Laboratory Animals (1980) 14, 137-142

137

A closed marine culture system for rearing Octopus joubini and other large-egged benthic octopods
J. W. FORSYTHE & R. T. HANLON
Marine Biomedical Institute, University oJ Texas Medical Branch, Galveston, Texas 77550, United States oj America

Summary The system consists of 2 adjoining 150 litre aquaria, one functioning as the water-conditioning tank and the other as the principal rearing tank. Water quality remained high with biological and mechanical filtration, physical adsorption and ultraviolet-light disinfection taking place exclusively in the conditioning tank. The pH ranged from 758 to 800 and ammonia and nitrite levels never exceeded 0004 mg/l and 0198 mg/l, respectively. Nitrate levels were maintained at 40 mg/l or less with no adverse affects. Adult octopuses readily mated and females produced 50-150 eggs, with 95% hatching success. When fed small live crabs, the octopus hatchlings were reared to sexual maturity either in groups or individually in about 120-150 days. Growth rates (4% bodyweight/day) and food conversion efficiences (30-400A) were as high as those obtained in open systems by previous workers. The works of Young (1971), Wells (1978) and others have established the value of the octopus as an important and reliable research animal, and additional interest has developed in its mariculture potential
(Voss, 1973; Van Heukelem, 1976; Hanlon, 1977). To

undeserved reputation for being difficult to rear or maintain: species that produce large, telolecithal eggs and large benthic young (i.e. direct development with no larval stage) are, in fact, reasonably easy to grow. In this paper we describe systems and methods we have developed for the closed system rearing of the pygmy octopus Octopus joubini Robson, 1929 (Fig. 1) during the past 3 years. O. joubini is a small benthic octopod inhabiting shallow water along the east coast of Georgia, the east and west coasts of Florida, the north central Gulf of Mexico and the northwestern Caribbean (Voss, Opresko & Thomas, 1973; Kraeuter & Thomas, 1975; Lipka, 1975). This species is well suited to laboratory culture since it has a maximum length of only 10-15 cm and produces 2 generations per year. The techniques described for O. joubini should be applicable to any of several other largeegged benthic octopods that have been successfully reared in open systems. System design and rearing methodology The basic system consists of 2 all-glass aquaria (each 150 litre; 122 x 32 x 42 em) situated end-to-end (Figs

date, most maintenance and rearing work on benthic octopods has been done in open or flow-through marine culture systems (Boletzky, 1974). These systems are proven and reliable, but they require a coastal location with good water quality, and they offer little or no control over temperature and salinity fluctuations, disease organisms, turbidity and pollutants. Recently, the understanding and implementation of the closed seawater system approach has increased (Spotte, 1973, 1979; Siddall, 1974), making it an attractive alternative to researchers by providing a stable, controllable and reproducible marine environment. With the use of artificial seawater, closed marine systems can be operated in any locale. The general reluctance to deal with marine research animals derives primarily from the misconception that maintaining small-volume marine systems is too difficult, expensive, or labour-intensive. In addition, the cephalopods-squids and octopuses-have a partially
Received 2 October 1979. Accepted 4 December 1979.
Fig. I. Adull OClopusJoubinJ. Mature male. about 180 days old and 15 g body weight, grown from eggs laid in the laboratory.

138

Forsythe & Hanlon

REARING TANK

WATER CONDITIONING

TANK

Fig. 2. Basic closed marine culture system for rearing OctopusJoubini. A water exchange pump. B return syphons. C crushed oyster sheUfilter bed. D air system. E air-lift. F auxiliary tank filter. G protein skimmer. H heater. I ultraviolet light filter. J hatching tank. K tray tank. L small rearing chambers in rack. M large rearing chamber. J and K can be incorporated into the basic double tank system. It is not usual for
aU rearing components to be in use simultaneously as shown.

REARING

TANK

WATER

CONDITIONING

TANK

Fig. 3. Diagrammatic representation of culture system. For key to lettering see Fig. 2.

