1. Habitat modification and fragmentation are key factors responsible for fish population decline worldwide. Previous assessments documented a total of 72 species extinctions for the sole class of Actinopterygii. However, global extinctions are difficult to monitor or study based on fossil records. By contrast, local extinctions occurring at the population level are easier to study. Given this context, an important question relates to whether extinction dynamics studied at the local scale can provide useful information to understand extinctions occurring at larger scales. This would be the case if local extinctions were not balanced by recolonisation as in a classic metapopulation. Our aim is thus to explain the observed regional (per basin) persistence of 252 fish populations by testing contribution of local extinction rates and more generally metapopulation dynamics components.
2. To address this aim, we used the annual extinction probability of 252 regional populations of up to 14 species inhabiting 18 coastal rivers, which became isolated c. 8,500 years ago. We specifically compared extinction probabilities obtained by seven theoretical models to investigate whether regional extinction rates (i.e. loss from a river system) were correlated to local extinction rates (i.e. loss from an occupied site) and the role of metapopulation dynamics to explain regional persistence.
3. Using empirical data, we showed the importance of variables related to metapopulation dynamics to explain extinction rates across the 18 river systems. As expected, the regional extinction rate decreased with the colonisation rate, area, metapopulation size, and percentage of occupied localities. By contrast, an inconsistent relationship emerged between regional and local extinction rates, as species with high local extinction rates were not particularly prone to regional extinction.
4. Our results provide strong support for the contribution of colonisation rates to explain persistence. Overall, our results show that the equilibrium number of occupied localities could be a good predictor of the long‐term persistence of metapopulations in rivers. Finally, our results suggest the importance of connectivity to maintain sustainable populations within the river system.

中文翻译：

---

迁移动力学解释河流系统鱼类持久性的重要性

1. 栖息地的改造和破碎化是导致世界鱼类种群减少的关键因素。先前的评估记录了唯一的放线opter纲动物共有72种物种灭绝。但是，基于化石记录很难监测或研究全球灭绝事件。相反，在人口层面发生的局部灭绝更容易研究。在这种情况下，一个重要的问题涉及在地方尺度上进行的灭绝动力学是否可以提供有用的信息，以了解在更大范围内发生的灭绝。如果局部灭绝不能像传统的聚居种群那样通过重新定殖来平衡，那就是这种情况。
2. 为了实现这一目标，我们使用了居住在18条沿海河流中的多达14种物种的252个区域种群的年度灭绝概率，这些物种被孤立了c。8500年前。我们专门比较了通过七个理论模型获得的灭绝概率，以调查区域灭绝速率（即河流系统的损失）是否与当地灭绝速率（即从被占领地点的损失）相关，以及种群迁移动力学在解释区域持久性中的作用。
3. 使用经验数据，我们表明了与种群动态相关的变量对解释18个河流系统灭绝率的重要性。正如预期的那样，区域灭绝率随定居率，面积，种群数量和所占地区的百分比而降低。相比之下，区域灭绝速度与区域灭绝速度之间存在不一致的关系，因为具有高局部灭绝速度的物种并非特别容易发生区域灭绝。
4. 我们的结果为殖民率对解释持久性的贡献提供了有力的支持。总体而言，我们的结果表明，被占领地区的均衡数量可以很好地预测河流中种群的长期持续存在。最后，我们的结果表明了连通性对于维持河流系统内可持续人口的重要性。