rate produced by stimulation of the in-
hibitor nerves in the four hearts before picrotoxin perfusion was 20.2 beats per
minute. During and after picrotoxin perfusion the decrease was 8.3 beats per
388
RALPH A. PAX AND RICHARD C. SANBORN
minute. We have no data concerning changes in contraction strength during
stimulation of inhibitor nerves while perfusing with picrotoxin.
The reduced effectiveness of the inhibitor nerves outlasts the perfusion with
picrotoxin. As can be seen from Figure 5 the mean relative rate obtained by
stimulation 20 minutes after the end of perfusion with picrotoxin was still 0.80
(SD = 0.09), 0.60 unit greater than the mean relative rate obtained before picro-
toxin treatment.
Since picrotoxin is effective in blocking the function of the cardioinhibitory
nerves of Linndus, its ability to antagonize the action of applied GABA was also
tested. Four isolated hearts were used in these experiments. Since from one
preparation to the next there is considerable variation in the response to a given
concentration of GABA, a control perfusion of 100 ml. of GABA was made for
I 0.6
o
p
o
•••••• *•••
GABA
PICROTOXIN
PICROTOXIN •»• OABA
10 13 20
TIME (MINI
FIGURE 7. Contraction strength changes in a heart perfused with GABA alone (2 X 10 5 M )
and with GABA plus picrotoxin (1 X lO"3 M). See text for details.
each heart prior to picrotoxin treatment. After the heart had recovered from the
GABA perfusion by perfusing for one-half hour with drug-free saline, treatment
with picrotoxin was begun. After perfusion with 100 ml. of 10~3 M picrotoxin in
saline, a second 100-ml. portion containing the same concentration of GABA as
that previously given was perfused. In this way GABA at concentrations of 5, 10
and 20 X 10~6 M was tested against picrotoxin at 10~3 M.
The response of the heart to GABA is not significantly altered by picrotoxin.
Figure 6 presents the results for one of the four hearts. In the example shown the
mean decrease in rate was 8.7 beats per minute during GABA perfusion prior to
picrotoxin treatment. During GABA perfusion after picrotoxin treatment the mean
decrease in rate was 7.7 beats per minute. Not only is the decrease in rate almost
identical in the presence or absence of picrotoxin, but the time course of the response
GABA, ANTAGONISTS, AND LIMULUS HEART 389
to GABA is essentially unaltered. Although there were differences between relative
rates obtained during GABA perfusion before and after picrotoxin treatment in
individual hearts, the mean relative rate for the four hearts during GABA perfusion
prior to picrotoxin treatment was 0.30, exactly the same as the mean value obtained
during GABA perfusion after picrotoxin treatment (Mean Difference — 0.00 ;
SD = 0.05).
In one of the hearts in which the interaction between GABA and picrotoxin was
tested, data about strength changes were also obtained (Fig. 7). During perfusion
with GABA alone the minimal relative contraction strength was 0.43. When
GABA and picrotoxin were perfused together it was 0.50. The time course of the
inhibition in both cases was approximately the same.
DISCUSSION
We have considered the evidence that GABA acts as a synaptic transmitter in
the cardioinhibitory pathway of Limiilus. It is worthwhile comparing our observa-
tions with those on other arthropod systems in which GABA is believed to be a
junctional transmitter.
In crustacean inhibitory motor neurons, GABA clearly appears to be the natural
transmitter. It duplicates the effects of activation of the inhibitory neurons on
muscle (Dudel, 1965; Takeuchi and Takeuchi, 1966), is present in the inhibitory
axons and the synthetic machinery is present in such axons (Kravitz et al., 1963).
Nearly as conclusive evidence exists that GABA is the natural transmitter for
cardioinhibition in crustaceans. While it has not been isolated from this site,
application to the ganglion cells of the crustacean heart has been shown to mimic, in
all respects, the action of the natural transmitter (Florey, 1957; Maynard, 1961).
On the other hand, although the crustacean stretch receptor has been shown to
be inhibited by GABA it does not appear to be the transmitter in this system
(Kuffler and Edwards, 1958; Edwards and Kuffler, 1959).
Unequivocal proof that a given compound is the endogenous transmitter at a
given junction is not easily obtained. Short of actual demonstration that the
supposed transmitter is liberated by activity in the presynaptic fibers and that it,
when applied in physiological concentrations, reproduces the conductance changes
which occur during synaptic transmission (Terzuolo and Edwards, 1962), some
doubt about the identity of the transmitter will exist. Because of the anatomical
arrangement at many junctions it is difficult, if not impossible, to produce such
direct evidence about the nature of the transmitter.
In view of this difficulty a number of other sets of criteria have been proposed
which do not rely on such direct evidence. One such set is that of McLennan
(1963) : (1) The substance occurs in presynaptic structures. (2) An enzymatic
mechanism for synthesis of the substance is present. (3) An enzyme system for
inactivation of the substance is present. (4) Application of the substance mimics
stimulation. (5) During stimulation the substance is detectable in perfusates. (6)
Pharmacological agents which interfere with operation of the neuron similarly affect
the action of the substance artificially applied.
If a given chemical is to be seriously considered to be the endogenous trans-
mitter at a given junction then it must meet each of these criteria. Conversely, if
a given chemical does not meet one or more of these criteria, it is doubtful that it is
390 RALPH A. PAX AND RICHARD C. SANBORN
the endogenous transmitter at that junction. Our evidence shows that GABA fails
to meet two requirements. First, picrotoxin which effectively blocks the action of
the inhibitor nerves is without effect upon the slowing of the heart rate caused by
GABA (Criterion 6). Since the exact site of action of picrotoxin at the inhibitory
junction is unknown, failure to meet this requirement alone is not sufficient to
eliminate GABA as a possible transmitter at this junction. However GABA also
fails to meet a second requirement, namely that it mimic stimulation of the inhibitor
nerve (Criterion 4). Although stimulation and GABA both slow the heart rate,
they have quite different effects upon the pattern of neural activity in the cardiac
ganglion. Stimulation of the inhibitor nerves decreases the number of units dis-
charging in the cardiac ganglion during each burst of electrical activity, as well
as the total duration of each burst (Heinbecker, 1933). GABA produces neither
of these changes in the pattern of the burst.
We believe, therefore, that even if GABA were to meet some of the other
criteria listed above, it could not be seriously considered as a natural transmitter
in the Limulus cardioinhibitory pathway.
The authors wish to express their thanks to the National Science Foundation
for support for certain phases of these studies.
SUMMARY
1. GABA (5 X 10-7 to 5 X 1Q-5 M) perfused through the isolated Limulus
heart mimics stimulation of the cardioinhibitory nerves by decreasing rate and
strength of beating of the heart.
2. GABA, unlike activity in the cardioinhibitory nerves, decreases neither the
number of units discharging nor the total duration of each burst of electrical activity
in the cardiac ganglion.
3. Picrotoxin (1 X 10~3 M) blocks the function of the cardioinhibitory nerves.
4. Picrotoxin (1 X 1O3 M) blocks neither the rate nor the strength-decreasing
effects of applied GABA.
5. Since GABA does not mimic the action of the inhibitor nerves and its action
is not blocked by an agent blocking the function of the inhibitor nerves, we believe
it is probable that GABA is not a transmitter in the Limulus cardioinhibitory
pathway.
LITERATURE CITED
BURGEN, A. S. V., AND S. W. KuFFLER, 1957. Inhibition of the cardiac ganglion of Limulus
polyphetnus by 5-hydroxytryptamine. Biol. Bull., 113: 336.
CARLSON, A. J., 1905. The nature of cardiac inhibition with special reference to the heart of
Limulus. Amcr. J. Physiol, 13: 217-240.
CARLSON, A. J., 1907. On the mechanism of the stimulating action of tension on the heart.
Amer. J. Physiol., 18: 149-155.
CARLSON, A. J., 1909. Vergleichende Physiologic der Herznerven und der Herzganglien bei
den Wirbellosen. Ergcbn. Physiol., 8: 371-462.
CHAD, I., 1933. Action of electrolytes on the dorsal median nerve cord of the Limulus heart.
Biol. Bull., 64: 358-382.
DUDEL, J., 1965. Presynaptic and postsynaptic effects of inhibitory drugs on crayfish neuro-
muscular junction. Pflilgers Archil'., 283: 104-118.
GABA, ANTAGONISTS, AND LIMULUS HEART 391
EDWARDS, E., AND S. W. KUFFLER, 1959. The blocking effect of 7-aminobutyric acid (GABA)
and the action of related compounds on single nerve cells. /. Neurochcm., 4: 19-30.
FLOREY, E., 1957. Further evidence for the transmitter function of factor I. Naturwissenschaften,
44: 424-425.
HEINBECKER, P., 1933. The heart and median cardiac nerve of Liimilus pol\phemus. Amer.
J.Physiol., 103: 104-120.
KERKUT, G. A., AND M. A. PRICE, 1964. Chromatographic separation of cardio-accelerators
(6-HT and a mucopeptide) from Carcinns heart. Comp. Biochem. Physiol., 11: 45-52.
KRAVITZ, E. A., S. W. KUFFLER AND D. D. POTTER, 1963. Gamma aminobutyric acid and other
blocking compounds in Crustacea III. Their relative concentrations in separated motor
and inhibitory axons. /. Neurophysiol., 26: 739-751.
KUFFLER, S. W., AND C. EDWARDS, 1958. Mechanism of gamma aminobutyric acid (GABA)
action and its relation to synaptic inhibition. /. Neurophysiol., 21: 589-610.
MAYNARD, D. M., 1961. Cardiac inhibition in a decapod crustacean. In: Nervous Inhibition,
E. Florey, Ed., Pergamon Press, New York, pp. 148-178.
MCLENNAN, H., 1963. Synaptic Transmission. W. B. Saunders Co., Philadelphia.
PAX, R. A., AND R. C. SANBORN, 1964. Cardioregulation in Limulus. I. Physiology of inhibitor
nerves. Biol. Bull, 126: 133-141.
TAKEUCHI, A., AND N. TAKEUCHI, 1966. On the permeability of the presynaptic terminal of
the crayfish neuromuscular junction during synaptic inhibition and the action of
7-aminobutyric acid. J. Physiol., 183 : 433-449.
TERZUOLO, C. A., AND C. EDWARDS, 1962. Excitation and synaptic transmission. Ann. Rev.
Physiol., 24:325-356.
VAN DER KLOOT, W. G., J. ROBBINS AND I. M. COOKE, 1958. Blocking by picrotoxin of
peripheral inhibition in crayfish. Science, 127: 521-522.
CARDIOREGULATION IN LIMULUS. III. INHIBITION BY
5-HYDROXYTRYPTAMINE AND ANTAGONISM BY
BROMLYSERGIC ACID DIETHYLAMIDE
AND PICROTOXIN
RALPH A. PAX i, 2 AND RICHARD C. SANBORN
Department of Biological Sciences, Purdue University, Lafayette, Indiana 47907
Cardioinhibition in Limulus appears to be chemically mediated. The decrease
in heart rate resulting from stimulation of the inhibitor nerves is not tightly coupled
to the stimulation, a time lag in the response occurring both at the beginning and
at the end of the stimulation periods (Carlson, 1905 ; Pax and Sanborn, 1964).
The nature of the chemical mediator of inhibition is not known. A variety of
pharmacological agents have been tested since the neurogenic nature of the heart
beat was first shown by Carlson (1904). Of these, only three have been reported
to cause a decrease in heart rate: ergot (Carlson, 1906), 5-hydroxytryptamine
(5-HT) and gamma-aminobutyric acid (GABA) (Burgen and Kuffler, 1957).
A study of GABA as the possible inhibitory neurotransmitter in the Limulus
heart has previously been reported (Pax and Sanborn, 1967). Although this
compound decreases rate and strength of beating when applied artificially, it does
not decrease the number of units discharging or the total duration of each burst of
electrical activity in the cardiac ganglion as does stimulation of the inhibitor nerves.
Moreover, picrotoxin, though effective in blocking the function of the inhibitor
nerves, is not an effective antagonist to GABA activity. It appears, therefore, that
GABA is not involved as the inhibitory neurotransmitter in the Limulus heart.
5-HT, like GABA, is found in a wide variety of animals (Welsh and Moorhead,
1960). It has been shown to have physiological significance in such diverse animal
groups as flatworms and vertebrates (Mansour et al., 1960). In contrast to its
reported inhibitory action on the Limulus heart (Burgen and Kuffler, 1957) it has
an excitatory effect on the crustacean neurogenic heart (Florey and Florey, 1954;
Maynard and Welsh, 1959; Kerkut and Price, 1964; Cooke, 1966). We report
here results of experiments exploring more fully the possibility that 5-HT or a
5-HT-like compound is the cardioinhibitory transmitter in the neurogenic Limulus
heart.
MATERIALS AND METHODS
Materials and methods are as previously described (Pax and Sanborn, 1967).
RESULTS
5-Hydroxytryptamine
Perfusion of 5-HT through the isolated heart results in a decrease in heart
rate. A typical result of 5-HT perfusion is shown in Figure 1. In Figure 2 the
1 Predoctoral Fellow of the National Institute of General Medical Sciences, N.I.H.
2 Present Address : Department of Zoology, Michigan State University, East Lansing,
Michigan.
392
5-HT INHIBITION OF LIMULUS HEART
393
2 MIN
FIGURE 1. Response of the isolated heart to perfusion of 5-HT. During the time between
the arrows 100 ml. of 5 X 10~a M 5-HT were perfused through the heart.
relationship between concentration of 5-HT perfused and relative heart rate is
plotted for 18 perfusions of 5-HT in 12 different hearts. The solid line on the
graph is the regression line for these data as determined by the method of least
squares. The standard error of this line is indicated by the dashed lines on either
side of the regression line. The slope of this regression line is —0.34; the standard
error 0.17. The threshold for rate changes, as determined by solving the equation
e 9 50 10
83 JO 100
»-MT ((j X 10-' 1
-7,
FIGURE 2. Relation of relative heart rate to concentration of 5-HT perfused through the
isolated heart. Each point represents a single perfusion. The solid line is the regression line
determined by the method of least squares and the dashed line is the standard error of the
regression line.
394
RALPH A. PAX AND RICHARD C. SANBORN
for the regression line, is 4.9 X lO8 M while at 4.1 X 1O5 M 5-HT a relative
heart rate of zero would he expected.
