Pires Mateus M, Grech Marta G, Stenert Cristina, Maltchik Leonardo, Epele Luis B, McLean Kyle I, Kneitel Jamie M, Bell Douglas A, Greig Hamish S, Gagne Chase R, Batzer Darold P
Universidade do Vale do Rio dos Sinos (UNISINOS), 950 Unisinos av, São Leopoldo, RS, Brazil.
Centro de Investigación Esquel de Montaña y Estepa Patagónica (CONICET-UNPSJB), Roca 780, Esquel, Chubut, Argentina.
Ecol Indic. 2021 Jun;125. doi: 10.1016/j.ecolind.2021.107437. Epub 2021 Feb 15.
The efficiency of biodiversity assessments and biomonitoring studies is commonly challenged by limitations in taxonomic identification and quantification approaches. In this study, we assessed the effects of different taxonomic and numerical resolutions on a range of community structure metrics in invertebrate compositional data sets from six regions distributed across North and South America. We specifically assessed the degree of similarity in the metrics (richness, equitability, beta diversity, heterogeneity in community composition and congruence) for data sets identified to a coarse resolution (usually family level) and the finest taxonomic resolution practical (usually genus level, sometimes species or morphospecies) and by presence-absence and relative abundance numerical resolutions. Spearman correlations showed highly significant and positive associations between univariate metrics (richness and equitability) calculated for coarse- and finest-resolution datasets. Procrustes analysis detected significant congruence between composition datasets. Higher correlation coefficients were found for datasets with the same numerical resolutions regardless of the taxonomic level (about 90%), while the correlations for comparisons across numerical resolutions were consistently lower. Our findings indicate that family-level resolution can be used as a surrogate of finer taxonomic resolutions to calculate a range of biodiversity metrics commonly used to describe invertebrate community structure patterns in New World freshwater wetlands without significant loss of information. However, conclusions on biodiversity patterns derived from datasets with different numerical resolutions should be critically considered in studies on wetland invertebrates.
生物多样性评估和生物监测研究的效率通常受到分类鉴定和量化方法局限性的挑战。在本研究中，我们评估了不同分类和数值分辨率对来自北美和南美六个地区的无脊椎动物组成数据集一系列群落结构指标的影响。我们特别评估了粗分辨率（通常为科级）和实际可行的最精细分类分辨率（通常为属级，有时为种级或形态种级）以及存在-缺失和相对丰度数值分辨率下识别的数据集在指标（丰富度、公平性、β多样性、群落组成异质性和一致性）方面的相似程度。斯皮尔曼相关性显示，粗分辨率和最精细分辨率数据集计算的单变量指标（丰富度和公平性）之间存在高度显著的正相关。普罗克汝斯分析检测到组成数据集之间存在显著一致性。无论分类水平如何，具有相同数值分辨率的数据集的相关系数更高（约90%），而跨数值分辨率比较的相关性始终较低。我们的研究结果表明，科级分辨率可以用作更精细分类分辨率的替代，以计算一系列常用于描述新世界淡水湿地无脊椎动物群落结构模式的生物多样性指标，而不会有显著的信息损失。然而，在湿地无脊椎动物研究中，应审慎考虑从具有不同数值分辨率的数据集中得出的生物多样性模式结论。
