par
G. Brunel
C.T.E. Laurent Bouillet
15, Rue du 18 janvier 1952
Tunis, Tunisie
Résumé
La C.T.E. Laurent Bouillet a procédé durant l'année 1974 à l'aménagement d'une station expérimentale d'aquaculture marine. Ces installations, qui comprennent une écloserie polyspécifique et divers bassins de grossissement sont destinées à l'élevage de poissons (loups, daurades) et de crevettes pénéides (Penaeus kerathurus). Les premières expérimentations (hiver 1974–75) ont été consacrées à la reproduction induite et à l'élevage larvaire du loup (Dicentrarchus labrax) et ont permis une petite production de 12 000 alevins de 90 jours.
Les travaux se poursuivent actuellement avec diverses expériences de production de post-larves de crevettes (Penaeus kerathurus). Les méthodes et résultats obtenus à ce jour, plus particulièrement sur le loup sont revus dans ce document.
Abstract
The firm of C.T.E. Laurent Bouillet commenced, in 1974, the development of an experimental mariculture station in Tunis. The installations, which include a multi-purpose hatchery and a number of rearing tanks, are to be used for the rearing of fish (sea bass and sea bream) and shrimp (Penaeus kerathurus). Initial tests (winter 1974/75) were aimed at the induced breeding and larval rearing of sea bass (Dicentrarchus labrax) and result in the production of 12 000 juveniles. Experiments aimed at the production of post-larvae of shrimp (P. kerathurus) are also being conducted. The paper reviews the methods used and results obtained to date with emphasis on sea bass.
Depuis 1973, la C.T.E. Laurent Bouillet a aménagé une ferme expérimentale d'élevage de crevettes (Penaeus kerathurus) et de poissons (Dicentrarchus labrax et Chrysophrys auratus).
Le but de cette expérimentation est de réaliser le cycle total de l'élevage (reproduction - elevage larvaire - croissance jusqu'à la commercialisation). La phase expérimentale de ce travail est prévue sur deux ans, et sera consacrée à la maîtrise des techniques et à déterminer, pour les divers élevages, un avenir rentable.
Dans ce rapport, nous développerons essentiellement notre expérimentation de reproduction et d'élevage larvaire du loup, et nous nous limiterons à résumer les techniques que nous utiliserons pour Penaeus kerathurus.
Les installations de la C.T.E. Laurent Bouillet à Porto Farina se composent de deux ensembles principaux:
Une écloserie polyspécifique.
Une série de bassins de croissance pour les divers juvéniles produits par l'écloserie.
2.1 L'Ecloserie
2.1.1 Installations
L'écloserie est composé de quatre éléments :
2.1.1.1 Bassins de stockage et de ponte des géniteurs
Au début des opérations (novembre 1974) les géniteurs de loup étaient stockés dès leur capture dans des bassins de béton (9 m × 5 m × 2 m). Ces bassins, alimentés en eau en permanence, n'étaient pas thermorégulés et ne permettaient pas d'études sur la maturation. Depuis, de nouveaux bassins ronds ont été construits (diamètre 3,5 m; = 1,10 m) dans ce but et pourront avoir des cycles indépendants de température et de photopériode.
Pour la ponte, une série de quatre bassins de 1 m3 ont été utilisés. Une paroi vitrée permet une observation parfaite des femelles dans les heures précédant la ponte.
2.1.1.2 Bassins d'élevage larvaire
Différents types de bacs et bassins ont été étudiés au cours de la saison 1974–1975 de reproduction du loup:
Bac cylindrique à fond conique de 350 l.
Bassins ronds à double fond de 8 m3.
Bassins en béton de 90 m3.
Ce sont ces mêmes bassins qui seront utilisés pour l'élevage des larves des pénéides (Penaeus kerathurus). Ces différents bassins peuvent être alimentés en eau de mer filtrée soit grossièrement (sur filet planctonique à vide de maille 200 microns) ou très finement (de l'ordre de 3 microns). Un chauffage par résistance électrique (3000 Watts) immergées permet d'éviter les brusques variations thermiques (chute nocturne, tempête) et de maintenir la température à une valeur fixée.
2.1.1.3 Unité de production de plancton (phytoplancton et zooplancton)
Cette unité a pour but de fournir aux stades larvaires des espéces élevées une nourriture vivante proportionnée à leur taille. Elle est constituée d'une salle de culture d'algues (cultures mixtes flagellés et diatomées) et d'un ensemble d'incubateurs à Artemia salina, de bacs d'élevages de copépodes et de rotifères.
2.1.1.4 Laboratoire d'analyse et de contrôle
Ce laboratoire est équipé pour suivre l'évolution des différents paramètres physicochimiques des milieux d'élevages (température, salinité, O2 dissous, pH, NO3, NO2) et pour surveillez les différentes cultures et élevages (algues, zooplancton, oeufs et larves de poissons).
2.1.2 Cycle d'utilisation
2.1.2.1 Choix des espèces
L'aménagement de l'écloserie a été initialement prévu pour l'élevage larvaire de Penaeus kerathurus, dont la période de ponte s'étale de fin mai à octobre, avec un pic aux mois de juin et de juillet.
Les larves de Penaeidae sont conservées environ trente jours en bassin à l'écloserie et sont alors transférées en bassin de croissance.
Nous avons donc cherché, afin d'utiliser au maximum les installations de l'écloserie, à compléter notre programme de travail avec des espèces dont la période de reproduction se situe entre septembre et mai, et dont l'approvisionnement en alevins à partir du milieu naturel apparaît délicat à grande échelle.
La daurade (Chrysophrys auratus): période de reproduction septembre à novembre.
Le loup (Dicentrarchus labrax): période de reproduction novembre à février.
2.1.2.2 Cycle de l'écloserie (Fig. 1).
2.2 Bassins de croissance
Les bassins de croissance sont actuellement au nombre de trois:
Deux bassins de 1 ha (100 m × 100 m) aménagés pour le grossissement des crevettes Penaeus kerathurus;
Un bassin de deux ha plus spécialement consacré aux élevages de loups, daurades.
2.2.1 Aménagements
2.2.1.1 Bassins à crevettes
Ces bassins ont été construits sur une lagune par une grue équipée en dragline. Les digues limitant les bassins sont constituées par les sédiments prélevés par l'engin, qui se composent de sable, argile et vase. Cette méthode de construction a été choisie par ce qu'elle confère au fond du bassin un profil particulier: un plateau central de 70 × 70 et de 1,20 m de profondeur, bordé d'un fossé profond de 2,50 m et large de 15 m. Cette particularité est très intéressante, pour l'élevage des Penaeidae:
Le plateau central sert en effet de “Litière” surélevée aux crevettes, pendant la journée. La nourriture est distribuée dans la fossé périphérique. Le surplus non consommé de cette nourriture, en se décomposant, crée sur le fond une couche d'eau et de sédiment très pauvre en oxygène dissous, qui serait nocive pour les crevettes si elles étaient obligées de s'enfouir à ce niveau.
Ce bassins sont alimentés en eau de mer grâce à une station de pompage (500 m3/h). L'évacuation se fait par le fond.
2.2.1.2 Bassins à poissons
Ils sont aménagés de la même manière que les bassins à crevettes. Diverses séparations sont placées pour éviter les interactions entre les différentes classes d'âges et pour permettre une pêche fractionnée des animaux en bassin.
De plus, diverses zones ont été draguées à -3,00 m de façon à permettre l'installation de cages flottantes, qui serviront essentiellement de tests comparatifs de croissance.
2.2.2 Cycle d'utilisation
2.2.2.1 Bassins de croissance des crevettes
La saison de reproduction des crevettes (Penaeus kerathurus) débute vers la fin du mois de mai et se prolonge jusqu'à octobre.
Les post-larves sont placées en bassin de grossissement au stade P20 (fin juin pour les premières pontes). Elles ne sont pas lachées directement sur la totalité du bassin, mais sont stockées pendant environ six semaines en parc de pré-grossissement.
Le cycle thermique des eaux du golfe de Tunis permettant la croissance (température supérieure à 20°C jusqu'à la fin novembre, la récolte se situera à ce moment (poids individuel espéré: 15 g environ).
Différents essais seront tentés afin d'utiliser ses bassins pendant l'hiver et le début du printemps. On fait pondre les dernières femelles matures (octobre) et on garde en écloserie les post-larves, jusqu'à P30.
Début décembre, après la récolte de novembre, on peut placer les post-larves dans les grands bassins, à faible densité (15 individus/m2). La croissance hivernale est certes pratiquement nulle, mais ainsi les juvéniles profitent du début du printemps (avril/mai/débutjuin). Cette expérience sera tentée fin 1975.
2.2.2.2 Bassins de croissance des poissons
A proprement parler, on ne peut mettre en évidence le cycle d'utilisation des bassins d'élevage de poissons, plusieurs facteurs importants devant être déterminés:
Temps nécessaire pour atteindre la taille-portion (daurade, loup).
Type d'élevage (cage, parc).
Les installations de Porto Farina ont été effectivement fonctionnelles au courant du mois d'octobre 1974. Le premier sujet de travail a donc été la reproduction du loup (Dicentrarchus labrax).