The strength of heart heat also decreases when 5-HT is perfused through the
isolated heart (Fig. 1). In Figure 3 the relationship between 5-HT concentration
and relative contraction strength is plotted for 14 perfusions of 5-HT in nine differ-
ent hearts. The calculated threshold concentration for strength changes is 5.6
X 10~8 M, about the same as that calculated for rate changes, but the calculated
regression lines for rate and strength changes are not parallel (slope = —0.34 for
rate, —0.31 for strength).
0.2
\0
\
1.0 2.9 $.0 10 23 60 100
S-HT (M X I0'r)
FIGURE 3. Relation of relative contraction strength to concentration of 5-HT perfused
through the isolated heart. Each point represents a single perfusion. The solid line is the
regression line determined by the method of least squares and the dashed lines are the standard
error of the regression line.
Both of the above effects of 5-HT are readily reversible. Perfusion with drug-
free saline for five minutes following drug treatment is usually sufficient to bring the
rate and strength of beating within 10% of their pre-treatment levels.
Neither 5-hydroxytryptophan — the precursor of 5-HT — nor 5-hydroxyindole-
acetic acid — its major metabolite — at 10~4 M had any detectable effects on rate or
strength of beating of the isolated heart.
Electrical activity of the isolated cardiac ganglion is also affected by 5-HT treat-
ment. The rate of rhythmic discharges decreases. The number of units discharging
in each burst is reduced and the total duration of each burst is lessened. The
pattern of a typical burst of electrical activity recorded from the fourth segment of
the isolated cardiac ganglion before treatment and after treatment with 1 X 10~6 M
5-HT is shown in Figure 4. In this experiment the relative heart rate during
5-HT INHIBITION OF LIMULUS HEART
395
perfusion of 5-HT was 0.51. The changes in the pattern of electrical activity in
the ganglion during a particular burst are readily apparent.
After treatment of isolated ganglia with 5-HT, one minute of bathing in drug-
free saline returns the rate to the pre-treatment level.
Bromlysergic acid diethylamide
Bromlysergic acid diethylamide (BOL) is a potent and specific antagonist of
5-HT in other animals (Gyermek, 1961). Since 5-HT appears to mimic the action
of the inhibitor nerves of the Limulns heart we have studied the interaction between
BOL and the inhibitor nerves. The function of the inhibitor nerves was tested in
four animals before, during, and after perfusion with 1.6 X 10~5 M BOL. These
experiments w7ere performed in a manner parallel to that used in testing the inter-
action of the inhibitor nerves and picrotoxin (Pax and Sanborn, 1967). Nerves
were stimulated for 20 seconds out of every five minutes. During the first five
I
0.1 MV
j* | fH f fttj|l '((""([j
0.5 SEC
il'Li {LI.-. ..._ 1 MI
ifcHr
(!
FIGURE 4. Changes in the pattern of electrical activity in the isolated ganglion resulting
from 5-HT treatment. The upper trace is a representative burst before drug treatment, the
lower trace a representative burst after bathing the ganglion for one minute in 1 X 10"* M 5-HT.
five-minute stimulation intervals, drug-free saline was perfused. During the next
two five-minute intervals 100 ml. of BOL were perfused and during the last six
five-minute intervals drug-free saline was again perfused.
BOL (1.6 X 10'5 M) alone causes a slight increase in heart rate, the mean rate
for 15 different hearts being 30.3 beats per minute before BOL treatment and 31.8
beats per minute after BOL treatment. The relative rate for each of the 13
stimulation periods was computed by taking the ratio of the rate during the stim-
ulation period to the rate just previous to that same stimulation period.
BOL is an effective antagonist of inhibitor nerve action in the Limulus heart.
The mean relative rate during stimulation of the inhibitor nerves for each of the 13
stimulation periods is shown in Figure 5. Stimulation of the inhibitor nerves
before BOL treatment resulted in a mean decrease in rate of 19.8 beats per minute.
The relative rate was 0.28 (SD = 0.11), i.e., stimulation reduced the rate by 72%.
After BOL treatment the decrease in rate was 12.0 beats per minute and the
relative rate was 0.66 (SD = 0.12), i.e., stimulation reduced the rate by only 34%.
396
RALPH A. PAX AND RICHARD C. SANBORN
A "t" test for the difference between the two relative rates showed the inhibitor
nerves significantly less effective in decreasing heart rate after BOL treatment
(P > 0.95). Function of the inhibitor nerves does not begin to return to the pre-
BOL perfusion level even after perfusion with drug-free saline for as long as 30
minutes (Fig. 5).
Since BOL blocks the function of the cardioinhibitory nerves in Limulus, it
should also antagonize the action of artificially applied 5-HT, if 5-HT is acting at
a junction in the cardioinhibitory pathway. The ability of BOL to antagonize the
action of 5-HT was tested on four isolated hearts.
These experiments were performed in a manner parallel to our experiments
testing the interaction of GABA and picrotoxin. One hundred ml. of saline con-
08
06
<
<T
0.4
0.2
BOL
10
2.0
30
TIME (WIN)
40
30
60
FIGURE 5. Effect of BOL on the function of the cardioinhibitory nerves. Each point
represents the mean relative heart rate for four hearts during stimulation of the inhibitor nerves.
Vertical lines extend one standard deviation on either side of the mean. During the time
between the two arrows 100 ml. of 1.6 X 10~5 M BOL were perfused.
taining 5-HT were initially perfused through each heart to calibrate its response.
After one-half hour of perfusion with drug-free saline to eliminate the effects of
the 5-HT, 100 ml. of 1.6 X 1Q-5 M BOL were perfused. This perfusion was im-
mediately followed by perfusion of 100 ml. of 1.6 X 10~5 M BOL to which had
been added the same concentration of 5-HT as that previously given during the
calibration perfusion. 5-HT at concentrations of 1 and 5 X 10~6 M was tested in
this way against BOL at 1.6 X lO'5 M.
Our experiments show that this concentration of BOL is an effective antagonist
of the rate-decreasing effects of 5-HT. The mean relative rate with 5-HT perfusion
prior to BOL treatment was 0.57 (mean decrease in rate, 12.9 beats per minute)
while after BOL treatment it was 0.93 (mean decrease in rate, 2.1 beats per
5-HT INHIBITION OF LIMULUS HEART
397
1.0
o.e
04
06
5-MT
PICROTOXIN PICROTOXIN + S-HT
10 15 20
TIME (MiN )
FIGURE 6. Rate changes in a heart perfused with 5 X 10 8 M 5-HT alone and with 5-HT
plus 1.6 X 10-5 M BOL. See text for details.
minute). A "t" test for the difference between the two relative rates showed 5-HT
significantly less effective in reducing heart rate after BOL treatment (P > 0.99).
The results of a typical experiment are shown in Figure 6.
BOL (1.6 X 10~5 M) is also an effective antagonist of the strength-decreasing
effects of artificially applied 5-HT. The mean relative strength with 5-HT per-
fusion prior to BOL treatment was 0.41 while after BOL treatment it was 1.04. A
0.6
0.4
I.O
0.8
. . * . .
*
B
BOL + 5-HT
10 15 20
TIME I M I N J
25
FIGURE 7. Changes in contraction strength with perfusion of 5 X 10~6 M 5-HT alone and
with 5-HT plus 1.6 X 10~5 M BOL. See text for details.
398
RALPH A. PAX AND RICHARD C. SANBORN
"t" test for the difference between these means showed it significant (P > 0.99).
Results of a typical experiment are shown in Figure 7.
Pier o toxin
Picrotoxin has previously been shown to be effective in blocking the action of
the inhibitor nerves in Limnlus (Pax and Sanborn, 1967). Since it blocks the
inhibitor nerves it should also antagonize the action of applied 5-HT, if 5-HT is
acting as a neurotransmitter in the cardioinhibitory pathway. We have therefore
tested the ability of picrotoxin to block the action of applied 5-HT on four isolated
hearts.
The experiments were performed in a manner parallel to that described for
testing the interaction of 5-HT and BOL. One hundred ml. of saline containing
5-HT were initially perfused through each heart to determine a control response.
After one-half hour of perfusion with drug-free saline to eliminate the effects of
1.0
• . «•
5-HT
BOL
BOL * S-MT
13 80
TIME (HIN )
FIGURE 8. Rate changes in a heart perfused with 5 X 10~8 M 5-HT alone and with 5-HT
plus 1 X 10~3 M picrotoxin. See text for details.
the 5-HT, 100 ml. of 10~3 M picrotoxin were perfused followed immediately by 100
ml. of 10"3 M picrotoxin to which had been added the same concentration of 5-HT
as that given during the control perfusion. 5-HT at concentrations of 1, 5, and
10 X 10~6 M was tested in this way against picrotoxin at 10~3 M.
As with BOL, there is antagonism between 5-HT and picrotoxin. The mean
relative rate with 5-HT perfusion prior to picrotoxin treatment was 0.34 (mean
decrease in rate, 18.2 beats per minute) while after picrotoxin treatment it was
0.75 (mean decrease in rate, 11.5 beats per minute). A "t" test for the difference
between the two relative rates showed the relative rate to be significantly higher
after picrotoxin treatment (P > 0.99). The results of a typical experiment are
presented in Figure 8.
By contrast, the effects upon relative contraction strength are quite different.
In this variable, picrotoxin and 5-HT show synergism rather than antagonism.
S-HT INHIBITION OF LIMULUS HEART
399
Though no measurable change in contraction strength is brought about by perfusion
of picrotoxin alone, when picrotoxin is perfused with 5-HT a greater decrease in
contraction strength occurs than when 5-HT alone is perfused. The mean relative
strength of four hearts with 5-HT perfusion prior to picrotoxin treatment was 0.51
while after picrotoxin treatment it was 0.26. A "t" test for the difference between
these means showed the relative contraction strength significantly lower after
picrotoxin treatment than before (P > 0.95). The results of a typical experiment
are presented in Figure 9.
0 S
0.6
0.6
PICROTOXIN PICOOTOXIN 4 5-HT
10 1} 20 23
Tl ME I MIN )
FIGURE 9. Changes in contraction strength with perfusion of 5 X 10 * M 5-HT alone and
with S-HT plus 1 X 10~3 M picrotoxin. See text for details.
BOL and GAB A
BOL blocks the function of the cardioinhibitory nerves. GABA, although it
decreases the rate and strength of beating of the intact heart, does not alter the
pattern of electrical activity in the cardiac ganglion. Thus, GABA does not mimic
stimulation of the inhibitor nerves and appears to produce its effects at some site
other than the cardioinhibitory pathway in the Lhnulus heart (Pax and Sanborn,
1967). If BOL is blocking the action of the inhibitor nerves by acting specifically
at a junction in the cardioinhibitory pathway, and GABA is acting at some site
other than this, then there should be no interaction between simultaneously applied
BOL and GABA.
We have tested this in eight isolated hearts. One hundred nil. of saline con-
taining GABA were initially perfused through each heart. After one-half hour of
perfusion with drug-free saline, 100 ml. of 1.6 X 10~5 M BOL were perfused. This
400
RALPH A. PAX AND RICHARD C. SANBORN
was followed immediately by 100 ml. of 1.6 X 10~5 M BOL to which had been
added the concentration of GABA previously given during the control perfusion.
GAB A at concentrations of 5, 10, and 50 X 10~6 M was tested in this way against
BOL at 1.6 X 10-5 M.
There is no apparent interaction between BOL and GABA so far as rate is
concerned. The mean relative rate resulting from GABA perfusion prior to BOL
treatment was 0.34 (mean decrease in rate, 19.7 beats per minute) while after BOL
treatment it was 0.44 (mean decrease in rate, 15.4 beats per minute). In four out
of eight hearts tested in this manner the relative rate resulting from GABA
perfusion was higher after BOL treatment than before. In the other four it was
lower. The mean difference between the relative rate prior to, and following BOL
treatment was 0.10 (SD = 0.18). A "t" test for the difference between the relative
rate before BOL treatment and after BOL treatment showed it to be non-significant
(P<0.90).
TABLE I
Summary of the responses of the Limulus heart to treatment
•with various inhibitors and antagonists
Inhibitor
Antagonist
5-HT
BOL
5-HT
Picrotoxin
GABA
BOL
Mean relative rate
No. animals tested
Inhibitor alone
Inhibitor with antagonist
Difference ± 1 SD
Mean relative contraction strength
No. animals tested
Inhibitor alone
Inhibitor with antagonist
Difference ± 1 SD
4
0.57
0.93
0.36 ± 0.09
P < 0.01
4
0.41
1.04
0.63 ± 0.21
P < 0.01
4
0.34
0.75
0.41 ± 0.18
P < 0.01
4
0.51
0.26
-0.25 ± 0.11
0.01 < P < 0.05
8
0.34
0.44
0.10 ± 0.19
0.10 < P < 0.20
5
0.67
0.65
-0.02 ±0.07
0.60 < P < 0.70
Similarly, there is no apparent interaction between BOL and GABA with
respect to strength of contraction. For five hearts the relative contraction strength
resulting from GABA perfusion prior to BOL treatment was 0.67 while after BOL
treatment it was 0.65. A "t" test for the difference between these means showed
it not significant (P < 0.90).
The results of the various treatments are summarized in Table I.
DISCUSSION
In the first paper of this series (Pax and Sanborn, 1964), we presented our
reasons for believing that a chemical transmitter was involved in the inhibition of
the neurogenic Limulus heart. At this point it is appropriate to examine the known
and possible components of the Limulus cardioinhibitory system in order to visualize
the sites at which chemical transmitters might operate.
5-HT INHIBITION OF LIMULUS HEART 401
Both the decapod crustacean cardiac ganglion and the Lint it! us cardiac ganglion
possess two cell types (Heinhecker, 1936). The primary difference between the
Liiiiiilns heart and the crustacean heart appears to he in the number of cells involved
and it would seem reasonable to assume that the mechanism by which the rhythmic
discharge is originated is common to both hearts ( Maynard, 1955).
In the crustacean cardiac ganglion the burst is usually initiated by the smallei
cells, the pacemakers. The larger cells are the major motor neurons (followers)
and appear to be only relays which increase the number of impulses. Feedback, if
any, from the followers to the pacemakers is small since only long subthreshold
current pulses to the followers or a long series of follower cell impulses are necessary
for modification of the rhythm of the pacemaker cells (Otani and Bullock, 1959).
The inhibitor fibers make connections with both the pacemaker and the follower
cells (Terzuolo and Bullock, 1958).
If the Li in 11 1 us cardiac ganglion has an arrangement of functional units similar
to the decapod heart, then we may diagram the inhibitory pathway as in Figure 10.
