Aquaculture Research, 2007, 38, 1539^1553 doi:10.1111/j.1365-2109.2007.01814.x

Improved techniques for rearing mud (Estampador 1949) larvae

Truong Trong Nghia1, MathieuWille2,Tran Cong Binh1,HoangPhuocThanh1,NguyenVanDanh1 & Patrick Sorgeloos2 1College of Aquaculture and Fisheries, Can Tho University, Can Tho City,Vietnam 2Laboratory of Aquaculture & Artemia Reference Center, Ghent University, Ghent, Belgium

Correspondence: T Trong Nghia, College of Aquaculture and Fisheries, Can Tho University,3rd February avenue, Campus II, Can Tho City,Vietnam. E-mail: [email protected]

Abstract rate from Z1to crab1^2 stages in large rearing tanks of10^15% (maximum 30%). A series of rearing trials in small 1L cones and large tanks of 30^100 L were carried out to develop opti- Keywords: Scylla paramamosain, rearing techni- mal rearing techniques for mud crab (Scylla parama- ques, water exchange, micro-algae, larval density, mosain) larvae. Using water exchange (discontinuous life food density,prophylaxis partial water renewal or continuous treatment through bio¢ltration) and micro-algae (Chlorella or Chaetoceros) supplementation (daily supplementation at 0.1^0.2 million cells mL À1 or maintenance at Introduction 1^2 millions cells mL À1), six di¡erent types of rearing systems were tried. The combination of a Aquaculture of mud , Scylla spp., contributes a green-water batch system for early stages and a recir- large proportion to the world production of the culating system with micro-algae supplementation (FAO 1999). Moreover, mud crabs represent a valuable for later stages resulted in the best overall perfor- component of small-scaled coastal ¢sheries in many mance of the crab larvae. No clear e¡ects of crab countries in tropical and subtropical Asia, for which stocking density (50^200 larvae L À1) and rotifer there has been a general trend of increased exploita- (30^60 rotifers mL À1)andArtemia density (10^ tion in recent years (Angell 1992; Keenan 1999a). In 20 L À1) were observed. A stocking density of Vietnam, the mud crab Scylla paramamosain is the 100^150 zoea 1 (Z1)L À1, combined with rotifer second most important marine species next to of 30^45 mL À1 for early stages and Artemia feeding shrimp, being cultured widely in the coastal area. at 10^15nauplii mL À1 for Z3^Z5 seemed to produce However, mud crab farming currently relies entirely the best performance of S. paramamosain larvae. Op- on the wild for seed stock and the main obstacle for timal rations for crab larvae should, however, be ad- expansion is the unavailability of hatchery-reared justed depending on the species, larval stage, larval seed (Liong 1992; Rattanachote & Dangwatanakul status, prey size, rearing system and techniques. A 1992; Keenan 1999a; Shelley & Field 1999; Mann, practical feeding schedule could be to increase live Asakawa, Pizzuto, Keenan & Brock 2001; Xuan food density from 30 to 45 rotifers mL À1 from Z1 to 2001). Z2 and increase the number of Artemia nauplii mL À1 Rearing techniques, disease and nutrition are the from10to15from Z3 to Z5. Bacterial disease remains three main areas of research that have supported one of the key factors underlying the high mortality commercial production of marine ¢sh and crusta- in the zoea stages. Further research to develop safe cean larvae (Sorgeloos & Le¤ger1992). These three as- prophylactic treatments is therefore warranted. Com- pects are to a large extent interconnected and bined with proper live food enrichment techniques, developing hatchery techniques for a‘new’species is application of these ¢ndings has sustained a survival not possible unless all the areas are addressed. The

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd 1539 Improved rearing techniques for mud crab larvae T T Nghia et al. Aquaculture Research, 2007, 38, 1539 ^ 1553 design of rearing systems covers more than purely (30 g L À1) and ammonia and nitrite levels. Every technical aspects. Sub-optimal rearing conditions other day, the crabs were placed in a 50 mLLÀ1 for- (e.g. physical stress, lack of oxygen or sub-optimal malin bath for 1h to reduce or prevent infestation of water quality) a¡ect larval health and can cause the eggs with fungi and bacteria. During egg incuba- mass mortality due to disease outbreaks. Similarly, tion, the crabs were not fed. system design in£uences (live) food quality and its One to two days before hatching, the berried availability to the predator larvae. female was moved to a 500 L ¢breglass tank. When There has been a great deal of progress in marine the hatching process was completed, larvae were larval rearing technology since its beginning in the selected based on their phototactic behaviour. Aera- 1960s (Shelbourne 1964; Howell, Day, Ellis & Baynes tion in the hatching tank was turned o¡ for several 1998). Many of these technical improvements devel- minutes and the active larvae swimming up to the oped over the past decades could be applied for mud surface were collected by gentle scooping. crab with some modi¢cations. An overview of the The larvae were then transferred to the rearing rearing systems currently applied for larviculture of containers. Acclimation was performed by placing mud crabs was presented by Davis (2003). Although the larvae in a 50 L plastic mesh bucket and slowly knowledge has been obtained from these systems, rinsing them with water from the larval rearing con- there is a need to further optimize rearing techniques tainers for 20^30 min before release. in order to maximize larval survival and quality. Furthermore, techniques should be adapted for each Scylla species (Keenan, Davie & Mann 1998; Keenan Food and feeding 1999b) in relation to local conditions (seawater source, status of hatchery management, local Micro-algae culture resources). The aim of this research is to adapt the Starting cultures of the micro-algae Chaetoceros calci- existing rearing systems for larviculture to mud crab trans and Chlorella vulgaris were maintained indoors species (S. paramamosain being the test case) and to with Walne solution in seawater of 30 g L À1 at 25 1C. improve other techniques in order to maximize larval Large-scale production was performed indoors in survival and quality. 500 L tanks under a transparent roof. A haemocyt- ometer was used to count micro-algal densities.

Materials and methods Rotifer culture and enrichment Source of larvae The same Brachionus plicatilis L-strain with a lorica Gravid crabs were bought from local markets and length and width of 164 Æ 22 and 120 Æ 22 mm, re- transported to the hatchery. Before stocking in the spectively, was used in all the experiments. Rotifers hatchery, the crabs were subjected to a bath of were cultured indoors in 100L ¢breglass tanks oper- 100 mLLÀ1 of a 40% formalin solution for 1h. The ated in a batch mode, following the procedure de- crabs were stocked individually in 100 L compart- scribed in Sorgeloos and Lavens (1996). Rotifers were ments of a roofed 2 Â 2 Â 0.5 m cement tank, initially grown on baker’s yeast, but later on fed s equipped with a bio¢lter. Rearing water of Culture Selco (INVE Aquaculture, Dendermonde 30 Æ 1gLÀ1 salinity was diluted from brine (90^ Belgium) before feeding to the larvae. Temperature 110 g L À1) with tap water and chlorinated before use. and salinity were controlled at 25 1Cand25gLÀ1 Ambient temperature £uctuated slightly around respectively. They were harvested through a 60 mm 28 1C. Every crab was fed a daily ration of 10^15g of screen and rinsed thoroughly. fresh marine squid, bivalve or shrimp meat alternately. Rotifers were enriched with micro-algae or arti¢- After 3^5 days of acclimation, unilateral eyestalk cial enrichment media before being fed to the crab ablation was applied to induce spawning. After larvae. Enrichment with Chlorella was performed at spawning, berried crabs were again subjected to a a density of 5 Â 10 6 cells mL À1 for 3 h (Dhert 1996). s 10 0 mLLÀ1 formalin bath for 1h and transferred to a Rotifers were also enriched with Dry Immune Selco s 70 L plastic tank connected to a bio¢lter for egg incu- (DIS , INVE Aquaculture), using two separate doses bation. Daily management consisted of siphoning out of 0.05g L À1 at a 3-h interval. Enrichment was per- waste materials and shedded eggs from the tank bot- formed at a density of 500rotifersmL À1. The water tom and controlling the temperature (30 1C), salinity in the enrichment vessel was slowly heated to

r 2007 The Authors 1540 Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al.

Table 1 Overview of larval rearing systems applied in this study based on the method of water exchange and micro-algae supplementation

Water exchange

Continuous water Discontinuous manual treatment through the Algae supplementation partial water renewal use of a biofilter

No micro-algae supplemented (indoors) Clear-Batch system Clear-Recirc system Micro-algae supplemented at low levels to provide Algae-Batch system Algae-Recirc system extra food for live preys (indoors or outdoors) Micro-algae supplemented at a high concentration Green-Batch system Green-Recirc system and self-sustainable under natural sunlight as an (Combination of Green-Batch and extra food for live prey and water conditioning (outdoors) Algae-Recirc system at early and late larval stages respectively)

29^30 1C to avoid exposing the rotifers to thermal Larval rearing experiments: objectives and shock when they were added to the larval rearing experimental design tanks. Before being fed to the larvae, enriched rotifers In experiments1,2 and 3, the e¡ect of di¡erent water were rinsed and re-suspended in clean seawater at exchange schemes and the addition of micro-algae the same temperature as the crab-rearing tanks. on larval survival and development were evaluated. In experiments 4^8, other culture aspects such as Artemia culture and enrichment Z1stocking density, live food density and the e¡ect of di¡erent prophylactic treatments were investigated. Artemia nauplii (Vinh Chau strain) were hatched as The water quality management schemes tested in ex- described byVan Stappen (1996). Both newly hatched periments 1^3 are summarized in Table 1. An over- or enriched Artemia nauplii were used in the experi- view of the experimental design and culture ments of this study. Artemia were enriched with conditions of all the experiments is presented in Chaetoceros in the same micro-algal density as for ro- Table 2. The small-scale experiments (1^30 L) were tifer enrichment.The nauplii were also enriched with s carried out in a temperature-controlled room DIS (using two separate doses of 0.3 g mL À1 at a 6-h (28^30 1C). The experiments in 100L tanks were interval).Water temperature and salinity were main- performed outdoors at ambient temperature (27^ tained at 30 1Cand30gLÀ1, respectively, during 31 1C). The source and the disinfection procedure of Artemia enrichment. The density of Artemia during the seawater for larval rearing were similar to those enrichment was 200 mL À1. Before feeding to the used for broodstock rearing. Formalin at a concentra- crab larvae, the Artemia were rinsed with disinfected tion of 20 mLLÀ1 was applied every other day as a seawater and suspended at a known density in prophylactic treatment in experiments1^6. seawater. Experiment1 Feeding Larval survival and growth in a clear water system Rotifers were fed to the crab larvae from 0 to 6 days with daily partial water exchange (Clear-Batch) was after hatch (DAH 0^6) [roughly corresponding to compared with those in a clear water recirculating zoea 1 (Z1)^Z2 stages]. Newly hatched Artemia or system (Clear-Recirc). In the ¢rst rearing system, Artemia meta-nauplii were o¡ered from DAH 6 (Z3 30^50% of the water was replaced daily. In the recir- stage) onwards. Rotifers and Artemia were added culating system, all rearing tanks were connected to daily at 30^45 and 5^10mL À1 to the rearing tank a central bio¢lter.Water was recirculated at a rate of respectively (experiments 1, 2, 3, 4, 7 and 8). For approximately100% of the tank volume every 3^4 h. experiments 5 and 6, live feed were fed at the Live food and crab larvae were retained in the rearing required prey densities based on the planned tanks with by a mesh screen of 70 and 300 mmdur- treatments. Whenever the crab larvae were fed en- ing the rotifer and Artemia feeding stage respectively. s riched live feed, algae- or DIS -enriched live feed Larger mesh screens (250 and 500^1000 mmforroti- were used on alternate days. fer and Artemia stage respectively) and higher £ow

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1541 mrvdraigtcnqe o u rblarvae crab mud for techniques rearing Improved 1542

Table 2 Overview of the experimental conditions and water quality parameters (mean Æ standard deviation) in experiments 1^8

Rearing Container No. of Stocking Ammonium Nitrite Experiments Factor system volumeà (L) replicates density (Z1L À1) (mg L À1 )w (mg L À1 )w

1 Rearing system Clear-Batch 30 5 50 0.35 Æ 0.14a 0.14 Æ 0.11a Clear-Recirc 0.03 Æ 0.07b 0.11 Æ 0.09a b b ora Compilation Journal 2 Rearing system Clear-Recirc 100 8 100 0.02 Æ 0.04 0.04 Æ 0.01 Algae-Recirc 0.07 Æ 0.08b 0.10 Æ 0.08b Green-Batch 1.72 Æ 1.30a 0.57 Æ 0.33a 3 Rearing system Green-Batch 100 4 100 1.54 Æ 1.50a 0.43 Æ 0.13a

Green-Recirc 0.11 Æ 0.07b 0.15 Æ 0.10b Nghia T T 4 Z1 density Green-Batch 100 3 50 1.50 Æ 1.09a 0.53 Æ 0.39a 100 1.07 Æ 0.62a 0.43 Æ 0.35a r a a 150 0.98 Æ 0.68 0.43 Æ 0.29 al et 07BakelPbihn Ltd, Publishing Blackwell 2007 200 0.73 Æ 0.48a 0.34 Æ 0.26a . 5 Rotifer density (30, 45 Clear-Recir 30 5 50 0.06 Æ 0.05 0.06 Æ 0.06 and 50 rotifers mL À 1 ) 6 Artemia density (10, 15 Clear-Recir 30 5 100 0.04 Æ 0.05 0.13 Æ 0.09 and 20 Artemia mL À 1 ) 7 Prophylactic treatment (control, Clear-Batch 1 4 100 ND ND qautr eerh 2007, Research, Aquaculture formalin and oxytetracycline) 8 Duration of direct ozonation (control: Clear-Batch 1 3 100 ND ND without ozonation and ozonation for

qautr Research, Aquaculture 2, 4, 6, 8 and 10 min)

ÃCylindro-conical ¢breglass tanks of 30^100L and plastic cones of 1L. wValues within an experiment in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05). For a description of rearing systems, refer to Table 1.

r ND, not determined. 07TeAuthors The 2007 38, 38, 59^1553 ^ 1539 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al. rates were used upon daily £ushing out of uneaten (experiment 4) or a Clear-Recirc system (experiments live food and waste. 5 and 6) as described above.