3.1 Elevage du loup
Le but de l'opération pour la saison 1974–1975, était d'obtenir l'induction de la ponte, de mettre au point une méthode d'élevage larvaire permettant un développement à grande échelle et de produire pour cette première campagne un nombre significatif d'alevins.
3.1.1 Géniteurs
3.1.1.1 Pêche
Pour cette année, nous avons dû commencer l'approvisionnement en géniteurs alors que ceux-ci approchaient de la période de reproduction. Nous n'avons donc pas étudié le problème de la maturation. Les géniteurs ont été capturés au filet trémail sur la plage voisine. Cette méthode n'est pas à recommander, car les poissons souffrent beaucoup (30 à 50 pour cent de mortalité consécutive à la pêche). La pêche à la ligne ou à la bordigue semble beaucoup plus adéquats et n'occasionne qu'un très faible pourcentage de mortalité.
3.1.1.2 Stockage
Quand les animaux arrivent à l'écloserie, ils sont placés dans des cages d'observation situées dans le bassin de stockage. En effet, le fort taux de mortalité impose une surveillance des individus pendant les premiers jours suivant leur capture, afin de conserver uniquement les animaux sains et non blessés. Passée cette période les loups sont lachés dans le bassin.
L'eau du bassin est aérée et renouvelée en permanence pendant la journée (environ trois volumes par jour). Le bassin est couvert et non chauffé. Le cycle de température suit donc à peu prés le cycle de la mer.
La charge du bassin n'a jamais dépassé 200 kg pour les 90 m3.
3.1.1.3 Alimentation
Les animaux sont nourris six fois par semaine avec des déchets de pêche. Les espèces trop grandes sont coupés en morceaux.
Espèces principalement utilisées:
Gobbie, athérines, mulets, sardines.
Seiches et poulpes en morceaux, Palaemon sp.
Les quantités distribuées varient suivant la température (diminution des doses journalières en dessous de 12–13°C).
Il est nécessaire d'habituer le loup à la nourriture “morte”, qu'il ne consomme pas naturellement. Cette adaptation se révèle délicate lorsque l'animal est capturé peu avant le moment de sa reproduction, période où il se nourrit peu.
3.1.1.4 Evolution du rapport gonado-somatique (R.G.S.)
Sur les quelques loups n'ayant pas supporté les stresses répétés de la pêche, du transport et de l'acclimatation à la captivité, nous avons effectué des mesures de R.G.S.
Ces mesures nous ont permis de définir plus précisément le cycle de maturation du loup en Tunisie et de programmer les premières injections hormonales pour l'induction de la ponte. Il est possible de suivre l'évolution de maturation chez les femelles en appréciant la convexité abdominale. Cette méthode, qui requiert une certaine habitude, est cependant moins précise, mais évite la dissection de l'animal.
Nous avons pu aussi sélectionner plusieurs femelles dont l'abdomen présentait un gonflement bien apparent. Les mesures de R.G.S. effectuées parallèlement (moyenne 4,5) nous ont conduit à provoquer la ponte chez ces femelles par injection hormonale.
3.1.1.5 Induction de la ponte
Les femelles qui ont été retenues pour la ponte sont placées dans de petits bassins de 1 m3 à paroi vitrée (ce qui facilite l'observation) à raison de trois femelles par bassin. Deux mâles fluents sont ajoutés. On pratique sur les femelles une injection hormonale intramusculaire (Gonadotrophine chorionique H.C.G.) à raison de 1 000 UI/kg. Ce dosage d'hormone, qui peut sembler un peu fort, a été choisi car notre but était d'obtenir d'abord une ponte et par là même de pouvoir étudier l'élevage larvaire. Par la suite, nous avons essayé plusieurs dosages plus faibles avec des réponses identiques. Il est possible que l'hormone agisse selon une loi du “Tout ou rien”, car on n'observe pas de relation nette entre les temps de réponse à l'injection et les concentrations utilisées.
La température du bac est maintenue à environ 15°C. Trente à cinquante heures après l'injection, on peut observer un net gonflement abdominal qui est un signe très positif d'évolution vers la ponte. Celle-ci peut être obtenue environ 24 heures après cette observation.
Il a été montré que la période de fécondabilité des ovules est très courte (5 à 6 h). Il importe donc de bien déterminer cette période en suivant l'évolution des ovules (prélèvement par catheter) dont la taille s'accroît beaucoup (de 600 à 1 100 microns) au cours de la vitellogénèse.
Nous avons pu ainsi provoquer des émissions d'oeufs viables par pression abdominale.
3.1.1.6 Résultats (Tableau I)
3.1.2 Elevage larvaire
3.1.2.1 Fécondation
Les oeufs recueillis sont placés dans un cristallisoir et mis en contact étroit avec du sperme (prélevé sur un mâle fluent) par une agitation modérée.
TABLEAU I. INDUCTION DE LA PONTE
| GENITEURS | INJECTION | OVULATION | Observations | ||||||
| No. | Poids (kg) | Date Heure | Prod. | Dose UI | Date Heure | Temps Latence (h) | Nb. Oeufs | % Fec. | |
| 1 | 0,800 | 2.12.74 12 h 30 | HCG | 1 000 | - | - | - | - | Pas d'évolution |
| 2 | 1,500 | 4.12.74 17 h | HCG | 1 500 | 6.12.74 15 h | 46 | 100 000 | 50 | - |
| 3 | 1,000 | 4.12.74 17 h | HCG | 1 000 | 6.12.74 23 h | 54 | 80 000 | 0 | Oeufs opalescents |
| 4 | 1,000 | 7.12.74 10 h | HCG | 1 100 | 9.12.74 23 h 30 | 57.30 | 60 000 | 60 | - |
| 5 | 2,800 | 7.12.74 10 h | HCG | 3 000 | 9.12.74 23 h 30 | 61.30 | 230 000 | 0 | Oeufs non transparents et non circulaires |
| 6 | 1,600 | 19.12.74 11 h | HCG | 1 500 | 22.12.74 6 h | 67 | 140 000 | - | - |
| 7 | 1,2000 | 4.1.75 9 h 30 | HCG | 1 250 | 7.1.75 8 h 30 | 71 | 100 000 | 70 | - |
| 8 | 0,800 | 4.1.75 9 h 30 | HCG | 1 000 | 7.1.75 14 h 30 | 77 | 120 000 | 60 | - |
| 9 | 1,000 | 4.1.75 9 h 30 | HCG | 1 000 | 7.1.75 7 h | 69.30 | 200 000 | 45 | - |
| 10 | 3,000 | 23.1.75 10 h 30 | HCG | 3 000 | 25.1.75 14 h 30 | 52 | 350 000 | 20 | |
Environ 15 minutes après, de petites quantités d'eau de mer sont ajoutées. Dès que les premiers critères de fécondation sont observés (rétraction vitelline, première division cellulaire) les oeufs sont mis dans des incubateurs à fond conique de 350 1. Plusieurs essais concernant le milieu d'incubation ont été tentés.
Milieu 1: eau de mer filtrée à 5 microns et brassée par aérateur.
Milieu 2: culture d'algues (chloroflagellés g. Tetracelmis sp.), agitée par aération.
Milieu 3: culture d'algue sans aération.
La salinité varie de 36,5 ‰ à 37,5 ‰. La température de 15°C à 16°C. Dans ces conditions, les oeufs flottent à la surface. Les oeufs morts ou non fécondés sédimentent peu à peu.
Des différents milieux d'incubation expérimentés, les milieux enrichis en algues donnent de meilleurs résultats, avec peu de différence entremilieux aérés et milieux non aérés.
L'éclosion survient au bout de 45 h à 55 h aux températures de 15–16°C, avec des pourcentages très forts pour le milieu 3 (algues sans agitation jusqu'à 95 pour cent).
3.1.2.2 Conditions de milieu d'élevage larvaire
Comme pour l'incubation, nous avons expérimenté les 3 milieux précités. Rapidement, nous avons pu constater que l'aération même faible des milieux 1 et 2 avait des conséquences néfastes sur les larves qui étaient précipitées plus ou moins violemment contre les parois des bacs, ou bien qui étaient rassemblées en agrégats par différents mouvements de convexion du liquide.
Chaque tentative d'aération ou de renouvellement de l'eau du bac provoquait ces phénomènes. Nous avons donc adopté pour la suite de nos élevages le milieu 3/culture d'algues (chloroflagellés et diatomées) en eau stagnante.
Les conditions de milieu furent les suivantes:
Température: de 15°C à 18°C les 30 premiers jours puis ensuite le cycle naturel des températures.
Salinité: de 36,8 ‰ à 37,6‰.
Aération: nulle pendant 15 jours.
O2 dissous: 9 à 11 mg/l.
Eclairement: naturel.
Algues: diatomées (Nitzschia closterium - Chaetoceros sp. - Skeletonemia costatum) flagellés (Tetracelmis sp.).
3.1.2.3 Alimentation
Juste après l'éclosion, la larve ne se nourrit pas. Sa bouche n'est pas ouverte et elle vit sur ses réserves vitellines. Vers le quatrième jour à 15°C, la bouche s'ouvre et la larve peut commencer à s'alimenter. Elle doit le faire avant que ne s'épuisent ses réserves et doit donc dès ce moment disposer d'une nourriture qui soit adaptée à la taille de sa bouche et qui soit attractive (couleur-mouvement).
