van Proosdij André S J, Raes Niels, Wieringa Jan J, Sosef Marc S M
Biosystematics Group, Wageningen University, Wageningen, the Netherlands.
Naturalis Biodiversity Center, Leiden, the Netherlands.
PLoS One. 2016 Dec 29;11(12):e0169200. doi: 10.1371/journal.pone.0169200. eCollection 2016.
In conservation studies, solely widespread species are often used as indicators of diversity patterns, but narrow-ranged species can show different patterns. Here, we assess how well subsets of narrow-ranged, widespread or randomly selected plant species represent patterns of species richness and weighted endemism in Gabon, tropical Africa. Specifically, we assess the effect of using different definitions of widespread and narrow-ranged and of the information content of the subsets. Finally, we test if narrow-ranged species are overrepresented in species-rich areas. Based on distribution models of Gabonese plant species, we defined sequential subsets from narrow-ranged-to-widespread, widespread-to-narrow-ranged, and 100 randomly arranged species sequences using the range sizes of species in tropical Africa and within Gabon. Along these sequences, correlations between subsets and the total species richness and total weighted endemism patterns were computed. Random species subsets best represent the total species richness pattern, whereas subsets of narrow-ranged species best represent the total weighted endemism pattern. For species ordered according to their range sizes in tropical Africa, subsets of narrow-ranged species represented the total species richness pattern better than widespread species subsets did. However, the opposite was true when range sizes were truncated by the Gabonese national country borders. Correcting for the information content of the subset results in a skew of the sequential correlations, its direction depending on the range-size frequency distribution. Finally, we find a strong, positive, non-linear relation between weighted endemism and total species richness. Observed differences in the contribution of narrow-ranged, widespread and randomly selected species to species richness and weighted endemism patterns can be explained by the range-size frequency distribution and the use of different definitions of widespread or narrow-ranged. We call for a reconsideration of the use of widespread species as an indicator of diversity patterns, and advocate using the full ranges of species when assessing diversity patterns.
在保护研究中，通常仅将分布广泛的物种用作多样性模式的指标，但分布范围狭窄的物种可能呈现出不同的模式。在此，我们评估了分布范围狭窄、广泛或随机选择的植物物种子集在非洲热带地区加蓬的物种丰富度和加权特有性模式方面的代表性如何。具体而言，我们评估了使用分布广泛和狭窄的不同定义以及子集信息含量的影响。最后，我们测试分布范围狭窄的物种在物种丰富地区是否被过度代表。基于加蓬植物物种的分布模型，我们根据非洲热带地区和加蓬境内物种的分布范围大小，定义了从分布范围狭窄到广泛、从广泛到狭窄以及100个随机排列的物种序列的连续子集。沿着这些序列，计算了子集与总物种丰富度和总加权特有性模式之间的相关性。随机物种子集最能代表总物种丰富度模式，而分布范围狭窄的物种子集最能代表总加权特有性模式。对于根据其在非洲热带地区的分布范围大小排序的物种，分布范围狭窄的物种子集比分布广泛的物种子集更能代表总物种丰富度模式。然而，当分布范围大小被加蓬国界截断时，情况则相反。校正子集的信息含量会导致连续相关性出现偏差，其方向取决于分布范围大小的频率分布。最后，我们发现加权特有性与总物种丰富度之间存在强烈的正非线性关系。分布范围狭窄、广泛和随机选择的物种对物种丰富度和加权特有性模式贡献的观察差异，可以通过分布范围大小的频率分布以及对分布广泛或狭窄的不同定义的使用来解释。我们呼吁重新考虑将分布广泛的物种用作多样性模式指标的做法，并主张在评估多样性模式时使用物种的完整分布范围。
