Norwegian Institute for Nature Research (NINA), Oslo, Norway.
Conserv Biol. 2011 Jun;25(3):577-86. doi: 10.1111/j.1523-1739.2010.01636.x. Epub 2011 Feb 1.
The most comprehensive data on many species come from scientific collections. Thus, we developed a method of population viability analysis (PVA) in which this type of occurrence data can be used. In contrast to classical PVA, our approach accounts for the inherent observation error in occurrence data and allows the estimation of the population parameters needed for viability analysis. We tested the sensitivity of the approach to spatial resolution of the data, length of the time series, sampling effort, and detection probability with simulated data and conducted PVAs for common, rare, and threatened species. We compared the results of these PVAs with results of standard method PVAs in which observation error is ignored. Our method provided realistic estimates of population growth terms and quasi-extinction risk in cases in which the standard method without observation error could not. For low values of any of the sampling variables we tested, precision decreased, and in some cases biased estimates resulted. The results of our PVAs with the example species were consistent with information in the literature on these species. Our approach may facilitate PVA for a wide range of species of conservation concern for which demographic data are lacking but occurrence data are readily available.
最全面的数据来自于科学收集的许多物种。因此，我们开发了一种种群生存力分析（PVA）方法，其中可以使用这种类型的发生数据。与经典的 PVA 相比，我们的方法考虑了发生数据中的固有观测误差，并允许估计生存力分析所需的种群参数。我们使用模拟数据测试了该方法对数据空间分辨率、时间序列长度、采样力度和检测概率的敏感性，并对常见、稀有和受威胁的物种进行了 PVA。我们将这些 PVA 的结果与忽略观测误差的标准方法 PVA 的结果进行了比较。在标准方法没有观测误差的情况下，我们的方法可以对种群增长项和准灭绝风险进行现实估计。在我们测试的任何采样变量值较低的情况下，精度降低，并且在某些情况下会导致有偏差的估计。我们对示例物种进行的 PVA 的结果与这些物种的文献信息一致。对于缺乏人口统计数据但易于获得发生数据的广泛保护关注的物种，我们的方法可以促进 PVA。