Nous avons placé dès le troisième jour une grande quantité d'oeufs et de trocophores de moules (taille 60–80 microns; concentration 5 à 10/ml). Chaque jour, jusqu'au dixième jour, de petites quantités de trocophores et de véligères de moules sont ajoutées afin de conserver la concentration initiale.
La larve se développant, elle peut ingérer des proies plus grosses du huitième jour, un apport de zooplancton filtré calibré à 150 microns de nauplii de copépodes Tisbe sp. complète les trocophores et véligères de moules. Le zooplancton filtré et les nauplii de copépodes sont fournis seuls à partir du dixième jour. La concentration de ces proies est plus faible (1 à 2 ind/ml).
Le quinzième jour, de petites quantités de nauplii de balanes, de nauplii d'Artemia salina sont ajoutées et vont constituer 50 pour cent du régime jusqu'au trentième jour. Le zooplancton filtré, calibré à 250 microns, les stades copépodites de Tisbe complètent la séquence alimentaire.
A partir du trentième jour, les nauplii d'Artemia salina sont supprimés. Le zooplancton filtré sur 300 microns continue à être distribué, en même temps que de petites quantités de crabe vert (Carcinus moenas) finement broyés.
3.1.2.4 Croissance
Cf. courbe Fig. 2.
Sur les différentes séries d'élevage que nous avons eues, nous avons pu observer diverses variations de croissance des larves et alevins. Nous avons représenté une courbe moyenne réalisée en bac de 350 l, avec cultures d'algues.
Les principales causes de ces variations peuvent être de différent ordre:
Conditions de milieu:
Essentiellement la température:
Une température de 20°C–22°C permet une croissance nettement plus rapide, mais provoque une mortalité plus importante. La température que nous avons choisie semble être un compromis satisfaisant (croissance-survie).
L'équilibre chimique du milieu d'élevage larvaire a une grosse importance; une concentration trop forte de NH4, NO2, H2S, une variation de pH ou de salinité peuvent avoir des conséquences irréparables, car il est très délicat de procéder à d'importants renouvellements de milieu sans léser les larves qui sont très fragiles.
Conditions biologiques:
Nourriture: quantité et qualité
La séquence alimentaire pendant l'élevage larvaire à une grande importance, car elle conditionne tout d'abord la survie et ensuite la croissance: des lots de larves n'ayant pas eu une concentration suffisante (1 ind/ml) en proie dans leur bac d'élevage ont présenté des retards de croissance en taille de 2 mm au trentième jour, sur d'autres lots bien nourris.
Pour la prochaine saison de reproduction du loup, mous avons prévu un programme d'étude de la croissance larvaire en fonction de divers paramètres (température, photopériodes, séquences alimentaires, concentration).
3.1.2.5 Mortalité
Au cours du développement larvaire (de l'éclosion à 30–35 jours), différentes périodes de mortalité brutale sont observées, quelques soient les conditions de l'élevage.
De l'éclosion au dixième jour environ, il y a une mortalité régulière, vraisemblablement due à des malformations génétiques.
Du dixième jour au quinzième jour en peut noter une mortalité très forte de 10 à 75 pour cent du stock, qui semble correspondre au passage des réserves vitellines à la nutrition active: les larves mortes ont le tube digestif vide, non coloré. Un certain nombre de larves ne paraît pas pouvoir se nourrir. Les causes peuvent être de différent ordre:
Malformation de la bouche ou du tube digestif.
Epuisement des réserves vitellines trop rapide (avant que la larve ne puisse être stimulée activement par la nourriture proposée).
Un deuxième passage très critique se situe du vingt-cinquième au trentième jour. La mortalité à ce stade est très régulière. Les larves ont un comportement très particulier; elles tournoient en remontant à la surface et leur oeil réfléchit la lumière avec éclat. Il est possible que ce comportement, qui conduit à la mort de la larve, soit dû à une malformation d'organe d'équilibration (vessie natatoire).
Vers le trentième jour, ces phénomènes de mortalité massive disparaissent et le stock d'alevins évolue alors sans grande érosion.
Cette saison de travail sur la reproduction et l'élevage larvaire du loup nous a donc permis de mettre au point une technique de production d'alevins, applicable dès maintenant à une grande échelle: 12 000 alevins de 90 jours ont pu être produits cette année.
Les recherches complémentaires que nous aborderons l'an prochain concerneront:
La maturation: augmentation de la période de reproduction.
La ponte: passage complet à la ponte naturelle.
La croissance des larves et alevins (optimisation des conditions d'élevage).
Ces différentes expérimentations seront menées parallèlement à une production de masse d'alevins.
3.2 Elevage des crevettes (Penaeus kerathurus)
Les techniques d'élevage que nous utiliserons sont essentiellement dérivées des techniques mises au point au Japon par Hudinaga et Shigeno.
Nous nous contenterons de résumer notre programme de travail pour 1975, l'année 1974 ne nous ayant permis d'effectuer que quelques expériences limitées.
3.2.1 Reproduction et élevage larvaire
3.2.1.1 Capture des géniteurs
Les femelles matures sont pêchées dans le golfe de Tunis au filet trémail. Elles sont immédiatement ramenées au laboratoire dans des bacs aérés, à densité faible (30 pour 60 l environ). Dès leur arrivée, elles sont tirées et placées dans les bassins de ponte.
3.2.1.2 Ponte
Pour provoquer la ponte, l'eau des bassins est progressivement chauffée (résistances électriques immergées) jusqu'à la température de 26°C–27°C. Cette température est maintenue pendant toute la nuit. Le lendemain matin divers prélèvement sont effectués pour étudier la présence d'oeufs et en déterminer approximativement la concentration. Si celui-ci apparaît suffisant, les femelles sont sorties du bassin. Dans le cas contraire, leur stabulation est poursuivie 24 h.
3.2.1.3 Eclosion et elevage larvaire
Nous ne développerons pas ce paragraphe qui a déjà fait l'objet de nombreuses communications.
3.2.1.4 Alimentation
Le schéma d'alimentation peut se résumer ainsi:
Stades Zoë: Diatomées (Nitzschia closterium, Skeletonemia costatum).
Stades Mysis: Trocophores de moules ou d'hûitres.
Stades Post-larvaires: Nauplii d'Artemia salina (jusqu'à P5).
P5 à P20: Chair de mollusque (Donax) finement broyée.
Au delà de P20, stockage en bassin de croissance.
3.2.2 Croissance
3.2.2.1 Prégrossissement
Les P20 sont stockées pendant 5 à 7 semaines dans un parc de prégrossissement (sable fin-aération forte-renouvellement d'eau important) ou elles sont nourries 3 fois par jour avec du crabe (Carcinus moenas) et des mollusques (Donax sp.) broyés.
3.2.2.2 Grossissement
Dès que les juvéniles ont atteint une taille de 2 à 3 cm, elles sont lachées dans l'ensemble du bassin de croissance. Elles sont nourries une fois par jour (crabe et poissons broyés), le soir vers 17 h. Tous les matins, les restes de nourriture sont examinés, afin d'éviter une distribution d'aliment trop importante, qui entraînerait rapidement une pollution du fond.
3.2.2.3 Récolte
La pêche des crevettes en bassin commence au filet fixe. Dès que les prises diminuent, le bassin est vidangé et la récolte s'effectue à la main en fouillant le sable des litières.
Fig. 1. Cycle de l'ecloserie.
Fig. 2. Courbe de croissance d'alevins de loup du 19 décembre 1974 au 30 mars 1975.
by
M.K. Kelleher and M. Vincke
Projet PNUD/FAO Vulgarisation de la Pisciculture
Centre Piscicole de la Landjia
Bangui, Central African Republic
Abstract
Clarias lazera can be induced to breed throughout the year by means of a special spawning technique. Fry survival presents a major problem. Results of survival experiments on five day old fry stocked in ponds for a period of one month are presented. Tentative conclusions are drawn. Work in progress and intended future research are mentioned.
Résumé
La reproduction de Clarias lazera peut être obtenue toute l'année par induction. La survie des alevins est un problème majeur. Les résultats d'essais de survie d'alevins de cinq jours, élevés en étang durant un mois, sont présentés dans cette note, ainsi que des conclusions provisoires. On mentionne également les travaux en cours et les recherches futures.
The feasibility of raising Clarias lazera in ponds has been investigated over a number of years (Micha, 1973); Hastings, 1973; Vincke, 1974). It is an ideal fish for culture in many ways; as a hardy omnivorous it accepts artificial feed and is resistant to handling.
This species ranges widely throughout most of the fresh waters of Africa, from Lake Galilee in the Middle East throughout the Nile, Congo and West African river system (Boulenger, 1912; Pellegrin, 1923; Daget, 1959; Blache, 1965) Clarias lazera ranks among the largest of the 27 species of the genus Clarias on the continent with one specimen caught in Lake Edward weighing 12.8 kg and measuring 131 cm in total length (C.T.F.T., 1972).
