Original ArticlesBiodiversity shortcuts in biomonitoring of novel ecosystems

- Presence/absence data can replace abundance data.
- Genera and species data can be a strategy for the monitoring of new ecosystems.
- The use of substitute groups was inconsistent.
- Different zooplankton groups should be monitored.

Hydropower reservoirs are novel ecosystems that present different challenges for the design of biomonitoring programs. To ensure long-term programs and wide spatial coverage, it is important to test the reliability of different cost-saving strategies that have been widely evaluated among researchers, such as taxonomic sufficiency, numerical sufficiency and surrogate groups. Using data on zooplankton composition, our objective was to test whether these strategies could be applied to increase the efficiency of biomonitoring programs in reservoirs. Zooplankton data were collected at the Santo Antônio do Jari Hydroelectric Plant, which is located between the states of Pará and Amapá (Amazon region, Brazil), over 23 months between 2012 and 2015. The data were organized in different taxonomic groups (cladocerans, copepods, rotifers and testate amoebae) and matrices by decreasing the taxonomic resolution (from species to genera and families) and the numerical resolution (from species abundance to species presence/absence) of the data. The ordination patterns obtained with Principal Coordinate Analysis for the different matrices were compared using Procrustes analyses. Our results suggest that ordination patterns using genus-level data were similar to those obtained with species-level data. However, analyses based on family-level data were often unable to reproduce results based on species-level data. Ordination patterns using presence/absence data were similar to those obtained from abundance data. We also found that the strengths of the relationships between ordinations derived from different taxonomic groups (e.g., rotifers and cladocerans) were low and often not significant. We conclude that the use of zooplankton genera and presence/absence data may be a reliable strategy to monitor reservoirs. However, our results highlight the need to monitor different zooplankton groups, as the ordination patterns depicted by a given group were poorly related to those generated by a second zooplankton group.