ON THE ECOLOG Y OF

COENOBIT A CL YPEATUS IN CURA<;AO

 ON THE ECOLOGY OF
COENOBITA CLYPEATUS IN CURAQAO
with reJerence to reproduction, water economy
and osmoregulation in terrestrial hermit crabs

ACADEMISCH PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE

WISKUNDE EN NATUURWETENSCHAPPEN AAN DE
UNIVERSITEIT VAN AMSTERDAM, OP GEZAG VAN DE
RECTOR MAGNIFICUS DR. A. DE FROE, HOOGLERAAR IN DE
FACULTEIT DER GENEESKUNDE, IN HET OPENBAAR TE
VERDEDIGEN IN DE AULA DER UNIVERSITEIT (TIJDELIJK
IN DE LUTHERSE KERK, INGANG SINGEL 411, HOEK SPUl)
OP WOENSDAG 7 NOVEMBER 1973, DES NAMIDDAGS
TE 13.30 UUR

door

PETER ARNOUD WILLEM J ACOBUS DE WILDE

GEBOREN TE HEEMSKERK

Springer-Science+Business Media, B.V 1973

 Promotor: PROF. DR. A. PUNT

Co-referent: PROF. DR. J. H. STOCK

Dit proefschrift verschijnt tevens in Studies on the Fauna of Cura<;ao and

other Caribbean Islands Vol. 44.

ISBN 978-94-017-6702-6 ISBN 978-94-017-6768-2 (eBook)

DOI 10.1007/978-94-017-6768-2
STUDIES ON THE FAUNA OF CURA~AO AND OTHER
CARIBBEAN ISLANDS: No. 144.

ON THE ECOLOGY OF COENOBITA CLYPEATUS

IN CURAQAO
WITH REFERENCE TO REPRODUCTION, WATER ECONOMY
AND OSMOREGULATION IN TERRESTRIAL HERMIT CRABS

by

P. A. W. J. DE WILDE
(Carmabi, Cura<;ao I Dierfysiologisch Lab.,
Univ. v. Amsterdam)

CONTENTS

Page
GENERAL INTRODUCTION • . . • • . • • . • . 3
Remarks on the species - Acknowledgements

FIELD WORK

1. MIGRATION AND REPRODUCTION

Coastal and inland animals . 7
Size classes. . . . . . . . 9
Migration and Reproduction 9
Methods . . . . . . . . . 11
Tags - Locality of field experiments - Time - General proceedings
Experiments and results . . . . . . . . . 18
Discussion . . . . . . . . . . . . . . . 31
Reproductive season and mating places 32
Migrations . . . . . . . . . . . . 34
Mating. . . . . . . . . . . . . . 38
Egg production and egg development 40
Egg deposition . . . . . . . . 43
Successive reproduction periods . . . 47
Retreat to inland habitats . . . . . 48

II. ADDITIONAL NOTES ON THE ECOLOGY OF COENOBITA

Habitat . . 50
Temperature 52
Food. 53
Water . . . 54
 LABORATORY EXPERIMETS

Hr. WATER MANAGEMENT

Introduction 56
Material . . . . . . 57
Mefuods . . . . . . 58
General water economy 60
Uptake . . . . . . 61
Preference experiments. 62
Soi! humidity - Air humidity - Salinity
Retainment and 1055. . . . . . . . . . . 64
Evaporation on the respiratory surface - Evaporation by the
tegument - Loss of water by excretion - Loss of water by
ecdysis
Nocturnal activity. . . . . . . . . . . . . . . . . . . . . . 65
Introduction - Method - Results and interpretation
Limestone as a source of water . . . . . . 70
Introduction - Experiments and results
The shell as a protection against desiccation 73
Discussion . . . . . . . . . . . . . . . . . 78
IV. SHELL WATER AND ITS REGULATION
Introduction . . . . . . . . . . 83
Uptake and storage of shell water . 84
Experiments . . . . . . . . . . 88
Uptake of water of a fixed concentration . 88
Method - Results and interpretation
Uptake of water in the drinking apparatus 93
Method - Results and interpretation
Regulation by mixing 0% and 100% sea water . . . . . . . . . 96
Introducdon and mefuod - Differences between small and large
animals - Mixing of other concentrations - Time as a factor
influencing regulation - Influence of air humidity on regulation
Field observations. 105
Final conclusions 107
V. OSMOREGULATION
Introduction . . 109
Experiments and observations. . . . . . . . . . . . . . . . 110
The effect of dehydration on concentration of the body fluid 113
Methods - Results - The effect of reviving - Discussion
Osmoregulation, when water in any form is present . . . . 116
Introduction - The effect of osmotic stress on mortality - The
effect of osmotic stress on body fluid concentration - Endurance
test 1 - Endurance test 2 - The non-electrolyte component of
the blood - Osmotic concentration in the blood - Isotonicity of
the urine
Final conclusions 130
Summary - Resumen. 132
REFERENCES • • • 136
 GENERAL INTRODUCTION

Among Crustacea terrestrial species are unusual, and only small

numbers of species from various taxa have been more or less success-
ful in occupying the land. In doing so they deviated from the normal
evolutionary path from the primitive marine environment through
brackish and fresh water to marshy and land areas; in this case
some groups went straight from sea to land.
For several reasons connected with locomotion, gas exchange,
water regulation, ion regulation and osmoregulation and the ex-
treme variability of the external environment, a terrestrial way of
life presents many problems. When a direct step is made from sea to
land, particularly under hot and arid climatological conditions,
such problems occur simultaneously.
All kinds of transitional stages may be distinguished. Among the
Decapoda in the Netherlands Antilles only two brachyuran species
and a hermit crab are genuine land crabs. For many days these
animals can stay in dry habitats without any visible uptake of
water. They are adapted to live almost completely on fresh water
and are able to roam far inland. For the development of the larvae,
however, they are still dependent on the sea.
Earlier papers on the physiology and ecology of terrestrial deca-
pod Crustacea are scarce. Mainly by the work of GROSS, during the
last decade, the subject finally received the attention it deserved on
ac count of its promising possibilities. A symposium on 'Terrestrial
adaptations in Crustacea' (1968) held in 1967 did not include
research on hermit crabs.
 4

From an evolutionary point of view, land hermit crabs are

considered to be a relatively young group. Systematic and morpho-
logical characters provide evidence in favour of this view. As far as
known no fossils have ever been found which could be identified
with certainty as belonging to this group.
During the author's stay in Cura<;ao from 1965 to 1968 (DE WILDE,
1969) special attention was paid to problems concerning water
economy and osmotic regulation in Coenobita clypeatus (Herbst).
The Caribbean Marine Biological Institute (Carmabi) at Willemstad
offered adequate laboratory facilities. Field experiments were main-
ly carried out on the easternmost part of the island (Oostpunt). The
lack of details on both ecology and major aspects of the life history
of the species, such as reproduction, justified additional research.
Though at first the field research was only a side-line, Chapter I
is mainly devoted to its results. In this chapter full stress is laid
upon the way those crabs manage to keep alive in their inhospitable
habitats in Cura<;ao. Chapter II gives further information on the
ecology of Coenobita. It appeared from visits to Bimini (Bahamas),
the Miami area, Florida Keys, Jamaica, Aruba, Bonaire, Trinidad
and Barbados, that Coenobita clypeatus may be found in widely
diverging habitats. All the same a habitat as studied on Cura<;ao is
considered to be fairly typical for the species.
In Chapters III to V - on Water management, Shell water and
Osmoregulation respectively - adjustment problems are discussed
more closely, mainly based on results obtained by relevant labo-
ratoryexperiments.

REMARKS ON THE SPECIES

The small family Coenobitidae, which is mainly restricted to the
tropics, comprises two genera only: Birgus with a single species
(B. latro L., the coconut- or robber-crab), and Coenobita (the land
hermit crabs) with some seven species.
This paper mainly deals with Coenobita clypeatus (Herbst), the
only species in the western Atlantic area. (For a description of the
species see CHASE & HOBBs, 1969). Though the material is considered
to be uniform all over the area, a great many specimens were seen
 5

on Bimini which could be distinguished from the Cura<;ao an im als

especially by the conspicuous yellow colour of ambulatories and
major cheliped, and by the golden yellow hairs around the male
gonopore. Among the many thousands of animals examined in
Cura<;ao there were some brick-red colour varieties and occasionally
an albino.
According to PROVENZANO (1959) the species' distribution ranges
from Bermuda, southern Florida and the Bahamas, through the
West Indies to southern Brazil (CHASE & HOBBS, 1969, think that
the Brazilian records need verification). It occurs in the coastal
regions on to several miles inland, including considerable altitudes,
e.g. up to the summit of the 887 m high Mountain of Saba. The
largest numbers, however, are found in the more arid parts of its
distribution al area. As far as known the species seems to avoid
places with a high rain fall and a luxuriant vegetation; it is rare or
absent along the 'green coast' of the South-American continent.
In the Leeward Islands of the Netherlands Antilles Coenobita
clypeatus - locally called "soldaatje" (little soldier) - is a very
common and conspicuous an im al. It is also common in the Wind-
ward Islands of the Netherlands Antilles (St. Martin, Saba and St.
Eustatius), situated about 800 km to the North.

Part of the material on whieh this study has been based is preserved in the
Zoologie al Museum of the State University, Utreeht.

ACKNOWLEDGEMENTS
The author wishes to express his gratitude to Prof. Dr. A. PUNT for his guidanee,
aid and adviee in eonnection with all phases of the work. Sineere thanks are also
due to Prof. Dr. J. H. STOCK for his valuable eomments on the manuscript, to
Dr. W. VERVOORT for his critical reading ofthe manuscript, to Dr. L. B. HOLTHUIS for
solving some problems of nomenc1ature and to Dr. 1. KRISTENSEN for directing my
attention to the group of terrestrial decapods and for introducing me to the Carmabi
on Cura9ao.
To Dr. F. CREUTZBERG and the staff of the Institute I offer my thanks for the
hospitality I reeeived during my stay on the island. Prof. dr. A. J. PROVENZANO Jr.
of the University of Miami I thank for his interest and valuable discussion, and Mr.
R. F. MATHEWSON, resident director of the Lerner Marine Laboratory, for his kind
aid and hospitality.
I owe a debt of gratitude to Mr. RUBEN ANGEL, who assisted me during the
 6

nocturnal wanderings on Oostpunt, and to Mr. J. P. MAAL, the owner of the estate,
who gave me permission to enter his properties.
Appreciations are due to the management and my colleagues of the Netherlands
Institute for Sea Research for their patience in offering me the required time and
facilities.
Miss S. M. VAN DER BAAN translated the manuscript into English, Mr. C. MARTIS
turned the summary into Spanish, and Mr. H. HOBBELINCK made the drawings and
the laboratory photographs.
In particular I am indebted to Dr. P. WAGENAAR HUMMELINCK, and to Drs.
LOUISE J. VAN DER STEEN, both from the Zoological Laboraty of the University of
Utrecht, for smoothing the long and difficult path of preparing the manuscript for
the press, and for reading the proofs.
This study has been made possible by financial support from the Netherlands
Foundation for the Advancement of Tropical Research (WOTRO). I acknowledge
the valuable contacts with Dr. J. H. WESTERMANN. A stay in Bimini and some
short visits to Florida and Jamaica were supported by a grant of the U.S. Naval
Research (O.N.R. nr. 552-07).
 FIELDWORK

1. MIGRATION AND REPRODUCTION

COASTAL AND INLAND ANIMALS

The populations of Coenobita clypeatus observed in Cura<;ao may

be roughly subdivided into two different ecological types: coastal
and inland.
Typical coastal animals occupy the habitats a~ng the sea co ast
and, in places, along the banks of the so-called inner bays and
lagoons. Food is rather scarce on the shorelines, and numerous
other crustaceans such as the quicker and more active semi-
terrestrial grapsoid crabs, are formidable food competitors. Popu-
lations of coastal animals are characterized by the large percentages
of very small and small specimens, while the larger ones are often
badly housed, occupying shells which are too small for their inhabi-
tants or shells with large holes, cracks or worn interiors. Shells of
quite unsuitable types, coming from various species such as Murex,
M elongena, Thais, Latirus, Bulla, etc. are also common. Moreover
lots of shells are often attacked by boring sponges or are covered by
thick cmsts of calcareous algae. Coastal animals always look more
or less shabby.
Coastal animals depend mainly on sea water for their water
supply; water stored in the shells was found in only a negligible
number of cases. As will be explained more fully hereafter, coastal
animals are more dependent on the sea.
Typical inland animals, on the other hand, give an impression of
prosperity. Small animals are relatively fewer in numbers. The
majority of the occupied shells came from the species Livona pica
 8

(L.) (PI. Ia, IVa). Shells of Livona (= Cittarium) are certainly the
preferred ones in Cura<;ao, as was proved in a choice-experiment
on a small scale.
The inland habitats are characterized by the occurrence of fresh-
water drin king spots, shady trees and much bett er food conditions.
Inland animals fully depend on fresh water for their needs. Shell
water was usually found to be present to a certain amount.
Both types, with all sorts of transitional stages occurring in be-
tween, become distinctly separated in the dry season after long
periods of drought. In fact the existence of the two types of popu-
lations is largely based on the possession of weIl-fitting and reliable
shells. Very small specimens of Coenobita and also the larger ones
with ill-fitting shells are supposed to be unable, or at least to have a
poor chance, of ever reaching the favourable inland habitats. Lack
of shell water in the dry season makes travelling increasingly diffi-
cult. Breakage of shells will be fatal in most cases, as the chance of
finding a new one in the interior is extremely small, and lethaI
desiccation of the unprotected animals will be only a matter of
hours. In the inland region animals are worse off after moulting than
those on the co ast which have a variety of shells at their disposaI.
It will be evident that only properly equipped animals are able to
invade the favourable free inland habitats and to occupy them
successfuIly. In particular a good shell will provide the possibility
for land hermit crabs to live in such an advanced stage of terrestrial
life. Inland animals originate from coastal animal populations but
on the other hand an inland animal mayaIso go back to the stage of
a coastal animaI.

Für the Pacific Coenobita clupeatus auct. (= C. brevimanus Dana) HARMS (1932)
also distinguished between two types of animals, which he calls "Formenkreisen."
The terrestrial form inhabiting the dry coral foreshores, e.g. in the Moluccas up to
an altitude of 80-100 m, may be distinguished from the coastal form - living in
the zone next to the line of low water - by a number of modifications, due to the
transition from an aquatic to a terrestrial way of life. According to HARMS even
morphological differences may be observed, for instance in the covering membranes
of the statocysts. However, the observations described in the present paper, con-
cerning the related species C. clypeatus (Herbst) raise some doubt as to the system-
atic value of the differences described by HARMS.
 9

SIZE CLASSES

Old specimens of Coenobita clypeatus, living in shells of Livona or

Astraea species may be as big as a man's fist, re ac hing gross weights
up to 500 g, whereas small, but already sexually mature animals in
Tudora, Tectarius or other sm all and light shells only re ach gross
weights of about two grammes.
Both, for field observations and laboratory experiments, it proved
useful to subdivide the animals into classes, on account of the widely
divergent weights and sizes. In practice the subdivision into 5
different size classes - strictly gross weight groups - was most
sa tisfactory.

SIZE CLASS ABBREVIATION GROSS WEIGHT OF

Livona - ANIMALS

very small v.s. <lOg

sm all s. 10--20 g
medium m. 20-50 g
latge I. 50-100 g
very large v.I. >100g

Though this division is based on weights of land hermit crabs in

shells of the gastropod Livona pica (L.) only, this does not compli-
cate matters, since the majority of the bigger animals occupies
shells of this species and nearly all experiments concern hermit
crabs in Livona-shells. Anyway, individuals in shells of Astraea
tecta (Lightfoot) and Astraea caelata (Gmelin), respectively the
second and third choice of larger Coenobita in Curayao, also fit
satisfactorily into the above mentioned system.
With some experience it is wen possible, as was proved by re-
peated experiments, to distinguish at sight between Livona and
Astraea-animals with relatively few mistakes.

MIGRATION AND REPRODUCTION

Since the migration of land hermit crabs is known to occur all

over the Caribbean, it is very remarkable that the phenomenon has
received so little attention from scientists. Scattered over the whole
 10

area there are natives, in particular among elderly fishermen, who

ean tell stories about the clattering noise of huge numbers of soldier
erabs, proeessions of several hundreds of meters long, at bright
moonlight or, as others will have it, at piteh-dark nights. In the
Netherlands Antilles a traditional story is told about invaders who
took flight after they heard - literally by putting their ears to the
ground - that many soldiers (soldier erabs) were on the march.
Beeause of the soft abdomens of the animals, which seem to be
exeellent bait for fishing, most Antlllean fishermen show mueh
interest in these ereatures and know when they are absent, or at
least diffieult to find, during several months of the year in those
plaees where they are usually found. From our own experienee we
know that the experimental animals were hard to obtain from about
J uly till Oetober.
There is no doubt that the terrestrial hermit erabs of the Family
Coenobitidae, which have all beeome adapted to life on land and
often spend the major part of their life far - up to several miles -
from the eoast, reproduee by means of larvae liberated into the sea.
Based on animals reared in the laboratory, following two hatehings
in 1960 on 22 J uly and 24 August, and two more hatehings in 1961,
both on 10 August, PROVENZANO (1962) gives a deseription of the
eomplete larval development of Coenobita clypeatus. From his ex-
periments it is clear that the development of the larvae takes plaee
in normal seawater.
VERRILL (1908) reports the occurrence of Coenobita clypeatus on sand hills in
the Bermudas, far from the coast. PEARSE (1915) already mentioned a procession
of about 500 individuals, moving over an area of about 200 m square, apparently
migrating from the ocean through the desert of Santa Marta (Colombia), on August
29, 1913.
In a - rather imaginative - popular publication CARPENTER & LOGAN (1945) give
an eyewitness account of two occasions on which they observed a "jamboree" of
Coenobitas at Mona Island, halfway Puerto Rico and Hispaniola. They described
how towards the end of August ten thousands of animals struggled along to one
sandy beach.
None of the collected female specimens on Dominica in January, February,
March and June by CHASE & HOBBs (1969) were ovigerous.
All the aforesaid communications show that the events dealing with the repro-
duction of Coenobita clypeatus are to be expected in aperiod ranging from about
the middle oi July till the end oi August.
 ON THE ECOLOG Y OF
COENOBIT A CL YPEATUS IN CURA<;AO
 ON THE ECOLOGY OF
COENOBITA CLYPEATUS IN CURAQAO
with reJerence to reproduction, water economy
and osmoregulation in terrestrial hermit crabs

ACADEMISCH PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE

WISKUNDE EN NATUURWETENSCHAPPEN AAN DE
UNIVERSITEIT VAN AMSTERDAM, OP GEZAG VAN DE
RECTOR MAGNIFICUS DR. A. DE FROE, HOOGLERAAR IN DE
FACULTEIT DER GENEESKUNDE, IN HET OPENBAAR TE
VERDEDIGEN IN DE AULA DER UNIVERSITEIT (TIJDELIJK
IN DE LUTHERSE KERK, INGANG SINGEL 411, HOEK SPUl)
OP WOENSDAG 7 NOVEMBER 1973, DES NAMIDDAGS
TE 13.30 UUR

door

PETER ARNOUD WILLEM J ACOBUS DE WILDE

GEBOREN TE HEEMSKERK

Springer-Science+Business Media, B.V 1973

 Promotor: PROF. DR. A. PUNT

Co-referent: PROF. DR. J. H. STOCK

Dit proefschrift verschijnt tevens in Studies on the Fauna of Cura<;ao and

other Caribbean Islands Vol. 44.

ISBN 978-94-017-6702-6 ISBN 978-94-017-6768-2 (eBook)

DOI 10.1007/978-94-017-6768-2
STUDIES ON THE FAUNA OF CURA~AO AND OTHER
CARIBBEAN ISLANDS: No. 144.

ON THE ECOLOGY OF COENOBITA CLYPEATUS

IN CURAQAO
WITH REFERENCE TO REPRODUCTION, WATER ECONOMY
AND OSMOREGULATION IN TERRESTRIAL HERMIT CRABS

by

P. A. W. J. DE WILDE
(Carmabi, Cura~ao I Dierfysiologisch Lab.,
Univ. v. Amsterdam)

CONTENTS

Page
GENERAL INTRODUCTION . . • • • • . • • • • 3
Remarks on the species - Acknowledgements

FIELD WORK

I. MIGRATION AND REPRODUCTION

Coastal and inland animals . 7
Size classes. . . . . . . . 9
Migration and Reproduction 9
Methods . . . . . . . . . 11
Tags - Locality of field experiments - Time - General proceedings
Experiments and results . . . . . . . . . 18
Discussion . . . . . . . . . . . . . . . 31
Reproductive season and mating places 32
Migrations . . . . . . . . . . . . 34
Mating. . . . . . . . . . . . . . 38
Egg production and egg development 40
Egg deposition . . . . . . . . 43
Successive reproduction periods. . . 47
Retreat to inland habitats . . . . . 48

11. ADDITIONAL NOTES ON THE ECOLOGY OF COENOBITA

Habitat . . 50
Temperature 52
Food. 53
Water . . . 54
 LABORATORY EXPERIMETS

IH. WATER MANAGEMENT

Introduction 56
Material . . . . . . 57
Methods . . . . . . 58
General water economy 60
Uptake . . . . . . 61
Preference experiments. 62
Soil humidity - Air humidity - Salinity
Retainment and loss. . . . . . . . . . . 64
Evaporation on the respiratory surface - Evaporation by the
tegument - Loss of water by excretion - Loss of water by
ecdysis
Nocturnal activity. . . . . . . . . . . . . . . . . . . . . . 65
Introduction - Method - Results and interpretation
Limestone as a source of water . . . . . . . . . . . . . . . . 70
Introduction - Experiments and results
The shell as a protection against desiccation 73
Discussion . . . . . . . . . . . . . . . . . 78
IV. SHELL WATER AND ITS REGULATION
Introduction . . . . . . . . . 83
Uptake and storage of shell water . 84
Experiments . . . . . . . . . . 88
Uptake of water of a fixed concentration . 88
Method - Results and interpretation
Uptake of water in the drinking apparatus 93
Method - Results and interpretation
Regulation by mixing 0% and 100% sea water . . . . . . . . . 96
Introduction and method - Differences between small and large
animals - Mixing of other concentrations - Time as a factor
influencing regulation - Influence of air humidity on regulation
Field observations. 105
Final conclusions 107
V. OSMOREGULATION
Introduction . . 109
Experiments and observations. . . . . . . . . . . . . . . . 110
The effect of dehydration on concentration of the body fluid 113
Methods - Results - The effect of reviving - Discussion
Osmoregulation, when water in any form is present . . . . 116
Introduction - The effect of osmotic stress on mortality - The
effect of osmotic stress on body fluid concentration - Endurance
test 1 - Endurance test 2 - The non-electrolyte component of
the blood - Osmotic concentration in the blood - Isotonicity of
the urine
Final conclusions 130
Summary - Resumen. 132
REFERENCE s. . . 136
 GENERAL INTRODUCTION

Among Crustacea terrestrial species are unusual, and only small

numbers of species from various taxa have been more or less success-
ful in occupying the land. In doing so they deviated from the normal
evolutionary path from the primitive marine environment through
brackish and fresh water to marshy and land areas; in this case
some groups went straight from sea to land.
For several reasons connected with locomotion, gas exchange,
water regulation, ion regulation and osmoregulation and the ex-
treme variability of the external environment, a terrestrial way of
life presents many problems. When a direct step is made from sea to
land, particularly under hot and arid climatological conditions,
such problems occur simultaneously.
All kinds of transition al stages may be distinguished. Among the
Decapoda in the Netherlands Antilles only two brachyuran species
and a hermit crab are genuine land crabs. For many days these
animals can stay in dry habitats without any visible uptake of
water. They are adapted to live almost completely on fresh water
and are able to roam far inland. For the development of the larvae,
however, they are still dependent on the sea.
Earlier papers on the physiology and ecology of terrestrial deca-
pod Crustacea are scarce. Mainly by the work of GROSS, during the
last decade, the subject finally received the attention it deserved on
account of its promising possibilities. A symposium on 'Terrestrial
adaptations in Crustacea' (1968) held in 1967 did not include
research on hermit crabs.
 4

From an evolutionary point of view, land hermit crabs are

considered to be a relatively young group. Systematic and morpho-
logical characters provide evidence in favour of this view. As far as
known no fossils have ever been found which could be identified
with certainty as belonging to this group.
During the author's stay in Cura<;ao from 1965 to 1968 (DE WILDE,
1969) special attention was paid to problems concerning water
economy and osmotic regulation in Coenobita clypeatus (Herbst).
The Caribbean Marine Biological Institute (Carmabi) at Willemstad
offered adequate laboratory facilities. Field experiments were main-
ly carried out on the easternmost part of the island (Oostpunt). The
lack of details on both ecology and major aspects of the life history
of the species, such as reproduction, justified additional research.
Though at first the field research was only a side-line, Chapter I
is mainly devoted to its results. In this chapter full stress is laid
upon the way those crabs manage to keep alive in their inhospitable
habitats in Cura<;ao. Chapter II gives further information on the
ecology of Coenobita. It appeared from visits to Bimini (Bahamas),
the Miami area, Florida Keys, J amaica, Aruba, Bonaire, Trinidad
and Barbados, that Coenobita clypeatus may be found in widely
diverging habitats. All the same a habitat as studied on Cura<;ao is
considered to be fairly typical for the species.
In Chapters III to V - on Water management, Shell water and
Osmoregulation respectively - adjustment problems are discussed
more closely, mainly based on results obtained by relevant labo-
ratory experiments.

REMARKS ON THE SPECIES

The small family Coenobitidae, which is mainly restricted to the
tropics, comprises two genera only: Birgus with a single species
(B. latro L., the coconut- or robber-crab), and Coenobita (the land
hermit crabs) with some seven species.
This paper mainly deals with Coenobita clypeatus (Herbst), the
only species in the western Atlantic area. (For a description of the
species see CHASE & HOBBS, 1969). Though the material is considered
to be uniform all over the area, a great many specimens were seen
 5

on Bimini which could be distinguished from the Curayao animals

especially by the conspicuous yellow colour of ambulatories and
major cheliped, and by the golden yellow hairs around the male
gonopore. Among the many thousands of animals examined in
Curayao there were some brick-red colour varieties and occasionally
an albino.
According to PROVENZANO (1959) the species' distribution ranges
from Bermuda, southern Florida and the Bahamas, through the
West Indies to southern Brazil (CHASE & HOBBS, 1969, think that
the Brazilian records need verification). It occurs in the coastal
regions on to several miles inland, including considerable altitudes,
e.g. up to the summit of the 887 m high Mountain of Saba. The
largest numbers, however, are found in the more arid parts of its
distributional area. As far as known the species seems to avoid
places with a high rain fall and a luxuriant vegetation; it is rare or
absent along the 'green coast' of the South-American continent.
In the Leeward Islands of the Netherlands Antilles Coenobita
clypeatus - locally called "soldaatje" (liWe soldier) - is a very
common and conspicuous an im al. It is also common in the Wind-
ward Islands of the Netherlands Antilles (St. Martin, Saba and St.
Eustatius), situated about 800 km to the North.

Part of the material on whieh this study has been based is preserved in the
Zoologie al Museum of the State University, Utreeht.

ACKNOWLEDGEMENTS
The author wishes to express his gratitude to Prof. Dr. A. PUNT for his guidanee,
aid and adviee in eonnection with aIl phases of the work. Sineere thanks are also
due to Prof. Dr. J. H. STOCK for his valuable eomments on the manuseript, to
Dr. W. VERVOORT for his eritieal reading of the manuseript, to Dr. L. B. HOLTHUIS for
solving some problems of nomenclature and to Dr. I. KRISTENSEN for direeting my
attention to the group of terrestrial deeapods and for introdueing me to the Carmabi
on Curas;ao.
To Dr. F. CREUTZBERG and the staff of the Institute I offer my thanks for the
hospitality I reeeived du ring my stay on the island. Prof. dr. A. J. PROVENZANO Jr.
of the University of Miami I thank for his interest and valuable diseussion, and Mr.
R. F. MATHEwsoN, resident director of the Lerner Marine Laboratory, for his kind
aid and hospitality.
I owe a debt of gratitude to Mr. RUBEN ANGEL, who assisted me during the
 6

nocturnal wanderings on Oostpunt, and to Mr. J. P. MAAL, the owner of the estate,
who gave me permission to enter his properties.
Appreciations are due to the management and my colleagues of the Netherlands
Institute for Sea Research for their patience in offering me the required time and
facilities.
Miss S. M. VAN DER BA AN translated the manuscript into English, Mr. C. MARTIS
turned the summary into Spanish, and Mr. H. HOBBELINCK made the drawings and
the laboratory photographs.
In particular I am indebted to Dr. P. WAGENAAR HUMMELINCK, and to Drs.
LOUISE J. VAN DER STEEN, both from the Zoological Laboraty of the Vniversity of
Vtrecht, for smoothing the long and difficult path of preparing the manuscript for
the press, and for reading the proofs.
This study has been made possible by financial support from the Netherlands
Foundation for the Advancement of Tropical Research (WOTRO). I acknowledge
the valuable contacts with Dr. J. H. WESTERMANN. A stay in Bimini and some
short visits to Florida and Jamaica were supported by a grant of the V.S. Naval
Research (O.N.R. nr. 552-07).
 FIELDWORK

I. MIGRATION AND REPRODUCTION

COASTAL AND INLAND ANIMALS

The populations of Coenobita clypeatus observed in Curayao may

be roughly subdivided into two different ecological types: coastal
and inland.
Typical coastal animals occupy the habitats along the sea co ast
and, in places, along the banks of the so-called inner bays and
lagoons. Food is rat her scarce on the shorelines, and numerous
other crustaceans such as the quicker and more active semi-
terrestrial grapsoid crabs, are formidable food competitors. Popu-
lations of coastal animals are characterized by the large percentages
of very small and sm all specimens, while the larger ones are often
badly housed, occupying shells which are too small for their inhabi-
tants or shells with large holes, cracks or worn interiors. Shells of
quite unsuitable types, coming from various species such as Murex,
Melongena, Thais, Latirus, Bulla, etc. are also common. Moreover
lots of shells are often attacked by boring sponges or are covered by
thick crusts of calcareous algae. Coastal animals always look more
or less shabby.
Coastal an im als depend mainly on sea water for their water
supply; water stored in the shells was found in only a negligible
number of cases. As will be explained more fully hereafter, coastal
animals are more dependent on the sea.
Typical inland an im als, on the other hand, give an impression of
prosperity. Small animals are relatively fewer in numbers. The
majority of the occupied shells came from the species Livona pica
 8

(L.) (PI. Ia, IVa). Shells of Livona (= Cittarium) are certainly the
preferred on es in Cura<;ao, as was proved in a choice-experiment
on a small scale.
The inland habitats are characterized by the occurrence of fresh-
water drin king spots, shady trees and much better food conditions.
Inland animals fully depend on fresh water for their needs. Shell
water was usually found to be present to a certain amount.
Both types, with all sorts of transitional stages occurring in be-
tween, become distincdy separated in the dry season after long
periods of drought. In fact the existence of the two types of popu-
lations is largely based on the possession of well-fitting and reliable
shells. Very small specimens of Coenobita and also the larger ones
with ill-fitting shells are supposed to be unable, or at least to have a
poor chance, of ever reaching the favourable inland habitats. Lack
of shell water in the dry season makes travelling increasingly diffi-
cult. Breakage of shells will be fatal in most cases, as the chance of
finding a new one in the interior is extremely small, and let hai
desiccation of the unprotected animals will be only a matter of
hours. In the inland region animals are worse off after moulting than
those on the co ast which have a variety of shells at their disposaI.
It will be evident that only properly equipped animals are able to
invade the favourable free inland habitats and to occupy them
successfully. In particular a good shell will provide the possibility
for land hermit crabs to live in such an advanced stage of terrestrial
life. Inland animals originate from coastal animal populations but
on the other hand an inland animal mayaiso go back to the stage of
a coastal animal.

