• Mussel saxitoxin increased with Alexandrium relative abundance. • 1.24% (87th percentile) Alexandrium relative abundance aligned with elevated saxitoxin. • The 87th percentile level across 14 taxa was then compared to chl-a values • 4.6 µg/L chl-a in fall corresponded with elevated levels of Heterosigma, Dinophysis, and Heterocapsa Harmful Algal Blooms (HABs) are a hazard for coastal environments worldwide; identifying screening thresholds of chlorophyll- a (chl- a) associated with increased risk of HABs is a management priority. Molecular surveillance of coastal phytoplankton and bivalve biotoxins could be used to link chl- a with HAB risk, but requires an understanding of whether the HAB risks increase uniformly as chl-a rises, or whether some taxa are disproportionately favored, and if these relationships vary by season. In this study, we present a novel use of molecular abundance data to investigate the scientific bases for estuarine chl- a thresholds protective against HABs. In San Francisco Bay (SFB), California, the relationship between molecular relative abundance (as measured by 18S metabarcoding) of nine different HAB taxa, absolute quantitative polymerase chain reaction (qPCR) abundance, and mussel toxin concentrations of a subset of the taxa were investigated for thresholds as a function of increasing chl- a . Our results show most HAB taxa did not increase in absolute or relative abundance during SFB’s spring bloom interval, when chl- a levels were highest (>10 µg/L) but the assemblage was dominated by non-harmful diatoms. However, several flagellated, mixotrophic taxa did increase above their molecular baseline in fall, and the combined probability of any HAB occurring above baseline was elevated when chl- a reached ∼4.6 µg/L in the fall. This work demonstrates the promise of molecular approaches in disentangling the seasonally complex interplay between stressors and phytoplankton/HAB community responses and has the potential to provide clearer, more cost-effective monitoring and mitigation strategies for managers.
Killam et al. (Sat,) studied this question.