Experiment 2 Experiment 7

A Clear-Recirc system was compared with two In experiment 7, the e¡ect of prophylactic chemicals systems where micro-algae were added. Rearing con- on the survival of the larvae was investigated. Three ditions for the Clear-Recirc system were similar to treatments, consisting of a control (no chemicals those described in experiment 1. In the Algae-Recirc used), dailyaddition of formalin at 20 mLLÀ1and dai- system, micro-algae were added daily to the recircu- ly addition of oxytetracycline at 10mg L À1,wererun lating system at a low concentration ranging from 0.1 in 1L plastic cones. All cones were placed in a water to 0.2 million cells mL À1. The operation of the rear- bath in order to maintain the rearing temperature at ing tanks was similar to the Clear-Recirc treatment. 30 1C. Water was replaced almost completely daily. In the Green-Batch treatment, a classical ‘green- Upon water exchange, the survival was determined. water’system, micro-algae concentrations in the cul- ture tanks were kept at a tenfold higher level of Experiment 8 1^2 million cells mL À1. In this system, the culture tanks were initially only ¢lled to 50% of their capa- To avoid the use of drugs as a prophylaxis, direct ozo- city and gradually increased to 100% by the end of nation of the culture tanks was tested in aerated 1L the Z2 stage by adding water and algae daily. Later plastic cones. Ozone (O3) was injected directly via an on, 10^30% of the rearing water was replaced daily airstone into every larval rearing cone upon chan- by clean seawater and/or algae, depending on the ging water and feed daily. Six treatments with three density of micro-algae remaining in the rearing replicates were arranged consisting of a control tanks. Upon water exchange, uneaten live food was (without O3 application) and O3 injection for 2, 4, 6, also £ushed out through a mesh screen (mesh sizes 8 and 10min (equivalent to a residual O3 level in the À1 as described in experiment 1). The same amount of water of 0.06,0.12,0.15,0.17 and 0.19 mg L as mea- live food (30^45 rotifers mL À1 and 5^10 Artemia sured by a test kit upon ¢nishing the injection). Other nauplii mL À1) was fed in all the treatments. In the rearing conditions and daily management were simi- systems using algae, Chlorella was used for Z1^Z3 lar to those described for experiment 7. stages (which is unsuitable as a food source for Artemia); from Z4 onwards, Chlorella was gradually replaced with Chaetoceros. Evaluation criteria The survival rates in the experiments using large Experiment 3 (30^100 L) containers (experiments 1^6) were esti- In this experiment, a Green-Batch and a Green-Recirc mated by volumetric sampling. Depending on the system were set up in order to further evaluate the tank volume and the density of the surviving larvae, application of micro-algae on the performance of triplicate 300^1000mL samples were taken from crab larvae. The ¢rst rearing system was a batch sys- each tank. Megalopae (M) (DAH 15^18) and ¢rst tem with addition of high concentrations of algae as crabs (C1) (DAH 22) were counted individually.In ex- described in experiment 2. The second system con- periments 7 and 8 (using small cones), the average sisted of a combination of the Green-Batch system survival rate was calculated by individuallycounting for early crab stages (Z1^Z2) and a Algae-Recirc sys- all surviving larvae in each replicate. tem for later stages (Z3 onwards). Larval development was monitored every 3 days but daily in experiments 7 and 8 by identifying the average zoeal instar stage of a sample of larvae and Experiments 4^6 assigning it a value: Z1 51, Z2 52, etc. Megalopa Inthese experiments, the e¡ect of Z1stockingdensity stage was assigned a value of 6.To compare the larval (50, 100, 150 and 200 L À1, experiment 4), rotifer development in each treatment, an average larval feeding densities at 30, 45 and 60 mL À1 for Z1^Z2 stage index (LSI) was calculated from the average (experiment5)andArtemia densities at 10, 15 and LSI value of all replicate containers in the same treat- 20 mL À1 for Z3 onwards (experiment 6) was evalu- ment. For large containers (experiments 1^6), ¢ve or ated. These experiments were run in a Green-Batch 10 larvae (in 30 and 100 L tanks respectively) were

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1543 Improved rearing techniques for mud crab larvae T T Nghia et al. Aquaculture Research, 2007, 38, 1539 ^ 1553

Table 3 Experiment1: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae cultured in two di¡er- ent rearing systems

Days after hatch

Treatment 369121518

Survival rate (%)Ã Clear-Batch 85 Æ 6a 79 Æ 9a 70 Æ 6a 64 Æ 7a 42 Æ 6b 32 Æ 5b Clear-Recirc 84 Æ 4a 78 Æ 8a 72 Æ 5a 70 Æ 5a 63 Æ 9a 47 Æ 6a LSIÃ Clear-Batch 1.5 Æ 0.2a 2.7 Æ 0.1a 3.5 Æ 0.4a 4.0 Æ 0.0a 4.6 Æ 0.2a ND Clear-Recirc 1.5 Æ 0.2a 2.7 Æ 0.1a 3.6 Æ 0.3a 4.2 Æ 0.3a 4.8 Æ 0.1a ND

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05). For treatment descriptions, refer to Table 1. ND, not determined. sampled from each tank to calculate the average LSI. Batch system (both at Po0.01) (Table 3). Although The sampled larvae were staged under a dissecting slightly higher in the recirculating system, LSI was microscope. In experiment 8, using small containers, not signi¢cantly di¡erent between treatments. The all larvae were staged visually upon counting daily better larval performance in the Clear-Recirc system survival. was accompanied by signi¢cantly lower average In this research, six larval rearing systems were ammonia levels (Po0.01) and slightly lower nitrite applied for the experiments. Each rearing system levels (seeTable 2). had its own features, i.e. water quality and ‘ease of operation’. Therefore, these features in combination made up a treatment as a type of rearing system.They Experiment 2 were not considered as variables. OnDAH9,larvalsurvivalintheClear-Recircsystem was signi¢cantly lower (Po0.05) than in both treat- ments with micro-algae supplementation (Algae-Re- Statistical analysis circ and Green-Batch systems) (Table 4). On DAH 12,

One-way analysis of variance (ANOVA) was used to survival in the Clear-Recirc treatment was lower compare data. Homogeneity of variance was tested (Po0.05) than in the Algae-Recirc system, whereas with the Levene statistic (P or a value was set at the Green-Batch system had intermediate results. 0.05). If no signi¢cant di¡erences were detected be- The LSI values on DAH 15 show a similar trend, tween the variances, the data were submitted to a although not signi¢cantly di¡erent. one-way ANOVA. Tukey’s honestly signi¢cant di¡er- The average levels of ammonia and nitrite in the ence post hoc analysis was used to detect di¡erences Clear-Recirc and Algae-Recirc systems were signi¢- between means and to indicate areas of signi¢cant cantly lower (Po0.01) than those in the Green-Batch di¡erence. If signi¢cant di¡erences were detected system (see Table 2). In the Green-Batch system, between variances, data were transformed using the peaks of ammonia and nitrite concentrations of 3 À1 arcsine-square root (for percentage data, i.e. survival and 1mg L , respectively, were recorded at the end rate) or logarithmic transformations (for LSI value) of the experiment. (Sokal & Rohlf 1995). All analyses were performed using the statistical program STATISTICA 6.0. Experiment 3

Table 5 presents the larval performance of the crab Results larvae cultured in two rearing systems. The survival rates and LSI values of both rearing systems were not Experiment 1 signi¢cantly di¡erent. However, the survival rates on Survival in the Clear-Recirc system at Z4^Z5 stages later days (from DAH 12^22) in the Green-Recirc sys- on DAH 15 and in the megalopa stage on DAH 18 tem tended to be higher than those in the Green- was signi¢cantly higher than those in the Clear- Batch system. The bio¢lter had a positive impact on

r 2007 The Authors 1544 Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al.

Table 4 Experiment 2: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae cultured in three di¡erent rearing systems

Days after hatch

Treatment 3691215

Survival rate (%)Ã Clear-Recirc 74 Æ 12a 63 Æ 7a 44 Æ 6b 26 Æ 11b 8 Æ 7a Algae-Recirc 74 Æ 12a 63 Æ 9a 61 Æ 7a 43 Æ 7a 15 Æ 8a Green-Batch 74 Æ 11a 67 Æ 9a 58 Æ 9a 35 Æ 9ab 13 Æ 6a LSIÃ Clear-Recirc 1.9 Æ 0.1a 2.7 Æ 0.2a 3.9 Æ 0.1a 5.0 Æ 0.1a 5.1 Æ 0.1b Algae-Recirc 2.0 Æ 0.1a 2.8 Æ 0.3a 4.0 Æ 0.1a 5.0 Æ 0.1a 5.6 Æ 0.2a Green-Batch 2.0 Æ 0.1a 2.8 Æ 0.2a 4.0 Æ 0.1a 5.0 Æ 0.1a 5.1 Æ 0.1ab

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05). For treatment descriptions, refer to Table 1.

Table 5 Experiment 3: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae cultured in two dif- ferent rearing systems

Days after hatch

Treatment 369121522

Survival rate (%)Ã Green-Batch 94 Æ 6a 88 Æ 9a 80 Æ 3a 66 Æ 15a 44 Æ 20a 9 Æ 1a Green-Recirc 94 Æ 6a 89 Æ 8a 80 Æ 5a 68 Æ 11a 56 Æ 11a 12 Æ 3a LSIÃ Green-Batch 1.4 Æ 0.3a 2.7 Æ 0.1a 3.8 Æ 0.4a 5.0 Æ 0.0a 5.2 Æ 0.2a ND Green-Recirc 1.4 Æ 0.2a 2.6 Æ 0.2a 3.9 Æ 0.3a 5.0 Æ 0.0a 5.3 Æ 0.1a ND

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05). For treatment descriptions, refer to Table 1. ND, not determined. water quality in the second part of the experiment, that LSI was always the lowest in treatment with signi¢cantly reduced ammonia (Po0.05) and 30 rotifers mL À1 and the highest in treatment nitrite (Po0.01) concentrations as a consequence 60 rotifers mL À1. (seeTable 2).