2 and 3). One tank served as the water-conditioning tank and the other as the principal rearing tank. A water exchange pump (A) and return siphons (B) provide continuous water movement between tanks at a rate of 65 l/min. In the conditioning tank the seawater is continuously subjected to biological and mechanical filtration, physical adsorption and ultraviolet-light disinfection. A 6 cm deep crushed oyster shell filter bed (C) is the principal site of biological filtration. A low-pressure, high-volume air system (D) powers 4 air-lifts (E) which draw water through the filter bed (C) at a rate of 15 l/min, allowing nitrifying bacteria living in the oyster-shell bed to oxidize toxic ammonia (NH3) to nitrite (N02), and finally to less toxic nitrate (N03). Mechanical filtration of suspended particulate matter and detritus takes place in the filter bed (C) and in side-mounted auxiliary

tank filters (F). Each of these auxiliary filters (E. G. Danner Mfg Inc., 160 Oval Drive, Central Islip, New York 11722, USA) is powered by a pump that draws water at a rate of 65 l/min through 2 layers of polyester filter fibre and a 4 cm layer of activated carbon. The fibre reduces water turbidity and the level of organic colloids. The carbon provides physical adsorption of dissolved organics; half of it is changed every 2-4 weeks, depending upon the animal load in the system. A protein skimmer (G) removes surfaceactive organics by foam fractionation. An 8 W, 2537 nm ultraviolet-light filter unit (1; Hawaiian Marine Imports, 465 Town & Country Village, Houston, Texas 77024, USA) provides ultraviolet-light disinfection to reduce populations of bacteria, algae and protozoans that are often introduced with wild-caught food organisms. The ultraviolet-light filter unit receives

Breeding Octopus in the laboratory water from an auxiliary tank filter pump, but because of back pressure in the filter, flow rate is reduced from 65 to 45 l/min. A standard 150 W immersible aquarium heater (H) maintains constant water temperature. Natural and artificial seawater ('Instant Ocean'; Aquarium Systems Inc., 33208 Lakeland Boulevard, Eastlake, Ohio 44094, USA) have been used in this system. Natural seawater is always passed through a 1 .urn diatomaceous earth filter prior to use, and artificial seawater is made with deionized water to avoid tap-water contaminants. Water quality is monitored regularly to assess the effectiveness of the conditioning system. Temperature, salinity and pH are measured several times weekly and estimates of nitrite and nitrate concentrations are made once a week using field-kit chemical tests (Hach Chemical Co., PO Box 907, Ames, Iowa 50010, USA). Periodic detailed water chemistry tests are performed according to the methods of Rand, Greenberg & Taras (1976) to determine more precisely ammonia, nitrite and nitrate concentrations. When concentrations surpass the recommended limits proposed by Spotte (1973), 25% of the seawater is replaced. Deionized water is added several times a week to replenish that lost by evaporation. The rearing tank may be used for stocking adult and juvenile O. joubini as well as rearing the hatchlings. The bottom is covered with a 1 cm layer of crushed oyster shell that serves only as a bottom substrate. Clay pots, polyvinyl chloride (PVC) pipe fittings and shells provide shelter or 'dens' for individual octopuses older than 120 days and greater than 3 g wet weight. At this size the octopuses are large enough to use the items provided for dens, whereas smaller octopuses often hide in the oyster shell substrate, where it is difficult to locate them. They are fed live crabs ad libitum, and food debris and faeces are syphoned out daily. Adult and juvenile octopuses should not be released in the filter tank as they will burrow into and under the filter bed, disrupting filtration. Several adults of both sexes maintained in the same rearing tank will mate readily. When eggs are laid, the den containing the brooding female and eggs is moved to the hatching container (Fig. 21), a 10-15 I rectangular clear-plastic aquarium (32 x 18 x 18 cm) placed in the system (in Fig. 2 it is placed in a separate water conditioning tank). A pump and return syphon provide continuous water exchange and fine plastic screening (approximately I mm2 mesh) is placed over the syphon intakes to prevent hatchling escape. If provided with moderate water circulation, eggs separated from the mother will develop normally in the hatching container. Females generally lay 50-150 eggs; development takes 30-40 days at 25C, with hatching occurring over 2 weeks (Boletzky & Boletzky, 1969; Opresko & Thomas, 1975).

139

Fig. 4. Juvenile Octopus joubini, 7 weeks old, in a small rearing chamber. Showing mesh sides, crab remains (right) and sheil fragments (beneath animal and left) provided as den material. Line represents 10 mm.

