A variety of tagging techniques are now available to monitor fish behaviour, physiology and their environmental experience. Tagging is frequently used in aquaculture research to monitor free-swimming individuals within farmed populations. However, for information gathered from tagged fish to be representative of farmed populations, tagging must not fundamentally affect fish behaviour, physiology or survival. Here, we systematically review studies that used tags to monitor farmed fish behaviour and test factors that affect tag retrieval and tag-related mortality. Most studies using tags assessed movement and swimming behaviour in salmonids, predominantly in Europe and North America. Mortality of tagged fish was 10 times higher in sea-cages (mean = 25%, range = 0–61.5%, n = 22 studies) than in tanks (mean = 2.5%, range = 0–17%, n = 23 studies), while mortality of tagged fish in sea-cages was markedly higher in longer trials (from 4% in single day trials to 36% after 100 days). Higher-than-usual mortality rates among tagged fish, together with largely unknown sublethal effects on behaviour, should caution against using tagging studies to make decisions related to farm management. Moreover, key metrics such as mortality rates of tagged and untagged fish or evidence of sublethal effects are often unreported. We make several recommendations to improve future tagging studies and increase transparency in reporting. A greater insight into the causes of tagged fish mortality in sea-cages is required to secure animal welfare and data validity in studies that use tags to assess fish behaviour in aquaculture.

中文翻译：

---

水产养殖中用于监测鱼类行为的标签：收益、问题和解决方案综述

现在可以使用多种标记技术来监测鱼类的行为、生理和它们的环境体验。标签经常用于水产养殖研究，以监测养殖种群内自由游泳的个体。然而，为了从带标签的鱼收集到的信息代表养殖种群，标签不能从根本上影响鱼的行为、生理或生存。在这里，我们系统地回顾了使用标签来监测养殖鱼类行为和测试影响标签检索和标签相关死亡率的因素的研究。大多数使用标签的研究评估了鲑鱼的运动和游泳行为，主要是在欧洲和北美。在海笼中标记鱼的死亡率高 10 倍（平均值 = 25%，范围 = 0–61.5%，n = 22 研究）比在坦克中（平均值 = 2.5%，范围 = 0–17%，n = 23 项研究），而在更长的试验中，海笼中标记鱼的死亡率明显更高（从单日试验的 4% 到 100 天后的 36%）。标记鱼的死亡率高于正常水平，加上对行为的亚致死影响很大程度上未知，因此应谨慎使用标记研究来做出与养殖场管理相关的决策。此外，标记和未标记鱼的死亡率或亚致死效应的证据等关键指标往往未被报告。我们提出了几项建议，以改进未来的标签研究并提高报告的透明度。在使用标签评估水产养殖中鱼类行为的研究中，需要更深入地了解海网箱中标记鱼类死亡的原因，以确保动物福利和数据有效性。