Aquaculture 263 (2007) 68 74 www.elsevier.

com/locate/aqua-online

A recirculated maturation system for marine ornamental decapods

CCMAR Universidade do Algarve, Campus de Gambelas, 8000-117 Faro, Portugal b Templo Aqutico, Rua de Santa Marta, No. 27, 1150-291 Lisboa, Portugal Universidade Lusfona, Departamento de Cincias Naturais e Biolgicas, Campo Grande, 376-1749 Lisboa, Portugal Received 17 August 2006; received in revised form 9 October 2006; accepted 10 October 2006

Abstract The present work describes an easy to operate recirculated maturation system for different types of marine ornamental decapods that: i) demands shorter periods of time to perform routine tasks, while allowing better water quality for broodstock keeping, ii) eliminates the need to capture ovigerous females (or euhermaphrodites) before larval release, minimizing the risks of disrupting reproductive pairs, iii) separates newly hatched larvae from the reproductive pair, impairing adults from preying on larvae and iv) allows live prey to be provided to larvae immediately after hatching if needed. Breeding pairs of the following species were used to test the maturation system: cleaner shrimp Lysmata amboinensis, fire shrimp L. debelius, Monaco shrimp L. seticaudata, peppermint shrimp L. boggesii, cleaning rock pool shrimp Urocaridella antonbruunii, sexy shrimp Thor amboinensis, dancing shrimp Rhynchocinetes durbanensis, boxer shrimp Stenopus cyanoscelis, S. hispidus, hermit crabs Clibanarius tricolor, C. erythropus and the emerald crab Mithraculus sculptus. Larger species displaying strong agonistic behavior toward conspecifics (L. debelius and S. hispidus) had to be kept in larger divisions (0.20 m 0.30 m 0.15 m), while all other species were successfully kept in smaller divisions (0.20 m 0.15 m 0.15 m). All tested species were able to successfully mate and produce consecutive larval batches during the experimental period, and routine tasks (e.g. checking for and collecting newly hatched larvae, monitoring molts of breeding pairs, recording the presence of specimens carrying embryos about to hatch, tank siphoning, filters cleaning and water changes) were daily performed in about 1 h. The unidirectional water flow inside each maturation tank, as well as the presence of actinic light in the front glass, allowed newly hatched caridean, stenopodid, anomuran and brachyuran larvae to be successfully removed from the chamber containing the reproductive pair. The simple use of 150 m mesh size screens inside maturation tanks allowed larval prey (Artemia nauplii) to be provided to larvae immediately after hatching, avoiding the negative effects of early larval starvation. The use of suitable maturation and larviculture systems will play a vital role for the successful development of profitable commercial scale culture protocols for the most heavily collected marine ornamental decapod species. 2006 Elsevier B.V. All rights reserved.
Keywords: Recirculated systems; Maturation; Marine ornamental decapods

1. Introduction In the last 10 years a growing research effort has addressed the captive culture of marine ornamental decapods (Calado et al., 2003a,b). Considerable breakthroughs in the larviculture of these organisms were achieved after the development of suitable rearing

Corresponding author. Templo Aqutico, Rua de Santa Marta 27K, 1150-291 Lisboa, Portugal. Tel.: +351 91 629 1417; fax: +351 21 355 9030. E-mail address: [email protected] (R. Calado). 0044-8486/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.aquaculture.2006.10.013

