Longfin yellowtail (Seriola rivoliana) larval rearing: skeletal development and effects of increasing dietary DHA levels at weaning phase

Longfin yellowtail (Seriola rivoliana) larval rearing. Skeletal development and effects of increasing dietary DHA levels at weaning phase. Seriola rivoliana Antonio Mesa Rodríguez Seriola rivoliana Seriola rivoliana "Longfin yellowtail (Seriola rivoliana) larval rearing. Skeletal development and effects of increasing dietary DHA levels at weaning phase" Antonio Mesa Rodríguez Doctorado en Acuicultura: Producción Controlada de Animales Acuáticos. Escuela de Doctorado Universidad de Las Palmas de Gran Canaria Tesis para optar al Grado de Doctor Las Palmas de Gran Canaria 2017 Directores: Ph. D. Francisco Javier Roo Filgueira Ph. D. Carmen Mª Hernández Cruz Por ti, por mí... Gracias Mamá Descomponiendo mi sinsentido... descomponiendo mi Matriz "Mi sinceridad es un regalo que NO todo el mundo se merece.... a partir de ahora procuraré NO ser tan Generoso" "Nadie nace con ello... pero se cultiva y crece desde dentro. Procura disfrutar de tu propia cosecha" "La Felicidad es un estado excepcional muy difícil de alcanzar debido a la inconformidad social NO natural del ser humano" " Y nos dejó.... dejando todo su amor en nosotros. Somos fruto de su influencia en vida.... y siendo tal como somos, estará siempre con nosotros... Toda una Vida" Antonio Mesa Rodríguez Summary According to the increasing DHA levels in inert diets, the inclusion of dietary DHA in feed up to a 3.17% (dw) improved larval resistance to air exposure. Also, DHA did not significantly affect fish final growth or final survival. On the contrary, high levels of DHA tend to increase the incidence of skeletal anomalies in S. rivoliana larvae, albeit no significant differences were observed. Furthermore, the occurrence of severe anomalies such as kyphosis and lordosis, was mainly associated to the larvae fed with the highest levels of dietary DHA. Additionally, an adequate ratio of the omega 3 (n3) long chain polyunsaturated fatty acids (LC-PUFA) DHA and eicosapentaenoic acid (20:5n-3; EPA) should be above 3.1. This ratio and the role of n-3 LC-PUFA in bone formation in S. rivoliana should also be studied in order to understand the requirements of these emerging fast growth species. Moreover, further studies of EFA requirements should be studied to enhanced S. rivoliana larval production. - II - Acknowledgements Several institutions supported the present Thesis. In these sense, I would like to thank the Viceconsejería de Pesca del Gobierno de Canarias through the programme “Proyecto piloto para el cultivo de especies de rápido crecimiento en Canarias (RAPCREC)"; the Agencia Canaria de Investigación, Innovación y Sociedad de la Información (ACIISI) and Fondo Europeo de Desarrollo Regional (FEDER) through the program “Mejora de las técnicas de cría de larvas de (Seriola rivoliana): Determinación de requerimientos de ácidos grasos esenciales en su etapa larvaria y optimización de la secuencia alimentaria (METCSERProID20100094)”. I would also like to thank the Universidad de Las Palmas de Gran Canaria (ULPGC) for financing my PhD. Many thanks to all IMARES group from Yerseke (Wageningen University & Research, Netherlands) for letting me the opportunity to work at their facilities with RAS systems, specially to Dr. Wout Abbink. My gratitude goes to Dr. Clara Boglione, for sharing all her "bone knowledge" without inconvenience at any time. At this point... I would like to thank in my mother language. Many Thanks... Me gustaría inaugurar estos agradecimientos con mis dos pilares científicos que compartieron sus conocimientos y confiaron en mí, mis directores de Tesis: Dr. Javier Roo, llegados a este punto no sé ni por dónde empezar, ya que son muchos años aguantando mis "cagadas" y mis nervios. Mis agradecimientos por tu completa dedicación siempre me sabrán a poco... y después de este último año largo y complicado, sólo puedo estar eternamente agradecido. No todos los directores te acogen en su casa para ultimar correcciones de artículos, pósters o tratamiento de datos. Gracias por guiarme. Gracias por tanto. Dra. Carmen Hernández-Cruz, no me cabe duda que soy de los pocos afortunados que ha tenido el placer de tenerte como directora, y es que más que una directora, desde el principio fuiste muy clara y sincera (valores que comparto 100%) y que entre tantos momentos vividos y consejos recibidos, has conseguido que te sienta como mi madre científica. NUNCA olvidaré todo lo que hiciste por mí durante mi ausencia. Además, muchas collejas, muchas ciruelas y mucho mucho cariño. Gracias por tanto. Gracias de Corazón. - III - Acknowledgements De forma generalizada, me gustaría agradecer a todos los miembros del GIA, ya que de una manera o de otra, son muchas las técnicas, consejos de manejo y situaciones que te hacen mejor técnico, adquiriendo muchos conceptos y mejorando profesionalmente. Más detalladamente, me gustaría agradecer a: Dr. Pipo, me "bautizaste" de la mejor manera posible durante mi primer congreso... mostrándome que siempre hay un momento para pasarlo bien y echarnos unas risas. Gracias por dedicarme en cualquier momento un minuto para debatir y compartir buenos consejos. Qué bueno se hace trabajar con alguien tan implicado, sincero y transparente... demostrando tu calidad humana. Sin duda, de la misma escuela que Carmen !! Mis más sincera admiración y respeto. Dr. Dani Montero, siempre disponible para todo el grupo, siempre resolutivo ante cualquier situación. Gracias por tu apoyo y cercanía, y no solo en los momentos difíciles. Dr. Juan Socorro, tu calidez y presencia destaca por donde quiera que vas... priorizando siempre a la persona y demostrando lo natural y gran persona que eres. Buen rollo. Gracias por tus palabras y tu apoyo. Dra. Gercende, siempre afortunado de poder compartir buenos ratos y risas. Siempre sonriente. Gracias. Dr. Rubén, que sólo con tu experiencia, el grado de Dr. lo tienes más que ganado !!. Gran profesional, minucioso (a veces algo perreta), amigo y mejor persona aún. Gracias por aguantar tantas jodiendas y batallas mías. Muchas horas compartidas en el "invernadero". Si tuviera que agradecerte en cervecitas todo lo que me has ayudado, bufff.... sas ke no !!! Dra. Marisol, muchas gracias por la oportunidad de formar parte de este grupo. He tenido la suerte de aprender muchísimo. También me gustaría agradecer ESPECIALMENTE de todo corazón a Sara, Nacho y Anais, que intentaron hacerse cargo de mi último experimento durante mi ausencia. Eso NUNCA lo olvidaré !!! Agradecer también a: Ada, la GRANDÍSIMA Lore, Dominique, JuanMa, Rafa, Mapi, Carmen, Yurena, Silvia, Tibi, Andrea, Fran Robaina, Fran Otero, Samira, David ("bones"), Reda, Nabil, Rayco, Javi y el inigualable y siempre disponible Chano. Gracias a Sergio por tantas soluciones en las instalaciones. Y por último, gracias a Tere y María por hacer del centro un lugar más limpio. - IV - Acknowledgements Además, me gustaría destacar que a lo largo de todos estos años, son muchas cosas las que he aprendido, indudablemente a nivel científico, pero lo que es más importante (por lo menos para mí).... a nivel personal. He visto y sufrido situaciones las cuales me han hecho madurar como persona, y me han ayudado a determinar qué es lo que quiero en mi vida, y especialmente qué tipo de situaciones y personas NO quiero en ella. Por todo ello, Gracias. Me gustaría poder agradecer a todos mis amigos que siempre me mostraron su apoyo, más aún en el momento más difícil de mi vida: Doc. Mauricio (Morenitooohh), Antonio La Barbera, Bruno, mi Desi (tqm), María, Faty, Guacimerss, Alfonsillo, Antonio, Iñaki, Jesús (Bajo), JuamPe, Áureo, Juanillo, J. Daniel Camacho (Gracias Compañeroooo !!) Víctor (Wassa), JuanJo, Cap. Jason R. Pipe ("echa caballa"), Manolys, Mikel, Ainhoa (Nocillits), David (Fishing), Agustín Cabrera (Maestro), Manolo Ramírez, J. Luis Beistegui, Cris (Gracias por tu ayuda espiritual), Berta, Cristo y Álvaro. ESPECIALMENTE a: Monique (Cómo agradecerte tanto que me has ayudado ?? Son muchos años de amistad y compartiendo horas de trabajo, de darte el coñazo... pero tu sobreesfuerzo en los momentos más difíciles hacen que esté eternamente agradecido, siempre estaré para lo que necesites); Darius y Naxits (mis SoulSurfers de Siempre, hermanos de agua... siempre han estado ahí, en las buenas y en las malas. Les quiero mucho. Gracias por tanto !!); Adri, Tomicius y Ancosito (qué fácil es pasar buenos ratos juntos, cualquier excusa es buena para sumar razones de vida. Gracias!!); Juanito (My friend, me has dado la oportunidad de hacer grandes debates científicos, hacerme crecer mucho como investigador, la oportunidad de ser Codirector de Tesis de Máster, y lo que es más importante aún, darme la oportunidad de forjar una grandísima amistad, que es de las mejores cosas que me llevo de todos estos años. Gracias por tanto apoyo en los momentos más difíciles. Tqm.), Noe (Gracias mi pequeñina, son tantas las vivencias que sólo tú sabes lo importante que eres para mí. Tqm.), Dava (Hermano, gracias por haberte cruzado en mi camino, son muchas las historias vividas y otras tantas que nos quedan. Siempre sintiendo tu gran apoyo. Tqm.) y Rut ("Serendipia", gracias por tanto, gracias de corazón). Carmen, a pesar de las vueltas que da la vida, has sido una parte importante de mi vida, y por ello sólo tengo sinceros y sentidos agradecimientos por haber aguantado tantas situaciones estresantes durante todos estos años. Gracias. Y para ir terminando, a todos mis familiares, especialmente a: Abuelo y Abuela (gracias por tener siempre un tupper con potajes), Ana prima, Ana (tití), Cruci (tití), Félix, Nieves, primo Félix, Marco, Miguel, Luisillo, J. Manuel y Gregorio. - V - Acknowledgements Por mucho que quiera expresar con palabras todo lo que han sido y son para mí, nunca será suficiente. Siempre unidos, siempre una piña. Gracias Dani, gracias por ser tú y tirar del carro cuando fue necesario. Por estar siempre ahí, por luchar juntos. Pero nunca olvides lo esencial en esta vida... tqm. Pi, la bella mariposa de mamá... mi bella mariposa también. Gracias por tanto. Gracias por tus diseños, por plasmar un sentimiento en mi conexión con el mar... tqm. Papá, gracias por cuidarnos en los momentos más difíciles. Por demostrarme y enseñarme que los valores inculcados nos mantendrán unidos. Por darnos todo de tí cuando la vida así lo decidió. Gracias a tu esfuerzo hemos conseguido terminar la Tesis. Este logro es de todos. Tqm. Y es ahora cuando se me nubla la vista, cuando me tiembla el alma más que nunca. Y es que irónicamente, el mayor apoyo, la mayor comprensión, a quien le debo la vida y todo mi ser, quién me ayudó a ser la persona que soy, quien me daba fuerzas, quien dedicó mucho esfuerzo para la consecución de esta Tesis.... sencillamente y desgraciadamente no está para verlo. La persona más importante en mi vida. Mi más emotivo e infinito agradecimiento. Siempre conmigo. Te quiero Mamá Y para todo el que lo lea, procura ser Feliz... es lo único que nos llevamos. ....Gracias - VI - List of Figures Figure 1. S. quinqueradiata (Temminck & Schlegel, 1845) Japanese production (FAO, 2014b).............................................................................................................................................5 Figure 2. World production of farmed greater amberjack (S. dumerili). FAO, 2014. FishStat....6 Figure 3. World production of farmed longfin Yellowtail (S. rivoliana). FAO, 2014. FishStat...........................................................................................................................................7 Figure 4. Larval rearing classification Divanach (1985)...............................................................8 Figure 5 (a, b). Intensive larval rearing tanks mainly used at ULPGC-ECOAQUA facilities for research (a:0.2m3 ;b: 2m3)..............................................................................................................8 Figure 6. Ancient Hawaii fishpond "Menehune", on the island of Kauai (Hawaii)......................9 Figure 7 (a, b). Mesocosms tanks (40m3) in the ULPGC-ECOAQUA facilities in Telde (Canary Islands, Spain)................................................................................................................................9 Figure 8 (a, b). S. rivoliana indoor RAS culture; Recirculating system used for larval rearing of Mediterranean species (Blancheton, 2000)..................................................................................10 Figure 9 (a, b). Cannibalism in S. rivoliana larvae culture.........................................................12 Figure 10 (a, b). Grading S. rivoliana larvae; larvae grader.......................................................13 Figure 11 (a, b). Faeces packets of undigested Artemia in 30 dah S. rivoliana larvae...............19 Figure 12 (a, b). P. dentex larvae with kyphosis (mid) and scoliosis (lower); S. rivoliana larvae with LSK syndrome.....................................................................................................................22 Figure 13 (a, b). Crossbite in S. rivoliana larvae; operculum reduction in P. dentex larvae............................................................................................................................................23 Figure 14. S. rivoliana distribution (Fishbase)............................................................................37 Figure 15. Adult of S. rivoliana...................................................................................................38 Figure 16 (a, b). S rivoliana adult; S. rivoliana gonadal biopsy inserting catheter (Kruuse, Langeskov, Denmark) into the gonadal cavity.............................................................................39 Figure 17 (a, b). S. rivoliana eggs; S. rivoliana larvae of 1dph.................................................40 Figure 18 (a, b). Semi-intensive 40m3 tank; Intensive 2m3 tank................................................41 - VII - List of Figures Figure 19 (a, b). Intensive 200l tank...........................................................................................41 Figure 20 (a, b). Artemia cyst; Artemia cysts separator (SEP-Art technology, INVETM)..........42 Figure 21 (a, b). Compressing and compressed paste; grounded pellets in sizes.......................43 Figure 22 (a, b). Defatted squid meal extraction with vacuum pump.........................................43 Figure 23. Meristic length measurements of S rivoliana larvae..................................................45 Figure 24. Regions of stained S. rivoliana larvae.......................................................................52 - VIII - List of Tables Table I. Domestication aquaculture levels (Telechea & Fotaine, 2014; modified from Bilio, 2007) for Seriola sp........................................................................................................................4 Table II. Seriola dumerili seed (units) production in Spain..........................................................6 Table III. Seriola sp. rearing conditions......................................................................................15 Table IV. DHA requirements for Seriola sp. larvae....................................................................28 Table V. S. rivoliana feeding sequence.......................................................................................44 Table VI. Staining protocol according to Vandewalle et al. (1998)............................................46 Table VII. Pectoral and pelvic fins skeletal abbreviations (Cubbage & Mabee, 1996)..............47 Table VIII. Skeletal elements abbreviations for vertebral column, caudal, dorsal and anal fins (Manod, 1968; Matsuoka, 1985; Suda, 1996)..............................................................................48 Table IX. Viscerocranial structures abbreviations grouped into regions according to their functionality (Matsuoka, 1985; Collette & Gillis, 1992; Cubbage & Mabee, 1996; Suda, 1996; Faustino & Power, 2001). ............................................................................................................49 Table X. Meristic counts for vertebral column, caudal, dorsal and anal fins structures..............50 Table XI. Meristic counts of the Viscerocranial skeleton. Single (S) or paired (P) structures.......................................................................................................................................51 Table XII. Skeletal abnormalities classification (Boglione et al., 2001)....................................53 - IX - List of Abbreviations A Artificial AA Amino Acids ABARE Australian Base Aquaculture AR Alizarin Red ARA Araquidonic Acid Art Artemia BW Body Weight CH Cephalic Height D Dark dah Days after hatch DHA Docosahexaenoic Acid dw Dry weight E Absolute Ethanol ED Eye Diameter EFA Essential Fatty Acids EPA Eicosapentaenoic Acid FA Fatty Acids FAMEs Fatty Acid Methyl Esters FAO Fisheries and Aquaculture Organization Fig. Figure GIA Grupo de Investigación en Acuicultura Gly Glycerine GnRHa Gonadotropin Releasing Hormone analogue HUFA High Unsaturated Fatty Acid - XI - bottlenecks for increased a stable production are related to and steady supply of juveniles in halibut and slow reproductive maturation in captivity of wreckfish (Mylonas & Robles, 2014). The success of diversification is directly related to domestication level, which is, closing its life cycle in captivity and modifying its physiological and production characteristics to suit human needs (Telechea & Fotaine, 2014). These authors stated five domestication levels, in which levels 1,2 and 3 are a transitory production that depend on the wild source availability; whereas levels 4 and 5 are those species in which life cycle is closed in captivity without wild inputs and selective breeding programs (growth rate, flesh quality, etc). Well known marine finfish species such as gilthead sea bream (Sparus aurata), European seabass (Dicentrarchus labrax) or meagre (Argyrosomus regius) (domestication levels 5, 5 and 4, respectively) are well domesticated in captivity, but the high market demand requires a variety of species like the wild stocks offers, and fish that grow faster will achieve market size more quickly. In this sense, fast growth species become the main objective for aquaculture diversification. Fast growth species such as