Full text of "The biology, ecology and mariculture of Mithrax spinosissimus, utilizing cultured algal turfs"

See other formats

a: > s Xiftosqy

a n_LI B RAR I ES "SMITHSONIAN INSTITUTION " NOlinillSNI NVINOSHIIWS^SB I BVB 8 II^LI B RAR I ES^SMITHSONIAf

~ 5 _ _ w , — ^ 2 \ CO _ _ _ = oo

2 _

ION NOliniliSNI NVINOSHIIIAIS S3 I B VB 8 13 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IW!

- w ro 2 . ^ « J&c-

CD
33
>

m ■wc XJ ro 'viissa _ m '<C0L2i>'' j" Xcc/uiis^JZ m

co \ _ to _ co _ co.

an LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIIAIS S3IBVHai3 LIBRARIES SMITHSONIA

Z co 2 ,v* co 2 _ _ _ CO 2

2 ,>W\ 1 — I 2 ../ / 2 Z&fWW'A -I . ,v

X .

2 x > '^oii^y' 2 s X >

CO 2 - 00 *2 CO 2 CO *2

ION NOlinillSNI NVIN0SH1I/JS S3IBVBai3 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IW

^ _ —? . v . tO _ ^ _ ^ 2! _ - — ^ ^ _

n - w ^ co “ ,• „ «<* ^

dl < 1st as) h n < 9s) z* (sfe 9s < >/

O — NfQiass^ O pC^X “ O pC

BIl^LIBRARI E S 2 SMITHSON IAN-* INSTITUTION NOlinillSNI ""NVINOSHIMS SaiBVBail^LIBRARI ES^ SMITHSONIA

I j/hL 5 I 2 I I 1 4*^

03
33
>

m x^«os'v^2 'LL rn m ^ m

co fE co _ co — co

ION NOlinillSNI NVIN0SH1IWS S3IBVBai3 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IW

CO 2 co 2 s CO 2 » CO 2

!•* ‘ * jg '<S^P' |- 5 | xCkjssX g v^* |.* ^ ^ g

a II2 LI B RAR I ES^SMITHSONIAN INSTITUTION NOlinillSNI _N\/INOSHllWs‘/’s 3 I d VB a IT^LI B RAR I ES^SMITHSONIA

~ CO “ co _ _ 2 \ CO — CO

Q x^yosv^y^ — v»^yA^>2>y — V>Ai>^ q

2 -I 2 -I 2

'ION NOlinillSNI NVINOSHIIIAIS S3IBVB3n LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IW
2 I — 2 _ z f- ^ 2 r-

O — .Ss.XX O 2Tu4VaT\ . X^VCSO/^X o _ o ..

<*y Xh'yt&kSr^J ^ co m SJ rn Vgvoc^y xioiu^x m

CO*-. ^ co — CO — co

an LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3IBVB8IT LIBRARIES SMITHSONIA

2 .... CO 2 ,v CO 2 _ _ CO 2 .... CO

v ^ ^ 2 2 x- -

* " z ^ -C 2 H , ><>/ 2 -* A" 2 MVx -C

o x o |#\ x . .//A o 2 . '/m

CO CO 00 2 CO CO "*

\ ° ^1/ X o X

KX 2 \-tV,afe^7 t— > m 2 t v>-«r« — X42SW — wx —

5 v'-, > 5 Xo^DC^/ 5> 2 \ >

riON^NOlinillSNI NVINOSHIIIAIS^SS I BVB a H LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IW

^ ~ .,v . tO _ 2; _ ^ 2! _ ^ _ 2

A\ ■ . i ><jcni>v lil ><u^O*7N. Itl yziCviTX *•

Q DL^ _ Q

ail LI B RAR I E S~ SMITHSONIAN- INSTITUTION NOlinillSNI ^NVINOSHIIIAIS ^S 3 IBVBail^LIBRARI ES^ SMITHSONIA

2 _ 1- 2

O Z /

/o'

33 /S
^ > P
X) W

— vv.- .

m \V^osi\v

MWIKIDCU I IIAJC 53 I JlWHfl 11

m > w pn

co ' — co

IIRRARIF^ RMITMQnwiAM I MRTITI ITIHW KinimillQWI

MWIWORH I I IAJ

s ■ VP > 2 •• 2 XCftos*12 > 2 VP > x^^i:

co ** Z co y“ 2 co z co 2

ITUTION NOIinillSNI_ NVIN0SH1IIAIS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI _ NVINOSh

(O _ \ to ^ _ _ CO ^ _ CO _ _ _ ^ c

H W
- VJ
O

~7 I

avaan libraries Smithsonian institution NoiiniusNi nvinoshiiws saiavaan libraries smithsc

i S .,,,/ 1 2 ^ I /^VT^N 2 /gm&i, l ,,,/

m xc.1* 0x2 zs m x^u gs^x m vt »; xcuusj^ rn

co _ co _ to X E co

ITUTION NOlinillSNI NVIN0SH1IIMS S3ldVaan LIBRARIES SMITHSONIAN INSTITUTION NVINOSI

co 2 _ to 2 ... CO Z .v. co 2

> ' 2 Xijscs^ > '<pp~ 2 '\ 2» X^OiuM^ 2 J>

2 CO 2 CO ‘.2 - Jo k 2

avasn libraries Smithsonian institution NoiiniusNi nvinoshiiiais S3iavaan libraries smithsc

co 5 _ _ _ CO _ — ,v co 5 _ _ _ co

Ul fk LjJ X JiSSA UJ 2^*X 7k UJ

fj» *

WsC#

_ _ _ o

ITUTION NOlinillSNI^NVINOSHlIlMS S3 1 a Vd a n ^Ll B RAR I ES^ SMITHSON IAN-1 IN ST ITUTION NOlinillSNI “NVINOSI-
2 r~ , 2 r- 2 r- 2 r~ .

-A CD

^^4 'Xa y,

— VvJCHJW

m xqvqsvpx ^ pi XCftos*aX "L‘ xjvAi.»^x m

co “ co E to £ co

avaan libraries Smithsonian institution NoiiniusNi nvinoshhi/ms saiavaan libraries smithsc

E £ - - - 2 CO 2 CO 2 w

w §t i < /4^x 2 sMy,. < ^ ?

5 SJr > 5 XV > 2 Xi^iX > 5

co 2 co •'* ..2 co 2 co

ITUTION NOlinillSNI NVIN0SH1IINS S3ldVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSI

,n — •- in - (/) — "■> ~

avaan libraries Smithsonian institution NoiiniusNi nvinoshiiims shi avaan libraries smithsc

2 f~ _ 2 1-2-1-2

CD
73
>

m \ji QX2 " pi ~ m z pi

ITUTION ^ NOlinillSNI- NVIN0SH1I IMS S3 I d Vd 8 n~LI B R AR I ES ^SMITHSONIAN “INSTITUTION ^ NOlinillSNI NVINOSI

,v CO 2 5 X-- _ _ _ 2 ^ CO

2 AfX-

avaan libraries Smithsonian institution NoiiniusNi nvinoshiums SBiavaan libraries smithsc

CO 7Z _ CO _ — ... CO — CO

o X^M PCX ~ 'W* O Xl* 0X2 _ XjOj 1X^2 Q NX. — XOlmS^X o

ITUTION NOlinillSNI “JNVIN0SH1IIMS SaidVdaiT^LIBRAR I E S^ SMITHSONIAN“‘lNSTITUTION NOlinillSNI ^NVINOSE
2 r-,2 r- 2 <2 z c~

O 2^vtItI7X _ ,w O ~ 2f"i>ov5x O m ° /5otS^

33 /i/

> p(

— V

PI ^ pi ^ PI

CO — CO E CO E CO

avaan libraries smithsdnian institution wniiniiisKii nwinoshiiiais siiavaan libraries smithsc

The Biology, Ecology and
Mariculture of Mithrax spinosissimus
Utilizing Cultured Algal Turfs

Walter H. Adey
Marine Systems Laboratory
Smithsonian Institution

Washington, D.C.

Dennis F. Farrier
The Mariculture Institute
Smithsonian Institute
Washington, D.C.

January 1989

THE MARICULTURE INSTITUTE

P. O. Box 84136
Los Angeles, California 90073
United States

No part of this book may be reproduced in anyway or by any means
without permission in writing from The Mariculture Institute.

Printed in the United States of America.

TABLE OF CONTENTS

PREFACE

INTRODUCTION

ACKNOWLEDGEMENTS

SECTION I. ALGAL TURF PRODUCTION

Harvest Production of Coral Reef Algal Turfs. W.H. Adey
and J . M . Hackney . . . . . . .

SECTION II. BIOLOGY AND ECOLOGY OF MITHRAX SPINOSISSIMUS

Morphology and Relative Growth of Mithrax spinosissimus .

A. H. Biddlecomb, M.P. Craig and J.M. Iglehart .............

Population Density and Structure of M^. spinosissimus at Six
Caribbean Sites and The Florida Keys. J.M. Iglehart, R.V.
Ruark and K.H. Koltes . . . . .

Reproductive Biology, Fecundity and Embryonic Development of

M. spinosissimus . M.P. Craig, K.L. Porter and

R.V. Ruark . . .

Post-Larval Growth of Cultured M^ spinosissimus . K.L. Porter
J.M. Iglehart, R.V. Ruark, M.P. Craig, A. H . Biddlecomb, D.M.
Baudler and W.H. Adey . . . . .

SECTION III. MARICULTURE TECHNIQUES

Mariculture Techniques for Mithrax spinosissimus including
Broodstock, Larval and Post-Larval Rearing and Growout.
K.L. Porter, J.M. Iglehart, M.P. Craig and W.H. Adey ....

PREFACE

This publication consists primarily of an integrated set of
papers treating the efforts of the Marine Systems Laboratory from
1983-1986 to develop a low cost, itinerant fishermen mariculture
for Mithrax spinosissimus , the West Indian Red Spider Crab or
Caribbean King Crab. Efforts to develop such a mariculture
derive directly from our increasing understanding of the
potential for the utilization of cultured marine algae in coastal
Caribbean waters. A discussion of the biology and culture
methods developed for algal turfs, as well as methods for feeding
this algal production to Mithrax spinosissimus crabs, is also
presented along with the biology, ecology and mariculture of the
crab itself.

For many decades, it has been accepted that primary
production on shallow water tropical coral reefs is among the
highest of any natural community on earth. Indeed, at a level of

— 9 — 1

5-25 g dry m d , it generally exceeds that of the best
agriculture (Lewis, 1977; Adey and Steneck, 1985). Since most
coral reefs occur in seas with extraordinarily low nutrient
concentrations, it has generally been assumed that rapid
recycling processes, or perhaps local input of nitrogen and
phosphorus is occurring to allow such high levels of production
(Adey, 1987) . Unfortunately, given both the recycling
requirement, as well as the sensitivity of coral reef communities
to perturbations of any nature, including fishing and elevated

nutrients, it has seemed unlikely to most reef scientists and
mariculturists that high reef production was of more than
academic interest.

During the past decade, a series of laboratory studies, with
coral reef microcosms, and field studies, using a variety of
techniques, have demonstrated that the source of high primary
production is in the dense algal turf mat that covers a major
part of most shallow water reef substratum. These studies have
also shown that elevated production levels are real (i.e., not an
artifact of methodology) , can be harvested and under the high
physical energy environmental conditions common to coral reefs,
are indefinitely sustainable under low nutrient conditions (Adey,
1982, 1983, 1987; Adey and Steneck, 1985; Adey and Goertemiller ,

1987). Algal turfs are highly diverse (20-30 species on a single
square meter screen), and a typical, cultured community includes
taxa of most major algal groups. Polyculture harvest of algal
turfs is appealing as a potential mariculture because it avoids
the built-in disadvantages (particularly inherent instability) of
monocultural agriculture (Doyle, 1985).

The possibility of a mariculture based on the relatively
simple and low cost artificial substratum-grown algal turf
technology was quite interesting in terms of potential
adaptability to local populations in third world, tropical
countries. During early 1983, support from the Bioenergy

Council, Earthwatch and the United States Department of State for
the potential development of a low technology mariculture in
Caribbean waters led to the Agency for International
Development's (AID) interest in developing a low cost algal turf
mariculture .

Earlier Marine Systems Laboratory microcosm and field work
had identified several turf grazers within reef environments
which could have mariculture potential. One of these, Mithrax
spinosissimus (the West Indian Red Spider Crab or Caribbean King
Crab), a large and tasty animal, is a gourmet fishery item of low
volume interest in several Caribbean and South American
countries. It has been the subject of some previous research
(Munro, 1974; Hazlett and Rittschoff 1975; Bohnsack, 1976;
Brownell et al. , 1977; and Provenzano and Brownell, 1977). While

all of these authors had investigated Mithrax spinosissimus for
potential fishery or mariculture utilization, the general
conclusion was that the animal was not abundant enough to form a
fishery and grew too slowly to have viable mariculture potential.
Based on Marine Systems Laboratory preliminary research in both
the field and laboratory, JL spinosissimus was deemed hardy, both
in the wild and in microcosm culture, with growth rates exceeding
those presented in the literature (Bohnsack, 1976) . With regard
to Bohnsack' s studies, and MSL follow up studies partially
described herein, there was and still remains some concern that
the Florida population is a separate species or sub-species.

In 1983, the Marine Systems Laboratory submitted a proposal
to AID and was eventually funded by that agency, specifically to
attempt to develop a low cost, itinerant fishermen mariculture in
several Caribbean countries.

This treatise summarizes the information available to date
regarding this species and its preferred food, algal turfs. It
also presents the results of a three-year effort in a variety of
Caribbean locations and environments on the biology, ecology and
mariculture of the Caribbean King Crab. The approaches presented
here are not the only possibilities for a mariculture of this
animal. We specifically avoided any extensive efforts to develop
a mariculture that appeared to have limited value for an
itinerant fishermen mariculture or did not utilize algal turfs as
a food source. Commercial modifications to the methods described
herein would appear to have considerable potential.

Mithrax spinosissimus is a large spider crab that is
primarily a herbivore. In the wild, it feeds mostly on algal
turfs and, like a number of similarly adapted crabs, has
specially-developed claws for handling these small algae (Coen,
1987). It also posesses specialized grinding teeth in a
"gizzard" or anterior stomach. In addition, larval rearing is
brief (40-72 hours to settling at megalops and 70-90 hours to
first crab) and relatively simple (planktonic feeding regimes and
required conditions for such feeding appear to be unnecessary) .

As adults these crabs also have a harem-type social structure in
which excessive aggressive behavior is generally avoided. It is
relatively docile in character. Its primary defenses lie in
nocturnal behavior, an ability to cling tightly to the substratum
and its large adult size. The latter two characteristics provide
strong and extensive body musculature ( i . e . , considerable body
meat) . When crowded, juvenile Mithrax crabs sometimes kill many
of their cohorts, perhaps in response to a common tendency for
heavy local settlement when conditions for propagation are ideal.
This may provide an area of concern for mariculture. On the
other hand, fecundity is very high, each female brooding about
300,000 eggs during her reproductive life.

In culture, growth rates for It spinosissimus are relatively
high (12-15 months to maturity). Considering the highly cryptic
nature of immatures of the species, it would appear that rapid
growth is to a large extent genetically determined and achieves a
major species survival element in the large size of the adults.

M. spinosissimus populations vary widely from locality to
locality, suggesting minimum gene transfer and a relative
plasticity in a shallow water organism with a very short
planktonic phase. Since the animal breeds easily in captivity,
stock improvement with breeding seems assured.

All of these characteristics together suggest considerable
mariculture potential without the great difficulties that
accompany culture in most crabs and lobsters.

INTRODUCTION

The basis for the Mithrax spinosissimus (Caribbean King
Crab) mariculture described in this set of papers lies in the
screen culture of algal turfs. There is little question that the

production of a highly diverse algal turf, at a level of 8-18 g

— —1 , . .
dry m d , is routinely possible at a wide variety of Caribbean

sites at which Mithrax crab mariculture can be practiced. Algal

production is affected by the characteristics of the screen

material and by the way in which it is oriented in the water

column. It is affected strongly by turbulence (the strength of

ambient current and surge) , depth, harvest rate and continuity as

well as atmospheric cloudiness. These factors are now reasonably

well understood in field practice by MSL. The role of carbonate

sedimentation in limiting algal turf production is now better

understood and managed (in terms of site selection and harvest

criticality) on more of a qualitative basis. A more quantitative

understanding of the role of sedimentation in limiting algal turf

culture is needed.

Carbon levels appear low (15-20%) in algal turfs cultured in
open Caribbean waters and nitrogen levels (6-8%) are relatively
high. Low carbon content is also a characteristic of many of the
individual species that make up the turf. Blue-green algae have
largely proteinaceous walls and diatoms have siliceous frustules.
Both of these groups are major turf components. On the other

hand, laboratory cultured turfs with few diatoms have relatively
high levels of carbon (28-30%) . Carbonate siltation is certainly
a factor in low organic carbon, though, it does not explain the
low carbon levels occurring in samples of individual plant
components. Preparation and preservation procedures for these
"volatile" plants is strongly suspect and should be further
examined .

In any case, algal turfs are highly productive, rich in
protein and generally lack protective "toxic" compounds and the
tough walls and skeletons that characterize many macroalgae.

They also appear to be preferred by Mithrax spinosissimus . as
well as a wide spectrum of vertebrate and invertebrate grazers.
The high growth rates consistently found in culture and given
below for juvenile Mithrax spinosissimus are based primarily on
algal turf feeding. Work currently underway at Grand Turk using
pure algal turf cultures have given similar high growth rates
that point to to an adult crab in 12-15 months.

In order for this new form of algal production to be used as
fodder, an efficient transfer method from algal culture to grazer
culture is essential, regardless of the absolute level of plant
production available. The system that we have devised and
describe in detail in this set of papers averages a transfer rate
of about 5-8 screens (0.5 m2 each, cultured for 12 days) every
three to four days for 50 older juvenile crabs (greater than 30
mm) . This process is labor intensive; however, the development

of a "low tech," low capital cost mariculture was a primary
objective of this project. Considerable improvements can be made
in the transfer process. For example, the current transfer
methods can certainly be improved greatly through the mass
movement of screens using specially-designed small outboard-
powered boats (catamarans, e.g.) equiped with simple power
winches such as those that are used widely in lobster fisheries.

Mithrax spinas issimus is a large, tasty, meat-rich animal
that already forms the basis of a small gourmet fishery in
several Caribbean countries. At most Caribbean sites for which
we have adequate population data, average adult males exceed 1.5
kg in weight and average adult females exceed 1 kg in weight.
Antigua crabs are a little smaller and Florida crabs are much
smaller. Some negative comment has been received by MSL from
North American crustacean marketing specialists with regard to
the color, thickness and encrusted nature of the carapace, which
they feel may prevent it from being an "ideal" or top value
restaurant product. However, these observations refer to the
Florida collected crabs that are more greenish brown in color
(and are a much smaller animal) . They also refer to an adult
animal that is on the average several months post-molt. Cultured
Caribbean crabs are quite red in color and have rather thin
shells. Since, in culture, they would be harvested relatively
soon following a molt, they would lack encrustations.

In the Caribbean/West Indian region, wild Mithrax

spinosissimus crab populations, rather uniform at each site vary
widely from island to island. At most locales, the collection
process has been extensive and varied. It is highly unlikely
that the population characteristics derive from sampling bias.
This variation most likely results from their short larval time
and therefore limited gene transfer from pool to pool. Florida
"Mithrax spinosissimus . 11 somewhat isolated from the Caribbean
populations and subject to very different yearly weather and mean
habitat conditions, is much smaller than its Caribbean
counterparts. It differs more subtly in a wide variety of ways
and may be a different species or subspecies. In any case,
because breeding is so easily accomplished with Mithrax
spinosissimus . the possibility of considerable stock
"improvement" in the future certainly exists.

Mithrax spinosissimus populations tend to have a weak
"harem" type of social structure, with scattered or even sparsely
distributed "colonies" of a few to many females accompanied by
several males. These animals are nocturnal and tend to remain
close to their home crevice or cave. "Bachelor" males wander
more widely and, while still crevice and nocturnally oriented,
they are frequently found alone on open pavement. Our
observations on hundreds of adults from several sites show that
this species as adults accomodates well to cage culture. While
some "jostling" for space and male competition for females

occurs, it seems clear that adult intra-species aggression is not
an important mortality consideration. There appears to be
juvenile mortality during molts resulting from crowded conditions
in cages and we are now in the process of developing shore based
experimental systems at Grand Turk, so as to allow careful
observation of growing dense juvenile populations.

Mithrax spinosissimus mate in the hard shell state, after
the molt to maturity. It seems likely that multiple
fertilizations, as well as routine sperm storage, do occur. In
captivity, an average female produces four to five broods with a
mean of 60,000 eggs each. The female is ovigerous for about 30
days, producing a brood about every two months. Experience with
hundreds of animals from several sites has demonstrated that with
20-25 adult crabs, roughly 70-80% female, a hatch per week can be
routinely achieved and easily managed in a cage culture system.
The methods for accomplishing this process are described in
detail in the text.

Many thousands of hatchlings have been raised in cage
culture in this project to 100-120 days of age in a standard two-
step process that is described in detail. Most recently, 200-300
animals per hatch were routinely produced at 120 days, in a
procedure that at Buen Hombre , Dominican Republic, was easily
understood and managed by the local fishermen working with the
project. Considering the ease by which this can be accomplished,

this is certainly sufficient at this time for a successful
itinerant fishermen mariculture . In any case, steady improvement
in the survival rate has been attained and we predict that 500-
600 animals per hatch at 120 days of age can be achieved for this
crab, with present methods. In the text, we discuss several
problems and the steps by which this improvement can be achieved.
For a commercial mariculture or a government supported fisherman
mariculture, a hatchery process can probably be developed which
would improve considerably 100-day survivorship.

In the last six months of the project, we routinely achieved
growth rates of about 0.21 mm/day (carapace length) in a growout
from first crab through seven to nine instars at 100 days. From
that time to about 300 days, mean growth rates of about 0.35 mm
per day were achieved through another six to seven instars.

These crabs were fed primarily algal turfs with an occasional
supplement of a variety of macroalgae. Recent work at Grand Turk
has achieved the same growth rates with a diet of pure algal
turf. We have little statistically significant data from that
point on, but these rates strongly indicate that for a marketable
adult of 120 mm carapace length, 360-500 days would be required.
We believe that growout to a marketable adult can be routinely
achieved in an itinerant fishermen cage culture in about 400
days, though this has yet to be demonstrated in large numbers.

Steady mortalities of about 80% per 100 days have reduced
the prepubertal crabs that we have raised in cage culture to a

few individuals at the molt to adult. There are no indications

of serious disease or nutrition factors involved in these
mortalities. Ryther, et al . (1987) found disease a serious

problem in culture. However, nitrate levels of 35 mg/1 (compared

, . e ...

with typically 1 to 7 X 10 mg/1 in the wild) indicate serious

water quality problems. At two sites, Carriacou and Grand Turk
(in the 1983-1987 period), logistics, finances and timing
prevented crab growout beyond a three to six month period.

Nonsuch Bay, Antigua was our only environmentally poor site,
where we were restricted in the later stages of growout, also for
reasons not related to crab biology or mariculture management.

At Buen Hombre, Dominican Republic sufficient time was available
to gain minimum experience in the 300-500 day range and some
crabs were brought to reproductive maturity by local fishermen
participating in the project. However, even here, logistic and
political problems minimalized that experience.

Our successful experience and continual improvement in
reducing mortalities in bringing 200 to 300 Mithrax spinosissimus
per hatch to 100 days in open water cage culture is promising.

We are continuing to work on providing in cage molt protection in
a variety of forms and a more efficient turf transfer process.

It seems likely that a survivorship of greater than 50% per 100
days can be achieved from 100-400 days. This would render the
process successful, at least on a minimal basis. With more

experience in managing the details of the culture process, it is
likely that within a relatively short period of time, 30%
survivorship can be achieved from first crab to 100 days (to
produce 600 juveniles per hatch for adult growout) , as well from
that time to market size (for a total of 180 market animals per
hatch) . This is sufficient to provide a successful and
competitive process for the itinerant fisherman who can easily
produce a hatch several times per month.

We feel strongly that all indicators remain positive for the
completion of a successful itinerant fishermen mariculture of
Caribbean King Crab, based on algal turf culture. Success is
likely to lie in improving the details of cage structure and
operation (the removal of waste without predator introduction and
in providing molt protection for individuals against their
cohorts and rough weather) and in the improved management of
algal screen transfer. Ultimately, low-cost, mass-produced
plastic cages would greatly reduce the time and cost to the
individual fisherman.

Note that in this volume we chose to measure carapace length
(CL) rather than carapace width (CW) , which is the standard for
crab biology. This was done to provide a more accurate
measurement for this species when hundreds of individuals, some
quite small need to be sized. A relationship for CW = f (CL) is
provided in figure 5 (Biddlecomb, et al . , this volume) .

LITERATURE CITED

Adey, W.H. 1982. U.S. Pat. Doc. 4,333,263.

Adey, W.H. 1983. The microcosm: A new tool for coral reef
research. Coral Reefs, 1: 193-201.

Adey, W.H. 1987. Food production in low nutrient seas: bringing
tropical ocean deserts to life. Bioscience 37(5): 340-348.

Adey, W.H. and Steneck. 1985. Highly productive eastern

Caribbean reefs: synergistic effects of biological, chemical
and geological factors. In: M.L. Reaka (ed.). The ecology
of coral reefs. Symposia Series for Undersea Research,
Volume 2. National Oceanic and Atmospheric Administration,
Rockville, Maryland, U.S. A.

Adey, W. H. and T.R. Goertemiller . 1987. Coral reef algal turfs

master producers in nutrient poor seas. Phycologia 26(3):

374-386.

Bohnsack, J.L. 1976. The spider crab, Mithrax spinosissimus : an
investigation including commercial aspects. Florida
scientist., 39(4): 259-266.

Brownell, W.M. and A.J. Provenzano, Jr. and M. Martinez. 1977.
Culture of the West Indian Spider Crab (Mithrax
spinosissimus) at Los Roques, Venezuela. J. of World Mar.
Soc .

Coen, L.D. 1987. Plant-animal interactions: ecology and

comparative functional morphology of plant-grazing decapod
(Brachyaran) crustaceans. Ph.D. dissertation, University of
Maryland. 241 pp.

Doyle, J. 1985. Altered Harvest: agriculture, genetics and the
fate of the world's food supply., New York., 502 pp.

Hazlett, B.H. and D. Rittschof. 1975. Daily movements and home
range Mithrax spinosissimus (Majidae, Decapoda) Mar. Behav.
physiol., 3: 101-108.

Lewis, J.B. 1977. Processes of organic production on coral
reefs. Biol. Rev. Cambridge Philos. Soc., 52: 305-347.

Munro, J.L. 1974. The biology, ecology, exploitation and

management of Caribbean Reef Fishes. Part V. The biology,
ecology and bionomics of Caribbean Reef fishes: Crustaceans
(spiny lobsters and crabs) Res. Rept. Zool. Dept. Univ. West
Indies . , 3 : 39-48 .

Provenzano, A. J. , Jr. and W.N. Brownell. 1977. Larval and early
post-larval stages of the West Indian spider crab, Mithrax
spinosissimus (Lamarck) (Decapoda :Maj idae) . Proceed. Biol.
Soc. of Wash., 90(3): 735-752.

Ryther, J., R. Winfree, J. Holt, R. Creswell, W. Lellis, J.
Chaiton, C. Kovach and F. Prahl . 1987. Antigua Crab

Mariculture, Annual Progress Report. Harbor Branch
Oceanographic Institution, Fort Pierce,, Fla; July 15, 1987.

7 8 pp .

ACKNOWLEDGEMENTS

This volume is dedicated to the memory of Louis Petersen.
Although his primary role with the laboratory was that of an
aviator and mechanic, two roles in which he was superb, no task
at the Marine Systems Laboratory was beneath him or too much for
him. It would have pleased Louis greatly to finally see the
"Caribbean King Crab" in production.

So many people have participated with us on our Mithrax
project that it would be impossible to mention them all. Our
field fellows included Bill and Kathy Bernard, Kimberly Peyton
and Kim Moller , who are still actively involved in attempts to
commercialize the crab farming process. In addition, Tim
Goertemiller , Mitch Yadven, David Robicheaux, John Nader, John
Tschirky and Dave Warren provided technical as well as hands-on
physical support. Many Earthwatch volunteers helped to build and
maintain the turf screens and crab cages. Special thanks to
Richard Lederer for his contributions and friendship.

Dennis Farrier worked tirelessly to secure commercialization
of the crab, while Kate Hartley and Janice Byrum were our
administrators for the project. Greta Rosenzweig acted as
production coordinator/editor of the volume with help from
Charlotte Johnson. In addition, Charlotte was the artist
responsible for all of the figures in this publication excluding
those done by Gustavo Hormiga who graciously completed the ink

drawings of the eggs, larvae and morphological subjects.

Earthwatch and the Bioenergy Council funded the earliest
project stages. Much help in the Dominican Republic was provided
by the Fundacion Natura. The Agency for International
Development, with Maria Hatziolis and Jim Hester as project
managers, provided most of the funding.

We thank Anson Heines, Majorie Reaka and James McVey for
reviewing this manuscript. Their comments were provocative and
quite helpful. However, we take full responsibility for any
remaining problems in the text.

SECTION I: ALGAL TURF PRODUCTION

THE COMPOSITION AND PRODUCTION OF
TROPICAL MARINE ALGAL TURFS IN
LABORATORY AND FIELD EXPERIMENTS

THE COMPOSITION AND PRODUCTION OF TROPICAL MARINE
ALGAL TURFS IN LABORATORY AND FIELD EXPERIMENTS

W. H. Adey and J. M. Hackney

Abstract

Laboratory and field studies in the Caribbean region with
cultured algal turf communities have demonstrated consistent
primary production levels of 8 to over 15 g (dry) /sq. m/day. A
wide variety of environmental factors that control this
production, including wave action, current, harvest rate and
substratum type, are discussed in detail.

Tropical screen-cultured algal turfs are highly diverse
(typically 30-40 species) . Algal screens are dominated by
benthic diatoms, during early development. At maturity, under a
7-20 day harvest regime, blue-green and red algae, with epiphytic
diatoms, come to dominate the community.

Carbon percentages, as measured by CHN analyses, are
relatively low (14-30%) in algal turfs, in large part due to the
siliceous cell walls of diatoms and the proteinaceous cell walls
of blue-green algae. Also, in the field, even properly-managed
algal turf screens contain some carbonate silt. On the other
hand, protein levels are high (8-10%) in algal turfs and the high
growth and reproduction rates of crabs fed pure algal turf
suggest that the problems that terrestrial grazers face in
degrading and utilising higher plant cellulose are greatly
reduced with turf grazers in the marine environment.

INTRODUCTION

Most traditional fisheries and maricultures are based on
phytoplankton production. Recent world-scale analyses of primary
production demonstrate that this floating and mostly pelagic
community has low efficiency, well below most land communities,
but nevertheless is the dominant component of total oceanic
biological energy conversion (Bunt, 1975; Ryther, 1959, 1969 ;

Adey, 1987a) . On the other hand, coral reefs often have been
regarded as the most productive of world ecosystems (Lewis, 1977;

Sournia , 1977; Adey and Steneck, 1985), and the source of that

production has been attributed mostly to a thin, low biomass,
grass-like layer of algae on the reef surface. Unfortunately,
little serious effort has been made to understand the mechanisms
of this particular type of primary production and how it might be
used. It is now strongly suspected that these diminutive reef
algae can consistently achieve high levels of plant production
and are generally not limited by the factors that typically
restrict phytoplankton production (Adey, 1987a) . It has become
common practice to refer to this assemblage of reef algae as
"turfs" . This paper examines a series of laboratory and field
experiments on primary production by a coral reef algal turf
assemblage .

The term "algal turf" has been employed frequently without
reference to any particular morphological, ecological, or
systematic definition. Most authors have regarded turf as a
collection of short filamentous or foliose algae that grow to
form a thick, dense mat (Dahl, 197 2 ; see also review by Stewart,
1982) , a group definition that remains applicable to a range of
plant associations. Neushul and Dahl (1967) utilized the term to
describe subtidal algae that varied from delicately branched
forms such as Griff ithsia , Antithamnion » and Pterosiphonia , to
considerably larger species of foliose Halvmenia and the more
heavily branched Botrvocladia . These algae grew both as

epiphytes of kelps and as cover on hard or soft substrata beneath
a macroalgal canopy. The authors reported that this turf form
displayed marked seasonal changes in distribution and composition
in temperate Pacific waters. Algal turf also has been identified
as a major component of the intertidal. Stephenson and
Stephenson (1972) noted its presence throughout most of the
temperate and tropical regions of their survey. These turfs
encompassed blue-green filaments, crustose forms, articulated
corallines, and a diversity of fleshy macroalgae that included
Bostrvchia . Codium, Crvptonemia . Dictvota , Gelidium. Giaartina .
Laurencia . Plocamium . Pterocladia . Rhodvmenia . and Turbinaria .
Turfs from the low intertidal of temperate and subtropical waters
were described by Stewart (1982) as composite structures of
various anchor species (e.g. , Corallina . Lithothrix . Hvpnea . and
Pterocladia) , as well as, epiphytic species (e.g. , Ceramium,
Laurencia . Chondria, Heterosiphonia) . The anchor taxa in these
turfs, though often multi-generic composites, displayed similar
morphological characteristics and persisted throughout the year,
while the epiphytic species tended to fluctuate seasonally.
Aggregate assemblages of a few algal species have been labeled
turfs on shallow coral reef flats and slopes (Hay, 1981, referred
to Dictvota . Halimeda , and Laurencia) , and on temperate
intertidal rocky shores (Taylor and Hay, 1984, referred to
Corallina . Lithothrix . Gelidum . and Rhodoqlossum, ) . These algae
possess both prostate and upright branches that form a tightly

compacted morphology which is capable of resisting both
desiccation, occuring during periodic emersion in either
environment, and intense herb ivory encountered on coral reefs.
These factors probably also prevent epiphytism by more productive
species. Lower portions of the coral reef turf thalli exhibit
decreased rates of apparent photosynthesis and dark respiration,
which enables persistance as resting stages during periods of
physical stress and maintains substratum coverage by the
individual populations (Hay, 1981) . Species forming turfs in the
temperate environment are able to adjust the degree of thalli
compaction under varied levels of environmental stress (Taylor
and Hay , 1984 ) .

In addition to the wide variety of turf forms of the
examples cited above, investigators have begun to recognize a
turf assemblage on coral reefs that is distinguished less by
characteristic morphologies or systematics than by a high
efficiency in solar energy capture and growth. Whole-system
estimates of daily coral reef gross primary productivity have

O — , 1

ranged from 3.4-20.0 (mean = 10) g C m d (see reviews by
Lewis, 1977, and Sournia, 1977), with yearly estimates ranging
from 1800-4200 g C m“2 yr-1. Gordon and Kelly (1962) have
reported a maximum of 11,680 g C m-2 yr-1 gross primary
productivity for a fringing reef in Hawaii, which rivals systems
under intense agricultural management. Algal assemblages that

have been described as turfs have been identified as providing
the major portion of the carbon fixation occurring on coral reefs
(Odum and Odum, 1955; Borowitzka et al_. , 1978; Borowitzka, 1981).
As much as 70-80% of the total productivity of reefs (Brawley and
Adey, 1977; Adey and Steneck, 1985) have been attributed to algal
turfs. Though the importance of algal turf to the coral reef
ecosystem and the great magnitude of its productivity have been
acknowledged in the last decade, the literature has not
recognized yet a specific description of the community most
responsible for these productivity levels.

For the past several years, the Marine Systems Laboratory
has studied the dominant plant assemblage of the Caribbean and
western tropical Atlantic coral reefs in field, microcosm, and
productivity chamber studies. We refer to this assemblage as a
coral reef algal turf. We have demonstrated that this turf is
capable of high rates of productivity both in the wild and in
culture. We describe here the various characteristics which we
feel distinguish it from assemblages that are more broadly
defined in the literature.

The Coral Reef Algal Turf Assemblage

In this paper, a coral reef algal turf is a multi-specific
association of benthic, subtidal, free-living algae subjected to
emersion only during extremely low tides. The assemblage
typically persists as coverage of dead coral colonies, loose

rubble, interstitial surfaces, and other areas of reefs that are
subjected to high levels of grazing. Although seasonal shifts in
biomass and abundance occurs in this assemblage (Adey and
Steneck, 1985) , the lack of more pronounced seasonal changes on
submerged portions of tropical coral reefs renders inappropriate
a distinction of annual vs. perennial forms. These turf species
are predominately unicells, uniseriate filaments, simple
branching filaments or weakly corticated filaments that generally
range in height from a few mm to a maximum of several cm. The
height of this assemblage at any point in space and time depends
largely on the intensity of grazing by herbivores. Smaller
members of more anatomically complex macroalgae having pseudo-
parenchymatous thalli are generally less important but persistent
members of this assemblage.

In a similar grouping on an Australian fringing reef,
Morrissey (1980) noted an imprecise boundary between turf species
and larger macroalgae, particularly those genera with early
developmental stages of short stature. An additional
distinguishing characteristic for many of these turf species is
the frequent presence of an extensive, stoloniferous basal system
that supports multiple, upright axes (Brawley and Adey, 1977) .

In this regard, this assemblage closely resembles the
herpophytes, described by Setchell (1924) as minute, crawling
algae with prostrate axes that attach to the substratum by
fascicles or rhizoids and bear erect or ascending lateral

branchlets. A similar classification, the Hemichamaephyceae , was
made by Nasr (1946) for perennial tropical algae of this form
that possess apical growth zones (see also Feldman, 1966) .

The thalli arising from this stolonif erous basal system
conform generally to the Littler and Littler (Littler et al . ,
1980; Littler et al . . 1983) and the Steneck and Watling (1982)
filamentous functional-form groups as commonly uniseriate or
multiseriate , lightly corticated filaments. However, composition
of this coral reef turf assemblage is fairly complex and includes
many species which do not fit any one particular functional-form
or anatomical group classification. A number of diatoms
(centrate, pennate, unicellular, and filamentous) , coccoid and
filamentous blue-green algae, and benthic dinof lagellates
dominate in early stages of turf development and persist
generally through later stages, adhering to a thin layer of
detrital scum which persists among the basal holdfasts. Such
components of the turf probably contribute to formation of the
scum by mucilage production. A variety of bacteria, protozoans,
and occasional metazoans (e.g. , nematodes, small annelids,
microcrustaceans) are also associated with this scum layer.
Tubular or sheet-like species of Enteromoroha . which attach to
the substratum with a single holdfast, are often present, though
only in small size. With the reduced stature of this assemblage,
epiphytes are limited to blue-green algae, diatoms, or minute

branching filaments such as Asterocvtis or Erythrotrichia .

Table 1 contrasts the diversity of species that are found
consistently in this assemblage with those genera that most
frequently composed the remainder of the benthic flora in surveys
of a number of Caribbean/West Indian coral reefs (Adey et al. ,
1979; Adey and Goertemiller , 1987; Adey and Steneck, 1985; Connor
and Adey, 1977; Peyton et al. , 1987). Each genus in this table
is categorized as: 1) including species that are common,
persistent components of the coral reef turf assemblage; 2)
including both the turf species listed and other species which do
not persist within the turf assemblage; or 3) including no
persistant turf species and considered as either an encrusting
coralline or a macroalga in this study. The genera are divided
further into a series of anatomical groupings ranging from
simple, single-celled forms to the more tightly packed cellular
arrangement of parenchymatous thalli. A category of pseudo-
parenchymatous construction was designated in this summary to
account for those algae which lack extensive three dimensional
cellular division, yet achieve some mass and degree of
morphological complexity by cortication and other differentiation
about a basically filamentous form.

As defined by the organization of Table 1, the coral reef
algal turf assemblage encompasses genera of simpler anatomical
organization in each of the phylogenetic groups. The unicellular
forms representing green, brown and red algae are limited to

spores which are observed often to adhere to the mucilagenous
scum layer at the base of the assemblage. Among the filamentous
eucaryotic components, branched forms predominate generally and,
as defined here, include the common Licmophora , a diatom cell
complex supported on repeatedly branched mucilaginous stalks.
Those turf genera that achieve a pseudo-parenchymatous thallus
construction through multiseriate, polysiphonous , or corticated
growth are predominately members of the Rhodophyceae , a class
with an essentially filamentous organization. These more
complexly corticated red algal species are often in genera that
contain other species not persistent within the turf assemblage.
In such cases, the turf elements tend to be more minute species
of the genus (e . g . , Gelidium pusillum. Laurencia caraibica .
Amphiroa f raqill issima . Jania capillacea) . Species in this
assemblage from "turf/non-turf " genera are organized clearly with
greater thallic structural support than simpler filamentous turf
algae, but they have mature thallus diameters that are limited to
under 300 uM. This is less than the diameter achieved by certain
uniseriate turf elements such as Cladophora. On the other hand,
larger examples of simple anatomies (e.g. , Valonia,

Dictvosphaeria . certain species of Chaetomorpha and Cladophora)
are rarely encountered among the turf assemblage. On this basis,
we have concluded that it is a small adult plant stature, rather
than a strictly anatomical or systematical classification, that

is the primary factor determining composition of the coral reef
turf assemblage. The ability to grow rapidly, to enter
reproductive cycles quickly and to withstand repeated grazing
through basal structures or rapid spore settlement are necessary
for successfully exploiting a small adult stature of this nature.

Grazing Pressures and Turf Maintenance

A more unified definition of coral reef algal turf
composition is provided by examining the macroherbivore grazing
pressures which are encountered normally in reef communities and
contribute to maintenance of this assemblage. (Macroherbivores
are defined here to include fish, gastropods, urchins, and larger
crustaceans such as crabs.) In brief, the turf is composed
primarily of those algae which successfully withstand the high
levels of grazing by rapid growth, rapid reproduction and basal
persistence rather than by protective mechanisms such as toxicity
or thallus strength. The reduced thallus size within this
assemblage generally has allowed minimal differentiation of
photosynthetic tissue to structural material. Since allocation
of materials to structural components increases with the increase
of structural complexity (Littler and Arnold, 1980) , high grazing
pressure would maintain surviving plants in a generally early
stage of regrowth and increase the proportion of
photosynthetically active cells in the corticated or partially
calcified turf species included in Table 1. Such high ratios of

photosynthetic tissue to structural material would provide for
high biomass-specific rates of productivity and growth (Littler
and Littler, 1980; Littler et al. , 1983). The extensive basal
holdfast systems of algal turfs often persist after upright axes
have been grazed, since macroherbivores, even parrot fish, face
increased energy expenditures after removal of a certain quantity
of plant material and substratum (Brawley and Adey, 1981) . Many
of the algal species composing this turf persist by vegetative
reproduction through fragmentation of thalli, while others
display very short cycles of sexual reproduction.

Each of these factors would assist a rapid regeneration of
the turf following grazing, and it is likely that the generally
resistant, prostrate growth form has enabled survival of the
assemblage under intense macroherbivore grazing (Dethier, 1981;
Hixon and Brostoff , 1981) . Indeed, it is now understood that
macroherbivore grazing actually benefits the coral reef algal
turf assemblage, much as the grazing of large mammals can
maintain grassland. In this case, it enables the turf to
maintain its commonly widespread coverage of dead portions of
coral colonies, loose rubble and various interstitial areas.

(Odum and Odum, 1955; Dahl, 1972; Wanders, 1976a; Morrissey,

1980) .

Among the turf algae, listed in Table 1, are a number of
macroalgae which, once established, may be able to escape grazing
with structural or chemical defenses (Ogden et al . , 1973; Ogden,

1976; Ogden and Lobel, 1978). Certain of these macroalgae, which
often display appparent parenchymatous thallus construction,
often maintain a scattered, non-persistant distribution amidst
the turf assemblage. However, sporeling or germling stages of
these plants are often consumed indiscriminately by reef
macroherbivores that graze among the turf assemblage. Since
these immature stages lack adaptations for rapid recovery from
such disturbance, this grazing decreases macroalgal recruitment
while increasing short-term fitnesses of turf species (Sammarco,
1982; Carpenter, 1981, 1984). Consequently, such grazing
pressure promotes the observed persistence of this turf
assemblage over large portions of substratum and maintains the
benthic algal component of coral reefs in a characteristic high-
turnover, early successional stage with a low standing crop of
small stature and patchy distribution (Hatcher and Larkum, 1983;
Ogden and Lobel, 1978; Dahl, 1972; Marsh, 1976).

The experimental removal of macroherbivore grazing commonly
results in temporary increases in the biomass of algal turf
(Randall, 1961; Carpenter, 1981, 1984). However, the simple,
filamentous growth form predominant in this assemblage lacks the
structural support necessary for the development of large
standing stocks. The dense packing of thalli that quickly
develops from lack of grazing eventually decays from senescence,
as a result of self-shading and reduced circulation of

surrounding seawater (Stephenson and Searles, 1960; Wanders,

1977; Carpenter, 1984). Following this decomposition, upright
thalli may regenerate from persisting holdfasts to maintain the
assemblage in an unstable stage of transition, but eventually the
turf is displaced by the slower growing, more heavily corticated
or parenchymatous macroalgae listed in Table 1 (Sammarco, 1982;
Sammarco et al. , 1974; Ogden and Lobel, 1978; Hatcher and Larkum,
1983; Carpenter, 1981, 1984). Large standing crops of such
macroalgae (e.g., Gracilaria . Halimeda . Padina, Sargassum.
Turbinaria) often develop naturally on coral reef algal ridges
and wave-swept flats or pavements, where wave surge prevents easy
access by macroherbivores (Wanders, 1976b; Adey, 1978; Adey et
al . . 1977; Connor and Adey, 1977).

The diversity of benthic algal species on Caribbean coral
reefs has been shown to decrease eventually during an absence of
macroherbivore grazing (Sammarco, 1982, 1983; Carpenter, 1981,
1984). Relatively few macroalgae can come to dominate the
community. On the other hand, extremely heavy grazing pressures
can also decrease diversity by allowing the replacement of both
the turf assemblage and larger macroalgae with encrusting
coralline genera (see Table 1 and Brawley and Adey, 1977; Hay,
1981; Sammarco, 1982; Carpenter, 1984). These responses by the
benthic algal component suggest that macroalgae are prevented
from excluding (through eventual overgrowth and shading) the
smaller, more prostrate turf species by intermediate levels of

generalist macroherbivore grazing, a pattern encountered
frequently in the marine intertidal (Lubchenco, 1980) .

McNaughton (1984) reviews such changes in species composition,
diversity and growth form as a consequence of grazing common to
both marine and terrestrial ecosystems. The coral reef algal
turf might be further compared to terrestrial grassland
assemblages which require either regular biological or physical
disturbance (most frequently grazing and fire, respectively) to
prevent eventual exclusion by larger shrubs and woody plants (see
Pellew, 1983) .

The algal turf assemblage is subjected also to grazing by
certain microherbivores, which compound the effects that have
been discussed thus far. Brawley and Adey (1981) have
demonstrated that an amphipod crustacean found on Caribbean reefs
is capable, when undeterred by predation, of heavily exploiting
the assemblage, eventually clearing the substratum for growth by
a larger macroalga of a later successional stage. The macroalga
in these experiments, Hvonea spinella . was protected from
microherbivore grazing by its large size. On coral reef
substrata inaccessible to fish, due to experimental caging or
wave surge, the eventual displacement of turf species by
macroalgae may therefore be further facilitated by the grazing of
such microherbivores (see also Fenwick, 1976; Lobel, 1980;
Kennelly , 1983 ) .

Algal Turf in Water Quality Control

Coral reef algal turfs have played an important role in

maintaining a series of marine and estuarine microcosms and

mesocosms established by the Marine Systems Laboratory (Adey,

1983; Williams and Adey, 1983; Tangley, 1985; Adey, 1987b). The

development of these functioning closed model ecosystems has

required the balancing of diel community metabolic processes.

Many shallow water communities often have overall P/R

(photosynthesis/respiration) ratios of 1.0 or less. Even when

P/R is 1.0 or greater, considerable daily deviation from this

value can occur with non-conservative metabolites. For example,

in a study of community metabolism on Robin Reef, St. Croix, Adey

and Steneck (1985) measured a respiration rate at sunset of
. _ 1

approximately 2g02mhata depth of approximately 1 m.

Even when waters overlying the reef are supersaturated with
oxygen during the day, it is evident that with this magnitude of
community respiration, an oxygen debt would be incurred by the
reef within a few hours of darkness. However, with the cessation
of photosynthetic activity each evening, turbulent mixing of
ocean water provided by trade winds and water currents
replenishes oxygen, preventing community stress. Simple
diffusion across a still air-water interface could not supply the
demonstrated oxygen demand by itself (Smith and Marsh, 1978) .
Equally important, metabolic products such as C02 or ammonia

would be expected to accumulate to toxic levels without the
flushing provided by exchange with the open ocean. These
considerations likely contribute to the frequency of well-
developed shallow water coral reefs on the windward side of their
adjacent landforms (Adey and Burke, 1976) . The requirement for
support of a benthic ecosystem by ocean, or open river or bay
waters is even more pronounced in ecosystems that show a
consistent P/R less than 1.0 due to the import of organic
materials from other ecosystems (or mud flat, for example) .

Attempts to establish aquaria or microcosms of ecological
communities in enclosed aquaria have for the most part simulated
this open ocean exchange through traditional means of water
quality control, i . e . , bacterial or mechanical filtration with
subsidiary water sterilization. However, these water treatments
are inadequate for all but the smallest of microcosm systems and
essentially extend the basic degenerative processes of
respiration without proper restoration of the oxygen
concentrations or pH values or nutrient levels required to
prevent community deterioration. Unfortunately, the more
obvious strategy of an open aquarium system that provides
continual circulation of fresh seawater is hindered often by poor
water quality due to the siting of the laboratory or to the
changes that result from pumping, transport, or storage of the
water supply.

The critical problem of matching high quality open waters in

microcosm and mesocosm maintenance can be solved by
circulating water through a sub-unit of the system where
photosynthetic activity is promoted during hours of darkness to
counteract the high rate of community respiration (Adey, 1983).
This supplemental photosynthesis, which clears respiratory
products from the water column while restoring pH and oxygen
concentrations, can be most efficiently supplied by cultures of
algal turfs. The utilization of this turf form provides numerous
advantages over alternative plant assemblages. Phytoplankton
cultures are difficult to maintain in adequate densities and
would require continuous filtration from the seawater being
recycled to the microcosm. Cultures of larger macroalgae would
display generally lower rates of carbon fixation and being
"leaky" would likely release a variety of dissolved organic
compounds into the water column (Khailova and Burlakova, 1969;
Sieburth and Jansen, 1969; Spotte, 1979). The algal turf
assemblage, on the other hand, being opportunistic, with little
stored food, is not characterized by "leakiness." Also, it is
established in culture easily on an inexpensive, uniform
substratum of plastic-coated fiberglass screening which provides
porous substratum for the algal bases. When grown in shallow
"sea tables" and provided with intense metal halide irradiance, a
continuous circulation of water, and protection from grazing, the
very productive turf assemblage provides a highly efficient,

controllable and easily harvested unit of biological water
quality control. We refer to these separate water quality
control units as "algal turf scrubbers" (Adey , 1987) . The growth
of the algal turf assemblage upon a section of scrubber screen is
represented in Figure 1.

