Anthropogenic sound can affect fish behaviour and physiology which may affect their well-being. However, it remains a major challenge to translate such effects to consequences for fitness at an individual and population level. For this, energy budget models have been developed, but suitable data to parametrize these models are lacking. A first step towards such parametrization concerns the objective quantification of behavioural states at high resolution. We experimentally exposed individual Atlantic cod (Gadus morhua) in a net pen to the playback of seismic airgun sounds. We demonstrated that individual cod in the net pen did not change their swimming patterns immediately at the onset of the sound exposure. However, several individuals changed their time spent in three different behavioural states during the 1 h exposure. This may be translated to changes in energy expenditure and provide suitable input for energy budget models that allow predictions about fitness and population consequences.

人为的声音会影响鱼类的行为和生理，从而可能影响鱼类的健康。然而，将这种影响转化为对个人和人群健康的影响仍然是一个重大挑战。为此，已经开发了能量预算模型，但是缺少合适的数据来参数化这些模型。进行此类参数化的第一步涉及以高分辨率对行为状态进行客观量化。我们实验性地暴露了单个大西洋鳕鱼（Gadus morhua），用网笔播放地震气枪的声音。我们证明了网笔中的单个鳕鱼并没有在暴露声音后立即改变其游泳方式。但是，几个人在1小时的暴露过程中改变了在三种不同行为状态下花费的时间。这可以转化为能源支出的变化，并为能源预算模型提供适当的输入，从而可以进行有关适应度和人口后果的预测。