Seriola sp. became a strong candidate for aquaculture diversification, due to its high market value and quality fillet (Mazzola et al., 2000; Nakada, 2002). These carangids are found in the Atlantic, Indian and Pacific oceans developing in tropical and subtropical waters. A few species have a worldwide distribution (such as greater amberjack, S. dumerili; longfin yellowtail, S. rivoliana; and yellowtail kingfish, S. lalandi) and others limited to regional distribution. Currently, 9 species of Seriola (Fishbase) have been described (Guinean amberjack, S. carpenteri; S. dumerili; lesser amberjack, S. fasciata; samson fish, S. hippos; S. lalandi; fortune jack, S. peruana; Japanese amberjack, S. quinqueradiata; S. rivoliana and banded rudderfish S. zonata), and very few are considered partially domesticated (Table I). Table I. Domestication aquaculture levels (Telechea & Fotaine, 2014; modified from Bilio, 2007) for Seriola sp. Scientific names Common names Domestication level References S. quinqueradiata Yellowtail 2 Bilio, 2007. S. lalandi Yellowtail kingfish 2 - 3 Kolkovski & Sakakura,2004. S. dumerili Greater amberjack 4 Hong & Zhang, 2003. S .rivoliana Longfin yellowtail 2 - 3 Roo et al., 2012. S. carpenteri Guinean amberjack 0 - S. fasciata Lesser amberjack 0 - S. hippos Samson fish 0 - S. peruana Fortune jack 0 - S. zonata Banded rudderfish 0 - - 4 - 1.1.1. Seriola Aquaculture Production Considering its relative long farming history, the yellowtail or Japanese amberjack, S. quinqueradiata (Temminck & Schlegel, 1845) is the mayor relevant carangid species. The commercial farming of yellowtail started in Japan by the 1940's and expanded rapidly in the 1960's. Nowadays, the production peak is ranged between 140.000 and 160.000 tonnes (t) (Fig. 1), due to the consideration of fish farmer to maintain a significant level of wild juveniles in order to preserve their seasonally seed's supply captured from the natural environment (Mojako). Due to the markedly influence of Japanese gastronomy, Seriola culture (S. quinqueradiata, S. dumerili and S. lalandi) accounts with the 57% of all Japanese marine fish aquaculture (Ohara et al., 2005). Figure 1. S. quinqueradiata (Temminck & Schlegel, 1845) Japanese production (FAO, 2014b). With an estimated annual production of 13.200 t in China (FAO, 2014a), 4558 t in Japan (Benedetto Sicuro & Umberto Luzzana, 2016) and 3000-4000 t in Australia (ABARE, 2009), Yellowtail kingfish (S. lalandi Valenciennes, 1833) is the second highest production of the Seriola's family. Other countries such as New Zealand (Kolkovski & Sakakura, 2004; Symonds et al., 2014), USA (Stuart & Drawbridge, 2011; Buentello et al., 2015), Chile (Orellana et al., 2014), Netherlands (Abbink et al., 2011; Garcia et al., 2015), México (BAJASEAS, 2016), South Africa and Namibia (O'Neill et al., 2015) are promoting the Yellowtail kingfish culture, but still limited productions. As well as the yellowtail, the greater amberjack (S. dumerili Risso, 1810) farming begun late in the 1970's in Japan and continued in the Mediterranean Sea by the 1980's with the on-growing of wild caught juveniles (Fig.2). - 5 - Major Mediterranean greater amberjack producers have been Spain (Grau et al., 1999), Italy (Lazzari et al., 2000), Malta (FAO, 2016b), Croatia (Benovic, 1980) and Turkey (Yilmaz & Şereflişan, 2011), as well as China (Rongxing et al., 2008), Vietnam (Ottolengui et al., 2004) and Taiwan (Lu et al., 2012). However, most of this farming activity stopped due to the lack of seed supply and pathological disease. Nowadays, the main producer is Japan, with 46.000 t of wild on-growed juveniles (FAO, 2014c). Figure 2. World production of farmed greater amberjack (S. dumerili). FAO, 2014c. FishStat. Nowadays, pilot S. dumerili culture programs are being developed according to national (PNA-SERIOLA) and international (DIVERSIFY) aquaculture programs. In Spain, the production of seasonally fingerlings of greater amberjack is still uncertain (Table II). Further studies on larval rearing and nutritional requirements need to be carried out to obtain fries of good quality. Table II. Seriola dumerili seed (units) production in Spain. 2012 2013 2014 2015 Source Information Andalucía 20.000 13.000 50.000 210.000 JACUMAR (2015)*1 Canary Islands 5.000 10.000 13.000 25.000 GIA Facilities (2016)* 2 * 1 JACUMAR (2015).Ministerio de Agricultura, Alimentación y Medio Ambiente. Spain. *2Grupo de Investigación en Acuicultura (GIA), IU-ECOAQUA, Universidad de Las Palmas de Gran Canaria (Unpublished data). - 6 - Longfin yellowtail (S. rivoliana, Valenciennes 1833) has a brief farming history, finding the first larvae rearing records at Ecuador (Blacio et al., 2003; Blacio, 2004) and Hawaii (Laidley et al., 2004). In Hawaii, longfin yellowtail has been commercially cultured since 2005 (Sims & Key, 2011) with a total of 400-500 t/year FAO, 2014a (Fig.3). Moreover, in the Canary Islands (Spain), pilot production scale of S. rivoliana was developed according to regional funding for aquaculture diversification (Roo et al., 2011; 2012). Figure 3. World production of farmed longfin Yellowtail (S. rivoliana). FAO, 2014a. FishStat 1.2. Marine Fish Larvae Production In natural environment, larval fish mortality varies from 20% per day after hatching to >99% over the entire larval period (Bailey & Duffy-Anderson, 2001). On the contrary, even that predation risk and starvation conditions are absent, major mortality occurs at early developmental stages (Blaxter, 1986; Houde, 1989; Papandroulakis et al., 2005; Roo et al., 2010a, 2012). The obtention of a constant amount of larvae in emerging new species is still complicated. Several problems have been encountered in juvenile's production on S. lalandi (Benetti et al., 2005), S. rivoliana (Roo et al., 2012), S. dumerili (Papandroulakis et al., 2005). Whitmore, the dependence on the availability of seeds in S. quinqueradiata is one of the critical factors for the commercial success of Seriola production in Japan (Masumoto, 2002; Nakada, 2002). Moreover, the state of Seriola sp. seeds production is still at the initial phases, with the incessant needs of larvae and juvenile culture and nutritional requirements research. Divanach (1985), established first a larval rearing classification based on larval density and tank volume. Thus, three main categories could be established: intensive, extensive and mesocosms systems (Fig. 4). Moreover, these systems can vary depending on the localization and the type of facilities, being opened, semi-opened or closed systems. In this sense, systems - 7 - such as extensive are majorly closed and exposed to environmental conditions, whereas other systems such as intensive or RAS can be open, semi-opened or closed, with high recirculation rates and highly controlled water conditions. Figure 4. Larval rearing classification Divanach (1985). Intensive Systems: the complexity that implies the intensive systems denote in consequences such as cannibalism and aggressive behaviour (Hecht et al., 1996; Baras & Jobling, 2002), low survival and growth under stressful conditions (Hernández-Cruz et al., 1999; Roo et al., 2007, 2010a, b) as well as anomalies occurrence (Andrades et al., 1996; Roo et al., 2005a, b). In this way, the intensive systems (Fig.5a,b) techniques require sophisticated facilities and elevated investments, being less suitable for the culture of new species (Divanach et al., 2002). Figure 5 (a, b). Intensive larval rearing tanks mainly used at ULPGCECOAQUA facilities for research (a:0.2m3 ;b: 2m3). a b - 8 - Extensive Systems: the success of extensive systems (Fig.6) is usually higher than intensive systems, showing larvae with natural biological patterns (Pitta et al., 1998) and excellent fry quality without problems such as abnormal swim bladder inflation, skeletal deformities, pigmentation anomalies or deviations in natural behaviour (Divanach et al., 1996). Figure 6. Ancient Hawaii fishpond "Menehune", on the island of Kauai (Hawaii). Mesocosms Systems: the results obtained with mesocosms system (Fig. 7a,b) are commonly better than intensive or extensive techniques (Papandroulakis et al., 2004), with weaning totally achieved at one month, a generally fry survival ranging between 40-90% after weaning , deformities incidence around 5-10%, swim bladder inflation up to 95%, low size dispersion and low cannibalism incidence (Divanach & Kentouri, 2000). This culture system has been used favourably for more than 25 marine fish species and 5 hybrids for fry production (Divanach & Kentouri, 2000). Figure 7 (a, b). Mesocosms tanks (40m3) in the ULPGC-ECOAQUA facilities in Telde (Canary Islands, Spain). a b - 9 - Recirculating Aquaculture Systems (RAS) The recirculation aquaculture systems (RAS) are systems in which water is (partially) re-used after undergoing treatment (Rosenthal et al., 1986). Each treatment step reduces the system water exchange needs. This system has been developed to satisfy the increasing environmental regulations in countries with limited access to land and water (Fig. 8a,b). Figure 8 (a, b). a) S. rivoliana indoor RAS culture; b) Recirculating system used for larval rearing of Mediterranean species (Blancheton, 2000). Compared to common aquaculture system flows, several positive effects are defined. In RAS, not only the reduction of water consumptions (Verdegem et al., 2006) and consequently the reduction of aquaculture waste water impact, but also the conservation of heat (avoiding seasonal changes), the reduction of disease due to its environmental control and the optimization of growth rates and fish health by the surveillance of water quality (Blancheton et al., 2009; Martins et al., 2010). Another important aspect of RAS is the establishment of microbial control in the cultivation tanks by stabilizing the substrate to an adequate bacteria ratio. Well-balanced concentrations of substrate induce growth of a stable, slow growing and more beneficent bacterial community (Vadstein et al., 1993; Skjermo et al., 1997; Salvesen et al., 1999). In RAS conventional water treatments, solid particles larger than 40-60μm are removed to prevent the mineralization and production of smaller components (Chiam & Sarbatly, 2011). More sophisticated techniques in RAS systems utilize a biofilm membrane bioreactor (BF-MBR). This membrane enhances the reduction of fine suspended solids, colloidal particles and nutrient a b - 10 - from the system, with the significantly reduction of the turbidity, number of colloidal particles and bacterial concentration, as well as more stability in the water system (Holan et al., 2014). Since the introduction of RAS, the production and diversity of species has significantly increased (Martins et al., 2005). In Europe, the most established RAS technology is found in The Netherlands and Denmark with indoors production of African catfish (Clarias gariepinus) and eel (Anguilla anguilla), and semi-closed and out-doors production of trout, respectively. Actually, more than 10 species are produced in RAS, being seabass and sole as the major marine species, and African catfish, trout and eel as major freshwater species cultured in Europe (Martins et al., 2010). Fast growth species, such as S. lalandi, have been cultured in RAS on a research scale (Partridge et al., 2003). Due to its rapid growth rate to market size (3kg in 1 year), market value and life cycle closed in captivity, yellowtail kingfish was considered a species with great potential for RAS culture (Abbink et al., 2012). Recently, limited amounts of yellowtail kingfish are commercially produced with RAS in the Netherlands (Garcia et al., 2015), Chile (Orellana et al., 2014) and Mexico (BAJASEAS, 2016). 1.2.1 State of culture techniques for Seriola. The establishment of adequate rearing conditions for new aquaculture species, mainly stated by the high market price and commercial demand, became complicated. Seriola culture, as an emerging fast growth genus, requires culture techniques that differ from the well-known commercial intensive techniques used for other marine finfish such as Sparus aurata or Discentrarchus labrax. Thus, the selection of an appropriated rearing technology depends on the species specific characteristics and its susceptibility to different parameters (larval density, tank volume, turbidity, temperature, photoperiod, light quality and intensity, live prey regimes, etc.) directly affecting the larval rearing success. Thereby, mesocosms systems seems to be an adequate technique for the initial phases of new candidate species, due to its similar conditions to the natural environment with large water volumes and low larval density (Papandroulakis et al., 2005; Roo, 2009a; Roo et al., 2012). 1.2.1.1. Larval rearing of Seriola. Two critical periods have been identified in the culture of Seriola, with a high mortality from hatching to the first feeding period and the second being related to aggression behaviour and cannibalism. Thus, a considerable difference at larval hatching time of fast growth species, may produce disparity at first exogenous feeding and consequently, size heterogeneity, encouraging aggressive behaviour and cannibalism at later larval stages. The onset of cannibalism behaviour and its development in Seriola (Fig. 9a,b) was first described by - 11 - Sakakura & Tsukamoto (1996). These authors stated the onset of cannibalism after metamorphosis and it occurs until the apparition of schooling behaviour, outcome by distance to the nearest neighbour. The "J-posture" is determined as a precursor to aggressive behaviour (Sakakura & Tsukamoto, 1996) and is assumed to be some kind of intimidation behaviour after notochord flexion. Larvae that show this conduct frequently become dominants, existing a positive correlation between dominant larvae, "J-posture" and aggressive behaviour (Sakakura & Tsukamoto, 1999). Figure 9 (a, b). Cannibalism in S. rivoliana larvae culture. A high level of uncontrolled aggressive behaviour has been associated with the appearance of mass mortality in S. quinqueradiata (Sakakura & Tsukamoto, 1996; 1999; Sakakura et al., 1998), S. lalandi (Ebisu & Tachihara, 1993; Yamazaki et al., 2002; Moran, 2007; Stuart & Drawbridge, 2013), S. mazatlana (Benetti, 1997), S. dumerili (Shiozawa et al., 2003; Papandroulakis et al., 2005; Miki et al., 2011; Mesa-Rodriguez, Unpublished data) and S. rivoliana (Blacio, 2004; Mesa-Rodriguez, Unpublished data) larvae, mainly the smallest specimens. These dead specimens are mostly subordinated larvae, which are continuously subjected to stressful situations being chased by the larger (total length) dominants larvae (Sakakura et al., 1998), causing food depravation or physical trauma (Moran, 2007). Other factors, such as live prey regimes could indirectly affect the onset of cannibalism. Moran (2007) stated that the introduction of Artemia sp. as a food source was correlated with the increase of size heterogeneity and aiming behaviour in S. lalandi larvae, being a precursor of aggressive behaviour. Moreover, restricted live prey feeds amplified the size heterogeneity, thereby significantly exacerbating aggressive behaviour in S. dumerili larvae (Miki et al., 2011). In the natural environment, the floating seaweed prevents the aggressive behaviour and works as a shelter from other predators. The utilization of shelter or floating objects to promote the schooling behaviour and avoid the cannibalism in culture tanks has been used in S. quinqueradiata (Sakakura & Tsukamoto, 1996; 1999) and S. rivoliana (Blacio, 2004; MesaRodriguez, Unpublished data) culture. a b - 12 - In order to prevent size heterogeneity, with the consequently social rank and aggressive behaviour, the grading of fish larvae becomes an alternative (Fig. 10a,b). This method has previously reported to significantly improve the survival and production of S. lalandi (Ebisu & Tachihara, 1993; Moran, 2007), S. quinqueradiata (Yamazaki et al., 2002), S. dumerili (Shiozawa et al., 2003; Miki et al., 2011; Mesa-Rodriguez, Unpublished data) and S. rivoliana (Mesa-Rodriguez, Unpublished data) larvae. The early grading of fish larvae may reduce the size heterogeneity consequences. Figure 10 (a, b). a) Grading S. rivoliana larvae; b) larvae grader. Moreover, larval quality is directly related to biotic and abiotic rearing factors. 1.2.1.2. Larval density and tank volume. The larval stocking density as well as rearing systems (intensive and semi-intensive systems) of Seriola and emerging fast growth species is directly related to larval growth and survival. High larval densities employed in intensive and semi-intensive (up to 50 egg/l) systems may decrease the theoretical prey availability and increase larval interaction and competition for prey. Furthermore, the incoming size heterogeneity, together with the visual acuity development and increment of larval interaction may incite aggressive behaviour and cannibalism. The comparison between different larval densities has been previously studied. Roo and collaborators (2012), compared semi-intensive/mesocosms system (SIS-M: 4.5 egg/l, 40m3) and intensive system (IS: 125 egg/l, 2m3) in S. rivoliana larval rearing, showing better average survival under SIS. Moreover, this better survival could also be associated with the use of larger water volume tanks. In this sense, Stuart & Drawbridge (2013) stated that S. lalandi larvae should be moved into a large diameter (3.6 m), shallow (0.9 m), flat-bottom 6000 l fibreglass tank after 10 days after hatch (dah) reared in incubators tanks due to their very strong a b - 13 - 1.3. Importance of Skeletal Development One of the bottlenecks of finfish aquaculture production is the presence of morphological deformities. In this sense, the knowledge and the study of skeletal developmental pattern of reared species represent an important contribution for the optimization of larval rearing. The occurrence of skeletal deformities directly depends on the rearing methodology applied, obtaining great differences between intensive and semi-intensive conditions which also affect the external shape of the fish (Boglione et al., 2001). These authors observed that in S. aurata culture, a range between 15-50% of the juvenile's present different deformities which result in important economic losses due to mortality, reduction in growth and market rejection by the final consumer. The occurrence of abnormalities in cultured marine finfish has been widely reported (European sea bass, Dicentrarchus labrax: Koumoundouros et al., 2002; barramundi, Lates calcarifer: Fraser et al., 2004;Atlantic halibut, Hippoglossus hippoglossus: Hamre et al., 2005; cod, Gadus morhua: Grotmol et al., 2005; red porgy, Pagrus pagrus: Roo et al., 2005b; Diplodus sargus, Saavedra et al., 2009; Atlantic salmon, Salmo salar: Witten et al., 2009). 