DEVELOPMENT OF INTENSIVE POND FARMING TECHNIQUES

FOR THE MUD CRAB SCYLLA SERRATA (FORSKAL)

IN NORTHERN AUSTRALIA.

September 1986 - December 1988

FIRTA PROJECT 86/9

FINAL REPORT

by

D.S. FIELDER, D.L.MANN & M.P. HEASMAN

(Sea Hatcheries Limited.)

1
DEVELOPMENT OF INTENSIVE POND FARMING TECHNIQUES FOR THE MUD CRAB
SCYLLA SERRATA (FORSKAL) IN NORTHERN AUSTRALIA.

INTRODUCTION
A large demand for crab meat on both Australian and
S.E.Asian markets has provided incentive for the development of a
mud crab Scylla serrata farming industry in tropical Australia.
Cannibalism of newly-moulted. softshelled crabs however
constitutes a major constraint to profitable pond farming.
The main aim of the project was to observe moult-related
cannibalism of mud crabs and define causative factors of this
phenomenon and then to develop pond design and management
techniques which alleviated this problem.
The project was initiated in September 1986 when
construction of Sea Hatcheries Ltd.' commercial multi-species
hatchery at Mourilyan Harbour. North Queensland (fig. 1)
commenced. Facilities for larval rearing had been established by
December 1986 and the first attempt to produce juvenile crabs was
carried out.
By mid year of 1987 the main hatchery building was completed
and a controlled environment laboratory was made available for
clearwater behavioural studies. Experimental tank systems for
replicated trials were set up in this room and time-lapse video
equipment. for continual surveillance. was purchased and
installed.
Initial behavioural studies had used wild caught juvenile
mud crabs. however the numbers collected were small and their
availability was erratic which made experimental design
difficult.

2
 After many, largely unsuccessful. attempts at rearing larvae
in order to obtain juvenile crabs for the projected behavioural
trials, it was decided to postpone the project for six months
(between 16.10.87 - 31.3.88) while a more sophisticated rearing
system was constructed.

This was carried out and all subsequent rearing attempts

have been successful. These hatchery reared crabs were used in
intensive small-scale trials and techniques have been identified
which allow high stocking density with minimal mortality due to
moult-related cannibalism.

SCOPE OF IBE PROJECT

The project work can be divided into three main areas:

1: Larval rearing
�: Clearwater laboratory observations
J: Small-scale growout trials
Results of these studies are reported separately below.

LARVAL REARING
.L.. Broodstock Maintenance
Over the course of the project eighty-eight (88) mature
female mud crabs (mean C.W. = 15.5 + 1.1cm) were collected from
Mourilyan Harbour and Conn Creek, Hinchinbrook channel (fig. 2)
by set pot or dilly. A histogram comparing the size range of
mature female mud crabs collected in northern (this study) and
southern Queensland (Heasman,1980) is presented in figure 3. This
shows that female mud crabs reach maturity at a smaller size

3
 �
�
,
38
0
�

u rt
0
r
�
m
,
, m
0
rtr
�rt

�m ~rt
� I
� m
�u
m�
�-
ru

rt

m
Figure 2. Mourilyan Harbour and Hinchinbrook Channel,
North Queensland. Collection sites for mature
female mud crabs Scylla serrata.

:4--.. 1-t1wc.l.l.1� c.Roail

l..1-lll,,JiJ '-1..
 Figure 3. The size frequency of mature female mud crabs Scylla
serrata collected in Southern (Heasman,1980) and Northern
(present study) Queensland.

50
Heasman, 1980
45 Moreton Bay
Southern Queensland
40 x C.W. - 16.65 cm
(n=339)
35
F
R 30
E
Q
u 25
E
N
C 20
V

J.5

J.0

- 0
20
Present stu dy
J.8 Mourilyan Harbour
Tropical Queensland
J.6
x C.W. - 15.5 cm
J.4 (n=ll8)
F
R
E J.2
Q
u J.0
E
N
C 8
V
6

0
> > > > > > > > > > > > > > > > > > > > > > > > > > > > >
J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. J. 1 J. 1 J. 2 2 2 2
. . . . . . . . . . . . .
2 2 3 3 3 4 4 4 5 5 5 5 6 6 6 ? ? 7 8 8 8 8 9 9 9 0 0 0 J.
. . . .
6 9 2 5 8 J. 4 7 0 3 6 9 2 5 8 J. 4 ? 0 3 6 9 2 5 8 J. 4 ? 0
. . . . . . . . . . . .
SIZE CLASS <cM)
(carapace width) in tropical Australia.
Initially the crabs were maintained individually in buoyant
clothes baskets held within floating Barramundi (Lates
calcarifer) broodstock cages in Mourilyan Harbour. This system
proved inappropriate as servicing (feeding and cleaning) was
difficult and water quality parameters such as salinity.
temperature and turbidity were variable. Consequently a land
based crab broodstock holding facility was established.
This consisted of a 2.4m diameter (5000 litre) above-ground
fibreglass tank (plate 1) which had a perforated false bottom
fitted to allow water circulation through a sand substrate.
Approximately 20cm depth of calcareous beach sand was supplied as
substrate to permit crab burying as well as provide biofilter
media for reconditioning of sea water.
No more than seventeen (17) adult fem�le mud crabs were held
communally in the broodstock tank at any one time and all had the
propodus of each cheliped clipped to reduce damage sustained
during any antagonistic encounters.
Crabs were fed once daily a mixed diet of trawled crabs
(Portunid .e.E..�J. prawns (Penaeus £E...:...) and pilchards (Sardiops
neopilchardus) and any uneaten food was removed the following
morning.
The temperature in the tank closely followed that of ambient
seawater as heating facilities were not available and fresh
seawater from the harbour was exchanged daily. A screen of 70%
shade cloth was placed over the tank to prevent excessive algal
blooms.

4
.6.1.. Induction of Spawning
Methods used to induce female crabs to spawn (extrude eggs)
included bilateral and unilateral eyestalk ablation (Heasman and
Fielder, 1983) and environmental stimulation. Table 1 shows the
effect of the three husbandry techniques on induction of spawning.
Both physical eyestalk removal techniques induced crabs to
spawn (60% and 70% for bilateral and unilateral ablation
respectively) however bilateral ablation caused high mortality
(22.5%) of crabs in the 24 hours following eyestalk removal. This
was most likely due to excessive loss of haemolymph and other
stress related factors.

Table 1. The number of mature female mud crabs, Scylla serrata

which - extruded - eggs when subjected to -th
- re-e -(3) - spawn
induction techniques.
Spawn No. Crabs No. Crabs Mean No. No.
Induction Subjected Spawning Days to Died
Method Spawn± sd

Bilateral
Eyestalk 40 24 12.4 + 4.4
Ablation (60%)

Unilateral
Eyestalk 10 7 21.9 + 13 1
Ablation (70%)

Eyestalks 38 7 27.4 + 29
Intact (18.4%)
Tank
Environment
-----------------------------------------------------------

5
 On the other hand, crabs subjected to unilateral eyestalk
ablation showed a low mortality level (10%) equivalent to that of
intact crabs held communally. Similar results have been reported
for Penaeus monodon (Liao and Chen, 1982) where mortality for
bilaterally ablated prawns was higher than for those unilaterally
ablated and bilateral ablation was therefore not considered a
practical procedure for induction of spawning in this species.
The mean time taken for intact crabs, held in the broodstock
tank, to spawn naturally was 27.4 + 29 days (n = 7). Eyestalk
ablation markedly reduced this time and increased the percentage
of crabs which extruded eggs. Plate 2 shows a berried female mud
crab holding approximately 1 million eggs.

