Área de Probabilidad y Estadística, Centro de Investigación en Matemáticas AC, Col Valenciana, CP 36240 Guanajuato, Gto, México.
J Theor Biol. 2013 Sep 21;333:166-73. doi: 10.1016/j.jtbi.2013.05.015. Epub 2013 May 31.
Rates of extinction can be estimated from sighting records and are assumed to be implicitly constant by many data analysis methods. However, historical sightings are scarce. Frequently, the only information available for inferring extinction is the date of the last sighting. In this study, we developed a probabilistic model and a corresponding statistical inference procedure based on last sightings. We applied this procedure to data on recent marine extirpations and extinctions, seeking to test the null hypothesis of a constant extinction rate. We found that over the past 500 years extirpations in the ocean have been increasing but at an uncertain rate, whereas a constant rate of global marine extinctions is statistically plausible. The small sample sizes of marine extinction records generate such high uncertainty that different combinations of model inputs can yield different outputs that fit the observed data equally well. Thus, current marine extinction trends may be idiosyncratic.
灭绝率可以根据目击记录来估计，许多数据分析方法都假设其是隐含不变的。然而，历史目击记录却十分稀少。通常，唯一可用于推断灭绝的信息是最后一次目击的日期。在本研究中，我们基于最后一次目击记录，开发了一种概率模型和相应的统计推断程序。我们将该程序应用于最近的海洋灭绝和绝灭数据，旨在检验灭绝率不变的零假设。我们发现，在过去的 500 年中，海洋中的灭绝事件一直在增加，但增加的速度不确定，而全球海洋灭绝的恒定速率在统计学上是合理的。海洋灭绝记录的样本量很小，导致这种不确定性非常高，以至于模型输入的不同组合可以产生同样拟合观测数据的不同结果。因此，当前的海洋灭绝趋势可能是特殊的。