1.3.1. Generalities of skeleton development. The finfish skeletal system has multiple functions. Bones and cartilages provide support for the body, structural integrity, protection for major organs, provide attachment for ligaments and muscles, and serve as a mineral reservoir (Hall, 2005). Four classes of mineralized tissues (and related cells) can be identified: bone (osteoblast, osteocytes and osteoclast), cartilage (chondroblast, condrocytes and chondroclasts), dentine (odontoblast, odontocytes and odontoclasts) and enamel (ameloblats) (Hall & Witten, 2007). Moreover, Teleost fish display a large range of intermediate skeletal tissues. - Bone is a specialized mesenchymal vascularized tissue formed by cells (osteoblasts, osteocytes and osteoclast), a mineral phase (mainly composed of calcium phosphate forming hydroxyapatite crystals) and an organic mineralized extracellular matrix. The major organic component of bone is collagen type I, and the degree of matrix mineralization is variable and seems to depend on type of bone (cellular o acellular bone), life style or fish habits (Meunier & Huysseune, 1992; Danos & Staab, 2010; Sfakianakis et al., 2011; Dean & Shahar, 2012). - Cartilage is mainly composed of chondrocytes, proteoglycans and collagen type II forming an extracellular matrix (Witten et al., 2010). - 20 - In fish, according to the species and skeletal elements, there are three bone formation mechanisms: endochondral, perichondral and intramembranous ossification. - Endochondral ossification: most of bones that ossify endochondrally originate from embryonic mesoderm. This involves a cartilaginous template, which is replaced or remodelled into bone (Hall, 2005). - Perichondral ossification: is the most common in fish. Larval fish essentially only have perichondral bone formation. Perichondral ossification produces the bone that surrounds the cartilage (Benjamin, 1989). Witten & Huysseune (2007) reviewed that, perichondral bone formation is a basic process of ossification of the endoskeleton fin. - Intramembranous ossification: this type of bone development has been described in many teleosts. Mesenchymal cells differentiate into osteoblasts and form bone without a cartilaginous template (Franz-Odendaal et al., 2006). Bones formed in this way are defined as dermal or membrane bones. Taking into account the different mineralized tissues and the bone formation mechanisms, skeleton can be divided into two systems: i) dermal skeleton, that usually develops intramembranously, being mesenchymal precursor cells developing directly into bone (teeth, scales, fin rays, dermal skull bones, etc.); ii) endoskeletal, these are bones developed by endochondral ossification, where cartilaginous elements are replaced or remodelled by bone (branchial arches, axial and appendicular skeleton, etc.). 1.3.2. Finfish common skeletal deformities. Skeletal anomalies in reared fish can affect all type of skeletal tissues, but alterations of the notochord, cartilage and bone are the most relevant for the aquaculture industry. The skeletal anomalies can appear throughout the culture as long as the skeletal elements are in incessant bone remodelling under the influence of rearing and/or nutritional factors. In this sense, skeletal anomalies can be ranged from slight internal anomalies (not affecting morphological shape) to severe internal abnormalities that directly affect the external body shape. Thus, externally detectable deformations such as lordosis, kyphosis, scoliosis (vertebral column abnormalities), shortened opercula and cranial deformity (head abnormalities) have a great economic impact in the final production. 1.3.2.1. Vertebral column abnormalities. Vertebral column is composed by the vertebral body (centrum), neural and haemal arches. In teleost's, the mineralization of the notochord sheath establishes the identity of vertebral bodies and species, and is initiated by direct mineralization first, and intramembranous - 21 - mineral deposition after around the notochord sheath (Inohaya et al., 2007). Vertebral column abnormalities have been widely reported in several species (S. aurata, Afonso et al., 2000; D. labrax, Koumoundouros et al., 2002; red sea bream, Pagrus major: Matsuoka, 2003; common pandora, Pagellus erythrinus: Sfakianakis et al., 2004; L. calcarifer, Fraser et al., 2004; P. pagrus, Roo et al., 2009a; S. salar, Witten et al., 2009; Pseudocaranx dentex, Mesa-Rodriguez et al., 2012). Vertebral column deformities involve lordosis (V-shaped dorsoventral curvature of vertebral column), kyphosis (ventraldorsal curvature), scoliosis (lateral curvature) (Fig.12a) and LSK syndrome (combination of Lordosis-Scoliosis-Kyphosis, Fig. 12b). Moreover, vertebral dislocation, fusion, shortening, deformation, compression or supernumerally vertebral bodies can take place. Figure 12(a, b). a) P. dentex normal larvae (upper) with kyphosis (mid) and scoliosis (lower); b) S. rivoliana larvae with LSK syndrome. 1.3.2.1. Head abnormalities Cephalic deformities are frequently found in hatcheries. Several forms of mouth abnormalities have been reported, such as crossbite (Fig. 13a), pugheaded, sucker mouthed, prognatism, or reduction of dentale (Boglione et al., 2001, 2003) and reduction or twisting of the operculum (Fig. 13b). a b - 22 - Figure 13 (a, b). a) Crossbite in S. rivoliana larvae; b) operculum reduction in P. dentex larvae. 1.3.3. Abnormalities in Seriola. Even that some Seriola species are commercially produced, limited information is available about incidence of skeletal anomalies incidences. Cobcroft and collaborators (2004) reported several jaw malformations during S. lalandi larval rearing in New Zealand. Ma et al. (2014) compared the occurrence of jaw deformities from two sets of different broodstocks, suggesting the need for selective breeding to reduce occurrence of jaw deformity in S. lalandi larvae. At ULPGC-ECOAQUA facilities (Canary Islands, Spain), S. rivoliana and S. dumerili larvae reared under semi-intensive and intensive systems showed some jaw and operculum abnormalities (Mesa-Rodriguez, unpublished data). Other studies found that infections with Myxobolus buri affect several regions of the brain, including the olfactory and optic lobes, cerebellum and the 4th ventricle, inducing disturbances in the central nervous system and leads to severe scoliosis in cultured S. quinqueradiata in Japan (Egusa, 1985). These findings suggest that lesions in the brain are related to abnormal swimming behaviour and deformity in the vertebral column of the diseased fish (Egusa, 1985; Sakaguchi et al., 1987; Maeno & Sorimachi, 1992; Maeno et al., 1995). Recently, and due to the high incidence of jaw malformations, heritability of lower jaw malformation, nasal erosion and deformed operculum has been also studied (Nguyen et al., 2016). a b - 23 - 1.4. Marine Fish Larvae Nutrition In the natural environment, marine fish larvae are able to hunt and feed on a wide range of prey and sizes. Nevertheless, the difficulties for mass production of wild marine zooplankton have augmented the interest in controlled production of rotifer and Artemia sp. Even though rotifer and Artemia sp. are the main live preys used as larval feed, enrichment become necessary due to their poor nutritional quality. The Seriola larvae grow extremely rapid, being the content of essential fatty acids (EFA), amino acids (AA), minerals and vitamins an important issue in the live prey's enrichment in order to fulfil its nutritional requirements. Nevertheless, the live feed post-enrichment biochemical composition seems to be unstable, mainly when they are stored or left at rearing tank for long time, losing their nutritional value. According to this, adequate Artemia sp. enrichment with long chain polyunsaturated fatty acids (LC-PUFAs) such as docosahexaenoic acid (DHA; 22:6n-3), becomes difficult due its rapid catabolism (Danielsen et al., 1995; Evjemo et al., 2001; Naz, 2008). On the other hand, Naz (2008) stated that rotifers preferentially utilized other fatty acids rather than DHA as energy sources during starvation periods, being rotifers enriched with DHA a suitable live prey for marine fin fish requirements as they can conserve more efficiently the enrichment. As long as endogenous feeding is undergoing, the yolk sac and oil droplet have an essential role in larvae performance by covering their primarily nutritional requirements. At these early stages, broodstock nutrition plays an important role in adequate larvae performance at first stages (Watanabe et al., 1984), as it is directly related to poor hatching quality when broodstock feeds are imbalanced. Indeed, the parental EFA reserve supplemented to the egg is essential for larval survival as well as development of body organs and structures (Izquierdo et al., 2000; Izquierdo & Koven, 2011). Other dietary compounds such as protein, vitamins or carotenoid levels directly affect the spawning, hatching and larval quality. Roo and collaborators (2015) used mackerel (Scomber japonicus) as feeding base for S. rivoliana broodstock, increasing the hatching rates up to 20%, as well as the number of viable eggs and hatched larvae in comparison to the broodstock fed with regimes containing the commercial diets, suggesting a dietary protein level over 70% for S. rivoliana broodstock feeds. In this sense, taurine, one of the most abundant free amino acids in fish and implicated in osmoregulation, antioxidation and hormone release, has been previously researched in Seriola (Matsunari et al., 2003, 2006), and the inclusion of at least 1.0% of taurine to the diet of S. quinqueradiata broodstock resulted on improved egg qualities (Matsunari et al., 2006). Carotenoids have an important role in sexual maturation. Since Verakunpiriya and collaborators (1996) observed the presence of the carotenoid zeaxanthin in S. quinqueradiata eggs, which is converted from the dietary astaxanthin, its inclusion in feeds has been tested - 24 - (Verakunpiriya et al., 1997a; Agius et al., 2001; Vassallo-Agius et al., 2002), obtaining better spawning results, egg quality and final number of normal larvae. Moreover, inclusion of paprika powder as carotenoid source results on even better results than pure astaxanthin inclusion (Vassallo-Agius et al., 2002). Vitamin levels have also been tested in Seriola. Spawning quality was improved in S. quinqueradiata when dietary vitamin E was supplemented in the diet (Watanabe & VassalloAgius, 2003). Contrary to what was expected, the inclusion of krill meal in S. quinqueradiata broodstock diet was not beneficial. Fish fed without krill meal supplementation showed a better hatching rate, number of fertilized eggs and final normal larvae (Verakunpiriya et al., 1997b). 1.4.1. Lipids in marine fish larvae nutrition. As a source of metabolic energy, precursors of essential metabolites and components of biological membranes, the lipids in marine fish larvae play an essential biological role (Sargent et al., 1989). Being crucial for biomembranes structure and energy supply, dietary lipids provide the phospholipids (PL) into the larvae metabolism. Furthermore, dietary lipids are involved in the absorption of fat-soluble vitamins A, D, E and K, and skeletal development (Cahu et al., 2003, 2009; Villeneuve et al., 2005; Roo et al., 2009a; Izquierdo et al., 2010, 2013), as well as being components of hormones or precursors for synthesis of various functional metabolites, such as prostaglandins. However, dietary lipids utilization by body larvae is directly or indirectly affected by several morphological and physiological changes occurring during larval ontogeny and development. 1.4.1.1. Importance of essential fatty acids. Lipids are constituted by fatty acids (FA), being the LC-PUFA the most abundant in fish. The EFA, particularly DHA (22:6 n-3), eicosapentaenoic acid (EPA, 20:5 n-3) and arachidonic acid (ARA, 20:4 n-6) are PUFA, and are known to be essential components of the cellular membranes, modulating physiological mechanisms as membrane transports and enzymatic activity, especially at the first larval developmental stages (Izquierdo, 1996). In this sense, the importance of dietary lipids for larval rearing performance is extremely associated with their high growth rates, energy demands and structural components needs. Indeed, EFA deficiencies are frequently related to elevate marine fish larval mortalities (Izquierdo, 1996). These EFA need to be incorporated in the diet, due to a very low enzymatic capacity (Δ5 and Δ6) to desaturate and elongate its precursors to from DHA, EPA and ARA. Contrarily, freshwater fish are enzymatically able to produce DHA, EPA and ARA from their precursors such as linoleic (18:2 n-6) and linolenic (18:3 n-3) acids. However, competitive interactions exist between DHA, EPA and ARA due to their marked chemical similarities, especially - 25 - between DHA and EPA molecules using the same enzymes to esterify FA into PLs structures (Sargent et al., 1999). According to this, amount of individual PUFA and their ratios must be considered, being important for well-balanced feeds and import for the proper growth and fish larval development (Izquierdo, 1996; Sargent et al., 1999). Watanabe (1993) stated that DHA has greater potential as an EFA for marine fish larvae than EPA, and it requirements being more limiting for growth and survival than those for n-3 PUFA (Izquierdo, 1996). 1.4.1.2. Importance of docosahexaenoic acid. The particular structure of DHA, with a 22-carbon chain and six cis double bonds (22:6 n-3), makes it determinant for many important functions in marine fish metabolism (Watanabe, 1993; Izquierdo, 1996, 2005), and its essentiality has been corroborated by its retention in starved or low EFA fed fish (Koven et al., 1989; Madsen et al., 1999) as well as greater accumulation in ovary, egg and early larval stages (Rønnestad et al., 1998, Mourente et al., 1999; Laurel et al., 2010; Rodríguez-Barreto et al., 2012). During early larval developmental stages, DHA content in marine fish larvae rapidly decreases, and its continuous incorporation must be supplied in order to maintain adequate levels for adequate larval performance (Watanabe, 1993). However, larvae DHA supplementation via exogenous live preys feeding becomes difficult, due to long starving periods in rearing tank and its rapid catabolism (Danielsen et al., 1995; Evjemo et al., 2001; Naz, 2008). DHA deficiencies may cause deleterious effects in marine fish larvae, such as alterations in neurological system with impaired vision or abnormal behaviour (Bell et al., 1995; Brandsen et al., 2005; Benítez-Santana et al., 2007), skeletal deformities (Cahu et al., 2003; Roo et al., 2009a, 2010a; Izquierdo et al., 2010, 2013), delay early mineralization (Izquierdo et al., 2013; Saleh et al., 2015), reduce tolerance to stressful conditions (Izquierdo, 2005; Jalali et al., 2008; Mesa-Rodríguez et al., 2012; Saleh et al., 2013, 2015), reduce larval survival (Copeman et al., 2002; Rezek et al., 2010; Mesa-Rodríguez et al., 2012; Saleh et al., 2013, 2015), among others. In the other hand, an excess in dietary DHA without adequate amount of antioxidant nutrients, may suffer oxidative processes and produce skeletal deformities (MesaRodríguez et al., 2012; Izquierdo et al., 2013), supernumerary vertebrae (Villeneuve et al., 2006), appearance of muscular lesions (Betancor et al., 2011) among other detrimental effects on larvae (Bradsen et al., 2005; Villeneuve et al., 2005; Izquierdo et al., 2010, 2013). Marine fish larvae are visual feeders, and the adequate development of their visual capacity directly affects their water column behaviour (Izquierdo, 2005). According to the importance of DHA in larval eye development, Bell & Dick (1993) determined that rods and - 26 - cones photoreceptors retain and accumulate DHA in the external segments. Roo and collaborators (1999) stated that the most important changes in the eye structure occur during lecitotrophic stages, underlining the importance of broodstock nutrition and the particular importance of DHA in neural and retinal tissue functions. Moreover, the brains of fish contain large amounts of DHA (Mourente et al., 1991; Mourente & Tocher, 1993), and Bell et al. (1995) suggested that DHA deficiency impairs retinal development and visual capacity, as has been reported in S. quinqueradiata larvae (Masuda et al., 1999). Possible functions of DHA involve the constructions of synapses and formation of neural networks. According to this, DHA deficiencies may result in poor development of central nervous system and the improper development of learning ability, including feeding and schooling behaviour. For Seriola, few studies on DHA requirements have been published (Table IV). As long as dietary DHA is considered to be essential for adequate brain development at early larval stages, studies with S. quiqueradiata larvae fed with enriched DHA Artemia suggest the essentiality of this EFA for the adequate development of schooling behaviour (Masuda et al., 1998, 1999; Ishizaki et al., 2001). The swim bladder inflation is considered to be a critical factor for Seriola sp. larval survival, and DHA plays an important role for its proper development. In this sense, Matsunari et al. (2012a,b) determined that S. dumerili larvae fed rotifers enriched with DHA showed improved swim bladder inflation, as well as growth and final survival. The increase of larval growth and survival due to the elevation of DHA content in live preys enrichment has also been reported for S. quinqueradiata (Furuita et al., 1996; Ishizaki et al., 1996, 1997, 1998; Takeuchi et al., 1998 ), as well as increased the tolerance to stressful conditions in S. rivoliana larvae (Mesa-Rodríguez et al., 2014a). - 27 - Table IV. DHA requirements for Seriola larvae. Specie Tested Parameter Objective Feed Levels Optimum Requirement Reference S. quinqueradiata DHA & EPA Larval requirements Enrich. Artemia sp. DHA: 0 - 2.63% EPA: 0.5 - 8.9% DHA: 1.39-2.63% EPA: 3.65% Furuita et al., 1996. S. quinqueradiata DHA + Euglena sp. & Vit.E Larval performance Enrich. Rotifer - - Ishizaki et al., 1996. S. quinqueradiata ≠ levels EPA + DHA Substitutability between EPA/DHA Enrich. Artemia