Experiment 6 Experiment 4 Table 8 presents the survival and development rate of Table 6 shows the survival and development rate of crab larvae fed Artemia nauplii at three densities (10, crab larvae stocked at four di¡erent Z1 densities (50, 15 a n d 20 m L À1) from the Z3 stage onwards. No sig- 10 0, 150 a n d 20 0 L À1). The survival rates were not ni¢cant di¡erences were observed between the treat- signi¢cantly di¡erent among treatments. Only on ments. There seemed, however, to be a weak trend DAH 6 was a negative correlation between LSI and towards higher survival and LSI with increasing larval density observed. Artemia density towards the end of the trial.

Experiment 5 Experiment 7

Table 7 shows the survival rates and the LSI values of Table 9 shows the survival and development rate to crab larvae fed three di¡erent rotifer densities in the the megalopa stage (DAH 22) of larvae receiving dif- Z1^Z2 stages. No signi¢cant di¡erences were found ferent prophylactic treatments. From DAH 6 on- for any of the parameters. It can, however, be noticed wards, the survival rate of larvae in the treatment

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1545 Improved rearing techniques for mud crab larvae T T Nghia et al. Aquaculture Research, 2007, 38, 1539 ^ 1553

Table 6 Experiment 4: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae stocked at four di¡er- ent Z1densities (Z1L À1)

Days after hatch

Treatment 36 9121522

Survival rate (%)Ã 50 79 Æ 9a 56 Æ 19a 42 Æ 16a 31 Æ 17a 28 Æ 12a 4 Æ 6a 100 80 Æ 7a 74 Æ 6a 71 Æ 10a 56 Æ 11a 45 Æ 8a 5 Æ 4a 150 79 Æ 2a 57 Æ 12a 45 Æ 9a 31 Æ 12a 28 Æ 10a 5 Æ 1a 200 85 Æ 5a 53 Æ 17a 42 Æ 8a 34 Æ 3a 30 Æ 5a 5 Æ 1a LSIÃ 50 1.7 Æ 0.2a 3.0 Æ 0.1a 3.9 Æ 0.1a 5.0 Æ 0.1a ND ND 100 1.8 Æ 0.1a 3.0 Æ 0.1ab 4.0 Æ 0.0a 5.0 Æ 0.1a ND ND 150 1.8 Æ 0.2a 3.0 Æ 0.1ab 3.9 Æ 0.2a 5.0 Æ 0.1a ND ND 200 1.8 Æ 0.2a 2.7 Æ 0.1b 3.7 Æ 0.2a 4.8 Æ 0.1a ND ND

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05). ND, not determined.

Table 7 Experiment 5: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae-fed three di¡erent rotifer densities (rotifers mL À1)fromday0today6afterhatch

Days after hatch

Treatment 3691215

Survival rate (%)Ã 30 89 Æ 7a 53 Æ 10a 30 Æ 8a 13 Æ 7a 10 Æ 7a 45 87 Æ 6a 58 Æ 9a 35 Æ 7a 18 Æ 9a 14 Æ 8a 60 87 Æ 5a 55 Æ 7a 32 Æ 6a 16 Æ 11a 11 Æ 10a LSIÃ 30 1.8 Æ 0.2a 2.7 Æ 0.1a 3.6 Æ 0.2a 3.8 Æ 0.2a 4.0 Æ 0.4a 45 1.8 Æ 0.2a 2.8 Æ 0.2a 3.8 Æ 0.1a 3.9 Æ 0.1a 4.3 Æ 0.5a 60 1.8 Æ 0.2a 2.8 Æ 0.2a 3.8 Æ 0.2a 4.0 Æ 0.1a 4.4 Æ 0.5a

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05).

using antibiotics was signi¢cantly higher than those the survival of treatment Ozon2 became the highest in the remaining treatments.The survival rates of the (52%); however, this was not statistically di¡erent control and formalin treatments were similar on from the control (25%). On DAH9, there were no sig- most days. From DAH 6 onwards, the LSI values of ni¢cant di¡erences in survival in all treatments. the formalin treatment were generally higher than for the other treatments (not always signi¢cant). On DAH 15 and 18, the antibiotic treatment resulted in Discussion lower LSI values (Po0.01). Rearing system

Recirculation Experiment 8 Water recirculation through a bio¢lter in the Clear- Table 10 presents the survival at the Z2, Z3 and Z4 Recirc system positively a¡ected larval performance stages of larvae that were daily exposed to di¡erent compared with manual partial water replacement in levels of O3.OnDAH3(Z2stage),survivalinthecon- the Clear-Batch system (experiment 1). The advan- trol and the treatment with O3 injection for 2 min tages of recirculating systems in commercial ¢sh (Ozon2) was higher (67^78%) compared with the and larval production have been proven other treatments (24^45%). On DAH 6 (Z3 stage), before for other species. Research into recirculating

r 2007 The Authors 1546 Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al.

Table 8 Experiment 6: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae-fed three di¡erent instar-1Artemia densities (Artemia mL À1) from 6 days after hatch

Days after hatch

Treatment 91215

Survival rate (%)Ã 10 26 Æ 10a 12 Æ 5a 8 Æ 3a 15 30 Æ 6a 13 Æ 7a 10 Æ 6a 20 32 Æ 8a 19 Æ 9a 18 Æ 9a LSIÃ 10 3.1 Æ 0.2a 3.7 Æ 0.4a 4.3 Æ 0.5a 15 3.1 Æ 0.1a 3.7 Æ 0.2a 4.3 Æ 0.5a 20 3.2 Æ 0.1a 3.8 Æ 0.3a 4.6 Æ 0.3a

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05).

Table 9 Experiment 7: survival rates and larval stage index (LSI) values of Scylla paramamosain larvae treated daily with prophylactic chemicals

Days after hatch

Treatment 3 6 9 12151822

Survival rates (%)Ã Control 85 Æ 3a 64 Æ 7b 48 Æ 8b 34 Æ 13b 28 Æ 11b 17 Æ 7b 9 Æ 5b Formalin 84 Æ 7a 66 Æ 8b 47 Æ 7b 34 Æ 12b 26 Æ 8b 13 Æ 10b 11 Æ 8ab Antibiotics 91 Æ 4a 80 Æ 2a 74 Æ 4a 66 Æ 8a 52 Æ 6a 34 Æ 3a 21 Æ 5a LSIÃ Control 1.8 Æ 0.1a 2.6 Æ 0.3a 3.4 Æ 0.2ab 4.5 Æ 0.3a 5.1 Æ 0.1a 5.5 Æ 0.1b 5.8 Æ 0.2a Formalin 1.8 Æ 0.1a 2.8 Æ 0.1a 3.7 Æ 0.1a 4.3 Æ 0.3a 5.1 Æ 0.1a 5.8 Æ 0.1a 6.0 Æ 0.1a Antibiotics 1.9 Æ 0.0a 2.7 Æ 0.2a 3.3 Æ 0.2b 4.3 Æ 0.1a 4.9 Æ 0.1b 5.3 Æ 0.2b 5.6 Æ 0.2a

ÃSurvival rates or LSI values in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05).