In addition to providing homeostatic control of the physical
and chemical parameters of microcosm water, when utilized for
coral reef microcosm management, the algal turf cultures are
employed on the reef itself to maintain a balance in the trophic
exchange that occurs with the system. Coral reef microcosms
maintained in this way are self-supporting communities in which
negative diel variation is controlled by metal-halide irradiance
and algal turfs. The primary productivity occurring within the
reef and lagoon tanks of this system is consumed by grazers on
several trophic levels and is distributed amongt more than 300
species of consumers in many of the same energy exchange patterns
observed on natural coral reefs. The only external organic input
provided is a daily addition of a very small quantity of brine
shrimp and dried krill, to simulate an open-ocean input to the
diet of the fish and some invertebrates. The periodic (seven
day) harvest of the coral reef algal turf culture compensates for
the addition of this biomass to the microcosm, develops a
balanced import/export ratio, and maintains the scrubber culture
in an early, more highly productive stage of successional growth.

Wave Surge and Cultured Turf Productivity

As benthic plants attached to either natural reef substrata
or fiberglass screening, algal turfs would be expected to
demonstrate productivity which responds to the strength and form
of passing currents (Gerard and Mann, 1979; Wheeler, 1980, 1982;
Norton et al . . 1982 ; Madsen and Sondergaard, 1983). Such currents
replenish nutrients in the water overlying these stationary
plants and reduce boundary layers, which enhances cellular
exchange and promotes efficient plant metabolism (Wheeler, 1980,
1982). Preliminary scrubbers designed to provide simple
continuous flow demonstrated relatively low harvest production
and proved ineffective in controlling nutrient levels within a
reef microcosm. Later, scrubbers were provided with wave action.
Examination of the continously recirculated seawater of one
scrubber (Figure 2) illustrates the relationship between oxygen
production and wave action. As shown, the blocking of the wave
bucket to provide a simple continuous flow of seawater results in
a decreased rate of oxygen production. This trend is reversed as
wave action is restored and production increases until the
original rate is attained. This oxygen production pattern, with
an absence of an overshoot, suggests that wave surge is not
acting simply to dislodge gas accumulating around algal
filaments, but is instead promoting photosynthesis to the maximum
rate possible given the physiological constraints imposed by the
micro-environment of this scrubber.

Nutrients and Plant Production in the Ocean

Adey (1987a) recently reviewed the basis of our
understanding of marine primary production and the role of
nutrients, waves , currents and light in determining the level of
that production by algal turfs. In this section, for continuity,
that discussion is briefly summerized.

In an examination of the "Potential Productivity of the
Sea", Ryther (1959) concluded that a net "production of organic

— 9—1

matter of some 10-20 g (dry) m d may be expected" m the
oceans. He suggested that a maximum of about 25 g (dry) xn d
could be attained under ideal conditions, without nutrient
limitation and with maximum irradiance. Later work by Ryther and
others (e.g., Goldman et al. . 1975) established that actual
productivity approaching 50% or more of these proposed maximal
rates can be achieved in large-scale culture when the critical
parameters, especially nutrient availability and mixing, are
optimized.

Traditionally, it has been accepted that net phytoplankton
production in the sea ranges from about 0. 3-2.0 g (dry) m d

(0.15-1.0 g C m~2 d-1), with the lowest values applying to
tropical open seas. More recent physical and chemical
oceanographic investigations have suggested that 14C-based
studies of ocean primary production are in error and
underestimate production by 5-10 times. Epply (1982) set a

maximal level for net plankton production in tropical and
subtropical oligotrophic seas at about 1-2 g (dry) m-2 d-1 (0.5-1
g C m“2 d"1) .

Ryther (1959), accepting severe nutrient limitation of
phytoplankton production, questioned such limitation in benthic
communities: "The fact that they (nutrients) are continually
being replenished as the water moves over the plants probably
prevents their ever being limiting." While later studies have
shown a large production potential for benthic algae (e.g. , La
Pointe and Tenore, 1981) , unrestricted nutrient availability has
been regarded as crucial.

A recent review of coral reef primary productivity by Lewis
(1977) revealed a wide range of gross primary production (GPP)
rates, covering nearly an order of magnitude (3.4-20.0 g
C m~2 d-^ ) . Kinsey (1979) took 7 g C m-2 d-^ as the modal rate
of organic carbon production for reef systems in general. Adey
and Steneck (1985) demonstrated that under ideal conditions
(determined by the state of geological development, wave and
current action and intense continuous irradiance) , considerably
higher rates of GPP (20-30 g C m“2 d-1) are routinely possible.
Data that corresponds closely to published values has been
obtained in recent studies of large reef communities (Atkinson
and Grigg, 1984) when various modeling techniques are combined
with primary productivity rates based upon standard oxygen
production data and known rates of predation.

With regard to potential nutrient limitations on primary-
production in coral reefs, some studies have shown that reefs
actually can export nitrogen (Wiebe et al. , 1975) . Other
researchers have been able to demonstrate that certain benthic
communities rich in heterocystic blue-green algae are able to fix
nitrogen (Wilkinson and Sammarco, 1983) . While older studies
suggested a "tight recycling" or retention of phosphorus, more
recent studies carried out across very broad reef flats have
indicated phosphorus uptake and depletion from the overflowing
ocean water (Atkinson, 1981) .

In spite of these well-known facts, it often has been
accepted in recent years that reef communities are strongly
nutrient limited, which has lead to numerous studies for
identifying potential sources of nutrient input (e.g., D'Elia et
ad., 1981; Andrews and Gentien, 1982; Meyer et ad., 1983; Andrews
and Muller, 1983).

Growth of Algal Turfs on Artificial Substrata in Tropical Seas

For several years, the Marine Systems Laboratory has grown
algal turfs on artificial substrata in a wide variety of marine
environments, most frequently employing plastic screens in a
range of sizes, shapes, colors, pore sizes and densities. This
work has been concentrated in the waters of the Caribbean and
southwestern tropical Atlantic, where nutrient concentrations of
nitrate/nitrite rarely exceed 0.5 ug-at/1 and often are below 0.1

ug-at/1 . Algal harvests obtained from these screens generally
indicate the extraordinary levels of productivity previously
implied by whole-system coral reef community metabolism studies.
More typically, harvests reach roughly 50% of gross primary
production by the reef. When in- situ losses due to community
respiration, thalli fragmentation, and micro-grazing are
considered, these values are quite reasonable and confirm
previous estimates of high reef production. More importantly, in
the context of the present volume, these harvests are attained
repeatedly in situations where significant recycling is either
impossible or highly unlikely.

Given intense solar energy input to an environment with
considerable wave-induced turbulence, it seems clear that high
rates of metabolic exchange and production are possible. Since
algal turfs have the ability for efficient extraction at low
concentrations of the required nutrients, it is conceivable that
only a depletion of nutrients from the overlying water column
could lead to limited production. With generally constant,
strong seawater flow in trade wind seas generated by both
equatorial currents and local wave action, nutrient depletion is
rarely a serious factor in production limitation.

Algal Turf Research

In this paper several years of research both in the
laboratory and the field are summarized, particularly with regard

to turf community structure, succession and productivity as
controlled by a wide variety of environmental factors and by the
characteristics of the artificial substrata used.

MATERIALS AND METHODS

LABORATORY STUDIES

These studies were conducted on a side loop of a 12 kl coral
reef microcosm (Adey, 1983; 1987a). A series of attached sea-
tables were constructed of polyester-coated plywood to provide
shallow wave surge tanks for the cultivation of the experimental
algal turf assemblage. One end of these algal turf scrubbers
supports a trough-like wave bucket (approximately three liters
volume) designed to fill with a continuous inflow of water pumped
from the microcosm. The wave buckets tip into the sea-tables on
off-centered axes, and create a periodic surge of seawater which
crosses the table to drains that direct the return of the outflow
back to the microcosm. These drains are fitted with standpipes
so that a 2 cm column of seawater is maintained in the bottom of

each scrubber during periods between wave generations. Two

. . . . ?
squares of plastic-coated fiberglass window screening (0.5 m, 1

mm mesh) are mounted in plexiglass frames across the bottom of

each scrubber, which divides the bottom surface area into an

upstream screen, nearest the wave bucket, and a downstream screen

of equal area. Each screen is positioned beneath a 400 watt

multivapor, high intensity metal halide lamp providing ca. 1000

. ? «...
uEin/m /sec to the screen surface. With the intense irradiance

and continuous circulation of coral reef microcosm water through

the scrubbers, algal spores, zygotes and vegetative fragments

settle quickly and grow on the screens. The passage of each

generated wave oscillates sharply the algal filaments, though

wave intensity diminishes markedly towards the far end of the

scrubber, providing a weaker surge for the turf on the downstream

screen. The design of the scrubbers was described in general by

Adey (1982) .

Standard Conditions of Turf Growth ♦ During the period of
study, the salinity of the microcosm water ranged from 35.0-36.0
parts per thousand, temperature from 25-28°C, pH from 8. 2-8. 3,
and dissolved oxygen from 5. 5-8. 3 mg/1. The concentrations of
nitrate/nitrite averaged 1.0-1. 5 ug-at/1, somewhat higher than
the 0. 1-1.0 ug-at/1 common to many Caribbean reefs (Adey and
Steneck, 1985; Adey and Goertemiller , 1987). Typically, nitrogen

concentrations increase sharply within the system following the
introduction of blocks of reef carbonate containing organisms or
fish from field collections and lower gradually with continued
removal of biomass from the scrubbers (see below) . High nutrient
concentrations for experimental work were induced by including a
number of large Mithrax spinosissimus in a 800 1 tank
interconnected with the reef microcosm.

The primary production of turf algae growing on upstream and

downstream screens was monitored over a period of several months
to assess the influence of different factors in the scrubber
environment. During this period of study, a set of scrubber
parameters was selected as standard for purposes of comparison
and included a 12 day harvest period, a light intensity of 1000
uEin/m /sec, a 2 cm depth of water overlying the screens, a
screen pore size of 1 mm, and a flow of 13.5 1/min to the wave
buckets. The standard photoperiod provided to both the reef
tanks and the algal scrubbers was 14 h light/10 h dark. The light
period for the scrubbers commenced as the various lamps over the
reef exhibit began to switch off, providing essentially inverse
photoperiods. The length of the light period is typical of a
trade wind/tropical summer day. While the scrubber light
intensities of 1000 uEin/m /sec were somewhat below those
encountered typically on a Caribbean coral reef flat at
approximately 1 m depth, the total incoming energy on a daily
basis is about the same.

Measurement of Primary Production . Primary production of
the algal turf community was determined during this study as the
weight of biomass harvested every 12 days. During harvests of
the turf assemblage, the screens were placed in a shallow
plexiglass collecting tray and scraped with a beveled plexiglass
blade. Scraping of the screens in this manner removed the major
portion of the upright filaments while the basal holdfasts or

rhizoids remained attached to the inner surfaces of the mesh
pores ( Brawley and Adey, 1981) . In addition to the material
collected by scraping the screens, any biomass growing on the
sides and bottom surface areas of each scrubber also was removed
at time of harvest for purposes of microcosm management.

However, only that portion removed from the circular area of
screen directly beneath the halide lamp reflectors was employed
in production calculations. After harvest, the collected biomass
was drained briefly, dried to constant weight (approximately four
days at 35°C) and weighed to determine production as g dry
wt m“2 d-1.

For comparison, the biomass-specific primary productivity of
the algal turf also was measured as the rate of carbon fixation

in a series of short-term incubations (Hackney, 1984) . Samples

• . . • • ?
of screening with attached algal turf were divided into 4 cm

squares and incubated for 2 h under metal halide irradiance (1000

uEin/m4/sec) m filtered, constantly stirred seawater with added

14C-NaHCC>3 (5 x 10-3 uCi/ml) . Following incubation, the turf

squares were dried, weighed and converted to g dry wt biomass

after subtracting the standard weight of a square of cleaned

screen. Individual turf squares were oxidized by the method of

Van Slyck et al . (1951) to 14C02, which was reduced to soluble

ion by passage through a 1.0 N solution of KOH. Replicate

aliquots of base from each oxidation were counted by liquid

scintillation, with average counts converted to rates of total

carbon fixed during incubation (mg C/h/g) .

Turf Composition . Prior to each of the screen harvests in
this study, three randomly placed samples of the turf assemblage
were removed with tweezers and preserved in 5% formalin. Samples
from each of the screens were combined and examined periodically
to identify the various species of algae that compose the turf
assemblage .

Long-term Observations . For a series of observations
covering a period of eight months, parameters were established in
each of seven separate scrubbers to conform to the above
standards with the exception of one of the following
substitutions (i.e., the test conditions): 1) a variation in the
depth of water overlying the screens ( 1 or 4 cm); 2) a shortened
harvest period (7 or 10 days); 3) an increased light period (16 h
light: 8 h dark); 4) a varied screen pore size (210, 710, or 1400
um) and 5) a change in intensity of irradiance (800 or 1200
uEin/m2/sec) . The intensity of light received by scrubber
screens, as measured by an integrating photometer (Li-Cor Model
LI-188B) at the water surface, was varied by raising or lowering
the position of the metal halide lamps above the scrubbers.

Paired Scrubber Studies . In addition to the long-term
observations, two scrubbers were paired for a series of tests
lasting 5-11 weeks to compare concurrent 12 day harvests under
varying flow rates. The adjustment of flow rates to the

scrubbers affected directly the frequency of wave generation,
although one comparison employed a shallower wave bucket
(approximately 2 1 volume) to examine the impact of two different
frequencies under a single high rate of flow.

Blockage of Wave Surge . The effect of totally blocking wave
generation was tested directly in one scrubber that otherwise
shared the standard parameters described above. For a period of
six weeks, seven day periods of harvest were monitored with a
functional wave bucket provided for one week, alternated by a
week during which the bucket was blocked from rotation to create
a continuous flow of water that spilled across its leading edge.
This design tested the effect of providing an identical flow rate
in both the presence and absence of wave generation and
controlled for any unaccounted environmental parameters that
might affect biomass production within these scrubbers.

Monitoring of Nutrient Concentrations . Effects of changing
nitrate/nitrite concentrations on production were observed by
monitoring a series of harvests from one scrubber after
collections of invertebrates were introduced to the reef
microcosm. As part of this examination, the productions of both
scrubber screens, subjected to standard parameters, were recorded
over a 26-week period as N- (N02 + N03 ) concentrations fell from
nearly 20 uM to less than 2 uM.

FIELD STUDIES

Mayaquana . The studies undertaken at Mayaguana were
published by Adey and Goertemiller (1987) . To provide continuity
to this volume, that work is summarized here.

The waters off Mayaguana Island (22° 20' N; 773° W) , in the
southeastern Bahamas (Figures 3, 4) were chosen as a site because
they provided a well-developed coral reef that faces open, deep
and incoming North Equatorial Current water. In addition, this
site was logistically feasible, with minimum expense and was not
subjected to winter extratropical swell. During most of the
year, Mayaguana lies within the outer trade wind belt (Adey
1978) . Abraham Bay reef on the south side of the island provides
low nutrient concentration lagoonal waters and open ocean waters
within close proximity.

During the course of the project, no significant differences
in nutrient concentration between lagoon and open ocean waters
could be discerned. The maximum concentration of nitrate/nitrite
was 0.13 ug-at/1, and on some occasions was below detection
limits .

. 9

The surfaces supporting algal turf growth were 1 m plastic
screen material of either single or double layers. The bottom of
the double screen was made of black polypropylene screen (1.6 x
4.8 mm mesh), while the upper layer was polyester (1 mm mesh).
Both layers of the double screen were sewn together with fine
plastic filament to increase strength. The single screens were

standard polypropylene (1.6 x 4.8 mm).

The turf screens were attached to individual rafts made of
7.6 cm diameter PVC plumbing pipe and were suspended at 15, 30
and 40 cm depth. The pipes at the top of the frames were sealed
shut to allow flotation. Water entered through holes drilled in
the legs and bottom sections of the frame to serve as ballast and
prevent capsizing (Figure 5) . A set of three double-screen
rafts, one for each depth (15, 30 and 45 cm), was placed at both
the lagoon and open ocean sites. A similar arrangement was
developed for single screen rafts, with two replicates of the
single screen set for both the lagoon and the open ocean sites.

Another two sets of single fiberglass and polypropylene
screens (0.5 m x 0.5 m) were suspended directly on anchored line
rather than on rafts. These open ocean screens were suspended
between the reef and the oceanic drop-off in a water depth of
about 17 m (Figure 6) . Placing of these screens allowed
examination of variation in algal turf production with depth in
an area with a much deeper bottom.

After three weeks of initial algal turf growth, all screens
were scraped and harvested. A regular scraping interval of seven
days began at the end of these three weeks. All scrapings were
done as described above. The top and bottom of each screen was
harvested and the results pooled. The algae were oven dried
until weight variation over a 12 h period was less than 0.5 g

(total) . A variation in screen design was provided by attaching
the screen directly to a polyester-resined plywood sheet. It was
not successful, but is mentioned here and briefly discussed below
because it has interesting implications.

Grand Turk. The algal turf studies undertaken at Grand Turk
have been separately prepared in manuscript form and are now in
review for publication (Peyton et al. , 1987). To provide
continuity in this volume, this paper also is summarized briefly.
At Grand Turk, additional testing with regard to the effects on
algal turf production of a variety of additional factors (e.g. ,
screen type, irradience level and harvest rate) was undertaken.

From January to September, 1984, an algal turf production
study was conducted from the R/V Marsvs Resolute in the reef
lagoons adjacent to Grand Turk Island (Figures 3, 7). The
island, which is well within the trade wind belt (Adey 1978), is
3.2 km wide by 13 km long and is oriented north to south. The
lagoon used for these studies lies on the eastern or windward
side of the island, and is approximately 10 km long (north to
south) by 2 km wide. The east lagoon faces directly into the
North Equatorial Current and is protected partially from the
trade wind sea by a patch or boiler-type complex of coral reefs
and algal ridges. The impinging waters can be characterized as
essentially tropical open ocean. Silver Bank and Mouchoir Bank
lie to the east, which is south of the general equatorial flow to

Grand Turk. The lagoon is 1-4 m deep, with scattered small patch
reefs, while the remaining bottom is covered by calcareous sand
(approximately 50% of total area) and seagrass (approximately
45%, primarily Thalassia) . The floating rafts used to support
the algal screens were essentially the same as those described
above. The controls consisted of white polyester monofilament
screens (1000 um pore size) hung at 30 cm depth.

Three study sites were established in lagoonal waters off
the east and southeast sides of the island. A study of the
effect of screen types was conducted at site 1 (Figure 7) in 4 . 0
m water depth, over the partially sandy bottom of a patch reef
grazing halo. Four of the seven screen types tested were white
polyester monofilament (500, 710 and 1000 um (control) pore
size) . Other substrata tested included nylon weave (200 um pore
size) , blue multiweave (2mm thick) and a black polypropylene
molded screen ( 2 X 3 mm pore size).

The harvest rate study and turf community study were
conducted at site 2, at 4 m water depth. The bottom consisted of
sandy sediment sparsely covered with siphonaceous green algae.
Screens were harvested at 4 , 7, 12 and 20 day intervals.

The development of the turf community upon two screens (1000
um pore size) was studied under controlled conditions. One
screen was harvested every 12 days, while the other was never
harvested. With the exception of black polypropylene screens of
2 X 3 mm mesh, all screens in this study were sampled at each

harvest for the first 97 days to observe possible variation in
community development among screens.

Site 3 was located between an algal ridge and a Thalassia
bed in 2.3 m water depth, over a white calcareous sandy bottom.
The effects of irradiance (see light measurement) on biomass
production as a function of water depth was investigated at this
location. Six screens of blue and white multiweave (2 mm thick)
screens were hung at the surface (0-3 cm) and at 10, 20, 30, 40
and 100 cm water depth, and harvested every seven days. In
addition, single, double and triple-layered black polypropylene
screens ( 1 X 3 mm mesh) were suspended at 30 cm water depth and
harvested every seven days. Multilayered screens were sewn
together with monofilament fishing line.

At Grand Turk Island, algal turf was harvested as described
above. Due to differences in irradiance received by the top and
underside of the screen, the turf scrapings were processed
separately. Harvested biomass was oven dried at 80°C for three
days and transfered to a 100° C drying oven for 24 hours or until
the variance in weight between hourly weighings was less than
0.1 g .

Prior to harvesting, four 1.0 cm samples were collected
randomly from the screens. Samples were collected from each side
of the screen and preserved in 3% buffered formalin to determine
the turf algal species composition and relative abundances of

algal genera for each screen substratum. Both the direct count
method and point counts were used in these calculations.

Photosynthetically active radiation (PAR) was measured on
cloudless days using a LI-COR Model 511 photometer with a flat-
topped cosine corrected sensor (Licor, Inc.), which averaged PAR
over a 10 sec period. Irradiance values of incident light and of
light reflected from the bottom were recorded every 10 cm through
the water column at each study site. Incident and reflected PAR
at 30 cm below the surface were noted every hour from sunrise to
sunset on four occasions. Incident and reflected light also were
measured through a screen supporting seven days of algal turf
growth and again immediately after harvesting.

Irradiance was measured in specific wave bands at site 3
using an IL 1500 series photometer with a SEA015 detector
(International Light, Inc.).

Concentrations of nitrate/nitrite and orthophosphate were
determined by standard methods using a Beckman DU-2
spectrophotometer. Salinity, temperature, and pH of ambient
water were recorded also. All samples were collected at 30 cm
water depth at each study site.

Antigua . From February to July, 1985, experimental studies
of algal turf growth were conducted in the easternmost part of
Nonsuch Bay, Antigua (17° 5 ' N ; 61° 4 1 ' W ; Figures 8, 9). The
basic methods and equipment used were the same as those employed

at Grand Turk.

In addition to seeking corroboration of previous results at
an entirely different, high island site, the work at Antigua
further extended the study of the effect of different screen
characteristics, specifically screen color (i.e., black or
white) , on algal turf production. Depth and protection factors
also were examined further, along with the effect of screen
orientation (i.e., horizontal vs. vertical). In this section,
turfs were harvested every seven days to maximize the number of
test events.

Nonsuch Bay is a relatively narrow, steep-sided, east-west
oriented bay cut into shelf limestones raised during the late
Tertiary (Figure 9) . It is protected to the east by a bank
barrier reef, which blocks much of the constant easterly trade
wind swell at this latitude. There are two main passes into the
Bay from the east, one running diagonally north/south through the
reef and the other running over a submerged reef south of Green
Island and entering west of Green Island. There are no
significant exits to the south-west or north. As a result of
this configuration, water flow forced by trade winds over the
reefs into Nonsuch Bay can exit only back into the general
equatorial current. Antigua is a moderately elevated island and
has greater rainfall and run-off than Mayaguana and Grand Turk.
Thus, the inner end of Nonsuch Bay is somewhat stagnant and
highly turbid with suspended sediment. The outer end of the bay,

where the experimental work was conducted, is significantly
clearer, though even at this site turbidity is greater under
certain conditions than at previous algal turf research sites.

Screen sets were placed 1) at Rat Island; 2) in the lee of
Bird Island; 3) at Devils Bridge and 4) to the north of Green
Island. One set of single thickness black polypropylene screens
(2X3 mm mesh) were placed at each of the sites to examine the
effects of location ( i . e . , availability of protection, amount of
suspended sediment) on production. A double set of screens was
placed at Devils Bridge. All other testing was carried out at
the Green Island site in 3 m of water, over a light sand bottom.

The Rat Island site had intermediate protection, providing
relatively turbid water and turbulence from waves generated
within the bay itself. The site lay at 6 m watere depth over a
silty sand bottom. The Bird Island site, at 3 m water depth, was
slightly less turbid than Rat Island. The lee side of Bird
Island had a small patch of mangroves that indicated considerable
protection from intense wave action. However, this site was
subject to moderate currents that changed with tide and wind
direction. Screens at the Devils Bridge site were placed at 3 m
water depth in clear water overlying patch reefs and a coral
rubble bottom. Turbulence at this site generally was intense due
to waves coming at a variety of angles around the patches and
reflecting off limestone cliffs to the northwest.

The Green Island site was the standard work area for the

studies in Antigua. This site received moderate wave-driven
surge and flow over the reef at most times, thus sharing
characteristics with the Grand Turk and Mayaguana sites.

However, due to the protection provided by Green Island, the site
on occasion could be relatively calm.

Since previous tests on the effects of screen color ( i . e . ,
potential reflection or absorption of light and/or heat) were
inconclusive, additional testing was conducted at the Green
Island site. Sets of single layer black polypropylene (2 X 3 mm
mesh) and white polyester monofilament (2 mm mesh) were paired
and placed at the surface, and at 0.1, 0.2, 0.4, 1.0 and 2.0 m
water depth and monitored for a period of 14 weeks.

At the Green Island site, additional testing of multilayered
black polypropylene screens (2 X 3 mm mesh) was conducted using
double, triple and quadrupal layers for a period of nine weeks.
Single layers of these same screens also were placed horizontally
at 30 cm and observed for nine weeks. These were placed on
standard float lines, similar to those used for feeding crabs
(see below) .

Carriacou , Grenada . From March thru June, 1986, algal turf
production studies on standard plastic screen substrata were
carried out in Grand Bay, Carriacou (12 °, 28' N; 61°, 26' W;
Figures 8, 10). As a relatively high energy study site, interest

in algal turf studies here was directed towards comparing
findings with data from Mayaguana, Grand Turk and Antigua.

Carriacou lies on the large and moderately deep Grenadines
Shelf. It is subject to trade winds of high constancy and
relatively high velocity (Adey, 1978) . The shelf is
characterized also by strong tidal flows. Grand Bay is oriented
north to south and the east and south sides are protected by a
nearly continuous bank barrier reef, the crest of which lies near
or slightly below mean low water. Continuous waves of
approximately 0.2-0. 6 m height occur on the reef apron that lies
adjacent to the outer lagoon. The primary reef lagoon pass is
located near the island in the southwest corner of Grand Bay.
Though a well-developed lagoon, the reef apron site at Grand Bay
is nevertheless one of the most turbulent lagoon sites in the
eastern Caribbean.

Work at Grand Bay was carried out at five sites. Two of
these lay in the more turbulent eastern part of the bay. Site 1
was situated on the reef apron itself, while site 2, located
immediately off the apron, was adjacent to a patch reef. Both
sites were in very clear water over coarse sands, with a water
depth of about 5m.

The remaining sites lay in the western half of the bay on
the island sediment apron. Site 3 lay near the eastern margin of
the apron in fairly clear and moderately turbulent water,
overlying fine sand at a water depth of about 8m. Sites 4 and 5

lay nearer inshore on the island sediment apron, at water depths
of 5 and 4 m, respectively. At these sites, the water was quite
turbid. While waves were slightly higher on average at the inner
sites, total water flow from waves and tide-driven currents was
less .

Standard harvest procedures were employed at Carriacou, as
described above. All screens used at this site were black
polypropylene (2 X 3 mm mesh). Standard screens (0.92 m2 ) were
placed horizontally at 30 cm water depth and vertically, with the
top at 10 cm below the surface. In addition, smaller screens
(0.57 m ) were placed horizontally and vertical arrays of eight
screens were fitted in a cage size frame.

Buen Hombre , Dominican Republic . Extensive algal turf
productivity studies were not carried out at Buen Hombre (19 °,
40* N; 71°, 20* W ; Figure 11). However, since extensive crab
mariculture studies were carried out at the Buen Hombre site, a
limited number of comparative algal turf production studies were
undertaken there from May thru September, 1985. These studies
were conducted using the standard mariculture float/feed lines
and black polypropylene screens (2 X 3 mm mesh) (see below) that
were placed vertically. The harvest and drying procedures were
standard. All screens were double-layered and 0.56 m (0.61 X
0.92 m) . Four screens were anchored by a single attachment
point, four were anchored by two attachment points, and two

screens were hung one over the other (tandem) .

Experimental screens grown in the wild at one locality show
wide and synchronous variation in algal turf production that is a
function primarily of wave action, current and available light
(cloudiness) . Because of this synchronous variation,
demonstrations of the significance of other variables is often
difficult to demonstrate with available statistical techniques,
even when a plot of the data clearly shows a visually significant
difference. On four test lines of 10 screens each recently
established in Grand Turk lagoon (May-September , 1987) , each test
line was successively examined for the sign of change (+,-)
relative to each other line with each harvest. The results were:
34/11; 37/8; 35/10 and 37/8 (change in same direction as line
tested/ change in opposite direction) . This procedure clearly
demonstrates that much of the variation against which
significance is tested is due to synchronous and not random
variations. In the following presentation of results and
discussion, where tests of significance are negative but close
and where the plotted data indicate the likelyhood of a
difference, the results are simply stated along with the nature
of the test.

RESULTS

LABORATORY STUDIES

Scrubber Turf Composition . Thirty (36) species of algae
were identified as consistently present in the scrubber turf
(Table 2). As in the field turf assemblage, the scrubber turf is
a complex of species from each of the major benthic algal groups.
At any one time, observations indicated that 30-50% of the
biomass of this turf was composed of blue-green algae, most
frequently species of Calothrix . Qscillatoria . and Schizothrix .
The eukaryotic genera most dominant in the community were
Ceramium . Cladophora , Ectocarpus , Enteromorpha . and Polysiphonia .
Smithsoniella earleae . a persistent, occasionally common
component of the scrubber turf assemblage, is encountered rarely
in the field. Vegetative fragments of non-turf, macroalgal
species were found frequently to have settled within the scrubber
assemblage. Without periodic harvests, these fragments are
capable of maintaining active growth on the scrubber screens.

Long-term Observations . The harvests in the long-term
observations provided production values ranging from 0.3-21.6 g
m~2 (-j-2 (mean = 7.2, ±4.0 S.D., n=277 harvests from seven
scrubbers) . The lowest productions were measured in the first
harvests of the study, following introduction of screens to the
scrubbers, and typically increased over a two-three week period
as the turf assemblage became fully established upon the screens.
The C/H/N ratios for samples of scrubber algal turf were

determined by using a Perkin-Elmer Model no. 240 elemental
analyzer. Mean compositions of 25.9% carbon (±3.0, range = 17.0-
31.2, n = 61 replicates from 27 turf samples) and 2.7% nitrogen
(±0.5, range = 1.4-3. 4, n - 61) were calculated. When combined
with harvest data from this study, the carbon composition value

. . —9 —1

would indicate that 0.8-5. 6 g C m d are fixed by algae m
these scrubbers. In comparison, rates of carbon fixation
measured during 14C incubations averaged 5.32 mg C/h/g dry wt
(±3.46, range = 1.35-23.45, n = 216 samples from 12 incubations).
When these data are calculated alternately in units of areal
production by employing the value of 4 cur for each sample of
turf-covered screen, a range of 0.15-0.64 g dry m h is
obtained (mean = 0.28, ±0.09).

The harvests of both upstream and downstream screens from
all scrubbers undergoing long-term observation were compared to
examine the impact of wave surge upon turf biomass production.
Mean production of downstream screens, 4.5 g dry m-2 d-1 (±1.88,
n = 98 harvests from 5 scrubbers) was significantly lower (paired
t-test, P < 0.01) than the mean upstream screen production, 8.5 g
dry m'2 d"1 (±3.8, n=99) .

Harvests from screens of varying pore size demonstrated that
pore size may be decreased (fiber surface area may be increased)
to a point where carbonate accumulation is promoted, which
interferes in the establishment and subsequent growth of the turf

assemblage. As a result, screens of 210 urn pore size failed to
support adequate biomass accumulation and were subsequently
discontinued. While it has been observed that turf growth may
decline on the decreased surface area provided by markedly larger
pore sizes, no significant differences were detected between
harvests from standard screens (1000 urn) and 710 or 1400 urn
screens .

There were no significant differences detected between
scrubber treatments when water depth, harvest schedule,
photoperiod, or light intensity were varied within the scrubbers.
However, it is likely, based on field studies, that insufficient
parameter variation was applied.

Paired Scrubber Studies . The summary of data from the
series of concurrent harvests (Table 3) shows that in the first
test the two scrubbers provided with flow rates of 5.5 and 16.0
1/min produced significantly increased biomass on both upstream
and downstream screens under the higher flow. Furthermore, under
the 16.0 1/min flow rate, average productions from both screens
were equal statistically (comparison of 1 c to 1 d, paired t-
test, p < 0.05). In the second test, average upstream harvests
under 13.5 1/min (2a) were equal statistically to average
harvests from both screens receiving 16.0 1/min in the preceeding
tests ( lc and Id each compared to 2a, P < 0.05). However, an
increase to 32.5 1/min led to greater harvests on downstream
screens only (2d), which resulted in a mean production value that

equaled statistically the average upstream production under
either flow rate (2d compared to both 2a and 2c, P < 0.05). In
the third test, the increase from 13.5 to 32.5 1/min resulted in
no significant increase for either screen. This test employed a
smaller wave bucket at the higher flow rate, creating a 5 sec
wave period. The average downstream production in this test,
lower than observed on upstream screens at both flow rates, was
lower statistically than average downstream production under the
same flow rate in the preceeding test (comparison of 3d to 2d, P
< 0 . 05) .

Blockage of Wave Surge. The results of the blocked wave
surge tests (Figure 12) illustrate that, while production varied
considerably throughout the six weeks of harvests, the blockage
of wave surge affected significantly lower production for both
the upstream and downstream screens (paired t-test, P < 0.01).

The production of downsteam screens was consistently lower than
that of upstream screens in both the presence and absence of wave
surge .

Monitoring of Nutrient Concentrations . Nitrate/nitrite
concentrations measured during the 12 days prior to each harvest
varied by over an order of magnitude, from 0.8-20.2 ug-at/1
(Figure 13). Although there was no statistically significant
relationship detected between the nutrient data and harvest,
production averaged 11.2 g dry m”2 d-1 (N = 18) at N

concentrations < 2.0 ug-at/1, 7.2 g dry m~2 d_1 (N =10) at
concentrations of 2-7 ug-at/1, and 5.0 g dry m“2 d”1 (N = 2) at
concentrations of > 15 ug-at/1. If there is a relationship, it
is an inverse one.

FIELD STUDIES

Mavaquana . Plywood-based screens tended to accumulate fine
carbonate sediment and produced algal turf growth very slowly.
These screens lacked the constant animal activity of a hard reef
surface, and were not subjected to cycloidal wave currents.
Consequently, these screens resembled broad patches of protected
reef pavement and contained more turf-bound sediment than the
largely sediment-f ree open carbonate surfaces that characterize
more irregular and/or turbulent reef sites. These rafts were
eventually discontinued and detailed data are not presented. The
experiments are mentioned because they demonstrate that the
nature of the substratum by itself may limit production
significantly even when potential production is very large.
Subsequent studies have shown that finer screens (less than 0.5
mm mesh) also can accumulate significant fine carbonate sediment
in some lagoonal sites, particularly on upper surfaces (see
below) , resulting in greatly reduced algal production. In the
Mayaguana study, microscope observations on selected sample
scrapings indicated the existence of only a very small proportion
of sediment.

Data from the remainder of the screens is given in Table 4.
The screens reached maturity, or sub-climax production, after
four to six weeks of harvesting (Figure 14) . Both the single
layered black polypropylene (1.6 X 4.8 mm mesh) and the attached
plastic-coated fiberglass window screen initially appeared to
have the same potential for supporting algal growth. The single
screen, however, lacked the support and protection of double
layers and tended to degenerate, a process that ultimately lead
to the tearing and loss of entire pieces of screen in rough
water.

About 50% of the turf species identified on the reef were
also found on mature raft screens (Table 5) , although a long term
study would probably lead to identification of further species.

A number of diatom species, belonging to several genera,
constituted the first algal colonizers. The diatoms appeared
within a few days and developed to form a heavy, white-yellow
fuzz within several weeks. After five to seven weeks, and
several harvests, a mixture of blue-green algae and diatoms
dominated the screens. After seven to eight weeks, the typical
screen turf assemblage consisted of roughly equal quantities of
blue-green algae and red algae, with only epiphytic diatoms and
very few filamentous browns.

It is obvious that screen depth is critical to harvest
production (Figure 15) . As one might expect, algal turf
production generally declines with depth. However, on the 1 m

screens, production also clearly declines near the water surface,
displaying a consistent production peak at about 30 cm water
depth. In all of the major raft localities (both lagoon and
ocean) , screens were placed at 15, 30 and 45 cm. At all non¬
lagoon sites for which harvests at all three depths were
conducted, the peak of mean harvest production occurred at 30 cm,
with marked reductions at both 15 and 45 cm (Figure 16) . Also,
out of 34 harvests conducted at all three depths, 19 showed peak
production at 30 cm while 15 showed peak production at 45 cm. In
only one case did the peak occur at 15 cm, and this was on a
single screen raft, the most unstable design type. A two-way
analysis of variance on a randomized block design revealed a
highly significant difference between production at 15 and 30 cm
(P < 0.01). The difference between 30 and 45 cm is significant
at a probability of 0.05. However, the very consistent and
smooth drop in production levels on screens placed in deeper,
open ocean water indicates a pattern of production that peaks
near 30 cm and declines with increased depth. Simple comparisons
of standard deviation bars tend to cover up the wide and parallel
production variation over time.

Even though there were significant and consistent
differences in harvest values between the lagoon and open ocean
sites for all shallow depths of screen placement, the patterns of
increase or decrease over time are quite similar in both areas.

For example, in 27 out of 33 instances, harvests of the deeper
(45 cm) and shallower (15 cm) screens followed both increases and
decreases of harvests from screens placed at 30 cm water depth.

On the other hand, lagoon and open ocean harvests seemed to be
uncoupled. After the screens matured, the change in lagoon
harvest production over time was obvious and consistent on each
raft. Harvest was high in April, low in May, and high again in
June (Figure 16) .

Grand Turk.

Species Composition . Algal taxa from the Rhodophyta (reds) ,
Cyanophyta (blue-greens) , Chlorophyta (greens) , and Chrysophyta
(golden browns) were observed in the screen algal turf
assemblages (Table 6) . Spatial distribution and attachment of
species observed in the intial 168 days of screen turf
development can be divided into three ecological types: 1) the
mat forming species which occur as mucilagenous colonies, on
stalks, or as chains or filaments of cells (primarily Chrysophyta
and Cyanophyta); 2) filamentous plants growing through the
matted layer and producing a canopy of erect and creeping species
with holdfasts, prostrate branches, and rhizoidal outgrowths
entangled around the screen mesh (almost exclusively Rhodophyta)
and 3) conspicuous epiphytes, growing on the canopy plants,
primarily Chrysophyta and a few Cyanophyta. Two distinctive

communities developed, due to differences in incident light

between the top and underside of the screen. Although each side
had similar species composition, percent composition of those
species varied (Figure 17, 18). Benthic diatoms colonized the
screens within 24 hours after they were placed in the lagoon and
during the initial two months, diatoms continued to dominate the
pioneer community. After one week, both sides had developed a
layer of mucilage. This sticky layer consisted primarily of
diatoms, bacteria and their secretions, which included some
detrital matter. Thus, a bio-adhesive layer, on the basal
plastic screen substratum, proceeded spore setttlement of most
green, brown and red algae.

All major diatom species were present after 10 days on the
top sides of the screens. Eighty percent of the turf community
was the genus Syneda sp.B. Development on the bottom side was
slower and included a greater number of diatom species, including
those of Tabelaria sp, Licmophora sp, Svnedra sp.A and a single
unidentified species.

A notable successional pattern occured over the first 25
days on the top side of the harvested screen and about 50 days on
the bottom. On the top, the growth of the dominant Svnedra sp.B
was followed by that of Licmophora and finally Tabellaria . The
same pattern occured on the bottom, except that during the
25-50 m day interval, two new species settled and successively
developed into major components. Growth of Svnedra sp.A was
followed by that of an unidentified diatom. For the first 25

days, the field count percentages and generic composition did not
significantly differ between the harvested and unharvested
screens. By the 84th day, coccoid and filamentous red and blue-
green algae appeared in significant numbers. For the first 69
days, diatoms dominated the top side of the screens, representing
55% of the community and over the next 30 days their numbers
decreased to 40% of the community. Later epiphytic diatoms,
representing 60% of the community and largely growing on the red
canopy, dominated the remaining 78 days of the study.

On the underside of the harvested screens the diatom
colonization stage ended at 54 days, representing 34% of the
community. At this point, the diatoms were succeeded by blue-
greens, primarily Anacvstis . For the next 50 days, the blue-
greens fully dominated the underside. Finally, reds with a thick
cover of epiphytic diatoms became primary elements. Mature
screens discussed in this paper, have a whitish/yellow color on
the upper side and a reddish brown color on the bottom side,
which indicate the relative importance of the diatoms and the
red algae.

On the top sides of the unharvested test screen, diatoms
decreased from 60% at 49 days to 41% at 64 days. Cyanophyta
increased in point count percentages from 7% at 49 days to 20% at
60 days. There was no significant change in composition of red
algae between 49 days (20%) and 64 days (19.6%). Rhodophyta

began to dominate the community, increasing from 23% at 49 days,
to 39% at 59 days, to 55% five days later. Macrophytes, such as
Laurencia sp. and Dasyopsis antillarium, dominated the
unharvested community for the remainder of the study, clearly
demonstrating that the relationships between harvested and
unharvested screens are similar to those of wild reefs.

Unfortunately, poor substratum material (1000 urn, single
layered screen) was chosen to model potential turf community
development. Due to the relative thinness (0.5 urn thick) of the
1000 um screen, algal holdfasts were frequently removed on
harvest, which resulted in a poorly developed canopy structure.

In the screen type and depth array studies, the multiweave screen
had a high surface area to screen thickness (2000 um) ratio which
allowed for sufficient unscraped surface area for holdfast
development. The lOOOum mesh screens maintained high diatom
concentrations. With each harvest the mat layer was almost
entirely removed from the substratum, except for small quantities
which were forced between the pores of the mesh. Thus, the
substratum was partly re-exposed and diatom colonization began
again .

Substratum Area and Type . The screen mesh size studies
conducted at site 1 indicated a relationship between substratum
type, algal community, and biomass production. Of the seven
single layered mesh types tested, the multiweave screen had the
highest mean production rates at 10.6 g dry m~2 d_1 (see Table

7) . While not significantly different (using ANOVA at 5% level)
than the multiweave screen, screening with a finer mesh (125,

500, 710 um) had a lower mean production (7.8, 7.7, 7.8 g dry m“2
d”l , respectively) . The course black screen (2x3 mm) produced at
levels close to the multiweave screen when in a single layer, but
produced at much higher levels when doubled and tripled.

Site 1 at Grand Turk was protected from the prevailing trade
winds. However, a shift in wind direction to the southeast would
result in carbonate silt and detrital matter settling on the top
sides of the screens. This silt coating hindered algal turf
development and growth, especially on the finer mesh screens, and
lowered production rates. The screens with greater porosity, had
less of a build-up of silt and so algal turf production was much
greater. Although mean production rates in the finer mesh screen
were lower during the initial harvests, ultimately they reached

production values equal to that of coarser screens for the same

, — 1

harvest period (200 um mean - 14.6 g dry m d ; multiweave mean

- 14.5 g dry m-2 d-1) .

Due to the silt problem at site 1, site 3 was chosen for a
multilayered (single, double, triple) screen study. There were
significant differences in biomass production between the single
layered screen and both the double and triple layered screens
(Scheffe Procedure at the 5% level) . Although no significant
differences were recorded between the double and triple layered

screens, the triple layered screen yielded a higher mean
production (17.8 g dry m“2 d”1) than the double layered screen
(14.7 g m“2 d*"1) .

Small, double layered, black (2x3 mm) screens were

introduced at site 3 to examine the possibility that edge to

surface area ratios affects the biomass production, possibly due

to water motion around the edges of artificial substrata

(Borowitzla, et al. , 1978). A 0.25 m2 screen averaged 5.1 g dry

— ? — 1 . .

md, while a similar screen of 0.75 m (introduced and
harvested at identical intervals) averaged 12.2 g dry m~2 d”1.
Biomass production based on edge to surface area ratios was to be
the inverse of what was expected.

Harvest Rate . In order to determine harvest rates at which
algal turf production could be optimized, biomass was collected
from the screens at varied intervals. The mean biomass produced
during 4, 7, 12 and 24 day harvest regimes was 5.3, 7.4, 7.6, 9.4
g dry m~2 d_1, respectively.

Depth and Light . In depth/production studies, mean biomass
yields increased with an increase in water depth from the surface

— O _ 1 . ...

(9.7 g dry m d ) up to 30 cm depth, with no significant
differences between 30, 40 and 100 cm depth (maximum 16.0 g dry
m”2 d_1; Table 9, Figure 19). The highest production values for
a given harvest occurred at 30 cm (29.5 g dry m~2 d”1) and 40 cm
(28.9 g dry m“2 d-1) . Mean biomass production on the top side of
the screen at 30, 40 and 100 cm water depth was not significantly

different, while the shallower screens placed at the surface and
10 and 20 cm below the water surface had lower biomass values on
average. Average underside production showed little variation
with depth (3.5-4. 6 g dry m-2 d”1)

The peak irradiance for incident and refected PAR, measured
over 13 h, from sunrise to sunset (0630 - 1830 HRS), occured
between 1130 and 1330, with incident light averaging 1700

, 9 , , 9

uEm/m^/sec. Reflected light averaged 200 uEm/m^/sec m 2.0 m
water depth, overlying a sandy bottom. Average incident PAR over

• 9

the 13 h was 986 uEm/m /sec and average refected PAR was 100

. 9 ...

uEin/m /sec. The spectral quality of available light over the
top one meter of the water column in which the depth/production
studies were carried out is shown in Figure 19.

At 30 cm water depth, incident PAR measurements were reduced

* 9

to 655 uEm/m /sec when made through a 1000 um mesh screen. PAR
was reduced further by 40%, when made through a similar screen,
with seven days of turf growth. Harvesting the screen did not
significantly change the incident PAR intensity ( 400uEin/m2/sec) ,
which passed through the screen.

Nutrients . Nutrient analyses were done in July and August
at Grand Turk. At site 1, mean nitrite/nitrate was 0.06 ug-at/1,
site 2 the mean averaged 0.25 ug-at/1 and at site 3 the mean was
0.12 ug-at/1. These results are consistent with the flow pattern
of equatorial currents over the reef and through the lagoon,

which gradually pick up nitrogen released from the lagoon
sediments. At all raft locations, orthophosphate concentrations
were less than 0.03 ug-at/1 below the detectable limits of the
method used.

Antigua . Algal community structure was not formally tallied
at Antigua. It was apparent, however, that the typical pattern
of successive diatom, blue-green, red algal dominance was
followed as a succession with time. However, it is the opinion
of all observers at Antigua that algal diversity on the screens
was markedly lower than at other sites.

Mean algal turf production with time is shown in figure 20
for both black polypropylene ( 2 X 3 mm mesh size) and white
polyester monofiliment (2mm mesh size) screens. The same data
are plotted with depth in figure 21. It is apparent that the
white screens show a significant increase in production
(approximately 15% increase). This difference appears
consistently when considering time, as well as depth. With
depth, there is a clear production peak levels at 30-40 cm, with
a sharp rise from surface values of about 75% of the peak and
then with a slow drop off in algal growth with greater depth.

Table 10 shows the results of the location, multilayer and
horizontal/vertical tests at Antigua. At dry weight production
means of 7.3 g m~2 d-1 and 4.2 g m-2 d-1, the protected sites of
Rat Island and Bird Island, respectively, are significantly below

. . — o — 1

the dry weight production mean of 18.2 g m d found m the
more open and turbulent Green Island and Devils Bridge sites.

The more protected sites were also characterized by gradual
sediment deposition and a general degeneration of production with
time to virtually zero values at 100 days.

The use of double-layered screens over single screens
significantly increases algal production, as has been
demonstrated at a number of sites. At Antigua, the increase in
production from single screens to double layered screens was over
50%. On the other hand, triple and quadruple layers of screening
reduced production. Algal growth was significantly lower when
the quadruple layered screens were employed.

The vertical single screens tended to produce at mean levels
of about 12% below that of the horizontal screens, which is
consistent with earlier studies. However, the variation with
time was typically large and, as a result, the difference is not
statistically significant. The screens that were placed
vertically developed an intermediate community between the
typical diatom dominated top and red/blue-green dominated bottom
communities .

As discussed below, apparent production levels were larger
at the Antigua Green Island site than at any other location.

This resulted in large measure from a high volume of included
sediment .

Carriacou . Thirty-five algal species were tallied on the
turf test screens at Carriacou (Table 11) . While this is more or
less typical of sites previously tallied, it is probably low
because of the relatively short time that work was carried out at
this site.

The productivity data collected at Carriacou are shown in
Table 12 and plotted with time in Figure 22. The pooled means
taken at each site included the small and vertical screens, since
they did not provide harvests that were significantly different
from the standard screens. The results from the tandem
arrangement are omitted from the means at sites 1 and 2, since
they provided mature harvests well below the stand alone screens.

The mean mature screen harvests for outer lagoon sites
(sites 1 and 3) with maximum flow and turbulence, are 16.3 and

— 9 — 1 . ...

14.7 g m . d respectively. These values are significantly
higher than other sites, except for Antigua with its heavy
sediment load, and reflect the greater turbulence levels in Grand
Bay. Wind levels (taken three times/day on the ship's
anemoneter) at Carriacou generally averaged above 15 knots.
However, during the 40-53 day interval, wind speeds averaged
below 10 knots. This is reflected in the general production dips
for sites 1 and 3 at 45 and 55 days. It is also probably
responsible for the apparent peaks in production on the island
apron sites due to settled sediment during the calm period.

The Island apron sites, with less turblence in general

a _ O „ 1

showed production levels of 12.9, 8.6 and 9.7 g m d , which
are significantly below those in the outer lagoon and reef apron.

Buen Hombre , Dominican Republic. The harvest production
levels achieved at Buen Hombre are shown in Table 13. The mean

— 9—1

value for all screens was 9.1 g m d , which is to be expected
from this moderately quiet site. Additional factors in the
moderate production level are a relatively short harvest interval
and vertically oriented screens. No significant difference
between the separate tests could be discerned. The prime
interest in the data is that it extends through the summer
period, a time when continuous data from other sites is not
available .

DISCUSSION

LABORATORY WORK

The results of the laboratory experiments suggest, that
given adequate light levels, of all the variables tested,
turbulence is the primary factor controlling biomass production
within the algal turf scrubbers. When provided with standard
flow (13.5 1/min . ) , scrubber biomass production rates ranged
above 20 g dry m-2 d”1, indicating carbon fixation rates of
over 6 g C m-2 d-1 . Assuming a 14 h light period and a set
proportion of carbon lost to dark respiration each day, the range
of these harvest data coincides well with 14C-based specific

These

productivity rate estimates of 0.15-0.64 g C m-2 h-1
biomass productivity rates were not significantly affected by the
tested ranges of variation for most of the factors in this study.
Water depth appears irrelevant to scrubber production so long as
screens are continually submerged, though increases past a
certain depth might interfere with turbulent mixing. There is
also some indication that water depth may control the impact of
UV toxicity in the field (see below) . Although self-shading
undoubtedly decreases rates of weight-specific productivity,
results here suggest that for a period of between one and two
weeks the exact harvest schedule is not critical to biomass
production. While longer scrubber light periods conceivably
could encroach upon the light cycle of the reef tanks, resulting
in increased competition for carbon dioxide and other nutrients
between scrubber and reef algae,, no such hinderance to production
was evident with the 2 h increase tested here. Scrubber light
levels tested here are approximately 60% of the values recorded
at levels of maxium turf production in the western tropical
Atlantic and are probably well below saturation levels
(Carpenter, 1985) . It would be be expected, therefore, that
increased irradiance would promote turf production in the
scrubbers. The lack of a discernable response in this study
probably reflects the narrow range of intensities tested, a
limitation of the light source employed.