sp. DHA: 1.4% EPA: 2.3 - 5.5% DHA: 1.4 % EPA: 4.0 % Ishizaki et al., 1997. S. quinqueradiata ≠ levels ARA + DHA ARA & DHA effect in Larval performance Enrich. Artemia sp. DHA: 2.1 - 2.5% ARA: 4.0% DHA: 2.1 - 2.5 % ARA* Ishizaki et al., 1998. S. quinqueradiata DHA Behaviour development Enrich. Artemia sp. DHA: 1.46 - 2.40% - Masuda et al., 1998. S. quinqueradiata DHA Larval requirements Enrich. Rotifer Enrich. Artemia sp. DHA: 0 - 1.6 - 2.6% DHA: 1.6-2.6% Takeuchi et al., 1998. S. quinqueradiata DHA Schooling Behaviour Ontogeny & Brain Enrich. Artemia sp. DHA: 0 - 2.13 - 2.53% DHA: 2.13-2.53% Ishizaki et al., 2001. S. dumerili - Larval requirements; mass production Enrich. Rotifer Enrich. Artemia sp. DHA Rot: 1.6-4.0% DHA Art: 0.7-1.3% - Yamamoto et al., 2008. S. dumerili - Rotifers quality in larval rearing Tanks Enrich. Rotifer E.R*: 0.8-6.0% T.R*: 3.9-10.3% - Yamamoto et al., 2009. S. dumerili Algae with ≠ DHA content Larval performance & SB Inflation Enrich. Rotifer DHA: 0.04-0.6-1.32% - Matsunari et al., 2012a. S. dumerili Algae with ≠ DHA content Larval performance & SB Inflation Enrich. Rotifer DHA: 0-0.4-1.0-1.9% DHA: 1.5 % Matsunari et al., 2012b. S. rivoliana DHA Larval performance & Stress resistance Experimental Microdiet DHA: 3 - 6.5% DHA: 3.7-4.5 % Mesa-Rodríguez et al., 2014a. - 28 - 1.4.2. Microdiets for marine fish larvae. Supply of an optimal feed which fulfils larval nutritional requirements implies difficulties. As previously mentioned, the utilization of enriched live preys such as Artemia sp. becomes difficult due to its rapid catabolism (Danielsen et al., 1995; Evjemo et al., 2001; Naz, 2008). The early substitution of live prey with a compound diet is a suitable option in order to reduce production costs and avoid unbalanced feeds. Contrarily, early feeding with inert diets resulted in low larval performance, possibly due to its composition, palatability or physical characteristics (Person Le Ruyet et al., 1993) or unable to digest it (Kolkovski et al., 1993; Zambonino-Infante & Cahu, 1994; Kolkovski, 2001). However, the combination of live prey and inert diet (co-feeding) improve larval performance (Fernández-Díaz & Yúfera, 1997; Kolkovski et al., 1997; Hernández-Cruz et al., 2007; Roo et al., 2010a; Sandel et al., 2010). According to this, the use of experimental microdiets in aquaculture research to identify the marine finfish larvae nutritional requirements has being widely reported (Betancor et al., 2011, 2012a, b; Izquierdo et al., 2013; Saleh et al., 2013, 2015; Hernández-Cruz et al., 2015). However, very few studies based on microdiets have been reported for Seriola. Mesa-Rodríguez et al. (2014a) studied the effect of increasing dietary DHA content in microdiets for S. rivoliana larvae, in which the better resistance to acute stress and the incidence of skeletal deformities was correlated with the DHA content. By the way, the incidence of skeletal anomalies has a very important economic implication in aquaculture. The most important anomalies are those affecting the opercula complex, neurocranium and vertebral column, due to its high visual impact. The appearance of skeletal deformities has been widely studied for several marine aquaculture finfish species (Koumoundouros et al., 1997a,b, 2001a,b, 2002; Boglione et al., 2001, 2003; Sfakianakis et al., 2003, 2004; Roo et al., 2005a,b, 2009a). In this sense, the hatchery phase remains the bottleneck of the sector, and the good understanding of species specific larval requirements and bone development becomes indispensable in order to increase fry quality and reduce production costs. - 29 - 3. Material and Methods 3.1. Biological aspects of Seriola rivoliana S. rivoliana belong to the Carangidae family. Carangids are mainly marine teleost's, mostly distributed along tropical and subtropical waters of Atlantic, Indian and Pacific Oceans. Generally present compressed body and small cycloid scales, which often modify into spiny scutes along the lateral line. Carangid juveniles and adults present 2 dorsal fins, with 3-9 spines in the anterior dorsal fin and 1 spine with 18-37 soft rays in the posterior dorsal fin, 3 anal spines with 15-31 soft rays and a widely forked caudal fin. In the Canary Islands, 4 species of Seriola genus are described (S. dumerili, S. rivoliana, S. fasciata and occasionally S. carpenteri). With the common name of longfin yellowtail or Almaco jack, S. rivoliana is a pelagic and benthopelagic species, rarely found at inshore waters. Its circumglobal distribution (Fig.14; Fishbase, 2016), its aquaculture production interest. It can be found in the Eastern Central Atlantic regions from Portugal (Azores and Madeira), Canary Islands to Cape Verde but also sporadically in the Mediterranean Sea (Castriota et al., 2002). Figure 14. S. rivoliana distribution (Fishbase, 2016). As a fast swimming predator, its diet is mainly based on fish. The body colour is uniformly brownish to olivaceus green with lighter colours at the sides and belly. A dark band is commonly present through the eye to the upper back and sometimes an amber line extends from the eye up to the tail. As a main characteristic, the first rays of the second dorsal fin are about twice longer than the rest dorsal spines (Fig. 15). - 37 - Figure 15. Adult of S. rivoliana. The maximum size registered is 119 cm (TL) and 59.87 kg (IGFA, world record), but common size its around 55 cm fork length (Smith-Vaniz, 2002). Histological studies indicated that the ovarian development of Seriola sp. occurs according to a group synchronous pattern (Marino et al., 1995), forming aggregations when spawning takes place. S. rivoliana presents seasonal maturation, directly influenced by increasing temperature. In the Canary Islands, hormonal injection from late spring (May) to late summer (October) permitted to obtain successful spawns of S. rivoliana (Roo et al., 2009b, 2012, 2015). In captivity, females tend to grow faster than males, increasing their weigh twice faster every two years, and 66% of the females are considered sufficiently mature (oocytes over 500μm) to be hormonally induced after 4 years (Roo et al., 2015) with weights over 6 kg. S. rivoliana eggs obtained in captivity are transparent and spherical, with an average egg diameter of 1.1 mm and a single oil droplet of 0.24 ± 0.02 mm (Roo et al., 2012). 3.2. Experimental conditions The experiments and samples analysis described in this document were carried out at the GIAQUA (University of Las Palmas de Gran Canaria, ULPGC) facilities, in Telde, Gran Canaria (Spain). - 38 - 3.2.1. Broodstock S. rivoliana broodstock were collected as sub adults from the wild (Fig. 16a) in 2009 with an average weight of 1.7±0.25 kg, and adapted to captivity in 10m3 squared glass fibber tanks (3x3x1.5m depth). Fish were kept under natural photoperiod, temperatures ranging from 18 to 24ºC and fed twice a week with commercial pellets (Vitalis Repro™; Skretting, Burgos, Spain) corresponding to 1% of the body weight (BW), supplemented once a week with frozen squid (Illex argentines) and mussels (Mytilus galloprovincialis) at 2% of BW. The maduration stage was assessed using gonadal biopsy, oocytes from females were taken in vivo and placed in Serra's solution (6:3:1, 70% ethanol, 40% formaldehyde and 99.5% acetic acid) to be measured using a profile projector (Mitutoyo Pj-3000A, Kanagawa, Japan). Mean diameter of the largest oocytes were determined. When the oocyte diameter was over 500μm (Fig. 16b), whole individuals were injected with gonadotropin releasing hormone analogue (GnRHa, desGly 10, [D-Ala6]; Sigma-Aldrich, St. Louis, MO, USA) at a dose of 20 μg/kg of body weight (Mylonas et al., 2004). Hormonal treatment was applied every two weeks during the spawning season. Figure 16 (a, b). a) S rivoliana adult; b) S. rivoliana gonadal biopsy inserting catheter (Kruuse, Langeskov, Denmark) into the gonadal cavity. 3.2.2. Egg stocking Floating S. rivoliana eggs were collected 32h after hormonal injection. The spawning quality was measured according to the number of fertilized eggs and 3 days old hatched larvae (Fig. 17a,b), following the methodology described by Fernández-Palacios et al. (1995). Eggs were directly stocked in the rearing tanks (40m3 - 2m3). a b - 39 - Figure 17(a, b). a) S. rivoliana eggs; b) S. rivoliana larvae of 1dph. 3.2.3. Larval rearing tanks The different larval rearing tanks used in the present Thesis were chosen according to the culture systems and volumes. - Semi-intensive (SIS). SIS tanks presented a cylinder-conical shape with a diameter of 5m, 2.35m depth and a total volume of 40m3 (Fig. 18a). The water entrance was located in the lateral bottom part of the tank at fist larval stages and modified to the surface with the age of the larvae. As well as the water entrance, output waters could be located at the bottom and/or the lateral side of the tank, and modified along the larval stage. Airlift is located all around the tank. - Intensive systems (IS). This cylinder-conical tanks with a diameter of 1.5m, 2.10m depth and a total volume of 2m3(Fig. 18b), are frequently used for high larval densities. As well as the SIS tanks, water entrance and output can be modified according to the culture requirements. In both systems, larval rearing was conducted under 12:12 (12 h light:12h dark) photoperiod using mixture of artificial fluorescent lights (Mod. TLD 58W/54-765; Philips, Lyon, France) and natural sun light with intensity between 1.000-3.500 lux. Moreover, water conditions were daily measured (temperature: 24.03 ± 0.26 ºC; oxygen levels: 6.22 ± 0.21 ppm; OxyGuard, Denmark). a b - 40 - Figure 18 (a, b). a) Semi-intensive 40m3 tank; b) Intensive 2m3 tank. - Experimental tanks. For specific feeding trials, larvae were settled in 200 l fibreglass cylinder tanks with conical bottom and painted in a light grey colour (Fig. 19a,b). Water conditions were daily measured (temperature: 22.5 ± 0.6 ºC; oxygen levels: 6.5 ± 0.3 ppm; OxyGuard, Denmark). Photoperiod was kept at 12:12 (12 h light:12 h dark) by fluorescent daylights at 1700lux (digital Lux Tester YF-1065; Powertech Rentals, Osborne, Australia). Figure 19 (a, b). Intensive 200l tank. 3.2.4. Phytoplankton and live preys culture Pseudo-green water technique was used for both SIS and IS, adding live phytoplankton (Nannochloropsis sp.) to maintain a concentration of 250.000 cell/ml. In order to preserve this concentration in the rearing tanks, additional phytoplankton culture was maintained. For this massive phytoplankton production, transparent polyethylene bags of 50, 230 and 460 l were conducted in "Bach" system culture, following the phytoplankton culture protocol described by Roo et al. (2009b). a b a b - 41 - Rotifers mass culture was carried out on cylinder conical fibreglass tanks of 1.700 l total volume, with a mixture of fresh and seawater to achieve a salinity of 25 ppt (Roo, 2009b). Rotifers (Brachionus sp.; L-strain) were enriched during 6 hours before harvesting with DHA Protein Selco (INVETM), following manufacturer instructions. Artemia sp. cysts used during the study were ready to hatch without decapsulation phase. This cysts have been treated with SEP-Art technology (Fig. 20a), giving a magnetic coating on the cyst which after pass through separator tube containing passive magnets (Fig. 20b) attract the iron coated shells and obtaining clean nauplii after hatching. Afterwards, Artemia sp. nauplii were enriched with Easy DHA Selco (INVE, Dendermonde, Belgium) in a tank at a concentration of 250-300.000 nauplii/l during 18hours. Figure 20 (a, b). a) Artemia sp. cyst; b) Artemia sp. cysts separator (SEP-Art technology, INVETM). 3.2.5. Microdiets (Formulation and Preparation) Two different types of inert diets were used all along the study. During the first larval experiences (Study I), the weaning protocol included hand feeding a commercial diet (Gemma Micro, Skretting, Vervins, France) at first weaning days, and with automatic feeders afterwards. For the DHA nutritional requirement experience (Study IV), five isoproteic (54.8%) and isolipidic (24.1%) microdiets, which varied in their DHA content from 0.5-5.0% dry weight (dw), were elaborated. The microdiets were manufactured according to Liu et al., (2002), by mixing squid powder with water soluble compounds, lipids and fat soluble vitamins. Vitamin, mineral and attractant mixes were added according to Betancor et al., (2012a,b). Diluted gelatine was used as a binder. The obtained paste was compressed and dried at 38ºC for 24h (Fig. 21a,b). Next, dried pellets were ground and sieved in two size ranges (250-500μm and 500-710μm). a b - 42 - Figure 21(a, b, c). a) Compressing processes of the paste; b) compressed paste; c) grounded pellets in sizes. In order to evaluate the fatty acid effect and guarantee the desired lipid content, squid powder (SkrettingTM) was defatted thrice with a chloroform:meal ratio of 3:1 (Fig. 22a,b). The defatted meal was separated from the chloroform fraction after extraction with a vacuum pump, spreaded out in a tray and the remaining solvent was evaporated during 12 h at 38ºC. Figure 22(a, b). a) Defatting processes of squid meal; b) extraction of chloroform fraction with a vacuum pump. 3.2.6. Feeding Regimes and Protocol Both systems (SIS and IS) followed the same rearing protocol (Table V) according to the larval rearing techniques established by Roo et al. (2010a). S. rivoliana eggs were settled at a density of 4.5 eggs/l in SIS tanks and 125 egg/l in IS tanks. All experimental tanks were supplied with filtered and UV sterilized sea water, being clear water fist up to 2dah and pseudogreen water after until 30dph, at a concentration of 250.000 cells/ml. Water exchange was a c a b b - 43 - progressively increased from 15% to 100% of tank volume per day. Enriched rotifers were added twice a day (08:00; 14:00) to maintain a live prey density of 4-5 rot/ml in SIS and 7.5-10 rot/ml in IS from 2 to 25 dah. Enriched Artemia sp. nauplii were added first once a day at 15dah (0.25 Art/ml) and increased up to three times a day after in both systems. Table V. S. rivoliana feeding sequence. For the dietary DHA experience, 90 larvae per tank (in triplicate) with 30 dah were settled in 200 litre tanks. Diets were manually supplied every hour from 8:30 to 19:30. Initially, S. rivoliana larvae were fed twice a day during the first 5 days with un-enriched rotifers and Artemia to ensure the adaptation to the new rearing tanks. The daily amount of microdiet was gradually increased from 1.5gr to 2.5gr per tank. 3.3. Measurements The aim of the different experiences determined the type of measurements carried out as well as the sampling points. At the different samplings larvae were sacrificed by immersion in water and ice according to the current regulations (Spanish Royal Decree 1201/2005) which were accepted by the Spanish Ethic Welfare Committee (Comité Ético del Bienestar) of the University of Las Palmas de Gran Canaria (ULPGC) in 2011. 3.3.1. Larval Length, Weight and Final Survival In order to evaluate for first time the larval growth of S. rivoliana larval growth under two different rearing conditions (SIS and IS), total length (TL) and dry weight (DW) of 25 larvae per tank were measured every 5 days from hatching to 30 dah. TL and other larval body - 44 - meristic characters such as standard length (SL), pre-anal length (PAL), eye diameter (ED), cephalic height (CH), yolk sac length (YSL) and lipid globule diameter (LGD) were evaluated (Fig. 23) using a profile projector (Mitutoyo PJ-3000A, Kanagawa, Japan) Figure 23. Meristic length measurements of S rivoliana larvae. Larval DW was determined by measuring (in triplicate) the whole body weight of 10 larvae washed with distilled water and dried in a glass slide in an oven (Jouan EU 28, S. Herblain, France) at 100ºC during 24 hours until constant weight in a precision balance (Mod. Mettler, AG 204, Ohio, USA). At the beginning (30dah), intermediate (40dah) and final (50dah) points of the experimental microdiets trial, growth was assessed by estimating the TL of S. rivoliana larvae. Final survival was determined at 30 dah (SIS and IS) or 50 dah (microdiets trial) counting the remaining alive larvae in the experimental tanks. 3.3.2. Activity Test An air exposure test was performed on 20 and 30 dah larvae from SIS and IS tanks. On day 20, larvae (n = 45) were individually exposed to air for 15, 30 or 60 seconds in a 500 μm nylon mesh screen (Izquierdo et al., 1989). After the air exposure, larvae were transferred to an aerated 2l beaker and survival was recorded 24 h later. At 30 dah, a new set of stress tests were performed. At this point, two clove oil doses (1 and 2 ppt), as anaesthetic, were evaluated in combination with different air exposure times (15, 30, 60, 75 and 90 s). The same procedure (30 seconds of air exposure) was performed at 42 and 50 dah S. rivoliana specimens (n = 15) from dietary DHA content experience. Larval survival was recorded 24 h after each stress test. - 45 - 3.3.5. Deformities characterization For the skeletal anomalies characterisation, a total of 15 larvae (50 dah) per tank were fixed in 10% buffered formalin and stained with alizarin red according to the methodology of Vandewalle et al. (1998). The different regions were divided (Fig. 24) and evaluated (Table XII) according to Boglione et al. (2001). Figure 24. Regions of stained S. rivoliana larvae. - 52 - Table XII. Skeletal abnormalities classification (Boglione et al., 2001). Code Description Region A Cephalic vertebrae (carrying epipleural ribs) B Pre-haemal vertebrae (carrying epipleural and pleural ribs and open haemal arch, without haemal spine) C Haemal vertebrae (with haemal arch closed by haemal spine) D Caudal vertebrae (with haemal and neural arches closed by modified spines) E Pectoral fin F Anal fin G Caudal fin H Dorsal spines I Dorsal soft rays L Pelvic fin Anomalies 1 Kyphosis 2 Lordosis 3 Partial vertebral fusion 3* Total vertebral body fusion 4 Vertebral malformation (deformation, ossification ridges, marked reduction in length or elongation, intervertebral bony plate) 5 Malformed neural arch and/or spine 5* Supernumerary neural elements. Absence of neural elements 6 Malformed haemal arch and/or spine 6* Supernumerary haemal elements Absence of haemal elements 7 Malformed rib 7* Supranumerary pleural rib 8 Malformed pterygophores (deformed, absent, fused, supernumerary) 9 Malformed hypural (deformed, absent, fused, supernumerary) 9* Deformed or broken parahypural or fused with hypural/hemaspine 10 Malformed epural (deformed, absent, fused, supernumerary) 11 Malformed ray (deformed, absent, fused, supernumerary) 12 Swim-bladder anomaly 13 Presence of calculi in the urinary ducts 14 Malformed maxillary and/or pre-maxillary 15 Malformed dentale 16 Other cephalic deformities (glossohyal, neurocranium, ..) 