Table 10 Experiment 8: survival rates of Scylla paramamosain larvae treated daily by ozone for di¡erent durations of time (min)

Treatment DAH 3 (Z2) DAH 6 (Z3) DAH 9 (Z4)

Control 78 Æ 6a 25 Æ 9ab 9 Æ 6a Ozon2 67 Æ 6ab 52 Æ 14a 11 Æ 10a Ozon4 40 Æ 17abc 33 Æ 11ab 5 Æ 2a Ozon6 45 Æ 25abc 27 Æ 15ab 5 Æ 5a Ozon8 31 Æ 2bc 19 Æ 3b 0 Æ 0a Ozon10 24 Æ 15c 12 Æ 9b 0 Æ 1a

Survival rates in the same column followed by the same superscript letter are not statistically di¡erent (P  0.05). DAH, days after hatch; Z, zoea; Control, without ozonation; Ozon2, 4, 6, 8 and 10, duration of ozone injection from 2 to 10min, which is equivalent to 0.06, 0.12, 0.15, 0.17 and 0.19 mg L À1 of the residual ozone respectively. systems has also been identi¢ed as a priority for tain a high biological carrying capacity in relatively shrimp culture (Lawrence & Lee 1997). In these sys- little space (Quillere, Marie, Roux, Gosse & Morot- tems, water exchange is minimized through the use gaudry 1993; Twarowska, Westerman & Losordo of biological, chemical and/or mechanical ¢ltration 1997). For crab larviculture, recirculating systems to maintain good water quality continuously.As they also appear to warrant further investigation in order provide less stress and confer constant good water to decrease labour requirements and seawater con- quality to the larvae, these systems are able to main- sumption, providing a more stable culture medium

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1547 Improved rearing techniques for mud crab larvae T T Nghia et al. Aquaculture Research, 2007, 38, 1539 ^ 1553 and thus reducing larval stress. If the system design dually ¢ll up the tanks with fresh seawater, algae and is kept simple, recirculating systems could also be sui- rotifers than £ushing out old rotifers in the recircula- table for large-scale production. tion system. In the recirculating system, the young larvae may be prone to physical damage and may Role of supplemented micro-algae spend considerable energy trying to swim up against the current. Early crab larvae are delicate due to their The addition of micro-algae to the recirculation small size and the three long spines on the carapace systems resulted in both higher survival and faster that are easily damaged when they are entrapped on development in this study. Micro-algae have been the mesh screen during £ushing out of uneaten feed proven to be bene¢cial by various modes of action. in the recirculation system (Davis 2003). The nutri- They could help maintain the quality of live food. As tional e¡ect of micro-algae is probably also more pro- in the culture of marine ¢sh larvae, unconsumed nounced during the rotifer feeding stage than during rotifers may reside in the tanks for several days and the Artemia feeding stage. Furthermore, it is not ne- their nutritional value may become severely reduced cessary to recirculate water during these ¢rst days, (Makridis & Olsen 1999). Furthermore, according to as the concentrations of ammonia and nitrite are still these authors, poorly fed rotifers were more sensitive low. Using the Algae-Recirc system in later stages is to starvation than well-fed rotifers, as their nitrogen more favourable for reducing the increasing ammo- content decreased at a higher rate. nia and nitrite concentrations as more waste materi- Micro-algae also play an important role in stabiliz- al is produced by the crab larvae. Moreover, as the ing water quality via either ammonia uptake or oxy- larvae develop into more e⁄cient predators, feed is gen production (Tseng, Huang & Liao 1991). Because consumed faster, and maintenance of optimal feed the Clear-Recirc system already provided optimal quality is less of an issue. Many studies successfully water quality, it is unlikely that the stabilizing e¡ect applied a similar combined rearing technique due to on water quality is responsible for the improved per- its bene¢t for the larvae and convenience for man- formance in the algae-supplemented system. In agement, particularly for large rearing containers. batch culture systems, this e¡ect would probably be Under green-water culture conditions, water is not much more pronounced. A direct comparison exchanged for the ¢rst 3 days. Thereafter, water ex- between a green and clear water batch system was, change is slowly increased from 10^20% day À1 for however, not made in this study. Z2^Z3 to between 40% and 50% day À1 at the end of In a study on the e¡ect of Chlorella on the popula- the rearing cycle (Z4^M) (Mann, Asakawa & Pizzuto tion of luminous bacteriaVibrio harveyi, no luminous 1999; Quinitio, Parado-Estepa, Millamena, Rodriguez bacteria were recovered on days 2 and 3 in £asks & Borlongan 2001). In Japan, a mesocosm system is with Chlorella, while those without the micro-algae used for culturing larvae in larger tanks (410 m 3). still harboured luminous bacteria at day 3 (Tendencia The tanks are partially ¢lled with green water at Z1 & dela Pena 2003). Also, the diatom Chaetoceros has (20^25% volume), tanks are then ¢lled up with clean been shown to produce natural antibiotics and high seawater during the course of the Z2^Z3 stages and concentrations of this marine diatom will eliminate during the Z4 and M stages water is exchanged on a Vibrio vulni¢cus and other pathogenic bacteria, which £ow-to-waste basis (Hamasaki, Suprayudi & Takeuchi contribute to the propagation of viruses in the shrimp 2002). production environment (Wang 2003). In conclusion, micro-algae in mud crab larval rear- ing may playa role in improving and maintaining live Other rearing techniques food quality and controlling bacteria levels. Z1stocking density Choice of system No signi¢cant e¡ect of larval density was observed In experiment 3, the Green-Recirc system (which is a from 50 to 200 Z1 L À1. This would suggest that the combination of a Green-Batch system during the ro- larvae can be grown at 200 Z1L À1. Variation in the tifer feeding stage and a Algae-Recirc system there- ¢nal survival between replicate tanks also seemed to after) seemed to be better than the Green-Batch decrease at higher densities. For S. paramamosain, system. The Green-Batch system seems to be more Djunaidah, Mardjono, Wille, Kontara and Sorgeloos appropriate for early stages of crab larvae (Z1^Z2) as (2001) found a tendency of increased survival to Z5 it is less stressful for the early zoeae and easier to gra- as a function of the Z1 stocking density (i.e. survival