In contrast, adjustment to flow rate appears to directly

affect biomass production within the scrubber environment by
controlling the frequency of wave generation. This conclusion is
based upon the disparate production values separating upstream
and downstream screens, the harvests contrasting production under
presence and absence of wave action, and the responses of
production to increased flow rate.

Within the confines of the scrubbers, the cyclical dumping
of the trough-like wave buckets is viewed most accurately as
generating three liter displacements of water which cross the
turf assemblage in the form of pulses, or waves. As this
displaced water travels across the scrubber, it encounters
friction in contacting algal filaments and quiescent water and
converts in part to an oscillating flow that travels across the
surface of the turf increasing disturbance of the boundary layer
(Nowell and Jumars, 1984).

Diffusion gradients form continually within the boundary
layers surrounding the thalii of metabolically active alga as a
result of the low coefficients of molecular diffusion in water
(Leyton, 1975; Norton et al. , 1982). Thus, any disturbance to

these boundary layers, particularly an oscillating disturbance,
will increase molecular exchanges between the plant cells and the
surrounding water, ultimately influencing metabolic rates. By
increasing the rates of simple, unidirectional currents in
laboratory cultures of various aquatic plants, investigators have

succeeded in reducing the thickness of boundary layers and
raising rates of respiration, photosynthesis, nutrient uptake,
and growth (e . g . , Dromgoole, 1978a, b ; Madsen and Sondergaard,
1983). However, turbulent flow, more typical of that encountered
in the field, is particularly effective for disruption of
boundary layers (Anderson and Charters, 1982). It is now
recognized that certain marine algae may even have adaptive
morphologies that serve to generate disruptive eddies as water
flows across their thalli (Anderson and Charters, 1982; Norton et
al . , 1982) .

Since the waves provide a turbulent flow that lessens in
intensity with passage, the metabolic efficiency of a turf alga
could be expected to vary with location in the scrubber. In
addition, because diffusion gradients undoubtedly are re¬
established between wave passages, production would be influenced
further by wave frequency, which results directly from flow rate.
The importance of magnitude and frequency of wave surge most
likely explains the significantly higher rates of production
observed generally on upstream screens, specifically when wave
action was restored.

In a preliminary examination of potential nutrient
limitation within a scrubber receiving the standard flow rate
(13.5 1/min) , the nitrate/nitrite concentrations measured in
upstream water filling the wave bucket and downstream water
entering the drain pipe demonstrated a constant drop from about

1.0 to 0.75 ug-at/1 . Similar measures of orthophospate failed to
detect significant differences. While such data indicate an
ability for rapid clearance of nitrogen, it appears doubtful that
the availability of these particular nitrogen forms is a major
controlling factor in turf biomass production. Based upon a 2.7%
nitrogen composition, the mean value for upstream production of
8.53 g dry m d represents a requirement of about 16.3 mmole
N/m /day . In comparison, under the standard flow rate and
photoperiod (14 h illumination/day ) , the nitrate/nitrite
clearance in the monitored scrubber indicates an uptake of only
about 4.7 mmole N/m /day. The plot of production values against
nitrate/nitrite concentrations also suggests that these species
may not represent the major source of nitrogen and may have
negligible control of biomass production within these scrubbers.
With the exception of the one point that represents the lowest
production and the highest nitrate/nitrite concentration, a
possible outlier value, these data form a nearly vertical plot,
indicating no relation between production and nitrogen
concentrations over the ranges of concentration studied. Should
the outlier be included within the analysis, an inverse relation
is suggested. A relationship which these data clearly do not
represent is the conventional interpretation of limitations to
marine productivity, i . e . , a decreased production under lower
nutrient levels.

The scrubber turf algae may rely predominately upon an
alternate source of nitrogen, ammonium, which has been measured
at levels of around 0.3 ug-at/1 within the microcosm. This
concentration, appreciably higher than those which typify shallow
Caribbean reefs (Meyer et al. , 1983; Williams, 1984), undoubtedly
reflects accumulation due to the closed system circulation and
probably serves as a major nutrient source that is present at
consistently high levels. However, an analysis of the specific
ammonium clearance within an individual scrubber has yet to be
performed. Fixation of atmospheric nitrogen by blue-green algae,
an important turf component, might provide another source that is
distributed through micrograzer activity or pathways. Should
such fixation occur, it probably could provide large amounts of
nitrogen to the assemblage, given the mixing enabled by shallow
depths and wave surge within the scrubbers.

The results of the paired harvests under varied flow rates
(Table 3) reguire some analysis to determine whether the wave
freguency or the nutrient supply limits production. Increased
harvests for both screens and egual productions between screens
were observed under the higher flow rate of the first test.

Under even more greatly increased flow in the second test,
harvests increased on downstream screens to equal the upstream
harvests under either flow rate. When considered together, these
results suggest that any limit to production by flow rate past
the upstream screens would be alleviated with a flow of around

13-16 1/min. The increase in downstream productions with 32.5
1/min regime to equal those upstream screens which remained
unchanged, indicate that nutrients were not a limiting factor for
production on the upstream screens at either flow rate in the
second test. Similarly, the lower downstream production under
the 13.5 1/min flow likely resulted from wave frequency that was
too low to maintain as effective a disruption of diffusion
gradients downstream, given the lowered turbulence reaching this
portion of the scrubber.

The third test of 13.5 and 33.5 1/min, with a smaller wave
bucket at the higher flow rate, resulted in no significant
increases for either screen. Again, production on upstream
screens may have neared the maximum for this scrubber environment
under a flow rate of 13.5 1/min. A wave bucket with 1/2 the
normal volume decreased the amount of disruption provided even
further, by each wave that reached the downstream turf. Such
decreased disruptive wave surge might have prevented a higher
production on downstream screens, even with the greatly increased
flow and resulted in the significantly lower value when compared
to the mean obtained under the same flow rate in the second test.
Thus, both volume and frequency appear to affect the disruptive
capacities of wave surge. As particularly suggested by the wave
generation test, where lack of surge flow nearly equalizes the
differences between upstream and downstream screens, the nutrient

supply aspect of flow rate does not appear as a significant
factor limiting production in the scrubbers.

While the results of the six week study that examined the
effect of intermittently blocked wave generation clearly support
the contention that wave surge is the primary controlling factor
in turf production, a decreased production nevertheless persisted
on the downstream screen despite the maintenance of constant flow
(13.5 1/min) . A close examination of flow patterns arising from
blocked wave buckets shows that they fail to provide a continuous
advance of seawater across the width of the scrubber, allowing
patches of reduced flow to develop over portions of the
downstream screen. These relatively stagnant areas develop as
the flow follows paths of least resistance towards the sides (or
unharvested portions) of the scrubbers and probably explains the
decreased production on downstream screens.

FIELD WORK

The mature algal turf community that develops on cultured,
harvested screens in turbulent Caribbean waters is quite diverse,
with over 30 species typically present. Except in the case of a
few widespread and common species such as Centroceras clavulatum .
Sohacelaria tribuloides . Polvsiphonia sphaerocarpa and Wranqelia
argus . the turfs do not generally have common species from site
to site. That this is a real difference, in that species have

been selected from stable but different population pools at each
site, seems unlikely. It seems more likely that the pattern is a
result of either widespread geographic and time variability in
turf species throughout the region, morphological variation that
makes accurate identification difficult, or real problems in
taxonomy at the species level. In any case, at the generic
level, there is considerable uniformity from site to site.

Genera that tend to be common on the screens at most sites are:
the diatoms, Licmophora and Navicula; the blue-greens, Anacvstis .
Oscillatoria and Schizothrix : the greens, Brvopsis . Cladophora .
Derbesia and Enteromorpha ; the browns, Gif fordia and Sphacelaria
and the reds, Callithamnium. Gentroceras . Ceramium, Polvsiphonia ,
Herposiphonia . Lophosiphonia . Wranaelia . Griff ithsia . and
Laurencia .

Diatoms are the primary early colonizers of algal turf
screens at all sites. In most cases, they are largely replaced
by blue-greens and early members of the canopy-forming reds and
greens by 30 days. After that time, the diatoms that remain are
primarily epiphytic. At Grand Turk, diatoms persisted as a major
biomass element for longer than other sites but this may relate
to the type of screen used for those particular tests and
overharvesting. The percentage diatom composition varied for
screen types tested, with benthic diatoms responsible for 30 to
50% of the biomass yielded. This suggests that diatoms can be as
productive in the tropics as they are considered to be in the

temperate zones. In a recent mangrove system study, Littler, et
al . (1985) reported that a gelatinous species of naviculoid

diatom was a major primary producer in the total community
productiviy. Further, certain diatom species have demonstrated a
resistance to ultraviolet light which could make them
particularly suited for growth on screens suspended near the
surface .

Perhaps the greater abundance of diatoms on the harvested
screens as compared to the unharvested test screens and the wild
reef was partially due to a shorter (4-12 days) harvest schedule
on the screens. Harvesting would expose new substratum
continuously, hinder red algal development, and allow benthic
diatoms to proliferate. Benthic diatoms which initially
colonized the screening also rapidly respond to harvesting
(grazing pressure) . Diatom dominance, although not
quantitatively sampled, was noted to be particularly obvious at
the four day harvest interval. Also, community structure tests
were performed on single layered screens which resulted in
overharvesting or, in ecological terms, high disturbance levels.

It would be difficult to quantify the biomass attributed to
the major algal groups as the screens develop without major time
expenditure. However, experience at all sites, with screen
appearance and in tallies on microscope slides, indicates that
there is an intermediate time from about 20-40 days in which

blue-green algae dominate in biomass. After that time, it is
typically the species of red algae that dominate in biomass as
well as diversity (Figure 23). It is almost invariably the case
that on the undersides of the horizontal screens the turf takes
on a dark red brown color. On the topside of the screen, diatoms
and blue-greens remain importent , resulting in a more light brown
color. The green and brown algae, although almost invariably
present on mature screens, are rarely important in terms of
biomass .

Production of algal turfs on screens, as a function of
depth, was examimed at Mayaguana, Grand Turk and Antigua. In all
cases, production levels at or near the water surface were lower
than at a depth of several tens of centimeters. At Mayaguana,
the peak production level was at 20-30 cm and at Antigua, it was
30-40 cm with lower production in deeper water. At Grand Turk,
however, no decline in production could be discerned up to a
depth of 1 m.

Reduced algal turf production at the shallowest depths, with
peak production occurring somewhat deeper, is to be expected
because of ultraviolet toxicity, as described in phytoplankton
studies. However, the tropical ocean/plankton production peak
typically occurs at 10-30 meters (Steemann-Nielsen, 1955) . For
example, in Oscillatoria thiebautii (a blue-green phytoplankton
in the Caribbean Sea) , peak production occurs between 300 and 600
uEin/m2/sec (Li et al . , 1980) , whereas in the algal turfs

. . . . o

investigated here, it occurred over 1600 uEin/i /sec. Carpenter
(1985) found similar results in chamber studies of coral reef
algal turfs.

It is apparent from observations of algal turfs in a wave
surge environment that the position of any given cell relative to
incoming light is constantly changing, and that in most cases,
individual cells are receiving widely varying light levels. The
apparent increased resistance to ultraviolet radiation, with a
concomitant ability to utilize most of the available visible
radiation, may derive in part from the shading/light-flashing
effects on individual algal cells in a regime of alternating wave
surge ( Falkowski , 1984). As we discussed above, with regard to

the laboratory studies and with further consideration below,
there is little question of the close association between high
algal turf production and turbulence, including both current and
wave surge .

Successive harvest production rates on the topsides of the
screens were least variable at 100 cm depth. However, production
in intense irradiance regimes ranging from the surface to 40 cm,
revealed a high variability in biomass production from harvest to
harvest. The turf community at less than 40 cm, while adapted to
intense PAR and relatively high UVB was perhaps less
photo synthetically efficient, and therefore, relatively more
sensitive to weather related fluctuation in irradiance values.

The two ends of the spectrum, red/infrared and ultraviolet A
+ B, show significant reductions in intensity in the upper 30-40
cm (Figure 19B) and this corresponds with lower topside biomass
production levels at the same depths. Recent work by Jokiel and
York (1984) suggests that ultraviolet light should be considered
as a photoinhibitor of primary productivity. However, Carpenter
(1985) reported minimal UV effects on coral reef algal turf
respiration rates. The algal community on screens may well have
developed a chromatic adaptation to ultraviolet light. Certain
species of diatoms (mostly from tropical waters) have been shown
to have a resistance to ultraviolet radiation (Jokiel and York
Jr., 1984), and several species of Rhodophyta (Tsujino and Saito,
1961; Yoshida and Sivalingam, 1970) and Cyanophyta (Shibata,

1969) contain ultraviolet absorbing substances.

. . — 9

At Mayaguana, a maximum algal production of 19.6 g dry m

—1 . . „9 —1

d was achieved with a mean of 12.6 g dry m d for 30 cm
depth, double layered screens. Applying the same approach to
Grand Turk, with similarly low nutrient concentrations, a maximum

. « 9 — 1

yield of 30.8 g dry m d was produced with mean biomass of
18.0 g dry m d . The increase in production at Grand Turk as
compared to Mayaguana, in large part derives from an increased
screen thickness and spatial heterogeneity, thereby preventing

overharvest .

Algal turf production, on individual screens, varied
significantly over time, apparently with no distinctive pattern,

and often exhibiting dramatic fluctuations in production from one
harvest to the next. However, production rates of the depth
array screens, at both Mayaguana, and Grand Turk, introduced at
the same site and harvested on the same days, had consecutive
peaks and troughs of production for individual screens. This
continuity of biomass production among screens, from one harvest
to the next, is quite similiar to that shown by simultaneously
harvested screens at all sites, and suggests a real and
consistent photosynthetic sensitivity to short term changes in
physical parameters such as irradiance and water motion.

When the biomass production rates of the topside and the
underside of the screens were considered separately, it is
evident that the topside produced a higher proportion of the
biomass values. For all screens tested at Grand Turk, the
underside was consistent, as it yielded a mean of 5 g dry m d
1 , with only a small deviation from the mean. Although the
species composition of the underside screen community was more
diverse, many of the same species were identified on both sides.
The species composition of turf algae cultivated on screening did
not seem to have a significant effect on the variations in turf
biomass production.

Visually, the topside of most horizontal screens can be
distinguished from the characteristically darker underside, which
receives less PAR (as refected and transmitted light) than the

topside. At 30 cm, with 1700=1800 uEin 2 sec being received

. . . — 9 — 1

on the top side of the screens, approximately 400 uEm sec

would be transmitted through a standard black screen to the
. . 8 —1

bottom side. With 200 uEin sec being reflected back from the
bottom at 4m (over a sand bottom) the bottoms of the screens are
receiving nearly 50% of that received on the topside. In the
Grand Turk experiments, where top and bottom production was
separated, the bottoms developed 50% of surface production at 20
cm depth and 30% at 30-40 cm depth. Algae can vary
photosynthetic pigment ratios or the total amount of pigments as
chromatic adaptation for various irradiance intensities. Under
reduced irradiance, the plants are more photosynthetically
efficient and yeild lower but more consistant production values ,
regardless of long term irradiance fluctuation.

At Nonsuch Bay, Antigua, a 100 day study provided a
consistent and significantly higher level of production on white
translucent screens, as compared to the standard black screens.
Based on the above discussion, this is clearly the result of
increased transmission of PAR through the substratum screen.

In the screen type study at Grand Turk, both single-layered
screens and finer mesh screens produced significantly less than
multilayered and complex mesh screens. As discussed above,
overharvesting as a result of insufficient substratum surface on
single-layer screens and therefore protection for algal rhizoids
and attachments seems to be clearly responsible. On the other

hand, the finer mesh screens tend to accumulate fine carbonate
sediment, which may initially result in relatively large apparent
harvests. Eventually however, depending upon the availability of
wave energy to remove the superficial sediment settling on the
screens, sedimentation reduces turf production to low levels and
in some cases to virtually no algal production at all.

The most significant and obvious difference in algal harvest
at Mayaguana was that observed between ocean and lagoon rafts.
Average production of the mature double lagoon screens at 30 cm
depth was 13.8 g dry m“2 d-1, over 2 1/2 times the ocean double

. <— O -1

screen production of 5.0 g dry m d for the same depth. The
"ocean array" system likewise achieved lower production levels.

As mentioned above, nitrogen concentrations tested both in
our laboratory and in an independent laboratory were in the range
of 0.100-0.130 ug-at/1 over a period of several weeks in
April/May. No significant difference between ocean and lagoon
could be found. While it is possible that another source of
nitrogen is available in the lagoon (e.g. , NH4+) , it is unlikely
to be a limited factor independent of phosphorus, particularly
considering that abundant nitrogen fixation on these screens is
quite likely. It is also unlikely that another source of
moderate to elevated nitrogen would be available in the absence
of higher concentrations of nitrate.

At Mayaguana, "ocean"/shelf sea conditions are only moderate

for trade wind seas because of the partial protection offered by
the East/West trending island. It is unfortunate that no open
ocean or open shelf production data are available at other sites
to assess offshore trade wind sea production levels. However,
what is most important in this context is that at Mayaguana, the
trade wind sea, when driven over the reef by wave action, greatly
increases in turbulence and in flow rate, due to depth
compression.

In the course of the three year investigation, some current
meter readings were taken with a Marsh McBirny , electromagnetic
meter that measures both current and surge. However, without
continuous readings over many months, energy guantif ication is of
little value. On the other hand, the very process of working
algal screen lines on a daily basis provides a qualitative
measure of energy levels available at each site. Thus, the
direct relationship between production and wave and current
energy, as shown in figure 24, can be regarded as a reasonable
semi-quantitative measure. The single anomalously high value for
Green island, Antigua deserves further comment.

The harvest biomass of maximum production screens (lagoon
doubles), at Mayaguana, had a mean of 15.5% carbon (C) (S.D =
±1.97; N = 22), based on C/H/N analysis. This compares to a
figure of 19.8% (S.D. = +3.62; N = 14) for all turf algae
(including blue-greens, greens and browns, but not diatoms)
provided by Atkinson and Smith (1983). The somewhat lower carbon

value for Mayaguana could be related to sediment inclusion,
though little could be seen on microscopic examination of turf
samples. On the other hand, the Mayaguana "ocean" equivalent
screens had a mean carbon level of 14.7% (S.D. = +0.71; N = 8) ,
not significantly below the lagoon value. If sediment were a
contributing factor to apparent production in this case, and it
might be argued that this is the reason for the lower "ocean"
production, "ocean" carbon values would be significantly higher.
Blue-green algae, with their largely proteinaceous cell walls,
are a major component of algal turfs; this is undoubtedly
partially responsible for the relatively high protein content (8-
10%) of algal turfs. Likewise, diatoms with siliceous walls are
major elements of the turfs. With these factors in mind, the
relatively low carbon percentages do not appear unreasonable. It
is likely that insufficient drying and/or dry storage aboard ship
for several months in the tropical climate provided for some
fungal and bacterial respiration losses in the samples. The
upstream, mature laboratory algal turfs subjected to C/H/N
analysis provided much higher carbon percentages (C = 27.9%; S.D.
= +1.29; N = 16), even at relatively low ambient nitrogen levels
(near 1 uM) . Diatoms are a much smaller part of the laboratory
turf flora, and this is likely partially responsible for the
higher carbon percentages; laboratory levels are also well above
the values obtained by Atkinson and Smith (1983) for field turf

species. However, a general preparation problem of algal turfs
with regard to low carbon percentage cannot be ruled out. It is
to be further noted that the data of Atkinson and Smith show no
significant difference for carbon percentage in algal turf and
small algae for low nutrient and high nutrient waters.

The carbon percentage values provided by Atkinson and Smith
(1983) show means of 19.8% for algal turfs, 30.3% for small macro
algae (e.g., Ulva . Dictvota . and Hypnea) and 33-35% for large
algae (e.g. , S areas sum ) and marine flowering plants. The
relatively low carbon values probably result in part from less
structural carbon in walls (cellulose and its algal relatives) .
They may also result in part from greater difficulty in
perserving organics intact (considering the larger surface areas)
until analysis is performed. Laboratory turf samples with a
quite similar algal community, though perhaps significantly
lacking in diatoms, showed carbon percentages close to that for
macroalgae. The subject needs further investigation.

Unfortunately C/H/N analyses were not performed on the
Antigua harvests. It is clear from visual and microscopic
observations that levels of carbonate sediment in the Green
Island samples were high, and this could account for the
anomalously high weights. On the other hand, it has been our
experience from other sites (Grand Turk, Carriacou) that where
low energy and a corresponding increased sediment deposition is
involved, that algal production is very low. Also, it was found

that screens placed vertically as compared to screens placed
horizontally (which should greatly reduce sediment settlement) at
the same site at Antigua, provided harvests only about 12% lower.

Nonsuch Bay, Antigua, unlike all other sites, has no ready
lagoon outflow down wind. Waters that are driven over the reef
with their accumulated carbonate sediment as well as water and
sediment derived from land run-off, must largely depart from the
Bay through two channels, both trending eastward into the
tradewind sea. As a result, the waters of Nonsuch Bay are
generally more turbid than other sites that were utilized in this
investigation. Even though wave energies and current flow are
moderate at the Green Island reef site, suspended fine sediment
is typically high. Thus, it is tentatively concluded that
increasing sediment loads generally reduce algal turf production
(real and apparent) in cultured screen situations. Where both
moderate wave and current energy are available and high loads of
fine sediment also are present, sediment is driven into and
imbedded in the basal, blue-green and diatom, mucilagenous layers
that comprise the algal turf community. As long as the sediment
is fine-grained and moderate in abundance, turf production is not
significantly decreased, but simple dry weight algal production
is increased. Judging by the considerable difficulties in
sustaining the growth of grazing crabs and in reducing crab
mortality at this site, as compared to other sites, the sediment

load in the screens under these special circumstances appeared to
provide a significant problem for these animals. This matter is
discussed in greater detail under Mariculture (this volume) .

General Discussion

Unfortunately, quantitive algal turf production work was
undertaken only during the Winter, Spring and Summer periods,
between the months of February to August. Based on the coral
reef production work of Adey and Steneck (1985) at St. Croix,
reductions in algal turf production are to be expected during the
relatively cloudy calm and radiation minimum autumn period.
Qualitatively, this early "winter" reduction was experienced at
several sites, in terms of the utilization of turf production to
feed grazing crabs. However, a more interesting characteristic
of the long term production of algal turfs on plastic screens,
which cannot be substantiated by quantitative studies at this
time, is the apparent tendency for continuously used screens at
some sites to accumulate coralline algal crusts and sometimes
bryozoans, anemones and ascidians. Unfortunately, this results
in a lowered algal production. This problem appeared to be more
characteristic of the lower energy, more turbid sites (Antigua,
Buen Hombre) than the clearer more turbulent areas. Brief drying
and re-stringing of the turf screen lines quickly solves the
problem. In addition, at all sites, continuous harvest, in the
10-20 day range, is essential to keep the screens highly

productive with algal turfs. Screens, un-harvested for several
months tend to develop a patchy macroalgal growth, often
including species characterized by toxic compounds (e.g.,
Laurencia . Acanthophora) and by significantly increased attached
epifauna. These screens have greatly lowered production rates
(presumably equivalent to that of a mature forest community) and
greatly reduced value for feeding grazers (much as results when a
forest replaces grazing farm land) . Unfortunately, it is very
difficult to return such screens to mature algal turf production
levels. Thorough drying and scraping is the only method that we
recognize at this time.

In cases where relatively low algal production was
experienced, variability with time was also considerably less.
Also, production rates often continued to slowly rise throughout
these studies. This phenomenon occurred at the deep ocean sites
at Mayaguana, the protected sites at Carriacou and the moderate
energy situations at Buen Hombre. The reasons for these
relationships and the level of the eventual peak values that
could be obtained under these conditions are not known and should
be further investigated.

CONCLUSIONS

There is little question that given adequate wave and
current motion and intense solar irradiance, reef algal turf
communities on artificial screen-type substrata, in nutrient poor

seas, can consistently produce at harvest rates of 8 to over 15 g

_ O _ 1 , ...

dry m d . A number of physical and biotic variables are
involved in determining the level of this production. These
include: 1) overharvest potential as related to substratum type
and harvest frequency; 2) underharvest potential as related to
colonization by macroalgae and animals; 3) water depth, screen
type and screen orientation as related to ultraviolet and/or
perhaps infrared radiation and to the transmittance and
reflectance of light and 4) sedimentation level and quality as
related to wave action.

While mat-forming benthic diatoms are the primary colonizers
of these substrata and remain important elements in all cases,
mat-forming blue-green algae and finally "canopy"-forming red
algae become critical biomass-producing elements as the screens
mature. Red algae become the dominant producers in the presence
of some shading and where short term harvesting is not carried
out.

Wild, screen-cultured algal turfs are highly diverse with
approximately 30-40 species of algae typically being significant
elements of the community as it develops and matures. This
diversity could very well be of critical importance to the
stability of biomass production for mariculture, as it is
unlikely to be disturbed significantly by disease or micro
predation. The tendancy for human agriculture to heavily pursue

monocultures and to use a battery of modern tools to reduce
genetic diversity has considerable implications with regard to a
general sensitivity to large scale crop failure.

Carbon percentages based on C/H/N analyses are relatively
low in field algal turfs as compared to laboratory algal turfs,
macro algae and flowering marine plants. This relates only
partially to included carbonate silt levels. Algal turfs are
rich in blue-green algae with proteinaceous walls and are
therefore relatively rich in protein as compared to cellulose.
Dried algal turfs arre also rich in silica as a result of high
diatom abundances. Sample preparation and preservation may also
be a critical factor. The very nature of the algal turf, in
which species compete by rapid growth and reproduction with
little investment in structure or defense through armor or toxic
compounds, suggests a composition subject to rapid breakdown by
fungi and bacteria and alternatively easily available for
utilization as a food source. Also, many grazing animals,
including the Mithrax crab, discussed in the depth in this
volume, appear to prefer algal turfs and are adapted to their
harvest, even when in the wild this results in a digestive system
that is rich in sediment and detritus.

Most important, numerous and repeated successful broodings
of Mithrax females utilizing algal turfs have occured in this
study (see Reproduction, this volume). In addition, and probably
more critical, Mithax soinsosissimus fed pure algal turfs or

algal turfs supplmented with small quantities of wild macro algae
in culture have provided considerably higher growth rates (see
Morphology, this volume) than crabs fed on diets of macro algae,
meat and commercial feeds (Ryther, et al. , 1987) This indicates
that algal turf is a rich food source for this predominantly
herbivorous animal and is likely a similarly rich food source for
other reef grazers.

LITERATURE CITED

Adey , W.H. 1978. Algal ridges of the Caribbean Sea and West
Indies. Phvcoloqia . 17: 361-367.

Adey, W.H. 1982. Algal Turf Scrubber. U.S. Pat. Doc. 4,333,263

Adey, W.H. 1983. The microcosm: A new tool for reef research.
Coral Reefs, 1: 193-201.

Adey, W.H. 1987a Food production in low nutrient seas: bringing
tropical ocean deserts to life, Bioscience 37(5).

Adey, W.H. 1987b. Marine Microcosms In Jordan, W. , J. Aber and
M. Gilpin (eds.) Restoration Ecology: Progress toward a
science and art of ecological healing. 1987.

Adey, W.H., P. Adey, R. Burke and L. Kaufman. 1977. The

Holocene reef systems of eastern Martinique, French West
Indies. Atoll Res. Bull., 87:95-109.

Adey, W.H. and R. Burke. 1976. Holocene Bioherms (algal ridges
and bank barrier reefs) of the eastern Caribbean. Geol.

Soc. Am. Bull., 87:95-109.

Adey, W.H., C. Rogers, R. Steneck and N. Salesky. 1979. The
south St. Croix reef. Report to the Department of
Conservation and Cultural Affairs, U.S. Virgin Islands.

64pp. and Appendix.

Adey, W.H. and R.S. Steneck. 1985. Highly productive eastern
Caribbean reefs: synergistic effects of biological,
chemical, physical and geological factors. In: M.L. Reaka
(ed.). The ecology of coral reefs. Symposia Series for
Undersea Research, Volume 2. National Oceanic and
Atmospheric Administration, Rockville, Maryland, U.S. A.

Adey, W.H. and T. Goertemiller . 1987. Coral reef algal turfs:

master producers in nutrient poor seas. Phycologia 26(3):
374-386.

Anderson, S.M. and A.C. Charters. 1982. A fluid dynmics study
of seawater flow through Gelidium nudifrons . Limnol.
Oceanogr. , 27(3): 399-412.

Andrews, J. and P. Gentien. 1982. Upwelling as a source of

nutrients for the Great Barrier Reef ecosystem: a solution
to Darwin's question. Mar. Ecol. Prog. Ser. , 8: 257-269.

Andrews, J. and H. Muller. 1983. Space-time variabilities of

nutrients in a lagoonal patch reef. Limnol . Oceanogr. , 28:
215-227.

Atkinson, M. 1981. Phosphate flux as a measurement of coral reef
productivity. In Proceedings of the Fourth International
Coral Reef Symposium. The Great Barrier Reef Committee,
Brisbane, Australia. 1: 417-418.

Atkinson, M. and R. Grigg. 1984. Model of a coral reef

ecosystem. II. Gross and net primary production at French
Frigate Shoals, Hawaii. Coral Reefs, 3: 13-22.

Atkinson, M. and S. Smith. 1983. C:N:P ratios of benthic marine
plants. Limnol. Oceanogr., 28: 568-574.

Bakus, G.J. 1967. The feeding habits of fishes and primary

production at Eniwetok, Marshall Islands. Micronesica, 3:

135-149 .

Borowitzka, M. A. 1981. Algae and grazing in coral reef
ecosystems. Endeavor, 5: 99-106.

Borowitzka, M.A. , A.W.D. Larkum and L.J. Borowitzka. 1978. A
preliminary study of algal turf communities of a shallow
coral reef lagoon using an artificial substratum. Aquat.

Bot., 5: 365-381.

Brawley, S.H. and W.H. Adey. 1977. Terittorial behavoir of

threespot damselfish ( Eupomacentrus planif rons) increases
reef algal biomass and productivity. Environ. Biol. Fishes,
2(1): 45-51.

Brawley, S.H. and W.H. Adey. 1981. The effect of micrograzers
on algal community structure in a coral reef microcosm.
Marine Biology. 61: 167.

Bunt, J.S. 1975. Primary productivity of marine ecosystems. In
H. Lieth and R.H. Whittaker (Eds.), Primary Productivity of
the Biosphere . Springer-Verlag, New York, pp. 169-183.

Carpenter, R.C. 1981. Grazing by Diadem antillarum (Philippi)
and its effects on the benthic algal community. J. Mar.

Res. , 39(4) : 749-765.

Carpenter, R.C. 1984. Herbivores and herbivory on coral reefs:
Effects on algal community biomass, structure and function.
Ph.D. thesis, Univ. Georgia. 175pp.

Carpenter, R.C. 1985. Relationships between primary production
and irradiance in coral reef algal communities. Limnol.
Oceanogr. , 30: 784-793.

Carpenter, R.C. 1986. Partitioning herbivory and its effects on
coral reef algal communities. Ecol. Monogr . 56: 345-363.

Connor, J.L. and W.H. Adey. 1977. The benthic algal

composition, standing crop and productivity of a Caribbean
algal ridge. Atoll Res. Bull., 211: 1-40*

Dahl, A . L. 1972. Ecology and community structure of some

tropical reef algae in Samoa. In K. Nisizawa (Ed.), Proc.
7th Int. Seaweed Symp. 1: 36-39. New York: J. Wiley and
Sons, Inc.

Dethier, M.N. 1981. Heteromorphic algal life histories: The
seasonal pattern and response to herbivory of the brown
crust, Ralf isia californica . Oecologia (Berl . ) , 49: 333-
339 .

Dromgoole, F.I. 1978a. The effects of oxygen on dark
respiration and apparent photosynthesis in marine
macroalgae. Aquat. Bot. , 4: 281-297.

Dromgoole, F.I. 1978b. The effects of pH and inorganic carbon
on photosynthesis and dark respiration of Carpophvllum
(Fucales, Phaeophyceae) . Aquat. Bot. 4: 11-22.

D'Elia, C. , K. Webb and J. Porter. 1981. Nitrate-rich ground

water inputs to Discovery Bay, Jamaica: a significant source
of nitrogen to local reefs? Bull. Mar. Sci., 31: 903-910.

Eppley, R. 1982. The PRPOOS program: a study of plankton rate
processes in oligotrophic oceans. EOS., 63: 522-523.

Falkowski, P. 1984. Physiological response of phytoplankton to
natural light regimes. Journal of Plankton Research, 6:
295-307.

Feldmann, J. 1966. Les types biologiques d'alques marines

benthiques . Mem. Soc. Bot. Fr. (Collogue de Morphologie,
1965) , pp. 45-60.

Fenwick, G.D. 1976. The effect of wave exposure on the amphipod
fauna of the algal Caulerpa brownii . J. Exp. Mar. Biol.

Ecol . , 25 : 1-18 .

Gerard, V.A. and K.H. Mann. 1979. Growth and production of

Laminaria lonqicruris (Phaeophyta) populations exposed to
different intensities of water movement. J. Phycol., 15:

33-41.

Goldman, J. , J. Ryther and L. Williams. 1975. Mass culture of
marine algae in outdoor cultures. Nature, 254: 594-595.

Gordon, M.S. and H.M. Kelly. 1962. Primary production of an
Hawaiian coral reef: A critique of flow respirometry in
turbulent waters. Ecology, 43: 473-480.

Hackney, J.M. 1984. The impact of photorespiration on the

productivity of a coral reef algal turf community. Ph.D.
thesis, Georgetown University. pp. 349.

Hatcher, B.G. and A.W.D. Larkum. 1983. An experimental

analysis of factors controlling the standing crop of the
epilithic algal community on a coral reef. J. Exp. Mar.
Biol. Ecol., 69: 61-84.

Hay, M.E. 1981. The functional morphology of turf-forming
seaweeds: persistence in stressful marine habitats.

Ecology, 62(3): 739-750.

Hay, M.E., T. Coburn and D. Downing. 1983. Spatial and temporal
patterns in herbivory on a Carribean fringing reef: the
effects on plant distribution. Oecologia (Berlin), 58: 299-
308 .

Hixon, M. A. and W.N. Brostoff. 1981. Fish grazing and community
structure of Hawaiian reef algae. In. Proceedings, 4th Int.
Symp. on Coral Reefs, Vol 2. Committee on Coral Reefs, Int.
Assoc, of Biol. Oceanogr. , Manila, pp. 21-25.

Jokiel , P. and R. York. 1984. Importance of ultraviolet

radiation in photoinhibition of microalgal growth. Limnol .
Oceanogr. 29: 192-199.

Kennelly, S.J. 1983. An experimental approach to the study of
factors affecting algal colonization in a sublittoral kelp
forest. J. Exp. Mar. Biol. Ecol., 69: 257-276.

Khailov, K.M. and Z.P. Burlakova. 1969. Release of dissolved

organic matter by marine seaweeds and distribution of their
total organic production to inshore communities. Limnol.
Oceanogr., 14: 521-527.

Kinsey, D. 1979. Carbon turnover and accumulation by coral
reefs. PhD. thesis, Univ. Hawaii, pp. 248

LaPointe, B. and K. Tenore. 1981. Experimental outdoor studies
with Ulva fasciata . I. Interaction of light and nitrogen
on nutrient uptake, growth and biochemical composition. J.
Exp. Mar. Biol. Ecol . , 53: 135-152.

Lewis, J, B. 1977. Processes of organic production on coral
reefs. Biol. Rev. Cambridge Philos. Soc. , 52: 305-347.

Leyton, L. 1975. Fluid Behavoir in Biological Systems .

Clarendon Press, Oxford, pp. 235.

Li, W. , H. Glover and I. Morris. 1980. Physiology of carbon
photoassimilation by Oscillatoria thiebautii in the
Caribbean Sea. Limnol. Oceanogr. 25: 447-456.

Littler, M.M. and K.E. Arnold. 1980. Sources of variability in
macroalgal primary productivity: Sampling and interpretative
problems. Aquat. Bot. , 8: 141-156.

Littler, M.M. and D.S. Littler. 1980. The evolution of thallus
form and survival strategies in benthic marine macroalgae:
Field and laboratory tests of a functional form model. Am.
Nat. , 116: 25-44.

Littler, M.M. , D.S. Littler and P.R. Taylor. 1983. Evolutionary
strategies in a tropical barrier reef system: Functional-
form groups of marine macroalgae. J. Phycol . , 19: 229-237.

Littler, M.M. , P. Taylor, D. Littler, R. Sims and J. Norris.

1985. The distribution, abundance and primary productivity
of submerged macrophytes in a Belize barrier-reef mangrove
system.

Lobel, P.S. 1980. Herbivory by damsel fishes and their role in
coral reef community ecology. Bull. Mar. Sci . , 30: 273-289.

Lubchenco, J. 1980. Algal zonation in the New England rocky

intertidal community: An experimental analysis. Ecology, 61
333-344.

Madsen, T.V. and M. Sondergaard. 1983. The effects of current
velocity on the photosynthesis of Callitriche stagnalis
Scop. Aqua. Bot., 15: 187-193.

Mann, K. 1973. Seaweeds: their productivity and strategy for
growth. Science, 182: 975-981.

Mann, K.H. 1982. Ecology of Coastal Waters . University of
California Press, Berkeley. pp.322.

Marsh, J.A. , Jr. 1976. Energetic role of algae in reef
ecosystems. Micronesica, 12: 13-21.

McNaughton, S.J. 1984. Grazing lawns: Animals in herds, plant
form and coevolution. Am. Nat., 124: 863-886.

Meyer, J.L., E.T. Schultz and G.S. Helfman. 1983. Fish schools:
An asset to corals. Science, 220: 1047-1049.

Morrissey, J. 1980. Community structure and zonation of

macroalga and hermatypic corals on a fringing reef flat of
Magnetic Island (Queensland, Australia). Aquat. Bot. , 8:
91-139.

Nasr, A . H . 1946. The biological forms of some marine algae from

Ghardaqa. Bull. Inst. Egypte. 28: 203-213.

Neushul, M. and A.L. Dahl. 1967. Composition and growth of

subtidal parvosilvosa from Californian kelp forests. Helg.
Wiss. Meeresunters . , 15: 480-488.

Norton, T.A. , A.C. Mathieson and M. Neushul. 1982. A review of
some aspects of form and function in seaweeds. Bot. Mar.,
25(11) : 501-510.

Nowell, A.R.M. and P.A. Jumars. 1984. Flow environments of
aquatic benthos. Ann. Rev. Ecol. Syst. , 15: 303-328.

Odum, H.T. and E.P. Odum. 1955. Trophic structure and

productivity of a windward reef coral community on Eniwetok
Atoll. Ecol. Monogr. , 25: 291-320.

Ogden, J.C. 1976. Some aspects of herbivore-plant relationships
on Caribbean reefs and seagrass beds. Aquat. Bot., 2: 103-
116.

Ogden, J.C., R. Brown and N. Salesky. 1973. Grazing by the
echinoid Diadema antillarum Phillipi: Formation of halos
around West Indian patch reefs. Science, 182: 715-717.

Ogden, J.C. and P.S. Lobel. 1978. The role of herbivorous
fishes and urchins in coral reef communities. Environ.

Biol. Fishes, 3: 49-63.

Pellew, R.A.P. 1983. The impact of elephant, giraffe and fire
upon the Acacia tortilis woodlands of the Serengeti. Afr.

J. Ecol . , 21: 41-74.

Peyton, K. , K. Moller and W. Adey. (In Review) Community

structure, development and biomass production of algal turfs
grown on artificial substrata in an oligotrophic sea.
Botanica Marina.

Randall, J.E. 1961. Overgrazing of algae by herbivorous marine
fishes. Ecology, 42(4): 812.

Ryther, J.H. 1959. Potential productivity of the sea. Science,
130: 602-608.

Ryther, J.H. 1969. Photosynthesis and fish production in the
sea. Science, 166: 72-76.

Ryther, J.H, R. Winfree, J. Holt, R. Creswell , W. Lellis , J.
Chaiton, C. Kovach and F. Prahl . 1987. Antigua Crab

Mariculture, Annual Progress Report. Harbor Branch
Oceanographic Institution, Fort Pierce, Fla; July 15, 1987.

7 8 pp .

Sammarco, P.W. 1982. Effects of grazing by Diadema antillarum
Philippi (Echinodermata : Echinodea) on algal diversity and
community structure. J. Exp. Mar. Biol. Ecol., 65: 83-105.

Sammarco, P.W. 1983. Effects of fish grazing and damselfish
territorality on coral reef algae. I. Algal community
structure. Mar. Ecol. Prog. Ser. , 13: 1-14.

Sammarco, P.W., J.S. Levinton and J.C. Ogden. 1974. Grazing and
control of coral reef community structure by Diadema
antillarum Philippi (Echinodermata: Echinoides) : A
preliminary study. J. Mar. Res., 32(1): 47-53.

Setchell, W . A . 1924. American Somoa. Part 1: Vegetation of

Tuila Island. Carnegie Inst. Washington Publ., 341: 1-188.

Sieburth, J. McN. and A. Jansen. 1969. Studies on algal

substances in the sea. III. The production of extracellular
organic matter by littoral marine algae. J. Exp. Mar. Biol.
Ecol., 3: 290-309.

Smant-Froelich, A. 1985. Functional aspects of nutrient cycling
in coral reefs. NOAA Symp. Ser. for Undersea Res., 1(1):
133-139.

Smith, S. and J. Marsh. 1978. Organic carbon productioon on the
windward reef flat of Eniwetok atoll. Limnol . Oceanogr. , 18:
953-961.

Sournia , A. 1977. Analyse et bilan de la production primaire
dans les recif coralliens. Ann. Inst. Oceanogr. (Paris)
53(1): 47-74 .

Spotte, S. 1979. Seawater Aquariums , The Captive Environment .
John Wiley and Sons, New York. pp. 413.

Steemann Nielsen, E. 1955. Production of organic matter in the
oceans. J. Mar. Res., 14: 374-386.

Steneck, R.S. and L. Watling. 1982. Feeding capabilities and
limitation of herbivorous molluscs: A functional group
approach. Mar. Biol. (Berlin), 68: 299-319.

Stephenson, T . A. and A. Stephenson. 1972. Life Between

Tidemarks of Rocky Shores . W.H. Freeman, San Francisco,
pp. 425.

Stephenson, W. and K.B. Searles. 1960. Experimental studies on
the ecology of intertidal environments at Heron Island. I.
Exclusion of fish from beach rock. Aust. J. Mar.

Freshwater Res., 11; 241-267.

Stewart, G. 1982. Anchor species and epiphytes in intertidal
algal turf. Pac. Sci . , 36 (1): 45-59

Tangley, L. 1985. And live from the East Coast - a miniature
Maine ecosystem. Bioscience 35: 618-619.

Taylor, P.R. , and M.E. Hay. 1984. Functional morphology of

intertidal seaweeds: Adaptive significance of aggregate vs.
solitary forms. Mar. Ecol. Prog. Ser. , 18: 295-302.

Tsujino, I. and T. Saito. 1961. Studies on the compounds

specific for each group of marine algae. I. Presence of
characteristic ultraviolet absorbing material in
Rhodophyceae . Bull. Fac. Fish. Hokkaido Univ. 12: 39-58.

Van Slyke, D.D., J. Plazin, and J.R. Weisiger. 1951. Reagents
for the Van Slyke-Folch wet carbon combustion. J.Biol.
Chem., 191: 299-304

Wanders, J.B.W. 1976a. The role of benthic algae in the shallow
reef of Curacao (Netherlands Antilles). I. Primary
productivity in the coral reef. Aquat. Bot. , 2: 235-270.

Wanders, J.B.W. 1976b. The role of bethnic algae in the shallow
reef of Caracao (Netherlands Antilles). II. Primary
productivity of the Saraassum beds on the north-east coast
submarine plateau. Aquat . Bot. , 2: 327-335.

Wanders, J.B.W. 1977. The role of benthic algae in the shallow
reef of Curacao (Netherlands Antilles). III. The
significance of grazing. Aquat. Bot., 3: 357-390.

Wheeler, W.N. 1980. Effect of boundary layer transport on the
fixation of carbon by the giant kelp Macrocvstis pyrifera .
Mar. Biol. (Berl.), 566: 103-110.

Wheeler, W.N. 1982. Response of macroalgae to light intensity,
light quality, temperature, C02 , HC03-, 02, mineral
nutrients, and pH. IN A. Mitsui and C.C. Black (Eds.), CRC
Handbook of Biosolar Resourses , Vol . 1 , Part 1 , Basic
Principles . CRC Press, Boca Raton (Florida), pp. 157-1844.

Wiebe, W. , R. Johannes and K. Webb. 1975. Nitrogen fixation in
a coral reef. Scoence, 188: 257-259.

Wilkenson, C. and P. Sammarco. 1983. Effects of fish grazing
and damselfish territoriality on coral reef algae. II.
Nitrogen fixation. Mar. Ecol. Prog. Ser. , 12:15-19.

Williams, S.L. 1984. Uptake of sediment ammonium and
translocation in a marine green macroalga Caulerpa
cupressoides . Limnol . Oceanogr. , 29(2): 374-379.

Williams, S.L., and W.H. Adey. 1983. Thalassia testudonum Banks
ex Konig seedling success in a coral reef microcosm. Aquat.
Bot., 16: 181-188.

Yoshida, T and P. Sivalingam. 1970. Isolation and

characterization of 337 mu UV-absorbing substances in the
red alga Porphvra vezoensis . Plant and Cell Phusiol . , 11:
427-434.

FIGURE 1.

FIGURE

LEGENDS

Typical laboratory/mesocosm algal turf growing on a
one mm2 mesh screen. The plastic strands of the
screen are 330 um in diameter. Present on this
screen, as drawn are: the diatoms, Licmophora sp.
and Navicula sp. ; the blue greens, Anacystis
dimidiata . Calothrix Crustacea and Qscillatoria
submembraanacea ; the greens, Cladophora
fascicularis . Smithsoniella earleae and Derbesia
vaucheriaeformis ; the browns, Ectocarpus
rhodochortonoides and Sphacelaria tribuloides ; and
the reds, Asterocvtis ramosa . Ceramium corniculatum ,
Polvsiphonia havanensis and Herposiphonia secunda .

2. Oxygen concentration of seawater undergoing direct,
continuous recirculation through an algal turf
scrubber versus elapsed time of incubation. Note
change in rate of oxygen production as wave action
blocked and restored.

3. Location of Mayaguana Island, Bahamas.

4. Abraham Bay, Mayaguana research site. Depths in
fathoms show sharp drop off near rafts. Local trade
winds and currents constantly bring ocean water to
the experimental site.

5. Standard pvc pipe raft used to hold screens for algal
culture. The screen dimensions are 1 m2 and the PVC
pipe is 7.6 cm dia.

6. "Ocean Array" of 1/4 m screens established to
determine rates of algal production with depth.

7. Grand Turk shelf, reef and research sites. See
figure 3 for the location of Grand Turk.

8. Eastern Caribbean sea, showing islands with field
research sites.

9. Detail of Nonsuch Bay, Antigua study sites 1-4.

10. Detail of Grand Bay, Carriacou and field study sites
1-5.

11. North Coast of Monte Christi Province, Dominican
Republic, showing the location of the Barrier Reef,
Buen Hombre and the algal research site.

R. 6

FIGURE

12 . Scrubber biomass production rates for one week
harvest periods under the presence or absence of
disruptive wave surge. Production rates indicated
for upstream (triangles) and downstream (circles)
screens within one scrubber that had wave action
blocked during alternate weeks.

13. Scrubber biomass production rates for 12 day harvest
periods vs. N-(N02 + N03) concentrations in
circulating microcosm water. Values plotted for
upstream screens (triangles) and downstream screens
(circles) for one scrubber. Values correspond to
average of nitrogen concentrations measured during
harvest period. Range in measured concentrations
indicated, when possible, by bars.

14. Mean harvest production rates of screens at
Mayaguana with time (harvest every seven days) .

15. Harvest production rates of screens with depth
(harvest every seven days; means beginning with
forth week for one meter screens; means beginning
with fifth week for ocean array) .

16. Relative abundance of diatoms on test screens at
Grand Turk for the first 41 days of algal turf
development .

17. Relative abundance of major algal groups on test
screens at Grand Turk for 48-168 days.

18. Biomass production of algal turf as a function of
water depth and screen size at Grand Turk. Also,
included is a typical, mid day solar energy spectrum
as a function of depth.

19. Algal turf production with time on white
(translucent) screens and black screens at the same
site on Antigua.

20. Algal turf production on white (translucent) and
black screens as related to depth at Antigua.

21. Algal turf production with time and site on
standard screens at Carriacou. Dashed lines
indicate period of heavy sedimentation on island
apron screens.

22. Number of species of major algal groups on algal
turfs as a function of time at Grand Turk.

FIGURE 23.

Qualitative relationship between turbulence (wave
and current energy) and mean harvest production at
research sites. This is the result of an
independent survey taken of scientists who spent a
considerable amount of time working with algal
screens at at least several of the sites in
question .

TABLE 1.

Algal genera consistently present in various surveys
of Caribbean and western tropical Atlantic coral
reefs. Species listed within genera are common,
persistent components of the coral reef algal turf
assemblage. Genera designated by parentheses
contain no persistent turf components and are
classified as encrusting coralline or macroalgae in
this study. Asterisks designate genera that contain
both the listed turf component species and other
species which do not persist within the assemblage.

Plus signs designate calcareous genera.

TABLE 2.

Common, persistent components of the coral reef

algal turf assemblage in microcosm scrubbers. Species were

identified using Taylor (1960), Humm and Wicks

(1980) and Sears and Brawley (1982).

TABLE 3.

Biomass production rates calculated from 12 day,
concurrent harvests of paired scrubbers. Mean
production rates, +S . D . , range and harvest numbers
listed for both upstream and downstream screens in
each of three flow rate comparisons.

TABLE 4.

Algal turf harvest with time, depth and screen type
at Abraham Bay, Mayaguana, Bahamas Islands.

TABLE 5.

Turf Algae of Mayaguana

TABLE 6.

Algal species identified on turf culture screens at

Grand Turk, Turks and Caicos Islands.

TABLE 7.

Screen type study at site 1 and site 2, Grand Turk.

TABLE 8.

Results of harvest rate study at site 2, Grand Turk.

TABLE 9.

Results of depth array study at site 3, Grand Turk.

TABLE 10.

TABLE 11.

TABLE 12.

TABLE 13 .

. y — i

Algal production (g m d ) in location and screen
type tests in Nonsuch Bay, Antigua.

Algal species tabulated on turf growth rafts at
Grand Bay, Carriacou, Grenada.

Algal turf production (g m”2 d as a function of

location, screen type and time at Grand Bay,
Carriacou .

Algal Harvest Data from Buen Hombre, Dominican
Republic .

Figure 1

■■V

(I/Bui) NOIlVaiNBONOO zO

Figure 2

TIME IN MINUTES

Figure 3

Figure 4

Figure 5

DEPTH IN METERS

Figure 6

.Coral

Head

A00'rn

Figure 7

LOCATION MAP

ANTIGUA AND CARRIACOU ISLANDS

Figure 8

Venezuela

Figure 9

Figure 10

Figure 11

1 -

0 1

ON OFF

1 — i - ~r

2 3 4

ON OFF ON

TIME IN WEEKS

Figure 12

▲
A A

■ . I i y " y— » ' | ™T— T8 T 1 | l 1 ' I l | i

5 10 15 20

NITROGEN CONCENTRATION (NOj + NOi), //M

Figure 13

20 -

Figure 14

wonoa

^Aep^.ui (Ajp)6 1S3AUVH

Figure 15

PERCENT COMPOSITION PERCENT COMPOSITION

70 -I

TOPSIDE

50-

40-

30-

i i i i i - r - — i - n— - 1 - r - 1 —

48 60 72 84 96 108 120 132 144 156 168

DAYS

Figure 16

PERCENT COMPOSITION PERCENT COMPOSITION

100 -i

Figure 17

//Em'

130-

120-

110

100

90-

80-

70-

60-

Ultraviolet B 308 5 nm
Ultraviolet A 355 nm

~ T~

—r~

90 100

DEPTH IN CENTIMETERS

18 B

White Screen

Figure 19

TIME FROM PLACEMENT IN DAYS

Aep/,IU/1M Ajp 6 Noiionaoad

Figure 20

DEPTH IN cm

CARRIACOU calm weather

L O

if)

If)

(Aep/2ui/6] NOIlOnaOdd

Figure 21

DAYS

TOPSIDE

saioads jo aaawriN

Figure 22

DAYS

V-

Low High

QUALITATIVE TURBULENCE SCALE

Figure 23

U CJ OOH

U G
CD 0)
Cl U

tn eg

•H £i

u a

<u <y

<D ‘X3
<D CD
u u

• £1
aj

*H 03
•u U
G CD
03 G

•H 03 U O

Cfl H (U

C OJ
03

<D 14-1
^ 0)
X> 0)
•H 5-4

O U C/3 *H
03 O
WO) <D

^ j= "a a
o d hi u
> in

3-4 4_| U-4 AJ

3 o -H c

C/3 C/3 QJ

C/3 C/3 G
C A-> 03 O
•H G rH a
(DUE
•u C O

C O CD U

ai a u

C/3 E CQ U— 4

(DO G

u ti u u ffl
C C/3 c c/3 ^

O CD
a a> to
6 -e

O 4-1 AJ <D
£5-4 0 5-4

E 3 JO 03
O AJ U

C (D
O a-»
O 03

. fcC

W -H
4-J -n
PQ AJ
< *H

H 5

C CD C
*H C W)
03 (D *H
4J Of W

Table 1

I 1 ^qdoi<

TABLE 2. COMMON, PERSISTENT COMPONENTS OF CORAL REEF ALGAL TURF
ASSEMBLAGE IN MICROCOSM SCRUBBERS. SPECIES WERE
IDENTIFIED USING TAYLOR (1960), HUMM AND WICKS (1980)
AND SEARS AND BRAWLEY (1982).

Bacillariophyta

Licmophora sp.

Navicula sp.

Nitzschia sp. /

Thalassiothrix sp.

Cyanophyta

Anacvstis dimidiata (Kutzing) Drouet and Daily
Calothrix Crustacea Schousboe and Thuret
Entophvsalis sp.

Microcoleus Ivngbvaceus (Kutzing) Crouan

Oscillatoria submenbranacea Ardissone andd Straf forella

Schizothrix sp.

Chlorophyta

Brvopsis hvpnoides Laxnouroux
Cladophora crvstallina (Roth) Kutzing
Cladophora delicatula Montagne
Derbesia vaucheriaef ormisa (Harvey) J. Agardh
Derbesia sp.

Enteromorpha linqulata J. Agardh
Enteromorpha prolif era (Muller) J. Agardh
Smithsoniella earleae Sears and Brawley

Phaeophyta

Ectocarpus rhodocortonoides Borgesen
Gif f ordia rallsiae (Vickers) Taylor
Pvlaiella antillarum (Grunow) De Toni
Spacelaria tribuloides Meneghini

Rhodophyta

Acrochaetium sp.

Asterocvtis ramosa (Thwaites) Gobi
Banqia fuscopurpurea (Dillwyn) Lyngbye
Callithamnion sp.

Centroceras clavulatum (C. Agardh) Montagne
Ceramium corniculatum Montagne
Ceramium f laccidum (Kutzing) Ardissone
Ervthrocladia sudinteqra Rosenvinge

Ervthrotrichia carnea (Dillwyn) J. Agardh
Gymnothamn imn eleqans (Schousboe) J. Agardh
Herposiphonia secunda (C. Agardh) Ambronn
Polvsiphonia havanensis Montagne
Spvridia sp.

TABLE 3. BIOMASS PRODUCTION RATES CALCULATED FROM 12 DAY ,
CONCURRENT HARVESTS OF PAIRED SCRUBBERS. MEAN
PRODUCTION RATES +S . D . , RANGE AND HARVEST NUMBERS
LISTED FOR BOTH UPSTREAM AND DOWNSTREAM SCREENS ON EACH
OF THREE FLOW RATE COMPARAISONS .

FLOW RATE

BIOMASS PRODUCTIONS

(g m~^ d-1)

COMPARISONS

UPSTREAM SCREENS

DOWNSTREAM

SCREENS

b .

5 . 5 1/min

X— 6 . 7 (±1.4 S.D.)

X=2 . 9 (±1.3)

(1 wave/70 sec)

Range-5 .3-8.5

1.4-4 . 3

N— 4 harvests (12 days

N=4

growth per harvest
period)

VS.

c .

d .

166.0 1/min

X=ll . 4 (±3.4) *

X=12 . 6 (±2.2)

(1 wave/ 2 3 sec)

8.8-16.2

9.7-14 . 7

N— 4

N=4

a .

b .

13.5 1/min

X=ll . 9 (±3.4)

X=5 . 5 (±1.3)

1 wave/27 sec

6.4-15.4

3 . 6-6.8

N=5

VS.

c .

32.5 1/min

X— 13 . 7 (±3.3)

X=12 . 4 (±3.9)

1 wave/10 sec

10.9-19.3

6.0-16.8

N=5

a .

b .

13.5 1/min

X=8 . 9 (+2.1)

X=5 . 5 (±2.0)

1 wave/27 sec

5.5-11.8

2 . 9-8 . 3

N=7

VS.

c .

32.5 1/min

X=10 . 6 (±2.4)

X=5 . 6 (±1.0)

1 wave/5 sec,

7 . 3-14 . 1

3 . 9-6 . 6

wave bucket 1/2
normal volume

N=7

* = statistical significance between harvests on the same screen
within the comparison, P < 0.05

TABLE 5. TURF ALGAE OF MAYAGIJANA

Chlorophyta (Green Algae)

Ulotrichales

Ulvaceae

*Enteromorpha chaetomorphoides Borgesen

Cladophorales

Cladophoraceae

*Cladophora laetenvirens

Siphonocladiales

Valonniaceae

Cladophoropsis macromeres Taylor
Cladophoropsis membranacea (C. Agardh) Borgesen

Siphonales

Derbesiaceae

Derbesia marina (Lyngbye) Kjellman

Bryopsidaceae

Brvopsis sp.

* Bryopsis pennata Lamouroux
*Pseudobrvopsis sp.

Phaeophyta (Brown Algae)

Ectocarpales

Ectocarpaceae

*Gif f ordia sp.

Sphacelariales

Sphacelariaceae

*Sphacelaria sp .

*Sphacelaria tribuloides Meneghini

Rhodophyta (Red Algae)

Gelidiales

Gelidiaceae

*Gelidium sp.

*Gelidium pusillum (Stackhouse) Le Jolis

Cryptonexniales

Corallineae

*Amphiraa sp.

*Amphirea f ragilissima (Linnaeus) Lamouroux
*Jania sp.

Jania adherens Lamouroux
Jania capillacea Harvey
Jania pumila Lamouroux
Jania rubens (Linnaeaus) Lamouroux
Rhodymeniales

Champiaceae

Coelothrix irregularis (Harvey) Borgesen

Ceramiales

Cerramiaceae

Crouania attenuata (Bonnemaison) J. Agardh
*Grallatoria reptans Howe
*Antithamnion sp.

Dohrniella antillarum (Taylor) Feldman-Mazoyer
Wrangelia argus Montagne
*Wrangelia pencillata C. Argardh
*Callithamnion sp.

Callithamnion haliiae Collins
*Mesothamnion caribaeum Borgesen
*Grif f ithsia globulifera Harvey
Spermothamnion investens (Crouan) Vickers
*Ceramium sp.

*Ceramium f laccidum (Kutzing) Acdissm
Ceramium nitens (C. Agardh) J. Agardh
Ceramium breizonatum H.E. Petersen V. Caraibica
*Centroceras clavulatum (C. Agardh) Montagne

Delesseriaceae

Taenioma macrourum Thuret

Dasyaceae

*Dasva sp.

Dasva rigidula (Kutzing) Ardissone

Heterosiphonia wurdemanni (Bauley ex Harvey) Falkenberg
Halodictvon mirabile Zanardini

Rhodomelaceae

Falkenbergia hillebrandii (Bornet) Falkenberg

(This is the tetrasphorophyte stage of Asparagopsis
arroata )

Polvsiphonia sp.

Polvsiphonia atlantica Kapraun and Norris
Polysiphonia binnevi Harvey

Polvsiphonia denudatum (Dillwyn) Kutzing
Polvsiphonia exilis Harvey
Polvsiphonia ferulaceae Shur
Polvsiphonia f laccidisma Hollenberg
Polvsiphonia sphacerocarpa Borgesen
Polvsiphonia simplex (Wulfen) C. Agardh

Herposiphonia sp.

*Herposiphonia pectin-veneris (Harvey) Falkenberg
*Herposiphonia secunda (C. Agardh) Ambronn
Herposiphonia tenella (C. Agardh) Ambronn
*Lophosiphonia cristata Falkenberg
*Laurencia sp.

Laurencia microcladia ( Kurtz ing)

Cyanophyta (Blue-Green Algae)

Nostocacceae

*Calothrix

*Nostoc
*Anabaena
Oscillator iaceae

*Oscillatoria

*Schizothrix

Chrysophyta

Bacillariophyceae (diatoms)

(several genera and species occuring particularly on young
screens - not tabulated)

* Occurs on algal turf rafts

TABLE 6. ALGAL TURF SPECIES IDENTIFIED ON TURF CULTURE SCREENS
AT GRAND TURK

CHRYSOPHYTA
Navicula sp.

Nitzschia sp.

Svnedra sp. A
Svnedra sp. B
Tabellaria sp.

Diatom M (unidentified)
Licmophora

mat forming & epiphytic
mat forming & epiphytic
mat forming & epiphytic
mat forming & epiphytic
mat forming & epiphytic
mat forming & epiphytic
mat forming & epiphytic

CYANOPHYTA

Anacvstis sp .

Schizothrix calcicola
(C. Agardh) Gomont
Schizothrix mexicana Gomont

mat forming

epiphytic

mat

CHLOROPHYTA

Cladophora sp . canopy

coccoid green mat

RHODOPHYTA

Asterocvtis ramosa (Thawaites) Gobi
Centroceras clavulatum (C. Agardh)
Montagne

Ceramium sp .

Ceramium bvssoideum Harvey
Ceramium comptum Borgesen
Ceramium cruciatum Collins & Harvey
Ceramium f astoiatum (Roth) Harvey
Champia sp .

Dasvopsis antillarum Howe
Chondria collinsiana Howe
Dohrniella antillarum (Taylor)
Feldman-Mazoyer

Griff ithsia Schousboei Mantage
Herposiphonia sp.

Laurencia sp .

Laurencia obtusa (Hudson) Lamouroux
Laurencia Poitei (Lamouroux) Howe
Polvsiphonia sp . A (unidentified)
Polysiphonia sphaerocarpa Borgesen
Wrangelia sp .

Wranqelia araus Montagne
Wrangelia penicillata C. Agardh

mat forming

canopy

mat forming

canopy

& epiphytic

TABLE 7

SCREEN TYPE STUDY AT SITE 1 AND SITE 2

PORE SIZE (urn)

MEAN la m— d— X

STD ERROR

MONOFILAMENT WOVEN SCREENS AT SITE 1

200

500

710

710x

lOOOu

multiweave

8.8

298

415

644

1
1
0

0.783

0.582

0.820

MULTILAYER EXTRUDED BLACK SCREENS 2 X 3 mm SITE 3

single layer
double layer
triple layer

10.1b
14 . 8a
17.8a

0.872

0.996

1.812

a,b values with same subscript are not significant at 5% level
using Scheffe procedure

1 coarser screen (monofilament wider diameter)

TABLE 8. RESULTS OF HARVEST RATE STUDY AT SITE 2, GRAND TURK.

HARVEST RATE fdavs) MEAN

4 daya 5 . 3

7 daya'b 7.4

12 day la'b 7.8

12 day 2a'b 7.2

12 day 3a'b 7.9

20 dayb 9.4

(q m— d— X STD. ERROR

0.449

0.672

0.762

0.614

0.922

1.200

a , b means with same superscript are not significantly different
at 5% level using Scheffe procedure

TABLE 9

RESULTS OF DEPTH ARRAY STUDY AT SITE 3

DATA IN g m 2 d 1

DEPTH TOPSIDE UNDERSIDE ENTIRE NUMBER

MEAN

RANGE

MEAN

RANGE

MEAN

RANGE

HARVESTS

sur

6.1

4. 1-9.1

3.5

2.9-4. 6

9.7

7.1-12.

7.9

5.2-16.5

4.2

3. 0-5. 8

12 . 1

8 . 2-21.

7.9

3 . 0-15.9

4 . 1

3 . 2-4 . 9

12 . 0

6.8-20.

11.2

5.6-22 . 0

4.5

3. 2-7. 5

15.7

9.5-29.

11.7

3.7-23.7

4.2

2. 7-5. 5

15.9

8 . 8-28 .

100

11.4

6.4-19.2

4 . 6

3.4-6. 1

16.0

14 . 2-24

. 0

ro o

M O

m ro

ro n

cn ro

cn >

cn m

M <

ro w

ro h3

ro ro

O M

> w

> o

o ro

z z

i-a

tv)

to I—1

t— * i— •

i— *

vo un

* «

•

UJ UJ

H- •

t to

H— 4

t-j

•

H-4

u»

tv)

►— 4

tv)

t— 1 *

O '

-o un

•

-4

to U>

tv)

I—4

4a.

•u

un as

4s.

•

f— 4

u»

to un

•

H-*

tv)

H=» K-*

-o

ov -o

•

U»

4S.

JS*

to VO

tv)

to to

f— 4

h-4

tv)

to ov

•

•o

-j

fv)

u> as

t-j

t t—

I—4

-o

H-4

•

t— •

5— 1

l— ' '

U» U)

•

@ *

•

4a.

u»

V©

-O vO

h«*

1 to

-j

•

tv)

i to

•o

-J

•

h-*

! to

■o

I—1

•

H*4

1ABLE 10, Algal production in location and screen type tests in Nonsuch Bay.

TABLE 11. ALGAL SPECIES TABULATED ON TURF GROWTH RAFTS AT GRAND
BAY, CARRIACOU, GRENADA.

Chlorophyta

Cladophora sp .

Cladophora fascicularis
Cladophoropsis sp.

Halicvstus osterhoutii
Brvopsis plumosa

Phaeophyta

Dictvota dentata

Dictvota ciliolata v. bermudensis
Dictvota dichotoma
Gif fordia mitchellae
Padina vickersiae

Rhodophyta

Ceramium byssoideum
Ceramium subtile
Callithamnion sp .

Griff ithsia alobulifera
Laurencia sp .

Acanthophora spicifera
Polvsiphonia subtilissima
Polvsiphonia denudata
Lophosiphonia sp ,

Wranalia penicillata

Chrysophyta (bacillariophyceae)
Asterionella sp .

Cosinodiscus sp.

Isthemia sp .

Nitzschia sp . A
Nitzschia sp . B
Licmophora sp .

Striatella sp .

Thalassionema sp .

Grammatophora sp .

Cyanophyta

Anabeana sp .

Anacvstis sp .

LYUSfeYa sp^.

Nostoc sp .

Oscillatoria sp .

Schizothrix mexicana

TABLE 12. Algal turf production as a function of location,
screen type and time at Grand Bay, Carriacou, Grenada.

DAYS (g m“2 d”1)

STATION

ST1

1.63

1.54

7.55

17.09

10.27

14.63

12.86

17.70

ST2

1.74

5.54

14.69

23.51

10.58

16.17

13 .31

18.04

REEF

APRON

1BF

1.93

18.90

18.00

19.63

12.21

15.53

17.45

1BH

—

.47

6.87

17.09

25.15

13.95

15.02

17.08

5 m
depth

4.99

11.55

13.24

9.32

1.7

2.4

12 . 0

17.8

16.4

13 . 3

14.2

17.6

OUTER

STS

—

0.24

6.63

9.59

9.21

19.77

8 . 17

13.98

ISLAND APRON

8 m

ST4

0.40

4.08

11.21

10.22

20.56

11.02

15.52

depth

0.3

5.4

10.4

9.7

20.2

9.6

14.7

STS

—

2 . 11

15.42

14.50

7.32

13.77

14.45

18.86

PATCH

ST6

1.86

7.35

18.94

7.73

13.72

17.42

21.84

5 m

depth

3BF

2.59

11.61

16.23

14.49

16.39

15.41

17.97

3BH

3 .30

5.16

12.83

11.81

14 . 38

11.88

14 . 52

—

4.07

3.73

9.47

—

10.39

—

2.5

9.9

15.6

10.3

15.3

14.8

18.3

ISLAND

ST7

0.51

6.76

6.46

5.94

13 . 02

9.74

11.43

APRON

ST8

0.41

2.7

6.82

5.06

13 .20

5.89

8.73

5 m
depth

0.5

4.7

6.6

5.5

13 . 1

7.8

10.1

INNER

ISLAND

APRON

ST9

1.12

7.21

16.44

9.24

6.07

8.40

10.79

4 m

ST10

1.30

7.75

15.52

7.54

7 .38

6.33

8.52

depth

1.2

7.5

16.0

8.4

6.7

7.4

9.7

COMBINED

STATIONS

2 . 5

10.9

16.7

13.4

14 . 3

14 . 5

18 . 0

1 AND 3

TABLE 13, Buen Hombre, Dominican Republic Algal Harvest Data.
Each value represents two pooled screens of area 0.56 m2 each.
Growth time equals seven days.

TOTAL

DRY HARVESTS

(all 2

screens

pooled)

„ _-2
g m

d“

1 SV

6/10

6.7

6/17

8.2

6/24

10.3

7/1

6 . 6

7/8

8.5

7/15

8.6

7/22

10.0

105

7/29

8.2

8/5

9.4

8/12

9.0

8/19

9.6

8/26

8.8

9/2

12.3

101

112

9/9

10.2

72.7

71,4

67.8

69.9

73 . 6

Mean harvest

70.98

for 7

9.3

days ; X =

9.1 8.6

SUMMATION OF
‘ 9.1 g m 2 d

8.9 9.4

ALL HARVESTS

1, S.D. = ±2.0,

Mean daily
production,

N = 68

•

SECTION II: BIOLOGY AND ECOLOGY OF MITHRAX SPINOSISSIMUS

MORPHOLOGY AND RELATIVE GROWTH OF MITHRAX
SPINOSISSIMUS (DECAPODA: BRACHYURA: MAJIDAE)

MORPHOLOGY AND RELATIVE GROWTH OF MITHRAX (MITHRAX)
SPINOSISSIMUS (DR CAPO BA: BRACHYURA; MAJIDAE)

A.H. Biddlecomb, M.P. Craig and J.M. Iglehart

Abstract

In this study of MX. spinosissimus collected on the reefs of
Antigua, twelve instars from first crab stage (1.5 mm CL)

to instar 15 (83,0 mm CL) are identified from culture. All 15
stages were obtained by rearing eggs taken from berried females
caught from the reef environment. Measurements of ventral
propodus length, carapace length and carapace width were used to
define the growth phases of Mj, spinosissimus . The average growth
per molt for crabs less than 80 mm CL is 35.7%, and 23.7% for
those greater than 80 mm CL.

In this work, the point at which the molt of puberty occurs
(+105 mm CL) is decidedly different from the results of a study
in the Florida Keys (±80 mm CL) (Bohnsack, 1976) . It is
suggested that there are several morphologically distinct post-
pubertal male instars and the molt of puberty is not the terminal
molt.

INTRODUCTION

Mithrax (Mithrax) spinosissimus (Lamarck, 1818) is placed in
the Family Majidae (infraorder Brachyura) , commonly known as the
spider crabs. Milne-Edwards (1832) amended the taxonomic
classification to the present genus and specices, from the
previously classified Mala spinosissima (Lamarck, 1818) . The
genus Mithrax designates crabs as being characterized by long
slender walking legs, a nearly oval shaped carapace with the
front formed of two small often pointed rostral horns. The
orbital margins are generally more or less spinous or
tuberculate, and the merus of the external maxillipeds are broad.

The abdomen of the male is formed of seven free segments. M.
spinosissimus is sometimes referred to as the West Indian Red
Spider Crab or, because of its large size, the Caribbean King
Crab. In this investigation, we have succeeded in obtaining a
number of successive developmental stages of this species. These
collected specimens have served as the basis for describing the
life history and growth phases of M^_ spinas issimus .

Morphology of Adults

Mithrax spinosissimus is distinct from other tropical Majids
because of its large size. It is the largest species in the
genus; the largest adult previously reported had a carapace
length (CL) of 167 mm (Rathbun, 1925) and larger animals are
reported in this study. Considerable size variation does,
however, occur throughout the geographic range (see Iglehart, et
al . , this volume) . M^_ spinosissimus is distinguished from other

Caribbean/tropical Atlantic species by the presence of eight to
nine spines on the dorsal edge of the manus, the presence of two
spines on the basal part of the antennae and its nearly naked
carapace. The hepatic and cardiac regions of the carapace are
distinctly delimited.

The carapace of an adult Mithrax spinosissimus is large,
naked and approximately equal in width and length. The entire
carapace is rough with short spines. In the center, the spines

are blunt , while elsewhere they are sharp. With a deep cervical
suture , the rostral horns are narrow, truncate at the tips and
are separated by a U-shaped sinus of equal length and breadth.
Mithrax epinosiss irons is strongly sexually dimorphic. The
chelipeds of mature males are larger and longer than their
walking legs. A large tooth-like structure emanates from the
male dactyl (Figure 1) . Spines on the chelae of old males tend
to be blunt and tuberculiform, while those on the merus tend to
retain their spiny character. Chelipeds of both sexes are armed
with numerous stout spines. Chelipeds of the mature female are
no longer and not much stouter than their first walking legs.

The manus tapers distally and the fingers are narrowly gaped with
numerous denticles on the inner edges. The tips of the chelae
are hollowed and spoonlike with serrated edges in both sexes,
thus allowing the animals to feed on the abundant algal turfs and
small macroalgae present in reef environments.

The male abdomen is triangular through abdominal segments 1,
2 and 3 , and nearly rectangular through segments 4 , 5 , 6 and 7
and hangs loosely in mature males. The abdomen of the male
covers only a small portion of the width of the sternum, while
the abdomen of the mature, female is broad and nearly covers the
entire sternum.

Morphology of Juveniles

Rathbun (1925) described medium sized juveniles as having a
carapace much longer than wide and covered with setae. At this
age, the spines are sharper than the spines of adult crabs,
including the tips of the rostral horns which also curve slightly
inward. Both sexes, as juveniles, have small chelipeds with a
gape extending one-half the length of the fingers.

In younger individuals, spines appear sharper than those of
older juveniles and longer in relation to carapace size. The
rostral horns are also longer, one-fifth as long as the carapace.
There are two spines on the suborbital margin outside the
antennal segment. The chelipeds of the young are no longer or
stouter than the first pair of legs.

Morphology of Larvae

Provenzano and Brownell (1977) described the larval and
first crab stages. Wilson, et al . (1979) discussed morphological

differences and similarities in larvae of M. forceps, M.
spinosissimus and M. pleuracanthus (Yang, 1967) . The latter
species was fully described by Goy, et al. , (1981), in which the

authors proposed a reassessment of the genus because of
similarities between M. forceps and M. pleuracanthus (of separate
subgenera) , both of which differ from M^ spinosissimus . Goy, et
al. (1981) states that larval M. spinosissimus are "clearly less
related" to other Mithracinae. Other developmental descriptions

of larval stages for species in the genus are for M. corvphe
(Scotto and Gore, 1981) , M. hispidus (Fransozo and Hebling, 1982)
and M. verrucosus (Bolanos and Scelzo, 1981) .

Relative Growth of the Chelae

The process of reaching maturity, the size at maturity and
the presence or absence of a terminal molt are significant
elements in considering the mariculture potential of M.
spinosissimus . Physical and sexual maturation bring about
changes in the relative growth between various body parts. An
aspect of the relative growth of M. spinosissimus with biological
and economic significance is the large size of the chelipeds of
mature male instars. Useable meat to total weight ratios are
considerably higher for mature males with the largest chelae. At
the molt of puberty, chelae of the male M. spinosissimus become
substantially larger, while in the females the abdomen increases
in width so as to cover the sternum, thereby making a protected
brooding space. In addition, gonads in both sexes begin to
mature. The female pleopods, gonopores and abdominal locking
mechanism also noticably change at the molt of puberty (see
Craig, et al. , this volume) . By relating the "claw length" (VPL)
to the carapace width of the crabs collected in Florida, Bohnsack
(1976) suggested that male M. spinosissimus molted to maturity at
approximately 80 mm CW, because at this point the chelae length
increases at a proportionately higher rate than carapace length.

Tessier (1935), Vernet-Cornubert (1958) and Hartnoll (1963,
1965) compared ventral propodus length (VPL) and carapace length
(CL) or width (CW) on a log-log plot to determine specific
changes in allometry, especially at maturation. These three
authors all concluded that in the Majidae, ecdysis and therefore
growth during molt ceases after the molt to puberty.

Hartnoll (1965) investigated the life history of five
tropical spider crab species. He concluded that the molt to
maturity was the terminal molt based on: 1) evidence of pre¬
pubertal instars molting in captivity, i.e., observations of
several dissected pre-pubertal instars preparing to molt and the
lack of molting or indications of molting among the post-pubertal
instars; 2) pre-pubertal instars developed limb buds upon losing
an appendage, while post-pubertal instars with autotomized
appendages formed calcified stumps and were not observed to form
limb buds; and 3) the "epifauna of the integument," which attach
as larvae (i.e., barnacles and serpulid worms), was observed to
be more abundant and older on post-pubertal instars, indicating
at least, a much longer intermolt period.

Tessier (1935) described a critical molt in male instars of
the species Maia scruinado prior to the molt of puberty. The
molt, called the molt of pre-puberty, marked the first
differentiation of relative growth rates between sexes. In Pisa
tetraodon . Vernet-Cornubert (1958) was able to show that the pre-

pubertal instar has a great range of carapace lengths, and the
molt of puberty may be from one to three instars after the first
pre-pubertal instar. Tessier (1935) showed the molt of pre¬
puberty to occur at approximately 70 mm CL in Maia squinado and
the molt of puberty to occur three instars later. The first
slight proportional increase in growth of the chelae over the
carapace was considered by Tessier (1935) , Vernet-Cornubert
(1958) and Hartnoll (1963) to be the beginning of the range of
carapace lengths where the molt of pre-puberty occurs in the
species they studied. Hartnoll (1963) gives extensive
consideration and discussion to a pre-pubertal molt in the "Manx"
spider crabs, but the Majids he studied in Jamaica are not
mentionned (Hartnoll, 1965) .

Hartnoll (1965) found a wide size range of post-pubertal
Majids in both sexes. From samples of Mithrax sculptus .
differences in post-pubertal CL were calculated showing that the
largest mature female was 162% larger than the smallest mature
female, and the largest mature male was 314% larger than the
smallest mature male (Hartnoll, 1965).

In most Majids there is only a single type of post-pubertal
male instar. However, Hartnoll (1963) and Vernet-Cornubert
(1958) , found that among Pisa tetraodon . Pisa gibbsi and Inachus
leptochirus . there are two morphologically distinct types of
post-pubertal males which vary in chela length (the latter two)
and in chelae breadth (all three) , so that the two distinct types

in each species have different levels of allometry, i . e. ,
separate phases of maturity.

Hartnoll (1974, 1978) described the growth of organs
functioning as primary or secondary sexual characteristics
relative to carapace growth by the allometric growth equation,
y=Bxa where y = variable dimension (organ size) ; x = reference
dimension (carapace length or width) ; a = the regression
coefficient (or rate of growth) ; and B = y intercept (or
proportional difference between organ and body) . The allometric
growth equation expressed logarithmically is: log y = log B + a
log x. The relative growth of Brachyuran chela, abdomen and
first pleopods may then be compared by using the regression
coefficient, "a", or the level of allometry (Hartnoll, 1974).
These changes in levels of allometry at specific instars indicate
important changes in the function of particular body parts
relative to sexual maturation.

In a comparison of the positive allometry levels of the
adult male chela and the female abdomen, Hartnoll (1974) also
found a positive allometry in the pre-pubertal male phase and
considerable size increases at the pubertal molt. At the molt of
puberty, the relative growth of the male chela increases to a
higher level of positive allometry as does the relative growth of
the abdomen of the female whose terminal molt coincides with the
molt of maturity. In those females that continue to molt after

the molt of puberty , the abdomen width to the carapace length
decreases isometrically .

METHODS

From 1985-1986 , 157 M*, spinosissimus were collected with
standard West Indian lobster traps off the eastern coast of
Antigua in the Caribbean/tropical Atlantic (Lat. 17° 10' N , Long.
61° 43' W) . In addition, 107 measured crabs were reared from
larvae at Nonsuch Bay in eastern Antigua (see Porter, et al . .
this volume) . The cultured crabs were reared from broods
generated by captured animals from wild populations. Nearly all
crabs less than 80 mm were cultured. Photographs of post-larval
instars 1-15 are presented from culture.

Ventral propodus length (VPL) was measured from the most
distant tip of the fixed finger to the end adjacent to and
projecting slightly below the carpal hinge (A/B Figure 1) . All
measurements were made to the nearest 0.1 mm using precision dial
calipers. Plots of the measurements of CL vs VPL were used to
identify size at maturity. Measurements were taken from the
right cheliped unless it was missing, in which case, the left was
used. Carapace length (CL) was measured, exclusive of rostral
horns, from the middle of the rostral sinus to the most posterior
edge of the carapace. Carapace length is a more precise

measurement than the width due to its ease of measurement and
standardization of measurement procedure in terms of replication.
Carapace width (CW) , measured from just in front of the fourth
branchial spines, was also taken to compare CW to CL and provide
a reference to previous work. A least squares regression was
used to determine the relationship between CW and CL.

RESULTS

Of 264 crabs measured in Antigua, 107 were cultured and 157
were captured locally. The crabs ranged in size from 19.8 to
146.5 mm CL. The VPL and CL data for individual crabs given in
figure 2 shows that males and females are dimensionally
indistinguishable up to approximately 55 mm CL. Up to that
size, both sexes are showing a slight positive allometry or
proportional increase of the propodus relative to the carapace.

At that point, however, sexual dimorphism begins to be evident
with the males showing proportionally even longer chelae. At 75
to 80 mm CL there is a distinct inflection in the relationship
between ventral propodus length and carapace length for pre¬
pubertal males with even greater chela elongation. Beginning at
about 100 mm CL, the VPL of mature males increases markedly
relative to carapace size. This point marks the instar size at
which the cheliped becomes longer than the carapace in
approximately 85% of the post-larval instars in the 100-120 mm CL
range. Finally, at about 130 mm CL, the allometry of the VPL

decreases slightly for mature males. These allometric
relationships are rigidly adhered to in the crab population we
studied. A single male, 114.8 mm CL, 76.4 mm VPL, had both
chelae equivalent in length to females of the same carapace size.

The relationship between ventral propodus length and
carapace length of females is also positively allometric (+1.17),
except for a number of the largest adults, which tend to show
relatively "shortened" chelae, or a slightly negative allometry,
like the oldest males. A study of the allometric relationships
of the abdomen width would probably show roughly inverse
male/female relationships.

Examination of the percent incease in carapace length
increment at each molt shows a total mean increase of 30.4% (S.D.
+9.4%,* N-44) (Figure 3). The data were separated at 8 0 mm CL
because of the definite differentiation of males from females at
this point as determined from figure 2. The mean CL increment of
instars less than 80 mm CL is 35.7% (S.D. ±6.8 %, N=26) , and
above 80 mm CL, 23.7% (S.D. +5.8%, N~18) .

Eleven, wild, immature crabs molted in captivity in Antigua.
Four of the seven males and each of the four females molted to
maturity. The females averaged a 24.6% increase in CL and an
average of 31.1% increase in VPL. Of the four females that
molted to maturity, the largest pre-pubertal instar was 86.2 mm
CL, 53.3 mm VPL and the smallest post-pubertal instar was 94.4 mm

CL, 64.3 mm VPL. The males that molted to maturity were among
the largest of the immature crabs captured but they molted only
to the lower end of the range of mature males, from means of
101.4 mm CL, 76.6 mm VPL to 119.3 mm CL, 114.6 mm VPL,

(averages). The males averaged a 17.7% increase in CL and a
49.9% increase in VPL. The other three males molted to a size
somewhat larger than the pre-pubertal size of those that became
mature. They experienced an average increase of 26.1% in CL and
39.7% in VPL. These relationships are plotted as means on figure
2.

Of the 264 JL spinosissimus examined, the relative growth of
the chela of pre-pubertal males exhibit a positive allometry that
is higher than that of the pre-pubertal females, especially
between 80 and 105 mm CL (Figure 4) (Table 1) . The male VPL
significantly increases at the molt to maturity as represented by
the break in the distributions, with a subsequently higher level
of positive allometry. The female VPL, however, appears to have
nearly isometric chelar growth in relation to the CL for
specimens greater than 105 mm CL. Photographs of the maturation
of male and female chalae are shown in figure 6.

The relationship between CL and CW (Figure 5) (CW ~ -4.84 +
1.06 CL; P <0.005) depends on the stage of development. At the
early post-larval instars, CL is greater than CW, whereas in
adult M^_ spinosissimus . CW exceeds CL. The transition from the
elongated body shape (CL > CW) of the juvenile poost-larval

instars to the anteriorly-posteriorly flattened body shape (CW >
CL) of the adults occurs at approximately 80.7 mm CL. At
approximately 80.7 mm, CW is equal to CL. Simply stated, the
juvenile carapace is longer than wide. At 80.7 mm CL as the
crabs enter full puberty, it is virtually round. Finally, as a
sexually mature adult, the carapace becomes wider than long.

Developmental Morphology

All the developmental stages observed were obtained by
rearing larvae from eggs obtained from gravid females caught in
the reef environment. Fifteen successive instars, from first
instar to the last pre-pubertal instar are depicted in figures 7-
25. The terminal part of the first post-larval instar is similar
to the megalops stage as the telson is loosely tucked to the
abdomen. From the second post-larval instar to the eighth post-
larval instar (Figures 8 to 13) , additional spines become
apparent and existing spines increase in length. Also, the
orbital region initially increases in size in relation to the
body, and then proportionally decreases in size through those
instars as the portion of the carapace shielding the eyes
diminishes in size. At instar 4, the second rostral horns
develop. The carapace becomes more oval shaped and wider
posteriorly through instars 5-7. In addition, the carapace shows
increased spine development, with the third branchial spine

bifurcating. The orbital region shortens and the rostral horns
turn in slightly at the tips through instar 8. From instar 9
(Figure 14) , which generally occurs from 14 to 24 mm CL, the
basic form and shape of the adult carapace has been assumed, and
the species is easily distinguished. Prior to instar 9 or 10. M.
spinosissimus is similar in appearance to the adult Mithrax
acuticornis and juveniles of other Mithrax species such as
Mithrax verrucosus . Sex is readily determined at instar 10,
although the abdomen does not achieve its mature shape until the
pubertal molt. The third branchial spine is longer while the
rostral horns are noticeably thicker and shorter in proportion to
the carapace length. At instar 11, there is a shortening of the
orbital region and the L/W ratio approximates 1.08 which remains
constant or decreases slightlty through the rest of the instars.
Instar 13 (Figure 20) lacks hooked setae on the carapace. On the
dorsal region of the carapace, these setae are absent but are
present along the carapace edges. The propodus and dactyl
segments of the walking legs have abundant setation even on
mature crabs. The chelipeds and abdominal regions exhibit marked
allometric growth from the fourteenth instar. At instars 14-16
the spines on the carapace become shorter and more rounded.

There are fewer setae on the carapace.

DISCUSSION

Accompanying growth in Mithrax spinosissimus are changes in
carapace and limb proportions. When very young, the juveniles
are slightly elongated and quite spiny. Prior to about instar 7
they are decorators and very difficult to see in an algal
environment. As they approach pre-puberty, they become slightly
wider than long and the sharp setae gradually become blunter.

The mean increment of molt size increase of approximately 30.4%
for carapace length is of little value in the morphometics of
growth. This increment although highly variable is clearly
larger (approximately 35%) in young juveniles and decreases and
narrows in range in older juveniles. In pre-pubertal crabs, the
molt interval drops to about 25% and finally, as the terminal
molt approaches, extends below 20%. The development of the
chelae in the males is slightly allometric as young juveniles
becomes strongly allometric and proportionally larger at the
molts of pre-puberty. Hartnoll (1965) determined that the sharp
alteration in relative size of the chelae in male spider crabs is
the primary and most reliable index of the the molt of puberty.
Thus, in It spinosissimus the relationship between ventral
propodus length and carapace length indicated by the sharp
inflection and break in the line at approximately 105 mm CL
indicates the point at which the male molt of puberty occurs.
Figure 6 shows the chelae of a male and female through a series

of molts. Though the relationship between ventral propodus
length and carapace length becomes sharply allometric in males at
about 100 mm CL, and becomes isometric in females at about 80-90
mm CL, as both apparently molt to even larger sizes, growth
becomes negatively allometric and cheliped size shortens relative
to the body length. Although the primary growth and
morphological relationships seem quite clear through the young
adults in the Antiguan population, the post-puberty patterns and
the relationship of the Antiguan crabs to those elsewhere in the
Caribbean leave some questions.

The relationship between VPL and CL from M_;_ spinosissimus
collected in Antigua indicates that the morphological
relationships shown in figure 4 are significantly different from
those previously sampled in Florida (Bohnsack, 1976) . There
appears to be some Caribbean-wide regional variation in
morphological characteristics as mature crabs caught at other
sites were larger than those in Antigua, which in turn were
considerably larger than the Florida population (see Bohnsack,
1976) .

Very large pre-pubertal and post-pubertal male instars were
collected in the Dominican Republic, measuring 131.0 and 180.0 mm
CL respectively. The largest pre-pubertal and post-pubertal
female instars were collected in Grand Turk, measuring 105.6 and
158.2 mm CL respectively. This coincidental trend suggests that
the large pre-pubertal crabs molt to the largest adults. The

large immature male (131.0 mm CL) had morphologically immature
chelae, and after molting, it measured 153.3 mm CL (no VPL was
recorded for this crab) .

The molt increment data in figure 4 indicates a sharply
decreasing trend in molt increments for crabs molting at a CL
longer than 80 mm. A general decrease in increment might be
attributed to conditions of culture since much of this data is
derived from crabs cultured from eggs. However, the low molt
increments are consistent with that found from wild crabs only
observed in culture for short periods. Also, this relationship,
as expressed by the "Hiatt Growth Diagram," is well known for
other crabs (Mauchline, 1976) .

The observed molt increments suggest that it is improbable
for an Antiguan crab to molt from the largest CL in the size
range of pre-pubertal crabs to the largest in the size range of
post-pubertal crabs. This would be a carapace length increase of
about 34% for males, whereas we predict an increase in CL
increment at each molt of only about 15-20% based on figure 3 and
the molts achieved after capture. Comparable figures are
available for Hvas coarctatus (brachyura) with an average
increase in CL among laboratory specimens (n=8) of 21.5 (Hartnoll
1963a) . With regard to carapace/propodus relationships, female
maturation is not so clear. Based on those animals that did molt
to maturity in captivity, a 25% increment is found, and this

essentially agrees with figure 3. On the other hand, the
required female CL molt increment would have to be 37.4% between
the largest pre-pubertal instar (92.4 mm CL, 59.0 mm VPL) and the
largest post-pubertal instar (127.0 mm CL, 81.1 mm VPL).

It is suggested that the terminal molt may not occur at the
molt to maturity in female Mithrax spinosissimus as evidenced by
the wide variation in range of carapace length for mature females
and the consistent variation in molt increments for crabs of the
same carapace length. On the other hand, the wide variation in
carapace length observed in cultured crabs at the same instar and
the inability to be certain whether some females have indeed
achieved the full reproductive state make it very difficult to
establish the state of some crabs. Also, great differences in
size increase between crabs that molted to maturity and the
increments predicted based on the distribution, may occur due to
environmental variables, such as temperature, light, amount and
type of food available, and competition for food and space. The
difference may also be due to genetic variation, in which case
the cited crabs that molted to maturity in captivity are an
insufficient sample.

Until recently the ma j id crab Chinoecetes opilio (the snow
crab) was thought not to have a terminal molt at the molt to
maturity (Davidson, et al . , 1985) . While it has now been
demonstrated otherwise, it is to be noted that in that species
there is considerable overlap in male carapace size between the

pre-pubertal animals and the adults (Watson, 1980) . This differs
significantly from the data found for It spinosissimus in this
study (Figure 2). In addition, from Figure 2, there is an
apparent reduction in positive allometry for those males over 130
mm VPL. Males with a VPL greater than 130 mm are offset from and
slope slightly less than the smaller mature males. Thus, in
this species there remains considerable doubt that the terminal
molt always occurs at the molt to maturity, at least as that
maturity is evidenced by the size of the propodus.

Juvenile instars of 80 mm CL exhibit positive chelar
allometry. Allometry levels calculated from logarithmically
transformed values for ventral propodus length and carapace
length of instars show the relative growth of Mithrax
spinosissimus to be similar to other Brachyurans (Hartnoll,

1974) . The allometry level of VPL for the pre-pubertal male of
this species and for pre-pubertal male Brachyurans in general is
respectively, 1.27 and 1.26. Post-pubertal Mithrax spinosissimus
males have a higher level of allometry than post-pubertal
Brachyurans (1.73 to 1.53, respectively). The VPL allometry
level for pre-pubertal females is also higher than that for their
Brachyuran counterparts, 1.17 and 1.11, respectively. However,
the VPL allometry level for post-pubertal females decreases
considerably from that calculated for other mature Brachyuran
females, 0.97 compared to 1.10.

Hartnoll (1965) states that mature ma j ids do not molt based

on evidence of epizoeal growth. However, the post-pubertal M.
spinosissimus were relatively clean of epizoal growth. Because
of this, observations of the growth of epizoal organisms to gauge
relative age of the exoskeleton may not be easily applicable for
these crabs. When collected, the mature Antiguan crabs had very
little epizoeal growth on the carapace. During captivity,
however, all crabs developed epizoal organisms and some
eventually attained very well developed red crustose algae on the
carapace. In addition, gooseneck barnacles developed around the
base of the chelipeds.

CONCLUSIONS

The morphometric characteristics of mithrax spinosissimus
change considerably from first crab, through juvenile and pre¬
pubertal instars to adult. The shape and growth, relative to the
carapace, of the chelipeds provides a standard morphological
reference point for Mithrax spinosissimus . Maturity is reached
in 16 or 17 molts. The size range of carapace length among
individual populations from different geographic locations
throughout the Caribbean/West Indian region varies widely,
although it is rather narrow within populations. Different mean
carapace sizes at each study site may be the result of specific
ecological effects perhaps including available burrow size,
available food, predation pressure and temperature in the case of

Florida. However, considering the extremely short swimming
period of the larval stages and the great depths of water present
between many Caribbean Islands, genetic differences between
isolated populations must be suspected.

From our observations, we cannot establish with certainty
that the molt of puberty is the terminal molt. It appears that
the expected molt increment of 20-25% is too small to allow for
all pre-pubertal instars to molt to the largest post-pubertal
instars. In addition, at Antigua the lack of attached epifaunal
growth on the carapace of all post-pubertal instars coupled with
the observation of male crabs with slightly enlarged chelipeds
molting lead us to suggest that another instar can occur after
the pubertal molt, that being the terminal molt.

Except for the Florida populations, which in mariculture
might molt to non-commercial sizes, Mithrax spinosissimus appears
to offer maximum potential for marketing in that a choice could
be made between relatively small, low meat to weight ratio
females, the large, high meat to weight ratio males and an even
larger size male post-larval instar. Unfortunately, we have no
information at this time on the time interval to a second adult
molt. It may or may not be consistent with the later molts of
pre-puberty and the molt to adult.

LITERATURE CITED

Bohnsack, J.L. 1976. The spider crab, Mithrax spinosissimus : an
investigation including commercial aspects. Florida
Scientist - Florida Academy of the Sciences 39(4): 259-266.

Bolanos, J. and M. A. Scelzo. 1981. Larval development of the spider
crab Mithrax verrucosas Milne-Edwards , reared in the
laboratory (Decapoda: Brachyura: Majidae) . Am. Zool. 21 (4) :
989, abstract 436.

Brownell, W.N., A.J. Provenzano, Jr. and M. Martinez 1977. Culture
of the West Indian Spider Crab Mithrax spinosissimus at
Los Roques, Venezuela. Jour. World Mar. Soc. 8: 157-167.

Coen, L.D. 1987. Dissertation: Plant animal interactions:

Ecology and comparative functional morphology of plant
grazing decapod (Brachyuran) crustaceans. 241 pp.

Davidson, K. , J. Roff and R. Elmer. 1985. Morphological,

electrophoretic, and fecundity characteristics of Atlantic
Snow Crab, Chionoecetes opilio , and implications for
fisheries management. Canadian J. Fish. Aquat. Sci. 42. :
474-482.

Fransozo, A. and N.J. Hebling. 1982. Larval development of
Mithrax hispidus (Decapoda Majidae) in the laboratory.

Cienc. Cult. (Sao Paulo) 34(3): 385-395.

Goy, J.W. , C.G. Bookhout and J.D. Costlow, Jr. 1981. Larval
development of the spider crab Mithrax pleuracanthus
Stimpson reared in the laboratory (Decapoda: Brachyura:
Majidae). Journal of Crustacean Biology 1(1): 51-62.

Hartnoll, H.G. 1963. The biology of the Manx spider crabs.

Proc. Zool. Soc. London 141(3): 423-496.

Hartnoll, H.G. 1965. The biology of spider crabs: A comparison
of British and Jamaican species. Crustaceana 9: 1-16.

Hartnoll, H.G. 1974. Variation in growth pattern between some
secondary sexual characters in crabs (Decapoda Brachyura) .
Crustaceana 27(2): 131-136.

Hartnoll, H.G. 1978. The determination of relative growth in
Crustacea. Crustaceana 34(3): 281-293.

Mauchline, J. 1976. The Hiatt growth diagram for Crustacea.
Marine Biology 35: 79-84.

Milne-Edwards . 1832. Magasin de zoologie 2(2).

Munro, J.L. 1976. The biology, ecology, exploitation and

management of Caribbean Reef Fishes. Part V. The biology,
ecology and bionics of Caribbean Reef fishes: Crustaceans
(lobster and crabs) Res. Rept. Zool. Dept. Univ. West
Indies 3(6): 39-48.

Provenzano, A.J., Jr. and W.N. Brownell. 1977. Larval and early
post-larval stages of the West Indian spider crab, Mithrax
spinosissimus (Lamarck) (Decapoda: Majidae). Proceedings of
the Biological Society of Washington 90(3): 735-752.

Rathbun, M.J. 1925. The spider crabs of America. Bulletin of the
United States National Museum 129: 1-613.

Scotto, L.E. and R.H. Gore. 1980. Larval development under

laboratory conditions of the tropical spider crab Mithrax
(Mithraculus) corvphe (Herbst, 1801) (Brachyura: Majidae).
Proceedings of the Biological Society of Washington 93(3):
551-562.

Tessier, G. 1935. Croissance des variants sexualles chez Maia
scruinado. Trav. Sta. Biol. Roscoff 13: 93-130.

Vernet-Cornubert , G. 1958. Biologie general de Pisa tetraodon
(Pennant). Bull. Inst. Oceanogr. Monaco 1113: 1-52.

Warner, G.F. 1977. The biology of crabs. Van Nostrand Reinhold
Co. N.Y. , N.Y. 10001, 90-92 pp.

Watson, J. 1970. Maturity, mating, and egg laying in the spider
crab, Chinoecetes opilio . J. Fish. Res. Bd. , Canada 27:
1603-1616.

Wilson, K.A. , L.E. Scotto and R.H. Gore. 1979. Studies on Decapod
Crustacea from the Indian River region of Florida XIII.

Larval development under laboratory conditions of the
spider crab, Mithrax forceps (A. Milne-Edwards, 1875)
(Brachyura: Majidae). Proceedings of the Biological Society
of Washington 92(2): 307-327.

Yang, W.T. 1967. A study of zoeal, megalopal and early crab
stages of some oxyrhynchous crabs (Crustacea: Decapoda) -
Ph.D. dissertation. University of Miami. Coral Gables, FLA.
1-459 pp.

LEGENDS

Figure X .
Figure 2 .

Figure 3 .

Figure 4.

Figure 5.
Figure 6.
Figure 7 .

Figure 8.

Figure 9 .

Figure 10.

Figure 11.

Figure 12 .
Figure 13 .

Position of male chelar VPL measurement (A to B.)

VPL vs CL in mm for wild and cultured Mithrax
spinosissimus from Antigua. Solid lines represent
average increment increase of A) four males that
molted to maturity , B) three males that molted to
pre-pubertal instar and C) four females that molted
to maturity.

Relationship of increase in size of Mithrax
spinosissimus carapace length (CL) as a function of
pre-molt carapace length.

Growth phases for juvenile and adult Mithrax
spinosissimus . CL^carapace length; CW=carapace
width; VPL™ventral propodus length. Log/ log plot.

Relationship of Mithrax spinosissimus carapace width
to length (mm) .

Chelae of immature and mature male and female Mithrax
spinosissimus .

Zooea, megalops and first crab (instar I) of mithrax
spinosissimus hatched in culture. 1) Pre-zooea, 2)
Zooea lf 3) Zooea 2, 4) Megalopa, 5) first crab
(Instar I , CL = 1.5 mm) .

Instar 2, 2,5 mm CL. The extended carapace eye
shields are gone and a spine on the eyestalk is more
prominent.

Instar 4 , 5.0 mm CL. The second rostral horns and
carapace spines are more developed.

