Histological and biochemical procedures were employed to study liver histomorphology and plasma levels of aspartate aminotransferases (AST), alanine aminotransferase (ALT), triglyceride (TRIG), cholesterol (CHOL), glucose (GLU) and total protein (TP) in cultured (CSB) and wild sea bass (WSB) (Dicentrarchus labrax) from the northern part of the Adriatic Sea. Histopathological changes in liver included varying degrees of infiltration and lipid degeneration of hepatocytes in examined cultured fish. No significant differences between median AST values of CSB (44 IU) and WSB (45 IU) were observed. Values of ALT were <5 IU in both groups. TRIG, CHOL, GLU and TP levels were higher in CSB (2.08 mmol/L, 3.67 mmol/L, 10.66 mmol/L and 49 g/L, respectively) than in WSB (0.67 mmol/L, 2.74 mmol/L, 3.68 mmol/L and 36 g/L, respectively). The study showed differences between plasma biochemical parameters and liver histomorphology of CSB and WSB. This can be explained as a consequence of different diets (artificial in contrast to natural foods), which influence energy metabolism.

A comparative study of blood chemistry and histology was conducted on two groups of mullets (Mugilidae) living under different conditions with different feed sources. The aquaculture influenced mullet group (AIM), was collected near fish farms and the control group of mullet (CM) was caught in the waters without any aquaculture activities. Histological and biochemical procedures were employed to study liver histomorphology, plasma aspartate and alanine aminotransferase (AST, ALT), triglyceride (TRIG), cholesterol (CHOL), glucose (GLU) and total protein (TP) of both AIM and CM. Moderate histological changes (lipid infiltration) were observed in the liver of AIM. Significant changes in plasma variables were observed in AIM. Blood chemistry variables measured proved to be good indicators of artificial feed effects. Classical statistical approaches were applied to the blood chemistry and histopathology data. For the first time machine learning techniques were used to generate comprehensible classification models and to explore blood chemistry variable importance, strength, their mutual interactions or dependencies, and to investigate reliability of particular variables within the groups.

SummaryThe purpose of this study was the identification and quantification of biochemical parameters over a 1-year cycle and to provide a detailed picture of seasonal changes in plasma metabolites and enzymes. Using the novel methods of machine learning techniques, the authors created and generated for the first time comprehensible classification models for exploring the importance of blood chemistry parameters, strength, mutual interactions or dependencies, and reliability of particular parameters within the seasonal groups.

The health status of rudd (Scardinius erythrophthalmus hesperidicus H.) in Lake Vrana, the largest Croatian karstic lake, was evaluated. Studies comprising parasitological, haematological and bacteriological surveys were conducted over a 2‐year period. Parasitological examination revealed a light infestation of 27% of the examined fish, mostly Dactylogyrus and Ichthyophthirius species. Haematological studies showed that haematocrit values were lower than the physiological limit. A haematocrit coefficient of correlation in all research periods was higher than 15%, indicating that the majority of fish in the study were susceptible to the development of bacterial and other diseases. Indeed, a diverse array of bacteria were isolated from rudd, mainly Flavobacterium spp. and Aeromonas spp., but also some specific fish pathogens, notably Pasteurella piscicida, Yersinia ruckeri, and Edwardsiella ictaluri, were identified. Under stress conditions, detected bacterial species can give rise to disease outbreaks.