17L/R Malformed left/right opercle 17*L/R Malformed, absent, fused branchiostegal ray 18 Predorsal bones anomalies 19 Hypural with decalcifications 20 Decalcifed pterygophore 21 Deformed epipleural ribs 22 Deformed dorsal ribs 23 Deformed pleural ribs 24 L/R Decalcified left/right opercular plate 25 Epural with decalcifications 26 Supernumerary bone 27 Decalcified urostyle 28 Ossification defects in vertebrae 29 Deformed postcleithrum S Scoliosis CL L/R Malformed left/right cleithrum COR L/R Malformed left/right coracoid SBS Saddle-back syndrome - 53 - 3.4. Biochemical Analysis 3.4.1. Proximate Analysis For biochemical analysis, samples of feeds (rotifer, Artemia sp. and microdiets) were collected in all the feeding trials. Larval biochemical composition was analyzed collecting samples at the beginning (30 dah) and the end (50 dah) of the dietary DHA experiment after a starving period of 12 h, washed with distilled water and kept at -80ºC. 3.4.2. Moisture and Ash content Moisture and ash content was determined according to the Association of Official Analytical Chemists (A.O.A.C., 1995). Samples (about 100mg) were dried at 110ºC during 24 h until constant weight to obtain moisture content. The ash content was determined by drying the samples in an oven at a temperature of 450ºC until a constant ash weight was attained. The moisture and final ash content was obtained applying following the equations: Moisture (%) = 100−(B−A)− (C−A) B−A Ash (%) =100 x Ws Wa Where: A = weight of empty container; B = weight of wet sample + container; C = weight of dry sample + container; Ws = weight of sample; Wa = weight of ash. 3.4.3. Total Lipid and fatty acid content The total lipids were extracted following the method of Folch et al. (1957), by homogenising the samples in an Ultra Turrax (IKA-Werke, T25 BASIC, Staufen, Germany) with a solution of 5 ml of Chloroform:Methanol (2:1) and 0.01% of BHT, and filtered after adding KCL to increase water phase polarity. The remnant solvent was dried under nitrogen atmosphere until obtain constant lipid weight. Fatty acid methyl esters (FAMEs) were obtained following the method of Christie (1982) by transesterification of total lipid with 1% sulphuric acid in methanol (H2SO4). The reaction was conducted in dark conditions under nitrogen atmosphere for 16 h at 50ºC. Afterwards, fatty acid methyl esters were extracted with hexane and purified by adsorption chromatography on NH2 Sep-pack cartridges (Waters S.A., Massachusetts, USA) as described by Fox (1990). FAMEs were separated by GLC (GC-14A, Shimadzu, Tokyo, Japan) as described Izquierdo et al. (1989). - 54 - 3.5. Statistical Analysis All the data were statistically treated using SPSS Statistical Software System ver 15.0 (SPSS, Chicago, IL, USA). A t-test for simple mean comparison analysis (P < 0.05) (Sokal & Rolf, 1995) was applied to compare differences between rearing systems. When data were not normally distributed, arcsine-transformation was applied, and then Kolmogorov– Smirnov nonparametric test was applied to the non-transformed data. Results are presented as mean values ± SD. - 55 - Study I First results of spawning and larval rearing of longfin yellowtail Seriola rivoliana as a fastgrowing candidate for European marine finfish aquaculture diversification. Published in: Aquaculture Research, 2012, 1 - 12; doi: 10.1111/are.12007 - 57 - First results of spawning and larval rearing of longfin yellowtail Seriola rivoliana as a fast-growing candidate for European marine finfish aquaculture diversification. J. Roo, H. Fernández-Palacios, C.M. Hernández-Cruz, A. Mesa-Rodriguez, D. Schuchardt & M. Izquierdo. Abstract The present study describes the adaptation of longfin yellowtail Seriola rivoliana as broodstock and first larval rearing trials under intensive and semi-intensive conditions. Fifteen sub-adults were captured in the South coast of Gran Canaria (Canary Islands, Spain) in June 2007. Fish (initial weight 1.76 ± 0.25 kg) reached a weight of 6.0 ± 1.1 kg in July 2010. Once a year, fish were sampled to determine individual growth in weight and size. In addition, the state of sexual maturity was established based on gonadal biopsies. On the basis of repeated hormonal injection (GnRHa, 20 lg/kg), 10 successful spawns were obtained between July and October 2009, with 92.5 ± 5.5% and 72.6 ± 17.2%, fertilization and egg viability respectively. First results of larval rearing under semi-intensive conditions, showed an average survival at 30 DAH of 2.5% as compared with 0.5% under intensive conditions. The low survivals under the two rearing conditions in addition to their failure to pass a stress test could be attributed to deficiencies in essential fatty acids as could be seen in both eggs and feeds. Morphometric parameters showed no significant difference between the two rearing systems in 30 DAH larvae. Keywords: diversification, seriola, broodstock, spawning, larval rearing, rearing techniques. - 59 - 1. Introduction Longfin yellowtail Seriola rivoliana as other seriola species is considered as one of the most important emerging marine finfish species in Japan, Australia and the United States. In contrast, this species has not been under development for mariculture in Europe. This circumtropical carangid species can be found in Eastern Central Atlantic regions from Portugal (Azores and Madeira), Canary Islands to Cape Verde (Fischer, Bianchi & Scott 1981), whereas some individuals were caught sporadically in the Mediterranean sea (Castriota, Greco, Marino & Andaloro, 2002). As other seriola species, S. rivoliana is well known for its fast growth, reaching a maximum standard length (SL) of 160 cm and a maximum weight of 59 kg (IGFA, 2001) and its high market value of 7–10 USD per kg (Nakada, 2002). However, the bottleneck to S. rivoliana mass production is the unreliable supplies of juveniles resulting from poor spawns and low hatchery survival. Therefore, the industry in Japan, Australia and USA is relying on the collection of fingerlings from the wild, to be ongrown in tank and cages (Nakada, 2002; Yamamoto, Teruya, Hara, Hokazono, Hashimoto, Suzuki, Iwashita, Matsunari, Fuguita & Mushiake, 2008). A reproduction protocol for this species in terms of culture conditions, maturation and use of hormonal treatment is still not available. Furthermore, larval rearing studies of this species are scarce and limited to some reports in Ecuador (Benetti, 1997; Blacio, Darquea & Rodríguez, 2003) and Hawaii (Laidley, Shields & Ostrowksi, 2004). On the other hand, reproduction and larval rearing protocols have been developed for similar species, such as Japanese yellowtail Seriola quinqueradiata, greater amberjack Seriola dumerili or yellowtail king fish Seriola lalandi (Benetti, 2000; Poortenaar, Hooker & Sharp, 2001; Nakada, 2002; Papandroulakis, Mylonas, Maingot & Divanach, 2005). To enhance the development of S. rivoliana for aquaculture diversification in Europe, different experimental activities, including broodstock management and larval rearing, are being conducted in the Canary Islands (Spain). - 60 - The objective of this study was to test mesocosms or semi-intensive techniques, which have been previously tested and reported successful in the larval rearing of difficult-to-rear fish species, which could be applied later to more intensive, commercial systems to improve biological performance and system productivity (Papandroulakis et al., 2005; Jerez, Samper, Santamaría, Villamandos, Cejas & Felipe, 2006; Roo, Hernández-Cruz, Socorro, FernándezPalacios & Izquierdo, 2010). The comparison between the intensive and semi-intensive techniques, for larval rearing of S. rivoliana, will contribute to better understand of the husbandry needs of the species regarding future application in commercial production. 2. Material and Methods 2.1. Broodstock Fifteen S. rivoliana sub-adults (1.76 ± 0.25 kg) were captured at the South coast of Gran Canaria(Canary Islands, Spain), transported to land facilities and adapted to captivity in 10 m3 squared glass fibber tanks (3 m 9 3 m 9 1.5 m depth). Fish were kept under natural photoperiod and natural sea water with 37 g L-1 salinity and temperature ranging from 18 to 24°C year around. After capture, all fish were weighed, sized, individually tagged with PIT tags of 0.1 g and 152 x 12 mm in length (EID Ibérica SA – TROVAN, Madrid, Spain), and sexed by gonadal biopsy inserting a 1.3 mm internal diameter catheter (Kruuse, Langeskov, Denmark) into the gonadal cavity and applying gentle aspiration. Fish were fed twice a week with commercial pellets (13 mm, Vitalis ReproTM; Skretting, Burgos, Spain) corresponding to 1% of the body weight (BW), supplemented once a week with frozen squid (Illex argentines) and mussels (Mytilus galloprovincialis) at 2% of BW. Once a year (June), the whole population was anaesthetized with clove oil (Guinama S.L, Valencia, Spain; 50 ppm) and standard length (SL), body weight and condition factor (CF) were recorded (Table 1). The maturation stage was assessed using gonadal biopsy, oocytes from females were taken in vivo and placed in Serra’s solution (6:3:1, 70% ethanol, 40% formaldehyde and 99.5% acetic acid) to be measured using a profile projector (Mitutoyo PJ-3000A, Kanagawa, Japan). Mean diameter of the largest oocytes - 61 - species to mature in captivity although, no spontaneous spawns were obtained. Lack of oocyte development and maturation in other species in captivity is generally associated with inadequate environmental (photoperiod, temperature) conditions or other stress factors, such as fish density or rearing volume (Micale, Maricchiolo & Genovese, 1999; Benetti, 2000). Indeed, lower fish densities and bigger tanks used in comparison with the present study, for same species in Hawaii and Ecuador, lead to successful spontaneous spawns (Blacio et al., 2003; Laidley et al., 2004). In the present study, the use of hormonal induction of oogenesis and egg maturation with the GnRHa analogue was associated with a high rate of egg fertilization without negative effects on brood fish survival. The dosage applied was based on Mylonas et al. (2004) for greater amberjack and Duncan, Estevez and Mylonas (2007) for meagre (Argyrosomus regius). The fact that in S. rivoliana, in the present study, was induced to spawn at a size of about 5.0 kg would suggest that this seriola species has a practical advantage over other seriola species, such as S. dumerilli, as the latter needs very large tanks as to reach sexual maturity (Mylonas et al., 2004; Jerez et al., 2006). 4.2. Larval rearing Survival rates obtained in this study (0.5–2.5%) were similar to those reported for the same species in Ecuador (Blacio et al., 2003), Hawaii (Laidley et al., 2004) and other seriola species, such as S. lalandi (Tachihara, El-Zibdeh, Ishimatsu & Tagawa, 1997) and S. dumerili (Papandroulakis et al., 2005; Hamasaki, Tsuruoka, Teruya, Hashimoto, Hamada, Hotta & Mushiake, 2009). In the present study, the highest larval mortalities were observed in the early developmental stages, from 8 to 12 DAH. Different studies attributed the early larval mortalities to an unstable bacterial flora in the rearing water and opportunistic bacteria colonization, the larval digestive tract (Hansen & Olafsen, 1999; Makridis, Fjellheim, Skjermo & Vadstein, 2000; Verner-Jeffreys, Shields, Bricknell & Birkbeck, 2003; Reid, Treasurer, Adam & Birkbeck, 2009) causing massive mortalities mainly when intensive larval rearing systems are utilized - 68 - (Skjermo & Vadstein, 1999). The improved larval survival in the SIS in comparison with IS would suggest that seriola larval rearing should be performed following the SIS regime. This could be associated with the use of larger water volumes, with a low renewal rate and low larval density. This regime might help promote the development of a more stable environment, probably associated with a more mature microbial flora, similar to mature waters obtained in recirculation systems (Roo et al., 2010; Attramadal, Salvesen, Xue, Øie, Størseth, Vadstein & Olsen, 2012). Indeed, the use of microbial matured water has being reported to improved, larval survival and improved feeding incidence at early stages in different marine species, such as Atlantic halibut Hippoglossus hippoglossus (Skjermo, Salvesen, Øie, Olsen & Vadstein, 1997) or Atlantic cod Gadus morhua, among others (Attramadal et al., 2012). Furthermore, theorical prey availability in relation with initial larval population was tenfold lower in the IS, with around 160 rotifers/larva/day, whereas more than 1600 rotifers/larva/day could be available in the SIS. Similar data where previously reported by Roo et al. (2010) in red porgy larvae. Nevertheless, this fact seems not to affect feeding incidence, measured in 5 DAH (50%) and 8 DAH (80%), in the present study in both system regimes. Furthermore, our observations on seriola larval feeding were lower than those reported in similar species, such as S. dumerilli (>70% first feeding at 4 DAH; Hamasaki et al., 2009), suggest that initial larval mortalities could be also related to the low success for first feeding in S. rivoliana under both culture techniques and suggesting a strong dependence on maternal reserves during the early development. Similar results were reported for S. lalandi (Chen, Qin, Kumar, Hutchinson & Clarke, 2006) and could be related to nutritional imbalances in broodstock diets, and hence, deficiencies in essential nutrients in the endogenous reserves of the larvae, nutritional imbalance of live prey at first feeding or inadequate culture conditions. In fact, it was suggested that broodstock diets for Seriola sp. may need an extra supply of neutral lipids as an important energy source for early larval development (Hilton, Poortenaar & Sewell, 2008). In the present study, B. plicatilis L-strain seems to be adequate for first feeding of S. rivoliana larvae, which is in agreement with the results reported by Hamasaki et al. (2009). These authors reported that larval survival in S. dumerilli was independent of rotifer size of different strains, which had a - 69 - range of lorica length from 135 to 211 lm. In the present study, copepods (Harpacticoida, Tisbidae) produced endogenously in the semi-intensive systems tanks were observed in the larval digestive tract (Plates 1e, 1f, 1g). This was particularly true for the small naupliar stages of marine copepods, suggesting the preferences for this type of prey by Seriola larvae. These results are similar with those reported by Van der Meeren (1991) when turbot larvae were offered rotifers and copepods. In general, copepods are a suitable prey-size organisms for firstfeeding larvae that are rich in a number of biochemical components, such as lipids, highly unsaturated fatty acids, digestible phospholipids, protein, protein-bound amino acids, free amino acids, pigments or vitamins (Støttrup & Norsker 1997; Van der Meeren, Olsen, Hamre & Fyhn, 2008), which are essential nutrients and as such would helped promote seriola larval survival under semi-intensive conditions. In older larval stages (>20 DAH), there was a continuous mortality with weak larvae floating at the water surface. Seriola larvae at this age also succumbed after stress test. The presence of larvae with faeces, including undigested and even live Artemia, demonstrated the low digestibility of this prey (Plate 1i), which leads to larval malnutrition when Artemia is the main source of food (15–30 DAH). Moreover, striking differences were found in the fatty acid profiles of eggs and different feeds utilized in this study, suggesting a potential fatty acid deficiency. This is in agreement with the shock syndrome (sudden mortality) observed after the stress test or even at capture as it was reported by Izquierdo et al. (1989). Also, sudden massive mortalities were recorded along the different trials in this rearing period. The relation between larval welfare and essential fatty acids, such as 20:5n-3 (EPA), 22:6n-3 (DHA) and 20:4n-6 (ARA), has been emphasized in early stages of development for different marines species (Izquierdo 1996), with ‘shock syndrome’ being one of the indication of EFA limitations in the diets (Izquierdo et al., 1989). Previous results from Yamamoto et al. 2008; suggest that in Artemia feeding stages at least 1.4–2.6% of DHA and 2.3–4.1% of n-3 HUFAs are needed in the diet of the yellowtail S. quinqueradiata. Present results showed that S. rivoliana eggs contents 16.5% total fatty acids (TFA) (about 4.3% DW) in DHA and n-3 HUFA were as high as 26.9% TFA (7.1% DW) suggesting even higher EFA requirements for this species, than for S. quinqueradiata. In agreement with these potential high - 70 - EFA requirements, high mortalities could be related with the low DHA and n-3 HUFA content in rotifers (10.3–19.6%), Artemia nauplii (5.3% and 13.3%) and dry feeds (5.7–9.9%). In addition, the ratios of DHA/EPA and oleic/DHA, used to evaluate the essential fatty acids requirements (Izquierdo 1996), greatly differed for feeds (0.93–1.76 DHA/EPA; 1.8-4.1 oleic/DHA) and eggs (2.45 and 1.22 respectively). It is noteworthy to mention, that not only fatty acid composition, but also lipid classes should be taken into consideration in larval feeds. Thus, commercial live prey enrichment, such as the ones utilized in present experiment provides most of EFA as neutral lipids (NL) particularly as triglycerides (TG) and free fatty acids (FFA) forms (data not shown). On the contrary, natural live preys, such as copepods, are rich in phospholipids (PLs) (Van der Meeren et al., 2008). It is well known that in marine fish larvae, PL is structural constituents of bio-membranes and therefore highly demanded in the fastgrowing larvae. However, there are several indications that fish larvae are unable to efficiently synthesize PL in a rate fast enough to cover their high demand and therefore PL needs to be included in the diet (Izquierdo & Koven, 2011), suggesting that nutritional modification in S. rivoliana larval feed could be beneficial for improved larval survival. Total length of newly hatched larvae averaged 2.54 ± 0.01 mm, which is similar to previous studies in this species (Blacio et al., 2003) and S. dumerilli (2.88 mm TL) (Papandroulakis et al., 2005), but shorter than S. lalandi larvae (4.30 mm TL) (Chen et al., 2006). Nevertheless, both absolute growth was similar to the reported in these Seriola species (0.45–0.51 mm TL/day) and in the range of values reported for Thunnus species (0.44–0.68 mm TL/day) according to Kaji, Tanaka, Oka, Takeuchi, Ohsumi, Teruya and Hirokawa (1999). Notochord flexion started at 15 DAH (5.6–6.1 mm TL), similar to S. dumerilli (5.5 ± 0.52 mm TL) (Papandroulakis et al., 2005), being completed at 20 DAH (7.7–8.2 mm TL). This change in larval external morphology may be used as an indication of digestive system maturity, as it occurs synchronously with the appearance of digestive glands, as it was reported by Abreu (2010), suggesting changes in larval behaviour and feeding habits, similar to other species, such as red porgy (Pagrus pagrus) (Roo, Socorro, Izquierdo, Caballero, Hernández-Cruz, Fernández - 71 - & Fernández-Palacios, 1999). Thus, the appearance of gastric glands marked the formation of a functional stomach with a higher capacity to utilize proteins (Govoni, Boehler & Watanabe, 1986) and the capacity to perform a successful early weaning onto dry diets, as it has been found in S. lalandi (Chen et al., 2006). By day 30 (15.0–16.0 mm TL), juveniles had a wet weight ranging from 0.15 to 0.3 g in wet weight with the characteristic striped pigmentation (Plate 1j). At this point, the elevated mortality recorded in the stress test trials can be reduced by the use of anaesthesia based on the use of low clove oil doses, which was associated with 100% larval survival even after a 90 s air exposure. This protocol could be applied in routine manipulations, such as grading or juveniles transportation. Finally, at 90 DAH juveniles from semi-intensive and intensive systems reached 26.7 ± 4.7 and 14.2 ± 5.2 g in wet weight, respectively, denoting the rapid growth of this species. In summary, the results of this first experience of S. rivoliana culture in Europe showed that this species adapts well to captivity conditions and even with dry commercial feeds, responds with good quality spawns to GnRHa hormonal injection, obtaining similar larval survival and growth behaviour as other Seriola species. Moreover, larval survival results and biochemical analysis of eggs and feeds suggests that this species might have higher n-3 HUFA and particularly DHA requirements during larval stages as compared with other commercial species. Further studies are being conducted to improve the culture performance and the study of EFA requirements of S. rivoliana larvae. 