Instar 5 , 5.1 mm CL. The carapace is noticeably
spinier.

Instar 6, 7,9 mm CL. The posterior width of the
carapace is more pronounced.

Instar 7, 10.3 mm CL. Greater spine development.

Instar 8, 13.8 mm CL. The orbital region is
shortened and the rostral horns are turned in
slightly at the tips.

Pronounced lengthening of the

Figure

14.

Instar

14.8

CL.

third branchial

spine .

Figure

15.

Instar

10,

21.5

CL.

shorter

and

. the

carapac

Figure

16.

Instar

11,

28.5

CL.

more truncate and

turn

Figure

17.

Instar

11,

ventral

•

Figure

18 .

Instar

12,

dorsal '

view ,

rounder

Figure

19.

Instar

12,

ventral

•

Figure

20.

Instar

13,

48.0

CL.

thicker

and

, shorter in

Figure

21.

Instar

14,

60.5

CL.

of the

carapace

•

Figure

22.

Instar

15,

83.0

CL.

Figure 23.
Figure 24.

proportion to the CL and the spines on top of the
carapace are low and rounded.

Instar 16, 101.2 mm CL.

Detail of immature Mi thrax spinosissimus showing limb
buds that form upon loss of an appendage.

Figure 1

180

160

140

120

100

N - 264
# FEMALE
O MALE

§6*

L&O
o O

20 40 60 80 100

CL Imm)

120

140

Figure 2

0°r

50 H

• •

3SV3U0NI 10 [%) J.N30d3d

Figure 3

PRE-MOLT CL Imml

[uiui] ndA

CL [mm]

Figure 4

180

160

140

120

100

N = 150

»•

20 40

60 80 100 120 140 160

CL (mm)

180

Figure 5

Figure 6

scale bars; 1mm

Figure 7

Figure 8 & 9

Figure 10 & 11

Figure 12 & 13

1 5

Figure 14 &

Figure 16 & 17

Figure 18 & 19

Figure 20 &

Figure 22 &

Figure

SECTION II

BIOLOGY AND ECOLOGY OF MITHRAX SPINOSISSIMUS

POPULATION DENSITY AND STRUCTURE OF M^ SPINOSISSIMUS
AT SIX CARIBBEAN SITES AND THE FLORIDA KEYS.

POPULATION DENSITY AND STRUCTURE OF MITHRAX SPINOSISSIMUS

J.M. Iglehart, R.V. Ruark and K.H. Koltes

Abstract

The population density and structure of the Caribbean King
Crab, Mithrax spinosissimus , was examined for mariculture
purposes at seven sites from the Florida Keys to the eastern
coast of Antigua. A significant size difference was found in the
carapace length when comparing Florida crabs to the Caribbean
populations. Individuals from the Florida Keys were smaller,
with a greener coloration and a relatively thinner carapace.
Temperature differences, fishing pressure and the species short
planktonic stage may contribute to the variation in crab density,
size and sex ratios found between sites and seasons. The
differences observed between sites, particularly Florida versus
all other sites, suggests that present-day populations have been
isolated for a long period of time, and represent separate stocks
of the same species.

INTRODUCTION

In response to the declining Alaskan King Crab fishery,
based primarily on Paralithodes camtschatica . investigations of
crab species suitable to support mariculture projects and
fisheries have increased (Idyll, 1971). Preliminary studies have
included investigations of the West Indian giant spider crab, or
Caribbean King crab, Mithrax spinosissimus (Munro, 1974; Hazlett
and Rittschof, 1975; Bohnsack, 1976, Provenzano and Brownell,
1977; Porter, et aj^. , in review). The crab is found throughout
the tropical Western Atlantic from the Florida Keys and Bahamas
to the West Indies, Venezuela and Nicaragua (Williams, 1984) at

depths of 2-200 meters (Colin, 1978). A reported range of
Mithrax spinosissimus extending as far north as the Carolinas
(Rathbun, 1925) seems unlikely.

Two populations have been described previously. Munro
(1974) conducted a survey of 212 Mithrax spinosissimus caught in
traps off the Jamaican coast. In that study, males outnumbered
females almost 2 : 1 and had a mean carapace width (CW) of 133.4
mm. Females had a mean CW of 122.8 mm.

The Florida population was sampled by Hazlett and Rittschof
(1975) and by Bohnsack (1976). Bohnsack (1976) reported carapace
widths of 96 mm (male) and 86 mm (female) for crabs in the South
Florida Keys, with females outnumbering males 2:1. Hazlett and
Rittschof (1975), sampled 115 crabs from a canal at Little Torch
Key, Florida and found none to exceed 110 mm CW.

Little information exists concerning seasonal trends in the
relative abundance of Mithrax spinosissimus . However, crabs have
been collected every month of the year. Munro (1976) found a
seasonal variation in crab abundance based on a two-year survey
of trapping rates in Jamaica. Catches were zero or negligible in
January and February, 1970, 1971 and 1972. Catch rates reached a
maximum in March or April, June or July and November or December,
1970 and 1971. Catch rates were generally greatest in areas of
intensive fishing pressure.

Several factors appear to affect the local abundance of

crabs. Hazlett and Rittschof (1975) and Bohnsack (1976) found
that crab density was directly proportional to crevice density in
their studies of the Florida population. Patterns of activity
and size of home range were found to differ between males and
females in these studies (Hazlett and Rittschof, 1975) and may
influence both sex ratios and overall abundance.

This paper reports results of field studies conducted in the
Florida Keys, the southeastern Bahamian archipelago and in
several Antillean sites from the Dominican Republic to Antigua.
The purpose of these investigations was to identify and define
parameters of Mithrax spinosissimus populations throughout the
region. Conclusions about population characteristics are drawn
from both these data and from previous studies.

METHODS

Seven sites in the Caribbean were investigated (Figure 1) .
Site 1, in the Florida Keys ( 24°80 ' N ; 81°05 ' W) , was surveyed in
January, 1984, July, 1986 and July, 1987. The census was
conducted along manmade canal walls and old quarries particularly
at Grassy Key. One canal at Big Pine Key and three canals at
Little Torch Key also were surveyed in July, 1987. These data
were combined with those from Grassy Key. Site 2 was located at
the main reef off Abraham Bay along the south coast of Mayaguana,
Bahamas (22° 0* N ; 73° O' W) and was surveyed between February
1983 and May, 1983. Site 3, located at the east end of the

Caicos Bank, just west of Long Cay, South Caicos, Turks and
Caicos Islands, British West Indies (22°05'N; 71°30 1 W) , was
repeatedly surveyed from December, 1984 to May, 1986. Site 4 was
situated at the reef drop off, on the western side of Grand Turk,
Turks and Caicos (21°90'N; 68o80'W). Crab population surveys
there were conducted between January 1984 and March 1985. An
active crab research program continues at Grand Turk and data
from this site include information collected through the summer

of 1987. Site 5 crabs were collected from the inner and outer

reefs, off the north coast of the Dominican Republic, at Buen
Hombre (19°10'N, 71°20'W) from March 1985 to September, 1986.

The crabs from Site 6, located on the south coast of the

Dominican Republic, at Azua ( 18°20 ' N ; 70°50 ' W) , were collected

over the same time period. Site 7 crabs were collected from reef
areas along northeast Antiqua, WI (61°43'W; 17°10'N), between
May, 1985 and July, 1986. With the exception of sites 1 and 5,
crabs were surveyed by sampling all individuals in a 3.05 m
diameter circle (area = 7.3 m2 ) established at random during
diving operations.

Crabs were collected by three methods. Those at site 1 were
collected at night from the shore using a pole net or by free
diving. At sites 2, 3, 4, 5 and 7, SCUBA was used to conduct
surveys of crab populations and to collect individuals. Isolated
coral heads, rubble and shoal areas and drop offs were examined

from depths of 3-35 meters. Most dives were conducted at dusk or
in the early morning hours. Lights were used to illuminate
crevices and caves. Wire mesh traps also were used to catch
crabs at sites 5, 6 and 7. At sites 2, 4, 5 and 7, extensive,
less formal searches were carried out by snorkel.

Approximately 600 crabs were sampled in this investigation.
Carapace length (CL) (distance from the rostral sinus to the most
posterior point of the carapace) and weight were measured. In
addition, crabs were sexed and state of maturity determined.

Males were considered adult if the ventral propodus length (VPL)
was equal to or greater than the CL, based on VPPL = f (CL) (see
Biddlecomb, et al. , this volume) . VPL is the ventral distance of
the propodus from the tip of the chela to the posterior end of
the socket containing the carpal hinge (see Figures 1 and 2;
Biddlecomb et a 1 . , this volume) . Females whose abdomen covered
the entire sternum were considered adult. Lost appendages and
other physical characteristics also were recorded. Captured
crabs were used to determine population parameters for each site
and to establish the percentage of individuals with a number of
missing appendages. Data on trapped crabs from Sites 5 and 7
were used to determine seasonal fluctuations, sex distribution
and reproductive status.

CL measurements on male and female crabs from six of the
seven sites (Site 6 was omitted) were used in an analysis of
variances to test for differences between sexes. Data from crabs

trapped at Sites 5 and 7 were used to determine sex ratios (t-
tests) and seasonal patterns in crab abundance. Data from both
trapped and hand caught crabs at Sites 5 and 7 were also used for
histogram plots of the CL (mm) for adult males and females.

Pooled data (all sites) were used to compare the percentage loss
of appendages for adult crabs. Chelae loss was considered
separately from loss of walking legs because of their importance
in overall body weight and their presumed role in aggression
and/or courtship (particuarly in males) .

RESULTS

General Population Characteristics

Significant variation in the size of male and female Mithrax
spinosissimus was observed both within and between populations at
the seven sites (Table 1) . Overall mean CL for adult males (N -
166) was 136.5 mm (S . D. = 17.5) with a range of 79.6 mm CL (Site

1) to 190.5 mm CL (Site 2). Adult females (N = 281) averaged

120.4 mm CL (S.D. = 16.4) with a range of 65.7 mm CL (Site 1) to

165.0 mm CL (Site 2). More than 50% of the wild caught females

were gravid at the time of capture.

Results from an analysis of variance (Table 2) showed a
significant difference (F TEST; P <0.05) between sites. The
greatest difference was observed between the Florida (Site 1) and
Caribbean populations (Sites 2-7) . Florida crabs were

significantly smaller (X = 98.8 mm CL for males; X - 83.5 mm CL
for females) than those at all other sites (Duncan's Multiple
Range ; P <0.05).

Antiguan (Site 7) crabs (both male and female) were the
smallest of the Caribbean population (Sites 2-7) with males
averaging 131.4 mm CL and females 108.2 mm CL, while crabs from
Mayaguana (Site 2) were the largest (X - 166.9 mm CL for males;

X = 140.6 mm CL for females). Males from the Turks and Caicos
Islands (Sites 3 and 4) and from Buen H ombre (Site 5) were
similar in size (with Grand Turk males averaging slightly larger)
and intermediate among the populations. Females from Grand Turk
were slightly smaller in size than Mayaguana (site 2) females,
the largest females among the sites. South Caicos (site 3) and
Buen Hombre (site 5) females were intermediate in size.

The distribution of CL (Figure 2, 3) for adult males and
females from Sites 5 and 7 showed some overlap in size between
the sexes, but males were consistently larger than females at all
sites (Table 2), (t-test; P<0 . 001) . An interesting observation
in the population from Site 5 is that an abrupt attenuation in CL
distribution for females occurs at the 135 mm CL size class
(Figure 2) . The reason for this is not known, but it suggests
either differential mortality or catch rates (primarily trapping)
at the higher size classes. It could also represent a bimodal
curve with two overlapping adult instars.

The range of CL of immature crabs overlapped the range of
adult CL for both males and females for all sites except Site 4
(females; N = 2) and Site 6 (females; N = 3) , probably due to the
small number of individuals collected. At Site 7 (N — 42) there
was a bimodal distribution of CL for juvenile males (Figure 4)
indicating two instars in the juvenile size range. These sizes
correspond with those found for instars 14 and 15 (Biddlecomb et
al . , this volume) .

Individual Sites

Site 1 ; Florida . Grassy Key

A total of 87 crabs were collected at Site 1 using a pole
net or by free diving. Ten adult males and 18 adult females had
a mean CL of 98.8 and 83.5 mm respectively. The remaining 59
crabs were immature.

The average size of adults was smaller than those determined
for all other sites. Except for the large mean carapace length
among Mayaguana crabs (unfortunately with a low N; N=ll) , the
size range between all non-Florida sites was only about one-half
of that between the Florida sites and the next nearest non-
Florida site (in size). The Florida crabs also displayed other
characteristics which differed from the Caribbean/West Indian
populations. They had a dull dark green color and in some adult
males, the chela had a bluish hue. Offspring of these crabs in
culture displayed a coloration similar to that of the adults as

well. The females were much hardier than the males when
maintained in cages. Other differences included a relatively
thinner carapace and the absence of male chela puncture scars on
the female crabs.

Five juveniles at Site 1 were found feeding in the shallows
with adult females in January, 1984. Their feeding movements and
algal species preference were identical to the adult females.
However, when pursued, they moved more rapidly and with more
direction than the adults.

Site 2 : Bahamas . Mavaquana

At Site 2, a total of 30 crabs were captured using SCUBA of
which 12 were measured. CL averaged 166.9 mm (N = 6) for adult
males, and 140.6 mm (N = 5) for adult females. The largest male
measured 190.5 mm CL, while the largest female measured 165.0 mm
CL. Only one juvenile was captured, a female measuring 54.0 mm
CL. An extensive search for juveniles or small adults was
carried out throughout the lagoon and reef areas by snorkel and
diving. Approximately 100 man-days of effort was devoted to this
search, and a number of coral heads of various spatial
arrangements were totally dissected.

Thirty hours of post-sunset/night SCUBA surveys were
conducted on the Abraham Bay reef , along the forereef spur and
groove region, at depths of 20-32 m. The grooves or sand
channels 13 -18m wide, extended between the spurs to a wide sand

plain (30 m) at the top of the dropoff. The high, porous and
sometimes cavernous spurs ran perpendicular to the reef crest.

The overlying substratum was covered by live corals (Mpntastrea
sp . ) and gorgonians. Macroalgae dominated areas of eroded
limestone. Macroalgal species included Halimeda sp. . Lobophora
sp. , Microdictvon sp . , Piety osphaeria cavernosa . Valonia sp . ,
Padina sp. with occasional Saraassum sp . The currents were
predominantly longshore. Water temperatures ranged from 25.2 -
29.8° C during the period of study.

Twenty-two of the 30 crabs caught during the study period
were taken from a single cave at the seaward end of a spur (25 m
water depth) . The cave was large and cavernous, approximately
2 o 5 x 10 m with small (0.8m x 1.2m) interconnecting crevices
hollowed out of the coral rock. These, and other smaller
crevices led further in to and out of the spur. Light levels in
the cave were low. Many of the crabs clung to the cave ceiling,
but when approached by divers, retreated into crevices.
Approximately 30-35 large crabs were observed in this cave and
adjoining crevices. Four others were captured from the dropoff
wall while grazing on algae, and one adult male was found in the
shallow forereef (1 m water depth) .

Site 3 : Turks and Caicos , Caicos Bank , Long Cav

Forty crabs were captured at Site 3 by free diving and
SCUBA. Nine adult males (X = 138.3 mm CL) and 26 adult females

(X = 114.2 mm CL) were taken. Five juveniles were also captured.

Site 3 adults were characterized by a heavy growth of
encrusting algae on the carapace. Carapace color was a light
venous red, with blue in the chela of the older adult males.
Carapace thickness was relatively moderate. The crabs were very
docile, and easily handled.

During 12 hours of underwater survey, approximately 25 coral
heads were examined. Two coral heads were found with large
populations of crabs. One small head (approximately 35 m2)
contained well over 50 crabs, mostly females and juveniles. On
surrounding heads, single males were common but females were
rarely encountered. The second head, not quite as densely
populated, was located some 300 m away. Again, crabs were sparse
on the surrounding heads. Most of the surrounding heads were
similar in size and coral composition. Local fishermen do not
normally harvest crabs. This site was visited on numerous
occasions over a two-year period and the pattern of crab
distribution remained static.

Site 4 : Turks and Caicos , Grand Turk

A total of 74 crabs were captured at Site 4 using SCUBA and
free diving. The 25 adult males had a mean CL of 146.8 mm
(largest = 180.0 mm), and the 59 adult females a mean CL of 136.5
mm (largest = 158.2 mm).

The carapace and appendages were a deep venous red and had

little or no calcareous encrustation. The adult males had
comparatively large chela and although intimidating in
appearance, they were very docile. Generally scarless, the
chitin was relatively thicker than all but the Site 5 population.
Most were caught in >20 m water depth.

SCUBA Survey, site 4

A SCUBA survey was conducted along the west coast of Grand
Turk (Figure 5) . A deep vertical coral covered "wall" parallels
the coast. Starting with the spur and groove region at a depth
of 10-20 m at the wall crest, it drops to a small shelf at
approximately 100 m before dropping down to a depth of more than
1 km. The current flow is primarily offshore.

The wall was surveyed in two sectors, one north and one
south of the island's main shipping point (South Dock). The
northern sector (Figure 5; a) was 4 km in length. Approximately
50 hours of survey were conducted using SCUBA. This section of
the wall included areas of rubble, Montastrea so. . dense soft
corals (gorgonia forests) , open areas with small caves and mixed
coral populations. All crabs were located between depths of 17-
35 m and 16 of the 24 caught were found in a fairly open area
(al) , with many small, interconnected caves. This area,
approximately 50 m in length, was the only one of its kind along
both the north and south sectors. Crabs were found either at the
entrances to crevices, or just inside them. No two crabs were

found in the same crevice.

The southern sector (Figure 5; b) ran 3 km south from South
Dock. Forty seven hours of SCUBA were used to survey this
section. Shoreward of the wall, the reef community was dominated
by Montastrea sp. . This gave way to a diversified coral reef
community along the edge of the wall. The wall itself drops
vertically and is characterized by a mixed community of sponges
and gorgonians.

All 15 crabs were found between depths of 15-28 m, along a
300 m section (bl) of a diverse coral community dominated by
Montastrea sp . Crabs were located in small crevices in the wall,
between plates of the deeper Montastrea sp . and in small heads at
the edge of the drop-off. In two instances, female pairs were
found together in a small head. No other crabs were found
together and no other crabs were found outside the 300 m section.

Juveniles were seen on numerous dives. Compared to adults,
they were extremely quick and agile. All were observed on or in
coral and numerous molt shells (<40 mm CL) were found at the base
of small coral heads. Extensive search patterns by snorkel and
diving were also carried out through the east lagoon area for
juveniles and adults.

Site 5 : Dominican Republic , Buen Hombre

A total of 201 crabs were collected at Site 5 by SCUBA and
trapping over an 18-month period. The 57 adult males had a mean

CL of 140.2 mm and mean CL for the 125 adult females was 122.9
mm. Fourteen juvenile males and five juvenile females were
captured. The largest male and female measured 168.4 mm CL and
141.7 mm CL, respectively.

Male chela were not proportionately as large as those at
other sites; however, the chitin was extremely thick. It was
common for males to have one or more puncture wounds on both the
chela and legs. These males were more aggressive than those
found at any of the other sites. About 20% of the females had
puncture scars. They too, were more aggressive than females
found at other sites. There appears to be a large crab predator
population at Site 5, which includes hogfish, groupers, and
humans. Newly molted adults at Site 5 have a carapace which is
dark, dull red in color. The new spines are light colored with
white tips. In both sexes, the carapace is soon encrusted with
crustose or coralline algae, serpulid worms, and soft algae. All
crabs captured at the site, except two newly molted individuals,
had small barnacles encrusting the carapace at the gill outlet.

Trapping, Site 5

Among the total number of crabs trapped, adult females
outnumbered adult males 3.5 ; 1, and 81.7% were gravid at
capture. The percentage of gravid females varied from a low of
55.6% in July to a high of 100% in April. Juvenile males made up
3.8% and juvenile females 1.3% of the total. Catches varied from

a low of six crabs in November to a high of 34 in March, No crabs
were caught during the months of December and January because of
unfavorable weather conditions or technical difficulties.

SCUBA survey, Site 5

In Buen Hombre , 105 hours of SCUBA were spent examining 130
coral heads over an area divided into nine sectors (Fig. 6) . The
heads in sectors A, B, C and D rose from an average depth of 18 m
to an average depth of approximately 12 m below the surface.

These heads were approximately 50-60 m , and covered with
macroalgae and a few soft corals present. Sectors E and F
contained numerous heads with an area of approximately 7 0-8 0 m2 .
These heads rose from a sandy bottom at 11-14 m and were covered
with heavy coral growth, reaching up to an average depth of 3 m
below the surface. Sectors G and H contained heads similar in
size and growth to sectors A, B, C and D, but rose from a grassy
bottom at 11-14 m. Area I contained a large diversity of heads
and shoal areas in depths ranging from 3-10 m. Of the 130 coral
heads examined, 19 were found to be occupied by M^. spinosissimus
crabs. A total of 65 crabs were found over the entire area of
sectors A - I .

In the 13 coral heads found to contain adult males, there
were only two instances in which more than one male was present
(Table 3) . In the 11 heads with a predominantly female
population, seven contained more than one male. Of the six

juveniles located, four were found on heads with other juvenile
crabs, but only one was found on a head with adult males. Three
heads (two in sector B; one in sector I) had large crab
populations, but with the exception of sector I, could not be
relocated. A head in sector I was revisited on numerous
occasions and several adult crabs were removed for study.

Three crabs were found dead and a fourth dying. All four
were adult, one was male, and all were found at different
locations and times. All had apparently died where they were
found. One female was on a ledge outside a crevice, the other
three crabs were out in the open, more than a meter from cover.
The chelae on all four were extremely worn. The chela (one chela
was missing) on the moribund crab was worn to points. Upon
dissection and examination of the dying individual, complete
atrophication of the musculature had occurred.

In one dive outside of the nine sectors, five large males

(140+) were found wandering in a sandy plain (>1 ha.) littered
• • ?

with dispersed small (10 m ) heads. No females were observed.

As at Site 4, juveniles were located on numerous dives. They
too, were observed to be extremely quick and agile. All were
found on and in coral with the exception of one, which was ('35
mm CL) found on a gorgonian at night. Numerous molt shells (<40
mm CL) were found at the base of small coral outcrops. Molt
shells were also located along the forereef area at this site,

but in no particular pattern. Numerous small crabs (15-30 mm CL)
were found in grouper stomachs.

In addition to the above formal search processes and trap
catches, numerous snorkel searches were undertaken throughout the
extensive lagoon/mangrove complex. These were undertaken
primarily in a search for appropriate macroalgal feed, but any M.
spinosissimus crabs sighted were taken.

Site 6 : Dominican Republic . off Azua

At Site 6, 47 adult females, three juvenile females, and 12
juvenile males captured in traps were brought to the laboratory.
The adult females had a mean CL of 122.0 mm, and the largest was
141.1 mm CL. Data for trapped adult males was unobtainable
because they were marketed by the fishermen.

Crabs at this site resembled Site 5 crabs in color, but
generally lacked scars. The crabs surveyed had an encrusting
growth of barnacles on the carapace.

Site 7 : Antigua , northeast reefs

A total of 168 crabs were collected at Site 7 by SCUBA and
trapping over a 15 month period. Of 105 males, 45 were
juveniles. The 60 adults had a mean CL of 131.4 mm. Forty-four
adult females averaged 108.2 mm CL. There were 16 juvenile
females caught. The largest male and female measured 157.4 mm CL
and 127.0 mm CL respectively.

The mean of Site 7 crabs were slightly smaller in size

compared to the Site 3 crabs. The VPL for adults having
undergone normal molting was significantly longer than CL.
Coloration was deep, venous red. Carapaces were usually free of
significant encrustation by coralline algae. Epifaunal species
consisted of bryozoans, small serpulid worms, an occasional
sponge, and a few goose barnacles along the gill areas. Chela
wear was negligible, and carapace spines were sharp. Males were
generally slow, quite docile, and easily handled. Females often
exhibited strong defensive behaviors, using legs and body
movements to aid chelae effectiveness.

Trapping, Site 7

Of the total 168 crabs collected, 148 were caught in traps.
Trapped adult males outnumbered adult females by 1.07:1. A large
segment of the total were juveniles: 29.7% male and 10.1% female.
A high percentage (62.8%) of the adult females were gravid and
ranged from zero (N=3 ) caught in May to a high of 84.6% of those
caught in November. No crabs were caught during the months of
August and September for logistic reasons. The greatest number
were caught in November (N=19) , with females outnumbering males
about 2:1.

SCUBA Survey, Site 7

Over 120 hours of SCUBA and skin diving surveys were
conducted at several locations around Site 7. Crabs or crab

remains were sighted at all locations except in algal ridges due
east. Three small "pockets" of crabs were found northeast, north
and west of the island. At the northeast area inside the mouth
of Parham Sound (15-25 m depth), five males and one female (all
adult) were captured on a 15-25 m length of an old algal ridge
(dead coral) with 2-3 m (diameter) boulders. The water was
turbid. At nearby spur and groove formations, no crabs were
found. In a trapping area leeward of the well-developed algal
ridge, two crabs were found dead in traps, but none were seen on
the ridge, or in the mounds of dead Acropora sp and
Montastrea sp .

At the northern end of the island, shallow (1-3 m depth)
patch reefs were examined. Six males were located among the
patches of Acropora sp . and Millepora sp .

On the west coast of Antigua, shallow boulders have fallen
from a cliff face fronting the sea at Dickerson Bay, providing a
refuge for spinosisimus . In the 1-2.5 m crevices in these
boulders, five males and one female were captured. The area
experiences strong wave action along with wakes from boating
activity associated with nearby resorts.

These three small "pockets" of crabs produced a total of 16
males and two females. Crabs were not captured by SCUBA at any
other locations, although four were sighted, two in a single
coral head.

Seasonal Variation

Trapping results by month at Sites 5 and 7 (N > 140) showed
variation in catch rates by month, sex and site (Fig. 7) . At
Site 5, peak catches of females occurred in March, May and
November. Male catch rates were lower overall and varied less
than those for females with modest peaks in March and August.

A somewhat different pattern occurred at Site 7. Overall,
male catches exceeded female catches, with peaks in April, June
and December. Female catches peaked in November.

Sex Ratios

Data on trapped crabs from Sites 5 and 7 showed a high
degree of variation in the sex ratio between populations (see
Figure 7) . Females were significantly more abundant than males
( t-test ; P<0„01) at Site 5 by a factor of 3.5:1 (115 females: 33
males). No significant difference (t-test; P>0.10) was found in
the ratio of females to males (46 females: 46 males) at Site 7.

Appendage Loss

Three hundred twenty five of the crabs sampled were
recorded as to the total number of appendages present. Appendage
loss in populations from all sites ranged from zero (complete) to
six (Table 4). Of the 197 females recorded, 67 (34%) had all
appendages present and of the 128 males recorded, 38 (29.7%) had
all appendages present. Female crabs missing one appendage

(N=68 ) were 3% lighter than the complete female crabs. Those
missing two appendages were 7.5% lighter. Chelae loss was
considered separately from loss of walking legs in males. Males
with all appendages present (N-38) had a mean CL of 137.4 mm
(weight = 1607 g) . They were 2.2% heavier than males with one
missing leg (N=25) and 14.8% heavier than males missing both a
walking leg and a chelae (N=7) . We found that 26.6% of all males
recorded were missing at least one cheliped (N=34) .

DISCUSSION

Archaeological studies indicate a long-term presence of
Mithrax spinosissimus in the Caribbean region. Collins and
Morris (1976) found evidence of Mithrax spinosissimus in Pliocene
and Pleistocene formations in Barbados while Rathbun (1923) found
shell fragments of M._ spinosissimus in Pleistocene formations in
Haiti. The present-day distribution of It spinosissimus is
somewhat in question, particularly in regard to the northern
limit of its range. Rathbun' s (1925) description of two
chelipeds found at an undetermined locality in North or South
Carolina do not agree with the known ranges from these and
previous studies. The possibility that these chelipeds were
transported from a more southerly site should be considered.

Results from this study showed a high degree of variability

in the populations of Mithrax spinosissimus throughout its range,
and yet a rather narrow range of characteristics within a
population. This suggests that fishing pressure, geographic
isolation, genetic drift and adaptation to local conditions have
played an important role in shaping the characteristics of the
individual populations. Differences were particularly evident
between the Florida and Caribbean populations where the Gulf
Stream may act as a barrier to dispersal of crabs from the
southern populations, particularly given the brief planktonic
stage. While samples taken in Florida and South Caicos might
have been biased by limited collection time, searches at the
other sites were extensive, and it is highly unlikely that major
segments of the M_i_ spinosissimus populations could have been
totally overlooked.

Despite the small sample size for some sites, results showed
that the Florida population was unique in several aspects from
all other populations investigated. Crabs from South Florida
taken in this study were similar in size to those of previous
studies (Hazlett and Rittschof, 1975; Bohnsack, 1976) and were
significantly smaller than those at all other sites.

Additionally, Florida crabs were different in coloration and
carapace thickness. These physical characteristics may be
attributed to lower water temperatures and/or different habitats,
particularly adaptation to living in crevices in turbid waters.

Crabs also appear to exhibit different behavior in the

Florida vs. Caribbean populations. Hazlett and Rittschof (1975)
reported that in cases of multiple occupation by crabs of a
crevice in South Florida, pairs most often consisted of a male
and a female and least frequently of a female and female.

Bohnsack (1976) recalculated these data, correcting for small
sample size, and concluded that female pairs were not rare, but
male-female pairs were still more common. Bohnsack' s (1976) own
data agree with those of Hazlett and Rittschof (1975) in
determining male-female pairs to be the most common, but he found
significantly more multiple occupations of crevices by females.
Pairs of males were rare. This finding differs from our
observations in the Caribbean populations (Sites 2-7) in which
the most common pair or group consisted of multiple females with
none or a few males. These large phenotypic and behavorial
differences between the Florida and Caribbean populations suggest
the Florida crabs may be a subspecies.

The sizes of Caribbean crabs (Sites 2-7) found in this study
were consistent with Monroe's (1974) data from Jamaica where the
mean size of males was 133.4 mm CW (-130.4 mm CL) and 122.8 mm CW
(=120.4 mm CL) for females. Jamaican crabs were intermediate in
size among those investigated in this study and were not
significantly different from the populations at Antigua (males)
or Buen Hombre (females). Size does not seem to be directly
related to temperature since the largest (Site 2) and the

smallest (Site 7) crabs found in the Caribbean region were at the
northern (cooler) and eastern (warmer) extremes respectively.
However, between Mayaguana and Antigua, mean minimum temperatures
are not likely to be more than a degree or two centigrade.

Florida crabs were consistently smaller than any of the Caribbean
crabs. They would also be subjected to winnter temperatures on
the order of 5°C lower than the Caribbean sites.

Density of crabs varied by site, season and sex. Crab
densities appear to be correlated with both habitat (resource
availability) and social factors. In the Florida population,
where crabs occupy canal wall crevices, crevice density appears
to be the limiting factor in crab abundance (Hazlett and
Rittschof, 1976; Bohnsack, 1977). Our studies suggest that
"crevice" availability may be important in at least some areas
where competition for space exists.

Fishing pressure may account for some of the observed
variation in population density. In areas experiencing heavy
fin-fishing pressure, crab abundance was found to be greater.
Intensive fishing pressure occurs in the Dominican Republic (Site
5) , Antigua (Site 7) and along the Caicos Banks (Site 3) where a
commercially important fishery for spiny lobster exists. Fishing
pressure may reduce crab predator abundance as suggested by Munro
(1974) for Jamaica, or reduce competition for other resources
such as crevice availability or food resources. By contrast,
crab densities were low in areas experiencing little fishing

pressure such as that found along the west coast of Grand Turk
(Site 4) , an area protected from fishing. While large adult
crabs, as nocturnal grazers, are probably relatively free from
heavy fish predation, few juveniles are likely to recruit to
adults when fish predation is heavy.

Seasonal variation in trapped crab abundance has been
observed by Munro (1974) for Jamaican crabs and at Sites 5 and 7
in this study. Munro (1974) found higher catches for crabs in
March/April, June/ July and November/ December over a two year
trapping period. Catches at Sites 5 and 7 showed seasonal
fluctuations, particularly at Buen Hombre (Site 5) . Part of the
fluctuation at site 5 is due to the shift by fishermen in Buen
Hombre to agricultural and other landbased occupations in the
fall and winter (Stoffle, 1986) . The decline in catch rates,
particularly for females, probably reflects this declining effort
by the fishermen versus an actual seasonal variation in the
population density.

Results from trapping studies at Site 5, in which females
were caught at a rate of 3.5:1, indicate a bias towards females
in this population. Assuming a 1:1 sex ratio at recruitment,
this suggests that there is a differential mortality rate for the
sexes at this site or that "excess" males move away from the
area. Hazlett and Rittschof (1975) found in the Florida
population that males moved more frequently, had a larger home

range and a greater activity radius than females. In our
studies, "bachelor" males were more often found far from their
crevices, generally alone. Trap effect may also account for some
of the variability if inter-male aggression influences the
probability of more than one male entering the trap. Only one
recorded instance of two males in a trap occurred at Site 5,
although Munro (1974) reports several males captured at a time in
his Jamaican trap study.

Not all sites showed the same sex ratios. There was no
difference found in the sex ratio at Antigua (Site 7) and males
outnumbered females almost 2:1 in trap catches in Jamaica (Munro,
1974) . Variation in the sex ratio between sites suggests other
factors such as fishing pressure or intraspecific competition
contribute to a population's sex ratio. Fishing pressure may be
of particular importance in the Dominican Republic and South
Caicos where large male It spinosissimus are locally marketed.

M. spinosissimus is generally gregarious, usually with pairs
or several individuals occupying a cave or crevice. Bohnsack
(1976) found that about 55% of the occupied holes in his study
contained crabs in clusters of 2-11 individuals. On the basis
of his results, he suggested that agression occurs between males
since only one of 42 clusters examined had more than one male.
Hazlett and Rittschof (1975) found a negative relationship
between average day-to-day movement for males and male density, a
relationship not observed for females. Our studies also support

the concept of a gregarious social structure with males
apparently holding a "harem" of one to several females. Multiple
occupation of crevices by females was commonly observed at Sites
2, 3, 4 and 5. At Site 5, female-female pairs were encountered
most often, with as many as five females found in a crevice.

Males were found more than one to a crevice in only two
instances: the cave at Site 2, and a large hole (1 m x 2 m) on
head II, Site 5. Divers examining "harem" heads after trapping
commonly found continued high female populations, with three or
four females sharing a small coral outcrop. At one of these
heads, a single unbaited trap was placed at the edge of the
grazing ring, resulting in seven females (six gravid) caught
after a two-day set. Females were seldom caught individually,
but as previously noted, two males in a trap were rare.

Mariculture experiments at Sites 4, 5 and 7 showed little
aggression in the captive populations. Crabs were reared from
egg and stocked in large growout cages (1.3 m x 1.3 m x 2.6 m)
with as many as 25-35 crabs per cage. Brood stocks of wild crabs
were maintained in holding cages (1.3 m x 1.3 m x 0.75 m) and
were stocked with 10-12 crabs, usually consisting of one male and
several females. One instance of mortality assumed to have
resulted from aggression was noted at Grand Turk, when a recently
introduced male was found dead in a holding cage containing one
male and several females. Otherwise, aggression was not observed

in the captive adult populations.

A considerable percentage (66% females; 69.1% males) of the
adult population is found without walking legs and/or chelipeds ,
suggesting some non-mortal predation and the ability to continue
with less mobility and feeding potential. The frequency of
highly worn claws and encrusted shells and algae suggests a
relatively long life span once an animal has reached adulthood.

A relatively large number of old, apparently moribund crabs also
suggests moderate predation of adults. The population data
occasionally suggest a second molt in males.

The differences observed among the populations in this study
suggest that present-day populations have been isolated for a
long period of time and represent separate stocks of this
species. The large differences between the Florida (Site 1) and
Caribbean (Sites 2-7) populations may be the result of
environmental, geographic and hydrologic factors coupled with the
species short planktonic stage. The Caribbean region is
characterized in general, by volcanic islands (Antillean), or
groups of islands on broad, shallow banks (Bahamian) separated by
deep channels such as that between South Caicos (Site 3) and
Grand Turk (Site 4) and between these sites and the Domincan
Republic (Site 5) . These natural barriers presumably restrict
interchange between the populations.

The possibility of larval exchange between the populations
is limited by the short duration (< 125 hr) of the free-swimming

planktonic stage (Provenzano and Brownell, 1977; Porter et al . ,
1987) which permits little opportunity for dispersal of the
larvae over long distances. Sastry (1983) states that restocking
of benthic crustacean populations generally occurs from the
larvae retained within the geographic range of the species.

Given the extremely brief planktonic stage of this species and
limited movement of the benthic population, it seems likely that
restocking of spinosissimus populations generally occurs from
larvae retained within the restricted range of each individual
population.

Recruitment, then, must come from the offspring of the local
adult population. At all sites, juveniles were found to inhabit
the same areas as adults. They were abundant at Sites 5, 6 and
7, where trapping of crabs was successful. Scarcity of juveniles
reported in previous studies such as those of Munro (1974) may be
due to the cryptic nature of the juveniles. Juvenile M.
spinosissimus smaller than 20 mm CL generally decorate their
carapaces with algae and inhabit small crevices, making them
difficult to detect. Munro (1976) failed to locate juveniles by
diving or capture in traps but reported the occurrence of several
juveniles in the stomachs of the red hind Epinephelus quttatus .
Juveniles were located in the stomachs of groupers (Site 5) in
this study. Traps were generally successful for capturing larger
(> 60 mm CL) juveniles, but were used only at three sites.

Our studies show that variation in size, sex, social

structure, abundance and presumably recruitment of M.
spinosissimus must be considered for individual populations in
any fishery model of this crab in the Caribbean. The variation
observed throughout the region suggests little exchange between
the populations. If each population is a separate stock, the
potential exists for rapid decline in local populations
experiencing heavy exploitation. On the other hand, the crabs
are easily bred, and considerable opportunities exist for stock
" improvement . "

CONCLUSIONS

Information on the size, sex ratio, abundance and ecology of
M. spinosissimus has been reported. A high degree of variation,
particularly between the Florida (Site 1) and Caribbean (Sites 2-
7) populations, was observed. This suggests populations have
been isolated for a long period of time and little genetic
exchange occurs between the separate stocks. The Florida
population is significantly smaller than any other population
sampled and differs from Caribbean/West Indian ft spinosissimus
in a number of morphological and behaviorial features.

In the Caribbean/West Indian region, adult Mithrax
spinosissimus populations are bimodally distributed with males
approximately 15-25% larger (CL) than females. These nocturnal
crabs inhabit caves or crevices in coral reef habitats and

usually consist of a single male with a "harem" of several
females. In spite of this social structure, serious adult inter¬
male aggression seems rare, even when confined to cages.

Crab abundance varies by site, sex and season. Factors
affecting abundance appear to be crevice availability (directly
proportional) and fishing pressure (inversely proportional) . A
large number of females are found gravid year-round. Juveniles
are rarely seen in the wild, probably due to their cryptic habits
and in the earliest instars algal decoration.

LITERATURE CITED

Biddlecomb, A. B. Morphology and relative growth of M.
spinas issimus . This Treatise.

Bohnsack, J.L., 1976. The spider crab, Mithrax spinosissimus
an investigation including commercial aspects. Florida
Scientist. 39(4): 259-266.

Colin, P. 1978. Caribbean reef Invertebrates and Plants. T.F.H.
Publications Inc. Neptune City, N.J.

Collins, J.S.H. and S.F. Morris, 1976. Tertiary and Pleistocene
crabs from Barbados and Trinidad. Palaeontology, 19(1):
107-131.

Hartnell, R.G. , 1982. Growth, The Biology of Crustacea, p. 111-
256. Academic Press, Inc.

Hazlett, B. A. and D. Rittschof, 1975. Daily movements and home
range in Mithrax spinosissimus (Majidae, Decapodae) Mar.
Behav. Physiol. 3: 101-118.

Idyll, C.P., 1971. The crab that shakes hands. Nat. Geogr.

Mag. 139: 254-271.

Munro , J.L. 1974. The biology, ecology, exploitation and
management of Caribbean Reef Fishes. Part V. The
biology, ecology, and bionomics of Caribbean Reef Fishes :
Crustaceans (spring lobster and crabs) Res. Rept. Zool.

Dept. Univ. West Indies 3 VI pp 39-48.

Porter, K.P. , J.M. Iglehart, W.H. Adey and M.W. Yadven. Cage
culture of the Caribbean King crab (M^_ spinosissimus .

Lamarck) in conjunction with algal turfs. In review. Proc.
of Caribbean Aquaculture Symp.

Provenzano, A.J. and W.N. Brownell, 1977. Larval and early
post-larval stages of the West Indian spider crab,

Mithrax spinosissimus (Lamarck) (Decapoda: Majidae)

Proc. Biol. Soc. Wash. 90(3): 735-752.

Rathbun, M.J., 1923. Fossil crabs from the Republic of Haiti.

Proc U.S. Natn. Mus. 63, 6pp., plsl, 2 .

Rathbun, M.J., 1925. The spider crabs of America. United States
National Museum Bulletin 129, 613 pp.

Stoffle, R.W. , 1986. Caribbean fisherman farmers, a social
assesment of Smithsonian king crab mariculture. Inst.

Soc. Res., U. of Mich. Ann Arbor, Michigan.

Sastry, A.N., 1983. Pelagic larval ecology and development.

The Biology of Crustacea, Vol.7, p. 213-282. Acadamic Press,
N.Y., N.Y.

Williams, A.B., 1984. Shrimps, lobsters and crabs of the
Atlantic Coast of the eastern United States, Maine to
Florida. Smithsonian Institution Press, Washington, D.C.,
550 p.

LEGENDS

FIGURE 1.

Map of Caribbean, showing seven study sites. 1 =

Grassy Key, Florida; 2 = Mayaguana, Bahamas; 3 = South
Caicos, Turks and Caicos, B.W.I.; 4 = Grand Turk,

Turks and Caicos Islands, B.W.I.; 5 = Buen Hombre,

Dominican Republic; 6 = Azua, Dominican Republic; 7 =
Antigua, W.l.

FIGURE 2.

Histogram plot of CL (mm) for adult male and female
crabs from Buen Hombre, Dominican Republic (site 5) .

FIGURE 3.

Histogram plot of CL (mm) for trapped adult male and female
crabs from Antigua, W.l. (site 7).

FIGURE 4.

Histogram plot of CL (mm) for immature trapped crabs
from Antigua, W.l. (site 7).

FIGURE 5.

Map of Grand Turk, showing survey areas along western
wall; a = "north sector"; b - "south sector." a = 11
females, 10 males captured; a2 = 3 females captured;

— 6 females, 9 males captured.

FIGURE 6.

Map of coral reef (forereef, reef crest and back
reef) area around Buen Hombre, Domincan Republic
showing nine sectors surveyed by SCUBA (site 5) .

(Sectors A-D: 18 meters depth; E-H: 12 meters depth;

I; 10 meters) See text for description of sectors.

FIGURE 7.

Number of trapped male and female crabs by month at
sites 5 and 7.

TABLE 1.

Number, S.D., and range of CL for each site.

TABLE 2.

Results of ANOVA.

TABLE 3.

Results of SCUBA survey.

TABLE 4.

Percent appendage loss.

Figure 1

35 -

•<*

*3°

■<*

<3*

f—

r“

T“

V*“

T—

T“

▼“

T™

T“

T-

115-

120-

125-

130-

135-

140-

145-

T-

155-

CARAPACE LENGTH (mm)

Figure 2

160-164

Male

savao do aaawnw

Figure 3

CARAPACE LENGTH (mm)

10i

691.-99I-

*91-091

691--99I.

*91.-091.

6*L-9*I.

PW-OPl

681.-901.

*81.-081.

621.-981.

*21.-021.

6U.-9U

m-ou

60L-90L
*01.-001.
66-96
*6-06
68-98
*8-08
6Z-9 Z
*Z-0Z

"1 — r

3iiN3Anr

nnav

savuo 3iviAi jo aaaiAinN

Figure 4

CARAPACE LENGTH (mm)

GRAND TURK

Figure 5

Figure 6

MONTH

Figure 7

AUG.

TABLE 1

Mean, S.D. and range of CL (mm) for adult male and
female Mithrax spines issimus for each site. See text
for description of sites.

Site

sex

mean (mm)

Range

(mm)

98.8

14.8

79.6

110.0

83 . 5

9.9

65.7

94 . 9

166.9

17.3

141.0

190.5

140.6

20.9

114.3

165.0

138.3

8.3

127.0

148 . 0

114.2

10.0

93.8

136.6

146.8

22 . 0

104 . 1

180.0

136.5

9.7

114 . 6

158.2

140.4

9.7

121.0

168 . 4

126

122.9

8.9

91.6

141.7

122 . 0

8.7

106.3

141.1

131.4

8.7

112 . 1

157 . 4

108.2

8.7

90.3

—

127 . 0

TABLE 2. Results of analysis of variance used to test for

differences in CL (mm) for adult males and females
between sites (excluding site 6) . Subsets (S) with
different letters are significantly different (Duncan's
multiple range; P< 0.05); DF=5 . An asterisk (*)
indicates that males and females within a site are
significantly different in size (t-test; P< 0.05).

Sites

MALE

FEMALE

Mean

98.8

83 . 5

166.9

140.6

138.3

114.2

146.9

136.5

140.4

122.9

131.4

108.2

Q W Pm U ffi H

TABLE 3

SECTOR

Total :

Results of 105 hours of SCUBA survey in Buen Hombre,
Dominican Republic (site 5) .

# OF MALES

CORAL HEADS

FEMALES JUVENILES TOTAL

CRABS

2-17

9-29

4-27

1-2

2- 13
1

3- 28
1-2

4- 10
130

3
2

4
4
8
4
1
0
1
1
1
0

TABLE 4

CRAB

MALE

FEMALE

Percent appendage loss for adult male and female crabs
from all sites. Chelae and walking legs are treated
separately for males only.

N % CHELAE

LOST

1 2

128 20.3 6.3

197

% WALKING LEGS LOST
0 12 3 4

29.7 23.4 16.4 2.3 1.6

34.0 35.0 16.8 9.1 4.6

5 6

— 0.5

SECTION

II: BIOLOGY AND ECOLOGY OF MITHRAX SPINOSISSIMUS

REPRODUCTIVE BIOLOGY, FECUNDITY AND EMBRYONIC
DEVELOPMENT OF M. SPINOSISSIMUS.

REPRODUCTIVE BIOLOGY , FECUNDITY AND EMBRYONIC DEVELOPMENT

OF MITHRAX SPINOSISSIMUS

M.P. Craig, K.L. Porter, R.V. Ruark and J.M. Iglehart

Abstract

The reproductive morphology, size of female and clutch at
maturity, fecundity, seasonality of reproductive effort,
embryogenesis and duration of egg incubation were investigated
for the tropical Majid crab, Mithrax soinosissimus . Crabs were
collected from the Dominican Republic, Antigua, the Turks and
Caicos Islands, and the Florida Keys.

Mithrax soinosissimus is sexually dimorphic. The male
prepubertal instar ranges in carapace length (CL) from 80.0-115.0
mm. The mean CL of ovigerous crabs ranges from 83.5-137.5 mm
depending on the study site. The mean CL of mature males ranged
from 93.9-146.7 mm. Reproductive capability was determined to be
year round. Captive crabs were observed to mate during the
intermolt period while positioned ventrally to each other.

Female It soinosissimus have the capability to store spermatazoa
for long periods of time. Mean clutch size is 6.0639 X 104 eggs
(S.E.-3025; N— 20) .

Embryogenesis is divided into five distinctly observed
stages which develop over a mean time period of 29.5 days
(S . D=0 . 5 ; N—4 ) . All larvae were hatched within 12-16 hours. The
mean interval of time between hatching of eggs and the spawning
of a new batch was 61.9 days (S . D.=19 . 6 ; N=247) . Each female had
an average 3.75 hatches over a 10 month period in captivity. It
is concluded that the reproductive capacity of Mithrax
soinosissimus appears relatively unaffected by conditions of
captivity. All reproductive characteristics of If soinosissimus
are suitable for mariculture.

INTRODUCTION

The molt of puberty or maturation molt denotes reproductive
capability in Ma j idae (Hartnoll, 1963; Ingle, 1983) and may be
attained after 16-20 postlarval molts in Brachyura (e.g.,

Callinectes saoidus or blue crabs: Van Engel, 1958) . Hartnoll

(1965) found female members of the Majidae from tropical waters
(including Mithrax sculptus) only able to physically mate and
reproduce after the maturation molt. Terminal anecdysis or the
cessation of molting is said to coincide with the maturation molt
in Majidae (Hartnoll , 1963, 1965, 1974). In Mithrax sculptus .
the maturation molt (Hartnoll, 1965a) is accompanied by distinct
external morphological changes in the chelae of males and by a
variety of changes in the shape and function of the abdomen,
pleopoda, sternum and gonopores of females. An internal change
occurs in the males' sperm ducts, the posterior part of the
genital tract. They characteristically appear pale and
translucent when immature, whereas in mature males, the sperm
ducts become swollen and assume an opaque white appearance
indicating the presence of mature sperm. In females, the
maturation of the ovaries is evident by marked changes in the
developing ova. The diameter of the ova increases as much as
eight times before spawning and the color changes as the yolk is
deposited (Hartnoll, 1965a) .

Usually, maturation of the gonads is considered to coincide
with the development of certain anatomical characteristics.
However, Hartnoll (1965b) mentions that some of the larger
prepubertal Mithrax sculptus and Microphrvs bicornutus . both
Majids, had partly swollen sperm ducts which contained ripe
sperm. In males of the Majid crabs Inachus and Macropodia and in
females of Pisa tetraodan, the gonads of some prepubertal crabs

were maturing before the maturation molt and in Hvas coarctatus
the females' gonads always began to mature prior to the
maturation molt (Hartnoll, 1963, 1965b). Little information has
been published concerning maturation of M_j_ soinosissimus . but it
has been assumed to be characteristic of the order and family.

Teissier (1935) describes a critical molt preceeding the
maturation molt in Maia scruinado and considered this "molt of
pre-puberty" or the pre-pubertal instar, to mark the first
differentiation of relative growth rates between the sexes. The
final molt, the maturation molt occurs three molts later. In
Pisa tetraodan , Vernet-Cornubert (1958) was able to show that the
maturation molt occurs over a range of carapace lengths (CL) and
may be from the first to third molt after the pre-pubertal
instar. In a boreal crab Hvas coarctatus the maturation molt
occurs during particular seasons of the year. Where mating
occurs only during a particular season, Hartnoll (1963) found the
time of year and the size of the individual at the instance of
molting contributing factors in determining whether a crab will
make a further normal molt after pre-pubertal instar or will
undergo the maturation molt directly. Crabs of similar size were
found to undergo a normal molt or reach the pre-pubertal instar
at the same time, indicating that other factors may contribute to
the timing of the maturation molt. In four Majid species Hvas
coarctatus . Inachus dorsettensis . Macronodia rostrata and

Microphrvs bicornutus . it was found that at the same time and
within the same population of each species, specimens of very
different size were undergoing the maturation molt (Hartnoll ,
1963) .

