A reliable response of bioindicators to environmental variation is a cornerstone for effective bioassessment and biomonitoring. Fish and aquatic macroinvertebrates are widely used as bioindicators of different human impacts in freshwater ecosystems, but the cost-effectiveness of their usage can be improved through the use of surrogates. We investigated congruence patterns between using different taxonomic and numeric resolutions for aquatic macroinvertebrates and fish to assess community-environment relationships. We also tested whether dataset characteristics (e.g., area sampled, species pool) could explain the variation in the effectiveness of using different taxonomic and numerical resolutions. We used a Brazilian nationwide database encompassing multiple datasets with a gradient of riparian deforestation each. Our findings suggest that families and genera can effectively represent macroinvertebrate genera and fish species, respectively, when using community matrices for assessing community-environment relationships, with an acceptable loss of information. EPT (Ephemeroptera, Plecoptera, and Trichoptera) and Characiformes or Siluriformes may be used as a surrogate, in some cases, for the entire assemblages of macroinvertebrates and fish, respectively, but their use may result in higher loss of information. Presence-absence data also presented a minimal loss of information compared to abundance data, for both macroinvertebrates and fish. The variation in congruence levels among macroinvertebrate datasets was less predicted by dataset characteristics than fish. Across distinct resolutions, on average, 10% and 19% of the variation in community composition of macroinvertebrates and fish, respectively, was explained by broad-scale environmental variables, and the effect size was negatively affected by the dataset's sample size and spatial extent for fish. Whereas identification at species (fish) and genus (macroinvertebrates) level and quantification of all individuals still provide the best scenario, we provide evidence that coarser taxonomical resolution and presence-absence data can be used as cost-effective alternatives to facilitate biomonitoring and bioassessment of freshwaters in the Neotropical region impacted by deforestation.
Read full abstract