R. Calado et al. / Aquaculture 263 (2007) 6874

69

systems (Calado et al., 2003c), allowing the first commercial scale larviculture protocols to be established (Calado et al., 2005a; Penha-Lopes et al., 2005). However, to assure a commercial scale production of any marine ornamental decapod species, a large number of larvae must be produced by broodstock pairs in a regular and feasible basis. Broodstock used in marine ornamental decapods research studies, and some commercial enterprises, is generally kept in pairs in isolated aquariums, where they are fed a variety of frozen and fresh diets to promote gonad maturation. Although it has already been successfully used (Zhang et al., 1997), there is no need to employ the damaging technique of eyestalk ablation (Lin and Zhang, 2001; Lin and Shi, 2002) to trigger ovarian maturation in marine ornamental decapods. The presence of late stage embryos in the abdomen of females (or euhermaphrodites, in the case of ornamental shrimp Lysmata (see Bauer and Holt, 1998)) is regularly monitored and, once detected, these specimens are captured, removed from the broodstock tank and placed in mesh baskets inside larval rearing tanks until larval hatching takes place. After larval release, females (or euhermaphrodites) are returned to their broodstock aquarium, in order to re-establish the reproductive pair (e.g. Palmtag and Holt, 2001). Apart from being time consuming, these practices present several disadvantages: stress induction to females (or euhermaphrodites) carrying late stage embryos due to their capture from broodstock tanks; disruption of the pair-bonding behaviour of established pairs of certain species that may trigger agonistic responses once the pair is reunited (namely in cleaner shrimp Lysmata (Rufino and Jones, 2001), boxer shrimp Stenopus (Johnson, 1969, 1977) and dwarf reef lobsters Enoplometopus (Cushing and Reese, 1998)); induction of osmotic stress to females (or euhermaphrodites) and late stage embryos when removing them from the broodstock tank and placing them inside the larval rearing tanks, negatively affecting their molting and hatching mechanisms, respectively; predation by females (or euhermaphrodites) on their newly hatched larvae, if kept in the same container without any physical barrier; and females (or euhermaphrodites) molting soon after larval release, while still in the larval rearing tank, missing the opportunity to successfully mate and produce the next embryo batch. In addition, the use of individual aquariums to keep breeding pairs, instead of a recirculated system equipped with suitable filtration, may result in poorer water quality, variation of water parameters between breeding pairs of the same species (that may affect embryo production and larval quality) and longer periods to perform routine tasks, such as water quality monitoring and water changes.

The present work describes an easy to operate recirculated maturation system for different types of marine ornamental decapods that: i) demands shorter periods of time to perform routine tasks, while allowing good water quality for broodstock keeping, ii) eliminates the need to capture ovigerous females (or euhermaphrodites) before larval release, minimizing the risks of disrupting reproductive pairs, iii) separates newly hatched larvae from the reproductive pair, impairing adults to prey on larvae and iv) allows live prey to be provided to larvae immediately after hatching if needed. 2. Materials and methods Twenty glass aquariums (0.60 m long 0.30 m wide 0.30 m high, total volume 54 L) were connected in parallel to a recirculation system composed of a 270 L sump equipped with a wet-dry filter with bioballs (assuring biological filtration), a protein skimmer and two 36 W ultra-violet sterilizer. Water was pumped from the sump with a pump rated at 5800 L h 1 and was distributed to the culture tanks through 25 mm polyvinyl chloride (PVC) pipes. Each maturation tank had two 10 mm diameter inlets positioned side by side in the top of the back glass of the aquarium, each equipped with a valve to regulate water inflow at 50 L h 1. Round Nitex mesh screens (90 mm diameter with 500 m or 150 m mesh size) were mounted in a PVC T connected to a 20 mm PVC standpipe fitted to the bottom of aquarium. Standpipes kept a 0.15 m water column in the tank, resulting in a water volume of approximately 27 L per aquarium. A movable plastic mesh (5 mm2 mesh size) was positioned inside each aquarium, isolating the third of the aquarium containing the two inflows from the two-thirds of the tank where the outflow and mesh screen were located. A movable opaque plastic division was placed perpendicularly to the movable plastic mesh, dividing the isolated third of the aquarium in two divisions (0.20 m 0.15 m 0.15 m), each with a water inflow (Fig. 1). Maturation tank water passed through the mesh screens and standpipes and was returned to the sump through a 50 mm diameter PVC pipe equipped with a 5 and a 20 m mesh bag to retain any uneaten food and other large particles. The tanks were illuminated from above with fluorescent white light, with an intensity of 1000 lx at the water surface and a photoperiod of 14 h light:10 h dark. Actinic lights (58 W) were enclosed in a 32 mm PVC pipe with 25 mm holes, so that each maturation tank had one of such holes facing its front glass, allowing the blue light to illuminate that area. Actinic lights were only switched on after the white fluorescent lights were switched off.

