Performance of richness estimators for invertebrate inventories in reservoirs

• Published in: Environmental monitoring and assessment 2021-10, Vol.193 (10), p.686-686, Article 686
• Main Authors: Brito, Pablo Gouveia, Jovem-Azevêdo, Daniele, de Araújo Campos, Magnólia, Paiva, Franciely Ferreira, Molozzi, Joseline
• Format: Article
• Language: English
• Subjects: Accuracy; Aquatic insects; Aquatic invertebrates; Aquatic organisms; Atmospheric Protection/Air Quality Control/Air Pollution; Benthos; Bias; Earth and Environmental Science; Ecology; Ecotoxicology; Environment; Environmental Management; Environmental monitoring; Environmental science; Estimators; Geographical distribution; Identification; Inventories; Invertebrates; Littoral environments; Littoral zone; Macroinvertebrates; Monitoring/Environmental Analysis; Nonparametric statistics; Pattern analysis; Performance evaluation; Reservoirs; River basins; Rivers; Semi arid areas; Semiarid lands; Semiarid zones; Species; Species richness; Zoobenthos
• ISSN: 0167-6369; 1573-2959
• DOI: 10.1007/s10661-021-09487-z

Biological inventories combined with the estimation of species richness represent a useful tool for the analysis of the pattern of species distribution in different regions. This study aimed to (i) comparatively evaluate the performance of non-parametric richness estimators for invertebrate inventories in reservoirs between ecoregions and (ii) to assess whether the efficiency (bias, precision and accuracy indices) of the estimators is altered when applied to sites from different ecoregions. The study was conducted in the ecoregions Central Pediplano of the Borborema Plateau (Paraíba River basin) and Northern Sertaneja Depression (Piranhas-Assu River basin), semiarid region of Brazil. Six reservoirs were selected and benthic macroinvertebrates were sampled at 141 sites distributed along the littoral zone, in 4 periods (June, September, December 2014 and March 2015). The organisms were identified to the family level, except for Chironomidae, identified to the genus level. We comparatively analyzed six non-parametric richness estimators: Jackknife 1 , Jackknife 2 , Chao1 , Chao 2 , ICE , and Bootstrap , and three performance indicators: bias, precision, and accuracy. ICE and Jackknife 2 had more stable results for total species richness, but with different performance between ecoregions for bias, precision, and accuracy. Variation in performance of the estimators may be associated with differences in species richness and frequency between ecoregions. ICE and Jackknife 2 proved to be the best estimators for biological inventories of aquatic invertebrates in reservoirs in studies comparing data from different ecoregions, due to accuracy and precision, while Bootstrap is the least indicated, given greater bias and less accuracy and precision.