Research at the Research Institute for Farm Animal Biology (FBN) Facing problems of the Embryo-Larval-Threshold in Pikeperch Aquaculture

To gain general insights into growth patterns during the problematic ontogenetic stages, we conducted a morphometric analysis. In specimens of different embryonic and larval age stages, 18 parameters were measured, including parameters related to endogenous nutrition. The results indicate a steady and uniform growth in the embryonic stages, while in larval stages increasing size differences were observed between specimens of the same age. In addition, there was an intermediate threshold stage with dormant length growth but increased energy expenditure.

Subsequent molecular analysis of 19 selected genes associated with myogenesis and further developmental processes revealed two phases of increased muscle gene expression at the observed age stages. A first increase occurred in the late embryonic stages that prepared for basic locomotion and formed a genetic stem cell pool. A second increase in muscle gene expression took place during larval development when muscle mass increased, further enhancing locomotion and prey hunting abilities. Due to increasing size differences and improved locomotion, cannibalistic activity rises during the larval stages. To analyse the cause of differential growth, we examined the gene expression patterns in different size classes at early and late larval stages, as the genetic mechanisms may provide a clue to differential growth that could be influenced in aquaculture.

In pikeperch aquaculture, efforts are being made to prevent high mortality rates through size sorting. To date, this requires continuous manual size control and additional stressful handling of larvae. To support large-scale rearing, we started to develop a video monitoring system that will use image analysis software to automatically detect larval numbers and sizes. For this purpose, pikeperch hatchlings were placed in an aquarium surrounded by cameras. The video sequences obtained were analysed using single frames (OpenCV-software) and served as the basis for training a neuronal net (Tensorflow) to automatically extract the number of larvae using video analysis algorithms.

The obtained results will help to improve the management of pikeperch rearing in aquaculture facilities. In addition, the developed monitoring system can be a first step towards digitalisation, and thus long-term improvement of pikeperch rearing.