70

R. Calado et al. / Aquaculture 263 (2007) 6874

Fig. 1. Maturation tanks: A) upper view, B) lateral view. 1 divisions for broodstock pairs (0.20 m 0.15 m 0.15 m), 2 movable opaque plastic division (allowing the use of a larger single division (0.20 m 0.15 m 0.30 m) for larger and more aggressive species) 3 movable plastic mesh (5 mm2 mesh size, isolating the third of the aquarium containing the inflows and breeding pairs from the two-thirds of the tank with the outflow and mesh screen), 4 mesh screen (500 m or 150 m mesh size), 5 water outlet with PVC T connected to standpipe fitted to the bottom of aquarium, 6 actinic light illuminating front glass, 7 water inlet, 8 water column (0.15 m, resulting in a water volume of approximately 27 L per tank). Light grey arrows represent unidirectional water flow inside the maturation tank, while the black arrow represents water inflow.

Three breeding pairs of each of the following species were used to test the present maturation system: cleaner shrimp Lysmata amboinensis (De Man, 1888), fire shrimp L. debelius (Bruce, 1983), Monaco shrimp L. seticaudata (Risso, 1816), peppermint shrimp L. boggesii Rhyne and Lin, 2006, cleaning rock pool shrimp Urocaridella antonbruunii (Bruce, 1967), sexy shrimp Thor amboinensis (De Man, 1888), dancing shrimp Rhynchocinetes durbanensis Gordon, 1936, boxer shrimp Stenopus cyanoscelis Goy, 1984, S. hispidus Olivier 1811, hermit crabs Clibanarius tricolor (Gibbes, 1850), C. erythropus (Latreille, 1818), and the emerald crab Mithraculus sculptus (Lamarck, 1818). Broodstock was fed ad libitum 4 times per day with Marine Cuisine, a frozen food for marine aquarium organisms manufactured by San Francisco Bay Brand, supplemented with minced squid and shrimp and Cyclop-Eeze flakes. The presence of newly hatched larvae, adult molts and dead adults was daily checked in the morning, before routine tank siphoning was performed. Plastic mesh commonly used for fruit bags was provided to each division containing a breeding pair, providing shade, shelter and a surface to specimens

to cling upside-down (a favored position by many ornamental decapods). This material was preferred over live rock (tested in a previous trial) due to the accumulation of uneaten food under the pieces of rock, and consequently the need to remove it to properly siphon the breeding pair division, which always disturbed the broodstock. Additionally, undesired organisms were commonly introduced in the system when using live rock (namely amphipods, glass anemones Aiptasia sp., hydroids and fire worms Hermodice sp.), which were commonly seen preying on newly hatched larvae. Artificial seawater was prepared using freshwater purified by a reverse osmosis unit and mixed with the salt Crystal Sea produced by Marine Enterprises International, following the instructions of the manufacturer. Salinity was maintained at 35 1 and temperature was kept at 26 1 C through a heating/ cooling system. Ammonia and nitrite were monitored daily and maintained bellow detectable levels. Nitrate and pH showed average values ( standard deviation, S.D.) of 5 ( 2.5) mg L 1 and 8.1 ( 0.1), respectively. Every other day, a water change of 40 L was performed

