TAXONOMIC SURROGACY OF BENTHIC MACROINVERTEBRATES ALONG DEPTH AND SUBSTRATE GRADIENTS IN THE PUGET SOUND

Taxonomic surrogates are frequently used in paleontology due to limited preservation of the entire fauna and generally come in two forms: coarser resolutions than species-level data (‘sufficiency’, e.g., genus-level analysis) or a subset of the whole fauna (‘subsetting’, e.g., taxa with skeletal hardparts). However, the ability of taxonomic surrogates to reliably detect similar environmental gradients as the whole fauna remains underexplored. Here, the power of both methods was evaluated using an infaunal dataset from 10 stations sampled annually since 2016 within Puget Sound, Washington. Three taxonomic sets were used: the whole macrobenthic fauna, polychaetes-only (representing most species and individuals), and bivalves-only (common both alive and in fossil assemblages). Each taxonomic set was tested at five resolutions (species, genus, family, order, functional groups) along environmental gradients for water depth, grain size, and natural concentrations of organic carbon and nitrogen. Cluster analysis identified three general environments within Puget Sound: shallow-sandy, shallow-silty, and deep-silty. Community compositions ordinated along gradients with distance-based redundancy analysis (dRDA) showed strong correlation among sets at the species level. These patterns were preserved up to the family level for the whole fauna, but only survived genus-level coarsening for polychaetes and bivalves. Gradients strongly correlated to richness (number of taxa), with more taxa at sandy than at silty stations. Evenness (distribution of individuals among taxa) and Shannon index (summary of both richness and evenness) varied with no consistent pattern. Bivalves and polychaetes are both effective surrogates for the whole fauna at the species and genus levels, retaining a strong environmental signal, but family-level coarsening likely does not work due to the analysis of too few taxa. The power of taxonomic surrogacy in Puget Sound’s relatively natural depth- and substrate-driven system contrasts with our previous analyses on the southern California shelf where ordinal-level data effectively captures strong anthropogenic pollution signals. These results suggest that taxonomic surrogacy requires assessment across numerous scales and settings in order to converge on general rules.