Horizontal hydroacoustics for the Estimation of fish Parameters in sea Bass (Dicentrarchus labrax) farming Facilities

Horizontal hydroacoustics for the estimation of fish parameters in sea bass (Dicentrarchus labrax) farming facilities. PhD thesis. Carlos Orduna Marín. Directors: Lourdes Encina Encina, Amadora Rodríguez Ruiz and Victoria Rodríguez Sánchez. Department of Plant Biology and Ecology, Faculty of Biology, University of Seville. Seville 2024. Hidroacústica horizontal para la estima de parámetros de peces en sistemas de cultivo de lubina (Dicentrarchus labrax). Tesis doctoral. Carlos Orduna Marín. Directoras: Lourdes Encina Encina, Amadora Rodríguez Ruiz y Victoria Rodríguez Sánchez. Departamento de Biología Vegetal y Ecología, Facultad de Biología, Universidad de Sevilla. Sevilla 2024. A Vicky. Index/Índice Introduction. ........................................................................................... 9 1. Aquaculture. ............................................................................... 9 1.1. The origins of aquaculture. ....................................................... 9 1.2. Geographic and technological diversification in the 20th century………………………………………………………………………………… 12 1.3. Global aquaculture today. ....................................................... 13 1.4. Aquaculture in Spain. ............................................................. 16 1.5. The future of aquaculture. ...................................................... 19 2. Hydroacoustics. ........................................................................ 20 2.1. The origin of hydroacoustics. ................................................. 20 2.2. The functioning of hydroacoustics. ........................................ 22 2.3. Horizontal hydroacoustics. ..................................................... 22 3. The use of hydroacoustics in aquaculture. .............................. 23 Results. .................................................................................................. 25 Introducción. ......................................................................................... 29 1. La acuicultura........................................................................... 29 1.1. Los orígenes de la acuicultura. ............................................... 29 1.2. Diversificación geográfica y tecnológica en el siglo XX. ........ 32 1.3. La acuicultura mundial en la actualidad. ................................ 33 1.4. La acuicultura en España. ....................................................... 37 1.5. El futuro de la acuicultura. ..................................................... 39 2. La hidroacústica. ...................................................................... 40 2.1. El origen de la hidroacústica. .................................................. 41 2.2. El funcionamiento de la hidroacústica. .................................. 42 2.3. La hidroacústica horizontal. ................................................... 43 3. El uso de la hidroacústica en la acuicultura. ........................... 44 Resultados. ............................................................................................ 45 Article/Artículo I. ................................................................................. 47 Invention/Invención I. .......................................................................... 61 Article/Artículo II. ................................................................................ 77 7 Figure 2. Production percentages of different groups of organisms produced through aquaculture worldwide in 2022 (FAO, 2024). On the other hand, focusing on the top 10 most cultivated species worldwide, the top three in production are algae. The first is kombu, represented by various species of the genus Laminaria, which is widely consumed in Japan (16%); followed by eucheuma or guso (Eucheuma spp.), an alga consumed in Indonesia and the Philippines and a raw material to produce carrageenan, which is important in the cosmetics industry (12%). The third group is Gracilaria spp., red algae used as human food, seafood feed, and to produce agar (11%). In fourth place is a crustacean, shrimp (10%), followed by a mollusc, oysters (9%). From the fifth position onwards, we find the first fish species: the grass carp (Ctenopharyngodon Idella Valenciennes, 1844, 9%), the silver carp (Hypophthalmichthys molitrix Valenciennes, 1844, 7%), and the Nile tilapia (7%), all three being freshwater species. The eighth most produced species in the world is again a mollusc, the Japanese clam (Ruditapes decussatus Linnaeus, 1758, 7%), followed by two fish species, the catla (Catla catla Hamilton, 1822, 6%) and the common carp (6%) (Figure 3) (FAO, 2024). It is noteworthy that of the 10 most cultivated aquatic species in the world, only two can typically be found in a Spanish fish market: shrimp Aquatic plants 28% Crustaceans 10% Diadromous fishes 5% Freshwater fishes 39% Marine fishes 3% Miscellaneous aquatic animals 1% Molluscs 14% 14 and the Japanese clam, with no marine fish or species commonly consumed in Spain or Mediterranean countries present on the list. Figure 3. Production percentages of the 10 most cultivated aquatic species worldwide in 2022 (FAO, 2024). Regarding global aquaculture production distribution, there are significant differences across the various regions of the planet. Asia is the global leader in aquaculture, accounting for 91.4% of global production in 2022. China, as the largest producer, dominates with 58.0 % of the total, thanks to the cultivation of algae, carp, and tilapia. This dominance is nearly absolute when referring to aquatic plant cultivation (99.3% of global production), which in turn represents 43.5% of its total aquaculture production. The perfection of techniques such as polyculture and the adoption of advanced technologies have been key to its success. Other countries like India, Vietnam, and Indonesia are also major producers, contributing 22.9% of Asian production, particularly in species such as tilapia, carp, and panga (Pangasius hypophthalmus Sauvage, 1878). Latin America and the Caribbean contribute 3.3% of global production, focusing on high-value export species. Chile is the second-largest global Japanese kelp 16% Eucheuma seaweeds 12% Gracilaria seaweeds 11% Whiteleg shrimp 10% Cupped oysters 9% Grass carp 9% Silver carp 7% Nile tilapia 7% Japanese carpet shell 7% Catla 6% Common carp 6% 15 producer of salmon, while Ecuador leads in shrimp production. Brazil and Mexico are in the process of expanding with Nile tilapia farming. Europe accounted for 2.7% of global production in 2022, standing out for its high level of technology and strict environmental regulations. Norway is the world's largest producer of salmon, and Mediterranean countries like Spain, Greece, Italy, and Turkey excel in the production of sea bass and gilthead sea bream. Spain is also one of the largest producers of mussels (Mytilus galloprovincialis, Lamarck, 1819). In Africa, aquaculture accounts for only 1.9% of global production but has great potential to combat food insecurity in rural areas. Egypt leads the continent's aquaculture production, focusing on tilapia, while Nigeria has developed farming systems for African catfish (Clarias gariepinus, Burchell, 1822) and tilapia. North America contributes 0.5% of global production, focusing on highvalue species such as salmon, mollusc, and rainbow trout. Canada follows a model similar to Norway, while in the United States, rainbow trout and catfish (Ictalurus punctatus Rafinesque, 1818) farming predominates. Oceania, with 0.2% of global production, focuses on mollusc and salmonid farming in Australia and New Zealand, with the latter standing out for its production of green mussels (Perna canaliculus, Gmelin, 1791) and environmentally sustainable practices. 1.4. Aquaculture in Spain. Aquaculture in Spain has its roots in Roman times, with evidence of the existence of cetariae, nurseries in estuaries that the Romans used for the fattening and preservation of marine species such as gilthead seabream and mullet (Mugil cephalus Linnaeus, 1758). These nurseries were common along the Mediterranean coasts, especially in areas like Cádiz, where fishing was combined with captive breeding to ensure a constant supply of fresh fish for trade and local consumption (Bernal Casasola, 2014; Vargas Girón, 2020). The true rise of modern aquaculture in Spain occurred in the second half of the 20th century, with the introduction of intensive farming techniques for marine species in floating cages, the transformation of estuarine salt flats, and the development of raft cultivation systems, particularly in Galicia with mussels. Since then, Spain has established itself as one of Europe's leading aquaculture producers, thanks to the adoption of new technologies and driven by the European Union’s legislative framework. 16 Today, Spain is one of the largest aquaculture producers in the European Union, with a consolidated and modern industry. In 2022, 326,520 tons of aquaculture products were produced, representing 23% of the total production in the European Union (APROMAR, 2023). Regarding the groups of organisms cultivated in Spain, the most recent FAO data from 2022 show that molluscs are the most cultivated, accounting for 75% of production and 227,674 tons. They are followed by marine fish, with 19% and 61,235 tons, and freshwater fish, with 6% and 16,470 tons (Figure 4). As with global aquaculture, despite the common perception that aquaculture is associated with marine fish farming, in Spain, 3 out of 4 kilos of aquaculture products are molluscs, with the vast majority (98.9%) being mussels (APROMAR, 2023). In terms of the most cultivated species (Figure 5), mussels hold a dominant position, representing 75% of Spain’s aquaculture production. Their cultivation is concentrated mainly in Galicia, which accounts for 97% of the country's total production. Mussel seed is collected from rocks or through collector ropes and then planted in rafts, floating platforms from which ropes with mussels hang, or in long-line systems. The mussels are fed naturally by filtering water until they are harvested. The second most cultivated species is sea bass, accounting for 8% of production. Their cultivation begins with the production of fry, generated in breeding centres from broodstock under controlled conditions. Most of these fry are stocked in cylindrical cages in the sea, while a smaller portion are reared in coastal earth ponds, where they grow to market size, feeding on commercial feed. 17 Figure 4. Production percentages of different groups of organisms cultivated through aquaculture in Spain in 2022 (APROMAR, 2023). Figure 5. Percentages of the most cultivated aquatic species in Spain in 2022 (APROMAR, 2023). Freshwater fishes 6% Marine fishes 19% Molluscs 75% Sea mussels 74% European seabass 8% Rainbow trout 5% Atlantic bluefin tuna 4% Gilthead seabream 3% Turbot 3% Meagre 2% Red swamp crawfish 1% 18 Rainbow trout, the only freshwater fish with significant production, follows with 5% of total production. It is cultivated in concrete tanks. Bluefin tuna (Thunnus thynnus Linnaeus, 1758), accounting for 4% of production, bases its farming system on the capture of wild specimens that are confined in floating farms, considered marine nurseries, where they are fattened and later harvested. Other species include gilthead seabream, accounting for 3% of production and farmed similarly to sea bass, and turbot (Scophthalmus maximus Linnaeus, 1758), which is produced using a similar method for fry but is mainly raised in circular concrete tanks in coastal facilities. Meagre (Argyrosomus regius Asso, 1801) accounts for 2% of production and is farmed in a manner similar to sea bass and gilthead seabream. Other species cultivated on a smaller scale in Spain include sole (Solea senegalensis Kaup, 1858), eel (Anguilla anguilla Linnaeus, 1758), yellowtail (Seriola dumerili Risso, 1810), sturgeon (Acipenser naccarii Bonaparte, 1836 and Acipenser baerii Brandt, 1869), tench (Tinca tinca Linnaeus, 1758), oysters (Ostrea edulis Linnaeus, 1758 and Crassostrea gigas Thunberg, 1793), clams (Ruditapes decussatus Linnaeus, 1758 and Ruditapes philippinarum Adams & Reeve, 1850), shrimp, and some species of macro and microalgae (APROMAR, 2023). 1.5. The future of aquaculture. The future of global aquaculture points to sustained growth, playing a relevant role in global food supply. It is shaping up as a key sector to meet the growing demand for aquatic products, driven by both population growth and the overexploitation of extractive fishing grounds. Among the main challenges aquaculture faces are environmental sustainability, the efficient management of natural resources such as water and feed, and the need to develop systems that minimize ecological impacts, such as pollution and biodiversity loss. In this context, innovation and the development of new techniques and technologies are crucial to improving production, which remains the sector’s main goal, while addressing the significant challenge of ensuring sustainable resource management. Specifically, this work focuses on the need for fish farms to obtain accurate data regarding the size, density, and biomass of fish in cultivation. The more precise this data are, the better the optimization of the management of these facilities in aspects such as feed requirements, feed conversion ratios, medication administration, sales forecasts, early detection of fish loss 19 due to mortality, theft or escapes, etc. (McCallum, 2005; Soliveres et al., 2014; Pérez et al., 2015; Føre et al., 2016; Hofmeester et al., 2016). 2. Hydroacoustics. Hydroacoustics is the branch of physics that studies the sound in water. Since ancient times, rivers, lakes, seas, and oceans have fascinated people from different societies worldwide. The aquatic depths, invisible to the naked eye, have continuously posed a challenge to understanding the natural world. Unlike terrestrial ecosystems, where direct observation is relatively easy, the aquatic environment creates a physical barrier that complicates the study of the species living within it. The study of fish presents unique challenges due to their habitat. The apparent transparency of water is deceiving; its density, currents, and changing light and temperature conditions create a dynamic, threedimensional environment that makes direct observation difficult. Throughout the centuries, humans have sought ways to "see" through water, whether through fishing, diving, or, more recently, the development of advanced detection technologies aimed at studying organisms capable of thriving in an environment that is, in many ways, impregnable to us. It is precisely this difficulty that has stimulated the development of more sophisticated and non-invasive techniques for aquatic research. Hydroacoustics is one of the most efficient, offering a way to "listen" to fish using sound waves, thus overcoming the visual barrier imposed by water. This technology not only reveals the presence of fish but also provides key information about their size, biomass, and behaviour without needing to physically interact with them. In an environment where human intervention can disrupt the ecosystem's balance, hydroacoustics emerges as an effective solution for studying the underwater world without disturbing it (Simmonds & MacLennan, 2007; Kubecka et al., 2009). 2.1. The origin of hydroacoustics. The first documented evidence of scientific interest in underwater sound dates back to Leonardo Da Vinci, who in 1490 described how, by using a submerged tube, one could hear ships from a great distance. However, the first formal experiment occurred in 1826 when physicists Daniel Colladon and Charles Sturm used a long tube to listen to sounds emitted by a bell underwater at a known distance. In doing so, they measured the speed of sound in Lake Geneva (between France and Switzerland), recording 1,435 m/s in water at 1.8°C, only 3 m/s less than the speed accepted today and much 20 faster than the speed of sound in air, which is 332.4 m/s at the same temperature (Rodríguez-Sánchez, 2015). At the beginning of the 20th century, technological advances allowed for the practical application of hydroacoustics. In 1918, at the end of World War I, the ASDIC (Anti-submarine Detection Investigation Committee) emerged under the auspices of France and Great Britain to counter the German naval dominance based on their submarine fleet. As a result, SONAR (Sound Navigation and Ranging) was developed, a submarine detection system that works by emitting sound pulses and measuring their echo when reflected off submerged objects (Granado-Lorencio, 1996). This technology evolved to be used in civilian applications such as fishing, for detecting fish starting in the 1930s. In 1929, Japanese researcher Kimura was a pioneer in detecting fish using this method, with the first successful detection experience (Simmonds & MacLennan, 2007). In the following decades, hydroacoustic research advanced significantly with studies improving the accuracy of fish biomass estimates. Researchers like Foote and Love (Love, 1971; Foote, 1980), developed equations that related the energy reflected by fish to their size, facilitating the use of this technology in fisheries management, which has greatly improved its accuracy through the development of various applications of the technique and specific equations to relate fish sound to their size and weight (Frouzova et al., 2005; Draštík et al., 2009; Godlewska et al., 2012; Rodríguez-Sánchez, 2015; Rodríguez-Sánchez et al., 2015, 2016b, 2018b; Balk et al., 2017; Baran et al., 2017). The main advantage of hydroacoustics in ichthyofauna sampling is its great efficiency, allowing the collection of large amounts of data and the coverage of extensive bodies of water in short periods, providing accurate information about the distribution, abundance, size, and biomass of fish (Encina & Rodríguez-Ruiz, 2002). Moreover, hydroacoustics is non-invasive, as it does not require the capture of fish, avoiding their stress (in the bestcase scenario) and the alteration of their behaviour. Another advantage is that it allows work in turbid waters or at great depths, where traditional methods present significant limitations. Additionally, the obtained data can be integrated into geographic information systems (GIS), facilitating largescale analysis. Although the technique has limitations in species identification, it remains a powerful and efficient tool for the ecological study of fish. 21 2.2. The functioning of hydroacoustics. Hydroacoustics uses sound waves to detect and measure the presence of objects underwater by using an echo sounder. This device acts both as a transmitter and a receiver of acoustic signals. The sound waves travel through the water mass and strike objects and organisms in their path, generating an echo. This echo returns to the echo sounder, which captures and processes it. The time it takes to return is used to calculate the distance, while the echo intensity is used to estimate the size of the object, known as target strength (TS) (Simmonds & MacLennan, 2007). Through data acquisition software, this information is translated into an echogram, visually representing the underwater ecosystem (Figure 6) (RodríguezSánchez, 2015). Figure 6. Graphic representation of the functioning of hydroacoustics. The red cylinder shows the signal-emitting and receiving device. On the right is an echogram, which graphically represents the received sound (Rodríguez-Sánchez, 2015). To translate this TS into biological parameters, conversion equations are needed to assign each signal a size or weight (Lucas & Baras, 2000; Rodríguez-Sánchez et al., 2015). 2.3. Horizontal hydroacoustics. Conventional hydroacoustics proposes a vertical use of the acoustic beam direction, emitting sound waves perpendicularly to the water surface and insonifying fish dorsoventrally. This technique is widely used and has a high degree of refinement in deep marine ecosystems, but it presents certain limitations when used in shallow ecosystems. Firstly, the acoustic beam is 22 cone-shaped with a specific opening angle, which in the first few meters may be too small to gather a representative amount of data for the ecosystem. Secondly, there is the near-field effect. This concept refers to an unstable acoustic wave zone located just after the sound emitter, making the measurements unreliable (Dawson et al., 2000; Knudsen et al., 2004; Simmonds & MacLennan, 2007). This problematic measurement area can cause density estimates in shallow ecosystems or ecosystems with heterogeneous fish distribution in the water column to significantly deviate from reality (Kubecka & Duncan, 1998; Knudsen & Saegrov, 2002; Kubecka et al., 2012). To solve these issues, vertical sampling must be supplemented by using horizontal hydroacoustics, with the acoustic beam pointing parallel to the water surface, sampling the surface area not covered by vertical hydroacoustics (Yule, 2000; Knudsen & Saegrov, 2002). However, horizontal hydroacoustics presents a challenge due to the angle at which the acoustic beam strikes the fish, causing significant differences in the echo intensity returned. Fish tend to swim horizontally or at shallow ascent/descent angles, so when vertical hydroacoustics insonifies them dorsoventrally, the surface area of their bodies exposed to the sound beam does not vary significantly. In contrast, when fish are insonified horizontally, they swim in all directions, so the acoustic beam may hit them head-on or tail-first, exposing a small portion of their bodies, or hit them completely laterally, exposing the largest part of their bodies. The larger or smaller exposure of the fish to the acoustic beam results in higher or lower TS signals. Therefore, to increase the accuracy of these measurements, it is necessary to use equipment that can detect the angle of incidence of the fish with the acoustic beam and conversion equations that take this orientation into account when assigning biological parameters to detections (Hazen & Horne, 2003; Simmonds & MacLennan, 2007; Rodríguez-Sánchez et al., 2015, 2016b, 2018b). 3. The use of hydroacoustics in aquaculture. This doctoral thesis focuses on applying hydroacoustic technology to improve fish aquaculture production systems. Specifically, it centers on the production of sea bass, the most cultivated marine fish species in Spain and southern Europe, distinguishing between the two most common types of facilities: land-based ponds and open sea cages. One of the main areas where these farming systems can improve is in monitoring fish size during the fattening phase (Soliveres et al., 2014). More precise control would allow for the optimal adjustment of the amount of feed 23 distintas especies con el objetivo de incrementar las poblaciones y facilitar su posterior captura (Beveridge & Little, 2002; Costa-Pierce, 2022). Los restos arqueológicos más antiguos vinculados a la acuicultura fueron descubiertos en el yacimiento de Jiahu, en la provincia de Henan, China, y datan de entre los años 6.200 y 5.700 a.C. Estos hallazgos muestran a la carpa común (Cyprinus carpio Linnaeus, 1758) como el primer organismo cultivado de forma documentada. Los peces se mantenían vivos en aguas confinadas, reguladas por personas, donde desovaban de manera natural y las crías crecían aprovechando los recursos disponibles, confirmándose la práctica de la acuicultura ya en el neolítico, 8.000 años atrás (Nakajima et al., 2019). China se destaca indudablemente como uno de los primeros y más influyentes centros de desarrollo de la acuicultura, con la implementación de complejas técnicas de policultivo en estanques de peces integrados con los sistemas de irrigación de los arrozales. Estos sistemas beneficiaban tanto a los cultivos como a la cría de organismos acuáticos, incrementando la producción general de alimentos, con un enfoque particular en la proteína, esencial para el crecimiento de las poblaciones urbanas de la época. Un ejemplo significativo es el tratado publicado por Fan Li hace 2.500 años, en el que se describe el cultivo de carpas con tal detalle que convierte al documento en una prueba inequívoca del desarrollo avanzado de la acuicultura en la antigüedad (Li & Mathias, 1994). Otro punto importante en cuanto al origen de la acuicultura se encuentra en el antiguo Egipto, donde los peces tenían a la vez un papel prosaico y sagrado en la sociedad, sirviendo de alimento pero estando también vinculados con las fuerzas cíclicas que daban vida al río Nilo. En concreto, la tilapia del Nilo (Oreochromis niloticus Linnaeus, 1758) estaba vinculada a la diosa Hathor y al concepto de reencarnación (Desroches-Noblecourt, 1954). Se cree que la pesca con caña era una práctica habitual en todos los estratos de la sociedad egipcia, reservándose para la nobleza la pesca en balsas artificiales construidas en jardines de la clase pudiente, cuyo interés residía más en los rituales religiosos asociados con la muerte y la reencarnación que en el ocio o la alimentación (Brewer & Friedman, 1989). En cuanto al cultivo de tilapia en el antiguo Egipto, Chimits (1957), reproduce un bajorrelieve de la tumba de Thebaine realizado hace 4.000 años en el que se muestra a un ciudadano perteneciente a la nobleza sentado en su jardín, pescando con una caña de doble línea, con su mujer sentada detrás, quitando los anzuelos a los peces, junto a lo que parece ser un estanque artificial (Figura 1). El bajorrelieve es considerado de notable importancia, ya que también muestra el cultivo de loto y frutales recogidos por sirvientes e irrigados con el agua 30 del estanque, representando un ejemplo de acuicultura integrada en la antigüedad (Costa-Pierce, 2022). Figura 1. Bajorrelieve de la tumba de Thebaine, en el delta del Nilo (2.000 a.C.) (Chimits, 1957), redibujado por Costa-Pierce (2022). En Europa, los primeros indicios de acuicultura se remontan a los etruscos y romanos, quienes construían barreras de tierra permanentes o semipermanentes en lagunas costeras de los mares Adriático y Tirreno, creando ecosistemas productivos controlados en los que posteriormente instalaban trampas para la captura de los peces, en un sistema conocido hoy en día como “vallicoltura”(Beveridge & Little, 2002). Marco Terencio Varrón, (116-27 a.C.) caballero y polígrafo romano, describió la preferencia de los romanos por los peces de mar, enfatizando que los peces continentales de estanques eran considerados vulgares, destinados a los plebeyos, documentando así la existencias de balsas de acuicultura de agua dulce en esa época (Balon, 1995). Es relevante destacar el aprecio de esta sociedad por los organismos marinos, lo que llevó a Cayo Sergio Orata (siglo I a.C.), inventor romano, a construir depósitos de agua salada lejos del mar, llamados piscinae. Estos depósitos permitían a la alta sociedad tener un suministro permanente de productos marinos frescos, independientemente del clima o el éxito de las capturas. Una característica innovadora de estos depósitos era la capacidad de controlar la temperatura del agua, lo que facilitaba el cultivo de doradas 31 (Sparus aurata Linnaeus, 1758), una especie muy sensible a las bajas temperatura invernales. Este sistema constituye el primer hito documentado de acuicultura marina con un considerable nivel de sofisticación y lejos de la costa. Así mismo, Cayo Sergio Orata también diseñó un sistema de cultivo para ostras (Crassostrea spp.), producto muy apreciado por los romanos, estableciendo un modelo de producción semi-intesiva muy similar al utilizado en la actualidad para el cultivo de moluscos (Balon, 1995; Kron, 2008). Tras el colapso del Imperio romano y el establecimiento del cristianismo, los relativamente complejos sistemas de acuicultura desarrollados hasta entonces prácticamente desaparecieron. Fue a lo largo de la Edad Media y el Renacimiento cuando la acuicultura de peces continentales se expandió significativamente, presentándose como práctica habitual en los monasterios, que encontraron en la cría de peces una solución adecuada para cumplir con las necesidades alimenticias sin violar las prohibiciones religiosas, que vetaban el consumo de carne durante una parte importante del año (Nash, 2011). Los monjes desarrollaron sofisticados sistemas de estanques en los que criaban mayormente carpas, que se convirtieron en una parte esencial de la dieta de muchas comunidades monásticas en Europa central y del este. Con el tiempo, estas prácticas se extendieron más allá de los monasterios, y los estanques de peces comenzaron a proliferar en propiedades privadas y públicas, especialmente en regiones alejadas del mar. El caso de la expansión de la carpa común en Europa es un claro ejemplo de cómo la acuicultura pasó de ser una práctica local a una actividad extendida en buena parte del continente. 