Given that diatoms are photosynthetic algae, they are restricted to the sunlight zone, i.e. the depth of the water in a lake or ocean that is exposed to sufficient quantities of sunlight to allow for survival. They are highly sensitive to any environmental changes such as light availability, temperature, salinity etc. In general, diatoms prefer cold, nutrient rich waters. This is what makes them so valuable as indicators for water quality. The specific composition of diatom communities is a very sensitive instrument to measure changes in aquatic environments.
Diatoms have been regularly used as bioindicators to assess water quality of surface waters, especially in developed countries. Many of the widely used diatom indices have been developed from studies of European rivers and they are integrated in policies such as the European Water Framework Directive.
However, Diatom-based indices require unambiguous taxa identification to species level and that is challenging. Morphological approaches require expert taxonomic knowledge and often expensive infrastructure as many of the characters can only be detected by scanning electron microscopy or similar high-resolution technologies.
It comes to no surprise that researchers working with diatoms are looking into the application of DNA Barcoding to overcome the difficulties of identification. The community is still discussing which marker to use but to me it seems they slowly gravitate towards a fragment of the 18S rDNA (V4 region) and thereby following the suggestions of the protist working group. A good example for this is a new study by a group of German researchers:
We here investigate how identification methods based on DNA (metabarcoding using NGS platforms) perform in comparison to morphological diatom identification and propose a workflow to optimize diatom fresh water quality assessments.
Samples from seven different sites along the River Lusatian Neisse and the River Odra were taken and split into three subsamples. One of those was used for next generation sequencing of the 18S V4 region, the second for morphological analysis, and the third for the establishment of clone cultures from individual cells. The colleagues found that next generation sequencing almost always led to a higher number of identified taxa, which was subsequently verified by morphology. Taxa retrieval varies considerably but not necessarily because of natural variation but more as the result of varying taxonomic coverage in available reference databases. The authors conclude:
Next-generation sequencing based eDNA barcoding is not a swiss army knife, but provides a more comprehensive insight into diatom diversity or other protist communities and therefore could be the basis for the ecological projection of global diversity. If thoroughly conducted, the here presented approach not only bears the potential to supplement and improve the old identification system, but beyond that opens up many new opportunities and challenges: diversity data from NGS eDNA barcoding of environmental samples can easily be compared and combined on different spatial (α-, β-, γ-diversity), temporal and taxonomical levels. Therefore, it is applicable for large scale biomonitoring and the quality management of water bodies, for example under governmental frameworks.