Spontaneous rhythmic activity in the pacemaker cell ( P ) produces postsynaptic
FIGURE 10. Diagram showing possible organization of the nerve cells and the cardioinhibitory
connections in the cardiac ganglion of Limulus. Symbols explained in the text.
potentials in the follower cell ( F ) . These postsynaptic potentials result in propa-
gated action potentials which produce contraction of the myocardium (M). A
block in transmission across the neuromuscular junction (C) would result in only
a decreased strength of contraction. A block in transmission at junction "B"
would give the same results.
Activity in the inhibitor nerve through its action at junction "A" would pro-
duce a decrease in the rate of spontaneous bursting in cell "P" and thus cause a
decrease in heart beat rate. Some lesser effect on contraction strength might occur
if activity in the inhibitor also lessens the number of discharges in the pacemaker
during any particular burst of activity. At junction "D," activity in the inhibitor
nerve would produce a decrease in contraction strength by reducing the number
of action potentials in the follower cell.
Turning now to the results of the experiments described here we find that 5-HT
decreases both rate and strength of beating. This could be due to activation of
junction "A" alone or of both junction "A" and "D." BOL blocks the action of
the inhibitor nerves but does not otherwise disrupt heart function and thus probably
acts at junction "A" alone or at both "A" and "D." Since this compound also
402 RALPH A. PAX AND RICHARD C. SANBORN
blocks the action of applied 5-HT it appears probable tbat 5-HT acts at these
same junctions.
Picrotoxin also blocks the action of the inhibitor nerves without otherwise
markedly disrupting heart function. Thus it probably also acts at junction "A"
or both "A" and "D." The rate-decreasing action of 5-HT is blocked by picro-
toxin so again it would appear that 5-HT is acting at these same junctions. In
contrast to the antagonism shown between BOL and 5-HT as far as strength-
changing abilities are concerned, picrotoxin enhances the strength-decreasing ability
of 5-HT. Such a pattern of responses could occur if junction "D" possesses
pharmacological properties which are slightly different from those of junction "A."
GABA interacts with neither picrotoxin nor BOL. Thus it appears to act at
neither junction "A" nor "D." It probably does not act at junction "B" or "C"
since the rate-reducing effects of GABA are produced in the isolated ganglion and
no change in burst parameters is noted with application of GABA. At this time
we have no way of assessing the significance of feedback from the followers to the
pacemakers (dashed line and junction "E" in Figure 10). Perhaps the major
site of action of GABA is in this pathway.
Whether the endogenous inhibitory transmitter of the heart of Lumdns is 5-HT
or some related compound is open to question. Cogeners of 5-HT such as 5,6-
dihydroxytryptamine, 6-hydroxytryptamine, or other substituted hydroxytrypta-
mines have not been tested and may be as potent as 5-HT. In the crustacean
heart 5-HT, 5,6-dihydroxytryptamine and 6-hydroxytryptamine are all potent
cardiotropic agents and all have been detected in tissue extracts (Carlisle, 1956;
Maynard and Welsh, 1959; Kerkut and Price, 1964; Belamarich and Terwilliger,
1966).
We wish to thank the National Science Foundation for supporting some aspects
of these studies. Professor Tom S. Miya of the Purdue Department of Pharma-
cology has been generous with advice and contributed the BOL used in these studies.
SUMMARY
1. Heart rate in Liinuhts is slowed by 5-hydroxytryptamine (5-HT). The
threshold for this inhibition is 4.9 X 1O8 M.
2. The strength of beat is also reduced in 5-HT solutions. The calculated
threshold for this effect is 5.6 X 1O8 M.
3. Both of these effects are readily reversible.
4. Neither 5-hydroxytryptophan (10~* M} or 5-hydroxyindole acetic acid
(10~* M) have any detectable effects on rate or strength of beating.
5. Applied to the isolated cardiac ganglion, 5-hydroxytryptamine (10~6 M)
decreases the rate of rhythmic discharge, reduces the number of neurons discharg-
ing in each burst, and lessens the duration of each burst. All of these effects are
also reversible.
6. Bromlysergic acid diethylamide (BOL), 1.6 X 1O5 M, decreases the ability
of the cardioinhibitory nerves to influence heart rate.
7. BOL prevents the rate and strength changes engendered by exogenous 5-HT
applied to the isolated heart.
5-HT INHIBITION OF LIMULUS HEART 403
8. Picrotoxin antagonizes the decrease in heart rate produced by application of
5-HT, but synergizes with 5-HT with respect to its strength-decreasing ability.
9. No interaction between BOL and y-aminobutyric acid (GAB A) could be
demonstrated.
LITERATURE CITED
BELAMARICH, F. A., AND R. TERWILLIGER, 1966. Isolation and identification of cardio-excitor
hormone from the pericardial organs of Cancer borcalis. Amcr. ZooL, 6: 101-106.
BURGEN, A. S. V., AND S. W. KUFFLER, 1957. Inhibition of the cardiac ganglion of Litmdns
Polyphemus by S-hydroxytryptamine. Biol. Bull., 113: 336.
CARLISLE, D. G., 1956. An indole-alkylamine regulating heart beat in Crustacea. Biochem. J.,
63: 32-33 P.
CARLSON, A. J., 1904. The nervous origin of the heart-beat in Limulus and the nervous nature
of co-ordination in the heart. Amer. J. Physiol., 12: 67-74.
CARLSON, A. T., 1905. The nature of cardiac inhibition with special reference to the heart of
Liinuhis. Amcr. J. Physiol., 13: 217-240.
CARLSON, A. J., 1906. On the point of action of drugs on the heart with special reference to the
heart of Limitlus. Amcr. J. Physiol., 17: 177-210.
COOKE, I. M., 1966. The site of action of pericardial organ extract and 5-hydroxytryptamine in
the decapod crustacean heart. Amcr. ZooL, 6: 107-122.
FLOREY, E., AND E. FLOREY, 1954. Uber die mogliche Bedeutung von Enteramin (5-oxy-Trypta-
min) als nervoser Aktions-substanz bei Cephalopoden und dekapoden Crustacean. Z.
Nahirforsch.,9: 58-68.
GVERMEK, L., 1961. 5-Hydroxytryptamine antagonists. Pharmacol. Reviews, 13: 399-439.
HEINBECKER, P., 1936. Potential analysis of a pacemaker mechanism in Limitlus polyphemus.
Amcr. J. Physiol., 117: 686-700.
KERKUT, G. A., AND M. A. PRICE, 1964. Chromatographic separation of cardioaccelerators
(6-HT and a mucopeptide) from Carcinus heart. Comp. Biochem. Physiol., 11: 45-52.
MANSOUR, T. E., E. W. SUTHERLAND, T. W. RALL AND W. BUEDING, 1960. The effect of
serotonin (5-hydroxytryptamine) on the formation of adenosine 3', 5'-phosphate by
tissue particles from the liver fluke, Fasciola hcpaiica. J. Biol. Chcm., 235: 466-470.
MAYNARD, D. M., 1955. Activity in a crustacean ganglion. II. Pattern and interaction in burst
formation. Biol. Bull., 109: 420-436.
MAYNARD, D. M., AND J. WELSH, 1959. Neurohormones of the pericardial organs of Brachy-
uran Crustacea. /. Physiol., 149: 215-227.
OTANI, T., AND T. H. BULLOCK, 1959. Effects of presetting the membrane potential of the
soma of spontaneous and integrating ganglion cells. Physiol. ZooL, 32: 104-114.
PAX, R. A., AND R. C. SANBORN, 1964. Cardioregulation in Limulus. I. Physiology of in-
hibitor nerves. Biol. Bull, 126: 133-141.
PAX, R. A., AND R. C. SANBORN, 1967. Cardioregulation in Limulus. II. Gamma aminobutyric
acid, antagonists and inhibitor nerves. Biol. Bull., 132 : 381-391.
TERZUOLO, C. A., AND T. H. BULLOCK, 1958. Acceleration and inhibition in crustacean ganglion
cells. Arch. Hal. Biol., 96: 117-134.
WELSH, J. H., AND M. MOORHEAD, 1960. The quantitative distribution of 5-hydroxytryptamine
in the invertebrates, especially their nervous systems. /. Ncurochem., 6: 146-169.
SEMINAL LOSS IN REPEATEDLY MATED FEMALE
AEDES AEGYPTI *
ANDREW SPIELMAN, SR. M. G. LEAHY AND VALERIE SKAFF
Department of Tropical Public Health, Harvard School of Public Health,
Boston, Massachusetts 02115
Critical data are not available regarding the frequency of insemination of indi-
vidual female Acdes aegypti. Males may inseminate as many as 6 or more females
(Roth, 1948; Jones and Wheeler, 1965), and various authors have noted that
females may mate several times. Subsequent matings, however, are of shorter
duration than the first (Spielman, 1964). In a brief abstract, Vandehey and Craig
(1958) indicated that multiple insemination may occasionally occur in caged
populations.
Multiple matings in certain other mosquitoes have been studied more com-
pletely. Female Cnlcx pipiens appeared to utilize sperm from only one of two
genetically marked males with which they were confined (Kitzmiller and Laven,
1958; Spielman, 1956). On the other hand. Anopheles gouibiae (Goma, 1963)
and A. quadrimaculatus (French and Kitzmiller, 1963) occasionally accepted sperm
from more than one male. The significance of these findings is limited in that
caged populations may be abnormal in their mating behavior. Gillies' (1956) ob-
servation of multiple masses of semen in the genital atria of wild-caught A. gambiae
females may indicate that multiple insemination occurs in nature.
The objective of the present study was to determine whether more than one
semen mass is accepted and retained by female A. acgypti.
MATERIALS AND METHODS
A strain of Acdcs aegypti obtained from Grand Bahama Island in 1965 was
used. In addition, one experiment employed males of a genetically marked strain
(Gold Mesonotum ) obtained from Dr. George B. Craig, Jr. Larvae were reared
in tap water at 22 ± 1 ° C. and fed Purina Rabbit Chow Pellets. Length of day
was maintained at 16 hours and mosquitoes were manipulated at about the middle
of the day.
LTnless otherwise noted, males were 2-5 days of age when first mated. Mating
of free-flying pairs occurred in glass lantern chimneys (18 X 10 cm. ) at 22 ± 1° C.
and 76 ± 2% R.H. The chimneys were rotated on their sides in order to induce
continuous flight by the mosquitoes within. After mating began, rotation was
stopped and duration of genital contact was noted. Mosquitoes were discarded if
no copulation occurred within 2 minutes. To mate tethered specimens the female
was etherized and glued to a slide. Males were anesthetized with gas (99%
nitrogen and \% hydrogen), decapitated, and held in vacuum forceps. Genitalia
1 This study was supported in part by U. S. Public Health Service Grant AI 00046 from the
National Institute of Allergy and Infectious Diseases.
404
SEMINAL LOSS IN FEMALE AEDES
405
of these males were then brushed against those of the females until copulation
occurred.
Mosquitoes were removed from chimneys with a breath-operated aspirator tube;
those to be frozen were then blown into test tubes immersed in a mixture of alcohol
and solid CO.,. Techniques of dissection, preparation of whole mounts and sec-
tioning (employing Newcomer's fixative and hematoxylin-azure II-eosin stain)
followed methods described by Spielman (1964). Techniques for transplanting
tissue into mosquitoes have been described by Leahy and Craig (1965).
RESULTS
1. Filling of the copulatory bursa
The rate of inflation of the copulatory bursae of mating females was determined.
Virgin pairs which were permitted to mate in lantern chimneys were removed and
quick-frozen at •- 70° C. at intervals after copulation began. Females were fixed
while frozen, dissected, and mounted to permit a lateral view of the bursa. The
outline of each bursa was sketched with a camera lucida (X 225) and the area
of the sketch measured with a polar planimeter. (Measurement of the lateral
profile of the bursa provided an adequate measure of degree of distention because
inflation did not markedly affect the transverse dimension.) After 4 seconds of
mating, bursae were approximately as distended as those in females that were
allowed to mate without interruption ( Fig. 1 ) .
The duration of copulation necessary for the retention of semen after mating
was studied in females whose mating was interrupted mechanically. This is in
cr>
tr
3
ffi
>-
(T
o
Q.
O
O
u.
o
0.6
< 0.5
III
K
-I Q4
0.3
0.2
0.1
I I
123456 8
(COPULATION INTERRUPTED BY FREEZING)
10
12.0- 42.9
(NOT
INTERRUPTED)
SECONDS FROM BEGINNING OF COPULATION UNTIL FREEZING
FIGURE 1. Rate of inflation of the copulatory bursa during mating. When copulation was not
interrupted, females were frozen immediately after withdrawal of the male.
406 A. SPIELMAN, SR. M. G. LEAHY AND VALERIE SKAFF
contrast to interruption by quick-freezing as practiced in the previous experiment.
Five minutes after separation from the males, the females were chloroformed and
dissected. Those that had mated for 6 seconds contained semen (Table I).
Subsequently, all experiments involving successive matings were done with
females known to have mated for at least 6 seconds. Duration of copulation among
virgin specimens generally exceeded 10 seconds; the mean for 128 pairs was 17.6
seconds (Table II).
2. Behavior during successive matings
The mating behavior of virgin and non-virgin females was compared. Single
virgin females were placed with two males in lantern chimneys. Females were
removed after copulation and held for varying periods of time before re-exposure
to males. Females appeared to mate somewhat more readily when first exposed
to males than during second exposure (Table II). Thus, nearly 90% of 3-day-old
virgin females mated for 6 seconds or more while less than half mated a second
TABLE I
Presence of semen in female A. aegypti according to duration of interrupted mating
Duration of mating (seconds)
Proportion of females containing semen
1
2
3
4
5
6
0/2
3/10
2/6
4/9
4/5
6/6
time when re-exposed during the next week. Virgin females of comparable age
mated readily. Abortive matings (i.e., duration less than 6 seconds) occurred with
greatest frequency in the previously-mated group. Of those females that mated,
most did so during the first minute of exposure to males.
3. Appearance of the copulatory Inirsa after mating
The copulatory bursa of freshly inseminated females was distended. Its con-
tents included motile sperm and many clear globules intermixed with fine granules.
Sperm were most numerous in a clear region at the periphery. The bursal walls,
which were approximately 2 microns thick before mating, generally became vacuo-
lated and as much as 20 microns thick after mating (Fig. 2). During the ensuing
day, the bursa gradually lost most of its volume, globules disappeared, sperm be-
came quiescent, and the bursal walls again became thin. Within the first 10 min-
utes after mating, the genital atrium contained a rapidly undulating mass of sperm.