3: Hatching
Although broods of eggs were obtained in a shorter period of
time when crabs were subjected to eyestalk ablation, the quality
of first stage zoea larvae (Z 1) obtained at hatch was variable.
Broods extruded after bilateral ablation produced very poor Z1
with only 25% of the hatches providing larvae suitable for
culture i.e.. vigorous and feeding at the water surface. Hatching
was often extended over a 24 hour period resulting in prezoea or
Z1 larvae which lacked vigour and fell out of the water column.
Total mortaltiy occured within 24 hours of hatch.
In contrast, broods extruded after unilateral ablation or
natural spawning showed strong hatching synchrony (0600-0800 hrs)
at 29 ± 1°c and produced good quality larvae in 71% of··. the cases.
Samples of larvae from these hatches were used in larval rearing
trials.

6
Plate 1 Fibreglass broodstock holding tank. The tank is
4.5 sq.m and has afalse bottom covered in 20 cm
depth of calcareous sand.

Plate 2. A 'berried' female mud crab Scylla serrata.

The brood contains approximately 1 million eggs.
1..:.. Larval Rearing
Thirteen (13) attempts at rearing mud crab larvae to first
stage crab (C1) were carried out however only the final three (3)
were successful in producing large numbers (1000 total) of
juvenile crabs for projected observational studies.
Figure 4 shows a typical larval survival curve for the
successful and unsuccessful attempts. The first ten larval runs
were carried out using experimental rearing tanks situated inside
the main hatchery building. Sea water was drawn from the main
rearing system and was pretreated by sterilization followed by
reconditioning through ultra-violet and biological filters.
Unfortunately continued water quality problems were
experienced (as seen by mass mortality of commercial numbers of
Barramundi larvae) and after major mortality at Z1 a steady
decline in larval numbers occurred. Similar mortality patterns
have been reported for S.serrata and Portunus trituberculatus
larvae cultured in Japan (Cowan, 1984). The major causes were
attributed to poor quality of the newly hatched Z1; deteriorating
water conditions with time and inadequate larval nutrition.
Because satisfactory numbers of juvenile mud crabs had not
been obtained by October, 1987 a decision was made to postpone
the project while a new larval rearing system was established. A
modified system based on a design by Heasman and Fielder (1983)
was constucted.
This overcame the water quality problems previously
experienced and larval survival, particularly during the early
zoeal stages, was greatly increased. Larval nutrition for the
early stages (Z1 , Z2 and Z3) was also improved however mortality

7
 1-•
, ..t:,., (..,r, t:J'.i -.J (:Cl 1...0 C.�
1::::;::1 13;) ($) ,:s::i C5:I ,::s:i ($1
-��-�.........--.1.,..--L--.J

i.Ci
r.::

!J!
:- �-;
n .<
n ·1:i
m :-:.
!fi n
Ul !li
...h !M-:

\l:
!M.:o
<
�:J

,-,
'-'
tn ffO!--:

::'�t��
!3:)•
I�•� !·•--

u;0·:a = r.::r
::c

lf!
1 n::
nJ -1;
!t::t. [ !l! ·1
-1 D.1
II' c. i-'• r-T
!.:I: i c. !l !
I' :J
,.r.)
.,r
!Citi
;\
"..
"' t
t..
!M ...� �!r·r-J ..!l1
-r r :-r <
t�i [] - m !J.:
..§
M: ,:
: IT!
1-4- ,-::r,
er:, !.n �-
II :.::i
•.!.JI llfll •.,�. Cl
o:::i cb
�-..i
o:�
persisted during the latter larval stages (Z4, Z5).
Exuvia entrapment at the first metamorphosis from Z5
Megalopa caused excessive mortality. Affected animals died during
ecdysis after being unable to free their appendages and eyes from
the exuvia or alternatively some individuals shed their exuvia
but lacked vigour and died soon after. Similar symptoms have been
experienced for larvae of Macrobrachium rosenbergii and Palaemon
serratus larvae and juvenile lobsters, Homarus §.E..:.. (Brock, 1982).
It is thought that this is largely due to inadequate larval
nutrition.
Thesurvival rate from Z1 to C1 was very low and averaged
approximately 0.5%. This is in the order of survival rates (4-9%)
claimed by Japanese technicians for culture of Scylla (Cowan,
1984) but is much lower than the 26% survival recorded by Heasman
and Fielder, (1983) indicating that site variation may contribute
significantly to the success of mud crab larval rearing.
Improvement of larval nutrition needs further investigation
before commercial numbers of juvenile mud crabs will be reared.

CLEARWATER LABORATORY OBSERVATIONS

Observations of moult associated cannibalism were carried

out under clearwater conditions in tanks contained in a
controlled environment room (29 + 1°c; 14/10 hr Day/Night;
0
salinity 30 + 2 /00 ).
Investigations included a series of replicated trials
determining the effect of substrate, food quality and quantity
and stocking density on survival of juvenile mud crabs. In

8
addition, non-replicated continual surveillance trials which
utilized time-lapse video recording, enhanced data obtained for
the above trials with crab behavioural observations.

L Simulated Tidal Trial

Introduction
Heasman (1980) observed that juvenile mud crabs commonly
moulted (shed exuvia) in seclusion under rocks and root systems
in the intertidal zone of mangrove flats. Evidence for this was
the discovery of exuvia and occasionally newly moulted,
softshelled crabs when rocks or leaves were lifted at low tide.
Hill et.al. (1982) also recorded that juvenile mud crabs (20-80
mm carapace width) remained in the mangrove zone at both high and
low tide.
It was thought therefore that secluded intertidal sites may
be essential for ecdysis of juvenile mud crabs and a trial was
established to simulate this in the laboratory.

Materials and Methods and Results

A simulated tidal system based on principles developed by
Quinn and Fielder (1978) was established in a glass aquarium
(fig. 5). In this, an intertidal zone, comprising half of the
bottom substrate, was exposed and flooded at six (6) hourly
intervals. Four (4) shelters, made from offcuts of 40 mm pvc
pipe, were provided with two (2) shelters positioned in the
intertidal and subtidal zone respectively. Both of the intertidal
shelters were capable of holding water when emersed, thus

9
F" i (;i !...I. i·- E·! ��:j ,; i�·::1 5;. j_ in u. 1 -��!. t. E· d l:::! f:.-:� -73. C r--1 ·::::- )/ ;::. L t-:: I! 1 ( i·-1 .=:1 -::;;. f:.:.·, r·i c:, n 2-. cl e �-:. j_ i;} rt !-.••
,..! yI
Ciu_j_ nn
c.-tnd r:-:tE:J.c!E�f:-� J_C)/':3) :1

c..
b

1-\,.9.h.. \JQ l-c-J --- ---- - - - -­

i
e

. . . .•... .·:.
.....
.. . .....
· .. -��
. .. ..··'·.... - ·.,
. ·�.. ..: :· ...
..
.. . .... .. ..
• • '
i

a & c, perforated false walls; b, sand compartment; d, airlift

controlled by time switch (h) and aquarium air pump (g); e,
constant level siphon; f, 30 litre reservior; i, artificial
shelters.
 Figure 6. The mean hourly activity of 2 juvenile mud crabs Scylla
serrata held in a simulated tidal system for 5 days.

60
"\\'i)� OU( "t"tt>E ltJ •nbe. 0� "t"l\)£:: UJ
"'M'j WG-1-\-1.