1.2. Diversificación geográfica y tecnológica en el siglo XX. A lo largo del siglo XX, la acuicultura no solo se expandió geográficamente, sino que también experimentó una diversificación en términos de especies cultivadas y técnicas empleadas. En este periodo, países como Japón destacaron en la innovación acuícola, particularmente en el cultivo de ostras y algas. Japón además fue pionero en el desarrollo de técnicas de cría en jaulas y el uso de piensos compuestos, lo que facilitó la expansión de la acuicultura marina (Nash, 2011). En América, el impulso llegó en la segunda mitad del siglo XX. En países como Estados Unidos y Canadá, la acuicultura se desarrolló principalmente en torno a especies como la trucha (Oncorhynchus mykiss Walbaum, 1792) y el salmón (Salmo salar Linnaeus, 1758), inicialmente como apoyo a la pesca deportiva. Sin embargo, la cría del salmón en jaulas marinas pasó con el tiempo a convertirse en una industria comercial masiva, con Noruega, Chile 32 y Escocia como grandes productores, y trasformando el mercado global de pescado (Costa-Pierce, 2022). El cultivo de langostino (Penaeus vannamei Boone, 1931) es otro ejemplo destacado de diversificación e industrialización de la producción acuícola de finales del siglo XX, especialmente en América Latina y el Sudeste Asiático. Con una rápida expansión en el cultivo en balsas continentales y beneficiado en gran medida por el alto valor de exportación, países como Ecuador, Brasil y Tailandia se convirtieron rápidamente en líderes de una industria internacional en auge, desarrollando tecnologías avanzadas en alimentación y manejo de la producción (Nash, 2011; APROMAR, 2023). En Europa, la acuicultura también experimentó un crecimiento notable durante este periodo, impulsado en gran medida por la modernización tecnológica y el aumento de la demanda de productos acuáticos. El desarrollo del cultivo de lubina (Dicentrarchus labrax Linnaeus, 1758) y dorada en países mediterráneos como España, Italia y Grecia fue especialmente significativo, ya que estas especies se convirtieron en productos de alto valor tanto para el mercado interno como para la exportación. Paralelamente, países nórdicos, con Noruega como principal exponente, lideraron la producción de salmón en jaulas marinas, convirtiendo a esta especie en uno de los productos acuícolas más apreciados de Europa y del mundo. Cabe destacar que la Unión Europea promovió políticas que incentivaron la investigación y el desarrollo tecnológico en el sector, lo que contribuyó a la expansión de la acuicultura intensiva y a la diversificación de especies cultivadas (APROMAR, 2023). 1.3. La acuicultura mundial en la actualidad. La producción total de animales acuáticos ha aumentado progresivamente a lo largo de las últimas décadas, pasando de 19 millones de toneladas de peso vivo, registradas en 1950, a un récord histórico de más de 185 millones de toneladas en 2022, con un ritmo de crecimiento medio anual del 3,2%. Si se incluyen en la producción todos los organismos acuáticos, el récord asciende a 223 millones de toneladas de peso vivo (FAO, 2024). De hecho, 2022 se enmarca como el primer año en el que la producción acuícola de animales acuáticos superó a la producción de la pesca extractiva. De los 185 millones de toneladas de animales acuáticos producidos, el 51% (94 millones de toneladas) procedió de la acuicultura, mientras que el 49% (91 millones de toneladas), de la pesca de captura. Si se considera el total de la producción de organismos acuáticos (incluye a las algas como factor diferencial) el sorpasso se habría producido ya en el año 2013, con una producción total de 186 millones de toneladas, de las cuales el 51% (95 33 millones de toneladas) fue aportado por la acuicultura, y el 49% (91 millones de toneladas) por la pesca extractiva (FAO, 2024). En cuanto a los grupos de organismos cultivados, los datos más recientes de la FAO muestran que en 2022 el cultivo de peces de agua dulce fue el mayor en cuanto a producción, con un 39%. Le siguen las plantas acuáticas (28%), los moluscos (14%) y los crustáceos (10%). A pesar de la percepción predominante en España sobre la acuicultura, que relaciona habitualmente a este sector con la producción de peces de agua salada, los peces diádromos (principalmente el salmón) y los peces marinos ocuparon las últimas posiciones, con un 5% y un 3% de la producción, respectivamente (Figura 2) (FAO, 2024). Figura 2. Porcentajes de producción de los distintos grupos de organismos producidos mediante acuicultura en el mundo en el año 2022 (FAO, 2024). Por otro lado, si se pone el foco en las 10 especies más cultivadas en el mundo, las 3 con mayores producciones son algas. La primera es el kombu, representado por varias especies del género Laminaria, muy consumida en Japón (16%); seguida por la eucheuma o guso (Eucheuma spp.), un alga consumida en Indonesia y Filipinas, y materia prima para la producción de carragenano, importante en la industria cosmética (12%). El tercer grupo son Plantas acuáticas 28% Crustáceos 10% Peces diádromos 5% Peces de agua dulce 39% Peces marinos 3% Otros animales acuáticos 1% Moluscos 14% 34 las garcilarias (Garcilaria spp.), algas rojas utilizadas como alimento humano, de mariscos y para la producción de agar (11%). En cuarta posición se encuentra un crustáceo, el langostino (10%) seguido de un molusco, la ostra (9%). Es a partir de la quinta posición cuando encontramos a las primeras especies de peces, la carpa herbívora (Ctenopharyngodon idella Valenciennes, 1844, 9%), la carpa plateada (Hypophthalmichthys molitrix Valenciennes, 1844, 7%) y la tilapia del Nilo (7%), las 3 de agua dulce. La octava especie más producida en el mundo vuelve a ser un molusco, la almeja japónica (Ruditapes decussatus Adams & Reeve, 1850, 7%) seguida por dos peces, la catla (Catla catla Hamilton, 1822, 6%) y la carpa común (6%) (Figura 3) (FAO, 2024). Cabe destacar que, de las 10 especies de organismos acuáticos cultivados en el mundo, en una pescadería española únicamente podríamos encontrar 2 de ellas, el langostino y la almeja japónica, sin encontrarse en la lista ningún pez marino ni de consumo habitual en España o en los países mediterráneos. Figura 3. Porcentajes de las 10 especies de organismos acuáticos con más producción mediante acuicultura en el mundo en el año 2022 (FAO, 2024). Kombu 16% Eucheuma 12% Gracilarias 11% Langostino vannamei 10% Ostras 9% Carpa herbívora 9% Carpa plateada 7% Tilapia del Nilo 7% Almeja japónica 7% Catla 6% Carpa común 6% 35 En cuanto a la distribución mundial en términos de producción de la acuicultura, hay grandes diferencias en las distintas regiones del planeta. Asia es líder mundial en la acuicultura, representando el 91,4% de la producción global en 2022. China, como mayor productor, domina con un 58,0% del total, gracias al cultivo de algas, carpas y tilapia. Este enorme dominio es prácticamente total si nos referimos al cultivo de plantas acuáticas (99,3% de la producción mundial) que a su vez representa el 43,5% del total de su producción acuícola. El perfeccionamiento de técnicas como el policultivo y la adopción de tecnologías avanzadas han sido clave para su éxito. Otros países como India, Vietnam e Indonesia también son importantes productores, con un 22,9% de la producción asiática, destacando en especies como la tilapia, la carpa y el panga (Pangasius hypophthalmus Sauvage, 1878). América Latina y el Caribe aportan el 3,3% de la producción global, centrada en especies de alto valor comercial para la exportación. Chile es el segundo mayor productor mundial de salmón, mientras que Ecuador es líder en la producción de langostinos. Brasil y México están en proceso de expansión con el cultivo de tilapia del Nilo. Europa representó el 2,7% de la producción mundial en 2022, destacando por su alto nivel tecnológico y estrictas regulaciones ambientales. Noruega es el mayor productor mundial de salmón, y países mediterráneos como España, Grecia, Italia y Turquía sobresalen en la producción de lubina y dorada. España es, además, uno de los mayores productores de mejillón (Mytilus galloprovincialis Lamarck, 1819) . En África, la acuicultura representa solo el 1,9% de la producción mundial, pero tiene un gran potencial para combatir la inseguridad alimentaria en áreas rurales. Egipto lidera la producción acuícola en el continente, centrada en la tilapia, mientras que Nigeria ha desarrollado cultivos de bagre africano (Clarias gariepinus Burchell, 1822) y tilapia. América del Norte contribuye con un 0,5% de la producción global, enfocándose en especies de alto valor como el salmón, los moluscos y la trucha. Canadá sigue un modelo similar al de Noruega, mientras que en Estados Unidos predominan el cultivo de trucha y bagre (Ictalurus punctatus Rafinesque, 1818). Oceanía, con un 0,2% de la producción mundial, se centra en el cultivo de moluscos y salmónidos en Australia y Nueva Zelanda, destacando esta última por la producción de mejillón verde (Perna canaliculus Gmelin, 1791) y sus prácticas ambientalmente sostenibles. 36 1.4. La acuicultura en España. La acuicultura en España tiene sus raíces en la época romana, con evidencias de la existencia de cetariae, viveros en los esteros que los romanos utilizaban para el engorde y conservación de especies marinas como la dorada y el mújol (Mugil cephalus Linnaeus, 1758). Estos viveros eran comunes en las costas del Mediterráneo, especialmente en áreas como Cádiz, donde se combinaba la pesca con la cría en cautividad para asegurar un suministro constante de peces frescos para el comercio y consumo local (Bernal Casasola, 2014; Vargas Girón, 2020). El verdadero despegue de la acuicultura moderna en España ocurrió en la segunda mitad del siglo XX, con la introducción de técnicas intensivas de cría de especies marinas en jaulas flotantes, la transformación de las salinas de esteros, y los sistemas de cultivo en bateas, especialmente en Galicia con el mejillón. Desde entonces, España se ha consolidado como uno de los principales productores acuícolas de Europa, gracias a la adopción de nuevas tecnologías y la regulación del sector, impulsada por el marco legislativo de la Unión Europea. En la actualidad, España es uno de los mayores productores de acuicultura de la Unión Europea, con una industria afianzada y moderna. En 2022 se produjeron 326.520 toneladas de productos acuícolas, representando el 23% de la producción total de la Unión Europea (APROMAR, 2023). En cuanto a los grupos de organismos cultivados en España, los datos más recientes de la FAO, de 2022, presentan a los moluscos como los más cultivados, con un 75% de la producción y 227.674 toneladas. Le siguen los peces marinos, con un 19% y 61.235 toneladas y los peces de agua dulce, con un 6% y 16.470 toneladas (Figura 4). Al igual que ocurre con los organismos cultivados a nivel mundial, pese a que la percepción más habitual por parte de la población en cuanto a la acuicultura evoque al cultivo de peces marinos, en España 3 de cada 4 kilos de productos acuícolas son moluscos y en su gran mayoría (el 98,9%) de mejillones (APROMAR, 2023). En cuanto a las especies más cultivadas (Figura 5), con una posición predominante se encuentra el mejillón, representando un 74% de la producción española. Su cultivo se concentra principalmente en Galicia, que aporta el 97% de la producción total del país. La semilla del mejillón se recolecta de rocas o mediante cuerdas colectoras, y luego se siembra en sistemas de long-lines o bateas, plataformas flotantes de las que cuelgan cuerdas con los moluscos. Los mejillones posteriormente se alimentan de manera natural, filtrando el agua, hasta su posterior recolección. 37 Figura 4. Porcentajes de producción de los distintos grupos de organismos cultivados mediante acuicultura en España en el año 2022 (APROMAR, 2023). Figura 5. Porcentajes de las especies de organismos acuáticos con más producción mediante acuicultura en España en el 2022 (APROMAR, 2023). Mejillón 74% Lubina 8% Trucha 5% Atún rojo 4% Dorada 3% Rodaballo 3% Corvina 2% Otros 1% Peces de agua dulce 6% Peces marinos 19% Moluscos 75% 38 La segunda especie más cultivada es la lubina con un 8% de la producción. Su cultivo se inicia con la obtención de alevines, producidos en centros de reproducción a partir de individuos reproductores en condiciones controladas. Estos alevines son sembrados en su mayor parte en jaulas cilíndricas en el mar, y en menor medida en estanques de tierra en entornos costeros, donde engordan hasta la talla comercial alimentándose de pienso. Le siguen la trucha arcoíris (5% de la producción), único pez de agua dulce con una producción significativa, que se cultiva en tanques de hormigón; y el atún rojo (Thunnus thynnus Linnaeus, 1758, 4% de la producción), que basa su sistema de cultivo en la captura de ejemplares salvajes que son confinados en granjas flotantes, consideradas viveros marinos, en los que se les alimenta para su engorde y posterior sacrificio. Con un 3% de la producción cuentan la dorada, con un sistema de cultivo similar al de la lubina; y el rodaballo (Scophthalmus maximus Linnaeus, 1758) con un sistema de obtención de alevines similar pero cuyo cultivo posterior se realiza mayormente en tanques circulares de hormigón en instalaciones costeras. Con un 2% de la producción se encuentra la corvina (Argyrosomus regius Asso, 1801) con un cultivo también muy parecido a lubinas y doradas. En cuanto a otras especies cultivadas de manera minoritaria en España, se encuentran el lenguado (Solea senegalensis Kaup, 1858), la anguila (Anguilla anguilla Linnaeus, 1758), la seriola (Seriola dumerili Risso, 1810), el esturión (Acipenser naccarii Bonaparte, 1836 y Acipenser baerii Brandt, 1869), la tenca (Tinca tinca Linnaeus, 1758), las ostras (Ostrea edulis Linnaeus, 1758 y Crassostrea gigas Thunberg, 1793), las almejas (Ruditapes decussatus Linnaeus, 1758 y Ruditapes philippinarum Adams & Reeve, 1850), el langostino y algunas especies de macro y microalgas (APROMAR, 2023). 1.5. El futuro de la acuicultura. El futuro de la acuicultura mundial apunta a un crecimiento sostenido, desempeñando un papel relevante en la alimentación global. Se perfila como un sector clave para satisfacer la creciente demanda de productos acuáticos, impulsada tanto por el aumento de la población como por la sobreexplotación de los caladeros de pesca extractiva. Entre los principales desafíos a los que se enfrenta la acuicultura, están la sostenibilidad ambiental, la gestión eficiente de recursos naturales como el agua y los piensos, y la necesidad de desarrollar sistemas que minimicen los impactos ambientales, como la contaminación y la pérdida de biodiversidad. En este contexto, la innovación y el desarrollo de nuevas técnicas y tecnologías son fundamentales para mejorar la producción, que 39 compara las relaciones longitud-talla de lubinas en dos instalaciones de cultivo diferentes, de manera estacional, para comprobar si existen variaciones significativas que deban ser tenidas en cuenta para realizar estimas de biomasa con hidroacústica. 46 Article/Artículo I. Hydroacoustics for density and biomass estimations in aquaculture ponds. Carlos Orduna, Lourdes Encina, Amadora Rodríguez-Ruiz, & Victoria Rodríguez-Sánchez. (2021). Aquaculture, 545, 737240. https://doi.org/10.1016/J.AQUACULTURE.2021.737240 47 48 Aquaculture 545 (2021) 737240 Available online 27 July 2021 0044-8486/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Hydroacoustics for density and biomass estimations in aquaculture ponds Carlos Orduna a , b , * , Lourdes Encina a , Amadora Rodríguez-Ruiz a , Victoria Rodríguez-S´ anchez b a Department of Vegetal Biology and Ecology, University of Seville, Seville, Spain b EcoFishUS Research S.L.L. Seville, Spain ARTICLE INFO Keywords: Aquaculture Hydroacoustics Sea bass Gilt-head sea bream Ponds ABSTRACT The use of hydroacoustics is currently being studied and developed as a promising non-intrusive methodology to monitor and manage fish stocks in aquaculture farms. The main objective of this study was to develop an acoustic method for the estimation of fish density and biomass in inland aquaculture farms and test the accuracy and precision of the estimates with real data provided by the company. The study was conducted in sea bass (Dicentrarchus labrax) production ponds located in Seville (Southern Spain). A Simrad EK60 echosounder with two split-beam circular transducers operating simultaneously at 200 kHz was used for hydroacoustic surveys. Two different hydroacoustic designs were considered: central trajectories and zigzag trajectories. The accuracy and precision of the estimates were examined in order to select the best sampling design. Due to a nonhomogeneous fish distribution in the pond caused by the avoidance behaviour, as a response to the sampling disturbance presented by fish, acoustic density and biomass were corrected by applying sampling theory according to the probability of fish detection. When density and biomass were corrected, the estimates became highly accurate and precise with respect to real data, which confirms that the proposed method is adequate. Similarly, acoustic estimates of fish weight were highly in agreement with real data, due to the use of specific equations developed “in situ” for the study. Although no significant differences were recorded in the density and biomass estimates with regard to the trajectory used (central vs. zigzag), it was observed that the most accurate agreement and precision were always obtained in central trajectories. Therefore, central design is proposed as the most appropriate design for hydroacoustic measurements in inland ponds. The results obtained in this study provide estimates of density and biomass that accurately match the real data, supporting the use of hydroacoustics as a potentially valid tool to manage inland aquaculture farms. 1. Introduction Aquaculture is a growing sector in Spain and other countries. In southern countries, sea bass (Dicentrarchus labrax), gilt-head sea bream (Sparus aurata), and turbot (Psetta maxima) are the most produced species, amounting to over 80% of the total market (Apromar, 2018). Most of these come from offshore facilities. Only a small percentage of them comes from inland aquaculture farms, although this sector is expanding (Espinosa et al., 2015; FAO, 2016; Magalh˜ aes et al., 2017; RodríguezS´ anchez et al., 2018). In order to be properly managed, aquaculture companies require accurate data of the abundance, biomass and average weight of the fish farmed in their facilities. The more accurate these data are, the better the decision-making for the fish farm will be in terms of feed requirements, growth rate and food conversion calculations, medication administration, early detection of fish losses due to deaths, robberies or escapes, splitting of farming units, economic forecasting, etc. (McCallum, 2005; Soliveres et al., 2014; P´ erez et al., 2015; Hofmeester et al., 2016; Føre et al., 2018). Production management techniques currently used are highly intrusive and do not offer the accuracy required to avoid inefficient operations in fish farms. These techniques involve an excessive handling of fish resulting in negative effects upon them, such as stress, deterioration of their immune system, decreased appetite and growth rate, diseases, etc. (Hatziathanasiou et al., 2002; HSUS, 2008; Di Marco et al., 2017; Li et al., 2020). Furthermore, these processes are highly laborious and increase production costs. Thus, one of the main goals in the aquaculture sector is to develop a more effective and profitable method that is easy to implement and maintain in order to monitor fish growth and evaluate their biomass. Different alternatives are currently being studied and developed where non-intrusive methodologies are used to * Corresponding author at: Department of Plant Biology and Ecology, Faculty of Biology, University of Seville, PO Box 1095, E-41080 Seville, Spain. E-mail addresses: [email protected] (C. Orduna), [email protected] (L. Encina), [email protected] (A. Rodríguez-Ruiz), [email protected] (V. RodríguezS´ anchez). Contents lists available at ScienceDirect Aquaculture journal homepage: www.elsevier.com/locate/aquaculture https://doi.org/10.1016/j.aquaculture.2021.737240 Received 9 April 2021; Received in revised form 19 July 2021; Accepted 25 July 2021 49 Aquaculture 545 (2021) 737240 monitor fish stock in aquaculture facilities. One of these non-intrusive methodologies consists of using hydroacoustic techniques to estimate fish abundance and biomass in production units. Hydroacoustics has become a technique which provides accurate and robust estimates of population density with an adequate balance between costs and results (Mehner and Schulz, 2002; Mackinson et al., 2004; Boswell et al., 2007; Koslow, 2009; Trenkel et al., 2011; Cushing, 2013; Zenone et al., 2017; Egerton et al., 2018; Føre et al., 2018). One of the most interesting advantages of applying this technique to aquaculture is that it is a noninvasive technique that allows for the estimation of fish abundance without any manipulation. In order to achieve this, this technique applies echosounders, which transmit a sound pulse of known characteristics within a water column and record the characteristics of the sound or echo returned by the transducer. The intensity of the returned echoes can be translated into estimates of fish density in the ensonified volume by applying the appropriate acoustic conversion equations (Simmonds and MacLennan, 2005; Knudsen, 2009; Winfeld et al., 2009; Cox et al., 2011). At present, hydroacoustic exploration of inland farm ponds poses an unprecedented challenge as a method to obtain reliable estimates of fish abundance and biomass. In previous studies, hydroacoustics has been used to estimate fish density in open sea cages with vertical hydroacoustics (placing the transducer with the main beam perpendicular to the water surface) (Espinosa et al., 2002; Espinosa et al., 2006; Espinosa et al., 2015; Knudsen et al., 2004; De La G´ andara and Espinosa, 2012; Soliveres et al., 2010; Soliveres et al., 2014; Soliveres, 2015). However, inland aquaculture ponds are shallow and as such, vertical hydroacoustics cannot be applied. Thus, horizontal hydroacoustics (placing the transducer with the main beam parallel to the water surface) must be used in these environments. Horizontal hydroacoustics works well in shallow systems, but its requirements are different from those of vertical hydroacoustics (Kubeˇ cka et al., 2000; Yule, 2000; Balk, 2001; Boswell et al., 2007; Draˇ stik et al., 2009; Gy¨ orgy et al., 2012; Zenone et al., 2017; Johnson et al., 2019). Foremost among these, fish size estimates change depending on the fish aspect i.e. the swimming angle of the fish with respect to the axis of the transducer (Baran et al., 2017). Therefore, to estimate fish size, equations where these variations are considered must be used or developed (Lilja et al., 2000; Frouzov´ a et al., 2005; Boswell et al., 2009; Furusawa and Amasuku, 