The sperm were directed toward the spermathecal eminence and the spermathecal
vestibule, forming a flickering mass that later dispersed. A distended bursa, filled
with motile sperm and clear globules, was accordingly taken as evidence of recent
insemination.
SEMINAL LOSS IN FEMALE AEDES
407
FIGURE 2. Copulatory bursa of female after being inseminated for the first time. Clear
globules are present in the center of the semen mass; a swirling mass of sperm (F) is in the
ventral portion of the bursal orifice and the bursal wall is thick and vacuolated.
FIGURE 3. Copulatory bursa of female that had mated twice and retained semen from both
matings. Bursa was dissected after the second mating which followed one week after the first.
The clear globules and swirling sperm of the second mating are present, together with the dark
remnants of the initial semen mass (R). The wall of the bursa is thin and membrane-like.
4. Presence of semen after mating
The frequency with which virgin females became inseminated was studied in
females of various ages. Those that mated for at least 6 seconds were chloro-
formed and the copulatory bursae examined within 5 minutes of mating. Semen
was almost invariably present after young virgin females mated (Table III).
In contrast, previously mated females generally did not contain fresh semen
after second mating (Table III). This effect was especially marked when one or
TABLE II
Readiness to mate and duration of mating of individual females
during consecutive or single exposures to males
Number of females individually exposed to
males for 2 minutes
% of females mating for 6 seconds or more
Mean number of contacts of less than 6 seconds
per mated female
Mean seconds until mating (6 seconds or more)
commenced
Mean seconds duration of mating
Consecutively mated group
First mating of
3-day-old females
128
87.5
1.5 ± 0.3*
35.1 ±4.2
17.6 ± 0.5
Second mating at
5-10 days of age
Ill
46.9
2.5 ± 0.4
45.0 ±4.7
12.6 ± 1.3
Control; single-
mated group
First mating of
5-10-day-old
females
29
100
1.7 ± 0.6
47.5 ± 6.5
19.4 d= 1.3
* Standard error.
408
A. SIM KL MAX, SR. M. G. l.KAHY AND VALERIE SKAEF
two days had elapsed between the first and second matings. Tlie Imrsae of females
that retained semen from the second mating invariably contained identifiable rem-
nants of the first seminal mass (Fig. ^). The first mass was darker than the
second, clearly demarcated, and located apically. It was thus possible in each
instance to confirm that prior insemination had occurred. Furthermore, in more
than half (8 ont of 14) of the twice-inseminated females, the walls of the bursa
remained thin. The bnrsal walls of females that were mated for the first time
almost invariably (95 ont of 96) became thick.
TABLE III
Comparison of retention of semen by females after second mating with that observed
in once-mated control group of comparable age
Previously mated
Control groups of same age
with no previous mating
Time to second
mating
No. females
% retaining semen
after second mating
No. females
% retaining semen
after initial mating
M hrs.
22
27
22
100
5 hrs.
16
19
16
100
1 day
15
0
14
100
2 days
27
7
25
96
1 week
27
11
15
93
2 weeks
25
24
15
93
3 weeks
12
8
9
78
5. Loss of semen following second mating
The absence of fresh semen in most females following second mating suggested
that insemination might not have occurred despite observed genital contact of
sufficient duration to ensure effective mating of virgin females. Accordingly,
mosquitoes were frozen during copulation to compare semen transfer in first and
second matings. Eighteen females were frozen during their first mating ; of these,
14 were prepared as whole mounts and 4 were sectioned. Eleven additional
females were re-mated two days after initial mating and similarly frozen ; 7 of this
group were mounted whole and 4 sectioned. Semen was present in the copulatory
bursae of all 29 females in both groups (Fig. 4). Thus, insemination of virgin
females and of previously mated females occurred with equal frequency.
Copulation of virgin and of once-mated females was compared. Genital union
in all cases was firm and corresponded closely to descriptions by Spielman (1964).
The presence of semen in the copulatory bursa during second mating and its
absence after separation indicated that semen must have been expelled following the
mating. This process is illustrated by the female in Figure 5, one which was for-
tuitously frozen as it was separating from the male. The paraprocts and claspers
of the male were in normal copulating position and in contact with the female. The
aedeagus, however, was retracted. The female's genital parts, too, were in copulat-
ing position, but the copulatory bursa was contracted and a mass of semen was
present between the genital lips. Taken together, these observations suggest that
the seminal mass was expelled from the copulatory bursa following withdrawal
of the aedeagus.
SEMINAL LOSS IN FEMALE AEDES
409
CB
DV-
FIGURE 4. Whole mount of pair that was frozen during copulation. Although the aedeagus
of the male (left) is only partially extended, the copulatory bursa is distended with semen (S).
The spermathecae contain sperm from a previous mating.
FIGURE 5. Whole mount of copulating female that was frozen as the aedeagus was with-
drawn. The copulatory bursa tCB) is empty and a mass of semen (S) was present between
the genital lips and adherent to the post-genital plate. Sperm from a previous mating is present
in the spermathecae but not visible in the figure.
Seminal expulsion following insemination of previously mated females was also
observed directly when mosquitoes were mated manually. Ten 5-day- old females,
mated two days previously, were tethered and brought into contact with virgin
males. A mass of semen was visible externally on 9 of the females as the pairs
were separated 10-20 seconds later. The tenth female contained semen as did
all 10 of a group of virgin females that were similarly mated as controls.
6. Loss of semen following mutiny of rinjin females implanted icith mole tissue
The preceding observations suggested that loss of semen in twice-mated females
might be due to some attribute of semen transferred to the female during first
copulation. To explore this possibility, organs removed from 3- or 4-day-old
virgins were implanted in the thoraxes of 3- or 4-day-old virgin females as follows :
male accessory gland (Group A), testis (Group B), and ovary (Group C). Con-
trols included normal female virgins of the same age ( Group 1) ) and normal females
TABLE IV
Seminal retention after mating in virgin females that had received
tissue implants 2 days previously
Females
f^
T* 1 h - 1
Number
% retaining semen
A
Male accessory gland
28
21
B
Testis
16
81
C
Ovary-
10
100
D
No implant (virgin)
10
100
E
No implant (mated 2 days previously)
22
14
410
A. SPIELMAN, SR. M. G. LEAHY AND VALERIE SKAFF
which had mated at the time that groups A-C received implants (Group E). All
were individually exposed to males two days after groups A-C received implants.
The post-mating results (Tahle IV) showed that females in groups A and E tended
to lose semen, that loss of semen was less frequent in Group B, and that semen
was retained by all mosquitoes in Groups C and D. The findings suggest that
male accessory gland and, to a lesser extent, the testis contain materials that may
be responsible for seminal loss.
The effect upon the wall of the copulatory bursa produced by implantation of
male accessory glands was studied in 10 virgin females which were examined ap-
proximately one hour after receiving the implants. In 9 the bursal walls were
vacuolated and more than 2 microns thick ; in two of these the bursal walls were
highly vacuolated and about 20 microns thick. By comparison, the bursal walls
were non-vacuolated in 10 females implanted with ovarian tissue, and in only one
of these females was the bursal wall thicker tban 2 microns.
TABLE V
Effect of blood feeding upon seminal retention in non-virgin females. At beginning of experiment,
females were mated, then fed blood and re-mated after 1 week. In control groups,
initial mating and /or blood feeding was omitted
Treatment at beginning of experiment
Females
Number
% retaining semen after
subsequent mating
Mated. Blood-fed
22
28
Mated. Not exposed J:o host
Not exposed to males. Blood-fed
Not exposed to males. Not exposed to host
23
14
17
13
100
100
7 . Utilization of sperm received in second mating
Although semen was occasionally retained by twice-mated females, it was not
known whether the sperm from the second mating were utilized. This was tested
by mating genetically marked (Gold Mesonotum) males with wild-type females 5
hours after the females had first been inseminated by males of their own type.
Although Gold Mesonotum has frequent penetrance in heterozygous females (Craig
and Vandehey, 1962), it is not completely dominant, and the resulting families
would contain fewer marked females than actual heterozygotes. Of 22 families
produced, 4 contained females of both genetic types. Thus, semen retained from
a second mating is capable of fertilizing eggs.
8. Effect of blood feeding upon seminal retention in non-virgin females
The influence of blood meals and resulting ovarian development on retention
of semen in previously mated females was studied. Three-day-old virgin females
were mated individually in lantern chimneys ; one hour later they were permitted
to feed on a guinea pig. Blooded mosquitoes were re-mated one week later. In
control groups of similar virgin females, initial mating, the blood meal, or both,
were omitted. Raisins were provided as supplemental food. Retention of semen
SEMINAL LOSS IN FEMALE AEDES 411
was universal among females mated for the first time, regardless of food regimen
(Table V). Retention after second mating, however, appeared to he somewhat
enhanced by blood feeding.
DISCUSSION
In most insects, males transfer sperm to the female by means of spermatophores
from which the sperm escape after the copulating pair separates (Khalifa, 1949;
Davey, 1960). Other animals may instead possess seminal gels that harden within
the female, and "it is sometimes supposed that in vertebrates such plugs assist
insemination by preventing loss of semen from the female genital tract" (Hinton,
1964, p. 96). Diptera do not have spermatophores, and it seems likely that some
special device may aid in the retention of semen after mating. Copulation in A.
aeg\ptl is accomplished through the superficial apposition of genital parts, semen
being extruded into a chamber formed by the mating pair (Spielman, 1964). This
arrangement might result in loss of semen as the male withdraws unless the seminal
mass becomes altered in some way during or immediately after insemination. In-
deed, the appearance of clear globules in the copulatory bursa during the first
minute of a first mating may be associated with such a change. In addition, semen
appears to be expelled unless held within the female by the aedeagus of the male
for a few seconds. The formation of a "mating plug" following insemination of
various mosquito species has been described (Gillies, 1956; Lum, 1961).
Loss of semen after a second mating occurred more frequently than after the
first, especially if the period between matings exceeded a few hours. This sug-
gests that prior mating may interfere with the normal reaction of the female's
genital tract to semen. Under these conditions, the mass of semen flows freely
and is lost when the female's genital orifice is vacated by the aedeagus.
Vacuolization of the bursal wall is one reaction of the female's genital tract to
the semen (Spielman, 1964; Jones and Wheeler, 1965). That vacuolization fre-
quently did not occur following a second mating of female A. acgypti suggests that
the wall of the bursa may have some role in seminal retention. However, vacuoli-
zation is not an absolute prerequisite to seminal retention, because in some females
that retained semen, vacuolization was not observed.
Loss of semen deposited in a second mating appears to be due to some com-
ponent of the semen of the first mating, and the male's accessory glands, which
elaborate a major portion of the semen, are the probable source of this factor.
There need not he direct contact with the bursa, for seminal loss and vacuolization
of the bursal wall occurred after the male tissue was implanted in the thoraxes of
virgin females. Interestingly, male accessory glands have also been shown to
contain a material that enhances oviposition in A. acg\pti females (Leahy and
Craig, 1965).
The present observations indicate that effective multiple insemination of female
A. acgvpti may be infrequent in nature. Indeed, if the second mating does not
occur during the same day as the first, it appears unlikely that semen from both
would be retained.
SUMMARY
1. The effectiveness of mating of female Aedes aegypti that had previously
been mated was compared to that of virgin females. Non-virgin females mated
412 A. SPIELMAN, SR. M. G. LEAHY AND VALERIE SKAFF
less readily than virgin females and copulation was of somewhat shorter duration.
Cienital union was firm, and insemination occurred in both virgin and non-virgin
females, yet semen was generally not retained in the copulatory bnrsa of females
that had previously been mated. This effect was most evident when one or two
days had elapsed between matings. Multiple insemination, with utilization of
sperm from both matings, however, was occasionally effective when less than 5
hours separated the first and second matings.
2. Factors derived from the accessory glands of the male and, to a lesser extent,
the testes appeared to induce this loss of semen.
3. It was suggested that semen normally gels during mating and that loss of
semen following second mating may result from a defect in this process.
4. These data indicate that female A. acg\ptl in nature may normally utilize
sperm from but one male.
ADDENDUM
Recently, George1 observed female A. aegypti that were sequentially mated to
irradiated and to non-irradiated males. He suggested that, "copulation may occur
repeatedly, (but) the only effective one is the first" (p. 85 ).
LITERATURE CITED
CRAIG, G. B., AND R. C. VANDEHEY, 1962. Genetic variability in Aedes aegypti (Diptera:
Culicidae). 1. Mutations affecting color pattern. Aim. Ent. Soc. America, 55: 47-58.
DAVEY, K. G., 1960. The evolution of spermatophores in insects. Proc. Ro\. Ent. Soc. London
(A). 35: 107-113.
FRENCH, W. L., AND J. B. KITZMILLER, 1963. Tests for multiple fertilization in Anopheles
quadrimaculatus. Proc. New Jersey Mosq. E.rtenn. Ass., 50: 374-380.
GILLIES, M. T., 1956. A new character for the recognition of nulliparous females of Anopheles
gambiae. Bull, ll'orld Health Org., 15: 451-459.
GOMA, L. K. H., 1963. Tests for multiple insemination in Anopheles qainbiae Giles. Nature,
197: 99-100.
HINTON, H. E., 1964. Sperm transfer in insects and the evolution of haemocoelic insemination.
In: Insect Reproduction. Syinp. Roy. Ent. Soc. London, 2: 95-107.
JONES, J. C., AND R. E. WHEELER, 1965. Studies on spermathecal filling in Aedes aegypti (Lin-
naeus). I. Description. Biol. Bull, 129: 134-150.
KHALIFA, A., 1949. Spermatophore production in Trichoptera and some other insects. Trans.
Roy. Ent. Soc. London, 100: 449-479.
KITZMILLER, J. B., AND H. LAVEN, 1958. Tests for multiple fertilization in Culc.v mosquitoes
by the use of genetic markers. Ainer. J. II yg., 67: 207-213.
LEAHY, SR. M. G., AND G. B. CRAIG, JR., 1965. Accessory gland substance as a stimulant for
oviposition in Aedes aegypti and A. alhopictus. Mosq. News, 25: 448-452.
LUM, P. T. M., 1961. The reproductive system of some Florida mosquitoes. II. The male
accessory glands and their role. Ann. Ent. Soc. America, 54: 430-433.
ROTH, L. M., 1948. A study of mosquito behavior. An experimental laboratory study of the
sexual behavior of Acdcs aegypti (Linnaeus). Ainer. Midi. Natural., 40: 265-352.
