Explore the vast range of titles available.

A wide range of applications in marine ecology extensively uses underwater cameras. Still, to efficiently process the vast amount of data generated, we need to develop tools that can automatically detect and recognize species captured on film. Classifying fish species from videos and images in natural environments can be challenging because of noise and variation in illumination and the surrounding habitat. In this paper, we propose a two-step deep learning approach for the detection and classification of temperate fishes without pre-filtering. The first step is to detect each single fish in an image, independent of species and sex. For this purpose, we employ the You Only Look Once (YOLO) object detection technique. In the second step, we adopt a Convolutional Neural Network (CNN) with the Squeeze-and-Excitation (SE) architecture for classifying each fish in the image without pre-filtering. We apply transfer learning to overcome the limited training samples of temperate fishes and to improve the accuracy of the classification. This is done by training the object detection model with ImageNet and the fish classifier via a public dataset (Fish4Knowledge), whereupon both the object detection and classifier are updated with temperate fishes of interest. The weights obtained from pre-training are applied to post-training as a priori. Our solution achieves the state-of-the-art accuracy of 99.27% using the pre-training model. The accuracies using the post-training model are also high; 83.68% and 87.74% with and without image augmentation, respectively. This strongly indicates that the solution is viable with a more extensive dataset.

Aquaculture requires precise non-invasive methods for biomass estimation. This research validates a novel computer vision methodology that uses a signature function-based feature extraction algorithm combining statistical morphological analysis of the size and shape of fish and machine learning to improve the accuracy of biomass estimation in fishponds and is specifically applied to tilapia (Oreochromis niloticus). These features that are automatically extracted from images are put to the test against previously manually extracted features by comparing the results when applied to three common machine learning methods under two different lighting conditions. The dataset for this analysis encompasses 129 tilapia samples. The results give promising outcomes since the multilayer perceptron model shows robust performance, consistently demonstrating superior accuracy across different features and lighting conditions. The interpretable nature of the model, rooted in the statistical features of the signature function, could provide insights into the morphological and allometric changes at different developmental stages. A comparative analysis against existing literature underscores the competitiveness of the proposed methodology, pointing to advancements in precision, interpretability, and species versatility. This research contributes significantly to the field, accelerating the quest for non-invasive fish biometrics that can be generalized across various aquaculture species in different stages of development. In combination with detection, tracking, and posture recognition, deep learning methodologies such as the one provided in the latest studies could generate a powerful method for real-time fish morphology development, biomass estimation, and welfare monitoring, which are crucial for the effective management of fish farms.

This research has been carried out within the framework of the ECOFISH project: ecoinnovative strategies for sustainable fishing in the Gulf of Cadiz SPA. This initiative has been supported by the Biodiversity Foundation, the Ministry for Ecological Transition and Demographic Challenge, through the Pleamar Program, co-financed by the European Maritime and Fisheries Fund (EMFF) [grant number: 2019-016/PV/PLEAMAR18/PT; 2020-013/PV/PLEAMAR19/PT; 2020- 055/PV/PLEAMAR20/PT; 2021/PV/PLEAMAR20-21/PT; 2021-060/PV/PLEAMAR21/PT].

Coral reefs are under increasing threat, and the loss of reef‐associated fishes providing valuable ecosystem services is accelerating. The monitoring of such rapid changes has become a challenge for ecologists and ecosystems managers using traditional approaches like scuba divers performing underwater visual censuses (UVC) or diver operated video recording (DOV). However, the use of small, low‐cost robots could help tackle the challenge of such monitoring, provided that they perform at least as well as diver‐based methods. To address this question, tropical fish assemblages from 13 fringing reefs around Mayotte Island (Indian Ocean) were monitored along 50 m‐long transects using stereo videos recorded by a semi‐autonomous underwater vehicle (SAUV) and by a scuba diver (Diver Operated stereo Video system, DOV). Differences between the methods were tested for complementary fish assemblage metrics (species richness, total biomass, total density, Shannon diversity and Pielou evenness) and for the number and size of nine targeted species. SAUV recorded on average 35% higher biomass than DOV which in turn recorded on average 12% higher species richness. Biomass differences were found to be due to SAUV monitoring larger fishes than DOV, a potential marker of human‐related fish avoidance behaviour. This study demonstrates that SAUV provides accurate metrics of coral reef fish biodiversity compared to diver‐based procedures. Given their ability to conduct video transects at high frequency, 100 m depth range and at a moderate cost, SAUV is a promising tool for monitoring fish assemblages in coral reef ecosystems. New tools are needed in light of exosystemic global change monitoring, therefore we tested the possibility to use robots to perform coral reef fishes monitoring surveys. Thirteen fringing reefs around Mayotte Island (Indian Ocean) were surveyed along 50 m‐long transects using stereo videos recorded by a semi‐autonomous underwater vehicle (SAUV) and by the standard approach using a scuba‐diver operated stereo video system (DOV). SAUV recorded on average 35% higher biomass than DOV which recorded on average 12% higher species richness. The biomass differences were found to be due to SAUV monitoring larger fishes than DOV, a potential marker of human‐related fish wariness.

The Indian hilsa shad, Tenualosa ilisha (Hamilton, 1822), is a highly valued fish of the Indo Pacific region. This paper represents the length-weight relationships (LWR), condition factor ( K ), relative condition factor ( K n ), and otolith dimension of hilsa, throughout the Hooghly River (West Bengal) from January to December 2020. The value of exponent “b” in the equation W = cL b was 2.6 for the species. It suggests the growth rate almost follows the cube law. Most of the population under study complied with the ideal b-value. Regression analysis indicated a high positive correlation between length and weight in the species ( R 2 = 0.76). The K n shows their general well-being to be good in the river. The relationships between total fish weight and otolith biometry parameters such as long arm ( L.A ) length, short arm ( S.A ) length, L.A / S.A , and diameter ( D ) were examined. Linear regressions have represented fish weight-otolith biometry relationships. Regression analysis indicated a strong positive relationship between body weight and otolith length ( R 2 = 0.848), otolith diameter ( R 2 = 0.803), and short arm length of otolith ( R 2 = 0.678), but no relationship was found between L.A / S.A of otolith and fish weight. This strong relationship between fish weight and otolith length and diameter of T. ilisha , reflects that otolith length and diameter could become the best predictor of fish weight.

The Strait of Messina (Central Mediterranean Sea) has always been known for the stranding of marine organisms, especially during the spring. We came across an extraordinary event of mass stranding in April 2025, with 453 specimens of Macroramphosus sp. found through a single night. A total of 571 post-larvae and juvenile individuals stranded between February and May 2025 were examined for this study. Clear morphological differences related to the size, especially in post-larvae, were highlighted. The relationship between Body Length (BL) and other morphometric measurements, such as Dorsal Spine Length (DSL), Snout Length (SNL), and Body Height (BH), were studied, with the aim of identifying without any doubt the species Macroramphosus gracilis. A slightly negative allometric relationship between length and weight shows that it grows faster in length and slower in weight. This study aims to improve the state of knowledge on slender snipefish M. gracilis, and particularly on the somatic features of different post-larval development stages, such as the presence of spinules in various parts of the body. All these morphological changes could give us a hint at the ecological adaptation to the habit shift, as relates to development.

A trial was performed to study the effects of a partially defatted Hermetia illucens (L. 1758) larvae meal (HIM) to replace a fish meal in the diet of an adult gilthead sea bream. The fish were fed for 120 days with diets containing HIM: 0% (CTRL), 25 (HI25), 50 (HI50), and 75% (HI75), corresponding to inclusion levels of 0, 9.2, 18.4, and 27.6%. A digestibility trial was carried out to evaluate the apparent digestibility coefficients (ADC). Growth performance, morphometric and biometric indexes, gill cortisol, status of gut mucosa, and the gut enzymes’ (alkaline proteases, lipase, and amylase) activity were evaluated. Morphometric and biometric indexes, digestibility, gill cortisol, and gut enzymes were not affected by the diet. Conversely, the 27.6% HI inclusion level worsened the specific growth rate, feed conversion ratio, and protein efficiency ratio. Histology showed no significant differences between CTRL and HI25 groups. Conversely, in many of HI50 and HI75 fish, anatomo-functional changes in the gut were observed. No statistical differences were found among diets for ADC of dry matter, crude protein, ether extract, and gross energy. The results showed that defatted HIM can be used as an alternative protein source up to 18.4% of the inclusion level without impacting the growth performance, gill cortisol, morphometric and biometric indexes, ADCs, and gut enzymes.