5. Acknowledgments This study was funded by Viceconsejería de Pesca del Gobierno de Canarias through the programme “Proyecto piloto para el cultivo de especies de rápido crecimiento en Canarias (RAPCREC). Dr J. Roo, thanks the financial support provided by the Spanish Ministry of Science and Innovation and European Social Funds through the programme “Incorporación de técnicos de apoyo”. Final thanks to Dr Amos Tandler for his kind revision and valuable comments to the manuscript. - 72 - 6. References Abreu N. (2010) Desarrollo ontogénico de las larvas de corvina (Argyrosmus regius) y medregal negro (Seriola rivoliana). Master Thesis, University of Las Palmas de Gran Canaria, Spain 70pp (in Spanish). AOAC (1990) Official Methods of Analysis, 15th edn, vols. I and II. Association of Official Analytical Chemists, Arlington, VA, USA. 1298pp. Attramadal K.J.K., Salvesen I., Xue R., Øie G., Størseth T. R., Vadstein O. & Olsen Y. (2012) Recirculation as a possible microbial control strategy in the production of marine larvae. Aquacultural Engineering 46, 27–39. Benetti D.D. (1997) Spawning and larval husbandry of flounder (Paralichthys woolmani) and Pacific yellowtail (Seriola mazatlana), new candidate species for aquaculture. Aquaculture 155, 307–318. Benetti D.D. (2000) Aquaculture of Pelagic Marine Fish: I. Yellowtail amberjacks (S. quinqueradiata, S. lalandi, S. dumerili and S. mazatlana). Global Aquaculture Advocate 4, 20. Blacio E., Darquea J. & Rodríguez S. (2003) Avances en cultivo de huayaipe, Seriola rivoliana, en las instalaciones del Cenaim. El mundo acuícola 9, 21–24. Castriota L., Greco S., Marino G. & Andaloro F. (2002) First record of Seriola rivoliana Cuvier, 1833 in the Mediterranean. Journal of Fish Biology 60, 486–488. Chambers C.B. & Ernst I. (2005) Dispersal of the skin fluke Benedenia seriolae (Monogenean Capsalidae) by tidal current and implications for sea-cage farming of Seriola sp. Aquaculture 250, 60–90. Chen B.N., Qin J.G., Kumar M.S., Hutchinson W. & Clarke S. (2006) Ontogenetic development of the digestive system in yellowtail kingfish Seriola lalandi larvae. Aquaculture 256, 489–501. Christie W.W. (1982) Lipid Analysis. Pergamon Press, Oxford, 207pp. Duncan N.J., Estevez A. & Mylonas C.C. (2007) Efecto de la inducción hormonal mediante implante o inyección de GnRHa en cantidad y calidad de puestas de corvina (Argyrosomus regius). In: Actas del XI Congreso Nacional de Acuicultura, (ed. by A. Cervinõ, A. Guerra & C. Pérez), pp. 731–732. Vigo, Spain. - 73 - Fernández-Palacios H., Izquierdo M.S., Robaina L., Valencia A., Salhi M. & Vergara J.M. (1995) Effect of n-3 HUFA level in broodstock diets on egg quality of gilthead sea bream (Sparus aurata L.). Aquaculture 132, 325–337. Fischer W., Bianchi G. & Scott W.B. (1981) FAO Species Identification Sheets for Fishery Purposes. Eastern Central Atlantic: Fishing Areas 34, 47, Vol. 1. Canada Fundsin-Trust, Ottawa, Canada. Folch J., Lees M. & Stanley G.H.S. (1957) A simple method for the isolation and purification of total lipids from animal tissues. Journal of Biological Chemistry 226, 497–509. Fox C. (1990) Studies on polyunsaturated fatty acid nutrition in larvae of marine fish the herring, Clupea harengus L. PhD Thesis, University of Stirling, Scotland, UK, 196pp. Govoni J.J., Boehler G.W. & Watanabe Y. (1986) The physiology of digestion in fish larvae. Environmental Biology of Fishes 16, 59–77. Hamasaki K., Tsuruoka K., Teruya K., Hashimoto H., Hamada K., Hotta T. & Mushiake K. (2009) Feeding habits of hatchery-reared larvae of greater amberjack Seriola dumerili. Aquaculture 288, 216–225. Hansen G.H. & Olafsen J.A. (1999) Review article: bacterial interactions in early life stages of marine cold water fish. Microbiology Ecology 38, 1–26. Hilton Z., Poortenaar C. & Sewell M. (2008) Lipid and protein utilisation during early development of yellowtail kingfish (Seriola lalandi). Marine Biology 154, 855–865. Hirayama T., Kawano F. & Hirazawa N. (2009) Effect of Neobenedenia girellae (Monogenea) infection on host amberjack Seriola dumerili (Carangidae). Aquaculture 288, 159–165. IGFA (2001) Database IGFA Angling Records Until 2001. IGFA, Fort Lauderdale, USA. Izquierdo M.S. (1996) Essential fatty acid requirements of cultured marine fish larvae. Aquaculture Nutrition 2, 183–191. Izquierdo M.S. & Koven W. (2011) Lipids. In: Larval Fish Nutrition, (ed. by J. Holt), pp. 47–82. Wiley Blackwell, John Wiley and Sons Publisher, West Sussex, England. - 74 - Izquierdo M.S., Watanabe T., Takeuchi T., Arakawa T. & Kitajima C. (1989) Requirement of larval red sea bream Pagrus major for essential fatty acids. Nippon Suisan Gakkaishi 55, 859–867. Jerez S., Samper M., Santamaría F.J., Villamandos J.E., Cejas J.R. & Felipe B.C. (2006) Natural spawing of greater amberjack (Seriola dumerili) kept in captivity in the Canary Islands. Aquaculture 252, 199–207. Kaji T., Tanaka M., Oka M., Takeuchi H., Ohsumi S., Teruya K. & Hirokawa J. (1999) Growth and morphological developments of laboratory reared yellowfin tuna Thunnus albacares larvae and early juveniles, with special emphasis on the digestive system. Fisheries Science 65, 700–707. Laidley C.W., Shields R.J. & Ostrowksi A.O. (2004) Progress in amberjack culture at the Oceanic Institute in Hawaii. Global Aquaculture Advocate 7, 42–43. Makridis P., Fjellheim A.J., Skjermo J. & Vadstein O. (2000) Colonization of the gut in first feeding turbot by bacterial strains added to the water or bioencapsulated in rotifers. Aquaculture International 8, 367–380. Van der Meeren T. (1991) Selective feeding and prediction of food consumption in turbot larvae (Scophthalmus maximus L.) reared on the rotifer Brachionus plicatilis and natural zooplankton. Aquaculture 93, 35–55. Van der Meeren T., Olsen R.E., Hamre K. & Fyhn H.J. (2008) Biochemical composition of copepods for evaluation of feed quality in production of juvenile marine fish. Aquaculture 274, 375–397. Micale V., Maricchiolo G. & Genovese L. (1999) The reproductive biology of the amberjack, Seriola dumerilii (Risso, 1810): I. Oocyte development in captivity. Aquaculture Research 30, 349–355. Mylonas C., Papandroulakis N., Smboukis A., Papadaki M. & Divanach P. (2004) Induction of spawning of cultured greater amberjack (Seriola dumerili) using GnRHa implants. Aquaculture 237, 141–154. Nakada M. (2002) Yellowtail culture development and solutions for the future. Reviews in Fisheries Science 10, 559–575. Papandroulakis N., Mylonas C., Maingot E. & Divanach P. (2005) First results of greater amberjack (S. dumerili) larval rearing in mesocosm. Aquaculture 250, 155–161. - 75 - Poortenaar C.W., Hooker S.H. & Sharp N. (2001) Assessment of yellowtail kingfish (Seriola lalandi lalandi) reproductive physiology, as a basis for aquaculture development. Aquaculture 201, 271–286. Reid H.I., Treasurer J.W., Adam B. & Birkbeck T.H. (2009) Analysis of bacterial populations in the gut of developing cod larvae and identification of Vibrio logei, Vibrio anguillarum and Vibrio splendidus as pathogens of cod larvae. Aquaculture 288, 36–43. Ricker W.E. (1958) Handbook of computations for biological statistics of fish populations. Bulletin of the Fisheries Research Board of Canada 119, 1–300. Roo J., Socorro J., Izquierdo M.S., Caballero M.J., Hernández-Cruz C.M., Fernández A. & Fernández-Palacios H. (1999) Development of red porgy Pagrus pagrus visual system in relation with changes in the digestive tract and larval feeding habits. Aquaculture 179, 499–512. Roo J., Hernández-Cruz C.M., Socorro J.A., Fernández-Palacios H. & Izquierdo M.S. (2010) Advances in rearing techniques of Pagrus pagrus, (Linnaeus, 1758): comparison between intensive and semi-intensive larval rearing Systems. Aquaculture Research 41, 433– 449. Skjermo J. & Vadstein O. (1999) Techniques for microbial control in the intensive rearing of marine fish larvae. Aquaculture 177, 333–343. Skjermo J., Salvesen I., Øie G., Olsen Y. & Vadstein O. (1997) Microbial maturated water: a technique for selection of a non opportunistic bacterial flora in water that may improve performance of marine larvae. Aquaculture International 5, 13–28. Sokal R.R. & Rolf S.J. (1995) Biometry. In: The Principles and Practice of Statistics in Biological Research, (3rd edn), pp. 419. W.H. Freeman and Company, NewYork, USA. Støttrup J.G. & Norsker N.H. (1997) Production and use of copepods in marine larviculture. Aquaculture 155, 231–247. Tachihara K., El-Zibdeh M.K., Ishimatsu A. & Tagawa M. (1997) Improved seed production of goldstriped amberjack Seriola lalandi under hatchery conditions by injection of triiodothyronine (T3) to broodstock fish. Journal of the World Aquaculture Society 28, 34–44. Verner-Jeffreys D., Shields R.J., Bricknell I.R. & Birkbeck T.H. (2003) Changes in the gut-associated microfloraduring the development of Atlantic halibut (Hippoglossus hippoglossus L.) larvae in three British hatcheries. Aquaculture 219, 21–42. - 76 - Yamamoto T., Teruya K., Hara T., Hokazono H., Hashimoto H., Suzuki N., Iwashita Y., Matsunari H., Fuguita H. & Mushiake K. (2008) Nutritional evaluation of live food organisms and commercial dry feeds used for seed production of amberjack Seriola dumerili. Fisheries Science 74, 1096–1108. - 77 - Skeletal development and mineralization pattern of vertebral column, dorsal, anal and caudal fin complex in Seriola rivoliana (Valenciennes, 1833) larvae. A. Mesa-Rodríguez, C.M. Hernández-Cruz, J.A. Socorro, H. Fernández-Palacios, M.S. Izquierdo and J. Roo. Abstract Bone and fins development in Seriola rivoliana were studied from cleared and stained specimens from 3 to 33 days after hatching. The vertebral column began to mineralize in the neural arches at 4.40 ± 0.14 mm standard length (SL), continued with the haemal arches and centrums following a cranial-caudal direction. Mineralization of the caudal fin structures started with the caudal rays by 5.12 ± 0.11 mm SL, at the same time that the notochord flexion occurs. The first dorsal and anal fin structures were the hard spines (S), and lepidotrichium (R) by 8.01 ± 0.26 mm SL. The metamorphosis was completed by 11.82 ± 0.4 mm SL. Finally, the fin supports (pterygiophores) and the caudal fin were completely mineralized by 16.1 ± 0.89 mm SL. In addition, the meristic data of 23 structures were provided. Results from the present study might be used as a practical guide for future studies on this field with S. rivoliana or in related species. Key words: Amberjack, hatchery, abnormalities, osteology, skeleton. - 85 - 1. Introduction Longfin yellowtail, Seriola rivoliana (Valenciennes, 1833) is one of the species proposed for marine aquaculture diversification, mostly due to its fast growth rate [1,2] and worldwide distribution. This specie belongs to Carangidae family, along with other popular species like Seriola dumerili (greater amberjack), Seriola lalandi (yellowtail king fish) and Seriola quinqueradiata (Japanese yellowtail). Even though S. rivoliana is commercially produced [3], studies about its biology are scarce and only few reports on larval rearing have been conducted in Ecuador [4-6], Hawaii [7] and more recently in the Canary Islands [8]. In contrast, numerous studies of the genus Seriola have been published related to the feeding requirements and nutrition [9-17], reproduction biology [18-21,2] and culture needs [22-26,8]. Regarding osteology studies, previous reports illustrate the bone structure development for other Seriola species. The osteological development of the greater amberjack have been described by different authors [27,28] These authors obtained distinct results probably associated to different environmental conditions and/or the number of samples. Also, the caudal skeleton development of the S. lalandi has been reported [29]. In addition, numerous studies have described the osteological development of other marine finfish species, such as Sparus aurata [30,31], Pagrus pagrus [32,33], Solea senegalensis [34,35], Dentex dentex [36], Argyrosomus regius [37], Epinephelus septemfasciatus [38] or Dicentrarchus labrax [39]. The objective of the present study was to chart the ossification of the vertebral column, dorsal, anal and caudal fin complex in S. rivoliana larvae cultured under semi intensive system conditions (mesocosms, [8]). Larvae culture under this type of system usually performed better than those cultures under intensive conditions [32]. The identification of bony structures and mineralization pattern will serve as a tool for future studies, where different factors (zootechnical, nutritional, environmental parameters, etc) may affect the apparition of skeletal abnormalities. - 86 - 2. Material and methods S. rivoliana eggs were obtained from induced spawning (hormonally injection, GnRHa; SigmaAldrich ™), based on the reported dosage [8]. Larvae were reared under mesocosms rearing system (4.5 eggs.l-1 in two 40m3 tanks [8]), kept under natural photoperiod and filtered natural sea water with 37 g/L salinity and temperature of 23.0 ± 0.9°C. Green water technique was used adding live phytoplankton (Nannochloropsis sp.) to maintain a concentration of 250000 cells ml-1 in the rearing tanks. From 2 to 25 days after hatching (dah) rotifers, Brachionus sp., Lstrain enriched with DHA Protein Selco (INVE ™), were added twice a day (08:00 and 14:00 h). Artemia feeding starts at 15 dah, and were enriched with A1 Easy Selco (INVE ™). Weaning protocol included manual feeding from 20 dah (Genma Micro, Skretting ™) to 25 dah and automatic feeding afterwards. Larval growth was assessed measuring the standard length (SL) of 25 larvae, every 2 days using a profile projector (Nikon V-12A, NIKON™). A total of 75 specimens were individually stained (3.28 ± 0.15 - 16.1 ± 0.89 mm SL). To study the bone ossification, all specimens were fixed in 10% buffered formalin, from hatching to 33dah. Fixed larvae were cleared and stained with alizarin red [40]. Larvae were individually examined using stereomicroscopy. Drawings of the different developmental stages were made using the Adobe Photoshop CS3-10.0 (1990-2007 Adobe System Incorporated, United States) directly from digital photographs. Bone description, followed the terminology suggested by different authors [41-43], and their abbreviations, are illustrated in the Table I. The angles of the spine were measured from the beginning of the vertebral body to the tip of the spine. A total of 10 S. rivoliana reared juveniles were soft X-ray monitored (Mod. Senographer-DHR, General electric’s, USA) for meristic counts. - 87 - 3. Results 3.1. Vertebral column In the present study, S. rivoliana vertebral column mineralization was initiated with the neural arches (Na1-Na3) by 4.40 ± 0.14 mm SL (Fig. 1-A), followed by the haemal arches (Ha1-Ha3) and the cephalic vertebrae (Ce1-4) by 4.74 ± 0.27 mm SL (Fig. 1-B). The ossification of the vertebral column followed a cranial-caudal direction, being totally ossified by 11.82 ± 0.4 mm SL (Fig. 1-G). This size marked the end of metamorphosis. The notochord flexion was initiated at 5.12 ± 0.11 mm SL (Fig. 1-C), at the same time that the caudal complex mineralization was initiated. Initially, urostyle was formed by two independent structures (Ur1-Ur2) that fused by 10.23 ± 0.26 mm SL (Fig. 1-F). The neural spine (Ns23), the haemal spine (Hs13) and the Ce22Ce23 were the last structures that mineralized. At 11.82 ± 0.4 mm SL (Fig. 1G), four types of articulation processes were mineralized: anterior neural zygapophyses (Anz), posterior neural zygapophyses (Pnz), anterior haemal zygapophyses (Ahz) and posterior haemal zygapophyses (Phz), (Fig. 1G*). The vertebral bodies mineralization in the cephalic (without parapophyses) and prehaemal (with parapophyses) region (Ce1-4 and Ce5-10 respectively) proceeded from dorsal to ventral direction and from the surface to internal bone layers (Fig. 2A-B), whereas in the haemal (with Hs) and caudal region (Ns and Hs modified to support caudal fin complex) the mineralization pattern proceeds in both directions dorsal and ventrally (Ce11-21 and Ce22-23 respectively), joining in the middle of the centra (Fig. 2C). Exceptionally, in Ce9 and Ce10, the mineralization of the vertebral bodies differed from other vertebral structures of the prehaemal region, proceeding dorsally first and ventrally later on. The Na and Ha developed from centrum and fused in the middle, forming the neural and haemal canals with rounded shape, later developing into the Ns and Hs (Fig.2A-C). The Ns and Hs angle in relation to the vertebral body varied along the vertebral column, increasing in cranialcaudal direction (Fig. 1G**). The Hs developed according to the angle of the first anal pterygiophore, and decreasing afterward. - 88 - The parapophyses (Pp) was first observed with the mineralization of the Pp6-Pp5 by 5.12 ± 0.11 mm SL (Fig. 1C). These structures had a caudal-cranial development and were fully ossified at 11.82 ± 0.4 mm SL (Fig.1G). The Pp structures become larger form Ce4 to Ce10. The pleural ribs (Plr) were observed for the first time at 10.23 ± 0.26 mm SL (Plr1-Plr3) with the caudal development (Fig.1F). Plr4-Plr7 developed at 11.82 ± 0.4 mm SL (Fig. 1G). The dorsal ribs (Eb) were first seen at 11.82 ± 0.4 mm SL (Fig.1G), with the mineralization of the Eb1Eb3,following the caudal development. 