2010; Rodríguez-S´ anchez et al., 2015; Rodríguez-S´ anchez et al., 2016a; Rodríguez-S´ anchez et al., 2016b; Balk et al., 2017; Rodríguez-S´ anchez et al., 2018). Other relevant problems when studying fish populations in farming ponds are not directly derived from hydroacoustics, but they are inherent to the sampling of wild animal populations in their natural habitats. Several authors have highlighted that the sampling has a direct or indirect effect on fish detectability and, ultimately, on its abundance or biomass estimation (Brehmer et al., 2004; Ona et al., 2007; MacNeil et al., 2008; Guillard et al., 2010; Kulbicki et al., 2010; Bozec et al., 2011; Kaartvedt et al., 2012; Glennie et al., 2015; Prato et al., 2017; Pais and Cabral, 2018; Brehmer et al., 2019). Specifically, two types of bias can occur in acoustic explorations: 1) positive bias, when fish move toward the transducer which leads to an overestimation of abundance and/or biomass; 2) negative bias, when fish move in the opposite direction of the transducer (avoidance), which leads to an underestimation of density and biomass. Avoidance is the most frequent behaviour, and its intensity can change from species to species (Lucas et al., 2002; Vabø et al., 2002; Gerlotto et al., 2004; Brehmer et al., 2019). When fish distribution in the explored systems is uniform, density and biomass adjust efficiently to those provided by the echosounder based on the volume or area ensonified by the transducer (Draˇ stik and Kubeˇ cka, 2005). However, a non-uniform distribution can lead to biased density and biomass results, as may occurs due to the shock of sampling on the fish population (Mitson and Knudsen, 2003; Jørgensen et al., 2004; Marques, 2004; Marques, 2009; Marques et al., 2013; De Robertis and Wilson, 2010; De Robertis et al., 2010; De Robertis and Handegard, 2013). In these cases, it is crucial to correct the bias in the data to obtain correct density and biomass estimates. This is especially important in inland production ponds since they are extremely shallow and relatively small (compared to a natural ecosystem) and, therefore, fish can be found in high densities, tending to gather themselves in groups (Zhao and Ona, 2003; Draˇ stik and Kubeˇ cka, 2005; Godlewska et al., 2009; Wheeland and Rose, 2015). Likewise, they are highly likely to present unwanted behaviour during the sampling process, such as escapes or burials, which results in a non-uniform distribution in relation to the transducer. Therefore, it is necessary to verify the existence of density gradients which occurred during the sampling to add them to the estimation method and thus avoid potential bias in the estimates (Hjellvik et al., 2008; Cox et al., 2011; Marques et al., 2013; Pais and Cabral, 2017). In light of the above, the motivation behind this study is to contribute not only to the improvement of the aquaculture sector, but also to advance the related scientific and methodological fields. The intention of this study is to develop an estimation method to obtain the density, biomass and average weight of fish farmed in shallow aquaculture facilities using hydroacoustic techniques. The main milestones in this study are: 1) To develop an efficient sampling design for this facility type, 2) to study fish distribution patterns during the hydroacoustic exploration, verifying the existence or not existence of density gradients, 3) To prove a model that includes density gradients to correct the bias caused by a non-uniform distribution during the sampling, 4) to verify the accuracy and precision of the estimates obtained by employing hydroacoustic methods. We generally do not know about the abundance or biomass when estimates are made. Without knowing the true density or biomass of fish in a sampled area, a true accuracy cannot be determined. In our study, actual density data, biomass and average weight from sowing and fish harvesting was provided by the fish farmers for all sampled ponds. In this sense, this experiment presents a great opportunity to validate our hydroacoustic methodology. Developing a reliable and non-intrusive method to accurately determine density and biomass in aquaculture inland farms will result in better control and management of the fishery production and greater efficiency in the aquaculture sector. 2. Material and methods This study was conducted in the sea bass production ponds of the company Pesquerías Isla Mayor S.A. (located in Seville, Spain). The ponds were rectangular (230 m ×12 m ×2 m) with a surface area of 2700 m 2 and a volume of 3150 m 3 . A total of four production ponds were studied, one of them corresponds to a just planted pond (P-1; small fish), one corresponding to intermediate fish size (P-2) and two corresponding to big fish (P-3 and P-4) that were fished after the hydroacoustic survey. At the end of the acoustic study, the company provided us with reliable abundance data, biomass and average weight of the fish planted in the P1 pond, those of the fish extracted from P-3 and P-4 ponds, and those estimated from the rutinary control of the P-2 pond. Pond P-1 was surveyed two times because the company supplied information on a high mortality rate occurring in this pond two months after the planting. Although the company did not provide us with new density or biomass data, we thought it would be relevant to include this pond in the study in order to verify if the decrease in the population caused by such deaths could be detected acoustically. For hydroacoustic surveys, we used a Simrad EK60 echosounder (Simrad Kongsberg Maritime AS, Horten, Norway) with two split-beam circular transducers operating simultaneously at 200 kHz (ES200-7C). Both transducers were mounted on a stainless-steel frame fixed to the side of a boat, with the beam aligned horizontally, perpendicular to the navigation axis, and with each transducer, considered as channel 1 and channel 2, aimed in opposite directions. The positioning of the transducers enabled horizontal sampling, perpendicular to the direction of the boat movement. The sailing speed remained constant at around 6 km⋅h −1 using a 600 W electric outboard motor. The transducer was C. Orduna et al. 50 Aquaculture 545 (2021) 737240 placed 1 m below the surface. The pulse duration was 0.128 ms, and the repetition rate was 10 pings per second. The acoustic unit was calibrated with a calibration copper sphere using the standard calibration method (Simrad, 2004). Two different hydroacoustic designs were considered to determine the most appropriate one for this approach. These designs were named central design (C) and zigzag design (ZZ). In the central design, the transducer moved straight through the centre of the pond from one end to the other. In the zigzag design, the transducer moved from one end to the other following a zigzag trajectory. Given that the transducers were aimed in opposite directions, the entire pond could be scanned. Eight to ten passes were recorded in each pond, following the same GPS navigation route for both the central and zigzag designs. Data were stored on a PC and later processed with the Sonar-5 Pro analysis software (Balk and Lindem, 2011). Raw data were converted with the 40logR function. In order to reduce the noise coming from unwanted signals, a threshold of −60 dB was selected. Moreover, a strict criterion was selected to distinguish single targets: a minimum echo duration of 0.80 ms and a maximum of 1.6 ms (rel. to the pulse length). The maximum gain compensation was −3 dB (one-way), and the maximum phase deviation was 5. Target Strength values compensated for angular location in the beam (TS, dB re 1 m 2 ) were used for the analysis. Echo counting method was used and Single Echo Detections (SEDs) were analysed 2 and more metres away from the transducer, avoiding a possible TS deviation caused by the effect of the near field of the fish and the transducer (Tichy et al., 2003; Dawson et al., 2000; Boswell et al., 2009; Rodríguez-S´ anchez et al., 2016b; Johnson et al., 2019; Koliada et al., 2019). SEDs within −55 and −25 dB were selected for the analysis since this was the size range of the fish farmed in the ponds. All of the surveys included in this study exhibited Nv values less than 0.1, corresponding to a 1% probability of accepting multiple targets as single ones, indicating suitable conditions for fish density and biomass estimations (Warner et al., 2002; Ona and Barange, 1999; Sawada et al., 1992). Sonar-5 Pro provides fish density (hereafter “acoustic density”) both in units of volume (fish/m 3 ) and in units of area (fish/ha). We worked with units of area to make it possible to compare the data collected with those provided by the company. Sonar-5 Pro also allowed us to estimate fish biomass (kg/ha) based on the calculated density by incorporating TS-length and length-weight conversion equations to the program, as well as the average weight of the fish detected. For estimating fish length from acoustic size, we applied the De-convolution method for aspect correction. This method is applied to mobile horizontal survey where fish aspect cannot be obtained from the tracks, assuming random orientation. The SEDs obtained from the echogram are classify as follows, the largest size class contains echoes from the largest fish seen from the side aspect, the second largest size class contains the second largest fish plus the largest fish with some aspect, and so on (Duncan and Kubeˇ cka, 1995). The horizontal TS-length (Standard length, SL) and length (SL)-weight (W) conversion equations used in this study were those developed specifically for sea bass in the same facilities by Rodríguez-S´ anchez et al. (2018): TS =27.10logSL −101.23,for side aspect TS =26.96logSL −111.42,for Head and Tail aspects W=3.50⋅10−5SL2.88 with the Standard length in millimetres and the weight in grams. Before the analysis to estimate the density and biomass of the fish present in the ponds, a study was conducted to verify fish distribution in the ponds since, in case of a non-homogeneous distribution, the estimates must be corrected accordingly. Firstly, the behavioural response of fish in relation to the movements of the boat from one end to the other within the longitudinal axis of the pond was checked. The pond was divided into four parts and no statistical differences in terms of fish density were detected between these divisions. However, differences were found in the transversal axis. For that reason, the acoustic density and biomass approach proposed by Draˇ stik and Kubeˇ cka (2005) was used, regarding the comparison of an acoustic measurement of fish density and fish biomass at different distances in relation to the position of the transducer. The procedure was as follows: considering that a pass is an observation band, each pass was analysed by dividing it into layers with a thickness of one metre and perpendicular to the acoustic beam (Fortuna, 2001) (Fig. 1). In central design passes, a maximum of 4 layers from 2 to 6 m was established. In zigzag design passes, a maximum of 6 layers from 2 to 8 m was established. Density and biomass were estimated in each layer and the results were represented in a histogram which linked both parameters to the distance to the transducer to determine if fish were homogeneously distributed or if, on the contrary, there was a density gradient. When the distribution is homogeneous, the acoustic density obtained in the pass is an unbiased estimator of the density in the pond (δa =N/A; where δa is the acoustic density in the pass, N is the total number of fish and A is the area). Thus, we can use the average acoustic density obtained in that pass and extrapolate it to the total of the pond, thereby obtaining its fish abundance (N). On the contrary, when the distribution is not homogeneous, N must be estimated as n/P, where P is the average probability to detect a fish in the sampled area and n is the number of fish quantified in the sampling (Nichols et al., 2000; Farnsworth et al., 2002; Bart and Earnst, 2002; McCallum, 2005). This probability is related to the detection function, g(x), which describes the probability of detecting an animal depending on the distance perpendicular to the pass (Buckland et al., 2001; Buckland et al., 2013; Buckland et al., 2015; Marques, 2009; Marques et al., 2010; Thomas et al., 2010; Martella et al., 2012; Marques et al., 2013). In this study, one of the so-called ad hoc models was selected to correct biomass and density estimates. This model is based on the distance perpendicular to the line of the pass and uses a function based on the maximum value of the number of observed fish. This correction was applied because of its simplicity and because it is not affected by the type of probability distribution obtained (normal, binomial, etc.) (Nichols et al., 2000; Fortuna, 2001; Farnsworth et al., 2002; Cupul-Maga˜ na, 2009). The selected method calculates the observed visible proportion or fraction of the population (P) based on the acoustic density gradient obtained in the sampling: P=∑ n d=2 δad/l δmax where d is the distance from the layer to the transducer, δa d is the acoustic density obtained in the layer corresponding to that distance, l is the total number of layers in the passes, and δmax is the maximum density recorded in the pass. Thus, the estimated density (δe) of fish in the pond would be: δe =δa/P Biomass estimates (β) were calculated following the same process. The variations in the acoustic biomass and density depending on the distance to the transducer and the comparison between the estimated biomass and density in each sampling design (central vs. zigzag trajectories) were studied using an analysis of variance (ANOVA). All statistical analyses were conducted using IBM SPSS Statistics 18.0 (IBM, 2011). A significance level of 0.01 was used to contrast the null hypothesis. To select the best sampling design, we examined two factors: accuracy and precision. Accuracy concerns to how close the density and C. Orduna et al. 51 Aquaculture 545 (2021) 737240 biomass estimate are to the true population mean; precision concerns to the variability around the estimates (which may or not be accurate) (Samoilys and Carlos, 2000; Kritzer et al., 2001; Cupul-Maga˜ na, 2009; Gallardo et al., 2010; Kowalewski et al., 2015; Pais and Cabral, 2017). To verify the accuracy of the density and biomass estimated with the method developed in the study, the results were compared with the data of fish density and biomass provided by the production company (δm and βm, respectively). An agreement index was calculated which relates the estimated density and biomass to the density and biomass provided by the farmers managers (δe/δm) (Gallardo et al., 2010; Johnson et al., 2019). Bias was calculated as the absolute difference from density and biomass data provided by farmers and expressed as a proportion (Pais and Cabral, 2017): |δe −δm|/δm The Relative Standard Error (RSE) of the mean, i.e., the Standard Error of the mean (SE) divided by the mean (SE/mean, expressed as a percentage) was calculated from each set of passes of each survey and it was used to determine the precision of the values obtained for the estimates of fish density and biomass (Johnson et al., 2019). 3. Results The analysis of fish density in relation to the distance to the transducer showed that, in all ponds and all cases (passes and sampling designs), fish distribution was not homogeneous during the sampling process (Fig. 2), and the acoustic density was significantly affected by the distance to the transducer (ANOVA, p <0.01). In samples with central trajectories, a gradual increase in fish density with distance was observed, which was probably caused by the fish escape behaviour to the pond edges when the vessel approached. This behaviour was observed in all passes and all ponds regardless of fish size. In samples with zigzag trajectories, the functions that link fish density to distance were more diverse, albeit they also reflected a non-uniform distribution of density in relation to the distance to the transducer. These results were the same as those obtained in the analysis of fish biomass detected acoustically (Fig. 3). Both results confirm that acoustic estimates of average density and biomass are not homogeneous in the pond and are biased by the effect of the sampling. To correct this deviation, we calculated the probability of detecting a fish in relation to the distance to the transducer (P) in each pass. Based on the data provided by the echosounder and the P value, we calculated the estimated values of density (δe) and biomass (βe) in each pond. Table 1 presents the results for the density in each analysed pond and each sampling design. In addition to the mean acoustic density (δa) and mean estimated density (δe) values, the density values provided by the managers of the aquaculture facilities (δm) have also been included. There are no significant differences in the mean acoustic density obtained in the zigzag or central samples (ANOVA, p >0.01). Regarding fish density, the probability of detection (P) ranged between 0.49 and 0.78, and the average of all ponds was 0.6. No significant differences in P between sampling designs were recorded in any pond (central against zigzag; ANOVA, p >0.01), neither for estimated density values (δe; ANOVA, p >0.01). It can be observed that correcting acoustic density (δa) with the probability of detection P results in density values increasing significantly (δe), which indicates that acoustic density is being underestimated due to the fish avoidance behaviour caused by the disturbance of the boat during the sampling process. The mean RSE for the fish density estimate was 12.5% and ranged from 4.1% to 25.7%, being higher for zigzag designs (13.9%) with respect to central ones (11.1%). Table 2 presents the values obtained for mean acoustic biomass (βa), the estimated biomass once the correction P has been incorporated (βe) and the biomass values provided by the managers of the aquaculture facilities (βm). No significant differences were found neither in the mean Fig. 1. Analysis layering of the acoustic survey for central (a) and zigzag (b) design. C. Orduna et al. 52 Aquaculture 545 (2021) 737240 acoustic biomass, P nor estimated biomass for any case in any of the ponds with respect to the trajectory of the sampling (zigzag vs central; ANOVA; p >0.01). P ranged between 0.44 and 0.74 with an average of 0.57 for the whole group of ponds. As in the case of density, the same occurs between the biomass values calculated based on the acoustic data (βa) and those estimated incorporating the correction P (βe): the estimated values (βe) considerably increase with the correction, which shows that the acoustic value was being underestimated due to the effect of the sampling. The mean RSE for the fish biomass estimate was 15.15% and ranged from 5.9% to 22.4%, also being higher for the zigzag designs (17.5%) with respect to the central ones (12.8%). For both, the estimated density and biomass, the average precision, measured as RSE, was greater in central designs than in zigzag. In addition to the mean values estimated for density and biomass in each pond (δe and βe, respectively), Tables 1 and 2 show the data provided by the company for each variable (δm and βm, respectively) and their accuracy to the estimated values provided by the agreement index and the bias value. Fig. 4 shows these estimated density and biomass values and their adjustment for both types of trajectories in all studied ponds. Likewise, they also show the density and biomass values supplied by the company. Although no significant differences were recorded in the estimates with regards to the trajectory used (ANOVA, p >0.01), it was observed that the most accurate agreement between the estimated values of Fig. 2. Fish density in relation to the distance to the transducer. Unfilled circles for channel 1 and filled squares for channel 2. C. Orduna et al. 53 Aquaculture 545 (2021) 737240 density and biomass and those provided by the company was always obtained in central trajectories (especially in the case of density). Table 3 presents the average weight values estimated based on the individual detections (SEDs) of the hydroacoustic explorations conducted in all ponds and compares them with the values provided by the company. The average weight obtained with hydroacoustic methods matches the average weight provided by the company, with an agreement index close to 1 in all cases. The mean RSE was 12.1%, being similar for both sampling designs. Regarding the fish mortality in pond P1 (small-sized sea bass), Fig. 4 clearly shows a decrease in fish stock recorded between the exploration conducted immediately after the planting and that conducted after the deaths. This decrease is evident both in central and zigzag samples, with no significant differences in the average value of the density estimated regarding the trajectory of the sampling (ANOVA, p >0.01). The difference between the estimates recorded before and after the deaths in the pond presents fish mortality values of around 40%. Specifically, the mortality rate recorded was 39.4% in central samples and 46.2% in zigzag samples. Regarding biomass, it can be observed how it increased during two months between both explorations due to the growth of the surviving fish in the pond. However, the detected biomass growth was lower than expected, without deaths reaching mean weight and sowing density of fish. This would translate into a biomass loss of 23.3% in the estimates Fig. 3. Fish biomass in relation to the distance to the transducer. Unfilled circles for channel 1 and filled squares for channel 2. C. Orduna et al. 54 Aquaculture 545 (2021) 737240 from the central samples, and of 35.5% in the estimates from the zigzag ones. As in the previous cases, both sampling designs (central and zigzag) delivered similar average biomass values, without any significant differences (ANOVA, p >0.01). 4. Discussion Currently, split-beam hydroacoustic equipment allows reliable estimates of fish stock in large volumes of water to be obtained using vertical hydroacoustics (with the acoustic beam oriented perpendicular to the water surface). However, in small and shallow systems such as the one analysed in this study, this hydroacoustic technique faces greater limitations and requires further development to verify the reliability of the estimates obtained. This is why this study posed an unprecedented challenge in the use of hydroacoustic techniques in this kind of farming system, which not only presents problems derived from shallow depths, but also from fish farming in production systems with medium to high densities. The main issue of horizontal hydroacoustics lies within the fact that there is an important variation in the relationship between the echo returned by the fish and its size depending on the aspect in which it is ensonified. However, this problem can be solved by using split-beam equipment and adding equations that incorporate the required Table 1 Mean acoustic density (δa), mean probability of fish detection (P), mean estimated density (δe), relative standard error (RSE), density value from the aquaculture managers (δm), agreement index (δe/δm) and bias for each survey design and pond. Pond Survey design δa (fish/ha) P δe (fish/ha) RSE (%) δm (fish/ha) Agreement index Bias (%) P1 Central 52,560.3 0.49 109,903.9 10.0 110,000 0.99 0.08 Zigzag 46,473.5 0.63 78,298.0 20.9 0.71 28.82 P2 Central 28,638.5 0.50 62,852.1 25.7 70,000 0.89 10.21 Zigzag 32,709.3 0.62 53,266.3 9.4 0.76 23.90 P3 Central 35,536.3 0.78 45,151.9 4.1 47,000 0.96 3.93 Zigzag 32,612.8 0.66 48,802.7 7.7 1.04 3.83 P4 Central 25,335.7 0.52 49,336.1 4.7 49,000 1.00 0.68 Zigzag 23,604.1 0.64 37,068.3 17.4 0.75 24.35 Mean values Central 0.60 11.1 0.96 3.73 Zigzag 0.60 13.9 0.82 20.23 Table 2 Mean acoustic biomass (βa), mean probability of fish detection (P), mean estimated biomass (βe), relative standard error (RSE), density value from the aquaculture managers (βm), agreement index (βe/βm) and bias for each survey design and ponds. Pond Survey design βa (Kg/ha) P βe (Kg/ha) RSE (%) βm (Kg/ha) Agreement index Bias (%) P1 Central 2963.6 0.45 7161.8 18.3 9708 0.73 26.22 Zigzag 3699.9 0.74 5348.5 22.4 0.55 44.90 P2 Central 10,190.6 0.44 24,368.1 20.4 24,500 0.99 0.54 Zigzag 12,235.7 0.63 20,245.1 18.0 0.82 17.36 P3 Central 29,981.0 0.68 44,371.9 5.9 39,500 1.12 12.33 Zigzag 27,530.1 0.57 49,235.3 11.8 1.24 24.64 P4 Central 27,432.7 0.51 53,338.4 6.6 49,400 1.08 7.97 Zigzag 20,775.7 0.58 36,160.9 17.8 0.73 26.80 Mean values Central 0.50 12.8 0.98 11.77 Zigzag 0.60 17.5 0.84 28.43 0 20 40 60 80 100 120 140 Before C Aer C Before ZZ Aer ZZ 0 2 4 6 8 10 12 14 Before C After C Before ZZ After ZZ 1000 Fish/ha 1000 Kg/ha Fig. 4. Estimated density and biomass recorded between the exploration conducted immediately after the planting and that conducted after the deaths in pond P1. C for central and ZZ for zigzag design. Table 3 Acoustic average weight ( ω a) estimated from single echo detections (SED), relative standard error (RSE), average weight given by the aquaculture company ( ω m), agreement index and bias for each pond and survey design. Pond Survey design ω a (g) RSE (%) ω m (g) Agreement index Bias (%) P1 Central 83.1 23.3 87 0.95 4.48 Zigzag 103.3 8.1 1.18 18.73 P2 Central 399.9 7.4 350 1.14 14.25 Zigzag 332.5 8.8 0.95 5.00 Zigzag 881.5 9.6 1.03 3.70 P4 Central 987.1 13.6 1000 0.98 1.29 Zigzag 918.6 20.7 0.92 8.14 Mean Values Central 12.4 1.03 6.60 Zigzag 11.8 1.02 8.89 C. Orduna et al. 55 62 e 12 OFICINA ESPAÑOLA DE PATENTES Y MARCAS ESPAÑA IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII1 1 III ()Número de publicación: 1 279 955 8 Número de solicitud: 202131907 O Int. Cl.: G1OK 11/34 (2006.01) SOLICITUD DE MODELO DE UTILIDAD U O Fecha de presentación: 24.09.2021 O Fecha de publicación de la solicitud: 22.10.2021 O Solicitantes: UNIVERSIDAD DE SEVILLA (70.0%) Paseo de las Delicias S/N, Pabellon de Brasil 41013 Sevilla (Sevilla) ES y ECOFISHUS RESEARCH S.L.L. (30.0%) 72 Inventor/es: ENCINA ENCINA, Lourdes ; RODRÍGUEZ RUIZ, Amadora; RODRÍGUEZ SÁNCHEZ, Victoria; ORDUNA MARÍN, Carlos y CID QUINTERO, Juan Ramón 74 Agente/Representante: PONS ARIÑO, Ángel &Título: ESTRUCTURA DE ECOSONDEOS ES 1 279 955 U ES 1 279 955 U 19 OFICINA ESPAÑOLA DE PATENTES Y MARCAS ESPAÑA 11 21 Número de publicación: 1 279 955 Número de solicitud: 202131907 51 Int. CI.: G10K 11/34 (2006.01) 12 SOLICITUD DE MODELO DE UTILIDAD U 54 Título: ESTRUCTURA DE ECOSONDEOS 71 Solicitantes: UNIVERSIDAD DE SEVILLA (70.0%) Paseo de las Delicias S/N, Pabellon de Brasil 41013 Sevilla (Sevilla) ES y ECOFISHUS RESEARCH S.L.L. (30.0%) 72 Inventor/es: ENCINA ENCINA, Lourdes ; RODRÍGUEZ RUIZ, Amadora; RODRÍGUEZ SÁNCHEZ, Victoria; ORDUNA MARÍN, Carlos y CID QUINTERO, Juan Ramón 74 Agente/Representante: PONS ARIÑO, Ángel 22 Fecha de presentación: 24.09.2021 43 Fecha de publicación de la solicitud: 22.10.2021 63 ES 1 279 955 U DESCRIPCIÓN ESTRUCTURA DE ECOSONDEOS 5 OBJETO DE LA INVENCIÓN El objeto de la presente invención se relaciona con el campo técnico correspondiente a las ciencias medioambientales, concretamente perteneciente al sector de explotación pesquera. 