r 2007 The Authors 1548 Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al. rates of 27%, 39% and 63% being obtained at densi- tion by larval stages in this way compensates for ties of 50,75 and100 Z1L À1 respectively). Baylon and the increased ingestion of crab larvae as they grow Failaman (1999) also reported higher survival and (Baylon, Bravo & Manigo 2004). For early larvae, metamorphosis of at 50 Z1L À1 com- however, food amount cannot be reduced to their pared with lower densities of 10 and 25 Z1L À1.In- maximum ingestion potential as they are quite ine⁄- creased survival at higher larval densities somehow cient predators and therefore might require a mini- seems contradictory. However, indirectly, food ration mal density to maximize encounter. might be responsible. Excess food in treatments with Similar to our study,most studies investigating the low larval densities may pollute the water and may e¡ect of rotifer density added the live food in one sin- thus cause mortality. In our study, we noted higher gle ration. Under these circumstances, theoretic den- concentrations of ammonia and nitrite in the treat- sities are only attained upon feeding and gradually ment having 50 Z1L À1 (see Table 2). For the highest decrease as larvae consume the prey. Optimal live stocking densities tested in our study (200 Z1L À1), food quantities cannot, however, be separated from the larval development rate seemed slightly impaired. feeding frequency.Because zoea larvae can consume This high stocking density may have caused competi- their optimal ration within 1h, Genodepa, Southgate tion for feed, resulting in slower development. There- and Zeng (2004) suggested that they can be fed once fore, Z1 stocking densities in the range of a day. Because of the severe reduction in the nutri- 100^150 Z1L À1 might be optimal. tional value of rotifers with longer retention times in rearing containers (Makridis & Olsen 1999) and the fact there is a minimum prey density needed for the Rotifer density for feeding early larval stages (Z1^Z2 passive feeding behaviour of zoea larvae (Heasman & stages) Fielder1983; Zeng & Li1999), the interaction between Although there was a trend towards increased survi- the optimal ration and feeding frequency should be valand growthwith increasing rotiferdensity,no sig- further investigated. ni¢cant di¡erences in larval survival or growth were found between the di¡erent rotifer densities tested. Artemia for feeding later larval stages (from Z3 Although not signi¢cant, the highest survival was onwards) generally observed at 45 rotifers mL À1; while a den- sity of 60 rotifers mL À1 resulted in the fastest larval We found no di¡erence between feeding Z3 a daily development.The di¡erences were, however, not very feed ration of 10, 15 or 20 Artemia nauplii mL À1. marked, and moreover such high feeding rates might Especially in later larval stages (Z4^Z5), there was, be economically unrealistic. We can therefore con- however, a tendency towards higher survival with clude that feeding 30 rotifers mL À1 is enough for op- increasing ration. In this respect, it might be bene¢- timal larval performance. In practice, however, the cial to increase theArtemia density bycrab stage from intermediate density of 45 rotifers mL À1 was fre- 10 t o 15 m L À1.Highlivefeeddensitieswouldin- quently used for feeding early larval stages. Other crease the chance for early larvae to encounter and studies indicated that high rotifer densities capture feed organisms (Zeng & Li 1999) and there- (30^80 mL À1) are required for optimal growth and fore would improve the larval performance (Brick survival of S. paramamosain (Djunaidah, Mardjono, 1974; Heasman & Fielder 1983; Quinitio et al. 2001). Lavens & Wille 1998; Zeng & Li 1999) and S. serrata On the other hand, older larval have a higher inges- (Suprayudi, Takeuchi, Hamasaki & Hirokawa 2002). tion capacity.Optimal rations should therefore be de- For S. paramamosain larvae, feeding 30 and termined for each larval stage separately. In this 60 rotifers mL À1 resulted in a signi¢cantly higher respect, studies on individual larvae are very useful survival compared with feeding only to determine prey consumption. According to our 15 ro t i f e r s m L À1 (Djunaidah et al. 2001). These previous experiments (Nghia 2004), each Z3, Z4, Z5 authors found that the individual dry weight of Z5- and megalopa larva was capable of consuming on fed 15rotifers mL À1 was signi¢cantly lower than average 15, 25, 37 and 114 newly hatched those of Z5 fed with higher rotifer densities. Practi- Artemia day À1 respectively. Therefore, at a stocking cally, feeding 30 rotifers mL À1 at Z1 and increasing of 100 larvae L À1, the dailyArtemia feeding densities gradually to 45 mL À1 at Z2 proved to be su⁄cient theoretically should be at least 1.5, 2.5, 3.7 and for a stocking densityof100Z1L À1inourtrialsinlar- 11.4 m L À1for Z3, Z4, Z5 and megalopa stages respec- ger rearing tanks (500^1000L). Increasing the ra- tively. For Z1, Z2 and Z3 stages of S. serrata,the

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1549 Improved rearing techniques for mud crab larvae T T Nghia et al. Aquaculture Research, 2007, 38, 1539 ^ 1553 number of Artemia nauplii ingested by the larvae at a culture systems used in other experiments.The small lower food density of 2.5 mL À1 was comparable to nauplii size of the Artemia strain (Vinh Chau strain) that at 5 mL À1, and for Z4^Z5, at 5 mL À1,itwas used in our study could be another reason that led comparable to 10mL À1 (Baylon et al. 2004). In that to increased ingestion. In practice, (larger sized) study, Artemia was, however, co-fed with rotifers at a highly unsaturated fatty acid-enriched Artemia were density of 15^20 mL À1.IfArtemia was the only food, normally used in order to reduce the prey amount to the optimal Artemia ration would therefore probably 5^10mL À1. Megalopa probably should be fed more be higher than 2.5^5 mL À1. In another study on S. frequently and, towards the end of that developmen- serrata, a daily optimum food concentration of 10 Ar- tal stage, a non-moving food may be better. temia nauplii mL À1 was established for zoea survival (Brick1974). In the mass seed production of S. serrata, Prophylactic chemicals newly hatched Artemia are given starting late Z2 at 0.5^3 mL À1 and 5^7-day-old Artemia are routine fed Laboratory cultures of crab larvae often su¡er severe from late Z5 to early megalopa (Quinitio & Parado- mortality from disease, particularly from epibiotic Estapa 2003). Older Artemia provided a larger-sized bacteria and larval mycosis (Armstrong, Buchanan prey for zoeae to megalopae and hence, the density & Caldwell1976; Hamasaki & Hatai1993a, b). A study was reduced. on S. serrata indicated a signi¢cantly higher survival For megalopae of S. paramamosain, we found a up to DAH 7 (over 90%) when using oxytetracycline, threefold higher number of ingested newly hatched whereas almost complete mortality occurred in the Artemia nauplii compared with Z5 (114 and 37 control treatment (Mann 2001). The author consid- Artemia respectively) for a similar prey density ered that potentially up to 80% of the larval mortality (Nghia 2004). This means that megalopae are vora- could be attributed to bacteriological causes. The re- cious predators, capable of chasing their preyactively sults of our study also indicated that bacteria are, and consume large amounts of feed in a short time. more than any other factor tested, a main cause of From this, it could be bene¢cial if megalopae are fed larval mortality. Antibiotics more than doubled sur- frequently smaller rations in order to optimize feed vival up to the crab stage. quality and reduce cannibalism. Genodepa et al However, antibiotics have not always been used in (2004) similarly indicated that in contrast to earlier a responsible manner in aquaculture. A major conse- larval stages, which can be fed once per day, S. serrata quence of using antibiotics has been the proliferation megalopae may need to be fed more often to maxi- of resistant bacteria and the transmission of resis- mize ingestion. These authors found no signi¢cant tance to other bacterial species (Benson 1998). The di¡erences in the ingestion rate of megalopae fed mi- development of antibiotic resistance by pathogenic crobound diets at rations ranging from 12.5% to bacteria is considered to be one of the most serious 100% of the standard ration (equivalent to 5 Artemia risks to human health at the global level (FAO 2002). nauplii mL À1 in 1h). Baylon et al (2004) also found a Formalin is more acceptable than antibiotics as it high increase in Artemia ingestion in the ¢rst few shows no accumulation in tissues (Jung, days of the planktonic phase of the megalopa stage. Kim, Jeon & Lee 2001). Recently, however, Japan has Later on, megalopae become more benthic as they strictly banned the use of formalin in aquaculture as prepare for the second metamorphosis to ¢rst crab. it may cause cancer in humans, reduces oxygen Swimming Artemia are no longer accessible, and levels in the water and causes algae to die o¡ (VASEP minced shrimp or mussel meat are a more suitable 2003). Moreover,in our experiments, formalin did not feed. signi¢cantly improve larval survival compared with In conclusion, a ration of 10 Artemia nauplii mL À1 the negative control. Pathogenic bacteria are consid- appears to be su⁄cient for the optimal performance ered to be one of the most serious causes for the high of Z3 larvae. An increase in prey density in the Z4^Z5 mortality of early crab larvae. It can be safely as- stage may, however, be bene¢cial. These food sumed that all inputs (seawater, broodstock, live feed amounts appear to be higher than what most other and daily management in hatcheries) into the cul- studies recommend (2.5^10 Artemia nauplii mL À1; ture tank are potential sources of infection (Black- Brick1974; Baylon et al.2004). Perhaps the recirculat- shaw 2001). Strict hygiene at all steps is always ing system used in this study resulted in a greater loss advised for hatchery activities. However, this advice of prey organisms (e.g. more Artemia were entrapped is not always followed, especially in backyard on the over£ow screen) than in the small batch hatcheries. Therefore, other techniques should be

r 2007 The Authors 1550 Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al. investigated as alternatives for the use of chemicals. products, moreover, are not encouraged for commer- Ozonation and probiotics could be interesting in this cial mud crab larviculture as they are unsafe. Direct respect (Davis 2003; Nghia 2004). Ozone is a power- ozonation as an alternative to prophylactic chemicals ful oxidant and is becoming more and more popular is worth investigating. in various aquaculture systems for disinfection and improving water quality by oxidation of inorganic and/or organic compounds (Tango & Gagnon 2003). Acknowledgments In our study, direct application did not signi¢cantly improve survival compared with a negative control. This study was supported by the European Commis-

