Microtaxonomy is the study of classifying and identifying microorganisms based on their genetic and phenotypic characteristics, such as morphology, biochemistry, and physiology. On the other hand, metagenomics is a field that involves the study of the genetic material recovered directly from environmental samples, including the DNA of microbial communities.
In terms of identifying microbiota, both metataxonomics and shotgun metagenomics can be successful and reliable, depending on the research question and the samples being analyzed. Metataxonomics typically involves amplifying and sequencing a particular gene, such as the 16S rRNA gene, which is commonly used to identify and classify bacterial species. Shotgun metagenomics, on the other hand, involves sequencing all the DNA present in a sample, allowing for a more comprehensive analysis of the microbial community. The choice between metataxonomics and shotgun metagenomics will depend on the specific goals of the study, the complexity of the microbial community being analyzed, and the available resources. For example, metataxonomics may be preferred if the goal is to identify specific bacterial species in a relatively simple microbial community. However, shotgun metagenomics may be more appropriate if the microbial community is complex, and a more comprehensive analysis is required.
In short, it depends on the questions you want to answer.
If you only care about microbial taxonomy composition in your samples and you have a lot of them, then Metataxonomics can be a good choice since it is cheaper and easier to analyse.
However, Metagenomics is the more powerful alternative since it can show not only which taxa are present, but also what they are doing there (functions). Usually this approach is more expensive and requires more time to analyse.
There are a lot of papers that can describe it in more details. For example: Article Best practices for analysing microbiomes
Metataxonomics can be based on 16S rRNA, but not necessarily restricted to it and can also be apllied to shotgun data. I guess the termin itself is too broad and therefore may be a little bit confusing.