3.2. Dorsal and anal fins development The formation and mineralization of the dorsal and anal fins of the longfin yellowtail followed a cranial-caudal direction. The first dorsal fin structures were the hard spines (S) and lepidotrichium (R) by 8.01 ± 0.26 mm SL (Fig. 1D), which initiated its mineralization from the base to the tip of the structure. Predorsal bones (Pd1-Pd3) and proximal pterygiophore (Pr) had a dorsal-ventral mineralization pattern (Fig. 1F). In the anal fin, the two hard spines (S1-S2) were first seen at 8.01 ± 0.26 mm SL (Fig. 1D), same as the anal lepidotrichia (R) and Pr. The S1-S2 fused into the Pr1 by 9.92 ± 0.84 mm SL (Fig. 1E). The S1-S2 and R mineralized from the base to the tip of the structure, whereas the Pr followed a ventral-dorsal pattern (Fig. 1F). 3.3 Caudal fin development The first caudal complex structures in mineralized were the upper and lower caudal lepidotrichia (PCR) by 5.12 ± 0.11 mm SL (Fig. 3A). Then, hypurals (Hy) initiated their mineralization as fused structures, first Hy1+Hy2, continues with Hy3+Hy4 and finally parhypural (Ph) by 5.38 ± 0.11 mm SL (Fig. 3B). At the same time, the first upper caudal dermatotrichia (SCR) started to mineralize, following a base-tip mineralization pattern. The last hypural (Hy5) delayed its mineralization to 8.01 ± 0.26 SL (Fig.3C). By 9.92 ± 0.84 mm SL uroneurals (Un1+Un2) started to mineralize and fused forming a single structure (Uroneural) (Fig. 3D-E). Finally, the last caudal complex structures in mineralized were the epurals (Ep1-3) by 11.82 ± 0.4 mm SL(Fig. 3E). - 89 - 3.4 Meristic characters Meristically, S. rivoliana had a total number of 23 vertebrae (urostyle not included), 23 neural spines, 13 haemal spines, 16 pleural ribs, 12 dorsal ribs and 6 parapophyses. In the dorsal region, 3 predorsal spines, VII+I hard spines and a variable number (30-34) lepidotrichia were identified. Besides, VII hard spine proximal pterygiophores, 30 to 34 distal radial and proximal pterygiophores were also observed. Within the anal region, II+I hard spines and 19-21 lepidotrichia, same number of distal radial and proximal pterygiophores were identified. Finally, in the caudal complex, 1 parahypural, 5 hypurals, 3 epurals, 2 uroneurals, 10+9 caudal lepidotrichia and 10+9 caudad dermatotrichia were observed (Table II). 4. Discussion This study reported for first time S. rivoliana skeletal development and mineralization. The comparison of present results with other species from the same family and genus, such as S. dumerili, suggest some correspondence. Thus, [27] describes first mineralized structure in the vertebra centra for S. dumerilli at 6.6 mm (NL) while other result for the same specie [28] identified neural spine (Ns) and centra at 4.8mm (TL). This pattern agrees with present data for S. rivoliana, where similar mineralization was obtained (4.74 ± 0.27 mm SL). The differences between vertebra centra mineralization timing for S. dumerili could be explained by the different environmental conditions, such as temperature, or rearing protocols applied in those studies [32]. In fact, mineralization pattern is more accurately described when larval growth is used as reference instead of larval age [33]. Also, the present study showed a similar vertebra centra mineralization timing in comparison with other carangid species, such as Caranx crysos [44] and Selene setapinnis [45], , suggesting that some developmental events during mineralization process may be common for many species. For instance, the dorsal flexion at the posterior end of the notochord could be an external indicator of the initiation of the internal column mineralization for this and other species. In fact, these events also occur in other species such as S. aurata larvae (5.7 – 6.0 mm, SL) [30], Solea senegalensis larvae (4.7 mm, SL) [35], Pagrus pagrus larvae (6.0 ± 0.5 mm, TL) [32] or Argyrosomus regius larvae (5.42 – 6.01 mm, TL) [47]. - 90 - In most Perciforms, the vertebral column follows a bidirectional mineralization pattern (Pagrus major, [42]; S. aurata, [30]; Dentex dentex, [36]; Diplodus sargus, [47]; Pagrus pagrus, [32]). However, in S. rivoliana the vertebral column followed a unidirectional mineralization pattern, in agreement with data reported in S. dumerili [28] and in A. regius [46]. According to the centrum mineral deposition, three complementary models occur in the vertebral region: in a dorsal-ventral direction (D-V), in a ventral-dorsal direction (V-D) or simultaneously (DVS). In S. aurata [30] and D. sargus [47], two centra mineral deposition models occur, the first one takes place in a D-V direction from the Ce1 to Ce21 and the second one in a V-D direction in Ce22 and Ce23. In S. rivoliana, the mineralization expands in a D-V direction from Ce1 to Ce23, following the same pattern as S. dumerili between the Ce1-Ce19 [28]. Additionally, S. rivoliana had simultaneously DVS mineralization from Ce8 to Ce23, whereas S. dumerili [28] presents this simultaneous DVS mineralization pattern from Ce20 to Ce23. Other marine finfish such as A. regius [47] showed a D-V mineral deposition from Ce1-Ce4, while the remaining vertebrae had simultaneous DVS mineralization. About to urostyle (Ur) structure of S. rivoliana larvae and other marine finfish such as P. major, C. crysos and S. dumerili [48, 44, 28], this was formed by the fusion of two elements (Ur1+Ur2). In contrast, at least three elements were necessary to form this structure in S. lalandi [29]. The results of the present study suggest that the fusion of different structures to form the urostyle is nonspecific of the genus Seriola sp. The development of the parapophyses (Pp) of S. rivoliana followed a caudal-cranial development, in concordance with S. dumerili [28] and many other perciforms such as S. aurata [30], Lates calcarifer [49], Diplodus sargus [47] or Pagrus pagrus [32]. The correlation between the present study and many other marine finfish suggest that the developmental patter for the parapophyses may be common in perciforms. In many marine finfish species, the mineralization of the anal and posterior dorsal fins starts prior to the anterior dorsal fin [31, 32, 50-52]. Unlike this developmental pattern, but in accordance with S. dumerili [28], S. rivoliana dorsal and anal structures followed a cranialcaudal development, developing the anterior dorsal fin prior to posterior dorsal fin. However, - 91 - despite S. rivoliana and S. dumerili had the same developmental pattern in dorsal, anal and caudal fins; some differences in structures development have been observed. For instance, in S. rivoliana, firsts structures in mineralized were hard spines (S) and lepidotrichium (R) (present study), whereas in S. dumerili [28] the dorsal fins development starts with the mineralization of the proximal pterygiophore. During the process of the caudal complex mineralization of S. rivoliana, the fusion of hypurals (Hy1+Hy2 and Hy3+Hy4) was observed. This developmental pattern is common in other carangid species such as S. lalandi [29], S. setapinnis [47], C. crysos [44] and S. dumerili [27, 28]; as well as other perciforms such us Coryphaena equiselis [53], P. major [42], S. aurata [30] and D. dentex [36]. The development of three distinct structures (Hy1+Hy2, Hy3+Hy4, Hy5) could remain as a characteristic of carangids and Coryphaena [44]. In the present study, the development of three epurals and two uroneurals were observed. The number of epurals in the caudal complex of Carangoidei species varies between species[44]: 3 epurals in S. dumerili [28], between 3-4 epurals (usually 3 epurals) in S. lalandi [29], 2 independent epurals that fused during ontogeny in C. equiselis [53] and 3 epurals for S. rivoliana (present study). Other authors [54] considered that the presence of uroneurals is a characteristic of the Teleost. The presence of two uroneurals in S. rivoliana caudal fin complex is in concordance with other species from the same genus such as S. lalandi [29] and S. dumerili [28]. Meristically, the vertebral column of longfin yellowtail (S. rivoliana) was characterized in this study. Similar results have been reported in S. dumerili [28]. Nevertheless, in other carangid species such as S. setapinnis [45], Hemicaranx amblyrhynchus [55] or Trachurus japonicas [43] a total number of 24 vertebral structures were observed, and the first haemal arch was observed at the 10th vertebra [45] instead of at the 11th vertebrae in S. dumerili [27, 28] and S. rivoliana (present study), indicating that this could be a conserved feature among the genus Seriola (Table II). Concerning the caudal complex, S. rivoliana presented similar results than those observed in S. lalandi [29], S. setapinnis [45], C. crysos [44] and S. dumerili [27, 28], although the number of - 92 - caudal fin rays is a characteristic for each species. Thus, in this study, S. rivoliana had 10+9 caudal lepidotrichia and 10+9 caudad dermatotrichia, while 9+9 caudal lepidotrichia and 1113+10 caudal dermatotrichia where reported in S. dumerili [28] or 9+8 caudal lepidotrichia and 6+5 caudad dermatotrichia were observed in S. setapinnis [45]. The importance of the meristic characterization is widely known for the identification not only for marine finfish species, [56, 57] but also in cultured fish species [52, 58, 59]. Results from the present study might be used as practical guide for future studies on this field with S. rivoliana or in related species. 5. Acknowledgements This study was funded by the Agencia Canaria de Investigación, Innovación y Sociedad de la Información (ACIISI) and Fondo Europeo de Desarrollo Regional (FEDER) through the program “Mejora de las técnicas de cría de larvas de (Seriola rivoliana): Determinación de requerimientos de ácidos grasos esenciales en su etapa larvaria y optimización de la secuencia alimentaria (METCSERProID20100094)”. 6. References 1. Mazzola, A., Favaloro, E., Sará, G. (2000). Cultivation of the Mediterranean amberjack, Seriola dumerili (Risso, 1810), in submerged cages in the Western Mediterranean Sea. Aquaculture 181, 257-268. 2. Jerez, S., Samper, M., Santamaría, F J., Villamandos, J.E., Cejas, J.R., et al. (2006). Natural spawing of greater amberjack (Seriola dumerili) kept in captivity in the Canary Islands. Aquaculture 252, 199-207. 3. Sims, N.A., Key, G. (2011). Fish without footprints. OCEANS´11-MTS/IEEE Kona, Program Book. Article number 6107321. - 93 - 59. Boglione C, Costa C, Di Dato P, Ferzini G, Scardi M, et al. (2003). Skeletal quality assessment of reared and wild sharpsnout sea bream and pandora juveniles. Aquaculture 227:373–394. 60. Bañón, R. & Mucientes, G.R. (2009). First record of Seriola fasciata (Carangidae) from Galician waters (NW Spain). A new northernmost occurrence in the NE Atlantic. Cybium 2009, 33(3): 247-248. 61. Andaloro, F., Falautano, M., Sinopoli, M., Passarelli, F.M., Pipitone, C., et al. (2005). The lesser amberjack Seriola fasciata (Perciformes: Carangidae) in the Mediterranean: A recent colonist?. Cybium 29:141–145. 62. Castriota, L., Falautano, M., Greco, S., Andaloro, F. (2004). Second record of Seriola rivoliana (Carangidae) in the Mediterranean. Cybium 28:265–266. 63. Castriota, L., Greco, S., Marino, G., Andaloro, F. (2002). First record of Seriola rivoliana Cuvier, 1833 (Osteichthyes: Carangidae) in the Mediterranean. Journal of Fish Biology 60:486– 488. - 100 - 7. List of tables and figures. Table I. Skeletal elements and their abbreviations. Region Skeletal elements Abbreviations Vertebral Column Vertebra centra Ce Notochord No Urostyle Ur Neural Arch Na Neural Spine Ns Haemal Arch Ha Haemal Spine Hs Dorsal Ribs Eb Pleural Ribs Plr Parapophyses Pp Anterior neural zygapophysis Anz Posterior neural zygapophysis Pnz Anterior haemal zygapophysis Ahz Posterior haemal zygapophysis Phz Caudal Fin Hypurals Hy Parhypural Ph Epurals Ep Uroneurals Un Caudal lepidotrichia PCR Caudal Dermatotrichia SCR Dorsal Fin Predorsal Pd Hard Spines S Lepidotrichium R Proximal Pterygiophores Pr Distal Radial Dr Anal Fin Hard Spines S Lepidotrichium R Proximal Pterygiophores Pr Distal Radial Dr - 101 - Table II. Meristic counts in different carangid species. ( - ) no data; ( + ) and; ( / ) between. Species Regions References Vertebral column Caudal fin Dorsal Fin Anal Fin Vertebra centra Urostyle Neural spine Haemal spine Dorsal Ribs Pleural Ribs Parapophyses Hypurals Parhypural Epurals Uroneurals Lepidotrichia Dermatotrichia Predorsal Hard Spines Proximal Pterygiophores Lepidotrichia Proximal Radial Distal Radial Hard Spines Lepidotrichia Proximal Radial Distal Radial S. rivoliana 10+13 1+1 23 13 6+6 8+8 6 5 1 3 1+1 10+9 10+9 3 VII+I 7 30 / 34 30 / 34 30 / 34 II+I 19 / 21 19 / 22 19 / 22 Present Study S. dumerili 10+13 1+1 23 13 6+6 8+8 6 5 1 3 1+1 9+9 12+10 3 VIIVIII 7-8 31 / 34 35 32 III 19 / 21 20 / 23 20 / 23 Laggis et al.,2010 S. fasciata - - - - - - - - - - - - - - VIII+I - 28 / 31 - - II+I 18 / 20 - - Bañón & Mucientes, 2009 C. crysos - 1+1 - - - - - 5 1 3 1+1 - - - - - - - - - - - - Hilton & Johnson, 2007 S. fasciata - - - - - - - - - - - - - - VIII+I - 28 / 31 - - II+I 18 / 20 - - Andaloro et al., 2005 S. rivoliana - - - - - - - - - - - - - - VII+I - 30 - - II+I 21 - - Castriota et al., 2004 S. rivoliana - - - - - - - - - - - - - - VII+I - 32 - - II+I 22 - - Castriota et al., 2002 S. dumerili 10+13 1 22 13 - 8+8 - 5 1 3 1+1 9+7 12+11 3 VIIVIII 31 / 32 38 - IIIII 20 21 - Liu, 2001 S. lalandi - 1+1+1 - - - - - 5 1 3-4 1+1 - - - - - - - - - - - - Kohno, 1997 S. setapinnis 10+14 - 24 14 - - - 5 1 - - 9+8 5+6 - - - - - - - - - - Katsuragawa, 1997 H. amblyrhynchus 10+14 - 24 14 - - - - - - - 9+9 8+8 - VII+I - 28 - - II+I 25 - - Flores-Coto et al., 1998 T. japonicus 10+14 1 24 14 7+7 10+10 6 5 1 2 1 17 - 3 8 8 27 / 35 27 / 35 27 / 35 II+I 25 / 31 25 / 31 25 / 31 Suda, 1996 - 102 - Figure 1. Development of the vertebral column (A - G) in S. rivoliana (painted areas, mineralized structures). Structures: Ahz, anterior haemal zygapophysis; Anz, anterior neural zygapophysis; Ha, haemal arch; Hs, haemal spine; No, notochord; Na, neural arch; Ns, neural spine; Pp, parapophyse; Plr, pleural rib; Pd, predorsal; Ph, parhypural; Phz, posterior haemal zygapophysis; Pnz, posterior neural zygapophysis; Pr, proximal pterygiophore; R, lepidotrichium; Dr, distal radial; S, hard spine; Ce, vertebral centra; Ur, urostyle. - 103 - Figure 2. Vertebra mineralization (A - C) (painted areas, mineralized structures). Ha, haemal arch; Hs, haemal spine; Na, neural arch; Ns, neural spine; Pp, parapophyses. - 104 - Figure 3. Development of S. rivoliana caudal complex (A - E) (painted areas, mineralized structures). Structures: Ep, epurals; Hy, hypurals; PCR, caudal lepidotrichia; Ph, parhypural; SCR, caudal dermatotrichia; Ur, urostyle; Un, uroneural. - 105 - Study III Bone development of the skull, pectoral and pelvic fins in Seriola rivoliana (Valenciennes, 1833) larvae. Published in: Fish Physiology & Biochemistry, 2016. DOI: 10.1007/s10695-016-0257-8 - 107 - Bone development of the skull, pectoral and pelvic fins in Seriola rivoliana (Valenciennes, 1833) larvae. A. Mesa-Rodríguez, C.M. Hernández-Cruz, M. B. Betancor, H. Fernández-Palacios, M.S. Izquierdo and J. Roo. Abstract Skull, pectoral and pelvic fins bone structures in longfin yellowtail Seriola rivoliana were studied from 3.43 ± 0.15 to 16.20 ± 0.73 mm standard length (SL) specimens. The S. rivoliana skull started to mineralize with the appearance of the parasphenoid and maxillary by 3.43 ± 0.15 mm SL at the neurocranium and jaw regions respectively. The first pectoral structure to mineralize was the cleithrum at 3.75 ± 0.14 mm SL, shortly followed by supracleithrum and posttemporal. The pelvic fin started by 6.16 ± 0.32 mm SL with the spine and continued with the soft rays and basipterygium. The present study determined the onset of the skull, pectoral and pelvic fins mineralization. These results might be used as a reference for future studies in S. rivoliana or related species. Key words: Amberjack, hatchery, abnormalities, osteology, skeleton. - 109 - Chloroscombrus orqueta, Caranx caballus, Caranx sexfasciatus (Sumida et al. 1985); Selene setapinnis, Katsuragawa (1997), Hemicaranx amblyrhynchus (Flores-Coto et al. 1998); Parastromateus niger (Hilton et al. 2010) was not identified nor in S. rivoliana (present study) neither in similar species such as S. dumerili (Liu 2001). This type of skeletal element together with opercular spines has been identified as protective structures (Morgan 1989), and might greatly differ between species particularly related to aspects such as type of environment were the larvae growth (estuarine, open ocean) or predators pressure and need of predator avoidance . Concerning the ossification timing, even within the same species or genus many differences could be addressed. Thus, Liu (2001) described the start of the neurocranium development at 6.6 mm (LN) with the mineralization of F, Inf, Ps and V simultaneously in S. dumerili, whereas S. rivoliana (present study) began neurocranium mineralization by 3.75 ± 0.14 mm (SL) starting with Ps, and followed by