Min Wan has completed her Ph.D in 2010 from Karolinska Institutet, Sweden, and continued her
postdoctoral studies in Karolinska Institutet. At present, she is working as a distinguished professor
and the Vice Dean of Fisheries College in Ocean University of China. The research interests of Min
Wan focus on the host-microbe interaction in fish, and aim to clarify how gut microbiota influences
the metabolism and immunity in fish. She has received a series of competing external funding to
support her research and published nearly 40 papers in reputed journals.

Abstract
As one of the main aquaculture species, fish plays an important role in the food supply of aquatic products. Fish behavior can indirectly reflect their living state, such as the difference in the behavior of fish under environmental stress, hunger, fighting, reproduction, and other conditions. Currently, the behavior of fish under various situations is still in the form of artificial observation, which has the disadvantages of strong subjectivity, time-consuming and labor-intensive, and cannot be quantitatively analyzed. Additionally, fluorescent labeling or sensor labeling is a commonly used method to observe fish behavior, but the implantation of labels often causes stress reactions to fish and interferes with the experimental results. Therefore, we established multiple fish behavior datasets, and proposed approaches of fish behavior analysis based on computer vision to detect reproductive behavior, quantitative statistics of ammonia nitrogen stress behavior, and classification of feeding intensity based on multi-modal information fusion. These approaches can not only be used for qualitative analysis of fish behavior, but also for quantitative analysis, and have the advantage of non-invasive. We believe that using computer vision-based methods to research fish behavior can further improve fish welfare, farming efficiency and yield, and empower aquaculture automation and intelligence.

Abstract 
The ongoing climate changes generally worsen the oxygen conditions in the sea due to increased temperature and eutrophication. In open sea cages stocked salmon and trout, low dissolved oxygen (DO) concentration can be a potential problem, especially in late summer and autumn at peak water temperature and high fish density. Over the last ten years, use of so-called lice skirts shielding salmon from sea lice attacks, has been widely applied. Unfortunately, deep lice skirts reduce the water exchange through the cages and reinforce the oxygen deficit problems. A minimum concentration of ca. 80% DO throughout the growth cycle is generally considered adequate to ensure good fish welfare and growth conditions. There are two principle ways to improve the oxygen conditions in open or lice shielded cages: by direct injection of pure oxygen or by pumping of oxygen-rich deep water to the surface layer. Oxygenation, i.e. adding on-site produced oxygen or liquid oxygen by a diffuser network in salmon cages, has demonstrated reduced mortality and maintained feed appetite during warm periods with critical DO deficit. Pumping of water from 15-20 m depth using specially designed air-lifts run by large air bubbles is a promising method to improve the DO levels in the upper parts of salmonid cages.

Biography

Asbjørn Bergheim is a former senior researcher at IRIS – International Research Institute of Stavanger. At present, he is a research leader at Bio Marine (www.biomarine.no). He is Dr scient from NMBU – The Norwegian University of Life Sciences. He has published more than 50 papers in international journals and has been editorial member of the journal Aquacultural Engineering since 1996.