Hartnoll (1965a) observed throughout the Majidae studied
that the maturation molt occurred over a large range of carapace
lengths. Among a group of Mithrax sculotus . differences in mm CL
between the largest and the smallest postpubertal crab,
represented as the largest being a percentage of the smallest,
was 162% for females and 314% for males (min 8 mm, max 13 mm? min
7 mm, max 22 mm) . The length of post-pubertal specimens has been
shown to vary between separate populations of a species within
the same region (e.g., Hvas coarctatus Leach around the Isle of
Man) but even when a single population was studied from a very
limited area (e.g., Microphrvs bicornutus in Kingston Harbor,
Jamaica) , these same large variations were still apparent
(Hartnoll, 1965b). Munro (1974) found ovigerous Mithrax
spinosissimus females from Jamaica to average 125.3 mm carapace
width (CW) and he believed this to be the mean size after the
maturation molt.

A considerable increase in chelae size relative to carapace
length occurs during the maturation of IL spinosissimus males,
but is not noted in the females. Bohnsack (1976) decided that
Mithrax spinosissmus males from a Florida canal population may be
mature based on increased cheliped size, once carapace width (CW)

is near 80.0 mm, though he did not examine sperm ducts for mature
sperm. Brownell, et al. (1977) reported a 94 mm CL male specimen
from Venezuela with small chelae to be sexually active before it
molted to a 115 mm CL having proportionally larger chelae.

Courtship in the Brachyura may involve a series of visual,
tactile, auditory and/or chemical signals. The male
distinguishes a female from a potential aggressor male crab by
the female cheliped size and more than likely, also by olfactory
signals or pheromones which are released by the sexually
receptive female (Warner, 1977) .

Copulation in Brachyuran crabs occur during different phases
of the molting cycle, the phases being species specific. Females
of Carcinus maenas and Cancer paqurus , Cancrid crabs, copulate
soon after molting while their integument is soft. Females of
Hvas coarctatus and Maia squinado , both Majids, mate in either
the soft or hard shelled state (Hartnoll, 1965b). Other Majid
crab species, i.e., Pisa tetraodon and corvstes . mate when
hard shelled (Hartnoll, 1963, 1969; Ingle 1983).

In matings where both male and female have hard
exoskeletons, duration of copulation is measured in minutes
rather than in days, a characteristic of those species copulating
only when the female has just molted (Hartnoll, 1969) . Corystes
cassivelaunus . a Corystid crab, can only mate for a 12-20 day
period during the intermolt when the opercula of the gonopores

become decalcified (Hartnoll , 1968b) . No data is available in
the literature with regard to the conditions required for Mithrax
spinosissmus mating.

During copulation, many crabs face each other ventrally ,
generally with the male clasping the female. The male is
positioned so that the posterior of his body is between the
females abdomen and sternum. His abdomen is unfolded and his
first pleopods are inserted into the genital openings of the
female (Warner, 1977) . In Brachyurans, the female may receive
spermatozoa as packets or spermatophores for storage in
spermathecae . Consecutive egg broods can be fertilized using the
stored sperm. For example, one female Mithrax hispidus was
observed to have laid three batches of eggs in the absence of a
male of that same species (pers. obs.). Van Engel (1958) found
sperm to survive for at least one year in the seminal receptacles
of several crab species. Moreover, females of Corvstes
cassivelaunus can store sperm for a year and then mate again
(Hartnoll, 1968b) . Munro (1974) suggested, due to the capability
of Mithrax spinosissimus to store spermatozoa in spermathecae for
multiple fertilizations of egg broods, it may only be necessary
for mating to occur only once in its lifetime. Hartnoll (1965b) ,
suggested for the Majids he studied, one copulation can fertilize
a succession of spawnings and is probably sufficient for all the
eggs a female will produce during her lifetime.

The potential number of batches of eggs carried during a year

is related to the incubation time and the time interval between

spawnings. The interval between mating and spawning may range
from days f Corvstes cassivelaunas) to months ( Cancer paemrus)
(Ingle, 1983). huaehus and Macropodia . Majid crabs, incubate
eggs for approximately three months and breed continuously,
producing at least three sets of eggs per year (Hartnoll, 1963).

There was no indication of seasonal breeding activity in the
five Jamaican Majids studied by Hartnoll (1965a) . In nearly
every sample collected from December to July, all of the post-
pubertal females were ©vigerous. He also suggests there is
continuous breeding in these tropical Majids, with a succession
of incubations extending from the maturation molt to the death of
the crab. In four of the five species where successive sets of
eggs were recorded (during an undescribed part of an eight month
experiment) , he found Macrocoeloma t rispinosum to have three sets
with an incubation time of 13 to 14 days each? Microphrvs
bicornutus . four sets with 10 to 13 days incubation? Mithrax
sculptus , two sets with 11 to 13 days incubation? and
Stenorhvnchus seticornis . three sets with 12 days incubation.
There was little delay between hatching and spawning.

Mala scminado . a boreal crab of similar carapace length to
Mithrax spinosissimus » carries only one batch of eggs per year
(Hartnoll, 1963) . However, Mediterranean populations of Pisa
tetraodon have been observed to carry six to eight batches of

eggs per year ( Vernet-Cornubert , 1958; Ingle, 1983). Brownell,
et al. , (1977) observed an individual M_j_ spinosissimus female to

produce three consecutive broods of eggs at one month intervals
while in captivity.

Several authors report observing ovigerous ft spinosissimus
during particular months of the year. Ovigerous crabs were
observed in January (Brownell and Provenzano, 1977) ; February to
August (Brownell, et al . , 1977); May and June (Rathbun, 1925);
August through November (Bohnsack, 1976) ; and year-round (Munro,
1974) .

For some tropical crabs, usually land crabs, reproduction
follows the lunar cycle (29.5 days between full moons or 14.8
days between spring tides) with larval hatching timed to coincide
with spring tides (Warner, 1967; Gifford, 1962).

Majid crabs generally produce large numbers of eggs. A 130
mm CL specimen of Maia souinado produced approximately 156,000
eggs at its annual spawning (Hartnoll, 1963), while a 127 mm CW
specimen of Cancer paaurus . a Cancrid crab, produced 1,000,000
eggs (Edwards, 1978) . Munro estimated It spinosissimus brood
size at 50,000 eggs for females of 122.8 mm CW caught in waters
off Jamaica (Munro, 1974) .

The proteinaceous yolk and lipid vesicles containing
carotenoid pigments give crab eggs their characteristic
coloration throughout incubation (Anderson, 1982) . The eyes and
pigment spots appear first, followed by the outlines of the

abdomen and cephalothorax (Warner , 1967) .

Little published data has been available concerning Mithrax
spinosissimus size at maturity , reproductive morphology,
fecundity , seasonality of reproductive effort, embryogenesis and
duration of egg incubation. This study was carried out, in part,
to provide answers to these critical questions.

METHODS

The Mithrax spinosissimus specimens examined in this study
were collected in traps and by SCUBA divers from five study sites
with depths ranging from 6-10 meters on the north coast of the
Dominican Republic (19° 80' N? 71° 20* W) (Site 1), 1-180 m in
Antigua (17° 10 » N; 61° 43' W) (Site 2), 10-25 m in Grand Turk
(21° 90* N; 71° 10* W) (Site 3), 2-5 meters in South Caicos (21°
80* N ; 71° 30* W) (Site 4) and 1-3 meters at Grassy Key, Florida
(24° 80' N? 81° 05' W) (Site 5). Each crab was weighed, measured
and tagged. For identification, colored plastic cable or Hzip
ties11 were fastened around the merus of one or two walking legs.
Lost appendages and other distinctive physical characteristics
were recorded, as well. Specimens were sexed and the state of
maturity was determined for each crab. For purposes of table
and graph construction, male maturity was determined to be the
point when ventral propodus length (VPL) becomes approximately
equal to or greater than the carapace length (CL) after molting

(Iglehart, et al. , this volume). Females were considered mature
if their abdomen completely covered their sternum. It was also
determined whether or not the mature females were gravid; for
gravid crabs the egg mass characteristics were recorded.

After initial data collection, crabs were introduced into
4.0* X 4.0' X 8.0' wooden cages covered with 1/4-1" mesh black
plastic screen. The cages floated with their tops at the
surface, and were anchored in protected lagoonal or back reef
environments. Crabs were fed diets of screen-grown turf algae
(Adey and Hackney, this volume) .

Non-gravid females were kept with mature males until the
females were observed to have eggs. None of the mature females
were isolated to determine the number of broods from one
copulation. Ovigerous females were transferred to cages without
males and egg development was observed to determine the time to
hatching. When close to hatching (one to three days) , as
determined by the egg color, the gravid female was put into a
"hatch cage" separate from other females. After hatching, the
female was then returned to a cage containing males and other
females. Several ovigerous crabs were dissected to examine
reproductive structure and to gather data on brood
characteristics and egg development.

Brood size was determined by two methods; 1) individually
counting all of the eggs in a brood from four individual crabs
and 2) by replicate calculations of dry and wet weight

determinations of individual eggs and the entire egg mass. Dry
weights (24 hours at 80°C) of an entire brood and of five samples
of 20 eggs were determined from the egg masses of six more female
broods. Wet weight was determined after blotting eggs for 10
seconds. A Mettler electrobalance (model PC 8000) was used for
weight determinations. Results were presented as individual egg
weights (dry) and total brood egg number, calculated from the
total brood dry weight to individual egg weight proportion. A
wet weight to dry weight ratio was also determined from mean
individual egg wet and dry weights. The egg mass and total
number of eggs for 20 females was determined by weighing the
females both pre and post hatch. The difference was considered
the brood wet weight. The brood size was determined using the
calculated wet weight to dry weight ratios and the mean (dry)
individual egg weight.

Approximate intervals for marked changes in gross egg
characteristics were determined from observations of live and
preserved fertilized eggs selected from broods throughout their
incubation. Eggs were removed every two days, until the day of
hatch, from the broods of several ovigerous females from just
after extrusion onto the pleopodal endopodites. These were fixed
in 5% buffered formalin and then transferred to 70% isopropyl
alcohol. External changes in the eggs were studied using a
dissecting microscope.

Duration of egg incubation was determined as the length of
time between the exact date of extrusion of the eggs onto the
pleopods and the hatch date. Data for time intervals between any
successive viable hatches from individual tagged females was used
to determine mean interval of time between hatching and spawning
of a new batch.

RESULTS

Size at Maturity

Mithrax spinosissimus crabs are sexually dimorphic. The sex
of juveniles cannot be determined visually until the crabs grow
to a carapace length (CL) of approximately 25 mm. At this size,
the females' abdomen begins to be noticeably wider than the
males' through the fourth, fifth and sixth abdomenal segments.

At first, immature females are characterized by an oblong
triangular abdomen that covers a similar proportion of the
sternum through each molt. Finally, the female undergoes a molt
resulting in the maximum increase in width of the abdomen and at
this point, for our purposes, was considered sexually mature,
though ovaries of immature and newly molted mature crabs were not
examined. Females in captivity reaching this stage of
development were never observed to molt again.

Assuming that female M*. spinosissimus is characterized by a
terminal molt of puberty (see discussion Biddlecomb, et al . , this

volume) , it can be concluded that an ovigerous crab is mature and
will not molt again. In terms of the allometric growth equation
(y = Bxa) , as more typically applied to various dimensions of the
male chelae (Hartnoll, 1974), the female abdomen compared to
carapace length is positively allometric (a > 1.0) throughout its
juvenile development (beginning at 25 mm CL) and becomes strongly
positive at the maturation molt (Figure 1) .

Males maintain a uniform abdomen shape throughout their molt
history. The allometry of the abdomen in relation to the
carapace length is nearly isometric (Figure 2) . However, in
contrast to females, maturing male crabs are characterized by a
significant increase in size and shape of the chelipeds. When
comparing log carapace length (CL) to log ventral propodus length
( VPL) of Antiguan crabs, the males' VPL, as well as the cheliped
in general, gradually begins to become proportionately larger
than the females' at approximately 80 mm CL. The VPL then
becomes markedly larger at a subsequently increased rate at
approximately 105 mm CL (see Biddlecomb, et al . , this volume) .
This first gradual inflection is likely the beginning of the size
range for the pre-pubertal instar of this species. The pre¬
pubertal instar appears to range in CL between 80 mm CL and 115
mm CL. The second inflection, assumed to represent male crabs
having undergone the maturation molt, begins as low as 102 mm CL
and ends for the Antiguan population at 147 mm CL (Refer to
Figure 2, Biddlecomb, et al. , this volume). These appendages

were observed to aid in agonistic displays and may also aid in
courtship communication and defense. Biddlecomb, et a .1 . (this
volume) discuss the possibilty of a single post pubertal molt in
M. spinosissimus .

Natural population characteristics categorized by sex, size
class, egg bearing state and state of maturity for each of the
sites are presented in Table 1. The mean size (CL) of male and
female crabs, designated mature, are different at all sites (see
Iglehart, et al . , this volume) . Mature male and female crabs
were largest from site 3 (Grand Turk, BWI) while those from site
5 (Florida) were smallest. The largest population samples were
from the Dominican Republic and Antigua, with 56 and 57 males and
126 and 44 females, respectively. Table 2 and Figure 3 show the
size frequency distributions of female crabs caught from the five
study sites.

Gravid crabs were caught every month of the year both in
traps and by SCUBA divers, indicating reproductive capability
throughout the year. The mean percentage of mature ovigerous
females caught in traps at the Dominican Republic was 81.3%

(S.D.= 12.4%) for February through November (no traps were set in
December and January), with a mean of 11.5% (S . D. - 5.8%) mature
females caught each month. At the Antigua site 65.7% (S.D.=
17.4%) of crbas trapped were ovigerous with an average of 4.8%
(S.D.= 3.4%) ovigerous females per month. No crabs were

collected during the months of January, May and November (Figure
4) .

Mating

Mating of captive Mithrax spinosissimus while in aquaria was
observed on four separate occasions. In each instance, both
female and male were hard shelled and the females were barren.

The crabs faced each other ventrally, the male dorsal side down
and underneath the female. The female's abdomen was opened and
outside the carapace of the male, whereas the male's abdomen was
only slightly parted from his sternum and the first pleopods were
inserted into the gonopores of the female. The female used her
legs for support, while the male held the female by interlocking
legs and/or by clasping her appendages with his chelae. Crabs
were observed to remain in this mating position for less than one
hour. The observed matings occurred in January and February.

Reproductive Morphology

The configuration of the ovaries in female Mithrax
spinosissimus differs from that of the Brachyuran crabs as
described in Warner (1977) . Ovaries are paired and anastomose
directly posterior to the cardiac stomach. They have lobes
leading both anteriorly on each side of the stomach to the
frontal margin under the eyes and posteriorly underneath the
pyloric stomach, as in other crabs (Figure 5) .

Each of two gonopores located on the sternites of the sixth
thoracic segment and covered with pliable chitinous flaps, opens
to an approximately one cm thin chitinous tube, which ends at the
junction of each ovary and each of the paired spermathecae. This
configuration, like that in other Majids, is "concave" as
described by Hartnoll (1968) , with the spermathecae projecting
above the ovary-gonopore tube junction, rather than being between
the ovary and gonopore tube (Figure 5) . From this junction, a
section of each ovary runs dorsally to each of the main paired
bilobal sections. The spermathecae are barrel shaped with
dimensions of approximately 15 X 10 mm in a crab of 106 mm CL.

The spermathecae consists of two parts. The dorsal half is
composed of two white waxy bodies pressed together, possibly
sperm plugs (Hartnoll, 1969), while the ventral half, at the
ovary-gonopore tube junction, consists of a more viscous
brown/white liquid.

Of the 28 mature female crabs dissected, all had pale orange
unfertilized eggs throughout the lobes of the ovaries and all had
spermathecae with the anatomy described above, regardless of
whether fertile eggs were being actively brooded at the time. In
one dissected female (79.4 mm CL) judged to be one or two molts
from maturity, the gonopores were barely discernible (pinholes) .
Chitinous tubes leading to the ovaries from the gonopores did not
exist and ovaries and spermathecae were not obvious, as they are
in mature females. Protrusions on the fifth thoracic sternite

fit into sockets in the sixth abdominal segment in this immature
crab. Pleopodal setae were only slightly developed and sparse as
compared to those on mature females.

The eight pleopods of mature females have very setose
exopods that cover the area between the abdomen and sternum when
the abdomen expands as the brood develops. The setae on the
endopodite of each pleopod are stouter, spirally organized and
more sparse than those on the exopods. There are approximately
100 sets of three to six setae on each of eight endopodites.

Each encased fertile egg has an individual strand that runs along
a portion of each clumped set of setae and then twists around the
set at the junction of the strand and egg. The strands of a
clump of eggs may also twist around each other before separating
and running along the set of setae (Figure 6) .

Male reproductive structure was examined in less detail. In
the largest mature males, spermataphores which appear as white
waxy bodies (closely resembling those found in the spermathecae
of the female) were observed in each of the paired sperm ducts,
probably indicating the presence of mature sperm. The sperm
ducts of smaller "mature" males and those of immature males were
not found.

Clutch Size

Clutch size in ft spinosissimus was determined using dry and
wet weights and weights of samples of individual eggs from those

masses. The mean dry weight of five replicate samples of 20 eggs
from each of six females was 4.7 X 10~3 gms (S.E.= 0.07 X 10“3
gms; range = 4.3 - 5.1 X 10“3 gms; n=30) , which yields a mean dry
weight of 0.23 X 1 0 “ 3 gm/egg. The mean wet weight of these
samples was 19.7 X 10”3 gms (S.E. = 0.09 X 10”3 gms? n~ 30).

Based on this estimate, each dry-weight gram of Mithrax
spinosissimus egg mass contains about 4348 eggs, assuming that
the egg clutch is filled with eggs of uniform size and weight.
However, the egg mass ordinarily contains some interfollicular
connective tissue, as well as a very small portion of eggs at
various stages of embryogenesis . The wet weight of the above
samples is 4.2 times the dry weight.

Twenty ovigerous crabs were weighed pre and post hatch. The
difference in weight was taken to be the total egg mass wet
weight. Using the calculated wet weight to dry weight ratio and
the egg number per dry weight gram, the total egg number per wet
brood was estimated. Mean clutch size was 6.0639 X 104 eggs for
crabs with a mean CL of 125.5 mm (SE = 3025; range — 55,249 -
66,026? N=20) . The linear regression of the crab carapace length
to derived clutch size was determined (r = 0.65? p< .05? N=20)
and plotted (Figure 7) .

Counting the eggs in four broods, using the two methods
described, from four females of differing carapace lengths
revealed a mean total of 7.1446 X 104 eggs per brood. From the

wet weights of these four broods and the wet weight to dry weight
ratio, the dry weight of each brood was estimated and the number
of eggs expected from the dry weights calculated using 4348 eggs/
g (dry wt.)* the predicted and actual brood number varied less
than 1000 eggs in each of the four cases.

Embryogenesis

Unfertilized eggs in the ovaries are usually pale orange.
After spawning the fertilized eggs (about one mm in diameter) are
attached to the pleopods and are orange in the early stages of
development. Subsequently, they change to orange-red, red,
amber-red and at the day of hatching, they change from clear-
amber to a translucent opaque white color. From a sample of four
female crabs, where the exact dates of spawning and subsequent
larval release are known, incubation time was 29.5 days (S.D.=
0.5).

Minor differences in egg color were observed depending on the
collection site of crabs. The "red" eggs from Antiguan crabs are
actually dark red or raspberry color, while the "red" in eggs
from Dominican Republic and South Caicos crabs are light red or
strawberry color.

The development of the eggs up to hatching has been divided
into the five most distinctly observed stages from visual and
microscopic examination of the eggs of seven female crabs
(Figures 8, 9).

Stage I

At the time of spawning, the yolk is pale to bright orange.
Over the next six to nine days, the yolk divides into oblong
cells and the color gradually changes to cil> IT* d C* lor (Figure 8a) .

Stage II

After 10-12 days, the yolk, 95% of the egg, is prominent and
composed of large oblong red cells with small rounder red cells
closest to the developing larva. The larva appears as a small
amber-clear segmented line. As the egg develops up to hatching,
the yolk cells become smaller but they do not completely
disappear. The eye spots are not yet visible. The heartbeat is
not present at the beginning of this stage (Figure 8b) .

Stage III

By the 15-18th day, the eyes (not yet full ovals as in later
stages) , thorax, abdomen, telson, appendage buds and beating
heart are obvious. The yolk, now comprising 50% of the egg,
appears microscopically as four contiguous lobes around the
cardiac region and, in reference to the larval portion, appears
visually as a dorsolaterally positioned solid crescent shape. A
few black, orange or yellow chromatophores around the cardiac
region and along the abdomen are also visible (Figures 8c-8g) .

Stage IV

From 19-27 days, the yolk still appears dorsolateral ly, as a
solid crescent and coloration continues to disappear gradually as
the larva develops up to and through this stage, so that the
colored yolk encompasses only 25-30% of the egg. Yolk cells
without color are noticeably smaller than those with color. The
now fully oval eyespots are pink-red and the thoracic appendages
become more developed, though still not extended (as in the
hatched zoea) . To the unaided eye, the larval portion is clear-
amber with both a few very small chromatophores and the eyespots
visible (Figure 9a-9c) .

Stage V

At 26 -27 days, three to four days prior to hatching, the
yolk color separates at the dorsal midline into distinct spheres
on each side of the cardiac region. At the beginning of this
stage the yolk occupies 10-15% of the egg. The thoracic
appendages are clearly visible and periodically beat rapidly.
Haemolymph can be seen rushing with each heartbeat around the
inside of the larva. The cardiac region has been seen in healthy
eggs to beat as fast as 250 beats/minute. The larva portion is
still amber-clear to the unaided eye. Visually, the eyespots are
pink and can be confused with the colored portion of the
remaining yolk until the day of hatching, when little or no yolk
color is visible in the morning and no yolk color is visible by

dusk. Occasionally, healthy larvae strongly flex their abdomens
during the few days prior to hatching. At the day of hatching,
the egg cases change subtly from clear-amber to translucent
opaque white (Figure 9d-9k) .

Hatching

Nocturnal hatching was the norm, usually beginning at dusk or
early evening. Several hatches occurred in the afternoon but
rarely as early as 1400 hours. Eggs removed and put into a
beaker of seawater during the day prior to the night of hatching
would hatch only after dusk. The occurrence of hatching in
relation to its proximity to the new or full moon was examined
with no apparent pattern observed. For all sites combined,
hatches occurred on any given day of the year regardless of the
time of the month.

Throughout incubation, females were occasionally observed to
lower their abdomen and pleopods as one unit and then
successively raise each layer of pleopods (four layers). They
would start with the innermost layer, followed by the next layer,
until the abdominal flap itself returned to its original closed
position, at which point they would then repeat the pumping
action. At hatching, the abdomen and pleopods are pumped
continuously in the same manner to expel the prezoea. In
addition, those larvae that are able to do so, swim out of the
brood space. Occasionally, one or both of the chelae were

observed to be briefly placed into the brood space. Complete
hatching time varies from two hours to as long as 36 hours, but
usually was complete for viable hatches between 12-16 hours.

Results of the occurrence of captive crab hatches are
summarized in Table 3. Thirty-eight crabs held in captivity for
two months averaged 1.03 hatches each and four crabs held in
captivity 10 months averaged 3.75 hatches each.

Of a sample of 26 females collected in Antigua, 70% were
gravid when caught. Of the 30% not gravid when caught, 87%
became gravid within 30 days. Forty-seven percent of crabs
caught gravid or becoming gravid within 30 days (12 crabs) had
three or more hatches while in captivity. Of those 47%, 70% died
before having a fourth hatch. One female produced seven
successful hatches over the course of a year in captivity.

The mean interhatch interval for 37 individual females
studied was 61.9 days (S.D. ™ 19.6, n = 247). Samples not
included were those where hatches failed due to the brooding
females death prior to hatching or that were aborted. The
shortest interval was 33 days, only three days from the hatch to
the extrusion of her subsequent brood. Another female had a new
brood the day after hatching but the new eggs were not attached
and washed out easily when the crab was handled. The longest
interval of time between release of an egg brood and the spawning
of a new batch of eggs was 127 days. Sixty percent of the
intervals ranged between 50-70 days.

DISCUSSION

A noticeable change occurs in the relationship between male
ventral propodus length (VPL) and carapace length (CL) at the
pre-pubertal instar. Most importantly, chelae begin to enlarge
prior to the pre-pubertal instar and continue to enlarge
significantly at an increased rate up to the maturation molt.

This indicates a single maturation molt to adult morphology
occurs in male Mithrax spinosissimus . although the increase in
cheliped size during the pre-pubertal instar suggests sexual
activity prior to the maturation molt. Brownell, et ad., (1977),
mentions an immature crab being sexually active prior to molting
to maturity and Hartnoll (1963) presents evidence from his
observations that males of different Majid species, including
Mithrax sculptus . have developed some mature sperm prior to the
maturation molt. Since there was no data collected on the
internal morphology of pre-pubertal instars and mature male crabs,
we have no evidence other than external characteristics to
suggest there may also be sperm maturation prior to the
maturation molt. Future reproductive studies should focus on the
necessity for collecting information on male internal anatomy at
a variety of morphological stages.

We have little evidence to indicate that females mature
during the pre-pubertal range, since external changes are not
evident and specific examination of eggs from immature females

was not conducted. An examination of an immature female of 79.4
mm CL showed no obvious internal reproductive structures. It is
possible, however, that the duct between the gonopore and
spermathecae-ovary junction was soft and so separated from the
carapace upon examination. This duct may become chitinous only
at the maturation molt. This limited evidence suggests females
do not sexually mature internally and do not obviously mature
externally until the maturation molt.

We have never observed a female to molt once the abdomen
reached maximum width or if she was ever gravid. When dissected,
mature females always had eggs at some stage of development in
their ovaries and may direct their energy into egg production
rather than to prepare for further ecdysis and subsequent growth.
This suggests females have a terminal molt to maturity. With
males, however, the evidence is not as clear. We have observed
only one apparently mature male crab greater than 120 mm CL to
molt (131.0 mm CL to 153.3 mm CL); the individual was collected
in the Dominican Republic where unusually large immature male, as
well as female crabs, were recorded. Though the percent change
of molt increments generally decreases after 80 mm CL (see
Biddlecomb et ad. , this volume) , there is one example of a male
crab molting from 89.0 mm CL to 125.0 mm CL, which is an increase
of 43%. Hartnoll (1963) also presents evidence that the Majids
which generally have a terminal molt, often have a wide range
between the smallest and the largest mature crabs.

On the other hand, as discussed in detail in Biddlecomb, et
al. (this volume) , the percentage size increase in CL of recorded
moltings constantly drops from a mean of over 35% to a mean of
23% or less at the maturation molt. Considering the
discontinuity in data at the maturation molt and the slight
decrease in VPL allometry of the largest adult males, a second
molt after the maturation molt may be indicated. This may not
occur for a year or more and may not occur in all surviving
males. Further investigation is necessary to support this
theory.

Ninety-five percent of mature females collected in Antigua
produced eggs while in captivity. Forty-seven percent of a
sample of 26 crabs collected from Antigua had three or more
hatches. All dissected females brooding fertilized eggs were
also developing another batch of eggs in their ovaries, thus if
caught gravid, we could expect at least two hatches. Moreover,
the occurrence of three or more hatches while captive suggests
that adequate nutritional material was being made available using
the standard algal turf feeding process, supplemented
occasionally with various macroalgae. The occurrence of one crab
having seven broods in captivity appears rather exceptional. The
relationship between the frequency and the number of hatches per
crab is shown in Figure 7. Four to five hatches are the normal
maximum to be expected from a mature female Mithrax

spinosissimus .

The broods of crabs gravid or becoming gravid soon after
collection were large and healthy. In some cases, the ensuing
broods were smaller while in others, a greater number of
undeveloped eggs were noticed. Whether this suggests decreasing
or deteriorating internal sperm stores or deteriorating culture
conditions is not known. Though we did not record pre-and post¬
hatching weights consistently for females, general observations
of relative brood size for captive crabs appeared to be
consistent for successive broods, except for the last brood
before the death of the crab where the last brood tended to be
unusually small. Since a reduction in fecundity with time is
not apparent in captivity, it would appear that the conditions of
cage life, including feeding quantity and quality are adequate
for a mature crab producing approximately 5-10% of her body
weight in eggs.

The broods of several females kept in closed aquaria for two
to four weeks became infested with filamentous epiphytes,
microscopic isopod-like animals running around the egg surface,
as well as various Ciliophorans . If infested early during
incubation, most or all eggs in a brood would die and appear
opaque-brown. One brood infested approximately nine days prior
to the expected time of larval hatching appeared quite viable but
died by the expected hatch date. However, one brood with
infestation obvious at only four to six days prior to hatching,

survived well. Broods of females kept in open water cages never
became infested. The few broods (3) that died while in open
cages occurred in ovigerous females collected by fishermen and
had been inappropriately transported, i.e., out of water or lying
in bilge water. These broods appeared opaque grey-white, and
subsequently decayed in the abdomen space, turning pleopods
black. Subsequent broods from these females were not affected.

In the Dominican Republic, a mean of 81.3% of mature females
caught in traps over a 10 month period were ovigerous. If this
percentage is representative of the wild population, it suggests
that mature females are consistently reproducing throughout the
year, i.e., seasonality does not affect reproduction and females
constantly reproduce throughout their adult life. If the percent
becoming gravid within 30 days, as determined from a sample of
Antiguan crabs, is applied to the Dominican Republic population,
then we might expect an additional 16% of the total collected
females to become gravid within 30 days. This large number of
wild gravid crabs or crabs becoming gravid shortly after
collection (97%) strongly suggests that the consistent egg
production rates observed in captivity could represent a natural
characteristic of the species.

The mean interval between brood release and spawning of the
next brood was 61.9 days with a standard deviation of 19.6 days.
The large variation in this interval is due to a number of

factors. Of 14 female crabs tranported to Carriacou in aquaria
with constant sea water exchange, seven had intervals greater
than 80 days between the initial hatch and their next hatch. The
mean intervals for these is 105.0 days (S.D. = 20.4). Of the
remaining seven, five had a mean interval of 61.4 days (S.D. =
7.4) for that particular interval. The last two females died
before their next hatch after transport. These rare and
unusually long intervals, and the death of two of the transported
females before successfully hatching, suggests stress due to the
process of transportation. The hatches of those crabs surviving
the pre- and post- transportation interval were successful. The
mean for all intervals recalculated without those seven
individuals is 58.4 days (S.D. = 14.2).

Of the 37 females examined to determine intervals, a
specific group of 11 females, having more than two hatches
consistently produced hatches whose individual intervals varied
within by a range of 10 days. For example, one female's
intervals were 53, 62 and 59 days. This consistency suggests
that a fertile female, healthy and well fed, will produce eggs at
a specific biologically determined interval. The variance in
intervals among 11 females examined was much greater than the
variance within each female (mean = 51.5 days; S . D . = 5.1, n =

22). That is, while one female's intervals may vary consistently
with a mean of 50 days, another female's may vary at 60 or 40
days .

No relationship between lunar cycle and hatch times was
noted, though the mean incubation time is 29.5 days and the mean
between brood interval is 61.9 days, suggesting that hatching
could potentially correlate with the lunar cycle. On the other
hand, the large variation in time interval, if representative of
the species in the wild, suggests no correlation to any definite
cycle. Since the ovaries of both brooding and non-brooding
females were always full of eggs, presumably developing, the time
intervals may solely be due to the genetically determined
unspawned egg maturation time, specific to each female. However,
some variation in the maturation time is probably due to
energetic and/or nutritional constraints. If food supply is
limited, it may affect yolk deposition into the unspawned eggs.

The few times mating was observed, both male and female
Mithrax spinosissimus were in the hard shell state with the
female dorsal to the male. In adult Mithrax spinosissimus the
gonopores are covered by a moveable flap and it is likely that
copulation can occur at any time after maturation. The similar
shape and texture of the white, waxy bodies found in the
spermathecae of the female (two in each) and in the male
reproductive tract (one in each) suggest that spermatophores are
generally present in the spermathecae of mature females.

It appears that egg bearing capacity increases with crab
size as measured by carapace length. However, the variance in

the measured crabs was quite high and is attributed to the
difference in egg mass weight due to the state of embryonic
development, the number of spawnings since mating and the
condition of the female.

The pre-hatch larvae or embryo changes drastically during
the second week of development from what appears to be a mass of
yolk to a distinct larva. The percentage of yolk color present
is a characteristic that can be used to determine the approximate
time to hatching, but only during the last week can the hatch
date be predicted within a day. When the yolk appears to split
into two colored spheres, it is usually three days before
hatching. Crabs from Antigua had darker red eggs, suggesting
phenotypic variation between populations. Darker eggs make
determination of the hatch date considerably easier.

Eggs observed in a beaker during the day prior to a night of
hatching, hatched at dusk. This suggests that hatching is
controlled at least partially by the larvae. Also, the subtle
change in egg case color on that same day could indicate internal
chemical changes initiated by the larvae. The strong flexing of
larvae prior to hatch suggests they help to liberate themselves
from their cases. However, synchronization of the liberation of
entire broods within two to twelve hours may indicate at least
some internal biological control or biological rhythm by the
female as to the hatch date and duration.

CONCLUSIONS

Aspects of the reproductive biology of Mithrax spinosissimus
have been examined and the results reported. As with some other
Majids, this species appears to have a prepubertal instar just
prior to the maturation molt at least in males. Females appear
to cease molting after the maturation molt. Our evidence is not
conclusive for males? some indications for a second molt exist
and are discussed in Biddlecomb et al . (this volume) . Limited
evidence suggests the male may begin to mature prior to the
maturation molt. At the maturation molt the male’s chelae
significantly increase in size and the female's abdomen increases
in width to cover the sternum.

The reproductive morphology and mating behavior is similar
to other described Majidae. A female can store sperm in
spermathecae and potentially fertilize many and perhaps all
spawnings during its reproductive life from one mating. As well,
it appears capable of multiple matings once mature.

Like other tropical Majids, this species was found to be
ovigerous throughout the year with no obvious seasonality and can
reproduce continuously, approximately every 62 days, throughout
its mature life. The maximum potential reproductive life of the
female appears to be one to one and a half years after the
maturation molt. Incubation of eggs attached to the pleopods

requires approximately 30 days to mature and hatch.

From the pattern of embryogenesis , exact hatch dates can not
be determined until the last week of incubation. Based on
morphological evidence presented in this paper, an estimate can
be made at that time to within a day. Hatching in the sea, in
cages, can be very reliable with no indication of disease
problems .

More comprehensive data are needed, such as, 1) examination
of the internal male reproductive anatomy at a variety of
morphological stages; 2) positive determination of a second molt
after the maturation molt in male Mithrax spinosissimus ; 3) the
weights of successive hatches from newly mature to death to
indicate reproductive potential and effects of captivity; 4) the
weights of females themselves through time in captivity; 5) an
accurate determination of numbers of hatches from a fertilized
but isolated female; 6) a determination of the age of mature
crabs at the time of collection to indicate reproductive life
span; and 7) determinations of any variations in fecundity with
changes in environmental and geographical parameters.

From the present study, we can conclude that Mithrax
spinosissimus appears relatively unaffected by conditions of
captivity, with a large percentage of mature females consistently
producing viable broods. The fecundity of this species may
actually be enhanced in an open water mariculture facility by
stable feeding conditions, protection of larvae from predators.

and a thorough understanding of the reproductive biology and
embryonic development.

LITERATURE CITED

Anderson, D.T. , 1982. Embryology. In The biology of

Crustacea: Embryology, morphology and genetics, Vol 2
Ed. Abele, L.G. Academic Press 1982.

Bohnsack, J. L. , 1976. The spider crab, Mithrax

spinosissimus : an investigation including commercial
aspects. Florida scientist. V 39, (4) pp 259-266.

Brownell, W.N. and Provenzano, A.J., 1977. Culture of the

West Indian Spider crab (Mithrax soinosissimus) at
Los Roques, Venezuela. J. of World Mar. Soc.

Edwards E. , 1978. The edible crab and its fishery in British
Waters. Fishing News Books LTD. Surrey. 142 pp.

Gifford, C.A. , 1962. Some observations on the general biology
of the crab, Cardisoma quanhumi (Latreille) , in South
Florida Biol. Bull. Mar. biol. Lab., Woods Hole 123:
207-23 .

Hartnoll, R.G., 1963. The biology of Manx spider crabs.

Proc. zool. Soc. Lond. 141:423 -96.

Hartnoll, R.G., 1965a. Notes on the marine grapsid crabs of

Jamaica. Proc. Linn. Soc. Lond. 176: 113-47

Hartnoll, R.G., 1965b. The biology of spider crabs: a
comparison of British and Jamaican species.

Crustaceana 9: 1-16.

Hartnoll, R.G., 1968a. Morphology of the genital ducts in
female crabs. J. Linn. Soc., (Zool.) 47: 279-300.

Hartnoll, R.G., 1968b. Reproduction in the burrowing crabs,
Corvstes cassivelaunus (Pennant 1777) (Decapoda,
Brachyura) Crustaceana 15: 165-170.

Hartnoll, R.G. , 1969. Mating in Brachyura. Crustaceana 16:
161-181.

Hartnoll, R. G . , 1974. Variation in growth pattern between

some secondary sexual characteristics in crabs ( Decapoda ,
Brachyura) . Crustaceana 27: 131-136.

Ingle, R.W. , 1983. Shallow Water Crabs - Synopses of the
British Fauna (new series). Kermack, D.M. and R.S.K.

Barnes (eds . ) . Cambridge University Press. Cambridge.

206 pp.

Munro, J.L., 1974. The biology, ecology, exploitation and

management of Caribbean Reef Fishes. Part V. The
biology, ecology, and bionomics of Caribbean Reef
fishes: Crustaceans (spiny lobsters and crabs) Res
Rept. Zool. Dept. Univ. West Indies 3: 39 -48.

Rathbun, M. J. , 1925. The spider crabs of America. United
States National Museum Bulletin 129, 613 pp.

Teissier, G. , 1935. Croissance des variants sexuelles chez
Mai a scminado. Trav. Sta. biol. Roscoff, 13: 99-130.

Van Engel, W.A. , 1958. The blue crab and its fishery in the
Chesapeake Bay. Part I - reproduction, early
development, growth and migration. U.S. Fish. Wildlife
Serv. Vol. 20, 17 pp.

Vernet- Cornubert, G. 1958. Biologie general de Pisa

tetraodon (Pennant). Bull. Inst, oceanogr. Monaco 1113,
1- 52.

Warner, G.F., 1967. The life history of the mangrove tree
crab Aratus pisoni . J. Zool. Lond. 153: 321-335.

Warner, G.F., 1977. The biology of crabs. Van Nostrand
Reinhold Co. London.

LEGENDS

FIGURE 1.

Adult female ventral and posterior view
indicating abdomen size difference and appearance
of the inside of the abdomen with eggs on the pleopods

FIGURE 2.

Adult male ventral and posterior view.

FIGURE 3.

Histogram plot of number of ovigerous females
captured pooled from all sites.

FIGURE 4.

Percent of total females trapped from the North
coast of the Dominican Republic (site 5) and the
Northeast coast of Antigua (site 7) for the year
beginning October 1985.

FIGURE 5.

Illustration of the paired ovaries, and the gonopore
tube-spermathecae-ovary complex. The outlines of
the carapace and sternum are of the dorsal view.

FIGURE 6.

Examples of incubating eggs attached to the
clumped setae of the endopodites of the pleopods.

FIGURE 7.

Relationship between the carapace length of It
spinosissimus and potential clutch size in
thousands of eggs. Based on size specific dry
weight of total egg mass and mean dry weight of
individual eggs (0.0047 gm/20eggs : S.E. =

0 . 0007 gm; n = 20) .

FIGURE 8.

Stages I - III of incubating eggs.

FIGURE 9.

Stages IV and V of incubating eggs.

TABLE 1.

The mean carapace length of mature males and
females by sites and source of collection.

Carapace width (CW) is approximately 1.05 carapace
length (CL) in adult Caribbean populations.

TABLE 2.

The frequencies of ovigerous females captured by
study site and 10 mm carapace length
increments .

TABLE 3.

The average number of hatches per crab for
specific time intervals of captivity. Data for (N)
crabs is from 49 total females observed.

■

Figure

Figure 2

CARAPACE LENGTH IN mm

Figure 3

ino Sdvai on

ino sdvai on

Zco
< 00
~3 ~

ino Sdvai on

ino sdvai on

OCT

*85

r r — n — — r" i i i — i i r

00000000007,

OOJOOl'-CDlO^tCOCM1^®

aiAvao onv aaddva± ®

S31VW33 HV101 30 1%) ±N30U3d f

■*-»

Figure 4

Gonopore

Spermatheca

Spermatheca
Gonopore Tube

Sternum

Gonopore

Figure 5

Figure 6

CLUTCH SIZE IN THOUSANDS OF EGGS

CARAPACE LENGTH IN mm

Figure 7

Figure 8

//a

Figure 9

TABLE 1: The mean carapace length of mature males and females by
sites and source of collection. CW = 1.05 CL for all
adults except from Florida.

MEAN SIZE OF CRABS (MM CL)

LOCATION

ALL ADULT
(FEMALES)

GRAVID

(MATURE

FEMALES)

MATURE

(MALES)

SOURCE

TOTAL

POP.

SAMPLE

DOMINICAN

REPUBLIC

122.7
+/- 0.8

N— 12 6

122.5
+/- 0.9

N— 97

140.8

+/-1-4

N=56

THIS

STUDY

203

ANTIGUA

107.8
+/“ 1.3
N“44

108.6
+/“ 1.6

N— 27

131.2
+/- 1.2

N"57

THIS

STUDY

162

GRAND TURK

140.2

+/“ 2.5

N™ 19

137.5
+/- 3.7
N—10

146.7
+/- 6.1
N—15

THIS

STUDY

SOUTH CAICOS

113.7

+/- 1.9

N-27

113.7
+/“ 3.2

N— 12

135

+/“ 3.7
N=10

THIS

STUDY

FLORIDA

81.7

+/“ 3.6

N~3 3

83 . 5
+/-5.1

N— 18

93.9
+/“ 4.6
N=ll

THIS

STUDY

JAMAICA

125.3 (CW)
N=7 1

122.8 (CW)
N=3 5

133 .4 (CW)
N=14 1

MUNRO

(1974)

212

FLORIDA

86.0 (CW)
N=73

96.0 (CW)
N— 28

BOHNS ACK 103
(1976)

TABLE 2 : The frequencies of females captured ovigerous by study
site and 10 mm. carapace length increments.

SIZE CLASS OF OVIGEROUS FEMALES BY STUDY SITE

SIZE

CLASS

CL (MM)

GRAND

TURK

SOUTH

CAICOS

DOMINICAN

REPUBLIC

FLORIDA

ANTIGUA

TOTALS

60-69

70-79

80-89

90-99

100-109

110-119

120-129

130-139

140-149

150-159

TOTALS

TABLE 3: The average number of hatches per crab for specific time
intervals of captivity. Data for (N) crabs is from 49
total females observed.

NUMBER OF HATCHES FROM CAPTIVE FEMALE CRABS

MONTHS

AVG # OF

NUMBER OF

% OF

TOTAL

HATCHES/

CRABS

TOTAL

HATCHES

CAPTIVITY

CRAB

(N)

0.6

100

1.03

77. 5

1.85

67.3

2.70

42.8

3.23

26.5

3.75

8.2

2.0

SECTION II: BIOLOGY AND ECOLOGY OF MITHRAX SPINOSISSIMUS

POST-LARVAL GROWTH OF CULTURED M. SPINOSISSIMUS

POST-LARVAL GROWTH AND SURVIVORSHIP OF CULTURED
MITHRAX SPINOSISSIMUS

K.L. Porter, J.M. Iglehart , R.V. Ruark, M. Craig, A. Biddlecomb,

and W.H. Adey

Abstract

Mithrax spinosissimus rearing studies were conducted to
develop simple and inexpensive hatching and growout techniques.
Crabs were hatched in 0.25 cubic meter cages enclosed by 0.5 mm
plastic mesh. Post larval crabs were transferred to larger cages
upon reaching carapace lengths of 10 mm and again at 20-25 mm.
Initial post-larval first crab (Instar I) densities are estimated
at 2000/0.25 cubic meter cage.

Fed on an experimental diet of primarily cultured algal
turfs, a total of 56 hatched egg broods were reared through 100
days post hatch or longer. Crab growth ranged from 0.11-0.19
mm/ day at 60 days post-hatch to 0.10-0.21 mm/day at 100 days
post-hatch and to over 0.5 mm/day at 280 days post-hatch. Crab
survivorship averaged 22% to 60 days, 23.6% from 60-120 days and
18.5 % from 120 to 300 days.

Growth rate and feeding data strongly indicate that under
optimum environmental conditions in cages, growth rates are high
and molt stages largely pre-determined. The relatively high
mortalities in this study may result in part from under-feeding,
but more likely result from a complexity of factors including
predation by intruders in the cage environment, wave damage
during molting and aggressive behavior by some individuals.

Proper cage construction and management can likely greatly reduce
juvenile mortality.

INTRODUCTION

The tropical western Atlantic crab Mithrax spinosissimus is
not extensively exploited by commercial fishermen, due in large
part to its scattered distribution and relatively low population
levels (Munro, 1974), though it forms an important gourmet
restaurarnt food in some Caribbean countries. On the other hand,

its brief larval life (90-140 hours to first crab, (Provenzano
and Brownell, 1977; Porter, et al. , in review), herbivorous
lifestyle (Coen, 1987; Colin, 1978), high fecundity, and
potential economic value make it an ideal choice for mariculture
(Brownell et al. , 1977) .

Little has been known of the growth rate of spinosissimus
either in the wild or in captivity . Laboratory rearing studies
have been limited to describing the morphology of the larval
stages (Provenzano and Brownell, 1977). Brownell, et al . , (1977)

conducted a single experiment of IL spinosissimus juvenile crab
culture in a fine meshed (0.363 mm) open water cage which
resulted in rates of growth of 0.06 mm/day CL. The authors
suggested conditions in the cage adversely affected the juvenile
crabs and depressed their rate of growth. The attached algal
microflora on the screen mesh surface served as the sole food
supply. A single crab was grown at a rate of 0.11 mm/day CL for
over 175 days in the laboratory (Brownell, et al. , 1977).

In a field study of crabs grown in unattended screened cages
for 60 days (similar to the methods of Brownell, et al. , 1977)
Porter, et al. (in review) found crab growth (at a mean of 0.11
mm/day CL) to be much greater than in the earlier work by
Brownell. Iglehart, et al. (1987) found that crab growth under
conditions of supplemental feeding of cultivated algal turfs in
cages increased crab growth rates to 0.17 - 0.20 mm/day CL for
100 day cultures.

The objective of our work was to develop baseline growth
information that would allow the development of simple and
inexpensive rearing techniques for post-larval crabs. It was
hoped that such techniques could be used by itinerant fishermen
in under-developed Caribbean nations. The effects of crab
density, diet and rearing systems on the growth and survival of
juveniles of the Caribbean King Crab, It spinas issimus in open
water cage culture are described.

MATERIALS AND METHODS

Crabs

In this investigation, crab culture was carried out, "in
situ", in cages in tropical Caribbean coastal lagoons at ambient
water conditions. The field study sites, occupied at various
times from 1984-1986, were 1) Buen Hombre, northwestern Dominican
Republic (19° 80' N, 71° 20* W) ; 2) Nonsuch Bay, eastern Antigua
(17° 10' N, 61° 43' W) ; and 3) Carriacou, Grenada (12° 60'N, 61°

4 0 ' W ) (Figure 1). In this investigation, 154 ovigerous M.
spinosissimus were collected from the coastal waters of the study
sites. Gravid crabs nearing hatch release were placed in
separate cages. The procedures for managing breed stock and egg
development in the ovigerous female are discussed in detail by
Craig, et al . (this volume) . The ovigerous female crab was
removed from the cage upon completion of the hatching process and

the larvae were left undisturbed until they settled out as post
larval instar I or first crabs. Survival of larvae from egg to
the first post-larval crab stage has averaged 3.0% in earlier
studies (Iglehart, et al. , in manuscript). The rate of survival
is based on the mean brood size of female crabs (6.0639 x 104
eggs; S.E. + 3025: Craig, et al. , this volume) and the resulting
survival rate as found in hatch cages similar to those used in
this study. Therefore, initial post-larval crab densities in the
hatch cages in this study are assumed to be approximately 2000
crabs/cage (Porter et al. , in review) . Crabs were reared in
cages throughout the experiment. All survivorships are
calculated from first crab or from the stated interval.
Experimental food

Algal turfs cultured upon rigid, black plastic screens (0.61
x 0.92 m; 2 x 3 mm mesh) (Figure 2) were introduced as feed to
the post-larval crabs. Algal screens were exchanged every two to
six days to ensure an adeguate supply for crab consumption.
Baseline data for algal turf community structure, methodology and
productivity are discussed in Adey and Hackney (this volume) .

Rearing systems

The rearing system cages were constructed of wood frames
enclosed by plastic screening of various mesh sizes. Polyester
fiberglass resin was used to coat the wood frames before
assembly. The hatch cage (1.0 m Lx .35mWx .70 m H) (Figure

3) was covered by white, 0.5 mm polyester monofilament screen.
Cage dimensions varied among the field sites from 0.3 m3-0 . 4 m3
(Tables 1, 3) . The cages were anchored and suspended in
protected lagoonal waters at 2.5-4 m depths so as to reduce or
eliminate any surface wave action from affecting the cage. The
intermediate cage (1.0 m L x 1.0 m W x 1.3 m H) was identical in
construction to the hatch cage but covered with 1-1.5 mm mesh
white polyester monofilament screen and anchored near the hatch
cages in reef lagoons at depths of 1-2 meters from the bottom.

The growout cage (2.6 m L x 1.0 m W x 1.0 m H) (Figure 4) was 3.5
m covered with 1/4" black polypropylene plastic mesh and
anchored so as to float at the lagoon surface. A more complete
discussion of cage design, construction, placement and mooring is
given in Porter et al. (this volume) .

The cage system used for each crab hatch included a hatch
cage and a growout cage. Crabs were hatched and reared during
early larval stages in unattended cages suspended in the water at
field sites 1-3 (Tables 1-3) . From approximately 15-25 days
post-hatch, cultivated algal turf screens were introduced into
the juvenile crab cages on a regular schedule. Juvenile crabs
were transferred to a larger cage (Tables 1-3) when the average
carapace length of the group was 10 mm or greater. At site 1 an
additional intermediate cage was used as crabs were transferred
again at 100-125 days (Table 1) . At all sites juvenile crabs
were transferred to the growout cage when the crabs had attained

sizes of approximately 20-25 mm CL.

The crab cages were regularly sampled to determine carapace
length, survival rates and to assess general culture conditions.
Sixty hatches were examined, 56 carried through 100 days post¬
hatch, and 22 broods through as many as 280 days of growth in the
growout cage. Measurements of 23,000 post-larval crabs over a
total span of 472 days were obtained to derive the data discussed
herein .