The high growth rate and predatory capacities of C. lazera have given excellent results both in monoculture and in polyculture with Tilapia nilotica. At the Centre Piscicole National de la Landjia, maximum growth rate, during the first 35 to 120 days of life, has been recorded at 2.9 g/day. In experimental scale tests conducted by Hastings (1973), maximum production exceeded 16 tons/ha/year in monoculture at a density of 2 fish/m2 with pelleted artificial feeds. Miller and Maletoungou (in preparation) report productions ranging from 7–10 tons/ha/year in polyculture with Tilapia nilotica in association with pigs and on a production scale.
The induced spawning technique now used at the Centre Piscicole National was first elucidated by Micha (1973). This involved intraperitoneal injections of the synthetic hormone Deoxycorticosterone Acetate (DOCA). Chorionic gonadotrophin has been tried without success at the Centre (Micha, 1973; Miller, personal communication).
Attempts have been made to correlate an index of fecundity of female Clarias lazera (numbers of fry produced per month by induced reproduction) with climatological data (Delince, in preparation). However, climatological data for 1974 on which part of the study is based were atypical.
Natural reproduction of C. lazera has been achieved with varying success at the Fisheries stations at Landjia, Bambari and Bouar (Central African Republic).
Although large numbers of five day old fry have been produced, mortalities beyond this stage have raised a serious problem (Micha, 1973).
After one year of intensified spawning efforts at the Centre, with an average of 24 attempts at induced spawning per month, it is apparent that Clarias lazera can be reproduced artificially throughout the year. Thus the main problem remaining concerns the survival of the fry.
Considerable variations has occurred in the numbers of fry available per month. This and other difficult local conditions have resulted in somewhat heterogeneous data on the survival of Clarias lazera fry in ponds at the Centre National Piscicole.
This paper attempts to summarize and interpret, in so far as possible, this heterogeneous information.
2.1 Stocks of Clarias lazera
The brood stock used at the Centre were originally procured from the Oubangui River; second or third generation fish are now being used. Morphological data on these fish are given by Micha (1973) and Vincke (1974).
2.2 Injection Technique
Ripe females are selected on the basis of release of eggs by manual pressure. Extruded eggs are not all mature. Ripe females vary in size from 28 to 65 cm and in weight from 175 to 1,600 g. These females are weighed, measured, injected and placed in the spawning tanks as early as possible in the morning. Five mg of DOCA (Laboratoires Biergon, Liege, Belgium) per 100 g of fish are injected intraperitoneally on the side, some distance behind the pectoral fin.
2.3 Spawning Tanks
Concrete spawning tanks used in this research measure 70 cm × 80 cm with depth of 47 cm. Stones without sharp edges are used as a spawning substrate. These are placed in half of the bottom of the tank one layer deep. Stones are approximately 2–6 cm in size. The tanks are filled to a depth of about 20 cm before the female is introduced.
2.4 Mating
For the purposes of this article placing of the male with the female is referred to as mating. Males are selected to be smaller or roughly the same size as females. It has been found that when large males are placed with small females the death of the latter often results. Males are also chosen on the basis of aggressivity or non-aggressivity and controls which have not been put in combat are also used. Studies on the selection of males are still in progress and details of the methods used are given by Delince (in preparation).
The uninjected male is introduced approximately 10 hours after the injected female. The tank is rapidly filled and water is allowed to flow through the tank overnight. Injections are usually completed by 08:00 hours and the mating performed between 18:00 and 19:00 hours. There is always only one male and one female by spawning tank.
2.5 Egg Laying and Hatching
Spawning usually occurs before morning following which both the male and female are removed from the tank, weighed, measured and any injuries noted.
A number of stones are retrieved from the spawning tank and the total number of eggs counted. These are left in enamel dishes beside the tank until the following day, at which time the fry are counted to determine the hatching rate.
On the fourth day after hatching the spawning tanks are drained and all the fry counted using small dip nets constructed from wire and mosquito netting. The counted fry are then stocked in ponds. Total length of these five day old fry is about 10–12 mm.
2.6 Rearing Ponds
Rearing ponds used at Landjia range in size from 100 to 400 m2 and vary widely as to type of soil and vegetation.
While many attempts were made to maintain an overall experimental plan throughout these survival trials, for numerous reasons these plans could not be implemented. Availability of adequate ponds was extremely limiting, while similar stocking densities could not be maintained because of lack of sufficient fry from the same female. Lack of fertilizers also contributed to the variability of these preliminary results.
2.7 Stocking and Harvesting
All rearing ponds were stocked with fry five to six days old. All the trials were terminated approximately one month after stocking. Ponds were treated according to one of the following procedures:
These various approaches are discussed in more detail in the Results section.
It must be stressed that though every effort was made to perform repetitions of identical experiments, conditions prevailing did not lend themselves to this end. Firstly, the experiments themselves are diverse in nature and secondly, the results, in spite of numerous repetitions in some cases, showed great variations.
3.1 Experiments with Similar Fertilization
The largest block of similar survival experiments were done with a fertilizer of constant composition. The ingredients of this fertilizer were as follows:
| Peanut oil cake | 5 kg/100 m2 |
| Bone meal | 5 kg/100 m2 |
| Chicken manure | 2 kg/100 m2 |
| Dried blood | 2 kg/100 m2 |
| Inorganic fertilizers containing ammonium and phosphorus | 1.045 kg/100 m2 |
| Total | 15.045 kg/100 m2 |
The amount of fertilizer applied was virtually identical in each of these 32 experiments. The densities at which the fry were stocked varied as indicated in Table 1.
From these 32 trials, details of the four experiments which gave the best survivals are given in Table 2 along with a similar series performed the following month in an effort to duplicate the results obtained. It can be seen from this Table that the results differed widely. The mean survival of the density group 0.7–5.6 fry/m2 given in Table 1 would appear to be high due to the four exceptional results listed in Table 2.
3.2 Survival under Varying Conditions
Table 3 presents the results of numerous pilot trials. Both experiments and results differed to such an extent that it is meaningless to include average stocking densities and mean survivals for some of these summarized results.
3.3 Density of Stocking as a Factor in Survival of Clarias lazera fry
It appears from the results listed in Table 4 that stocking density plays an important role in survival of Clarias lazera fry. A closer examination of this factor is made in Section 4.
3.4 Numbers of Fry from November 1973 to July 1975
The total number of fry obtained per month by the induced spawning technique and the total number of approximately 35 day old fry harvested per month is given in Table 5.
The number of injections done per month varied throughout the above period. These results are discussed below.
It is evident that Clarias lazera can be reproduced by an induced spawning technique throughout the year. The problem of fry survival remains unsolved. The preliminary and heterogeneous nature of the results presented here must be emphasized. However, the number of experiments carried out (84) may perhaps compensate somewhat for the spatial, temporal and experimental diversity involved.
4.1 Density Dependence
Evidence of density dependence in the survival of fry of Clarias lazera is given in Table 4. A survival of 20.9% has been recorded at a stocking density of over 40 fry/m2.
The effect of handling and time required to count large spawns appear to be of significant importance. With large spawns, counting of fry at Landjia may continue until mid-day. It is with these large spawns that high stocking densities have been effected. Because of the human error involved and higher environmental temperatures prevailing when many thousands of fry are counted, this apparent density dependence may well be related to handling mortalities. Thus while low stocking densities are advisable on methodological grounds, to conclude that they are necessary to achieve high fry survival would be imprudent.
Work is in progress to resolve this fry handling problem and other questions relating to technique.
4.2 Fertilization
Canal water at Landjia has a CaCO3 hardness of less than 15 ppm. In some experiments (see section 3.1) bone meal, which contains a high percentage of calcium, was used. From the relative success of these studies (Table 1) it appears that availability of calcium may be a limiting factor in the survival of Clarias lazera fry at Landjia.
Four trials with high quantities (25 to 63 kg/100 m2) of organic fertilizer suggest from the poor survivals obtained (maximum 3.6%), that dissolved oxygen may be another limiting factor. However, water quality was not monitored in these studies.
Pham (1975) has obtained good survivals of Clarias lazera fry up to the 15 day old stage using concrete tanks with aerated water. Due to the limited number of trials conducted in his work, further research is necessary to evaluate this important parameter.
4.3 Feeding
Micha (1973) indicates, that up to the age of 15 days, the stomach contents of Clarias lazera are entirely of zooplankton; the diet changes to larvae of aquatic insects thereafter. Brewery waste does not appear in the stomach contents until the fry are 40 days old (ibid).
This primary growth period (upto 15 days of age) seems critical. At the C.P.N. an artificial foodstuff of adequate particle size has not been found. Work is in progress using locally available foodstuffs such as dried brewer yeast and peanut oil cake. These studies involve the use of ponds, concrete tanks, aquaria and “happas”.
In pond experiments with feeding, the highest survival rates recorded were 5.5% and 40% at the respective stocking densities of 1.6 fry/m2 and 50 fry/m2. Feeding used in these trials included brewery waste and a powdered mixture of wheat, maize, soya and powdered milk in the former, and peanut oil cake with the above powdered mixture in the latter.
No conclusions can be drawn from the studies which involved both feeding and fertilization without further clarification of the separate effects of each.