M
I 50
N
s
A 40
C
'I
I
u
I 30
T
y

/' 20
H
0
u
R 1.0

0 ........_________________________________,
0 0 0 0 0 0 0 0 0 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 2 2 ·2 2 2
1. 2 3 4 5 6 ? 8 9 0 1. 2 3 4 5 6 ? 8 9 0 1. 2 3 4
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -0 0 0 0 0 0
HOURLY INTERUAL
emmulating a pool of water typically found beneath a rock or leaf
in the mangrove zone at low tide.
Two (2) wild caught C6 mud crabs (av.C.W.= 21 + 1.45mm) were
used in each trial and were fed twice daily on minced crab
(Portunid .§.E...:..). The trial was repeated three (3) times and each
trial was run until one of the crabs moulted (5-6 days). Closed­
circuit television and time-lapse video recording were used to
continually observe crab behaviour.
The state of the water level appeared to have a marked
effect on juvenile mud crab behaviour (fig. 6). The number of
antagonistic contacts and time spent actually moving over the
bottom and foraging for food was much greater at "high tide". At
"low tide" the crabs were mostly inactive i.e. either buried in
the sand in both subtidal and intertidal zones or occupying an
artificial shelter.
Three (3) crabs were observed to moult. Two (2) of these
crabs moulted at high tide between 0507 and 0607 hrs. whereas the
third crab moulted at low tide at 1500 hrs.
All crabs moulted in the subtidal zone and were positioned
on the sand substrate. They did not occupy the artificial
shelters which were large enough to accomodate a moulting crab.
All of the newly moulted (softshelled) crabs were attacked
and consumed by their hardshelled conspecifics, almost
immediately following ecdysis. It appeared that the hardshelled
crabs were able to detect the presence of a moulted crab by a
chemical stimulus. On several occasions the hardshelled crab,
which had been buried and had the sight path blocked by a
shelter, became active and made its way directly to the

10
softshelled crab. The chemical involved was most likely moulting
hormone (Fielder, pers. comm.).
From these results obtained by small-scale tank trials, it
appears that the imposition of a simulated tidal system is not an
effective method for reducing cannibalism of softshelled juvenile
mud crabs.
Other factors against this system of management include:
(a) subadult (70-150 mm C.W.) and adult (150-200 mm C.W.)
crabs, which represent the marketable product, migrate into the
intertidal zone at high tide but mostly retreat back to the
subtidal zone with the falling tide (Hill et.al.,1982). Therefore
stocking density would increase at low tide and hardshelled as
well as softshelled cannibalism could account for major
mortality.
- - . - (b) - good-·- quality brackish/sea water would be·· necessary
constantly for water replacement
(c) High pumping costs.

L Substrate Preference Trial

Introduction
The above tank trial had shown that juvenile mud crabs had
essentially occupied a different habitat for hardshelled
(intermoult) and softshelled (moulted) behaviour. The intermoult
crabs spent their inactive periods either buried in the sand or
secluded in the artificial shelters whereas moulting occured on
top of the sand substrate.
It was postulated therefore that the mud crabs may prefer to
moult in open spaces to facilitate escape from potential

11
predators. Heasman (1980) showed that recently moulted crabs are
able to flee from or repel (by threat postures) other hardshelled
crabs within several hours (1 - 2% of the intermoult duration) of
moulting.
This trial was set up to determine if juvenile mud crabs had
preferred intermoult and moulting sites when provided with
naturally occurring substrates.

Materials and Methods

The substrates used in the trial were coral rubble, detritus
(leaf litter found in the crabs natural habitat) and calcareous
sand. The coral rubble and detritus provided darkened cavities
for seclusion and the sand a fluidised substrate for either
burying or open space for moulting.
Two (2) substrates were provided in equal proportions in
·each of three·treatment·substrate combinations. The· ·combinations·
were coral rubble/sand, detritus/sand and sand/sand (control).
Each treatment combination was replicated three (3) times in a
completely randomized design. Tanks were 0.11 m2 and each tank
was stocked
with 3 C5 or C6 wild caught mud crabs (stocking
density = 27 m-2).
Observations of intermoult crab and exuvia position were
noted every 2 hours from 0800 to 1800 hrs for seven (7) days.

Results
Hardshelled crabs significantly (P < 0.001) preferred coral
rubble and detritus substrates to sand as sites for normal
intermoult behaviour during the period of observation (tables
2 & 3).

12
Table 2. Total number of juvenile mud crabs found occupying the
coral rubble or sand for 6 observations during 0800 and
1800 for 7 days. Three (3) treatment replicates with 3
crabs per rep.
TREATMENT
RUBBLE SAND
DAYS (total number crabs) (total number crabs)
Rep 1 Rep 2 Rep 3 Rep 1 Rep 2 Rep 3

1 18 17 18 0 1 0

2 17 13 18 1 5 0

3 17 12 17 1 6 1
4 15 7 9 3 11 9
5 13 13 10 5 5 8
6 14 13 8 4 5 10
7 9 13 11 9 5 7

Subtotal 103 88 91 23 38 35
TOTAL 282 *** 96

*** Significant at 0.1% level

Table 3. Total number of juvenile mud crabs found occupying the
detritus or sand for 6 observations during 0800 and
1800 for 7 days. Three (3) treatment replicates with 3
crabs per rep.
TREATMENT
DETRITUS SAND
DAYS I
I (total number crabs) I
I (total number crabs)
---------------------------------------------------------
Rep 1 Rep 2 Rep 3 Rep 1 Rep 2 Rep 3
I
I I

1 17 16 17 1 2 1
2 15 13 16 3 5 2
3 16 18 13 2 0 5

4 15 17 13 3 1 5
5 17 17 17 1 1 1
6 16 18 13 2 0 5

---------------------------------------------------------
7 15 14 15 3 4 3
Subtotal 111 113 104 15 13 22
TOTAL 328 *** 50
*** Significant at 0.1% level
 When crabs moulted in the coral/sand treatment 83.3% (n= 12)
of the exuvia were found on the sand surface. Thus there appeared
to be a distinct preference for different sites to carry out
intermoult and moulting activity. This was a very desirable
attribute because by simple preference for a substrate the
vulnerable softshelled stage had removed itself from the vicinity
of the predatory hardshelled crabs.
There did not appear to be any preference for the sand
substrate as a moulting site when combined with detritus. Only
56% (n= 9) of the exuvia were found on the sand.
A possible reason for the difference between detritus and
coral rubble may concern the type of cavity available for
seclusion. Although the coral had interstitial cavities suitable
for hardshelled seclusion they may not have been large enough for
ecdysis. thus forcing the crab to leave the coral. On the other
hand, the nature of the detritus (organic debris) was such that
the cavity size was undefined i.e. the substrate could be easily
moved by both hard and softshelled crabs to create larger
cavities if desired.
Although the treatment combinations caused different
behavioural patterns there was no effect on cannibalism at this
stocking density as survival for all treatments was 100%. Further
trials investigating the treatment effect at higher stocking
density were not carried out for the following reasons.
If a rubble system were to be established in a pond
situation several problems could arise.
(a) The rubble would have to provide intersticies large
enough to accomodate subadult and possibly adult crabs. This

13
would only be accomplished with large rocks/bricks which in most
cases would have to be trucked onto site. This would be a costly
procedure.
(b) Siltation, which would reduce the efficiency of the
system by reduction of cavity volume, could be a major problem.
Desilting would require removal. at least in part, of the rubble
which would be labour intensive and therefore expensive. The
frequency of this procedure would depend largely on the quality
(amount of suspended solids) of the incoming exchange water.
(c) Harvesting could present problems as seine netting would
be impossible and ineffective over the rocky bottom; drain
harvesting would again require the removal of the rubble to
capture secluded crabs. Trapping would be the only other
alternative and this can lead to bias sampling as well as
incomplete population harvest.
(d) By providing a wide range of cavity sizes the smaller
crabs may remain within the rubble to moult. thus becoming
vulnerable to larger intermoult crabs also occupying this zone.
The use of organic detritus in a pond situation would not be
possible due to the potential water quality problems. however it
would be possible to provide inert artificial materials which
created a similar environment.
The following trials were based on development of this idea.

