A taxonomic resolution assessment for deep-pelagic fish assemblage analysis in a high-diversity ecosystem

The species number of mesopelagic (200–1000 m depth) fishes is extremely rich in the Gulf of Mexico. The aim of this study is to evaluate whether coarser taxonomic resolutions (genus, family, order) can provide informative and reliable measures of fish assemblage structure when species-level identification is not available, and to assess how taxonomic resolution and data transformation influence the ability to discriminate among assemblages across well-established ecological gradients of depth and diel cycle in an oceanic ecosystem. A better understanding of these effects can inform analytical decisions about taxonomic sufficiency when using data that are not, or cannot be, resolved to species level, a common occurrence in deep-sea systems whose faunal inventories are incomplete. This study analyzed assemblage structure patterns in deep-pelagic fishes collected with a large-mesh, commercial-sized, high-speed rope trawl in the Gulf of Mexico from the epipelagic, mesopelagic, and upper bathypelagic zones (0– 1800 m, collectively). Assemblage compositions were compared at four taxonomic levels (species, genus, family, order) and six data transformations between sampling events collected during the day and at night, and across two depth strata. Assemblage structures were analyzed using taxonomic ratios, ANOSIM analyses, and a 2nd Stage resemblance coefficient analysis. The results show that genus-resolution data provided similar assemblage discrimination (94% effectiveness) as species data, with family and order data providing less suitable substitutions (65% and 8% respectively). Family- and order-level analyses were unable to discriminate between daytime and nighttime assemblages based on depth as effectively as species- and genus-level data. Second-stage analyses indicated that dissimilarities between taxonomic levels were dependent on the transformation applied, with the strongest data transformations leading to greater dissimilarity between taxonomic levels. Overall, these results suggest that in many cases where species-resolution data are not available, genus-level data can be used as a meaningful proxy to characterize deep-pelagic fish assemblages with respect to major ecological drivers, with the proviso that strong data transformations can amplify taxonomic differences.