4.4 Number of Fry
Attempts at induced spawning have been carried out at an average of 24 per month during the lqst year, yet the number of 5 day old fry produced varied widely per month (Table 5). This problem will be the subject of further studies.
An attempt to correlate survival in ponds of 5 day old fry with the time of year has given inconclusive results.
4.5 Predation
High predation by frogs on the fry of Clarias lazera in ponds has been suggested by Micha (1973). However, five studies performed at the C.P.N. indicate that amphibian predation accounts only for 10% mortalities (Nugent, in preparation).
Though amphibians, aquatic insects and, occasionally, wild fish certainly prey on Clarias lazera fry, lack of adequate nutrition is probably the single greatest cause of mortality.
As a result of varying growth rates of individual fry during the 5 to 35 day period, some cannibalism may occur but no quantitative observations concerning this aspect have been made.
4.6 Survival of Clarias lazera under Production Conditions
Survival beyond the 35 day old stage presents no serious problem. At the C.P.N. survival rates in the 60 to 75% range have been recorded regularly with varying artificial feeds in both monoculture and polyculture with tilapia.
Present research at Landjia is aimed at defining optimal conditions for growth and survival within the 35 to 65 day old period.
Research carried out in the Central African Republic (C.A.R.), Ivory Coast and Cameroun indicates that survival of Clarias lazera fry present a major problem. Handling, feeding and predation appear to be the main causes of mortality. Further research is necessary before Clarias lazera can be raised on a production scale. Such research is at present being conducted in the Central African Republic.
The authors wish to express their thanks to the staff of the Centre National Piscicole de la Landjia, in particular Mr R.R. Koyemtan who supervised the counting and stocking of fry.
Blache, J., 1964 Les poissons du bassin du Tchad et du bassin adjacent du Mayo Kebbi. Etude systématique et biologique. Mem.O.R.S.T.O.M., Paris
Boulenger, G.A., 1909–1916 Catalogue of the fresh water fishes of Africa in the British Museum. Vol.II. British Museum (Natural History), London
C.T.F.T., 1966 (Centre Technique Forestier Tropical), Premières directives pour l'introduction de Clarias lazera en pisciculture. Annexe 8 au rapport final du C.T.F.T., FAO FI:SF/RAF/66/054
Daget, J., 1959 Les poissons du Niger supérieur. Bull.Inst.Fr.Afr.Noire (A):21 (2), 664–88
Hastings, W.H., 1973 Expérience relative à la préparation d'aliments des poissons et à leur alimentation. Rapport préparé pour le Projet Régional de recherche et de formation piscicoles FAO FI:DP/RAF/66/054/1
Micha, J.C., 1973 Etudes des populations piscicoles de l'Ubangui et tentatives de sélection et d'adaptation de quelques espèces à l'étang de pisciculture. C.T.F.T., Nogent-sur-Marne, 110 p.
Pellegrin, J., 1923 Poissons d'eaux douces de l'Afrique occidentale. Larose, Paris, 373 p.
Pham, A., 1975 Données sur la production en masse d'alevins de Clarias lazera Val. à la Station de Bouaké (Côte d'Ivoire). C.T.F.T., Nogent-sur-Marne
Vincke, M., 1974 Bilan des premiers essais de reproduction et d'alevinage de Clarias lazera Val. Mimeo. Rapport. Centre National Piscicole de la Landjia, Bangui, R.C.A.
Table 1
Survival of Clarias lazera fry in ponds
with similar fertilization
| Number of trials | 15 | 10 | 7 |
| Range of stocking densities (fry/m2) | 0.7–5.6 | 10–10.84 | 15–50 |
| Average stocking density (fry/m2) | 2.9 | 10.0 | 25.7 |
| Range in survival rate (percent) | 0.29–67.5 | 0.18–29.3 | 0.07–7.8 |
| Mean survival rate (percent) | 20.63 | 10.09 | 3.96 |
Table 2
Individual Results of Survival of Clarias lazera fry in ponds with similar fertilization
| Pond Number | F1 | F2 | F3 | F4 | F1 | F2 | F3 | F4 |
| Pond area (m2) | 200 | 200 | 200 | 200 | 200 | 200 | 200 | 200 |
| Fertilization (kg/100 m2) | 15.045 | 15.045 | 15.045 | 15.045 | 13.4 | 13.4 | 13.4 | 13.4 |
| Date of stocking | 18.2.74 | 18.2.74 | 18.2.74 | 18.2.74 | 21.3.4 | 21.3.74 | 21.3.74 | 21.3.74 |
| Age of fry stocked (days) | 4–5 | 4–5 | 4–5 | 4–5 | 5 | 5 | 5 | 5 |
| Number of fry stocked | 459 | 459 | 459 | 459 | 540 | 540 | 540 | 540 |
| Stocking density (fry/m2) | 2.28 | 2.28 | 2.28 | 2.28 | 2.7 | 2.7 | 2.7 | 2.7 |
| Age of fry harvested (days) | 35 | 35 | 35 | 35 | 39 | 39 | 39 | 39 |
| Number of fry harvested | 227 | 310 | 256 | 280 | 33 | 42 | 34 | 21 |
| Survival rate (percent) | 49.4 | 67.5 | 55.7 | 61.0 | 6.1 | 7.7 | 6.2 | 3.9 |
Table 3
Summary of results of survival of Clarias lazera fry in ponds under various conditions
| Treatment of ponds | Number of trials | Stocking density (fry/m2) | Average stocking density (fry/m2) | Survival range (percent) | Mean survival (percent) |
| Without treatment | 5 | 5–20 | 16.4 | 0.6–31.5 | 9.88 |
| With feeding | 7 | 1.6–50 | 22.6 | 0.27–5.5 | 2.99 |
| In association with pigs | 4 | 6.7–20 | 16.6 | 0.41–20.1 | 8.41 |
| Diverse fertilization | 13 | 5–32.25 | - | 0.1–13.8 | - |
| Fertilization and feeding | 9 | 6.03–50 | - | 0.06–20.9 | - |
Table 4
Survival of Clarias lazera in ponds
stocked at different densities
| Density (fry/m2) | 0.5–5 | 5–10 | 10–20 | 20 + |
| Number of trials | 17 | 27 | 14 | 13 |
| Range in survival rate (percent) | 0–67.5 | 0.05–29.3 | 0.91–20.1 | 0.01–20.9 |
| Survival: | ||||
mean | 20.3 | 6.92 | 5.41 | 4.23 |
mode | 6–8 | 0–1 | 0–1 | 0–1 |
Table 5
Survival of fry of Clarias lazera from November 1973 to July 1975
| Year and month of stocking | Number of 5 day-old fry stocked | Number of 35 day-old fry harvested | Survival (percent) | |
| 1973 - | Nov. | 12,000 | 1,366 | 11.4 |
| 1974: | ||||
| Feb. | 1,836 | 1,073 | 58.4 | |
| March | 30,573 | 2,981 | 9.76 | |
| April | 7,449 | 297 | 3.99 | |
| May | 2,033 | 82 | 4.03 | |
| June | 34,883 | 2,747 | 7.87 | |
| July | 6,321 | 542 | 8.6 | |
| August | 51,280 | 6,752 | 13.17 | |
| Sept. | 19,700 | 179 | 0.98 | |
| Nov. | 13,378 | 267 | 2.00 | |
| Oct. | 1,782 | 66 | 3.70 | |
| Dec. | 450 | 65 | 14.44 | |
| Total (Nov.73 to Dec.74) | 181,685 | 16,410 | 8.7 | |
| 1975: | ||||
| Jan. | 1,025 | 59 | 5.76 | |
| Feb. | 29,056 | 935 | 3.22 | |
| March | 65,227 | 1,700 | 2.61 | |
| April | 55,136 | 1,146 | 2.08 | |
| May | 20,177 | 305 | 1.51 | |
| June | 8,511 | 219 | 2.57 | |
| July | 19,907 | 289 | 1.45 | |
| Total for 1975 (up to July) | 199,039 | 4,653 | 2.5 | |
| TOTAL | 380,724 | 21,063 | 5.5 | |
| (Nov. 1973 to July 1975) | ||||
by
J.D. Ardill1 and R.K. Thompson2
1 Divisional Scientific Officer, Ministry of Fisheries, Port Louis, Mauritius
2 Volunteer, U.S. Peace Corps - Smithsonian Environmental Programme,
Ministry of Fisheries, Camaron Hatchery, Trou d'Eau Douce, Mauritius
Abstract
The phasing of the project leading to the introduction and mass culture of M. rosenbergii in Mauritius is described. The project is now in the pre-investment phase. Hatchery facilities and procedures are described, as well as the results of larval rearing, and difficulties encountered in hatchery operation. The selection of sites for the culture of juvenile prawns to market size, the construction of ponds, their stocking and management are described. Finally, a report is given on the response from the private sector and future plans of the Ministry of Fisheries.
Résumé
Les phases du projet ayant pour objet l'introduction et la culture industrielle de Macrobrachium rosenbergii à l'Ile Maurice sont décrites. Le projet est maintenant dans la phase du préinvestissement. L'écloserie et son mode d'opération sont décrites, ainsi que les résultats obtenus en production de crevettes juvéniles. Les difficultés encourues dans son opération, les critères de sélection de sites pour la construction de bassins d'élevage d'adultes destinés à la commercialisation, la construction de ces bassins et les méthodes de culture sont décrites. Pour terminer, un rapport est fait quand à la suite donnée à ce projet par le secteur privé, et par le Ministère des Pêches.