SPIELMAN, A., 1956. The inheritance of autogeny in the Cule.v pipiens complex of mosquitoes.
Thesis, Johns Hopkins School of Hygiene and Public Health.
SPIELMAN, A., 1964. The mechanics of copulation in Aedes aegypti. Biol. Bull.. 127: 324-344.
VANDEHEY, R. C., AND G. B. CRAIG, 1958. Multiple fertilization demonstrated in Aedes acgvpti.
Bull. Ent. Soc. America, 4: 102.
1 George, J. A., 1967. Effect of mating sequence on egg-hatch from female Aedes aegypti
(L.) mated with irradiated and normal males. Mosq. News, 27: 82-86.
CAROTENOID PIGMENTS IN THE CELLULAR SLIME MOLD,
DICTYOSTELIUM DISCOIDEUM l>'
SUZANNE O. STAPLES3 AND JAMES H. GREGG
Department of Zoology, University of Florida, Gainesville. Florida 32601
Five distinct stages of development may be recognized in the life-cycle of the
cellular slime mold, Dictyostelinin discoideiiui (Bonner, 1944; Raper, 1937, 1939,
1940). The first four stages are relatively colorless, but the fifth, or fruiting body,
stage is marked by a change in the color of the sori from pale buff to bright lemon
yellow. Since a change in color in an organism may reflect changes in metabolic
events, the nature of the pigments was investigated to determine : ( 1 ) whether this
color transition represents dc noi'o synthesis by the spore cells or the mere accumu-
lation of substrate pigments from the medium, as suggested by Whittingham and
Raper (1956); (2) the nature of the pigment; (3) the effect of diphenylamine,
which specifically inhibits carotenogenesis (Goodwin, 1952, 1954; Haxo, 1955;
Kharasch, 1936; Turian, 1950; Zalokar, 1957).
METHODS AND MATERIALS
Culture procedures and harrcstimj of tissue
All cultures of D. discoiilenin and the bacterial associate, Escherichia coll, were
maintained on an agar medium (Bonner, 1947) in Petri plates. Cultures were
incubated in darkness at 22° C. in an environment ranging from 55 to 90% relative
humidity. Following various intervals of incubation, cultures of D. discoidenin
were harvested for dry weight determination and pigment assay. Cultures were
scraped into tared flasks and weighed wet. An aliquot of known wet weight was
removed for dry weight determination and the remaining wet sample was extracted
for pigment. The aliquot for dry weight determination was freed of E. coli in a
0.55 A//0.95 M sucrose gradient, washed, dried, and weighed. This procedure
permitted calculation of the dry weight of the sample used for pigment extraction.
In order to extract the pigment, cultures of the desired stages were scraped into
95% ethanol, allowed to stand several hours, and filtered. The residue, from
greyish-white to pale yellow in color, was discarded, and the yellow filtrate (ethanol
extract) was stored in darkness at 10° C. Extracts obtained by this procedure
were either assayed for pigment as described below or purified further.
1 Presented to the Graduate Council of the University of Florida in partial fulfillment of
the requirements for the degree of Master of Science (S. O. S.).
- This investigation was supported in part by a Public Health Service Research Career
Programs Award 5-K3-HD-15,780 from the National Institute of Child Health and Human
Development and Research Grants E-1452 and GM-10138 from the National Institutes of Health
(J. H. G.).
3 Present address : Department of Pharmacology and Therapeutics, College of Medicine,
University of Florida, Gainesville, Florida.
413
414 SUZANNE O. STAPLES AND JAMES H. GREGG
Assay of pigment
The optical density (O. D.) at 390 m//. of ethanol extracts was proportional to
the O. D. at 390 m/A of the more purified preparations described later, and was
suitable for gross quantitative assay of pigment.
Purification of p'ujment for analysis
When large quantities of pigment were needed for chemical analyses, only cul-
tures of mature sorocarps were harvested, and were extracted as described pre-
viously. Cultures containing only E. coli and nutrient agar contained no pigment
and were not further studied. The ethanol extract was saponified with 10% KOH
(w/v) at 65° C. for 2 hours. Following hydrolysis, the pigment solution was
transferred to a separatory funnel with an equal volume of diethyl ether and sufficient
water to effect separation of the two phases. Acetic acid was added to transfer all
pigment into the ether phase. The epiphase was then washed with water, dried
over anhydrous Na,SO4, transferred to an Erlenmeyer flask, and taken to dryness
under reduced pressure (Residue I). Small portions of hexane were added to
Residue I and then decanted from the flask until further additions of solvent re-
mained colorless. The hexane extracts were combined, reduced in volume under
vacuum, and poured onto a powdered sucrose column. The column was first devel-
oped with hexane, and then washed successively with diethyl ether and methanol.
The major yellow fraction was eluted with ether, and is designated Fraction I.
The hexane eluate, which was also yellow, was rechromatographed on MgO:Celite
(1:1). Several pigments were subsequently eluted from this column with hexane
and ether, but the very small quantities present precluded further analysis. Residue
I was subsequently extracted with ether, and the ether-soluble pigments were also
chromatographed on powdered sucrose. Developing the column with ether eluted
a pigment designated Fraction 1 1 -a. The column was then washed with methanol,
eluting Fraction Il-b. Following removal of the hexane and ether-soluble pigments
from Residue I, the remaining pale yellow residue was taken up in methanol (Frac-
tion III).
Chemical analyses of fractions
The presence of an acidic function was tested by noting differences in the dis-
tribution behavior of the salt and the free acid between two solvents (Fox, 1953;
Zalokar, 1957). The absorption spectra of the acidic and basic forms of the pig-
ment were also compared (Zalokar, 1957).
Several qualitative color tests were used to detect polyene structure. Fractions
I through III were taken to dryness, and a drop of concentrated H,SO4 was added
to the residues. In addition, concentrated H2SO4 was layered under ether solutions
of Fractions I and Il-a (Karrer and Jucker, 1950). A few crystals of dithionite,
a reducing agent, were added to Fractions I and Il-a in ether and to Fractions II-b
and III in methanol. Color changes were noted. Antimony trichloride was added
to chloroform solutions of Fractions I and Il-a (Carr and Price, 1926; Karrer and
Jucker, 1950). This reaction could not be carried out on Fractions II-b and III
because they were insoluble in chloroform.
SLIME MOLD PIGMENTS
415
Spectrophotometric analyses
All spectral data were obtained with either a Beckman DK-2 or a Bausch and
Lomb Spectronic 505 recording spectrophotometer.
The effects of diphenylamine
Culture media were prepared by adding to the agar a stock solution of W'2 M
diphenylamine (DPA) in 95% ethanol to give concentrations of DPA from
5 X 1O6 M to 5 X lO5 M. Controls were prepared by adding the appropriate
volume of ethanol to the culture media. Plates were then inoculated, incubated for
various intervals, and examined for relative number of sorocarps and intensity of
coloration.
RESULTS
Studies relating p'njment concentration to developmental stage
The course of pigmentation is shown in Figure 1. The concentration of pig-
ment was low initially and did not change appreciably for 60 hours. It then in-
creased slowly until fruiting began at about 72 hours. At 84 to 96 hours, fruiting
was morphologically complete and the pigment concentration continued to increase.
Major pigment accumulation did not occur until after fruiting was complete.
0>
*
o
=<.
E
O
ffi
z
u
o
60 Ti 64 96
TIME OF INCUBATION (HOURS)
FIGURE 1. Increase in pigment concentration during development. VM -- Vegetative myxamoe-
bae. Ag = Aggregates. MP = Migrating pseudoplasmodia. MS = Mature sorocarps.
416
SUZANNE O. STAIM.KS AND JAMES H. GREGG
1.0
.9
.8
(D
UJ
_J
O
Q.
O
.5
.4
.3
.2
.1
600
395
r
340 323
FIGURE 2. Absorption spectra of D. discoidciiin and E. coli extracts. A — D. discoidcnin
ethanol extract. B = E. coli ethanol extract. C := D. discoidcmn ether phase. D — E. coli
ether phase. E = Control hydrolysate.
Comparison of spectra of D. cliscoicleum, E. coli, anil nutrient injar ethanol extracts
The ethanol extracts of D. discoideuin, E. coli, and nutrient agar were examined
spectrophotometrically. The D. discoideuin extract showed a peak at 398 m/j, with
shoulders at 414 m/x and 381 m/z. The spectra of the E. coli and agar extracts
showed no ahsorption of the visible region (cf. Curves A and P>, Figure 2). Fur-
ther, the ether phases obtained after hydrolysis of ethanol extracts of D. discoideuin,
E. coli, and agar were compared. As shown in Figure 2, the strong absorption
band seen in the ethanol extracts of D. discoideuin was resolved to yield absorption
SLIME MOLD PIGMENTS
417
maxima at 417 m/x, 397 m/x, and 377 m/x (Curve C). On the other hand, the
spectra of the colorless E. coll and agar extracts were essentially the same as that
of the control.
Spectral characteristics of pigment fractions
The absorption maxima of the pigment fractions are summarized in Table I.
Chemical tests
Test for an acidic function
Partitioning an ether solution of Fraction I against 2 N NaOH moved the pig-
ment to the interface where it was visible as a yellow layer. Addition of methanol
distributed the pigment between the two phases. Acidification restored the pigment
to the ether phase.
TABLE I
Comparison of literature and experimental maxima at \i
Compound
Solvent
Maxima at Xi
zeta-carotenea
Hexane
425
400
378 360
295
zeta-caroteneb
Petroleum ether
418
396
376
OH-zeta-caroteneb
Petroleum ether
417
396
376
di-OH-zeta-caroteneb
Petroleum ether
420
397
378
0-carotenec
Not specified
421
397
375.5
Fraction I
Diethyl ether
418
397
* 378* 360
338
Fraction 1 1 -a
Diethyl ether
418
396
* 375*
Fraction II-b
Methanol
414
397
* 378
Fraction III
Methanol
397
Calculated with Kuhn's formula
None
408
* Indicates Xmax-
a Nash, 1945.
b Jensen, 1958.
c Haxo, 1955.
When the ether solution of Fraction Il-a was partitioned against alkali, the
pigment was distributed between the two phases. Acidification moved all pigment
into the epiphase. Hence Fraction Il-a may contain at least two components, only
one of which has an acidic group.
When partitioned against ether and alkali, all pigments in Fractions II-b and
III were hypophasic. Addition of methanol moved very little pigment into the
epiphase. Acidification moved the pigments into the epiphase. Therefore, the
pigments of Fractions II-b and III possess acidic functions.
Test for an acidic function in conjugation with the chromophore
The positions of the maxima of Fractions I, Il-a, and II-b in the visible region
were not altered by pH. The intensity of absorption, however, increased in basic
solutions and decreased in acidic solutions.
The maximum of Fraction III occurred at 392 m/x in alkaline methanol but was
418
SUZANNE O. STAPLES AND JAMES H. GREGG
at 397 m/i in acidic methanol. The absorption increased in alkaline methanol and
decreased in acidic methanol. The spectral shift of 5 m/x, is consistent with the
data for neurosporaxanthin (Zalokar, 1957) and indicates that the acidic function
is conjugated with the chromophore.
Qualitative tests for polyene structure
The tests in Fractions I and Il-a were all weakly positive ; treatment with
H2SO4 produced traces of blue which rapidly gave way to relatively stable brown
colors. The Carr-Price reaction yielded deep bluish-orange colors. The color of
these fractions was not entirely abolished by dithionite, although absorption in the
visible disappeared. On the other hand, the brilliant blue colors obtained by treat-
ing Fractions Il-b and III with H2SO4 (Haxo, 1949; Karrer and Jucker, 1950).
and the complete decolorization of these two fractions by dithionite showed polyene
structures.
TABLE II
The effect of various concentrations of DP A upon mature sorocarp formation
and pigment synthesis in D. discoideum
Concentration of DPA in media
Relative number of sorocarps
Relative pigment concentration
5 X 10-* M
None
—
4 X 10-B M
Very few
Colorless
3 X 10-6 M
Not abundant
Colorless to pale yellow
2 X 10-B M
Abundant
Colorless to pale yellow
1 X ID"6 M
Abundant
Pale yellow
5 X 1C-6 M
Abundant
Only slightly less yellow
than controls
Ethanol control
Abundant
Bright lemon yellow
Control
Abundant
Bright lemon yellow
The results suggest that Fractions I and Il-a contained carotenoid pigments,
with, however, colorless impurities and non-carotenoid pigments of unknown nature.
The data point to many similarities between these non-carotenoid pigments and the
lipofuschins described by Fox (1953). Unquestionably the pigments in Fractions
Il-b and II are carotenoids.
The effect of diphenylamine on pigment synthesis
The effects of several concentrations of DPA upon pigment synthesis are given
in Table II. At 2 X 1O5 M to 3 X 10'5 M it significantly inhibited pigment syn-
thesis without interfering with growth. This concentration range was very critical.
Growth was severely limited at 5 X 1O5 M DPA, but in the presence of 5 X 1Q-6 M
DPA the mature sorocarps were practically indistinguishable from the controls.
DISCUSSION
Whittingham and Raper (1956) have suggested that pigmentation in D. dis-
coideum sori depends upon environmental factors, such as the substratum or the
bacterial associate upon which the slime mold feeds. Other studies have estab-
SLIME MOLD PIGMENTS 419
lished that sorocarp color may be influenced by the incorporation of vital dyes
(Bonner, 1952) or pigmented foodstuffs (Raper, 1937). In these instances pig-
mentation occurs as a result of the accumulation and retention of soluble or par-
ticulate exogenous pigment.
This investigation indicates that sorocarp pigmentation arises by de novo syn-
thesis. Absorption spectra of ethanol extracts of D. discoideum showed in the
visible region a characteristic peak which was absent from ethanol extracts of both
E. coli and nutrient agar (Fig. 2). Further, a sharp separation exists between the
feeding period and the morphogenetic phase of the life cycle (Bonner, 1947, 1959).
Although food intake ceases at the beginning of aggregation (Bonner, 1959), the
colorless pseudoplasmodia could be transferred to a non-nutrient agar surface where
pigmented fruiting bodies subsequently occurred. Further evidence that the yellow
pigment is not merely accumulated was obtained from the studies relating pigment
concentration to development (Fig. 1). At the time pigment appears in sig-
nificant quantities, the spore cells are supported in the air by a stalk, and thus are
removed from immediate contact with any exogenous supply of pigment.