14
Artificial Substrate Trials
Introduction
Preliminary observations of large numbers of C1 C4 mud
crabs held in tanks supplied with two (2) types of artificial
material, namely fibreglass fl¥Wire and nylon onion bags, had
shown that both materials provided highly desired habitats.
Juvenile mud crabs spent most of their time secluded
underneath or within folds of the mesh and only left to forage
for food. Antagonism between conspecifics was reduced and this
appeared to be related to the amount of artificial material
present.
A large proportion of the exuviae were found on the mesh and
it seemed that the moulting crabs required a substrate which
they could grasp to facilitate extraction from the exuviae. Many
of the exuviae were firmly attached with the dactyl of .their
perieopods (walking legs) wrapped around the matrix of the mesh.
When C4 crabs were observed in the process of moulting it
also appeared that extraction from the exuvia was accomplished
faster on a mesh substrate than on a bare tank surface.
Although both artificial materials provided similar
substrate qualities the onion bag was chosen for further
investigations for the following reasons.
(a) Onion bags were much cheaper than comparitive areas of
flymesh and they were
(b) readily available.
(c) The open weave of the bags allowed better water
circulation and did not accumulate detritus as quickly as
flymesh.

15
 (d) The bags were purchased in a form suitable for
establishment in a pond situation whereas the fl¥Wire required
work such as cutting and folding/stitching to create darkened
cavities.

Artificial Shelter/Depth Trial

Due to the above observations a trial was set up to
establish quantitatively if provision of increased amounts of
artificial shelter affected survival of juvenile mud crabs.
Of particular interest was whether the typically benthic
juvenile crabs would utilise artificial substrate positioned in
the water column as Heasman (1980) observed that threatened C1 -
C3 mud crabs commonly swam vertically upwards and remained in the
water column for several minutes.
Materials and Methods and Results
The trial consist-ed -of four (4) non-replicated-- shelter
treatments and was carried out in the experimental tank system as
described in figure 7.
Horizontal layers of onion bag were provided at intervals of
112 mm depth and treatments 1,2,3 and 4 each had 1,2,3 and 4
layers of mesh respectively.
Each tank was stocked with 9 C5 (C.W.= 15.5mm) and 9 C6
(C.W.= 20.6mm) crabs which represented a stocking density of 65
crabs/m2 . The crabs were fed to satiation three times daily
(0800, 1700 and 2300 hrs) on a diet of minced crab (Portinid .§.E..:..)
and prawn (Penaeus .§.P..:..).

The tanks were checked twice daily for any exuviae and the
position in which they were found was recorded. The number of

16
surviving crabs was recorded every 2-3 days. The trial was
conducted for 31 days.
The amount of available shelter had a marked effect on
survival of the crabs (table 4). On the whole. the number of
surviving crabs was related to the number of layers of shelter.
Survival was lowest (22%) in the treatment with 1 shelter and
highest (61%) in the treatments with 2 and 4 layers. The number
of mortalities was closely related to the number of moults which
occurred on the bottom level (figure 8). Food was placed onto
this level so it is likely that most of the crabs present there
were foraging and were therefore potential cannibals.
This may explain why survival in treatment 2 was the same as
treatment .4. 71% and 77% of the moults occured above the bottom
layer for treatments 2 and 4 respectively whereas only 55%
occurred above the bottom layer in treatment 3. This may indicate
that the position where moulting occurred was largely
opportunistic. The treatments were not replicated so no real
conclusions about site preference can be drawn.
The data does show however that juvenile crabs occupied
artificial shelter which was present at varying depths throughout
the water column.

17
Figure /. Experimental tank �;�L=m t.he
effect of artificial shelter m1...td

J
7
C

----..::::r-==:::::::::
I.4-
,.
d - j::3__ -
I-
'-;- ..
,_
,_ .

.
._

'
..
.
b
.'
'
' '
a.

a, coralline biofilter; b, 100 litre circular experimental tank;

c, airlift pump; d, overflow return line.
Table 4. The number of moults and mortalities experienced in
tanks which had horizontal layers of artificial
substrate provided at depth intervals of 112.0mm.
Tanks were initially stocked with 9 C5 and 9 C5 crabs.
-------------------------------------------------------------
Number of Number of Number of No.Mortality %
Shelter Moults Mortality No. Moults Surv.
Layers %

1 17 14 82 22
2 24 7 29 61
3 22 11 50 39
4 26 7 27 61

Figure 8. The position, number of moults and mortality that took

place in 4 treatments of varying levels of artificial
shelter. Tanks were stocked with 9 C5 and 9 C6 mud crabs
initially.

Treatment 1 . . Number· Number

Moults mortality
17 14

Treatment 2
-I , -,. - -_ - I
,_J, -(
17

\ •: -." - .- . : : -: - · I
7 7

Treatment 3
4

·1 ------. · · · · �. -
•• : •. •' ••••• :
;i 8
10 11

I
Treatment 4 5
..............
.............
7
8
6 7

18
Artificial Shelter/Sand Trial
Introduction
The above trial had shown that crabs would utilise
artificial shelter for both intermoult and moulting behaviour
however the shelter was not tested in conjunction with a natural
substrate.
It was thought that the mud crabs may have had a greater
preference for a natural fluid substrate in which they could
bury, and therefore the inclusion of artificial shelter in a pond
would be pointless if the crabs did not use it.
Two trials were run in concert to determine the effect of
artificial shelter with a sand substrate on crab behaviour and
survival.
Continual Surveillance Trial
Materials and Methods
This trial utilised continuous video observation and was
carried out in a single 100 litre (Nally Pty Ltd) container,
which was established with two levels of artificial shelter (225
and 337 mm depth) and a bottom level of calcareous sand. Each
layer of substrate occupied 1/3 of the container surface area
(plates 3 & 4).
The artificial layers were constructed from plastic
cardboard covered in nylon mesh bag. Excess onion bag was bunched
to create folds in which the crabs could hide. The sand layer was
positioned centrally and was approximately 5 cm deep.
Water was air-lifted from the tank into a coralline
biofilter and reconditioned water was returned to the tank below
the sand layer. This prevented turbulence and allowed undisturbed

19
Plate 3. Surveillance camera used in conjunction with time­
lapse video recorder for observation of juvenile
mud crab behaviour.

