Species surrogacy in environmental impact assessment and monitoring: extending the BestAgg approach to asymmetrical designs

The mainstream approach to define effective surrogates for species in routine biomonitoring focuses on the use of coarse levels of taxonomic resolution. A new approach to species surrogacy, the Best Practicable Aggregation of Species (BestAgg), which uses pilot data to generate null models of decreasing information at increasing levels of aggregation of variables, has been recently proposed. The approach has been tested in different environmental contexts, providing several advantages compared to the analysis at taxonomic levels higher than species. However, BestAgg still lacks a framework for its application to asymmetrical designs, thus limiting its general use in environmental impact assessment. Here, we implemented a new procedure to fill this gap and provided a specific R code for BestAgg in asymmetrical analysis. We presented an example of the application of this procedure to a real case study assessing the potential impact of harbour activities on subtidal sessile assemblages. Results from taxonomically fine-resolved data showed a significant effect of harbour activity on assemblage structure, although variable in time. Such patterns were consistent up to order level. Results based on surrogates from BestAgg were also aligned with those obtained at fine taxonomic resolution, but led to retention of much more information on original patterns and increased timesaving in sample processing compared to a classic approach based on the use of higher taxonomic levels. BestAgg represents a more formal procedure to species surrogacy than the empirical determination of the sufficient taxonomic resolution. The approach increases cost-efficiency while maximizing ecological information, and can be used under a wide range of experimental settings, now including asymmetrical designs.