These pigments are mainly carotenoids. At a suitable concentration, DPA,
a well known inhibitor of carotenoid synthesis, either decreased or completely in-
hibited pigmentation without affecting growth. In addition, the massive bands
in the visible region of the absorption spectra are characteristic of carotenoid pig-
ments. This band, believed to arise from the oscillation of pi electrons from one
end of the conjugated polyene structure composing the chromophore to the other
(Dale, 1954), often exhibits fine structure usually manifested by three maxima or
two maxima and a shoulder. The peak with the highest intensity is referred to
as the Amas, while the whole band, including its fine structure, is referred to as the
fundamental band or X1 (Dale, 1954; Zechmeister, 1960). For a given solvent, the
position of this band and its degree of fine structure depend upon the length of
the chromophore (Zechmeister, 1960). One of the most important determinants
of the length of the chromophore and hence the position of Ax is the number of
conjugated double bonds. This relationship has been worked out both theoreti-
cally and empirically so that by the position of Aa, one can estimate rather closely
the number of conjugated double bonds in the chromophore. From curves relating
the number of conjugated double bonds and maxima at Ax (Dale, 1954; Nash,
1948), it was estimated that pigments in Fractions I, Il-a, Il-b, and III possessed
seven conjugated double bonds. Then, Kuhn's empirical formula was used to
calculate the wave-lengths near which a system with seven double bonds should
display maxima (Table I). Kuhn's formula (Dale, 1954) is:
157
A* —
- 0.922 cos
where n = the number of conjugated double bonds and s = the band order. It can
be seen that close agreement exists between observed and calculated values at A1?
despite the dependence of the position of this band upon both the solvent and the
presence or absence of an oxygen atom conjugated with the chromophore. The
observed maxima at Ax are compared with literature values of known carotenoids
in Table I.
420 SUZANNE O. STAPLES AND JAMES H. GREGG
The evidence that these pigments are acidic is of interest, for relatively few
acidic carotenoids have been described. In the native state these pigments exhibited
limited solubility in water. It is unlikely that combination with protein contributed
in any substantial way to this water-solubility, for a dramatic spectral shift (Fox,
1948) was never observed when the pigments were subjected to procedures which
would hydrolyze a protein moiety and liberate pigment. On the other hand,
esterification with a sugar residue may be an important factor in conferring this
limited water-solubility. Until the pigments were hydrolyzed, no amount of acid
forced the pigments from the ethanol extract into the ether phase.
The partition behavior of these pigments was not studied extensively. How-
ever, the insolubility of Fractions Il-b and III in such non-polar solvents as hexane,
benzene, and carbon disulfide suggested the presence of polar groups. The possi-
bility that these pigments were xanthophylls is of interest with respect to the
observed DPA inhibition. Haxo (1955) reported that DPA inhibited formation
of xanthophylls in Mycobacterium phlei, and Turian and Haxo (1952) found this
inhibition to be most marked at the terminal synthetic steps, i.e., at the conversion
of neutral hydrocarbons into acidic compounds.
The evidence that these pigments are carotenoids and that they contain seven
conjugated double bonds leads to the conclusion that they probably belong to the
zeta-carotene series.
Functional significance of these pigments should be explored. Most attempts
to establish a function of carotenoids in fungi have pointed to their mediating photo-
kinetic responses (Goodwin, 1954). Migrating pseudoplasmodia in D. discoideum
exhibit a strong positive photoactic response (Bonner, 1952) ; and Francis (1964)
has recently shown this action spectrum to have a major peak near 425 m/x and
a minor peak near 550 m/x. He found that the absorption spectrum of "slime"
from spore heads also peaked about 425 m^.. On the basis of these two findings,
he has suggested that the slime sheath may contain the receptor system for photo-
taxis. The absorption spectrum of sori "slime" which Francis has reported is
similar to the spectra of pigments studied in this investigation, and it seems plausible
that these components are identical. Although Francis has demonstrated a photo-
tactic response in the migrating pseudoplasmodia, he did not study the absorption
spectra of this stage. In the present study no absorption peak near 395 m/z, and
consequently no yellow pigment, was found in migrating pseudoplasmodia. In
order to demonstrate an association between phototaxis and a pigment, one should
be able to correlate the absorption spectrum maximum with the same develop-
mental stage as that in which the phototactic response occurs. Evidence obtained
by other investigators (Goodwin, 1952, 1954) suggests that in some instances as
little as 1-2% of the usual carotene content may be sufficient for phototactic action;
or alternatively, that the more saturated polyenes mediate a photokinetic response.
Another possibility is that the pigment and phototactic response have nothing to
do with each other.
There is a great deal of circumstantial evidence in the literature pointing to,
but never specifically defining, a reproductive function for carotenoids. In D. dis-
coideum, carotenoids accumulate in the spores, i.e., in the structures directly con-
cerned with reproduction. The pigmentation reaches a peak after fruiting and
sharply decreases in the vegetative phase. These findings are similar to those of
SLIME MOLD PIGMENTS 421
other investigators (Emerson and Fox, 1940; Fox, 1948; Goodwin, 1950; Mur-
neek, 1934) which indicate that the highest concentrations of carotenoids in plants
and animals are found in tissues and secretions associated with reproduction, and
suggest that carotenoids for some reason may be associated with reproduction in
D. discoideum although no conclusions about their function can be drawn from
the present evidence.
We are extremely indebted to Dr. James A. Olson of the Department of Bio-
chemistry and Dr. Robert M. DeWitt of the Department of Zoology, University of
Florida for their invaluable suggestions, advice and aid during the course of this
investigation.
SUMMARY
1. The lemon-yellow pigmentation in the mature sori of the cellular slime mold,
Dictyostelium discoideum, was shown to arise by de novo synthesis and not by
accumulation from an exogenous source. Pigment synthesis reached a peak after
fruiting and then sharply declined in the vegetative phase.
2. The major pigments appeared to be related to the zeta-carotenes. Inhibition
of pigment synthesis by diphenylamine, which specifically inhibits carotenogenesis,
indicated the pigments were carotenoids. Chemical and spectral analyses of the
pigments indicated polyene structures with seven conjugated double bonds.
3. Most of the pigments contained acidic functions. The acidic function of
one pigment appeared to be conjugated with the chromophore.
LITERATURE CITED
BONNER, J. T., 1944. A descriptive study of the development of the slime mold Dictyostelium
discoideum. Amer. J. Bot., 31 : 175-182.
BONNER, J. T., 1947. Evidence for the formation of cell aggregates by chemotaxis in the
development of the slime mold Dictyostelium discoideum. J. Exp. Zool., 106: 1-26.
BONNER, J. T., 1952. The pattern of differentiation in ameboid slime molds. Amcr. Nat., 86:
79-89.
BOXXER, T. T., 1959. The Cellular Slime Molds. Princeton University Press, Princeton, New
Jersey.
CARR, F. H., AND E. A. PRICE, 1926. Color reactions attributed to vitamin A. Biochem. J., 20:
497-501.
DALE, J., 1954. Empirical relationships of the minor bands in the absorption spectra of polyenes.
Acta Chem. Scand., 8: 1235-1256.
EMERSON, R., AND D. L. Fox, 1940. Gamma-carotene in the sexual phase of the aquatic fungus
Allomyces. Proc. Roy. Soc. London, Ser. B, 128: 275-293.
Fox, D. L., 1948. Some biochemical aspects of marine carotenoids. Fortschr. Chem. org.
Naturstoffe,S: 20-39.
Fox, D. L., 1953. Animal Biochromes and Structural Colors. Cambridge University Press,
Cambridge.
FRANCIS, D. W., 1964. Some studies on phototaxis of Dictyostelium. J. Cell. Comp. Phvsiol.,
64: 131-138.
GOODWIN, T. W., 1950. Carotenoids and reproduction. Biol. Rev. Camb. Phil. Soc., 25: 391-
413.
GOODWIN, T. W., 1952. Studies in carotenogenesis 3. Identification of the minor polyene com-
ponents of the fungus Phycomyces blakesleeanus and a study of their synthesis under
various cultural conditions. Biochem. J., 50: 550-558.
GOODWIN, T. W., 1954. Carotenoids Their Comparative Biochemistry. Chemical Publishing
Company, Inc., New York.
422 SUZANNE O. STAPLES AND JAMES H. GREGG
GOODWIN, T. W., M. JAMIKORN AND J. S. WILMER, 1953. Studies in carotenogenesis 7. Further
observations concerning the action of diphenylamine in inhibiting the synthesis of beta-
carotene in Phycomyces blakesleeanus. Biochcm. J., 53: 531-538.
HAXO, F. T., 1949. Studies on the carotenoid pigments of Neurospora. I. Composition of the
pigment. Arch. Biochem., 20: 400-421.
HAXO, F. T., 1955. Some biochemical aspects of fungal carotenoids. Fortschr. Chem. org.
Naturstoffe, 12: 169-197.
JENSEN, S. L., 1958. The path of carotenoid synthesis in a photosynthetic bacterium. Biochim.
ct Biophys. Acta, 29: 477-498.
KARRER, P., AND E. JUCKER, 1950. Carotenoids. Elsevier Publishing Company, New York.
KHARASCH, M. S., 1936. Some chemical factors influencing growth and pigmentation of certain
microorganisms. /. Bad., 32 : 533-540.
MURNEEK, A. E., 1934. Relation of carotenoid pigments to sexual reproduction in plants.
Science, 79: 528.
NASH, H. A., 1945. Absorption spectrum of zeta-carotene. Arch. Biochem., 7: 305-312.
NASH, H. A., 1948. Studies on the structure of zeta-carotene. /. Amcr. Chem. Soc., 70:
3613-3615.
RAPER, K. B., 1937. Growth and development of Dictyostclium discoideum with different bac-
terial associates. /. Agric. Res., 55: 289-316.
RAPER, K. B., 1939. Influence of culture conditions upon the growth and development of Dic-
tyostelium discoideum. J. Agric. Res., 58: 157-198.
RAPER, K. B., 1940. Pseudoplasmodium formation and organization in Dictyostelinm discoideum.
J. Elisha Mitchell Sci. Soc., 56 : 241-282.
TURIAN, G., 1950. Recherches sur la biosynthese des carotenoids chez un Bacille paratuber-
culeux. III. Inhibition de la pigmentation par la diphenylamine. Hclv. Cliim. Acta,
33: 1988-1993.
TURIAN, G., AND F. HAXO, 1952. Further use of diphenylamine for the study of carotenoid bio-
synthesis in Mycobacterium plilei. J. Bad., 63: 690-691.
WHITTINGHAM, W. F., AND K. B. RAPER, 1956. Inhibition of normal pigment synthesis in
spores of Dictyostclium purpurewn. Amcr. J. Bot., 43: 703-708.
ZALOKAR, M., 1957. Isolation of an acidic pigment in Neurospora. Arch. Biochem. Bioph\s.,
70: 568-571.
ZECHMEISTER, L., 1960. Cis-trans isomeric carotenoid pigments. Fortschr. Chem. org. Natnr-
stoffe,l8: 223-349.
INDEX
Acid phosphatase in Arbacia eggs, 1.
Acropora, modifications of by commensal crab,
56.
Action spectrum for spawning of Ciona, 222.
Activity of corpora allata in larval Thermobia,
277.
Adaptation to darkness of goldfish, 200.
to different salinities by Nereis, 362.
Adaptive nature of migration rhythm in
Hantzschia, 44.
Adipohemocyte counts in Galleria, 211.
Aedes, repeatedly mated female, seminal loss
in, 404.
spermatocysts in, 23.
ALBARRACIN, C. See F. D. BARBIERI, 209.
Ambient temperature changes, responses of
bats to, 320.
water oxygen content, in relation to respira-
tory stress in sea urchins, 16.
Ambystoma, water economy of, 126.
Amino acids, uptake of by starfish tissues, 161.
as excretion products of tropical sea
urchin, 34.
Ammonia excretion by tropical sea urchin, 34.
Amount of cytoplasmic DNA in sea urchin
egg, 68.
Amphibian, water economy of, 126.
development, amylase and glycogenolysis in,
299.
Amylase in amphibian development, 299.
Anatomy of genital segments of Spirorbis, 91.
of holothurian digestive system, 337.
of Nereis nephridium, 362.
of Paramonostomum, 133.
of Uniserialis, 266.
Annelid, endoskeletal cartilage in, 244.
genital segments of, 91.
morphology of nephridia of, 362.
Antheraea, cholinesterase in brain of, 108.
Arbacia eggs, distribution of enzymes in, 1.
Area of brine shrimp embryos, relation of to
respiration, 156.
Artemia embryos, respiration during hatching
of, 156.
Arthropod, cardioregulation in, 381, 392.
Ascidian, effect of light on spawning of, 222,
292.
genetic and developmental studies on, 229.
Assay of carotenoid pigments of Dictyostelium,
413.
Association-formation between photic and geo-
physical stimulus patterns, 311.
Asterias, utilization of dissolved exogenous
nutrients by, 161.
Axons, vesiculated, in haemal vessels of holo-
thurian, 329.
B
BARBIERI, F. D., J. S. RAISMAN AND C.
ALBARRACIN. Amylase and glycogenolysis
in amphibian development, 299.
Bats, responses of to changes in ambient
temperature, 320.
Behavior, feeding, of commensal crab, 56.
of repeatedly mated female Aedes, 404.
Benthic microflora, persistent migration
rhythms in, 44.
Bird parasite, morphology and taxonomy of,
266.
BLACK, R. E. See C. JACKSON, 1.
Body temperature of bats, in relation to am-
bient temperatures, 320.
Botryllus, genetic and developmental studies
on, 229.
Brain of silkmoth, cholinesterase in, 108.
BRANDT, C. L. See C. C. LAMBERT, 222.
Brine shrimp embryos, respiration during
hatching of, 156.
Bromlysergic acid diethylamide, effect of on
Limulus heartbeat, 392.
BROWN, F. A., JR., AND Y. H. PARK. As-
sociation-formation between photic and
subtle geophysical stimulus patterns — a
new biological concept, 311.
Bufo, amylase and glycogenolysis during de-
velopment of, 299.
Bursa fabricius, parasite of, 266.
Carrasius, retinomotor rhythms in, 200.
Carbohydrases of holothurian gut, 354.
423
424
INDEX
Carbohydrate metabolism of amphibian em-
bryos, 299.
Cardioregulation in Limulus, 381, 392.
Carotenoid pigments in Dictyostelium, 413.
Cartilage, endoskeletal, in polychaete, 244.
Cecropia silkmoth, cholinesterase in brain of,
108.