Plate 4. 100 litre experimental container. High and Low

levels of artificial shelter and a bottom sand
layer.
observation with the video equipment.
The tank was stocked with 9 C6 mud crabs (3 2 crabs/m 2 ) which
were fed 3 times daily on minced crab and prawn. The food was
placed on the sand substrate.
Video records were taken for 5 days after a 7 day
acclimation period. Analysis of the video consisted of
observation of crab position at 15 minute intervals. Data was
analysed for significance using ANOVA.
Results
Results showed that over the 5 day period of observation the
juvenile crabs significantly (P<0.001) preferred the artificial
substrate for intermoult activities and that more crabs spent
their time secluded on the higher bag shelter than the lower
(figure 9).
Crabs were noted to be absent from the sand on 2 4.8% of the
observations (n = 464), however the mean for total hourly
observations over 5 days showed that there were always some crabs
present on the sand. On only 2 occasions were crabs observed to
be buried. The rest of the time they were exposed and actively
moving (foraging) over the sand surface.
Although the crabs were not individually identified it
appeared that they underwent a feeding rotation. There were no
apparent peaks of activity (increased number of crabs) at feeding
time. Crabs would leave the shelter of the bags and move onto the
sand where they consumed food or on occasion retreated back to
the bags holding food. A maximum·of 4 crabs were observed on the
sand at any one time.
Moulting did not occur during the 5 day observation period

20
 Figure 9. The mean number of Juvenile mud crabs found occupying
the substrate layers per hour (total of 4 by 15 minute
observations). Substrate consisted of 2 layers of onion bag
at 225 <LOW) and 337mm <HIGH> depth respectively and a
bottom sand layer.The tank was stocked with 9 C6 mud crabs
initially and the trial ran for 5 days.

25 +----'c�A�Y'--___...,____�N�'���--r-______ -L.___ o�A�i___-l

N 20
u
M
B
E
R J.5

0
F
C 1.0
R
A
B
s
5

0
�
\f? :9
J. J. J. J. 1. 1. 1. 1. 2 2 2 2 0 0 0 0 0 0 0 0 0 0 1. 1.
2 3 4 5 6 7 8 9 0 l. 2 3 0 1. 2 3 4 5 6 7 8 9 0 1.
3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

-
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 � 0 0 0
HOURLY INTERUAL
HIGH SHELTER-++- SAND -e- LOW SHELTER
however 8 exuvia were found during the 7 day acclimation period.
Five (5) moults occurred on the 337 mm depth shelter, 1 on the
225 mm depth layer and 2 on the sand. Mortality did not occur
over the course of the trial.
The trial indicated that artificial shelter was actually
preferred for seclusion over the natural fluid sand substrate.
Inactive crabs also preferred to be as far away as possible from
the feeding site.

Artificial Shelter/Sand Survival Trial

Materials and Methods
This trial was established to determine if the presence of
vertical bag shelter affected mud crab growth and survival.
The trial consisted of 2 treatments with 2 replicates per
treatment and was carried out in the tank system as in figure 7.
The first treatment (control) provided a calcareous sand
substrate only while the second treatment provided both sand and
vertically suspended artificial bag substrate. Two (2) onion bags
were cut in half and placed into each of the 2 treatment tanks.
This represented 14.5 bags/m 2 . To ensure that the bags were
suspended vertically in the water column a piece (10 cm 3) of
polystyrene was tied to one end while the other end was weighted
down with a piece of coral. The sand substrate was placed on top
of a false bottom and water was circulated (by an air-lift pump)
through the sand to prevent anoxia. The depth of each tank was
approximately 450 mm (plate 5).
Each tank was stocked with 25 crabs (S.D. = 92.5 crabs/m 2 )
ranging in size from C4 to C7. An equivalent ratio of each stage

21
Plate 5. Experimental 100 litre tank. Substrate provided
consisted of vertically suspended artificial
onion bags and a bottom sand substrate.
crabs were placed into each tank. Crabs were fed as in the
previous trial.
Observations of any exuvia and their position were noted
twice daily. A population census was taken approximately every
third day. The trial was run for 23 days.

Results
The results showed that addition of the vertically suspended
bags significantly (P<0.001) increased survival of the juvenile
mud crabs (table 5).
As a consequence of increased survival, the biomass after 23
days was also significantly (P<0.01) greater in the plus bag
treatment. There was no difference between the mean weight of
individual surviving crabs however the range in the plus bag
treatment was larger and both replicates of this treatment
-comprised Cg-·cra-bs:- -There were no C9· crabs-·present -±n · the---mmd -­
only treatment as that these larger crabs were consumed, as they
attempted to moult, by smaller conspecifics.
This is contrary to the result obtained by Holland et.al.
(1971) when blue crabs, Callinectes sapidus were held
communally in a tank with a sand bottom only. They found that
crabs became larger as the population became smaller (due to
cannibalism) and their trial was terminated when the population
was comprised of one large crab.
These authors did report however that a lessening of
cannibalism (as was expected) was not observed as the population
in each aquarium was diminished. This was also observed in our
trial (figure 10).

22
 Fi(;JLtr... t:::· :l() ., -rf1E� E1f-f<::-. :<ct �.J•;· "L!!t.:= r-::1·-·e.·=-1::::nce c1f \/�:-:.1 1---t.j.c:.::·:�11·-y-- �::;1...1.�::;pencJi::·:-:1cl t::i.:;:·•. g
shelter □n growth (success of moulting) and surv1va1 of
juvenile mud crabs. 25 mud crabs ranging from L4 to C7 were
stocked initially. The trial ran for 23 days.

50.00
M
E
A
N
N 40c00
0

M
0
u 30.00
L
T
....,
C

2() e (}Q

T
A 10c00
, ___ *
----
L ---- =
I --¾---------$-�
T
V
'
0:00-+-tit"----------,----------.---------.--------,
,
0= (H) 12 ... i){) 18= i)(i 24: ()(}

□ NO BAG - MEAN NUMBER MOULTS

� NO BAG - MEAN NO. MORTALITIES
◊ PLUS BAG MEAN NUMBER MOULTS
� PLUS BAG - MEAN NO. MORTALITIES
 Cannibalism of hardshelled crabs accounted for a major
proportion of mortality in the sand only treatment, as indicated
by a greater number of mortalities than moults.
Aggression between the few remaining crabs in the sand only
treatment was significantly (P<0.05) high as 50% of the crabs
were missing appendages. In the plus bag treatment however a mean
of 24.7% of a larger population of crabs were missing appendages.
Approximately equal numbers of exuviae were found in the bag
treatment replicates and about half of the exuviae were
positioned either in or on the mesh bags (table 6). The other
half were lying on top of the sand. This data suggests that there
was no strong preference to moult on the bags however the
observations included only the position of exuviae after the
crabs had moulted. Because the bags were suspended vertically
through the water column it is highly likely that some exuviae
had fallen off onto the sand.

Table 6. The number of exuviae found on the sand and the mesh
bags in treatment 2.
SAND MESH TOTAL
Repl 20 19 39
Rep2 20 15 35
TOTAL 40 34

Discussion
Results of the above 2 trials show conclusively that the
addition of vertically suspended artificial shelter greatly
increases juvenile mud crab growth and survival.