10 La presente invención se refiere a una estructura para la sujeción de transductores, adaptable en embarcaciones neumáticas, para realizar tareas de prospección hidroacústicas mediante ecosondeos con fines científicos y/o comerciales. 15 ANTECEDENTES DE LA INVENCIÓN En los últimos años, las técnicas y herramientas pesqueras y de estudio y gestión de las poblaciones de peces se han ido orientando hacia la tecnología, especialmente hacia métodos rápidos y eficientes en la obtención de datos, como son las técnicas hidroacústicas. 20 En un principio el uso de estos equipos estaba limitado por su tamaño y coste a sectores muy especializados, como buques pesqueros, militares o científicos. Con el desarrollo de las científicas portátiles, su uso se generalizó a masas de agua más 25 reducidas como son ecosistemas continentales, y a usuarios con recursos más reducidos, como pequeñas embarcaciones pertenecientes a particulares, grupos de investigación, etc., siendo una de las metodologías con mayores perspectivas de futuro en el muestreo y la gestión de las masas de agua no vadeables. 30 Su incorporación a embarcaciones de cierto calado y con cascos rígidos no suponen ningún problema, ya que cualquier tipo de estructura puede ser atornillada a la embarcación para la sujeción de los transductores, que deben ser introducidos en el agua. Sin embargo, la problemática que surge es el empleo de estos equipos en embarcaciones 35 pequeñas, mucho más versátiles en los ecosistemas de agua dulce, debido a que las 2 U202131907 24-09-2021 ES 1 279 955 U DESCRIPCIÓN 64 ES 1 279 955 U embarcaciones pequeñas no incorporan un casco o estructuras rígidas a las que se pueda fijar la estructura. La sujeción de los transductores es vital para el funcionamiento óptimo del equipo y la 5 obtención de datos hidroacústicos precisos. La sujeción afecta directamente a las tareas de prospección hidroacústicas, por ejemplo, afecta al mantenimiento de la distancia lateral de la embarcación de forma constante y a la profundidad en la que se encuentran los transductores inmersos en el agua durante la navegación. 10 DESCRIPCIÓN DE LA INVENCIÓN La estructura soporte de ecosondeos objeto de la invención resuelve la problemática descrita debido a que no requiere ser fijada a un casco rígido de una embarcación, sino que es adaptable, debido a su geometría, al flotador lateral de una embarcación neumática, pudiendo 15 llevar a cabo labores de ecosondeos en lugares donde no pueden acceder embarcaciones de mayor tamaño que cuentan con un casco rígido. La estructura soporte comprende un bastidor, de sector longitudinal curvado que es acoplado y apoyado en el flotador lateral de la embarcación neumática. 20 Comprende también un marco, que puede ser de geometría rectangular, desplazable transversalmente y de manera guiada respecto a dicho bastidor, un eje sumergible en el agua acoplado sobre la placa y con posibilidad de ser desplazable verticalmente respecto al bastidor de forma guiada y un soporte acoplado a la parte inferior del eje, preferiblemente 25 mediante unión rosca -tuerca, que alberga los transductores para realizar las tareas de ecosondeos cuando el eje está sumergido. Asimismo, el bastidor está dotado de unos arcos curvos unidos longitudinalmente por unas primeras barras y un chasis fijado mediante elementos de fijación a los arcos curvos. 30 El chasis adicionalmente comprende unos perfiles huecos que parten de dichos arcos curvos, sustentados por otras barras, que son verticales y extendidas desde los arcos curvos para evitar la flexión de dichos perfiles huecos. Entre las barras verticales del chasis situadas en la zona delantera del bastidor se encuentra fijado un refuerzo longitudinal, siendo dicho refuerzo 35 una placa plana de la que parten unas barras hacia el marco. 3 U202131907 24-09-2021 ES 1 279 955 U 65 ES 1 279 955 U Por un lado, el marco está fabricado, al igual que el bastidor, por perfilería y barras cilíndricas, componiéndose dicho marco de un perfil superior, un perfil inferior y unos perfiles laterales. A su vez, en los extremos del perfil superior están fijadas unas barras extendidas hacia el bastidor y que encajan dentro de los perfiles huecos del mismo, con capacidad de 5 desplazamiento telescópico para regular la distancia entre el marco y el bastidor. Por otro lado, el marco incorpora otros perfiles huecos incrustados transversalmente en el perfil inferior y que recibe a las barras del chasis que parten desde el refuerzo longitudinal permitiendo el desplazamiento de dichas barras a través de ellos. Las barras que parten del 10 chasis adicionalmente incorporan una serie de orificios pasantes que reciben unos elementos fijadores, como pueden ser unos pernos de sujeción u otro tipo de elemento para fijar la posición del marco respecto al bastidor. Finalmente, en el marco se encuentran incorporados, a su vez, unos elementos guía que 15 facilitan el desplazamiento vertical del eje respecto del bastidor; y una vez se alcanza dicha posición vertical óptima para llevar a cabo las tareas de ecosondeos, es fijado, pudiendo ser dichos elementos de guía unos cierres de presión. Unido al eje, está localizado el soporte, que es preferiblemente una pletina con geometría de 20 L, en donde dicha unión puede realizarse mediante una tuerca posicionada en una pared superior de dicho soporte y que encaje con el extremo del eje, que puede ser roscado. El soporte también comprende una pared lateral que parte de la pared superior y donde están ubicados unos orificios en los que se sujetan los transductores para realizar las tareas de 25 ecosondeos y obtener datos hidroacústicos precisos y unas ranuras que permiten el paso de los cables conectados a los transductores para que se localicen alrededor del eje y faciliten las labores de ecosondeo. Los elementos que componen la estructura soporte son de acero inoxidable para poder dotar 30 a dicha estructura soporte de la rigidez necesaria para aumentar el tiempo de vida útil de la misma. La presente invención logra, además de la adaptabilidad a embarcaciones neumáticas mediante el empleo de un bastidor con geometría semicircular, una variación de la distancia 35 horizontal y vertical de los transductores que se encuentran incorporados en la estructura, 4 U202131907 24-09-2021 ES 1 279 955 U 66 ES 1 279 955 U dotando de mayor viabilidad y aumentando la eficiencia de recogida de datos durante las labores de ecosondeo. DESCRIPCIÓN DE LOS DIBUJOS 5 Para complementar la descripción que se está realizando y con objeto de ayudar a una mejor comprensión de las características de la invención, de acuerdo con un ejemplo preferente de realización práctica de la misma, se acompaña como parte integrante de dicha descripción, un juego de dibujos en donde con carácter ilustrativo y no limitativo, se ha representado lo siguiente: 10 Figura 1.- Muestra una vista en perspectiva de la estructura de ecosondeos. Figura 2.- Muestra una vista en detalle del bastidor y la placa desplazable. Figura 3.- Muestra una vista en detalle del eje desplazable. Figura 4.- Muestra una vista en detalle del soporte de la estructura de ecosondeos. 15 REALIZACIÓN PREFERENTE DE LA INVENCIÓN Con ayuda de las figuras 1 a 4 se muestra un ejemplo de realización de la estructura de ecosondeos objeto de la invención. 20 La Figura 1 muestra una vista en perspectiva de la estructura de ecosondeos, comprendiendo dicha estructura un bastidor (1) dotado de un sector longitudinal curvado con una zona trasera (26) y una zona delantera (27), destinado a estar acoplado y apoyado en un flotador lateral de una embarcación neumática. 25 Adicionalmente la estructura de ecosondeos comprende un marco (2) que es desplazable transversalmente y de manera guiada respecto al bastidor (1), un eje (3) sumergible acoplado sobre la placa (2) con posibilidad de ser desplazable verticalmente y de manera guiada respecto al bastidor (1), dotado dicho eje (3) de un extremo inferior (20); y un soporte (4), 30 preferentemente con geometría de pletina en L, acoplado al extremo inferior (20) del eje, destinado a albergar unos transductores para realizar ecosondeos cuando el eje (3) está sumergido. La Figura 2 muestra una vista en detalle del bastidor (1) y del marco (2) desplazable, en donde 35 se observa que el bastidor (1) comprende unos arcos curvos (5), unas primeras barras (6) que 5 U202131907 24-09-2021 ES 1 279 955 U 67 ES 1 279 955 U unen a los arcos curvos (5) longitudinalmente y un chasis (7) fijado mediante elementos de fijación a los arcos curvos (5). Dicho chasis (7) adicionalmente comprende unos primeros perfiles huecos (8) que parten de 5 los arcos curvos (5) y extendidos horizontalmente desde la zona trasera (26) del bastidor (1) hasta la zona delantera (27), unos perfiles verticales (9) que parten desde los arcos curvos (5) y que están fijados a los primeros perfiles huecos (8) y están destinados a evitar la flexión de dichos primeros perfiles huecos (8), un refuerzo longitudinal (10) dispuesto en la zona delantera (27) y fijado entre al menos dos perfiles verticales (9), y unas segundas barras (11) 10 extendidas desde el refuerzo longitudinal (20) hacia el marco (2). Al mismo tiempo, el marco (2), preferiblemente de geometría rectangular, está dotado de un perfil superior (12), un perfil inferior (13) y unos perfiles laterales (14), comprendiendo adicionalmente unas terceras barras (15) incorporadas en los extremos del perfil superior (12) 15 y que desplazan de forma telescópica por el interior de los primeros perfiles huecos (9) donde encajan, unos segundos perfiles huecos (16) incrustados transversalmente en el perfil inferior (13) y que permiten el desplazamiento por su interior de las segundas barras (11) del chasis (7) que reciben. 20 Incorporados en los perfiles superior e inferior (12, 13) del marco (2) se encuentran unos elementos guía (17), que pueden ser unos cierres de presión, que facilitan el desplazamiento vertical del eje (3) respecto del bastidor (1) y, una vez se ha conseguido la posición deseada, dichos elementos guía (17) tienen la capacidad de fijar la posición de dicho eje (3). 25 La Figura 2 también muestra como las terceras barras (15) comprenden unos primeros orificios (18) que están destinados a recibir a unos elementos fijadores (19) destinados a su vez a fijar la posición transversal del marco (2) respecto al bastidor (1). Con ayuda de la Figura 3 se observa el al eje (3) que comprende un extremo inferior (20) que 30 puede ser roscado y, acoplado a dicho eje (3) se encuentra el soporte (4) tal y como expone la Figura 4. Asimismo, la Figura 4 muestra una vista en detalle del soporte (4), dotado de una pared superior (22) y una pared lateral (23) que parte de perpendicularmente de dicha pared superior 35 (22). Ambas paredes (22, 23) están dotadas de unos segundos orificios (24) que permiten el posicionamiento de unos transductores para realizar tareas de ecosondeos; y, finalmente, de 6 U202131907 24-09-2021 ES 1 279 955 U 68 ES 1 279 955 U unas ranuras (25) destinadas a permitir el paso de unos cables conectados a los transductores. Finalmente, el eje (3) comprende un extremo inferior (20) que puede ser roscado y el soporte 5 (4) puede incorporar una tuerca (21) en la pared superior (22) para recibir a dicho extremo inferior (20) del eje (3) para garantizar la unión entre ambos elementos. 7 U202131907 24-09-2021 ES 1 279 955 U 69 ES 1 279 955 U REIVINDICACIONES 1.- Estructura de ecosondeos caracterizada porque comprende: - un bastidor (1) dotado de un sector longitudinal curvado con una zona trasera (26) y 5 una zona delantera (27), destinado a estar acoplado y apoyado en un flotador lateral de una embarcación neumática, un marco (2) desplazable transversalmente y de manera guiada respecto a dicho bastidor (1), un eje (3) sumergible acoplado sobre el marco (2) con posibilidad de ser desplazable 10 verticalmente y de manera guiada respecto al bastidor (1), dotado dicho eje (3) de un extremo inferior (20), y - un soporte (4) acoplado al extremo inferior (20) destinado a albergar unos transductores para realizar ecosondeos cuando el eje (3) está sumergido. 15 2.- Estructura de ecosondeos según la reivindicación 1 en el que el bastidor (1) comprende: unos arcos curvos (5), unas primeras barras (6) que unen los arcos curvos (5) longitudinalmente, fijadas a dichos arcos curvos (5), y - un chasis (7) fijado a los arcos curvos (5) mediante elementos de fijación. 20 3.- Estructura de ecosondeos según la reivindicación 2 en el que el chasis (7) está dotado de: - unos primeros perfiles huecos (8) que parten de los arcos curvos (5) y extendidos horizontalmente desde la zona trasera (26) del bastidor (1) hasta la zona delantera (27), 25 unos perfiles verticales (9) que parten desde los arcos curvos (5) y que están fijados a los primeros perfiles huecos (8) y están destinados a evitar la flexión de dichos primeros perfiles huecos (8), un refuerzo longitudinal (10) dispuesto en la zona delantera (27) y fijado entre al menos dos perfiles verticales (9), y 30 unas segundas barras (11) extendidas desde el refuerzo longitudinal (20) hacia el marco (2). 4.- Estructura de ecosondeos según la reivindicación 3 donde el marco (2) está dotado de un perfil superior (12), un perfil inferior (13) y unos perfiles laterales (14), comprendiendo dicho 35 marco (2) adicionalmente: 8 U202131907 24-09-2021 ES 1 279 955 U REIVINDICACIONES 70 ES 1 279 955 U unas terceras barras (15) fijadas en los extremos del perfil superior (12) y con dimensión suficiente para encajar en los primeros perfiles huecos (8) y desplazables longitudinalmente por el interior de dichos primeros perfiles huecos (9), unos segundos perfiles huecos (16) dispuestos transversalmente en el perfil inferior 5 (13) con dimensión suficiente para permitir el desplazamiento de las segundas barras (11) por el interior de dichos segundos perfiles huecos (16), y unos elementos guía (17) situados en el perfil superior (12) y en el perfil inferior (13) que facilitan el desplazamiento vertical del eje (3) respecto del bastidor (1) y tienen capacidad de fijar la posición de dicho eje (3) respecto de dicho bastidor (1). 10 5.- Estructura de ecosondeos según la reivindicación 4 donde las terceras barras (15) comprenden unos primeros orificios (18) destinados a recibir a unos elementos fijadores (19) y fijar la posición transversal del marco (2) respecto al bastidor (1). 15 6.- Estructura de ecosondeos según la reivindicación 1 en el que el soporte (4) comprende una pared superior (22) y una pared lateral (23) que parte perpendicularmente desde pared superior (22). 7.- Estructura de ecosondeos según la reivindicación 6 donde la pared superior (22) y la pared 20 lateral (23) comprenden unos segundos orificios (24) destinados a posicionar los transductores para realizar ecosondeos. 8.- Estructura de ecosondeos según la reivindicación 7 donde la pared superior (22) y la pared lateral (23) comprenden adicionalmente unas ranuras (25) destinadas a permitir el paso de 25 unos cables conectados a los transductores. 9.- Estructura de ecosondeos según la reivindicación 1 donde el extremo inferior (20) del eje (3) es un extremo roscado. 10.- Estructura de ecosondeos según la reivindicación 9 donde el soporte comprende 30 adicionalmente una tuerca (21) que recibe al extremo inferior (20) del eje (3). 9 U202131907 24-09-2021 ES 1 279 955 U 71 78 Aquaculture 545 (2021) 737242 Available online 27 July 2021 0044-8486/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Testing of new sampling methods and estimation of size structure of sea bass (Dicentrarchus labrax) in aquaculture farms using horizontal hydroacoustics Carlos Orduna a , b , * , Lourdes Encina a , Amadora Rodríguez-Ruiz a , Victoria Rodríguez-S´ anchez b , * a Department of Vegetal Biology and Ecology, University of Seville, Seville, Spain b EcoFishUS Research S.L.L. Seville, Spain ARTICLE INFO Keywords: Aquaculture Hydroacoustics Sea bass Offshore ABSTRACT In aquaculture, monitoring fish size and density is fundamental to improve management and profitability of fish farms. The aim of this study is to test whether horizontally aimed 200-kHz transducers are adequate to obtain fish size structure in open-sea cages in order to apply horizontal hydroacoustics as a non-intrusive and innovative technique in sea bass (Dicentrarchus labrax) farming. Several sampling strategies have been tested by placing the transducer in two different positions: outside the cage and inside the cage. In addition, two sampling approaches have been implemented: placing the transducer at a fixed position or moving it vertically. The results show that horizontal hydroacoustics is a useful technique for monitoring size distribution of sea bass in farming cages. The most adequate sampling method consists of using a vertically moving transducer positioned outside the cage, since it exhibits the narrowest size distributions with the lowest variance estimates, which matches the data provided by fish farmers. 1. Introduction Aquaculture has become a very important field on a global scale. It currently surpasses extractive fishing in terms of tons produced. It is also a growing field where both the production and number of farmed species increase every year (Apromar, 2018). This makes it necessary to improve the production processes to increase efficiency while achieving sustainability (OPP, 2009). In order to accomplish this, different factors must be optimised within farming, not only for improving economic return, but also for minimising potential ecological impacts. Among these factors, feeding strategies should be prioritised along with farming monitoring and its growth (Espinosa et al., 2006). Sea fish farming in the Mediterranean Sea area is mainly performed on two species: the sea bass (Dicentrarchus labrax, Linnaeus 1758) and the gilt-head bream (Sparus aurata, Linnaeus 1758), with a production of 81,852 T and 83,186 T in, 2018, respectively. It is mainly performed in open-sea cages (Apromar, 2018). In this kind of facility, monitoring fish size is essential to improve production control (Soliveres et al., 2014). Such information is crucial to feed optimisation and potential cannibalism detection, which is the main cause of mortality during the early farming stages and can be reduced by lowering fish density inside the cages (Hatziathanasiou et al., 2002). Although feeding efficiency has improved in recent years based on diet research of these species (Carbone and Faggio, 2016; Di Marco et al., 2017; Magalh˜ aes et al., 2017), production monitoring methodologies are still lacking. Currently, capture-dependent sampling methods are occasionally used. These are both expensive and erroneous. Additionally, they cause a high level of stress to the fish, potentially contributing to an increase in fish mortality (Espinosa et al., 2002; Conti et al., 2006; HSUS, 2008). Hydroacoustics is one of the most efficient techniques used in fish studies which enables fish detection in a capture-independent way (Simmonds and MacLennan, 2005; Kubeˇ cka et al., 2009). It has been used in numerous field studies of fish (Fabi and Sala, 2002; Neilson et al., 2003; Axenrot et al., 2004), as well as to monitor sea bass and gilt-head bream in farming cages (Soliveres et al., 2014; Soliveres, 2015). One of the major concerns when estimating fish abundances in aquaculture using hydroacoustics is the non-linear effects produced when linearity fails at very high fish densities (Simmonds and MacLennan, 2005). The shadow effect can attenuate the acoustic signals, primarily when monitoring dense shoals of fish. The fish nearest to the transducer * Corresponding authors at: Department of Plant Biology and Ecology, Faculty of Biology, University of Seville, PO Box 1095, E-41080 Seville, Spain. E-mail addresses: [email protected] (C. Orduna), [email protected] (L. Encina), [email protected] (A. Rodríguez-Ruiz), [email protected] (V. RodríguezS´ anchez). Contents lists available at ScienceDirect Aquaculture journal homepage: www.elsevier.com/locate/aquaculture https://doi.org/10.1016/j.aquaculture.2021.737242 Received 15 January 2021; Received in revised form 22 July 2021; Accepted 25 July 2021 79 Aquaculture 545 (2021) 737242 attenuate the acoustic energy in such a way that the more distant fish contribute less to the received signal (Simmonds and MacLennan, 2005; Løland et al., 2007). In addition, multiple scattering is an important feature in high fish densities involved in aquaculture (Simmonds and MacLennan, 2005). In previous studies, vertical hydroacoustics was applied (placing the acoustic beam perpendicular to the water surface) (Soliveres et al., 2014; Soliveres, 2015). However, this method presents some logistical challenges when applied in real production cages such as the ones in this study. Vertical sampling involves placing the transducer inside the cage, either at the water surface or below the cage, at a considerable depth. When the transduced is located at the surface position, faces logistical issues due to the large solid floating structure located in the middle of the cage to support the net used to protect the fish from bird predation. Second, vertical sampling from the bottom position also involves some major problems. Firstly, professional scuba divers are needed to install the transducer, since the cage depth is usually 10–15 m underwater. This translates into additional time and resources costs. Likewise, because the cage net is usually tied at the bottom (see Material and Methods), it may interfere and bias the recordings. Finally, vertical surveys require continuous recordings for a prolonged amount of time to obtain reliable estimates, due to strong daily differences in the estimated fish density (Soliveres et al., 2014; Soliveres, 2015). Given the above-mentioned difficulties inherent to vertical surveys, horizontal hydroacoustics can be an alternative technique to obtain more accurate fish size estimates in the production cages. Horizontal hydroacoustics presents important advantages compared to vertical samplings in this case. First, it solves any potential issue caused by the horizontal stratification of fish within the vertical sampling. Further, horizontal surveys can be conducted from either inside the cage, with hardly any interference with the nets, and from outside the cage, which prevents any kind of manipulation of the net or any other structure. In addition, there is no need for scuba divers, which considerably reduces installation costs. A relevant benefit obtained from this technique is the sampling speed. Unlike vertical sampling that implies handling with the protection net construction, or hard installation below cages, the horizontal survey requires less staff, without external support or interfering with the normal operation of the aquaculture facility, and, additionally, this sampling entails a short period of recording time and then requires a single equipment to analyse several cages, which greatly reduces costs. The main objective of this study was to select the most appropriate sampling technique using horizontal hydroacoustics, in order to develop a sampling and analyse protocol for size structures of fish in cages estimates. This paper describes the methodology used and the results obtained from observations of sea bass in open-sea cages at farming facilities. Various sampling methods have been tested with the aim to provide fish farmers with a functional and efficient tool to monitor the size structure of the fish in their cages. 