Sea bass is a fish widely produced, consumed and appreciated in Italy. Its intensive rearing system provides the consumption of valuable fish to a wider population. Thanks to the use of an appropriate feed, it is possible to obtain reared sea bass which are richer in total lipid with a majority presence of polyunsaturated fatty acids, such as n-3 and n-6 series. In this study, a total of 75 specimens of European sea bass coming from three different origins (two farmed and one wild) were considered, with 25 fish from each origin. Biometry traits were valued as of the chemical and fatty acid profile of fillets. Biometric indices, proximate composition and fatty acid percentage were significantly affected by the rearing system. Fishes from the intensive rearing system (IRS) showed the highest value of relative profile and condition factor, a higher content of lipid and total n-6 that influenced the n-6/n-3 ratio and the atherogenic indexes, and values that indicated their flesh for human consumption as a healthy alternative to the wild fishes.

Northeast Arctic cod, saithe and haddock are among the most important fisheries resources in Europe, largely shipped to various continental markets. The present study aimed to map the presence and distribution of larvae of parasitic nematodes in the Anisakidae family which are of socioeconomic and public health concern. Fishes were sourced from commercial catches during winter or spring in the southern Barents Sea. Samples of fish were inspected for nematodes using the UV-press method while anisakid species identification relied on sequencing of the mtDNA cox2 gene. Anisakis simplex (s.s.) was the most prevalent and abundant anisakid recorded, occurring at high infection levels in the viscera and flesh of cod and saithe, while being less abundant in haddock. Contracaecum osculatum (s.l.) larvae, not found in the fish flesh, showed moderate-to-high prevalence in saithe, haddock and cod, respectively. Most Pseudoterranova spp. larvae occurred at low-to-moderate prevalence, and low abundance, in the viscera (Pseudoterranova bulbosa) and flesh (Pseudoterranova decipiens (s.s.) and Pseudoterranova krabbei) of cod, only 2 P. decipiens (s.s.) appeared in the flesh of saithe. Body length was the single most important host-related factor to predict overall abundance of anisakid larvae in the fish species. The spatial distribution of Anisakis larvae in the fish flesh showed much higher abundances in the belly flaps than in the dorsal fillet parts. Trimming of the flesh by removing the belly flaps would reduce larval presence in the fillets of these gadid fish species by 86–91%.

Over the past decade, concerns over microplastic pollution in the marine ecosystem has increasingly gained more attention, but research investigating the ingestion of microplastics by marine fish in Malaysia is still regrettably lacking. This study investigated the microplastic presence, abundance, and morphological types within the guts of four species of commercial marine fish ( Atule mate, Crenimugil seheli, Sardinella fimbriata and Rastrelliger brachysoma ) caught in seawater off the coast of Malaysia’s Northwest Peninsular. A total of 72 individual commercial marine fish guts from four species (fish per species n  = 18) were examined. Remarkably, this study found that 100% of the samples contained microplastics. A total number of 432 microplastics (size < 5 mm) from the four species were found in the excised marine fish guts. The most common type of microplastic discovered was fragment, which accounted for 49.5% of all microplastics present. The gut microplastic content differed between species. Sardinella fimbriata recorded the greatest amount of microplastic ingestion, with an average microplastic count of 6.5 (±4.3) items per individual fish. However, there were no statistically significant differences found when comparing study species and different locations. SEM-EDX analysis confirmed the presence of microplastic particles by identifying the chemical elements found in the samples. Since the four studied species of commercial marine fish are popular protein sources in Malaysians’ daily diet, this study suggests potential microplastic exposure to humans via contaminated fish consumption in Malaysia, which was previously unknown. Based on previous scientific evidence, this study also demonstrates the high probability of microplastic ingestion in marine fish in the Malaysian seawater, which could have an adverse effect on fish health as well as marine biota.