RESULTS

Growth

In this investigation, the mean carapace length (CL) as
measured at 20 day intervals up to 280 days post-hatch is plotted
for each study site to compare overall growth rate data
regardless of location variance. Crab growth rates during the
first 60 days (three sampling intervals) ranged from 0.11-0.19
mm/day for all sites, with crabs cultured at site 3 having the
highest rate. Carapace length (CL) vs. time up to 280 days post¬
hatch is shown in figure 5.

Crabs 60 days post-hatch had a mean carapace length in mm of:
site 1) 7.76;+ 0.49; n=14382 ; site 2) 6.42; + 0.25; n=384 1 and

site 3) 11.6 mm; + 0.80; n=3000. At 100 days post-hatch the

mean CL from site 3 was greatest. The maximum period of culture
was 472 days (site 1) where specimens averaged 79.88 mm CL (SE +
2.43; n =5), with a maximum of 91.0 mm CL. Because the project

at Buen Hombre was closing down for the last 100 days, these
crabs were poorly fed and growth rates were low. Some cultured
crab populations averaged similar or larger sizes (66-84 mm CL)
at much earlier periods post-hatch (e.g., 230-318 days post¬
hatch). At site 2 a mean CL of 76.88 mm CL+ ( n--5) was measured
after a 318 day period of culture. The deviation of crab
carapace length from the mean among a population increased with
age. For those cultured crab populations with large numbers of
measured individuals, from 0-100 days post-hatch the mean growth
rate increases from 0.09 mm/day to 0.15 mm/day or greater.

Beyond 100 days post-hatch, the mean growth rate began to climb
at two of the sites for the remainder of the study period; of
0.31 mm/day (site 1); 0.38 mm/day (site 2); 0.13 mm/day (site
3). Beyond 180 days at sites 1 and 2 (combined) growth rates
exceeded 0.50 mm/day.

Experimental Diet

Cultured algal turfs were provided to each cage every two to
six days, depending on the algal production rates of each site
(see Adey, et al . , this volume) . Algal turf screens were
exchanged more frequently at site 2 (necessitating a larger
number of screens per cage) to offset the lower algal turf
production levels. Table 1 presents the amount of algae turf
fodder provided to crab culture cages per day at each site.

Algal turf growth was greatest at site 3, as was the rate of

crab growth. The algal turf growth at sites 1 and 2 was
moderate, necessitating an increase in the number of screens per
cage and an increase in the number of times that the screens are
changed. However, we were unable to construct a sufficient
number of plastic screens during the study period to offset the
lower algal turf production levels. This is especially evident
in the first 120 days of post-hatch growth as the lower levels of
algal turf production at sites 1 and 2 depressed early crab
growth rates. On the other hand, after 120 days at site 3, the
project was closing down, less time was available for feeding
and growth rates fell off.

The feeding rate in the growout cages and the intermediate
cages was much greater than in the hatch cages and the crab
density per cage was greatly reduced, thus increasing the level
o-f survival. These factors promoted rapid growth and a lower
relative mortality rates for all crabs in culture from 100 days
to 300 days post-hatch. For those reasons, the Carriacou growth
rates are used in a composite curve (Figure 6) to 100 days and
then the Buen Hombre and Antigua rates are used. The molt
intervals are calculated from Biddlecomb, et al . (this volume) ,
figure 3.

Survival

The average gravid female crab carries 60,000 eggs (Craig, et
al. , this volume) thus the density of zoea in the hatch cages

exceeded 45 crab larvae/cm of cage screen surface area, in the
best of circumstances. Approximately 2000 crabs survive the
larval period to five days post-hatch using the hatch cage
techniques employed in this study (Xglefaart, et al., in
manuscript) . Though the larval density greatly affected larval
survival to the first post-larval stage, instar I stage (3.3%
survival), the non-technical , low cost "in situ" lagoon cage
hatchery methods, provided an adequate number of stage I
postlarval crabs (n=2000) for continued growout. Mean survival
to 100 days in hatches conducted during the beginning of the
research study was 4.93% and in the later studies increased to
10.9% (n=Xl hatches at site 1).

Upon transferring crabs from the hatch cages to the larger
mesh cages, survival rates increased to 64% from 100-150 days
post-hatch (site 1); 63% from 100-160 days post-hatch (site 2);
and 39% from 80-120 days post hatch (site 3). The mortality rate
beyond 150 days post-hatch at all sites is considerably less than
in the first 150 days post-hatch. By the 300th day of post
larval culture, the number surviving was 16-28% of the total
number of crabs initially placed in the growout cage. Of the
hatches initiated earlier in the study at site 1, while we were
still developing procedural methods for larval rearing, the total
number of crabs from each hatch surviving beyond 300 days was not
sufficient to analyze statistically.

DISCUSSION

Crabs cultured at site 3 exhibited the greatest rate of
early growth, attaining a mean size of 11.6 mm CL at 60 days. In
general, when the crabs attained sizes of 10 mm CL or at 80-100
days post-hatch, they were transferred from the hatch cages to
larger mesh cages. The larger screen mesh effectively decreased
detrital buildup in the cage. At site 3 (compared with sites 1
and 2) crabs were larger throughout the entire culture period of
180 days. Up to 120 days, growth showed a log rate of increase
(Figure 7) . However, mean growth rates started to fall off after
100 days (less labor was available for feeding). In several
cases (4 hatches) at site 3 the cultured crab population achieved
growth rates up to 0.40 mm/day CL based solely on algal turf
feeding. This can be partly attributed to the greater quality
and quantity of algal turf biomass cultivated per screen per unit
time and the ensuing benefits derived from higher feeding rates.
This suggests that growth in crabs cultured at sites 1 and 2 was
retarded through at least 100 days post-hatch. Crab growth rates
increased at sites 1 and 2 to a comparable level attained at site
3 after 150-180 days culture, most likely due to lower crab
density per cage (31 crabs/ m ) and greater feeding rates.

Managing young larvae is not particularly time consuming, so
that at sites with lower algal turf productivity we suggest that
the easiest method to increase crab growth rates may be by
splitting hatches into several cages (thereby decreasing crab

density/cage) , increasing algal turf screen rate of exchange into
and out of the cages, and increasing the hatch cage size.

Zoea had metamorphosed to the first post-larval stage crabs
by 144 hours post-hatch at all study sites. For the first 20
days post-hatch the fine meshed screen initially allowed adequate
exchange of seawater through the cage. Eventually however, the
fine meshed screen became clogged with microfloral algal growth
as well as detrital material. This apparently decreased the
water quality and exchange of clean water into the cage. As the
crab biomass increased with time, mortality exponentially
increased (Figure 8) . In addition, if the algal turf screens
that were placed into a cage were not fully grazed, the ungrazed
algae then would die and slough off the plastic screen to collect
at the cage bottom. This buildup of detrital matter (e.g., algal
matter plus crab feces and crab molt shells) is thought to have
substantially contributed to the fouling of the cages and
increased crab mortality rates in the cages. A cage design
allowing the removal of detritus, without damaging crabs, would
greatly improve survivorship.

Crabs remained in the hatch cages for as long as 120 days in
some cases. The resident crab population in each cage decreased
with time. The first 60 days of cage culture resulted in
approximately 13-26% survival at all sites. By the 70th day
post-hatch, if the crabs were left in the hatch cage, crab

mortality greatly increased. This necessitated the development
of procedures to transfer crabs from the hatch cage to the larger
mesh cages at the 80-100 day mark or at approximately 10 mm CL.
The coarser screen mesh greatly enhanced the exchange of seawater
through the cage and removal of detrital material from the cage.
Crab survival to 100 days post-hatch was 5.29% (site 1); 2.92%
(site 2); and 6.7% (site 3).

At site 3 survival decreased by X/3rd from 60 days (survival
= 23%) to 80 days (survival - 14%) post-hatch. It is believed
that the greater cultured algal biomass per screen at site 3
contributed to a greater detrital buildup in the cage. The
greater quantities of algal turf per crab increased crab survival
rates, crab growth rates, and thus greater amounts of metabolic
byproducts (feces, molt shells). The total detrital buildup in
the cage appeared to reach a saturation point at approximately 60
days post-hatch, when a mass mortality of crabs occurred. Upon
opening several cages (site 3) at 60 days post-hatch, l/4th to
l/3rd of the total crab population was found to be dead, and
lying on the bottom of the cages. This reinforces the concept
that the waste products, due to increased biomass level (as
compared to crab survival and size at sites 1 and 2) at 60 days
post-hatch substantially increased the fouling levels of detrital
material entrapped by the fine meshed screen. As mentioned
above, a new cage designed to reduce detrital material might be
critical to increasing survivability. Overall however, the

greater number of crabs surviving the first 120 days of culture
at site 3 is attributed to the increased feeding rate and earlier
transfer of crabs to a larger mesh cage as compared to sites 1
and 2. Crab survival from 120 days post-hatch to the end of the
culture period remained at a high level at all sites. Site 2 had
the poorest level of survival (12%) from 120 days to 300 days,
while the numbers surviving at sites 1 and 3 changed very little
from 120-300 days post-hatch. At the end of the culture period
the survival rates were 18% (300 days) and 28% (175 days) at
sites 1 and 3 respectively.

In the investigation period reported here, growth time was
not long enough to bring any crabs to market size, though the
largest sizes attained (91.0 mm CL) (site 2) were only one to two
molts away. During later unmonitored studies at site 1, several
crabs were brought to. adult size.

Evaluation of Rearing Systems

The dimensions and therefore surface area of the hatch cages
was found to have the greatest influence in terms of success for
these mariculture techniques. The initial larval and first post-
larval stage crab densities in the hatch cages were much too high
for the cage sizes used in this study. Several alternatives
exist to ensure adequate feeding rates and space requirements.

By enlarging the cage, the screen surface area may be increased.
Hatch success therefore depends upon using a larger hatch cage

and/or transferring five day post-hatch crabs from the hatch cage
to several cages of similar design. This should increase the
feeding rate as well as decrease the crab density and lessen the
detrital buildup as the crab biomass is greatly reduced. All of
these factors increase crab survival and growth rates, allowing
the crabs to be transferred to a larger meshed cage at an earlier
age .

Following 100 days post-hatch, the culture of crabs in open
water cages approached levels of growth and survival necessary
for mariculture operations. We suggest that the redesigning of
the hatch cage to allow the removal of detrital matter etc. ,
combined with proper, consistent algal turf feeding rates will
significantly improve crab survival, as well as allow crab growth
rates to attain mean values of 0.30 mm/ day CL through 400 days
post-hatch as achieved in several individual brood rearings.

Local fishermen were hired to work on the project at Site 1.
Their initial level of marine husbandry skills was very low, but
as the artisinal fishermen worked on the project, the level of
skill and knowledge greatly increased. This is directly evident
in Figure 9, where the relationship of the survival of crabs at
100 days post-hatch relates to the advancement of time of the
study period. By day 500 of the study period, survival rates had
increased by a factor of 10. Using the growth rates of figure 5
and the length to weight ratios developed by Porter et al . (this

volume) , after 100 days cultured crabs double in weight every 25
days .

Although adult crabs in cages show little aggression towards
each other, and very few incidents of dismemberment have been
encountered, some of the relatively few juveniles kept in aquaria
for observation have attacked and dismembered their cohorts.
Rarely has such an attacked crab been eaten. The extent of this
problem is not known and as a standard practice we have included
pvc pipe habitats in cages to provide protection during molting.

It seems clear that if environmental quality is high, M.
spinosissimus maintain relatively high, probably genetically
fixed growth rates. Sometimes in this study the algal feed
supplied was insufficient and this may have resulted in
relatively low survivabilities. However, the often very high
growth rates suggest another mechanism. In spite of the comments
of Ryther et a 1 . (1987) , disease does not seem to be a factor in

our work. The level of nitrate at 35 mg/1 in the Ryther study is
extremely high as compared to the wild environment (approximately
10,000 times higher) and suggests poor environmental conditions
in that investigation. Even though adult crabs show little
aggression in captivity, there is some evidence for mortal
aggression in juveniles (see also Ryther et al. , 1987). While we
have typically provided molt habitats in many of our cages, these
may well be inadequate. We are now trying a wide variety of
habitat types.

CONCLUSIONS

Using the algal turf feeding methods and crab culture
techniques described In this volume for spinosissimus , mean

growth rates of approximately 0.27 mm/ day CL from hatch to at
least 280 days were achieved during the last six months of the
study . Although from 180 to 280 days, growth rates exceeded 0.45
mm/day , beyond 280 days, data is minimal and only suggests a
possible decrease in growth rates. Nevertheless, it should be
possible to grow cultured crabs to maturity in less than 400
days .

Crab survivorship averaged 22% to 60 days, 23.6% from 60 to
120 days and 18.5% from 120 to 300 days. Methods of improving
survivorship are discussed. In particular, increasing the
cultured algal turf quantities available to each juvenile crab,
providing a means of periodically and safely removing detritus
from the juvenile cages and providing a variety of molt habitats
in the adult cages should increase survivorship significantly.

LITERATURE CITED

Brownell, W.M. , A. J. Provenzano, Jr. and M. Martinez. 1977.

Culture of the West Indian Spider Crab Mithrax spinosissimus
at Los Roques, Venezuela. J. of World Mariculture Soc.

Coen, L.D. 1987. Dissertation: Plant animal interactions:
Ecology and comparative functional morphology of plant
grazing decapod (Brachyuran) crustaceans. 241 pp.

Colin, P.I. 1978. Caribbean reef invertebrates and plants.

T.F.H. Publications Inc. Neptune City, N. J.

Iglehart, J.M. , K.L. Porter and W. H. Adey. In manuscript .

Caribbean King Crab (Mithrax spinosissimus , Lamarck) Cage
Culture for the Artisanal Dominican Fishfarmer.

Munro , J.L. 1974. The Biology, ecology, exploitation and

management of Caribbean Reef Fishes. Part V. The biology,
ecology and bionics of Caribbean reef fishes: Crustaceans
(lobster and crabs) Res. Rept. Zool. Dept. Univ. West Indies
3 (6): 39 -48.

Porter, K.L., J.M. Iglehart, W.H. Adey, and M.W. Yadven. In

Review. Cage Culture of the Caribbean King Crab (Mithrax
spinosissimus . Lamarck) Using Algal Turfs for Feed.

Provenzano, A.J., Jr. and W.N. Brownell, 1977. Larval and early
post-larval stages of the West Indian spider crab, Mithrax
spinosissimus (Lamarck) (Decapoda: Majidae). Proceedings of
the Biological Society of Washington 90(3): 735-752 .

LEGENDS

Figure 1

Figure 2

Figure 3

Figure 4
Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Map of Caribbean showing coastal research field sites.
Buen Hombre, Dominican Republic (site 1) ; Nonsuch Bay
Antigua (site 2) ; and Grand Bay, Carriacou, Grenada
(site 3) .

Plastic screens (0.61 x 0.92 m, 2 . 0 x 3 . 0 mm mesh) for
algal turf cultivation. Screens suspended vertically
in the water column from the float line.

Hatch cage/ intermediate cage. Hatch cage dimensions
vary from 1.0 Lx .35 W x .70 H, to 1.0 Lx .60 W x
.70 H with 0.5 mm mesh. Intermediate cage (1/4"
mesh) identical in design to hatch cage but with
dimensions as follows 1.0 L x 1.0 W x 1.3 H.

Growout cage. Dimensions are 2.6 Lx 1.0 Wx 1.0 H
and hold 23 screens. Mesh size is 1/4" - 1/2".

Relationship of carapace length to age showing growth
of algal turf fed It soinosissimus at sites 1-3.

Points represent mean CL of n=5745 (site 1) ; n=2086
(site 2) ; n=479 (site 3) .

Projected growth rate curve based on rates routinely
achieved during the last six months of the project.
Molt numbers and sizes based on figures 2 and 3
(Biddlecomb, et al . , this volume) and figure 7 (this
section) .

Linear relationship of carapace lengths of crabs (site
3) at successive intervals of time (Carapace lengths
transformed to logs) (r = 0.96); y = 1.249 + 0.016 x;
n-479 ) .

Relationship between age and survival of post-larval
crabs. All post-larval crabs transferred to larger
mesh cage at 100 days (site 1) , 100 days (site 2)
and 80 days (site 3). Site 1 crabs transferred
again at 125 days. (Dashed lines represent transfer
of crabs to new cages) .

Relationship showing increase in success of culture
technique refinement of post-larval culture of 100 day
old crabs. # of hatches = 25; y = 108.7; SE + 22.54

Table 1.

Table 2 .

Table 3 .

Amount of algal turf fodder provided to crab culture
cages and crab dimensions. Site 1.

Amount of algal turf fodder provided to crab culture
cages and crab dimensions. Site 2.

Amount of algal turf fodder provided to crab culture
cages and crab dimensions. Site 3.

STUDY FIELD SITES

Figure 1

Figure 2

SIDE VIEW WITH LID

TOP VIEW

Figure 3

SIDE VIEW END VIEW

Figure 4

2.6 m

NUMBER OF CRABS MEASURED

20 60 100 140 180 220 260 300

AGE IN DAYS

Figure 5

Carriacou Antigua and Buen Hombre

lulu Nl H10N31 BOVdVdVO

Figure 6

100 200 300 400 500

AGE IN DAYS

CARAPACE LENGTH IN mm

Figure 7

Buen Hombre
Site 1

IVAIAdflS {%) lN30d3d

Figure 8

DAYS

Figure 9

TABLE 1

SITE 1

AMOUNT OF ALGAL TURF FODDER PROVIDED TO CRAB CULTURE CAGES

AND CAGE DIMENSIONS

Hatch cage

Intermediate cage

Growout

Dimensions 1

(meters)

. 0 x .35 x .70

1.0 x 1.0 x 1.3

1.2 x 1.2 x 2

Cage surface

area

2.6 sq. m

7.2 sq. m

13 sq. m

Cage volume

0.3 cu . m

0.5 cu. m

3.6 cu. m

Screen feeding

rate 2

screens/ 5 days

6 screens/5 days

4 screens/5
days

Total Algal 98
Turfs

Initial Algal
Feed Rate

.3 gm (dry) /5 days

655 gm (dry) /6 days

364 gm (dry ) /
days

crab/day 0.

Algal Turfs
per crab per

05 gm/crab/5 days

6.36 gm/crab/5 days

5.6 gm/crab/
5 days

day 0 .

Initial Density
Crabs/area

01 gm/crab/day

1.27 gm/crab/day

1.1 gm/crab/
day

of cage surface

769/sq. m

14.3/sq. m

5/sq. m

TABLE 2

SITE 2

AMOUNT OF ALGAL TURF FODDER PROVIDED TO CRAB CULTURE CAGES

AND CAGE DIMENSIONS

Hatch cage

Growout cage

Dimensions

(meters)

1.0 x .35 x .70

2.6 x 1.0 x 1.0

Cage surface
area

2.6 sq. m

12.4 sq. m

Cage volume

0.3 cu. m

1.8 cu . m

Screen feeding
rate

2 screens/ 3 days*

6 screens/4 days*

Total Algal
Turfs

117.9 gm (dry) /3 days

874 gm(dry)/4 days

Initial Algal
Feed Rate
crab/day

0.06 gm/crab/3 days

15.3 gm/crab/4 days

Algal Turfs
per crab per
day

0.02 gm/crab/day

3.8 gm/crab/day

Initial Density

Crabs/area

of cage surface

769/sq. m 4.6/sq. m

-- Site 2 algal turf production was heavily biased by entrapped
sediment, therefore the true algal turf production may be
considerably less. See Adey and Hackney in this Volume.

TABLE 3

SITE 3

AMOUNT OF ALGAL TURF FODDER PROVIDED TO CRAB CULTURE CAGES

AND CAGE DIMENSIONS

Hatch cage

Growout cage

Dimensions

(meters)

1.0 x .60 x .70

2.6 x 1.0 x 1.0

Cage surface
area

3.4 sq. m

12.4 sq. m

Cage volume

0.4 cu. m

1.8 cu. m

Screen feeding
rate

3 screens/ 6 days

9 screens/5 days

Total Algal

Turfs

305 gm (dry) /6 days

1368 gm(dry)/5 days

Initial Algal
Feed Rate
crab/day

0.15 gm/crab/6 days

5.0 gm/crab/5 days

Algal Turfs
per crab per
day

0.03 gm/crab/day

1.0 gm/crab/day

Initial Density

Crabs/area

of cage surface

581.4/sq. m 22 . 3/sq. m

SECTION III: MARICULTURE TECHNIQUES

MARI CULTURE TECHNIQUES FOR MITHRAX SPINOSISSIMUS
INCLUDING BROODSTOCK LARVAL AND POST-LARVAL
REARING AND GROWOUT TECHNIQUES

MARI CULTURE TECHNIQUES FOR MITHRAX SPINOSISSIMUS
INCLUDING BROODSTOCK LARVAL AND POST-LARVAL REARING

AND GROWOUT TECHNIQUES

K.L. Porter, J.M. Iglehart, M. Craig and W.H. Adey

INTRODUCTION

Innovations developed over the past 50 years in the fishing
industry have produced a world fish catch totaling approximately
70 million metric tons annually (Pillay, 1985; 74.8, Simon and
Kahn, 1984; 74.0, Brown, 1985). The use of electronic equipment
and modern fishing vessels caused a six percent increase per year
in the total catch in the 1950's and 1960's. In the last decade,
however, the total worldwide fish catch has increased less than
one percent per year, accompanied by serious reductions of fish
stocks in some important fisheries. This worldwide depletion of
the fishing stocks has resulted in a more concerted effort to
develop aquaculture and mariculture. Annual worldwide output of
aquaculture products now exceeds 10.5 million metric tons
(Pillay, 1985) .

In the late 1970 ' s and early 1980 's, Smithsonian scientists
of the Marine Systems Laboratory (MSL) , while conducting
microcosm research, developed a technique for culturing algal
turfs on artificial substrata to manage the water quality in
closed systems (Tangley, 1985; Adey, 1983; Adey, 1987). This
technique reproduces, in manageable form, the high algal turf

production levels that characterize these benthic algae
communities in the wild. In these "scrubber" systems, laboratory
production levels of five to over 15 g dry m-2 d”1 are common.
Algal turf production levels of 8-18 g dry m“2 d-1 are attained
when similarly designed "algal scrubbers" (in this case,
suspended plastic screens) are anchored in turbulent tropical
Caribbean waters (Adey and Goertemiller , 1987; Figure 1).

The highly diverse and productive algal turfs are consumed
by a wide variety of animal grazers in the wild. However, a
direct use of algal turfs for humans has yet to be developed.

Many modern maricultures and aquacultures are capital intensive,
require considerable technological expertise and utilize
expensive industrial food preparations. It was hoped that a
system could be developed that would allow local fishermen in
underdeveloped countries the means to tap this rich and readily
available food source.

Mithrax spinosissimus (Figure 2), a large tropical spider
crab, which consumes smaller benthic algae and algal turfs
naturally, was chosen as a potential candidate for this new
approach to mariculture. A full life cycle mariculture for this
Caribbean King Crab has been developed for utilization in
developing countries with low technological capabilities. While
brooding, hatching and juvenile growout have been quite
successful for a "low tech" mariculture based on early pilot
studies, late juvenile mortalities have been higher than

predicted. Efforts directed at improving late juvenile growout
techniques are continuing. No published efforts have been
undertaken previously to develop an operational mariculture for
Mithrax spinosissimus . although several preliminary biological
studies have been carried out. These previous studies are
discussed in detail in the earlier sections of this volume.

For the past three years, the development of Mithrax
spinosissimus mariculture by MSL has been funded by the United
States Agency for International Development. As part of this
project, Mithrax spinosissimus cage culture has been conducted in
a number of Caribbean/West Indian islands, including the Turks
and Caicos Islands, the Dominican Republic, Antigua and in
Carriacou, Grenada (Figure 3) .

Mariculture production techniques for Mithrax spinosissimus .
as developed by MSL, have been separated into care of brooding
females, hatching, juvenile, older juveniles and immature adult
phases. Gravid females, obtained from the wild and brooded in
captivity, have provided a regular and reliable supply of newly
hatched juvenile crabs. A gravid female, which produces an
average of 60,000 eggs every two months (Craig, et al. , this
volume) , is placed in a sea-cage prior to hatching. Such "hatch
cages" (90 X 40 X 110 cm) are made of a wood frame, coated with
polyester resin and covered by a fine (0.5 mm) plastic mesh
screen. Newly hatched crabs (approximately 2000 first post

larval instar crabs at five days post-hatch) are left in the
"hatch" cage until 60 days post-hatch. At that time, they are
transferred to an identically designed cage, which is covered
with a coarser mesh screen for the next 40-60 days. These
methods have produced 100-400 juveniles from each hatch surviving
to 100-120 days old. Crab growth rates and survival following
hatch depend in large part upon the amount of algal food made
available and the density of juvenile crabs in relationship to
algal turf screen area. Care in construction of cages and in
management of stocking and feeding is also required to avoid the
introduction of predators or algal turf competitors, which are
primarily small invertebrates.

Growout to harvest of Mithrax sninosissimus is carried out
in similar wood frame/plastic cages (2.2 X 1.0 X 1.0 m) enclosed
with plastic mesh ( 1/4-1/2 " ) . When the crabs are 100-120 days
old, they are transferred to these juvenile/adult growout cages
which are designed to hold 30-50 harvestable crabs. To date, our
work indicates that a crab requires 12-15 months of growth to
reach a harvest size of 0.8-1. 3 kgs, though it is quite likely
that further refinement can both shorten the time and increase
the size of the harvested animals. Crab survival rates in these
growout cages need to be about 50%, yielding 50 harvestable crabs
per cage to be successful. Chronic mortality has led to a
survival rate of 15-25% per 100 days. Small predators in the
cages, excessive wave action during molting and aggressive

interactions of the crabs appear to be responsible. This last
phase to improve survivorship requires additional research
effort. It is thought that the problems can be solved by cage
design and efforts are underway at Grand Turk to achieve this.

In the MSL mariculture technique, turf algae is the primary
food source for cultured Mithrax spinosissimus crabs. Though
labor intensive, it is replenishable and easily grown as well as
highly productive and highly nutritious. Laboratory analysis
shows the algal turfs to contain 8-10 grams of protein per 100
grams (dry) of algae.

During the latter part of the early juvenile stage and
during the entire growout stage, algal covered plastic mesh
screens are fitted into the cages. As the crabs consume the
algae, the screens are replaced every three to four days. The
algae is cultivated on screens suspended from lines that are
floated in turbulent waters. Wave action, current and sunlight
in a low nutrient water environment provides optimal conditions
conducive to algal turf growth on these plastic screens. Before
being placed in a crab cage the plastic screens must have
sufficient algal growth, which takes 20-40 days to initiate.
However, repeated feedings typically require 7-14 days for
sufficient algal re-growth. Algal growth rates are variable
during the year due to changes in weather and other environmental
factors. Sunny, windy weather, typical of the "trade" wind

islands, produces maximum algal growth. Environments
consistently rich in suspended carbonate sediment need to be
avoided as continual sediment accumulation on the screens must be
removed or algal production will suffer.

The changing of screens and the tending of the juvenile crab
and growout cages is labor intensive. A one man operation
utilizes 24 growout cages with accompanying juvenile and
intermediate cages and about 800 screens. The operation requires
six to eight man hours per day. However, as in most fisheries
and in agriculture, it is desirable to carry out a larger
operation using teams.

In several Caribbean countries, including the Dominican
Republic and Belize, the wild harvest of Mithrax spinosissimus in
traps forms a small but important fishery. This animal is
generally regarded as a delicacy and is served in seafood
restaurants. Although older crabs often have a tough shell,
cultured animals, especially those newly molted to maturity, have
a thin shell. The Caribbean King Crab has a sweet, textured
meat, concentrated in the outer body, legs and in the large
chelae of the males. At 18-35% of total body weight, harvestable
meat is considerable.

BIOLOGY OF CULTURED CARIBBEAN KING CRAB

This study treats the species of crab commonly known as both
the "West Indian Giant Red Spider Crab" and the "Caribbean King
Crab." In Spanish speaking Caribbean countries, it is generally
referred to as "Centolla. " The species' name is Mithrax
spinosissimus and it is one of the evolutionarily more advanced
members of the Mithrax family (Goy et al . , 1981) . The biological
and ecological aspects of this species pertaining to mariculture
are briefly reviewed.

Caribbean King Crabs are generally found throughout the
tropical Western Atlantic, from southern Florida and the Yucatan
through the West Indies, to eastern Venezuela at depths of 2-200
meters. Individuals may be encountered at night as they feed on
reef or pavement surfaces, during the day they tend to be
confined to caves and crevices at depths of 2-30 meters often in
small "communities" or harems. In these daytime hideaways, there
are usually several females in close proximity to one another,
often with one or two males. On the other hand, additional
"bachelor" males tend to be isolated and distantly dispersed.

The abundance of Mithrax spinosissimus in the Caribbean/West
Indian area varies greatly from island to island and from site to
site. Although a quantitative survey of the entire Caribbean has
not been done, it appears that crab distribution and abundance
are a function of the availablity of appropriately-shaped caverns
or crevices and probably the proximity of good algal feeding

grounds. (See notes from population surveys in Jamaica: Munro,
1976; Florida: Bohnsack, 1976; Hazlett , 1975; Dominican Republic,
Antigua, Turks and Caicos and Mayaguana : Iglehart , et ad., this
volume; and Belize: Koltes, personal communication.) It is
likely that the abundance of fish, crab and octopus predators are
also crucial in determining Mithrax distribution and abundance.

With knowledge of the local reef terrain, Mithrax
spinosissimus is most easily obtained by SCUBA from dusk into the
evening. Crabs may also be caught in traps. Wild Mithrax
spinosissimus population sizes vary from reef to reef (Iglehart,
et al . , this volume) . While the very real likelihood of stock
improvement through breeding exists, the initial breed stock is
quite important. In this context, crabs from the Florida region
are by far the smallest and therefore least desirable (Iglehart,
et al . , this volume) .

The extensive observation of both wild and captured crabs
during the process of feeding, the growth and development of the
crabs in culture on a diet of algal turfs and the examination of
crab stomachs and the structure of both the chela and gastric
mill all support the conclusion that fL spinosissimus is a
facultative herbivore on smaller algae. In the course of
grazing, organic detritus, sediment, coral bryozoans, and other
small organisms (eg. , amphipods) are almost invariably ingested.
While Mj_ spinosissimus is not particularly discriminating in its

choice of algae, some macroalgae are strictly avoided. When a
wide variety of algal turf, along with some macroalgae are
presented on cultured screens, the turfs are almost invariably
eaten before the macroalgae. Mithrax spinosissimus will eat meat
(conch, urchin, e.g.) when provided. However, it has not been
demonstrated that meat is necessary in any way for growth or
reproduction. The highest known growth rates and maximum egg
production have been achieved on a pure algal turf diet.

The large claws or chelipeds terminate in long narrow fingers
with crenulated spoon shaped tips. These aid in the digging and
sifting of the substratum for algal material and also provide
strong sharp tools for the cutting, tearing and pulling of algal
holdfasts and stalks. The paired appendages associated with the
endostome "teeth" hold and tear food and assist in respiration
(For more information on feeding mechanics, see Coen, 1987)

The gastric mill, located inside the mouth between two
stomachs, is well-developed for algal turf feeding. The
urocardiac ossicle has a single large "grinding" tooth and two
smaller, more pointed ones. The opposing zygocardiac ossicles
are ridges and most likely perform crushing and chewing motions.

A small row of soft spines are located opposite the urocardiac
ossicle side of each zygocardiac ossicle. These are not
connected directly to either ossicle and most likely are utilized
to transfer food in and out of the gastric mill.

Research to date has not focused on providing phytoplankton

to the larval stages. While good planktonic feed might improve
their survivorship at 30-50 hours post-hatch, survival of 2000
post-larval crabs from a single brood is normal without special
plankton feeding. After settling, they shift over to micro-
benthic algae in algal turfs (particularly diatoms) , through the
megalopae and early crab stages. From 1-10 mm CL, the young
crabs feed upon diatoms, blue-green algae, organic detritus and
smaller turf algae. As they grow to carapace lengths of 10-25 mm
CL, the larger algal turfs form the majority of the diet. For
crabs above 25 mm to 80 mm CL, the natural diet is primarily
composed of larger algal turfs with some included macroalgae.

The final period of growth is characterized by a diet including a
wide variety of benthic algae, excluding many of the algal
species that are mildly toxic, or protected by carbonaceous or
similarly tough cortex. Note that similar crabs in Florida and
apparently some other areas in the northern and western
Caribbean, especially in those growing in mangrove communities,
are much smaller (Iglehart, et al. , this volume). They may be a
separate species or subspecies.

In natural waters, Caribbean King Crabs attain carapace
lengths of 120-180 mm, weighing 0.8-3 kgs. Under culture
conditions, crab growth rates are a function of density and algal
turf feeding rates, as well as the food conversion rates of the
crabs. The daily linear growth rates of ft spinosissimus , under

conditions of constant algal turf and macroalgal feeding, have
been found to be initially 0.10-0.20 mm CL/day ranging up to over
0.50 mm/day at 200 days, thus attaining a mature weight of 0.8-
1.8 kgs and 100-150 mm CL at 12-15 months growout (Figure 4).

Wild crabs are known to achieve a weight exceeding 2 kgs,
probably at an 18th molt. It is not known whether this can be
achieved in culture.

M. spinosissimus undergo a molt of puberty which is generally
the final molt, wherein both males and females take on their
final adult secondary sexual characteristics. There is some
question as to whether or not they undergo a second molt. This
could be important to the mariculture of very large animals
(Biddlecomb, et al . , this volume) . Copulation has been observed
among hardshell crabs that have passed through the molt of
puberty.

On average, every 60 days, females produce a batch of eggs.
Subsequent egg clutches are fertilized by sperm which are stored
in spermathecae . The sperm remains viable in the spermathecae
for an extended period of time, though secondary mating may
occur. Mating and fertilization occurs easily in captivity in
both aquaria and cages.

Crabs may be ovigerous at any time of the year. Egg
development requires approximately 30 days, at which time 40,000-
70,000 eggs are released (Table 1). In the early stages of
development, the fertilized eggs (about 1 mm in diameter) are

attached to the female pleopods and are orange in color. As
embryogenisis proceeds, they change to an orange-brown, red and
finally amber-red color . Towards the end of the egg development
cycle, the eggs are actively brooded by the female and the
frequency of cleaning and aeration of the brood is increased.
Normally, it is possible to determine time of release, within a
day or two (Craig, et al . , this volume) , by sampling and
observing the eggs closely. The eggs almost always are released
at night. The entire process of maintaining a breeding
population and bringing the eggs to a successful hatch in a cage
situation in the water is relatively simple and has been
accomplished approximately 150 times using the methods outlined
below.

Newly hatched rt spinosissimus larvae normally develop
through: 1) a non-swimming prezoeal stage; 2) swimming first

(two hours post hatch) and second zoeal (36-48 hours post hatch)
stages; 3) a non-swimming, benthic megalops (40-72 hours post-
hatch) ; and then 4) first crab (70-90 hours post-hatch) (Figure
5) .

Crab growth is dependent upon temperature, molt frequency,
food quantity and quality and the stage in development. Molting
is the primary method of growth. Increments of molt show an
average of 30-40% increase of carapace length per molt in the
early juveniles decreasing to 10-25% for the pre-adult and adult

molts. After the early juvenile stages, considerable overlap in
carapace sizes between instar classes occurs.

Growth is rapid in juvenile crabs. Second instar to eighth
instar molt frequencies decrease from about 4-6 day intervals
( ie . , between first and second instar) to 18-20 days by the 60th
day post-hatch. Juvenile crabs are spiny, rather elongate and
often heavily decorated with algal turfs. Not until the eighth
crab instar does the overall appearance resemble that of the
immature adult form (Biddlecomb et al . , this volume) .

Not including "Florida crabs," puberty is estimated to be
attained at the 16- 17th post-larval molt or under optimum growing
conditions, approximately 8-12 months post hatch. Growth of
younger juvenile males and females is only slightly, i . e. , all
the parts increase in size at roughly the same rate. However, as
they approach and then become sexually mature, the growth of the
males is allometric (with regard to their large chelae) and the
females' growth is allometric with regard to their abdomen width.
The range of sizes (CL) at which sexual maturity is attained
varies considerably. Excluding "Florida crabs" and similar small
populations, the mean size of sexually mature females from many
sites is 124.5 mm CL, 0.9 kgs; and for males 144.8 mm CL, 1.5
kgs .

SELECTION OF CRAB MARI CULTURE SITES

General

Two basic factors are crucial in site selection for Mithrax
spinosissimus mariculture as described in this paper: 1) a
shallow, turbulent (wave and current) and moderately sediment-
free locality for growing algal turfs; this should be in
reasonable proximity to 2 ) a moderately quiet lagoon, with some
current. We have found that sites where a coral reef breaks the
surface at low tide are ideal as they provide sufficient
protection of a lagoon area in which both algal turfs and crab
cages can be kept and easily worked from small boats.

Ideally, such a site should have a broad back reef flat or
sandy reef apron of two to four meters depth with constant wave
surge and wave driven currents for algal growth. It is then
desirable to locate crab cages in somewhat calmer shallow
lagoonal water of two to six meters depth, not too far from the
algal screens. Moderate currents are desirable. A sandy bottom
substratum allows maximum reflective light for algal turf growth.
However, water of relatively high clarity is desired. Excessive
suspended sediment entrains within the algal turfs and eventually
reduces algal growth. If this occurs only occasionally, under
storm conditions, it can be managed by lightly brushing the
screens. If it occurs frequently, algal production is
considerably reduced. Water temperature should average 23-30°C
year round with a salinity of 33-38 ppt. In addition, the more

remote a site from boat traffic the better, as the extensive
algal turf screen and crab cage system requires a large area.

Numerous sites that fulfill these general requirements exist
throughout the Caribbean and West Indies. Of the five sites we
have worked at extensively, four were quite good (Mayaguana,

Grand Turk, Buen Hombre and Carriacou) . Nonsuch Bay, Antigua was
generally too turbid.

Requirements for algal turf growth

Mature algal turf screen cultivation results in biomass
productivity rates of 8-18 g dry m“2 d-1 over 7-14 days growth.

An appropriate algal turf community develops four to eight weeks
after emplacement of the screen on the water. It is desirable to
scrape the screens with a hand held piece of plastic once or
twice during this period. After the start-up or colonization
period, periodic harvesting of the algae every 7-14 days mimics
grazing and results in high, continuous biomass production. The
harvesting (or grazing in natural, benthic algal turf
communities) prevents overgrowth and competition for available
space, light and nutrients. Generally, the turf screens must be
harvested on a regular basis or undesirable algae and animals
colonize the screen. If screens have been allowed to overgrow
with undesirable organisms, it may be necessary to remove, dry
and clean them before the full efficiency of algal turf growth
can be achieved again. If algal turf screens are not to be

regularly used for feeding animals, it is necessary to regularly
scrape them to keep production level high.

Algal turfs are dense mats of small, anatomically simple
algae (usually less than several centimeters in height) belonging
to all major groups of benthic marine algae. The algal species,
which compose the complex turf associations, tend to be
anatomically simple as compared to the more morphologically
differentiated macroalgal types, although a few complex miniature
macroalgae are important turf elements as well. With almost all
cells of the algal thallus participating in metabolite exchange
and active photosynthesis, the energy efficiency of algal turfs
is higher than in other more complex plants. Since these algae
rely on rapid growth and reproduction to survive, they are not
characterized by protective skeletons or toxic defense chemicals.

a) Light

Algal turfs in a area of strong current and wave induced
oscillatory flow are able to maximize production in high light
intensities. However, in these brillantly-lit situations,
ultraviolet light levels are also high, and U.V. usually has a
detrimental affect on organisms, including algae. Considerable
testing (Adey and Hackney, this volume) has demonstrated that
algal turf production levels at the surface in tropical seas are
well below those at 20-100 cm. Also, since a screen has two
sides (equivalent to the extensive surface area of a reef) and

production on the underside of horizontal screens is somewhat
reduced, the transmission of light through the screen and
reflection off shallow sandy bottoms is also critical to
maximizing production. It has been shown that white, translucent
screens are more productive than black screens. In practice,
since horizontally suspended screens must be hung from four
points (thus doubling the time required to remove and re-string a
screen from a line) , the screens are hung vertically. Other
factors being equal, algal turf production on vertical screens is
slightly less than the production from horizontally suspended
screens. However, the time factor involved in the management of
the horizontally suspended screens more than outweighs the
decrease in algal production.

b) Water Motion

Trade winds and their seas drive ocean water over reefs and
across the back reef flats to the lagoon. In addition, the
shallowing water in these localities compresses the flow of the
equatorial current, raising flow rates from that source. Waves
approaching 15-50 cm in height on the backreef or lagoon side,
with currents of 5-35 cm/sec, are ideal for growing algal turfs
on screens. Yet these conditions are not generally so rough that
the maintainence of those screens without serious losses due to
overly rough sea conditions is a serious problem.

Screens must be held rigid to some extent in order to force
contact with the turbulent water and to maximize algal turf
growth. Thus, the screens are hung vertically and perpendicular
to the general movement of the waves and currents. A moderately
sized weight, such as a piece of 1/4" reinforcing rod, or sand
fill in the p.v.c. pipe frame, on the lower side of each screen
frame assists in providing "rigidity" against the flow and surge.
As the algal turf develops, the screens become more buoyant and
gradually lift to about an angle of 45°. Thus, once familiar
with the strength of the current at a site, a glance at the angle
of the screen line is usually enough to assess the maturity of
the algal growth.

c) Sediment

Algal turfs do not require the presence of fine carbonate
sediment to provide maximum production rates, and in laboratory

. . . . 9

culture, with wave surge and lighting (exceeding 1000 u/m / sec) ,
production rates of 5-18 g dry m d have been achieved without
significant carbonate inclusion. In reef environments with low
sediment loads, production rates near to or slightly higher than
those in the laboratory are achieved. At higher concentrations
of suspended sediment, apparent algal production rates (as
measured by dry weight) initially appear higher (to over 30 g dry
m-2/d-1) (Peyton et al . , in review) , but after several weeks to
months production tends to drop radically to levels of 3-8 g dry

— p _ 1 ,

m d . At very high sediment loads, real algal production can
be virtually nil.

Site Requirements for Crab Growth

As mentioned above, the mariculture of Mithrax spinas issimus
can be treated in four distinct phases: 1) breed stock; 2) larval
development; 3) juvenile development and 4) adult growout. To
some extent, different conditions are required for each of these
phases. In all cases, cages should be located in areas of
moderate current flow (0.5-5 cm/sec) and minimum sedimentation.

a) Hatchlings

Newly hatched crabs are particularly sensitive to strong
wave action. We have found that the damaging effects of waves or
chop can be avoided by sinking the hatching cages to several
meters depth. Unfortunately, this renders the exchange of algal
turf screens difficult, and is to be avoided if possible.

b) Juveniles and Adults

As far as we are aware, the older crab instars (over 20 mm
CL) are not particularly sensitive to light wave action and any
increased water movement is usually desireable. Larger waves
(approximately greater than 30 cm) at molt may cause some
mortalities. For growout, practical limit to roughness at sites
is determined by the ease of working algal screens and the crabs
themselves from a small boat.

Description of Crab Mariculture Research Sites

In the last several years, crab mariculture research has
been conducted at a variety of sites in the north central and
eastern Caribbean. In the algal chapter of this volume, detailed
maps of each of these sites are presented, so they will not be
repeated here. However, it would be helpful in terms of
analyzing potential crab mariculture sites if we briefly
described the advantages and disadvantages of each site for
Mithrax spinosissimus mariculture. The following is a list of
the sites in approximate order of site quality from best to
worst :

1) Grand Turk, east lagoon. The water quality at this site
is excellent, as it is derived directly from the North Equatorial
current. The reef and algal ridge to the east is rather patchy,
but continuous enough to force current compression from breaking
waves and to block the larger seas. Tidal currents also provide
a component of north/south water movement, and the entire east
side of the lagoon is sufficiently free from suspended sediment
to provide for maximum algal growth. The western portions of
this very large lagoon have too much sediment for good algal
growth, especially within a few hundred meters of the island
itself. Southwest or northwest winds, although quite infrequent,
can cause sediment entrainment in turf screens on the east side
of the lagoon. The lagoon waters can be rough at times, and

considerable practical working experience is required to learn
how to keep lines from becoming seriously frayed or entangled
with the combination of moderate seas and rotating tidal
currents. Crab cages for both juveniles and adults can be
effectively managed anywhere in this lagoon. In the winter, the
lagoon seas can sometimes be too rough for hatchling cages and
cause mortality in juvenile crabs. However, the protection
provided in the lee of the eastern cays provides excellent
hatchling andn winter growout localities.

2) Grand Bay, Carriacou. This was the most turbulent of all
our work sites. The constant flow and wave chop across the reef
crest provided excellent algal growth. Sea conditions, however,
were too rough for hatchling cages and thus these were anchored
in the middle of the lagoon at a depth of three meters from the
surface. . This caused an increase in labor intensity of the hatch
phase, but was quite successful. In addition, it was frequently
difficult to work the larger crab cages because of the strength
of the sea and current, although the sea itself did not appear to
affect the juvenile or adult crabs. We feel that a specialized
boat, perhaps a small catamaran, could overcome the sea
conditions. We understand that a continuing crab mariculture in
the somewhat more protected bay just to the North (Watering Bay)
has retained most of the algal growth and yet achieved an easier
work situation.

3) Buen Hombre , North Coast, Dominican Republic. This very
large lagoon is one that could support many crab mariculture
operations. It is oriented East/West and as a result, it tends
to be rather quiet. The sea conditions are conducive to raising
crabs. Winter northers, however, provide rough conditions on
occasion, and on many summer afternoons the trade wind combines
with onshore winds to provide quite rough seas in the lagoon.
However, an adjustment in the work day ( i . e . , early morning) will
offset these conditions. Probably, the most difficult practical
situation is that in which the working conditions are generally
so good that it is easy to be careless and lose cages and screens
on the occasionally very rough days. While this site is
certainly a good site, in the quiet lagoon waters' suspended
sediment reduces algal growth to moderate levels. The western
portion of this 35 mile long reef lagoon might very well be
closer to an ideal site.

4) Nonsuch Bay, Antigua. This bay is "closed" against the
island itself. The outflow channels are northeast and southeast
and they flow back into the trade wind seas, thus restricting the
outflow of water from the bay. The sediment load is therefore
quite high. Although this site was relatively quiet and easy to
work in and there were no direct energy related problems
associated with hatchling juvenile or adult cages, the algal
production was not good. Furthermore, although initial apparent
algal weights were high, due to included sediment, the screens

did not remain even moderately productive for long periods of
time. Encrusting animals and undesirable algae provided frequent
problems. Nonsuch Bay was an undesirable site. However, the
numerous other bays on the northeast side of Antigua would
probably provide many suitable sites.

MITHRAX MARICULTURE

Overview

Caribbean King Crab culture, as developed in this project,
requires the practice of some critical elements of animal
husbandry, as well as the cultivation of algal turf fodder. We
have attempted to develop the operation of a low technology
mariculture system based on units. The size of the operation
determines the number of units required. The operation, which
will be briefly described, is based on a single person unit of
labor. As in most husbandry, continuous attention to the
operation is required and in general two individuals (or two
units) would be the normal minimum level of production.

A unit operation consists of three phases, including six
hatchery cages, six intermediate cages and 24 growout cages. This
process allows for 20-30% downtime for each cage for potential
repairs. Six hatchery cages provide space for stocking a crab
brood every 32 days. At five days post-hatch, the crab density
is thinned by transferring half of the post-larval crabs to a

second hatch cage. At 60-80 days post-hatch, juvenile crabs are
transferred to the intermediate cage. The amount of crabs
produced in the intermediate cages will sufficiently stock two
growout cages with 100 crabs (20 mm CL, at 100-120 days age) each
month. To achieve a harvest of 100 crabs per month in a unit
operation, it is necessary to achieve a survivorship to adult of
50% during adult growout. To achieve one hatch each month, a
brood stock of about nine females and three males are required.

Cages stocked with crabs are constructed to accept plastic
screens of uniform size. Algal laden screens are exchanged with
those in the cage after the crabs have fully grazed the algae
from them. These fully grazed screens are then removed from the
cage, scraped to remove undesirable algae and then rehung on the
float lines at the algal turf growing site. Approximately 10-12
days later, newly developed turf growth reaches a level at which
the screens may be exchanged back to a cage. Each crab cage has
three to five sets of plastic screens that are rotated. One set
of screens is in the cage and the other sets are in various
stages of growth. Therefore, with a regular schedule of screen
changes, the crabs in the cage are assured of a continual supply
of algae. The culture of algae is as important as that of the
crabs. To be certain of sufficient feeding it is desirable that
some algae remain in each screen when it is removed. If the
crabs are "overgrazing" relative to the fed algal turf then
either a larger number of screens, a shorter time in the cage or

a longer growing period is necessary. In general, it is
necessary to manage the algal turf carefully for optimum
results. A rigid schedule for screen rotation rather than random
selection or visual choice will usually provide maximum
production in the long run.

Breeding Stock

Female crabs can be collected using SCUBA or fish traps.
Coral reef spur and groove, large coral heads or patch reefs
where water depths are 20-100* are likely areas. Controlling
reproduction is not necessary, as spinosissimus spawn
throughout the year. Females bear eggs roughly every two months
and are egg-bearing for about 30 days. A typical female bears
four to five sets of eggs during her lifetime, although several
more are possible. We have found the average breed stock female
to bear 2-3 sets of eggs while in captivity. Colored plastic
tags facilitate the establishment of an individual code for each
crab. Tags can be attached to the legs.

To sustain the requirements of a small operation with 24
large growout cages and a hatch each month, approximately nine
mature female crabs are needed. These female crabs should be
maintained in several large cages rather than a single one
because sufficient feeding to all members of the brood stock can
be more easily monitored, making weather and loss of a single
cage less crucial. Females crabs should be checked/ inspected on

a regular basis, at least every five days.

Breed stock cages are the standard adult growout cages. A
smaller cage with dimensions of 90 X 78 X 35 cm (Figure 6) can be
used for a "female in waiting." This smaller cage is useful as
it provides for easier observation of ovigerous crabs. It is
constructed of a 2" X 2" wood frame covered with 2 mm screen and
fiberglass resined for strength on the wooden framing. Eyebolts
from both ends allow several cages to be strung together from one
anchor. The tops should be hinged in the center with a simple
latch mechanism installed. A crab operation that requires one
egg hatch per month should have several broodstock cages
containing two to six crabs each.

Special dietary care of Broodstock

Special care is required for successful broodstock
production. The female crabs should be fed ample quantities of
fully mature cultured algal turfs, as well as periodic
supplements of smaller macroalgae that are collected from reefs,
pavements or rock ridges. An algal monoculture diet for the crab
(i.e., a single species of algae grown in culture), while
possible, is extremely risky in terms of potential disease or
predation and also is likely to preclude the full reproductive
capacity of the species. Algal turfs, by the nature of their
community composition (i.e., algal species from nearly every
major algal group, forming a community of 30-60 algal species)

combine to form a highly nutritious and complete feed.

