Addressing assumptions: variation in stable isotopes and fatty acids of marine macrophytes can confound conclusions of food web studies

Studies that use biomarkers to elucidate consumer diets often must assume that these signatures are relatively invariant in space and time. We tested this assumption for multiple stable isotopes (MSI: delta C-13, delta N-15, delta S-34) and fatty acids in 10 marine macrophytes (macroalgae and seagrass) on 3 different dates, and also quantified MSI at 3 sites in the coastal northeast Pacific. For all comparisons, we found significant variation in biomarkers among species, sites, and dates; furthermore, there were always significant site x species and date x species interaction terms, indicating that biomarkers do not change consistently across species among dates or sites. Despite this variation within and among species, biomarkers could readily distinguish macrophyte phyla. To observe how variation could affect conclusions about diets, we used a Bayesian mixing model to evaluate scenarios for a theoretical consumer given diverse diets, and with varying assumptions about the way it integrates foods over sites and seasons. Accuracy of the model runs (predicted diet versus simulated diet) increased with the number of biomarker variables, and depended strongly on initial assumptions about diets. The lowest accuracy occurred when biomarker values were based on macrophytes sampled only from 1 season but the consumer integrated foods across multiple seasons, a situation commonly seen in biomarker literature. Contrary to the notion that natural biomarker variation reduces insight into food web structure, exploring the potential mechanisms behind this variation should provide a more realistic view of coastal ecosystem dynamics.