For the Pacific Coenobita clupeatus auct. (= C. brevimanus Dana) HARMS (1932)
also distinguished between two types of animals, which he calls "Formenkreisen."
The terrestrial form inhabiting the dry coral foreshores, e.g. in the Moluccas up to
an altitude of 80-100 m, may be distinguished from the coastal form - living in
the zone next to the line of low water - by a number of modifications, due to the
transition from an aquatic to a terrestrial way of life. According to HARMS even
morphological differences may be observed, for instance in the covering membranes
of the statocysts. However, the observations described in the present paper, con-
cerning the related species C. clypeatus (Herbst) raise some doubt as to the system-
atic value of the differences described by HARMS.
 9

SIZE CLASSES

Old specimens of Coenobita clypeatus, living in shells of Livona or

Astraea species may be as big as a man's fist, reaching gross weights
up to 500 g, whereas small, but already sexually mature animals in
Tudora, Tectarius or other small and light shells only re ach gross
weights of about two grammes.
Both, for field observations and laboratory experiments, it proved
useful to subdivide the animals into classes, on account of the widely
divergent weights and sizes. In practice the sub division into 5
different size classes - stricHy gross weight groups - was most
satisfactory.

SIZE CLASS ABBREVIATION GROSS WEIGHT OF

Livona - ANIMALS

very small v.s. <lOg

small s. 10-20 g
medium m. 20-50 g
large I. 50-100 g
very large v.l. >100g

Though this division is based on weights of land hermit crabs in

shells of the gastropod Livona pica (L.) only, this does not compli-
cate matters, since the majority of the bigger animals occupies
shells of this species and nearly all experiments concern hermit
crabs in Livona-shells. Anyway, individuals in shells of Astraea
tecta (Lightfoot) and Astraea caelata (Gmelin), respectively the
second and third choice of larger Coenobita in Curac;ao, also fit
satisfactorily into the above mentioned system.
With some experience it is well possible, as was proved by re-
peated experiments, to distinguish at sight between Livona and
Astraea-animals with relatively few mistakes.

MIGRATION AND REPRODUCTION

Since the migration of land hermit crabs is known to occur all

over the Caribbean, it is very remarkable that the phenomenon has
received so little attention from scientists. Scattered over the whole
 10

area there are natives, in partieular among elderly fishermen, who

ean tell stories about the clattering noise of huge numbers of soldier
erabs, proeessions of several hundreds of meters long, at bright
moonlight or, as others will have it, at piteh-dark nights. In the
Netherlands Antilles a traditional story is told about invaders who
took flight after they heard - literally by putting their ears to the
ground - that many soldiers (soldier erabs) were on the march.
Beeause of the soft abdomens of the animals, whieh seem to be
exeellent bait for fishing, most Anhllean fishermen show mueh
interest in these ereatures and know when they are absent, or at
least difficult to find, during several months of the year in those
plaees where they are usually found. From our own experienee we
know that the experimental animals were hard to obtain from about
J uly till Oetober.
There is no doubt that the terrestrial hermit erabs of the Family
Coenobitidae, which have all beeome adapted to life on land and
often spend the major part of their life far - up to several miles -
from the eoast, reproduee by means of larvae liberated into the sea.
Based on animals reared in the laboratory, following two hatehings
in 1960 on 22 J uly and 24 August, and two more hatehings in 1961,
both on 10 August, PROVENZANO (1962) gives a description of the
complete larval development of Coenobita clypeatus. From his ex-
periments it is clear that the development of the larvae takes plaee
in normal seawater.
VERRILL (1908) reports the occurrence of Coenobita clypeatus on sand hills in
the Bermudas, far from the coast. PEARSE (1915) already mentioned a procession
of about 500 individuals, moving over an area of about 200 m square, apparently
migrating from the ocean through the desert of Santa Marta (Colombia), on August
29, 1913.
In a - rather imaginative - popular publication CARPENTER & LOGAN (1945) give
an eyewitness account of two occasions on which they observed a "jamboree" of
Coenobitas at Mona Island, halfway Puerto Rico and Hispaniola. They described
how towards the end of August ten thousands of animals struggled along to one
sandy beach.
None of the collected female specimens on Dominica in January, February,
March and June by CHASE & HOBBS (1969) were ovigerous.
All the aforesaid communications show that the events dealing with the repro-
duction of Coenobita clypeatus are to be expected in aperiod ranging from about
the middle of J uly till the end of August.
 1967 I 1966
DU MOON mARKS UGOONS w ,LL 1 DIR ,CTIEST RAN(
IfRESH HAlf fULL IIftESH I HiLf fUll I fRESR HI I rULL IfR ESH I Hilf I JULL MO ON

I-
•

f--
o
20 11~9

f-
•
30 -
I- 0
•
40

f-
o
50 ,1..;10
•
f-

60
- •
I- 0

70

I-
o
80 1-11 •
f-

•
90

I- 0

~100

f--
o
10 1- 12
1
•

..
t-

'

t-
o only one , :
:.:
'

animal _ll ,

130121-1:1

COENOBITA CLYPEATUS (HERBST)

Fig.5. SUCCESSIVE WAVES OF EGG PRODUCTION AND DEVELOPING EGGS at Oostpunt

(1967) and Directiestrand (1966/7). - Numbers: percentages oi ovigerous iemales
with freshly spawned, half-developed and fully developed eggs; x x x: records
oi which no percentages are known, or which are based on very small numbers;
dotted parts oi bars: extrapolations.
 28

T ABLE 5
SEX RATlOS IN Coenobita clypeatus AS FOUND NEAR THE
LAGOONS AT OOSTPUNT, 1967
Upper taUe: north of the Lagoons, at landside.
Lower table: south of the Landbridge, near the sea, at the time of egg-depositing, showing
100 per cent females.
Ratio = number of males divided by number of females, per size class.
Size classes: v.s. = very smalI, s. = smalI, m. = medium, 1 = large, v.!. = very large
Number of sexed individuals in italics.
a
8.VIII 14.VIII 18.VIII 22.VIII 29.VIII 8.IX lUX 15.IX
v.s. 0.75 7 0.0650 0.2550 0.2353 0.1950 0.25 25 0.0425 0.04 75
s. 0.25 5 0. 1425 0.5625 0.39 25 0.75 I4 0.67 25
m. 1.0 I4 0.92 25 4.0 25 1.4 24 0.42 I7
1. 3.0 4 7.3 25 6.5 15 5.5 13
v.l. 11.0 I2

b
lUX 15.IX
v.s. 0.0 5°
s. 0.0 20 0.0 20
m. 0.0 2I 0.0 I4
1. 0.0 2 0.0 2
v.l.

a cont'd

18.IX 27.IX 6.X 13.X 20.X 27.X 8.XII 29.XII

v.s. 0. 142 4 0.2728 0.23 I6 0.19 I9 0.75 7
s. 1.2220 0.47 2 5 0.38 I I 0.0 6 0.13 9 0.0 7 0.3726 0.0 6
m. 2.0 3° 0.7528 0.25 2 5 0.5 I8 2.0 I2 0.78 I6 0.8 9 6.0 7
1. C\:) 4 7.0 8 0.0 2 C\:) I 3.0 8
v.!. C\:) I

b cont'd

6.X 13.X 20.X 27.X

v.s. 0.0 53 0.0 48
s. 0.0 3 2 0.0 I7 0.0 I 0.0 3
m. 0.0 I6 0.0 22 0.0 I
1. 0.0 I 0.0 I
v.!.
 29

Field observations in the nights of August 22 and 29 and of September 8, and the
results of subsequent sampling hardly produced any new data. There was still a
clear increase of the number of animals present in the vicinity of the Lagoons in
general, and in the surroundings of the Landbridge and the Delta area, in particular.
An aggregation in these parts was also demonstrated by the movements of the
orange (new) series, which were released on the eastside of Lagoon A. A fair number
of these animals were recovered later on near Lagoon C. Estimations showed that
the total number of animals near the Lagoons was now some ten thousand, perhaps
even over one hundred thousand.
Ey now the colour of the eggs carried by a fair number of the fe males had changed
gradually from dark reddish brown, through to light reddish brown, light grey and
greyish blue. Dark eyespots indicated very distinctly the advanced development of
the eggs.
Again all possible suitable parts of the coastline, both on the north and south
coasts of Oostpunt were searched for other concentrations of hermit crabs. Apart
from the concentrations already known to exist around Well 1, not exceeding 200
animals, nothing could be detected.
Just as on October 2, 1966, when masses of hermits on the Directiestrand had
been noticed, weather-conditions on September 3, 1967 were equally dull, with a
drizzling rain. Visiting the Directiestrand in the morning again thousands of
very small animals, mostly living in Tectarius shells, were seen on the coral debris,
limestone boulders and under the manchineel trees. SampIes produced lots of
ovigerous females with developing eggs, but nothing special could be detected.
On September 11 under quiet weather conditions (partly clouded, moon in its
first quarter, low tide) thousands of hermit crabs, not previously observed in this
spot, were seen struggling down over the Landbridge in the direction of the sea
coast. Clusters of animals in which all sizes were represented were distributed
between the tide marks over a distance of about 100 m. Here at low tide with little
wave action a number of quiet, seawater-filled rockpools occur. Hundreds ofhermits
frequented the edges of the rockpools, under and against stones covered with algae
and also on the still moist coral debris, but never, neither completely nor partly,
submerged in the water. Clusters of fully developed eggs were attached to the
stones above water level. Sampies of water from the rock-pools and off the coast
were found to contain young larvae.
No males were present among egg-depositing females in the tockpools, neither
were there any crossing the Landbridge. SampIes of animals from the Landbridge
and the adjacent part of the coast showed sex ratios of 0.0 in all size classes as is
shown in Table 5.
This phenomenon, namely that the majority of the ovigerous females with fully
developed eggs separate themselves from the main group about the time of egg
deposition, thus leading to distinctly lower sex ratios than when the animals are
collected closer to the sea, was used later on to find evidence for other possible
hatching places. Sampies were taken on the eastside of Lagoon A and near Well 1,
where hatching pI aces were expected to exist. Sex ratios from these places, however,
did not give any information to this effect.
In 1967 the phenomenon occurred again on the Directiestrand, though not as
clearly as near the Landbridge.
 30

Based on the observations on the Directiestrand in 1966 and in

1967 and also at Oostpunt ne ar the Lagoons, near WeIl 1 and again
on the Directiestrand the appearance of another group of females
with freshly spawned eggs was expected and did indeed occur to-
wards the end of the first wave of egg production and development,
so that there must be a second wave (vide Fig. 5). Further obser-
vations showed the occurrence of a third, and at one time even of a
fourth wave. The strongly decreased numbers of ovigerous females
of which this third (and fourth) wave(s) were composed, made it
impossible to follow the complete waves in detail.
Further visits to Oostpunt on September 15 and 18 showed a
gradual decrease of the percentages of females with fully developed
eggs and subsequently an increase of the percentages of females
carrying freshly spawned eggs. On September 27 only fresh or partly
developed eggs were found.
In the meantime, and following the end of the first wave, the
total number of animals near the Lagoons decreased markedly
(Fig. 4).
The second time that great activity on the Landbridge was no-
ticed, was on October 6.
The same low sex ratios, showing alm ost 100 per cent females,
were found in sampIes from the sea coast. Later on females with
freshly spawned eggs were observed and a furt her decrease of the
total of hermits near the Lagoons. During the second half of October
only a few animals were left.
Observations of diffuse migrations in several places in the grounds
of Oostpunt indicated the re-population of the inland. Though none
of the recovered tags in the surroundings of WeIl 4 had been
recorded earlier on the southcoast, the growing population near
WeIl 4 was obvious enough (Fig. 4).

In addition, a fair number of well-sized animals was found in a small machineel

grove, N.E. of Awa Blancu and about 1500 m west of the Landbridge. No animals
had been observed here in the preceding months. Two recoveries of orange (new)
tags on October 27 and December 29 in this place proved that the animals came
from the Lagoon area. So did a yellow (new) tag, recovered on October 20 in a
place midway between WeHs 2 and 3 on the north coast.
During the last visits to the Lagoons within the scope of these experiments in
December 1967 only a limited number of individuals could be counted. At that
 31

time, however, the population of WeIl 4 was obviously not yet at its fuIl strength.
Visiting Oostpunt on June 25 1968 some feIlow-workers of the Institute observed
two tagged animals, a black and a red one, near WeIl 4. They did not observe
whether they were originally tagged animals or whether moulting or changing of
shells had occurred in the meantime. A similar observation near WeIl 4 followed in
October 1968 at which occasion three tagged animals were seen; apart from a violet
and a black tag also an orange (new) one from the lagoon was found.
Finally it must be recorded that in J une 1970 an assistant of the Carmabi collected
some 100 animals near WeIl 4. Three years after the original tagging experiments
a quarter of the animals present was found to be still in possession of tagged shells.

DISCUSSION

From the tagging experiments described above and the chrono-

logical field observations carried out in the second half of 1967 and
the supplementary data from 1966 a satisfactory picture could be
assembled concerning the seasonal migrations and other activities
connected with reproduction of the land hermit crab Coenobita
clypeatus Herbst in Curac;ao. Though important details as copu-
lation, sexual behaviour, etc., were not observed, the story as a whole
has become sufficiently c1ear.
There are only few publications on migration and sexual
behaviour of Coenobita clypeatus and c10sely related species and
therefore the data described here cannot be compared with the
findings of the other authors. For the present it must be kept in
mind that some of the data recorded here may perhaps apply to the
islands of the Leeward Group only and other comparable islands
and coastal areas with a dry climate and same hinterland within the
dispersal area of Coenobita.
Typical "coastal animals", at least as far as their reproduction is
concerned, do not show any migrations as described here. Flourish-
ing populations such as occur on the very small Keys off Kingston
Harbour, Jamaica, e.g. Drunkenmans Cay which were visited in
November 1967, have no possibillities for migration whatever.
Coenobita living under quite different ecological circumstances,
such as occur in Bimini (Bahamas) or Saba (Windward Group) may
show patterns different from those discussed in this paper.
 32

Reproductive season and mating places

Reproduction of Coenobita clypeatus in the Caribbean takes place

in summer and autumn, dispersed over several months. In 1967
first indications of migrations in Curayao were noticed early in
J uly; the last ovigerous fern ale was found on December 19.
Judging from the freshly spawned eggs in 1966 and 1967, first
noticed on October 2 and August 18 respectively, considerable
changes in time may occur.
Since PROVENZANO (1962) mentioned hatchings on July 22 and
August 10 and 24, there must have been freshly spawned eggs at
least since the end of June in the Florida region. It is possible that in
the northernmost part of the dispersal area of this species, reproduc-
tion may even start some weeks earlier. The absence of any sign of
eggs or larvae, during my visit to Bimini, early in November 1967,
also points in that direction.
As Coenobita often lives in habitats far from the sea coast and its
larvae need the open ocean for their development, the aim of the
summer migrations is dear. They appear to migrate to preferential
spots along the coast, which are not necessarily dose to the shore.
The striking feature is the directedness of the movement.
PEARSE (1915) mentions amigration of Coenobita consisting of
about 500 individuals, moving over an area of about 200 square
meters, away from the ocean in North Colombia. Eleven animals
were collected;and preserved; all animals proved to be fern ales and 8 of
them were carrying eggs. CARPENTER & LOGAN (1945) described mass
movements of land hermit crabs to one sandy beach, measuring
three acres, in Mona Island. In Saba (personal communications of
Mr. A. J. GEERLINK, 1965) fishermen annually observe large num-
bers of hermit crabs descending certain gullies from the mountain
slope. The present paper describes the migration of the large ma-
jority of the hermit populations living in the eastern part of Cura-
yao towards a small area on the south coast of the same island.
No doubt difficulties such as those caused by the steep diffs, in
combination with the extreme exposure to wave action along the
north coast of Cura<;ao restrict suitable places for egg deposition by
 33

the females and further development of the young larvae; but it

cannot be explained how the adult animals are aware of these com-
plications. On the other hand sheltered inlets as occur e.g. in the St.
J oris Baai on the north co ast and in most parts of the south coast,
with moderate wave action, including open in lets as Awa di Oost-
punt and Fuik Baai seem to offer more suitable spots than the
100 m wide strip near the Landbridge.
Some preliminary experiments already showed a higher mortality of newly
hatched larvae in hypersaline water: for example a combination of 125% sea water
and a temperature of 28°C showed an almost 100% mortality within 24 hours.
PROVENZANO (1962) mentions successful rearings of larvae under laboratory
conditions in filtered sea water of a salinity of about 36.0%0 and a temperature of
29°C. Therefore, water of higher salinity in inlets and lagoons seems to be less
suitable for the development of the larvae. There is no evidence that the larvae
of Coenobita are equaHy weH adapted to estuarine environments with steep salinity
gradients as those of the brachyuran land crab Cardisoma guanhumi Latreille
(KALBER & COSTLOW, 1968).

Since Coenobita was observed to regulate the salinity of its shell

water by diluting it with fresh water if necessary, the presence of
drinking places along the northern banks of the Lagoons seems to be
the main factor why Coenobita prefers these particular localities.
From the moment of arrival up to the time of liberation of the larvae
by the females - aperiod of at least five weeks - which makes high
demands upon the water management of the ovigerous fern ales in
particular, the animals have to remain in an arid and barren area.
As far as known, the surroundings of the Lagoons are the only
places along on Oostpunt to ofter both quiet, c1ear sea water of the
right salinity, and relatively fresh water, within easy reach for
Coenobitas.
The large number of animals round about the drinking places
show the need for fresh water, which unquestionably is a seepage of
rain water originating from the limestone terraces above Ceru
Port'i Camber and Ceru Palibandera or even perhaps from the
Ceru Blancu table land. By mixing with the salt lagoon water, in the
outcrop on the Lagoon banks just above the water level, it obtains
its brackish character (see Table 4). Low salinities such as l.6%0
and l.8%0 from the sampies taken on October 27 show that the
available water is derived from fresh water. High water levels in
 34

the Lagoons as observed on November 6 cause temporary inunda-

tions of some of the drin king places, which make them unsuitable
for Coenobita.
Except for WeIl 1 no other spawning places where ovigerous fe-
males could be observed were found on the north coast of Oostpunt,
though it has been searched thoroughly. Just as in the Wells 2 and 3
on the north coast suitable drinking water is offered here at a short
range from the ocean, however, the ocean is not by any means a
quiet water.
The development of the eggs near WeIl 1 went on smoothly.
Deposition of eggs on the nearby coast during the favourable nights
of September 15 and 18 could not be observed while at the same
time it was in progress on the South coast.
As stated before the population on the Directiestrand is inter-
mediate between real "CoastaI" and "Inland" -animals. The situa-
tion on the Directiestrand is hardly comparable with that at Oost-
pu nt. Although fresh water is not permanently available, the water
conditions seem to be sufficient for reproducing Coenobita. Shady
trees and thick layers of mould and old leaves provide shelter.
Manchineel fruits and waste water apparently provide a sufficient
supply of moisture.

Migrations

From field observations and tagging experiments it must be

concluded that the migrations down to the Lagoons from WeIl 4 are
straight and directional. Not a single tag was recovered near the
seven other check points, all situated around WeIl 4; not a sign of
movements elsewhere than near Cer'i Palibandera and east of Awa
Blancu was noticed.
At present a random swarming, followed by a concentration by
trial and error in a suitable spawning place seems very unlikely.
Starting from a larval development of at least some weeks, of
which the greater part is planctonic, and also considering the strong
currents both in the hatching place and in this part of the Carib-
bean, we may suppose that part of the larvae will be passively
 35

transported over considerable distances. Probably part of the

animals in Cura<;ao will have come from the Bonairean coast or
from more eastward islands (Aves Islands).
PROVENZANO (pers. comm.) believes, however, that at least part
of the litt oral larvae are retained in their hatching area and that
the question of transport is not as simple as is generally supposed.
For instance EMERY (1972) proved that on Barbados a re-stock-
ment of the decapod population of shallow water takes place from
locally produced larvae, in spite of strong and steady currents
impinging upon that island from the East. Their taking off is pre-
vented by aseries of eddies and counter-currents, which previously
were left out of consideration.
The colonization, more or less by chance, of the inland habitats by
the larger and properly equipped animals, favoured by proper
weather conditions, is quite understandable. On the other hand a
direct migration to a fixed breeding place by a crustacean, hatched
far away, transported passively by sea currents, moved upon the
shore in any possible place and finding an appropriate habitat by
chance only, requires a lot of our imaginative faculties on the point
of homing and directional orientation of land hermit crabs. Obser-
vations made on Oostpunt point to a guidance of young and un-
experienced animals by experienced animals.
Whatever the mechanism of orientation may be, by some internal
or external signal, migration towards the spawning places starts in
the early summer in Cura<;ao and adjacent areas.
On July 5, 1967, when the first series of tagged animals were released ne ar WeIl 4,
migrations had probably already started. From the number of recovered black
animals and the number of caught untagged animals on the next day, a very rough
calculation may be made, giving some idea ab out the strength of the population
present around WeIl 4. Assuming that the population did not change very much
in the course of one day, and also that tagged and untagged animals occurred at
random, the number was estimated at 7-800 individuals (very smaIl animals not
counted). A similar calculation from the recovered black animals on July 7 gave
ab out twice as many of them present near WeIl 4. Another calculation based on
recovered red animals on July 14, two days after the release, gave a number of
over 400.
Recoveries of violet and yellow tagged animals near WeIl 4, released 500 m
westward and eastward of the Weil, proved that medium-sized animals moved
over at least 500 m in a single night. Previous measurements, based on walking
velocity over relatively short distances of about 50 m, gave theoretical values of
 36

about 1000 m per night. Assuming that the lower value is the correct one, hermits
from Well 4 can cover the distance down to the Landbridge, the shortest distance
being about 4000 m, in eight days, but probably in a shorter time. In theory they
can do that without any water supply on the way. As a matter of fact such a journey
will take several weeks at leisure, time spent in looking for food and water included.
A yellow animal released on July 5 was recovered on July 21 near Palibandera,
thus covering 3 km in 16 days. The first observed tagged animal near the Lagoons,
between Landbridge and Delta, also a yellow one, made a jouney of 40 days and the
first of the red animals, released on July 12, which was observed near the Lagoons
had taken 36 days to cover a distanee of 4 km.
It is supposed that favourable weather and moist terrain conditions will highly
stimulate migration.
As to the preferred route between Well 4 and the Lagoon-area we cannot say
anything as yet. It seems plausible that the animals follow the esearpment of the
Ceru Blaneu in the direetion of Cer'i Palibandera, and from there through the gully
southwards; others perhaps will take Rooi Sjon Tata and go from Ceru Patia down
to the eoast. Hardly any aetivity was notieed on the traek between 'Well4 and Awa
Blaneu and the migrating animals did not show any interest in the fresh water
basin of Poos Grandi along the track.

Only animals in good condition and fitted with a reliable sheH

migrate towards the spawning places. The great majority of the
non-migrating hermits, observed in the inland habitats e.g. those of
WeH 4, proved to be of the smaHer size classes or were obviously
badly housed. Quite a few of the examined females proved to be
deficient, which means that they had underdeveloped pleiopods.
Typical small-sized and shabby coastal animals as usually ob-
served round We1l3 and in several other places, scattered along the
coast line, not showing migration, were also found to have high
percentages of deficient fern ales as is shown in the two examples
foHowing:
1) A sampie of 42 sm all- and medium-sized animals from the co ast
east of Fuikbaai, taken on August 29, proved to consist of 52.4%d'd',
14.3% normal ~~, and 33.3% deficient ~~.
2) Another sampie of 60 small- and medium-sized animals from
WeH 3, which was taken on September 8, consisted of 56.7%d'd',
6.6% normal n, and 36.7% deficient n.
Tagging experiments with black, violet, yeHow (old) and red
series, as mentioned before, gave certainty about the migrations
from WeH 4 towards the Lagoons. Considering the vastness and
inaccessibility of the area, the number of recoveries was not dis-
appointing. In Table 6 the number of animals recovered in 1967
 37

T ABLE 6
RECOVERED TAGS OF COENOBITAS RELEASED NEAR WELL 4,
OOSTPUNT, 1967
In addition to the recovered totals, the data for recoveries near the Lagoons are
given separately.

series number recovered total recovered

released near Lagoons
number percentage
number percentage

black 168 46 27.4 8 4.8

yellow (old) 118 13 11.0 8 6.8
violet 130 12 9.2 4 3.1
red 184 35 19.0 7 3.8

total 600 I06 I7·7 27 4-5

divided into the total numbers and those from the Lagoons only,
are summarized. Tags observed twice or more often are not included.
I t cannot be concluded from these figures that any of the colours
should be more easily noticed in the field than the other colours;
apparently the observation of tagged animals is mainly based on the
recognition of a contrasting disk on the shell of the individuals.
Comparing the percentages of the recovered tags of all the series
near the Lagoons it appears that they are within the same order of
magnitude, and that they do not show any disadvantage for the
violet and yellow (old) animals, which were released at a distance of
500 m from Well 4.
When studying the numbers of recovered animals, obvious
differences appear between the various size classes.
Though perhaps for the relatively small numbers the conclusions
are somewhat speculative, the figures of Table 7 seem to indicate
reduced ability of the small animals to perform the trip to the
spawning places on the coast. On the other hand, this would be also
in entire conformity with the assumption of special inland popu-
lations, being characterized by the large percentages of animals in
the larger size classes.
 38

T ABLE 7
RECOVERED COENOBITAS FROM THE LAGOONS DlVIDED INTO
SIZE CLASSES

Black, yellow, violet and red tags. Number of released

individuals in parentheses.

dass number percentage

small 1 (105) 1.0

medium 11 (283) 3.9
large & very large 15 (212) 7.1
total 27 (600) 4·5

Mating

Animals in the Lagoons settle on the northern banks in the dose

vicinity of the drinking places, and in the course of time concentrate
in the area between the Landbridge and Delta. Huge numbers of
weIl-housed animals, possibly something like a hundred thousand,
at the height of the season, lead us to assurne the presence of
numerous inland populations existing in the grounds of Oostpunt;
the population of WeH 4 being just one of them.
From this crowd of animals 27 different specimens out of a
number of 600 tagged animals from Wen 4, could be recovered. The
number of tags observed more than once was low. Only once the
same tag was recovered in a different place on a different day. Three
other repeated observations were always made in exactly the same
spot, under the same stone or in the same hole.
From the data obtained it is possible to make a rough calculation
which percentage of the population of WeIl 4 succeeded in ac-
complishing the tour to the coast. The small numbers of green and
orange (old) tags found near the Lagoons indicate that in any case
the chance to do this from WeIl 2 or 3 is much smaller.
At daytime the Lagoon banks are mostly empty, the animals
hiding in the labyrinth of crevices and holes in the limestone
deposits, or between the twisted roots of the numerous Conocarpus
 39

trees. At night they are active, refilling the shell-water and as-
sembling in certain, more or less flat places on the bank, often
together with animals of about the same size. Though copulation
was not observed, it seems very likely that this will happen on
these special "playing grounds." It is suggested that mating takes
place between animals of roughly the same size group only. It
seems impossible for sexually mature, very small animals, for the
greater part females, living in Tectarius- or Tudora-shells, with a
gross weight of 1 g to mate with giant animals up to a gross weight
of 500 g. The rough division on the "playing grounds" into size
groups which have no sharp limits and partially overlap, depends on
the ability to succeed in mating. In particular groups of very small
animals in Tectarius- and Tudora-shells, and in shells of the smaller
N erita-species are obvious. In other places groups of large and very
large specimens are numerous, whereas congregations of medium-
sized animals are most common. The demonstrated differences in
sex ratio between smaller and larger animals indicate that there
exists only arestricted supply of females in the larger size groups.
This will certainly result in interesting differences in sexual be-
haviour within the various size groups.
Animals on the "playing grounds" often stay in the same place
for a long time; they are active, have raised bodies, and continuously
move the antennae and antennules. However, no attention was paid
to special patterns of behaviour, associated with mating as describd
by HAZLETT (1966a) for a number of marine hermit crabs of Curac;ao.
In his observations on the social behaviour of Coenobita (1966) no
sexual behaviour is mentioned.
Small numbers of animals from the "playing grounds" observed
in glass aquaria in the laboratory did not show any sign of actual
mating.