Cellular slime mold, carotenoid pigments in,
413.
Changes in ambient temperature, responses ot
bats to, 320.
in hemocyte picture of Galleria, 211.
Chemistry of marine polychaete cartilage, 244.
Chick embryo, induction of immunological
tolerance in, 38.
Cholinesterase in brain of Cecropia silkmoth,
108.
Chromophore, role of in spawning of Ciona,
222.
of Dictyostelium carotenoid pigment, 413.
Chromosome number in Botryllus, 229.
Ciona, effect of light on spawning of, 222, 292.
Circadian rhythm in goldfish, 200.
Circularity of cytoplasmic DNA in sea urchin
eggs, 68.
CLARKE, B. J. See A. M. MUN, 38.
Coelomic oxygen tensions in sea urchins, 16.
Cold, responses of temperate-region bats to,
320.
Colonial corals, modifications of by commensal
crabs, 56.
Colony fusion in Botryllus, 229.
Color changes in Dictyostelium, 413.
Commensal crab, effects of on host, 56.
Cones of goldfish retina, rhythms in behavior
of, 200.
Control of sex characteristics in Rivulus, 174.
Copulation, effect of on number of spermato-
cysts in Aedes, 23.
Copulatory bursa of repeatedly mated female
Aedes, 404.
Coral host of commensal crab, modifications
of, 56.
Corpora allata of Thermobia, growth and
activity of, 277.
Crab, commensal, effects of on host, 56.
CRITTENDEN, L. B. See A. M. MUN, 38.
Cucumaria, digestive system of, 337, 354.
haemal vessels, vesiculated axons in, 329.
Culture methods for Botryllus, 229.
Cyclic activity in goldfish retina, 200.
Cysts of Artemia, respiration during hatching
of, 156.
Cytochrome c, possible role of in spawning of
Ciona, 222.
Cytology of gametogenesis in Spirorbis, 91.
Cytoplasmic DNA in sea urchin eggs, 68.
Cytoplasmic volume of corpora allata cells in
Thermobia, 277.
D
DNA, cytoplasmic, in sea urchin eggs, 68.
DANFORTH, C. G. Northern Pacific Gigan-
tione (Isopoda), 147.
Darkness, effect of on retinomotor rhythms in
goldfish, 200.
role of in migration rhythm of Hantzschia,
44.
in spawning of Ciona, 222, 292.
DAVIS, W. H., AND O. B. REITE. Responses
of bats from temperate regions to changes
in ambient temperature, 320.
Desmognathus, water economy of, 126.
Developing Artemia, relation of surface area
to respiration of, 156.
chick, induction of immunological tolerance
in, 38.
Development of amphibians, amylase and gly-
cogenolysis in, 299.
of Nassarius, relation of temperature to, 253.
Developmental stage of Dictyostelium, as cor-
related with pigment, 413.
studies on Botryllus, 229.
Diadema, nitrogenous excretion in, 34.
Diapause of silkmoth, cholinesterase in brain
during, 108.
Diatom, diurnal migration rhythm in, 44.
Dictyostelium, carotenoid pigments in, 413.
Differential hemocyte counts in Galleria, 211.
Digenetic fluke Paramonostomum, anatomy of,
= 133.
trematode, morphology, life-history and sys-
tematics of, 266.
Digestive enzymes of holothurian gut, 354.
Digestive system of holothurian, 337.
Diphenylamine, effect of on Dictyostelium pig-
ments, 413.
Dissolved nutrients, utilization of by starfish,
161.
Distribution of absorbed exogenous nutrients
in starfish tissues, 161.
of digestive enzymes in holothurian gut, 354.
of enzymes in Arbacia egg, 1.
Diurnal rhythm in Hantzschia, 44.
Domecia, studies of with relation to feeding
behavior and modifications of host, 56.
DOYLE, W. L. Vesiculated axons in haemal
vessels of an holothurian, Cucumaria, 329.
Dugesia, orientation of, 311.
E
Echinoderm, oxygen uptake and respiratory
stress in, 16.
tropical, nitrogenous excretion in, 34.
INDEX
425
eggs, cytoplasmic DNA in, 68.
distribution of enzymes in, 1.
gut, histochemistry of, 354.
haemal vessels, vesiculated axons in, 329.
Echinoderms, utilization of dissolved exo-
genous nutrients by, 161.
Ecology of Hantzschia, 44.
of Nereis, in relation to morphology of
nephridia, 362.
of sea urchins, in relation to respiratory
stress, 16.
Effect of light on spawning of ascidians, 222,
292.
Eggs, Arbacia, distribution of enzymes in, 1.
Bufo, glycogenolysis in, 299.
Rivulus, production of primary male gono-
chorists from, 174.
sea urchin, cytoplasmic DNA in, 68.
EICHENBAUM, D. M. See D. G. SHAPPIRIO,
108.
Electrocardiograms of bats, 320.
Electron microscopy of Cucumaria vesiculated
axons, 329.
Electrophysiology of Limulus heart, 381, 392.
Embryo, chick, induction of immunological
tolerance in, 38.
Embryology of Botryllus, 229.
Embryonic development, glycogenolysis dur-
ing, 299.
Embryos of brine shrimp, respiration during
hatching of, 156.
Emergence of Artemia embryos, respiration
during, 156.
EMERSON, D. X. Surface area respiration dur-
ing the hatching of encysted embryos of
the brine shrimp, Artemia, 156.
Encysted brine shrimp, respiration during
hatching of, 156.
Endocrine activity in larval Thermobia, 277.
Endopeptidase in holothurian gut, 354.
Endoskeletal cartilage in polychaete, 244.
Environmentally controlled induction of male
gonochorist fish, 174.
Enzyme secretion sites in holothurian gut, 354.
in silkmoth brain, 108.
Enzymes, amphibian, during development, 299.
digestive, of holothurian gut, 354.
hydrolytic, distribution of in Arbacia eggs, 1.
Epicarid crab, new species of, 147.
Epidermal absorption of exogenous nutrients
by starfish, 161.
Eptesicus, low-temperature responses of, 320.
Esterase in Arbacia eggs, 1.
in holothurian gut, 354.
Eudistylia, endoskeletal cartilage in, 244.
Exchange of water with soil by salamanders,
126.
Excretion, nitrogenous, in tropical sea urchin,
34.
Exogenous nutrients, utilization of by starfish,
161.
Feeding, effect of on seminal retention in fe-
male Aedes, 404.
habits of Domecia, 56.
Female Aedes, repeatedly mated, seminal loss
in, 404.
genital segments of Spirorbis, 91.
FERGUSON, J. C. Utilization of dissolved ex-
ogenous nutrients by the starfishes As-
terias and Henricia, 161.
Fertilization of Botryllus eggs in vitro, 229.
Fine structure of Cucumaria, 329.
Firebrat, larval, growth and activity of cor-
pora allata in, 277.
Fish, retinomotor rhythms in, 200.
self-fertilizing hermaphroditic, production of
primary male gonochorists by, 174.
FISH, J. D. The digestive system of the holo-
thurian, Cucumaria, 337, 354.
Food, role of in activity of corpora allata in
Thermobia, 277.
of Botryllus, 229.
GABA, effect of on Limulus heartbeat, 381.
Galleria, changes in hemocyte picture of, 211.
Gamete release by Botryllus, 229.
by ascidians, effect of light on, 222, 292.
Gametogenesis in Spirorbis, 91.
Gastropod larval development, relation of tem-
perature to, 253.
Genetic studies on Botryllus, 229.
Genital segments of Spirorbis, 91.
union in repeatedly mated female Aedes, 404.
Geographic orientation of planarians under
experimental conditions, 311.
Geophysical forces and photic stimulus pat-
terns, association-formation between, 311.
Gigantione, new species of, 147.
Glucose, exogenous, uptake of by starfish tis-
sues, 161.
Glycogenolysis in amphibian development, 299.
Goldfish, retinomotor rhythms in, 200.
Gonads of Spirorbis, 91.
Gonochorists, primary male, environmentally
controlled induction of, in Rivulus, 174.
Grafts, intracoelomic, induction of immuno-
logical tolerance by in chick embryo, 38.
GREGG, J. H. See S. O. STAPLES, 413.
Growth of corpora allata in larval Thermobia,
277.
426
INDEX
rate of Nassarius larvae at different tem-
peratures, 253.
Gut of holothurian, structure of, 337.
H
Hantzschia, diurnal migration rhythm in, 44.
HARRINGTON, R. W., JR. Environmentally
controlled induction of primary male gono-
chorists from eggs of the self-fertilizing
hermaphroditic fish, Rivulus, 174.
Hatching of brine shrimp, surface area respira-
tion during, 156.
Heartbeat, regulation of in Limulus, 381, 392.
of bats, in relation to ambient temperature,
320.
Hemal system of holothurian, structure of, 329.
Hemocyte changes in Galleria, 211.
Henricia, utilization of dissolved exogenous
nutrients by, 161.
Hermaphroditic fish, production of primary
male gonochorists by, 174.
Hibernating bats, responses of to changes in
temperature, 320.
Histochemistry of holothurian gut, 354.
of silkmoth brain, 108.
Histology of Spirorbis genital segments, 91.
of larval Thermobia, 277.
of marine polychaete, 244.
of Nereis nephridia, 362.
Holothurian, digestive system of, 337, 354.
hemal vessels, vesiculated axons in, 329.
Hormonal activity in larval Thermobia, 277.
Horseshoe crab, cardioregulation in, 381, 392.
Host of commensal crab, modifications of, 56.
Hyalophora, cholinesterase in brain of, 108.
Hydration, role of in excystment of Artemia
embryos, 156.
Hydrolytic enzymes, subcellular distribution of
in Arbacia eggs, 1.
5-Hydroxytryptamine, inhibition of cardioreg-
ulation in Limulus heart by, 392.
Immunological tolerance in chick embryo, 38.
Induction of immunological tolerance in chick
embryo, 38.
by light of gamete-shedding in ascidians, 222,
292.
Inhibition of Limulus heartbeat, 381, 392.
Insemination of repeatedly mated Aedes, 404.
Inter sexuality in fish, 174.
Intracoelomic grafts, induction of immuno-
logical tolerance by, in chick, 38.
Invertase of holothurian gut, 354.
Iodine reactions of Bufo egg glycogen, 299.
Isopod, new species of, from northern Pacific,
147.
JACKSON, C, AND R. E. BLACK. The sub-
cellular distributions of some hydrolytic
enzymes in unfertilized eggs of the sea
urchin, Arbacia, 1.
JOHANSEN, K., AND R. L. VADAS. Oxygen
uptake and responses to respiratory stress
in sea urchins, 16.
JOHN, K. R., M. SEGALL AND L. ZAWATZKY.
Retinomotor rhythms in the goldfish Car-
rasius, 200.
JONES, J. C. Changes in the hemocyte picture
of Galleria, 211.
Spermatocysts in Aedes, 23.
JONES, M. L. On the morphology of the
nephridia of Nereis, 362.
Juvenile hormone activity of Thermobia cor-
pora allata, 277.
Laboratory culture of Botryllus, 229.
LAMBERT, C. C., AND C. L. BRANDT. The ef-
fect of light on the spawning of Ciona, 222.
Larval development of Nassarius, relation of
temperature to, 253.
Galleria, changes in hemocyte counts of, 211.
Thermobia, growth and activity of corpora
allata of, 277.
Lasiurus, temperature responses of, 320.
LEAHY, SR. M. G. See A. SPIELMAN, 404.
Lepismatid, larval growth and activity of cor-
pora allata in, 277.
LEWIS, J. B. Nitrogenous excretion in the
tropical sea urchin Diadema, 34.
Life-duration of Rivulus in the laboratory, 174.
Life-history of Uniserialis, 266.
Light, effect of on retinomotor rhythms in
goldfish, 200.
role of in migration rhythm of Hantzschia,
44.
in orientation of planarians, 311.
Light-induction of gamete-shedding in ascid-
ians, 222, 292.
Limulus, cardioregulation in, 381, 392.
Lipase of holothurian gut, 354.
LOCKE, B. R. See D. G. SHAPPIRIO, 108.
Loss of semen in repeatedly mated Aedes, 404.
Low temperature, effect of on sex-determina-
tion in Rivulus, 174.
responses of bats, 320.
Lunar influence on orientation of planarians,
311.
Lytechinus eggs, cytoplasmic DNA in, 68.
M
Magnetic fields, role of in orientation of
planarians, 311.
INDEX
427
Male genital segments of Spirorbis, 91.
gonochorists, environmentally controlled in-
duction of, in Rivulus, 174.
Maltase in holothurian gut, 354.
Marine polychaete, endoskeletal cartilage in,
244.
MATHEWS, M. B. See P. PERSON, 244.
Mating, effect of on number of spermatocysts
in Aedes, 23.
of female Aedes, effect of on seminal re-
tention, 404.
Maturation of sperm in Aedes, 23.
Metabolism of hatching brine shrimp, 156.
of sea urchins, in relation to their ecology,
16.
Metamorphosis of silkmoth, cholinesterase in
brain during, 108.
Microflora, benthic, migration rhythms in, 44.
Migration rhythms in Hantzschia, 44.
MILKMAN, R. Genetic and developmental
studies on Botryllus, 229.
Mitochondrial DNA in sea urchin eggs, 68.
Modifications of coral host by commensal
crab, 56.
Moisture tension, role of in water economy of
salamander, 126.
Molecular structure of Dictyostelium carote-
noids, 413.
Molgula, light-induction of gamete shedding
in, 292.
Mollusc, relation of temperature to larval de-
velopment of, 253.
Molting of Thermobia, role of corpora allata
in, 277.
Monochromatic light, effect of on spawning
of Ciona, 222.
Morphogenesis of frog, glycogenolysis in, 299.
Morphology of genital segments of Spirorbis,
91.
of Gigantione, 147.
of holothurian digestive system, 337.
of Nereis nephridia, 362.
of Paramonostomum, 133.
of Uniserialis, 266.
Mosquito, repeatedly mated female, seminal
loss in, 404.
spermatocysts in, 23.
Moth, wax, changes in hemocyte picture of,
211.
Mouthparts of commensal crab, 56.
Mudsnail, relation of temperature to larval
development of, 253.
Multiple insemination of Aedes, 404.
MUN, A. M., L. B. CRITTENDEN AND B. J.
CLARKE. Induction of immunological tol-
erance by intracoelomic grafts in the 4-
day chick embryo, 38.
Myotis, temperature responses of, 320.
