Investigation of the prevalence and characterisation of infection by Kudoa thyrsites and K. paniformis in South African marine fish species

[摘要] ENGLISH ABSTRACT: Post-mortem myoliquefaction and formation of black pseudocysts in marine fish muscle are associated with infection by Kudoa thyrsites and/or K. paniformis. These quality defects become apparent only upon filleting, resulting in economic losses. Methods to detect infection include microscopy and polymerase chain reaction (PCR) techniques. These methods are time consuming, expensive, and destructive to the sample. Near-infrared (NIR) spectroscopy is a rapid, non-destructive method, used for identifying nematodes-infected fish. The aims of this study were to investigate the prevalence of K. thyrsites and K. paniformis in South African (SA) sardine (Sardinops sagax ocellatus), kingklip (Genypterus capensis) Cape hake (Merluccius capensis and M. paradoxus), and to use NIR spectroscopy to discriminate between infected and uninfected samples.A real-time quantitative PCR (qPCR) method was designed for use as qualitative determination of K. thyrsites. A total of 536 fish samples (296 SA sardine, 70 kingklip and 170 Cape hake) were analysed with qPCR. The SA sardine samples had K. thyrsites prevalence of 92%, Cape hake 79%, and SA kingklip 40%. Results showed that there was no significant difference in K. thyrsites prevalence between male and female (SA sardines, kingklip and Cape hake), season (SA sardines) and area of capture (SA sardines). Cape hake samples showed conflicting results for the relationship between K. thyrsites prevalence and size. A higher prevalence for smaller (39.63 ± 10.88 cm) than larger sized (49.46 ± 13.94 cm) samples was found in one study, while the other independent study showed no significant relationship to size. A limited number of Cape hake samples were used, resulting in larger sized fish to be poorly represented in this study.Infection by K. thyrsites had no significant effects (P > 0.05) on moisture and ash contents of SA sardine. Similarly, the moisture, ash and protein contents of infected Cape hake did not differ from uninfected samples. Moisture, ash and protein contents were therefore not useful in prediction of K. thyrsites infection.No NIR spectral differences between K. thyrsites infected and uninfected fish sample were observed when NIR spectroscopy and principal component analysis (PCA) were investigated. In addition, NIR spectroscopy, in combination with soft independent modelling of class analogy (SIMCA) and partial least square discriminant analysis (PLS-DA), was unable to distinguish infected from uninfected samples. Fish samples used in this study were frozen, resulting in water loss, protein degradation and aggregation, thus possibly masking the chemical and textural changes associated with K. thyrsites infection. NIR spectroscopy, in combination with SIMCA and PLS-DA, respectively, was therefore not a useful method to predict K. thyrsites infection in frozen fish samples.Further studies are suggested where the relationships between level of infection (number of K. thyrsites spores), myoliquefaction, and texture analyses, in combination with NIR spectroscopy, are investigated. However, since SA sardine has a high prevalence of infection, such a study may require more than 500 samples in order to ensure an equal amount of infected and uninfected samples for the development of reliable NIR classification models. For Cape hake, it is suggested to investigatethe use of imaging spectroscopy in an attempt to differentiate between infected and uninfected samples based on optical properties of fish muscle and black pseudocysts.

[发布机构] Stellenbosch University