V, F and Exo at 5.38 ± 0.11 mm (SL). Besides, at 16.8 mm (LN) the Sph, Pto, Eoc, Pro, Soc, Exo and Bo were the last structures to mineralize in S. dumerili (Liu, 2001), whereas in S. rivoliana larvae the E was the last structures that started to develop by 12.41 ± 0.43 mm (SL). These developmental patterns and mineralization timing are clearly different between S. dumerili (Liu 2001) and S. rivoliana (present study). In this regard, not only the species but also the environmental conditions (Boglione et al. 2009; Cobcroft et al. 2012), and zootechniques applied (Roo et al. 2010a,b) might delay or advanced the larval ontogeny, promoting differences in the larval stages of stained specimens that would explain these differences. The ontogeny and mineralization of feeding and breathing structures should be combined with those skeletal elements involved in swimming activity, related to prey capture and escape abilities. Thus S. rivoliana larvae initiated the ossification of the dorsal and anal fin at 8.01 ± 0.26 mm SL (Mesa-Rodriguez, et al. 2014), and were shortly followed by the pelvic fins at 8.60 ± 0.66 mm SL during the notocord post-flexion stage. This developmental pattern occur after the mineralization of the main jaw and breathing structures between 3.75 ± 0.14 mm SL and 6.16 ± 0.73 mm SL, supporting the importance of feeding and breathing. Moreover, bones - 116 - involved in main functions such as feeding and breathing have been conserved among many different Teleost (Danio rerio, Cubbage & Mabee, 1996; Betta splendens, Mabee & Trendler, 1996; Gadus morhua and Pseudopleuronectes americanus, Hunt von Herbing, 2001; Solea solea, Wagemans & Vandewalle, 2001; Epinephelus septemfasciatus, Nagano et al. 2007). The onset of similar ossification pattern in swimming structures such as the dorsal, anal and pelvic fins between S. dumerili (Laggis et al. 2010), S. lalandi (Sumida et al. 1985) and S. rivoliana (Mesa-Rodriguez, et al. 2014) at post-flexion stage suggest that this characteristic is common for the Seriola genus. Therefore, identification of pre-flexion and post-flexion stages are useful external indicators of the initiation of the internal column mineralization (Mesa-Rodriguez, et al. 2014) and fins. This fact improves larval swimming ability to capture prey and predator avoidance. In addition, notochord flexion is paired with the appearance of the first gastric glands and new photoreceptors in the eye (rods) providing a higher digestive capacity and improving visual acuity respectively (Roo, personal com.), which is probably related with changes in feeding habits and environmental needs, as happens with other species (Roo et al. 1999). Furthermore, notochord flexion was also an indicator to distinguish different genus from the carangid family during the formation of pelvic fin elements (Laroche et al. l984; Ahlstrom & Sumida 1985). For example, the presence of single Scl and Pt spines in the pectoral fin support of S. rivoliana is in concordance with that found for S. setapinnis, S. vommer (Katsuragawa 1997) and C. sexfasciatus (Sumida et al. 1985), but differs from Trachurus lathami, Decapterus punctatus, Chloroscombrus chrysurus (Katsuragawa 1990), C. caballus, C. orqueta and S. lalandi (Sumida et al. 1985), in which the number of spines of each structure varies. According to the presence and number of spines, there is no tendency between carangid species. In this regard, obtention of meristic data of S. rivoliana are also relevant to identify deviation from the regular pattern under culture conditions. In fact, meristic counts are considered quality descriptors of reared juveniles (Boglione et al. 2001). Generally, cultured juveniles display a higher variability of meristic characters than natural populations. In this study, S. rivoliana were cultured under semi-intensive system conditions, a condition where juvenile finfish were previously reported to be more prone to present similar characteristics to wild juveniles (Roo et - 117 - al. 2009). The meristic data obtained for S. rivoliana identified a number of 7 paired Bsr, similar to other carangid species such as S. dumerili (Liu 2010) and P. niger (Hilton et al. 2010), and in common with A. davidsonii and X. californiensis (Watson & Walker 1992). Also, the number of spines in the pelvic fin (I+5) and soft rays in the pectoral fin (20-23), are common in several carangids such as S. setapinnis (Katsuragawa 1997), S. dumerili (Liu 2001; Laggis et al. 2010), P. niger (Hilton et al. 2010), S. lalandi (Sanchez-Ramirez & Flores-Coto 1993), S. rivoliana (Castriota et al. 2004; present study), Seriola fasciata (Bañón & Mucientes 2009), C. caballus, C. orqueta and C. sexfasciatus (Sumida et al. 1985). Whereas for many other marine finfish the soft rays from the number of pectoral fin varies from 14-22 (S. emarginatum, Beckley 1989; A. davidsonii and X. californiensis, Watson & Walker 1992; Dicentrarchus labrax, Marino et al. 1993; D. dentex, Koumoundouros et al. 2000; S. aurata, Faustino & Power 2001; D. sargus, Koumoundouros et al. 2001; P. erythrinus, Sfakianakis et al. 2004). These results may indicate that the total number of 20-23 pectoral fin soft rays is characteristic from the carangid fish. Results from the present study determined that the onset of structures ossification is directly related with the main functionalities of the different structures, suggesting that some developmental events during mineralization process may be common for many different species. In addition, the first description of meristic data and mineralization timing for this species might be a useful practical guide for future studies on this field with Seriola genus or related species. 5. Acknowledgements This study was funded by the Agencia Canaria de Investigación, Innovación y Sociedad de la Información (ACIISI) and Fondo Europeo de Desarrollo Regional (FEDER) through the program “Mejora de las técnicas de cría de larvas de (Seriola rivoliana): Determinación de requerimientos de ácidos grasos esenciales en su etapa larvaria y optimización de la secuencia alimentaria (METCSERProID20100094)”. - 118 - 6. References Ahlstrom EH & Sumida BY(1985) Early life history studies of eastern Pacific carangids. NOAA, NMFS, SFC, Admin. Rep. LJ - 85-02.25 p. Bañón R & Mucientes GR (2009) First record of Seriola fasciata (Carangidae) from Galician waters (NW Spain). A new northernmost occurrence in the NE Atlantic. Cybium 2009, 33(3): 247-248. Beckley LE (1989) Larval development of Spondyliosoma emarginatum (euvier & Valenciennes) (Pisces: Sparidae) from southern Africa. S. Afr. 1. Zool. 1989, 24(3) Blacio E, Darquea J & Rodríguez S (2003) Avances en cultivo de huayaipe, Seriola rivoliana, en las instalaciones del Cenaim. El mundo acuícola 9,21–24. Blacio EM (2004) Outdoor tank culture of almaco jacks in Ecuador. Global Aquaculture Advocate, April 2004. Boglione C, Gagliardi F, Scardi M & Cataudella S (2001) Skeletal descriptors and quality assessment in larvae and post-larvae of wild-caught and hatchery-reared gilthead sea bream (Sparus aurata L. 1758). Aquaculture. 192, 1-22. Boglione C, Costa C, Di Dato P, Ferzini G, Scardi M & Cataudella S (2003) Skeletal quality assesment of reared and wild sharp snout sea bream and Pandora juveniles. Aquaculture 227, 373-394. Boglione C, Marino G, Giganti M, Longobardi A & De Marzi P (2009) Skeletal anomalies in dusky grouper Epinephelus marginatus (Lowe 1834) juveniles reared with different methodologies and larval densities. Aquaculture 291, 48–60. Castriota L, Falautano M, Greco S & Andaloro F (2004) Second record of Seriola rivoliana (carangidae) in the Mediterranean. Cybium 2004, 28(3): 265-266. - 119 - Cobcroft JM & Battaglene SC (2013) Skeletal malformations in Autralian marine finfish hatcheries. Aquaculture. Vol: 396-399, 1 June 2013, Pages: 51-58. Cobcroft JM, Pankhurst PM, Poortenaar C, Hickman B & Tait M (2004) Jaw malformation in cultured yellowtail kinfish (Seriola lalandi) larvae. New Zealand Journal of Marine and Freshwater Research. Vol: 38, Issue 1, March 2004, Pages: 67-71. Cobcroft JM, Shu-Chien AC, Kuah MK, Jaya-Ram A & Battaglene SC (2012) The effects of tank colour, live food enrichment and greenwater on the early onset of jaw malformation in striped trumpeter larvae. Aquaculture 356–357 (2012) 61–72. Collette BB & Gillis GB (1992) Morphology, systematics, and biology of the double-lined mackerels (Grammatorcynus, Scombridae). Fishery Bulletin, U.S. 90, 13–53. Cubbage CC & Mabee PM (1996) Development of the Cranium and Paired Fins in the Zebrafish Danio rerio (Ostariophysi, Cyprinidae). Journal of Morphology 229: 121-160. Faustino M & Power DM (1999) Development of the pectoral, pelvic, dorsal and anal fins in cultured sea bream. Journal of Fish Biology 54, 1094 –1110. Faustino M & Power DM (2001) Osteologic development of the viscerocranial skeleton in sea bream: alternative ossification strategies in teleost fish. Journal of Fish Biology (2001) 58, 537– 572. Fernández-Palacios H, Schuchardt D, Roo J, Hernández-Cruz CM, Izquierdo M (2015) Spawn quality and GnRHa induction effiency in longfin yellowtail (Seriola rivoliana) broodstock kept in captivity. Aquaculture. Vol: 435, 167-172. Flores-Coto C, Sánchez-Ramírez M, Zavala-García F (1998) Morfología externa del desarrollo larvario de Hemicaranx amblyrhynchus (Pisces: Carangidae) del sur del Golfo de México. Rev biol trop 46. - 120 - Garcia-Gomez A (2000) Recent advances in nutritional aspects of Seriola dumerili. Paper presented at: Recent advances in Mediterranean aquaculture finfish species diversification. Zaragoza: CIHEAM-IAMZ, 2000. pp. 249–257: 6 graphs, 1 table, bibliography pp. 254–257, summaries (En, Fr) (Cahiers Options Méditerranéennes; v. 47). Seminar of the CIHEAM Network on Technology of Aquaculture in the Mediterranean on ‘Recent advances in Mediterranean aquaculture finfish species diversification’, 1999/05/24–28, Zaragoza (Spain). Gavaia PJ, Dinis MT, Cancela ML (2002) Osteological development and abnormalities of the vertebral column and caudal skeleton in larval and juvenile stages of hatchery-reared Senegal sole (Solea senegalensis). Aquaculture. Vol: 211, Issue 1-4, 23 August 2002, Pages: 305-323. Hamasaki K, Tsuruoka K, Teruya K, Hashimoto H, Hamada K (2009) Feeding habits of hatchery-reared larvae of greater amberjack Seriola dumerili. Aquaculture 288 (2009) 216–225. Hattori M, Sawada Y, Takagi Y, Suzuki R, Okada T & Kumai H (2003) Vertebral deformities in cultured red sea bream, Pagrus major, Temminck and Schlegel. Aquac. Res. 34, 1129-1137. Hilton EJ, Johnson GD & Smith-Vaniz WF (2010) Osteology and Systematics of Parastromateus niger (Perciformes: Carangidae), with Comments on the Carangid Dorsal GillArch Skeleton. Copeia, Vol. 2010, No. 2 (May 20, 2010), pp. 312-333. Hunt von Herbing I (2001) Development of feeding structures in larval fish with different life histories: winter flounder and Atlantic cod. Journal of Fish Biology (2001) 59, 767–782. Izquierdo MS, Socorro J & Roo J (2010) Studies on the appearance of skeletal anomalies in red porgy: effect of culture: effect of culture intensiveness, feeding habits and nutritional quality of live preys. J. Appl. Ichthyol. 26 (2010), 320–326. - 121 - Jerez S, Samper M, Santamaría F J, Villamandos JE, Cejas JR (2006) Natural spawing of greater amberjack (Seriola dumerili) kept in captivity in the Canary Islands. Aquaculture 252, 199-207. Katsuragawa M (1990) Estudo sobre o desenvolvimento, a distribuição e a abundância de larvas de carangídeos da costa sudeste do Brasil. Ph.D. Dissertation, University of São Paulo, Oceanographic Institute. 115 p. Katsuragawa M (1997) Larval development of the Atlantic Moonfish Selene setapinnis (Osteichthyses, Carangidae) from southeastern Brazil. Bulletin of Marine Science, 61(3): 779789. Kohno H (1997) Osteological development of the caudal skeleton in the carangid, Seriola lalandi. Ichthyological Research 44, 219-221. Koumoundouros G, Divanach P & Kentouri M (2000) Development of the skull in Dentex dentex (Osteichthyes: Sparidae). Marine Biology (2000) 136: 175±184. Koumoundouros G, Sfakianakis DG, Maingot E, Divanach P, Kentouri M (2001) Osteological development of the vertebral column and of the fins in Diplodus sargus (Teleostei: Perciformes: Sparidae). Marine Biology (2001). 139: 853-862. Kozul V, Skaramuca B, Glamuzina B, Clavic N, Tutman P (2001) Comparative gonadogenesis and hormonal induction of spawning of cultured and wild mediterranean amberjack (Seriola dumerili Risso 1810). Scientia Marina 65:215–220. Laggis A, Sfakianakis DG, Divanach P, Kentouri M (2010) Ontogeny of the body skeleton in Seriola dumerili (Riso, 1810). Italian Journal of Zoology 77(3): 303-315. Laidley CW, Shields RJ, Ostrowksi AO (2004) Progress in amberjack culture at the Oceanic Institute in Hawaii, Glob. Aqua. Advoc. 7(1): 42-42. Laroche WA, Smith-Vaniz WF & Richardson SL (1984) Carangidae: development. In H.G. Moser, W.J. Richards, D.M. Cohen, M.P. Fahay, A.W. Kendall, Jr., and S.L. Richardson - 122 - (eds.), Ontogeny and systematics of fishes. Spec. Publ. No. I. Amer. SOC. Ichthyol. Herpetol., p. 510-522. Liu CH (2001) Early osteologycal development of the yellow tail Seriola dumerili (Pisces: Carangidae). Zoological Studies 40, 289-298. Mabee PM & Trendler TA (1996) Development of the cranium and paired fins in Betta Splendens (Teleostei: Percomorpha): Intraspecific variation and interspecific comparisons. J. Morphol. 227: 249-287. Marino G, Boglione C, Bertolini B, Rossi A, Ferreri F, Cataudella S (1993) Observations on development and anomalies in the appendicular skeleton of sea bass, Dicentrarchus labrax L. 1758,larvae and juveniles. Aquaculture and fisheries Management, 1993, 24, 445-456. Marino G, Mandich A, Massari A, Andaloro F, Porrello S (1995a) Aspects of reproductive biology of the Mediterranean amberjack (Seriola dumerilii Risso, 1810): Gonadal development. Paper presented at: Marine aquaculture finfish species diversification/Diversification des espèces de poissons en aquaculture marine. Zaragoza: CIHEAM-IAMZ, 1995. pp. 115–124: 7 ill., 19 ref (Cahiers Options Méditerranéennes; v. 16), Seminar of the CIHEAM Network on Technology of Aquaculture in the Mediterranean (TECAM), 14–17 June 1995, Nicosia Cyprus. Marino G, Mandich A, Massari A, Andaloro F, Porrello S (1995b) Aspects of reproductive biology of the Mediterranean amberjack (Seriola dumerilii Risso) during spawning period. Journal of Applied Ichthyology 11:9–24. Matsuoka M (1985) Osteological development in the Red Sea Bream, Pagrus major. Japanese Journal of Ichthyology. Vol. 32. N,1. McKenney TW, Alexander EC and Voss GL (1958) Early development and larval distribution of the Carangid fish, Caranx crysos (Mitchill). Bulletin of Marine Science of the Gulf and Caribbean [8(2),1958]. - 123 - Mesa-Rodríguez A, Hernández-Cruz CM, Socorro JA, Fernández-Palacios H, Izquierdo MS & Roo J (2014) Skeletal Development and Mineralization Pattern of the Vertebral Column, Dorsal, Anal and Caudal Fin Complex in Seriola Rivoliana (Valenciennes, 1833) Larvae. J Aquac Res Development 2014, 5:6. ISSN: 2155-9546 JARD, an open access journal. Morgan SG (1989) Adaptive significance of spination in estuarine crab zoeae. Ecology 70: 462482. Mylonas CC, Papandroulakis N, Smboukis A, Papadaki M, Divanach P (2004) Induction of spawning of cultured greater amberjack (Seriola dumerili) using GnRHa implants. Aquaculture 237:141–154. Nagano N, Hozawa1 A, Fujiki W, Yamada T, Miyaki K, Sakakura Y & Hagiwara A (2007) Skeletal development and deformities in cultured larval and juvenile seven-band grouper, Epinephelus septemfasciatus (Thunberg). Aquaculture Research, 2007, 38, 121-130. Papadakis IE, Chatzifotis S, Divanach P, Kentouri M (2007) Weaning of greater amberjack (Seriola dumerilii Risso 1810) juveniles from moist to dry pellet. Aquaculture International 16:13–25. Papandroulakis N, Mylonas CC, Maingot E, Divanach P (2005) First results of greater amberjack (Seriola dumerili) larval rearing in mesocosm. Aquaculture 250:151–161. Roo J, Socorro J, Izquierdo MS, Caballero MJ, Hernández-Cruz CM, Fernández A & Fernández-Palacios H (1999) Development of red porgy Pagrus pagrus visual system in relation with changes in the digestive tract and larval feeding habits. Aquaculture 179,499–512. Roo FJ, Hernández-Cruz CM, Fernández-Palacios H & Izquierdo MS (2005) Development of skeletal deformities in gilthead sea bream (Sparus aurata) reared under different larval culture and dietary conditions. Larvi’05FISH & SHELLFISH LARVICULTURE SYMPOSIUM. C.I. Hendry, G. Van Stappen, M. Wille and P. Sorgeloos (Eds). European Aquaculture Society, Special Publication No. 36, Oostende, Belgium, 2005. - 124 - Roo FJ, Hernandez-Cruz CM, Socorro JA, Fernandez-Palacios H, Montero D, Izquierdo MS (2009) Effect of DHA content in rotifers on the occurrence of skeletal deformities in red porgy Pagrus pagrus (Linnaeus, 1758). Aquaculture 287, 84–93. Roo J, Socorro J, Izquierdo MS (2010a) Effect of rearing techniques on skeletal deformities and osteological development in red porgy Pagrus pagrus (Linnaeus, 1758) larvae. J. Appl. Ichthyol. 26: 372-376. Roo FJ, Hernández-Cruz CM, Socorro JA, Fernández-Palacios H, Izquierdo MS (2010b) Occurrence of skeletal deformities and osteological development in red porgy Pagrus pagrus larvae cultured under different rearing techniques. Journal of Fish Biology. Vol: 77, Issue 6, October 2010, Pages: 1309-1324. Roo J, Fernández-Palacios H, Hernández-Cruz CM, Mesa-Rodriguez A, Schuchardt D & Izquierdo M (2012) First results of spawning and larval rearing of longfin yellowtail Seriola rivoliana as a fast-growing candidate for European marine finfish aquaculture diversification. Aquaculture Research, 2012, 1–12. Sánchez-Ramírez M & Flores-Coto C (1993) Larval development and identification key of some species of the carangidae (Pisces) family of southern Dulf of Mexico. Contribución No. 736 del Instituto de Ciencias del Mar y Limnología. Sims NA & Key G (2011) Fish without footprints. OCEANS´11-MTS/IEEE Kona, Program Book. Article number 6107321. Sfakianakis DG, Koumoundouros G, Divanach P, Kentouri M (2004) Osteological development of the vertebral column and of the fins in Pagellus erythrinus (L. 1758). Temperature effect on the developmental plasticity and morpho-anatomical abnormalities. Aquaculture 232 (2004), 407-424. Suda Y (1996) Osteology and muscular attachments of the Japanese jack mackerel, Trachurus japonicus. Bulletin of Marine Science 58, 438–493. - 125 -