N
Nassarius, relation of temperature to larval
development of, 253.
Nephridia of Nereis, morphology of, 362.
Nereis, morphology of nephridia of, 362.
Neurogenic heartbeat of Limulus, 381, 392.
Neurophysiology of Cecropia brain during dia-
pause, 108.
Neurosecretory action in silkmoth brain, rela-
tion of cholinesterase to, 108.
in holothurian, possible evidence for, 329.
New species of Gigantione, 147.
of Uniserialis, 266.
Nitrogenous excretion in tropical sea urchin,
34.
Northern Pacific Gigantione, 147.
Nucleic acid distribution in Arbacia egg, 1.
of sea urchin egg mitochondria, 68.
Numbers of spermatocysts in Aedes at vari-
ous stages, 23.
Nutrients, dissolved exogenous, utilization of
by starfish, 161.
Nutrition, role of in activity of corpora allata
in larval Thermobia, 277.
Oenocytoid counts in Galleria, 211.
Oogenesis in Spirorbis, 91.
Osmotic relations in Nereis, 362.
Osteoid tissue in marine polychaete, 244.
Ova, sea urchin, cytoplasmic DNA in, 68.
Oxygen consumption of hatching brine shrimp,
156.
of sea urchins, 16.
Pacific, new species of Gigantione from, 147.
PALMER, J. D., AND F. E. ROUND. Persistent,
vertical-migration rhythms in benthic mi-
croflora. VI, 44.
Paramonostomum, anatomy of, 133.
Parasite of birds, morphology and taxonomy
of, 266.
PARK, Y. H. See F. A. BROWN, JR., 311.
Patterns, stimulus, association-formation be-
tween photic and geophysical, 311.
PATTON, W. K. Studies on Domecia, a com-
mensal crab, 56.
PAX, R. A., AND R. C. SANBORN. Cardio-
regulation in Limulus, 381, 392.
Persistent migration rhythms in Hantzschia,
44.
rhythm in goldfish retina, 200.
PERSON, P., AND M. B. MATHEWS. Endo-
skeletal cartilage in a marine polychaete,
Eudistylia, 244.
428
INDEX
Photic and geophysical stimulus patterns, as-
sociation-formation between, 311.
Physical properties of sea urchin egg cyto-
plasmic DNA, 68.
Picrotoxin, action of on Limulus heart, 381.
Pigmentation of Botryllus, inheritance of, 229.
Pigments, carotenoid, in Dictyostelium, 413.
PIKO, L., A. TYLER AND J. VINOGRAD.
Amount, location, priming capacity, cir-
cularity, and other properties of cytoplas-
mic DNA in sea urchin eggs, 68.
Pipistrellus, temperature responses of, 320.
Planarians, orientation of, 311.
Plasmatocyte levels in Galleria, 211.
Plethodon, water economy of, 126.
Polychaete, endoskeletal cartilage in, 244.
genital segments of, 91.
morphology of nephridia of, 362.
Polyene structure of Dictyostelium carotenoid
pigments, 413.
POTSWALD, H. E. Observations on the genital
segments of Spirorbis, 91.
Primary male gonochorists, environmentally
controlled induction of, in Rivulus, 174.
Priming capacity of cytoplasmic DNA in sea
urchin eggs, 68.
Prohemocyte levels in Galleria, 211.
Proteases in holothurian gut, 354.
Pseudotriton, water economy of, 126.
Puerto Rico commensal crab, effects of on
host, 56.
Pupal Aedes, spermatocyst counts of, 23.
diapause of silkmoth, cholinesterase in brain
during, 108.
R
RN-ase in Arbacia eggs, distribution of, 1.
Radiocarbon-labeled exogenous nutrients, up-
take of by starfish, 161.
RAISMAN, J. S. See F. D. BARBIERI, 299.
Regulation of heartbeat in Limulus, 381, 392.
REITE, O. B. See W. H. DAVIS, 320.
Relationship of temperature to larval develop-
ment of Nassarius, 253.
Repeatedly mated Aedes, seminal loss in, 404.
Reproductive cycle of Botryllus, 229.
Respiration during hatching of Artemia, 156.
Respiratory stress in sea urchins, 16.
Responses of bats from temperate regions to
changes in ambient temperature, 320.
Retinomotor rhythms in goldfish, 200.
Rhythms, retinomotor, in goldfish, 200.
vertical-migration in Hantzschia, 44.
Rivulus, production of primary male gono-
chorists by, 174.
Rods of goldfish retina, rhythms in behavior
of, 200.
ROUND, F. E. See J. D. PALMER, 44.
Salamanders, water economy of, 126.
Salinity, in relation to morphology of Nereis
nephridia, 362.
Samia, cholinesterase in brain of, 108.
SANBORN, R. C. See R. A. PAX, 381, 392.
Scales, role of corpora allata in development
of, in Thermobia, 277.
SCHELTEMA, R. S. The relationship of tem-
perature to the larval development of
Nassarius, 253.
Sea urchin, tropical, nitrogenous excretion in,
34.
eggs, cytoplasmic DNA in, 68.
distribution of enzymes in, 1.
Sea urchins, oxygen uptake and respiratory
stress in, 16.
SEGALL, M. See K. R. JOHN, 200.
Self-fertilizing hermaphroditic fish, production
of primary male gonochorists by, 174.
Seminal loss in repeatedly mated female Aedes,
404.
Sex determination in Rivulus, 174.
Sexual activity of Aedes, as related to num-
ber of spermatocysts, 23.
SHAPPIRIO, D. G., D. M. EICHENBAUM and B.
R. LOCKE. Cholinesterase in the brain of
the Cecropia silkmoth during metamor-
phosis and pupal diapause, 108.
Shedding of gametes in ascidians, as induced
by light, 222, 292.
Silkmoth, cholinesterase in brain of, 108.
SKAFF, V. See A. SPIELMAN, 404.
Skin grafts in chicks, 38.
Slime mold, carotenoid pigments in, 413.
Soil, role of in water economy of salamander,
126.
Spawning of ascidians, as induced by light,
222, 292.
of Nassarius, relation of temperature to, 253.
Species-differences in responses of bats to
ambient temperatures, 320.
Sperm, loss of in repeatedly mated female
Aedes, 404.
Spermatocysts in Aedes, 23.
Spermatogenesis in Spirorbis, 91.
Spherule cell counts in Galleria, 211.
SPIELMAN, A., SR. M. G. LEAHY AND V.
SKAFF. Seminal loss in repeatedly mated
female Aedes, 404.
SPIGHT, T. M. The water economy of sala-
manders: Exchange of water with the
soil, 126.
Spirorbis, genital segments of, 91.
STAPLES, S. O., and J. H. GREGG. Carotenoid
pigments in the cellular slime mold Dic-
tyostelium, 413.
INDEX
429
Starfish, utilization of dissolved exogenous
nutrients by, 161.
Stimulus patterns, association-formation in,
311.
Strongylocentrotus, oxygen uptake and res-
piratory stress in, 16.
eggs, cytoplasmic DNA in, 68.
Structure of holothurian gut, 337.
Studies on trematode genus Paramonostomum,
133.
STUNKARD, H. W. The morphology, life-
history, and systematic relations of the
digenetic trematode Uniserialis, 266.
Studies on the trematode genus Paramono-
stomum, 133.
Subcellular distributions of hydrolytic enzymes
in unfertilized Arbacia eggs, 1.
Sulfatase in Arbaca eggs, 1.
Surface area respiration during hatching of
Artemia, 156.
Synthesis of pigment in Dictyostelium, 413.
Systematics of genus Paramonostomum, 133.
of Gigantione, 147.
of Uniserialis, 266.
Taxonomy of Gigantione, 147.
of Paramonostomum, 133.
of Uniserialis, 266.
Teleost, retinomotor rhythms in, 200.
self-fertilizing hermaphroditic, production of
primary male gonochorists in, 174.
Temperate-region bats, responses of to changes
in ambient temperature, 320.
Temperature, relationship of to development
of Nassarius, 253.
role of in sex-determination of Rivulus, 174.
responses of bats, 320.
Thermobia, larval, growth and activity of
corpora allata in, 277.
Thysanuran, larval, growth and activity of
corpora allata in, 277.
Tidal rhythm in Hantzschia, 44.
Time of shedding of ascidian gametes, in re-
lation to light, 222, 292.
Toad, glycogenolysis and amylase in develop-
ment of, 299.
Tolerance, immunological, in chick embryo, 38.
Trematodes, studies on, 133, 266.
Tropical sea urchin, nitrogenous excretion in,
34.
TYLER, A. See L. PIKO, 68.
U
Ultrastructure of Cucumaria, 329.
Unfertilized Arbacia eggs, distribution of en-
zymes in, 1.
Uniserialis, morphology, life-history and sys-
tematics of, 266.
Uptake of amino acids by starfish, 161.
of oxygen by sea urchins, 16.
Utilization of dissolved exogenous nutrients
by starfishes, 161.
VADAS, R. L. See K. JOHANSEN, 16.
Vascular system of Botryllus, 229.
tissue of holothurian, electron microscopy
of, 329.
Vesiculated axons in haemal vessels of holo-
thurian, 329.
VINOGRAD, J. See L. PIKO, 68.
W
Water economy of salamander, 126.
temperature, relation of to larval develop-
ment of Nassarius, 253.
WATSON, J. A. L. The growth and activity
of the corpora allata in the larval firebrat
Thermobia, 277.
Wave-length of light, role of in gamete-shed-
ding of ascidians, 222, 292.
Wax moth, changes in hemocyte picture of,
211.
WHITTINGHAM, D. G. Light-induction of
shedding of gametes in Ciona and Mol-
gula, 292.
Yolk DNA in sea urchin eggs, 68.
ZAWATZKY, L. See K. R. JOHN, 200.
Zeta-carotenes in Dictyostelium, 413.
Zoogeography of Domecia, 56.
Volume 132 Number 1
THE
BIOLOGICAL BULLETIN
PUBLISHED BY
THE MARINE BIOLOGICAL LABORATORY
Editorial Board
JOHN M. ANDERSON, Cornell University JOHN H. LOCHHEAD, University of Vermont
T i T *•*. / TI 1*1. LEONARD NELSON. Emory University
JOHN B. BUCK, National Institutes of Health
MELVIN SPIEGEL, Dartmouth College
PHILIP B. DUNHAM, Syracuse University WM RANDOLPH TAYLOR) University of
W. D. RUSSELL HUNTER, Syracuse University Michigan
ANNA R. WHITING, Oak Ridge National
SHINYA INOU£, University of Pennsylvania Laboratory
J. LOGAN IRVIN, University of North Carolina CARROLL M. WILLIAMS, Harvard University
DONALD P. COSTELLO, University of North Carolina
Managing Editor
FEBRUARY, 1967
fvhrne Riclog'.oj! Laboratory
LIBRARY
MAR 1 4 1967
WOODS HOLE, MASS.
Printed and Issued by
LANCASTER PRESS, Inc.
PRINCE & LEMON STS.
LANCASTER, PA.
INSTRUCTIONS TO AUTHORS
The Biological Bulletin accepts papers on a variety of subjects of biological interest. In
general, however, review papers (except those written at the specific invitation of the Editorial
Board), very short papers, preliminary notes, and papers which describe only a new technique or
method without presenting substantial quantities of data resulting from the use of the new method
cannot be accepted for publication. A paper will usually appear within three months of the date
of its acceptance.
The Editorial Board requests that manuscripts conform to the requirements set below;
those manuscripts which do not conform will be returned to authors for correction before review
by the Board.
1. Manuscripts. Manuscripts must be typed in double spacing (including figure legends,
foot-notes, bibliography, etc.) on one side of 16- or 20-lb. bond paper, 8£ by 11 inches. They
should be carefully proof-read before being submitted and all typographical errors corrected
legibly in black ink. Pages should be numbered. A left-hand margin of at least li inches
should be allowed.
2. Tables, Foot-Notes, Figure Legends, etc. Tables should be typed on separate sheets and
placed in correct sequence in the text. Because of the high cost of setting such material in type,
authors are earnestly requested to limit tabular material as much as possible. Similarly, foot-
notes to tables should be avoided wherever possible. If they are essential, they should be indi-
cated by asterisks, daggers, etc., rather than by numbers. Foot-notes in the body of the text
should also be avoided unless they are absolutely necessary, and the material incorporated into
the text. Text foot-notes should be numbered consecutively and typed double-spaced on a sepa-
rate sheet. Explanations of figures should be typed double-spaced and placed on separate sheets
at the end of the paper.
3. A condensed title or running head of no more than 35 letters and spaces should be included.
4. Literature Cited. The list of papers cited should conform exactly to the style set in a
recent issue of The Biological Bulletin; this list should be headed LITERATURE CITED,
and typed double-spaced on separat ; pages.
5. Figures. The dimensions of the printed page, 5 by 7| inches, should be kept in mind in
preparing figures for publication. Illustrations should be large enough so that all details will be
clear after appropriate reduction, but not larger than 15 X 22 inches. Explanatory matter should
be included in legends as far as possible, not lettered on the illustrations. Figures should be pre-
pared for reproduction as line cuts or halftones ; other methods will be used only at the author's
expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper, good
quality tracing cloth or blue-lined coordinate paper; those to be reproduced as halftones should be
mounted on Bristol Board, and any designating numbers or letters should be made directly on the
figures. All figures should be numbered in consecutive order, with no distinction between text-
and plate-figures. The author's name should appear on the reverse side of all figures, as well as
the desired reduction.
6. Mailing. Manuscripts should be packed fiat. All illustrations larger than 85 by 1 1 inches
must be accompanied by photographic reproductions or tracings that may be folded to page size.
Reprints. Reprints may be obtained at cost ; approximate prices will be furnished by the
Managing Editor upon request.
THE BIOLOGICAL BULLETIN
/
THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and
Lemon Streets, Lancaster, Pennsylvania.
Subscriptions and similar matter should be addressed to The Biological Bulletin, Marine
Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and
Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers,
$3.75. Subscription per volume (three issues), $9.00, (this is $18.00 per year for six issues).
\
Communications relative to manuscripts should be sent to Dr. Donald P. Costello, Marine
Biological Laboratory, Woods Hole, Massachusetts, between June 15 and September 1, and to
Dr. Donald P. Costello, P. O. Box 429, Chapel Hill, North Carolina, during the remainder of
the year.
Copyright © 1967, by the Marine Biological Laboratory
Second-class postage paid at Lancaster, Pa.
MBL WHOI LIBRARY
UH 1B1M N
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11