23
MEAN FOOD CONSUMED <xBWt)

1-n. i•.•.:. m

n:!�) I-"- 1:��� (,�,) t•� IC,,.H'j (Jl"'N .....� 0:) •.,Q n.i�
;-,.:. :i

::r:·:· m
�-
1::L ::::!
m
n !:u _,!
·:

::i .....
,.C! :,
C
..J:::: ::::
'""r'
_,

-:; 11

:...i.
!
J
: I
-,
__:

CL !.i1
U'l !--

-,-, 1T.!
C C:L

o_ :�:
[1.ir·l-
..,( 1:
!.11

···<
r-=·
 () '���' i,-,.. 1-• t·J
Q OJ t•,J r.:r- C:•
t,,._.1
,:) () 0 0
l.11 i.J" ,_,,
1' +
+ ,, ·I·
<
I
,. ·I-
,, /T
,, i
+
,'i' !'.IJ
+ + ,,
,,I
I-'
,' -1...1., :!:L!::i
+ /+
()
::i:: ()
rn
J,-..j
,, //
G"I
:1: ,/ +
1'
ll
-..j
Ci
·+:
'C)
"1 /'/
1 + l �ji -h
:1•,
I·•.) ,,:; .....
3: / m Cl
r- --� 0..
t:t i::::i
1'
()
1/
('") ,,
;1J
l�
1;1:1 ,,.,/
.......
J.Ci I
,'
....... ,.
+ ,I
(,•J
I'
i:::::i
()

,'/(
/ ..ffj,
+ 1,
,
·-·
,· o. rn
,I +
fll1
!-,•I+

,I
·'1'
_,,
J.t
-h ··!<

••••••i••,1:::
 The crabs showed a strong preference for the artificial
shelter, especially in the shallower regions, for normal
intermoult activity.
Both hardshelled and softshelled cannibalism was
significantly reduced by providing sites (cavities) for seclusion
in which crabs could either carry out moulting or escape to after
having already moulted. The crabs were able to remain here until
their new shell hardened and they could continue normal
intermoult activity.
Provision of the artificial shelter markedly reduced contact
and therefore antagonism between conspecifics which resulted in
fewer crabs losing limbs. This is very beneficial as generally
loss of appendages results in smaller increments of moult
(smaller sized animals) and may prolong intermoult duration
_______ . dep_ending _on the _ number o .f . 1 imbs __ requiring regener.afi.on
(Hartnoll, 1982).

Therefore by suspending artificial shelter through the water

column, another dimension (depth) which is otherwise unoccupied
by the mud crabs can successfully be utilised to increase
intensity of production.

Feeding Trials
Several trials were set up to determine the quantity of food
consumed by juvenile crabs and the effect of feeding frequency on
survival.

24
L Individually contained crabs
The first trial investigated food consumption of individual
juvenile crabs. Fifteen (15) wild caught juvenile mud crabs
ranging in size from 0.007 - 3.679g were held individually in 1
litre containers of sea water (salinity = 31.5ppt; Temp =25 °c).
Each crab was presented a known quantity of 20% agar bound prawn
pellet at 4 hourly intervals. The food was placed on a 53 um mesh
screen to facilitate retrieval. After 4 hours any remaining
pellet was removed from the container, drained for 1 minute on
blotting paper and weighed (Mettler AE 166 balance). The amount
of food consumed was calculated by difference. Preliminary trials
had shown that pellet weight did not change after 4 hours
immersion.
The feeding trays were cleaned and a new cube of food placed
into the crab container. Water was exchanged daily and the trial
ran for 6 days after a 7 day acclimation period.
Results
More food was consumed during 1000 - 2200 hrs however some
food was consumed during all 4 hourly observations (figure 11).
Food consumption (dry wt. food/wet wt.crab * 100) was highly
variable from day to day and between crabs. The mean amount of
food consumed by 15 crabs over the 6 day period was 27.2 + 22.6
%Biomass/day.
There was a distinct reduction in the quantity of food
consumed as the crabs became larger (figure 12). Smaller mud
crabs (C2 - C5) usually moult every 5 - 6 days (at 29 - 30 °c)
whereas the larger crabs in the trial (C7) moult approximately
every 14 days. Therefore the demand for food by the smaller crabs

25
may be greater due to the shorter intermoult duration.
An optimal feeding strategy would be to supply food
constantly because although a feeding peak was observed, some
food was consumed during every 4 hour period. Also because the
amount of food consumed varied daily, a strict feeding regime
could not be set. Better management would involve constant
observation of the amount of food present in a pond and altering
the quantities supplied accordingly.

� Feeding Frequency Trial

The quality of food available can significantly influence
the growth (intermoult duration and moult increment) of
crustaceans (Hartnell, 1982). Food quality can vary for reasons
such as leaching of water soluble nutrients and the rate at which
fats/oils become ·rancid-:-Toi13- trial was e--stab'tished··to tll:ftertnit1e
if frequency of feeding affected survival and growth of
juvenile mud crabs.
Materials and Methods
The trial consisted of 2 feeding frequencies of 2 times/day
and 6 times/day in which food was provided at 0800 and 1700 hrs
and 0800,1030,1300,1530,1730 and 2200 hrs respectively.
A total of approximately 60% (dry wt.food : wet wt.crab*l00)
of minced crab, Portunid .§.h was fed to each tank daily. (30%
BWt. was fed initially but it was being consumed rapidly.
Increased rearing temperatures of 28 ± 1 0
C was the probable
cause of the higher consumption.) This was a sufficient quantity
to allow satiation with little remaining food so that good water

26
quality was maintained. Prior to each feeding session any
remaining food and waste was siphoned from the tanks.
The experimental design was completely randomized with 4
replicates of each treatment. Experimental containers were 0.11
m 2 and each received reconditioned water from a central coralline
biofilter.
Each tank was stocked with 10 hatchery reared
(C.W.=15.9mm) mud crabs. An equal amount of bunched artificial
material was provided for shelter.
Daily observations of the number of remaining crabs and any
newly moulted crabs were carried out. The weight of the newly
moulted crabs was determined and the amount of food given to each
tank was adjusted accordingly. The trial ran for 28 days. Results
of the trial were subjected to ANOVA to determine any treatment
significance.
Results
Table 7 shows that frequency of feeding had no effect on
survival or growth of the mud crabs. This indicates that the
minced crab had a similar quality after 2.5 and 9 hours
immersion. The food was usually noticed to be present but in
diminishing quantities as the day progressed.

27
Table 7. The effect of 2 frequencies of fee2ing on survival of 10
juvenile (C5) crabs held in 0.11 m tanks. Each tank was
fed a total of 60% biomass per day for 28 days.
Frequency of Feeding
2 times/day 6 times/day
Rl R2 R3 R4 Rl R2 R3 R4

Survivors 9 3 4 8 8 5 4 6

-----------------------·
mean
-----------------------------------
6.0 n.s mean 5.0
= =

Tot. Wt(g) 31.5 12.4 17.8 30.2 25.6 22.2 17.1 28.9
crabs
mean = 23.0 n.s mean = 23.5

Mean 29.3 30.5 32.2 30.0 28.7 32.0 31.9 32.9

C.W. (mm)
mean = 30.5 n.s mean = 31.4

Whether the food quality was good or bad (due to leaching

etc) was no-t determined chemically-;· However··it -was-~-highly­
palatable and produced excellent growth rates which implies that
it had remained in good condition.
Pelletized, formulated feeds were not tested as Portunid
crab specific feeds do not currently exist in Australia. It is
likely however that retention of quality after immersion will
pose major problems to the development of suitable formulated
feeds. Bordner et.al. (1986) found while testing synthetic feeds
for the clawed lobster Homarus americanus, that the degree to
which soluble nutrients leached was highly variable. This was
dependent on the composition of the pellets and was more related
to the quantity of fish and shrimp meal present rather than the
amount of binder.

28
 Therefore in order to overcome potential quality problems
with pelletised foods the most practical feeding scheme would be
one of frequent application of small quantities of food.

� SMALL-SCALE GROWOUT TRIALS

Introduction
As a continuation of the previous experiments a larger-scale
trial, incorporating the tank design and management systems
deemed most suitable for intensive mud crab culture, was set up.
Unfortunately only one tank was available in which to carry out
the trial, however it has provided excellent information on the
potential production of mud crab culture.