The curious descriptions of CARPENTER & LOGAN (1945) may concern mating on
Mona Island: "At sunset they scattered on the beach. Then came the startling event.
The jamboree began and it was as though the grandfather of all hermit crabs had
called through his megaphone: 'change shells with the lady on your right'. The hermit
crabs came out of their portable houses and began to exchange them. It was moving
day, dance marathon and bargain basement rolled into one. A good many were
soon looking the worse for wear; but the brawl went on ... Less than a week after
 40

the start, the sun rose on an empty beach. The hermit crabs had all gone back into
sec\usion on the top of their fortress rock."
It is difficult to fit CARPENTERS observations into the sequence of events as
they were observed on Curayao.

On some occasions shell-fighting as described by HAZLETT (1966) was observed,

but never the exchange of houses on a large scale. The rapid retreat of the hermits
to the inland area of Mona Island is completely different from our observations.

In 1966 the activity on the "playing grounds" in Oostpunt lasted for several
weeks, starting in the middle of August and terminating towards the middle of
October. The peak of activity was observed some days before full moon. According
to the activity measurements in the Laboratory (Chapter III) diurnal activity
on the "playing grounds" and near the drinking places was also greatest in the
early evening, slowing down towards midnight.
When very small and small females from the "playing grounds" are removed
from the shell the swollen, dark reddish-brown gonads can often be seen through
the transparent skin of the abdomen. In older females this is less obvious because
of the thicker tegument.

Sometimes, by pressing the abdomen of a male, long spermatophores may be

produced from the gonopores (PI. Va); the latter are widely opened and the tips
are covered with black or dark brown hairs. These sperma threads are whitish in
colour and sometimes have thickenings like a rope of beads. They are very elastic
and sticky, and in old males reach a length of over 10 cm.

After preliminary precopulatory behaviour the copulation itself, on the analogy

of HAZLETTS observations in marine species, is supposed to be a partial moving out
of the shell- perhaps as observed by CARPENTER. When gonopores of both animals
are opposed, parts of the spermatophores will be transferred to the female, possibly
with the aid of the fifth pereiopodes. Soon afterwards both animals will pull back,
after which postcopulatory behaviour may be shown.

Egg production and egg development

The freshIy spawned eggs are firmIy attached to the thin but
extremely tough hairs of the branched pleopodes at the Ieft side of
the abdomen (PI. Vb).
With a freshIy caught femaIe, placed into an artifical house of
transparent glass, the role of the modified fifth pereiopodes in
attaching the eggs could clearly be seen (PI. Vlb).

In 1967 the first freshly spawned eggs were observed near the Lagoons on August
18; one day before fulI moon. Only very few of the females were then carrying
 41

eggs. A sampie of 50 very small animals, collected at random, held only 1 (= 2%)
ovigerous female. Sampies from the northern bank of Lagoon C four days later
on August 22, already showed 21 % of the very small females carrying eggs; for the
small animals this percentage was 38, for the medium sized animals 46 and for the
larger females about 75%. Sampies taken on August 29 showed percentages in the
same order of magnitude.
It is assumed that in Coenobita clypeatus sexual maturity is not attained before
the second year. Then many of the very small ovigerous females occupy Tectarius
shells. Two sampies of 50 very small animals each, allliving in Tectarius, irom a
"playing ground" near Lagoon C showed gross weights ranging from 0.6 to 3.0 gr,
with over 80% between 1.2 and 2.2 gr. In the meantime many much smaller
specimens of Coenobita, not yet mature, are to be found. These miniatures have
gross weights between 0.1 and 0.7 gr. Possibly they belong to an earlier year class,
probably of the first year.
The number of eggs carried by these young females fluctuates around one thou-
sand. Eggs of ten ovigerous animals were counted; numbers of 796, 859, 869, 871,
923,957,986, 1016, 1044 and 1260 were found. In the large animals these quantities
may be 40 to 50 times larger.

Freshly spawned, still compact egg masses are dark reddish brown
in colour. Gradually the eggs become paler, changing to grey or
light blue, the structure of the egg masses becoming much looser.
From structure and colour of the egg masses a good impression
may be obtained by the naked eye about the state of development
of the eggs. Fully developed eggs, when placed in seawater, burst
open immediately, releasing the young zoea larvae.
A few times females were observed, with sets of eggs of which the
eggs on the frontal part of the abdomen had already fully developed
and the eggs on the distal part were still brown and distinctly un-
developed. The gradual change in colour gives the impression that
this may be caused by physiological factors, such as differences in
temperature or oxygen supply, rather than that one or more new
sets of eggs have been attached later on by the female.

Sets of eggs, attached to the female and enclosed by a proper shell

are weIl protected against damage and desiccation. The relative
humidity in the shells is supposed to be near a hundred percent ;
this, however, was not measured. There is always the moist skin of
the abdomen and usually shell water is also present, but not in such
quantities as observed in normal, non-ovigerous, animals. Ovigerous
females in transparent artificial shells usually showed small amounts
 42

of water present in the apices of the shells, often moistening the

eggs but never permanently bathing them.

In a small number of ovigerous females, caught on August 21 and 29 near Lagoon

C, it was possible to shake out enough shell water for conductivity measurements
in the laboratory. The salinity of this water proved to be quite normal, only a
little below the salinity of 100% sea water (Table 8).

T ABLE 8
SALINITY OF REMAINS OF SHELL WATER

measured in 7 ovigerous females from Lagoon C.

nr. date, 1967 eggs salinity %0

1 21.VIII freshly laid 34.4
2 21.VIII freshl y laid 35.0
3 2l.VIII freshly laid 31.3
4 29.VIII half developed 35.0
5 29.VIII half developed 32.1
6 29.VIII half developed 35.8
7 29.VIII half developed 33.5

It is quite comprehensible that in females, carrying nearly developed or fully

developed eggs, there is no water in the shells, since absence of water prevents
hatching of the eggs in the shell cavity. Only 4 out of a sampie of 50 ovigerous
females, caught southwards of the Landbridge on September 11 and all carrying
mature eggs, showed traces of water in the shell, which may also have been some
water from the branchial cavities.

Ovigerous females in captivity often lose the eggs prematurely, which could be
observed in particular when females were placed into too small shells, deforming
the loose clusters of developing eggs into compact masses.

Under circumstances as met with in Cura<;ao egg development in

the field takes 3 tot 4 weeks. Interrupted periods of egg-production
and also, to a certain extent, differences in physiological and eco-
logical factors will playa role in the observed differences in the
duration of maturing of the eggs. When clusters of three weeks old
eggs are placed in sea water they yield only small percentages of
swimming larvae. If the same is done with eggs which are a few days
older, large quantities of young larvae are released.
 43

Ovigerous females pass over the Landbridge only if nearly all the
eggs are fully developed.
Determinations of sex ratios on the Landbridge as well as in the
adjacent part of the coast gave an extremely low percentage of
males in all size classes. The majority of the fe males carried eggs,
sm aller numbers were normal females without eggs and a small
percentage of the females proved to be deficient. It is not quite clear
why males should be absent and deficient females, along with egg-
less females, should occur in sm aller or larger numbers. Perhaps the
distance between hiding places and beach are too short to permit a
good separation. Moreover part of the animals here may be looking
for sea water for drin king purposes.

Egg deposition

When strong movements towards the sea were observed on the

Landbridge in the night of September 11, thousands of Coenobitas
were already present on the shore, partiallY hiding under loose
stones and in the shoreline, as well as active on the wet coral stone
near the water level. The weather on that day was very quiet; the
mo on nearly in its first quarter and the tide was low. From the
scattered clusters of eggs - rather inconspicuous on the buH under-
ground - it was clear that egg deposition was going on. It was also
clear that none of the animals was in direct contact, not even
partially, with the water of the rock-pools or the open sea. A simple
washing away of the ripe eggs (larvae) from the oostegites did not
occur. For completeness' sake it may be mentioned here that
fishermen on the island of Saba ins ist that the local soldier crabs
migrate into the sea in September to "wash off the eggs" (Mr A. J.
GEERLINK, pers. comm.).
Just as with the incorrect representation of the uptake of shell
water, suggested to be a simple filling of the shell by the waves on
the shore, the water management and salt regulation of Coenobita
clypeatus does not allow such a coarse procedure. As a matter of fact
complete inmersions caused by a sud den high wave or a fall may
occur during the egg deposition, but always by accident.
 44

What actually happens is that the female, with the help of the
fifth pereiopodes, picks up sm all clusters of fully developed eggs from
the masses attached to her abdominal appendages and takes them
further out with the maxillipeds and chelipeds. In this way the eggs
become spread over the wet stones and later on, during high tide, the
larvae are washed into the ocean. In females hiding under stones,
higher up on the shore, we often noticed that only part of the eggs
were removed, and it is supposed that liberation of the larvae
occurs on a number of consecutive nights, enabling the fern ale to
release only the completely developed eggs.
Several days after the fern ales have deposited their egg-clusters,
they are still recognizable by the whirled and coiled setae on the
pleopodes and the yellow coloured remains of lute or fragments of
the egg shells.

In order to learn more about the method of egg deposition considerable numbers
oi ovigerous iemales were brought to the laboratory. A few tim es it was possible to
observe females kept in a wire netting cage. bringing out bits of the egg masses
from the shell cavity. Then the eggs were twisted into a slightly spherical shape
between the maxillipeds and afterwards attached to the tip of one of the two
chelipeds. A short forward movement hurled the eggs away.
Unfortunately normal "egg-shooting" by females in artificial glass houses could
not be observed, but it was sufficiently clear that the eggs are loosened from the
setae of the pleopodes by the small scissors-like pincer of the left and right 5th
pereiopod.
Some experiments described below were carried out to study the effect of the
"egg-shooting." Glass aquaria of about 30 X 45 X 30 cm were divided into two
compartments by a partition of steel gauze with a square mesh of 2 cm. At one
side medium-sized females with fully developed eggs were placed and at the other
side, 1 cm from the gauze a square plastic tray (20 X 20 cm) filled with normal sea
water. Next day lots of swimming larvae were observed in the water, but also
clusters of eggs beside the tray, attached to the gauze and to the walls and bottom
of the animal compartment.
Evidently the eggs are not shot off in any definite direction and the water does
not seem to have any special attraction.
More elucidating are the experiments in which the water of the tray was replaced
by a layer of moist filter-paper, on which the eggs stay attached to the very place
where they hit the paper (Fig. 6).
The maximum distance observed over which they were shot from a maximum
elevation oi about 25 cm, as was made possible in the experiments, amounted to
about 20 cm. The numbers in Figure 7 reier to the numbers of eggs in each cluster
on the paper. Again there is no question of a directional "shooting".
Yet this manner of egg deposition is quite. functional for the animals. safeguarding
them against entering the ocean. Theoretically Coenobita clypeatus would be able
 45

Fig.6. EXPERIMENTAL SET UP FOR DEMONSTRATION OF 'EGG SHOOTING' by Coenobita

clypeatus. - After fernales with fully developed eggs had been put into the left
cornpartrnent, srnall clusters of larvae were observed on rnoist filter paper in the
plastic tray to the right.

T ABLE 9
LARVAL STAGES IN Coenobita clypeatus FROM
LABORATORY REARINGS

after PROVENZANO, 1962. - Larval duration in nature

rnay be shorter. - All stages are planctonic, except for
the glaucothoe which shows partly a non-swirnrning
behaviour.

stage totallength duration

first zoea 2.6-2.9 rnrn 3-5 days

second zoea 3.2 rn rn 3-5 days
third zoea 3.8rnrn 3-4 days
fourth zoea 4.6rnrn 3-4 days
fifth zoea 4.7-5.5 rnrn 4-6 days
sixth zoea 4-6 days
glaucothoe 4.3rnrn over 1 month
 46

I. •91
l 1

. " 111

I
il •
, #
~I
6

• ;, •
iI .
,...
\01
EI
2\. • 21 i\
21
c:m • 11

.
. (JO),(l,)1 1

• ,
1
(2o}1 2\ 81, • il '1\
'2 \
(20)1 31 (2si 15 7
3 81'
1 •61
. ;\1
• 21. 141 · · 4 "1 l' (30)!
..~-:·.· el. !\
"I 14

Fig.7. EXAMPLE OF EGG SHOOTING RESULTS on wet paper. - The distance between
the steel gauze (cf. Fig. 6) and the base line of the photograph was 2 cm. The
numbers refer to the numbers of eggs in an egg cluster; estimated numbers in
parentheses.

to drop the eggs frorn dry places above sea level, e.g. cliffs, etc., with fairly good
results.
According to PROVENZANO (1962) larvae of Coenobita clypeatus hatch in a first
zoeal stage, after which a number of three, four or five other zoeal stages are to be
completed before they moult into the glaucothoe stage. Data from his laboratory
rearings at a ternperature of about 29°C and a salinity ranging from 33-36%0 are
shown in Table 9.
A specimen which lived for 31 days as a glaucothoe and very closely approached
the moult to the first crab stage, assumed a non-swimrning behaviour about one
week before it died.
Thus the total duration of the free-swimming larval stages in the laboratory
covers periods frorn 40 to over 60 days. As the duration of the larval stages may be
greatly influenced by diet as weIl as by temperature this period is probably shorter
und er natural conditions.
 47

Successive reproduction periods

From the observations in 1966 and 1967 it was evident that thc
first group of animals, completing a sequence of mating - egg
production - egg development - egg deposition, was followed by a
partially overlapping second group of breeding animals, which in its
turn was followed again by a third one. Studying the two observed
periods in 1966 on the Directiestrand and three others in 1967 from
both the Directiestrand and Oostpunt, a distinct lunar cycle is to be
noticed. Figure 5 gives the range of each period and also the phases
of the moon. Depending on the duration of the development of
the eggs (3-4 weeks), the liberation of the larvae takes place in the
week preceding full moon. In all cases the first young eggs were
observed round about full moon. Egg production on the 2 places at
Oostpunt and the Directiestrand seems to be synchronized. It is
supposed that the introductory mating activities are primarily
induced by the moon phase, roughly in the week before full moon.
Correlations between moon phase and other stages of the reproduc-
tion are therefore secondary.
The numbers of an im als observed always gave the impression that
the first group of breeding animals was the most important one; the
second and third groups of decreasing importance; however, no
exact data about the first and second groups can be given. In any
case the third group is very small in number. The observation of an
ovigerous fern ale in December 1966 is considered to be quite an
exception.
Other re cords of consecutive breeding periods of Coenobita clypea-
tus are absent, but it is acceptable that this phenomenon will only
occur in the central, warmer parts of its range, whereas in the
northern parts, as Florida, the Bahamas, Bermuda, only one
period occurs. Comparable data on other related species are also
lacking.
 48

Retreat to inland habitats

From the decreasing numbers of animals near the Lagoons

around the middle of September, coinciding with the end of the
first breeding period, it might be concluded that this first group
disappears into the interior. It does not seem likely that females of
this group stay and breed for a second time. Probably this holds
true for males too. Sex ratios do not show a marked decrease of
males. A number of the sampies from the Lagoons, however, gave
deviating sex ratios for the larger size classes in the period after
about September 15, showing an increase of fern ales or a relative
decrease of males. This of course will be influenced in the first
place by the direct depart of females towards the sea co ast in the
periods of egg deposition, which is also demonstrated in a very
distinct way by comparing the sex ratios found on both sides of the
Landbridge (Table 10).

T ABLE 10
SEX RATlOS IN Coenobita clypeatus POPULATIONS
OF THE LAND BRIDGE, 1967
Ratio = number of males divided by number of females, per size class. Size classes:
v.s. = very small, s. = small, m. = medium, 1. + v.l. = large + very large.
Average values computed from a great number of Oostpunt observations. Number
of sexed individuals in italies.

land side of Landbridge

27.IX 6.X 13.X 20.X 27.X average
values

v.s. 0.2323 0.50 2 0.0 2 0.04-0.25

s. 0.1120 1.14 I5 0.0 I2 0.1120 0.2024 0.25-0.75
m. 0.0520 0.36 I5 0. 1229 0.08 I3 0.1326 0.75-3.0
1. + v.I. 0.33 4 N 3 3.0 3 3.O-N
se a si d e of Landbridge

27.IX 6.X 13.X 20.X 27.X

allclasses 0.0 I02 0.0 88

together
 49

About the retreat from the breeding places towards the inland
habitats most details are absent. Probably this takes more time and
goes on in a more dispersed and less directional fashion. There is no
direet evidence that an im als originating from WeH 4, visiting the
breeding place near the Lagoons, returned again to the same WeH,
though 18 tagged animals were observed here after the period
coinciding with the breeding season, in which not a single one could
be observed near WeH 4.
Of the 304 orange (new) tagged animals, caught and released
near the Lagoons, only 2 could be traced afterwards, again more
westwards in a manchineel grove near Awa Blancu. Not a single of
the orange or yellow animals released near the Lagoons was ob-
served ne ar WeIl 4.
In general the way back to the inland will be far more difficult
to trace. The wet season has started now and rainfaH in November
and December in Cura<;ao is high er and more frequent in comparison
to the beginning of the breeding season. Drinking places do not
specially attraet Coenobita, making observation not easy during
that time.
Sooner or later the animals will return to their favourite inland
habitats. As could be proved, even three years later tagged animals
were still numerous in the hermit population near WeIl 4. According
to observations by PEARSE (1929), however, there is apparently no
evidence of a homing instinet directed towards fixed hiding places.
Within a certain habitat an animal "beds down" in the nearest
convenient shelter and returns to the same spot more than onee
by chance only.
 H. ADDITIONAL NOTES ON THE
ECOLOGY OF COENOBITA

Land hermit crabs are animal species characteristic for coasts

and islands in tropical and sometimes also in subtropical areas, the
latter when washed ashore by warm currents. That they are still
bound to the sea is primarily due to the fact that the fertilized eggs,
at first attached to the female's abdomen, have to be brought to a
marine environment to permit further larval development. More-
over, the sea is also the main provider of the larger gastropod shells,
which serve as exchangeable houses. Another factor may be the
presence of salt (chloride ion) in the environment (BOUSFIELD,
1968). In arid areas the often higher relative humidity of the coastal
air may be of importance, while sea water of normal salinity always
remains a safe and stable source of water, to fall back on in cases of
emergency.

HABITAT

As to its habitat Coenobita clypeatus is not very particular. The

areas concerned comprise such widely divergent sceneries as desert-
like coasts, and shady parklands with shrubs and trees, sometimes
up to a considerable altitude. The common feature in all these
areas is that the sea is not farther away than 10 km, or 15 km at
the utmost. Because of their way of locomotion land hermits prefer
open grounds, preferably with a dry and hard soil; dense vege-
tations of herbs and grasses are rarely visited. Contrary to what
might be expected hermit crabs avoid permanently humid areas,
marshland, muddy banks of salinas and freshwater pools.
 51

In Chapter I the characteristics of the area inhabited by Coenobita

on the Leeward Group have been discussed in some detail. Of
other places visited by the author especially J am ai ca and Bimini
must be mentioned because of thcir unusual character.

In Ja mai c a flourishing and numerous populations were found on the keys off
Kingston Harbour. On Drunkenmans Cay, the size of which is probably not over
one hectare, partly rocky with a vegetation of Rhizophora, partly a sand-bank, with
a sparse vegetation of halophytes, many hundreds of medium-sized an im als were
found on the fallen leaves under thick mangroves.
During day-time the inactive animals had no further shelter. Some of them
moved about or sat with the first antennae projecting. The peculiar thing is that
the animals live and prosper in such a small area. There is no fresh water, except
when it rains. The climate in this part of J amaica is much the same as in Curayao.
Besides water taken up with the food - which is thought to consist mainly of edible
matter from the shoreline - water is probably taken up from the moist humus as
weil. The moist micro-climate under the hothouse-like mangroves implies a low
evaporation. The rest will have to co me from the sea water.
Actually these animals should be classed as "coastal animals," however, the
shabbiness, as known from the Curayao animals, is lacking altogether. Large
Livona shells can be found everywhere. There is no hinterland and therefore no
inland migration. Since all reproductive activities must be performed along the
restricted and isolated coastline this population would be admirably suited for
further research on reproduction.

Thenumerousanimalsfoundon North Bimini in the monotonous Casuarine

groves also merit closer attention. The Bimini group is situated on the western
edge of the Great Bahama Bank, approximately 60 miles from the Florida coast.
The climate is more temperate than that of Curayao. Early November 1967, during
the author's visit, it frequently rained during the night, with temperatures slightly
over 20°C. By day it was sunny, with temperatures around 25°C. Smaller animals,
mostly in white Polinices shells, lived in holes in the thick layer of fallen Casuarina
needles. Larger animals practically all in possession of Phalium, Tonna and es-
pecially Fasciolaria shells led more nomadic lives and were even active during day
time, in blazing sunlight. It could be concluded from the fact that unicellular
green algae grew in the siphon al groove of the occupied shells, that this diurnal
activity must have been going on over long periods. The animals were strongly
attracted by a refuse dump of an adjoining hotel. Some small animals foraged in
the branches of shrubs.
In the southern part of North Bimini the - mostly small - animals were to be
found in the neglected gardens around the Lerner Marine Laboratory, both by
night and day. At night considerable numbers of small animals were seen on the
sandy beach at the western side of the island.
In Key Largo (Florida), Barbados, and Trinidad solitairy specimens of Coenobita
were seen accidentally, but only during the night.
 52

TEMPERATURE
The distribution of terrestrial decapods, and particularly of
Coenobitidae, is restricted to areas with hot climates, thus re-
flecting their great dependence on temperature. Especially low
temperatures limit their activity and determine their natural
boundaries. Being poikilothermous animals, they more or less
conform to the temperature of their surroundings, however, by
choosing the correct miero-habitat and by a certain amount of
individual regulation of temperature (Buss, 1968) they can avoid
extreme body temperatures for a certain time. The optimal temper-
ature range for Coenobita clypeatus - probably between 22 and
34°C - is practically the same as the general range of temperatures
in the whole Caribbean area, where those boundaries are only
exeptionally exceeded. Extreme temperatures of 37°C and more,
as recorded for the coastal areas of Venezuela and Colombia, seern
less suitable. When such temperatures were offered in laboratory
experiments a distinct decrease of activity was observed, which
mayaiso be connected with retention of water. Though it is not
exactly known what the lethai temperatures are - whieh anyhow
is only a doubtful concept without a time factor - it does not seem
likely that in their natural habitats the animals are ever exposed
to such temperatures. Their strong tendency to seek adequate
shelter, together with their predominantly nocturnal habits protects
them from extreme temperatures during the day. Therefore it is
not necessary to consider the inevitable evaporation as having an
important function in keeping the body temperature below a
certain limit (Buss, 1968). The lower temperature delimitation is
found at 20-22°C, where all movements get much slower. Animals
kept at 18°C for a longer time became more or less lethargie.
The northern boundary of the distribution of Coenobita clypeatus
closely reflects the restriction of its chance of life at lower temper-
atures. On the North American continent the species is found
from the east co ast of Florida to about latitude 27°N (PROVENZANO,
pers. comm.). In the more temperate Bahamas they occur farther
to the north and in the even more oceanic Bermudas they reach
their northernmost limit at 32°N. With average February temper-
 S3

atures of 17°C and lowest temperatures recorded of ab out 4°C the

species here is at the margin of its existence. According to recent
information (W. E. STERRER in litt. 1971; ABBoTT, 1972) Coenobita
is now extremely rare on the Bermudas, possibly due to the absence
of suitable big Livona shells. This may be doubted, however, since
even specimens in very small shells reach sexual maturity.

The diurnal activity of Coenobita on Bimini may be explained by the consider-

able differences between day and night temperatures. If at night the temperature
falls below the optimum range, while during the day it stays within that range, the
normal night-and-day rhythm may be discontinued, humidity conditions per-
mitting. If in the winter months the temperature during the day also remains too
low, the animals do not come out at all. Good shelter offers the only opportunity
for withstanding relatively low winter-temperatures.
According to WÜST (1969) in the warrnest months (August-September) - the
time when the larval development of Coenobita clypeatus is in progress - sea surface
temperatures around 28°C are to be found all over the Caribbean. In the waters
around the Bermudas, however, the temperatures remain some degrees lower.
The somewhat earlier and shorter reproductive season in the more northern areas
as compared to Cura<;ao is a question of temperature.

FOOD

Land hermit crabs are omnivorous animals feeding on all kinds of

vegetable matter as weIl as on protein-rich food, especially on
carrion. If they do not get the right kind of treatment in the
laboratory, cannibalism may occur frequently. Birgus latro, the
coconut- or robber crab, feeds especially on protein-rich Cocos and
Pandanus-fruits (HARMS, 1932, 1937). In Cura<;ao the food of
Coenobita comprises fruits, berries, seeds or germinating plants,
whatever is to be found in the field. In such pI aces sometimes large
concentrations of animals are to be found, especially under the
fruit trees in sm all orchards. All kinds of wild fruit, e.g. cactus-
fruits or the "apples" of the manchineel tree, which are poisonous
to most animals, are taken eagerly. In the dry season the disks of
Opuntia and the trunks of Cereus are also gnawed, probably be-
cause of the moisture they contain. N ewly-deposited droppings of
horses and cows are taken as a source of food and water. Carrion is
veryattractive.
 54

On Oostpunt a dead donkey provided food for hundreds of animals for weeks
on end. Even when only an empty dried skin was left, animals were still entering
it through the various holes. The dry skin funetioned as asounding box and the
seuttling of all these animals eould be heard over a great distanee.
Coenobita is useful as a seavenger; not only on the waterfront but also around
human habitations, refuse dumps, kitehen gutters, ete. Aeeording to loeal fishermen
soldier erabs ean best be lured by grounds of eoffee.

WATER

In other chapters the role of both fresh and salt water in the
environment will be discussed in detail. The occurrence of Coenobita
in more inland habitats is closely connected with the availability
of fresh or brackish water in these areas. Therefore it is the more
remarkable that the animals are usually absent in places where there
is plenty of fresh water. Even Poos Grandi and Poos Manzanilla, the
two main rainwater basins on Oostpunt, filled with fresh water for
the greater part of the year, do not seem to have any attraction
at all. Neither near the other man-made reservoirs on Curac;ao,
where rainwater is retained behind dikes of earth, hermit crabs
were practically ever found. They are completely absent in plains
which are inundated in years with a high rainfall, where often
luxurious swamp vegetations have developed, while those areas
are still inhabited by Cardisoma. In the same way land hermits
seem to be rare, or absent altogether, in any other area with a high
rainfall. Fresh water seems to become attractive only if it is present
in scarce quantities in an otherwise arid area. The water manage-
ment of Coenobita is apparently based on an equilibrium main-
tained by, on the one hand, the pressure of a constant strong
evaporation and on the other hand a subtle method of uptake from
the environment (Chapters IU and IV). Heavy rains may disturb
the water regulation within a short time (Chapter IV). Water for
the regulation of shell water is taken up in proportion to the salt
content. The still hypothetical or partly revealed mechanisms for
uptake of water from moist substrates (this paper; BLISS, 1968) or
even from the atmosphere may be only an inconvenience to the
animal in a wet environment.
It is generally accepted that the presence of salts has an essential
function in active water transport. This might even be of still
 ss
greater importance to aland hermit crab with its delicately balanced
water management. Just as in our endurance tests with Coenobita
(Chapter V), BLIss, WANG & MARTINEZ (1966) demonstrated for
Gecarcinus lateralis that the animals remained in a better condition
if the available drinking water contained a certain percentage of
salto The presence of both fresh water and a salt spray carried by
the wind into the inland regions of Curac;ao might create a favour-
able environment for Coenobita, similar to the situation suggested
by BLIss (1968) for Gecarcinus. It is another example of the extent
to which the animals depend on the sea even in inland regions.
 LABORATORY EXPERIMENTS

IH. WATER MANAGEMENT

INTRODUCTION

Even the most terrestrial species of decapods are still not com-
pletely independent of the sea and thus cannot be considered
completely successfulland animals. It cannot be doubted that in the
general course of evolution the animals are only in a transitionary
stage and it is still an open quest ion whether they will ever succeed
in obtaining a complete independence of their environment as far
as water is concerned.
It is certain that the direct step hom sea to land is not an easy
one. One might expect that such a step would stand a better chance
of success in a moist climate. Still, terrestrial crabs are often typical
of tropical coastal areas in arid regions. Among terrestrial hermit
crabs Coenobita scaevola is a very common species of the arid coasts
of the Red Sea. Coenobita clypeatus is found especially on the arid
northern co ast of the South American continent and a great many
islands in the Caribbean.
Terrestrial insects have solved the problems of living very
successfully in dry air, mainly by economizing their water manage-
ment to a high degree. Evaporation is restricted to a minimum by
the complex construction of the tegument and by the existence of
adjustable spiracula. Water is imbibed by drinking or with the
food, but it may also be taken up from the air, or originate from
metabolism. Water loss through excretion is very small due to
rectal re absorption (PROSSER & BROWN, 1962).
It is to be expected that a terrestrial crab, being originally an
 57

aquatic animal, will have problems in its water- and salt-manage-

ment. It appears that for this group their success as terrestrial
species depends on a great number of morphological, physiological
and ecological adaptations as weH as on a special behaviour,
adapted to an economical use of water.

MATERIAL

For the greater part of the year Coenobita clypeatus is easily

obtainable in Cura<;ao; only during moist weather conditions when
the usual drinking spots are obviously of little attraction, the
animals disperse in the fields and are hard to collect. During the
reproductive season, ranging from July till October, most animals
live in the neighbourhood of the breeding places along the coast.

Animals used for laboratory experiments were mainly collected in the sur-
roundings of the old country house "Santa Catharina" on the north coast and in
the yard of the country house "Klein Si nt Joris" on Oostpunt. Both habitats are
very similar. There is plenty of (fresh) drinking water and food. Moreover, in both
places there are suitable hiding places and shady trees, including fruit trees. The
greater part of the animals collected here belong to the type of inland animals;
however, since the distance to the north coast is rather short, there are relatively
high numbers of smaller individuals. Very large specimens are absent or rare,
probably due to fishermen, who now and again collect hermit crabs for bait.
Unless stated otherwise only medium-sized animals were collected and used for
laboratory experiments. Newly collected animals were placed either in wo oden
boxes with access to fresh or/and sea water, or in the outdoor crab pen. The latter
was a 40 cm high, gauze-covered, concrete construction of 4 m 2, situated under the
machineel trees in the grounds of the Institute. There were a 401 water trough ip.
one corner and against the two opposite walls long rows of hiding places constructed
of concrete cable-tiles. A weIl-projecting iron roof, about 2 m high over the pen,
gave shelter against rainfall.
In order to be informed on the main ecological conditions an electric temper-
ature, as weIl as an electric humidity probe, were mounted in the pen and eonnected
by cables with an automatie recorder in the laboratory.
Land hermit erabs survived very weIl in the pen, providing that sufficient food
and water were supplied and, even at densities of several hundreds of medium-
sized animals, mortality was found to be very low.
In most eases animals used in experiments were treated beforehand in various
ways, in order to obtain a more uniform and properly adapted material. Depending
on the way of treatment the following types of treated animals are to be dis-
tinguished:
normal-animals: animals which have had access to both fresh-(tap) and sea
water for at least two weeks;
 58

aqua destillata (A.D.) animals: animals throroughly rinsed and partially im-
mersed in distilled water for 24 hours and afterwards kept in clean glass containers
with access to distilled water only;
sea water (S.W.) animals: like A.D.-animals, but rinsed and immersed in 100%
S.W.only;
desiccaled (D.) animals: kept for at least 7 days without any access to water.