However, there was a tendency of a residual O3 con- sion (INCO-DC) through the project grant No ICA4- centration of 0.06 mg L À1 to improve larval perfor- CT-2001-10022 ‘Culture and management of Scylla mance. Variability within this treatment was, species’, the Flemish Inter-University Council (VLIR- however, very high, which could be indicative of the IUC) and the International Foundation for Science fact that dosing was not careful enough. Longer O3 (IFS) through the research grant agreement No A/ exposure times (4^10-min exposure, equivalent to 2505-1 ‘Improvement of larviculture of the mud crab À1 0.12^0.19 mg L residual O3) all decreased overall (Scylla paramamosain) in the Mekong Delta,Vietnam’. survival. These high O3 concentrations probably caused physical damage to the crab larvae. In Penaeus À1 monodon juveniles, 0.34^0.5 mg L residual O3 References caused loss of balance, immobility and destruction of the gill lamellar epithelium (Meunpol, Lopinyosiri Angell C.A. (1992) The mud crab In: Report of the Seminar on & Menasveta 2003). Ozone treatment should there- the Mud Crab Culture and Trade (BOBP/REP/51, Bay of fore be investigated further, with determination of Bengal Programme, Madras, India), Surat Thani,Thailand, 5^8 November1991 (ed. by C.A. Angell), 246pp. proper doses for each larval stage. Ultimately, the mi- Armstrong D.A., Buchanan D.V. & Caldwell R.S. (1976) A crobial £ora will need to be controlled and there is mycosis caused by Lagenidium spp. in laboratory reared evidence that this can be achieved using recirculat- larvae of the Dungeness crab, Cancer magister,andpossi- ing systems in which O3 treatment is combined with ble chemical treatments. Journal of Invertebrate Pathology the inoculation of the biological ¢lter with selected 28, 329^336. nitrifying and probiotic bacteria (Gatesoupe 1991; Baylon J.C. & Failaman A.N. (1999) Larval rearing of mud Rombaut, Suantika, Boon, Maertens, Dhert,Top, Sor- crab Scylla serrata in the Philippines. In: MudCrabAqua- geloos & Verstraete 2001). culture and Biology, Proceedings of an International Scienti- ¢c Forum, Darwin, Australia, 21^24 April 1997, ACIAR Proceedings No. 78 (ed. by C.P. Keenan & A. Blackshaw), pp.141^146. Australian Centre for International Agricul- Conclusions and suggestions tural Research, Canberra, Australia. Baylon J.C., Bravo M.E.A. & Manigo C. (2004) Ingestion of The combination of a green-water batch system for Brachionus plicatilis and Artemia salina nauplii by mud early stages and a recirculating system with micro- crab Scylla serrata larvae. Aquaculture Research 35,62^70. algae supplementation for later stages, a stocking Benson S.M. (1998) Reservoirs of resistance. Today’s Life À1 density of 100^150 Z1L , feeding density of 30^ Science10. 1 45 rotifers mL À for early stages and 10^15 Artemia BlackshawA.W.(2001)Larval culture of Scylla serrata:main- nauplii mL À1 for later stages are recommended for tenance of hygiene and concepts of experimental design. larval rearing of S. paramamosain. Asian Fisheries Science14,239^242. The optimal ration for crab larvae should, however, Brick R.W. (1974) E¡ect of water quality, antibiotics, phyto- be adjusted depending onvarious factors, e.g. species, plankton and food on survival and development of larvae larval stages, larval status, prey size, rearing system of Scylla serrata (Crustacea: ). Aquaculture 3, 231^244. and rearing techniques. A feeding regime with Davis J.A. (2003) Development of hatchery techniques for the frequent addition of small quantities of feed is worth mud crab Scylla serrata (Forsskl) in South Africa.PhDthe- investigating. sis, Faculty of Agricultural and Applied Biological Antibiotics improved larval survival, proving Sciences, Ghent University, Belgium,163pp. again that bacterial interference is one of the major Dhert P.(1996) Rotifers. In: Manual on the Production and Use causes of mortality. Formalin could not signi¢cantly of Live Food for Aquaculture. FAO Technical Paper No. 361 improve survival compared with the control. Both (ed. by P. Lavens & P. Sorgeloos), pp. 61^100. Food and

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1551 Improved rearing techniques for mud crab larvae T T Nghia et al. Aquaculture Research, 2007, 38, 1539 ^ 1553

Agriculture Organization of the United Nation, Rome, Keenan C.P. (1999a) Aquaculture of the mud crab, genus Italy. Scylla ^ past, present and future. In: MudCrabAquaculture Djunaidah I.S., Mardjono M., Lavens P. & Wille M. (1998) and Biology, Proceedings of an International Scienti¢c For- E¡ects of light and feeding regimen on culture perfor- um, Darwin, Australia, 21^24 April 1997, ACIAR Proceed- mance of mud crab (Scylla spp.) larvae. In: Culture of Por- ings No. 78 (ed. by C.P.Keenan & A. Blackshaw), pp. 9^13. tunid Crabs, Extended Abstracts of the International Forum, Australian Centre for International Agricultural Boracay, Philippines,1^4 December 1998 (ed. by C. Keen- Research, Canberra, Australia. an, O. Millamena & E.T. Quinitio), pp. 27^28. SEAFDEC, Keenan C.P. (1999b) The fourth species of Scylla.In:Mud Tigbauan, Philippines. Crab Aquaculture and Biology, Proceedings of an Interna- Djunaidah I.S., Mardjono M.,Wille M., Kontara E.K. & Sorge- tional Scienti¢c Forum, Darwin, Australia, 21^24 April loos P. (2001) Investigations on standard rearing techni- 1997, ACIAR Proceedings No. 78 (ed. by C.P. Keenan & A. ques for mass production of mud crab Scylla spp. seed. A Blackshaw), pp. 48^58. Australian Centre for Interna- research review. In: Mud Crab Rearing, Ecology and Fish- tional Agricultural Research, Canberra, Australia. eries, Book of Abstracts of 2001Workshop, Can Tho Univer- Keenan C.P., Davie P.J.F.& Mann D.L. (1998) A revision of the sity, Vietnam, 8^10 January 2001 (ed. by L. Le Vay, P. genus Scylla de Haan,1833 (Crustacea: : Brach- Sorgeloos, M. Wille, T.T. Nghia & V.N. Ut), p. 11. Institute chyura: Portunidae). The Ra¥es Bulletin of Zoology 46, for Marine Aquaculture (IMA) ^ Can Tho University, Can 217^245. Tho City,Vietnam. Lawrence A.L. & Lee P.G. (1997) Research in the Americas. FAO (1999) Aquaculture production statistics 1988^1997.FAO In: Crustacean Nutrition (ed. by L. D’Abramo, D. Conklin ¢sheries Circular No.815, Revision11, Rome, Italy,203pp. & D. Akiyama), pp. 566^587.World Aquaculture Society, FAO (2002) The state of world ¢sheries and aquaculture. See Baton Rouge, LA, USA. http://www.fao.org/docrep/ Liong P.C. (1992) The fattening and culture of mud crab Gatesoupe F.J. (1991) The e¡ect of three strains of lactic bac- (Scylla serrata) in Malaysia. In: Report of the Seminar on teria on the production rate of rotifers, Brachionus plicati- the Mud Crab Culture andTrade (BOBP/REP/51, Bay of Ben- lis, and their dietary value for larval turbot Scopthalmus galProgramme,Madras,India), Surat Thani,Thailand,5^8 maximus. Aquaculture 96,335^342. November1991 (ed. by C.A. Angell), pp.185^190. Genodepa J., Southgate P.C. & Zeng C. (2004) Diet particle Makridis P.& OlsenY. (1999) Protein depletion of the rotifer size preference and optimal ration for mud crab, Scylla Brachionus plicatilis during starvation. Aquaculture 174, serrata, larvae fed microbound diets. Aquaculture 230, 343^353. 493^505. Mann D. (2001) The in£uence of microbiology on the suc- Hamasaki K. & Hatai K. (1993a) Experimental infection in cess of mud crab larval culture. In: Mud Crab Rearing, the eggs and larvae of the swimming crab Portunus tritur- Ecology and Fisheries, Book of Abstracts of 2001 Work- biculatus and the mud crab Scylla serrata with seven fun- shop, Can Tho University, Vietnam, 8^10 January 2001 gal strains belonging to Lagenidiales. Bulletin of the (ed. by L. Le Vay, P. Sorgeloos, M. Wille, T.T. Nghia & V.N. Japanese Society of Scienti¢c Fisheries 59, 1059^1066. Ut), pp. 18^21. Institute for Marine Aquaculture (IMA) ^ Hamasaki K. & Hatai K. (1993b) Prevention of fungal infec- Can Tho University, Can Tho City,Vietnam. tion in the eggs and larvae of the swimming crab Portunus Mann D., Asakawa T. & Pizzuto M. (1999) Development of a trituberculatus and the mud crab Scylla serrata by bath hatchery system for larvae of the mud crab Scylla serrata treatment with formalin. Bulletin of theJapanese Society of at the Bribie Island Aquaculture Research Center. In: Mud Scienti¢c Fisheries 59,1067^1072. Crab Aquaculture and Biology, Proceedings of an Interna- Hamasaki K., Suprayudi M.A. & Takeuchi T. (2002) Mass tional Scienti¢c Forum, Darwin, Australia, 21^24 April mortality during metamorphosis to megalops in the seed 1997, ACIAR Proceedings No. 78 (ed. by C.P. Keenan & A. production of mud crab Scylla serrata (Crustacea, Deca- Blackshaw), p. 153^158. Australian Centre for Interna- poda, Portunidae). Fisheries Science 68,1226^1231. tional Agricultural Research, Canberra, Australia. Heasman M.P. & Fielder D.R. (1983) Laboratory spawning Mann D.L., Asakawa T., Pizzuto M., Keenan C.P.& Brock I.J. and mass larval rearing of the mangrove crab Scylla ser- (2001) Investigation of an Artemia-based diet for larvae of rata (Forsskl), from ¢rst zoea to ¢rst crab stage. Aquacul- the mud crab Scylla serrata. Asian Fisheries Science 14, ture 34,303^316. 175 ^ 184. HowellB.R.,DayO.J.,EllisT.&BaynesS.M.(1998)Early Meunpol O., Lopinyosiri K. & Menasveta P. (2003) The stages of farmed ¢sh. In: The Biology of Farmed Fish (ed. e¡ects of ozone and probiotics on the survival of black by K.D. Black & A.D. Pickering), She⁄eld Academic Press, tiger shrimp (Penaeus monodon). Aquaculture 220, She⁄eld, UK, 415pp. 437^448. Jung S.H., Kim J.W., Jeon I.G. & LeeY.H. (2001) Formaldehyde Nghia T.T. (2004) Optimization of mud crab (Scylla paramamo- residues in formalin-treated olive £ounder (Paralichthys sain) larviculture inVietnam. PhD thesis, Faculty of Agri- olivaceus), black rock¢sh (Sebastes schlegeli), and seawater. cultural and Applied Biological Sciences, Ghent Aquaculture194,253^262. University, Belgium,192pp.