The giant freshwater prawn, Macrobrachium rosenbergii, was first introduced into Mauritius in early 1972 with the objective of developing a prawn culture industry. This species has several advantages over the locally occurring freshwater prawn M. lar, in that larval rearing techniques have been developed and rapid growth occurs under pond conditions when prawns are given supplemental feed.
A project leading to commercial production of prawns, both for the local market and for export, was envisaged and was divided into three phases as discussed below. The successful completion of each phase was to be the justification for the beginning of the next phase.
Phase I - Introduction of M. rosenbergii into Mauritius:
Importation of parent female stock carrying eggs, together with males for fertilization of these females at subsequent egg layings.
Rearing of larvae hatched from these eggs to the juvenile stage at the pilot hatchery at Trou d'Eau Douce.
Stocking of these juveniles in a wide variety of ponds in order to assure the availability of a breeding stock for Phase II.
Phase II - Experimental pond management:
Expansion of Trou d'Eau Douce hatchery for increased production of juveniles and year-round operation.
Studies on nutrition and feeding methods.
Maximum density stocking of half-to one-acre ponds under simulated commercial conditions for yield assessment.
Phase III - Commercial operation:
Construction of a permanent hatchery and of growing ponds.
Production of juveniles in the hatchery; stocking and management of ponds.
Marketing of adult prawns.
Phase I of the project was successfully completed by the end of 1972 (Ardill et al., 1973). Phase II is now in progress and this paper represents a report on the present status of the prawn-rearing programme in Mauritius. A detailed description of the population dynamics and production of prawns in Government ponds No. 1 and No. 2 at Pamplemousses will be published at a later date.
2.1 General
The natural history of M. rosenbergii has been well documented by Ling (1967a, 1967b). The adults live and breed primarily in fresh water whereas larval development requires brackish water for its successful completion. This type of reproductive pattern makes any aquaculture venture biphasic, i.e.:
a hatchery with brackishwater facilities is needed for rearing larvae from the time of hatching from the egg to post larvae or juveniles; and
freshwater outdoor ponds are needed for rearing juveniles to marketable size adult prawns.
This type of development allows complete control of the culture cycle, avoiding the type of problem encountered in tilapia ponds where excess reproduction results in a pond filled with small fish. As no reproduction occurs in the growing ponds, the size of the prawn population can be maintained at the optimum level by controlling stocking rate and the growth rate and survival can be assessed in conjunction with differing management practices and environmental conditions. The rearing cycle of M. rosenbergii is shown in Fig. 1.
2.2 Adults
After juvenile prawns reach 5 cm in length, sexual dimorphism, as indicated by the presence of the male genital pore at the base of the fifth walking leg, becomes obvious. The majority of juvenile prawns placed in freshwater rearing ponds reach sexual maturity after 7–9 months but a few females have been observed to carry eggs on their pleopods as early as 3–4 months of age. Adults mate and normal egg production and embryonic gestation takes place in fresh water. However, upon hatching, the larvae must reach brackish water in 4–5 days (Ling, 1967a) or they will die.
Adult prawns are benthic omnivores and are considered to have reached marketable size at 12 cm in body length, as measured from base of eyestalk to tip of telson, and about 50 g in weight. In a pond kept for reproductive purposes, one male individual was caught which weighed 525 g after three years growth (Baissac, personal communication).
2.3 Larvae and juveniles
Adult females with eggs are seined from the freshwater rearing ponds, brought to the hatchery and placed in brackish water. Upon hatching, the planktonic larvae (2.0–2.2 mm length) are transferred to larval rearing tanks. The larvae moult 11 times in 35–45 days depending on environmental conditions, before metamorphosing into benthic post larvae which look like miniature adults 0.8–1.2 cm long. These developmental stages have been described in detail by Ling (1967), and by Uno and Kwon (1969). Additional morphological changes take place in the next few moults. The young juvenile prawns are put into fresh water and then transferred into outdoor rearing ponds.
A prototype hatchery at Trou d'Eau Douce was operated from January 1972 until July 1972 and a full description of it has been published (Ardill et al., 1973). This was completely dismantled and construction of a permanent building of concrete block and cement commenced in November 1972 and was completed by March 1973.
Figure 2 shows the present layout of this hatchery. Fresh water is pumped from an adjacent spring by a 200 l/min plastic centrifugal pump. Sea water is pumped from a brackish-water inlet, through a sand filter (depth 75 cm) which removes the larger planktonic organisms into holding tanks 13, 14 and 15. There it is aged for at least three days in order to allow the elimination of toxic contaminants from the decomposition of organic matter in the inlet. From these tanks the water is pumped into the phytoplankton tanks where it is diluted to a salinity of 15–18 ppt and seeded with plankton-rich water. The water is fertilized by the metabolic wastes of male tilapia reared in these tanks. While the value of phytoplankton as food for the larvae is established, it is believed to be a vital water-conditioning factor for successful larval rearing.
During 1973 and early 1974 tanks 9–12 were used to produce phytoplankton, the light necessary for plant growth passing through a roof of transparent fibreglass sheets. Throughout this period, however, growth of phytoplankton was poor and this affected hatchery juvenile production adversely. After many water fertilization trials, including the addition of trace elements, it was decided to try phytoplankton production out of doors in full sunlight, and tank 16 was built. Results were so encouraging that tanks 17–20 have been added and have been in use since October 1974. The only constraint to this method is that the water from outdoor tanks has to be allowed to adjust to the right temperature before being transferred into the larval rearing tanks.
With the exception of tanks 2–7, which are the original wooden fibreglass-lined tanks used in the prototype hatchery, all the tanks are made of concrete aggregate blocks, faced interiorly with cement and painted with either polyester resin or with swimming-pool paint, in order to prevent the leaching out of toxic substances from the cement. Water is transferred between tanks by flexible polythene piping with a 200 l/min pump. Each tank has a stand pipe drain fitted interiorly with a Nitex filter1 in order to prevent larvae from being washed out when changing water.
All the interior tanks have a continuous air supply from a blower providing about 355 l of air/min at 0.1 kg/cm2. The phytoplankton tanks have a separate air supply from two diaphragm pumps. An adequate air supply for both oxygenation and circulation of water is a vital necessity for successful larval rearing.
4.1 Treatment of females
In Mauritius, berried females are present in the population throughout the year. Records of the breeding stock in Pamplemousses pond No. 1, from November 1972 to December 1973, show that the percentage of berried females ranged from 14.2–36.7 percent with the highest incidence occurring during May. The smallest berried female recorded was 7.7 cm in length and was captured in February 1974.
For each cycle 400 000 larvae are needed if all the tanks are to be stocked at 12 larvae/l. In order to get sufficient numbers of larvae hatching synchronously, 20–40 berried females are brought into the hatchery from Pamplemousses pond No. 1 despite the fact that this number should produce at least twice as many larvae as are needed. A larger number of berried females may be needed if they are of small size, as there is a direct relationship between body weight and number of eggs.
1 Eastside Net Shop 14207 100th NE, Bothell, Washington 98011, U.S.A.
Freshly deposited eggs are bright orange in colour. As the embryo develops, the colour of the eggs changes from orange, to yellow, to grey. The total period of embryonic development is 19–22 days at 26–29°C (Ling, 1967a). This colour change permits a choice of females whose eggs will hatch when larvae are required for stocking. Females with orange eggs are kept in fresh water and fed fish scraps. The water is changed daily. Females with grey eggs are then transferred to water at a salinity of 6–10 ppt at a temperature of 28–29°C.
4.2 Larval tank management
According to the procedure followed at present, tanks 2–7 are stocked with the freshly hatched larvae at a concentration of about 25/l. Larval numbers are estimated from the hatching tanks by counting larvae from 30 sub-samples of known volume. This method is only possible if the volume of the hatching tank is fairly small and if the distribution of larvae is assumed to be homogeneous by action of vigorous aeration of the water.
The larvae are kept in these tanks for 10 days after which, when most of the larvae will have moulted four times (about stage 5 or 6), they are transferred to tanks 1, 8, 9, 10, 11 and 12. They are stocked at a density of 12/l for the remaining part of their larval development.
The larvae are fed with tuna fish flesh and brine shrimp (Artemia salina). The size of the feed particle and size of prawn larvae are directly related hence the size of the food is changed with the age of the larvae. Commencing after day 5, the prawn larvae are fed freshly hatched 2 day- or 3 day-old brine shrimp. Tuna fish flesh is broken up into small muscle fibres by using water pressure to force the flesh through a 3.2 or a 6.4 mesh/cm2 stainless-steel seive. The amount fed is determined by direct observation of the rate of consumption by prawn larvae, i.e., feeding upon demand. The water in all tanks is flushed every two days and the detrital matter is siphoned out of the bottom of each tank twice daily and examined for the presence of dead larvae. Using this method, survival through metamorphosis into juvenile prawns of 50–60 percent of larvae stocked is currently achieved.