METHODS

Osmoconcentrations of fluids always were determined by the total number of

dissolved particles, both electrolytes and non-electrolytes. Here the osmoconcen-
tration of blood, urine and sometimes of shell water was measured by means of
freezing-point determinations. For this purpose the micro-cryoscopic method and
apparatus as described by HOHEN DORF (1963) were used (Fig. 8).
The precision of the thermometer readings is about 0.005°C. In 1965, working
at the Institut für Meereskunde in Kiel, with exactly the same apparatus as used
by HOHENDORF, the same degree of exactitude as mentioned by hirn was reached
by the author.
Blood was always sampled by means of a small glass capillary penetrating the
membranous joints of the walking legs. Urine was sampled with a small curved
pipette, which was brought into the urine bladder through the porus at the base

Fig. 8. DEVICE FOR MEASURING FREEZING POINT DEPRESSIONS. - Small sampies

of body fluids in glass capillaries are glued to a copper fork (front) and frozen in
a Kryomat (left) . Melting ice crystals are observed in a inelting chamber by means
of a microscope and melting points read on a thermometer.
 59

of the second antennae. Neither blood nor urine were centrifuged before use, but
as soon as possible they were prepared for cryoscopic determinations. As a rule
the blood starts to clot soon after sampling and in 10 to 15 minutes it becomes
gelatinous and therefore useless for proper determinations.
Theoretically sampIes of 0.3-3 y are sufficient for one determination but in
practice sampIes of 10-100 Y are more efficient for a proper filling of the capillaries.
Fresh sampIes of blood and urine were put immediately in watch glasses under a
high quality liquid paraffin, after which as so on as possible glass capillaries with a
diameter of 50-100 [L were filled with alternating drops of blood (or urine) and
paraffin, in such a way that every capillary held about 10 drops from the same
sampIe. A maximum of four capillaries, holding 4 different sampIes could be ex-
amined simultaneously. The sampIes were frozen in a "Lauda Tisch-Kryomat"
type T.K. 30 D, at -36°C for aperiod of one hour. Melting ice crystals were ob-
served in a melting chamber with a Zeiss Stereo Microscope III with a zoom system.
A very slow and gradual rise of temperature in the melting chamber is absolutely
necessary for accurate readings. Therefore the rise in temperature was always
fixed at O.OloC per 1-2 minutes. Readings were made at the moment of disap-
pearance of the last ice crystal of each drop. The average value of the first 5 or 6
fully melted drops per capillary was considered to be the freezing point depression
of the sampIe.
For shell water and sometimes also for the larger quantaties of blood - if concen-
trations of the total amount of electrically active particles were desired - determi-
nation of the electrical conductivity was performed in the same way as used by
HOHENDORF (1963), with the only difference that a Philips Direct Reading Con-
ductivity Meter, type P.R. 9501 was used. The micro-measuring cells permitted
volumes of about 0.04-0.05 mm 3 . Since, however, quantities 2 of 3 times as large
were more satisfactory, the measuring chamber could first be rinsed with the
liquid before measuring. All determinations were carried out at a constant temper-
ature of 25°C; the results were recorded in milli-Siemens cm- 1 and afterwards
converted into salinities per thousand.
Blood was sampled in a similar way as described for cryoscopic determinations,
but never stored under paraffin. Quick measurements are needed to prevent clotting
in the measuring chamber.
Shell water was knocked out oi the shells, caught in solid watchglasses and
measured immediately afterwards, or - when sampling occurred in the field -
stored in 2 ml glass jars with plastic caps and measured the next day.
For measuring drinking water and other water sampIes of a sufficient quantity,
a Philips' Macro-measuring-cell, type P.R. 9513 was used, again at a measuring
temperature of 25°C.
In order to be able to compare the results of both cryoscopic and conductivity
measurements, most values were converted into salinities per thousand. To this
end conversion-diagrams were made based on sea water dilutions and concen-
trations of known salinities by Mohr's titrations.
Beside the use of salinities, percentages of sea water (% S. W.) are often used to define
concentrations of drinking water offered to Coenobita, in particular because the prepa-
ration of dilutions or concentrations of sea water is simple. The saline contents of
the Caribbean sea water proved to be a little higher than "standard" sea water.
(Compare LA.P.O. Standard Sea water with a salinity of 34.95%0)' In general the
surface salinity of the sea water around Curas;ao fluctuates between 35.80 and
 60

36.80%0 (WÜST, 1964). In this paper 100% S.W. (normal sea water) is always
based on a salinity of 36.00%0'
Temperature was usually determined by sensitive mercury thermometers up to
aprecision of O.OI°C.
lf permanent readings were required two Philips miniature Pt. 100 Ohms re-
sistance bulbs, type PZ, ± 0.1 at O°C, connected with a Philips automatie recorder
PR 3210 u/OO, were at my disposal. lf humidity and temperature were to be meas-
ured and recorded simultaneously, two nickel resistance thermometers Philips PR
6002 B were used (PHILIPS, 1963).
Permanent information about relative humidity of the air in experiments under
laboratory conditions was obtained by using a combination of the aforementioned
nickel resistance thermometers and Lithium-chloride Dew Point Hygrometers,
Philips, type HA, both connected with the recorder.
In 1961 a Karl Weiss' Electric Hygrometer, "Hygronom," type nr 1381, was
used for field work, whereas in so me preference experiments in the humidity-
apparatus dewpoint meters were used, according to the principle of the conden-
sation hygrometer of Regnault, as designed by Professor A. PUNT of the University
of Amsterdam.
For general information about elimate and weather conditions in Cura<;:ao the
data of the Meteorological Service Netherlands Antilles, Dr. A. Plesman Airport,
Cura<;:ao, were consulted. The meteorological station is situated near the North
coast of the island, about 7 km North of the Marine Biological Institute. Especially
data on daily rainfall measured in a number of rain-gauges in the eastern part of
Cura<;:ao were most valuable.
Description of more specific apparatus and methods is given in the chapters in
question.

GENERAL WATER ECONOMY

Only a few representatives of the class Crustacea have succeeded

in solving the difficulties of a terrestriallife in a more or less satis-
factory way. Most of these and related problems are compilated
and summarized by EDNEY (1957, 1960).
Within the Decapoda only two groups show successful terrestrial forms: Land
hermit-crabs of the Family Coenobitidae and True crabs of the Family Gecarcinidae.
Semi-terrestrial forms within the Decapoda are much more common and a rather
large number of species can stay for short times out of the water and show this
behaviour in their normal way of life.
More or less semi-terrestrial species always keep water elose at hand; either
they can stay within a short distance from the shore, or they have burrows reaching
down to the ground water. By immersing their bodies from time to time, at which
occasion water mayaiso very weIl be taken up, these animals avoid serious water
problems (GROSS, 1955).
Genuine land crabs such as Birgus latro, various species of Coenobita, and Gecar-
cinus lateralis have become far more adapted to life on land (HARMS, 1929, 1932,
1937; GROSS, 1964; BLISS, 1968). They can stay for many days without taking up
 61

water and are able to roam far from their sourees. This enables them to live nearly
completely on fresh water. For the development of the larvae, however, they are
still dependent on the sea and mature females have to go to the ocean to release
their offspring. As is mentioned in this paper, it is not necessary for Coenobita
clypeatus to enter the water for this purpose.

Water economy roughly involves three major parts: uptake,

retaining and loss. Land crustaceans permanently suffer loss of
water by evaporation. Provisions leading to economy in water
consumption will be very important.

Uptake

Usually most water is taken up by drin king or with the food.

Moreover, Coenobita is able to store water in its shell. This water
may be used again partially as drinking water. Whether this
water is also taken up by permeation through the soft tegument
of the abdomen is not known.
Drinking and uptake of shell water in Coenobita clypeatus takes
place in an active way. For other species another (passive) way of
uptake of shell water is described.
By dipping the tips of the chelipeds, in particular of the minor
right chela, in water, drops are lifted, handed over to the third
maxillipeds and put into the mouth and gill cavity. Tufts of setae
on chelae and maxillipeds playa role in this proces. This way of
water uptake is very efficient and the slightest amounts of water
e.g. moisture on surfaces of objects, occasional raindrops, etc. are
utilized. Sometimes a more indirect way of drinking is found in
which the tips of both chelipeds, held dose together, are put into
the water. By shoveling movements of the chelipeds, which bear
setae on their innerside, capillary water rises to the mouthparts.
Finally the maxillipeds bring the water into the mouth. VÖLKER
(1965) described the same behaviour for Coenobita scaevola.
Coenobita clypeatus prefers drinking water of a rather low salinity
but also water of higher concentration, up to a salinity of ab out
36%0 (= 100% S.W.) or even higher, is used. The higher concen-
trations, however, are unsuitable for use as shell water.
 62

The way in which Coenobita clypeatus is able to trace water and

the discrimination between water of different salinities points to a
highly developed sensory perception.

Preference experiments

In order to obtain indications as to the way in which Coenobita clypeatus detects

the presence of water and to which extent it discriminates between drinking water
of different salinities, some preliminary preference experiments were carried out,
which are briefly discussed below.

Soil humidity
Preference experiments on soil humidity were conducted in a narrow and rather
high tank of a length of 220 cm, in which a slope of sieved coral sand had been set
up. A considerable time after the sand had been moistened a tolerable gradient of
soil humidity had established itself along the sand's surface. The process was
checked by determining the wet and dry weights of sand samples. Along the sides
of the tank alternating transverse partitions had been made, to provide the neces-
sary shelter for the animals, in such a way that in this respect any part of the tank
was just as attractive as the other parts. The whole apparatus could be shut off
with shutters. Moreover an air current could be sent down the whole length of the
sand's surface by turning on a small ventilator.
With the ventilator turned off regular counts of the number of animals in the
respective compartments showed that they clearly preferred a zone at a distance
of about ! from the lower end of the tank. The lowest part, where the sand felt wet,
was avoided. The fact, however, that this observed orientation is shattered by
turning on the ventilator, led to the conclusion that the animals do not use the
soil humidity as such, but rather the gradient of humidity in the air just over the
soH. With an air current this orientation is disturbed or prevented fram establishing
itself, so that orientation is made impossible.

Air humidity
The orientation on relative air humidity was tested in another. apparatus. In an
elongated deep tank bowls with evaporators as well as bowls with so me odourless
hygroscopic liquid (sulphuric acid) were put in such a way that a humidity gradient
between about 50 and 100% R.H. was established. Over the bowls an experimental
gutter was constructed from fine meshed wire, also provided with alternating
transverse partitions. It was covered by a sheet of plateglass, with the dew-point
meters screwed in at regular intervals. All experiments were carried out in the dark.
From the frequency distributions observed in the gutter compartments it could
be concluded that without restriction the animals orientated themselves on the
relative humidity of the air in such a way that they always went to the place with
the highest humidity.
 63

Salinity
In a third series of experiments it was tried to find out in how far Coenobita is
able to discriminate between available drinking water of different salinities. To
this end the basal part of the drinking apparatus described in Chapter IV was
encircled by a round run of wire, divided by partitions into five compartments

TABLE 11
PREFERENCE EXPERIMENTS IN COENOBITA CLYPEATUS

Difference in salinity expressed in % sea water.

Differences in concentration in the order of 1 per cent are probably still perceived.
X2 are calculated from results in the columns pos. and neg.
Totals in italics.

difference range treatment of number of results

in salinity animals animals pos. neg. indif. X2
0.5% S.W. 0-0.5 dist. water 7 2 5 2.0
5-5.5 dist. water 25 10 10 5 0
99.5-100 dist. water 15 8 1 6 5.44
99.5-100 sea water 16 8 8 0
120-120.5 sea water 75 26 39 10 2.60

I38 52 60 26 0·57

I%S.W. 0-1 dist. water 11 4 3 4 0.14

5-6 dist. water 35 20 9 6 4.17
10-11 dist. water 31 17 8 6 3.24
99-100 sea water 17} 16 7 6 3.52
99-100 desiccated 12
120-121 sea water 18 13 4 4.76

I24 70 3I 23 7·53

2%S.W. 0-2 dist. water 23 11 6 6 1.47

10-12 dist. water 3 2 1 0.33
98-100 seawater 4 4 4.0

30 I3 II 6 0.I7

3%S.W. 0-3 dist. water 11 7 3 1.60

10-13 dist. water 3 3 3.0
97-100 sea water 4 3 1.0
I8 I3 4 I 4.7 6

4%S.W. 0-4 dist. water 12 6 3 3 1.0

5%S.W. 120-125 125% sea water 10 7 2 1 2.78
6%S.W. 0-6 dist. water 46 27 8 11 10.31
 64

of equal size. Every compartment held two drinking places, connected with long
and slender reservoir tubes. The water in the tubes differed in salinity. One animal
was put into each compartment, and after a day the quantity of the water used from
each of the reservoirs was measured. If the consumption from the two tubes differed
less than 0.1 ml the animal was considered to have shown no preference.
In most cases the animals used in these experiments were given a preliminary
treatment, to enforce a better choice, if the two salinities were nearly the same.
One might assurne that an A.D.-animal, or in other words an animal with a shortage
of ions, will discriminate more acutely in the region of the lower salinities, 0-0,5%
S.W., while e.g. a D.-animal or S.W.-animal profits most by a right choice in the
region of the higher salinities, 120-120.5% S.W.
From the results obtained, which are best explained by table 11 one is led to
conclude that differences in sea-water concentrations between t and 1% sea-water,
or in the order of 0.18-0.36%. salinity may still be perceived.
Doubtlessly land hermits, which are omnivorous animals, will take considerable
quantities of water with their food: fruit, succulent parts of plants, carrion, etc.
It is evident, however, that these sources of water will be highly reduced in periods
of drought.
As was observed in the field and subsequently also proved experimentally,
limestone containing water may serve as a source of water in the dry season. By
eating small pieces of limestone the animals are able to meet the need of water, to
a certain extent.
Coenobita stores large amounts of orange-coloured fat or oily substances in its
abdomen. Yet it seems that quantities of water formed in the process of fat me-
tabolism will contribute only in a minor way to the water economy of the animals.

Retainment and 10ss

Both are closely related to each other, as a less economic management will lead
to a more extensive loss of water.

Evaporation on the respiratory surfaces

Though the typical gills of the marine hermit crabs have undergone a number of
modifications to make them more adapted to a terrestrial environment, such as a
reduction of the number of gills, sclerotization, etc., still the respiratory surfaces
have to be kept moist to enable a proper respiration. Moreover in this group of
animals the respiration is highly supported by vascularisation of the walls of the
gill cavities (HARMS, 1929; EDNEY, 1960). Air ventilation is supported by lifting
the pleural margins of the carapace. Finally the most important feature in these
animals is probably the skin respiration by the soft and moist tegument of the
abdomen. Of course a proper gas exchange is possible only through more or less
moist body surfaces; but this has the dis advantage of a serious water loss.
Adaptations to reduce evaporation on the respiratory surfaces appear in the
shape of a reduced number of gills and also in a diminishing of the effective surface
of the gills. The gill cavities are no longer filled with water, neither completely nor
 6S

partially. Well-developed glands in the branchial cavity help to moisten the gills
(HARMS, 1929). The gills act as a lung. The cavity has narrow connections with
the air outside the anima!. The moist abdomen is enveloped by a shell, the latter
being carefully selected as to quality and shape (DE WILDE, unpub!.; VÖLKER,
1967).

Evaporation by the tegument

Evaporation through the exoskeleton of terrestrial crustaceans is highly affected

by temperature, humidity and wind. It acts more or less as if it were a physical
body (EDNEY, 1960). The tegument of terrestrial crustaceans, however, seems to
be less permeable to water than that of non-terrestrial related species. According
to HARMS (1932) the better ca1cification of the exoskeleton of Birgus should reduce
the permeability. Layers of a waxlike substance, as are known in insects and
spiders, do not occur (EDNEY, 1960).

Loss of water by excretion

Decapod crustaceans possess paired segmental excretory organs; the antennal or

green glands. The excretion product is urine. End products of the nitrogen me-
tabolism are discharged in this way and may playa role in the osmoregulation of
the animals. The osmoconcentration of the thin, liquid urine of Coenobita clypeatus
was found to be isosmotic with the haemolymph. Little is known about the quanti-
ties of water which are thus discharged. Loss of water by defaecation seems very
low in times of water shortage. Faecal pellets are sometimes very dry and resorption
of water in the end gut seems most likely.

Loss of water by ecdysis

Although loss of water during the moult will be considerable, this event, which
only takes place at long intervals, needs not to be considered in the scope of this
paper. The animals hide themselves as much as possible. The moult largely takes
place inside the shell. It was observed how a moulting animal in a glass-blown
shell managed to retain a certain amount of shell water in the shell during the
moult. Recently moulted, still pale-pink coloured Coenobitas are always soft and
moist to the touch and will desiccate very soon without the shelter of the shell and
the hiding place.

Nocturna1 activity

Besides direct provisions to reduce the 10ss of water a number of

more or 1ess secondary adaptations, which serve the water economy,
must be mentioned here.
 66

In the dry southern parts of the Caribbean Coenobita is almost

completely nocturnal. At night the temperature is somewhat lower
and consequently the relative humidity of the air is higher. The
strong radiation of the sun is absent. Therefore the evaporation
will be much lower at night. During daytime Coenobita stays
motionless in its hiding place: in natural crevices, existing holes
and tree stumps, under boulders, heaps of mould and decaying
leaves, etc. Burrowing in loose earth as described for other species
(VÖLKER, 1965) occurred in captivity, but was never observed in
the natural habitats. Favourable places are sometimes inhabited
by dozens of animals, staying in compact clusters. Such hiding
places, though often found in very dry places, may show high
humidities. In burrows of the terrestrial species Gecarcinus, BLISS
(1956) found relative humidities of 95-100 per cent.
Provided land hermit crabs have a shell of the right size and the
right shape, the animals are able to withdraw completely in the
shell and close the aperture of the shell with chelipeds and legs.
VÖLKER (1965) gives a detailed description of the closing mechanism
as shown by various groups of hermit crabs, in particular of the
terrestrial species.
In Cura<;ao some of the adaptations mentioned here or mecha-
nisms concerning the water management of Coenobita clypeatus
were further investigated.

Method

Coenobita clypeatus is considered to be nocturnal in its activity (PALMER, 1971).

This was evident from own observations in the Leeward Group of the Netherlands
Antilles as well as from the scarce notes on this subject by other authors, a.o.
PEARSE (1916), CARPENTER & LOGAN (1945). HAZLETT (1966), who made obser-
vations on the social behaviour of the species in Cura~ao says: "Tn Cura~ao land
hermit crabs were found at night, except for a few individuals found in the caves
on the north side of the island."

Since a nocturnal activity involves certain advantages concerning the water

management of terrestrial animals living in dry habitats in general, attempts were
made to study the activities of Coenobita clypeatus in the laboratory. To this end
the activity of about 200 tagged animals, living in the outdoor crab pen, was
observed. The animals used had already spent ten days for acclimatization in the
 67

pen, prior to the first observations on April 23, 1967. Every day at about 17.00 h
cooked rice, bread and fruit, and sometimes fish or other food rich in proteins was
placed in the pen. The drinking trough held always fresh water. Obviously the
animals were in optimal condition and mortality was observed to be negligible.
A temperature- and humidity probe, placed in the centre of the pen, about 20 cm
over the bottom, and connected with the recorder, gave permanent information
on temperature and humidity.
All visible animals in the pen were considered to be active; non visible specimens,
e.g. animals hiding under the tiles, were considered to be at rest. This rather simple
criterion as an indication for the activity of Coenobita was most useful in practice.
Counting the visible specimens takes only little time and therefore hardly disturbs
active animals. In particular this was an advantage at night, when a small light
source was used. Moreover, a small number of observations was made in order to
specify the activity of the visible animals as to feeding, drinking, locomotion,
attitude activity (the animal is nearly immobile, but the body is raised and protrudes
partially out of the shell, often the antennulae are moving) and inactivity (the
animal is drawn back in the shell and the shell is closed). These observations showed
a very small percentage of inactive animals. Of course motionless animals which
were observed to remain for a long time in the same place, were also considered to
be at rest.
Additional inspections to observe the hidden animals in the compartments, by
suddenly lifting the covering tiles, usually demonstrated inactive Coenobitas.
Exceptionally active animals were also observed in the compartments, e.g. shortly
before sunset or during sudden showers. In such cases, however, nearly all animals
in a compartment were active, apparently ready to leave their hiding place.

Results and interpretation

The results of the described experiment are combined in Figure 9.

Here the activity, as indicated by the total numbers of visible
(active) animals and the relative humidity of the air in percentages
are plotted against time. Night and day (from 6.15-18.45 h) and
the occurrence of rainfall are given as weIl.
It was a favourable coincidence that during the course of the
experiment weather conditions proved to be rather variable.
Beside completely dry days - the normal type of weather in this
month of the year - a number of short showers also occurred.
April 26, and to a lesser extent May 5, were even wet days with
drizzling rain over rather extended periods.
The activity curve shows a distinct rhythm; activity being
high at night and low during the day. As an effect of the differences
in temperature between day and night, the humidity curve in
Number acllve ommals Relative humidily %
90
1~~
120 ~:illli~~~L
3 rMal W
r::I::::::::::::::::::1 rl 1"'::::::::::::1=::::::::::::::::1 !:::::;:;:;'{:/::",:V ~_ ';'&!IM"'I/~ .1' 0fu': , 80
70
60
40
.Cl 60
50
40
2~ j~.~~.r,:T ,Y, 1/:: ;'·:···:f:<::::;:·}/J ~ , , 1, 'E/:T?i\{:i: J;~;:-;, , , c, H:;:';'::d':///f?J """ r::::,;:·,::;::l- 30
o 6 12 18 24 6 12 18 24 6 12 18 24 6 12 ;0 nou,"
90
:~~ j:;'fWJ~ /f!ill:::::J::::::::::::::::::1 j: : : : : : : L:: : :': : : : : I ·~' F:::::::::::::k:::::::::::::::::l i 'i : f 80
70
60 60
40 50
40
2~ 3:.~!~:~:\: :t\-d !:::::::::.:·:::,r :::::::::::::::::l-.. 1.7~ .' "1": :.,.:.:.: :.: ] ~.,..-;' 1::::::':.:::'::::::::::::::::::::::::::1 1':':':':'::;'<::1
o ' '6 12 18 ""24" "6" " '12' '18"" 24" '6" 12 18"" '24" "6 '12 ""18 'tiours' 30 0"-
90 CO
80
I~~ lmfttJ '. [141 Ewi~ü:jmt\. [1 fl:w~gH~g;IfIl @Mi@#hil~Itll !UMf:f. 70
60 60
40 50
20 40
0 30
120 90
100 80
80 70
60 60
40 50
20 - 40
0 30
0 6 12 18 24 6 12 18 24 6 12 18 24 6 12 18 hou rs
 69

general gives the same daily fluctuations, i.e. lower R.H. during
the day and high er humidity at night.
Presumably the basic pattern of the activity of Coenobita is
based on anormal 24 hours cyde. P ALMER (1971) found that
nocturnalism in this species is under control of a biological dock.
About noon the activity is lowest. Activity increases a liWe during
the afternoon and shows a sudden and pronounced rise shortly
after sunset (18.45 h). The highest activity is reached between
20.00 and 21.00 h. From now on little by little the activity decreases
until the low values are reached about noon.
It was also obvious from field observations that activity of the
species re ach es its maximum in the first part of the night. Studying
the activities of Coenobita on Oostpunt in Curac;ao, during the
reproductive season 1967, places along the water wells and lagoons
were often found to abound with animals shortly after sunset,
whereas some hours later all had disappeared.
Besides the basical day- and night-rhythm it may be expected
that the relative humidity also affects the activity of Coenobita.
In most cases a sudden rise of humidity brings on an obvious
increase of activity. On April 26, a day with rainy weather and an
almost constantly high humidity, the greater part of the activity-
curve is shifted to a higher level. The same is seen on May 5 and
again in the nights of May 6 and 9. Sudden and often considerable
rises of humidity nearly always occur near sunset or as a result of
rainfall. In both cases a fall in temperature occurs. A simple linear
relation between activity and humidity, however, seems to be
absent. Compare e.g. a relatively small rise in humidity in the night
of May 8-9, effecting an enormous rise in activity with the effect
of a much larger rise in humidity in the night of May 4-5. Ap-
parently also other factors, perhaps the inclination or condition of
the animals itself, will play an important part in this matter.
The secondary character of the humidity in respect to the diurnal

Fig. 9. PATTERNS OF CIRCADIAN ACTIVITY (solid line) expressed in numbers of

active animals in relation to day and night rhythm and relative humidity of the
air in percentages (broken line) during a number oi days in April and May 1967.
- Periods oi rainiall (hatched columns in horizontal bar) generally induce sud den
rises in activity (thickened parts oi solid line).
 70

rhythm is weIl shown by the steady decrease of the activity during

the second part of the night, whereas on the contrary the relative
humidity in general reaches its highest values shortly before
sunnse.

Any direet influenee on the aetivity eaused by ehanges in temperature was not
tested and is left out of eonsideration. However, if such an influenee exists, the
effeet of temperature, as met with under natural eonditions in Curas;ao, will proba-
bly be small and the reverse of the effeet of the relative humidity.
Hermit erabs kept under laboratory eonditions at low temperatures, 15-18°C,
showed very little aetivity; even in eombination with high relative humidities. At
a temperature of 18°C, for a longer time, they beeame more or less lethargie. As
will be expeeted eombinations of high temperatures (28-32°C) and high humidities
(over 90%) indueed high aetivity; high temperatures and low humidity (10-20%
R.H.) minimal aetivity. Appropriate equipment enabling experiments at different
temperatures under eonstant eonditions as to the relative humidity was not availa-
ble.

Limestone as a source of water

It is weIl known that terrestrial crustaceans are often found In

the neighbourhood of caves, in particular in the limestone caves
along the north coast of Cura<;ao.
Investigations in December 1960 in a number of caves and
excavations in the steep escarpment of Rooi Rinc6n near Hato,
Cura<;ao, did indeed reveal higher humidities (up to 95%) of the
air, as compared with the much drier air outside (about 70% R.H.).
There was, however, no water to be found, except for the moisture
incorporated in the rocks. Of much more interest, however, were
the formations of a very soft and porous lime-sinter stone, occurring
in numerous places along the escarpment. Under conditions of
sufficient daylight the limestone is locally covered by thick, velvety
layers of green or blue algae, which doubtless indicate the high
content of moisture of the stone. Wet and dry (24 hours at 110°C)
sampies of limestone, which were collected on December 8, 1960
during dry weather conditions, indicated weight percentages of
water, present in the material, of at least 20-25%. As the attention
of the terrestrial crabs appeared now to be focussed on these sinter
formations, aseries of laboratory experiments and observations
 71

was started to learn more about the advantages of limestone for

hermit crabs.

Experiments and results

The experiments were carried out in two aquaria, each divided in two com-
partments by a double partition of smaIl-meshed iran gauze and moreover provided
with a grid of the same material, constructed about 3 cm over the bottom. In one
of the two compartments of each aquarium a good-sized piece of !imestone was
placed, which beforehand had been carefully cleared of algae, whereas in both
compartments D.-animals were housed ; three experimental animals with access
to limestone and three control animals without any access to that material, but
staying under exactly the same conditions as to the other factors such as tempera-
ture (28 D C) and humidity (90-100% R.H.). To prevent rapid desiccation of both
animals and limestone as weIl as to guarantee similar humidity conditions in both
compartments, a weIl fitting glass plate covered the whole aquarium . All the same
it could not be avoided that the water conte nt of the limes tone decreased gradually

UI
100 100
CI
t7I
o
1:
GI
~
8. 95 95
.E
,,
, -', ._-.
" 90
,,
".-. - ,
,
,,
" , .,
85 85

A
Z Z
0 I i , i i i i i i 01 i , i i i 'i i ,
0 5 10 15 0 5 10 15
doys doys
Fig. 10. LIMESTONE E X PERIMENT I. - Weight changes in percentages of 3 ex-
perimental animals (solid !ines) and 3 controls (broken !ines) . A : experiment with
soft porous limestone ; B: experiment with hard crystalline limestone.
 72

by desiccation to a value of 10-15% near the end of the experiment. Besides the
experiment with soft and porous limesinter another experiment was set up in
which hard and crystalline limestone with a water content of only 0.3-2% was
offered to land hermit crabs. Changes in gross weights in all animals were checked
by daily weighings. The first experiment lasted 14 days. Results are given in Fig.
10. Gross weights are converted into percentages of the net weight during the
first day of the experiment.
Weight decrease in the control animals is gradual and well comparable in both
aquaria. The graphs for the experimental animals with access to soft limestone
represented in Fig. 10 A, shows obvious changes; total decrease in weight as
compared to the controls is remarkably smaller and irregular, whereas distinct
gains in weight occur now and again. In Fig. 10 B (hard limestone) differences in
weight loss between experimental and control animals are shown to be relatively
small. One animal shows a somewhat higher loss.