R. Calado et al. / Aquaculture 263 (2007) 6874

71

using new artificial seawater, to assure good water quality. The suitability of division size to hold broodstock was ascertained through a comparison between the performance of three Lysmata debelius breeding pairs (since territory limitation may trigger agonistic behaviors even among members of a matted pair (personal observation)) kept in a single division (0.20 m 0.15 m 0.15 m) and three other pairs kept in the area of both divisions (0.20 0.30 0.15) (through the removal of the opaque plastic division placed perpendicularly to the movable plastic mesh). The same experimental design was repeated using breeding pairs of Stenopus hispidus. The possibility of supplying prey to larvae immediately after hatching was evaluated by placing newly hatched Artemia nauplii in maturation tanks containing paired L. amboinensis carrying embryos 1 day from hatching. These tanks were provided with 150 m mesh size screens in the standpipes, preventing larval prey to be drained. The gut appearance of larvae collected in the morning after from those tanks was compared under a stereo-microscope with those of larvae that were not provided any type of food after hatching. The efficiency of larval removal from the division containing the broodstock was quantified by placing batches of 100 newly hatched larvae in a division with (n = 5 batches) or without (n = 5 batches) the breeding pair and monitoring the number of larvae swimming near the actinic light illuminating the front glass of the aquarium after 30, 120 and 300 s. Since the morphology of decapod larvae may affect their swimming ability (Queiroga and Blanton, 2005), batches of larvae of distinct decapod groups were used: the caridean larvae of the shrimp L. amboinensis, the stenopodid larvae of S. hispidus, the anomuran larvae of Clibanarius erythropus and the brachyuran larvae of M. sculptus. Results on the efficiency of larval removal were compared by a multivariate analysis of variance (MANOVA) using the software Statistica (version 6.0), after checking the assumptions. Whenever significance was accepted at p b 0.05, the Tukey multiple comparison test was used (Zar, 1996). 3. Results All tested species were able to successfully mate and produce consecutive larval batches (Table 1). Routine tasks, such as checking for and collecting newly hatched larvae, monitoring molts of breeding pairs, recording the presence of specimens carrying embryos about to hatch, tank siphoning, filters cleaning and water changes, were daily performed each morning in about 1 h.

Table 1 Average carapace length (CL), cephalic shield length (CSL) or carapace width (CW) (mm) (standard deviation, S.D.) and average number of larvae ( S.D.) produced per batch by females or (euhermaphrodites in the case of Lysmata) of marine ornamental decapod crustaceans Species Carapace length, Average number cephalic shield of larvae produced length or carapace per batch (S.D.) width (mm) (S.D.) 1162 115 (n = 18) 1633 282 (n = 18) 558 153 (n = 18) 784 146 (n = 18) 123 23 (n = 6) 56 12 (n = 6) 322 24 (n = 6) 430 128 (n = 6) 633 140 (n = 6) 261 22 (n = 6) 203 31 (n = 6) 411 56 (n = 6)

CL = 10.2 0.2 (n = 6) Lysmata debelius1 CL = 12.1 0.2 (n = 6) Lysmata seticaudata CL = 7.3 0.2 (n = 6) Lysmata boggesii CL = 8.4 0.2 (n = 6) Urocaridella antonbruunii CL = 5.2 0.1 (n = 3) Thor amboinensis CL = 3.4 0.1 (n = 3) Rhynchocinetes durbanensis CL = 6.4 0.3 (n = 3) Stenopus cyanoscelis CL = 5.7 0.2 (n = 3) Stenopus hispidus1 CL = 6.5 0.1 (n = 3) Clibanarius tricolor CSL = 3.7 0.1 (n = 3) Clibanarius erythropus CSL = 3.1 0.1 (n = 3) Mithraculus sculptus CW = 15.2 0.2 (n = 3)

Lysmata amboinensis

Carapace Length (CL) measured from the postorbital eye socket to the posterior median edge of the cephalothorax; Cephalic Shield Length (CSL) measured from the rostrum anterior end to the posterior median edge of the cephalic shield; Carapace Width (CW) measured across the widest region of the carapace. 1 Breeding pairs were placed in double divisions (0.20 m 0.30 m 0.15 m) due to agonistic behavior (see Results).

