We propose to combine experiments, molecular ecology and modelling to investigate influences of hypoxia on habitat choice, boldness, recovery from stress and swimming activity in relation to individual genotype and physiology in fish. Comparisons both wit hin (among different populations) and across species will be used to test the hypothesis that hypoxia has unequal effect on the behaviour of different haemoglobin genotypes. We have chosen fish as model organisms because they are known to respond directly to local abiotic factors such as oxygen, temperature, salinity and pH, and we select Atlantic cod (Gadus morhua), three-spined stickleback (Gasterosteus aculeatus) and two-spotted goby (Gobiusculus flavescens) because these species are found in a range o f variable aquatic and marine environments under low and high renewal rates of oxygen. For cod there is documented spatial heterogeneity in the frequencies of haemoglobin genotypes among populations, but also between genotypes among populations, but also between genotype variations in feeding behaviour. And the responses may reflect different physiological demands as the haemoglobin molecules differ in their efficiency of oxygen transport from the gas exchange surface at the gi lls. This may lead to unequal behavioural responses to the availability of oxygen in the local environments. A consequence of this could be that habitat choice, how well individuals cope with hypoxic stress as well as activity level, will vary among genot ypes and across populations in habitats that differ in oxygen level. These aspects have not yet been studied and causes of spatial heterogeneity in haemoglobin genotype distributions in cod are still unresolved. Studies on whether it could be associated w ith gradients in temperature have been inconclusive and further research will therefore be needed. Our aims are to clarify some behavioural mechanisms associated with oxygen availability. Molecular techniques will reveal possible heterogeneity in haemoglobin genotypes across populations and species. Experiments will provide input values to a bio-energetic model that will be developed to describe functional relationships between oxygen level and oxygen uptake for haemoglobin genoty pes. The work will be the first step towards developing an individual-based model for studies of how spatial and temporal heterogeneity in fish depends on haemoglobin genotype and environmental dynamics.