From the experiment the preliminary conclusion may be drawn

that lime-sinter stone for some reason enables a better water
management. Two more experiments, each with ten animals
confirmed this conclusion. In these experiments the animals were
seen to pick up small particles from the stone, using the sm aller
cheliped, and to transfer them to the mouth. This was observed at
various occasions. In addition evidence was found that the lime-
stone particles passed, to a certain extent, the alimentary tract,
since the faeces of the experimental animals yielded ash contents
between 8S and 90%, while the faeces of the controls had ash
contents of some 50%.
In this connection the existence of a mechanism for water
uptake, which BLIss (1963, 1968) mentions for the terrestrial crab
Gecarcinus lateralis is interesting. Here the protruding pericardial
sacs appear to transport water from moist substratum towards the
branchial cavities, where it maintains a high relative humidity
of the air and moistens the gills through direct contact between
pericardial sacs and gills (GRUBER & SHOUP, 1969).
Convolutions of the membrane covering the pericardial sacs seem
to reduce the surface tension of waterdrops which are in contact
with the membrane. The water flows along the folds in the mem-
brane and the posterior edges of the carapace into the branchial
chambers. Abdominal setae along the posterior edge of the carapace
facilitate the water movement (BLIss, 1968). Uptake of moisture
from a damp substraturn by means of tufts of setae between the
 73

second and third pairs of walking legs is also described for several
species of Ocypoda (BLISS, 1968).

The shell as a protection against desiccation

Very young specimens of Coenobita clypeatus in the glaucothoe

stage, when still living in the ocean, were already seen to search
for a shell to protect the vulnerable abdomen (see also PROVENZANO,
1962). Immediately when going ashore the an im als have the
advantages of this sheHer: protection against predators, protection
against damage and chafing and, last but not least, protection
against desiccation.
Though in the case of land hermit crabs the value of the shell for
water conservation is obvious, as has been mentioned by a number
of authors, the extent of this protection has never been examined.
In the next experiment the benefits of a weH-fitting shell for Coenobita were
proved by comparing desiccation in adequately housed animals and in the ones
deprived from such housing. Twenty animals, freshly caught near Santa Catharina
were used after all traces of shell water had been removed. Ten animals with a
shell were compared with ten animals of which the shell had been removed by
carefully cracking it in a vice. All animals were placed separately in small aquaria
without access to food or water, exposed to temperatures of 28°C and a R.H. of
about 75%, both quite normal conditions in Cura~ao. Decrease in weight as found
from daily or more frequent measurements was considered to be brought about
mainly by loss of water. Loss of weight caused by defaecation proved to be negli-
gible. Every series of weighings ended with the death of the anima!. Finally, the
empty shells were weighed to calculate the weight of the anima!. For a proper
comparison of the results net weights were converted into percentages of the net
animal weight on the first day. Survival times proved to be extremely different for
the two groups: 3-18 days, with an average of 7.6 days for the sheltered animals,
and 1-2 days, with an average of about 30 hours, for the exposed ones. The rather
large variations in survival time are due to the initial conditions of the animals,
e.g. the uptake of water in time, the shape and size of the shell, the activity of the
animal during the experiment.
It appears from this experiment that for properly equipped animals the time of
survival is at least six times as long as for animals without shells. Moreover it will
be clear that, in case shell water is present, the survival times will extend to several
more days. To Coenobita clypeatus, living in a hot and rather dry climate, well-
fitting and reliable shells are a matter of life and death.
The same experiment revealed that, for the majority of the animals in question,
independent of their possession of a shell, desiccation of about 15-20% of the
original net weight is lethaI. Thus it is clear that survival in the first place is de-
 74
24
23

8
7

i i i i

8
5
rn
i
.§
61
4
3
2
~ 1
g. 0
11 12
days
 75

termined by the rate of desication, more so than by the time during which the
animals are exposed to desiccation. As will be seen later, desiccation is always
accompanied by an increase of the osmoconcentration of the body fluids and it is
likely that when a critical value is exceeded, there will be increasing difficulties
in the maintenance of vital biological processes.
Figure 11 represents the results of a second desiccation experiment, in which the
changes in net weight in thirty, for the greater part medium sized animals, were
measured under the same environm.ntal conditions of 28°C and approx. 75% R.H.
During the experiment, after I, 4, 7 and 10 days, so me animals, taken at random,
were sacrified for determination of the concentration of the body fluid. Two animals
still surviving on the 10th day were replenished with fresh water, after which
both the amount of retained water and the concentration of the body fluid were
measured. For results concerning concentrations, however, the reader is referred
to Chapter 5.
In an additional desiccation experiment with another 30 fairly large animals,
some were removed every day and their loss in weight, as compared to the initial
weight was determined. This was followed by determination of the dry weight
(24 hours, 100°C). Thus the average daily loss in weight could be expressed in
percentages of the total amount of free body water present in the animal (Fig. 12).

According to the experiments desiccation becomes lethal at

weight losses of about 15 per cent or more of the initial body
weight or of about 30 per cent of the initial body water. Environ-
mental conditions as provided in the experiment bring on death in
periods ranging from 5 to at least 10 to 12 days, with a rough
average of 8 days. As the period of desiccation itself, as weIl as
the loss in weight in percentage during this period are weIl compa-
rable in both the largest and smallest of the tested animals, the
actualloss of water in grammes per day must be sm aller in the laUer.
The regression in the upper graphs (Fig. 11) is larger than in the
lower graphs which represent animals of half the body weight of
the former animals. In general the loss in weight in each animal
is remarkably similar. During the first two days the evaporation
may be somewhat higher, to decrease slightly in the next days and
to increase again during the last few days before the animal expires.
As has already been described by other authors for a number of
terrestrial brachyuran crabs, the almost gradual and constant

Fig. 11. A. WEIGHT LOSS BY DESICCATION in 30 specimens of Coenobita. - Relative

air humidity about 75%; temperature about 28°C. 0: used for conductivety
determination of the blood; t: died of desiccation. On the 10th day of desiccation
two animals were revived with water. - B. MORTALITY BY DESICCATION.
 76

%
35

30

25

20

15

10 0

o 2 3 4 5 6 7 8 9 10 11 12 13
days
Fig. 12. W ATER LOSS in percentages of total initial body water. - 0: weight
determination in one specimen. t: died of desiccation. Dotted line showing influence
of traces of shell water. R.H. 75%; 28°C.

deerease in weight in desieeating hermit erabs points to a passive

way of dehydration : in other words to an evaporation similar to
that oi inanimate physical bodies (EDNEY, 1960).
 77

In their normallife, however, land hermit crabs are enclosed by

a more or less weIl-fitting sheIl, which may also contain a consider-
able amount of shell water, and therefore, though the principle
of a fully passive dehydration for the animal itself may be justly
applied, some complications in water loss by the animal in its shell
are to be expected. Dependent on individual differences in size and
shape of both the animal and its sheIl, the actual evaporation of
each animal, kept under constant environment al conditions, may
be expected to differ slightly, even when animals of equal net
weight and equal activity are compared. To eliminate such vari-
ations in water loss as induced by animal-shell interaction, finally
some desiccation experiments were carried out, in which the water
loss under constant environmental conditions, during a certain
period, was measured in several consecutive periods. In each series
of measurements precautions were made to stop desiccation in
time; then the animals were replenished with fresh water, after
which they were exposed to desiccation again, and so on.
In general the presence or absence of shell water will represent
a complication rat her difficult to deal with; it seems very likely
that the shell water will affect the pattern of desiccation. As soon
as the animals are touched, they react by a deep retraction into
the shell and in most cases this will cause a loss of water; the more
shell water there was present originally, the more water is wasted.
When, however, only moderate or sm all quantities of shell water
are present, reliable results are possible, if the animals are held
upside down during the procedure of weighing.
The great similarity in shape and gradient of the consecutive
desiccation curves derived from a certain animal under equal
conditions, is most striking (see Fig. 13). Dependent on environ-
mental conditions the gradient of the curves changes. Besides the
linearity abend, occurring in the first or second day of each period
of desiccation, is obvious. Most likely the latter represents the
disappearance of the last remains of the shell water; or it may be
caused by differences in the activity of the animal. From now on
the water loss proceeds markedly slower. Linearity and equal
regressions point to patterns of dehydration, weIl comparable to
the evaporation of inanimate objects.
 78

gross weight 9

45

44

43

42
o 50 100 150 200 hours
Fig. 13. EXAMPLE OF WATER LOSS by one Coenobita, in 5 consecutive periods of
desiccation of about 8 days each. - Graphs 1, 2 and 3 at 22°C and 75% R.H.;
graphs 4 and 5 at 28°C and 70% R.H.

DISCUSSION

Desiccation in semi-terrestrial crustaceans received fairly wide interest, cf.

EDNEY, 1960 and BLIss, 1968. Comparable information on other decapod crus-
taceans deals mainly with brachyuran species. For Gecarcinus lateralis, one of the
most purely terrestrial crabs, BLIss (1968, in Bimini, Bahamas, tempo 30°C, R.H.
78%) found survival periods of 89 hours on the average; 21 % of the original weight
was lost by the time of death. Similar data for Cardisoma guanhumi are 53 hours
and 16% weight loss and for Ocypode quadrata 20 hours and 14% weight loss. Thus
more terrestrial species can withstand a greater desiccation as compared to semi-
terrestrial species, whereas the rate of water loss is lower. BLIss also found that all
sizes of Gecarcinus lateralis are equally resistant to dehydration. In the other two
species the smaller individuals proved to be more vulnerable. Evaporation of the
exoskeleton in both terrestrial and semi-terrestrial crustaceans is highly influenced
by temperature, humidity and wind (EDNEY, 1960). Waxlike layers, such as occur
 79

in insects and other land arthropods, have never been demonstrated in crustaceans,
and water loss in terrestrial crustaceans always tends to be much greater as com-
pared to insects.
For an appropriate discussion of the water loss in land hermit crabs, in the first
place the biological observations and morphological facts observed by HARMS
(1929, 1932) in Birgus latro and the Indo-Pacific species Coenobita clupeatus (= cf.
Coenobita brevimanus Dana) and C. cavipes Stirnpson, and in the second place, the
more ecologically directed studies of VÖLKER (1965) on Coenobita scaevola are
important.
HARMS (1929) emphasizes the great advantages for animals living in a sheIl, as
regards protection against desiccation. The major chela and to a lesser extent the
minor chela and partly also the flattened ambulatories together form a shield,
similar to an operculum, which in weIl-fitting shells serves as an extremely useful
elosing mechanism (see for details on this subject VÖLKER, 1965). Those parts in
the active animal protruding further outside the shell show increasingly thicker
teguments and heavier chitinization. HARMS (1932) states that the importance of
the abdominal respiration in Coenobita exceeds the function of the (modified) gills.
The abdominal cuticles are very thin, showing, in particular along the dorsal side,
invaginations resembling tracheae, and they appear to be moistened by special
glands.
VÖLKER (1965) determined water loss and survival times in Coenobita scaevola
under constant conditions in the laboratory. Three animals kept at 25°C and 75%
RH. survived for periods of about 9 days and on the average lost 18% of their
original weight. At conditions of 25°C and 40% R.H. water loss proved to be 2!
times higher. Survival times in two experiments, carried out at conditions of 29°C
and 40 or 55% R.H. proved to be shorter (up to two days) in small animals (2-6 g
net weight) as compared with periods ranging from 5-6 days in larger animals
(8-12 g). The animals succumbed as weight los ses by desiccation exceeded 20-30%
of the original net weights. Moreover, VÖLKER calculated for his animals water
loss in milligrammes per hour per gram animal weight and noticed, notwithstanding
constant environmental conditions, that dehydration occurred irregulary. After
initial high water loss an obvious slow-down was observed, which, however, did
not tend to any constant value but went up and down alternatingly in a daily
rhythm or with longer interva1s. Finally, just before death, a final water 10ss of a
greater volume occurred.
An explanation for these observations is not given. VÖLKER coneludes that
Coenobita scaevola does not possess any morphological structures to reduce dehy-
dration. The only remaining possibilities are to elose the shell and to become
inactive.

From VÖLKER'S experiments in Coenobita scaevola and our

observations in Coenobita clypeatus it appears that the periods of
survival may show rat her large differences depending on: a) the
strength of desiccation, b) the size and physiological conditions of
the animals, c) the ability to dose the shell adequately, d) the
degree of activity of the an im als during the experiment and e) the
availability of shell water. Survival periods of one week or more are
 80

quite normal in both species of terrestrial hermits. In general these

periods proved to be considerably longer than those found by
BLISS (1968) for the genuine land crab Gecarcinus lateralis (e.g. 4
days at 30°C and 78% R.H.). Since Coenobitidae are able to stand
periods of drought without any visible uptake of water, their
success as land animals is apparent. Considering the dehydration
which the species can stand during the desiccation, however,
Coenobita clypeatus seems to take only a central place in the range
of semi-terrestrial and terrestrial species. The extremely high
values in Coenobita scaevola are remarkable (Table 12).

T ABLE 12
W ATER LOSS OF 'TERRESTRIAL' DECAPODS

expressed in percentages of original (net) weight at death.

species % author

Coenobita scaevola 25-30 VÖLKER, 1965

Gecarcinus lateralis 21 BLISS, 1968
Coenobita clypeatus 15-20 DE WILDE, this paper
Cardisoma guanhumi 16 BLISS, 1968
Ocypode quadrata 14 BLISS, 1968

Thus, when comparing Coenobita clypeatus and Gecarcinus

lateralis, two of the most purely terrestrial West-Indian species,
which very often occur in exact1y the same habitats, the lethai
desiccation appears to be be more or less similar. On the other hand
the survival periods occurring during drought are twice as high
in Coenobita as in Gecarcinus. It will be evident that the latter
difference is mainly due to the adoption of a shell.
Another point worth discussing is the passive and uncontrolled
water loss by semi-terrestrial and terrestrial crustaceans and by
terrestrial hermit crabs in particular. If the teguments of these
animals are indeed freely permeable to water and no regulating
mechanism is present, the water loss may be bett er described as
evaporation, rather than by the generally used term of tran-
spiration, which suggest some form of an active process. Heavier
chitinizations or calcification in terrestrial crabs may produce a
 81

decreased dehydration, but they do not provide a mechanism of

regulation. Description of water loss in crustaceans as a pure
physical process (EDNEY, 1960), observations by HERREID (cf.
BLIss, 1968, p. 369) in dead (semi-)terrestrial crabs in which water
loss continued at 75-95% of the rate at which they lost water
when alive and the linearity of the desiccation curves in Coenobita
clypeatus, all point to an evaporation process. Yet it is necessary
to consider the quest ion more closely. Although water is lost over
the entire exoskeleton there is no reason to suspect that in semi-
terrestrial or terrestrial crabs the respiratory membran es are
particularly susceptible to dehydration (BLIss, 1968). To enable a
proper functioning the membranes must be thin-walled in structure
and moreover they ought to be kept moist. HARMS (1929) attributes
the moistening of the respiratory membran es in Coenobita to
special complex glands, which seem to be well-developed in the
walls of the gill chambers, the head region and in certain ap-
pendages. In Birgus the secretion products of the "lungs" mix
easily with water, forming a mucous fluid, which does not easily
evaporate (HARMS, 1932). Microscopic slides of the abdominal
tegument of Coenobita rugosus showed sub-tegumental glands,
emptying into the transverse chitinized folds (HARMS, 1932).
Although HARMS does not mention any function of these glands,
a function concerning water secretion seems most likely. Assuming
that HARMS' observations are right, it is weIl possible that there
is some regulation in the loss of water induced by the activity of
the glands.
Under constant environment al conditions weight loss by eva-
poration in a physical body is proportional to the surface of that
body and further to the moisture content of that surface and the
nature of the evaporating fluid. In living individuals the main
source of evaporation - the stretched abdominal skin - shrinks in
a very distinct way during desiccation; the often tightly stretched
abdominal bladders fade completely away. Moreover the abdomen
itself becomes drier, both at sight as weIl as to the touch, and
finally the evaporating fluid continuously increases in concen-
tration.
Thus a gradual decrease in the rate of water loss in desiccating
 82

hermits is to be expected. Although some of the curves tend to

concavity, such a general decrease in water loss was not proved
for Coenobita clypeatus, nor for C. scaevola in the study of VÖLKER.
This leads to the conelusion that the rate of water loss is mainly
determined by the out er barrier, e.g. the shell and the "operculum."
Another anomaly in the desiccation pattern of Coenobita is the
frequently observed rise in water loss some days before the death of
the animals (Fig. 11, A). The same was observed by VÖLKER (1965).
The suggestion is that the animals lose their ability to elose the
shell adequately. In addition there is a reason to suspect that
the alternating periods of higher and lower water loss in C. scaevola
are caused by periods of higher and lower activities. The effectivity
of the elosing mechanism is absent in periods of high activity.

Summarizing dehydration and patterns of water loss in terrestrial

hermit crabs we may conelude that the shell ranks first in im-
portance. Initial higher loss is due to the presence of shell water.
Hereafter a remarkably constant water loss, mainly determined by
the outer barrier (shell and operculum) occurs. A gradual decrease
in water loss as is suspected by the shrinkage of the main eva po-
rating surface, the abdomen, is not elear. Water secretion or
secretion of mucus seems to be a possibility, but any special evi-
dence for this assumption was not found. Variations in activity
play an important role. Increasing water losses before the death
of the animals are due to a less efficient elosure of the shell.
 IV. SHELL WATER AND ITS REGULATION

INTRODUCTION

It is a well-known fact that littoral and semi-terrestrial crustaceans, when staying

ashore for a shorter or longer time, often carry some external water, usually retained
in the branchial cavities (WOLVEKAMP & WATERMAN, 1960; EDNEY, 1960). The
main purpose seems to be to moisten the respiratory surfaces and thus to enable
breathing. Species such as Grapsus, Ocypode and Uca, as weil as sometimes more
aquatic species such as Eriocheir, vigorously aerate the water present in the bran-
chial cavities. In more terrestrial species such as Gecarcinus, Cardisoma and Coeno-
bita branchial water is not an absolute necessity; more or less moist respiratory
surfaces seem to be sufficient.
The role of water in the branchial cavity in osmotic regulation, as discussed by
VERWEY (1957) is probably of little significance.
In land hermit crabs of the genus Coenobita the dermal respiration of the soft
and not-calcified skin of the abdomen has become very important and for this
reason this part of the body must also be kept moist.
Land hermit crabs often store considerable amounts of water in the shells in
which they live. Following GROSS (1964) this water is called shell water.
The occurrence of shell water in the Pacific species Coenobita perlatus (Edwards)
was already noticed much earlier by SEURAT (1904), who connected its function
with moistening of the gills. HARMS (1932) and VÖLKER (1965, 1967) mention
shell water in two more species of Coenobita, pointing to the importance of keeping
the abdomen moist.
In the case that rather large volumes of shell water are stored, which facilitate
respiration as weil as general water economy, enabling the animals to survive in
dry inland habitats, its role in osmoregulation cannot be neglected. As far as known
there is no information about the permeability for salt and water of the abdominal
skin in land hermit crabs and although nothing is known concerning the osmore-
gulatory abilities of this part of the body wall either, one can imagine that such
extremely thin teguments as described by HARMS (1932), continously bathed by
shell water, will have some consequences to this effect. Shell water of a more or less
constant and adequate salinity would be extremely valuable to the animals, but
on the other hand this would imply some kind of mechanism to regulate the salinity
of shell water.
 84

Except for some observations by GROSS (1964) on the Pacific

species Coenobita perlatus and C. brevimanus no attempts to study
the concentration of shell water are known.
As recorded here, measurements of the concentration of shell
water of Coenobita clypeatus, both in the laboratory and in the
field in Cura<;ao, yielded salinities of a remarkable constancy,
pointing to the existence of some mechanism of regulation. A
number of factors of biological, physical and climatological nature,
which were assumed to influence the stability of shell water,
received special attention.

UPTAKE AND STORAGE OF SHELL WATER

Two different methods for taking up shell water may be distin-

guished, viz.: passive and active uptake. They depend on the
hermit crab's way of life and may very weIl be characteristic for the
species.
I. Passive uptake: The shell water enters passively into the shell. GROSS (1964)
observed Coenobita perlatus H. Milne Edwards in Eniwetok atol, entering sea- and
brackish water at night, which resulted in the adopted shells being filled with water.
According to SE URAT (1904), who is also quoted here by WIENS (1962), Coenobita
perlatus in Tahiti goes to the Lagoon shares every evening at sunset "to let the
waves wet them and renew their snpply of water." A similar behaviour was observed
by VÖLKER (1965) for Coenobita scaevola Forskäl in the Red Sea area. This species
approached the waterline, so that it was wetted now and again by the waves. The
animal squatted here for some time with the entrance of the shell facing the sea.
After so me minutes, when the shell was filled, the animals returned.

z. Active uptake: The active way of uptake is related with normal drinking, with
the help of the chelae as described in Chapter III for Coenobita clypeatus.

Desiccated animals living in transparent glass shells (PI. Vlb)

in the laboratory were placed on a grid of sm all meshed wire netting,
some millimeters over a drinking bowl which was filled with red
water, coloured by a tasteless dye. The animal could touch the
water only with the sm aller right-hand chela, so that every possible
way of passive uptake by capillary rising was excluded. It was very
clear that the glass blown shells were filled, using the active way
of water uptake.
If an animal wants to store water in its sheIl, soon after the water
 85

is inspected and brought up to the mouth, red-coloured water is

seen to flow into the branchial cavities. This observation is possible
because the pleurae of the carapace are thin and slightly tran-
parent. Soon afterwards the animal raises the posterior margins of
the carapace and the water runs backwards by capillary action and
also helped by vigorous pulsating movements of the abdomen
itself, filling the space between the inner wall of the shell and the
abdomen and also the empty space in the shell behind the animal.
Although the shells are sometimes full of shell water the animals
walk ab out with the water without spilling any. In hermits in
transparent shells it was observed that the water is not only held
by capillary action but, also, that the skin of the anterior part of
the abdomen is pressed firmly against the inner wall of the shell to
prevent spilling. When, however, the animals are disturbed and
have to withdraw deep into their shells a large portion of water is
lost.
Concerning the water stored inside the animal's body attention
has to be given to two bladder-like folds on the postero-ventral
side of the abdomen. When drinking water of the right concen-
tration abounds, these bladders become transparent after some
time, bulging with body fluid.
To obtain information about the amount of retained water in
eight Coenobitas the exact quantity was measured afterwards.
For this purpose D.-animals were used, which had no access to
water for 13 days previously and lived in prepared Livona-shells.
These shells were sawn through longitudinally and fixed together
with paraffin wax so that they could easily be opened at any time.
The dry weight of the prepared shells (24 hours in exsiccator) was
known. The gross weight of the desiccated animal was measured
before bringing it to the water; afterwards the gross weight (now
including the shell water) was measured again, and also the net
weight of the carefully wiped hermit. Thus it was easy to find out
which part of the water was taken up by the animal itself and which
part was stored as shell water.
Figure 14 gives the results of the experiment. Net weights of the
experimental animals in the last three days of aperiod of desiccation
are shown. Fresh water is supplied on the 13th day. The gain in
 86

net animal weighf

g sfored wafer
60
58

-.
56
54
......................
52

36
34

32

30

22
. . -.. -~311.0
20

--------.- 0 I. 0 0
I.
18 ,_. 2 __

16 ------···_._l/I 2----------
14
---.---..--./.h: ;11-.0
12
10
'--'--
--"---::-:::::':::::::.1 //2 1
D1_.0
=-____: _____

i r - -....
i _ _i....
I
~

10 11 12 13 14 days
Fig. 14. UPTAKE OF FRESH WATER AND ITS STORAGE by 8 specimens of Coenobita
after a 13 day's period of desiccation. - Dark columns: water stored in animal;
white columns: shell water.

weight on the next day indicates the total of the water uptake for
each animal, whereas the black and white columns represent the
water penetrated into the body, respectively into the shell.
It is clear that considerable amounts of water are taken up here.
At a later stage it was observed that in some cases the amounts of
retained water might be much greater and that the amount of shell
water alone could amount to a third or even half of the net body
weight. Evidently the concentration of the offered water and the
 87

condition of the animals determine whether shell water will be

retained and if so in which quantities.
An experiment as described above in which 100% S.W. or water
with a somewhat higher concentration was offered also showed
uptake of water, but now only in very small amounts. In these
cases shell water is often not retained at all. Under natural con-
ditions in the field it was also obvious, that animals depending on
sea water store much smaller amounts of shell water than inland
animals with access to fresh water.
From the observed differences in the stored quantities of shell
water, depending on access either to fresh- or to sea water, it is
obvious that Coenobita clypeatus discriminates between water of
different salinities and moreover that water of a random salinity
may be unfit to be used as shell water.
When, however, the concentrations of shell water were measured
in a number of land hermits, which had only had access to fresh
water, it was observed that in most cases the concentration of the
shell water was much higher than that of the available drinking
water.
This led to aseries of laboratory experiments on the relations
between drinking water and shell water. Essential in all these
experiments was that water of various salt concentrations was
offered to the animals, in such a way, that the volume of absorbed
water could be exactly registered. For this purpose drinking
fountains, as used for cage birds, were applied. In modified fountains
with long tubular reservoirs, which were calibrated, water con-
sumptions of 0.1 ml could be read off easily. Unless otherwise
specified all experiments were performed under normal indoor
conditions for Cura<;ao with temperatures between 26 and 32°C
and a relative humidity between 70 and 80%. During the day the
room was kept in twilight.
 88

EXPERIMENTS

Uptake of water of a fixed concentration

Method

In a first experiment the total water consumption in a number of animals was

measured when water of different salt concentrations was offered. The experimental
set-up is shown in Fig. 15.

Fig. 15. EXPERIMENTAL SET-UP FOR MEASURING CONSUMPTION OF WATER of different

salt concentrations. - The long calibrated reservoirs are fixed with plasticine on a
plastic drinking trough. The first tank on the left does not contain any animals and
is used for determination of normal evaporation.

In 7 glass tanks long tubular drinking fountains were put, holding water of
respectively 0, 0, 10, 50, 100, 125 and 150% S.W. When the experiment started
four A.D.-animals of equal size were placed in each tank, with the exception of
tank nr I, which was used for measuring the normal evaporation of the cisterns.
The initial blood concentration of these A.D.-animals was known from determi-
nations in other, completely comparable, animals. Every day the water con-
sumption in the various tubes was checked by reading off the water levels; the
reservoirs were replenished if necessary.
Except for 2 animals in tank II, with 0% S.W. and for the 4 animals in tank VII,
 89

with 150% S.W., all other animals survived for 50 days, the duration of the ex-
periment.
Concentration of both shell water (by conductivity) and blood (by conductivity
and freezing point determination) was measured at the end of the experiment.

Results and interpretation

For 5 successive periods of ten days each, the water consumption, expressed in
ml per animal, is calculated. A correction for the evaporation in the drinking
fountains has already been made.
Looking at Fig. 16, first period, the hermits in all tanks have a rather moderate
consumption; nevertheless a distinct preference for 100% S.W. is shown, whereas
distilled water seems to be the least suitable. During the second and following
periods the absorbed volumes of 0, 10 and 150% S.W. still remain small, but now
the uptake of 50% S.W. has increased markedly, whereas uptake of 100 and 125%
S.W. has increased to extraordinarily high levels.
At first sight these results are in flat contradiction with the previous observations
in which was stated that animals, having access to fresh water, always store large
amounts of shell water, while on the other hand, shell water is scarce or absent in
animals which drink 100% S.W. or water of a higher salinity. However, considering
the behaviour of the animals, the observed differences are easily comprehensible.
A.D.-animals encounter two major problems, which are the adjustment of the low
concentration of the body fluids and the filling-up of the unavoidable loss of water
by evaporation.
The data of the first period show that the animals in tank II only take very
small amounts to avoid further desiccation but do not store any quantity of shell
water, which would bring about further ion loss. Doubtlessly these animals are
in very difficult circumstances, which is also demonstrated by the death of an
animal in the first period and another in the second.
The same holds true more or less for tanks III and IV. As will be demonstrated
below, the concentration of 100% S.W. is only slightly higher than the optimal
concentration of the shell water and therefore animals in tank V are not only able
to bring up the salt concentration of their body fluids to a normal level, but will
also store certain amounts as shell water.
Animals in tanks VI and VII take so me water to prevent further desiccation and
adjust the concentration of the body fluids at the same time; uptake of shell water,
however, does not take place, for the concentration of the available drinking water
is far higher than the optimal concentration of the shell water.
In the second period the overall picture of the water uptake has changed. For
the animals in tanks II and III the situation seems to be similar; at 0% S.W.
another hermit crab has died, whereas the amounts of absorbed water are still
very small.
In spite of the small amounts here and also in tank IV salt accumulation takes
place caused by a repeated uptake and evaporation of small quantities of water
of low salinity. Of course this will produce the smallest effect and will take the
longest time in the case of 0% S.W. It was often observed how in these circum-
stances, Coenobitas stimulate a good evaporation by going abnormally far out
 90

ml
ml Period I

-"
9-19 March 1966

1
40
20
0 150 •
mr Period 2
250
19-29MarchI966
100

- .-------
200
ml
350 50

150
300 0
\
\
100 \
Period 3 \
29March-8April 1966 \
250 \
\
50 \

~" \
\

-" 1-
200

0
I
150 \\
\
\
\
100 I
\
ml Periad 4 I
I
8 -18 April 1966 I
\
50 200 I
\
I

o 150
't
\
\
\
\
\
100 Period 5 \
18-28 April 1966 \
\\
50 \
~. \
\
\
•
o i--,.-----..--------,..----.,....-----;.'t S.W.
001. 10% 50% 100% 125% 15001.
Tube n m m 1l":m E1
Fig. 16. CORRECTED USE OF WATER PER ANIMAL in 10 days, in ml. - Offered are
concentrations of 0, 10, 50, 100, 125 and 150% sea water. In period 4 the con-
sumption of 125% S.W. is more than 150 times that of 0% S.W. Columns drawn
only in this period. 150% S.W. is not suitable as drinking water.
 91

of the shell and sitting with legs extended on a well exposed spot, e.g. against the
wire netting covering the tanks.
Considering the increased uptake of water of the animals in tank IV it seems
most likely that these animals already have the right kind of shell water at their
disposal.
Animals in tanks V and VI with respectively 100% S.W. and 125% S.W. behave
quite differently. Such animals are also confronted with the steady increase of
the concentration of the body fluids ,p,nd - if present - of the shell water. When the
concentrations of the body fluid or the shell water surpass the permissible levels,
the animals start drinking again. From now on, however, they do not retain any
quantity of this water, which would only add more salt to the animals, but they
let it circulate through the gill cavities and the shell, while bathing the abdomen.
After a shorl time the water flows out of the shell and the animal starts "drinking"
again. In this way they spend many hours a day around the water fountains,
pushing each other away, in order to replenish the old shell water and to keep the
salt concentration of the body fluids as low as possible. The used volumes of water
as read on the calibrated water reservoirs proved to be unexpectedly high.
Water of 150% S.W., as offered in tank VII is too salt to be used in a circulation
as described. Only very small quantities of this water are used. One animal from
tank VII did not survive the second period.