Breed stock care

Maintained in these cages with ample fresh food supplies, an
average female crab will produce two to three consecutive broods
in a 200-300 day period. Forty-seven percent of a captive crab
population (n = 26) produced three or more broods while in
captivity (Craig et al . . this volume). Each brood reguires three
to five weeks to undergo embryonic development. Approximately
60,000 eggs are released by a female (this varies with both the
female body size and age) . Approximately 70 gms (wet weight) of
algae per crab per day should be delivered to each cage. Old or
uneaten algae should be discarded at the time that the new algae
are put in. Thus, with nine female crabs, each producing two to
three broods over a year, one can be assured a hatch every month.

It is critical in this mariculture process to accurately
time an impending hatch. Egg brood coloration is the simplest
indicator of time to hatch (Table 2) . Microscopic examination of
individual eggs is a more accurate indicator. An egg brood
nearing release (one to three days pre-hatch) will appear clear-
amber in color. The individual prezoea will appear active inside
the egg and the regular pulsing of the animal circulatory system
will be readily apparent. The eyespots will appear predominant
while the presence of the brightly-colored yolk will be barely
discernible (Craig et al. , this volume).

Hatching

Ovigerous female crabs are placed in the fine meshed hatch
cages (Figure 7) , preferably within 24 hours before hatch. This
hatch box, with several bare screens, is placed in the water
several days before the female is introduced to allow a fine
algal (diatom) growth to begin development. Following hatch, the
female crab is removed from the cage. The larvae are then left
undisturbed for a period of 5-20 days post-hatch. It is then
generally desirable to transfer a portion of the crabs from the
hatch cage to another to reduce density. This can be done
without handling the very small crabs by simply transferring the
bare screens previously placed in the cage. After another 5-20
days, plastic screens containing algal turfs are periodically
exchanged to replenish the supply of algae to the young crabs.

It is crucial when the Mithrax crabs are very small to avoid
bringing in potential predators, especially small carnivorous
crabs, on the algal feed screens.

Rearing to 60 days and 10, mm CL (Figure 7 . 0 . 5 mm screen)

The rearing cages for crabs to 60 days post-hatch are small,
easily handled and labor intensive. The following operational
procedures should be observed for managing the rearing system
from several days post hatch to 60 days.

1) The post larval crabs should be thinned to separate cages
to a density of no more than 500 crabs/m2 of surface
area at 10-20 days post-hatch.

2) Algal screens should be exchanged every four to six days .
The grazed screens should be returned to the algal screen
float line for a re-growth period of 12-15 days. These
screens should be single-layered young screens, rich in
diatom growth, that have been in the water no longer than
two months .

3) Crabs reaching 10 mm CL should be transferred to the
intermediate cage.

Rearing 60-120 days ( to approximately 25 mm CL) (Figure 7 , 1 . 5-2
mm screen)

1) Algal feeding rate remains the same.

2) Crab density should be on the order of 50-100/m2.

3) Upon reaching a size of 20 mm CL crabs are transferred to
the adult growout cages. This can be done by gently
"flicking" the crabs off the algal feeding screens, where
they tend to congregate, or by individually picking them
out of the intermediate box where necessary.

Rearing of Juvenile Crabs to Harvest Figure 8 , 1/4-1/2" mesh

1) The growout cages are considerably longer (1.0 X 1.0 X
2.6 m) than the hatch and intermediate cages. Crab
density therefore is much less (100 when young and 40-50

when grown out) .

2) Algal feeding rate is greatly increased, with rates of
five to eight double-layered algal screens, every 3-4
days for 30-40 crabs - rate of feeding adjusted by noting
remaining algae on screens returning to the line.

Some turf should remain to avoid overgrazing.

3) Growout period for 250-400 days, at which point
harvestable crabs should attain a size of over 120 mm CL.

CAGE DESIGN, FABRICATION AND MANAGEMENT
The effects of the marine environment on cage, screen line
and anchor systems, designs and materials has been examined at
six coastal field sites across the northern and eastern
Caribbean. Due to the variability of marine conditions at each
field site, a unique set of designs and material specifications
for mariculture equipment suited to each site has evolved.
However, there are many characteristics that are common to all
sites and considerable effort has been made to standardize
equipment and techniques.

Three types of primary crab cages are used: hatch,
intermediate and growout. Small "female in waiting" cages can
facilitate the hatching process. The hatch and intermediate
cages, with the exception of mesh size, are identical. They are
constructed to accept the smaller screens, and their smaller

dimensions allow for rapid inspection, removal of material from
the cage and ease of handling. The growout cages are larger,
more crudely constructed and used to grow crabs from 20-25 mm CL
to market size. The three cages are designed to be easily
constructed and provide a manageable means of feeding algal turfs
to crabs.

The most important factor in the survival of very young
Mithrax spinas issimus in an "in situ" mariculture (to 100 days)
is the avoidance of predators. This is especially critical in
the early stages of growth, when the crabs are essentially
defenseless. The presence of competing invertebrate herbivores
is also undesireable . Careful cage assembly techniques play the
greatest role in prevention of the entrance of other organisms
into the cages.

The cages are relatively simple in design. They consist of
a frame constructed with wooden 2 x 4 "s and 2 x 2 "s nailed
together with 4" galvanized nails and covered with plastic mesh
screen. The lids are removable or hinged and the entire cage is
anchored. Slots inside the cage hold the algal turf screens in
place. Spaces between the slots allow for movement of crabs
between the screens.

Tools necessary to build the cages are common carpenters
tools including hammers, saws, pliers, screwdrivers and nails.

The cage should be assembled so that all joints are smooth and
without gaps. Once assembled, all finished edges of the cage

should be sanded smooth to insure a tight fit between the lid and
the attached screen. The wood frame should be carefully painted
with fiberglass resin to avoid numerous sharp points of hard
resin. Any gaps in the wood joints will provide hiding spots for
unwanted organisms and so these gaps should be filled.

It is particularly important that the cage lids are well
fitted to prevent the entry of unwanted animals into the cage, or
the cultured crabs out of the cage. Gaskets must also be fitted
on the removable lids of the hatch and intermediate cages. To
rigidly hold the lids to the frame on the hatch and intermediate
cages, bolts and nuts are necessary. On the larger growout
cages, latches should be used.

Hatch cage

The hatch cage is relatively small to allow for easy
handling, particularly with regard to the ability to haul it into
a small boat. However, it should be as large as possible for
maximum crab and screen space. We have generally used cages of
90 X 40 X 110 cm (Figure 7) . It is assembled from wood 2 x 2"s
nailed together. Assembly of this case requires special care to
insure that all joints are exact and tightly fitted so that there
are no open spaces between them. Slots or grooves are cut into
the inside of the cage frame members to hold four to eight, algal
turf screens. The wooden framework is then coated with
fiberglass resin before the cage is enclosed.

The water-tight frame is covered with a fine plastic screen
of 0.5 mm mesh. This prevents newly hatched zoea from escaping
and other organisms from entering. If the pore size is any
smaller , the flow is restricted unnecessarily. During cage
assembly, the enclosing screen mesh should be stretched tight
onto the wooden frame work, stapled and then resined directly to
the wood surface to form a complete seal of the screen to the
cage. After the screen has been resined to the frame, all the
gaps between the frame and the screen inside the cage should be
filled. When completed there should be no cavities in which
predator or competing organisms can hide.

The lid is a separate piece that is fitted and attached to
the cage frame by long threaded bolts and nuts. A gasket of
silicone plastic sealant is formed by compressing the uncured
silicone with the lid to form a complete seal of the cage when
the nuts are tightened on the lid. Silicone is used because it
retains its shape, does not shrink from the pressure of the
bolted lid and lasts for a long time under marine conditions.

To form the gasket, the cage lid is prevented from sticking to the
silicone by stretching SaranR wrap or similiar plastic over it
on the first clamping.

If meticulous care is taken and the cage is well
constructed, the screens will fit tightly and easily. There will
be no places for intruders to enter or hide and it will provide

years of service. Maximizing the number of slots per cage is •
important so that more screens can be added before handling the
screens or crabs that are already in the cage. We have used 24“

X 36" screens for this cage and typically have allowed two to
four slots. After the basic cage is finished, four eyebolts are
put in the bottom and ropes with clips are tied to make two
attachment lines. This design tends to reduce pitch and roll
when waves or currents affect the cages. This technique also
makes the cage pivot at the point of attachment to the anchoring
system in a seaway, rather than at the attachment on the cage.
Wooden blocks nailed to the cage ends serve as handles.

Hatch Procedures

The ovigerous female should be placed in the hatch cage one
to two days before release. She should be adequately fed up to
the time of her removal from the breed stock cage. Once in the
hatch cage, no algal turf food for the female crab should be
introduced, since it may also introduce predators. In any case,
the female does not eat on the night of the egg release.

Almost always, at dusk or into the evening, the brooding
female will actively release the eggs by flexing her
abdomen/pleopods . The entire procedure takes about three to four
hours. The spent female crab should be removed from the hatch
cage as soon as possible after hatching. The quality of the egg
release can be determined by taking a one liter water sample and

determining the percentage of viable zoea under the microscope.
The larval swimming stages last two to four days.

Viable larval crab densities in these hatch cages should be
on the order of 2000 crabs per cage to start. In the instance of
a typical size hatch, 2000 1st crabs, of about 1.5 mm carapace
length, will remain in the hatch cage at five days post-hatch.
From that point until 60 days post-hatch, mortality using these
methods will reduce the crab population to 400-600 crabs. Since
a nine crab breed stock will provide a hatch a month, this is far
more crabs than one fisherman could handle, if they were all to
be raised to adults.

At the time that the berried female crab (prior to hatching)
is placed in the cage, it is important that the heavy growth of
algae or sediment and detritus on and in the hatch cage screen
surface be removed. The larval stages need a high exchange rate
of sea water through the cage for both adequate supplies of
phytoplankton and to carry nitrogenous wastes out of the cage.

It has been found that the best way to do this is to remove the
hatch cage from the water before the gravid female is placed in
it. The hatch cage should be brushed, rinsed and dried to remove
all fouling algae from the screen surface both inside and out.

The cage should be placed back in the water about five days prior
to the gravid female's placement in it. This will allow a light
growth of diatom rich turf to develop, providing food for the
megalops and early crabs without blocking water flow.

Upon reaching the first crab stage, Mithrax spinosissimus
become entirely benthic and settle out on the screen surface.

They will consume the diatom turf which has colonized the cage
screen surface during the 8-10 days before they reach that state.
By about the 10th to 20th day, new cultivated algal turf screens
should be emplaced in the cage, as the original growth on the
cage itself will be exhausted.

The cage, constructed with four to eight slots should
receive two single-layered screens at first, two more, four days
later and then two additional screens on the next changing day
(remove the first two grazed screens at this time) . From that
point on, from two to four screens should be placed in the cage
at each screen changing date. These should all be singled-
layered screens. The grazed screens should be returned to an
algal screen line for new growth. Depending upon hatch density
and the rate of algal turf growth in a particular locality, hatch
cage screens should be changed every four to five days. Hatch
cage screens should be young screens, rich in diatom growth that
have had no more than eight to ten weeks in the water. After
that amount of time, they should be dried and brushed clean.

When algal turf screens are exchanged, the screens that have
been in the cage may have many small crabs still attached to the
screen, particularly at the earliest changes. Unfortunately,
hand removal or brushing increases mortality and is time

comsmning . Therefore, the screens should be left in the cage and
new screens put in the unused screen slots. If a dense hatch has
been achieved, some of these screens and their crabs can be
safely and quickly transferred to a second cage, that has been
tied up alongside the boat. This procedure will also thin the
density of crabs. Each cage should then be provided with a new
set of algal laden screens. If a hatch has not been dense enough
to justify a split, the older screens can be left in until the
next change, at which time most of the young crabs will have
moved to the new screens.

The algal turf screens are rich in a resident invertebrate
microfauna which can amount to 5-10% of the total screen biomass.
Most of these are very small and may well provide some food to
the crabs along with the algal turf. However, the larger of
these animals must be removed from the screen before it is placed
in a cage. It is imperative that exposure of the young M.
spinosissimus crabs to other invertebrate species be minimized.
These unwanted invertebrates may prey on or compete for food with
the young Mithrax spinosissimus crabs. Eliminating other larger
invertebrates can be accomplished by careful inspection, shaking
of the screens and if necessary by crushing the animals by
rolling a "rolling pin" across the screen surface several times.

After a hatch cage is used and the animals transferred to
the intermediate cage, it must be cleaned and reconditioned for
use. It should be brushed with a coarse brush inside and outside

at the same time, either when quite wet or when completely dry.
The cage should then be allowed to dry at least four days,
turning it over to be sure of thorough drying. If water has
seeped underneath the coating of resin or has saturated the
corners, the resin should be peeled off, the wood allowed to dry
and the resin replaced. The lid seal should also be checked and
redone, especially if drying has warped the top.

Intermediate cage

The intermediate cage design for the second phase is
identical for that of the first. Only the cage screen mesh size
is different. The mesh size is ideally 1. 5-2.0 mm. This
promotes a rapid exchange of "green" water through the cage to
clean the growing detritus load, while retaining crabs in the
cage. The screen must be attached in the same fashion, that is,
resined to the framework. Although perhaps not quite so
critical, the same care must also be given to the predator
control methods. The same gasket type and lid securement is
recommended .

Growout cage

Growout cages need to be considerably larger than the hatch
or intermediate cages; generally, we have constructed our units
with dimensions of 2.4 X 1.0 X 1.0 m. These are also constructed
of 2" X 4 " s , using the same polyester resin coating. While some

care should be taken in cage construction, so that it is able to
withstand a sea conditions for several years, joint and screen
tightness is not so crucial (Figure 8) . These cages hold 20
screens which are oriented vertically and are "guided" by slots
cut into the cage structure. Approximately 5" of space between
each screen allows the growing juvenile crabs ample feeding
space. There is a space (minimum of six inches) under the
screens to allow the crabs an area to move from screen to screen.
No space is allowed on top of the screens, since the lid is
designed to hold the screens tight and to keep them from moving.
In the bottom of the cage, "molt" compartments of 1/4" plastic
mesh, of 35-90 mm size, are constructed for the crabs.

In the growout cage a double plywood top is used. This
provides some darkness and protection from wave chop. It is
hinged at the center and opens to one side at a time for screen
exchange. There is also a small door cut in one side of the top
for easy placement of macro algae supplemental feed. Heavy rope
( >3/8 "polypropylene) is used for this and rubber strips from used
tires serve as hinges.

Because the cages remain in the water for long periods, they
are given three coats of fiberglass resin. The cage corners
should be reinforced with fiberglass cloth as well. To be able
to stock a 20-25 mm CL crab, a mesh size of 1/4-1/2" or less is
required. One-quarter inch mesh is ideal. The plastic screening
should be attached to the cage with fence nails, then resined and

covered with wood or plastic strips for added protection.

Cage Placement

Water flow and quality is the most important condition to
consider. Sediments falling from screens, as well as feces, dead
crabs, and detritus from inefficient feeding of Mithrax crabs,
and other inhabitants, potentially contribute to poor water
quality. On the other hand, particularly for hatching crabs, too
much flow presses zoea to the screen inside the cage and
increases cage maintenance requirements due to racking of the
cage structure and chafing of the lines.

In most localities with good flow and wave chop for adult
cages and algal screens, nearby areas in the lee of a reef
structure or cay can be found to provide the slight additional
protection needed for hatch cages. Where this has not been
possible, such as at the Carriacou research site, we have
achieved the needed hatch cage conditions by submerging them.
However, this requires either a diver to enter the water and
release them from their anchorage or a system to lower and raise
the box.

Anchoring systems

The smaller cages may be anchored either to a chain
suspended between two anchors, or anchored individually. The
growout cages should be anchored individually, and situated so

that the current runs parallel to the screens in the cage. The
four lines attached to the cage bolts should tie off to a single
heavy line 3/8" or greater, which in turn ties off to a chain
from the anchor. The chain is shackled to the anchor. All rope
connections are made with hose to reduce line chafe. The line
and chain length to depth ratio should be seven to one, or more
under particularly turbulent situations.

The water depth in which the cage is placed is critical. A
depth that is too shallow increases the effect of waves and
swell, which causes the cage to hit bottom, thus exposing the
crabs to excessive sand and debris. If current and wave motion
are particularly strong, cages should be attached with shackles
or 3/8" rope using heavy rubber tubing to prevent chafing from
the metal eyebolts on the cage. Alternately, chain can be used
throughout, although this greatly increases cost. At the
Carriacou research site, where the current was very strong, 1/4"
chain was used with six inch longline clips and 1/4" shackles and
a chain stretched between two 30 lb. Danforth anchors served as
the bottom attachment. Weights were necessary along the chain
between cages since the buoyancy of the cages and the strength of
the current pulled excessively on the bottom chain. In
situations of strong daily reversing tide currents, it is
necessary to run the line on the bottom parallel to the current
so that opposite anchors will hold alternately as the tide
switches .

DISCUSSION

Algal Turf Cultivation

Algal turf growth on plastic screens develops in three
distinct phases. The artifical substratum on which the algae
grows should be scraped 10 days after being put in the water, and
every 10-12 days thereafter. It takes about four to eight weeks
for a screen to develop a mature algal turf community. Diatoms
and their mucilage sheaths appear in the first week. Shortly
before the second scraping (at about 20 days) , blue-green algae
dominates and appears as soft, sometimes mucilagenous brownish
tufted filaments up to two centimeters long. After about three
weeks of cultivation, the other algal types, primarily red algae,
begin to colonize the screen. If the plastic screen surface is
not scraped during this initial period of turf cultivation, the
turf will not develop as rapidly or fully. Also, the community
may have a disproportionate macroalgal component and gradually
develop sponge and calcareous animals that are not palatable to
the crabs. Only through a continual scraping/grazing schedule
will a high diversity of rapidly growing turf algal growth be
established and maintained.

Screen designs and screen fabrication may be tailored to a
particular site or to accomodate available materials. Whichever
type and size is chosen, it is important that all the screens,

for the intermediate and adult cages, are exactly the same
dimensions. The algal turf screen mesh size should be no larger
than three millimeters and no smaller than one millimeter. We
have found a 2 X 3 mm black polypropylene plastic screening
material to be adequate. This screen is able to withstand
continued scraping from both the initial algal turf cultivation
techniques and the crabs. It has been demonstrated that a
transparent or translucent screen material significantly
increases production. However, a testing program to identify a
transparent screen that also has the right mesh and toughness has
not been carried out.

For hatch cages, single-layered screens should be used to
reduce hiding spaces for predators. For intermediate and adult
cages, algal growth is considerably increased by constructing the
algal turf screen of two layers of screen material which provides
a greater surface area for the algal holdfasts. In addition,
optimum algal production is achieved when the plastic screen is
rigidly suspended in the water column. A vertical, two point
mounting system (Figure 10) optimizes algal production and is
easy to handle. The screen frame, constructed of 3/4" PVC pipe
is non-corrosive, as well as non-toxic. Filling the lower PVC
pipe with sand provides greater rigidity against wave surge and
currents. When determining a size for the screen frame, the wet
weight of the algae growing on the screen should be considered.

A screen measuring two feet by three feet can weigh up to 25

pounds with substantial algal growth; therefore the total surface
area of the plastic screen should be kept below one meter. We
have used "schedule 40" 3/4" PVC with standard sleeve-type
elbows, carefully glued with PVC cement. The plastic screen is
wrapped around the ends of the PVC screen frame and tied to it
with 60 lb test monofilament fishing line or attached by plastic
cable ties. The other two edges of the screen which lay inside
the pipe framework are stitched together to prevent tearing.

This simply constructed screen and frame is then fitted with 1/4"
polypropolene line which extends 12" from each corner of one end
of the PVC screen frame. A piece of rubber hose is fitted on the
rope so as to prevent the chafing of the rope on the PVC pipe
edge. A simple, strong knot such as the "fishermen's" knot
should be used, so that the rope loop tightens around the pipe.
Due to their air/water tight construction, weight and buoyancy,
the screens hang vertically from the water surface with the sand
filled pipe weighing the screen down at one end.

Twenty to thirty algal screens typically have been placed on
double anchored lines perpendicular to current and wave surge.
This "screen" line has regularly spaced tied loops (Figure 11) .
The length of the screen determines the distance between loops.
The distance between screens is approximately one foot. The
screens are then simply tied to the "screen lines" with their tie
lines. They should be tied as close as possible to the loops to

keep them from bouncing into each other. A float at either end
of the line marks and supports the ends of the screen line.

Anchor lines extend from the floats (Figure 11) . The major
benefit of this system is its mobility. If the area chosen does
not produce expected amounts of algae, or in the advent of a
major storm, the line can easily be moved to a better location.

Larval and Post-Larval Rearing and Growout

The 0-60 day or hatching phase is the most critical in the
crab's life. The crabs go through metamorphosis from a
planktonic zoea to a benthic crab stage and then through several
molts. Growth and survival rates are related to initial hatch
quality, predation, water quality and flow, food type and
availability and handling.

The viability of an egg brood may be discerned by
examination of a few eggs prior to hatching. Viable eggs show
active larvae with "heart beats" up to 250 bpm and occasional
flexing of the abdomen. If the eggs appear cloudy or grey they
are usually dead or are of a poor quality and survivability will
be reduced.

As the eggs develop, they change colors. As discussed
above, examining the eggs as they are brooded can determine just
how close they are to hatching and when to put the female into
the hatch cage. It is important to minimize the length of time
the female is left in the hatch cage, because she is not fed at

this time. Eggs may be removed from the brooding crab using
forceps and placed in a vial with seawater for later examination.

A female approaching a hatch should be well fed prior to
putting her into the hatch cage so that food does not need to be
added during her stay there. It is also important that the female
be examined for attached predators that could hide in the
crevices in her skeleton or among the pleopods or eggs. When
transporting the female from the box where she lived to the hatch
cage, styrofoam coolers of fresh seawater are ideal.

If the female crab must stay in the hatch cage for more than
three days, feedings of macroalgae should be made. However, any
such feedings (usually a pressed handful) must be thoroughly
checked for invertebrates. Females previously well fed have
lasted at least 10 days and have produced successful hatches.
Females that have released only part of their hatch can be kept
for a second day in the hatch cage, but usually those remaining
are not very viable.

After the female has been put in the hatching box with the
top sealed and the cage attached, she will eventually position
herself in a way that enables her to pump her abdomen which will
aerate and eventually liberate the hatched zoea over a two to
twelve hour period. On the day after the female was expected to
release the eggs the cage should be checked in the late morning
or early afternoon. Zoea can be seen through the screen that has

begun to accumulate diatoms and sediment. Occasionally, if the
female is checked early enough, empty egg cases are left on the
pleopods and can be mistaken for unhatched larvae. With a good
hatch, many zoea can be easily observed to be swimming about the
cage. If a cage is raised and brought to a boat for examination,
care should be taken to move it slowly so as to not drive the
zoea against the cage screen. It should not be fully lifted from
the water, or if absolutely necessary, lifted very slowly for the
same reason.

Mithrax spinosissimus starts as a prezoea and goes through
two zoeal stages and a megalops stage before metamorphosing into
a crab. These stages to first crab last about five days. When
examining the contents of the hatching boxes during the
planktonic stages one should see larval crabs swimming as well as
a mixture of live and dead zoea and empty molts on the cage
floor. Often, when removing the female from the cage after
hatching, thousands of zoea can be seen on the cage bottom, but
this phenomenon doesn't necessarily indicate poor survival.

Living zoea may be distinguished by their red eyes, while dead
ones and molts have white ones.

The first crab stage is entirely benthic and negatively
phototropic. When settled the first crabs space themselves
evenly on all of the inner surfaces of the cage, including the
top, for the first thirty days. Generally the mortality to the
first crab stage from an average brood of 60,000 is 95-97%; the

range is dependent on brood fertility, viability and water
characteristics. From first crab to 60 days post-hatch, normal
mortality will reduce the crab population to 400-600 crabs. The
diatoms and other algal spores settling on the cage screen
surface provides initial food to the post-larval crabs. But this
same settling algae also reduces water flow.

After settling, the crabs will molt at about four, eight and
twelve days post-hatch, at which time they have begun to deplete
the algae on the inside of the cage. The average growth rates
through this 60 day period vary from site to site and seasonally
from 0.11 to 0.30 mm CL/day so that the average CL varies from 6
to 12 mm at 60 days. It is not uncommon to see dead crabs at
various stages, and it is very common to see molt shells
cluttering the bottom of the cage as well as feces and algal
debris. Later cage designs have included a valved shute and well
at the bottom of the cage to drain off debris with minimum crab
loss .

When the algal layer growing on the outside of the cage
begins to slough off and it is obvious from looking on the inside
of the cage that the algae there has been eaten, then screen
feeding should begin. Screen feeding should be initiated at 10-
20 days post-hatch. Food preferences appear to be benthic
diatoms after settling at megalops, with an increasing emphasis
on blue-green algae and by 60 days the red, brown and green algae

that dominate a mature algal turf community. Different
strategies of feeding have been employed in the first 45 days
post-hatch. Ideally, minimum disturbance and handling affords
maximum survival, but on the other hand the crabs must be fed.
Strategies have ranged from putting one single layer screen into
the cage on the day the female was removed after hatching to
waiting for 80 days before the first screens were added. Single
layer screens added the first day also supply more settling space
for first crabs. Generally, screens were first added at 14-30
days. Cages with many screen slots are most desirable since
screens can be added for sometime before any are removed, or at
least allowed to remain in the cage until most of the small crabs
had migrated to newer more algae-laden screens. When screens are
added prior to 30 days, they should be assembled with a single
layer of screen. This reduces the chance of adding small
predators in the folds of the screen.

In the standard hatch cage, after the initial screen or two,
screens should be added every four to seven days up to 30 days.
These should be primarily colonized by a diatom growth that takes
5-10 days to become established on a new screen. After 30 days
and up to 60 days post-hatch, two or more screens should be added
on a regular basis of three to five days.

As new screens are added, the older ones are examined for
algal growth and the presence of Mithrax crabs. If they have
neither, they are removed. Eventually, it will be necessary to

remove crabs from the screens being removed. If the screens are
single layered, a gentle shake or tapping will remove them
easily, especially if no algae is remaining on the screen. If
the screens are double layered, holding the screen out of the
water for a few seconds then tapping the screen will usually
cause the crabs to run towards the bottom of the screen where
they will jump off or can be easily removed by hand;
alternatively, the screen can be gently agitated in and out of
the water to remove them. A cleaner screen affords easy removal
by tapping, but a screen with algae still growing on it affords
easier removal by hand, since the crabs are not firmly attached.

Another strategy for minimum handling, especially if a hatch
is particularly large, is to transfer crab-laden screens from the
original hatch cage to an unused cage so that more algae-laden
screens can be added to both cages. This technique is also
useful if the first cage becomes clogged or too detrital laden.

Intermediate phase

At 60 days post-hatch, the crabs in the hatch cage are
transferred to the intermediate cage. This cage (25.5" X 27.5" X
40.5") is identical in design and size to the hatch cage.

However, it is covered with a wider screen mesh size of 1.5-2 mm.
This greater mesh size allows detritus to fall out of the cage
and maintains a high water quality state in the cage. The
intermediate cage should be anchored and, if waves and currents

are strong, submerged in a manner similar to that of the hatch
cages. Algal screens are exchanged into and out of these cages
every 3-4 days.

These intermediate cages are stocked with no more than 600
sixty day post-hatch crabs. If the number of crabs surviving
from a hatch at 60 days post-hatch is greater than 600, the
excess above 600 should be placed in a separate cage. At about
120 days, the crabs should be transferred to the growout cages.

A typical mature intermediate cage will yield 100-300 crabs at
this point in time, each approximately 25 mm CL. Management of
the feeding requires changing algal screens on a routine basis.
The intermediate cage screens require a 12-15 day recovery period
for optimum algal growth. There should be equal numbers of
intermediate and hatch cages.

We have found that survival rates are less and growth rates
are slower for crabs left in the original cage for more than 60
days. Crabs greater than 8 mm are large enough to withstand the
stress of transfer handling, with only about a 5% population
loss. The greatest numbers we have experienced at 60 days is 600
crabs of 10-12 mm. When these crabs molt again, the
proportionately large increase in body mass seems to increase the
demand for food to a point that exceeds the capacity of the 0.3
cubic meter cage and of the feeding regime (up to 8 screens per
week) . The subsequent large increase in fecal material, dead

crabs, molt and algal debris and the established populations of
other organisms, including predators and competitors seems to
contribute as well to unfavorable conditions and increased
mortality .

Thus, it is necessary to decide to either split the hatch
using unused cages similar to the hatch cages or transfer all the
crabs to clean intermediate cages with larger mesh size. Either
method will work, but a larger mesh size increases water and
waste exchange potentially improving growth rate and survival.

If the cages are examined carefully at each feeding, predators
are not a problem in the intermediate cages of slightly larger
mesh sizes. No more than 400 crabs should be placed in each cage
of the second phase. Three hundred seems to be ideal. During
the next 60 days, after a split hatch, losses of crabs
transferred would approach 25-35% so that by 100-120 days post¬
hatch, 250-300 (20-25 mm) crabs would be left. Another viable
strategy, if growth rates are exceptionally high, involves moving
only the largest crabs (25 mm CL or more) from the hatch cage
directly to the growout cage. Smaller remaining crabs tend to
show increased growth rates as competition for food and space is
decreased .

In this second phase, algae screens should be changed twice
weekly regardless of their condition. These crabs become very
selective and sometimes a screen with the appearance of good
growth will not be touched or will have been picked over.

Screens on which with bluegreen algae are dominant satisfy most
of the intermediate cage crabs for approximately the first three
weeks; after that, fully developed algal-turf screens are
desirable.

Initially, the screens are placed in the cage four at a time
in an six-eight slot cage, so that each set remains for a one
week period. If at each change, all the screens are cleaned of
growth, then as many screens as possible are changed. By the
time the crabs reach 25 mm CL in size, they are able to defend
themselves against most small predators that might inadvertantly
enter the cage and are able to withstand handling.

Crabs of 20-25 mm or greater, should be transferred into a
growout cage. One to two hundred crabs may be transferred to
each growout cage. We have found that as the 20 mm CL crabs are
removed from the hatch cage and intermediate cages, the remaining
crabs are under less competition and reach 25 mm CL in 10-30
days. The intermediate and hatch cages should be removed from
the water, cleaned and dried if they have been in the water for
more than 80 days.

Growout Phase

The third and final phase involves only variations of
previously discussed techniques. The emphasis is on strategies
of maximum feeding with attention to increasing varieties of
algae fed. Less attention needs to be paid to marine life that

may enter the cage. However, on occasion, we have discovered an
octopus in our cages. These predators raise havoc with the young
crabs .

Screens added to the growout cage are exchanged every rwo to
three days. Screens must be scraped after each feeding and
returned to the screen line. At times the screens will need to
be cleaned, dried and restarted. It seems clear that a source of
older juvenile mortality is insufficient feeding. This is best
judged by the amount of algae remaining on the screens at the
time of removal. If the screens are nearly bare, algal feeding
is probably insufficient.

Crabs in the adult growout cages molt another 7-8 times
until they reach sexual maturity in about a year. The actual
time to maturity can be determined by the growth rate. As the
crabs grow they become very easy to handle with little or no risk
of damage. Heavy wave action can apparently damage molting
crabs, and we are now experimenting with screen annd pipe
habitats to provide molting crabs protection against aggressive
cage mates.

Wild crabs should not be placed with cultured crabs, since
wild ones are often aggressive and may cause mortalities. Mixing
sizes even in cultured crabs also may be risky. Because
predators do establish themselves and the polyester resin does
not keep out all the shipworms, the cages themselves need to be

rotated. A cage frame that is pulled out of the water every 12
months, re-resined and repaired should last for at least three
years. The effects of hurricanes may be severe if adequate steps
are not taken prior to a storm. Important considerations are
anchoring methods as well as location and maintenance of gear.
During periods of stormy weather, cages may be anchored so as to
be submerged beneath the sea surface.

Three of our field sites have been in the direct paths of
hurricanes with wind speeds up to 70 knots. The first experience
met with cage and crab losses of nearly 30-40%. Subsequent
experiences have resulted in minimal losses. These losses have
primarily been amongst individual crabs themselves. The design
of the cages prohibited the crabs from getting a strong foothold
in the cage resulting in many crabs being throwwn about the cages
by the motion of the water. Recent cage modifications are
expected to prevent these types of losses in the future.

CONCLUSIONS

The mariculture described in this volume was specifically
developed in an attempt to utilize a newly found base resource
(cultured algal turf) for a local fishermen mariculture. The
possiblility of growing Mithrax spinosissimus economically in a
sophisticated hatchery and factory growout situation may very
well exist. However, we have specifically avoided approaches

that seemed unlikely to be adaptable to the itinerant fisherman.

We have repeatedly demonstrated in the Caribbean lagoon
environment of the West Indian fisherman that Mithrax
spinosissimus can be hatched in inexpensive cages "in situ."

After working with the itinerant fishermen of Buen Hombre on the
northwest coast of the Dominican Republic for nearly a year with
this mariculture , we feel confident that these totally uneducated
and isolated fishermen are quite capable of carrying out the
hatching process as well as the remainder of the mariculture.
Indeed after the project was officially closed at this site, the
fishermen continued to work with our remaining stock and brought
a number of animals to maturity.

Regardless of the likely economics of a Mithrax commercial
market, a harvest of 1000 adult crabs per year would provide a
significant economic boost, or even a total livelyhood for many
Caribbean fishermen. Furthermore, this is an environmentally
non-destructive process, and with reasonable controls on the
adult populations, it is unlikely to be self-destructive. By
keeping a breed stock of no more than 20-30 adult crabs in
several cages, a hatch rate of from one per week to one per month
can be easily achieved. We have accomplished this at all project
sites. The keeping of this minimum breed stock is no more than
5% of the total work effort.

Once the vagaries of dealing with wave destruction of the
cages and predator control were managed, our results have

repeatedly shown that 200-300 crabs per hatch can be raised to
120 days or about 25 mm CL in size. Thus, we have securely
demonstrated that in the itinerant fisherman environment, it is
reasonably possible to bring roughly 2000-8000 crabs per year,
depending upon the number desired, to 120 days of age. Our
success at 120 days has continued to improve with time, and it
would appear that with the same basic methods, 500-600 crabs per
hatch can be routinely achieved at 120 days post-hatch.

If adult growout from 120 days to 12-15 months
(approximately 9-12 months in an adult growout cage) can achieve
a 30-50% survivorship, the entire algal turf/crab culture method
described in this paper would be technially feasible.

To date, we have not achieved this kind of survivorship.
Unfortunately, it is very difficult to separate the biology and
technology of the process from the very significant logistical
and political problems that have beset the project. At our
longest running research site, Buen Hombre, research funding has
been repeatedly cut off for periods of up to two months, making
significant progress extremely difficult. Each site experienced
similar or even more difficult problems not related to the basic
biological problem of reducing late growout mortality.

At this point, late growout survivorship, despite obvious
nonbiological problems, has averaged 18% per 100 days. During
the late stages of cage growout, increased mortalities apparently

occur at periods of molting. Male/female losses are more or les
equal. Intercrab competition may be important during post molt
periods, though no extensive evidence of such has been
consistently observed. Ryther et al . . (1987) observed

significant dismemberment and death of some animals by "dominant"
crabs. However, rarely does cannabalism appear to be a factor.
Rough water at molt obviously causes mortalities and modified
cage design may greatly increase survivorship. We have developed
a new flow-thru sea water system for observing crabs in cage
environment and have begun a redesign process for cage
microhabitats. Disease may be a possibility for juvenile
mortality, though dead animals have been dissected and obvious
symptoms or disease sites have not been located. Most important,
survivorship has tended to improve throughout the project,
suggesting that the answer to a successful itinerant fishermen
mariculture utilizing algal turfs lies in continuous improvement
of techniques and apparatus.

During the last six months of the project, overall growth
rates of about 0.20 mm/day have routinely been achieved to about
100 days, and a mean rate exceeding 0.41 mm/day for potentially a
400 day total growout was achieved at both Buen Hombre and
Antigua. Thus, Mithrax spinosissimus mariculture is not likely
to be limited by crab growth rates, but rather by survivorship
and economics.

We have currently begun an effort at Grand Turk in the Turks

and Caicos Islands to concentrate on the problems of late growout
mortality. This is certainly one of the best growout sites, so
that with a more secure logistic/political situation, the
techniques to bring this mariculture to a fully successful
conclusion biologically can be developed.

Economically, several hurdles will have to be overcome.

While crab sales to local hotels and restaurants would
undoubtedly provide some minimum income to the

fisherman/entrepreneur, a high volume mariculture would need to
develop an export market to North American and perhaps European
cities. However, as long as a sufficient source were available,
namely enough fishermen working and coordinated in a well-defined
market structure, such a market can probably developed fairly
easily. Finally, it is necessary for the cost of cage and screen
construction to be reduced. This is not an unreasonable goal,
and given adequate volume, molded or stamped plastic cages and
screens could likely be made cheaply in large numbers. Thus,
following biological success, it would be necessary to coordinate
development of the supply and marketing process.

LITERATURE CITED

Adey, W.H. 1983. The microcosm: A new tool for reef research.
Coral Reefs, 1: 193-201.

Adey, W.H. 1987. Food productionn in low-nutrient seas.
Bioscience. 37(5): 340-348.

Adey, W.H. and T. Goertemiller . 1987. Coral reef algal turfs;

master producers ini nutrient poor seas. Phycologia (1987)
26(3) : 374-386.

Bohnsack, J.L. 1976. The spider crab, Mithrax spinosissimus :
an investigation including commercial aspects. Florida
Scientist. 39(4): 259-266.

Brown, L. 1985. State of the World. Maintaining world
Fisheries. Norton. 301 pp.

Coen, L. D. 1987. PhD. Dissertation: Plant-animal interactions:
ecology and comparative functional morphology of plant -
grazing decapod (Brachyura) crustaceans.

Hazlett, B . A . and D. Rittschof. 1975. Daily movements and home
range in Mithrax spinosissimus (Majidae, Decapodae) Mar.
Behav. Physiol. 3: 101-118.

Munro, J.L. 1974. The bilogy, ecology, exploitation and

management of Caribbean Reef Fishes. Part V. The biology,
ecology and bionomics of Caribbean Reef Fishes: Crustaceans
(spring lobster and crabs) . Res. Rept. Zool. Dept. Univ.
West Indies 3(6): 39-48.

Peyton, K. , K. Moller, and W.H. Adey. (In review) Community
structure, development and biomass production of algal
turfs grown on artificial substrates in an oligotrophic sea.
Botanica Marina.

Pillay, T.V.R. 1985. Some recent trends in aquaculture

development. In: Status and Prospects on Aquaculture
Worldwide (Proc. from Aquanor 85, Trondheim, Norway.

Ryther, J . , R. Winfree, J. Holt, r. Creswell, W. Lellis, J.
Chaiton, C. Kovach and F. Prahl . 1987. Antigua Crab

Mariculture, Annual Progress Report. Harbor Branch
Oceanographic Institution, Fort Pierce, Fla; July 15, 1987.
78 pp .

Simon, J and H. Kahn. (Eds.). The resourceful Earth. Blackwell.
585 pp.

Tangley, L. 1985. And live from the East Coast - a miniature
Maine ecosystem. Bioscience 35: 618-619.

LEGENDS

FIGURE 1.

Screen line commonly used for growing algal turfs in
reef/lagoon environments in trade wind seas.

FIGURE 2.

Male Mithrax spinosissimus . 10-20 days after final
molt and at 1-2.5 kgs, an ideal siize for harvest.

FIGURE 3.

Map of eastern Caribbean showing mariculture work
sites for caae culture of Mithrax spinosissimus .

FIGURE 4.

Relationship between size (CL) and live weight of
Mithrax spiinosissimus .

FIGURE 5.

Larval stages showing prezoea, first zoea , second
zoea and megalops. (Photos to be added later) .

FIGURE 6.

Cage for observing females approaching release.

FIGURE 7.

Hatch cage (0.5 mm screen) and intermediate cage
(1.5-2 mm screen).

FIGURE 8.

Growout cage ( 1/4-1/2" mesh).

FIGURE 9.

Detail of algal screen assembly.

FIGURE 10.

Screen line.

FIGURE 11.

"Mushroom" type anchor design, fabrication and set
up .

TABLE 1.

Size and fucunditv of Mithrax spinosissimus

TABLE 2.

Assumptions to derive cage number and algal turf
screens for production level of x crabs/15 months.

TABLE 3.

Essential requirement of hatch cage rearing.

■

Figure 1

CENTIMETERS
INCHES

Figure 2

STUDY FIELD SITES

Figure 3

180

160

140

120

100

o MALE

• FEMALE 2 ,

o °o

500

1000 1500 2000

LIVE WEIGHT IN GRAMS

2500

1 - 1 T I

3000

Figure 4

Figure 6

SIDE VIEW WITH LID

END VIEW WITH LID

I - .35 m

TOP VIEW

Figure 7

SIDE VIEW END VIEW

Figure 8

2.6 m

Figure 9 & 10

Figure 11

TABLE 1

SIZE AND FECUNDITY OF SELECTED FEMALE M. SPINOSISSIMUS

Total Weight Weight Weight of Crab: Total Number

CL of of Weight of Egg of

Length Crabs Egg Mass mass Eggs

(mm) (gins) (gins)

105

550

13 : 1

42,000

115

800

16:1

51,000

125

1090

19:1

60,000

135

1400

21:1

69,000

*Xnterpolated from Data in Craig, et al . , this volume.

•

TABLE 2.

BASIC BIOLOGICAL CHARACTERISTICS TO DERIVE CAGE NUMBER AND ALGAL

TURF SCREEN SETS FOR PRODUCTION LEVEL OF 50+ CRABS PER CAGE PER

15 MONTHS

1. It is a dual system of crab culture and algal turf culture

2. Average female crab produces 60 , 000 eggs per hatch.

3. Average time between hatches for any one female crab is 60
days .

4. Average female will produce 3-4 broods while in captivity.

5. The initial density of larval crabs is 60 , 000/sq. m of cage
screen surface which is equivalent to 0.2 zoea/cc of cage
volume.

6. The rate of survival of larvae is 3.3% (to 5 days post¬
hatch.

7. The rate of stocking of post-larvae (5-20 days post-hatch) is
500 - 800 crabs/sq. m of cage volume.

8. The rate of survival to 6-10 mm (CL) is 30% (approximately 60
days post-hatch) .

9. Algal turf screen sets per hatch cage,, number 4 sets of 4
single-layer screens each. Algal feeding rate should be
greater than 0.15 gm (dry) /crab/4 day (assuming each screen set
is exchanged into the cage for grazing for 4 days and thus is
in the lagoon proper undergoing algal regrowth for 12 days) .

An area of rough , flowing water for growing ample quantities
of cultured algal turfs.

10. Young crabs of 10 ran CL or larger are transferred to an
intermediate cage of 1-2 mm size mesh,

11. The feeding rate for post-larval crabs of 10 mm CL to 25 mm
CL cultured in the intermediate cage is increased to 5-10
gins (dry) /crab/day

12. The rate of survival of rearing young crabs from 10 mm to 25
mm CL is 20-40%

13. Upon attaining 25 mm CL , the crabs are transferred to the
growout cages at a stocking density of 8-15 crabs/sq. m. of
cage surface area.

14. The number of algal screens required is 60-80, which then
comprise a total of 3-4 sets that are periodically exchanged
into the cage every 4-6 days.

15. A crab is considered marketable after attaining a size of
120 - 140 mm CL and a weight of 0.8-1. 5 kg.

16. The yield of marketable crabs per cage is 50+ at a period of
15 months post-hatch.

TABLE 3

ESSENTIAL REQUIREMENTS OF HATCH CAGE REARING

1) Breedstock crabs in adult cages, fed meticuously .

2) A reasonably calm, well flushed lagoonal area.

3) Hatch cages covered with plastic screening of 0.5 mm mesh
openings and meticuously constructed so as to keep
unwanted marine life from entering. Calm water or sunken
cage.

4) Cultured algal turfs in ample quantities to feed the post-
larval crabs up to 60-80 days post-hatch

5) The hatch cage screen mesh should be scrubbed, flushed and
rinsed of all algal/detrital material that clogs the screen
mesh from previous usage

6) The hatch cage seal should be reinspected to insure its
integrity

7) The cage anchoring location should be shallow (10-15*) calm
or protected, and situated so as to receive clear, highly
oxygenated water of 25 * C, 34-38 ppt salinity. The shallow
pavement area of a backreef is ideal

8) The hatch cage should be anchored 4-5 days prior to

the expected hatch to provide the right diatom-rich screen
and reduce the potential for presence of predators

] ll\ I B RAR I ES^SMITHSONIAN^INSTITUTION^NOlinillSN^NVINOSHilWS^Sa I 3 V d a I^L I B R A R I E S^SMITHSONIAN

5 <o „ 5 ^ ^ c° . _ ^ z \ „ _ ^ ^ co

‘ co u w /£223e&\ ^ co V ^

;*2SKS|\

2 < (Si

Tjyw ~ 07^# - > - - t'/wr. «-.»-# rv- v««rw . * .'>. W O' \^\

O X&us*^

2 _j 2 -J ^ -J 2 _

ON NOlinillSNI NVIN0SH1IIAIS S3l3V3ai1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1MS
z r - z r- z

m " v^missx m x? ^xz z; m xzicx^ " nwu,^x m

co \ £ CO £ co £ co

111 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1I/JS SBlUVHaH LIBRARIES SMITHSONIAN

Z . , .CO z .v' co z _ _ _ co z - , . co

< Zx\ . 2 <f .vVN . e < zTvtsoZx 2 < a\ .. 2

CO * Z - 00 ■ z CO Z CO V z

ON NOlinillSNI NVIN0SH1IWS S3IHV3an LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS

co — co iz co ^ CO —

O x _ z ^

nJU B RAR I ES^ SMITHSONIAN- 'INSTITUTION NOlinillSNI ^NVINOSHIIWS SaidVaail^LIBRARI ES^ SMITHSONIAN
r~ z r- z (— , z r* Z

CO
30
>

m co * m ^ ^x^a^Z^ m ~ m

co £ co £ co £ co _

on NoiiniiiSNi NviNosHims S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws

co z CO z s z V co z

1 I (sgjl 1 i JPife 1

y j 1# § 9 ^ 1

o „ ,.v

£ * *‘S / ' •_

> • 2 X'ftosqsz > 2 ' > x^vA^r^' 2 ’’ xn5^’ > ' 2

3 n_LI BRAR I ES^SMITHSONIAN JNSTITUTION ^ NOlinillSNI _ NVINOSHIIWS^ S3 IBVHail^LIBRARI ES^SMITHSONIAN

Ha H

ION NOlinillSNI NVIN0SH1MS S3l3VHan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS

o / r- \-[/<i ^

' m 'CiJcoSz' m " xicuus^z rn

co t: co £ co £ co

BIT LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1I1NS SBIBVaan LIBRARIES SMITHSONIAN

z .... co 2 co z _ _ _ co z .... co

— AS vo 2 .< 2 .<

t i «, s ■■'t'/y* &

Jf% i€m i

v > I x > i 2 ^ > N®I

CO V Z - 00 *■„ z CO Z CO * Z

ION NOlinillSNI NVIN0SH1IIMS S3ldVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS

CO — CO — CO ~ _ CO

o 0<^z _ '^7'r o _ _ _

311 L I B R A R I E S~” SMITHSON IAN-* INSTITUTION NOlinillSNI ^NVINOSHIIWS S 3 I a V a 3 IT L I B R A R I E S^SMITHSONIAN
r- z z z r~ Z

O .edlfos VqX CD ZoaVtoDJx n w O Zv^T,r(7/X “ ZoTJi\ O

CD
X)

o/ PO w\

_ j4y' — vv^cjlipS^y

m x^os^z ^ rn x^vq^jz z; x^va^>^z m >vw ^ X^vas^z m x^osiiyz

co £ co £ co £ co

ION NOlinillSNI NVIN0SH1IINS S3IHVaai1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IIMS

UTION ^ NOlinillSNI ^NVINOSHIIWS^S 3 I dVaan2LIBRARI ES^SMITHSONIAN^INSTITUTION ^ NOlinillSNI _NVINOSHl

CO _ \ 00 - _ _ CO -3> _ 00 , _ ^ 2 ’\

/Bail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IWS S3IBVdan LIBRARIES SMITHSOI'
z f~ _ _ _ z r* z

O “

' ~ ^ - V\ — CO

o >

^ & *>

^ m rn “

tO £ CO £ CO ± CO

'UTION NOlinillSNI NVIN0SH1IIMS SBIHVtiQn LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH.

i .i 1 ,1 4Sk I i I .1 ^ :

^ § XCLHic-'' > <§;?** 2 X >

k 2,,„ to 2 to v Z <o ‘ 2 CO

^an LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3ldVd8n LIBRARIES SMITHSOf
to x _ to — ... to ~ CO

O XCVjoC^ “ o ~sd,> o

‘UTION ^ NOlinillSNI “NVINOSHlIftlS^SB IdVdaiT^LIBRARI ES 2 SMITHSONIAN-* INSTITUTION 2 NOlinillSNI NVINOSHi
z r-» z i- Z r- 2 f~

_ -—-£?/ — »• ' — \ 't~^v.tS.3Sy,rv/

m '"W ^ nv*vasv^z rn X^vovi^z m X'^os*^/ xjvam^jx m

to ' £ co £ to £ co

/Ban LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3 1 dVda IT LIBRARIES SMITHSOI

2 <2 „ _ ^ 2 CO 2 <0 . Z . to

^ E < a 2 E ^ 2

^ | 1 |J | '§51 \.*n$ 1 vkiWfe/ */•' fc :w 2 ^

S 2 'VV^ > * 2 x^oshJ-z >’ 2 '

UTION ^ NOlinillSNI NVIN0SH1HNS°°S3 I dVd 3 n\| B R AR I E S^SMITHSONIAN INSTITUTION W NOIinillSNI_ NVINOSH.
co _ \ w _ _ _ ^ co rr _ to _ jS \

tr
. <
cc

co — Q Xitos^Z - o

2 _l Z

tfdan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS SSIdVdail LIBRARIES SMITHSOI
2 <“ . _ _ 2 m ... z r- z

CD
50

> ■’S
50

m vgv oxz r; m rn ^ ^ X£ i^z m

co £ to £ co X £ co

rUTION NOlinillSNI NVIN0SH1II/MS S3ldVaan LIBRARIES SMITHSONIAN INSTl UTION NOlinillSNI NVINOSH.

CO 2 _ to 2 ..... to Z to 2

<
z
o

CO
X

^ 2 2 > Xto^y 5 ^ > '<wr 2

Vdan“LIBRARIESWSMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS^S 3 t d V d a IT LIBRARIES SMITHSOr
^ ^ . « . . (/) ^ ^

M ^ ^ , co w u w -

-i A* — < /^fCIrJikA .. /> /v°iWk% H ./

o __ Q o

rUTI0N2N0linillSNI_JNVIN0SHlllMS2:S3 I dVd 3 IT ^Ll B RAR I ES^ SMITHSONIAN^INSTITUTION NOlinillSNI NVINOSHJ
2 r- , Z r- 2 i~ z »~

tfdail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IIAIS S3ldVdail LIBRARIES SMITHSOI

(SON AN

3 9088 00923 9138

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The biology, ecology and mariculture of Mithrax spinosissimus, utilizing cultured algal turfs"

See other formats