Development of vision and larval feeding responses in southern bluefin tuna and yellowtail kingfish

Southern bluefin tuna, Thunnus maccoyii and yellowtail kingfish, Seriola lalandi, are marine finfish species currently cultured in Australia. High early mortality and larval malformations have hindered the successful production of quality juveniles. Rearing conditions experienced during early larval culture are critical for the production of high quality seed stock. My study described the visual capacity of T. maccoyii and S. lalandi to feed under a variety of abiotic and biotic conditions and examined retinal morphology and physiology to identify species-specific adaptations to help explain the observed feeding behaviour. The morphological development of the visual apparatus of T. maccoyii and S. lalandi is described by histological analysis and microspectrophotometry (MSP) and the visual ability of larvae is examined through behavioural experimentation. Larvae were visually challenged to feed under a range of conditions in short-duration (4 h) first-feeding experiments. Feeding performance was measured as the proportion of larvae feeding and the intensity of feeding. In T. maccoyii the first-feeding performance was positively affected by increasing prey density and lower turbidities and unaffected by light intensity, tank colour, turbulence, prey size and larval density. In contrast, S. lalandi showed greater limitation in terms of feeding performance, and in the number of variables that were conducive to feeding, which indicated a narrower environmental window‚ÄövÑvp for first-feeding success. Light intensity, tank colour, turbulence, larval density and prey density all significantly affected the feeding response in early-feeding S. lalandi. Feeding experiments on 3, 6 and 9 dph larvae revealed that as T. maccoyii aged, lower light intensities significantly increased feeding performance, indicating increased photopic sensitivity. In contrast, S. lalandi continued to display better feeding performance with increasing light intensity and age. Histological studies revealed significant differences between the retinal indices of T. maccoyii and S. lalandi larvae. Thunnus maccoyii displayed high cell densities in the ventral region of the eye, possessed a low convergence of cone cells onto ganglion cells, with relatively large cones at first feeding (almost twice the size of S. lalandi first-feeding larvae) and exhibited retinal pigment epithelium (RPE) pigment migration at an early age (12 dph). Seriola lalandi displayed high cell densities in the dorsal retinal region and on average possessed twice the convergence of cones onto ganglion cells. Retinal pigment epithelium migration in S. lalandi was not seen until the development of rods at 21 dph. MSP and behavioral feeding experiments under coloured lights showed that T. maccoyii displayed peak spectral sensitivity in the blue spectrum, whereas S. lalandi displayed sensitivity in the red and green spectrum. My study has also identified important differences between the species that have culture implications. The strong first-feeding response of T. maccoyii across a broad range of abiotic and biotic factors indicates that major mortality during their early life history is not associated with a failure to initiate feeding. Although, T. maccoyii are likely to show decreased feeding and increased surface and sinking mortality, when the larvae are exposed to light intensities commonly encountered in the culture of fish larvae. The use of low-light intensity during larval rearing has the potential to dramatically improve larval survival during the first two weeks of culture. In comparison, while S. lalandi displayed improved feeding with increasing age, the narrow set of parameters that were conducive to first-feeding highlighted the need for strict control of culture parameters during this critical stage.