The hatching container is checked daily and the new hatchlings removed either into the shallow clear-plastic tray tank (76 x 30 x 10 cm, Fig. 2K) for mass rearing or into individual compartments (Fig. 2, Fig. 4). The tray tank is connected to a conditioning tank by a pump and return syphons and has a water depth of 4-5 cm. Squares of plastic artificial grass inverted and floated in the tank provide shelter for the hatchlings, and the close proximity of the grass to the tank bottom allows individual octopuses easily to observe and capture prey (Hanlon, 1977). An excess of live food is provided, and uneaten debris is removed daily. After 3-4 months, the young octopuses are large enough to be transferred to the rearing tank. Individual octopuses can be reared in separate chambers for any part or all of their life cycle. Octopus hatchlings and young octopuses are isolated in small rearing chambers made from disposable 100 ml plastic beakers (7 cm deep x 5 cm diameter) whose walls have been replaced with fine mesh screening to allow water circulation while preventing animal escape (Fig. 4). Groups of up to 40 chambers are supported in rigid plastic racks at the water's surface in either the rearing tank or water-conditioning tank. At a wet weight of 2-3 g, young octopuses are moved to larger rearing chambers (Fig. 2M), plastic trays (28 x 17 x 13 em) whose sides have been replaced with 3 mm2 mesh plastic screen. The same screening is used to divide the chamber, and an octopus is placed in each half. 4 of these chambers can be supported in the rearing or water-conditioning tank. All chambers are supplied with den material such as shells, and daily maintenance involves only the syphoning out of uneaten food remains from each chamber and the addition of new food animals.

140 Results This closed system has proved reliable for long-term rearing and study of O. joubini. Water quality has remained consistently high, with pH 758-800 and ammonia and nitrite levels never exceeding 0004 mg/I and 0198 mg/I respectively. Nitrate levels gradually rise with increased animal load and it is necessary to monitor nitrate accumulation regularly. We found that O. joubini could tolerate nitrate levels up to 150 mg/I for short periods with no obvious adverse affects. However, when nitrate levels exceeded 40 mg/I, 25% of the water was changed: this was required once or twice a month. Water temperature was maintained at 25C and salinities at 34-37 parts per thousand. The simple methods and the accessibility of the apparatus resulted in minimal maintenance requirements, less than an hour each morning, or slightly longer when water quality was measured or water changed. O. joubini has been successfully reared through its life cycle in this system. All rearing work with this species relied on live food, principally fiddler crabs (Uca spp.) and xanthid mud crabs ranging in size from 2 mm carapace width for octopus hatchlings, to full grown crabs (maximum carapace width 3 cm) for adult octopuses. Small penaeid shrimp (Penaeus spp.), palaemonid shrimp (Palaemonetes pugio) and mysid shrimp (Mysidopsis almyra) of an appropriate size can be used, though less preferred by the octopuses. The small marine gammarid amphipod Orchestia grillis was readily accepted by octopus hatchlings and seems to be as acceptable a live food as small crabs. Dead food such as crab, fish or shrimp meat was accepted only reluctantly, but can be used to maintain adult octopuses temporarily if no live food is available. Octopus hatchlings conditioned to eat dead food have poor survival and little growth (personal observations; Bradley, 1974; Opresko & Thomas, 1975). Wild-caught adults have been maintained for up to 5 months, laying numerous broods of eggs. Hatching success of both brooded and unbrooded eggs has been 95% or better. The hatchlings have been reared through their life cycle both in groups and singly up to an age of 331 days, with a live wet weight of 33 g and a mantle length of 46 cm. In one experiment 50 hatchlings were stocked in the tray tank with a 20% survival rate to sexual maturity (120-150 days). Most deaths occurred in the first 30 days because of an inadequate food supply. 3 females in this group laid fertile eggs and the resultant F 2 generation was reared another 3 months until rearing studies were intentionally ended. In a 2nd experiment 20 individual octopuses were reared simultaneously in separate rearing chambers with 85% survival to sexual maturity. The use of rearing chambers allowed the growth of individual octopuses to be followed. Precise weighing and measuring was accomplished by nar-