r 2007 The Authors 1552 Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 Aquaculture Research, 2007, 38, 1539^1553 Improved rearing techniques for mud crab larvae T T Nghia et al.

Quillere I., Marie D., Roux L., Gosse F. & Morot-Gaudry J.F. Tendencia E.A. & dela PeoØa M. (2003) Investigation of some (1993) An arti¢cial productive ecosystem based on a ¢sh: components of the greenwater system which makes it bacteria: plant association. 1. Design and management. e¡ective in the initial control of luminous bacteria. Aqua- Agriculture Ecosystems and Environment 47, 13^30. culture 218,115^119. Quinitio E.T., Parado-Estepa F., Millamena O.M., Rodriguez E. Tseng K.F.,HuangJ.S. & Liao I.C. (1991)Species control of mi- & Borlongan E. (2001) Seed production of mud crab Scylla croalgae in an aquaculture pond. Water Research 25, serrata juveniles. Asian Fisheries Science14,161^174. 1431^ 1437. Quinitio E.T. & Parado-Estapa F.D. (2003) Biology and Twarowska J.G., Westerman P.W. & Losordo T.M. (1997) hatchery of mud crabs. Scylla spp. SEAFDEC. Tigbauan, Water treatment and waste characterization evaluation Philippines, 42pp. of intensive recirculating ¢sh production system. Aqua- Rattanachote A. & Dangwatanakul R. (1992) Mud crab culture Engineering16,133^147. (Scylla serrata Forsskl) fattening in Surat Thani province. Van Stappen G. (1996) Artemia.In:Manual on the Production In: Report of the Seminar on the Mud Crab Culture andTrade and Use of Live Food for Aquaculture. FAO Technical Paper (BOBP/REP/51, Bay of Bengal Programme, Madras, India), No. 361 (ed. by P.Lavens & P.Sorgeloos), pp.101^170. Food Surat Thani, Thailand, 5^8 November, 1991 (ed. by C.A. and Agriculture Organization of the United Nation, Angell), pp.171^177. Rome, Italy. Rombaut G., Suantika G., Boon N., Maertens S., Dhert P.,Top VASEP. (2003) Japan banned on using formalin in aquacul- E., Sorgeloos P. & Verstraete W. (2001) Monitoring of the ture (in Vietnamese). Magazine of Vietnam Association of evolving diversity of the microbial community present in Seafood Exporters and Producers,May2003. rotifer culture. Aquaculture198,237^252. Wang J.K. (2003) Conceptual design of a microalgae-based Shelbourne J.E. (1964) The arti¢cial propagation of marine recirculating oyster and shrimp system. Aquacultural ¢sh. Advances in Marine Biology 2,1^83. Engineering 28,37^46. Shelley C. & Field D. (1999) The culture of Portunid crabs. Xuan T.T. (2001) The status and the prospect of mudcrab Info¢sh International 4, 46^49. culture in Ca Mau province, Vietnam. In: Mud Crab Sokal R.R. & Rohlf F.J. (1995) Biometry: The Principles and Rearing, Ecology and Fisheries, Book of Abstracts of Practice of Statistics in Biological Research,3rdedn. 2001 Workshop, Can Tho University, Vietnam, 8^10 W.H. Freeman and Company,NewYork, NY, USA,887pp. January 2001 (ed. by L. Le Vay, P. Sorgeloos, M. Wille, Sorgeloos P.& Lavens P.(1996) Manual on the Production and T.T. Nghia & V.N. Ut), p. 28. Institute for Marine Aqua- Use of Live Food for Aquaculture. Fisheries Technical Paper, culture (IMA) ^ Can Tho University, Can Tho City, Vol.361. Food and Agriculture Organization of the United Vietnam. Nation, Rome, Italy pp.9^100. Zeng C. & Li S. (1999) E¡ects of density and di¡erent combi- Sorgeloos P.& Le¤ger P.(1992) Improved larviculture outputs nations of diets on survival, development, dry weight and of marine ¢sh, shrimp and prawn. Journal of the World chemical composition of larvae of the mud crab Scylla Aquaculture Society 23,251^264. paramamosain.In:Mud Crab Aquaculture and Biology, Suprayudi M., Takeuchi T., Hamasaki K. & Hirokawa J. Proceedings of an International Scienti¢c Forum, Darwin, (2002) E¡ect of Artemia feeding schedule and density on Australia, 21^24 April 1997, ACIAR Proceedings No. 78 the survival and development of larval mud crab Scylla (ed. by C.P.Keenan & A. Blackshaw), pp.159^166. Austra- serrata. Fisheries Science 68,1295^1308. lian Centre for International Agricultural Research, Tango M.S. & Gagnon G.A. (2003) Impact of ozonation on Canberra, Australia. water quality in marine recirculation systems. Aquacul- tural Engineering 29, 125^ 137.

r 2007 The Authors Journal Compilation r 2007 Blackwell Publishing Ltd, Aquaculture Research, 38, 1539^1553 1553