The only dead specimens recorded during the breeding trials were among the pairs of L. debelius and S. hispidus kept in a single division (0.20 m 0.15 m 0.15 m). In all these trials the smallest specimen (the male, in the case of S. hispidus) was always killed by the larger mate (the female, in the case of S. hispidus) after the molt. No agonistic behavior or signs of aggression (e.g.: cut antenna or damaged pereiopods) were ever detected in the specimens placed in the double division (0.20 m 0.30 m 0.15 m). Newly hatched larvae L. amboinensis that had Artemia nauplii available in the maturation tank displayed their gut with an orange coloration when observed in the morning after hatching under a stereo-microscope,

72

R. Calado et al. / Aquaculture 263 (2007) 6874

Table 2 Average percentage of larvae (standard deviation, S.D.) of Lysmata amboinensis, Stenopus hispidus, Clibanarius erythropus and Mithraculus sculptus being removed from an empty division or a division containing a parental pair after 30, 120 and 300 s Species Empty division (n = 5 batches of 100 larvae per species) 30 s Lysmata amboinensis Stenopus hispidus1 Clibanarius erythropus Mithraculus sculptus
1

Parental pair present (n = 5 batches of 100 larvae per species) 300 s 99.4 0.5 c 98.8 1.1 c 99.0 0.7 c 98.6 0.9 c 30 s 22.6 1.9 a 23.7 1.2 a 23.1 1.9 a 23.8 1.6 a 120 s 49.6 2.1 b 50.8 1.3 b 49.5 2.7 b 50.2 1.6 b 300 s 98.0 1.6 c 97.0 1.0 c 97.4 1.1 c 97.2 2.2 c

120 s 53.0 3.9 b 51.8 2.6 b 48.6 2.7 b 51.8 2.3 b

24.6 1.1a 25.6 1.3 a 23.4 1.9 a 24.7 1.0 a

Different superscript letters represent significant differences ( p b 0.05). Test was performed in double divisions (0.20 m 0.30 m 0.15 m) due to agonistic behavior of parental pair (see Results).

revealing the capture and ingestion of larval prey. On the other side, larvae that were collected from maturation tanks without Artemia nauplii had empty guts, and some dead larvae were recorded in the bottom of the tank. The unidirectional water flow inside the maturation tank, as well as the actinic light placed in the front glass, allowed newly hatched larvae to be quickly removed from the chamber containing the reproductive pair, either when broodstock was kept in a single or double division (in the case of S. hispidus). Over time, a significant higher ( p = 0.001) number of larvae of any morphological type was removed from the division with or without the parental pair (Table 2). After 30 s between 22.6 1.9% and 25.6 1.3% of larvae had already been removed from the division, while after 300 s 97.0 1.0% to 99.4 0.5% of released larvae were collected near the actinic light. After the same period of time (30, 120 or 300 s), there were no significant differences ( p N 0.05) between the percentages of the four types of larvae (caridean, stenopodid, anomuran and brachyuran) being removed from the parental division. Although percentages of collected larvae were generally slightly lower in the trials where the parental pair was present (Table 2), no significant differences ( p N 0.05) were detected when comparing the percentage of larvae being removed, after the same period of time, from a division with or without the breeding pair. 4. Discussion Ovarian maturation will naturally take place if a mated pair of a marine ornamental decapod is kept under stable conditions, in good quality water and fed a varied diet several times, with embryos being commonly produced after 2 to 3 weeks. However, maturation can still be triggered if water quality is not adequate, but several problems can occur: a reduced