2. Material and methods 2.1. Experimental design The experiment was performed at the facilities of the company Seaculture S.A., located in Sines, Portugal. Three cylindrical farming cages were used with sea bass (Dicentrarchus labrax) of three different sizes. According to the information provided by the fish farmers after the surveys were conducted, cage 1 contained large, market-size fish (238 mm standard average length after harvest), cage 2 also contained fish of large size (210-250 mm standard average length), and cage 3 contained smaller fish (175-200 mm standard average length). Each cage had a floating circular ring from which the net was suspended, and the net closed at the bottom in a cone shape. Cage 1 was 25 m in diameter and 13 m height, cage 2, 25 m in diameter and 14 m height, and, finally, cage 3, 12 m in diameter and 11 m height. Acoustic data was obtained using a Simrad EK60 echosounder (Simrad Kongsberg Maritime AS, Horten, Norway) with an ES200-7C circular split-beam transducer operating at 200 kHz with a transmitting power of 150 W. Before the recordings, the hydroacoustic equipment was calibrated using a copper sphere with a diameter of 13.7 mm and a known reference target strength (TS =-45 dB). Water temperature and salinity were added following the calibration protocol of the manufacturer. The mean deviation of TS, backscattered by the calibration copper sphere, was below 0.55 dB. In order to test the potential effect of the net on the acoustic signal, the calibration sphere was placed inside and outside an empty cage and recorded horizontally at different distances. There were no significant differences in the TS recorded in these positions (ANOVA; p >0.05). The sampling was conducted under favourable weather conditions with a slightly wavy surface in the water and during daylight hours (Balk et al., 2017). The transducer was mounted on a custom structure with two guide ropes placed on its ends, hanging from a boat. This special structure is aimed at maintaining the beam orientation facing horizontally the centre of the cage in a fixed position. The patent on said custom structure is pending. The distance from the transducer to the cage in the outside position remained stable by placing two boats parallel to each other firmly tied to the cage floating structure. The acoustic signals were recorded placing the transducer horizontally (with the beam axis parallel to the water surface) in two different positions in relation to the cage (Fig. 1). In position 1, the transducer was placed outside the cage, five metres away from the net, hanging from a boat. In position 2, the transducer was placed inside the cage, next to the net and pointing towards its centre. For each position, data was obtained following two sampling Fig. 1. Position of the transducer in relation to the cage. Position 1: outside the cage. Position 2: inside the cage, next to the net. C. Orduna et al. 80 Aquaculture 545 (2021) 737242 methods: either with the transducer placed at a fixed depth or with the transducer moving vertically at a controlled and constant speed through all the depth of the cage. In total, four different sampling plans were analysed combining these two variables. For the fixed sampling, the transducer took five one-minute recordings, at three different depths: 3 m, 6 m and 9 m. Thus, fifteen oneminute samples were obtained for each cage. Data from all three depths was pooled to represent one sample of the whole column. For the vertically moving sampling method, the transducer moved from the bottom to the top of the cage and back down to the bottom at a constant speed through the whole cage, thus ensonifying all its depth. Each journey up and down, lasting one minute, was considered a sample. Ten samples were taken for each cage. 2.2. Data processing Acoustic data were analysed using Sonar5-Pro (CageEye AS, Oslo, Norway). Single echo detections (SED) were stored from echograms since the fish distribution patterns did not allow for isolation of individual fish tracks. A strict data setting for single echo detection (SED) was selected in order to avoid unwanted echoes coming from noise, multiple targets, etc. The minimum target size was -70 dB. The pulse length selected was 0.128 ms. The minimum and maximum echo-lengths were 0.7 and 1.3, respectively, and the ping rate was 1 ping∙s −1 . To avoid including echoes from multiple targets placed in the same range, a phase deviation of 0.4 was selected and multiple peak echoes were rejected. Finally, in order for echoes detected in the same ping to be accepted, a separation of 100 mm was necessary between them. In order to avoid bias in the analysis due to the near-field effect, i.e., the area near the sound source where the wave instability can affect the measurements (Medwin and Clay, 1998; Dawson et al., 2000; Tichy et al., 2003; Rodríguez-S´ anchez et al., 2016), the recordings were analysed at a minimum distance of 1.3 m from the transducer. Additionally, erroneous echoes, such as ropes and net folds, recognized as fixed, repetitive and hard sounds in the echogram, located inside the cage were excluded. SEDs were classified in nine categories based on their TS, with intervals of 3 dB. Regarding the angular correction due to the swimming path of the fish, it was assumed that fish swam in a side-aspect orientation, with the incidence angle of the acoustic beam ranging from 70◦to 110◦. This assumption was made because fish at this kind of facility swim in circles, parallel to the net of the cage, mostly exposing their side aspect (Fig. 2). Such behaviour was directly observed from the surface of the cages and from video recordings obtained from cameras installed by the managers into the cages, thus supporting our assumption. Fish TS depends on morphological parameters such as length, weight, fat content, gonadal development and swim bladder. The latter is responsible for most part of the returned sound (Foote, 1980; Ona, 1990; Hazen and Horne, 2003; Knudsen et al., 2004; Frouzova et al., 2005; Rodríguez-S´ anchez et al., 2015). In order to relate the abovementioned TS categories to the length of the studied fish, we used the TS-length conversion equation developed by Rodríguez-S´ anchez et al. (2018) for the horizontal and lateral positions for sea bass (Dicentrarchus labrax) (70◦-110◦), which is TS =27.10logSL −101.23 (SL, standard length). Finally, the fish average size obtained with hydroacoustics was validated with the farmers’ estimates. 2.3. Statistical analysis The Kruskal-Wallis test was used to compare the distribution of SEDs in the different samples of each sampling plan and to detect potential errors in their repeatability. To obtain an estimate of the presumed shadow effect occurring at high fish densities (Zhao et al., 1993), echoes were analysed in terms of amount (number of echoes/beam volume) and intensity (mean TS), differentiating metre by metre. In all studied cages, it was observed that the number of detected echoes decreased progressively from the beginning to approximately the middle of the cage. In the second half of the cage, the number of detected echoes was too low, which did not correspond with the actual state of the cage (Fig. 3). Subsequently, a Mann–Whitney U test was conducted, dividing the cages in two parts: the first half comprised the distance from the front to the centre of the cage and the second half comprised the distance from the centre to the end of the cage. The heterogeneity in the number and TS values of the echoes obtained from both halves were compared, which allowed for the detection and evaluation of the potential distance effect on the intensity and number of echoes in the recordings. A permutational multivariate analysis of variance (PERMANOVA) was performed to assess the differences found in TS distribution in each sampling plan, comparing the recordings obtained with the transducer positioned inside and outside the cage and using both the fixed and vertically moving sampling methods. A Bray-Curtis Resemblance matrix was subsequently created for each studied cage using PRIMER 6 & Permanova+. The mixed model PERMANOVA (maximum permutation =9999) was used to test each data set providing “out-in” or “vertically moving-fixed” as main factors. An analysis of variance (ANOVA) was conducted to analyse the differences in the mean TS found in each cage. Each pair of cages was compared using the Bonferroni method to analyse their differences. The significance level used for all performed analyses was 95%. To compare TS data dispersion after the PERMANOVA analysis, Levene’s test was used to assess the equality of variances. The Kruskal-Wallis test was used to compare the mean TS obtained for each sampled cage using the plan with vertically moving transducers, outside the cage and the Mann-Whitney U test was performed for the pairwise comparison. All statistical analyses were performed using SPSS Statistics (Version 24.0, IBM Corp., New York, US) and Primer 6 & Permanova+(Version Fig. 2. Representation of the behaviour of fish swimming in circles through the acoustic beam, seen from above. -38 -37 -36 -35 -34 -33 -32 -31 -30 -29 -28 0 100 200 300 400 500 600 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Mean TS (dB) Volume density SED (SED/1000m3) Range (m) Fig. 3. Representation of Volume density SED (SED/1000m 3 ) in black bars, and Mean TS (dB) in grey line, vs Range (m) of the inside of an example cage. C. Orduna et al. 81 Aquaculture 545 (2021) 737242 6.0, PRIMER-E Ltd., Auckland, New Zealand). 3. Results The analysis of the SEDs obtained for each sample in each sampling plan showed no significant differences in their distributions (KruskalWallis; p >0.05). Given that all individual samples were comparable with each other, all of them were used for the subsequent analyses. For each sampling plan, the density and mean TS of the echoes obtained in the first half of the cage were compared with those obtained in the second half and it was observed that they presented significant differences (Mann-Whitney U; p <0.05). The SED density in the first half was significantly higher than in the second in all cages (Table 1 and Fig. 3). Likewise, the echoes received in the first half presented a TS significantly lower than those of the second half in all cages (Mann-Whitney U; p < 0.05, Table 1 and Fig. 3). These results suggest that the echoes of the first half of the cage had non-linear effects on those of the second half. Thus, only the first halves of the cages were selected in subsequent analyses. Nv (mean number of fish per sampled volume) values of the first halves of the cages were between 0.05 and 0.23 (Table 2). Those maximum values of 0.23 correspond to 11% probability of accepting multiple targets as single ones (Ona and Barange, 1999). Considering the specific characteristics of the experiment, ensonifying a high-density farming cage, this was considered as desirable multiple targets probability to work with. Fig. 4 displays the TS frequency distributions for the three studied cages based on the position of the transducer (inside or outside the cage) and each sampling method (fixed or vertically moving sampling). PERMANOVAs revealed significant differences in TS distribution depending on the position of the transducer (p <0.05, Table 3). Moreover, the mean TS obtained for each cage was different regardless of whether the recordings were obtained with the transducer placed outside or inside the cage (ANOVA, p <0.05). The Bonferroni test revealed differences between the mean TS obtained from all pair of cages (p <0.05). Two of the three acoustic samples taken inside the cages presented a broader range of TS (fish size) classes when compared to the samples recorded with the transducer outside the cage (Fig. 4). Similarly, the variance was lower when the sampling was performed from outside the cage (Levene’s test p <0.05, Table 2). The results show that the samples obtained with the transducer outside the cage present less dispersion in TS distribution and, therefore, they were selected for the subsequent analyses. Fig. 5 shows the TS frequency distributions for each studied cage with the transducer placed outside the cage, differentiating whether the sampling was performed in motion or in a fixed way. PERMANOVA showed significant differences in the distribution of size classes between the fixed and vertically moving sampling methods (Table 4). All sampled cages presented differences in the mean TS obtained with the different sampling techniques (ANOVA; p <0.05). The Bonferroni test revealed differences between the mean TS values obtained from all cages (p < 0.05). The size distribution showed slight differences regarding the size groups calculated with each sampling technique. The variance of the mean TS in the samples taken in motion was lower than in those obtained with the fixed approach in cage 2 and 3, and very similar in cage 1 (Table 2). Table 5 presents the mean TS values and the estimated standard length obtained for each sampled cage with the transducer placed outside the cage and moving vertically. The Kruskal-Wallis test shows significant differences in the mean TS obtained for each cage (X 22.58 = 30.431; p <0.05). The pairwise comparison revealed significant differences between the cages (Mann-Whitney U; p <0.05). Cage 1 and 2 contained larger sea bass than cage 3, and the size structure in cages 1 and 3 matched the range provided by the farmers (Table 5). Fish in cage 1 were collected for sale during the weeks following the sampling. Thus, the size data presented in Table 5 represents the actual average length of the fish. The length data from cages 2 and 3 are ranges of fish size estimates provided by the farmers and, therefore, they can include inaccuracies and possible deviations from the actual range. 4. Discussion Our results show that horizontal hydroacoustics is a useful technique to study fish in farming cages. The sampling method with the transducer outside the cage and moving vertically to cover the whole range of cage depths has proven to be the most adequate to monitor fish at these facilities since it exhibits the narrowest size distributions and variances of size estimates. Besides, the size estimates derived from this approach match the actual size ranges described by the farmers. An important concern in our analyses is that acoustic data are affected by non-linear effects such as shadow effect and multiple scattering, since echoes are more numerous and of a lower intensity within Table 1 Mann-Whitney test results for volume density of SED (SED/1000m 3 ) and TS (dB) comparison between first and second half of the three cages using two different positions of the transducer (position 1: outside the cage; and position 2: inside the cage) and fixed and vertically moving sampling types. 1st Half 2nd Half Cage Transducer Position Type Volume density SED (SED/1000m 3 ) TS (dB) Volume density SED (SED/1000m 3 ) TS (dB) Mann-Whitney U 1 Out Fixed 190.97 −40 33.47 −34 p <0.01 V.moving 264.36 −38 40.85 −32 p <0.01 In Fixed 151.56 −38 75.78 −33 p <0.01 V.moving 151.11 −31 41.02 −30 p <0.01 2 Out Fixed 168.50 −35 43.62 −33 p <0.01 V.moving 181.01 −36 32.81 −33 p <0.01 In Fixed 131.81 −30 61.39 −31 p <0.01 V.moving 161.77 −38 51.38 −27 p <0.01 3 Out Fixed 254.99 −40 57.61 −38 p <0.01 V.moving 318.81 −40 48.06 −37 p <0.01 In Fixed 1280.27 −37 230.88 −41 p <0.01 V.moving 1161.85 −40 350.21 −39 p <0.01 Table 2 Mean target strength (TS), standard deviation (SD) and mean number of fish per sampled volume (Nv) values for fixed or vertically moving and out or in position of the first half of the three cages. Fixed Vertically moving Cage Transducer Position Mean TS (dB) SD (dB) Nv Mean TS (dB) SD (dB) Nv 1 Out −35.58 2.61 0.13 −37.07 2.91 0.15 In −32.44 12.78 0.12 −35.49 2.05 0.06 2 Out −33.95 4.65 0.17 −35.71 1.01 0.20 In −38.51 9.20 0.05 −38.02 3.71 0.05 3 Out −38.86 3.39 0.23 −40.02 1.02 0.16 In −34.20 9.39 0.19 −39.05 2.22 0.23 C. Orduna et al. 82 Aquaculture 545 (2021) 737242 the first half of the cage, whereas fewer echoes with higher intensities are found within the furthest half. A similar attenuation in the acoustic signal caused by the shadow effect produced by the fish shoal was also reported in previous studies where shoals of known densities were analysed (Røttingen, 1976; Furusawa et al., 1984) or in studies conducted in natural environments (Appenzeller and Leggett, 1992). When comparing the detections obtained in both halves of the cage, our results show that the fish shoal reduces the number of echoes by 74% (average percentage of the three analysed cages). In turn, the average TS of the few echoes returned in the second half of the cage increased by TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 1 VERTICALLY MOVING TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 1 FIXED TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 2 VERTICALLY MOVING TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 2 FIXED TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 3 VERTICALLY MOVING TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 3 FIXED Fig. 4. TS distribution of the SEDs of the three cages analysed depending on the position of the transducer in relation to the cage (inside, in grey; outside, in black) for each sampling plan, differentiating between fixed and vertically moving. C. Orduna et al. 83 Aquaculture 545 (2021) 737242 8%. The decrease in the number of SEDs and the increase in TS with range are probably due to the increased uncertainty in the returned phase signal. In high-density aggregations, scattered echoes coming from fish interfere with each other when relayed back to the transducer. This may lead to a scatter signal attenuation, i.e., the loss of some SEDs, as well as to an increase within the signal amplitude, i.e., an increase in TS. Therefore, the analysis of fish size distribution was limited to the first halves of the cages studied. The recordings taken in the second halves were discarded due to the interactions in the acoustic signal. After comparing samples from inside and outside the production cages, size frequency histograms from inside were more dispersed with more size groups. This higher dispersion may be caused by the bias induced by the near-field effect i.e., because of the proximity of fish to the transducer in the inside position. In spite of having removed enough distance to avoid the near-field effect, the volume of the acoustic beam in the first ensonified metres is low due to its opening angle being 7◦. This reduced ensonification space implies that the beam might not encompass the whole fish. Thus, there can be a significant proportion of echoes coming from parts of the fish, resulting in echoes with a TS lower than the actual one (Rodríguez-S´ anchez et al., 2016), and the volume density of SED detected can be biased. Conversely, when ensonifying from a distance of five metres outside the cage, the acoustic beam is much larger inside the cage. Therefore, this problem would be greatly reduced since the whole fish can be ensonified, resulting in more representative and unbiased echoes. This can be observed in Table 1, where the presented volume density of SED from inside the cage in cage 3 is much higher than that from the outside. Cage 3 is twice as small as cage 1 and 2, and thereby the ensonified volume is too small, and measurements become unreliable. Furthermore, the fish farming system at this kind of facility involves planting fry from the same cohort in the cages. This means that they grow evenly and, therefore, they should not present large differences in size within the same cage. In light of the above, the sampling approach with the transducer outside the cage was selected as the most adequate one, both because of its higher-quality results and because of its logistical advantages and implications. The fixed and vertically moving sampling methods showed significant (albeit small) differences both in the distribution of the size histograms and in the average fish size. Samples taken in motion presented a lower dispersion than the fixed ones and the average size obtained in the different cages matched the size ranges provided by the farmers. In case of cage 1, this average length was obtained by measuring the fish directly since the fish in this cage were collected for sale during the weeks following the sampling. The difference between this length and the standard average length obtained by means of hydroacoustics was only 5 mm (Table 5), which can be considered a very reliable value. The standard-length values from cage 2 and 3 were estimations made by the farmers and, therefore, they were less precise. Cage 2 was especially problematic due to several issues which occurred during the farming process. The fish in this cage were collected 18 months after sampling, which hindered the estimation process conducted by the fish farmers. These data does not match the length range obtained by means of hydroacoustics. Cage 3 had recently been sown when the sampling took place. At that moment, farmers had good information about the length of the fish introduced into the cages. This length range matches the one obtained with hydroacoustics. Moreover, the vertically moving system allows for a whole scanning of the whole depth of the cage, unlike the fixed sampling, where the whole of the cage is estimated based on the data measured at three depths, with the potential deviations that might occur in these Table 3 PERMANOVA test performed on single echo detection (SED) distribution in vertically moving recordings for the position of the transducer regarding the cage: inside or outside, for the three cages. Cage Factor df Sum. Sq. Mean Sq. PseudoF p Unique perms 1 Transducer Position 1 2700 2700 5.576 0.019 999 2 Transducer Position 1 3024 3024 13.201 0.001 999 3 Transducer Position 1 1120 1120 4.931 0.009 998 TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 1 TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 2 TS (dB) -26/-23-29/-26-32/-29-35/-32-38/-35-41/-38-44/-41-47/-44-50/-47 SED (%) 50 40 30 20 10 0 CAGE 3 Fig. 5. TS distribution of the SEDs of the three cages analysed with the transducer placed outside the cage, differentiating between fixed (grey) and vertically moving (black). C. Orduna et al. 84 Aquaculture 545 (2021) 737242 calculations. Based on the results obtained in this experiment, the vertically moving system was selected as the most adequate method to conduct the ensonification in production cages. An additional advantage of placing the transducer outside the cage is that it avoids interactions with the farmed fish, thereby reducing the risks derived from stress. When using vertical hydroacoustic methods, the acquisition of acoustic data is performed within the cage and with the transducer placed in a vertical position (Espinosa et al., 2002; De La G´ andara and Espinosa, 2012; Soliveres et al., 2014; Soliveres, 2015). This is a major issue since the effect of the stress on fish has been proven to have negative consequences on their growth and immune system, which increases disease emergence (Pickering and Pottinger, 1989; Caruso et al., 2005; HSUS, 2008; Iwama et al., 2011). In addition, monitoring from outside the cage facilitates the sampling logistics to a great extent since it does not require manipulating the anti-bird nets used in this kind of facility. It also makes it possible to perform the sampling in a faster, easier and more economical way. We consider that our sampling protocol provides reliable results for sea bass facilities, which has been corroborated with actual size data. This method could also be easily applied to production systems of gilthead bream since they are farmed at similar facilities. It would only be necessary to use the TS-length conversion equation, also suitable for this species (Rodríguez-S´ anchez et al., 2018). This monitoring system could improve aspects such as food dose optimization, which constitutes more than 50% of the production costs (Soliveres, 2015). It could also provide relevant information to evaluate the potential effect of cannibalism, which occurs at times during the first farming stages, since the size difference has been proven to be a conducive factor in fry planting that can lead to serious predation issues (Gersanovich, 1983; Giles et al., 1986). For this predation to occur, the predator must be twice as large as the prey (Katavi´ c et al., 1989). Therefore, having a tool that provides information on fish size distribution during the first farming stages can be of great help to fish farmers when assessing losses caused by this problem. To conclude, our results have enabled us to develop a specific sampling protocol to monitor fish size structures in this type of open-sea aquaculture production system which can be implemented in an easy and quick manner. This study lays the foundation to develop a system that will allow us to know the amount and biomass of the caged farmed fish in the coming years. Funding This work was supported by the company EcoFishUS Research S.L.L., the Spanish Ministry of Economy through the program Torres Quevedo (PTQ-16-08221) and the Internal Research and Transfer Plan (PPIT) of the University of Seville. Declaration of Competing Interest None. Acknowledgements The assistance offered by the company Seaculture S.A. has been essential for this project. They allowed us to perform our experiments at their facilities, which is essential to the progress of science in this field. The company Simrad Spain also contributed to this research by supplying important material. Finally, we would like to thank Juan Ram´ on Cid Quintero for his invaluable help with field work on sampling and Miguel Tejedo Madue˜ no for his valuable comments and input. Our most sincere gratitude. References Appenzeller, A.R., Leggett, W.C., 1992. Bias in hydroacoustic estimates of fish abundance due to acoustic shadowing: evidence from day–night surveys of vertically migrating fish. Can. J. Fish. Aquat. Sci. 49, 2179–2189. Apromar, Asociaci´ on Empresarial de Acuicultura de Espa˜ na, 2018. La Acuicultura en Espa˜ na Espa˜ na, p. 2018, 94 pp. Axenrot, T., Didrikas, T., Danielsson, C., Hansson, S., 2004. Diel patterns in pelagic fish behaviour and distribution observed from a stationary, bottom-mounted, and upward-facing transducer. ICES J. Mar. Sci. 61, 1100–1104. Balk, H., Søvegjarto, B.S., Tuˇ ser, M., Frouzov´ a, J., Muˇ ska, M., Draˇ stík, V., Baran, R., Kubeˇ cka, J., 2017. Surface-induced errors in target strength and position estimates during horizontal acoustic surveys. Fish. Res. 188, 149–156. https://doi.org/ 10.1016/j.fishres.2016.12.017. Carbone, D., Faggio, C., 2016. Importance of prebiotics in aquaculture as immunostimulants. Effects on immune system of Sparus aurata and Dicentrarchus labrax. Fish Shellfish Immunol. 54, 172–178. https://doi.org/10.1016/j. fsi.2016.04.011. Caruso, G., Genovese, L., Maricchiolo, G., Modica, A., 2005. Haematological, biochemical and immunological parameters as stress indicators in Dicentrarchus labrax and Sparus aurata farmed in off-shore cages. Aquac. Int. 13, 67–73. https:// doi.org/10.1007/s10499-004-9031-5. Conti, S.G., Roux, P., Fauvel, C., Maurer, B.D., Demer, D.A., 2006. Acoustical monitoring of fish density, behavior, and growth rate in a tank. Aquaculture. https://doi.org/ 10.1016/j.aquaculture.2005.06.018. Dawson, J.J., Wiggins, D., Degan, D., Geiger, H., Hart, D., Adams, B., 2000. Point-source violations: split-beam tracking of fish at close range. Aquat. Living Resour. 