Forsythe & Hanlon

cotizing octopuses for 60-90 s in a 15% solution of ethanol in seawater. Through the first 16 weeks an average growth rate of 4% bodyweight/day was obtained at temperatures of 23 to 25C. This is the same growth rate obtained by Opresko & Thomas (1975) rearing Octopus joubini in open system aquaria. Food conversion efficiencies, averaging 3040%, were consistent with those determined for other benthic octopod species reared in open systems (Nixon, 1966; Borer, 1971; Mangold & Boletzky, 1973; Van Heukelem, 1976). For a small species like O. joubini, the tray tank is capable of holding 30-40 hatchlings and their food animals with assured water quality for the first 3 months before transfer to the rearing tank. The rearing tank can in turn hold 30-40 transferred animals for another 2 months, but as the animals attain full size (15-35 g) no more than 15 adults should be kept in a single rearing tank. Space is the major constraint on the number of animals that can be held in rearing chambers. If both the rearing and water-conditioning tanks are used, the system can accommodate either 75 small rearing chambers or 8 large rearing chambers. Other octopus species have been kept in this system. Large adult individuals of O. vulgaris of up to 1 kg body weight have been held for up to 4 months. A wild-caught juvenile O. defilippi was reared from approximately 12 mm mantle length to 41 mm mantle length in 151 days with infertile eggs being laid on day 143 (Hanlon, Hixon & Forsythe, 1980). This was the first recorded rearing of O. defilippi. Discussion This work represents the first reported rearing of any octopus species in closed-system aquaria for its full life cycle. It was possible to duplicate all growth and rearing work done with O. joubini in open systems with comparable results. Daily maintenance and animal care was neither time-consuming nor complicated, and all equipment and materials used were commercially available and easily obtained. O. joubini is a hardy and durable species well suited to laboratory experimentation. Hatchlings reared in this closed system were found suitable for longduration (12 h) respiration and carbohydrate uptake studies (Castille & Lawrence, 1978). At any age it can easily be narcotized to permit precise weighing and measuring, or minor operative techniques. Being small, it requires minimal laboratory space for rearing through the relatively short life cycle. It is tolerant of handling, and transport of eggs or gravid females in sealed plastic bags filled with oxygenated seawater is simple and reliable (personal observations; Bradley, 1974; Van Heukelem, 1976), and could be used in starting a laboratory culture. The closed system described here should be suitable for rearing large-egged benthic octopods that

Breeding Octopus in the laboratory

141

have been successfully reared in open systems, such as Hapalochlaena maculosa (Hoyle) from the Indo-west Pacific (Tranter & Augustine, 1973), Octopus briareus Robson, 1929 from the Caribbean (Wolterding, 1971; Hanlon, 1977), O. maya Voss & Solis, 1967 from the Gulf of Mexico (Van Heukelem, 1976), O. bimaculoides Pickford & McConnaughey, 1949 from California (R. A. Underhill, personal communication, 1973) and Eledone moschata (Lamarck, 1799) from the Mediterranean (Boletzky, 1975). Holding capacity for hatchlings and young juveniles of these species should be about the same as those mentioned for Octopus joubini. At 3-5 months of age, all except H. maculosa surpass the adult size of O.joubini (15-35 g) and grow on to 1-2 kg. 3 adults of this size could be maintained adequately in the rearing tank, but a secure top would be needed to prevent them from escaping. In addition to octopuses, there are 13 species of sepioid squids that are benthic and produce large eggs (Boletzky, 1974) that might also be suitable for rearing in this system. The system would have to be duplicated

or otherwise enlarged to increase the holding capacity. We conclude that rearing large-egged, benthic octopods in closed systems is a feasible alternative in situations where open systems are either unavailable or undesirable. The use of this system, which is convenient, reliable and repeatable, can facilitate the use of these highly evolved molluscs in such diverse fields as physiology, pharmacology, behaviour and genetics.

Acknowledgements We thank Professor G. L. Voss, Dr D. V. Aldrich and R. F. Hixon for critically reviewing the manuscript. D. A. McConathy prepared Figs 2 and 4. John W. Forsythe thanks the Graduate College of Texas A & M University for allowing the use of that part of the material that is being incorporated in his M.S. thesis. This work was supported in part by the Marine Medicine General Budget account #7-11500-765111 of the Marine Biomedical Institute.

References

Boletzky, S. v. (1974). Elevage de cephalopodes en aquarium. Vie et Milieu 24,309-340. Boletzky, S. v. (1975). Le developpement d'Eledone moschata (Mollusca, Cephalopoda) elevee au laboratoire. Bulletin de la Societe zoologique de France 100,361-367. Boletzky, S. v. & Boletzky, M. V. (1969). First results in rearing Octopus joubini Robson, 1929. Verhandlungen der Naturforschenden Gessellschaft in Basel 80, 56-61. Borer, K. T. (1971). Control of food intake in Octopus briareus Robson. Journal oj Comparative and Physiological Psychology 75, 171-185. Bradley, E. A. (1974). Some observations of Octopusjoubini