number of embryos are produced, molts still carrying embryos are discarded and larvae do not inflate with water after hatching, remain curled and commonly die (Calado et al., 2005b). In the present work, such warning signs of poor water quality were not recorded, with the filtration system of the recirculated maturation systems proving to be adequate to promote good water quality, enabling all stocked species to successfully undergo gonad maturation, mate and produce consecutive larval batches. The movable plastic mesh (5 mm square mesh size) positioned inside each aquarium, isolating the third of the aquarium containing the two inflows from the remaining two-thirds of the tank, was ideal to form breeding pairs of species with agonistic behavior towards conspecifics, since they could be physically isolated while still able to touch each other and start to develop the pair bond. Once agonistic responses became less frequent, the plastic mesh was removed to allow both specimens to be together and their behavior monitored for 60 to 90 min. When no agonistic response was recorded the specimens were paired in the division with the inflows. The size of the division tested to stock breeding pairs (0.20 m 0.15 m 0.15 m) proved to be suitable to most tested decapod species, with the exception of those displaying strong agonistic behavior towards conspecifics, including their mate (e.g. L. debelius and S. hispidus). The complex social behaviors involved in pair formation of such species (see Johnson, 1969; Zhang et al., 1998), the easiness of its disruption and their larger size (when compared to the other tested species, namely the larger S. hispidus versus the smaller S. cyanoscelis), can explain why a larger double division (0.20 m 0.30 m 0.15 m) was needed to properly stock the breeding pairs. Breeding pairs of large Stenopus species, namely S. spinosus Risso, 1827 and S. pyrsonotus Goy and Devaney, 1980, as well as dwarf reef lobsters Enoplometopus should

R. Calado et al. / Aquaculture 263 (2007) 6874

73

always be placed in a larger double division (0.20 m 0.30 m 0.15 m) to prevent aggression. Despite the existence of facultative primary lecithotrophy (see Thessalou-Legaki et al., 1999) among some ornamental decapod crustacean larvae, the importance of providing adequate food to larvae immediately after hatching has already been highlighted (Simes et al., 2002). This system allows larval prey to be provided to larvae immediately after hatching, avoiding the negative effects of early starvation (Calado et al., 2005a,b). The present maturation system is also highly effective for collecting and concentrating newly hatched larvae of ornamental decapods displaying different morphological features. In less than 5 min, most newly hatched larvae were successfully removed from the division containing the breeding pair. Since there were no significant differences between the number of collected larvae, in the same time frame, from divisions with or without the breeding pair, we can assume that this system minimizes the problem of adults cannibalizing their newly hatched larvae. Additionally, the selection of larvae displaying the strongest positive phototactic response is facilitated, since newly hatched larvae readily cluster near the area of the front glass of the aquarium illuminated by the actinic light. 5. Conclusion Although larviculture and juvenile grow out are commonly regarded as the bottlenecks that may impair the profitability of commercial scale culture of marine ornamental decapods, the feasible production of good quality larvae is also of vital importance. Any enterprise or research center addressing the culture of marine ornamental decapods will certainly benefit if a maturation system similar to the one described is implemented. Apart from providing similar conditions to all breeding pairs (ideal to test the suitability of maturation diets), important technical resources (e.g. larviculture tanks) can be better managed, since the described system allows breeding pairs to be easily monitored and larval hatching to be predicted (e.g. observation of embryo coloration). The possibility of providing larval prey to larvae immediately after hatching, as well as selecting the ones displaying strong positive phototactic behavior, assures that good quality larvae will be used during larviculture trials. With current research efforts and with suitable maturation and larviculture systems now available, profitable commercial scale culture of the most heavily collected marine ornamental decapods will soon be achieved.