13, 291–295. De La G´ andara, F., Espinosa, V., 2012. Estimaci´ on del número y la biomasa de individuos por m´ etodos acústicos en el cultivo del atún rojo (Thunnus thynnus). Collect. Vol. Sci. Pap. ICCAT 68 (1), 276–283. Di Marco, P., Petochi, T., Marino, G., Priori, A., Finoia, M.G., Tomassetti, P., Porrello, S., Giorgi, G., Lupi, P., Bonelli, A., Parisi, G., Poli, B.M., 2017. Insights into organic farming of European sea bass Dicentrarchus labrax and gilthead sea bream Sparus aurata through the assessment of environmental impact, growth performance, fish welfare and product quality. Aquaculture 471, 92–105. https://doi.org/10.1016/j. aquaculture.2017.01.012. Espinosa, V., Ramis, J., Alba, J., 2002. Evaluaci´ on de la sonda ultras´ onica EY-500 de Simrad para el control de explotaciones de dorada Sparus auratus Linnaeus, 1758. Bol. Inst. Esp. Oceanogr. 18, 15–19. Espinosa, V., Soliveres, E., Estruch, V.D., Redondo, J., Ardid, M., Alba, J., Escuder, E., Bou, M., Vicente, D., 2006. Acoustical monitoring of open mediterranean sea fish farms: problems and strategies. In: EAA European symposium on Hidroacoustics. Gandia. FAO, pp. 1–75. Fabi, G., Sala, A., 2002. An assessment of biomass and diel activity of fish at an artificial reef (Adriatic Sea) using a stationary hydroacoustic technique. ICES J. Mar. Sci. 59, 411–420. https://doi.org/10.1006/jmsc.2001.1173. Foote, K.G., 1980. Importance of the swimbladder in acoustic scattering by fish: a comparison of gadoid and mackerel target strengths. J. Acoust. Soc. Am. 67, 2084–2089. https://doi.org/10.1121/1.384452. Frouzova, J., Kubecka, J., Balk, H., Frouz, J., 2005. Target strength of some European fish species and its dependence on fish body parameters. Fish. Res. 75, 86–96. https://doi.org/10.1016/j.fishres.2005.04.011. Furusawa, M., Ishii, K., Miyanohana, Y., Maniwa, Y., 1984. Experimental investigation of an acoustic method to estimate fish abundance using culture nets. Jpn. J. Appl. Phys. 23, 101–103. https://doi.org/10.7567/JJAPS.23S1.101. Table 4 PERMANOVA test performed on single echo detection (SED) distribution for the transducer in vertically moving or fixed position, from outside the cage, for the three cages. Cage Factor df Sum. Sq. Mean Sq. PseudoF p Unique perms 1 Transducer V. Moving/Fixed 1 8416 8416 87.958 0.003 424 2 Transducer V. Moving/Fixed 1 2859 2859 26.420 0.001 857 3 Transducer V. Moving/Fixed 1 423 423 18.801 0.009 217 Table 5 Mean target strength (TS) for the vertically moving surveys performed from outside the cage for the three cages, average Standard Length (SL) of the acoustical estimates, average Standard Length (SL) given by fish farmers and corresponding standard deviation (SD). Cage Mean TS (dB) Acoustical average SL (mm) Fish farmers average SL (mm) SD (mm) 1 −37.07 233 238 35 2 −35.71 262 210–250 – 3 −40.02 181 175–200 – C. Orduna et al. 85 Aquaculture 545 (2021) 737242 Gersanovich, A.D., 1983. Factors determining variation in growth rate and size distribution in groups of young paddlefish, Polydon spathula (Polydonidae). J. Ichthyol. 23, 56–61. Giles, N., Wright, R.M., Nord, M.E., 1986. Cannibalism in pike fry, Esox lucius L.: some experiments with fry densities. J. Fish Biol. 29, 107–113. Hatziathanasiou, A., Paspatis, M., Houbart, M., Kestemont, P., Stefanakis, S., Kentouri, M., 2002. Survival, growth and feeding in early life stages of European sea bass (Dicentrarchus labrax) intensively cultured under different stocking densities. Aquaculture 205, 89–102. https://doi.org/10.1016/S0044-8486(01)00672-X. Hazen, E.L., Horne, J.K., 2003. A method for evaluating the effects of biological factors on fish target strength. ICES J. Mar. Sci. 60, 555–562. HSUS, 2008. The humane Society of the United States: the welfare of animals in the aquaculture industry. Hsus Reports 5, 24. Iwama, G.K., Pickering, A.D., Sumpter, J.P., 2011. Fish Stress and Health in Aquaculture. Cambridge University Press. Katavi´ c, I., Jug-dujakovi´ c, J., Glamuzina, B., 1989. Cannibalism as a factor affecting the survival of intensively cultured sea bass (Dicentrarchus labrax) fingerlings. Aquaculture 77, 135–143. https://doi.org/10.1016/0044-8486(89)90197-X. Knudsen, F.R., Fosseidengen, J.E., Oppedal, F., Karlsen, Ona, 2004. Hydroacoustic monitoring of fish in sea cages: target strength (TS) measurements on Atlantic salmon (Salmo salar). Fish. Res. 69, 205–209. https://doi.org/10.1016/j. fishres.2004.05.008. Kubeˇ cka, J., Hohausov´ a, E., Matˇ ena, J., Peterka, J., Amarasinghe, U.S., Bonar, S.A., Hateley, J., Hickley, P., Suuronen, P., Tereschenko, V., Welcomme, R., Winfield, I.J., 2009. The true picture of a lake or reservoir fish stock: a review of needs and progress. Fish. Res. 96, 1–5. https://doi.org/10.1016/j.fishres.2008.09.021. Løland, A., Aldrin, M., Ona, E., Hjellvik, V., Holst, J.C., 2007. Estimating and decomposing total uncertainty for survey-based abundance estimates of Norwegian spring-spawning herring. ICES J. Mar. Sci. 64, 1302–1312. https://doi.org/10.1093/ icesjms/fsm116. Magalh˜ aes, R., S´ anchez-L´ opez, A., Leal, R.S., Martínez-Llorens, S., Oliva-Teles, A., Peres, H., 2017. Black soldier fly (Hermetia illucens) pre-pupae meal as a fish meal replacement in diets for European seabass (Dicentrarchus labrax). Aquaculture 476, 79–85. https://doi.org/10.1016/j.aquaculture.2017.04.021. Medwin, H., Clay, C.S., 1998. Fundamentals of Acoustical Oceanography. Academic Press, Boston, p. 712. Neilson, J.D., Clark, D., Melvin, G.D., Perley, P., Stevens, C., 2003. The diel vertical distribution and characteristics of pre-spawning aggregations of Pollock (Pollachius virens) as inferred from hydroacoustic observations: the implications for survey design. ICES J. Mar. Sci. 60, 871. Ona, E., 1990. Physiological factors causing natural variations in acoustic target strength of fish. J. Mar. biol. Assoc. United Kingdom 70, 107–127. Ona, E., Barange, M., 1999. Single target recognition. ICES Coop. Res. Rep. 235, 28–43. OPP, 2009. Organizaci´ on de Productores Piscicultores: SustainAqua-Integrated Approach for a Sustainable and Healthy Freshwater Aquaculture. A Handbook for sustainable aquaculture. OPP, 122 pp. Pickering, A.D., Pottinger, T.G., 1989. Stress responses and disease resistance in salmonid fish: effects of chronic elevation of plasma cortisol. Fish Physiol. Biochem. 7, 253–258. Rodríguez-S´ anchez, V., Encina-Encina, L., Rodríguez-Ruiz, A., S´ anchez-Carmona, R., 2015. Horizontal target strength of Luciobarbus sp. in ex situ experiments: testing differences by aspect angle, pulse length and beam position. Fish. Res. 164, 214–222. https://doi.org/10.1016/j.fishres.2014.11.020. Rodríguez-S´ anchez, V., Encina-Encina, L., Rodríguez-Ruiz, A., S´ anchez-Carmona, R., 2016. Do close range measurements affect the target strength (TS) of fish in horizontal beaming hydroacoustics? Fish. Res. 173, 4–10. https://doi.org/10.1016/ j.fishres.2015.03.020. Rodríguez-S´ anchez, V., Rodríguez-Ruiz, A., P´ erez-Arjona, I., Encina-Encina, L., 2018. Horizontal target strength-size conversion equations for sea bass and gilt-head bream. Aquaculture 490, 178–184. https://doi.org/10.1016/j. aquaculture.2018.02.034. Røttingen, I., 1976. Studies on the long-term storage of living Saithe, Pollachius virens Linnaeus, 1758; On the Relation Between Echo Intensity and Fish Density. FiskDir. Skr. Ser. HavUnders 16, 301–304. Simmonds, D.M., MacLennan, E.J., 2005. Fisheries acoustics: theory and practice. In: Fish and aquatic resources series, second ed. 10 Blackwell science, Oxford. Soliveres, E., 2015. Estimaci´ on de biomasa de peces en granjas marinas mediante ultrasonidos. PhD Thesis. Universitat Polit´ ecnica de Valencia, Valencia, Spain. Soliveres, E., Puig, V., Perez-Arjona, I., 2014. Acoustical biomass estimation results in Mediterranean aquaculture sea cages. In: UA2014 - 2nd Int. Conf. Exhib. Underw. Acoust, pp. 1423–1428. Tichy, F.E., Solli, H., Klaveness, H.K., 2003. Non-linear effects in a 200-kHz sound beam and the consequences for target-strength measurement. ICES J. Mar. Sci. 60, 571–574. https://doi.org/10.1016/S1054. Zhao, X., Chen, Y., Jin, X., 1993. Study of the feasibility of measuring shadowing effect encountered in dense fish school by SIMRAD EK400-QD system. Mar. Fish. Res. 14, 47–53. C. Orduna et al. 86 Invention/Invención II. Hydroacoustic sampling equipment in deep sea. Equipo de muestreo hidroacústico en profundidad marina. Patent/patente. Carlos Orduna, Amadora Rodríguez Ruiz, Lourdes Encina, Juan Ramón Cid Quintero. (2024). Oficina Española de Patentes y Marcas. Nº application/solicitud P202230152, Nº publication/publicación ES2949836. 87 (15) DECLARACIONES DE PRIORIDAD: PAÍS DE ORIGEN: CÓDIGO PAÍS: NÚMERO: FECHA: (16) REMISIÓN A UNA SOLICITUD ANTERIOR: PAÍS DE ORIGEN: CÓDIGO PAÍS: NÚMERO: FECHA: (17) AGENTE DE PROPIEDAD INDUSTRIAL: APELLIDOS: PONS ARIÑO NOMBRE: ANGEL CÓDIGO DE AGENTE: 0499/5 NÚMERO DE PODER: (18) DIRECCIÓN A EFECTOS DE COMUNICACIONES: DIRECCIÓN ASOCIADA AL PRIMER SOLICITANTE DOMICILIO: LOCALIDAD: CÓDIGO POSTAL: PAÍS RESIDENCIA: CÓDIGO PAÍS: TELÉFONO: FAX: CORREO ELECTRÓNICO: MEDIO PREFERENTE DE COMUNICACIÓN (19) RELACIÓN DE DOCUMENTOS QUE SE ACOMPAÑAN: DESCRIPCIÓN: [✓] N.º de páginas: 5 REIVINDICACIONES: [✓] N.º reivindicaciones: 6 DIBUJOS: [✓] N.º de dibujos: 1 RESUMEN: [✓] N.º de páginas: 1 FIGURA(S) A PUBLICAR CON EL RESUMEN: [✓] N.º de figura(s): ARCHIVO DE PRECONVERSION: [✓] DOCUMENTO DE REPRESENTACIÓN: [ ] N.º de páginas: JUSTIFICANTE DE PAGO (1): [✓] N.º de páginas: 1 JUSTIFICANTE DE PAGO (2): [✓] N.º de páginas: 1 LISTA DE SECUENCIAS PDF: [ ] N.º de páginas: ARCHIVO PARA LA BUSQUEDA DE LS: [ ] OTROS (Aparecerán detallados): (20) EL SOLICITANTE SE ACOGE A LA REDUCCIÓN DE TASAS PARA EMPRENDEDORES PREVISTA EN EL ART. 186 DE LA LEY 24/2015 DE PATENTES Y, A TAL EFECTO, APORTA LA SIGUIENTE DOCUMENTACIÓN ADJUNTA: [ ] (21) NOTAS: (22) FIRMA: FIRMA DEL SOLICITANTE O REPRESENTANTE: 50534279J ANGEL PONS (R: A28750891) LUGAR DE FIRMA: MADRID FECHA DE FIRMA: 24 Febrero 2022 94 Justificante de pago de TASA en materia de Propiedad Industrial CODIGO 511 Modelo 791 Importe en Adeudo en NRC Asignado: Ingreso 87,03 7915111222934000000001 Fecha: Firma: Declarante 24/02/22 16:03 ANGEL PONS ARIÑO Número Tipo Unidades: Clave: Modalidad Expediente: Concepto: Titular del expediente si es distinto del pagador: Año: Importe: Autoliquidación Código de barras: UNIVERSIDAD DE SEVILLA - ECOFISHUS P IE01 2022 SOL. DE INVENCION O REHABILITACIÓN POR INTERNET 1 87,03 µ9099921002001882654044415Ä 909992100200188265404441 7915111222934 2022 Ejercicio: Localizador: Pagador: NIF: Apellidos y Nombre o Razón Social: Identificación [email protected]Correo electrónico: ANGEL PONS ARIÑO Código de Agente o Representante 0499 Apellidos y Nombre o Razón Social: NIF: 50534279J Agente o Representante legal El pago de esta tasa no implica la efectiva realización del trámite, si éste exige, además, la aportación de una solicitud o escrito (como por ej.: renovación de marca, división de solicitud, cesiones, etc.). El pago de una anualidad de patente o de modelo de utilidad no requiere la presentación de una solicitud, pues el pago es suficiente siempre que se haya realizado conforme a la legislación vigente. Modelo 791 NIF: Q2820005C TBP OEPM, Paseo de la Castellana, 75, 28071 Madrid Pasarela2 95 Justificante de pago de TASA en materia de Propiedad Industrial CODIGO 511 Modelo 791 Importe en Adeudo en NRC Asignado: Ingreso 593,65 7915111222943000000001 Fecha: Firma: Declarante 24/02/22 16:05 ANGEL PONS ARIÑO Número Tipo Unidades: Clave: Modalidad Expediente: Concepto: Titular del expediente si es distinto del pagador: Año: Importe: Autoliquidación Código de barras: UNIVERSIDAD DE SEVILLA - ECOFISHUS P IE04 2022 SOLICITUD I.E.T. INTERNET 1 593,65 µ909992100200188265404667QÄ 909992100200188265404667 7915111222943 2022 Ejercicio: Localizador: Pagador: NIF: Apellidos y Nombre o Razón Social: Identificación [email protected]Correo electrónico: ANGEL PONS ARIÑO Código de Agente o Representante 0499 Apellidos y Nombre o Razón Social: NIF: 50534279J Agente o Representante legal El pago de esta tasa no implica la efectiva realización del trámite, si éste exige, además, la aportación de una solicitud o escrito (como por ej.: renovación de marca, división de solicitud, cesiones, etc.). El pago de una anualidad de patente o de modelo de utilidad no requiere la presentación de una solicitud, pues el pago es suficiente siempre que se haya realizado conforme a la legislación vigente. Modelo 791 NIF: Q2820005C TBP OEPM, Paseo de la Castellana, 75, 28071 Madrid Pasarela2 96 D E S C R I P C I Ó N EQUIPO DE MUESTREO HIDROACÚSTICO EN PROFUNDIDAD MARINA OBJETO DE LA INVENCIÓN 5 El objeto de la presente invención se relaciona con el campo técnico de las ciencias medioambientales, concretamente, con el sector correspondiente a la investigación de ecosistemas acuáticas. 10 La presente invención se refiere a un equipo de muestreo acústico con capacidad de flotación sobre la superficie del agua y que, mediante medios de regulación, permite el desplazamiento en un plano vertical de una estructura que sustenta una serie de transductores, sin balanceo horizontal de la misma; y que se encuentra sumergida en el agua para realizar muestreos en jaulas de acuicultura de mar abierto, con el fin de 15 desarrollar un sistema de estimas de biomasa y densidad de peces en ese tipo de instalaciones. ANTECEDENTES DE LA INVENCIÓN 20 La acuicultura o acuacultura es el conjunto de actividades, técnicas y conocimientos de crianza de especies acuáticas vegetales y animales. Es una importante actividad económica en lo referido a la producción de alimentos, materias primas de uso industrial y farmacéutico y organismos vivos para repoblación. 25 Los sistemas de cultivo son muy diversos, de agua dulce o agua de mar, desde el cultivo directamente en el medio acuático (mar abierto) hasta en instalaciones cuyas condiciones de entorno son totalmente controladas. Los cultivos más habituales corresponden a organismo planctónicos, macroalgas, 30 moluscos y crustáceos. Los profesionales encargados de realizar estas actividades son ingenieros pesqueros, zootecnistas, acuícolas y biólogos marinos. Por otro lado, el desarrollo de la acuicultura se realiza empleando, principalmente, jaulas para el control de la cantidad de fauna y flora marina en una ubicación concreta, siendo 35 97 éstas las primeras estructuras usadas por los pescadores para mantener vivos a los peces capturados durante la jornada de trabajo hasta su venta en el mercado. Dichas jaulas también sirven para mantener controlado un ecosistema de especies que podrían estar en riesgo de extinción, dentro del mar abierto, en un entorno en el que se 5 puedan medir diversas variables. Sin embargo, el sector de la acuicultura en jaulas de mar abierto encuentra su mayor dificultad para el control de la producción en el desconocimiento de la cantidad y peso de los peces que cultivan en sus jaulas. La medición de estos parámetros es realizada 10 mediante estimaciones, dando lugar a fallos numéricos que alteren el análisis de las condiciones que se pretenden medir. Las estimaciones son necesarias debido a que los aparatos empleados para la medida de las variables mencionadas están sujetos a cambios de posición fruto de la fuerza del 15 oleaje, debido a que dichas mediciones se realizan en mar abierto. DESCRIPCIÓN DE LA INVENCIÓN La presente invención pretende resolver los problemas planteados mediante la 20 implementación de una plataforma flotante en la que se encuentra una estructura de sustentación de ecosondas cuya orientación horizontal no varíe de manera que los cálculos realizados para realizar un análisis de los parámetros de cantidad de fauna y flora, densidad de población y/o peso no sean erráticos. 25 Es necesario que, para obtener esos datos con el menor error posible, el transductor, pieza de la ecosonda que emite los ultrasonidos, se desplace en orientación horizontal fija, a lo largo de un plano vertical por dentro de la jaula y a una velocidad constante. Esta configuración es necesaria para obtener datos hidroacústicos de calidad que puedan ser analizados con el fin de ofrecer estimas de densidad y biomasa de los 30 cultivos de fauna y flora objetivos. El equipo objeto de la invención comprende una plataforma de geometría esencialmente rectangular y, de al menos, 1,5m2 de superficie que está formada por un panel náutico superior que puede ser de madera y, acoplado inferiormente a dicho panel de madera, 35 98 otro panel de poliestireno expandido, dotando a la plataforma de la flotabilidad característica sobre la superficie del mar, adicionalmente también presenta una estructura soporte sumergible y con posibilidad de ser desplazable en un plano vertical respecto de dicha plataforma y un soporte de ecosondas acoplado mediante elementos de sujeción a la estructura soporte y que aloja unas ecosondas con el fin de obtener 5 datos sobre la fauna y flora marina de manera precisa. Asimismo, la estructura soporte está constituida por al menos una placa de sustentación, fijada al soporte de ecosondas; y unos perfiles verticales extendidos en un plano vertical desde los extremos de dicha placa de sustentación. 10 Acoplados en la plataforma, al menos se localiza un regulador de altura, ubicado en la zona central de dicha plataforma; y una serie de dispositivos de guiado, situados en las zonas laterales de dicha plataforma, pudiendo ser el regulador de altura un cabestrante eléctrico y los dispositivos de guiado unos cabestrantes manuales. 15 Los cabestrantes presentan una serie de cuerdas enrolladas a su eje giratorio y fijadas a él por uno de sus extremos, y, mediante giro de dicho eje se enrolla o desenrolla las cuerdas tensándolas o destensándolas, extendidas verticalmente hacia la profundidad del mar, contando la plataforma una serie de orificios a través de los cuales traspasan 20 las cuerdas de los cabestrantes manuales y del cabestrante eléctrico hacia el fondo marino mencionado. Las cuerdas de los dispositivos de guiado, a su vez, traspasan a los perfiles verticales de la estructura soporte ejerciendo de guía de movimiento del soporte de ecosondas y en consecuencia de la estructura soporte. 25 Es necesario resaltar que las cuerdas de los dispositivos de guiado presentan en su extremo opuesto al que están vinculadas una serie de contrapesos que apoyan sobre la superficie del fondo marino, a modo de ancla, fijando la posición de flotación de la plataforma y tensando dichas cuerdas laterales ejerciendo de guía 30 Con esta configuración, dichas cuerdas junto con los contrapesos que apoyan sobre la superficie del fondo marino combinado con la tensión de las cuerdas laterales de manera manual por parte de un usuario para fijar completamente la posición de flotación de la plataforma del equipo establecen la orientación fijada de una de las caras de la placa de sustentación de transductores en el plano horizontal, y mediante el movimiento de 35 99 ascenso y descenso de la cuerda vinculada al cabestrante eléctrico se consigue regular la altura del soporte de ecosondas de manera guiada, y, por tanto, del transductor, posibilitando una captación de datos de forma directa y precisa debido a que la orientación de los transductores se produce en un plano horizontal y no en un plano vertical, facilitando las labores hidroacústicas y la propagación de las ondas de una 5 manera más eficiente que si estuvieran dichos transductores orientados verticalmente respecto a la superficie marina. DESCRIPCIÓN DE LOS DIBUJOS 10 Para complementar la descripción que se está realizando y con objeto de ayudar a una mejor comprensión de las características de la invención, de acuerdo con un ejemplo preferente de realización práctica de la misma, se acompaña, como parte integrante de dicha descripción, un juego de dibujos en donde, con carácter ilustrativo y no limitativo, se ha representado lo siguiente: 15 Figura 1.- Muestra una vista en perspectiva del equipo de muestreo hidroacústico en profundidades marinas. REALIZACIÓN PREFERENTE DE LA INVENCIÓN 20 Con ayuda de la Figura 1 se describe una realización preferente del equipo objeto de la invención. La Figura 1 muestra una vista en perspectiva de un equipo de muestreo hidroacústico 25 destinado a estar situado sobre una jaula de cultivo de fauna y flora marina que comprende una plataforma (1) destinada a flotar sobre la superficie del agua dotada de una zona central (2) y unas zonas laterales (3), una estructura soporte (4) sumergible y con posibilidad de ser desplazable verticalmente respecto de la plataforma (1), y un soporte de ecosondas (5) acoplado mediante elementos de sujeción a la estructura 30 soporte (4) y que está destinado a alojar unas ecosondas (19) para realizar muestras hidroacústicas de la fauna y flora marina. Acoplado en la plataforma (1), inferior o superiormente al menos un regulador de altura (6) acoplado en las proximidades de la zona central (2) de la plataforma y asociado al 35 100 soporte de ecosondas (5) mediante un primer elemento de sujeción (13), regulando la altura de sumersión del soporte de ecosondas (5) y, por tanto, de la estructura soporte (4) respecto de la plataforma (1). Al mismo tiempo, el equipo de muestreo hidroacústico comprende adicionalmente unos 5 dispositivos de guiado (14) acoplados en las zonas laterales (3) de la plataforma (1) y vinculados a los laterales de la estructura soporte (4) mediante unos segundos elementos de sujeción (16), posibilitando el desplazamiento de manera guiada en un plano vertical de dicha estructura soporte (4) sin movimiento de balanceo transversal de la estructura soporte (4) respecto de dicho plano vertical. 10 La estructura soporte (4) está dotada de, al menos, una placa de sustentación (7) que comprende un primer extremo (8) y un segundo extremo (9), que está acoplada al soporte de ecosondas (5), y unos perfiles verticales (10) huecos asociados a sendos extremos (8 ,9) de la placa de sustentación (7) y que son traspasados por los segundos 15 elementos de sujeción (16). La Figura 1 también ayuda a observar como el equipo de muestreo hidroacústico está dotado de unos contrapesos (18) vinculados a los segundos elementos de sujeción (16) que están destinados a situarse sobre el fondo marino evitando el movimiento de la 20 plataforma (1) por acción del oleaje del mar y a tensar a dichos segundos elementos de sujeción (16). Finalmente, el regulador de altura (6) es, preferentemente, un cabestrante eléctrico que está dotado de un primer eje giratorio (12) y el primer elemento de sujeción (13) es una 25 cuerda enrollable alrededor de dicho primer eje giratorio (12); y los dispositivos de guiado (14) son, preferiblemente, cabestrantes manuales dotados de unos segundos ejes giratorios (15) y los segundos elementos de sujeción (16) son cuerdas enrollables alrededor de dichos segundos ejes giratorios (5). 30 A modo de facilitar el guiado y el acople del regulador de altura (6) y de los dispositivos de guiado (14), la plataforma (1) puede comprender unos orificios (17) pasantes localizados en las proximidades de la zona central (2) y zonas laterales (3) de dicha plataforma (1) y que permiten el paso de los primeros y segundos elementos de sujeción (13 ,16) del regulador de altura (6) y de los dispositivos de guiado (14) respectivamente. 35 101 R E I V I N D I C A C I O N E S 1.- Equipo de muestreo hidroacústico en profundidades marinas destinado a tomar muestras de flora y fauna marina que comprende:  una plataforma (1) destinada a flotar sobre la superficie del agua dotada de una 5 zona central (2) y unas zonas laterales (3),  una estructura soporte (4) sumergible y con posibilidad de ser desplazable en un plano vertical respecto de la plataforma (1),  al menos un soporte de ecosondas (5) acoplado mediante elementos de sujeción a la estructura soporte (4) y que está destinado a alojar unas ecosondas (19) 10 para realizar muestras hidroacústicas de la fauna y flora marina, y  al menos un regulador de altura (6) acoplado en las proximidades de la zona central (2) de la plataforma y asociado al soporte de ecosondas (5) mediante un primer elemento de sujeción (13), regulando la altura de sumersión del soporte de ecosondas (5) y, por tanto, de la estructura soporte (4) respecto de la 15 plataforma (1); caracterizado el equipo de muestreo hidroacústico por que comprende adicionalmente unos dispositivos de guiado (14) acoplados en las zonas laterales (3) de la plataforma (1) y vinculados a los laterales de la estructura soporte (4) mediante unos segundos elementos de sujeción (16), posibilitando el desplazamiento de manera guiada en un 20 plano vertical de dicha estructura soporte (4) sin movimiento de balanceo transversal de la estructura soporte (4) respecto de dicho plano vertical. 2.- Equipo de muestreo hidroacústico según la reivindicación 1 en el que la estructura soporte (4) está dotada de: 25  al menos una placa de sustentación (7) que comprende un primer extremo (8) y un segundo extremo (9), que está acoplada al soporte de ecosondas (5), y  unos perfiles verticales (10) huecos asociados a sendos extremos (8 ,9) de la placa de sustentación (7) y que son traspasados por los segundos elementos de sujeción (16), 30 3.- Equipo de muestreo hidroacústico según la reivindicación 1 que comprende adicionalmente unos contrapesos (18) vinculados a los segundos elementos de sujeción (16) que están destinados a situarse sobre el fondo marino evitando el movimiento de 102 la plataforma (1) por acción del oleaje del mar y a tensar a dichos segundos elementos de sujeción (16) 4.