The strikingly different behaviour observed in animals which

have access to drinking water of low salinity at one side and high
"alinities at the other, was also observed in the three other periods.
Except for tank VII, in which the remaining animals all died
during the third period indicating the total unsuitability of 150%
S.W., no further mortality occurred. Obviously the animals of
tank II had now passed the critical period.
At the end of the fifth period the remaining 18 animals were
checked as to the presence of shell water and its concentration was
measured. Both animals of tank II had very sm all quantities of a
moderate salinity; animals in tanks III and IV proved to have
large amounts of shell water of normal concentration. Although
actually shell water was absent in the animals in tanks V and VI,
in a number of cases it was possible to determine the concentration
of the circulating water dripping out of the shells. At 100% S.W.
the same concentrations as in the available drinking water was
measured; at 125% S.W. the concentration of shell water was
considerably higher. In this way the remarkable S-shaped regu-
lation curve (Fig. 17) of shell water against available drinking
water is found. Data on blood concentrations are incorporated in
Fig.32.
 92

equiv.S%o
60
shellwafer
: drinking wafer
55

50

45

40

35

30

25

20

15 I
10

0
o 10 50 100 125 150
drinking wafer, %SW

Fig. 17. SALINTY OF SHELL water in relation to the concentration of the available
drinking water. (100% S.W. = 36%0 S.).

Without entering into details two important principles playing a

role in the regulation of the salt concentration of shell water in
Coenobita clypeatus are introduced here:
I) Accumulation of salts by repeated uptake and subsequent
evaporation of sm all amounts of drinking water of low concen-
trations.
2) Circulation of large amounts of water in animals having access
only to water of a relatively high concentration.

The results of the experiment mentioned above reflect the main

 93

difference between coastal and inland animals. The amount of

absorbed water and the time needed for its uptake in coastal
animals are much more than those on the inland animals, which
use fresh water.
On ac count of their unsatiable need of circulating water, coastal
animals are compelled to stay in the coastal regions.

Uptake of watt'[ In the drinking apparatus

Method

A second series of experiments describes hermits which are placed in the drinking
apparatus. In this apparatus the animals can choose between sea water of different
concentrations; from the absorbed quantities it is possible to conclude which
concentration is the most attractive for Coenobita.
The apparatus consisted of a wooden standard, around which 10drinking foun-
tains, provided with long and slender reservoirs, were placed in a circle. The standard
was enclosed by a cup board with a door in front and a perforated bottom and top
to permit a good ventilation. By revolving the standard the position of the fountains
and reservoirs, which were fastened on the standard by means of tool-clips, could
be changed; this in order to eliminate a preference for a particular direction and
place. The experiments were performed under the same indoor conditions as
described for experiment 1. The fountains were filled with 0, 2, 10, 25, 50, 75, 90,
100, 110 and 125% S.W. Thirty A.D.-animals were released in the cupboard. The
experiment lasted for 40 days, or for 8 periods each of 5 days. The water available
to the animals was constantly renewed, but measurements were only made in the
Ist, 2nd, 4th, 6th and 8th periods. The experiment has been repeated a second time
under the same conditions, with fresh animals.

Results and interpretation

The picture of the water consumption in both experiments is roughly the same:
During the first period the A.D.-animals, which have a serious salt-deficiency,
do not greatly discriminate in the uptake of water. Different concentrations in a
fairly wide range are absorbed; however, moderate concentrations of 25, 50 and
75% S.W. seem to be most favoured, but a distinct preference is still lacking.
Concentrations over 100% S.W. are seen to be less favoured. (The low amounts
of 10% S.W. in particular when compared with the used quantities of and 2%
S.W. are difficult to explain).
°
During the second period a more distinct preference for 50% S.W. is obvious.
In the next period a gradual shift to the use of ever fresher water is shown; finally
resulting in a distinct preference for water of the lowest concentrations. In Figure
 94

ml

150

n
100 , : 071.92)
~/1"
: .-.
i""-.... :m:
. ~l
, 056.70)
:__ ~:__ .___ .

50
E
. ~~
::
058.27)
__ l __•___•
'\' I ~ 1 I i

.--.--.----- -'~j i
1ZlII
031.57)
o I I i
_.--.-.
i i i i
02 10 25 50 75 90 100 110 125 %SW.
, i i
01 3.6 9.0 18.0 27.0 32.4 36.039.6 45.0 sal.%o
0.72
Fig. 18. USE OF WATER in ml per period of 5 days by 30 animals in drinking appa-
ratus with 10 fountains filled resp. with 0,2, 10,25,50,75,90, 100, 1 IO and 125% sea
water. - Dotted lines: concentrations of normal shell water and normal sea water;
between parentheses: total amounts of water absorbed in ml per period.

18, giving the results of one of the experiments, the consumption of water is plotted
for periods of 5 days.
A check on the presence and concentration of shell water, carried out at the end
of the experiment, indicated that 28 out of 30 animals had considerable amounts
of water in their shells. The abdomens and insides of the shells of the two remaining
 95

animals proved to be only slightly moist. In 26 cases the concentration of shell

water could be measured; see Fig. 19.
Though the animals unmistakably prefer drinking water of low concentrations,
the concentration of their shell water is still maintained at much higher values,
which correspond to 75-100% S.W.

number

o
Fig. 19. DISTRIBUTION OF SALINITY OF TRE SRELL WATER of 26 animals from
experiment 2. - Broken line: 100% sea water.

Thus Coenobita takes up water of low concentrations in order to

keep the concentration of the retained shell water on a more or less
constant and much higher level. In other words: the shell water,
which by evaporation grows salter continuously, is kept on a
correct level by dilution with fresh water.
It is assumed that as soon an A.D.-animal is released in the
drinking apparatus it will start to adjust its salt balance, either by
taking up water of the right concentration or by a combined
absorption of water of high and low salinity in the right pro-
portions. In addition an extra supply of water of the right concen-
trat ion will be retained in the form of shell water. Since its concen-
tration will now increase by evaporation, the animal takes up a
certain amount of fresh water for dilution. The fresher the absorbed
water, the smaller the volume which will be necessary to obtain
the desired effect. The uptake of water with higher concentrations
(over 25% S.W.) shows a very distinct decrease during the suc-
cessive periods, which on the other hand is compensated by the
increased uptake of water of 10, 2, and also 0% S.W.
Moreover it is seen how during successive periods, the total
amounts of absorbed water decrease from about 172 ml in the first
period to 131.5 ml at the end of the experiment.
 96

Considering the results of the previous experiments, it is inter-

esting to speculate upon the process of passing from a marine to a
terrestrial environment by coastal animals. Generally speaking any
establishment in a terrestrial habitat will involve a passage from
salter to fresher water and moreover this water will not be as
easily available as sea water iso Such animals have not only lost
the necessity to have sea water circulating continuously, but, as
the water becomes less salty, they can also manage with ever
sm aller amounts of drinking water. In this way they become less
dependent on water, which implies a high er degree of terrestrial
adjustment.

R e g u 1a t ion b y mix i n gof 0 an d 100% s e a w a t e r

Introduction and method

After the existence of various ways of regulation of shell water

in Coenobita had been recognized, the next step was to investigate
the in- or external factors, which might influence the process of
regulation, such as the actual limits between which the concen-
tration of shell water is kept, and wh ether the existence of such an
optimal concentration could be proved.
. To this end it was assumed that a simple device as shown in
Fig.20, inset, representing a tank with two drinking fountains,
filled with 0 and 100% S.W. respectively, would be sufficient to
offer to the experimental animals a favourable opportunity to
compose and to retain shell water of a desired concentration.
After shaking out all shell water, normal animals from the
storage box were placed in groups of ten, under conditions of 28°C
and about 75% R.H., in the testing tank. Four or five days later
the shell water was shaken out again and tested. A hundred and
ten out of 120 tested animals had retained enough shell water to
permit determination of the electrical conductivity.
A diagram showing the distribution of this shell water concen-
tration, converted into salinities, is given in Fig. 20.
Although theoretically the salinity of the determined shell water
 97

number
,
30 I
I
I
I
~
u)
'$.
o
20 Q

10

o
10 20 30 36
SO/00

Fig.20. SALINITY RANGE OF SHELL WATER in specimens of Coenobita which had

access to distilled water (0% sea water) and normal (100%) sea water, under
laboratory conditions. - Broken line: 100% S.W. Inset: experimental set-up.

concentrations rnight range frorn 0%0 S (dist. water) to over 36%0 S

(normal sea water), the lowest value found here is a salinity of
10%0; the highest 35%0. A very distinct peak occurs at a salinity
of 30-31%0 S. All salinities are less than 100% S.W.
Since aperiod of 4 or 5 days is much too short for building-up
shell water concentrations as discussed in experiment 1 by the
uptake and evaporation of low-concentrated water only, we have to
conclude here again that in the first place the regulation consists of
a simple mixing of water of different concentrations in the right pro-
portions. A possible release or exchange of salts or ions by the animal
is left out of consideration, though it very likely takes place.
 98

Differences between small and large animals

Two additional experiments, fully identical to the one described in the last
section, were performed with 2 groups of small- and of large-sized animals re-
spectively (gross weights of the small animals ranging from 10-20 g; large ones
ranging from 50-100 g) . The intention was to investigate whether the size of
Coenobita could be correlated with the concentration of shell water retained by the
animals. Such a correlation was found indeed, as could be concluded from the
results of two series of experiments in which shell water of a total of 120 specimens
was tested. Small animals build up shell water of a somewhat higher concentration
than larger animals do in the same period (Fig.21).

number

8
6 small

4
2
0
25 30 35I 40
6 I
I
large I
4 I
I
I
2
0
?P 25 30 :35 40 %. salinity

number
4
2

I
0

4L
25 30 35 40 45

large i
; ~ ~ ~ _ _r-f""I"'tj---f"'~~~--
20 25 30 35 40 %osalinity

Fig.21. DrFFERENCES IN DISTRIBUTION OF THE CONCENTRATION OF RETAINED SHELL

WATER comparing small and large specimens of Coenobita. - Large individuals
show distinctly lower values. Broken line: 100% sea water.
 99

Mixing of other concentrations

°
As in the previous experiment, normal animals were allowed to choose between
two concentrations and 100% S.W., after all the shell water had been removed.
In two more tanks combinations of 0% and 150% S.W., and 0% and 200% S.W.
were offered to normal animals under the same conditions. After aperiod of 5
days the shell water of all animals was shaken out and measured. In addition an

numbers ,
I.
o and 100% S.w. ~
10 - contral-

0
15 20 45
""0 end 100% S.w.

0
15

0
15 45
o end 200% S.w.

0
15 20 25 30 35 40 45
saL%o

° ° °
Fig.22a. EQUAL DISTRIBUTION OF SHE LL WATER CONCENTRATION in 8 groups of
Coenobita having access to resp. combinations of and 100, and 150, and and
200% S.W. - Contro! represents the shell water of animals from the storage box.
- Measurements in first 4 groups after intervals of 5 days .
 100

equal number of animals was kept back in the storage-box for control, where they
also had access to both 0% and 100% S.W. The shell water of the latter animals
was also examined at the end of the experiment.
As represented in Fig. 22a, the results in all cases show well-comparable frequency
distributions of the salinity of shell water. In the same way as shell water is com-
posed from 0% and 100% S.W., Coenobita is able to mix 0% and 150% S.W. or
0% and 200% S.W. in the right proportions to obtain the right kind of shell water.
The most frequently occurring values for the salinity are those of about 32%0 S,
which is distinctly below the salinity of 100% S.W. (S = 36%0)' but somewhat
higher than the values of 30-31 %0 S, found for the shell water concentration in the
previous experiments.
Moreover the distribution curves derived from the animals in the testing tanks
show a somewhat larger spread and are more negatively skewed, as compared to
the curve of shell water of the animals in the storage box.
The same experiment was repeated once more. The shell water of the animals
in the three testing tanks, however, was now measured after intervals of one day.
Moreover, shell water of an equal number of animals, kept in the storage box, was
measured again for comparison. So the shell water of this last group, which had
been kept in the box for at least a week was measured after a considerably Ion ger
interval.

Considering the results (Fig. 22b), the check (upper histogram)

shows that the most frequently occurring shell water concen-
trat ions are those with salinities of 34-35%0' which is higher than
the most frequently occurring salinities of the three lower histo-
grams, which are only 31-32%0 S. Furthermore there is also a
larger spread.
On the one hand the results confirm the previous findings, since
an equally successful composition of shell water takes place if
water of a different concentration in combination with fresh water
is available; on the other hand, the factor time also seems to
influence the degree of stability attained in the shell water. Over a
longer time the composition and regulation of shell water seems to
become more stabilized.
From the frequency distributions of shell water in Coenobita
derived from the experiments, it may be seen that, although the
figures show distinct maxima which are lower than the concen-
trat ion of 100% S.W., the place of the maxima is somewhat varia-
ble, ranging from salinities of ab out 30-31 %0 up to salinities of
about 35%0 S. The same holds true for the total spread and conse-
quently for the minimum and maximum values.
Assuming that there is only one concentration which represents
 101

numbers
10
o and100% sw.
-conlrol -

0
15 20 25 30 35: 40 45
,
I
I
I
I
5 o ond 100% Sw. I
I

0
45

5 o and 150% Sw.

0
40 45

5 o ond 200% S,w.

0 I
15 20 25 30 35 40
SOI%~

Fig.22b. Same as Fig. 22a. - Measurements in second 4 groups after intervals of

1 day, except the contro!. Broken line: concentration of 100% S.W.

the optimal condition, the observed deviations in the found shell

water concentrations, are caused by imperfections in the way of
regulation itself, e.g. the attainment of the optimal concentration
will take more time than the experiment permits. I t is also assumed
that the optimal concentration may change, depending on certain
in- or extern al factors, such as the size of the animals, temperature,
relative humidity, etc., or the combination of factors such as
temperature and humidity.
 102

Time as a factor influencing regulation

In a following experiment the influence of the time-factor on the final concen-

tration of shell water was further investigated. For this purpose the concentration
of shell water was determined for various groups of animals, staying under similar
external conditions and having access to both 0% and 100% S.W. The time-
intervals for the determination were I, 2, 3, and 5 days. After determination of the
concentration of shell water, the animals were discarded and replaced by new
animals from the storage box, which were similar as to origin and previous adap-
tation. To provide an undisturbed uptake of water at the drinking fountains,
never more than six animals were tested simultaneously in one tank. As a result of
the restricted equipment, which did not permit the use of more than four experi-
mental tanks, the animals in which the shell water could be measured after a 5-day
interval, were not as numerous as those in which shell water was measured after
shorter intervals. Thus in the first experiment, which lasted for 10 days, 58 animals
were measured after 1 day, 29 after 2 days, 18 after 3 days and only 12 after an
interval of 5 days. For a second series, lasting 15 days, the numbers in the groups
were 74, 41, 30 and 17 respectively. In spite of this imperfection the average value
of the measured shell water concentrations in each group increased from 30.3%0 S.,
over 30.6 to 32.0 and 31. 9%0 S. in the first series and from 31.1 %0 S. over 33.4 and
34.3 to 35.6%0 S. in the second series.
It is clear that shell water concentrations tend to become higher after longer
intervals. In the first place this must be ascribed to evaporation. New water,
however, is taken up by the animals in the course of time. As only relatively few
animals were found with low shell water concentrations it must be assumed that
the rise in concentration is also brought about by a more balanced uptake of water
as the animals grew more familiar with the experimental tanks and the available
shell water.

As the factor time has a (direct) effect on the shell water concentration, due to
higher loss by evaporation, a similar effect may be expected to be caused by higher
temperatures of the environment; lower relative humidity of the air and stronger
air movements (wind) mayas weil affect evaporation positively. Uptake of the
right amounts of drinking water of 10w concentrations will bring the concentration
back to the correct value.

It will be clear that in this way a system of shell water regulation

is encountered in which a fixed concentration may never be ex-
pected. lust after a dilution the shell water will have its lowest
value; little by little, depending on the rate of evaporation, the
concentration will rise. This will continue until the animals start
to experience the concentration as being unfavourable. If possible
they will now start to dilute the shell water again.
Somewhere between the lowest and highest values the shell water
 103

will have its optimal concentration. It is, however, not absolutely

necessary that this optimal concentration coincides with the maxima
in the frequency distributions of the concentration of the shell
water.
Under favourable conditions most animals will be able to keep
the differences between lowest and highest concentration of the
shell water rat her small. In Fig. 20, 88% of the measured values lie
within the salinity range or 27-33%0 S. Preliminary experiments
proved that the majority of the animals took up new water every
night.
To get more information about the drinking frequency of Coenobita clypeatus
the activities of over 200 tagged animals in the outdoor crab pen were followed
continuously. Observations during part of the night in the end of April 1967, at
rather dry weather conditions, indicated that as many as 50-65% of the animals
visited the drinking trough between sunset and midnight and quite a few of them
were seen twice near the trough. During the rest of the night a considerable number
of animals would find their way to the water. Occasionally drinking animals were
seen during daytime. From these data it was estimated that 75-100% of the animals
took up water every day.

When circumstances get less favourable the differences between

initial and final shell water concentrations will of course increase.
For instance when the animal fails to dilute the shell water in time,
the final concentration may increase to 45%0 S. As a rule, however,
they will get rid of the water before this situation is reached. It is
already known that water with salinities of over 36%0 (= 100%
S.W.), was hardly ever retained as shell water; typical coastal
animals were seldom found to carry shell water.
Apart from the quest ion of the optimal concentration, which will
be discussed later, other interesting data are the lowest concen-
tration to which Coenobita does not object, and the highest concen-
tration at which it is desirable for the animals to dilute or to
remove the shell water.
As could be demonstrated in a subsequent series of experiments,
the salinity of the initial concentration also depends on external
environment al conditions. The ability to adapt the rate of tolerance,
as far as the maximum (and minimum) concentration is concerned,
must be very valuable indeed to the animals, as it will enable
them to utilize a certain amount of water over a longer period under
 104

less favourable circumstances, but on the other hand it will make a

great demand on their osmoregulation.

Influence of air humidity on regulation

This experiment describes the differences in concentration of the shell water in
two groups of Coenobita, exposed to different degrees of evaporation, induced by
high and low relative humidities of the air. Other factors, which are supposed to
affect the evaporation, such as temperature (indoor temperatures of 28°C) and
wind, were kept as constant as possible. As a suitable experimental set-up 2 similar
tanks were chosen as described, but now both were enc10sed by a much larger
(80 X 50 X 50 cm) eternite container, covered with an air-tight plate-glass window.
Through small holes in the side walls of the outer container cables were fitted for
temperature- and humidity probes, which were mounted under the wire-mesh
covers of the inner tank. Both types of protes were connected with arecorder.
The bottom surface of one of the containers which was not occupied by the inner
tank, was filled with a number of shallow trays with distilled water and provided
with cottonwool evaporators reaching up to the roof of the container. On the
bottom of the twin container shallow trays holding about 2 kg of dry silica gel were
placed.
Under temperatures of 28 ± 2°C, the relative humidity in the wet tank fluctuated
around 95 ± 5% R.H. after so me hours; at the same temperature the R.H. in
the dry tank was 40 ± 10%. At night (24.00 h) six normal animals, with all shell
water shaken out, were released in both tanks. Next morning the concentration
of the shell water present, taken up du ring the night, was determined. Three nearly
identical series of experiments, in which respectively 120, 48 and 72 animals were
used, gave higher averages of the concentrations in the wet tanks than in the dry
tanks. Results are given in Fig. 23. Apart from the differences demonstrated here
in the measured concentration of shell water - which seem to be controlled by the
external environment, in this case the difference in R.H. of the air - it was also
interesting to notice the quite different behaviour of the two groups of animals.
"Dry animals" proved to be much inclined to take up shell water. After each night
the animals were found huddling in the corners of the inner tank. The shells were
well-c1osed with chelipeds and ambulatory legs. During the removal of the shell
water for conductivity measurements it was noted that the shells were filled to
the brim with shell water. The "wet animals" on the other hand showed little
inclination to retain shell water. Though the shells of these animals were moist to
the touch on the outside, shell water itself was found to be present in considerably
smaller quantities and often not at all. In the "wet animals" the number of un-
successful shell water determinations due to too small quantities of shell water
was 28 out of a total or 120 tested animals. In the "dry animals" this number
amounted to only 8 out of 120 animals. In general the "wet animals" behaved
much more fidgety; walking about or sitting with ambulatories widely extended.
As a more or less logical sequel to this experiment attempts were also made in a
series of experiments to investigate the effect of various temperatures - 20°, 28°
and 36°C, at constant relative humidities - on the composition and maintenance
of the shell water concentrations.
 105

numbers
I
I
humidily high ~
cri
10 o::.!!Q

OJV----------------~

10
humidily low

o
20 25 30 35

Fig.23. DISTRIBUTION OF SALINITY OF THE SHELL WATER in Coenobita kept at

high (95 ± 5%) and low (40 ± 10%), r elative humidity both at 28°C. - Broken
line: 100% S.W.

Due to the lack of suitable equipment in the Institute at Curayao it was im-
possible to obtain a simple experimental device in which two or more different
constant temperatures could be maintained under equal relative humidities.
Several devices, utilizing evaporators of various sizes failed to give satisfactory
results, for the small but unavoidable changes in temperatures always induced
llluch larger changes in hUlllidity.
As the effect of the R.H. on the concentration of the shell water had already
been studied the experiments were not continued. Taking the observed lower
concentrations in the previous experiment to be areaction of the "dry animals"
to the high evaporation caused by low humidity of the environment, higher temper-
atures are expected to cause a similar effect.

FIELD OBSERVATIONS

Compared with the laboratory experiments the number of

animals in the field of which the concentration of shell water was
measured is rather small. In 1966 on two occasions shell water was
measured in small, medium and large animals near Punta Tera at
Oostpunt, Cura<;ao. Most of them were collected in the gully which
 106

~ 14 Nov. 1966
numbers Punta Tera t Well4
10
8
6
4
• 19 Nov. 1966
the same, but now
after heavy rains

2
0
15 20 25 30 35 40
12 I
I
30May 1967 I
10 I
Punta Tera, Weil 4 I
8
6
3.0-3.9
4
2
0
20 25 30 40
8
8 Sept. 1967
6
East from Lagoon 0 287
4
2
0
20 25 30 35 1 40
10 I
I
11 Sept. 1967 I
8 I
near Lagoon C I
6 I
I
17.5
4
2
0
I
15 20 25 30 35 40
SO/oo

Fig.24. DISTRIBUTION OF THE CONCENTRATION OF SHELL WATER from Coenobita

collected in the field. - Concentration of shell water is regulated on a level of about
32-33% S. Heavy rains lower the concentration. Vertical arrows: concentration of
available drinking watet.

is situated eastwards of WeIl 4. In November the terrain was

rat her dry; 5 days later heavy rainfaIl occurred. The results of
both sampling days are given in Fig.24, upper histogram.
 107

In 1967 shell water was collected in the field three times. On

May 30 animals were collected near Punta Tera again, now in the
plain immediately east of the weIl. Lots of animals fed here on the
fruits of a kind of thron-apple (Datura metel). On September 8
shell water was sampled in a small number of Coenobitas eastwards
of Lagoon D and on September 11 in animals congregating at the
northern bank of Lagoon C. Histograms of the results of the last
group of she11 water measurements are also presented in Fig.24.
Data on the she11 water of Coenobita clypeatus sampled in Bimini,
Bahamas, will be published separately.

As mentioned above salinity of the shell water of (inland-)animals

sampled in some of the natural habitats in Cura<;ao ranges from
24%0 up to values of 37%0 S. The most frequently occurring values
are those in the range of 31-34%0 S. Again the majority of the
examined concentrations of she11 water lies below the concen-
tration of normal (100%) sea water in the Caribbean.
Due to heavy and frequent rain showers the she11 water may
show a decrease in concentration, as is obvious from the data of
November 19.

FINAL CONCLUSIONS

The fact that terrestrial hermit crabs must be counted among

the most successful land inhabiting decapods, is for the greater
part due to the role played by the adopted shel1. Besides some
obvious advantages, such as protection against damage and preda-
tion, its chief functions are the prevention of desiccation and the
retention of she11 water. Under favourable circumstances the animal
not only retains a considerable stock of water, but is also able to
keep the salinity of that water at a more or less constant value
of 32 to 33%0 S. The import an ce of water management for the
animal is obvious. Moreover, in this way it creates an intermediate
environment between internal and external media, which contri-
butes greatly to the facilitation of osmoregulation.
It must be noted that by no means all hermit crabs under natural
conditions possess shell water. As a rule small individuals, or
 108

animals living near the sea, do not store water. So it is not indis-
pensable. On the other hand shell water is practically always found
in larger individuals of the inland animals. Observations indicate
that the variation in concentration of this shell water is sometimes
limited (Fig. 24). The regulatory mechanism, however, often com-
prises much greater fluctuations. Besides, a more or less skew
frequency distribution of the measured concentrations may be
expected. Individually the concentration of"shell water will show
a gradual rise, followed by a sudden fall at every dilution, yielding
a kind of serrated line.
From laboratory determinations on hundreds of animals it
appeared that the lowest value was a salinity of about 10%0; the
highest values were found at ab out 44%0 S. Various diagrams in
Figs. 20, 22 a, band 24 also show a distinctly skew distribution.
Generally the diagrams show a fairly distinct upper boundary at
36-37%0 S. There is no c1ear lower limit. Depending on various
causes, such as different sizes of the animals or differences in time
available for stabilization, or in humidity of the environment, the
maxima of the diagrams of the various experiments shift between
about 30%0 and 35%0 S. It is by no means certain that those maxima
also represent optimal values, values at which osmoregulation
requires the least possible energy, or in which a constant gradient
with the environment is maintained. The decisive factor is wh ether
an animal can renew its shell water in time.
 V. OSMOREGULATION

INTRODUCTION

For the greater part of their life the truly terrestrial crabs never
return to an aquatic life. Strictly speaking, these animals should
have no osmotic problem, since they lack an aqueous environment.
Sometimes, even on dry land, terrestrial crabs may still have a
concealed extern al environment in the shape of water retained in
the branchial cavities and, as is the case in hermit crabs, the shell
water. Continuous water losses, however, both from the animal
and the retained external storages, cause an increase in concen-
tration of body fluids as weIl as of extern al water.
According to various authors terrestrial crabs are capable of
osmoregulation, or at least of maintaining the osmoconcentration
üf the body fluids within certain limits. Depending on the environ-
ment in experiments or in the field they show either hypo- or
hyperregulation (see e.g. GROSS, 1957; Buss, 1968).
Here the concept of osmoregulation is used in its widest sense:
that of keeping the body fluids at a fairly stable level, induding
the possibility of hyporegulation in avoiding or diminishing high
concentrations, caused by evaporation. It appears that in adult
land crabs the gills, pericardial sac and gut are particularly involved
in the uptake, storage and redistribution of salt and water, whereas
the antennal gland are mainly concerned with the ion-regulation.
From all terrestrial species the terrestrial brachyuran decapods
have been most thoroughly investigated as far as osmoregulation
is concerned. Unfortunately far less attention has been paid to
 110

terrestrial anomuran decapods and, though recent studies on other

decapods may be of general application, the knowledge about
terrestrial hermits is mainly based on older publications. Such
references, however, are scarce.
In the Tortugas, PEARSE (1932) measured the depression in freezing point of
the blood of Coenobita clypeatus and found that the range was 1.90 to 2.21 °C (average
2.09°C), which is slightly higher than the value of 2.04°C for normal sea water with
a salinity of 36.05%0 in that area. A species of the marine genus Petrochirus gave
comparable values, from which the author concluded that both Coenobita and its
marine ancestors have blood contents equalling normal sea water. For terrestrial
brachyuran species, such as Gecarcinus and Cardisoma, however, distinct1y lower
values were obtained, ranging from 1.55 to 1.79°C.
In Birgus and Coenobita, GROSS (1955) measured various concentrations of body
fluids, dependent on the salinity of the available drinking water. When both fresh
water and sea water was offered to Birgus, the concentration of the body fluids
ranged over a value equivalent to 85 to 92% sea water. When only fresh water
was offered, concentrations equivalent to 65 to 74% S.W. were found; when only
normal sea water was available, the concentrations in the body might increase to
the equivalent of 118 to 128% sea water.
When given the choice between fresh water and sea water, animals with low
blood concentrations prefer the latter, and with high concentrations the former.
In this way Birgus will select the proper amounts of water of different salinities to
adjust its body fluids till the optimal value for the species is reached. Terrestrial
hermit crabs show a similar behaviour; as will appear from this paper, they are
capable of adjusting the concentration of their retained shell water in a similar way.
Remarkably constant values for the blood serum of Birgus latro were found by
HARMS (1932), who obtained freezing points of - 1.42 to - 1.49°C. In addition
HARMS mentions comparable blood concentrations for two Coenobita species from
the same area.

EXPERIMENTS AND OBSERVATIONS WITR COENOBITA CLYPEATUS

Attempts were made to study the concentrations of the blood, the

urine, and also the shell water, as weIl as the osmoregulatory
abilities of Coenobita. Moreover the relations between the osmo-
concentrations and the environment al conditions as met by the
animals in their natural habitats were investigated.