Metamorphosis of all the larvae in a tank may sometimes take from 10–15 days, particularly if the larvae have been stressed during their development (temperature changes, inadequate feeding, poor oxygenation of the water, etc.). It has been observed that a high rate of cannibalism takes place during moults and if the post larvae are kept for more than 10 days in the hatchery up to 10 percent are lost every seven days. At Trou d'Eau Douce, therefore, 10 days after the appearance of post larvae in a tank, the remaining larvae are siphoned into another tank to complete their metamorphosis and the juveniles are removed, counted, and sent to the culture ponds.
4.3 Juvenile production
From February 1973 to July 1974 the production of juvenile prawns at Trou d'Eau Douce was 146 108. This production is far below the theoretical capacity of the hatchery and the possible reasons for this are listed below, together with some of the remedies applied:
It was discovered that the sea water had to be aged for at least three days before use, possibly to eliminate hydrogen sulphide produced by the decomposition of organic material in the brackishwater inlet and saltwater piping.
One cycle was lost through a sudden temperature drop when the doors and windows had not yet been fitted to the building. Since then, temperatures are recorded in all tanks at 2-h intervals.
Overfeeding of tuna flesh (6–9 times a day) resulted in bacterial blooms in the detritus of the tank floors. Feeding is now carefully monitored and the larvae are fed on demand 2–3 times a day, the amount dependent on the larval stage.
Production of phytoplankton in indoor tanks was never satisfactory and this is thought to have been one of the main factors causing poor production. Phytoplankton is now grown in outdoor tanks.
An infection of hydroid medusae (believed to be a new species of Odessia (Limnomedusae), Kay W. Petersen, personal communication) caused heavy mortality of several cycles. Larvae were seen with nematocysts from the medusae stuck into their exoskeletons. All the hatchery system was cleaned with orthophosphoric acid or 15 percent chlorine solution. Attention to cleanliness of piping, filters, and tanks has prevented recurrence of the problem.
The Trou d'Eau Douce hatchery was closed down for renovation and repairs in August 1974. The hatchery resumed operation in mid-October 1974 and in four months the production has been 330 000 juveniles. With these juveniles, 5 ac have been stocked. It is expected that when the hatchery ceases production in winter (mid-June) another 500 000 juveniles will have been produced or sufficient for 7 ac of ponds.
5.1 Site selection
The criteria for the selection of pond sites are as follows:
a climatic region having a temperature ranging between 18 and 35°C, with an optimum range of 27–31°C;
the availability of uncontaminated fresh water with sufficient flow to fill the ponds initially, to flush the ponds if necessary, and to maintain a flow of 225 l/min/ha of pond surface on a year-round basis. Poor results have been obtained in Hawaii in ponds using a hard water source and it is recommended to use water with a minimum calcium and magnesium content, in no case exceeding 1 000 ppm total hardness;
in order to avoid expensive excavation, the choice of fairly level ground is to be preferred. Rocky soils should be avoided in order to reduce excavation costs, particularly as level pond bottoms are required for ease of seining, and rocky pond bottoms are difficult to seal.
Pond layout depends on site characteristics, but each pond should receive its water directly from the common source by conduits equipped with some means of regulating the water flow, such as valves, baffles, etc. Similarly, pond outlets should be into a drainage canal, not into other ponds.
5.2 Pond construction
Figure 3 shows pond plans prepared under the guidance of T. Fujimura (personal communication). Although the linear dimensions can be modified to suit site characteristics, the rectangular ponds shown are easier to manage. The length of the pond can be increased at will to obtain larger water areas but an increase in width makes seining and feeding difficult. While most of the ponds built in Mauritius have been of about 0.2 ha area, ponds of 1.25 ha are being tried in Hawaii. This does reduce the space lost in dividing walls but the gain may be outweighed by management problems.
The maximum depth of the pond is not critical except in relation to seining operations when the fishermen should have secure footing. Areas shallower than those indicated should be avoided, however, as the water can become overheated in warm weather with a consequent reduction in oxygen tension. The plan shows the bottom of the pond to have a gradient of 0.3 m. This is desirable in order to drain the water out when reconditioning the pond.
It should be noted that, in order to obtain a pond with an average depth of 1 m and height of 1.5 m between the pond bottom and the top of the walls, the soil need only be excavated by 0.3–0.6 m. In areas of shallow topsoil, it is recommended to scrape all the topsoil into the centre of the pond area, to build the retaining walls with available rubble, and afterwards to spread out the topsoil again on the pond bottom. By this means, better pond fertility and sealing is achieved especially in rocky and sandy areas where seepage can be a major problem. Further measures to reduce seepage are as follows:
compacting of pond walls and bottom with heavy rollers or bulldozers;
“puddling” of pond bottoms with a rotovator and the addition of clay;
heavy fertilization to produce phytoplankton blooms - the planktonic detritus rapidly seals leaks.
Proper design of the inside slopes of the pond walls (usually 3:1) is extremely important for two reasons: steeper walls make seining difficult; and they are more subject to erosion damage by surface waves. The prawns often tend to stay near the banks of ponds and, unless the slope is sufficiently gentle to allow a seine to maintain contact with the bottom, many prawns escape around the edges and under the seine.
The placement of the inlet pipe and outlet sluice gate on diagonally opposite ends of the pond is thought to help the circulation of water throughout the pond, avoiding stagnant spots. In practice, it has been observed that wind action on the pond surface is the major water circulation agent and, for this reason, it is preferable to place the outlet so that the prevailing wind carries the detritus out of the pond.
Figure 4 shows the recommended outlet design. Concrete sluices are to be preferred to stand pipes which frequently become obstructed with mud. The sluice design, with a gap under the first set of sluice boards, allows water to be drawn from the bottom of the pond, while the water level is controlled by the second set of boards. The insertion of a third set creates a holding area which can be used to contain prawns for sampling purposes or for marketing. A sluice gate also permits rapid flushing or drainage of ponds.
Screening (1-mm mesh plastic or metallic screening) of both inlet and outlet water is needed to control the entry of predators and exit of prawns. The latter problem can also be controlled by placing the inlet canal or pipe above the water surface as prawns prefer to move upstream. Water splashing down from the inlet pipe also helps to oxygenate the pond.
Saron screening is indicated across the pond surface. This is normally only needed for a period after the pond has been stocked in order to provide shade and shelter for the juvenile prawns. When a good phytoplankton bloom has been obtained in a pond, the screening can be removed. Undesirable algal growth on the pond bottom can also be controlled by shading with saron screening.
6.1 Pond preparation
Prior to the first stocking of juvenile M. rosenbergii, a pond should be drained and cleaned out of all predators such as fish or dragonfly nymphs. A fertilizer treatment -NPK at the rate of 65–125 kg/ha - is applied to the pond bottom and the pond is filled with water 7–10 days before stocking. This should allow time for a phytoplankton bloom to develop inside the pond; an adequate bloom is obtained when visibility is reduced to about 45 cm. Once the bloom is established, it can be maintained by placing bags of insecticide-free manure (chicken, duck, cow, etc.) near the inlet side of the pond. A flow of water at a rate of 225 l/ha/min is maintained in the pond, partly for oxygenation, and partly in order to cope with evaporation and seepage. The phytoplankton bloom should not be allowed to get too heavy as this can result in conditions of oxygen depletion in early morning hours. Control measures can include the removal of fertilizers and flushing out of the pond by an increased water flow.
Dragonfly nymphs are predatory on juvenile M. rosenbergii, and as a measure of biological control Lebistes reticulatus, which eat dragonfly eggs and nymphs, are stocked in the pond prior to the introduction of the prawns. It is thought that these small fish which breed rapidly are also an important food item in the diet of large prawns. Prawns have been observed to eat the fish in daytime but catch them more easily at night when the prawns are actively feeding and the fish less active.
6.2 Stocking
The recommended stocking rate for maximum production of 3 500 kg/ha/year is 175 000 juveniles/ha (Fujimura, personal communication). Lower densities result in faster growth rates but lower total production. Juveniles are transported from the hatchery to the growing ponds in sealed polythene bags containing oxygen, with 1 500–2 000 juveniles per bag. The juveniles are very sensitive to rapid temperature changes and the bags are normally immersed in the pond to equilibrate to the pond temperature before the juveniles are released. In order to reduce the chances of predation by fish, and also to avoid problems of temperature stratification which sometimes occur in ponds, the juvenile prawns are released on the pond bottom. The possibility of stocking ponds at night should also be envisaged, especially when adult prawns are already present.
6.3 Feeding
Under these semi-intensive rearing conditions, the prawns require supplemental feed. As M. rosenbergii is largely a scavanger, the feed composition does not appear to be critical (Fujimura, 1970) but the form in which the feed is supplied is important. Although pelleted feed should ideally be supplied, pressure pelleting is inadequate as the pellets dissociate rapidly in water. The use of a binder (high-gluten wheat flour, alginate, etc.) is therefore necessary to hold the pellets together. In Mauritius, this type of feed is not commercially available and prawn farmers have been using broiler starter (chicken feed) in powdered form. This feed has to be wetted and compacted into balls prior to introduction in the ponds as it will otherwise float. The feed is always placed in the same 5 or 6 positions around the pond edges, generally in the early evening. This type of feed, however, disperses rapidly and a large part is unavailable to the prawns.