Materials and Methods

A · 5000 1 i tre -·circular ( 2 .4 m - diameter) fibregl-ass· tank --was · - ·
used as the rearing vessel. Eight (8) reefs each consisting of 5
onion bags were provided for shelter. In order to suspend the
bags vertically a piece of polystyrene was placed in the closed
end of each bag (plate 6). The tank bottom was left bare (no sand
was provided) to enable easy removal of waste and uneaten food.
Thirty-three (33) sibling mud crabs ranging in size from Ca
(C.W.=36.5mm) to C10 (C.W.=63.7mm) were placed into the tank.
This represented an initial stocking density of 7.2 crabs/m 2 . The
food type varied but on the whole consisted of minced crab
Portunid §E..:_ and formulated fish pellet. This was mixed together
to form a moist dough. Records of food consumption were kept to
determine conversion rates. The amount of food present in the

29
tank was observed regularly (every 1- 2 hours) and provided when
necessary. Feeding commenced at 0730-0800 and continued to 2 400

hrs.
The tank water was aerated continually and at least 100% of
the volume was exchanged daily. Four (4) 2 00 Watt aquarium
heaters were used during the cooler winter months in an attempt
to raise the water temperature. Temperature ranged from 21.5 to
0
2 9.5 C.
At 3-4 day intervals the tank bottom was siphoned to remove
waste. At approximately 14 day intervals the tank was completely
drained and the crabs were measured (C.W.mm) and weighed to
determine biomass increase. The tank was checked twice daily for
any exuviae. The trial ran for 1 2 3 days.

Results and Discussion

- Figure· ·13 shows the-percent survival-and·number---o·f-- mou-lts­
recorded over the 1 2 3 day growout period. Sixty-six (66) moults
and therefore instances in which the crabs were softshelled and
highly vulnerable to cannibalism were observed. Twelve
(i.e.36.4%) mortalities were recorded during the trial and
reduced the stocking density from an initial 7. 2 crab/m 2 to a
final 4.6 crab/m 2 .
Many of the exuviae and newly moulted crabs were found
inside folds of the bags demonstrating that the larger mud crabs
continued to use the artificial shelter for moulting and
seclusion.
Growth of the mud crabs was exceptional with the first C1 2
crabs of 100mm C.W. ( 2 00g) obtained in 77 days (137 days from

30
 Figure 13. The µbrL�11� survival and number of □ 2
□
juvenile mud crabs grown for 123 days in rc�_,} ��.11]·=·__:,:· s,,,-.,· \,��--.!.i;'.�._•, ! -� ·_!.•--�
.r·• .• �:. �
•
. :.
1
:·,�· ·- • ·•1�..::
. '._:
1 1

tank. Vertically suspended artificial substrate was provided

•

as substrate. �� 0 �r�h� ,r� to C10> were stocked initially

2 b
representing 7.; c;��;;m :

100. OO-r----'"'I

s
u
R 8(J .. ()(>
V
I
V
A
L
60.00
-t4aa:!

-!tfl!!ili
itl+W

N '

u ,w#
M 40.00 #"
,-,
-lf't
'.i ..
E �:l!:i
,H+!i-#
�..\Hlf,. ·
•·

.uHt+
2t). (H)
+
#--!ttit-
T ..,.rii".,i..\HfT
s *+
o. oo,-+1----------,�-------........---------,--------�
0.00 40.00 8(): (1{) 120.00 160.00

DAYS

SURVIVAL
'l.
+ TOTAL NUMBER OF MOULTS
metamorphosis). This size crab represents the marketable Asian
product (Baliao et.al.,1981). An example of this sized crab is
shown in plate 7. The first C13 crab of 120mm (300g) was obtained
in 118 days (182 days from metamorphosis).
Figure 14 shows the growth rate of hatchery reared crabs in
this study was greater than that recorded by Heasman (1980). The
mean temperature experienced in both trials was similar i.e.
26.4 + 2.1 and 27.0 ± 0.5 °c respectively.
The intermoult duration for successive instars was similar
in both trials however the moult increment in the present study
was larger in the later stages (Cg - C12>- In fact the carapace
width of a C 12 crab in this trial was equivalent to that of a C13
crab in Heasmans trial.
The high sustained moult increments (30%) of the latter
stages in the present trial may be explained by differences in
feeding strategy and amount of food consumed. In our trial food
was supplied ad libitum in small quantities from 0800 to 2400
each day. The food was generally consumed within 1 2 hours
after which time more fresh food was provided. An average of 10%
(wet wt. food : total biomass) was consumed each day.
In Heasmans trial however food was provided in excess once
daily and uneaten food was removed about 12 hours later. The
mean amount of food consumed per day was 3.3% total body weight.
This 3-fold reduction in consumption when fed once daily to
excess (in this case prawn, fish and mussel) may indicate that
the quality of the food had deteriorated rapidly leading to
smaller consumtion rates. A deficiency in either quality or
quantity of food depresses the rate of crustacean growth by

31
Plate 6. Small-scale growout tank (4.5 sq.m). 40 onion
bags were suspended vertically through the water
column.

Plate 7. Stage 12 (C.W.- 100mm) hatchery reared female

mud crab Scylla serrata. This is a marketable sized
crab in Asia (Balioa et.al.1981)
 Figure 14. The growth rate of hatchery reared Juve�ile mud crabs
(present study and Heasman, 1980)
conditions. Crabs were reared at approximately 27°C

150.00

I
N
s
T 120.00
A

C
A
R 90.00
A
p
A
C
E 6(} c: (}()

D
T
''
u 30.00
m
m

0.00 75 ,,; ()() 15(}: i){) 225 .. (J(J 300.00

DAYS FROM METAMORPHOSIS

X PRESENT TRIAL
+ HEASMAN (1980)
either reducing the increment or lengthening the intermoult
(Hartnoll, 198 2 ). In this case with mud crabs it appeared that
only moult increment was affected.
Therefore by providing high quality food constantly,
consumption and consequently growth rates can be increased.
Table 8 shows that production (5580 kg/ha) in the present
trial was approximately 6 times greater than that of mud crabs
grown in experimental earthern ponds at SEAFDEC (Baliao et.al.,
1981). Crabs were able to be held at much higher initial stocking
density (7. 2 crabs/m 2 ) with relatively low mortality (36.4%) when
provided with artificial shelter whereas crabs grown in earthern
ponds at 2 crabs/m 2 sustained major mortality (69%) with
subsequent low net production (948 kg/ha).
The relative growth increment (g/day) of the SEAFDEC crabs
was greater than that of crabs grown in our trial. It is most
likely that the higher temperatures experienced in the earthern
ponds (25 34 °c) contributed to this by shortening the
intermoult duration.
Autotomy (the sacrificial loss of limbs during antagonistic
encounters) was experienced throughout the growout period and
increased as crabs became larger (figure 15). Walking and
swimming legs were autotomized more often than chelipeds. At the
termination of the trial 14.3% of the crabs were missing one
cheliped and 2 9.0% were missing other limbs.
The 14.3% missing chelipeds would not be suitable for market
as approximately 60% of the edible flesh is contained within the
claws. These crabs would have to be on grown to the next instar
when the claws would be regenerated (smaller than the original)

32
Table 8. Growth and productivity of mud crabs grown in tanks with
artificial substrate (present study) and in experimental
earthern ponds at SEAFDEC (Baliao et.al., 1981).

Trial S.D. % Mean Growout Relative Feed Product.