Fig. 25. RELATION BETWEEN WEIGHT LOSS BY DESICCATION AND BODY FLUID
CONCENTRATION, ßI, as observed in four Coenobitas (a, b, c, d). - Solid lines:
weight loss by evaporation; broken lines: increase in body fluid concentration
derived from freezing point determinations (0) in °C; dotted lines: theoretical
concentrations of body fluid in °C calculated from the concentration indicated by
arrow and daily water loss; t: death of animal.
 111

gross wgt ,g
67 Ai:C
1.8

1.9

20

2.1

2.2

2.3

2.4

25
r
2.6

27

2.8

2.9

o 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 days

gross wgt ,g
68

1.9

20

2.1

2.2

23

24

2.5

2.6

27

2.8

2.9

3.0

3.1
, , , , , , , , , , , , , , , , , ,
0 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 21 days
Fig.25a.
 112

.b.j ,oe
1.9

2.0

2.1

2.2

2.3

2.4

t 2.5

2.6

2.7

2.8

2.9

3.0

6 2 :3 4 5 6 t 1\ 9 10 12 13 14 15 16 17 18 days

.b.j ,oe
gross wgt,g 2.0
63
~
\~ 2.1

\~
62
2.2
61 "-
\\~ \ 2.3
60

\\~
2.4
59
2.5
58
1\ ° "
57
\ '. "-'-. 2.6

\.,. t 2.7
~.
0,.
56 ~ .....
0.
2.8
\"';",
"",
,..
o'\'~" 2.9
"'. ...
\",
\
o \~. 3.0
\ .
\

3.1
d
0 1 2 3 .\ 5 6 t 8 9 10 1'1 12 13 14 15 16 17 18 days

Fig.25b.
 113

Effect of dehydration on concentration of body

fluid

In addition to the experiments described in Chapter III the

effect of water loss by evaporation on the concentration of the body
fluids of Coenobita was investigated in order to answer the im-
portant question whether these animals have some mechanism to
lower the concentration of their body fluids.
It has already been demonstrated how, and at which rate, these
los ses occur and it was shown that a water loss equalling 15-20%
of the net body weight is lethaI. It is not direct1y clear, however,
which organs are responsible for the hypo-osmoregulation, if no
free water is present, neither in the branchial cavities nor in the
shell.

Methods

In a number of desiccating animals the water loss under environmental conditions

of 28°C and a R.H. of about 75% was found by determination of the decreasing
gross animal weights twice a day. In addition a very small blood sampie was
collected every day for measuring freezing point depressions, by penetrating a
membraneous joint of one of the ambulatories with a finely drawn out glass cap-
illary. After sampling the animal was weighed again for determination of the weight
loss caused by the sampling. As the blood sampies averaged only 0.006 g each, this
loss was left out of consideration in the further calculations. Each of the four
experiments ended by the death of the anima!. Then wet and dry weights of shell
as well as the dry weight (24 h at 110°C) of the animal itself were determined.
From this information the water content of the animal in each step of the desiccation
process was calculated.

Results

In four cases Fig. 25 shows the decreasing gross animal weights

as weIl as the increasing body fluid concentrations, ~1, as calculated
from the freezing point depressions. Also the concentration curve
as ca1culated from water loss is given. From the moment the water
loss becomes more or less constant after the initial period of greater
 114

10ss, probably due to loss of shell water, the theoretical values of

the freezing points are calculated from the concentration (indicated
by arrow) and the daily loss of water.
It is clear that the effect of (hypo-)regulation becomes visible
when the measured curve deviates in an upward direction as
compared to the theoretical curve, derived from water loss. Since,
however, the two curves tend to coincide for the greater part, or
even the measured curve lies below the theoretical curve, such a
regulation is not apparent. Therefore a decrease in osmoconcen-
tration, i.e. a reduction of the total number of particles in the blood,
cannot weIl be attained by the animals under circumstances in
which they are exposed to desiccation and free water is not availa-
ble.

The effect of reviving

In the desiccation experiment discussed in Chapter III the increase in concen-

tration of the body fluid as a result of dehydration was determined by conductivity
measurements. After ten days of desiccation, the animals had lost approximately
15% of the net body weight. The body fluid concentrations, as far as the electrolytes
are concerned, had increased up to the remarkably high values of 170% of the
initial concentration. Then fresh water was offered, which was taken up avidly.
Determination of the net body weights showed that the animals had regained or
even surpassed their original weights within two days. The body fluid concentration
had again reached the original values (Fig. 26).

DISCUSSION

From these findings, combined with those of the previous ex-

periment, it may be concluded that osmoregulation in Coenobita
clypeatus stands or falls with the presence of water. It seems as
if the animals are only able to control the concentration of the body
fluids when free water is available, either taken up directly or
retained as shell water. Whether this happens by a direct uptake
by the mouth parts and penetration of the walls of the alimentary
tract, or by an exchange of water and salts from the body fluids
against water or salts in externally stored water, is still uncertain.
Thus is the absence of water, the animals must at any rate be
 115

tli I electrolytes
170%

0 160

150

140

130

120
weight %
110
110 0 0
100
100

',,-
IG)
1-
IC
I~
I
I I I I I I I I I I
2 3 4 5 6 7 8 9 10 11 12
days
Fig. 26. CHANGES IN BODY FLUID CONCENTRATION, ßt. as an effect of water loss
and reviving. - Net body weights (dots) and electrolyte concentrations (circles)
derived from conductivity measurements expressed in percentages of initial values
(= 100%). After 10 days of desiccation water was offered.

able to tolerate rather wide variations in internal concentration

or in other words behave more or less like osmoconformers.
When water is available the animals will turn to osmoregulation
and behave like osmoregulators. From the observations by GROSS
(1955) on Birgus it is clear that, even with water of different
salinities available, considerable variations in body fluid concen-
trations may still occur.
 116

Osmoregulation, when water in any form is present

As osmoregulation appears to fail or at least to be of minor

importance in animals deprived of water, the osmoregulatory
capacities in terrestrial hermits having access to water will be
discussed first.
The existing literature (GROSS, 1955) on the capacities of the
related species Birgus records rather diverging concentrations,
depending on the concentration of the available drin king water.
A good deal of useful supplementary information on this subject
was already obtained during my stay in Cura<;ao in 1961. Unfortu-
nately in that period neither a cryoscopical device, nor a micro-
method for conductivity measurements were available. Therefore
only determinations which required at least 0.5 ml blood could be
made, which meant that continued series of measurements with
the same individual were practically out of the question. In general,
experiments tended to require a good number of animals, since
every determination caused the death of the animal.

A preliminary experiment already demonstrated that GROSS' observations also

held for Coenobita clypeatus. When drinking water of various salinities is offered to
- somewhat desiccated - hermits, a notable change in the internal fluid concen-
tration results immediately. When normal seawater is offered the concentration of
the body fluids increases with 4-20% within a lapse of one day. When on the
other hand distilled - or tap water is offered, this will in the same time in du ce
decreases ranging from 4 to 24% of the initial values. From the rather quick effect
of the absorbed drinking water the impression is gained that water and salts will
pass the walls of the alimentary tract very quickly.
Of course experiments of such a short duration will only give very limited infor-
mation. In the first place the amounts of water absorbed differ largely, depending
on the salinity of the water. Moreover a certain amount of shell water may be
retained and when it is not, the osmotic possibilities of the abdominal :tegument
are for the greater part ruled out, if indeed they play any role at all. To avoid
these complications, at least to a certain extent, the animals were placed in tanks,
filled with a layer of about 2 cm of water of a given concentration, so that the
shells became filled and the entire body of the crab was sufficiently bathed in the
water. On the other hand enough of the animal's body emerged to warrant anormal
function of the respiratory surfaces. From their behaviour it was clear that the
animals were not comfortable. Most of them tried to extend their ambulatories
well over the surface of the water, or they would try to climb on top of each other
to get out of the water. From time to time clusters of animals had to be disentangled.
 117

The effect of osmotic stress on mortality

Ta study the effect of various osmotic stresses on Coenobita a mortality test was
carried out. A number of tanks were filled with water of a salinity of 35, 40, 45, 50,
55, 60, 65, 70, 80, 90, 100, 110, 120, 130, 140, and 150%0 resp., and ten animals
were placed in each. Every day the salinity of the water was checked and adjusted
if necessary. At first the number of dead animals was counted at intervals of 30
minutes, later on at langer intervals. Animals were considered to have died when
they did not react by pinching, when the hairs inside the large nipp er were touched,
and moreover, when they could easily be removed from the shells. The experiment
was continued for 45 hours. The temperature was kept at 28°C.
The main result was that all animals in concentrations of 35 and 40%0 survived .
In 50%. the survival rate was 50%, in concentrations of 70%0 and over, all the
animals died. In the highest concentrations the first animals were found dead after
I! to 2 hours. After about 8 hours the general picture of mortality was practically
stabilized .
At concentrations of 90%0 and more the mortality was 100%, the animals
dying more or less simultaneously after an interval of 6-10 hours (Fig. 27).