Fujimura (personal communication) recommends that M. rosenbergii should be provided with supplemental feed at a daily rate of 3 percent of the total biomass in the pond. Based on the observed growth rates and projected survival, the following indicative feed schedule is recommended to prawn farmers in Mauritius. The amount should however be adjusted to follow a “feeding on demand” schedule.
TABLE I
Indicative prawn feeding schedule supplied to farmers in Mauritius
| Age of prawns | Weight of feed (kg/day) | |
| 0.2 ha | 0.4 ha | |
| 0–15 days | 0 | 0 |
| 15 days-1 month | 0.125 | 0.250 |
| 1–2 months | 0.250 | 0.5 |
| 2–3 months | 0.5 | 1 |
| 3–4 months | 4 | 8 |
| 4–5 months | 5 | 10 |
| 5–6 months | 6.5 | 12 |
| 6–7 months | 8 | 16 |
| 7–8 months | 12 | 20 |
Harvesting of the prawns in Mauritius starts after the seventh of eighth month and the biomass of prawns in the pond does not increase greatly after this time if fishing pressure is maintained.
6.4 Growth and production
For maximum production from a prawn pond it is necessary to remove the prawns as soon as they reach marketable size as competition from large prawns greatly retards the growth of the smaller animals. In Mauritius, the marketable size is taken to be at a weight of 50 g or an average length of 12 cm, measured from the base of the eyestalk to the tip of the telson. Selection of this size range of individuals is done automatically by the use of a seine net with a mesh size of 5 cm stretched (25-mm bar). Monofilament nylon is used for the netting as the exoskeletal spines on the prawns catch in twisted yarn, retaining small prawns in the net. After the seventh month, by which time some 3 percent of the prawn population can be expected to have reached marketable size, the practice is to seine each pond twice a month. Only half the pond is seined on each occasion in order to avoid stressing the prawns by disturbing the whole pond bottom. Seining is carried out very slowly in order to avoid stirring up mud and to keep the leadline in close contact with the bottom. This also allows the small prawns to escape through the netting while keeping the large prawns ahead of the net where they are collected in the pocket.
In the experimental phase of pond management each pond is sampled regularly for size composition of the prawn population and for survival of the prawns. This is done by measuring and weighing a random sample of the population caught with a small-meshed net so that all size classes are included. At the same time, data on sex ratios, the number of berried females, evidence of disease, general condition, etc., are collected. Prawn survival is assessed by a mark-recapture method previously described (Ardill et al., 1973). It is hoped that, in the full commercial phase of the operation, less laborious methods will be evolved to monitor the prawn population pond performance but it is already apparent that, for successful culture, a periodic check must be maintained.
6.5 Control of aquatic plants and filamentous algae
In the management of prawn ponds a number of problems have been encountered with aquatic plants. Unless phytoplankton growth is very dense, preventing penetration of light to the pond bottom, most ponds develop growths of aquatic plants, mainly Hydrocharitaceae. These growths make seining difficult as the seine either becomes clogged with detached vegetation or lifts above the clusters of plants allowing the prawns to escape from the net underneath the leadline. However, it is probable that these plants provide shelter for the small prawns, and may also provide a substrate for invertebrates which are eaten by the prawns. In one pond (Pamplemousses pond No. 2) where the prawns were possibly not fed sufficiently for a period of time, plants disappeared completely, apparently having been eaten by the prawns. It is probable, therefore, that these growths are beneficial to the pond ecology, at least until the harvesting begins in the pond.
Filamentous algae (Spirogyra spp.) have also been observed to grow in prawn ponds and they are definitely undesirable. Heavy prawn mortalities have been observed in ponds having extensive growths of these filamentous algae (Baissac, in print). Deaths were thought to be due to the stress caused by the entanglement of the prawns in the algal filaments. In ponds with algal growth, prawns have also been found with pink spots, black spots, and perforated black-ringed areas on the carapace. It is not known whether these anomalies (possibly of bacterial or fungal origin) result in mortalities but they undoubtedly lower prawn market value. Finally, filamentous algae make seining very difficult as the nets become filled with algae which also entangle the prawns. Sorting through these masses of plant material is time-consuming and may result in the mortality of small prawns which are missed in the sorting or smothered before they can be found. The prawns in contaminated ponds are also seen to avoid algal accumulations, resulting in overcrowding of the remaining pond areas.
Several different methods have been tried to remedy the problem of filamentous algae. These include covering algal patches with polythene sheeting to screen out light, seining, and the introduction of male Tilapia macrochir into the ponds. These methods are still under evaluation and no specific recommendations can yet be made.
6.6 Control of bilharzia
In Mauritius, there exists a parasitic blood fluke, Schistosoma haematobium, which, although not affecting the prawns, could endanger the health of fishermen and others employed in the freshwater pond-rearing phase of M. rosenbergii. The disease, locally known as bilharzia, causes urogenital pathologies.
The life cycle of the parasite is complex and involves an intermediate host. The source of infection to man is small cercariae which are released into water after developing in the liver and gonads of a specific species of freshwater snail in Mauritius, Bulinus cernicus. The free-swimming cercariae are able to penetrate rapidly the human skin and enter the circulatory system, finally settling in the veins of the urinary bladder where they develop into adult worms of 1–2 cm in length. These adults in turn release eggs which penetrate the bladder wall and enter the lumen. A population of vector snail may not be harbouring cercariae, but urination of only one infected person in the water where the snail vector occurs is sufficient to start the entire parasite life cycle.
Two types of preventative measures are being recommended to prawn farmers: (i) prevention of entry of snail vector into rearing ponds and constant vigilance for presence of snail vector; and (ii) medical history and examination of fishermen and others entering the ponds.
7.1 Private sector
Response from the private sector to the mass culture of M. rosenbergii has been immediate. Within one year of the introduction of this prawn to Mauritius the WEAL group of sugar estates had set up their own pilot hatchery, run by technicians trained at the Trou d'Eau Douce government hatchery. Since that time, they have appointed an aquaculture chief on their staff who has been to Hawaii for training in Macrobrachium culture. This hatchery at Beau Rivage now has a capacity equal to that of Trou d'Eau Douce and their 22 ponds (0.2 ha each) should be fully stocked at a commercial level by June 1975.
Concurrently with this the Mauritius Sugar Producers' Association has formed a new company for the production of juvenile prawns. The hatchery, which is to be completed by late 1975, has a planned production of 10 million juvenile prawns annually. The parent sugar companies have already given a commitment to build 30 ha of prawn ponds in the suitable lowland areas.
7.2 Research
While less spectacular, response from non-sugar interests has also been good and there are sufficient potential prawn farmers for government to have decided to build a second large-scale multi-purpose hatchery at Albion. While this hatchery will also have a potential production of some 10 million juveniles, it is being planned with a view to eventual production of other crustaceans, molluscs and both fresh- and saltwater finfish. Total hatchery production in Mauritius will therefore permit the stocking of over 300 ac (123 ha) of prawn ponds.
With such a rapid development in mass culture of M. rosenbergii in Mauritius, data will rapidly be available from a large number of ponds covering a variety of climatic and topographical regions and it is to be hoped that this will become one of the major research centres for the culture of freshwater prawns, which in turn should help the development of other forms of aquaculture.
Ardill, J.D. et al., 1973 The introduction of the freshwater prawn, Macrobrachium rosenbergii (de Man), into Mauritius. Revue Agricole et Sucrière de l'Ile Maurice, 52:6–11
Baissac, de Boucherville P., 1975 Problems caused by the occurrence of the filamentous algae Spirogyra sp. in culture ponds of Macrobrachium rosenbergii (de Man), in Mauritius. Revue Agricole et Sucrière de l'Ile Maurice
Fujimura, T. and H. Okamoto, 1970 Notes on progress in developing a mass culture technique for Macrobrachium rosenbergii in Hawaii. IPFC 14th Session, Bangkok, Thailand, 18–27 Nov., 1970
Ling, S.W., 1967a General biology and development of Macrobrachium rosenbergii. FAO Fisheries Reports (57) Vol.3:589–606, Proceedings World Scientific Conference on Biology and Culture of Shrimps and Prawns, Mexico City, Mexico, 12–21 June 1967
Ling, S.W., 1967b Methods of rearing and culturing Macrobrachium rosenbergii. FAO Fisheries Reports (57) Vol.3:589–619, Proceedings World Scientific Conference on Biology and Culture of Shrimps and Prawns, Mexico City, Mexico, 12–21 June 1967
Uno, Yutaka and Kwon chin soo, 1969 Larval development of Macrobrachium rosenbergii (de Man) reared in the Laboratory. J. Tokyo Univ.Fish., 55(2):179–90
Fig. 1. Rearing cycle of Macrobrachium rosenbergii.
Fig. 2. Schematic representation of the hatchery at Trou d'Eau Douce (not drawn to scale). Tank capacities are given in both gallons (in brackets) and liters.
Fig. 3. Plans of prawn culture ponds. (dimensions are given in feet and inches).
Fig. 4. Detail of outlet sluice gate of prawn pond (dimensions are given in feet and inches).
