One of the main questions in microbial ecology is to understand what are the processes that structure and govern the species composition of communities. Answering to this question implies the need to detect bacterial species in environmental samples or at least to define ecologically meaningful “units of differentiation”. The dynamics and distribution of different broad taxonomic groups that constitute the bacterioplankton (at the phylum or class level) has been widely studied. The different large groups are known to follow different spatial distributions and seasonal cycles, but important variations also occur at lower levels (i.e. species or ecologically distinct populations). In this thesis we aimed to study some of these aspects for an important group of marine bacteria: the phylum Bacteroidetes. First at all, we evaluated the existing molecular tools in an updated phylogenetical and methodological context. Strengths and weaknesses of different probes used in different techniques were thoroughly studied and discussed, and alternatives were designed for some specific purposes related with the study of this group in marine environmental samples. Standing on these tools, we studied the phylum at different levels of resolution resulting from assorted molecular approaches. We related entire phylum abundances with environmental parameters as they vary spatially and temporally in the oceanic realm. We identified the most abundant phylotypes forming the Bacteroidetes assemblages and their ecological variability. Both the relative abundance and the changes in the phylotypes constituting the community followed clear spatial and seasonal patterns, which were thoroughly described. To increase the resolution of our study, we built extensive clone libraries including the 16S-ITS-23S ribosomal operon. The phylogenetic study of these sequences from different environmental samples showed that closely related phylotypes were organized in monophyletic clades of habitat specific sequences. This confirmed that similar bacteria (in terms of the 16S rRNA) seem to present different physiologies, which would drive their distribution and adaption to specific ecological niches. The study of the distributions of these clades would help to understand the ecology of this group. Finally, we integrated the above information to design several clade-specific qPCR primer sets targeting putative Bacteroidetes ecotypes. These primers were thoroughly tested as the best possible compromise between specificity and performance for the intended purpose, and they were extensively evaluated in silico. An optimization of the qPCR technique workflow and a demonstration of its usefulness was carried out using the NS2b group. These primers offer a toolbox for hierarchical quantitative studies characterizing Bacteroidetes populations in different environments. Such studies should help to explore and corroborate links of certain groups with specific habitats and environmental variables Descripción: