The size and age at which individuals mature is rapidly changing due to plastic and evolved responses to fisheries harvest and global warming. Understanding the nature of these changes is essential because maturity schedules are critical in determining population demography and ultimately, the economic value and viability of fisheries. Detecting maturity changes is, however, practically difficult and costly. A recently proposed biphasic growth modelling likelihood profiling method offers great potential as it can statistically estimate age-at-maturity from population-level size-at-age data, using the change-point in growth that occurs at maturity. Yet, the performance of the method on typical marine fisheries datasets remains untested. Here, we assessed the suitability of 12 North Sea and Australian species’ datasets for the likelihood profiling approach. The majority of the fisheries datasets were unsuitable as they had too small sample sizes or too large size-at-age variation. Further, datasets that did satisfy data requirements generally showed no correlation between empirical and model-derived maturity estimates. To understand why the biphasic approach had low performance we explored its sensitivity using simulated datasets. We found that method performance for marine fisheries datasets is likely to be low because of: (1) truncated age structures due to intensive fishing, (2) an under-representation of young individuals in datasets due to common fisheries-sampling protocols, and (3) large intrapopulation variability in growth curves. To improve our ability to detect maturation changes from population level size-at-age data we need to improve data collection protocols for fisheries monitoring.

中文翻译：

---

从年龄大小数据估计成熟度：现实世界的渔业数据集能胜任这项任务吗？

由于塑料和对渔业收获和全球变暖的进化反应，个体成熟的大小和年龄正在迅速变化。了解这些变化的性质至关重要，因为成熟时间表对于确定人口结构以及最终确定渔业的经济价值和生存能力至关重要。然而，检测成熟度变化实际上是困难和昂贵的。最近提出的双相增长建模可能性分析方法提供了巨大的潜力，因为它可以使用成熟时发生的增长变化点从人口水平的年龄大小数据统计估计成熟年龄。然而，该方法在典型海洋渔业数据集上的性能仍未得到测试。这里，我们评估了 12 个北海和澳大利亚物种数据集对可能性分析方法的适用性。大多数渔业数据集都不合适，因为它们的样本量太小或年龄差异太大。此外，满足数据要求的数据集通常显示经验和模型得出的成熟度​​估计之间没有相关性。为了理解为什么双相方法性能低下，我们使用模拟数据集探索了它的敏感性。我们发现海洋渔业数据集的方法性能可能较低，因为：（1）由于密集捕捞导致年龄结构被截断，（2）由于常见的渔业采样协议，数据集中的年轻个体代表性不足，以及（ 3) 生长曲线的种群内变异性大。