/ha. surv. C.W.mm Period growth conv. kg/ha
in. fin. days inc.
----------------------------------------------------------------
g/day

A 72000 64 45 89 119 0.82 6.1 5580

B 20000 31 52 97 90 1. 69 4.04 948
C 5000 88 52 107 90 2.28 1.72 908

A Present study
B & C - Treatments at SEAFDEC
S.D. - Stocking Density (crabs/ha.)
 Figure 15. The number(%) of juvenile mud crabs missini
appendages during 123 days of continued growout in a 4.5 m
fibreglass tank supplied with ar�ificial shelter. Crabs were
stocked initially at 7.2 crabs/m�.

Cl)
�
100
(C 90
�
u 80
� 70
0 60
I:ii:l
c.,
50
(C 40
foot
30
u 20
� 10
I:ii:l
� 0
1 14 24 38 67 77 100 108 123
DAYS IN GROWOUT TANK
or discarded if this was not economically viable.
At the end of the growout trial the population consisted of
crabs ranging in size (C.W.) from 67.8 - 1 2 0.0 mm or C10 - C13-
Therefore a proportion of the population had not reached a
marketable size. Recent research by Aiken and Waddy (1988) on
lobsters Homarus americanus has shown that they also display a
wide variation in growth rate when held communally in tanks.
When the lobsters were sorted into size groups and restocked
2
at 2 5/m at 60 day intervals the results showed that each group
reflected the growth pattern of the original population. That is
some of the lobsters in each group grew very rapidly while others
continued to grow slowly. This resulted in the faster growing
lobsters of the smaller group surpassing the slower growers of
the larger lobster group. This growth pattern was caused by
social pressure. That is, "growth rate for any individual is
determined by genetic potential and the relative size of other
individuals in the tank."
The size variability of the mud crabs may have been caused
by similar social pressure. Therefore the potential to increase
growth rates, by regular grading and restocking of ponds may
exist. This must be investigated further.

CONCLUSIONS

(1) The provision of vertically suspended artificial shelter

(onion bag) significantly increased survival and productivity of
mud crabs in experimental tanks by reducing moult related
cannibalism.

33
( 2) The mud crabs preferred the artificial material to sand for
both intermoult and moulting behaviour.

(3) The onion bags occupied depth and therefore increased

available pond substrate surface area. This allowed mud crabs to
be held at high stocking density (7/m 2) with minimal mortality.

(4) The floating onion bag system is easily established and

managed and does not interfere with harvesting. On the contrary,
extraction of bags could aid harvesting and sampling by capturing
inhabitant crabs.

(5) Growth rates were excellent with commercial sized crabs

(by Asian standards) obtained in 5 months from metamorphosis.
There is potential to increase this growth rate by regular
grading of crabs into size groups and growout at high (30 °c)
sustained temperatures.

(6) Juvenile mud crabs consumed food during both day and night
and food should therefore be provided continually each day.

(7) Quality of food can have a major effect on growth of mud

crabs. In order to overcome potential problems with deterioration
of food quality a feeding strategy of frequent. small quantities
should be adopted.

(8) The amount of food consumed by individual crabs each day was
variable and therefore regular observation of consumption rate
(using retrievable food trays) should be carried out and
adjustments made accordingly.

34
(9) It may be possible to achieve greater productivity in an
earthen pond with suspended artificial substrate. In the present
growout trial a sand or mud substrate was not provided. In a pond
situation this substrate would provide a further dimension for
seclusion as well as promote generation of natural benthic flora
and fauna. Supplementation of a formulated diet with naturally
occurring prey organisms may enhance overall nutrition which
could lead to faster growth rates.

Results of this study have shown the potential of mud crab

Scylla serrata culture in Northern Australia. Before any large­
scale production can be realized however, a commercial formulated
feed must be available. Sea Hatcheries Ltd. will be undertaking
further research into development of such a product as well as
continuing development of pond design and management techniques
for mud crab culture.

35
 REFERENCES

Aiken,D.E. and S.L.Waddy.1988. Strategies for max1m1z1ng growth

of communally reared juvenile American lobsters. World
Aquaculture 19(3) :61-63.
Baliao,D.D., E.M.Rodriguez and D.D.Gerochi.1981. Culture of the
mud crab Scylla serrata (Forskal) at different stocking
densities in brackishwater ponds. SEAFDEC. Aquaculture
Department Quarterly Research report Vol 5(1):10-15.
Bordner,C.E., L.R.D'Abramo, D.E.Conklin and N.A.Baum.1986.
Development and evaluation of diets for crustacean
aquaculture. Journal of the World Aquaculture society. Vol
17, Nos.1-4:44-51.
Brock,J.A.1982. Diseases (infectious and non-infectious),
metazoan parasites, predators and public health
considerations in Macrobrachium culture and fisheries. In
CRC Handbook of mariculture. Vol 1 Crustacean Aquaculture.
Ed. J.P.McVey. CRC Press Inc.:329-369.
Cowan,L.1984. Crab farming in Japan, Taiwan and the Phillippines.
Queensland Department of Primary Industries. Information
series Q184009.
Hartnoll,R.G.1982. In The Biology of Crustacea. Vol 2. ED.
L.G.Abele. Academic Press:111-196.
Heasman,M.P.1980. Aspects of the general biology and fishery of
the mud crab Scylla serrata (Forskal) in Moreton Bay,
Queensland. Ph.D. Thesis. University of Queensland.
Heasman,M.P. and D.R.Fielder.1983. Laboratory spawning and mass
rearing of the mangrove crab, Scylla serrata (Forskal), from
first zoea to first crab stage. Aquaculture, 34:303-316.
Hill,B.J., M.J.Williams and P.Dutton.1982. Distribution of
juvenile, subadult and adult Scylla serrata
(Crustacea:Portunidae) on tidal flats in Australia. Marine
Biology. Vol 69:117-120.
Holland,J.S., D.V.Aldrich and K.Strawn.1971. Effects of
temperature and salinity on growth, food conversion,
survival and temperature resistance of juvenile blue crabs,
Callinectes sapidus (Rathbun). Sea Grant Publication TAMU-
SG-71-222.
Liao,I.C. and Y.P.Chen.1982. Maturation and spawning of penaeid
prawns in Tunskang Marine Laboratory, Taiwan. In CRC
Handbook of Mariculture. Vol 1. Crustacean Aquaculture. Ed.
J.P.McVey. CRC Press Inc.:155-160.

36
Quinn,R.G. and D.R.Fielder.1978. A laboratory beach system for
prolonged maintenance of sand crabs, Mictvris
(Latreille,1806) and Scopimera (De Haan, 1833) (Decapoda,
Brachyura). Crustaceana 34(3):310-313.

37
Table 5. The effect of vertically suspended artificial bag
shelter on survival and growth of juvenile mud crabs.
Two replicates per treatment. An initial stocking
density of 25 crabs. The trial was run for 23 days.
No. Total Mean Crab Cumm. % Crabs
Surv. Biomass Crab Size Number Missing
Crabs (g) Weight Range Moults Limbs
(g) (g)

NO BAGS
REP 1 4 24.3 6.07 3.2-7.4 16 50
+2.0 C 7-C 8
REP 2 6 23.0 3.83 2.2-5.9 23 50
-----------------------------------------------------------------
+1.4 C5-C8
MEAN 5.0 23.7 4.95 19.5 50

PLUS BAGS
REP 1 15 76.0 5.07 2.1-11.1 39 20
+2.6 C1C9
REP 2 17 85.9 5.05 0.7-13.6 35 29.4
±4.0 C5-C9
MEAN 16.0 81.0 5.06 37.0 24.7
*** ** n.s * *
*** - significant at 0.1% level
** II
1 • 0% II

* II
5 • 0% II

38