Marine Protected Areas (MPAs) are increasingly used for biodiversity conservation and the management
of sustainable fisheries. The use of taxonomic surrogates in routine monitoring of the reserve effects on
fish assemblages may represent a promising method due to its substantial technical and economic benefits.
However, higher taxonomic ranks should be used as surrogates with caution, especially for fish,
where the approach is still virtually undocumented. This study aims to shed light on relationships
between taxonomic relatedness and ecological similarity, which is crucial to assess the relevance of species
surrogacy to reflect species-level information and detect changes in fish related to protection
regimes. By analyzing data from a Mediterranean MPA, we show that rather being related to taxonomic
relationships, the ability of higher taxa to reflect species-level patterns was explained in terms of aggregation
level and distribution of species within taxa. Null models using random aggregations of species
were created to identify the best surrogates able to depict changes in responses of assemblages to protection
observed at the species-level. Comparison of null model predictions with the more classical
higher-taxon approach revealed that the latter was not reliable because, unlike null model outcomes, surrogates
determined empirically were not relevant for other subsequent independent monitoring. The
effectiveness of species surrogates to depict changes in responses of assemblages to protection observed
at the species-level depended (1) on the numerical resolution of the aggregation and (2) the magnitude of
differences between protected and unprotected locations. Such findings suggest that surrogacy
approaches may be applied for routine monitoring of Mediterranean reef fish communities, although
mere empirical determinations of sufficient taxonomy seems to be not reliable, legitimizing the use of
alternative methods based on null models. Guidelines for the careful use of species surrogacy in the
ecological evaluation of MPAs on fish assemblages are provided