Di-nitrogen (N₂) fixers, also called diazotrophs, are able to reduce atmospheric N₂ into bioavailable nitrogen, giving them an advantage in open ocean regions with low dissolved inorganic nitrogen concentrations. The focus of this thesis are three lineages of symbiotic heterocystous filamentous types (het-1, het-2 and het-3), that associate with several genera of microalgae called diatoms (collectively referred to as Diatom Diazotroph Associations, DDAs). Other major cyanobacterial diazotrophs in the ocean are the filamentous Trichodesmium spp., and the unicellular UCYN-A, UCYN-B and UCYN-C. Although widespread in the tropics and subtropics, and first described in the early 20th century, the DDAs are an understudied group of diazotrophs. Hence, our knowledge of their distribution, abundance, activity, and how these are constrained by the environment is limited.

Initially we investigated the abundances and distributions of eight cyanobacterial diazotrophs, and two proposed micro-algal hosts of UCYN-A1 and A2, in the western tropical south Pacific (WTSP), using quantitative polymerase chain reaction (qPCR). Trichodesmium spp. was the most abundant diazotroph and het-1 was the most abundant DDA symbiont. Using correlation analysis a distinct vertical separation was observed between UCYN-A and the other diazotrophs (Trichodesmium spp., UCYN-B and DDA symbionts). The most influential environmental parameter on the diazotroph abundances in the WTSP was temperature, and in order to investigate this further we compiled qPCR data from 11 publicly available datasets from four ocean basins. Using a weighted meta-analysis we found that temperature was a robust factor governing the diazotroph abundances (except for UCYN-A) across ocean basins.

Attempting to identify differences in environmental impacts on two of the DDA symbiont strains (het-1 and het-2), we applied a new statistical tool called piecewise structural equation model, on qPCR abundance data from the western tropical North Atlantic. We saw that the two strains had a direct positive interaction between each other, but two parameters (salinity and dissolved inorganic phosphorous) differed. Based on a direct positive effect of salinity on het-1, and an indirect negative effect of salinity on het-2, we concluded that het-2 prefers intermediate salinities (30-35 PSU), which is consistent with where observations of het-2 blooms have been made.

Although DDA and UCYN-A symbionts both are major contributors of new N, and are symbiotic, they have several unique differences. The host partners differ in phylogeny (diatom vs prymnesiophyte), size (80-250 vs 7-10 µm) and the symbiotic life history (colonial vs solitary). The larger, colonial nature of DDAs make them difficult to collect, and hence they are often under-sampled and undetected. In fact, after reviewing 46 qPCR studies we found that < 30% of the studies (13 out of 46) quantified all three DDA symbionts, compared to UCYN-A (96%, 44 out of 46).

In order to study the DDA symbiont gene expressions we developed a highly specific DDA symbiont microarray (748 probes), which was applied on samples collected in the South China Sea. Although the gene expression levels were highly variable, we observed an upregulation of the nifH gene (for N₂ fixation) in the night. Investigating environmental impact on overall gene expression levels, we found that fluorescence, temperature and salinity was most important. Temperature and salinity also constrained abundances, but fluorescence could be seen as a proxy for either other phytoplankton or light availability, suggesting that daylight and host influence DDA symbiont gene expression levels.

The results of this thesis broaden our understanding of the DDAs and how their ambient environment influences them. It has also opened up new possibilities for in depth analysis of these complex environmental impacts. Lastly, it has provided new analysis tools for further development on how the symbionts and hosts potentially impact each other’s activities.