100% level

S%.
150
140
130
120
110
100
90
80
70
65
60

~~~~----~~~~==::::~::::::~====~S%.
55
SO . 45
45 0% level 40
40 35
35
o 5 10 15 20 25 30 40 45h
time

Fig.27. :MO RTALITY INCoenobita clypeatus AT VARIOUS SALINITIES of an imposed

external medium as a function of time.

From these results it is clear that against osmotic death, supposed

to be caused by permeating salts or by water loss, the tegument of
Coenobita gives only a very limited protection.
In this connection it is interesting to stress that in similar ex-
 118

periments with diluted seawater mortality only occurred after

much longer periods. Even a two-days stay in repeatedly renewed
distilled water did not result in any significant mortality. (Compare
also the treatment of the so-called "aqua dest" -animals). Since the
cause of death becomes controversial, the long-lasting experiments
in diluted sea-water are not discussed here. Yet they might lead
us to suppose that either tolerance against dilution of the body
fluids is much greater or that the prevention of either withdrawal
of salts or penetration of water in the reverse direction is more
effective than in the opposite case.
Still the concentrations of the body fluids are seriously affected
when the animals remain under imposed conditions of osmotical
stress.
The next series of experiments will show how in this case the
osmotic concentrations of the body fluids change in the course of
time.

The effect of osmotic stress on body fluid concentration

Use was made of external concentrations equivalent to salinities of 0.1 %0 (dis-

tilled water), ± 1%0 (tap water), 36.0%0 (normal sea water) and 50%0 (hypersaline
sea water). In the 175 animals originating from Klein St. Joris, Curayao, February
1961, and used in the above-mentioned concentrations of 0.1, 1.0, and 36.0%0' the
initial body fluid concentration, as found from conductivity measurements, aver-
aged an equivalent salinity of 22.5%0' In another fifty animals from the same
locality, but collected two months later and used in hypersaline sea water of 50%0'
the initial concentration averaged a salinity of 25.5%0' The experiments lasted 7
days. Every day the water in which the animals were kept was removed and
renewed, and the concentrations of both the old and the fresh water determined.
After periods of 1, 2, 3, 4, and 7 days respectively, ten animals were removed at
random from each of the four aquaria and the conductivity of their body fluid was
measured.

The results of these experiments are summarized and plotted in

Fig.28.
It appears from the diagrams that the osmoconcentration of the
blood, as far as the electrolytes are concerned, undergoes con-
siderable changes as a result of the different environments. Distilled
water and tap water induce a decrease in concentration, normal
 119

Aj,electrolytes, S%0
50
hyper saline
sea water
50%0
40

30

20

10
6 i
3 4
i
5
i
Ei -7
days
0 i
2
I
3
i
4
i
5 6
I i
7
days

Aj,electrolytes, S%o
40

30

t-~--~---~_~ _t-Y-rl-~--i
<::-i
20

10 distilledwate, tapwate, !I%.

6 2345670 2 3 4. 5 6 -7
days days
Fig. 28. OSMOTIC CONCENTRATION OF BODY FLUID, ßI. electrolytes of Coenobita,
expressed in equivalent salinity in %0' when various osmotical stresses are imposed
- Indicated are the mean, and the ranges on either side of the mean.
 120

sea water and hypersaline sea water an increase in concentration.

After an initial fall or rise, respectively, the lines for tap water and
normal sea water tend to run horizontally, pointing to osmoregu-
latory abilities. A tendency to stabilize the blood concentration at
a new level is far less apparent with the animals in distilled water
and absent in hypersaline sea water.
For animals in sea water the new level varies around an equivalent
salinity of 30%0' for animals in tap water around 17%0. It is to be
doubted whether such an equilibrium will be found with animals in
distilled water. On the seventh day of the experiment there is still a
considerable spread in the values in the diagram, and moreover a
heavy mortality begins to appear. With animals in sea water and
tap water no mortality occurred. With animals in hypersaline sea
water the curve starts to rise quickly right from the start, and
continues to rise during the next days. After the fourth day, when
the average concentration of the electrolytes in the blood is about
40-45%0 the mortality is very great indeed. On the seventh day,
the majority of 15 animals which had been kept in reserve for
replacement, and which were kept in the same environment, had
died, and it was very difficult to measure the conductivity of the
blood of the remaining animals. The abdomens of these crabs had
shrunk considerably and held only very sm all quantities of blood.
Two important conclusions may be drawn from this experiment:
in the first place that the animals possess abilities for both hypo-
and hyper-osmoregulation, provided that they stay in an aqueous
environment, and secondly that a considerable range in the concen-
trations of body salts can be tolerated, as concluded from con-
ductivity measurements, and therefore only concerning the elec-
trolytes.
High mortalities in the distilled water and hypersaline sea water
series indicate, however, that there exists an upper as weIl as a
lower limit to this tolerance. The upper limit is found at an equi-
valent salinity of ab out 45%0' the lower at ab out 15%0 s.
As has been stressed it must always be kept in mind that the
above data were found under conditions of osmotical stresses
imposed upon the animals. In its natural surroundings Coenobita
clypeatus will never be faced with such extreme conditions. The
 121

animals are never to be found in water, and shell water of either

low and high concentrations is avoided.
This is why another series of so-called endurance tests was
carried out, in which the long term effects were studied of a situ-
ation in which the crab was offered drinking water of a fixed
salinity over long periods. It was left to the animals whether they
would imbibe the water offered, retain it as shell water, or leave it
alone. Precautions were taken to prevent the crabs from falling
into their drinking bowls and thus creating a situation as had been
tested in the previous experiment.

Endurance test 1

In the first series three types of drinking water were offered, viz. fresh water
(S about 1.0%0)' normal sea water (S 36.0%0) and hypersaline sea water (S 50.0%0).
In large glass tanks the bowls, covered with a grid, were placed in such a way that
the crabs could only touch the water with their chelae. In the freshwater test 5
animals were tested over 105 days, the other two tanks contained 10 animals each.
The animals were tested in normal sea water for over 105 days, those with hyper-
saline water for 43 days. Every two days the gross weight of each crab was de-
termined, to be converted into the net weight afterwards and then again into the
percentage of the initial net weight. Fresh water and sea water were renewed every
day, hypersaline water every two days. The temperature was kept at 28°C and the
relative humidity at about 75%.
The results of the experiments are represented in figures 29, 30 and 31, summa-
rizing the weight percentages and including several other data, such as shell ex-
change, mortality, cannibalism, etc.
On the first day practically all the animals show a definite increase in weight, as
water is taken up. The fluctuating course of the graphs may be explained as follows:
1. Water uptake is followed by loss through evaporation (viz., in the parts of the
curves where the weight decreases gradually for some days). 2. It is known that
considerable quantities of water may be stored as shell water. It is hardly avoidable
that some of this water is lost when the animals are being weighed. The greater
part of the sudden drops are due to this loss. 3. When two animals exchange shells
mutually, the altered relation between volume of animal and volume of shell may
often result in a marked change in weight. In the figures such changes are indicated
by a vertical dash. 4. Other causes of different nature, e.g. seH-amputation of legs,
are also indicated.
In Figure 29 (fresh water) the weight curves show an obvious descent in the course
of time. This is especially clear with the animals I, 2, 3 and 5. The drawn lines have
been drawn by sight. During aperiod of 105 days 4 out of the 5 animals died and
three exchanges of shell took place.
In Fig. 30 (normal sea water) several different weight curves show a horizontal
 122

net wgt %
--I ••••
e. •• • ••
130 ••
120

110
1
I I0 ~ .&............. ..!..!•• 1; •.•.•.•.a.,. •..:..::... •
100-'
I i i
~.L.:-
i ,=+- 2

lT;"···.!.!.·!._• . ~
. . - .-
100 .! •••
90 _.1). ••• ••••
••
.. -.. • .-t 3
80-' i i i i

110 1...........
1 1. . . . . . . . . . . . • • • .~I··.-tl;...-...a.!.!-;-··.-o-O-.~.• 4
100·t'---.---.--~--~--~--~--~--~--~--r-

120
j ••• .I...·--.·~_·__• •__
• •• • • ·_·· . ..
.---!-. • • ....... ~.-.-.-.- . ---.------
110 ••••
•
t
,.- 5
100-' i i i i i i i i

o 10 20 30 40 50 60 70 80 90 100
days
Fig.29. ENDURANCE TEST WITH FRESH WATER (S = ± 1.0%0)' Net weight in
percentages of 5 Coenobitas versus time. - After initial increase of weight by
drinking during the first days all individuals gradually show loss of weight. Vertical
dashes: shell exchange; t: death of animal; 1): self-amputation of ambulatory;
2) death by cannibalism.

Fig.30. ENDURANCE TEST WITH NORMAL SEA WATER (S = 36.0%0)' Net weight in
percentages of 10 Coenobitas versus time. - Various weight curves show horizontal
or slightly upward trends. Vertical dashes: shell exchange; t: death of animal;
1): seH-amputation of ambulatory.
net wgt %
130
•••
120 ._-_.&.-
• •• •
•
e",;,_-.-;_ ---.;.._-••
•••
•
• • • •
~._._':;...---.
•
110
•..:.,.. .•'----.,----. 1
·
•

oo.,--.,--.,-----,r----r---r---T"""--r-.::-..t.t.
. ---":1.. ;-..........
120 1+·-:':"'~·-·~.·
• • • --·..·"--·-.-·-t-
110
100·'---r--.--.---.---.---.---r---r--"""T'""---.------.
2
J __ •__ e_ _ -_._._.-._ _.-.- • . - : . .-._-_.- •
110 ,.. •• •• •• •• • •••• •• .-.-.-.---;:-. • • • • • • • .t
100.'.-......,..---.----.---,---~-~-r---....----r---..-........, 3
.. . .. . . ..
130
120
•
.
.,....-!-••-.-•••-.,-",.-.:'--.-.- --~~-~ ~.--
• • ••
--
,- -:--- --~---
••• =-; e_ • .&..r--
110
4
100·t-----T""""----r------r-----.----,i••.--,------.----~-------r-----~---~
110 1 .. ... •· •• · • • • • f
••
-r.-. --r--=.....,.--__r---,----r--....,----. 5
1-.. _. .-._._._. -.__. ._.-
--,:-'.
100-,-· • i • ---'---.

•
120
110 • • • • •• 1)---!. • • • .!.L.• •• • •
......',:,"___
•••• fI7 - • e-e_ 6
• • • • •
....-~--1"'""""-~--,--~--~--.--....,

j-.-...~.!.
100·'----~-~-

110 •• -;.~.e;_;__.~.-.--._
•• ..
-·...;.·'1.r....·Io,· .•...,.....---.....• e--;-
• •
• 7
100-, •
I
100·j'.· •••••• ..,..-.
• •
. ••
.. • •• • ••• •••• •
-.-.T.-.-.r-.-.-'-;:--r-.--.--. ~
8
• •

. .. .' ... •
120 1':'·":-. ".-.-.-._.,.j'L!,._ .-._&•.:.~,&.."•..;';..
• ••
,-----:-
.' t
110
100-,
9
j•
.--r-------r------Ti----•.....,..---__r-------r------;-------.----~---~
..-r.......-!.a..--!..!.. ' ••• ' •• •
· · .,---.,...--..
•• ·.. -t
110 •• 10
1006-~10--2~'O--3~0-~40--5~'O--6~O--i~0-~80--9~'O--10~O--~1I0
days
 124

or even a slightly upward trend (drawn line). In 105 days 5 out of the 10 animals
died, 2 exchanges took place.
In Figure 31 (hypersaline sea water) practically all the curves show a downward
trend. Five out of the 10 animals died within aperiod of only 43 days. Three ex-
changes 01. shell took place.
A similar test in which six animals were offered hypersaline sea water of a salinity
of 120%0 resulted in the death of all animals within 24 hours.

When the results of these endurance tests are compared the

conclusion may be drawn that under the given circumstances
normal sea water offers the best chances of survival of these ani-

net w~ %

12°1·~
• ••
net wgt%
~,
··
110

. •••
.
6
12°1~ ..
'~:1~;.~
~
110 .-a-.~. 1

· .'.
100"1, "
1 "7t ' 7
110 1~_...:-....
1 ••
..
.~
2
100., i ,

Ilol;·~ ..

·
100-, , , ,.
1 •,
~-
·~8
1I01~·~·
· 3
100·'X'

120 1. ..~ .
."
.~..... .~.
100-, , , ,~.

l ~'~ ~~.~
110 • ••,
110 ••• 4 100·' ' , ,
t
,
9
100-, " ~-....
·I:C-•••
120
1<' •••• >:=I '.'...
110 1'. 5 110 ·~
100-'1~\-r,--r-,
t 10
100-'
o ·t5 o
- " T " i- - , , - i , , i
20 30 40 10 20 30 40
days days
Fig.31. ENDURANCE TEST WITH HYPERSALINE SEA WATER (S = 50%0). Net weight
in percentages of 10 Coenobitas versus time. - Nearly all individuals show a rapid
weight loss. Vertical dashes: shell exchange; t: death of animal.
 125

mals. Loss in weight and a tendency to exchange shells are also

assumed to be symptoms of a lowered well-being.

Endurance test 2

This test was set up with a greater diversity in the drinking water offered, since
water of concentrations of 0, 10, 50, 100, 150 and 200% S.W. was offered and
moreover all series were set up in duplicate, one series at 22°C and one at 30°C.
It was tried to keep the humidity as constant as possible during the experiment,
but under the given laboratory conditions in could not be avoided that the extreme
climatic conditions caused the humidity to fluctuate between 70 and 85% R.H.
Here all weight alterations agree with those of the previous experiment. Both
0% S.W. and hypersaline water cause considerable loss in weight. When water of
10, SOor 100% S.W. was available, the animals were able to maintain a more or
less constant weight over a prolonged period. Differences between the two temper-
ature series are not decisive.
The survival times from Table 13 give more information. Again hypersaline sea
water proves to be unsuitable for survival; 200% S.W. is worse than 150%. On
the other hand 0% S.W. yields remarkably long survival values. The influence of
temperature is of particular interest; there are obvious differences between the two
series. In the colder series, with low concentrations of sea water (10 and 50% S.W.)
survival times are twice as long as in the warmer series. With 100% sea water
survival is about the same for both temperatures. With hypersaline sea water of

T ABLE 13

ENDURANCE TEST OF COENOBITA CLYPEATUS AT TWO TEMPERATURES

Available water in concentrations of 0, 10, 50, 100, 150, and 200% S.W. Concen-
tration of body fluid, At, expressed in equivalent salinity in %0 by conductivity and
by freezing point depression (av. of two determinations).

mean general Al Al
survival trend bycond. byF.P.D.
time in weight
in days change
22°C 30°C 22°C 30°C 22°C 30°C 22°C 30°C

O%S.W. 248 153 )j )j 20.0 23.8 29.5 29.3

10% S.W. 253 136 --+ --+ 25.1 29.5 32.1 33.5
50% S.W. 326 164 --+ --+ 28.1 30.9 34.4 35.6
100% S.W. 174 173 --+ --+ 32.7 34.2 44.6 40.4
150% S.W. 35 69 )j )j

200% S.W. 18 16 )j )j
 126

150% the survival time is twice as high in the higher temperature. It is not sur-
prising that this trend is not continued in the last and most concentrated solution,
since this high concentration is wholly unsuitable for drinking water, as is proved
by the short survival period for both temperatures. So, with certain reserves, we
may conclude that low concentrations of sea water are better endured in combi-
nation with (relatively) low temperatures and that high (but not too high) concen-
trations are better tolerated if the temperature of the environment is high. These
results are opposite to the findings in Crangon crangon (L.) (e.g. SPAARGAREN
1971) where combinations of high temperatures and low salinity, and low temper-
ature and high salinities are better endured.
When considering the data of blood concentration ßi, as derived from freezing
points and conductivity and both expressed in equivalent salinities, in the first
place an increase in concentration with increasing salinity of the available water
is obvious. Moreover regulation is evident from the horizontal parts of the curves

total,22°
45 ~

40

_A-~ electrolytes,300
35
ty- - - - -~.. .. ...Oelectrolytes,

30
fa------ ____ -0------
I/.~ ~---...... 22°

I •

,
I
I
25
6
20 •
r---r---------------,--------------~
o 10 50 100
drinking water,% S.W.
Fig. 32. TOTAL OSMOTIC CONCENTRATION AND ELECTROLYTE CONCENTRATION OF
BODY FLUID in Coenobita expressed in equivalent salinity in %0' at two temperatures
as a function of the concentration of the available drinking water. - Concentration
of drinking water expressed in percentages of sea water. After endurance test 2.
 127

(Fig. 32). At higher temperatures the position of the two curves representing elec-
trolyte concentration and total particle concentration is higher than of the same
curves at the lower temperature. At 100% S.W. only the curve for total concen-
tration shows the reverse.

The non-electrolyte component of the blood

The principal electrolytes in the body fluid are the inorganic

ions Na+, K+, Ca++, Mg++, Cl- and S04-- (ROBERTSON, 1960;
PROSSER & BROWN, 1962). The total amount of electrically active
particles, including the small amount of electrically charged organic
molecules, can be measured by determination of electrical con-
ductivity. Though important, the electrolytes contribute to only
part of the total osmotic concentration, since this is connected with
the total number of dissolved particles.
The non-electrolyte componellt, formed by proteins, free amino
acids, (lipids and carbohydrates) cannot be neglected. Since the
freezing point depression also depends on the total number of
dissolved particles, irrespective of their nature, these data are most
useful in this respect. By comparing freezing point depressions and
electrical conductivity, the contribution of non-electrolytes to the
osmoconcentration of the blood can be easily estimated.
Fig. 33 gives the concentration of body fluid ßi, as calculated
from conductivity measurements and freezing-point depressions, for
animals in which the concentration of their shell water is known.
This shell water is considered to form the actual extern al environ-
ment, ß e of Coenobita. The curves in this figure have been drawn
at sight. The distance between the curves, as measured along the
vertical, represents the share of the non-electrolytes. Thus it may
be concluded that non-electrolytes account for about 30% of the
osmotic concentration; a percentage which is remarkably high as
compared to data obtained with other crustaceans.
It is tempting to speculate on the conformity with the most successful group of
terrestrial arthropods, the insects, in which comparably high fractions of organic
substances, particularly amino-acids, contributing to the total osmoconcentration,
are found (PROSSER & BROWN, 1962).
GRIMM (1969) states that proteins and other non-diffusable solutes in the blood
plasma of 2 species of shrimps are probably present in low concentrations only.
 128

Äi.S%o
60

50

40
o

30

20

10
• I
30
i • I
40
i
, , , , '60
Äe.S%o
Fig.33. BODY FLUID CONCENTRATION, 61, total and electrolytes, both expressed
in equivalent salinity in %0' as a function of the concentration of the retained shell
water, 6 e, also expressed in S%o.

PARRY (1954) estimated protein in the blood of Palaemon serratus to amount to

about 12%, a high value when compared with other decapods (ROBERTSON, 1960;
vide GRIMM, 1969). According to SPAARGAREN (1971) the concentration of non-
electrolytes in the blood of Crangon crangon and Crangon allmanni, depending on
temperature and salinity, shows variations of 0-20% of the total osmotic concen-
tration.
 129

Osmotie concentration in the blood

It is also apparent from the deviation of the curves from the

isosmotic line in Fig. 33, that apart from the regulation by means of
the shell water itself, the abilities of regulation of the body fluids
(ßi) against the shell water (ß e ) also play an important part.
From these figures evaluation of the exact shape of the regulation
curves and the effect of temperature on osmotic regulation is not
wen possible, notwithstanding the fact that data from animals
kept at temperatures of 22, 28 and 38°e are included. In the area
with low concentrations of shell water the osmoconcentration of
the blood is hypertonic, in the area of high concentrations of shell
water the blood is hypotonie. Since, however, the concentration
of the stored water is also subject to regulation and the majority
of the animals are able to maintain a shell water concentration of
about 32-33%0 S, it may be said that in most animals the osmo-
concentration is slightly hypertonie to the extern al environment
(shell water) and hypo- or more or less isotonie to normal sea water.
Furthermore, because the curves in Fig. 32 run parallel, it appears
that osmoregulation is practically completely due to the regulation
of the electrolytes, while the absolute concentration of the non-
electrolytes hardly changes.

Isotonicity of the urine

From experiments in whieh the concentration of the body fluids

as well as of the urine of desiccating Coenobita were measured by
means of freezing point determinations, it was proved that the
antennary glands of these animals do not have any osmoregulatory
function. Body fluids concentrations expressed in equivalent
salinities increase as result of water loss by evaporation; while at
the same time determinations of urine concentration showed
exactly the same rise. Figure 34 demonstrates the isotonicity of
the two liquids.
 130

50

40

30

.'

20
I i • r I I I I i , i i i I I
20 30 40 50 60
~i,S"loo

Fig. 34. CONCENTRATION OF URINE, Au, of Coenobita at various fluid concentrations.

FINAL CONCLUSIONS

Terrestrial hermit erabs exhibit a remarkable variation in

osmoregulatory possibilities. Dependent on environment al eon-
ditions they may be fairly good regulators, trying to keep the
eoneentration of body fluids as eonstant as possible, both in hypo-
tonie and hypertonie media. When, however, a moist external
 131

environment is not available, the animals are eonformers, able to

withstand enormous variations in internal eoneentrations.
Here osmoregulation is superimposed on regulation of the shell
water, which is the only aquatic environment with which the
animal is eoneerned, onee its larval development is completed.
In view of the variations in salinity of this shell water - from about
10 to 44%0 - and a general average blood eoneentration whieh for
terrestrial erabs is slightly hypotonie to normal sea water, both
hypo- and hyperregulation have adaptational values for this group
of animals. A well-funetioning regulation of the sheIl water will
restrict its eoneentrations to values from ab out 30 to 35%0' Onee
this is obtained it is easy to maintain an equivalent body fluid
eoneentration of 34 to 35%0 S by osmoregulation.
If no shell water is present Coenobita has no adequate means of
hypo regulation. Progressive desieeation, whieh may go as far as the
loss of about 30% of the internal water at the utmost, eauses the
eoneentration in the body fluids to rise to ineredibly high values,
equivalent to a salinity of over 60%0' A great nu mb er of adaptations
to restrict water loss as weIl as a wide range of possibilities to make
use of even the lowest amount of suitable drin king water protect
the animal against such exeeptional eoneentrations.
 SUMMARY

I. This paper deals with various aspects of the life-history, ecology, water
management and osmoregulation of the West-Indian land hermit crab Coenobita
clypeatus (Herbst) in Cura~ao, Netherlands Antilles.

2. Land hermit crabs belonging to the family Coenobitidae may be considered as

one of the most terrestrial forms of decapods. They are characteristic for tropical
co asts and islands. Though C. clypeatus may be found in a variety of habitats
they show a preference for areas with a relatively dry climate. In this respect
habitats as found on the Leeward Group of the Lesser AntilIes are representative
for the species' occurrence. In addition to populations of animals living on the
coast, the 'coastal animals,' there are also individuals living in the interior, the
'inland animals.' The latter generally are older specimens, living in well-fitting and
undamaged Livona-shells, and able to setUe and maintain themselves in habitats
rich in food, where there is a supply of fresh or brackish water. Though these
animals still maintain a close bond with the sea, they may be said to have reached
a more advanced stage of terrestrial life. The greater part of the research was
carried out with inland animals.

3. In July reproductive migration starts, adult and not yet fully grown
animals migrating towards a restricted number of spawning places on the southern
coast, probably following traditional pathways. These places are characterized by
the presence of good shelter, suitable drinking water and a quiet and accessible
coast. Usually animals of roughly the same size gather in separate groups. It is
assumed that fertilization occurs here. Soon afterwards, usually around the time
of full moon, the first ovigerous females may be observed. Under the circumstances
prevailing on Cura~ao, the fresh red-brown eggs develop in about three weeks.
The eggs then contain a full-grown larva, in the first zoea stage, which is liberated
as soon as the egg is brought into the sea. In a number of consecutive nights the
females deposit clusters of ripe eggs at the low water line, from which clusters the
larvae are carried away by the rising tide. Apart from females with fully developed
eggs another, second, wave of animals with freshly laid eggs may be observed, to
be followed in some cases by a third wave. In this way groups of larvae in con-
 133

secutive cycles, are brought to the sea throughout the summer season. The repro-
ductive period lasts from July to about November on the Leeward Group of the
Lesser Anti1les. A striking feature is the difference in sex ratio between younger
and older animals, the relative number of females decreasing with increasing age.

4. Land hermit crabs are frequently exposed to strong evaporation, therefore a

good water management is of primary importance. There are many factors
contributing to this maintenance, such as adaptations in anatomy, way of life and
behaviour, the powers of detection and uptake of water, the mainly nocturnal
life, the possession of a shell which can be closed, the shell water, the urge to seek a
suitable miero-habitat, etc. Under constant environmental conditions dehydration
always proceeds along the same lines in the same animal. At a temperature of 28°C
and a relative air humidity of about 75%, which for Cura~ao are the normal con-
ditions, average survival was 8 days, together with water losses of maximally 30%
of the initial amount of body water. Dehydration always entails an increase in
osmotic value of the body fluids; uptake of water makes the concentration decrease
again.

5. In favourable conditions Coenobita clypeatus actively stores a great amount of

water as shell water. By applying salt accumulation, washing, mixing or
dilution, in relation to the salinity of the available drinking water, the animal
effects a rough regulation, in which the shell water is kept more or less constantly
at a salinity fluctuating around 32-33%0 S, which is slightly below the value of
normal sea water (= 36%0 S). As a rule small animals maintain a slightly higher
concentration of shell water than large animals. Air humidity also influences the
concentration of shell water in such a way that in aperiod of drought shell water
of a slightly lower concentration is stored. Generally speaking the time factor has a
stabilizing influence on the concentration. When in an experiment drinking water
of various concentrations is offered, there is a tendency to take up water of ever
decreasing salinities, while the total amounts are decreasing too.

6. Osmoregulation fails in land hermit crabs that do not possess an external

environment of water. Due to evaporation and consequently dehydration of the
animal the osmoconcentration may reach very extreme, lethai heights, equivalent
to over 60%0 S. Dehydrating hermit crabs may be compared to osmoconformers.
If, however, such an external environment is actually present, Coenobita proves to
be a fairly good regulator, trying to keep its body fluid as constant as possible,
both in hypotonie and hypertonie media. Generally the role of the external environ-
ment is played by the shell water, which serves as an intermediate environment
between the available water and the internal medium. As this shell water is already
regulated, the osmoregulation proper must be considered as superimposed on that
of the shell water. An optimal value for osmoconcentration of the body fluid
approaches an equivalent of 34-35%0 S. Non-electrolytes contribute greatly - up
to 30% - to the total concentration. Osmoregulation however is mainly brought
about by regulation of the electrolytes.
 134

RESUMEN

I.Esta tesis trata sobre varios aspectos de la vida, de la ecologia, el manejo de

agua y la regulaci6n osm6tica deI cangrejo ermitafio terrestre Coenobita clypeatus
(Herbst) en Curayao, Antillas Neerlandesas.

2. Los cangrejos ermitafios terrestres que pertenecen a la familia de Coenobitidae

pueden ser considerados corno una de las especies mas tipicamente terrestres de los
decapodos. Son caracteristicos de las costas y las islas tropicales. Aunque puedan
ser encontrados en una gran variedad de ambientes, los C. clypeatus, prefieren
regiones con un clima relativamente arido. EI am b i e n t e que form an las Islas
de Sotavento es muy apropiado para estos animales. Ademas de los ejemplares de
estos animales que viven en la costa, los llamados "animales costales," hay tambien
otros grupos que viven en el interior, grupos que llamaremos "animales deI in-
terior." Estos ultimos generalmente son ejemplares mas viejos que viven en bien
conservados caracoles Livona pica; viven en lugares donde hay abundancia de
alimentos y de agua dulce 0 salobre. Aunque existe todavia un fuerte lazo que une
estos ultimos animales con el mar, se puede decir que han alcanzado un alto nivel
de vida terrestre. Gran parte de nuestra investigaci6n fue llevada a cabo con los
animales que viven en el interior.

3. En el mes de julio empieza la migraci6n relacionada con la reproducci6n.

Para esta fecha tanto los animales adultos corno los que todavia no pu eden ser
considerados corno tales, migran a un numero determinado de lugares de la costa
meridional, probablemente por sendas ya conocidas. Estos parajes se caracterizan
por abundancia de cobijo, cantidad suficiente de agua potable y una costa tranquila
y accesible. Los animales que tienen mas 0 menos el mismo tamafio suelen agruparse.
Se supone que esta epoca coincide con la fertilizaci6n. Poco tiempo despues, gener-
almente alrededor de la primera luna nueva siguiente, se puede notar la presencia
de huevos en las hembras. En las circunstancias reinantes en Curayao, los huevos
de un color entre rojo y castafio, maduran dentro de unas tres semanas. Para
entonces, cada huevo contiene una larva madura en la primera fase zoea, que sale
deI huevo tan pronto corno este entra en contacto con el agua deI mar. Durante
algunas noches consecutivas las hembras, aprovechandose deI reflujo, ponen
racimos de huevos completamente desarollados en el fondo de las rocas. EI altamar
luego se encarga de llevar las larvas. Ademas de las hembras con huevos desarollados,
se puede observar un segundo grupo, con huevos nuevos, y hasta un tercer grupo.
De este modo sucesivos grupos de larvas en diversos ciclos de evoluci6n son depo-
sitados en el mar durante el verano. EI periodo reproductivo abarca los meses de
julio hasta noviembre, mas 0 menos, en las Islas de Sotavento. Un rasgo muy
sorprendente es la diferencia numerativa en sexo entre los animales j6venes y los
viejos. EI numero de hembras disminuye con la edad.

4. Los cangrejos ermitafios terrestres frecuentemente estan expuestos a una im-

portante deshidrataci6n. Por eso un buen man e jod e ag u a es de primordial
importancia. Hay muchos factores que contribuyen a ese manejo de agua, tales
como adaptaciones anat6micas, modo de vivir, comportamiento, la capacidad de
encontrar y de poder abastecerse de agua, una vida casi exclusivamente nocturna,
 135

un caracol que se puede cerrar, el agua que puede conservar en el caracol, la prefer-
encia por un micro habitat, etc. En circunstancias ambientales constantes la
deshidrataci6n siempre se efectua segun los mismos patrones, tratandose deI mismo
animal. A una temperatura de 28° centfgrados y una humedad relativa deI aire de
unos 75%, cosas muy comunes en Cura~ao, el promedio de supervivencia era de 8
dias con una perdida de agua de a 10 sumo 30% de la cantidad inicial. La deshi-
drataci6n siempre acarrea un incremento deI valor osm6tico de los liquidos cor-
porales. EI tomar agua hace que disminuya nuevamente la concentraci6n.

5. En circunstancias favorables, Coenobita clypeatus almacena una buena cantidad

de ag u a e n s u ca r a c 0 1. Acumulando sal, lavandose, mezclando 0 diluyendo
de acuerdo con la salinidad deI agua potable disponible, el animal efectua cierta
regulaci6n por medio de la cual el agua en el caracol se mantiene a una salinidad
constante de mas 0 menos 32 a 33%0 S, 10 cual es algo menos que la salinidad
normal deI agua deI mar (36%0 S). Normalmente los animales pequeiios mantienen
una concentraci6n algo mas alta en el agua que se encuentra dentro de su caracol,
que los mayores. La humedad deI aire tarn bien influye sobre la concentraci6n
acuosa en el caracol en tal modo que en periodos de sequida se almacena una
cantidad de agua de concentraci6n mas baja. En general se puede decir que el
factor tiempo ejerce una influencia estabilizadora en la concentraci6n. Cuando
durante un experimento, les ofrecimos a los animales agua potable de varias con-
centraciones decrecientes, disminuye al mismo tiempo la cantidad total de sales.

6. La re g u 1a c i 6 nos m 6 t i c a faHa en la vida de los cangrejos ermitaiios ter-

restres que carecen de medio ambiente adecuado. A causa de la evaporaci6n que
trae consigo la deshidrataci6n deI animal, la concentraci6n osm6tica puede aIcanzar
un grado letal sumamente alto que sobrepasa el 60%0 S. Se puede comparar los
cangrejos ermitaiios, que se estan deshidratando, con animales no-reguladores.
Si, por el contrario, se halla presente un medio ambiente favorable, los Coenobitas
resultan buenos reguladores que tratan de mantener el liquido en su cuerpo a un
nivel mas constante posible, tanto en un ambiente hipot6nico corno en el hiper-
t6nico.
Generalmente es el agua que se conserva en el caracolla que desempeiia el papel
de ambiente externo y sirve corno enlace entre el aqua disponible y el medio am-
biente interno. Cuando el aqua deI caracol ha sido regulada, la funci6n osm6tica
propiamente dicha debe ser considerada corno mas importante que el contenido
acuoso deI caracol. EI grado 6ptimo de regulaci6n osm6tica en el liquido corporal
es de una equivalencia aproximada entre 34 y 35%0 S. Los no-electr6litos contribuyen
con un 30% de la concentraci6n total. La regulaci6n osm6tica, sin embargo, es
ocasionada particularmente por la regulaci6n de los electr61itos.
 REFERENCES

AB BOTT, R. T., 1972. The Bermuda "Wilke." NewsleIter Bermuda Biol. Sta. 2, I: 2.
BLISS, D. E., 1956. Neurosecretion and the control of growth in a decapod crus-
tacean. In: WINGSTRAND, Bertil Hansträm, Zool. papers in honour 01 his
sixty-lilth birthday: 56-75. Lund.
BLISS, D. E., 1963. The pericardial sacs of terrestrial Brachyura. In: WHITTINGTON
& ROLFE, Phylogeny and evolution 01 Grustacea: 59-78. MCZ Harvard.

BLISS, D. E., 1968. Transition from water to land in decapod crustaceans. Am.
Zoologist 8: 355-392.
BLISS, D. E. & WANG, S. M. & MARTINEZ, E. A., 1966. Water balance in the land
crab, Gecarcinus lateralis, during the intermolt cyc1e. Am. Zoologist 6: 197-
212.
BOUSFIELD, E. L., 1968. (Discussion following BLISS, Transition to land in Deca-
pods). Am. Zoolologist 8: 393-395.
CARPENTER, A. & LOGAN, W. H., 1945. Hermits don't always live alone. Natural
History A .M.N.H. S4: 286-287.
CHASE Jr., F. A. & HOBBS Jr., H. H., 1969. The Ireshwater and terrestrial decapod
crustaceans 01 the West Indies with special relerence to Dominica. Bull. U.S.
N.M. 292, 258 pp.
EDNEY, E. B., 1957. The water relations 01 -terrestrial arthropods, 109 pp. Monogr.
Exp. Bio!., Cambr. Uno Press.
EDNEY, E. B., 1960. Terrestrial adaptations. In: WATERMAN, The physiology 01
Grustacea, I: 367-388. Acad. Press, N.Y. & London.
EMERY, A. R., 1972. Eddy formation from an oceanic island: ecological effects.
Gar. J. Sei. I2; 121-128.
FUGLISTER, F. C., 1960. Atlantic Ocean atlas 01 temperature and salinity ... , 209 pp.
Woods Hole.
 137

GRIMM, A. S., 1969. Osmotic and ionic regulation in the shrimps Crangon vulgaris
Fabricius and Crangon allmanni Kinahan, 88 pp. Thesis, Glasgow.
GROSS, W. J., 1955. Aspects of osmotic regulation in crabs showing the terrestrial
habit. Am. Naturalist 89: 205-222.
GROSS, W. J., 1957. An analysis of response to osmotic stress in selected decapod
Crustacea. Biol. Bult. II2: 43-62.
GROSS, W. J., 1964. Water balance in anomuran land crabs on a dry atoll. Biol.
Bult. u6: 54--68.
GRUBER, M. & SHOUP, J. B., 1969. Crabs move ashore. Sea Frontiers I5: 364--375.
HARMS, J. W., 1929. Die Realisation von Genen und die Consecutive Adaption, I.
Phasen in der Differenzierung der Anlagenkomplexe und die Frage der
Landtierwerdung. Z. Wiss. Zool. I33: 211-397.
HARMS, J. W., 1932. Die Realisation von Genen und die Consecutive Adaption, 11.
Birgus latro L. als Landkrebs und seine Beziehungen zu den Coenobiten.
Z. Wiss. Zool. I4o: 167-290.
HARMS, J. W., 1937. Lebenslauf und Stammesgeschichte des Birgus latro L. von den
Weihnachtsinseln. Z. Naturwiss. Jena7I: 1-34.
HAZLETT, B. A., 1966. Observations on the social behaviorof the land hermit crab,
Coenobita c1ypeatus (Herbst). Ecology 47: 316-317.
HAZLETT, B. A., 1966. Social behavior of the Paguridae and Diogenidae of Cura~ao.
Stud. fauna Curafao 23: 1-143.
HOHENDORF, K., 1963. Der Einflusz der Temperatur auf die Salzgehaltstoleranz
und Osmoregulation von Nereis diversicolor O.F. Muel!. Kieler Meeresf.
I9: 196--218.
KALBER, F. A. & COSTLOW, J. D., 1968. Osmoregulation in the larvae of the land-
crab, Cardisoma guanhumi Latreille. Am. Zoologist8: 411-416.
PALMER, J. D., 1971. Comparative studies of circadian locomotory rhythms in four
species of terrestrial crabs. Am. Midland Nat. 85: 97-107.
PEARSE, A. S., 1916. An account of the Crustacea collected by the Walker Ex-
pedition to Santa Marta, Colombia. Proc. U.S. Nat. Mus. 49: 531-556.
PEARSE, A. S., 1929. Observations on certain littoral and terrestrial animals at
Tortugas, Florida, with special reference to migrations from marine to
terrestrial habitats. Papers Tortugas Lab. 24: 205-223.
PROSSER, C. L. & BROWN, F. A., 1962. Comparative animal Physiology (2nd. ed.),
688 pp. Saunders, Phila. & London.
PROVENZANO Jr., A. J., 1959. The shallow-water hermit crabs of Florida. Bull.
Mar. Sei. Gul! & Carib. 9: 349-420.
PROVENZANO Jr., A. J., 1962. The larval development of the tropicalland hermit
Coenobita c1ypeatus (Herbst) in the laboratory. Crustaceana 4: 207-228.
PHILIPS, 1963. Universeie I2-kanalen recorder PR 32IO U/OO. [Service manual].
Eindhoven.
 138

SEURAT, L. G., 1904. Observations biologiques sur les cenobites (Coenobita perlata
Edwards). Bull. Mus. Rist. Nat. IO: 238-242.
SPAARGAREN, D. H., 1971. Aspects of the osmotic regulation in the shrimps Crangon
crangon and Crangon allmanni. Neth. J. Sea Res. 5: 275-335. Thesis, Amster-
dam.
Statistiek van de meteorologische waarnemingen van de Nederlandse Antillen. I963-
I968. Curayao.
STOFFERS, A. L., 1956. The vegetation 0/ the Netherlands Antilles. Pub!. Found. Sei.
Res. Sur. & N.A. I5: 142 pp. Thesis, Utrecht.
Terrestrial adaptations in Crustacea. [Symposium, 27-29.XII.1967]. Am. Zoologist
8, 1968: 307-685.
TUCKER ABBOTT, R, 1972. The Bermuda "Wilke." Newsletter Bermuda Biol. Sta.
2, I: 2.

VERRILL, A. E., 1908. Decapod Crustacea of Bermuda. 1. Brachyura and Anomura.

Trans. Conn. Acad. Sei. I3: 299-474.
VERWEY, ]., 1957. A plea for the study of temperature influence on osmotic regu-
lation. Ann. Biol. 33: 129-149.
VÖLKER, L., 1965. Experimentelle Untersuchungen zur Ökologie des Landeinsied-
lerkrebses Coenobita scaevola Forskdl am Roten Meer, 115 pp. typescript.
Thesis, Techn. Hochschule Darmstadt.
VÖLKER, L., 1967. Zur Gehäusewahl des Land-einsiedlerkrebses Coenobita scaevola
Fors~ä.1 vom Roten Meer. J. exp. mar. Biol. Ecol. I: 168-190.

WIENS, H. J., 1962. Atoll environment and ecology, 532 pp. Yale Univ. Press.
WILDE, P. A. W. J. DE, 1969. Research on terrestrial crabs in Curayao, Netherlands
Antilles. WOT RO report jor the year I968: 38-42. The Hague.
WILDE, P. A. W. J. DE, 1972. Soldaatjeskrabben. Stinapa 6: 37-47. Curayao.
WOLVEKAMP, H. P. & WATERMAN, T. H., 1960. Respiration. In: WATERMAN, The
Physiology oj Crustacea I: 35-100. Acad. Press N.Y. & London.
WÜST, G., 1964. Stratijication and circulation in the A ntillen-Caribbean basins
(Part 1), 199 pp. Columbia Univ. Press.
 SAMEN V A TTING

1. Deze publicatie gaat over verschillende aspecten van de I e v e n s g e-

schiedenis, oecologie, waterhuishouding en osmoregulatie van de West-
indische Landheremietkrab Coenobita clypeatus (Herbst) op Curar,:ao, Neder-
landse Antillen.
2. Landheremietkrabben, behorend tot de familie der Coenobiti dae, moe-
ten tot de meest terrestrische vormen onder de Decapoden gerekend worden.
Het zijn kenmerkende dieren van tropische kusten en eilanden. Ofschoon
C. clypeatus in een verscheidenheid van I a n d s c h a p p e n aangetroffen kan
worden, hebben woongebieden met een relatief droog klimaat de voorkeur.
Habitats als op de Benedenwindse Eilanden kunnen in dit opzieht represen-
tatief genoemd worden. Naast populaties van langs de kust levende dieren,
"coastal animals," worden meer landinwaarts de z.g. "inland animals" on-
derscheiden. Over het algemeen bestaat deze laatste groep uit oudere exem-
plaren in goed passende en gave schelpen van de soort Livona pi ca, die zieh in
de voedselrijke habitats, met zoet of brak water, hebben kunnen vestigen en
handhaven. Niettegenstaande ook deze dieren een nauwe binding met de zee
onderhouden, worden zij beschouwd een geavanceerder stadium van land-
leven bereikt te hebben. Het meeste onderzoek werd aan "inland animals"
verricht.
3. In juli begint de v 0 0 r t p I a n tin g s-m i g r a ti e, waarbij volwassen en
onvolgroeide dieren - vermoedelijk langs traditioneie wegen - naar een be-
perkt aantal paaiplaatsen aan de zuidkust trekken. Deze plaatsen worden
gekenmerkt dOOf de aanwezigheid van goede schuilplaatsen, geschikt drink-
water en een rustig en accessibel kustgedeelte. Veelal verzamelen dieren van
ongeveer een zelfde grootte-klasse zieh op afzonderlijke plaatsen. Aangeno-
men wordt dat hier de bevruchting plaats vindt. Spoedig daarna, meestal
omstreeks volle maan, kunnen de eerste vrouwtjes met jonge, roodbruine
eieren, vastgehecht aan de pleopoden van het abdqmen, waargenomen wor-
den. Onder de op Curar,:ao vOOfkomende omstandigheden ontwikkelen de
eieren zieh in ongeveer drie weken tot een volgroeide larve in het eerste zoea
stadium. Klompjes rijpe eieren worden in een aantal opeenvolgende nachten
dOOf de vrouwtjes op de laagwaterlijn verspreid. In aanraking met water
barsten de eieren terstond open; de larven worden met het vloedwater mee-
gevoerd. Naast vrouwtjes met geheel ontwikkelde eieren wordt nu opnieuw
een tweede groep dieren met pas gelegde eieren waargenomen, die op zijn
beurt dOOf een derde groep gevolgd kan worden. In de loop van het voort-
plantingsseizoen worden aldus in een aantal opeenvolgende golven groepen
larven aan de zee toevertrouwd. De voortplantingsperiode loopt op de Be-
nedenwindse Eilanden van juli tot omstreeks november. Opvallend zijn de
verschillende verhoudingen in sexe tussen jonge en oude dieren, zodanig dat
het aantal vrouwtjes bij oudere dieren relatief sterk afneemt.
4. Landheremieten staan voortdurend aan sterke evaporatie bloot, zodat
een goede w a t e r h u i s h 0 u d i n g van primaire betekenis iso Tal van adap-
taties in bouw, levenswijze en gedrag, de fijne detectie- en opnamemogelijk-
heden van water, de overwegend nachtelijke levenswijze, het bezit van een
goed afsluitbare schelp, het schelpwater, en het opzoeken van de juiste miero-
 140

habitat leveren hiertoe een belangrijke bijdrage. Onder gelijke omstandig-

heden verloopt het waterverlies bij een bepaald dier steeds op dezelfde wijze.
Bij een temperatuur van 28°C en een relatieve luchtvochtigheid van onge-
veer 75%, dus onder voor Cura9ao normale omstandigheden, werden gemid-
delde overlevingstijden van 8 dagen gemeten, hetgeen gepaard ging met wa-
terverliezen van maximaal 30% van de initiele hoeveelheden lichaamswater.
Dehydratie brengt steeds een stijging van de osmotische waarde van de li-
chaamsvloeistoffen met zich mee; opname van water doet de concentratie
weer dalen.
5. Coenobita clypeatus slaat on der gunstige omstandigheden actief grote
hoeveelheden water als sc hel p wa t e r op. Door in afhankelijkheid van het
zoutgehalte vanhet beschikbare drinkwater op adequate wijze zoutaccumu-
latie, doorspoeling, menging of verdunning toe te pas·sen, geraakt het dier tot
een grove regulatie, waarbij de saliniteit van het schelpwater min of meer
constant gehouden wordt op een waarde schommelend rond 32-33 %0 S, dus
iets beneden de waarde van het normale zeewater (= 36 %0 S). Kleine dieren
onderhouden doorgaans een wat hogere concentratie van het schelpwater dan
grote dieren. De vochtigheid van de lucht blijkt eveneens van invloed op het
schelpwater te zijn, zodanig dat in een drogere omgeving water van een lagere
salin,iteit vastgehouden wordt. In het algemeen heeft de factor tijd een stabili-
serende invloed op de concentratie. Wanneer onder experimenteie omstan-
digheden drinkwater van verschillende concentratie aangeboden wordt, be-
staat de tendens om water van steeds lager zoutgehalte op te nemen, waarbij
de totaal opgenomen hoeveelheden afnemen.
6. Bij landheremietkrabben, die niet beschikken over een uitwendig wate-
rig milieu, faalt de 0 sm 0 r e g ul a t i e, waardoor tengevolge van evaporatie
en de da armee gepaard gaande dehydratie van het dier de osmoconcentratie
tot zeer extreme, letale waarden, equivalent met ruim 60 %os, kan stijgen. Uit-
drogende heremieten knnnen met osmoconformers vergeleken worden. Wan-
neer ech~r een dergelijk extern milieu wel aanwezig is, manifesteren Coeno-
bitas zich als tamelijk goede regulators, die er naar streven de concentratie
van de lichaamsvloeistof zo constant mogelijk te houden, zowel in hypo- als
in hyper~onische media.
Veelal vervult het schelpwater de rol van het externe milieu, waardoor het
als intermediair milieu tussen het beschikbare drinkwater en het interne mi-
lieu geschakeld wordt. Gezien de regulatie van het schelpwater, moet de
eigenlijke osmoregulatie van het dier als gesuperponeerd op die van het
schelpwater gedacht worden. Een optimum waarde voor de osmoconcentratie
van de lichaamsvloeistof ligt bij een equivalent van 34-35 %0 S. Niet-electro-
lieten leveren een aanzienlijke bijdrage - tot 30% - aan de totale concentra-
tie. Osmoregulatie geschiedt evenwel hoofdzakelijk door regulatie van de
electrolieten.
 PLATE I

la. 'Oostpunt' estate on Curac;:ao, as seen from the East towards the limestone
escarpments of Duivelsklip (62 m, left) and Tafelberg (196 m, centre). A heavily
eraded area with an open vegetation consisting mainly of thorny shrubs, with
cactuses, woody herbs, and small trees . (Phot. P. Wagen aar Hummelinck, 28.X.
1963) .

Ib. A large specimen of Coenobita clypeatus (Herbst) peeping fram its Livona-shell.
(Phot. M. Arnoldo Broeders).
PLATE H

Ha. Weil 1: A characteristic coastal habitat of Coenobita, showing a concrete water

through fed by a wind mill, at the base of a limestone escarpment on the north
coast of East-Cura~ao. The area is barren, and only scantily covered with Opuntia
wentiana and low shrubs.

!Ib. Punta Tera: A characteristic inland habitat of Coenobita, showing two concrete
water basins fed by a wind mill. The area is covered with shady trees, incl. Acacia
tortuosa, Hippomane mancinella and Swietenia mahogany.
 PLATE IIl

lIla. Aerial view of the Lagoon area along the south coast of East-Curayao, showing
part of Lagoon A (top-left), Lagoon B, and Lagoon C. Lagoon Bis separated by the
Landbridge (arrow) from Lagoon C. The semi-circular mudbank in Lagoon Cis the
so-called Delta.

IIIb. The Landbridge separating Lagoons Band C as seen towards the East. Note
the desolate surroundings, with coral stone boulders, and a few bushes of Conocarpus
erecta which represent the main hiding places for Coenobita .
PLATE IV

IVa. Coenobita clypeatus in shell of Livona pica, crawling around with extended
antennae.

IVb. Coenobita clypeatus, rernoved frorn its shell.

 PLATE V

Va. Male Coenobita clypeatus showing sperrnatophores protruding frorn the paired
gonopores.

Vb. Fernale Coenobita clypeatus, one of thern with egg clusters attached to the hairs
of the pleiopods.
PLATE VI

VIa. A most unsuitable type of 'shell', accepted by a Coenobita in a bungalow at

Savaneta, Aruba. (Courtesy Lago Oil & Transport Co. Ltd., 1961).

VIb. A glass-shell, accepted by a female Coenobita during experiments.

 STELLINGEN

1. Heremietkrabben uit de Familie der Coenobitidae kunnen beschouwd worden als

de voorlopers van een toekomstige succesvolle groep landdieren
Dit proefschrift.

2. Transpiratie als een mechanisme voor temperatuursregulatie bij landkrabben

moet als weinig zinvol beschouwd worden.
D. Bliss (1968), Am. Zool. 8: 364-368.

3. De voortplantingsmigratie van Coenobita c1ypeatus gaat langs traditioneie routes,

waarbij jonge en onervaren dieren door oude dieren geleid worden.
Dit proefschrift.

4. De combinatie van benthische micro-algen, detritus en bacterii~n vormt het

voedsel van deposit feeders in de estuariene gebieden.
R. C. Newell (1973), Biology of intertidal animals. Paul Elek Sc. Books,
Londen: 238-263.

5. Bij de meting van het energieverbruik door marie ne organismen in het kader van
productieonderzoek wordt onvoldoende rekening gehouden met het gedrag en de
levenswijze van de die ren.
P. A. W. J. de Wilde (1973), Neth. J. Sea Res. 6: 157-162.

6. De resultaten van de experimenten van Hesz vinden met het exosomenmodel van
Fox en Y oon (1970) een redelijke verklaring.
D. Hesz (1972), Naturwissenschaften 59: 348-355.

7. Het CARMABI moet in de eerste plaats als een instituut voor oecologisch
veldonderzoek gezien worden.

8. Er dient rekening mee gehouden te worden dat de veranderingen in soorten-

samenstelling en talrijkheid van evertebraten in Noord- en Waddenzee toege-
schreven kunnen worden aan andere dan natuurlijke oorzaken.

9. In het Nederlandse deel van de Waddenzee is plaats voor een zeehondenbestand

van ongeveer 2000 dieren.

10. De wettelijk toegestane raaptijd van kievitseieren moet ingekort worden tot 7
april; die van andere weidevogels dient geheel gesloten te worden.

11. In verband met ontoelaatbare geluidsoverlast en hinder moeten zo spoedig

mogelijk afdoende maatregelen getroffen worden om de snelle expansie van de
z.g. kleine luchtvaart boven het Waddengebied in te perken.

12. Nimrod is dood.

P. A. W. J. de Wilde, november 1973