Acknowledgements The authors would like to thank Fundao para a Cincia e a Tecnologia (scholarship SFRH/BPD/18009/ 2004) from the Portuguese government for their financial support. We also thank Templo Aqutico stuff, namely Srgio Dantas and Lus Simes for their technical support. References
Bauer, R.T., Holt, G.J., 1998. Simultaneous hermaphroditism in the marine shrimp Lysmata wurdemanni (Caridea: Hippolytiae): an undescribed sexual system in the decapod Crustacea. Mar. Biol. 132, 223235. Calado, R., Lin, J., Rhyne, A.L., Arajo, R., Narciso, L., 2003a. Marine ornamental decapods popular, pricey, and poorly studied. J. Crustac. Biol. 23, 963973. Calado, R., Narciso, L., Arajo, R., Lin, J., 2003b. Overview of Marine Ornamental Shrimp Aquaculture. In: Cato, J.C., Brown, C.L. (Eds.), Marine Ornamentals Collection Culture and Conservation. Iowa Press, USA, pp. 221230. Calado, R., Narciso, L., Morais, S., Rhyne, A.L., Lin, J., 2003c. A rearing system for the culture of ornamental decapod crustacean larvae. Aquaculture 218, 329339. Calado, R., Figueiredo, J., Rosa, R., Nunes, M.L., Narciso, L., 2005a. Larval culture of Monaco shrimp Lysmata seticaudata (Decapoda: Hippolytidae): effect of temperature, rearing density and larval diet. Aquaculture 245, 221237. Calado, R., Conceio, L.E.C., Dinis, M.T., 2005b. Early larval culture of marine ornamental decapods: getting it right from the start. Larvi 05 Fish and Shellfish Larviculture Symposium. Spec. Publ. Eur. Aquac. Soc. 36, 6770. Cushing, B.S., Reese, E., 1998. Hawk-like aggression in the Hawaiian red lobster, Enoplometopus occidentalis. Behaviour 135, 863877. Johnson Jr., V.R., 1969. Behaviour associated with pair formation in the banded shrimp Stenopus hispidus (Olivier). Pac. Sci. 23, 4050. Johnson Jr., V.R., 1977. Individual recognition in the banded shrimp Stenopus hispidus (Olivier). Anim. Behav. 24, 418428. Lin, J., Shi, P., 2002. Effects of broodstock diets on reproductive performance of the golden banded shrimp, Stenopus scutellatus. J. World Aquac. Soc. 33, 383386. Lin, J., Zhang, D., 2001. Effects of broodstock diet on reproductive performance of the peppermint shrimp Lysmata wurdemanni. J. Shellfish Res. 20, 361363. Palmtag, M.R., Holt, G.J., 2001. Captive rearing of fire shrimp (Lysmata debelius). Texas Sea Grant College Program Research Report. Penha-Lopes, G., Rhyne, A.L., Lin, J., Narciso, L., 2005. The larval rearing of the marine ornamental crab, Mithraculus forceps (A. Milne Edwards, 1875) (Decapoda: Brachyura: Majidae). Aquac. Res. 36, 13131321. Queiroga, H., Blanton, J., 2005. Interactions between behaviour and physical forcing in the control of horizontal transport of decapod crustacean larvae. Adv. Mar. Biol. 47, 107214. Rufino, M.M., Jones, D.A., 2001. Binary individual recognition in Lysmata debelius (Decapoda: Hippolytidae) under laboratory conditions. J. Crustac. Biol. 21, 388392. Simes, F., Ribeiro, F., Jones, D.A., 2002. Feeding early larval stages of fire shrimp Lysmata debelius (Caridea, Hippolytidae). Aquac. Int. 10, 349360.

74

R. Calado et al. / Aquaculture 263 (2007) 6874 Zhang, D., Lin, J., Creswell, R.L., 1997. Effects of eyestalk ablation on molt cycle and reproduction in the banded coral shrimp Stenopus hispidus (Oliver). J. Shellfish Res. 16, 363366. Zhang, D., Lin, J., Creswell, R.L., 1998. Mating behavior and spawning of the banded coral shrimp Stenopus hispidus. J. Crustac. Biol. 18, 511518.

Thessalou-Legaki, M., Peppa, A., Zacharaki, M., 1999. Facultative lecithotrophy during larval development of the burrowing shrimp Callianassa tyrrhena (Decapoda: Callianassidae). Mar. Biol. 133, 375398. Zar, J.H., 1996. Biostatistical Analysis. Prentice Hall, Upper Saddle River, New Jersey, USA. 481 pp.