- Equipo de muestreo hidroacústico según la reivindicación 1 en el que el regulador de altura (6) es un cabestrante eléctrico dotado de un primer eje giratorio (12) y el primer 5 elemento de sujeción (13) es una cuerda enrollable alrededor de dicho primer eje giratorio (12). 5.- Equipo de muestreo hidroacústico según la reivindicación 1 en el que los dispositivos de guiado (14) son cabestrantes manuales dotados de unos segundos ejes giratorios 10 (15) y los segundos elementos de sujeción (16) son cuerdas enrollables alrededor de dichos segundos ejes giratorios (5). 6.- Equipo de muestreo hidroacústico según la reivindicación 1 en donde la plataforma (1) comprende unos orificios (17) pasantes localizados en las proximidades de la zona 15 central (2) y zonas laterales (3) de dicha plataforma (1) y que permiten el paso de los primeros y segundos elementos de sujeción (13 ,16). 103 Fishes 2023,8, 227 2 of 10 TS–length equations for most of the farmed fish species have been developed, such as Atlantic salmon (Salmo salar), brown trout (Salmo trutta), perch (Perca fluviatilis), carp (Cyprinus carpio), gilt-head seabream (Sparus aurata), or sea bass (Dicentrarchus labrax) [22–24]. These equations allow us to use hydroacoustics to know the size of fish in a non-intrusive way. The second approach involves computer vision, which is rapidly developing as a system for obtaining fish measurements without any kind of manipulation, using stereo vision underwater cameras. This technology uses images to measure the fish size in aquaculture facilities through computer image analysis [ 3 , 12 , 25 – 29 ]. These two techniques have proven to be a good adjustment in fish size measures, but both usually need to use a previously calculated length–weight relationship to provide weight estimates from TS or stereoscopic images [ 11 , 30 ]. This length–weight relationship can vary with fish growth rate, due to different external or intrinsic factors [31]. The European sea bass represents one of the most important farmed fish species in southern Europe [ 32 ]. However, the potential effects of seasonal variation on length–weight equations and their implication for biomass estimation are largely unexplored. Water temperature and its significant seasonal differences are critical external factors that affect most biochemical and physiological processes in aquatic organisms, resulting in one of the most important environmental–physical factors. Increasing water temperature has a significant influence on fish growth, determining higher food intake and growth rates within the non-stressful thermal range for each species [ 33 – 39 ]. In this sense, adult sea bass species migrate to estuaries and coastal lagoons for spawning, where larvae can find optimal growth conditions and plankton prey due to the early warming of these shallow waters in spring [ 40 ]. Variation in salinity is a disturbing factor influencing the metabolism and growth of the European sea bass. Despite being a euryhaline fish, acute changes in water salinity can cause complex metabolic changes in sea bass. On the one hand, a decrease in salinity suggests a reduction in the costs associated with osmoregulation. However, salinity changes in the environment involve more than just osmoregulation costs. The gills are the main osmoregulation organ, but because of their involvement in respiration, they are at the same time the main site of water and ion leakage. Thus, as osmoregulatory costs are linked to metabolic activity through ventilation, the portion of energy necessary to compensate for the ventilation-related osmotic and ionic loss will increase, as fish metabolic expenditure rises [ 37 , 40 – 46 ]. Dissolved oxygen is another crucial factor for fish survival and growth in aquaculture farms. However, significant inputs of nitrogen and phosphorus due to the decomposition of uneaten feed, feces, and metabolic excretion by cultured animals can lead to substantial variation in oxygen levels. This leads to the production of oxygen by photosynthetic organisms. Furthermore, substantial amounts of oxygen are consumed in plankton and fish respiration. Therefore, dissolved oxygen concentration is a parameter closely monitored in aquaculture farms, as sudden drops in its concentration, especially at night, could result in significant fish mortality rates [35,36,40,47,48]. The aim of this study is to compare the seasonal length–weight relationships of sea bass (Dicentrarchus labrax) between two farming systems presenting different environmental conditions, to assess the benefits of using specific equations for each facility and season of the year. 2. Materials and Methods 2.1. Samplings Two different sea bass farming facilities were selected for this study (Figure 1). The first one was located in the Guadalquivir River estuary, with brackish water, in Seville, Spain. This farm system was based on rectangular inland ponds, with a surface area of 2.700 m2and a volume of 3150 m3. The second farming facility was in the Atlantic Ocean, in the Setúbal district, Portugal. This farm consisted of cylindrical offshore cages, of an approximate volume of 6.500 m 3 . In both cases, the fish farm managers set the seeding and management of the facilities aiming at obtaining a density of around 5 fish/m 3 . Moreover, fish were fed until they were satiated during the whole breeding process. Samples were 110 Fishes 2023,8, 227 3 of 10 obtained seasonally for one year. About 200 sea bass were measured for each sampling. Fish were measured in the field for standard length (SL) to the nearest millimeter, and weighed (W, wet weight) to the nearest gram. All the specimens were measured during the harvest process before sales, with all of them being approximately the same age. Daily records of temperature and salinity were provided by the managers of the facilities, except for offshore salinity due to its low variation. Previous studies monitoring water quality in the study area where the offshore facilities were located reported average salinity values of 36 PSU, with little interand intra-annual variability [ 49 , 50 ]. Thus, we considered this parameter as constant. Figure 1. Location of the sea bass farming facilities. 2.2. Data Analysis All statistical analyses were conducted in the statistical package R [ 51 ]. The power function W = aSL b is commonly used to predict fish weight from the fish length, where W is the total body weight of the fish in grams, SL is the standard length in millimeters, a is the regression intercept, and bis the regression slope. This length–weight model was transformed into a linear model by applying a log 10 transformation to both sides, resulting in the equation log(W) = log(a) + blog(SL). Using the logarithmic form of the equation, length–weight relationships were estimated for each season in each farming facility. The values of the regression slope indicate isometric (b= 3) or allometric (positive if b> 3, negative if b< 3) growth. In order to investigate the seasonal growth pattern in each farming facility, a Student’s t-test was performed to evaluate if the mean bvalue was significantly different from 3 and to identify the type of growth. Moreover, significant differences in the intercept and slope were analyzed among seasons within each facility using analysis of covariance (ANCOVA) with log(W) as the response variable, log(L) as the continuous covariate, and “season” as the categorical variable. Seasonal differences were also tested in length–weight relationships between facilities by fitting an ANCOVA separately for each season with “facility” as the categorical variable. The assumptions of normality and homoscedasticity were verified through the inspection of the quantile– quantile plot and the residual-against-fit plot, respectively. 111 Fishes 2023,8, 227 4 of 10 In order to verify the actual differences that may result from using different equations, a practical case was performed to estimate fish biomass through their application. Length– frequency distribution was simulated for an inland pond and an offshore cage of the two studied farming facilities. Fish length–frequency distribution was obtained by simulating random length values for each size interval of the target strength (TS) frequency distribution from previously collected acoustic data, specifying the length range and the number of individuals for each range [ 5 , 10 ]. The predictions of individual weight values and confidence intervals given the length values were achieved by fitting the specific regression for each season using the predict function from the “stats” R package. The predicted weights were then back-transformed to the original scale and corrected for bias using the logbtcf function from the “FSA” R package [52]. 3. Results The seasonal sea bass length–weight regressions of the two studied farming facilities are shown in Table 1and Figure 2. The r 2 values ranged from 0.844 for the inland farm in winter to 0.957 for the offshore farm in autumn and were all highly significant (p ≤ 0.001). Table 1. Seasonal seas bass length–weight regressions for the offshore and inland farms. Facility Season n SL (mm) W (g) log(a) 95% C.I. log(a) b 95% C.I. b r2Growth Offshore Winter 196 308 ±28 345 ±92 −4.740 −4.990; −4.490 2.920 2.820–3.021 0.944 * Isometric Offshore Spring 197 303 ±35 470 ±179 −5.292 −5.564; −5.019 3.201 3.091–3.311 0.944 * Allometric (+) Offshore Summer 200 300 ±30 426 ±120 −4.262 −4.528; −3.996 2.777 2.700–2.884 0.929 * Allometric (−) Offshore Autumn 197 309 ±37 397 ±148 −5.193 −5.424; −4.961 3.121 3.028–3.214 0.957 * Allometric (+) Inland Winter 198 346 ±22 678 ±120 −3.736 −4.132; −3.340 2.585 2.429–2.742 0.844 * Allometric (−) Inland Spring 200 331 ±22 575 ±107 −3.940 −4.305; −3.576 2.656 2.512–2.801 0.868 * Allometric (−) Inland Summer 166 336 ±34 526 ±149 −5.087 −5.366; −4.809 3.085 2.975–3.196 0.949 * Isometric Inland Autumn 200 364 ±32 587 ±157 −4.763 −5.057; −4.470 2.937 2.823–3.052 0.928 * Isometric Number of specimens (n); mean ± standard deviation; standard length (SL) and weight (W); intercept of regression line (a); slope of regression line (b); confidence interval (C.I.); regression coefficient (r2); * significant p≤0.001. Sea bass presented different growth types within each facility. Fish from the inland facility showed isometric growth (b= 3) in summer and autumn and negative allometric growth (b< 3) in winter and spring. On the other hand, sea bass from the offshore facility presented isometric growth (b= 3) in winter, negative allometric growth (b< 3) in summer, and positive allometric growth (b> 3) in spring and autumn (Table 1). Significant differences were found in the intercept and slope among seasons for the inland (ANCOVA, F12.791;p< 0.001) and offshore (ANCOVA, F13.414;p< 0.001) facilities. Seasonal differences test in the length–weight relationships between facilities for each season also showed significant differences for all of them: autumn (ANCOVA, F 5.772; p< 0.001), spring (ANCOVA, F 30.658 ;p< 0.001), summer (ANCOVA, F 15.642 ;p< 0.001) and winter (ANCOVA, F12.881;p< 0.001). The results of the total biomass simulation for two cases of both studied facilities, i.e., an inland pond and an offshore cage, are shown in Figure 3. 112 Fishes 2023,8, 227 5 of 10 Figure 2. Seasonal seas bass length–weight relationship for the offshore and inland farms. Green area indicates the 95% confidence interval. 113 Fishes 2023,8, 227 6 of 10 Figure 3. Biomass simulation for the two studied facilities: an inland pond ( a ) and an offshore cage (b). Error bars represent the estimated 95% confidence intervals. 4. Discussion In this study, we compared two different aquaculture facilities located only 250 km apart but presenting significant differences in water temperature and salinity. This is mainly because the offshore cages are situated in the ocean, while the ponds are located inland, specifically in an estuary area. In aquaculture facilities, such as those selected for this study, food and oxygen concentration in the water are not limiting factors. In offshore cages, automatic feeders and a camera system controlled by an operator stop feeding when leftover food is detected. While this system does not allow for the precise dosing of feed and results in significant feed waste, it ensures that the fish are fed until they are fully satiated. In inland ponds, fish have feeders that release food on demand when fish approach the feeding area and activate a floating ball. Workers also provide extra feed during times of high demand, ensuring that fish are fully satiated, although with little precision and significant waste. Moreover, the oxygen concentration was monitored in both types of facilities studied, with particular attention paid to the inland ponds, where compressed oxygen systems were available to inject oxygen into the ponds in case of a sudden drop that could have caused fish mortality. Temperature and salinity are the two most influential environmental factors on the growth and metabolism of farmed fish since they are hard to control [ 34 – 36 , 53 , 54 ]. In this regard, the water in offshore cages primarily consisted of coastal water from the Atlantic Ocean that was constantly renewed through the cage’s net wall. As a result, the annual average temperature remained around 15.9 ◦ C with a low fluctuation of 7.6 ◦ C, and the salinity remained constant at around 36 ppt (Figure 4), which is similar to the natural habitat of these fish. By contrast, estuaries are complex aquatic systems where river and sea dynamics interact, creating unique conditions that evolve in space and time across various scales [ 55 ]. This interaction varies continuously throughout the estuary with each tidal cycle, neap, and spring tides, and it is drastically altered during brief periods associated with the occurrence of river floods [ 56 ]. This environmental variation, combined with the fact that inland ponds have much lower water renewal rates (just through a channel that runs across the pond) and much shallower depths, can cause great fluctuations in temperature and salinity, as they lack the buffering effect of the ocean, even though the climates of both facilities are similar. In inland ponds, water temperature ranged from 10 ◦ C in winter to 30 ◦ C in summer, with an average temperature of 20.7 ◦ C, while salinity averaged 18 ppt and reached a maximum of 25 ppt in autumn (Figure 4), always being lower than in offshore cages. Previous experimental studies suggest that the preferred temperatures 114 Fishes 2023,8, 227 7 of 10 for this species range between 20 and 25 ◦ C [ 37 , 40 ]. Under experimental conditions, the sea bass growth rate increased with increasing salinity, reaching a maximum at 33 ppt [ 47 ]. However, acute changes in water salinity can rapidly increase fish metabolic rate due to osmoregulation and respiration [ 37 , 40 , 57 ]. In addition, sea bass presented different growth types within each facility, showing the important effect of seasonal variation on fish growth patterns. Positive allometric values, defined as cases in which fish became stouter with increasing body length, were only observed for offshore cages in spring and autumn, while fish from inland ponds presented isometric or negative allometric growth. Fish growth in offshore cages may be favored by relatively stable water temperatures around the preferred range and constant salinity conditions. Environmental conditions experienced in early life can also play an important role in the morphology and growth trajectories of sea bass [ 58 ], suggesting that seeding season might be an important factor to consider in farm management. Figure 4. Daily temperature ( a ) and salinity ( b ) variation in the two studied locations: inland ponds (red lines) and offshore cages (blue lines). Assuming that temperature and salinity variations are responsible for the observed differences in the length–weight relationship of fish among different facilities and seasons, it is crucial to account for such variations when estimating biomass based on length. In this regard, the practical case presented in Figure 3shows the variations in the estimated biomass for a cultivation unit in the two kinds of studied facilities. The data reveal significant differences, with variations reaching up to 31.5% between winter and autumn in inland ponds, and 28.5% in offshore cages between winter and spring, which are the seasons that exhibit the greatest differences, respectively. These substantial differences should be taken into consideration when estimating fish biomass. Increasingly precise methods are being developed to remotely estimate the length of fish, and the same precision should be the goal for the length–weight conversion, as the total biomass is the estimated value that really matters to fish farmers for the management of their companies. 5. Conclusions Our research highlights the need to use specific length–weight equations for aquaculture facilities that take into account seasonal environmental variations. Many studies aim to establish the most accurate possible length estimates of fish using remote detection methods, such as techniques involving digital imaging or hydroacoustics [ 59 – 62 ]. Avoiding capture-dependent methods that cause high levels of stress to fish can prevent potential increases in mortality [ 5 , 13 – 15 ]. However, it should not be assumed that the length–weight conversion is a minor issue, as the selection of a specific equation can cause significant variations in biomass estimates. This is particularly important in aquaculture, as having 115 Fishes 2023,8, 227 8 of 10 accurate estimates of density and biomass in these farms allows managers to obtain important information for optimizing feed dosage [ 5 , 6 ] (a major production cost for these companies) or managing stocks and sales, all of which are crucial aspects for the sector. Author Contributions: Conceptualization, C.O., L.E. and A.R.-R.; methodology, C.O. and J.R.C.-Q.; software, I.d.M.; validation, L.E. and A.R.-R.; writing—original draft preparation, C.O. and I.d.M.; writing—review and editing, L.E. and A.R.-R.; project administration, C.O. and J.R.C.-Q.; funding acquisition, C.O. and J.R.C.-Q. All authors have read and agreed to the published version of the manuscript. Funding: This study has been funded by the own funds of EcoFishUS Research, a spin-off of the University of Seville dedicated to knowledge transfer, and by the University of Seville through the Research and Transfer Plan, with a doctoral scholarship in collaboration with companies. Institutional Review Board Statement: Our study involved only fish that were caught and immediately sacrificed for the purpose of sale. We did not work with live animals in any way. As part of our research, we took measurements of the fish that had already been sacrificed for commercial purposes. Therefore, we believe that Ethics Committee or Institutional Review Board approval is not required for our manuscript. Data Availability Statement: Restrictions apply to the availability of these data. Data was obtained from two private fish production companies and are available on request from the corresponding author with the permission of the companies. Conflicts of Interest: The authors declare no conflict of interest. References 1. Brugére, C.; Ridler, N. Global Aquaculture Outlook in the next Decades: An Analysis of National Aquaculture Production Forecasts to 2030. FAO Fish. Circ. 2004,1001, 47. 2. Zhou, C.; Xu, D.; Lin, K.; Sun, C.; Yang, X. Intelligent Feeding Control Methods in Aquaculture with an Emphasis on Fish: A Review. Rev. Aquac. 2018,10, 975–993. [CrossRef] 3. An, D.; Hao, J.; Wei, Y.; Wang, Y.; Yu, X. Application of Computer Vision in Fish Intelligent Feeding System—A Review. Aquac. Res. 2021,52, 423–437. [CrossRef] 4. Goldburg, R.; Naylor, R. Future Seascapes, Fishing, and Fish Farming. Front. Ecol. Environ. 2005,3, 21–28. [CrossRef] 5. Orduna, C.; Encina, L.; Rodríguez-Ruiz, A.; Rodríguez-Sánchez, V. Testing of New Sampling Methods and Estimation of Size Structure of Sea Bass (Dicentrarchus labrax) in Aquaculture Farms Using Horizontal Hydroacoustics. Aquaculture 2021 ,545, 737242. [CrossRef] 6. Soliveres, E.; Puig, V.; Perez-arjona, I. Acoustical Biomass Estimation Results in Mediterranean Aquaculture Sea Cages. In Proceedings of the UA2014—2nd International Conference and Exhibition on Underwater Acoustics, Rhodes, Greece, 22–27 June 2014; pp. 1423–1428. 7. Pérez, J.M.; Sarasa, M.; Moço, G.; Granados, J.E.; Crampe, J.P.; Serrano, E.; Maurino, L.; Meneguz, P.G.; Afonso, A.; Alpizar-Jara, R. The Effect of Data Analysis Strategies in Density Estimation of Mountain Ungulates Using Distance Sampling. Ital. J. Zool. 2015 , 82, 262–270. [CrossRef] 8. Hofmeester, T.R.; Rowcliffe, J.M.; Jansen, P.A. A Simple Method for Estimating the Effective Detection Distance of Camera Traps. Remote Sens. Ecol. Conserv. 2016,3, 81–89. [CrossRef] 9. Føre, M.; Alver, M.; Alfredsen, J.A.; Marafioti, G.; Senneset, G.; Birkevold, J.; Willumsen, F.V.; Lange, G.; Espmark, Å.; Terjesen, B.F. Modelling Growth Performance and Feeding Behaviour of Atlantic Salmon (Salmo salar L.) in Commercial-Size Aquaculture Net Pens: Model Details and Validation through Full-Scale Experiments. Aquaculture 2016,464, 268–278. [CrossRef] 10. Orduna, C.; Encina, L.; Rodríguez-Ruiz, A.; Rodríguez-Sánchez, V. Hydroacoustics for Density and Biomass Estimations in Aquaculture Ponds. Aquaculture 2021,545, 737240. [CrossRef] 11. Rodríguez-Sánchez, V.; Rodríguez-Ruiz, A.; Pérez-Arjona, I.; Encina-Encina, L. Horizontal Target Strength-Size Conversion Equations for Sea Bass and Gilt-Head Bream. Aquaculture 2018,490, 178–184. [CrossRef] 12. Harvey, E.; Cappo, M.; Shortis, M.; Robson, S.; Buchanan, J.; Speare, P. The Accuracy and Precision of Underwater Measurements of Length and Maximum Body Depth of Southern Bluefin Tuna (Thunnus maccoyii) with a Stereo-Video Camera System. Fish. Res. 2003,63, 315–326. [CrossRef] 13. Espinosa, V.; Ramis, J.; Alba, J. Evaluación de La Sonda Ultrasónica EY-500 de Simrad Para El Control de Explotaciones de Dorada Sparus Auratus Linnaeus, 1758. Bol. Inst. Esp. Oceanogr. 2002,18, 15–19. 14. Conti, S.G.; Roux, P.; Fauvel, C.; Maurer, B.D.; Demer, D.A. Acoustical Monitoring of Fish Density, Behavior, and Growth Rate in a Tank. Aquaculture 2006,251, 314–323. [CrossRef] 15. The Humane Society of the United States. The Welfare of Animals in the Aquaculture Industry. Hsus Rep. 2008,5, 24. 116 Fishes 2023,8, 227 9 of 10 16. Godlewska, M.; Frouzova, J.; Kubecka, J.; Wi´sniewolski, W.; Szlakowski, J. Comparison of Hydroacoustic Estimates with Fish Census in Shallow Malta Reservoir—Which TS/L Regression to Use in Horizontal Beam Applications? Fish. Res. 2012 ,123, 90–97. [CrossRef] 17. Dennerline, D.E.; Jennings, C.A.; Degan, D.J. Relationships between Hydroacoustic Derived Density and Gill Net Catch: Implications for Fish Assessments. Fish. Res. 2012,123, 78–89. [CrossRef] 18. Greenstreet, S.P.R.; Holland, G.J.; Guirey, E.J.; Armstrong, E.; Fraser, H.M.; Gibb, I.M. Combining Hydroacoustic Seabed Survey and Grab Sampling Techniques to Assess “Local” Sandeel Population Abundance. ICES J. Mar. Sci. 2010 ,67, 971–984. [CrossRef] 19. Koliada, I.; Balk, H.; Tušer, M.; Ptáˇcek, L.; Kubeˇcka, J. Limitations of Target Detection in Horizontal Acoustic Surveys of Extremely Shallow Water Bodies. Fish. Res. 2019,218, 94–104. [CrossRef] 20. Simmonds, J.; MacLennan, D. Fisheries Acoustics: Theory and Practice; Blackwell Science Ltd.: Hoboken, NJ, USA, 2005; ISBN 978-0-632-05994-2. 21. Rodríguez-Sánchez, V.; Encina-Encina, L.; Rodríguez-Ruiz, A.; Sánchez-Carmona, R. Do Close Range Measurements Affect the Target Strength (TS) of Fish in Horizontal Beaming Hydroacoustics? Fish. Res. 2016,173, 4–10. [CrossRef] 22. Lilja, J.; Marjomäki, T.J.; Riikonen, R.; Jurvelius, J. Side-Aspect Target Strength of Atlantic Salmon (Salmo salar), Brown Trout (Salmo trutta), Whitefish (Coregonus lavaretus), and Pike (Esox lucius). Aquat. Living Resour. 2000,13, 355–360. [CrossRef] 23. Frouzova, J.; Kubecka, J.; Balk, H.; Frouz, J. Target Strength of Some European Fish Species and Its Dependence on Fish Body Parameters. Fish. Res. 2005,75, 86–96. [CrossRef] 24. Rodríguez-Sánchez, V.; Encina-Encina, L.; Rodríguez-Ruiz, A.; Monteoliva, A.; Sánchez-Carmona, R. Horizontal Target Strength of Cyprinus Carpio Using 200 kHz and 430 kHz Split-Beam Systems. Fish. Res. 2016,174, 136–142. [CrossRef] 25. Harvey, E.; Fletcher, D.; Shortis, M. Estimation of Reef Fish Length by Divers and by Stereo-Video. Fish. Res. 2002 ,57, 255–265. [CrossRef] 26. Costa, C.; Loy, A.; Cataudella, S.; Davis, D.; Scardi, M. Extracting Fish Size Using Dual Underwater Cameras. Aquac. Eng. 2006 , 35, 218–227. [CrossRef] 27. De La Gandara, F.; Espinosa, V. Estimación Del Número y La Biomasa de Individuos Por Métodos Acústicos En El Cultivo Del Atún Rojo (Thunnus thynnus). Collect. Vol. Sci. Pap. ICCAT 2012,68, 276–283. 28. Islamadina, R.; Pramita, N.; Arnia, F.; Munadi, K. Estimating Fish Weight Based on Visual Captured. In Proceedings of the 2018 International Conference on Information and Communications Technology (ICOIACT), Yogyakarta, Indonesia, 6–7 March 2018; pp. 366–372. [CrossRef] 29. Muñoz-Benavent, P.; Andreu-García, G.; Valiente-González, J.M.; Atienza-Vanacloig, V.; Puig-Pons, V.; Espinosa, V. Enhanced Fish Bending Model for Automatic Tuna Sizing Using Computer Vision. Comput. Electron. Agric. 