food by Octopus vulgaris. Journal oj Zoology, London 150,1-9. Opresko, L. & Thomas, R. (1975). Observations on Octopus joubini: some aspects of reproductive biology and growth. Marine Biology 31,51-61. Rand, M. C., Greenberg, A. E. & Taras, M. J. (eds) (1976). Standard methods Jor the examination oj water and wastewater, 14th ed. Washington: American Public Health Association. Siddall, S. E. (1974). Studies of closed marine culture systems. Progressive Fish Culturist 36,8-15. Spotte, S. H. (1973). Marine aquarium keeping. New York:

reared in an inland aquarium. Journal of Zoology,

London 173,355-368. Castille, F. L. & Lawrence, A. L. (1978). Uptake of amino acids and hexoses from sea water by octopod hatchlings. Physiologist, Washington 21 (4), 18. Hanlon, R. T. (1977). Laboratory rearing of the Atlantic reef octopus, Octopus briareus Robson, and its potential for mariculture. Proceedings oj the World Mariculture Society 8, 471-482. Hanlon, R. T., Hixon, R. F. & Forsythe, J. F. (1980). The "Macrotritopus Problem" solved: Octopus defilippi raised from a wild-caught, pelagic Macrotritopus. Bulletin oJthe American Malacological Union, 1979. (Abstract, in press). Kraeuter, J. N. & Thomas, R. F. (1975). Cephalopod mollusks from the waters off Georgia, U.S.A. Bulletin oj Marine Science 25, 301-303. Lipka, D. A. (1975). The systematics and zoogeography oj cephalopods Jrom the Gulf oj Mexico. Ph.D. Dissertation, Texas A & M University. Mangold, K. & Boletzky, S. v. (1973). New data on reproductive biology and growth of Octopus vulgaris. Marine Biology 19,7-12. Nixon, M. (1966). Changes in body weight and intake of

Wiley.
Spotte, S. H. (1979). Fish and invertebrate culture: water management in closed systems, 2nd ed. New York: Wiley. Tranter, D. J. & Augustine, O. (1973). Observations on the life history of the blue-ringed octopus Haplochlaena maculosa. Marine Biology 18,115-128. Van Heukelem, W. F. (1976). Growth, bioenergetics and life-span oj Octopus cyanea and Octopus maya. Ph.D. Dissertation, University of Hawaii, Honolulu. Voss, G. L. (1973). Cephalopod resources oJthe world. FAD Fisheries Circular 149. Voss, G. L., Opresko, L. & Thomas, R. (1973). The potentially commercial species oj octopus and squid oj Florida, the Gulf oj Mexico and the Caribbean area. Sea Grant Field Guide Series #2. University of Miami Sea Grant Program. Wells, M. J. (1978). Octopus: physiology and behaviour oj an advanced invertebrate. London: Chapman & Hall. Wolterding, M. R. (1971). The rearing and maintenance oj Octopus briareus in the laboratory, with aspects oj their behavior and biology. M. S. Thesis, University of Miami. Young, J. Z. (1971). The anatomy oj the nervous system oj Octopus vulgaris. London: Oxford University Press.

142

Forsythe & Hanlon

Ein geschlossenes marines Aufzuchtsystem und -verfahren fur Octopusjoubini und andere grosse Eier produzierende auf dem Meeresgrunde lebende Tintenfische J. W. FORSYTHE & R. T. HANLON
Zusammenfassung Oas System besteht aus 2 miteinander verbundenen 150Liter-Aquarien, wobei einer a1s Wasseraulbereitungsbehiilter und der andere als Hauptaufzuchtbehiilter eingesetzt werden. Eine hohe Wasserqualitiit konnte durch die biologische und mechanische Filtrierung, die physikalische Absorption sowie die UV-Licht-Sterilisation, welche nur im Wasseraulbereitungsbehiilter durchgefiihrt wurden, erreicht werden. Oer pH-Bereich lag zwischen 7,58 and 8,00 und die Ammoniak- resp. Nitratspiegel iiberschritten nie 0,004 resp. 0,198 mg/L. Nitratspiegel konnten

bei 40 mg/L oder weniger gehalten werden, ohne dass schiidliche Wirkungen festgestellt wurden. Erwachsene Octopi verpaarten sich problemlos und die Weibchen legten 50-ISO Eier, wobei die Schlupfrate 95% betrug. Bei der Verfiitterung von kleinen lebenden Krebsen wurden die geschliipften Octopi bei Gruppen- oder Einzelhaltung innerhalb von ca. 120-150 Tagen geschlechtsreif. Wachstums(4% Korpergewicht/Tag) und Futterverwertungsraten (3040%) waren ebenso hoch wie bei der Aufzucht in olTenen Systemen, welche von anderen Autoren verwendet wurden.