Pseudo-nitzschia is a cosmopolitan genus of marine diatoms known for its production of the neurotoxin domoic acid (DA) and resultant harmful algal blooms. The production of DA varies across species; of approximately 37 defined Pseudo-nitzschia species, at least 14 have produced DA in laboratory studies. Previous work based on ribosomal RNA sequences suggested the genus Pseudo-nitzschia could be divided into two major clades, or groups made up of an ancestor and its descendants. With the discovery of new species and identification of cryptic species complexes, the structure of these clades has eroded. We sought to integrate phylogeny, morphology and physiology of Pseudo-nitzschia, by re-examining their phylogeny, using both ribosomal LSU sequences and the rbcL gene, which encodes the large subunit of the carbon fixation enzyme, ribulose-1,5-bisphosphate carboxylase oxygenase (RuBisCO). Trees based on rbcL sequences provided higher resolution to resolve the branching orders of Pseudo-nitzschia species. By mapping physiological characteristics such as cell size, frustule morphology, pigment content and DA production onto the tree we can interpret the evolutionary history of these features and species distributions in the context of our presumed clades.

The cosmopolitan pennate diatom Pseudo-nitzschia is notable for production of the neurotoxin domoic acid. Recently, a virus was isolated on Pseudo-nitzschia multiseries that has a broad host range, infecting many genera of diatoms. This is contrary to previous reports of very specific host ranges amongst diatom viruses, and suggests a nested infection network whereby specialist viruses infect only the most permissive hosts, while generalist viruses can also infect hosts that are well-defended. To explore the infection patterns of Pseudo-nitzschia and its viruses, new Pseudo-nitzschia strains were isolated monthly at a coastal and estuarine location in the Pacific Northwest for one year. Over 100 new strains of Pseudo-nitzschia were identified by sequencing the ITS region. These new strains are then challenged with environmental viral communities that were sampled at each location concurrently. To date, 28 of 47 strains tested are susceptible to infection by at least one viral community. Three strains are lysed by the majority of viral communities, suggesting we have identified permissive hosts. This analysis will continue until all currently isolated strains are challenged. Seasonally, May and July contained viruses that infected the widest range of strains, suggesting either high concentration of generalist viruses in those months, or high concentration of both specialist and generalists together. Understanding the seasonal structure of these infection networks will inform future virus isolation attempts and provide insight into a key component of these diatom-virus interactions.