2018,150, 52–61. [CrossRef] 30. Martinez-De Dios, J.R.; Serna, C.; Ollero, A. Computer Vision and Robotics Techniques in Fish Farms. Robotica 2003 ,21, 233–243. [CrossRef] 31. Person-Le Ruyet, J.; Mahé, K.; Le Bayon, N.; Le Delliou, H. Effects of Temperature on Growth and Metabolism in a Mediterranean Population of European Sea Bass, Dicentrarchus labrax.Aquaculture 2004,237, 269–280. [CrossRef] 32. Llorente, I.; Fernández-Polanco, J.; Baraibar-Diez, E.; Odriozola, M.D.; Bjørndal, T.; Asche, F.; Guillen, J.; Avdelas, L.; Nielsen, R.; Cozzolino, M.; et al. Assessment of the Economic Performance of the Seabream and Seabass Aquaculture Industry in the European Union. Mar. Policy 2020,117, 103876. [CrossRef] 33. Elliot, R. Faking Nature. Inquiry 1982,25, 81–93. [CrossRef] 34. Hidalgo, F.; Alliot, E.; Thebault, H. Influence of Water Temperature on Food Intake, Food Efficiency and Gross Composition of Juvenile Sea Bass, Dicentrarchus labrax.Aquaculture 1987,64, 199–207. [CrossRef] 35. Deane, E.E.; Woo, N.Y.S. Modulation of Fish Growth Hormone Levels by Salinity, Temperature, Pollutants and Aquaculture Related Stress: A Review. Rev. Fish Biol. Fish. 2009,19, 97–120. [CrossRef] 36. Besson, M.; Vandeputte, M.; van Arendonk, J.A.M.; Aubin, J.; de Boer, I.J.M.; Quillet, E.; Komen, H. Influence of Water Temperature on the Economic Value of Growth Rate in Fish Farming: The Case of Sea Bass (Dicentrarchus labrax) Cage Farming in the Mediterranean. Aquaculture 2016,462, 47–55. [CrossRef] 37. Claireaux, G.; Lagardère, J.P. Influence of Temperature, Oxygen and Salinity on the Metabolism of the European Sea Bass. J. Sea Res. 1999,42, 157–168. [CrossRef] 38. Reynolds, W.W.; Casterlin, M.E. Behavioral Thermoregulation and the “Final Preferendum” Paradigm. Integr. Comp. Biol. 1979 , 19, 211–224. [CrossRef] 39. Zanuy, S.; Carrillo, M. Annual Cycles of Growth, Feeding Rate, Gross Conversion Efficiency and Hematocrit Levels of Sea Bass (Dicentrarchus labrax L.) Adapted to Two Different Osmotic Media. Aquaculture 1985,44, 11–25. [CrossRef] 40. Dalla Via, J.; Villani, P.; Gasteiger, E.; Niederstätter, H. Oxygen Consumption in Sea Bass Fingerling Dicentrarchus labrax Exposed to Acute Salinity and Temperature Changes: Metabolic Basis for Maximum Stocking Density Estimations. Aquaculture 1998 ,169, 303–313. [CrossRef] 41. Von Oertzen, J.A. Influence of Steady-State and Fluctuating Salinities on the Oxygen Consumption and Activity of Some Brackish Water Shrimps and Fishes. J. Exp. Mar. Biol. Ecol. 1984,80, 29–46. [CrossRef] 42. Rao, G.M.M. Oxygen Consumption of Rainbow Trout (Salmo gairdneri) in Relation to Activity and Salinity. Can. J. Zool. 1968 ,46, 781–786. [CrossRef] 117 Fishes 2023,8, 227 10 of 10 43. Farmer, G.J.; Beamish, F.W.H. Oxygen Consumption of Tilapia Nilotica in Relation to Swimming Speed and Salinity. J. Fish. Res. Board Canada 1969,26, 2807–2821. [CrossRef] 44. Boyd, C.E.; Torrans, E.L.; Tucker, C.S. Dissolved Oxygen and Aeration in Ictalurid Catfish Aquaculture. J. World Aquac. Soc. 2018 , 49, 7–70. [CrossRef] 45. Ma, Y.; Ding, W. Design of Intelligent Monitoring System for Aquaculture Water Dissolved Oxygen. In Proceedings of the 2018 IEEE 3rd Advanced Information Technology, Electronic and Automation Control Conference (IAEAC), Chongqing, China, 12–14 October 2018; pp. 414–418. [CrossRef] 46. Li, C.; Li, Z.; Wu, J.; Zhu, L.; Yue, J. A Hybrid Model for Dissolved Oxygen Prediction in Aquaculture Based on Multi-Scale Features. Inf. Process. Agric. 2018,5, 11–20. [CrossRef] 47. Dendrinos, P.; Thorpe, J.P. Effects of Reduced Salinity on Growth and Body Composition in the European Bass Dicentrarchus labrax (L.). Aquaculture 1985,49, 333–358. [CrossRef] 48. Eroldo ˇ gan, O.T.; Kumlu, M.; Akta¸s, M. Optimum Feeding Rates for European Sea Bass Dicentrarchus labrax L. Reared in Seawater and Freshwater. Aquaculture 2004,231, 501–515. [CrossRef] 49. Gomes, M.; Correia, A.; Pinto, L.; Sá, C.; Brotas, V.; Mateus, M. Coastal Water Quality in an Atlantic Sea Bass Farm Site (Sines, Portugal): A First Assessment. Front. Mar. Sci. 2020,7, 175. [CrossRef] 50. Mosqueira, M.; Pombo, A.; Borges, C.; Brito, A.C.; Zacarias, N.; Esteves, R.; Palma, C. Potential for Coastal and Offshore Aquaculture in Portugal: Insights from Physico-Chemical and Oceanographic Conditions. Appl. Sci. 2022,12, 2742. [CrossRef] 51. R Core Team. R: A Language and Environment for Statistical Computing; Austria R Foundation for Statistical Computing: Vienna, Austria, 2022. 52. Ogle, D.H.; Doll, J.C.; Wheeler, A.P.; Dinno, A. FSA: Simple Fisheries Stock Assessment Methods. 2023. Available online: https://cran.r-project.org/package=FSA (accessed on 30 March 2023). 53. Jobling, M. The Thermal Growth Coefficient (TGC) Model of Fish Growth: A Cautionary Note. Aquac. Res. 2003 ,34, 581–584. [CrossRef] 54. Jauralde, I.; Martínez-Llorens, S.; Tomás, A.; Ballestrazzi, R.; Jover, M. A Proposal for Modelling the Thermal-Unit Growth Coefficient and Feed Conversion Ratio as Functions of Feeding Rate for Gilthead Sea Bream (Sparus aurata, L.) in Summer Conditions. Aquac. Res. 2013,44, 242–253. [CrossRef] 55. Bramato, S.; Losada, M.A.; Contreras, E.; Polo, M. An Integrated Database Manager to Forecast Estuarine Dynamics and Water Quality in the Guadalquivir River (Spain). River Flow 2010,2, 1415–1420. 56. Contreras, E.; Polo, M.J. Measurement Frequency and Sampling Spatial Domains Required to Characterize Turbidity and Salinity Events in the Guadalquivir Estuary (Spain). Nat. Hazards Earth Syst. Sci. 2012,12, 2581–2589. [CrossRef] 57. Precht, H.; Christophersen, J.; Hensel, H. Temperatur Und Leben; Springer: Berlin/Heidelberg, Germany, 1955; p. 514. 58. Georgakopoulou, E.; Sfakianakis, D.G.; Kouttouki, S.; Divanach, P.; Kentouri, M.; Koumoundouros, G. The Influence of Temperature during Early Life on Phenotypic Expression at Later Ontogenetic Stages in Sea Bass. J. Fish Biol. 2007 ,70, 278–291. [CrossRef] 59. Boswell, K.M.; Wilson, M.P.; Wilson, C.A. Hydroacoustics as a Tool for Assessing Fish Biomass and Distribution Associated with Discrete Habitats in a Louisiana Estuary. Estuaries Coasts 2007,30, 607–617. [CrossRef] 60. Álvarez-Ellacuría, A.; Palmer, M.; Catalán, I.A.; Lisani, J.L. Image-Based, Unsupervised Estimation of Fish Size from Commercial Landings Using Deep Learning. ICES J. Mar. Sci. 2020,77, 1330–1339. [CrossRef] 61. Hsieh, C.L.; Chang, H.Y.; Chen, F.H.; Liou, J.H.; Chang, S.K.; Lin, T. Te A Simple and Effective Digital Imaging Approach for Tuna Fish Length Measurement Compatible with Fishing Operations. Comput. Electron. Agric. 2011,75, 44–51. [CrossRef] 62. Shi, C.; Wang, Q.; He, X.; Zhang, X.; Li, D. An Automatic Method of Fish Length Estimation Using Underwater Stereo System Based on LabVIEW. Comput. Electron. Agric. 2020,173, 105419. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. 118 General discussion. This industrial doctoral thesis, consisting of 3 scientific articles and 2 inventions, aims to estimate density and biomass in fish aquaculture systems using horizontal hydroacoustics. The industrial doctorate distinction is based on the business-oriented approach of the work, which seeks to provide solutions to companies in the aquaculture sector, allowing them to reduce production costs by creating an estimation service through EcoFishUS Research S.L.L., a spin-off company from the University of Seville, established for the transfer of knowledge generated at the University, and a co-financer of this doctoral thesis alongside the University of Seville and CTAQUA (Centro Tecnológico de Acuicultura de Andalucía). This work addresses the significant challenges that the application of hydroacoustics faces in fish farming systems, where high densities are confined in small spaces. Various alternatives are tested to overcome these challenges, leading to the achievement of accurate results in different types of sea bass farming facilities. In the first article, the use of horizontal hydroacoustics for estimating the density and biomass of sea bass in land-based farming ponds is proposed. In this case, not only were the technical challenges of high fish density and their orientation relative to the acoustic beam addressed, but the effect of a boat navigating through the pond was also considered, as it inevitably causes an escape response in the fish, which must be accounted for in the estimates. To overcome these difficulties, split-beam transducers were used, which allows determining the angle of incidence of the acoustic beam with the fish; along with specific equations for horizontal hydroacoustics and the species under study (Hazen & Horne, 2003; Simmonds & MacLennan, 2007; Rodríguez-Sánchez, 2015; Rodríguez-Sánchez et al., 2015, 2016b; a, 2018). Regarding the escape behaviour of the fish, despite using a very quiet electric motor, it was found that this escape effect must be considered, especially in shallow systems like the one studied (Buckland et al., 2001; Hjellvik et al., 2008; De Robertis & Wilson, 2011; Marques et al., 2013) therefore, the function for detecting individual fish was incorporated into the sampling (Williams et al., 2002; Draštík & Kubecka, 2005; Cox et al., 2011; Marques et al., 2013; Glennie et al., 2015; Wheeland & Rose, 2015). 119 126 Discusión general. Esta tesis doctoral industrial, compuesta por 3 artículos científicos y 2 invenciones, tiene como objetivo realizar estimas de densidad y biomasa en sistemas de acuicultura de peces utilizando hidroacústica horizontal. La mención de doctorado industrial se basa en el enfoque empresarial del trabajo, que busca ofrecer soluciones a empresas del sector acuícola que les permitan reducir costes de producción, mediante la creación de un servicio de estimas por parte de EcoFishUS Research S.L.L., empresa spinoff de la Universidad de Sevilla, creada para la transferencia de conocimientos generados en la Universidad, y financiadora junto a la Universidad de Sevilla y CTAQUA (Centro Tecnológico de Acuicultura de Andalucía) de esta tesis doctoral. En este trabajo se plantean las grandes dificultades que la aplicación de la hidroacústica tiene en sistemas de cultivo de peces, con altas densidades confinadas en espacios reducidos, y se prueban diferentes alternativas para solventarlas, que llevan a la obtención de resultados precisos en distintos tipos de instalaciones de cría de lubinas. En el primer artículo se plantea el uso de la hidroacústica horizontal para la estima de densidad y biomasa de lubinas en balsas de cultivo en tierra. En este caso no solo se afrontaron los desafíos a la técnica que representan la alta densidad de peces y la orientación de estos respecto al haz acústico, sino que se incluye el efecto de una embarcación navegando por la balsa, provocando un inevitable comportamiento de escape de los peces, a tener en cuenta para las estimas. Para superar estas dificultades, en cuanto a la orientación de los peces respecto al haz acústico se usaron transductores de haz partido, que permiten determinar el ángulo de incidencia del haz acústico con el pez; y ecuaciones específicas para hidroacústica horizontal y la especie de estudio (Hazen & Horne, 2003; Simmonds & MacLennan, 2007; Rodríguez-Sánchez, 2015; RodríguezSánchez et al., 2015, 2016b; a, 2018b). Respecto al comportamiento de huida de los peces, pese a utilizar un motor eléctrico muy silencioso, se ha comprobado que este efecto de escape debe ser tenido en cuenta, especialmente en sistemas someros como el estudiado (Buckland et al., 2001; Hjellvik et al., 2008; De Robertis & Wilson, 2011; Marques et al., 2013) por lo que se incorporó la función de detección de peces individuales en el muestreo (Williams et al., 2002; Draštík & Kubecka, 2005; Cox et al., 2011; Marques et al., 2013; Glennie et al., 2015; Wheeland & Rose, 2015). 127 Como contraposición a estas dificultades, el experimento contaba con una gran ventaja, y es el despesque posterior de las balsas, permitiendo obtener datos reales de producción, que pudieron ser comparados con los de las estimas realizadas, presentando un buen ajuste entre ellas, con una diferencia de solo el 3,73% en cuanto a densidad y el 11,77% en cuanto a biomasa en los muestreos con trayectorias centrales. Los resultados permitieron establecer un protocolo de muestreo y análisis que proporciona resultados de densidad y biomasa precisos en balsas de cultivo de lubina en tierra, con un ajuste muy valorado por los acuicultores debido a la enorme dificultad que les supone realizar este tipo de estimas. Se presenta por tanto una innovadora herramienta para el manejo de la producción en este tipo de instalaciones. Para llevar a cabo los muestreos de este experimento, se desarrolló una estructura de sujeción del transductor en la que el Grupo de Investigación llevaba trabajando ya varios años, y se procedió a su protección industrial. La sujeción del equipo emisor hidroacústico en embarcaciones pequeñas, mucho más versátiles que las medianas o grandes y única opción en ecosistemas someros, presenta una dificultad insalvable en cuanto a su fijación a la barca, ya que no existe un casco o estructuras rígidas a las que se pueda atornillar el equipo. La correcta sujeción afecta a un aspecto fundamental de este tipo de muestreos, como es mantener constante la distancia lateral entre la embarcación y el transductor, así como la profundidad que este mantiene dentro del agua durante la navegación. Se trata de una estructura tubular de acero inoxidable, que permite la sujeción del transductor de la ecosonda en una embarcación neumática sin dañarla, debido a su horma curva, que se ajusta al flotador lateral de la misma. Esta invención, junto con el primer artículo, permiten la realización de estimas de densidad y biomasa de lubinas en balsas de cultivo de tierra de una manera eficiente, y con resultados válidos comprobados. El segundo artículo prueba diferentes métodos de muestreo para la estima de estructuras de talla en jaulas de cultivo de lubinas en mar abierto con hidroacústica horizontal, como un primer paso para la estima de densidad y biomasa de peces en este tipo de cultivos. Se realizaron 3 tipos de muestreos en jaulas de cultivo reales de lubinas, y cada uno de ellos se llevó a cabo a su vez de dos formas: en movimiento y de manera fija. En este caso, no se presentaron grandes dificultades por la perturbación y escape de los peces, debido a que el sistema de cultivo cuenta con un mayor volumen de agua que las balsas en tierra, y no era necesario navegar 128 en ellas. Las dificultades encontradas residieron en la alta densidad de peces con la que impacta el haz acústico, dificultando en gran manera la obtención de datos precisos debido a la atenuación de la señal devuelta con la distancia (Røttingen, 1976; Furusawa et al., 1984; Appenzeller & Leggett, 1992). Por ello, se redujo la distancia de análisis de datos a la mitad de la jaula, aprovechando el comportamiento de las lubinas, que nadan en cardumen en paralelo a la red de la jaula, homogenizando la distribución de tallas en el eje horizontal de la misma y posibilitando esta división. Los resultados vuelven a presentar a la hidroacústica horizontal como una técnica útil y aplicable a este tipo de instalaciones, estableciendo como el mejor método de muestreo el realizado en movimiento y desde fuera de la jaula, en función de los resultados obtenidos y su comprobación con los datos aportados por los acuicultores, en base a estimas propias y a los despesques realizados para la venta de la producción. En base a los resultados de este artículo, se diseña y construye la segunda invención: el equipo de muestreo hidroacústico en profundidad marina. Se trata de un equipo formado por una plataforma con capacidad de flotación sobre la superficie del agua que, mediante una serie de poleas, permite el desplazamiento en un plano vertical del emisor acústico, sin balanceo horizontal ni vertical. Además, este dispositivo cuenta con un motor eléctrico que permite situar al transductor a una profundidad determinada, o realizar muestreos en movimiento en el eje vertical, a velocidad constante. En el tercer artículo se aborda un factor que se detecta como posible nueva variable en cuanto a las estimas de biomasa durante el estudio, y es la variabilidad estacional y ambiental en los sistemas de acuicultura en cuanto a la relación longitud-peso de los individuos cultivados (Person-Le Ruyet et al., 2004). Las ecuaciones específicas TS-longitud desarrolladas para la lubina nos permiten relacionar de manera precisa la señal acústica recibida con la longitud de los peces (Rodríguez-Sánchez et al., 2018), pero el paso de longitud a peso se hace con ecuaciones que, en muchas ocasiones, no pertenecen a los peces insonificados, ya que precisamente una de las grandes ventajas de la hidroacústica reside en ser un método no invasivo (Espinosa et al., 2002; Conti et al., 2006; The Humane Society of the United States, 2008). El artículo muestra como esta relación cambia en función tanto del lugar de cultivo como de la estación del año en la que se realizan los muestreos, presentándose como un factor relevante y a 129 tener en cuenta para este tipo de estimas, siendo conveniente el uso de ecuaciones específicas en base a estudios previos. En conjunto, los resultados obtenidos a lo largo de esta investigación representan un avance significativo en la gestión de la acuicultura, al demostrar la viabilidad técnica y científica de la aplicación de tecnologías hidroacústicas no invasivas para la monitorización del tamaño-peso de lubinas en diferentes tipos de instalaciones acuícolas. La combinación de nuevas metodologías y tecnologías avanzadas para el muestreo hidroacústico es capaz de aportar datos relevantes sobre los cultivos, permitiendo la reducción de costes, de impactos negativos asociados con el estrés por la manipulación física de los peces y optimizando los procesos de producción (Pickering & Pottinger, 1989; Espinosa et al., 2002; Caruso et al., 2005; The Humane Society of the United States, 2008; Iwama et al., 2011). La incorporación de estas tecnologías no solo responde a desafíos concretos de la industria, sino que también contribuye de manera significativa a la sostenibilidad económica y ambiental del sector acuícola. Al optimizar la toma de decisiones, se mejora la eficiencia en la gestión de recursos, lo que reduce tanto la contaminación por sobrealimentación como el riesgo de pérdidas asociadas a una inadecuada gestión de los stocks. Asimismo, estas innovaciones abren nuevas oportunidades para la investigación y el perfeccionamiento de los sistemas de cultivo acuícolas, impulsando su desarrollo y competitividad a largo plazo (Hatziathanasiou et al., 2002; McCallum, 2005; Espinosa et al., 2006; Soliveres et al., 2014; Føre et al., 2016). 130 Conclusiones. La aplicación de la hidroacústica horizontal en balsas de cultivo de lubinas permite realizar estimas fiables de densidad y biomasa de los peces cultivados. El método de muestreo más adecuado en balsas de cultivo de lubinas en tierra se realiza navegando con trayectorias centrales a lo largo de las balsas. La estructura para ecosondeos diseñada para el muestreo con hidroacústica desde embarcaciones neumáticas permite la toma de datos acústicos precisos y fiables. La aplicación de la hidroacústica horizontal en jaulas de cultivo de lubinas en mar abierto permite realizar estimas fiables de estructuras de talla de los peces cultivados. El mejor método de muestreo hidroacústico horizontal en jaulas de acuicultura de lubinas en mar abierto es con el emisor de sonido situado el exterior de la jaula y en movimiento. El equipo de muestreo hidroacústico en profundidad marina permite realizar muestreos con hidroacústica horizontal en jaulas de cultivo de lubina de manera precisa, permitiendo el desplazamiento vertical del emisor de sonido sin balanceo horizontal. Es importante el uso de ecuaciones de conversión longitud-peso específicas que tengan en cuenta la variación ambiental y estacional de los cultivos para obtener estimas precisas de biomasa de lubinas a partir de su talla. 131 Financiación. Este trabajo ha sido financiado por EcoFishUS Research S.L.L., empresa spin-off dedicada al estudio de ecosistemas acuáticos, fundada por el Grupo de Investigación de Ecología de Peces RNM-320 de la Universidad de Sevilla; por el Centro Tecnológico de Acuicultura de Andalucía (CTAQUA); y por el Plan Propio de Investigación y Transferencia de la Universidad de Sevilla (VI PPIT 2017 II.2B2; IV.9). 132 Referencias. APPENZELLER, A.R. & W.C. LEGGETT 1992. Bias in hydroacoustic estimates of fish abundance due to acoustic shadowing: evidence from day–night surveys of vertically migrating fish. Canadian Journal of Fisheries and Aquatic Sciences 49 (10): 2179–2189. APROMAR 2023. La acuicultura en España 2023. R. V. Garazi (ed.). Asociación Empresarial de Acuicultura de España. BALK, H., B.S. SØVEGJARTO, M. TUŠER, J. FROUZOVÁ, M. MUŠKA, V. DRAŠTÍK, R. BARAN & J. KUBECKA 2017. Surface-induced errors in target strength and position estimates during horizontal acoustic surveys. Fisheries Research 188: 149–156. BALON, E.K. 1995. Origin and domestication of the wild carp, Cyprinus carpio: from Roman gourmets to the swimming flowers. Aquaculture 129: 3–48. BARAN, R., T. JU ZA, M. TUŠER, H. BALK, P. BLABOLIL, M. CECH, V. DRAŠTÍK, J. FROUZOVÁ, A.D. JAYASINGHE, I. KOLIADA, T. MRKVICKA, M. MUŠKA, D. RICARD, Z. SAJDLOVÁ, L. VEJR ÍK & J. KUBECKA 2017. A novel upward-looking hydroacoustic method for improving pelagic fish surveys. Scientific Reports 7 (1): 1–12. BERNAL CASASOLA, D. 2014. Viveros y yacimientos haliéuticos. Importancia y fragilidad del patrimonio pesquero-conservero hispanorromano. M. Olcina Domènech and R. Pérez Jiménez (eds.). MARQ. Museo Arqueológico de Alicante. BEVERIDGE, M.C.M. & D.C. LITTLE 2002. The History of Aquaculture in Traditional Societies. In: Ecological Aquaculture. B. A. Costa-Pierce (ed.): 1–29. BREWER, D.J. & R.F. FRIEDMAN 1989. Fish and fishing in ancient Egypt. Aris & Phillips. BUCKLAND, S.T., D.R. ANDERSON, K.P. BURNHAM, J.L. LAAKE, D.L. BORCHERS & L. THOMAS 2001. Introduction to Distance Sampling: Estimating Abundance of Biological Populations. Oxford University Press. 133 CARUSO, G., L. GENOVESE, G. MARICCHIOLO & A. MODICA 2005. Haematological, biochemical and immunological parameters as stress indicators in Dicentrarchus labrax and Sparus aurata farmed in offshore cages. Aquaculture International 13 (1): 67–73. CHIMITS, P. 1957. Tilapia in ancient Egypt. FAO Fisheries Bulletin 10 (4): 1–5. CONTI, S.G., P. ROUX, C. FAUVEL, B.D. MAURER & D.A. DEMER 2006. Acoustical monitoring of fish density, behavior, and growth rate in a tank. Aquaculture 251 (2–4): 314–323. COSTA-PIERCE, B.A. 2022. The Anthropology of Aquaculture. Frontiers in Sustainable Food Systems 6 (843743): 1–20. COX, M.J., D.L. BORCHERS, D.A. DEMER, G.R. CUTTER & A.S. BRIERLEY 2011. Estimating the density of Antarctic krill (Euphausia superba) from multi-beam echo-sounder observations using distance sampling methods. Journal of the Royal Statistical Society Series C: Applied Statistic 60 (2): 301–316. DAWSON, J.J., D. WIGGINS, D. DEGAN, H. GEIGER, D. HART & B. ADAMS 2000. Point-source violations: split-beam tracking of fish at close range. Aquatic Living Resources 13 (5): 291–295. DESROCHES-NOBLECOURT, C. 1954. Poissons, tabous et transformations du mort: Nouvelles considerations sur les pelerinages aux villes saintes. Librairie Orientaliste Paul Geuthner. Paris. DRAŠTÍK, V. & J. KUBECKA 2005. Fish avoidance of acoustic survey boat in shallow waters. Fisheries Research 72 (2–3): 219–228. DRAŠTÍK, V., J. KUBECKA, M. CECH, J. FROUZOVÁ, M. R ÍHA, T. JUZA, M. TUŠER, O. JAROLÍM, M. PRCHALOVÁ, J. PETERKA, M. VAŠEK, M. KRATOCHVÍL, J. MATE NA & T. MRKVICKA 2009. Hydroacoustic estimates of fish stocks in temperate reservoirs: Day or night surveys? Aquatic Living Resources 22 (1): 69–77. ENCINA, L. & A. RODRÍGUEZ-RUIZ 2002. Patterns of seasonal distribution of the fish assemblage in a reservoir of recent construction. SIL Proceedings, 1922-2010 28 (2): 641–650. ESPINOSA, V., J. RAMIS & J. ALBA 2002. Evaluación de la sonda ultrasónica EY-500 de Simrad para el control de explotaciones de 134 dorada Sparus auratus Linnaeus, 1758. Boletín del Instituto Español de Oceanografía 18: 15–19. ESPINOSA, V., E. SOLIVERES, V.D. ESTRUCH, J. REDONDO, M. ARDID, J. ALBA, E. ESCUDER, M. BOU & D. VICENTE 2006. Acoustical monitoring of open mediterranean sea fish farms: problems and strategies. In: EAA European symposium on hidroacoustics. Gandia. FAO. 1–75. FAO 2024. El estado mundial de la pesca y la acuicultura 2024 El estado mundial de la pesca y la acuicultura 2024. FAO. Rome. FOOTE, K.G. 1980. Importance of the swimbladder in acoustic scattering by fish: A comparison of gadoid and mackerel target strengths. The Journal of the Acoustical Society of America 67 (6): 2084–2089. FØRE, M., M. ALVER, J.A. ALFREDSEN, G. MARAFIOTI, G. SENNESET, J. BIRKEVOLD, F.V. WILLUMSEN, G. LANGE, Å. ESPMARK & B.F. TERJESEN 2016. Modelling growth performance and feeding behaviour of Atlantic salmon (Salmo salar L.) in commercial-size aquaculture net pens: Model details and validation through full-scale experiments. Aquaculture 464: 268–278. FROUZOVA, J., J. KUBECKA, H. BALK & J. FROUZ 2005. Target strength of some European fish species and its dependence on fish body parameters. Fisheries Research 75 (1–3): 86–96. FURUSAWA, M., K. ISHII, Y. MIYANOHANA & Y. MANIWA 1984. Experimental Investigation of an Acoustic Method to Estimate Fish Abundance Using Culture Nets. Japanese Journal of Applied Physics 23: 101–103. GLENNIE, R., S.T. BUCKLAND & L. THOMAS 2015. The effect of animal movement on line transect estimates of abundance. PLoS ONE 10 (3): 1–15. GODLEWSKA, M., J. FROUZOVA, J. KUBECKA, W. WISNIEWOLSKI & J. SZLAKOWSKI 2012. Comparison of hydroacoustic estimates with fish census in shallow Malta Reservoir - which TS/L regression to use in horizontal beam applications? Fisheries Research 123–124: 90–97. GRANADO-LORENCIO, C. 1996. Ecología de peces. Secretariado de Publicaciones, Universidad de Sevilla. 135 A mis amigos, a los que siempre han estado ahí y sé que siempre estarán para disfrutar los momentos felices, en honor a los tiempos difíciles. Por último, me resulta necesario confesar el enorme dolor que me produce que ya no estén aquí dos de las personas más importantes que pasarán nunca por mi vida para verme cerrar este capítulo. Ni Vicky ni mi padre podrán nunca leer estos agradecimientos, pero tengo claro que sin ellos yo no sería la persona que soy, yo no estaría aquí. Gracias por haberme hecho mejor. 142 143